Principles and Practice of Brachytherapy using afterloading systems Edited by
C.A. Joslin Emeritus Professor of Radiotherapy, Leeds University, Tunbridge Building, Regional Cancer Treatment Centre, Cookridge Hospital, Leeds, UK
A. Flynn Head of Brachytherapy Physics, Department of Medical Physics and Engineering, Cookridge Hospital, Leeds, UK
and
E.J. Hall Professor of Biophysics, Radiology and Radiation Oncology, Director-Center for Radiological Research, Department of Radiation Oncology, Columbia University, New York, USA
A member of the Hodder Headline Group LONDON Co-published in the United States of America by Oxford University Press Inc., New York
First published in Great Britain in 2001 by Arnold, a member of the Hodder Headline Group 338 Euston Road, London NW1 3BH http://www.arnoldpublishers.com Co-published in the United States of America by Oxford University Press Inc., 198 Madison Avenue, New York, NY10016 Oxford is a registered trademark of Oxford University Press © 2001 Arnold All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronically or mechanically, including photocopying, recording or any information storage or retrieval system, without either prior permission in writing from the publisher or a licence permitting restricted copying. In the United Kingdom such licences are issued by the Copyright Licensing Agency: 90 Tottenham Court Road, London W1P OLP. Whilst the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors, editors nor the publisher can accept any legal responsibility or liability for any errors or omissions that may be made. In particular (but without limiting the generality of the preceding disclaimer) every effort has been made to check treatment schedules, instructions or ideas contained in the material herein. However it is still possible that errors have been missed. For these reasons, and because of rapid advances in the medical sciences, the reader is strongly urged to consult the latest references before utilising any of the treatment schedules, instructions or ideas contained in this book. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN 0 340 74209 7 1 2 3 4 5 6 7 8 9 10 Publisher: Joanna Koster Development Editor: Sarah de Souza Project Manager: Marian Haimes Production Editor: Lauren McAllister Production Controller: Martin Kerans Typeset in 10/12 Minion by Phoenix Photosetting, Chatham, Kent Printed and bound in Great Britain by the Bath Press
What do you think about this book? Or any other Arnold title? Please send your comments to
[email protected]
Contents
Contributors
v
Preface
PART I
viii
THE PHYSICS OF BRACHYTHERAPY
1
1
Sources in brachytherapy
3
2
Source specification and dosimetry
3
Calibration of sources
4
Systems of dosimetry
5
Computers in brachytherapy dosimetry
6
Dose specification and reporting: the ICRU recommendations
7
Afterloading systems
8
Quality assurance in low dose-rate afterloading
9
Quality assurance in high dose-rate afterloading
Edwin Aird J.M.Wilkinson
11
Colin H.Jones
19
Anne Welsh and Karen D'Amico
35
Robert van der Laarse and Robert W. Luthmann Andre Wambersie and Jan J. Battermann
A. Flynn
49 81 103
Eric D. Slessinger Colin H. Jones
112 133
10
Radiation protection in brachytherapy
A.M. Bidmead
147
PART II
THE RADIOBIOLOGY OF BRACHYTHERAPY
11
The radiobiology of low dose-rate and fractionated irradiation
12
Dose-rate effects with human cells
13
Radiobiology of high dose-rate, low dose-rate, and pulsed dose-rate brachytherapy
159 Joel S. Bedford
G. Gordon Steel and John H. Peacock
161 180
David J. Brenner, Roger Dale, Colin Orton, and Jack Fowler
189
14
Predictive assays for radiation oncology
205
15
Principles of the dose-rate effect derived from clinical data
John A. Cook and James B. Mitchell Eric J. Hall and David J. Brenner
PART III CLINICAL PRACTICE
215
223
16
Endobronchial brachytherapy in the treatment of lung cancer
17
Brachytherapy in cancer of the esophagus
18
High dose-rate afterloading brachytherapy for prostate cancer
19
Low dose-rate brachytherapy for breast cancer
20
Brachytherapy in the treatment of head and neck cancer
Burton L Speiser
A.D. Flores
225 243
P.J. Hoskin
Julia R. White and J. Frank Wilson A. Gerbaulet and M. Maher
257 266 284
iv Contents 21
22
High dose-rate interstitial and endocavitary brachytherapy in cancer of the head and neck Peter Levendag, Connie de Pan, Dick Sipkema, Andries Visser, Inger-Karine Kolkman, and Peter Jansen
296
Brachytherapy in the treatment of pancreas and bile duct cancer Srinath Sundararaman, and Margot Heffernan
317
Dattatreyudu Nori, Suhrid Parikh,
23
Brachytherapy for treating endometrial cancer
24
Low dose-rate brachytherapy for treating cervix cancer: changing dose rate
25
High dose-rate brachytherapy for treating cervix cancer
26
Brachytherapy for brain tumors
27
H.A. Ladner, A. Pfleiderer, S. Ladner, and U. Karck R.D. Hunter and S.E. Davidson
C.A Joslin
373
A.M. Nisar Syed and
Ajmel A. Puthawala 28
343 354
Maarten C.C.M. Hulshof and Jan J. Battermann
Interstitial brachytherapy in the treatment of carcinoma of the cervix
333
379
Interstitial brachytherapy in the treatment of carcinoma of the anorectum
Ajmel A. Puthawala and
A.M. Nisar Syed
387
29
High dose-rate brachytherapy in the treatment of skin tumors
C.A. Joslin and A. Flynn
30
Hyperthermia and brachytherapy Peter M. Corry, Elwood P. Armour, David B. Gersten, Michael J. Borrelli, and Alvaro Martinez
400
31
The costs of brachytherapy
410
32
Quality management: clinical aspects
33
Safe practice and prevention of accidents in afterloading brachytherapy
Graham Read C.A. Joslin
423
C.A. Joslin 34
Pulsed low dose-rate brachytherapy in clinical practice Index
393
A. Flynn, S.E. Griffiths, and 433
Patrick S. Swift
443 451
Contributors
Edwin Aird Medical Physics Department, Mount Vernon Hospital, Middlesex, UK
S.E. Davidson The Christie Hospital NHS Trust, Manchester, UK
Elwood P. Armour
Connie de Pan Department of Radiation Oncology, Dr Daniel den Hoed
Department of Radiation Oncology, William Beaumont
Cancer Centre, Rotterdam, The Netherlands
Hospital, Michigan, USA Jan J. Battermann
A.D. Flores 7955 E, Chaparral Unit 125, Scottsdale, Arizona 85250, USA
Department of Radiation Oncology, Academisch Ziekenuis Utrecht, The Netherlands
A. Flynn Medical Physics Department, Cookridge Hospital, Leeds, UK
Joel S. Bedford Department of Radiological Health Sciences, Colorado State University, Colorado, USA
Jack Fowler Department of Human Oncology K4/336, University of Wisconsin Cancer Center, Wisconsin, USA
A.M. Bidmead Physics Department, Royal Marsden NHS Trust Hospital, London, UK
A. Gerbaulet Brachytherapy Department, Institut Gustave-Roussy, Villejuif, France
Michael J. Borrelli Department of Radiation Oncology, William Beaumont Hospital, Michigan, USA
David B. Gersten Department of Radiation Oncology, William Beaumont Hospital, Michigan, USA
David J. Brenner Center for Radiological Research, Columbia University, New York, USA
S.E. Griffiths Department of Radiotherapy, Regional Cancer Treatment Centre, Cookridge Hospital, Leeds, UK
John A. Cook Radiation Biology Branch, National Cancer Institute, National Institutes of Health, Maryland, USA
Eric J. Hall College of Physicians and Surgeons Center for Radiological Research, Columbia University, New York, USA
Peter M. Corry Department of Radiation Oncology, William Beaumont Hospital, Michigan, USA
Margot Heffernan Tumor Registry, New York Hospital Medical Center of Queens, New York, USA
Roger Dale District Department of Medical Physics, Charing Cross Hospital, London, UK
P.J. Hoskin Marie Curie Research Wing, Mount Vernon Hospital, Middlesex, UK
Karen D'Amico Medical Physics Department, Cheltenham General Hospital, Cheltenham, UK
Maarten C.C.M. Hulshof Academisch Medisch Centrum, Amsterdam, The Netherlands
vi Contributors R.D. Hunter
Suhrid Parikh
The Christie Hospital NHS Trust, Manchester, UK
Radiation Oncology, New York Hospital Medical Center -
Peter Jansen
Cornell, New York, USA and New York Hospital Medical Center of Queens, New York, USA
Department of Radiation Oncology, Dr Daniel den Hoed Cancer Centre, Rotterdam, The Netherlands
John H. Peacock
Colin H. Jones
Surrey, UK
Radiotherapy Research Unit, Institute of Cancer Research,
Physics Department, Royal Marsden NHS Trust, London, UK A. Pfleiderer C.A. Joslin Leeds University, Department of Radiotherapy, Regional Cancer Treatment Centre, Cookridge Hospital, Leeds, UK
University Hospital for Women, Freiburg, Germany Ajmel A. Puthawala Department of Radiation Oncology, Long Beach Memorial Medical Center, California, USA
U. Karck University Hospital for Women, Freiburg, Germany
Graham Read Oncology Services, Royal Preston Hospital, Preston, UK
Inger-Karine Kolkman Department of Radiation Oncology, Dr Daniel den Hoed Cancer Centre, Rotterdam, The Netherlands
Dick Sipkema Department of Radiation Oncology, Dr Daniel den Hoed Cancer Centre, Rotterdam, The Netherlands
H.A.Ladner University Hospital for Women, Freiburg, Germany
Eric D. Slessinger Regional Cancer Center, Community Hospital Indianapolis,
S. Ladner
Indiana, USA
University Hospital for Women, Freiburg, Germany Burton L Speiser Peter Levendag
St Joseph's Hospital and Medical Center, Department of
Department of Radiation Oncology, Dr Daniel den Hoed
Radiation Oncology, Arizona, USA
Cancer Centre, Rotterdam, The Netherlands Robert W. Luthmann St Vincent's Medical Center, Department of Radiation Oncology, Florida, USA M. Maher Radiotherapy Department, Mater Private Hospital, Dublin, Ireland Alvaro Martinez Department of Radiation Oncology, William Beaumont Hospital, Michigan, USA
G. Gordon Steel Radiotherapy Research Unit, Institute of Cancer Research, Surrey, UK Srinath Sundararaman Radiation Oncology, New York Hospital Medical Center of Queens, New York, USA Patrick S. Swift Radiation Oncology, Alta Bates Comprehensive Cancer Center, California, USA A.M. Nisar Syed Department of Radiation Oncology, Long Beach Memorial
James B. Mitchell
Medical Center, California, USA
Radiation Biology Branch, National Cancer Institute, National Institutes of Health, Maryland, USA
Robert van der Laarse Nucletron BV, Veenendaal, The Netherlands
Dattatreyudu Nori Radiation Oncology, New York Hospital Medical Center Cornell, New York, USA and New York Hospital Medical Center of Queens, New York, USA
Andries Visser Department of Radiation Oncology, Dr Daniel den Hoed Cancer Centre, Rotterdam, The Netherlands
Colin Orton
Andre Wambersie
Gershenson Radiation Oncology Center, Harper Hospital and Wayne State University, Michigan, USA
Unite de Radiobiologie et de Radioprotection, Faculte de Medecine, Universite Catholiquede Louvain, Bruxelles, Belgium
Contributors vii Anne Welsh Medical Physics Department, Cheltenham General Hospital, Cheltenham, UK
J.M. Wilkinson North Western Medical Physics, Christie Hospital, Manchester, UK
Julia R. White Medical College of Wisconsin, Wisconsin, USA
J. Frank Wilson Medical College of Wisconsin, Wisconsin, USA
Preface
Brachytherapy was for many years in a state of decline, principally because of the radiation hazards to users and those associated with the management of patients. The introduction of afterloading machines in the 1960s provided the means to control the movement and position of individual radioactive sources and greatly reduced the radiation exposure to staff. As a result, brachytherapy underwent a renaissance and provided the necessary stimulus to promote the development of afterloading brachytherapy techniques. These developments have been further supported by the availability of nuclides, particularly cobalt-60, cesium-137, and iridium-192 and, more recently, radioactive seeds of iodine-125 and palladium 105. In parallel with the technological advances in afterloading machines, there have been major developments in imaging techniques and computerized planning. Cancer management generally has undergone major advances since the 1960s and brachytherapy has played an increasingly important role. The optimal management of cancer patients requires expert teams who specialize in certain cancer sites within which brachytherapy may have a specific place. Much of this work is now being provided on an outpatient or day-care basis and prolonged hospital stay is proving to be unnecessary. Clinical training is largely obtained by observation of and training from one's peers and also from supervised hands-on experience. In parallel with the development
of clinical experience, an understanding of the principles of radiobiology and physics is of great importance. It is also prudent that clinical radiation oncologists continue to update their state of knowledge with respect to current practice. The purpose of this book is not only to present to the trainee clinical oncologist the scientific background and principles of brachytherapy afterloading techniques, but also to update those who specialize in brachytherapy. The book is presented in three sections: physics, radiobiology, and clinical treatment. The sections attempt to cover the scientific principles, technical procedures, and clinical applications of'afterloaded' brachytherapy. The editors have aimed at a consistent presentation for the various chapters without attempting to interfere with the different styles of the individual authors. Some chapters will be found to be more extensive than others, which is mainly a reflection of the widespread application of brachytherapy techniques within the subject of those chapters. We hope that readers of this textbook will find the contents helpful in their work. The editors would like to express their appreciation to all authors for their well-prepared manuscripts and for their tolerance during the book's production. C.A. Joslin, A. Flynn, and Eric J. Hall
PART
The physics of brachytherapy
1 2 3 4 5 6 7 8 9 10
Sources in brachytherapy Source specification and dosimetry Calibration of sources Systems of dosimetry Computers in brachytherapy dosimetry Dose specification and reporting: the ICRU recommendations Afterloading systems Quality assurance in low dose-rate afterloading Quality assurance in high dose-rate afterloading Radiation protection in brachytherapy
3 11 19 35 49 81 103 112 133 147
I
This page intentionally left blank
1 Sources in brachytherapy EDWIN AIRD
1.1 1.1.1
INTRODUCTION Radium
Radium was discovered by Marie Curie in 1898. Within 3 years of this discovery the first patients were treated with radium implanted into their tumors. In the UK, St Bartholomew's Hospital received its first radium for clinical use in 1906. Early clinical experience with these sources led to radiation necrosis and it became clear that this was due, in part, to the intense beta-ray dose from the radium. It was not until 1920 that successful filtration of the beta-rays was achieved. Radium was then used extensively throughout the world. Physicists in the major clinical centers developed dosimetry systems for interstitial and intracavity brachytherapy. The Manchester and Paris systems are probably still the most widely used for interstitial therapy. However, in general radium has been replaced by other radionuclides because, although it has a long halflife, it has several disadvantages: • Radium and several of its daughter products, including radon, are alpha emitters. Radon is a noble gas which is soluble in tissue. This gas could escape through a hairline crack - not easily detected by a visual check - in the radium capsule. If an implanted radium source were to be ruptured within the patient's body, radium and its daughter products may become deposited more or less permanently in the bone. • There is also the possibility of damage - by incineration or mechanical means - when the sources are lost or while they are being processed, with the subsequent release of toxic radioactivity to the environment.
• The gamma radiation from a radium source is of higher energy than is necessary for brachytherapy. Radiation protection for these sources requires large thicknesses of lead, which can cause problems when it comes to: transporting sources in heavy containers using very heavy protective screens around the patient the need for a heavy rectal shield in applicators used for gynecological treatment. • The practical maximum activity concentration (the specific activity) of radium salt is low (approximately 50 MBq mm~3 of active volume). Sources of higher activity are therefore bulky and not suitable for afterloading systems. 1.1.2
Radium substitutes
This was the phrase used to describe the first set of new (artificial) radionuclides which were found useful for brachytherapy from about 1950 onwards, though it is only very recently that most radiotherapy centers have stopped using radium. It was found that there were very few radionuclides with the appropriate properties of the ideal brachytherapy source. These properties are as follows: (A full discussion of these points may be found in the British Journal of Radiology Supplement 21 (1987); an abbreviated set is stated here.) • Photon energy should be low to medium (0.03-0.5 MeV) to minimize radiation protection problems (with the proviso that low-energy radionuclides should not be used near bone because of the enhanced dose to bone at these energies). • For permanent stock, a long half-life is desirable such that the radioactive decay within the practical lifetime
4 Sources in brachytherapy
•
• • •
of the source and its container (typically 10 years) is small. For permanent implantation, a fairly short half-life is essential in order to minimize the time over which special precautions, towards relatives of a radioactive patient and members of the public, need to be in place. The nuclide should be available at high specific activity. There should be no gaseous disintegration product. The nuclide should be available in a form which does not powder or otherwise disperse if the source is damaged or incinerated.
The first sources to be used as alternatives to radium were cobalt-60, gold-198, cesium-137 and iridium-192. These are all described briefly below. The most commonly used sources at this time are cesium-137 and iridium-192, both of which are used in after-loading systems. Iridium-192 has the possibility of high specific activity, which allows it to be used as a high dose-rate (HDR) source.
1.1.3
New sources
The newer sources are not known as radium substitutes, mainly because they have very different properties from radium, namely very much higher specific activity (for example the HDR iridium-192 source) and very different energy. The only new source that has been accepted into routine clinical use in certain centers throughout the world is iodine-125. Palladium-103 is also now available as a standard commercial source. The other sources that are still at the research stage of development, to find out whether they can be of use clinically, are samarium-145, americium-241, and ytterbium-169.
1.2
PRODUCTION OF RADIONUCLIDES
The most common method of producing the radionuclides used in brachytherapy, apart from cesium-137 (which is a fission product), is by neutron bombardment in a nuclear reactor. The reaction is that of neutron capture, normally in the stable isotope of the element required (except for iodine-125, see below). Thus, for iridum the reaction is:
For cobalt-60 the reaction is:
This method of production has the disadvantage that the radioactive isotope cannot be separated from the
stable isotope and limits the specific activity possible. However, for iodine-125, the reaction proceeds in two stages, with xenon as the initial target element:
The radioactive xenon decays by beta emission, with a half-life of 8 days, to the required125/53I,which can then be chemically extracted as a pure radioisotope. To estimate the yield of a given radionuclude, it is possible to use the simple form (for short irradiation time and low fluxes) of the yield equation given in The Radiochemical Manual [ 1 ]:
where s is the reaction cross-section; (j is the neutron flux; n0 is number of atoms in target (=N 0 wq/A, where w is target mass, q is the isotopic abundance of the nuclide of interest in the target element, and A is the atomic mass of the target element, N0 = Avogadro's number); t = bombardment time; l = decay constant of the product element. The specific activity, S, of the target nuclide may be approximated to:
For long irradiation times:
and S ® Ssat, the maximum activity possible for a given neutron flux. Example. For indium-191, the neutron capture crosssection is 910 barns. (This is very high, compared with cobalt-60, for example, which has a cross-section of 43 barns.) If the neutron flux is high (typically 1013 rt.cm-2S-1), using the above equations, it is possible to show that the maximum theoretical specific activity for iridium-192 is about 29 TBq g-1, which equates to about 2.5 TBq for an HDR source of 0.086 g. In practice, although the neutron flux is probably higher:
The irradiation times are shorter, so that only a fraction of Ssat is reached. The typical specific activity produced for HDR sources is 4.3 TBq g-1 (for a 370 GBq source). It is interesting to compare the specific activities (in terms of activity per unit length) available for the different radionuclides used now in brachytherapy with those in the older sources. Iridium wire Iridium HDR source
37 MBq mm-1 74 GBq mm-1 actual source length
Brachytherapy sources used in afterloading systems 5
Cesium miniature cylindrical
Cobalt-60 beads
Compared with radium
333 MBq mm -1 actual source length (for manual afterloading source trains) 592 MBq mm-1 actual source length (in a set of active beads) 50 mg in 13.5 mm active length, 20 mm actual length, equates to 92.5 MBq mnT-1 actual length
Table 1.2
Examples of source trains suitable for Manchester
System
Medium vaginal ovoid Medium intrauterine Medium tandem a
3 5 11
1.7 2.1 5.4
127 158 412
The nominal output is expressed in terms of air kerma rate (AKR) at 1 m from the center of the source.
13 BRACHYTHERAPY SOURCES USED IN AFTERLOADING SYSTEMS Except where otherwise stated, reference data concerning sources come from the Amersham Catalogue of Radiation Source for Oncology.
13.1
Image Not Available
Cesium-137 (Table 1.1)
FORMS FOR MANUAL AFTERLOADING Miniature cylindrical sources (Figure 1.1) contain cesium-137 glass beads encapsulated in stainless steel. They are used in source trains in machine and manual afterloading systems for gynecological brachytherapy.
Figure 1.1 Cylindrical cesium-137 sources as used in the Amersham afterloading source train. (Reproduced by kind permission of Nycomed Amersham plc.)
AMERSHAM MANUAL AFTERLOADING SYSTEM In the Amersham Manual Afterloading System (Figure 1.2) a source train consists of a flexible stainless-steel holder containing miniature cylindrical sources separated by spherical steel spacers 1.8 mm in diameter. The
Table 1,1
Properties of cesium-137
Production
Half-life
A fission product Small quantities (less than 1% cesium-134 present [2], which decays with a half-life of about 2 years)
Image Not Available
30.17 years
Decay scheme Beta energies 0.512 MeV 1.173MeV
Emission probability-betas* 94.6% 5.4%
Photon energies 0.662 MeV
Emission probability - photons 90.1%
Barium X-rays 0.032-0.038 MeV
-7%
Beta filtration
0.5 mm of platinum or stainless steel
Half value layer in lead
6.5mm
* Data from The Radiochemical Manual [1].
Figure 1.2 Source train used in the Amersham afterloading system. (Reproduced by kind permission of Nycomed Amersham
Pic.)
6 Sources in brachytherapy
sources and spacers are retained in the holder by a steel spring, secured by a screwed-in end plug. They are designed to locate in the Amersham manual afterloading plastic applicators. The standard set of 11 source trains is suitable for the Manchester System of Gynecological tube dosage. Some examples are given in Table 1.2.
CESIUM-137 SOURCES FOR BUCKLER* AFTERLOADING
These are cylindrical sources used in the fixed ovoids of the Buchler Gynaecological System. The sources vary from 10.1 GBq with active dimensions 2 mm x 3.5 mm to fit applicators 6 mm diameter for low dose rates (LDRs), to 148 GBq with active dimensions 4.1 mm x 11.5 mm to fit applicators 8 mm diameter for HDRs.
WALSTAM-TYPE SOURCES [3] CESIUM-137 SOURCES FOR CURIETRON
These are short, cylindrical sources with hemispherical ends. They consist of cesium-137 in a ceramic matrix contained in a welded stainless-steel capsule. They are used in dome or cylindrical gynecological applicators. The sources approximate a point source of activity higher than that used in the Amersham (Manchester) System (Table 1.3).
These are cylindrical sources that are very similar to those of the Amersham Manual Afterloading System, but for use in a remote afterloading system in which the source trains are attached to a cable drive.
1.3,2 Table 1.3
Cobalt-60 (Table 1.4)
Waktnm-tvnr snurrtx;
FORM OF SOURCE
0.37-74 GBq
28.5-570.0 mGy h-1
REMOTE AFTERLOADING CESIUM-137 SOURCES
Although used in various forms in the past, the most common form in recent years is in 'bead' form, with a design very similar to that used for cesium-137 beads in the Selectron unit. However, the activity of cobalt-60 beads is higher and they are used for HDR brachytherapy.
Spherical Sources
Spherical sources are used in the Selectron (Nucletron BV) afterloading system (Figure 1.3). The cesium-137 is incorporated into a glass bead and encapsulated in stainless-steel ball bearings (referred to as 'beads' or 'pellets') which, together with inactive spacer beads, can be pneumatically loaded from the intermediate safe into a patient applicator along a plastic tube (nominal activity 1.48 GBq per bead, air kerma rate 112mGyh-1m2).
Table 1.4
Properties of cobalt-60
Production
By neutron activation of the stable isotope cobalt-59
Half-life
5.27 years
Decay scheme Beta energies 0.318 MeV
Emission probability - beta 99.9%
Photon energies 1.17 MeV 1.33 MeV
Emission probability - photon 99.9% 100.0%*
Beta filtration
Typical source wall thickness
Half value layer in lead 10mm * Data from The Radiochemical Manual [1].
Image Not Available
133
lridium-192 (Table 1.5)
FORMS OF IRIDIUM-192 Wire In Europe, platinum-covered iridium-192 wire is supplied in 500 mm length coils. The wire consists of an active iridio-platinum core, 0.1 mm thick, encased in a sheath of platinum, 0.1 mm thick. Figure 1.3 Spherical cesium-137 source as used in the Selectron afterloading system. (Reproduced by kind permission of Nucletron BV.)
* Buchler GmbH, Braunschweig, Germany.
Brachytherapy sources used in afterloading systems 7
Table 1.5
Properties of iridium-192
Production
By neutron activation of the stable isotope iridium-191; the process also produces quantities of iridium-194 (from the activation of iridium-193); because this has a half-life of only 17 h, it does not contribute a significant dose by the time the source is used in the patient
Half-life
73.83 days
Decay scheme Beta energies 0.079-0.672 MeV
Emission probability-betas 0.1-48.1%
Photon energies Range 0.2-1.06 MeV
Emission probability-photons
Effective photon energy 0.37 MeV (unencapsulated) 0.4 MeV (encapsulated)
Significant photon energies (>10%) for those greater than 10% 0.296 MeV 28.7% 0.308 MeV 29.8% 0.316 MeV 83.0% 0.468 MeV 47.7% Beta filtration
0.1 mm platinum
Half value layer in lead
4.5mm
Data from The Radiochemical Manual [1].
Iridium-192 wire is not classified as a 'sealed radiation source.' Because it is activated by neutron irradiation, its cladding remains slightly active. This is not significant in its clinical use. For radiation protection purposes, iridium wire is know as a 'closed radiation source.' Available source strength is shown in Table 1.6. Table 1.6
Available source strength of indium wire
1.11-37.00 MBqmnr
126 nGy h-1 mm-1-4.19 mGy h-1 mm-1
Wire is cut to the required lengths and loaded into plastic tubes or hypodermic needles.
Hairpins (Figure 1.4)
Platinum-covered iridium wire is supplied in the form of 'hairpin' or 'single-pin' shapes. The wire has a diameter of 0.6 mm to give it added strength; the beta filtration remains at 0.1 mm platinum. Hairpins are 131 mm overall length, with leg length 60 mm nominally (the legs can be cut to the required length) and with a range of source strength (Table 1.7).
Image Not Available
Figure 1.4 Platinum-covered iridium-192 wire hairpin (a) and slotted hairpin guide needles (b) as supplied by Nycomed Amersham pic. (Reproduced by kind permission of Nycomed Amersham plc.)
8 Sources in brachytherapy
Table 1.7
Available source strength of iridium hairpins
13.4
lodine-125(Table1.8[5])
FORMS OF SEED (Figure 1.6a, b) 1.48-11.10 MBq mm-1
168-1257 m G y - 1 mm-1
Single pins are 73 mm overall length with a nominal leg length of 60 mm and with a range of source strength the same as the hairpins. Slotted stainless-steel guides are used for implanting hairpins and single pins. lridium-192 'seeds'
In the USA these are used instead of wire. Two seed styles are commercially available: 1. 0.1 mm diameter core of active wire (30% iridium, 70% platinum) surrounded by 0.2 mm cladding of stainless steel (Best Industries, Springfield, VA). 2. 0.3 mm diameter core (10% iridium, 90% platinum) surrounded by 0.1 mm cladding of platinum (Alpha-Omega, Bellflower, CA). Both seeds are 3 mm active length and are supplied inside strands of nylon of 0.8 mm outside diameter. Normal spacing is 1 cm, but other spacings are available. Maximum overall length is about 18 cm. Air kerma strengths range from 120 to 650 MBq per seed. Miniature iridium-192 sources for high dose rate
There is a variety of types of these sources, ranging from 0.2 to 1.3 mm diameter and 1 to 20 mm active length, with typically up to 370 GBq activity (air kerma rate 42 mGy h-1). Figure 1.5 shows the HDR sources in use throughout the world at the present time. They are always permanently attached to a cable drive. The active wire is encased in stainless steel. There is now a 'new design source' for the Nucletron BV microSelectron-HDR machine [4] with slightly smaller dimensions (4.95 mm length, 0.90 mm diameter) and similar dose distribution, except for some improvement near the source tip and in the shadow of the cable assembly.
Type 6711 seeds (Nycomed Amersham plc, Amersham, UK) are used for permanent implant. Each seed consists of a welded titanium capsule containing iodine-125 adsorbed onto a silver rod (which also acts as X-ray marker). The active length is 3.0mm and diameter 0.5mm. The overall length is 4.5mm and diameter 0.8 mm. Sources are available with air kerma rates at 1 m of 0.13-7.58 [mGyh-1. The seeds are used with special applicators to introduce them into the patient a fixed distance apart. A new type of absorbable suture called Rapid Strand™ (Figure 1.7), from Nycomed Amersham plc, has become available that encases ten seeds at a fixed distance apart (1cm) in tissue until the suture dissolves (5mm or 15 mm is also available in this form). The suture material is braided Vicryl which is stiffened thermally and sterilized by ethylene oxide gas. It eventually dissolves in tissue. These seeds also emit silver characteristic X-rays Table 1.8
Properties of iodine-125
Production
Neutron activation of xenon-124 to xenon-125, which then decays to iodine-125
Half-life
59.4 days
Decay scheme
Iodine-125 decays by electron capture to the first excited state of tellurium-125, which undergoes internal conversion 93% of the time; the other 7% is occupied by the production of a gamma ray photon of 35.5 keV.
The electron capture and internal conversion processes give rise to characteristic X-rays as follows: (X-ray) Photon energy 27.4 keV 31.4keV
Decay photons emitted 15% 25%
Tenth value layer in lead
0.01 mm
Image Not Available
Figure 1.5 MicroSelection iridium-192 HDR sources. (Reproduced by kind permission of Nucletron BV)
Brachytherapy sources used in afterloading systems 9
Image Not Available
Image Not Available
Figure 1.6 (a) Type 6777 iodine-125 seed, (b) Type 6702 iodine125 seed. (Reproduced by kind permission of Nycomed
Figure 1.7 Rapid Strand. (Courtesy Nycomed Amersham plc.)
Amersham plc.)
of 22.1 and 25.5 keV. The average photon energy is taken to be 27.4 keV. Type 6702 seeds are used for temporary interstitial implants. These consist of a welded titanium capsule containing three resin spheres onto which the iodine125 is adsorbed by an ion exchange. Sources are available with air kerma rates at 1 m of 6.4-51.9 (mGyh-1. The effective energy of the photons from this seed is taken to be 28.5 keV. A further type of iodine-125 seed is available in North America from Best Industries. The Model 2300 contains radioactive iodine adsorbed on a tungsten wire that is encapsulated by two walls of titanium. This source offers the following advantages: Because it contains radioactive iodine on the ends as well as on the surface of the tungsten, it produces a more isostropic dose distribution than the other sources [6]. It is available in a wide range of source strengths and therefore suitable for both temporary and permanent implantation. The tungsten wire acts as a radiographic marker. The double-walled encapsulation reduces the risk of radioactive leakage. There is another new source of iodine-125 seeds on the US market, designated as MED 3631-A/S and manufactured by North America Scientific Incorporated, North Hollywood, California [7], This source has now been reconfigured (MED 3631-A/M) with the intent of providing greater facility for radiographic source identification while achieving reduced isotropy [8].
13.5
Palladium-103 sources (Table 1.9)
FORMS OF SEEDS
The active material is coated onto two graphite pellets 0.9 mm long and 0.6 mm in diameter. Between these is a 1 mm long lead marker for radiography. These seeds are encapsulated in a 0.05 mm thick titanium tube, laser welded, that is 4.5 mm long and 0.8 mm diameter (the same dimension as the iodine-125 seed).
Table 1.9 Properties of palladium-103 sources
,
Production
Palladium-103 is formed when stable palladium-102 absorbs a neutron
Half-life
16.97 days
Decay scheme
By electron capture, mostly to the first and second excited states of ruthenium-103
An excitation is almost totally by internal conversion, leading to the production of characteristic X-rays: Photon energy 20.1 keV 23.0 keV
Photons em itted 65.6% 12.5%
Effective energy
21 keV
Tenth value layer in lead
0.03mm
10 Sources in brachytherapy
13.6
Other proposed sources [9,11]
The properties of other proposed sources are shown in Tables 1.10, 1.11, and 1.12. Table 1.10
Properties of samarium-145
Photon energy range
38.2-61.4 keV
Mean photon energy
41keV
Half-life
340 days
Maximum specific activity
73 GBq mm-3 (compared with 370 GBqmm-1for iodine-125)
Tenth value layer in lead
0.2mm
Purpose
To improve dose distribution and shelf-life compared with iodine-125; in addition, it is noted that the photon energy emitted allowssensitization of biological cells to radiation damage by the addition of iodinated deoxyuridine; there are no commercially available sources at the present time
Table 1.11
Properties of ameridum-241
Mean photon energy
13.9-125 keV (but dominated by 59.5 keV) GOkeV
Half-life
432 years
Maximum specific activity
0.34 GBq mm-3
Tenth value layer in lead
0.42 mm
Purpose
Could be used as an alternative to cesium-137 for cancers of the cervix and endometrium
Disadvantages
An a emitter Only low specific activity available
Photon energy range
Table 1.12
Properties of ytterbium-169 [10]
Photon energy range
50-308 keV
Mean photon energy
93keV
Half-life
32.0 days
Maximum specific activity
340 GBq mm-3
Tenth value layer in lead Seed dimensions
1.6mm
Purpose
Similarto iodine-125 Possible benefit-less attenuation in tissue than iodine-125 or palladium-153 and higher specific activity
REFERENCES 1. Longworth, G. (ed.) (1998) The Radiochemical Manual. Harwell, UK, AEA Technology plc. 2. Godden, T.J. (1988) Physical Aspects of Brachytherapy, Medical Physics Handbooks 19. Bristol, Adam Hilger. 3. Walstram, R. (1965) Studies in therapeutic short distance and intracavitary gamma beam techniques. Physical considerations with special reference to radiation protection. Acta Radial., Supplement 236,1-129. 4. Daskalov, G.M., Loffler, E. and Williamson, J.F. (1998) Monte Carlo-aided dosimetry of a new high dose-rate brachytherapy source. Med. Phys., 25, 2200-8. 5. Nath, R., Anderson, L.L, Luxton, G., Weaverk, A., Williamson, J.F. and Meigooni, A.S. (1995) Dosimetry of interstitial brachytherapy sources: recommendations of the AAPM Radiation Therapy Committee Task Group No.43. Med. Phys., 22, 209-34. 6. Nath, R. and Melillo, A. (1993) Dosimetric characteristics of a double wall 1-125 source for interstitial brachytherapy. Med. Phys., 20,1475-83. 7. Wallace, R.E. and Fan, J.J. (1998) Evaluation of a new brachytherapy iodine-125 source by AAPM TG43 formalism. Med. Phys., 25, 2190-6. 8. Wallace, R.E. (1999) Report on the dosimetry of a new design iodine-125 brachytherapy source. Med. Phys., 26, 1925-31. 9. Battista, J.J. and Mason, D.L.D. (1994) New radionuclides for brachytherapy. In Brachytherapy from Radium to Optimization, ed. R.F. Mould, J.J. Battermann, A.A. Martinez and B.L Speiser. Veenendaal, The Netherlands, Nucletron International, 373-84. 10. Mason, D.L.D., Battista, J.J., Barnett, R.B. and Porter, A.T. (1992) Ytterbium-159: calculated physical properties of a new radiation source for brachytherapy. Med. Phys., 19, 695-703. 11. Williamson, J.F. (1995) Recent developments in basic brachytherapy physics. In Radiation Therapy Physics, ed. A.R. Smith. New York, Springer-Verlag, 247-302.
2 Source specification and dosimetry J.M.WILKINSON
2.1 2.1.1
SOURCE SPECIFICATION BY CONTENT Radium and radium mass
Early brachytherapy was practiced with two radionuclides from the uranium/radium series, namely radium-226 and its immediate daughter, radon-222. Both exist in equilibrium with later radionuclides in the series, and indeed their usefulness as brachytherapy sources arises from the gamma emissions that occur in the transitions from lead-214 (referred to at one time as radium B) to bismum-214 (radium C), and from bismuth-214 to polonium-214 (radium C'). Both radium and radon require heavy metal screenage (at least 0.5 mm of platinum or gold) to remove the particulate emissions, leaving practical brachytherapy sources with almost identical gamma spectra. Radium-226 sources were specified in terms of the mass of radium element, in milligrams, that each contained, and, given the very long half-life for the decay of radium, 5.85 x 105 days, it was usually regarded as unnecessary to correct the mass specification during a period of less than about 20 years, that is to say, during the normal working life of the source.
2.1.2
quially as the 'k' factor, and an assumed knowledge of the attenuating properties of the materials used in the source construction. The specific gamma ray constant for radium was defined as the product of the exposure rate, in roentgen per hour, and the square of the distance, in cm2, from a 1 mg point source, encapsulated in a platinum sheath of 0.5 mm thickness. Early workers adopted a value of 8.4 Rh-1cm-2mg-1 for this constant, but subsequently this was revised to what is now the generally accepted value of 8.25 Rh-1cm-2mg-1. To determine the exposure rate at a point near to a line source required an evaluation of the Sievert Integral [1]. This was originally expressed as an angular integration, as illustrated in Figure 2.1, and the integral itself is that given in equation 2.1:
The Sievert Integral
It was anticipated that the degree of radiation damage in tissue would be closely related to the magnitude of the exposure. The exposure rate at a point outside a radium source was determined by using a calculated value of the specific gamma ray constant, sometimes referred to collo-
Figure 2.1 The exposure rate at point P is obtained by the angular integration of the Sievert Integral from f1 to f2.
12 Source specification and dosimetry
where dX/dt is the exposure rate at the point P, M is the radium mass, k0 is the specific gamma ray constant corrected back to zero filtration, m, is an appropriate filtration coefficient, and the other symbols are as indicated in Figure 2.1. There is no analytic solution to the integral, but tabulated values have been published (see, for example, reference 2). This approach is not too unreasonable at relatively high gamma energies, but in fact represents a considerable simplification of the true physical problem, which becomes apparent when doses in the vicinity of a source are calculated by Monte Carlo techniques [3]. The classical Sievert Integral assumes that increased scattering at short distances will compensate for attenuation in the irradiated medium to within the precision tolerances required for practical brachytherapy work. Furthermore, there is no allowance for internal absorption in the source material itself and there is no obvious way to determine an appropriate value for the filtration coefficient for the sheathing material. Early workers used a single value of 0.2 mm-1 for filtration of radium gammas in platinum [4], but Whyte [5] suggested that a better approximation could be achieved by using decreasing values with increasing thicknesses to counter the hardening of the energy spectrum.
2*13
The milligram-hour concept
Systems of brachytherapy dosimetry for radium applications were devised which dictated the relative distribution of active material for different treatment geometries. Of these, the best known were the Manchester System [4, 6-9] and the Quimby System [10-12]. Such 'systems' recognized that, for a predetermined geometry and predetermined relative distribution of active material, the exposure rate at any point was proportional to the total amount of radium used, and that, for a complete treatment, the total exposure was proportional to the product of the amount (specified as radium mass in milligrams) and the duration (specified as time in hours). Hence the milligram-hour (which was the name given to both the quantity and its unit) became a key parameter in early brachytherapy dosimetry. It was assumed that this product would have to remain constant, for any particular source geometry, in order to achieve consistency in the observed clinical result. It is interesting to note that this assumption was, in effect, challenged at a very early stage by the use of time factor corrections [13, 14], but the application of such corrections, now generally called dose rate corrections, remains the subject of much debate. Brachytherapy systems are discussed in more detail in Chapter 4. They are introduced here to demonstrate how the radium mass specification was a fundamental component of these early clinical dosimetry procedures.
2.1.4
Radon and radium mass equivalent
A radon source was specified, at any particular instant in time, by its radium mass equivalent, defined as that mass of radium, encapsulated by a 0.5 mm thickness of platinum, which would give the same exposure rate at 1 cm from the axis of the source. Radon-222 has a half-life of 3.83 days and so both the radium mass equivalent and the exposure rate at any defined reference point decrease during an application. Radon seeds were used both for superficial mould treatments and for permanent implantation. The total exposure for an application was calculated by multiplying the initial exposure rate by an 'effective' treatment time, determined by integrating the area under the exponential decay curve. The generalized expression for 'effective' treatment time, teff, is that in equation 2.2, where l, is the decay constant and t is the duration of the application:
For a permanent implant of radon seeds, this reduced to teff = half-life/loge2, or 132.5 h. The use of radium and radon has now generally been discontinued due to radiation safety considerations. Radionuclides that have subsequently been substituted for radium and radon, for example cesium-137 or gold198, have also been specified in terms of the radium mass equivalent as this facilitates their use with the established radium systems that use the milligram-hour concept. However, such nuclides have different gamma emission spectra and may be encapsulated in different materials. Hence the dose distribution around a substitute source may be different from that around a radium or radon source of similar physical dimensions. In practice, the radium mass equivalent was frequently determined by comparing the source in question with a similar-sized radium source, of known mass content, in a well-type ionization chamber. This appears to have been satisfactory and there is no evidence to suggest that clinical results have been adversely affected by such practice. 2.1.5
Specification by activity content
An alternative quantity for specification by content is the activity of the radionuclide that is encapsulated in the source. The activity, A, of an amount of radioactive nuclide, in a particular energy state and at a given time, is defined by equation:
where dN is the expectation value of the number of spontaneous nuclear transitions from that energy state in the time interval dt [15]. The early unit of activity, the curie (Ci), was originally defined as the activity of 1 g of radium, or approximately
Specification by emission 13
3.7 x 1010 transitions per second. A subsequent redefinition of the curie made it exactly this figure. The curie is now obsolete as the unit of activity and has been replaced by the SI unit the becquerel (Bq). One becquerel is one transition per second. In order to calculate exposure rate at a point external to a source specified by its activity content, it was again necessary to assume a knowledge of the attenuating properties of the source and source encapsulation material, and also, now, to know the value of the exposure rate constant, Gd. This latter quantity replaces the specific gamma ray constant. The two constants are very similar in concept and for many nuclides are assigned the same value. However, the specific gamma ray constant does not allow for possible contributions from internal conversion X-rays, and these may be significant for low energy emitters. The definition of the exposure rate constant, as given in reference 16, is the quotient:
where(dX/dt)Gdis the exposure rate due to photons of energy greater than 5, at a distance / from a point source of a nuclide containing activity A. The exposure rate constant is characteristic of the particular radionuclide. For radium and radon, d is determined for a point source with a filter thickness of 0.5 mm platinum and hence is numerically equal to the specific gamma ray constant. For all other radionuclides, the constant is determined for unfiltered emissions and hence a correction is required to allow for source encapsulation.
2.1.6
Equivalent activity
In practice, it was difficult to determine the activity content of an existing source and so the equivalent activity, A eq , was often used in its place [17]. This was calculated by determining the exposure rate external to the source, and then using the exposure rate constant to obtain the activity of a hypothetical, unfiltered, point source that would give the same result. There is scope here for much confusion and possible error, particularly if those commercial computer systems that offer a brachytherapy dosimetry package fail to specify clearly whether it is the true activity content, or the equivalent activity, that is required when entering source data.
2.2
2.2.1
SPECIFICATION BY EMISSION
Air kerma strength
As an alternative to using a content quantity, such as radium mass, radium mass equivalent, or activity, a brachytherapy source may be specified directly in terms
of an emission property, so avoiding errors due to uncertainties in the exposure rate constant or any other similar parameter. At the same time dosimetry errors that may be made when allowing for the encapsulation material will be reduced. The benefits of this approach are summarized by Jayaraman et al. [18]. Specification in terms of a reference exposure rate was proposed by Wambersie et al. [19], and in the following year this was the subject of a formal recommendation by the National Council on Radiation Protection and Measurements [20]. However, as exposure rate has now been replaced by air kerma rate in many aspects of fundamental radiation dosimetry, specification quantities that are based on air kerma rate are now being recommended instead. The French Committee on Measurements of Ionising Radiations (CFMRI) [21] and the American Association of Physicists in Medicine (AAPM) [22,23] have each independently recommended a quantity that is defined as the product of the air kerma rate at a distance /, measured along the transverse bisector of the source, and the square of the distance /. The distance / must be large enough that both the source and the detector may be treated mathematically as points. The CFMRI called the quantity le debit de kerma normal, and the AAPM use the term air kerma strength. The latter term will be used in this chapter, but an international readership must be wary not to translate strength as force, which is the dictionary translation for some European languages. The AAPM have assigned the symbol U to the unit of air kerma strength, where, for a point source:
2.2.2
Reference air kerma rate
Various other national and international organizations [24-27] have defined an air kerma specification quantity as the air kerma rate at a reference distance of 1 m from the center of the source. The precise wording of the definition differs slightly in the different publications, but the quantity is, in practice, the same, and several of the reports and recommendations assign the name reference air kerma rate. The definition given in the BIR/IPSM recommendations [27] is that the reference air kerma rate is the kerma rate to air, in vacuo, at a reference point which is 1 m from the center of the source, and that for needles, tubes, and other similar rigid sources, the direction from source center to the reference point is that at right-angles to the long axis of the source. It was recognized by the authors of the report that measurements at 1 m, in vacua, and in scatter-free conditions, would not be possible, and that the magnitude of the reference air kerma rate for any given source would have to be derived from measurements made in other conditions, the most likely being ionization measurements made in air, and conceivably at distances of less than 1 m. In deriving the
14 Source specification and dosimetry
reference air kerma rate from such measurements, it will be necessary to convert the measured charge released to a statement of energy released by using a calculated value of the average energy required to produce one unit of ionization in air. It will also, in principle, be necessary to correct for attenuation and scattering in air, for the response of any detector that cannot be regarded as a point detector, and for any deviation from the inverse square law when extrapolating from the actual measurement distance to 1 m. Other techniques of measurement, and other methods of deriving the magnitude of the specification quantity from such measurements, although very unlikely, are not excluded by the BIR/IPSM definition. The recommended units for the reference air kerma rate are mGy h-1 for low dose-rate sources, i.e., those used in applications where treatment durations are quoted in hours, progressing to mGy min-1 and mGy s-1 where the treatment durations would be expressed in minutes or seconds respectively. The air kerma strength of a source, expressed in U, and the reference air kerma rate, expressed in (mGy h-1, although dimensionally different, will be numerically the same for all practical brachytherapy purposes. The discussions that follow on the relationship between the older content specifications and reference air kerma rate will therefore be equally applicable to air kerma strength.
2.2.3 Radium mass equivalent and reference air kerma rate All advocates of source specification by emission recommend that the practice of source specification by content should be discontinued. However, there is a practical problem here in that many commercial software packages, and some source suppliers, continue to use content quantities. It is therefore necessary, at least during a transition period, to convert from reference air kerma rate to milligram radium equivalent or to equivalent activity. Using a specific gamma ray constant of 8.25 Rh-1cm2 mg~' for radium with 0.5 mm platinum filtration, the average energy per unit charge released by ionization in air of 33.97 J C-1 [28], and taking the charge released per unit mass of air by one roentgen of exposure to be 2.58X 104 C kg-1 [15], then a point source containing 1 mg radium equivalent will give an air kerma rate of 7.23 mGy h-1 at 1 m.
2.2.4 Equivalent activity and reference air kerma rate To convert between reference air kerma rate and equivalent activity requires knowledge of the appropriate air kerma rate constant. With the demise of the quantities exposure and exposure rate, the air kerma rate constant has replaced the exposure rate constant in modern radi-
ation dosimetry. Unfortunately, the ICRU has assigned the same symbol Gd to both constants. The air kerma rate constant is defined by the quotient:
where (dK/dt)d is the air kerma rate due to photons of energy greater than 8, at a distance / from a point source of the nuclide containing activity A [15]. The weakness of this approach is that different parties may adopt different values for the air kerma rate constant. For iridium-192 several values have been proposed; for example, Godden [29] recommends 0.111 LlGyrr'm'MBq-1, whereas Dutreix et al. [30] suggest 0.1157 mGyh-1m2MBq-1. Clearly, great care is required to avoid significant systematic error.
2*3 DOSE-RATE CALCULATION FROM A REFERENCE AIR KERMA RATE SPECIFICATION
2*3*1 Reference air kerma rate and spherical sources with isotropic emission For small spherical sources with isotropic emission, the most commonly used expression for calculating the dose rate to water in water, dD(r)water/dt, at radial distance, r, is:
where (dfCair/dt)ref is the reference air kerma rate specification for the source, f(r) is a radial function describing the net effect of attenuation and scattering in water, the term in square brackets is the ratio of the mass energy absorption coefficient for water to the mass energy transfer coefficient for air, and (djr)2 gives inverse square scaling from the reference distance dr, (df equals 100 when r is in cm). The value of absorption coefficient to transfer coefficient ratio may be calculated from the data published by Hubbell [31], and for photon energies between 150 keV and 1.5 MeV is in the range 1.107-1.112. Hence a single value of 1.11 may be adopted without incurring serious error for most of the commonly used brachytherapy radionuclides. Strictly speaking, however, this ratio is a function of photon energy and care must be exercised when using this approach with nuclides of low energy emissions and where the energy spectrum will be further significantly degraded by scatter. Inverse square scaling will also break down when very close to a finite-sized source, but this has no practical dosimetric consequences.
2*3*2 Attenuation and scattering in the irradiated medium The net effect of attenuation and scattering in water has been investigated both experimentally and by Monte
The AAPM recommendations 15
Carlo techniques. Meisberger et al. [32] summarize the earlier work and recommend values for coefficients of third-order polynomials for the function f(r). The Meisberger polynomials became the most common correction method, but early Monte Carlo calculations [33] suggested that these polynomials were suspect. However, more recent Monte Carlo work, for example Sakelliou et al. [34], is in much closer agreement. The BIR/IPSM report [27] recommends polynomials based on Sakelliou's work. Klevenhagen [35] demonstrated that the polynomial approximation must break down both at very small and at very large distances, but, as the correction is usually very small when using the common radionuclides at short distances, this may be ignored in practice.
233
Seed sources
Small seeds containing, originally, radon, but more recently gold-198 or iridium-192, should strictly be considered as cylindrical sources. However, where a large number of such seeds are randomly orientated in a permanent implant, a more practical approach is to treat them as point sources but to include an anisotropy correction giving the average emissions over all angles. Anisotropy corrections may not be applicable, however, when the seeds are arranged in more controlled geometries, such as on a plaque for a superficial treatment. 2.3.4
Reference air kerma rate and
cylindrical line sources For a line source the BIR/IPSM recommendations [27] advocate an adaptation of the Sievert Integral evaluated by a summation of the contributions to the total dose rate from N contiguous line source elements, each no more than 1 mm in length. Each line element is subjected to a different inverse square scaling, to a different water absorption and scattering correction, and to oblique filtration corrections for both the source encapsulation material and also for the source material itself. With reference to Figure 2.2, the expression for the dose rate at radial distance r and angle 0 is:
Figure 2.2 Intregral evaluated as the sum of the dose contributions from many small contiguous line source elements.
expression (equation 2.7). The angle convention in the above equation has been changed from the original BIR/IPSM publication so as to be consistent with the AAPM formalism, which will be described later in this chapter. The BIR/IPSM approach represents an improvement over the original Sievert Integral in as far as there is now an allowance for self-absorption in the source material, and in that water attenuation and scattering are included, but there remains the problem of choosing appropriate values for the filtration coefficients. The BIR/IPSM report recommends the use of the linear absorption coefficients, as opposed to the linear attenuation coefficients, for the mean photon energy of the radionuclide concerned. For the higher energy emitters with stainless-steel encapsulation, this will be a good approximation, but will be less good when there is a low energy component and when significant thicknesses of high-density, high atomic number materials are involved. For iridium-192 sources, for example, there will be moderately large errors in local dose calculations at points close to the axis where the oblique filtration thicknesses in the source material itself are relatively large [36]. However, this is of academic interest only and will not significantly affect the calculation of treatment times for clinical applications.
2.4
THE AAPM RECOMMENDATIONS
2.4.1 Low energy emitters and the general AAPM formalism
where ts(qi) is the thickness of the encapsulation material at angle qi, and ta(qi) is the thickness of source material at the same angle measured from the source axis; ms and ma are the corresponding filtration correction coefficients. The ratio of water absorption coefficient to air transfer coefficient has been given the value 1.11, assuming that the source will be one of the higher energy emitters. The other symbols are as for the spherical source
As indicated in the previous section, the BIR/IPSM approach is not satisfactory with low energy emitters, and indeed is starting to be suspect with iridium-192. In North America, most interstitial brachytherapy is done using seeds of either iridium-192 or the very low energy emitter iodine-125. Two other low energy emitters, palladium-103 and ytterbium-169, have also been
16 Source specification and dosimetry
metry, at 1 cm from the center of a unit air kerma rate source of that type. Thus:
Figure 2.3 The geometry pertaining to the formalism recommended by the MPM Radiation Therapy Task Group.
investigated for brachytherapy applications (see, for example, Meigooni et al. [37], or Chiu-Tsao and Anderson [38] for palladium, and Perera et al. [39] or MacPherson and Battista [40] for ytterbium). The formalism recommended by the AAPM [23] attempts to solve the problem by incorporating 'a direct use of measured or measurable dose distributions produced by a source in a water equivalent medium.' However, the difficulties associated with measuring low dose rates, in very high dose gradients, and with finite-sized detectors which may or may not be energy dependent, should not be underestimated, and anyone attempting to implement this protocol should only use data that have been validated and approved by the appropriate AAPM task group. In practice, it would appear that much reliance is being placed on Monte Carlo calculations. A critical review of published work on Monte Carlo calculations and dose distribution measurements for those brachytherapy sources commonly used in interstitial treatments in North America has been included with the published AAPM recommendations [23]. The geometry for the AAPM formalism is shown in Figure 2.3. In addition to the source specification quantity, air kerma strength, symbol Sk, the general formalism introduces several other new quantities. These are the dose rate constant, A; the geometry factor, G(r,q); the radial dose function, g(r); and the anisotropy function, -F(r,q). For cylindrically symmetric sources, the expression for calculating the dose rate by this formalism, is:
The symbol dD(r,q)water/dt, for dose rate to water in water at radial distance r and angle q, has been retained so as to maintain consistency with the previous notation in this chapter.
2.4.2
The dose-rate constant, A
The dose-rate constant for a particular source type is the dose rate to water in water, in the radial plane of sym-
It is an absolute quantity which includes consideration of the source geometry, the spatial distribution of the active material within the source, self-absorption of the radiation and scattering within the source material, attenuation and scattering within the encapsulation material, and attenuation and scattering within the water medium. As the quantity is inversely proportional to the air kerma strength, any future systematic change in the air kerma strength specification for a particular source type, such as may arise from a change in calibration technique, must be accompanied by an equal and opposite change in the value of the dose rate constant. The other terms in the formalism are all relative quantities and normalize to unity at r=l cm and 0=71/2.
2.4.3
The geometry factor, G(r,q)
The geometry factor is included in the formalism to enhance the accuracy of interpolation between tabulated discrete values of both the radial dose function and the anisotropy function in regions of very high dose gradient. Its purpose is to remove the effects of the inverse square law on the dose distribution, and its use is perhaps most easily understood when considering the case of a small, spherical source with isotropic emission (i.e., when F(r,q) is unity for all values of r and 9). In such a case, the geometry factor takes the value of 1/r2 for all angles, leaving the radial dose function, g(r), as a very slowly varying function of radial distance describing only the net effect on the dose rate of attenuation and scattering in water. For a line source it takes the value of the Sievert Integral for zero filter thickness, thus:
2.4.4
The radial dose function, g(r)
The radial dose function describes the relative variations in the dose to water in water along the transverse axis of the source (i.e., in the radial plane of symmetry only). It excludes the effect of inverse square fall off, but includes the net effect of absorption and scattering in the medium and, for points close to the line source, any effects of oblique filtration in both the source and the source encapsulation materials. It may be defined mathematically as: g(r) = [dD(r,p/2) water /dfG(r = l,p/2)]/[dD(r =l,p/2)water/dtG(r,p/2)]
(2.12)
References 17
2.4.5
The anisotropy function, F(r,®jq)
The two-dimensional anisotropy function describes the relative variations in dose at points away from the transverse axis of the source. It may be defined mathematically as:
It allows for the effects of oblique filtration in both the source material and the sheathing material, the effects of internal scattering within the source, and the effects of attenuation and scattering in the surrounding water medium.
2.4.6 Anisotropy factor, q an (r) t and anisotropy constant, f)an The complete formalism using the anisotropy function describes the dose distribution around individual line sources. In practical cases where there are a large number of randomly orientated small seed sources, and where the distances involved are generally greater than the source dimensions, it may be more convenient to use a point source approximation. The formalism may then be simplified as follows:
where f an (r) is the so-called anisotropy factor. It gives the averaged dose rate over all angles at radial distance r, relative to the dose rate at the same radial distance, r, on the transverse axis. For the more common seed sources it is possible to replace the anisotropy factor with a distanceindependent anisotropy constant without any significant loss in accuracy. Typical values of the anisotropy constant range between 0.9 and 0.98, depending on the radionuclide concerned and details of the source construction. It should be noted, therefore, that when using this simplified formalism the calculated dose rates will be typically 2-10% less than those on the transverse axis, and this could result in significant error in techniques where the source orientation is controlled.
2.5
SUMMARY
It is unfortunate that two authoritative, but apparently contradictory, formalisms for brachytherapy dosimetry are currently being recommended. It is particularly unfortunate that there are two air kerma specification quantities, which have different names and different unit dimensions, but which, for practical purposes, are interchangeable. A recent European publication [41] uses the American formalism in conjunction with a reference air kerma rate specification, and perhaps this hybrid
approach will become more generally accepted in the future. The BIR/IPSM formalism may be used with confidence for steel-encapsulated cesium-137 sources, and will give very acceptable results for clinical dosimetry when using high dose-rate iridium-192 stepping sources, and for iridium wires and iridium seeds in ribbons. However, the Sievert Integral approach is certainly not satisfactory with the lower energy emitters and a formalism based on measured parameters has its attractions, but the difficulties encountered in making precise and accurate dose rate measurements in the immediate vicinity of low activity sources are considerable. Confidence in Monte Carlo calculations in brachytherapy suffered a setback when early attempts were subjected to criticism and revision, but the more recent code should be better and, when used in conjunction with measurements, offers a reasonable method of determining the parameters for use in the AAPM formalism.
REFERENCES 1. Sievert, R. (1921) Die Intensitatsverteilung der primaren Gammastrahlung in der Nahe medizinischer Radiumpraparate. Acta Radial., 1,89-128. 2. Shalek, R.J. and Stovall, M. (1990) Brachytherapy dosimetry. In The Dosimetry of Ionizing Radiations, Vol. Ill, ed. K.R. Kase, B.E. Bjarngard and F.H. Attix. San Diego, Academic Press. 3. Williamson, J.F., Morin, R.L and Khan, F.M. (1983) Monte Carlo evaluation of Sievert Integral for brachytherapy dosimetry. Phys. Med. Biol., 28,1021-32. 4. Paterson, R. and Parker, H.M. (1938) A dosage system for interstitial radium therapy. Br.J. Radiol., 1,252-340. 5. Whyte, G.N. (1955) Attenuation of radium gamma radiation in cylindrical geometry. Br.J. Radiol., 28,635-6. 6. Paterson, R. and Parker, H.M. (1934) A dosage system for gamma ray therapy. Br.J. Radiol., 7, 592-632. 7. Tod, M.C. and Meredith, W.J. (1938) A dosage system for use in the treatment of cancer of the uterine cervix. Br. J. Radiol.,11,809-24. 8. Tod, M.C. and Meredith, W.J. (1953) Treatment of cancer of the cervix uteri - a revised Manchester method. Br. J. Radiol., 26,252-7. 9. Meredith, W.J. (ed.) (1967) Radium Dosage: the Manchester System, 2nd edn. Edinburgh, Livingstone. 10. Quimby, E.H. (1932) The grouping of radium tubes and packs to produce the desired distribution of radiation. Am.J. Roentgenol. Radium Ther., 27,18-38. 11. Quimby, E.H. (1935) Physical factors in interstitial radium therapy. Am.J. Roentgenol. Radium Ther., 33, 306-16. 12. Quimby, E.H. (1947) Radium dosage in radium therapy. Am.J. Roentgenol. Radium Ther., 57,622-7. 13. Cowell, MAC. (1937) Research into time factors in radiotherapy. 14th Annual Report of the British Empire Cancer Campaign. London, British Empire Cancer Campaign, 97-103.
18 Source specification and dosimetry 14. Paterson, R. (1948) The Treatment of Malignant Disease by Radium and X-rays. London, Edward Arnold. 15. ICRU (1980) ICRU Report 33. Radiation Quantities and Units. Washington, DC, International Commission on Radiation Units and Measurements.
Sciences in Medicine. London, British Institute of Radiology. 28. Moretti, C.J. (1992) Changes in the National Physical Laboratory standard for X-ray exposure and air kerma. Phys. Med. Biol., 37,1181-3.
16. ICRU (1971) ICRU Report 19. Radiation Quantities and
29. Godden, T.J. (1986) Physical Aspects of Brachytherapy.
Units. Washington, DC, International Commission on Radiation Units and Measurements. 17. ICRU (1962) ICRU Report 10c. Radioactivity. Washington, DC, International Commission on Radiation Units and
Bristol, Adam-Hilger. Philadelphia. 30. Dutreix, A., Marinello, G. and Wambersie, A. (1982) Dosimetrieen Curietherapie. Paris, Masson. 31. HubbellJ.H. (1982) Photon mass energy absorption
Measurements. 18. Jayaraman, S., Lanzl, LH. and Agarawal, S.K. (1983) An overview of errors in line source dosimetry for gamma-ray brachytherapy. Med. Phys., 10,871-975. 19. Wambersie, A., Prignot, A. and Gueulette, J. (1973) A propos du remplacement du radium par le caesium-137 en Curietherapie. y. Radiol. d'Electrol. Med. Nud., 54, 261-70. 20. NCRP (1974) Report No. 41. Specification of Gamma Ray Brachytherapy Sources. Washington, DC, National Council on Radiation Protection and Measurements. 21. CFM Rl (1983) Recommendations pour la Determination des Doses Absorbees en Curietherapie. Rapport du Comite Francais'Mesuredes Rayonnements lonisants' No. 1. Paris, Bureau National de Metrologie. 22. AAPM (1987) Specification of Brachytherapy Source Strength. Report 21. Task Group 32 of the American Association of Physicists in Medicine. New York, American Institute of Physics. 23. Nath, R., Anderson, LL, Luxton, G. et al. (1995) Dosimetry of interstitial brachytherapy sources: recommendations of the AAPM Radiation Therapy Committee Task Group No. 43. Med. Phys., 22,209-34. 24. BCRU (1984) Specification of brachytherapy source. Memorandum from the British Committee on Radiation Units and Measurements. Br.J. Radiol., 57,941-2. 25. ICRU (1985) ICRU Report 38. Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology. Bethesda, Maryland, USA. International Commission on Radiation Units and Measurements. 26. NCORD (1991) Recommendations for Dosimetry and Quality Control of Radioactive Sources used in Brachytherapy. Amsterdam, Netherlands Commission on Radiation Dosimetry. 27. Bl R/l PSM (1993) Recommendations for Brachytherapy Dosimetry. Report of a Joint Working Party of the British Institute of Radiology and the Institute of Physical
coefficients from 1 keV to 20 MeV. Int.J. Appl. Radial Isot., 33,1269-90. 32. Meisberger, LL, Keller, K.J. and Shalek, R.J. (1968) The effective attenuation in water of the gamma rays of gold198, iridium-192, caesium-137, radium-226 and cobalt60. Radiology, 90, 953-7. 33. Webb, S. and Fox, R.A. (1979) The dose in water surrounding point isotropicgamma emitters. Br.J. Radiol., 52,482-4. 34. Sakelliou, L, Sakellariou, K., Sarigiannis, K. et al. (1992) Dose rate distributions around Co-60, Cs-137, Au-198, Ir-192, Am-241,1-125 (models 6702 and 6711) brachytherapy sources and the nuclide Tc-99m. Phys. Med. Biol., 37,1859-72. 35. Klevenhagen, S.C. (1993) Oral presentation at the British Institute of Radiology, London. 36. Williamson, J.F. (1996) The Sievert Integral revisited: evaluation and extension to 1251,169Yb, and 192lr brachytherapy sources. Int.J. Radial Oncol. Biol. Phys., 36,1239-50. 37. Meigooni, A.S., Sabnis, S. and Nath, R. (1990) Dosimetry of 103
Pd brachytherapy sources for permanent implant. Endocuriether. Hypertherm. Oncol., 6,107-17.
38. Chiu-Tsao, S.T. and Anderson, LL (1991) Thermoluminescent dosimetry for 103Pd (model 200) in a solid water phantom. Med. Phys., 18,449-52. 39. Perera, H., Williamson, J.F., Li, Z., Mishra, V. and Meigooni, A.S. (1994) Dosimetric characteristics, air kerma strength calibration and verification of Monte Carlo simulation for a newytterbium-169 brachytherapy source. Int.J. Radial Oncol. Biol. Phys., 28,953-70. 40. MacPherson, M.S. and Battista, J.J. (1995) Dose distribution and dose rate constant for new ytterbium-169 brachytherapy seeds. Med. Phys., 22,89-96. 41. Permattei, A., Azario, L, Rossi, G. etal. (1995) Dosimetry of 169 Yb seed model X1267. Phys. Med. Biol., 40, 1317-30.
3 Calibration of sources COLIN H.JONES
3.1
INTRODUCTION
Although commercial suppliers of brachytherapy sources provide a measure of source strength, it is unwise to rely solely on this value for patient dose calculations. Manufacturers usually specify source strengths within a broad range of activity. Most departments planning to provide brachytherapy should have the ability to verify source strengths independently and to improve the overall precision of the measurement. The radiation characteristics of an encapsulated source are strongly dependent upon the chemical composition of the radionuclide, the inert filler material, their distribution within the source, and the details of the source encapsulation. Also in relation to source calibration, the presence of radioactive impurities may require a storage period after initial production to allow for the decay of short half-life isotopes. Details of the construction of sources are given in Chapter 1. Such information is important, because attenuation in the source capsule may significantly alter the dose distribution around the source and affect the dose calibration in a variety of ways, especially when measurements are made with re-entrant ionization chambers. It is possible for two sources of different construction to have the same source strengths but significantly different radiation distributions close to the sources. The possibility that such differences might influence calibration measurements must be taken into account. Source specification is considered in Chapter 2. In summary, the source strength is specified as the air kerma rate, in air, at a reference distance of 1 m, corrected for attenuation and scatter in air. The unit to use for low dose-rate brachytherapy sources is (mGy tr1 and
mGys-1for high dose-rate (HDR) applications. The specification quantity is called the reference air kerma rate (RAKR), which is the name used by the ICRU [1].
3.2 REFERENCE STANDARDS AND TRACEABILITY The calibration of sources is traceable to national or international standards at various levels [2]. Direct traceability is established when a source or calibrator has been calibrated at a national standards laboratory or an accredited dosimetry calibration laboratory. Secondary traceability is established when the source is calibrated in comparison with a source of the same design and comparable strength which has direct traceability or when the source is calibrated using an instrument with direct traceability. Secondary traceability by statistical inference is a term that is used for multiple sources of the same activity from which a suitable random sample has been calibrated with secondary traceability. Remote traceability occurs if the user relies upon the manufacturer's calibration as the only standard, which may, or may not, be traceable to a national or international standard. Ideally, brachytherapy sources used clinically should have calibrations with direct or secondary traceability to national standards. In the UK, the traceability routes for the calibration of brachytherapy sources supplied by Nycomed Amersham plc have been summarized by Rossiter [3]. The first line of traceability for this supplier's cobalt-60 sources is in terms of the quantity 'activity.' Reference sources were
20 Calibration of sources
compared with a base standard of cobalt-60 whose activity was measured by absolute counting at the National Physical Laboratory (NPL). The air kerma rate of the base standard was calculated from the measured activity and associated energy fluence, making allowances for gamma ray absorption in the capsule, in the material itself, and in air. Radium-226 sources have been treated similarly, the activity being determined by comparison with the British National Radium Standard. Two additional methods have been used to confirm traceability to national standards of source air kerma rates. The first, used for both cobalt-60 and cesium-137, involved ionometric comparisons with a standard radium-226 source, and the second the measurement of the exposure rate of sources in a scatter-free area at NPL by a large volume chamber (200 mm diameter) directly calibrated against the national primary standard [4]. This work was repeated by Rossiter et al. [5] for cobalt-60, cesium-137, and radium-226, and extended to include iridium-192 wire sources. The results, shown in Tables 3.1 and 3.2, indicate good agreement between NPL and Amersham
source output measurements and provide assurance on source air kerma rate figures provided by this major source supplier and which are based on a traceability route to standards of activity. In the USA, the National Institute of Standards and Technology (NIST) maintains air kerma strength standards for sealed sources of iodine-125, iridium-192, and cesium-137. It should be noted, however, that the strength specification of sealed sources is in terms of air kerma strength (Sk), which is defined as the product of air kerma rate in free space, K(d), measured along the transverse bisector of the source, and the square of the measurement distance d:
The distance d must be large enough that both source and detector may be treated as mathematical points. Such standardization measurements are performed in air using air-attenuation corrections if needed. Sk has units of mGy m2 h-1 and these units are denoted by the
Table 3.1 Comparison of air kerma rate values in vacuo at1 m distance. (Rossiter, Williams, and Bass, 1991 [5].j
137 60
Cs(1802MC)
Co(HR117)
18.7.89
78.70
18.7.89
88.71
226Ra (S5)
18.7.89
36.56
192
13.3.90 13.3.90
29.44
lr(A49945) 192 lr(A49946)
29.39
77.79 87.77a
88.1 7b 36.52 29.25 29.22
1.012 1.011 1.006 1.001 1.006 1.006
' Comparison with60Coreference source. b Comparison with 226Ra reference source. NPL= National Physical Laboratory.
Table 3.2 Air kerma rate measurement uncertainties. (Rossiter, Williams and Bass, 1991 [5].)
60
Co, 137Cs, 226Ra Random Non-random
Determination of secondary standard calibration factor Source measurements Determination of secondary standard calibration factor (all energies) Measurement of pressure Measurement of temperature Measurement of distance Air attenuation correction Correction for finite chamber size
±0.4 ±0.4 ±1.2 ±0.1 ±0.2 ±0.1 ±0.2 ±0.2 ±2.5(60Co)
±1.4
Overalla
±1.8(137Cs) ±0.9 (226Ra)
192|
r
Non-random Overalla aQuadrature sum. NPL= National Physical Laboratory.
Uncertainties as above Weighting procedure for secondary standard factor
±1.4 ±0.4 ±1.5
Calibration methods 21
symbol U. A set of equations has been developed for unambiguously converting source strength estimates and renormalizing published dose-rate tables, which assume traditional quantities and units, into forms consistent with air kerma strength [6]. These authors list the factors to convert source strength of a selection of nuclides from apparent millicuries (mCi) to air kerma strength. The factors are independent of source geometry, but depend on the nominal exposure rate constant value selected by the vendor. Conversion factors applicable to mass of radium or true activity depend upon both source geometry and radionuclide identity. It should be noted that because many of these conversion factors depend upon vendor choices of physical constants and exposure rate constants, users should review source strength specification practices employed by the vendor. This is a requirement even when an independent calibration is made, because a comparison of the measured source strength with that provided by the vendor is a useful and necessary quality assurance procedure. Although an institution might accept the manufacturer's calibration, it is the responsibility of the institution to verify that the manufacturer's stated value is correct. If the measured source strength disagrees with the manufacturer's data by more than 5%, the source of disagreement should be investigated and any unresolved disparity should be reported to the manufacturer. In the case of a batch of sources, a 3% tolerance is probably more applicable, because individual sources may differ from the mean by a greater amount. Discrepancies greater than the accuracy limits specified by the manufacturer should always be explored further. For further reading on this subject, the reader is referred to reference 7.
33
CALIBRATION METHODS
There are three principal methods of calibrating brachytherapy sources. The most frequently used method employs a calibrated re-entrant ionization chamber. The second method makes use of an ionization chamber to measure the air kerma rate at a known distance from the source. In the former method, calibration of the reentrant chamber is actually achieved by use of a radiation source, the air kerma strength of which has been previously measured in air. The third method uses a solid phantom into which source(s) and ion chamber can be introduced in a convenient and reproducible way. In addition, experiments have been conducted on novel dosimetry methods [8], but are not suitable for routine calibrations at the present time. Re-entrant chambers are preferred for the calibration of conventional low-strength brachytherapy sources [9,10], and ionization chambers measuring the air kerma rate at a distance are preferred for HDR sources
[11,12]. However, ionization chambers have been used successfully for conventional dose-rate sources [13], and re-entrant chambers can be used for HDR sources [14]. Baltas et al. [15] report on a comparison of different calibration methods for HDR sources, and conclude that satisfactory results can be obtained by both re-entrant chambers and phantom methods.
3.3.1
Re-entrant ionization chambers
These instruments are characterized by a cylindrical well and an ion collection volume, which surrounds the source approximating at 4p measurement geometry. The re-entrant (well-type) ionization chamber should respond linearly throughout its measuring range; its energy response must be known and care must be taken to ensure that when measuring high activities there is no drop in sensitivity. The response of the chamber will be dependent upon the geometric configuration of the source, its filtration, and encapsulation. The use of such an instrument for intercomparison of sources requires great care and it is advisable for potential users to ascertain the characteristics of the chamber before embarking on measurements. The report of AAPM Radiation Therapy Committee Task Group 40 [2] describes the physical characteristics of a suitable calibrator. It is recommended that the reproducibility of the calibrator should be better than 2% and the signal-to-noise ratio greater than 100:1. The response of the chamber is dependent on the orientation of the source and its position in the well [16], so it is essential to devise a source holder that will reproduce the source positioning. It is also recommended that the scale factor and linearity of each scale used on the electrometer be determined and monitored. The collection efficiency should be better than 99% for commercial well chambers using conventional brachytherapy sources. The sensitivity of reentrant ionization chambers depends on the energy of the photons, thus a calibrated source of one radionuclide cannot be used to determine the source strength of another radionuclide. Similarly, because of dose anisotropy about the source, the relative orientation of the source axis is important for any calibrator. The source should be moved through the active volume of the chamber to verify and quantitate the extent of the change in sensitivity with source position. Figure 3.1 illustrates a typical re-entrant chamber and Figure 3.2 shows the variation of sensitivity with source location in the chamber well. The source-length dependence of the chamber should also be investigated: this is best achieved by determining the chamber response for wire sources of different lengths. The source-length dependence may also be a function of the radionuclide. Williamson et al. [16] showed that a calibrated source of one encapsulation may not be reliable for determining the strength of a source of the same radionuclide but different encapsu-
22 Calibration of sources
Figure 3.1 Selectron Source Dosimetry System (SDS): PTW-Freiburg re-entrant (well-type) chamber with Perspex holders for LDR/MDR Selectron sources and HDR microSelectron sources.
Figure 3.2 Relative response of SDS (PTW-Freiburg) chamber with iridium-192 source inserted at various depths.
lation. For example, two cesium-137 sources of equal strength and of similar size but encapsulated in platinum and stainless steel, respectively, might cause different chamber responses. In practice, it is usual to use a positioning device to assure reproducible positioning of the source close to the longitudinal chamber axis and where the chamber sensitivity is high but least dependent upon geometrical positioning of the source. AAPM Report 13 [17] describes the use of re-entrant (well-type) ionization chambers for measuring different types of brachytherapy sources. • Long-lived sources (cesium-137, cobalt-60 etc.)
1. For each radionuclide (and encapsulation) to be measured, one source should be identified as the standard source. The source should be marked or otherwise identified so that it can be recognized at a later date. It is appropriate to ensure that the source selected is typical of other sources in the batch. 2. The standard source should be sent to an appropriate calibration laboratory for calibration. 3. The standard may be used to calibrate all other similar sources by sequential placement of the standard source and the sources to be calibrated in the same geometry within the chamber and
Calibration methods 23
comparing readings. By correcting for decay of the source, it is also possible to use the standard source to check for long-term chamber stability and chamber malfunction. • Short-lived sources (iridium-192, gold-198 etc.) 1. Identify a long-lived source as the reference source. This source may be a standard source for another radionuclide. 2. Obtain a standard source of the appropriate short-lived isotope and compare this with the reference source. This intercomparison will be used to establish a baseline comparison of the relative sensitivity of the system to the two sources. 3. Submit the standard source to a suitable calibration laboratory for calibration. 4. There are two methods that can be used to transfer the calibration: (a) The chamber is calibrated with the shortlived standard source, and the reference source is used to check that the chamber is functioning properly. This requires temperature and pressure corrections to be made if unpressurized ambient air chambers are used. (b) A correction factor defined as the ratio of two measurements of chamber response using the standard source is calculated. The correction factor relates the response of the chamber to the short-lived standard source in terms of the response to the reference source. 5. Whichever method is used, the reference source is measured every time the chamber is used to calibrate the short-lived sources. 6. After decay of the standard source, the reference source is used for subsequent calibrations. Ideally, every radioactive source that is to be used in a patient should be calibrated. In practice, this is not always possible. For short half-life sources such as iodine-125 and iridium-192 seeds, traceability by statistical inference may be appropriate, depending upon the number of ribbons or seeds in the designated strength groupings under consideration. Kutcher et al. [2] recommend that if the grouping contains only a few seeds or ribbons, all seeds should be calibrated. For groupings with a large number of loose seeds, it is recommended that a random sample containing at least 10% of the seeds be calibrated. For a large number of seeds in ribbons, a minimum of 10% or two ribbons (whichever is larger) should be calibrated. For sources purchased in a sterile configuration, the report recommends purchasing and calibrating a single (non-sterile) seed for each designated strength grouping. The calibration of iodine-125 seeds is traceable to a calibration at the NIST. The dosimetry is more complex
than other brachytherapy sources as the nuclide decays principally by electron capture emitting characteristic X-rays at energies from 27.2 to 31.8 keV and 35.5 keV gamma rays. The construction of the source affects the mean energy of the emitted radiation and any calibration must be made with a calibrated source of the same design as the sources being investigated. All long half-life sources should be calibrated. The use of re-entrant chambers is illustrated further in the following sections, where specific calibration procedures are described.
3.3.2 In-air' method using ionization chambers The 'in-air' measurement technique is the primary method of determining the strength of a brachytherapy source. Unfortunately, the method has several inherent problems and, consequently, with the exception of the calibration of strong sources (used for HDR afterloading), its use is confined largely to specially equipped laboratories. For a low strength source, the air kerma rate at a large distance from the source will be low and difficult to measure; scattering from surroundings is also a problem, for which allowance must be made. Measurements can be made closer to the source, but as the measurement distance is decreased, the significance of the physical dimensions of both the source and the ion chamber increase. Positioning uncertainty also becomes a problem for shorter distances. In the absence of a calibrated re-entrant chamber, the 'in-air' ion chamber method is a convenient means of checking the relative strengths of individual sources by comparing the signal strength for the unknown source against that for a reference source of known air kerma strength [13]. The method is less dependent upon the effects of differences in encapsulation than for re-entrant chamber measurements. To measure low air kerma rates requires large volume chambers to achieve a signal-to-noise ratio better than 100:1 and a wide range of chamber types have been employed. One of the corrections that has to be applied to the electrometer reading is to allow for the dose gradient across the chamber response, which causes the chamber response to differ from that of a point detector. This effect depends upon the relative position of the chamber with respect to the source and the physical dimensions of the chamber. It has been analyzed by Tolli [18], who provides a means of calculating a factor to correct for the gradient effect. This factor is referred to in more detail when the calibration of HDR sources is described. The ideal jig for reproducible 'in-air' calibrations should provide mechanical rigidity without contributing scatter to the chamber. Mechanical devices are suitable for shorter distances of 100-200 mm, but for measure-
24 Calibration of sources
ments made at much longer distances, some form of optical alignment might be more appropriate. It is sometimes possible to make use of radiotherapy machine laser alignment devices for this purpose. Corrections for room-scattered radiation can be made by making measurements at various distances. Room scatter, which can be assumed to be constant over the distances measured, can be determined from examination of the data, assuming that the dose data set after correction for room scatter should comply with the inverse-square-law. The ion chamber and electrometer used for calibration purposes should have a traceable calibration for radiation of the same energy that is being investigated and preferably be relatively insensitive to changes in photon energy over a wide range.
source strengths (such as those used in the HDR Nucletron Selectron). • Single HDR sources (similar to those used in the HDR Nucletron microSelectron, the Gammamed HDR system, the CIS Curietron 192 HDR, and the Varian Varisource HDR remote afterloader). • Pulse dose-rate (PDR) sources.
The following source types are considered:
It is good working practice when calibrating sources used in remote after-loading brachytherapy equipment for a written procedure to be drawn up and followed. Measurements and observations must be fully documented. Whenever possible, the definitive calibration should be derived from two independent sets of measurements made by two physicists experienced in radiotherapy, using different dose instruments. The calibration measurement should be compared with the supplier's certificate of calibration. When necessary, data from the certificate of calibration should be re-calculated using the same units and conversion factors as those employed in the definitive calibration. Any difference between the corrected data and the definitive calibration must be reconciled. Using the clinical data calculated from these measurements, the response of a suitable dosimeter must be calculated for a different treatment time from that used in the calibration. In the case of equipment into which a value of source strength can be programmed, it is recommended that the calculated treatment time should be based on the displayed source strength. The definitive calibration is confirmed if the predicted reading is obtained within the limits of experimental uncertainty. In some situations an allowance will have to be made for any dose that is delivered during transit of the source. In practice, to comply with the above recommendations it is useful to have sufficient instrumentation to facilitate measurements with a calibrated re-entrant ionization chamber and also either an 'in-air' or phantomtype calibration measurement. It is appropriate to note that, despite international agreement that source strength should be specified in terms of the reference air kerma rate, it is recognized that some commercially supplied computer software requires source information in terms of activity. The use of such software requires great care in converting an RAKR specification to an effective content specification. An equivalent activity may be determined from the RAKR by using the following expression:
• Low dose-rate (LDR) sources in the form of wires or ribbons (such as those used in the Nucletron microSelectron). • Low dose-rate preloaded cesium-137 source trains (similar to those used in the CIS Curietron machine). • Multiple low dose-rate cesium-137 sources of similar strengths (such as those used in the LDR/MDR Nucletron Selectron). • Multiple high dose-rate cobalt-60 sources of similar
where Aeq is the equivalent activity, and 10-6 is the mGy to Gy conversion factor. Ft is a constant and equal to 1/3600,1/60 or 1 depending on whether the RAKR is in mGy h-1, mGy min-1 or mGy s-1, respectively; dr is the distance at which the RAKR is defined and is unity; and Gg is an appropriate air kerma rate constant in m2 GyBq-1s-l.
333
Source calibrations in solid phantoms
The third method of calibrating sources employs a solid acrylic phantom, which contains a centrally placed ion chamber with two or more cavities for source catheters positioned radially around the chamber. Source measurements in a solid phantom are more reproducible and straightforward than 'in-air' measurements. In practice, however, for reference air kerma rate measurements, allowance has to be made for phantom attenuation and scatter. Furthermore, at the point of measurement, it is necessary to take into account the replacement of the phantom material by the ionization chamber. These factors, which are characteristic of the measurement set-up, are difficult to determine with precision, so the overall accuracy of the measurement is reduced. Even so, once a solid phantom has been calibrated satisfactorily, the high reproducibility of the technique and its convenience make it very useful for routine quality assurance-type source measurements. Some phantoms are designed to be filled with water: this simplifies attenuation corrections, but the need for scatter and chamber displacement factors still applies.
3.4 CALIBRATION OF SOURCES USED IN REMOTE AFTERLOADING SYSTEMS
Calibration of sources used in remote afterloading systems 25
3.4.1 Low dose-rate sources in the form of wires or ribbons Remote afterloaders that are designed to treat interstitially with wire sources or catheters loaded with radioactive ribbons are equipped with up to 20 or so channels. These may be connected to flexible or rigid implants using iridium-192 wires, or ribbons of cesium-137 microseeds. A re-entrant-type chamber can be used to determine the activity of the individual seeds. In the case of the wire sources, it might be necessary to measure the source strength per unit length along the wire. Special scanning devices have been developed for this purpose; generally, these devices are collimated detectors that can be used to scan the length of the wire, which may be up to 135 mm. Usually, for calibration, a re-entrant ionization chamber is the instrument of choice. This is best achieved by using wire that has been measured at a national standards laboratory. The measurement of this wire provides a baseline value, which can be compared against a source with a longer half-life; this reference source can be used to monitor chamber response over the long time periods in-between the definitive source calibrations. With long wire sources, it is particularly important to use an appropriate source holder in order to maintain the source centrally in the chamber well: offaxis displacement can increase the chamber signal by up to 20%. As part of the calibration procedure, a reference curve should be constructed that shows the dependence of the chamber output upon the length of the source. This may be achieved by using a 10 mm piece of wire to measure the response for different positions of wire inside a thin, straight tube which is positioned precisely along the axis of the well. The characteristics of well-type chambers vary according to commercial design, but a typical response with a 10 mm wire source should be within about 2% over a 100 mm length [19]. Calibration measurements should be made on all sources after wires have been cut to length and sealed in catheters. Steggerda and Mijnheer [20] make reference to the use of a solid phantom in which a 0.6 cm3 Farmer-type ionization chamber is used for dose rate measurements. The Perspex phantom (see section 3.4.3) is 20 cm in diameter and 15 cm high. Three stainless-steel afterloading catheters are positioned in the phantom at a distance of 5.0 cm from the axis of the cylindrical phantom and at 120° angles. Although the phantom was designed originally for measuring Selectron sources, it can also be used for calibrating iridium-192 line sources and cesium-137 microseed trains. 3.4.2 Low dose-rate preloaded cesium-137 source trains A preloaded source train consists of one or more cesium-137 sources in the form of small source capsules
separated by inactive spacers (usually small ball bearings), all of which are contained in a flexible, stainlesssteel spring catheter. Before being used clinically, the location of individual capsules in the source train should be ascertained. One method of achieving this is by autoradiography and by densitometric scanning of the autoradiograph to locate the center of each source. The calibration of individual source capsules is not easily accomplished, especially when sources are very close to each other. It is important for potential users of preloaded source trains to liaise fully with manufacturers before source trains are made up. The manufacturer should be able to guarantee that the strength of each source loaded into the source train does not differ by more than 5% from that stated. It is useful to purchase one or more source capsules for reference purposes, which can be used to study the effect of the steel spring and also to examine the response of the re-entrant calibration chamber when the source is moved along its longitudinal axis. Individual sources encapsulated within a flexible spring can be measured satisfactorily against a reference source either in air using an ion chamber or in a re-entrant chamber. Source trains with multiple sources are more difficult to measure with high precision. The manufacturer is better placed to obtain information about the strength of individual sources prior to fabrication of the source trains. The user can check the relative distribution of activity by film dosimetry, thermoluminescent dosimetry (TLD), and by judicious use of a re-entrant chamber, but overall it is better to obtain as much information as possible about the strength of individual sources prior to loading of the source trains. The location and relative strength of individual sources in a source train can also be recorded by means of a device employing a highly collimated detector.
3.43 Multiple low dose-rate sources of similar strengths To calibrate cesium-137 sources such as those used in the Nucletron LDR/MDR Selectron machine, it is necessary to measure the strength of each source. This is best achieved with a suitably calibrated re-entrant ionization chamber. The sources are cesium-137 glass beads encapsulated in a 2.5 mm diameter stainless-steel pellet; a machine can take up to 48 pellets. The pellets are not readily identifiable and are used effectively in a random manner. Ideally, all pellets would have the same strength, but in practice this is not the case. Figure 3.3 shows two typical sets of source strength measurements. A reentrant chamber is usually calibrated in terms of the source strength (mGy h-1 at 1 m), which produces a current of a n A. When a chamber has not been suitably calibrated, an alternative method of determining source strength must be used.
26 Calibration of sources
3. From the ratio of the measured and calculated air kerma rates in water, the mean actual reference air kerma rate of the set of sources can be determined. Meertens [22] gives the following details about the air kerma measurements and calculations.
KERMA RATE MEASUREMENTS IN WATER
The air kerma rate in water, km, was determined from readings obtained with the ionization chamber and an electrometer using the following formula [23]: Figure 3.3 Source-strength frequency distribution for two batches of cesium-137 LDR/MDR Selectron sources.
Where M is the corrected instrument reading: Aukett [21] has described a method in which a Farmer ionization chamber is used for the direct measurement of the air kerma rate in air for small, spherical cesium137 sources at distances of 35-70 mm. A 2 mm Perspex build-up cap was used on the chamber, which was supported centrally, in air, equidistant from three stainlesssteel applicator tubes, each carrying six sources. Geometry correction factors were calculated for each group of sources. The resultant measurement was found to differ from expected values by 4.4%. An alternative method has been described by Meertens [22]. The method is based on the use of a phantom consisting of three parallel Perspex catheters with a wall thickness of 0.11 cm, fixed together in a cylindrical geometry at 120° angles. A Perspex support for a 0.6 cm3 graphite-walled Farmer-type NE 2505/3A ionization chamber was mounted between the three catheters parallel to their axes; the distance between each of the catheters and the chamber centre was 5 cm. The catheters and the ionization chamber were placed in a 36 x 32 x 20 cm (full scatter) water phantom. The technique was designed to position sources in all three catheters simultaneously in such a configuration that the dose gradients through the chamber volume were minimized. By using multiple sources (ten sources in each catheter), the dose rate was high enough to measure satisfactorily and, by eliminating dose gradients, the need to apply a correction factor for the finite size of the chamber is overcome. To obtain a uniform dose rate at the point of measurement, the ten sources in each catheter were distributed in two batches (2 x 5), separated by 50 mm above and below the level of the chamber center. The procedure designed to measure the mean source strength of a set of sources is as follows. 1. The air kerma rate in water at a point about 5.5 cm distance from a number of sources is measured. 2. The air kerma rate in water at the same point for the same set of sources is calculated for a reference air kerma rate of 100 (mGy h-1 at 1 m in free air for each source.
M=
MuncorPtPpPhum
Muncor is the uncorrected instrument reading; Pt> Pp, Phum are the air temperature, pressure, and humidity correction factors respectively; Pion is the ion recombination correction factor; and Ppol is the correction factor for polarity effects. The last three correction factors were assumed to be unity. Nk (Jkg-1) is the air kerma factor. The product of the correction factors pki (0.989) is given by:
where katt (0.990) is a correction for attenuation in the wall and build-up cap of the ionization chamber; km (0.999) is a correction for the difference in composition between the wall plus build-up cap and air; kst (1.000) is a correction for the stem effect for the employed field size; and kce (1.000) is a correction for the effect of the central electrode on the response of the chamber during calibration. The product of the correction factors ppi (0.972) to be applied to the measurement in water for cesium-137 photons is given by:
where pwall (1.002) corrects for the difference in composition between the ionization chamber wall and water; pd (0.970) is the displacement direction factor that corrects for the displacement of the effective center of the ionization chamber and is a function of the photon energy, the source to chamber distance, and the size of the ionization chamber. This value 0.970 for the 0.6 cm3 (0.3 cm radius) Farmer chamber was estimated from preliminary results of experiments with ionization chambers with radii ranging from 0.1 to 0.8 cm, Pce (1.000) corrects for the effect of the central electrode on the response of the chamber during the measurements in the water phantom. The integration time, t, applied for the ionization current measurements was 0.033 h (120 s).
Calibration of sources used in remote afterloading systems 27
KERMA RATE CALCULATIONS IN WATER
The contribution of one point source with a given reference air kerma rate, Ktefy in mGyh-1 at 1 m free air to the calculated air kerma rate Kc, in cGylr1 in water at a distance, d, in cm from the point source was calculated according to the following formula:
where S(d) is the absorption and scatter correction factor according to Meisberger et al. [24].
and A0 = 1.0091, B0 = -9.015 x 10-3 cnr-1, C0 = -3.459 x 10-4 cm-2, and D0 = -2.817 x 10-5 cm-3. Kc in the calibration point of the water phantom for sources in the three catheters, each with a KKf value of 100 mGyhr1. Meertens [22] also reports on the use of a Perspex calibration phantom smaller than the water phantom but of a similar design. From experiments, it was determined that the cylindrical Perspex phantom 20 cm diameter and 15 cm high gave results 4% lower than those obtained with the water phantom. The measurements in water had a reproducibility of 0.3% (1 SD), with an uncertainty that could not be evaluated by statistical methods of about 1.2%. Perspex phantoms similar to the one described ensure good reproducibility of set-up and, once an appropriate factor relating measurements made in Perspex to those in water has been determined, the smaller solid phantom can be used for routine calibration purposes. The method has also been used by Jones [25] with a different source configuration. Figures 3.4 and 3.5 show the water phantom and the resultant distribution using eight sets of five sources with a 40 mm separation between each set of sources.
straight applicator and measure the air kerma rate with the Farmer-type chamber at a distance of 500 mm. This method has been described and used by Chenery et al [26] and Messina et al. [27]. The chamber used should have a calibration traceable to a national calibration laboratory; it should be used with a build-up cap for cobalt60. For reproducible measurements, some form of fixation device should be used to ensure that the applicator and the ion chamber are parallel and held so that the applicator is not displaced by transfer of the sources. The applicator can be metal or plastic as long as it is rigid. If both types are available, measurements can be made to determine the attenuation effect of the metal applicator. Both applicator and ionization chamber should be at least 1 m from the floor, the walls, or any other large scattering medium. Room scatter will be characteristic of the local environment and an estimate should be made of its magnitude. In most situations it is likely to be between 3% and 4% for the source-chamber geometry described. The principal error is likely to be associated with the position of the sources inside the applicator: they are not constrained to lie perfectly on the axis of the applicator, but tend to zigzag, with displacements of over 0.5 mm occurring 16% of the time [26]. The positional errors, including those associated with the measurement of the applicator-chamber separation, produce an uncertainty in the measured dose rate
3.4.4 Multiple high dose-rate cobalt-60 sources of similar source strengths The cobalt-60 sources used in the Nucletron remote afterloading machine consist of 1.5 x 1.5 mm cylinders encapsulated in titanium spheres 2.5 mm diameter; there are 20 sources in each machine. The relative strength of each source may be determined with a reentrant ionization chamber. If the chamber has not been calibrated for these sources, a Farmer-type 0.6 cm3 ion chamber can be used. A measurement device and source configuration similar to that described in the preceding section might be used, replacing the set of four sources by a single source so that the chamber would be exposed to eight sources simultaneously. An alternative method is to use all 20 sources in a
Figure 3.4 Perspex/water calibration phantom with central cavity for Farmer chamber and four stainless-steel catheters distributed radially at 50 mm.
28 Calibration of sources
Figure 3.5 IGE (Target) computer calculation of dose distribution of four straight Selectron catheters each loaded with
2x5
LDR/MDR Selectron pellets with 40 mm separation between each set of pellets (1.4 GBq per pellet). The calibration chamber is positioned centrally in the region of low dose gradient. Units of distribution are cGyh-1; magnification factor (%) = WO.
of less than 1%. It should be noted that chamber leakage should be measured, for which an appropriate correction might be made. This leakage should not be greater that 0.1-0.2%.
3.4.5
Single high dose-rate sources
High dose-rate remote afterloading devices, such as those listed above (section 3.4), make use of an iridium192 source with an activity of up to 370 GBq. The source is typically in the form of a small pellet (0.5 mm diameter, 4 mm active length, with a 0.3 mm stainless-steel wall), connected to a wire that pushes and pulls it through a plastic catheter to guide it to the desired location. The half-life of iridium-192 is 73.83 days so source replacement is relatively frequent. The principal supplier of these sources is Mallinckrodt Diagnostica in the Netherlands, which provides a calibration certificate with each new source that states the overall uncertainty in activity to be ± 5%. An independent recalibration is required after installation of a source before the machine is used to treat patients. There are three methods of calibrating a single HDR source: 'in-air' measurements with a calibrated ion
chamber positioned at a distance of 10-20 cm from the source; measurements with a re-entrant ionization chamber that has a calibration traceable to a national standards laboratory; and measurements with an ion chamber and a solid phantom (or water phantom). IN-AIR'CALIBRATIONS A Joint Working Party of the BIR and IPSM recommended that iridium-192 HDR sources should be calibrated in air with a Farmer-type 0.6 cm3 ion chamber held at a distance of 100 mm from the source [28], and that the traceability route for such measurements should be through the external-beam standard. The recommended method has the advantage of being widely applicable and experience suggests that the method produces very reproducible results. It also has the advantage that the purchase of additional calibration instrumentation is not required. However, although the procedure is straightforward, a number of interrelated factors have to be taken into consideration. These include the effects of ion chamber geometry, source-chamber distance, positioning reproducibility, dose gradient, signal strength, and room scatter. Furthermore, because national calibration laboratories do not offer calibration of ionization
Calibration of sources used in remote afterloading systems 29
chambers with the gamma ray spectrum of iridium-192, some form of procedure must be employed to determine an appropriate factor for the ion chamber used in the measurement procedure. These matters have been considered comprehensively by Ezzell [29], Goetsch et al. [12], Piermattei [30], Grimbergen and van Dijk [31], and Biiermann et al. [32]. At a source-chamber distance of 100 mm, the correction required to allow for the fluence gradient across the Farmer (0.6 cm3) ion chamber is 0.9% [18]. For shorter distances, the positioning becomes more critical and a larger factor will be required to correct for the finite chamber size (Table 3.3). With a suitable measurement jig, it is possible to restrict dose errors due to positional uncertainties to less than 0.5%. Both source and detector need to be mounted well above floor level and well away from walls or other large structures so as to reduce the effects of room scatter to negligible levels. Several devices have been described in the literature, two of which are shown in Figure 3.6. The ideal jig for reproducible 'inair' calibrations would provide mechanical rigidity without contributing scatter to the chamber. In practice, a small correction will be required to allow for room scatter, including any scatter that might arise from the jig itself. If it is assumed that the air kerma rate at the point of measurement caused by room scatter does not depend on the distance from the source, then:
where d = distance from the source to the point of measurement; d0 = distance from the source to a reference point; X = total exposure at the point of measurement; X0 = primary exposure at the reference distance; and Xs = room scatter exposure (assumed constant for all d). The value for the room scatter correction factor at the distance d is then given by:
Table 3.3 Dose gradient correction factors for NE 2571 Farmer 0.6 cm3 ion chamber as a function of distance from a point source to chamber center.
1.0 2.0 5.0 10.0 15.0 20.0
(T6lli,1997,[18].)
1.342 1.116 1.026 1.009 1.005 1.004
Figure 3.6 Two 'in-air' calibration jigs: (a) Nucletron device, and (b) 'Royal Marsden Hospital device. Both jigs are shown with Farmer chamber without build-up cap.
Since X and d are measurable quantities and d0 is an arbitrarily chosen distance, the values for X0 and Xs may be determined by fitting a line to data relating X to d. Ezzell [29] used this method and obtained data at 200, 300, 400, and 500 mm from the source. The average reading at each point was corrected for leakage and timer error and the reference distance was chosen to be 200 mm. The four data points fell on a straight line with a correlation coefficient of unity. The author concluded that, for the calibration jig used and at a reference distance of 200 mm, room scatter was 0.6% of the measured dose. A similar series of measurements was made by Goetsch et al [12] using an Exradin A3 spherical chamber of 3.6 cm3 with air-equivalent plastic walls (including cap) at nominal distances from 100 mm to
30 Calibration of sources
396.4 mm from an iridium-192 source. The room scatter at a source-chamber distance of 200 mm was found to be 0.63%. Table 3.4 summarizes both sets of measurements. For relatively large distances, and in the typical hospital laboratory, room scatter may introduce errors that are not negligible. As an extended time period will be required to collect sufficient charge for acceptable precision, it is important that leakage for the electrometer system is measured and, if necessary, an appropriate correction made. Iridium-192 has a very complex emission spectrum which includes approximately 24 lines occurring in the energy range 9-885 keV. The lowest energy emissions are attenuated by the source capsule and do not influence measurements. The exposure-weighted average of the remaining lines is 397 keV, which falls approximately halfway between the cesium-137 gamma ray energy of 662 keV and the average energy (146 keV) of a 250 kVP medium-filtration X-ray beam (half value thickness = 3.2 mm Cu). For a beam of this radiation quality, the choice of calibration factor and choice of build-up cap are not straightforward. There are three factors to consider. First of all, Goetsch et al. [12] have demonstrated that for in-air measurements at short distances, it is necessary to exclude high-energy photoelectrons emitted from the source capsule. This can be achieved either by placing a build-up cap over the ion chamber or by introducing the source itself into a narrow-bore Perspex tube where a wall thickness of 1 mm will be sufficient to remove the electron contamination. Secondly, Goetsch et al [12] have also suggested that iridium-192 measurements require that the ionization chamber's wall thickness must be sufficient to provide charged particle equilibrium for the highest energy secondary electrons present. These authors conclude, from measurements made with ion chambers that have graphite caps of various thicknesses, that the total thickness of wall and cap should be at least 0.3 g cm-2 to assure charged particle equilibrium: a Farmer-type chamber has a wall of 0.065 g cm-2. The third consideration concerns the instrument calibration factor. There are two components: an intercom-
5.0 10.0 15.0 20.0 30.0 40.0 50.0
1.00 0.999 0.997 0.994 0.988
—
-
0.999 0.997 0.994 0.986 0.975
0.966
-
' For local conditions as described by Ezzell (1989) [29], and Goetsch et a I. (1991) [12].
parison ratio between the field instrument and a secondary standard, and a calibration factor for the secondary standard appropriate to this energy. The former component may be determined using an iridium source at sufficiently large distance to ensure that the correction for finite chamber size will be the same for the two detectors. The second component is more difficult to establish. The reason for this is that there is an absence of primary standards for radiation from iridium-192 sources and the calibration factor of the chamber has to be obtained by other means. In the UK, no factor for calibration against the primary standard is normally available between that for heavily filtered 280 kVP X-rays (for use with measurements made without a build-up cap) and that for 2 MV X-rays (for use with measurements made with a build-up cap). In the USA and in some European laboratories, a factor for calibration against cesium-137 (gamma ray energy of 662 keV) is also available. The subject has been discussed comprehensively by Grimbergen and van Dijk [31] and Goetsch et al. [12]. The method described by Goetsch is a linear interpolation between a medium filtered 250 kV X-ray quality and cesium-13 7, with correction for differences in wall attenuation between X-rays, cesium-137, and iridium192 radiation. The correction factor was derived from wall attenuation measurements with one type of ionization chamber. This method is used extensively in the USA and forms the basis of calibrating iridium-192 sources that are subsequently used in the calibration of re-entrant ionization chambers. Grimbergen and van Dijk describe the air kerma calibration procedure that can be provided by the Netherlands Measurements Institute (NMI), which is based on weighting the chamber response according to the air kerma spectrum of the iridium-192 source [5]. This 'energy response curve' method is probably the most accurate, but is time consuming and expensive to realize. The authors calculated the photon spectrum of iridium-192 source-type used in the microSelectron HDR afterloading equipment using the EGS4 Monte Carlo System. The energy response was calculated for two widely used ion chambers: the NE 2561, which is used in the UK as a secondary standard, and the NE 2571, which is used as a field instrument. The calibration of the chambers was performed against the primary standards in beams of X-rays, cesium-137, and cobalt-60 gamma radiation; during the measurements the chambers were fitted with build-up caps. To determine the chamber response at the energy of each iridium-192 spectrum line, a polynomial was fitted through the chamber calibration data (see Figures 3.7 and 3.8). A comparison was then made for each chamber: a calibration factor was derived using the 'energy response curve' method and also one was derived using the average of the medium filtered 250 kV X-ray and cesium-137 calibration factors. It was found that for both chambers the two-point method results in a 0.3% lower value than that derived by matching the energy response curves. The
Calibration of sources used in remote afterloading systems 31
Image Not Available
Figure 3.7 Energy response (E) of the NE2561 (S/N 051) with build-up cap. The error bars indicate two standard deviations. The line is the polynomial fit used to determine the response at the iridium-192 spectrum energies. (Reproduced with permission from Grimbergen and van Dijk, 1995 [31].)
Image Not Available
and build-up caps. These data are also presented in IAEATECDOC1079 [7] together with a description of an in-air calibration using the method based on the technique by Goetsch et al, [12]. The principle of the method for calibrating an 192Ir HDR source, is to calibrate it at an appropriate X-ray quality and at 137Cs, or in 60Co if a 137Cs beam is not available. With the knowledge of the air kerma calibration factors at these two energies, the air kerma calibration factor for 192Ir is obtained by interpolation making use of the wall correction factors. This method requires the total wall thickness to be the same at each quality that the chamber is calibrated. In the UK, a Joint BIR and IPSM Working Party [28] set up to report on brachymerapy dosimetry recommended that the calibration for use with iridium should be that for the highest available kilovoltage quality, that is, the factor for heavily filtered 280 kV X-rays. This recommendation was made in the absence of a primary calibration for iridium-192. It was also suggested that the measurements should be made with an NE 2 MV buildup cap, or a Perspex sheath around the source to remove photoelectrons emitted from the source capsule. The recommended values for electron filter corrections are 1.017 for the NE 2 MV build-up cap, or 1.004 for a 1 mm thick Perspex sheath around the source; more generally, for small thicknesses of low atomic number filter material, the recommended correction is 3% g-1 cm-2. It is estimated that, whichever method is used to ascertain the chamber factor, the difference between the 'energy response curve' method, the two-point method, or the simpler 250 kV factor method is less than 1%. Typical exposure times for an 'in-air' calibration at 100 mm using a nominal 10 mGy s-1 source will be about 300 s. If the mean reading is R then the RAKR, Kr in mGy s-1 is derived as follows:
Fc Figure 3.8 Energy response (E) of the NE 2571 (S/N 1436) with build-up cap. The error bars indicate two standard deviations. The line is the polynomial fit used to determine the response at the iridium-192 spectrum energies. (Reproduced with permission from Grimbergen and van Dijk, 1995 [31].)
uncertainty in the calibration factor for iridium-192 determined with the energy response curve method was estimatedat l%fortheNE2561 and l.l%for the NE 2571. It is concluded by Grimbergen and van Dijk [ 31 ] that, with respect to the chambers considered, the two-point method is a reasonable alternative to the energy response method. More recently, Ferreira et al. [33] have performed Monte Carlo calculations of chamber wall correction factors for 51 commercially available ionization chambers
F.c Ftp
Fs
Fg
Fe Fis
is the air kerma calibration factor for the secondary standard; is the intercomparison ratio between the field instrument and the secondary standard; is the temperature/pressure correction factor (= TIP x 1013/293.2), where T is the temperature in Kelvin and P the pressure in millibar; is the correction for room scatter and scatter produced in the jig and its support (= 0.998 for the Royal Marsden jig and support system); is the dose gradient correction factor for the ion chamber (= 1.009 for a 0.6 cm3 Farmer-type chamber at 100 mm); is the electron filter correction (= 1.017 for an NE build-up cap); is the inverse square scaling from 100 mm to 1 m
(= 10-2);
Fm is the gray to microgray conversion (= 106); t is the time, in seconds, for each reading.
32 Calibration of sources
In principle, there should also be corrections for air attenuation and scattering and for the source transit time, but both of these are taken as unity. RE-ENTRANT IONIZATION CHAMBER MEASUREMENTS
The most convenient means of measuring the strength of an HDR iridium-192 source is to use a re-entrant ionization chamber. The Standard Imaging HDR 1000 Ion Chamber is designed specifically for high strength sources [14]. The well of the chamber consists of a 35 mm diameter x 122 mm deep cavity into which a holder can be inserted to carry a catheter or rigid applicator for positioning the radioactive source. The responsivity of the chamber is affected by source position, but the variation is only ± 0.5% over a range of 25 mm near to the center of the chamber axis. Jones [34] used a modified Standard Imaging chamber with an NE 2571/1 electrometer as part of a qualityassurance programme for checking iridium-192 source strengths. The chamber measurements were found to correlate well with 'in-air' calibration measurements made over a 3-month source-decay period; the maximum discrepancy between the 'in-air' measurement and the re-entrant chamber measurement was 0.9% (Figure 3.9). A further investigation with a Nucletron Source Dosimetry System (PTW-Freiburg re-entrant ion chamber) calibrated at NIST and a modified Standard Imaging chamber showed that, over a 4-month period, 13 comparative measurements made with both chambers were found to be within 0.18% (1 SD). The air kerma strength measured using the NIST traceable Nucletron PTW chamber was found to agree within 0.3% of the 'in-air' calibration measurement. Goetsch et al. [29] also report measurements using three re-entrant ionization chambers manufactured by
Figure 3.9 The relative source strength of an iridium-192 source as a function of time measured with a modified standard imaging HDR WOO chamber over a 3-month period. The line corresponds to the decay curve; the points correspond to chamber measurements [34].
three different manufacturers. Agreement was found to be within 1%. In order to obtain maximum reproducibility, measurements in a re-entrant chamber should be made with the aid of a Perspex insert. For definitive calibration measurements, the chamber should be located away from objects that might cause radiation scatter (such as a laboratory wall). Consistent readings are obtained best when the chamber is used each time in the same position and in a reproducible manner. SOLID PHANTOM (OR WATER PHANTOM) MEASUREMENTS
The underlying principle of this type of measurement has been described in section 3.4.3, except that in the case of a single HDR source, the dose gradient through the centrally placed ion chamber has to be corrected for [18]. The method has been described by Ezzell [29], and Krieger [35]. Measurements in a phantom require corrections to be made for attenuation and scatter. Usually, the effective distance between the source and the attenuation is determined using a formula that applies to the condition of full scatter. Generally, the small dimensions of the solid (acrylic) phantom do not fulfil this requirement and a correction factor has to be applied to compensate for this lack of scatter. Ezzell [29] measured this to be 1.079 for a 130 mm diameter x 130 mm long acrylic cylinder. 3.4.6.
Pulse dose-rate sources
Pulse dose-rate (PDR) sources are of lower activity than HDR sources and are of different physical construction. PDR sources are typically 18-37 GBq in activity. In the case of iridium-192, the active length of a 37 GBq PDR source is 0.5 mm. The most convenient method of calibrating such a source is to use a re-entrant ion chamber. It will be necessary to ensure that the chamber has been calibrated for PDR sources rather than, or as well as, HDR sources. The strength of the source is such that an 'in-air' calibration with a 0.6 cm3 Farmer ion chamber is achieved best at a distance of about 50 mm; greater distances would require extended exposures. At this short distance, it is important to ensure precise alignment between the source and the chamber center. The location of the source center should be ascertained (by autoradiography) so that alignment can be achieved reproducibly. It is necessary to use a build-up cap (or a Perspex sheath over the source), as described in the section on HDR source calibration, to remove electrons emitted from the source capsule. A correction for the finite size of the chamber must be applied: this increases significantly at short distances. Tolli [18] has shown that, for a chamber of the same dimensions as a 0.6 cm3 Farmer chamber, the correction factor is 1.026. Exposure times to accumulate sufficient charge will be
References 33
long (typically 0.5 h) and an allowance must be made for any chamber leakage that occurs during the measurement period. REFERENCES
1. International Commission on Radiation Units and Measurements (1985) Dose and Volume Specification for Reporting Intracavitary Therapy in Gynaecology, Report 38. Bethesda, Maryland, ICRU. 2. Kutcher, G.J., Cola, L, Gillin, M. etal. (1994) Comprehensive QA for radiation oncology: Report of AAPM Radiation Therapy Committee Task Group 40. Med. P/7ys.,21(4),581-618. 3. Rossiter, M.J. (1990) The traceability of brachytherapy sources supplied byAmersham International. Br.J. Radiol., 63,663. 4. Read, L.R., Burns, J.E. and Liquorish, R.A.C. (1978)
calibration of iridium-192 high dose rate sources. Int. J. Radial Onc. Bio/. Phys., 24,167-70. 15. Baltas, D., Geramani, K., lonaddis, G.T. et al. (1999) Comparison of calibration procedures for 192lr high dose rate brachytherapy sources. Int.J. Radial Oncol. Bio/. Phys., 1,43(3), 653-61. 16. Williamson, J.F., Morin, R.L and Khan, F.M. (1983) Dose calibrator response to brachytherapy sources: a Monte Carlo and analytic evaluation. Med. Phys., 10(2), 135-40. 17. AAPM Report No. 13 (1984) Brachytherapy. In Physical Aspects of Quality Assurance in Radiation Therapy. New York, American Institute of Physics, 38-9. 18. Tb'lli, H. (1997) Ionisation chamber dosimetry for brachytherapy. Doctoral dissertation. University of Goleborg, 15-31. 19. Schaeken, B., Vanneste, F., Bouiller, A. et al. (1992) 192Tr brachytherapy sources in Belgian hospitals. Nucl. Insl Methods Phys. Res., A312,251 -6. 20. Steggerda, M.J. and Mijnheer B. (1994) Replacement
Exposure-rate calibration of small radioactive sources of
corrections of a Farmer-type ionisation chamber for the
cobalt-60, radium-226and caesium-137. IntJ.Appl. Radial /sof.,29,21-7.
calibration of Cs-137and lr-192 sources in a solid
5. Rossiter, M.J., Williams, T.T. and Bass, G.A. (1991) Air kerma rate calibrations of small sources of cobalt-60, caesium-137, radium-226and iridium-192. Phys. Med. fi/o/., 36(2), 279-84. 6. Williamson, J.F. and Nath, R. (1991) Clinical implementation of AAPM Task Group 32
phantom. Radiother. Oncol., 31,76-84. 21. Aukett, R.J. (1991) A technique for the local measurement of air kerma rate from small caesium-137 sources, fir. J. Radiol., 64,918-22. 22. Meertens, H. (1990) In-phantom calibration of SelectronLDR sources. Radiother. Oncol., 17,369-78. 23. Mijnheer, B.J., Aalbers, A.H.L, Visser, A.G. and
recommendations on brachytherapy source strength
Wittkamper, F.W. (1986) Consistency and simplicity in the
specification. Med. Phys., 18(3), 439-48.
determination of absorbed dose to water in high-energy
7. IAEA-TECDOC-1079 (1999) Calibration of Brachytherapy Sources. Vienna, International Atomic Energy Agency. 8. Mcjury, M., Tapper, P.D., Cosgrove, V.P. et al. (1999) Experimental 3-D dosimetry around a high dose rate 192lr source using a polyacrylamide gel (PAG) dosimeter. Phys. Med. fi/o/., 44(10), 2431-44. 9. Berkley, L.W., Hanson, W.F., and Shalck, R.J. (1981)
photon beams: a new code of practice. Radiother. Oncol., 7,371-84. 24. Meisberger, LL, Keller, R.J. and Shalek, R.J. (1968) The effective attenuation in water of the gamma rays of gold 198, iridium 192, cesium 137, radium 226, and cobalt 60. Radiology, 90,953-7. 25. Jones, C.H. (1991) Quality assurance in brachytherapy
Discussion of the characteristics and results of
using the Selectron-LDR/MDR and microSelectron-HDR.
measurements with a portable well-ionisation chamber
Activity, 5(4), 12-15. Leersum, The Netherlands,
for calibration of brachytherapy sources. In Recent Advances in Brachytherapy Physics, ed. D.R. Shearer, Monograph No. 7. New York, AAPM, 38-48. 10. Weaver, K.A., Anderson, LL and MeliJ.A. (1990) Source calibration. In Interstitial Brachytherapy: a Report by the Interstitial Collaborative Working Group, ed. L.L. Anderson, New York, Raven Press. 11. Flynn, A. and Workman, G. (1991) Calibration of a microSelectron HDR iridium-192 source. Br.J. Radiol., 64, 734-9. 12. Goetsch, S.J., Attix, F.H., Pearson, D.W. and Thomadsen, B.R. (1991) Calibration of Ir192 high dose rate afterloading systems. Med. Phys., 18,462-7. 13. Jones, C.H. (1988) Quality assurance in gynaecological brachytherapy. In Dosimetry in Radiotherapy Vol. 1, Vienna, IAEA, 275-90. 14. Goetsch, S.J., Attix, F.H., Dewerd, LA. and Thomadsen, B.R. (1992) A new re-entrant ionisation chamber for the
Nucletron BV. 26. Chenery, S.G.A., Pla, M. and Podgorsak, E.B. (1985) Physical characteristics of the Selectron high dose rate intracavitaryafterloader. Br.J. Radiol., 58,735-40. 27. Messina, C.F., Ezzell, G.A., Campbell, J.M. and Orton, C.G. (1988) Commissioning the Selectron HDR cobalt-60. Activity, 2, 5-10. Leersum, The Netherlands, Nucletron. 28. Aird,E.GA, Jones, C.H.,Joslin,C.A.F.efo/. (1993) Recommendations for brachytherapy dosimetry: Report of a Joint Working Party of the BIR and IPSM. London, BIR, 1-17. 29. Ezzell, G. (1989) Evaluation of calibration techniques for the microSelectron HDR. In Brachytherapy2, ed. R.F. Mould. Leersum, The Netherlands, Nucletron, 61 -9. 30. Piermattei, A., Azario, L, Soriani,A. et al. (1995) Reference air kerma rate determination for iridium-192 brachytherapy sources. Phys. Med. 9-15. 31. Grimbergen, T.W.M. and van Dijk, E. (1995) Comparison of
34 Calibration of sources methods for derivation of iridium-192 calibration factors for NE2561 and NE2571 ionisation chambers. Activity, Special Report No. 7, 52-6. Leersum, The Netherlands, Nucletron BV. 32. Biiermann, L, Kramer, H.M. and Selbach, H.J. (1995) Reference Air Kerma Rate Determination of an Iridium-192 Brachytherapy Source. Activity: Special Report No. 7. Nucletron-Oldelft, Leersum, The Netherlands, 43-47. 33. Ferreira, I.H., Marre, D., Bridier, A., Marechal, M.H. and de Almeida, C.E. (1999) Personal communication.
34. Jones, C.H. (1995) HDR microSelectron quality-assurance studies using a well-type ion chamber. Phys. Med. Biol., 40,95-101. 35. Krieger, H. (1991) Messungder Kenndosisleistung punktund linien-formiger HDR iridium-192 Afterloadingstrahlermiteinem PMMAZylinderphantom.Ze/f. Med. Physik., 1,38-41.
4 Systems of dosimetry ANNE WELSH AND KAREN D'AMICO
4.1
INTRODUCTION
Brachytherapists have always aimed to deliver the optimum treatment to the patients in their care. In order to achieve this ideal, the introduction of brachytherapy was soon followed by the development of dosimetry systems which were based on clinical experience. A paper on the physics of brachytherapy was published in 1923 by Coliez [ 1 ] and this was followed by many other publications. The well-known Manchester System [2] was published in 1934, followed by the Quimby System in 1944 and 1953 [3,4]. The early systems were designed for radium. Subsequently, the Paris System [5,6] was published as a modern dosimetry system which allows the brachytherapist to take full advantage of the technological improvement in source production, in particular the use of iridium-192 flexible wire sources. The aim of all the published dosimetry systems is simple: to provide a set of guidelines for the brachytherapist which, if followed, enable a prescribed dose to be delivered to the patient as accurately as possible. This chapter contains brief descriptions of the systems most commonly used and some sample data to enable a limited number of practical calculations to be performed. It is recommended that the original works describing the dosimetry systems are consulted before any of the systems described are set up for clinical use for the first time.
4.2 NON-GYNECOLOGICAL TREATMENT SYSTEMS 4.2.1
The Paris System
BASIC PRINCIPLES
The first account of this system was published in 1966 [5] and was followed by several more detailed publications, e.g., in 1978 and 1987 [6,7]. The Paris System is designed for modern sources, particularly iridium-192 wires, which have narrow diameter, are flexible, and may be formed to any length required. The system requires the brachytherapist to distribute sources over a predetermined target volume in accordance with the system rules. Dose rates are then calculated at defined points, known as basal dose-rate points, within the target volume. The isodose that closely envelopes the target volume may be found by calculating 85% of the average basal dose rate. The basic rules for positioning the sources are as follows: • Each source must be implanted parallel to the others. • Each source must be equidistant from adjacent sources. • The reference air kerma per unit length of source is constant for all the sources used in the implant. • Ideally, the centers of all the sources are contained in a single plane which perpendicularly bisects the
36 Systems of dosimetry
sources. This plane is called the central plane of the implant. If such a plane cannot be defined, the central plane is that plane perpendicular to the sources which passes as closely as possible to the source centers. The perpendicular distance between two adjacent sources is referred to as the source separation in this chapter. The source separation must lie in the range 8-15 mm for short wires (i.e., 50 mm or less) and 8-22 mm for long wires (greater than 70 mm). If the source separation exceeds the maximum permitted, there will be an unacceptably large high dose area around each source. This is illustrated in Figure 4.1. If the wire separations are less than 8 mm, it is difficult to implant the wires in a parallel fashion and the implant may not conform to the requirements of the Paris System.
Figure 4.1 Two implants each showing the area receiving the prescribed dose and twice the prescribed dose. The diameter of the high dose area is considerably larger for the implant with
POSITIONING THE SOURCES
the greater wire separation.
The first planning consideration is the size of the target volume. The target volume has three dimensions, which are usually referred to, in the Paris System, as the length, the width, and the thickness (Figure 4.2). The sources are positioned parallel to the length dimension and may be implanted in one or more planes, depending on the thickness of the treatment volume. Practical planning commences by determining the number of source planes and the source separation necessary for the satisfactory treatment of the patient. • For a single plane containing only two sources, the source separation is given approximately by: wire separation = thickness x 2.0
Figure 4.2 The relationship between the wire positions, the
• For a single plane implant containing three or more sources, the source separation is given approximately by:
safety margins, and the target volume dimensions.
wire separation = thickness x 1.7
The number of sources needed to implant a target volume can then be calculated from: width = wire separation x (number of wires - 1) + safety margin x 2
The safety margins for the different wire arrangements are summarized in Table 4.1. Some brachytherapists prefer to omit the safety margin from their planning calculations and implant the sources up to the
edge of their planned volume. The safety margin then fulfils the purpose of its name and provides a small additional margin around the volume. If the calculated wire separation exceeds the maximum permitted value, the implant cannot be carried out as a single plane of sources and two or more planes must be implanted. In multi-plane implants the sources may be implanted on a square lattice ('in squares') or an equilateral triangular lattice ('in triangles') (Figure 4.3).
Table 4.1 Safety margins as fractions of wire separations for implants planned in accordance with the Paris System rules
0.37
0.27
0.15
Non-gynecological treatment systems 37
source length = 1.4 x target volume length All the formulae given above are approximate, but are sufficiently accurate for clinical use. Detailed tables may be found in reference [7].
CALCULATION OF THE DOSE
Figure 4.3 Cross-sections through two double-plane implants, one implanted 'in triangles' and one implanted 'in squares'. The dots indicate the wire positions and the crosses indicate the basal dose points.
• For an implant in squares, the wire separation can be calculated from: thickness = wire separation + safety margin x 2 • For an implant in triangles, the wire separation can be calculated from: thickness = wire separation x 0.87 + safety margin x 2 • The width of an implant in squares is given by: width = number of squares x source separation + safety margin x 2 • The width of an implant in triangles is given by:
The isodose that encloses the target volume is known as the reference isodose. It is calculated from the basal doserates, which are the lowest dose rates found within the central plane. For a single-plane implant, the basal doserate points are positioned at the midpoint of each gap between the sources. The basal dose-rate points for an implant in squares are at the center of the squares, and for implants in triangles the basal dose rates are calculated at the points where the perpendicular bisectors of the sides of the triangles intersect. The dose rates can be calculated by entering the source position data into a suitable computer program, which is likely to be available in most oncology centers, or by hand calculation using graphs or tables of dose rate against distance from the source for unit activity. The dose rates taken from graphs or tables must be corrected to values for the actual mid-implant source strength. This will involve obtaining the source strength from the supplier's measurement certificate and correcting it for radioactive decay to the date of the middle of the implant. Sample data are given in Tables 4.2a and 4.2b. A simple example of an application of the Paris System is shown below.
EXAMPLE
width = number of triangles x source separation x 0.5 + lateral safety margin x 2 The length of the sources required to treat a target volume is almost independent of the other parameters and is given approximately by the expression:
A target volume 55 mm long, 8 mm thick, and 35 mm wide is to be given a dose of 30 Gy. Calculate the source lengths required and the dose rate assuming an air kerma rate source strength of 800 nGy h-1 mm-1 at 1 m halfway through the implant.
Table 4.2a Sample dose rates for iridium-192, 0.3 mm diameter wire sources, calculated according to the recommendations of the Joint BIR/IPSM Brachytherapy Working Party [8]
20 10 20 30 40 50 60 80 100 120
350 489 547 577 595 606 619 627 631
104 176 219 246 262 275 290 299 305
48.0 88.3 118 138 153 164 179 188 194
27.7 52.4 72.6 88.4 101 110 124 132 138
100
12.5 24.3 35.0 44.4 52.3 59.1 69.5 76.9 82.3
7.1 13.9 20.3 26.3 31.6 36.3 44.2 50.2 54.9
4.5 8.9 13.2 17.2 20.9 24.3 30.3 35.1 39.0
3.1 6.2 9.2 12.0 14.7 17.3 21.8 25.7 28.9
1.7 3.4 5.1 6.7 8.2 9.3 12.6 15.1 17.3
1.0 2.0 3.1 4.1 5.1 6.1 7.9 9.6 11.1
Dose rates a re in mGyh-1 for an air kerma rate source strength of 1 mGy h-1 mm-1 at 1 m. All the dose rates are for crossline = 0, i.e., in the plane through the center of the wire and perpendicular to it.
38 Systems of dosimetry
Table 4.2b Sample dose rates for iridium-192, 0.3 mm diameter wire sources, calculated according to the recommendations of the Joint BIR/IPSM Brachytherapy Working Party [8]
0 10 20 30 0 30 50
50 50 50 50 100 100 100
595 581 487 139 627 606 319
262 251 200 98.9 299 278 157
153 145 116 71.4 188 170 103
101 95.2 78.0 53.9 132 117 74.9
52.3 49.9 43.0 22.7 76.9 66.9 46.8
31.6 30.4 27.2 22.8 50.2 43.8 32.7
20.9 20.3 18.7 12.1 35.1 30.9 24.2
14.7 14.4 13.5 9.3 25.7 22.9 18.6
8.2 8.1 7.8 7.3 15.1 13.8 11.8
5.1 5.1 4.9 4.7 9.6 8.9 7.9
Dose-rates are in mGy h-1 for source strengths of 1 u.Gy h-1 mm-1 at 1 m. The dose rates are for different crosslines, i.e., in different planes perpendicular to the wire and at the crossline distance from the wire center.
Single plane implant. Source separation = 1.7 x thickness = 1.7x8 = 14 mm Width =2 x 0.37 x 14 +14 x (number of wires -1) = 2 x .37 x 14 + 14 x 2 = 38 mm This is the closest width to the required value and requires three wires to be used. Wire length = 1.4 x 55 = 77 mm - use 80 mm. The details of the distances and dose rate calculations for this calculation are given in Table 4.3. The mean basal dose rate for a source strength of 1 1-iGy h-1 mm-1 at 1 m, from Table 4.3, is
0.5 x (943 + 943) = 943 mGy h-1 Correcting this to true source strength available for use:
Reference dose rate
_ 85
~100 X
Treatment time may be calculated from this dose rate and the prescribed dose.
The isodose distribution for this implant is shown in Figure 4.4. PROBLEMS It is unlikely that an implant that conforms perfectly with the Paris rules can be achieved without the use of a template to aid the source positioning. If the implant does not comply precisely with the rules, the Paris System may still be used providing none of the individual basal dose rates differs from the mean basal dose rate by more than 10% and if, when using the triangular formation of sources, none of the triangles is obtuse (Figure 4.5). Further information on dealing with variants in source position which may be needed in practice may be found in the book Modern Brachytherapy [7].
average basal dose rate
Table 4.3 Dose rates calculation for sample iridium-192 imnlnnt
1 2 3
7 7 21
412 412 119 943
Dose rates in mGy h-1 calculated for wire of air kerma rate source strength 1 mGy h-1 mm-1 at 1 m.
Figure 4.4 The dose distribution for the example of the Paris implant showing isodoses of 60 Gy, 30 Gy, 22.5 Gy, and 15 Gy.
Non-gynecological treatment systems 39
4.2.2
The Manchester System
The first widely available set of rules that could be used to deliver a uniform dose to a target volume is often referred to as the Manchester System, after its place of origin and is fully described in the book Radium Dosage, The Manchester System [9]. Excluding gynecological treatments, three different types of brachytherapy implant were considered: moulds, interstitial planar implants, and volume implants. The systems used to distribute sources and calculate treatment times need to be discussed separately for each of these cases. MOULDS
The term mould is used to describe the situation in which the radioactive sources are positioned external to the patient, usually at a distance from the patient's skin known as the treatment distance and represented by the letter d (Figure 4.6). The treatment dose is prescribed to the plane which is at distance d from the sources and the dose in this plane will be delivered with a 10% accuracy if the rules are followed. The sources may be arranged over the treatment area in circles or squares, circles being the preferred arrangement. In both cases the target area should be bounded with radioactive sources arranged in the required shape. Ideally, the sources should be laid in a continuous line around the periphery, but gaps not exceeding d are acceptable. For small circular treatment areas, the sources around the periphery will treat the target area adequately, but large circles may require additional sources in the interior. A circle may be deemed to be small if the ratio of the
Figure 4.6 Source arrangement for a Manchester-type mould arrangement. Source positions are indicated by dots on the cross-section and lines on the wire positions viewed from above.
circle diameter to the treatment distance (d) is less than 3. The distributions required for larger circles are given in Table 4.4. Squares will be adequately treated by the peripheral sources only if the side of the square does not exceed twice the distance d. If the square does not meet this requirement, supplementary lines of sources, parallel to the side of the square, will be needed in the interior. The distance between the lines of sources must not exceed twice the distance d. If only one additional line of
Table 4.4 Distribution of total source activity for circular moulds Ratio of Diameter to Treatment Distance
<3
3-6
Percentage of total source activity on outer circle
100
95
Percentage of total source activity on a circle of diameter half that of the outer circle
-
Central spot
6
7.5
80
75
70
-
17
22
27
-
5
3
3
3
100
95
90
80
-
5
10
20
For small values of d Percentage of total source activity on outer circle Central spot
10
40 Systems of dosimetry
sources is required, the linear source strength of the interior sources should be half that of the periphery. If there are two or more additional lines, the linear strength of the interior sources should be two-thirds of the peripheral value. If the target area is rectangular rather than square, treatment times calculated for sources should be increased. Details of this increase are given in Table 4.5. Once the sources have been arranged and the total source strengths used are known, the treatment time can be calculated by looking up, in Table 4.6, the total product of total source strength and treatment time for the area in use and the treatment distance. Table 4.6 gives values necessary to deliver a dose of 1 Gy to the treatment plane, and the value looked up should be adjusted proportionately for the dose actually prescribed. The treatment time may then be calculated as follows: Treatment time for n Gy = n x
table value for total source strength x treatment time total source strength used
The dosage table is based on the original Manchester data, updated to allow for current values of the calculation parameters [10,11] and converted into SI units. EXAMPLE
A circular area is to be treated to a dose of 30 Gy with source of strength 1200 nGy h-1 mm-1 at 1 m. The treatment distance is 10 mm.
Table 4.5 Percentage increase to square implant treatment times for rectangular implants Ratio of side to base for rectangle Percentage to increase treatment time
2:1 5
3:1 9
4:1 12
Treat with a single circle 25 mm in diameter (from Table 4.4) Area = 5 cm2 Total source strength x treatment time is 237 microgray-hours at 1 m from Table 4.6. Total source strength is 25 x p x 1200 nGy h-1 at 1 m = 94.3 (mGy h-1 at 1 m.
INTERSTITIAL PLANAR IMPLANTS
The interstitial rules are based on those for moulds, but simplified to allow for the technical difficulties of implanting the sources into the patient's tissue. The implant can be effected most easily using the square or rectangular arrangement described for moulds. Parallel lines of sources are implanted through the center of the volume and across the ends of the target volume. The simplified distribution rules are given in Table 4.7. The treatment time for a treatment volume 10 mm thick may be obtained from Table 4.6 using a treatment distance of 5 mm. The target volume extends for 5 mm either side of the plane containing the wires and is thus 10 mm thick. If it is desired to sandwich the target volume between two parallel planes, the sources may be arranged in the same manner as for a single plane and the product of total source activity and treatment is again obtained for a treatment distance of 5 mm. The treatment time is then increased by the factors given in Table 4.8, which are dependent upon the separation of the two planes.
Table 4.7 Distribution of total source activity between periphery and interior for planar implants
Table 4.6 Dose table for area planned according to the Manchester System
interior
0-25 0 5 8 10 15 20 30 40 50 60 80 100
23 124 159 182 234 285 379 466 545 619 758 893
92 237 297 335 423 496 615 722 829 932 1473 1327
207 385 463 506 610 704 883 1041 1177 1301 1520 1730
368 572 661 714 838 947 1150 1339 1514 1686 1981 2235
Total air kerma source strength is expressed in mGy h-1 at 1 m and treatment time is in hours. The table gives the product of total source strength used and treatment time to deliver a dose of one Gy.
25-100 >100
67 50 33
33 50 67
Table 4.8 Factor by which treatment times for two-plane implants calculated for 5 mm treatment distance (e.g. from Table 4.6, column 2) must be increased for plane separations greater than 10mm
10 15 20 25
1.0 1.25 1.40 1.50
Non-gynecological treatment systems 41
The dose will be low in the plane halfway between the source planes for separations which exceed 20 mm. It is often impractical in an afterloaded implant to have two rows of sources perpendicular to the principal implanted direction. If it is not possible to cross the ends of the implant, the implanted area may be considered to be decreased by 10% for each uncrossed end. The lines of implanted sources then extend 10% of their length beyond the ends of the target volume for each uncrossed end. VOLUME IMPLANTS
Volumes may be considered to be spherical, cylindrical, or cuboid. Sources must be distributed throughout the volume in accordance with the rules given in Table 4.9. In general, volumes are considered to consist of an outer shell and an inner core. Within the shells and cores the source activity should be distributed as evenly as possible in compliance with the distribution rules, and gaps between sources should ideally be less than 15 mm. The treatment times can then be determined from the volume dosage in Table 4.10 in a manner analogous to that for the moulds and planar implants. Again, as for planar implants, it may well be impossible to afterload sources which are positioned perpendicularly to the main direction of implantation, and cylindrical volumes should be reduced by 7.5% for each uncrossed end.
Table 4.9 Distribution of source activity for volume implants planned in accordance with the Manchester rules
Sphere Cylinder Cuboid
Shell: 6 parts Core: 2 parts Belt: 4 parts Core: 2 parts Each side: 1 part Core: 2 parts
Each end: 1 part Each end: 1 part
Gaps between the needles should not exceed 10-15 mm. Table 4.10 Dose table for volumes planned according to the Quimby and Manchester systems
5 10 20 40 80 100 125 150
78 123 195 310 491 570 662 747
152 244 335 472 663 762 853 952
Total air kerma source strength is expressed in mGy h-1 at 1 m and treatment time is in hours. The table gives the product of total source strength used and treatment time to deliver a dose of 1 Gy for sources with a filtration equivalent to 0.5 mm of platinum.
4.2.3
The Quimby System
This system was developed by Edith Quimby from work she originally carried out in the 1930s [12]. The system is essentially an adaptation of the Manchester System and is described in her later works [3,4]. The Manchester System made severe demands upon the radium stocks available in the USA. In general, the activities were higher than those which could be obtained in the UK and the range of available activities was smaller. This made it difficult for oncologists to fulfil the requirements of the Manchester System for sources in the interior of the implants to have a lower linear source strength than those in the periphery. It was also difficult to achieve the required spacing of the sources within the implant without substantially changing the proportions of activity that the Manchester System assigned to the periphery and interior of the implant. The Quimby System is based on relaxing the Manchester System distribution rules so that uniform linear source strength sources may be used and the activity is distributed uniformly throughout the treatment volume. The homogeneity of dose within the treated volume is not as good as that achieved by the Manchester System, but the inhomogeneity was considered to be acceptable in clinical practice. Among the possibilities that Quimby considered were: 1. Using uniform-strength needles throughout the target volume, and achieving the desired proportions of activity on the periphery and interior of the implant by increasing the source separations within the implant beyond that allowed by the Manchester System. 2. Using uniform-strength needles throughout and maintaining the source separations recommended by the Manchester System and thus altering the proportions of activity allocated to the interior and periphery of the implant. Planar implant tables based on the Quimby system, with the sources spread uniformly across the target, have been calculated as for the Manchester System, and a sample set of data (Table 4.11) is included here, based on the work of Godden [13] and converted to SI units, in accordance with the recommendations of the BIR/IPSM Working Party report. The tables are for the number of mGy total source strength multiplied by treatment time in hours required to deliver a dose of 1 Gy to the target area. The source strength is specified in mGy h-1 at 1 m. These tables are analogous to the Manchester System tables, which were originally published in the units of milligramhours of radium required to deliver a dose of 1000 R, but have been corrected to the currently recommended SI units and current values of calculation parameters. The Quimby System deals with uncrossed ends in the same manner as the Manchester System, reducing the treated area by 10% for each end of the target volume not
42 Systems of dosimetry
Table 4.11 Sample dosage table for implanted areas planned according to the Quimby System Square applicators
10 20 50
35 61 190
107 143 309
224 263 457
381 434 652
Rectangular applicators
10x15
39
112
227
385
40x60 60x90
218 434
324 553
475 720
666 944
34 79 228
104 167 345
215 305 519
378 462 699
Circular applicators
10 30 60
Total air kerma source strength is expressed in mGy h-1 at 1 m and treatment time is in hours. The table gives the product of total source strength used and treatment time to deliver a dose of 1 Gy for sources with a filtration equivalent to 0.5 mm of platinum.
bounded by sources. The sources should thus extend well beyond the boundary of the target volume if it is not possible to place sources across the ends of the target volume. When considering volume implants, Quimby examined the effects of distributing the source activity uniformly throughout the volume. She published tables of total source strength multiplied by treatment time to deliver a treatment dose to the periphery of the target volume for volume implants, assuming the activity is uniformly distributed. Some data are reproduced in Table 4.10, corrected for modern units. The quantity total source strength multiplied by treatment time differs substantially from the Manchester System because these tables deliver a minimum treatment dose rather than an average dose to a plane.
4.2.4 Modern seed implantation, average dimension method BASIC PRINCIPLES
This method is usually employed to deliver a required dose to a small-volume permanent implant afterloaded with iodine-125 seeds. A substantial part of the early work was done at the Memorial Sloan-Kettering Cancer Center, New York, and was based on clinical experience with prostatic cancer [14].
Initially, the total implanted activity in millicuries was simply taken to be five times the average dimension in centimeters, the average dimension simply being the mean of the length, width, and thickness of the target volume. The factor 5 was based on previous clinical experience. It was expected that this system would result in a patient dose that varied inversely as the sixth root of the volume. Clinically, it is well known that small volumes can tolerate higher doses than large volumes, so this slow reduction of treatment dose was thought to correspond well to clinical need. In practice, it was found that the dose varied more rapidly than the sixth root of the volume and this was attributed to absorption of the lowenergy iodine-125 emissions by tissue. Tissue scatter and absorption are not usually clinically significant for the radionuclides used in the classical dose systems. The average dimension method was thus modified to increase the amount of activity used in larger volume implants. The factor 5 was replaced as follows: factor = 5
average dimension < 2.4 cm
factor = 3.87 (average
2.4 < average dimension < 3.24
dimension + 1.0)1293 factor = 2.76 (average
average dimension > 3.23
dimension + 1.0)1581 PRACTICAL USE
To simplify the calculations of activity required for an implant, a nomogram was produced which allowed the brachytherapist to calculate the necessary number of seeds and seed strength for a particular volume. The brachytherapist measures the dimensions of the volume to be implanted in the coronal, sagittal, and transverse planes and calculates the average dimension. A nomogram is then used to determine the number of seeds which should be implanted for the seed activity available. The activity is then spread uniformly throughout the volume in a three-dimensional matrix. Practical considerations usually constrain the separation of the columns of seeds to 'round' number separations, e.g., 10 mm, 15 mm, etc., and the nomogram will indicate the most appropriate seed spacing within the column. Modern practice, although based on the above general principles, dispenses with nomograms for prostatic cancer and is computer based. The seeds are positioned in the prostate under ultrasound guidance using a template. The use of the template restricts the seed positions; it usually has holes for placing columns of seeds every 5 mm. Typical implant spacing is 10 mm. The seeds must be positioned out to the edges of the treatment for a satisfactory implant. The expected seed position data are entered into a treatment planning computer and source activity is calculated to deliver the prescribed dose to the isodose which enclosed the target volume. Many computer programs used the data of Krishnaswamy [15] for
Non-gynecological treatment systems 43
this purpose. Recommendations for calculating the dose distribution around iodine-125 seeds have been published by the American Association of Physicists in Medicine (AAPM) brachytherapy working group [16]. These are the best current sources of data. The matched peripheral dose for implants, that is, the dose that encloses a volume of the dimensions to be treated, is normally in the range 100 to more than 200 Gy. The dose within the target volume will be considerably higher than this. Further discussion of the dose distribution and optimization techniques are discussed by Yang et al. [17].
4.2.5
Other dosimetric methods
THE 'NO SYSTEM' METHOD
It is possible to implant a patient without following any rigid dosimetric system such as those described previously. The brachytherapist may position the radioactive sources to cover the target volume as he or she thinks fit. The source catheter positions may then be calculated using orthogonal radiographs, isocentric films, or any other appropriate method available. The dose distribution is then calculated either by hand, using published data for the sources in use, or, more usually, with the aid of a computer program. Suitable programs are available in most radiotherapy departments. The brachytherapist then prescribes the treatment dose to the isodose that most appropriately covers the target volume. There are considerable dangers in this system, which may cause the unwary practitioner to give a considerable overdose to parts of the treatment volume. There is a very steep dose gradient around each wire, principally because of the operation of the inverse square law from every element of every source. It is essential that the sources are not so widely separated that, if the dose is prescribed to an isodose that envelops all the wires, the inevitable high dose region around each source is large enough to cause necrosis. The guidelines for maximum source separations used in the Paris System may be taken as a useful indicator. It is also essential that the brachytherapist makes a carefully considered choice of prescription isodose. An isodose which is at a large distance from the sources may appear to be a good choice in that it covers the required treatment volume, but if the dose is prescribed at a large distance from the sources, necrosis may occur in the vicinity of the sources. Again, the Paris System guidelines may be used to determine the maximum treatment thickness permissible with any particular source separation. The dose delivered by brachytherapy sources falls off toward the physical end of the source and for that reason the sources must always extend beyond the end of the target volume or, alternatively, the quantity of source activity at the boundary of the target volume must be increased by 'crossing the ends' as per the Manchester System. Once again, even if the brachytherapist is not
adhering strictly to a system, the Manchester or Paris rules on source length will provide satisfactory guidelines. It is worth noting that there is a small amount of disagreement between the Paris and Manchester systems about the distance beyond the target volume which the sources should extend. The Paris System recommends that the source length is 40% greater than the target volume length, whereas the Manchester System recommends the source lengths are increased an additional 10% of the volume length for each uncrossed end, i.e., 20% of the target volume length if the sources are implanted in one direction only. Consideration of the above leads to the conclusion that it is impossible to carry out any sort of brachytherapy work without following some basic guidelines or rules for source distribution and choice of prescription isodose, even if these are not quite so formally set out as for the systems described above. It is appropriate to emphasize at this point that systems should not be mixed. For example, one should never distribute the source activity in accordance with the Paris or Quimby systems and then use the Manchester System tables to determine the treatment time. If one chooses to use a particular system, then the rules for that particular system should be scrupulously followed. If the brachytherapist chooses not to use a system, then particular attention must be paid to the choice of source separation, prescription isodose, and length of treatment sources with respect to the treatment volume. DOSE-VOLUME HISTOGRAMS
Dose-volume histograms are, strictly speaking, not a dosimetric system, but a method of examining and analyzing the implant. The analysis may be used to change the treatment prescription and for that reason dosevolume histograms are discussed in this chapter. The simplest form of dose-volume histogram is obtained by calculating the dose at every point on a matrix which covers the dose volume and sorting the points into dose bins. Each point represents the dose in a volume element of the matrix and thus counting the points which fall into each dose band allows the volume receiving a minimum of any particular dose to be calculated. This will allow the volume receiving a high dose, for example twice the prescribed dose, to be calculated and may give some information about possible dangers of necrosis. There is little useful information to be obtained about uniformity of the implant. An alternative approach, known as 'natural' volumedose histograms, was developed by Anderson [18] in 1986. The basic aim of their work was to remove one large source of dose variation within the implant, the inverse square law, from the dose-volume histogram in order that the effect of other contributors to nonuniformity to the implant, i.e., poor source positions,
44 Systems of dosimetry
could be more clearly seen. They showed that for a single source plotting the volume in dose bins which varied with dose rate to the power —3/2 would result in every bin containing the same number of dose points or volume elements. However, for multisource implants, a peak in the graph is obtained at a dose rate approximately equal to the average value at points midway between adjacent sources. A typical multisource implant graph is shown in Figure 4.7. The more uniform the dose rate within the implant, the higher and sharper the peak will be. Anderson suggests that the dose rate at which the prescribed dose is delivered should be close to the dose rate at which the histogram peak occurs. If the treatment dose rate is substantially lower than the peak dose rate, undesirably large doses may be given within the implant, with the accompanying risk of necrosis, whereas if the treatment dose rate is significantly higher than the peak dose rate, a large volume outside the target region may be receiving a dose close to the treatment dose. In practice, it has been found that, for implants carried out following the Paris System rules, the peak of the volume-dose histogram corresponds closely to the average basal dose rate [19]. Thus, the position of the peak on the natural dose-volume histogram could be used as a guide to the value of the basal dose rate. For implants which follow the Manchester System, the treatment dose rate has been found to be approximately 60-90% of the peak dose rate. Because the Manchester System is based on delivering an average dose to a particular treatment plane, it is not surprising that the peak of the histogram does not reliably indicate the Manchester treatment dose rate. The position of the peak in the natural dose-volume histogram may be helpful in choosing a treatment isodose, particularly for those implants that were based on the Paris System, even if the final source distribution does not comply strictly with the Paris rules. The brachytherapist should be wary of routinely prescribing to 85% of the peak dose rate without first checking that this will not lead to a significant overdose in any part of the implant.
Arbitrary units on scale of (dose rate) -3/2
Figure 4.7 'Natural' dose-volume histogram for seven evenly spaced wires.
43 INTRACAVITARY THERAPY FOR GYNECOLOGICAL CANCER
4.3.1
Cervical cancer
THE MANCHESTER SYSTEM This system was initially developed for radium tubes, but was easily adapted to different afterloading systems. The Manchester System [20] specifies the spatial distribution of sources in specially designed applicators and their relative activity and defines reference points at which the dose is calculated. The applicators consist of a central tube inserted into the uterus and locating on the cervix by means of a flange, and a pair of ovoids, which normally sit in the vaginal fornices. To cope with individual variations in patient anatomy, different combinations of central tube length and ovoid size are used and the separation between the ovoids can also be adjusted. In the extreme case of the narrow vagina, it may be necessary to insert the ovoids in tandem along the vagina. In the Manchester System, the dose is specified at point A, which is defined as 2 cm superior to the mucous membrane of the lateral fornix (which, for practical purposes, is level with the external os) in the plane of the uterine tube and 2 cm lateral to the center of the uterine canal, as shown in Figure 4.8. The dose at point A is representative of the dose throughout much of the malignant tissue. In addition, point B is defined 5 cm laterally to midline at the same level as the A points and represents the dose to the pelvic wall. The dose at point B is approximately 20-25% of the dose at point A and is of importance when calculating the total dose when brachytherapy is combined with external-beam irradiation. There were two reasons for the choice of point A as the dose specification point: treatment was thought to be limited by the tissue tolerance of the paracervical triangle, and the dose rate at this point is not too sensitive to small variations in applicator position. When loaded according to the rules shown in Figure 4.9, the different combinations of central tube and ovoid deliver approximately the same dose rate to the A points (as shown in Table 4.12), and ensure a suitable ratio between the vaginal and intrauterine contributions to the dose at point A, with the ovoids contributing no more than one-third of the dose. Although point A was thought to be anatomically comparable between patients, it is in fact a point which is related to the geometry of the sources and not to patient anatomy. Consequently, for a non-ideal insertion in which the uterine canal is rotated, the A points also rotate as shown in Figure 4.10. A typical isodose distribution is shown in Figure 4.11. Applicators that reproduce the traditional Manchester geometry are available for manual and low and high dose rate remote afterloading systems. In order to minimize the risk of sources sticking in the afterloading
Intracavitary therapy for gynecological cancer 45
Figure 4.8 Position of A and B points for an ideal insertion, (a) Lateral view, (b) anterior-oblique view.
Figure 4.9 Manchester System: loading patterns for uterine tubes and avoids.
46 Systems of dosimetry
Figure 4.10 Position of A and B points for a non-ideal insertion.
applicators, they usually have a less flexible central tube and the geometry of the applicators is often more rigid. Some configurations, e.g., ovoids in tandem, which were possible with the traditional Manchester applicators may not be possible with all afterloading systems, although alternative applicators may be supplied. In the traditional system, for treatment using brachytherapy alone the dose prescribed to point A at conventional low dose rates is 74 Gy in 140 h split into two fractions with approximately 1 week between fractions. (Please refer to Chapter 25 for further details of cervix treatment.) Manchester applicators do not incorporate rectal shielding. The dose to this critical structure is minimized by careful positioning of the applicators and by careful packing of at least 1.5 cm of gauze behind
Figure 4.11 A typical isodose distribution for treatment of carcinoma of the cervix according to the Manchester System.
the ovoids. The packing also eliminates movement of the applicators during treatment. Strict adherence to the Manchester System means that the dose rate and hence the insertion time are taken from the standard tables for an ideal insertion. Many oncologists, however, prefer to calculate the dose rate, and hence the insertion time, at point A by computer. The use of computer planning and afterloading systems allows standard dose distributions to be modified if required by adjusting the source arrangement or dwell times of individual source trains. Caution must be
Table 4.12 Dose rates at point A for standard Manchester loadings
6-cm uterine tube 4-cm uterine tube 2-cm uterine tube Large ovoids Medium ovoids Small ovoids Large ovoids Medium ovoids Small ovoids Large ovoids Medium ovoids Small ovoids
6, 4, 4 units 6,4 units 8 units 9 units 8 units 7 units 9 units 8 units 7 units 9 units 8 units 7 units
1 -cm spacer 1 -cm spacer 1 -cm spacer Washer Washer Washer In tandem In tandem In tandem
The unit of source activity is 18 u.Gy h-1 (2.5 mg radium equivalent).
34.4 34.2 27.3 18.3 18.8 18.9 18.9 19.0 19.0 14.6 14.9 14.8
References 47
exercised when departing from the standard conditions. In particular, the radiobiological effect of different dose rates must be taken into account. Isodose distributions produced by computer can be used to establish afterloading source distributions which give rise to dose distributions comparable to those produced by conventional radium systems. A remote afterloading implementation of the Manchester System has been described by Wilkinson et al. [21]. OTHER DOSIMETRY SYSTEMS FOR CERVICAL CANCER
THE STOCKHOLM SYSTEM
The traditional technique involved packing the uterus with capsules containing radium sources known as Heyman capsules. The uterus was stretched as these were inserted until it was full. At the end of treatment, the carefully numbered sources had to be removed in reverse order from that in which they had been inserted. The dose specification point is 15 mm from the surface of the most lateral applicator. Lundberg et al. [25] describe an adaptation of this technique for remote afterloading.
433
Vaginal irradiation
The use of Fletcher-type applicators (where shielding is used in the colpostats to reduce the bladder and rectal dose) with afterloading systems has been considered by Marbach et al. [22]. Fletcher applicators require the placement of two sources in the vagina in a manner similar to that of the Manchester System, but the source applicators are of a different shape from Manchester ovoids. Shielding is placed in the upper and lower poles of the vaginal applicators to provide rectal and bladder shielding. The original system did not permit afterloading but there are now modified designs available for afterloading machines. Isodose distributions can also be used to design loading patterns for different types of applicator and technique, effectively tailoring the treatment volume as defined in the ICRU Report 38 [23] to the requirements of the brachytherapist. Excellent descriptions of the Gustave-Roussy technique and the Creteil System are given by Pierquin et al. [7]. The majority of techniques for brachytherapy of the cervix give rise to a pear-shaped irradiation volume. Some systems, e.g., the Newcastle System [24], are designed to give a more cylindrical distribution which can be matched more easily to external-beam irradiation. This is achieved by replacing the two ovoids with a single, cylindrical, vaginal applicator. It is possible to incorporate rectal shielding into the cylindrical applicator. The Newcastle System continues to use the Manchester definition of points A and B for dose specification.
The different dosimetry systems all have slightly different philosophies and it is important not to attempt to change from one system to another without giving due consideration to the consequences to the patient in such an event. The Manchester mould and interstitial system achieves a high degree of uniformity to a particular treatment plane if the rules are carefully followed, but the required source positions may sometimes be difficult to achieve in an afterloading system. The source positioning is simpler for the Paris and Quimby systems, but the dose specification is different. Whichever system is used, it is essential for accurate dosimetry that the source positions of any implant are known precisely. They can be obtained from orthogonal or isocentric radiographs, direct measurement from the patient or a template, shift films or possibly computed tomography. Poor reconstruction of source positions will inevitably lead to poor accuracy of dosimetry.
4.3.2
REFERENCES
Carcinoma of the body of the uterus
THE MANCHESTER SYSTEM
Traditionally, a 1 cm diameter uterine tube was used in conjunction with ovoids positioned in the vaginal fornices, the applicators again being loaded with radium. In the absence of a special-purpose afterloading applicator, a Manchester-type cervical applicator set can be used provided that it is possible to obtain the length (up to 12 cm in extreme cases) of uterine tube required. The loading patterns of the uterine tube and the ovoids differ from treatment of the cervix, with a higher proportion of the dose being delivered by the uterine source train.
Vaginal applicators include cylinders of different diameters, ovoids, and custom-made moulages, which may be used to deliver a surface dose (typically 60 Gy) for treatment given by brachytherapy alone.
4.4
PRACTICAL CONSIDERATIONS
1. Coliez, R. (1923) Les bases physiques de I'irradiation du cancer de col uterin par la CurietherapieJ. Radio!., 7, 201. 2. Patterson, R. and Parker, H.M. (1934) A dosage system for gamma ray therapy. Br.J. Radiol., 7, 592. 3. Quimby, E. (1944) Dosage tables for linear radium sources. Radiology, 43, 572-7. 4. Quimby, E. and Castro, V. (1953) Calculation of dosage in interstitial radium therapy. Am. J. Roentgenol., 70(5), 739-9. 5. Pierquin, B. and Dutreix, A. (1966) Pour une nouvelle
48 Systems of dosimetry
methodologie en curitherapie; le Systeme de Paris. Endo et plesio-radiotherapie avec preparation curietherapie non-radioactive. Ann. Radiol., 9,757.
18. Anderson, L. (1986) A natural volume-dose histogram for brachytherapy. Med. Phys. 13(1), 898-903. 19. Langmack, K.A. and Thomas, S.J. (1995), The application
6. Pierquin, B., Dutreix, A., Paine, C.H.et al. (1978) The Paris system in interstitial radiation therapy. Acta Radiol.
of dose-volume histograms to the Paris and Manchester Systems of brachytherapy dosimetry. Br.J. Radiol., 68, 42-8.
Oncol.,17,33. 7. Pierquin, B., Wilson, J. F. and Chassagne, D. (1987) Modern
20. Tod, M. and Meredith, W.J. (1953) Treatment of Cancer of
Brachytherapy, New York, Masson.
the Cervix Uteri - a Revised 'Manchester Method'. Br. J. Radiol., 26,252.
8. Aird, E.G.A., Jones, C.H.Joslin, CAR et al. (1993) Recommendations for Brachytherapy Dosimetry. London,
21. Wilkinson, J.M., Moore, C.J., Notley, H.M. and Hunter, R.D.
British Institute of Radiology. 9. Meredith, W.J. (ed.) (1949) Radium Dosage, The Manchester
(1983) The use of Selectron afterloading equipment to simulate and extend the Manchester System for
System. Baltimore, Williams and Wi I kins. 10. Gibbs, R.and MasseyJ.B. (1980) Radium dosage: SI units and the Manchester system. Br.J. Radiol., 53,1100-1.
intracavitary therapy of the cervix uteri. Br.J. Radiol., 56, 409-14.
11. MasseyJ.B., Pointon, R.C.S. and Wilkinson,]. (1985)The
Marbach, J.R., Stafford, P.M., Delclos, L. and Almond, P.R. (1984) A dosimetric comparison of the manually loaded
Manchester system and the BCRU recommendations for
and Selectron remotely loaded Fletcher-Suit-Delclos utero-
brachytherapy source specifications. Br. J. Radiol., 58, 911-12. 12. Quimby, E. (1932) Grouping of radium tubes in packs or plaques to produce the desired distribution of radiation. Am.J. Roentgenol., 27(1), 18-39.
22.
vaginalapplicators. In Brachytherapy 1984, Proceedings of the 3rd International Selectron Users Meeting 1984, ed. R.F. Mould. Veenendaal, Netherlands, Nucletron BV, 255-65. 23. International Commission on Radiological Units and
13. Godden, T. J. (1988) Physical Aspects of Brachytherapy.
Measurements (1985) Doseand Volume Specification for
Bristol and Philadephia, Adam Hilger. 14. Anderson, LL, Kuan, H.M. and Ding, I. (1981) Clinical
Reporting Intra-cavitary Therapy in Gynaecology, ICRU
dosimetry with 125I. In Modern Interstitial and
Report 38. Bethesda, MD, ICRU. 24. Dawes, P.J.D.K., Roberts, J.T., Dean, E.M., Lambert, G.D.
Intracavitary Radiation Cancer Management, 3rd edn., ed.
and Locks, S. (1988) The treatment of cancer of the uterine
F.W. George. New York, Masson, 9-16.
cervix using the Newcastle Afterloading System. In
15. Krishnaswamy, V. (1978) Dose distribution around an 1-125 seed source in tissue. Radiology, 126(2), 489-91. 16. Nath, R., Anderson, LL, Luxton, d.et al. (1995) Dosimetry of interstitial brachytherapy sources: recommendations of the AAPM Radiation Therapy Committee Task Group No. 43. Med. Phys., 22(2), 209-34. 17. Yang, G., Reinstein, L.E., Pai, S. and Zhigang, X. (1998) A
Brachytherapy 2, Proceedings of the 5th International Selectron Users Meeting 1988, ed. R.F. Mould. Veenendaal, Netherlands, Nucletron BV, 235-9. 25. Lundberg, L.M., Mattsson, 0. and Ragnhult, I. (1988) Remote afterloading of the uterine cavity using a microSelectron LDRunit. In Brachytherapy2, Proceedings of the 5th International Selectron Users Meeting 1988,
new genetic algorithm technique in optimisation of
ed. R.F. Mould. Veenendaal, Netherlands, Nucletron BV,
permanent 125 prostate implants. Med. Phys., 25(12),
223.
2308-15.
5 Computers in brachytherapy dosimetry ROBERT VAN DERLAARSE AND ROBERT W. LUTHMANN
5.1
INTRODUCTION
Historically, brachytherapy treatment planning involved the use of systems and lookup tables that allowed an empirical approach toward source placement and dose calculations [1-4]. With the introduction of treatment planning computers came the ability to display isodose distributions around an implant and to analyze an implant not only by visual evaluation of twodimensional and three-dimensional dose distributions, but also by dose-volume histograms. Compared to a conventional low dose-rate (LDR) treatment, high dose-rate (HDR) brachytherapy offers two additional degrees of freedom once the catheters or applicator have been placed: source position and dwell time. By manipulating these two parameters, one can, either interactively or by computer methods, optimize the dose distribution around the implant. Computerized optimization algorithms have been developed that explicitly define the desired target dose distribution and then calculate the 'optimal' dwell times and source positions required to obtain this distribution. Optimized HDR brachytherapy is a highly conformal type of cancer therapy. Current developments in the use of computers as applied to brachytherapy include (i) methods for dose calculation around sources, (ii) methods for source localization, (iii) three-dimensional visualization of target volume, organs at risk, and dose distribution
utilizing computerized tomography (CT) and magnetic resonance imaging (MRI) images, (iv) optimization of dose distribution to the target volume, (v) evaluation of dose distribution using dose-volume histograms, and (vi) combination of external-beam and brachytherapy distributions. 5.2 FORMALISMS FOR DOSE CALCULATION AROUND A SOURCE Computer-assisted dose calculations around a brachytherapy implant are currently based upon two formalisms: the traditional formalism [5,6] and the American Association of Physicists in Medicine (AAPM) Task Group 43 (TG43) formalism [7]. Both formalisms determine the dose to a point in medium from a single source. Basically, the AAPM Task Group 43 formalism combines several traditional dose calculation quantities into new ones, as discussed below. Using the traditional formalism, the dose rate at a point with coordinates (r,q), Dr(r,q), can be expressed as: D,(r,Q) =Aa (U/^/pr* [G(r,e)/G(r0,00)] T(r)F(r,Q)
(5.1)
where: Aa
(r,)x,
= apparent activity in mCi. = exposure rate constant [R h-1mCi-1cm2]: for medium and low energy isotopes like iridium-192, the value of (Gd)x depends on
50 Computers in brachytherapy dosimetry
the spectral distribution of the attenuation of radiation filtered through the encapsulation of the source. / = exposure to air kerma conversion factor [cGy R-1] (0.8764 cGy/R-1 in air). (m/p)tissue/air = ratio of the mass energy absorption coefficient of tissue to that of air, where (m/p)tissue/air=/(m/P)air,
G(r,q)
(r0,q0)
T(r)
F(r,q)
= a geometry factor for the dose in point (r,q) accounting for the distribution of the radioactive material. For a point source, G(r,q) = 1/r2. For a line source with active length I, G(r,q) = b/(L r sin q) with b the angle subtended by the active length L at point (r,9) (Figure 5.1). = the reference point that lies on the transverse axis of the source (q0 = 90°) at 1 cm distance from the source. Thus, for a point source, G(r0,q0) = 1 and, for a line source, G(r0,q0) = b/I. = tissue attenuation function (e.g., the Meisberger function [8]). This function accounts for absorption and scatter in tissue along the transverse axis. For medium or low energy isotopes, T(r) also depends on the filtering of the radiation through the source encapsulation. = the angular anisotropy function that accounts for the absorption and scattering in the medium and the source encapsulation. The function value on the transverse axis, F(r,90°), is defined as 1 for all r values.
For the source strength specification of sources, the reference air kerma rate (Kk) or the reference exposure rate (Xk) is usually given: Kk = reference air kerma rate (cGy h-1 m2), the kerma rate in air at 1 m, disregarding air attenuation and room scatter. Xk = reference exposure rate (R h-1 m 2 ), the exposure rate in air at 1 m, disregarding air attenuation and room scatter. In the case of a point source:
The factor 104 accounts for K k and Xk being defined at 1 m and (Fg)x at 1 cm. The AAPM Task Group 43 [7] focuses on the dosimetry of iridium-192, iodine-125, and palladium-103 seeds. However, the TG43 formalism is also applied to HDR and pulsed dose-rate (PDR) stepping sources. It recommends obtaining the dose rate in a point with coordinates (r,q) as follows:
Where: = air kerma strength (cGy h-1 cm2): the air kerma rate on the transversal axis at 1 cm distance from the source, based on a measurement at a large distance and converted to 1 cm distance by considering the source as a point source. Thus, this quantity is related to the reference air kerma rate Kk(cGy h-1 m2) by the inverse square law: S k =10 4 K k . A = specific dose rate constant: the dose rate in tissue per unit air kerma strength at 1 cm distance from the source. This constant is not based upon an idealized point source, but, instead, is dependent upon the physical configuration of the source. g(r) = radial dose function: this function accounts for absorption and scatter along the transverse axis. Note that g(r) = 1 for r = r0 (= 1 cm). Sk
For a point source, equation (5.3) can be obtained from equation (5.1) by combining several traditional dose calculation quantities into new ones:
Figure 5.1 Illustration of the geometry used in the dose calculation around a line source. L = active length of line source; P( r o>q 0 ) - reference point on the transverse axis of the source; P(r,q) = point to calculate the dose at; (3 = angle subtended by the active length L.
Note that Sk is the air kerma rate at 1 cm on the transverse axis of the source as if the source is a point source. SkA is the real dose rate in tissue at 1 cm on the transverse axis of the source. From this follows that A is also a function of the active length of the source.
Calculating of dose by computer 51
53
CALCULATING OF DOSE BY COMPUTER
To calculate dose to points around an HDR source, either equation (5.1) or (5.3), or a mixture of both, can be used by computer algorithms. The orientation of the source in free space must be known in order to apply the geometry factor G(r,q) and the anisotropy factor F(r,q). Localization techniques that supply this information are an essential part of any brachytherapy treatment planning system and are discussed in section 5.4. In this section, the different dose calculation parameters are discussed, with emphasis on sources used in afterloading.
53.1
Source strength
Typically, the source strength is derived from the measured air kerma rate at 1 m, a distance much larger than 1 cm, at which distance a source is practically a point source. The reference air kerma rate kk, defined at 1 m, is obtained directly from this measurement. The air kerma strength Sk is derived from Kk using Sk = 104 Kk for both a point source and a line source. Thus, Sk is the air kerma rate at 1 cm on the transverse axis of the source, as if the source is a point source. The apparent activity (Aa) follows from Kk using equation (5.2), Aa = 104 Kk/[(Gd)xf]. For iridium-192 sources, Kk or Sk values should be entered into the planning program instead of the derived Aa values, as there are different values published for (Gd)x. If a planning system requires the entry of apparent activity, then the same (Gd)x must be used when converting measured air kerma rate to apparent activity as when calculating dose values. Otherwise, a serious error in the dose given to the patient may occur.
53.2
Specific dose-rate constant (A)
The specific dose-rate constant (A) is defined as the dose rate in tissue per unit air kerma strength at 1 cm from the source center along the transversal axis of the source. It depends upon the physical configuration of the source, i.e., its active length, and upon the radiation spectrum. The latter influence is due to the tissue scattering and absorption factor at 1 cm, T(r0), which depends on the radiation spectrum. Thus, there will be different dose rate constants for identical isotopes in sources with different physical configurations.
533
Geometry factor, 6(r,q)
Depending on the active length of a source, either a point source or a line source should be assumed. For a point source, G(r,q) = 1/r2 and for a line source with active length I, G(r,q) = b(I r sin 0) with (3 the angle
subtended by the active length L at point (r,q). The modeling as a line source is of importance when calculating dose rates at distances shorter than 2 L. If HDR stepping sources with an active length of 3 mm or more are not modelled as a line source, doses calculated at distances shorter than 6 mm are less accurate [9-11].
53.4
Dose anisotropy function, F(r,q)
The dose anisotropy function, F(r,q), accounts for the anisotropic behavior of the dose distribution around the source, due to the self-absorption in the active material and the attenuation in the encapsulation of the source. The value of the anisotropy function at the distance r from the source along the transversal axis, F(r, 90°), is defined as 1.0. The anisotropy correction can be handled in several ways, depending upon the radiation spectrum of the isotope and the physical form and encapsulation of the source. Measured F(r,q) values, stored in tables, are often fitted to functions that are used in computer algorithms. For sources emitting high energy radiation, such as a cesium-137 source, the anisotropy correction for a point P(r,q) can be calculated by subdividing the active material in the source in small volume elements Da. For each element Da the path ra through the active material itself and the path rw through the source wall are determined. This gives a correction factor exp [-(ma ra + mw rw ], with ma the linear attenuation coefficient for the active source material and |iw the linear attenuation coefficient for the encapsulation material. By summing over all active volume elements Da, the correction factor C(r,q) for point P(r,q), is found. NowF(r,q) = C(r,q)/C(r,90°). In the traditional approach of dose calculation around radium and cesium tubes, the interval method is used. This method divides the source in a large number of point sources and calculates the attenuation of the rays from each point source by the source encapsulation [12,13]. Thus, there is no strict separation between the geometry factor and the anisotropy factor. For sources with medium or low range energies, an analytical correction is not possible due to the dependence of the tissue scattering and absorption on the radiation spectrum. For such a source, a table with measured values of F(r,0) or a function fitted to these values is used. The AAPM Task Group 43 report [7] gives tabulated values of -F(r,q) and G(r,q) for iodine-125, iridium192, and iridium-103 seeds, currently in use. Some authors present tables which are corrected for the inverse square law only and normalized to 1 at 1 cm distance from the source center on the transversal axis, thus representing F(r,q) g(r) values. The multiplication factor to the dose in the point (1 cm, 90°) to obtain the dose in point (r,0) then becomes F(r,q) g(r)/r2 [14]. For sources of medium range energies, such as iridium-192, there is less variation in the values of the
52 Computers in brachytherapy dosimetry
anisotropy function with distance r from the source. Therefore, the anisotropy function in some older planning systems is taken as F(q), with F(90°) again defined as 1. It is usually implemented as F(q)/r2, with F(q) a table with values for 0 between 0° and 180° [15]. Accurate dose values around an HDR source with an active length of 3.5 mm are obtained by using F(q) G(r,q), with G(r,q) applied for a line source of 3.5 mm. These dose values are even valid for short distances up to 1 mm to the source center [9]. If the orientation of an HDR stepping source in space is not known, then only the inverse square law with a fixed anisotropy factor, the anisotropy constant (jan, can be applied. The value of (jan is less than 1.
53.5 Tissue attenuation factors, T(r) andg(f) The tissue attenuation function, T(r), and the radial dose function, g(r), both account for the effects of absorption and scatter in tissue along the transverse axis of the source. In most computer algorithms, the tissue attenuation function is applied to represent the dose fall-off along the transverse axis, due to tissue absorption and scattering. The function T(r) is normally expressed as a polynomial in the form of:
The parameters a; most often used are those of Meisberger et al. [8], which are valid for the range of 1 cm through 8 cm. At depths beyond 8 cm, the above expression decreases sharply and, at those depths, T(r) is usually approximated by exp(m t r). A fit function with the least number of parameters is the modified Van Kleffens and Star function [16]:
For indium-192 the parameter a equals 0 and equation 5.8 reduces to T(r) = d(l + b r2) with 8 = 1.008 and b = 0.0012 cm-2. Equation 5.8 coincides with the Meisberger relation within 0.5% for sources with medium and high range energies, such as iridium-192, cesium-137, and cobalt-60. At depths beyond 8 cm, equation 5.8 decreases gradually and a separate exponential attenuation function is not needed when calculating clinical dose distributions around implants or applications.
5.4 RECONSTRUCTION OF SOURCE LOCALIZATION In order to obtain the dose distribution around an implant or application, the exact position of each source or dwell position in space must be known. For the reconstruction of the source localizations by a treatment planning program, different techniques are available.
5.4.1
Specification of coordinates
If the three-dimensional coordinates of the sources or dwell positions are known, keyboard entry of these data can be used. This, of course, is the most accurate description of the source positions.
5.4.2 Localization using film imaging techniques If the absolute coordinates are not known, imaging localization techniques must be used. These imaging techniques utilize either plane radiographs taken from different directions, or computed tomography (CT)/ magnetic resonance imaging (MRI) images of the implant. If an afterloading technique is used, it is necessary to simulate the position of the sources or dwell positions utilizing localization markers. These X-ray markers must be (i) easily discernible on the radiograph or CT/MRI slice, (ii) accurately depicting the source or dwell position, and (iii) coded such that the user can determine which markers correspond to a given catheter or applicator. They usually consist of a thin, braided, metal wire with markers of high Z metal at every centimeter. By counting from the catheter tip, corresponding images of a given X-ray marker are easily found. The images of these markers are called catheter describing points as the corresponding images of two plane radiographs are projections of the same point in a catheter. In implants with many catheters close to one another, it is sometimes difficult to follow these X-ray markers. In this case high Z wires can be inserted in each catheter up to the catheter tip. The images of these wires, starting at the catheter tip, are digitized from two-plane radiographs. The localization in space of each catheter can be reconstructed from its images on the two radiographs by dedicated software. This technique is called catheter image tracking. The points describing the curvature of an image are called catheter image points. Contrary to the catheter describing points, there is no direct link between the catheter image points on the two radiographs. ORTHOGONAL RECONSTRUCTION METHOD
The most widely used radiographic method for source localization is the orthogonal reconstruction method [17]. Two-plane radiographs of the implant are taken in a lateral and antero-posterior (AP) orientation. Either a radiotherapy simulator is used or a localization box with cross-wires on the faces of the box is placed over the patient (Figure 5.2). In the latter case, the beams are aligned such that the X-ray images of opposing crosswires coincide. The advantage of this technique is that AP and lateral radiographs are easily interpreted by the physician. A disadvantage is, however, that sources or X-ray markers in the lateral X-ray film are often difficult
Reconstruction of source localization 53
Figure 5.2 Orthogonal reconstruction. The beam set-up is obtained by calculation of the localization of the AP and lateral X-ray foci from the cross-wire images on the radiographs. Instead of a reconstruction jig to adjust the AP and lateral beams, a radiotherapy simulator with gantry angles of 0° and 90° can be
Figure 5.3 Semi-orthogonal reconstruction. The beam set-up is obtained by calculation of the localization of the AP and the lateral X-ray foci from the size and the displacement of both cross-wire images on the radiographs. I = center of box; P - point to be reconstructed.
used. I - isocenter; P = point to be reconstructed.
ISOCENTRIC RECONSTRUCTION METHOD
to distinguish, particularly in the pelvic region, due to the thickness of tissue and the overlying bony structures. SEMI-ORTHOGONAL RECONSTRUCTION METHOD
Truly orthogonal orientations for the AP and lateral film are not easily obtainable with portable X-ray units. In the semi-orthogonal reconstruction method, a localization jig with AP and lateral cross-wires is placed over the patient and two radiographs (a lateral and AP) are taken [18] (Figure 5.3). It is not necessary for these to be truly orthogonal, because the set-up information will be determined by the computer from the size and the relative distances of the cross-wire lead marker images on each of the two films. This method, therefore, accepts X-ray beams whose central axes do not intersect and are not perpendicular to one another. The only requirement is that the projections of the crosswires on the two corresponding box faces are visible on the radiographs. The semi-orthogonal reconstruction method has proven to be particularly advantageous in HDR endobronchial applications. Directly after insertion of the catheters, a portable radiographic X-ray unit can be used with the localization box to obtain the radiographs for localization of the catheters. This will shorten the time between the insertion of the catheters and the treatment of the patient considerably, because there is no localization session on a radiotherapy simulator required.
With an isocentric X-ray unit, such as a radiotherapy simulator, two images of the X-ray markers in each catheter can be obtained on a single, large-size radiograph [17] (Figure 5.4). The gantry angle of the first X-ray beam is -a and of the second beam is +a, with the
Figure 5.4 Isocentric reconstruction, a - reconstruction angle; FID -focus to isocenter distance; IFD = isocenter to film distance; I = isocenter; P=point to be reconstructed. This method requires an isocentric X-ray unit such as a treatment simulator. A large-size film is placed under the patient. The beam set-up is obtained by rotating the gantry over an angle of +a and -a, with a preferably in the range of 15°-30°.
54 Computers in brachytherapy dosimetry
total angle (2a) between the central axes of the projecting beams taken as large as possible. In order to visualize both images on this radiograph, it is essential that the isocenter is placed in the center of the implant and that the X-ray field is defined such that the two images on the radiograph do not overlap. Due to the equal angles between the left and right central axes with respect to the normal to the radiograph, lines between corresponding points on the left and right image all run parallel. This aids in the determination of individual seed or dwell positions from one image to the other. Usually, the gantry angle of each image with respect to the normal of the radiograph is between 15° and 30°, depending on the extension of the implant and the distance between the isocenter and the X-ray film. STEREO-SHIFT RECONSTRUCTION METHOD
The set-up is similar to that of the isocentric method, but, instead of the X-ray tube rotating, it is moved laterally over a given distance [ 17]. This method is applicable, for instance, with a ceiling-mounted X-ray unit where such a lateral movement is available. Usually, the angle between the two projecting beams is very small, typically 7°, which makes this method very sensitive to even small errors in the measuring of the source images or a small movement of the patient between the taking of the radiographs. If possible, this reconstruction method should be avoided. VARIABLE ANGLE RECONSTRUCTION METHOD
This method reconstructs the localization of the catheters from two radiographs taken with a therapy treatment simulator [19] (Figure 5.5). The only limitation is that the central axes of the projecting beams are not coinciding or opposing. The reconstruction algorithm requires that the angle, focus-isocenter distance, and isocenter-film distance of each radiograph are accurately known. The advantage of this technique is that the implant can be observed fluoroscopically at various gantry angles to define the two gantry angles that display the sources or localization dummies with the highest clarity and least obstruction. It is preferred that the total angle between the two projecting beams lies between 60° and 120°. Of course, the greatest accuracy is obtained when this total angle is 90°. The orthogonal reconstruction method is a special case, e.g., with gantry angles 0° and 90°. RECONSTRUCTION METHODS USING CORRESPONDENCE LINES
A digital image can be obtained by scanning a radiographic film with a film scanner or, in real time, by a portal imaging device mounted on the X-ray equipment. These images are then displayed on the treatment planning computer which has options to enhance them to
Figure 5.5 Variable angle reconstruction. ®a=gantry angle beam 1; b=gantry angle beam 2; FID 1 -focus to isocenter distance beam 1; IFD 1 = isocenter to film distance beam 1; FID 2 =focus to isocenter distance beam 2; IFD 2 = isocenter to film distance beam 2; I = isocenter. The two gantry angles a and b are determined such that the catheters are clearly visible on the image intensifier. The total angle a+b should preferably lie
between 60° and 120°. The radiographs are made with the film in the cassette holder on top of the image intensifier.
better delineate the source or X-ray marker positions, the target volume, and the surrounding organs. A problem exists with imaging equipment such as image intensifiers, where the image may be distorted by the cushion effect and the influence of the earth magnetic field, which will change with the gantry angle. To minimize these effects, the implant should be kept in the center of the image. Of course, the reconstruction methods based on two radiographs taken at different directions can still be used by replacing the digitizer with the mouse. However, if both digitized films can be displayed together on the monitor, an extension of the variable angle reconstruction method becomes possible. By pointing with the mouse to a point on one of the digitized films, the planning program can draw the ray from the corresponding X-ray focus to that point. The other X-ray focus will project that ray to the other film as a line over it. This line on the second film is called the correspondence line to the point on the first film (Figure 5.6). Thus, by moving the mouse over one of the digitized films to the image of an X-ray marker, the correspondence line moves over the other film until it intersects the other image of this marker. In this way the corresponding images in both radiographs are easily found and reconstruction methods based on catheter describing points can be used.
5.43
Tracking of catheter images
Sometimes, visually matching of X-ray markers on the two localization radiographs is not easily done, as in the
Reconstruction of source localization 55
Figure 5.6 Reconstruction using correspondence lines. C1 and C2 are images of the same X-ray marker in the catheter. [1] and [2] are the films obtained by variable angle reconstruction. They are digitized by a film scanner and displayed on the monitor. After entering marker image C2 on film 2, the ray of focus 2 to C2 is projected to film [1] as the correspondence line of C2. The intersection of this projection with the catheter image on film 7 gives the point C1. The intersection of the rays focus 2-C2 and focus 1-C1 gives the reconstructed localization of the marker in the catheter. Similarly, if C1 is entered, the corresponding image C2 on film 2 can be found. With this method, corresponding image points can be found without the use of an X-ray marker in the catheter.
case of implants with many catheters. In a lateral radiograph only a cloud of X-ray markers is visible, but not the braided metal cables. In such an implant, continuous radio-opaque wires should be used instead of the X-ray markers. As stated earlier, the isocentric reconstruction method connects corresponding catheter describing points between the two X-ray images with parallel lines. The localization of a catheter in space from its two images on the isocentric film is reconstructed as follows. First, the two X-ray images of the catheter are described with catheter image points. If on both X-ray images the catheter tip is clearly visible, the line connecting these tips is taken as the horizontal base line. This corrects for any patient movement between the taking of the two exposures. Otherwise, the line connecting the images of the isocenter is taken as the horizontal base line. With appropriate software, any set of variable angle radiographs taken with the same focus isocenter distance can be converted to a computer-generated isocentric 'radiograph' (Figure 5.7). Then each point with which a catheter image is described can be connected by the planning software to its corresponding point on the other image by a line parallel to the one connecting the images of the catheter tip. This conversion of the variable angle reconstruction to the isocentric one is not possible with the semi-
orthogonal reconstruction method. In the case of afterloaded sources or of stepping source dwell positions in a catheter, image tracking reconstruction can still be used [ 19]. Again, the tip of each catheter needs to be visible on both images. Subsequent points on each image are placed such that they describe the curvature of the image. Each one of the two catheter images is then digitized, starting with the tip, until a point beyond the last source or dwell position is entered. The advantage of this technique is that the points depicting the image on the first radiograph are not corresponding to the ones on the second radiographic image. Thus, any radio-opaque marker (e.g., any flexible metal cable) can be used to depict the catheter. This technique simplifies the data entry into the planning system, but is less accurate in handling any patient shift between the taking of the two radiographs, as is discussed in the section on reconstruction accuracy below. Digital images from CT or MRI scanners are also used for treatment planning. Usually, a series of parallel, transverse slices through the treatment volume and its surroundings is obtained. The digital images are displayed on the monitor of the treatment planning computer, which has options to enhance them to better delineate the source or X-ray marker positions, the target volume, and surrounding organs. They are distinguished by their position along the caudal-cranial axis along the CT table top. The coordinates of an image of a source or X-ray marker can be obtained with a pointing device such as a mouse. Reconstruction of the localization is straightforward; the x-coordinate and y-coordinate are obtained directly from the mouse coordinates and the 2-coordinate is given by the table position of the slice.
5*4.4
Reconstruction accuracy
ACCURACY OF THE DIFFERENT RECONSTRUCTION METHODS
The most accurate reconstructed catheter or source coordinates are obtained when the central axes of the two projecting beams are perpendicular to each other. This is achieved by the orthogonal reconstruction method or by the variable angle method when the gantry angles differ by 90° (or 270°). The accuracy remains high if the angle between the two beams lies in the interval (60°-120°). A reconstruction set-up which uses an angle outside that interval, thus smaller than 60° or between 120° and 180°, becomes sensitive to patient shifts between the taking of the radiographs and to digitizing errors. The stereo-shift reconstruction method is extremely sensitive to digitizing errors, due to the small angle between the two projecting beams, up to 10°, and should be used with great care [17]. The reconstruction accuracy depends also on the use of X-ray markers.
56 Computers in brachytherapy dosimetry
Figure 5.7 Variable angle reconstruction using isocentric pseudo-film. The intersections of the rays to the catheter image points with a computer-generated isocentric plane are determined. Corresponding images on the pseudo-film are all connected by parallel lines. This method is suitable for reconstruction of implants with flexible catheters using a therapy simulator. If the patient shifts between the taking of the radiographs, lines between corresponding images are parallel to the line connecting the images of the catheter tip.
RECONSTRUCTION USING a OR MRI SLICES
The main factor determining the accuracy of the reconstructed localization is the slice spacing, the distance between consecutive slices. A typical value of 4 mm results in an accuracy of 2 mm in that direction. The choice of the material for the catheters and the X-ray markers is also essential because dense high Z material will introduce artifacts in the scan. RECONSTRUCTION USING CATHETER DESCRIBING POINTS
The localization in space of a catheter describing point from its two X-ray images is defined by constructing the two rays projecting the catheter describing point to its X-ray images. These two rays will not intersect because there will always be some movement of the patient between the taking of the two radiographs. The reconstructed localization of the catheter describing point is taken halfway on the line connecting the two rays along their shortest distance (Figure 5.8). Thus, if the patient
moved a certain distance between the taking of the two radiographs, the resulting error in localization of the catheter describing point is only half that distance. RECONSTRUCTION USING CATHETER IMAGE TRACKING
The accuracy of image tracking reconstruction depends strongly on any movement of the patient between the taking of the two films. Because the projections of the tip of a catheter on both radiographs are usually visible, the shift of this catheter due to a patient shift between the taking of the two radiograph images can be determined. One of the images can then be adjusted by the planning system such that the rays from the X-ray foci to these first image describing points intersect. However, a rotation of the patient cannot be taken into account and will result in an error in the reconstructed localization of the source or wire. With catheter image tracking, the error in catheter localization due to a given patient movement is about twice the error obtained with catheter describing points.
Optimization techniques in stepping source brachytherapy 57
rithms with varying user-defined constraints may all deliver different dose distributions. Therefore, clinical experience will always be needed to judge the mathematically optimized dose distribution for actual patient treatment. Based on that judgement, changes may be made in the optimization constraints, resulting in a new optimized dose distribution, before the final dose distribution will be accepted for clinical use. 5.5.1
Figure 5.8 Correction for patient shift between thetakingof radiographs 7 and 2. To find the localization of an X-ray marker, the shortest distance between the rays to the corresponding images of this marker is determined. The reconstructed localization of this marker is placed midway on the line along this shortest distance.
5*5 OPTIMIZATION TECHNIQUES IN STEPPING SOURCE BRACHYTHERAPY Once the catheters are placed in the patient, stepping source brachytherapy offers two degrees of freedom: the dwell position and the dwell time. Usually, the dwell positions are placed in the sections of the catheters that are inside the target volume. Then optimization of the dose distribution is performed by manipulation of the dwell times either by the user or by dedicated software. Most optimization procedures do not determine the absolute dwell times for each dwell position. Instead, they result in a set of relative values for the dwell times in the range 0.0-1.0 with a corresponding set of relative dose values in the dose points. Another module of the treatment planning program calculates the absolute dwell times for the stated mean dose in these dose points. There are several factors which influence a mathematically optimized dose distribution. For example, due to the radial nature of radiation from a point source, it is not possible to obtain along the axis of a catheter a prescription isodose curve in the shape of a box. Also, if the placement rules for the catheters of a given target volume are not adhered to, it is difficult, if not impossible, to obtain a good dose distribution. Or, if an implant is not covering the target volume geometrically, manual changing of relative dwell times may be required to cover the target volume with the prescription isodose surface. Thus, a mathematically optimized dose distribution does not always represent the best possible one in and around an implant. Finally, different types of optimization algo-
Distance and volume implants
In mathematical optimization, two types of implants are distinguished: distance implants and volume implants. In a distance implant, dose points are placed at a given distance around the catheters. The mathematical optimization aims at determining the dwell positions and relative dwell times such that the prescription isodose surface passes through these dose points [ 16,20,21]. This is called 'optimization on distance.' Examples of distance implants are single catheters, double catheters, and single-plane implants. If digital imaging with transverse slices is used, these so-called dose points can be placed equidistantly on the target contours in the displayed cross-sections of the patient. With reconstruction by radiographic films, dose points can be placed only relative to the catheters. A volume implant contains one or more planes of catheters. Series of dose points are placed inside the target volume midway between the catheters and throughout the implant. In a volume implant, the relative dwell times are optimized to the same dose in these dose points. This is called 'optimization on volume' [16,21]. The prescription dose is defined as a given percentage of the mean dose in these dose points, with the prescription isodose surface encompassing the target volume as closely as possible. However, the above definition of optimization of volume is a gross simplification. The placement of dose points midway between the catheters alone is not sufficient for optimization (Figure 5.9). In cases in which dose point placement is complicated due to irregularities in the distances between implanted catheters, the dwell positions themselves can act as dose points for optimization. This technique is called 'geometric optimization' and can be performed either on distance or on volume [16,23]. 5.5.2
Rules of optimization
Once a stepping source implant is optimized, it can be evaluated by the following rules. Rule 1. The optimized isodose distribution should match the requirements specified by the physician. In the case of a distance implant, the prescription isodose surface should pass as closely as possible through each of the defined dose points. For a volume implant, the
58 Computers in brachytherapy dosimetry
Figure 5.9 Distance optimization of a volume implant. A two-plane implant with five parallel catheters is optimized to the same dose in rows of dose points, placed midway between the catheters. A perfect fit is obtained by only activating the central catheter. If the sum of the squares of the dwell times is minimized, the central catheter is switched off. Dwell positions 1-13 in each catheter; step length 0.5 cm; reference air kerma rate 4.0682 cGy h-1 m2, i.e., 10 Ci iridium-192. •: Active source dwell position; O: Inactive source dwell position; •: Dose points, midway between catheters along 6 cm active length.
dose in the volumes midway between the catheters should be as homogeneous as possible throughout the implant. Rule 2. The shape of the isodose surface close to the applicator or catheters should be smooth and resemble as closely as possible the shape of the isodose surfaces at distances further from the catheter. Rule 3. Active dwell positions should not extend outside the target volume. Rule 4. Optimization algorithms should be fast enough to keep the time between the application and treatment to a minimum. The first rule of optimization addresses the optimization goals of a distance and a volume implant. The second rule aims at the minimization of hot or cold spots in the implant, especially close to the active dwell positions. The third rule requires the active dwell positions to remain within the target volume, because extending beyond this boundary would increase dose to healthy tissue. To satisfy this rule, the most distal dwell positions must be increasingly weighted to account for the lack of contributions of dose from adjacent dwell positions. The fourth rule minimizes the hardship to the patient by keeping the time that the catheters or applicators remain in the patient to a minimum.
5.53 Optimization of distance implants by least square minimization The aim in optimization of a distance implant is to have the prescription isodose surface pass at a given distance from the dwell positions along the catheter(s). To accomplish this, dose points are placed at this distance from the implant, relative to the successive dwell positions. To obtain the best combination of dwell positions and relative dwell times, a mathematical object function must be minimized. This consists of the difference between the actual dose calculated at each dose point and the ideal dose requested. Usually, this optimization problem is addressed by least squares minimization. In that case, if there are N dose points, each receiving a dose contribution from M sources (active dwell positions), then the following least squares function is to be minimized [16,21]:
where N is the number of dose points, Dpi is the prescribed dose to point i, and Dci is the calculated dose to point i. (See the equations representing the prescribed doses, Dpi, in Figure 5.10.)
Optimization techniques in stepping source brachytherapy 59
Figure 5.10 Optimization on distance of the dwell times in a straight catheter to prescribed doses in dose points along the catheter. Dpi=prescribed dose in dose point i; Dcj = calculated dose in dose point i; Au = dose in point if mm dwell position] for tt = 1, see equation 5.10. The Chi-square function x2 is minimized by setting the derivatives 8%V8f,for each dose point i to 0 and solving the resulting set of equations.
The dose Dci to a dose point i from each source j is calculated by using equation 5.3 for a point source:
where Sj is the source strength of source j: Sj = (Sk A)j tj is the time that source j stays at distance ri,j C( r i,qi,j) combines the radial dose function and the anisotropy function:
Ai,j
is the dose contribution in point i from dwell position j for tj = 1.
In the case of a stepping source implant, the source strength Sj for all j sources is the same. Thus, the only variable that can be altered to minimize equation 5.9 is the dwell time tj at each dwell position j. x2 can be minimized by setting its derivative to each tj toO:
In this way, M equations are obtained which are linear in their M unknown tjs. There are a number of mathematical procedures available to solve these equations. This results in a set of values for the tp. A problem arises when there are fewer dose points than dwell positions. Then, the set of equations is underdetermined and many mathematical solutions exist. To arrive at a unique solution, an additional criterion must be supplied. Intuitively, a suitable criterion would be to minimize the sum of the squares of the relative dwell times,
thus
suppressing wildly varying values. The method of singular value decomposition (SVD) contains this additional criterion [24]. The SVD method minimizes the least square differences between prescribed and calculated dose in the dose points. If there are fewer dose points than relative dwell times (i.e., fewer equations than variables), SVD also minimizes the sum of the squares of the relative dwell times in order to arrive at a unique solution. DWELL TIME GRADIENT
When a set of equations linear in tj is solved, negative relative dwell times may result. For example, in Figure 5.11 (a), an endobronchial application is simulated by a single catheter with 25 dwell positions. It is optimized using 25 corresponding dose points at a lateral distance of 1 cm from each dwell position. The solution which gives exactly the prescribed dose to the dose points results in large positive and negative values for the relative dwell times in adjacent dwell positions. This is unacceptable because large fluctuations will cause hot and cold spots (a violation of the second rule of optimization) and negative relative dwell times are a physical impossibility. An unacceptable solution to this problem is to set all negative relative dwell times to zero. In doing so, the calculated dose in the dose points changes considerably, which will offset the requirement of an equal dose to each dose point. A solution to prevent these negative and strongly fluctuating relative dwell times was developed by Van der Laarse et al. [16,19-21]. By gradually suppressing large fluctuations of dwell times in adjacent dwell positions, the negative relative dwell times must eventually all become positive, because in the limit situation where all dwell times are equal, they are positive (Figure 5.1 Ib). To
60 Computers in brachytherapy dosimetry
dwell time gradient restriction is applied to equation 5.9 as follows:
where w is the weighting factor for the dwell time gradient restriction. This expression remains linear in tj so x2 can be minimized as given by equation 5.11. By increasing the weighting factor, w, the dwell time gradient in adjacent dwell positions is reduced. This, in turn, reduces the likelihood and magnitude of negative relative dwell times (Figure 5.lib). The minimum value for the weighting factor is the one that makes all values of the relative dwell times positive or zero. As previously stated, the concept of the dwell time gradient is based on the second rule of optimization, which requires that the isodose surfaces remain as close in shape as possible as they get closer to the catheters. Thus, by requiring a smooth transition of relative dwell times along the catheters, smooth isodose surfaces will be obtained with a minimum of hot and cold spots close to the catheters. POLYNOMIAL OPTIMIZATION
Figure 5.11 The influence of the dwell time gradient restriction on the dwell times of a straight catheter implant optimized on distance. By restricting the difference between the dwell times of adjacent dwell positions, an optimized solution with positive or zero dwell times is obtained, (a) Optimized relative dwell times where there is no dwell time gradient restriction (DTGR) imposed on dwell times of adjacent dwell positions. The maximum
Optimization of a distance implant by setting the derivatives of x2 to each tj to 0 results in M equations with M unknown tp. Thus, in an M x M coefficient matrix, with M being the number of active dwell positions, large implants may exceed the available memory of the planning computer. For example, an implant with 500 active dwell positions requires the planning computer to solve 500 equations with 500 unknown tjs. Also, the computation times may become unacceptably long. The dwell time gradient also offers a solution to this problem. Because the differences between successive relative dwell times are smoothed, the relative dwell times in a given catheter can be approximated by a continuous function t(x) with x the distance to the catheter tip [16]. Thus, the relative dwell time at dwell position j with a distance xj from the tip, is given by tj = t(xj). For t(x), a set of polynomial functions, Pm(x), to the order m can be taken, similar to describing a continuous curve by a Fourier series expansion to a given order. Thus:
difference of successive dwell times is nearly 2, the limit value. (b) Optimized relative dwell times with a small DTGR of 0.01. The maximum difference of successive dwell times is reduced to 1.2. (c) Optimized relative dwell times with a DTGR of 0.18. The maximum difference of successive dwell times is 0.4. All dwell times are positive or zero, so this represents the best possible fit of an isodose line through the dose points.
implement this limitation of the variance of relative dwell times between adjacent dwell positions, the socalled dwell time gradient restriction was developed. The
where m is the index, indicating the order of the polynomial Pm(x), am is the mth parameter, Pm(x) is the polynomial of order m, and p is the number of parameters required for an adequate approximation of tj by t(xj at all dwell positions x}. By inserting equation 5.13 into equation 5.12, x2 is now expressed as a function of the p parameters am instead of the M relative dwell times tj. Minimizing of x2 is again obtained by setting the derivatives dx2 da m to 0. This leads to a p x p coefficient matrix instead of the M x M one when the derivatives of x2 to tj are taken.
Optimization techniques in stepping source brachytherapy 61
Figure 5.12 Relative dwell times as in Figure 5.Tib, with just sufficient dwell time gradient restriction but now parameterized in terms of distance along the catheter. t(x) is approximated by polynomials P(x) to the degree m with m<M.
Take as an example a catheter with 33 active dwell positions (Figure 5.12). The value of p will be much smaller than 25 because, due to the dwell time gradient, the dwell times are interrelated. A suitable value of p is given by p = 2 m - 1. Thus, the set of 25 tjs can be described by nine parameters, am. This reduces the memory requirements by a factor of 8, from 25 x 25 to 9 x 9.
5.5.4 Optimization of distance implants by linear programming Linear programming, as applied to stepping source brachytherapy, solves the problem of minimizing a function linear in the dwell times, subject to a finite number of constraints. These constraints are again linear in the dwell times. The problem is formulated mathematically as follows [25]:
mized to the shortest overall time possible, with positive or zero dwell times and with the doses in the dose points at least equal to the prescription dose [26]. A disadvantage of this technique is that the dose distribution is solely defined by these dose points, with no regard to the dose distribution close to the catheters. This may conflict with rule 2 of optimization, which states that isodose surfaces near the catheters should be smooth and resemble as closely as possible the shape of the isodose surface through the dose points. Thus, implants optimized by equation 5.14 often show the occurrence of hot and cold spots at distances close to the catheters. Of course, this is not a problem when optimizing a vaginal cylinder application, where the dose inhomogeneities are within the plastic of the applicator. However, for an endobronchial implant, these volumes with high and low doses are within the target volume. Note. The concept of minimum dose points on circles with a given radius around the dwell positions can also be applied using least squares optimization. First, evenly spaced dose points are constructed on the circle around each dwell position, perpendicular to the catheter. The radius of the circle is the distance to the prescribed dose. Then, all relative dwell times of the active dwell positions are set to unity and the doses in these dose points are calculated. Finally, for each circle around an active dwell position, the point with the minimum dose is found and stored. These points with minimum dose will always lie in the area of lowest dose, even after optimization. They define the surface of the treatment volume. The points with minimum dose are now taken as dose points for the polynomial optimization technique, previously described. This technique has the advantage of minimizing the effort of placing the dose points, like in the linear programming approach, but overcomes the problem of irregular isodose surfaces close to the catheter by applying the dwell time gradient restriction.
5.5.5 Optimization of distance implants by simulated annealing In the case of a distance implant, one can construct circles with a given radius around each dwell position which are perpendicular to the catheter. The radius of the circles is the distance to the prescribed dose. Dose points are placed on the circles at equal angle intervals, e.g., every 15°. The calculated dose, Dci, in each dose point i is required to be at least equal to the stated prescription dose Drep thus Dci > Dref. Using equation 5.10, this translates in equation 5.14 to ay= C(r i,j qi,j) and b = Dref, again M being the number of dwell positions and N the number of dose points. If as object function M
is minimized, taking cj = 1, the treatment is opti-
Another optimization technique suitable for brachytherapy treatment planning is simulated annealing. This technique solves the optimization problem in a stochastical way, using a directed random search for the dwell times to seek the lowest value for an object function such as defined in equation 5.14 [27]. Simulated annealing proceeds iteratively as follows: 1. An initial solution (set of dwell times and/or source positions) is chosen and evaluated using the objective function. 2. A new solution is constructed from the current one by varying the dwell times in a random direction and by an amount which is initially large but decreases sufficiently slowly so that a global minimum can be found. If the new solution is
62 Computers in brachytherapy dosimetry
better, it replaces the current one. If it is worse, it is accepted with a probability that depends on the difference from the current one and on the size of the change in dwell times. 3. The process iterates with the amount of change in dwell times being reduced sufficiently slowly such that the system can search out the region in solution space containing the global optimum and converge to it. This method has been applied by Sloboda [28,29] for planning low dose-rate treatments with trains of active and non-active pellets in the catheters. In this case, a step is performed by switching an active position to an inactive one, or the reverse. An evaluation of different implementations of simulated annealing in brachytherapy is given by Wehrmann et al. [30].
5.5.6
Optimization of volume implants
A volume implant is any implant with two or more planes. Polynomial optimization of a volume implant aims to make the dose in the volumes midway between the catheters as homogeneous as possible. This optimization cannot be based on dose points alone. If dose points are placed only outside a volume implant, the inner part of the target volume will be underdosed because the outer dwell positions will be used mainly to deliver the required dose to these dose points. This is due to the inverse square dependency of the dose on the distance and to the minimization of the overall time by the optimization procedure. It is explained in a similar way: that the periphery will be underdosed if points were placed midway between the catheters. Of course, it is possible to place dose points inside the implant and around it at a given distance. However, except for very regular implants, it is not possible to determine the relation between the doses in the inner points and in the outer ones. It is obvious, therefore, that in optimization
of volume implants additional constraints are required. These are provided by applying polynomial optimization with constraints obtained by geometric optimization. GEOMETRIC OPTIMIZATION ON AMERICAN VOLUME IMPLANTS Geometric optimization is an alternative approach to optimization on dose points. It is based solely on the dwell positions on the assumption that they represent the target volume. Originally, it was applied only to those stepping source implants for which the spacing between the dwell positions (the intracatheter spacing) is about equal to the spacing between the catheters (the intercatheter spacing), typically 1-1.5 cm [16,23]. This type of implant is performed mainly in the USA. It results in an equidistant, three-dimensional grid of dwell positions in the target volume (Figure 5.13). The basic assumption underlying geometric optimization is that the dwell time in a dwell position is inversely related to the dose given in that position by the other dwell positions [23]. Take, for example, a dwell position at the border of the target volume. It requires a larger dwell time than a dwell position in the center, because it is further away from the other dwell positions which all contribute according to the inverse square of their distance (Figure 5.13). Geometric optimization in treatment planning software is based on the following two suppositions: (1) the dwell time at a dwell position is inversely proportional to the dose delivered by the other dwell positions; and (2) the dose given by another dwell position is inversely proportional to the square of its distance. So the influence of source geometry and tissue scatter is disregarded, i.e., the factor C(ri,j, qi,j) in equation 5.10 is assumed to be constant. Thus, the dose in a given dwell position i is inversely proportional to the sum of the inverse square of the distances from that point to all
Figure 5.13 Geometric optimization (GO) of a two-plane American-type implant with six catheters. Separation between the sources in the catheters is about equal to the distance between the catheters. Calculation is midway between the planes, (a) No optimization. Note the hot spot where the catheters converge, (b) Geometric optimization.
Optimization techniques in stepping source brachytherapy 63
other dwell positions j. So the relative dwell time tt at position i becomes
(I)
with ri,j the distance between dwell positions i and j, and M the number of dwell positions. Note the following: Figure 5.14 Geometric optimization on distance. The relative
1. Using the inverse square of distances as doses in equation 5.15 is a simplification, but acceptable for iridium-192 or cobalt-60 sources. Of course, the sum of the actual dose contributions from the other dwell positions can also be used. 2. The optimized dwell time in a given dwell position is mainly determined by the dose contribution from its nearest neighbors. 3. Only relative dwell times are determined and one or more dose points are still required to define the prescription dose. An obvious advantage of this technique is its simplicity. However, this algorithm relies on good geometry of the implant itself and on the intracatheter spacing being about equal to the intercatheter spacing. When catheters converge, the algorithm tends to overcompensate by reducing the dwell times too much. Therefore, the contribution from any dwell position at a distance less than a given threshold distance is ignored. On the other hand, when the intercatheter distance is larger than the intracatheter one, the intracatheter distances dominate the dwell time calculation and the separation between the catheters is insufficiently taken into account. GEOMETRIC OPTIMIZATION ON EUROPEAN DISTANCE IMPLANTS Interstitial brachytherapy in Europe is based on continuous wires of iridium-192, usually with interwire spacing of 1.5-2 cm [3,4]. In stepping source dosimetry, these continuous wires are replaced by a source stepping in the catheters with a step length of 2.5 mm or 5 mm. Thus, in a European-type implant with intracatheter spacing of 5 mm, the intercatheter spacing is two to four times larger. If geometric optimization is used as described above, then the nearest dwell positions are always located in the same catheter unless another catheter approaches within the step length. Because the optimized dwell time in a given dwell position is mainly determined by the dose contributions from its nearest neighbors, this results in a cylindrical dose distribution around each catheter, similar to the ones obtained with polynomial optimization on distance with dose points placed laterally at 1 cm. So the geometric optimization procedure as performed on a USA volume implant results in optimizing on distance on a European implant. When applied on European-type implants, this technique is called geometric optimization on distance (Figure 5.14).
dwell time = 1/dosefrom all other dwell positions. (1) High dose contribution from both catheters. (2) High dose contribution, mainly from both neighboring dwell positions. (3) Low dose contribution, mainly from left neighbors. Thus, t1= t2 < t3. Geometric optimization on volume. The relative dwell time = 1/dosefrom dwell positions in other catheters. Dose contributions by other catheter (1) - high, (2) = medium, (3) low. Thus, t1< t2 < t3.
GEOMETRIC OPTIMIZATION ON EUROPEAN VOLUME IMPLANTS To optimize a European-type volume implant, a variation called geometric optimization on volume was developed [21]. This variation uses in equation 5.15 only the distances between the dwell position i in the current catheter k and all dwell positions j in the other catheters. If a given dwell position in a catheter is at a large distance from the ones in the other catheters, the dose contribution will be small, resulting in a large dwell time. This will substantially increase the dose in the volume between the dwell position and the other catheters (Figure 5.14). The consequences of geometrical optimization in pulsed dose-rate brachytherapy on European volume and distance implants are discussed in detail by Berns et al. [22] Although geometric optimization on volume does not require dose points, they are still required for the definition of the prescription dose. For a regular volume implant, the best position for the dose points is in the central transversal plane midway between the catheters. The prescription dose is then based on a given percentage of the mean dose in these dose points. A percentage of 90 is required for a distance of about 3 mm between the outer catheters and the prescription isodose line in the central transversal plane. A more detailed discussion about the prescription dose of an optimized implant is given in section 5.7.6. Geometric optimization alone of a European volume implant does not always suffice (Figure 5.15). It is not strong enough to keep all dwell positions inside the target volume. At the outer ends of the catheters, active dwell positions are needed on or even outside the target surface. In the next section, a combination of geometric and polynomial optimization is presented where all active dwell positions are located inside the target volume.
64 Computers in brachytherapy dosimetry
Figure 5.15 Intraluminal implant, optimized on distance, (a) Polynomial distance optimization on dose points placed at 1 cm distance in region of minimum dose, away from the other catheter. Note that the WOO cGy prescription isodose runs through all dose points, (b) Geometric optimization on distance of a European-type implant. The 1000 cGy prescription isodose is taken as the mean of the dose values in the dose points. Note that the geometric optimization undercorrects when catheters coincide.
POLYNOMIAL OPTIMIZATION ON VOLUME As previously stated, European volume implants cannot successfully be optimized based on dose points midway between the catheters alone. However, if an additional constraint, supplied by geometric optimization on European volume implants, is added to the least squares function x2 in equation 5.12, optimization on volume is achieved [21]. This two-step process is called polynomial optimization on volume. In the first step, the geometric optimization on volume is used to obtain the total relative dwell time for each catheter. The additional constraint now requires that the total relative dwell time for each catheter remains equal to the one obtained by geometric optimization. Thus, the total dwell time in each catheter, determined by geometric optimization, is redistributed by polynomial optimization in such a way that the dose points midway between the catheters all receive the same dose. Polynomial optimization on volume is an essential part of the Stepping Source Dosimetry System, which is presented in the next section.
5.6 THE STEPPING SOURCE DOSIMETRY SYSTEM Before the use of computers in brachytherapy, the Paris Dosimetry System was developed as a low dose-rate dosimetry system using afterloading of iridium-192 wires with equal linear activity into thin flexible catheters or rigid needles [3,4]. For a given target volume, the Paris System gives rules on how to implant a
target volume V = L x W x T a s a function of L, W, and T, with L the length, Wthe width, and T the thickness of the target volume. The active lengths extend outside the target to correct for the bending of the prescription isodose surface in between the catheter ends. This is because the Paris System applies wires of constant linear activity and does not place a crossing catheter at the end of the implant, as is done, for instance, in the Manchester System with a needle implant. As already described, the high dose-rate treatment of volume implants is performed with a source stepping through a set of catheters or needles. Thus, the Paris Dosimetry System can easily be adapted to high dose rate by applying equidistant dwell positions with equal dwell times. If, however, the dwell times at the longitudinal ends of the catheters are increased by polynomial optimization on volume, the active dwell positions, even at the longitudinal ends, are kept inside the target volume. This adaptation of the Paris System is called the Stepping Source Dosimetry System (SSDS) [16,31]. The SSDS uses the same rules for implantation as the Paris Dosimetry System, except that the active lengths in the catheters remain within the target surface, even at the longitudinal ends. Dose points are placed midway between the catheters over the whole length of the implant. When an implant is very regular, for example when templates are used to maintain the prescribed distances between the catheters, the first and the last dose point of each row midway between the catheters should be discarded. The SSDS applies polynomial optimization on volume to obtain the same dose in these dose points. Originally, the SSDS defined the prescription dose as 85% of the mean dose in these dose points [16,31]. As
Dose-volume histograms 65
discussed in section 5.7.6, where non-optimized and optimized implants are compared, the prescription dose is best defined as 90% of the mean dose in the dose points.
5.6.1 Summary of the Stepping Source Dosimetry System
A comparison of an isodose distribution for a breast implant using the Paris System and the SSDS is shown in Figure 5.16. The optimized dose distribution shows a more homogeneous dose distribution inside the target volume and an appreciable dose reduction outside it. A more graphical and quantitative approach for the evaluation of a dose distribution is given by its differential or natural volume dose histogram, presented next.
The following parameters are used in the SSDS: L T S M
= = = =
length of the target volume thickness of the target volume spacing between the catheters the safety margin around an implant: it is the distance between the prescription isodose and the active lengths in the outer catheters in the central transversal plane AL = active length in a catheter: it is the distance between the first and the last active dwell position inside the target volume. The SSDS implantation rules are as follows. • For a short target volume, L < 5 cm, the catheter spacing S varies between 8 mm and 15 mm; for a long volume, L > 5 cm, S varies between 15 cm and 22 cm. • For a target thickness T < 12 mm, single-plane implants are applied, with the catheter spacing S T/0.6, which gives M 0.35 S. • For a target thickness T > 12 mm, double-plane implants are used. A double-plane implant with a catheter pattern in triangles must conform to S 271.3. The safety margin Mbecomes M 0.2 S. For a double-plane implant with a catheter pattern in squares, S 271.6 and M 0.3 S. • The active dwell positions at the longitudinal ends of the catheters are placed inside the target volume using the safety margin as given above, AL = L — 2M. • The dwell position spacing is 5 mm. • The dwell times are obtained by polynomial optimization on volume, using dose points midway between the catheters along their whole active lengths. The first and the last dose point of each row midway between the catheters are usually discarded. • The prescription dose is defined as 90% of the mean dose in these dose points. As they are all optimized to the same value, this definition is equivalent to 90% of the mean basal dose points in the central transversal plane. The above rules are guidelines on how to implant a given target volume. The resulting dose distribution must be evaluated by assessing the isodose lines in several transversal and longitudinal planes. In order for the prescription isodose surface to encompass the target volume as closely as possible, dwell positions may have to be activated or deactivated and dose points to be added or removed, both near the longitudinal ends of the catheters.
5.7
DOSE-VOLUME HISTOGRAMS
Dose-volume histograms (DVHs) play an important role in evaluating the dose distribution in and around an implant [32,35,36]. A DVH of a dose distribution is represented as a graph with a series of dose intervals on its horizontal axis, and, on the vertical axis, for each dose interval, a volume related to that dose interval. Such a dose interval in a histogram is called a bin. For example, if 1000 bins are taken for a dose range of 500-2500 cGy, the first bin, DD1 will be 499-501 cGy, the second bin, DD2, will be 501-503 cGy, etc., with the bin width AD being 2 cGy. A differential DVH is a graph with dose intervals ADj on the horizontal axis and on the vertical axis, for each dose interval DDi the ratio AV/AD with DVi the volume receiving a dose between Di - 0.5 AD and Di + 0.5 AD. In a clinically useful histogram, AD is much smaller than Di. Then the volume with a dose between Di and Dj is given by the area under the histogram between Di and Dj If a volume implant is optimized to the same dose midway between the catheters, all the volumes midway between the catheters will obtain that dose and the differential DVH will show a sharp peak for that dose. Based on this behavior, a differential DVH can be used to assess the homogeneity of the dose distribution of a volume implant. A cumulative DVH has the same horizontal axis of dose intervals Di;, with bin width AD, as the differential DVH. On the vertical axis, however, is given the volume receiving at least the dose Di - 0.5 AD for each DDi. In a clinically useful histogram with AD much smaller than Di, the histogram presents for each Di the volume encompassed by the isodose surface(s) of that dose. Thus, the cumulative DVH of the target volume can be used to determine those parts of the target volume that are either underdosed or overdosed. In a clinical case, the dose distribution around an implant is so complex that a DVH can only be determined numerically, thus a discretization of the threedimensional dose distribution in and around the implant must take place. It is common practice to construct a grid of equidistant dose points inside a rectangular box, placed around the implant with a given margin. A sufficient number of grid points is placed inside the box and the dose in each one of them is
66 Computers in brachytherapy dosimetry
Figure 5.16 Two-plane breast implant with seven catheters. The length of the target volume is 8 cm. The active length, i.e., distance between first and last dwell position, is 10cm for Paris System implant, and 7 cm for SSDS implant. Step size is 5 mm. The 500 cGy lines are the prescription isodose lines determined according to the rules of the Paris System and the SSDS System, respectively, (a) Dose distribution in the central transversal plane. The solid HneZ=0 indicates the central longitudinal plane, midway between the two planes, (b) Dose distribution in the central longitudinal plane. The solid line indicates the central transversal plane, (c) Comparison of the dose distributions in three longitudinal planes of the Paris-type and the SSDS implant. PlaneZ=0is the central longitudinal plane; plane Z=-8 is the plane through needles 4, 5, 6, and 7; plane Z=+8 is the plane through needles 1,2, and 3. The upper part of each dose distribution is given by the SSDS implant, the lower part by the Paris implant.
Dose-volume histograms 67
voxels ni with a dose value between Di — 0.5 AD and Di + 0.5 AD, multiplied by the voxel volume v; thus DVi = n- v. A cumulative DVH gives for a given dose interval (Di — 0.5 AD, Di + 0.5 AD) the corresponding volume Vi, defined as the sum over all voxels with a dose equal to or exceeding Di - 0.5 AD, thus Vi = When a three-dimensional equidistant grid of points over the implant is used, a large number of grid points, between 50 000 and 200000, is needed for an accurate DVH. This is due to the application of an equidistant grid over the regular geometry of the implanted catheters, which leads to a large redundancy of grid points. It can be proven statistically that a more efficient grid over a set of regularly implanted catheters is a grid where the x, y and z coordinates of each point are determined randomly [16,33]. The voxel size is then equal to the volume of the rectangular box around the implant, divided by the number of grid points inside the box. The number of randomly placed grid points needed for an accurate DVH lies between 10 000 and 50 000. The target volume is defined by the tumor and the margin around it. The minimum peripheral dose (MPD), is defined as the dose of the isodose surface that just encompasses the target volume, thus the highest dose still encompasses the target volume. The treatment volume is defined by the prescription isodose surface which is selected by the radiation oncologist when viewing the dose distribution. The prescription dose (PD) is the dose prescribed to the prescription isodose surface. If CT or MRI images with the contours of the target volume are not available, the target volume is considered to coincide with the treatment volume and the MPD is taken to be equal to the PD.
5.7.1 Differential dose-volume histogram of a single point source
Figure 5.16 cont.
calculated. Each grid point is the center of a cubical volume element, a voxel. The whole voxel is considered to receive the same dose as the grid point. When a threedimensional equidistant grid with spacing s is applied, the voxels are cubes with edge s, which are centered around the grid points. More explicitly, a differential DVH gives, for a given dose interval (Di - 0.5DD,Di+ 0.5 DD), the corresponding ratio DV/DD, with DVi the volume with a dose value in that interval. DVi is obtained from the number of
Understanding of the properties of a DVH of a single point source is essential for the evaluation of implants with more sources. If a point source is ideal, i.e., tissue scatter and absorption can be ignored, then D = S/r 2 , with D the dose at distance r from the source. The values of the differential DVH for an ideal point source can be calculated directly [16], because the isodose surfaces are spheres with the source as center, of which the volumes are easily calculated by V= (4/3) p r3. For an ideal point source with D = S/r2, Vcan be written as a function of D:
The numerical and analytical value of (DV/DD) for D = 1000 cGy will be calculated for an ideal point source with S = 1000 cGy cm2. The numerical calculation requires the distance r at which D = 1000 cGy: r = s/D = 1 cm. The value of (AWAD) for D = 1000 cGy is found by calculating D and V for r equal to 0.99 cm and 1.01 cm. For
68 Computers in brachytherapy dosimetry
r = 0.99 cm, D099 = 1000/0.992 = 1020 cGy and V099 = (4/3) n 0.993 = (4/3) p 0.997 cm3. For r = 1.01 cm, D101 = 1000/1.012 = 980 cGy and V,01 = (4/3) p 1.013 = (4/3) p 1.03 cm3. Thus, DD = D1.01 - D099 = -40.0 cGy and DV = Voi -V0.99= 0.08 p cm3 This results in (DV/DD) = -0.002 p = -0.0063 cm3 cGy-1 for D = 1000 cGy. Analytically, the differential DVH for an ideal point source with D = S/r2 is given by, using equation 5.16: For S = 1000 cGy cm2 and D = 1000 cGy, dWdD = -2p 10003/2 1000'5/2 = -0.0063 cm3 cGy-1. See reference 16 for more details. A differential DVH of an iridium-192 point source is given in Figure 5.17. As in the above example, the dose at 1 cm given by this source is 1000 cGy. At such a short distance, the tissue scatter and absorption factor is near unity. From the dependence of dWdD on 1/D5/2, it is clear that, as the dose decreases, dV/dD increases more than quadratically.
5.7.2 Differential dose-volume histogram of multiple point sources A differential DVH of a volume implant behaves as a single point source for very low doses and for very high doses. This is illustrated in the case of an implant consisting of N identical point sources each with strength S (Figure 5.18). Points far away from the implant receive
very low doses from the sources, and essentially respond to all sources as a single source with strength N S. Correspondingly, for these distant points the histogram is dWdD = -2 7p(NSY3/2 D 5/2 . For points of very high doses, thus very near to a source, only the dose contribution of that source will be seen, due to the inverse square dependence of the dose on distance. Now the histogram shows the behavior of N times that of a single source: dV/dD = - 2 p N S3/2 D-5/2, which equals, of course, the DVH of a single point source with strength N2/3 S. For a large number of sources, this strength is much smaller than that at a large distance from the implant (Figure 5.18). Consequently, the influence on the dose distribution of the catheter placement and the dwell time optimization is reflected most strongly in the middle section of the differential DVH, which lies between about 75% and 200% of the PD which encompasses the target volume. Visual evaluation of this influence is difficult because of the underlying inverse square law influence. In Figures 5.18 and 5.19, the differential DVH of a point source located in the center of the implant is also given. The difference between the point source curve and the implant curve indicates how much better the implant is compared with a single source. The strength of this point source is defined such that the dose value of the spherical isodose surface that just fits inside the rectangular grid box equals that of the similarly fitting isodose surface given by the implant. The margin of the grid box around the implant is 1 cm. To determine the dose value of that spherical isodose surface, the maximum dose occurring on the
Figure 5.17 Differential dose-volume histogram of an iridium-192 point source which delivers WOO cGy at 7 cm in tissue. The influence of tissue scattering and absorption is hardly visible, as is indicated by the histogram value for D = 300 cGy and the theoretical value, using dV/dD = -2p S3/2 D- 5/2 for an ideal point source. A shaded area between two dose values represents the volume between the corresponding three-dimensional isodose surfaces. For a discussion of the three shaded areas of equal size, representing three equal volumes, see 'Natural dose-volume histogram,' p.69.
Dose-volume histograms 69
Figure 5.18 Differential dosevolume histogram of the twoplane breast implant of Figure 5.16, according to the Paris Dosimetry System. The lower peak is related to the volumes between the outer catheters, the higher peak to the volumes between the inner catheters. The underlying curve is the histogram of the point source in the center of the rectangular box around the implant. The strength of the point source is such that the same maximum dose on the grid surface is obtained as given by the implant. Note that the peak of this histogram is not well defined and the location of the 85% peak dose value D85 is not at the base of the peak.
rectangular box surface must be found. From that maximum dose value on the box surface, the source strength of the point source in the center of the box is calculated. This explains why both curves in Figures 5.18 and in 5.19 start with the same dV/dD value for the minimum dose of 40.0 cGy. Figures 5.18 and 5.19 present various dose values around the peak dose D100 to judge the implant. D100 is defined as the largest difference between the dose histogram of the implant and that of the ideal point source. So D100 is the dose under the peak after correction for the slope by the ideal point source. D85, D95, D105, and D115 are dose values of 85%, 95%, 105%, and 115% of D100. Note that D100 is located in the center of the implant and that the dose range where the implant differs from the point source extends only about 15% from the peak dose D100. D85 is usually taken as the PD for an implant which is not optimized. Figure 5.19 indicates that, for an optimized implant, the implant histogram value for DS5 already approaches the point source histogram. As discussed in section 5.7.6, the PD for an optimized implant should be taken as 90% of the peak dose D100. It is possible to evaluate an implant by its differential DVH. In addition, figures of merit based upon DVHs
have been developed [39]. However, to evaluate an implant implies evaluating the differences between the histogram of the implant and that of the corresponding point source in its center (see Figures 5.18 and 5.19). To address this problem, Anderson [34] in 1986 developed the concept of the natural DVH and derived the Uniformity Index (UI) as a figure of merit of the implant. Based on the natural DVH, a new figure of merit was defined, the Quality Index (QI) [22,35,36].
5*73
Natural dose-volume histogram
Because the inverse square law has such a detrimental effect on interpretations of the differential DVH, Anderson [34] introduced a new dose unit, u, for the horizontal axis: u(D) = D-3/2. Now, for an ideal point source, using equation 5.17 and dV/dw = (dV/dD)/ (dD/dw), it follows:
which is independent of dose D, thus the natural histogram of a point source is a horizontal line (see Figure
70 Computers in brachytherapy dosimetry Figure 5.19 Differential dosevolume histogram of the twoplane breast implant of Figure 5.16, optimized according to SSDS. The single high peak is caused by the optimization to the same dose of all volumes midway between the catheters. The underlying curve is again the histogram of the point source in the center of the rectangular box around the implant (see Figure 5.18). Note that the 85% peak dose value D85 is not located at the base of the peak and therefore is unsuitable as prescription dose.
5.20). Note that the horizontal axis is linear in u, thus linear in D-3/2. Also, that for increasing dose, u decreases. Thus, when expressed in dose D, the low dose section of this axis is expanded and the high dose section is compressed. The area under the curve between Dl and D2 is proportional to the volume between the Dl isodose surface and the D2 isodose surface. Note the difference in appearance of three equally sized volumes in the differential DVH of Figure 5.17 and the natural DVH of Figure 5.20. The natural DVHs of the unoptimized and optimized two-plane breast implants discussed in section 5.6 are given in Figures 5.21 and 5.22. The peak of the histogram represents the volumes, midway between the catheters, that receive an approximately uniform dose. The narrower the peak, the more uniform the doses in the volumes between the catheters. The narrowest peak is obtained by an equidistant three-dimensional grid of dwell positions over the target volume with the dwell times optimized to the same dose in dose points between the dwell positions. The broader the peak, the less desirable the implant, until finally a horizontal line remains which represents a single point source in the center of
Figure 5.20 Natural dose-volume histogram of an iridium-192 source which delivers 1000 cGy at 1 cm in tissue. The - sign of dV/du is disregarded.
indicates the actual histogram.
indicates the theoretical value of 132.5 using equation 5.19. The three equal volumes of Figure 5.17 now show up as areas with equal base, indicating the compression of the dose axis for high dose values and the expansion for low dose values.
Dose-volume histograms 71
Figure 5.21 Natural dose-volume histogram of the two-plane breast implant of Figure 5.16, according to the Paris Dosimetry System. (See legend of Figure 5.18 for explanation of the occurrence of two peaks.) Line (3) represents the theoretical limit, (4/3) p (nS)3/2 of dV/du at a large distance from the implant, with n the number of dwell positions and S = D(r) r2 the source strength of the ideal point source. Line (2) lies midway between lines (1) and (3) and defines the so-called 'low dose.' Line (5) represents the theoretical limit (4/3) p p S3/2 of dV/du for very high dose values. Line (4) lies midway between line (1) and (5) and defines the so-called 'high dose.' The prescription dose (PD) is defined as 85% of the mean dose in the basal dose points. See p. 72 for explanation of the Uniformity Index and the Quality Index.
Figure 5.22 Natural dose-volume histogram of the two-plane breast implant of Figure 5.16, optimized according to the SSDS System. The prescription dose (PD) is defined as 90% of the mean dose in the basal dose points, i.e., 90% of the peak dose value. Note the relatively small change in the Uniformity Index compared to the unoptimized case (Paris System) in Figure 5.21 (2.26 versus 1.98). The horizontal tails at the left side (very low dose values) and the right side (very high dose values) are explained on p. 71. The prescription dose coincides with the natural prescription dose (NPD), which lies at the base of the peak at the LD side (see p. 74).
the target volume. It should be noted that only volume implants will show a peak in their natural histogram. As discussed in the section 'Differential DVHs of multiple point sources', the histogram behaves for very low
doses as if a single point source exists, and the natural DVH will thus display a horizontal line for these values. Also, for very high doses, the histogram behaves as if a point source, although with much less strength, exists
72 Computers in brachytherapy dosimetry
(Figure 5.22). Because of the strong contraction of the w-axis for high doses, the horizontal line at the high dose end is sometimes hardly visible and appears as if it runs straight down to this high dose limit value of dV/du (Figure 5.21). Anderson derived the Uniformity Index by taking the ratio of the volume under the peak, normalized to the w-scale, and the volume encompassed by the isodose surface of the target dose, again normalized to the w-scale. A related Figure of Merit is the Quality Index, which is independent of the target dose [22,35,36]. The quantities PD, LD, and HD are used in these indices. PD refers to the target dose prescribed by the radiation oncologist. Usually, it belongs to the isodose surface which encompasses the target volume with a margin of about 3-5 mm. LD refers to the dose value at the half height of the peak in the low dose region. This half height is measured from the limit value of the histogram for dose values approaching 0. HD refers to the dose value at the half height of the peak in the high dose region. This half height is measured from the limit value of the histograms for doses approaching infinity. UNIFORMITY INDEX The Uniformity Index is a quantitative index to assess how well the dose distribution covers the target volume. It is defined as the volume between the dose values of PD and HD, normalized to the w-interval between PD and HD, divided by the volume encompassed by PD, normalized to the w-interval between PD and infinity dose. Thus, the Uniformity Index is:
with V(PD) the volume encompassed by the prescription isodose surface, V(HD) the volume encompassed by the high dose isodose surface, w(PD) the w-value corresponding to the prescription dose, and w(HD) the w-value corresponding to the HD value. By substituting u = D-3/2 and using u( ) = 0, we get
The first term of the UI is the volume under the peak extended to the target dose, divided by the width of the peak extended to the target dose. The narrower the peak, the larger the first term will be. The effect of the second term is the opposite. For a single point source, there is no peak and the UI = 1. As already mentioned, the UI is dependent upon the PD chosen by the radiation oncologist. It is a measure of the quality of the dose distribution within the selected target dose. If a perfectly regular implant is not covering the target volume completely and therefore a lower PD must be selected, this is reflected by a lower value of the UI.
QUALITY INDEX To compare different geometries of implantation and different dwell time optimization schemes, another Figure of Merit is needed, which is independent of the PD. For this purpose, the Quality Index is introduced [ 16,39]. In the Quality Index, LD is substituted for PD in equation 5.20. Thus:
A detailed study of the differences between UI and QI in geometrically optimized breast implants is given in reference 22.
5.7.4
Cumulative dose-volume histogram
The cumulative dose-volume histogram (CDVH) presents for each dose value the volume encompassed by the isodose surface(s) of that dose [51,52]. It is widely used to determine if a part of the target volume is underdosed or if an organ at risk is overdosed. The CDVH of the target volume shows a distinct behavior (Figure 5.23). For dose values lower than the minimum peripheral dose, the complete target volume is covered and the CDVH runs horizontal. When the dose value exceeds the minimum peripheral dose, part of the target volume is not included by the isodose surface with that dose value, so the CDVH runs steeply downward with increasing dose. For high dose values, only the small volumes directly around the sources inside the target contribute and the CDVH value decreases slowly.
5.7.5 Evaluation of dose distributions with dose-volume histograms DVHs play an important role in evaluating the following aspects of dose distributions in brachytherapy [32]. • How homogeneous is the dose distribution of a volume implant? This can be assessed independently of the actual PD, either by visual inspection or by the QI of the natural DVH. Note that it is not possible to evaluate the homogeneity of a distance implant, because of the steep gradients around the catheters. The differential and natural DVHs assess the regularity of the catheters and the optimization of the dwell times, irrespective of the coverage of the target volume by the implant. • How well is the dose distribution covering the target volume? This depends on the PD selected. It can be assessed again by visual inspection. It is obtained from the CDVH of the target volume by looking at the amount of the target volume which is underdosed by a dose less than the PD. The cumulative DVH is also used to determine the amount of volume being overdosed by, for example, a dose greater than 2 x PD.
Dose-volume histograms 73
Figure 5.23 Cumulative dose-volume histogram of the target volume of a non-optimized prostate implant. On the horizontal axis is given the dose relative to the prescription dose (PD). On the vertical axis is given the encompassed volume for each dose value, relative to the target volume V. The minimum peripheral dose (MPD) is the highest dose still encompassing the target volume. If the PD was taken to be equal to the natural prescription dose (NPD), it is evident that the implant is not covering 10% of the target volume. Additional dwell positions must be activated in the missed volume (see p. 74). The dose/non-uniformity ratio for a non-optimized implant, defined as DNR(D) = V(1.5 D)N(D), will show a minimum for the PD in this histogram as the curve slope is the steepest in the dose range PD-1.5 PD (see p. 74).
The UI of the natural DVH, which is based on the PD, scores the combination of aspects (1) and (2) only if the PD equals the minimum peripheral dose of the target. Thus, much more detailed information is obtained when the dose homogeneity is evaluated with the QI of the natural DVH and the coverage of the target volume with the CDVH. • How much volume outside the target volume receives a high dose? If CT or MRI images of an organ at risk are available, the volume which receives a dose exceeding the maximum dose allowed in that organ can be determined. Then, a grid of dose points must be placed over the organ at risk and the CDVH of that organ be determined. Similarly, the difference between the treatment volume and the target volume can be assessed by the CDVH of the treatment volume. The last two aspects are strongly related to the value of the PD. In a volume implant with an optimized dose distribution, dwell time has been moved from the center of the implant to its periphery. This results in a steeper dose gradient around an optimized implant, which influences the definition of the PD and the treated margin around the outer dwell positions.
5.7.6 Definition of the prescription dose in non-optimized and optimized volume implants A non-optimized Paris-type dose distribution displays a low dose gradient which starts from the center of the
implant. The 85% of the mean basal dose in the central transversal plane coincides more or less with the minimum peripheral isodose, i.e., the isodose surface with the highest value, which still encompasses the complete implant. In the natural DVH of the unoptimized breast implant in Figure 5.21, the prescription dose PD lies about halfway on the left side of the peak. An SSDS implant shows a much more homogeneous dose distribution inside the implant, which corresponds to the pronounced peak in the natural DVH of Figure 5.22, and a steep inverse-square-law dose gradient around it, which corresponds to the horizontal histogram curves away from the peak (see, again, Figure 5.22). The mean basal dose is practically equal to the mean dose in all dose points midway between the catheters, because these points are all optimized to the same value. Figure 5.24 shows that the isodose surface with a dose value of 90% of the mean basal dose coincides with the highest dose value still lying in the steep dose gradient area around the outer needles of an implant, and maintains a margin of a few millimeters around the implant. Therefore, the PD of an SSDS implant is taken as 90% of the mean basal dose. This PD isodose surface shows a smaller margin around the outer dwell positions, due to the steeper dose gradient around an optimized implant and the higher percentage of the basal dose (see Figure 5.24). This margin is about 3 mm, whereas the treated margin around a non-optimized implant is about 5 mm. This definition defines a PD which for an optimized implant lies at the base of the peak of the natural DVH, at the LD side. To define the PD as the dose value at the base of the
74 Computers in brachytherapy dosimetry
Figu re 5.24 The definition of the prescription dose for an implant, according to the Stepping Source Dosimetry System. The 90% value of the mean basal dose (BD) in the central transversal plane is the highest percentage still lying in the steep dose gradient area around the implant, thus just at the base of the peak in Figure 5.22. The treated margin around the outer sources is about 3 mm. Note that the 95% isodose line shows deep bends between the catheter intersections.
peak of the natural DVH is generally valid for all volume implants. It is the value of the isodose surface inside which the optimized dose distribution lies and outside which the inverse square law predominates. This definition of PD is called the natural prescription dose (NPD). Note that this PD is defined on the dose distribution only. The PD that covers the target volume is the minimum peripheral dose (MPD). If the NPD is equal to the MPD, the implant is well placed over the target volume and the PD can be taken as equal to the MPD. The UI now correctly scores both aspects of the dose distribution, the dose homogeneity and the target coverage. If, in the natural DVH in Figure 5.22, the PD is taken with a value lower than 90% of the mean basal dose, the PD will shift more to the left on the horizontal tail for low dose values. This implies that the PD is taken at a distance so far away from the implant that the inverse square law dominates. The UI will decrease correspondingly. Thus, according to the dose distribution, the PD should coincide with the NPD at the base of the peak in the natural DVH (Figure 5.22). According to the target volume, the PD should coincide with the MPD, the highest dose value still encompassing the target volume. If the PD does not match these two requirements, the dose distribution does not cover the target volume adequately. This matching is evaluated by the natural dose ratio (NDR), which is defined as the ratio of the NPD and the PD:
In clinical practice, the PD is often taken as equal to the MPD. If the target volume is suitably covered by the implant, the NDR has a value nearly equal to one. If the target volume is not suitably covered by the implant, cold spots in the target volume will arise, around which the MPD runs. Or, stated differently, the volume encompassed by the NPD covers the target volume only partly. Thus, an ill-covered target volume prescribed to the same MPD as a well-covered target volume will receive an overall much higher dose. This translates into values of NDR greater than one. NDR equal to 1.4 means that the base of the peak of the natural DVH starts at 1.4 PD and the whole target will receive a 40% higher dose than when treated to the same PD but with an NDR equal to 1.0. The use of NPD, MPD, and NDR is of utmost importance for implants of the prostate [61]. If the NDR is larger than 1 or, stated differently, if the required prescription dose does not lie at the base of the peak of the natural DVH, a part of the target volume is not covered by dwell positions (or by seeds, in the case of permanent implants). All transverse slices should then be inspected visually for target areas with a dose lower than the NPD, and dwell positions in these areas should be activated (or seeds be placed).
5.7.7 Other methods for evaluation of dose distributions Another assessment of implant quality, utilizing only the CDVH, is given by the dose-non-uniformity ratio (DNR) [53,54] for the homogeneity of the dose distribution and the coverage index (CI) for the target coverage [38]. The DNR ratio is a graph for each dose value of the ratio of the volume receiving a dose larger by a given fraction, say 50%, to the volume enclosed by that given dose value. Thus, for a given dose value, D:
The behavior of the DNR versus dose plot can be correlated with implant quality in the target volume. For non-optimized implants, the fraction is usually taken as 50%. In Figure 5.23, it is indicated that the minimum value for DNR(D) is defined by the 50% dose range where the volume curve has the steepest slope and thus the difference between V(D) and V(1.5 D) is maximal. The steeper the CDVH curve between D and 1.5 D, the better the dose distribution and the smaller DNR(D) will be. However, the relation between the minimum value of DNR and the MPD is defined by the shape of the CDVH, and there is also no direct relation with the NPD, as the latter is based on the implant dose distribution and not on the target volume. DNR(PD) is also known as the Dose Homogeneity Index (DHI) [37].
Three-dimensional imaging techniques 75
For optimized implants, it is better to take the fraction as 25% [55]. This is explained as follows. The differential histogram in Figure 5.19 shows that the dose range where the optimized implant differs from a single point source in the center of the target is only ±15% of the peak dose. As discussed in section 5.7.6, the PD of an optimized implant is 90% of the peak dose; thus, the dose range PD-1.25 PD just covers the volume under the peak. From this follows that the DNR value will be smaller for dose values other than 90% of the peak dose, and the plot of DNR(D) versus D will show a minimum at D equal to 90% of the peak dose. (For more details, see reference 55.) When adapting the fraction to the amount of optimization performed, the minimum value of DNR(D) will not distinguish for different types of optimization. Depending on the regularity of the implant and the type of optimization performed, the differential (and natural) DVH will show different peak widths and, as a result, different fractions are required in relation (22), ranging from 25% for a regular, fully optimized implant to 50% for a non-optimized implant. A detailed discussion is given by Low and Williamson [55], in whose article different optimization schemes were applied to different implant types but a fixed fraction of 25% was taken. The Coverage Index (CI) scores the coverage of the target volume by the PD. It is defined as:
In the CVDH of the target volume, V(target) is equal to V(MPD). If PD is based on the minimum value of the DNR(D) plot, then CI will be less than 1, about 0.9 for a non-optimized implant [55] (see Figure 5.23). Thus, also the combination of the DNR for the dose distribution inside the target volume and the CI for the target coverage are not suitable, easy to apply, evaluation tools for HDR implants. Another index, based on the CDVHs of the target and critical structures, is the Conformal Index (COIN), which scores the target coverage together with the unwanted irradiation of normal tissues and parts or all of critical structures. (See reference 38 for an extensive discussion.)
5.8 THREE-DIMENSIONAL IMAGING TECHNIQUES CT-based brachymerapy treatment planning is based on the visualization of the tumor, the target volume, the catheters, and the surrounding anatomical structures in a series of two-dimensional CT slices or in three-dimensional views reconstructed from these CT slices
[40-3,51]. Unfortunately, compared to catheter and anatomical point reconstruction from two radiographs, the localization of catheters or individual source positions using their intersections with CT slices is more difficult. The entry of anatomical structures into the planning system is achieved either by digitizing these structures from a hard copy of the CT slices on a digitizer tablet or by delineating these structures by mouse on the computer display. Usually, only the target and critical structures are entered, because the other internal structures are of no clinical significance and are not taken into account by the dose calculation routines in brachymerapy. Reconstruction of source locations from CT images poses several problems, especially if the spacing between the slices is large. When an X-ray ruler is used to reconstruct a catheter, markers may fall between the slices and a high resolution scout view is then required to indicate which marker is visible in a given slice. For that reason, X-ray rulers are often used which consist only of a single radio-opaque wire. If the tip of such a marker wire falls between two slices, the intersections of the marker wire in the first two CT slices must be used to extrapolate the marker wire over the distance between the first slice and the tip of the marker wire. The reconstruction of a looping catheter requires the differentiation in several slices between the images of the two sections of the loop. In general, in order to get the best reconstruction of a looping catheter, a small interslice distance is required, resulting in the necessity of handling a large number of CT slices. Dwell positions in these needles are calculated by the planning system, using the distance of the first dwell position from the tip and the dwell step. If the patient can be reproducibly positioned on the treatment table of both a CT scanner and a treatment simulator, the localization of the target volume and critical organs can be obtained from the CT scans and the localization of the catheters from two radiographs. The intersection of the catheters reconstructed from radiographs can be displayed in the CT slices. In this way, any change in the patient localization between the CT session and the simulator session can be visually detected and corrected interactively. This method is of interest for HDR implants of the prostate. Ultrasound imaging and three-dimensional treatment planning tools have been used for evaluation of HDR prostate implants [51]. The dosimetric quality indicators are the CI, the MPD, and the HI, which is the fraction of the target volume receiving doses in the range of 1.0-1.5 times the PD. The combination of different imaging modalities in the treatment of prostate carcinoma with HDR iridium192 afterloading is becoming common practice [46]. The flexible HDR catheters are inserted under guidance of ultrasound imaging, but the treatment planning is per-
76 Computers in brachytherapy dosimetry
formed on the postneedle-placement CT images. The target MPD is optimized to conform to the prostate's peripheral shape as it changes from base to apex. The urethra's dose is limited to 120% of the MPD. MRI has excellent soft tissue imaging capabilities, but the alinearities in the image prohibit direct utilization in brachytherapy treatment planning. To get around this problem, the reconstruction of prostate implant catheters can be performed with radiography and the imaging of the prostate and the implant can be done with MRI [43]. In this paper, permanent seeds are implanted, but the technique is equally valid for HDR implants with catheters with X-ray rulers inserted for radiography and MR rulers for MRI. With radiography, the seed coordinates are reconstructed in the conventional manner from a pair of isocentric radiographs. The reconstructed seed distributions are verified on the AP and lateral radiographs. With MR scanning, a flat table top is used to allow reproducible patient positioning between MR scanning and radiography. After scaling of the MR dataset to the real-size seed distribution, corresponding seeds in the data set of reconstructed seed positions and signal voids in the MR images are interactively identified. These corresponding seeds are distributed over a cranial, central, and caudal transverse slice through the prostate. A total of about ten corresponding seeds is then used for matching with the corresponding signal voids in the MR images. The resulting rotation and translation to the MR images is then applied to all reconstructed seeds. Similarly, instead of radiography, CT imaging can be used for seed reconstruction, and image fusion is used for matching the MR images with the CT ones [44]. Summarizing, three-dimensional imaging of a brachytherapy implant is a valuable tool for reconstruction of target and organs, and for assessment of the resulting dose distribution with the differential or natural DVH. The CDVHs of the target and of the critical organs show quantitatively the part of the target volume which is underdosed, and the amount of the critical organ volumes which are overdosed. CT imaging can also be used to reconstruct the localization of the catheters or even sources, but only if proper software is available. Finally, the dose distribution in transaxial planes and even arbitrary user-defined planes can be displayed together with the patient structures, interpolated between the CT slices.
5.9 THREE-DIMENSIONAL DOSE CALCULATION ALGORITHMS Three-dimensional dose algorithms which incorporate the influence of tissue inhomogeneities and metal shields on the dose distribution around brachytherapy sources are just emerging. Promising approaches are the
Monte Carlo simulation of the radiation transport equation [47-50] and the convolution algorithms which are based on a scatter dose kernel calculated by Monte Carlo simulation [59]. An analytical approach is given by Daskalov et al [60]. The Monte Carlo aided three-dimensional dose calculations have become so accurate that they function as the standard against which other algorithms are compared. For example, recent data for the dose calculation around high dose-rate and pulsed dose-rate sources according to the AAPM TG43 formalism are based on Monte Carlo calculations [9,10,11]. The convolution algorithms are suitable for dose computations in heterogeneous geometries. They are similar in approach to the pencil beam modeling of external beams, but are much more complex. Current treatment planning computers are not yet fast enough to use these algorithms for clinical treatment planning. An overview of these algorithms is given by Williamson [59]. 5.10 DOSE SUMMATION OF BRACHYTHERAPY AND EXTERNAL-BEAM DOSE DISTRIBUTIONS In order for doses from brachytherapy and externalbeam treatments to be combined, both dose distributions must be based on the same patient localization in space. If the external beam and the brachytherapy treatment planning are both based on the same set of CT slices, the overlaying is straightforward. If this is not the case, common reference points in space must be defined for both modalities, or the brachytherapy patient orientation must be matched interactively to the external beam one. Once the matching is obtained, the proper spatial transformations can be performed and the summation of the doses simply becomes an additive process. Weighting factors based upon the biological effectiveness of each treatment modality must be applied. These weighting factors may be based upon radiobiological models, such as the linear-quadratic (LQ) model [56]. This model can be used for high dose-rate, low doserate, and pulsed dose-rate treatments [57,58]. The resulting dose distributions can be evaluated with the viewing techniques available with current treatment planning systems. Three-dimensional isodose surfaces can be viewed together with the target and critical anatomic structures [34]. For ease of viewing, the degree of transparency and color of each of these isodose surfaces can be adjusted. 5.11 RECENT DEVELOPMENTS IN BRACHYTHERAPY With all the computing power, display techniques, and optimization methods currently available, the degree to
Recent developments in brachytherapy 77
which a brachytherapy implant will be effective is determined not by how well the implant is optimized, but by how well the physician has physically placed the catheters or applicators. Simply stated, optimization software cannot provide a good dose distribution around a badly placed implant. Therefore, to assist in the correct implantation of the catheters or applicator, visualization of the target volume, the critical structures, and the catheters themselves is often essential. When a CT scanner with a gantry tilt option is available, a regular volume implant, such as a brain implant using a template and needles, can be performed interactively, with each row of needles fully displayed in a CT plane. Filmless planning is the integration of the X-ray or ultrasound imaging and the treatment planning process. In such an integrated brachytherapy unit, an isocentric radiographic localizer (such as a treatment simulator) with digital imaging capabilities is directly interfaced with the treatment planning computer. In the surgical theater, the localizer transfers on-line the image information to the treatment planning computer for reconstruction, optimization, and display of the dose distribution. However, if these digital images are obtained from an image intensifier, they must be corrected for the image distortion due to the curved surface of the image intensifier screen, the influence of the earth magnetic field, and any imperfections of the electro-optical system. As the required accuracy in reconstructing the catheters or sources is in the range of 0.5 mm, the distortion of a digital image must be corrected to the same extent. Such an integrated brachytherapy system, which corrects these image distortions to the required accuracy, is currently already available [45]. This allows an interactive assessment of a needle position during implantation, especially if, in the future, the localizer is also provided with a CT option for visualization of the target volume and the surrounding structures. Three-dimensional ultrasound-guided perineal implantation of the prostate, combined with real-time treatment planning using ultrasound images, is also becoming available [51]. Catheters, guided by a perineal template, are inserted into the prostate and connected to an HDR afterloader. The three-dimensional ultrasound unit provides real-time transverse and longitudinal slices through the prostate and its immediate surroundings. A longitudinal plane through the prostate and its surroundings can be oriented such that a catheter being inserted can be imaged real-time in that ultrasound plane. The ultrasound images with the catheters depicted are directly available to the planning computer, which is an integral part of the implantation system. The treatment planning process consists of three distinct stages, the preplanning, the live planning and the postplanning. In the preplanning stage, transverse ultrasound
images are used to determine the catheters to be placed through the template. Longitudinal ultrasound images are used to determine the depth of insertion. After the virtual catheters have been placed, the virtual dwell positions inside the target volume are activated by the planning system. Geometric optimization on volume is currently available to obtain a real-time optimized dose distribution. Dwell positions and dwell weights can be modified interactively to obtain the required dose sparing of the urethra, e.g., to 120% of the MPD. In the live planning stage, the virtual catheters are replaced one by one by the real inserted catheter. The real dwell positions are activated by the planning system and the dose distribution from the catheters already placed, and the virtual catheter left is displayed real time. In this way, deviations of the catheters from the planned position and the corresponding deviation from the planned dose distribution can be corrected with the catheters still to be placed. Finally, when all catheters have been placed and the dose distribution has been decided upon, the actual treatment can start. Postplanning is done one or more days after the actual treatment has started, to check the stability of the catheter positions in the target volume. A full, real-time optimization method for volume implants is still to be developed. Such a method should combine the ease of the geometrical method (no dose points to be placed midway between the catheters) with the full optimization quality of the SSDS and should allow one or more critical volumes inside and outside the target organ to be treated to a predefined value of the PD. This is important for the HDR treatment of the prostate, where the urethra is to be treated to a given fraction of the PD and where the rectal wall adjacent to the prostate is to be spared. Only when such a fast volume optimization is available, will realtime full optimization of HDR volume implants become possible. Considerable research is still to be conducted to develop a bioeffect dose model that can be applied clinically. Computer software to implement such a model in a computer planning system is readily available and will allow the display of radiobiological isoeffect distributions instead of physical isodose distributions.
REFERENCES 1. Paterson, R. (1963) The Treatment of Malignant Disease by Radiotherapy, 2nd edn., Rome, London, Butler & Tanner Ltd.
2. Meredith, W.J. (ed.) (1967) Radiation Dosage, the Manchester System, 2nd edn, Edinburgh, Livingstone. 3. Dutreix.A. and Marinello, G. (1987) The Paris System. In
78 Computers in brachytherapy dosimetry
Modern Brachytherapy, ed. B. Pierquin, J.F. Wilson and D.J. Chassagne, New York, Masson, 25-42.
Brachytherapy HDR and LDR, Proceedings Brachytherapy Meeting Remote Afterloading: State of the Art, ed. A.A.
4. Dutreix, A., Marinello, G. and Wambersie, A. (1982)
Martinez, C.G. Orton and R.F. Mould. Columbia,
Dosimetrieen Curietherapie, Masson, Paris.
Nucletron, 169-83.
5. Quimby, E.H. and Castro, V. (1953) The calculation dosage in interstitial radium therapy./Uw.y. Roentgenol., 70, 739-49. 6. Godden, T.J. (1988) Theory of sealed-source dosimetry. In Physical Aspects of Brachytherapy, Medical Physics
20. Van der Laarse, R. (1989) New developments in computerized brachytherapy treatment planning. In Changes in Brachytherapy, ed. K. Rotte and J. Kiffer. Neurenberg, Wachholtz, 57-71. 21. Van der Laarse, R. and Prins,T.P.E. (1994) Introduction to
Handbooks 19. Bristol, Philadelphia, Adam Hilger,
HDR brachytherapy optimization. In Brachytherapy from
41-75.
Radium to Optimization, ed. R.F. Mould, J.J. Battermann, A.A. Martinez etal. Veenendaal, The Netherlands,
7. Nath, R., Anderson, LA., Luxton, G., etal. (1995)
Nucletron International B.V., 331-51.
Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy
22.
Berns, C., Fritz, P., Hensley, F.W. and Wannenmacher, M. (1997) Consequences of optimization in PDR
Committee Task Group No. 43. Med. Phys., 22(2), 209-34. 8. Meisberger, LL, Keller, R. and Shalek, R.J. (1968) The effective attenuation in water of the gamma rays of 198Au, 192
lr, 137Cs, 226Ra and "to. Radiology, 90,953-7.
brachytherapy-is a routine geometrical optimization recommendable?int.y.Radial Oncol. Biol. Phys., 37, 1171-80. 23. Edmundson, G.K. (1994) Volume optimization: an American viewpoint. In Brachytherapy from Radium to
9. Williamson, J.F. and Li, Z (1995) Monte Carlo aided dosimetry of the microselectron pulsed and high dose-
Optimization, ed. R.F. Mould, J.J. Battermann, A.A.
rate 192lr sources. Med. Phys., 22(6), 809-19.
Martineze?al. Veenendaal, The Netherlands, Nucletron
10. Daskalov, G.M., Loffler, E. and Williamson, J.F. (1998) Monte Carlo-aided dosimetry of a new high dose-rate brachytherapy source. Med. Phys., 25,2200-8.
Numerical Recipes in C: the Art of Scientific Computing.
11. Wang, R. and Sloboda, R.S. (1998) Monte Carlo dosimetry of the VariSource high dose rate 192lr source. Med. Phys., 25,415-23. 12. Shalek, R.J. and Stovall, M. (1969) Dosimetry in implant therapy. In Radiation Dosimetry, Vol. Ill, ed. F.H. Attix and W.C. Roesch. New York, Academic Press, 743-807. 13. Breitman, K.E. (1974) Dose-rate tables for clinical sources, sheathed in platinum. Br.J. Radiol., 47,
International B.V., 314-18. 24. Press, W.H., Teukolsky, S.A., Vetterling, W.T. etal. (1992)
137
Cs
657-64.
New York, Cambridge University Press. 25. Chvatal, V. (1983) Linear Programming, New York, W.H. Freeman and Company. 26. Renner, D.W., Connor, T.P. and Bermudez, N.M. (1990) An algorithm for generation of implant plans for high-doserate irradiators. Med. Phys., 17,35-40. 27. Metropolis, N., Rosenbluth, M., Teller, A. etal. (1953) Equation of state calculations by fast computing machines,J Chem. Phys., 21,1087-92.
14. Nath, R., Meigooni, A.S., Muench, P. etal. (1993)
28. Sloboda, R.S. (1992) Optimization of brachytherapy dose
Anisotropy functions for 103Pd, 125I, and 192lr insterstitial brachytherapy sources. Med. Phys., 20,1465-73. 15. Baltus, D., Kramer, R. and Loeffler, E. (1992)
distributions by simulated annealing. Med. Phys., 19, 955-64. 29. Sloboda, R.S., Pearcy, R.G. and Gillian, S.J. (1993)
Measurements of the anisotropy of the new HDR iridium192 source for the MicroSelectron-HDR. In International Brachytherapy: Program me and Abstracts. 7th International Brachytherapy Working Conference, Baltimore. Veenendaal, The Netherlands, Nucletron International B.V., 290-306. 16. Thomadsen, B.R., Houdek, P.V., Van der Laarse, R. etal. (1994) Treatment planning and optimization. In High Dose Rate Brachytherapy: a Textbook, ed. S. Nag. Armonk, New York, Futura Publishing Company, Inc., 79-145. 17. Godden, T.J. (1988) Practical aspects of absorbed dose calculations. In Physical Aspects of Brachytherapy, Medical Physics Handbooks 19. Bristol, Philadelphia, Adam Hilger, 232-60.
Optimized low dose rate pellet configuration for intracavitary brachytherapy. Int.J. Radiat. Oncol. Biol. Phys., 26,499-511. 30. Wehrmann, R., Kneschaurek, P. and Hugo, C. (1994) Simulated annealing and optimization in brachytherapy. In Proceedings of theXlth International Conference on the use of Computers in Radiation Therapy, ed. A.R. Hounsell, J.M. Wilkinson and PC. Williams. Manchester, Christie Hospital NHSTrust, and Madison, Medical Physics Publishing, 34-5. 31. Van der Laarse, R. (1994) The Stepping Source Dosimetry System as an extension of the Paris System. In Proceedings of theXlth International Conference on the use of Computers in Radiation Therapy, ed. A.R. Hounsell, J.M.
18. Loeffler, E.L (1989) Quality control in brachytherapy. In
Wilkinson and P.C. Williams, Manchester, Christie
Changes in Brachytherapy, ed. K. Rotte and J. Kiffer, Neurenberg, Wachholtz, 71-84.
Hospital NHS Trust, and Madison, Medical Physics
19. Van der Laarse, R. and De Boer, R.W. (1990) Computerized high dose rate brachytherapy treatment planning. In
Publishing, 32-3. 32. Chen, G.T.Y. (1988) Dose volume histograms in treatment planning. Int.J. Radiat. Oncol. Biol. Phys., 14,1319-20.
References 79 33. Niemierko, A. and Goitein, M. (1990) Random sampling for evaluating treatment plans. Med. Phys., 17,753-62. 34. Anderson, LL (1986) A 'natural' volume-dose histogram for brachytherapy. Med. Phys., 13,899-903. 35. Van't Riet, A., Te Loo, H.J., Ypma, A.F.G.V.M. etal. (1992) Ultrasonically guided transperineal seed implantation of the prostate: modification of the technique and qualitative assessment of implants. Int.J. Radial Oncol. Biol.Phys., 24, 555-8. 36. Van't Riet, A., Te Loo, H.J., Mak, A.C.A., Veen, R.E. and Ypma, A.F.G.V.M. (1993) Evaluation of brachytherapy implants using the 'natural' volume-dose histogram. Radiother. Oncol., 26,82^1. 37. Yu, Y, Waterman, P.M., Suntharalingam, N. and Schulsinger, A. (1996) Limitations of the minimum peripheral dose as a parameter for dose specification in permanent 125I prostate implants. IntJ. Radial Oncol. Biol. Pfcys.,34,717-25. 38. Baltas, D., Kolotas, C, Geramani, K. etal. (1998) A Conformal Index (COIN) to evaluate implant quality and dose specification in brachytherapy. Int.J. Radial Oncol. Biol. Pfcys.,40,515-24, 39. Van der Laarse, R. and Prins, T.P.E. (1994) Comparing the Stepping Source Dosimetry System and the Paris System using volume-dose histograms of breast implants. In Brachytherapy from Radium to Optimization, ed. R.F. Mould, J.J. Battermann, A.A. Martinez etal. Veenendaal, The Netherlands, Nucletron International B.V., 352-72. 40. Hilaris, B.S., Tenner, M., Moorthy, C. etal. (1994) Threedimensional brachytherapy treatment planning at New York Medical College. In Brachytherapy from Radium to Optimization, ed. R.F. Mould, J.J. Battermann, A.A. Martinez etal. Veenendaal, The Netherlands, Nucletron International B.V., 307-13. 41. Kolotas, C., Birn, G., Baltas, D., Rogge, B., Ulrich, P. and Zamboglou, N. (1999) CT guided interstitial high dose rate brachytherapy. Br.J. Radiol., 72,805-8. 42. Martin,T., Kolotas, C., Dannenberg,T. etal. (1999) New interstitial HDR brachytherapy technique for prostate cancer: CT based 3D planning after transrectal implantation. Radiother. Oncol., 52,257-60. 43. Moerland, M.A., Wijrdeman, H.K., Beersma, R., Bakker, C.J.G. and Battermann, J.J. (1997) Evaluation of permanent 1-125 prostate implants using radiography and magnetic resonance imaging. Int.J. Radial Oncol. Biol. Phys., 37, 927-33. 44. Amdur R.J., Gladstone, D., Leopold, K.A. and Harris, R.D. (1999) Prostate seed implant quality assessment using MR and CT image fusion. Int.J. Radial Oncol. Biol. Phys., 43, 67-72. 45. Kolkman-Deurloo, I.K.K., Visser, A.G., Idzes, M.H.M. and Levendag, P.C. (1997) Reconstruction accuracy of a dedicated localiser for filmless planning in intraoperative brachytherapy. Radiother. Oncol., 44, 73-81. 46. Mate, P.M., Gottesman, J.E., Hatton, J., Gribble, M. and
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
Van Hollebeke, L (1998) High dose-rate afterloading 192 lridium prostate brachytherapy: feasibility report. Int.J. Radial Oncol. Biol. Phys., 41, 525-33. Williamson, J.F. (1990) Dose calculations about shielded gynecological colpostats. Int.J. Radial Oncol. Biol. Phys., 19,167-78. Kirov, A.S., Williamson, J.F., Meigooni, A.S. and Zhu, Y. (1996) Measurement and calculation of heterogeneity factors for an lr-192 high dose-rate brachytherapy source behind tungsten alloy and steel shield. Med. Phys., 23, 911-19. Weeks, K.J. (1998) Monte Carlo dose calculations for a new ovoid shield system for carcinoma of the uterine cervix. Med. Phys., 25,2288-92. Watanabe, Y., Roy,J., Harrington, P.J.and Anderson, L.L. (1998) Experimental and Monte Carlo dosimetry of the Henschke applicator for high dose-rate192lr remote afterloading. Med. Phys., 25,736-45. Kini, V.R., Edmundson,G.K., Vicini, FA, Jaffray, DA, Gustafson, G. and Martinez, A.A. (1998) Use of threedimensional radiation therapy planning tools and intraoperative ultrasound to evaluate high dose rate prostate brachytherapy implants. Int.J. Radial Oncol. Biol.Phys., 43, 571-8. Narayana, V., Roberson, P.L, Winfield, R.J. and McLaughlin, P.W. (1997) Impact of ultrasound and computed tomography prostate volume registration on evaluation of permanent prostate implants. Int.J. Radial Oncol. Biol. P&ys.,39,341-6. Saw, C.B. and Suntharalingam, N. (1991) Quantitative assessment of interstitial implants. Int.J. Radial Oncol. Biol.Phys., 20,135-9. Saw, C.B. and Wu, A. (1991) Interpretation of the dose nonuniformity ratio for interstitial brachytherapy. Med. Phys., 18, 605. Low, DA and Williamson, J.F. (1995) The evaluation of optimized implants for idealized implant geometries. Med. Phys., 22,1477-84. Orton, C.G. (1994) Mathematical models. In Brachytherapy from Radium to Optimization, ed. R.F. Mould, J.J. Battermann, A.A. Martinez etal. Veenendaal, The Netherlands, Nucletron International B.V., 34-8. Bleasdale, C. and Jones, B. (1994) Mathematical model for the time interval between external beam radiotherapy and brachytherapy. In Brachytherapy from Radium to Optimization, ed. R.F. Mould, J.J. Battermann, A.A. Martinezef al. Veenendaal, The Netherlands, Nucletron International B.V., 39-48. Deehan, C. and O'Donoghue, J A (1994) Biological equivalence of LDRand HDR brachytherapy. In Brachytherapy from Radium to Optimization, ed. R.F. Mould, J.J. Battermann, A.A. Martinez etal. Veenendaal, The Netherlands, Nucletron International B.V., 19-33. Williamson, J.F. (1995) Recent developments in basic brachytherapy physics. In Radiation Therapy Physics, ed. A.R. Smith, Berlin, Springer Verlag, 247-302.
80 Computers in brachytherapy dosimetry
60. Daskalov, G.M., Kirov, A.S. and Williamson, J.F. (1998) Analytical approach to heterogeneity correction factor calculation for brachytherapy. Med. Phys., 25,722-35.
Wijrdeman, H.K., Battermann, J.J. (2000) The combined
61. Moerland, M.A., Van der Laarse, R., Luthmann, R.W.,
prostatic seed implants. Radiother. Oncol., 57,279-84.
use of the natural and the cumulative dose-volume histograms in planning and evaluation of permanent
6 Dose specification and reporting:the ICRU recommendations ANDRE WAMBERSIE AND JAN J.BATTERMANN
6.1
INTRODUCTION
This chapter addresses the problem of dose specification for reporting in brachytherapy. It is based on the recommendations of the International Commission on Radiation Units and Measurements (ICRU), mainly ICRU Report 38, 'Dose and volume specification for reporting intracavitary therapy in gynecology' (1985) [ 1 ]1,and ICRU Report 58, 'Dose and volume specification for reporting interstitial therapy' (1997) [2] 2 . Exchange of clinical information between radiation oncology centers requires uniformity and agreement on the methods used to specify the doses and the volumes to which these doses are delivered. To avoid confusion, an agreement has also to be reached on definitions of the terms and concepts necessary for reporting treatments. Due to the limitations of the irradiation techniques, the doses delivered to the target volumes are in general not homogeneous. In external-beam therapy with pho-
1. The Reporting Committee for ICRU Report 38 was the following: D Chassagne and A Dutreix (Co-Chairmen), P Almond, JMV Burgers, M Busch, and CA Joslin (Members), M Cohen and T Landberg (Consultants). 2. The Reporting Committee for ICRU Report 58 was the following: D Chassagne and A Dutreix (Co-Chairmen), D Ash, WF Hanson, AG Visser and JF Wilson (Members).
tons and electrons, the differences between the maximum and the minimum doses in the target volume often reach 10%, 15%, and even 20%. Therefore, one can introduce large discrepancies, and thus confusion, depending on the criteria (or the dose levels) used for prescribing, recording, and reporting the treatment, e.g., maximum, minimum, or any mean dose or 'weighted' mean dose. Such discrepancies are in general much larger than the current dosimetric uncertainties. On the other hand, a difference in dose of 5% can be detected clinically for some radical treatments [3,4]. The ICRU recognized the importance of the problem many years ago and, in 1978, published Report 29, 'Dose specification for reporting external beam therapy with photons and electrons' [5]. This report was superseded, in 1993, by ICRU Report 50, 'Prescribing, recording and reporting photon beam therapy' [6]. A 'Supplement to ICRU Report 50' appeared recently [7]. In brachytherapy, the situation is even more difficult because very high doses are always obtained close to the sources, and there are actually no large volumes for which the dose is nearly homogeneous (reaches a kind of plateau), as in external-beam therapy. The problem is addressed in ICRU Reports 38 and 58. In order to retain as much consistency as possible, it is desirable to use, the same terms and wherever possible, in concepts and the same approach as in external-beam therapy. In particular, the definitions of the volumes are therefore the same for the two techniques. It is, however,
82 Dose specification and reporting: the ICRU recommendations
recognized that brachytherapy raises some specific problems which have to be taken into account. Other ICRU reports dealing with dose specification for reporting special techniques, such as electron, proton, and neutron beam therapy, are in preparation. For reporting in external-beam therapy, the dose is specified first at the ICRU reference point, which is located (always) in the central part of the clinical target volume, and (when possible) at, or near, the intersection of the beam axes. The maximum and the minimum doses to the planning target volume (or their best estimation) should also be reported. In brachytherapy, there are high dose gradients within the clinical target volume and the use of a single reference point is not, therefore, sufficient. It is, however, appropriate to consider the dose at points where plateaus of dose occur in the central part of the clinical target volume and where the bulk of the malignant cell population is generally located. This leads to the concept of the mean central dose. In addition, the whole of the clinical target volume must receive a certain minimum dose in order to achieve the desired clinical effect. It is therefore also important to record the minimum dose at the periphery of the clinical target volume, i.e., the minimum target dose. Several systems of brachytherapy have developed historically. Best known and most widely used (with or without modification) are the Manchester, Quimby and Paris systems [8-14]. The term 'system' denotes a set of rules which takes into account the source types and strengths, geometry and method of application to obtain suitable dose distributions over the volume(s) to be treated. The system also provides a means of calculating and specifying dose. It is important to remember that, whereas an implant may follow the source distribution rules of a system, it does not comply with the system unless the method of dose prescription and specification are also followed. In addition, if the implant rules are modified, the dose uniformity intended by the system may be compromised. The situation is more difficult in intracavitary therapy due to the steep dose gradient in the vicinity of the sources, i.e., throughout the target volume. Therefore, the specification of the target absorbed dose in terms of the absorbed dose at specific point(s), in the vicinity of the sources, becomes less meaningful and a different approach is required. Instead of a target-dose specification, a volume specification is an alternative and, in that respect, specification of an intracavitary application in terms of the 'reference volume' enclosed by a reference isodose surface of 60 Gy has been proposed in ICRU Report 38. The problems of interstitial therapy (and of brachytherapy in general) are discussed first. The specific problems encountered in intracavitary therapy, and especially in gynecology, are dealt with in section 6.6. Over the last two decades, technological developments in brachytherapy have seen the introduction of minia-
turized and highly flexible sources which can be used in afterloading devices with radionuclides of different activities that can produce a wide range of dose rates. At the same time, sophisticated three-dimensional source localization methods have been developed and can be linked to computerized methods of dose calculation and representation of dose distribution. These developments have led many clinicians to depart from the longestablished implant systems and it is for this reason that a common language is valuable to provide a method of dose specification for reporting which can be used and be common for all types of brachytherapy applications. It should be stressed from the beginning that it has not been the intention or the role of the ICRU to encourage users to depart from their current practice of brachytherapy and from their method of dose prescription. The aim of the ICRU recommendations is to develop a common language for reporting a treatment, based on existing concepts. The description of the treatment and the method of dose specification for reporting should be presented in a way that can be easily understood and closely related to the treatment outcome.
6.2
DEFINITION OF VOLUMES
The definition of volumes is of utmost importance, both in external-beam planning and in brachytherapy planning. The process of determining volumes for the treatment of a malignant disease consists of several distinct steps, during which different volumes may be defined. 6.2.1
Gross tumor volume
The gross tumor volume (GTV) is the gross palpable or visible/demonstrable extent and location of the malignant growth. The GTV may consist of the primary tumor ('GTV primary'), metastatic lymphadenopathy(ies) ('GTV nodal'), or other metastases. The GTV almost always corresponds to those parts of the malignant growth where the tumor density is largest. Due to the high density of the cancer cells in the GTV, an adequate dose must be delivered to the whole GTV in order to achieve the aim of therapy in radical treatments. According to the above definition, there is no GTV after complete 'gross' surgical resection. There is no GTV when there are only a few individual cells or 'subclinical' involvement (even histologically proven). From the origin of medical terminology, the latin word tumor was used to designate a swelling, which could be of various types. The shape, size, and location of the GTV may be determined by means of different diagnostic methods such as clinical examination (e.g., inspection, palpation, endoscopy), and various imaging techniques (e.g., X-ray,
Definition of volumes 83
computerized tomography (CT), digital radiography, ultrasonography, magnetic resonance imaging (MRI), and radionuclide methods). The methods used to determine the GTV should meet the requirements for scoring the tumor according to the TNM [15,16] and American Joint Committee on Cancer (AJCCS) [17] systems, and the definition of the GTV is then in full agreement with the criteria used for the TNM classification. The GTV (primary tumor, metastatic lymphadenopathy, other metastases) may appear to be different in size and shape, sometimes significantly, depending on what examination technique is used for evaluation (e.g., palpation versus mammography for breast tumors, CT versus MRI for some brain tumors). Therefore, the radiation oncologist should, in each case, indicate which method has been used for the evaluation and delineation of the GTV. A GTV may be confined to only part of an organ (e.g., a Tl breast cancer), or involve a whole organ (e.g., multiple metastases of the brain). The GTV may or may not extend outside the normal borders of the organ tissue involved. For reporting, the GTV should be described in standard topographical or anatomical terms, e.g. '18 x 12 x 20 mm3 tumor in the left lobe of the prostate adjacent but not reaching the capsule.' In many situations, a verbal description might be too cumbersome and, therefore, for the purpose of data recording and analysis, a classification system is needed. Several systems are proposed for coding the anatomical description, some of them are mentioned in ICRU Report 50 [6]. There are at least three reasons for identifying the GTV. First, accurate description of the GTV is needed for staging (e.g., TNM). Second, identification of the GTV is necessary to allow for recording of tumor response in relation to the dose and other relevant factors. It can be used (carefully) as a prognostic factor. Third, an adequate dose must be delivered to all parts of the GTV in order to obtain local tumor control in radical treatments.
6.2*2
Clinical target volume
Clinical experience indicates that around a GTV there is generally subclinical involvement, i.e., individual malignant cells, small cell clusters, or microextensions, which cannot be detected by the staging procedures. The GTV, together with this safety margin consisting of tissues with presumed or proved subclinical involvement, is defined as the clinical target volume (CTV). The tissues immediately surrounding the GTV usually have a high malignant cell density close to the edge of the GTV; the cell density decreases toward the periphery of the CTV (often a safety margin of about 1 cm thick is taken). This CTV is usually denoted CTV-T. If the GTV has been removed by radical surgery, but it
is still thought that radiotherapy is needed for the tissues that remain close to the site of the removed GTV, this volume is also usually designated as CTV-T (e.g., in breast-saving procedures). Additional volumes with presumed subclinical spread (e.g., regional lymph nodes) may also be considered for therapy. They are also defined as CTVs and may topographically be designated CTV-N1, CTV-N2, etc. The CTV is a tissue volume that contains a gross tumor volume and/or subclinical microscopic malignant disease. This volume has to be treated at an adequate dose level (and time-dose pattern) in order to achieve the aim of therapy - cure or palliation. Delineation of a CTV will require consideration of factors such as the local invasive capacity of the tumor and its potential to spread to, for example, regional lymph nodes. The CTV, like the GTV, is a purely clinical-anatomical concept. It must always be described, independently of the dose distribution, in terms of the patient's anatomy and the tumor volume. As a minimum recommendation, the physical dimensions of the clinical target volume are described in terms of its maximum diameters (cm) in three orthogonal directions. For reporting, the CTV must be defined in plain topographic terms and/or according to a corresponding code in conformity with the recommendations for the GTV. If different dose levels are prescribed, different CTVs have to be defined. This is the case, for example, in 'boost' therapy where the 'high-dose' volume (often containing the GTV) is located inside the 'low-dose' volume. It must be stressed that the descriptions of the GTV(s) and CTV(s) are based only on general oncological principles, and are independent of any therapeutic approach. In particular, they are not specific to the field of radiation therapy. For example, in surgery, a safety margin is taken around the GTV according to clinical judgement, and this implies the use of the same CTV concept as in radiation therapy. In brachytherapy, as in external-beam therapy, volumes to be irradiated are defined, and thus the same concept of CTV is applied. Furthermore, the CTV concept can be applied to other modalities, e.g., regional chemotherapy, hyperthermia, and photocoagulation. The definitions of GTV and CTV in brachytherapy are thus identical to the definitions given for external-beam radiotherapy in ICRU Report 50 [6] and Supplement to Report 50, ICRU Report 62 [7].
6.2.3
Planning target volume
In external-beam therapy, to ensure that all tissues included in the CTV receive the prescribed dose, one has, in principle, to plan to irradiate a volume geometrically larger than the CTV. This is the planning target volume (PTV).
84 Dose specification and reporting: the ICRU recommendations
The additional safety margin included in the PTV results from a number of factors: expected physiological movements (e.g., with respiration) and variations in size, shape, and position (e.g., stomach, bladder, rectum) of the CTV; all variations and uncertainties in beam geometry and patient-beam positioning. The PTV is a geometrical concept, used for treatment planning, and it is defined to enable selection of appropriate beam sizes and beam arrangements, taking into consideration the net effect of all the possible geometrical variations, in order to ensure that the prescribed dose is actually absorbed in the CTV. The dose distribution to the PTV has to be considered to be representative of the dose distribution to the CTV. As indicated in ICRU Report 50 for external-beam therapy, when delineating the PTV, consideration may also be given to the presence of any radiosensitive normal tissue (organs at risk) as well as to other factors such as the general condition of the patient. Delineation of the PTV is a matter of compromise, implying the judgement and thus the responsibility of the radiation oncologist. In brachytherapy, the PTV is in general identical to the CTV. There are only very few exceptions. For instance, with some techniques in which there are uncertainties of consistency of source position (high dose rate, moving sources, fractionated techniques) or alteration of source position (intracavitary applications, permanent implants) during the application, the PTV may be larger than the CTV to take these factors into account. In this chapter, as in ICRU Report 58 [2], the term clinical target volume is used rather than planning target volume. In external therapy, the two steps localization of CTV and treatment planning can always be dissociated and therefore checked separately. However, in interstitial therapy, the CTV is finally decided upon by the clinician at the time of implantation on the assumption that it is contained within the minimum target isodose surface (see section 6.4.4). This procedure cannot be recommended, and the CTV should be clearly described in the patient chart before the implant is planned.
6.2*4
Treated volume
The treated volume is that volume of tissue, based upon the implant as actually achieved, which will receive at least a dose selected and specified by the radiation oncologist as being appropriate to achieve the purpose of treatment (e.g., tumor eradication or palliation). The treated volume is thus encompassed by an isodose surface corresponding to that dose level, which is the minimum target dose (see section 6.4.4). This isodose surface should, ideally, entirely encompass the CTV.
6.3 TECHNIQUES OF BRACHYTHERAPY: CLASSICAL SYSTEMS The Paterson-Parker or Manchester System was developed to deliver a reasonable dose uniformity (±10%) throughout a region implanted with radium needles [10]. The system specifies rules for the geometrical arrangement of the sources, and for the linear activity required in order to cover a PTV with a sufficiently homogeneous dose (Figure 6.1). The system includes tables of milligram-hour needed to deliver specified doses for different sizes of implants or moulds. The proportion of activity on the periphery is specified according to the size of the implant; it is larger for smaller implants. The system is still used for single-plane and double-plane implants in many centers. The Quimby System is characterized by uniform source spacing and uniform source activity [11]. Consequently, this arrangement of sources resulted in a non-uniform dose distribution, higher in the central region of the implant (as in the Paris System; see Figure 6.2). This system was particularly used in the US centers. The Paris System of implant planning has been developed mainly with iridium-192 wire sources [13,18]. The sources are of equal linear activity, parallel, placed at equal distances, and arranged in such a way that their centers are in the same plane perpendicular to the direction of the lines (Figure 6.2a and b). This plane, called the central plane, is the midplane of the application (Figure 6.3a, b, and c). If the volume to be treated is large, more than one plane containing wires is used. Again, equidistance of the radioactive lines is required. This means that their intersections with the central plane are arranged according to the apices of equilateral triangles or squares (Figure 6.3b and 6.4a and b). This regular distribution of the wires results in a slight overdose at the center of the target volume. The dose rate at a point in the middle of a group of sources is called the basal dose rate (BD). This BD is always calculated from the position of the sources in the central plane and is the minimum dose rate between a pair or group of sources. The values of the isodose curves are expressed as a percentage of the BD. The reference dose rate is derived from the BD and is equal to 85% of the BD. It is used for calculating the total treatment time of the implant. Because the ends of the active wires are not crossed, as in the Manchester System, the active sources should be 20-30% longer than the target volume at both ends. The minimum thickness of a treated volume is 50-60% of source separation for single planes and 130-150% for two planes. Dosimetry according to the Paris System has many advantages. The use of equal linear activity, equal distance between the sources, and the fact that no cross
Techniques of brachytherapy: classical systems 85
Figure 6.1 Manchester System for application of radioactive sources with different loading. As an example, (a) is the localization film for a bladder implant with radium needles of different activity. Although the Manchester System was designed for radium sources, it can be used with other radionuclides as well, such as cesium-137 needles or iridium-192 wires. As an example of the application of the Manchester System with radioactive sources other than radium, (b) and (c) give the distribution of dose rates for a single-pi one implant with iridium wires of unequal linear activity in order to ensure dose uniformity throughout the implanted region. Wires 1, 4, 5, and 6 (peripheral) contain a linear activity of 60 MBq (1.6 md) per cm; wires 2 and 3 contain a linear activity of 37 MBq (1 md) per cm. Wires 1, 2, 3, and 4 are 6 cm long; wires 5 and 6 are 3.5 cm long, (b) The dose rates in the plane containing the wires; (c) the dose rates in a perpendicular plane.
needles are used make the application itself relatively easy. The relationship between the geometry of the implant, and the dimensions of the target volume can easily be determined. The dose rate can be quickly controlled with a planning computer, even in complicated implantations. Nowadays, most of the implant techniques are based on the original Paris System. For rapid planning, in some institutes, normograms have proven useful as approximate planning guides. Both for removable iridium implants and for permanent iodine implants, normograms were developed at the Memorial Sloan Kettering Institute [19]. However, individualized computer planning, in general is superior to
the above-described techniques, because the isodose lines generated by the computer allow a far more complete evaluation of the treatment plan. Both orthogonal and isocentric techniques are used to reconstruct the source coordinates. The isocentric reconstruction method is a variation of the stereo-shift method. With isocentric equipment, like a treatment simulator, the angle between the central axes of the projecting beams can be enlarged up to 60°, still obtaining two projections of the sources (carriers) on the same radiograph (Figure 6.5 a, b, and c). With the simulator, variable angles can also be chosen, such that sources are not obscuring one another.
86 Dose specification and reporting: the ICRU recommendations
Figure 6.2 lridium-192 wire implant according to the Paris System (single-plane implant). The wires are of equal linear activity, parallel, and arranged in such a way that their centers are in the same plane perpendicular to the direction of the wires (i.e., the central plane, see Figure 6.3). Figure 6.3 Central plane. In an implant where the source lines are rectilinear, parallel, and of equal length, the central plane is perpendicular to the direction of the source lines and passes throughout their centers. The mean central dose (DJ is the arithmetic mean of the local minimum doses D, (i = A, B ...) in the plateau region, (a) A single-plane implant; (b) a twoplane implant; (c) an actual single-plane implant where sources are not rectilinear: the central plane can be defined as in (a). (From ICRU Report 58 [21)
6,4 DESCRIPTION OF DOSE DISTRIBUTION IN INTERSTITIAL THERAPY
regions of high dose surrounding each source. However, within the volume of the implant there are regions where the dose gradient approximates a plateau (Figure 6.6).
6.4.1
1. In an interstitial implant, the regions of plateau dose are equidistant between adjacent neighboring sources, for sources of identical activity. They are regions of local minimum doses.
General concepts
In interstitial therapy, the dose distribution is nonhomogeneous and includes steep dose gradients and
Description of dose distribution in interstitial therapy 87
Figure 6.4 Dose planning for implants with iridium-192 wires contained in two parallel planes, following the Paris System. Examples of a breast implant in two planes, (a) The seven wires are equidistant and arranged in triangles (length of the wires 7 cm for the upper row and 8 cm for the lower row), linear activity 52 MBq cnr1 (1.4 md cnr1), application time 43.32 h for a reference dose of 20 Gy. (b) The six wires are equidistant and arranged in squares (length of the wires 6 cm for the upper row and 7 cm for the lower row), linear activity 52 MBq Cm-1 (1.4 mCi cm-1), application time 42.91 h for a reference does of 20 Gy.
2. Variations between these local minimum doses can be used to describe the dose uniformity of an implant. 3. A region of plateau dose is the place where the dose can be calculated most reproducibly and compared easily by different departments. Although in modern computer systems the threedimensional dose distribution can be computed and presented as isodose surfaces, these facilities are not yet available in all departments. In order to provide the minimum of information needed about the dose or dose rate distribution, the calculation of isodose curves in at least one chosen plane is necessary. If only one plane is chosen for isodose calculation, the central plane of the implant (as defined in section 6.4.2) should be chosen for this purpose. In order to assess the dose distribution in other areas of the implant, multiple planes for isodose calculation can be chosen, either parallel or perpendicular to the central plane.
6.4.2
The central plane
In source patterns in which the source lines are straight, parallel, of equal length and with the centers which lie in a plane perpendicular to the direction of the source lines, this plane is the central plane (see Figure 6.3a and b).
In an actual implant, all source lines may not necessarily be straight, parallel, and of equal length. In such cases, the central plane should be chosen perpendicular to the main direction of the source lines and passing through the estimated center of the implant (see Figure 6.3c). For more complex implants, it may be necessary to subdivide the target volume into two or more subvolumes for dose evaluation. In this event, a central plane may be defined for each of these subvolumes (Figure 6.7). The calculation of dose distributions in multiple planes throughout the target volume shows that a variation of a few millimetres in the position of the central plane is not critical.
6.43
Mean central dose
In interstitial therapy, the mean central dose is taken to be the arithmetic mean of the local minimum doses between sources in the central plane (or in the central planes if there are more than one). In the case of a single-plane implant, the mean central dose is, in the central plane, the arithmetic mean of the doses at mid-distance between each pair of adjacent source lines, taking into account the dose contribution at
88 Dose specification and reporting: the ICRU recommendations
Figure 6.5 Orthogonal AP (a), lateral (b) and isocentric (c) radiographs of Fletcher-Suit rigid applicator. Note lead wire in vaginal packing, contrast medium in balloon of Foley catheter, and air in distal rectum.
that point from all sources in the pattern (see Figure 6.3a). In the case of an implant with line sources in more than one plane, the mean central dose is the arithmetic mean of the local minimum doses between each set of three adjacent source lines within the source pattern (see Figure 6.3b). As seen in Figure 6.4a, the minimum dose lies at the intersection of perpendicular bisectors of the sides of the triangles (geometric center) formed by these source lines. This point is equidistant from all three source lines. In some complex implants, a single central plane may not bisect or even include all the sources. In these cases, a mean central dose based on one plane can be misleading and it is advisable to subdivide the volume and to choose a separate central plane for each sub volume (see Figure 6.7). Three practical methods are acceptable for determining the mean central dose: 1. If parallel lines are used, one can identify triangles consisting of three adjacent source lines for all the sources, so that the triangles formed constitute as
Figure 6.6 Plateau dose region between radioactive sources. In a plane perpendicular to linear and parallel sources, the dose distribution shows a plateau region of low dose gradient. In this example of three sources, 6 cm long and with 1.5 cm spacing, the dose varies by less than 2% in the gray region between the sources. (From Dutreix et a I. [18].)
Description of dose distribution in interstitial therapy 89
Figure 6.7 Central planes in a complex implant. It is sometimes necessary to plan the treatment in terms of two or more subvolumes. In the example shown, where all source lines are not of equal length, two central planes are identified: (a) for the longest source lines and (b) for the shortest ones. Two mean central doses are determined in the two subvolumes Dma and Dmb, respectively. Open circles are the intersections of the sources with the central planes, and closed circles are the points where the local minimum doses are calculated. (From ICRU Report 58 [2].)
many acute triangles as possible. The intersection points of the perpendicular bisectors of each triangle are determined and the local minimum doses are calculated at each of these points. The mean of these local minimum doses is the mean central dose. This method is the most precise one when parallel lines are used. 2. Evaluation of dose profiles: the dose profiles are calculated for one or more axes through the center of the implant expected to pass through as many local minima as possible. The local minimum doses are determined by inspection. The mean of these local minimum dose values is the mean central dose (Figure 6.8). In a single surface implant performed following a curved surface, a profile may lead to an underestimation of the mean central dose. In a complex implant, it may be difficult to find axes passing through the minima and profiles may lead to an overestimation of the mean central dose. However, experience shows that the error lies within acceptable limits. This method is sometimes preferred for seed implants. In a seed implant, such as the one presented in Figure 6.9, the dose should be calculated along several random profiles passing through the implant. 3. Inspection of dose distribution: the dose
Figure 6.8 Evaluation of dose profiles. Three profiles (b) are drawn along two orthogonal directions through a two-plane implant (a) with eight parallel line sources, 10 cm long, 1.8 cm spacing. The profiles are calculated in percentage of the minimum target dose (thick line) along axes XX, YY and YY in the central plane. The profile along the axis YY is the most representative to estimate the mean central dose. The mean of the local minimum doses is the mean central dose. The mean central dose is equal to 7 78% of the peripheral dose. (From ICRU Report 58 [2].)
90 Dose specification and reporting: the ICRU recommendations
Figure 6.10
Determination of mean central dose from
inspection of dose distribution. Dose distribution in the central plane of an implant with six parallel iridium-192 line sources, 6 cm long, 1.5 cm spacing, reference air kerma rate 14.5 mGyh-1 at 1 m. The dose varies by 5% between plotted isodose lines in the region of interest. The idodose values are 16, 19, 22, 24, 26, 28, 30, 31.5, 33, 35, 40, and 45 cGy h-1 The local minima, A, B, C, and D, can be easily estimated by inspection. DA and DD approximate 31 cGy h-1 and D6 and Dc approximate 34 cGy h-1 The estimated mean central dose is Dm - 32.5 cGy h-1 (From ICRU Report 58 [2].)
Figure 6.9 Seed implant with 68 iodine-125 seeds of 19.2 MBq (0.52 md), total activity 1310 MBq (35.4 md). (a) Radiograph of implant, (b) Dose distribution in the central plane.
distribution is plotted in the central plane. With isodose lines varying by 5% (at most 10%) of the local dose in the central region, the local minima can be determined by inspection. The mean of these local minima is the mean central dose (Figure 6.10). This method is often preferred for complex implants with line sources. 6*4.4
Minimum target dose
The minimum target dose is the minimum dose at the periphery of the CTV. It should be equal to the minimum dose decided upon by the clinician as adequate to treat the CTV. The minimum target isodose surface is the isodose surface corresponding to the minimum target dose. As indicated above, it defines the treated volume and should entirely encompass the CTV (see section 6.2.2). The minimum target dose corresponds to the prescribed dose in many instances.
The minimum target dose is known in some American centers as the 'minimum peripheral dose' [20]. It is known as the 'reference dose' in the Paris System, and is equal to about 90% of the prescribed dose in the Manchester System for interstitial therapy. 6.4.5
High-dose regions
In order to correlate radiation dose with late damage, the high-dose regions around sources should be assessed (Figures 6.4 and 6.11). There will inevitably be a high-dose zone around each source. Although this zone is often small and well tolerated, the exact tolerance dose and volume for interstitial therapy are not known. However, it is necessary, for intercomparison purposes, to agree on a way to describe the high-dose volumes. It has been suggested that a dose of approximately 100 Gy is likely to be significant in determining late effects. In those patients who receive 50-60 Gy as peripheral minimum dose or 60-70 Gy as mean central dose, 100 Gy corresponds approximately to 150% of the mean central dose. It is therefore recommended in ICRU Report 58 [2] to report the size of the region receiving more than 150% of the mean central dose. The high-dose regions should be defined as the
Description of dose distribution in interstitial therapy 91
Figure 6.11 Tongue implant, using five loops of 8 cm indium wires with activity of 68 MBq cm'1 (1.8 md cm-1). (a) Radiographs of the implant, (b) Dose distribution in the central plane of the implant.
regions encompassed by the isodose corresponding to 150% of the mean central dose around the sources in any plane parallel to the central plane where a high-dose region is suspected. The maximum dimensions of all regions in all planes calculated should be reported. 6.4.6
Low dose regions
A low dose region should be defined as a region within the CTV, encompassed by an isodose corresponding to 90% of the prescribed dose. The maximum dimension of the low dose region in any plane calculated should be reported. In implants for which the CTV is included within the minimum target dose isodose, the occurrence of a low dose region is exceptional. If the clinical target volume is not covered by the minimum target dose isodose, there will be low dose regions due to geographical miss. Low dose regions should be reported in order to correlate the local recurrence rate with the dose distribution.
6*4.7
Dose uniformity parameters
Several indices quantifying the homogeneity of the dose distribution have been proposed (see, for example, references 21-23).
Two parameters describing dose uniformity for interstitial implants are recommended in ICRU Report 58 [2]. They can be derived directly from the concepts of minimum target dose and mean central dose: 1. the spread in the individual minimum doses used to calculate the mean central dose in the central plane expressed as a percentage of the mean central dose; 2. the dose homogeneity index, denned as the ratio of minimum target dose to the mean central dose.
6*4*8 Additional representations of the dose distribution In order to obtain a full perception of the dose distribution of an implant, the use of volume-dose calculations has been advocated (see, for example, references 24-26). For this purpose, the CTV (or a larger volume including an additional margin) is subdivided in subvolumes (e.g., voxels) and the dose rate is calculated at the center of each subvolume. The volume receiving at least a specified dose is then defined as the sum of all subvolumes where at the center at least that dose is received. Examples of results are shown in Figure 6.12. Because of high dose gradients, significant differences in calculated volumes can be observed, depending upon the size of the elementary subvolumes. The size of the grid and of the
92 Dose specification and reporting: the ICRU recommendations
elementary subvolumes used in dose and volume calculations should be clearly stated. Volume-dose data can also be represented by means of histograms showing the distribution of fractions of the CTV receiving doses within chosen intervals, especially the natural volume-dose histogram (NVDH) as published by Anderson [27]. With this model, even small differences between implants can be revealed. The main characteristic of the NVDH is the peak that occurs with regular implant of several sources (see, for example, Figure 6.13). In fact, the peak dose reflects the basal dose of the Paris System. If the implant is less uniform, the peak is wider. So, the NVDH can be used for intercomparison between planned and realized source arrangements [28,29]. The value of these alternative representations of the dose distribution as a possible prognostic factor for treatment outcome has still to be established in clinical research.
6.5 RECORDING AND REPORTING INTERSTITIAL THERAPY Figure 6.12 Volume-dose curves. Volume (sum of subvolumes receiving at least a certain dose) versus dose, for two different patterns of parallel source lines: a two-plane implant with six sources 5 cm long (upper curve), and a cylindrical implant with seven sources 4 cm long (lower curve). The dose is expressed as percentage of the minimum target dose. The size of the voxel used for calculation is 1 mm3. For the upper curve, linear
Adequate information must be recorded in order to give a consistent description of any implant. The guidelines for reporting doses will make it possible to compare results of future brachytherapy practice and to better relate outcome to treatment. In order to report an implant, at least the following should be recorded [2].
sources are simulated by point sources (seeds) arranged in a linear fashion. (Bridier et al. [25])
Figure 6.13 Natural volumedose histogram of the tongue implant in Figure 6.11. Treatment dose rate of 1 Gy h-1 was chosen to deliver 60 Gy in 60 h. (From Anderson [27].)
Recording and reporting interstitial therapy 93
6.5.1
Description of volumes
6.5.4
Description of time-dose pattern
The description of volumes should include as a minimum the GTV, the CTV, and the treated volume.
The description of the time-dose pattern should include the type of irradiation with the necessary data on treatment and irradiation time, as described below. The information on dose and time should provide the necessary data to calculate instantaneous and average dose rates.
6.5.2
• Continuous irradiation: the overall treatment time should be recorded. • Non-continuous irradiation: both the overall treatment time and the total irradiation time should be recorded. • Fractionated, hyperfractionated, and pulsed irradiation: the irradiation time of each fraction, the interval between fractions, and the overall treatment time should be recorded. • When the irradiation times of the different sources are not identical, they should be recorded.
Description of sources
The description of the sources employed should include details of: • Radionudide used including nitration, if relevant. • Type of source used, i.e., wire, seed, seed ribbon, hairpin, needle, etc. • Length of each source line used. • Reference air kerma rate of each source (or source line): the reference air kerma rate of a brachytherapy source is the kerma rate to air, in air, at a reference distance of 1 m, corrected for air attenuation and scattering. The quantity reference air kerma rate is expressed in Gy s-1 at 1 m, or a multiple of this unit (in a convenient way, for low dose-rate brachytherapy, in microgray per hour, )m,Gy Ir1, at 1 m). The problem of specification of sources used in brachytherapy has been discussed by several authors, and the quantity reference air kerma rate has been increasingly adopted by different organizations or commissions [1,2,30-39]. • The distribution of the strength within the source should be described (uniform or differential loading, etc.) [40,41].
6.5.3 Description of technique and source pattern If the source distribution rules of a standard system have been followed, this must be specified. If it is not the case, the source pattern should be described completely and unambiguously. In addition, the following data should be recorded: • number of sources or source lines, • separation between source lines and between planes, • geometrical pattern formed by the sources with the central plane of implant (e.g., triangles, squares), where relevant, • the surfaces in which the implant lies, i.e., planes or curved surfaces, • whether crossing sources are placed at one or more ends of a group of linear sources, • the material of the inactive vector used to carry the radioactive sources, if any (e.g., flexible or rigid); whether rigid templates are used at one or both ends, • type of remote afterloading, if used.
Moving sources: • Stepping sources: step size and dwell time should be recorded if constant. Variation of the dwell times of a stepping source can be used for manipulating the dose distribution. If such a dose optimization is applied, this should be specified (e.g., optimization on dose points defined in the implant or geometrical optimization [42]. • Oscillating sources: speed in different sections of the vectors should be recorded.
6.5.5
Total reference air kerma (IRAK)
The total reference air kerma is the sum of the products of the reference air kerma rate and the irradiation time for each source. The TRAK is an important quantity which should be reported for all brachytherapy applications. It is a quantity that is simple to calculate and on which there can be no ambiguity. It is analogous to the milligrairrhour (mg.h) of radium. The conversion of the quantity mg.h to the TRAK is easy and straightforward (1 mg.h radium equivalent corresponds to 7.2 mGy h-1, at 1 m). In addition, the TRAK is proportional to the integral dose to the patient, and can also serve as a useful index for radiation protection of personnel. However, the simple determination of the TRAK does not allow one to derive, even approximately, the absorbed dose in the immediate vicinity of the sources (i.e., in the tumor or target volume).
6.5.6
Description of dose distribution
The following doses should be recorded. • Prescribed dose: if the dose is not prescribed at the level of either the minimum target dose or the mean central dose, the method of dose prescription should
94 Dose specification and reporting: the ICRU recommendations
be recorded. If, for clinical or technical reasons, the dose received differs from the prescribed dose, it should be noted. • Minimum target dose. • Mean central dose. The following additional information, when available, should be recorded: • • • •
Dimension of high dose region(s). Dimension of any low dose region. Any dose uniformity data. Additional representation of dose distribution, if any.
The above guidelines are based on the recommendations contained in ICRU Report 58 [2]. It should be stressed again that it is not the intention, or the role, of the ICRU to encourage radiation oncologists to depart from their current practice of dose prescription or technique of application. The ICRU reports aim to help radiation oncologists to report a given application in the same way, using the same definitions and concepts. One should avoid a situation in which the same application would be described differently in different centers or, conversely, in which the same reported dose would correspond to completely different actual dose distributions. For the purposes of this chapter, the prescribed dose is defined as the dose which the physician intends to give and which is entered in the patient's treatment chart. Depending on the system used, the approach for dose prescription in interstitial therapy may be different from center to center.
6.6 SPECIFIC PROBLEMS FOR INTRACAVITARY THERAPY IN GYNECOLOGY
6.6.1
Introduction
As the absorbed dose in soft tissues from intracavitary applications is so highly non-uniform throughout the target volume, the concepts of maximum, mean, median, and modal target absorbed dose, as defined in ICRU Report 50 [2], are not relevant. The minimum target absorbed dose is the only useful concept and is, by definition, equal to the treatment absorbed dose level. For external-beam therapy, it has been recommended that the target absorbed dose be defined as the absorbed dose at one or more specification points which are representative of the dose distribution throughout the target volume. These specification points could be established with respect to the target volume (center or central part) or to the beam axes, or both. In contrast, in intracavitary therapy, due to the steep dose gradient in the vicinity of the sources, i.e., throughout the target volume, the specification of the target absorbed dose in terms of the absorbed dose at specific
point(s), in the vicinity of the sources, is not at all meaningful and a different approach is required. Instead of a target-dose specification, a volume specification is recommended in ICRU Report 38 [1]. Specification of an intracavitary application in terms of the 'reference volume' enclosed by the reference isodose surface of 60 Gy is proposed. However, as the different isodose surfaces are close to each other, the indication of the reference volume must be supplemented, for safety reasons, by the indication of the TRAK. In addition, recording the absorbed dose at reference points related to organs at risk or to fixed bony structures is recommended. As for interstitial therapy, the ICRU recommendations contained in Report 38 [1] do not imply a modification of the method used for the calculation of the treatment duration, but they require the calculation of specific quantities for reporting. The recommendations presented in ICRU Report 38 [1] must be considered a minimum requirement for reporting. On the other hand, the reported parameters will be meaningful only to the extent that the technique of the particular intracavitary application has been completely described. ICRU Report 38 [1] deals mainly with the treatment of cervix carcinoma, for which the anatomical region of interest is similar for every patient and the possible variation in the position of the radioactive sources is limited. However, for other gynecological intracavitary applications, the same philosophy can be adopted, but some of the numerical values and definitions may need to be modified according to the type of application.
6.6.2
Description of the technique
It is recommended that the technique be described on the basis of the guidance given below. THE SOURCES
1. Radionuclide 2. Reference air kerma rates 3. Shape, filtration, etc. SIMULATION OF LINEAR SOURCES
When a linear source is simulated by a set of point sources, the activity of these point sources and their separation^) must be indicated [25,43]. When moving sources are used to simulate a set of different sources in fixed position, in order to produce an appropriate dose distribution, the following indications are required [44-47]: 1. type of movement (continuous or stepwise, step distance), 2. unidirectional or oscillating movement, 3. range of movement or oscillation,
Specific problems for intracavitary therapy in gynecology 95
4. speed in different sections of the applicator, or dwell times of the source at different positions. THE APPLICATOR
Reference to the applicator is sufficient when a complete description has already been published, provided that there is no significant difference between the applicator used and the one described in the literature. To avoid confusion, it is recommended that the applicator be described, including the name of the manufacturer. The description should include information on the following points: 1. rigid (or not), consequently with fixed known geometry (or not) of the complete applicator, 2. rigid uterine source with fixed curvature (or not), 3. connection between vaginal and uterine applicators, i.e., fixed, loose (semi-fixed), free, 4. type of vaginal sources, number and orientation of line sources, special sources (box, ring, etc.), 5. high atomic number shielding materials in vaginal applicator (or not).
6.63
Recommendations for reporting
Three sets of quantities are recommended in ICRU Report 38 [1] to specify intracavitary application for cervix carcinoma; they complement each other and should be combined. TOTAL REFERENCE AIR KERMA (TRAK)
The TRACK will always be reported (see section 6.5.5).
rates, the radiation oncologist has to indicate the dose level which he or she believes to be equivalent to 60 Gy delivered at the conventional low dose rate, and this should be clearly stated [48-50]. Reference volume: description of the pear-shaped volume
When the uterine source(s) is combined with vaginal sources, or when the uterine source is more heavily loaded at the lower end, the tissue volume to be described presents a pear shape, with its longest axis coincident with the intrauterine source (Figure 6.14). This reference volume is defined by means of three dimensions (Figure 6.15): 1. the height (dh)is the maximum dimension along the intrauterine source and is measured in the oblique frontal plane containing the intrauterine source; 2. the width (dw) is the maximum dimension perpendicular to the intrauterine source and is measured in the same oblique frontal plane; 3. the thickness (dt) is the maximum dimension perpendicular to the intrauterine source and is measured in the oblique sagittal plane containing the intrauterine source. The definitions of dh, dw and dt are proposed in order to minimize the number of calculations. These dimensions are usually expressed in centimeters. The volume estimated from the intersections of the surface of a pearshaped volume on two conventional planes does not necessarily represent the size of the true reference volume. However, for most applications they do not differ from the maximum dimensions of the reference volume by more than 1 or 2 mm.
DESCRIPTION OF THE REFERENCE VOLUME
The description of the reference volume, i.e., the tissue volume encompassed by a reference isodose surface, has been proposed for specification in reporting. The reasons for this approach are described in section 6.6.1. Dose level
An absorbed dose level of 60 Gy is widely accepted as the appropriate reference level for conventional low doserate therapy. When two or more intracavitary applications are performed, the absorbed dose to consider is that resulting from all applications. The time-dose pattern should be clearly stated. When intracavitary therapy is combined with external-beam therapy, the isodose level to be considered is the difference between 60 Gy and the dose delivered at the same location by external-beam therapy. For example, if a dose of 20 Gy were delivered to the whole pelvis by external-beam therapy, the isodose level to be considered would be 60 —20 Gy = 40 Gy. Nevertheless, it is recognized that the combined dose does not necessarily produce the same effect as a similar dose from intracavitary therapy alone. For intracavitary therapy at medium or high dose
ABSORBED DOSE AT REFERENCE POINTS
Several reference points are in current use. Some are relatively close to the sources and related either to the sources or to organs at risk; others are relatively far from the sources and are related to bony structures. The following definitions apply to the case where the doses are calculated from two perpendicular radiographs, anterioposterior (AP) and lateral. When other methods are used, such as stereographic X-ray films, oblique perpendicular radiographs or transverse sections (CT scans), the calculations need to be modified. Reference points close to the sources and related to the sources
As such points are located in a region where the dose gradients are high, any inaccuracy in the determination of distance results in large uncertainties in the absorbed doses evaluated at these points. Such calculated absorbed doses do not, therefore, seem an appropriate means of characterizing an intracavitary application and/or of reporting the target absorbed dose, particularly if rigid source combinations are not used. Such points are not recommended in ICRU Report 38 [ 1 ].
96 Dose specification and reporting: the ICRU recommendations
Figure 6.14 Dose distribution of Fletcher-Suit rigid applicator, as in Figure 6.5, using total activity of 606 MBq (16.4 md) cesium-137 and showing the pear-shaped tissue volume, (a) Plane perpendicular to Z-axis. (b) Plane perpendicular to X-axis.
Reference points relatively close to the sources but related to organs at risk
The determination and specification of the absorbed dose to organs at risk (bladder, rectum, etc.) are obviously useful with respect to normal tissue tolerance limits. However, such information will be meaningful only to the extent that it is obtained and expressed in precise and well-codified ways. • Calculated values: reference points for the expression of the absorbed dose to the bladder and the absorbed dose to the rectum (see Figure 6.16) have been proposed by Chassagne and Horiot [51]. The bladder reference point is obtained as follows. A Foley catheter is used. The balloon must be filled with 7 cm3 of radio-opaque fluid. The catheter is pulled
downwards to bring the balloon against the urethra. On the lateral radiograph, the reference point is obtained on an AP line drawn through the center of the balloon. The reference point is taken on this line at the posterior surface of the balloon. On the frontal radiograph, the reference point is taken at the center of the balloon. The point of reference for the rectal dose is obtained as follows. On the lateral radiograph, an AP line is drawn from the lower end of the intrauterine source (or from the middle of the intravaginal sources). The point is located on this line 5 mm behind the posterior vaginal wall. The posterior vaginal wall is visualized, depending upon the technique, by means of an intravaginal mould or by an opacification of the vaginal cavity with a radio-opaque gauze used for the
Specific problems for intracavitary therapy in gynecology 97
/
Figure 6.16 Determination of the reference points for bladder and rectum as proposed by Chassagne and Horiot [51].)
Reference points related to bony structures
Figure 6.15 Geometry for measurement of the size of the pearshaped 60 Gy isodose surface (broken line) in a typical treatment of cervix carcinoma using one rod-shaped uterine applicator and two vaginal applicators. Plane a is the 'oblique'frontal plane that contains the intrauterine device. The oblique frontal plane is obtained by rotation of the frontal plane around a transverse axis. Plane b is the 'oblique' sagittal plane that contains the intrauterine device. The oblique sagittal plane is obtained by rotation of the sagittal plane around the AP axis. The height (d J and the width (d J of the reference volume are measured in plane a as the maximal sizes parallel and perpendicular to the uterine applicator respectively. The thickness (dt) of the reference volume is measured in plane b as the maximal size perpendicular to the uterine applicator. (From ICRU Report 38 [1].)
packing. On the AP radiograph, this reference point is at the lower end of the intrauterine source or at the middle of the intravaginal source(s). Monitoring of the absorbed dose rate to the rectum: in addition to calculating the rectal dose, the dose, or dose rate, can be measured at different points along the anterior rectal wall to ensure that no area of the rectal mucosa receives a dose above the tolerance level. This type of measurement requires special care in positioning the measuring probe. An example is given in Figure 6.17.
• The lymphatic trapezoid is obtained as follows (Figure 6.18). A line is drawn from the junction of S1-S2 to the top of the symphysis. Then a line is drawn from the middle of that line to the middle of the anterior aspect of L4. A trapezoid is constructed in a plane passing through the transverse line in the pelvic brim plane and the midpoint of the anterior aspect of the body of L4 (from Fletcher [52]). A point 6 cm lateral to the midline at the inferior end of this figure is used to give an estimate of the dose rate to mid-external iliac lymph nodes. At the top of the trapezoid, points 2 cm lateral to the midline at the level of L4 are used to estimate the dose to the low para-aortic area. The midpoint of a line connecting these two points is used to estimate the dose to the low common iliac lymph nodes. • The pelvic-wall reference point [51] can be visualized on an AP and a lateral radiograph and related to fixed bony structures. This point is intended to be representative of the absorbed dose at the distal part of the parametrium and at the obturator lymph nodes (Figure 6.19). On an AP radiograph, the pelvic-wall reference point is intersected by the following two lines: a horizontal line tangential to the highest point of the acetabulum, and a vertical line tangential to the inner aspect of the acetabulum. On a lateral radiograph, the highest points of the right and left acetabulum, in the cranio-caudal direction, are joined and the lateral projection of the pelvic-wall reference point is located at the mid-distance of these points. Evaluation of the absorbed dose at reference points, related to well-defined bony structures and lymph node areas, is particularly useful when intracavitary therapy is combined with external-beam therapy. It is also useful in
98 Dose specification and reporting: the ICRU recommendations
Figure 6.17 Measurement of the rectal dose rate. The rectal dose is measured following the insertion of the source applicators, either preloaded (low-activity treatment) or manually loaded with low-activity sources identical in design to those used during highactivity afterloaded treatment. Method A: the measuring probe is moved relative to a rigid guide tube inserted into the rectum and held in position. The point of maximum rectal dose rate is noted and the distance d, in cm, from the anal verge deduced. Method B: the measuring probe is moved so that the tip of the probe is moved along the midline of the recto-vaginal septum until the point of maximum dose rate is reached. Distance is taken as a direct reading on the central tube and at the anal verge. The dose rate and distance are recorded. The major disadvantage of Method A is that the probe tip cannot follow the surface of the anterior rectal wall closely. However, a IIowa nee for the distance of the probe sensor from the vagi no-recta I septum needs to be taken into account. (From ICRU Report 38 [1].)
Figure 6.18 Determination of the lymphatic trapezoid. On the left is an anteroposterior view and on the right a lateral view (see text). (From Fletcher [52].)
helping to avoid an overdose when intracavitary therapy is to be followed by surgery. CALCULATION OF DOSE DISTRIBUTION
The present recommendations, in particular the description of the reference volume encompassed by the 60-Gy isodose surface, necessitate the computation of complete dose distributions in several planes.
The respective planes for which the dose distribution is to be computed will depend on the technique and the particular clinical situation. However, as a minimum requirement, it is recommended that the dose distributions be computed in two planes: the oblique frontal plane and the oblique sagittal plane, both containing the intrauterine source. When practicable, it is recommended that dose distributions be calculated in additional sets of planes and
Concluding remarks 99
Figure 6.19 Determination of the right (RPW) and left (LPW) pelvic wall reference points (see text). (From Chassagne and Horiot
[57].;
that these dosimetric data be correlated with those obtained from radiographs or CT sections, in order to determine the absorbed dose at any relevant anatomical point. While this additional information will be of value in assessing effects in any individual patient, it will also provide: 1. the possibility of comparing the methods of specification used in different centers and of evaluating their respective merits; 2. the possibility of comparing the methods of specification used in historical series (mg»h, points 'A' and 'B') with the methods recommended in ICRU Report 38 [1]; 3. the possibility of deriving new clinical and radiobiological data and correlations which could improve treatment techniques and develop further the method of specification. 6.6*4 Definition of the 60-Gy reference volume in special situations • One linear source only. In some situations, only one linear source is present: in the case of a narrow vagina with a uterine source protruding into the vaginal cavity, in the case of vaginal irradiation with a central source from a cylindrical applicator. In estimating the volume in this simple case, the width is equal to the thickness, as the dose distribution is symmetrical about the source axis. • Vaginal sources only. When only vaginal sources are present, width is the largest dimension from right to left in an oblique frontal plane through the main axis of the vagina. Thickness is the largest dimension in a direction perpendicular to the above oblique plane. Height is measured along the vaginal axis, and is commonly shorter than the other two dimensions. • Rigid applicator. Provided that there is a fixed connection between vaginal and uterine sources, pre-
calculated isodose surfaces can be obtained for given loadings of the applicator. Therefore, pre-calculated dimensions of height, width, and thickness can be given [53]. Uterine packing in endometrial carcinoma. In connection with uterine packing, the same definitions of height, width and thickness given in section 6.6.3 can be used. However, two facts need to be noted: width and thickness are usually located at the level of uterine fundus (the pear-shaped volume is reversed), height should be determined in the oblique frontal plane, which gives the maximum dimension.
6.7
CONCLUDING REMARKS
At the end of this chapter, it should be stressed again that the aim of the ICRU Reports 38 and 58 [1,2] is not to encourage the users to depart from their current practice in brachytherapy. Treatment prescription is the responsibility of the radiation oncologist (or team) in charge of the patient; it is based on the radiation oncologist's judgement and experience and implies his or her responsibility. Reporting a treatment is another issue. The aim of the ICRU efforts is to recommend a common language for reporting a treatment in such a way that the clinical information can be exchanged in a relevant way, avoiding misinterpretation and confusion between radiation oncologists and departments. However, the use of the same sets of definitions, concepts, and approaches for prescribing, recording, and reporting a treatment has obvious advantages in simplifying the issues and avoiding confusion. It could be a long term beneficial consequence of the ICRU efforts. ICRU Report 58 [2] on interstitial brachytherapy was published at the end of 1997, but Report 38 [ 1 ] was published in 1985. Significant changes took place during the 14 years in the field of brachytherapy, especially the development and dissemination of high dose-rate, and pulse dose-rate applications.
100 Dose specification and reporting: the ICRU recommendations
Revision of ICRU Report 38 [1] becomes necessary and is welcomed by the radiation oncology community, as indicated by a recent inquiry [54,55]. The ICRU initiated a revision of Report 38, in 1998, on the basis of the answers to a questionnaire sent to the ESTRO members and a large survey of the recent literature. The following topics are considered: HDR, MDR, PDR, exploitation of the patient data provided by the modern imaging techniques, localization of the different Volumes', and identification of the organs at risk, specification of dose, combination of external and intracavitary therapy, three-dimensional treatment planning, and radiobiological issues raised by the different dose rates which can now be applied. The use of the TRAK should be encouraged. The fact that several companies use this quantity to specify the sources they are manufacturing will certainly facilitate its general use (e.g., Amersham,1 CIS Bioindustries,2 Mallinckrodt Medical3). Finally, as in externa-beam therapy (see ICRU Report 50 [6]), several levels of complexity could be proposed for reporting the treatments in brachytherapy (see, for example, reference 56): • Level 1: basic techniques-minimum requirements. A standard applicator is used, with a fixed geometry. Unambiguous and simple definitions of reference points are required. Radiographs are taken to check the position of the applicator. • Level 2: advanced techniques-modern radiotherapy standards. An individual assessment of absorbed doses at reference points, in different volumes (GTV, CTV, PTV), organs at risk, and normal anatomy is needed, based on radiographs taken in well-defined geometry or CT sections. Computer-assisted dose calculations are performed in three planes at different levels/sections, indicating accurately the doses at the chosen reference points, reference volume(s), and other volumes of interest, e.g., high dose volumes. • Level 3: developmental techniques-clinical research. Planning and performance of brachytherapy according to proposed level 3 imply individualized three-dimensional, computer-assisted assessment of the patient anatomy based on sectional imaging (e.g., with CT, MRI, ultrasound). The different volumes, such as GTV, CTV, PTV, and organs at risk, can be identified and localized. This makes possible accurate three-dimensional dose computation at selected reference points and in different volumes of interest. Dose-volume histograms can also be computed.
1. Amersham International pic, Amersham Laboratories, White Lion Road, Amersham, Buckinghamshire HP7 9LL, UK. 2. CIS Bioindustries (Compagnie ORIS Industrie S.A.), CEN Saclay, 91190 Gif-sur-Yvette, France. 3. Mallinckrodt Medical B.V., Westduinweg 3, NL-1755 LE Petten, The Netherlands.
As in external-beam therapy, the limits between the three levels proposed for reporting in brachytherapy are not definitely fixed, but may vary, in time, with the development of the imaging, computation, and dosimetry techniques.
REFERENCES 1. International Commission on Radiation Units and Measurements (1985) Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology, ICRU Report 38,7910 Woodmont Avenue, Bethesda, Maryland 20814, USA. 2. International Commission on Radiation Units and Measurements (1997) Doseand Volume Specification for Reporting Interstitial Therapy, ICRU Report 58,7910 Woodmont Avenue, Bethesda, Maryland 20814, USA. 3. International Commission on Radiation Units and Measurements (1976) Determination of Absorbed Dose in a Patient Irradiated by Beams of X or Gamma Rays in Radiotherapy Procedures, ICRU Report 24,7910 Woodmont Avenue, Bethesda, Maryland 20814, USA. 4. Mijnheer, B.J., Battermann, JJ. and Wambersie, A. (1987) What degree of accuracy is required and can be achieved in photon and neutron therapy? Radiother. Oncol., 8, 237-52. 5. International Commission on Radiation Units and Measurements (1978) Dose Specification for Reporting External Beam Therapy with Photons and Electrons, ICRU Report 29,7910 Woodmont Avenue, Bethesda, Maryland 20814, USA. 6. International Commission on Radiation Units and Measurements (1993) Prescribing, Recording and Reporting Photon Beam Therapy, ICRU Report 50,7910 Woodmont Avenue, Bethesda, Maryland 20814, USA. 7. International Commission on Radiation Units and Measurements (1999) Prescribing, Recording and Reporting Photon Beam Therapy (Supplement to ICRU Report 50), ICRU Report 62,7910 Woodmont Avenue, Bethesda, Maryland 20814, USA. 8. Paterson, R. and Parker, H.M. (1934) A dosage system for gamma ray therapy. Br.J. Radio!., VII, 592. 9. Paterson, R. and Parker, H.M. (1952) A dosage system for interstitial radium therapy. Br.J. Radiol., 25,505-16. 10. Meredith, W.J. (1967) Radium Dosage: the Manchester System. Edinburgh, Livingstone. 11. Quimby, E.H. and Castro, V. (1953) The calculation of dosage in interstitial radium therapy. Am.J. Roentgenol., 70,739-49. 12. Pierquin, B. (1964) Precis de Curietherapie, EndocurietherapieetPlesiocurietherapie. Paris, Masson. 13. Pierquin, B., Dutreix, A., Paine, C, Chassagne, D., Marinello, G. and Ash, D. (1978) The Paris System in interstitial radiation therapy. Acta Radiol. Oncol., 17, 33^7.
References 101
14. Pierquin, B., Wilson, J.F. and Chassagne, D. (1987) Modern Brachytherapy, Masson, New York.
31. Dutreix, A. and Wambersie, A. (1975) Specification of gamma-ray brachytherapy sources. Br.J. Radiol.,48,1034.
15. International Union Against Cancer (1997) TNM Classification of Malignant Tumours, 5th edn, ed. L.H.
32. Comite Francais de Mesure des Rayonnements lonisants (1983) Recommendations pour la Determination des Doses
Sobin and C.H. Wittekind. New York, Wiley-Liss and Sons. 16. International Union Against Cancer (1990) TNM Atlas, Illustrated Guide to the TNM/pTNM, Classification of malignant tumours, 3rd edn, ed. 0. Spiessl et al. Berlin, Springer Verlag.
Absorbeesen Curietherapie, Rapport du Comite Francais 'Mesure des Rayonnements lonisants' No. 1. Paris, Bureau National de Metrologie. 33. BCRU (1984) Specification of brachytherapy sources, Memorandum from the British Committee on Radiation
17. American Joint Committee on Cancer (1988) Manual for Stagingof Cancer, 3rd edn, ed. O.H. Beahrs, D. Henson,
Units and Measurements. Br.J. Radiol., 57,941-2. 34. AAPM (1987) Specification of Brachytherapy Source
R.V.P. Mutter and M.H. Myers. Philadelphia, J.P. Lippincott. 18. Dutreix, A., Marinello, G. and Wambersie, A. (1982)
Strength, AAPM Report No. 21. New York, American Institute of Physics. 35. Netherlands Commission on Radiation Dosimetry (1991)
Dosimetrieen Curietherapie. Paris, Masson. 19. Anderson, LL, Hilaris, B.S. and Wagner, LK. (1995) A
Recommendations for Dosimetry and Quality Control of
normograph for planar implant-planning.
Radioactive Sources used in Brachytherapy, NCS Report 4,
Endocuriether./Hypertherm. Oncol., 1,9-15.
Netherlands Commission on Radiation Dosimetry,
20. AAPM (1993) Remote Afterloading Technology, AAPM Report No. 41. New York, American Institute of Physics. 21. Wu, A., Ulin, K. and Sternick, E.S. (1988) A dose homogeneity index for evaluating 192-lr interstitial implants. Med. Phys.,15,104-7. 22. Paul, J.M., Koch, R.F. and Philip, PC. (1988) Uniform analysis of dose distribution in interstitial brachytherapy dosimetry systems. Radiother. Oncol., 13,105-25. 23. Saw, C.B. and Suntharalingam, N. (1991) Quantitative assessment of interstitial implants. Int.J. Radial Oncol. Biol.Phys., 20,135-139.
Bilthoven. 36. British Institute of Radiology (1993) Recommendations for Brachytherapy Dosimetry. Report of a Joint Working Party of the BIR and the IPSM. London, British Institute of Radiology. 37. Nath, R., Anderson, LL, Luxton, G., Weaver, K.A., Williamson, J.F. and Meigooni, A.S. (1995) Dosimetry of interstitial brachytherapy sources: recommendations of the AAPM Radiation Therapy Committee Task Group No. 43. Med. Phys., 22,209-34. 38. Societe Francaise des Physiciens d'Hopital (1995) Controle
24. Neblett, D., Nisar Syed, A.M., Puthawala, A.A., Harrop, R.,
de Qua lite en Curietherapie par lridium-192 a Haul Debit
Frey, H.S. and Hogan, S.E. (1985) An interstitial implant technique evaluated by contiguous volume analysis. Hypertherm. Oncol., 1,213-21. 25. Bridier, A., Kafrouni, H., Houlard, J.P. and Dutreix, A.
deDose, Rapport No. 11, Commission de Curietherapie,
(1988) Comparison des distributions de dose en
Gamma-ray Sources, ICRU Report 18,7910 Woodmont Avenue, Bethesda, Maryland 20814, USA. 40. Bernard, M., Guille, B. and Duvalet, G. (1975) Mesure du
in Radiotherapy, IAEA SM-298/23, IAEA, Vienna.
debit d'exposition lineique nominal des sources a une
McCrae, D., Rodgers, J. and Dritschilo, A. (1987) Dose-
dimension, utiliseesen curietherapie./ Radiol. Electrol.,
volume and complication in interstitial implants for breast carcinoma. Int.J. Radial Oncol. Biol. Phys., 13, 525-9. 27. Anderson, LL (1986) A 'natural' volume-dose histogram for brachytherapy. Med. Phys., 13,898-903. 28.
Measurements (1970) Specification of High Activity
curietherapie interstitielle autour des sources continues et discontinues. International Symposium on Dosimetry 26.
Institut Curie, 26 rue d'Ulm, 75231 Paris Cedex. 39. International Commission on Radiation Units and
56,785-90. 41. Ling, C.C. and Gromadski, Z.C. (1981) Activity uniformity of 192lrseeds. Int.J. Radial Oncol. Biol. Phys., 7,665-9. 42. Kolkman-Deurloo, I.K.K., Visser, A.G., Niel, C.G.J.H, Driver, N. and Levendag, P.C. (1994) Optimization of interstitial
Laarse, R. van der and Prins, T.P.E. (1994) Comparing the
volume implants. Radioth. Oncol., 31,229-39.
Stepping Source Dosimetry System and the Paris System
43. Dutreix, A. and Wambersie, A. (1968) Etude de la repartition des doses autour de sources ponctuelles
using volume-dose histograms of breast implants. In Brachytherapy from Radium to Optimization, ed. R.F. Mould, J.J. Battermann,A.A. Martinez and B.L Speiser. Veenendaal, The Netherlands, Nucletron, 352-72. 29. Merrick, G.S., Butler, W.M., Dorsey, A.T., and Walbert, H.L (1997) Prostaticconformal brachytherapy: 125l/103Pd postoperative dosimetric analysis. Radial Oncol. Invest., 5,305-13. 30. National Council on Radiation Protection and Measurements (1974) Specification of Gamma-ray
alignees. Acta Radiol., 7,389--tOO. 44. Henschke, U.K., Hilaris, B.S. and Mahan, G.D. (1966) Intracavitary radiation therapy of cancer of the uterine cervix by remote afterloading with cycling sources. Am. J. Roentgenol.,96,45. 45. Joslin, C.A., Liversage, W.E. and Ramsay, N.W. (1969) High dose-rate treatment moulds by afterloading techniques. Br.J. Radiol.,42,108. 46. Joslin, C.A., Smith, C.W. and Mallik, A. (1972) The
Brachytherapy Sources, NCRP Report 41,7910 Woodmont
treatment of cervix cancer using high activity 60-Co
Avenue, Bethesda, Maryland 20814, USA.
sources., Br. J. Radiol., 45,257.
102 Dose specification and reporting: the ICRU recommendations 47. von Essen, C.F. (1980) Clinical application of the Brachytron: the San Diego technique for treatment of cancer of the cervix. In High Dose-rate Afterloading in the
48.
uterine cervix. In Textbook of Radiotherapy, ed. G.H. Fletcher. Philadelphia, Lea & Febiger, 720-851. 53. International Atomic Energy Agency (1972) Atlas of
Treatment of Cancer of the Uterus, ed. T.D. Bates and R.J.
Radiation Dose Distributions, Vol. IV, Brachytherapy
Berry. British Journal of Radiology Special Report 17, p. 117.
Isodose Charts. Sealed Radium Sources, ed. M. Stovall, L.H.
Hall, E.J. (1994) Radiobiologyfor the Radiologist,
Agency, Vienna.
Philadelphia, J.B. Lippincott Company. 49. van Limbergen, E., Chassagne, D. Gerbaulet, A. and Haie, C. (1985) Different dose rates in preoperative endocavitary brachytherapy for cervical carcinoma./ Eur. Radiother., 1,21-7. 50. Leborgne, F., Fowler, J.F., Leborgne, J.H., Zubizaretta, E. and Chappell, R. (1997) Biologically effective doses in medium dose rate brachytherapy of cancer of the cervix. Radial Oncol. Invest., 5,289-99. 51. Chassagne, D. and Horiot, J.C. (1977) Propositions pour une definition commune des points de reference en curietherapiegynecologique.y. Radiol. Electrol., 58,371. 52. Fletcher, G.H. (1980) Squamous cell carcinoma of the
Lanzl and W.S. Moos. Vienna, International Atomic Energy 54. Groupe Europeen de Curietherapie-European Society for Therapeutic Radiology and Oncology (1998) Workshop ICRU 38: The basis for a revision (Dir. R. Potter), Napoli, 11-13 May. 55. Potter, R., van Limbergen, E., Gerstner, N. and Wambersie, A. (2000) Survey of the use of the ICRU 38 in recording and reporting in brachytherapy of cervical cancer. Radiother. Oncol. 56. Potter, R., Kovacs, G. and Haverkamp, U. (1995) 3D Conformal Therapy in Brachytherapy, 8th International Brachytherapy Conference, Nice (France), 25-28 November 1995, Nucletron-Oldelft, PO Box 930,3900 AX Veenendaal, The Netherlands.
7 Afterloading systems A. FLYNN
7,1
INTRODUCTION
The aim of this chapter is to examine the various afterloading systems that are currently available and also some that are perhaps not strictly currently available but have been used in the recent past. It is not intended to be a complete itemization of the subject, but rather an overview of the main types of equipment and the uses to which they are generally put. It is anticipated that this chapter will help the newcomer to this field to see which systems are particularly suited to each treatment site and method of treatment. It is inevitable in an analysis of this nature that there will be omissions, as it is impossible to cover every aspect of each use of each piece of equipment or technique, and the writer apologizes in advance to any user of this equipment if his or her particular technique has been omitted. However, it is hoped that there is sufficient information for a user or potential user to perceive the relative clinical and technical advantages and disadvantages of each method or equipment. No attempt has been made to assess the relative costs of the equipment, as this will vary from one country to another depending on the local supply situation and servicing and delivery costs. However, it will become apparent that some types of equipment are designed to be specific for a particular site in the body, whereas others are more flexible in the way that they can be adapted for use in several body sites, and the financial considerations of whether a particular machine will be more or less cost-effective will depend, in part, on the anticipated number of applications and the case mix. It is stated at the outset that, whilst various suppliers of radioactive sources and manufacturers of brachyther-
apy equipment are mentioned, this should not be taken to imply any specific recommendation of the products of any individual company. The chapter does not make recommendations of the merits and demerits of similar treatment machines made by the different manufacturers. The reader will appreciate that machines which are similar (but not identical) in their mode of operation are available from different manufacturers and suppliers: an example of such a pair of similar machines is the microSelectron-HDR (supplied by Nucletron) and the Varisource (supplied by Varian). It is the responsibility of the prospective purchaser to decide which particular machine of a certain type is best suited to his or her purpose, bearing in mind the cost, safety aspects, supplier's service record, and all other appropriate considerations. The reader is referred to Chapter 1 for further details of the radioactive sources associated with these afterloading techniques, and to Chapters 8 and 9 for a full discussion of commissioning and quality assurance aspects of low, high, and pulsed dose-rate equipment. An excellent review of remote afterloading technology may be found in American Association of Physicists in Medicine (AAPM) Report No. 41 [1], and internationally agreed specifications for the safety requirements of remotely controlled afterloading equipment may be found in reference 2.
7.2 ADVANTAGES OF AFTERLOADING The accelerated development of afterloading from the 1960s onwards was initially driven by the desire to
104 Afterloading systems
improve the radiation protection environment of the staff involved in the provision of brachytherapy treatments. Readers of older textbooks will find pictures of implants using many radium needles, all of which were inserted manually in the operating theatre, with the consequent radiation exposure of the clinicians and other staff, and there are both documented and anecdotal reports of radiation injury to the fingers of radiotherapists. The technique of afterloading involves the placing of the non-radioactive needles, tubing, or applicators into the patient without the presence of the radioactive sources. Often, dosimetry radiography will be performed at this stage using non-radioactive marker inserts in the applicators. The radioactive material is inserted only when all of the preliminary procedures have been carried out and the operator is satisfied that the applicators are placed correctly within the treatment site. In the case of manual afterloading, the sources are introduced into the carrier needles or tubing directly by the operator, using forceps or other appropriate manipulation instruments. This may be done in the operating theatre (for example, in the case of iridium-192 hairpins) or perhaps on the ward (as may be the case for cesium-137 gynecological source trains or iridium-192 wires). In this way, the radiation exposure to operating room and radiography staff may be minimized. An important 'by-product' of the use of non-radioactive materials at this stage is that the radiotherapist is able to spend more time in the placing of the applicators and can therefore obtain an improved geometry within the implant site. Thus, afterloading confers a benefit to the patient, even with this minimal afterloading method. However, it is apparent that with manual afterloading the radiation protection advantage is gained only in the initial phases of the treatment and that, eventually, the sources themselves have to be inserted into the applicators by the operator and the patient has to be cared for by the nursing and clinical staff, with radioactive sources in position, for the duration of the treatment. The use of remote or machine afterloading extends the radiation protection advantage to the whole of the brachytherapy treatment in that not only are the sources placed initially in the applicators by the machine, but they may also be temporarily withdrawn from the patient into the machine's protected safe for the duration of any nursing procedures that the patient may require. In addition,.the treatment machine timer(s) controls the duration of the treatment to a high degree of accuracy. Whereas considerations of radiation protection were the main driving force for the development of afterloading, more recently other advantages have become apparent. For example, high dose-rate (HDR) and pulsed dose-rate (PDR) brachytherapy would not be possible (or would at least be highly inconvenient) without the use of afterloading devices. Also, the modern development of conformal brachytherapy using stepping or
oscillating source positions has been dependent on the availability of computer-controlled, accurate source positioning, which would not be possible without afterloading.
73 DEFINITION OF LOW, MEDIUM, HIGH, AND PULSED DOSE RATES There is no general agreement in the literature regarding the boundaries between low, medium, and high dose rate or even where the relevant dose rates are defined. Both the International Commission for Radiological Units (ICRU) [3] and the AAPM [1] base their definitions on the dose rate at the prescription point or prescription isodose. ICRU 38 is a document concerned solely with intracavitary therapy, so it may be reasonably assumed that it refers to the dose rate at or near point A of the Manchester System, but a similar inference cannot be made for the AAPM categories. The ICRU recognizes three categories: 0.4 Gy Ir1 to 2 Gy Ir1 2 Gy Ir1 to 12 Gy Ir1 greater than 0.2 Gy mhr1 (i.e., 12Gyh-') although it acknowledges that these definitions are debatable. On the other hand, the AAPM defines LDR in terms of '... conventional doses of about 10 Gy are delivered daily ...,' which implies a prescription dose rate of about 0.5 Gy Ir1. An HDR category is defined as having a prescription dose rate greater than 0.2 Gy min~', which is the same as the ICRU definition of HDR. MDR is defined as being 'between LDR and HDR,' but the boundary between LDR and MDR is not defined. It is interesting that the dose rate of about 1.5 Gy Ir1 that has been frequently used for cervix treatment since afterloading was introduced, and for which radiobiological considerations require a dose rate correction factor to be used, is actually less than the lower boundary for MDR as defined by the ICRU, and is therefore regarded as LDR by this authority, although it could be classed as MDR according to the AAPM definitions. However, this dose rate is often colloquially called MDR. Another definition may arise from the standpoint of radiation protection, where the interest is in the environmental levels of radiation around equipment rather than the clinical dose rates. For example, the Guidance Notes for the United Kingdom's Ionising Radiation Regulations [4] defines equipment giving a dose rate of less than 10 m Gy h-1 at 1 m as LDR; radiation levels greater than this are HDR. This boundary corresponds approximately to a dose rate at point A of 0.4 Gy mur1, which is not the same as the ICRU and AAPM boundary. In practice, these differences are somewhat academic, as HDR machines generally operate at dose rates well Low dose rate (LDR): Medium dose rate (MDR): High dose rate (HDR):
Manual afterloading systems 105
above these boundaries, typically at around 2 Gy mhr1, which is well within the HDR category as denned by all the aforementioned documents. The boundary between LDR and MDR is more questionable. In any event, the only safe practice when reporting radiotherapy is to state exactly the dose, dose rate, and fractionation used, as recommended in the ICRU Report 38 [3]. The idea behind pulsed dose-rate (PDR) brachytherapy is to replace continuous low dose-rate brachytherapy (CLDR) by a series of 'pulses' of higher dose-rate treatment. Brenner and Hall [5] and Fowler and Mount [6] have published analyses of studies of the radiobiology of these modalities, from which come recommendations that, to obtain equivalence to CLDR, the PDR should give the same overall dose in the same overall time, provided that the pulse interval is about 1 h, the length of each pulse should be not less than 10 min, and each pulse should give a dose of about 0.5 Gy, i.e., a dose rate of not more than about 3 Gy h-1 within the pulse. Some early studies of the use of PDR have been reported [7,8]. The main advantage of PDR is technical, in that it enables the equivalent of CLDR to be given with a single stepping source, thereby increasing the range of active lengths and dose distributions that may be obtained.
7A MANUAL AFTERLOADING SYSTEMS
7.4.1 Iridium wires Iridium wire has an external diameter of 0.3 mm and is generally supplied as a loosely wound coil containing a length of 500 mm, although other lengths may be available to special order. In cross-section there is a central 'core' of iridium/platimim alloy of diameter 0.1 mm, which is surrounded by a sheath of platinum of thickness 0.1 mm, giving the overall diameter of 0.3 mm, as stated above. It is available in a range of activities; for further details, the reader is referred to Chapter 1. In use, the wire is cut to the appropriate lengths required for the particular application. Before being afterloaded into the needles or 'outer' tubing implanted in the patient, it is normal practice for the wire itself to be encapsulated into the so-called 'inner' tubing. Once the wire has been fed into its encapsulating tubing, the latter is deformed slightly, either mechanically (for example by the Amersham crimping tool or the Amersham Iridium Wire Loader crimping tool) or by heat (for example, by the now obsolete TEM Iridium Wire Loader), to provide a seal in the tubing at each end of the wire which fixes the wire into the tubing. Nonactive ends of empty tubing may be left at each end of the wire and these may be used to control the position of the active material within the 'outer' tubing or needles. Equipment to facilitate the preparation of iridium wires is commercially available.
The implantation of the carrier tubing into the treatment site may be performed either with flexible nylon tubing or with rigid needles, depending upon the circumstances of the particular case. Surgical methods of placing the tubing in the implant site have been described by Pierquin et al. [9]. The encapsulated iridium wires are held in place for the duration of the implant by crimped lead discs separated from the patient's skin by nylon balls. These may be easily removed to allow removal of the wires from the patient at the end of treatment. The main problem with the use of these flexible tubes is that it is difficult to achieve good implant geometry, as the tubes do not generally remain straight in the patient. The use of rigid needles (and, where possible, templates) permits improved implant geometry: this method is preferred where possible.
7.4.2 Iridium hairpins These have a larger diameter than the wires and are therefore more rigid. The overall diameter is 0.6 mm, this being made up of a 0.4 mm diameter central 'core' of iridium/platinum alloy, surrounded by a platinum sheath of thickness 0.1 mm. They are available in a preformed 'hairpin' shape, as shown in Figure 7.1. The length of each 'leg' of a hairpin is 60 mm, and this may be cut down to the required length just prior to implantation. Single pins may be available to special order. Hairpins are implanted into the tissue with the aid of slotted hairpin guides. The guides are implanted and their positions checked by radiography. If they appear satisfactory, the hairpins are inserted into the guides, which are then removed, leaving the hairpins in the tissue. These are secured by a suture around the crosspiece of the hairpin.
Figure 7.1 Indium wire hairpin (courtesy Nycomed-Amersham pk).
106 Afterloading systems
Hairpins are particularly useful in head and neck implants where the implant site is often only accessible from one end. The radioactive 'bridge' across the top of the hairpin provides an effective 'crossing source,' which allows the reference isodose to be brought up to the level of the mucosa.
7*43 Iridium ribbons In North America, iridium-192 is available in the form of 'ribbons.' A ribbon consists of 12 seeds loaded into a nylon carrier, the seeds being placed at 10 mm intervals. They are used in a similar way to iridium wires in that the ribbon is afterloaded into previously positioned carrier tubing in the implant site. The ribbon may be shortened to the required active length by cutting between the seeds. One advantage over conventional wire is that it becomes unnecessary to cut through the active material itself when preparing the sources. Otherwise, the method of use is similar to that for the iridium wires already described. 7.4*4 Iodine seeds Iodine seeds are mainly used for non-afterloaded techniques and are therefore beyond the scope of this chapter. Examples of their use are for transperineal implantation of the prostate, which has been described by Blasko et al. [10], and for the manufacture of ophthalmic applicators. However, the treatment of brain lesions using removable afterloaded high activity seeds has also been described [11]. 7*4*5 Tantalum wire This material was similar in construction to iridium wire, which has now superseded it, owing to the latter's greater specific activity. It was used in the late 1940s and early 1950s as a source for interstitial implants.
the Manchester ovoids), but they are cylindrical in shape and incorporate tungsten rectal and bladder shielding. The colpostats are linked external to the patient and may move with a 'scissor' action to allow the separation between them to be varied. A further modification was reported and evaluated in 1985 [ 14] to enable the applicator system to be used with the Selectron-LDR afterloading unit. 7*4*7 Amersham Gynecological System A manual afterloading system is supplied by NycomedAmersham. The applicators are based on the Manchester System and are designed to allow the ovoids to be separated by a 'washer' or 'spacer' or be used 'in tandem.' There is a choice of three sizes of applicators. The applicators are made of semi-flexible plastic tubing and are supplied pre-sterilized and are intended to be disposed of after a single use. The uterine tube and ovoids are linked together, but they can slide longitudinally with respect to one another to accommodate differing anatomy. In contrast to the traditional Manchester System, the ovoids lie with their axes parallel to the vaginal axis. The source trains used with these applicators consist of a flexible helical spring, which is loaded with an arrangement of miniature cesium-137 sources and spacers. A number of standard source train arrangements are available and, if required, the supplier can make up trains to the customer's requirements at the time of purchase. Normally, therefore, a selection of trains would be needed to cover the variations in source requirements envisaged. The handle of each source train is marked with a code to aid identification, as shown in Figure 1.2 in Chapter 1. Figure 7.2 shows a typical applicator set. Other applicators are available based on the Fletcher and the Henschke systems, but these use the standard cesium-137 tubes. This system is described in more detail in Chapter 1.
7*4*6 Fletcher-Suit applicators These applicators for treatment of the uterine cervix developed from a non-afterloading system designed by Fletcher [12] and modified in the early 1960s by Suit [13]. Suit's modification allowed the use of afterloaded radium sources, now superseded by cesium-137 sources, held in a hinged carrier, which could be inserted into the tubes. The source arrangement has similarities to the traditional Manchester System (see Chapter 4) in that it employs a line source in the uterine canal and two vaginal sources, one placed in each lateral fornix. In the Fletcher-Suit system, the vaginal sources are called 'colpostats' rather than 'ovoids.' There are three sizes of colpostat, 20 mm, 25 mm, and 30 mm diameter (similar to
Figure 7.2 Amersham manual afterloading system for the uterine cervix; plastic applicator set (courtesy NycomedAmersham pic).
Lose dose-rate remote systems 107
7.5 LOW DOSE-RATE REMOTE SYSTEMS The reasons behind the introduction of remote afterloading systems are considered in section 7.2. However, it will also be apparent from the foregoing descriptions of manual afterloading systems using pre-prepared source trains that another significant disadvantage to this method is the limited availability of source arrangements. An institution would generally have only a small number of source trains, which limits the number of applicator combinations and dose distributions that may be employed. Remote afterloaders were therefore designed to attempt to overcome this restriction by increasing the number of available source patterns. In earlier machines, this was done by having a larger number of preloaded trains, whereas later machines allow the user to compose source trains in the required pattern at the time of use, or by using a single source whose movement through the applicators can be controlled to give the desired dose distribution. This section reviews some of the LDR afterloaders and their applications.
7.5.1 Curietron The Curietron is one of the older type of afterloaders. It was designed and manufactured in France and was used during the 1960s and 1970s, though it is now largely obsolete. The machine was used for the treatment of the uterine cervix. It employed pre-loaded flexible source trains, consisting of cesium-137 sources and spacers in various combinations. The trains were mechanically coupled to drive motors and up to three trains could be transferred to the patient applicators. The treatment exposure of each train could be independently timed and the treatment could be readily interrupted to allow for nursing care. The treatment unit contained a safe for the source trains, to which they were withdrawn during interruptions and at the end of the treatment. The capacity of the main treatment unit was limited to four source trains, so the Curietron also had a 'secondary' radiation sources safe, separate from the main treatment unit, which housed extra source trains, thereby increasing the range of treatment dose distributions that could be obtained. When the applicators and treatment requirements for a particular application were known, the appropriate source trains were transferred to the main treatment unit, from whence they were subsequently driven into the applicators, as described above. The radioactive sources were cesium-137 pellets of length 5.3 mm and diameter 1.8 mm. These were loaded into source holders, with spacers, to give a variety of active lengths. The use of this machine is described in reference 15.
7.5.2 Selectron-LDR/MDR Whilst a number of types of afterloading machines predate the Selectron-LDR (Nucletron), it could be argued that this machine was the first LDR afterloading machine to achieve worldwide acceptance. It was developed by Nucletron during the late 1970s and has been in extensive clinical use since around 1980. There are over 100 installations around the world and the reader will find numerous references in the literature relating to its clinical use. It was designed initially for the treatment of the uterine cervix [16], but it has also been used for intraluminal and surface applicator treatments. The Selectron-LDR was designed to circumvent the aforementioned difficulty by allowing the user the flexibility of being able to construct source trains as required for a particular insertion. To this end, it contains up to 48 cesium-137 sources of external diameter 2.5 mm (see Figure 1.3 in Chapter 1) and a large number of inactive spacers, also of diameter 2.5 mm. The sources and spacers are initially stored in their respective compartments of the main safe. When a source train is programmed and composed by the user, the machine selects sources and spacers in the correct order, as required, and places them in a vertical column in the so-called 'intermediate safe.' Three-channel and six-channel versions of the machine are available, so this process is repeated until all the required channels have been prepared. When all channels have been composed, they may be pneumatically driven through flexible transfer tubes into the treatment applicators. The trains may be withdrawn into the intermediate safe during planned treatment interruptions, under alarm conditions, and finally at the end of the treatment. Each channel may be independently timed. Source activities between 20 mCi (740 MBq) and 40 mCi (1480 MBq) are available. Most users opt for the higher activity sources, which typically give a dose rate to the Manchester Point A of about 1.5-1.7 Gy tr1, putting it in what might be called the MDR category. The gynecological applicators are constructed of stainless steel, the tubing being 6 mm external diameter. They are available in various configurations, including the Manchester set, the Fletcher set (which incorporates shielding in the ovoids), the Henschke set, and a ring applicator set in which the vaginal component is in the form of a ring of sources around the cervical os. There are also applicators for vaginal and endometrial treatments. The open ends of the applicators are mechanically coded to ensure that they connect to the correct transfer tubes. The Selectron-LDR has also been used for the treatment of the oesophagus [17] and nasopharynx [18].
7.5.3 microSelectron-LDR High dose-rate afterloaders have now mainly superseded this machine. It is an LDR system that can position
108 Afterloading systems
radioactive sources into up to 18 treatment catheters simultaneously, and has been typically used for LDR implant therapy. The catheters have an external diameter of 2 mm and may be either flexible tubes or rigid needles. Flexible transfer tubes connect the catheters to the treatment unit. As with most other afterloaders, each catheter may be timed independently. A choice of source systems is available. Originally, the microSelectron-LDR was used with iridium wires. These were made from the standard coils of iridium wire (as described in section 7.4.1) and had to be made up to the required lengths and attached to the drive cables, using a preparation station supplied with the machine. A range of lengths would be made up to ensure that wires suitable for any proposed treatments were available. The treatment unit itself could store up to 18 wires; any further lengths were stored in a separate sources safe, from where they could be transferred to the treatment unit when required, in a manner reminiscent of the Curietron. The main disadvantage of these was the short half-life of iridium-192, which meant that new sources had to be prepared every few weeks. Later, miniature cesium-137 source trains were introduced, which overcame this disadvantage. Reference 19 describes an example of the use of this system.
7*5.4 Buchler System This machine has been used in both low dose-rate and high dose-rate versions, and in single-channel and threechannel configurations, and is intended principally for the treatment of the uterine cervix. It uses either cesium 137 or iridium-192 sources. Each channel is treated by a single source, rather than a train of sources as used in the machines described above. All sources are mechanically afterloaded, but the central source of a three-channel unit (or the only source of a single-channel unit) is mechanically coupled to a drive system which controls the position of the source in an oscillating manner within its catheter, the range and pattern of movement of the source being used to provide the required dose distributions. An eccentric cam within the drive system, the shape of which determines the position of the source at any instant, controls the oscillating movement of this source. The main advantage of this system is its reproducibility, it being necessary only to select the appropriate cam corresponding to the dose distribution required. However, the cams have to be specially made for each dose distribution, so it is inflexible in use, as changes in dose distribution cannot be implemented at short notice. Typical source activities are 300 mCi (11.1 GBq) of cesium-137 for low dose-rate use, but activities up to 4Ci (148 GBq) of cesium-137 or 20 Ci (740 GBq) of iridium-192 were used in high dose-rate versions.
7.6 HIGH DOSE-RATE SYSTEMS High dose-rate afterloading has become increasingly popular in recent years. The main advantage is the increased rate at which treatments can be carried out, which is particularly important when a high patient throughput is required. Treatment may be given in minutes instead of several hours or days, and may be given on a day-case basis in many instances. The short treatment times allow rigid rectal retractors to be used, thereby reducing the rectal dose compared with LDR systems. However, these advantages must be offset to some extent by the fact that the treatment has to be fractionated, so a patient may need several of these, albeit shorter, treatments within a course of treatment. Also, there may be a clinical disadvantage in that a patient may need several anesthetic episodes during the course of treatment, depending on the insertion procedure being performed. Also, the treatment rooms required for HDR equipment need to be more substantial than for LDR equipment, due to the extra radiation protection required, making HDR installations generally more expensive. Reference is made in the clinical sections of this book to some of the treatment regimes used in HDR brachytherapy; these, of course, are very different from those used for the equivalent LDR therapy due to radiobiological considerations. Now that much experience has been gained in the clinical applications of HDR therapy, most new installations of afterloading equipment are of this type. The reader is referred to a recently published report of the AAPM Task Group 59, which considered the practice of HDR afterloading brachytherapy [20].
7.6.1 TEMCathetron Although most of these machines have now been decommissioned, the Cathetron is discussed here as it was the first HDR afterloader to be put into general use, particularly for the treatment of the uterine cervix, and considerable valuable data regarding treatment schedules for this condition were obtained using this machine. It was introduced in 1966 and its early clinical use is described by O'Connell et al. [21]. It consisted essentially of a spherical lead safe in which there were channels for nine source trains. The source trains were made up of cobalt60 pellets and inactive spacers; these were made up to the user's specification at the time of purchase and were therefore fixed for the useful life of the sources. The sources and spacers were held in helical steel springs (the source holders), which were welded on to the end of a Bowden cable, at the distal end of which was an 'eye' connector. When in the standby state, the sources rested close to the center of the safe, each in its own channel, with the end of the Bowden cable projecting from the rear of the safe.
High dose-rate systems 109
Outside the treatment room was the control and drive system. This provided three drive cables, each powered by its own electric motor and independently timed. These drive cables entered the treatment room (through a curved track or under the floor, to maintain radiation protection) and could be connected to up to three (of the nine) source trains required for use. At the front of the safe, three hollow transfer tubes were connected to the output ports of the appropriate source trains, and these transfer tubes led to the stainless-steel applicators in the patient. The three drive cables and the three transfer tubes were color coded to enable them to be matched throughout the system, and mechanical and electrical interlocks ensured that everything had to be connected correctly before a source transfer could be initiated. The safe was designed for a maximum content of 50 Ci (1850 GBq) of cobalt-60. Source trains would usually be made up to provide for the various lengths of intrauterine tube and ovoids sizes of the Manchester (cervix) System. Some users also used this machine for the HDR treatment of surface moulds and appropriate source trains would then be included, for example perhaps a single source pellet for use as the center spot of a mould.
7.6.2
Selectron-HDR
The Selectron-HDR (Nucletron BV) operates in a similar general manner to its stable-mate, the Selectron-LDR, in that source trains consisting of 2.5 mm diameter sources and spacers are composed at the time of use, and these are then transferred pneumatically to and from the applicators as required. However, there are differences in the source type and safe design, to accommodate the HDR requirements of this machine. In this case, the sources are cobalt-60 (see Chapter 1), each has a nominal activity of 500 mCi (18.5 GBq), and the machine can contain up to 20 such sources. There are three output channels, as the machine can only be used for one patient at a time. The system is controlled from the operator's desk situated outside the treatment room. The machine is designed specifically for gynecological treatments, and the treatment applicators are similar to (in fact, mechanically interchangeable with) those for the Selectron-LDR. However, for a busy center, much of the time advantage of HDR is lost if pre-treatment dosimetry has to be performed between performing the insertion and starting the treatment, and many users therefore use applicator systems whose geometry within the patient is predictable, allowing the use of pre-calculated treatment plans and standard treatment times. The Joslin-Flynn applicator (Figure 7.3) is an example of such an applicator system; this allows the operator to select one of two intrauterine tube lengths and one of two ovoid positions, and it also gives rectal dose sparing by means of a rigid retractor. Typically, dose rates of
Figure 7.3 Joslin-Flynn afterloading applicator (HDR) for the uterine cervix (courtesy Nucletron BV).
about 2 Gy min-1 at the Manchester Point A are obtained, and the rectal dose may be kept to less than 60% of the Point A dose. 7.63
Stepping source units
The availability of iridium-192 sources that are physically small but that contain typically an activity of 10-20 Ci (370-740 GBq) has led to the development of this type of treatment machine, in which a single source is sequentially stepped through a series of dwell positions in all the treatment applicators in turn, thereby removing the need for several sources or source trains to be present in the machine. There are several machines of this type now available; examples are the microSelectron-HDR (Nucletron), the Varisource (Varian), and various versions of the Gammamed (Isotopen-Technik Dr Sauerein). Whilst there are differences between the different system relating to source design, maximum catheter number and dimensions, number of dwell positions etc., they are sufficiently similar to be dealt with generically for the purpose of this present description. Generally, an encapsulated iridium-192 source is attached to the end of a drive cable. Usually, the machine also contains a 'check cable,' which is essentially a dummy source on its own drive cable. The purpose of the check cable is to be driven out through the transfer tubes and applicators before the source is transferred, in order to check the correct connection of all the components and also for obstructions or tight curves. The check cable may also be used as a simulated source for radiography in some systems. The source and check cable are driven out, when appropriate, by stepper motors to a claimed positional accuracy of ±1 mm. Each machine will treat a number of channels, for example up to 18 for the microSelectron-HDR and up to 24 for the Gammamed, and each channel may be 'treated' by a number of dwell positions, for example, up to 48 for the microSelectronHDR and up to 40 for the Gammamed. The interval
110 Afterloading systems
between the dwell positions depends on the type of machine, but is typically 2.5 mm, 5 mm, or 10 mm. The source diameter is small, typically about 1 mm, so the catheters through which it travels can also be narrow; an external diameter of 2 mm (6 French gauge) or less is common. Also, the length of travel of the source (and check cable) is long, from 1.5 to 2 m depending on the machine. The combination of these two features makes this type of machine very adaptable and it may be used for intraluminal, interstitial, and intracavitary therapy. Each of the individual dwell times in each of the catheters may in general be different, and this gives the user the opportunity to optimize dose distributions to suit the target volume. Optimization is dealt with more fully in Chapter 5. With older versions of these machines, data from the treatment planning computer had to be manually entered into the treatment unit; with multicatheter treatments and many dwell positions per catheter, this required a lot of data to be entered and was prone to errors. Later machines used a program card system to transfer the data, and the latest machines incorporate combined treatment planning and machine control systems into one computer. Iridium-192 has a half-life of 74 days, and source exchanges are required at (usually) 3-month intervals. Frequent source calibrations are therefore required - this and other quality assurance aspects are fully dealt with in Chapter 9. All the machines have various fail-safe systems built into them to reduce the possibility of errors and accidents. At the time of writing, machines of this type have been used experimentally for the emerging technique of intravascular brachytherapy. This is a developing field, so it is inappropriate to go into detail here, but the reader is referred to references 22 and 23 for further information.
2. BSI 5724 Section 2.17 (equivalent to IEC 602-1-17) (1990) Specification for Remote-controlled Automatically-driven Gamma-rayAfterloading Equipment. London, British Standards Institute. 3. ICRU Report 38 (1985) Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology. Bethesda, MD, International Commission on Radiation Units and Measurements. 4. N RPB (1988) Guidance Notes for the Protection of Persons against Ionising Radiations arising from Medical and Dental Use. Didcot, UK, National Radiological Protection Board. 5. Brenner, D.J. and Hall, E.J. (1991) Conditions for the equivalence of continuous to pulsed low dose rate brachytherapy. Int.). Radial Oncol. Biol. Phys.,20, 181-90. 6. Fowler, J.F. and Mount, M. (1992) Pulsed brachytherapy: the conditions for no significant loss of therapeutic ratio compared with traditional low dose rate brachytherapy. Int.J. Radial Oncol. Biol. Phys., 23,661-9. 7. Mazeron, J.S., Boisserie, G., Gokarn, N. and Baillet, F. (1994) Pulsed LDR brachytherapy: current clinical status. In Brachytherapy from Radium to Optimisation. Veenendaal,The Netherlands, Nucletron BV. 8. Swift, P.S., Fu, K.K., Phillips, T.L, Roberts, LW. and Weaver, K.A. (1994) Pulsed low dose rate interstitial and intracavitary therapy. In Brachytherapy from Radium to Optimisation. Veenendaal, The Netherlands, Nucletron, BV. 9. Pierquin, B., Wilson, J.F. and Chassagne, D. (1987) Modern Brachytherapy. New York, Masson. 10. Blasko, J.C., Ragde, H. and Schumacher, D. (1987) Transperineal percutaneous iodine-125 implantation for prostatic carcinoma using trans-rectal ultrasound and template guidance. Endocuriether. Hypertherm. Oncol., 3,
7.7 PULSED DOSE-RATE SYSTEMS The only commonly available PDR system is the microSelectron-PDR (Nucletron), which is an adaptation of the microSelectron-HDR machine. Its external appearance, mode of operation, and safety systems are similar. However, there are two major differences. First, the iridium-192 source contains less radioactivity, typically having an activity of 0.5 Ci (18.5GBq) to 1.0 Ci (37 GBq). Consequently, it is also physically smaller, having an active length of 0.5 mm and an overall length of 2.7 mm. Second, the operating software is different, allowing the source movement to be programmed for the pulsed nature of the treatment, as described in section 7.3.
131-9. 11. Liebel, S.A., Peschel, R.E., Hilaris, B.S., Gutin, P.M. and Wara, W.M. (1990) Principles of Implantation for Brain Tumours, Interstitial Collaborative Working Group. New York, Raven Press. 12. Fletcher, G.H., Shalek, R.J. and Cole, A. (1953) Cervical radium applicators with screening in the direction of bladder and rectum. Radiology, 60,77. 13. Suit, H.D., Moore, E.B., Fletcher, G.H. and Worsnop, B. (1963) Modification of Fletcher ovoid system for afterloading using standard radium tubes. Radiology, 81, 126. 14. Marbach, J.R., Stafford, P.M., Delclos, L and Almond, PR.
(1985) A dosimetric comparison of the manually loaded and Selectron remotely loaded Fletcher-Suit-Delclos utero-vaginal applicators. In Brachytherapy 1984. Veenendaal, The Netherlands, Nucletron BV.
REFERENCES
15. Jackson, A.W. and Davies, M.L (1983) In Radiation Treatment Planning, ed. N. Bleehan, E. Glatsein and J.
1. AAPM (1993) Remote Afterload ing Technology, AAPM Report No. 41. New York, American Institute of Physics, for The American Association of Physicists in Medicine.
Haybittle, New York, Marcel Dekker. 16. Wilkinson, J.M., Moore, C.J., Notley, H.M. and Hunter, R.D. (1983) The use of Selectron afterloading equipment to
References 111
17.
18.
19.
20.
simulate and extend the Manchester System for intracavitary therapy of the cervix uteri. Br.J. Radial., 56, 404-14. Rowland, C.G. (1985) Treatment of carcinoma of the oesophagus with a new Selectron applicator. In Brachytherapy 1984. Veenendaal, The Netherlands, Nucletron BV. Flores, A.D. (1989) Remote afterloading intracavitary irradiation for cancer of the nasopharynx. In Brachytherapy 2, Veenendaal, The Netherlands, Nucletron BV. de Ru, V.J., Hofman, P., Struikmans, H., Moerland, MA Nuyten-van-Deursen, M.J.H.and BattermannJ.J. (1994) Skin dose due to breast implantation for early breast cancer. In Brachytherapy from Radium to Optimisation. Veenendaal, The Netherlands, Nucletron BV. Kubo, H.D., Glasgow, G.P., Pethel, T.D., Thomadsen, B.R.
and Williamson J.F. (1998) High dose-rate brachytherapy treatment delivery: report of the AAPM Radiation Therapy Committee Task Group No. 59. Med. Phys., 25(4), 375-403. 21. O'Connell, D., Joslin, C.A., Howard, N., Ramsay, N.W. and Liversage, W.E. (1967) The treatment of uterine carcinoma using the Cathetron. fir. J. Radiol., 40,882-9. 22. Schopohl, B., Liermann, D., Pohlit, LJ. etal. (1996) 192lr endovascular brachytherapy for avoidance of intimal hyperplasia after percutaneous translumenal angioplasty and stent implantation in peripheral vessels: 6 years of experience. Int.]. Radial Oncol. Biol. Phys, 36,835-40. 23. Nath, R.,Amols, H.,Coffey, C.etal. (1999) Intravascular brachytherapy physics: report of the AAPM Radiation Therapy Committee Task Group No. 60. Med. Phys., 26(2), 119-52.
8 Quality assurance in low dose-rate afterloading ERICD.SLESSINGER
8.1
INTRODUCTION
The advent of remote afterloading brachytherapy devices has required that the scope of quality assurance in brachytherapy be broadened substantially. This is due to the fact that these devices are designed to perform the basic tasks that previously had been directly controlled by staff: source selection, transport and positioning in the treatment applicators, the monitoring of elapsed treatment time, and treatment termination. These remote functions should practically eliminate the radiation exposure to anyone involved in the treatment and care of brachytherapy patients. A comprehensive quality assurance program is necessary to verify that these devices perform in accordance with the manufacturers' specifications to deliver treatment accurately and safely [1]. This chapter describes the methodology for achieving those ends. The process starts with planning for the equipment and the treatment facility, the application for authorization to use the equipment, the coordination of the facility construction and the equipment installation. The facility and equipment must be thoroughly tested, new treatment procedures must be established, and personnel trained and an ongoing quality assurance and equipment maintenance program must be developed. A successful quality assurance program requires a team approach, incorporating the radiation oncologist, medical physicist, device manufacturer, brachytherapy technologist, nurse, dosimetrist, health physicist, and service technician. The experience of this author in low doserate (LDR) remote afterloading has been exclusively with
devices manufactured by the Nucletron Corporation BV, specifically the Selectron-LDR and the microSelectronLDR. The principles of quality assurance that have been applied for those devices serve as a basis for comparable systems.
8.2 PREPARATION FOR A LOW DOSE-RATE REMOTE AFTERLOADING PROGRAM 8.2.1
Equipment selection
The selection of appropriate equipment is one of the first steps toward the establishment of a remote afterloading program. The type of brachytherapy procedures to be performed (intracavitary, interstitial, intraluminal etc.) with remote afterloading should be determined and then matched to the most suitable device. These devices offer a variety of features for consideration, including source types and activities, applicator systems, the number of treatment channels, the means for programming treatment times, and the ability to customize source configurations. Other basic device design features include the mechanism for source transport, the monitoring of correct source positioning, the safety interlocks and fail-safe systems, and the shielded source storage containers. For a more complete description of remote afterloading systems, see Chapter 7. The review, selection, and specification of applicator designs must be done carefully, consistent with the anticipated treatments. This process will directly influence the selection of the source activities and possibly the
Preparation for a lose dose-rate remote afterloading program 113
selection of the treatment machine as well. If a radiation oncology department is converting from manual afterloading to remote afterloading, maintaining similar applicator styles and source activities is important. For example, if the Fletcher-Suit afterloading tandem and ovoid system had been used, then a comparable applicator set, compatible with the remote afterloading device, should be requested. In addition, applicators for treating the vagina, endometrium, oesophagus, bronchus, and nasopharynx should also be considered. With a remote afterloader like the Selectron-LDR, linear sources are replaced by uniform activity spherical sources configured with inactive spherical spacers. These spherical cesium-137 sources have an outer diameter of 2.5 mm and are available with a maximum activity of 40 mCi (1480 MBq or 115 U) per source. When specifying the source inventory for this type of remote afterloader, the following questions should be addressed: • What is the range of prescription dose rates for the different clinical applications that are anticipated? • How many channels may be in use at one time? • How will the active and inactive source pellets be configured to achieve suitable dose distributions? • What is the recommended maximum activity for the main and intermediate source safes? These questions should aid in deciding the total number of sources and the activity of each source. Combinations of eight sources and spacers can be used to compose 2-cm long sequences that can replace 2-cm linear sources. The medical physicist should determine the ideal source strengths to deliver the prescribed dose rates for each clinical application as specified by the radiation oncologist, and a treatment planning computer should be used to verify the design of the source-spacer configurations. At the Mallinckrodt Institute of Radiology, St Louis, Missouri, a six-channel device with a 36-source inventory (36 U/source) has proven to be adequate for enabling two simultaneous intracavitary treatments. Further discussion of spherical source configurations is presented in the clinical implementation section of this chapter. The specification of sources for a remote afterloader that uses drive motors to transport source-cable assemblies requires a different rationale, because the source trains usually cannot be easily reconfigured, if at all. A list of required source trains, their activities, and active lengths must be established that will satisfy most clinical situations. Certain applicators, such as the intrauterine tandem, may require source trains composed with differential linear activity. Their designs can be based on the non-uniform linear activities that had been used for the manually afterloaded treatments. The position of the source in the applicator can also be specified by the user, but the physical constraints of source rigidity and the curvature of the source transport channels may not necessarily allow ideal source alignment. In that case, dosimetric analysis should be used to evaluate the clinical
utility. Several papers [2-5] have reported on source inventory specification, source assembly design, and source-spacer configurations. For interstitial work, the user can select rigid metal needles or flexible plastic needles or catheters. The availability of templates from the manufacturer should be explored, as well as the compatibility of existing templates with the remote afterloading device. For interstitial applications, a permanent inventory of fixed cesium seed source trains or an inventory of temporary iridium wires or seed ribbons, or a combination of both, may be used. The purchase of cesium seed source trains is practical if one can anticipate an appropriate source inventory. The cesium seed source trains are expensive, but can be used over a long time period, whereas the iridium, though requiring frequent purchases, allows the user more options. A rationale for choosing a specific cesium seed ribbon inventory has been reported [6]. Whether cesium or iridium is used for interstitial application, the positioning of the active length within the needle or catheter must be well understood, because these techniques can use varied active and inactive source lengths as well as varied needle or catheter lengths. Special procedures must be established for iridium source preparation and for verification of accurate interstitial source placement. These procedures are reviewed in greater detail later in this chapter.
8.2.2
Site preparation
LOCATION
Careful planning for a remote afterloading facility is very important to achieve convenience and safety for the patient and staff, and to avoid unnecessary costs. Physicians, physicists, nursing personnel, and the manufacturer should all be involved in the planning. Typically, LDR brachytherapy is given in a private patient room on a surgical or cancer nursing division. It is ideal if the brachytherapy room is near to the nurse station and to the brachytherapy preparation laboratory, while being as far as possible from unrestricted patient rooms. Multiple LDR treatment rooms should be adjacent to each other. Figure 8.1 shows an example of a floor plan of an extensive LDR remote afterloading facility. SHIELDING CONSIDERATIONS
The shielding design must be based on the anticipated maximum source activity loadings and duration to limit exposure rates and total exposure in unrestricted areas to less than the maximum level allowed by the user's regulatory agency. In the USA, for example, the United States Nuclear Regulatory Commission (USNRC) limits the radiation exposure to 2 mrem
114 Quality assurance in low dose-rate afterloading
Figure 8.1 Floor plan for a four-treatment device remote afterloading facility at Barnes-Jewish Hospital in St Louis, Missouri. The equipment shown consists of two microSelectron-LDRsfor intracavitary applications, a six-channel Selectron-LDR for simultaneous treatments in two adjacent rooms, and a microSelectron-LDRfor interstitial Indium located in the room opposite to the source preparation laboratory where the two air compressors for the systems are housed.
(0.02 mSv) in an hour to any non-monitored person, with an annual limit of 100 mrem (1 mSv) [7]. The USNRC ALARA (as low as reasonably achievable) program limits exposure in restricted areas to 500 mrem (5 mSv) per year. For LDR brachytherapy rooms that were previously used for manual afterloading, it is possible to use rolling bedside shields to protect unrestricted areas on the same floor, but lead thicknesses in the range of 0.6-1.3 cm should be anticipated for the floors above and below. New construction is best served with shielded walls. The recommendations for room shielding should be made after studying appropriate blueprints and considering the actual location of the brachytherapy patient within the room, as well as the weight limitations of the room [8]. Empirical radiation survey testing of existing room shielding may also be performed to determine if shielding deficiencies are present before installing remote afterloading equipment [9]. The National Council on Radiation Protection and Measurements (NCRP) Report 49 [10] should be used to guide the determination of the shielding specifications that will achieve the required exposure limits. For
more information on this subject, the reader should refer to Chapter 10. FACILITY DESIGN
Room design details for remote afterloading brachytherapy are summarized in Table 8.1. Electrical power to the device should be provided from a dedicated circuit that is also served by emergency power. If compressed air is required, then the decision whether to use in-house or free-standing compressors must be made. Although it may seem simpler to use the hospital air, a dedicated compressor specifically suited to the device can give greater assurance as long as it can be situated far enough from the patient so that its noise is not disturbing. It is possible to install an air compressor in a small hallway enclosure or in the source preparation laboratory, provided there is adequate space for inspection and servicing. An area radiation monitor with an independent secure power supply should be installed adjacent to the implant patient's bed for maximum sensitivity. A remote
Preparation for a low dose-rate remote afterloading program 115
Table 8.1 Treatment room details for low dose-rate remote afterloading brachytherapy Area radiation monitor (with a dedicated check source and battery pack) and remote display Closed circuit TV (monitor at nurses' station) Compressed air system Electric power on emergency power circuit Emergency container and equipment Remote controls with treatment status displays in the hallway and at the nurses' station
tical, so that one patient can be safely visited without disturbing the treatment in the adjacent room. Door locks should be installed to secure the treatment room when not occupied by a patient, because the remote afterloading device and its source inventory generally remain in the room. 'Radioactive Material' and 'Radiation Area' warning signs must be posted on the room door or at the entrance to the brachytherapy ward. These warnings can conveniently be embossed on a hanging door placard designed also to display treatment forms, nurse instructions, and the room diagram.
Treatment tubing support device Door-mounted board for posted warnings and instructions Door interlock system Door lock Intercom to hallway and nurses'station Rolling bedside shields Permanent shielding in floor and above Wall pass-through for two simultaneous treatments from one device
display for the area monitor is necessary, situated either outside or just inside the doorway to alert anyone entering the room about exposed sources before approaching the implant patient. The area monitor is intended to indicate the safe status of the sources in the entered room only. Therefore, it may be necessary to mount the area monitor against a small lead shield on the wall so that it is not sensitive to radiation from an adjoining brachytherapy room. The remote controls for the afterloaders are usually installed in the corridor, adjacent to the patient's room door. Typically, the treatment room door is closed during treatment and a door interlock switch is installed to automatically retract all sources into the intermediate safe in the event that the treatment interrupt button is not activated prior to entering the room. A window in the door can provide only limited visual contact with the patient. It is recommended to install a closed circuit television camera for remote patient observation from the nurse station. In addition to the normal intercom system from the nurse station to the patient, a two-way hallway intercom to the patient is also practical for limiting the number of treatment interruptions. A remote treatment status display at the nurse station is also recommended, which can indicate whether treatment is progressing, is interrupted, or if an alarm status exists. Display of the remaining treatment time should also be available. If one remote afterloading device is capable of treating two patients simultaneously in adjacent rooms, then a wall pass-through is needed for the source transport tubing. The wall opening should be devoid of sharp edges that could damage the tubing. Each treatment room should have its own independent controls installed, when prac-
AUTHORIZATION FOR REMOTE AFTERLOADING
In some countries, it is necessary to apply for authorization to use an afterloader. Registration of the afterloader may also be necessary depending upon the policies of the agency that regulates its use. The regulatory agency should specify what information must be submitted before authorization and/or registration can be considered. If the agency is not very familiar with afterloading technology, the device manufacturer should be able to suggest which information is required to submit, because the manufacturer will have also gone through a similar submission of specifications and information in order to obtain the approval to market the device. Table 8.2 lists the points that should be addressed when submitting a license amendment proposal to the USNRC. Generally, one should not commit in the application to any procedures that are not required by the licensing agency [8]. However, the licensing agency will be seeking assurance that the proposed treatment program will be safe and has been well thought out. INSTALLATION
The installation of remote afterloaders should be coordinated by a physicist or biomedical engineer. The installation process will proceed smoothly as long as the needs of the manufacturer's installation engineer, in-house engineering, clinicians, nurses, and physicists are understood. The process should begin with a pre-installation meeting, at which time the responsibilities of the manufacturer and the hospital are defined and agreed upon. If existing shielded brachytherapy rooms are to be converted for remote afterloading, temporary relocation of brachytherapy patients may be required during the installation process. Appropriate shielding similar to that in regular use should be provided for these temporary areas. If no supplemental temporary shielding is used, vacating adjacent areas will be necessary unless the exposure rates are within the allowable limits. Therefore, radiation surveys are necessary to verify acceptable exposure levels in the unrestricted areas around the temporary brachytherapy rooms. The manufacturer's engineer initially should review and verify receipt of all equipment. The project coordinator should oversee this process, maintain a file of all
116 Quality assurance in low dose-rate afterloading Table 8.2 Information to submit to the licensing agency when requesting authorization for remote afterloading Remote afterloader model and manufacturer Source description (radionuclide, size, the activity, manufacturer, and physical construction) Certification of federal registration of the device
dures and results should be carefully reviewed by the responsible physicist to determine which tests should be repeated and which additional tests need to be added to the acceptance testing procedure.
83
ACCEPTANCE TESTING
The intended clinical applications The intended users, their training, certification, and experience Radiation-detection instruments to be used Facility floor plan and elevation Calculations that demonstrate acceptable exposure levels in unrestricted areas Describe area security, including access to operating keys, door interlocks, and radiation warning systems Describe patient viewing and communication Describe dosimetry equipment, calibration procedures and frequency, leak test procedures and frequency, and the qualifications of those performing the tests Emergency procedures and frequency of mock emergency drills Personnel-monitoring program Describe the titles and locations of procedure manuals Procedures to prevent multiple treatment devices from operating simultaneously Quality assurance program Training of the source exchangers Training and frequency for operators Disposal of decayed sources Adapted from AAPM eport No. 41 [8].
equipment receipts, and verify that the complete order has been received. The engineer should test the equipment prior to installation, verifying such items as correct battery function, electrical parameters, air pressure sensors, and switches. The installation process includes cable routing between the machine and the remote controls outside the treatment room, wiring the door interlock mechanism, and installation of the radiation monitors. Additional cables are routed to the nurses' station to connect to the remote treatment status display and the closed circuit television monitor. If remotely located air compressors are to be used, a pipefitter will install the appropriate lines to a compressed air outlet near to the treatment machine. The manufacturer's engineer should verify that all cables are properly connected and functioning correctly. If several rooms require installation, it may be possible to schedule the work from one room to the next, with the manufacturer's engineer testing and verifying the installation as it proceeds. This will also allow the physicist to begin the evaluation of the installation, possibly prior to the completion of all the rooms. The device manufacturer's equipment test proce-
8.3.1
Introduction
Acceptance testing for a new remote afterloading device must be performed prior to clinical implementation in order to certify that the device performs in accordance with the manufacturer's specifications. If the treatment facility is also new, then it must also be carefully evaluated. This initial testing also establishes a starting point for machine performance evaluation and provides the basis for the development of an ongoing quality assurance program. There have been several acceptance test approaches reported in the literature [8,11-16]. AAPM Report 41 divides the acceptance testing process into several broad categories, including evaluations of the radioactive sources, the mechanical and electrical operation of the remote afterloading device, the radiation monitors and facility, the applicators, and the treatment planning computer. The following sections of this chapter address many of the issues that comprise a thorough acceptance-testing program. The testing sequence begins with new source evaluations. Once the sources are loaded into the machine, radiation surveys are performed. This is followed by the evaluation of the safety features and interlocks, machine performance and applicator function, and finally the treatment planning system. Some of these tests will yield quantitative results, however much of the testing involves verification of correct function.
83.2
Brachytherapy source testing
Several excellent reviews of brachytherapy source testing can be found in the literature [8,12,17]. The process begins with a careful review of the new source certificates. The isotope, its activity or air kerma strength, the uncertainty of the calibration, the active and physical lengths, and source encapsulation must all be verified to assure that the correct sources have been received. The documentation of acceptable leak tests (<5nCi or 200 Bq removable activity) within the previous 6 months should also be checked. The sources should be visually examined through shielded glass, looking for any signs of damage. Any engraved serial numbers or other identifying details should be recorded. If there is any suggestion of damage to any sources or if the sources have not been leak tested within the previous 6 months, new leak tests must be performed. After the sources have been loaded into the remote afterloader, periodic leak testing
Acceptance testing 117
can be accomplished by wiping the inner surfaces of the source transport tubes or, if accessible, the inner surface of the storage container positions. The manufacturer should provide detailed source diagrams indicating the source construction, dimensions, and materials used. These source documents are very important and should be copied and filed so that working copies are available for fast reference and the original documents are safely stored with other source certificates. A bound logbook should be obtained for each remote afterloader for recording tests and measured results, and for documentation of the routine quality assurance testing. The initial entries into this logbook should pertain to source documentation, listing the source model numbers and serial numbers, activities, active and physical lengths, encapsulation material and thickness. Copies of the source certificates and diagrams should contain all of this information. The assigned storage container position for each source should also be recorded when relevant. The new sources should be autoradiographed to evaluate the uniformity of the activity distribution and its location within the physical length. The results are then compared to the source diagrams. The autoradiographs are obtained by placing the sources in uniform close contact with film that is wrapped in light-tight paper. The source position must be maintained for a brief time period, usually 30 s or less, depending on the film sensitivity and source strength. Kodak Ready-Pack localization film is suitable for this purpose. Catheters or other narrow source applicators can be used to maintain precise source positioning. Reference marks such as pinpricks are used to indicate the source location on the developed film, and the area of relatively high optical density can be evaluated in relation to those marks. Jones [12] has described an autoradiograph technique that uses moulded wax to position and support the sources or applicators, with lead foil markers embedded between the wax and the film to provide a radiographic image of identification marks and scales for precise comparisons. These films can be evaluated by visual inspection, using a ruler to compare the stated active length against the film image and the activity location relative to the physical ends of the source. When possible, radiographs of the sources provide additional verification of source construction details and, when combined with the autoradiograph exposure, the image of the physical structure superimposed over the autoradiograph image provides the most complete evaluation. The use of a scanning optical densitometer could provide additional graphic information to identify accurately the center of the source activity and to evaluate the activity uniformity. Typically, the results of the source evaluations should appear practically identical for all sources of the same model. If any variants are discovered, those sources should be isolated and the source manufacturer contacted. The subject of brachytherapy source calibration is covered in Chapter 3.
833
Radiation surveys
Once the documentation and testing of the new brachytherapy sources have been satisfactorily completed, the source inventory is loaded into the main safe of the remote after-loading device. Radiation surveys are then performed to monitor the exposure rate on the surface and at 1 m away from the main and intermediate source safes. By wrapping the surface area around the safes with a sensitive film wrapped in light, tight paper, shielding defects can be detected and the areas where the surface exposure is relatively high can be located. Average and maximum exposure rates at the surface and at 1 m should be documented and should be consistent with the manufacturer's specifications and national guidance. Radiation protection aspects are dealt with in more detail in Chapter 10.
83.4
Safety features
The safety features of the remote afterloading device and the facility must be thoroughly evaluated. Table 8.3 lists Table 8.3 Safety features of a remote afterloading treatment program that should be tested or established and available Source leak testing Radiation exposure rate surveys in compliance with the regulations Radiation warning lights inside and outside the treatment room function properly Independent radiation monitor in the room and its remote display function properly using a check source with normal and with back-up battery power and during treatment Closed circuit TV system and intercom systems Door interlock causes source retraction and cannot resume treatment until the on/off switch is activated and sources cannot be driven out of the safe with the door open Treatment controls Treatment status indicators, outside of treatment room and at the nurses'station Back-up storage batteries used during power failure Disconnected applicator interlock Compressed air lines maintain adequate pressure under load and for treatment duration Back-up compressed air reservoir for source retraction when normal air compression fails Emergency-off switches and equipment for source removal, handling, and storage Treatment bed and bedside shields arranged to optimize protection Hand-held radiation exposure survey meter Radiation warning signs and posted instructions Room and afterloader secured when not in use
118 Quality assurance in low dose-rate afterloading
the safety features to be tested, equipment that should be available, and procedures to be established. These include the treatment controls, safety interlocks, and the response to an electrical power disruption or an air compression failure. The area radiation monitor should be evaluated with its normal power supply and with its back-up battery power. The sensitivity of the radiation monitor to a minimal source loading should be verified as well as its insensitivity to radiation from an adjacent room. The operation of the closed circuit television camera, monitor system, and intercom system should also be checked.
83.5
Source transport and applicator tests
Applicator tests are performed to verify functionality and to document source position within the applicator. Applicators should be easily coupled to and decoupled from the source transport tubing and should be designed to detect when an applicator is not adequately connected. The integrity of the source guide tubes should also be evaluated. Cable-driven source assemblies should have source guide tubes of constant length to assure 1 mm source position accuracy. For pneumatic source transport, the transport tubing should be inspected for leaks, constrictions, and other obstacles. If specific transport tubes are designed to connect only to specific applicators, this should be tested, as well as all possible combinations of applicators and transport tubes to detect faulty connectors. The applicators should be visually inspected to evaluate the mechanical integrity and radiographed with dummy sources in place to clearly document source position with respect to the applicator surfaces and to verify agreement with the applicator design. For certain applicators, more than one radiograph may be necessary to view the source position along different applicator axes. The location and size of internal applicator shields should also be documented. To illustrate a potential problem, Figure 8.2 shows the dummy source position within the microSelectron-LDR shielded ovoid. Due to the rigidity of the source, it assumes the tipped alignment relative to the ovoid axis. This was not considered to compromise its clinical utility in this case. Autoradiographs with sources loaded in the applicators demonstrate the reliability of the radiographic markers to indicate actual source positions. Dosimetry measurements should also be made to evaluate the effects and adequacy of the shields and to determine if any transmission corrections are indicated. The speed of source transport can be determined indirectly by methods based on ionization measurements or can be observed using video equipment. These techniques have been described by Meigooni et al. [18]. The resultant velocity can be used to predict the transit dose to tissues that are close to the source transport tub-
Figure 8.2 Radiographic evaluation of the source position in the microSelectron-LDR shielded ovoid. The dummy source shown here is constructed to appear identical to the actual source. An autoradiograph (not shown) demonstrated agreement between the applicator positioning of the dummy and actual sources. Solder wire is used to delineate the surface of the plastic caps that enlarge the surface diameter to 2.5 and 3.0 cm. These plastic caps do not extend the anterior ovoid surface.
ing. Alternatively, direct measurement of the transit radiation dose can be made using thermoluminescent dosimetry. These measurements should be repeated at least annually. The many treatment interruptions that can occur during LDR remote afterloading treatments should be considered in order to determine the total dose that normal tissues may receive. However, the magnitude of this transit dose should not be significant. To illustrate this point, assume that transported sources come into close proximity to a normal tissue point for effectively 1 s in each transport direction. For a 40-h treatment, the number of interruptions (2 s of close exposure per interruption) to give 1% of the total treatment dose to normal tissue would have to be close to 720 and, depending on how close this point is to the implant site, this transit dose may be relatively low when compared to the dose received from the sources while in the treatment position. If transparent applicators are available, the position of simulated dummy sources and radioactive sources can be observed via closed circuit television within an accuracy of 1 mm. Certain disposable applicators, such as the Norman-Simon capsules, and flexible interstitial needles require quality assurance testing to verify consistent insertion depth. The Norman-Simon capsules should be tested to verify source insertion capability and autoradi-
Acceptance testing 119
ographs obtained to demonstrate the actual source location relative to the capsule tip. For interstitial applications, the appropriate gauge needle or catheter must be established. A 20-cm long, 15-gauge, plastic flexiguide needle has been used for perineal applications using the microSelectron-LDR. All new flexiguide needles should be checked for length accuracy and trimmed if necessary. A special needle obturator can be marked to indicate the 20-cm insertion depth and serve as a guide if trimming is necessary. If new plastic materials are being used, the effects of gas sterilization should be investigated. Shrinkage of the plastic during sterilization can compromise the source transport and reduce the insertion depth.
83*6
Functional tests
Functional tests to be performed are listed in Table 8.4. Additionally, a list of suitable tests can be found in reference 19. For afterloader systems that utilize air compression for source transport, simulated treatments should be performed to evaluate the reliability of the air compression supply under full demand and the minimum air compression that the compressors will allow. Airflow
Table 8.4 Operations that should be verified by functional testing of a remote afterloading device Console functions: key switches, main power, battery power, source on/off, door open/close Programmability of treatment Correct console displays and treatment data printout Printer function Timer accuracy and transit time measurements Sources retract to safe at the end of a preset time, when treatment is interrupted, when electrical power or air compression is lost, when the room door is opened, or when the emergency off button is activated Treatment cannot proceed if applicators are not connected correctly or are obstructed Source selection or configuration (autoradiograph) Source transport and positioning (autoradiograph) Air compressors maintain adequate compressed air Response to loss of air compression Maintenance of remaining treatment time when treatment is interrupted or after a power failure Battery voltage under load is adequate and operating functions are retained under battery power Return of sources to intermediate safe and main safe Accuracy of source decay computations Proper operation of radiation detectors in the afterloader Manual means for source retraction if the remote system fails
sensors in those systems are used to detect whether or not a source loading has been completely transported into the treatment applicator. A control pellet at the distal end of the source assembly or source train should seal the inner tube of the source container, preventing airflow into the source tube when the sources are completely extended into treatment. If this does not occur within a certain time period, the sources should be automatically withdrawn into the intermediate safe, eliciting an alarm condition. Autoradiographs are very important for verification of correct source selection, source configuration, and source positioning. The applicator must be secured as close as possible to the film. This technique is similar to the one described earlier for initial source testing, but now the intent is to determine exactly where the sources are positioned within the treatment applicators during treatment, and this should be verified on a regular basis. The International Standard recommended by the International Electronical Commission (IEC) specifies a maximum variation in source position within an applicator set to be ±2 mm [20]. Williamson [17] has suggested that geometric accuracy of ±1 mm should be achievable. Radiographs with dummy sources loaded should clearly delineate source position in all applicators. Comparisons between simulator radiographs with dummy sources in place and autoradiographs with applicator surfaces delineated provide the basis for establishing source position. However, the distance between the sources and the film can limit the measurement precision of source position reproducibility, because as the applicator diameter increases, the distance to the film also increases, causing the image of the source activity to be less discrete. Autoradiographs for each applicator should be marked so that localization of source position can be correlated to the dummy source radiograph. If the remote afterloader and a radiographic unit can be used together, combining the radiograph and autoradiograph onto one image makes the evaluation of actual source position within the applicator direct, simple and precise. The comparison of this film to the radiograph of the dummy source marker position within that applicator also provides very precise positioning verification. The position of radiographic markers should vary by less than +1 mm. Prior to commencing a treatment, an autoradiograph should be obtained to verify correct source selection and the channel and applicator that will contain each source. This autoradiograph is especially critical for interstitial iridium seed ribbon assemblies, where the correct source design for each channel must be verified. The most direct autoradiograph process involves sources transported into the same treatment applicators as those in the patient. If a more general autoradiograph device is used, the relation between source extension in an applicator and its position in the autoradiograph device must be established. This involves the use of dummy sources that
120 Quality assurance in low dose-rate afterloading
can either be visualized and measured directly when inserted or radiographed to determine where the sources extend relative to reference fiducial marks. For the interstitial sources, this enables not only verification of the number of seeds per channel but also documentation of the inactive length. Tests should be performed to verify that interlocks operate effectively. These tests should include attempting to treat without an applicator properly connected. If an applicator is designed to be connected only to a particular channel, that specificity should be tested. A dummy source used with a device such as the microSelectronLDR can be intentionally unsnapped from the drive cable to test that the absence of the source will be detected upon source retraction. The door interlock must also be tested, by verifying that treatment is prevented unless the door is completely closed, and by verifying that treatment is interrupted should the door be opened. The treatment timing mechanisms must be tested for accuracy, and source transit time should be determined. These tests can be performed using a thimble ionization chamber that is positioned in close proximity to sources in an applicator. Once the sources are transported into treatment, an integrated charge can be measured over a certain time period to determine the static ionization current (coulombs/s). Additional measurements of charge for different programmed treatment times are then used to determine the actual 'measured' time by dividing the measured charge for the programmed treatment time by the static ionization current [17]. A curve-fitting routine is then used to relate the time measured to the time set by the linear relation:
83*7
Acceptance test report
The results of all tests performed during the commissioning of a remote afterloading device and facility should be summarized in a report, which should serve as a benchmark from which the quality assurance program can continue. In addition to the test results, the report should document operational procedures and describe the basic mechanics of the system. If the report is for a new remote afterloading facility, treatment-related procedures and a quality assurance program should also be specified in the report prior to clinical implementation. Some of these clinical procedures are discussed in the following sections. Table 8.5 gives an example of the contents of an acceptance test report.
Table 8.5 Example table of contents for a remote afterloading device acceptance report 1. Institution and location 2. Facility name 3. Remote afterloading device 4. Sources and applicators 5. Tests and results
Source calibration Radiation safety Mechanical Programming a treatment
6. Procedures
Machine start-up and daily quality assurance procedures Nurse's instructions Radiographic localization Computed clinical dosimetry Emergency procedures Treatment completion procedures Patient instruction Room assignment Quality assurance program Physician loading and unloading procedures
Timemeas = (Timeset)(Timer accuracy slope) + Transit time
In general, the measured time should agree with the machine-set time within ±2%, but, as treatment time decreases or the number of interruptions increases, the dose effects due to transit time error increase. Williamson has reported timer error to be of the order of a few seconds, while absolute timer accuracy is usually within ±1%. The distance between the dosimeter and the sources should be kept constant and as small as possible, because this parameter will affect the timer error determination. Another way to determine transit time is similar to the method used for cobalt teletherapy units [21], where integrated charge is measured for one exposure period and then compared to the same total exposure time administered over several on/off sequences. For both methods described, the results will be clinically applicable if the distances between the source and chamber are similar to a typical treatment depth. Additional functional tests are included in Table 8.4, many of which can be verified simply by observation.
7. Summary and recommendations 8. Appendix
Source data Radiation surveys Applicator radiographs Source position autoradiographs Tim ing tests Special forms and techniques License application Manufacturer's field service reports Machine specification Error codes
Quality assurance procedures and clinical implementation 121
8.4 QUALITY ASSURANCE PROCEDURES AND CLINICAL IMPLEMENTATION
8.4.1 Conversion from manually loaded tube sources to source-spacer pellet configurations A methodology for converting manually loaded, 2-cm long sources to source-spacer configurations has been described by Grigsby et al [3]. This approach replaces the 2-cm long sources by groups of eight 2.5-mm source-spacer pellets. To duplicate the inactive ends of the tube source, the first and eighth positions are reserved for spacer pellets. The remaining six positions contain the active sources, distributing them as uniformly as possible. Isodose comparisons of linear sources and active/inactive pellet configurations generated by the treatment planning system should be used to match the two dose distributions. The Fletcher-Suit ovoids designed for the Selectron have six active source positions between the bladder and rectal shields. The active sources are therefore arranged within those positions. Table 8.6 lists some of these source configurations. The concept of using six source positions to specify an active length close to 14 mm is also supported by the analytic approach reported by Williamson [22], which states that a linear array of n point sources, spaced at center-to-center intervals S, approximates a
Table 8.6 Cesium source and spacer arrays developed by Grigsby et al. [3] to replace the previously used 2-cm tube sources
144 U 72 U
-2 - 4 5 - 7 7 _ -2
72 U 144 U 180 U 216 U
1 1 1
2- 3 2 3 2 3
Each 2.5 mm source pellet is nominally 36 U (5 mg radium equivalent).
- 54 - 6 - 5 6 4 5 6
line source of active length n x S, i.e., (6) x (2.5 mm) = 15 mm. Others have also reported on source-spacer pellet configuration schemes for the replacement of an intracavitary Manchester radium source technique [4] and for source optimization in vaginal cylinder applications using a simulated annealing algorithm [5]. The limitations of point source algorithms to model Selectron source configurations have also been reported, and measured dose distributions have demonstrated where the effects of attenuation by inactive spacers, other sources, and applicators are the greatest [23,24]. This is discussed further in the dose computation section.
8.4.2 Conversion from manually loaded tube sources to motor-driven source-cable assemblies The specification of source activities and active lengths for motor-driven remote afterloaders is a relatively straightforward process. Because the sources cannot be reconfigured, one must list all the sources anticipated to be needed. An example of an intracavitary cesium source inventory for a 15-channel, cable-driven, remote afterloader is shown in Table 8.7. Each source should have a documented assigned location in the main storage container so that prescribed sources can be easily identified. The design of a remote afterloader will sometimes limit how closely one can replicate manually loaded brachytherapy technique. For example, if the path of the source transport tube through an applicator such as a Fletcher-Suit-style ovoid will not permit the use of a 2-cm long rigid source, then a shorter source length may be necessary. This was the case when an 8.5-mm active length source replaced a 14-mm active length cesium tube source and, as a consequence, the dose rate on the lateral surface of the ovoid increased by 6% to 15% as the ovoid diameter decreased from a 3 cm to the 1.6 cm mini ovoid [2].
Table 8.7 The intracavitary cesium source inventory for a microSelectron-LDR1
Norman-Simon Capsule Mini ovoid Shielded ovoid Shielded ovoid Tandem Tandem Tandem a
12 2 2 2 1 1 1
8.5 8.5 8.5 8.5 40.0 60.0 60.0
9.11 9.74 18.93 28.50 19.12-9.54 9.54-19.12-9.54 19.12-9.54-9.54
1-12 19,20 21,22 23,24 38 39 40
Purchased by the Mallinckrodt Institute of Radiology for treatments given at Barnes-Jewish Hospital, St Louis, Missouri. MLDR, microSelectron-LDR.
1-12 13,14 13,14 13,14 15 15 15
122 Quality assurance in low dose-rate afterloading
It is important to note a significant treatment limitation with the fixed-style cable-source assembly. There is no adjustment of the source position in the applicator. With manual afterloading, sometimes a spacer of length between 5 mm and 15 mm at the cephalad aspect of the tandem is used to optimize the position of the most distal tandem source in relation to the ovoid sources. Without that option, the flange of the tandem must be set after evaluating the sounding of the uterine depth and reviewing the source lengths that are available. External markings on the tandem can indicate the distal extent of the available source trains for the tandem and the most appropriate source train can be selected.
8.4.3
Iridium seed ribbon preparation
Conventional iridium seed ribbons reinforced with internal plastic filaments can be configured for remote afterloading as long as they have not been previously used for manual afterloading. The depth of the inner lumen of the implanted catheters or needles must be known in order to accurately position the active length at the prescribed treatment area. The depth information is determined either by pre-cutting the catheters for a constant inner depth or by measuring the insertion depth of each catheter with a special 'measuring dummy' or comparable commercial measuring device. A special ribbon preparation station is necessary to prepare the source assembly safely and accurately [25]. The process begins with trimming away any active portion of the ribbon not to be used. This trimming involves cutting the interseed plastic filament leaving a small (2-3 mm) inactive tip that is heated with a special element to achieve a bullet-shaped tip for smooth transport. The inactive proximal end is then cut to the length that will position the active length to span the prescribed treatment length. Accurate ribbon preparation must also account for the ribbon length that inserts into the control pellet and any gap between the control ball and the opening of the needle or catheter. The total source length should be prepared so that at least 2 mm of gap is maintained between the ribbon tip and the inner depth limit. This gap is necessary to ensure that the control pellet seats completely, enabling treatment to proceed. Clinical experience can help to determine the optimal gap. The control pellet also serves as the connection for the source ribbon and the drive cable. The drive-cable length controls the positioning of the source assembly within the shielded area when situated in the intermediate safe. Therefore, the drive-cable length must be determined based on the sum of the active and inactive lengths of each source ribbon. Finally, the completed source assembly is transferred from the preparation station to the source storage container, from where it is subsequently transferred to a channel of the intermediate safe and from there to the treatment catheter. An autoradiograph
must be obtained to verify the details of the source construction as well as to ensure that each source resides in the correct machine channel. A current source inventory list must also be maintained as new sources are created and old ones disposed of. Because of the many steps involved in iridium source preparation, a source preparation form is very practical to ensure that all source details are incorporated. It is important for a user to experiment and establish a system that prepares an iridium ribbon so that, when it is finally configured, it delivers the correct active length to the prescribed treatment length of the catheter or needle. The process of ribbon preparation can be simplified if certain source dimensions can be kept constant, but in many clinical situations that is not possible. Another approach for interstitial source preparation is to maintain a stock of fixed-length sources and vary the length of the catheter or needle. However, this will cause the transfer tubes to have varying radii of curvature, increasing the likelihood of source transfer difficulties [26]. Although interstitial applications can be performed for almost any body site, LDR remote afterloading may be better suited for perineal and breast applications, whereas head and neck applications are generally more difficult. Special attention also needs to be paid to assuring that templates designed to hold the implanted catheters securely are also fixed securely to the patient and, finally, never allow flexible catheters to be cut after the measurements for source preparation have been obtained.
8*4*4
Treatment planning
LOCALIZATION RADIOGRAPHS
The localization radiographs for the evaluation of applicator placement, the determination of applicator source loadings, and the computation of dose distributions are obtained in a way similar to that used for manual afterloading. The radiographic source markers should be inserted in the operating room once the applicators have been secured, and the patient has arrived for simulation ready for filming. A quick inspection of the applicators is advised to verify that the radiographic markers are still completely inserted. The configuration of dummy markers should be documented for general reference. The physicians, simulator technologists, and dosimetrists should all be familiar with the dummy source configurations so that, while reviewing the radiographs, errors are not made due to misunderstanding which markings represent the actual source positions. Documentation of the position of the dummy sources and/or actual sources within applicators is also useful for reference, especially for ovoids, because those dummy source markers will not always be distinct on lateral localization radiographs. Figure 8.3 shows the dummy source positions in the Selectron-LDR Fletcher-Suit-style ovoids and demon-
Quality assurance procedures and clinical implementation 123
Figure 83 Radiograph of the shielded Fletcher-Suit-style ovoid for the Selection demonstrating the available source positions. Positions 1-6, indicated by the square and round radiographic markers, are shielded in the directions toward the bladder and the rectum. This radiograph can serve as a reference for determining source positions for dose distribution computations because the source positions are usually not distinct on the lateral source localization film.
strates that source positions 1-6 are situated between the bladder and rectal shields. Standard active source positions can therefore be established, because the positions relative to the shielding are fixed. The plastic caps that enlarge the ovoid diameter should be evaluated to document the distance from the sources to the vaginal mucosa and to the applicator surfaces adjacent to the rectum and bladder (which are not necessarily equal). For interstitial applications, verification measurements of catheter insertion depth should be made while the patient is in the operating room. If catheters can be premeasured and cut to known lengths or marked in a way that can guide precise cutting, that is a simpler process than measuring catheters of varying lengths. Catheters should extend at least 5 cm from the patient's surface to allow connection to the transfer tubes. DOSE COMPUTATIONS
Low dose-rate remote afterloading does not require special treatment planning equipment. The radiographs used for source reconstruction should be obtained following whichever techniques are convenient and compatible with the user's planning system. However, if a new planning system is obtained specifically for a remote afterloading system, acceptance testing of the treatment planning computer must be performed before any brachytherapy dose calculations are obtained for patient treatments. Evaluations of treatment planning algorithms for remote afterloading have been reported with reference to source reconstruction methods [27] and dose calculation [23,24,28,29]. Results from one report are shown in Figure 8.4. There are several methods for source reconstruction. The accuracy of each technique
must be evaluated before it is used clinically. Source positional accuracy should be within ±2 mm. Details of source calculation algorithms can be found in Chapter 5. The clinical evaluation of computed isodose distributions is no different for manual or remote afterloading. However, once the dose distributions for remote afterloading have been reviewed and the final decisions on source configuration and treatment duration have been documented, it may be necessary to reprogram the treatment times or possibly modify a source configuration. However, unlike manual afterloading, there is no attempt to calculate the precise time of day the implant will be completed, because treatment interruptions occur during the implant course for various reasons, such as patient care and visitors. It has been reported that on average 14% more time is required to complete the treatment, with 4% due to mechanical failures and the other 10% to patient care [2]. TREATMENT PRESCRIPTION
Treatment prescriptions for remote afterloading differ from those for manual afterloading only with respect to the source types and the specification of the treatment machine. This assumes that the source activities available are similar to the manual sources that were used previously. To prescribe a Selectron treatment, a diagram of the 48 available source positions for each of the three channels is helpful. The physician can darken the prescribed active sources with ink, and the number of hours for each channel can be documented adjacent to the source column of each channel. Figure 8.5 shows a prescription form that was designed for both remote and manual afterloading.
124 Quality assurance in low dose-rate afterloading
Image Not Available
Figure 8.4 Dose profiles for a single source in a Selectron applicator (profiles 1, 2, and 3) and for three sources in the applicator (profile 4) for geometries shown in the right-hand inset. The solid profiles labeled 'a' are measured below the applicator; the dashed profiles labeled 'b' are calculated. Profile 4b is the sum of profiles Ib, 2b, and 3b and is valid for the geometry of profile 4a except that the attenuation effects of the stopping screw and spacers are ignored. The left-hand inset shows a correction factor to the calculated dose for the four dose profiles. (Reprinted with permission from Conrado Pla, PhD (1987) in the International Journal of Radiation Oncology, Biology and Physics, 13, 1761-6.)
8.5 THE ONGOING QUALITY ASSURANCE PROGRAM
8.5.1
Treatment verification
Treatment verification procedures must be performed for all remote afterloading treatments. The prescribed treatment is programmed by a brachytherapy technologist, who refers to a copy of the physician's written directive. A physician should check the treatment data before treatment begins. The treatment verification process includes reviewing the machine console displays and printout tape, indicating source selection or configuration, channel selection, and treatment duration. When applicable, the source storage container positions from which sources were drawn for the prescribed channels should also be checked. An autoradiograph should be obtained for each treatment so that source configuration or source selection can be verified for each treatment channel and applicator. It should be noted that source storage containers might allow manual manipulation of
the source locations, which can allow accidental source relocation to a new container position. This could only be detected by means of the autoradiograph. Devices for autoradiography may be commercially available, but a simple device that secures the applicators in close contact to a Ready-Pack film can be easily designed and constructed. It is also possible inadvertently to connect a treatment tube to the wrong channel connector. Verification that source transfer tubes are connected to the corresponding treatment channels is therefore necessary on some afterloaders. Connections should be verified during patient connection by tracing each transfer tube back to the channel locking mechanism. Therefore, the ideal autoradiograph device should be capable of detecting not only source relocations in the storage container, but also errors in transfer tube connections. In comparison, autoradiographs of the Selectron are less likely to reveal source configuration errors that could not have been detected from the machine displays or printout tape, but rather, they serve primarily to document correct source and spacer configuration. Autoradiographs should be reviewed by a physicist on the same day
The ongoing quality assurance program 125 Figure 8.5 This intracavitary prescription form was developed at the Mallinckrodt Institute of Radiology, St Louis, Missouri, for remote afterloading and manual afterloading.
that the treatment is started and filed as a source verification document. Figures 8.6 and 8.7 show two excellent autoradiograph devices that were developed by Dr Ali Meigooni and Dr Jeffrey Williamson. Sometimes, it is necessary to reprogram selected channels that require longer total time. If the reprogramming is done during the nighttime, it may be difficult to have it double-checked. As soon as possible, the reprogramming should be verified by reviewing the treatment history printout. When possible, reprogramming should be scheduled to occur at a time when immediate verification is possible. Procedures should also be established to verify the dose calculations on which the treatment prescription is based. Ideally, a qualified person who has not made the initial calculation performs this check. Items to check
include: a review of the written directive, recalculation of any arithmetic, data transfer from tables and graphs, and correct use of data. Computer-generated dose calculations should be verified by confirming correct source and geometric input parameters [16]. An alternative method compares the manual calculation of a point dose to the computed dose at that same point.
8.5.2
Shield placement and surveys
Prior to initiating treatment, portable bedside lead shields should be positioned, when necessary, as indicated on the established floor markings or room diagram. Radiation surveys in restricted and unrestricted areas contiguous to the treatment room must be
126 Quality assurance in low dose-rate afterloading
Figure 8.6 Verification of correct source selection and position for the microSelectron-LDR channels designated for the tandem and avoids is accomplished with a special autoradiograph device (a). The three applicators can only be coupled to designated transfer tubes and the same is true for the patient's applicators. The resultant autoradiograph (b) can be overlaid directly onto the dummy marker film template by aligning the fiducial pinholes. The development of this device was necessary due to the possibility of changing source positions in the storage containers which could result in an incorrect source loading. (Courtesy of Ali Meigooni and Jeffrey Williamson.)
Figure 8.7 A special autoradiograph device (a) was developed for the Selectron-LDR to facilitate quality assurance testing and to verify the prescribed treatment configuration of active cesium pellets and spacers. The 48 available source positions are indicated by the radiographic markers and the three fiducial pinholes enable a resultant autoradiograph (b) to overlay the device radiograph and easily check the positions where active sources have been sequenced. (Courtesy of Ali Meigooni and Jeffrey Williamson.)
performed immediately after the treatment has begun to verify compliance with the regulations of the governing radiological health agency. The survey results must be documented, including the model and serial number of the survey instrument used and the initials of the person performing the survey.
8.5,3
Machine-testing Schedules
The AAPM [8] recommends that the quality assurance testing schedule be designed by the physicist in charge in accordance with the established regulations and recommendations, and advises that the schedule of testing be
The ongoing quality assurance program 127
frequent enough to guarantee that the equipment will work properly during a therapy session. Prior to starting a treatment, the brachytherapy technologist or physicist should perform tests to verify normal operations of critical components and, once all tasks have been performed, the treatment can be programmed and independently verified by the physician. A listing of daily quality assurance tests is included in Table 8.8. It is useful to establish a form that lists the daily test items and allows for the entering of the treatment date, the test results, and the initials of the individual performing the tests. These daily test results should be maintained in a special quality assurance binder and reviewed regularly. Daily air compressor testing includes checking the oil level, bleeding off moisture build-up, and venting the
system. The pressure levels that activate and terminate the compressor cycle should be recorded. The area radiation monitor and its back-up battery power are checked using a low-activity source (20-40 KBq), with the normal electrical power disconnected. The source transfer tubing should be inspected for defects and cleanliness and the applicator-transfer tube coupling mechanism should operate smoothly and easily. During treatment, loose protective tubing or plastic wrapping should be used to help protect the transfer tubing from becoming soiled. The applicator connectors should be cleaned promptly after each treatment. While verifying the function of the door interlock, other machine functions are also tested, including programmability of a treatment, correct response of the air compression sensors and
Table 8.8 Quality assurance review schedules for low dose-rate remote afterloaders
(Technologist) Air compressors Area radiation monitor Paper-tape supply Treatment indicator lights Door interlock Treatment interrupt switch Intercom system CCTV system Coupling cleanliness Treatment tubing integrity Bedside shields positioned Operating instructions
(Physicist) General Key switch Channel connectors Coupling mechanisms Audio/visual communications Air compressors Room condition Safety Operating instructions Emergency instructions Radiation warning sign Regulatory agency notices Emergency instructions Treatment autoradiograph Functional tests Door interlock Dooroperability Indicator lights Audible signals Bed location Daily quality assurance log Functional Air loss test Power I oss test Alarm indicators Timer versus stopwatch Transit time Time/date printout Source inventory printout Autoradiograph Correct storage locations Correct source sequencing (Maintenance technician) Maintenance review Complete treatment cycle Interlock operations Power supply voltage Air pressure levels Air compressor operation
(Physicist)
Source Leak test (semi-annual) Calibration Treatment-planning parameters Facility Radiation survey Area radiation detectors CCTV system Air compressors Treatment device Main safe radiation survey Intermediate safe radiation survey Area radiation monitor Autoradiograph Source selection Position in applicator Complete treatment cycle Safety Interlock systems Power loss back-up system Timer linearity Air loss back-up system Applicator tests Air leaks Ease and specificity of coupling to transfer tube Total source inventory Source and dummy positioning agreement Internal shield position and dosimetry effects Temporal Timer accuracy Transit time Dose computations Source para meters Point dose comparisons
128 Quality assurance in low dose-rate afterloading
valves, and treatment status indicators. Other items to check include the paper tape printer, the intercom, closed circuit television system, and the accessibility of the operating instructions and emergency instructions. If the outcome of any of the daily tests is not satisfactory, the physicist should be notified to determine whether a repair is required or if the treatment must be cancelled. For example, if the door interlock is disabled, the remote afterloader should not be used. Finally, just prior to treatment, the autoradiograph of the programmed sources is obtained, the bedside shields are properly positioned, and the physician reviews the programmed treatment parameters and initials the paper tape printout if everything is correct. Instructions should be reviewed with the patient regarding the limited range of motion due to the connection to the machine, as well as the importance of minimizing treatment interruptions. With only the patient remaining in the room, the door is then closed and the treatment switch activated, sending the sources into treatment. The radiation exposure in the adjacent unrestricted areas is then measured and recorded. If excessive exposure is measured, the shield position or patient position should be rechecked and, if necessary, the adjacent areas vacated, posted, and secured. Testing by a physicist should be done on a monthly or quarterly schedule. The frequency of the physicist reviews should be based on machine reliability. Initially, a monthly schedule is appropriate, but, if the machine performance proves to be stable and reliable, a quarterly schedule is justified as long as the equipment performance does not diminish. The items included in the physics reviews are listed in Table 8.8. Although some of the tests overlap those included in the daily testing protocol, additional tests are performed to monitor machine performance more extensively. Functional tests of the device, such as response to the loss of air compression and electric power, are checked. These tests should be performed in a manner that ensures that any personnel radiation exposure is kept as low as possible. With the physicist remaining in the treatment room, a spacer train or a minimal number of sources is transferred into treatment. The physicist then disconnects the air supply to the machine and verifies that the back-up compressed air reservoir is activated and retracts the source train immediately and completely. The compressed air line is then reconnected and the back-up power system is checked in a similar way, by disconnecting the power cable from the wall outlet and verifying that the back-up battery system executes the source retraction and also retains the treatment history. Once power has been restored, the treatment should be ready to resume without the necessity of reprogramming. During these tests, the display of the appropriate error codes and alarm conditions should also be verified. Techniques for testing the timer system accuracy and determining source transit time are described above, in
the acceptance testing section. Simpler tests can be performed on a quarterly or monthly basis. Integrating ionization charge over three different time settings and normalizing the charges to one time setting can check timer linearity. The relative charges and time settings are then compared. The agreement should be within 2%. Stopwatch measurements should be compared to the machine timer and the machine calendar should also be checked by inspecting the date and time listed on the treatment printout tape. An autoradiograph should be obtained documenting the accountability of the complete source inventory and that each source is located in its assigned container position. For devices that can sequence sources and spacers, the autoradiograph can also confirm correct sourcespacer sequencing. A quarterly maintenance review by a qualified service engineer is another important component of the quality assurance program. However, it is very important to distinguish between different types of machine maintenance, and who is authorized to perform those services. For example, the USNRC states that authorized personnel are required for installation, replacement, relocation, or removal of sealed sources. In addition, any adjustment to any mechanism on the afterloading device, treatment console, or interlocks that could expose the source, reduce shielding around the source, or affect the source drive controls can only be performed by authorized personnel [16]. Therefore, it is necessary to clarify with the appropriate regulatory agency the maintenance tasks that can only be performed by an authorized service engineer. The maintenance and repair history for each remote afterloader should be well documented. Table 8.8 also includes the items to be reviewed on the quarterly maintenance schedule. The results of the physics and maintenance reviews should be documented on suitable forms that are filed in the logbook designated for each afterloader. On an annual basis, an in-depth physics review includes source calibrations, radiation surveys, applicator tests, radiographic dummy marker evaluations, and treatment dose computations (see Table 8.8). It should also report on the source leak test results that are done at the prescribed intervals. A report is then generated based on the annual review and the treatment history for the year. In addition to this extensive quality assurance testing schedule, the physicist must also determine the appropriate tests to perform following the repair of an afterloader before the device can resume clinical operation. The schedule of routine quality assurance and maintenance testing should be based on the ongoing performance history of the device, and the tests to be performed should also be evaluated. The pre-treatment tests may reveal the most obvious machine problems, but more subtle problems may go undetected. The more extensive quarterly tests should expose those problems. The con-
The ongoing quality assurance program 129
tent of the quarterly review should be based on the frequency of device mishaps. Williamson [17] has reported that quality assurance testing rarely reveals problems involving accuracy of source position or treatment timing, although he does acknowledge that interstitial techniques are much more complex and require more effort to verify correct source positioning. However, he has found that the quality assurance testing can reveal system failures involving interlocks, source transport, source or channel recognition, and back-up battery power. He has proposed a quarterly schedule for physics testing and a quarterly schedule for maintenance review.
8*5*4
Documentation
Treatment documentation can be quite extensive for remote afterloading. The treatment autoradiograph and the treatment printout tape should be maintained for that purpose. The treatment tape lists the entire treatment history, complete with the initial programming, treatment start time, treatment interruptions, error codes, and treatment completion time. The rest of the treatment documentation is applicable to all brachytherapy and includes the prescription/treatment form, computed isodose distributions, point dose computations, and a dose summary report. The physician should date and sign the written record before the treatment has begun. The record should list the radioisotope, treatment site, total source strength, and exposure time or total dose [16]. A quality assurance checklist from start to finish can also serve as a valuable document. One critical item on that checklist should document that a survey was performed to verify that all sources were transported to the storage safe of the remote afterloader after the completion of treatment. The exposure rate measured at the surface of the patient's implanted area should be recorded. It is convenient to file copies of some of the treatment documents in a special 'Quality Management' binder, which can be used to facilitate internal auditing that may reveal weaknesses in the quality management program and thereby continually improve treatment quality.
A continual program of ongoing training for the staff is also very important. The device manufacturer should provide annual refresher training, but this should be supplemented with intramural reviews. The training of all involved personnel must include the requirement for implementing the instructions of the authorized users, consistent with the quality management program and government regulations. The training should be specific to the device model and include radiation protection principles and practice, operating and emergency procedures, and system design. In addition to the formal training in the use of brachymerapy sources that attending physicians and residents have been given, they must also understand how the afterloaders function so that their involvement in treatment verification, programming, and treatment termination procedures is performed correctly. Physics procedures should also be reviewed regularly. The nursing staff need training in the operation of the treatment controls as well as in interpretation of the treatment status displays and the area radiation monitor. They need to know who to contact in the event of a treatment problem as well as emergency procedures. They must also understand the need to control the number and duration of treatment interruptions. A videotape is a convenient way for nurses to review these procedures. Table 8.9 gives a complete listing of the items to be covered in an in-service training program for nurses. Emergency procedures should be established that ensure that the medical physicist or radiation safety officer be notified immediately if sources fail to retract completely. The names of emergency personnel and the means for contacting them while on duty or off duty should be posted at the operating console. The location of emergency equipment should be specified. Emergency equipment should include a portable radiation monitor, shielded storage containers, and long source-handling tools. Supplies for removing applicators should be Table 8.9 Nurse in-service topics for remote afterloaders Remote control button operation Hallway intercom operation
8*5*5 Training personnel for remote afterloading Initially, the manufacturers of a new remote afterloading device normally provide training for the users, operators, and supervisors of the equipment, as part of the installation contract. The training should include applicator description, the function and operation of the devices under normal and emergency conditions, all safety features, radiation protection procedures, suggested quality assurance procedures and, when applicable, a review of the computerized treatment planning system [8].
Utilize closed circuit TV system Procedures for patient control Procedures to control treatment interruptions by coordinating with housekeeping and dietary personnel and limiting interruptions due to visitors, clergy, and medical staff Notify physics staff if machine error occurs (give telephone and beeper numbers) How to recognize a source How to interpret the area radiation monitor How to handle a dislodged source How to use a survey instrument Radiation warnings and information posted on the room door
130 Quality assurance in low dose-rate afterloading
readily available. In an emergency situation, the area should be posted with radiation warning signs to prevent accidental exposures. Procedures should be carefully designed to minimize personnel radiation exposure, tested regularly, and be clearly displayed at the control console and the room entrance. Certain practical details should also be reviewed during personnel training. These include scheduling procedures. If more than one type of remote after-loader is available, the scheduling should ensure that the appropriate machine is available for each patient. As usual, good communication between the radiation oncologist, the referring physicians, and the brachytherapy technologist is very important. Another practical point to review is that flexible interstitial catheters should never be shortened after measuring to determine the necessary active and inactive source lengths has been done. An undetected shortened catheter will result in either a source positioning error or a source transfer alarm condition. One final practical point is that applicator openings should always be capped to prevent fluid or dirt from getting inside. The applicators should only be uncapped when inserting or removing dummy sources and when connecting to the transfer tubes for treatment. Foreign material, if carried by the source or source assembly into the afterloader interior, can damage optical sensors and cause premature mechanical deterioration. It is the responsibility of the physicists and brachytherapy technologists to schedule and provide the ongoing training. The reviews must be given to all involved personnel at least once per year and promptly for new staff. A reference binder containing descriptions of established procedures and instructions for all involved personnel should be readily available at the afterloader treatment areas (Table 8.10).
8*5.6
Practical operational considerations
The introduction of new brachytherapy technology not only demands that new treatment procedures be established, but sometimes also requires modification of certain design features of a new device in order to achieve smooth operation in the field. Some of these issues can be resolved by further product development by the
Table 8.10 Contents of a procedures manual available at the nurses'station Procedures for machine programming and initiating treatment Procedures for uncoupling and terminating treatment Procedures for cleaning treatment tubing and connectors Emergency procedures Error codes, their interpretation, and corrective actions to be taken
manufacturer. A close and cooperative relationship between the user and the manufacturer can facilitate these improvements. This type of experience has been reported with regard to the microSelectron-LDR [2]. A brief review of some of those experiences is presented in the following paragraph so that additional insight into the mechanics of remote afterloading can be gained. If difficulties in source assembly transport into treatment are encountered, the first thing to check is the curvature of the transfer tubing and flexible applicators, especially at the applicator-transfer tube interface. Irregularly shaped source tips have also been determined to cause source transport problems, especially when the source encounters a curvature just before it traverses the transfer tube-applicator interface. This is why iridium ribbon preparation includes the use of the special heating element to achieve a smooth, bullet-shaped, plastic tipOther difficulties have been associated with flexible catheters used for perineal implant sites. Flexing or constricting of the catheters will result when the patient shifts position in the bed or sits up. This may cause source assemblies and drive cables to uncouple during source retraction, leaving the sources in the catheters. An alarm condition will occur. By inspecting the machine console display, the errant channels can be identified. The patient's position should be adjusted to relieve the pressure exerted on the catheters so that source recovery can be undertaken. The transfer tube should be uncoupled at its connection to the intermediate safe and a source recovery tool passed down the transfer tube to retrieve the source assembly and return it to the source storage container. This problem can be minimized if the patient's position is kept at a low angle in the bed. The patient can be instructed via the intercom to assume this position just before source retraction is activated. The use of rigid implant needles should also reduce the frequency of this problem. These corrective actions can also be used to facilitate source transport into treatment. Obviously, the audible alarm system is very important in afterloaders. Strengthening the connection of the drive cable to the source assembly has also been attempted to prevent uncoupling. This may be a reasonable solution if the same source assemblies are used routinely, but requires that those sources be stored indefinitely in the intermediate safe. However, this approach exposed a malfunction that otherwise might have never been detected. On occasion, sources in the intermediate safe that were not programmed for treatment would actually be transferred into treatment along with the programmed sources, even without an applicator or transfer tube in place. The discovery of that malfunction forced those treatment machines to be removed from clinical service until the problem was corrected. Another potential problem involves perineal applications. When the patient adjusts his or her position in the
References 131
bed, the applicator system can become dislodged. This is caused by the patient moving away from the transfer tubing without the tubing being free to follow, as in the situation in which the patient inadvertently places one foot over the tubing while adjusting position. It is essential that the applicator systems be adequately secured to the implant site, and the patients also need to be reminded that they are tethered to the treatment machine and care must be taken to avoid applying forces that can disengage the implant systems. The nurses and physicians should regularly check the source transport tubing to prevent tangles and to verify that the applicator system has not become dislodged. Although not all remote afterloaders are plagued with all the problems described above, a functional problem is always possible due to the number of treatment interruptions over the long duration of the treatments. Certain personnel, generally physicists, must be assigned to respond to a machine failure at any time so that these treatment-related problems can be resolved quickly and safely. The nursing staff cannot be expected to be responsible for this, although sometimes only a telephone call to the physicist on call is necessary to resolve a relatively simple problem. The adjustments that one must make for remote afterloading are related to the new source types, the changes in the control of source placement and treatment timing, the fact that the patient is tethered to a machine, and the uncertainty of not knowing exactly when the treatment will be completed. One must realize that reliance on a machine to control a treatment requires a comprehensive quality assurance program and ongoing training and, despite it all, there will inevitably still be machine failures at some time.
REFERENCES
intravaginal brachytherapy. Int.J. Radial Oncol. Biol. Phys., 26(3), 499-511. 6. Glasgow, G.P. (1989) An inventory of cesium-137 seeds for multiple site interstitial brachytherapy with a microSelectron. Endocuriether. JHypertherm. Oncol., 5, 175-9. 7. US Nuclear Regulatory Commission (1993) Title 10 Chapter 1, Code of Federal Regulations-Energy, Part 20: Standards for Protection Against Radiation. Washington, DC, Government Printing Office. 8. Glasgow, G.P. (1993) Remote Afterloading Technology, Report of the American Association of Physicists in Medicine Task Group No. 41. New York, American Institute of Physics. 9. Klein, E.E., Grigsby, P.W., Williamson, J.F.etal. (1993) Preinstallation empirical testing of room shielding for high dose rate remote afterloaders. IntJ. Radial Oncol. Biol. Phys., 27(4), 927-31. 10. National Council on Radiation Protection and Measurements (1976) Report No. 49: Structural Shielding Design and Evaluation from Medical X-rays and Gamma Rays of Energies up to 10 MV. Bethesda, Maryland, National Council on Radiation Protection and Measurements. 11. Pipman.Y., Jamshidi, A. and Sabbas, A. (1989) Commissioning of a Selectron LDR remote afterloader tests and measurements (Abstract). Med. Phys., 16(3), 494.
12. Jones, C.H. (1990) Quality assurance in brachytherapy. Med. Phys. World, 6,4-11. 13. Mesina, C.F., Ezzell, G.A., Campbell, J.M. etal. (1988) Acceptance testing for the Selectron high dose rate remote afterloading cobalt-60 unit. Endocuriether./ Hypertherm. Oncol., 4,253-6. 14. Slessinger, E.D. (1990) A quality assurance program for low dose rate remote afterloading devices. In Brachytherapy HDR and LDR, ed. A. Martinez, C. Orton and R. Mould. Columbia, Maryland, Nucletron Corporation, 160-8.
1. Nath, R., Anderson, LL, Meli, J.A., Olch, A.J., Stitt, JA and
15. Slessinger, E.D. (1990) Selectron-LDR quality assurance.
Williamson, J.F. (1997) Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Task Group No. 56. Med. Phys., 24,1557-98. 2. Grigsby, P.W., Slessinger, E.D., Teague, S.P. etal. (1995)
Selectron Brachy therapy J., 4(2), 36-40. 16. Idaho National Engineering Laboratory (1994) Quality Management in Remote Afterloading Brachytherapy. US Washington, DC, Nuclear Regulatory Commission.
Clinical evaluation of an interstitial remote afterloading
17. Williamson, J.F. (1991) Practical quality assurance in low-
device for multichannel intracavitary irradiation. Int. J.
dose rate brachytherapy. In Quality Assurance in
Radial Oncol. Biol. Phys., 31(4), 875-81.
Radiotherapy Physics: Proceedings of an American College of Medical Physics Symposium, ed. G. Starkschall and J.
3. Grigsby, P.W., Williamson, J.F. and Perez, C.A. (1992) Source configuration and dose rates for the Selectron afterloading equipment for gynaecologic applicators. Int. J. Radial Oncol. Biol. Phys., 24(2), 321-7. 4. Wilkinson, J.M., Moore, C.J., Motley, H.M. etal. (1983)The use of Selectron afterloading equipment to simulate and extend the Manchester System for intracavitary therapy of the cervix uteri. Br.J. Radio!., 56(666), 409-14. 5. Sloboda, R.S., Pearcey, R.G. and Gillan, S.J. (1993) Optimized low dose rate pellet configurations for
Morton. Madison, Wisconsin, Medical Physics Publishing Company, 139-82. 18. Meigooni.A.S., Williamson, J.F. and Slessinger, E.D. (1993) Practical quality assurance tests for positional and temporal accuracy of HDR remote afterloaders. Endocuriether./Hypertherm. Oncol., 9,46. 19. Thomadsen, B.R. (2000) Quality management for low and medium dose rate afterloaders. In Achieving Quality in Brachytherapy. Bristol, IOP Publishing.
132 Quality assurance in low dose-rate afterloading
20. International Electrotechnical Commission (1989) International Standard IEC 601-2-17 Medical Electrical Equipment Part2. Geneva, Bureau Central de la Commission Electrotechnique International. 21. Orton, C.G. and Siebert, J. (1972) Instrument non-linearities and therapy unit timer error. Phys. Med. Biol., 17,198-205. 22. Williamson, J.F. (1986) The accuracy of the line and point source approximations in 192lr dosimetry. Int.J. Radial Oncol. Biol. Phys., 12(3), 409-14. 23. Siwek, R.A., O'Brien, P.F. and Leung, P.M.K. (1991) Shielding effects of Selectron applicator and pellets on 24.
isodose distributions. Radiother. Oncol., 20(2), 132-8. Pla, C, Evans, M.D.C. and Podgorsak, E.B. (1987) Dose distributions around Selectron applicators. Int.J. Radial Oncol. Biol. Phys., 13(11), 1761-6.
25. Slessinger, E.D. (1995) Commissioning of non-stepping source remote afterloaders. In Brachytherapy Physics, ed.
J.F. Williamson, B.R. Thomadsen and R. Nath, Medical Physics Publishing, Madison, Wl, 503-22. 26. Slessinger, E.D. (1995) Clinical implementation of LDR remote afterloading. In Brachytherapy Physics, ed. J.F. Williamson, B.R. Thomadsen and R. Nath, 521^0. 27. Slessinger, E.D. and Grigsby, P.W. (1989) Verification studies of 3D brachytherapy reconstruction techniques, in Brachytherapy 2, ed R.F. Mould. Leersum, The Netherlands, Nudetron Trading BV, 130-5. 28. Meertens, H. and van der Laarse, R. (1985) Screens in ovoids of a Selectron cervix applicator. Radiother. Oncol., 3(1), 69-80. 29. Steggarda, M.J., Moonen, L.M., Damen, E.M. and Lebesque, J.V. (1997) An analysis of the effect of ovoid shields in a Selectron-LDR cervical applicator on dose distributions in rectum and bladder. Int.J. Radial Oncol. Biol. Phys., 39,237-45.
9 Quality assurance in high dose-rate afterloading COLIN H.JONES
9.1 INTRODUCTION A quality assurance program should give assurance to the user that predetermined specific objectives are being met. There are two fundamental requirements in brachytherapy and these are: first, to deliver a prescribed radiation dose within acceptable limits of accuracy; and, second, to ensure that, in so doing, patient, staff, and public are not irradiated unnecessarily. The uncertainty in dose specification can be subdivided into that due to clinical procedures, such as the uncertainty in the indication of the target volume on a localizing radiograph, and that due to physical procedures. The latter is principally the uncertainty in dose determination due to calibration of the source and dosimetric calculations. The Netherlands Commission on Radiation Dosimetry (1991) recommended that the uncertainty in the dose specification due to these physical procedures (defined as one relative standard deviation) should be less than 5%. This is consistent with the recommendations of the International Commission on Radiological Protection (ICRP) [1] and the World Health Organisation [2]. In the USA, the Nuclear Regulatory Commission (NRC) define a misadministration as a dose differential of 20% between the prescribed and administered doses; a dose delivered to the wrong treatment site is also considered to be a misadministration. In order to achieve a high level of accuracy and to minimize radiation exposure beyond the treatment volume, it follows that a quality assurance system must cover every aspect impinging upon the treatment of the patient and that each part of the treatment process should be evaluated critically. The
subject has been considered by several groups [3-5]. Chapter 8 describes the quality assurance measures required in low dose-rate (LDR) afterloading. Many of these measures apply equally well to high dose-rate (HDR) procedures, and the contents of both chapters may be considered to be complementary to each other. There are four important elements which are common to quality assurance programs: 1. The appointment of a person experienced in radiation physics to be responsible for drawing up the quality assurance program, for training staff, and for ensuring compliance with the program. 2. The program should be documented in detail, including the procedures which must be followed, the tests which must be carried out, and the frequency of these tests. The results of these tests in terms of compliance or non-compliance should be recorded. 3. Radioactive sources should only be used in compliance with local, or national and/or international recommendations. 4. Incidents which have, or might have, affected the precision of treatment or the safe use of sources should be noted: the program can then be modified in the light of experience. The overall program should be reviewed periodically. Quality assurance for remote afterloading devices with HDR sources are considered here under the following headings: 1. Regulatory requirements 2. Facility design
134 Quality assurance in high dose-rate afterloading
3. Machine function tests 4. Tests relating to treatment precisions 5. Quality assurance documentation.
9.2 REGULATORY REQUIREMENTS Remote-controlled, automatically driven, gamma ray afterloading equipment is made and finished with a degree of uniformity and manufacturers are required to comply with generally accepted principles of sound and safe practice. This applies to the radiation source and to all components of the equipment, including, for example, any programmable electronic system used as a controlling timer, or any interlocks that are used in the equipment. Consequently, the machine features designed by different manufacturers are often similar and quality assurance procedures required for different types of remote afterloading equipment are correspondingly of a common nature. The fabrication of the radioactive source and its containing capsule must conform with BS 5288 in the UK and the source housing within the equipment must conform to regulatory requirements. The International Standard IEC 601-2-17 [6] specifies the requirements with regard to the design, function, testing, and use of remote afterloading equipment. This standard specifies requirements for equipment which gives air kerma rates up to 500 mGy rr1 at 1 m from the radioactive source or sources in use. For equipment operating outside this range, special precautions may be necessary. In the UK, the use of such equipment is also covered by the Ionising Radiation Regulations [7]. In the USA, the NRC specifies a number of requirements with regard to the design, function, testing, and use of such equipment. The IEC Standard 601-2-17 [6] specifies requirements for equipment intended to be: used under the supervision of qualified persons, maintained at predetermined intervals, and subjected to regular checks by the user. The requirements of the standard are based on the assumptions that an irradiation treatment prescription is available which prescribes appropriate values of the treatment parameters and the air kerma rate at 1 m from the radioactive source(s) in the equipment is known. These requirements are intended to ensure that the prescribed values of the treatment parameters can be achieved by the equipment, in particular that the selected radioactive source (or sources) is positioned or moved within the source applicator in the selected configuration relative to the source applicator. The indication that the position of any source or source train within an applicator supplied with the equipment is to be as selected will not be given unless such positions are within ±2 mm of those programmed, in any direction. It is a requirement of the standard that the equipment is provided with a controlling timer for each channel or
each group of channels where channels are grouped to operate together. Furthermore, controlling timers must have a mean percentage average error not exceeding 1. In order to comply with these requirements, it is necessary to undertake appropriate tests and these must form part of a quality assurance program. International regulatory requirements are designed to ensure that, when equipment is used in compliance with these standard recommendations, radiation hazards are minimized and treatment delivery can be achieved safely. Clearly, good radiation protection practice consistent with regulatory requirements should be part of the quality program.
93
FACILITY DESIGN
High dose-rate brachytherapy covers a wide range of different kinds of treatment, including the use of surface moulds, intracavitary and interstitial techniques using HDR treatments and pulse dose-rate (PDR) treatments. The duration of a treatment might range from a few minutes, as in the case of conventional HDR treatments, to a few days for PDR treatments. Although it is possible to install an HDR machine in a modified radiotherapy treatment room, it is better to install such equipment in a room designed specifically for the treatment techniques envisaged. The details of the design brief for such a room will depend upon the type of brachytherapy being planned. There are, however, various considerations that are common to most requirements: • The treatment room should be constructed to allow safe implementation of the prescribed treatment. • The room should have sufficient shielding or be so isolated that healthcare personnel, patients, and members of the public will not receive incidental radiation exposure in excess of recognized specified limits. The radiation levels in and around the room should be measured or calculated and staff working in the area should be monitored. Areas where the dose rate is more than 7.5 uvSv h'1 should be designated as controlled areas: a supervised area where the dose rate is less than 7.5 (iSv h-1 but more than 2.5 |lSv h-1 might in some circumstances also be identified. Radiation warning notices should be posted identifying the controlled area. • The layout of the room should allow patients to be nursed safely, efficiently, and with as little inconvenience and discomfort to the patient and staff as possible and also allow all the procedures associated with the medical care of the patient to be accomplished. • The room should be equipped so that any necessary medical emergency procedures can be implemented speedily and effectively.
Facility design 135
• Particular attention should be paid to the need for coping with radiation emergencies that might occur during the course of treatment. It is important that the radiation protection adviser and/or experienced radiation physicist should be involved in the room design concept at an early stage in the planning process. HDR sources are principally either cobalt-60 or iridium-192. In both cases, the high dose rates associated with the high-activity sources necessitate a room design which has either some form of maze entrance or a very heavy lead door. HDR treatment rooms require walls of thickness 40-60 cm of concrete, depending upon the size of the room and the energy and activity of the sources that are being used. Table 9.1 lists the physical characteristics of cobalt-60 and iridium192. In practice, a room with a maze-type entrance requires more floor space, but allows economies to be made in terms of cost and convenience. If adequate space is available, it is also advantageous to include some form of X-ray facility within the room so that applicator positions can be checked prior to treatment without the need for moving the patient. Protective lead doors and also any barrier to the maze entrance must be interlocked with the treatment unit. A comprehensive radiation protection survey should be made prior to using the facility and this should be kept for reference purposes: if alterations are made subsequently to the room structure or layout, these should be recorded and the survey repeated. The air kerma rate around the source storage container should also be surveyed. At any position 50 mm from the surface of the storage container or any other surface permanently fixed to it, the air kerma rate mSv should not exceed 0.01 |0,Svh-1.The air kerma rate at any position 1 m from the surface of the storage container should not exceed 1 (iiSv h-'. It is a requirement of the IEC standard that these measurements should be made with the combination of radioactive sources that is possible within the specifications given by the manufacturers and is the least favorable with regard to the magnitude of the dose rates. For measurements at 50 mm, the air kerma rate should be averaged over an area up to but not exceeding 10 cm2; for measurements at 1 m, the air kerma rate should be averaged over an area up to but not exceeding 100 cm2. The room containing the source safe should be lockable so that access is restricted when the machine is not in use.
The room should be equipped with the following: • Closed circuit television (CCTV) cameras with monitors at the control console: it is preferable that one of these systems should be in color so that the patient's clinical appearance can be viewed satisfactorily. • Two-way patient intercommunication. • Radiation warning lights: these should be fitted inside the treatment room, at the door entrance, and at the control console, and should be activated when the radiation source is exposed. Controlled area signs should be posted to indicate areas where the dose rate is greater than 7.5 (J,Sv h-1; these signs should preferably be linked to the machine so that they will be illuminated when the radiation sources are exposed. • It is advisable to use an audible time delay interlock so that the machine can only be switched on within a predetermined time after the time delay has been activated: the time setting should be adjusted so that this is optimal, giving just sufficient time for personnel to leave the treatment room and initiate treatment. • There should be an independent audible radiation alarm inside the treatment room. • Near the treatment console, there should be a handheld dose-rate meter and a personal integrating dose meter for use in emergencies. • Within the treatment room, there should be a shielded safe, which should be available for storing the radioactive source(s) in the event of an emergency or when the machine is being serviced. • There should be a clearly labelled 'emergency stop' adjacent to the console. • If a diagnostic X-ray unit is available within the treatment room for localizing radiographs, warning lights should indicate when this machine is being used and there should be adequate radiation protection for the operator. Many of these items will require checking at regular intervals as part of a quality assurance program. HDR equipment should not be used as mobile equipment. In the USA, the NRC requires licensees to comply with the above requirements, and relocation of the remote afterloading device to another area is prohibited without
Table 9.1 Properties of principal brachytherapy nudities
"Co 92 lr
5.27 years 74 days
1.17,1.33 0.3-0.6
309 113
HVT = half value thickness; TVT = tenth value thickness.
12 4.5
40 15
206 113
136 Quality assurance in high dose-rate afterloading
prior NRC approval. The dedicated treatment room must be equipped with continuous viewing and intercom systems to allow for patient observation during treatment and, if the systems do not have a back-up system to be used if the primary system fails, the licensee must commit to suspending treatments until the primary system is repaired. In common with the IEC requirements, the NRC specifies that areas or rooms used to store remote afterloading devices or source containers housing a source must be secured; dedicated treatment rooms must have an electrical interlock system and restricted area controls. If there are other radiation-producing devices located in the treatment room, the licensee must institute mechanisms to ensure that only one device can be placed in operation at any one time.
•
•
9.4 MACHINE FUNCTION TESTS The function of the remote afterloading device should be evaluated before the radioactive source is installed. When possible, a number of tests should be performed with a dummy source substituted for the actual source to test the source drive mechanism and safety interlocks. These tests should include the following: • • • • •
All door and safety interlocks. Emergency return mechanisms. Controlling timers. The security of couplings and connections. Catheters and any other accessories that are going to be used during treatment should be checked to ensure that all have closed ends so that the radioactive source cannot be lost within a patient should source encapsulation fail. • The limitations of channels and source applicators must be tested and documented: in practice, acute directional changes of source transfer tubes should be avoided to prevent 'sticking' of the source in transit, and it is useful to carry out such tests using a dummy source. • The operation of the door warning lights, CCTV systems, intercoms, and the audible time delay interlock should be checked for function and also adjusted to work optimally. • The console display should be checked: test all button functions by programming and carrying out a simulated treatment; verify that all displays are correct; verify that the data on any printout agree with the programmed data; and, when appropriate, check program card data input function. Once the radioactive source has been installed into the treatment device, the above tests should be repeated in addition to the following tests: • With the source in its safe position, a survey should be made with a portable radiation dose-rate meter around the remote afterloading device to ensure that
•
•
•
the shielding of the source safe complies with regulatory requirements. With the source exposed, exposure rates in accessible areas outside the treatment room should be measured. The maneuverability of the unit should be taken into account: in small treatment rooms it might be possible for the machine to be near to the treatment room door, which could result in high dose rates outside the room. On the basis of the predicted workload, it will be necessary to calculate the monthly radiation exposures to staff operating the machine to ensure that the dose levels are acceptable. The user is required to estimate the dose delivered to the patient during transit of the radioactive sources into the treatment position. Transit doses are partly determined by the time taken by the transit movements, but they are also determined by the strength of the radioactive sources selected by the user and by the number of interruptions during treatment. To calculate the total transit dose, it is necessary that the user has information from which the transit doses may be estimated and takes appropriate account of them. The IEC standard relates to measurements or calculations for two conditions for each channel: the air kerma at a position 20 mm from the axial center of the source applicator, and the air kerma at a position 1 m from the axial center of the channel. For each condition, the air kerma should be measured or calculated for one specified radioactive source and should be at the positions that are least favorable with regard to this requirement. For measurements at 20 mm the air kerma should be averaged over an area of up to but not exceeding 2 cm2, and for measurements at 1 m the air kerma should be averaged over an area up to but not exceeding 100 cm2. In the event of an emergency, it might be necessary for a user to enter the treatment room with the source in the exposed position. It is useful to have knowledge of the radiation distribution inside the treatment room so that the line of entry can be optimized to reduce personnel exposure. The radiation monitor used to check dose rates outside the treatment room and also inside the room and around the machine should be tested over the entire range of radiation levels that might be present. Some monitors will saturate and fail if directly exposed to high dose rates, so it is appropriate to use a survey instrument with a wide range from 1 mSv tr1 to at least 1000 mSvh-1. Although remote afterloading devices make use of radioactive sources in a sealed system, users are obligated to check for leakage and/or contamination of the source. It is not possible to check for contamination directly and it is therefore appropriate periodically to check the catheters through which the radioactive sources pass. Sometimes, in the
Tests relating to treatment precision 137
preparation of encapsulating the source, radioactive contamination adheres to the source capsule and it is appropriate to check for this free radioactivity by passing the source through a catheter several times and then checking for any radioactivity with a sensitive radiation detector capable of detecting less than 200 Bq. • The radioactive source should be calibrated and the strength of the source (or sources) should (when appropriate) be entered into the treatment machine. The output from the printer should be checked to ensure that the correct source activity is displayed and, if the machine corrects automatically for source decay, it will be necessary to check that this calculation is within acceptable limits.
9,5 TESTS RELATING TO TREATMENT PRECISION Accurate delivery of doses using HDR systems depends on knowing the strength of the radioactive sources at the time of treatment (see Chapter 3), the precision and consistency of the timer, and the ability of the treatment machine to position the source at the proper location along the catheter or treatment applicator. Reference has already been made to the requirement that the equipment should provide a controlling timer for each channel which should have a mean percentage average error not exceeding 1. Timers should be checked by selecting five pre-set times (not smaller than 1% of the maximum pre-setable time) to cover the range possible with the treatment device - taking the mean percent average error at the five pre-set times. In practice, some equipment manufacturers arrange for these tests to be carried out as part of a routine maintenance contract. Even so, it is important for the user to recognize that such tests have to be made. Timer errors can occur [8]. Likewise, machines that use a single source which can be programmed into a number of dwell positions should be checked to ensure that the programmed dwell times are within the required precision. Before describing specific quality assurance tests, it is appropriate to mention that there are three principal categories of HDR equipment. These are:
1. single HDR iridium-192 source machines 2. PDR single HDR iridium-192 source machines 3. machines that have multiple HDR cobalt-60 sources. Details of these machines are shown in Table 9.2. Whereas the majority of quality assurance checks are common to all systems, some of the design features are different and allowance has to be made for such variations in the quality assurance program that is adopted. These may be summarized as follows: • In the case of single HDR iridium-192 source machines, the high-activity source is physically very small and attached to a cable which is driven by a stepping motor so that the required radiation distribution can be achieved. The half-life of iridium is relatively short, so the source has to be changed at approximately 3-monthly intervals. The physical characteristics of the radioactive source are such that very fine catheters can be used, with the result that very high local doses are given to tissues surrounding the catheters. Precise control of the stepping motion is necessary if reliable radiation distributions are to be achieved. The length of the cable to which the source is attached must be constant and when sources are replaced it is necessary to check that this length has not altered. Providing the stepping function of the machine works satisfactorily, then reliably high precision radiation treatments can be achieved. The versatility of the machine allows a wide range of different types of treatments to be achieved, including those using applicators, needles, and catheters. The position of the source within these treatment devices must be checked before clinical use. When the source is changed, different tests must be carried out to ensure that the catheter length has not been altered in any way. Several different commercial HDR iridium-192 source machines are commercially available and such equipment has largely replaced multiple-source HDR devices. • PDR single HDR iridium-192 source machines also use a stepping source, but are designed specifically to use a lower activity source and to deliver the radiation dose over a much longer period of time. For example, a treatment might consist of 30-40 fractions over a
Table 9.2 Details of sources used in remote afterloading equipment
HDR PDR HDR
192
400
192
20-40 4-20
lr
lr
60Co
1 1 20 variable
HDR = high dose rate; PDR = pulsed dose rate.
Cylinder Cylinder Pellet
1.1 1.1 2.5
Cable Cable Pneumatic; cable
138 Quality assurance in high dose-rate afterloading
period of 3-5 days or maybe the dose will be delivered with a series of fractions given each hour throughout the day. The technique that is used might vary from one center to another and, in some situations, it is possible to disconnect the treatment machine from the patient to allow the patient greater comfort during the period when radiation is not being given. The use of multiple fractions over a longer period of time can, in some circumstances, be logistically difficult and it is important to ensure that appropriate quality assurance measures are adopted to prevent applicators within the patient being displaced. This is especially so when patients are disconnected from the machine and then recoupled to the machine many times over the treatment period. One very important consideration of PDR brachytherapy is the safety aspect. Based purely on health and safety considerations directly related to the instantaneous exposure rate from PDR machines, it is necessary to ensure the continuous availability of a trained person during the delivery of the radiation dose. Should the device fail with or without alarm generation, due either to mechanical failure or facility power failure, the dose delivered to the patient could be considerable unless immediate removal of the source by manual means by an experienced member of staff can be achieved. Other quality assurance measures are similar to those for a conventional HDR treatment machine. • HDR machines using multiple sources of cobalt-60 are similar in design to LDR or medium dose-rate multiple source cesium-137 machines and the quality assurance measures are similar to those described in Chapter 8. The source catheter is made up of active sources and inactive spacers and the size of each of these is greater than that of the cable-driven source. The sources and spacers are spherical and driven pneumatically from a treatment safe into a specially designed treatment catheter. Stacking of sources occurs in a similar way to that shown in the radiograph of the dummy source train in Figure 9.1. The fact that there are multiple sources means that there is a range of source activity and there is no control over which source is used for a particular treatment. The consequence is that the precision with which radiation doses can be delivered using such a system is less than that for a single high-activity stepping source.
Figure 9.1 Part of a radiographic marker (magnified) for HDR (MDR) Nudetron Selectron showing 'stacking' of dummy source pellets in photograph (a) in comparison with photograph (b).
each day of use and that the precision of reproducibility should be within 1 mm. The techniques can be used separately or conjointly to provide information about the distribution of radioactive material within the source container and positional information about individual sealed sources in treatment trains. The autoradiography technique is simplified by having a PMMA or wax support which has a recess with the same dimensions as the source (Figure 9.2). Autoradiographs are useful for checking the uniformity of radioactivity and are also a useful means of checking the relative strength of individual sources in a source train. Visual inspection of the autoradiograph might indicate lack of uniformity at the 10% level, but for more reliable assessment, densitometric scanning is to be preferred. In the case of positional studies, it is useful to incorporate lead markers into the wax support: secondary
9.5*1 Positional reproducibility Sources used inside applicators, catheters, and needles should be autoradiographed and radiographed. This is to establish the precise location of the source with respect to the end of the applicator and to check the machine functions. The NRC requires that the reproducibility of the source positioning should be checked
Figure 9.2 Wax disc for applicator and source autoradiographs.
Tests relating to treatment precision 139
electron emission from the lead results in an autoradiograph of the markers and provides the required positional data. This is particularly useful for checking the position of sources loaded into applicators or catheters, for which it is sometimes difficult to identify the precise end or tip. The method allows precise comparison to be made with different applicators and provides a record of the location of the radioactive sources inside loaded applicators (Figure 9.3). Envelope-wrapped Kodak X-Omat Verification film is suitable for these studies. Uniform pressure should be maintained over the film and source to keep both in close contact. For high-activity sources such as those used in HDR equipment, film exposure is only a fraction of a second and is suboptimal because the transit time is of comparable magnitude. Gafchromic film is a useful alternative. This is a thin radiation film which is colorless, grainless, and offers high spatial resolution (1200 line pairs mm'1). When exposed to high doses of radiation, typically 200 Gy, a dye in the film turns blue - the density of which depends upon the absorbed dose. This means that exposures of about 20 s are required for autoradiography, which results in better control of the image quality. The film is not light sensitive and produces high-quality images. Figure 9.4 illustrates the response of Gafchromic film to iridium-192 radiation, and Figure 9.5 illustrates a multiple exposure of an HDR iridium-192 source in a test jig designed for the measurement of source position [5]. Detex paper is a cheaper alternative to Gafchromic. This paper is used in the printing industry and changes color when exposed to high radiation doses. Before exposure, the paper is yellow, but under irradiation, hydrochloric acid is released from the ink and the yellow azo dye turns red. The shade of red does not indicate the dose, but rather that a certain level of dose has been achieved. Detex labels, for example, are used to indicate whether a product has received a sterilizing dose of radiation. Detex paper can be used in situations in which the dose falls within 1-100 kGy.
Figure 9.3 HDR microSelectron source autoradiograph showing position of source in relation to top of applicator.
Figure 9.4 Density-dose response curve for Gafchromic exposed to iridium-192 source, using light of wavelength 600 nm.
Figure 9.5 (a) Photograph of test jig. (b) Autoradiographs of HDR iridium-192 source in test jig-
140 Quality assurance in high dose-rate afterloading
In summary, autoradiography forms an important part of commissioning and quality assurance and may be used to: • examine the distribution of radioactivity in the source capsule or the distribution of activity amongst sources • record the position of the source with respect to the end of the transfer cable • record the relative position of each individual source in a source train • check the reproducibility of source positions when inserted into catheters, needles, or applicators. It is important that every such device which is used clinically should first be tested and the autoradiographic test data should be used as a baseline for subsequent quality assurance investigations. There is also a place for X-ray radiography in quality assurance. For example, it is necessary to check the spatial distribution of sources in the patient for dosimetry purposes and this might require the use of radiographic markers. These markers are inserted into empty applicators prior to loading the radioactive sources into the patient so that the geometrical position of the applicators can be located accurately. In the course of commissioning new equipment, radiographic checks should be made to ensure that these markers are reproducible and that their position within a set of applicators is clearly defined in relation to the radioactive sources. The reproducibility of a set of markers should be within ±1 mm (Figure 9.6). The precision of the method used to localize the inserted sources should be checked for each type of technique to be used. The accuracy of the reconstruction
Figure 9.6 Radiographs of markers in HDR (MDR) Nudetron Selectron applicators showing relative positions of sources.
program used to determine the geometrical configuration of the sources should also be checked. In estimating the overall accuracy of a particular technique, some estimations should be made of the likelihood of source displacement which might occur during the course of treatment. In the case of HDR machines with a single source, the location of the source can be determined by alternative methods. The simplest of methods is to use a video camera system, a transparent plastic applicator, and a linear scale. On a daily basis, a check ruler (such as that available for the HDR microSelectron) may be used to check the accuracy of source positioning. The ruler is attached to the treatment unit through a treatment transfer tube. It consists of a scale and a marker rod, which is moved by the source cable as it moves along the ruler. One can evaluate the motion of the source by checking the position of the rod against the programmed position of the source. However, the technique only indicates the motion of the end of the marker rod. Speiser and Hicks [9] have modified a check ruler to incorporate a diode radiation detector which indicates source position (Figure 9.7). The diode is a 1 mm2 photodiode placed within a few millimeters of the source path. Its position corresponds to a fixed distance from the treatment head. The diode is connected via coaxial cables to an electrometer. As the source gets near the diode, the reading on the diode increases. The electrometer readings can be correlated to source position with an accuracy of 0.1 mm. The authors found this accuracy sufficient to detect differences in length of transfer tubes, as well as the variation of path lengths with relaxation or change of curvature of the transfer tube. As part of the basic equipment package, some manufacturers include a 'QA Physics Package.' This comprises a well chamber and an electrometer for calibrating the radioactive source, a camera and scale assembly, which enables the source wire to be imaged against a scale at specified distances of source travel. One such system is the VariSource (manufactured by Varian Oncology Systems) remote afterloading machine. A frame-grabber captures the images of the calibration device and displays the position of the source wire on the VariSource computer monitor (Figure 9.8). The system is convenient to use and takes only 3 min to check and, if necessary, recalibrate the drive positioning system and to obtain an automatic printout for quality assurance records; consequently, this can be done prior to each use of the machine. A well-type ionization chamber can also be used for checking the source position at the end of a transfer cable. By means of a specially designed quality assurance insert which fits into the well of the ionization chamber, it is possible to test not only source positioning but also the timer and its consistency. Quality assurance inserts for use with the Standard Imaging well chamber have
Tests relating to treatment precision 141
Figure 9.7 Top and side views of modified check ruler as described by Speiser and Hicks [9].
been described by Jones [10] and De Werd et al. [11]. One of these inserts for use with an HDR microSelectron is shown in Figure 9.9. The insert consists of a lead cylinder 30 mm in diameter, 115 mm long, housed in a 2-mm thick brass shield with a 100-mm deep central cavity of 2-mm diameter, which takes a bronchus source catheter. The insert has a brass flange through which it is fixed by a screw to the top face of the ion chamber. Four 2-mm diameter holes pass radially through the 15-mm thick lead cylinder: the apertures are in a plane at 72 mm from the base of the insert and intersect centrally at right angles to each other. The collimator insert modifies the response of the chamber as the radiation source is moved along its longitudinal axis. With the collimator in place, the response of the chamber increases as the
Figure 9.8 Frame-grabber image of VariSource source wire against a scale at 130 cm travel. (Courtesy of Varian Oncology Systems.)
source comes in line with the four radial apertures. The variation in response over the first 20 mm from the bottom of the catheter cavity of the insert (positions 1 to 6) is within ±0.4% and one of these positions may be used for source calibration measurements. When the source is at position 1, the source center is 6.5 mm from the tip of the catheter: each dwell position corresponds to an increment of 2.5 mm. The position of the source can be programmed to be in the plane of the apertures and the resultant data may then be used as reference information against which subsequent measurements of source position can be checked. This is achieved by a combination of altering the length of the source cable and/or entering the dwell position of the source. The cable length can be altered in 1 mm increments and the source positions can be changed in steps of either 2.5 mm or 5 mm: by combining both of these, it is possible to alter the source position in increments of 0.5 mm. This procedure may be used to move the source step by step across the plane of the aperture. The resultant profile enables the position of the aperture to be determined in relation to the programmed position of the source. The method is useful for measuring any changes that might occur when source cables are replaced. Typical profiles are shown in Figure 9.10: in practice, it is not necessary to record a complete profile because the location of the maximum chamber response can be determined satisfactorily by means of four or five measurements made at 0.5 mm intervals around the region of the aperture. The method is very convenient and more precise than autoradiography unless densitometric measurements are used: the method is able to determine a source positional change of less than 0.5 mm. Comparative measurements using autoradiography and the well-type chamber insert have shown good correlation.
142 Quality assurance in high dose-rate afterloading ;
igure 9.9 (a) The Standard waging HDR 1000 re-entrant ?A7 chamber with collimator nsert. (b) The collimator insert hawing one of the four 'pertures.
Figure 9.9 shows a PMMA extension scale fixed to the collimator insert. This is designed to facilitate withdrawal of the source catheter by a measured distance so that alternative source positions can be programmed and checked. For example, if the aperture position is found to be at a programmed distance of 54 mm corresponding to a source position of 18 and a cable length of 919 mm, then by withdrawing the catheter 15 mm the new aperture position should be at a source position of 12 for the same cable length. De Werd et al. [11] have described an alternative quality assurance insert which is made of lead and has two acrylic spacers, one of which is 4 mm thick and the other 1 mm thick. The device has been used for checking positional accuracy and also for checking dwell times and source activity. As the source is moved across the 1 mm spacer, the response of the ionization chamber results in a profile similar to that shown in Figure 9.10, with a full width at half maximum of 13 mm ±0.5 mm, with the peak also falling within ±0.5 mm of the same location. The authors conclude that the device is very easy to use and gives more accurate results than conventional radiographic film, and the measurement of dwell times is more precise than measurements made using a stopwatch. The positional accuracy was found to be ±1 mm
and the relative dwell time accuracy was ±5% for a 10 s dwell time.
9.5.2 Positional reproducibility in interstitial brachytherapy Interstitial and intraluminal brachytherapy with HDR iridium-192 and PDR iridium-192 afterloading machines raise special problems in quality assurance. Each catheter or needle implanted in the patient is attached to the treatment machines via a transfer tube. The precise position of the radioactive source is determined by the programmed length and the dwell position selected for a particular treatment. The source position within the needle or catheter is dependent upon the precise length of the needle and also upon the length of the transfer tube. In practice, the lengths of individual needles vary, as also do the lengths of individual transfer tubes. Tube-to-tube variation in length should be no more than ±0.5 mm and similar variations can occur in the length of individual needles. Various methods of checking the length of the needle and tube have been described. Williamson [12] has dealt comprehensively with some of these issues and has described three
Quality assurance documentation 143
Figure 9.10
Typical
profiles obtained by moving the HDR microSelectmn source along the chamber axis in front of the insert apertures. Top profile: 345 GBq source; lower profile: 145 GBq source.
methods for correlating treatment lengths and dwell position number. These are: 1. The use of a source-like cable with tungsten seed markers spaced at 1 mm intervals which is inserted into the treatment tube applicator and then radiographed. The marker centers function as individual catheter rulers with 1 cm gradations, from which the programmed treatment length can be read. This method has the advantage of convenience of use, but multiple markers are required so that individual markers can be inserted into each of the implanted needles (or catheters). The basic check consists of comparing the marker positions of a fully inserted simulation marker with the position of the actual radioactive source when programmed to dwell at positions corresponding to the marker seeds. In practice, one marker is used as a calibration standard against which other markers are measured: this may be achieved by inserting markers sequentially into a transparent catheter tube and marking the seed positions on graph paper. The calibration of the marker consists of superimposing a transmission radiograph of the marker upon the autoradiograph of a source itself. 2. The applicator orifice method: this method is based on the use of the insertion of a dummy marker relative to the applicator orifice which is constrained by a small cap on the end of its proximal end. The simulation marker acts as a 'sound' which is used to measure the inner length of the needle or catheter, from which the dwell position of the source can be calculated. The method assumes that all transfer tubes have the same length and that their average length is consistent with the cap to distal-node seed position of the marker set. By independently
measuring the distance from the applicator orifice to the 995 mm seed center of a calibrated radiographic marker (using graph paper and a transparent applicator), the actual transfer tube length may be calculated. Tube-to-tube variations in length should be no larger than ±0.5 mm. 3. The applicator tip localization method: this method uses the closed end of each applicator to localize the dummy marker relative to the index frame of reference in combination with the radiographic marker sounding of each transfer tube applicator combination. Simulation markers are fully inserted into each implanted catheter and radiographs are obtained: the appropriate transfer tubes are attached to the applicators and each transfer-tube-applicator assembly is 'sounded' using a calibrated marker to obtain the offset variation for each assembly. In summary, the objectives of any quality assurance program are to ensure that the length of individual needles or catheters that are used and also the length of individual transfer tubes are known. Williamson [12] found, over a 12-month period, the PDR flexible-catheter treatment tube length to be constant within ±0.5 mm on average. Over a 3-year period, semi-flexible HDR transfer tubes maintained their length within 1 mm, with tube-to-tube length variations within the 0.5 mm range.
9.6 QUALITY ASSURANCE DOCUMENTATION In practice, each user will develop his or her own quality assurance program, but certain minimum requirements are common to all.
144 Quality assurance in high dose-rate afterloading
The calibration procedure should include details of the method used to determine the air kerma strength of the source. It is required that calibrations be performed following installation of a new source before patient treatment is resumed and they are recommended at monthly intervals thereafter. The following list of quality control checks based on NRC recommendations might be considered to be a minimum requirement: • The afterloading device should be tested to determine the accuracy of source positioning. Source positioning within the catheter guide tube should be accurate to within ±1 mm of the programmed position. A record of the test should be maintained and should include the date of the test, the programmed position, the actual position of the source following activation of the device, and the initials of the individual who performed the test. Ideally, the record should include the radiograph or autoradiograph used to determine the source position. • Timer accuracy and linearity. • For devices that use a cable and/or wire to transport the source, measurement of the source guide tube to confirm the length to 1 mm of accuracy. • The back-up battery for the remote afterloading device should be tested, in accordance with the manufacturer's instruction, to verify emergency source retraction capability upon power failure. The minimum requirement for this test should consist of a function test with the mains power disconnected. • A record of these tests should be maintained for a period of at least 3 years and should include the date of the check, the results of the check, and the initials of the individual who performed the check. In practice, it is useful to document quality assurance information in the form of a monthly quality assurance chart (Figure 9.11). The NRC also recommends that at the beginning of each day of use the following checks should be performed in accordance with the manufacturers' instructions: • The permanent radiation monitor fitted within the treatment room should be checked for proper operation. • The television monitor and intercom should be checked to verify proper operation. • The treatment console operational functions should be checked, testing all indicator lamps, other status and operational displays and, if appropriate, check the printer and data which it displays. • Source status indicators ('safe' or 'unsafe'), including those which are integral to the afterloading device as well as any additional indicators installed at the treatment console or room entrance, should be checked.
• Electrical interlocks installed at the room entrance should be tested for proper operation. Records of these tests should be maintained for a period of 3 years. • The mechanical integrity of all applicators, source guide tubes, and connectors to be used should be determined by visual inspection and/or radiographs. The presence and correct placement of any internal shields and other essential internal components should be determined. • The records of all the checks specified above should be maintained for a period of 3 years and should include the date of the check, the results of the check, and the initials of the individual who has performed the check. One important aspect of documentation relating to the use of HDR equipment is that related to the emergency procedures that need to be implemented should the source fail to return to its safe position. At a minimum, these procedures should address the following: • The procedures should specify the circumstances in which they are to be implemented, such as any circumstances under which the source cannot be retracted to a fully shielded position in the afterloading device. • The action specified for emergency source removal should give primary consideration to minimizing radiation exposure to the patient and healthcare personnel while maximizing safety to the patient. • The procedures should specify step-by-step actions for equipment failure and specify the individual(s) responsible for implementing the actions. The procedure should clearly specify which steps are to be taken in different scenarios (for example, source decoupling versus a jammed source). The procedure should specify situations in which surgical intervention maybe necessary and the steps which must be taken in the event that surgical intervention is required. • In the event of an emergency, the procedures should specify the names of authorized personnel who should be informed, including the Radiation Safety Officer and/or the Radiation Protection Adviser. • There should be requirements to restrict and 'post' the treatment area with appropriate signs to minimize the risk of inadvertent exposure of personnel not directly involved in the emergency source recovery. • It is a requirement that the location of emergency source recovery equipment should be specified and the equipment that might be necessary for various equipment failures should be readily available, and their use described in the emergency procedure. At a minimum, emergency equipment should include shielded storage containers, remote handling tools, and, if appropriate, supplies necessary to remove applicators or sources from the patient, including scissors and cable cutters.
References 145
Figure 9.11 A monthly quality assurance HDR iridium-192 chart, Royal Marsden NHS Trust, London, UK. The Gafchromic autoradiograph is photocopied onto the chart as a record of positional accuracy of the source.
REFERENCES
Jones, C.H. (1991) Quality assurance using the SelectronLDR/MDRand microSelectron-HDR./4rf/V/Yy, 5(4), 12-16. Veenendaal, The Netherlands, Nucletron BV.
1. ICRP Publication 44 (1985). Protection of the patient in radiation therapy. Ann. ICRP, 15,2. 2. Quality Assurance in Radiotherapy (1988) Institute of Radiation Hygiene and World Health Organisation. Geneva, WHO. 3. AAPM Report No. 13 (1984) Physical Aspects of Quality Assurance in Radiation Therapy. American Institute of Physics for the AAPM. Chapter 6. 4. Ezzel,G.A. (1991) Acceptance testing and quality
6. I EC 601 -2-17,1989 (1990) Specifications for Remotecontrolled Automatically Driven Gamma-ray Afterloading Equipment. London, British Standards Institute. 7. The Ionising Radiation Regulations (1999) London, HMSO. 8. Chenery, S.G.A., Pla, M. and Podgorsak, E.B. (1985) Physical characteristics of the Selectron high dose rate intracavitary afterloader. Br.J. Radiol., 58,735-740. 9. Speiser, B.L. and Hicks, J.A. (1994) Safety programmes for
assurance for high dose rate afterloading systems. Activity, 5(4), 2-6. Veenendaal, The Netherlands,
remote afterloading brachytherapy: high dose rate and pulsed low dose rate. In Brachytherapy: from Radium to
Nucletron BV.
Optimisation, ed. R.F. Mould, J.J. Batterman,
146 Quality assurance in high dose-rate afterloading A.A. Martinez and B.L. Speiser. Veenendaal.The
B.R. (1995) Quality assurance tool for high dose rate
Netherlands, Nucletron International B.V., 270-84.
brachytherapy. Med. Phys., 22(4), 435-40.
10. Jones, C.H. (1995) HDR microSelectron quality-assurance studies using a well-type ion chamber. Phys. Med. Biol., 40,95-101. 11. De Werd, LA, Jursimic, P., Kitchen, R. and Thomadson,
12. Williamson, J.F. (1995) Simulation and source localisation procedures for pulsed and high dose rate brachytherapy. Activity Report, 7,57-65. Veenendaal, The Netherlands, Nucletron-Oldelft.
10 Radiation protection in brachytherapy A.M.BIDMEAD
10.1
INTRODUCTION
A fundamental requirement of radiation protection when brachytherapy is given is that the patient receiving treatment, the hospital staff, and the general public are not irradiated unnecessarily as a result of the brachytherapy. Protective measures should therefore be used to keep the dose levels as low as reasonably achievable (ALARA). Dose limits and recommendations are detailed in the Recommendations of the International Commission on Radiological Protection (ICRP), 1977 and 1978, and the International Commission on Radiation Units and Measurements (ICRU) and Institute of Physical Scientists in Medicine (IPSM) [1-5]. An extract from these recommendations, indicating the current and the previous dose limits (Table 10.1) is shown in reference 6. Each hospital in which brachytherapy is carried out is required to have a comprehensive radiation protection Table 10.1
policy, based on the appropriate national or state regulations. In the UK, these are the Ionising Radiations Regulations (IRR) 1999 [7]. These regulations were issued early in 2000 and some aspects of their detailed implementation still need clarification, particularly those aspects relating to the protection of the patient. The previous regulations were IRR 1985 [8], together with the POPUMET Regulations [9], Guidance Notes [10] and approved codes of practice [11,12]. IRR 1985 dealt mainly with the safety of employees and the public at large, whereas the new regulations include issues relating to the protection of patients, previously included in POPUMET. The hospital policy should describe the responsibilities of the hospital, departmental heads, radiation protection adviser and supervisors, occupational employment medical advisers, and other staff involved with the use of ionizing radiation. The regulations require all staff concerned with the use of radiation to have adequate training. Radiotherapists
Dose limits
Effective dose
20 mSv per year averaged
1 mSv in a yearb
50 mSv
5 mSv
15mSv
150mSv
15 mSv
over defined periods of 5 years3 Annual equivalent
150mSv
dose in the lens of
the eye a b
With the further provision that the effective dose should not exceed 50 mSv in any single year. In special circumstances, a higher value of effective dose could be allowed in a single year, provided that the average over 5 years does not exceed 1 mSv per year.
148 Radiation protection in brachytherapy
must have an Administration of Radioactive Substances Advisory Committee (ARSAC) license to practice brachytherapy, as described in the MARS regulations [ 13]. Previously, outside employees - that is, employees of another organization or company temporarily working in the hospital (e.g., service engineers etc.) -were covered by the Ionising Radiation (Outside Workers) Regulations 1993 [ 14], but this aspect is now included in the new regulations. The issues relating to pregnant staff are also covered in the new regulations. One new aspect in IRR 1999 is the specific requirement for formal risk assessments to be performed and documented. Although this was probably done previously in a less formal manner, it is now specifically required. There are two main protection issues to address: • The design of protected rooms, for the protection of the patient and staff during treatment and for the protection of staff preparing and handling sources. • The tracking and recording of the whole treatment process, starting with the progress of the source from its protected environment, through to the treatment delivery, continuing until the source is returned for storage or disposal. This ensures a source can be traced at every stage [15,16]. Radiation protection surveys of the designated rooms and preparation areas, and monitoring of staff radiation doses provide data on the effectiveness of the radiation protection. Protection of the patient includes the prevention of gross treatment delivery errors, whether they are as a result of device malfunction or of human error in the design, evaluation, and execution of the brachytherapy procedure. The development and maintenance of a good quality assurance program (preferably to a recognized, auditable standard) are a major factor in effective radiation protection in brachytherapy.
10.2 QUALITY ASSURANCE ISSUES IN BRACHYTHERAPY PROTECTION
10.2.1 Source identification and description Sources should be kept safely, with manufacturers' data sheets or test reports appended to local documentation, so that in the event of loss or damage as much information as possible, relevant to each particular source, is available [ 17]. The physical and chemical composition of the radioactive source should be noted, including the presence of any radioactive impurities. Encapsulated sources with a long half-life should be clearly identifiable and distinguishable from each other. A closed circuit television (CCTV) camera with a closeup lens is useful for this purpose.
The following information should be recorded: 1. 2. 3. 4. 5. 6. 7.
the radionuclide, energy, emissions due to decay the source encapsulation the activity on a given date the serial number or other distinguishing mark the date of receipt the normal location of the source the recommended working life of the source (when appropriate) 8. the date and manner of disposal (when appropriate). Some sources (e.g., iridium-192) require a storage period after initial production to allow the decay of short-lived impurities: the user should ensure that such procedures are followed. Sources (such as cesium needles and tubes) should be assessed to determine whether they are safe to use or should be replaced.
10.2.2 Source integrity checks: leakage and contamination tests Before sources are used clinically, and subsequently at regular intervals, checks should be made to ensure that they are not leaking and that the distribution of radioactivity within the source is as expected and acceptable for clinical use, and that this distribution does not change with the course of time. Long-lived brachytherapy sources are doubly encapsulated for mechanical strength and to prevent leakage of radioactive material in the event of source damage. Sources obtained from manufacturers are issued with leakage test certificates, which describe the tests that have been carried out: these include immersion tests at different temperatures and wipe tests [18]. For cesium sources manufactured in the UK, the safety level is taken to be 200 Bq (5 nCi). It should not be assumed that new sources are necessarily free from surface contamination. Long-lived sources in clinical use should be leak tested at least every 2 years; annual tests should be made on sources that have been in use for several years. New, encapsulated sources should be wiped with a swab or tissue moistened with water or ethanol and measured using a Geiger-Muller or scintillation counter capable of detecting 200 Bq (5 nCi). The method used should keep radiation exposure to a minimum and it should swab the outside of the source without causing abrasions to the source capsule. When using iridium (iridium-192) and gold (gold198) wires and seeds, the principal hazards are those caused by scoring of the surface and particulate fragmentation when the wire is manipulated or cut without appropriate equipment. Handling tools and cutting equipment should be monitored and decontaminated regularly and cutting should only take place under controlled conditions in protected areas. In the case of beta-ray sources such as strontium-90/
Source handling and associated protection issues 149
yttrium-90 ophthalmic applicators, special care must be taken because the surface of the applicator is particularly delicate. Leak tests must be performed at least annually.
pellets (which are used to identify active source positions for treatment planning) provides information to crossreference active sources and dummy source positions.
10*23
103 SOURCE HANDLING AND ASSOCIATED PROTECTION ISSUES
Source strength measurements
Before being used clinically, sources should be calibrated by the user. A useful measurement device is a re-entrant ionization chamber (i.e., an isotope calibrator). In the case of high-activity sources, accurate calibration can also be achieved with an ion chamber. The calibration of both these instruments should be traceable to a national standards laboratory. To achieve the required accuracy in the prescribed dose, source calibration accuracy should be better than ± 5% of the true strength. Most manufacturers are unable to provide calibrations of this accuracy and the user must carry out independent measurements, but reference should always be made to the manufacturer's calibration certificate (agreement within ± 10%).
10*2.4
Autoradiography and radiography
These techniques can be used together or separately to provide information about the distribution of radioactive material within the container and positional information about individual sealed sources in radioactive source trains. Autoradiography is useful for checking the uniformity of wire sources and ribbons or radioactive seeds and pellets. When radioactive wires are cut, it provides a means of recording particulate contamination. Needle sources might have two or more cells: the source activity and distribution in each cell can be checked with this method. Autoradiography and radiography should be used to check the configuration of single and multiple sources in preloaded source trains. In the case of afterloading machines whose source configuration can be programmed, autoradiographic checks should be carried out at commissioning and after machine service or source or catheter replacement to ascertain the precise location of each source and the integrity of software and machine function. Applicators into which sources are loaded, either by hand or automatically by machine, should be checked by autoradiography before being put into clinical use, and thereafter annually. Moulded wax is useful for positioning and supporting the applicators. Lead-foil markers embedded into the surface of the wax can be used to provide identification marks and scales which are imaged on film by electron emission. The method allows precise comparisons to be made of different applicators and provides a radiographic record of the location of the radioactive sources inside loaded applicators. Radiography of the same applicators containing dummy
The steps required for the safe use of radioactive sources from storage, preparation, transportation, insertion, removal, and cleaning are discussed in the following paragraphs
103*1
Storage of sealed sources
Clean sources should be kept in a locked, radiationshielded safe designed to allow the safe visualization of sources and identification marks: the safe should be compartmentalized to permit easy and fast access for removing individual sources and for carrying out stock checks. The safe should be situated close to the source preparation bench for easy access. The storage safe or container should be swab tested annually and any radioactive contamination found should be removed and its source of origin identified. A detailed inventory of the number, type, and activity of sources in the store must be kept in addition to details of sources being used in patients. When there is a large number of long-lived sources, it is helpful to use a display board or computer spreadsheet to track and record the whereabouts of each individual source. An audit should be carried out at regular intervals for every source in storage or in use. In the UK, an independent audit should be made annually by a senior person nominated by the employer. Lead carrying pots should be monitored after transfer of sources to ensure that all sources have been removed. To reduce the chances of source loss, it is useful to have gamma-ray alarms at the exits to areas where sealed sources are routinely used.
103.2 Preparation of sources and applicators for clinical use Radioactive sources should not be issued or used clinically without a written request from an authorized person: transfer of the source(s) should be recorded, and responsibility for the source taken by different signatories during each stage of the source transfer, implant, and source return. Manipulation of sources should be with long, lowpressure forceps to avoid mechanical damage; the forceps should be monitored and cleaned after use. Wire sources should be cut only with an appropriately designed cutter. Wire sources with cut ends should be
150 Radiation protection in brachytherapy
sealed in plastic tubing before being inserted into body tissues. The tools used to prepare wire sources (cutters etc.) should be labelled and used only for this purpose, checked for contamination at least twice a year, and kept sharp. All handling must be carried out behind protective barriers which reduce the radiation to the abdomen and chest. A protected observation screen is also useful to reduce the dose to the head and eyes. The speed and skill of the operator are also important. It is good practice to limit the number of sources out of the protected storage area at any one time. When necessary, sources should be sterilized before clinical use. The efficacy of the process must be checked. The source manufacturer should be consulted about the effect of sterilization on source integrity: the sterilization process must not be detrimental to the containment of the radioactive source. Sources such as cesium needles and tubes should not be exposed to temperatures above 180°C. Iridium-192 is baked 'dry' at 150°C for 1 hour. Some brachytherapy techniques make use of empty applicators, needles, or catheters into which the sources are after-loaded. These devices should be checked before and after use to ensure that they are mechanically sound and free of contamination. Figure 10.2 Long-handled carrying pot.
1033
Transportation of sources
Sources are carried from the preparation bench to the patient in a specially designed lead pot. For iridium-192 wire, preloaded in plastic tubes, the pot is designed to be sufficiently protective whilst not being too heavy to carry (Figure 10.1). Alternatively, iridium hairpins and cesium sources can be transported in a long-handled, lead-lined pot which uses the principles of distance in addition to protective material to provide a reduction in radiation dose (Figure 10.2). Associated documentation accompanies the sources.
Figure 10.1 Iridium wire-carrying pot.
10.3.4
Insertion of sources into patients
Mobile protective lead shielding barriers are used wherever practicable around the patient's bed. Optimal thickness is 2-2.5 cm lead, which reduces the dose recorded on film monitors worn at waist level to 50% of the dose recorded when worn on the chest. Faulkner et al. [19] recommend wearing monitors at chest level when lead shields are used, but at abdominal level when remote after-loading systems are universally used. Using long-
Source handling and associated protection issues 151
handled forceps, the physicist and radiotherapist load the active sources into the patient as quickly as possible, fixing each source in place as it is positioned. The function of the radiation detector in the treatment room is checked and warning notices posted outside the protected room for the duration of treatment (in addition to illuminated signs warning of radiation dose in controlled areas). A lead pot and long-handled forceps are left in the room with the patient in case of emergencies. If the patient is moved at any time from the protected room (such as for X-ray), a radiation warning notice should be prominently displayed on the trolley or chair in which they are transported. The use of small sealed sources, such as iodine-125 seeds, has great protection advantages (provided that the sources are not lost), as shown by Hilaris et al. [20,21]. Exposures of the order of 2 |LlSv h"1 per 37 MBq at 0.75 m from a patient with a prostate implant have been measured by Liu and Edwards [22]. The greater protection hazard occurs if all sources cannot be accounted for. It is good policy to X-ray patients as soon as possible after implantation and to keep good account of the total number of sources used.
103.5
Treatment delivery
The time that the sources were inserted into the patient and the proposed removal time (dependent on dosimetry calculations)) are recorded. The nurse in charge of the patient accepts responsibility for the custody of the radioactive sources whilst they are in the patient and ward by signing the appropriate paperwork. The time that the nurse may spend with the patient (daily nursing time) is calculated from Table 10.2 (used for iridium implants) and is dependent on the total activity implanted. Permissible times, indicated by radiation warning notices, are intended as a guide to nursing staff. Nursing procedures can safely be carried out, but unnecessary time must not be spent close to the patient while the warning notice is displayed. The times given in Table 10.2 are such that a nurse remaining at a distance of 50 cm from the patient for the Table 10.2 Times for different total activities of iridium administered
0.3-0.6 GBq (300-600 MBq) 0.6-1.3 GBq (600-1300 MBq) 1.3-2.5 GBq (1300-2500 MBq) 2.5-3.8 GBq 3.8-6.5 GBq 6.5-13 GBq 13-20 GBq
1h 30min 15min 10min 5min 3 min 1 min
In these cases the time to be indicated remains the same for each day the radioactive sources are in position.
time indicated each day would, after 5 consecutive days, have received a dose equal to his or her average weekly dose limit. The nurse in charge of the ward is responsible for ensuring that the time on the warning notice is set correctly each day in accordance with Table 10.2 and for removing the notice on the day indicated. While the brachytherapy treatment is taking place, the protected room becomes a controlled area and staff enter under local rules only. Emergency procedures (e.g., unplanned removal of source, patient bleed) are documented in the local rules and are dependent on the site and activity of implant. Visitors are discouraged, but if absolutely necessary are allowed to spend no more than the daily 'nursing time' with the patient. 103*6
Removal of sources from patients
When sources have to be removed from a patient, either at the end of treatment or, occasionally, if a source has been inadvertently displaced, they should be removed carefully to avoid patient trauma and source damage. Sources should be placed in a lockable, shielded container lined with a plastic pot containing bactericidal fluid. When removing iridium wire sources contained in plastic tubes, great care must be taken not to cut the wire when removing the sources. The patient must be checked with a Geiger-Miiller monitor after removal of sources from the treatment room to confirm that all radioactivity has been removed from the patient. The sources should be returned to the laboratory/ store as soon as possible after removal from the patient, where they should be counted, checked for damage, and stock records completed. Responsibility for the sources is transferred from the ward nurse to the person who transfers the sources and then to the source curator. Only after all sources have been accounted for should the patient be allowed to leave hospital. 103*7
Source and applicator cleaning
It is necessary to clean sources that have come into contact with body tissues before they can be stored and/or re-used. Immersion in bactericidal fluid on removal from the patient prevents biologically active material reaching the laboratory store. Source manufacturers should advise about cleaning procedures: damage to source capsules can occur as a result of chemical attack if inappropriate cleaning agents are used. Wire sources that are to be re-used should be inspected for damage and re-measured prior to use. After removal from the patient, applicators (catheters etc.) should be immersed in bactericidal fluid, cleaned, and inspected for damage. Radioactive sources are often inserted into patients in sealed plastic tubing, plastic applicators or stainless-steel
152 Radiation protection in brachytherapy
tubes. It is good practice to check these source-carrying devices for radioactivity with a Geiger-Miiller monitor after each use.
that treatment times can be reduced (low-medium dose rate treatments). Patients can be treated as out-patients for fractionated high dose rate treatments (with cesium137 and iridium-192).
10.4
10*4.2
AFTERLOADING
10.4.1
Manual afterloading
Radiation exposure can occur during:
General
Afterloading in brachytherapy has achieved the biggest improvement in radiation protection of staff and, indeed, also of patients. It has minimized the exposure to the technical staff involved in the preparation and transportation of sources and it has reduced the dose to medical staff. Staff on the wards who nurse patients have their doses reduced by the use of automatic afterloading. During afterloading techniques, applicators or catheters are inserted into a patient and subsequently loaded with radioactive sources. This method allows optimal positioning of the applicators (or catheters) without any radiation exposure to the clinician or support staff. Afterloading of the radioactive sources can either be achieved by hand or remotely with specially designed equipment that drives one or more of the sources into the patient. In both situations, afterloading occurs after the patient has had localizing radiographs and, in the case of conventional brachytherapy, has returned to the ward or room where treatment is to be delivered. Although manual afterloading largely eliminates exposure of the clinician, some staff have still to work in significant radiation fields in the course of nursing the patient. The evolution of afterloading has been linked closely with the improved availability of flexible wire sources and cobalt-60, cesium-137, and iridium-192 sources designed specifically for use in remote afterloading equipment. Remote afterloading systems have their own in-built storage safes, the design of which ensures that the maximum exposure rate on the surface of the safe is less than 25 |lSv h~'. Table 10.3 summarizes the source types that are currently available. Remote afterloading permits the use of sources of higher air kerma rate so
transfer of source from storage and preparation for use transfer from preparation bench to patient irradiation of patient removal of sources from patient and transfer to storage removal from applicators, cleaning, returning to storage.
10.4.3
Remote afterloading
Radiation exposure can occur when: machine/door interlocks do not function sources are changed sources stick. Regular quality assurance of the remote afterloading machine helps to prevent machine malfunction and unnecessary radiation exposure.
10.5 DESIGN ASPECTS OF BRACHYTHERAPY TREATMENT ROOMS
10.5.1
Types of treatment room
Brachytherapy covers a range of different forms of treatment and includes surface moulds as well as intracavitary and interstitial techniques. The duration of treatment can range from several days as in low dose-rate (LDR) interstitial implants, or to 1-2 h per day (pulse dose rate, PDR) or to a few minutes as in high dose-rate (HDR) brachytherapy. The treatment may be continuous or fractionated. A
Table 10.3 Properties of principal brachytherapy nudities
'"Co 125,
137
Cs lr 198 Au 192
5.27 years 60 days 30 years 74 days 2.7 days
1.17,1.33 0.035 0.66 0.3-0.6 0.41
309
33 78 113 55.5
HVT = half value thickness; TVT = tenth value thickness.
12 0.03 6.5 4-5 3
40 0.1 21 15 11
206 157 113 135
Design aspects of brachytherapy treatment rooms 153
large number of different nuclides are available for clinical use, some of which have short half-lives and can be implanted permanently; nuclides may be of low energy and require little shielding, or of high energy requiring specially designed rooms with thick concrete walls. The wide range of techniques used and the large selection of available nuclides necessarily influence the design of a particular treatment facility. In the past, patients have been treated in open wards, but this should be avoided whenever possible. A specially designed and protected single room is optimal, but a single room with appropriate shielding added or a shared room with mobile shielding and/or inter-bed shielding can be used. For HDR treatments where source activities are 37-370 GBq, the room might be specially designed or a modified shared radiotherapy treatment room. Experience has shown that it is advantageous to have a physics room (it need not be large) nearby with facilities for storage and cleaning of sources, applicators, and accessories. Ideally, the brachytherapy facility should be designed according to the types of treatment envisaged. This section describes factors that affect such a design. Table 10.3 summarizes the properties of the nuclides frequently used.
10.5*2
design concept at an early stage in the planning process. It is usual for hospital planning (or estate management) to be the interface between the consultant architect and other interested professional groups within the hospital. The physicist must ensure that the planning team understands the need for a room with adequate radiation protection and should ensure throughout the planning, building, installation, and commissioning periods that the room is constructed and equipped accordingly. At the design stage, consideration should be given to the following. THE INTENDED USE OF THE ROOM This will be influenced by the current techniques in use, the number of patients to be treated simultaneously, and whether or not the facility will also be used by nonradioactive patients who might have special requirements. The total amount of radioactivity and the energy of photon emissions will determine the radiation shielding requirements. The long-term use of the room should be considered, especially in relation to techniques that might require additional shielding. The storage of radioactive sources, prior to or immediately after removal, should be considered.
The design brief THE LOCATION OF THE ROOM
The details of this brief will depend upon the type of brachytherapy room being planned. There are various considerations that are common to most brachytherapy requirements: • The treatment room should be constructed to allow implementation of the prescribed treatment with precision and safety. • The room should have sufficient shielding or be so isolated that hospital personnel, other patients, and members of the general public will not receive incidental radiation exposure in excess of specified limits. Radiation levels in and around the room should be measured and calculated and staff working in the area should be monitored. • The layout of the room should allow patients to be nursed safely, efficiently, and with as little inconvenience and discomfort to the patient and staff as possible, and to allow all the procedures associated with the medical care of the patient, including any necessary emergency procedures, to be accomplished. • Particular attention should be paid to the facilities for removing sources and storing them until they can be removed to a more permanent store. • The room should be constructed and equipped so that it can be used for nursing non-brachytherapy patients. It is important that the radiation protection adviser or an experienced radiation physicist is involved in the
The room should be reasonably accessible from facilities such as the operating theatre, X-ray, and computed tomography (CT). If the room is to be used for inserting radioactive sources by hand, it should ideally be located within reach of the source preparation laboratory. The location of the room must also be considered in relation to surrounding areas, public corridors, and waiting areas. Areas immediately adjacent and areas above and below the proposed site must be surveyed early on in the planning process to ensure that adequate shielding can be incorporated. THE SIZE AND LAYOUT OF THE ROOM
The size of the room will be determined by the available space, and whether or not en-suite facilities are to be provided. The detailed requirements will also depend upon the type of treatments given. In general terms, it will be necessary to ensure adequate access for patients on beds and/or trolleys, to provide enough space for local shielding, such as bed shields, and for inserting and removing radioactive sources into patients from behind protective shields. En-suite facilities may be provided so that patients with implants can be confined to a clearly defined 'controlled area.' Thought should be given to the arrangements to be made for contingencies such as medical emergencies that might require the rapid removal of radiation sources.
154 Radiation protection in brachytherapy
10.53
Radiation protection requirements
It will be necessary to estimate the likely radiation levels and the associated hazards to occupationally exposed staff and also members of the general public. ICRP 60 (1990) [23] specifies that the dose limits should be: occupationally exposed staff:
20 mSv per year
general public:
1 mSv per year
and the dose limits in the UK regulations (IRR 1999) are based on these. In situations that require an estimate of the dose likely to be received or where room, wall, or door protection calculations are undertaken, occupancy factors can be used for various areas in the vicinity of the brachytherapy room. Occupancy factors indicated in NCRP Report 49 [24] are as follows: T= 1
work areas such as laboratories and occupied space in adjacent buildings
T = 1/4 T= 1/16
doors, staff rooms, lifts, parking areas waiting rooms, toilets, staircases.
It is sometimes policy to work to lower levels than those specified and, apart from local reasons, consideration should always be given to the possibilities that: current radiation exposure limits will be reduced; brachytherapy techniques and technology might change; the use of rooms above and below and adjacent to the facility might change. It is therefore appropriate, whenever possible, to reduce the dose rate outside the room to 2.5 (iSv h~' (the maximum dose rate allowable for the general public). Within the room, the dose rates will be such that 'it is necessary for any person who enters or works in the area to follow special procedures designed to restrict significant exposure to ionising radiation' (IRR 1999 [7]), and therefore the room will be designated as a 'controlled area' under those regulations. Brachytherapy sources are not collimated and shielding calculations should take into account the primary photon energy of the emission. Radiation emanating from sources within a patient will
be attenuated by the patient's body, but, except for lowenergy emitting sources such as iodine-125, this attenuation is only 10-20%. Table 10.4 shows the thickness of concrete required to reduce the dose rate to 7.5 (iSv Ir1 and 2.5 jiSv h'1 at 3 m from various activity sources allowing for 10% absorption in the patient. Room doors need to be lead lined, with a maximum thickness of lead of about 6 mm (any more is too heavy). The position of the patient's bed should not be in a direct line with the door, so that door protection can be kept within acceptable limits, as can the exposure of nursing staff as they approach the patient. Access to the room should be observable from the nursing station, and remote viewing of the patient by CCTV is useful. The use of a two-way intercom also helps to reduce the time spent by the nurse in the radiation environment. If more than one protected room is required, rooms can be built adjacent to each other so that the total shielding requirements can be optimized. This arrangement, however, is not always advantageous when nursing expertise for certain types of patients (gynecological, head and neck) is highly specialized and focused on specific wards.
10.5.4
Additional requirements
There are certain requirements that are common to all brachytherapy patient facilities. These are: • A definitive area around the patient (where special procedures are required under IRR 1999 [7]) must be designated as a controlled area. A supervised area, where dose rates need to be kept under review, might in some circumstances be identified. Staff working within the controlled area should be monitored. • Members of the general public who might wish to visit a brachytherapy patient undergoing treatment may do so provided they comply with the IRR 1999 regulations. Visiting in general is discouraged. • The dose rate in the area of the bedside should be displayed visually. A beta/gamma mini-monitor type of indicator is suitable for this purpose. • A portable dose rate Geiger-Muller monitor should be available near to, but outside, the treatment area so
Table 10.4 Thickness of concrete (cm) required to reduce the dose rate to the specified values at 3m from various sources, assuming 10% attenuation in the patient
Implants Gynecological intracavitary MDRafterloading HDRafterloading Implants PDRafterloading HDRafterloading
37
Cs Cs 37 Cs fl Co 92 lr 92 Ir 92 Ir 37
1.85 7.4 22.2 185 1.85 37 370
50 200 600 5Ci 50 1Ci 10 Ci
8 19 28 68 10 31 44
19 28 36 77 18 36 51
Design aspects of brachytherapy treatment rooms 155
that bed linen etc. can be checked for unsuspected sources and also the patient can be checked to ensure that all sources have been removed. • Radiation warning notices should be posted. • Suitable source container(s) and handling tools should be available. • Preferably, there should be CCTV monitoring and a two-way intercom to the patient.
10.5*5
out in Figure 10.4 shows that the afterloading machine is housed in a separate, small, lockable room. The source transfer tubes pass through the wall of each treatment room into the treatment catheters. The air compressor is located in a cupboard away from the patient area to minimize disturbance. The dose rates shown correspond to 15 GBq of cesium-137 in each patient. The concrete wall thickness and lead door thicknesses are as shown.
Typical treatment rooms
A) FOR GYNECOLOGICAL INTRACAVITARY TREATMENTS
Figure 10.3 shows a typical layout for a gynecological intracavitary patient. The walls are 30 cm concrete, which reduces the primary radiation from a cesium-137 source by the equivalent of two tenth value thicknesses. Mobile lead bed shields are useful for reducing dose rates. A typical shield might be of 2.5 cm thickness, weigh 100-200 kg, and have a shielding area of 100 x 600 cm. Usually, shields are on lockable wheels. Inter-bed shielding can be used to reduce the dose rate between patients. This is usually 1 cm or 1.5 cm thick, but requires steel frame supports. Under-bed shielding is also feasible if the dose rate in the area immediately below the patient is too high. B) LOW DOSE-RATE/MEDIUM DOSE-RATE REMOTE AFTERLOADING
In the case of remote afterloading, the treatment room door can be interlocked with the machine. The room lay-
Figure 10.4 Two-bed Selectmn-LDR suite.
l37
Cs
60
Co
Ratio
A
1. 9
5.6
2.9
B
2.7
7.8
2.8
C
1. 3
3.8
2.9
D
1. 3
1. 5
1. 5
E
I.I
1. 7
1. 5
F
1. 4
2.2
1. 6
G
I.O
2.0
2. 1
H
24.0
45.0
1. 8
I
1 9.0
33.0
1. 7
J K
39.0
78.0
2.0
1 90.0
780.0
4.0
Figure 10.3 Dose rates at various positions around an afterloading room (in mv h-1 for cesium-137 and cobalt-60 sources of AKR WOO uGy /?"' at 7 m).
156 Radiation protection in brachytherapy
Position
376 376 94 94 32 19 9 2 I <0.5
A B C D E F G H I J
1 867 1 502 5I5 494 75 36 19 <0.5
28! I 3004 3I76 1 502 1 70 79 36 8 2 <0.5
Figure 10.5 Dose rates in u5v h ' at specific points for a 370 GBq iridium-192 source in a tissue-equivalent phantom placed in positions 1, 2, and 3.
C) HIGH DOSE-RATE REMOTE AFTERLOADING
HDR treatment rooms require concrete walls of thickness 50-80 cm, depending upon the size of the room and the nuclide being used (see Table 10.4). Figure 10.5 shows a purpose-built radiotherapy theatre, indicating dose rates in |J,Sv h'1 at specific points for a 370 GBq iridium-192 source in a tissue equivalent phantom placed at positions 1, 2, and 3. Useful general references for brachytherapy protection are to be found in reference 25.
REFERENCES 1. ICRP (1977) Recommendations of the International
Medical Examination or Treatment) Regulations 1988, Statutory Instruments 1988 No. 778, London, HMSO. 10. N RPB, HSE DoH (1985) Guidance Notes for the Protection of Persons Against Ionising Radiations arising from Medical and Dental Use. London, HMSO. 11. Health and Safety Executive (1985) Approved Code of Practice: The Protection of Persons against Ionising Radiation arising from any Work Activity, The Ionising Radiations Regulations 1985. London, HMSO. 12. Hea Ith a nd Safety Executive (1972) Code of Practice for Protection of Persons against Ionising Radiation from Medical and Dental Use. London, HMSO. 13. The Medicines (Administration of Radioactive Substances) Regulations 1978, Statutory Instruments 1978 No. 1006. London, HMSO. 14. Ionising Radiations (Outside Workers) Regulations 1993. London, HMSO.
Commission on Radiological Protection. Ann. ICRP, 1(3). 2. ICRP (1978) Statement from the 1978 Stockholm meeting
15. International Standards Organisation (1980) Sealed Radioactive Sources-General, IS01677. Vienna, ISO
of the International Commission on Radiological Protection. Ann. ICRP, 2(1). 3. National Radiation Protection Board (1993) Radiation Exposure of the UK Population -1993 Review. NRPB R263. London, HMSO. 4. ICRU (1993) Quantities and Units in Radiation Protection
Publications. 16. The Radioactive Material (Road Transport) (Great Britain) Regulations 1996, Statutory Instruments 1996, No. 1350. London, HMSO. 17. International Standards Organisation (1980) Sealed
Dosimetry, ICRU Report 51. Bethesda, Maryland, ICRU
Radioactive Sources - Classification, ISO 2919-1980.
Publications Office.
Vienna, ISO Publications.
5. IPSM (1986) Radiation Protection in Radiotherapy, IPSM
18. International Standards Organisation (1979) Technical Report, Sealed Radioactive Sources - Leak Test Methods,
Report 46. York, IPSM Publications. 6. Council Directive 97/43/Euratom, 30 June 1997. 7. Ionising Radiations Regulations 1999 Statutory Instrument 1999 No. 3232. London, HMSO. 8. Ionising Radiations Regulations 1985 Statutory Instrument 1985 No. 1333. London, HMSO. 9. Ionising Radiations (Protection of Persons Undergoing
ISO/TR4826-1979 (E). Vienna, ISO Publications. 19. Faulkner, K., James, H.V., Chappie, C.L and Rawlings, D.J. (1996) Assessment of effective dose to staff in 20.
brachytherapy. Health Phys., 71(5), 727-32. Hilaris, B.S., Holt, J.G. and Germain, J. St (1975) The Use of lodine-125 Interstitial Therapy. US Department of Health,
References 157 Education and Welfare Publication 76-8022, Rockville, Maryland. 21. Germain,]. St (1975) Iodine seed brachytherapy. In Handbook of Interstitial Brachytherapy, ed. B.S. Hilaris. Acton, MA, Publishing Sciences Group, 117-28. 22. Liu, J. and Edwards, P.M. (1979) Radiation exposure to medical personnel during iodine -125 seed implantation of the prostate. Radiology, 132,748.
23. ICRP Publication 60 (1990) Recommendations of the International Commission on Radiological Protection, Ann. ICRP, 21(1-3), 11-201. 24. NCRP (1976) Structural Shielding Design and Evaluation for Medical Use of X-rays and Gamma Rays of Energies up to 10 MeV. Report No. 49. NCRP Publications. 25. Godden, T.J. (1988) Physical Aspects of Radiotherapy, Bristol, Adam Hilger.
This page intentionally left blank
PART
Theradiobiologyof brachytherapy
11 Theradiobiologyof low dose-rate and fractionated irradiation 12 Dose-rate effects with human eel Is
161 180
13 Radiobiology of high dose-rate, low dose-rate, and pulsed dose-rate brachytherapy 14 Predictive assays for radiation oncology
189 205
15 Principles of the dose-rate effect derived from clinical data
215
II
This page intentionally left blank
11 The radiobiology of low dose-rate and fractionated irradiation JOELS. BEDFORD
11.1
INTRODUCTION
For exposure to sparsely ionizing radiations such as X-rays or gamma-rays, the degree of a biological effect produced can depend as much on the dose rate as on the total dose received. The importance of dose rate and dose fractionation effects has been recognized for more than 70 years. Studies of Regaud and his collaborators were perhaps the first to show the potential therapeutic advantages of dose fractionation in the treatment of patients with cancer by radiation [1,2]. Since that time, the evolution of treatment regimes involving dose time variations have increasingly improved cancer radiotherapy and the evolution continues even today. Not only are dose rate and dose fractionation important factors in cancer radiotherapy, but also in connection with the mutagenic and oncogenic hazards of radiation exposure [3-5]. Generally, reducing the dose rate decreases the biological effectiveness. Put in another way, decreasing the dose rate generally increases the dose necessary to yield the
same level of effect. A number of factors can contribute to the dose rate or dose fractionation effect, depending on the conditions and cell or tissue system involved. For example, in a tissue or tumor exposed over a period of weeks or months, cells may migrate into or out of the radiation field, or the oxygenation status may change to alter the intrinsic ratiosensitivity of the cells during the course of treatment. Such factors are not considered in this chapter. Radiation responses in normal tissues and tumors largely reflect the collective effect on individual cells comprising them, and the discussion below focuses on three major processes governing dose rate effects at the cellular level: repair, cell-cycle-dependent radiosensitivity, and perturbations in cell proliferation. Repair is usually defined as a 'restoration to good, sound, or healthy condition.' Various more or less stringent definitions of the term have been used regarding repair of radiation damage and it can apply to organs, tissues, cells, subcellular structures, or molecules which comprise them. It is useful to have at least an outline of the broad perspective for a subject so important to radiobiology as repair. It is among the most important of
162 The radiobiology of low dose-rate and fractionated irradiation
many contributions of radiation biology to biology in general, ranking with other contributions such as the first line of evidence that DNA and not protein is the genetic material (based on absorption versus action spectra), and the first demonstration of mutagenesis using X-rays and the utility of such mutants for genetics.
11.2
THE GOOD OLD DAYS
Repair of radiation damage, in the sense of 'restitution' of chromosome breaks after exposure, was recognized early in the history of radiation cytogenetics, perhaps most notably by Sax [6]. Further, mathematical descriptions of the dose rate effect in terms of the dependence of chromosomal aberration yields on the duration of exposure were presented by Lea and Catcheside [7], and Marinelli and coworkers [8] and discussed in detail by Lea in his classic book, Actions of Radiations on Living Cells [9]. As we shall see later, this is very pertinent to the dose rate effect for cell killing that largely governs time-dose relationships in radiotherapy. The first observation of the enzymatic repair of radiation damage to DNA had its origins in the reports of photoreactivation of ultraviolet damage leading to lethality in spores of the fungus Streptomyces griseus by Kelner in 1949 [ 10], and in the same year by Dulbecco in the T group of coliphages [11]. About a decade later, this was shown in Hemophilus influenzae and Saccharomyces ceravisae to result from a light-mediated enzymatic (photoreactivating enzyme or photolyase) splitting of cyclobutane-type pyrimidine dimers in DNA induced by the ultraviolet [12,13]. More relevant for mammalian systems was the discovery of excision repair by Setlow and Carrier in 1964 [14] and also in the same year by Boyce and Flanders [15]. These and other repair processes, involving (at first) ultraviolet damage in DNA, preceded work with repair of ionizing radiation damage in mammalian cells. Ionizing radiation produces negligible quantities of pyrimidine dimers, but does produce other lesions such as base damage and DNA strand breaks (which are discussed below).
shoulder region of the curve (survivals below about 10%) would contain sublethal damage capable of interacting with further damage to become lethal. They questioned whether this sublethal damage might remain in surviving cells, in which case their dose response at some later time would not be 'shouldered'. Alternatively, if the sublethal damage were repaired, the cells would be expected to respond as if they had never been irradiated, i.e., the surviving cells would display the same shouldered survival curve for subsequent irradiation. The latter was found to be the case, as is illustrated in Figure 11.1 from their early work. The curve indicated by filled circles in Figure 11.1 illustrates a dose-response curve for irradiations requiring only a few minutes each - high dose rate (HDR) or 'acute' exposures - over a range of doses from 0 to about 12.5 Gy. Curves starting at a dose of 5.05 Gy illustrate dose-response curves for cells surviving a first dose of 5.05 Gy followed by various additional doses given either immediately after the first dose (filled circles) or 18 h following the first dose (open circles). During the time interval between the first and second doses, the surviving cells 'restored themselves to good (original) condition.' They had repaired this so-called sublethal damage so they again had to accumulate damage for cell killing. This sublethal damage repair (SLDR) is a repair process operationally defined in terms of the observa-
113 SPLIT-DOSE RECOVERY FROM SUBLETHAL DAMAGE IN MAMMALIAN CELLS For ionizing radiation damage in mammalian cells, the first direct demonstration of a cellular repair process affecting cell killing that could explain dose rate and dose fractionation effects seen in mammalian tissues or tumors was provided by Elkind and Sutton [16]. They reasoned that because the shouldered survival curves for mammalian cells exposed to X-rays or gamma-rays indicate the involvement of a damage accumulation process in cell killing, then cells surviving a dose beyond the
Figure 11.1 Initial survival curve (closed circles) and fractionation survival curve (open circles) for 'clone A' cultured Chinese hamster cells. The fractionation survival curve was determined 18.7 h after 5.05 Gy, and the curve is normalized to the survival corresponding to 5.05 Gy. RE.- plating efficiency. (Reproduced with permission from Elkind and Sutton [16].J
The cell-cycle complication: a heterogeneous population 163
tions demonstrating the phenomenon, i.e., the increase in the fraction of cells surviving. It says nothing about what is being damaged and repaired. While survival curve shapes were discussed at great length in terms of target theory in the early literature, it was appreciated by nearly everyone in the field some 30 years ago or more that the simplifying assumptions necessary for target theory to be useful were so overwhelmed by such complexities as various repair processes and heterogeneity of radiosensitivity within cell populations that we could expect to learn very little about radiobiological mechanisms by such analyses. This does not mean target theory has not or cannot be useful or informative in certain instances. What it does mean is that it is not very useful for determining the nature, size, and number of critical targets whose damage can kill cells after exposure to sparsely ionizing radiations.
11.4
RECOVERY HAS ITS UPS AND DOWNS
One interesting aspect of the repair process, studied by Elkind, was the reaction kinetics of sublethal damage repair. By giving two 'acute' X-ray doses separated by various periods of time, he and his colleagues, and later others, found that the repair process was essentially complete by about 2 h in randomly dividing log phase cultures of V79 Chinese hamster cells. This is shown in Figure 11.2. All the cultures received the same total dose,
Image Not Available
Figure 11.2 Survival curve (closed circles) and two-dose fractionation curve (open squares) of 'clone A' cultured Chinese hamster eel Is. The two-dose fractionation curve should be read along the lower abscissa. P.E. = plating efficiency. (Reproduced with permission from Elkind and Button [16].j
given in two fractions of 5.05 Gy followed by 4.87 Gy, but were separated in time by periods ranging from 0 up to 30-40 h. What was immediately apparent from such 'split-dose recovery' curves was that repair was essentially complete by 2-3 h. After this, however, survival decreased during the next few hours, followed by another increase. Aside from a few cells in mitosis or late G2 at the time of the first dose, no appreciable cell division occurred, at least during the early period from 0 to about 6 h after the 5.05 Gy first dose. Not until the cell cycle dependence of radiosensitivity was discovered and studied did the explanation for this peculiar cyclic 'split-dose recovery curve' become clear. Prior to this, survival studies in vitro all involved the use of randomly dividing cell cultures where the population consisted of mixtures of cells in all different phases of the cycle (M, Gl, S, and G2).
11.5 THE CELL-CYCLE COMPLICATION: A HETEROGENEOUS POPULATION In the early 1960s, Terasima and Tolmach first showed with synchronized cultures of HeLa cells that cellular responses varied greatly throughout the cell cycle [17,18]. This is illustrated in Figures 11.3 and 11.4, where the experiments were carried out as follows. During mitosis, cells become very loosely attached to the surface of the culture vessel and these were collected by a 'shakeoff method, leaving the interphase cells behind in the flask. Appropriate numbers of mitotic cell populations were inoculated into dishes. After various periods of incubation, different sets of the synchronously progressing cells were irradiated when they were (for the most part) at a particular stage of the cycle. When the dose was the same for all cultures, but the time after mitotic shake-off was varied, the proportion surviving to form colonies varied as shown in Figure 11.3 in the upper panel. Parallel cultures were flash labelled with tritiated thymidine (3H TdR) to monitor the synchronous progression of cells into and out of S phase. For irradiation of mitotic cells (Oh), survival was low, indicating a high sensitivity for this cell cycle phase. As cells progressed into mid-Gl (-2-6 h), the cells were more resistant. At around the Gl/S border and in early S phase, cells were again more sensitive, and as cells progressed toward late S phase and early G2, the cells again became more resistant. The changes were examined in more detail in other experiments by determining complete dose-response curves at different times after mitotic harvest, as shown in Figure 11.4 where the '0-h' curve represents largely mitotic cells, the 4-h curve Gl, the 13-h curve early-tomid S phase, and the 19-hour curve late S to G2. Because there is some variation from one cell to the next in the cell cycle transit times, particularly through Gl, there is an increasing decay in synchrony and therefore the
164 The radiobiology of low dose-rate and fractionated irradiation
Image Not Available Image Not Available
Figure 11.4 X-ray survival of HeLa S3-91V cells at four different times during the division cycle measured from mitosis (0 h). (Reproduced with permission from Terasima and Tolmach [18].J
Figure 11.3 The upper curve shows the fraction of HeLa S3-91V cells surviving after 300 rod of X-rays administered at different times in the division cycle. Mitosis is taken as 0 h. Each symbol represents data from a separate experiment, the time scales of which have all been normalized to a minimum interdivisional time of 18 h. The lower curve shows the fraction of cells in DNA synthesis as found in three separate experiments with synchronized cells. The fraction of labeled cells after 20 min in medium containing H3-thymidine is plotted against time after mitosis (normalized as in the upper curve). (Reproduced with permission from Terasima and Tolmach [18].J
resolution of experimental data on cycle-dependent radiosensitivity with time. Nevertheless, there is clearly a large variation in the radiation response of cells through the cell cycle. Other cells have shown similar cell-cycle-dependent variations in radiosensitivity (e.g., see references 19-26), although the peak of resistance in Gl is not well resolved experimentally in cells with very short Gl transit times. An example of cell cycle variations in sensitivity for Chinese hamster V79 cells, such as those used by Elkind and his coworkers for the split-dose SLDR studies, is shown in Figure 11.5 [20]. Special allowance was made in the curves shown in this figure to eliminate the effect of cross-contamination of resistant S-phase cells from the more sensitive M, late G2, and Gl cells. The sensitivity of cells in different parts of G2 is difficult to determine by the synchronization procedure described above because of synchrony decay during the passage of the
starting population of mitotic cells through their first Gl and S phase, and because G2 transit times are relatively short (about 1-2 h). A modification of the technique, however, allows a much greater resolution for studying
Image Not Available
Figure 11.5 Single-cell survival curves for Chinese hamster cells (V79 line) in different phases of the cell cycle, M, Gp early S (ES), late S (LS), obtained by mitotic selection plus 3HTdR treatment to kill contaminating S-phase cells, except where responses for S-phase cells themselves were studied. A dose-modifying factor of 2.5, which might be expected fora noxic conditions, applied to mitotic cells is shown dotted. (Reproduced with permission from Sinclair [20].)
The ups and downs are now explained 165
G2 sensitivity. This is sometimes called 'retroactive synchronization': cells are first irradiated and then, as a function of time, cells arriving in mitosis are harvested by mitotic shake-off and plated for survival. In this way, Griffiths and Tolmach [24] showed for HeLa cells, and Dewey and his colleagues [25,26] as well as Schneiderman [26] for Chinese hamster CHO cells, that early G2 cells are as radioresistant as late S cells, and late G2 cells are nearly as sensitive as mitotic cells. A sharp transition in radiosensitivity occurs around the so-called X-ray transition point (now often called a 'check point') for G2 cell cycle delay.
11,6 RADIATION AFFECTS CELL-CYCLE PROGRESSION ITSELF Radiation effects on cell cycle progression are yet another factor that influences dose rate effects, and for present purposes are the final area for discussion concerning our present understanding and explanation of observed variations in cellular effects related to dose fractionation and dose rate. That ionizing radiation reduces the mitotic index within a short time after exposure (mitotic delay) has been known for some time (e.g., see references 9,27). This delay has been studied extensively in more recent times, and the timing for the reduction in mitotic index and subsequent recovery clearly indicates the delay is reversible and occurs some time during G2 [28,29]. The production of this effect is very radiosensitive. Appreciable proportions of the cells are delayed by doses of the order of tens of cGy [28-30]. The G2 delay increases with dose and frequently corresponds to about 1-3 h Gy'1 depending on the particular cells and on the stage in the cycle when the cells are irradiated. Most of the extensive work on cell cycle progression delays in cultured mammalian cells was carried out in the 20-year period between about 1965 and 1985 using 'transformed' or tumorigenic cell lines. Delays in Gl or S phase were relatively minor and, in many cases, undetectable in the 0-5 Gy dose range. As it turned out, the generalization or extrapolation of the results to normal or untransformed cells was unwarranted. Some investigators during this period, even as early as 1968 [31], reported appreciable delays in the progression of 'nontransformed' cells from Gl into S phase or in the transition from the non-cycling GO to the cycling state after low dose or low dose-rate (LDR) irradiation [31-34]. More recently, the importance of p53 status in ionizing radiation-induced Gl delay vis-a-vis transformed versus non-transformed or tumor versus normal cells was discovered by Kastan and his coworkers [3.5]. In any event, knowledge of the cell cycle dependence of radiosensitivity with respect to cell killing as well as the effects of radiation on cell cycle progression itself
allowed the observations on the cellular basis of dose fraction and dose rate effects to be largely understood.
11.7 THE UPS AND DOWNS ARE NOW EXPLAINED In a split-dose experiment, such as that discussed in connection with Figure 11.2 using randomly dividing log phase cells, the first dose kills a fraction of the cells, but this fraction is not the same in all portions of the cell cycle. Survival for cells in the most sensitive phases will be much lower and, in resistant phases, much higher than the average. Thus, after the first dose, the population of cells surviving will not be distributed around the cell cycle as it normally is, but will be highly enriched in cells from more radioresistant phases. It is these surviving cells that determine the further reduction in survival measured by the second dose. If the first dose is of sufficient magnitude to bring the survival down to, say, 10% or less, then these surviving cells will still contain sublethal damage capable of interacting with an additional dose. Thus, if the additional dose were given immediately after the first, the survival reduction would effectively continue down along the single dose survival curve. With a time delay, however, three things happen. First, the sublethal damage begins to repair, and the halftime for this process is relatively fast (by most cell cycle transit time clocks) being 0.5-2 h depending on the system. The effect of this repair process on the surviving cells is to make them more resistant to a second dose, so the proportion surviving will increase with an increasing time interval between the first and second doses. This process is 90% or more complete within about 2-4 h. Second, the cells surviving the first dose which were already in the more resistant phases of the cycle begin to progress and, at least for the first few hours, this progression can only be toward a more sensitive state. For initially log phase populations, it is no longer surprising, then, that with increasing time, between about 3 to 6 or 7 h after the first dose, the survival after the second dose actually decreases. The first dose also produces a mitotic and division delay, so the increase in number of surviving colonies with increasing time before the second dose is not due to an actual increase in numbers of surviving cells from cell division, at least for the first few hours. For example, after a first dose of 5 Gy, there would be essentially no cell division for some 5-10 h, depending on the cells. Third, after the mitotic delay, cell division would resume, so instead of having only one viable cell per surviving colony, as would be the case immediately after the first dose, some, and eventually all, would have two or more viable cells, both of which would have to be killed to prevent colony formation at that locus. The influence of SLDR alone, without the complication of cell cycle effects, can best be seen by the simple
166 The radiobiology of low dose-rate and fractionated irradiation
expedient of using noncycling cells. One way is to reduce the temperature between dose fractions to about 20-24°C, where it has been shown that the rate of cell cycle progression is greatly reduced while still allowing the repair process to operate. This is shown in Figure 11.6, also from the work of Elkind and his colleagues [36]. The reduction in survival between about 2 and 6 h, followed by the subsequent rise from 6 to 10 h seen for incubation at 37°C are both eliminated for incubation at 24 °C. Especially appropriate for cell culture applications are 'normal' or 'nontransformed' cells, which form so-called contact-inhibited monolayers. In such monolayers, the cells enter a noncycling G0 state, where they are no longer a heterogeneous population with respect to the radiosensitivity of subpopulations and, of course, where cell cycle progression and cell division during treatment do not complicate the picture. One additional issue that does arise with the use of contact-inhibited monolayer systems as well as organized tissues in vivo is that another, perhaps related, repair process known as 'potentially lethal damage' repair (PLDR), also plays an important role. This is very different from SLDR, at least in an operational sense.
a cell survival experiment, the flasks must, of course, be subcultured and plated at a low enough density to allow surviving cells to form colonies for the surviving fraction to be assessed. As it turns out, the proportion of irradiated cells surviving a single acute dose in such cultures depends greatly on whether the cells are subcultured, diluted, and plated for the colony forming assay immediately after irradiation, or the subculture is delayed for some hours, in which case the survival is much higher. This is illustrated in Figure 11.7 for human fibroblasts.
Image Not Available
11.8 WHAT IS THIS THING CALLED POTENTIALLY LETHAL DAMAGE?
When contact-inhibited monolayers of nontransformed cells (or, for that matter, even plateau phase densityinhibited cultures of transformed cells) are irradiated for
Image Not Available
Figure 11.6 Split-dose survivals for Chinese hamster cells exposed to 2.5-MvX-rays and incubated at the temperatures shown between doses. The split-dose survivals for 24° C were adjusted to the same total dose as for 37° C. (Redrawn from F/fom/etal.[36].;
Figu re 11.7 Survival response to various doses of X-rays exhibited by density-inhibited normal (AG1522) and ataxia (AT5BI) fibroblasts. Irradiated cultures were either subcultured and plated for survival immediately (O, AG1522; A, AT5BI), or aftera24hdelayat37°C(O,AG1522; ,AT5BI). Thesurvival data were fitted to an expression of the form $=read+bdal>wusing a weighted least-squares method. Closed symbols represent survival predicted on the basis of the mean number of metaphase aberrations per cell for AG1522 (%) andforATSBI ( ). Bars indicate 95% confidence limits for the sampling distribution of mean survival estimates from up to four separate experiments. The dashed line represents the predicted survival if cells completely lacked the ability to rejoin any initially sustained chromosome breaks as measured by premature chromosome condensation (PCC) analysis and any excess fragment were lethal. The D0 is approximately 0.17 Gy, suggesting that while A-Tcells may be deficient in the amount of potentially lethal damage (PLD) they can repair, they nevertheless must be capable of repairing a large proportion of the chromosome breaks initially produced. (Reproduced with permission from Cornforth and Bedford [44].;
What is this thing called potentially lethal damage? 167
The interpretation of this phenomenon is that because damage is lethal in some cells under one set of circumstances (e.g., immediate subculture) but is not under another set (e.g., delayed subculture), such damage must be considered not as 'inevitably lethal' but only 'potentially lethal,' depending on the circumstances. Early work in this area was summarized for density-inhibited or plateau phase cultures by Little [37], but the general phenomenon was first demonstrated by Tolmach and his coworkers for cells whose cell cycle progression was inhibited by drugs for various times after irradiation [38]. It was also demonstrated at about the same time for post-irradiation treatment in phosphate-buffered saline [39] and later for post-irradiation treatment in anisotonic salt solutions [40-42]. Even earlier than any of these studies, a similar phenomenon known as liquid holding recovery was described whereby the survival of bacteria could be increased by post-irradiation incubation in nutrient-deficient phosphate-buffered saline [43]. Again, it is of interest to know about the kinetics or rate of PLDR. This is shown in Figure 11.8, using the same non-cycling contact-inhibited normal human fibroblasts
as for the experiments illustrated in Figure 11.7 [44]. In all samples for survival estimates shown in Figure 11.8, contact-inhibited cultures receive a single dose of 6 Gy of X-rays (0.55 Gy mhr1). Either immediately or at various later times, cells were subcultured and plated at lower densities for a survival assay by colony formation. The increase for subculture from 0 up to 6 or more hours after irradiation was from a survival of about 0.01-0.10 (i.e., a factor of 10) and the rate of increase corresponded to a repair half-time of about 1.7 h [44]. There has been some confusion about the apparent similarity in the survival increases in split-dose SLDR and delayed plating PLDR experiments and about the processes underlying these phenomena. Even without a knowledge of the mechanisms involved, however, a difference is clearly illustrated by split-dose experiments carried out under conditions in which the full amount of PLDR occurs in all samples but different amounts of SLDR occur after the first dose, when the second dose is given. Figure 11.9 illustrates this point. In this case,
Image Not Available Image Not Available
Figure 11.8 Survival as a function of post-irradiation incubation time before subculture}cor AG1'522 cells. Densityinhibited cultures were given 6.0 Gy of X-rays, then incubated at 37° Cfor varying lengths of time before subculture and plating. The solid line represents the increase in survival predicted on the basis of the interphase chromosome (PCC) break induction and rejoining kinetics. Bars are approximate 95% confidence limits about the mean number of colonies scored per petri dish. (Reproduced with permission from Cornforth and Bedford [44].)
Figure 11.9 The increase in survival ofAG1522 cells as a function of time between two 6-Gy doses. The solid circles are observed survivals and represent the mean response of two independent experiments carried out in parallel on the same day. The open squares show the survivals predicted from the average number of chromosome aberrations per cell measured for the 0-h and 10-h split doses only. All monolayers were given a 24-h postirradiation incubation period prior to subculture for the estimation of survival or aberration frequencies. The solid line represents the predicted survival as a function of time, t (h), between two equal split doses, where l is the rejoining rate constant for X-ray-induced chromosome breaks. Dotted lines indicate survival predicted within 2 SD of X. Bars represent approximate 95% confidence limits associated with either the mean aberration frequencies ( ) or the mean survival estimates (•). (Reproduced with permission from Bedford and Cornforth
[45].;
168 The radiobiology of low dose-rate and fractionated irradiation
non-cycling contact-inhibited cultures of normal human fibroblasts were given two doses of 6 Gy, each separated by times ranging from 0 to 10 h [45]. In each case, the cultures were maintained in the contact-inhibited state for 24 h after irradiation so that all of the PLDR that could occur from each dose was allowed to occur before subculture for the survival assay. Still, survival increased from about 0.003 for a 0-h separation between doses (the value expected from Figure 11.7 for a 12 Gy dose with full PLDR) up to about 0.015 for a 10-h separation between doses. The half-time for the SLDR process is about 1.7 h. This is perhaps the best way an experiment to distinguish SLDR from PLDR can be carried out without seeing the varying contributions of both PLDR and SLDR, and perhaps other processes, together. If, for example, the cells were subcultured immediately after the second dose, then the survival would increase both because different amounts of PLDR would occur from the damage inflicted by the first dose and because different but increasing amounts of SLDR would reduce the effectiveness of the second dose, for which no extra time for PLDR would be allowed. Thus, studying these processes in contact-inhibited systems offers a way to isolate their individual contributions to dose fractionation and dose rate effects. Another factor for the study of cellular radiation responses relevant to normal tissue effects is that virtually no normal tissue contains cells existing in the abundant nutrient conditions of in-vitro culture and which are proliferating with growth fractions near 1.0 and doubling times of 12-24 h. Perhaps intestinal crypt stem cells come as close to this unusual situation as any in vivo. The non-cycling contact-inhibited state for normal cells in culture may fail to simulate all conditions in vivo, but the conditions are perhaps a little closer in general to those in most cell renewal tissues, and much closer with respect to the cell cycling status.
11.9
MORE AND MORE DOSE FRACTIONS
To this point, we have not dealt with multiple dose fractionation or the case of continuous radiation exposures at reduced intensities, i.e., with effects of reducing the instantaneous dose rate. The latter situation represents the limiting case of dose fractionation where the total dose is spread over longer times with increasingly large numbers of smaller and smaller doses per fraction. The effect of multiple dose fractionation with different doses per fraction and of continuous irradiation at lower and lower dose rates is shown in Figures 11.10 and 11.11, respectively, for untransformed contact-inhibited noncycling mouse C3H 10T1/2 cells where the effect of repair alone can be assessed without the simultaneous contribution of cell cycle redistribution and cell proliferation effects during treatments [33,34]. There are three
Image Not Available
Figure 11.10 Dose fractionation effects in plateau-phase cultures of 7 OB cells. MuInfraction survival curves were determined for doses per fraction ranging from 7.7 to 0.8 Gy delivered at 12-h intervals 7 days per week. The fractionated survival curves were compared in each case to a corresponding acute dose survival curve generated in that particular experiment. Each point represents the mean of two independently determined survival estimates. Whereas muInfraction survival curves should ideally be exponential, one apparent discrepancy is that, at the larger doses per fraction, the survival curves for the slowly cycling untransformed cells were observed to bend toward lower survivals. (Reproduced with permission from Zeman and Bedford
[34].;
main points to gain from the results shown in Figures 11.10 and 11.11. First, the dose fractionation or dose rate effect is large, requiring nearly three times higher total doses at the lowest dose rates or doses per fraction to produce the same level of effect (for survivals < 10-3) than that required for single acute doses delivered at the highest dose rates. Second, the dose rate effect is most pronounced (change in dose for a given level of effect with change in dose rate) in the range just above and below 2 Gy h'1 (3.3 cGy mhr1). Third, there is a limit below which no further reduction in effect per unit dose occurs with further reduction in dose per fraction or dose rate. There is a residual nonrepairable component of damage under these conditions. PLDR is not a factor in any of these experiments because the contactinhibited cells were not subcultured for the survival assays until at least 12 h after the total doses were delivered and, therefore, full PLDR occurred in every case. Further, it was also shown that the intrinsic
More and more dose fractions 169
Image Not Available
radiosensitivity of the cultures was not changing during the periods up to 4 weeks required to deliver the highest doses at the lowest dose rate. An example of the way cell cycling may influence the dose rate effect is illustrated in Figure 11.12. In this case, randomly dividing log phase cultures were irradiated at
Image Not Available
Figure 11.12 Survival curves for log-phase cultures of C3H 70Tj cells exposed to cesium-137y-rays at dose rates of 55.6, 1.42, 0.49, and 0.29 Gy /r'. Surviving fraction was determined by dividing the number of viable cells present at the end of the exposure by the number of viable cells at the beginning of the exposure. (Reproduced with permission from Wells and Bedford [33].)
Figure 11.11 Dose-rate effects for density-inhibited cultures of C3H WJ1/2 cells. Subculture and plating for colony formation were delayed for 8-12 h at the end of each exposure to ensure that the same opportunity for full potentially lethal damage repair would occur for all dose increments during the long or short exposures. (Reproduced with permission from Wells and Bedford [33] J
different dose rates for different periods of time to correspond to the various total doses, and the surviving fractions were obtained by dividing the number of viable cells present at the end of each exposure by the number present initially [33]. When extensive cell proliferation occurs during exposure, specification of the dose is a little unclear because cells present at the end of an exposure did not actually exist except as components of the medium at the beginning of the exposure. Nevertheless, we will specify dose as the average dose delivered to the culture as a whole. There is a large dose rate effect due to repair as the dose rate was reduced from 55.6 Gy h-1 to 1.4 Gy tr1, and relatively little proliferation or cell cycle redistribution would have occurred because less than one cell cycle would elapse during even the longest exposure. For a further reduction in dose rate to 0.49 Gy tr1, there appears to be little further reduction in effect per unit dose. In fact, for the higher doses this dose rate is at least as effective, and perhaps more so, as the higher dose rate of 1.42 Gy h"1. There is a clear effect of cell cycling and cell cycle redistribution here. Apparently only a few cells were capable of dividing at this dose rate and were blocked in a sensitive portion of the cell cycle. One complete cell cycle would have elapsed after a dose of 12 Gy at 0.49 Gy h-1, after which the remainder of the doses would have been delivered to (mostly) G2 blocked cells. When the dose rate was reduced even further to 0.29 Gy tr1, the division of some viable cells actually allowed the 'apparent survival' to increase above 1.0 and even to remain as high as 0.7-0.8 for doses up to 15 Gy, corresponding to exposures lasting about 2 days. Then, as proliferation ceased and a G2 block became effective, the
170 The radiobiology of low dose-rate and fractionated irradiation
reduction in survival with dose nearly paralleled the rate of reduction for the higher dose rate of 0.49 Gy rr1. That these principles are not a peculiarity of rodent systems such as the mouse C3H 10T1/2 cells is shown in Figures 11.13 and 11.14. Figure 11.13 illustrates an experiment for multiple-dose fractionation using contact-inhibited normal human fibroblasts [45]. Here, the different total doses were given either as single acute doses or as 4, 2, 1, or 0.5 Gy doses per fraction, with 6 h allowed between fractions. Again, there is a large dose fractionation effect, which corresponds to a factor of nearly three in total dose to reach a given level of effect (in the higher dose regions), and there appears to be a limit or non-repairable component. Figure 11.14 illustrates, for cycling 'transformed' human tumor cells (HeLa), a very pronounced influence of cell cycle redistribution and proliferation during exposures at lower dose rates. In this study [30] as the dose rate was decreased from the baseline high dose rate of 143 cGy min~' (8580 cGy Ir1) down to 154 cGy Ir1, (Figure 11.14, right panel), there was a sizeable reduction in effectiveness per unit dose, but decreasing the dose rate to 74 cGy h"1 did not further reduce the effective-
ness. Further reducing the dose rate to 55 cGy h' and 37 cGy h'1 actually increased the effectiveness. Extensive experimentation with this HeLa human tumor cell system [30,46] has shown that this reversal of the dose rate effect, which more than counteracts the sparing from repair, results from a block in a radiosensitive G2 + M-like state. With reductions in the dose rate to 15 cGy h'1 and 10 cGy tr1, the G2+M block and subsequent redistribution are reduced or disappear and the reduction in effectiveness per unit dose is then governed principally by killing due to the dose-rate-independent, nonrepairable component of damage, along with the occurrence of cell proliferation or the division of viable cells during exposure.
11,10
IS THIS CLINICALLY RELEVANT?
The possibility that these factors may have some importance for radiotherapy has been discussed on numerous occasions. The question generally centers on the dose rates or doses per fraction necessary to produce this G2 block during treatment, and the cell cycle transit times and growth fractions in various tumors. The general assumption is also made that proliferation and cell cycle redistribution during low dose-rate irradiation are not a factor for dose-limiting 'slow turnover' tissues largely responsible for late-effects complications. Although there is no universal agreement on this issue, it is fair to say that, because of the enormous diversity and variation among patients and tumors, the only generalization that is likely to be correct is that any single generalization about all tumors is almost certain to be incorrect.
Image Not Available 11.11 WHAT IS HAPPENING AT THE SUBCELLULAR LEVEL?
Figure 11.13 Survival response of plateau-phase ACT522 cells to X-mys delivered at 0.5 Gy'/fraction (H), 1 Gy/fraction (O), 2 Gy'/fraction (O), and 4 Gy/fraction (A,), as well as by a single dose (O). As the dose per fraction decreases, survival approaches that predicted by the a component (dashed line), which represents the dose-rate-independent or dose-fractionation-independent component of X-ray-induced cell killing. All cultures were allowed 24-h post-irradiation incubation at 37 °C prior to plating for survival. (Reproduced with permission from Bedford and Cornforth [45].j
With our current more detailed knowledge of the workings of biological systems at the subcellular and molecular level, it is worth some discussion of what is a reasonable hypothesis about the mechanism of the dose rate and dose fractionation effect for mammalian cells. Perhaps the most important progress in this area has come from cytogenetics. Considerable evidence has accumulated over the years indicating that gross chromosomal aberrations leading to large genetic losses in the progeny of irradiated cells are the principal cause of cell reproductive death [44,45,47-55]. One of the most convincing sets of experiments on this subject was carried out by Revell and his colleagues [52,53], who showed, for living normal diploid hamster cells, that virtually every cell died where an acentric fragment yielding a micronucleus was present in at least one daughter cell after irradiation of a Gl cell in the previous cycle.
The answer is: chromosome breaks! The question is: what is sublethal damage? 171
Image Not Available
Figure 11.14 Survival curves for log-phase S3 HeLa cells irradiated at different dose rates of j radiation. The right panel illustrates survival curves for continuous irradiation where no cell division occurs during any of the exposures. The left panel illustrates survival curves where in some cases (10 and 15 rod h~1) cell division was occurring during the irradiation. Vertical arrows on the curves indicate the accumulated dose after a time equivalent to one generation time in unirradiated cultures. PE - plating efficiency. (Reproduced with permission from Mitchell el al. [30]. j
Conversely, virtually every irradiated Gl cell that did not yield a micronucleated daughter cell lived. If some other independently produced but unseen lesion (rather than the fragment-producing chromosomal aberration) were actually responsible for cell killing, then some cells with micronucleated daughter cells should have survived, and also some which did not yield micronuclei should have died. This did not happen. Currently, there is no evidence for any proposed killing process or mechanism for any other chemical or physical agent that comes close to being as compelling as the one-to-one correspondence demonstrated by this group for ionizing radiation exposure of mammalian cells. Other experiments Revell and his colleagues carried out on tetraploid and aneuploid cells tended to show that in these cases more than one fragment-producing aberration was necessary to kill a cell [52,53]. Higher ploidy levels apparently result in greater tolerance to chromosome fragment loss, but the basic process or mechanism of killing did not change. Apoptosis is a process of 'programmed cell death.' There is evidence that, for some cell types, particularly those of lymphoblast origin, but for others as well, radiation can trigger a signal transduction pathway leading to apoptosis. In these instances, the contribution of apoptosis to radiation-induced cell death adds appreci-
ably to that resulting from chromosomal aberrations. (This is discussed in more detail below.)
11.12 THE ANSWER IS: CHROMOSOME BREAKS! THE QUESTION IS: WHAT IS SUBLETHAL DAMAGE? If cells are killed by ionizing radiation mainly as a result of acentric fragments generated from chromosomal aberrations, it then requires only a few short steps to explain the dose fractionation and dose rate effect along with the shape of the survival curve for acute high doserate exposures to sparsely ionizing radiations. Figure 11.15 shows a few examples of the kinds of chromosomal aberrations produced by ionizing radiations. It is worth noting that only very few agents, namely those that produce prompt DNA double-strand breaks (dsbs), give this kind of cell-cycle-related pattern of aberration types at the first mitosis after treatment, i.e., chromosome types after GO or Gl treatment and chromatid types after S or G2 treatment. Treatment with agents that produce single-strand and base damage but not DNA dsbs, such as ultraviolet light or alkylating agents,
172 The radiobiology of low dose-rate and fractionated irradiation
Image Not Available
Figure 11.15 Examples of the formation of some chromosome-type and chromatid-type aberrations appearing at the first mitotic metaphase following exposure to ionizing radiation. Chromosome-type asymmetrical exchanges yield acentric fragments including both chromatids of the chromosome or chromosomes. Examples shown in this diagram are the dicentric plus its acentric fragment (asymmetrical interchange) and the interstital deletion (asymmetrical intra-arm intrachange) shown in the light chromosome along with another unaffected (dark) chromosome. Interstitial deletions are mostly acentric rings, but are usually small so the 'hole' in the ring cannot be seen. The other category of asymmetrical exchange is the centric ring plus its acentric fragment (asymmetrical inter-arm intrachange), which is not shown in this diagram. It is seen at much lower frequencies than the aberrations shown. Only two chromosomes (non-homologous) are shown per cell, because that is the minimum required for an interchange. (Reproduced with permission from Bedford [55].)
yield chromatid types even when treatment is in GO or Gl. Chromatid-type aberrations can also kill cells, but more, on average, have to be produced in order that both daughter cells experience the necessary genetic loss. For the analysis below we will, therefore, confine ourselves to the simpler basic case of diploid cells irradiated in GO or Gl, which would apply for most critical normal cells governing normal tissue responses and closely approximate that for most tumor cell populations. Nearly all chromosome-type aberrations that generate acentric fragments are exchange types requiring at least two breaks for their formation, either both in one chromosome or one in each of two different chromosomes. Figure 11.16 shows dose-response curves for the production of acentric-fragment-producing chromosome-type aberrations in human fibroblasts irradiated in a contact-inhibited GO state and then held for full 'PLDR' before subculture and collection for scoring at the first post-irradiation mitosis [44]. Together, interstitial deletions and dicentrics with their associated acentric fragments make up the great majority of total aberrations, at least for higher doses, and the curve relating their increase as a function of dose is not linear.
If each requires the production of two chromosome breaks by the passage of two independent electron tracks, then their frequency would increase as the square of the dose.1 This can be seen as follows. The actual number of total breaks within the cell nucleus (ND) increases linearly with dose. In other words, the number of breaks 1
The formation of so-called 'complex' aberrations, which involve three or more breaks in two or more chromosomes, was at one time thought to occur with negligible frequencies even at relatively high single acute (high dose-rate) doses of 5-10Gy. Recent studies indicate that the fraction of complex aberrations begins to be appreciable for acute X-ray or gamma-ray doses of 3-4 Gy and the relative frequency and degree of complexity increase fairly rapidly with further increases in acute dose. For densely ionizing radiations, the fraction that is complex appears to be very high, even for lower doses. While there is little argument that the relationship between dose and the production of aberrations yielding acentric fragments is well approximated by a linear-quadratic function of dose for sparsely ionizing radiations, there is ample and convincing evidence to argue against the simplified notion that two breaks from two electron tracks account entirely for the curvilinearity often referred to as the dose-squared or quadratic component. Nevertheless, for acute dose fractions of 2 Gy or less and for continuous low dose-rate exposures pertinent to brachytherapy, the simplified picture is perfectly adequate to show the nature of the fractionation and dose-rate effect.
The answer is: chromosome breaks! The question is: what is sublethal damage? 173
Image Not Available
Now, exchanges can also be produced as a result of two breaks occurring close enough together along one electron or other charged particle track and the number of these, along with the proportion actually forming the exchanges in question, would increase in direct proportion to the dose giving a yield Y = gD. Also, breaks that do not rejoin at all (usually a small fraction) produce terminal deletions, adding to the acentric fragment category yield as Y = hD so the total 'single track' production of acentric fragments, Yp increases as Yl = gD + hD or if a = g+h, Yj - aD. Thus, for both two-track and one-track modes of acentric fragment production, the yield (Y) will be Y= aD+f3D2. The total is shown in Figure 11.17 for the sum of the curves (the same cells and conditions) shown in Figure 11.16. In this case, the total yield of lethal fragment-producing aberrations with dose is closely fitted by a curve of the above form where a = 1.5 x 10-' Gy-1 and P = 2.5 x 10"2 Gy~2 [45]. The curve marked 'A-T fibroblasts' is for the very radiosensitive cells of patients with ataxia telangiectasia. This genetic disease is
Figure 11.16 Chromosome-type aberrations per cell in the first mitosis after irradiation of GO human fibroblasts, as a function of dose. Bars around data points indicate 95% confidence limits. (Reproduced with permission from Cornforth and Bedford [44].J
is directly proportional to the number of electron tracks, and the latter is directly proportional to dose. To have an exchange, a second break must occur within some maximum distance, say r, of any given break, i.e., it must occur within the volume 47irY3, which itself occurs within a cell nucleus of radius R. The number of these additional or second breaks occurring within the interaction range would then be approximately (47cr3/47cR3)(ND-l) = (r3.R3)(ND-l). The same consideration must be given to each of the other ND—l breaks and all these N cases summed up to give the total pairs of breaks, Yp, capable of interacting to form an exchange which is (Nu) (A/D-1) (rVtf 3 ) or, if N is fairly large, this will be approximately N^rV-R3). However, as we have already said, the number of chromosome breaks, ND, within the nucleus is directly proportional to dose, so ND = bD and if the ratio rY£3 = a then Y p =a(bD) 2 = jD2 where j=ab2. Because only some constant fraction, f, of break pairs will interact to form an asymmetric fragment-producing exchange (while the others will either form a nonlethal symmetrical exchange or, by far the majority, will restitute), then the yield of two-break asymmetrical exchanges produced by independent electron tracks Y2 will be fjD 2 or, if k = fj, Y2 = kD2. Of course, there is more than one kind of asymmetric exchange and the value of k for each kind will differ, but these constants will also sum, so: Y2 = k,D2+k2D2+ ... knD2=(k,+k2+ ... +kn)D2=$D2 (11.1)
Image Not Available
Figure 11.17 Total chromosome-type aberrations per cell as a function of X-ray dose for normal (ACT 522) and ataxia fibroblasts. Density-inhibited cultures were irradiated and then subcultured 24 h later. Data from the normal cell strains were fitted by a weighted least-squares method to an equation of the form Y = C + aD + pD2 (C = 0.070; a = 7.54 x 70-'; (3 = 2.52 x 70^, where D is the dose in gray. Bars indicate 95% confidence about means. (Reproduced with permission from Cornforth and Bedford [44].,)
174 The radiobiology of low dose-rate and fractionated irradiation
the phenotype of individuals who are homozygous recessive for the A-T gene. A number of other artifically produced hypersensitive mutants are also now available which should greatly help in the understanding of the molecular mechanisms controlling radiosensitivity in general and dose rate effects in particular [57-63]. From the above considerations, it is easy to see how a repair process could operate to reduce the biological effect with dose fractionation or a reduced dose rate. Immediately after a given radiation dose there will be a number of chromosome breaks located too far away from any other break for an exchange to be possible, so the only possibilities would be restitution (repair) or failure to rejoin (lack of repair). Failure to rejoin is relatively infrequent, as seen from the curve for terminal deletions in Figure 11.16. These open lesions or breaks too far from another for rejoining can be considered as sublethal damage and their rejoining or restitution as SLDR. If an additional dose is given immediately after the first, before the isolated breaks have restituted, then they are available for interaction with the additional breaks if such occur near enough. If the additional dose is given at some later time, after the isolated breaks from the previous dose have rejoined, they will not be available and the effectiveness of the additional dose will be lessened. For continuous LDR irradiation, a similar consideration applies. If the dose rate is high so the total dose is delivered in a time that is short compared to the half-time for rejoining of breaks, then the maximum number of break-pairs for potential exchanges under these conditions will be produced. On the other hand, if the dose rate is very low, a second break will never be produced by an independent electron track before the one produced by an earlier track has disappeared by restitution or rejoining. Thus, the '|3 term' in the above equation will approach zero at low dose rates. In this case, the only break-pairs for potential exchange will be those produced simultaneously along the same electron track. This 'a-term' or 'a-component' is dose rate independent.
Y = aD + $D\ but Y is the average number per cell. Each cell does not have exactly Y aberrations. Some have a few more, some have a few less, and it is the ones that have none that are of interest in this case of normal diploid cells irradiated in GO or Gl because it is only these that will survive. The Poisson distribution:
allows us to calculate the proportion of cells that have exactly x events (aberrations) occurring in them when the mean number is |i. Thus, the proportion with exactly zero aberrations or the proportion surviving, S, will be:
In this case, the mean (m = Y, so the surviving fraction
is:
The way in which aberration yield and survival change with dose fractionation and dose rate has been reported and discussed on many occasions, both relatively early in the history of the field by Lea [9] and Neary (64) and also more recently [45,54,55]. For a damage-interaction mechanism in general, by far the most comprehensive treatment of this and other aspects has been given by Kellerer and Rossi [65]. Figures 11.18 and 11.19 illustrate the way in which chromosome aberration frequencies and the fraction of cells surviving change when a total dose of 8 Gy is given for all cells but the delivery is spread out in time by giving one acute dose of 8 Gy, two acute doses of 4 Gy, four acute doses of 2 Gy, eight acute doses of 1 Gy, or 16 acute doses of 0.5 Gy with a time between fractions of 6 h in each case [45]. This illustrates the gradual and parallel loss of the f3 component of aberration production and cell killing.
11.14 DOES POTENTIALLY LETHAL DAMAGE PLAY A ROLE? 11.13 HOW DOES A LINEAR-QUADRATIC INCREASE IN DAMAGE LEAD TO AN EXPONENTIAL DECREASE IN SURVIVAL? The damage killing the cells is being produced in a way that increases both linearly with dose for one component and as the square of the dose for the other component. How does it happen, then, that the fraction of cells surviving decreases as a complex exponential function of dose? This follows because the damage which develops in the form of lethal chromosome aberrations is distributed randomly and independently among cells (of uniform sensitivity) in a way that can be approximately described by a Poisson distribution. The formula for the total yield of the lethal aberrations per cell, Y, is
Earlier another well-known cellular damage repair process referred to as potentially lethal damage repair was discussed. How does this relate to chromosomal aberrations and the dose rate effect? Actually, it is connected in a rather simple (but hypothetical) way under the current framework or hypothesis of aberrations and cell death. PLD can be considered as a break-pair with a certain probability of break interaction to form a lethal aberration. If conditions change, such as post-irradiation holding in the contact-inhibited state versus subculture stimulating cells into the cycle, or changing the osmolarity of incubation medium from isotonic to hypertonic or hypotonic, then the probability of break-pair interaction may change. During incubation under one set of condi-
Changes with linear energy transfer 175
Image Not Available Image Not Available
Figure 11.18 Net frequency of different aberration types scored in AG1522 cells as a function of the number of fractions used to delivers Gy of X-rays. An additional 24-h incubation followed the last fraction. Note that while the frequencies of rings plus dicentrics and interstitial deletions decreased with increasing fraction number, terminal deletions appeared to be independent of dosefractionation. The dashed line traces the reduction in total aberration frequency predicted from a chromosome break rejoining half-time of 1.7 h. Solid lines were drawn by eye. Error bars indicate 95% confidence limits about the mean number of total aberrations per cell. (Reproduced with permission from Bedford and Cornforth [45].;
tions, individual breaks comprising break-pairs (including those produced by two breaks along single tracks, known as the single-hit or a component) can rejoin, misrejoin, or not rejoin at all in a way characteristic of this incubation condition, but if the condition changes, the relative frequencies or probabilities of these may change so that survival is either higher or lower. It may even be that certain conditions, such as those known to change chromatin structure, could cause a change not in actual number of total breaks at risk, but in the number of breaks close enough to another, i.e., break-pairs, for an exchange to occur. In any case, the issue of PLD versus SLD is in no way a dilemma in regard to the dose rate effect or cellular radiobiology in general.
11.15 CHANGES WITH LINEAR ENERGY TRANSFER Another issue worthy of comment is the question of radiation quality and relative biological effectiveness
Figure 11.19 Survival of plateau-phase AG1522 cells as a function of the number of X-ray fractions used to deliver a fixed dose of 8 Gy. Open circles (O) show measured survival as the means of two separate experiments. Crosses (+) were derived from the best-fit survival curves shown in Figure 11.13. Solid squares (•, ±95% confidence about the means) show the survival calculated from the expression S = e~Vi where Yt is the average number of chromosome-type aberrations per cell from Figure 11.18. Cultures received a 6-h incubation at 37°C between each fraction, and a 24-h post-irradiation incubation following the final fraction, before being subcultured and plated for survival or chromosomal analysis. Survival expected after a large number of very small doses is indicated following the break in the abscissa and was derived from the best estimate of a for this experiment. (Reproduced with permission from Bedford and Cornforth [45]J
(RBE) in the current context. The radiation dose delivered by an X-ray or gamma-ray exposure is delivered almost exclusively by the production of ionizations occurring along the tracks of electrons. The ionization density or linear energy transfer (LET) is much lower for such electrons than for heavier charged particles of modest energy such as alpha particles or recoil protons arising from neutron irradiation. Obviously, if the spacing of ionizations along a single electron track is close enough together occasionally to form a break-pair for potential exchange formation, then this should occur at a correspondingly higher frequency if the spacing between ionizations is even closer, as it is for higher LET radiations. Accordingly, this should result in a greatly increased contribution from a dose-rate-independent a component of damage for high LET radiation. In fact, this is the case and there is not only little or no dose rate or dose fractionation effect for radiations such as a particles, but the a component of damage is so greatly increased that the dose-response curve for exchange aberrations is linear and completely overshadows any intertrack P component. Likewise, the dose-survival curve becomes dominated by the a component and is a steeper simple exponential function of dose. It is of interest that with respect to cell killing at high versus low dose rates, certain X-ray-sensitive mutants behave for
176 The radiobiology of low dose-rate and fractionated irradiation
Image Not Available
Figure 11.20 The survival responses of plateau phase CHO 10B2 (open symbols) and irs-20 (dosed symbols) cells exposed at various dose rates. Cells were incubated for 6-12 h before replating to determine survival. O, •, 0.75 Gym/Tr'/D, •, 0.50 Gy/r'; A, A, 0.12 Gy h~1; 4, 0.06 Gy h~\ (Reproduced with permission from Stackhouse and Bedford [56].,)
low LET radiation in a way that is similar to that for their wild-type counterparts with high LET radiations. This is illustrated in Figure 11.20, where a large gamma-ray dose rate effect is seen for plateau phase CHO wild-type cells but there is no dose rate effect for the CHO irs-20 mutant cells [57]. High LET data are not available for CHO irs-20 cells.
11.16 MORE ON APOPTOSIS AND SIGNAL TRANSDUCTION While there are still many other issues and controversies concerning the mechanisms of cell killing, and even chromosomal aberration induction itself, the above treatment is, to a first approximation, accurate enough to account for most of the observations of importance for dose rate and dose fractionation effects. As mentioned earlier, one particularly interesting subject that has recently moved into prominence for biology in general, and naturally into radiation-induced cell killing, is the
process of apoptosis. This process of programmed cell death appears to be induced to a greater or lesser degree in some cells by the triggering of a signal transduction pathway by ionizing radiation. Regarding radiotherapy, cells of lymphocytic origin appear more radiosensitive to this process of killing than the much reduced contribution to killing in other cells. Currently, however, the contributions of apoptosis to radiation-induced cell killing have not been carefully assessed in a quantitative way, as has been done for aberration induction, but these may not be altogether unrelated processes for some cells. For example, apoptosis need not always occur before the first post-irradiation mitosis. The fact that cells can be killed by the loss of a large acentric chromosome fragment after mitosis does not mean that the death such cells eventually experience cannot be apoptotic. There are several quantitative issues that need to be addressed and resolved to assess the role of signal transduction and apoptosis triggered by ionizing radiation, and its relevance in regard to dose fractionation and brachytherapy. One important question is whether radiation triggers signal transduction in all cells but to a greater or lesser degree depending on dose, or whether the fraction of cells triggered or not triggered changes with dose in a way that corresponds with the fraction killed or not killed, respectively. Recently, the kinds of data necessary are beginning to emerge. For example, R. Mikkelsen and R. Schmidt-Ullrich and their colleagues are utilizing fluorescent reporter molecules to register the triggering of signal transduction pathways in individual cells, and are finding that in some cases the fraction of cells triggered and not the degree of transduction increases with radiation dose in the 0-10 Gy range [69]. Another important question is the time course of damage development. Of major concern for brachytherapy (or teletherapy) are the late effects in dose-limiting normal tissues. In tissues that turn over very slowly, late effects can be understood, at least to some degree, if they result from a mitotic-linked death from lethal chromosomal aberrations. Similar reconciliation with long-delayed signal transduction pathways and apoptosis are needed to better assess their role in the late tissue effects that essentially define the limits of radiation treatment.
11.17 SELECTION OF MUTANTS BY LOW DOSE RATE Low dose-rate irradiation turns out to be a very effective way to isolate repair-defective (or misrepair-proficient) cells and also to map and isolate the gene(s) involved. For example, irs-20 cells, a mutant CHO line, complete only two or three divisions during irradiation at 6 cGy h"1, a dose rate that has no effect on the growth of wildtype CHO cells [56]. The defect in these mutant cells can be corrected by fusion with a normal human cell and
References 177
repeated selection at LDR. This allows a FISH mapping of the gene to human chromosome 8qll, which is the location of the human homolog of the SCID gene [66]. The exact nature of the mutation, a substitution of lysine for glutamic acid, has been determined [67,68]. Thus, LDR irradiation can be a powerful tool in gene discovery.
REFERENCES
1. Del Regato, J.A. and Regaud, C. (1993) In Radiological Oncologists: the Unfolding of a Medical Specialty. Reston VA, Radiology Centennial Inc., 53-63.
damage in surviving Chinese hamster cells. Radial Res., 13, 556-93. 17. Terasima, T. and Tolmach, L.J. (1961) Changes in x-ray sensitivity of HeLa cells during the division cycle. Nature, 190,1210-11. 18. Terasima, T. and Tolmach, L.J. (1963) Variations in several responses of HeLa cells to X-irradiation during the division cycle. Biophys.J., 3,11-33. 19. Dewey, W.C. and Humphrey, R.M. (1963) Survival of mammalian cells irradiated in different phases of the life cycle as examined byautoradiography. Nature, 198, 1063-6. 20. Sinclair, W.K. (1968) Cyclic X-ray responses in mammalian cells in vitro. Radial Res., 33,620-43. 21. Bedford, J.S. and Mitchell, J.B. (1973) Dose rate effects in synchronous mammalian cells in culture. Radial Res., 54,
2. Regaud, C. (1922) Influence de la duree d'irradiation sur
316-27.
les effects determines dan le testicule par le radium. Compt. Rend. Soc. Biol., 86,787-90.
22.
3. Committee on the Biological Effects of Ionizing Radiation (1990) Health Effects of Exposure to Low Levels of Ionizing Radiation. Washington, DC, National Academy of
23.
Sciences/National Research Council. 4. Russell, W. (1965) Studies in mammalian radiation genetics. Nucleonics, 23, 53-62. 5. Ullrich, R.L, Jernigan, M.C., Satterfield, L.C. and Bowles, N.D. (1987) Radiation carcinogenesis: time-dose relationships. Radial Res. 111,179-84. 6. Sax, K. (1938) Chromosome aberrations induced by x-rays. G<wr/cs,23,494-516. 7. Lea, D.E. and Catcheside, D.G. (1942) The mechanism of the induction by radiation of chromosome aberration in Tradescantia. J. Genet., 44,216-45. 8. Marinelli, L.D., Nebel, B.R., Giles, N.H. and Charles, D.R. (1942) Chromosomal effects of low x-ray doses on five-day Tradescantia. Am.). Bot., 29,866-74. 9. Lea, D.E. (1955) Actions of Radiations on Living Cells, 2nd edn. Cambridge, UK, Cambridge University Press. 10. Kelner, A. (1949) Effect of visible light on the recovery of Streptomycesgrieusconidia from ultraviolet irradiation injury. Proc. NatlAcad. Sci. USA, 35,73-9. 11. Dulbecco, R. (1949) Reactivation of ultraviolet inactivated bacteriophage by visible light. Nature, 163,949-50. 12. Rupert, C.S., Goodgal, S.H. and Herriott, R.M. (1958) Photoreactivation in vitro of ultraviolet inactivated Hemophilus influenzae transforming factor./ Gen. P/7ys/o/.,41,451-71. 13. Rupert, C.S. (1960) Photoreactivation of transforming DMA by an enzyme from baker's yeast./ Gen. Physiol.,43, 573-87. 14. Setlow, R.B. and Carrier, W.L (1964) The disappearance of thymine dimers from DMA: an error-correcting mechanism. Proc. NatlAcad. Sci. USA, 51,226-31. 15. Boyce, R.P. and Flanders, H. (1964) Release of ultraviolet
Mitchell, J.B., Bedford, J.S. and Bailey, S.M. (1979) Doserate effects on the cell cycle and survival of S3 HeLa and V79 cells. Radial Res., 79, 520-36. Dewey, W.C., Stone, LE., Miller, H.J. and Giblak, R.E. (1971) Radiosensitization with 5' bromodeoxyuridine of Chinese hamster cells X-irradiated during different
phases of the cell cycle. Radial Res., 47,672-88. 24. Griffiths, T.D. and Tolmach, L.J. (1976) Lethal response of HeLa cells to X-irradiation in the latter part of the generation cycle. Biophys.J., 16,303-18. 25. Dewey, W.C. and Highfield, D.P. (1976) G2 block in Chinese hamster cells induced by X-irradiation, hyperthermia, cyclohexamide, oractinomycin-D. Radial Res., 65, 511-28. 26. Schneiderman, M.H., Dewey, W.C., Leeper, D.B. and Nagasawa, H. (1972) Use of the mitotic selection procedure for cell cycle analysis: comparison between the X-ray and G2 markers. Exp. Cell Res., 74,430-8. 27. Canti, R.G. and Spear, F.G. (1929) Effect of gamma-rays on mitosis in vitro. Proc. R. Soc. B, 105,93-8. 28. Yamada, M.and Puck, T.T. (1961) Action of radiation on mammalian cells. IV. Reversible mitotic lag in the S3 HeLa cell produced by low doses of X-rays. Proc. NatlAcad. Sci. L/S/4,47,1181-91. 29. Doida, Y. and Okada, S. (1969) Radiation-induced mitotic delay in cultured mammalian cells (L5178Y). Radial Res., 38,513-29. 30. Mitchell, J.B., Bedford, J.S. and Bailey, S.M. (1979) Doserate effects in mammalian cells in culture. III. Comparison of cell killing and cell proliferation during continuous irradiation for six different cell lines. Radial /tes.,79,537-51. 31. Little, J.B. (1968) Delayed initiation of DMA synthesis in irradiated human diploid cells. Nature (Lond), 218, 1064-5. 32. Nagasawa, H.and Little J.B. (1983) Comparisons of
light-induced thymine dimers from DMA in E. coli. Proc.
kinetics of X-ray induced cell killing in normal, ataxia
NatlAcad. Sci. USA, 51,293-300.
telangiectasia and hereditary retinoblastoma fibroblasts.
16. Elkind, M.M. and Sutton, H. (1960) Radiation response of mammalian cells grown in culture. I. Repair of x-ray
Mutal tes.,109,297-308. 33. Wells, R.L. and Bedford, J.S. (1983) Dose-rate effects in
178 The radiobiology of low dose-rate and fractionated irradiation
mammalian cells. IV. Repairable and non-repairable damage in non-cycling C3H 10T1/2 cells. Radial Res., 94, 105-34. 34. Zeman, E.M.and Bedford, J.S. (1985) Dose fractionation effects in plateau-phase cultures of C3H 10T1/2cellsand their transformed counterparts. Radial Res., 101, 373-93. 35. Kastan, M.B., Onyekwere, 0., Sidransky, D., Vogelstein, B. and Craig, R.W. (1991) Participation of p53 protein in the cellular response to DNA damage. Cancer Res., 51, 6304-11. 36. Elkind, M.M., Sutton-Gilbert, H., Moses, W.B., Alescio, T. and Swain, W. (1965) Radiation response of mammalian cells grown in culture. V. Temperature dependence of the repair of X-ray damage in surviving cells (aerobic and hypoxic). Radial Res., 25,359-76. 37. Little, J.B. (1969) Repair of sub-lethal and potentially lethal radiation damage in plateau phase cultures of human cells. Nature (Lond), 224,804-6. 38. Phillips, R.A. and Tolmach, LJ. (1966) Repair of
mammalian cells: emphasis on repair. Proc. NatlAcad. Sci. USA, 68,667-71. 49. Bedford, J.S., Mitchell, J.B., Griggs, H.G. and Bender, M.A. (1978) Radiation-induced cellular reproductive death and chromosome aberrations. Radial Res., 76, 573-86. 50. Davies, D.R. (1963) Radiation-induced chromosome aberrations and loss of reproductive integrity in Tradescantia. Radial Res.,2Q, 726-40. 51. Carrano, A.V. (1973) Chromosome aberrations and radiation-induced cell death. II. Predicted and observed cell survival. Mutal Res., 17,355-66. 52. Revel I, S.H. (1983) Relationship between chromosome damage and cell death. In Radiation-Induced Chromosome Damage in Man, ed. T. Ishihara and M.S. Sasaki. New York, Liss, 215-33. 53. Joshi, G.P., Nelson, W.V., Revell, S.H. and Shaw, C.A. (1982)
potentially lethal damage in X-irradiated HeLa cells.
X-ray-induced chromosome damage in live mammalian
Radial Res. ,29,413-32.
cells, and improved measurements of its effects on their colony forming ability. InlJ, Radial Biol., 41,161-81.
39. Belli, J.A. and Shelton, M. (1969) Potentially lethal radiation damage: repair of mammalian cells in culture. Science, 1654,490-2. 40. Dettor, C.M., Dewey, W.C., Winans, LF. and Noel, J.S. (1972) Enhancement of X-ray damage in synchronous Chinese hamster cells by hypertonic treatments. Radial 41.
human cells in vitro. Proc. NatlAcad. Sci. USA, 44, 772-80. 48. Dewey, W.C., Miller, H.H. and Leeper, D.B. (1971) Chromosomal aberrations and mortality of X-irradiated
Res., 52,352-72. Raaphorst, G.P. and Kruuv, J. (1977) Effect of salt solutions on radiosensitivityof mammalian cells. III. Treatment with hypertonic solutions. InlJ. Radial Biol., 32,
109-26. 42. Utsumi, H., Hill, C.K., Ben-Hur, E. and Elkind, M.M. (1981) 'Single-hit' potentially lethal damage: evidence of its repair in mammalian cells. Radial Res., 87, 576-91. 43. Stapleton, G.E., Billen, D. and Hollander, A. (1953) Recovery of X-irradiated bacteria at suboptimal incubation temperatures./ CellComp. Physiol., 41, 345-57. 44. Cornforth, M.N. and Bedford, J.S. (1987) A quantitative
54. Cornforth, M.N. and Bedford, J.S. (1993) Ionizing radiation damage and its early development in chromosomes. In Advances in Radiation Biology, 17, DNA and Chromatin Damage Caused by Radiation, ed. J.T. Lett and W.K. Sinclair. San Diego, Academic Press, 423-96. 55. Bedford, J.S. (1991) Sublethal damage, potentially lethal damage, and chromosomal aberrations in mammalian cells exposed to ionizing radiations. InlJ. Radial Oncol. Biol.Phys., 21,1457-9. 56. Stackhouse, M.A. and Bedford, J.S. (1993) An ionizing radiation sensitive mutant of CHO cells: irs-20.1. Isolation and initial characterization. Radial Res. 136,241-9. 57. Stackhouse, M.A. and Bedford, J.S. (1993) An ionizing radiation sensitive mutant of CHO cells: irs-20 cells. II. Dose-rate effects and cellular recovery process. Radial Res., 136,250-4. 58. Biedermann, K.A., Sung, J., Giaccia, A.J., Tosto, L.M. and
comparison of potentially lethal damage repair and the
Brown, J.M. (1991) Scid mutation in mice confers
rejoining of interphase chromosome breaks in low
hypersensitivity to ionizing radiation and a deficiency in
passage normal human fibroblast. Radial Res., 111,
DNA double-strand break repair. Proc. NatlAcad. Sci. USA, 88,1394-7.
385-405. 45. Bedford, J.S. and Cornforth, M.N. (1987) Relationship between the recovery from sublethal X-ray damage and the rejoining of chromosome breaks in normal human
59. Stamato, T.D., Weinstein, R., Giaccia, A. and Mackenzie, L. (1983) Isolation of cell cycle-dependent gamma raysensitive Chinese hamster ovary cell. Somal Cell Genet., 9,
fibroblasts. Radial Res., 111,406-23. 46. Bedford, J.S., Mitchell, J.B. and Fox, M.A. (1980) Variations
165-73. 60. Jeggo, P.A. and Kemp, L.M. (1983) X-ray-sensitive mutants
in responses of several mammalian cell lines to low dose irradiation. In Radiation Biology in Cancer Research, ed. R.E. Meyn and H. Rodney Withers. New York, Raven Press, 251-62. 47. Puck,T.T. (1958) Action of radiation on mammalian cells. III. Relationship between productive death and induction of chromosome anomalies by X-irradiation of euploid
of Chinese hamster ovary cell line. Isolation and crosssensitivity to other DNA-damaging agents. Mutal Res., 112,313-27. 61. Whitmore, G.F., Varghese, A.J. and Gulyas, S. (1989) Cell cycle responses of two X-ray sensitive mutants defective in DNA repair. InlJ. Radial Biol., 56,657-65. 62. Thacker, J. and Wilkinson, R.E. (1991) The genetic basis of
References 179
63.
64.
65. 66.
resistance to ionizing radiation damage in cultured mammalian cells. Mutat. Res., 254,135-42. Zdzienicka, M.Z., van Wessel, N. and van der Schans, G.P. (1992) Afourth complementation group among ionizing radiation-sensitive Chinese hamster cell mutants defective in DMA double-strand break repair. Radial Res., 131,309-14. Neary, G.J. (1965) Chromosome aberrations and the theory of RBE. I. General considerations. Int.J. Radiat. fi/o/.,9,477-502. Kellerer, A.M. and Rossi, H.H. (1972) The theory of dual radiation action. Curr. Top. Radiat. Res. Q., 8,85-158. Lin, J.Y-D and Bedford, J.S. (1997) Regional gene mapping
using mixed radiation hybrids and reverse chromosome painting. Radiat. Res., 148,405-12. 67. Lin, J.Y-D., Muhlmann-Diaz, M.C., Stackhouse, M.A. etal. (1997) An ionizing radiation-sensitive CHO mutant cell line: irs-20. IV. Genetic complementation, V(D)J recombination and thesdd phenotype. Radiat. Res., 147, 166-71. 68. Priestley, A., Beamish, H.J., Gell, D. etal. (1998) Molecular and biochemical characterisation of DMA-dependent protein kinase-defective rodent mutant irs-20. Nucl. Adds Res., 26(8), 1965-73. 69. Mikkelsen, R.B. (2000) Personal communication.
12 Dose-rate effects with human cells G. GORDON STEELANDJOHN H. PEACOCK
The term 'dose-rate effect' refers to the change in sensitivity or tissue response when the dose rate of irradiation is modified. Dose-rate effects are common in mammalian cell systems, including human tumors and normal tissues. The response of these tissues is complex, depending in part on the radiosensitivity of the stem cells (or 'clonogenic' cells) of the tissue, but also on the modifying effects of cell proliferation and such physiological parameters as oxygenation and growth factors. Whereas in rodent cell systems it is possible to compare effects seen in vivo with effects on cells brought into tissue culture, such a comparison is difficult for human tissues. There is a body of relevant clinical experience, but this does not provide the necessary detail and precision to allow theories to be tested in the way that can be done in experimental laboratory systems. As a result, attention
is restricted in this chapter to dose-rate effects seen in human cells studied in tissue culture. Although such cell systems are necessarily artificial, it can be argued that they do allow the major components of the dose-rate effect to be demonstrated.
12.1
THE TIME-SCALE OF RADIATION ACTION
The time-scale of biological effects of ionizing radiation is illustrated in Figure 12.1. It is the operation of some of the processes represented in this chart that gives rise to dose-rate effects. Immediately after exposure, free-radical processes take place leading to damage to many constituents of the cell. Because of its vital nature and the
Figure 12.1
Time-scale of
the effects of radiation exposure on biological systems. (From Steel [18].J
Mechanisms of the dose-rate effect 181
relative uniqueness of its genetic message, DNA is the most important of these damaged molecules. Under physiological conditions the rapid free-radical reactions are complete within around 1 ms. During the subsequent few minutes, enzymatic processes begin to operate on the damaged molecules. Some of these act to repair the damage; others leave the molecules in a changed but stable form and we describe this as 'misrepair.' Within a few hours, these enzymatic processes will be complete. Repair of radiation damage to DNA is highly effective in most cell types: a 1 Gy dose will induce upwards of 1000 DNA strand breaks in every irradiated cell. Roughly half of the cells will survive this dose, so strand-break rejoining must be a remarkably error-free process. Most strand breaks are to one strand only of the double helix, but a small proportion can be recognized as affecting both DNA strands (double-strand breaks, dsb). There is evidence that these are much more serious for the viability of the cell. Even so, the great majority of dsb are also successfully repaired. Of particular importance are dsb that arise from clusters of ionizations at the end of the tracks of secondary electrons: these can involve severe damage to the DNA molecule (so-called 'multiply damaged sites') and it may be that these events have a relatively low probability of successful repair and a correspondingly high likelihood of leading to cell death or mutation. At longer intervals after irradiation (Figure 12.1), cell proliferation will take place within tissues, leading to the replacement of radiation-damaged cells. In tumors this may lead to recurrence or to a reduced likelihood of success as a result of subsequent treatment. In normal tissues, proliferation may prevent tissue breakdown and the observed early effects of irradiation will then be minimal. However, if the level of cell killing is greater and of such a severity that it cannot be counteracted by proliferation, then serious tissue damage may appear. At even longer time intervals after irradiation (months to years), the very long-term effects will become apparent, including tissue failure, formation of new tumors, and mutational effects in germ cells.
12.2 MECHANISMS OF THE DOSE-RATE EFFECT Observed dose-rate effects derive from the operation of the processes just described. Clinical external-beam treatments are usually given within a few minutes. These brief exposures are long enough for the initial chemical effects of irradiation to be complete, but are too short for the subsequent enzymatic and proliferation processes to take place. As radiation dose rate is lowered, the irradiation time for a given dose increases and it becomes possible for such processes to take place during radiation exposure. These will modify the extent of damage and thus lead to a dose-rate effect. Four main processes lead in this way to the dose-rate effect. They are the '4Rs of radiobiology': repair, redistribution, repopulation, and reoxygenation. Among these, repair is the fastest. As indicated below, the time required to repair half the induced damage is often in the region of 1 h. This means that as soon as the duration of exposure becomes a significant fraction of an hour, some repair will take place during irradiation. At the other extreme, repopulation is a much slower process: repopulation requires cell multiplication and human cells cannot divide in less than about a day. Repopulation will therefore only have a significant effect when the exposure time is a day or more. Redistribution and reoxygenation probably have a speed that is intermediate between these two processes. Figure 12.2 illustrates the range of dose rates over which each of these processes might be expected to influence radiation action. For dose rates in excess of a few gray per minute, none of the processes will take place significantly during irradiation and there will be no dose-rate effect due to them. At much higher dose rates than illustrated a further process, the consumption of oxygen by radiochemical reactions leading to partial hypoxia, may have an effect [1]. At dose rates around 1 Gy min-1, sometimes used for 'high dose rate' or 'acute' irradiations, there may be a small amount of repair
Figure 12.2 The range of dose rates over which repair, reassortment, and repopulation may influence radiation effects.
182 Dose-rate effects with human cells
during irradiation and (as illustrated in the figure) such treatments will be slightly less effective than if given at a higher dose rate. Even in this high dose-rate region there may therefore be significant implications for practical radiation therapy and care must be taken to make suitable corrections for a change in dose rate. Note that the effect of repair increases smoothly over roughly two decades of dose rate (from around 100 to 1 cGy min"1). In the middle of this range the effect is changing most rapidly with dose rate. This has important implications for high dose-rate brachytherapy: when such dose rates are used the biological effects may depend rather critically on the precise dose rate employed in a particular study. The lines drawn in Figure 12.2 to represent the effects of repopulation or reassortment are diagrammatic. Repopulation is a much slower process than repair and only when the exposure time becomes a significant proportion of a cell cycle time (perhaps 1-4 days in human tumor and normal tissue cells) will it have a significant effect during the period of irradiation. Reassortment (otherwise known as redistribution) refers to effects that derive from the movement of surviving cells through the cell cycle after a first dose or dose increment of radiation. These effects may modify the response of a tissue or cell system to subsequent irradiation and occur over a doserate range that is somewhere intermediate between those of repair and repopulation. The comparative effects of repair and repopulation are further illustrated in Figure 12.3. This figure shows actual calculations for a typical human cell line, based on a repair half-time of 0.85 h and an oc/|3 ratio of 3.7 Gy. Curves are drawn for four different cell population doubling times and the calculations show the radiation doses (i.e., ED50
100
Figure 12.3 In human cell systems proliferation probably affects radiation response for dose rates below about 1 Gy h'1 (or a few cGy min'1). For a typical human tumor cell line the figure shows isoeffect curves in the absence of proliferation or with proliferation at a doubling time of 5, 10, 20, or 50 days. The calculations are based on a simple model of exponential growth, ignoring radiation effects on the rate of cell proliferation.
values) for a survival of 0.01. For these parameter values there is no effect of proliferation at dose rates above 1 cGy min"1, but as dose rate is lowered to 0.01 cGy min'1 dramatic effects are predicted, depending on the cell population doubling time. The implication for brachytherapy is that above 1 cGy min'1 repopulation effects can be ignored, but below this dose rate they can, under some circumstances, predominate over effects due to incomplete repair. These calculations also show that in proliferating cell systems it is not possible to find a dose rate at which repair is complete but no proliferation occurs: there is significant overlap between these two processes.
123 DOSE-RATE EFFECTS IN HUMAN TUMOR CELLS Pioneering experimental studies of the dose-rate effect were made in a number of publications by Hall, Bedford and Mitchell (see reviews by Steel et al. [2,3]). The experiments were performed on a variety of cell lines, mainly derived from experimental animals but also including the long-established HeLa cell line (derived from a human cervix carcinoma). They showed that the doserate effect mainly appeared over the range of dose rates from 1 Gy min"1 down to 0.1 cGy min"1. There was considerable variation in the magnitude of the dose-rate effect (i.e., the relative radiosensitivities at high and low dose rates). Steel et al. [2] analyzed these data and showed that derived values for the half-time for repair of radiation damage ranged widely: from below 0.1 h to greater than 1 h. Studies on human tumor cell lines taken from a variety of tumor types were reported by Steel et al. [4]. Most of the cell lines were newly established. In some cases the cells were taken directly from human tumors that had first been grown as xenografts in immune-deficient mice; other studies were made on cell lines established in tissue culture. They were irradiated with cobalt-60 gamma-radiation at dose rates ranging from 1 to 150 cGy min"1 at body temperature and under conditions of controlled oxygenation. Cell survival was measured using a colony assay, either in soft agar or in monolayer, depending on the growth characteristics of the cell line. Data on four cell lines are shown in Figure 12.4, covering the range of responses seen in a larger group of human tumor cell lines. Figure 12.4a shows results at high dose rate. The data are fitted by a linearquadratic equation; there is a well-defined initial slope to the data, which are clearly consistent with a continuously bending relationship. The range of sensitivities is considerable. The doses required for a survival of 0.01 range from 3.6 Gy in the HX142 neuroblastoma to 10.9 Gy in the RT112 bladder carcinoma (i.e., by a factor of 3). In the initial dose region the factor is greater. Figure 12.4b shows the results for the same cell lines at the low dose
Models of the dose-rate effect 183
icr Figure 12.4 Cell survival curves for four human tumor cell lines irradiated at (a) 150 cGy min-1 or (b) 7.6 cGy min~\ HX142, neuroblastoma; HX58, pancreas carcinoma; HX156, cervix carcinoma; RT112, bladder carcinoma. (From Steel [18].j
rate of 1.6 cGy min '. The curves have fanned-out and become straight or almost so on the semi-logarithmic plot. It can be seen that at low dose rate the lines seem to extrapolate the initial slopes of the high dose-rate curves. The range of sensitivities among the cell lines is now larger: by a factor of approximately 10. The data shown in Figure 12.4 indicate the range of sensitivities seen among tumors of different histological types. Less information is available about the range of sensitivities among tumors of the same type, from different patients. Kelland and Steel [5] studied five cell lines newly established from human cervical carcinomas. They found that at high dose rate the dose to produce a surviving fraction of 0.01 ranged from 5 to 10.5 Gy. The dose-rate sparing factors (the dose at 1.6 cGy min"1 compared with the dose at 150 cGy min"1) ranged from 1.1 to 1.6. This showed that among tumors of the same type there were considerable radiobiological differences that could be clinically significant. There may be a number of causes of failure in brachytherapy and these include the inherent insensitivity of the tumor cells to radiation. A so-far insufficiently explored aspect of brachytherapy is the attempt to develop predictive tests of radiosensitivity in order to identify patients most at risk of recurrence. The data in Figure 12.4 clearly indicate that such tests should be made at low dose rate, where the differences
among cell lines are greatest. Bjork-Eriksson et al. [6] reported that dose-rate experiments were useful to discriminate between radiosensitive and radioresistant tumors. This confirmed the earlier work of Amdur and Bedford [7].
12.4
MODELS OF THE DOSE-RATE EFFECT
Theoretical attempts to simulate the dose-rate effect were made by Oliver [8], by Szechter and Schwartz [9], and by Dale and Jones [10]. Thames [11] extended the work of Oliver to the linear-quadratic model and developed what he called the Incomplete Repair (IR) model. He formulated this not only for the case of continuous low doserate exposure but also for fractionated irradiation with fractions too close together to allow full repair. The model is empirical in the sense that it is not based on underlying mechanistic assumptions. It has been very successful in fitting a wide range of radiobiological data, both in experimental and clinical studies. Curtis [12], based on earlier work of Pohlit and coworkers, developed the Lethal-Potentially Lethal damage (LPL) model. This is conceptually different from the IR model. It is a mechanistic model which assumes that radiation induces two
184 Dose-rate effects with human cells
types of intracellular lesion: lethal lesions (L) and potentially lethal lesions (P). The L lesions are non-recoverable. The P lesions may be repaired but they may also be 'fixed', leading to cell death. The probability of fixation is assumed to occur by binary interaction between P lesions. At low dose rate the probability of such interactions will be low (because the lesions are produced far apart in time and will be repaired). This model thus predicts that the slope of the low dose-rate curves shown in Figure 12.4B is due to the sensitivity of the cells to the production of L-type lesions. This is a mechanistically interesting hypothesis which is consistent with current ideas about the production of 'clustered lesions' by low linear energy transfer (LET) radiation (see reference 13). Although the IR and LPL models have very different starting points, they work equally well in most situations. Thames [11] showed mathematically that they are equivalent for cell survival down to about 0.01. Data on human tumor cells are well fitted by both these models [4]. Half-times for repair of radiation damage were deduced to range from 0.4 to 2.3 h. An example is shown in Figure 12.5. Human melanoma cells (HX34) were irradiated at dose rates ranging from 1.6 to 150 cGy min-1. Figure 12.5a shows the survival curves at the extreme dose rates. These data, together with data at intermediate dose rates, have been simultaneously fitted by the IR model giving the following parameter values: a - 0.27 Gy-1, P = 0.046 Gy2, half-time - 0.92 h. The model allows an isoeffect curve to be drawn as a function of dose rate (Figure 12.5b). This is a sigmoid curve of the type that has been reproduced in Figures 12.2 and 12.3. The steepest change of sensitivity with dose rate occurs at around 10 cGy min'1 (if the half-time were longer, the curve would lie to the left, and if it were shorter, the curve would lie to the right of the present one). Current models suggest that, in the absence of proliferation, cellular radiosensitivity at low dose rate approaches a limiting value, indicated by the straight line in Figure 12.5a. The steepness of this line is crucially important for the success of low dose-rate brachytherapy. There is considerable research interest at the present time in the determinants of low dose-rate radiosensitivity. According to the LPL model, this depends on the probability of induction of non-repairable lesions in cells. Such lesions could arise by a variety of processes, including ion cluster effects, as mentioned above, or recombinational effects during the processing of DNA damage. Values of radiosensitivity parameters for some human tumor cell lines are given in Table 12.1. These include the surviving fraction for 2 Gy given at high dose rate (SF2) and the sensitivity of the cells at a low dose rate of 1-2 cGy min"1 (the slope of the linear survival curve is given by the equivalent a value, Gy1). For tumor cell lines in which detailed studies have been made of the dose-rate effect [4], the parameters of the LPL model are also given.
Figure 12.5 The dose-rate effect in a human melanoma cell line, HX34. (a) Cell survival curves at high dose rate (150 cGy min'1) and low dose rate (1 cGy min'1). (b) Isoeffect curve for a survival of 0.01 as a function of dose rate. The full line is calculated using the LPL model.
12.5 EFFECT OF IRRADIATION ON CELL-CYCLE PROGRESSION Irradiation at high dose rate blocks cell entry into mitosis. The cell cycle may be interrupted at a number of socalled 'check-points', and the biochemical processes involved in these arrests are the subject of intense laboratory research at the present time. At high dose rate there are two reasons why proliferation effects during irradiation are unimportant: irradiation times are too short, and the cells are subject to mitotic delay and therefore inhibited from proliferating. As dose rate is reduced, both these factors become less severe and cell cycling
Cell killing around an implanted radiation source 185 Table 12.1 Radiosensitivity parameters* for a number of human tumor cell lines
HX34 HX118 HX58 HX156 HX147 HX148 HX149 MGH-U1 RT112 GCT27 HX138 HX142 HC12 IN859 IN1265 SB
Melanoma Melanoma Pancreas carcinoma Cervix carcinoma NSCLC NSCLC SCLC Bladder carcinoma Bladder carcinoma Teratoma Neuroblastoma Neuroblastoma SCLC Glioma Glioma Glioma
0.47 0.43 0.25 0.59 0.82 0.55 0.26 0.57 0.73 0.40 0.11 0.13 0.39 0.57 0.73 0.47
0.30
0.27
0.046
0.092
0.35 0.53 0.32 0.20 0.07 0.38 0.26 0.19 0.50 0.84 0.98 0.47 0.22 0.30 0.15
0.33 0.47 0.30
0.038 0.052 0.022
0.23 0.80 0.54
0.26 0.12 0.46
0.17 0.027 0.036
0.80 0.85 0.34
0.48
0.17
0.54
0.17 0.28 0.15
0.055 0.027 0.034
0.49 0.58 0.97
* SF2 is the surviving fraction for a dose of 2 Gy at high dose rate. LDR sens, indicates the slope of the survival curve at 1-2 cGy min"1 (Gy1). a (Gy1), P (Gy2) and Ti (h) are parameter values of the Lethal-Potentially Lethal model, determined from low dose-rate experiments. NSCLC = non-small-cell lung cancer; SCLC = small-cell lung cancer.
takes place during irradiation, thus counteracting the effect of irradiation. At what dose rate does this happen? Hall [1] concluded that HeLa cells did not proliferate above 0.2 cGy min'1. Mitchell et al. [14] estimated the critical dose rate for cell proliferation in six mammalian cell lines: this ranged widely, from 0.1 cGy min~' in a lymphoma cell line to over 4 cGy mirr1 in pig kidney and V79 Chinese hamster cells. A study by Skladowski et al. [15] illustrates the complexity of the concept of a dose-rate effect on cell proliferation. Human bladder carcinoma cells (MGH-U1) were irradiated at dose rates ranging from 0.1 to 1.4 cGy min"1. The mitotic index was scored at frequent intervals and flow-cytometry was used to quantify the number of cells in the Gl, S, and G2 phases of the cell cycle. Mitotic activity was immediately suppressed, to a greater degree at the higher dose rates, but this recovered after irradiation had continued for about 24 h and, surprisingly, the time of this recovery was independent of dose rate. Thereafter, the mitotic index overshot to above the normal level and at the higher dose rates there was evidence of synchrony. So proliferation in these cells, though initially disturbed, did continue when the dose rate was as high as 1.4 cGy min'1. Cell survival data were also available for this cell line as a function of dose rate. It was possible to calculate that irradiation for 24 h would lead to a surviving fraction of approximately 0.2 at 0.4 cGy min"1, 0.1 at 0.7 cGy min 1 , 0.03 at 1 cGy min"1, and 0.008 at 1.4 cGy min-1. So the irradiation just during the initial 24 h of mitotic arrest led to sterilization of more than 90% of the clonogenic
cells at dose rates above 0.7 cGy min-1. The flowcytometric data on cell-cycle progression may therefore be misleading, for, at the higher dose rates, most of the cells observed by this method may be dead from a colony-forming point of view. It is impossible by such methods to learn about the cycling of the surviving colony-forming cells on which tumor recurrence will depend. Skladowski et al. [15] concluded that cell-cycle effects in tumor cells are unlikely to be of any great significance in relation to the cell-killing effect at different distances from an implanted radiation source. Overall treatment times in brachytherapy tend to be short compared with external-beam treatment and proliferation effects are correspondingly of less significance.
12.6 CELL KILLING AROUND AN IMPLANTED RADIATION SOURCE The non-uniformity of radiation field around an implanted source has important radiobiological consequences. Close to the source, the dose rate is high and the amount of cell killing will be close to that indicated by the acute-radiation survival curve. As the distance^from the source is increased, two changes take place: cells will be less sensitive at the lower dose rates, and within a given period of implantation the accumulated dose will also be less. These two factors lead to a very rapid change of cell killing with distance from the source [16]. This is illustrated in Figure 12.6 for the case of a point
186 Dose-rate effects with human cells
curves derives in part from the underlying assumed Poisson relationship between the average number of surviving cells per shell and the control probability. As is the case with tumor control by external-beam irradiation, there will in reality be factors that make the tumor control curves less shallow: heterogeneity, for instance. Within tissues (tumor or normal) that are close to the source, the level of cell killing will be so high that cells of any radiosensitivity will be killed. Further out, the effects will be so low that even the most radiosensitive cells will survive. Between these extremes there is a critical zone in which differential cell killing will occur. In this critical region the radiation dose rate will be low. For this reason we would argue that the low dose-rate survival curves as shown in Figures 12.3 and 12.4 are more clinically realistic than the high dose-rate curves, certainly for brachytherapy [13]. Figure 12.7 contrasts this situation with externalbeam radiotherapy, where the aim is to deliver a uniform
Figure 12.6 The likelihood of cure varies steeply with distance from a point radiation source. The radius at which failure occurs depends upon the steepness of the survival curve at low dose rate (upper panel). (From Steel [18].)
radioactive source. A source strength was chosen that gives 75 Gy in 6 days at a range of 2 cm. Three different tumor-cell sensitivities were assumed, as shown in the upper panel. It is the low dose-rate sensitivities that matter for this calculation. For spherical shells containing 109 clonogenic cells at different distances from the source, it was possible to calculate the surviving fraction from 6 days' irradiation, the absolute number of surviving clonogenic cells, and thus the probability that all cells in the shell would be killed. The results are shown in the lower panel. For cells of any given level of radiosensitivity there will be cliff-like change from high to low local cure probability, taking place over a radial distance of a few millimeters. Note that the order of the lines in the upper and lower panels of this figure is reversed: very sensitive tumor cells (lines A) can be cured out to a greater radius than less sensitive cells (B) or very radioresistant cells (C). The steepness of the tumor control
Figure 12.7 Variation of cell kill around a point source of radiation. The source gives 0.87 Gy min~1 at 2 cm (i.e. 75 Gy in 6 days); there are 109 cells per cm3, for which -a - 0.35 Gy1, P = 0.035 Gy2, ha If-time for recovery is 1 h. (a) The hatched area indicates the volume within which the surviving fraction is below 10~20. The stippled area indicates the volume where survival is between 10-20 and 10-6, which is the critical region for tumor control. For comparison, panel (b) shows the type of profile that would be aimed for with external-beam radiotherapy.
References 187
radiation dose across the tumor. Only in a narrow zone around an implanted source (where the surviving fraction changes from, say, 10-20 to 10-6) will radiobiological considerations be of interest or importance in relation to tumor control. The same principle will apply to normaltissue damage: serious damage to normal structures depends on making sure that they are outside the corresponding 'cliff.'
12.7 IMPLICATIONS FOR CLINICAL BRACHYTHERAPY The radiobiology of low dose-rate irradiation is now fairly well understood. Although data are not available on a wide range of human tumors, the data that we have do indicate the range of responses that are seen for human cells in tissue culture. It is likely that these will be realistic for effects on well-oxygenated cells in the patient. Much less is known about the effects of low dose-rate irradiation on hypoxic cells in vivo. These are, of course, less sensitive to high dose-rate irradiation. The work of Ling et al. [17] showed that the sparing effect of low dose-rate irradiation as a function of oxygen concentration was complex. Lowering the dose rate initially had more effect on the oxic cells than on the hypoxic cells. Further lowering of dose rate then had more effect on the hypoxic cells. Although for such reasons there is much that still needs to be understood about the tumor effects of brachytherapy, some simple conclusions can be drawn: 1. In the dose-rate range from a few Gy min-1 down to a few cGy mnr1, repair of radiation damage is the main modifying process on radiosensitivity. The effects are large, leading to a change in the isoeffective radiation dose by a factor of 2 or more. Below 1 cGy mhr1, cell proliferation will play an increasingly strong role in making tumors or normal tissues less sensitive to radiation damage. 2. There is evidence for a dose-rate effect in the region of 1 Gy mhr1. If, in external-beam radiotherapy, a change of machine or of source-skin distance leads to a substantial lowering of dose rate, then a doserate correction should be considered. 3. The biological effect of irradiation changes rapidly at dose rates around 10 cGy min'1. This may mean that greater precision in dosimetry and dose prescription is required in high dose-rate brachytherapy than when a low dose rate is used. 4. Tumor cells of different origins show very different response to low dose-rate irradiation. Theoretical calculations suggest that as we move out from an implanted radiation source the local tumor control probability will change rapidly, i.e., there will be sudden failure to eradicate all clonogenic tumor cells. The prediction that the range at which this
occurs will depend strongly on the low dose-rate radiosensitivity of the tumor cells could be clinically important [6]. There is a strong case for predictive testing of tumors that are to be treated with curative intent by brachytherapy in order to predict those that require a greater or lesser range of dosedistribution.
REFERENCES 1. Hall, E.J. (1972) Radiation dose-rate: a factor of importance in radiobiology and radiotherapy. Br.J. /tarf;o/.,45,81-7. 2. Steel, G.G., Down, J.D., Peacock, J.H. and Stephens, T.C. (1986) Dose-rate effects and the repair of radiation damage. Radiother. Oncol., S, 321-31. 3. Steel, G.G. (1997) The dose-rate effect: brachytherapy. In Basic Clinical Radiobiology, ed. G.G. Steel. London, Edward Arnold, 163-72. 4. Steel, G.G., Deacon, J.M., Duchesne, G.M., Norwich, A., Kelland,LR. and Peacock, J.H. (1987) The dose-rate effect in human tumour cells. Radiother. Oncol., 9, 299-310. 5. Kelland, LR. and Steel, G.G. (1988) Differences in radiation response among human cervix carcinoma cell lines. Radiother. Oncol., 13,225-32. 6. Bjbrk-Eriksson,T., West, C., Karlsson, E. and Mercke, C. (1998) Discrimination of human tumor radioresponsiveness using low-dose rate irradiation. Int.J. Radiat. Oncol. Biol. P/jys.,42,1147-53. 7. Amdur, R.J. and Bedford, J.S. (1994) Dose-rate effects between 0.3 and 30 Gy/h in a normal and a malignant human cell line. IntJ. Radiat. Oncol. Biol. Phys., 30, 83-90. 8. Oliver, R. (1964) A comparison of the effects of acute and protracted gamma-irradiation on the growth of seedlings of vivia faba. Int.J. Radiat. Biol., 5,475-88. 9. Szechter, A. and Schwarz G. (1977) Dose-rate effects, fractionation and cell survival at lowered temperatures. Radiat. /?«., 71, 593-613. 10. Dale, R.G. and Jones, B. (1998) The clinical radiobiology of brachytherapy. Br.J. Radial., 71,465-83. 11. Thames, H.D. (1985) An 'incomplete-repair' model for survival after fractionated and continuous irradiations. IntJ. Radiat. Biol.,47,319-39. 12. Curtis, S.B. (1986) Lethal and potentially lethal lesions induced by radiation -a unified repair model. Radiat. Res., 106,252-70. 13. Steel, G.G. (1991) Cellular sensitivity to low dose-rate irradiation focuses the problem of tumour radioresistance. Radiother. Oncol., 20,71-83. 14. Mitchell, J.B., Bedford, J.S. and Bailey, S.M. (1979) Doserate effects on the cell cycle and survival of S3 HeLa and V79 cells. Radiat. Res., 79, 520-36.
188 Dose-rate effects with human cells 15. Skladowski, K., McMillan, T.J., Peacock, J.H., Titley, J. and Steel, G.G. (1993) Cell-cycle progression during continuous irradiation of a human bladder carcinoma cell line. Radiother. Oncol., 28,219-27. 16. Steel, G.G., Kelland, LR. and Peacock, J.H. (1989) The radiobiological basis for low dose-rate radiotherapy. In Brachytherapy 2, Proceedings of the 5th International
Selectron Users' Meeting 1988, ed. R.F. Mould. Leersum, The Netherlands, Nucletron International BV, 15-25. 17. Ling, C.C., Spiro, I.J., Mitchell, J. and Stickler, R. (1985) The variation of OER with dose rate. Int.}. Radial. Oncol. Biol. P/?ys.,11,1367-73. 18. Steel, G.G. (ed.) (1997) Basic Clinical Radiobiology. London, Edward Arnold.
13 Radiobiology of high dose-rate, low dose-rate, and pulsed dose-rate brachytherapy DAVID J. BRENNER, ROGER DALE, COLIN ORTON, AND JACK FOWLER
13.1 BIOPHYSICAL MODELING IN RADIOTHERAPY In the 1970s, before the observations about the differential response of early-responding and late-responding tissues had been made, the technique in widespread use for predicting alternative fractionation schemes was the Nominal Standard Dose (NSD) approach, developed by Ellis [ 1 ]. This concept introduced for the first time the idea of separating overall time from the number of fractions, but did not include the important differential response to fraction size/dose rate of early and late effects. Following the initial suggestion by Barendsen [2], since the 1980s, the NSD (or time-dose factor, TDF) approach has been largely replaced by the linearquadratic (LQ) model. The reasons for this change were reviewed by Fowler [3]. Essentially, the NSD model, which is an empirically based approach, was useful when used to interpolate regimens with fraction numbers comparable to those from which its parameters were derived, but unsuccessful when used outside this range. When used to extrapolate, neither the assumed dependence on overall time nor the fraction size proved reliable in late-responding normal tissues. The LQ model is different from the NSD approach in two fundamental ways:
1. The LQ model is based on mechanistic radiobiological notions about how cells are killed by radiation, so extrapolation of clinical results, though always to be viewed with caution, is a more reasonable option than with NSD, which represents an empirical fit of a convenient, but nonmechanistically based, formula to a limited data set. 2. The LQ approach allows a clear separation to be made between effects in early-responding and lateresponding tissues. As described in this chapter, after more than a decade of investigation and use, the basic ideas and parameters in the LQ model have held up surprisingly well in the clinic. The large dose gradients and inhomogeneous dose distributions associated with brachytherapy present problems in deciding where to specify the dose prescription point for such treatments. There are, however, two general points which may be made, and which simplify some of the considerations: 1. Tumor control probability (TCP) is determined mostly by the minimum dose received by the tumor. Provided all parts of the tumor are within an isodose surface, which is sufficiently large to achieve a reasonable TCP, and provided there are no pockets of underdosage within that critical surface, the
190 Radiobiology of HDR, LDR, and PDR brachytherapy
degree of dose variation within that surface is not a major consideration. 2. In all types of radiotherapy, the normal tissue complication probability (NTCP) is volume related. In the case of a brachytherapy treatment, the dose gradients ensure that there is no clearly delineated treatment volume, and the NTCP is therefore more likely to be related to the integral dose received by critical structures. Because of these two considerations, no 'one number' dose specification can give an adequate indication of both the likely TCP and the likely NTCP. By the same token, attempts to compare the results of several treatments in terms of 'one number' dose specifications are fraught with difficulty and may be exceedingly misleading. Figure 13.1 helps to quantify the nature of the problem. The curves show how the biologically effective dose (BED) required for 50% TCP changes with increasing tumor radius, firstly for tumors containing only 1% of clonogenic cells, and then for tumors consisting of 100% clonogens. The curves have been derived by direct application of the LQ model to concentric shells around a central source. The net TCP associated with any particu-
lar tumor is determined by the net surviving number of clonogens. Two points emerge from the graphs: 1. As in other types of radiotherapy, radiosensitivity and the percentage of tumor clonogens are the most important determinants of the dose necessary to cure a tumor with brachytherapy. 2. For a tumor of given radiosensitivity and percentage of clonogens, the BED required for a given TCP increases only slowly with increasing size, i.e., provided the dose (and hence BED) is specified at the tumor surface. To judge properly the quality of brachytherapy applications, and to allow some degree of treatment optimization, new methods and parameters are now being proposed. Some of the indices which have been suggested relate only to the physical dose uniformity [4-6], whereas others relate to both the dose uniformity and the degree of overlap between the reference isodose surface and the normal tissue volume [7]. Whilst such indices are currently related to the physical characteristics of brachytherapy, their future development will undoubtedly require extensions which involve radiobiological parameters, thus opening the way to biologically optimized brachytherapy.
Figure 13.1 Illustration to show how the tumor surface dose (Gy) required for 50% tumor cure probability varies relatively slowly with increasing tumor radius (cm). Three sensitivities (a values) are considered: a,=0.45 Gy7, a2=0.35 Gy', a3-0.25 Gy1. (a) Tumor volume consisting of 1% clonogenic cells, (b) Tumor volume consisting of 100% clonogenic cells.
The linear-quadratic model 191
13,2 THE LINEAR-QUADRATIC MODEL 13,2*1 Mechanistic basis of the LQ model Central to the LQ approach is a biological model of radiation action, which was spelled out in detail over 50 years ago by Lea and Catcheside [8,9], based on a mechanistic analysis of chromosomal-aberration induction in Tradescantia spores. The application of this formalism to radiation therapy has been reviewed by Hall [10], Thames and Hendry [11], Dale [12], Fowler [3], and many others. Central to the approach is that radiotherapeutic response is primarily related to cell survival (or perhaps survival of groups of cells). This concept is itself not necessarily true, but there is now a wealth of evidence that cell killing is the dominant determinant of radiotherapeutic response, for both early-responding and lateresponding endpoints [10,11]. In the most basic LQ approach, cellular survival, S, at a dose D is written as:
The mechanistic interpretation of equation 13.1 is that cell killing results from the interaction of two elementary damaged species - probably DNA doublestrand breaks - to produce a species - perhaps a dicentric chromosomal aberration - which may cause lethality. The two terms in equation 13.1 indicate that the two elementary damaged species may be produced by the passage of the same track of radiation (linear term in dose) or by two different tracks (quadratic term). Clearly, if some time elapses between the passage of the first and second tracks, there exists the possibility of the first damaged site being repaired before interacting with the second. This repair will result in the reduction of the second, quadratic term in equation 13.1 (but not the first), by a factor denoted G by Lea and Catcheside [8]:
where, for acute exposures, G—>l, and for very long exposures G—>0. In this context, 'acute' and 'long' are defined relative to the half-time for repair of sublethal damage (Ti). In general, the G factor in equation 13.2 will depend on the details of the temporal distribution of the dose, as well as on TI. As discussed before, for many simple cases, G can be calculated analytically. For example, for n short, equal fractions, where the separation between fractions is much longer than T|, G«l/n. Formulae for many other standard schemes have also been derived [8,12,13], as has a general formalism for any possible scheme [14]. It is important to note the mechanistic basis of equation 13.2 so that it is not simply an equation which happens to fit cellular survival curves. It has been suggested, for example, that the LQ model can be considered sim-
ply as the first two terms (i.e., dose and dose squared) of a power-series expansion [15]. If LQ were just another empirical model, there would be no good reason for considering the linear dose term to be independent of protraction/fractionation, and the quadratic term in dose to be fractionation dependent. This distinction between the linear and the quadratic terms is at the heart of the LQ model and its application. Based on equation 13.2, we can proceed in two ways. We can either try to equate schemes, i.e., produce a regimen with the same, say, tumor response, as a 'tried and tested' regimen, or we can try to predict absolute responses. Both routes are discussed here, though it is argued elsewhere in the chapter that equating schemes is a far more reliable procedure. Thus, assuming tumor repopulation (described below) is negligible, to match a new fractionation scheme (labeled n) to a given ('old') fractionation scheme (labeled o), we must calculate the dose (DJ in scheme n such that:
Assuming we know how to calculate G, and that we know a suitable value for a/f3, equation 13.3 can clearly be solved to yield Dn. If we wish to proceed to the other route, and actually calculate absolute tumor control probabilities (TCP) or normal-tissue complication probabilities (NTCP), we need models relating cellular survival (S) with TCP or with NTCP. The standard model derives from the suggestion of Munro and Gilbert [16] that TCPs can be calculated from the probability that, after radiation treatment, there are no remaining stem cells capable of initiating tumor regrowth. Let us suppose that a dose, D, delivered in a given fractionation pattern, produces a stem-cell survival probability S. We define K to be the initial number of potential stem cells in the tumor, i.e. the number of cells that have the independent capability to initiate tumor regrowth. Then each initial stem cell will have a probability of not initiating tumor regrowth of (1-S), and thus the TCP is simply:
which, for small values of S, is approximately
The surviving fraction, S, is given by the LQ model as equation 13.2 or 13.7, and so the TCP is:
This same formula may also be used to calculate NTCPs, except that now the parameter K does not refer
192 Radiobiology of HDR, LDR, and PDR brachytherapy
to the number of tumor cells which need to be sterilized, but rather to the number of groups of cells in the normal tissue ('tissue-rescuing units' [17]), whose destruction would result in the late complication. The problem with using formulae such as equation 13.6 to calculate absolute, de novo, values of TCP or NTCP is that the results are exquisitely sensitive to the parameter values, particularly the K parameter, and, in general, the utility of these equations for absolute, de novOy calculations is quite limited. Rather, the main application of the LQ model is for comparisons between regimens equation 13.3, which are much less sensitive to the LQ parameter values.
13*2.2 Our knowledge of LQ parameters Around 1980, Withers and colleagues [18] made the key observation that early-responding and late-responding tissues differed in their responses to fractionation (and, by implication, low dose rate, LDR). Essentially, Withers and colleagues showed that, for a given dose, increasing the fraction size (or, by implication, decreasing the number of fractions or increasing the dose rate) will increase late effects much more than it will increase tumor control. Conversely, decreasing the fraction size (or increasing the number of fractions, or decreasing the dose rate) will decrease late effects much more than it will decrease tumor control. Thus the 'therapeutic ratio' (ratio of tumor control to complications) will increase as the number of fractions increases, or as the dose rate decreases. In terms of the LQ model, these observations can be interpreted in terms of the a/p ratio. In terms of survival curves (see Figure 13.2), the a/b ratio essentially describes the degree of 'curviness' of the acute survival curve. A small value of a/b means that the P (dose squared) term is dominating at radiotherapeutic doses, resulting in a curvy survival curve (bottom curve in Figure 13.2). A large value of a/b means the a (linear in dose) term is dominating, resulting in a straighter survival curve. Now, as a first approximation, the dose-response relation for a fractionated (or LDR) regimen can be thought of as simply the result of multiple repeats of the initial part of the survival curve. Clearly (see Figure 13.2), repeating the early part of a curvy survival curve many times will result in far more sparing than repeating the early part of a straighter survival curve. Thus, late effects, which are very sensitive to changes in fractionation, are characterized by small values of a/b, and early effects (tumor control or early-responding normal sequelae) are characterized by large values of oc/p. As clinical data accumulated during the 1980s and 1990s from which a/P ratios can be derived, the dichotomy between a/b ratios for early and late effects, originally inferred from animal data, has held up remarkably well.
Figure 13.2 The dose-response curve for late-responding tissue is 'curvy', i.e., has a small a/b ratio; for early-responding endpoints such as tumor control, the dose-response curve is straighter, i.e., the a/b ratio is larger. Consequently, dose fractionation spares late-responding tissues more than earlyresponding tissues. (Adapted from reference 63.)
Consequently, when using the LQ model, it is essential to be clear about whether the calculation is designed to refer to early-responding or late-responding tissue. From equation 13.3, it is clear that use of different values of a/P will result in different answers for the isoeffect dose.
1.3.2.3 New extensions to the LQ model REPOPULATION
Equation 13.2, which is the basic LQ equation, addresses only the inactivation of a homogeneous population of cells. There is no doubt, however, that accelerated repopulation of tumor clonogens during radiotherapy is often an important factor in determining tumor control [19,20]. The LQ formalism can also take into account the effects of tumor repopulation, i.e., the effects on tumor control of changes in the overall time. Following the original formulation by Travis and Tucker [21], overall time is taken into account by increasing the surviving fraction by a factor exp(y[T-TD]), where T is the overall treatment time, and TD is the delay after the beginning of the treatment before tumor-cell proliferation begins. Then, the survival is given by:
where
The linear-quadratic model 193
The parameter y determines the speed of the repopulation, and is given by: y=0.693/7;,
(13.8)
where Tp is the effective doubling time of cells in the tumor. If we can ignore spontaneous cell loss, Tp is approximately the same as the measurable in-vitro doubling time of the tumor cells. In summary, then, to match a fractionation scheme (labeled 2) to a given fractionation scheme (labeled 1), we must calculate the dose (DJ in scheme 2 such that:
with the same convention as equation 13.7. It should be noted that, for late-responding tissues, y is effectively zero. In other words, late-responding sequelae do not vary significantly with changes in overall treatment time. REDISTRIBUTION AND REOXYGENATION
The LQ model as used in equation 13.7 incorporates sublethal damage repair and repopulation. These two factors are often referred to as two of the '4Rs' of radiotherapy. Recently, an extension has been proposed to the LQ model, termed LQR, to include also the other two 'Rs': cell cycle redistribution and reoxygenation [22]. In the LQR approach, redistribution (due to progression through the cell cycle) and reoxygenation are both regarded as aspects of a single phenomenon, termed resensitization [23]. Resensitization occurs when a radiation exposure preferentially kills the more radiosensitive cells in a diverse population, producing a decreased average radiosensitivity just after the dose is administered; subsequent biologically driven changes then tend gradually to restore the original population average radiosensitivity. The LQR model represents a generalization of the LQ model. In contrast to some multi-parametric approaches, it uses only two additional adjustable parameters, an overall resensitization time and overall resensitization amplitude. Its essential feature is to replace the LQ cell survival equation (equation 13.7) with the equation:
Here, all the terms except the last are the same as for the LQ survival equation (13.7); they again model lethality, repair, and repopulation. The new term is the last, + {(72 (jD2. This term contains the resensitization magnitude, which is positive and is written }a2. This resensitization magnitude is regarded as an average for the dominant resensitization effects present in a heterogeneous tumor. The influence of fractionation on resensitization is contained in the factor (j of the above equation and depends on a characteristic resensitization time, TS. When simplifying assumptions were introduced
for cell-cycle distribution and reoxygenation [22,23], G turned out to have exactly the same form as the G function for sublethal damage repair, i.e., in the term GfSD2, with the characteristic repair time replaced with the characteristic resensitization time, Ts. However, in contrast to repair, resensitization tends to increase sensitivity as the overall time increases. For example, tumor cells which were in a resistant part of the cell cycle at the time of one fraction and were thus preferentially spared may move to a more sensitive part of the cell cycle. This difference between repair and resensitization is manifest in equation 13.10 by the'+' sign in the resensitization term compared with a'—' sign in the repair term. While mechanistically driven, the model is designed to be sufficiently simple that it can be practically applied to isoeffect calculations in radiotherapy. Its idealizations in the consideration of resensitization parallel those inherent in the standard LQ model relating to repair. The model gives reasonable fits to relevant experimental data in the literature [22]. THE EFFECTS OF TUMOR SHRINKAGE
When there is a significant degree of ongoing shrinkage throughout a course of brachytherapy, and when the radiation sources are centrally situated within the tumor volume, the BED to the tumor may be higher than that calculated with standard equations. Because the physical dose in a brachytherapy treatment varies rapidly with distance from the source(s), radiation oncologists usually select a prescription point (itself an element of a threedimensional isodose surface designed to enclose the assumed tumor volume) at which the reference dose for the treatment is specified. If that reference surface diminishes in size as a result of tumor shrinkage, an increase in physical dose results, because those cells at the dose specification point will move closer to the source(s) during treatment, and will receive a higher dose than that prescribed. The radiobiological consequence of this has been examined in some detail by Dale and Jones [24,25]. By assuming that the linear dimensions of the tumor shrink exponentially with time, the effective BED of a high dose-rate (HDR) brachytherapy treatment consisting of N fractions of dose d (to the dose prescription point) is given by:
where X = 1 +(N/-1)zf
and z is the rate of linear tumor shrinkage (day1), and t is the time between fractions (days). For a continuous low dose-rate (CLDR) treatment delivering a dose rate of 1? Gy h'1 over T hours, the BED is:
where
194 Radiobiology of HDR, LDR, and PDR brachytherapy Y = 1 + zT/24
The similarity of form between equations 13.11 and 13.12 should be noted. For late-responding normal tissues adjacent to the tumor, there will be little or no repopulation during the treatment (i.e., K= 0). If the normal tissues move toward the source(s) as a result of the shrinking of the tumor, then the late-reacting BED will always increase with increasing fraction interval. The interesting point about equations 13.11 and 13.12 is that they indicate that an increase in BED is not an inevitable consequence of tumor shrinkage. In the case of HDR, increasing the time interval (?) between fractions allows for more shrinkage, but this will improve the BED only if the following inequality is satisfied:
Equation 13.13 confirms what is to be intuitively expected, i.e., that such favorable conditions are more likely to exist when the dose equivalent of the repopulation rate (fC) is small, and/or when the shrinkage rate (z) is large. Because equation 13.13 incorporates the ratio K/z, and not merely z, it serves as a reminder that ongoing shrinkage is not, by itself, enough to guarantee that the BED will be favorably increased by the use of larger fraction intervals, i.e., whether or not there is an increase in biological effect with shrinkage is dependent on the ratio of K/z. Dale and Jones [24,25] have examined the possible radiobiological consequences of tumor shrinkage, in terms of the optimal interfraction spacing of HDR brachytherapy. They concluded: 1. When the shrinkage rate (z) is small, then, largely irrespective of the value of K, there will be few benefits associated with the use of a long time gap (?) between fractions. For most values of Kthe BED always reduces with increasing time interval. 2. For higher values of z and moderate or low K values, BED can be usefully increased by increasing the time gaps. 3. Increasing t between fractions is beneficial only when the shrinkage constant and the daily repopulation factor are favorably related. When there is doubt about the radiobiological parameters, or in the absence of predictive assays (of SF2 and Tpot, for example), and if the tumor regression rate cannot be determined during the initial phase of treatment, then relatively close spacing of fractions minimizes the possibility of large variations about the prescribed value of BED. Close fraction spacing is therefore the 'safe' option in such cases. Following from conclusion (3), it is possible to investigate the radiobiological conditions which will allow the BED to increase with increasing fraction interval. In
mathematical terms, this is equivalent to saying that the slope of the curve of BED versus time - given by the differential d(BED)/dt - will be positive for all values of t in this condition. The conditions for which this is true lead to the following conclusion. For critical cases (high K value and/or small shrinkage rate), the ratio K/z will be large, and in such cases the interfraction intervals should be kept as small as possible. This is because the BED finally delivered in such cases will always be less than that intended (primarily because of the dominant effect of the high rate of repopulation), causing the loss of biological effect to increase as the interval between fractions increases, i.e., d(BED)/dt will usually be negative over the range of fraction intervals which are likely to be used in practice. These considerations concerning tumor shrinkage may also impact on the optimal time spacing between external-beam radiotherapy and brachytherapy. The precise time at which tumor shrinkage begins during a course of radiotherapy is an important consideration in judging how teletherapy and brachytherapy should be juxtapositioned. In practice, it is unlikely that dynamic tumor shrinkage will commence immediately after the first radiation delivery and, for this reason, it is likely that equation 13.11 overestimates the increase in BED when the K/z ratio is small. Similarly, there may be underestimation of the decrease in BED associated with a large K/z. The equations in this section may therefore have greater validity when the brachytherapy treatments have been preceded by teletherapy, or other anticancer modalities such as cytotoxic chemotherapy, which also causes tumor volumes to decrease exponentially [26]. It is therefore appropriate to consider separately brachytherapy used alone and brachytherapy used in combination with teletherapy. For radical brachytherapy, relatively large doses will be prescribed, and the conditions of the inequality in equation 13.13 are more likely to be satisfied because the product Nd is large, so that the tumor BED will probably increase with time once tumor shrinkage is initiated. Although difficult to implement in practice, there might be advantages in initially using relatively short interfraction intervals and then increasing them once the tumor regression has attained a steady state. However, even with a small K and/or large z (i.e., small K/z), excessively extended intervals may not be feasible if HDR applicators require to remain in situ throughout the whole course of treatment [27]. It is also possible that overextension of t may allow for accelerated repopulation in squamous cell carcinomas [19,20] or, alternatively, a reduction of the spontaneous cell-loss rate [28], either of these phenomena requiring the delivery of additional dose to restore isoeffectiveness. Brachytherapy is most commonly used in combination with external-beam radiotherapy. For high K/z ratios, the brachytherapy should be included within the overall time required for delivery of the teletherapy
The linear-quadratic model 195
regime. Where a small K/z ratio is known to be obtainable, the application of brachytherapy should be deferred until steady-state tumor shrinkage has been initiated, i.e., brachytherapy will always be more effective if it follows teletherapy in this case. Further improvements can be achieved by optimizing the time gap between the cessation of teletherapy and initiation of brachytherapy, but the overall TCPs remain dependent on tumor size [29,30]. If the tumor shrinkage rate following teletherapy is very small, the brachytherapy BED is unlikely to be significantly enhanced. In such cases, particularly for larger tumors, debulking surgery can be used to reduce clonogen number, followed by brachytherapy to the tumor bed. Alternatively, cytotoxic chemotherapy can be used to reduce tumor repopulation during continued tumor regression, and brachytherapy given at a later stage to the smaller tumor [29,31]. Other considerations may apply for clinical situations in which the tumor center is distanced away from the treatment catheter, as in the case of intraluminal HDR treatments for carcinomas of the bronchus and esophagus. This form of treatment geometry has been considered by Bleasdale and Jones [29], but the overall principles are the same as described in this chapter. With an offset tumor center, any regression will not produce the same degree of physical dose advantage, so that extensions of overall time will not always be advantageous. If the clinical constraints are such that HDR bronchoscopic applications can only be performed every 2 weeks, and the tolerable number of fractions is only three, then brachytherapy should be commenced relatively early within a 6-7-week teletherapy regime [32]. If the three HDR brachytherapy fractions can be delivered over a period of 1 month, when the repopulation factor is likely to be unfavorable, consideration should be given to shortening the external-beam therapy to the same overall treatment time. In all cases it must be remembered that tumor shrinkage may cause adjacent normal tissues to move closer to the treatment source(s). For radical treatments, any decision to increase the tumor BED by prolonging the fraction interval must be tempered by the possibility that the normal tissue tolerance dose may be reached or exceeded. Whether or not this is the case can only be determined by careful measurement or estimation of the movement of the critical normal tissues at the time of delivery of each dose fraction. It is clear that the development of improved brachytherapy treatment requires a greater knowledge of the radiobiological processes which govern treatment outcome, and in this chapter the importance of the opposing effects of tumor shrinkage and repopulation particularly have been highlighted. Predictive assays and serial imaging techniques are already available and are likely to be powerful tools in allowing the mathematical models to be more effectively applied for the benefit of
individual patients. The wider availability and use of such methods would seem to be the obvious next step in the further development of brachytherapy. MOVING ISODOSE SURFACES
Because brachytherapy doses are traditionally prescribed at a specific point, it has become commonplace also to compare radiobiological effects at a defined anatomical point. This has led to the notion that, if therapeutic ratio is not to be seriously compromised, considerable care must be exercised when replacing LDR techniques by fractionated HDR. Many articles have dealt with this aspect of radiobiology, but, as discussed later in this chapter, in certain special situations (of which brachytherapy of the uterine cervix is the most common), it may be demonstrated that HDR brachytherapy in small numbers of fractions may be less detrimental than is sometimes supposed. Dale [33], Brenner and Hall [34], and Orton [35] have shown that, even without relatively favorable radiobiological parameters, a modest extra amount of geometrical spacing of critical normal tissues at HDR allows the use of small numbers of fractions without loss of therapeutic ratio. Improved geometry is easiest to achieve in the case of intracavitary treatments, less so with intraluminal brachytherapy. A refreshing new analysis of the problem has recently been conducted by Deehan and O'Donoghue [36], who considered how the relative positions of isodose surfaces are changed in switching from LDR to HDR. Although the issue is seemingly complex, it may be summarized quite simply. Once a fractionated HDR regime has been designed to match an LDR regime at a particular reference point, the LQ model may be used to calculate the biological effects (BEDs) at other sites closer to, and further away from, the sources. Figure 13.3 illustrates a typical case. It will be noted that, at all points closer to the sources, the HDR treatment delivers a higher BED than the LDR treatment. Conversely, at those sites on the distal side of the matching point, the BEDs with HDR are always lower than in the LDR case. The important point here is that the gradients of the BED versus distance curves differ between HDR and LDR, even though the effects have been matched at the reference point. The separation between these curves, in fact, becomes greater as the fraction number is reduced. The individual elements making up the curves in Figure 13.3 are parts of smooth isoeffect surfaces surrounding the treatment sources. Thus, for sites close to the sources, any particular LDR isoeffect surface moves further out when HDR is used, i.e., the volume of enclosed tissue receiving more than a given biological dose is increased. Similarly, at surfaces beyond the matching point, the isoeffect surfaces move inwards, and the volumes of tissue receiving a given biological dose are reduced. Because the more proximal sites are likely to
196 Radiobiology of HDR, LDR, and PDR brachytherapy
volume effects. For intraluminal treatments in particular, consideration of surface movement shows that fractionated HDR may offer benefits over LDR.
133 HIGH DOSE-RATE VERSUS LOW DOSERATF RRAfHYTHFRAPY
13.3.1 From LDR to HDR: general principles
Image Not Available
Basic radiobiological principles tell us that the optimal strategy for any radiotherapeutic regimen requires: 1. Use of large numbers of fractions or LDR to maximize sparing of late-responding normal tissues, and to allow reoxygenation; 2. Use of short overall times to limit tumor repopulation; 3. Use of long overall times to reduce early normaltissue sequelae, especially to the skin and mucosa.
Figure 13.3 Illustration to show how radiobiological equivalence between LDR and HDR is achievable only at the chosen matching point. At sites closer to the sources, the HDR delivers a higher biological dose (BED) than LDR; at sites on the distal side of the matching point, the BED from the HDR is lower than from the LDR. The lateral separation between the two curves increases as the number of HDR fractions is decreased. Because the curves are two-dimensional representations of three-dimensional isoeffect surfaces, the switch from LDR to HDR brings about a (possibly useful) increase in the tissue volume irradiated to more than a given biological dose at sites close to the sources, but a favorable decrease in the volume receiving more than a given biological dose at sites beyond the matching point. (Reproduced from reference 36, with permission.)
be in the vicinity of tumor cells, and the more distal sites are likely to be in the vicinity of normal tissues, this relative shifting of the isodose surfaces will tend to enhance the integral dose (i.e., dose x volume product) for tumor, whilst diminishing that for the normal tissues. Such trends may be ideally expressed in terms of effectvolume histograms. Further extension of Deehan and O'Donoghue's work allows assessment of the linear displacements in the isodose surfaces, and hence of the changes in the included tissue volumes. It thus paves the way to a better understanding of radiobiological differences between HDR and LDR in terms of changes in irradiated volumes, and opens up a new avenue for investigating one of the most elusive aspects of brachytherapy - the modeling of
Because of the physical dose distributions in brachytherapy, the last point is not normally a major concern, and it becomes clear why brachytherapy is often the treatment of choice for those tumors which are reasonably accessible for an implant. It is also clear that moving from LDR to HDR must generally involve a loss in therapeutic advantage. To put it in terms of LQ calculations, if an HDR dose is calculated using, say, equation 13.3, based on producing equal tumor control to an LDR regimen, that HDR dose will not be isoeffective in terms of late effects, but will produce increased late sequelae. Conversely, if an HDR dose is calculated to produce, say, equal late sequelae to an LDR regime, the HDR dose would be expected to produce less tumor control than the corresponding LDR regime. Thus, in general, HDR represents a compromise between therapeutic advantage, which decreases, and some other factor (such as patient convenience, treatment repeatability, etc.), which may increase. 133.2
A special case: cancer of the cervix
It is clear from the previous discussion that many fractions are necessary in order for HDR treatments to be radiobiologically equivalent to LDR, all other things being equal. However, for the treatment of carcinoma of the uterine cervix, it has been well documented that equal, and maybe even superior, results (same local control and survival but with fewer complications) can be obtained with as few as about five fractions [33-35]. This apparent discrepancy between radiobiological theory and clinical evidence is probably because not all other things are equal for cervical cancer radiotherapy. Clearly, one of the major reasons why brachytherapy is so
High dose-rate versus low dose-rate brachytherapy
effective and therefore widely utilized in these treatments is that the radiation sources are placed in and around the tumor and away from the normal tissues most susceptible to late radiation damage, specifically the rectum and bladder. In the previous sections, it was assumed that the doses to tumor and normal tissues were the same, whereas, with cervical cancer brachytherapy, packing and/or retraction, 'optimal' source distributions, and applicator shielding are all strategies that can be applied to keep normal tissue doses below those applied to the tumor. As discussed earlier in this chapter, because cervical cancer brachytherapy dose distributions are so inhomogeneous, it is not really appropriate to consider the dose to just a single point, e.g., Point A, to represent the 'tumor dose' [37]. Doses to single rectal or bladder 'points' similarly do not truly represent 'normal tissue doses' [38]. What are required are the 'effective doses' to tumor and normal tissues derived by analysis of threedimensional dose-volume histograms, where the 'effective dose' is that dose which, if delivered uniformly to the tissue in question, would result in the same probability of effect (TCP or NTCP) as the inhomogeneous dose distribution present in that tissue. Several techniques, though not ideal, have been devised to reduce dosevolume histogram data to a single number, such as 'effective dose' [39]. For brachytherapy treatments for cervical cancer, it has been demonstrated that the vast majority of the cervix tissue receives substantially higher doses than most of the rectal and bladder tissues [38]. This is illustrated for rectal tissues in Figure 13.4. Hence, the 'effective doses' to normal tissues are significantly lower that the 'effective tumor dose.' This needs to be taken into account when applying radiobiological modeling, such as with the LQ theory, for comparison of LDR and HDR. However, such 'protection' of normal tissues is a benefit with both HDR and LDR treatments, so some additional 'protective' attributes of HDR must be contributing to the low complication rates observed. One of these is illustrated in Figure 13.4b, which shows that optimization of the dwell positions of the single HDR stepping source can significantly reduce the effective dose to rectal tissues compared to when 'fixed' LDR sources are used. A similar effect is observed for the bladder [37]. Hence, the almost infinite variety of source distributions obtainable with a stepping source makes it possible to achieve superior dose distributions to those attainable with conventional LDR fixed sources. (Note, however, that pulsed-LDR brachytherapy - PDR, discussed later in this chapter which also uses a single stepping source, will have a similar advantage.) A second potential advantage of HDR over LDR is the ability to make better use of packing or retraction. It has been reported frequently that the short duration of HDR treatments enhances the effectiveness of packing or retraction techniques. For example, gauze packing is known to shrink during long LDR irradiations. Also,
197
Image Not Available
Figure 13.4 Typical differential dose-volume histograms: (a) for cervix, and (b) for rectal tissues obtained using the University of Michigan CT-compatible Fletcher-type applicator for the treatment of carcinoma of the cervix [37]. The prescription called for a minimum dose of 11 Gy to the cervix, which, for a conventional LDR application, would correspond to a Point A dose of about 20 Gy [38]. The solid curves are the dose-volume histograms for non-optimized source loadings typical of those employed in LDR treatments. The dashed curves are the dosevolume histograms obtainable with the HDR stepping-source technology using optimization of dwell positions. (Reproduced with permission of Dr Mary Martel, personal communication.)
extensive retraction during the short HDR treatments can readily be tolerated, especially if the patient is anesthetized or medicated appropriately. Another less obvious potential benefit of HDR relates to differences in rates of repair between tumor and latereacting normal tissue cells. It appears that, on average, tumor cells tend to repair sublethal damage faster than normal cells [40]. This will be an advantage for HDR because almost no repair will be possible during short HDR exposures, whereas, during protracted LDR treatments, tumor cells will be better able to take advantage of the time available for repair, so repair at low dose rates will be enhanced more for tumor than for normal tissue cells.
198 Radiobiology of HDR, LDR, and PDR brachytherapy
These qualitative physical and biological potential advantages of HDR can be quantified by the use of the LQ model. LQ MODEL CALCULATIONS FOR CARCINOMA OF THE CERVIX: HDR VERSUS LDR
For radiobiological comparisons of LDR and HDR brachytherapy for cervix cancer, the following terminology will be applied. Let: • /represent the fractional geometrical sparing of normal tissues due to employment of brachytherapy, and therefore obtainable with both HDR and LDR, i.e.,/= effective normal tissue dose/effective tumor dose; • / be the extra fractional geometrical sparing of normal tissues attainable with HDR due to advantages of stepping source technology and better retraction or packing, i.e., j/' is the total geometrical sparing factor for HDR treatments; • |it and |l be the repair-rate constants for tumor and late-reacting normal tissue cells, respectively; • 'therapeutic advantage' (TA) of HDR over LDR be the ratio of the tumor log cell-surviving fractions, HDR versus LDR, for the same late-reacting normaltissue log cell surviving fractions, i.e.,
We can apply the basic LQ formula (equation 13.2) to determine the number of HDR fractions necessary to be equivalent to an LDR regime with effective tumor dose rateO.833 Gyh~'(60 Gyin72 h), assuming the same repair rate constant of 0.46 h~] for tumor and normal tissues (corresponding to repair half-time of 1.5 h) and no conventional brachytherapy geometrical sparing (/=!). This leads to the curves shown in Figure 13.5 for four values of /'. Thisfigureshows that the extra geometrical sparing that can be achieved with HDR, represented by the factor /', plays a very significant role. In this example, which illustrates the least favorable scenario for HDR (no normal geometrical sparing (/= 1) and no difference in repair rates), more than 17 fractions are needed before HDR becomes superior to LDR (TA>1) if /'=!, whereas this number of fractions reduces to 12,8, and 4 as/' decreases to 0.95, 0.90, and 0.85, respectively. With the additional advantages of faster repair of tumor cells and some degree of normal geometrical sparing of normal tissues (/
Figure 13.5 LQ model predictions of the therapeutic advantage of HDR over a 60 Gy in 72 h LDR regime as a function of the number of HDR fractions employed and the extra geometrical spa ring factor (Y) achievable with HDR. A therapeutic advantage >1 means that HDR is superior to LDR. Assumptions for this figure are: no normal geometrical sparing, i.e., f = 1; repair rates of tumor and late-reacting normal tissue cells are equal, with repair half-times 1.5 h; tumor o/P = 10 Gy; late-reacting normal tissue o/p = 3 Gy.
Figure 13.6 Combinations of f and f required for five fractions of HDR to be equivalent (therapeutic advantage = 1) to LDR regimes of either 60 Gy in 72 h (dose rate, R = 0.833 Gy h'1) or 60 Gy in 144 h(R = 0.417 Gy h~1), for two different tumor-cell repair rate constants (\i) 1.4 h-1 or 0.46 h~\ corresponding to repair half-times of 0.5 h or 1.5 h, respectively. For combinations of f and f below each curve, HDR is superior to LDR for the parameters assumed, and vice versa above each curve. Assumptions are: for late-reacting normal tissues, ji = 0.46 h~\ o/p = 3 Gy; for tumor, o/p = 10 Gy.
Each of the four curves in Figure 13.6 represents a different combination of tumor repair rate constant (|n) and LDR dose rate. For combination of / and /' below each curve, HDR is superior to LDR for the parameters assumed; and above, LDR is superior. For example, if |it is assumed to be 1.4 h~', and the LDR tumor dose rate is 0.833 Gy Ir1, HDR will be better than LDR for all combinations of/and/' below the top curve in Figure 13.6,
Low dose-rate versus pulsed dose-rate brachytherapy 199
i.e., for most values of / or /' less than unity, which icc highly likely. At the other extreme, if tumor cell repair rates are assumed to be the same as normal cells (mt = 0.46 h-1), and the LDR dose rate is only 0.417 Gy h'1, there are many combinations of/and/' for which LDR will be the superior treatment, although HDR will always be better if/'<0.8, regardless of the value of f. Clearly, replacing LDR with HDR is more difficult if the LDR dose rate is very low and/or the rate of repair of tumor cells approaches that of normal cells. Because it is desirable to obtain results at least as good as achievable with LDR at very low dose rates, and because repair rates of tumor cells are reported to vary widely but are not known for individual patients [41], it becomes essential to maximize geometrical sparing of normal tissues when converting from LDR to HDR. The superior results obtained with HDR for the treatment of cervical cancer are presumably due to this additional geometrical sparing of normal tissues, with possible enhancement for at least some patients attributable to a fortuitous combination of repair rates of normal tissue and tumor cells.
Figure 13.7 Gamma-ray doses for multi-fractioned stereotactic radiotherapy, which are calculated to be equivalent, in terms of tumor control, to single-fractioned radiosurgical doses. (Adapted from reference 63.)
1333 From HDR to LDR: fractionated stereotactic radiotherapy Primarily because of therapeutic convenience, there has been a trend in favor of moving from LDR to HDR. There is one situation in which the reverse is true, namely in fractionated stereotactic radiotherapy. The background to this development is the technique of single-treatment stereotactic radiosurgery, originally designed to treat arteriovenous malformations, which has been adapted to the treatment of primary and secondary brain tumors. For the reasons outlined above, namely because of the lack of sublethal damage repair and of reoxygenation in a single treatment, such singlefraction treatments are probably inappropriate for treating malignancies. The LQ model (in the form of equation 13.3) has been used to calculate fractionated regimes which are predicted to be isoeffective to currently used singlefractioned regimens (Figure 13.7). There is some question, however, of whether the LQ model actually applies in situations in which a single fraction of, say, 20 Gy is the baseline regime. Figure 13.8, for example, shows some well-known isoeffect results from Van der Kogel [42] for late-responding damage to the rat spinal cord, and from Douglas and Fowler [43] for acute damage in mouse skin. The form of the plots is such that, if the LQ formalism applies, the data would fall on a straight line [43]. Although there are more sophisticated methods available for assessing agreement with the LQ model [44], given the inherent uncertainties in the data, it is clear that all these data, including those for single fractions of ~ 20 Gy, are not inconsistent with the LQ model.
Figure 13.8 Isoeffect data for late response from three (D,O, A) different regions of the rat spinal cord [42], and for acute skin reactions ( ) in mice [43]. All the points at > 20 Gy/fraction correspond to single acute exposures. The data are plotted in a form [43] such that, if they follow a linearquadratic relationship, the points would fall on a straight line. The ratio of the slope to the intercept of this straight line then provides an estimate of the o/p ratio.
13.4 LOW DOSE-RATE VERSUS PULSED DOSERATE BRACHYTHERAPY
13.4.1 Introduction to PDR Because of its practicality and its logistic similarity to continuous low dose-rate (CLDR) brachytherapy, use of
200
Radiobiology of HDR, LDR, and PDR brachytherapy
pulsed dose-rate (PDR) brachytherapy is increasing [46-48]. In PDR, a CLDR brachytherapy regimen is replaced with one involving a series of high dose-rate pulses, typically (though not always) taking about 10 min h~' and typically (though not always) with the same overall dose and time as the corresponding CLDR regimen. PDR is achieved with a remote afterloader containing a single high-activity source which is stepped through the catheters of an implant, with dwell positions and times adjusted under computer control to achieve the required dose distribution. The advantages of PDR have been discussed elsewhere [41,49,50]. The patient has much more mobility - during the 'off' periods - than in a conventional CLDR regimen, during which nursing and visiting can be safely accomplished. There are two clinical advantages. First, by varying dwell times and locations of the source as it shuttles through the tumor, dose distribution can be optimized for the actual locations of the implanted catheters relative to the tumor and normal tissues. Second, the overall dose rate can be maintained even as the source decays, by increasing the length of individual pulses. Finally, from a practical viewpoint, the use of a single source has both logistic and radiation protection advantages compared with the usual inventory of sources.
13.4,2 'Daytime'PDR The original PDR protocols have been for day and night irradiation. It would clearly be advantageous to design pulsed-brachytherapy (PDR) protocols that are expected to be at least as clinically efficacious (in terms of both tumor control and late sequelae) as CLDR regimens, but that involve irradiation only during extended office hours. The LQ formalism has been used [49] to design PDR schemes in which pulses are delivered during 'extended office hours' (8 a.m. to 8 p.m.), with no irradiation overnight. Generally, the proposed PDR regimes last the same number of treatment days as the corresponding CLDR regimen, but the PDR treatment lasts longer on the final day (i.e., until 8 p.m.). PDR doses were calculated such as to produce a tumor control which is equivalent to standard CLDR protocols, and the corresponding predicted late complication rate was compared with that for CLDR. Ranges of plausible values for the half-times of sublethal damage repair for tumors and for late-responding normal tissues were considered. The efficacy of PDR relative to CLDR depends considerably on the repair rates for sublethal damage repair. The clinical and experimental evidence suggests that average repair half-times for early effects (e.g., tumor control) are less than about 0.5 h, and for late sequelae are more than about 1 h (but see below). If these estimates are correct, daytime PDR regimes can usually be designed which take the same number of days as the corresponding CLDR
regimen, but have comparable or better therapeutic ratios than responding tissues. The suggested protocols allow all of the advantages of a computerized, remotecontrolled afterloader while preserving the benefits of low dose rate. In addition, the protocols could allow the patient to go home overnight, or to stay overnight in an adjacent medical inn or hospital-associated hotel, rather than in a hospital bed - which could have major economic benefits. In such an economic situation, an extra treatment day for the daytime PDR could well be considered, which would virtually guarantee an improved clinical advantage relative to CLDR.
13*43 Equivalent regimens for PDR The key radiobiological question for PDR revolves around the question of equivalence between the results of CLDR and those of a corresponding PDR regimen. Initial calculations [41,50], based on equation 13.3 and LQ parameters from in-vitro systems, suggested that, as long as the time between, say, 10-min pulses was not increased much beyond 1 h, early-responding normal tissues would not show significant differences in response between CLDR and PDR (for the same overall dose and time). An example of the result of such a calculation is shown in Figure 13.9. Subsequent in-vitro experimental results [51,52] have corroborated this conclusion, as have in-vivo studies with an early-responding endpoint [53]. The limited clinical experience with PDR reported to date also suggests that early response is not markedly different from CLDR [45-48]. Several authors, however, have pointed to the need for caution with regard to late effects [41,54-57].
Figure 13.9 Combination of pulse widths and periods between pulses that will yield an equivalent survival to a continuous low dose irradiation of 30 Gy in 60 h. For this particular cell line, any combination of pulse width and period within the marked boundary is predicted to yield equivalent cell killing. The figure shows representative data for one of 38 cell lines analyzed in reference 41. (Redrawn from reference 41.)
Conclusions 201
Essentially, this is because of the fact (discussed above) that late-responding tissues are more sensitive than early-responding tissues to changes in fractionation patterns. These authors pointed out that changes in late effects when moving from CLDR to PDR are essentially determined by the rate of repair of sublethal lateresponding damage - and that these repair rates are simply not well known. Essentially, the trend, schematically illustrated in Figure 13.10, is that rapid repair rates in late-responding tissues would lead to increased late effects in PDR compared with CLDR [58]. On the other hand, slow repair rates would imply that PDR might well produce fewer late effects than the corresponding CLDR regimen. Experimentally, changes in late-responding sequelae are hard to quantify, particularly when these changes may well be relatively small. This is true both in the clinic and in the laboratory. In the clinic, a variety of reports have generally reported no significant difference in late sequelae between PDR and CLDR [45-48]. In a model late-responding system (cataract induction in the rat lens), no significant difference was observed between 15 Gy of X-rays delivered over 24 h and in various PDR regimens with the same overall dose and time [59]. Similar results have been obtained using as an endpoint rectal toxicity in the rat [60,61].
Figure 13.10 Calculated fractional change in cell survival for PDR compared with LDR as a function of the assumed half-time for sublethal damage repair. Both treatments consist of 30 Gy delivered in 60 h, either continuously (LDR) or in 60 10-min 50 cGy pulses delivered every hour (PDR). The calculated quantity is fSPDR - S Ldr) S LDR ; here, the survival (S) is calculated, using the linear-quadratic formalism (3,8), asS = exp(-ccD -GpD2) where D is the total dose, a and P are the linearquadratic formalism parameters, and G is the quantity describing sublethal damage repair (see equation 13.2), which depends on the half-time of sublethal-damage repair, TJ/2. Thus, the quantity calculated is actually exp/-(GPDR - GL JpD2/ -7. In the calculation, it was assumed that (3=0.025 Gy~2, though similarly shaped graphs are obtained for other values of p. (Redrawn from reference 58.)
The equality of late effects from CLDR and PDR in the laboratory must imply that sublethal damage repair is quite slow in this model late-responding system, in agreement with trends observed in the clinic for sublethal damage repair of late sequelae. Such trends would suggest that PDR is unlikely to produce significantly worse late effects than the corresponding CLDR regime, which is in agreement with early clinical data using PDR. Caution, however, is strongly indicated.
13.5 CONCLUSIONS In this chapter we have discussed the principles underlying the use of the LQ model, and some examples in designing equi-effect doses, for either early-responding or late-responding tissues. The main application of the LQ model is likely to be for comparisons of schema, or designing isoeffective schema. Such applications are much less sensitive to the values of LQ parameters than are absolute, de novo, predictions of TCP or NTCP. While it is important to be appropriately critical of the LQ model and its application to radiotherapy, it is equally important to recognize that it is the best model we have. It is a mechanistically based model of cell killing, with parameters that have a clear radiobiological interpretation, and there is a wealth of evidence that cell killing dominates radiotherapeutic response. Of course, the simplest form of the LQ model (equation 13.3) is not necessarily the most appropriate to apply. When repopulation is important, the LQ formalism can be appropriately modified [21]. If redistribution or reoxygenation is important, the LQ formalism can again be appropriately modified [22]. Similarly, if there is evidence that the LQ model is underpredicting survival at high doses, appropriate saturation-related modifications to the LQ formalism have been described [62,63]. A recent study [64] looked at the relationship of the LQ formalism to other commonly used radiobiological models, particularly in terms of their predicted time-dose relationships. It was shown that a broad range of radiobiological models is described by formalisms which result in the standard LQ relationship for dose fractionation/protraction, including the same generalized time factor, G (see equation 13.2). This approximate equivalence holds not only for the formalisms describing binary misrepair models, which are conceptually similar to LQ, but also for formalisms describing models embodying a very different explanation for time-dose effects, namely saturation of repair capacity. In terms of applications to radiotherapy, it was shown that a typical saturable repair formalism predicts practically the same dependencies for protraction effects as does the LQ formalism, at clinically relevant doses per fraction. Overall, use of the LQ formalism to predict dose-time
202 Radiobiology of HDR, LDR, and PDR brachytherapy
relationships is a notably robust procedure, depending less than previously thought on knowledge of detailed biophysical mechanisms, because various conceptually different biophysical models lead, in a reasonable approximation, to the LQ relationship including the standard form of the generalized time factor G.
GLOSSARY oc/P ratio The ratio of the parameters a and P in the linear-quadratic model (qv). a is considered to be a measure of the probability that a single radiation event will cause lethal cellular damage; (3 is a measure of the probability that two sublethal events will combine together to create lethal damage. In clinical radiotherapy, the value of a is the principal determinant of overall radiation sensitivity of a tumor or normal issue; the p component governs the fractionation and dose-rate characteristics of a particular cell line. In terms of the underlying cell survival curve, a is the initial slope and P is a measure of the downward 'bendiness' of the curve. Biologically effective dose (BED) A measure of biological effect as calculated by the LQ model. BED values are additive, and are therefore useful in calculating the overall effectiveness of treatments which consist of two or more components. The BED is tissue-specific as its value is dependent on the oc/P ratio and the recovery half-life. BED may be thought of as the dose necessary for a given effect when the treatment consists of an infinite number of infinitely small fractions. Early-responding tissues Those normal tissues in which the radiation damage is normally apparent within weeks or months of exposure; characterized by a relatively large oc/P ratio. High dose rate (HDR) A dose rate such that the time required to deliver a given dose is very short in comparison with the cellular recovery half-lives. Late-responding tissues Those normal tissues in which the radiation damage only becomes apparent months or years after exposure; characterized by a relatively small oc/P ratio. Linear-quadratic (LQ) model Cell-killing model based on the assumption that there are two components to cell killing: a linear component which is directly proportional to the delivered dose, and a quadratic component which is proportional to the square of the delivered dose. In the simplest case of a single instantaneous dose of magnitude dy the cell survival (S) is given by: S = exp(-ad - P^2). The model has been extensively developed to describe many other patterns of radiation delivery. Low dose rate (LDR) A dose rate such that the time required to deliver a given dose is long compared with the cellular recovery half-lives. Medium dose rate (MDR) A dose rate such that the
time required to deliver a given dose is comparable with the cellular recovery half-lives. Potential doubling time (Tpot) The predicted cell doubling time in the absence of cell loss. Pulsed dose rate (PDR) A technique whereby the biological effects associated with continuous LDR irradiation are simulated by the use of 'pulses' of HDR irradiation delivered at approximately 1 -h intervals over a long time period. Radiosensitivity A measure of the radioresponsiveness of a particular cell line. Possible measures of radiosensitivity are the surviving fraction after a single dose of 2 Gy (SF2), and the initial slope of the cell-survival curve (a). Recovery The process whereby the amount of injury to irradiated cells or tissues is able to reduce with time after irradiation. Recovery half-life The half-life which determines the (exponential) rate at which simple sublethal cellular injury may recover. Repopulation effect The concomitant growth of tumor clonogens during a course of radiotherapy. Repopulation factor (K) The daily dose necessary to offset tumor repopulation. In terms of the LQ model, K is defined as: K = 0.693/ccrpot. Sublethal damage That component of cellular injury which is capable of repair. When sublethal damage is accumulated within a given cell, lethal damage may result from an interaction between the sublethal components.
REFERENCES 1. Ellis, F. (1968) Dose, time and fractionation in radiotherapy. In Current Topics in Radiation Research, ed. M. Ebert and A. Howard. Amsterdam, North Holland Publishing Company, 359-97. 2. Barendsen, G.W. (1982) Dose fractionation, dose rate and isoeffect relationships for normal tissue responses. Int.J. Radial Oncol. Biol. Phys., 8,1981-97. 3. Fowler, J.F. (1989) The linear-quadratic formula and progress in fractionated radiotherapy. Br.J. Radiol., 62, 679-94. 4. Anderson, L.L. (1986) A'natural'volume-dose histogram for brachytherapy. Med. Phys., 13,898-903. 5. Saw, C.B. and Suntharalingham, N. (1988) Reference doserates for single-plane and double-plane iridium-192 implants. Med. Phys., 15,391-8. 6. Wu, A., Ulin, K. and Sternick, E.S. (1988) A dose homogeneity index for evaluating 192lr interstitial breast implants. Med. Phys., 15,104-7. 7. Meertens, H., Borger, J., Steggarda, M. and Blom, A. (1994) Evaluation and optimisation of interstitial brachytherapy dose distributions. In Brachytherapy: from Radium to Optimization, ed. R.F. Mould, J.J. Batterman, A.A. Martinez and B.L Speiser. Leersam, The Netherlands, Nucletron International, 300-6.
References 203 8. Lea, D.E. and Catcheside, D.G. (1942) The mechanism of the induction by radiation of chromosome aberrations in Tradescantia.J. Genet., 44,216-45. 9. Lea, D.E. (1938) A theory of the action of radiation on biological materials capable of recovery. I. The time intensity factor. Br.J. Radiol., 11,489-97. 10. Hall, E.J. (1994) Radiobiologyfor the Radiologist, 4th edn. Philadelphia, J.B. LippincottCo. 11. Thames, H.D. and Hendry, J.H. (1987) Fractionation in Radiotherapy. London, Taylor and Francis. 12. Dale, R.G. (1990) The potential for radiobiological modelling in radiotherapy treatment design. Radiother. Oncol., 19,245-55. 13. Dale, R.G. (1985) The application of the linear-quadratic dose-effect equation to fractionated and protracted radiotherapy. Br.J. Radiol., 58, 515-28. 14. Brenner, D.J., Huang, Y. and Hall, E.J. (1991) Fractionated high dose-rate versus low dose-rate regimens for intracavitary brachytherapy of the cervix: equivalent regimens for combined brachytherapy and external irradiation. Int.]. Radial. Oncol. Biol. Phys., 21,1415-23. 15. Yaes, R.J., Patel, P. and Maruyama, Y. (1991) On using the linear-quadratic model in daily clinical practice. Int. J. Radial Oncol. Biol. Phys., 20,1353-62. 16. Munro, T.R. and Gilbert, C.W. (1961) The relation between tumour lethal doses and the radiosensitivity of tumour cells. Br.J. Radiol., 34,246-51. 17. Hendry, J.H. and Thames, H.D. (1986) The tissue rescuing unit. Br.J. Radiol., 59,628-30. 18. Withers, H.R., Thames, H.D., Flow, B.L, Mason, H.A. and Hussey, D.H. (1978) The relationship of acute to late skin injury in 2 and 5 fraction/week gamma-ray therapy. Int.]. Radial Oncol. Biol. Phys., 4, 595-601. 19. Withers, H.R., Taylor, J.M.G. and Maciejewski, B.(1988) The hazard of accelerated tumor clonogen repopulation during radiotherapy. Acta Radiol., 27,131-46. 20. Fowler, J.F. and Lindstrom, M.J. (1992) Loss of local control with prolongation of treatment. Int.]. Radial Oncol. Biol. Phys., 23,457-67. 21. Travis, E.L. and Tucker, S.L. (1987) Iso-effect models and fractionated radiotherapy. Int.]. Radial Oncol. Biol. Phys., 13,283-7. 22. Brenner, D.J., Hlatky, L.R., Hahnfeldt, P.J., Hall, E.J. and Sachs, R.K. (1995) A convenient extension of the linearquadratic model to include redistribution and reoxygenation. Int.]. Radial Oncol. Biol. Phys., 32, 375-90. 23. Hlatky, L.R., Hahnfeldt, P. and Sachs, R.K. (1994) The influence of time-dependent, stochastic heterogeneity on the radiation response of a cell population. Math. Biosci., 122,201-20. 24. Dale, R.G. and Jones, B. (1994) The effect of tumour shrinkage on biologically effective dose, and possible implications for fractionated high dose rate brachytherapy. Radiother. Oncol., 33,125-132. 25. Dale, R.G., Jones, B. and Coles, I.P. (1994) The effect of tumour shrinkage on the biological effectiveness of
26.
27.
28. 29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
permanent brachytherapy implants. Br.J. Radiol., 67, 639-47. Barendsen, G.W. and Broerse, J.J. (1970) Experimental radiotherapy of rat rhabdomyosarcoma with 15 MeV neutrons and 300 kV X-rays. Effects of fractionated treatments. Eur.J. Cancer, 6,89-109. Smit, B.J., du Toit, J.P. and Groenwald, W.A. (1989) An indwelling intrauterine tube to facilitate intracavitary radiotherapy of carcinoma of the cervix. Br. J. Radiol., 62, 68-9. Fowler, J.F. (1992) Rapid repopulation in radiotherapy: debate on mechanism. Radiother. Oncol., 23,126. Bleasdale, C. and Jones, B. (1994) Mathematical model for the time interval between external beam radiotherapy and brachytherapy. In Brachytherapy from Radium to Optimization, ed. R.F. Mould, J.J. Batterman, A.A. Martinez and B.L. Speiser. Leersum, The Netherlands, Nucletron International BV, 39-48. Jones, B. and Bleasdale, C. (1994) Effect of overall time when radiotherapy includes teletherapy and brachytherapy: a mathematical model. Brit.}. Radiol., 67, 63-70. Jones, B., Kingston, R., Green, J., Adams, E. and Bleasdale, C. (1993) The effect of treatment duration in the local control of cervix cancer. Radiother. Oncol., 28,179-80. Speiser, B.L. and Spratling, L. (1994) Remote afterloading brachytherapy for the local control of endobronchial carcinoma. In Brachytherapy from Radium to Optimization, ed. R.F. Mould, J.J. Battermann, A.A. Martinez and B.L. Speiser). Leersum, The Netherlands, Nucletron International BV, 180-95. Dale, R.G. (1990) The use of small fraction numbers in high dose-rate gynaecological afterloading: some radiobiological considerations. Br.]. Radiol., 63,290-4. Brenner, D.J. and Hall, E.J. (1991) Fractionated high dose rate versus low dose rate regimens for intracavitary brachytherapy of the cervix. I. General considerations based on radiobiology. Br.}. Radiol., 64,133-41. Orton, C.G. (1993) High dose rate versus low dose rate brachytherapy for gynecological cancer. Semin. Radial Oncol., 3,232-9. Deehan, C. and O'Donoghue, J.A. (1994) Biological equivalence of LDR and HDR brachytherapy. In Brachytherapy from Radium to Optimization, ed. R.F. Mould, J.J. Battermann, A.A. Martinez and B.L. Speiser. Leersum, The Netherlands, Nucletron International BV. Martel, M.K., McShan, D.L and Fraass, B.A. (1995) Optimization based on dose-volume information for GYN applications. Med. Phys., 22,942. Schoeppel, S.L, LaVigne, M.L, Martel, M.K., McShan, D.L, Fraass, B.A. and Roberts, J.A. (1994) Three-dimensional treatment planning of intracavitary gynecological implants: analysis of ten cases and implications for dose specification. Int.}. Radial Oncol. Biol. Phys.,28, 277-83. Niemierko, A. and Goitein, M. (1991) Calculation of normal tissue complication probability and dose-volume
204 Radiobiology of HDR, LDR, and PDR brachytherapy
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
histogram reduction schemes for tissue with a critical element architecture. Radiother. Oncol., 20,161 -76. Hall, E.J. and Brenner D.J. (1994) Sublethal damage repair rates: a new tool for improving therapeutic ratios? InLJ. Radial Oncol. Biol. Phys., 30,241-2. Brenner, D.J. and Hall, E.J. (1991) Conditions for the equivalence of continuous to pulsed low dose rate brachytherapy. Int.}. Radial Oncol. Biol. Phys.,20, 181-90. Van der Kogel, A.J. (1985) Chronic effects of neutrons and charged particles on spinal cord, lung, and rectum. Radial Res., 104, S208-16. Douglas, B.G. and Fowler, J.F. (1976) The effect of multiple small doses of x rays on skin reactions in the mouse and a basic interpretation. Radial Res., 66,401 -26. Tucker, S.L. (1984) Tests for the fit of the linear-quadratic model to radiation isoeffect data. InlJ. Radial Oncol. Biol. Phys.,!0,1933-9. Levendag, P., Visser, A. and Jansen, P. (1994) Fractionated brachytherapy in cancer of the head and neck, Radiother. 0«co/.,31(Suppl.1),S35. Swift, P.S., Fu, K.K., Phillips, T.L, Roberts, LW. and Weaver, K.A. (1994) Pulsed LDR interstitial and intracavitary therapy, clinical experience at UCSF, San Francisco, USA. Activity (International Nucletron Radiotherapy Journal) Report #5,29-31. de Pree, C, Popowski, Y., Weber, D., Nouet, P., Rouzaud, M. and Kurtz, J.M. (1999) Feasibility and tolerance of pulsed dose rate interstitial brachytherapy. InlJ. Radial Oncol. Biol. Phys., 43,971-6. Fritz, P., Berns, C., Anton, H.W. etal. (1997) PDR brachytherapy with flexible implants for interstitial boost after breast-conserving surgery and external beam radiation therapy. Radiother. Oncol., 45,23-32. Brenner, D.J., Schiff, P.B., Huang, Y. and Hall, E.J. (1997) Pulsed-dose-rate brachytherapy: design of convenient (daytime-only) schedules. InlJ. Radial Oncol. Biol. Phys., 39,809-15. Fowler, J.F. and Mount, M. (1992) Pulsed brachytherapy: the conditions for no significant loss of therapeutic ratio compared with traditional low dose rate brachytherapy. InlJ. Radial Oncol. Biol. Phys., 23,661-9. Armour, E., Wang, Z.H., Corry, P. and Martinez, A. (1992) Equivalence of continuous and pulse simulated low dose rate irradiation in 9L glioblastoma cells at 37 degrees and 41 degrees C. InlJ. Radial Oncol. Biol. Phys., 22,109-14. Chen, C.-Z., Huang, Y., Hall, E.J. and Brenner, D.J. (1997) Pulsed brachytherapy as a substitute for continuous low dose rate: an in vitro study with human carcinoma cells. InlJ. Radial Oncol. Biol. Phys., 37,137-43.
53. Mason, K.A., Thames, H.D., Ochran, T. and Janjan, N. (1994) Comparison of continuous and pulsed low-dose rate brachytherapy: biological equivalence in vivo. InlJ. Radial Oncol. Biol. Phys., 28,667-71. 54. Fowler, J.F. (1993) Why shorter half-times of repair lead to greater damage in pulsed brachytherapy. InlJ. Radial Oncol. Biol. Phys., 26,353-6. 55. Fowler, J.F. and Van Limbergen, E.F. (1997) Biological effect of pulsed dose rate brachytherapy with stepping sources if short half-times of repair are present in tissues. InlJ. Radial Oncol. Biol. Phys., 37,877-83. 56. Fowler, J.F. (1995) Are half-times of repair reliably shorter for tumors than for late normal-tissue effects? InlJ. Radial Oncol. Biol. Phys., 31,189-90. 57. Millar, W.T., Hendry, J.H. and Canney, P.A. (1996) The influence of the number of fractions and bi-exponential repair kinetics on biological equivalence in pulsed brachytherapy. Br.J. Radial., 69,457-68. 58. Brenner, D.J., Hall, E.J., Huang, Y. and Sachs, R.K. (1995) Potential reduced late effects for pulsed brachytherapy compared with conventional LDR. InlJ. Radial Oncol. Biol. Phys., 31,201 -2. 59. Brenner, D.J., Hall, E.J., Randers-Pehrson, G. etal. (1996) Quantitative comparisons of continuous and pulsed low dose-rate regimens in a model late-effect system. InlJ. Radial Oncol. Biol. Phys., 34,905-10. 60. Brenner, D., Armour, E., Corry, P. and Hall, E. (1998) Sublethal damage repair times for a late-responding tissue relevant to brachytherapy (and external-beam radiotherapy): implications for new brachytherapy protocols. InlJ. Radial Oncol. Biol. Phys.,41,135-8. 61. Armour, E.P., White, J.R., Armin, A. etal. (1997) Pulsed low dose rate brachytherapy in a rat model: dependence of late rectal injury on radiation pulse size. InlJ. Radial Oncol. Biol. Phys., 38,825-34. 62. Zaider, M. and Rossi, H.H. (1980) Saturation effects for sparsely ionizing particles. In Radiation Physics, Biophysics and Radiation Biology, Radiological Research Laboratory Annual Report COO-4733-3. New York, Columbia University, 126-34. 63. Hall, E.J. and Brenner, D.J. (1993). The radiobiology of radiosurgery: rationale for different treatment regimes for AVMsand malignancies. InlJ. Radial Oncol. Biol. Pfcys.,25,381-5. 64. Brenner, D.J., Hlatky, L.R., Hahnfeldt, P.J., Huang, Y. and Sachs, R.K. (1998) The linear-quadratic model and most other common radiobiological models result in similar predictions of time-dose relationships. Radial Res., 150, 83-91.
14 Predictive assays for radiation oncology JOHN A. COOK AND JAMES B.MITCHELL
14.1
INTRODUCTION
In 1895, Wilhelm Roentgen contributed to the medical profession perhaps one of its most widely used and beneficial diagnostic tests. With the discovery of the X-ray and the rapid development of radiographs, Roentgen's discharge tube enabled physicians non-invasively to visualize anatomical structures quickly and establish sound diagnosis of a variety of medical problems. Diagnostic tests used in medical practice have progressed a long way since the early 1900s, both in type and sophistication. An elaborate array of tests now greatly aids the physician in making a rational diagnosis of a particular medical problem. Expeditious diagnosis is extremely important in the field of oncology and, unquestionably, if certain tumors are detected early enough, successful treatment and eradication of the tumor can be achieved. Unfortunately, not all tumors are detected early and, despite there being a wide variety of treatment options available, including surgery, chemotherapy, and radiation therapy, successful eradication of tumor with acceptable normal-tissue toxicity remains a major challenge to the practicing oncologist. Since the 1980s, radiation oncologists and biologists have recognized the need for additional assays on an individual patient basis that would select the most advantageous treatment approach [1]. We emphasize assays for individual patients for several reasons. First, the cellular radiation sensitivity of the tumor may differ among individuals, even for tumors of the same histological type. If the radiosensitivity of the individual's
tumor were precisely known, perhaps total radiation doses could be adjusted before the end of therapy to maximize tumor response. Alternatively, the option of using radiation sensitizers for 'radioresistant' tumors would have a more rational basis. Second, normal-tissue radiation sensitivity may differ among individuals. This is an important point because the total radiation dose that can be delivered to a patient's tumor is often limited by normal tissue tolerance. Stated differently, frequently radiation oncologists are compelled to treat a patient's tumor with radiation doses that are dictated not by tumor sensitivity but by normal-tissue tolerance, which in many instances results in inadequate dose to the tumor. If one assumes there is a Gaussian distribution of normal-tissue radiosensitivities among humans, then the most sensitive individuals in the population may well dictate radiation tumor doses utilized in the clinic. Because the radiation tumor control dose response curve is quite steep for many tumors, modest increases in the total radiation dose delivered would be expected greatly to enhance tumor control. If it were determined that the patient's normal-tissue radiation response were toward the 'radioresistant' edge of the Gaussian distribution, consideration could be given to administering higher radiation doses. Alternatively, if the patient's normal-tissue radiation response were toward the 'radiosensitive' edge of the Gaussian distribution, the use of radioprotectors could be considered. Unfortunately, selective normal-tissue radioprotectors have yet to be identified. Third, biological, environmental, and physiological factors of tumors may differ among individuals. Factors such as tumor pH, hypoxia, blood flow, and
206 Predictive assays for radiation oncology
growth of the tumor in terms of cell-cycle parameters and potential tumor doubling times (Tpot) can influence the overall radiation responsiveness of the tumor. If these factors were known prior to therapy, the use of hypoxic cell radiosensitizers or, in the case of Tpot values, alteration of fractionation/time schedules could be considered. Numerous predictive assays have been developed over the past two decades to address many of the points cited above and several have been evaluated in a clinical setting. This chapter briefly reviews the current status of several different predictive assays and discusses their advantages and shortcomings. These assays, while evaluated on patients receiving external-beam radiotherapy, are also highly relevant for patients receiving various forms of brachytherapy.
14.2 REQUIREMENTS OF CLINICALLY USEFUL PREDICTIVE ASSAYS A number of diagnostic tests already aid the oncologist in designing the course of treatment. These include: tumor type, histological grade, tumor biochemical markers, size and anatomical location of the tumor (which can be determined by various X-ray procedures), rate of tumor growth, receptor status, ploidy of the tumor cells, and patient performance status and age [2]. A major advantage of these tests is that they can be performed rapidly and are available when options for treatment are considered. The tests have proven to be predictive for both tumor responsiveness to therapy and ultimate survival of the patient. Ideally, predictive assays, particularly those for radiation oncology, should be rapid (ideally within a week) and predictive with low false negativity [3].
type [5]. As an example, cell lines isolated from glioblastoma tumors exhibit a broad range of SF2Gy values (0.2-0.9) [5], yet glioblastoma uniformly respond poorly to radiation. If heterogeneity in cellular radiosensitivity exists in human tumors of the same histological type, the need for an accurate individualized assessment of cellular radiosensitivity becomes extremely important if altered treatment approaches are to be considered in the clinic based on predictive assays. Raaphorst has pointed out that the reproducibility of SF2C}, determinations of an established cell line can be quite variable [6]. Using the radiosensitivity parameter of SF2Gy means that cell killing is usually confined to the first log of survival. Low doses of radiation which result in low levels of cell killing are difficult to resolve statistically due to the summation of statistical errors inherent in conducting survival assays for low radiation doses [7]. Thus, concerns relating to tumor-cell radiosensitivity assessments from a patient's tumors include: (a) whether the radiosensitivity of cells taken from the biopsy sample is representative of the entire tumor; (b) whether the single determination of radiosensitivity is so variable that it would not be useful; and (c) whether radiosensitivity would remain constant during a full course of 25-30 fractions of radiotherapy. While such concerns are reasonable (and perhaps sobering) with regard to tumor-cell biology, it is difficult to second-guess results until experiments are conducted. However, the studies discussed above formed the basis to explore the possibility of determining the inherent cellular radiosensitivity of primary cultures of tumor cells taken directly from the patient. There are several assays available to assess cellular radiosensitivity [8-17], but only those for which most clinical data have been obtained are discussed below.
143*2 Cell adhesive matrix assay 143
SURVIVAL ASSAYS
143*1 Tumor-cell radiosensitivity Intrinsic cellular radiosensitivity of cell lines established from patients' tumors has provided interesting correlations with respect to the radioresponsiveness of specific tumor types in the clinic. Fertil and Malaise, in an analysis of published radiation survival curves, found that cell lines derived from, for example, glioblastoma and melanoma exhibit a high surviving fraction at 2 Gy (SF2Gy) [4]. These tumor types are less responsive to radiotherapy than, for example, tumors such as lymphomas and small lung cancer, which exhibit low SF2Gy values [4]. While there is a semi-qualitative relationship between SF2Gy values of given tumor types and clinical radioresponsiveness, a range of radiosensitivities has been noted in cell lines derived from the same tumor
The cell adhesive matrix (CAM) assay is conducted on tissue culture dishes coated with a substance that facilitates cell attachment and cell growth [18]. Single-cell suspensions of tumor cells are plated onto CAM dishes, incubated for 24 h, and then exposed to varying doses of radiation (1-6 Gy). After a growth period of 10-14 days, the cell monolayer is stained with crystal violet. The density of cell monolayer is assessed by the amount of stain taken up by the surviving cells as recorded by image analysis techniques. By comparing the density of staining of unirradiated cultures to that of irradiated cultures, growth-rate-based cell survival curves can be obtained [18]. Using this assay, Brock et al, determined SF2Gy values of cells derived from biopsies of head/neck tumors prior to radiation therapy [19]. Figure 14.1 shows the cumulative frequency histograms of SF2Cy values for 72 patients evaluated [19]. The range in SF2Gy values in patients achieving local control was from 0.10 to 0.91.
Survival assays 207 Figure 14.1 Cumulative frequency histograms of SFXy values for 72 head and neck cancer patients (patients controlled locally + patients who failed locally) compared to patients with local recurrences. (Redrawn with permission from
Image Not Available
The plot also displays SF2Gy values from 12 patients with local recurrence following radiation treatment. SF2Gy values from this subset of patients ranged from 0.20 to 0.91. Note that patients' tumors with low and high SF2Gy values recurred. The authors concluded that SF2Gy values were not suitable prognostic indicators for this set of patients [ 19]. Preliminary data from Girinsky, et a/., who used the same assay to evaluate head/neck tumor-cell radiosensitivity, reached similar conclusions for SF2Gy values; however, using the calculated a values of the radiation survival curves, a correlation was found between patients with an a value > 0.07 Gy ' and local control [20]. That there was no correlation between tumor sensitivity and clinical response is particularly troublesome, because the assay is both simple and reasonable. A technical difficulty inherent in the CAM assay is the observation that normal host fibroblasts (included in the tumor biopsy) can also grow on the matrix-treated dishes, thus potentially complicating the interpretation of tumor-cell radiosensitivity [20].
1433
Courtenay-Mills soft agar assay
This clonogenic assay is conducted by plating single-cell suspensions from patients' tumors into medium containing soft agar [21]. Standard clonogenic radiation survival curves are conducted from which survival curve parameters such as SF2Gy can be determined. From the time of the initial biopsy to the evaluation of the radiation survival curve is approximately 4 weeks. Using the assay, West et a/., determined SF2Gy values of tumor cells taken from 88 patients with cervical carcinoma [22]. In
reference 19.)
the study, a significant correlation was found between SF2Gy values and both local control and survival, as shown in Figure 14.2. Patients with SF2Gy values > 0.4 had both a lower local control rate and a lower survival when compared to patients with SF2Gy values < 0.4 [23]. 5F2Gy values were independent of disease stage, tumor grade, and patient age [22]; however, there was a loose correlation with tumor volume and diploid status, which suggests that SF2Gy alone may not be a completely independent predictor of local control and survival. As shown for the CAM assay, fibroblasts have also been shown to grow in this soft agar assay [24]. Of all the cell-survival predictive assays presently under evaluation, the data from the Courtenay-Mills assay clearly provide the most encouragement. It will be most interesting to determine if SF2Gy values determined by this assay correlate with local control and survival for different tumors. Likewise, it would be informative to determine if the CAM assay applied to cervical carcinoma would give similar results to the Courtenay-Mills assay. A potential improvement that could be applied to both assays might be to use fractionated radiation (i.e., five 2-Gy fractions) to maximize the importance of repair, which is not apparent for a single 2-Gy dose of radiation [6]. Both assays require several weeks to obtain results. This represents a possible disadvantage, particularly if radiation sensitizers or protectors were to be considered an option as treatment commences. However, both assays can be completed before the end of therapy, thus permitting total tumor dose modification. The finding that host fibroblasts grow in both assays raises concern as to their contribution to the overall radiosensitivity
208 Predictive assays for radiation oncology
Image Not Available
Figure 14.2 Local control and survival probability as a function of time after treatment of cervical carcinoma patients. Local control and survival were further separated by examining patients with SF2Gy values greater than or less than 0.4. (a) Local control probability as a function of time after radiation, (b) Survival probability as a function of time after radiation. (Redrawn with permission from reference 23.)
assessment, although such a concern would only be important if tumor-cell and fibroblast radiosensitivities differ.
143*4 Normal-tissue cellular radiosensitivity Is there a correlation between the radiosensitivity of normal tissues and tumor cells from the same animal or patient? The answer to this question is not known for humans. If the answer were yes, then the determination of normal-cell radiosensitivity might be easier to perform and more reliable than tumor-cell radiosensitivity assessments, given that a homogeneous population (with respect to DNA content or chromosome stability) of cells from readily accessible normal tissues could be studied. Data from at least one animal model suggest that normal-tissue and tumor radiosensitivities are indeed similar. Budach et al. showed that tumors arising in severe combined immunodeficient (SCID) mice exhibited the same radiosensitivity as their normal skin fibroblasts [25]. The study, however, represents an extreme example, in that SCID mice are approximately threefold more radiosensitive than normal mice. In recent human studies, however, Stausbol-Gron et al. examined the SF2Gy of both fibroblasts and tumor cells from 71 head and neck patients and found no statistical correlation between the fibroblast and tumor SF2Gy [26]. The question as to whether there is a similarity of radiosensitivities between tumor and normal tissues in 'normal' mice and humans remains unanswered.
Another question that warrants consideration is whether normal-cell radiosensitivity (fibroblasts, lymphocytes, etc.) correlates with radiation-induced normal-tissue complications. Several studies have attempted to address this question [27-33]. Geara et al. evaluated the radiosensitivity of fibroblasts taken from 21 patients with head/neck tumors who subsequently received radiation therapy [34]. Fibroblast SF2Gy values were compared to the acute and late effects of skin and oral mucosa during and after radiation treatment, as shown in Figure 14.3 [34]. SF2Gy values correlated well with late normal-tissue reactions (Figure 14.3a). That is, patients whose fibroblast radiosensitivity was characterized by high SF2Gy values exhibited fewer late reactions. Likewise, low SF2Gy values correlated with more severe late effects. No correlation was found between SF2Gy values and acute reactions (Figure 14.3b). Russel et al. studied the sensitivity of dermal fibroblast from 79 breast cancer patients and attempted to correlate the SF2Gy with the degree of breast fibrosis [33]. Although there was significant variation in intrinsic radiation sensitivity, there was only a weak correlation between the SF2Gy and breast fibrosis. Other studies have suggested that acute effects may correlate with fibroblast radiosensitivity, but these studies are preliminary and require further verification [35]. Despite differences, collectively the studies suggest that radiation-induced,'late' normal-tissue reactions correlate to some extent with fibroblast radiosensitivity [32]. It is hoped that study of larger populations of patients will confirm and extend the preliminary findings to other tissues and organs at risk for radiation damage. Recently, the gene for ataxia telangiectasia (AT) has
Oxygen measurements and tumor response 209
Image Not Available
Figure 14.3 SF2Gy values as a function of the grade of either late or acute normal-tissue reactions, (a) Late reactions scored after the radiation treatments, (b) Acute reactions of the skin and mucosa scored weekly during the radiation treatments. (Redrawn with permission from reference 34.)
been cloned [36]. Fibroblast cultures derived from homozygous AT patients have long been known to exhibit extreme radiosensitivity [37]. Likewise, ataxia telangiectasia patients also exhibit severe normal-tissue reactions if given a standard course of radiotherapy [38]. Dunst et al. examined chromosome breaks in lymphocytes from cancer patients undergoing radiotherapy. This study included three individuals who had proven ataxia telangiectasia (AT homozygotes) [31]. It found a higher number of chromosomal breaks in the lymphocytes from patients who had extreme normal-tissue reactions to the radiation, while the ataxia patients had the highest break frequency of all. There is also evidence that ATheterozygotes exhibit slightly increased radiosensitivity [39] compared to fibroblasts from normal individuals. What makes the observation clinically important is that approximately 18% of patients with breast cancer may be heterozygous for the AT gene [40]. Application of molecular techniques may allow for the rapid identification of AT heterozygotes, thus allowing for a more precise attempt to correlate radiosensitivity and normaltissue damage. In the future, as more insight into human genetics and radiation responsiveness is gained, it maybe feasible to screen quickly for selective markers.
14.4 OXYGEN MEASUREMENTS AND TUMOR RESPONSE Inherent cellular radiosensitivity may not necessarily correlate with clinical radioresponsiveness because the tumor microenvironment may influence radiation response. A major modifier of the radiation response is
molecular oxygen [41]. Cells exposed to X-rays at low oxygen levels are more resistant than fully oxygenated cells by a factor of about three. Should hypoxic cells exist in tumors, they might pose a potential obstacle to successful tumor eradication. The ability to determine if hypoxic cells are present in human tumors has been facilitated by the use of sensitive oxygen electrodes. These tiny glass electrodes are inserted directly into the tumor and many oxygen measurements are made as the electrode is mechanically moved along a track of tissue (2-40 mm). The procedure can be done several times through different parts of the tumor, does not cause pain or discomfort to the patient, and can be completed in approximately 1 h. Oxygen levels in normal tissue can also be determined for comparison. Disadvantages of the technique include the invasive nature of the procedure and the inaccessibility of some tumors. Several studies using the oxygen-sensitive electrode to measure PO2 levels in head and neck carcinomas, cervical carcinomas, breast carcinomas, soft tissue sarcomas, and in squamous cell carcinoma metastases have been reported [42-46]. In squamous cell carcinoma metastases prior to radiation treatment, Gatenby et al. showed a relationship between tumor response and tumor PO2 levels [43]. Patients who achieved a complete response to radiation therapy had higher PO2 levels (< 26% of the tumor volume measuring < 8 mm PO2) in their tumors than did patients who did not respond (> 26% of the tumor volume measuring < 8 mm PO2). More recent clinical experience with oxygen electrodes has confirmed and extended the observations by Gatenby. Vaupel and Hockel published a series of studies in which PO2 levels were measured in patients with breast and cervix cancer and related to survival and/or recurrence-free survival
210 Predictive assays for radiation oncology
[44,45]. Examples from their studies are shown in Figures 14.4 and 14.5. Figures 14.4a and b showPO2 distributions for normal breast tissue (N = 16) and breast tumors (N = 15, Stage T1-T4), respectively [45]. The median PO2 level in normal breast was 65 mmHg compared to 30 mmHg for breast tumors. Notice that for breast tumors there was a significant number of values < 10 mmHg, whereas normal breast tissue had no measurements < 10 mmHg. A level of oxygen < 10 mmHg is within the range in which the oxygen enhancement ratio (OER) changes from a value of 1 (aerobic) to a maximum of 3 (hypoxic). Thus, values < 10 nimHg might be considered 'hypoxic.' The PO2 profiles shown in Figures 14.4a and b are mean values for a group of patients. Interestingly, PO2 profiles of individual breast cancer patients appear to have normal PO2 readings (Figure 14.5a) and yet other profiles that contain significant hypoxic readings < 10 mmHg (Figure 14.5b) [44]. Adams et al. examined the PO2 profiles of 37 head and neck cancer patients and also found a significant degree of hypoxia in these tumors as compared to subcutaneous PO2 readings [42]. Dunst et al. examined 49 cervical carcinoma patients both before and during radiotherapy
Image Not Available
0
100
Figure 14.5 P02 profiles from two individual breast cancer patients. In both panels, n = number of oxygen measurements made. (Redrawn with permission from reference 45.)
Image Not Available
Figure 14.4 P02 profiles from breast tumors and normal breast tissue. (A) P02 profiles of normal breast tissue (N = 16). (B) P02 profiles of breast tumors (M = 15). In both panels, n = number of oxygen measurements made. (Redrawn with permission from reference 45.)
[47]. They found PO2 changes occurred during radiotherapy, with a majority of those patients with either a pretreatment median PO2 > 10 mmHg or a median PO2 > 10 mmHg at 19.8 Gy having complete response to the radiation treatments. The example serves to reinforce the importance of individualized assessment of hypoxia in human tumors, particularly in the context of evaluating or considering the use of hypoxia-cell radiosensitizers. Figure 14.6 shows survival curves of patients with cervical carcinoma whose median PO2 profiles were either greater than or less than 10 mmHg. These preliminary data clearly show that survival is enhanced for those patients whose tumors had median PO2 profiles > 10 mmHg prior to treatment. This was true for patients receiving radiotherapy alone as well as for those who received surgery, chemotherapy, or combined therapies. The data, while of interest to radiation oncologists, are also of interest to tumor biologists, in that patients with tumors with PO2 profiles < 10 mmHg did not respond well to any therapy. The reason for the response profile is not clear, but it indicates that hypoxia is a marker of aggressive, poorly responsive tumors [48]. Likewise, the presence of hypoxia in soft tissue sarcomas may be predictive for metastatic potential [46]. Several other techniques to measure tissue PO2 levels
Cell-cycle analysis and tumor response 211
Image Not Available
Image Not Available
Figure 14.6 Survival probability of cervical carcinoma patients as a function of time after treatment (either radiation, surgery, or chemotherapy alone, or a combination of these treatments). Survival was separated based on whether the median ?02 was
Figure 14.7 Actuarial survival of head and neck patients as a function of time after treatment (either accelerated fractionation or conventional fractionation). Survival was
either less than or greater than 10 mmHg. (Redrawn with
(slow). (Redrawn with permission from reference 60.)
separated based upon a Jpot < 4 days (fast) or a T^, > 4 days
permission from reference 44.)
are currently under development [49]. Oxygen electrode measurements do not yield the complete PO2 profile of the entire tumor. Ideally, it would be advantageous to have a technique that could provide non-invasive assessment of PO2 levels for the entire tumor. The use of nitroimidazoles, which bind to macromolecules under hypoxic conditions [50,51 ], is one approach toward noninvasive imaging of hypoxic tissue. The nitroimidazole can be labeled with radioactive iodine-123 and evaluated by single photon emission computed tomography (SPECT) [52], labeled with an isotope of fluorine (fluorine-18) and evaluated by positron emission tomography (PET) [53] or detected histochemically [54]. Other techniques include the use of electron paramagnetic resonance (EPR) coupled with free radical probes [55,56] or India ink [57] and fluorine (fluorine-19) magnetic resonance spectroscopy [58]. These techniques hold great promise, but are still in the early stages of development.
14.5 CELL-CYCLE ANALYSIS AND TUMOR RESPONSE Because tumors are known to grow at different rates, it might be beneficial to have an assay to access the potential doubling time (Tpot) of the tumor. Such an assay has the potential to identify those patients with rapidly growing tumors who might benefit from accelerated or hyperfractionated radiation treatment. Technology has rapidly advanced over the past few years, making estimation of the rpot of a patient's tumors relatively simple and straightforward. The assay involves bolus injection of a halogenated pyrimidine (bromodeoxyuridine or iodo-
deoxyuridine). Over the next 2-4 h the halogenated pyrimidine is incorporated into the tumor-cell DNA. The tumor is biopsied and a single-cell suspension is prepared. The cells are then treated with a fluorescentlabeled antibody that recognizes the specific halogenated pyrimidine incorporated in the DNA. Following this treatment, the cells are analyzed by flow cytometry. From the DNA flow cytometry histogram, an estimate can be made of the rpot of the tumor [59]. Recent clinical studies have shown that local tumor control (using conventional fractionation) in patients with tumors with Tpot values < 4 days was significantly worse than that for tumors with Tpot values > 4 days, as shown in Figure 14.7 [60]. However, patients whose Tpot values were < 4 days and received accelerated fractionation achieved local control comparable to that of patients with Tpot values > 4 days who received either conventional or accelerated fractionation. Similar findings have been reported by other investigators [61]. In contrast, a recent study with 74 patients with head and neck tumors failed to show a correlation between Tpot values and local regional control using conventional radiotherapy [62]. These studies serve to demonstrate how important Tpot determinations might be in identifying candidates for altered fractionation schedules and perhaps high or low dose-rate brachytherapy. A cautionary note must be sounded because the number of patients analyzed in this fashion has been small, and verification with larger patient populations and different tumor types is needed. In addition, relying on a single biopsy for determination of Tpot values may be misleading. A recent study showed that when five biopsies were taken from individual esophageal tumors and rpot values determined, heterogeneity in T values were obtained [63].
212 Predictive assays for radiation oncology
14.6 CONCLUSIONS
clonogenic assay with chromosome aberrations scored using premature chromosome condensation with
Predictive assays are far from being incorporated into the routine radiation treatment decision-making process. Yet progress has been made and indeed much can be learned about tumor and normal-tissue biology and physiology through the development of such assays. Perhaps what will evolve from the cited initial studies is a battery of predictive assays that, when used together, will aid the radiation oncologist better to individualize treatment. There is no doubt that molecular techniques will aid ultimately in identifying new markers that will facilitate accurate and individualized predictive assays. There are indications that predictive assays have the potential to revolutionize the way in which the radiation oncologist will approach patient treatment. What is needed now is continued support by the radiation community of research and development that will lead to effective and practical assays.
fluorescence in situ hybridization. Int.J. Radiat. Oncol. Biol. Phys., 30,1127-32. 10. Olive, PL and Durand, R.E. (1992) Detection of hypoxic cells in a murine tumor with the use of the comet assay. J. Natl Cancer Inst., 84,707-11. 11. Fairbairn, D.W., Olive, PL and O'Neil, K.L (1991) The comet assay: a comprehensive review. Mutat. Res., 339,37-59. 12. Olive, PL, BanathJ.P.and MacPhail, H.S. (1994) Lack of a correlation between radiosensitivity and DNA doublestrand break induction or rejoining in six human tumor cell lines. Cancer Res., 54,3939^6. 13. Jones, L.A., Clegg, S., Bush, C., McMillan, T.J. and Peacock, J.H. (1994) Relationship between chromosome aberrations, micronuclei, and cell kill in two human tumour cell lines of widely differing radiosensitivity. Int. J. Radiat. Oncol. Biol. Phys., 66,639^2. 14. Bakker, P.J., Tukker, LJ., Stap, J., Veenhof, C.H. and Aten, J.A. (1993) Micronuclei expression in tumors as a test for radiation sensitivity. Radiother. Oncol., 26,69-72.
REFERENCES
15. Bush, C. and McMillan, T.J. (1993) Micronudeus formation in human tumour cells: lack of correlation with
1. Chapman, J.D., Peters, LJ. and Withers, H.R.(1989) Prediction of Tumor Treatment Responses, New York,
radiosensitivity. Br.J. Cancer, 67,102-6. 16. Wuttke, K., Streffer, C. and Muller, W.U. (1993) Radiation induced micronuclei in subpopulationsof human
Pergamon Press. 2. DeVita, V.T., Hellman, S. and Rosenberg, S.A. (1993)
lymphocytes. Mutat. Res., 286,181-8. 17. Wasserman, T.H. and Twentyman, P. (1988) Use of a
Cancer, Principles and Practice of Oncology, Philadelphia, J.B. LippincottCo. 3. Suit, H.D. and Walker, A.M. (1985) Predictors of radiation response in use today: criteria for new assays and methods of verification. In Prediction of Tumor Treatment Response, ed. J.D. Chapman, LJ. Peters and H.R. Withers. New York, Pergamon Press, 3-19. 4. Fertil, B. and Malaise, E.P. (1985) Intrinsic radiosensitivity of human cells lines is correlated with radioresponsiveness of human tumors: analysis of 101 published survival curves. Int.J. Radiat. Oncol. Biol. Phys., 11,1699-707. 5. Allalunis-Turner, M.J., Barron, G.M., Day, R.S., Fulton, D.S. and Urtasun, R.C. (1992) Radiosensitivity testing of human primary brain tumor specimens. Int.J. Radiat. Oncol. Biol. Phys., 23,339-43. 6. Raaphorst, G.P. (1993) Prediction of radiotherapy response using SF2: is it methodology or mythology? Radiother. Oncol., 28,187-8. 7. Al per, T. (1975) Cell Survival after Low Doses of Radiation: Theoretical and Clinical Implications. London, John Wiley and Sons. 8. Brown, J.M., Evans, J. and Kovacs, M.S. (1992) The prediction of human tumor radiosensitivity in situ: an approach using chromosome aberrations detected by fluorescence in situ hybridization. Int.J. Radiat. Oncol. Biol. Phys., 24,279-86. 9. Sasai, K., Evans, J.W., Kovacs, M.S. and Brown, J.M. (1994) Prediction of human cell radiosensitivity: comparison of
colormetric microtiter (MTT) assay in determining the radiosensitivity of cells from murine solid tumors. Int. J. Radiat. Oncol. Biol. Phys., 15,699-702. 18. Baker, F.L., Spitzer, G., Ajani, J.A. etal. (1986) Drug and radiation sensitivity measurements of successful primary monolayer culturing of human tumor cells using celladhesive matrix and supplemented medium. Cancer Res., 46,1263-74. 19. Brock, W.A., Baker, F.L, Wike, J.L, Sivon, S.L and Peters, L.J. (1990) Cellular radiosensitivity of primary head and neck squamous cell carcinomas and local tumor control. Int.}. Radiat Oncol. Biol. Phys., 18,1283-6. 20. Girinsky.T., Lubin, R. and PignonJ.P. (1992) Predictive value of in vitro radiosensitivity parameters in head and neck cancers and cervical carcinomas: preliminary correlations with local control and overall survival. Int.J. Radiat. Oncol. Biol. Phys., 25,3-7. 21. Courtenay, V.D. and Mills, J. (1978) An in-vitro colony assay for human tumours grown in immune-suppressed mice and treated in vivo with cytotoxic agents. Br.J. Cancer, 68, 819-23. 22. West, C.M.L, Davidson, S.E., Burt, PA. and Hunter, R.D. (1995) The intrinsic radiosensitivity of cervical carcinoma: correlations with clinical data. Int.J. Radiat. Oncol. Biol. Phys .,31,841-6. 23. West, C.M.L, Davidson, S.E., Roberts, S.A. and Hunter, R.D. (1993) Intrinsic radiosensitivity and prediction of patient response to radiotherapy for carcinoma of the cervix. Br.J. Cancer, 68,819-23.
References 213 24. Lawton, PA, Hodgkiss, R.J., Eyden, B.P. and Joiner, M.C. (1994) Growth of fibroblasts as a potential confounding factor in soft agar clonogenic assays for tumour cell radiosensitivity. Radiother. Oncol., 32,218-25. 25. Budach, W., Hartford, A., Gioioso, D., Freeman, J., Taghian, A. and Suit, H.D. (1992) Tumors arising in SCID mice share enhanced radiation sensitivity of SCID normal tissues. Cancer Res., 52,6292-6. 26. Stausbol-Gron, B., Bentzen, S.M., Jorgensen, K.E., Nielsen, O.S., Bundgaard, T. and Overgaard, J. (1999) In vitro radiosensitivity of tumour cells and fibroblasts derived from head and neck carcinomas: mutual relationship and correlation with clinical data. Brit.J. Cancer, 79,1074-84. 27. Loeffler, J.S., Harris, J.R., Dahlberg, W.K. and Little, J.B. (1990) In vitro radiosensitivity of human diploid fibroblasts derived from women with unusually sensitive clinical responses to definitive radiation therapy for breast cancer. Radial Res,, 121,227-31. 28. Geara, F.B., Peters, L.J., Ang, K.K., Wike, J.L and Brock, W.A. (1992) Radiosensitivity measurement of keratinocytes and fibroblasts from radiotherapy patients. Int.}. Radial Oncol. Biol. Phys., 24,287-93. 29. West, C.M.L, Elyan, S.A.G., Berry, P., Cowan, R. and Scott, D. (1995) A comparision of the radiosensitivity of lymphocytes from normal donors, cancer patients, individuals with ataxia-telangiectasia (A-T) and A-T heterozygotes. InlJ. Radial Oncol. Biol. Phys., 32, 1371-9. 30. Brock, W.A., Tucker, S.L., Geara, F.B. etal. (1995) Fibroblast radiosensitivity versus acute and late normal skin responses in patients treated for breast cancer. InlJ. Radial Oncol. Biol. Phys., 27,1173-9. 31. Dunst, J., Neubauer, S., Becker, A. and Gebhart, E. (1998) Chromosomal in-vitro radiosensitivity of lymphocytes in radiotherapy patients and AT-homozygotes. Sfra/7/On/ro/., 174,510-16. 32. Budach, W., Claben, J., Belka, C. and Bamberg, M. (1998) Clinical impact of predictive assays for acute and late radiation morbidity. Strahl. Onkol., 174,20^. 33. Russel, N.S., Grummels, A., Hart, A.M. et al. (1998) Low predictive value of intrinsic fibroblast radiosensitivity for fibrosis development following radiotherapy for breast cancer. InlJ. Radial Oncol. Biol. Phys., 73, 661-70. 34. Geara, F.B., Peters, L.J., Ang, K.K., Wike, J.L. and Brock, W.A. (1993) Prospective comparison of in vitro normal cell radiosensitivity and normal tissue reactions in radiotherapy patients. InlJ. Radial Oncol. Biol. Phys., 27, 1173-9. 35. Burnet, N.G., Nyman, J., Turesson, I., Wurm, R., Yarnold, J.R. and Peacock, J.H. (1994) The relationship between cellular radiosensitivity and tissue response may provide the basis for individualising radiotherapy schedules. Radiother. Oncol., 33,228-38. 36. Savitsky, K., Bar-Shira, A., Gilad, S. etal. (1995) A single ataxia telangiectasia gene with a product similar to PI-3 kinase. Science, 268,1749-53. 37. Taylor, A.M.R., Harnden, D.G., Arlett,C.F.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
Ataxia telangiectasia: a human mutation with abnormal radiation sensitivity. Nature, 258,427-9. Gotoff, S.P., Amirmokri, E. and Liebner, E.J. (1967) Ataxia telangiectasia, neoplasia, untoward response to x-irradiation and tuberous sclerosis. Am. J. Dis. Child., 114,617-25. Arlett, C.F. and Priestly, A. (1985) An assessment of radiosensitivity of ataxia-telangiectasia heterozygotes. Kroc. Found. Ser., 19,101-9. Swift, M., Reitnauer, P.J., Morrell, D. and Chase, C.L (1987) Breast and other cancers in families with ataxiatelangiectasia. N. Engl.J. Med., 316,1289-94. Hall, E.J. (1994) The oxygen effect and reoxygenation. In Radiobiologyfor the Radiologist. Philadelphia, J.B. Lippincott Co. 133-52. Adams, M.F.,Gabalski, E.C., Bloch, DA.etal. (1999)Tissue oxygen distribution in head and neck cancer patients. Head Neck, 21,146-53. Gatenby, R.A., Kessler, H.B., Rosenblum, J.S. etal. (1988) Oxygen distribution in squamous cell carcinoma metastases and its relationship to outcome of radiation therapy. InlJ. Radial Oncol. Biol. Phys., 14,831-8. Hockel, M., Schlenger, K., Knoop, C. and Vaupel, P. (1991) Oxygenation of carcinomas of the uterine cervix: evaluation by computerized 02 tension measurements. Cancer Res., 51,6098-102. Vaupel, P., Schlenger, K., Knoop, C. and Hockel, M. (1991) Oxygenation of human tumors: evaluation of tissue oxygen distribution in breast cancers by computerized 02 tension measurements. Cancer Res., 51,3316-22. Brizel, D.M., Scully, S.P., HarrelsonJ.M.tfo/. (1996) Tumor oxygenation predicts for the likelihood of distant metastases in human soft tissue sarcoma. Cancer Res., 56, 941-3. Dunst, J., Hansgen, G., Lautenschlager, C., Fuchsel, G. and Becker, A. (1999) Oxygenation of cervical cancers during radiotherapy and radiotherapy + cis-retinoic acid/interferon. InlJ. Radial Oncol. Biol. Phys., 43, 367-73. Hockel, M., Schlenger, K., Knoop, C. and Vaupel, P. (1996) Association between tumor hypoxia and malignant progression in advanced cancer of the uterine cervix. Cancer Res., 56,4509-15. Stone, H.B., Brown, J.M., Phillips, T.L. and Sutherland, R.M. (1993) Oxygen in human tumors: correlations between methods of measurement and response to therapy. Radial Res., 136,422-34. Urtasun, R.C., Chapman, J.D., Raleigh, J.A., Franko, A.J. and Kock, C.J. (1986) Binding of 3H-misonidazole to solid human tumors as a measure of tumor hypoxia. Int. J. Radial Oncol. Biol. Phys., 12,1263-7. Franko, A.J., Koch, C.J., Garrecht, B.M., Sharplin, J. and Hughes, D. (1987) Oxygen dependence of binding of misonidazole to rodent and human tumors in vitro. Cancer Res., 47, 5367-76. Parliament, M.B., Chapman, J.D., Urtasun, R.C.etal. (1992) Non-invasive assessment of human tumour
214 Predictive assays for radiation oncology
53.
54.
55.
56.
57.
58.
hypoxia with 123-iodoazomycin arabinoside: preliminary report of a clinical study. Br. J. Cancer, 65,90-5. Koh, W.J., Rasey, J.S. and Evans, M.L (1992) Imaging of hypoxia in human tumors with [F-18]fluoromisonidazole. IntJ. Radial Oncol. Biol. Phys., 22,199-212. Cline, J.M., Thrall, D.E., Page, R.L, Franko, A.J. and Raleigh, J.A. (1990) Immunohistochemical detection of a hypoxia marker in spontaneous canine tumours. BritJ. Cancer, 62,925-31. Halpern, H.J., Yu, C, Peric, M., Barth, E., Grdina, D.J. and Teicher, B.A. (1994) Oximetry deep in tissues with lowfrequency electron paramagnetic resonance. Proc. Natl Acad. Sci. USA, 91,13047-51. Murugesan, R., Cook, J.A., Devasashayam, N. etal. (1997) In vivo imaging of a stable paramagnetic probe by pulsed-radiofrequency electron paramagnetic resonance spectroscopy. Magn. Reson. Med., 38,409-14. Goda, F., O'Hara, J.A. and Rhodes, E.S. (1995) Changes in oxygen tension in experimental tumors after a single dose of x-ray irradiation. Cancer Res., 55,2249-52. Li, S.J., Jin, G.Y. and Moulder, J.E. (1991) Prediction of tumor radiosensitivity by hexafluoromisnidazole retention monitored by[1H]/[19F] magnetic resonance spectroscopy. Cancer Commun., 3,133-9.
59. Wilson, G.D., McNally, N.J., Dunphy, E., Karcher, H. and Pfragner, R. (1985) The labelling index of human and mouse tumors assessed by bromodeoxyuridine staining in vitro and in vivo and flow cytometry. Cytometry, 6, 641-7. 60. Begg, A.C., Hofland, I., Glabekke, M.V., Bartlink, H. and Horiot, J.C. (1992) Predictive value of potential doubling time for radiotherapy of head and neck tumor patients: results from the EORTC cooperative trial 22851. Semin. Radial Oncol., 2,22-5. 61. Corvo, R., Giaretti, W., Sanguineti, G. (1995) In vivo cell kinetics in head and neck squamous cell carcinomas predicts local control and helps guide radiotherapy regimen./ Clin. Oncol., 13,1843-50. 62. Eschwege, F., Bourhis, J. and Girinski, T. (1997) Predictive assays of radiation response in patients with head and neck squamous cell carcinoma: a review of the Institute Gustave Roussy experience. Int.). Radial Oncol. Biol. Phys., 39, 849-53. 63. Haustemans, K., Fowler, F., Geboes, K., Lerut, T. and van der Schueren, E. (1997) Do cell kinetics have prognostic and/or predictive value in oesphageal cancer treated by surgery? Eur.J. Surg. Oncol., 23,293-7.
15 Principles of the dose-rate effect derived from clinical data ERICJ. HALL AND DAVIDJ. BRENNER
15,1
INTRODUCTION
The dependence of the biological effect of a given dose on dose rate has been demonstrated unequivocally in experiments with cells cultured in vitro as well as in animal studies [1-3]. A decrease in biological effect as dose is protracted over time is universally observed for radiations of low linear energy transfer (LET) and for cells that die a mitotic as opposed to an apoptotic death. This is attributed to repair of sublethal damage and to repopulation (i.e., to cell birth) if the total exposure period is sufficiently long. Many experiments have been performed and the mechanisms worked out. In the case of clinical data, the situation is not as clear cut, despite the fact that interstitial implants typically involve the range of dose rates for which the variation of biological effect with dose rates is substantial in experimental systems.
15.2 DOSE-RATE CORRECTIONS IN THE ERA OF RADIUM NEEDLE IMPLANTS It was pointed out by Paterson in the 1960s [4] that the dose-limiting factor in the case of interstitial implants is the tolerance of normal tissues. His philosophy was to
push to the maximum the dose tolerated by the normal tissues in order to maximize tumor control. Paterson published a curve relating total dose to overall time, with limiting normal-tissue tolerance as the endpoint. Regarding 60 Gy in 7 days as the standard, he proposed that an implant of shorter duration should have a lower dose, and an implant of longer duration an augmented dose. The published curve represented his considerable clinical experience, accumulated over many years, and was unequivocally based on equalizing late effects in normal tissues. The experience was based on treatments with radium needles implanted according to the Manchester System. Ellis [5] proposed an essentially identical scheme for use in clinical practice; the data upon which his curve was based were attributed to TA Green. The curve of Paterson and Ellis is reproduced in Figure 15.1, together with theoretical curves relating equivalent dose to overall time based on radiobiological data for early-responding and late-responding tissues, normalized to 6000 cGy in 7 days. The oc/P and T{ for early-responding tissues were taken from the in-vitro data for cell lines of human origin [6]. The oc/P and T{ for late-responding tissues were taken from the limited experimental clinical data available, specifically from those of Turesson and Thames [7]. It is interesting to note that the calculated curve based on the radiobiological parameters for late-responding tissues is virtually indistinguishable from the curves of Paterson
216 Principles of the dose-rate effect derived from clinical data
of outer needles. Because all wires have the same linear activity, there is a correlation between implanted volume and dose rate, with larger volumes being associated with higher dose rates.
Figure 15.1 Dose equivalent to 60 Gy in 7 days calculated in different ways: based on clinical experience of normal tissue tolerance (Paterson [4], Ellis [5]); calculated from o/p and 7) values related to late effects; and calculated from the average oc/P and 7| values for cells of human origin cultured in vitro, i.e., early effects.
and of Ellis, who unequivocally based their judgment on obtaining equal late effects. The curve relating equivalent dose and overall time is steeper for late-responding than for early-responding tissues because of the smaller value ofot/p.
153 THE INTRODUCTION OF IRIDIUM WIRE IMPLANTS The introduction of iridium wire as a substitute for radium needles in interstitial brachytherapy allowed greater flexibility and patient comfort, but also resulted in a much larger variation of dose rate between individual implants. Two factors contribute to this: 1. As a consequence of the relatively short half-life of iridium-192 (74 days), the linear activity will vary significantly during the period of several months that the wires may be used or re-used. 2 The Paris System of dosimetry [8] developed for iridium-192 implants, where all sources have the same linear activity with varying separation between wires for different lengths (i.e. greater separation for larger wires), results in a wider range of dose rates than was characteristic of radium implants using the Parker-Paterson dosimetry system, where internal needles had half or two-thirds of the linear activity
The combination of those two factors results, in practice, in a threefold variation in the overall irradiation time for the delivery of a given tumor dose. Nevertheless, Pierquin and his colleagues [9] came to the conclusion that, in iridium-192 implants, the time factor (and therefore the dose rate) was unimportant or at least less important than others had suggested. Consequently, the Paris school recommended the same prescribed dose irrespective of overall time within the range 3-8 days. They were careful to point out that their conclusions were preliminary, but nevertheless concluded: 'We can however say with certainty that the variation in overall treatment time for the same tumor dose from 3 days to 8 days does not appear to influence the frequency or recurrence or necrosis.' Based upon this conviction, many hundreds of patients were treated with iridium-192 implants using standard doses uncorrected for treatment time or dose rate, despite the fact that this conflicts with the previously published clinical experience of Paterson and of Ellis, and does not agree with the experimental radiobiological data that would predict a substantial dose-rate effect over the dose-rate range in question.
15.4 DOSE-RATE EFFECTS FROM A RETROSPECTIVE ANALYSIS OF THE IRIDIUM IMPLANT DATA The large series of patients treated with iridium wire implants has been followed carefully over the years and constitutes an invaluable resource. Two important papers have appeared in recent years describing a retrospective analysis of these data. In the first, Mazeron et al. [10] studied the incidence of local tumor control and necrosis in T1 and T2 squamous cell carcinoma of the mobile tongue and floor of mouth treated with interstitial iridium-192. The data are shown in Figure 15.2. This represents one of the largest cohorts of patients that has been analyzed after treatment with a standardized implantation technique, but the numbers still preclude a detailed analysis of dose rate and can distinguish only between dose rates above and below 0.5 Gy h'1. Two principal conclusions can be drawn from this analysis: 1. There is little or no difference in local control between the two dose-rate ranges provided a sufficiently high total dose is used (65-70 Gy), but there is a clear separation at lower doses (around 60 Gy), with the lower dose rate being significantly less effective.
The bias of tumor size and dose rate 217
Figure 15.2 Incidence of local tumor control and necrosis for T 1-2 squamous cell carcinoma of the mobile tongue and floor of mouth, treated with interstitial radiotherapy using iridium192. (Based on the published results of Mazeron et al. [10].j
2. Over the entire range of doses used, there was a higher incidence of necrosis associated with the higher dose-rate range.
oc/p and T| for early-responding and late-responding tissues. The radiobiological predictions mirror the essential characteristics of the clinical data, i.e., adequately explain what was observed. In a second study, Mazeron et al. [11] analyzed data from a large group of patients with carcinoma of the breast who received iridium-192 as a boost to externalbeam radiotherapy. These results allow an assessment of the effect of dose rate on tumor control, but provide no information on the effect of dose rate on normal tissues, because there was only one case that involved necrosis, probably because the interstitial implant comprised only part of the radiotherapy. A fixed standard total dose was used, regardless of the dose rate, and there is a clear correlation between the proportion of recurrent tumors and the dose rate, as illustrated in Figure 15.4. For a given total dose, there were markedly fewer recurrences when the radiation was delivered at higher as opposed to lower dose rates. A clear difference in tumor control could be seen between 0.3 and 0.9 Gy h"1, as predicted from radiobiological experiments with cells in vitro.
The clinical data are in line with the predictions that would be expected based on radiobiological considerations, particularly the more critical dependence on dose rate of late-responding tissues. Figure 15.3 shows the theoretical expectations using representative values for
Figure 15.4 Fraction of patients who showed no local recurrence as a function of dose rate amongst patients with breast carcinoma treated by a combination of external-beam radiation plus iridium-192 interstitial implant. The implant was used to deliver a dose of 37 Gy; the dose rate varied by a factor of 3 due to different linear activities of the iridium-192 wire, and to different size volumes implanted. (Drawn from the data of Mazeron et a I. [11].,)
Figure 153
Local tumor control and necrosis rate at 5 years as
a function of dose in patients treated for T1-2 squamous cell carcinomas of the mobile tongue and the floor of the mouth with interstitial iridium-192 implants. The patients were grouped according to whether the implant was characterized by a high dose rate (above 0.5 Gy h~') or low dose rate (below 0.5 Gy h~1). The necrosis rate is higher for the higher dose rate at all dose levels. Local tumor control did not depend on dose rate provided the total dose was sufficiently large. (Redrawn from the data of Mazeron et a I. [10].,)
15.5 THE BIAS OF TUMOR SIZE AND DOSE RATE A complication and confounding variable in the interpretation of clinical data relating dose to produce an equivalent effect to implant time (and therefore to dose rate) is the fact that, for interstitial implants, the dose rate tends to increase as the size of the implant increases. This correlation is particularly true for implants using iridium-192 wires, as used in the Paris System, which are
218 Principles of the dose-rate effect derived from clinical data
all of the same linear activity, but less so when there is a variation in linear activity, as in the Parker and Paterson system [12]. The bias of larger tumors and larger volumes being associated with higher dose rates, while smaller tumors and smaller treatment volumes are associated with lower dose rates, was pointed out by Pierquin and his colleagues [9]. Larger tumors of course require a larger dose for a given level of local control, whereas the maximum dose that can be tolerated by normal tissues decreases as the volume implanted increases. This volume/dose-rate bias has an interesting effect on the isoeffect curves, as illustrated in Figure 15.5. The lower curves are isoeffect curves for late-responding normal tissues, while the upper curves relate to an earlyresponding tissue, which includes tumor control. The isoeffect curve for late-responding normal tissues is steeper than for tumor control, showing a greater dependence on overall time (or dose rate) in keeping with the smaller oc/f3 ratio characteristic of such tissues as described by Withers et al. [13]. The dashed lines in Figure 15.5 illustrate, in a qualitative way only, the effect of the variation of dose rate with tumor size and therefore with implanted volume. At one extreme, higher dose rates are associated with large tumors and therefore large implanted volumes. Larger tumors require a larger dose to produce local control, while larger volumes can tolerate lower doses to produce a given level of necrosis. The bias of higher dose
rates being associated with larger tumors (and, conversely, lower doses associated with smaller tumors) has opposite effects on the isoeffect curves for tumor control and for normal-tissue tolerance. It will tend to flatten the isoeffect curve for tumor control and steepen the isoeffect curve for normal-tissue tolerance, as shown by the dashed lines in Figure 15.5. It should be stressed that the dashed lines are qualitative only, showing a trend, but cannot be quantitative at this time because data relating tumor control dose to tumor volume, or tolerance dose to treatment volume, are very limited. Based on these considerations, then, it is clear why the Paris school and the Paterson/Ellis school differed so radically in their prescriptions for dealing with dose-rate changes. Firstly, the Paterson/Ellis recommendations were based on equalizing only late effects, where there is a clear change of equi-effect dose with dose rate (solid curves in Figure 15.5). However, the Paris recommendations were based on an attempt to equalize late and early effects (with hindsight, it certainly seems that when the dose rate changes it is not possible to match both late and early effects). Secondly, the Paterson/Ellis recommendations date from the era of radium needles when there was less correlation between volume and dose rate because needles of varying linear activity were available. However, the Paris recommendations were based on iridium wire implants, for which there is strong correlation between tumor volume and dose rate, which would tend to make the equi-effect curve for tumor control vary even less with dose rate.
15.6 EARLY-RESPONDING AND LATERESPONDING TISSUES
Figure 15.5 The correlation between implant volume and dose rate has an influence on the shape of the isoeffect curves for both early and late effects. Solid lines: calculated isoeffect curves for in-vitro data representing tumors and for in-vivo data corresponding to late-responding normal tissue. Dotted lines illustrate qualitatively the way in which the isoeffect curves change owing to the bias of higher dose rates being associated with larger tumors. Larger tumors require a larger dose for a given level of local control; this tends to flatten the isoeffect curve for tumor control. By contrast, the isoeffect curve for late effects in normal tissues is steeper, because larger implanted volumes can tolerate smaller doses.
A basic tenet of radiotherapy is that delivering the overall dose in many fractions, or at low dose rates, improves therapeutic outcome. The rationale for this originally empirical observation was provided by Withers and colleagues [13], who showed that, for a given dose, increasing the number of fractions will decrease late effects much more than it will decrease tumor control. Withers and colleagues quantified the difference in response between late-responding tissues and early-responding tissues, such as tumors, in terms of the linear-quadratic formalism, and the parameter oc/p (see below), which is generally smaller for late-responding than for acutely responding tissue. What goes for multifraction protracted treatments applies too to continuous low dose rate. Because in most instances tumor cells are cycling rapidly and qualify as early responding, whereas the dose-limiting normal tissues are late responding, a protracted treatment regime consisting of many fractions of external beam, or continuous low dose rate in an implant or intracavitary procedure, leads to an improved therapeutic ratio. There is, however, at least one excep-
Exploiting differences in repair rates to optimize brachytherapy 219
tion to this general rule, namely carcinoma of the prostate. This is a very slowly growing tumor for which the oc/P ratio is about 1.5 Gy, similar to that for lateresponding normal tissues. This value was derived by Brenner and Hall [14] from an analysis of two sets of mature data on the radiotherapeutic control of prostate cancer, one using external-beam therapy and the other permanent seed implants. Because the tumor and doselimiting normal tissues have similar oc/(3 values, high dose-rate (HDR) brachytherapy should be a highly appropriate modality for treating prostate cancer.
15.7
SUBLETHAL DAMAGE REPAIR RATES
In the past decade, it has become increasingly clear that there is a second type of difference in radiation response between late-responding and early-responding tissues, namely the rate of repair of sublethal damage. If lateresponding normal tissues and early-responding tissues such as tumors do repair at different rates, this could be used to produce improved radiotherapeutic protocols. Specifically, just as differences in cc/|3 are currently exploited to design schemes with improved therapeutic advantage between early and late damage, it might be expected that differences in repair rates could be exploited to the same end. The suggestion that repair rates might be slower in late-responding compared to early-responding tissue originated with Thames, Withers, and Peters in 1984 [15]. In rodents, they compared half-times for sublethal damage repair (TO for damage to bone marrow, jejunum, and colon (early-responding tissues) with those for lung and spinal cord (late-responding tissues). They concluded that Ti for the late-responding tissues were significantly greater than 1 h, whereas for the earlyresponding tissues the T{ were less than 1 h. This suggestion was subsequently corroborated in the clinic by analyzing the results of hyperfractionated radiotherapy, where treatments were given at intervals of only a few hours. Following some early indications from Nguyen etal. [16],Edsmyr etal. [17], Morgan eta/. [18], and Marcial etal. [19] that treatments with 2-h to 4.5-h intervals were producing excess late effects, Cox et al. [20] produced a definitive analysis of the results of the Radiation Therapy Oncology Group (RTOG) protocol 8313; this protocol allowed hyperfractionation intervals of 4-8 h for treatment of cancers of the upper respiratory and digestive tracts. The results were divided into interfraction intervals of < 4.5 h versus > 4.5 h. Both acute toxicity and tumor control were unaffected by the interfraction interval, suggesting a relatively short Ti of, roughly, < 100 min. On the other hand, the < 4.5-h group showed a significant increase in late toxicity, suggesting that repair was not complete between fractions, implying a T{ of, roughly, > 200 min.
Further evidence from the clinic comes from the results of Turesson and Thames [7] on early-responding and late-responding skin damage after fractionated radiotherapy. They found two-component repair processes, both early-responding and late-responding damage having an estimated fast repair TI of ~ 25 min. However, the slow repair for early-responding damage had an estimated T} of ~ 75 min, whereas the corresponding estimated slow Ti for late-responding tissue was ~ 250 min, with confidence limits from 210 to 320 min. In both cases, the slow and fast repair channels each repaired about half the repairable damage. The interpretation of laboratory-based animal data on repair rates in late-responding tissues is still a matter of some debate. Recent laboratory evidence has emerged from analyses of late radiation damage in rodents in which, following the clinical results of Turesson and Thames, a possible two-component (fast and slow) mechanism for late tissue damage repair was investigated. Ang et al. [21] found two components of repair for spinal cord damage in rodents (T{ values estimated at 0.7 h and 3.8 h) with an estimated 62% of the repair occurring through the slow channel. Moulder and Fish [22] have reported two components of repair for kidney damage in rats (Tf estimated at 30-40 min and 130-200 min), with an estimated 70-75% of the repair proceeding through the slow channel. Finally, van Rongen et al. [23], re-analyzing the data of Travis et al. [24] for rodent lung damage, also found two components of repair (Ti estimated at 0.4 h and 3.6 h), with an estimated 24% of the repair proceeding through the slow channel.
15.8 EXPLOITING DIFFERENCES IN REPAIR RATES TO OPTIMIZE BRACHYTHERAPY SCHEDULES Differences in repair rates between early-responding and late-responding tissues, if real, open up new possibilities for optimizing brachytherapy schedules: in this situation, it can be shown that continuous low dose rate (CLDR) no longer represents the optimal schedule producing the maximum therapeutic ratio between tumor control and late sequelae. For example, pulsed brachytherapy, with doses separated by several hours, would result in a bigger differential between tumor and normal tissue than CLDR [25]. It can also be shown that loading the early and late parts of brachytherapy schedules with larger doses, whilst keeping the overall dose and time fixed, should improve therapeutic ratios still more. These ideas suggest that the difference in repair rates between early-responding and late-responding tissues can be exploited to produce clinically practical protocols of considerably greater therapeutic advantage than currently achieved. In this regard, the considerations here complement the concepts expounded by Withers and
220 Principles of the dose-rate effect derived from clinical data
colleagues, who showed that the difference in the cc/P ratio between early-responding and late-responding tissues could be exploited to advantage. In this case, however, it is the difference in repair rates, as well as that in oc/P, which can be exploited. Under the assumption that late-responding tissues repair more slowly than do early-responding tissues, for the same overall dose and time, it is possible to design optimized protocols which have significantly less lateresponding tissue cell killing, and slightly more earlyresponding cell killing [26]. Thus, in comparison with conventional CLDR protocols of the same overall dose and time, the optimized protocols designed here should produce: (a) significantly fewer late-tissue complications, and (b) comparable or slightly increased tumor control and early normal-tissue sequelae. Temporal optimization - that is, systematic optimization of fractionation protocols - is a new concept in radiotherapeutic design. There is an extensive literature on spatial optimization in radiotherapy, in terms of delivering maximum dose to the tumor relative to normal tissue. Temporal optimization, however, has not received adequate consideration, and may allow major therapeutic gains. Finally, we add several caveats. First, the magnitude of the potential gains that can be made depends on the rate of repair for late-responding damage. The evidence for a component of repair of several hundred minutes is summarized above, and is reasonably convincing. However, details, such as the ratio of a slow to fast component of damage, are not well known. Second, the concepts described here will not allow improvement of conventional external-beam radiotherapy, where the time between fractions (-24 h) is much longer than all known repair times. Rather, the ideas here are relevant to all accelerated protocols, be they brachytherapy or accelerated hyperfractionated external-beam radiotherapy. Finally, it is stressed that not all potential fractionation protocols can be usefully optimized. In particular, as discussed above, pulsed protocols with long interfraction intervals (such as ~15 h overnight gaps) probably can never produce a therapeutically advantageous therapeutic ratio, although optimization will still produce the best possible therapeutic advantage for such protocols. In conclusion, there is likely to be another biological parameter, sublethal damage repair rates, which we may exploit to produce a therapeutic advantage, and which would open up a variety of new directions for brachytherapy.
REFERENCES 1. Hall, E.J. and Bedford, J.S. (1964) Dose rate: its effect on the survival of HeLa cells irradiated with gamma rays. Radial Krc.,22,305-15.
2. Bedford, J.S. and Mitchell, J.B. (1973) Dose-rate effects in synchronous mammalian cells in culture. Radial. Res., 54,316-27. 3. Fu, K., Phillips, T.L, Kane, LJ. and Smith, V. (1975) Tumour and normal tissue response to irradiation in vivo: variation with decreasing dose-rate. Radiology, 114, 709-16. 4. Paterson, R. (1963) The Treatment of Malignant Disease by Radiotherapy, 2nd edn. London, Edward Arnold. 5. Ellis, F. (1968) Dose time and fractionation in radiotherapy. In Current Topics in Radiation Research, Vol. 4, ed. M. Elbert and A. Howard. Amsterdam, North Holland, 359-97. 6. Brenner, D.J. and Hall, E.J. (1991) Conditions for the equivalence of continuous to pulsed low dose rate brachytherapy. Int.J. Radial Oncol. Biol. Phys., 20, 181-90. 7. Turesson, I. and Thames, H.D. (1989) Repair capacity and kinetics of human skin during fractionated radiotherapy: erythema, desquamation, and telangiectasia after 3 and Syears'followup. Radiother. Oncol., 15,169-88. 8. Pierquin, B. (1971) Dosimetry: the relational system. Proceedings of a Conference on Afterloading in Radiotherapy. Rockville, New York. US Department of Health, Education and Welfare, Publication number (FDA) 72-8024,204-27. 9. Pierquin, B., Chassagne, D., Baillet F. etal. (1973) Clinical observations on the time factor in interstitial radiotherapy using iridium-192. Clin. Radiol., 24, 506-9. 10. Mazeron, J.J., Simon, J.M., Le Pechoux. C. etal. (1991) Effect of dose-rate on local control and complications in definitive irradiation of TI squamous cell carcinomas of mobile tongue and floor of mouth with interstitial iridium-192. Radiother. Oncol., 21,39-47. 11. Mazeron, J.J., Simon, J.M., Crook, I etal. (1991) Influence of dose-rate on local control of breast carcinoma treated by external beam irradiation plus iridium-192 implant. InlJ. Radial Oncol. Biol. Phys., 21,1173-7. 12. Meredith, W.S. (ed.) (1967) Radium Dosage: The Manchester System, 2nd edn. Baltimore, Williams and Wilkins. 13. Withers, H.R. Thames, H.D. and Peters, LJ. (1982) Differences in the fractionation response of acutely and late-responding tissues. In Progress in Radio-oncology, Vol. II, ed. K.H. Karcher, H.D. Kolgelnikand G. Reinartz. New York, Raven Press, 287-96. 14. Brenner, D.J. and Hall, E.J. (1999) Fractionation and protraction for radiotherapy of prostate carcinoma. Int. J. Radial Oncol. Biol. Phys., 43,1095-101. 15. Thames, H.D.Jr, Withers, H.R. and Peters, LF.(1984) Tissue repair capacity and repair kinetics deduced from multifractionated or continuous irradiation regimens with incomplete repair. BrilJ. Cancer, 49,263-9. 16. Nguyen, T.D., Demange, L, Froissart, D., Panis, X. and Loirette, M. (1985) Rapid hyperfractionated radiotherapy. Cancer, 56,16-19. 17. Edsmyr, F., Andersson, L, Espoti, P.L, Littbrand, B. and
References 221
18.
19.
20.
21.
Nilsson, B. (1985) Irradiation therapy with multiple small fractions per day in urinary bladder cancer. Radiother. Oncol., 4,197-203. Morgan, DAL, Bradley, P.J. and MacLennan, K.A. (1987) Radiotherapy of advanced laryngeal cancer using three fractions per day (abstract). Brit.J. Radial., 60,606. Marcial, V.A., Pajak, T.F., Chang, C., Tupchong, L and StetzJ. (1987) Hyperfractionated photon radiation therapy in the treatment of advanced squamous cell carcinoma of the oral cavity, pharynx, larynx, and sinuses, using radiation therapy as the only planned modality: (preliminary report) by the Radiation Therapy Oncology Group. Int.}. Radial Oncol. Biol. Phys., 13,41-7. Cox, J.D., Pajak, T.F., Marcial, V.A., Coia, L, Mohiuddin, M. and Fu, K.K. (1991) Interfraction interval is a major determinant of late effects, with hyperfractionated radiation therapy of carcinomas of upper respiratory and digestive tracks: results from Radiation Therapy Oncology Group protocol 8313. InLJ. Radial Oncol. Biol. Phys., 20, 1191-5. Ang, K.K., Guttenberger, R., Thames, H.D., Stephens, L.C., Smith, C.D. and Geng, Y. (1992) Impact of spinal cord
22.
23.
24.
25.
26.
repair kinetics on the practice of altered fractionation schedules. Radiother. Oncol., 25,287-94. Moulder, J.E. and Fish, B.L (1992) Repair of sublethal damage in the rat kidney (abstract) In Radiation Research: a Twentieth-century Perspective, Vol. 1, ed. J.D. Chapman, W.C. Dewey and G.F. Whitmore. San Diego, Academic Press, 238. van Rongen, E., Thames, H.D. and Travis, E.L (1993) Recovery from radiation damage in mouse lung: interpretations in terms of two rates of repair. Radial Res., 133,225-33. Travis, E.L, Thames, H.D., Watkins, T.L. and Kiss, I. (1987) The kinetics of repair in mouse lung after fractionated irradiation. Int.J. Radial Biol., 52,903-19. Brenner, D.J. and Sachs, R.K., (1995) Potential reduced late effects for pulsed brachytherapy compared with conventional LDR. Int.J. Radial Oncol. Biol. Phys., 31, 201-2. Brenner, D.J., Hall, E.J., Huang, Y. and Sachs, R.K. (1994) Optimizing the time course of brachytherapy and other accelerated radiotherapeutic protocols. Int.J. Radial Oncol. Biol. Phys., 29,893-901.
This page intentionally left blank
PART
Clinical practice
16 Endobronchial brachytherapy in the treatment of lung cancer 17 Brachytherapy in cancer of the esophagus 18 High dose-rate afterloading brachytherapy for prostate cancer 19 Low dose-rate brachytherapy for breast cancer 20 Brachytherapy in the treatment of head and neck cancer 21 High dose-rate interstitial and endocavitary brachytherapy in cancer of the head and neck 22 Brachytherapy in the treatment of pancreas and bile duct cancer 23 Brachytherapy for treating endometrial cancer 24 Low dose-rate brachytherapy for treating cervix cancer: changing dose rate 25 High dose-rate brachytherapy for treating cervix cancer 26 Brachytherapy for brain tumors 27 Interstitial brachytherapy in the treatment of carcinoma of the cervix 28 Interstitial brachytherapy in the treatment of carcinoma of the anorectum 29 High dose-rate brachytherapy in the treatment of skin tumors 30 Hyperthermia and brachytherapy 31 The costs of brachytherapy 32 Quality management: clinical aspects 33 Safe practice and prevention of accidents in afterloading brachytherapy 34 Pulsed low dose-rate brachytherapy in clinical practice
225 243 257 266 284 296 317 333 343 354 373 379 387 393 400 410 423 433 443
Ill
This page intentionally left blank
16 Endobronchial brachytherapy in the treatment of lung cancer BURTON LSPEISER
16.1
INTRODUCTION
The treatment of lung cancer by radiation has consisted primarily of the use of standard external-beam radiation. The use of endobronchial brachytherapy (EBBT) in the form of manual afterloading low dose-rate (LDR) began to emerge in the early to mid 1980s. It was not until the mid and late 1980s that high dose-rate remote afterloading brachytherapy began to be utilized increasingly. At the present time, it is the most prevalent means of treating endobronchial disease with brachytherapy throughout the world. Endobronchial brachytherapy for lung malignancies consists of the placement of a catheter into the involved bronchus, with the radiation source placed with remote or manual afterloading. After-loaded implants can deliver LDR, intermediate dose-rate (IDR), or high dose-rate (HDR) ionizing radiation at less than 1 Gy h"1 (low dose rate, LDR), 1-12 Gy h'1 (IDR), or greater than 2 Gy miir1 (HDR). A typical afterloading system utilizes hollow catheters and iridium-192 seeds embedded in nylon ribbon for LDR, or a single iridium-192 source (nominal activity of 10 Ci or 370 GBq), which 'dwells' within the lumen of the treatment catheters. Endobronchial brachytherapy, however, has been used
primarily for palliation, for recurrent disease. Although it is used curatively, this usage is not as prevalent. Even less common is its use for occult carcinomas of the lung, where, as the sole modality, or used in conjunction with either photodynamic therapy or limited external radiation, it can lead to high cure rates with low morbidity.
16.2
BACKGROUND
Brachytherapy was first used at the turn of the twentieth century when the Curies gave a small radium tube to Dr Danlos for insertion into a malignancy. In the ensuing years, further progress was made in the area of brachytherapy. Kernan [1] first reported the implantation of radon-222 seeds into carcinomas of the trachea and bronchus utilizing a rigid bronchoscope. He reported on ten cases treated with radon-222 seed implantation combined with diathermy treatment. Of interest was his comment in the report that, 'It was possible to destroy the tumors with diathermy and Radon implantation, and there have been no local recurrences as yet, although one case has been followed for five years.' For the next 10 to 20 years, work was being performed on standardization of dose calculation and prescription
226 Endobronchial brachytherapy in the treatment of lung cancer
for brachytherapy. The publication of the Manchester System of Paterson, Parker, and Meredith (1934) set the foundation for brachytherapy dosimetry. This was further advanced with the publication of the manual, Radium Dosage [2]. Pool [3] reported on radon-222 used at Memorial Hospital, New York City, in a total of 42 patients implanted during the time period of 1936-1960. The implanted seeds consisted of a sealed gold capillary tube filled with radon-222 gas. The technique reported by Pool, 'proved to benefit patients with primary tracheal tumors, and patients with bronchial stump recurrences following pulmonary resection.' Radon-222 seed implantation was also used during thoracotomy, with the first reported case by Graham and Singer [4]. In order to decrease the radiation safety problems with radon-222, efforts were made to identify radioactive isotopes of shorter half-life and lower decay energies, as well as developing afterloading techniques, in order to reduce medical personnel and patient exposures. Henschke [5] introduced standardized afterloading techniques with gold-198 and, subsequently, iridium-192. With the shorter half-life of iridium-192 and afterloading techniques, interstitial implantation showed a rapid increase in medical utilization, both in its range of application and its overall usage in the USA. However, despite this rapid increase, there were very few changes in the limited use of brachytherapy for carcinoma of the lung. Early transbronchial implantation techniques, initially utilizing radon-222 and then iodine125 through the rigid bronchoscope, were difficult and associated with a considerable expenditure of physician time and effort. With the advent of the flexible bronchoscope, a new flexible applicator system was designed by Martinez and colleagues [6]. The use of intraluminal brachytherapy by the placement of an afterloading flexible applicator bronchoscopically was first reported in the American literature by Mendiondo et al [7]. In addition, the new neodymium-YAG laser was introduced into clinical use for the treatment of tracheal and endobronchial obstructions for both primary and metastatic lung malignancies in 1984. It is this convergence of medical technologies that led to a rapid increase in the use of intraluminal brachytherapy utilizing the afterloading techniques, usually in concert with YAG laser photoresection.
163
ENDOBRONCHIAL CATHETERS
A variety of commercial catheters is available for use. Basically, the physical characteristics of each catheter must allow for easy placement and removal from the endobronchial tree and immobilization when in place. The catheter must fit through the biopsy channel of the endoscope and the internal diameter must be source
compatible (most commonly used are 5 or 6 French with a 1.7 or 2 mm internal diameter, respectively). Finally, the distal end of the catheter is closed-end to prevent dummy seeds or the active source from coming into contact with tissues and/or fluids. The placement of the catheter is conducted after assessing the cardiac and pulmonary status of the patient to ensure his or her ability satisfactorily to tolerate an endobronchial procedure. With a pulmonologist, the catheter is placed in an endoscopy suite. This is performed with constant vital sign monitoring, the use of topical anesthesia and intravenous sedation. Under video bronchoscopy, the malignant lesion is identified by the pulmonologist and radiation oncologist. Photographic documentation is acquired and the anatomical characteristics of the malignancy are noted, including distance from anatomical landmarks such as bronchial branch points and carina. The extent of the malignant lesion can be further localized utilizing radio-opaque markers placed on the patient's thorax corresponding to the most distal and proximal extents of the malignancy as identified by the bronchoscope under fluoroscopy. After this visual inspection and fiuoroscopic confirmation, a guidewire in a catheter is placed through the biopsy channel. Its placement is confirmed visually and fluoroscopically. The bronchoscope is removed and, simultaneously, the catheter is positioned such that the tip is several centimeters distal to the most distal point of the malignancy. After complete removal of the bronchoscope, the proximal end of the catheter is secured to the nose. The guidewire is removed and replaced by a set of markers. Orthogonal treatment planning simulation films are obtained and the markers in the catheter and fiducial points are digitized into the treatment planning computer.
16.4
TREATMENT PRESCRIPTION
The prescription depth is calculated for the threedimensional volume by multiple points perpendicular to the axis of the catheter or source train to which the minimum target dose is prescribed. The prescriptions that are reported in the literature range from 0.5 cm to 2 cm. To ensure treatment of the entire tumor volume, with consideration for possible source or tumor movement secondary to respiratory excursions, the maximal distance from the source center to the margin of the malignancy must be considered during three-dimensional brachytherapy planning. A longitudinal margin of 2 cm proximal and distal to the malignant margins is commonly used. A prescription depth within the range 0.5-2 cm, which allows for effective dose distribution to encompass the tumor volume depth without exceeding bronchial mucosa tolerance dose, is employed, with a recommendation that dose standardization be used
Strategies for treatment of occult carcinomas of the endobronchus 227
(i.e., use or report the 1 cm depth from source axis for comparison to other studies).
16.5 STRATEGIES FOR TREATMENT OF OCCULT CARCINOMAS OF THE ENDOBRONCHUS Occult carcinomas of the lung are a subpopulation defined as carcinomas diagnosed by sputum cytology, and bronchoscopy using brushings, washings, and/or biopsy. Less commonly, patients with cough and/or hemoptysis undergo bronchoscopy and are diagnosed in that fashion without a prior sputum cytology being positive. Their defining concept is that they cannot be detected by conventional radiographic means before or immediately after the initial diagnosis. In 1974, Sanderson et al. published 'Bronchoscopic localization of radiographically occult lung cancer' [8]. In 1980, Cortese et al. published their study, 'Roentgenographically occult lung cancer' [9]. In the same year, Martini and Melamed published 'Occult carcinoma of the lung' [10]. Initially, the treatment of choice was surgery, either lobectomy or pneumonectomy. However, as experience with photodynamic therapy (PDT) and then EBBT increased, the 'menu' of treatment modalities increased. This subpopulation of roentgenographically occult carcinomas of the lung is associated with interesting characteristics. First, the time interval from the initial abnormal sputum cytology to bronchoscopic confirmation, as reported by the Mayo Lung Project [9], ranged from 1 to 1014 days (median, 70 days; 75th percentile, 169 days). Second, the disease is most often Tis, Tl and NO [11] (Saito etal. [12] found that of 94 patients, 17% were Tis and 77% were Tl). Third, most cases are squamous cell carcinomas; in a significant number, dysplasia initially had been the only finding [9]. Fourth, adverse prognostic factors (i.e., weight loss) that predict lower survival rates, are rarely present [13-16]. Finally, synchronicity and metachronicity are significant. In a surgical series, Nagamoto et al. [17] reported a rate of 1.09 lesions per patient, Kato etal. [18] found 1.21 lesions per patient, and Saito et al. [12] found 1.2 lesions per patient. In the Mayo Lung Project Study [9], a metachronous rate of 5% per year was reported. Saito et al. [19] reported a rate of 0.022 lesions per patient-year; the rate was 0.041 lesions per patient-year when synchronous and metachronous tumors were combined. If a patient had a second lesion, there was a 47% probability that, within 5 years, a third lesion would be identified, at a rate of 0.11 lesions per patient-year [19]. The 5-year survival rate for patients with a single lesion and no evidence of synchronous or metachronous lesions was 90%. If, however, other metachronous or synchronous lesions were present, the 5-year survival rate was 59%. In 108 patients
who underwent surgical resection for occult carcinoma, Nagamato etal. [17] identified ten (9.3%) who had additional squamous cell carcinomas <1 mm in size. These lesions were associated with either dysplasia or marked atypia. Another important prognostic factor is the size of the lesion. Many studies have found that lesions <10 mm in size are associated with the most favorable outcomes. In a surgical study of 127 patients [20], 55 patients had lesions of this size, and no metastatic lymph nodes were identified. Of 46 patients with lesions >10 mm but <20 mm, there were four patients (9%) with nodal metastasis. Of 26 patients with lesions that were >20 mm but <55 mm, four patients (15%) had nodal disease. Overall, for lesions >10 mm, the incidence of nodal involvement was 11%. In an earlier study by Saito [12], extrabronchial invasion was documented by pathological analysis in 16 (17%) of 94 patients. Five (31%) of these 16 patients had metastatic spread to nodes. Only one (1%) of 78 patients had nodal disease without evidence of extrabronchial invasion. No recurrences were identified in 75 patients who had intrabronchial disease with no lymphatic spread who underwent a complete resection. Overall, the cause-specific 5-year survival rate was 93.5% (versus 80.4% for all causes combined). Kato et al. [18] treated 45 lesions fulfilling the criteria for occult carcinomas in 40 patients (1.13 lesions per patient). PDT was the only treatment used for 30 lesions in 20 patients, and the complete response rate was 100%. Three patients (15%) had recurrences, one of whom (5%) died of the disease. An additional nine patients (45%) died of unrelated causes. Considerably fewer patients with occult carcinomas have been treated with EBBT than with PDT or surgery. Sutedja et al. [21] reported two patients with Tl squamous cell carcinoma who were treated with high doserate EBBT. Three fractions of 10 Gy were delivered at a 1 cm depth. Both patients were alive without disease at follow-up examinations at 54 and 25 months. Tredaniel et al. [22] treated 29 patients with a diversity of lesions, whose common denominator was that their carcinomas were limited to the bronchus (radiographically occult), such that the disease could be encompassed by intraluminal brachytherapy. In contrast to other reported series, however, these patients had undergone prior treatment, which included surgery, external radiation, and/or chemotherapy. The patients were treated with high dose-rate EBBT using a dose of 7 Gy calculated at a 1 cm depth for six fractions (42 Gy). The median actual survival of these patients had not been reached after 23 months of follow-up. Saito et al. [12] treated 49 occult carcinomas in 41 patients (1.2 lesions per patient) with external-beam radiation using 40 Gy in 20 fractions plus EBBT of 25 Gy in five fractions. Doses were customized and the prescription point ranged between 3 mm and 9 mm depth,
228 Endobronchial brachytherapy in the treatment of lung cancer
based on the average diameter of the airway being treated. With a median follow-up of 24.5 months, only two patients (5%) experienced recurrences. A prospective non-randomized study reported by Perol and coworkers [23], utilized the following selection criteria in the treatment of occult lung cancer with EBBT. All cases were proximal non-small cell lung carcinomas, <1 cm, in an area not previously irradiated. All lesions were roentgenographically occult, and the patients had severe chronic respiratory failure or had already had surgery or external radiation for previous lung carcinoma. An escalating dose protocol was employed, and doses were prescribed at 1 cm. The first two patients received three fractions of 7 Gy each, the next four patients received four fractions of 7 Gy each, and the last 13 patients received five fractions of 7 Gy each. Two months after the completion of treatment, 15 (83%) of the 18 evaluable patients were locally controlled with negative biopsies. At 1 year, 12 (75%) of 16 evaluable patients revealed no evidence of disease. Actuarial 1-year and 2-year survival rates were 78% and 58%, respectively, with a median survival of 28 months. Two patients who received five fractions of 7 Gy developed necrosis of the bronchial wall. Two of these patients died of hemoptysis, one with no evidence of carcinoma. The substantial synchronous and metachronous rates and the finding of additional small lesions with dysplasia or marked atypia all lead to the concept of a 'field defect.' It is quite likely that the entire bronchial mucosa is at risk, with a high probability of more than one lesion developing. Any treatment strategy should address this basic issue. Although lobectomy and/or pneumonectomy will cure a certain percentage of patients, the remaining lung will continue to be at risk. Another strategy for properly selected lesions (i.e., those <10 mm, no
evidence of extrabronchial extension, and squamous cell histology) is to consider therapies designed to preserve pulmonary function (PDT or EBBT). Lesions >10 mm or with evidence or suggestion of extrabronchial extension or non-squamous histologies, however, should be considered for surgery, if patients are medically operable. If they are inoperable, prophylactic nodal external radiation plus EBBT should be employed. A suggested strategy is outlined in Figure 16.1. Lesions that are <10 mm and with squamous cell histology would be treated by one of the two therapeutic modalities most conserving of pulmonary function. Such a strategy would facilitate the clinical development of EBBT for occult lung cancers. At our institution, we have treated more than 600 patients on protocols utilizing EBBT, which are described in detail later in the chapter. Of these patients, only 19% were treated by the curative intent protocol and, of these, only five (4%) met the criteria to be classified as radiographically occult carcinomas (Table 16.1). The symptom index score is listed in Table 16.6. This is a semi-quantitative symptom index (described later in the chapter), which allows for description and scoring of the symptoms of the patients during their medical course. Five patients are included, with a total of six lesions treated, for a rate of 1.2 lesions per patient. Two out of five patients (40%) died from intercurrent disease. One patient died of recurrence at 676 days. At the time of analysis, two of the patients were alive with no evidence of disease (NED) at 1228 and 650 days. In contrast to other reported studies, these patients were treated with a more conventional approach utilizing the current curative intent protocol described later in detail, which allowed for the delivery of 60 Gy or 64 Gy, with external radiation therapy given concurrently with their EBBT. In this group of five patients treated with the combination
Figure 16.1 Treatment schema for a proposed clinical trial for occult lung carcinoma with randomization between photodynamic therapy (PDT) and endobronchial brachytherapy (EBBT). (Ext. = external radiation.)
Endobronchial brachytherapy for stages I, II, and III or recurrent lung cancer 229
Table 16.1
Brachytherapy for radiographically occult carcinoma
1
TisRUL T1RLL
Dyspnea (1) Cough (1)
PDTx4
7.5Gyx3/60Gy
2
T1 LUL
Hemoptysis (2) Dyspnea (2) Cough (1)
COPD- severe
7.5Gyx3/60Gy
3
T1 RUL
Hemoptysis (1) Dyspnea (4) Cough (1)
COPD -severe
5Gyx4/60Gy
676
Died -local recurrence None
4
T2RLL
Cough (2)
COPD -severe
5 Gy x 3/64 Gy
650
Alive NED
None
5
T2RMS
Hemoptysis (1) Dyspnea (3) Cough (4)
COPD -severe
5Gyx3/64Gy
104
Died-intercurrent
None
202
1228
Died-intercurrent
None
Alive NED
None
a
The numbers in parentheses refer to grade and symptom as described in Table 16.6. Rx = radiation treatment; NED = no evidence of disease; RUL = right upper lobe; RLL = right lower lobe; LUL = left upper lobe; RMS = right main stem; COPD = chronic obstructive pulmonary disease.
of external-beam radiation and EBBT, there were no complications. 16.6 RESULTS OF ENDOBRONCHIAL BRACHYTHERAPY FOR STAGES I, II, AND III OR RECURRENT LUNG CANCER Bronchogenic carcinoma often leads to symptoms secondary to airway involvement. These can include obstructive pneumonia, atelectasis, hemorrhage, cough, or interference with airflow. YAG laser photo resection has immediate results, but is restricted to central airways and to use by highly experienced operators. Unfortunately, conservative photoresection of the tumor without other interventions usually results in a fairly rapid re-obstruction of the airway. The principle of endobronchial brachytherapy is that a very high dose of radiation is delivered in a short period of time to the tumor, with sparing of normal adjacent tissues. Historically, it is of interest that as early as 1922, Yankauer [24] placed capsules of radium through a rigid bronchoscope into the region of bronchogenic carcinoma in two patients with airway obstruction. In 1933, Kernan [1] used radon seed implantation. With the advent of fiberoptic bronchoscopy, new techniques became possible. The first reported use for transbronchial implantation was by Moylan [25] in 1983, using gold-198 seeds. Also in 1983, Mendiondo [7] used an afterloading tube placed with the aid of the fiberoptic bronchoscope and then placed an iridium-192 source into the tube. A total of 3000 cGy delivered at a 5 mm depth was prescribed, with an average treatment time of 10 h. Using this technique, it was noted that patients had 'significant improvement' in bronchial obstruction. Schray et al. [6,26] reported on afterloading with iridium-192 in 1985 and again in 1988. A single
catheter was used in 65 patients. Fifty-nine of his patients had either prior or concurrent external radiation, and 40 of the patients underwent YAG laser photoresection. Approximately 60% of the patients showed a response, 20% were stable, and 20% had progression at follow-up bronchoscopy. In his second report in 1988, 11 patients had developed either fistulas or massive hemorrhage, seven of which he felt were secondary to treatment. He also noted that six of the seven patients had undergone YAG photoresection and that this may have contributed to the complications. High dose-rate (HDR) remote afterloading was first reported in the American literature by Seagren et al. in 1985 [27]. All patients in this reported series had endobronchial carcinoma and had previously received a minimum external-beam radiation therapy dose equivalent to 5000 cGy prior to brachytherapy. Each patient had bronchoscopically documented disease with local symptoms and a Karnofsky performance status of 50 or greater. The HDR remote afterloading unit was a Brachyton using a 3 mm diameter cobalt-60 source with an average strength of 0.7 Ci. The unit had the ability to oscillate the source up to a maximum of 16 cm. A single catheter was used and a dose of 1000 cGy in a single fraction was delivered at a prescription depth of 10 mm. The total time for treatment ranged between 12 and 27 minutes. Seagren et al. reported on a total of 20 patients treated between 1982 and 1983. Four patients had first received YAG laser photoresection prior to the brachytherapy procedure. Complete palliation of symptoms was seen in 25% of the patients, partial in 69% for a combined partial, or complete palliation of symptoms in 94% of the patients. In six of these patients, palliation was long lasting, with no recurrence of symptoms. In 12 patients, these symptoms recurred or progressed, with a mean time to recurrence of 4.3 months.
Table 16.2 Compilation of endobronchial HDR studies
Seagren(AI) Macha (A2) Nori (A3) Burf (A4) Pass (A5) Miller (A6) Stout" (A7) Khanavkar(AS) Aygun (A9) Bedwinek(A10) Gauwitz(A11) Mehta(A12) Sutejda(A13) Speiser(A14)
1985 1987 1987 1990 1990 1990 1990 1991 1992 1992 1992 1992 1992 1993
Zajac(A15) Tredaniel(A16) Chang(A17) Gollinsa(A18)
1993
Cotter (A1 9) Goldman (A20) Marsh (A21)
1993
Nori (A22) Pisch (A23) Macha (A24) Huber(A25)
1993
Gustafson (A26) Sur(A27) Speiserc (A28)
1995
1994 1994 1994
1993 1993
1993 1995 1995
1995 1995
20 56 15 50 15 88 100 12 62 38 24 31 31 144 151 82 51 76 406
65 (17 intermediate, 48 HDR 20 12 32 39 365 93 46 14 485 total 47IDR 144 HDR 151 HDR
10at10mm 7.5 at 10 mm 20 at 10 mm 15-20 at 10 mm 5-36 at 10 mm 10at10mm 15-20 at 10 mm
Sat 5 mm 5at10mm 6 at 10mm 15at10mm 4 at 20 mm 10at10mm 10at10mm 7.5 at 10 mm 10^7 at 10 mm 7 at 10 mm 7 at 10 mm 10-20 at 10 mm
1 3 3 1 1-6 3 1 2-8 3-5 3 2 4b 3 3 3 1-5 2-8 3 1 (94%) 2 (6%) 2-4 1 1
94 74 80 50-86 75 NA 50-86 67 NA 76 88 88 82 85-99
NA 88 88 46 NA NA 46 NA 36 64 83 71-100 NA NA
100 75 NA 88 NA 80 NA 100 76 82 100 85 NA 80
82 55-85 79-95 88 (H) 62 (C) 60 (D) 92 (S) 66 (PS) 37(C)89(D)100(H) 92 (tumor response)
NA NA NA 46
74 84 87 NA
46 58
63 55
28 7 0 0 0 0 0 50 15 32 4 3 32 7 8 0 10 4 8
2.7-10 at 10 mm 15at10mm 26-53 at 10 mm; high activity125I 4-5 at 10 mm 10at10mm 5-7.5 at 10 mm 3.8 at 10 mm or 7.2 at 10 mm 7 at 10 mm 10at10mm
3-4 1-2 1-6 4 2 3 1-2
100(H)86(C)100(P) 93 (H) 80 (C) 20 (P) 69 Improved local control
Local control NA NA
83 NA NA
0 3 21 21
74 100(H)
69 NA
92 NA
7 7
10 at 5 mm 10at10mm
3 3
99 (H) 99 (P)
NA
53 (CR)
4.4 (CUR)
29 (PR) 82 (TR)
7.3 (PAL)
7.5 at 10 mm
3
86 (D)
2 0 8
9.1 (REC)
143 HDR
Saito (A29) Delclos (A30)
1996 1996
40 81
Huber(A31)
1997
98 42 56
Corsa (A32) Perol (A33)
1997 1997
29 19
Ornadel (A34)
1997
117
Ofiara (A35)
1997
30
Hennequin (A36)
1998
149
Total
1985-1998
3176
7.5at10mm 5at10mm 5 at 3-9 mm 15at6mm
3 4 5 2
7.3 (all)
85 (C) NA 32 excel lent 31 moderate 21 minimal 84 total median survival
External only 60" External + internal Internal 4.8 at 5 mm 5-15at10mm 76 at 10 mm
22-30
30 week
2 1-4 3-5
15at10mm
1
43 week 100(H)70(D)46(C) 79 biopsy (-) actual survival 1 year = 78 2 years = 58 median = 28 months 46 (C) 50 (D) 62 (H) 54 (PS)
8at10mm
3
4-7 at 5-1 5 mm
2-6
NA
100
NA
NA
0
0
14
83
46 (C) 33 (D) 20 (P) 79 (H) 60
43 79
20-100
36-88
79
63
55-100
19 6.9 10.5
9.4 20 2 years actuarial NR 6.7 (Multivariate analysis, endobronchial tumor length was significant p = 0.02) 0-50 Mean 10.1
C = cough, CR = complete response, CUR = curative, D = dyspnea, H = hemoptysis, IDR= immediate dose rate, P = pneumonia, HDR = high dose rate, NR = not recorded, PAL = palliative, PR = more than 50% response, PS = performance status, REC = recurrent, S = stridor, TR = CR + PR. Patients in more than one publication were counted once. a Same institution. b Four treatments in 2 days. c Speiser reporting different protocols. d Mean dose ~ 50 Gy ± 13 Gy. For references A1-A36, see p. 241-2.
232 Endobronchial brachytherapy in the treatment of lung cancer
Joyner and colleagues [28] first reported on the use of iridium-192 solid wire afterloading for endobronchial treatment. They treated 14 patients with stage III nonsmall cell lung cancer with a combination of neodymium-YAG laser photoresection followed by endobronchial radiation and external-beam radiation therapy treatments. Brachytherapy treatment times were approximately 8-20 h to deliver 3000 cGy at a prescription depth of 5 mm. Rooney and colleagues [29] discussed their protocol for anesthesia in the use of HDR endobronchial treatment utilizing a Gamma Med unit with an iridium-192 source. This technique allowed for the stepping sequence of the source in up to 12 dwell positions spaced at either 0.5 or 1.0 cm. The patients were treated on a weekly basis, receiving 600 cGy per week at a prescription depth of 1 cm with a maximum of five treatments. This was a technique paper, and clinical results were not presented. Macha et al. [30] published a report in 1987 of HDR afterloading with an iridium-192 source in which the treatment was fractionated into three treatments through a single catheter. The total dose was 1500 cGy at a 5 mm prescription depth. Patients had laser, external radiation, and/or chemotherapy and had a high complication rate. Patients who present with atelectasis may particularly benefit from application of this approach. Re-expansion of the lung after endobronchial brachytherapy may allow for better-tailored external-beam radiation fields. Bastin et al. [31] reported their ability to spare an average of 32% of the ipsilateral lung volume using this technique. In most of the studies described above, the procedure consisted of a single catheter for intraluminal brachytherapy. Additional physics factors must be taken into account, including the choice of the radioactive isotope, the dose at the prescription depth relative to the radius and inverse square law, as well as correction factors for the attenuation in water equivalent tissue versus air. Iridium-192 can be fabricated in very small physical sizes of high activity, allowing passage through catheters with an internal diameter of 1.5 mm or less. In addition, iridium-192 has virtually no significant correction for attenuation in water versus air up to a distance of 5 cm
from the source. This makes iridium-192 an ideal radioactive isotope to be used within tissues with mixed air/water density interfaces, such as intraluminal brachytherapy. The major drawback of iridium-192 is its relatively short half-life of 74 days, such that over a period of months, its activity changes considerably. However, this disadvantage can be addressed by the frequent replacement of the sources to keep the source strength well within the high dose-rate range. Table 16.2 is a compilation of most, but not all, of the studies published on HDR remote afterloading endobronchial brachytherapy. It covers the years 1985 to 1998 for a total of 3176 patients. The reader can see in the tabulation the different doses used per fraction, the wide range of fractions, as well as the dose prescriptions at various depths. Of interest is that, while the variation for fatal hemoptysis is extremely large (ranging from 0% to 50%), the mean is 10.1%. This rate appears to depend on multiple factors, such as total number of patients studied, extensiveness of follow-up, and aggressiveness of the treatment. For a small series, such as that of Khanavkar [32], the complication rate - 50% incidence of fatal hemoptysis in 12 patients - is probably not representative. Similarly, there are reports of no fatal hemoptysis, such as that of Burt et al. [11], reporting on 50 patients. However, when the same medical group updated their series and reported on a total of 406 patients [33], their rate of fatal hemoptysis was 8%. Some authors have attributed fatal hemoptysis to aggressive treatment, such as Bedwinek [34], who reported a 32% rate of hemoptysis. However, Macha [35] and Speiser [36] reported on 365 and 485 patients with lower rates of fatal hemoptysis, 21% and 8% respectively. Further information indicated that many of these cases were secondary to biologic progression of the carcinoma itself, suggesting that the treatment may not have been aggressive enough. In 1986, Speiser and Spratling developed a series of scoring systems (Tables 16.3-16.6) for selecting patients for treatment, as well as for outcome analysis. Their initial report [37] was based on experience with an intermediate dose-rate unit, which was a modification of a low dose-rate unit, to allow a higher specific activity of
Table 16.3 Influence of performance status on patients with inoperable lung cancer
1 2 3 4
100 80-90 60-70 40-50 20-30
Asymptomatic Symptomatic: fully ambulatory Symptomatic: in bed less than 50% of the day Symptomatic: in bed more than 50% of the day but not bedridden Bedridden
Protocol 233 Table 16.4 Weight loss score based on percentage of loss of body weight within 6 months preceding diagnosis
0 1^1.9 5-9.9 10-19.9 >20.0
0
1
2 3 4
16.7
Table 16.5 Obstruction score
Trachea Main stem Lobar bronchi
10 6 2
5 3 1
2 1 -
Atelectasis/pneumonia received additional 2 points per lobe.
Table 16.6 Symptom index scoring system
Dyspnea 0 None 1 Dyspnea on moderate exertion 2 Dyspnea with normal activity, walking on level ground 3 Dyspnea at rest 4 Requires supplemental oxygen Cough 0 1 2 3 4
iridium-192 for more rapid delivery of dose. In some of the low dose-rate protocols, delivery of radiation took as long as 60 h. The increase of iridium-192 activity allowed the treatment time to be decreased to 1.5-4 h. This range was based on (1) shorter times when the sources were new versus longer times after the sources decayed, and (2) the use, for the first time, of multiple catheters to deliver treatment. All treatments were performed in the outpatient setting and followed a protocol that is outlined in the next section.
None Intermittent, no medication necessary Intermittent, non-narcotic medication Constant or requiring narcotic medication Constant, requiring narcotic medication but without relief
Hemoptysis 0 None 1 Lessthan2/week 2 Less than daily but greater than 2/week 3 Daily, bright red blood or clots 4 Decrease of hemoglobin and/or hemotocrit > 10%; greater than 150 cm, requiring hospitalization or transfusion Pneumonia/elevated temperature 0 Normal temperature, no infiltrates, white blood count less than 10 000 1 Temperature greater than 38.5 °C and infiltrate, white blood count less than 10000 2 Temperature greater than 38.5 °C and infiltrate and/or white blood count greater than 10000 3 Lobar consolidation on radiograph 4 Pneumonia or elevated temperature requiring hospitalization
PROTOCOL
The protocol alluded to earlier in the chapter was initiated in 1986 when EBBT was transformed from low dose-rate manual afterloading to medium dose-rate remote afterloading procedures. This was a transitory step of short duration, lasting for 9 months. The HDR remote afterloader was, in fact, a Nucletron Selectron low dose-rate remote afterloader that was modified to accept a longer source train and a higher level of radioactivity. The activity was typically maintained at greater than 740 MBq, or 20 mCi cm"1. Dose rates initially were calculated at a 5 mm depth perpendicular to the source train, and were in the range of 5-10 cGy min-1.
16.7.1
Eligibility
Eligibility for the protocol included the following. 1. Disease must involve the trachea, main stem, or lobar bronchi. Involvement of the segmental bronchi without involvement more proximal was not considered sufficient for entry into the protocol. 2. The central airway disease must be intraluminal, visualized and biopsied via bronchoscopy. Patients requiring transbronchial biopsy were ineligible for the protocol. 3. Patients must have significant symptomatology within the four symptom groups consisting of cough, dyspnea, signs and symptoms of obstructive pneumonia, and/or hemoptysis. Evaluation of the patients meeting eligibility criteria for the EBBT protocol schedule was reviewed within the context of all patients diagnosed with lung cancer in the referral area from 1986 through 1996. This involved the greater Phoenix/Maricopa County area and, based on the Tumor Registry, an incidence of approximately 9000 cases of lung cancer during the 10-year period was calculated. Of these, only 19% received radiation and 16% of that group, or 3% of all patients, were treated on protocol, while an additional 11% of patients receiving radiation (or 2% of all patients) were treated with brachytherapy off protocol. Thus, 27% of all patients receiving radiation, or 5% of all diagnosed lung cancer
234 Endobronchial brachytherapy in the treatment of lung cancer
patients, received brachytherapy. For patients on the curative protocol, these figures were 3% and 0.5%, respectively.
16.7.2 Indications Indications for treatment are outlined in Table 16.7.
16.7.3
Protocol 1.0 curative intent
To be eligible for this protocol, patients must not have had prior radiation within the thoracic area, which would preclude the adequate delivery of a full dose of external radiation. Patients must be inoperable and have a primary lung carcinoma with non-small cell histology. Stages accepted were Tl,2,3, Nl,2, MO. These correspond to stage groupings I, II, and Ilia. Performance status using the East Coast Oncology Group (ECOG) fourtiered system must be 0, 1, or 2 and weight loss using a four-tiered weight loss system, likewise, must be 0, 1, or 2 and correspond to weight losses of 0, less than 5% or less than 10%, respectively, of the patient's weight in the 6 months prior to diagnosis. The rational for selection of this level of weight loss is described in 'Oncologic assessment using the four-tiered scoring system' [38].Patients were treated within groups 1-4, with dose modifications as described in Table 16.8.
16.7.4
Protocol 2.0 palliative intent
PROTOCOL 2.1 Eligibility for these patients includes: primary lung cancer with non-small cell histology, and stage T4, N3 and/or Ml disease. These corresponded to stage groupings Illb and IV. In addition, the patients ineligible for protocol 1.0 because of performance scores of 3 or 4 or a weight loss of 3 or 4 (>10% >20%) were reallocated to this protocol. Patients were treated within groups, characterized by dose, as described in Table 16.8. PROTOCOL 2.2
Primary lung cancer consisting of small cell histology, both limited and extensive; primary lung cancer with contralateral metastatic disease involving the endobronchial mucosa; and non-lung primaries with metastases primarily to the mucosa were treated within this category. Patients were treated within the group characterized by dose as described in Table 16.8.
16.7.5
Protocol 3.0 recurrent patients
All patients who had received prior radiation for a curative intent for carcinoma of the lung were included within this category. Patients were treated within the
Table 16.7 Indications for treatment with endobronchial brachytherapy
Tumors must be seen and biopsied by bronchoscopy (intraluminal)
Tumors presenting with extrinsic compression of the airway as seen by bronchoscopy and the biopsy must be performed transbronchially (extraluminal)
Intraluminal brachytherapy delivers a very high dose to tumor close to the source axis; extraluminal disease due to its much greater distance from the axis, would lead to unacceptable doses to the bronchial mucosa and surrounding structures
Tumors must be in the central airways which are defined as the trachea, main stem, and lobar bronchi
Tumors in peripheral airways which are defined as segmental bronchi or beyond
Significant symptomatology is most often caused by disease in central airways; treatment of small peripheral airways leads to stenosis of those airways
Tumors in central airways causing significant symptomatology
Patients with significant pre-existing dyspnea unrelated to carcinoma; patients with dyspnea secondary to effusion, or large extrinsic masses
Patients with symptoms second to disease other than central airway disease are not expected to improve with intraluminal brachytherapy
In-situ carcinoma for inoperable patients
Patients entered into national protocols using other modes of treatment, i.e., photodynamic
Preserves lung and pulmonary function; excellent treatment for multifocal disease
Pre-op. for submucosal spread from a peripheral/central lesion
Patients should be good candidates for lobectomy or pneumonectomy
Treatment provides a clear margin for surgery
Protocol 235 Table 16.8 Modification of doses by year for the curative, palliative and recurrent protocols
Curative protocol 1 2 3 4 Palliative protocol 1
1986-1988 1988-1990 1990-1992 1992-1994
6000 6000 6000 6400
30 30 30 32
1000 1000 750 500
5 10 10 10
3 3 3 3
MDR HDR HDR HDR
1986-1988
3750
15
2
1988-1990
3750
15
3
1990-1992
3750
15
4
1992-1994
3750
15
1000 1000 1000 1000 750 750 500 500 750
5 5 10 10 10 10 10 10 10
3 3 3 3 3 3 3 4 3
MDR MDR HDR HDR HDR HDR HDR HDR HDR
Recurrent protocol 1 2 3 4
1986-1988 1988-1990 1990-1992 1992-1994
1000 1000 750 500
5 10 10 10
3 3 3 4
MDR HDR HDR HDR
group characterized by dose, as described in Table 16.8. Group I patients were treated with medium dose rate. In the palliative protocol, the brachytherapy was constant and the use of external radiation was optional, at the discretion of the treating oncologist. Its use was restricted to patients with extrinsic disease that caused a significant contribution to the level of obstruction and/or symptomatology.
16.7.6
Results
The following results incorporate 600 patients treated in the curative, palliative, and recurrent protocols outlined previously. In each of the successive periods of the operation of the protocol, the eligibility factors for the curative, palliative, and recurrent protocols have remained constant. All patients treated in curative protocols with external radiation received 2 Gy per fraction and, in palliative protocols, 2.5 Gy per fraction. If patients received concurrent brachytherapy and external radiation, the two treatments were not given on the same day. For the curative protocol, brachytherapy was delivered during weeks one, three and five. For palliative or recurrent protocols, brachytherapy was delivered weekly for three or four fractions, depending on the protocol. The distribution of patients into the protocol groups was as follows: curative 19%, palliative 48%, and recurrent 33%. The age distribution of the patients had a median of 68 years and a mean of 67.1 years. Most of the
patients fell within the range of 60-80 years old. The gender distribution was 62% male and 38% female. The percentage of female patients increased from 28% in Group 1 to 41% in Groups 3 and 4. The breakdown for male/female patients was similar to that of all patients presenting with carcinoma of the lung within the geographical treatment area. Squamous cell carcinoma is by far the most common cell type, overall, in the study (49%), and even to a greater extent for those treated in the curative protocol (70%). This percentage is considerably higher than is currently being seen in newly diagnosed outpatients with lung cancer (27%). This fits with prior observations that squamous cell carcinomas tend to be more central and adenocarcinomas more peripheral. The use of laser photoresection predated the wide use of HDR brachytherapy for airway carcinoma. In this study there was a gradual decrease in photoresection from an initial 32% to 16% in the latter part of the study. It is currently estimated that less than 5% of patients with central airway disease require laser photoresection. The protocol required that patients must have one or more of the four primary symptom complexes in order to be included in this study. The incidence of the symptoms in the study were: cough, 99%; dyspnea, 97%; hemoptysis, 64%; and the signs and symptoms of obstructive pneumonia, 49%. Using the Four-tiered symptom index as outlined in Table 16.6, the severity of the symptoms was weighted and the total weighted scores was subsequently normalized to 100%. Response for each symptom score is related to each brachytherapy
236 Endobronchial brachytherapy in the treatment of lung cancer
procedure and the first follow-up bronchoscopy (Figure 16.2). Hemoptysis had the most dramatic and rapid of the responses with improvements of 70%, 90%, and >99% at each intervention point. Pneumonia improvement was only slightly less dramatic, with responses of 57%, 85%, and >99%. Improvement in dyspnea occurred in 36%, 54%, and 86% respectively. The fourth symptom, cough, showed improvements of 32%, 52%, and 85%, respectively. The improvements in hemoptysis and pneumonia were commonly seen within the first 24 h following the first brachytherapy procedure. Patients, who were admitted to the hospital with obstructive pneumonia and/or sepsis, or with severe bleeding requiring transfusion, generally had a prompt response. In the palliative protocol, the use of concurrent external radiation with brachytherapy was optional. When the weighted responses were measured for brachytherapy only, versus brachytherapy and external radiation, the results in terms of improvement at follow-up were as follows: hemoptysis, 94% and 97%; pneumonia, 86% and 82%; dyspnea, 54% and 48%; and cough, 51% and 57%. There was no statistical difference in response for each symptom group for each of these two therapies. The use of brachytherapy only was sufficient to provide palliation, without the need to add supplemental external radiation.
Airway obstruction scores (as seen in Figure 16.3), were analyzed in a different fashion. All of the scores were converted into median scores, which were normalized to 100%. These were obtained for each brachytherapy procedure and at the first follow-up bronchoscopy. The median score was normalized to 100% and the residual level of obstruction expressed as a percentage. Any tissue including inflammatory tissue was included as part of the obstruction score. The curative patients and the palliative groups fared better than the recurrent group, with scores of 12%, 12.5%, and 19% respectively (Figure 16.3). This is not unexpected, considering that patients with recurrent carcinoma have had previous external radiation, which may select for a slightly more radioresistant carcinoma. It is interesting that neither the use of concurrent external radiation nor the use of laser photoresection led to improved clearing of obstruction. Thus, as in the symptom index results, the addition of external radiation or laser resection did not add to clearing endobronchial disease. The survival of patients by protocol group was, for the curative patients, 10%, and for palliative-recurrent, 5% at 5 years. The cause of death shows a significant local failure rate in all categories. These rates were 31% and 30%, respectively, for the curative and palliativerecurrent protocols. As has been seen in numerous other studies, despite gradual increasing doses of radiation Figure 16.2 Response in the symptom index as measured by the decrease of the symptom index scores comparing the initial score at the first brachytherapy (normalized to 100%) and the scores at subsequent encounters as the percent residual.
Figure 16.3 Response in the obstruction score with the mean initial score normalized to 100% at the first brachytherapy and the percent residual at the first bronchoscopicfolloW-up (F/U).
Protocol 237
over the last several decades, local disease continues to be a significant problem. Survival curves for curative versus the palliativerecurrent patient illustrated in Figure 16.4 are calculated from the date of diagnosis and date of the first brachytherapy procedure. There was no statistically different result between these two groups when analyzed
from the date of diagnosis (p = 0.1). However, from the date of the first brachytherapy treatment, thep-value was <0.0001 comparing the curative versus the palliativerecurrent patients. The recurrent patients are selfselected by the fact that they lived long enough to develop a symptomatic recurrence. This bias was eliminated when analyzing survival by date of treatment.
Figure 16.4 Survival cures and p values for curative, palliative, recurrent, and combined group measured from diagnosis and from date of first treatment.
238 Endobronchial brachytherapy in the treatment of lung cancer
16.8
COMPLICATIONS
The complication most often described when discussing endobronchial carcinoma and its treatment is that of fatal hemoptysis. This is defined as bleeding which is usually caused by erosion into the right or left pulmonary artery, with subsequent exsanguination within the tracheal bronchial tree. This is a known complication of the progression of untreated or inadequately treated carcinoma involving the tracheal bronchial tree, usually in the region of the upper lobe bronchi, due to their proximity to the pulmonary arteries. With the advent of intraluminal treatment, it must also be considered as a possible complication of this modality. In the study of Speiser and Spratling [37], the overall rate of fatal hemoptysis is 6%, with the median time from diagnosis until death of 14 months. However, when measured from the date of the first brachytherapy until death, the median for the entire group is only 5 months. Recurrent patients, as would be anticipated, have the highest rate of fatal hemoptysis, at 9%, while the curative and palliative patients have rates of 5%. Of interest is that, in the different dose groups, there was an increase in fatal hemoptysis as the dose was increased from Group 1 to 2. However, there was paradoxically a slight further increase in the rate of hemoptysis as the dose was initially reduced, and then a decrease in the rate of fatal hemoptysis (2%) with further dose reduction. No clear dose-response relationship could be identified. In the palliative protocol, 41% of the patients were treated with brachytherapy only, and 59% with brachytherapy and external radiation. The rate of fatal hemoptysis was 5.5% in the brachytherapy-only group, versus 4% in the combined group who received higher doses of radiation. The most common side-effect related to intraluminal treatment is radiation bronchitis and stenosis. This was first described by Speiser and Spratling [37].Table 16.9 is a definition of the various grades of radiation bronchitis and stenosis. The incidences of radiation bronchitis and stenosis by group were 9%, 12%, 14%, and 14%, respectively. By protocol, they were 23%, 12%, and 8%, respectively, for the curative, palliative, and recurrent protocols. The rate is clearly highest in the curative patients. However, the only significant factor predicting for this response was length of follow-up. Although the true incidence of this complication is not known for patients receiving external radiation only, it appears to be a complication primarily of intraluminal brachytherapy, with its very high mucosal doses. For palliative patients, this complication was studied in those patients receiving brachytherapy only versus those receiving brachytherapy with concurrent external radiation, and the incidences were 17% and 10%, respectively. Whereas the incidence was slightly higher in the brachytherapy-only group, the median time to occurrence was slightly longer in this group, at 3.6 months
Table 16.9 Grades of radiation bronchitis and stenosis (RBS)
1 2
3
4
Fibrinoid membrane without significant luminal obstruction: no symptoms Increase of exudation and fibrous membrane with mild obstructive symptoms requiring therapeutic intervention such as simple debridement or medical treatment Characterized by severe inflammatory response with marked membranous exudate including fibrosis requiring multiple debridements A greater degree of fibrosis resulting in stenosis with decreased luminal diameter requiring laser photo resection, balloon or bougie dilation, and/or stent placement
versus 2.5 months for the patients receiving combined treatment.
16.9
MANAGEMENT OF COMPLICATIONS
The treatment of brachytherapy-related hemoptysis is the same as that of hemoptysis from other causes. This includes bed rest, codeine, and/or transfusion for hemodynamic stability, avoidance and/or reversal of anticoagulants, and vascular embolization, electrocautery or lasocautery to reduce bleeding volume. One of the largest reported series of 406 patients treated with palliative brachytherapy alone is from the Christi Hospital in Manchester, England [39]. Of these patients, 322 with inoperable non-small cell lung cancers were treated with a single fraction of HDR, with a total dose of 15-20 Gy delivered intraluminally, with the dose prescription 1 cm from the central axis of the catheter. Patients were evaluated 6 weeks after completing the HDR treatment regarding their symptoms, including stridor, hemoptysis, cough, dyspnea, pain, and pulmonary collapse. In addition, at various times following the brachytherapy procedure, 83 bronchoscopies were conducted on 55 patients. Massive hemoptysis leading to death occurred in 8% of the patients (32/406). Cox multivariate analysis revealed that treatment-related factors associated with subsequent massive hemoptysis were brachytherapy dose >15 Gy, prior laser therapy, second brachytherapy treatment, and concurrent external-beam radiation therapy. Twenty of the 25 patients whose deaths were assessable and related to hemoptysis had recurrent and/or residual tumor suspected at the hemoptysis site. The chronology of the massive hemoptysis leading to death occurred between 9 and 12 months after completion of the HDR procedure. This was in stark contrast to deaths from all other causes, which usually occurred 3-6 months after completion of the HDR procedures [40].
Pros and cons 239
Brachytherapy-related bronchitis and stenosis are managed depending on the level of severity of the reaction and/or stenosis. This can include observation for mild treatment-related bronchitis for the least symptomatic presentation, versus active treatment for its more debilitating form, with oral and/or aerosol administration of steroids, aerosol-administered bronchodilators, codeine-based or narcotic-based cough suppressants, and antifungal or antibiotic therapies as indicated. More aggressive interventional management for debilitating and/or life-threatening levels of bronchitis and/or stenosis may be managed with balloon and/or bougie dilatation, laser photoresection, bronchoscopic debridement, and/or placement of intraluminal stents. Although lung brachytherapy has been advocated by radiation oncologists for the past 20 years, recent technological developments in the area of HDR brachytherapy, such as the design of small, high-activity iridium-192 sources and remote afterloading machines, have prompted renewed interest in HDR endobronchial brachytherapy. The specific role of lung EBBT is not clearly defined within the standard and/or uniform community practice. There is an ongoing evolution for the selection criteria to identify those patients most likely to benefit from EBBT as part of definitive therapy. The American Brachytherapy Society (ABS) HDR Consensus Guidelines [41] currently state that, although endobronchial brachytherapy has demonstrated efficacy for the symptomatic relief of bronchial obstructions and hemoptysis, either alone or in combination with externalbeam radiation therapy, the curative benefit of brachytherapy in addition to conventional externalbeam radiation therapy and/or chemotherapy has not been proven. The ABS recommends that brachytherapy for the definitive treatment of lung cancer is done within the context of controlled clinical trials. Outside of clinical trials, the ABS suggests that brachytherapy be reserved for palliative treatments alone. Although the guidelines do not clearly state the indications for additional externalbeam radiation in newly diagnosed lung cancer patients, EBBT alone is recommended for recurrences after fulldose external-beam radiation therapy treatments have been administered. No single-dose fractionation scheme has been identified which provides a superior therapeutic ratio. Dose specifications to be prescribed have been recommended to a depth of 1 cm from the source center for uniform prescription dosimetry comparisons. A study by the Radiation Therapy Oncology Group (RTOG) evaluated the palliation provided by externalbeam radiation to patients with newly diagnosed nonmetastatic, non-small cell lung cancers [42].This study, by Simpson et al, demonstrated that a short course of external-beam radiation therapy delivering 30 Gy in ten fractions provided relief of hemoptysis in 74% of patients, of cough in 55% of patients, and of dyspnea in 43% of patients. Median survival was around 6 months. Compared to this RTOG study, brachytherapy alone
appears to provide palliation equivalent to or better than external-beam radiation, with a similar survival outcome. Given this fact, brachytherapy may give more prompt symptomatic relief of obstructive symptoms, in a more cost-effective manner. In a small group of 19 patients treated with HDR to a total dose of 15 Gy [43], a detailed assessment with rigorous testing was performed both before and after administration of the HDR brachytherapy. This enlisted chest X-rays, computerized tomography scan of the thorax, direct bronchoscopic evaluation, objective obstruction index scoring, 5-min walking stress tests, isotope ventilation and profusion lung scanning, and formal pulmonary function tests with maximum inspiratory and expiratory full volume measurements. Symptomatic relief was reported in 17 of the 19 patients. Atelectasis of a collapsed lobe or lung reported in 13 patients was demonstrated to have reinstituted ventilation in nine cases by radiographic imaging. Bronchoscopic evaluation of luminal patency demonstrated improvement in 18 of the 19 patients. Isotope lung scans showed significant increase in the percentage of total lung ventilation and perfusion in the abnormal lung. This rigorous study demonstrated the high correlation between objective and subjective improvement of the presenting symptomatology in these patients. In addition, it confirmed the palliative benefit of brachytherapy, which has been described in larger groups of patients. Further prospective studies of brachytherapy and external-beam radiation therapy are clearly needed to rigorously document treatment efficacy and toxicity, as well as cost-benefit and quality of life analyses in this setting.
16.10
CONCLUSIONS
Endobronchial brachytherapy is an excellent method of palliative treatment for patients who are symptomatic from endobronchial disease as part of definitive therapy for curative intent patients; however, a survival advantage is not shown. This group of patients should be randomized to external-beam radiation therapy versus external-beam plus intraluminal radiation to evaluate prospectively any survival advantage. Further studies will also be necessary to determine the optimal dose and number of fractions that will provide the greatest patient benefit, the lowest morbidity, and the lowest cost of treatment.
16/11
PROS AND CONS
Pro Endobronchial brachytherapy is ideal for delivering very high doses of radiation to neoplastic tissue in or within a 1 cm radius of the main airway.
240 Endobronchial brachytherapy in the treatment of lung cancer Con Because brachytherapy doses depend on the inverse square law (that is, the dose decreases by the square of the distance), neoplastic tissue >1 cm from the airway is not effectively treated. Pro Endobronchial brachytherapy provides relief of airway obstruction to a greater extent and significantly faster than external radiation. Con Obstruction of airways by extrinsic compression is best treated by external radiation. Pro Endobronchial brachytherapy can be used in the trachea main stem, lobar and segmented bronchi to an extent greater than feasible with the YAG laser. Con YAG laser works immediately (endobronchial brachytherapy takes 4-24 h for a partial response) and is sometimes necessary for very bulky exophytic tumors. Pro For palliation of airway signs and symptoms due to intrinsic disease, endobronchial brachytherapy provides excellent relief with minimal morbidity. Con Extrinsic compression such as nodal or parenchymal masses or pleural effusion can nullify benefits of brachytherapy, unless these other factors are adequately dealt with. Pro Endobronchial brachytherapy can cure occult cancer of the lung. Con Occult cancer has a high rate of synchronous lesions that must also be identified and treated.
9.
10. 11.
12.
13.
14.
15.
16. 17.
REFERENCES 18. 1. Kernan, J.D. (1933) Carcinoma of the lung and bronchus. Treatment with radon implantation and diathermy. Arch. Otolaryngol., 17,457-75. 2. Meredith, W.J. (ed.) (1967) Radium Dosage. The Manchester System. Edinburgh, Livingstone. 3. Pool, J.L (1961) Bronchoscopy in the treatment of lung cancer. Trans. Am. Bronchoesoph. Assoc., 41,128-36. 4. Graham, E.A. and Singer, J.J. (1933) Successful removal of an entire lung for carcinoma of the bronchus. JAMA, 101, 1371^. 5. Henschke, U.K. (1959) Interstitial implantation in the treatment of primary bronchogenic carcinoma. Am. J. Roentgenol. Radium Ther. Nucl. Med., 79,981-7. 6. Schray, M.F., McDougallJ.C, Martinez, A. etal. (1985) Management of malignant airway obstruction: clinical and dosimetric considerations using an iridium-192 afterloading technique in conjunction with the neodymium-YAG laser. Int.J. Radial Oncol. Biol. Phys., 11 (2), 403-9. 7. Mendiondo, O.A., Dillon, M. and Beach, LJ. (1983) Endobronchial brachytherapy in the treatment of recurrent bronchogenic carcinoma. Int.J. Radial Oncol. Biol. Phys., 9, 579-82. 8. Sanderson, D., Fontana, R., Woolner, L etal. (1974)
19.
20.
21.
22.
23.
24. 25.
Bronchoscopic localization of radiographically occult lung cancer. Chest, 65,608-12. Cortese, DA, Pairolero, P.C., Bergstraih, E.J. etal. (1980) Roentgenographically occult lung cancer, y. Thorac. Cardiovasc. Surg., 86,373-80. Martini, N. and Melamed, M.R. (1980) Occult carcinoma of the lung. Ann. Thorac. Surg., 30,215-23. Burt, M.E., Pomerantz, A.H., Bains, M.S.efo/.(1987) Results of surgical treatment of stage III lung cancer invading mediastinum. Surg. Clin. North. Am., 67, 997-1000. Saito, Y., Nagamoto, N., Ota, S. etal. (1992) Results of surgical treatment for roentgenographically occult bronchogenic squamous cell carcinoma./ Thorac. Cardiovasc. Surg., 104,401-7. Durci, M., Komaki, R., Oswald, M.J. etal. (1991) Comparison of surgery and radiation therapy for nonsmall cell carcinoma of the lung with mediastinal metastasis. Int.J. Radial Oncol. Biol. Phys., 21,629-36. Feinstein, A. and Wells, C. (1990) A clinical-severity staging system for patients with lung cancer. Lung Med., 69,1-33. Maki, E. and Feld, R. (1991) Prognostic factors in patients with non-small cell lung cancer: a critique of the world literature. Lung Cancer, 7,27-34. Pater, J. and Loeb, M. (1982) Nonanatomic prognostic factors in carcinoma of the lung. Cancer, 50,326-31. Nagamoto, N., Saito, Y., Sato, M. etal. (1993) Lesions preceding squamous cell carcinoma of the bronchus and multicentricity of canceration - serial slicing of minute lung cancers smaller than 1 mm. TohokuJ. Exp. Med., 170,11-23. Kato, H., Kawate, N., Kinoshita, K. etal. (1991) Photodynamic therapy of early-stage lung cancer: a critique of the world literature. Lung Cancer, 7,27-34. Saito, Y., Sato, M., Sagawa, M. etal. (1994) Multicentricity in resected occult bronchogenic squamous cell carcinoma. Ann. Thorac. Surg., 57,1200-5. Usuda, K., Saito, Y., Kanma, K. etal. (1993) Resected roentgenographically occult bronchogenic squamous cell carcinoma tumor size, survival, and recurrence. Nippon Geka GakkaiZasshi, 94, 631-6. Sutedja, G., Baris, G., van Zandwijk, N. etal. (1993) Highdose rate brachytherapy has a curative potential in patients with intraluminal squamous cell lung cancer. Respiration, 61,167-8. Tredaniel, J., Hennequin, C., Zalcman, G. et al. (1994) Prolonged survival after high-dose rate endobronchial radiation for malignant airway obstruction. Chest, 105, 767-72. Perol, M., Caliandra, R., Pommier, P. etal. (1997) Curative irradiation of limited endobronchial carcinomas with high-dose rate brachytherapy. Chest, 111(5), 1417-23. Yankauer S. (1922) Two cases of lung tumor treated bronchoscopically. N YMed.J., 21,741. Moylan, D., Strubler, K., Unal, A. etal. (1983) Work in progress. Transbronchial brachytherapy of recurrent
References for Table 16.2 241
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36. 37.
bronchogenic carcinoma: a new approach using the flexible fiberoptic bronchoscope. Radiology, 147,253—4. Schray, M.F., McDougallJ.C, Martinez, A. etal. (1988) Management of malignant airway compromise with laser and low dose rate brachytherapy. Chest, 93,264-9. Seagren, S.L, Harrell, J.H. and Horn, R.A. (1985) High dose rate intraluminal irradiation in recurrent endobronchial carcinoma. Chest, 88,810-14. Joyner, L.R.Jr., Maran, A.G.,Sarama, R.etal. (1985) Neodymium-YAG laser treatment of intrabronchial lesions. Chest, 87(4), 418-27. Rooney, S.M., Goldiner, P.L., Bains, M.S. etal. (1984) Anesthesia for the application of endotracheal and endobronchial radiation therapy.7. Thorac. Cardiovasc. Surg., 87(5), 693-7. Macha, H.N., Kock, K.,Stadler, M.rta/.(1987) New technique for treating occlusive and stenosing tumors of the trachea and main bronchi: endobronchial irradiation by high dose iridium-192 combined with laser canalization. Thorax, 42, 511-15. Bastin, K., Mehta, M. and Kinsella, T. (1991) Quantitative analysis of lung-sparing following endobronchial boost in stage III non-small cell lung carcinoma treated curatively [abstract]. Endocuriether. Hypertherm. Oncol., 7,225. Khanavkar, B., Stern, P., Alberti, W. etal. (1991) Complications associated with brachytherapy alone or with laser in lung cancer. Chest, 99,1062-5. Burt, PA, O'Driscoll, B.R., Motley, H.M. etal. (1990) Intraluminal irradiation for the palliation of lung cancer with the high dose rate microselectron. Thorax, 45,765-8. BedwinekJ., Bruton, C, Petty,k.eta\. (1991) High dose rate endobronchial brachytherapy and fatal pulmonary hemorrhage. Int.). Radial. Oncol. Biol. Phys.,22,23-30. Macha, H.N., Wahlers, B., Reichle, C. etal. (1995) Endobronchial radiation therapy for obstructing malignancies: ten years'experience with iridium-192 high dose radiation brachytherapy afterloading technique in 365 patients. Lung, 173,271-80. Speiser, B. (1995) The role of endobronchial brachytherapy in patients with lung cancer. Clin. Pulm. Med., 2(6), 344-52. Speiser, B. and Spratling, L (1990) Intermediate dose rate remote afterloading brachytherapy for intraluminal control of bronchogenic carcinoma. Int.J. Radiat. Oncol. Biol. Phys., 18,1443-8.
38. Speiser, B. (1995) Oncological assessment using the fourtiered scoring system. Curr. Oncol., 2(2), 54-9. 39. Gollins, S., Burt, P., Barber, P. etal. (1994) High dose rate intraluminal radiotherapy for carcinoma of the bronchus: outcome of treatment of 406 patients. Radiotherapy, 33,31^0. 40. Gollins, S.W., Ryder, W.J., Burt, P.A. etal. (1996) Massive haemoptysis death and other morbidity associated with high dose rate intraluminal radiotherapy for carcinoma of the bronchus. Radiat. Oncol., 39,105-16. 41. Nag, S., Abitbol, A.A., Anderson, LL etal. (1993) Consensus guidelines for high dose rate remote brachytherapy in cervical, endometrial, and
endobronchial tumors. Int.J. Radiat. Oncol. Biol. Phys., 27,-\24-\-4. 42. Simpson, J.R. (1985) Palliative radiotherapy for inoperable carcinoma of the lung: final report of a RTOG multi-institutional trial. Int.]. Radiat. Oncol. Biol. Phys., 11,751-8. 43. Goldman, J., Bulman, A., Rathmell, A. etal. (1993) Physiological effect of endobronchial radiotherapy in patients with major airway obstruction. Thorax, 48, 110-14.
REFERENCES PORTABLE 16.2 A1
Seagren, S.L, Harrell, J.H. and Horn, R.A. (1985) High dose rate intraluminal irradiation in recurrent endobronchial carcinoma. Ches\, 88, 810-14. A2 Macha, H.N., Koch, K., Stadler, M. et al. (1987) New technique for treating occlusive and stenosing tumours of the trachea and main bronchi: endobronchial irradiation by high dose iridium-192 combined with laser canalization. Thorax, 42, 511-15. A3 Nori, D., Hilaris, B.S., Tome, M. et al. (1987) Intraluminal irradiation in bronchogenic carcinoma. Surg. Clin. North. Am., 67(5), 1093-102. A4 Burt, P.A., O'Driscoll, B.R., Notley, H.M. et al. (1990) Intraluminal irradiation for the palliation of lung cancer with the high dose rate microselectron. Thorax, 45, 765-8. A5 Pass, D.E., Armstrong, J. and Harrison, L.B. (1990) Fractionated high dose rate endobronchial treatment for recurrent lung cancer. Endocuriether. Hypertherm. Oncol., 6, 211-15. A6 Miller, J.I. and Phillips, T.W. (1990) Neodymium: YAG laser and brachytherapy in the management of inoperable bronchogenic carcinoma. Ann. Thorac. Surg., 50,190-6. A7 Stout, R. (1993) Endobronchial brachytherapy. Lung Cancer, 9, 295-300. A8 Khanavkar, B., Stern, P., Alberti, W. et al. (1991) Complications associated with brachytherapy alone or with laser in lung cancer. Chest, 99,1062-5. A9 Aygun, C., Werner, S., Scariato, A. et al. (1992) Treatment of nonsmall cell lung cancer with external beam: radiotherapy and high dose rate brachytherapy. Int. J. Radiat. Oncol. Biol. Phys., 23,127-32. A10 Bedwinek, J., Bruton, C., Petty, A. et al. (1991) High dose rate endobronchial brachytherapy and fatal pulmonary hemorrhage. Int. J. Radiat. Oncol. Biol. Phys., 22, 23-30. A11 Gauwitz, M., Ellerbroek, N., Komaki, R. etal. (1992) High dose endobronchial irradiation in recurrent bronchogenic carcinoma. Int.J. Radiat. Oncol. Biol. Phys., 23, 397-400. A12 Mehta, M.P., Petereit, D., Chosy, L. et al. (1992) Sequential comparison of low dose rate and hyperfractionated high dose rate endobronchial
242 Endobronchial brachytherapy in the treatment of lung cancer
A13
A14
A15
A16
A17
A18
A19
A20
A21
A22
A23
A24
radiation for malignant airway occlusion. Int.}. Radial Oncol. Biol. Phys., 23,133-9. Sutedja, G.( Baris, G., Schaake-Koning, C. et al. (1992) High dose rate brachytherapy in patients with local recurrences after radiotherapy of non-small cell lung cancer. Int. J. Radial Oncol. Biol. Phys., 24, 551-3. Speiser, B. and Sprattling, L (1993) High dose rate brachytherapy for the local control of endobronchial carcinoma. Int. J. Radial Oncol. Biol. Phys., 25, 579-88. Zajac, A.J., Kohn, M.L, Heiser, D. et al. (1993) High-dose rate intraluminal brachytherapy in the treatment of endobronchial malignancy. Radiology, 187, 571-5. Tredaniel, J., Hennequin, C., Zalcman, G. et al. (1994) Prolonged survival after high-dose rate endobronchial radiation for malignant airway obstruction. Chest, 105, 767-72. Chang, LF.L, Horvath, J., Peyton, W. et al. (1994) High dose rate afterloading brachytherapy in malignant airway obstruction of lung cancer. Int. J. Radial Oncol. Biol. Phys., 28, 589-96. Collins, S., Burt, P., Barber, P. et al. (1994) High dose rate intraluminal radiotherapy for carcinoma of the bronchus: outcome of treatment of 406 patients. Radiotherapy, 33, 31-40. Cotter, G.W., Larisey, C., Ellingwood, K.E. et al. (1993) Inoperable endobronchial obstructing lung cancer treated with combined endobronchial and external beam irradiation: a dosimetric analysis. Int. J. Radial Oncol. Biol. Phys., 27, 531-5. Goldman, J., Bulman, A., Rathmell, A. et al. (1993) Physiological effect of endobronchial radiotherapy in patients with major airway obstruction. Thorax, 48, 110-14. Marsh, B.R., Colvin, D.P., Zinreich, E.S. et al. (1993) Clinical experience with an endobronchial implant. Radiology, 189,147-50. Nori, D., Allison, R., Kaplan, B. et al. (1993) High dose rate intraluminal irradiation in bronchogenic carcinoma. Chest, 104,1006-11. Pisch, J., Villamena, P., Harvey, J. et al. (1993) High dose-rate endobronchial irradiation in malignant airway obstruction. Chest, 104, 3721-5. Madia, H.N., Wahlers, B., Reichle, C. et al. (1995) Endobronchial radiation therapy for obstructing malignancies: ten years' experience with iridium-192 high-dose radiation brachytherapy afterloading technique in 365 patients. Lung, 173, 271-80.
A25
A26
A27
A28
A29
A30
A31
A32
A33
A34
A35
A36
Huber, R., Fischer, R., Hautmann, H. et al. (1995) Palliative endobronchial brachytherapy for central lung tumors. Chest, 107(2), 463-70. Gustafson, G., Vincini, F., Freedman, L et al. (1995) High dose rate endobronchial brachytherapy in the management of primary and recurrent bronchogenic malignancies. Cancer, 75(9), 2345-50. Sur, R., Mahomed, G., Pacella, J. et al. (1995) Initial report on the effectiveness of high dose rate brachytherapy in the treatment of hemoptysis in lung cancer. Endocuriether. Hypertherm. Oncol., 11,101-6. Speiser, B. (1995) The role of endobronchial brachytherapy in patients with lung cancer. Clin. Pulmonary Med. 2(6), 344-52. Saito, M., Yokoyama, A., Kurita, Y. et al. (1996) Treatment of roentgenographically occult endobronchial carcinoma with external beam radiotherapy and intraluminal lowdose rate brachytherapy. Int. J. Radial Oncol. Biol. Phys., 34,1029-35. Delclos, M.E., Komaki, R., Morice, R.C. et al. (1996) Endobronchial brachytherapy with high-dose rate remote afterloading for recurrent endobronchial lesions. Radiology, 201(1), 279-82. Huber, R.M., Fischer, R., Hautmann, H. et al. (1997) Does additional brachytherapy improve the effect of external irradiation? A prospective, randomized study in central lung tumors. IntJ. Radial Oncol. Biol. Phys., 38(3), 533^0. Corsa, P., Parisi, S.S., Raguso, A. et al. (1997) High-dose brachytherapy in endobronchial neoplastic stenoses. Radiol. Med. (Torino), 94(1-2), 94-9. Perol, M., Caliandro, R., Pommier, P. et al. (1997) Curative irradiation of limited endobronchial carcinomas with high-dose rate brachytherapy. Results of a pilot study. Chest, 111(5), 1417-23. Ornadel, D., Duchesne, G., Wall, P. et al. (1997) Defining the roles of high dose rate endobronchial brachytherapy and laser resection for recurrent bronchial malignancy. Lung Cancer, 16(2-3), 203-13. Ofiara, L, Roman, T., Schwartzman, K. et al. (1997) Local determinants of response to endobronchial highdose rate brachytherapy in bronchogenic carcinoma. Chest, 112(4), 946-53. Hennequin, C., Tredaniel, J., Chevret, S. et al. (1998) Predictive factors for late toxicity after endobronchial brachytherapy: a multivariate analysis. Int. J. Radial Oncol. Biol. Phys., 42(1), 21-7.
17 Brachytherapy in cancer of the esophagus A.D. FLORES
17.1
INTRODUCTION
It is estimated that, in the year 2001,13200 new cases will be diagnosed with carcinoma of the esophagus in North America and, of these, 12 500 will die as a result of the disease during the same year. This number corresponds to only 1 % of all cancer cases seen in this particular year [ 1 ]. Whereas the incidence of cancer of the stomach in the last 40 years has been declining, the incidence of adenocarcinomas arising in the esophagus has substantially increased, and now comprises 20-40% of all esophageal malignancies seen in North America and Western Europe [2-5]. Similarly, there has been a significant increment in the incidence of cardioesophageal lesions in relation to carcinomas developing in the distal portion of the stomach [2-5]. Higher incidence rates (40-50 per 100000 population) have been reported in certain regions of Iran, South Africa, China, and the former Soviet Union. Although etiological factors are still unknown, many environmental factors (alcohol, smoking, dietary, etc.) have been associated with the disease [6]. The clinical presentation, behavior, and prognosis of cancers arising in the esophagus and/or the cardioesophageal junction are similar. Unfortunately, conventional treatments have not altered the poor prognosis these patients have and survival rates of 5-7% have remained unchanged in the last four decades [7-9]. Most patients have advanced disease when first diagnosed and only palliative treatment is available to them.
Only 20% of all new patients seen could be eligible for treatment with an intention of cure. However, the treatment for these patients presenting with early disease is controversial. Although, historically, surgical treatments have produced a poor overall survival similar to that of radiation treatments [7-9] trials designed to compare these two treatments have been difficult, not enough patients could be recruited and they therefore had to be abandoned [10]. Results of investigations using preoperative or postoperative conventional radiotherapy have been mixed [11-15]. Cooperative efforts employing multimodality conventional treatments have also been disappointing. Chemotherapy regimens have not affected metastasis or stopped the development of metastasis in recent studies. Similarly, encouraging preliminary results using chemotherapy as an adjuvant treatment to surgery or radiation have been associated with increased toxicity and have resulted in neither significant improved survival nor better quality of life [16-28]. Clearly, new innovative treatment protocols will be required to enhance the cure rate for cancers arising in the esophagus and cardioesophageal junction.
17.2
NATURAL HISTORY OF THE DISEASE
The majority of patients with carcinoma of the esophagus are diagnosed with dysphagia, which is a manifestation of
244 Brachytherapy in cancer of the esophagus
extensive local disease causing malignant obstruction. In the western world, 80% of patients are diagnosed when the tumor is larger than 5 cm or extended beyond the esophageal wall (T2-T3). Even in early operable cases, lymphatic spread is recognized in 70% of the resected esophageal specimens 9 Anatomically, the esophagus does not have a serosal layer and the neoplasm could easily reach the adjacent peri-esophageal tissues and spread through the rich submucosal lymphatics to the most proximal esophageal wall and to the mediastinal, perigastric, and cervical lymph nodes. The study of patterns of failure in several institutions confirms that the overwhelming majority of patients (more than 80%) with esophageal cancer fail and die with persistent local or loco-regional disease and metastasis [26,27,32-36]. From these facts, it is reasonable to assume that if local control of the disease could be enhanced by innovative treatments, better quality of life and possibly even longer survival will result. It is, therefore, imperative that new protocols be designed with the specific aim of enhancing the local therapeutic ratio of external irradiation. 173 HISTORY AND TREATMENT RATIONALE FOR INTRALUMINAL BRACHYTHERAPY The potential therapeutic value of intraluminal brachytherapy for esophageal cancer was already recognized at the beginning of the twentieth century. As early as 1909, some investigators had reported useful palliation of dysphagia by directly placing radium beads in the esophagus housed in a nasogastric plastic tube [37,38]. It was, however, not until 1969 that successful intraluminal brachytherapy was reported in a selected group of eight patients by Rider and coworkers in Toronto [39]. Although these investigators stressed the usefulness of brachytherapy for esophageal malignancies, this type of procedure was unfortunately not actively pursued. The lack of interest was due to many factors, among them: the introduction of megavoltage units and deviation of interest towards external irradiation, technical difficulties of the brachytherapy procedure, and, most importantly, the poor tolerance of patients due to longer treatment times. The Second World War and legitimate concerns regarding radiation exposure also delayed the development of new radioactive isotopes for medical use until 1948, when cobalt-60 was introduced. In 1953, Henschke developed the idea of afterloading and designed plastic tube devices in an attempt to reduce the radiation exposure of the medical personnel [39a]. In more recent years, the production of safer and higher specific activity radioactive sources (Table 17.1) facilitated the fabrication of miniaturized sources, significantly decreased treatment times, introduced automatic remote control radiation delivery, and completely eliminated the problem of radiation exposure. This improved
Table 17.1 Linear activities of radioactive sources
Cesium-137 tubes Cesium-137 pellets lridium-192(10Ci/4 mm)
21 mCi cm-1 or 770 MBq cnr1 1 126 mCi cm or 4660 MBq crrr1 20 Ci cm-1 or 74000 MBq cm-1
technology made brachytherapy in general a more acceptable treatment modality for esophageal cancer and malignancies at other sites. The basic rationale for choosing brachytherapy in esophageal cancer is that while the amount of irradiation that can be given by external irradiation is limited by the tolerance of the normal tissues in the chest (lung, spinal cord, mediastinal structures, etc.), intraluminal brachytherapy places the highest concentration of irradiation directly into the intraluminal disease but delivers significantly lower doses to the adjacent normal tissues (due to the rapid dose fall-off). In view of this physical advantage, therefore, it is possible to increase the dose to the cancer with the use of brachytherapy without affecting the adjacent normal tissues, thus enhancing the therapeutic ratio.
17.4 CLINICAL STAGING AND PRETREATMENT INVESTIGATIONS The TNM classification for esophageal malignancies has changed. Tumor size and degree of circumferential involvement are no longer criteria to define the extent of the primary tumor. The new staging classification is based on the histopathological assessment of the depth of tumor penetration in the esophageal wall (Table 17.2). Because it is not always possible to obtain a representative sample of the full thickness of the esophagus by
Table 17.2 TNM classification: esophagus Primary tumor (T)
TX TO Tis T1 T2 T3 T4
Primary tumor cannot be assessed No evidence of primary tumor Carcinoma in situ Tumor invades the lamina propria or submucosa Tumor invades the muscularis propria Tumor invades adventitia Tumor invades adjacent structures
Lymph node (N) NX Regional lymph nodes cannot be assessed NO No regional lymph node metastasis N1 Regional lymph node metastasis Distant metastasis (M) MX Presence of distant metastasis cannot be assessed MO No distant metastasis M1 Distant metastasis
Therapy decision process 245
endoscopic biopsy, pretreatment clinical staging of this cancer has been difficult. Recently, endoscopic ultrasonography, a new diagnostic procedure, has produced satisfactory imaging of the different layers of the esophageal wall and regional lymphatics. Several investigators have demonstrated good clinico-pathological correlation, with ultrasonography suggesting superiority over computerized tomography (CT) or magnetic resonance imaging (MRI) scans in the clinical staging of esophageal cancer [40]. The routine use of endoscopic ultrasonography to assess the extent of esophageal malignancies has been endorsed by the International Society of Gastroenterology [41]. If the diagnosis of cancer of the esophagus is suspected, the initial evaluation should include routine blood tests and full physical examination, including an evaluation of the lymph node regional bearing areas. Then, adequate radiological studies of the upper gastrointestinal tract and endoscopic examination for direct biopsy of a suspicious area should follow. If the lesion is located above the tracheal bifurcation, a bronchoscopic examination at the same setting is advisable. This will permit a better evaluation of the tumor extension and exclude invasion to the membranous portion of the adjacent bronchus or trachea. Once the diagnosis of a malignancy has been confirmed, CT scan of the chest and upper abdomen is performed to assess the extent of the primary disease and exclude metastasis. As discussed earlier, endoscopic ultrasound is desirable to determine with certainty the extent of the local and regional disease and for clinical staging prior to definitive treatment. Only 20% of the patients diagnosed with cancer of the esophagus have their disease confined to the esophageal wall (T1NO, T2NO) and could be suitable for treatment with curative intent. In most patients, the esophageal disease has already extended beyond the wall and cannot be resected adequately (T3, T4). The presence of regional lymph node metastasis (Nl, N2) is also a sign of poor prognosis, even if the esophagus is resected, as only 15% or fewer patients with positive nodes will survive [7,8,30,31]. Because, overall, 5-year survival rates with conventional treatments are extremely poor, a careful and adequate assessment of the clinical staging prior to treatment is essential. This information is extremely useful when selecting the most suitable treatment according to prognosis and to avoid unnecessary complications associated with radical treatments [42,43]. The addition of intraluminal brachytherapy to external irradiation is a good treatment strategy, as it is simple, well tolerated, and can enhance the local control and quality of life of most of these patients. 17.5
TREATMENT STRATEGIES
Attempts to improve local control by increasing the amount of external irradiation have failed due to poor
tolerance of the normal tissues adjacent to the esophagus. Newer schedules of fractionated radiotherapy (accelerated, superfractionation, etc.), concurrent chemotherapy, chemical modifiers, or sensitizers aiming to enhance the effect of external irradiation are also likely to fail for the same reasons. Because the local failure is high, it is sensible to assume that better control can be obtained by combining radiotherapy and surgery. Results of studies using postoperative irradiation have been negative and those of preoperative irradiation trials have been mixed. Two European phase III studies [11,12], comparing esophagectomy alone and preoperative conventional external irradiation showed no advantage for either group. In these trials, the radiotherapy schedule was unusual (4000 cGy in ten fractions) and a high rate of complications resulted. Three other studies [13-15], comparing esophagectomy alone with a more conventional preoperative external irradiation, reported significant benefits in the preoperative irradiation arm. Due to concerns regarding perioperative complications, current trials have not considered a preoperative irradiation treatment arm. Instead, newer programs involve studying the value of neoadjuvant or concurrent chemotherapy added to either surgery or external irradiation. However, the impact of these new programs should be expected to be minimal and a larger number of patients will be required to show differences between the treatment arms. The esophagus above the tracheal bifurcation lies contiguous and posterior to the membranous portion of the trachea and the proximal aspect of the left main bronchus. These anatomical characteristics show the obvious weakness of esophagectomy alone as a rational treatment for cancers of the esophagus. Furthermore, the morbidity and mortality associated with esophagectomy are higher for upper esophageal lesions. The lower end of the esophagus is relatively free and more accessible for adequate removal. Conventional external irradiation produces similar results to surgery in early cases and may be better for palliation in more advanced tumors. The morbidity associated with external irradiation compared to surgery is mild and similar for all levels of the esophagus, and no additional toxicity should be expected with intraluminal brachytherapy [44].
17.6
THERAPY DECISION PROCESS
Patients with cancer of the esophagus could be candidates for either palliative or curative treatment, depending on the extent of their disease, performance status, associated illnesses, and age. Patients with distant metastasis have an average lifespan of only 3-6 months, and the palliative treatment should be short, simple, and specifically designed to improve their quality of life in terms of swallowing and pain. The same philosophy
246 Brachytherapy in cancer of the esophagus
should be applied to elderly patients with locally advanced disease and poor condition. Because dysphagia and nutrition are the main concerns at presentation, most patients require dilatations and/or parenteral feeding prior to their treatment for palliation or curative intent. In palliative conditions, a single intraluminal treatment alone or combined with a short course of external irradiation (1800 cGy in three fractions given by parallel and opposed fields) may be sufficient to restore and maintain swallowing and improve the quality of life of these patients. Patients with only loco-regional disease (as per CT scan and endoscopic sonogram findings) and in good overall condition are suitable for a form of radical treatment designed to control the disease. They could be eligible for trials of investigation using radiotherapy or esophagectomy and adjuvant treatments. Patients with esophageal malignancies located above the tracheal bifurcation may be best treated with radiotherapy programs alone or in combination with chemotherapy in prospective trials. Lower esophageal lesions can be treated also by radiotherapy regimens, reserving esophagectomy for persistent or recurrent disease. Primary esophagectomy should only be considered in prospective clinical trials and opposed to radiotherapy programs. Radiotherapy with curative intent requires adequate planning and several weeks of external irradiation using multiple portals. Intraluminal brachytherapy can be given before, during, or after the course of external irradiation. Treatment planning should be tailored to each patient needs and to whether concurrent chemotherapy is used. It would be preferable to give part of the external treatment by parallel AP/PA opposed portals to prevent lung toxicity when chemotherapy is given concurrently. Tumor doses given by external irradiation are usually 4000 cGy in 15 fractions, or 5000 cGy in 20 fractions in 4 weeks to an 8 x 8 x 16 cm treatment volume. The intraluminal brachytherapy dose is estimated at 1 cm of the axis of an 8-12 cm linear source and consists of 3000 cGy in 48 h with low dose-rate sources (radium, cesium), or a single 1500 cGy with intermediate and high dose-rate sources (in 1.5h or a few minutes respectively). The radiobiological reasons for these doses are discussed below. General guidelines for esophageal brachytherapy have been published recently [45].
tube. Catheters of 8 mm or less external diameter can be easily placed through the nose and are, in general, better tolerated. Larger diameter catheters require the oral route and adequate sedation (xylocaine spray and valium 3-5 mg intravenously is usually sufficient). With either route, the suctioning of oral secretions and/or salivation is essential to maintain patient comfort and avoid aspiration. The following is a step-by-step plan and outline of the procedure. 1. The precise anatomical location and extension of the cancer are determined by reviewing the barium swallow X-rays, endoscopic and CT scan findings. 2. Localization and planning of the treatment area are performed in the simulator unit using a limited barium swallow study, at least 24 h prior to the procedure. The treatment centers and fields for external irradiation and intraluminal brachytherapy are chosen, and points of reference (bony landmarks, skin tattoos, etc.) identified (Figure 17.1).
17.7 INTRALUMINAL BRACHYTHERAPY TECHNIQUE The placement of the endoesophageal catheter for intraluminal brachytherapy is a very simple outpatient procedure and it is usually performed under local anesthesia and mild sedation. The catheter can be placed via the oral or nasal route, depending on the diameter of the
Figure 17.1 Simulator film showing treatment center and fields for external and intraluminal brachytherapy. Note bony landmarks and skin tattoo.
Intraluminal brachytherapy technique 247
3. In a fluoroscopy room, the patient is sedated and his or her nose and throat are adequately anesthetized in the sitting or supine position. It is important to have a suction unit available to assist the patient during the procedure. 4. A thin, cut-end (French 8 or 10) nasogastric tube containing a fine, soft, atraumatic Teflon-coated guide wire is passed or swallowed into the stomach, under fluoroscopy. If the obstruction is significant, the nasogastric tube is passed only to the level of stricture, and then, under fluoroscopy, the guide wire alone is maneuvered through the stenosis to the stomach and its position anchored in the body of this organ (Figures 17.2, 17.3 and 17.4). 5. The esophageal stricture is dilated, either by a balloon or a Savory dilator, through the guide wire if required before the placement of the applicator tube for brachytherapy (Figure 17.5). 6. The exact positioning of the intraluminal treatment is verified using dummy sources according to plan, and the applicator position is secured to the mouth guard or taped to the nose. X-ray films are then obtained for planning and dosimetry (Figures 17.6 and 17.7). 7. The patient is treated, preferably using a remote afterloading medium dose-rate (MDR; i.e., 1000 cGy h-1) or a high dose-rate (HDR; i.e., 50000-20000 cGy h-1) unit (Figures 17.8 and 17.9).
Figure 17.2 An 8-mm nasogastric tube containing a guide wire at the level of strictured esophagus.
Figure 17.3 The guide wire is negotiated through the stenotic esophagus.
Figure 17.4 The position of the guide wire in the stomach is verified.
248 Brachytherapy in cancer of the esophagus
Figure 17.5 Balloon dilatation.
Figure 17.6 Passage of the intraluminal applicator through the esophagus, using the guide wire. Note the dummy cesium (MDR) pellets in the applicator, used as reference for the intraluminal treatment position.
17.8 RADIOBIOLOGICAL AND CLINICAL CONSIDERATIONS
Prior to 1980, intraluminal brachytherapy for esophageal malignancies was only sporadically used, and mainly to treat patients with recurrent disease. In the 1980s, the Vancouver Clinic utilized radium and later cesium tubes (four tubes, each with 2 cm active length and 10 mg radium equivalent). These elements were placed in tandem within an esophageal plastic tube with the aim of delivering a dose of 3000 cGy at 1 cm from the axis in 48 h, and a low dose rate (LDR) of 75 cGy Ir1 (Figure 17.10). This treatment was used only in selected patients who had not responded to conventional external radiotherapy. Although this treatment required hospitalization, it was well tolerated and an adequate temporary palliation of the dysphagia was achieved. An esophageal applicator [46] became available in February 1985 that
could be connected to a remote afterloading unit (Selectron). This unit operated radioactive cesium pellets, each of 2.5 mm in diameter and 40 mCi of radium equivalent; 40 pellets placed in tandem generated a linear source of 10 cm in length with a dose rate of 1000 cGy h'1 at 1 cm from the axis (MDR). Similar applicators are now available for HDR iridium-192, which uses a stepping but smaller radioactive source of only 1.1 mm in diameter, which produces a higher dose rate (50000-20000 cGy h-1) (see Figures 17.7 and 17.9). Because higher biological effects should be expected with higher dose rates, a linear quadratic model [47] was used to estimate equivalent doses to standard LDR brachytherapy. Assuming a recovery time to sublethal effects of irradiation of 2 h and an a/p ratio of 4 for late effects and 10 for acute effects, this model showed that 3000 cGy given with LDR radioactive materials were equivalent to 1500 cGy given by HDR sources (Figure 17.11). It can also be seen that, according to this formula,
Radiobiological and clinical considerations 249
Figure 17.7
Verification of position of the intraluminal
applicator using dummy sources for an iridium (HDR) stepping source, in a similar patient to Figure 17.6.
tissue effects do not change for dose rates higher than lOOOcGylr 1 . The isodose distribution of intraluminal brachytherapy and rapid dose fall-off in depth can be appreciated from Figure 17.12. The outer diameter of the applicator is also important when considering intraluminal brachytherapy. It determines the dose to the surface of the tumor or normal mucosa in contact with the applicator. The dose is usually prescribed at 1 cm from the axis of the source. The actual dose to the mucosa lying adjacent to the surface of an applicator with smaller outer diameter (i.e., 6 mm) will be significantly higher than when a larger outer diameter applicator is used (i.e., 10 mm). Higher morbidity, mucositis, ulceration, and fibrosis should be expected with smaller outer diameter applicators if the patient survives. Therefore, the use of larger outer diameter applicators and confining the brachytherapy boost only to the area affected by the disease could minimize complications. Intraluminal brachytherapy for esophagus can be used alone for palliation of dysphagia in advanced or metastatic cases. It can also be used as a complement to external irradiation in earlier cases with the intention of cure. In these situations, brachytherapy can be given before, during, or after external irradiation. If the obstruction is significant, it may be advantageous to start with brachytherapy as dysphagia could be improved and tolerance to external treatment thus enhanced. On the other hand, intraluminal brachytherapy may be more effective after external irradiation, as the bulk of the disease will be reduced and the depth dose given to the base of the tumor by brachytherapy is significantly better. In a study conducted in Vancouver there was, however, no
Figure 17.8 A patient undergoing treatment with an oral-esophageal applicator. Note the head support and mouth route. MDR cesium pellets.
250 Brachytherapy in cancer of the esophagus Figure 17.9 A patient undergoing treatment with a naso-esophageal applicator. Note the patient comfort and suction unit. HDR iridium source.
Figure 17.11 Dose equivalencies for different dose rates according to the linear quadratic equation, (i - recovery time to sublethal effects of irradiation.)
Figure 17.10 X-ray film showing an endoesophageal tube containing four radium tubes in tandem. LDR treatment and 8 cm active length.
repeated discomfort and the probability of more complications due to trauma and inconvenience for a patient whose quality of life has already been significantly affected by the disease.
17.9 TREATMENT RESULTS significant difference in outcome when brachytherapy was used either before or after external irradiation. A single brachytherapy boost of 1500 cGy (MDR or HDR) had been preferred at the Canadian Vancouver Cancer Clinic, and proved to be an effective, well-tolerated treatment in more than 700 patients. Brachytherapy has been given in three or more fractions in other centers, but multiple applications need to be balanced against
In China, where the incidence of esophageal carcinoma is high, a mass screening program was conducted in the rural areas of Linxian in 1970-1974. This program detected a large number of patients with esophageal malignancy, not all of whom could be treated conventionally with external irradiation due to the scarcity of machines [48]. Thus, 203 patients were treated only with
Treatment results 251 Figure 17.12 Isodose distribution for a linear radioactive source.
intraluminal brachytherapy using cobalt-60 wires with two to four applications and variable doses. Seventeen out of 203 patients survived 5 years or more and were apparently cured [49]. A randomized study of 200 patients, performed at the Shanxi cancer hospital between 1982 and 1986 [50], comparing external radiotherapy alone versus external plus intraluminal brachytherapy, showed significant difference in favor of the treatment arm with brachytherapy (17% versus 10% 5-year survivals, respectively). Hishikawa, from Japan, also using a cobalt linear source and external radiotherapy, reported an 18% 5-year survival rate in a group of 66 patients with disease limited to the esophagus [51]. In Vancouver, a phase I-II study was conducted between 1985 and 1987 to evaluate the toxicity and efficacy of a combined treatment consisting of 4000 cGy in 15 fractions given by external irradiation plus 1500 cGy at 1 cm given by intraluminal brachytherapy. All patients (171) seen during that period, except for those with impending tracheal or bronchial fistula, were treated in that manner. This was a feasibility study and to assess the toxicity and response of additional intraluminal brachytherapy to external irradiation. It was concluded that this combined treatment was feasible, well tolerated, and could be done as an outpatient procedure. Morbidity was only related to temporary mucositis, and there was no associated treatment-related mortality. A quality-of-life assessment after treatment in relation to patients' performance status, ability to swallow, weight, and pain demonstrated improvement in all of these parameters [29]. A subsequent phase III study, with a similar group of patients, but designed to compare the value of brachytherapy before and after external irradiation in 219 patients, showed no significant difference
between these two treatment arms. From 1985 to 1993, over 700 patients with cancer of the esophagus and cardia were treated with intraluminal brachytherapy at the Cancer Agency in Vancouver. The first 150 patients were treated with a linear source made of radioactive cesium pellets that generated a dose rate of 1000 cGy tr1 (MDR) at 1 cm. All subsequent patients were treated using a stepping radioactive source of iridium (HDR), producing a dose rate between 50 000 and 20 000 cGy Ir1 (depending on the activity/age of the source) at 1 cm. The treatment dose chosen for both radioactive sources was the same or 1500 cGy at 1 cm from the axis, as estimated by the linear quadratic model [46]; however, the treatment time was significantly shorter with the HDR unit. The clinical evaluation of the acute and late effects of patients treated by LDR (radium or cesium), MDR (cesium pellets), or HDR (iridium) showed similarity of effects, suggesting good correlation with the predicted values by the linear quadratic equation for the different dose rates. Two hundred and ninety-seven patients with cancers of the esophagus and cardia treated in Vancouver with external and intraluminal irradiation were eligible for an analysis and a minimum follow-up to 5 years. Ninetythree patients (31%) had an upper esophageal cancer above the carina, and 204 had esophageal tumors below the tracheal bifurcation. Forty upper esophageal cancers had only palliative treatment (five of them had distant metastasis and 35 had advanced disease and were in poor condition). Only two of the 93 patients with upper esophageal cancers survived 5 years. Of the 204 (69%) patients who had lower esophageal cancers, 114 (56%) had an inoperable disease (29 had distant metastasis and 85 locally advanced disease). Eight out of 85 (9.4%)
252 Brachytherapy in cancer of the esophagus
patients with local advanced disease survived 5 years, whereas none of the patients with distant metastasis survived. Ninety patients were explored, but only 66 (73%) were resected. Of the 24 patients who were considered inoperable at the time of exploration, none survived. Only three patients had palliative resection (all had liver metastasis), none survived more than 4 months. Of the 63 patients who had curative esophagectomy after external and intraluminal irradiation, 31 (49%) had survived the disease for 5 years or more after treatment. A summary of the clinical presentation, treatment, and results for all patients is shown in Table 17.3. The study of the 63 pathological specimens after external and intraluminal radiotherapy permitted an evaluation of the changes induced by brachytherapy in the esophageal tissues. The macroscopic effect in the specimen was obvious, and the localized changes seen in the mucosa reflected the effect of the intraluminal treatment. The historical examination revealed marked pyknotic changes, with no evidence of viable tumor at different levels of the wall of the esophagus (Figure 17.13). The radiation effects and the depth of tumor penetration were determined separately for each specimen and subsequently correlated with local control and survival. The depth was expressed according to the four layers of the esophagus: level 1 when penetration was no deeper than the muscularis mucosa; level 2 when there was more penetration but no deeper than the submucosa; level 3 when it was deeper than the muscularis propria; and level 4 when it involved the peri-esophageal tissues [52]. A ratio of 1 (radiation effect/tumor penetration) was present in 43 cases and this correlated with local tumor control in 95% of them. The survival was superior when there was complete sterilization of malignancy in the surgical specimens, as can be seen in Table 17.4. The histopathological findings of radiation effects in regional lymph nodes were surprising. Periesophageal nodes considered to have been involved by the cancer were pathologically sterilized in 11 cases by the treatment combination and this was also reflected in survival, as can be seen in Figure 17.14. The combined treatment of external and intraluminal radiotherapy was well tolerated, morbidity was low, and
Figure 17.13
Pathological specimen of esophagus showing
esophageal layers and radiation effect of intraluminal brachytherapy.
there was no associated mortality in over 700 patients. A summary of treatment complications is shown in Table 17.5. Acute radiation mucositis, although common, was usually transient. Persistent mucositis was due to extensive disease or associated factors (candidiasis, etc.) and dysphagia was related to persistent disease. Bleeding as a complication following brachytherapy is extremely rare in our experience; however, it should be expected if the tumor is advanced or significant trauma occurred during the insertion of the applicator. Fibrosis and stenosis
Table 17.3 Cancer of the esophagus: treatment results in 297patients
Above carina Advanced Below carina Advanced Below carina Resected3 Number of patients with distant metastasis Number of patients explored and found inoperable Palliative resection (liver metastasis)3 All cases (overall) 3
88 85 63 34 24 3 297
Total esophagectomy after external + intraluminal brachyth< rapy. External + intraluminal brachytherapy, Vancouver (1985-198!J).
2 8 32 0 0 0 42
2 9.4 49
14
Conclusions 253
are common in long-term survivors, and occurred in 42% of our patients with advanced cancers surviving more than 5 years. Dilatations at regular intervals were required in most of these cases. From the results obtained, it can be seen that intraluminal brachytherapy enhances the therapeutic ratio when added to external irradiation for this disease. A comparison of results with other recent combined treatment modalities showed superiority of external and intraluminal radiotherapy over esophagectomy plus chemotherapy or external irradiation combined with adjuvant chemotherapy (Table 17.6).
17.10
CONCLUSIONS
Because the prognosis and survival of patients with esophageal cancer are poor, an adequate evaluation of these patients prior to any treatment is essential. The emphasis of treatment in advanced cases should be palliation and improving the quality of life of these patients. A short course of external irradiation alone or combined with a single intraluminal brachytherapy may be sufficient for most patients. A recent prospective, randomized trial has shown no
Table 17.4 Cancer of the esophagus: radiation pathology
Sterilized (no recognizable disease) Ratio 1 (radiation damage to all levels) Ratio < 1 (radiation damage to superficial layer only)
16 24 23
3 7 16
1 3 2
12 14 5
75 58 21
Totals
63
26
6
31
49
NED = no evidence of disease. Table 17.5 Cancer of the esophagus: complications of external irradiation + brachytherapy
Radiation esophagitis Radiation pneumonitis Broncho-esophageal fistula
Mild
Moderate!
Severe
358
72 (14%)
60 (12%) 1a 30 (6%)b
a
Unrelated to intraluminal brachytherapy. Related to disease; occurred 6 months or more after treatment. Mild: transient, requires symptomatic medication only. Moderate: transient but may require supplementary care. Severe: requires hospitalization or intervention. b
Table 17.6 Cancer of the esophagus: summary of treatment results
Non-surgical (a 11 cases) XRT + chemotherapy XRT +brachytherapy Extensive disease Limited disease XRT + brachytherapy Extensive disease Surgical combined (early cases or limited) Preoperative chemotherapy Preoperative chemotherapy + XRT Preoperative XRT + brachytherapy ? Data are not available. [*] Vancouver series. XRT = irradiation.
[18]
61
2
30
?
82 66
0 0
7 37
0 18
138
0
33
48 43 63
11 8 0
25 50 65
[51]
[*]
[23] [20]
7.2
18 34 49
254 Brachytherapy in cancer of the esophagus
retrospective study of esophageal cancer presenting to an endoscopy unit. Gut, 32, A555. 6. Parkin, D., Pisani, P. and Ferlay, J. (1999) Global cancer statistics. CA Cancer J. Clin., 49(1), 33-64. 7. Earlam, R. and Cunnha-Melo, J.R. (1980) Esophageal squamous carcinoma: a critical review of radiotherapy. Br. J.Surg., 67,457-61. 8. Earlam, R. and Cunnha-Melo, J.R. (1980) Esophageal squamous carcinoma: a critical review of surgery. Br.}. SHrg.,67,381-90. 9. Matthews, H.R. and Waterhouse, J.A. (1987) Cancer of the Esophagus. Clinical Monograph, Volume 1. Basingstoke, Macmillan Press. 10. Earlam, R. (1991) An MRC prospective randomized trial of radiotherapy versus surgery for operable squamous carcinoma of the esophagus. Ann. R. Coll. Surg. Engl., 73, 8-12. 11. Launois, B., Delarue, D., Campion, J. and Kerbaol, M. (1981) Preoperative radiotherapy for carcinoma of the Figure 17.14
Survival versus peri-esophageal lymph node
status.
benefit from chemotherapy prior to esophagectomy in early operable cases [28]. The combination of external and intraluminal radiotherapy has obvious advantages over other treatments in esophageal cancer: it is simple, cost effective (hospitalization is not required), and has lower morbidity than adjuvant chemotherapy and/or esophagectomy. As it does not increase the operative morbidity associated with esophagectomy, it should be the treatment of choice in early cases, radical surgical treatment being reserved only for persistent or recurrent disease. Future prospective trials must consider combined external and intraluminal brachytherapy as the main treatment arm in early operable esophageal cancer cases to determine the best local therapy. These future trials must also evaluate morbidity and the quality of life of survivors.
REFERENCES
esophagus. Surg. Gynecol. Obstet, 2,690-2. 12. Gignoux, M., Russell,A., Paillot, B.etal. (1987) The value of preoperative radiotherapy in esophageal cancer. Results of the EORTC World J. Surg., 11,426-32. 13. Huang, G., Gu, X., Wang, L etal. (1988) Combined preoperative irradiation and surgery for esophageal carcinoma. In International Trends in General Thoracic Surgery: Esophageal Carcinoma, ed. E.W. Wilkins and J. Wong. Philadelphia, Saunders, 315-18. 14. Wang, M., Gu, X.Z., Yin, W.B., Huang, G.J., Wang, LJ. and Zhang, D.W. (1988) Randomized clinical trial on the combination of preoperative irradiation and surgery in the treatment of esophageal carcinoma; report on 206 patients. Int.J. Radial. Oncol. Biol. Phys., 16,325-7. 15. Nygaard, K., Hagen,S., Hansen, H.S.etal. (1992) Preoperative radiotherapy prolongs survival in operable esophageal carcinoma. A randomized, multicenter study of preoperative radiotherapy and chemotherapy. The second Scandinavian trial in esophageal cancer. World J. Surg., 16,1104-9. 16. Schlag, P., Herrman, R., Raeth, V.etal. (1988) Preoperative chemotherapy in esophageal cancer. A Phase II study.4cta Oncol., 27,811-14. 17. Roth.JA, Pass, H.I., Flanagan, MM. etal. (1988) Randomized clinical trial of preoperative and postoperative adjuvant chemotherapy with cisplatin, vindesine, and bleomycin for carcinoma of the
1. Greenlee, R.T., Hill-Harmon, M.B., Murray, T. and Thun, M. (2001) Cancer statistics 2000. CA CancerJ. Clin., 51,15-36. 2. Blot,W.J., Devesa,S.S., Kneller, R.etal. (1991) Rising incidence of carcinoma of the esophagus and cardia. JAMA, 265,1287-9. 3. Powell, J. and McConkey, C.C. (1990) Increasing incidence of adenocarcinoma of the gastric, cardia and adjacent sites. Br. J. Cancer, 62,440-3. 4. Reed, P.I. (1991) Changing patterns in esophageal cancer. Lancet, 338,178. 5. Johnson, B.J., Hill, M.J. and Reed, P.I. (1991) Fifteen years
esophagus./ Thorac. Cardiovasc. Surg., 96,242-8. 18. Herskovic,A., Martz, K., Al-Sarraf, M.etal. (1993) Combined chemotherapy and radiotherapy compared with radiotherapy alone in patients with cancer of the esophagus. N. Engl.J. Med., 326,1593-7. 19. John, M.J., Flam, M.S., Monry, P.A. etal. (1989) Radiotherapy alone and chemoradiation for non metastatic esophageal carcinoma. A critical review of chemoradiation. Cancer, 64,2397-03. 20. Forastiere, A.A., Orringer, M.B., Perez-Tamayo, C., Urba, S.G. and Zahurak, M. (1993) Preoperative chemoradiation
References 255
followed by transhiatal esophagectomy for carcinoma of the esophagus: final report./ Clin. Oncol., 11(6), 1118-23. 21.
Urba, S.G., Orringer, M.B., Perez-Tamayo, C., Bromberg, J. and Forastiere, A. (1992) Concurrent preoperative chemotherapy and radiation therapy in localized
esophageal adenocarcinoma. Cancer, 69(2), 285-91. 22. Roth, J.A., Ajani, J.A. and Rich T.A. (1990) Multidisciplinary therapy for esophageal cancer. Adv. Surg., 23,239-60. 23.
36. Ruol, F., Segalin, A., Castoro, C. etal. (1987) Patterns of neoplastic recurrence after radical and palliative resection of the esophagus. In Diseases of the Esophagus, ed.J.R. Siewert. Berlin, Springer-Verlag714-16. 37. Guisez, J. (1909) Essais de traitement de quelques cas d'epithelioma de I'esophage par les applications locales directes de radium. Bull. Soc. Med. Hop. Paris, 27, 717-22. 38. Harriman, F.R. (1927) Malignancy of the larynx and the esophagus treated by radium emanation. Laryngoscope,
Kelsen, D.P., Bains, M. and Burt, M. (1990) Neoadjuvant
treatment for cancer of the esophagus. Semin. Surg. Oncol., 6,268-73. 24. Coia, L.R., Engstrom, P.P., Paul, A.R., Stafford, P.M. and Hanks, G.E. (1991) Long term results of infusional 5FU, mitomycin-C and radiation as primary management of
37,660-3. 39. Rider, W.D. and Mendosa, M.D. (1969) Some opinions on treatment of cancer of the esophagus. Am.]. Roentgenol., 105,514-17. 39a. Henschke, U.K. (1960) 'Afterloading'applicator for radiation therapy of carcinoma of the uterus. Radiology,
esophageal carcinoma. Int.J. Radial Oncol. Biol. Phys., 20,29-36. 25. Araujo, C.M., Souhami, L, Gil, R.A. etal. (1991) A
74,834. 40. Tio, T.L, Coene, P.P.L.O., Luiken, G.J.H. and Tytgat, G.N.J.
randomized trial comparing radiation therapy versus
(1990) Endosonography in the clinical staging of
concomitant radiation therapy and chemotherapy in
esophagogastric carcinoma. Gastrointest. Gastroenterol.,
carcinoma of the thoracic esophagus. Cancer, 67,2258-61. 26. DenhamJ.W., Burmeister, B.H., Lamb, D.S.etal. (1996) Factors influencing outcome following radio-
36(2),S2-S10. 41. Clinical applications of endoscopic ultrasonography in gastroenterology. State of the Art 1993. Results of a Consensus Conference, Orlando, Florida. Endoscopy,
chemotherapy for esophageal cancer. The Trans Tasman Radiation Oncology Group. Radiother. Oncol., 40(1), 31-43.
1993; 25,358-66. 42. Gaspar, L.E., Qian, C., Kocha, W.I., Coia, L.R., Herskovic, A. and Graham, M. (1997) A phase l/ll study of external
27. BossetJ.F., Gignoux, M.,Triboulet,J.P. etal. (1997) Chemotherapy followed by surgery compared with
beam radiation, brachytherapyand concurrent
surgery alone in squamous-cell cancer of the esophagus.
chemotherapy in localized cancer of the esophagus (RTOG 92-07): preliminarytoxicity report. Int.J. Radiat.
N.Engl.J.Med.,337,161-7. 28. Kelsen, D.P., Ginsberg, R., PapkJ.f.etal. (1998) Chemotherapy followed by surgery compared with surgery alone for localized esophageal cancer. N. Engl.J. Med., 339,1979-84.
Oncol. Biol. Phys., 37(3), 593-9. 43. Coia, L.R., Minski, B.D.John, M.J. etal. (1999) The evaluation and treatment of patients receiving radiation therapy for carcinoma of the esophagus: results of the 1992-1994 Patterns of Care Study. Cancer, 85(12),
29. Flores, A.D., Nelems, B., Evans, K. etal. (1989) Impact of new radiotherapy modalities in the surgical treatment of esophageal cancer. Int.J. Radiat. Oncol. Biol. Phys., 17, 934-44.
44.
2499-505. Flores, A.D., Nelems, B., Evans, K.G., Stoller, J. and Hay, J. (1987) Combined primary treatment of cancer of the
30. Muller, J.M., Erasmi, H., Stelzner, M., Zieren, N. and
esophagus and cardia by intracavitary and external
Pichmair, H. (1990) Surgical therapy of esophageal carcinoma. Br.J. Surg., 77,845-57. 31. Desai, P.B., Vyas, J.J., Sharma, S., Deshpande, R.K. and
irradiation. In Diseases of the Esophagus, ed. J.R. Siewert.
Badwe, R.A. (1989) Current status of surgical treatment of cancer of esophagus. Semin. Surg. Oncol., 5,359-64. 32. Kavanagh, B., Anscher, M., Leopold, K.etal. (1992) Patterns of failure following combined modality therapy for esophageal cancer, 1984-1990. Int.J. Radiat. Oncol. Biol. Phys., 24, 633-42. 33. Mantravadi, R.V.P., Lad, T., Briele, H. and Liebuer, E.J. (1982) Carcinoma of the esophagus. Sites of failure. Int.J. Radiat. Oncol. Biol. Phys., 8,1897-901. 34. Isono, K., Onoda, S., Ishikawa, T., Sato, H. and Nakayama, K. (1982) Studies on the causes of death from esophageal carcinoma. Cancer, 49,2173-9. 35. Mandard, A.M., Chasle, J., Marnay, J. etal. (1981) Autopsy findings in 111 cases of esophageal cancer. Cancer, 48, 329-35.
Berlin, Springer-Verlag, 741-9. 45. Gaspar, L.E., Nag, S., Herskovic, A., Mantravadi, R. and Speiser, B. (1997) American Brachytherapy Society (ABS) consensus guidelines for brachytherapy of esophageal cancer. Clinical Research Committee, American Brachytherapy Society, Philadelphia, PA. Int.J. Radiat. Oncol. Biol. Phys., 38(1), 127-32. 46. Rowland, C.G. and Pagliero, K.M. (1985) Intracavitary irradiation in palliation of carcinoma of the esophagus and cardia. Lancet, 2,981 -3. 47. Dale, R.G. (1985) The application of the linear quadratic dose-effect equation to fractionated and protracted radiotherapy. Br.J. Radiol., 58, 515-28. 48. Wei-bo, Y. (1990) Personal communication. 49. Wei-bo, Y. (1989) Brachytherapy of carcinoma of the esophagus in China. In Brachytherapy 2 Proceedings of the 5th International Working Conference, ed. R.F. Mould,
256 Brachytherapy in cancer of the esophagus Leersum, The Netherlands, Nucleotron Corporation, 139-41. 50. Yan-jun, M.,Xian-zhi,G., Wei-bo \.etal. (1982) Intracavitary irradiation in the treatment of esophageal cancer. Chin.J. Oncol., 4,45-7. 51. Hishikawa, Y., Kurisu, K., Taniguchi, M., Kamitonya, N. and Miura, T. (1991) High dose rate intraluminal
brachytherapy for esophageal cancer: 10 years experience in Hyogo College of Medicine. Radiother. Oflco/.,21,107-14. 52. Berry, B., Miller, R.R., Luoma, A., Nelems, B., Hay, J.H. and Flores, A.D. (1989) Pathological findings in total esophagectomy specimens, after intracavitary and external beam radiotherapy. Cancer, 64,1833-7.
18 High dose-rate afterloadingbrachytherapy for prostate cancer P.J. HOSKIN
18.1
INTRODUCTION
Brachytherapy as a treatment modality for prostate cancer has been used with varying degrees of enthusiasm for many years. Early techniques used direct open implantation with live source permanent implants, typically gold grains or iodine seeds [ 1 ]. The relatively poor results of this treatment from the 1960s led to disenchantment with this approach, but in the last decade developments in both imaging techniques and brachytherapy equipment have led to this form of high dose radiation therapy becoming an important integral part of the management of localized prostate cancer. Manual afterloading using iridium wire to deliver a medium dose-rate (MDR) implant has been employed in prostate cancer. The largest published series from Long Beach, California, reports 450 patients treated with 30-36 Gy external-beam radiotherapy in 3.5-4 weeks followed by an iridium interstitial implant boost delivering 30-40 Gy in 40-60 h. A local control rate of 92% with a 76% 10-year disease-free survival and 6% moderate to severe morbidity rate was obtained [2]. High dose-rate (HDR) afterloading enables small diameter catheters of around 2 mm external diameter to
be used for interstitial implantation, with the great flexibility of dose delivery inherent in the stepping source system. Catheter placement exploiting the advantages of transrectal ultrasound and computed tomography (CT) imaging to provide accurate positioning and reconstruction of the implant can ensure high-quality implants and precise dosimetry related to anatomical structures. The advantages of high dose localization inherent in brachytherapy can therefore be fully exploited and, in a situation in which a dose response has been demonstrated for radiation therapy [3], an improvement in local rates of control without additional normal tissue morbidity is a realistic expectation. Delivery of radiation at high dose rate, however, carries biological implications which demand careful attention to dose fractionation and dose distribution. The challenge with HDR brachytherapy in prostate cancer is to deliver a safe, effective dose using a technique which will enable several fractions of treatment to be given over several days whilst retaining the high quality of the implant throughout. Unlike low dose-rate (LDR) brachytherapy for prostate cancer, a biological advantage cannot be claimed for HDR over external-beam treatment. The better geographical localization of dose with brachytherapy is
258 High dose-rate afterloading brachytherapy for prostate cancer
a major advantage, but its clinical implementation is probably optimized by using it in conjunction with external-beam treatment. The natural history of prostate cancer includes early invasion of the prostatic capsule and seminal vesicles. A high proportion of patients presenting with apparently localized disease will already have microscopic disease within the immediate vicinity of the prostate gland. Predictors for this include a prostate-specific antigen (PSA) level greater than 10, a Gleason score greater than 7, and a clinical stage of T2b or higher, all of which would carry a greater than 30% chance of regional microscopic disease [4]. In this scenario, external-beam treatment is undoubtedly superior in covering the wider field necessary to incorporate these patterns of spread. In a manner analogous to many other sites where brachytherapy is shown to be a vital component of successful radical treatment, HDR afterloading brachytherapy can then be used as a high dose localized boost treatment to the primary bulk of tumor.
3. The skin is prepared and full sterile theater precautions should be used throughout for the implant to avoid the introduction of infection. 4. The prostate gland is subject to considerable movement, in particular rotation about its axis, if steps are not taken to avoid this. Fixation needles are inserted as the first step in most implant procedures. Positions in the template away from those likely to be immediately implanted are chosen. Two needles, one in each lateral lobe, are usually adequate, although some advocate a third needle anteriorly. 5. A blunt needle to define the length of implantation and position of the bladder neck has also been advocated. The advantage of this is that it gives an additional gauge against which the applicator position can be checked, particularly if imaging quality is variable, as can happen where there is interference from adjacent planes on the ultrasound as the implant builds up.
18.2
18.4 FLUOROSCOPIC IMPLANTATION PROCEDURE
IMPLANT TECHNIQUES
Open procedures are no longer used for prostate implantation. The common route is transperineal, although transrectal approaches are also described. Transperineal implants typically use a template on the skin to define catheter position of entry. A common approach in current use is to combine this with transrectal ultrasound to provide direct real-time imaging as the implant is inserted. Two types of applicator are in common use: rigid needles or flexible plastic catheters. The rigid needles may have the advantage of easier positioning, being less readily deflected by the tissues after skin entry, but they may be more traumatic to the tissues and less well tolerated when an implant remains in place for several days. Fixation of the needles on the perineal skin is also a challenge. 183 PROCEDURE: GENERAL CONSIDERATIONS 1. All patients will require general or spinal anesthesia. The patient is placed in the lithotomy position with the pelvis tilted anteriorly as far as possible. This aids the passage of the applicators below the pubic arch into the more anterior portion of the prostate gland. However, for patients who are elderly, perhaps with degenerative hip disease, care must be taken in positioning the patient while anesthetized to avoid damage to the pelvis and hips. 2. An indwelling urethral catheter will be required and will be retained until removal of the catheters at the end of the last fraction of treatment. This is both to facilitate urinary drainage and also to provide a marker for urethral position, which is an important consideration in catheter placement and dosimetry.
1. The C-arm of the imaging intensifier requires careful positioning to enable imaging of the prostate implant to be seen together with the bladder base. As it is difficult during the procedure to move the C-arm, an antero-posterior or lateral view will be chosen. In general, the lateral view provides better information, particularly for multiplane implants, although some prefer an oblique or direct anteroposterior view. 2. The urinary catheter balloon will be filled with hypaque or a similar contrast-enhancing medium so that it can be readily visualized on the fluoroscope screen. This then defines the bladder base. Ideally, previous CT images will have been obtained in a diagnostic setting to define the position of the prostate in relation to the bladder base. Often, the prostate will be seen to impinge above a catheter balloon in the bladder and it is important this is known so that coverage of the apex of the gland is adequate. 3. Applicators will then be inserted transperineally through a template, if used, and their cranial position, defined by their relation to the bladder base, marked by the catheter balloon. Fluoroscopy will help enable the catheters to be inserted in parallel, although this may become more difficult to judge as sequential planes are built up. 18.5 TRANSRECTAL ULTRASOUND IMPLANTATION TECHNIQUE 1. Transrectal ultrasound (TRUS) is now the standard means of guiding applicators accurately into the
Catheter insertion and fixation 259
prostate gland. For implantation, a probe mounted on a frame with a stepping platform enables reproducible positioning of the catheters. Onto the frame is fixed a template, the image of which can be superimposed on the ultrasound images to relate a catheter position to the anatomy seen, as shown in Figure 18.1. A further feature is to have a dual crystal probe which can produce both transverse and longitudinal images. This then enables catheter placement to be followed in both planes. 2. The role of a pre-implant ultrasound study to define the treatment volume and design an ideal implant prior to the procedure is not mandatory with this technique. Whilst necessary for LDR iodine seed implants, for which the seeds have to be ordered and loaded in a fixed array, the flexibility of HDR afterloading dosimetry means that, provided catheters are inserted in a fixed format - that is, in parallel rows as defined by the template - covering the prostate volume, post-implant dosimetry is adequate to define the dwell positions for coverage of the volume. Pre-implant imaging and dosimetry are not necessary for HDR afterloading, although diagnostic CT or MR images are of value both to give information on the size of the prostate gland and also to predict pelvic arch interference, notwithstanding their role in providing accurate staging information. 3. Setting up of the TRUS prior to implantation is very important and, if not done carefully and accurately, will greatly detract from the final quality of the implant. Important criteria to observe are that the inferior border of the gland is as flat as possible, and considerable variation in gland contour can be achieved simply by altering the angle of the probe, as shown in Figure 18.2. A flat inferior border with the most inferior row of applicators parallel and just inside the gland is required. 4. The urethra should be identified and positioned between vertical rows of applicators around the center
Figure 18.1 Transrectal ultrasound probe and stepping frame set up as used for HDR prostate implant showing template, stabilizing needles, and HDR applicators.
Figure 18.2 Transrectal ultrasound images of the prostate gland showing how shape may change with angle of probe at the same position in the gland.
of the implant and followed throughout its course into the bladder neck to ensure that the prostate is lying straight in its axial plane along the probe.
18.6
CATHETER INSERTION AND FIXATION
Typically, the stepping TRUS technique is used with a 1 cm grid template developing parallel rows of catheters in a square multiplane array. A second template maybe used in addition to the one attached to the ultrasound frame to guide skin entry. Manipulation of the catheters after skin entry may be achieved manually, using the ultrasound images to guide direction. Parallel multiplane implants are generally required; typically, three or four planes will be necessary to cover the depth of the gland. The pubic arch and pelvic side walls may present physical obstacles in placing the most peripheral catheters. Whereas the template technique generally improves geometry, a rigid template against the perineal skin reduces the ability to move the catheter slightly out of plane to avoid the pubic arch or other bony interference. The advantage of flexible catheters is that they may be seen to bend around the pubic arch and then regain
260 High dose-rate afterloading brachytherapy for prostate cancer
their plane in the more anterior aspects of the prostate gland. Extreme pubic arch interference should be predicted prior to the implant from CT images or a preplanned ultrasound, and in some circumstances this may be a contraindication to proceeding with the implant if a satisfactory source distribution cannot be anticipated. Once in position, fixation of the catheters to ensure reproducible positioning on sequential fractions is one of the more difficult aspects of this technique. A rigid applicator may be sutured to the skin and rigid needles fixed into this using a locking device within the template. Rigid needles may be individually sutured to the skin, but where a typical implant may comprise 12 to 20 needles, this becomes somewhat tiresome and difficult. 18.7 MOUNT VERNON APPLICATOR AND TEMPLATE TECHNIQUE At Mount Vernon Hospital we have developed an HDR implant technique using a flexible template and flexible HDR applicators. A standard TRUS system with stepping unit is used for imaging, with the template attached to the ultrasound frame, as shown in Figure 18.1. The template has been modified by increasing the diameter of the guiding holes to enable the flexible catheters to pass through them so that the applicator maybe removed over the proximal ends of the catheters at completion of the implant. Once the ultrasound is placed in position and set up as defined above, a flexible latex template is placed against the perineal skin. This is aligned to be exactly matching the ultrasound template. The flexible template has rubber 'O' rings incorporated in it in a 1 cm grid identical to that of the ultrasound template. It is fastened to the skin using adhesive, the skin having previously been shaved to make removal less painful. The implant procedure then continues with placement of the applicators guided by the ultrasound template passing through the corresponding 'O' ring of the flexible template against the perineal skin. An added advantage of this approach is to give clearance between the two templates which enables digital guidance of the applicators where necessary, the flexible template allowing movement at skin entry. At completion of the implant the ultrasound probe is removed and the ultrasound template is removed over the proximal ends of the flexible applicators. These are then held in position, gripped by the 'O' rings of the flexible template. Retaining sutures in each corner of the flexible template are used to prevent it buckling when the legs are brought together, and this can be trimmed to improve comfort around its edges. The flexible catheters are kept capped to avoid contamination of the HDR channel, their length from the 'O' ring to their distal connecting end is carefully measured, and each catheter is labeled with its grid position. This length is carefully documented and is vital in maintaining quality assur-
ance for the implant as it is measured on each occasion treatment is given to identify any catheter movement. A completed implant in situ is shown in Figure 18.3.
18.8
IMPLANT RECONSTRUCTION
Following recovery from the anesthetic, reconstruction of the implant is required. Although this may be done using the TRUS images obtained during the procedure, we have found that post-implant CT imaging provides better reconstruction of the implant, allowing volume definition on the basis of the anatomical information on the CT scan. It also readily interfaces with the planning system to provide accurate volume and catheter position transfer. Transrectal ultrasound images in general will require manual digitization to import the volumes into the planning system. There are changes following implantation and, in particular, prostate volume increases are recognized when comparing post-implant CT images to TRUS taken during the implant, which may need to be taken into account, and CT gives accurate positioning of the rectum which may be very different once the TRUS probe has been removed. Conventional orthogonal film reconstruction of the implant is, of course, possible and will give adequate information regarding the catheter positions, but none
Figure 18.3 HDR implant of prostate.
Dose prescription 261
in relation to the prostate gland and rectum. Although adequate, therefore, in terms of implant reconstruction, it does not provide the information required for optimization of the implant in relation to the soft tissue. Once the implant has been reconstructed and the catheter positions defined, the dose distribution can begin to be calculated. All the modern HDR systems have associated planning programs with them to facilitate this process. The principles of dosimetry are to deliver a homogeneous dose within the denned volume, with rapid fall-off particularly posteriorly towards the anterior rectal wall, and with a relative cold spot around the urethra. Catheter placement usually takes into account these requirements, with peripheral grid positions being filled, but leaving one or two central positions empty around the urethra. Dwell positions will be defined along the catheters, initially using one of two conventions, either Paris dosimetry or Manchester dosimetry. From this baseline, individual optimization may then be required [5]. Paris dosimetry has the advantage of simplicity in that uniform linear activity, achieved by equal times in each dwell position, is the requirement. Inevitably, however, because an ideal Paris implant requires catheters longer than the volume to be treated, it will, unless the distal end of the implant has been taken well beyond the apex of the prostate, be cold around the apex. In contrast, Manchesterbased dosimetry, with weighting of the dwell positions at the ends and periphery of the implant volume typically aiming for a 2:1 weighting, reproducing the Manchester rule requiring two-thirds of the dose to be delivered to the periphery, will give a more homogeneous dose to the limits of the catheters. Current planning systems have more complex optimization techniques which can then be introduced, although they may not give superior results to conventional individualized planning and often result in marked ranges of dwell time along the catheter. Once satisfactory dose distribution has been obtained, this will be transferred into the HDR control program as a series of dwell times for each catheter. A typical distribution with corresponding dwell times is shown in Figure 18.4. The treatment delivery is relatively simple, although careful and meticulous attention to identify any catheter movement and having a rigorous protocol to ensure that the correct channels are attached to the appropriate catheter are vital. Typical treatment times are very short, most catheters having a total dwell time of only a few seconds, and total treatment times for the implant being of the order of 5 min, depending upon the total number of catheters.
18.9 COMPUTED TOMOGRAPHY-BASED THREE-DIMENSIONAL PLANNING A novel technique has been described from the group in Offenbach [6], relying on three-dimensional reconstruc-
tion of an implant inserted under ultrasound control using four rigid steel needles placed transrectally. Two crossing needles are placed in the upper and lower left and two in the upper and lower right peripheral regions of the gland. An optimized distribution is then defined using three-dimensional reconstructed CT images for HDR afterloading. Because only four needles are used to cover the entire volume, extensive high dose regions are introduced, with up to 50% of the volume receiving 200% of the reference prescription dose. Early clinical results have not revealed significant additional morbidity. 18.10
DOSE PRESCRIPTION
HDR afterloading implants for prostate cancer are rarely, if ever, used as sole treatment. Their use is generally designed to be part of a treatment program in which external beam accounts for two-thirds of the total dose, equivalent to a dose sufficient to eliminate microscopic disease, followed by the remaining one-third of the radical dose delivered by the implant. However, a review of the actual doses prescribed in the various centers using this technique reveals considerable variation, particularly in relation to the implant dose. There appears to be general agreement that a dose of 40-45 Gy in 4-5 weeks or its equivalent is an appropriate external-beam schedule, no doubt reflecting experience from other sites where this approach to radical treatment is successful, for example the head and neck, and cervix. As can be seen in Table 18.1, however, the implant doses typically given in two to three fractions vary enormously [7-12]. This may, in part, be explained by the lack of conformity in dose definition for interstitial implantation and, perhaps more critical in assessing the dosimetry, is the actual dose distribution and relative dose to normal structures compared to the high dose volume. A better comparison of the different dose fractionation schedules may be achieved using the biological equivalent dose (BED) formula, acknowledging the limTable 18.1 Prostate HDR brachytherapy doses
Michigan [7] Oakland,CA [8]
18
Seattle [9]
3
16.5
3
Goteborg[12]
20
2
Kiel [10]
30
2
Berlin [11]
18
2
Offenbach [6]
28
4
Melbourne"
20
4
MVH
17
2
" Personal communication, Professor G. Duchesne. NB: where centers have quoted a range of doses, the most recent schedules have been stated here.
262 High dose-rate afterloading brachytherapy for prostate cancer
Figure 18.4 Dose distribution (a) and catheter dwell times in nominal seconds (b) for completed implant.
Dwell position 1 2 3 4 5 6 7 8 9 10 II 12 13 14 15
1
2
3
4
5
6
0 5 3 3 3 3 3 3 3 2 1 2 3
5 4 3 3 3 3 3 3 3 3 3 4 2
5 3 4 3 3 3 3 3 3 3 3 4 4 2
5 4 1 0 2 2 . 1 1 0 1 1 4 4
4 4 0 1 0 1 0 1 0 0 0 0
0 5 4 3 3 3 3 3 4 3 4 3 2
Channel c.hanrtel 7 8
9
10
II
12
1
5 3 1 0 1 0 0 0
5 3 1 0 1 0 0 0 0 0 0
0 0 3 2 2 2 0 0 0 1 2 2 3 3
5 2 3 2 2 1 1 1 1 1 1 2 2 3
0 0 0 4 3 2 2 1 2
0
5 3 3 3 3 3 3 3 3 4 3 4 I 3
2 3 4 3 3 3 3 3 3 3 3 4 3
1 1
0 4
1
1
3
itations of this when applied to a two-fraction or threefraction implant using large doses per fraction. This is shown in Table 18.2, together with equivalent 2 Gy fraction doses, which are perhaps a more familiar format to consider the dose equivalence of the different schedules. The rectal tolerance in this setting is also uncertain. The nearest analogy would be to gynecological brachytherapy, in which rectal dose is also one of the major limiting normal tissue effects. Indeed, the externalbeam and HDR combination doses used in the prostate cancer schedules are very similar to those for cervical cancer, where the limit for rectal dose would be a 2 Gy equivalent of 64-66 Gy. This, of course, does not take into account the volume effect, whereby a much higher dose may be tolerated to a small volume provided there is ade-
2 2 2 2 3
quate recovery in surrounding areas. This may be more apparent in prostate implantation where 5 mm CT slices or TRUS images are being used to define the dose distribution. The general rule of thumb, however, is to keep the rectal dose within 60% of the tumor dose. This is reflected in the Mount Vernon schedule, in which, from a tumor dose of 8.5 Gy using the HDR implant, the target rectal dose is 5 Gy or less. Equivalent rectal doses from the published schedules are shown in Table 18.3. Similarly, urethral tolerance is poorly denned. It would appear that the urethra is able to tolerate far higher doses than the rectum, but experience from LDR iodine-125 implants confirms that urethral problems can arise if steps are not taken to reduce the central dose of the implant [13].
Complications and toxicity 263
Table 18.2 HDR brachytherapy doses: biological equivalent doses (BED) and 2 Gy equivalents for different oc/P ratios
Michigan [7] Oakland,CA[8]
90.0
38.6
48.0
28.8
28.8
24.0
77.0
33.0
46.7
28.0
25.6
21.3
Goteborg[12]
153.3
65.7
86.7
52.0
40.0
33.3
Kiel [10]
330
141.4
Berlin [11]
126
54.0
72.0
Offenbach [6]
158.7
68.0
86.7
37.2
113.3
48.6
Seattle [9]
Melbourne3 Mount Vernon Hospital b
75.0
62.5
43.2
34.2
28.5
93.3
56.0
47.6
39.7
53.3
32.0
30.0
25.0
65.2
39.1
31.5
26.3
180
108
Personal communication, Professor G. Duchesne.
Table 18.3 Rectal dose from HDR schedules (assuming 60% of tumor dose and o/p = 3)
Michigan [7] Oakland, CA [8]
28.8
17.3
103.3
62.0
Seattle [9]
28.0
16.8
103.0
61.8
52.0
31.2
135.3
81.2
64.8
191.3
114.7
43.2
25.9
115.2
69.1
Offenbach" [6]
56.0
33.6
131.0
78.6
Melbourne*
32.0
19.2
108.7
65.2
Mount Vernon Hospital
39.1
23.5
107.5
64.5
Goteborg[12] Kiel [10]
108
Berlin [11]
"Quoted mean rectal dose is 3 Gy, i.e., 42.8% of prescription dose. ''Personal communication, Professor G. Duchesne.
The disparity and uncertainty regarding dose prescription for this technique make it imperative for all centers undertaking this work to carefully document and report normal tissue toxicity in addition to tumor control rates.
18,11
TREATMENT RESULTS
Reporting of results from this technique is still at an early stage, with most data being relatively immature, particularly in a tumor which has a natural history spanning 10 or 15 years untreated. A summary of the published data is shown in Table 18.4. The only conclusions that can be drawn are that this is an effective form of treatment, that high control rates from locally advanced disease as well as early prostate cancer can be achieved, and that, to date, morbidity is within acceptable limits for a high dose radical radiotherapy schedule, although later normal tissue effects require clarification.
While single center reports are of value, the ultimate question that will arise is how this technique relates in its outcome to LDR iodine-125 seed implantation and optimized external-beam treatment alone using conformal and possibly intensity modulated radiotherapy techniques. Currently, only one randomized trial is underway at Mount Vernon Hospital, comparing a standard external-beam technique of 55 Gy in 20 daily fractions with the HDR schedule shown above.
18.12
COMPLICATIONS AND TOXICITY
In general, this technique appears well tolerated, with no additional complications other than those which would be anticipated from high dose pelvic radiotherapy. The implant procedure itself appears straightforward, notwithstanding the inevitable risks of anesthesia in a patient population which is predominantly elderly. This is, of course, amplified with techniques and programs
264 High dose-rate afterloading brachytherapy for prostate cancer
Table 18.4 Clinical results
Michigan
33 [7]
T2:26 T3: 7
91
Bowel:
86
59 [14]
5
7
Oakland, CA [8]
110
85
Rectal: Urinary:
1 4
Seattle [9]
104
84
Bowel: Urethra I:
0 7
72:110 T3: 59
T2:89 T3: 85
Bowel: Bladder:
3 7
82
T2:21 T3:61
All: 53
Bowel:
3.6
59
Not reported
Goteborg[12]
171
Kiel [10] Berlin [11] Offenbach [6] Melbourne
Not reported
3
Mount Vernon Hospital
None
Not reported Not reported
° Personal communication, Professor G.Duchesne.
which deliver successive fractions with a new implant on each occasion. During the implant itself, there may be minor discomfort, but this is usually controlled with simple or moderate analgesia. Indeed, after the first hour or so of recovering from the anesthetic, most patients require no additional anesthesia. Hematuria is common, but rarely of great consequence. Rectal symptoms are few. Some disturbance of bowel function often follows, our own schedule incorporating an enema preoperatively to empty the bowel followed by a constipating regime for 24 h to prevent bowel motions whilst the implant is in situ. Many men then find it takes a week or two for bowel motions to return to a normal pattern and may require a gentle laxative for a few days. Urinary symptoms once the catheter is removed may persist for a week or two, with mild dysuria and hematuria. Occasionally, patients who have had significant outflow obstruction with symptoms prior to the implant may require catheterization for a period of a week or two after the implant whilst the initial edema settles. Long-term catheterization, however, has not been encountered in our experience. Urinary incontinence as a complication is not reported. Experience from LDR implants suggests that transurethral resection is a risk factor for urinary symptoms and incontinence after implantation [13] and we regard this as a contraindication at present, although data on HDR implants after transurethral resection (TURP) are not currently forthcoming. The incidence of expected late complications such as rectal fibrosis and stenosis, bladder telangiectasia, and urethral stricture is unknown on examination of the literature. The most mature data to report this from Kiel [10] and Berlin [11] suggest the incidence will be of the order of 3% for severe bowel complications, with per-
haps a similar or slightly higher incidence of severe urinary symptoms.
18.13
CONCLUSION
HDR afterloading implants to the prostate gland can be achieved with a high level of technical accuracy and enable the delivery of a concentrated high dose radiation treatment to a defined volume encompassing the prostate gland. Doses to the rectum can be limited to remain within normal tissue tolerance and adjustment of catheter placement and dwell positioning can also minimize the dose to the urethra. In combination with external-beam treatment, this approach offers a treatment program which combines a moderate radiation dose sufficient to control microscopic disease along patterns of regional spread with a central high dose treatment to known sites of macroscopic tumor. As such, it is both conformal and intensity modulated, and early results suggest it is highly effective in the management of localized prostate cancer.
REFERENCES 1. Khan, K.( Crawford, D.E. and Johnson, E.L (1983) Transperineal percutaneous iridium-192 implant of the prostate. Int.J. Radial Oncol. Biol. Phys., 9,1391-5. 2. Puthawala, A., NisarSyed, A.M., Austin, PA.etal. (1996) Combined interstitial iridium-192 implant and external beam irradiation in the treatment of carcinoma of the prostate. Radiother. Oncol., 39 Suppl. 1, S1. 3. Hanks, G.E., Kerring, D.F. and Kramer, S. (1985) Patterns of
References 265
4.
5.
6.
7.
8.
9.
care studies: dose-response observations for local control of adenocarcinoma of the prostate. Int.J. Radiat. Oncol. Biol.Phys., 51,153-7. Partin, A.W., Subong, E.N.P., Walsh, P.C. etal. (1997) Clinical stage and Gleason score to predict pathological stage of localized prostate cancer. JAMA, 277,1445-51. Hoskin, P.J. and Rembowska, A. (1998) Dosimetry rules for brachytherapy using high dose rate remote afterloading implants. Clin Oncol., 10,226-30. Martin, T., Kolotas, C, Dannenberg, T. etal. (1999) New interstitial HDR brachytherapy technique for prostate cancer: CT based 3D planning after transrectal implantation. Radiother. Oncol., 52,257-60. Stromberg, J., Martinez, A., Gonzalez, J. et al. (1995) Ultrasound-guided high dose rate conformal brachytherapy boost in prostate cancer: treatment description and preliminary results of a Phase l/ll clinical trial. Int.J. Radiat. Oncol. Biol. Phys., 33,161-71. Rodrigues, R. and Demanes, D.J. (1997) HDR brachytherapy: ultimate in conformal radiotherapy for the treatment of prostate cancer. In New Developments in Interstitial Remote Controlled Brachytherapy, ed. N. Zamboglou. Munich, WZuckschwerd undVerlag, 119-25. Mate, T., Gottesman, J., Hatton, J. et al. (1998) High dose rate afterloading iridium-192 prostate brachytherapy:
10.
11.
12.
13.
14.
feasibility report. Int.J. Radiat. Oncol. Biol. Phys., 41, 525-33. Kovacs, G., Wirth, B., Bertermann, H. etal. (1996) Prostate preservation by combined external beam and HDR brachytherapy at nodal negative prostate cancer patients - an intermediate analysis after ten years experience. Int. J. Radiat. Oncol. Biol. Phys., 36 (Suppl.) S80,198. Dinges, S., Deger, S., Koswig, S. etal. (1998) High-dose rate interstitial with external beam irradiation for localized prostate cancer - results of a prospective trial. Radiother. Oncol., 48,197-202. Borghede, G., Hedelin, H., Holmang, S. etal. (1997) Irradiation of localized prostatic carcinoma with a combination of high dose rate iridium-192 brachytherapy and external beam radiotherapy with three target definitions and dose levels inside the prostate gland. Radiother. Oncol., 44,245-50. D'Amico, A.V. and Coleman, C.N. (1996) Role of interstitial radiotherapy in the management of clinically organconfined prostate cancer: The jury is still out. J. Clin. 0ncol.,14,304-15. Rodriguez, R.R., Demanes, D.J. and Altieri, G.A. (1999) High dose rate brachytherapy in the treatment of prostate cancer. Hematol. Oncol. Clin. North Am., 13, 503-23.
19 Low dose-rate brachytherapy for breast cancer JULIA R. WHITE AND J. FRANKWILSON
19.1
INTRODUCTION
Low dose-rate (LDR) brachytherapy has had a long history of use for the treatment of breast cancer. Its primary use has been as a boost following whole breast radiation therapy (WBRT) as part of breast conserving treatment (BCT) for early-stage disease. Additionally, it has had some broader application for the treatment of locally advanced as well as locally recurrent breast cancer. As a result of both technical and philosophical changes in the radiation treatment of breast cancer, the use of LDR brachytherapy has declined significantly. Despite this, there remains a role for LDR brachytherapy in breast cancer treatment.
19.2
HISTORICAL PERSPECTIVE
LDR brachytherapy played an important part in the dramatic practice shift over the past three decades toward BCT as the preferred local therapy for eligible breast cancer patients [1]. While most of the transition to BCT happened over the last two decades, it was actually just shortly after the turn of the twentieth century that visionary investigators began the insertion of radium needles into primary and recurrent breast cancer as an alternative to surgical resection [2]. As early as 1910, Dr Henry Janeway, a surgeon at Memorial Hospital in New York, gained experience treating primary and metastatic breast cancer with the insertion of capillary glass seeds containing radon [3]. In 1924, a British surgeon, Sir
Geoffrey Keynes, began systematically treating primary breast cancer with radium needle insertion instead of radical mastectomy. He routinely included the breast, as well as the supraclavicular, infraclavicular, axillary, and internal mammary regional nodal areas in his treatment volume (Figure 19.1) [4]. In 1929, reporting his results of treating 90 women with definitive interstitial irradiation, he proclaimed, 'at the present time we regard treatment by buried needles as preferable to operation,' for operable breast cancer. Subsequently, in 1937 Keynes modified his technique to include tylectomy of the primary breast mass prior to radium needle insertion [5]. Similarly, in 1932, Dr George Phaler, a Philadelphia radiologist, reported his experience using radium 'for intensive local effect by interstitial implantation' in combination with roentgen rays for primary treatment of 127 breast cancer patients [6]. He reported an 81% control rate at 'well over 5 years' for 40 patients with resectable breast cancer treated with radiation alone because they had either refused surgery or were medically inoperable. This early work with interstitial LDR implants became the foundation of experience upon which BCT would be built, and preceded the first published reports of external-beam therapy with or without tylectomy by at least 10 years [7,8]. The development of better X-ray generators and greater understanding of radiobiologic principles, combined with frustration with the stagnant results and physical deformity from radical mastectomy, led to the emergence of more reports utilizing external radiation for BCT [9-11]. Shortly after the Guy's London Trial [12] comparing tylectomy and breast radiation to radical mastectomy was initiated in 1955, Pierquin et al. began
Early-stage breast cancer 267
Figure 19.1 Distribution of radium needles used by Sir Geoffrey Keynesfor the definitive treatment of operable breast cancer between 1924 and 1929 [4].
routinely treating operable breast cancer with radical radiation therapy alone, consisting of whole breast external-beam irradiation (WBRT) and an LDR interstitial implant into the tumor-bearing quadrant [13]. Similar work combining WBRT and brachytherapy after tumor excision for BCT began in the USA as well [14]. These single institution experiences, by consistently demonstrating comparable survival and good local control in comparison to treatment with radical mastectomy, helped further the acceptance of BCT for treatment of early-stage breast cancer. Their work ultimately led to the prospective randomized trials [15-20] which established BCT as the preferred alternative to mastectomy in eligible patients. LDR brachytherapy remains a part of the radiation management of breast cancer some 80 years after its inception.
193
EARLY-STAGE BREAST CANCER
193.1 Low dose-rate brachytherapy as boost treatment The predominant use of LDR brachytherapy has been in the setting of BCT to deliver a 'boost' or supplemental dose of radiation to the tumor-bearing quadrant of the
breast after WBRT. This was an ideal application for an interstitial implant, given that the anatomy of the breast is readily accessible to an implant and the resulting dose distribution allows a localized area of high dose, while sparing much of the remaining breast, underlying chest wall, and lung. Until linear accelerators capable of generating high-energy electron beams became generally available, an interstitial LDR implant was the method of choice for delivering the boost dose for patients treated with BCT. The wide acceptance of LDR brachytherapy as a boost technique is evidenced by the fact that in two of the randomized trials comparing BCT to mastectomy for early-stage breast cancer, an LDR interstitial implant was part of protocol radiation [18,19]. There has been considerable variation in terms of the surgery, radiation, and implant techniques utilized among different institutions (Table 19.1). However, in general, 10-30 Gy was delivered by LDR irradiation over 1-3 days, either before or after WBRT of 45-50 Gy had been given over a 4-5week time period. This yielded local breast recurrence rates varying from 3.7% to 16%, after 5-15 years of follow-up (Table 19.2). Over the past decade, many institutions have abandoned using an LDR implant for the boost dose. Results from the 1983 Radiation Oncology Patterns of Care Study in the USA for definitive breast irradiation demonstrated 29.8% of boost doses were delivered with an LDR implant [32]. In comparison, the 1989 US Patterns of Care Survey included 449 cases from academic (30%), hospital-based (38%), and free-standing (32%) radiation oncology practices. Ten percent, 2%, and 13% of these respective practices used an LDR interstitial implant for the boost technique [33]. This change in practice pattern probably reflects a combination of the availability and convenience of high-energy electron technical capabilities and the controversy associated with routinely boosting the lumpectomy bed after WBRT. The rationale for boosting the tumor-bearing quadrant is supported by the work of Holland et a/., who performed histologic examination of 282 mastectomy specimens from breast cancer patients with solitary tumors less than 5 cm in size who would otherwise have been good candidates for treatment with BCT. The amount and distribution of microscopic tumor foci outside the index mass in the adjacent normal breast tissue were mapped. This demonstrated that in 177 (63%) of cases, tumor foci were found outside of the index lesion, with 20% of these being within a distance of 2 cm, and 43% at distances greater than 2cm [34]. In view of the high risk of residual microscopic tumor burden in the 2-4 cm of tissue surrounding the index mass, and the cosmetic deformity of such a large excision margin in most cases, boosting the affected area with higher localized doses of radiation seemed a logical solution. Multiple prospective and retrospective series utilizing more limited surgery in combination with WBRT and a boost have demonstrated breast recurrence rates of
Table 19.1 Techniques from various institutions using LDR brachytherapy as a boostfollowing whole breast treatment for breast conserving therapy
Netherlands Cancer Institute [21 -24]
WLE
Positive Close Negative Unknown
9 15 46 30
Positive Negative Unknown
15 56 2.9
50/2
15-25
Paris3
2
15
2.5
45/1.8
15-20
0.30-0.5
2
17.5
N/A
Thomas Jefferson University/ University of Kansas [25]
WLE
Hopital Henri Mondor, Institut Gustave-Roussy [26]
Tumorectomybfor T size <3 cm
N/A
25-37
25 37
Paris
2
N/A
N/A
Harvard University, JCRTC [27,28]
Excisional biopsyd
N/A
45-50/1.8-2
10-30
0.4-0.5
1-2 2
24
N/A
West Mead Hospital Australia [29]
Excisional biopsyquad rantectomy
N/A
50/2
30
0.4-0.5
1
7.6
3
EORTC1081[18]
WLE
50/2
25
Paris
2
N/A
NCI [17]
Excisional biopsy
48.6/1.8
15-20
N/A
N/A
N/A
William Beaumount Hospital [30]
WLE
45-50/1.8
15
0.625
N/A
N/A
WBRT, whole breast radiation therapy; WLE, wide local excision - removal of tumor with 2 cm margin of normal adjacent breast tissue; N/A information not available. 'Paris: reference dose rate was prescribed according to the Paris System, i.e., 85% of the basal dose rate [31]. Tumorectomy - excision of tumor with a margin of normal tissue not exceeding 1 cm. The Joint Center for Radiation Therapy. d Excisional biopsy - complete gross tumor excision.
19
5
Table 19.2 Outcomes from various institutions using LDR brachytherapy as a boost following whole breast treatment for breast conserving therapy
Netherlands Cancer Institute [21 -24]
774 (1026)
66
T1 T2
64 36
N/A
3.1
91
86
Thomas Jefferson University/University of Kansas [25]
655
48
T1 T2
71 29
88
V
93
79
Hopital Henri Mondor, InstitutGustaveRoussy [26]
245
T1 T2 T3
25 56 19
Harvard University (JCRT) [27,28]
688C (783)
80
T1 T2
54 46
60
83
T1 T2
69 31
N/A
T1 T2
38 62
67
West Mead Hospital, Australia [29] EORTC1081[18]
180b
131 456
72
13
11 N/A 8.5
83
77
10
NCI [17]
98 (121)
120
T1 T2
43 51
N/A
16
William Beaumont Hospital (30)
277
81
T1 T2
68 32
89
lodine-1252 lridium-1927
"Actuarial. b Minimum follow-up. C 24 patients had mixed electron + implant boost. "8-year survival. N/A Not available.
T1 T2 T3
66
77
86
81d
63 51 28
270 Low dose-rate brachytherapy for breast cancer
3-10% at 5 years, and 10-18% at 10 years [15-21,25,29,35-42]. This is significantly less than would be predicted from limited excisions alone by Holland's data and has supported the concept that radiation can eradicate microscopic residual tumor burden left in the remaining breast. The most convincing evidence against the routine use of a boost dose after WBRT comes from the NSABP B06 trial that compared modified radical mastectomy, lumpectomy, and lumpectomy with radiation for breast cancers up to 4 cm in size [17]. For this trial, the radiation dose was 50 Gy, delivered over 25 fractions to the whole breast without systematic boosting of the tumor-bearing quadrant. After 12 years of follow-up, the 10% breast relapse rate after lumpectomy and WBRT is comparable to that in other trials in which boosting of the tumor bed was routinely employed [15-20]. The NSABP B06 recurrence rates may reflect that negative microscopic margins of resection were a prerequisite or a mastectomy was performed. There are retrospective series which have corroborated the NSABP B06 experience, demonstrating good local control without boosting for patients with negative resection margins [35,36]. So far, the preliminary results from three randomized trials addressing this issue have begun to clarify whether a boost to the tumor bed should be routinely used when resection margins are negative. In the Lyon, France, trial, 1024 women were randomized to receive a 10 Gy boost delivered with electrons after WBRT of 50 Gy given over 20 fractions. With a median follow-up of 3.3 years, the local failure is 3.6% versus 4.5% (p=0.044), with and without the addition of the boost, respectively [43]. In contrast, a similar trial from Nice, France, found no difference in local recurrence after 6 years follow-up in 664 women randomized after 50 Gy WBRT in 5 weeks with cobalt-60 to either a 10 Gy boost or observation: 4.3% versus 6.8% (p = 0.13) local recurrence, respectively [44]. An EORTC randomized trial (22881/1882) has been done to identify the optimal radiation dose after lumpectomy in 5569 patients. Patients with negative margins (n = 5318) of resection were randomized to no boost versus 16 Gy boost after 50 Gy/2 Gy fractionation WBRT. After 5.1 years followup, the 5-year actuarial recurrence rate is 6.8% with the boost versus 4.3% without one (p = 0.0001). The full results from this trial are anxiously awaited. Boosts to the tumor-bearing quadrant are consistently recommended in the setting of microscopic margin involvement after local excision of the breast cancer. While many reports support that there is an increased risk of local failure after BCT when microscopic surgical margins are positive [35,37-40,46], others have found no correlation between positive microscopic margin status with subsequent breast relapse [41,42]. This may be partially explained by the relatively higher cumulative total doses to the excision cavity from boosting when microscopic margins were positive in these studies [41,42]. The association of adverse pathologic features
with higher rates of breast relapse has been reported from prospective randomized trials [44-48], as well as many retrospective series [21,22,27,31,37,38]. For example, a review of the pathologic findings from the NSABP B06 trial revealed a non-significant correlation between the presence of lymphovascular invasion and increased local breast recurrence (8% versus 3%, p=0.07) [49]. Other factors which have been associated with a higher likelihood of breast relapse after BCT when present include tumor size, nuclear grade, histologic grade, and an extensive intraductal component [21,27,46-48]. Along with adverse pathologic features, other patientrelated and tumor-related factors, such as young age and hormone receptor-negative tumors, have been associated with higher breast relapses after BCT [21,27,38]. All of these factors need to be taken into consideration when determining the planned final dose to the tumor-bearing quadrant. What is known is that the addition of a boost is safe and will not compromise cosmesis if administered appropriately, with either an implant or electrons (Tables 19.2 and 19.3). The characteristics of electron externalbeam therapy, with the sharp drop-off in dose beyond its useful range, makes it ideal for treating superficial volumes. It is very suitable for delivering a boost dose to a tumor bed located in a small to moderate sized breast or in the superficial aspect of a large breast. Typical electron energies used for delivering a boost range from 8 to 12 MeV, with the goal of encompassing the volume within the 80% isodose line. Energies higher than 12 MeV have been associated with a poorer cosmetic outcome [50], related to more frequent telangiectasis and fibrosis. Electron boosts also provide the important advantages of not requiring a surgical procedure or a hospital stay, as is necessary for an LDR interstitial implant. A few institutions have examined their experience retrospectively to evaluate the influence of the type of boost on the outcomes of local breast recurrence and cosmesis (Tables 19.3 and 19.4). The breast recurrence rates were comparable whether an LDR implant or an electron boost had been delivered, despite the differences between the series regarding surgical techniques, radiation doses, and degree of resection margin assessment. Perez et al, in their series from Washington University, USA, observed improved local control for a small subset of patients with T2 lesions when boosted with an LDR implant to a cumulative dose of 70.2 Gy (1/24, 4%) versus lower doses (4/14, 28.6%) [52]. Mansfield et al. reported an advantage in local control for stage II patients who received an LDR implant boost versus electrons [25]. Similarly, Deore et al. noted more local recurrences among T2 patients who received an electron boost (4/13,31%), versus an LDR implant (15/157,10%) [55]. In contrast, Wazer et al. reported more local recurrences at 7 years in patients treated with an implant boost versus non-implant boost (9% versus 2.2%) [56].
Table 19.3 Comparison of outcomes from implant versus electron boost for breast conserving therapy
Hopital Tenon, Hopital Pitie, Salpetriere [51]
82
45-46/2-2.5
Washington University [52]
67.2
48-50/1.8-2.0 119
169
15.3
10-20
Thomas Jefferson University, University of Kansas [25]
40
45/1.8
654
William Beaumont Hospital [30]
81
45-50/1.8
87(iodine-125) 15 190(iridium-192)
Beth Israel Medical Center [54] 'Actuarial recurrence. N/A Not available.
46
50/2
35
1 5-20
15
8.1 6.7
61
160
14.6
81.5
493
10-20
3
7 (5 years) 91 12 (10 years) 2a
89
13.5
82.5
6
81 3
416
20
8 (5 years) 95 19 (10 years)
104
N/A
5a
90
45
15
0
N/A
7 8.6
N/A
272 Low dose-rate brachytherapy for breast cancer
However, implants were performed only for cases with indeterminate or final microscopic margins < 2 mm. Breast relapses were not significantly different between implant versus non-implant boosts when this higher risk group of patients was analyzed separately. In these series, cosmetic results were equivalent, irrespective of which boost method was employed. Touboul et al. found poorer cosmesis in their group of patients who had an LDR implant boost, but attributed this difference to the fact that patients who received an implant boost had WBRT with Cobalt-60, which delivers a greater skin dose than the 4-6 MV photons used to treat the patients boosted with electrons [51]. Olivotto etal. also reported an initially worse cosmetic outcome in association with larger volume implants that was defined surrogately based upon the number of iridium-192 seeds utilized [28]. A later series, which included this cohort of patients, failed to demonstrate a negative impact on cosmetic outcome associated with an implant boost [57]. As seen in Table 19.3 and reported by others [58,59], cosmetic results are not compromised by an LDR implant boost when careful attention is paid to technical and dosimetric factors. Two distinctly different randomized trials have evaluated the outcomes from delivering an LDR implant versus an electron boost. The first of these is RTOG 83-06, which accrued 285 patients between 1983 and 1988 in an attempt to compare electron versus LDR implant boost for breast cancer patients with pathologically negative axillary nodes and Tl, T2 tumors [60]. The breast surgery was an excisional biopsy to obtain clear gross margins of > 1 cm. After WBRT was delivered to a total dose of 46 Gy over 23 fractions using 4 or 6 MV photons, a 16 Gy or 20 Gy boost was delivered, depending on the status of the microscopic margins of resection. Unfortunately, to date no results have been reported. The other trial, performed at the Institute Curie, was conducted in more advanced cases of primary breast cancer with tumor sizes of 3-7 cm treated with radiation therapy alone after incisional or core needle biopsy for diagnosis. After 57.6 Gy at 1.8 Gy fractionation WBRT, patients were randomized to receive an additional 37 Gy with either an iridium-192 LDR implant or cobalt-60 external therapy. After a median follow-up of 8 years, there was a higher breast relapse rate in patients who received the cobalt-60 external therapy boost compared to those boosted by an implant: 45/129 (35%) versus 28/126 (22%), (p = 0.024) [61]. Therefore, while the boost modality probably does not influence outcome after lumpectomy and WBRT, in the setting of radical radiation, an LDR interstitial boost appears necessary to improve local control. Despite the efficacy and convenience of electrons, there is a specific role for LDR implants as a boost after WBRT. Brachytherapy is the preferred boost method for a deep tumor within a large breast. Tumor beds that are greater than 4-5 cm from the skin surface are generally
beyond the useful range for electron irradiation. Electron boosting in this setting leads to significantly higher doses to both the overlying skin and underlying chest wall and lung in comparison to an LDR implant [62]. A brachytherapy boost is advocated by some investigators for patients at a higher risk of local failure after CSRT [25,51-53]. This generally includes patients with involvement of microscopic resection margins, an extensive intraductal component, or other poor pathologic risk features. The main rationale for an implant in these higher risk patients is that it permits intensification of radiation doses into a localized area with less morbidity than is seen with electrons or photons. Krishnan et al. have proposed that there is, in addition, a radiobiologic advantage to boosting with an LDR interstitial implant, particularly when performed at the time of breast surgery [63,64]. Specifically, these theoretical advantages include the dose differential of greater than 50% at the center of the implant versus the periphery, continuous irradiation provides greater dose per cell cycle than conventionally fractionated external-beam radiation, and greater recruitment of cells into more radiation-sensitive phases of the cell cycle [63]. At The Medical College of Wisconsin, the patient charges for an electron boost are nearly half of those generated for an LDR implant. It should be noted that patient billing practices vary widely among institutions. However, when indicated, the use of an LDR implant may represent an overall economic advantage when the cost of treating a local recurrence or poor cosmetic outcome is factored into the equation.
193.2 Brachytherapy as the sole radiation treatment The 5-6-week treatment duration necessary for WBRT makes BCT prohibitive for certain women who either live a great distance from a radiation center or are elderly and/or disabled. In order to reduce the overall treatment time and improve utilization of BCT, phase I/II trials utilizing interstitial implants to the tumor-bearing quadrant as the sole method of radiation after lumpectomy have been conducted [65-69]. The rationale for this approach is based upon the local failure pattern within the breast after lumpectomy with or without breast irradiation. Multiple randomized trials examining treatment with lumpectomy alone versus lumpectomy and radiation therapy consistently demonstrated a significantly reduced risk of local recurrence with the addition of radiation therapy [17,20,45,70,71]. In examining the local failure patterns in these trials, it is noted that the local recurrences occur most commonly at the site of the original tumor excision and recurrences elsewhere in the breast are relatively uncommon [70-72]. Following this argument, it may be that the dominant benefit derived by the addition of radiation therapy is primarily
Early-stage breast cancer 273
from reducing local recurrences at the original excision site. Radiation to the tumor-bearing quadrant alone following lumpectomy was evaluated in the Christie Hospital Breast Conservation Trial, which randomized breast cancer patients with tumors smaller than 4 cm to tumor bed radiation with electrons versus WBRT after lumpectomy [65]. No attempt was made to obtain clear surgical resection margins. With a 65-month median follow-up, the overall survival is similar between the two groups. The local-regional recurrence rate was significantly higher in the local-field-alone arm (19.6% versus 11%, p = 0.0008), particularly in patients with infiltrating lobular histology or an extensive intraductal component. The authors concluded that, with more stringent selection criteria, radiation to the tumor-bearing quadrant alone might be sufficient. The necessity for careful selection of patients to receive treatment to the tumor-bearing quadrant alone after lumpectomy is illustrated by the Guy's trial [66]. After tumorectomy with no attempt made to achieve clear resection margins, an iridium LDR implant of the tumor bed was performed delivering 55 Gy over a 5-day time period to 27 patients. The mean tumor size was 3.5 cm and 10/27 (37%) had margin involvement. After a 6-year median follow-up, the local failure rate was 36%. This unacceptably high rate of local failure probably reflects the poor selection criteria and limited surgical excision. Other investigators have used stricter selection criteria for doing an interstitial implant as the sole radiation method after lumpectomy. At the Oschner Clinic, 51 patients have received an implant after lumpectomy as the sole radiation treatment [67]. Twenty-six of these were with an iridium-192 LDR implant administering 45 Gy in 3.5-6 days. At 20 months' median follow-up, no local recurrences have been observed. At William Beaumont Hospital, 50 patients have received an LDR implant with iodine-125, giving 50 Gy at 0.52 Gy h~' after lumpectomy [68]. There have been no local regional failures after 47 months' median follow-up. In each of these trials the average tumor size was 1.5 cm and negative microscopic resection margins were a prerequisite. Excellent or good cosmesis was noted in 71% and 98% of cases at the Oschner Clinic and William Beaumont Hospital, respectively. A similar pilot study using high dose-rate (HDR) brachytherapy has been performed at the London Ontario Regional Cancer Center using comparable selection criteria but with somewhat smaller volume implants. It reports one local recurrence in 39 patients after 20 months' median follow-up [69]. These results are promising, but longer follow-up is needed to determine if this method yields results comparable to WBRT after surgery. The Radiation Therapy Oncology Group (RTOG) has recently completed a Phase II trial to test the reproducibility of these results in a multi-institutional setting. Patients eligible for this trial had tumors < 3 cm in size,
none to three axillary lymph nodes involved, and must have negative surgical margins after lumpectomy. Tumors with an extensive intraductal component and invasive lobular histologies were excluded. Implants were either LDR or HDR based upon institutional preference. The LDR dose was 45 Gy/3.5-6 days. 193.3 Brachytherapy for local recurrences after breast-conserving therapy Standard treatment for local recurrence after conservative surgery and radiation therapy is mastectomy. When a patient is medically inoperable or refuses mastectomy, brachytherapy alone or after a simple excision represents a potential alternative. Maulard et al. reported their experience with 38 patients utilizing brachytherapy either alone (n = 23) for tumors > 3 cm or after excision (n = 15) for tumors < 3 cm in size [73]. A total dose of 30 Gy was delivered by a single LDR implant after excision. For treatment with brachytherapy alone, two separate insertions were done, delivering cumulative total doses of 60-70 Gy. For the entire group, the second local recurrence rate was 21% after a mean follow-up of 40 months. With a mean follow-up of 48 months, 4/15 (27%) of the excision and brachytherapy group had a second local recurrence, in comparison to 4/23 (17%) in the brachytherapy-alone group after a mean follow-up of 36 months. The 5-year overall and disease-free survivals were 61% and 31% respectively, for the entire population. Local control in the group treated with brachytherapy after excision might have been better had a higher dose than 30 Gy been used. Toxicity included necrosis in one patient and chronic breast pain in another, both requiring mastectomy. In comparison, Kurtz et al. reported a 32% second local recurrence rate after a median follow-up of 51 months for 50 patients with breast failure after BCT treated with local excision alone [74]. While mastectomy remains the recommended treatment for recurrence after BCT, brachytherapy after local excision represents a potential alternative in selected cases.
19.4
LOCALLY ADVANCED BREAST CANCER
In an attempt to expand the benefits of BCT to patients with locally advanced breast cancer, LDR interstitial implants have been used as a boost after induction chemotherapy and WBRT alone or in combination with lumpectomy (Tables 19.4 and 19.5). The total implant dose delivered tended to be slightly higher in this setting, ranging from 25 Gy to 40 Gy (Table 19.4). Local control rates from these series range from 74% to 91%, with the majority of patients treated with exclusive radiation. Breast preservation rates were 78-94% of attempted cases. This held true even when larger tumors and more
274 Low dose-rate brachytherapy for breast cancer Table 19.4 Techniques from various institutions using LDR brachytherapy as a boost following whole breast treatment for more locally advanced breast cancer
Hopital Tenon, Hopital Pitie Salpetriere [75]
Induction:3 4 cycles Adjuvant:1312 cycles
None 33 WLE 27 MRM 37
45/1.96
25-30
Centre Hospitaller et Universitaire[76]
Induction:0 3 cycles Adjuvant:" 6 cycles
None 32 WLE 45 MRM 81
45/2.25
15-35
Neckar Hospital [77]
Induction:6 4-6 cycles Adjuvant:'5-12 cycles
None
23/5.0-6.3
20-30
Hopital Pitie Salpetriere [78]
Induction only
None
45/1.8 23/1.5-6.5
25-30
Memorial Medical Center of Long Beach [79]
None
Incisional biopsy
50/2.0
Breast 30-40 Axil la 20-30
a
5-Fluorouracil, adriamycin, cyclophosphamide. Cyclophosphamide, methotrexate, 5-fluorouracil. c Mitoxantrone, vincristine, cyclophosphamide, 5-fluorouracil. "As above, with epirubicin. 'Vinblastine, thiotepa, methotrexate, 5-fluorouracil, prednisone with adriamycin. WLE, wide local excision; MRM, modified radical mastectomy. b
advanced stages were treated. Survival was comparable to those series with similar chemotherapy regimens which routinely employed mastectomy [80]. As reviewed earlier, the Institute Curie has demonstrated improved local control from an LDR interstitial implant versus an electron boost in a prospective randomized trial [61] when treating primary breast cancers 3-7 cm in size with radiation therapy alone without definitive surgery. Therefore, in this setting, an LDR interstitial implant is the recommended and preferred boost method.
19.5
TECHNIQUE
Iridium-192 is the most commonly used isotope for LDR breast implants (see Table 19.1). Alternatively, a few institutions have used iodine-125 and reported clinical advantages in comparison to iridium-192 in terms of easier shielding and dose optimization [53]. However, iodine-125 requires significant physics time and expertise for the assembly and dismantling of custom ribbons. In most cases, a minimum of two planes is necessary for appropriate coverage by the implant of the lumpectomy site. A double-plane implant can generally be used to treat volumes with thicknesses up to 2.5-3 cm. For treatment volumes with thicknesses > 3-3.5 cm, a three-plane implant is generally warranted to improve dose homogeneity [81]. Single-plane implants are reserved for treatment volumes with a thickness of 1 cm
or less, which may occur in very small breasts or in the periphery of the breast, particularly the upper inner quadrant. Spacing between planes and the individual ribbons within a plane typically varies between 1 and 2 cm. The planes can be arranged so that the needles in opposing planes are parallel, creating a square configuration between needles in different planes. For breast implants, the needles in adjacent planes are commonly staggered. This means that the needles in one plane are situated at half the in-plane separation of the opposing plane, so that there is a triangular configuration between needles in different planes (Figure 19.2a). This latter arrangement has been used extensively in the Paris System for interstitial radiation therapy, for which the recommended spacing between individual needles in the same plane is 15, 18, or 20 mm and the interplanar separations are 13, 16, and 18 mm, respectively, or nearly equidistant [82]. There is substantial clinical experience supporting the efficacy of the Paris-type arrangement for LDR interstitial implants in the breast (see Table 19.2). In using the Paris System the operator should be aware that, as the distance between needles and planes increases to cover larger target volumes, so does the relatively higher dose region in the center of the implant [83]. For this reason, others have recommended varying needle and plane separations to improve dose homogeneity [84,85]. As an example, Zwicker et al. use a Quimby-like system which maintains a constant needle spacing of 1 cm within the plane as the interplanar spacing varies with target thickness in
Table 19.5 Outcomes from various institutions using LDR brachytherapy as a boost following whole breast treatment for more locally advanced breast cancer.
97
IIIA-IV
6.6
86
78
78
16/60 (17)
80
69
Centre Hospitalier et Universitaire [76]
158
T2-T3
5.6
38
48.7
N/A
6/77
73.2a
-
Necker Hospital [77]
137 94
IIA-B IIIA-B"
3-15
62
94
85
337196 (17)
IIA KB IMA 1MB
95 80 60 58
-
Hopital Pitie-Salpetriere [78]
135
T3-T4b
8.5
96
90
72
26d/135 (19)d
All T3 T4
64 66 56
50 52 47
7V85
47
Hopital Tenon, Hopital F'itie-Salpetriere[75]
(8)
Memorial Medical Center of Long Beach [79]
78
III
N/A
48-84
90
65
(8) 'Disease-free survival. "Including inflammatory limited to part of the breast. C 27 isolated with metastasis, 23 breast, 4 axillary. d 11 with DCIS. e 5 breast failures, 2 axillary. BCT, breast conserving treatment. N/A Not available.
276 Low dose-rate brachytherapy for breast cancer
Figure 19.2 Isodose distributions in (a) transverse, (b) coronal, and (c) sagittal planes for a rigid implant demonstrating staggered interplane needle arrangement.
order to reduce the high-dose region at the center of the implant [85]. Interstitial breast implants are generally placed with the patient under general anesthesia. In some cases, conscious sedation combined with local infiltration is adequate. The patient is positioned to match what was utilized for her preplanning, or otherwise is generally positioned so that the ipsilateral arm is abducted to approximately 90°. The entire breast, with a generous margin, should be sterilely prepped and draped. In general, 15-20 cm hollow, stainless-steel needles are used. The deep plane is generally placed first. The orientation of the needles depends on the location of the tumor bed within the breast. Care is taken that the superficial plane is sufficiently deep to the skin (0.5-1 cm) to avoid exces-
sive skin dose. Once all the needles are in position, they can be replaced with flexible plastic catheters. Flexible catheters have the benefit of improving patient comfort. The major disadvantage is the potential for change in the implant geometry and dosimetry with patient movement [86]. To maintain ideal geometry, some have advocated that the needles or catheters be fixed in position by either endplates [82] or a bridge [87] (Figure 19.2). Once the implant is completed, a small amount of antibiotic ointment can be placed around the needle puncture sites and the implant covered with loose-fitting gauze until loading. After the patient has recuperated sufficiently from her anesthesia, post-implant treatment planning proceeds. In general, prophylactic antibiotics are not required unless the patient has a comorbidity such as
Treatment planning 277
cardiac valve disease. Usually, the patient can be comfortably maintained with mild analgesics throughout her treatment. Patients are allowed bathroom privileges, but activity should be restricted so that the implant geometry is maintained. It has been shown that the implant dosimetry can change with the patient position [86]. Once the radiation is completed, removal of the implant is done at the bedside, using a sterile technique.
19.6
TREATMENT PLANNING
Typically, an LDR interstitial implant boost can be performed 1-2 weeks before or after initiation of WBRT. Significant time delays between WBRT and the implant should be avoided, as this was correlated with a greater risk of local breast recurrence by Dubray et al. in the Creteil experience. In that analysis, the mean time delay between the WBRT and the LDR implant was prolonged at 5.9 ± 1.7 weeks [88]. Because most patients in this series were treated by exclusive radiation without breast surgery, it is not clear whether this finding is applicable to WBRT plus implant following excision as well. For patients receiving an LDR implant as the sole radiation modality following lumpectomy, implants should probably be performed within 4-6 weeks of the final breast surgery. Performing the implant at the time of excision or reexcision has the important advantage of allowing direct visualization of the surgical bed, ensuring that it is covered by the needle geometry. Furthermore, implants done at the time of the re-excision or the axillary node dissection have the added benefit of avoiding a separate surgical procedure and anesthesia. The disadvantage of implanting at the time of the breast surgery is that the presence of any adverse pathologic feature and status of the final resection margins are unknown. For this reason, delayed loading of the implant for 48 h after periexcision placement is recommended, to allow for wound healing and availability of the pathologic review and margin assessment [89]. Others have loaded the implant within 6 h after peri-excision placement and reported no greater occurrence of wound complications and have adjusted total dose for any adverse pathologic features found at the time of pathology review [56,90]. Another disadvantage of performing the implant at the time of the excision or re-excision is that it is limited to those patients who are treated at centers with both surgical and radiation oncology capabilities on site. At many large referral centers, this would preclude an interstitial implant for the numerous patients who have had their breast surgery performed elsewhere. Patients who do not have the interstitial implant performed at the time of breast surgery require careful localization of the excision cavity. This includes all implants done at the time of the axillary node dissection
before WBRT, as well as those performed after WBRT. De Biose et al. found that clinical examination (physical examination, operative report, and mammograms) underestimated the full extent of the lumpectomy cavity 87% of the time in comparison to preoperative and intraoperative ultrasound definition [91]. Sedlmayer et al. demonstrated a potential error rate of 51% from estimating the location of the tumor bed based upon the clinical criteria of preoperative mammography, surgical reports, and palpation of postoperative indurative changes in comparison to radiographs of surgical clips outlining the excision cavity [92]. This potential error rate rose to 78.5% for large pliable breasts. As a result, this group advocates that the excision cavity be marked with surgical clips. Demarcating the lumpectomy site with surgical clips has also been shown to be equally important when planning WBRT and electron boost volumes to ensure adequate coverage [93,94]. For this reason, it is our belief that the lumpectomy cavity should always be marked with surgical clips to avoid geographic misses by any radiation modality, not just in the setting of an implant. When the lumpectomy cavity has not been demarcated with surgical clips, localization with either computerized tomography (CT) scan and/or ultrasound is necessary. Both of these modalities are dependent on the presence of a seroma and/or hematoma to outline the cavity. Therefore, it is necessary to perform these studies for localization purposes as close to the breast surgery as possible, as postoperative changes can resolve within a few weeks. Prior to going to the operating room, the treatment volume to be implanted should be determined and mapped on the skin of the breast. Also, the depth of the deepest plane necessary to adequately cover the excision cavity should be planned preoperatively as this is frequently also underestimated by clinical examination [91]. One proposed method to ensure accurate tumor localization and intraoperative reproduction of a preplanned implant is to make a mould of the breast that will provide fixed geometry for the implant. Teo and Chung first reported using a cobex cast of the breast to aid in the planning of LDR implants that were a component of radical radiation for locally advanced breast cancer [95]. Once the cobex cast of the breast was made, a CT scan was performed for tumor localization with the cast in place. A geometrically ideal implant for tumor coverage was planned from the CT scan, and the entry and exit points for the needle placement drilled into the cast preoperatively. Intraoperatively, the cast was refitted to the patient's breast and the implant constructed via needle placement through the predrilled holes. This technique was modified by Perera et al. for use together with surgical clips marking the lumpectomy cavity. The surgical clips were added because of difficulty separating the tumor bed from normal dense breast tissue with the use of CT alone [96]. Their technique used the simulator first
278 Low dose-rate brachytherapy for breast cancer
to localize the tumor bed based on the position of the clips within the breast. This area was then mapped out on a breast mould and marked by radio-opaque catheters attached to the mould. Similarly, after CT localization, the optimal implant for coverage of the treatment volume was planned and the exit and entry points for the implant needles placed on the mould in relation to the radioopaque markers. Reconstruction of the implant was then done intraoperatively after the mould had been refitted to the patient's breast. Vicini et al. have used a similar technique to preplan implants using ultrasound to initially map the lumpectomy cavity prior to a CT done with radio-opaque markers placed directly on the breast instead of a mould or cast. The positions of the radioopaque markers were subsequently drawn directly on the breast. Using a three-dimensional treatment planning system, they were able to create a volumetric model of their fixed template system and virtually plan the ideal position of the implant on the CT. Digitally reconstructed radiographs of the skin surface were generated demonstrating the needle exit and entry points relative to the radio-opaque markers to guide intraoperative placement [97]. Sedlmayer reported a technique for preplanning implants based solely on simulator localization of the clips marking the excision cavity to determine the depth and trajectory of the implant planes [92]. This information is then mapped directly on the affected breast with the patient in the same position to be used at the time of surgical placement. At The Medical College of Wisconsin, we have adopted a similar technique, preoperatively mapping the position of clips onto the skin of the breast and estimating the depth of the planes in the simulator with the patient in a position that will mimic what will be used intraoperatively. In addition, we routinely use a C-arm fluoroscopic unit intraoperatively to confirm inclusion of the clips marking the excision cavity within the needle geometry after the first few needles have been placed prior to completing the entire implant. Multiple dosimetry systems for preplanning interstitial implants have been formulated. These are meant to guide seed activity and spacing, as well as intraplanar and interplanar needle spacing, so as to deliver a certain dose rate over a given volume. These include a nomogram [84], which specifies ribbon spacing based upon the linear activity of the sources and target volume, and planning tables [85], which determine the seed activity and interplanar separation needed to achieve a desired dose rate at the boundaries of the treatment volume of dimensions specified in the table. These preplanning tables and monographs have been particularly useful for implants performed at the time of lumpectomy. They should not replace post-implant treatment planning as the actual geometry of the implant performed at the time of surgery can vary significantly from what was preplanned. Post-implant treatment planning should consist of a minimum of two variable angle films, typically 60-90°
apart. Ribbons containing radio-opaque dummy seeds are loaded into all the implanted needles or catheters and, by their different appearances, allow identification of the separate needles. The positions of the dummy seeds and the surgical clips demarcating the cavity on the orthogonal films are digitized into the planning system, allowing reconstruction of the implant. The radiation oncologist should specify the volume to be covered by the reference dose rate on the treatment planning films. Evaluation of isodose curves in the sagittal and coronal as well as in the transaxial planes gives a better understanding of the volume irradiated and the dose heterogeneity (Figure 19.2a-c). Ideally, isodose curves should be generated in more than just the central plane. It is particularly important to evaluate other planes if a nonrigid implant was performed, as central plane dosimetry will not reliably represent any convergence or divergence of the needles that may have occurred. CT scan-based dosimetry is very useful for planning and documenting tumor volume coverage of multiplane implants [98]. It reduces the time necessary for reconstruction from the orthogonal radiographs, reliably demonstrates areas of uneven spacing, and can evaluate the isodose coverage of the lumpectomy site. Dose prescription includes selection of the dose rate and total dose to be delivered to a specified volume. Doses for boost therapy after lumpectomy and WBRT range from 10 Gy to 25 Gy, with dose rates typically varying from 0.4 Gy to 0.8 Gy h~' (see Table 19.1). Doses for boost therapy after WBRT in cases of exclusive radiation therapy for locally advanced disease tend to range from 20 Gy to 40 Gy given over similar dose rates (see Table 19.4). For the Paris System, the reference dose rate is 85% of the basal dose, or the average minimum dose between the sources [31]. A dose-rate effect for local control was reported by Mazeron et al. in a review of 398 T1-T3 adenocarcinomas of the breast treated by exclusive radiation therapy without surgery. An LDR interstitial implant of 37 Gy was done after WBRT of 45 Gy. Significantly higher rates of breast relapse were noted when the implant dose-rate was < 0.6 Gy rr1 [99]. Deore et al. demonstrated a similar effect for LDR implant boosts after lumpectomy and WBRT [55]. In this series of 273 patients, an increased rate of local recurrence was seen in a small number of cases in which the dose rate of the implant boost was 0.2-0.29 Gy (4/17, 24%) versus >0.3 Gy h-1 (16/253, 6.3%). However, little information is available in this report regarding the presence of other adverse prognostic features in this group of 17 patients. Overall, dose rates > 0.4 Gy h"1 have yielded very high local control rates after lumpectomy (see Table 19.2). Special attention should be paid to the skin dose during the implant treatment planning. Excessively high skin doses may result in telangiectasis and skin retraction as a late outcome. For this reason, the most superficial sources should be at least 5-10 mm beneath the skin surface. Implants with large interplane separations
References 279
may need up to 10 mm to avoid excessive skin dosage. Van Limbergen studied the location of the dermal vascular structures of 30 healthy skin samples from ten mastectomy specimens and found that the dermal vascular structures responsible for telangiectasia are located in the first 5mm under the skin surface [100]. Thermoluminescence dosimetry (TLD) measurements made over several points on the skin surface of the breast during the implant and WBRT suggested that doses > 50 Gy to the skin are associated with a higher probability of late telangiectasia [86]. At The Medical College of Wisconsin, two to three BBs are placed over the skin at the time of post-treatment planning in the simulator. These are then digitized into the planning system along with the dummy seeds and the surgical clips. This provides us with an estimate of skin dose. While it is necessary that the desired dose accurately covers the treatment volume to secure local control, it should be kept in mind that the parameters of total dose and treatment volume can influence the cosmetic and toxicity outcomes. The irradiated boost volume and total dose significantly influence the formation of late fibrosis. A very carefully executed study evaluating breast fibrosis was performed in 404 patients with early-stage breast cancer who underwent BCT [23]. This study demonstrated that the odds of developing moderate to severe fibrosis were dependent on overall dose, particularly for doses > 70 Gy. Furthermore, for any given dose, there was a proportional increase in late fibrosis with increasing irradiated volumes, (see Figure 19.3). Specifically, for each 107 cm3 added to the irradiated volume, the risk of fibrosis increased by a factor of 4. Similarly, McRae et al. correlated soft-tissue complications with the implant volume in 56 breast implants [ 101 ]. In contradiction to these findings, others have not found a correlation between implant volume and cosmetic or toxicity outcome [59].
Implicit in the prescription of dose with LDR interstitial implants are those volumes of tissue that will receive markedly higher doses due to dose heterogeneity. Several investigators have proposed methods for quantifying dose uniformity. These include the dose homogeneity index (DHI), dose non-uniformity ratio (DNR), and double dose volume ratio (DDVR) [81,102,103]. The DHI represents the percentage of the target volume receiving dose rates between 100% and 150% of the reference dose rate [81]. The DNR is defined as the ratio of the high-dose volume relative to the reference volume [ 102]. Lastly, the DDVR is the ratio of the volume of tissue receiving twice the prescribed dose divided by the volume of tissue receiving the prescribed dose [103]. Acceptably homogeneous implants are recommended to have a DHI of > 0.8, a DDVR of < 0.1, and the minimum DNR. The value of these indices is that they allow comparison of implants in terms of dose actually delivered versus prescribed dose. Ideally, the 150% isodose line should be broken up and visible around individual ribbons only. The importance of the interaction between total treatment volume and dose heterogeneity for determining cosmetic and toxicity outcomes further demonstrates the need for careful treatment planning.
19,7 CONCLUSION Low dose-rate brachytherapy for the treatment of breast cancer has a well-established track record for providing high levels of local control and good cosmetic outcomes. Careful quality control with special attention to target localization and dose specification has proved crucial in the success of LDR implants. While its role as a boost after WBRT is diminishing in cases of BCT, newer applications are emerging. In particular, brachytherapy as the sole radiation modality following lumpectomy appears promising. Even as HDR brachytherapy becomes more prevalent for the treatment of breast cancer, the wealth of LDR experience and data provides an important benchmark for success in terms of local control and late outcomes.
REFERENCES 1. NIH Consensus Development Conference (1991) Treatment of early-stage breast cancer./4M/4,265(3), 391-4. 2. Vaeth, J. (1983) Historical aspects of tylectomy and radiation therapy with the treatment of cancer of the breast. In 37 Front Radiat. Ther. One., Karger Basel, Figure 19.3 Relationship of dose and implant volume to probability of breast fibrosis. (Adapted with permission from Borger et al. [23].)
1-10. 3. Ewing, J. (1934) Early experiences in radiation therapy, Janeway Memorial Lecture. Am. J. Roentgenol., 31, 153-63.
280 Low dose-rate brachytherapy for breast cancer 4. Keynes, G. and Cantab, F. (1929) The treatment of primary carcinoma of the breast with radium. Acta Radiol., 1 0,393-402. 5. Keynes, G. and Camb, F. (1937) Conservative treatment of cancer of the breast, fir. Med.)., 2,643-7. 6. Pfahler, G. (1932) Results of radiation therapy in 1,022 private cases of carcinoma of the breast from 1902-1928. Am.J. Roentgen. Radial Ther., 27(4), 497-508. 7. Baclesse, F. (1949) Roentgen therapy as the sole method of treatment of cancer of the breast. Am. J. Roentgen. Radial 7/M?r.,62(3),311-19. 8. Pendergrass, E. and Modes, P. (1938) Further observations on carcinoma of the breast. Am. J. Roentgenol.,42(3), 393-403. 9. Porritt, A. (1964) Early carcinoma of the breast. Br.j. Swig., 51 (3), 214-16. 10. Guttmann, R. (1962) Survival and results after 2-million volt irradiation in the treatment of primary operable carcinoma of the breast with proved positive internal mammary and/or highest axillary nodes. Cancer, 15, 383-6. 11. Rissanen, P. (1969) A comparison of conservative and radical surgery combined with radiotherapy in the treatment of stage I carcinoma of the breast. Br. J. Radiol., 42,423-6. 12. Hayward, J. (1974) The conservative treatment of early breast cancer. Cancer, 33, 593-9. 13. Pierquin, B., Owen, R., Maylin, C. etal. (1980) Radical radiation therapy of breast cancer. Int. J. Radial On col. Biol. Phys., 6,17-24. 14. Hellman, S., Harris, J., Levene, M. etal. (1980) Radiation therapy of early carcinoma of the breast without mastectomy. Cancer, 46,988-94. 15. Arriagada, R., Le, M., Rochard, F. etal. (1996) Conservative treatment versus mastectomy in early breast cancer: patterns of failure with 15 years of followup data,/ Clin. Oncol., 14,1558-64. 16. Blichert-Toft, M. and Rose, C. (1992) Danish randomized trial comparing breast conservation therapy with mastectomy: six years of life-table analysis. 7. Natl. Cancer /nsf., 11,19-25. 17. Fisher, B. and Anderson, S. (1995) Reanalysis and results after 12 years of follow-up in a randomized clinical trial com pa ring total mastectomy with lumpectomy with or without irradiation in the treatment of breast cancer. N. Engl.J. Med., 333,1456-61. 18. Jacobson, J., Danforth, D., Cowan, K. etal. (1995)Tenyear results of a comparison of conservation with mastectomy in the treatment of stage I and II breast cancer. N. Engl.J. Med., 332,907-11. 19. van DongenJ., Bartelink, H., Fentiman, \.etal. (1992) Factors influencing local relapse and survival and results of salvage treatment after breast conservation compared with mastectomy in TNM stage I and II breast cancer. Eur. J. Cancer, 2Sk,W\-5. 20. Veronesi, U., Salvador!, B., Greco, L etal. (1995) Breast conservation is a safe method in patients with small
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
cancer of the breast. Long-term results of three randomized trials on 1,973 patients. Eur.J. Cancer, 31 A, 1574-9. Borger, J., Kemperman, H., Hart, A. etal. (1994) Risk factors in breast-conservation therapy./ Clin. Oncol., 12(4), 653-60. Bartelink, H., Borger, J. etal. (1988) The impact of tumor size and histology on local control after breastconserving therapy. Radiother. Oncol., 11,297-303. Borger, J., Kemperman, H., Smith, \\.etai (1994) Dose and volume effects on fibrosis after breast conservation therapy. Int.}. Radial Oncol. Biol. Phys., 30(5), 1073-81. Kemperman, H., Borger, J., Hart, A. etal. (1995) Prognostic factors for survival after breast conserving therapy for stage land II breast cancer. The role of local recurrence. Eur.J. Cancer, 31A(5), 690-8. Mansfield, C., Komarnicky, L, Schwartz, G. etal. (1995) Ten-year results in 1070 patients with stages I and II breast cancer treated by conservative surgery and radiation therapy. Cancer, 75,2328-36. Pierquin, B., Huart, J., Raynal, M. etal. (1991) Conservative treatment for breast cancer: long-term (15 years). Radiother. Oncol., 20,16-23. BoyagesJ., Recht, A., Connolly, \.etal. (1990) Early breast cancer: predictors of breast recurrence for patients treated with conservative surgery and radiation therapy. Radiother. Oncol., 19,29-41. Olivotto, I., Rose, M., Osteen, R. etal. (1989) Late cosmetic outcome after conservative surgery and radiotherapy: analysis of causes of cosmetic failure. Int. J. Radial Oncol. Biol. Phys., 17,747-53. BoyagesJ., Bosch, C., Langlands, k.eia\. (1992) Breast conservation: long-term Australian data. IJROBP, 24, 253-60. Vicini, FA, Horwitz, E.M., Lacerna, M.D. etal. (1997) Long-term outcome with interstitial brachytherapy in the management of patients with early-stage breast cancer treated with breast-conserving therapy. Int.J. Radial Oncol. Biol. Phys., 37(4), 845-52. Pierquin, B., Dutreix, A., Paine, C.H.etal. (1978) The Paris System in interstitial radiation therapy. Acta Radiol. Oncol., 17(1), 33-48. Solin, L.J., Fowble, B.L., Martz, K.Letal. (1991) Results of the 1983 Patterns of Care Process Survey For Definitive Breast Irradiation. InlJ. Radial Oncol. Biol. P/?ys.,20(1), 105-11. Kutcher, G., Smith A., Fowble, B.etal. (1996) Treatment planning for primary breast cancer: a Pattern of Care Study. InlJ. Radial Oncol. Biol. Phys., 36(3), 731-7. Holland, R., Veling,S., Mravunac, M.etal. (1985) Histologic multifocality of TIS, T1 -2 breast carcinomasimplications for clinical trials of breast-conserving surgery. Cancer, 56,979-90. Smitt, M., Nowels, K., Zdeblick, M. etal. (1995) The importance of the lumpectomy surgical margin status in long term results of breast conservation. Cancer, 76, 259-67.
References 281 36. Pezner, R., Wagman, L, Ben-Ezra, J. etal. (1994) Breast conservation therapy: local tumor control in patients with pathologically clear margins who receive 5000 cGy breast irradiation without local boost. Breast Cancer Res. Treat, 32,261 -7. 37. Burke, M., Allison, R.,Tripcony, L. etal. (1995) Conservative therapy of breast cancer in Queensland. Int.). Radial Oncol. Biol. Phys., 31(2), 295-303. 38. Ryoo, M., Kagan, A., Wollin, M.etal. (1989) Prognostic factors in recurrence and cosmesis in 393 patients after radiation therapy for early mammary carcinoma. Radiology, 172,555-9. 39. Spivack, B., Khanna, M., Tafra, L. et at. (1994) Margin status and local recurrence after breast-conserving surgery. Arch. Surg., 129,952-7. 40. Anscher, M., Jones, P., Prosnitz, L. etal. (1993) Local failure and margin status in early-stage breast carcinoma treated with conservation surgery and radiation therapy. Ann. Surg., 218,22-8. 41. Clarke, D., Le, M., Sarrazin, D.etal. (1985) Analysis of local-regional relapses in patients with early breast cancers treated by excision and radiotherapy: experience of the Institut Gustave-Roussy. Int. J. Radiat. Oncol. Biol. Phys., 11,137-45. 42. Solin, L, Fowble, B., Schultz, D. etal. (1991) The significance of the pathology margins of the tumor excision on the outcome of patients treated with definitive irradiation for early stage breast cancer. Int.J. Radiat. Oncol. Biol. Phys., 21,279-87. 43. Romestaing, P., Lehingue.Y., Carrie, C.etal. (1997) Role of a 10-Gy boost in the conservative treatment of early breast cancer: results of a randomized clinical trial in Lyon, France./ Clin. Oncol., 15(3), 963-8. 44. Teissier, E., Hery, M., Ramaioli, A. etal. (1998) Boost in conservative treatment: 6 years results of randomized trial. Breast Cancer Symposium Scientific Poster Presentations, 21 st Annual - San Antonio, TX, December 12-15. 45. Ptaszynski, A., Van den Bogaert, W., Van Glabbeke, M. et al. (1994) Patient population analysis EORTC trial 22881/10882 on the role of a booster dose in breastconserving therapy. Eur. J. Cancer, 30A(14), 2073-81. 46. Bartelink, H., Collette, L, Fourquet, A. etal. (2000). Impact of a boost dose of 16 Gy on the local control and cosmesis in patients with early breast cancer: the EORTC 'boost versus no boost' trial. Proceedings of the American Society for Therapeutic Oncology 42nd Annual Meeting, 48(3), Suppl., 111. 47. Clark, R., Whelan, T., Roberts, L. et al. (1996) Randomized clinical trial of breast irradiation following lumpectomy and axillary dissection for node-negative breast cancer: an update./. Natl. Cancer Inst., 88(22), 1659-64. 48. Magee, B., Swindell, R., Harris, M. etal. (1996) Prognostic factors for breast recurrence after conservative breast surgery and radiotherapy: results from a randomized trial. Radiother. Oncol., 39,223-7. 49. Fisher, E., Anderson, S., Redmond, C. etal. (1993)
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
Pathologic findings from the National Surgical Adjuvant Breast Project Protocol B-06. Cancer, 71,2507-14. Ray, G. and Fish V. (1983) Biopsy and definitive radiation therapy in stages I and II adenocarcinoma of the female breast: analysis of cosmesis and the role of electron beam supplementation. Int.]. Radiat. Oncol. Biol. Phys., 9,813-18. Touboul, E., Belkacemi, Y., Lefranc, \.etal. (1995). Early breast cancer: influence of type of boost (electron vs. iridium-192 implant) on local control and cosmesis after conservative surgery and radiation therapy. Radiother. Oncol., 34,105-13. Perez, C., Taylor, M. etal. (1996) Brachytherapy or electron beam boost in conservation therapy of carcinoma of the breast: a nonrandomized comparison. IntJ. Radiat. Oncol. Biol. Phys., 34(5), 995-1007. Vicini, F., White, J., Gustafson, G. etal. (1993) The use of iodine-125 seeds as a substitute for iridium-192 seeds in temporary interstitial breast implants. Int.J. Radiat. Oncol. Biol. Phys., 27, 561-6. Nobler, M. and Venet, L (1985) Prognostic factors in patients undergoing curative irradiation for breast cancer. Int.J. Radiat. Oncol. Biol. Phys., 11,1323-31. Deore, S. and Sarin, R. (1993) Influence of dose-rate and dose per fraction on clinical outcome of breast cancer treated by external beam irradiation plus iridium-192 implants: analysis of 289 cases. Int.J. Radiat. Oncol. Biol. Phys., 26,601 -6. Wazer, D., Kramer, B., Schmid, C. etal. (1997) Factors determining outcome in patients treated with interstitial implantation as a radiation boost for breast conservation therapy. Int.J. Radiat. Oncol. Biol. Phys., 39(2), 381-93. Rochefordiere, A., Abner, A., Silver, B. etal. (1992) Are cosmetic results following conservative surgery and radiation therapy for early breast cancer dependent on technique? Int. J. Radiat. Oncol. Biol. Phys., 23,925-31. Sarin, R., Dinshaw, K., Shrivastava, S. etal. (1993) Therapeutic factors influencing the cosmetic outcome and late complications in the conservative management of early breast cancer. Int.J. Radiat. Oncol. Biol. Phys., 27,285-92. Wazer, D., DiPetrillo,T., Schmidt-Ullrich, R.etal. (1992) Factors influencing cosmetic outcome and complication risk after conservative surgery and radiotherapy for early-stage breast carcinoma./ Clin. Oncol., 10(3), 356-63. Danoff, B. and Griffin, T. (1983) Phase III protocol for adenocarcinoma of the breast Stage T1, T2, NO, MO comparing boost with electrons versus implant. RTOG Protocol 83-06. Fourquet, A., Campana, F., Mosseri, V. effl/. (1995) Iridium-192 versus cobalt-60 boost in 3-7 cm breast cancer treated by irradiation alone: final results of a randomized trial. Radiother. Oncol., 34,114-20. DenhamJ., Hamilton, C., Cross, P. etal. (1991) Breast conservation, the problem of treating the excision site
282 Low dose-rate brachytherapy for breast cancer
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
effectively: physical criteria for the choice of technique used. Clin. Oncol., 3,250-6. Krishnan, E., Krishnan, L, Cytaki, E. etal. (1990) Radiobiological advantages of an immediate interstitial boost dose in conservative treatment of breast cancer. Int.J. Radial. Oncol. Biol. Phys., 18,419-24. Krishnan, L, Jewell, W., Krishnan, E.etal. (1992) Breast cancer with extensive intraductal component: treatment with immediate interstitial boost irradiation. Radiology, 183,273-6. Ribeiro,G., Magee, B., Swindell, R.etal. (1993) The Christie Hospital Breast Conservation Trial: an update at 8 years from inception. Clin. Oncol., 5,278-83. Fentiman, I., Poole,C, Winter, P.etal. (1996) Inadequacy of iridium implant as sole radiation treatment for operable breast cancer. Eur.J. Cancer, 32A(4), 608-11. Kuske, R.( Bolton, J., Wilenzick, R. etal. (1994) Brachytherapy as the sole method of brest irradiation in TIS, T1, T2, NO-1 breast cancer. Int. J. Radial Oncol. Biol. P/7ys.,30(Suppl.1),245. Vicini, F., Vijay, R.,Chen, P.etal. (1999) Irradiation of the tumor bed alone after lumpectomy in selected patients with early-stage breast cancer treated with breast conserving therapy./ Surg. Oncol., 70,33-40. Perera, F., Engel, J., Holliday, R. etal. (1997) Local resection and brachytherapy confined to the lumpectomy site for early breast cancer: a pilot study. J.Surg. Oncol., 65,263-8. Clark, R., McCulloch, P., Levine, M.etal. (1992) Randomized clinical trial to assess the effectiveness of breast irradiation following lumpectomy and axillary dissection for node-negative breast cancer./ Natl. Cancer Inst., 84(9), 683-9. Liljegren, G., Holmberg, L, Westman, A. etal. (1994) Sector resection with or without postoperative radiotherapy for stage I breast cancer: five-year results of a randomized trial.). Natl. Cancer Inst, 86(9), 717-22. Fisher, ER., Sass, R., Fisher, B. etal. (1986) Pathologic findings from the National Surgical Adjuvant Breast Project (Protocol 6). Cancer, 57,1717-24. Maulard, C., Housset, M., Brunei, P. etal. (1995) Use of perioperative or split-course interstitial brachytherapy techniques for salvage irradiation of isolated local recurrences after conservative management of breast cancer. Am.J. Clin. Oncol., 18(4), 348-52. Kurtz, J., Jacquemier, J. and Amalric, R. (1991) Is breast conservation after local recurrence feasible? Eur.J. Cancer, 27(3), 240-4. Touboul, E., Buffat, L, LefrancJ.ef al. (1996) Possibility of conservative local treatment after combined chemotherapy and preoperative irradiation for locally advanced noninflammatory breast cancer. IJROBP, 34(5), 1019-28. Calais, G., Berger, C., Descamps, P. etal. (1994) Conservative treatment feasibility with induction chemotherapy, surgery, and radiotherapy for patients
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
with breast carcinoma larger than 3 cm. Cancer, 74(4), 1283-8. Jacquillat, C., Weil, M., Baillet, F. etal. (1990) Results of neoadjuvant chemotherapy and radiation therapy in the breast-conserving treatment of 250 patients with all stages of infiltrative breast cancer. Cancer, 66, 119-29. Baillet, F., Rozec, C. etal. (1985) Treatment of locally advanced breast cancer without mastectomy-5 and 10 year results of 135 tumors larger than 5 centimeters treated by external beam therapy, brachytherapy, and neoadjuvant chemotherapy. Ann. N.Y. Acad. Sci., 264-70. Syed, A., Puthawala, A. etal. (1984) Primary irradiation in the management of early and locally advanced carcinoma of the breast. Br.J. Radiol., 57,317-21. Bonadonna,G.,Valagussa, P., Moliterni, A. etal. (1995) Adjuvant cyclophosphamide, methotrexate, and fluorouracil in node-positive breast cancer. N. Engl.J. Med.,332,901-6. Wu,A., Ulin, K.,Sternick, E.ef al. (1988) A dose homogeneity index for evaluating 192lr interstitial breast implants. Med. Phys., 15(1), 104-7. Pierquin, B., Wilson, J.F. and Chassagne, D. (1987) Female breast-clinical review. In Modern Brachytherapy. Paris, Masson, 263-71. Gillin, M.T., Kline, R.W., Wilson, J.F. etal. (1984) Single and double plane implants: a comparison of the Manchester System with the Paris System. IJROBP, 10, 921-5. Anderson, LL, Hilaris, B.S. and Wagner, LK. (1985) A nomogram for planar implant planning. Endocuriether. Hypertherm. Oncol., 1,9-15. Zwicker, R., Schmidt-Ullrich, R., Schiuller, B.etal. (1985) Planning of lr-192 seed implants for boost irradiation to the breast. IntJ. Radiat. Oncol. Biol. Phys., 11, 2163-70. Habibollahi, F., Phil, Mayles, H.M.O. etal. (1988) Assessment of skin dose and its relation to cosmesis in the conservative treatment of early breast cancer. Int.J. Radiat. Oncol. Biol. Phys., 14,291-6. Johnson, D.W. (1984) An adjustable breast bridge for use in multiple-plane interstitial breast implants. Br.J. Radiol., 57,159-61. Dubray, B., Mazeron, J., Simon, J. etal. (1992) Time factors in breast carcinoma: influence of delay between external irradiation and brachytherapy. Radiother. Oncol, 25,267-72. McCormick, B., Wesson, M., Cox, L. et al. (1988) Iridium192 implants for primary breast cancer: experience with placement at the time of wide local excision. Int. J. Radiat. Oncol. Biol. Phrys., 15, 745-8. Mansfield, C.M. (1990) Intraoperative lr-192 implantation for early breast cancer- techniques and results. Cancer, 66(1), 1-5. De Biose, D., Horwitz, E., Martinez, A. etal. (1997) The use of ultrasonography in the localization of the
References 283
92.
93.
94.
95.
96.
97.
lumpectomy cavity for interstitial brachytherapy of the breast. Int.J. Radial Oncol. Biol. Phys., 38(4), 755-9. Sedlmayer, F., Rahim, H., Kogelnik, D. etal. (1996) Quality assurance in breast cancer brachytherapy: geographic miss in the interstitial boost treatment of the tumor bed. Int.]. Radial. Oncol. Biol. Phys., 34(5), 1133-9. BedwinekJ. (1993) Breast conserving surgery and irradiation: the importance of demarcating the excision cavity with surgical clips. Int.J. Radial Oncol. Biol. Phys., 26,675-9. Solin, L, Danoff, B.etal. (1985) A practical technique for the localizing of the tumor volume in definitive irradiation of the breast. Int.J. Radial Oncol. Biol. Phys., 11,1215-20. Teo, P. and Chung, D. (1987) An accurate method of 192 Ir write implantation for locally advanced carcinoma of the breast. Acta Oncol, 26,273-6. Perera, F., Chisela, F., Engel, J. etal. (1995) Method of localization and implantation of the lumpectomy site for high dose rate brachytherapy after conservative surgery forT1 and T2 breast cancer. Int.J. Radial Oncol. Biol. Phys ,31(4),959-65. Vicini, F., Jaffray, D., Horwitz, E. etal. (1998) Implementation of 3D-virtual brachytherapy in the management of breast cancer: a description of a new
98.
99.
100.
101.
102.
103.
method of interstitial brachytherapy. Int.J. Radial Oncol. Biol. Phys., 40(3), 629-35. Regional Medical Physics Department (1986) The use of simulator-based computed tomography in iridium-192 dosimetry. Br.J. Radiol., 59,1035-6. MazeronJ.J., Simon, J. and Crook, J. (1991) Influence of dose rate on local control of breast carcinoma treated by external beam irradiation plus iridium 192 implant. Int. J. Radial Oncol. Biol. Phys., 21,1173-7. Van Limbergen, E., Briot, E. and Drijkoningen, M. (1990) The source - skin distance measuring bridge: a method to avoid radiation telangiectasia in the skin after interstitial therapy for breast cancer. Int.J. Radial Oncol. Biol. Phys., 18,1239-44. McRae, D., RodgersJ.and Dritschilo, A. (1987) Dosevolume and complication in interstitial implants for breast carcinoma. IntJ. Radial Oncol. Biol. Phys., 13, 525-9. Saw, C.B., Suntharalingam, N. and Wu, A. (1993) Concept of dose nonuniformity in interstitial brachytherapy. Int. J. Radial Oncol. Biol. Phys., 26, 519-27. Metcalf, D.R., Lewinsky, B.S., Fingerhut, A.G. etal. (1988) A ratio of volumes technique for analyzing interstitial implants. Int.J. Radial Oncol. Biol. Phys., 15 (Suppl. 1), 143.
20 Brachytherapy in the treatment of head and neck cancer A.GERBAULETANDM.MAHER
20.1
HISTORICAL BACKGROUND
Since the discovery of radium in 1898 by Pierre and Marie Curie, there have been many advances in the use of radioactive sources for brachytherapy procedures. The first reported cases of interstitial therapy were published in 1914 from Dublin, Ireland [1-3]. Subsequent development of radium needles allowed easy implantation of skin tumors, lip tumors, and cancers of other accessible sites. Advances in dosimetry saw the development of units such as threshold erythema dose, millicurie destroyed and milligram-hours. In the 1930s Patterson and Parker devised a series of rules for single-plane, double-plane, and volume implants governing the use of radium needles. Milligram-hours were at that time the unit of measurement in common use for such applications [4,5]. In head and neck cancers the most common sites to be treated by such implants were lip, tongue, floor of mouth, and buccal mucosa. Radium needles had the disadvantage of significant radioprotective problems. Artificial radioactivity, however, was discovered by Irene Curie and Frederic Joliot in 1933 and shortly thereafter new materials suitable for clinical use became available. The introduction of iridium-192 wire brought with it greater flexibility and a major reduction in the hazards posed to both staff and patients by radium brachytherapy. From the 1960s onwards, the Paris school led by
Pierquin, Dutreix, and Chassagne [5-7] laid the foundations of the Paris system of implantation. Enormous clinical experience coupled with scientific endeavor resulted in what became known as the Paris System. This refers to a system of implantation and dosimetry which is predictive, reliable, consistent, and clinically safe [8,9]. With the development of linear accelerators in the 1950s and 1960s, the popularity of brachytherapy as a treatment modality began to decline. However, brachytherapy is currently undergoing a resurgence in use due to the widespread availability of isotopes which offer more flexibility and few radioprotective problems and the introduction of remote afterloading machines into many treatment departments [6,9].
20.2 PRETREATMENT ASSESSMENT AND INVESTIGATIONS
20.2.1
Primary tumor
Pretreatment assessment is without doubt the single most important step in any planned brachytherapy procedure. In order to be able to decide whether a brachytherapy procedure is appropriate, one must be able to define with confidence the limits of the tumor volume and the limits of the proposed target volume [1,5,7,9].
Treatment methods 285
In most instances, brachytherapy is contraindicated as an initial therapeutic procedure for T3-T4 cancers of the head and neck. On the other hand, brachytherapy is, in the majority of clinical situations, perfectly appropriate as an initial or indeed exclusive local therapeutic procedure for T1-T2 cancers of the head and neck. The key, therefore, to the correct and effective use of brachytherapy in head and neck cancers lies in the pretreatment assessment and investigations. Pretreatment assessment includes detailed history, full physical examination, appropriate investigations, and multidisciplinary consultation. With regard to the primary tumor, this process refers to loco-regional tumor evaluation, routine search for metastatic disease, investigation for synchronous second primaries and, in the case of oral cavity implants, gingival/dental examination and recommendations [5,10]. Loco-regional tumor evaluation defines the exact limits of the primary tumor, allowing the brachytherapist precisely to appreciate the tumor volume. This may be an entirely clinical assessment for lesions on the lip, buccal mucosa, floor of mouth, or mobile tongue. In some situations, however, further delineation is provided by computerized axial tomography (CAT) scanning and/or magnetic resonance imaging (MRI) scans, which may more accurately assess the primary tumor volume and the integrity of adjacent structures such as bone and cartilage [8,11-13]. 20.2.2
Neck assessment
For patients with oral cavity tumors for whom the planned treatment is brachytherapy, assessment of nodal neck disease is primarily clinical. Even if clinically negative, all such patients treated in Gustave-Roussy systematically undergo elective neck dissection provided they are deemed fit for general anesthesia and surgery. In situations in which patients are deemed to be unfit for general anesthesia, the draining neck nodes may be assessed clinically and by serial CAT scanning. 20.23
Dental assessment
Gingival and dental assessment is of vital importance prior to any brachytherapy procedure which will result in radioactive sources lying in close contact with the teeth, gums, or mandible. Teeth with extensive caries and gross gingival disease are better extracted prior to brachytherapy if osteoradionecrosis is to be avoided. Following extraction, sufficient time to allow adequate gum healing must elapse (usually 2-3 weeks) before the brachytherapy treatment is carried out. For patients with healthy teeth and gums, efforts must be made to protect the dentition. In our institution, a shielding system made of 2 mm of lead is used to cover the teeth and gums. An identical denture shield made of
acrylic resin, which is radiotransparent, is put in place when the orthogonal X-ray films for dosimetry are taken following the brachytherapy procedure. This allows reproduction of the actual source positions during irradiation with the dental radioprotective lead shield in place. This shielding system confers a reduction in dose of 50% to the teeth and mandible.
203
203.1
TREATMENT METHODS
lridium-192
Iridium-192 is the radioisotope used in our institution for head and neck low dose-rate (LDR) brachytherapy. Some institutions continue to use cesium needles in certain clinical situations, such as lip and tongue cancer. Iridium-192, however, is flexible, pliable, and can be cut to prerequisite lengths. It is available as a continuous wire coil of iridium-platinum alloy (25-75) and can be purchased in diameters of 0.3 or 0.5 mm, depending on the technique for which it will be employed. Iridium has a half-life of 74 days, is a gamma ray emitter with average energy of 0.35 MeV, and has a half value layer of 2 mm lead [4-7].
203.2
Iridium-192 hairpins
Iridium-192 is also available in the form of single pins or hairpins. The hairpin consists of two lengths of iridium wire connected by a bridge, rather like the three sides of a rectangle. The 'legs' can be cut to any desired length and the separation between them is a constant 12 mm, thereby assuring parallelism. This iridium has a diameter of 0.5 mm. Iridium hairpins are implanted using specially made stainless-steel guide gutters. These guide gutters have a similar geometric shape to the hairpin, but the 'legs' are hollow and bevelled. The bridge of the guide gutter is everted and can be gripped by a long-handled or shorthandled Pierquin forceps, which allows better leverage for accurate placement. Guide gutters/iridium-192 hairpins are ideally suited to brachytherapy for floor of mouth and lateral border of mobile tongue tumors. For implantation of such tumors, the patient sits in a dental chair and the procedure is carried out under local anesthetic. Ideally, the theater should be equipped with a Carm fluoroscopic X-ray control unit. The area to be implanted is anesthetized, the guide gutters are introduced, and their geometric arrangement is verified by fluoroscopy. When found to be in a satisfactory position, a silk suture is passed underneath the bridge of the guide gutter, taking in a generous amount of tissue. The active iridium hairpin is then slipped into the bevelled edges of the guide gutter and advanced until the bridge reaches the surface tissue. A skin hook is placed over the hairpin
286 Brachytherapy in the treatment of head and neck cancer
bridge to maintain its position while the guide gutter is being withdrawn. The silk suture is then tied to ensure the hairpin does not become displaced during the irridation. Orthogonal X-ray films are taken, with, if appropriate, the previously described acrylic dental protector in place. Dosimetry is determined by both computer and manual planning [5,7,8,14].
2033
Plastic tube technique
The plastic tube technique is an elegant system of implantation, but requires considerable operator skill and expertise. It is an afterloading technique which fixes the tissues encompassed by loops of plastic tubing. It is a technique of implantation which is suited to tumors of the floor of the mouth spreading onto the ventral surface of the tongue, and vice versa, for tumors of the lateral and posterior tongue, the tonsillar fossa, tonsillar pillars, soft palate, and uvula. The technique is described here in relation to a midline Tl tumor of the soft palate. Having confirmed the indication for brachytherapy treatment by preoperative assessment and investigations, the patient is given a general anesthetic. The procedure is carried out under aseptic conditions with the patient in the supine position. A small tumor of the midline of the soft palate is implanted by looping two hollow plastic tubes from the submandibular neck region on one side, through the tonsillar pillars, across the substance of the soft palate, and exiting by the tonsillar pillar route on the opposite side of the neck. Step 1 is to pass a hollow stylette vertically through the skin of the neck and posterior floor of the mouth, so that it exits intraorally at the foot of the posterior tonsillar pillar. A length of nylon monofilament is threaded through the hollow stylette from outside and is recuperated intraorally. The stylette is then withdrawn, leaving the monofilament in place. A length of hollow plastic tubing (which will eventually hold the radioactive iridium wire) is threaded over the monofilament and advanced to the skin entry point on the neck. This plastic tubing has an outer diameter of 1.6 mm and an inner diameter of 1.2 mm. Forceps are applied at the distal end of the monofilament, which is overlapped by the plastic tubing. The intraoral monofilament is thereafter carefully pulled through the tissues in vertical fashion until the plastic tubing emerges at the foot of the tonsillar pillar. Great care must be taken when pulling the monofilament/plastic tubing through the tissues so that tearing and subsequent bleeding are avoided. A Reverdin forceps (which is curved) is passed from the contralateral side through the substance of the soft palate to emerge at the superior pole of the posterior tonsillar pillar. The monofilament is gripped by the Reverdin forceps and pulled back through the soft palate. The plastic tubing is then carefully advanced through the soft palate tissue, following the same principles as before.
To exit the tubing and complete the loop, a hollow stylette is passed as before in vertical fashion from the submandibular neck area to emerge at the foot of the contralateral posterior tonsillar pillar. The monofilament is threaded through from inside to emerge externally and the stylette is then withdrawn. The plastic tubing is pulled through very carefully as before, and the monofilament is then withdrawn. The tubing is now in place, forming a loop which passes from one side of the neck through the posterior tonsillar pillar, across the substance of the soft palate, and emerging via the contralateral posterior tonsillar pillar to the opposite neck. The plastic tubing is flushed through with a heparin solution to ensure it remains patent. A second length of plastic tubing is implanted in exactly the same manner, but on this occasion the anterior tonsillar pillar is the intraoral point of entry and exit. Although the actual physical process of implantation has been described above, no mention has been made of the relation of the tubes to each other or, indeed, to the tumor. Obviously, these factors are of vital importance if the Paris System rules are to be observed. The loops must be implanted so that they comfortably sandwich the tumor between them, while simultaneously remaining equidistant and parallel with respect to each other. In terms of dosimetry, the optimal distance between the two tubes is approximately 12-15 mm. This distance is maintained externally as follows: plastic tube spacers which have been drilled with holes separated by a distance equal to the separation between the two hollow plastic tubes are passed over the free ends of the tubes and advanced so that they come to lie in contact with the skin of the neck. Dummy wire is then threaded through the plastic tubes and is held in place by lightly clamping metal buttons which slide over the plastic tubes. Orthogonal X-rays are taken for dosimetric purposes [4,5,15,16].
203*4
Hypodermic needle technique
Hypodermic needles are most often used for cancers of the lip and nose, for which an element of rigidity in the source carriers is required. They have an external diameter of 0.8 mm and are bevelled at both ends. When the target volume has been implanted and the positions are deemed satisfactory, the needles are kept in place by a template system. The templates, which are made of perspex, are cut and drilled to the implant requirements and are then slipped over the ends of the needles. Lead caps are placed over the bevelled ends of the needles on one side of the implant and are crimped to secure them. Appropriate lengths of radioactive iridium-192 wire are manually afterloaded into the opposite ends of the
Treatment planning 287
needles and lead caps are put in position and crimped as before [5,7,17].
203.5
high dose uniformity within the target volume, and a rapid fall-off in dose outside it [5,6,8,19].
Silk suture technique 20.4
This technique is used predominantly for small tumors on the skin and eyelids. Braided silk (4.0) into which 0.3 mm diameter iridium wire can be threaded is the vector for implantation. The silk thread is first canalized by a small steel wire and then hardened by dipping into an organic compound. The steel wire is withdrawn and replaced by iridium-192 of 0.3 mm diameter, which is maintained in position by a knot on the silk thread. The tumor to be implanted is treated by passing the silk thread through the target tissue so that the portion containing the iridium wire lies in an appropriate position. Depending on the size of the lesion to be treated, several silk sutures may be threaded through the target tissue, ensuring at all times that the rules of the Paris System are respected [6,18].
203.6
The Paris System
20.4.1
TREATMENT PLANNING General assessment
As previously emphasized, the general assessment includes an evaluation of the patient, the patient's tumor under consideration, and other relevant factors such as dentition. When a brachytherapy procedure is being proposed, consideration has to be given to the manner of anesthesia. Local anesthesia is quite adequate for well limited tumors of the lip, for small tumors of the floor of the mouth or lateral border of the tongue for implantation by the guide gutter technique. Advantages of local anesthesia are that normal muscle tonus is maintained and, with a conscious patient check, fluoroscopy screening is easily carried out. Patients undergoing intraoral brachytherapy procedures under general anesthesia must be intubated by the nasal route to allow adequate room for the brachytherapist to work [1,7-9].
To respect the rules of the Paris System, several criteria have to be observed. 1. The linear activity of the sources must be uniform for all the sources used and must be uniform throughout the length of each source. 2. Source must be implanted parallel and equidistant to each other. 3. The distance between the sources may vary from implant to implant, but, once decided, must be constant for each individual case. 4. The plane which passes through the midpoints of each source at right-angles to the axis of the implant is defined as the central plane and is used for calculating the basal dose rates. 5. The basal dose rate is the minimum dose rate between lines. When three straight lines are implanted for a small skin carcinoma on the cheek, there will be two basal dose rates. When three lines are implanted in triangular fashion for a small lip carcinoma, there will be just one basal dose rate. Depending on the number of active wires and their configuration, most head and neck implants can be subdivided geometrically into triangles. The basal dose rate for each triangle is calculated. The results are added and divided by the number of triangles, giving an average basal dose rate for the entire implant. 6. The reference dose rate is equal to 85% of the mean basal dose rate. This isodose envelops the target volume and is the isodose on which the treatment is prescribed. A good implant that follows the rules of the Paris System will give precise coverage of the target volume, a
20.4.2
Tumor evaluation
At first clinical assessment the brachytherapist can generally envisage the treatment method to be employed. In rare instances, the predicted implant technique may need to be modified due to further information gained from CAT scans or MRI scans. Guide gutter, hypodermic needle and silk suture techniques require immediate loading in the operating theater. The operator does this using long-handled forceps manipulated from behind lead screens. The plastic tube technique, on the other hand, is an afterloading system, which allows dosimetry to be calculated prior to loading of the active wire. All of these factors have to be kept in mind by the brachytherapist when planning the treatment. In this respect, the advantage of the Paris System can be exploited. With clinical assessment of the tumor volume to be encompassed, a knowledge of the Paris System of implantation allows for provisional estimation of the number of lines to be implanted, the separation between the lines, and the lengths/activity of iridium to be used. Tumor evaluation therefore includes the method, spacing, and length/activity of sources to be used in accordance with the size, shape, and position of the tumor. The tumor and a margin of security are enveloped by the target volume. The target volume is therefore the volume of tissue to which we intend to deliver the prescribed dose. The treated volume is that included by the 85% reference isodose curve, and should be at least equal to the target volume [5,10,20].
288 Brachytherapy in the treatment of head and neck cancer
20.43 Geometric configuration When a series of loops or hairpins are implanted in precise geometric fashion, parallel and equidistant, the following volume applies: V=L x WxT
where V = volume, I = length, W= width, and T= thickness. Length treated = 0.8 x half total active length.
The activity of the radioactive source may have a bearing on ultimate outcome, and several studies have demonstrated a link between dose rate and local control. Such factors, therefore, impinge on the brachytherapist's therapeutic decisions. After-loading techniques such as the plastic tube method allow the brachytherapist to take great care over the geometry of the arrangement, thereby increasing the possibility of obtaining a perfect implant with uniform dose distribution [5].
Width treated = 1.55 x spacing. Thickness treated = distance between outermost branches + 0.5 x spacing.
As can be deduced from the above, the active wires must extend beyond the target volume at each end in order to compensate for dose fall-off at the limits of the implanted sources. The reference isodose dips sharply between the implanted wires or tubes at the ends of the implant due to the linear source strength of the iridium, and therefore the active ends must extend beyond the target volume to ensure the reference isodose curve equates with the volume to be implanted. Naturally, this results in a volume of tissue outside of the target volume receiving an appreciable dose of radiation, but this in turn is offset by the rapid fall-off in dose as one moves away from each source. With practice, the predictive dosimetry pertaining to each proposed implant can be envisaged, as can the influence of varying the number and spacings between the sources. Dosimetric planning consists of determining the dimensions of the target volume and choosing the method of implantation that best assures an optimal dose distribution within the treated volume. For single-plane implants, the three volume dimensions - length, width, and thickness - need to be considered. Length depends on the length of the iridium used, whereas width and thickness depend on the spacing between the sources.
20.4*4
Radioactive sources
For the specification of radioactive sources, the International Commission on Radiation Units and Measurements (ICRU) has recommended the use of kerma. The kerma rate defines the strength of a radioactive source and is measured in air at a distance of 1 m from the midpoint of the radioactive source. The unit of measurement for kerma rate is cGy rr1 m2. For LDR brachytherapy, radioactive sources must be of small diameter to allow them to be introduced into suitable carrier systems, such as plastic tubing. Sources must be to some degree pliable to conform to implant geometry and must not pose difficult radioprotective problems. Iridium is an element which fulfills these criteria.
20.5 20.5.1
TREATMENT Lip
For lip cancers, the hypodermic needle or plastic tube technique is generally used. Lip cancers are easy to assess clinically and allow one to visualize the implant technique and geometric arrangement prior to the procedure. Choice of technique depends on tumor volume, site, and anatomy. Large tumors and those involving the labial commissure are better treated by the plastic tube technique. Implantation follows the axis of the lip and preperforated templates may be introduced at both ends to maintain the geometric arrangement. General recommendations about dental protection apply and an appropriate shielding device should be worn by the patient during irradiation [1,4,5,7].
20.5.2
Nose
Cancer of the nasal vestibule may be implanted by plastic tubes, guide gutters/hairpins, or hypodermic needles. For large tumors infiltrating the columnella or nasal cartilage, plastic tubes may be used in a horizontal arrangement of several planes which transfix the nose and conform to the Paris System. Preperforated templates are used to maintain parallelism. Infiltration of the columnella may be treated by the guide gutter/hairpin technique [6,17,21].
20.5.3
Skin, ear, eyelid
The silk suture or the plastic tube techniques may be used for carcinomas of the skin, ear and eyelid. As the iridium contained within the silk suture has a diameter of 0.3 mm, this technique is suitable for areas such as the eyelid where the vector for the radioactive source needs to be of the smallest possible diameter. Carcinomas of the pinna and the external auditory canal can be treated by a mould system which conforms to the irregular anatomy [5,8,18,22].
The Institut Gustave-Roussy results 289
20*5.4
Mobile tongue
Carcinomas of the mobile tongue may be treated by either the guide gutter/hairpin or plastic tube technique. The guide gutter/hairpin method is suitable for small tumors and carries the advantage of avoiding general anesthesia. It is also a speedy procedure, which results in good parallelism and excellent dose distribution. The plastic tube technique is generally reserved for bigger tumors, requires general anesthesia, and transfixes the tongue/floor of mouth in its loops. As it is an afterloading system, there is less exposure to the operator and other personnel than that encountered by the guide gutter/hairpin technique, which requires 'live' loading in the operating theater. Choice of implant system is dependent on tumor site, volume, accessibility, and physician preference and experience [5,7,12,23-34].
20.5.5
Floor of mouth
Either the guide gutter/hairpin or the plastic tube technique can be used in the treatment of cancer of the floor of the mouth. As the separation between the 'legs' of the hairpin is a constant 12 mm, this method is reserved for small tumors. Care has to be taken that the active sources do not lie too close to the gingiva or mandible if subsequent late complications such as necrosis are to be avoided. As a general rule in this institution, brachytherapy is contraindicated if the implant demands that more than two radioactive lines lie in close contact with the mandible. The plastic tube technique can be equally applied in floor of the mouth cancers, but as this method transfixes the tongue and requires general anesthesia, the guide gutter/hairpin method is preferred providing tumor dimensions are appropriate [6,11 -13,27,39,40,42].
20.5.7
20.5.8 palate
Tonsil, tonsillar pillar, and soft
Nowadays, treatment of these sites by brachytherapy is confined almost exclusively to the plastic tube technique. Implantation follows the principles described earlier. For lesions confined to the tonsil, the loop may just involve the ipsilateral tonsillar bed. For central lesions of the soft palate, a looping system that passes from one side of the neck to the other may be used. In some instances, the loop may be fashioned in such a way that the tumor and the full width of the soft palate are enveloped by the loop, thereby confining the skin exit and entry points to one side of the neck [8,15,16,35,44].
Base of tongue
Whereas in the past the guide gutter/hairpin method was employed for base of tongue tumors, they are now treated by the plastic tube technique. The plane of implantation is generally sagittal (three loops) completed often by a frontal loop. In phase 1, the entire base of the tongue is the target volume when treating exclusively by brachytherapy. Phase 2 involves boosting the tumor and tumor bed volume by leaving the relevant sources in place for a longer duration of time, appropriate to the dose prescribed [14,19,35-38].
20.5.6
pin method is appropriate for non-infiltrating small tumors. The implantation of plastic tubes can be either in the vertical oblique or horizontal planes, depending on tumor orientation and infiltration [6,43].
Buccal mucosa
Again, the plastic tube or guide gutter/hairpin method can be applied. Choice of implant system is dictated by tumor dimensions and position. The guide gutter/hair-
20.5.9
Nasopharynx
Brachytherapy for nasopharyngeal carcinoma is a plesiotherapy technique. It is indicated in very few patients more often as a salvage treatment, as the maximum thickness of tissue that can be irradiated by this method without significant complications is 10 mm. Treatment involves taking an impression of the nasopharyngeal cavity using a very fine dental algate, and from the negative of this an individualized mould is constructed. The impression of the nasopharynx will reveal the position and dimensions of the tumor and from this the brachytherapist can decide where the plastic tubing source carriers should be positioned in the mould. The sources are afterloaded into the plastic tubes, which are appropriately positioned on the inside of the hollow acrylic applicator [6]. Tumor evaluation is of critical importance when this technique is to be considered. CAT scanning and/or MRI scanning are imperative in the assessment of the tumor dimensions.
20.6 THE INSTITUT GUSTAVE-ROUSSY RESULTS The following results are from experience at Institut Gustave-Roussy in treating head and neck cancers with LDR brachytherapy. In all cases, the rules of the Paris System were respected. The figures given are broken down into population base, age, sex, TNM status, implantation method used, and clinical results. Treatment protocol 'A refers to initial brachytherapy to the primary lesion ± (if indicated) cervical node dissection + (if indicated) external-beam irradiation.
290 Brachytherapy in the treatment of head and neck cancer
Treatment protocol 'B' refers to initial external-beam irradiation followed by brachytherapy as a boost procedure [6].
20.6.1
Lip [6]
Population: 231 patients. Mean age: 65 years (range 28-90). Males: 85%; females: 15%. TNM distribution: TI = 82%; T2 = 13%; T3 = 3%; NO = 80%. Treatment protocol: A = 97%; B = 3%. Method: plastic tube technique = 40%; hypodermic needles = 56%; silk threads = 14%. Mean dose delivered = 76 Gy. Clinical results: overall 5-year disease-free survival = 66% local control rate at 5 years = 95%. Complications: mucosal necrosis = 13%.
20.6.2 Nose [6] Population: 36 patients. Mean age: 66 years (range 44-82). Males: 83%; females: 17%. TNM distribution: TI = 45%; T2 = 45%; T3 = 10%; NO = 93%. Treatment protocol: A = 93%; B = 7%. Method: plastic tube technique = 20%; hypodermic needles = 54%; guide gutters/hairpins = 40%; silk threads = 6%. Mean dose delivered: 72 Gy. Clinical results: overall 5-year disease-free survival = 68% local control rate at 5 years = 86%. Complications: grade 1 and 2 = 33%; grade 3 = 18%.
20.6.3
Mobile tongue [5,6,11]
Population: 269 patients. Mean age: 55 years (range 25-87). Males: 77%; females: 23%. TNM distribution: TI = 31%; T2 = 55%; T3 = 14%; NO = 81%. Treatment protocol: A = 72%; B = 28%. Method for protocol A: plastic tube technique = 21%; guide gutters/hairpin = 79%. Mean dose delivered: 71 Gy. Method for protocol B: plastic tube technique = 51%; guide gutters/hairpin = 49%. Mean dose delivered: 27 Gy. Clinical results: overall 5-year disease-free survival: A = 62%; B = 30% local control rate at 5 years: A = 87%; B = 49%
complications: mucosal necrosis = 11%; bone necrosis = 8%.
20.6.4
Floor of mouth [6,11-13]
Population: 206 patients. Mean age: 53 years (range 31-85). Males: 72%; females: 28% TNM distribution: TI = 42%; T2 = 50%; T3 = 6%; NO = 70%. Treatment protocol: A = 87%; B = 13%. Method for protocol A: plastic tube technique A = 21%; guide gutters/hairpin = 79%. Mean dose delivered: 65 Gy. Method for protocol B: plastic tube technique A = 19%; guide gutters/hairpin = 81%. Mean dose delivered: 27 Gy. Clinical results: overall 5-year disease-free survival: A = 74%; B = 30% local control rate at 5 years: A = 89%; B = 59%. complications: mucosal necrosis = 11%; bone necrosis = 21%.
20.6.5
Oropharynx [6,19,37]
Population: 312 patients. Mean age: 58 years (range 38-86). Males: 85%; females: 15%. TNM distribution: TI = 25%; T2 = 40%; T3 = 35%; NO = 50%. Treatment protocol: A = 11%; B = 89%. Method: plastic tube technique = 26%; guide gutters/ hairpin = 74%. Mean dose: 75 Gy (including external-beam radiation) Clinical results: overall 5-year disease-free survival: base of tongue = 26%; tonsil = 37%; anterior oropharynx = 40%; soft palate = 37%. Local control rate at 5 years: base of tongue = 68%; tonsil = 79%; anterior oropharynx = 69%; soft palate = 93%. Complications: grade 1 and 2 = 27%; grade 3 = 7%.
20.6.6
Nasopharynx [6]
Population: 47 patients. Mean age: 42 years (range 26-57). Males: 70%; females: 30%. TNM distribution: 33 patients treated as 'boost' following external irradiation (45 Gy); 14 patients treated for recurrence in previously irradiated area. Method: moulded applicator. Mean dose for 'boost' treatment: 30 Gy. Mean dose for salvage treatment: 60 Gy.
Future directions 291
Clinical results: overall 5-year disease-free survival for 'boost' treatment = 42% overall 5-year disease-free survival for salvage treatment = 17% local control rate at 5 years for 'boost' treatment = 74% local control rate at 5 years for salvage treatment = 50%.
20.6.7
Pediatric head and neck
malignancies [6] Population: 39 patients. Mean age: 5 years (range 3 months-15 years). Main tumor sites: nasolabial sulcus = 31%; oral cavity = 21%; neck = 15%; ear = 10%. Percentage of cases with rhabdomyosarcoma: 70%. TNM distribution: TI - 61%; T2 = 36%; TX = 3%; NO = 56%; NI = 41%; NX = 3%. Treatment protocol: for children, the approach was quite different from that with adults and included chemotherapy in most cases; external radiotherapy was given in 31% of cases. Brachytherapy procedure: brachytherapy was indicated in two different situations: first-line brachytherapy = 64%; salvage brachytherapy = 36%. Plastic tubes ± hypodermic needles and/or guide gutters: 95% Guide gutters ± hypodermic needles: 5% Brachytherapy performed per-operatively: 31%. Mean dose: first-line brachytherapy = 68 Gy; salvage brachytherapy = 56 Gy. Results: 5-year disease-free survival: first-line brachytherapy = 76%; salvage brachytherapy = 50% local control rate: first-line brachytherapy = 84%; salvage brachytherapy = 64%. severe complications: first-line brachytherapy = 24%; salvage brachytherapy = 21%.
20.7
LITERATURE REVIEW
A literature review and comparison of results are presented in classified tables (Tables 20.1, 20.2, and 20.3) according to the different tumor sites. In these tables, the following abbreviations are used: Results: O = origin, TS = tumor site, NP = number of patients, TP = treatment protocol, CND = cervical node dissection, EB = externalbeam irradiation, BT = brachytherapy, PS = Paris System LDR, LC = local control, CP = complications, STN = soft-tissue necrosis, ORN = osteoradionecrosis, SV = survival, OSV = overall survival, SSV: site-specific survival, DPS: disease-free survival; tumor sites: BM = buccal mucosa, BT = base of tongue, FM = floor of mouth, GPS = glosso palatine sulcus, L = lip, MT = mobile tongue, NP = nasopharynx, NV = nasal vestibule, OC = oral cavity, OP = oropharynx, P = pinna, SP = soft palate, TO = tonsil, ToPi = tonsil + pillars.
20.8
FUTURE DIRECTIONS
While the techniques of implantation described for LDR brachytherapy for head and neck carcinomas are well established, efforts have been made to further refine them. In an attempt to diminish the exposure to staff, automatic afterloading devices for LDR isotopes are now available. However, various technical problems have hampered their universal introduction and manual afterloading continues to be widely practiced. In particular, with a plastic tube technique that involves a loop, the source needs to be able to negotiate a curved path. Such technical problems require further investigation and refinement. Similarly, dosimetry pertaining to the Paris System is also well established and its precision is reflected in the high rates of local control and low rates of complications. However, with the advent of three-dimensional planning, there exists an opportunity to better define the treatment volume. Computerized dosimetric calculation with exact definition of the target volume may afford an opportunity to ameliorate the various source arrangement possibilities and thus arrive at an optimized dose distribution within a given volume. The generation of
Table 20.1 Nasal vestibule, pinna
Levendag, P.C. and Pomp, J. 21]
NV
MazeronJ.J. [17]
Nose
Debois,J.M.[22]
P
MazeronJ.J. [18]
P
63
T1 NO 36 T2 NO 24
EB + BT BT
OS 65 DPS 80
T1 97 T279
BT
PS Iridium
93.6
140
BT
Cesium
95
70
BT
PS
94
1676
I, II, III 6
20
292 Brachytherapy in the treatment of head and neck cancer
Table 20.2 Oral cavity
MazeronJ.J. [17-19] FM
Pernot.M. [16]
FM
117
T1 47 T270
BT±CND +EB/N
PS
T1 NO 94 93.5 T2 NO 61 74.5 T2N1-N228 65
207 T185 T2 99 T316T44Tx3 NO 172
BT102 EB + BT105
PS
SSV T1 88 T2 47 T336
T197T272 T351
117,1112 III 6, IV 0.5
BT92 EB + BT 8
PS
DFS42 24
88 36
I, II 35 III 5
PS
DFS42 24
88 36
I, II 35 III 5
T1 94 T2 91 T371
STN20 ORN13
Volterrani, F. [41]
FM
175 T147T287 T319 T4 22
Benk,V. [24]
MT
110
II
BT±CND85 EB + BT25
Hareyama, M. [26]
MT
110
II
BT99 EB + BT25
MazeronJ.J. [17-19]
MT
166 T1 NO 70 T2 NO 83 T1-T2 N1-N213
Pernot, M. [29]
MT
147 T2 NO
Pernot.M. [16]
FM
207 T185 T2 99 T316T44TX3 NO 172
Iridium
PS
52 44 8
87 92 69
BT±CND70 EB + BT 77
PS
SS 62.2 34.7
89.8 50.6
BT102 EB + BT105
PS
SSV T1 88 T2 47 T336
T1 97 T2 72 T351
117,11 12,1116, IV 0.5
184 Ma 78 lib 72
Spf 85 Exoph 79 Infill 45
1138 1114
Shibuya, H. [30]
MT
370
I 90 Ma 196 lib 84
BT±CND EB + BT
Wang, C.C. [20]
MT
143
T1NO T2 NO
EB + BT
Wendt, C.T. [32]
MT
103 T1 NO 18 T2 NO 85
BT18 EB + BT 77 EB8
149
BT + ENI
Iridium
50 86 54
lOOorthoKv 100elec.eB
BT
Dearnaley, D.P. [25] MT FM Thomas L [31]
MT FM
Volterrani, F. [41]
OC
40 T1 NO T2 NO 406 T1 132.T2245, T329
29 T1/T2
BeitlerJ.J. [47]
TO
Matsuura K. [33]
MT
173
Fujita, M. [34]
MT
207 93 T1 114T2
Rudoltz, M.S. [42]
OC
55 T1 16 T226 T38 T45
75 Stage I 98 Stage 1 1
66 92 38 T1 T2 90
Iridium-cesium Radium
BT
52
65
T>20 mm
Radium
41.4
T1-279 T354
ORN22
BT+EB
Iridium 125
-
92
EB + BT 66 BTalone107
Iridium hairpins
Stage 1 93 (5 years) Stage II 78 (5 years)
EB + BT 82 BTalone125
Iridium hairpins
T193 T277
PS BT EB + BT3
EB + BT(HDR)
dose-volume histograms will, in time, allow correlations to clinical outcome, which should theoretically increase rates of cure and local control while simultaneously maintaining or diminishing current complication rates.
11.5
79 (2 years) 87forT1/2 47forT3/4
Other areas of worthwhile future investigation include biological studies involving proliferation rates, oncogene and tumor suppressor gene expression, which may, in time, better define the population base likely to benefit
References 293 Table 20.3 Oropharynx
Pernot, M. [10]
Esche, B.A. [15]
VT
361
SP
T190.T2141 T3119.T42, Tx91, NO 230
43
BT 18 EB + BT 343
EB + BT
80
PS
Specific SV
Iridium
63
PS
37
92
121,112, 1112, IV 0.2
GGPT Lusinchi.A. [37] Puthawala,A. [38]
BT
108
TO
80
T1 18T2 39 T3 51 111-IV81%
EB + BT
PS
26
T1 85, T2 50 T369
EB + BT
GGPT Iridium
72
84
from exclusive brachytherapy treatment. In our institution, the place of ultra-LDR brachytherapy is currently under investigation for patients suffering recurrence of previously irradiated head and neck cancers.
6
as serial reports confirm its usefulness in this regard [46].
REFERENCES 20.9
CONCLUSION
The aim of this chapter was not to be exhaustive but to try to show the efficacy of brachytherapy based on LDR technique. Taking into account the Institut Gustave-Roussy experience and most of the results collected from Group European Curietherapy (GEC) multicentric studies, two categories of patients can be identified: those treated with brachytherapy alone (60% of cases), in which the results are: disease-free survival (DPS) 65-75%, local control 80-90%, complications 10%; and those treated with brachytherapy boost: DPS 35-45%, local control 60-70%, complications 20%. To conclude, LDR brachytherapy plays a major role in the management of head and neck cancer: the local control rate is high, the complication rate is low. This treatment is well tolerated by the patient and can be delivered in a short time. However, successful brachytherapy depends much upon careful assessment of the patient, close collaborative efforts, and finally on the entire brachytherapy experience, which itself is based on precise planning and technique respecting the rules of the system used to implant and calculate the dose distribution, thus being able to report the results and compare them with those of other treatment approaches. It is likely that in the future, brachytherapy may be combined with chemotherapy in certain clinical situations or as an intraoperative endeavor [45]. In addition, it will continue to hold its proven role in the treatment of recurrent carcinoma in previously irradiated patients
1. Fletcher, G.H. (1992) Textbookof Radiotherapy, 3rd edn. Philadelphia, Lea and Febiger, 305-15. 2. Murnaghan, D. (1988) History of radium therapy in Ireland: the 'Dublin Method' and the Irish Radium Institute./ Irish Coll. Phys. Surg., 17(4), 174-6. 3. Stevenson, W.C. (1914) Preliminary clinical results on a new and economical method of radium therapy by means of emanation needles. Br. Med.J.,9-10. 4. Hilaris, B. et al. (1988) Atlas of Brachytherapy. New York, Macmillan. 5. Pierquin, B., Wilson, F. and Chassagne, D. (1987) Modern Brachytherapy. New York, Masson. 6. Gerbaulet, A., Haie-Meder, C, Marsiglia, H. et al. (1994) Role of brachytherapy in treatment of head and neck cancers: Institut Gustave-Roussy experience with 1140 patients. In Brachytherapy: from Radium to Optimization, ed. R.F. Mould. Leersum,The Netherlands, Nucletron International, 101-20. 7. Gerbaulet, A. and Pomp, J. (1995) The role of radiotherapy in the treatment of tumours of the oral cavity. In Textbook of Oncology, 1,1001-8. Oxford, OUP. 8. Le Bourgeois, J.P., Eschwege, F. and Chavaura, J. (1992) RadiotherapieOncologique. Ed. Hermann. Paris. 9. Mendenhall,W.M., Parson, J.T., Mendenhall, N.P. etal. (1991) Brachytherapy in head and neck cancer: selection criteria and results at the University of Florida. Oncology, 5(1), 87-93; 5(2), 44-60. 10. Pernot, M., Malissard, L, Hoffstetter, S. et al. (1994) Influence of tumoral biological and general factors on local control and survival of a series of 361 tumors of the velotonsillar area treated by exclusive irradiation (external beam irradiation + brachytherapy or
294 Brachytherapy in the treatment of head and neck cancer
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
brachytherapy alone). Int.]. Radial Oncol. Biol. Phys., 30, 1051-17. Haie-Meder, C, Marsiglia, R., Mamelle, G.,Wibault, P., Lartigau, E. and Gerbaulet, A. (1995) Curietherapie des carcinomes epidermoides de la cavite buccale. Bull. Cancer Radiother., 82,186-90. Marandas, P., Gerbaulet, A. and Luboinski, B. (1991) Tumeurs malignes du plancher de bouche. Med. Chir. Editions Techniques Otorhinolaryngologie, Paris, 20627, D10. Marsiglia, H., Haie-Meder, C., Mamelle, G. and Gerbaulet, A. (1992) Resultats de la curietherapic. A propos de 206 cancers du plancher buccal. Cancer Radiother., 79,391. Harrison, L.B., Zelefsky, M.J., Sessions, R.B.etal. (1992) Base of tongue cancer treated with external beam irradiation plus brachytherapy: oncologic and functional outcome. Radiology, 184,267-70. Esche, B.A., Haie, C., Gerbaulet, A., Eschwege, F., Richard, J.M. and Chassagne, D. (1988) Interstitial and external radiotherapy in carcinoma of the soft palate and uvula. Int.}. Radial Oncol. Biol. Phys., 15,619-25. Pernot, M., Malissard, L.Taghian,k.etal. (1992) Velotonsillar squamous cell carcinoma: 277 cases treated by combined external irradiation and brachytherapy. Results according to extension, localization, and dose rate. Int.). Radial Oncol. Biol. Phys.,23,715-23. MazeronJ.J., Chassagne, D., Crook, \.etal. (1988) Radiation therapy of carcinomas of the skin of nose and nasal vestibule: a report of 1676 cases bytheGroupe European de Curietherapie. Radiother. Oncol., 13,165-73. MazeronJ.J., Ghalie, R.,ZellerJ.rto/. (1986) Interstitial therapy for carcinoma of the pinna using iridium 192 wires: a series of 70 patients. Int.]. Radial Oncol. Biol. Phys., 12,1757-63. Crook, J., MazeronJ.J., Marinello, G.etal. (1998) Combined external irradiation and interstitial implantation forT1 and T2 epidermoid carcinoma of base of tongue: the Creteil experience. Int.]. Radial Oncol. Biol. Phys., 15,105-14. Wang, C.C. (1990) How essential is interstitial radiation therapy in curability of head and neck cancer? Int.J. Radial Oncol. Biol. Phys., 18,1529-30. Levendag, P. and PompJ. (1990) Radiation therapy of squamous cell carcinoma of the nasal vestibule. Int.}. Radial Oncol. Biol. Phys., 19,1363-7. DeboisJ.M.efo/. (1995)Curietherapyof epithelioma of the external ear. A series of 140 patients. Acta ORL Belg., 43,239-50. BachaudJ.M., Delannes, M., Allouache, N.etal. (1994) Radiotherapy of stage I and II carcinomas of the mobile tongue and/or floor of the mouth. Radiother. Oncol., 31, 199-206. Benk,V., MazeronJ.J., Grimard, Lrto/. (1990) Comparison of curietherapy versus external irradiation combined with curietherapy in stage II squamous cell carcinoma of the mobile tongue. Radiother. Oncol., 18, 339-48.
25. Dearnaley, D.P., Dardoufas, C., A'Hearn, R.P. and Henk, J.M. (1991) Interstitial irradiation for carcinoma of the tongue and floor of mouth: Royal Marsden Hospital Experience 1970-1986. Radiother. Oncol., 21,183-92. 26. Hareyama, M., Nishio, M.,Saito, k.etal. (1993) Results of cesium needle interstitial implantation for carcinoma of the oral tongue. Int.]. Radial Oncol. Biol. Phys., 25,29-34. 27. Lefebvre, J.L, Coche-Dequeant, B., Huisset, E., Mirabel, X., Ton Van J. and Prevost, B. (1994) Management of early oral cavity cancer. Experience of the Centre Oscar Lambret. Eur.J. Cancer, 30B, 216-20. 28. Mendenhall, W.M., Parsons, J.T., Stringer, S.P., Cassidi, N.J. and Millon, R.R. (1989) T2 oral tongue carcinoma treated with irradiation alone: comparison of two treatment techniques. Radiother. Oncol., 16,275-81. 29. Pernot, M., Malissard, L, Aletti, P., Hoffstetter, S., Forcard, J.J. and Bey, P. (1992) lridium-192 brachytherapy in the management of 147 T2NO oral tongue carcinomas treated with irradiation alone: comparison of two treatment techniques. Radiother. Oncol.,23,223-8. 30. Shibuya, R., Hoshina, M., Takeda, M., Matsumoto, S., Suzuki, S. and Okada, N. (1993) Brachytherapy for stage I and II oral tongue cancer: an analysis of past cases focusing on control and complications. Int.J. Radial Oncol. Biol. Phys., 26, 51-8. 31. Thomas, L, PigneuxJ., Richaud, P. etal. (1988) Lespetits cancers de la portion mobile de la langue et du plancher buccal traites par Curietherapie seule.y. fur. Radiother., 9,9-15. 32. Wendt, C.D., Peters, L.J., Delclos, L. etal. (1990) Primary radiotherapy in the treatment of stage I and II oral tongue cancers: importance of the proportion of therapy delivered with interstitial therapy. Int.}. Radial Oncol. Biol I. Phys., 18,1287-92. 33. Matsuura, K., Hirokawa.Y., Fujita, M.etal. (1998) Treatment results of stage I and II oral tongue cancer with interstitial brachytherapy: maximum tumor thickness is prognostic of nodal metastasis. Int.J. Radial Oncol. Biol. Phys., 40(3), 535-9. 34. Fujita, M., Hirokawa,Y., Kashiwado, K.etal. (1999) Interstitial brachytherapy for stage I and II squamous cell carcinoma of the oral tongue: factors influencing local control and soft tissue complications. Int.J. Radial Oncol. Biol. Phys., 44(4), 767-75. 35. Horiuchi, J., Takeda, M., Shibuya, H., Matsumoto, S., Hoshina, M. and Suzuki, S. (1991) Usefulness of 198 Au grain implants in the treatment of oral and oropharyngeal tumours. Radiother. Oncol., 21,29-38. 36. Housset, M., Baillet, F., Delanian, S. etal. (1991) Split course interstitial brachytherapy with a source shift. The results of a new iridium implant technique versus single course implants for salvage irradiation of base of tongue cancers in 55 patients. Int.J. Radial Oncol. Biol. Phys., 20, 965-71. 37. Lusinchi, A., Eskandari,J.,Son, Y.etal. (1989) External irradiation plus curietherapy boost in 108 base of tongue carcinomas. Int.J. Radial Oncol. Biol. Phys., 17,1191-7.
References 295 38. Puthawala, A.A., Syed, A.M.N., Eads, D.L., Neblett, D., Gillin, L. and Gates, T.C. (1988) Limited external irradiation and interstitial 192 iridium implant in the treatment of squamous cell carcinoma of the base of tongue. Int.}. Radial. Oncol. Biol. Phys., 14,839-48. 39. Mazeron, J.J., Grimard, L, Raynal, M. etal. (1990) Iridium 192 curietherapy for T1 and T2 epidermoid carcinomas of the floor of mouth. Int.). Radial. Oncol. Biol. Phys., 18(6): 1299-306. 40. Pernot, M., Hoffstetter, S., Peiffert, D. etal. (1995) Epidermoid carcinomas of the floor of mouth treated by exclusive irradiation: statistical study of a series of 207 cases. Radiother. Oncol., 35,177-85. 41. Volterrani, F.,Tana, S.,Trenti, H.etal. (1984) Results and indications of radiotherapy in the treatment of carcinoma of the mouth floor. Radiol. Med., 70,7-8. 42. Rudoltz, M.S., Perkins, R.S., Luthmann, R.W. etal. (1999)
43.
44.
45. 46.
High dose rate brachytherapy for primary carcinomas of the oral cavity and oropharynx. Laryngoscope, 109(12), 1967-73. Gerbaulet, A. and Pernot, M. (1985) Le carcmome epidermoide de la face interne de joue. A propos de 748 ma lades. J. Euro. P. Radiother., 6,1-4. Amdur, R.J., Mendenhall, W.M., Parsons, J.T., Million, R.R. and Cassisi, N.J. (1987) Carcinoma of the soft palate treated with irradiation: analysis of results and complications. Radiother. Oncol., 9,185-94. Vikram, B. (1999) Brachytherapy in cancer of the head and neck. Hematol. Oncol. Clin. North Am., 13, 525-9. Krull, A., Friedrich, R.E., Schwarz, R.etal. (1999) Interstitial high dose rate brachytherapy in locally progressive or recurrent head and neck cancer. Anticancer K«.,19(4A), 2695-7.
21 High dose-rate interstitial and endocavitary brachytherapy in cancer of the head and neck PETER LEVENDAG, CONNIE DE PAN, DICK SIPKEMA, ANDRIESVISSER, INGER-KARINE KOLKMAN, AND PETERJANSEN
21.1
INTRODUCTION
The development of manual afterloading techniques for radioactive sources and the introduction of artificial radionuclides have received considerable attention over the last few years and have contributed significantly to the renaissance of brachytherapy. The total elimination of radiation exposure to the nursing, medical, and physics staff as well as to patients' relatives by remote controlled afterloading devices was a logical extension of the manual afterloading concept. In 1961, the first high dose-rate (HDR) remote-controlled afterloader was introduced in the Memorial Sloan Kettering Cancer Center group, using high-activity cobalt-60 pellets. When the manufacture of small, cylindrical (<1 mm), high-activity (100-400 GBq) iridium-192 sources became feasible, the so-called 'stepping-source technique' was introduced. In this way, 'isodose-volumes' can be created very flexibly by combining careful placement of the afterloading catheters or applicators and adjustment of the dwell positions and dwell times of the iridium-192 point sources (i.e., optimization). In the University Hospital Rotterdam/Daniel den Hoed Cancer Center (UHR/DDHCC), the first HDR afterloader (microSelectron-HDR) using such a highactivity iridium-192 point source was introduced for
interstitial and/or endocavitary brachytherapy in cancer of the head and neck in 1990. Early on, it was recognized, however, that there might be a radiobiological disadvantage in applying large HDR (1-3 Gy min"1) single fractions, even if the catheters or applicators are carefully positioned and good immobility is achieved. Realizing this, one can, of course, try to fractionate the HDR brachytherapy to some extent. Ultimately, it was proposed to try to combine the radiobiological advantage of low dose rate (LDR), as opposed to HDR, with the practical advantages of computerized stepping-source HDR afterloading techniques in 'pulsed dose-rate' (PDR) brachytherapy, i.e., using multiple small fractions (pulses) per day with a fixed (small) interval in between the pulses. This necessitated the availability of a second afterloader; this microSelectron-PDR was obtained in 1992. In principle, the design of the PDR afterloading machine is quite similar to that of the microSelectron-HDR, except for the fact that the iridium-192 source for the PDR machine is about ten times weaker (37 GBq), enabling one to deliver multiple (e.g., every 0.5-3 h), small (e.g., 0.5-3 Gy) fractions to a composite implant. However, confusion emerged as to what fraction size at what interval should be considered optimal. We took the opportunity to design and test two types of fractionation schedules, one to be used in conjunction with the microSelectron-HDR
Brachytherapy schedules in Rotterdam 297
('fractionated HDR schedule;' i.e., a regimen with a relatively low degree of fractionation), and one with the microSelectron-PDR ('PDR schedule'; i.e., a regimen with a relatively high degree of fractionation); for details, see section 21.2. These two fractionation schedules have been consistently clinically tested with regard to tumor control and side-effects for some years (and, in fact, will remain under investigation for a number of years to come). This chapter summarizes the rationale for using either the microSelectron-HDR or microSelectron-PDR in our department. Some aspects regarding the brachytherapy techniques per se and the clinical brachytherapy protocols of the Rotterdam Head and Neck Cooperative Group are illustrated, and the preliminary findings obtained with fractionated HDR and PDR brachytherapy in cancer of the base of tongue, mobile tongue cancer, nasal vestibule (mould and interstitial brachytherapy), tonsillar fossa and soft palate (interstitial brachytherapy), and cancer of the nasopharynx (endocavitary brachytherapy) are reported. Finally, future possibilities regarding the implementation of brachytherapy in the head and neck using an integrated brachytherapy unit (IBU) are discussed.
21.2 BRACHYTHERAPY SCHEDULES IN ROTTERDAM 21.2.1 Rationale for choice of type of remote-controlled afterloaders As stated in the previous section, in the DDHCC we have consistently used for all brachytherapy cases in the head and neck either the microSelectron-HDR or the microSelectron-PDR, i.e., we have departed from LDR totally since August 1990. For both types of remote-controlled afterloaders, different fractionation schedules were designed: either using a fraction size of 3 Gy, twice daily, with an interval of 6 h (fractionated HDR; daytime regimen; HDR afterloader), or 1, 1.5 or 2 Gy per fraction, four or eight fractions per day, 3-h interval (PDR; daytime regimen in the case of four fractions or 24-h per day schedule in the case of eight fractions; PDR afterloader). For the calculation of the fractionated HDR and PDR schedules to be equivalent to LDR in tumor effect, the linear-quadratic (LQ) model was used in an incomplete repair formulation (for details, see section 21.2.2). The advantages of fractionated HDR and/or PDR brachytherapy over LDR are summarized in Table 21.1. The reasons for the implementation of two afterloading machines in conjunction with the different types of fractionation schedules are summarized in Table 21.2. The choice of when to use which afterloading machine is somewhat arbitrary; however, given the potentially sig-
Table 21.1 Advantages of fractionated HDR or PDR brachytherapy schedules 1. Afterloading-elimination of radiation hazard 2. High-activity point source (37-400 GBq) - no source preparation 3. Stepping source-dose optimization (NPS/PLATO) 4. Disconnection of patients during interval between fractions: easy doctor/nursing/family/care increased flexibility for patients (e.g., outpatient brachytherapy) potential for application of other modalities during interval (e.g., interstitial hyperthemia, hypoxiccell cytotoxins) 5. Continuation/automation brachytherapy over weekend (PDR)
nificant differences in biological outcome, the choice was made at least in a structured (protocolized) fashion. In summary: when the microSelectron-PDR is available, our preference will be the PDR schedule ('small fraction size' and potential for continuation of the irradiation over the weekend); if the PDR afterloader is connected to another patient or if it is felt that clinically and/or technically the patient (i.e., implant) is more suitable for fractionated HDR, we opt for the microSelectron-HDR (for details, see section 21.3). Of course, not every department can implement (afford) two types of afterloading machines, in particular if the brachytherapy caseload is relatively small (e.g. <100 cases/year). As, logistically and biologically speaking, the fractionated HDR and PDR schedules are quite different types of brachytherapy treatments, we hope that at some point in the near future we will be able to arrive at some kind of conclusion which can contribute to making the 'right' departmental choice as to which fractionation schedule (and type of afterloader) is to be preferred (see also section 21.2.2).
21.2.2 Radiobiological considerations and dose fractionation for fractionated HDR and PDR schedules The incentive to initiate fractionated HDR and/or PDR brachytherapy, and to test its feasibility as a replacement for 'classical' LDR brachytherapy, is the wish to combine the (mainly radiobiological) advantages of LDR brachytherapy with the (mainly physical and socioeconomic) advantages of modern HDR afterloaders, as summarized in the previous section. The conditions under which PDR brachytherapy may simulate LDR brachytherapy are being investigated in experimental models, both in vitro and in vivo in different laboratories [1]. This section deals with the choice of treatment schedules for fractionated HDR (with two fractions per
298 HDR interstitial and endocavitary brachytherapy in cancer of head and neck Table 21.2 Considerations that might favor having two afterloaders (e.g., microSelectron -HDR and -PDR) in conjunction with two different types of fractionation schedules - two fractions per day (daytime regimen, fractionated HDR), or eight fractions per day (24-h regimen, PDR) - in a single radiation therapy department*
Some tumor sites benefit typically from an HDR single-fraction treatment (e.g., palliation in carcinoma of bronchus, esophagus) With integrated brachytherapy unit, HDR afterloader essential From clinical experience, some implants (i.e., patients) are not suitable for a 24-h PDR schedule (implant technical and/or medical reasons, e.g., implants brain, nasal vestibule) Medical/physics support off-hours regarding safety precautions favor daytime HDR (depending also on national regulations) Brachytherapy on outpatient basis favors daytime HDR PDR
If LDR is radiobiologically more permissive regarding late side-effects, strong fractionation can be of importance (in fact, the use of small fractions, e.g., 1-3 Gy, with short intervals, e.g., 1-3 h has been proposed (PDR) For some (volume) implants, when using small fractions, transit times of high-activity point source (e.g., 100-400 GBq) are short; therefore, in the case of PDR (small fraction size), weaker point source (e.g., 37 GBq) needed Surgical demands might force brachytherapy procedure into end of week; because HDR afterloader can only be used at daytime during week days, fractionated HDR brachytherapy as opposed to PDR sometimes cannot start until after the weekend
HDR and PDR Large and variable brachytherapy caseload * For details, see text.
day) and for PDR brachytherapy (with more than two fractions per day) as alternatives to 'classical' LDR treatments. RADIOBIOLOGICAL MODEL
The model which is used to calculate fractionated HDR and PDR schedules is the linear-quadratic model in an incomplete-repair formulation as given by Brenner and Hall [2]. As stressed by these authors, the model is essentially based on the formalism described by Lea and Catcheside (1942), published more than 50 years ago [3]. For a full description of the model, the reader is referred to the paper of Brenner and Hall; only the main features are summarized here. The expression used for the survival of clonogenic cells is the familiar linear-quadratic one [4,5]. In order to take into account incomplete repair of sublethal damage, either during LDR irradiation or in the (limited) time between HDR/PDR fractions and during the (finite) duration of each radiation fraction, the quadratic term in the expression for the surviving fraction S(D) after a dose D is assumed to be reduced by a factor G:
Brenner and Hall have given the expressions for two specific cases of interest, that is, LDR brachytherapy with a continuous irradiation and PDR brachytherapy with a series of equal fractions at equal intervals. The expression for LDR irradiation is identical to the one given by
Dale [6] and Thames, et al. [7]. Treatment schedules with varying pulse intervals or combinations of different treatment schedules can also be evaluated. In order to compare treatment schedules for clinical practice, the results of model calculations are usually expressed in the quantity: BED =
InS
(21.2)
This quantity (with the unit dose) was designated by Barendsen [4] as the ETD (extrapolated target [or tolerance] dose) and by Fowler [8] as the BED (biologically effective dose). We follow the latter nomenclature here. Apart from comparing treatment schedules in terms of survival level, the effect of a treatment can also be expressed in terms of tumor control probability (TCP) or, if normal tissue effects are concerned, in normal tissue complication probability (NTCP). These quantities have the advantage of being directly related to data obtained from clinical studies, i.e., local control probability and complication probability. The TCP/NTCP model is based on the assumption that the probability of local control follows Poisson statistics, i.e., is an exponential function of the mean number of surviving clonogenic cells (Ns) after treatment:
where N0 is the average initial number of clonogenic cells. Regarding the probability of normal tissue compli-
Brachytherapy schedules in Rotterdam 299
cations, the analogous quantity could be the number of 'functional subunits' without which the organ would lose its function. In order to derive clinical HDR or PDR brachytherapy schedules, we have applied the following procedure. Because more clinical data are available for external radiotherapy, it seems adequate to choose an external radiotherapy schedule as a reference schedule, giving one fraction of 2 Gy per day and five fractions per week. As a practical example, we will study a case in which a brachytherapy boost has to be given which should be equivalent in tumor effect to an external radiotherapy schedule of 11 fractions of 2 Gy, i.e., 22 Gy physical dose. As we are dealing with the replacement of LDR brachytherapy schedules, subsequently LDR schedules equivalent to this reference external radiotherapy schedule were calculated. It should be stressed that we consider two treatment schedules 'equivalent' if both have the same BED value for a specific end-point (tumor effect or normal-tissue effect). That means that we have not applied the additional condition that the total physical dose and/or the overall time of a PDR treatment must be identical to the values for the corresponding LDR treatment regime. This additional condition was applied in proposals for PDR schedules by Brenner and Hall [2] and by Fowler [8]; however, we see no significant reason why (slight) variations in physical dose and overall time between PDR and LDR brachytherapy schedules should not be allowed for. A longer overall treatment time might lead to an increased risk of tumor cell repopulation, but for the rather short treatment schedules in brachytherapy (mostly not longer than 1 week), the probability of increased tumor repopulation seems limited.
The BED of an LDR schedule is determined using:
The model parameters used are given in Table 21.3. For the LDR schedules a constant dose rate of 50 cGy rr1 was taken. In our case, an LDR schedule of 46.3 h application time (physical dose 23.2 Gy) should have a tumor effect equal to that of the reference external radiotherapy schedule of 22 Gy. To determine an equivalent HDR or PDR schedule, first a fraction size and a fixed interfraction interval are chosen and the number of fractions can then be calculated, yielding a BED closest to the BED of the reference LDR schedule. The results are shown in Table 21.4. It can be observed that this approach results in different physical doses, i.e., with the LDR schedule a physical dose of 23.2 Gy is given, and with the fractionated HDR schedule a lower dose of 21 Gy, while for the PDR schedule a somewhat higher dose of 24 Gy is applied. EFFECTS OF DIFFERENT REPAIR HALF-TIMES AND OF DIFFERENT cc/p VALUES
The values of the LQ parameters and the repair time constant used in the calculations are given in Table 21.3. Unfortunately, limited data are available for human tumors and normal tissues and, furthermore, a large variation can be expected for different tumors. Brenner and Hall [2] presented the following values (arithmetic mean ± SD) for a set of 36 human cell lines in vitro: a = 0.36 ± 0.21 Gy-1, oc/p = 7.3 ± 5.4 Gy, repair half-time (TO = 0.54 ± 0.91 h. The human data regarding recovery capacity ( a / b ) and repair kinetics (TO have been
Table 21.3 Values used for the model parameters
50 cGy h-1
LDR dose rate Early effect/tumor late effects
o/P
10 Gy 3Gy 0.3 G>r1
a Half-time of repair of sublethal damage
Early effect/tumor late effects
1.0 h 3.0 h
Table 21.4 Fractionation schedules for fractionated HDR treatments (i.e., two HDR fractions per day with a minimum interval of 6 h) and PDR treatments (continuous treatment with a constant time interval of 3 h between pulses)*
11x2 = 22 Gy
26.4
46.3 h = 23.2 Gy
7x3.0 = 21.0Gy
27.1
24x1.0 = 24.0 Gy
* The schedules are intended to be equivalent in terms of tumor effects, applying the parameters as listed in Table 21.3.
27.0
300 HDR interstitial and endocavitary brachytherapy in cancer of head and neck
reviewed by Thames et al. [7]. The results show that the differences observed for cc/P between early and late effects in animals also hold for humans; for early reac-
tions cc/P is in the of range 7-11 Gy, and for late reactions a/P is in the range of 2-4 Gy. Data on repair kinetics are scarce, but the available data indicate that repair in human normal tissues might be slower than in rodents. For early skin reactions the repair half-time is about 1 h, whereas for late telangiectasia it may be as long as 3 h. Pop et al. (9) have pointed out that, from the available clinical experience with head and neck cancer treatments with LDR brachytherapy on the one hand and with external radiotherapy on the other hand, constraints can be derived for the possible values of the repair half-times. The assumption of a long repair halftime (i.e. > 3 h) in combination with a high dose of LDR brachytherapy would lead to extremely high BED values which are rather improbable. Due to the uncertainties, the choice of a single set of parameters is necessarily rather arbitrary. Therefore, we have chosen not only to calculate equivalent schedules for a chosen 'plausible' set of parameter values but also to evaluate the effects of varying the parameter values over a wide range. For this purpose, an approach proposed by de Boer [10] has been followed. By calculating equivalent HDR or PDR schedules for a large range of values, graphical 'maps' of the PDR/LDR dose ratio can be constructed as a function of the oc/P ratio (Figure 21.1, horizontal axis, ranging from 1 to 15 Gy) and the repair half-time (Figure 21.1, vertical axis, ranging from 0.1 to 5 h). For each 'map', a specific LDR reference schedule is used, e.g., 23.2 Gy given in 46.3 h, with a dose rate of 50 cGy h"1. Next, a specific equivalent schedule (fractionated HDR or PDR) is chosen by specifying the fraction dose and the interval between fractions. For each combination of oc/P and TI values, the number of fractions is calculated, giving an effect equal to that of the LDR schedule (i.e., equal BED values). The curves in Figure 21.1 represent the ratio of the PDR and LDR doses, expressed in percentages. These 'maps' of the PDR/LDR dose ratio can be utilized to estimate the effect of different values for oc/P and T^: if the PDR/LDR dose ratio is increasing going from tumor effects (a/p = 10) to late effects (cc/P = 3), this means that the chosen schedule can be considered 'safe,' because the late effects in normal tissues appear not to be dose limiting. If, however, the PDR/LDR dose ratio is decreasing going from tumor effects to late effects, one
Figure 21.1 Ratio of total PDR (or fractionated HDR) dose to LDR dose as a function of oc/P (Gy) and the half-time for repair of sublethal damage (7), in h)for different PDR/HDR schedules. The reference LDR schedule is the same for all three panels: 23.2 Gy in 46.3 h (i.e., a dose rate of 50 cGy h~1); this reference LDR schedule is a boost schedule, intended to have tumor effect equal to an external-beam radiation therapy (ERT) schedule of 22 Gy in 11 fractions of 2 Gy. Over the area of the 'map', the PDR or fractionated HDR dose has been calculated which would give an effect equal to that of the reference LDR schedule. Upper panel: PDR schedule with a time interval of 3 h between fractions and a fraction dose of 1.0 Gy. Central panel: 'daytime PDR' schedule applying four fractions of 2 Gy per day, with a time interval of 3 h between fractions. Lower panel: fractionated HDR schedule with two fractions of 3.0 Gy per day, with a time interval of 6 h between the two fractions on the same day. The hatched regions indicate parameter ranges considered plausible for early effects (a/P of about 10 Gy) and late effects (o/P of about 3 Gy), with repair half-times varying, respectively, from 1 h to 3 h and from 1.5 h to 3 h.
Brachytherapy schedules in Rotterdam 301
should be cautious because late-responding normal tissues may be overdosed, i.e., the probability of late effects in normal tissues may increase compared to the reference LDR treatment schedule. In Figure 21.1, PDR/LDR dose ratio maps are shown (see legend to Figure 21.1). Plausible ranges for the values of a/ (3 and TI are indicated by a cross mark and hatched rectangular areas, respectively: for early effects (including tumor effects), a/(3 is taken to be 10 Gy and TI is taken to be 1 h, as indicated by the cross mark. For late effects, we assume that a/b will be about 3 Gy and Ti slightly longer or equal to the value for early effects, i.e., a range of 1.5-3 h. From Figure 21.1, one can conclude that when choosing a PDR schedule with equal tumor effect, the PDR/LDR dose ratio is somewhat higher than 100%, depending on the value of Tj chosen. In other words, depending on the assumptions for the values for cc/P and T{, the total dose of the PDR schedule is somewhat larger than that of the reference LDR schedule for an equal tumor effect. With regard to late effects in normal tissues, the PDR/LDR dose ratio for these PDR schedules varies from 110% to 120%. This seems satisfactory: the PDR/LDR dose ratio for late effects is higher than for tumor effects and therefore with this type of PDR schedules late effects in normal tissues appear not to be dose limiting. In the central and lower panels of Figure 21.1, similar dose ratio maps are shown for schedules with fewer fractions, i.e., 'daytime PDR' schedules (four fractions of 2 Gy during daytime) and fractionated HDR schedules (two fractions of 3 Gy per day), respectively. It can be concluded that with less fractionation the dose ratio for late effects becomes closer to the dose ratios for early (tumor) effects, but even for the HDR schedule, there are no indications for overdosing late-responding normal tissues. A general trend in these maps is that, provided that the repair half-time for late effects is equal to or longer than that for tumor effects, the dose ratio for late effects is equal or greater than the dose ratio for early effects. The (PDR/LDR) dose ratios for late effects will decrease if the repair half-time for late effects is smaller than for tumor effects. However, current radiobiological data for repair kinetics point to longer repair times for late normal-tissue effects [11-13]. OVERALL EFFECT OF EXTERNAL RADIOTHERAPY (ERT) AND A BRACHYTHERAPY (BT) BOOST
It should also be kept in mind that in clinical practice the majority of patients receiving some form of brachytherapy are treated with a combination of external radiotherapy and brachytherapy. This will dilute the potential differences between fractionated HDR, PDR, and LDR, in particular in terms of risks of normal-tissue complications. This effect has been investigated in Figure 21.2, where again dose ratio maps have been calculated, this
time for the ratio (ERT + BT)/ERT, i.e., a combined treatment schedule in comparison to external radiotherapy alone. The reference external radiotherapy schedule is 35 fractions of 2 Gy given in 7 weeks. In the combined schedule an external radiotherapy series of 23 fractions of 2 Gy is followed by a brachytherapy boost (see also the legend to Figure 21.2). It can indeed be observed that the differences between the two brachytherapy regimes tend to be smaller in the combined treatments, i.e., the dose ratio for late effects for both HDR and PDR schedules is less than 5% lower than the dose ratio for early effects (at the cross mark).
Figure 21.2 Ratio of total dose for a combined ERT + BT schedule and a treatment schedule using ERT alone, as a function of o/p (Gy) and the half-time for repair of sublethal damage (7i, in h)for two different BT schedules, i.e. fractionated HDR and PDR. The reference ERT schedule is 35 fractions of 2 Gy in 7 weeks. The ERT series in the combined treatment is 23 fractions of 2 Gy. For each combination of a/$ and 7|, the number of fractions of the BT schedule (fractionated HDR or PDR) is calculated which would have a BED equal to that of the reference ERT schedule. Upper panel: a PDR schedule with a time interval of 3 h between fractions and a fraction dose of 1.0 Gy. Lower panel: fractionated HDR treatment schedule applying two fractions of 3 Gy per day, with a time interval of 6 h between fractions.
302 HDR interstitial and endocavitary brachytherapy in cancer of head and neck
21.2.3 Dose fractionation: table of reference, fractionated HDR and PDR As already mentioned, for each fractionation schedule (fractionated HDR or PDR) the fraction size, interfraction interval, and total number of fractions are fixed. We had chosen at the time to tailor the total dose of brachytherapy according to T-stage (Tl/2 versus T3/4). In case of re-irradiation, we have deliberately coned down the dose up-front, because it was felt from previous experience that we are only alleviating symptoms, i.e., we are aiming for (temporary) local-regional control as only rarely can patients be cured [17]. For cancer of the nasal vestibule, due to the severe acute (skin) reactions caused by brachytherapy (with tumors located in the skin) with otherwise excellent local control rates, we also lowered the total dose somewhat. Table 21.5 summarizes the fractionation regimens in use as of January 1995 until 2000. As of 2001 we have again slightly modified our brachytherapy schedules. This was done after rigorous evaluation of the clinical results, including side effects. For clinicians interested in our currently used table of reference, the reader is invited to contact the first author of this chapter (
[email protected]).
21.3 BRACHYTHERAPY TECHNIQUES AND PRELIMINARY CLINICAL RESULTS
213.1 Introduction: clinical brachytherapy For afterloading fractionated HDR or PDR brachytherapy, either commercially available standard afterloading catheters (outer diameter 2 mm), in the case of interstitial implants, or a mould technique, for example in the
case of cancer in the nasal vestibule and nasopharynx, are used. In general, brachytherapy techniques are not (very) dissimilar to those used for LDR (see also Chapter 20). The following paragraphs briefly summarize some relevant aspects of our techniques, dosimetry guidelines, and preliminary clinical results in different clinical sites in the head and neck.
213.2 Base of tongue and Mobile tongue: technique of interstitial volume implant Implants of the base of tongue (and mobile tongue) are interstitial volume implants. The dose specification of these implants can be considered as a variation of the well-known Paris System used for LDR treatments. That is, we specify the dose as being 85% of the average dose in the local minima of the central plane. When treating this type of implant with an HDR or PDR afterloader using a single stepping-source, a dwell position separation of 5 mm is chosen. A sagittal view of a base of tongue implant is shown in Figure 21.3 (taken from reference 18). This type of implant usually consists of three sagittal planes, each containing a 'looping' catheter running over the dorsum of the base of tongue and two or three blindending catheters with buttons sutured to the looping catheter. After reconstruction from orthogonal radiographs, a central plane is chosen more or less perpendicular to the main direction of the blind-ended catheters through the center of the active part of the implant, as indicated in Figure 21.3. The minimal dose in the central plane will, in general, be located in the geometric centers of triangles which are constructed from the intersections with the catheters. These centers of the triangles are used as reference points and are designated as 'basal dose points.' The reference dose is specified as 85% of the average dose in the basal dose points. In order to increase
Table 21.5 Table of reference for fractionated HDR and PDR brachytherapy
Nasal vestibule T1 Nasal vestibule T2, 3 Nasopharynx After 60 GyERT After 70 GyERT Any other site T1/2 booster dose T3/4 booster dose Any other site T1/2 full course T3/4full course Re-irradiation
50 54
60 64.8
52.5 56.7
60 64.8
16x3 17x3
61.9 65.7
20 14
24 16.8
21.1 14.8
24 16.8
6x3 4x3
23.2 15.5
22 26
26.4 31.2
23.2 27.4
26.4 31.2
7x3 8x3
27.1 30.9
24x1.0 26x1.0
27 29.2
66 68 58
79.2 81.6 69.6
69.3 71.4 60.9
79.2 81.6 69.6
20x3 21x3 18x3
77.3 81.2 69.6
44x1.5 46x1.5 39x1.5
78.3 81.9 69.4
Total number of fractions for cancer of the nasal vestibule and other tumor sites relative to T stage. (*): in the case of booster doses, generally 46 Gy are given by means of external-beam radiation therapy (ERT). Full course: radiation is only given by means of brachytherapy. The booster dose for cancer in the nasopharynx is given after either 60 Gy ERT (endocavitary brachytherapy 6 * 3 Gy) or 70 Gy ERT (endocavitary brachytherapy 4*3 Gy). Please note that after reviewing clinical experience, we have slightly modified this table of reference. For detailed information regarding the use of these experimental brachytherapy schemes, including our current schedules, refer to
[email protected].
Brachytherapy techniques and preliminary clinical results 303
Figure 21.3 Schematic view of the sagittal plane of a base of tongue implant, showing a catheter running partly over the dorsum of the tongue as well as catheters sutured to this so-called 'looping' catheter. (From reference 18, with permission.)
the dose uniformity, the dose distribution is optimized using geometric optimization, i.e., the volume mode [18,19]. In this optimization method, the dwell time in a dwell position is inversely proportional to the dose contribution from all other dwell positions except the ones in the same catheter [20-22]. This optimization method is sensitive for deviations in spacing and/or divergence of catheters. An example of the result is presented in Figure 21.4 (taken from reference 18), showing a sagittal view of an optimized and a non-optimized dose distribution of the base of tongue implant. After geometric optimization, the overdosed region receiving at least 200% of the reference dose is reduced, especially at the intersection of the blind-ended catheters and the looping catheter. Base of tongue cancers are, in general, treated with a combination of a volume implant of the base of tongue and an external radiotherapy dose of 46 Gy (conventional fractionation) to the base of tongue and neck. In case of positive neck nodes, the external radiotherapy of the neck is followed by a therapeutic (radical) neck dissection at the time of the implant. The same policy is followed for N+ mobile tongue cancer patients. For Tl,2 NO mobile tongue cancers, an interstitial implant of the mobile tongue is combined with a functional neck dissection. With regard to dose fractionation of base of tongue and mobile tongue implants, the reader is referred to the guidelines given in Table 21.5. At the time of writing, the number of volume implants treated by fractionated HDR and PDR brachytherapy in our cancer center, in particular for mobile tongue cancer, is small. At present, an extensive analysis in terms of local-regional control and acute and/or late side-effects is being performed for base of tongue cancer. For preliminary results, however, the reader is referred to references 23 and 24. Current treatment schedule (as of January 2001) for all T-stages (Tl-4): external radiotherapy 23 X 2 Gy. Brachytherapy in case of fractionated HDR 4 Gy + 4 times 3 Gy + 4 Gy (total dose 20 Gy; 2 fractions per day).
Figure 21.4 Sagittal plane through middle catheter running over dorsum of base of tongue; (a) non-optimized and (b) optimized dose distribution. (From reference 18 with permission.)
PDR: 2 Gy + 28 times 1 Gy + 2 Gy (total dose 22 Gy; 8 fractions per day).
2133 Nasal vestibule: technique of interstitial single-plane implant Although technically not obligatory, in the DDHCC, general anesthesia is preferred when implanting the nasal vestibule. After decongestion of the nose (R/xylometazoline), the points of entry for the trocar are demarcated on the skin. A hollow, stainless-steel guide needle and trocar are introduced interstitially. After removal of the trocar, afterloading catheters are inserted approximately 4 cm into the guide needle; subsequently, the guide needle is removed. Catheters are fixed and sutured to the skin by standard buttons. The number of catheters (generally four to five) is defined by the target volume. If the tumor is extending into the upper lip, an extra catheter is inserted 'horizontally' through the lip. Due to the anatomy of the nose per se and the geometry of the target, the catheter separation frequently varies over the target volume after implantation (Figure 21.5). The
304 HDR interstitial and endocavitary brachytherapy in cancer of head and neck
Figure 21.5 Schematic diagram of an interstitial implant in the nasal vestibule. (See also corresponding X-ray films in Figure 21.6.)
geometrical configuration of the implant is visualized on orthogonal X-rays postimplantation (or on the image intensifier in the Integrated Brachytherapy Unit (IBU), see section 21.4), with the patient still under general anesthesia. If the dose distribution of the implant is not satisfactory (i.e., anticipated coldspots and/or hotspots), additional catheters can be introduced (Figure 21.6). The dose is usually prescribed in so-called dose points at a distance of 0.5 cm from the catheters. This clinical example demonstrates the advantage of optimization (compare B and C to A, and D to E in planes I and II of Figure 21.7); however, it also demonstrates that, even with optimization, a 'poor' implant can never become a 'good' implant (compare A and C in planes I and II). After optimization of the implant (Figure 21.7), fractionated HDR is applied (for number of fractions, see Table 21.5). Current treatment schedule for all T-stages as of January 2001 is: fractionated HDR 15 x 3 Gy (total dose 45 Gy; 2 fractions per day). Moreover, the mould technique has been abandoned and only an implant using standard afterloading catheters is being used. In cancer of the nasal vestibule, only very rarely is the neck involved (less than 10%); we therefore do not treat the neck electively.
213A
Nasal vestibule: mould technique
In selected cases (e.g., contraindication for general anesthesia, small and very superficial lesions of the mucosal lining of alae or septum, patient preference), a mould technique can be used. After topical anesthesia of the nasal vestibule, the hair in the nasal cavity (nostrils) is cut away and posteriorly the nasal airway is blocked by a cotton-wool plug. Subsequently, an impression is made with alginate material (R/CA 37 superiorpink). From this provisional model a 'negative impression' of plaster of Paris is constructed. Three flexible aluminium wires (outer diameter 2 mm) are positioned and fixed into the plaster cast at the sites where afterloading catheters are to
Figure 21.6 Lateral X-ray films of a patient with an interstitial implant in the nasal vestibule (left ala). During the operation, it was realized that the implant with four catheters (A) was suboptimal in terms of anticipated dose distribution. Subsequently, an extra catheter was implanted (B). Lead demarcation visualizes tumor core lateral surface nose (skin). The stippled zone represents the planning target.
be positioned (see Figure 21.8). Using this mould technique, about midway through the series of fractions, the insertion of the mould becomes painful and thus less accurate. In our center, therefore, the interstitial technique is definitely the technique of choice.
213*5 Clinical results: nasal vestibule brachytherapy Brachytherapy of cancer of the nasal vestibule seems very rewarding. Between 1990 and 1994, 18 primary squamous cell carcinomas were treated, nine interstitially and nine by a mould technique. Out of 18 patients, two experienced a local relapse and one had a regional relapse on both sides of the neck. All relapses were salvaged surgically. It might be of interest that both local failures were found in nasal vestibule cancers treated by a mould technique. Although the mould technique seems simple and is more convenient to the patient, the positioning of the mould during each fraction is critical, in particular, as stated before, when the nose (vestibule) becomes more sore (Figure 21.9). Both techniques, when performed properly, give excellent cosmesis.
Brachytherapy techniques and preliminary clinical results 305 Figure 21.7 As an example, planes I and II were taken through the periphery of the target. (See also the legend for Figure 21.6.) Five types of dose distributions (A, B, C, D, and E) are compared. A: dose distribution with four catheters - constant dwell times and dwell positions (= nonoptimized, 'LDR equivalent'); B: dose distribution with four catheters, geometrically optimized; C: dose distribution of four-catheter implant, optimized on dose points 0.5 cm from catheters; D: dose distribution of five-catheter implant, optimized on dose points 0.5 cm from catheters; E: dose distribution with five catheters, constant dwell times and dwell positions (= nonoptimized, 'LDR equivalent'). Fraction size was 300 cGy (fractionated HDR schedule).
Figure 21.8 Plaster of Paris mould of carcinoma in the nasal vestibule ('negative' impression) with aluminium wires (outer diameter 2 mm) in situ. Subsequently, the definitive 'positive' silicone mould is manufactured with the aluminium wires replaced by afterloading catheters (see Figure 21.9). The mould is placed in situ and held by a head strip (cord) during the irradiation. Dosimetry and dose prescription are similar to interstitial implant of the nasal vestibule.
213.6 Tonsil and soft palate: technique of single-plane interstitial implant The implant of the tonsillar fossa and/or the soft palate can be performed as a single-plane or volume implant.
Figure 21.9 Final silicone mould for carcinoma of the nasal vestibule suitable for fractionated HDR brachytherapy on an outpatient basis. Three afterloading catheters are visible in the silicone mould.
Before embarking on such an implant, it is of paramount importance to investigate whether the target volume can be covered adequately by the implant. This means that computerized tomography (CT) and/or magnetic resonance imaging (MRI) of the tumor mass and neighboring structures (e.g., parapharyngeal space, nasopharyngeal side of the soft palate, retropharyngeal nodes etc.) are mandatory. If the target is considered suitable for implantation, the primary visible lesion is tatooed
306 HDR interstitial and endocavitary brachytherapy in cancer of head and neck
under general anesthesia. The next question regards the neck. In principle, for T1/T2 tonsillar carcinoma we treat the ipsilateral neck electively (NO) by either surgery (functional neck dissection) for small/superficial lesions of the primary tumor or external radiotherapy to a dose of 46 Gy (neck and primary) in case of bulky primary tumors. Subsequently - that is, after the functional neck dissection or external radiotherapy (46 Gy) - the primary tumor mass is implanted. The same policy applies for NO soft palate tumors, except that the neck is treated bilaterally. For N+ in the great majority of patients, the neck and primary are treated by external radiotherapy 46 Gy, followed by a neck dissection in conjunction with an implant of the primary tumor (soft palate and/or tonsillar fossa). The implant per se is very straightforward. Basically, by means of hollow guide needles, catheters are introduced into the tonsillar fossa (anterior and posterior tonsillar pillars) and soft palate. In general, two to three parallel running catheters are needed (Figure 21.10). The target is demarcated by permanent implantation of surgical clips or metallic seeds. For dose prescription, dose points are positioned at a distance of 0.5-0.75 cm from the catheters and the implant is subsequently optimized. On rare occasions (e.g., retropharyngeal nodes) extra dose points can be positioned at specified locations. Tumors are treated by either fractionated HDR or PDR. (For the total number of fractions, see Table 21.5).
213.7 Clinical results: tumor control of tonsillar fossa and soft palate From November 1990 until December 1994, 40 patients with primary squamous cell carcinoma originating from the tonsillar and fossa soft palate were treated by interstitial radiation therapy with or without external radiotherapy to the primary. Eighteen patients were NO, 22 had neck nodes at admission (N+). Of the implanted primary tumors, five were staged as Tl (13%), 22 as T2 (55%), 11 as T3 (28%) and 1 as T4 (3%). The neck was treated either by elective neck irradiation or functional neck dissection only (NO; «=18), or by a therapeutic neck dissection combined with external radiotherapy (N+; n=22). The functional results and local-regional control rates so far are excellent, in particular given the fact that 13 out of 40 (33%) patients had very advanced (T3/T4) tumors. Only four patients experienced a local relapse (one T2, NO [soft palate], one T3, NO [tonsillar fossa]; one T3, N3 [tonsillar fossa]; one T2, N2a [tonsillar fossa), one of which (T2, N2a [tonsillar fossa]) was salvaged. Three relapses in the neck occurred. The actuarial local control rate after brachytherapy at 4 years for all Tl-4, NO, + tumors was 89% (Figure 21.11), whereas an overall survival of 69% was observed. No difference was found for tonsillar fossa and soft palate (Figure 21.12) or fractionated HDR (n=21) versus PDR (n=19) (Figure 21.13). For a recent and more detailed analysis of local-regional control rates, the reader is referred to reference 25. As of January 2001, the treatment schedule has been slightly modified. For all T-stages (Tl-4), a dose of 4-6 Gy by external radiotherapy is used. The surdosage by brachytherapy for the primary consists of fractionated HDR 4 Gy + 4 times 3 Gy + 4 Gy (total dose 20 Gy; 2 fractions per day). Or, in the case of PDR, 2 Gy + 10 times 1 Gy + 2 Gy (total dose 22 Gy; 8 fractions per day).
Figure 21.11 Actuarial local control rate for patients with Figure 21.10 Anteroposterior X-ray film of an interstitial
carcinoma of the tonsillar fossa and soft palate, treated
implant in the tonsillar fossa and soft palate.
between 1990 and 1994 with either fractionated HDR or PDR.
Brachytherapy techniques and preliminary clinical results 307
Figure 21.12 Actuarial local control rate for patients with carcinoma of the tonsiliar fossa versus tumors of the soft palate, treated between 1990 and 1994 with either fractionated HDR or PDR.
Figure 21.13 Actuarial local control rate for patients with carcinoma of the tonsi liar fossa and/or soft palate, treated between 1990 and 1994 with fractionated HDR versus PDR brachytherapy.
213*8 Clinical results: early and late sideeffects for tonsillar fossa and soft palate With regard to acute and late side-effects, all patients were (prospectively) scored using a modified RTOG scoring system regarding skin, subcutaneous tissue, mucosa, dysphagia, salivary gland function (xerostomia), spinal cord, and pain. In summary, no significant lasting untoward acute and late side-effects were observed to be different from those of the LDR treatment era. Figure 21.14 demonstrates, as an example, the actuarial late complication-free probability of mucosal reactions grade 3 and 4 for fractionated HDR and PDR. Again, for a recent update of side-effects using this technique for tonsillar fossa and soft palate tumors, see reference 25.
Figure 21.14 Actuarial late complication-free probability of mucosal reactions grade 3 and 4 for carcinoma of the tonsillar fossa and soft palate, treated by fractionated HDR and PDR between 1990 and 1994.
radiotherapy to a dose of 60 Gy (Tl-3) or 70 Gy (T4 and/or parapharyngeal space extension). Neck nodes are generally boosted to a dose of 70 Gy [26,27]. After completion of the external radiotherapy part of the treatment, a booster dose is applied by brachytherapy. With regard to the endocavitary brachytherapy, a silicone mould, the Rotterdam Nasopharyngeal Applicator (Figure 21.15), is used for the treatment of the nasopharynx. Figure 21.16 shows schematically how this applicator is positioned. Planning of the brachytherapy treatment of the nasopharynx using the intracavitary mould is based on orthogonal radiographs. Patient points, representing tumor points and normal-tissue points, which will be used for dose optimization, are indicated on the lateral radiograph as indicated in Figure 21.17. Following our protocol, the points are transferred to the anteroposterior (AP) radiograph at a prescribed distance from the midline. Points are selected which should receive the reference dose. Usually, these are the nasopharynx points (Na) and the Rouviere node (R). The dose distribution is optimized such that these points receive the prescribed dose,
213*9 Nasopharynx: technique of endocavitary brachytherapy As reported previously, in the case of cancer of the nasopharynx we treat the primary cancer by external
Figure 21.15 The silicone Rotterdam Nasopharynx Applicator: outer diameter 5.5 mm and inner diameter 3.5 mm.
308 HDR interstitial and endocavitary brachytherapy in cancer of head and neck Figure 21.16 After decongestion (R/xylometazoline HC11%) and topical anesthesia (R/Concaine hydrochloride 7%) of the nasal mucosa and nasopharynx, guide tubes (outer diameter 2 mm) are introduced through the nose and exit through the mouth. The Rotterdam Nasopharynx Applicator (RNA) is guided intraorally over the guide tubes (GT) by pulling on the nasal part of the guide tubes. The applicator is finally placed in situ into the nasopharynx and nose. To facilitate positioning of the applicator into the nasopharynx, gently pushing the oral parts of the guide tubes intraorally using some standard type of forceps can sometimes be of additional help, as in C. By using a silicone flange, the Rotterdam Nasopharynx Applicator is secured in the correct position for the duration of the treatment, for example 3-4 days.
Figure 21.17 Schematic diagram of a lateral X-ray film of a patient with cancer of the nasopharynx and of the Rotterdam Nasopharyngeal Applicator in situ. Patient points that is, normal tissue points (CC, Pi, OC, Re, BOS, N, Pa) as well as tumor tissue points (Na, R) - are indicated. For explanation see also Table 21.6.
Brachytherapy techniques and preliminary clinical results 309
Table 21.6
Example of the planning of an endocavitary treatment of the
nasopharynx
Na-r Na-1 C R Pi OC Re-r Re-1 BOS-r BOS-1 N-r N-1 Pa-r Pa-1
104 96 30 117 22 15 18 19 43 41 140 151 86 83
100 100 26 100 23 15 18 20 42 41 142 168 84 85
101 99 27 102 19 12 13 13 40 39 102 103 75 73
The second column shows the non-optimized situation. In the third column, the dose has been optimized taking into account both nasopharynx points and the node of Rouviere. In the last column, the dose has been optimized such that both nasopharynx points, both nose points, and the node of Rouviere receive the reference dose. Na: nasopharynx; C: cord; R: Rouviere node; Pi: pituitary gland, OC: optic chiasm; Re: retina; BOS: base of skull, N: nose; Pa: palate.
i.e., an optimization on patient points [20]. If, during evaluation, the dose in specific points seems unsatisfactory, these points can be included in the optimization procedure as well. In that case, each patient point can be assigned a weighting factor for dose requirements (Table 21.6). The planning procedure is illustrated by the example mentioned in Table 21.4 [27]. Brachytherapy can be given on an outpatient basis in fractions of 3 Gy each, twice a day, with an interval of at least 6 h, using the HDR afterloader (see also Table 21.5). Currently, as of 2001, using the UICC/AJCC 1997 classification system, the protocol for the brachytherapy boost has been slightly modified. For Tl-2a tumors after external radiotherapy 60 Gy a booster dose of 4 Gy + 3 times 3 Gy + 4 Gy is given (total brachytherapy dose 17 Gy; 2 fractions per day). For good responding T2b tumors, a dose of 4 Gy + 3 Gy + 4 Gy (total dose 11 Gy; 2 fractions per day). Poorly responding T2b tumors and T3, 4 cancers are preferentially treated by Stereotactic Radiation.
213*10
nal radiotherapy only, 65 patients treated with a minimum dose of 60 Gy were recruited and analyzed for comparison purposes. Figures 21.18 and 21.19 depict the local relapse-free survival and overall survival, respectively, for both patient groups. It seems that boosting the primary tumor to a high dose is of benefit. However, there is a significant excess death rate due to intercurrent disease and/or second primaries. Moreover, comparing Figure 21.20 and Figure 21.21, we can observe that the gain in local control is mainly confined to the Tl-3 patient category. Figure 21.22 shows the overall survival
Nasopharynx: clinical results
From March 1991 until December 1994,49 patients with primary squamous cell carcinoma of the nasopharynx were treated according to the protocol combining external radiotherapy and fractionated HDR. Out of those 49 patients, ten were excluded because of re-radiation (n=5), distant metastasis at presentation (n=l), and histology other than squamous cell carcinoma (n=4). Subsequently, from a database of cancers of the head and neck treated in the DDHCC from 1978 to 1988 by exter-
Figure 21.18
Local relapse-free survival for patients treated for
Tl-4 carcinoma of the nasopharynx by ERT only (between 1978 and 1988) to a minimum dose of 60 Gy, as opposed to patients treated by ERT plus endocavitary brachytherapy (between 7997 and 1994).
310 HDR interstitial and endocavitary brachytherapy in cancer of head and neck
Figure 21.19 Overall survival for patients with Tl-4 cancers of
Figure 21.22 Overall survival for T1-3 cancers in the
the nasopharynx. (See also the legend to Figure 21.18.)
nasopharynx. (See also the legend to Figure 21.18.)
and side-effects using external radiotherapy in conjunction with the Rotterdam Nasopharynx Applicator, the reader is referred to reference 28. 21,4 INTEGRATED BRACHYTHERAPY UNIT (IBU): THE FUTURE
Figure 21.20 Local relapse-free survival for Tl-3 tumors originating in the nasopharynx. (See also the legend to Figure 21.18.)
A new development in brachytherapy is the concentration of implantation, implant reconstruction, dose planning, and irradiation in a shielded operating room, i.e., an IBU. In this way one can achieve an improvement of the implant quality and a shortening of the overall procedure as reconstruction and planning are performed immediately. Finally, it offers the possibility of applying intraoperative brachytherapy. An IBU should be equipped with an HDR afterloader and a dedicated brachytherapy localizer, consisting of an L-arm in combination with a C-arm (Figure 21.23), which is on-line connected to a treatment planning system. The design of the localizer enables one to view the implant from any chosen direction. The main features of the IBU are: 1. on-line filmless planning enabling real-time feedback of the dose distribution of the implant during the implantation procedure; 2. intraoperative irradiation using the HDR afterloader.
Figure 21.21 Local relapse-free survival for T4 cancers in the nasopharynx. (See also the legend to Figure 21.18.)
rate comparing Tl-3 patients treated by external radiotherapy alone with those treated by external radiotherapy and endocavitary brachytherapy. Except for a few patients developing synechiae, no significant difference in morbidity was observed for patients treated by external radiotherapy only compared with those treated by external radiotherapy plus brachytherapy. For a more elaborate, recent analysis of local-regional control rates
The procedure in an IBU is summarized in Figure 21.24 (taken from reference 29). After implantation of the catheters or applicators, reconstruction images are made using fluoroscopy. Both the video images from the image intensifier and the settings of the localizer, i.e., L-arm angle, C-arm angle, and the isocenter image intensifier distance, are on-line transferred to the treatment planning computer. After correction of the geometric distortions present in fluoroscopy images and, if necessary, some image processing, the geometry of the implant is reconstructed followed by dose planning and dwell-time optimization. The resulting dose distribution
Integrated brachytherapy unit: the future 311 Figure 21.23 The isocentric IBU localize? consisting of an L-arm in combination with a C-arm. The localize? is in the 'parkingposition,' i.e., rotated over 40 °, enabling maximum access to the operating table during implantation.
21*4*1 Brachytherapy of the neck in an IBU using a flexible intraoperative template
Figure 21.24 Intraoperativefilmless planning procedure in an IBU. (For explanation, see text.)
In October 1994, an IBU was installed in the DDHCC. To date, the IBU has been mainly used in attempts to improve on the quality of standard brachytherapy procedures; that is, the implant geometry is visualized (and/or changed) intraoperatively using fluoroscopy (see, for example nasal vestibule), and X-ray films for planning purposes are taken with the patient still under general anesthesia, thus reducing the 'time to start' of the actual irradiation. The next step will be research and development of routine filmless planning (see above). The intraoperative brachytherapy procedure is illustrated by the following case report. CASE REPORT: RE-IRRADIATION OF THE NECK
is presented on a monitor in the operating room for evaluation. In the case of an unsatisfactory dose distribution, one has the ability technically to optimize the implant. In other words, this real-time feedback of the resulting dose distribution during implantation allows dose optimization not only by dwell-time optimization, but also by modification of the implant geometry. Concerning the irradiation, the patient can either receive a high, single intraoperative fraction in the IBU, e.g., using a flexible intraoperative template (see next section), or a relatively small fraction (e.g., 4 Gy), to be followed postoperatively by fractionated HDR or PDR treatment, e.g. for a volume implant of the base of tongue. In some cases, the IBU system is only used for verification of the position of catheters or applicators and/or for on-line reconstruction and dose planning.
Our experience with re-irradiation in cancer of the head and neck has been previously reported [30]. A variety of tumor sites, including the neck, appeared to have been reirradiated by means of brachytherapy ± external radiotherapy ± surgery; 55% of the patients obtained tumor control of the re-irradiated site. It was concluded that reirradiation can be of benefit when pursuing long-lasting local-regional control and, in particular, the use of brachytherapy (albeit LDR at the time) was advocated. Regarding specifically the technique of implantation of the neck, after dissection the tumor bed was implanted using a variable number of catheters (usually six to nine) sutured onto the soft tissues in the neck (Figures 21.25 and 21.26). As of 1990, similar techniques have been used for recurrent (after previous external radiotherapy and/or surgery) tumors in the neck in combination with fractionated HDR brachytherapy (total dose of fractionated HDR
312 HDR interstitial and endocavitary brachytherapy in cancer of head and neck
Figure 21.25 Schematic diagram of a single-plane implant in the neck using standard afterloading catheters (outer diameter 2 mm) sutured onto the soft tissues in the neck (tumor bed). The defect in the neck is reconstructed as a one-step procedure at the time of implantation. Catheters remain in situ for a number of days (3-10), depending on factors such as the condition of the patient, fractionation schedule, and total dose.
impediment of the implant per se at the time of the reconstructive procedure, to the catheters having to remain in situ for quite a number of days, with the potential risk for infections in the area of the newly reconstructed defect in the neck, as well as the risk of wound breakdown if, for technical reasons, the catheters have to be positioned (too) close to the covering skin (high skin dose; this, by the way, might also have implications regarding the type of reconstructive procedure chosen). 21.4.2 Technical aspects of brachytherapy of the neck using a flexible intraoperative template
Figure 21.26 X-ray film of a single-plane implant tumor bed in the neck for re-irradiation by LDR brachytherapy in a patient with recurrent tumor after previous surgery and/or ERT. (See also the legend to Figure 21.25.)
54-60 Gy). In a preliminary analysis, only four out of 13 (31%) had failed at the site of the fractionated HDR implant (data not published). Although these control rates seem satisfactory given the poor population subset, the procedure is laborious and, for the reconstructive surgeon, not particularly gratifying. This is partly due to the
To eliminate the disadvantages of conventional implants of the neck by individual catheters (see Figure 21.25), a novel technique was developed in our institution. In principle, it consists of a flexible silicone template to be used as a temporary single-plane implant intraoperatively. This silicone flexible intraoperative template has a thickness of only 0.5 cm (flexible), and can be easily cut into different shapes and sizes (custom made). In the center of the template, catheters are positioned 1 cm apart in prefixed channels (allowing for fixed spacing). In predetermined pinpoint holes (1 cm apart), metallic buttons can be introduced during the brachytherapy procedure to delineate to target when using fluoroscopy (Figure 21.27). If one combines the use of a flexible intraoperative template technique with an IBU, some extra advantages can be envisaged. First, the custom-made flexible intraoperative template is, in most cases, easily positioned (Figure 21.28) and, after on-line computer planning, a single dose can be delivered intraoperatively using the HDR afterloader in the IBU with the patient still under
Integrated brachytherapy unit: the future 313
Figure 21.27 Photograph of a flexible intmoperative template with catheters and metallic buttons in situ, illustrating to some extent the flexibility in curvature and shape (silicone - easy to cut and bend) of the template. (For explanation, see text.)
Figure 21.29 Photograph of neck dissection and a flexible intraoperative template in situ (including metallic buttons to demarcate target) in a patient with recurrent sarcoma in the right neck. (For explanation, see text.)
general anesthesia. Second, after irradiation, the flexible intraoperative template is removed and reconstructive surgery can be performed without limitations due to the brachytherapy procedure per se (i.e., no remaining insitu catheters). Also, the covering skin is not at risk for too high doses of irradiation, i.e., there is less risk of wound breakdown. Finally, optimization of the dose distribution can be performed (see below).
decided to re-resect the remaining tumor mass and irradiate the tumor bed intraoperatively by means of a flexible intraoperative template (delivering a single fraction of 10 Gy at 1 cm; see also Figures 21.29 and 21.30), followed by postoperative external radiotherapy (conventional fractionation of 2 Gy/day, total dose 46 Gy). The technique of the brachytherapy procedure has been explained in section 21.4.1. The tailoring of the flexible intraoperative template itself, positioning of the template, the computer planning, and the actual irradiation took approximately 3 h. To demonstrate the advantage of optimized dose distributions using a flexible intraoperative template over a classical non-optimized LDR implant (in which catheters are sutured onto the tumor bed; see also Figures 21.25 and 21.26), selected
CASE REPORT: BRACHYTHERAPY OF THE NECK IN AN IBU USING ON-LINE PLANNING AND A FLEXIBLE INTRAOPERATIVE TEMPLATE
Recently, two patients presented with a recurrent sarcoma in the neck; the head and neck tumor board
Figure 21.28 Schematic diagram showing a flexible intraoperative template for intraoperative re-irradiation of the neck. (For explanation, see text. Compare also to Figure 21.25.)
314 HDR interstitial and endocavitary brachytherapy in cancer of head and neck
Figure 21.30 X-ray film of an implant in the neck using a flexible intraoperative template in an IBU. (See also text and the legend to Figure 21.29.)
isodose patterns are compared in Figure 21.31 (panels A-E) for different single-plane brachytherapy techniques. That is, typical examples are shown (in a central plane, I, as well as in a plane taken more in the periphery of the implant, II) for a non-optimized LDR implant (see also Figure 21.31, panel A), for an implant optimized on dose points positioned at 1 cm from catheters (see also Figure 21.31, panel B), for a dose calculation of ('standard') flexible intraoperative template (thus disregarding the possibly severe curvatures of a template in a real anatomical situation) optimized on dose points at 1 cm from the catheters with the dwell times transferred to the actual flexible intraoperative template (Figure 21.31, panel C), for an optimized flexible intraoperative template used in the actual patient with dose points at 1 cm from catheters (Figure 21.31, panel D), and, finally, for an optimized flexible intraoperative template as used for a typical situation with dose points at 1 cm in region 1 (e.g., at 'shallow depth') and dose points at 2 cm in region 2 (e.g., dose prescription to 'deeply located' dose points, for example due to anatomical constraints) (Figure 21.31, panel E). In conclusion, in general, better dose distributions can be obtained with flexible intraoperative template as opposed to classical LDR catheter implants (compare
Figure 2131 Comparison of different dose distributions for implant of the neck. Examples of isodose patterns are shown for a central plane (plane I) as well as for a plane taken more in the periphery of the implant (plane II). These dose distributions were calculated for a typical case of an LDR nonoptimized implant (panel A); optimized catheter implant with dose points at 7 cm from catheters (panel B); optimized standard flexible intraoperative template with dose points at 1 cm from the center of the implant (panel C); optimized flexible intraoperative template used in the patient in Figure 21.30, with dose points at 1 cm from catheters (panel D); and optimized flexible intraoperative implant to be used for situations with dose points at different depths, that is, with dose points at 1 cm (region 1; R-1) and 2 cm (region 2; R-2) (panel E). (For detailed explanation, see text.)
References 315
panels A and B to C in Figure 21.31). Also, a dose calculation using a standard flexible intraoperative template generates sufficiently adequate dose distributions as long as the curvature of the actual template is not too strong (compare also panel C to D in Figure 21.31). In some instances, with a flexible intraoperative template one can even divert to dose points at different levels, thereby arbitrarily dividing the implant into different 'dose point regions' (panel E). As with data concerning other applications of intraoperative irradiation such as those generated by linear accelerators, we have to await and see whether, with the use of large, single fractions of brachytherapy, we will run into severe late side-effects.
21.5 THE FUTURE OF BRACHYTHERAPY: A ROTTERDAM VIEWPOINT
vestibule (mould and interstitial brachytherapy) and tonsil and soft palate (interstitial brachytherapy). Although base of tongue and mobile tongue cancers are routinely treated in our institution by external radiotherapy in combination with fractionated HDR or PDR interstitial brachytherapy, clinical outcome has not been analyzed in detail to date. For this reason, we elected not to digress on base of tongue in this chapter, except for referring to papers and preliminary outcome data. Finally, some aspects regarding the way we are presently implementing (fractionated) HDR brachytherapy using an integrated brachytherapy unit are discussed.
REFERENCES 1. Armour, E., Wang, Z., Corry, P. and Martinez, A. (1990)
It is not within the scope of this chapter to elaborate on the future of brachytherapy. In fact, this chapter represents a comprehensive departmental Vision.' That is, given the technical means and infrastructure of the department, the authors have tried to present a 'how I do it' type of overview. Needless to say, there are many other ways to go about brachytherapy. Recently, the department expanded by purchasing a brachytherapydedicated CT scanner. With regard to head and neck cancer, in the next few years major efforts will be made to define the role of brachytherapy planning and to implement it as a routine procedure. Another interesting subject is the further extension of radiobiological models [31]. In fact, as of 1997, we routinely use fractionation schedules in head and neck brachytherapy, with the first as well as the last fraction size topped up by 1 Gy [32]. This way we hope to shorten the overall treatment time as well as to make the schedule more 'effective' [33].
Equivalence of continuous and pulse simulated low dose rate irradiation in 9Lgliosarcoma cells at 37° and 41 °C. InLJ. Radiat. Oncol. Biol. Phys., 22,109-14. 2. Brenner, D.J.and Hall, E.J. (1991) Conditions for the equivalence of continuous to pulsed low dose rate brachytherapy. IntJ. Radiat. Oncol. Biol. Phys., 20,181-90. 3. Lea, D.E. and Catcheside, D.G. (1942) The mechanism of the induction by radiation of chromosome aberrations in tradescantia.y. Genet, 44,216-45. 4. Barendsen, G.W. (1982) Dose fractionation, dose rate and iso-effect relationships for normal tissue responses. IntJ. Radiat. Oncol. Biol. Phys., 8,1981-97. 5. Thames, H.D. (1985) An 'incomplete-repair' model for survival after fractionated and continuous irradiations. IntJ. Radiat. Oncol. Biol. Phys., 47,319-39. 6. Dale, R.G. (1985) The application of the linear-quadratic dose-effect equation to fractionated and protracted radiotherapy. Br.J. Radio!., 58, 515-28. 7. Thames, H.D., Bentzen, S.M., Turesson, I., Overgaard, M. and Van den Bogaert, W. (1990) Time-dose factors in
21.6
SUMMARY
Single stepping-source, computer-controlled afterloading devices have changed classical low dose-rate radiobiological thinking as well as the application techniques per se in brachytherapy considerably. In the Head & Neck Cancer Cooperative Group of the University Hospital Rotterdam/Daniel den Hoed Cancer Center, fractionated HDR or PDR brachytherapy schedules have been initiated and clinically tested since August, 1990. In total, from August 1990 until December 1994, 176 patients were treated by either fractionated HDR or PDR. This chapter reviews our brachytherapy rationale, techniques, and dose specification methods as well as the clinical (brachytherapy) protocols in use. The first clinical results in terms of local tumor control and complications (acute and late side-effects) are reported for cancer of the nasopharynx (endocavitary brachytherapy), nasal
radiotherapy: a review of the human data. Radiother. On<:o/.,19,219-35. 8. Fowler, J.F. (1989) The linear-quadratic formula and progress in fractionated radiotherapy. Br.J. Radiol., 62, 679-94. 9. Pop, L.A.M., van den Broek, J.F.C.M. and van der Kogel A.J. (1994) Clinical considerations in the design of new treatment strategies for HDR- and PDR-brachytherapy as an alternative for LDR-brachytherapy. Radiother. Oncol., 31,(Suppl.1),S38. 10. de Boer, R.W. and LebesqueJ.V. (1991) Equivalent doses of low-dose-rate brachytherapy and pulsed-dose-rate brachytherapy. First ESTRO Biennial Meeting on Physics in Clinical Radiotherapy, (Budapest, October 1991, abstract 14).
11. Ang, K.K., Guttenberger, R., Thames, H.D., Stephens, L.C., Smith, C.D. and Feng, Y. (1992) Impact of spinal cord repair kinetics on the practice of altered fractionation schedules. Radiother. Oncol., 25,287-94.
316 HDR interstitial and endocavitary brachytherapy in cancer of head and neck Veenendaal, The Netherlands, Nucletron International BV, 325-44.
12. Turesson, I. and Thames, H.D. (1989) Repair capacity and kinetics of human skin during fractionated radiotherapy: erythema, desquamation and telangiectasia after 3 and 5
24. Levendag, P.C., Vikram, B., Flores, A.D. and Yin, W.B. (1994) High dose rate brachytherapy for cancer of the
years' follow-up. Radiother. Oncol., 15,169-88.
head and neck. In High Dose Rate Brachytherapy: a
13. van Rongen, E., Thames, H.D. and Travis, E.L (1993)
Textbook, ed. S. Nag. Amonk, NY, Futura Publishing
Recovery from radiation damage in mouse lung: interpretations in terms of two rates of repair. Radial Res., 133,225-33. 14. Fowler, J.F. (1993) Why shorter half-times of repair lead to greater damage in pulsed brachytherapy. Int.}. Radiat. Oncol. Biol. Phys., 26,353-56. 15. Turesson, I. (1990) Radiobiological aspects of continuous low-dose-rate irradiation and fractionated high-dose-rate
Company, 237-73. 25. Levendag, PC., Schmitz, P.I.M.Jansen, P.P.etal. (1997) Fractionated high-dose-rate brachytherapy: first clinical experience in squamous cell carcinoma of the tonsillar fossa and soft palate. Int.J. Radiat. Oncol. Biol. Phys., 38(3), 497-506. 26. Levendag, P.C., Visser, A.G., Kolkman-Deurloo, I.K.K., Eijkenboom, W.M.H. and Meeuwis, C.A. (1994) HDR
irradiation. Radiother. Oncol., 19,1-16.
brachytherapy with special reference to cancer of the
16. van den Aardweg, G.J.M.J. and Hopewell, J.W. (1992) The kinetics of repair for sublethal radiation-induced damage
nasopharynx. In Brachytherapy from Radium to
in the pig epidermis: an interpretation based on a fast
Optimization, ed. R.F. Mould, J.J. Battermann, A.A.
and a slow component of repair. Radiother. Oncol., 23, 94-104. 17. Levendag, P.C., Meeuwis, C.A., Wijthoff, S.J.M. and Visser,
27.
Martinez and B.L. Speiser. Veenendaal, The Netherlands, Nucletron International BV, 121-31. Levendag, P.C., Peter, R., Meeuwis, C.A., Visch, L.L.,
A.G. (1992) Reirradiation of recurrent head and neck
Sipkema, D., de Pan, C. and Schmitz, P.I.M. (1997) A new
cancers: external versus interstitial radiation therapy.
applicator design for endocavitary brachytherapy of cancer in the nasopharynx. Radiother. Oncol., 45,95-8.
Activity, 3,32-9. 18. Kolkman-Deurloo, I.K.K., Visser, A.G., Niel, C.G.J.H.,
28.
Levendag, P.C., Schmitz, P.I., Jansen, P.P. etal. (1998)
Driver, N. and Levendag, P.C. (1994) Optimization of
Fractionated high-dose-rate brachytherapy in primary
interstitial volume implants. Radiother. Oncol., 31,
carcinoma of the nasopharynx./ Clin. Oncol., 16(6),
229-39. 19. Thomadsen, B.R., Houdek, P.V., van der Laarse, R., Edmundson, G.K., Kolkman-Deurloo, I.K.K. and Visser, A.G. (1994) Treatment planningand optimization. In HDR
2213-20. 29. Kolkman-Deurloo, I.K.K., Visser, A.G., Idzes, M.H.A. and
Brachytherapy: A textbook, ed. S. Nag. Amonk, NY, Futura
operative brachytherapy. Radiother. Oncol., 44,73-81. 30. Levendag, PC., Meeuwis, C.A. and Visser, A.G. (1992)
Publishing Company, 79-145. 20. van der Laarse, R., Edmundson, G.K., Luthmann, R.W. and
Levendag, P.C. (1997) Reconstruction accuracy of a dedicated localiser for filmless planning in intra-
Reirradiation of recurrent head and neck cancers: external and/or interstitial radiation therapy. Radiother.
Prins, T.P.E. (1991) Optimization of HDR brachytherapy
Oncol., 23,6^5.
dose distributions. Activity, Selectron Brachyther.J., 5(2), 94-101. 21. Edmundson, G.K. (1991) Geometry based optimization for stepping source \mp\ants.Activity, Selectron Brachyther.J., 5(4), 22-6. 22. Edmundson, G.K. (1994) Volume optimization: an American viewpoint. In Brachytherapy from Radium to
31. Visser, A.G., van den Aardweg, G.J.M.J. and Levendag, P.C. (1996) Pulsed-dose-rate and fractionated high-dose-rate brachytherapy: choice of brachytherapy schedules to replace LDR treatments. Int.J. Radiat. Oncol. Biol. Phys., 32.
34(2), 497-505. Levendag, P.C. and Jansen, P.P. (1997) Fractionated high-
Optimization, ed. R.F. Mould, J.J. Battermann, A.A.
dose rate brachytherapy in cancer of the head and neck.
Martinez and B.L. Speiser. Veenendaal, The Netherlands,
9th International Brachytherapy Conference, Palm Springs, USA, September 3-6,1997.
Nucletron International BV, 314-18. 23. Levendag, P.C. and van Putten, W.LJ. (1990) Brachytherapy in head and neck cancer: Rotterdam low dose rate experience. In Brachytherapy HDR and LDR, ed. A.A., Martinez, C.G. Orton and R.F. Mould.
33.
Brenner, D.J., Hall, E.J., Huang, Y. and Sachs, R.K. (1994) Optimizing the time course of brachytherapy and other accelerated radiotherapeutic protocols. Int.J. Radiat. Oncol. Biol. Phys., 29(4), 893-901.
22 Brachytherapy in the treatment of pancreas and bile duct cancer DATTATREYUDU NORI, SUHRID PARIKH, SRINATH SUNDARARAMAN AND MARGOT HEFFERNAN
22.1
INTRODUCTION
Although radical surgery can cure patients with pancreatic or biliary tract carcinoma, few patients present with resectable disease. The majority of patients with pancreatic carcinoma have lesions in the head of the pancreas that are locally advanced at the time of diagnosis or are technically unresectable because of local extension to adjacent structures. Surgical options for patients with unresectable lesions are limited to palliative biliary bypass alone or in combination with elective gastroenterostomy. Similarly, surgical resection in the very small minority of resectable candidates and palliative stenting in unresectable cases have been the only treatment considerations for bile duct carcinoma. However, over the past decade, radiation therapy has been used more frequently in locally advanced pancreatic and biliary carcinomas. While the initial intent of radiotherapy was palliative in most cases, there is growing evidence that dose escalation - using a combined approach of external-beam radiation plus conformal brachytherapy - can actually improve survival in a select group of these inoperable patients. This chapter highlights the role of brachytherapy in the treatment of (a) pancreatic carcinoma, and; (b) bile duct carcinoma.
22.2
PANCREATIC CANCER
The incidence of pancreatic carcinoma appears to be increasing, though it is unknown whether this is an out-
come of poor previous reporting or new influences related to the environment or genetic predisposition [ 1 ]. Today, carcinoma of the pancreas accounts for 3% of the annual cancer incidence in the USA and 5% of cancer deaths. This translates into 27000 new cases (13000 males and 14000 females) each year. However, despite the advances in diagnosis, staging, and therapy, more than 90% of these patients die within a year, with an annual mortality of 25900 cases (12400 males and 13 500 females). Less than 2% of the patients are alive at 5 years [2,3].
22.2.1
Pretreatment assessment
At the time of diagnosis, more than 60% of the tumors have extended beyond the pancreas, with at least 20% of patients having clinically evident distant metastases [4,5]. A significant number of patients do exhibit the classic triad of symptoms associated with cancer of the pancreas - pain, weight loss, and progressive jaundice but these usually imply advanced disease. Back pain is a particularly ominous sign because it signifies infiltration of the celiac plexus - these patients are usually unresectable. It is unfortunate that, early in the course of the disease, when it is more likely to be resectable, the clinical picture is generally ill-defined and vague [6-8]. The traditional approach to the patient with non-metastatic pancreatic carcinoma was a surgical exploration. Preoperative staging was of limited importance because surgery provided the opportunity to obtain a
318 Brachytherapy in the treatment of pancreas and bile duct cancer
pathological diagnosis, assess resectability, and offer preemptive surgical palliation (4). However, a relatively small number of patients are truly resectable (< 20%), and palliation can often be easily achieved by endoscopic or other non-operative approaches. Also, a multimodality approach is playing a greater role in the treatment of these patients. Thus, a proper staging of the disease extent prior to laparotomy is essential, not only to spare a large number of patients a needless exploration, but also to identify patients suitable for these multidisciplinary approaches. Imaging tests of the pancreas are considered the cornerstone of the diagnosis of pancreatic carcinoma to assess operability and resectability [2,9]. A high-resolution, spiral, computerized tomography (CT) with intravenous contrast, employing a 'pancreatic protocol', is usually adequate in delineating the loco-regional anatomy to allow a preoperative selection of potentially resectable patients. CT-portography may be added if there are doubts about the invasion of the major vascular structures around the pancreas. A pre-laparotomy laparoscopy (with a peritoneal lavage) is very useful in diagnosing peritoneal/omental seeding below the resolution of the CT scan. Selected cases may also require one or more of the following: ultrasonography, endoscopic retrograde pancreatography, endoscopic ultrasonography, or celiac angiography. The current staging system is based on the AJCC staging classification (Table 22.1). Table 22.1 AJCC staging classification for cancer of the exocrine pancreas T TX TO T1 T1a T1 b T2 T3
Primary tumor Primary tumor cannot be assessed No evidence of primary tumor Tumor limited to pancreas Tumor 2 cm or less in greatest dimension Tumor more than 2 cm in greatest dimension Tumor extends directly to any of the following: duodenum, bile duct, peri pancreatic tissues Tumor extends directly to any of the following: stomach, spleen, colon, adjacent large vessels
N NX NO N1
Regional lymph nodes Regional lymph nodes cannot be assessed No regional lymph node metastases Regional lymph node metastases
M MX MO M1
Distant metastases Presence of distant metastases cannot be assessed No distant metastases Distant metastases
Stage grouping Stage 1 Stage II Stage III Stage IV
T1 T2 T3 AnyT AnyT
NO NO NO N1 AnyN
MO MO MO MO M1
22*2.2 General management (treatment methods) SURGERY Although surgical resection employing a Whipple's procedure (pancreatico-duodenectomy) is the standard treatment for patients with pancreatic carcinoma, 5-year survival rates have remained dismal (5-year survival rates are rarely consistently more than 5%); morbidity and mortality also remain high [10-17]. Furthermore, even in the resected cases, up to 85% fail locally [18,19], and local recurrence is the ultimate cause of death in almost 50% of patients [13,20,21,86]. Because less than 20% of all patients have resectable disease, the majority of patients who are explored may not be candidates for resection. On the other hand, almost 40% of patients have locally advanced disease, in the absence of obvious distant metastases. Standard postoperative regimens have done little to improve the outcome for these patients. Thus, these patients with potentially limited disease merit consideration for innovative approaches, including intraoperative brachytherapy. CHEMOTHERAPY
5-Fluorouracil (5-FU) used to be the only agent with any demonstrated responses in pancreatic cancer, but it does not confer any survival benefit. Combination chemotherapy may produce more objective tumor responses, but does not appear to prolong median survival beyond that expected with single-agent regimens [22]. The arrival of gemcytabine on the chemotherapy scene has raised new hopes for these patients. In one randomized study comparing 5-FU with gemcytabine in previously untreated patients, gemcytabine was associated with a significantly better response rate and median survival as well as palliation of disease-related symptoms (pain, weight loss, performance status) [23]. RADIATION THERAPY
Attempts at definitive external-beam radiotherapy have met with little success. Due to the high incidence of local failure in these patients, attempts have been made to escalate the radiation dose. A major deterrent to this is the close proximity of dose-limiting structures such as the liver, duodenum, stomach, kidney, and spinal cord. Although recent developments in three-dimensional radiation therapy allow the delivery of higher doses to the tumor volume while minimizing radiation damage to the surrounding structures [24], local control is still a major problem and the survival benefit is uncertain. The role of conventional external-beam radiation therapy in pancreatic carcinoma is essentially limited to palliation of pain; an occasional patient with localized, but unresectable, disease may obtain a survival benefit. Combined modality treatment using external radiation therapy and 5-FU has been
Pancreatic cancer 319
demonstrated to increase median survival rates in patients with locally unresectable disease. The Gastrointestinal Tumor Study Group (GITSG) randomized trial, which treated inoperable patients with a curative intent, provided the first definite evidence that chemoradiation could actually prolong survival in this otherwise fatal disease. However, even in that trial, the median survival was only 35 weeks with combined chemoradiation [25]. There were no long-term survivors. Local failure still remains a major problem [26]. INTRAOPERATIVE RADIATION THERAPY
ability to conform and sharply limit the radiation dose to the implanted area, thus allowing delivery of a very large radiation boost to the site of gross disease. Brachytherapy in pancreatic cancer has been implemented using several different approaches (Table 22.2). This chapter focuses exclusively on intraoperative interstitial implants. This approach can be utilized both in unresectable and in resected carcinomas of the pancreas and these intraoperative implants can be combined with postoperative standard external-beam radiation and/or chemotherapy. INTRAOPERATIVE BRACHYTHERAPY
Several institutions have studied the role of intraoperative electrons to boost the dose to the tumor bed; in this setting, a single large fraction of radiation can be delivered in the operating room, while retracting all the vital structures away from the radiation field [26-28]. The results suggest a definite impact on local control, but overall survival is not affected. Because of the wide range of intraoperative radiation therapy (IORT) doses and the adjuvant therapies that have been given with it, interpretation of results is difficult [28]. Thus, for patients with resectable disease and especially those with unresectable disease, the current methods of treatment have had little impact on median survival in the last 15 years. However, the available information from external-beam dose escalation protocols and the experience with IORT suggest that pancreatic cancer does exhibit a dose-response relationship with radiation. Administration of higher doses can improve local control, and this may translate into better survival. Data also suggest that the concomitant administration of chemotherapy and high-dose radiation increases median survival. Interstitial brachytherapy offers us the ideal avenue for maximal dose escalation. The concept of brachytherapy in pancreatic cancer is not new. Handley first performed an interstitial pancreatic implant back in 1934 [29] and is quoted as having said'. .. a surgeon who, on opening the abdomen, finds an irremovable cancer is not doing his duty to the patient unless he subjects the growth to interstitial irradiation' [30]. One of the major advantages of brachytherapy is the
Intraoperative brachytherapy is a radiation technique that delivers a high local dose to the tumor volume while sparing normal surrounding tissue. Theoretically, the application of interstitial brachytherapy to pancreatic cancer should offer a better chance for loco-regional control [31] because: (1) the physical dose distribution favors a higher total dose delivered at a higher dose rate to the tumor than to the adjacent normal tissues; (2) biologically, the delivery of a continuous therapeutic dose over a defined interval of time may offer a therapeutic advantage [32]; (3) it allows for the delivery of a relatively higher dose to the center of the implant where more resistant hypoxic tumor cells may exist; and (4) brachytherapy can be combined with external radiation with no additional morbidity. Indications for intraoperative brachytherapy
Basic radiotherapeutic principles dictate that candidates for intraoperative brachytherapy have pancreatic tumors that are unresectable but localized, with no evidence of metastatic liver, omental or peritoneal disease. The primary tumor should have minimal, if any, peripancreatic extension, to allow for the inclusion of all the known tumor volume within the implant. If regional nodal metastases are present and an implant is feasible, an interstitial implant may be done in order to relieve pain, but the intent would clearly be palliative. A histologic diagnosis by frozen section should be available. Patients with portal hypertension and tumors larger than 6 cm. are not candidates for intraoperative brachytherapy.
Table 22.2 Brachytherapy approaches employed in pancreatic cancer
Intraoperative externalbeam therapy Orthovoltage Electron beam Intraoperative high dose-rate brachytherapy HAM applicator Catheter based
Via naso-pancreatictube for selected cases of ampullary cancers
HAM = Harrison-Anderson-Mick; USG = ultrasound.
Interstitial Intraoperative lodine-125 Palladium-103 Percutaneous CT guided USG guided Infusional
320 Brachytherapy in the treatment of pancreas and bile duct cancer
22*23
Available isotopes
The suitability of a radionuclide for interstitial brachytherapy depends on its half-life, its photon energy spectrum, and the number of photons it produces per decay (Table 22.3). Radionuclides with short half-lives are advantageous for permanent brachytherapy applications in pancreatic tumors because the rapid dose delivery will aid in controlling fast-growing tumors. The radiation hazard from the patient will also decrease more rapidly [33]. Choice of photon energy determines the penetrating ability of radiation in tissue [34]. Photon energies from isotope sources have a wide range, from 20 to 1060 KeV (Table 22.3). Gold-198 was one of the earlier isotopes used, as a substitute for radon [35]. Its short half-life and higher photon energy increase the radiation exposure to hospital personnel and result in excessive radiation doses to the surrounding normal structures. We do not consider this as a suitable radionuclide for pancreatic implants today. Iodine-125 was the preferred isotope for several years [36-42]. This was primarily because of its low photon energy, ranging from 27 to 35 KeV [36], which minimizes the radiation to any of the surrounding vital structures. The major disadvantage of iodine-125 as a permanent implant isotope is the low dose rate, of about 7-8 cGy tf1 (as a result of a long half-life of 60 days coupled with sources of low specific activity). This is obviously a disadvantage when dealing with aggressive, rapidly growing tumors wherein tumor proliferation can easily outstrip radiation-induced cell kill. Palladium-103 is a new isotope developed specifically to address this issue. Its half-life of 17 days with a high specific activity yields dose rates of 18-20 cGy h~'. At the same time, the energy spectrum from palladium-103, of 20-35 KeV [43,44], is low enough to spare the surrounding sensitive and dose-limiting structures, like the stomach, small bowel, anastamotic sites, etc. These properties make it more favorable in the treatment of rapidly growing tumors such as pancreatic carcinoma because most of the dose radiation is delivered over a short period (8 weeks) and prolonged radiation exposure of adjacent normal tissue is minimized. We have employed palladium-103 as a substitute for iodine-125 in intraoperative brachytherapy of pancreatic carcinoma in a phase I-II study [44]. The findings of this study are discussed under the results section.
Treatment planning and technique
Intraoperative permanent interstitial implants, in which radioactive sources are permanently placed in the tissue, are the preferred approach because of the relative simplicity and easy applicability to deep-seated tumors such as pancreatic cancers. Temporary implantation in this location is associated with a number of technical difficulties related to proper catheter placement and subsequent removal, and is not recommended in this setting. The successful outcome of a brachytherapy treatment depends upon careful preoperative planning and a precise implementation of the technique. An accurate delineation of the tumor volume and proper spacing of sources are essential to avoid overdosage and necrosis within the tumor volume or 'cold spots' and consequent underdosage of the tumor. The basic permanent interstitial technique is used for tumors of the pancreas (Figure 22. la-f) [45]. It consists of two steps: the insertion of unloaded needles and the sub-
Figure 22.1 a A CTscan of the abdomen in a patient with localized unresectable carcinoma in the head of the pancreas. This tumor was unresectable because of superior mesenteric artery invasion.
Table 22.3 Isotopes for permanent implants in pancreas rnrrinnmn
Gold-198 lodine-125 Palladium-103
2.7 60.2 17.0
0.412 0.028 average 0.021 average
2.5 0.025 0.008
Figure 22.1 b Mobilization of the pancreas and bimanual examination. Also shown is retraction of stomach by the metal retractor.
Pancreatic cancer 321
Figure 22.1 c Insertion of a 75 cm long afterloading needle into the head of the pancreas.
Figure 22.1 e Mobilization of the head of the pancreas with insertion of afterloading 16-gauge needles for insertion of radioactive pellets.
Figure 22.1 d Measuring the depth-wise dimensions of the
Figure 22.1f A postoperative CT scan showing optimal
tumor with the help of a metal ruler.
placement of radioactive pellets into the head of the pancreas.
sequent afterloading with radioactive isotope. An exploratory laparotomy allows an accurate determination of the extent of disease, making possible a decision as to resectability or the alternative option of permanent interstitial implant with iodine-125 or palladium sources. Satisfactory tumor exposure is essential to a good implant. Mobilization of the stomach and upward retraction will expose the anterior surface of the pancreas, allowing the proper insertion of the unloaded needles. In lesions of the pancreatic head, mobilization of the duodenum (Kocher's maneuver) will aid in the determination of the posterior extent of the tumor. The tumor dimensions are measured in X, Y, and Z directions - the length and width are measured using caliper, while the anteroposterior dimension is estimated by carefully placing a 15 cm long, 17 gauge needle into the tumor and seeing how much of the needle still protrudes above the pancreas. The total activity to be implanted is based on the average dimension (X+Y + Z -5- 3) of the region to be treated. A minimum peripheral dose of 160 Gy is prescribed using iodine-125 sources, while 110 Gy is prescribed for palladium-103 (because of its shorter half-life, and higher dose rate). The total activity
and number of radioactive sources required to deliver this dose are determined from a nomograph (Figures 22.2 and 22.3), based on the estimated average dimension of the tumor. The nomograph also gives information on the total number of afterloading needles required to place the required number of sources and, thus, the source spacing. The predetermined number of needles are inserted into the pancreatic tumor as parallel to each other as possible, usually with an interneedle spacing of 1.0-1.5 cm, and the sources are introduced into the needles with the help of a permanent implantation applicator. The radioactive seeds should be placed at least 0.5-1.0 cm distant from the surface of the pancreas in order to avoid radiation damage to the adjacent stomach and duodenum. A segment of omentum should be positioned over the implanted tumor to increase the distance between the implanted seeds and the bowel and to prevent leakage of pancreatic fluid. Small surgical hemoclips should be placed on the anterior surface of the pancreatic tumor to define the region on future localization films for dosimetry evaluation, as well as to help delineate the target volume for any planned external-beam
322 Brachytherapy in the treatment of pancreas and bile duct cancer
therapy. Stereo shift or orthogonal films are taken when the patient becomes ambulatory, usually on postoperative day 3 or 4 (Figures 22.Ig and 22.4), isodose contours are generated, and the minimum and maximum tumor doses can be estimated.
Postoperative external-beam radiotherapy to the primary and regional lymphatics may be added, with or without chemotherapy. This is usually instituted 3-6 weeks following surgery, to allow for complete recovery from the operation. External-beam therapy is usually Figure 22.1 g A postoperative CT scan showing the combined dose distribution from permanent implant and external-beam radiation to 4500 Gy in 5 weeks. In this example the head of the pancreas and peripancreatic areas received a combination dose (external beam and implant) between 9000 and 10 000 cGy, with a significantly lower dose given to the adjacent normal structures.
Figure 22.2 Nomograph for palladium-103.
Pancreatic cancer 323
Figure 22.3 Nomograph for iodine-125.
delivered through a three-field or four-field plan, delivering a dose of 4500-5040 cGy in 180 cGy fractions, and 5-FU ± mitomycin-C may be added in the first and last week of radiotherapy. RESULTS
Borgelt reported on the experience at the M.D. Anderson Cancer Center using gold grain implants for
Figure 22.4 Complications as a function of isotope used. The complications are significantly higher for gold-198 (Au-198) compared with iodine-125 (1-125), and least with palladium-103 (Pd-103).
unresectable pancreatic cancer [46]. A median survival of 8 months was reported. Major morbidity (26%) and mortality (13%) were very high. The author ascribed the high incidence of complications to the dosimetric characteristics of gold-198, with the higher photon energy and excessive radiation dose to adjacent normal tissues being the main causes of the postoperative complications. In an earlier study, Peretz et al. reported on 98 patients with biopsy-proven unresectable adenocarcinoma of the pancreas treated with intraoperative iodine-125 implants [37]. Thirty patients had Tl NO MO disease, 47 patients had T2-3 NO MO disease, and 21 patients had significant regional lymph node involvement (Tl-3 Nl MO). Of the 57 patients who presented with pain, 37 (65%) were free of pain following the implant. Sixty-two patients had one or more follow-up localization films to assess tumor response; a 30% or greater reduction in the tumor size was seen in 28 (45%). A multivariate analysis showed that four factors significantly affected survival: T stage (p = 0.002), N stage (p = 0.017), administration of chemotherapy (p = 0.002), and greater than 30% reduction in the size of the implant volume on follow-up films (p = 0.03). Whereas the median survival for the entire
324 Brachytherapy in the treatment of pancreas and bile duct cancer
group was 7 months, a subgroup of patients with Tl NO stage disease who received postoperative chemotherapy survived 18.5 months. This advantage of post-implant chemotherapy has also been reported by other groups [47]. Thus, the cumulative experience with pancreatic implants using iodine-125 sources for unresectable pancreatic tumors demonstrates a significant palliation of symptoms, with circumstantial evidence of local control of the primary tumor in a high proportion of the patients (as evidenced by a reduction in the implant volume on follow-up CT scans). Post-implant chemotherapy enhances the final outcomes and, in a select subset of patients, this multimodality approach may even prolong survival (the 18.5 months' median survival compares favorably to several surgical series with little of the attendant morbidity). As discussed above, one of the disadvantages of iodine-125 is its long half-life time and the consequent very low dose rate of radiation. Palladium-103 overcomes this major disadvantage by virtue of its short half-life (17 days); this permits the use of higher total activities, with higher dose rates for a given total dose. To evaluate this clinically, we implemented a phase I/II trial to study the feasibility, toxicity, and palliative efficacy of palladium-103 as an intraoperative, interstitial isotope for localized but unresectable pancreatic adenocarcinoma [44]. Fifteen patients with biopsy-proven unresectable adenocarcinoma of the pancreas were implanted with interstitial palladium-103 during laparotomy. The technique of implantation and the instrumentation are the same as with iodine-125. The palladium-103 nomograph assures the minimum peripheral dose of 11 000 cGy. In addition, all patients underwent biliary and gastric bypass. A mean of 45 palladium-103 sources was implanted; the mean total activity (to obtain a minimum peripheral dose of 11000 cGy) was 68.9 mCi. All patients received postoperative external-beam radiation (4500 cGy in 5 weeks). This combined treatment, consisting of intraoperative brachytherapy using palladium-103 and postoperative external-beam radiation, was well tolerated in all patients. There was no treatment-related mortality, and no serious complications such as bleeding or fistula formation. Pain relief was obtained within 3-6 weeks in ten out of 12 patients presenting with pain. Survival ranged from 6 to 24 months, with a median survival of 10 months. The absence of complications is particularly gratifying considering other reports in the literature [48]; we believe that appropriate patient selection and meticulous attention to technical details can help minimize the complication rate in this rather challenging anatomic site. This study suggests that palladium-103 can be considered an alternative source to iodine-125 for interstitial brachytherapy for unresectable carcinoma of the pan-
creas. Because of its shorter half-life, palladium-103 can be implanted at higher total activities. The surrounding normal tissue, such as the stomach, small bowel, and anastamotic sites, is spared from radiation reactions more effectively because of the low gamma ray energy spectrum of palladium-103 as compared to that of iodine-125. This study also suggests a faster rate of pain palliation compared to the iodine-125 series, although median survival rates are similar to those observed in patients implanted with iodine-125. Because of the phase I/II nature of the study, a particularly unfavorable group was selected for this protocol (Figure 22.5), making interpretation of survival data difficult. However, one patient with biopsy-proven unresectable pancreatic adenocarcinoma did survive almost 5 years; at 3.5 years post-implant, he had a second laparotomy for adhesive small bowel obstruction, and biopsies of his primary site were completely free of disease [49]. The patient eventually succumbed to disseminated intra-abdominal disease.
Figure 22.5 Comparative median survival data shown in bar graphs for different isotopes used for unresectable carcinoma compared with surgical results for resectable carcinoma of the pancreas.
22.2A Summary/conclusions It is apparent that standard therapies fail to control the local progression of disease in patients with unresectable pancreatic carcinoma; this may be a significant factor in their ultimate death. Intraoperative brachytherapy offers the potential of maximal dose escalation in these patients; significant tumor responses have been reported in a number of clinical series to date [37,40,47]. Current data indicate that this procedure is most effective in a subgroup of patients with early-stage, unresectable disease, especially when coupled with postoperative systemic chemotherapy. Thus, intraoperative brachytherapy, when properly implemented, can deliver a high radiation dose and achieve faster pain palliation and better local control without significantly increasing complications or operative mortality.
Bile duct carcinoma 325
223
BILE DUCT CARCINOMA
Biliary duct carcinoma is a relatively rare disease in the USA and the Western hemisphere. Approximately 4000 new cases are reported in the USA each year and less than 10% of these achieve a prolonged 'cure' [50].
223.1
Table 22.4 AJCC staging classification for cancer of the extra hepatic biliary duct cancers T TX Tis Tl T2 T3
Pretreatment assessment
Patients present with painless obstructive jaundice, weight loss, pruritus, and a Courvoisier gallbladder. Positive physical findings are few in the early stages of the disease, and a palpable right upper quadrant mass, a periumbilical mass, or a rectal shelf usually indicate latestage disease [51]. The majority of patients with bile duct carcinoma will have elevations of bilirubin and alkaline phosphatase. Although many diseases can mimic bile duct carcinoma, the distinction between bile duct carcinoma and another process can be made on the basis of clinical presentation and results of imaging studies such as ultrasonography, high-resolution spiral CT, magnetic resonance imaging (MRI), percutaneous transhepatic cholangiography (PTCA), and endoscopic retrograde cholangiopancreaticography (ERCP) [52]. STAGING
These tumors pose a therapeutic challenge, because they are often multicentric, or can spread extensively along the ductal system by contiguity, placing the entire biliary tree at risk. The walls of the biliary tree are very thin, and the outer layers have a rich lymphatic system. This probably accounts for the fact that the majority of the patients present with locally advanced disease, with adherence to, or encasement of, the periductal vascular structures, and/or regional lymphadenopathy [53]; less than 20% of patients with bile duct carcinoma are candidates for a potentially curative resection [54]. However, hematogenous dissemination with distant metastases is uncommon, at least at presentation. The staging is based on the AJCC/UICC classification system (Table 22.4).
Primary Tumor Primary tumor cannot be assessed Carcinoma in situ Tumor invades mucosa or muscle layer Tumor invades perimuscular connective tissue Tumor invades adjacent structures: liver, pancreas, duodenum, gallbladder, colon, stomach
N NX NO N1 N1A
Regional lymph nodes Regional lymph nodes cannot be assessed No regional lymph node metastasis Regional lymph node metastasis Metastasis in cystic duct, pericholedochal, and/or hilar lymph nodes (i.e., in the hepatoduodenal ligament) N1B Metastasis in peripancreatic (head only), periduodenal, periportal, celiac, and/or superior mesenteric lymph nodes
M MX MO M1
Metastasis Metastasis cannot be assessed No metastasis Metastasis present
Stage grouping Stage I Stage 1 1 Stage I II Stage IVA Stage IVB
T1 T2 T3 AnyT AnyT
NO NO NO N1 Any N
MO MO MO MO M1
from 7% to 44%, especially with extended hepatic resections (54). Hence, most of the patients are managed with palliative procedures, such as stenting, with an occasional patient undergoing a palliative local or limited resection [54]. Even after curative resection, the local recurrence rate is high and the mean survival is 16-20 months [56]. The majority of patients die from uncontrolled loco-regional disease [57] with biliary obstruction, sepsis, and liver failure. CHEMOTHERAPY
223.2
General management
SURGERY
A complete surgical resection with pathologically negative margins probably offers the only hope for cure in these patients. However, only about 10-20% of the patients are operable at presentation and, of these, 40-50% cannot undergo a curative resection at the time of laparotomy, due to invasion of adjacent major blood vessels, the presence of extensive intraductal involvement, peritoneal seeding, or regional lymph node spread [51,53,55]. Operative mortality is high; reports range
Few chemotherapeutic agents show significant activity against biliary carcinomas. Although the reason for this lack of sensitivity is unknown, it may be due to inherent cellular resistance to the available drugs or poor delivery of the chemotherapeutic agent to the tumor due to intense peritumoral fibrosis [57]. The relatively small patient population may also account for the paucity of chemotherapeutic protocols in biliary cancer, when compared to other more common gastrointestinal malignancies [54]. Investigators have had the broadest experience with 5-FU, which has response rates of approximately 14% [58-60]. Single-agent activity has been noted with other drugs, such as adriamycin, but
326 Brachytherapy in the treatment of pancreas and bile duct cancer
clinical results have been disappointing [54]. Currently, no combination regimen has proven sufficiently encouraging to become established therapy [57]. RADIATION THERAPY
Given the overwhelming problem of loco-regional disease, radiation therapy would be the logical modality for these patients. However, despite lack of good supporting data, bile duct carcinomas have been traditionally considered radioresistant. It is only since the mid 1970s that a number of reports have appeared in the literature indicating that radiotherapy has a useful palliative role and may even prolong survival in selected patients with unresectable disease [57]. As with pancreatic cancer, however, the proximity of the liver, duodenum, stomach, spinal cord, and kidney limit the dose that can be safely delivered using conventional approaches.
as a radio-sensitizing agent may further improve results. In 1990, we reported our experience with 5-FU chemotherapy in combination with radiation therapy (50 Gy to the tumor bed and lymph nodes); most patients received a low dose-rate intraluminal boost of 15-20 Gy. This treatment was well tolerated and the overall 3-year actuarial survival was 50% [55]. However, the use of low dose-rate (LDR) brachytherapy was associated with logistical, technical, and significant radiation exposure problems. The feasibility and morbidity of external-beam radiotherapy with or without intraluminal irradiation have been reported in a series of 38 patients [83]. It was concluded that accelerated external-beam radiotherapy with or without intraluminal treatment was feasible and associated with acceptable toxicity. BRACHYTHERAPY
INTRAOPERATIVE RADIATION THERAPY
The application of IORT, which allows a substantive, single dose of radiation to be delivered to the tumor exposed at laparotomy, has been used alone and in combination with external-beam radiation therapy in the treatment of bile duct tumors [61], Damage to the small bowel is minimized by the strategic use of retractors, while the dose profile of the electron beam spares the deeper tissues such as the kidneys and spinal cord. Various doses of IORT have been employed in different series [61-65]. There is a suggestion of slight survival benefit in some series [63]; clinical estimates of tumor control may, however, be overestimated [51], as there is no pathological confirmation of local disease eradication, and the majority of the patients still succumb to locoregional disease. The other major limitation of this technique lies in the radiobiology of large-fraction radiotherapy. The single large dose of radiation is associated with a high risk of late normal-tissue damage; fibrosis of the biliary duct can occur after single doses of 15 Gy and doses greater than 30-40 Gy can cause secondary biliary stenosis. Major mucosal edema followed by stenosis and even perforation will occur at 30 Gy after 6 weeks [57]. A variety of other severe side-effects, such as duodenal ulceration and hemorrhage, have also been described. Iwasaki et al. reported four fatal complications in a series of 19 patients as a result of this treatment [62]. MULTIMODALITY TREATMENT
Progress in the treatment of bile duct carcinomas is anticipated from a multimodality approach. Because loco-regional disease seems to dominate the clinical picture, advances in optimizing radiation therapy offer the potential for the greatest clinical gains. This could be achieved by the combination of external-beam therapy with a local boost from either intraluminal sources or intraoperative external-beam radiation. The use of 5-FU
Intraluminal brachytherapy is an important component in the multimodality approach to bile duct carcinomas. The objective of this treatment is to deliver a high local dose of radiation to the tumor while sparing surrounding normal tissues. The treatment can be safely adapted for right and left hepatic duct as well as for common bile duct lesions. Several different techniques have been described; most of these are variants on one of two themes, employing either conventional strength iridium192 sources at a low dose rate (LDR) or the high doserate (HDR) remote afterloading technique. It has been reported that there does not appear to be any difference in survival or complications between low dose-rate and high dose-rate brachytherapy after biliary drainage, although local failure was a continuing problem [84]. The approach to source placement is where these techniques differ - thus, the plastic catheters (through which the radiation sources are loaded) can be placed at the time of the radiological studies (PTCA), or at laparotomy, or via an endoscope during an ERCP. The transhepatic catheter placement (during PTCA) has the advantage of providing both internal drainage across the tumor and external drainage via the proximal end of the catheter. Occasionally, peritumoral edema may block the passage of the tube into the duodenum; a few days of external decompression of the biliary tree allows the edema to resolve and a second attempt is usually successful in passing the tube into the duodenum. In the face of a refractory obstruction, intraluminal brachytherapy to the proximal tumor may further help open up the channel to the duodenum [66]. This is probably the preferred approach in most cases. Intraoperative catheter placement (T-tubes and several other variants) can be performed by the surgeon after biopsy, bypass, or resection. It is very important to coordinate the procedure with the radiation oncologist to ensure an appropriate location of the catheters in relation to the tumor, tumor margins, and bowel segments,
Bile duct carcinoma 327
as well as to ensure that the lie of the catheter has no acute angulations which can preclude successful afterloading [67,68]. Catheter insertion can also be accomplished by the gastroenterologist at the time of the ERCP [69,70]. This technique has the advantages of avoiding a percutaneous liver puncture and a laparotomy. It is essential that the gastroenterologist avoid the creation of a redundant loop in the duodenum which is difficult for the HDR source to negotiate [71,72]. The necessity for multiple ERCPs, difficulty in anchoring the catheter in the biliary tree, accidental irradiation of the duodenum, in-transit irradiation of the upper gastrointestinal structures, problems with the source negotiating the long path created by the transnasal catheter all the way to the bile duct, and the need to remove, and then replace, the biliary endoprosthesis are all problems with this approach. TREATMENT PLANNING AND TECHNIQUE The following general principles are applicable for percutaneous biliary irradiation regardless of the specific technique employed. 1. The site and length of the malignant stricture are identified by review of the cholangiograms (Figure 22.6). 2. Biliary drainage is established with an adequatesized catheter (10 French for LDR techniques, and 12-14 French for HDR techniques). 3. The patient is pretreated with antibiotics (usually ciprofloxacin) the day before the procedure, and the antibiotics are usually continued for a day following the procedure. 4. The outer end of the biliary drain is opened, cleaned well with Betadine, and flushed with 50 ml of normal saline. 5. The sterile, blind-ended brachytherapy applicator catheter is passed through the biliary drainage catheter; this passage may be eased by keeping a
6.
7.
8. 9.
guide wire within the brachytherapy catheter to 'stiffen' it and help negotiate any curves or bends. However, it should be remembered that the transhepatic drain has numerous side-holes, both proximal and distal to the stricture, and the brachytherapy catheter may be 'deflected' out through one of these holes, especially if it encounters resistance to smooth passage along the biliary drain. This could result in penetration of the liver parenchyma by the catheter, and should be suspected any time the patient complains of discomfort or pain during the negotiation of the catheter. The use of fluoroscopy with or without biliary contrast injection can help in such situations. A dummy source is then passed into the brachytherapy catheter and orthogonal radiographs are obtained for computerized dosimetry (Figure 22.7). The target volume is carefully delineated on the simulation films, using the original cholangiograms as a guide. Following the generation of a computer-assisted plan, the patient is treated to the prescribed dose. With continued aseptic precautions, the brachytherapy catheter is withdrawn, the biliary tree is flushed again with 50 ml of sterile normal saline, and external biliary drainage is reinstituted.
TARGET VOLUME
Given the multicentric nature of the tumor, the propensity for submucosal or intraluminal spread, and reports of marginal failures in the early literature, a wide variety of margins have been employed proximal and distal to the cholangiographic abnormality, while defining the target length. Herskovic et al. [73] even recommended irradiating a length extending from the ampulla to at least 5 cm proximal to the hepatic bifurcation. Levitt et al. [72] recommend a source length 1.5 times the length of the stricture. We recommend a margin of 2-3 cm on either side of the cholangiographic abnormality and have not noted any marginal failures with this approach. DOSE SPECIFICATION
Figure 22.6 Cholangiogmm of a 60-year-old white male showing obstruction of common hepatic duct.
The diameter of the extrahepatic biliary tree can vary from 3 mm to 25 mm [73]. Traditional cholangiographic imaging is barely adequate in defining the presumed proximal and distal extent of disease; it gives no indication as to the lateral extent of the tumor. Various prescription points have been employed - distances of 0.5-1.5 cm with the reference being either the source train or the exterior catheter wall, or even the bile duct surface; sometimes, the reference point is not even specified. This results in a great variation in reported delivered doses and dose rates. It also makes it difficult to compare results from different series; an analysis of complications and derivation of a dose-response
328 Brachytherapy in the treatment of pancreas and bile duct cancer Figure 22.7 The right and left transhepatic drainage catheter and a dummy source ribbon in the left catheter.
relationship are a real challenge, given the heterogeneity of data. In order to standardize dose prescription and make inter-institutional comparisons meaningful, it is recommended that the dose be prescribed at 1 cm from the central axis of the source. DOSE-FRACTIONATION SCHEMES
Traditionally, a 2-4-week gap has been allowed between the completion of the external-beam radiotherapy and the institution of brachytherapy, although there are no hard data to support the use of any break at all. While the doses reported in the literature vary widely, a single fraction of 2000 cGy, prescribed at 1 cm from the source axis, is usually employed with LDR brachytherapy, when it is used as a boost following external-beam radiotherapy. Similarly, for HDR brachytherapy, we currently employ four weekly fractions of 500 cGy, at a distance of 1 cm from the source axis, as a boost following 45-50 Gy external-beam therapy. In palliative settings, the therapy obviously needs to be tailored to the individual patient. RESULTS OF LOW DOSE-RATE THERAPY
Although several series have reported results of LDR intraluminal brachytherapy [69,70], none has conclusively shown a survival advantage. Using the percutaneous approach, Molt et al. [69] reported a median survival of 4.5 months amongst 15 patients undergoing treatment with intraluminal brachytherapy. Similarly, Johnson et al. (70) combined external-beam radiation and intraluminal brachytherapy in the treatment of bile duct cancer and reported a median survival of 5.5 months. Results with endoscopic management of inoperable cholangiocarcinoma using the LDR iridium-192 source technique were reported by Ede et al. [71]. The authors
justified their endoscopic approach by the high rate of morbidity and mortality associated with percutaneous stent insertion in malignant obstructive jaundice. This approach necessitated hospitalization and two ERCPs for visualization of the biliary tree, localization of the tumor, insertion of a stent, and administration of brachytherapy. Complete obstruction or strictures interfered with this treatment in three out of 14 patients, necessitating a percutaneous transhepatic drainage. A dose of 6000 cGy was prescribed to 0.5 cm from the axis of the source. The indwelling time ranged from 77 to 116 h (median 85 h), depending on the length and activity of the available source. No deaths were reported within the first 30 days following the treatment. The complications related to this treatment included transient increase in bilirubin level noted in four patients due to interference with bile drainage during the procedure, and ascending cholangitis in 3/14 cases (21.4%) within 4 days following the treatment. The rate of ascending cholangitis was 30%, according to authors using a similar procedure [58]. The patients in Ede's series were discharged from the hospital 2.5 days on average (range 0-28) following the procedure. The overall median survival was 10.5 months. The combination of external-beam radiation with LDR internal irradiation via an indwelling biliary catheter, reported by Herskovic et al., resulted in a very high complication rate [73]. Seven out of 16 patients (44%) experienced severe life-threatening complications within 1-9 months following the procedure, such as septic shock, septicemia, abscesses, endocarditis, cholangitis, hemobilia, duodenal and gastric ulcers. The median survival in this series was 9 months. One of the major problems inherent in the LDR approach is the prolonged interference with biliary drainage, resulting in a high risk for cholangitis and
Conclusion 329
other septic complications. There is also the problem of radiation exposure to the treating personnel; the need for hospitalization is another issue in today's costconscious environment. All these issues are easily addressed by an HDR approach, as described below. RESULTS OF HIGH DOSE-RATE TREATMENT
We reported the results on a total of 46 HDR brachytherapy sessions in 14 patients with inoperable bile duct carcinomas [74,75]. The combined treatment, consisting of EBRT and 5-FU and a boost given by conformal HDR percutaneous transhepatic intraluminal cholangioirradiation (PTICI) was well tolerated by all 14 patients. The general principles of transhepatic intraluminal brachytherapy, as described earlier, were implemented. 12-French gauge hepatic drainage catheters were inserted in the right and in the left hepatic ducts. Following simulation, and definition of the target length, the number of dwell positions, their optimal separations from each other, and the length of time the sources should remain in each of the dwell positions to deliver a dose of 500 cGy to a depth of 1 cm from the catheter are determined with the aid of a computer. A total of 2000 cGy is delivered in four weekly fractions. The computerized optimization program is very useful in treating complex lesions involving the right, left, and common hepatic ducts, as the dosimetry at the bifurcation is quite complex. One month after the last treatment, the patients were evaluated by the interventional radiologist for placement of an internal stent and removal of the biliary drainage catheter. Kamada and colleagues analyzed the treatment of 145 consecutive patients treated by intraluminal iridium-192 irradiation either alone or in combination with externalbeam therapy (mean dose 83 Gy) with an expandable metallic biliary endoprosthesis to establish an internal bile passage. Median survival for patients for whom an endoprosthesis was used was 14.9 months, compared to 9.3 months when no endoprosthesis was used. The overall results indicated a survival advantage for patients not suited to surgical resection [81]. Conformal HDR brachytherapy, combined with external-beam radiation therapy and 5-FU, appears to be a safe, well-tolerated and efficient treatment for unresectable cholangiocarcinoma (Figure 22.8). Eight
patients reported mild to moderate gastrointestinal symptoms, due to the external-beam radiation therapy, which were treated symptomatically. None of the patients had any severe treatment-related complications necessitating admission. The median survival for the entire group was 19 months. Decrease in jaundice was reported by all patients. Symptomatic improvement, such as increase in appetite and weight gain, was reported by nine patients.
22.4
CONCLUSION
High dose-rate percutaneous transhepatic intraluminal cholangio-irradiation, when combined with externalbeam radiation and administration of 5-FU chemotherapy, is a well-tolerated procedure that may prolong the median survival of selected patients with unresectable bile duct carcinoma. This technique is simple, safe, reproducible, and is associated with a significantly lower complication rate and no exposure to personnel. The entire procedure is done on an outpatient basis. Also, in selected cases with complex tumor anatomy, the technique allows for a more versatile optimization of the dose distribution, as well as offering the option to vary the dose prescription as the tumor regresses between fractions (Table 22.5). As with any other brachytherapy
j-A-HDR PTICI [36] -a-LDR PTICI [59] -•-Conformal RT [6I]| Figure 22.8 Computerized disease-related survival with LDR and HDR brachytherapy compared with high dose conformal external-beam radiation.
Table 22.5 Comparison of LDR and HDR brachytherapy for bile duct carcinoma
Treatment duration Ability to optimize Complications Prescribed dose Dose prescription point Median survival Cost
5-10min Very versatile Few (little interruption of biliary drainage) SOOcGyxS 1.0 cm from source axis 19 months Low (outpatient procedure)
24^8 h Little, if any High (prolonged interruption of biliary drainage) 2000cGyx1 1.0 cm from source axis 13 months High (inpatient procedure)
330 Brachytherapy in the treatment of pancreas and bile duct cancer
procedure, good collaboration with the surgeon/radiologist and meticulous attention to detail are essential for a safe and successful outcome. Further advances, including a better definition of the target volume, and innovative adjuncts like hyperthermia, may improve on these results [76-87].
adenocarcinoma of the pancreas: an update. World J. Surg., 12, 658-61. 17. Grace, PA, Pitt, HA.Tompkins, R.K.rto/.(1986) Decreased morbidity and mortality after pancreoduodenectomy. Am. J. Surg., 151,141-9. 18. Tepper, J., Nardi, G. and Suit, H. (1986) Carcinoma of the pancreas: review of the MGH experience from 1963-1987, analysis of surgical failure and implications
REFERENCES
for radiotherapy. Cancer, 37,1519-24 19. Morrow, N., Hilaris, B. and Brennan, M.F. (1984) Comparison of conventional surgical resection, radioactive implantation and bypass procedure for
1. Tempero, M.A. (1994) Pancreas. In Current Therapy in Cancer, ed. J.F. Foley, J.M. Vose and J.O. Armitage. Philadelphia, Saunders. 2. DiMagno, E. (1991) Pancreatic adenocarcinoma. In Textbook of Gastroenterology, ed. Y. Tadataka, Philadelphia, Lippincott, 1893-910.
exocrine carcinoma of the pancreas. Ann. Surg., 199,1-5. 20. Trede, M., Chir, B. and Schwall, G. (1988) The complications of pancreatectomy. Ann. Surg., 207(7), 39-47. 21. Griffin, J., Smalley, S.R., Jewell, W. et al. (1990) Patterns of failure after curative resection of pancreatic carcinoma. Cancer, 66, 56-61.
3. Parker, S.L., long. T., Bolden, S. and Wingo, P.A. (1997) Cancer statistics, 1997. CA CancerJ. Clin., 47, 5-27. 4. Warshaw, A.L and Fernandez-del Castro, C. (1992) Pancreatic carcinoma. N. Engl.J. Med., 326(7), 455-65. 5. Murr, M.M., Sarr, M.G., Oishi, A.J. and van Heerden, J.A. (1994) Pancreatic cancer. CAJ. Clinic., 44,304-18. 6. Haskell, C., Selch, T. and Ramming, K. (1990) Pancreatic adenocarcinoma. In Exocrine Pancreas in Cancer Treatment, ed. C. Haskell, Philadelphia, Saunders, 259-66. 7. Merrick, H.W. and Dobelobower, R.R. (1990) Aggressive treatment for cancer of the pancreas. Gastroenterol. Clin. North Am., 19,935-61. 8. Vezeridis, M.P. and Wanebo, H.J. (1994) Pancreatic cancer -diagnosis and treatment in 1994. Rhode Island Med., 77, 115-18. 9. Fernandez del Castillo, C. and Warshaw, A.L. (1990) Diagnosis and preoperative evaluation of pancreatic cancer, with implications for management. Gastroenterol. Clin. North Am., 19,915-933. 10. Condie, J.D., Nagpal, S. and Peebles, S.A. (1989) Surgical treatment for ductal adenocarcinoma of the pancreas. Surg. Gyn. Obstet., 168,437-45. 11. Gudjonsson, B. (1987) Cancer of the pancreas. Cancer, 60, 2284-303. 12. Livingston, E., Welton, M. and Reber, H. (1991) Surgical treatment of pancreatic cancer-the United States experience. Int.). Pancreatol.,9,153-7. 13. Connolly, M.M., Dawson, P.J., Michelassi, F., Moossa, A.R. and Lowenstein, F. (1987) Survival in 1001 patients with carcinoma of the pancreas. Ann. Surg., 206,366-70. 14. Jones, B.A., Langer, B., Taylor, B.R. and Girotti, M. (1985) Periampullary tumors: which ones should be resected? >l/iU Surg., 149,46-52. 15. Lerut,J.P.,Gianello, P.R.,Otte,J.B.and Kestens, P.J.(1984) Pancreaticoduodenal resection. Surgical experience and evaluation of risk factors in 103 patients. Ann. Surg., 199, 432-7. 16. van Heerden, J.A., Mcllrath, D.C., llstrup, D.M. and Weiland, LH. (1988) Total pancreatectomy for ductal
22.
Hansen, R., Quebbeman, E., Ritch, P., Chitambar, C. and Anderson, T. (1988) Continuous 5-fluorouracil (5FU) infusion in carcinoma of the pancreas: A phase II study. Am.J. Metf. So'., 295, 91-3.
23. Moore, M., Andersen, J., and Burris H. et al. (1995) A randomized study of gemcitabine vs 5-fluorouracil in patients with inoperable pancreatic adenocarcinoma. Proc. Am. Soc. Clin. Oncol. (Abstract) 14,199. 24. Stevens, K.R. (1994) The pancreas. In Moss' Radiation Oncology, ed. J.D. Cox, St Louis, Mosby, 440-51. 25. Moertel, C.G., Frytak, S., Hahn, R.G. etal. (1981) Therapy of locally unresectable pancreatic carcinoma: a randomized comparison of high dose (6000 rads) radiation alone, moderate dose radiation (4000 rads) + 5-fluorouracil and high dose radiation (6000 rads) + 5-fluorouracil. Cancer, 48,1705-10. 26. Roldan, G.E., Gunderson, L.L, Nagorney, D.M. etal. (1988) External beam versus intraoperativeand external beam irradiation for locally advanced pancreatic cancer. Cancer, 16,1110-16. 27. Shipley, W.U., Wood, W.C., Tepper, J.E. et al. (1984) Intraoperative electron beam irradiation for patients with unresectable pancreatic carcinoma./Ann. Surg., 200, 289-94. 28. Gunderson, L.L, Martin, K.J., Kvols, LT. etal. (1987) Intraoperative and external beam irradiation +/- 5-FU for locally advanced pancreatic cancer. Int.J. Radial Oncol. fi/o/.Pr»ys.,13,319-24.
29. Dobelbower, R.R., Loeffler, R.K., Merrick, H.W. etal. (1991) Radiation therapy in cancer management: new frontiers. In Comprehensive Textbook of Oncology, 2nd edn, ed. A.R. Moossa, S.C. Schimpff and M.C. Robson, Baltimore, Williams and Wilkins, 502-22. 30. Handley, W.S. (1934) Pancreatic cancer and its treatment by implanted radium. Ann. Surg., 100,215-23. 31. Montemaggi, P., Dobelbower, R., Crucitti, F. etal. (1991) Interstitial brachytherapy for pancreatic cancer: report of seven cases treated with 1-125 and a review of the literature. Int.J. Radial Oncol. Biol. Phys., 21,451-7.
References 331 32.
Fu, K., Ling, C., Nath, R. and Kim, J. (1990) Radiobiology of brachytherapy. In Interstitial Brachytherapy- Physical
Biological, and Clinical Considerations, ed. L.L. Anderson, R. Nath, K.A. Weaver, D. Nori, T.L Phillips and Y.H. Son. New York, Raven Press, 47-61. 33. Weaver, K., Anderson, L and Meli, J. (1990) Source characteristics. In Interstitial
Brachytherapy-Physical,
Biological, and Clinical Considerations, ed. L.L. Anderson, R. Nath, K.A. Weaver, D. Nori, T.L Phillips and Y.H. Son. New York, Raven Press, 3-13. 34. Nath, R. (1993) New directions in radionuclide sources for brachytherapy. Semin. Radial Oncol., 3,278-89. 35. Khan, P.M. (1994) Brachytherapy. In The Physics of Radiation Therapy, ed. W. Passano, III. Baltimore, Williams and Wilkins, 418-73. 36. Iyer, P. and Shanta, A. (1994) Update of radionuclides seed in endocurietherapy. Endocuriether./Hypertherm. Oncol., 10,161-5. 37. Peretz,T., Nori, D., Hilaris, B.etal. (1989) Treatment of
localized unresectable adenocarcinoma of the pancreas. Int.J. Radial Oncol. Biol. Phys., 14,79-84. 48. Raben.A., Mychalczak, B., Brennan, M.etal. (1996) Feasibility study of the treatment of primary unresectable carcinoma of the pancreas with Pd-103 brachytherapy. Int.J. Radial Oncol. Biol. Phys., 35,351-6. 49. Parikh, S. and Nori, D. (1998) Unresectable pancreatic adenocarcinoma treated with palladium-103 implant and chemoradiation./ Clin. Gastroenterol., 26(2), 158-60. 50. Lotze, M.T., Flickinger, I.e. and Carr, B.I. (1993) Hepatobiliary neoplasms. In Principles and Practice of Oncology, ed, V.T. DeVita, S. Hellman and S.A. Rosenberg. Philadelphia, Lippincott, 883-914. 51. Gunderson, L.L. and Willet, C.G.(1992) Pancreas and hepatobiliary tract. In Principles and Practice of Radiation Oncology, ed. C.A. Perez and L.W. Brady, New York, Lippincott, 985-99. 52. Yeo, C.J., Pitt, H.A. and Cameron, J.L (1994) Surgical
primary unresectable carcinoma of the pancreas with
resection. In Hepatobiliary Malignancy-Its
1-125 implantation. Int.J. Radial Oncol. Biol. Phys., 17, 931-5.
Multidisciplinary Management, ed. J. Terblanche, London,
38. Syed, NAM., Puthawal, A.A. and Beblett, D.L (1983) Interstitial iodine-125 in the management of unresectable pancreatic carcinoma. Cancer, 52,646-50. 39. Fortner, G.F. (1970) Iodine 125 implantation for unresectable cancer of the pancreas. Postgrad. Med., 15, 226-30. 40.
D. and Rosato, F. (1988) Combined modality treatment of
Nori, D. and Hilaris, B.S. (1988) Intraoperative brachytherapy in pancreatic cancer. Cancer Invest., 6, 695-704.
Edward Arnold, 404-15. 53. Pitt, H.A., Dooley, W.C., Yeo, C.J. and Cameron, J.L (1995) Malignancies of the biliary tree. Curr. Prob. Surg., 32(1), 1-100. 54. Algren, J.D. (1994) Neoplasms of the hepatobiliary system. I n Medical Oncology - Basic Principles and Clinical Management of Cancer, 2nd edn, ed. P. Calabresi and P. Schein, New York, McGraw Hill, 713-35. 55. Minsky, B.D., Wesson, M.F., Armstrong, J.G. etal. (1990) Combined modality therapy of extrahepatic biliary
41. Shipley, W.U., Nardi, G.L, Cohen, A.F. and Ling, C.C. (1980)
system cancer. Int.J. Radial Oncol. Biol. Phys., 18(5),
Iodine-125 implant and external beam irradiation in patients with localized pancreatic cancer. Cancer, 45, 709-14.
1157-63. 56. Meerwaldt, J.H., Veeze-Kuijpers, B. and Visser, A.G. (1989) Combined modality radiotherapy in the treatment of bile
42.
Dobelbower, R.R., Merrick, H.W., Ahuja, R.K. and Skeel, R.
duct carcinoma. In Brachytherapy2, ed. R.F. Mould,
(1986) 1-125 implant, precision high dose external beam
Veenendaal, The Netherlands, Nucletron, 576-83.
therapy, and 5-FU for unresectable adenocarcinoma of pancreas and extrahepatic biliary tree. Cancer, 58, 2185-95. 43.
Meigooni, A., Sabnis, S. and Nath, R. (1990) Dosimetry of
palladium-103 brachytherapy sources for permanent implants. Endocuriether./Hypotherm. Oncol., 6,107-17. 44. Nori, D., Merimsky, 0., Osian, A.D., Hefernan, M., Cortes, E. and Turner, J.W. (1996) Palladium-103: a new radioactive source in the treatment of unresectable carcinoma of the pancreas: a phase l-ll study./ Surg. Oncol., 61,300-5. 45. Nori, D. (1988) Brachytherapy for cancer of the pancreas.
Management, ed. J. Terblanche. London, Edward Arnold, 468-77. 58. Falkson, G., Maclntyre, J.M. and Moertel, C.G. (1984) Eastern Cooperative Oncology Group experience with chemotherapy for inoperable gallbladder and bile duct cancer. Cancer, 54,965-70. 59. Haskell, C.M. (1980) Cancer of the liver. In Cancer Treatment, ed. C. Haskell, Philadelphia, Saunders, 645-60. 60. Davis, H.L, Ramirez, G. and Ansfield, F.J. (1974)
\nAtlasof Brachytherapy, ed. B. Hilaris, D. Nori and L.
Adenocarcinoma of the stomach, pancreas, liver and
Anderson, New York, Macmillan, 204-14.
biliary tracts: survival of 328 patients treated with fluoropyrimidine therapy. Cancer, 33,193-7.
46. Borgelt, B. (1980) Radiation therapy with either gold grain implant or neutron beam for unresectable adenocarcinoma of the pancreas. In Pancreatic CancerNew Directions in Therapeutic Management, ed. I. Cohn. New York, Masson, 256-65. 47.
57. Price, P. (1994) Radiotherapy and cytotoxic therapy. In Hepatobiliary Malignancy-Its Multidisciplinary
Mohiuddin, M., Cantor, R., Bierman, W., Weiss, S., Barbot,
61. Abe, M., Shibamonto, Y., Takahashi, M., Manabe, T., Tobe, T. and Inamoto, T. (1987) Intraoperative radiotherapy in carcinoma of the stomach and pancreas. World J. Surg., 11,459-64. 62. Iwasaki, Y., Todoroski, T., Fukao, K., Ohara, K., Okamura,
332 Brachytherapy in the treatment of pancreas and bile duct cancer
63.
64.
65.
66.
67. 68.
69.
70.
71.
72.
73.
74.
75.
T. and Nishimura, A. (1988) The role of intraoperative radiation therapy in the treatment of bile duct cancers. WorldJ.Surg., 12,91-8. Todorki, T., Iwasaki, Y. and Okamura, T. (1980) Intraoperative radiotherapy for advanced carcinoma of the biliary system. Cancer, 46,2179-84. Rich, T.A., Cacy, B., McDermott, W.V., Case, K.R., Chassey, J.T. and Hellman, S. (1988) Orthovoltage intraoperative radiation therapy: a new look at an old idea. Int. J. Radial Oncol. Biol. Phys., 10,1957-63. Busse, P.M., Stone, M.D., Shelton, T.A. etal. (1989) Intraoperative radiation therapy for biliary duct carcinoma. Results of a 5 year experience. Surgery, 10, 724-33. Son, Y.H., Nori, D., Leibel, S.A. etal. (1990) Biliary tree. In Interstitial Collaborative Working Group. Interstitial Brachytherapy, Programme and Abstracts. 7th International Brachytherapy Working Conference. New York, Raven Press, 153-6. Nunnerly, H.B.and Karani.J.B. (1990) Intraductal radiation. Radiol. Clin. North Am., 28,1237-40. Fletcher, M.S., Brinkley, D., Dawson, J.L et al. (1981) Treatment of high bile duct carcinoma by internal radiotherapy with iridium-192wire. Lancet,2, '\72-A. Molt, P., Hopfan, S., Watson, R, Botet, J.F. and Brennan, M.F. (1986) Intraluminal radiation therapy in the management of malignant biliary obstruction. Cancer, 57, 536-44. Johnson, D.W., Safai, C. and Goffinet, D.R. (1985) Malignant obstructive jaundice: treatment with external beam and intracavitary radiation therapy. Int.J. Radial Oncol. Biol. Phys., 11,411-16. Ede, R.J., Williams, S.J., Hatfield, A.R.W., Mclntyre, S. and Mair, G. (1989) Endoscopic management of inoperable cholangiocarcinoma using iridium-192. Br.y. Swrg., 76, 867-9. Levitt, M.D., Laurence, B.H., Cameron, F. and Klemp, P.F.B. (1988) Transpappilary iridium-192 wire in the treatment of malignant bile duct obstruction. Gut, 29,149-52. Herskovic, A.M., Engler, M.J. and Noell, K.T. (1985) Radical radiotherapy for bile duct carcinoma. Endocuriether./ Hypertherm. Oncol., 1,119-24. Nori, D., Nag, S., Rogers, D. etal. (1994) Remote afterloading high dose-rate brachytherapy for carcinoma of the bile duct. In High Dose Rate Brachytherapy: a Textbook, ed. S. Nag. Armonk, NY, Futura Publishing Co., Inc., 331-8. Merimsky, 0., Nori, D., Rogers, D. and Osian, A. (1995)
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86. 87.
Conformal high dose-rate percutaneous transhepatic intraluminal cholangio-irradiation for unresectable cholangiocarcinoma. Endocuriether./Hypertherm. Oncol., 11,115-20. Robertson, J.M., Lawrence, T.S., Dworzanin, l.M.etal. (1993) Treatment of primary hepatobiliary cancer with conformal radiation therapy and regional chemotherapy. J. Clin. Oncol., 11,1286-93. Minsky B., Botet, J., Derdes, H.etal. (1992) Ultrasound directed extrahepatic bile duct intraluminal brachytherapy. Int. J. Radial Oncol. Biol. Phys., 23,165-7. Coughlin, C.T., Wong, T.Z., Ryan, T.P. etal. (1992) Interstitial microwave-induced hyperthermia and iridium brachytherapy for the treatment of obstructing biliary carcinomas. Int.]. Hyperthermia, 8, 57-71. Foo, M.L, Gunderson, L.L., Bender, C.E. and Buskirk, S.J. (1997) External radiation therapy and transcatheter iridium, in the treatment of extra hepatic bile duct cancer. Int.). Radial Oncol. Biol. Phys., 39(4), 929-35. McMasters, K.M., Tuttle, T.M., Leach, S.D. etal. (1997) Neoadjuvant chemoradiation for extra hepatic cholangiocarcinoma. Am.J. Surg., 174(6), 605-8, discussion 608-9. Kamada,T.,Saiton,H.,Takamura,A., Nojima.T. and Okushiba, S.I. (1996) The role of radiotherapy in the management of extra hepatic bile duct cancer; an analysis of 145 consecutive patients treated with intraluminal and/or external beam radiotherapy. Int.J. Radial Oncol. Biol. Phys., 34(4), 767-74. Montemaggi, P., Morganti, A.G., Dobelbower, R.R. et al. (1996) Role of intraluminal brachytherapy in extra hepatic bile duct and pancreatic cancers: is it just for palliation? Radiology, 199(3), 861-6. Vallis, K.A., Benjamin, I.S., Muaro, A.J.^o/.(1996) External beam and intraluminal radiotherapy for locally advanced bile duct cancer: role and tolerability. Radiother. Oncol., 41(1), 61-6. Leung, J.T. and Kuan, R. (1997) Intraluminal brachytherapy in the treatment of bile duct cncer. Aust. /tad/o/., 41 (2), 151^1. Schuricht, A.L, Spitz, F., Barbot, D. and Rosato, F. (1998) Intraoperative radiotherapy in the combined modality management of pancreatic cancer. Am. Surg., 64(11), 1043-9. Thomas, PR. (1996) Radiotherapy for carcinoma of the pancreas. Semin. Oncol., 23(2), 213-19. Dobelbower, R.R. and Montemaggi, P. (1996) Brachytherapy for pancreatic cancer: a review. Hepatogastroenterology, 43(8), 333-7.
23 Brachytherapy for treating endometrial cancer H. A. LADNER, A. PFLEIDERER, S. LADNER, AND U. KARCK
23,1
CLINICAL ASPECTS
During the last 30 years, technical advances as well as clinical developments have produced a significant change in the diagnosis and therapy of endometrial cancer. The increase in absolute numbers of endometrial cancers [1,2] and the increase of patients over 65 years of age cause a number of problems, including accompanying illness, appropriate radiation fractionation schedules and techniques [3,4]. Progress in anesthesiology and postoperative care now allows surgery in nearly all cases of endometrial carcinoma. As a result of these developments, staging can now be based on postoperative histopathological assessment. Relevant prognostic factors of endometrial cancer include stage (FIGO), depth of myometrial invasion, extrauterine extension, histological subtype, grade, ploidy, and levels of progesterone receptors. Pathogenetically, two different types are described [5,6]. The estrogen-dependent type A is characterized by a low histologic grade 1, high progesterone receptor value, endometrioid cancer in obese patients, and has an excellent prognosis. Type B cancers show no hormonal stimulation, are high-grade cancers [3] without progesterone receptor activity, and have a poor prognosis. A review of the recent literature shows almost general agreement that the cornerstone of curative treatment for endometrial carcinoma is abdominal hysterectomy, bilateral oophorectomy, and facultative lymphadenectomy [7]. Modern facilities now provide for surgery in obese, elderly women with vascular illness, which has dramatically reduced the number of 'inoperable' patients. Only a few patients with inoperable endometrial carcinoma now receive primary radiotherapy, with
a 5-year survival rate of about 20%. In 1995, Weiss et al. [8] reported only 21 patients who had undergone primary irradiation during the preceding 6 years. The changes in therapy, as described above, have developed slowly over the last 20 years. According to other authors [9,10], the difference in survival rate between surgical and radiotherapy treatment was, in the decade until 1985, of the order of 20% (i.e., 80% survival versus 60%), and it is difficult to evaluate the optimal indications for irradiation. Some of the management aspects are now described, together with the technical progress that has been made. In this context, it appears important to outline the best technical application of afterloading therapy in the future on the basis of the interesting review by Joslin [11].
23*1*1
Brachytherapy
There are three main requirements for the best appropriate use of brachytherapy: 1. Spatial dose distribution must be adapted to the individual anatomical situation. 2. Temporal dose distribution should be short enough to avoid clinical complications without reducing the therapeutic ratio. 3. The technique should be simple to apply and allow for reproducible source positioning. It should also be safe to use and allow for adequate radiation protection of staff. In the past, these requirements were not always satisfactorily met. The classic radium methods such as the Paris System, the Manchester System, and the method of Fletcher et al.
334 Brachytherapy for treating endometrial cancer
(1980) [12] involved protocols similar to those used for cervical carcinoma, with some degree of individualization. The Stockholm technique used flexible applicators which formed the basis of the classical Heyman packing technique of the uterine cavity with radium sources. It allowed the dose distribution to be adapted to the various shapes and sizes of uterine cavity in an optimal way, especially in cases with exophytic tumors. Furthermore, a relatively uniform dose distribution was delivered to the entire myometrium. High dose-rate (HDR) afterloading and its development for irradiation of endometrial carcinoma are associated with the name of Ulrich Henschke [ 13]. Following his guidelines for remote HDR treatment of gynecologic carcinomas, a variety of techniques in brachytherapy of uterine cancer was established in the following years [14-21,82]. Due to the use of combined treatments (brachytherapy and external-beam radiation therapy, EBRT), the situation was and still is complicated when looked at from the radiobiology viewpoint [11]. This is due to the fact that intracavitary irradiation is not delivered homogeneously throughout the treated volume. Further, in consequence of the steep dose gradient from the brachytherapy, the normal tissues outside the brachytherapy target volume receive relatively low doses of irradiation and the brachytherapy target volume is small in comparison to the external-beam target volume [11].
23.1.2 Dose rates and the choice of nuclide The reference dose rate according to the International Commission on Radiation Units (1985) is defined as follows: HDR treatment with more than 12 Gy h'1, medium dose-rate (MDR) treatment with 2-12 Gy h"1, and low dose-rate (LDR) treatment with less than 2 Gy h"1. There are different spatial arrangements of sources in use, such as a single oscillating source, a single source moving in steps (so-called train of different weighted pellets), or a train containing active sources and spacers. Simon and Silverstone [22] recommended a modified Heyman packing technique using hollow capsules, afterloaded by low-activity cesium-137 sources. However, they were confronted with the clinical consequences of a relatively long treatment time and lack of radiation protection for staff and visitors. A prerequisite for the achievement of optimal spatial and temporal dose distribution in combination with radiation protection and the advantages of Heymanstyle packing with remote HDR afterloading is the availability of small sources and computerized treatment planning. One good example seems to be the Wurzburg system, with small iridium-192 sources (diameter 1 mm) in conjunction with a MicroSelectron HDR afterloading machine, which allows for different kinds of dose opti-
mization when carrying out treatment planning (for technical details, see Herbolsheimer et al. [23]). Optimization according to the individual case can be performed by altering the dwell times of the six to 18 different capsules. When comparing the advantages of cesium-13 7, cobalt-60 and iridium-192 in regard to size, specific activity, and gamma energy, cesium-137 meets the special requirements best. This was reported in 1977 by Walstam [24], who suggested cesium-137 as the most suitable nuclide for brachytherapy in general.
23.2 MANAGEMENT AND CLINICAL PRACTICE The technique used by M. Herbolsheimer and K. Rotte in the U K F Wurzburg (Germany) is one which we advocate and now describe. One of the most important requirements for optimal brachytherapy is to be able to provide an individual dose distribution according to the size and share of the uterine cavity. A modified packing system such as the classical Heyman packing system using HDR afterloading can only be used when a sufficient number of small sources is available. This modified method was introduced by Herbolsheimer and Rotte (Wurzburg, Germany) in 1988. The use of small iridium-192 sources (diameter 1 mm) allows the uterine cavity to be packed with up to a maximum of 18 hollow plastic capsules, each capsule having a diameter of 4 mm. Following the placing of the capsules in the uterine cavity, orthogonal pelvic radiographs showing the distribution of the capsules, the measuring probes in the organs at risk and the bony structures are taken and digitized. A computer-controlled planning system provides the dose distribution in three dimensions and calculates the doses to the reference points of the organs at risk. In order to adapt the reference isodose to the uterine surface, the doses were prescribed to the so-called point 'MY' (myometrium) during the early years following the introduction of the system. 'MY' is situated 2 cm lateral to the central uterine axis and 2 cm below the most proximal capsule, which means 2 cm below the uterine fundus. Because the uterine shape is not visible on conventional radiographs and due to the fact that the point 'MY' is only an approximation, magnetic resonance tomography has been introduced into the planning procedure since 1993. Double-angulated slices (Figures 23.la and 23.Ib) are digitized and several reference points on the uterine surface are defined individually. In order to combine these points located on the serosa of the uterus by the reference isodose, it is necessary to distinguish each capsule from another and to assess them separately. A binary code of the dummies allows one to perform an optimization of the dose distribution. Five treatment fractions are given, each of 10
Management and clinical practice 335 Figure 23.1 Digitized slices from a scan illustrating the isodose distribution in relation to various reference points. Isodoses are in centigrays and axis scale dimensions are in centimeters.
Gy separated by 10 days. These treatments are scheduled within a course of megavoltage external-beam treatment to the pelvic lymphatics using bisegmental-biaxial arc rotation. Twenty-five fractions, each of 2.0 Gy, are delivered to the maximum dose point according to a threedimensional plan based on computer tomography. The 80% isodose should at least enclose the target volume. The total dose to the center of the pelvis, that is to say to the uterus, is delivered by brachytherapy alone, thus external-beam treatment is restricted to the lateral pelvic lymphatics. As a consequence, the mean dose to the tumor will be considerably higher in comparison to a treatment strategy using brachytherapy only as a boost after homogeneous irradiation of the whole pelvis. Moreover, the exposure of the organs at risk is lower. This has been shown by analysis of more than 200 patients [23]. The new Wurzburg results (1988-1994) demonstrate a 3-year actuarial survival rate of 82% in
FIGO stage I-III patients (n=68, 9% recurrence rate). In bulky tumors, which may be expected to contain a large fraction of hypoxic cells, non-restricted brachytherapy is used only as a boost. Compared with other afterloading methods, the advantage of this technique is that it provides for individual planning of doses.
23.2.1
Forms of applicator
Techniques based on the use of rigid applicators do not allow sufficient adaptation to individual anatomical situations. Only the applicators using iridium developed by Bauer et al. [25], the cobalt bulb techniques of Bjornsson and Sorbe [26], the flexible tube combined with vaginal ovoids [27], and the endouterine'umbrella' [21] provide a reasonable compromise with individual dose distribution because they use semi-flexible applicators. Other
336 Brachytherapy for treating endometrial cancer
authors perform uterine packing [17,28], sometimes combined with ovoids in the vaginal vault [29].
23*2.2
Remarks about dose rate
There is no general agreement about what provides the best remote-control brachytherapy method (low, medium, or high dose rate). An important advantage of LDR techniques remains the greater therapeutic ratio in comparison to HDR regimes based on the repair capacity of the tissues involved. Sublethal radiation damage recovery of normal-tissue damage may occur during the exposure time if the dose rate to any tissue or organ at risk is similar to classic radium therapy [30]. The reduced therapeutic ratio of HDR treatment in comparison with LDR brachytherapy requires adequate fractionation based on mathematical models which relate to time-dose relationships and take into account the time factors for repair of normal tissue [31,32]. The question of which treatment method is the most preferable, HDR or LDR afterloading, cannot be answered [33]. Both strategies have their specific advantages. In general, clinical studies comparing HDR and LDR treatments confirm that there is no significant difference between the two modalities of brachytherapy [10,28,34,35]. At present, there is a clear tendency in favor of HDR brachytherapy. However, LDR techniques still retain their place, which is not confined only to gynecologic tumors. It seems to be a question of habituation or availability of the afterloading machines. In Germany, many radiotherapy institutions prefer HDR brachytherapy because it is also possible to use the machine for nongynecological patients requiring HDR. There is no agreement in regard to the optimal dose rates.
23.23 Preoperative and postoperative irradiation Preoperative radiotherapy, partly combined with additional postoperative irradiation, has been used at several institutions [12,21,35,36]. The value of preoperative radiotherapy is discussed controversially [36,37]. Others, such as de Waal and Lochmuller [38] and Calais et al. [39], saw no improvement in the results after preoperative radium or brachytherapy of endometrial carcinoma compared to postoperative application. The timing of radiotherapy (brachytherapy and EBRT), i.e., preoperative versus postoperative, was not a significant factor in univariate or multivariate analysis [40,41]. The important question remains: for which group of patients will afterloading therapy lead to an improvement in overall survival? Several studies suggest that in some cases the application of adjuvant afterloading therapy in combination with EBRT will improve survival in patients exhibiting unfavorable prognostic factors such as low differentiation (grade 2), stage III disease, certain
histologic subtypes (e.g., clear cell adenosquamous and serous-papillary carcinoma), lymph node metastases, peritoneal spread, etc. However, controlled studies are missing. Further, it remains unclear whether adjuvant radiotherapy is helpful in cases of incomplete surgery. The results of therapy can only be evaluated by an exact analysis of recurrences and all forms of relapsing disease and metastases, including cases of complete failure. Recurrences in FIGO stage I or II disease are of particular interest. Following a review of a large number of recent publications, it is apparent that several reports suggest and discuss the benefits of adjuvant irradiation in endometrial cancer.
23.2.4 vagina
Postoperative brachytherapy of the
The incidence of vaginal recurrences after primary hysterectomy and bilateral salpingo-oophorectomy ranges from 5% to 20% if no postoperative radiation is performed [18,42]; the median rate is about 10% [43]. Many authors have reported a considerable reduction of this rate by including vaginal irradiation in the primary management using iridium, cesium, or cobalt [15,17-19,28,39,44-52), but there are also controversial opinions as to whether vaginal irradiation should be done. Sometimes its value is questioned: Malkasian et al. [53] reported 10% recurrences (stage I, EBRT for highrisk patients); Hording and Hansen (54) reported 19% recurrences (stage I, surgery without irradiation). Recurrences occur not only in the vaginal cuff, but also in the lower vagina, particularly in the suburethral area. Therefore, we also irradiate the lower vagina to the hymenal ring. A similar procedure is reported by others [42-44,52]. In contrast, Pipard et al. [55] and Pernot et al [21] do not recommend including the entire length of vagina because they foresee a potential risk to healthy tissues. Randall etal. [56] described the role of an intravaginal cuff boost of 30-50 Gy surface dose after adjuvant external-beam irradiation in early-stage disease treated from 1971 to 1983 with reduction of recurrence rates from 23% to 13% in those cases receiving a vaginal cuff boost.
23.2.5 Fractionation schemes and different dose rates Kob et al, using iridium, discuss fractionation regimes of: 4 x 10 Gy, 5 x 8 Gy, and 8 x 5 Gy at reference point A [34]. Sorbe and Smeds refer to 4 x 9 Gy, 5 x 6 Gy, 6 x 5 Gy, and 6 x 4.5 Gy [18]. Pettersson et al used cesium. Their reference dose to the rectum was 17 Gy or more, and since 1981 has been less than 17 Gy [49]. Kucera used iridium, giving a 10 Gy single dose since 1983: 3 x 7 Gy total dose to the surface of the vagina, and 2 x 7 Gy (at 2 cm distance from the applicator axis) in combination with EBRT [19]. Different LDRs are discussed by Haie-Meder et al [57].
Management and clinical practice 337
In the Freiburg data, vaginal recurrence was distinguished from the combination of vaginal and pelvic recurrence and from primary vaginal involvement, stage III or IV (FIGO, 1988). Recurrences within 6 months or failure of therapy was classified as progressive disease. Pelvic recurrence is defined as tumor growth limited to the pelvis, not involving the vagina, and occurring later than 6 months after primary treatment. The definition of distant metastasis is that of proven distant disease. The 5-year and 10-year survival rates use the data of Kaplan and Meier (SPSS program).
23.2*6 Disease recurrence rates after surgery and irradiation Table 23.1 demonstrates the recurrence rates after 5 years as published in the current literature. The differences in the quoted rates appear to originate from the fact that in some cases all stages were analyzed together, whereas in other reports only the early stages were included, i.e., only stage I [18]. Other confounding factors are differences in age structure of the patients and the incidence of accompanying illnesses. When looking at all stages together in an unselected population, a recurrence rate of about 10% appears to be achievable. Figures substantially higher than 10% ought to stimulate a re-evaluation of the therapeutic regimen used.
FREIBURG RESULTS
The prognosis of recurrent disease depends on the tissue site involved. Isolated vaginal recurrence (n=52) has a favorable prognosis, with a 5-year survival rate of 56.5% and at 10 years of 40%, whereas for patients who in addition had pelvic recurrence («=65), only 29% survived 5 years, which at 10 years was reduced to 17%. Similar results were obtained from cases with primary involvement, with a 5-year survival of 29% (n=67). Other 5-year survival results following radiotherapy for isolated vaginal recurrence have been reported (Table 23.2). These report results similar to our own. The application of brachytherapy, teletherapy, and vaginal surgery has been individualized [45,51,52,60]. Thus, results can be compared in order to assess the possible effects of differing therapeutic approaches which might become apparent. The Freiburg data suggest that the combination of hysterectomy and postoperative irradiation of the entire vagina (done in 925 patients, 76% of 1215 outpatients, 1969-1990) as primary management lowers the frequency of all types of recurrence (vagina: 1.5%, pelvis: 2%, distant: 5.4%) for all stages. In patients demonstrated to have distant metastases («=103) after the completion of primary therapy, there was a 5-year survival rate of 44.3% (at 10 years, 20%). This relatively high survival rate may be caused by the delay in the progression of distant metastasis following
Table 23.1 Frequency of recurrences following surgery and irradiation, as published in the current literature
Reference
5-year recurrence rate (%)
Poulsen and Roberts [62] Lancia no etal. [40] Kuceraeffl/. [19] SorbeandSmeds[18] Randal I era/. [56] Kleineefo/. [63] Petterssonefo/. [49] Calais efo/. [39]
14 10
2.2 4 16 8 10
23.2.7 Therapy and outcome of recurrent disease in Freiburg, Germany DEFINITIONS An isolated vaginal recurrence is a tumor up to 20 mm in size, limited to the vagina, occurring later than 6 months after primary therapy. The definition of isolated vaginal recurrences was adopted from Perez et al. [58] and slightly modified in later publications [46,52,59,60].
Table 23.2 Published 5-year survival rates after radiotherapy of isolated vaginal recurrence and mean follow-up time
3-19 4 4
M\dersetal. [46] (1984), 1960-1976, all stages Mandell etal. [43] (1985), 1969-1980, stage I Nori [17] (1987), 1969-1979, stage I Greven and Olds [66] (1987), 1970-1982, stages I and II Curran et.al. [59] (1988),* 1970-1985, all stages Kuten etal. [61] (1989),* 1959-1986, all stages Vavra etal. [51] (1993), 1973-1987, stage l-lll Elliott etal. [52] (1994),* 1964-1985, stages I and II
42 12 18 18 55 17 40 27
3(0 10
Searsetal. [60] (1994),* 1973-1991, all stages Ladner, UHW Freiburg (1995),* 1969-1990, all stages
45 52
7 11
226
5
Weighted mean, publications with * only
3-10 5 4.8
24 40 50 33 31 40 54 (3a) 23 10(10a) 44 56 40(10a) 41
338 Brachytherapy for treating endometrial cancer
primary therapy because, when looking at the 5-year survival rate after the diagnosis of distant metastasis, it goes down to 21% (at 10 years, 15%). These results also indicate that, in the majority of isolated vaginal recurrence cases, cure can be achieved by consequent brachytherapy and/or surgery. There is no reason for fatalism for patients suffering recurrent endometrial cancer.
23.2*8
Primary irradiation
For patients who are inoperable, for whatever reason, radiotherapy should be individualized. However, there are, today, few patients, whether due to age or to medical disorder, who should be considered inoperable. The mean ages of patients in various publications addressing this problem were68.5years [68],68years [69] and70years [70]. A comparison of 5-year survival data reported in the literature is difficult, because of patient selection, differences in technique and dosage, and inexact staging. In women who undergo combined brachytherapy and external-beam irradiation, the 5-year survival rates are now reported to range from 50% to 88% for stage I tumors and between 35% and 60% for stage II tumors; when all stages are combined, the 5-year survival rate averages about 55% (Table 23.3). A breakdown of the Freiburg results is given in Table 23.4. For patients undergoing primary irradiation, mortality rates after therapy due to intercurrent disease range between 36% and 60% [8,21,29,71,80,82]. These reports demonstrate the need to evaluate correct survival data [10,21,29,82]. 23*2.9
Recurrent disease
The published rates of pelvic recurrences and distant metastases after primary irradiation cover a wide range. According to Glassburn et al. [36], de Vita et al. [42], and Pernot et al. [21], who summarized the literature, disease
Table 233 The published 5-year survival rates following primary radiotherapy for stage I and stage II disease and for all stages
Stage I
Varietal. [71] Sorbertfl/. [73] Kucera ero/. [19,75] Rouaneteffl/. [69] Kupelianeffl/. [68] Petereiteffl/. [77] Ladnerefo/. (1996)
Iridium Cobalt
57 47 67 58
87(stagel-ll) 69 (3 years) 71 (see Table 23.4)
Stage II Grigsby et al. [72]
Taghianeffl/. [74] Booth by efo/. [76] Kucera [19,75] Ladnerefo/. (1996) All stages Taghianefo/. [74] Onnlsetal. [78] Pernot etal. [21] Vahrson[10] Rouanetefo/. [69] Rotte [79] Lad neretal. (1996)
71
Radium
56 36 47
Iridium Iridium
58 (see Table 23.4)
Cesium
52
Iridium
57 55 62 58 72
Iridium
55 (see Table 23.4)
relapses ranged from 8% to 25% for stage I and II disease: Boothby etal. [76], 19%, stage II; Kucera etal. [75], 12%; Huguenin etal [4], 19%; Kupelian etal. [68], 14%; Rouanet et al. [69] (for all stages), 31%. The rate of distant metastases was about 30-35%. Because of the high frequency of recurrent disease, several authors consider a higher total dose of irradiation or a higher number of fractions should be considered in an attempt to improve the outcome [19,75].
Table 23.4 Five-year and 10-year survival rates after primary irradiation with high voltage (HV) irradiation and/or brachytherapy (BT) in the FIGO Stages I-IV. UHW Freiburg, Germany, 1969-1990 (15% of all patients, 392/2530)
I
BT + HV BTonly
47 202
71.4 55.8
66.4 31.2
II
BT + HV BTonly
12 60
58.3 51.2
38.9 31.6
III
BT + HV BTonly
38 7
21.1 0
14.0 0
IV
BT+HV BTonly
10 5
30.0 0
30 0
I-IV
BT+HV BTonly
82 312
55.1 49.5
45.4 28.9
Conclusion 339
23.2*10
Complications
Fistulas involving the rectum or bladder, ureteric stricture, vaginal necrosis or stenosis, and contracted bladder are reported to affect from 3% to 19% of patients (Table 23.5). 23.2.11
Prognostic factors
Recently, we performed a retrospective analysis of all 2533 patients treated for endometrial carcinoma at the Freiburg University Hospital from 1969 until 1990 in order to identify risk factors for survival, recurrence, and adjuvant radiotherapy. Based on a profile of risk factors, patients were classified as either low risk (adenocarcinomas in FIGO stage la with grading 1 or 2, stage Ib only with grading 1) or high risk (all tumors with grading 3 or higher than Ib G2 and all non-adenocarcinomas) in an approach similar to that of the study of Shumsky et al (1997) [86]. The median follow-up for our patients was 12 years. In the Freiburg patients, we found surgical stage, grade, and patient age to be significant predictors (p<0.001) for 5-year survival and recurrence (Table 23.6). The recurrence rates were 6.1% for low-risk and 13.1% for high-risk patients, as defined above. Shumsky et al. [86] found in their 435 patients a recurrence rate of 4.1% in low-risk and of 23.4% in high-risk patients. Therefore, we propose, together with Greven et al. [84], Shumsky et al. [86], Garcia-Domenech et al. [83] and Rakowsky et al. [85], that low-risk patients do not need to be maintained on a close, routine follow-up and that an intensified schedule ought to be used for high-risk patients.
Table 23.6 Frequency of recurrences and prognostic factors in 2533 patients with endometrial carcinoma treated at Freiburg University Hospital, Germany, between 1969 and 1990
VR VPR PR DM All recurrences
2.7% (41)
1.2% (12)
3.4% (51)
1.3% (13)
1.7% (26)
1.5% (16)
5.3% (80)
2.1% (22)
13.1%
6.1% (63)
(198)
"n = 1500. "n = 1033. VR-vaginal recurrence, VPR-combined vaginal-pelvic recurrence, PR = isolated pelvic wall recurrence, DM = distant metastasis.
Further analysis of the data indicates that, in patients with stage I disease, postoperative radiotherapy improved the 5-year survival independently of histological grading (SPSS program, COX-regression). However, this might be due to patient selection and requires further analysis. Postoperative brachytherapy alone had the same effect.
23.3
CONCLUSION
In conclusion, we believe the therapeutic challenge in endometrial cancer continues to be the careful identification of high-risk prognostic features [81] and the individualized use of the respective therapeutic modalities, in particular adjuvant postoperative radiotherapy.
Table 23.5 Published results of the different complication rates following radiotherapy
Variaefo/. [71] Booth by efo/. [76] Kupelianeffl/. [68] Roseetal. [29] Petereitrtfl/. [77] Taghianefo/. [74] Kuceraefo/. [75] Rouanetefo/. [69] Remoteo/. [21] Rauthetfo/. [81] Hugueninefo/. [4] Chaorto/.[70] Rotte [79] MurrellandOrton [64] Pettersonefo/. [49] Calais etal. [39] Norirto/. [65] Onsrudefo/. [6] Chen etal. [67]
Iridium
Iridium Iridium Cesium Cesium
Iridium
Cobalt
10 19 8 11 11 15 7 12
9 3 11 4 7 13 3 14 8 9 12 1.3
(3% grade 3 and 4) Severe
(No fistula)
Preoperative Postoperative
1 fistula
340 Brachytherapy for treating endometrial cancer
REFERENCES 1. Mestwerdt, W. and Kranzfelder, D. (1980) Gegenwartige Gesichtspunkte zur Epidemiologie und Atiologie des Korpuskarzinoms. Mediz. Klin., 75,333-8. 2. Cramer, D.W. (1986) Epidemiologie and statistical aspects of gynecologic oncology. In Gynecologic Oncology, ed. C. Knappand R.S. Berkowitz. 425-43. 3. Ladner, H.A. (1988) Some aspects of afterloading therapy in endometrium cancer. In High Dose Rate Afterloading in the Treatment of Cancer of the Uterus, Breast and Rectum, ed. H. Vahrson and G. Rauthe. Munich, Baltimore, Urban & Schwarzenberg, 40-2. 4. Huguenin, P.U., Glanzmann, C., Hammer, F. and Liitolf, M. (1992) Endometrial carcinoma in patients aged 75 or older: outcome and complications after postoperative radiotherapy or radiotherapy alone. Strahlenther. Onkol., 168, 567-72. 5. Bokhmam, J.V. (1983) Two pathogenetic types of endometrial carcinoma. Gynecol. Oncol., 15,10-17. 6. Onsrud, M., Marthinssen, A.B.L and Brosvik, D. (1992) Post-operative vaginal irradiation for endometrial cancer: a study of late reactions. In International Brachytherapy, Baltimore, 6-8 Sept, 214-18. 7. Pfleiderer, A. (1991) Die Therapie des Endometriumkarzinoms. Geburtsh. Frauenheilk., 51, 787-97. 8. Weiss, E., Arnold-Bofinger, H. and Hirnle, P. (1995) Primare Strahlentherapie beim Korpuskarzinom. Zbl. Gym/to/., 117,481-5. 9. Rotte, K. (1975) Das Endometriumkarzinom. Mediz. Klin., 70,169-73. 10. Vahrson, H. (1991) Die Strahlentherapie des Korpuskarzinoms. In Giessenergyndkologische Fortbildung, ed. W. Kiinzel and M. Kirschbaum. Berlin, Heidelberg, New York, Springer, 163-79. 11. Joslin, CAR (1991) A place for high dose rate brachytherapy in gynecological oncology: fact or fiction? Activity Selectron Brachytherapy Journal, Suppl. 2,3^4. 12. Fletcher, K.H., Rutledge, F.N. and Delclos, L (1980) Adenocarcinoma of the uterus. In Textbook of Radiotherapy, 3rd edn, ed. G.H. Fletcher. Philadelphia, Lea & Febiger, 789-808. 13. Joslin, C.A.F. (1989) Henschke memorial lecture. Endocuriether./Hypertherm. Oncol., 5,69-81. 14. Simon, N. (1977) The role of radiotherapy. In Endometrial Carcinoma and its Treatment, ed. Gray Springfield, IL, Thomas, 81-7. 15. Glaser, F.H., Rauh, G., Grimm, D. etal. (1977) Das Decatron-remote Afterloading mit hoher Dosisleistung in der Kontakt-Curie-Therapie. Radiobiol. Radiother. (Berlin), 18,707-16. 16. Snelling, M. (1977) The use of radium substitutes and of after- and remote loading techniques in intracavitary therapy. Strahlentherapie, 153, 594-7. 17. Nori, D. (1987) Principles of radiation therapy in the management of carcinoma of the endometrium. In
18.
19.
20.
21.
Radiation Therapy of Gynecological Cancer, ed. D. Nori and B.S. Hilar. New York, A.R. Liss, 115-46. Sorbe, B. and Smeds, A.C. (1989) Postoperative vaginal irradiation by a high dose rate afterloading technique in endometrial carcinoma stage \.Acta Oncol., 28,679. Kucera, H. (1990) Therapieformen und Techniken bei der Bestrahlungdes Korpuskarzinoms. In Interdisziplindre Behandlungsformen beim Mammakarzionom und bei gyndkologischen Malignomen, ed. K. Baier, M. Herbolsheimer and 0. Sauer, Wiirzburg, Bonitas-Bauer, 125-9. Ladner, H.A. (1992) Zur Strahlentherapie des Endometriumkarzinoms. In GynakologischeRadiologie, ed. H.A. Ladner, A. Pfleiderer and C. Profous. (eds): Berlin, Heidelberg, New York, Springer, 35-44 and 112-19. Pernot, M., Hoffstetter, S., Peiffert, F. etal. (1994) Preoperative, post-operative and exclusive irradiation of endometrial adenocaranoma.Strahlenther. Onkol., 170, 313-21.
22. Simon, N. and Silverstone, S.M. (1976) Afterloading with miniaturized 137-Cs sources in the treatment of cancer of the uterus. Int.J. Radiat. Oncol. Biol. Phys., 1,1017-21. 23. Brier, M., Herbolsheimer, M. and Sauer, 0. (eds.) (1990) Interdisziplindre Behandlungsformen beim Mammakarzinom und bei gyndkologischen Malignomen. Wiirzburg, Bonitas-Bauer, 29-54. 24. Walstam, R. (1977) Strahlenphysikalische Voraussetzungen derferngesteuerten Afterloadingbestrahlung. Strahlentherapie, 153,802-6. 25. Bauer, M., von Fournier, D., Fehrentz, F., Kuttig, H., zum Winkel, K. and Neidner, F. (1981) Afterloading-Methode zur Simulation der intrauterinen Packmethode beim Korpuskarzinom. Strahlentherapie, 157,793-800. 26. Bjornsson, M. and Sorbe, B. (1982) Intracavitary irradiation of endometrial carcinomas of the uterus in stage I using a'bulb technique'. Physical basis. Br.J. Radial., 55, 56-9. 27. Aldrich, J.E., Filbee, J.F. and Tompkins, M.G. (1989) Flexible applicators for the treatment of gynaecological malignancies, early clinical results. In Brachytherapy 2, ed. R.F. Mould. Leersum,The Netherlands, Nucletron International, 272-6. 28. Perez, C.A., Disaia, P.J. and Knapp, R.C. (1985) Gynecologic tumors. In Cancer: Principles and Practice of Oncology, ed. V.T. DeVita, S. Hellman and S.A. Rosenberg. Philadelphia, J.B. Lippincott, 1013-81. 29. Rose, P.G., Baker, S., Kern, M. etal. (1993). Primary radiation therapy for endometrial carcinoma, a case controlled study. Int.J. Radiat. Oncol. Biol. Phys., 27, 585-90. 30. Kummermehr, J. (1990) Strahlenbiologische Grundlagen der Brachytherapie. In Interdisziplindre Behandlungsformen beim Mammakarzinom und bei gyndkologischen Malignomen, ed. K. Baier, M. Herbolzheimerand 0. Sauer. Wurzburg, Bonitas-Bauer, 29-54. 31. Orton, C.G. (1987) What minimum number of fractions is
References 341
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45. 46.
required with high dose-rate remote afterloading? Br.y. Radiol., 60,300-1. Fowler, J.F. (1993) Why shorter half-times of repair lead to greater damage in pulsed brachytherapy. Int.]. Radial Oncol. Biol. Phys., 26,353-6. Battermann, J.J. (1992) Dose rates of the future. In International Brachytherapy, ed. R.F. Mould. Veenendal, The Netherlands, Nucletron International, 128-31. Kob, D., Kloetzer, K.H., Kriester, A. and Sommer, H. (1988) Prospektive randomisierte Studie zur Fraktionierung der Afterloading-Therapie von Zervix-und Endometriumkarzinomen. StrahlentherOnkol., 164, 708-13. Rotte, K. (1985). Klinische Ergebnisse der AfterloadingKurzzeittherapie im Vergleich zur Radiumtherapie. Strahlentherapie, 161,323-8. Glassburn, J.R., Brady, LW. and Grigsby, P.W. (1992) Endometrium. In Principles and Practice of Radiation Oncology, 2nd edn, ed. C.A. Perez and LW. Brady. Philadelphia, Lippincott, 1203-21. Eifel, P., Ross, J., Hendrickson, M., Cox, R.S., Kempson, R. and Martinez, A. (1983) Adenocarcinoma of the endometrium. Analysis of 256 cases with disease limited to the uterine corpus: treatment comparisons. Cancer, 52,1026-31. de Waal, J. and Lochmuller, H. (1982) Praoperative Kontaktbestrahlung beim Endometriumkarzinom. Geburtsh. Frauenheilk., 42,394-6. Calais, G., Vitu, P., Descasmps, G. etal. (1990) Preoperative or postoperative brachytherapy for patients with endometrial carcinoma stage I and II. Int.}. Radial Oncol. Biol. Phys., 19,523-7. Lanciano, R.C.,Curran, W.J., Greven, KM. etal. (1990) Influence of grade, histologic subtype, and timing of radiotherapy on outcome among patients with stage II carcinoma of the endometrium. Gynecol. Oncol., 39, 368-73. Calkins, J.D., Stehman, F.B., Sutton, G.P., Reddy, S., Horn back, N.B. and Ehrlich, C.E. (1986) Adenocarcinoma corpus et colli: analysis of diagnostic variables. Int. J. Radial Oncol. Biol. Phys., 12,911-16. de Vita, V.T. Jr, Hellman, S. and Rosenberg, S.A. (1993) Cancer. Principles and Practice of Oncology, 4th edn. Philadelphia, Lippincott, 1195-225. Mandell, L.R., Nori, D. and Hilaris, B.S. (1985) Recurrent stage I endometrial carcinoma, results of treatment and prognostic factors. Int.}. Radial Oncol. Biol. Phys., 11, 1103-9. Joslin, C.A.F. (1978) The indications for post-operative radiotherapy. In Endometrial cancer, ed. M.G. Brush, R.J.B. King and W. Taylor. London, Bailliere Tindall, 158-62. Ladner, H.A. (1983) Zur Prognose und Strahlentherapie des Korpuskarzinoms. Radiologie, 23,12-19. AaldersJ., Abeler,V. and Kolstad, P. (1984) Recurrent adenocarcinoma of the endometrium: a clinical and histological study of 379 patients. Gynecol. Onco\., 17, 85-103.
47. Mandell, L, Nori, D., Anderson, L and Hilaris, B. (1985a) Postoperative vagina I radiation in endometrial cancer using a remote afterloading technique. Int.}. Radial Oncol. Biol. Phys., 11,473-8. 48. Marchetti, D.L., Piver, M.S., Tsukada, Y. and Reese, P. (1986) Prevention of vaginal recurrence of stage I endometrial adenocarcinoma with postoperative vaginal radiation. Obstel Gynecol., 67,399-402. 49. Pettersson, B., Jansson, H., Nilsson, A., Schmidt, M. and Stenson, S. (1989) Prophylactic vaginal irradiation with low dose afterloading technique in women with endometrial cancer. In Brachytherapy2, ed. R.F. Mould. Leersum, The Netherlands, Nucletron International, 240-4. 50. Vergote, I., Kjorstad, K., Abeler, V. and Vossli, S. (1991) Postoperative vaginal irradiation by high-dose-rate cobalt afterloading in stage I endometrial cancer: experience from the Norwegian radium hospital. In Therapiedes Endometriumkarzinoms, ed. W. Kleine, H.G. Meerpohl, A. Pfleiderer, C.Z. Profous. Berlin, Heidelberg, New York, Springer, 103-7. 51. Vavra, N., Denson, U., Kucera, \\.etal. (1993) Prognostic factors related to recurrent endometrial carcinoma following initial surgery. Acta Obstel Gynecol. Scand., 73, 205-9. 52. Elliott, P., Green, D., Coates, A. etal. (1994) The efficacy of postoperative vaginal irradiation in preventing vaginal recurrence in endometrial cancer. Int.}. Gynecol. Cancer, 4,84-93. 53. Malkasian, G.D., Annegers, J.F. and Fountain, K.S. (1980) Carcinoma of the endometrium stage I. Am. J Obstet Gynecol., 136,872-88. 54. Hording, U. and Hansen, U. (1985) Stage I endometrial carcinoma, a review of 140 patients primarily treated by surgery only. Gynecol. Oncol., 22, 51-8. 55. Pipard, G., Milsztajn, M., Jaquet, P. and Sentenac, I. (1985) Low dose-rate postoperative irradiation for adenocarcinoma of the endometrium, treating all or part of the vagina. In Brachytherapy 1984, ed. R.F. Mould. Leersum, The Netherlands, Nucletron, 29-31. 56. Randall, M.E., Wilder, J., Greven, K. and Raben, M. (1990) Role of intracavitary cuff boost after adjuvant external irradiation in early endometrial carcinoma. Int.}. Radial Oncol. Biol. Phys., 19,49-54. 57. Haie-Meder, C, Kramar, A., Castaigne, D.etal. (1995) Comparison of two low-dose rates in vaginal prophylactic pre-operative brachytherapy for stage l-ll endometrial carcinoma. Results of a randomized trial. Int.J. Radial Oncol. Biol. Phys., 32 (Suppl. 1), 223. 58. Perez, C.A. and Camel, H.M. (1982) Long term follow-up in radiation therapy of carcinoma of the vagina. Cancer, 49, 1308-15. 59. Curran,W.J.,Whittington, R., Peters, A.J. and Fanning,]. (1988) Vagina I recurrences of endometrial cancer: the prognostic value of staging by a primary vaginal carcinoma system. Int.}. Radial Oncol. Biol. Phys., 15,803-13. , 60. Sears, J.D., Greven, K.M., (1994) Prognostic factors and treatment outcome for
H.
342 Brachytherapy for treating endometrial cancer
patients with locally recurrent endometrial cancer. Cancer, 74,1303-8. 61. Kuten, A., Grigsby, P.W., Perez, C.A., Fineberg, B., Garcia, D.M. and Simpson, J.R. (1989) Results of radiotherapy in recurrent endometrial carcinoma: a retrospective analysis of 51 patients. InlJ. Radial. Oncol. Biol. Phys., 17,29-34. 62. Poulsen, M.G. and Roberts, S.T. (1988) The salvage of recurrent endometrial carcinoma in the vagina and pelvis. Int.). Radial. Oncol. Biol. Phys., 15,809-13. 63.
Kleine, W., Giese, E., Spiegelhalter, E. and Pfleiderer, A. (1994) Das Rezidiv beim Endometriumkarzinom. Arch. Gynecol., 255,996-7.
64. Murrell, D.S. and Orton, C.G. (1988) Survival and complications in stage I carcinoma of corpus uteri receiving post-operative irradiation. Radiother. Oncol., 12,281-8. 65. Nori, D., Merimsky, 0., Batata, M. and Caputo, T. (1994) Postoperative high dose-rate intravaginal brachytherapy combined with external irradiation for early stage
74. Taghian, A., Pernot, S., Hofstetter, S., Luporsi, E. and Bey, P. (1988) Radiation therapy for medically inoperable patients with adenocarcinoma of the endometrium. Int. J. Radial Oncol. Biol. Phys., IS, 1135-40. 75. Kucera, H., Vavra, N. and Weghaupt, K. (1990) Zum Wert deralleinigen Bestrahlungdesallgemein inoperablen Endometriumkarzinoms mittels high-dose rate-iridium192. Geburtsh. Frauenheilk., 50,610-13. 76. Boothby, R.A., Carlson, J.A., Neiman, W. etal. (1989) Treatment of stage II endometrial carcinoma. Gynecol. Oncol., 33,204-8. 77. Petereit, D.G.,Sarkaria,J.N.,Schink,J.efo/.(1995) High dose rate brachytherapy for medically inoperable stage I endometrial cancer. InlJ. Radial Oncol. Biol. Phys., 32, 223. 78. Onnis, A., Marchetti, M. and Maggino, T. (1993) Endometrial cancer treatment today. Eur.J. Gynaec. Oncol., 14,160-3. 79. Rotte, K. (1990) Techniques and results of HDR
endometrial cancer: a long term follow-up. lnt.J. Radial
afterloading in cancer of the endometrium. In
Oncol. Biol. Phys., 30,831-7.
Brachytherapy HDR and LDR, ed. A.A. Martinez, C.G.
66. Greven, K. and Olds, W. (1987) Isolated vaginal recurrence of endometrial adenocarcinoma and their management. Cancer 60, 419-21, 67. Chen, P., Gibbons, S., Vicini, F. etal. (1995) Postoperative high dose rate intravaginal irradiation in endometrial cancer: a safe and effective outpatient treatment. lnt.J. Radial Oncol. Biol. Phys., 32 (Suppl. 1), 225. 68. Kupelian, PA, Eifel, P.J., Tornos, C, Burke, T.W., Delclos, L and Oswald, M.J. (1993) Treatment of endometrial carcinoma with radiation therapy alone. lnt.J. Radial Oncol. Biol. Phys., 27,817-24.
Orton, R.F. Mould. Veenendal, The Netherlands, Nucletron International, 20-4. 80. Boronow, R.C. (1980) Discussion to Malkasian G.D. etal. (•\98Q)Am.J. Obstel Gynecol., 136,885-6. 81. Rauthe, G., Vahrson, H. and Giers, G. (1988) Five-year results and complications in endometrium cancer HDR afterloading versus conventional radium therapy. In High Dose Rate Afterloading in the Treatment of Cancer of the Uterus, Breast and Rectum. Ed. H. Vahrson, G. Rauthe. Munich, Urban & Schwarzenberg, 240-5. 82. Bond, M.G., Workman, G., Martland, J. etal. (1997)
69. Rouanet, P., Dubois,J.B.,Gely,S.and Pourquier, H. (1993)
Dosimetric considerations in the treatment of inoperable
Exclusive radiation therapy in endometrial carcinoma. InlJ. Radial Oncol. Biol. Phys., 26,223-8. 70. Chao, C.K.S., Grigsby, P.W., Perez, C.A. etal. (1995)
endometrial carcinoma by a high dose rate afterloading
Brachytherapy-related complications for medically inoperable stage I endometrial carcinoma. lnt.J. Radial Oncol. Biol. Phys., 31,37-42. 71. Varia, M., RosenmanJ., Halle, J., Walton, L.CurrieJ. and Fowler, W. (1986) Primary radiation therapy for medically inoperable patients with endometrial carcinoma stage l-ll. InlJ. Radial Oncol. Biol. Phys., 13,11-15. 72. Grigsby, P.W., Kuske, R.R., Perez, C.A. et al. (1987) Medically inoperable stage I adenocarcinoma of the
packing technique. Clin. Oncol., 9,41-7. 83. Garcia-Domenech, R.V., InestaJ.M., Asins, E., Aznar. I. and Llixiona,J.(1997) Prognostic factors in endometrial carcinoma: risk groups and adjuvant radiotherapy. Eur.J. Gynaecol. Oncol., 18,164-70. 84. Greven, K., Corn, B., Case, B., Ourser, P. and Lanciano, R. (1996) Do the number of risk factors influence the outcome of patients with endometrial cancer? Proceedings of 38th Annual ASTRO Meeting. IntJ. Radial
endometrium treated with radiotherapy alone. InlJ. Radial Oncol. Biol. Phys., 13,483-8.
Oncol. Biol. Phys., 64, Suppl. 177. 85. Rakowsky, E., Shfelter, T., Fenig, E., Katzenelsen, R. and Sulkes, A. (1997) Prognostic factors in stage 1 and stage 11 endometrial carcinoma. 10th Meeting of Gynaecology.
73. Sorbe, B., Frankendal, B. and Risberg, B. (1989) Intracavitary irradiation of endometrial carcinoma stage
Gynecol. Oncol., 106,253-4. 86. Shumsky, A.G., Brasher, P.M.A., Stuart, G.C.E. and Nation,
I by a high dose-rate afterloading technique. Gynecol.
J.G. (1997) Risk-specific follow-up for endometrial
Oncol., 33,135-45.
carcinoma patients. Gynecol. Oncol., 65,379-82.
24 Low dose-rate brachytherapy for treating cervix cancer: changing dose rate R.D. HUNTER AND S.E. DAVIDSON
24.1
HISTORICAL BACKGROUND
Gynecological radium techniques were developed empirically initially in Paris and Stockholm. The Manchester Radium System was an important and popular development of those systems and the last major revision of this system using radium was described by Tod and Meredith in 1953 [1]. The Manchester System was designed to deliver a constant dose rate to defined points in the paracervical tissues near the cervix, irrespective of variation in the dimensions of the uterus and vagina, and achieved this by differential loading of two standard-sized intrauterine tubes and three intravaginal 'ovoid' sources to allow for normal variations in uterine and vaginal size. In this system, an application was specified in terms of the 'dose' in roentgens delivered at a specified ideal geometry point, which was called Point 'A' (Figure 24.1.) The paracervical tissue lateral to the cervix was considered to be of prime importance in the radiation tolerance of the central pelvic gynecological tissues and, because the dose distribution was inhomogeneous and the gradients steep, a point rather than a volume was defined. Point A is defined as being 2 cm lateral to the central canal of the uterus and 2 cm superior to the mucous membrane of the lateral vaginal fornix, in the axis of the uterus. A second dosage reference point, point B, was denned as being 3 cm lateral to point A and at the same level 5 cm from the midline and was used when
Figure 24.1 The Manchester System. Definition of point A, a point 2 cm lateral to the central canal of the uterus and 2 cm superior to the mucous membrane of the lateral fornix, in the axis of the uterus. Point B is defined as being in the transverse axis through point A, 5 cm from the midline.
344 Low dose-rate brachytherapy for treating cervix cancer: changing dose rate
X-ray therapy was combined with intracavitary therapy (see Figure 24.1). The dose received at point A is therefore a measure of the dose in the paracervical tissue. The cervix, upper vagina, and uterus are remarkably tolerant to radiation, but the tissues in the paracervical region, which include the uterine artery and the ureters, are more vulnerable. It was demonstrated that the expression of radium dosage as a dose at point A showed a correlation with failure of local control and the incidence of late normaltissue damage in the pelvis, thus justifying the choice of this particular reference point for intracavitary dosimetry [ 1 ]. The Manchester System is based on a strict set of rules concerning the intracavitary applicators (in particular, the choice and positioning of the vaginal ovoids, their separation, and the relative strengths of the intrauterine and intravaginal sources). The differential loading patterns ensured that, irrespective of the use of different lengths of intrauterine tubes and of the size of the vaginal ovoids, a constant dose rate is achieved at point A in ideal geometry and, therefore, provided the applicators were placed correctly, in the patient. There was an acceptance that, in practice, the dose rates would vary by up to 10% of that prescribed in individual patients, but clinical studies on control and morbidity have always used this philosophy. Radium was far from the ideal material for intracavitary therapy. The ideal radionuclide should produce only a single gamma ray spectrum with an energy of around 0.5 MeV. Below this level, attenuation, rather than the inverse square law, dictates tissue absorption and consequently influences dose distribution. At energies above 0.5 MeV, radiation protection becomes increasingly difficult and expensive. The nuclide should have a long halflife, be cheap and easily produced, preferably solid, and with stable solid decay products and, in particular, have a high specific activity. The closest to the ideal radionuclide for low-dose intracavitary therapy at the present time is radioactive cesium-137. It remains the best radionuclide for remote after-loading systems operating at dose rates up to eight times those of the old radium system. In intracavitary brachytherapy, the dose rate is defined by ICRU 38 as 'at the point or surface where the dose is prescribed.' In the classical radium systems, this was typically at Point A and lay between 0.4 (Paris), 0.53 (Manchester) and 1.0 Gy h'1 (Stockholm). ICRU 38 [2] chose to include dose rates up to 2 Gyh-1in low dose rate (LDR) and the studies in changing dose rate described in this chapter involve, therefore, a change from one LDR to a higher LDR and do not quite reach medium dose rate (MDR). In Manchester, dose rates of 1.4-1.8 Gy h'-1 have been used and, by definition, this constitutes LDR treatment. Preloaded applicators gave rise to significant levels of radiation exposure to the radiotherapist, the technicians who prepared the sources, and nursing staff in the theater and the wards. These sources had to be placed in
the uterus and vagina as swiftly as possible and, occasionally, due to disturbed geometry, had to be removed and re-introduced on another day. Alternatively, although the applicator's position was less than perfect, the radiotherapist might have decided to accept a compromise in the treatment rather than repeat the procedure. This requires experience and judgment and sometimes the radiotherapist accepted a less than satisfactory technical insertion for other clinical reasons. This solution to the problem gave rise to the concept of afterloading, in which empty applicators are introduced initially without the need for undue hurry and/or compromise. Only when the geometric relationship of these applicators is known to be ideal are the radioactive sources inserted into the applicators. These small-caliber tubes must be lightweight for the patient's comfort and capable of easy sterilization. They should not be adversely affected by exposure to gamma radiation, and there should be minimal attenuation of gamma rays by the walls of the applicators. Many afterloading techniques have been devised and, in general, they are described as manual or remote afterloading. As the term implies, manual afterloading involves the manual insertion of radioactive sources into suitably placed applicators, after verification of the geometry of the applicators. This is an advantage to the patient and to the radiotherapist, but does very little to reduce the radiation exposure to the nursing staff. This problem can be overcome by remote afterloading, for which the empty applicators have been placed in position and the radioactive sources are remotely introduced from a protective safe. A variety of sophisticated remote afterloading machines have been developed, each offering a range of dosimetric options. In Manchester, an LDR remote afterloading system has been preferred in which the source strengths and geometry allow radium-like fractionation and dose distributions to be maintained. In this system, the applicators (Figure 24.2) are inserted under a general anesthetic in theater and the conscious patient is then transferred to the shielded treatment room. The empty applicators are then attached to the afterloading machine. Sources can be transferred between the safe and the applicators appropriately by remote control from outside the room by trained nursing staff.
24.2 PRETREATMENT ASSESSMENT AND INVESTIGATIONS Radiotherapy plays a significant role in the management of patients with all stages of carcinoma of the cervix. The results of radiotherapy in early-stage cervical carcinoma have always been considered comparable to those of Wertheim's hysterectomy. In practice, radiotherapy can be available for almost all patients, whereas surgical
Pretreatment assessment and investigations 345
the vaginal walls are superior to any imaging modality in determining the presence of vaginal extension. General clinical examination should assess the patient's suitability for general anesthesia and for fractionated externalbeam radiotherapy.
24.2.1
Figure 24.2 Remote afterloading applicators: (a) standard applicators, central tube with fixed flange and small avoids arranged as in treatment; (b) central tube with vaginal cylinder to replace avoids in tandem; (c) manual insert to check the position of individual sources after insertion of applicators.
A chest X-ray should always be done. It is unlikely to reveal metastatic disease, but may help in pulmonary and cardiac assessment for anesthesia and may be useful long term in the assessment of any changes that appear. Magnetic resonance imaging (MRI) or computerized axial tomography (CAT) is not indicated for FIGO staging of carcinoma of the cervix [9], but one of them will often give useful information for planning purposes. MRI scanning, and particularly T2-weighted images, gives better demonstration of the tumor and its local extension in the pelvis than CAT scanning, though there remains a difficulty in demonstrating early parametrial or vaginal involvement [10]. It is also better for demonstrating intrapelvic nodes. Both techniques give similar results for detecting enlarged lymph nodes outside the true pelvis. MRI scanning gives a larger field of view and has the advantage of multiplanar imaging. The disadvantages are that MRI scanning is time consuming and some patients cannot tolerate this examination due to claustrophobia. Whatever the final choice of investigation, the para-aortic nodes, kidneys, and ureters must also be examined.
24.2.2 treatment necessitates selection of patients, even in those with early disease. Close collaboration with gynecologists is essential to ensure the optimum approach for each patient. Probably the most important aspect of the treatment is accurate staging to determine whether the primary tumor is limited to the cervix or whether it has spread beyond the cervix, particularly into the parametrium or toward the bladder or rectum. It is also important to know whether the pelvic lymph nodes are involved, as this determines the overall approach to management. In general clinical practice, FIGO tumor stage remains the most important prognostic factor for disease outcome, closely followed by tumor volume and lymph node status [3-8]. At examination under anesthetic, inguinal, vaginal, and rectal examination should be performed. This will allow the extent of involvement of the cervix, fornices, and vagina to be determined. The uterus is also assessed for size and mobility. Rectal examination gives a more accurate assessment of the parametrial tissues, especially with regard to the extension of disease into the lateral pelvis. Rarely, rectal infiltration of the mucosa may be noted. Direct visual inspection and palpation of
Imaging
Biopsy
Usually the diagnosis is obvious, but a simple punch biopsy will provide histological proof of malignant disease. Biopsy proof should be considered mandatory in any patient fit for radical therapy.
24*2*3
Full blood count
Because of the regularity of bleeding, anemia is often present. The hemoglobin of patients on treatment should be maintained above 10.5 g dl-' and should, if necessary, be corrected by blood transfusion prior to commencing radical radiotherapy. A significantly raised white blood count may indicate a pyometrium, pelvic abscess, or large infected primary tumor. Drainage of pelvic infection and treatment with broad-spectrum antibiotics may be necessary prior to treatment.
24*2*4
Biochemical profile
A biochemical profile may give early indication of or confirm impaired renal function.
346 Low dose-rate brachytherapy for treating cervix cancer: changing dose rate
24.2.5
Cystoscopy
Cystoscopy is carried out for patients with locally advanced disease, particularly that involving the anterior fornix or the anterior vagina, if any urinary symptoms are present and if MRI scanning suggests bladder wall involvement.
243
TREATMENT METHODS
Over the past 10 years, and especially in the last 5 years, there has been a decline in the numbers of patients referred to the Christie Hospital, Manchester, for primary radical radiotherapy. This has been due, in part, to the decline in the incidence of invasive cervical carcinoma within the region (as in the whole of the UK) and also to a change in gynecological practice. Fewer patients are referred with FIGO stage IB and IIA disease as these patients are being treated by primary surgery. The majority of patients referred for consideration of radiotherapy now have locally advanced disease. This means that, in contrast to the situation 20 years ago, only a small number of patients are treatable with brachytherapy alone. The majority are treated with external-beam radiotherapy followed as quickly as possible by a single intracavitary treatment. The anatomy of the cervix, the ease with which applicators can be placed in the uterus and vagina, and the high tolerance of the vagina, cervix, and uterus to radiation make this an ideal site for intracavitary therapy. Because of the rapidly falling gradient of radiation dose laterally towards the pelvic side wall from the central radiation source, it is necessary to supplement the lateral pelvic dose in any patient whose primary disease is more than 4 cm diameter (stage IB2 [ 11 ]) or involves the parametrium by external-beam radiation. Carcinoma of the cervix regularly metastasi/es to the pelvic lymph nodes (parametrial, obturator, hypogastric, external iliac, common iliac, and presacral nodes) and then in a sequential manner to the para-aortic nodes. The potential lymph node drainage areas present a considerable volume of tissue and the dose which can be given is limited by the many normal tissues involved, particularly the tissues of the gastrointestinal tract. In a series of prospective clinical trials between 1949 and 1980 [12,13] at the Christie Hospital, it was demonstrated that, unless there are problems with severe hemorrhage from the tumor when an initial hemostatic limited intracavitary treatment should be carried out, external-beam radiotherapy should be given first. This approach allows all tumors time to regress and improves the geometry of the insertions in the majority of patients. External-beam radiotherapy can be homogeneous or the central pelvis can be shielded for part or all of the treatment, but if shielding is used, it should be
designed carefully to try to achieve some matching with the intracavitary dosimetry [14]. The acute reactions of smaller-volume external-beam treatments, e.g., true pelvis, are better tolerated and are therefore easier to integrate with intracavitary techniques while maintaining overall treatment times. Overall treatment time has recently been shown to be of importance in the treatment of carcinoma of the cervix by radiotherapy. Treatment prolongation results in an approximately 1% loss of disease control per day beyond 30 days [15]. As a consequence, it is recommended, in recently published UK guidelines, that patients with carcinoma of the cervix do not have their treatment prolonged unless there are good clinical reasons [16] for doing so. The dose to points A and B from both treatment modalities can be used as a general guide to matching the different treatment schedules, though simple addition of intracavitary and external-beam dosages is not reliable in the assessment of tolerance, particularly when different (non-classical) dose rates are being employed.
243*1
External-beam techniques
In Manchester for four decades a technique was employed for patients with small-volume disease and parametrial involvement which used a parallel anteroposterior (AP) pair of megavoltage fields in which a special, graduated, vertical central wedge was placed in the beam. This wedge allowed 50% of the prescribed dose to point A and 100% to the pelvic side wall (point B). The dose given was 3250 cGy in 16 fractions over 21 days. This was then followed by two insertions 7-10 days apart, giving 6000 cGy to point A. In a trial of this treatment versus a homogeneous pelvic external-beam technique with no central shielding, no significant difference was seen in survival, pelvic disease control, or morbidity between the two different philosophies [13]. Nowadays, for patients in whom enlarged pelvic lymph nodes are demonstrated by MRI or CT, an external-beam technique is used which treats the whole pelvis and its lymph node drainage areas from below the lowest extent of the disease in the vagina to the top of the fifth lumbar vertebra. It is a four-field technique and the AP fields are rectangular or hexagonal in shape. It can be designed to cover the primary tumor and nodes whilst reducing, where possible, the total volume of tissue to be irradiated. The lateral fields in the four-field box technique are the same length as the AP fields and are 10-11 cm in width. An 8-20 MeV linear accelerator is used for this treatment and no pelvic shielding is employed. The dose given to the pelvis is 4000 cGy in 20 fractions over 28 days, giving equal contributions from the AP and lateral fields. The intracavitary dose used is 3250 cGy, with a single remote afterloaded cesium insertion (formerly
Treatment methods 347
3750 cGy with radium) to point A as soon as possible after the external-beam treatment. For patients with bulky central disease and no evidence of nodes on scanning, an alternative, smaller fourfield box is used to treat the pelvis homogeneously. This treatment is set up isocentrically to cover the true pelvis with a 14-15 x 10-12 cm AP parallel pair and with a 9-10 x 10-12 cm lateral parallel pair. The size of the fields depends upon the extent of the primary tumor and the size of the pelvis. The dose given is 4000-4500 cGy, depending on the volume employed, and it is given in 20 fractions over 28 days. This is followed by a single intracavitary insertion giving 2000-2250 cGy to point A (previously 2500 cGy using radium). There is no proven advantage to any particular one of these treatment techniques, but the choice is made according to the age of the patient, size of the tumor and evidence of nodal disease, fitness for anesthesia, and suitability for intracavitary treatment.
24.3.2 The intracavitary technique Utilizing lightweight metal applicators mimicking the size and shape of the radium applicators, the distribution of 40 mCi cesium-137 pellets in the afterloaded applicators of the new afterloading system has been arranged so that the resulting isodose patterns are as near identical as possible to those of the standard Manchester Radium System. This LDR system has been giving a dose rate of 1.4-1.8 Gy h-1, i.e., 2.9-3.5 times the dose rate of the radium treatment, depending on the decay in the activity of the sources. In order to prevent increased morbidity with this increased dose rate, it is now known, on the basis of clinical trials, that the actual dose must be decreased [17]. The trials in the 1980s were set up as randomized prospective trials with the principal aim of isolating the dose rate factor as an independent variable. The doses now employed in Manchester (2000), are those which have been shown to give the best disease control with acceptable treatment morbidity and these generally employ a dose reduction of 10-17% (see below).
24.33
Low dose-rate remote afterloading
The main advantage of remote afterloading LDR techniques for intracavitary treatment, compared to using radium or cesium-137 preloaded sources, is the reduction of exposure to the medical and nursing staff in theaters and the nursing staff on the wards. Using a remote afterloading LDR system such as that described above also shortens the treatment time significantly for the patients in comparison with the times for treating patients with conventional radium or cesium137. Another advantage is that it is easier with rigid applicators to attain and hold better geometry of the
insertions compared to preloaded sources. It is easier, however, to cause physical damage to the patient using rigid applicators by stretching the uterine and vaginal tissues or to produce vaginal mucosal tears when inserting gauze packing around the applicators. Compared with high dose-rate (HDR) intracavitary brachytherapy, LDR patients, who are now treated in one insertion, do not have the repeated procedures including sedation/anesthesia when treatments are given in multiple fractions, in practice often three to 12 times the number of classical LDR fractions.
243.4
Treatment planning
Treatment must be planned according to the extent of the disease because this determines the minimum volume of tissue to be irradiated. The presence of any illness which may impair the patient's tolerance to radiation, e.g., diabetes, hypertension, or previous major pelvic surgery, may influence the decision regarding radiotherapy technique and dosage. In the planning of this treatment, the two most important factors are the volume of the tumor and the tolerance of the normal tissue in the radiation field. As stated above, small-volume disease - stage I and smallvolume stage IIA - can be treated with intracavitary therapy alone in two insertions 7-10 days apart. The dose given is 6500 cGy (formerly 7500 cGy with radium). This dose should be reduced to, for example, 6250 cGy or 6000 cGy in those patients who have undergone a recent cone biopsy as the particular susceptibility of this group to late morbidity was confirmed in the afterloading trials (see below). For more advanced disease, a combination of external-beam and intracavitary treatments is used, as outlined in the preceding section. 243.5 Treatment The afterloaded applicators are inserted in theater with the patient under general anesthetic, or spinal anesthetic if the patient's medical condition makes general anesthesia inadvisable. A self-retaining catheter is placed into the bladder after drainage of urine. An equal volume of Urografin 150 (30%) is diluted with sterile water and the urinary catheter balloon is inflated with 7 ml of the 1:2 diluted Urografin. The cervical canal is found by gentle probing and dilated to Hagar dilator Number 6 or 7 to allow the insertion of the uterine applicator tube. An inert gold seed marker is placed into healthy tissue in the cervix or upper vagina, which allows the position of the applicators relative to the cervix to be checked radiographically at any time during the subsequent treatment. The uterine canal length is measured with the uterine sound and the appropriate intrauterine tube (4 cm or 6 cm with a flange) is inserted. The majority of patients are treated using an intrauterine tube angled at
348 Low dose-rate brachytherapy for treating cervix cancer: changing dose rate
40° to the vagina and fitted with a fixed flange which can sit at the cervix without obscuring the view of the vaginal vault (see Figure 24.2). The deliberate angle in the tube draws the uterus, in most patients, into a central position in the pelvis away from the pouch of Douglas, the sigmoid colon, and the anterior rectal wall. The uterine tubes initially employed (Fletcher type) had varying degrees of curvature of the intrauterine section. If the curvature is small, the uterus is forced by the applicator to become partially retroverted and the fundus comes in close proximity to the sigmoid colon. Such tubes may have contributed to bowel damage by bringing the sources closer to the large bowel [18]. The vaginal source tubes are mounted into ovoids of dimensions similar to those of the rubber preloaded applicators (small, medium, and large) or 'long small' ovoids, which have an extra 0.5 cm of packing built into the ovoid. If these 'long small' ovoids do not fit comfortably, this means that there is not enough room for the normal small ovoid and satisfactory packing at the vault. The Manchester afterloading applicator includes a screw clamp on the intrauterine tube which, when tightened, draws the three applicators into an ideal physical position. It is possible to use this applicator in about 95% of patients. For patients with a fixed retroverted uterus, this applicator can be replaced with one with a moveable clamp which allows the uterine tube to be rotated through 180° and also allows the intrauterine length to vary. The use of this flexible applicator requires considerably more skill and can create difficult decisions about dosage and dose distribution. Gauze packing is placed firmly and carefully behind the ovoids, anteriorly between the ovoids and the base of the bladder, and around the applicator tubes down to the level of the introitus. In most patients, this packing will keep the applicators in position. For those patients with a small vaginal vault or extension of disease down the vagina, a single-line source is used. The intrauterine tube is loaded below the cervix and down to the introitus if required, to give isodoses which approximate to those of small radium sources in tandem. The vaginal walls are distanced from the tube by plastic cylinders, with a choice of diameter, including cylinders with extra in-built packing of 0.5 cm both anteriorly and posteriorly, which ensure that the vaginal mucosa does not come in contact with the applicator tube itself. To prevent displacement of the applicator during the treatment, a suture is normally placed at the introitus when a single in-line source is used. At the end of positioning the applicators, a measurement is made of the dose coming through to the anterior rectal wall. An insert mounted with cesium-137 pellets and simulating the treatment positions is placed in the applicators and a reading is made of the anterior rectal dose by exploring the lower rectum with an ionization chamber. The dose rate, as measured by the chamber in contact with the anterior rectal wall, is kept below two-
thirds of the dose rate to point A. If the reading is higher than two-thirds of the point A dose rate, then the gauze packing is removed and replaced and a further assessment of the anterior rectal dose is made. If the readings remain high despite repacking, an adjustment to the treatment may be necessary. The urinary catheter is drawn down and secured against the applicator stems to ensure that the balloon of the catheter is located at the bladder neck. Urine is drained and 120 ml of air is inserted into the bladder. AP and lateral radiographs are taken to check the position of the applicators with dummy trains of sources and markers in place, before the patient returns to the ward. Direct measurements of bladder dose have never established a place in Manchester. A study of bladder base dose was carried out on a series of 20 patients undergoing intracavitary treatment using CAT scanning and this scan information was transferred to a treatment planning computer [19]. The ICRU bladder reference dose was calculated from the radiographs. This study showed that the ratio of the maximum bladder base dose to the ICRU reference dose rate varied considerably, from 1.01 to 3.59. In half the patients, the maximum bladder dose was not in the midline. The variation in the size and position of the bladder relative to the applicators in the uterus and the vagina is very considerable, due to different lengths of vagina, vault capacity, and tumor bulk. The ICRU reference dose never represented the maximum bladder dose in this study and no other single fixed point within the bladder could be identified. All patients treated with radical intracavitary treatments alone, and certainly any patient for whom the treating doctor had difficulty with the placement of the uterine applicators, have a CAT scan carried out to check the position of the applicators. This principally looks at the intrauterine tube to see its position within the uterus and to ensure that no perforation has taken place. It is also possible to determine if the tube is eccentrically placed within the uterus and if the uterine wall is thinned over the tip of the applicator tube. If a perforation is seen, the applicators are removed immediately, the patient is commenced on antibiotics, and a further attempt at intracavitary treatment is carried out after 7-10 days. If the tip of the uterine tube appears to thin the uterine wall, the active pellet loading can be altered and sources removed from the upper uterus. In a series of 650 patients scanned after an apparently straightforward insertion, unexpected uterine perforation was seen in 3% of patients on CAT scan [20]. In the treatment room, when the patient has fully recovered from the anesthetic and the CAT scans show satisfactory positioning of the applicators, the intracavitary applicator tubes are linked to the cesium-137 line sources in the protective safe by transfer tubes. The remote afterloading system used at present utilizes cesium-137 as 1.5 mm glass spheres incorporated in stainless-steel spheres of 2.5 mm diameter. The maxi-
Treatment results 349
mum strength is 150 mBq. The actual treatment 'sources' are made from a combination of active and dummy spheres arranged in a line (see Figure 24.2) and there is no difficulty in producing applicator loading patterns that will reproduce the dose distributions of the Manchester Radium System. The source pattern construction takes place within the protected machine and the selected source 'lines' are transferred to the intrauterine and vaginal applicators by compressed air. The reliability of this two-way transfer system is very important. In practice, angulations up to 90° can be negotiated effectively by the pellets. The patient is nursed on her back or side during the treatment and, while attached to the machine, movement is permitted under supervision. Traction on the treatment lines has to be prevented as this could displace the applicators. A marker system to identify the treatment applicator position relative to the patient's thigh is used as an extra safeguard (Figure 24.3). During treatment, which may last for 24 h, trained nursing staff supervise the equipment. This system allows regular treatment interruptions for nursing care. The movement of the radioactive sources is controlled completely from outside the treatment room once the
Figure 24.3 Afterloading applicators in position. The position of the flange relative to the patient is drawn on the thighs as illustrated.
computer has been programmed and set. A record of the treatment interruptions is made by the machine to ensure that the patient has the correct treatment time before it is terminated. Present clinical experience has shown that preloaded radium and cesium-137 insertions can be safely replaced by remote afterloaded cesium-137. This eliminates all the hazards of preloaded sources and, using higher dose rates than those of the Manchester Radium System, has practical advantages for the patient and the hospital. From the randomized prospective trials carried out in Manchester, the intracavitary dose must be reduced to allow for the increased biological effectiveness of the higher dose rate treatments when dose rates three to four times that of the radium treatments are employed. The correction factor used is a minimum of 10%, but may be as much as 20%, depending on the clinical situation. This is based on the treatment results from the Selectron trials (see next section).
24,4
TREATMENT RESULTS
Acute radiation reactions (within 2 months of treatment) and late pelvic changes are inevitable with radical radiotherapy to the pelvis. In general, the likelihood of complications increases with large treatment volumes and high-dose treatments, but some changes of bowel and bladder function are seen in all patients. Bowel symptoms depend on which area of bowel is affected; the terminal small bowel, sigmoid colon, rectum, or all levels may be affected. During the second half of treatment and for 6 weeks following treatment, patients experience frequency and irritability of the bowel, which may produce varying severities of diarrhea, tenesmus and, very rarely, acute perforation. The majority of patients can control these symptoms with bulk aperients and antidiarrhea therapy, but occasionally patients' treatment may need to be halted. Mild cystitis producing dysuria and frequency may occur and this usually settles by 6 weeks following treatment. All patients who are premenopausal will have an induced menopause because of the radiation dose to the ovaries. Hormonal replacement therapy can be given safely to premenopausal patients following radiotherapy to the pelvis. Acute radiation reactions may involve only the upper vagina or may involve the whole length if this has to be treated. Adhesions will form in the acute phase if these are not prevented. Our patients are advised to use a vaginal dilator daily for the first 6 weeks after treatment and then two to three times a week over the next 3 months to avoid permanent vaginal fusion. Local estrogen cream may be used to help delayed healing at the vaginal vault. All normal tissues included in the high-dose pelvic treatment volume may show late effects. The rectum,
350 Low dose-rate brachytherapy for treating cervix cancer: changing dose rate
sigmoid colon, and bladder are most commonly affected. Bowel complications may appear from 6 months to 10 years after treatment and bladder complications from 18 months to 7 years. The effects that appear earlier are usually hemorrhagic and necrotic and the later effects are fibrotic and hence stenotic. Much of the morbidity changes and symptoms can be controlled conservatively, but symptoms that do not resolve or are sufficiently severe require surgery. Late complications are sometimes also seen on the anterior rectal wall close to the vaginal source position. Bleeding per rectum may settle with topical steroids and laxatives, but heavier or chronic bleeding may be a problem and may even require surgical resection. Sigmoid colon injury may give rise to lower abdominal or left iliac fossa pain and intermittent constipation, and increasing stenosis may lead to obstruction and thus surgery. Bladder injury is usually seen on the posterior wall in the midline close to where the vaginal and uterine sources were placed. A few patients may have generalized changes in the bladder leading to contraction of the bladder or hemorrhagic changes. Chronic bleeding may require cautery, but if the fibrotic process is severe, reducing the bladder volume to less than 200 ml, surgery may be required to augment or remove the bladder [14].
24.5 CLINICAL TRIALS OF DOSE-RATE CORRECTION FACTORS IN CARCINOMA OF CERVIX
24.5.1
Intracavitary therapy alone
A total of 563 patients with stage I and small-volume Ha disease were entered into randomized prospective trials in Manchester comparing manually preloaded radium/ cesium (dose rate 0.53 Gy h"1) with afterloaded cesium137 using the Selectron LDR after-loading system, as described above (dose rates 140-180 cGy h"1) from 1980 to 1988. The actual corrections employed were 0%, 6.5%, 13%, 20%, and 10%. All other treatment parameters were controlled between the groups. When considering randomized dose, the mature overall survival at 5 years for all patients shows no significant difference between groups (p = 0.56). From Figure 24.4 it can be seen that the 5-year primary disease-free survival of the group of patients with small-volume disease ranges from 96% to 84% across all the four phases of the trial, with the results showing a significant fall in primary disease-free survival only with the 20% reduction in dose (6000 cGy to point A).
Figure 24.4 Graph of primary disease-free survival for 564 patients treated with Intracavitary sources In the trials of remote afterloading versus preloaded sources in Manchester. The significance of the dose received in all the trials is due to the significant decrease in survival in the group who received 6000 cGy. Each dose group is indicated by the different lines, as shown in the key, with numbers of patients in parentheses. (With permission of all the consultants in the Clinical Oncology Department, Christie Hospital, Manchester, and those involved with the treatment of the gynecology patients: Dr R.D. Hunter, Dr R. Stout, Dr E. Sherrah Davies, Dr M.L Button, Dr R. Gattamaneni, and Dr B. Hancock.)
Clinical trials of dose-rate correction factors in carcinoma of cervix 351
Morbidity for the different dose-rate correction arms is shown in Figure 24.5. These results were updated in 1993, with minimum follow-up of 5 years, and graded using the Franco-Italian Glossary, which is now used in many European centers to record morbidity for all patients receiving radiotherapy for cervical carcinoma [21]. The randomized dose was the major factor in the morbidity experienced by these patients. The morbidity can be seen to be ranked by dose as expressed at point A utilizing the Manchester philosophy (see above). The majority of morbidity was evident by 2 years posttreatment, though some further bladder morbidity was seen beyond 5 years post-treatment. Modeling the relationships between dose and (a) primary control of tumor, and (b) grades 2, 3 and 4 morbidity has allowed an estimate of the optimum dose-rate correction factor to be made from the clinical data. The optimum estimated dose is close to 6500 cGy, a dose-rate correction of 13.3%. This dose, in the trials, gave 5% serious (grade 3/4) late morbidity with no significant decrease in overall and disease-free survival. This level of serious morbidity is not significantly different from that experienced by the patients treated with classical doserate cesium-13 7/radium at the same time.
disease with external-beam therapy followed by intracavitary therapy. This trial of megavoltage parametrial radiotherapy given with two intracavitary insertions was carried out between 1983 and 1989. Patients with bulky stage IB, stage IIB, or stage III disease received parametrial radiotherapy using a graduated wedge in the beam (as described in section 24.3.1, 'External-beam techniques') giving 3250 cGy in 16 fractions over 21 days. A total of 596 patients were randomly allocated to three treatment schedules: classical radium/cesium-137 (6000 cGy H-1), Selectron 5.5 (5500 cGy h-1) and Selectron 5.0 (5500 cGy H-1). The dose-rate corrections are therefore 8.5% and 17%. The mature 5-year results showed no difference in overall survival (p — 0.35) or pelvic disease-free survival (p - 0.35) between the three groups. Morbidity graded prospectively by the FI glossary is shown in Table 24.1. This shows no difference between the three different groups for any level of morbidity. The overall 5-year survival in this study was 65% for stage II and 38% for stage III. The grade 3/4 morbidity in this group of 596 patients was 6.5%. An optimum dose-rate correction factor cannot, therefore, be calculated for this study. A compromise of a value in the range 10-15% is the safest conclusion from these data.
24.5.2 External-beam (wedged) and intracavitary therapy
25*53 External-beam (unwedged, pelvic box type) and intracavitary therapy
In parallel with these very pure intracavitary therapy trials in which primary control and morbidity can only be a consequence of the single treatment modality, the Christie department also conducted a study of dose-rate correction in patients being treated for more advanced
No formal trials have been conducted in this situation, which is unfortunate because the use of intracavitary therapy alone and wedged external-beam therapy is now unusual both in Manchester and internationally. The only contemporary data that we have, come from a
Figure 24.5 Graph of time to complication for all trials treated by randomized dose. The different dose groups are shown as in the key, with patient numbers in parentheses. There is a highly significant association with dose received and complications, p = 0.00005.
352 Low dose-rate brachytherapy for treating cervix cancer: changing dose rate
Table 24.1 The number of patients allocated to each treatment group and the 5-year mature prospective grade 2, grade 3/4, and grade 4 morbidity
Radium/cesium-137 Classical 6.0 Selectron 5.5 Selectron 5.0
201 193 202
retrospective audit in 1997 of 123 sequential patients treated between 1982 and 1992 by a pelvic box technique (4000-4500 cGy in 16-20 fractions in 21-28 days) and followed by a single intracavitary afterloaded 'high LDR' cesium-137 insertion and a 10-20% dose-rate correction. The majority of the patients had stage III disease and were unsuitable for the wedged field approach. This group of patients had a 5-year survival of 37% and a grade 3/4 morbidity rate of 3.2%.
24.5.4 Since 1988 Following the completion of the formal trials, the department has replaced preloaded radium therapy with afterloaded cesium-137 at the higher dose rates described above. Dose-rate correction factors have varied slightly, depending on the technique employed, but they have consistently been maintained in the range 10-20%. The most recent data that we have available were obtained in 1999 and included all cases with cervical carcinoma treated radically at the Christie Hospital in 1993. This analysis is part of a UK national audit of treatment results and morbidity of patients treated in 53 radiotherapy centers in the UK (in press). One hundred and eighty-three patients were identified and clinical outcome data were available for 181 patients. Five-year survival for stage IB disease (n = 45) was 71% (72%), for stage IIB (n = 70) 52% (63%), and for stage III (n = 50) 34% (41%). The numbers in parentheses are the FIGO 5-year survival figures for 1990-1992. The complication rate, grade 3/4 morbidity (all sites), in the patients included in this audit was 3.4%.
24.6
INTERNATIONAL EXPERIENCE
Other series reported in the literature show the possibility of an increase in late radiation damage in groups of patients treated at the higher low and medium dose rates possible with the new afterloading systems when compared with retrospective series [22-24]. The only other randomized prospective trial exploring this particular issue employed a higher dose rate (220 ±10 cGy h-1) and concluded that, while the different levels of correction factor did not influence pelvic control, a dose-rate cor-
7 12 8
13(2) 13(2) 14(0)
rection of 12.5% was insufficient when moving from classical dose rates to this level, and that a correction factor of 'around 30%' was associated with an acceptable level of morbidity [25]. Dose-rate correction factors between these values were not explored. In practice, the majority of users of high LDR and MDR brachytherapy have evolved their own schedules to compensate for the dose-rate effect. In many situations this has involved changes in dose distributions and fractionation as well as in dose rate [26]. These experiences have led some authors to suggest that the linearquadratic (LQ) model can now be used to predict the optimum dose and schedules that may be employed [27]. This remains to be proven in clinical practice. One hundred years after the discovery of radium, gynecological intracavitary radiotherapy systems have contributed to the cure of many patients. The modern systems are more controlled, precise, and safer, particularly from the point of view of the staff treating the patient, than the original techniques, but evidence of therapeutic gain is not clearly identifiable. However, internationally there continues to be an enormous range of systems, modifications of systems, and personal prejudice dictating treatment, with no real evidence of specific benefits from one as compared to another. Treatment in many centers is 'custom and practice' modified with experience and amended by Ionizing Radiation Legislation. The Manchester System of the early twenty-first century is a direct descendant of the original radium system and can be tracked back to it through the successful completion of the series of trials and studies over that period. One direct consequence of moving from the radium system of the 1940s to the present high 'LDR' remote afterloading cesium-137 system has been the demonstration of the importance of carefully controlled dose-rate correction factors. Morbidity continues to dog the new systems, as it did the old ones, and is a consequence of primary tumor damage, the proximity of uninvolved normal tissues, particularly in early disease, radiation dose, and individual radiosensitivity. The modern systems are not worse than their predecessors, but they are also not significantly better than them. Everyone would wish to see total control without morbidity, but realistically that is not achievable, given the starting position with the majority of patients, but technical means to optimize treatment and biological means
References 353
to increase the efficacy on the tumor and decrease the efficacy on normal tissue need urgent exploring.
radiotherapy treatments in Stage III carcinoma of the cervix. Clin. Radiol., 37,23-7. 14. Fyles, A., Keane, T.J., Baron, M. and Simm, J. (1992) The effect of treatment duration in the control of cervix
REFERENCES
cancer. Radiother. Oncol., 25,273-9. 15. Royal College of Radiologists (1996) Guidelines for the
1. Tod, M.C. and Meredith, W.J. (1953) Treatment of cancer of the cervix uteri - a revised Manchester method. Br. J. Radial,, 26,252-7. 2. ICRU Report 38 (1985) Dose and volume specification for reporting intracavitary therapy in gynecology.
Management of the Unscheduled Interruption or Prolongation of a Radical Course of Radiotherapy. London, Royal College of Radiologists. 16. Stout, R. and Hunter, R.D. (1989) Clinical trials of changing dose-rate in intracavitary low dose-rate therapy. In
International Commission of Radiation Units and Measurements, Bethesda, Maryland, USA. Section 2.1.3. 3. Perez, C.A., Grigsby, P.W., Nene, S.M. etal. (1991) The
Brachy'therapy2, ed. R.F. Mould. Leersum, The Netherlands, Nucletron, 219-22. 17. Sherrah-Davies, E. (1985) Morbidity following low dose-
effect of tumour size on the prognosis of carcinoma of the uterine cervix treated with radiation alone. Cancer, 69,2796-806. 4. Kovalic, J.J., Perez, C.A., Grigsby, P.W. and Lockett, M.A. (1991) The effect of volume of disease in patients with carcinoma of the uterine cervix. Int. J. Radiat. Oncol. Biol. Phys., 21,905-10.
rate Selectron therapy for cervical cancer. Clin. Radial., 36,131-9. 18. Hunter, R.D., Wong, F., Moore, C., Notley, H.M. and Wilkinson, J. (1986) Bladder base dosage in patients undergoing intracavitary therapy. Radiother. Oncol., 7, 189-97. 19. Makin, W.P. and Hunter, R.D. (1988) CT scanning in
5. Lanciano, R.M., Won, M., Coia, LR. and Hanks, G.E. (1991)
intracavitary therapy: unexpected findings in 'straight
Pre-treatment and treatment factors associated with improved outcome in squamous cell carcinoma of the
forward' insertions. Radiother. Oncol., 13(4), 252-5. 20. Hunter, R.D. (1991) Female genital tract. In The
uterine cervix: a final report of the 1973 and 1978 patterns of care studies. Int.J. Radiat. Oncol. Biol. Phys., 20,667-76. 6. Magee, B.J., Logue, J.P., Swindell, R. and McHugh, D. (1991) Tumour size as a prognostic factor in carcinoma of the cervix-assessment by TRUS. Br.J. Radial., 64,812-14. 7. Thorns, W.W., Eifel, P.J., Smith, T.L.etal. (1992) Bulky
Radiotherapy of Malignant Disease, 2nd edn, ed. R.C.S. Pointon. Berlin, Springer-Verlag, 279-308. 21. Chassagne, D., Sismondi, P., HoriotJ.C.^o/. (1993) A glossary for reporting complications of treatment in gynaecological cancers. Radiother. Oncol., 26,195-202. 22. Jones, R.D., Symonds, R.P., Habeshaw, T., Watson, E.R., Laurie, J. and Lament, D.W. (1990) A comparison of remote afterloading and manually inserted caesium in
endocervical carcinoma: a 23-year experience. Int.J. Radiat. Oncol. Biol. Phys., 23,491-9. 8. Collins, C.D., Constant, 0., Fryatt, I., Blake, PR. and Parsons, C.A. (1994) Relationship of computed
the treatment of carcinoma of the cervix. Clin. Oncol., 2, 23.
tomography tumour volume to patient survival in carcinoma of the cervix treated by radical radiotherapy. Br. J.Radiol., 67,252-6. 9. RCR Working Party (1995) Making the Best Use of a Department of Clinical Radiology: Guidelines for Doctors,
N.S. and Lament, D.W. (1996) A linear quadratic analysis of gynaecological brachytherapy. Clin. Oncol., 8,90-6. 24. Newman. G. (1996) Increased morbidity following the introduction of remote afterloading, with increased dose rate, for cancer of the cervix. Radiother. Oncol., 39,
3rd edn. London, The Royal College of Radiologists. 10. Hawnaur, J.M., Johnson, R.J., Buckley, C.H., Tindall, V. and Isherwood, I. (1994) Staging, volume estimation and
193-8. Dodds, D., Symonds, R.P., Deehan, C., Habeshaw, T., Reed,
25.
97-103. Patel, F.D., Negi, P.S., Sharma, S.C., Kapoor, R., Singh, D.P.
assessment of nodal status in carcinoma of the cervix:
and Ghoshal, S. (1998) Dose rate correction in medium dose rate brachytherapy for carcinoma on the cervix.
comparison of magnetic resonance imaging with surgical
Radiother. Oncol., 49,317-23.
findings. Clin. Radiol., 49(7), 443-52. 11. Creasman, W.T. (1995) Editorial: new gynecologic cancer staging. Gynecol. Oncol., 58,157-8. 12. Cole, M.P. (1973) Radiotherapy for cervical cancerX-rays. In Cancer of the Uterine Cervix, ed. E.C. Easson. London, W.B. Saunders, 53-75. 13. Hunter, R.D., Cowie, V.J., Blair, V. and Cole, M.P. (1986) A clinical trial of two conceptually different radical
26. el-Baradie, M., Inoue, T., Murayama, S., Tang, J.T., Yazamaki, H.and Fournier-Bidoz, N.(1997) HDRand MDR intracavitary treatment for carcinoma of the uterine cervix. Strahlenther. Oncol., 173,155-62. 27. Leborgne, F., Fowler, J.F., Leborgne, J.H., Zubizaretta, E. and Curochquin, R. (1999) Medium dose rate brachytherapy for carcinoma of the cervix. Int. J. Radiat. Oncol. Biol. Phys., 43,1061-4.
25 High dose-rate brachytherapy for treating cervix cancer C.A.JOSLIN
25.1
HISTORICAL BACKGROUND
Among those who introduced high dose-rate (HDR) intracavitary afterloading for the treatment of cancer of the cervix were O'Connell et al. in 1965. They used dose escalation studies followed by surgery to carry out an assessment of the histological and acute clinical effects of treatment compared with established radium treatment [1-4]. Others were also developing remotely controlled afterloading techniques from the 1960s using either HDR or dose rates in line with conventional low doserate (LDR) radium, following the introduction of the concept of afterloading by Henschke in 1960 [5-8]. The author's experience of using HDR remote afterloaded source applicators for treating cervix cancer began during the 1960s [1-4] and this treatment was developed in combination with external-beam irradiation for the radical treatment of cervix cancer from 1967 onwards [9,10]. During the 1970s the use of dose rates above those for radium but much lower than those for HDR were introduced and are discussed in Chapter 24. A significant difference between those techniques and HDR was that LDR intracavitary irradiation was not necessarily com-
bined with external-beam irradiation, particularly for early-stage disease, whereas HDR intracavitary irradiation was. The principal reason for this was that to use HDR alone entailed using an impractical number of fractions. Various combined fractionation regimes have been introduced since the 1970s and the treatment results published. Strongly held views regarding the advantages and disadvantages of HDR versus LDR therapies have been expressed. A review in 1990 concluded that the optimal technique and dose fractionation scheme had yet to be established through systematic trials [11]. From 1967, the author used HDR in conjunction with external-beam irradiation on a radical treatment basis to treat patients with early and advanced cancer [9]. The method used followed the established Manchester System, which included a central intrauterine tandem and two small vaginal ovoids. The external-beam irradiation was given using parallel opposed fields and a central wedge, which was tapered to compensate for the fall-off in dose from the intracavitary treatment. The experience gained from treating a large elective group was published in 1972 and with extended follow-up in 1991 [9,10]. Others have reported the results of comparative studies and a few controlled clinical trials
Lymph node status 355
[12-20,59,61,91]. Among one of the early reported series was that of Shigematsu et al., who compared HDR with manual LDR treatment for treating advanced cervical cancer. They reported higher local disease control rates, but also a higher rate of complications for the HDR group when compared to historical LDR treatment [12]. The various publications identified minor differences in results, but no significant major differences were reported. A general review of the published data by Fu and Phillips in 1990 reported that most non-randomized studies showed similar survival, local control, and complication rates using fractionated HDR remote afterloaded intracavitary irradiation combined with external-beam irradiation compared with historical or concurrent LDR controls. It was pointed out that the techniques as well as the dose fractionation schedules used in the different institutions were variable and that the optimal technique and dose fractionation regime had yet to be established through systematic clinical trials. The conclusion drawn was that much clinical research effort was needed to establish the optimal technique for the delivery of HDR afterloaded intracavitary brachytherapy for treating carcinoma of the cervix [11]. That situation remains largely unchanged and is likely to remain so in the immediate future. The purpose of this chapter is initially to consider those aspects of cervix cancer that affect management by radiotherapy and then the application of HDR remotely controlled brachytherapy in combination with externalbeam irradiation for treating early-stage and late-stage disease. It also compares some of the reported results of HDR with LDR treatment.
25.2
AGE GROUPS AFFECTED
Cancer of the cervix in the UK has increasingly become a disease of young and old women. The biphasic age incidence and frequency of the disease have affected management mainly because, in general, the younger woman presents with early-stage disease and the older woman with advanced disease. Whereas both age groups were extensively treated by radiotherapy until the late 1980s, since that time patients with early disease are increasingly receiving surgical treatment.
253
TUMOR SIZE
Patients with squamous cell cancer of the uterine cervix present with a primary tumor that may vary in size from microscopic subclinical disease to an enlarged tumor that may extend into the parametrium, reach the pelvic side wall(s), invade the lower vaginal tissues, and infiltrate the body of the uterus. Invasion of adjacent organs
or tissues such as rectum and bladder may also occur. This wide range of primary tumor extent means that to deliver a curative dose at the periphery of the primary tumor may require external-beam irradiation in addition to intracavitary irradiation. In order to study the effects of treatment and give a prognosis, it is important to classify patients according to the disease extent. In clinical practice, the FIGO system has proven to be of great practical value and remains the most common method of staging. The FIGO staging system involves clinically assessing the extent of the primary tumor with the aid of an intravenous pyelogram, and cystoscopy. However, in addition to these standard assessment procedures, computerized tomography (CT) and magnetic resonance imaging (MRI) scans, transrectal ultrasound (TRUS), and surgical staging are increasingly being used to assess the disease extent. Computerized tomography is now routinely used by some for tumor localization and to assess parametrial infiltration. It can be used to detect hydronephrosis and enlarged lymph nodes. When compared with CT, MRI has been reported to be more accurate for assessing the primary tumor, but comparable for assessing lymph nodes. TRUS has limitations when assessing advanced cancer [21-24,70,72]. These additional assessment aids raise the important question of how best to provide appropriate reporting procedures for staging, especially if used to optimize treatment on an individual patient basis. The size of the primary tumor, including the extent of any local invasive component, determines the type of treatment. Views differ regarding the efficacy of intracavitary irradiation alone to cure tumors more than 3 or 4 cm in diameter and of external-beam irradiation to cure patients with extensive lymph node spread of disease. In general, some form of combined intracavitary and external-beam irradiation has become the treatment of choice.
25.4
LYMPH NODE STATUS
The size of the primary tumor and the incidence of spread to pelvic lymph nodes are related- When the primary tumor is less than 2.0 cm in diameter, lymph node involvement occurs in about 6% of cases, increasing to 18% for tumors greater than 2.0 cm in size. For tumors greater than 3.0 cm in diameter, over 30% of patients have lymph node involvement. A review of the published data of lymph node involvement against clinical stage of disease reported a rate of 19.8% for stage Ib, 36.1% for stage lib, and 42.7% for stage Illb [25]. The lymph node groups most commonly affected are the obturator, internal iliac, external iliac, and common iliac nodes, with the risk of involvement being least for
356 High dose-rate brachytherapy for treating cervix cancer
the last group. The position of these lymph node groups within the pelvis extends from about the lower border of the fourth lumbar vertebra to the lower border of the obturator foramen and laterally to between 1 and 2 cm wide of the inner pelvic brim. Bipedal lymphangiography can distinguish nodes with filling defects and CT can detect enlarged nodes. Neither of these investigations will clearly distinguish enlargement specifically due to cancer, nor will they identify microscopic disease. However, it has been reported that CT scans are of prognostic value for assessing outcome and that the cause-specific disease-free survival is related to lymph node size [26-28,70].
25.5
EARLY-STAGE DISEASE
The alternative treatments, in which HDR brachytherapy may form part, include the following.
25.5.1
Intracavitary irradiation alone
Despite the fact that pelvic lymph nodes may be involved by cancer, LDR intracavitary therapy alone has been an established method of treatment for well over 60 years. Such treatment ignores lymph node status, yet is reported to produce results that are comparable to those of other reported treatments. HDR brachytherapy alone has not become an established alternative method of treatment, mainly due to the practical problem of the number of treatment fractions required in order to restrict the risks of late reacting tissue damage. However, the use of HDR brachytherapy alone has been reported for treating cervical intraepithelial neoplasia (GIN III) lesions [32], and Newman et al. reported treating stage la cervical cancer by HDR brachytherapy alone [74].
25.5.2 Intracavitary irradiation followed by Wertheim hysterectomy The usual technique is to treat the primary tumor giving a dose less than radical followed by surgery. This form of management was in use for many years with radium and LDR afterloading and some would still advocate it. The reason for mentioning it in this chapter is that some of the earliest work using HDR brachytherapy involved efficacy studies by histological assessment of surgical specimens following a few HDR fractions [3,19].
25.5.3 Combined intracavitary and external-beam irradiation When compared with Wertheim's operation, there are advantages but also some disadvantages. Although the
overall survival and morbid effects are comparable to those of surgery alone, many gynecological surgeons now favor surgery alone for the younger woman, and that is a view also supported by many clinical oncologists. However, radiotherapy is still in use as the principal method of treating early-stage disease and a considerable amount of HDR experience has now been reported.
25.6
LATE-STAGE DISEASE
The risk of lymph node involvement in late-stage disease, i.e., stages lib and Illb, has been reported to exceed 30%. Also, depending upon the extent of spread of disease into the parametrial tissues, the outer limits of the primary tumor may receive insufficient irradiation from intracavitary treatment alone to be cured. In general, therefore, treatment is given using a combination of external-beam and intracavitary irradiation. The precise combination will differ from one center to another, but the reported results are similar [11]. The manner of combining the two treatment modalities includes: 1. external-beam irradiation to the pelvis to deliver a homogeneous dose followed by one or more boost doses of HDR intracavitary irradiation to the cervix. 2. external-beam irradiation to the pelvis to deliver treatment designed to compensate for the fall-off in dose from the HDR intracavitary treatment. The external-beam irradiation is delivered through a beam-shaping wedge in order to modify the midpelvic dose distribution to partly compensate for the rapid fall-off in dose from the intracavitary irradiation. The alternative is to use a simple lead block to overlie the position of the intracavitary treatment. This will partially protect the bladder, rectum, and lower vagina from the external-beam effects. Treatment is usually given daily for 4 days each week, and intracavitary irradiation is given once weekly for several weeks [9,10]. It is not recommended that combined treatment should be given on the same day unless there is a minimum gap of 6-8 h between treatments, principally because the total radiobiological equivalent dose may increase the risk of late normal tissue damage.
25.6.1 The case for external-beam irradiation The external iliac nodes are the most laterally placed group, being situated at about 5-6 cm from the pelvic midline. The dose received at this distance from the intracavitary treatment will depend upon the system used, and for a Manchester System would be from 25%
Intracavitary treatment 357
to 30% of the point A dose. That order of dose is unlikely to be sufficient to cure lymph node disease except in some cases of microscopic disease. Therefore, additional external-beam irradiation is necessary to boost the dose to the external and common iliac nodes to a curative level. Whether external-beam irradiation will effectively control lymph node disease remains an important question. Rutledge and Fletcher reported that, in a lymphadenectomy series, a high percentage of disease control was obtained in external iliac and obturator nodes with doses ranging from 50 to 60 Gy in 5-7 weeks [98]. Others have reported that external-beam irradiation before surgery can effectively reduce the expected number of involved nodes. Lagasse et a/., in a randomized study, found that patients receiving preoperative irradiation had nodal involvement in 12% of cases, compared with 22% who received surgery alone. The latter then had postoperative irradiation, producing a 5-year survival of 74%, compared with 88% in the former group [99]. The published reports of pelvic recurrence rates following external-beam irradiation for the different stages of disease have been less than the expected lymph node involvement rate, and that evidence is the substantial argument made to support external-beam irradiation. Also, in the case of a primary tumor exceeding 5 cm in size, a curative dose at the tumor periphery from intracavitary radiation alone is difficult to achieve without putting the central pelvic organs at high risk of radiation damage. As a consequence, shrinkage of the primary tumor by external-beam irradiation before intracavitary irradiation is indicated [47]. The claimed advantages and disadvantages from using remote afterloaded HDR intracavitary irradiation include: 1. no radiation exposure to staff during the setting up and delivery of treatment; 2. the ability easily to retain treatment applicators in position during the short treatment time; 3. delivering treatment in minutes as opposed to hours or days: this is to the patient's advantage as well as making it easier to deliver treatment in a consistent manner; 4. delivering treatment at dose rates comparable to external-beam irradiation so that the radiobiological effects are dependent on dose per fraction, number of fractions, and treatment volume; 5. no need for an indwelling bladder catheter; 6. minimal risk of ischemia of vaginal vault epithelium due to prolonged pressure from ovoids and packing; 7. a high throughput of patients who, in some situations, can be treated on a daycare basis; 8. it has been reported by some that they find no need to treat patients under general anesthesia.
The disadvantages include: 1. an increased number of treatment fractions and anesthetics; 2. an increase in theater work load; 3. the need to combine external-beam with intracavitary irradiation for early-stage disease. This argument is based on only 20% of stage Ib cases having lymph node involvement and external-beam irradiation being unnecessary in 80%. However, similar arguments can be claimed against lymphadenectomy. On balance, it has been the author's experience that the advantages outweigh the disadvantages.
25.7
INTRACAVITARY TREATMENT
The uterine cervix is ideally suited for intracavitary irradiation because: 1. the endocervical canal and vaginal vault form a suitable vehicle to carry radioactive sources; 2. the normal cervical tissues and vaginal vault epithelium are relatively radioresistant and tolerate high doses of irradiation; 3. the intensity of irradiation rapidly falls off with distance from the intracavitary sources. This restricts the amount of irradiation received by normal tissues beyond the cervix region. The majority of preloaded systems (classically, Stockholm, Paris, and Manchester) are adaptable for afterloading. In particular, the Manchester System has been developed to provide for afterloading with low, medium, and high strength sources. High strength sources involve several hundred giga-becquerels of activity in total. The applicator system should be sufficiently rigid to prevent distortion when inserted. The range of source lengths and ovoid applicators should provide a number of predetermined treatment options. In terms of dose distribution, most afterloaded systems are designed to be compatible with conventional manual systems, although there may be differences between systems in the definition of prescriptive reference point(s) and, in turn, dose delivered. Related to this is the sensitivity of the position of reference points close to the sources and disposition of those sources in regions of HDR gradient [33-35]. However, the advent of independent control of each source position and dwell time provides for a considerable variation in dose distribution and specification that can be important when combining intracavitary and external-beam therapy, particularly when optimized treatment is planned [33]. Treatment optimization may be indicated on the basis of the extent of local disease or the proximity of normal tissues to one or more sources. Optimization can be achieved by the use of CT or MRI
358 High dose-rate brachytherapy for treating cervix cancer
scans in conjunction with a computer planning system [31,34,36]. For further discussion, the reader is referred to the physics section of this book.
25.8
SOURCE POSITION RELATIONSHIPS
When intracavitary source applicators are independently positioned, the geometrical relationship between them will depend on the size and shape of the tumor. This may result in a portion of the primary tumor being underdosed and normal tissue overdosed. The later effect may be further increased by the addition of external-beam irradiation. Alternatively, one may use a system in which the individual source carriers are fixed in a predetermined fashion to provide a standard-shaped isodose volume. If necessary, the shape of the isodose volume can be modified by altering the source dwell times and altering the relative position of the applicators to each other [33,40,41 ].
25.9
DOSE DISTRIBUTION
In order that the intracavitary dose distribution may match the classical Manchester Radium System, the size of the intrauterine tube and vaginal ovoids should be as per the Manchester System. This means that for HDR treatment the sources need to be small and of high specific activity, such as cobalt-60 or iridium-192. Our initial experience was gained from using cobalt-60 sources that were too large to pass through a tight radius of curvature and the ovoids were placed in line with the vaginal axis. The introduction of iridium-192 sources meant that their relatively small size permitted a relative tight radius of curvature of the carrier tubes and, for some systems, to have the vaginal ovoids at an angle to the vaginal axis. The bladder and rectal doses have been measured and calculated at an ovoid angle of 18° and 60° and calculated for 0° [37]. It was found that altering the ovoid angle from 60° to 18° resulted in a reduction of the dose rate at point A and no significant difference in the bladder or rectal dose. The calculated dose to the bladder with the vaginal ovoids at 0° was found to be reduced at the expense of an increased dose to the rectal wall. Resulting from this work, and because using ovoid sources in line with the vaginal axis did not produce any excess major rectal or bladder problems using cobalt-60, we elected to continue with a similar source configuration following the introduction of iridium192 sources. The size of ovoid and uterine tandems used has been shown for LDR systems to affect dose distribution, and large ovoids result in a lower bladder and rectal dose but higher point B dose. Also, longer uterine tandems produce a lower rectal dose and higher point B dose [86]. Increases in age and stage of disease have been
reported to result in less use of large ovoids and long tandems and consequent higher dose to the rectum and bladder and lower dose at point B. Starting with large ovoids may result in small ovoids becoming neccessary as treatment progresses. The regular use of small ovoids throughout all applications overcomes the majority of these changes in dose distribution that may occur. Geometric variation may occur between HDR treatments. This has been reported more commonly to occur when colpostats rather than tandems are used. Also, vaginal packing was shown to have a consistent effect on dose distribution [87]. In the case of a small vagina, Sherrah-Davies reported that LDR treatment could be effected using small ovoids mounted on straight tubes. This meant that the long axis of the ovoids was in line with the vagina and vaginal packing was easier. The geometry was less ideal, but in practice seemed likely to replace using half-ovoids [75]. Our experience since the 1960s is that small ovoids in line with the vaginal axis are acceptable for cases treated with HDR brachytherapy. An alternative system is one that uses a single-line source. The position of the source(s) within the intrauterine vaginal applicator is altered, either in a stepwise or cyclical manner, in order to provide a predetermined dose distribution. The advantage of a central vaginal applicator is that, in general, an applicator of larger diameter than an ovoid can be used. While this will provide for a lower mucosal dose and greater depth dose below the vaginal wall, it does not provide a higher relative dose to the parametrial tissues. An important advantage claimed for a single-line system is that patients can conveniently be treated as outpatients [63]. The use of external-beam irradiation with a single-line LDR or HDR source(s) has been reviewed and reported to achieve good local control for non-bulky tumors [39]. Other alternatives include a modified ring-shaped applicator [73,80,81]. Comparison of the ring applicator and Fletcher applicator has been published, with the caution that one cannot transfer Fletcher-recommended loadings to the ring applicator without radically altering the dose distribution pattern that would have been expected with Fletcher [81]. A similar argument will apply to high activity source loading. The dose distribution for those systems that have predetermined fixed geometrical configuration will be known. However, for systems in which individual sources are independently positioned, the dose distribution can only be calculated from suitable radiographs done manually or by computer. This may pose a particular problem for HDR treatment, particularly if the patient is anesthetized, because the time taken to carry out the dose calculations and modify the treatment time(s) may greatly exceed the treatment time itself. The ideal situation is one where the source configuration is fixed and the dose distribution known and supported by a dose atlas for reference purposes [33,85].
Special considerations 359
25.10
TARGET VOLUME
The regional extent of cervical cancer can be separated into two components: (1) the primary tumor, and (2) the pelvic lymph nodes. These are essentially two separate but interrelated clinical tumor volumes. The two clinical tumor volumes can be superimposed within a common planning target volume, one being large and the other relatively small. The different normal tissues within each target volume will, because of their different radiobiological factors and volume of tissue irradiated, require different dose fractionation restrictions to provide the best therapeutic ratio. To deliver a maximum tumor dose, consistent with a high chance of effective tumor control and minimum risk of late normal-tissue damage, will entail using a carefully designed combination of intracavitary and external-beam irradiation. In particular, a combination of the two modalities should optimize the dose distribution for treating either early or late disease. Determining the extent of the clinical tumor volume is at best a combination of objective and subjective assessment. An improved assessment can be achieved by using conventional lymphography, ultrasonography, CT and MRI scans [22-24,27-30,72,101]. The relative merits and limitations have been covered in various publications, but currently the most accurate single imaging process is MRI [30,31,100,101]. This has the additional advantage that it can also be used to monitor tumor shrinkage during treatment [30]. A detailed study of the value of individualizing the volume required to cover pelvic lymph nodes by using lymphography has reported that a change of 1.0 cm in width of the treatment fields can determine whether lymph nodes are optimally covered or not. A change of 1.0 cm in width of the treatment volume represented an average change in target volume of 340 cm3.. This could increase morbidity risks for those likely to benefit from smaller fields. Similarly, for those needing wider fields, there would be a greater chance of recurrent nodal disease [29]. CT and sagittal MRI also provide the means of geographically identifying the position of the pelvic lymph node groups in order to determine the outer limits of the planning target volume. This is not only important laterally, but also posteriorly at the S2-S3 interspace and anteriorly at the anterior edge of the pubic symphysis [101]. These levels have been identified as a potential problem in four-field treatment systems and both CT [22,57] and sagittal MRI [31,101] offer a more accurate method of planning.
25.11
DOSE-VOLUME SPECIFICATION
The planning target volume to be treated to a specified dose should contain the clinical tumor volume, although
not necessarily having a similar shape. For intracavitary therapy it is usual to prescribe treatment at either a dose point (point A) or isodose (reference isodose surface), usually taken as 60 Gy for LDR or its considered radiobiological equivalent for HDR. A volumetric description, in three-dimensional terms, of the reference isodose surface should approximate to the planned treatment volume and for fixed source systems will remain constant. This is discussed in Chapter 6 and elsewhere [40]. In a reported study for LDR, the 60 Gy reference isodose volume for a constant point A dose prescription increased with ovoid size, but showed no consistent variation with length of uterine tube. Applicator geometry affected the isodose volume, which could also change during treatment. The 60 Gy isodose volume was considered potentially useful for dosimetric comparison, but had a limited role in predicting clinical response [41]. The majority of HDR systems use small ovoids and this has been our treatment policy since the 1960s. Also, when a fixed source configuration and loading pattern are used, the isodose reference volume, when combined with external-beam irradiation, will provide a consistent and reproducible system for carrying out dosimetric comparison.
25.12
SPECIAL CONSIDERATIONS
For LDR systems it is possible, using radiographic images, to calculate the dose at a prescriptive point before treatment is completed and, if indicated, to modify the treatment time to deliver the prescribed dose. As previously discussed, this is not practical when high activity source treatments are used that last only a few minutes. As a result, unless a fixed source configuration is used, treatment is effectively being prescribed in terms of time and quantity (i.e., milligram-hours). Also, when using orthogonal or stereoshift radiographs to calculate the dose to the Manchester point A, the calculated dose will be dependent upon the way in which point A is defined. If the usual current method of relating point A to the flange of the uterine tube is used, as opposed to the vault surface of the vaginal ovoid, a considerable difference in dose rate at point A may occur [33]. Any change of distance between the ovoids and flange from one treatment to another will have a greater effect on the point A dose when referenced from the flange. The difference between the two alternative dose calculation systems, for a patient prescribed to receive a dose of 8.5 Gy once weekly for 4 weeks, is illustrated in Figure 25.1. The prescribed treatment was based on a previous standard calibration. It can be seen that the calculated dose varied greatly from the prescribed dose, irrespective of the system used, being more consistent when the ovoid surface was used. Such problems will, of
360 High dose-rate brachytherapy for treating cervix cancer
Figure 25.1 Values of point A dose using two alternative definitions for five separate insertions in a single patient.
course, not apply to a fixed source configuration, which was a major reason for us introducing a rigid source carrier system (see Figure 7.3, p. 109). The method of specifying absorbed dose to organs at risk is important when comparing different methods of management. The biological effects are known to depend upon total dose, dose rate, and dose per fraction, which will change with distance from the sources. Attempts should, therefore, be made to determine the dose per fraction when reporting the maximum dose received by a particular tissue. The volume of tissue receiving the maximum dose is also important, and planning systems in future may need to provide routine dose-volume histograms (see Chapter 5) [42,79].
25.13
TREATMENT PLANNING
25*13*1 Combined intracavitary and external-beam irradiation For the reasons already outlined, HDR intracavitary irradiation is usually combined with external-beam irradiation. For the treatment of early disease, the emphasis is usually on intracavitary irradiation, with external-beam irradiation being used to boost the dose to the pelvic lymph nodes to what is considered to be a maximum tolerance level. For advanced and bulky stage Ib disease, external-beam irradiation is usually intended to shrink the primary tumor to a size that can adequately be encompassed with a boost volume of intracavitary irradiation. This particularly applies to late stage lib and Illb cancers of the cervix. Also, such cases will have a 30-50% risk of lymph node involvement, being largely dependent upon external-beam irradiation to deliver a radical dose. To encompass the pelvic lymph nodes adequately, the width of the pelvic template used is typically 15 cm wide. This will mean that, for the average patient, the planned target volume will extend 1-2 cm lateral to the pelvic
brim. The height of the template will depend upon the disease extent and was historically based on anatomical levels, but a typical height is 14 cm. This may be extended to 18-20 cm when the lower vagina and lower para-aortic nodes are included. Normally, the lower edge of the planned volume should be just above the lower border of the obturator foramen and the upper edge at the lower border of the fourth lumbar vertebra. Posteriorly, the treatment volume should be 3 cm anterior to the posterior point of the hollow of the sacrum and, anteriorly, 2 cm anterior to the surface of the lumbar vertebrae for early disease, but may require modifying in advanced disease. However, as previously discussed and especially in advanced disease, patients may be inaccurately planned unless aided by CT or MRI [22,29,100,101]. Despite the complexity of some of the treatment systems used, irradiation of the pelvis may be adequately achieved in most patients using two parallel opposed fields, provided the energy used is 4-10 MV. In situations in which pelvic separation exceeds 20 cm, it is often preferable to treat the pelvis with a four-field 'brick' technique. If the latter is used, consideration should be given to CT or, preferably, sagittal MRI planning to reduce the risk of treating with inadequate anterior and posterior margins [31,57,101].
25*13*2 Homogeneous external-beam irradiation with high dose-rate intracavitary boost irradiation Although this technique may be used for treating all stages of disease, the major use is for treating bulky stage Ib and stage lib and Illb cases. External-beam irradiation is delivered to the pelvis without central shielding, to deliver a homogeneous dose distribution. The reported dose fractionation schedules vary, but in general the schedule is from 45 to 50.4 Gy in 20 to 28 fractions, respectively, treatment being given daily for 5 days each week. Intracavitary therapy follows within 1-2 weeks from completion of intracavitary irradiation. A dose of 7.5 Gy to the Manchester point A is delivered on two occasions, one week apart; alternatively, a dose of 7.0 Gy on three occasions is given. (The reader is referred to section 25.13.11, 'Optimal dose regime').
25.13*3 Intracavitary high dose-rate irradiation with shielded external-beam boost irradiation When the emphasis of treatment is on intracavitary irradiation with external-beam irradiation to boost the dose to the pelvic lymph node areas, some form of shielding to reduce the dose to normal tissues and organs receiving intracavitary irradiation is necessary. This treatment schedule is usually restricted to patients with early dis-
Treatment planning 361
ease, and this has been our practice since 1974. However, our best results for treating advanced disease were obtained using this treatment method from 1967 to 1974 [9,10]. The most critical part of the combined treatment volume will occur in the region where the steep dose gradient from the intracavitary irradiation combines with the external-beam irradiation. This occurs particularly in the region of points such as Manchester point A and especially if a centrally positioned lead block is used for screening, as opposed to a wedge that has been designed partially to compensate for the fall-off in dose from the intracavitary sources. One advantage of an HDR intracavitary system is that the dose per fraction is similar to that for external-beam irradiation. This does not resolve the radiobiological problems of determining the effective dose for the two treatment modalities, but does restrict the radiobiological problem to dose per fraction and total dose. (The reader is referred to the radiobiology section of this book for further reading.) The various dose combinations of external-beam and intracavitary irradiation for treating cervical cancer have been largely developed as a result of clinical experience gained from using external-beam irradiation with radium intracavitary systems. In the latter situation, it has generally been accepted that the intracavitary dose at a specific point, e.g., point A, can be added to the dose of external-beam irradiation at the same point. This is irrespective of the fact that adding doses for different doserate and dose per fraction schedules may result in different radiobiological effects for the same total dose. Considerable care should therefore be given to applying this approach when comparing HDR with LDR systems. The author has used a combined treatment regime in which the external-beam irradiation is given through parallel opposed fields with a centrally placed beamshaping wedge. The wedge was designed to compensate partly for the fall-off in dose from the intracavitary treatment in both lateral and longitudinal directions. Various modifications have been made since it was first introduced in 1971, but the basic principle has remained unchanged. The design was based on matching the average radiation distribution from the intracavitary sources, obtained from a computer program to produce isodose plots, carried out on several patients. One treatment aspect that has always been considered to be extremely important has been to restrict the overall volume of normal tissues included in the external-beam irradiation in order to restrict the risk of bowel injury. For the regime used, this has involved screening off the corners of the fields to provide irregular octagonal fields [9].
25*13*4
Wedge position
The position of the treatment wedge will normally be dependent upon the position of the intracavitary source
system. Its position is usually centered over the Manchester A-B line, 2.0 cm above the cervix marker or applicator flange, as determined by check X-rays. With most intracavitary carrier systems, the position of the sources and uterus can be brought to within 1.0 cm of the central sagittal plane of the pelvis. One exception to this is when the uterus is fixed to other structures, either due to tumor or some other pathology. For stage I and II cases, lateral displacement due to cancer is not usually a problem.
25*13.5
Treatment procedure
The procedure follows a relatively standard approach, with patients being set up in the lithotomy position on a mobile treatment table. Patients are usually anesthetized, although some described methods do not use anesthetics but sedation. When anesthetized, setting up is normally carried out on a treatment trolley in an operating theater and, following completion of the setting-up procedure, the patient can be moved on the trolley to the radiation treatment room. Where indicated, and depending upon the treatment protocol, check radiographs may be done for dose calculation purposes. The procedure for insertion of the afterloaded applicators will normally follow a similar pattern to that for LDR systems. There are, however, some essential differences in that treatment only lasts for a few minutes and an indwelling bladder catheter is not necessary; although some catheterize the bladder as part of the initial theater procedure, others consider a full bladder is preferable. Also, pyometra is not, generally, a contraindication to continue with HDR irradiation that only lasts from 5 to 20 min. Following treatment, antibiotics are given and, if indicated, a suitable intrauterine drain may be inserted. Clearly, care is necessary against perforating the uterus in the presence of pyometra, particularly as the uterine wall may be thin. By carefully taking this approach, the author has not experienced any radiation or medical problems. The use of a rectal retractor eliminates the need for vaginal packing and reduces the risk of disturbing the anatomical relationship of the uterus to other pelvic structures. Also, clamping volsella forceps to the cervix and attaching the forceps to the applicator system external to the introitus using a spring under slight tension allow the uterus to be positioned at a low level in the pelvis. This provides a means of similar positional placement on further intracavitary treatment occasions. The insertion of an inert metallic 'seed' in the anterior lip of the cervix will, on X-ray, identify the relative position of the cervix to the applicator flange. Depending on the treatment schedule, this can also be used to check any change in position of the uterus between receiving intracavitary treatment and external-beam irradiation. Difficulties may arise toward the completion of a
362 High dose-rate brachytherapy for treating cervix cancer
course of treatment due to tumor shrinkage and fibrosis, but this has not proven to be a major dosimetry problem. Also, the position of the applicators can, unless the uterus is fixed or tethered to the pelvic wall, be positioned in the midline of the pelvis. A common method of checking rectal dose is by using a suitable rectal dose measuring probe. This is not a practical proposition for HDR treatment unless equivalent low-activity monitor sources are used which allow manual loading of the treatment applicators. When measurements are made, the distance of the maximum dose from the anal verge should be recorded. This will provide a check for the position of treatment applicators within the pelvis and should be a similar location on each treatment occasion [3,9]. The clinical value of bladder dose measurement and calculation is contentious. A major reason for questioning the value of bladder dose measurement is discussed in the previous chapter, and the topic is discussed further below. For patients who receive external-beam irradiation in conjunction with a beam wedge or screening block, the position of the treatment wedge or screening block is normally centered over the site of the inert metallic cervical marker, overlying the position of the high-dose intracavitary treatment. Before the initial intracavitary treatment commences, check radiographs can be done, usually in the treatment room. Depending on the treatment schedule, these can be used for planning external-beam irradiation. For most applications, patients are lightly anesthetized during treatment delivery and a patient-monitoring system is essential. This will include cardiovascular, respiratory function, and closed circuit television monitoring. Also, a two-way speech system allows communication with the patient in order to provide reassurance as necessary. Once irradiation is completed, the treatment applicators are removed. The entire procedure normally takes less than 30 min.
25.13.6
Rectal dose
Rectal dose measurement carried out before treatment will normally follow one of the established methods as
used for LDR systems, whereas dose calculation before treatment is not practical [9]. In general, we have followed a relatively standard procedure for restricting the maximum measured rectal dose as for low-activity manually loaded systems. This has meant restricting the dose to no more than 60% of the prescribed point A dose for any individual treatment fraction. For those systems that have a fixed source configuration with standard source loading and the rectal tissues are separated from the ovoids by a rigid fixed retractor of known thickness, which forms part of the treatment applicator, the rectal dose can be predetermined (see Figure 7.3, p. 109). We use rectal retractors of 10 mm, 12.5 mm, and 15 mm thickness and the dose received by the rectal wall will vary from 50% to 65% of the prescribed point A dose (Table 25.1). As a result, it is not our routine practice to measure rectal dose. However, it needs to be stressed that for systems that do not ensure a predetermined known dose to the rectum, dose measurements are essential. Those dose measurements are made immediately before treatment using appropriate manually loaded low-activity sources. An alternative to dose measurement is to carry out dose calculations using orthogonal or stereoshift X-rays at one or more rectal points. However, as previously discussed, this is not a practical routine procedure for HDR intracavitary irradiation. Also, modification of the source loading may alter the dose distribution and rectal dose. This needs to be taken into account if the dose distribution is to not exceed normal-tissue tolerance [86]. Transverse CT images can be used to estimate a threedimensional dose distribution for critical structures and other points of interest [36]. Using this approach for LDR treatment, it was shown that the maximum rectal dose value obtained from a transverse CT slice near the top of the vagina was representative of the rectal dose burden and offered a means of correlating with rectal complications. Future developments are likely to involve applicator systems specifically designed to be CT and MRI compatible [34]. Commercial planning systems are available that provide for the calculation of pretreatment doses to point A, rectum, and bladder wall. It is possible that, in future, this sophisticated system will be practical for HDR intracavitary therapy.
Table 25.1 Using a Joslin-Flynn applicator and rectal retractor (see Figure 7.3) to limit the rectal dose. The dose 5 mm below the surface of the retractor is expressed as a percentage of the point A dose. Values are approximate and are averaged for ovoid positions 1 and 2
45 65
65 65
55 55
50 50
Treatment planning 363
25.13.7
Bladder dose
As for the rectum, it is not possible effectively to spare the bladder from receiving a significant dose of irradiation, but it should be minimized. The early afterloaded applicators did not routinely provide ovoid screens, but, during the early 1980s, screens were introduced for the LDR Selectron applicator. Using screens, the bladder dose was reduced by 15%, being similar to a distance increase of 5 mm between ovoid and bladder wall [45]. Using small ovoids in which the long axis is in line with the vaginal axis makes the use of ovoid screens impractical. The routine measurement of bladder dose is normally advocated and is included in the ICRU 38 recommendations, as discussed in Chapter 6 [40]. Crook et al. have reported finding that, using LDR brachytherapy, the dose-injury relationship is less reliable then for rectum because of the smaller number of complications, and the bladder reference point is less reliable. They also advised that pelvic floor laxity, due to age and multiparity, leads to considerable variation in the orientation of the bladder in the pelvis, with a consequent effect in the relationship of the maximum mucosal dose and the standard reference point [42]. In a comparison of calculating the bladder dose by balloon or chain for an HDR system, Kuipers and Visser found that the dose predicted with the aid of a chain was 11% higher than that determined by balloon. They considered alternative points to correct for that deficit [43]. Hunter et al. carried out CT scans on patients receiving LDR brachytherapy. They found that the ratio of the maximum bladder dose compared with the ICRU bladder reference point varied by a factor of 1.01 to 3.59. They advised that the inhomogeneous distribution of bladder dose depended upon the anatomy, disease extent, applicator design, and technical details of the insertion. The important factor was to identify the patient at risk of major injury [44] (see also Chapter 24). The procedure of bladder dose measurement is not made more difficult by the use of HDR techniques. However, because of the short treatment times for HDR, it is not practical to carry out routine CT scans in order to calculate bladder dose distribution at the time of treatment. Also, sectional CT images taken as part of a treatment planning procedure for external and combined brachytherapy, i.e., with or without treatment applicators in position, do offer some way forward for determining dose distribution to the bladder and other pelvic organs. A major problem is that of possible changes in bladder shape and position between fractional treatments, and between CT position and treatment position. 25.13.8 control
Factors affecting pelvic disease
Kapp and colleagues [65] found few studies that had analyzed prognostic factors in patients treated with HDR
regimes, whereas this had been done for LDR regimes. They confirmed that prognostic factors for 181 patients treated with HDR brachytherapy compared with a previously treated LDR brachytherapy group were similar. The significance of pretreatment hemoglobin level, of tumor size, pelvic and para-aortic lymph nodes as assessed by CT scan in contrast to clinical staging (FIGO) has also been demonstrated [26,70]. A relationship between outcome, tumor bulk, stage, lymph node status and overall treatment time in patients treated by a combination of external-beam with LDR or HDR intracavitary treatment has been reported [39,70,93]. For patients treated by a combination of external-beam irradiation and LDR brachytherapy, Perez et al. reported on the 10-year actuarial pelvic failure rate. It was 5% for stage Ib tumors less than 2.0 cm in size and 34% for unilateral or bilateral bulky tumors greater than 5 cm in size. Also, for a given stage and size of tumor, the pelvic failure rate decreased with increasing dose. For stage lib cancer, with total doses of 70 Gy to point A, the pelvic failure rate was about 50%, whereas for doses over 80 Gy it was 30% for bulky and 20% for non-bulky tumors [46]. In a retrospective analysis of 659 patients treated by external-beam irradiation (50 Gy) and HDR brachytherapy (20-34 Gy), Ito et al. reported no correlation between radiation dose and pelvic failure except in advanced stage III cases. In these cases, an HDR dose of 24 Gy or less to point A correlated with a higher pelvic failure and lower survival [47]. Computed tomography has been used to visualize the volume and shape of deep-seated cervix tumors as a predictive test of treatment outcome in a quantitative manner [70,71]. The initial volume, measured at the beginning of radiation therapy, was not a significant prognostic guide for local tumor control, but the volume measured immediately following completion of external irradiation was. Of patients whose second volume was less than 38 cm3, 90% were locally controlled 3 years later, whereas for those with volumes larger than 38 cm3, local control was 74%. Five-year actuarial survivals were 53% and 26%, respectively [71]. Using HDR intracavitary treatment, our experience of stage I cases is that for tumors less than 2 cm in size it is unusual to fail to control local disease. It has also been our experience that control of the primary tumor is related to its size. Table 25.2 gives the pelvic disease control rates for various LDR and HDR regimes. There are no significant differences to note, and this also applied to the comparative and prospective studies carried out. Our experience, from using two different dose fractionation regimes in which the biologically equivalent dose (BED) values, both acute (GylO) and late (Gy3), were similar and overall treatment time was 21 or 29 days, was that for stage I cases pelvic disease control was achieved in 96% and 81% of cases, respectively. Others
Table 25.2 Published reports of 5-year survival and pelvic disease control rates for carcinoma of the cervix treated by combined external-beam and HDR or LDR intracavitary irradiation
Joslin(1972) Khoury (1991) 1 5 year results Kuipers(1984) Vahrson(1988) Comparative study Cikaric(1988) Comparative study Khoury (1991) (Literature review mean %) Chen (1991) Arai (1992) Hammer (1993) Comparative study Teshima(1993) Randomized study Selke(1993) Sarkaria(1994) Comparative study Patel(1994) Randomized study MacLeod (1997) el-Baradie(1997) Prospective study Tan (1997) Kapp(1997) Wang (1997) Takeshi (1998) Kagei (1998) Single-line source Petereit(1999) Prospective study *3-year study.
[9] [10] [59] [60] [61] [10] [64] [69] [88] [20] [68]
*[15] [16] [62] *[17] [39] [65] [67] [66] [63]
*[18]
HDR HDR HDR LDR HDR LDR HDR
94 93 80 71 74 86 (855 cases)
HDR HDR HDR LDR HDR LDR HDR HDR LDR HDR LDR HDR HDR MDR HDR HDR HDR HDR HDR
85 88
HDR LDR
86 82
79 85 85 93 72
78 73 78 62 68 79 93 79 -
— — — 90 — — —
— — — 85 — — —
— – – 70 – –
—
—
–
—
—
92 —
60 —
53
— — 66
— —
– –
62 54 71 76 53 54 70 66 (1830 cases) 70 77 67 63 53 73 78 65 66 77 stages I to III 66 stages I to III 64 62 72 42 All stages combined
37 34 43 62 24 37 43 46 (1841 cases) 49 53 52 24 41 43 43 53 47 66 45
61 57 59 77
31 46 41 51 50
88
65 58
33 58
85 91
50 43 29
—
83 78 77 overall 80 overall 75.8 overall 79.7 overall — 67 overall 74 overall 69
88
40
—
45
– 94
87 87
72 71 64
80 78
44 75
Normal-tissue effects 365
have reported that prolongation of treatment time is associated with decreased local disease control and survival in stage lib and Illb patients treated with HDR and external-beam irradiation. This constituted the most significant prognostic factor, staging being second. However, there was no correlation between late complications and prolongation of treatment time [48].
25.13.9 Pelvic failure and distant metastasis Our first radical dose fractionation regime failed to control pelvic disease in 26% of stage II cases, of whom 18% developed metastasis, as did 22% of those with a clear pelvis. For stage III cases, pelvic recurrence occurred in 44% of cases, of whom 27% developed metastasis and 25% of those with a clear pelvis. From this it would appear that failure to control pelvic disease is not necessarily associated with an increased risk of metastatic disease. Toita et al. found that distant metastases-free rate was strongly correlated with nodal status as a suitable independent predictor [70].
25.13.10
Survival rates
The 5-year survival rates of patients who received HDR intracavitary brachytherapy combined with externalbeam irradiation have been variously reported, but all reports indicate similar results [11-13,17,18,88-90] (Table 25.2). Overall, the HDR results are similar, stage for stage, for the different dose fractionation regimes used. Table 25.2 also includes the results for a few LDR series, some of which were historical controls and others either concurrent or controlled studies. Information on the long-term actuarial survival is important and, for our series of stage I-II and III cases treated from 1967 to 1974, at 5, 10, and 15 years it was 63.4%, 58.8%, and 57.9% respectively. When separated into stage I, II, and III, the 5-year survivals were 94.5%, 62.6%, and 37.3%, and at 15 years were 92%, 53%, and 33%, respectively. Clearly, survival to 5 years is an indication of a high probability of cure. A review by Khoury et al. of several thousand cases reported 5-year survival rates of 86%, 66%, and 46% for stages I, II, and III, respectively [10].
25.13.11
Optimal dose regime
An American survey of radiotherapy regimes for cervical cancer reported a wide variation in doses used. Of 315 responders reporting on 4892 cases, 24% used HDR brachytherapy. The median external-beam doses were 48 Gy and 50 Gy and the median intracavitary dose 6 Gy on five occasions. This compared with LDR median external-beam doses of 45 Gy and 50 Gy; the LDR doses
were 42 Gy and 45 Gy for early and advanced cancer, respectively. These dose regimes represent what might be described as an 'optimal' reference regime on which to base further dose fractionation studies [84]. Petereit and Pearcey reviewed the literature in order to determine if there was an optimal fractionation schedule. They used the linear quadratic equation to determine the biologically effective tumor dose for an alpha/beta ratio of 10, at point A. The median externalbeam fractionation schedule was 40 Gy in 20 fractions and the HDR schedule was 28 Gy in four fractions. For stages Ib, lib, and Illb, the BED values at point A were 96, 96, and 100 Gy, respectively. They advised that the optimal fractionation regime was still unknown and currently can only be based on the recommendations of single institutions with significant experience [55].
25.14
NORMAL-TISSUE EFFECTS
The price of failure to completely control pelvic disease is usually fatal. As a result, the major treatment emphasis has been on cure. With a shift of public concern to morbidity issues and the technical advancements that have taken place, including attention to quality control and assurance, the price of cure is also receiving greater attention. The addition of external-beam to intracavitary irradiation has increased the incidence of small bowel injury and this has become a major risk for normal-tissue injury. The commonly recognized morbidities are well documented in the literature. They include parametrial fibrosis, which may be associated with obstruction to a lower ureter, extrinsic rectal fibrosis causing rectal stenosis, vaginal stenosis, or occlusion. A localized high dose may cause rectovaginal and vesicovaginal fistulae. Occasionally, a stricture may occur in the sigmoid colon, particularly in the terminal ilium, and, less commonly, perforation or fistula. The dose delivered from intracavitary irradiation to the above-mentioned structures will depend, for a particular system, on the disposition of the sources and their geometrical relationship to the structures in question. The only situation in which control of dose can be accurately achieved and maintained is for rectal tissues. Because of the difficulty of dose calculation to critical tissues, the present practice of dose checking by measurement for those tissues that are accessible has much to commend it. However, where both calculation and measurement can be done, they should complement each other. The last two decades have seen increased attention being given to these problems [10,11,14,26,42,46,53,54, 58,67,76,77,80,88,89,97]. Unfortunately, one of the major issues of recording and reporting morbidity has been the lack of an internationally accepted and applied
366 High dose-rate brachytherapy for treating cervix cancer
system of morbidity grading. This has made it extremely difficult to compare data from different centers. A review by Sismondi et al. of 96 articles reported no classification in 59, and the remainder used different criteria [49]. A French-Italian glossary for reporting complications, suitable for a computerized format, was published in 1993. This basically followed a standard approach of mild (Gl), moderate (G2), severe (G3), fatal (G4) complications, extended to include, for each grade, a series of subdivisions based on signs and symptoms [50,51]. The system was independently assessed in 1996 and reported as useful [52]. Reference is made to this glossary because of its specific application to gynecological cancer. 25*14.1
Early tissue damage
Early reactions commonly affect patients receiving a radical course of pelvic irradiation. In particular, diarrhea affects up to 50% of patients receiving external-beam irradiation in addition to intracavitary irradiation. In some cases, this may be associated with small bowel colic. The onset often occurs during the second week of treatment and generally settles within 1-2 weeks following completion of treatment. Conservative treatment usually produces an easing of symptoms. Our policy has been to rest the older patient if symptoms do not settle within 1 or 2 days. Patients with severe proctitis or abdominal colic should be rested from treatment until symptoms settle. If the symptoms return on resuming radiotherapy, the management policy should be reviewed, including possible surgery. Urinary symptoms are, in our experience, less common and when they occur may be due to infection and accordingly investigated and treated. Early reactions were fully recorded in our first series of 144 cases treated by HDR and external-beam irradiation from 1967 and published in 1972 [9]. Acute effects following HDR intracavitary therapy are not generally dealt with in detail in the literature and it is not possible to provide any meaningful comparison with LDR regimes. However, it has been suggested that the criteria for producing an equivalent HDR to LDR dose protocol should be based on matching early rather than late effects [95]. 25.14.2
Delayed tissue damage
Delayed normal-tissue damage may involve any tissue or organ within the irradiated volume, but is usually restricted to tissues within the planned treatment volume. The major tissues and organs at risk include rectum, sigmoid colon, small bowel, bladder, and vagina. Fractionation of the HDR brachytherapy is essential and at least single weekly fractions are normally given. Twice-weekly brachytherapy with a consequent reduction in dose per fraction has been reported to reduce the
risk of normal-tissue injury [12,74,90,94]. Twice-daily HDR fractions have been reported to show similar complications to LDR when made biologically equivalent, the major advantage being a shortened hospital stay [91]. Others have reported on the giving of concomitant HDR with daily external-beam irradiation on three occasions each week for 3-4 weeks. On the concomitant treatment days, a stepped wedge was used in conjunction with external-beam irradiation. The complication rates were reported to compare favorably with LDR experience [92]. Again, the importance of a time gap to allow for sublethal damage tissue recovery between fractions should be stressed. Patient follow-up for at least 5 years is important, during which time 80% of patients suffering symptoms and signs of moderate/severe morbidity will have come to light (Table 25.3). Also, patients may die without presenting with late tissue injuries and it is important to provide data on an actuarial basis [9,51].
25.14.3
Rectum and sigmoid colon
Although the reporting of rectal complications has become more prevalent in recent publications, many of those reports relate complications to total dose rather than to the dose fractionation schedule received by the rectal tissues. This approach would appear to be based on external-beam studies of tolerance for a fraction size of 1.8 Gy and 2.0 Gy [76]. Calculating and reporting the BED, using the ICRU rectal reference point, for combined external-beam and HDR intracavitary treatment, may overcome some of the difficulties of correlation of dose fractionation to morbidity. However, the rectal dose as calculated at a defined point may not be the point of maximum dose as determined by CT [36]. Correlation between rectal complication and calculated rectal dose has been reported [53,54,90]. Using the ICRU criteria, Ogino et al. advised that the incidence of rectal complications was from 5% to 10% at time-dose fractionation (TDF) values from 104 to 124 and BED (Gy3) values from 119 to 146. Grade 4 complications were not observed with a TDF below 130 or BED below 147 [53]. Petereit and Pearcey reported no correlation between point A BED (Gy3) and risk of late rectal and bladder complications. They questioned the quality of the current HDR brachytherapy literature in providing details of dose fractionation schedules rather than the linear-quadratic (LQ) model [55]. Le Pechoux et al. reported that no factor was predictive of local control, but the dose of external-beam irradiation significantly influenced the risk of complication [56]. Some of the reported results of severe rectal morbidity, mostly graded as 2 or 3, are shown in Table 25.3. These results are for different HDR and LDR regimes, including comparative studies in the same centers. It can
Normal-tissue effects 367 Table 25.3 Complication rates reported by various centers using HDR and LDR brachytherapyfor cervix cancer. The grading criteria are not standardized to a particular system Reference
First author
Date
Treatment type Stages
[96]
Sato
(1984)
HDR
Small bowel
Rectum
Bladder
14.9%
9.2%
13.6%
7.5%
I-IV
LDR
[59] Kuipers Comparative study
(1984)
[61]
(1988)
Cikaric
HDR
7%
LDR
6%
LDR
II and III II and III
HDR
l-lll
HDR
3.5% 3.3%
G2 4%
G3 3%
N.R.
G3 6%
[80] Rotte Comparative study
(1989)
[10]
Khoury
(1991)
HDR
l-lll
[68]
Selke
(1993)
HDR
l-lll
[69]
Arai
(1992)
HDR
I-IV
HDR
l-lll l-lll
10% of patients 4% of patients
LDR
l-lll l-lll
6% G 3 andG 4 0.4% 20% G 3 andG 4 2.4%
LDR
[20] *Teshima Randomized study
(1993)
[16] Patel Randomized study
(1994)
[67]
Wang
(1997)
HDR
l-lll
[62]
MacLeod
LDR HDR
(1997)
HDR
l-lll
[39] Tan Single-line source
(1997)
LDR
l-lll
[65]
Kapp
(1997)
HDR
I-IV
[66]
Takeshi
(1998)
HDR
III
[63] Kagei (1998) Linear source arrangement
HDR
II and III
G3 4.0%
G3 0.7%
G2 6%
G3 2%
G3 1.4
1.1
G3 6.2
2.2
G2 3%
G, 0.5%
G2 4.0% G3 0%
7.6%G 3 andG 4 1.1%
4.1%
1.4%
G 3 andG 4 9%
G 3 andG 4 6%
5% overall late severe morbidity G2 4.3%
G3 1.4%
G2 2.7%
G3 0.6%
8%
2% severe 8.3% 1%
(major)
G3
+
2.6%
(major)
1%
2% G4
*3-year study. N.R. = not recorded; G = grade.
be seen that the results appear to favor HDR in some cases and LDR in others. Among the reasons that the results from the various centers may not be intercomparable are differences in dose specification and systems, of follow-up [ 11 ]. As so often occurs when reviewing various reports, without having the necessary information accurately to equate all the various parameters affecting outcome, no significant differences between the morbidity outcome of treatment for LDR and HDR systems are likely to be identified. Sigmoid colon narrowing has been reported as a complication following HDR and LDR treatments. In the case of HDR, it is usually in association with external-
beam irradiation. A long intrauterine tandem loaded to the tip may result in a high dose to the sigmoid colon [38,75]. Our experience does not support this view, except possibly when long tubes are used throughout treatment (Table 25.4). In the mixed situation, a long tube was used for the first two insertions and a mediumlength tube for the remaining three insertions. As a result, the author has made it a policy to restrict the use of long tubes to two insertions. Also, the tendency to push the flange of the central tube against the cervix and inadvertently move the uterus deep into the pelvis should, as previously discussed, be counterbalanced.
Table 25.4 Morbid effects using a long intrauterine tube when treating stage I and II cancers of the cervix, Leeds series 1974-1983
Long Medium Mixed
6 148 167
2(33.3) 10(6.7) 13(7.8)
9(6.1) 4(2.4)
7(4.7) 11(6.6)
368 High dose-rate brachytherapy for treating cervix cancer
25.14.4
Bladder
The reported symptoms include urinary frequency, dysuria, and hematuria. Late radiation changes may cause varying amounts of atrophy of the bladder mucosa, often associated with telangiectasia. Occasionally, a vesicovaginal fistula will occur. In a total of 187 stage Ib cases treated with HDR intracavitary irradiation, we saw one case of vesicovaginal fistula. Although most cases will present with bladder symptoms within 2 years, cases can first present at up to 10 years or later. The risk of hematuria recurring is relatively low and affects up to 5% of patients. Our experience has been that only rarely are patients so severely affected that urinary diversion is necessary. Similarly, a constricted bladder sufficiently severe to warrant an ileal loop bladder is unusual. Because of the problems of accurately determining bladder dose, comparison of the late effects for the different techniques used is difficult. Most reported series of complication rates only provide details of dose fractionation at a specific point, e.g., point A. The value of the ICRU bladder reference point being the point of maximum bladder dose has not been supported by Hunter et al. [44], whereas Pourquier et al. found that the maximum dose to the trigone showed a significant difference in patients with or without complications [77]. Specific details of the dose fractionation breakdown of the maximal dose (either measured or calculated) to the bladder are rarely reported. The majority of calculations relate to the trigone area that is fixed relative to the lower gynecological tract, whereas the maximum dose may be at some other point [43,44]. In general, the various studies that have reported bladder injuries show no significant difference between rates for LDR and HDR techniques either between or within institutions (see Table 25.3). Future developments based on dose area plots of the bladder mucosa or dose-volume plots may provide a better dose-injury risk correlation [42].
25.14.5
Small bowel
The overall reported rates of small bowel injury are variously reported as affecting 25% and more of patients. However, when restricted to moderate and severe morbidity, the rate falls to 5-15%. These rates apply whether LDR or HDR treatment in conjunction with external-beam irradiation is given. The risks of developing small bowel injury have been reported to be dependent on total dose, dose per fraction, dose rate, and the size of the planned treatment volume. A history of previous laparotomies and pelvic inflammatory disease increases the risk of small bowel injury [58], The volume effect on clinical tolerance has also been demonstrated. An analysis of radiation-related small bowel complications indicated that, if the volume of irradiated small bowel was increased by a factor of 2,
the total dose had to be reduced by 17% for the same complication risk [79]. The major contributing cause of small bowel injury is external-beam irradiation, although intracavitary irradiation will contribute from 20% to 30% of the point A dose at the level of the pelvic brim. This will mean that an HDR fraction of 6-7 Gy at point A will be approximately 2.0 Gy at the pelvic brim. This dose per fraction is of a similar order to that from external-beam irradiation. Thus, the terminal ilium may receive a similar dose per fraction from combined treatment. However, small bowel may lie within the pouch of Douglas or be stuck to the surface of the uterus. In these situations, the affected portion of bowel may receive a high dose with each fraction of HDR brachytherapy. Mobile small bowel may alter in position between fractional doses of HDR brachytherapy, and the risk of repetitive significant doses of irradiation to a particular segment of bowel will be reduced. In our experience, more than 60% of late large and small bowel complications occur within 2 years of treatment, with a median time of 14 months, being similar to other reports [10]. However, patients may initially present with small bowel morbidity at up to 10 years from treatment and, if previously discharged from follow-up at 5 years, these may not be recorded [10]. Various treatment techniques have been studied in order to minimize the volume of small bowel included in the external-beam volume. We have applied radiation shields to shape the anterior and posterior pelvic fields to the perimeter of the pelvic lymph node groups. This entails screening off all corners of the standard treatment field to produce an octagonal shape. Also, we normally treat patients in a prone position in an attempt to reduce the volume of irradiated small bowel. A prospective study of different treatment techniques has identified compression in the prone position combined with bladder distension as a single reproducible procedure. It was found that the irradiated volume of small bowel, especially following pelvic surgery, correlated with acute gastrointestinal effects. Late effects also correlated with prior pelvic surgery and volume of small bowel receiving more than 45 Gy [82]. Radiation enteritis is invariably progressive and further complications often occur following initial symptoms and signs. Prevention is virtually impossible for the range of dose fractionation schedules used for treating cervical cancer, but steps should be taken to minimize the risk [78]. The subject of radiation morbidity to the gastrointestinal tract following radiation has been extensively covered elsewhere [83].
25.15
CONCLUSION
An attempt has been made to discuss, in outline, some of the important issues that apply when treating cervix
References 369
cancer by HDR brachytherapy and external-beam irradiation. The various published reports of pelvic disease control rates, cure rates, and morbidity reveal a lack of standardization of several aspects of management, although these problems are being steadily overcome with improved imaging, treatment planning, accuracy of treatment delivery, and clinical management. However, in clinical practice, anatomical variation, size, and shape of tumor will mean that standard treatment will be inadequate for some patients. How to overcome some of the physical problems is compounded by the relationship between physical dose parameters and radiobiological effects [102]. Whether HDR or LDR is the better treatment remains an open issue and is likely to remain so for the foreseeable future.
REFERENCES
1. 0' Connell, D., Howard, N., Joslin, CAR et al. (1965) A new remotely controlled unit for the treatment of uterine carcinoma. Lancet, 2, 570-1. 2. 0' Connell, D., Howard, N., Joslin, CAR et al. (1967) The treatment of uterine carcinoma using the Cathetron, Part I. fir./ Radiol., 40,887. 3. Joslin, CAR, 0' Connell, D. and Howard, N. (1967) The treatment of uterine carcinoma using the Cathetron, Part III. Clinical consideration and preliminary results. Br.J. Radial.,40,895. 4. Joslin, CAR and Smith, C.W. (1970) Use of high activity cobalt-60 sources for intracavitary and surface mould therapy. Proc. R. Soc. Med., 63,1029. 5. Henschke, U.K. (1960) Afterloading applicator for radiation therapy of carcinoma of the uterus. Radiology, 74,834-9. 6. Henschke, U.K., Hilaris, B.S. and Mahan, G.D. (1964) Remote afterloading with intracavitary applicators. Radiology, 83,344. 7. Walstan, R. (1962) Remotely controlled afterloading radiotherapy apparatus. (A preliminary report.) Acta Radio!., Suppl. 236. 8. Joelsson, I. and Backstrom, A. (1970) Applicators for remote afterloading techniques for optimum pelvic dose distribution in carcinoma of the uterine cervix. Acta Radio!. Ther. Phys. Biol., 9,233. 9. Joslin, CAR, Smith, C.W. and Mallik, A. (1972) The treatment of cervix cancer using high activity Co60 sources. Br.J. Radio!., 45,257-70. 10. Khoury, G.G., Bulman, A.S. and Joslin, CAR (1991) Long term results of Cathetron high dose rate intracavitary radiotherapy in the treatment of carcinoma of the cervix. Br.J. Radial., 64,1036-43. 11. Fu, K. and Phillips, T.L (1990) High dose rate versus low dose rate intracavitary brachytherapy for carcinoma of the cervix. Int.J. Radial Oncol. Biol. Phys., 19,791-6.
12. Shigematsu, Y., Nishiyama, K., Masaki, N. et al. (1983) Treatment of carcinoma of the uterine cervix by remotely controlled afterloading intracavitary radiotherapy with high dose rate; a comparative study with low dose rate system. Int.j. Radial Oncol. Biol. Phys., 9,351-6. 13. Himmelmann, A., Holmberg, E., Oden, A. and Skogsberg, K. (1985) Intracavitary irradiation of carcinoma of cervix, stage IB and MA. A clinical comparison between a remote high dose rate afterloading system and a low dose rate manual system. Acta Radiol. Oncol., 24,139-44. 14. Taina, E. (1981) Complications following high and low dose-rate intracavitary radiotherapy for stage l-ll cervical carcinoma. A comparison of remotely afterloaded Co60 (Cathetron) and conventional radium therapy. Acta Obstel Gynecol. Scand., Suppl. No. 103, 39-49. 15. SarkariaJ.N., Petereit, D.G., Stitt,J.A. et al. (1994) A comparison of the efficacy and complication rate of low dose rate versus high dose rate brachytherapy in the treatment of uterine cervical carcinoma. Int.J. Radial Oncol. Biol. Phys., 30(1), 75-82. 16. Patel, RD., Sharma, S.C., Negi, P.S. et al. (1994) Low dose rate v high dose rate brachytherapy in the treatment of carcinoma of the uterine cervix; a clinical trial. Int. J. Radial Oncol. Biol. Phys., 23(2), 335-41. 17. el-Baradie, M., Inoue, T., Murayama, S. et al. (1997) HDR and MDR intracavitary treatment for carcinoma of the uterine cervix. A prospective randomised study. StrahlentherOnko 1., 173(3), 155-62. 18. Petereit, D.G., SarkariaJ.N., Potter, D.M.and SchinkJ.C. (1999) High dose rate versus low dose rate brachytherapy in the treatment of cervical cancer; analysis of tumour recurrence in the University of Wisconsin experience. Int. J. Radial Oncol. Biol. Phys., 5,1267-74. 19. Mayer, A., Nemeth, G., Nemesher, C. and Thalacker, U. (1993) Pre-operative low dose rate versus high dose rate brachytherapy in Stage IB cervix cancer. An efficacy study. StrahlentherOnkol., 169(12), 716-20. 20. Teshima,T., Inoue, T., Ikeds, H.et al. (1993) High dose rate and low dose rate intracavitary therapy for carcinoma of the uterine cervix. Final results of Osaka University Hospital. Cancer, 72(8), 2409-14. 21. Subak, LI., Hricak, H., Powell, C.B. et al. (1995) Cervical carcinoma: computed tomography and magnetic resonance imaging for pre operative staging. Obstel Gynecol., 86(1), 43-50. 22. Kim, S.H., Choi, B.I., Han, J.K. et al. (1993) Preoperative staging of uterine cervical carcinoma: comparison of C.T. and M.R.I. in 99 patients.J. Compul Assist. Tomogr., 17(4), 633-40. 23. Yang, W.T., Walkden, S.B., Ho, S. et al. (1996) Transrectal ultrasound in the evaluation of cervical carcinoma and comparison with spiral computed tomography and magnetic resonance imaging. Br.J. Radiol., 69(823), 610-16. 24. Ho, CM., Chien, T.Y., Jeng, C.M. et al. (1992) Staging of
370 High dose-rate brachytherapy for treating cervix cancer
25.
26.
27. 28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
cervical cancer: comparison between magnetic resonance imaging and computed tomography and pelvic examination under anaesthesia.7. Formos. Med. Assoc., 91 (10), 982-90. Van NagellJ.R. and Higgins, R.V. (1992) Clinical invasive carcinoma of cervix: clinical features, diagnosis, staging and pre-treatment evaluation. In Gynecological Oncology, Vol. 1,2ndedn, ed. M. Coppleson. Edinburgh, Churchill Livingstone, 663-71. Kapp, K.S., Stuecklschweiger, G.F., Kapp, D.S. et al. (1988) Prognostic factors in patients with carcinoma of the uterine cervix treated with external beam irradiation and lr-192 high dose rate brachytherapy. Int.J. Radial Oncol. Biol. Phys., 42(3), 531-40. Walsh, J.W. (1992) Computed tomography of gynecologic neoplasms. Radiol. Clin. North Am., 30(4), 817-30. Scheidler,J., Hricak, H., Yu, K.K.et al. (1977) Radiological evaluation of lymph node metastases in patients with cervical cancer by meta-analysis. JAMA, 278 (13), 1096-101. Thomas, R., Wilkins, M., King, D.M. et al. (1996) Carcinoma of the cervix individualising the external beam treatment volume with lymphangiography. Proceedings of the B.I.R. brachytherapy clinical dosimetry and integration of therapies of gynaecological cancer. Br.J. Radiol., 69, 578-80. Hricak, H. (1991) Cancer of the uterus: the value of MRI preand post irradiation. Int.J. Radial Oncol. Biol. Phys., 21,1089-94. Thomas, L, Chacon, B., Kind, M. et al. (1997) Magnetic resonance imaging in the treatment planning of radiation therapy for carcinoma of the cervix, treated with the four field pelvic technique. Int.J. Radial Oncol. Biol. Phys., 37(4), 827-32. Ogino, I., Kitamura, T., Okajima, H. and Matsubara, S. (1998) High dose rate brachytherapy in the management of cervical and vaginal intraepithelial neoplasia. Int. J. Radial Oncol. Biol. Phys., 40(4), 881-7. Joslin, CAF. and Flynn, A. (1984) Dose specification and associated problems. Proceedings of the 3rd International Selectron Users Meeting 1984, Innsbruck, Austria, 218-26. Weeks, K.J. and Montana, C.S. (1997) Three dimensional applicator system for carcinoma of the uterine cervix. Int.). Radial Oncol. Biol. Phys., 37(2), 455-63. Fowler, J.F. and Mould, R.F. (1991) HDRafterloading consequences of a tilt of the uterine tube. Activity. Gynaecological High Dose Rate Brachytherapy Supplement No. 2, ed. R.F. Mould. Leersum, The Netherlands, Nucletron, 19-24. van den Bergh, F., Meertens, H., Moonen, L. et al. (1998) The use of a transverse C T image for the estimation of the dose given to the rectum in intracavitary brachytherapy for carcinoma of the cervix. Radiother. Oncol., 47(1), 85-90. Flynn, A. (1985) Leeds experience using the Selectron 1981-84: Part II aspects of bladder dosimetry.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
Proceedings of the 3rd International Selectron Users Meeting, 1984 Innsbruck, ed. R.F. Mould. Leersum, The Netherlands, Nucletron, 43-7. Maruyama, Y., van Nagell, J.R., Wrede, D.E. et al. (1976) Approaches to optimization of dose in radiation therapy of cervix cancer. Radiology, 120,389-98. Tan, L.T., Jones, B., Gee, A. and Kingston, R.E. (1997) An audit of treating carcinoma of the uterine cervix using external beam radiotherapy and a single line source brachytherapy technique. Br.J. Radiol., 70,1259-69. ICRU Report 38 (1985) Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology. Bethesda, MD, Commission on Radiation Units and Measurements. Jacob, R., Bidmead, A.M. and Blake, P.R. (1999) 60 Gy isodose volumes in carcinoma of the cervix and their relation to the geometry of uterine tube and ovoids. Br. J. Radiol.,72(854), 168-72. Crook, J.M., Esche, B.A., Chaplain, G.et al. (1987) Dosevolume analysis and the prevention of radiation sequelae in cervical cancer. Radiother. Oncol., 8,321-32. Kuipers, T.J. and Visser, A.G. (1986) Technical aspects of bladder dosimetry in intracavitary irradiation of cervix cancer. Radiother. Oncol., 7,7-12. Hunter, R.D., Wong, F., Moore, C. et al. (1986) Bladder dose dosage in patients undergoing intracavitary therapy. Radiother. Oncol., 7,189-97. Meertens, H. and van der Laarse, R. (1985) Screen in ovoids of a Selectron cervix applicator. Radiother. Oncol., 3,69-80. Perez, C.A., Grigsby, P.W., Chao, K.S. et al. (1998) Tumour size, irradiation dose and long term outcome of carcinoma of uterine cervix. IntJ. Radial Oncol. Biol. P/jys.,41(2),307-17. Ito, R., Kutuki, S., Nishiguchi, I. et al. (1994) Radiotherapy for cervical cancer with high dose rate brachytherapy, correlation between tumour size, dose and failure. Radiother. Oncol., 31(3), 240-7. Chatani, M., Matayoshi, Y., Masaki, N. and Inoue, T. (1997) High-dose-rate intracavitary irradiation for carcinoma of the uterine cervix. The adverse effect of treatment prolongation. Strahlenther. Onkol., 173(7), 379-84. Sismondi, P., Sinistrero, G., Zola, P. et al. (1989) Complications of uterine cervix carcinoma treatments: the problem of uniform classification. Radiother. Oncol., 14,9-17. Chassagne, D., Sismondi, P., Horiot,J.C.etal.(1993) Glossary for reporting complications of treatment in gynaecological cancers. Radiother. Oncol., 26,195-202. Sinistrero, G., Sismondi, P., Rumore, A. and Zola, P. (1993) Analysis of complications of cervix carcinoma treated by radiotherapy using the Franco-Italian glossary. Radiother. Oncol., 26,203-11. Rodrigus, P., De Winter, K., Leers, W.H. and Kock, H.C.LV. (1996) Late radiotherapeutic morbidity in patients with carcinoma of the uterine cervix: the application of the French-Italian glossary. Radiother. Oncol., 40,153-7.
References 371 53. Ogino, I., Kitamura, T., Okamoto, N. et al. (1995) Late rectal complication following high dose rate intracavitary brachytherapy in cancer of the cervix. Int.J. Radial. Oncol. Biol. Phys., 31(4), 725-34. 54. Orton, C.G. and Wolf-Rosenblum, S. (1986) Dose dependence of complication rates in cervix cancer radiotherapy. Int.J. Radial Oncol. Biol. Phys., 12,37-44. 55. Petereit, D.G. and Pearcey, R. (1999) Literature analysis of high dose rate brachytherapy fractionation schedules in the treatment of cervical cancer: is there an optimal fractionation schedule? Int.J. Radial Oncol. Biol. Phys., 43(2), 359-66. 56. Le Pechoux, C., Akine, Y., Sumi, M. et al. (1995) High dose rate brachytherapy for carcinoma of the uterine cervix: comparison of two different fractionation regimes. Int.J. Radial Oncol. Biol. Phys., 31(4), 735-41. 57. Kim, R.Y., McGinnis, L.S., Spencer, S.A. et al. (1995) Conventional four-field pelvic radiotherapy technique with computerised treatment planning in cancer of the cervix: potential geographic miss and its impact on pelvic control. Int.J. Radial Oncol. Biol. Phys., 31(1), 109-12. 58. Lanciano, R.M., Martz, K., Montana, G.S. and Hanks, G.E. (1991) Influence of age, prior abdominal surgery, fraction size, and dose on complications after radiation therapy for squamous cell cancer of the uterine cervix. Cancer, 69(8), 2124-30. 59. Kuipers, T. (1984) High dose rate intracavitary irradiation: results of treatment. In Brachytherapy, ed. R.F. Mould. Leersum, The Netherlands, Nucletron Trad ingBV, 169-75. 60. Vahrson, H. and Romer, G. (1988) 5-year results with HDR afterloading in cervix cancer: dependence on fractionation and dose. In High Dose Rate Afterloading in the Treatment of Cancer of the Uterus, Breast and Rectum, ed. H. Vahrson and G. Rauthe. Munich, Baltimore, Urban & Schwarzenberg, 139-46. 61. Cikaric, S. (1988) Radiation therapy of cervical carcinoma using either HDR or LDR afterloading: comparison of 5-year results and complications, ed. H. Vahrson and G. Rauthe. Munich, Baltimore, Urban &Schwarzenberg, 119-22. 62. MacLeod, C., Fowler, A., Dalrimple, C.et al. (1997) High dose rate brachytherapy. I. The management of carcinoma of the cervix: an 8 year Australian experience. Aust. N.Z.J. Obstel Gynaecol., 37(3), 342-7. 63. Kagei, K., Shirato, H., Nishioka, T. et al. (1998) High-doserate intracavitary irradiation using linear source arrangement for stage II and III squamous cell carcinoma of the uterine cervix. Radiother. Oncol., 47(2), 207-13. 64. Chen, M.S., Lin, F.J., Hong, C.H. et al. (1991) High dose rate afterloading technique in the radiation treatment of cervical cancer: 399 cases and 9 years experience in Taiwan. Int.J. Radial Oncol. Biol. Phys., 20(5), 915-19. 65. Kapp, K.S., Stuecklschweiger, G.F., Kapp, D.S. et al. (1997) Carcinoma of the cervix: analysis of complications after
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
primary external beam radiation, lr-192 HDR brachytherapy. Radiother. Oncol., 42(2), 143-53. Takeshi, K., Katsuyuki, K., Yoshiaki,T.etal.(1998) Definitive radiotherapy combined with high dose rate brachytherapy for stage III carcinoma of the uterine cervix: retrospective analysis of prognostic factors concerning patient characteristics and treatment parameters. Int.J. Radial Oncol. Biol. Phys., 41(2), 319-27. Wang, C.J., Leung, S.W., Chen, H.C.,etal.(1997) Highdose-rate intracavitary brachytherapy (HDR-IC) in treatment of cervical carcinoma: 5 year results and implication of increased low-grade rectal complication on initiation of an HDR-IC fractionation scheme. Int. J. Radial Oncol. Biol. Phys., 38(2), 391-8. Selke, P., Roman, T.N., Souhami, L et al. (1993) Treatment results of high dose rate brachytherapy in patients with carcinoma of the cervix. Int. J. Radial Oncol. Biol. Phys., 27(4), 803-9. Arai, T., Nakano,T., Morita, S. et al. (1992) High-dose-rate remote afterloading intracavitary radiation therapy for cancer of the uterine cervix. A 20-year experience. Cancer, 69(1), 175-80. Toita, T., Nakano, M., Higashi, M. et al. (1995) Prognostic value of cervical size and pelvic lymph node status assessed by computed tomography for patients with uterine cervical cancer treated by radical radiation therapy. Int.J. Radial Oncol. Biol. Phys., 33(4), 843-9. Arimoto, T. (1993) Significance of computed tomography measured volume in the prognosis of cervical carcinoma. Cancer, 72(8), 2383-8. Mayr, N.A., Tali, E.T., Yuh, W.T. et al. (1993) Cervical cancer: application of MR imaging in radiation therapy. /torf/o/ogy,189(2),601-8. Bahena, J.H., Martinez, A., Yan, D. et al. (1998) Spatial reproducibility of the ring and tandem high dose rate cervix applicator. Int.J. Radial Oncol. Biol. Phys., 41(1), 13-19. Newman, H., James, K.W. and Smith, C.W. (1983) Treatment of cancer of the cervix with a high-dose-rate afterloading machine (the Cathetron). InlJ. Radial Oncol. .Biol. Phys., 9,931-7. Sherrah-Davies, E. (1985) Morbidity following low-dose rate Selectron therapy for cervical cancer. Clin. Radiol., 36,131-9. Potish, R.A. (1990) Factors predisposing to injury. In Radiation Enteritis, ed. R.B. Galland, and J. Spencer. London, Edward Arnold. Pourquier, H., Delard, R., Achille, Let al. (1987) A quantified approach to the analysis and prevention of urinary complications in radiotherapeutic treatment of cancer of the cervix. Int.J. Radial Oncol. Biol. Phys., 13, 1025-33. Galland, R.B. and Spencer, J. (1985) The natural history of clinically established radiation enteritis. Lancet, i, 1257-8. Letschert, J.G.J., Lebesque, J.V., de Boer, R.W. et al. (1990)
372 High dose-rate brachytherapy for treating cervix cancer
80.
81.
82.
83.
84.
85. 86.
87.
88.
89.
90.
Dose-volume correlation in radiation-related late smallbowel complications: a clinical study. Radiother. Oncol., 18,307-20. Rotte, K. (1989) Comparison of high dose rate afterloading and classical radium techniques for radiation therapy of cancer of the cervix uteri. In Brachytherapy HDR and LDR Proceedings, ed. A. A. Martinez, C.G. Orton and R.F. Mould. Red Brand Road, Columbia, Nucletron, 110-18. Wollin, M., Kagan, A.R., Olch, A. and Belloti, J. (1991) Comparison of the ring applicator and the Fletcher applicator for HDR gynaecological brachytherapy. Activity, Suppl. 2,25-7. Gallagher, M.J., Brereton, H.D., Rostock, R.A. et al. (1986) A prospective study of treatment techniques to minimise the volume of pelvic small bowel with reduction of acute late effects associated with pelvic irradiation. Int.}. Radial Oncol. Biol. Phys., 12,1565-73. Smalley, S.R. and Evans, R.G. (1991) Radiation morbidity to the gastrointestinal tract and liver. In Complications of Cancer Management, ed. P.M. Plowman, T.J. McElwain, and A.T. Meadows. Oxford, Butterworths and Heinemann, 272-308. Nag, S., Orton, C, Young, D. and Erickson, B. (1999) The American Brachytherapy Society survey of brachytherapy practice for carcinoma of the cervix in the United States. Gynecol. Oncol., 73(1), 111-18. Flynn, A. and Joslin, C. (1990) TheJoslin-Flynn cervix applicator system. 'Selectron' Brachyther. J., 4(4), 109-11. Senkus-Kouefka, E., Kobierska, A., Jassen, J. et al. (1997) Patient- related factors determining geometry of intracavitary applicators and pelvic dose distribution during cervical cancer brachytherapy. Stralenther. Oncol., 173(6), 323-9. Kim, R.Y., Meyer, J.T., Plott, W.E. et al. (1995) Major geometric variations between multiple high-dose-rate applications of brachytherapy in cancer of the cervix: frequency and type of variation. Radiology, 195(2), 419-22. Hammer, J., Zidl, J.P., Altendorfer, C. et al. (1993) Combined external and high dose rate intracavitary radiotherapy in the primary treatment of cancer of the uterine cervix. Radiother. Oncol., 27(1), 66-8. Caishu-Mo, WangXiang-E, Wang Qi (1989) High dose rate afterloading in the treatment of cervical cancer of the uterus. Int.J. Radial. Oncol. Biol. Phys., 16,335-8. Utley, J.F., van Essen, C.F., Horn, R.A. and Moeller, J.H. (1984) High dose rate afterloading brachytherapy in
91.
92.
93.
94.
95.
96.
97.
98.
99.
100. 101.
102.
carcinoma of the uterine cervix. Int. J. Radial Oncol. Biol. Phys., 10,225-6. Hsu, W.L, Wu, C.J., Jen, Y.M. et al. (1995) Twice per day fractionated high versus continuous low dose rate intracavitary therapy in the radical treatment of cervical cancer: a non-randomised comparison of treatment results. InlJ. Radial Oncol. Biol. Phys., 32(5), 1425-31. Han, I., Malviya, V., Chuba, P. et al. (1996) Multifractionated high-dose-rate brachytherapy with concomitant daily teletherapy for cervical cancer. Gynecol. Onto/., 63(1), 71-7. Garipagaoglu, M., Yalvac, S., Kose, M.F. et al. (1999) Treatment results and prognostic factors in inoperable carcinoma of the cervix treated with external plus high dose brachytherapy. Cancer Lett., 136(1), 17-26. Stitt, J.A. (1999) High dose rate brachytherapy in the treatment of cervical carcinoma. Hematol. Oncol. Clin. North Am., 13(3), 585-93, vii-viii. Brenner, D.J., Huang, Y., Hall, E.J. (1991) Fractionated high dose-rate versus low dose rate regimens for intracavitary brachytherapy of the cervix. Equivalent regimens for combined brachytherapy and external irradiation. InlJ. Radial Oncol. Biol. Phys., 21(6), 1415-23. Sato, S., Yajima, A. and Suzuki, M. (1984) Therapeutic results using high-dose-rate intracavitary irradiation in cases of cervical cancer. Gynecol. Oncol., 19,143-7. Akine, Y., Tokita, N., Ogino, T. et al. (1990) Dose equivalence for high-dose-rate to low-dose-rate intracavitary irradiation in the treatment of cancer of the uterine cervix. InlJ. Radial Oncol. Biol. Phys., 19, 1511-14. Rutledge, F.N. and Fletcher, G.H. (1958) Transperitoneal pelvic lymphadenectomy following supervoltage irradiation for squamous cell carcinoma of the cervix. Am.J. Obstel Gynec., 76,321. Lagasse, L.D., Smith, M.L, Moore,J.G.etal.(1974) The effect of radiation therapy on pelvic lymph node involvement in stage I carcinoma of the cervix. Am.J. Obstel Gynec., 119,328-34. Khoo, V.S. (2000) MRI-'Magic radiotherapy imaging'for treatment planning? Br.J. Radiol., 73,229-33. Zunino, S., Rosato, 0., Lucino, S. et al. (1999) Anatomic study of the pelvis in carcinoma of the uterine cervix as related to the box technique. InlJ. Radial Oncol. Biol. Phys., 44 (1)53-9. Joslin, C.A. (1996) Commentary: Brachytherapy-clinical dosimetry and the integration of therapies of gynaecological cancer. Br.J. Radiol., 69,689-92.
26 Brachytherapy for brain tumors MAARTENC.CM.HULSHOFANDJANJ. BATTERMANN
26.1
INTRODUCTION
Primary tumors of the central nervous system comprise 1.5% of all malignant disease. Astrocytomas represent about 40% of all brain tumors of which the majority (75%) are malignant (anaplastic astrocytoma and glioblastoma multiforme) [1]. For patients with a glioblastoma multiforme, which compromise 75% of the malignant astrocytomas, surgery alone results in a survival of approximately 4-6 months and, in combination with conventional external irradiation, survival is about 6-9 months [2]. Combinations with chemotherapy or radiosensitizers for patients with a glioblastoma multiforme did not result in a prolonged survival. For anaplastic astrocytomas, the median survival is somewhat better, but still less than 10% survive 5 years. There is a good rationale for increasing the radiation dose with brachytherapy in malignant gliomas: radiation therapy is so far the most effective treatment and a dose-effect relationship up to 60-70 Gy has been demonstrated [2,3]. However, more than 90% of malignant astrocytomas fail treatment within 2 cm from the original tumor site, and distant metastases are hardly ever seen [4]. Further control by local treatment failed when the external dose was raised to 75 Gy [5] and this increased dose exposes the patient to a significant risk of normal-brain necrosis [6]. The radiation dose in brachytherapy has a rapid dose fall-off, thereby sparing normal surrounding brain tissue. In addition, the use of external heavy particle therapy has failed to improve the treatment results, mainly because of the toxicity to normal brain tissue [7]. With the development of accurate stereotactic techniques, the
principles of interstitial radiotherapy became available for brain tumor patients. In Europe, a long tradition of brachytherapy exists for high-grade as well as low-grade gliomas [8]. In the last two decades, the use of brachytherapy for brain tumors increased all over the western world, and nowadays most publications and studies come from the USA and Canada.
26.2
TREATMENT TECHNIQUE
Both computerized tomography (CT) and magnetic resonance imaging (MRI) scans are used to define the tumor localization, although it is generally accepted that MRI better outlines the tumor contour. Under local or general anesthesia, a stereotactic head frame is fixed at four points to the patient's skull and a contrastenhanced scan is obtained, with images every 3 or 5 mm (Figure 26.1). The target volume is outlined on each individual image and generally includes the contrastenhancing area with or without a margin of 5 mm. After digitizing the target geometry using three-dimensional software, the most suitable catheter trajectory and number and length of catheters are determined with a threedimensional preplanning. The preplanning will try to optimize the dose distribution, aiming at a dose rate of 40-100 cGy h-1 at the reference points, with as small a number of catheters as possible and without unacceptable overdosage in the tumor center. These calculations and preplanning procedure take about 1-1.5 h. For that reason, local rather than general anesthesia for fixation
374 Brachytherapy for brain tumors
Figure 26.1 CT image of cerebrum with Leksell base ring during the calculation of target points.
of the base ring is preferred, both by physicists and clinicians. In the operation room, catheters are introduced percutaneously and parallel to each other at the calculated positions using burr or twist drill holes (Figure 26.2). Catheters are directly fixed to the skin or via a template, which is sutured to the skin. Usually within 24 h after implantation, a contrast-enhanced scan is obtained for analysis of the catheter configuration (Figure 26.3). The accuracy of these stereotactic techniques is stated to be within 1-2 mm [9,10]. Actual length of sources is
determined for each catheter on this scan. After reconstruction of catheter position and determination of source position by X-ray film in two directions, a threedimensional dosimetric plan is made. Most institutes use temporary sources of high-activity iodine-125 seeds or iridium-192 wires. Final loading can be done manually or with a remote afterloading system. The advantage of remote afterloading is the better radiation protection for personnel, but the disadvantage is the greater immobility of the patient. Another treatment technique is the permanent implant of low-activity iodine-125 seeds. This can be done during open brain surgery, whereby source carriers are implanted in the margins of the tumor after debulking of the tumor, or stereotactically, whereby seeds are implanted after three-dimensional preplanning of seed positions. Prophylactic antibiotics and corticosteroids are mostly given as a routine measure. Removal of the catheters can be done without anesthetic and is never described as a clinical problem. There is good general agreement on the eligibility criteria for brachytherapy in the different reported series. Tumors should be unifocal, supratentorial, not larger than 5 cm (in largest diameter), and without ventricular, corpus callosum, or brainstem infiltration. Patients should have a good performance status.
263
263.1
RESULTS
Recurrent malignant gliomas
The first reported results of brachytherapy in malignant glioma concerned recurrences after surgery and external Figure 26.2 Leksell stereotactic system and intracerebral catheters during implantation.
Results 375
anaplastic glioma treated with brachytherapy had equal survival results to recurrence of a glioblastoma multiforme [15]. There is little known about the effect of brachytherapy on the quality of life of these patients. Leibel et al. showed that there was no significant reduction in deterioration during the 6 months after implantation [14]. However, symptomatic radiation necrosis needing re-operation occurred in 40-50% of cases [14,17]. These series used high dose-rate (HDR) brachytherapy with temporary implants. Studies involving resection of the recurrent tumor followed by permanent low dose-rate (LDR) implants suggest that this approach decreases the incidence of symptomatic necrosis [15,18,19].
263.2 Newly diagnosed malignant gliomas The longer than expected survival results for recurrent tumors have led to the use of brachytherapy as an adjuvant treatment in newly diagnosed gliomas. Initial results of brachytherapy as a boost after external irradiation for newly diagnosed malignant gliomas showed encouraging survival rates in non-randomized trials (Table 26.1) [19-23]. Loeffler et at showed a very promising result of 27 months' median survival after brachytherapy in 35 patients with a glioblastoma multiforme compared to 11 months for a matched control group [22]. The Northern California Oncology Group (NCOG) study 6G-82-2 demonstrated an improved median survival of 88 weeks for patients with a glioblastoma multiforme. Patients with non-glioblastoma multiforme anaplastic gliomas, however, had a median survival of 155 weeks, which was not different from other reported series without brachytherapy [21]. Selection criteria will have a significant influence on survival results and thus it is difficult to draw conclusions from non-randomized studies. Florell et al. studied the influence of selection by retrospectively identifying patients as either eligible or ineligible for brachytherapy. Eligible glioblastoma multiforme patients lived much longer than the ineligible patients (13.9 versus 5.8 months) and they concluded that the improved survival is at least partly the result of patient selection [24].
Figure 26.3 CT scan of cerebrum for analysis of catheter position. Note the catheter positions at the periphery and center of the tumor.
irradiation. It has been estimated that median survival after recurrence of a malignant glioma without further treatment is 3.2 months [11]. Re-operation can be effective in selected patients (performance status, tumor size, and age), but the beneficial effect in terms of survival is disappointing, with median survival after re-operation ranging between 19 and 36 weeks [12,13]. Results from series of patients treated for a recurrent malignant glioma with brachytherapy at median doses of 50-70 Gy range between 47 and 64 weeks [14-16], and they all claim improved survival compared to historical data. However, after 2 years almost all patients have died. Recurrence of
Table 26.1 Results of brachytherapy after external irradiation for newly diagnosed glioblastoma multiforme
Laperriere et al. [26]* wen et al. [33] Sneedefo/. [28] Voges et al. [20] Malkin [23] Kootetal. [30] * Randomized trial.
71 56 159 27 20 21
50 59 59 10-40 60 60
60 50 55 60 59 40
42 22 26 25 37 48
+ + -
31 64 51 0 43 33
13.8 18 18 15.6 22 17
376 Brachytherapy for brain tumors
In 1986 and 1987, two randomized studies were conducted in the USA and Canada. The first results of Trial 8701 of the Brain Tumor Cooperative Group («=272) have only been published in an abstract form so far [25]. Patients randomized to brachytherapy received a temporary implant delivering a total dose of 6000 cGy at the tumor periphery with a mean dose rate of 40 cGy h-1, after a mean external dose of 6020 cGy. Preliminary comparison has shown a significant survival advantage of 3.5 months for the implant group for patients with a glioblastoma multiforme [25]. The Canadian trial randomized 140 selected patients to external-beam therapy of 50 Gy only or external-beam plus a minimal interstitial boost of 60 Gy with a median dose rate of 70 cGy h-1. Inclusion criteria were as follows: supratentorial malignant astrocytoma, age 18-70 years, Karnofsky Performance Status (KPS) > 70, no involvement of corpus callosum, and maximum tumor diameter < 6 cm. Brachytherapy was given with one or more parallel catheters of temporary high-activity iodine-125 seeds. Median survival for the implant group was 13.8 months, compared to 13.2 for the no-implant group, which was not a significant improvement [26]. Re-operation was performed in 31% of the implant group and in 33% of the no-implant group. In both randomized studies, as in most retrospective series, patients who underwent a reoperation for a recurrence and/or necrosis after brachytherapy did significantly better than those without re-operation [15,21,22,25,26]. Other factors that improved survival were young age, high performance status, and smaller tumor size. Patients under the age of 30 particularly had an improved survival [22,27]. The pattern of recurrence and toxicity after interstitial boost treatment is an interesting secondary endpoint. Even after a combined interstitial and external dose of 100 Gy or more, recurrences after brachytherapy occurred in about 90% at the primary site [26,28-30]. A few studies reported lower local recurrence rates, but still about 70% [31,32]. However, the pattern of recurrence has changed. Distant relapses within the central nervous system are described in 22-28% of interstitially treated patients with a glioblastoma multiforme [28,32,33] compared to about 10% after external irradiation only [4]. In the randomized study there was a higher incidence of multifocal recurrences in the implant arm [26]. Studies from the Boston group described a higher incidence of marginal recurrences (37%), with only 35% failing within the brachytherapy volume [33]. However, it remains difficult to distinguish radiologically between marginal and local treatment failures. Smaller tumors (< 25 cm3) and 'adequate' implants were shown to have a higher local control rate [28,31,32]. The higher rate of marginal and distant relapses that occur when local treatment is improved confirms the fact that glioblastomas have a diffuse and widespread growth pattern. A dose-response relationship was demonstrated by Sneed et al., but interstitial doses of more than 50 Gy were also related to increased risk of life-threat-
ening necrosis [34]. Sneed recommended a conformal interstitial dose of 45-50 Gy in conjunction with conventional external-beam radiotherapy. The acute toxicity of brachytherapy in brain tumors is low. Overall toxicity of stereotactic management in midline brain lesions was found to be 4.2% [35]. The risk of arterial bleeding caused by the implantation of catheters is 0-2% [20,21,26,30] and infections are also rare. Late toxicity, however, is a clinical problem. Necrosis leading to a re-operation is a major problem in all series. We think that after doses of more than 100 Gy, with much higher doses in the center of the implant volume, necrosis can be viewed as integral to this procedure and not as a complication. The described incidences of re-operations of 35-64% [26,28,33] can be considered as an underestimation of the risk for brain necrosis because some of the patients do not live long enough to develop necrosis and others are in too bad a condition for re-operation. Median time to re-operation was about 40-50 weeks [21,28]. Apart from necrosis, histology after re-operation also showed, in most cases, vital tumor cells [28,31], confirming the fact that glioblastoma cells are highly radioresistant. Bernstein et al. reported severe vascular occlusion in 4% of their implant group [36]. In our series in the Academisch Medisch Centrum, Amsterdam (AMC), three out of 21 patients developed a sudden palsy 6-12 months after brachytherapy, mimicking a cerebrovascular accident within the implant volume [30]. The overall complication rate, excluding re-operation for necrosis, was about 25% in the randomized series, which is in the range of other series [10,24,30]. The effect of brachytherapy on quality of life should be an important endpoint. Gutin et al. described a continuing dependency on corticosteroids to combat the edema of focal radiation necrosis, which caused the reduction in mean KPS in their series [21]. In 13 patients from the same group who survived more than 3 years, mean Karnofsky had decreased from 95 at the time of brachytherapy to 75 at the time of last follow-up [28]. In the only published randomized trial, there was a significant increase in dexamethasone dosage for the implant group, but performance status did not differ between the implant and no-implant groups [26].
2633
Low-grade tumors
There are no randomized trials of brachytherapy for low-grade or benign brain tumors. The largest retrospective series comes from Freiburg in Germany, describing 455 patients with inoperable (often deepseated) low-grade tumors treated with temporary or permanent interstitial iodine-125 [37]. Selection criteria were signs of tumor progression, circumscribed tumors not exceeding 5 cm, KPS > 70, and no corpus callosum infiltration. Dose rates of preferably < 10 cGy h-1 were
References 377
given up to reference doses of 60-100 Gy to the outer rim of the tumor. The 5-years survival rates for pilocytic astrocytomas, grade II astrocytomas, oligoastrocytomas, oligodendrogliomas, and gemistocytic astrocytomas were, respectively, 85%, 61%, 49%, 50%, and 32%. These results are comparable with the survival data of the European Organization for Research and Treatment of Cancer (EORTC) study [38], taking into account the more favorable prognostic factors of the implanted group (smaller tumors, good performance). Radiogenic complications were observed in 39 out of 515 patients at the same institute. The most important risk factor was the volume of the 200 cGy isodose [39]. Good local responses were described with LDR implants in inoperable brainstem gliomas [40]. Brachytherapy can be considered as a treatment option for progressive, small, and inoperable low-grade gliomas.
Perez and L.W. Brady. Philadelphia, Lippincott-Raven, 777-828. 2. Walker, M.D., Alexander, E.Jr, Hunt, VJ.E.etal.(1978) Evaluation of BCNU and/or radiotherapy in the treatment of anaplastic gliomas.J Neurosurg., 49,333-43. 3. Walker, M.D., Strike, T.A. and Sheline, G.E. (1979) An analysis of dose-effect relationship in the radiotherapy of malignant gliomas. Int.J. Radiat. Oncol. Biol. Phys., 5, 1725-31. 4. Wallner, K.E., Galicich, J.H., Krol, G., Arbit, E. and Malkin, M.G. (1989) Patterns of failure following treatment for glioblastoma multiforme and anaplastic astrocytoma. InLJ. Radiat. Oncol. Biol. Phys., 16,1405-9. 5. Salazar, O.M., Rubin, P., Feldstein, M.L and Pizzutiello, R. (1979) High dose radiation treatment of malignant gliomas; final report. Int.J. Radiat. Oncol. Biol. Phys., 5, 1733^0. 6. Sheline, G.E., Wara, W.M. and Smith, V. (1980) Therapeutic irradiation and brain injury. Int.J. Radiat. Oncol. Biol.
26.4
FURTHER DEVELOPMENTS
To further improve local tumor control in glioblastoma multiforme, some centers started to combine interstitial radiation with interstitial hyperthermia, using the same catheters. A randomized trial comparing brachytherapy boost alone versus brachytherapy boost plus interstitial hyperthermia in patients with newly diagnosed glioblastoma multiforme demonstrated a small but significant improvement in median survival of 9 weeks for the hyperthermia group [41]. Further technical developments can improve the level, homogeneity, volume, and control of the temperature in the target volume and thus could improve the clinical results [42]. A phase I-II study is ongoing in the AMC, Amsterdam, with this new technique for newly diagnosed glioblastoma multiforme patients. Dose escalation by focused stereotactic external radiation (radiosurgery) has become available and the advantages in dose distribution with this technique can be compared with brachytherapy. The first clinical comparisons between radiosurgery and brachytherapy in newly diagnosed and recurrent brain tumors resulted in a similar survival for both treatment options [43,44]. Radiosurgery has the advantage of being an outpatient, non-invasive therapy, with the possibility of fractionation. A phase III trial is being conducted in Europe randomizing between conventional radiotherapy with or without a stereotactic external boost in newly diagnosed, selected glioblastoma multiforme patients. REFERENCES
Phys., 6,1215-28. 7. Battermann,J.J.(1980) Fast neutron therapy for advanced brain tumours. Int.J. Radiat. Oncol. Biol. Phys., 6(3), 333-5. 8. Mundiger, F. and Weigel, K. (1984) Long-term results of stereotactic interstitial curietherapy. Acta Neurochir., 33 (Suppl.), 367-71. 9. Willis, B.K., Heilbrun, M.P., Sapozink, M.D. et al. (1988) Stereotactic interstitial brachytherapy of malignant astrocytomas with remarks on postimplantation computed tomographic appearance. Neurosurg., 23, 348-54. 10. Salcman, M., Sewchand, W., Amin, P.P. and Bellis, E.H. (1986) Technique and preliminary results of interstitial irradiation for primary brain tumours.J Neurooncol., 4, 141-9. 11. Rostomily, R.C., Halligan, J.B., Keles, G.E., Spence, A.M. and Berger, M.S. (1993) Management of adult recurrent supratentorial gliomas. Neurosurgery, 3,219-52. 12. Harsh, G.R., Levin, V.A., Gutin, P.H., Seager, M., Silver, P. and Wilson, C.B. (1987) Reoperation for recurrent glioblastoma and anaplastic astrocytoma. Neurosurgery, 21,615-21. 13. Dirks, P., Bernstein, M., Muller, P.J. and Tucker, W.S. (1993) The value of reoperation for glioblastoma. Can.J. Surg., 36,271-5. 14. Leibel, S.A., Gutin, P.H., Wara, W.M. et al. (1989) Survival and quality of life after interstitial implantation of removable high-activity iodine-125 sources for the treatment of patients with recurrent malignant gliomas. Int.]. Radiat. Oncol. Biol. Phys., 17,1129-39. 15. Halligan, J.B., Stelzer, K.J., Rostomily, R.C., Spence, A.M., Griffin, T.W. and Berger, M.S. (1996) Operation and permanent low activity 125I brachytherapy for recurrent high-grade astrocytomas. Int.J. Radiat. Oncol. Biol. Phys.,
1. Wara, W.M., Bauman, G.S., Sneed, P.K., Larson, DA and Karlsson, U.L (1997) Brain, brain stem and cerebellum. In Principles and Practise of Radiation Oncology, ed. C.A.
35, 541-7. 16. Bernstein, M., Laperriere, N., Glen, J., Leung, P., Thomason, C.and Landon,A.E. (1994) Brachytherapy for
378 Brachytherapy for brain tumors
17.
18.
19.
20.
21.
22.
23. 24.
25.
26.
27.
28.
29.
30.
recurrent malignant astrocytoma. Int.J. Radiat. Oncol. Biol. Phys., 30,1213-17. Sneed, P.K., Gutin, P.M., Stauffer, P.R. et al. (1992) Thermoradiotherapy of recurrent malignant brain tumours. Int.J. Radiat. Oncol. Biol. Phys., 24, 583-91. Larson, G.L,Wilbanks,J.H., Dennis, W.S., Permenter,W.D. and Easley, J.D. (1990) Interstitial radiogold for treatment of recurrent high grade gliomas. Cancer, 66,27-9. Zamorano, L, Yakar, D., Dujovny, M. et al. (1992) Permanent iodine-125 implant and external beam radiation therapy for the treatment of malignant brain tumours. Stereotact. Funct. Neurosurg., 59,183-92. Voges, J., Treuer, H., Schlegel, W., Pastyr, 0. and Sturm, V. (1993) Interstitial irradiation of cerebral gliomas with stereotactically implanted iodine-125 seeds. Acta Neurochir., 58 (Suppl.), 108-11. Gutin, P.H, Prados, M.D., Phillips, T.L et al. (1991) External irradiation followed by an interstitial high activity iodine125 implant 'boost' in the initial treatment of malignant gliomas: NCOGstudy6G-82-2. Int.J. Radiat. Oncol. Biol. Phys., 21,601-6. Loeffler, J.S., Alexander, E., Wen, P.Y. et al. (1990) Results of stereotactic brachytherapy used in the initial management of patients with glioblastoma.y. Natl. Cancer,inst.,82,1918-21. Malkin, M.G. (1992) Interstitial irradiation of malignant gliomas. Rev. Neurol., 148,448-53. Florell, R.C., Macdonald, D.R., Irish, W.D. et al. (1992) Selection bias, survival, and brachytherapy for glioma. J. Neurosurg., 76,179-83. Selker, R.G., Shapiro, W.R., Green, S.et al. (1995) A randomized trial of interstitial radiotherapy (IRT) boost for the treatment of newly diagnosed malignant glioma; Brain Tumor Co-operative Group (BTCG) trial 8701. [abstract] Congress of Neurological Surgeons 45th Annual Meeting Program, San Francisco, 94-5. Laperriere, N.J., Leung, P.M., McKenzie, S. et al. (1998) Randomised study of brachytherapy in the initial management of patients with malignant astrocytoma. IntJ. Radiat. Oncol. Biol. Phys., 41,1005-11. Sneed, P.K., Prados, M.D., McDermott, M.W. et al. (1995) Large effect of age on the survival of patients with glioblastoma treated with radiotherapy and brachytherapy boost. Neurosurgery, 36,898-903. Sneed, P.K., Gutin, PH., Larson, D.A. et al. (1994) Patterns of recurrence of glioblastoma multiforme after external irradiation followed by implant boost. Int.J. Radiat. Oncol. Biol. Phys., 29,719-27. Agbi, C.B., Bernstein, M., Laperriere, N., Leung, P. and Lumley, M. (1992) Patterns of recurrence of malignant astrocytoma following stereotactic interstitial brachytherapy with iodine-125 implants. Int.J. Radiat. Oncol. Biol. Phys., 23,321-6. Koot, R.W., Maarouf, M., Hulshof, M.C.C.M. et al. (2000) Brachytherapy; results of two different treatment strategies for patients with primary glioblastoma. Cancer, 88,2796-2802.
31. Schupak, K., Malkin, M., Anderson, L, Arbit, E., Lindsley, K. and Leibel, S. (1995) The relationship between the technical accuracy of stereotactic interstitial implantation for high grade gliomas and the pattern of tumour recurrence. Int.J. Radiat. Oncol. Biol. Phys., 32, 1167-76. 32. Loeffler, J.S., Alexander, E., Hochberg, f.H.etal.(1990) Clinical patterns of failure following stereotactic interstitial irradiation for malignant gliomas. Int. J. Radiat. Oncol. Biol. Phys., 19,1455-62. 33. Wen, P. Y., Alexander, E., Black, P.M. et al. (1994) Long term results of stereotactic brachytherapy used in the initial treatment of patients with glioblastomas. Cancer, 73, 3029-36. 34. Sneed, P.K., Lamborn, K.R., Larson, D.A. et al. (1996) Demonstration of brachytherapy boost dose-response relationships in glioblastoma multiforme. IntJ. Radiat. Oncol. Biol. Phys., 35,37-44. 35. Zamorano, L, Ausman, J.I., Chorni, D. and Dujovny, M. (1992) Stereotactic management of midline brain lesions. Stereotact. Funct. Neurosurg., 59,142-50. 36. Bernstein, M., Lumley, M., Davidson, G., Laperriere, N. and Leung, P. (1993) Intracranial arterial occlusion associated with high-activity iodine-125 brachytherapy for glioblastoma.y. Neurooncol., 17,253-60. 37. Kreth, F.W., Faist, M., Warnke, PC., Rossner, R., Volk, B. and Ostertag, C.B. (1995) Interstitial radiosurgery of lowgrade gliomas. J. Neurosurg., 82,418-29. 38. Karim, A.B.M.F., Maat, B., Hatlevoll, R. et al. (1996) A randomised trial on dose-response in radiation therapy of low-grade cerebral glioma: European Organization for Research and Treatment of Cancer (EORTC) Study 22844. IntJ. Radiat. Oncol. Biol. Phys., 36, 549-56. 39. Kreth, F.W., Faist, M., Rossner, R., Birg, W., Volk, B. and Ostertag, C.B. (1997) The risk of interstitial radiotherapy of low-grade gliomas. Radiother. Oncol., 43,253-60. 40. Mundinger, F., Braus, D.F., Krauss, J.K. and Birg, W. (1991) Long-term outcome of 89 low-grade brain-stem gliomas after interstitial radiation therapy./ Neurosurgery, 75, 740-6. 41. Sneed, P.K.,Stauffer, PR., McDermott, M.W.etal.(1998) Survival benefit of hyperthermia in a prospective randomised trial of brachytherapy boost ± hyperthermia for glioblastoma multiforme. IntJ. Radiat Oncol. Biol. Phys., 40,287-95. 42. Crezee, J., Kaatee, R.S.J.P., van der Koijk, J.F. and LagendijkJ.J.W. (1999) Spatial steering with quadruple electrodes in 27 MHz capacitively coupled interstitial hyperthermia. IntJ. Hyperthermia, 15,145-56. 43. Stea, B., Rossman, K., KittelsonJ. et al.(1994) A comparison of survival between radiosurgery and stereotactic implants for malignant astrocytomas. Acta Neurochir., 62,47-54. 44. Shrieve, D.C., Alexander, E., Wen, P.Y. et al. (1995) Comparison of stereotactic radiosurgery and brachytherapy in the treatment of recurrent glioblastoma multiforme. Neurosurgery, 36,275-82.
27 Interstitial brachytherapy in the treatment of carcinoma of the cervix A.M.NISARSYEDANDAJMELA. PUTHAWALA
27.1
INTRODUCTION
It is estimated that approximately 16000 new cases of invasive carcinoma of the cervix were being diagnosed in the United States in 1996, with an estimated 4900 deaths as a result of cervical cancer. The death rate per 100 000 population due to cervical cancer decreased by almost 50% from 1960 to 1990. The relative 5-year survival improved from 58% in 1960 to 69% in 1989 among whites, and from 47% in 1960 to 57% in 1989 among blacks [ 1 ]. The prognosis for patients with cervical cancer mainly depends upon their age, the clinical stage, nodal status, tumor volume, and lymphovascular invasion [2-5]. The overall 5-year survival in patients with stage I and IIA disease is 85-90% [6-9], but only 30-50% with stage III disease. As the tumor volume increases, so does the risk of nodal metastases with poor prognosis. The risk of pelvic recurrence is 8-10% in patients with stage I and IIA disease, as compared to 45-50% in patients with stage III disease [10,11]. Lociano et al [12] reported, from the Patterns of Care Study, an increased risk of in-field failure and poor survival with increasing tumor bulk within stage. Among 1558 patients treated between 1973 and 1978, Perez et al. [13] reported that the tumor size as a single variable was directly related to pelvic failure rate and inversely related to disease-free survival. Lowrey et al. [14] found that tumor size was a
strong independent predictor of pelvic control, distant relapse, and disease-free survival. Although surgery or primary irradiation, or a combination of both, yields 70-90% cure rates in patients with stage I and IIA disease; in locally advanced stage IIB, III, and IVA carcinomas, which are surgically unrespectable, external irradiation with conventional intracavitary application fails to provide an adequate dose to the target volume without extensive bladder and rectal doses. Waterman et al [15] in 1947 and Prempree and Scott [16] in 1978, reported transvaginal interstitial radium needle implant technique in the treatment of cancer of the cervix stage IIIB, with 31% 5-year survival and 78% loco-regional control, respectively. This technique had the disadvantage of excessive exposure to personnel and higher complication rates. We have reported an afterloading technique of interstitial-intracavitary implants using the 'Syed-Neblett' template in the treatment of carcinoma of the cervix [17-20] 27.2
PRETREATMENT WORK-UP
The following is an optimal work-up to determine the histology, tumor grading, volume, local tumor invasion, lymphovascular invasion, staging, etc., for designing appropriate treatment. 1. History and physical examination.
380 Interstitial brachytherapy in the treatment of carcinoma of the cervix
2. Pelvic examination, cystoscopic and proctosigmoidoscopic examination, preferably under anesthesia. 3. Complete blood count (CBC), liver function tests, and serum electrolytes; additionally, a human immunodeficiency virus (HIV) test for patients younger than 25 years of age. 4. Chest X-ray. 5. Intravenous pyelogram and/or computed tomographic (CT) scan of the abdomen and pelvis. 6. Biopsy of cervical tumor. 7. Bipedal lymphangiography and/or retroperitoneal pelvic and para-aortic lymph adenectomy, only for patients on specific protocols.
273 TECHNIQUE The interstitial-intracavitary implant technique utilizes an intracavitary tandem (whenever possible), as in conventional intracavitary applications, but the conventional intravaginal ovoids have been replaced with interstitial ovoids, i.e., transperineally implanted multiple guide needles through the paravaginal and parametrial tissues [17-20]. The interstitial-intracavitary Syed-Neblett applicator consists of a perineal template, vaginal obturator, 17-gauge hollow guide needles, and a tandem (Figure 27.1). The template is made of silicone and has a 2-cm diameter central hole to accommodate the vaginal obturator, and 34 holes drilled 1 cm apart in concentric circles to accommodate the guide needles. The vaginal obturators are 2 cm in diameter and have three different lengths, 12, 15, and 18 cm. The vaginal obturator has a central tunnel to accommodate a tandem and six longitudinal grooves on the surface for guide needles, and an embedded screw at its distal end to secure the tandem. The guide needles are 17 gauge and 20 cm in length. Each needle has the proximal end tapered and closed for easy penetration of the skin and tissues, and a met al ring close to the distal end to prevent
Figure 27.1 Assembled Syed-Neblett applicator.
it from sliding through the template and being left in the tissues when the implant is removed. The pre-implant preparation of the patient is the same as for the conventional intracavitary procedures, i.e., nothing by mouth after midnight and a Fleet's enema on the morning of the procedure. The procedure is performed either under general anesthesia or, more frequently and preferably, under epidural block so that the patients can control the pain following the procedure themselves. The patient is placed in the lithotomy position and an abdominal-pelvic examination is performed to determine the size and extent of the residual tumor and/or metastases, and any anatomical distortions. A proctosigmoidoscopic examination is performed to evaluate any abnormalities and radiation reaction, and also to clean out the feces from the rectum. The perineum and vagina are prepped with Betadine solution and the area is draped as for any surgical procedure. A Foley catheter is inserted into the bladder and the balloon is filled with 7 ml of Hypaque for X-ray localization films. The cervix and vagina are visualized with a vaginal speculum and two gold marker seeds are implanted into the cervix at the two and eight o'clock positions, and another at the lower end of the vaginal extension of the tumor, using an MD Anderson marker applicator. The endometrial canal is sounded and the length measured. The endocervical and endometrial canals are then dilated using Hank or Hegar dilators. A Fletcher or Henschke tandem is inserted into the endometrial canal with the met al flange fixed at the appropriate level according to the length of the endometrial canal. First, a guide needle (which has no guard ring at the distal end) is inserted into the remains of the anterior or posterior lip of the cervix to a depth of 3.5-4 cm. The distance between the distal end of the tandem and the first guide needle is measured, as this distance has to be maintained at the completion of the procedure. The vaginal obturator is inserted into the vagina while the tandem is threaded through its central canal. The template is held against the perineum and the vaginal obturator with the tandem, and the guide needle is threaded through its central hole. A rubber O-ring 2 cm in diameter is threaded over the vaginal obturator and placed into the groove on the template to secure the obturator and the first needle into position. An appropriate number of needles are implanted through the perineal template transperineally on both sides to encompass the initial extent of the tumor, i.e., 20 to 36 needles. The depth of insertion of these guide needles is the same as that of the first guide needle [20] (Figure 27.2). Lateral pressure is exerted on the needles between the perineum and the template while the needles are implanted to prevent central coning of the needles. The needles are dipped in alcohol for easy insertion through the template as alcohol acts as a lubricant for the template material.
Loading and unloading of radioactive sources 381
Figure 27.3 Implant completed for carcinoma of the cervix, stage IVA.
27.5 LOADING AND UNLOADING OF RADIOACTIVE SOURCES
Figure 27.2 Tandem, vaginal obturator, and perineal template are positioned (a), and guide needles are implanted (b).
The tandem is now pushed into the uterus and secured in position by tightening the screw at the distal end of the vaginal obturator with an Allen wrench. The template is secured in position by 2-0 silk sutures through the perineal skin and anterior two corners of the template. The space between the perineum and the template is filled with vaginal gauze, usually soaked in antibiotic cream or saline. A piece of vaginal gauze soaked in barium paste, or a rectal marker, is then inserted into the rectum for X-ray localization films (Figure 27.3).
27.4
LOCALIZATION FILMS
The tandem and the guide needles are loaded with inactive dummy sources, and anteroposterior and lateral orthogonal X-rays are obtained for computerized dose distribution plotting and volume analysis (Figure 27.4).
The tandem is loaded with radioactive cesium-137 and the parametrial guide needles with iridium-192 sources in the patient's room, with usual radiation precautions. The tandem is usually loaded with three sources of cesium-137 of 10 mg Ra-eq at the tip, and two sources of approximately 5 mg Ra-eq each distally, with appropriate spacers according to the length of the endometrial canal. Each of the guide needles is usually loaded with a plastic ribbon containing seven seeds of iridium-192 sources spaced 1 cm apart, having an activity of 0.3-0.45 mg Ra-eq each. The dose to the parametria can be optimized either by differential unloading of tandem and central guide needles or by using a higher activity of iridium-192 sources in the lateral guide needles. The dose can be optimized by utilizing 'remote afterloads' with a single iridium-192 source with their software and computer treatment planning systems for continuous low dose rate (LDR), pulse low dose rate (PDR), medium dose rate (MDR), or high dose rate (HDR). It is desirable to keep the dose rate to the medial parametria, ie, Point A, under 80 cGy h-1 in LDR treatment protocols to minimize complications. The radioactive cesium-137 and iridium-192 source ribbons are unloaded from the tandem and parametrial guide needles after the desired dose is delivered. The radioactive sources are placed in an appropriate lead container, or withdrawn by the remote afterloader, and taken to the isotope storage room.
382 Interstitial brachytherapy in the treatment of carcinoma of the cervix
The patient and the room are surveyed with the Victoreen survey meter for any radioactivity and the implant is then removed. The patient receives morphine
sulfate 10 mg intramuscularly 15 min before removal of the implant. The packing and the perineal sutures are removed. The screw holding the tandem to the vaginal obturator is withdrawn. The guide needles and tandem are removed in one motion by pulling the template from the perineum. The minimal bleeding from the perineum usually stops with the use of gentle pressure with gauze. The Foley catheter is removed and the patient is ambulatory within an hour.
27.6 TREATMENT PROTOCOL Patients receive external irradiation to the pelvic nodes, cervix, and vagina to 5040 cGy in 28 fractions, or 5000 cGy in 25 fractions, 180 or 200 cGy per fraction and five fractions per week, usually with the four-field technique. The rectum and bladder are shielded after 4000 cGy or 3960 cGy, using a midline block. The first interstitial-intracavitary implant is usually performed 1 week following completion of the external irradiation. The second application is performed 2 weeks following the first implant.
27.7 DISCUSSION In cervical cancer, the tumor size, even in the early stages of disease, has been found to be the most significant independent prognostic factor in multivariate analysis in several series [3,5,11].
Figure 27.4 (a) Anteroposterior X-ray localization film with isodose distribution plot overlaid, (b) Lateral X-ray localization film with isodose distribution plot overlaid.
Discussion 383
The treatment of choice for patients with locally advanced cervical cancer, i.e., stages IIB, III, and IVA, is radiation therapy with or without systemic chemotherapy and possible pelvic exenteration for patients with stage IVA disease. However, a combination of external and conventional intracavitary irradiation results in 40-50% pelvic failures, with significant complications. However, in most series, pelvic failures increased proportionately, i.e., 40-60%, with advanced stages of disease treatment with a combination of external and conventional intracavitary irradiation [11,21-23]. These treatment failures have essentially been due to large tumor volume and inadequate doses delivered by the intracavitary applicators. The radioactive sources in the tandem and ovoid could not deliver high enough doses to the target volume without excessive doses to the rectum and bladder. Anatomical distortions, even in early disease, i.e., narrowing of vagina, obliterated fornices, or inability to use tandem, can also cause failure to deliver adequate doses to the tumor volume. Waterman et al. [13,24] reported a 31% 5-year survival for patients with stage IIIB carcinoma of the cervix using a transvaginal interstitial radium needle implant technique. Prempree and Scott [16] reported 78% local tumor control using a similar technique in 1978 in stage IIIB carcinoma of the cervix. This technique of interstitial implant has several inherent technical and radiationexposure problems. We reported the technique of interstitial-intracavitary applicators using the 'Syed-Neblett' template with preliminary results in advanced carcinoma of the cervix in 1978 [17-20]. This technique involves the use of an intracavitary tandem and interstitial ovoids, i.e., multiple guide needles inserted into the parametria transperineally through the template. The technique is easily reproducible and provides excellent dose distribution to the tumor volume with relative sparing of critical structures, i.e., rectum and bladder.
The technique lends itself to HDR treatments utilizing commercially available remote afterloaders. Table 27.1 reflects the results during the evolution of the technique by several authors. Geddis et al. [25] reported, in 1983, a 14% rate of severe complications. These higher complications occurred during the evolution of the technique and have been mainly due to: (1) extensive necrotic tumors; (2) lack of availability of computer dosimetry, so the dose was based on milligram-hours, thus delivering much higher doses; (3) high dose rates, i.e., 120-200 cGy h~', to point A; and (4) point A received a total of 10000-14000 cGy by a combination of external and interstitial irradiation with higher doses to the rectum and bladder. Aristizabal et al. [26] reported 76% and 74% pelvic control in stages IIB and IIIB, respectively, and reduced the complication rate from 33% to 6% by modification in implant geometry and reduction of total milligram-hours. Ampuero et al. [27] reported 38% local recurrence and 29% complications in 24 patients with stage IIB and IIIB cervical cancer. Martinez et al [28] published results of 70 patients with locally advanced cervical carcinoma treated by transperineal implants using the Martinez Universal Perineal Implant Template (MUPIT) and achieved 66% local control and 14% severe complications. We reduced the dose rates and also the doses to the rectum and bladder by dose optimization and by differential unloading or using higher activity sources in the lateral parametrial needles (0.4-0.5 mg Ra-eq) and lower activity sources (0.2-0.3 mg Ra-eq) in the medial parametrial needles. The complication rate in subsequent series of patients has been reduced to 3% while maintaining the loco-regional control at 78% in stage IIB, III, and IVA carcinomas of the cervix (Tables 27.2 and 27.3). This finding is supported by others [33] who report that grade IV complication rates can be reduced from 14% to 3% when dose rates are reduced to below 70% cGy h~' without jeopardizing disease control rates.
Table 27.1 Interstitial irradiation (afterloading technique) for carcinoma of the cervix
Pitts and Waterman, 1940
IIIB
Prempree and Scott (1978) [16]
IIIB
110
49
31 (5-year)
13
60 (median 24 months)
8
78
Feder, Syed and Neblett (1978) [18] IIIB
Syed-Neblett
38
Geddisrto/. (1983) [25]
I, II, III, and IV
Syed-Neblett
84
71
Aristizabal et al. (1983)
Advanced Hand IIII
Syed-Neblett
21
85
Ampuero et al. (1983) [27]
IIBandlllB
Syed-Neblett
24
72
29
Aristizabaltfo/. (1985) [26]
Advanced Hand IIII
Syed-Neblett
118
75
21
Syed et al. (1986) [20]
I, II, III, and IV
Syed-Neblett
60
78
' Severe complications include severe proctitis, cystitis, rectovaginal fistula and/or vesicovaginal fistula.
14 58
58
35
3
384 Interstitial brachytherapy in the treatment of carcinoma of the cervix
Table 27.2 Carcinoma of the cervix, Syed-Neblett template: complications
IB IIAandllB IIIAandlllB IVA
6 23 26 0
0 2 1 0
0 2 1 0
0 0 2 0
0 0 0 0
All
60
3(5%)
3(5%)
2(3%) 0
RV, rectovaginal; W, vesicovaginal. Table 27.3 Carcinoma of the cervix, Syed-Neblett template: patterns of failure
IB IIAandllB IIIAandlllB IVA
6 23 26 5
0 2(+2Y 4 1
0 2 4 0
0 5 5 3
All
60
9 (1 5%)
6(10%)
13(22%)
'Two patients had only central recurrence; the other two failed only in the parametria.
cumulative dose of 7706 cGy at point A using a standard Fletcher-Suit technique. The interstitial group received a mean external dose of 5050 cGy and two interstitial implants using a transperitoneal Syed-Neblet template with a mean tumor dose of 2239 cGy and 1942 cGy for each application, respectively. No statistical difference could be detected in survival for stage III and IVA patients, but for stage II patients the intracavitary results were better, with more relapses in the interstitial group. However, the intracavitary group received a larger dose than the interstitial group (4608 versus 3504 radium milligram-hours equivalent) because a tandem was only used in 24% of the interstitial implants. Complications occurred in 21% of patients in each group. We have published the guidelines to be followed while employing the Syed-Neblett template technique to minimize the complications and maximize the local tumor control [20]. Bloss et al. [31] reported improvement in local control by utilizing the interstitial-intracavitary applicator with radiofrequency hyperthermia for radiopotentiation in patients with locally advanced and necrotic cervical cancers. The following are our conclusions and current indications for the use of interstitial-intracavitary applications:
1. Interstitial-intracavitary application using the SyedNeblett technique is safe and easily reproducible. The clinical outcome for HDR interstitial brachyther2. Loco-regional control of 70-77% can be achieved in apy in combination with external-beam irradiation [34] stage IIB and III cancers, with less than 4% severe has also reported satisfactory local and regional control complications. results. Six fractions were used of 5.5-6.0 Gy interstitial 3. This technique lends itself to manual, continuous therapy in combination with external-beam irradiation, LDR, pulse LDR, MDR, and HDR utilizing remote with central shielding, to a dose of 50 Gy to the pelvic afterloaders. side walls. The patients chosen had locally advanced dis4. Interstitial hyperthermia, i.e., radiofrequency or ease which precluded satisfactory tandem and ovoid microwave, can be utilized with the template insertion. This is perhaps an indication for interstitial technique for radiopotentiation. treatment that requires further consideration. 5. It is preferrable to use this technique in patients Hockel and Muller [29] modified the Syed-Neblett with: (a) cervical cancer stages IIB, III A, IIIB, and template for HDR brachytherapy of gynecological IVA, and in patients with stage IB and IIA with malignancies to reduce the rectal and bladder doses. The distorted anatomy, i.e., narrow vagina and template's assembly allows cystoscopic and rectoscopic obliterated fornices; (b) inability to use tandem in all control of needle positions. Wolkov et al. [30] constages; (c) carcinoma of the cervical stump; and (d) cluded, from the results of 14 patients with locally recurrent carcinoma of the cervix who have not advanced cervical cancer and review of the literature, been properly selected. that the loco-regional recurrence rate utilizing a transperineal template technique is approximately 50% less than the traditional intracavitary irradiation, with an REFERENCES overall complication rate of 18%, comparable to that reported by the Patterns of Care Study for similar stages 1. Parker, S.L., Tong, T., Bolden, S. and Wingo, P.A. (1996) of cervical cancer. Cancer statistics, 1996. CA Cancer J. Clin., 46, 5-10. Monk and colleagues [32] have reported different 2. International Federation of Gynecology and Obstetrics results. They carried out a retrospective assessment of (1991) Annual report on the results of treatment in the experience of two institutions, one of which used gynecological cancer. Int.J. Gynaecol. Obstet, 36(Suppl.), intracavitary brachytherapy in combination with 27-30. teletherapy and the other used interstitial brachytherapy 3. KovalicJ.J., Perez, C.A., Grigsby, P.W.etal.(1991) The with teletherapy. Patients from the two groups were simeffect of volume of disease in patients with carcinoma of ilarly matched. The intracavitary group received a mean
References 385
the uterine cervix. Int.J. Radial. Oncol. Biol. Phys., 21, 905-10. 4. White, CD., Morley, G.W. and Kumar, N.B. (1984) The prognostic significance of tumor emboli in lymphatic or vascular spaces of the cervical stroma in stage IB
17. Syed, A.M.N. and Feder, B.H. (1977) Technique of afterloading interestitial implants. Radiolog. Clin., 48, 458-75. 18. Feder, B.H., Syed, A.M.N. and Neblett, D.L (1978) Treatment of extensive carcinoma of the cervix with the 'transperineal parametrial butterfly:'a preliminary
squamous cell carcinoma of the cervix. Am.J. Obstet. Gynecol., 149,342-9. 5. VanBrommel, P., VanLindert, A., Kock, H.et al. (1987) A review of prognostic factors in early stage carcinoma of the cervix (FIGO IB and MA) and implications for treatment
report on the revival of Waterman's approach. Int. J. Radial Oncol. Biol. Phys., 4,735-42. 19. Syed, A.M.N. and Neblett, D.L. (1997) Interstitialintracavitary applicator in the management of gynecologic malignancy. Proceedings of the Pacific
strategy. Eur.J. Obstet. Gynecol. Reprod. Biol., 26,69-84.
Endocurietherapy Society, Winter Meeting, December
6. HorrotJ.C, PyneuxJ., Pourquier, H.et al. (1988)
5-7, Mazatlan, Mexico.
Radiotherapy alone in carcinoma of the intact uterine cervix according to GH Fletcher guidelines: a French cooperative study of 1383 cases. Int. J. Radial. Oncol. Biol.
20. Syed, A.M.N., Puthawala, A.A., Neblett, D.L et al. (1986) Transperineal interstitial-intracavitary 'Syed-Neblett' applicator in the treatment of carcinoma of the cervix.
Phys., 14,605-11.
Endocuriether./Hyperlherm. Oncol., 2,1-13.
7. Komaki, R., Brickner, T.J., Hanlon, A.L, Owen, J.B. and Hanks, G.E. (1995) Long-term results of treatment of cervical carcinoma in the United States in 1973,1978 and 1983: patterns of care study (PCS). Int.J. Radial. Oncol. Biol. Phys., 31,973-82. 8. Liebel, S., Bauer, M., Wasserman, T. et al. (1987)
21. Aristizabal, S.A., Giever, R.J., Duque, A. and Surwit, E. (1981) Invasive carcinoma of the cervix treated primarily with radiation therapy - University of Arizona experience. Ariz. Med., 38,613-16. 22.
Molina, J., Guillen, R., Rodriguez, E. and Velasquez, J.
Radiotherapy with or without meconidazole for patients
(1980) Radioterapia en cancer cervico uterino en etapa
with stage 1MB or stage IVA squamous cell carcinoma of
clinica 1MB (Spanish). ResulladosaCincoAnos GinecObstet Mex,47,419-31.
the uterine cervix: preliminary report of a Radiation Therapy Oncology Group randomized trial. Int.J. Radial. Oncol. Biol. Phys., 13, 541-9.
23.
9. Runowicz, C.D., Wadler, S., Rodriguez, L et al. (1989) Concomittant cisplatin and radiotherapy in locally advanced cervical carcinoma. Gynecol Oncol., 34,395-401. 10. Jampolis, S., Andras, J. and Fletcher, G.H. (1975) Analysis
Kapp, D.S., Fischer, D., Gutierrez, E. et al. (1983) Pretreatment prognostic factors in carcinoma of the uterine cervix: a multivariate analysis of the effects of age, stage, histology and blood counts on survival. InlJ. Radial Oncol. Biol. Phys., 9,445-55.
of sites and causes of failure of irradiation in invasive
24. Waterman, G.W. and Raphael, S.I. (1952) The role of interstitial radium therapy in the treatment of cancer of
squamous cell carcinoma of the intact uterine cervix. Radiology,115,681-5.
25. Geddis, D., Morrow, C.P., Klement, V., Schlaerth, J. and
the cervix uteri. Am. J. Roentgenol., 68, 58-62.
11. Montana, G.S., Martz, K.L and Hanks, G.E. (1991) Patterns
Nalick, R.H. (1983) Treatment of cervical carcinoma
and sites of failure in cervix cancer treated in the USA in 1978. IntJ. Radial Oncol. Biol. Phys., 20,87-93. 12. Lociano, R.M., Won, M. and Hanks, G.E. (1992) A reappraisal of the International Federation of Gynecology
employing a template for transperineal interstitia!192lr brachytherapy. Int.J. Radial Oncol. Biol. Phys., 9, 819-27. 26. Aristizabal, S.A., Valencia, A., Ocampo, G. and Surwit, E.A.
and Obstetrics staging system for cervical cancer: a study
(1985) Interstitial parametrial irradiation in cancer of the
of the patterns of care. CA Cancer J. Clin., 69,482-7.
cervix stage MB and 1MB: analysis of pelvic control and complications.Endocuriether./Hypertherm. Oncol.,'1,
13. Perez, C.A., Kurman, R.J., Stehman, F.B. et al. (1992) Uterine cervix. W.J. Huskins, C.A. Perez and R.C. Young. In Principles and Practice of Gynecologic Oncology, ed.
41-8. 27. Ampuero, F., Doss, L.L, Khan, M., Skipper, B. and Hilgers,
W.J. Huskins, C.A. Perez and R.C. Young, Philadelphia,
R.D. (1983) The Syed-Neblett interstitial template in
JBLippincott Co., 591-662. 14. Lowrey, G.C., Mendenhall, W.M. and Million, R.R. (1992) Stage IB or IIA,B carcinoma of the intact uterine cervix treated with irradiation: a multivariate analysis. Int.J. Radial Oncol. Biol. Phys., 24,205-10. 15. Waterman, G.W., Dileone, R. and Tracy, E. (1947) The use of long interstitial radium needles in the treatment of cancer of the cervix. Am. J. Roenlgenol., 57,671 -8. 16. Prempree, T. and Scott, R.M. (1978) Treatment of stage 1MB carcinoma of the cervix: improvement in local
locally advanced gynecological malignancies. InlJ. 28.
Radial Oncol. Biol. Phys., 9,1897-903. Martinez, A., Cox, R.S. and Edmundson, G.K. (1984) A multiple site perineal applicator (MUPIT) for treatment of prostate, anorectal and gynecologic malignancies. Int.J. Radial Oncol. Biol. Phys., 10,297-305.
29. Hockel, M. and Muller, T. (1994) A new perineal template assembly for high-dose rate interstitial brachytherapy of gynecologic malignancies. Radial Oncol., 31,262-4. 30. Wolkov, H.B., Manjat, J., Ordoridca, E. and Trelford, J.
control by radium needle implant to supplement the
(1987) Interstitial templates for locally advanced cervix
dosetotheparametria. Concur, 42,1105-13.
cancer. Med. Dosimetry, 12(1), 21-4.
386 Interstitial brachytherapy in the treatment of carcinoma of the cervix 31. Bloss, J.D., German, M.L, Syed, A.M.N. et al. (1992) Treatment of advanced carcinoma of the uterine cervix with interstitial radiotherapy and hyperthermia. Endocuriether./Hypertherm. Oncol., 8,145-50. 32. Monk, B.J., Tewari, K., Burger, R.A., Johnson, M.T., Montz, F.J. and Berman, M.L (1997) A comparison of intracavitary versus interstitial irradiation in the treatment of cervical cancer. Gynecol. Oncol., 67(3), 241-7.
33. Gupta, A.K., Vicini, F.A., Frazier, A.J. et al. (1999) Iridium192transperineal interstitial brachytherapy for locally advanced or recurrent gynaecological malignancies. Int. J. Radial Oncol. Biol. Phys., 43(5), 1055-60. 34. Demanes, D.J., Rodriguez, R.R. and Ewing, T.L (1999) High dose ratetransperineal interstitial brachytherapy for cervical cancer: high pelvic control and low complication rates: Int.J. Radial Oncol. Biol. Phys., 45(1), 105-12.
28 Interstitial brachytherapy in the treatment of carcinoma of the anorectum AJMELA. PUTHAWALAANDA.M. NISARSYED
28.1
INTRODUCTION
Surgery has played a dominant role in the definitive treatment of carcinoma of the lower rectum and anus, which has typically required abdominoperineal resection with permanent colostomy. However, over the past two decades a more conservative approach has been sought in an attempt to preserve the organ and its function. This is especially true for carcinoma of the anus, for which a combined modality, i.e., radiation therapy and systemic chemotherapy, has yielded up to 85% local control and preservation of the sphincter function in 65% of patients [1-5,22,23]. Newer surgical techniques, i.e., using an endorectal staple gun or pull-through procedure, as well as laser fulguration of the lower rectal cancers allow the preservation of the sphincter function [6-10]. The use of preoperative radiation therapy as well as chemotherapy has also facilitated organ preservation for borderline cases which would otherwise require abdominoperineal resection and permanent colostomy [11-13]. As early as 1915 Cade [14] reported the use of an intracavitary radium applicator for lower rectal cancers. In 1975, Papillon [15] reported a large series of patients who were treated successfully using endocavitary radiation for early-stage cancer of the lower rectum. In 1982, his updated results were published in the book entitled Rectal and Anal Cancers. Conservative Treatment by Irradiation: an Alternative to Radical Surgery [16]. For the past two decades we have used definitive radia-
tion therapy with a combination of external-beam irradiation and interstitial iridium-192 afterloading implants in the treatment of carcinoma of the anus and lower rectum in lieu of surgery [ 17,18]. Our experience is based on treating most of the patients with large bulky disease who were considered unresectable and/or inoperable because of advanced age and medical problems. We reported in 1982 the treatment results of 40 such individuals who underwent a combination of external-beam irradiation and interstitial iridium-192 implant [19]. Seventy percent of the patients achieved local tumor control with median follow-up of 36 months. Since then, at our institution, we have extended the use of definitive radiation therapy to include patients with relatively early stages of cancer as an alternative treatment to abdominoperineal resection. In 1987, Papillon and Montbarbon [20] reported a 3-year local control rate of 80% among 276 patients with epidermoid carcinoma of the anal canal treated by a combination of external-beam and interstitial iridium-192 implant radiation. Ninety percent of these patients had retained normal anal function. 28.2 PRETREATM ENT ASSESSM ENT AN D INVESTIGATIONS The pretreatment work-up should include complete history and physical examination, assessment of the regional lymph nodes, proctosigmoidoscopy, biopsy, as well as metastatic work-up including chest X-ray, liver
388 Interstitial brachytherapy in the treatment of carcinoma of the anorectum
function studies, complete blood count, urinalysis, coagulation profile, and serum human immunovirus. Transanorectal ultrasonography may be useful to evaluate the thickness of the lesion and penetration of any pelvic organs. Computer tomographic (CT) and magnetic resonance imaging (MRI) may be useful in some cases.
283
TREATMENT PROTOCOL
For epidermoid carcinoma of the anal canal, if there is no contraindication for systemic chemotherapy, these patients should first be treated with a combination of external-beam irradiation and continuous 5-fluorouracil intravenous infusion for 3-5 days, starting with the first day of external irradiation, with or without mitomycin-C or Cisplatin. All patients, except patients with early stage Tl tumors of the anorectum, receive a minimum dose of 4500 cGy delivered on megavoltage unit using anteroposterior (AP) and posteroanterior (PA) parallel opposing ports to encompass the primary disease and the draining lymphatics, using 1.8-2 Gy per fraction, five fractions per week. Patients with recurrent tumors after definitive surgery who have not received either preoperative or postoperative adjuvant radiation therapy should also be treated with external irradiation as above.
28.4
INTERSTITIAL BRACHYTHERAPY
The patient should have at least a 2-3-week rest period after completion of external-beam irradiation. One or two applications of interstitial brachytherapy may be required, depending upon the bulk of the original lesion. Each implant application may deliver a total tumor dose of 15-25 Gy over 30-60 h. Preferable dose rate for the anorectal implant is 0.4-0.5 Gy h'1. If two implants are planned, then the interval between implants should be at least 3-4 weeks. The implants are usually performed under general anesthesia; however, epidural or spinal anesthesia is preferable and the indwelling epidural catheter could also be left in for the duration of the implant for pain management.
28.5
EQUIPMENT
The equipment includes the following: Syed-Neblett disposable rectal template (Figure 28.1), a set of 15-20 cm long 17-gauge, hollow, stainless-steel needle guides, number 36 rectal tube or, preferably, plastic chest tube, marking pen, ruler, non-radioactive gold marker seeds, wire stylette, one 20 cm long 17-gauge open needle for marking seed insertion, 1-0 silk suture with non-cutting Gl needle, 2-inch roller gauze, 5 ml of Hypaque, triple sulfa cream, and Foley catheter 16 French, as well as alcohol to soak the template and to lubricate the guide
Figure 28.1 Syed-Neblett disposable rectal template (right) and classic (multiple use) rectal template (left).
needles so they can be easily inserted through the template (the needles are locked into position in the template after complete evaporation of alcohol). If the non-disposable template is used, then Allen-head screws should be loosened to allow the needles to pass through the template. After insertion of all the needles, screws should be tightened to fix the needles in place [13]. 28.6 INTERSTITIAL BRACHYTHERAPY TECHNIQUE After induction of satisfactory anesthesia, the patient is placed in the lithotomy position. The abdomen and perineum are prepped and draped in the usual fashion. The patient is then catheterized with a Foley catheter with the bulb inflated with 5 ml of Hypaque for localization of the bladder. The Foley catheter should be attached to the drainage tube and to the collecting bag. For male patients, the scrotum and penis should be pulled up with gauze and should be held by the abdominal drape with an Alice clamp. Prior to prepping and draping of the patient, a proctosigmoidoscopy should be carried out to ensure the satisfactory evacuation of the rectosigmoid colon and, more importantly, to register any undue reactions to the previously given radiation, as well as to locate the tumor and to take the measurements, i.e., extent of the tumor along the longitudinal axis in relation to the anal verge as well as circumferential involvement along the lumen. A bimanual examination should also be carried out and the tumor location should be marked on the perianal skin in relation to its circumferencial involvement (Figure 28.2). Gold marker seeds are then inserted transperineally with the finger in the rectum to define the superior-most and inferior-most extent of the palpable tumor as well as the lateral-most extent of the tumor. Usually, four marker seeds are implanted to help define the tumor volume on the radiographs. The first guide needle is then inserted transperineally through the perianal skin, guiding it with the finger in the rectum, keeping the needle just under the mucosa. The first needle should be placed approximately
Interstitial brachytherapy technique 389
at the center of the tumor on the longest extension of its vertical axis. The tip of the needle should be advanced at least 2-3 cm beyond the palpable lesion. The depth of insertion of this needle should then be measured from the perineal skin (Figure 28.2) and a Kelly clamp should be placed on the needle near the skin. The rectal tube or, preferably, a number 36 French plastic chest tube, is then
inserted in the rectum to a depth of at least 10-12 cm. A1 -0 silk suture should then be taken through the perianal skin and through the tube to secure the implant in place. The sutures should be placed opposite to the implant area so the suture will not be cut off while the needles are being inserted. A disposable Syed-Neblett rectal template is then placed against the perineum, letting the first guide needle pass through the appropriate hole in the innermost circle and the rectal tube pass through its large central hole (Figure 28.3). The template is then held against the perineum in a fixed position, and two or three more needles are inserted to a similar depth on the side of the perineal area where the tumor extension was marked on the skin earlier. In the second circle encompassing the circumference with a 1-cm margin on either side should be completed, i.e., six to nine guide needles, and in the third circle, again, six to eight needles are inserted to a similar depth through alternate holes. The needles should be dripping wet with alcohol to facilitate their passage through the template. 1-0 silk sutures are then taken through the perineal skin and the upper outer corner holes on the template to keep it in position. Two-inch roller gauze impregnated with triple sulfa cream is then applied lightly between the template and the perineal skin (Figure 28.4). The patient is then sent to the recovery room.
Figure 28.3 Plastic tube (No. 36 French) in rectum, and first guide needle through the template which is held against the perineum.
Figure 28.4 Template is sutured to the perineal skin after insertion of all required afterloading guide needles.
Figure 28.2 Tumor location is marked on the perineal skin. The first guide needle is placed with the finger in the anorectum. Total depth of insertion is measured from the anal verge.
390 Interstitial brachytherapy in the treatment of carcinoma of the anorectum
Prior to starting the interstitial implant, it is recommended that the patient receives 1 or 2 g of a broadspectrum antibiotic intravenously, and antibiotics should be continued for the duration of the implant. AP and lateral orthogonal X-rays with dummy sources in all
guide needles are taken for computer dosimetry. The number of iridium-192 seeds to be loaded in each guide needle is decided in the operating room after taking the measurements, with usually a 2-3 cm margin given superiorly and inferiorly. If the tumor extends through the perianal skin, the sources are usually brought down beyond the skin; otherwise, the sources should be kept under the skin to avoid a brisk, moist skin reaction. The implant is then loaded with iridium-192 ribbons, either manually or using the remote afterloading device. The computerized isodose distribution on both the X and Z axes at 5-mm intervals is obtained. The X and Z axes are usually placed through the center of the implant. The isodose planes at X = 0 and Z - 0 are then directly overlaid on the AP and lateral orthogonal radiographs (Figure 28.5). The isodose line encompassing the tumor volume adequately is chosen as minimum dose rate [21]. The dose-volume histograms are also obtained (Figure 28.6). The duration of the implant is then calculated by dividing the total dose to be delivered by dose rate. At the completion of interstitial irradiation, the radioactive sources are removed, again either manually or by remote afterloading device. The sutures are removed from the perineal skin and the template is removed by pulling it away from the perineum so all the needles will be removed at once. Betadine solution is then applied to the skin and light pressure is exerted with dry gauze for a few minutes to stop the bleeding. The applicator, the patient, and the room should be checked with a Geiger counter to ensure that all sources have been removed and placed in the lead container. If the patient has no continuous epidural pain medication, intramuscular Demerol (meperidine hydrochloride), 50-100mg, should be given at least 10-15 min prior to removal of the implant. Figure 28.5 Anteroposterior (a) and lateral (b) orthogonal radiographs with computergenerated isodose distribution overlaid for dose calculation.
Results 391
Figure 28.6 Dose-volume histogram.
28.7
AFTERCARE
The patient should be advised to take sitz baths once or twice a day for the first week to 10 days. The patient may also need steroid cream or Cortifoam enema to help alleviate the pain and discomfort from ensuing proctitis. Usually, acute proctitis subsides in 2-3 weeks on conservative management.
tion of this treatment had extensive tumors (T3 or T4) or they were treated for recurrent cancer (Table 28.2). Local tumor response has been assessed on clinical examinations, including proctosigmoidoscopy and repeat biopsies, if indicated. Most lesions resolved 2-3 months after completion of this treatment. Occasionally, induration or mucosal thickening may persist for a long time at the site of the original tumor secondary to fibrosis. In the past, some of these patients had undergone repeated biopsies because of concern about persistent or recurrent tumor and had developed ulceration and necrosis. Most recurrences or persistent disease are clinically manifested by 12-18 months and, rarely, after 24 months following this treatment. Histologic grade and type had no significant impact on the rate of local tumor control. The overall complication rate has been 19.6% (Table 28.3). The most common complication of this treatment is rectal ulceration or necrosis. Tumor-associated necrosis usually resolves on conservative management if treatment is successful. Hyperbaric oxygen therapy is usually successful in patients who develop persistent ulceration and necrosis. It is advisable for these patients to undergo temporary Table 28.1 Anorectal carcinoma stage distribution
T1 T2 T3-4 Recurrent Total
28.8
15 50 38 60 163
RESULTS Table 28.2 Anorectal carcinoma local control at 3 years
One hundred and sixty-three patients with a diagnosis of carcinoma of the anorectum were treated between 1974 and 1992: 103 patients were treated for primary cancer and 60 patients were treated for recurrent cancer (Table 28.1); 132 patients had adenocarcinoma and 31 patients had squamous cell carcinoma. A cumulative local tumor control was achieved in 130 (80%) of 163 patients at 3 years of follow-up (Figure 28.7). Most patients who had either persistent or recurrent disease after comple-
n T2 T3-4 Recurrent
15/15(100%) 46/50 (92%) 30/38 (79%) 39/60 (65%)
Total
130/163(80%)
Table 28.3 Anorectal carcinoma complications
Rectal ulceration and necrosis Anal stricture and fibrosis (with or without incontinence) Hemorrhage Perianal infection or ischio-rectal abscess RectovaginaI fistula Total Figure 28.7 Cumulative local control. REC, recurrent.
19/163(11.7%) 7/163(4.3%) 2/163(1.2%) 3/163(1.8%) 1/163(0.6%) 32/163(19.6%)a
'24/163 (14.7%) patients required colostomy because of complications.
392 Interstitial brachytherapy in the treatment of carcinoma of the anorectum
colostomy, and they should also have no evidence of persistent or recurrent disease proven on biopsy. Anal stricture and fibrosis with incontinence are rare complications and are often associated with circumferential implants. Twenty-four (14.7%) out of 163 patients required palliative colostomies following treatment because of persistent ulceration, necrosis, or anal stricture and tenesmus. Most of these patients had already been treated for extensive disease. This treatment-related complication rate can be minimized if patients are selected carefully who have limited disease which does not require more than half the circumference implant and active length of more than 7 cm. Salvage surgery is feasible for a substantial number of patients who may fail this treatment regimen. It often requires abdominoperineal resection with permanent colostomy. In our series, five (42%) out of 12 patients who failed this regimen were salvaged by abdominoperineal resection. The ideal lesions to be treated with definitive radiation therapy using a combination of external-beam irradiation and interstitial brachytherapy: (1) are located within 8 cm from the anal verge; (2) involve less than half the circumference of the lumen; (3) measure less than 3 cm in thickness and less than 6 cm in length; (4) have no complete fixation to the pelvic bones or visceral invasion; and (5) have no extensive ulceration.
REFERENCES 1. Cummings, B.J., Harwood, A.R., Kean, T.J., Thomas, G.M.
8. Wolmark, N. and Fisher, B. (1986) An analysis of survival and treatment failure following abdominoperineal and sphincter-saving resection in Dukes' B and C rectal carcinoma: a report of the NSABP clinical trials. Ann. Surg., 204,480-9. 9. Saadia, R. and Schein, M. (1988) Local treatment of carcinoma of the rectum. Surg. Gynecol. Obstel, 166, 481-6. 10. Yeatman,T.J.and Bland, K.I. (1989) Sphincter-saving procedures for distal carcinoma of the rectum. Ann. Surg., 209,1-18. 11. Kligerman, M.M. and Urdanetta-Lafec, N. (1974) Observation on 15 operable/nonresectable cases of rectal cancers given preoperative irradiation. Am. J. Roentgenol. Radium Ther. Nucl. Med., 120,624-6. 12. Roswit, B., Higgins, G.A. and Kahn, R.J. (1975) Preoperative irradiation for carcinoma of the rectum and rectosigmoid colon: report of a national Veterans Administration study. Cancer, 35,1597-602. 13. JessupJ.M., Bleday, R., Busse, P.et al. (1993) Conservative management of rectal carcinoma: the efficacy of a multimodality approach. Semin. Surg. Oncol., 9,39-45. 14. Cade, S. (1950) Malignant Disease of the Rectum and Anus, Vol 3: Textbook on Malignant Disease and its Treatment by Radium. Baltimore, Williams and Wilkins. 15. Papillon, J. (1975) Intracavitary irradiation of early rectal cancers for cure: a series of 186 cases. Cancer, 36,696-701. 16. Papillon, J. (1982) Recta I and Ana I Cancers. Conservative Treatment by Irradiation: an Alternative to Radical Surgery. New York. Springer-Verlag. 17. Syed, A.M.N., Puthawala, A.A., Neblett, D.L. et al. (1978) Primary treatment of carcinoma of the lower rectum and anal canal by a combination of external irradiation and
and Rider, W.D. (1980) Combined treatment of squamous cell carcinoma of the anal canal: radical radiation therapy with 5-fluorouracil and mitomycin-C: a preliminary report. Dis. Colon Rectum, 23,389-91.
interstitial implant. Radiology, 128,199-203. 18. Syed, A.M.N. and Feder, B.H. (1977) Technique of afterloading interstitial implants. Radiol. Clin., 46,
2. Cummings, B.J., Thomas, G.M., Kean,T.J.ef al. (1982) Primary radiation therapy in the treatment of anal canal carcinoma. Dis. Colon Rectum, 15,778-82. 3. Nigro, N.D. (1984) An evaluation of combined therapy for squamous cell cancer of the anal canal. Dis. Colon
458-75. 19. Puthawala, A.A., Syed, A.M.N., Gates, T.C. and McNamara, C. (1982) Definitive treatment of extensive anorectal carcinoma by external and interstitial irradiation. Cancer, 50,1746-50.
Rectum, 27,763-6. 4. Sischy, B. (1985) The use of radiation therapy combined
20. Papillon, J. and Montbarbon, J.F. (1987) Epidermoid carcinoma of the anal canal: a series of 276 cases. Dis.
with chemotherapy in the management of squamous cell
Colon Rectum, 30,324-33.
carcinoma of the anus and marginally resectable adenocarcinoma of the rectum. Int.J. Radial Oncol. Biol. Phys.,! 1,1587-93. 5. John, M.J., Flam, M., Lovalvol, L.J.etal.(1987) Feasibility of non-surgical definitive management of anal canal carcinoma. Int.J. Radial Oncol. Biol. Phys., 13,299-303.
21. Neblett, D.L, Syed, A.M.N., Puthawala, A.A., Harrop, R., Frey, H.S. and Hogan, S.E. (1985) An interstitial implant technique evaluated by contiguous volume analysis. Endocuriether./Hypertherm. Oncol., 1,213-22. 22. Sandhu, A.P., Symonds, R.P., Robertson, A.G., Reed, N.S., McNee, S.G. and Paul, J. (1998) Interstitial iridium-192 implantation combined with external radiotherapy in
6. Bacon, H.E. (1956) Abdominoperineal proctosigmoidectomy with sphincter preservation: Sand 10
anal cancer: ten years experience. Int.J. Radial Oncol.
years after 'pull through' operation for cancer of the
Biol. Phys., 40(3), 575-81.
rectum.JAMA, 160,628-34. 7. Rich, T.A., Weiss, D.R., Mies, C. et al. (1985) Sphincter preservation in patients with low rectal cancer treated
23.
Broens, P., Van Limbergen, E., Penninckx, F. and Kerremans, R. (1998) Clinical and manometric effects of
with radiation therapy with or without local excision or
combined external beam irradiation and brachytherapy
fulguration. Radiology, 156, 527-31.
for anal cancer. Int.J. Colorectal Dis., 13(2), 68-72.
29 High dose-rate brachytherapy in the treatment of skin tumors CAJOSLINANDA. FLYNN
29.1
INTRODUCTION
Epidermal skin in fair-skinned people is particularly at risk of developing skin cancer. In recent times, the risk has greatly increased with changes in social habits such as increased exposure to sunlight [ 1,2]. The frequency of skin cancer approaches 10% of all cancers, depending upon the country and its ethnic population. The risk increases as age increases [3], the commonest age group affected being 60 years and older. The site and extent of the disease, the histology, and the medical fitness of the patient can determine the type of treatment used. Basal cell carcinomas (or rodent ulcers) most commonly affect the skin of the head and neck regions, remain localized, and require local radical treatment. Less commonly, rodent ulcers occur in sites other than around the head. Squamous cell carcinoma may also involve the skin of the head and neck region, but more commonly involve the skin over the dorsum of the hand, particularly when these areas are exposed to ultraviolet light over long periods [1,2]. Radiation therapy can be an effective and satisfactory treatment for the majority of cases of non-melanoma skin cancer. In a review of the literature, Halpern advised that, with efficient methods of dose fractionation and delivery of radiotherapy, skin cancers could be controlled in over 90% of cases treated [4]. He also advised that, in general, the cosmetic appearance and function are better preserved under most circumstances. How-
ever, other forms of treatment such as curettage and cautery are effective against small rodent ulcers in the majority of patients. Surgery may also be used, but this can cause disfigurement in larger lesions, although sometimes a combination of surgery and radiotherapy is indicated. The type of radiotherapy and method of application are often determined by their practicability and the risk of exposure to staff and other people. One method in use for many years involved surface treatment moulds loaded with radium-226 as the active material [5]. More recently, radium-226 has been replaced by cesium-127 or iridium-192. However, there remains the potential radiation hazard arising from preparation of the sources, the loading of the mould, and also from nursing the patient during treatment. Because of this, the use of manually preloaded surface moulds went out of fashion, and nowadays most treatment is given using either superficial X-rays or low-energy electrons [4]. However, when the treatment volume includes cartilage and tendons, where the superficial tissues are thin and overlie bone, or the treatment area is large, a case for using surface mould therapy can be made. This will particularly apply when the only alternative treatment is to use superficial X-ray therapy when the absorbed dose to bone will be greater than in muscle. With the advent of remotely controlled afterloading equipment, the potential radiation hazard can be virtually eliminated and the place for using a surface treatment mould needs further consideration. The purpose
394 High dose-rate brachytherapy in the treatment of skin tumors
of this chapter is to discuss the use of afterloaded treatment moulds rather than superficial X-rays or electrons for treating skin cancer, particularly with regard to high dose-rate (HDR) remotely afterloaded surface moulds. However, when indicated and for completeness, reference will be made to other brachytherapy systems. The authors' experience was originally gained with the use of surface mould therapy using sources of radium226, gold-198, cesium-127 or iridium-192 placed on individually made shells or wax blocks. The spatial distribution and quantity of the radioactive sources were determined by the Paterson-Parker Rules [5-7]. The amount of radioactive material depended on the size of the treatment area and the distance between the plane of the sources and the treated surface. For large treatment areas, considerable amounts of source activity were needed, with consequent restriction on the amount of time that could be spent positioning the device on the patient and on nursing and visitors' time spent with the patient. In particular, the technical staff who prepared the mould before application, and dismantled it after use, could be exposed to a significant radiation hazard. The use of a remote afterloading machine offered a possible solution to those difficulties [8], and subsequently other reports of this technique were published [9-11]. A recent literature search has shown that there continues to be an interest in the technique [13-20].
Figure 29.1
Percentage depth dose characteristics of a 40 mm
diameter surface mould (SM) at 10 mm and 30 mm treating distances (TD): curves SM 10 mm TD and SM 30 mm TD. For comparison, superficial X-rays of 2 mm aluminum half value layer (Al HVL) and 8 mm Al HVL at a focus-to-skin distance (FSD) of 150 mm are also shown: curves X-ray 80 kV and X-ray 140 kV. (Data from references [7] and [12].;
29.2 HIGH DOSE-RATE AFTERLOADING SYSTEMS One author (CAJ) described a suitable remotely controlled system in 1969 [9], which was also being developed by the second author (AF), in conjunction with Dr A.J. Ward, in another center (Cookridge Hospital, Leeds) at about this time. Both these treatment systems used an HDR cobalt-60 unit, the Cathetron, previously described [8]. However, the methods used are adaptable to any afterloading machine with suitable small sources and source delivery system. More recently, Selectron-LDR (low dose-rate), microSelectron-HDR, and pulsed doserate (PDR) machines have been adopted for use in afterloading surface moulds [13-15,17-20].
293
ABSORBED DOSE DISTRIBUTION
Due to the short source to skin distance, which is typically between 5 mm and 30 mm, a rapid fall-off in dose with depth from the skin surface of the applicator occurs. Figure 29.1 shows a typical depth dose profile for a source to skin distance of 10 mm and 30 mm, for a 40 mm diameter applicator. The curves shown are for radium-226, but are not substantially different for other isotopes in common use. Also, for comparison, depth
dose data for 150 mm focus-to-skin distance (FSD) superficial X-rays at 80 kV and 140 kV are shown, the data being taken from the British Journal of Radiology, Supplement 26 [12]. In the case of a surface mould, the reduction in dose with depth within the irradiated tissues is principally due to the increase in distance from the radiation source(s), and the relative effect of tissue attenuation and scatter is small. In contrast, absorption and scatter, depending upon the quality of the beam, will be more important for superficial X-rays. The net result is that for water (or muscle tissue) the depth dose characteristics of the two modalities are similar, despite the differences of photon energy and source to skin distance/FSD. However, for tissues of higher atomic number, such as cartilage and bone, the photoelectric effect will dominate in the case of superficial X-ray therapy, but not in the case of a surface mould when using an isotope emitting high-energy photons. When these tissues are included in the treatment volume, there will be a relatively greater absorption of energy for superficial X-rays than for a surface mould, which may lead to a greater absorbed dose than that prescribed. This, in turn, will increase the risk of severe late normal-tissue injury, which the higher energy radiation from a surface mould will largely avoid.
Mould production 395
29,4
SELECTION OF TREATMENT DISTANCE
depth of the tumor. The fall-off in dose from the skin surface is greatly affected by the treatment distance. Figure 29.1 shows that, for a treating distance of 10 mm, using cobalt-60, the percentage depth dose falls to 20% at 30 mm beneath the skin surface. This needs to be taken into consideration when prescribing treatment which, for skin tumors, has historically been prescribed to the skin surface. With the elimination of exposure risks to staff, it may be tempting to improve the dose homogeneity by increasing the treating distance to 20 mm or 30 mm. Care should be taken if this is intended, because the usual method of treatment delivery does not normally provide any form of collimation. While improved collimation can be achieved to some extent by surrounding the treatment area with appropriate lead shielding, this may produce practical problems because of the weight of lead required and the need to provide some sort of supporting structure.
As shown in Figure 29.1, the depth dose characteristics of a surface mould depend strongly on the treatment distance, a greater distance giving a higher relative depth dose. Before any part of the mould device is actually constructed, it is necessary to decide on the appropriate treating distance. This will depend mainly on the thickness of the tumor to be treated and the dose required at the deepest part of the target volume. Whilst it would normally be necessary to construct a full treatment plan for the mould when all of the parameters have been decided, it is useful initially to choose a treatment distance based on the information contained in the Paterson-Parker tables [7]. Although these tables are calculated for radium-226 sources, the depth doses that may be obtained from them are sufficiently accurate for this purpose when using other nuclides such as cobalt60, cesium-137, or iridium-192. The choice of treatment distance is a compromise between conflicting requirements, and usually a distance of 10 mm or 20 mm is chosen. A treatment distance of less than 10 mm may produce unacceptable dose variation on the skin surface if the latter has undulations that cannot be accurately followed by the source applicators. A treating distance greater than 20 mm may lead to long exposure times and to a wide penumbra effect outside the edges of the intended treatment area. This may result in a clinically meaningful dose to critical organs such as the eye. To some extent these effects can be overcome by suitable shielding at the edges of the mould, as illustrated in Figure 29.2 and discussed in references 9-11. An additional consideration is the thickness of the tumor. There is an approximate relationship between the diameter of a skin cancer and its thickness. Lesions that are up to 25 mm diameter are, in general, no more than 5-6 mm thick. This is an important relationship because a cancericidal dose will be necessary at the maximum
The first step in the production of a mould is to take a cast of the affected area. If a flat area is involved, the traditional plaster of Paris can be used. However, for more detail on an undulating surface it may be better to use one of the modern alginate impression materials. A close-fitting, transparent plastic shell is made from the cast, usually by vacuum forming. If the area to be treated is on a limb extremity, such as the dorsum of the hand, the shell can be attached to a baseplate to provide stability; otherwise, straps can be used to attach it to the patient. The area to be treated is marked on the skin and superimposed on the shell (Figure 29.3).
Figure 29.2 Cathetron HDR afterloading surface mould, showing lead shielding around the irradiated area.
Figure 29.3 HDR afterloading surface mould, with the area to be treated marked on the skin.
29.5 29.5*1
MOULD PRODUCTION Providing a cast
396 High dose-rate brachytherapy in the treatment of skin tumors
29.5.2
Disposition of sources
The proposed arrangement of the sources and dwell positions should be determined. Our usual procedure is to use the Paterson-Parker distribution rules as a guide to the arrangement of the source positions around the periphery of the treated area and, when necessary, in a line or series of parallel lines over the area itself. For some afterloading machines, the number and configuration of the source trains available are restricted. These may be used if the source train length matches the required treatment length; otherwise, a single active source can be stepwise positioned along the line of the treatment. With some current afterloading machines, the stepping source arrangement offers greater flexibility, providing for any active length that may be required, and allows the dose distribution to be optimized if required. This can be of particular value when irradiating an irregular surface or where the treatment area is not flat (Figure 29.4). 29.5.3
Applicator supports
Having decided on a provisional source(s) arrangement, the construction of the mould may be completed. On
our early moulds using the Cathetron machine, we used a Perspex frame to support the (then rigid) treatment catheters (Figure 29.5), but later we developed flexible catheters. These are now superseded by using 6-French bronchus-type catheters with the microSelectron, or similar. The catheters are supported by Perspex catheter supports of the appropriate height, attached to the surface of the shell. These supports have a hole drilled in them to accommodate the flexible catheters, as appropriate. They are of the correct height to provide the required treatment distance, allowing for the thickness of the shell material. The supports at the end of each catheter position contain a recess rather than a complete hole. This provides an end-stop to define the longitudinal position of the catheter and to prevent any catheter movement. It also ensures that the catheters are inserted in the correct position for treatment. In some cases, lead shielding is placed around the treated area to help to protect nearby critical organs, as already mentioned and illustrated in Figure 29.2. There are, of course, alternative methods of supporting the applicators, such as embedding them in a wax or acrylic sheet of the appropriate thickness.
29*5*4
Provisional treatment times
A provisional set of treatment times for the various source positions is then drawn up. This should be done by referring to appropriate data on the dose rate(s) and distribution(s) for the available source trains. For the microSelectron-HDR and other similar afterloaders, the calculation can be performed on the brachytherapy treatment planning system. The use of a dwell time optimization program may be considered to improve the dose homogeneity on the treated surface, but in practice we found this difficult to use as the position of the dose calculation points cannot easily be defined, particularly for a curved surface. For a small mould with only a few treatment applicators, we found it just as easy to adjust
Figure 29.4 MicroSelectron HDR afterloading surface mould, showing the source catheters (applicators) in parallel lines over the irradiated area. (Courtesy of Nucletron BV.)
Figure 29.5 Cathetron HDR afterloading mould, showing the catheter support Perspex frame.
Clinical practice 397
the dwell times empirically, in conjunction with dosimetry measurements, as described below.
29.5.5
Dosimetry measurement
The dosimetry should be checked before treatment is started by thermoluminescence dosimetry (TLD) or other suitable dosimetry system. Our own experience is with TLD. A number of thin sachets containing lithium fluoride crystals are taped to the underside of the shell on the treated surface, and the space beneath this is filled with a tissue-equivalent material. The mould is exposed according to the provisional treatment times, and the dose to the treated area is determined. A correction to the dose measured is made to allow for the thickness of the sachet, based on the depth dose characteristics as previously calculated. The treatment times are then adjusted as necessary. The dose distribution is regarded as being satisfactory if the range of doses measured is within ±10% of the mean. This is the range suggested in the Paterson-Parker Rules, which can often be improved on in practice.
29.6
CLINICAL PRACTICE
Most of our experience has been gained from treating patients who were diagnosed with basal cell carcinoma, squamous cell carcinoma, or intraepidermal carcinoma. However, a few cases of less common soft tissue sarcoma have been treated postoperatively, including a case of recurrent malignant melanoma in a young man and a primary malignant melanoma of the pinna in a geriatric patient. A variety of body sites were treated, including the scalp, dorsum of hand, chest, abdominal wall, and lower leg. The main constraint from a technical point of view has been that the treated area should not be so curved as to restrict the movement of the source train through the catheters. This restriction is less critical for current machines which use small stepping iridium-192 sources, such as the microSelectron-HDR; whereas, for machines for which the size of the radioactive source(s) is relatively large, the minimum radius of curvature will be restricted. One other constraint has been the need to restrict the radiation dose received by any critical normal tissue(s) adjacent to the target volume.
29.6.1
Dose fractionation schedules
For soft tissues, the fall-off in dose with distance below the skin surface is similar whether superficial X-rays or cobalt-60 treatment moulds are used. If a similar dose fractionation regime is used for HDR treatment moulds as for superficial X-rays, any difference in tissue effect(s) will be principally dependent on the radiation quality
alone. This is of particular importance where bone or cartilage underlies a tumor, when it is clearly advantageous to use high-quality radiation. In order to achieve a radiobiological effect in muscle similar to that used in time-established superficial X-ray treatment, the dose per fraction for cobalt-60 will need to be increased by about 10%. Compared to muscle, bone will absorb about four to five times more energy per gram from superficial X-rays as compared with cobalt-60, with obvious advantages in favor of cobalt-60. For many treatment situations, small treatment areas less than 3.0 cm in diameter should be restricted to a dose of 45 Gy in ten fractions. When treating areas larger than 3.0 cm diameter, increased fractionation is necessary if the risk of late normal-tissue damage is to be minimized. Among the situations in which careful consideration of the dose fractionation regime used is necessary, is treatment to areas with minimal subcutaneous thickness such as skin overlying the shin bone.
29.6.2
Dorsum of the hand and lower arm
The commonest tumor is an invasive squamous cell carcinoma. These tumors are more often seen in older patients who have already suffered skin changes due to chronic exposure to sunlight. The majority of lesions are flat and do not exceed 20 mm in diameter at presentation. Typical dose regimes are 45 Gy in ten fractions and 50 Gy in 15 fractions. Usually, a source to skin treating distance of 10 mm or 15 mm is suitable. Others have reported using a short-distance cobalt unit to give 55 Gy in 15 fractions over 3 weeks to fields less than 3.0 cm in diameter. They referred to lesions on the dorsum of the hand being radioresistant, with a higher recurrence rate compared to other sites. The skin overlying the treated area may become thin and atrophic. Such changes can be aggravated in a situation where skin damage is already present due to previous chronic exposure to sunlight. Problems may also arise in a patient who, following treatment, is exposed to a risk of traumatic skin injury. We have seen such a case in a sailor who suffered a skin laceration through the treated area 2-3 years following radiation which required plastic surgery. In the older patient with thin atrophic skin, increased fractionation is advisable and the alternative treatment by plastic surgery should be carefully considered. Occasionally, multiple lesions may be unsuitable for surgery or small-field radiotherapy. A technique using a large treatment field to deliver 60 Gy in 2.0-Gy fractions to an area covering more than half the circumference of the forearm has been described by Rudoltz and others [20]. They used a thermoplastic mould 5.0 mm thick with 22 silicone-rubber catheters longitudinally placed 20 mm apart. The given dose was prescribed to 8 mm depth in tissue, treatment being delivered using source
398 High dose-rate brachytherapy in the treatment of skin tumors
dwell positions at 1-cm intervals and the mould covered with a lead shield. The results were reported as satisfactory, the disadvantage being the extended treatment time and concern about radiation exposure to staff. This latter problem emphasizes the need to deliver treatment within a radiation-protected room, the purpose of lead shielding being to reduce whole-body irradiation to the patient.
29.63
Face and scalp
When treating sites in this area, it is important to pay particular attention to appropriate shielding of surrounding normal tissues. Treatment distances should be short in order to reduce the exposure time of the radioactive sources. If shielding cannot be easily provided, it is preferable to consider using electron-beam or photon-beam therapy. The areas most applicable to surface mould therapy are the forehead and temporal areas. Where areas are situated close to the hairline and because of the wide penumbra, the risk of alopecia is high and appropriate shielding should be used where possible.
29*6.4
Pinna
Carcinoma of the skin of the pinna will overlie cartilage and the benefits of high-energy photon radiation will particularly apply for the reasons discussed earlier. HDR afterloading using cesium-137, cobalt-60, or iridium192 is especially suitable. Treatment can be given using a single-plane applicator at a treatment distance of 10-20 mm (although the authors have found 15 mm to be the most practical distance) to a prescribed dose of 45-50 Gy in eight to ten fractions. The pinna is amenable to the protection of adjacent structures, which is important if hair loss is to be minimized. However, the use of radiation shielding of adjacent tissues may involve the practical problems of physically supporting a lead shielding block [10]. For thick lesions of the helix, a double-plane applicator can be used and planned according to the Parker-Paterson Rules, but the practicalities of protecting surrounding tissues may prove difficult.
29.6.5
Legs
In general, skin healing following radiation, particularly overlying the anterior tibia, is poor. Of 20 cases treated with 45-47.5 Gy in 10 or 11 fractions, poor healing affected three cases, with superficial necrosis affecting three other cases (one following injury). Considerable care is therefore necessary when considering dose fractionation regimes for this site, and increased fractionation is advisable. A review by Podd [21] reported an overall recurrence
rate of 4.6% and a radionecrosis rate of 9.2% when treating squamous and basal cell cancers of the lower limb in older patients. The treatment given was to areas less than 30 cm2, which corresponded to a 6 cm diameter applicator. Although either superficial or orthovoltage X-rays to a dose of 40 Gy in ten fractions were used, allowing for the difference in radiation quality, these reported findings were similar to our own series. The treatment of a large area carries an increased risk of failure to obtain complete skin healing. However, in a case report of LDR treatment in an elderly patient, a treatment area of 8 x 8 cm, extending over half the circumference of the leg, was considered more acceptable than either a single external-beam field or a parallel opposed pair of fields, particularly as the latter would have treated more than half the leg thickness [12]. We support reserving treatment of lesions of the lower limb using a treatment mould to situations for which alternative treatments have been carefully assessed and eliminated.
29.6.6
Trunk
The skin of the lower abdominal wall is generally more sensitive than that of the upper trunk to the effects of radiation. However, because of the rapid fall-off in dose beneath the skin surface, surface moulds can be extremely useful for confining the effects of radiation to within the abdominal wall tissues. For the treatment of skin tumors or secondary skin nodules, this form of therapy can be useful where the treatment of a relatively large surface area is indicated. The technique employed is similar to that for radium loaded chest wall moulds as used historically. Where large areas are to be treated, increased fractionation is necessary and, in general, this form of treatment application has been superseded by electron therapy. However, optimized treatment can provide better dose homogeneity over curved surfaces [10] than from abutted electron fields [13]. When using spatially positioned sources, by altering the position of the source pencils and adjusting treatment times for different source positions, it is possible to treat surface areas up to 200 cm2 without loss of uniformity of dose [11]. Others have since described the use of an HDR (iridium-192, 370 GBq) remotely loaded applicator for treating Kaposi sarcoma lesions. This entailed using a custom-built, ceiling-mounted immobilization device that secures the applicator on the surface of the patient. The applicators were made of tungsten/steel, 1, 2, or 3 cm diameter. The treatment distance was 15 mm and treatment sites included the head and neck and extremities, as well as the torso. The applicators required a plastic cap to eliminate electron contamination. For treating Kaposi sarcoma lesions, an optimal surface dose of 10-15 Gy in a single fraction, depending on the thickness of the lesion, was recommended [19].
References 399
29.7
CONCLUSION
11. Joslin, C.A.F. (1972) Afterloading methods in radiotherapy. In Recent Advances in Cancer and Radiotherapeutics, ed. K. Hainan. Edinburgh, Churchill
The majority of skin tumors are amenable to treatment by conventional means, including superficial X-rays, electron therapy, and surgery. This chapter has reviewed the literature and, coupled with our own experience, has identified and discussed the use of afterloaded treatment moulds in situations in which they offer potential treatment advantages to the patient.
for use in radiotherapy. Br.J. Radiol., 26. 13. Kitchen, G., Dalton, A.E., Evans, M., Pope, B. and Smith,
REFERENCES
14. Kitchen, G., Dalton, A.E., Pope, B.P., Smith, P.O. and Powner, M. (1991) Surface applicator for basal cell carcinoma of the right pinna: a case report. Activity [Selectron Activity Journal], 5,140.
1. Gloster, H.M.and Brodland, D.G. (1996) The epidemiology of skin cancer. Dermatol. Surg., 22(3), 217-26.
Livingstone. 12. Supplement 26 BJR. (1996) Central axis depth dose data
P.O. (1990) Selectron-LDR mould for large area basal cell carcinoma; a case report. Activity [Selectron Activity Journal], 4,72.
15. Perrozzo, M., Stabile, L, Ross, R., Moorthy, C, Tchelebi, A. and Hilaris, B.S. (1992) HDR remote afterloading as an alternative to electrons for therapy of superficial
2. Strom, S.S. and Yamamura, Y. (1997) Epidemiology of nonmelanoma skin cancer. Clin. Plast. Surg., 24(4), 627-36.
tumours. Activity [Selectron ActivityJournaf], 6,11. 16.
3. Wei ,Q. (1998) Effect of ageing on DMA repair and skin
superfractionated skin irradiations using large
carcinogenesis: a mini review of population based
afterloading moulds. lnt.J. Radial Oncol. Biol. Phys., 36(1)147-57.
studies.J Investig. Dermatol. Symp. Proc., 3(1), 19-22. 4. Halpern, J.N. (1997) Radiation therapy in skin cancer. A historical perspective and current applications. Dermatol. Surg., 23(11), 1089-93. 5. Paterson, R.P. and Parker, H.M. (1934) Dosage system for gamma ray therapy. Br.J Radiol., 7, 592. 6. Paterson Rand Parker H M. (1938) A dosage system for
17. Leung, J.T. (1997) Extensive basal cell carcinoma treated with the mould radiotherapy technique. Australas. /tad/o/.,41,20-1. 18. Svoboda,V.H.J., KovarikJ.and Morris, R(1995) High dose-rate microSelectron moulds in the treatment of skin tumors./ Radial Oncol. Biol. Phys., 31,967-72.
interstitial radium therapy. Br.J. Radiol., 9,252 and 313. 7. Meredith, W.J. (ed.) (1967) Radium Dosage. The Manchester
19. Evans, M.D., Yossa, M., Podgorask, E.B., Roman, T.N., Schreiner, L.J.andSouhami, L. (1997) Surface applicators
System. ES Livingstone, Edinburgh and London. 8. O'Connell, D., Howard, N., Joslin, CAR, Ramsey, N.W. and
for high dose rate brachytherapy in A.I.D.S related Kaposi sarcoma. lnt.J. Radial Oncol. Biol. Phys., 39(3), 769-74.
Liversage, W.E. (1965) A new remotely controlled unit for the treatment of uterine cancer. Lancet, 18, 570-1. 9. Joslin, C.A.F., Liversage, W.E. and Ramsey, N.W. (1969)
Fritz, P., Hensley, F.W., Berns, C., Schraube, P. and Wannenmacher, M.(1996) First experiences with
20.
Rudoltz, M.S., Perkins, R.S., Luthmann, R.W. et al. (1998) High-dose rate brachytherapy with custom surface mold
High dose-rate treatment moulds by afterloading
to treat recurrent squamous cell carcinoma of the skin of
techniques. Br.J. Radiol., 42,108-12.
the forearm./ Am. Acad. Dermatol., 38,1003-5.
10. Joslin, CAR and Smith, C.W. (1970) The use of high
21. Podd, T.J. (1992) Treatment of lower limb basal cell and
activity cobalt 60 sources for intracavitary and surface
squamous cell carcinomas with radiotherapy. Clin. Oncol.,
mould therapy. Proc. R. Soc. Med., 63(10), 1029-34.
4,44-5.
30 Hyperthermia and brachytherapy PETER M. CORRY, ELWOOD P. ARMOUR, DAVID B. GERSTEN, MICHAEL J. BORRELLI, AND ALVARO MARTINEZ
30.1
INTRODUCTION
Brachytherapy provides an obvious and sometimes ideal setting for combining hyperthermia with radiation therapy. Such combination therapy has been carried out in the past primarily using microwave technology, multiple antennae being placed intratumorally through plastic catheters previously inserted for this purpose [1,2]. There are a number of situations, particularly in the head and neck region, where this methodology is useful, but it does not lend itself easily to automation and does not adapt well to the simultaneous application of hyperthermia or to the use of radioactive source afterloaders. Another situation in which hyperthermia can and has been applied [3-5] is for implants that involve stainlesssteel needles to contain the radioactive materials. This approach is usually done in conjunction with a cutaneously attached template guidance apparatus [6]. In this case, radio-frequency (RF) power is applied directly to the stainless-steel needles themselves. Other approaches include needles heated with resistive electrical heating elements or hot water [7,8] ferromagnetic seeds contained within plastic catheters [9], and RFdriven, capacitively coupled, plastic-coated catheters [10]. Ultrasonic interstitial and intracavitary applicators promise more functionality and versatility, but are not yet in widespread clinical use [11]. From a clinical historical perspective, there have been
several randomized, prospective, clinical trials carried out in Europe over the past few years [12] which have demonstrated benefit associated with the addition of hyperthermia to conventional radiation therapy. The only similar clinical trial combining hyperthermia with brachytherapy, conducted by the Radiation Therapy Oncology Group (RTOG) in the USA, showed no such benefit [13]. Unfortunately, this trial, which was fraught with several quality assurance issues, was designed and mostly conducted prior to the development of the quality assurance guidelines which are now known to be essential to a positive result [14]. Only one patient of 86 in this study received what was considered an 'adequate' hyperthermia session. 30.2
BIOLOGICAL FACTORS
Over the past 20 years, the biological effects of elevated temperatures (40-43 °C) and their implications in cancer therapy have been extensively investigated. The following points summarize these investigations and constitute the rationale for the application of elevated temperatures in this setting. • Hyperthermia kills tumor cells directly and preferentially kills cells in macroscopic tumors. • If sufficient exposure is given, hyperthermia supraadditively sensitizes cells to the action of ionizing
Biological factors 401
radiation with a thermal enhancement ratio (TER) up to 3, and chemotherapeutic drugs with a TER up to 10. • Hyperthermia eliminates dose-rate effects by the inhibition of repair phenomena (Figure 30.1) [15]. This is particularly important in application with brachytherapy because the biological effects of low dose-rate (LDR) radiation and high dose-rate radiation (HDR) should be approximately the same in heated tumor tissues. In cooler normal tissues, however, LDR should have an advantage in terms of lesser toxicities. • Recently, other investigators have shown that temperatures in the 41-42°C range, when applied to solid tumors in rodents, can have a profound effect on tumor oxygenation and/or reoxygenation [16,17]. This reoxygenation greatly reduces the hypoxic fraction, significantly enhancing the effects of ionizing radiation in addition to the direct hyperthermic effects. Similar observations have been made in human tumors [18]. • Hyperthermia acts in a complementary manner to ionizing radiation and chemotherapeutic drugs. Those cells most resistant to radiation or drugs are the most sensitive to hyperthermic killing and sensitization. For many years, the existence of a phenomenon known as thermal tolerance was thought to contraindicate closely spaced heat fractions and the administration of chronic protracted heating. In fact, several publica-
Figure 30.1 Dose-rate effects in rat 9L gliosarcoma cells in culture. (Data adapted from reference 15.)
tions (e.g., [19-21]) support the contention that few, perhaps only 10%, of a tumor's cells are killed by hyperthermia and that number can be expected to vary by at least an order of magnitude in any clinical setting. These reports also support the notion that, while thermal tolerance develops, it is not a clinically limiting factor. Furthermore, hyperthermia is rarely, if ever, used as single-agent cancer therapy. Recent randomized clinical trials using the combination of hyperthermia and conventional fractionated radiation therapy have shown positive benefit to the addition of hyperthermia (e.g. [12]) demonstrate T90 values between 39°C and 41°C: temperatures far too low to accomplish any significant direct cytotoxicity. What has been shown by authors quoted above, as well as others, is that such temperatures for protracted periods of time continue to affect thermal radiosensitization (TERs of 1.6-2.0) and that such sensitization is essentially independent of the development of chronic thermal tolerance. Whereas the intrinsic thermal sensitivity of human cells varies over several decades of survival from cell line to cell line at these temperatures, thermal radiosensitization seems to be affected to only a minor degree, if at all. In addition to thermal radiosensitization, thermal intratumoral reoxygenation and hypoxic cell fraction reduction are probably playing an important role as well. Irrespective of the clinical heating methodology, acute pain associated with power application has been reported as the primary limiting factor in achieving temperature distributions with T90 greater than 41 °C for protracted periods [22-25]. This pain is often associated with elevated temperatures themselves, as well as direct power deposition within the involved tissues. This has been found to be the case for tumors with nerves encompassed by the tumor and is a particularly significant factor for advanced malignancies in the pelvis and abdomen. Fortunately, several publications (e.g., [15, 26-28]) have demonstrated that 41 °C for long durations can yield TERs between 1.5 and 2 (Figure 30.2) as well as the elimination of dose-rate effects. This is more often than not the maximum temperature achievable clinically. All but one of the clinical brachytherapy studies to date applied hyperthermia for 1 h at a target temperature of 43 ° C before beginning the LDR radiation, and for 1 h after completion of the radiation therapy course, usually 48-72 h. This was not done because those regimens have been demonstrated to be optimal, but was dictated solely by the practical limitations of the systems used for tumor temperature elevation. As may be easily seen on consideration of the data in Figure 30.3 little added benefit in terms of antitumor effect should be expected from this regimen. Consideration of the above data and observations yield several conclusions. The negative findings, relative to the addition of hyperthermia, of the RTOG randomized clinical trial came about because of two factors:
402 Hyperthermia and brachytherapy
Figure 30.2 Enhancement of LDR in rat 9L cells, at a dose rate of 50 cGy h~1, by continuous simultaneous heating at41°C. The acute X-ray dose was delivered at approximately 3 Gy min~\ (Data adapted from reference 27.)
(1) extensive thermometry was not employed and the actual thermal distributions were not known to an adequate degree of certainty, and (2) hyperthermia was administered only for 1 h before and after the interstitial radiation. Sparse thermometry has previously been shown to yield a false sense of intratumoral heating adequacy [22]. If the data of Figure 30.3 are representative of what is occurring intratumorally, little or no supraadditive enhancement of radiation cytotoxicity should be expected. Other important observations relative to thermal reoxygenation are: (1) at least a 1-h exposure at 41-42°C is required to optimally reduce the hypoxic fraction; (2) reoxygenation effects disappear between 12 and 24 h after the hyperthermic exposure; and (3) at higher temperatures, 43-44°C, reoxygenation is not observed due to intravascular coagulation. This last observation (intravascular coagulation) might seem to suggest that lower temperatures are better because more reoxygenation is observed. It must be remembered, however, that for each degree increase in temperature above 42 ° C, direct heat killing as well as thermal radiosensitization to radiation increase by approximately a factor of 2. One scenario that would explain clinical results, where broad temperature distributions from 39°C to 44 °C are the rule rather than the exception, is that in those portions of a tumor that are poorly heated (39-42 °C), thermal radiosensitization by reoxygenation predominates.
Figure 30.3 Effects of 1-h 43 °C heat exposures both before, after and before, and after LDR irradiation in rat 9L gliosarcoma cells in culture at a dose rate of 50 cGy h-1 There is no significant difference in the slopes of any of these plots. The thermal enhancement ratio for all three hyperthermia regimens is slightly less than 1.0, suggesting that there was no supraadditive interaction between the heat and radiation for these regimens.
In those portions of the tumor that are heated well (41-44°C), thermal radiosensitization by repair inhibition predominates. In the 41-42°C range, both effects are operative. This is a fortuitous situation because tumors have proven to be remarkably refractory to uniform heating, even with interstitial and intracavitary technology. Taken together, these factors lead to the conclusion that, for optimal antitumor effect, hyperthermia should be administered continuously and simultaneously during LDR irradiation. The use of pulse dose-rate (PDR) and HDR irradiation technology requires some further consideration. For PDR, the hyperthermia requirements are essentially identical [28]; however, because the radiation for PDR is administered in short pulses, perhaps 1 min h'1, it is not essential that hyperthermia be administered during the actual irradiation period. This factor may be of considerable practical importance. In the case of HDR procedures, in which radiation fractions are spaced over days or weeks, hyperthermia should be administered for at least a 1-3-h period immediately prior to irradiation and, if practical, during irradiation [29].
A system for simultaneous hyperthermia and brachytherapy 403
303
THERMOMETRY REQUIREMENTS
Quality assurance criteria and the need to control temperatures in three dimensions at some prescribed level dictate thermometry requirements. The prescription should be expressed in terms of the temperature distribution throughout the treatment volume and not in terms of a given temperature at any given point within the volume. A convenient way of describing the distribution is the percentage of intratumoral temperature points at or above a given index temperature [22,30]. The software operating the system must have the capability of computing and displaying this type of information in real time, providing the operator with the ability to assess compliance with the prescription continuously. This type of real-time analysis also permits alteration of the power deposition pattern to comply with the prescription if necessary. The RTOG in the USA has developed comprehensive quality assurance guidelines for interstitial hyperthermia which are sufficiently extensive to permit this type of analysis and control [14]. Table 30.1 summarizes the approximate recommended number of sensors as a function of tumor volume. Table 30.1 Recommended implanted sensors
5 10 50 100 200 500
12 15 18 24 30 48
These publications also outline the guidelines for placement of the sensors. Catheters containing multiple sensors or within which a single sensor will be scanned along its length must be placed to represent accurately the central and peripheral aspects of the tumor as well as sensing temperatures in tumor tissue central to the arrays of heating entities. It must be stressed again that these figures are recommended minima for quality assurance purposes and for many systems additional sensors will be necessary to achieve adequate threedimensional power deposition control. Clearly, systems of the future will require expanded data acquisition capabilities over those of the past. A 32-point measurement capability appears to be the minimum required number and, for versatility, 64 is highly desirable.
30.4
SYSTEMS CONSIDERATIONS
The primary considerations in the design of systems for the administration of hyperthermia and brachytherapy are based on the biological factors and thermometry
requirements discussed above, as well as on practical considerations. The following criteria are recommended for an ideal system: • It should be possible to deliver hyperthermia simultaneously prior to and throughout the course of radiation, irrespective of the brachytherapy modality in use - LDR, PDR, or HDR. • Compatibility with LDR after-loaders as well as PDR and HDR source loaders is required for optimal flexibility in source handling and to reduce exposure to personnel. • The thermometry system should have at least 32 temperature-measuring channels. Preferably, 64 should be available. The design should permit temperature measurements in tissue between heating elements if sensors are built into the heating elements themselves. • The system should permit dynamic alteration of the power deposition pattern in three dimensions throughout the course of treatment. • Power deposition control and data acquisition must be fully automated. Because prolonged exposures of hours to days may be required, remote monitoring and control are highly desirable. • For practical reasons, set-up complexity should be minimized. • For optimum utility, the system should be portable or at least easily movable. In practice, there are no systems that are available commercially or prototype systems that have been developed in research laboratories that can satisfy all of the above criteria. It is also highly unlikely that any one system will be appropriate for all anatomic locations. For example, while local current flow (LCF) technology, such as that described below, works well for tumors in the deep pelvis with transperineal template guidance, it adapts poorly to tumors in the head and neck. Ultrasound and microwave systems permit wider element spacing, and as a result fewer elements, than other technologies and provide superior three-dimensional control, but do not adapt well to simultaneous administration with radiation. Capacitively driven systems are very flexible and adaptable to irregular shapes and to tumors in the head and neck, but power deposition control is more difficult and simultaneous administration is precluded for most systems. In all cases, trade-offs have been required. As a result, there is no ideal system or method available for hyperthermia administration at present.
30.5 A SYSTEM FOR SIMULTANEOUS HYPERTHERMIA AND BRACHYTHERAPY To test the hypothesis that simultaneous administration of hyperthermia throughout the course of
404 Hyperthermia and brachytherapy
brachytherapy administration should yield superior results to protocols applying heat only before and after radiation, a new system was designed. This first step was the Martinez Universal Perineal Implant Template (MUPIT) [6], used to guide implants to incorporate printed circuit boards as integral components to effect connections to the needles. This hyperthermic universal perineal implant template (HUPIT) is shown in position in Figure 30.4 after the operative procedure for a patient under treatment. There are 59 positions, at 11 mm separations, for stainless-steel needles in a seven wide by nine high array, with one needle missing at each corner. Each needle connects both to the afterloader to insert/retract the radioactive isotope, as well as to the power-generating circuitry to induce hyperthermia. There are 48 positions centrally located between these needles for dedicated thermometry catheters. Connection to the power generator is via two miniature connectors, one at each end of the template (29 wires each). The microprocessorcontrolled power generator has the capability of applying the RF power to any, all, or none of the needles
simultaneously, and each needle can be connected at an RF phase angle or either zero or 180 degrees. By varying the pattern and phase angles of the connections, any desired power deposition pattern can be achieved, at least in two dimensions. As areas of the tumor warm to the set point, the duty factor for needles in the immediate proximity to the sensors is varied to maintain the desired temperature. The inherent flexibility of this approach minimizes the information necessary prior to doing the implant. Treatment planning software is used after the implant is done to determine the initial power deposition patterns, however. All parameters are under operator control at all times during therapy and can be varied to account for factors such as blood flow and patient tolerance which cannot be accurately determined in advance. This system satisfies all of the design criteria outlined above except one. Because the stainless-steel needles are not segmented, power is controlled only in two dimensions rather than the more ideal situation in which it would be controlled in three dimensions. To have control in three dimensions it would be necessary to develop needles that are segmented along their length. This system was used to treat a variety of tumors in 19 patients to test the feasibility of the approach. The goal was to achieve an intratumoral temperature distribution with a T90 of 41 °C and to maintain that distribution throughout the course of treatment (continuous mild hyperthermia, CMH). Prior to finishing the development of the technology necessary for CMH, 14 patients were treated with acute fractionated hyperthermia (AFH), which consisted of hyperthermia for 1 h before and 1 h after the interstitial irradiation. Clinical response data for the two groups of patients were compared. Figure 30.5 shows the therapy set-up for the patient in Figure 30.4. The distribution of anatomic sites for the tumor in all 34 patients is given in Table 30.2. Figure 30.6 shows a composite integral distribution of intratumoral and normal tissue temperatures for all 19 patients undergoing CMH. A similar distribution (not shown) for patients who underwent AFH did not differ significantly from that of those who underwent CMH. For comparison, the results of a prior study, with a system
Table 30.2 Patient characteristics
Figure 30.4 Electronic template shown fixed in position after the operative procedure was completed on a patient being treated for recurrent cancer of the uterine cervix. The stainlesssteel needles (N) connect to remote afterloaders to insert and retract the radioactive sources. These needles are also connected to the hyperthermia power source by spring-loaded collars (not shown) that in turn connect to the 37 pin connectors (C), mounted at each end of the template. In this instance, six dedicated thermometry catheters (T) were implanted and a total of 30 thermocouple sensors were inserted. V = vaginal cylinder.
Endometrium Cervix Colorectal Prostate Urethra Vagina Anus Breast Total
7 4 5 2 1 2 1 1 23
7 1 1 2 11
Females, 28; males, 6. Mean age, 64 years; age range, 34-88 years.
A system for simultaneous hyperthermia and brachytherapy 405
Figure 30.5 Thermobrachytherapy set-up in the brachytherapy treatment room for the patient shown in Figure 30.4. The box with the connectors (copper-constantan thermocouples) at the upper left is the thermometry system, with 32 individual channels. Located immediately below the thermometry system is the treatment control computer, which is connected to the institution-wide computer network for control and monitoring purposes. The two large tubes at the lower left are the umbilici for two afterloaders, each of which can contain 15 iridium-192
that did not permit power control at the single-needle level [22] are shown. The error bars in Figure 30.6 represent one standard deviation of the 19 points for each index temperature. Normal tissue temperatures within 1-2 cm of the tumor periphery averaged 1.5-2.5°C below those in the tumor itself. The results of clinical follow-up, which varied in duration from 6 months to 3 years, are summarized in Tables 30.3 and 30.4, and the observed toxicities for all 33 patients are summarized in Table 30.5. The definitions for complete response (CR), partial response (PR), and no response (NR) are those that are conventionally used in cancer treatment. The radiation dose for the 33 patients ranged from 15 Gy to 39 Gy and was dictated by prior radiation exposure of the treated region. There was no significant difference between the dose range for AFH and for CMH. The number of implanted needles ranged from 14 to 30, the maximum number being limited by the aggregate number of available sources provided by two afterloading systems (30). The number of thermometry sensors inserted into the implant varied from 18 to 32 and was limited by both the number of catheters implanted (three to six) and the number of channels available (32).
ribbon sources. Smaller tubes (electrically non-conducting) extend from the ends of the umbilici to the template itself. The
Table 30.3 Overall response rates
patient remains in this configuration for 48-72 h. On demand, patient-controlled analgesia is used to control discomfort and pain.
CR PR NR CR+PR
14/19(74%) 4/19(21%) 1/19(5%) 18/19(95%)
7/14(50%) 1/14(7%) 6/14(43%) 8/14(57%)
AFH, acute fractionated hyperthermia; CMH, continuous mild hyperthermia; CR, complete response; PR, partial response; NR, no response. Table 30.4 Responses by tumor volume
CR PR NR
88(7) 120(1) 254(6)
36-100 120 90-432
143(14) 138(4) 122(1)
8-450 45-288 122
* Tumor volume was computed as the product of the three measured dimensions of the tumor multiplied by Pi/6. For abbreviations, see Table 30.3. Table 30.5 Toxicities
Figure 30.6 Temperature distributions.
Pain during treatment Pain after treatment Drainage Perineal reaction Fistula
6 (22) 3(11) 5(18.5) 4 (14.8) 2(7.4)
406 Hyperthermia and brachytherapy
30.6
DIRECTIONS FOR THE FUTURE
Over the past few years, LDR technology has been, for the most part, replaced with HDR systems. These systems provide numerous benefits, which are outlined in other chapters of this text and will not be elaborated upon here. In practice, the adaptation of HDR to application with hyperthermia is simpler than that for LDR technology. Because the dose is delivered in short pulses (a few minutes), the need for protracted hyperthermia application is lessened. The heat dose can be delivered prior to and after irradiation without the absolute requirement for simultaneous delivery. Eliminating hyperthermia for the short interval during irradiation should have inconsequential biological effect. We have adapted the LDR system described above to HDR applications, but insufficient clinical data and follow-up are available to evaluate this approach objectively. The hyperthermia/brachytherapy setting also provides an ideal scenario for the testing and application of gene therapy in cancer treatment. Because the procedures are already invasive, access to deliver the interventional product is guaranteed. To assess the potential efficacy of such an approach, a series of experiments was initiated to mimic the use of the truncated human heat shock promoter (designated AHSP) controlling a reporter gene, the green fluorescent protein (enhanced green fluorescent protein, EGFP), when delivered by a viral vector, Adenovirus 5 (Adv), in this setting. The following experiments were carried out in collaboration with Dr Mark Dewhirst and Dr Rod Braun of Duke University, Durham, North Carolina. Figure 30.7 shows the results of two experiments varying the effective heat dose (30, 60, 90 minutes at 42 °C, assayed 21-23 h later) and the time after heating (0-42 h). These data clearly show a relationship between heat dose and expression as well as time after exposure. Quantitation of gene expression from the digitized video frames for the Adv-AHSP-EGFP samples is expressed as the ratio of the fluorescent signal in the sample to the control after subtracting out baseline autofluorescence. For the dose-response data (panels E-H: A, 0, 30, 60, 90min at 42°C), the values were 3, 28, 102, and 177 respectively. For panels C and D, the values were 165 and 204 for 21 and 42 h post-heating, respectively. Other experimental data show that, ideally, heat shock should be delivered 12-24 h after viral administration, that the heat-induced expression of recombinant products tapers off over the following 24 h, and that it may be reexpressed by heat pulses spaced at 24 h. These parameters adapt admirably to a 3-4-day course of HDR with four to six fractions of administered radiation. These results demonstrated that gene expression could be controlled over a significant range of expression and heat doses in a predictable manner. Also of significant importance was the fact that, in the absence of heat shock, gene
product from the virally introduced recombinant DNA was undetectable. There are few other systems available which permit conformal control of foreign recombinant DNA expression in such a predictable manner.
30.7
CONCLUSIONS
Whereas hard and fast conclusions are difficult with the limited number of patients described above, we believe that the trends in the data are encouraging. The response data for AFH in Table 30.3 show a pattern that is consistent with historical controls for brachytherapy alone. The response rates for CMH showed an improvement from a CR rate of 50% for AFH to 74% for CMH. The overall objective response rate improved from 57% to 95%. Striking differences are also seen when the data are stratified according to tumor volume for each response category. For AFH, there is a clear dependence on tumor volume, a characteristic of all studies that involve radiation as single-agent therapy. For CMH, however, there was no such pattern. The CR category actually was characterized with a wider range of tumor volume than the other two response categories. Statistically, no significant differences could be determined for the ranges of tumor volume in the various response categories; however, this could easily be a result of the small numbers involved. Nevertheless, the elimination of the tumor volume effect requires a TER in the neighborhood of three. While the numbers are small and the assumptions used to calculate the required TER can certainly be challenged, the possibility of obtaining radiation dose modification of this magnitude encourages enthusiasm. Observed toxicities were acceptable and did not vary significantly from what would be expected for brachytherapy alone, with the exception of treatment-related pain. In one case, pain after treatment was protracted for 3 months but subsequently resolved without intervention. To some degree, intra-treatment pain was noted for all patients and in six cases (22%) this pain limited the power that could be applied and required aggressive pain management. The temperature distribution data shown in Figure 30.6 show an improvement in the direction of an ideal distribution, uniform heat dose in the tumor and none outside it. When compared to a previous study in which dynamic control of the power deposition pattern at the level of a single needle was not available [22], more of the tumor is adequately heated and less is overheated. Unfortunately, parts of the tumor remain below the target temperature of 41 °C (10-20%) and a few points as high as 46 °C were observed. These temperature distributions probably will not improve until three-dimensional control of the power distribution pattern can be achieved. This shortcoming points out the need for further instrumentation and systems development.
Conclusions 407
Figure 30.7 Epifluorescence photomicrographs of rat MAC tumors grown in a window chamber and topically infected with Adv-AHSP-EGFP viral preparations. Panels A-D are for the same tumor and track EGFP expression and fluorescence as a function of time after heating for 90 min at 42 °C. Panels E-F are for three different tumors (rats) for heat exposures of 0, 30, 60, and 90 min at 42°C. (A) Immediately prior to infection; (B) 29 h post-infection and immediately prior to heating at 42°Cfor 90 min; (C) 21 h post-heating; and (D) 42 h post-heating; (E) 28 h post-infection, immediately prior to heating at 42 ° Cfor 30 min; (F) as in (E) but 23 h postheating; (G) 23 h post-heating at 42°Cfor 60 min; and (H) 27 h post-heating at 42°Cfor 90 min. The scale in the lower left portion of panel A shows a 1-mm scale photographed through the microscope. All photographs are at the same magnification and aspect ratios are preserved. In panel D, very intense fluorescence in the top portion saturated the video camera, causing underexposure in the remainder of the photograph. All photomicrographs were taken at identical fluorescence excitation levels and magnification.
As stated in the introduction, brachytherapy often provides an ideal setting for combination with hyperthermia. The invasive nature of brachytherapy itself reduces objections to the requirement for invasive thermometry. In the system described above, the needles required for introduction of the radioisotope are also used for hyperthermia, without the need for the introduction of additional invasive elements, other than for thermometry. The ability to administer hyperthermia simultaneously optimizes intratumoral sensitization to radiation both for direct hyperthermic sensiti/ation and from reoxygenation, both of which are required to optimize the antitumor efficacy. While the system described
above is effective for our purposes, particularly for transperineal templates, it is poorly adaptable to other anatomic areas, such as head and neck, where other systems and principles must be employed. Although the results described here are encouraging, multiinstitutional randomized trials with stringent quality assurance criteria and strict patient selection criteria are essential to accurately assess efficacy and benefit to the patient. Finally, this combined modality may provide an ideal setting for testing new modalities such as heatactivated, controllable gene therapy using bacterial proteotoxins and cytokines as well as hypoxic cell radiation enhancers such as the inducible nitric oxide synthases.
408 Hyperthermia and brachytherapy
ACKNOWLEDGMENT This work was supported in part by grant CA-44550 from the US National Cancer Institute.
REFERENCES 1. Perez, C.A. and Emami, B.A. (1985) A review of current clinical experience with irradiation and hyperthermia. Endocuriether./Hypertherm. Oncol., 1,257-77. 2. Couglin, C.T. and Strohbehn, J.W. (1989) Interstitial thermoradiotherapy. Radiol. Clin. North Am.,27(3), 577-88. 3. Kong, J.S., Corry, P.M. and Saul, P.B. (1986) Hyperthermia in the treatment of gynecologic cancers. In Diagnosis and Treatment Strategies for Gynecologic Cancers, ed. F. Rutledge. Austin, Texas, The University of Texas Press, 83-92. 4. Vora, N., Forrel, B., Cappil, J., Lipsett, J. and Archambeau, J.O. (1987) Interstitial implant with interstitial hyperthermia. Cancer, 50,2518-23. 5. Prionas, S.D., Kapp, D.S., Goffinet, D.R., Ben-Yosef, R., Fessenden, P. and Bagshaw, M.A. (1994)Thermometry of interstitial hyperthermia given as an adjuvant to brachytherapy for the treatment of carcinoma of the prostate. Int.). Radial. Oncol. Biol. Phys., 28(1), 151-62. 6. Martinez, A., Cox, R.S., Edmundson, G. et al. (1984) A multiple site perineal applicator (MUPIT) for treatment of prostatic, anorectal and gynecological malignancies. Int. J. Radial Oncol. Biol. Phys., 10,297-305. 7. Kapp, K.S., Kapp, D.S., Stuecklschweiger, G., Berger, A. and Geyer, E. (1994) Interstitial hyperthermia and high dose rate brachytherapy in the treatment of anal cancer: a phase I/I I study. Int.J. Radial. Oncol. Biol. Phys., 28(1), 189-99. 8. Moran, C.J., Marchovsky, J.A., Wipold, F.J., DeFord, J.A. and Fearnot, N.E. (1995) Conductive interstitial hyperthermia in the treatment of intracranial metastatic disease.). Neurooncol., 26, 53-63. 9. Stea, B., Rossman, K., Kittleson, J., Shelter, A., Hamilton, A. and Cassady, J.R. (1994) Interstitial irradiation versus interstitial thermoradiotherapy for supratentorial malignant gliomas. Int.J. Radial Oncol. Biol. Phys., 30(3), 591-600. 10. Kaatee, R.S.J.P., Kampmeijer, A.G., van Hooije, C.M.C. et al. (1995) A 27 MHz current source interstitial hyperthermia system for small animals. Int.J. Hypertherm., 11(6), 785-96. 11. Diederich, C.J. (1996) Ultrasound applicators with integrated catheter-cooling for interstitial hyperthermia: theory and preliminary experiments. Int.J. Hypertherm., 12(2), 279-97. 12. Overgaard, J., Gonzalez Gonzalez, D., Hulshof, M.C. et al. (1995) Randomised trial of hyperthermia as adjuvant to
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
radiotherapy for recurrent or metastatic malignant melanoma, Lancet, 345, 540-3. Emami, B., Scott, C, Perez, C.A. et al. (1996) Phase III study of interstitial thermoradiotherapy compared with interstitial radiotherapy alone in the treatment of recurrent or persistent human tumors. Int.J. Radial Oncol. Biol. Phys., 34,1097-104. Emami, B., Stauffer, P., Dewhirst, M.W. et al. (1991) RTOG Quality Assurance Guidelines for Interstitial Hyperthermia. InlJ. Radial Oncol. Biol. Phys., 20, 1117-24. Wang, Z., Armour, E.P., Corry, P.M. and Martinez, A. (1992) Elimination of dose rate effects by mild hyperthermia. InlJ. Radial Oncol. Biol. Phys., 24,965-73. Iwata, K., Shakil, A., Hur, W.J., Makepeace, C.M., Griffen, R.J. and Song, C.W. (1996) Tumor p02 can be markedly increased by mild hyperthermia. Br.J. Cancer, 74, S217-21. Song, C.W., Shakil, A., Osborn, J.L. and Iwata, K. (1996) Tumor oxygenation is increased by hyperthermia at mild temperatures. InlJ. Hypertherm., 12(3), 367-73. Brizel, D.M. (1997) Personal communication. Department of Radiation Oncology, Duke University, North Carolina, USA. Armour, E.P., McEachern, D., Wang, Z., Corry, P.M. and Martinez, A. (1993) Sensitivity of human cells to mild hyperthermia. Cancer Res., 53,2740-4. Raaphorst, G.P., Bussey, A., Heller, D.P. and Ng, C.E. (1994) Comparison of thermoradiosensitization in two human melanoma cell lines and one fibroblast cell line by concurrent mild hyperthermia and low-dose-rate radiation, Radial Res., 137,338-45. Rosner, G.L, Clegg, S.T., Prescott, D.M. and Dewhirst, M.W. (1996) Estimation of cell survival in tumours heated to non-uniform temperature distributions. Int.J. Hypertherm., 12(2), 223-39. Corry, P.M., Jabboury, K., Kong, J.S., Armour, E.P., McCraw, J.F. and Leduc, T. (1988) Phase I evaluation of equipment for hyperthermic treatment of cancer. Int. J. Hypertherm., 4, 53-74. Kapp, D.S., Fessenden, P., Samulski, T.V., Bagshaw, M.A., Cox, R.S. and Lee, E.R. (1988) Stanford University Institutional Report: Phase I evaluation of equipment for hyperthermic treatment of cancer. Int. J. Hypertherm., 4, 75-116. Sapozink, M.D., Gibbs, F.A., Gibbs, P. and Stewart, J.R. (1988) Phase I evaluation of hyperthermia equipment: University of Utah Institutional Report. InlJ. Hypertherm., 4,117-32. Shimm, D.S., Cetas, T.C., Oleson, J.R., Cassady, J.R. and Sim, D.A. (1988) Clinical evaluation of hyperthermia equipment: The University of Arizona Institutional Report for the NCI Hyperthermia Equipment Evaluation Contract. InlJ. Hypertherm., 4,39-52. Ling, C.C. and Robinson, E. (1988) Moderate hyperthermia and low dose rate radiation. Radial Res., 114,379-84.
References 409
27. Armour, E.P., Wang, Z., Corry, P.M. and Martinez, A. (1991) Sensitization of rat 9L gliosarcoma cells to low dose rate irradiation by long duration 41 ° hyperthermia. Cancer Res., 51,3088-95. 28. Armour, E.P., Wang, Z., Corry, P.M. and Martinez, A. (1992) Equivalence of continuous and pulse simulated low dose rate irradiation in 9L gliosarcoma cells at 37° and 41°. Int.J. Radial Oncol. Biol. Phys., 22(1), 109-14.
29. Armour, E.P., Wang, Z., Corry, P.M., Chen, P.Y. and Martinez, A. (1993) Hyperthermic enhancement of high dose-rate irradiation in 9L gliosarcoma cells. Int.J. Radial Oncol. Biol. Phys., 28,171 -7. 30. Sapareto, S.A. and Corry, P.M. (1989) A proposed standard format for hyperthermia treatment data. Int.J. Radial Oncol. Biol. Phys., 16,613-27.
31 The costs of brachytherapy GRAHAM READ
31.1
INTRODUCTION
Healthcare costs have shown a consistently rising trend worldwide during the last three decades. Thus, in the USA healthcare spending, expressed as a percentage of the gross national product (GNP), rose from 6% in 1965 to over 14% in 1992 and to 14.9% in the year 2000 approximately $1.5 x 1012 [1]. There is naturally, therefore, an increasing interest in cost-effectiveness and cost containment [2,3]. This will particularly focus on cancer as, if present trends continue, cancer will soon surpass cardiovascular disease as the leading cause of death in the USA, with an estimated annual cost of $82 x 109, approximately 20% of all healthcare costs [4]. In the UK, the National Health Service (NHS) reforms of 1991 [5] were driven by a vision of a cost-efficient healthcare system based upon an efficient contracting and purchasing process [6]. This has subsequently been replaced by a looser system of service and financial frameworks (SaFFs) to be supported by clinical governance and performance indicators, with regulatory bodies such as the National Institute for Clinical Excellence (NICE) and the Commission for Health Improvement (CHI) [7]. In Canada, the previously unchallenged healthcare system has recently come under close evaluation and financial scrutiny [8].
31.2
RADIOTHERAPY COSTS
Radiotherapy has been perceived, incorrectly as this chapter will show, as an expensive treatment modality [9,10]. Analyses of the costs of radiotherapy have generally been directed at treatments involving linear accelerators [11,12]. Although this in itself presents considerable difficulties, the costing of brachytherapy is even more complex. Firstly, there is a greater variation in the modalities used: some having a high capital component, such as high dose-rate (HDR) brachytherapy, while others have a high revenue cost, such as interstitial seed insertion. Secondly, there has been a greater change in practice with the abandonment of early techniques such as those employing radium and cobalt and the adoption of newer procedures, usually involving some form of remote afterloading. These changes have occurred, in many cases, for sound clinical reasons, for example the re-evaluation of different forms of radiotherapy or a movement to other forms of management such as surgery. Changes may also be brought about, however, for reasons other than clinical. For example, amendments and greater stringency in protection regulations have led to a greater use of afterloading methods as opposed to those involving the handling of live sources. In some instances, certain types of radioactive source have become unavailable. Lastly, as with the costs of cytotoxic drugs, there have been large
Principles of costing 411
changes in the costs of some sources: for example, the cost of gold seeds rose approximately ten times between 1980 and 1990. However, the principal difficulty in assessing the costs of brachytherapy remains the lack of published information available.
313
PRI NCI PLES OF COSTING
313.1
Variables
Costs may be divided into those which are fixed, semivariable or variable. Fixed costs remain unchanged regardless of the level of activity (Figure 31.1). These include many aspects of the basic administration of a hospital such as the provision of financial, catering, and general maintenance services. Semi-variable (semi-fixed or step) costs remain constant only for a certain level of activity (Figure 31.2). These include most of the staffing costs and also items of capital expenditure such as afterloading equipment. Thus, one low dose-rate (LDR) or HDR machine may be adequate for a certain number of patients, above which a further unit would be required. Variable costs vary in direct proportion to the level of activity (Figure 31.3). These include drugs and disposable sources such as iodine seeds.
313*2
Capital costs
Capital costs are the initial costs of any item of equipment and those which arise subsequently, namely interest and depreciation. Because costs and benefits in any department accrue at different rates over different periods, investment appraisal is generally taken into account by the use of investment procedures, or discounting. The object of these procedures is to enable costs and benefits to be evaluated as if they had occurred at the same point in time. Thus, the initial capital costs can be converted into a series of annual payments by using a given interest or discount rate. These annual payments are referred to as equivalent annual costs (EAC). The EAC can take into account situations in which different components of capital expenditure have differing lifetimes: thus, an item of after-loading equipment may be amortised (paid off) over a period of 15 years, whereas a building may have a lifetime of 60 years. Since 1991, all providers in the UK have been required to keep capital asset registers of items valued in excess of £1000 and pay interest and depreciation charges at the recommended discount rate for public sector projects of 6%, in accordance with UK Treasury guidelines [13]. Frequently, an item of equipment has little alternative use (when it could give rise to an opportunity cost) and is then referred to as a sunk cost. Sunk costs should not be included amongst annual capital costs.
3133
Revenue costs
Revenue or operating costs include all those costs which recur annually. In the case of afterloading equipment, these include maintenance on the equipment and renewal of sources. These costs are fixed, because they will be unchanged irrespective of the number of patients treated. The costs of source replacement clearly relate to the half-life of the source used and are, therefore, higher for machines employing iridium in comparison to those using cesium (Table 31.1). Other operating costs will be
Table 31.1 Fixed costs for LDR and HDR afterloading equipment
Machine3 Gynecological applicators
285077
14182
7191
First-year sources
40878
81756
20000
283293
396000
312268
42984
81887
223483
326277
477887
535751
Sources and service"
Variable costs.
299862
7191
Total initial costs
Total fixed costs
Figure 31.3
235224
"Prices relate to Nucletron equipment. Discounted at 5% over 8 years. Data in Canadian dollars (1992) from Jones et al. [36].
412 The costs of brachytherapy
variable, including operating room costs, admissions costs, overheads, and consumables. STAFF (LABOUR) COSTS These generally account for a substantial proportion of all costs - thus, in the UK, approximately 70% of all National Health Service costs relate to starring and, in the USA, Perez et al. calculated that they were 66% of their facility's global costs [ 14]. In the UK, general management, senior medical and nursing staff are regarded as fixed costs, and junior medical, nursing, physics, portering, and technicians are regarded as semi-fixed [15]. In contrast, in the USA, direct labor costs are regarded as variable. Staff costs are particularly important as brachytherapy procedures are complex and labor intensive. Thus, Perez et al. calculated that an intracavitary insertion required an average of 195 min of medical time, and an interstitial implant 263 min, in contrast to times for external-beam treatment planning of 23-45 min and simulation 29-45 min. 31.3.4
Marginal costs
Although estimates of average costs have frequently been used in healthcare accounting, it may be more useful in some circumstances to estimate the marginal costs; that is, those costs associated with a small change in healthcare activity [16]. The marginal costs incurred in treating, for example, one additional patient on an HDR machine will normally be very much smaller than those measured by using average cost estimates. Conversely, the savings made, for example, by reducing the number of fractions will be much smaller than an average cost estimate. Consequently, the amount of money saved by making small changes in clinical practice is often overestimated. Thus, many procedures will result in a reduction of the length of stay for inpatients, but the savings made will only be the marginal rather than the average costs. Substantial savings are only made in the rare situations in which it is possible to lay-off machines and staff and close wards. 31.3.5
Inpatient or outpatient costs
During the last decade, there has been a progressive trend to outpatient care. It has been estimated that
85-90% of all cancer care is now administered on an outpatient basis [17]. This has occurred because of the recognition of the high costs of inpatient care and, in the USA, because of changes in healthcare legislation which have affected patterns of reimbursement [18]. The relative costs of inpatient and outpatient treatments may have a significant effect in cost minimization analyses, for example when comparing HDR with LDR, as discussed below (Table 31.2). The relative benefits of outpatient treatments are so far poorly documented. For example, Yokes et al. found that, while outpatient chemotherapy was 53% less expensive than inpatient care, a significant proportion of patients refused outpatient care because of fear of malfunction of the equipment or inconvenience due to frequent clinic visits or restrictions in daily home activities [19].
31.3.6
Other costs
Although the focus of costing falls naturally on those services provided by the health service or healthcare provider, many costs are borne by patients and their families or carers. These include direct costs such as travel or absence from work, which may be considerable. Thus, in a study in the west of Scotland [20], the time patients receiving radiotherapy spent away from home each day varied from 35 min to 7 h, with a mean of 2 h 50 min; 16% had a relative who had to take time off work. Indirect costs may also result from psychological distress or the 'knock-on' consequences of loss of work and production.
31.3.7
Costs of failure
Further considerations which are frequently ignored are the costs evoked by either not treating the patient or by failing to cure. It is often assumed that the costs of supportive care are low in comparison with the costs of active treatment. This has been shown not to be the case. Perez demonstrated that the costs of curing a patient with a stage A2 (Tib - UICC TNM stage classification) carcinoma of the prostate with radiotherapy were
Table 31.2 Costs of LDR and HDR
Staff costs Variable costs Capital
1833.73 383.98 3327.78
Total
5545.49
33.1 6.9 60.0 100
Costs per patient C$1997 from reference [38].
1277.38 103.42 3173.03 4553.83
28.1 2.3 69.7 100
Charging for healthcare 413
$199320000' compared with $199347000 if the treatment was unsuccessful [21]. It is interesting to note that the equivalent surgical costs were $199325 000 and $199349 307, respectively.
31.4
CHARGING FOR HEALTHCARE
A simple question which is often asked is, 'How much does procedure x cost?' (Note this is not the same question as 'What is the price of procedure *?') Unfortunately, there is rarely a simple answer. In countries such as the UK where there is state funding or insurance, the overall costs of a department will be known, but the costs of individual procedures may be difficult to calculate. Where healthcare is charged directly to the patient or via an insurance company, some estimate of the costs of different procedures will have been performed to enable a charge to be made. This may not necessarily represent the true costs because there may be an element of profit included or the charge may subsidize or be subsidized by other costs. Payment patterns may have been arbitrarily set up over a period of years independent of cost factors. Thus, Perez et al. found that brachytherapy accounted for 10% of technical costs, but only 4% of revenue [14]. Costs for individual procedures may depend upon how quickly it is intended to recoup the initial capital expenditure (for an example, see Bastin et al. [22]).
31.4.1
UK funding
In the UK, no charging existed for several decades after the inception of the NHS and therefore costing was poorly developed and ill-understood. The 1990 National Health Service and Community Care Act split District Health Authorities (DHAs) in the UK into purchaser units which purchase services for patients from their own hospitals, from other authorities' hospitals, from self-governing trust hospitals, or from the private sector, and provider units, usually representing their district general hospitals (DGHs), which provide patient services. A further innovation was the creation of fundholding general practices which, with money top-sliced from the DHAs, were able to buy services from those provider units which in their view offered the best services. With a change of government in 1997, a further major revision has been initiated, with the creation of primary care groups (PCGs) of general practitioners serving approximately 100 000 people. Four possible levels of function of PCGs are envisaged, with the highest being a trust having total purchasing power for its population. 1
For simplicity all monetary amounts have been left in their original currencies with a suffix denoting either the year in which the calculations were made where this is stated in the original articles or else the year of the article.
Certain services, such as specialist cancer services, will again be purchased by Regional Health Authorities (RHAs), restoring some functions which had been previously eroded - indeed they had been merged into eight units in 1994 and complete abolition had been scheduled for 1996. Despite these reforms, the basic purchaser/ provider division has been retained. However, without adequate information on costs and outcomes, their ability to make informed decisions is limited. The 1991 reforms and the creation of an internal market have forced a change from management accounting to financial accounting. NHS Trusts have an obligation to: (a) deliver 6% return on their capital assets; (b) balance their budget; and (c) live within their external spending limit. This has led to very stringent financial control. Costs have had to be assigned to different procedures, usually on the basis of inadequate data, in order to draw up contracts. Financial management was previously carried out on a purely departmental basis, which meant that the costs of any one procedure were usually unknown. This sort of budgeting is known as an input budget and has no regard for the level of activity, or outputs, in a department. In order to make an estimate of the costs of a procedure, estimates have to made of the direct operating costs, including staff and consumables, and the indirect operating costs, such as cleaning and portering. The calculation of indirect costs will vary depending upon the method of recharging used, and capital costs will depend upon whether the procedure uses a purpose-built protection suite or areas shared with other users. Clearly, there will be considerable differences amongst centers and countries.
31.4.2
USA funding
In the USA, there was a progressive shift away from 'out of pocket' payments to funding from Federal Government or private insurance. The Social Security Act Amendments of 1983 shifted Medicare reimbursement away from retrospective payments to a prospective payment system based upon diagnosis-related groups (DRGs). In 1989, cost-effectiveness was proposed as a criterion of funding by the Health Care Financing Administration [23]. Increasingly, purchasing cooperatives have been formed with extensive databases on performance and cost. They have attempted to maintain or lower costs by the use of clinical protocols, disease and case management, and outcomes research [24].
31.4.3
Healthcare resource groups
The DRG system is now being emulated in the UK with the development of healthcare resource groups (HRGs) [25]. These are intended to define patient groups which are both clinically relevant and consume similar levels of resources [26]. Provisional groupings developed for
414 The costs of brachytherapy
radiotherapy and chemotherapy were tested at three sites in 1992 [27] and, after further piloting at eight additional sites, definitive recommendations have now been published for use commencing April 2000 [28]. The proposed groups for brachytherapy are shown in Table 31.3. Table 31.3 UK healthcare resource groups in brachytherapy (version 1)
Mechanical afterloading
HDR LDR
Manual afterloading Live source
31,5 31.5.1
COSTS OF BRACHYTHERAPY Cost areas
The principal areas are shown in Table 31.4. Capital costs will include the costs of the afterloading equipment and applicators. Increasing attention is being given to treatment planning and optimization. As the systems used frequently form part of a larger system for external radiotherapy planning, the actual costs are frequently hidden. Other important areas include operating theater costs, hospitalization and staff costs. Hospitalization costs are usually derived from estimates of the average cost of an inpatient day. The cost reduction from using brachytherapy, particularly HDR therapy, will often result from the savings of such costs. Whether these costs are actually realized will depend upon the alternative uses to which the bed is put, and cost savings should, therefore, be restricted to marginal rather than average costs. These savings should include the reduction in the number of specialized nurses necessary to supervise inpatient treatments. 31.5.2
Costs of manual techniques
These include both the insertion/implantation of active sources and the use of afterloading techniques. A
significant feature of the cost is that of the isotope. Unlike the older radionuclides radium and cobalt, which could be kept in stock and used for a large number of patients over a long period of time, newer isotopes have relatively short half-lives and may require a new order for each patient. The length of theater time will depend upon the type of procedure and the experience of the operator. An iridium implant might occupy an hour of theatre time, whereas a prostatic iodine seed implant may require two or three. Telliffe [29] estimated the cost of a breast implantation using iridium-192 wire to be £1990689. The major component of this expenditure was the cost of the iridium wire (£1990280) and hospitalization for 2 days (£1990240). One hour of operating time was included, but this estimate really represents only the marginal costs of the procedure as detailed estimates of the true operating theater and accommodation charges (which may include a purposebuilt protection suite) were not made. The duration of stay would clearly have a significant effect on the overall costs.
31.5.3
Costs of afterloading techniques
These differ significantly amongst LDR, medium doserate (MDR), and HDR machines. This is due, firstly, to the difference in the half-lives of the sources. Source costs may be included as part of the initial capital cost if the isotope is long lived. Short-lived isotopes may be included with revenue costs. Example costs for Nucletron HDR and LDR machines, building, and maintenance are shown in Tables 31.1 and 31.5. Secondly, LDR and MDR techniques require hospitalization, with its associated costs. Costs will also depend upon the operating room costs and the type of anesthesia or sedation employed. Jelliffe [29] also estimated the costs of brachytherapy using afterloading techniques in breast cancer patients on the basis of 100 patients per annum. Again, no allowance was made for building costs, and it is not clear whether the annual maintenance charge included depreciation of capital. He estimated the cost of an LDR implant using cesium-137 to be £1990618, of which £1990273 was attributed to the costs of the Nucletron LDR
Table 31.4 Cost areas in different brachytherapy treatments
Capital Isotope/source Anesthetic/theater Hospitalization Mould room Staff costs
++
++
+ + +
+
++
+
+
+
+
+
++ ++ +
+
+
Economic evaluation 415
Table 31.5 Fixed costs for afterloading equipment
Estimated cost Capital investment-initial HDR afterloading equipment Verification system options Treatment room Miscellaneous
300000 100000 50000-100000 5000
Capital investment-annual Maintenance Sources Supplies Verification system maintenance Total (7-year)
12000 16000 75000 10000
633500-798000
Data in American dollars (1993) from Bastin et al. [22].
machine based upon an initial cost of £1990180000, including 30 cesium-137 source trains and an annual maintenance cost £199010 340 after the first year. Two days' hospitalization was costed at £1990240 and 1 h operating theater time was allowed at £199080. Treatment with an HDR unit was estimated to cost £1990435. The cost of the HDR machine was taken as £1990280, based upon an initial cost of £1990140000, but with replacement iridium192 sources at £19906000 per annum. The difference in the costs of the LDR and HDR techniques was mainly accounted for by the lack of hospitalization with HDR. These average costs are clearly heavily dependent upon the number of patients treated per annum. Benn [30] attempted to calculate and compare costs in two hospitals with a differing workload. Again, building and depreciation costs were not included. He estimated the cost of a single LDR to lie between £1990510 and £1990672. He found the cost of an HDR fraction to lie between £199049.5 and £1990117. Bastin et al. [22] estimated that an HDR application should be charged at $19931008, but this was dependent upon the rate at which the initial capital costs were recovered.
31.6
ECONOMIC EVALUATION
A number of methods of analysis seek to relate the costs and consequences of healthcare programs. Some of these expressions are used as 'umbrella' terms to cover any form of cost analysis, e.g., cost-effectiveness or cost benefit. However, the more precise meanings defined below are generally accepted. Although cost analyses have been increasingly reported in medicine, the number relating to cancer is relatively few. Not unnaturally, because of the high costs, procedures in medical oncology have attracted the largest number [31] of studies, whereas those in brachytherapy are extremely few in number at the present time.
Some have argued that cost analyses are inappropriate in medicine [32]. Indeed, in brachytherapy, one may be faced with the situation in which, for example, the costs of an implantation may be greater than those of an equivalent external-beam treatment, but the benefits, such as improved cosmesis, may be very difficult to quantify in monetary terms.
31.6.1
Cost minimization
The method compares the total costs of two different strategies which have the same outcome. Unfortunately, in clinical practice it is rare to find two modalities which differ significantly (and therefore have significantly differing costs) for which there is general agreement that the outcomes are the same. In general, comparisons of modalities in the literature have focused on survival and side-effects and only in recent years has the specific issue of cost comparison been addressed. MANUAL VERSUS REMOTE AFTERLOADING BRACHYTHERAPY
Jelliffe [29] estimated the cost of a manual breast implantation using iridium-192 wire to be £1990689, compared to £1990435 and £1990618 for the HDR and LDR implants. Ostrowski [33] compared manual and remote afterloadine techniaues. LDR VERSUS HDR BRACHYTHERAPY
A number of studies have sought to compare LDR with HDR brachytherapy. Leaving aside considerations as to whether the clinical outcomes are truly similar, there remains considerable controversy as to the optimal technique [34]. Because the numbers of applications of each modality required to achieve the same endpoint may vary considerably [35], this has an important consequence upon the cost-minimization analysis. Thus, Bastin et al. [22] compared LDR in gynecological cancer consisting of two applications with 3 days' hospitalization with an HDR of five outpatient applications. The costs of LDR were 244% higher, primarily due to hospital and operating theater costs. The cost estimates of Jelliffe and Benn show, however, that the comparison is not simple. Taking only average patient costs into consideration, the cost will depend upon the number of patients treated. Furthermore, the number of fractions used varies. Considering the marginal costs, these are clearly smaller in the case of HDR than LDR. In a detailed analysis, Jones et al. [36] compared the costs of HDR and LDR machines using Nucletron LDR-3, LDR-6, and HDR units. They found that the total fixed cost of the HDR unit was C$1992209 474 greater than the cost of the LDR-3 unit and C$,99257 864 more than the LDR-6 unit (see Table 31.1). However, analysis of the operating costs showed that, for various schedules of
416 The costs of brachytherapy
LDR and HDR, the operating costs per patient were less for HDR than for LDR, except for the comparison of one LDR insertion with four HDR insertions. Savings increased in favor of HDR over LDR as the annual case load increased (Figure 31.4). They also concluded that, for small units treating up to 40 patients per year, the LDR-3 unit was the most cost-effective machine. In general, for a greater number of patients, HDR would be recommended, but they produced algorithms allowing more precise assessments to be made depending upon the annual number of patients, the ratio of LDR to HDR insertions, and other practical considerations. Their calculations assumed that the machines were used only for gynecological insertions. Cost sharing, for example for the treatment of cancer of the esophagus, lung, or breast, would affect the calculations significantly. Chenery et al. estimated that HDR would lead to a cost avoidance of C$19841700000, based upon an annual patient load of 85 cervical and 60 endometrial cancer patients over a period of 20 years [37]. More recently, Pinilla [38] calculated that an HDR regimen (two fractions) was 22% less expensive compared to an LDR (one fraction) regimen amounting to C$991.66 per patient (see Table 31.2). However, if LDR maintenance was done in house, then the LDR cost fell to 98% of the HDR cost. Konski et al. [39] carried out a meta-analysis of six options in stage I endometrial cancer based upon costs paid, and concluded that LDR was the most costeffective treatment, with no evidence of any difference in disease-free survival. Thus, a simple choice between LDR and HDR cannot be made as it depends upon the number of fractions used, the number of patients treated, and the maintenance costs. However, there appears to be a general agreement that, except for in the smallest centers, HDR is likely to prove more cost-effective than LDR.
BRACHYTHERAPY AND EXTERNAL-BEAM THERAPY
In many instances in oncology, it will be clear from the extent or site of the tumor that external-beam radiotherapy is the treatment of choice. In some instances, a brachytherapy insertion may be added at the conclusion of the external-beam treatment to supplement the radiation dose to an area of special risk. Here, the issue will be whether or not this treatment, and the implied additional cost, confers a benefit in terms of improved local control or survival. BRACHYTHERAPY AND ELECTRON THERAPY
Jelliffe compared the cost of a five-fraction electron boost to that of an interstitial implant [29]. His estimate of the cost of electron therapy was £1990200, which compared favorably with costs of £1990435-£1990689 for the implants, but he noted that the poor cosmetic results from electrons might justify the higher costs of the implant. BRACHYTHERAPY AND SURGERY
Comparisons between the use of brachytherapy and surgery are often highly controversial and fraught with difficulty. Patients treated surgically are frequently younger and fitter. The surgical modality may, by its very nature, provide more accurate staging information, rendering stage-by-stage comparison difficult. For example, Farndon et al. [40] reported that surgical resection in esophageal cancer was more cost-effective than other methods of treatment, including brachytherapy. However, as the survival amongst surgical patients was significantly greater, it is unlikely that the patients' staging was comparable. Although it is possible to compare
Figure 31.4 Annual cost differences in Canadian dollars (1992) for LDR compared with HDR for different numbers of patients (from Jones et al. [36]). Positive values indicate cost balance in favor of LDR, negative in favor of HDR.
Economic evaluation 417
the costs of the initial procedures, these frequently do not take into account the costs associated with failure of treatment or of any complications which may arise. External-beam radiotherapy as an outpatient treatment has been shown to be less expensive than comparable surgical procedures for a number of common cancers because of the avoidance of hospital, anesthetic, and operating theater costs [41]. Hanks and Dunlap determined the costs of treating prostate cancer by radical prostatectomy, lymph node dissection with iodine-125 implant, and external-beam radiotherapy [42]. The median cost of radical prostatectomy was $198614000, of lymph node dissection and iodine-125 implantation $198612000, and of external-beam radiotherapy $6750 before 1984 and $5600 after 1984. These costs were derived from the hospital fees, professional fees, and other major expense items as charged and are therefore highly influenced by the method of charging as well as by the true costs of the procedures. Thus, the apparent change in the cost of radiotherapy was entirely due to changes in billing. It is interesting to note that bills for apparently the same procedure varied by up to 82%, indicating the inherent flaws in this method of determining costs. By reviewing the current literature, they concluded that none of the methods showed a superior outcome.
cost of a city bus ride!). This was one or more orders of magnitude different from the costs of a year gained by, for example, coronary bypass (C$19926698), renal dialysis (C$199267345), or school testing for tuberculosis (C$199269634) [44].
31*63
Although cost-effectiveness analyses may be very useful, they are unable to compare different diseases or strategies where the outcomes cannot be measured in a common unit. This may arise where, for example, a healthcare purchaser wishes to decide between, say, allocating money to cataract surgery or cancer chemotherapy. Cost utility relates the cost of different medical procedures to the increased utility (the amount of wellbeing) they produce in terms of improved quantity and quality of life. In order to measure utility, various quality-of-life scales have been developed. From the clinical point of view, quality-of-life studies are difficult and time consuming, in contrast to more objective measures such as response or survival. Even with apparently 'objective' scales, assessments made by doctors, patients, and their relatives may differ [45]. 31*6*4
31.6.2
Cost utility
Cost benefit
Cost-effectiveness
This form of analysis relates the cost of a treatment to its outcome. Two or more treatments can be compared provided there is a common unit of outcome or effectiveness such as 'life-years gained,' 'pain-free days,' or 'positive cases detected.' Thus, in a cost-effectiveness analysis, a ratio of benefit to cost is derived for each option. The most cost-effective option, therefore, can be defined either as that which maximizes benefits for a fixed cost or as that which minimizes costs for a fixed benefit [43]. It should be noted that this is not necessarily the largest net benefit which represents the optimum choice. In assessing the benefits of a particular treatment, the question arises as to how to deal with costs which occur at different points in time. This is clearly of importance when considering the immediate benefits of a particular treatment versus its long-term risks. Traditionally, benefits have been discounted in a manner similar to that described above for dealing with costs, which has the effect of weighting the short-term benefits. Adopting a zero discount rate, whilst it may be appropriate for areas such as neonatal care, would lead in cancer to an undue preoccupation with late effects in preference to immediate gains and may not, therefore, be appropriate. A number of studies have shown that, in general, radiotherapy is remarkably cost-effective. Thus, Glazebrook was able to show that for external-beam radiotherapy the cost of a year of life gained was C$1992661 (C$19921.82 per day- which he compared to the
This form of analysis, in many ways the most difficult, seeks to determine whether the benefits of using a given therapy outweigh its costs. A monetary value has, therefore, to be assigned to each strategy or treatment, which may amount to deciding how much it costs to save a life or to enable a person to live a pain-free one. Some people have an ethical objection to putting a monetary value on a human life, but such decisions are regularly made in economic planning, even if they are not implicitly stated. In fact, cost-benefit analysis has been regularly used in the analysis of economic and social policy in the public sector for many years. Roberts et al. estimated in 1985 that the NHS could not afford more than £198414000 to save a life [46], and Rees felt that treatment costing less than £19911000 for an improvement in benefit was excellent value, whereas those costing £199110000 probably represented an unfair distribution of resources [47]. Various approaches have been adopted in benefit evaluation. Initial approaches were based upon 'human capital', such as the loss of income incurred by illness. Clearly, this does not take into account benefits for the retired or unemployed. Another approach is to evaluate whether we are willing to pay the stated cost for a particular procedure or service. This is often expressed as a proportion of average income, as an alternative to a straight monetary cost. Thus, in one study, Thompson found that patients with rheumatoid arthritis would be willing to pay 22% of their annual income for a hypothetical cure of the disease.
418 The costs of brachytherapy
These considerations are important in brachytherapy. Thus, supposing that brachytherapy for the treatment of breast cancer is as effective as external-beam therapy but is more expensive (as suggested by Jelliffe above), what level of cost is one prepared to pay for improved cosmesis? Such considerations are important and are worthy of further study. The National Radiological Protection Board (NRPB) has attempted to calculate the cost of the health detriment caused by irradiation of the general public by taking into account the frequency of deleterious effects and their respective costs [48]. These costs were estimated from the loss of economic output, the costs of medical treatment for fatal and non-fatal cancers, and the costs for hereditary defects.
31*6*5
Radiation protection
The raison d'etre for the development of afterloading in brachytherapy was the reduction in radiation exposure to the various members of staff- radiotherapists, nurses, and technicians - who were involved in the treatment of patients with active sources, principally radium. As an example, the treatment of patients with cancer of the cervix using the Manchester System [49] involved the insertion of an average of 75 mg radium, which remained in position for approximately 3 days. Prior to the introduction of afterloading systems, approximately 400 patients were treated per annum, totalling approximately 700 insertions carried out, at the Christie Hospital, Manchester. Although, by adherence to all appropriate working procedures, radiation exposure of staff was maintained below the then annual dose limit of
50 mSv, a significant number of people received doses within the range 30-90% of this limit. From an economic perspective, the question arises as to whether the reduction in radiation dosage to staff and visitors justifies the increased costs of afterloading. Fleishman et al. [50] calculated the costs of two schemes for the introduction of LDR afterloading systems: scheme 1 involved the construction of a three twin-bedded, purpose-built, single-storey building, and scheme 2 the construction of two fully protected treatment rooms adjacent to a ward area. The annual cost of the first scheme was £198253000 and of the second £198230000 (Table 31.6). In order to assess the cost benefit of these schemes, reference was made to work by the NRPB, which had attempted to calculate monetary valuations of radiation-induced heath detriment [51], as shown in Table 31.7. These estimates are, of course, highly contentious, involving judgments by the NRPB of the value of public and occupational exposures and other assumptions in estimating the doses received. Using these criteria, Fleishman et al. calculated that the total detriment cost to staff was £198258 000 and to visitors £198234 000, a total of £198292 000, and hence that the costs of introducing afterloading could be justified.
31.7 31.7.1
AREAS FOR FUTURE STUDY Lung cancer
An important development has been the use of intraluminal radiotherapy for the treatment of carcinoma of the bronchus [52]. In order to be able to assess the cost
Table 31.6 Annual costs of two afterloading schemes
Annual capital costs of
(i) equipment (ii) building
26000 22000
26000 13000
Annual service costs of
(i) equipment (ii) building
15000 8000
15000 2000
2000
-6000
-20000
-20000
53000
30000
Net change in operating costs Bed occupancy savings Total net annual cost Data in £1982 taken from Fleishman et al. [50].
Table 31.7 Recommended costs of unit collective dose for members of the public
<5x1Q- 5 5xI0-5-5xI0-4 5x10-4 –5x10-3 a
<1 1-10 10-100
For explanation, see reference 51. Data from National Radiological Protection Board [51].
2000 10000 50000
Areas for future study 419
effectiveness of this treatment, the optimal scheduling, especially in relation to external-beam radiotherapy, has to be determined. Thus, one schedule for radical lung treatments employed external-beam radiation to a dose of 60 Gy in 6 weeks with brachytherapy on weeks 1, 3, and 5, giving 7.5 Gy at 10 mm from the source on each occasion [53]. It is difficult to extricate the value of brachytherapy in such a complex schedule. In palliative radiotherapy, the situation is clearer. The same author showed that adequate relief of symptoms could be achieved with brachytherapy alone using three applications of 10 Gy at 10 mm from the source on weeks 1, 2, and 3. Furthermore, Collins et al. showed that a single fraction of intraluminal radiotherapy as a primary treatment was as efficient in palliating symptoms as more prolonged courses of external-beam radiotherapy [54]. Clearly, with the recent demonstration of the effectiveness of single fractions of external-beam radiotherapy [55], further evaluation of the role of intraluminal radiotherapy in lung cancer is required. Whether such studies are evaluated as cost minimization or costeffectiveness will depend upon whether the evaluation is in terms of equivalence or some other measure such as 'days free of hemoptysis.' 31.7,2
Cervical cancer
Several areas are waiting for evaluation. In non-bulky stage I disease, both primary surgery and radiotherapy provide equal success rates. Could one perform a costminimization analysis in such a situation, and should this then be the factor which determines the choice of treatment? Such an analysis should go beyond the costs of the primary procedure. It should also include the costs of complications, whether surgical or due to late radiation effect. Ovarian preservation might, for example, obviate the need for hormone therapy. How should patient preference be taken into account and the differences in the quality of life - if they be significant - be evaluated? These important areas of debate require much further work. In higher stage disease, the importance of intracavitary radiation in the treatment plan has been established. In a review, Petereit et al. noted pelvic control rates of 52-74% with intracavitary irradiation, as compared with 33-45% without [34]. Economic considerations, therefore, relate to the comparison of LDR and HDR, as discussed above. 31.73
Esophageal cancer
In curative situations, the cost-effectiveness of externalbeam radiotherapy has already been demonstrated [56]. Intraluminal radiotherapy offers a promising simple approach to address the frequent problem of palliation in advanced tumors [57]. In evaluating palliation in
esophageal cancer, the costs of the procedure have to be compared with those of alternative surgical procedures, such as insertion of a stent or laser excision, or externalbeam radiotherapy. Its effectiveness can be set against the costs of nursing, which, in the case of a patient with severe dysphagia, possibly requiring tube or intravenous feeding, are likely to be very considerable. Low and Paliero reported a small, randomized trial comparing brachytherapy with laser photocoagulation [58]. Although both treatments were effective, laser excision required more expertise and was associated with higher numbers of retreatments and complications. Further studies are clearly needed, but evidence so far suggests that the use of intraluminal radiotherapy in this situation is likely to prove highly cost-effective. 31.7.4
Breast cancer
Brachytherapy in breast cancer is frequently used to boost the site of the primary tumor. From an economic point of view, it needs to be established, first, that the boost dose increases local control and, second, that the use of brachytherapy is superior to other forms of radiotherapy such as electrons. As noted above, a short course of electron therapy is probably less expensive than an implant and, therefore, the difficulty of assessing the monetary value of any improved cosmesis arises. Of greater interest is the possibility of treating selected cases with brachytherapy only when the possibility that brachytherapy would prove more cost-effective than external therapy arises. 31.7.5
Head and neck cancer
As with breast cancer, the difficulty of assessing the value of an implant as part of an external-beam treatment arises. However, the use of brachytherapy alone, for example in the treatment of tongue cancer, is likely to prove cost-effective. As with esophageal cancer, palliative implants to improve swallowing or reduce pain are also likely to be cost-effective. 31.7.6
Prostate cancer
There has been a major resurgence of interest in recent years in radioactive seed implantation in prostate cancer. This has been fueled by the increase in diagnosis of early prostate cancer, especially in the USA, where it is the most commonly diagnosed cancer in men. Various methods have been used to implant patients with early prostate cancer - principally iodine-125 seeds but also gold-192 seeds and, more recently, iridium-192. The difficulty of obtaining a satisfactory arrangement of the seeds has been improved by planning techniques using transrectal ultrasound. Although it has not been shown
420 The costs of brachytherapy
in a randomized control trial that local control is equivalent to or better than external-beam radiotherapy, the procedure offers many advantages, particularly in terms of convenience as it is may be carried out as a single outpatient procedure, in comparison to up to 8 weeks of daily external-beam radiotherapy. Important side-effects such as impotence may be reduced and the patient may rapidly return to work. Hanks and Dunlap [42] found that iodine-125 seed insertion was more expensive than external-beam radiotherapy, but their costs included lymphadenectomy and were based on hospital billing. Again, further detailed cost-based studies are required. 31*7.7
intervention. Note this applies to a cohort not an individual patient, so that 1 LY can equal 52 weeks for one patient or 4 weeks for 13 patients. Management accounting Information necessary for decision making, and the production of financial or business plans. Marginal costs Costs incurred by a small increase in treatment activity such as one additional patient. Semi-variable costs Costs which are fixed for a given level of activity but may change if the activity rises or falls beyond a certain threshold. Variable costs Costs which vary in proportion to activity levels.
Benign conditions
A recent innovation has been the use of brachytherapy to prevent restenosis after coronary artery angioplasty [59,60]. Economic evaluation in this area presents particular difficulties [61].
REFERENCES 1. Yabro, J.W. (1991) Changing cancer care in the 1990s and the cost. Cancer, 67,1718-27. 2. Griffin, T.W. (1994) Cost containment in health care: effects on American radiation oncology. Int.J. Radiat.
31.8
CONCLUSION
Oncol. Biol. Phys., 29(5), 1187-8. 3. Bernier, J. and Peters, LJ. (1994) Cost containment in
With rising costs of healthcare, the economic analysis of treatment modalities will be of paramount importance in the future. Unfortunately, cost analyses of brachytherapy have so far attracted little interest in published papers and are scarcely mentioned in textbooks on the subject. As noted in this chapter, although the costeffectiveness of brachytherapy can often be inferred, this is not a substitute for detailed, carefully costed, formal analyses. It is likely that, in comparison with radical surgical procedures, brachytherapy will acquit itself well in studies of cost-effectiveness. This may also prove to be true in comparison with radical external-beam treatments. However, brachytherapy is expensive when compared to a short course of external-beam radiotherapy. In particular, the value of boost treatments, which might replace a few additional external-beam or electron treatments, should be carefully considered. A plea is therefore made that authors include more analyses of costeffectiveness in future studies.
health care: a transatlantic perspective on radiotherapy. Int. J. Radiat. Oncol. Biol. Phys., 29(5), 1189-90. 4. Broder, S. (1991) The human costs of cancer and the response of the national cancer program. Cancer, 67, 1716-18. 5. National Health Service and Community Care Act 1990. London, HMSO. 6. Mawhinney, B. (1993) The vision for purchasing. In Purchasing for Health. A Framework for Action. Leeds, NHS Health Publications Unit. 7. The New NHS: Modern, Dependable (1997) London, HMSO. 8. Editorial (1995) Canada's threatened healthcare system. Lancet, 345,333-5. 9. Maynard, A. (1989) Is high cost medicine effective? Phys. Med.Boil.,34,407-18. 10. Kearney, B.J. (1994) The use of poorly validated and expensive technology [letter]. Med. J. Aust., 160,382. 11. Goddard, M. and HuttonJ. (1988) The costs of radiotherapy in cancer treatment. Centre for Health Economics Discussion Paper 48, University of York, UK. 12. Read, G. (1994) Estimating the cost of radiotherapy. Clin.
GLOSSARY Discounting A method of evaluating costs as if they had occurred at one point in time. Equivalent annual cost An annual payment calculated by a discounting method. Financial accounting The system necessary within an organization for transactions to take place. Fixed costs Costs which do not vary with changes in activity levels over a given time span. Life year (LY) Amount of time gained by a particular
Oncol., 6,35-9. 13. HM Treasury (1982) Investment Appraisal in the Public Sector. London, HMSO. 14. Perez, C.A., Koeissi, B., Smith, B.D., et al. (1992) Cost accounting in radiation oncology: a computer-based model for reimbursement. Int.J. Radiat. Oncol. Biol. Phys., 25,895-906. 15. Cost Allocation - General Principles and Approach (1994) London, NHS(ME) HMSO. 16. Goddard, M.and HuttonJ. (1991) Economic evaluation of trends in cancer therapy. Marginal or average costs? Int.J. Technol. Assess. Health Care, 7(4), 594-603.
References 421
17. Sandrik, K. (1990) Oncology-who's managing outpatient programs? Hospital, 64,32-7. 18. Rubenstein, E.B. (1994) Costs and benefits of outpatient therapy. Supportive Care Cancer, 2,307-11. 19. Yokes, E.E., Schilsky, R.L, Choi, K.E.etal.(1989) A randomised study of inpatient versus outpatient continuous infusion chemotherapy for patients with locally advanced head and neck cancer. Cancer, 62,30-6. 20. Junor, E.J., Macbeth, F.R.and Barrett, A.. (1992) An audit of travel and waiting times for outpatient radiotherapy. Clin.Oncol., 4(3), 174-6. 21. Perez, C.A. (1993) Quest for excellence: the ultimate goal of the radiation oncologist. Int.J. Radial Oncol. Biol. Phys., 26(4), 567-80. 22. Bastin, K., Buchler, D., Stitt,J.,et al. (1993) Resource utilization. High dose rate versus low dose rate brachytherapy for gynecological cancer. Am.J. Clin. Oncol., 16(3), 256-63. 23.
Leaf, A. (1989) Cost effectiveness as a criterion for Medicare coverage. N. Engl.J. Med., 321(13), 898-900.
24. RobbinsonJ.C. (1995) Health care purchasing and market changes in California. Health Affairs, 15,117-30. 25. National Casemix Office (1993) What are Healthcare Resource Groups? London, IMG(ME) HMSO.
37. Chenery, S.G.A., Pla, M. and Podgosarak, E.B. (1984) Physical characteristics of the Selectron high dose rate intracavitaryafterloader. Br.J. Radiol., 58, 735-40. 38. Pinilla, J. (1998) Cost minimization analysis of high-doserate versus low-dose-rate brachytherapy in endometrial cancer. Int. J. Radiat. Oncol. Biol. Phys., 42(1), 87-90. 39. Konski, A.A., Bracy, P.M., Jurs, S.G. and Zeidner, S.P. (1997) Cost minimization analysis of various treatment options for surgical stage 1 endometrial cancer. Int.J. Radiat. Oncol. Biol. Phys., 37(2), 367-73. 40. Farndon, M.A., Wayman, J., Clague, M.B. and Griffin, S.M. (1998) Cost-effectiveness in the management of patients with oesophageal cancer. Br.J. Surg., 85(10), 1394-8. 41. Parker, R.G. (1991) Varying charges for comparably effective cancer treatments. Am.J. Clin. Oncol., 15,281-7. 42. Hanks, G.E. and Dunlap, K. (1986) A comparison of the cost of various treatment methods for early cancer of the prostate. Int.J. Radiat. Oncol. Biol. Phys., 12(10), 1879-81. 43. Sugden, R. and Williams, A. (1978) The Principles of Practical Cost Benefit Analysis. Oxford, Oxford University Press. 44. Glazebrook, G.A. (1992) Radiation therapy: a long term cost benefit analysis in a North American region. Clin. Oncol., 4,302-5.
26. National Casemix Office (1993) HRG Definitions Manual, Version 1.0. London, IMG(ME)G, HMSO.
45. Slevin, M.L., Plant, H., Lynch, D., Drinkwater, J. and Gregory, W.M. (1988) Who should measure quality of life,
27. Casemix Development (1993) Radiotherapy and Oncology Project. Report and Recommendations for HRG Groups.
the doctor or the patient? Br.J. Cancer, 57,109-12. 46. Roberts, C.J., Farrow, S.C. and Charny, M.C. (1985) How
London, NHS(ME), HMSO.
much can the NHS afford to spend. Lancet, i, 89-91.
28. National Casemix Office (1998) Radiotherapy Healthcare
47. Rees, G. (1991) Cancer treatment: deciding what we can
Resource Groups, Version 1. London, NHS Executive, HMSO. 29. Jelliffe, A.M. (1992) Cost implication of HDRand LDR brachy-
48. Clark, M.J., Fleishman, A.B. and Webb, G.A.M. (1981)
therapy for breast cancer treatment. Activity, 6(2), 63-6. 30. Benn, D. (1992) A comparison of costs and treatment practices for low dose rate and high dose rate afterloading. BSc Thesis. 31. Smith, T.J., Hillner, B.E., Desch, C.E. (1993) Efficacy and cost-effectiveness of cancer treatment: Rational allocation of resources based on decision analysis.J. Natl Cancer Inst., 85,1460-74. 32. Doubilet, P., Weinstein, M.C. and McNeill, B.J. (1986) Use and misuse of the term cost-effective in medicine. N. Engl.J. Med., 314(4), 253-6. 33. Ostrowski, M.J. (1991) Cost review study of manual versus remote afterloading therapy for the treatment of early 34.
breast cancer. Activity, 5,7-12. Petereit, D.G., Fowler, J.F. and Kinsella, T.J. (1994) Optimising the dose rate and technique in cervical carcinoma - balancing cases versus complications. Int. J. Radiat. Oncol. Biol. Phys., 29(5), 1195-7.
35. Jones, B., Pryce, P.L, Blake, PR. and Dale, R.G. (1999) High dose rate brachytherapy practice for the treatment of gynaecological cancers in the UK. Br.J. Radiol., 72,371-7. 36. Jones, G., Lukka, H. and O'Brien, B. (1994) High dose rate versus low dose rate brachytherapy for squamous cell carcinoma of the cervix: an economic analysis. Br. J. Radiol., 67,1113-20.
afford. Br. Med.J., 302,799-800. Optimisation of the Radiological Protection of the Public. (A provisional framework document for the application of cost-benefit analysis to normal operations.) Oxford, NRPB R120 National Radiological Protection Board. 49. Meredith, W.J. (ed.) (1967) Radium Dosage: the Manchester System. Edinburgh, Livingstone. 50. Fleishman, A.B., Notley, H.M. and Wilkinson, J.M. (1983) Cost benefit analysis of radiological protection: a case study of remote afterloading in gynaecological radiotherapy. Br.J. Radiol., 56,737-44. 51. National Radiological Protection Board (NRPB) (1981) Cost-Benefit Analysis in Optimising the Radiological Protection of the Public: a Provisional Framework (ASP4). Chilton, Oxford, NRPB. 52. Stout, R. (1993) Endobronchial brachytherapy. Lung Cancer, 9,295-300. 53. Speiser, R. (1990) Protocol for local control of endobronchial carcinoma using remote afterloading HDR brachytherapy. Activity, (Suppl. 1) 16-22. Leersum,The Netherlands, Nucletron BV. 54. Gollins, S.W., Burt, P.A., Barber, P.V. and Stout, R. (1994) High dose rate intraluminal radiotherapy for carcinoma of the bronchus: outcome of treatment of 406 patients. Radiother. Oncol., 33,31-40. 55. Bleehen, N.M., Girling, D.J., Machin, D. and Stephens, R.J.
422 The costs of brachytherapy (1992) A Medical Research Council (MRC) randomised trial of palliative radiotherapy with two fractions or a single fraction in patients with inoperable non-small cell lung cancer (NSCLC) and poor performance status. Br. J. Cancer, 65,934-41. 56. Walker, Q.J., Salkeld, G., Hall, J. et al. (1994) The management of oesophageal carcinoma: radiotherapy or surgery? Eur.J. Cancer Ctin. Oncol., 25(11), 1657-62. 57. Brewster, A.E, Davidson, A.E., Makin, W.P., Stout, R. and Burt, P.A. (1994) Intraluminal brachytherapy using the high dose rate microSelectron in the palliation of carcinoma of the oesophagus. Clin. Oncol., 7,102-5. 58. Low, D.E. and Paliero, K.M. (1992) Prospective randomised
clinical trial comparing brachytherapy and laser photocoagulation for palliation of oesophageal cancer. J. Thorac. Cardiovasc. Surg., 104,173-9. 59. Stitt, J.A. and Thomadsen, B.R. (1997) Innovations and advances in brachytherapy. Semin. Oncol., 24(6), 696-706. 60. Nori, D., Parikh, S. and Moni, J. (1996) Management of peripheral vascular disease: innovative approaches using radiation therapy. Int.J. Radial Oncol. Biol. Phys., 36(4), 847-56. 61. Weinstraub, W.S. (1996) Evaluating the cost of therapy for restenosis: considerations for brachytherapy. Int.J. Radial Oncol. Biol. Phys., 36(4), 949-58.
32 Quality management: clinical aspects CAJOSLIN
32.1
INTRODUCTION
Brachytherapy has been an established and clinically proven method of achieving satisfactory tumor control since the early part of the twentieth century. The published work provides an enormous background of historical data, although it is often difficult to be sure of the precise details of the treatment given. In particular, it is often difficult to determine the extent and histopathology of the tumors being treated or the effectiveness of treatment, especially in relation to normal tissue effects, except in the most basic of details. Professional and public demands make it important to be able to verify that a particular mode of treatment not only offers the greatest chance of success coupled with the least interference in quality of life, but is also cost-effective when compared against other treatments. In order to meet these requirements, it is necessary to be able to show, by means of audit, the end-results of a particular treatment. This will involve having accurate data, which meet with universal acceptance. The collection of such accurate data will entail having a quality assurance program in place. Quality assurance in radiotherapy should cover not only the physics and technical aspects, but also the clinical management aspects. This is particularly important because radiotherapy involves a large number of processes, all of which should be covered by an appropriate quality assurance procedure. It also involves a number of different professional groups and it is extremely important to have professional cohesion and communication so that no shortfall exists between the various areas of activity.
The International Atomic Energy Agency (IAEA) and the International Commission on Radiological Units and Measurements (ICRU) promoted the unification of standards and specification of the physical, mechanical, and clinical dosimetry in 1984 [1], In general, brachytherapy was included with teletherapy, but more recently it has become recognized that brachytherapy has additional and special requirements. In particular, radiation hazards, the use of afterloading machines, and brachytherapy techniques differ from those of teletherapy. However, in common with all forms of radiotherapy, quality assurance is about reducing uncertainties of a particular aspect of treatment to an agreed level with a degree of confidence. This will apply whether it is a quantity of radiation activity, rate of delivery of treatment, objectively identifying the limits of a tumor, or assessing the response to treatment. In common with this approach, all definitions and end-points should have international agreement. Also, details of the assessment of response to treatment and normal-tissue morbidity will be obtained with a greater degree of confidence in those situations in which patient treatment has been subject to quality assurance criteria. However, quality assurance is but one component part towards establishing quality management in a treatment program. In fact, three basic component parts can be identified, which form the supporting structure to achieving best practice in brachytherapy (Figure 32.1). These are: 1. quality control, 2. quality assurance, and 3. audit.
424 Quality management: clinical aspects
degree of confidence that the patient received the intended treatment. It is, in effect, the process of confirming that what we do is what we think we do. From the radiotherapy context the World Health Organization [2] has defined quality as: 'all those procedures that ensure consistency of the medical prescription and the safe fulfilment of that prescription as regards dose to the target volume, together with minimal dose to normal tissue, minimal expense of personnel and adequate patient monitoring aimed at determining the end result of treatment'.
Figure 32.1 Quality management as applied to a treatment facility involves a series of interlinked processes. Q.C. = quality control; Q.A.= quality assurance.
The importance of quality assurance measures has been reported by a number of authors and in particular the errors which may occur as a result of poor measures of quality assurance [2]. A questioning attitude toward the effectiveness of the procedures used and accuracy of the equipment as it applies to a particular treatment has been shown to be beneficial in quality assurance [3-6]. While much of this work has been done for teletherapy, these principles also apply to brachytherapy [7-9,18]. PHYSICS ASPECTS
32.1.1
Quality control
Quality control involves those processes which regulate the delivery of a service. PHYSICS ASPECTS In brachytherapy, these will involve such processes as source specification and decay corrections, source calibration, and dosimetry. Each of these processes will be affected by various factors, e.g., source calibration procedures will be dependent upon the type of source, the distance at which calibration is done, the specification quantity (activity or air kerma rate), and other correction factors such as half-life, attenuation, and scatter. National and international agreements on measuring radiation dose against a standard reference calibration have been used in teletherapy for many years. In common with this, brachytherapy calibration procedures should be traceable to a national standard. These aspects are reviewed in detail by Wilkinson and Jones, respectively, in Chapters 2 and 3. CLINICAL ASPECTS These involve such processes as tumor staging, histopathological classification, tumor site, the patient's medical condition and age. They will also involve those processes used to assess tumor response to irradiation and early and late normal-tissue effects.
32.1.2
Quality assurance
Quality assurance involves those processes which confirm the process of delivery in order to provide a high
These should entail independent checks on all those aspects that involve measurement(s) or calculation(s) of matters such as source calibration, dosimetry, treatment planning, dose calculation, and checks following servicing and repair of machines. These aspects are covered in the physics section of this book. CLINICAL ASPECTS These entail ensuring that the various clinical and pathological investigations and other assessment parameters have been satisfactorily carried out and recorded and that the intended treatment plan was satisfactorily completed. They also involve ensuring that the measurement of tumor and normal-tissue responses has been systematically recorded according to an established system.
32.1.3
Audit
Audit is the process of assessment of outcome resulting from a particular treatment. It provides a means of identifying best practice when comparing one treatment process with another and of the cost-effectiveness of treatment [15,19]. It will require having accurate information regarding tumor response to treatment, cancerfree interval, and details of any recurrent disease. It is important also to have details of all acute and late normal-tissue responses to treatment [29]. These should be graded according to an established system [10]. Unfortunately, there is no universal system which will adequately apply to all sites, and various attempts have been made to overcome this [33]. An alternative is to provide assessment on a site-by-site basis [34].
Clinical aspects of quality management 425
One illustration of how to determine the impact of research findings and evolving technology on radiotherapy practice for cancer of the uterine cervix was initiated by the Patterns of Care, Study Group of the American College of Radiology in 1973. That work has continued and various site reports of long-term management are now being published [32]. QUALITY MANAGEMENT Thus, quality management includes a series of interlinking processes involving assurance and effectiveness of treatment (Figure 32.1). It should be seen as a way of providing a high quality of service for all patients having routine treatment which will match that more generally recognized to apply to patients being treated in a controlled clinical trial. Also, the proper assessment of any new technique can only be achieved in a quality management environment, otherwise there is the risk of it becoming established but remaining unproven. Brachytherapy, with its rapidly expanding technological developments, particularly when applied to new and innovative treatment, is but one example of a situation in which quality management is essential. Brachytherapy clinical trials have an important role in establishing the best treatment for a particular situation [7]. However, in some situations this will require national or international cooperation. The specific aims of such trials were considered in 1988 [7] and the various aspects addressed under the following headings: 1. Provide for international exchange of scientific data. 2. Enhance cooperation of investigational brachytherapy. 3. Identify the role of brachytherapy and appropriate clinical trials. 4. Study the role of brachytherapy used alone or in combination with other modalities. 5. Standardize the nomenclature used. 6. Standardize brachytherapy techniques, quality control procedures, and reporting methods. 7. Develop and define specific end-point(s) to be used to quantify tumor response and normal-tissue effects. 8. Determine appropriate toxicity scoring methods. 9. Evaluate the clinical significance of the treatment data (volume, dose rate, total time, total dose, and implant techniques). Thus, even within the ambit of a controlled clinical trial, quality management procedures should be standardized in a manner that can also be applied as part of everyday practice. Quality assurance as applied to brachytherapy physics has already been discussed in Chapters 8 and 9. Some of the clinical aspects of quality management as it applies to brachytherapy, either alone or in combination with external-beam irradiation, are now discussed.
The major purpose of applying quality management in the clinical context should be to reduce subjective and increase objective assessment wherever possible. In common with this approach, the definitions and end-points used should follow international recommendations such as those of the ICRU-38 and ICRU-58, which deal with specifications and measurements in treatment planning and its delivery [8,9]. The clinical assessment of the disease being treated should be classified and staged as recommended internationally, using systems such as the TNM or FIGO [11,12]. The assessment of response to treatment and morbidity of normal tissues should also follow accepted practice [10].
32.2 CLINICAL ASPECTS OF QUALITY MANAGEMENT
32.2.1
General considerations
Whereas the treatment methods and techniques used for teletherapy are generally more defined than those for brachytherapy, the latter still tends to follow established procedures, although these may be modified by the individual beliefs and habits of the clinician, more so than for teletherapy. The advent of optimization of treatment planning has also brought with it treatment which can be individualized to the patient's needs. The treatment intent decided for a particular patient will involve a core plan of management. A core plan of management will normally be determined by those who are clinically responsible, but may well involve an input from the physicist, radiographer, and nurse specialist. The treatment objective relative to radical (i.e., curative) intent or palliative intent needs to be established once the full investigative procedures have been completed. Following this, it is necessary to provide a clear description of the intended treatment, which would normally be followed by prescription. This includes identifying the planned target volume in anatomical terms, the treatment method in radiotherapy terms, together with the irradiation parameters such as the intended tumor dose either to point(s) or to a given isodose contour. The method used should be such that a meaningful relationship exists between the dose prescribed and the target volume. The target volume, as for teletherapy, will contain the tumor and should include some allowance for the possible invasion of local surrounding tissues. Also, the limiting dose to critical organs or tissues within or close to the treated volume needs stipulating, and that such doses have not been exceeded needs to be confirmed by either measurement or calculation. For further information, the reader is referred to Chapter 6. All dosimetric calculations and measurements should be available as part of the clinical record in a summary of
426 Quality management: clinical aspects
the treatment given. Whatever particular method is used, it should be easy to define, easy to understand, easy to use, and universally acceptable.
32.2.2
The patient
A quality assurance programme starts with the identification of those key functions which affect the management process of a patient's treatment, but it is also important to consider what the patient might expect of his or her management and treatment [13]. In particular, patients face a number of uncertainties relative to their treatment and outcome. Any reduction of these uncertainties could, from the patient's point of view, be considered as providing a high quality of care and reassurance. From the time of diagnosis, patients require fast access to the most appropriate clinician specializing in their type of cancer. They normally expect to be fully informed of their condition, the treatment available, the side-effects, and risks of possible complications. Having agreed to a particular treatment, they should receive it without delay [14]. They also expect to make a full and rapid recovery, with a high expectation of being cured. The treatment results should match those produced by leading cancer centers.
32.23
Pretreatment assessment
Pretreatment investigations should provide a clear description of the clinical extent of the cancer and the anatomical structures affected. Table 32.1, lists the various steps to carry out a pretreatment assessment. Staging should follow either the TNM or FIGO system. In the opinion of the author, the latter has proven the most practical for gynecological staging, whereas for all other sites the TNM system should normally be applied [13]. The methodology used is important and, where radio-
Table 32.1 Pretreatment assessment (1) Site-specific description (2) Stage of disease • TNM • FIGO
(3) Histology • Tumor type • Tumor grade (4) Mensuration • Clinical • Radiological (5) Patient profile • Hematological • Biochemical • Medical history • Surgical history
logical and surgical staging has been carried out in addition to clinical assessment, this should be recorded. Histological assessment of the tumor type and grade is preferably made according to the International Classification of Diseases (ICD) system. In addition to clinical assessment, surgical staging is now advocated by many for early disease. The treatment and morbidity implications of this are that, when radiotherapy follows surgical staging, the risks of developing morbidity have been reported to be increased. It is important in this type of situation to determine the possible advantages of surgical staging against non-surgical staging. Special investigations appropriate to the cancer site, such as cystoscopy or lung function tests, which provide a means of assessment of organ function should be done systematically and at set intervals during and following treatment. In addition to clinical assessment, radiological investigations, including imaging such as computerized tomography (CT) and magnetic resonance imaging (MRI) scans, should be carried out where indicated. Routine blood tests, including biochemical measurements, should also be done. To complete the pretreatment profile, a clinical history of any previous surgical operations or medical conditions which may affect the response and/or outcome to irradiation should be recorded. In turn, this may result in a decision to modify a standard dose prescription. Patients who come into one of these categories can then easily be identified for audit purposes.
32.2.4
Treatment intent
This is, in effect, a tentative management plan and should provide the basis of the intended treatment. It will include the type of brachytherapy and whether this is to be combined with external-beam irradiation. The intended dose regime, including dose rate, dose per fraction, total number of fractions, total dose, and overall treatment time, should follow established guidelines when possible whether the intention is to use either brachytherapy alone or combined with external-beam irradiation. The prescribed dose will need to be stated as applying to a specific point or points or isodose volume. This may involve some preliminary planning in terms of optimization of treatment. In many instances, the intention will be to follow a specific protocol or guidelines, and this should be stated.
32.2.5
The treatment plan
For most brachytherapy treatment, a physical outline of the anatomical site is not clinically possible and is increasingly becoming dependent on some form of imaging. However, when combined with external-beam
Clinical aspects of quality management 427
irradiation, a physical outline may be part of the planning procedure for the external-beam component, as for example, when treating breast, pelvic, or head and neck cancers. The steps involved are outlined in Table 32.2 [18,20]. Table 32,2 The treatment plan (1) Physical outline of normal anatomy (if indicated) (2) Physical limits of tumor (tumor volumefs]) (3) Identify critical organs
When combined brachytherapy and external-beam irradiation are to be used, it will usually mean that there are essentially two target volumes, one containing the primary tumor and allowance for possible local extension, the other containing either known or suspected secondary cancer in regional lymph nodes. The former volume will often receive the major portion of treatment from the intracavitary or interstitial component and the latter from external-beam irradiation [8,9]. The reader is referred to Chapter 6, which discusses the recommendations of ICRU reports 38 and 58. Three-dimensional planning, conformal treatment and its delivery require a comprehensive set of guidelines which are directly applicable to clinical treatment planning on a site-specific basis [21,24]. One such set of guidelines is that produced by Task Group 53 of the Radiation Therapy Committee and the American Association of Physics in Medicine. This report provides a framework for radiation physicists and clinicians to help design and perfect a comprehensive and practical treatment planning quality assurance system [21]. In some situations, such as breast implants, there is a high potential for a geographic miss, and a radiographic delineation of the tumor bed by intraoperatively placed clips has been used to enable delivery of a high dose to zones at risk of containing a high residual tumor load [20].
32.2*6
Tumor volume
That the physical limits of the tumor should be known is important, and clinical and, where indicated, imaging assessment(s) should be carried out. The different imaging methods carry different advantages, and scanning for treatment planning purposes will differ in terms of quality assurance and control compared to diagnostic applications [18]. When determining the physical limits of a primary tumor volume, it is also important to identify, in relation to that volume, the position of any critical organ(s) or tissue(s) which will, by necessity, be within or close to the target or treated volume [21]. Where secondary spread of cancer is known or suspected, as for example to regional lymph nodes, the physical limits of this 'secondary' tumour volume should be identified [18].
32*2.7
Target volume
The planning target volume contains those tissues that are to be irradiated to a specified absorbed dose according to a specified time-dose pattern. For curative treatment, the target volume consists of the demonstrated tumor(s), if present, and any other tissue with presumed tumor. The ICRU reports (38 and 58) advise that the target volume(s) must always be described independently of dose distribution in terms of the patient's anatomy, topography, and tumor volume [32].
32.2.8 The prescription and treatment procedure The prescription and treatment procedure will be dependent upon the site being treated and the technique used. In general, the prescription and treatment plan should provide the information outlined in Table 32.3. Treatment planning in many situations remains two dimensional, with reconstruction in one or more planes of the third plane. Three-dimensional reconstruction is now in increasing use and the advantages in identifying areas of potential underdose and overdose have been shown when planning treatment for nasopharyngeal cancer [24]. This has resulted in the further need to provide appropriate quality assurance guidelines [21]. Table 32.3 Provision of a prescription and treatment plan (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Outline of intended treatment Planned dose to point(s), or isodose level Details of actual delivery Confirmation of source position(s), (radiographic) Dosimetry assessment(s) to critical organ(s); define method Dosimetry calculations; record method Summary of treatment delivered including modifications Confirmation of completed treatment Further treatment intention(s) Follow-up plan
The prescription should describe the technique using simple, explicit terms. Where the prescription follows a standard protocol, reference to the protocol is sufficient except when departing from it. The intended total dose, dose rate, number of treatment fractions, and total time should be specified, relative to a reference point(s) or isodose level at which the prescribed dose is intended. The treatment record should also clearly identify those patients whose prescription is modified from a standard protocol and also the reasons, because this may affect the treatment result [32]. Using a standard approach should provide a meaningful treatment record for audit purposes. It will also help when preparing material for publishing and make it easier for others to follow and introduce similar
428 Quality management: clinical aspects
technique(s) and to compare their results with those published [19]. When a non-standard approach is used, clinical dose specification should include source geometry deviation from a standard dose specification system [31].
32*2*9 Confirmation of delivery of treatment An outline of the guidelines necessary to provide a prescription and treatment program is shown in Table 32.3. Confirmation, by calculation or measurement, of the dose received at the prescription point(s) or by critical organs or tissues forms a vital part of any quality assurance treatment program in brachytherapy. Even where source positions are controlled by a template system to provide a known dose distribution, variation in anatomical relationships may make it important to have an accurate assessment of the dose received by normal tissues at risk. In the author's opinion, there is no place in current brachytherapy to rely on using dose-rate tables based on a standard treatment set-up as applied for many years to intracavitary irradiation for treating cervix cancer and to interstitial therapy. The use of imaging techniques such as CT scanning can, since the advent of afterloading, be used in many circumstances for treatment planning and dose estimations without fear of excessive radiation exposure. Such dosimetry measurements and/or calculations, including the method used, should be documented and form part of the treatment record (ICRU Report No. 38,1985, and No. 58, 1997) [8,9,32]. When external-beam irradiation is part of the treatment process, it is necessary to be aware of possible geometrical errors during treatment delivery. These include errors that may produce distortion of treatment distribution. Such errors may be the result of deviation in the treatment set-up, such as field size, gantry angle, use of shielding blocks, couch sag, and patient movement. The use of regular check films was reviewed by Blanco et al. in 1987, since when, portal imaging devices now provide for greater accuracy of treatment delivery [25—27].
32.2.10
Treatment summary
A summary of the treatment delivered, including any modification(s) to the treatment procedure, should be recorded. Confirmation of the total prescribed dose(s), dose rate, fractionation, and overall treatment time, including dose(s) received by critical organs, will form an important record of each patient's treatment. This in turn will provide confirmation that all treatments of a similar type are delivered to within set limits of a predetermined treatment plan.
32.2.11
Treatment optimization
Full details of any optimization procedure carried out should be documented in order to be able to relate both tumor and morbid effects of treatment to the process of optimization. In particular, it is important to appreciate that a poor brachytherapy technique cannot be satisfactorily corrected by altering source position or dwell times under the guise of optimization. Thus, optimization should only be used to work within restricted limits, relative to a particular technique, and the optimization procedure should be recorded. Analysis of dose distribution using dose-volume histograms has been reported to be useful [35,36].
32.2.12
Effectiveness of treatment
In order to carry out an assessment of the effectiveness of treatment a number of important steps need to be taken as part of the treatment program. Most of these steps have already been discussed and include the important requirement that the aim of treatment is to achieve maximum tumor control with the minimum risk of producing normal-tissue complications. The process of assessment of response to and outcome from treatment can be considered under four separate headings: 1. local tumor control, 2. normal-tissue effects (a) acute (b) late, 3. time scale of follow-up, 4. further treatment. LOCAL TUMOR CONTROL
Assessment of response should be done by objective measurements if possible. This may be by clinical or radiological means. It is also important that mensuration should be carried out using a similar technique on each occasion and for all similarly treated patients. Also, measurements should be done in the same tissue planes throughout the monitoring period and should be carried out at similar time intervals from treatment for all patients. Where complete tumor response occurs, followed by local tumor recurrence, it is important to know the period of the tumor-free interval. Such records may prove helpful when carrying out an audit and help establish possible reasons for or better understanding of the reasons for failing to achieve satisfactory tumor control. NORMAL-TISSUE EFFECTS
It is important to record both the signs and symptoms associated with and indicative of acute and late-reacting tissue reactions. Often, the effects produced are subjective, and appropriate methods of grading such effects are diffi-
Administrative requirements 429
cult. Where objective clinical, radiological, or pathological investigations are carried out, it is possible to achieve consistency in terms of the reporting method used. Clearly, it is preferable that, as far as objective reporting is concerned, the person making the report should be conversant with the methodology and terminology and preferably should be a member of the treatment team. Only by carrying out careful assessment of late-reacting tissue reactions will it be possible to quantify dose-effect relationships in a meaningful way. In particular, such procedure provides a means of radiobiologically estimating the a/(3 ratios for late-reacting tissue reactions [ 10,29,30,33,34].
time. Signs and symptoms of normal-tissue effects can then be assessed in a consistent fashion, including the use of a suitable grading system. FURTHER TREATMENT
Further treatment may become necessary due to recurrent disease either within or outside the original primary target volume. Such treatment may involve further irradiation, chemotherapy, or surgery, and as such should form part of the documentary record of management and assessment. In particular, such treatment may affect normal-tissue response from previous radiotherapy or chemotherapy.
TIME SCALE OF FOLLOW-UP
There are two major purposes for the continued followup of patients: 1. to provide a complete record of the effects of treatment in terms of local tumor control within the target volume; 2. to provide a complete record of the effects of treatment on normal tissues and organs. For most normal tissues, late-reacting tissue changes may take several years to become manifest. It is therefore important to have in place a follow-up system which leads to the identification of radiation morbidity problems and allows such events to be graded according to severity. Some may question the value of regular follow-up of patients in either a radiotherapy or joint cancer clinic. They may argue that, should patients develop signs and symptoms of morbidity, they should be referred directly by the general practitioner to a consultant specializing in the treatment of the affected organ, e.g., bladder to urogenital surgeon, and bowel to gastrointestinal surgeon. Such events often occur between follow-up attendance at a cancer review clinic and it is, unfortunately, not standard practice for the treating clinician or the general practitioner always to keep the clinical oncologist informed. However, by seeing patients regularly, clinical oncologists are more likely to learn of any treatments carried out for normal-tissue damage from either the consultant concerned or directly from the patient, following which they may obtain more specific details. In the case of a patient developing recurrent local disease, referral back to the consultant who treated the patient is often done by the general practitioner, although in the case of secondary spread of cancer patients may be referred elsewhere, depending upon the organs affected. It is suggested, therefore, that if a full and accurate record of a patient's history following treatment is to be maintained, the most appropriate person to do this is the clinical oncologist. This may incur carrying out an annual postal follow-up to the general practitioner using a standard questionnaire. In general, follow-up records should follow a consistent reporting pattern, carried out at set intervals of
323
ADMINISTRATIVE REQUIREMENTS
Table 32.4 sets out a list of requirements which will require careful consideration if a quality management program is to be assured of a reasonable chance of success. Although the requirements as set out were intended to apply to an external-beam radiotherapy service, they are equally applicable to brachytherapy.
323*1
Motivation and training
Motivation and training are an essential part of any quality assurance program and appropriate supportive Table 32.4 Staff and administrative requirements (1) Motivation Clear understanding of: • Objectives • Individual roles • Working inter-relationships (2) Trained staff • Clinical • Scientific • Nursing • Paramedical (3) Time Identifiable time for: • Documentation • Check procedures • Analysis (4) Resources • Therapy and simulation equipment • Treatment planning equipment • Monitoring equipment • Staff availability • Record system (5) Communication • Documented procedures • Consistent and standardized records • Clear definitions of mensuration and end-points • Standard codes of practice
430 Quality management: clinical aspects
training courses should be provided. An integrated interdisciplinary approach with full cooperation of all members of staff, each of whom has a clearly identified role and understands and supports the objectives of quality assurance, is necessary. Table 32.4 identifies the individual roles of the various members of an interdisciplinary team, each member being responsible for carrying out a particular function. While this may provide considerable specialized expertise relative to a specific procedure that includes an understanding of the appropriate quality assurance procedure, that professional knowledge needs to be extended to being conversant with the procedural and quality assurance aspects of the other team members. To some extent, this can be achieved by providing an appropriate job description for each staff member.
Included in the job description, the responsibility within a general quality assurance program should be defined. The precise role for each member of staff will vary, but the composite integration of all individual roles should be such as to provide a balanced quality assurance program. A useful publication applicable to high dose-rate brachytherapy is the report of the AAPM Radiation Therapy Committee Task Group 59 [22]. The coordination of a quality assurance program may require a designated quality assurance manager, responsible to the head of department. To provide quality assurance requires dedication and invariably increases the workload for staff. The requirements outlined in Tables 32.4 and 32.5 clearly show the diversity of responsibility necessary for individual staff
Table 32.5 Staff functions in brachytherapy
Diagnosis
Site, stage, and grade of disease
Surgeon Pathologist Radiotherapist
Treatment decision
Best current management
Interdisciplinary clinical team
Tumor localization (CT, radiology, simulation, clinical assessment)
Anatomical relationships Target volume limits Critical normal structures within treatment limits
Diagnostic radiologist Radiotherapist
Equipment calibration and dose monitoring
Meeting code of practice requirements
Physicist Technician
Patient safety
Equipment failure Fail-safe situations Patient monitoring and communication
Physicist Technician Radiographer Radiotherapist
Treatment plan
Determine form of brachytherapy and whether with or without teletherapy Dose-time plan
Radiotherapist Physicist
Treatment procedure (modifications with reasons)
Placement of source and source carrier system etc. Simulation Computation of dose distribution Dose assessments to critical organs etc. Films for verification/dose calculation Confirmation of dose-time prescription Repositioning source carriers as necessary
Radiotherapist Physicist Radiographer
Treatment
Dose-time calculations for treatment Dosimetry checks and treatment monitoring Confirm completion of treatment
Radiotherapist Radiographer Physicist
Personal safety
Appropriate shielded room or facility Personal monitoring Interlock systems Radiation monitoring of treatment area Wearing film badges when required
Physicist Technician Radiographer
Effect of treatment
Tumor response Acute reactions Follow-up arrangement
Radiotherapist Nurses
Supportive care (follow-up clinic)
Tumor control Late morbidity Patient's quality of life
Radiotherapist Nurses General practitioner
Evaluation of treatment method
Clinical audit
Clinical team
References 431
members. The risk is that the workload may result in a 'rule of thumb' approach for expediency that is not appropriate for quality assurance. Having the necessary time made available in order to carry out the various quality check procedures is essential. Unfortunately, these requirements may be unappreciated by administrators and it is essential they have an identified and responsible role in any quality assurance program. Preferably, that role should include being a signatory to the aims and objectives of the quality assurance programme [16,17].
323.2
Resource requirements
Appropriate back-up resources are essential to apply quality assurance. In particular, appropriate and current state of the art equipment is essential for dose measurement, calibration, and dose monitoring. An appropriate, fully documented log for all forms of monitoring measurements is essential. An often neglected aspect of resource provision is the increased workload placed on clerical and secretarial staff. The amount of work entailed can be quite considerable and appropriate facilities should be provided.
3233
Communications
Communications are vital and, to ensure that they are meaningful, it is important that all documented procedures are consistent and standardized. To promote a better understanding of the recorded data, the definition of any end-points used and of any associated measuring methods is important. Wherever a recommended national code of treatment practice exists, it should be followed in order to standardize treatment and provide a common database for cooperating centers to exchange information.
32.4
development of new procedures and techniques. Clearly, it is important that research and development for better treatment should not be stifled, but should be carried out under agreed clinical phase 1, 2, or 3 protocols [7,9]. Introducing a quality assurance system is a complex problem because of the complexity of the various sitespecific treatment methodologies. One proposed method of providing a national and multinational mechanism is that of institutional certification through a mechanism of peer review [23,28]. A summary of the staff functions from the point of diagnosis to evaluation of the treatment method used is provided in Table 32.5.
CONCLUSIONS
The primary purpose of a quality assurance program in brachytherapy will be similar to that for teletherapy and medical practice in general. It should provide: 1. for maintenance of an established practice, 2. the means for assessing compliance in carrying out a management procedure, 3. the means for monitoring and evaluation of treatment using hard data, and 4. the means for determining weaknesses in a treatment program to further the development of 'best current management procedures.' A possible disadvantage is that a strictly enforced quality assurance program may stifle initiative and the
REFERENCES 1. Quality assurance in radiation therapy: clinical and physical aspects. Proceedings of the First International Symposium on Quality Assessment in Radiation Oncology, Washington DC, 8-10 June (1983). Int.J. Radial Oncol. BiolPhys., 10, Suppl. 1 (1984). 2. Quality Assurance in Radiotherapy. (1988) A guide prepared following a workshop, organized jointly by the Institute of Radioactive Hygiene, Federal Health Office, Neuherberg, Germany, and the World Health Organisation. Geneva, WHO. 3. Johansson, K.A., Hanson, W.F. and Horiot, J.C. (1988) Meeting report: Workshop of the EORTC Radiotherapy Group on quality assurance in co-operative trials of radiotherapy: a recommendation for EORTC Co-operative Groups. Radiother. Oncol., 11,201-3. 4. Griffiths, S.E., Pearcey, R.G. and Thorogood, J. (1987) Quality control in radiotherapy. The reduction of field placement errors. Int.J. Radial Oncol. Biol Phys., 135, 1583-8. 5. Griffiths, S.E. (1986) Reproducibility in radiotherapy. Radiography, 52,167-9. 6. Granville-Wright, M. (1987) Quality assurance errors in radiotherapy: an overview. Radiography, 53,135. 7. Wilson, J.F., Chassagne, D. and Joslin, C.A.F. (1988) Brachytherapy trials. Int.J. Radial Oncol. BiolPhys., 14, S57-S63. 8. ICRU (1985) Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology, ICRU Report No. 38. Bethesda, ICRU. 9. ICRU (1997) Dose and Volume Specification for Reporting Interstitial Therapy, ICRU Report No. 58. Bethesda, ICRU. 10. WHO Handbook for Reporting Results of Cancer Treatment (1979) Geneva, WHO. 11. Hermanek, P. and Sobin, LH. (eds.) (1987) TNM Classification of Malignant Tumours, 4th (fully revised) edition. Berlin, New York, Springer-Verlag. 12. FIGO News (1987) Carcinoma of the cervix staging. InLJ. Gynaecol. Obstel, 25,87. 13. Young, J. (1987) Radiotherapy, the quality assurance of patient management. Radiography, 52, 57-63.
432 Quality management: clinical aspects 14. Kovacs, G. (1998) Medical aspects of quality assurance in brachytherapy. Strahleuther. Onkol., 174 (Suppl. 2), 47-9. 15. Farfus, B., Rons, F., S'anchez-Reyes, A., Ferrer, F., Rovirosa, A. and Biete, A. (1996) Quality assurance of interstitial brachytherapy techniques in lip cancer: comparison of actual performance with the Paris System recommendations. Radiother. Oncol., 38(2), 145-51. 16. Kubo, H.D., Glasgow, G.P., Pethel,T.D.,Thomadsen, B.R. and Williamson, J.F. (1998) High dose-rate brachytherapy treatment delivery: report of AAPM Radiation Therapy Committee Task Group No. 59. Med. Phys., 25(4), 375-403. 17. Purdy.J.A. and Perez, C.A. (1996) Quality assurance in radiation oncology in the United States. Rays, 21(4), 505-40. 18. Smaniotto, D., Mantello, G. and Valentini, V. (1996) Quality Assurance program in radiotherapy for carcinoma of the uterine cervix. Rays, 21(4), 641-8. 19. Brundage, M.D., Dixon, P.F., Machillop, W.J. et al. (1999) A real-time audit of radiation therapy in a regional cancer centre. Int.J. Radial Oncol. Biol Phys., 43(1), 115-24. 20. Sedlmayer, F., Rahim, H.B., Kogelnik, H.D. et al. (1996) Quality assurance in breast cancer brachytherapy: geographical miss in the interstitial boost treatment of the tumour bed. Int.J. Radial Oncol. Biol Phys., 34(5), 1133-9. 21. Fraass, B., Doppke, K., Hunt, M. et al. (1998) American Association of Physicists in Medicine Radiation Therapy Committee Task Group 53: quality assurance for clinical radiotherapy treatment planning. Med. Phys., 25(10), 1773-829. 22. Kubo, H.D., Glasgow, G.P., Pethel, T.D., Thomadsen, B.R. and Williamson, J.F. (1998). High dose rate brachytherapy treatment delivery: report of the AAPM Radiation Therapy Committee Task Group No. 59. Med. Phys., 25(4), 375-403. 23. Vitale, V., Buconte, G., Foppiano, F. et al. (1998) Introducing quality assurance in radiotherapy. Tumori, 84(2), 101-3. 24. Leung, T.W., Wong, V.Y., Tung, S.Y. et al. (1997) The importance of three dimensional treatment planning for nasopharyngeal carcinoma. Clin. Oncol., 9(1), 35-40. 25. Blanco, S., Lopez-Bote, M.A., Desco, M. (1987) Quality
26. 27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
assurance in radiation therapy. Systematic evaluation of errors during the treatment execution. Radiother. Oncol., 8,256-61. Thomason, C.C. (1998). Implementation and clinical use of portal imaging. Cancer Treat. Res., 93,69-99. Griffiths, S.E., Khoury, G.G. and Eddy, A. (1991) Quality control of radiotherapy during pelvic irradiation. Radiother. Oncol., 20,203-6. Reinstein, L.E., Peachey, S., Laurie, F. and Glicksman, A.S. (1985) Impact of dosimetry review program on radiotherapy in group trials. Int. J. Radial Oncol. Biol Phys.,10,1179-84. Emami, B., LymanJ., Brown, A. et al. (1991) Tolerance of normal tissue to therapeutic irradiation. Int.J. Radial Oncol. Biol Phys., 21,109-22. Visser, A.S. (1989) An intercomparison of the accuracy of computer planning systems for radiotherapy. Radiother. Oncol., 15,245-58. Eifel, P.J., Monaghan, J., Owen, J., Katz, A., Mahon, I. and Hanks, G.E. (1999) Patterns of radiotherapy practice for patients with squamous carcinoma of the uterine cervix: patterns of care study. InlJ. Radial Oncol. Biol Phys., 43(2), 351-8. Wambersie, A., Chassagne, D., Dutreix, A. et al. (1996) Quality assurance in brachytherapy: the role of the ICRU in achieving uniformity in dose and volume specification for reporting. Rays, 21(4), 541-58. Dische, S., Warburton, M.F., Jones, D. and Lartigau, E. (1989) The recording of morbidity related to radiotherapy. Radiother. Oncol., 16,103-8. Rodrigus, P., De Winter, K., Leers, W.H. and Kock, H.C.L.V. (1996) Late radiotherapeutic morbidity in patients with carcinoma of the uterine cervix: the application of the French-Italian glossary. Radiother. Oncol., 40,153-7. Vicini, F.A., Kestin, LL, Edmundson, G.K. et al. (1999) Dose-volume analysis for quality assurance of interstitial brachytherapy for breast cancer. InlJ. Radial Oncol. Biol Phys., 45(3), 803-10. Joslin, C.A. (1996) Commentary: Brachytherapy-dinical dosimetry and the integration of therapies in gynaecological cancer. (Also, proffered abstracts from the meeting.) Br.J. Radial., 69, 578-80 and 689-92.
33 Safe practice and prevention of accidents in afterloading brachytherapy A FLYNN, S.E. GRIFFITHS, AND C.A. JOSLIN
33.1
INTRODUCTION
In this chapter, some of the safety issues of the provision of afterloading brachytherapy are considered. The provision of an afterloading brachytherapy service must take full account of essential requirements for (a) the safe and accurate delivery of the treatment as prescribed, (b) the provision of a safe working environment for the staff of the facility, and (c) the fulfilling of statutory and environmental considerations relating to the receipt, safe custody, and disposal of radioactive materials. The requirements of both manual and remote afterloading are considered. However, most of the chapter deals with remote afterloading. The issues relating particularly to manual afterloading are considered in section 33.14. Safe practice in brachytherapy is dependent on many factors. In particular, when using remote afterloading, it is dependent to a large extent on safety features built into the hardware and software associated with the installation. However, it is also especially important to have
highly trained specialist staff, working according to rigorous protocols and systems of work. These issues are considered in detail in this chapter. Reference is made, where appropriate, to UK, European and North American legislation and codes on safety matters. In the UK, the then Institute of Physical Sciences in Medicine (now IPEM) has published a report on radiation protection in radiotherapy which includes a section on brachytherapy afterloading [ 1 ], and an excellent review of remote afterloading issues has been published by the American Association of Physicists in Medicine [2]. Other countries have their own regulations and codes of practice, but the general principles will be similar.
33.2
EQUIPMENT DESIGN
The first step in the provision of a safe installation occurs at the pre-purchase evaluation of afterloading equipment. It is at this stage that the safety features of competing equipment must be evaluated, as it is difficult or
434 Safe practice and prevention of accidents in afterloading brachytherapy
impossible to rectify inherent design faults at a later stage. There may be practical difficulties in this, as the equipment will not be on site at this stage, but a thorough examination should be made of the manufacturers' literature, technical specifications, and operating instructions. Other installations of a similar type should be visited if possible, and the experience and independent opinions of other users of the equipment should be sought and considered. Many design features are mandatory in many countries and, indeed, licensing by national authorities or statutory bodies is generally a requirement. For example, installations within the UK are required to comply with BS 5724 Section 2.17 [3], which is identical to IEC 6012-17, both of which give detailed requirements regarding the design of remotely operated afterloading equipment. In the USA, installations have to be licensed either with an Agreement State Agency or with the US Nuclear Regulatory Commission, and equipment not on the Registry of Sealed Sources and Devices cannot be licensed [4]. In Canada, treatment programs and the use of sealed sources have to be licensed by the Canadian Nuclear Safety Commission. In any country, it is the responsibility of the user to ensure that equipment complies with the appropriate regulatory requirements. The user interface should be designed so that programming the treatment, operating the unit, and checking its operation should be as simple and unambiguous as possible. For example, there should be a display which is operational at all times when the machine is switched on and which shows the current status of the equipment and its programmed treatment. This enables a comparison with the required treatment data from the prescription sheet. There must also be a printout from the machine to provide a permanent record of all data entries and source exposures. A facility for the direct transfer of data from the treatment planning system to the afterloader reduces the likelihood of transcription errors, although the data should still be manually checked against the original prescription. Equipment design should incorporate 'fail-safe' features to prevent exposure of the radioactive source (or sources) if a fault occurs with any part of the equipment. These safety systems may be designed into the hardware, software, or, more likely a combination of the two. For example, a problem such as a disconnected or damaged applicator or transfer tube, excessive applicator curvature etc., should be detected by the equipment before a source transfer is initiated. This is often done by an air pressure check in the case of pneumatically driven sources or by a check cable in the case of a mechanically driven source. It should be made impossible to initiate an exposure unless data for a complete treatment (or fraction of treatment if part of a fractionated course or pulsed treatment) have been entered, except under 'service mode' conditions, as described below.
The equipment should check as far as possible that a treatment has been given correctly. For example, the source(s) position should be maintained within a defined tolerance (± 2 mm is recommended in BS 5724 Section 2.17 [3]), the treatment exposure time should be controlled by one timer and checked by an independent back-up timer, and the return of the source(s) to the safe at the end of a programmed exposure should be confirmed. Any detected malfunction during an exposure should result in the immediate return of the source (s) to the safe and the generation of an appropriate warning. Preferably, the equipment should display a code to indicate the nature and location of the fault. Any interruption of a treatment should be indicated: this is especially important for low dose-rate (LDR) treatments, for which an extended delay could affect the radiobiological effectiveness of the treatment. The equipment should have a back-up power supply to provide, in the event of a power supply failure, first, the return of the source(s) to the safe and, second, the retention of data relevant to the treatment. A useful feature is to have the facility to restrict certain activities with the equipment by means of a 'service mode' option. With this feature, procedures such as source data entry, source removal and replacement, and other non-'day-to-day' operations can be restricted to certain qualified personnel. Often, a two-key arrangement is used: one key enabling normal use of the machine, and the second key being held by the person authorized to perform these special functions.
333
ROOM DESIGN
Once the supplier and type of equipment have been decided upon, it is necessary to consider the design of the room in which the equipment is to be housed. The type of room will depend on whether high dose-rate (HDR) or low dose-rate (LDR) equipment is to be installed. The ideal situation is one in which a treatment room is designed specifically as a new installation for a particular piece of equipment. Sadly, this is rarely the case, and frequently new afterloading equipment is installed into a pre-existing room with appropriate modifications. This may, in turn, lead to restrictions in the manner in which the equipment may be used. For example, sources driven out from a mobile afterloading unit or through flexible transfer tubes may have to be kept within a specified area within the room in order to maintain adequate radiation protection. In some circumstances, it may be necessary to install two afterloaders into one treatment room. If one or both of these is an HDR unit, interlock systems must be installed to prevent the use of both machines simultaneously. If the two afterloaders are both LDR units, it may be desirable, for patient scheduling and logistical rea-
Treatment planning 435
sons, in exceptional circumstances to treat with both machines at the same time. If this is the case, a rigorous system of work must be developed and adhered to in order to minimize any radiation hazard. A.M. Bidmead considers the design of brachytherapy rooms in detail in Chapter 10.
33.4
INSTALLATION
The installation of complex equipment such as an afterloading machine is normally the responsibility of the supplier, this being part of the purchase agreement. However, it is the responsibility of the user to: 1. agree an installation plan with the supplier; 2. arrange for appropriate utilities and services to be provided to the treatment room; 3. perform acceptance testing on the installation prior to a formal hand-over; 4. perform appropriate post-installation commissioning tests, source(s) calibration, and provision of treatment data for clinical use. E.D. Slessinger considers these aspects in detail in Chapter 8.
33.5
TREATMENT PROTOCOLS
When the conditions and sites to be treated have been decided, written procedures should be documented. These documents should describe in detail the methodology to be used. The chance of errors occurring is much reduced if 'one-off' treatments are avoided when possible and when all treatments are given according to a documented protocol. Each treatment site and method of application should have its own protocol. The scope of each protocol should include localization procedures on imaging equipment, including a description of any radiographic markers, operating room procedures, treatment planning method, treatment delivery, and verification and recording procedures. The documentation should be clear and unambiguous. Each protocol should be written with the participation of all the staff groups concerned. The protocols should indicate the various responsibilities of the different staff groups. Particular care is needed to elaborate on the interface between the various personnel, to avoid misunderstandings, errors of omission, or errors in the transmission of information from one person to another. An example of a summary protocol is shown in the Appendix (p. 441), which would be used in conjunction with detailed work instructions. The latter are not reproduced here for reasons of space.
33.6
TREATMENT PRESCRIPTION
The treatment prescription should be clear, complete, and unambiguous. It should specify exactly the treatment machine and technique to be used. Diagrams should be used routinely and particularly where they aid clarity. The type and size of applicators, radiation sources, prescribed radiation dose and its location should be clearly indicated. The prescription should be clearly and legibly signed and dated by the prescribing clinician. The treatment prescription sheet should contain unambiguous information regarding the treated region, such as left or right side of the patient, and any patient and applicator position details. For example, intraluminal applicators should have their length of insertion described, gynecological applicators should be described to permit reproducibility for fractionated treatments and to allow correct reconstruction for dosimetric purposes. The prescription sheet should also state the dose given to critical organs. With HDR treatments, it is often necessary to produce a treatment prescription in a short time, as the patient may be intubated and anesthetized whilst awaiting the start of the treatment. This is facilitated if data providing information on the current source strength, radioactive decay tables, and treatment times are readily available at the treatment machine. Errors are most likely to occur in circumstances in which staff are working under pressure to provide treatment data and calculations quickly. It is therefore recommended that anesthetized patients undergoing HDR brachytherapy be treated with standard treatment plans when possible, for which the data should be readily available. Such standard treatments should, of course, be fully described in the protocols. Whereas the primary purpose of the treatment prescription is to provide instructions for the person(s) giving the treatment, it is customary for it also to serve as a permanent record of the treatment given. The International Commission for Radiological Units (ICRU) has published recommendations for the reporting of intracavitary therapy in ICRU 38 [5], and interstitial therapy in ICRU 58 [6]. It is imperative that the recommendations contained in these documents be followed in the recording and reporting of brachytherapy treatments. A. Wambersie and J.J. Battermann cover this subject in detail in Chapter 6.
33.7
TREATMENT PLANNING
Once the clinician has decided on the appropriate form of treatment for a specific patient, the planning process may conveniently be divided into three phases.
436 Safe practice and prevention of accidents in afterloading brachytherapy
33*7.1 Transfer of information to the planning staff If an individualized treatment plan is required, the clinical requirements of the plan (treatment volume, prescribed dose, critical organ constraints, dose fractionation, etc.) must be transmitted unambiguously to the planning physicist or technician. A convenient and safe way to do this is to make use of the prescription sheet. It is worthwhile spending some time and effort on the design of prescription sheets for the various applications so that the appropriate information may be easily and clearly written. It is unsafe to rely on verbal information: all information should be written and its authorship should be clear.
33.7.2 plan
The production of the treatment
The final accuracy of the treatment plan depends on the quality of the physical data used for its production. It is important to ensure that the origins of the data for particular types of radiation source, individual sources, and other physical parameters are properly researched, well documented, and have proven accuracy. The origin of 'in-house'-produced tables and graphs should be referenced and documented. There are many software packages commercially available for the calculation of brachytherapy treatments. Chapter 5, by R. van der Laarse and R.W. Luthmann, considers in detail some of the computer algorithms and software used for this purpose. Although in a busy department it may be tempting to assume that the commercial software writers have got it right, it is, of course, ultimately the responsibility of the user to be certain that the algorithms operate correctly and that the values of the various physical constants and other parameters used by the software are correct and used appropriately. Suitable dose-checking procedures by manual calculation and measurement should be carried out to put the computed treatment to a practical test. It is not intended that this chapter should describe a complete system for software evaluation, rather that it should emphasize the importance of performing such checks. The user should check the accuracy of the results from a treatment planning system before putting new software into clinical use. For simple source arrangements, this may readily be done by comparison with manual calculation and experimental data. Although tedious, the checking of more complex treatment arrangements should also be done. A program of regular quality assurance checks on the planning procedure should be devised, performed, and recorded. This should include a daily check calculation of a simple arrangement of sources, to check the correctness of the basic source data and calculation method.
More rigorous checks, to include a test of the reconstruction algorithm, should be done at, say, monthly intervals. The reader is referred to reference 7 for further guidance on this topic. Treatment planning computers use a variety of calculation and reconstruction algorithms. It is incumbent on those responsible for the production of treatment plans to know and understand the computational processes involved, particularly with respect to the various correction and conversion factors used in the software. Also, there may be mathematical limits to some of the functions, for example the Meisberger function [8] for tissue attenuation and scatter corrections, which is only valid to a distance of 70 mm from the source. The suppliers of these systems must make information on such limits in the computation algorithms available in the accompanying manuals. Although systems supplied by reputable manufacturers will have been rigorously tested before being offered for clinical use, there may still be undetected software 'bugs' present and the user must be alert to this. If an anomalous result is detected at any stage in the process, the supplier should be notified immediately so that the software can be corrected and other users notified. It follows from the above that treatment planning systems must always be used with caution and that the final plan must be scrutinized for the presence of errors. Each plan must be individually checked. Preferably, a written checklist should be followed and a record of the check should be filed. Where practicable, a manual calculation of part or the entire plan should be done for comparison with the computed version. Ideally, a second qualified person should do the checking of a treatment plan, independently of the original calculation.
33.7.3 Transfer of information to the treatment unit The completed treatment plan contains the data necessary for the correct treatment of the patient. This information must be transmitted to, and be correctly interpreted by, the people responsible for delivering the treatment. Conventionally, this information is transferred to the prescription sheet from the treatment plan. Transcription errors may be reduced, particularly for complex treatments, if the treatment planning computer can print the information directly onto the prescription sheet. Alternatively, the planning computer printout may be attached to the prescription sheet: in this case, it must be clear that the computer printout refers to the correct patient, whose name and identification must be prominently displayed. As already mentioned, care must be taken in the design of prescription sheets and computer printouts so that the information written thereon is clear and unambiguous. Some types of equipment allow direct transfer of information from the planning computer to the treat-
Treatment delivery 437
ment unit. This may be either by a direct connection between the two components, or by the use of a disk or program card. The main advantage of this arrangement is that it reduces the possibility of transcription errors during programming of the treatment unit. This is particularly the case for a multichannel treatment using a stepping source, for which manual entry of the data may be tedious and prone to error. The correct operation of the data transfer should be checked as part of the treatment planning computer and treatment unit quality assurance programs, and at each clinical use an appropriate person should check the data printout. The remarks in section 33.7.1 also apply here: it is unwise to rely on the verbal transmission of data, and all printouts and written information should be signed by the person originating them.
33.8 33.8.1
PRETREATMENT CHECKS Equipment checks
There are certain checks that must be performed on the treatment machine immediately prior to its use for a patient. These form part of the wider quality assurance program, which is alluded to in section 33.10 of this chapter and is dealt with more fully in Chapters 8 and 9, by E.D. Slessinger and C.H. Jones, respectively. All of these procedures form part of the 'system of work' for the radiotherapy department and should be written down as part of a 'quality manual' or equivalent document for the Radiotherapy Institute. The exact nature of the pretreatment checks depends on the type of equipment, but at a minimum should include a check on the source(s) positioning accuracy in the applicator(s) and the completion of a short simulated treatment. A check of the source data activity for the date of use should be done; this could be a mathematical check of stored data, for example, rather than a full calibration check, although the latter should also be done at specified regular intervals. It is advisable to perform the machine checks before the patient undergoes any operative procedure associated with the treatment, so that a machine fault that could prevent treatment will show itself before the start of anesthesia induction or invasive procedure. As with the rest of the quality assurance program, appropriate documentation of the pretreatment checks must be kept. This should include a description of the tests and the expected results, together with acceptable limits on accuracy. The performance of the tests should be documented, and the record signed by the person performing the test. In a busy department performing many treatments each day, it will be convenient to perform these checks once daily at the beginning of the working day.
33.8.2
Procedural checks
The correct identification of the patient must be confirmed and recorded at each treatment occasion. For a conscious outpatient, this may be done by asking the patient to state his or her name and address. An inpatient will have an identification tag, which must be checked. A remote machine afterloading treatment requires the presence of at least two radiographers (radiation technologists) to permit independent checking of the treatment parameters. After one radiographer has programmed the treatment unit, either by manual data input or direct transfer from the planning system, the received data must be checked by the second radiographer. The connection of the treatment applicators to the transfer tubes may be done either by a clinician or by a radiographer: in either case, it should be checked by the second qualified person. Where possible, radiographic verification of the position of the applicators in the patient should be performed. Particular care should be taken to ensure that the correct transfer tube is connected to each applicator. Some machine systems have a mechanical code to ensure that this is done correctly, but this is not always the case, particularly with interstitial implant applicators. Use should be made, whenever possible, of standard applicator arrangements and source loading patterns. These standard arrangements must be documented and copies kept available at the treatment unit control. If the system allows, they should be stored in the equipment memory as 'standard treatments.' If any subsequent manipulation or recalculation of the standards is required (for example changing the dose from the standard value to a non-standard), the second radiographer should independently check these. In the case of a multifraction treatment, any changes made to the prescription as the treatment progresses should be clearly documented, signed and dated, and brought to the attention of all concerned.
33.9
TREATMENT DELIVERY
It is usual, in the case of machine afterloading, for the treatment to proceed automatically under computer control once the source(s) transfer has been initiated. However, the progress of the treatment delivery must be monitored. In the case of HDR treatments, the treatment console must be monitored throughout the treatment session so that any malfunction or interruption may be dealt with immediately. The patient should be observed via a closed-circuit television system. For conscious patients, a voice intercom system should be available to allow two-way conversation. In the case of an anesthetized patient, the anesthetist will require monitoring of vital functions, and this may be achieved using remote
438 Safe practice and prevention of accidents in afterloading brachytherapy
displays and television systems. Circumstances may necessitate the interruption of treatment, and the equipment must be designed so that, in this event, it retains appropriate information relating to the state of treatment and the remaining treatment time(s). The accuracy of the information retained under these circumstances should be checked as part of the routine quality assurance procedures. Such checks should also be done following a power failure to the equipment. Similar principles apply to LDR machines, but in this case it is not appropriate to observe the treatment control continuously. Generally, the treatment will be allowed to proceed to completion with appropriate interruptions for nursing observations and care. However, a routine check of the placement of the treatment applicators should be performed at intervals throughout the treatment: every 2 hours is suggested, but the frequency of checks may be reduced during the night if the patient remains asleep. The equipment should be designed to detect and signal any malfunction for which intervention may be required. The situation for pulsed dose-rate (PDR) equipment is similar to that of LDR equipment. One difference is that, if the condition of the patient permits, he or she may be disconnected from the treatment unit between pulses. There should be a system for disconnecting and reconnecting the transfer tubes in a manner which minimizes the possibility of reconnection errors. One suitable way is to use a device similar to that provided with the Nucletron microSelectron-PDR, which permits several transfer tubes to be disconnected or reconnected simultaneously and ensures that transfer tubes and applicators do not become mismatched. Treatments on LDR and PDR systems generally take longer than the normal working day, and arrangements to provide 'out-of-hours' cover will be required. The nursing staff, who will be in attendance throughout the treatment, should be trained to provide first-line cover for interruptions and alarms, and they should also know how to obtain rapid help in the case of a radiation emergency. However, there may be occasions when a malfunction will require the presence of a physicist and/or radiographer, and an 'on-call' arrangement will be required. The staffing requirements (section 33.12) should take into account that this cover will have to be provided on a rota basis, and there should be sufficient staff to provide this service. It should also be borne in mind that, if the equipment needs to be reprogrammed after an interruption, two qualified people will be required so that they may check each other.
33.10
MACHINE QUALITY ASSURANCE
Each treatment unit must have a fully documented quality assurance programme, which should consist of a
series of specific tests and repetition schedules with appropriate records. The reader is referred to Chapter 8 by E.D. Slessinger for LDR machines units and to Chapter 9 by C.H. Jones for HDR and PDR units.
33.11
SERVICING AND MAINTENANCE
Radiotherapy equipment, in common with any other piece of electrical and mechanical apparatus, will perform reliably only if it is serviced and maintained on a regular basis. It is imperative, therefore, that appropriate arrangements for servicing are made. For each piece of equipment, the user has to decide between servicing 'in-house', using the institute's own staff, or the purchase of a service contract. The appropriate decision will depend on local conditions and no specific recommendation can be made here. A service contract from an equipment supplier entails a large annual outlay, typically up to about 10% of the capital cost of the equipment, but this usually has the advantage of providing regular service visits and provision of replacement parts when necessary. It will also usually provide an 'on-call' service for breakdowns. Tn-house' servicing will appear to be less expensive, but this must be offset against the cost of the salaries and training requirements of the staff performing the service, perhaps the provision of extra staff to provide breakdown cover, and the purchase of spare parts. The last mentioned may be a particular difficulty, as parts may be expensive and they may need to be purchased at short notice, and contingency funds will have to be kept aside for this eventuality. In any event, servicing and maintenance schedules must be drawn up and adhered to. If the servicing is to be done 'in-house,' advice must be taken from the supplier regarding the frequency of services, replacement of certain items, and the items to be checked on a regular basis. For machines using short half-life radiation sources, a regular program of source exchanges is required. For example, iridium-192 sources (half-life 74 days) used in afterloading machines are generally replaced at 3-monthly intervals. A source calibration by the user is required after a source exchange; Chapter 3, by C.H. Jones, considers this aspect in more detail. It is recommended that quality assurance checks be performed after a period of equipment servicing. It is helpful to design the quality assurance program around the servicing requirements so that the appropriate items and operations are checked following adjustment or replacement. This also has the advantage of reducing the machine downtime, allowing servicing, source replacements, and quality assurance sessions to be coordinated. The importance of maintaining a service logbook that clearly identifies all servicing and faults occurring on any
Manual afterloading 439
one individual machine is essential. Not only does this provide a record of the reliability of the unit, but it also provides information about repetitive faults and can provide a basis for feedback to the manufacturers. It is recommended that the logbook should contain a record of every 'event' on the equipment, including treatment sessions, quality assurance, servicing, calibrations, equipment malfunctions, etc. In this way, the user has on hand a history of the use of the equipment and any work and modifications done to it.
33.12
STAFFING CRITERIA
There must be adequate staff of all professional groups to allow safe provision of a brachytherapy service. For some groups of staff, there are recommended minimum staffing levels for certain functions. For example, Table 33.1 shows the UK IPSM (now IPEM) recommendations for medical physicists [9], but it should be noted that these are for the service provision only, and do not allow for other functions such as education, training, research, and absences due to leave. Such absences need to be taken into account when planning a facility. A minimum of two experienced radiographers/radiation technologists is required on an afterloading treatment machine in order to provide the necessary checks of each part of the treatment process, and to be able to manage an emergency situation. As with physicists, the department as a whole must have more than this minimum, to allow for leave absences and training. The number of nursing and medical staff will depend on the particular circumstances pertaining, for example the number and type of afterloaders, the amount of manual afterloading, working patterns, the sites treated, the extent to which operating room procedures are used, and whether overnight and weekend working is utilized. There is a strong argument in favor of the concept of a 'brachytherapy suite' in which members of staff from each profession specialize in the practice of brachytherapy, preferably utilizing a specially designated geograph-
Table 33.1 Institute of Physical Sciences in Medicine minimum physicist staffing levels for brachytherapy
Major item (e.g., HDR unit, planning system) Minor item (e.g., LDR unit) Each 100 new patients per year 3
0.4 0.2 0.2a
Should be increased if a high proportion involve complex techniques such as iridium-192 wires or iodine-125 implants. These are for service provision only, and do not include training, research, etc. [9].
ical area within the hospital. This leads to the development of a team approach, with each member of the team being aware of the range of his or her responsibilities and those of other members of the team. Many brachytherapy treatments are complex, and the development of an interdisciplinary group of trained and experienced staff will greatly reduce the risk of errors.
33.13
LOCAL RULES AND SYSTEMS OF WORK
In the UK, it is a legislative requirement [10] that each department utilizing ionizing radiation develop 'local rules.' These are locally written documents which describe the key working instructions which must be followed by people working in or otherwise entering controlled or supervised areas of the department, the purpose of which is to ensure that legislation is complied with and that safe working practices are used. The regulations also require employers to perform risk assessments and develop contingency plans for work with ionizing radiation. Other countries have their own legislation and regulations which must be followed, but the general principles will be similar. The documentation should include a description of all the procedures to be followed during normal operation of the afterloading equipment or, in the case of manual systems, during normal calibration, manipulation, usage, and cleaning of sources and appliances. It should also include contingency plans to cover all reasonably foreseeable emergency situations. A review of safety issues in relation to remote afterloading has been written by Glasgow [11]. Radiation protection issues are dealt with more fully in Chapter 10 by A.M. Bidmead.
33.14
MANUAL AFTERLOADING
The same general principles apply to manual afterloading as to machine afterloading. There is the same need for the provision of unambiguous prescriptions, documentation of treatment protocols, quality assurance programs for equipment, and the provision of local rules and systems of work. However, certain aspects of working with manual afterloading techniques require particular consideration. This section considers two areas of manual afterloading: the use of iridium wire, and the use of preloaded source trains for gynecological treatments. As with the use of any type of small, sealed radioactive source, there must be suitable arrangements for storage, calibration, and manipulation of the radioactive material. Controlled areas, local rules, and systems of work must be defined and adhered to. There must be appropriate documentation for recording the location, use, and eventual disposal of the radioactive material. Although the control documentation and other safety
440 Safe practice and prevention of accidents in afterloading brachytherapy
procedures should be designed to detect any misplacing, loss, or damage to a source, a regular and frequent audit of the source stock must be performed and documented by the person responsible for its custody. Iridium-192 wire poses particular difficulties in stock control in that a single length of wire as supplied will be cut into a number of smaller lengths for use in a treatment application. Whereas the stock control of discrete sources, such as gynecological cesium-137 tubes, may be simplified simply to counting the number of sources of a particular type, this is not possible with iridium wire as the number of sources making up the stock is variable. One method of accounting that has been used successfully is to keep a complete record of the history of each 'delivery' (i.e., coil or length as received) of wire, showing its initial length and activity, then subsequently documenting the preparation and use of each individual cut length, finally recording the disposal of the pieces. In this case, the record will show, at any particular time, the disposition of each element of this particular batch, finally recording that the length as supplied is finally disposed of. Appropriate arrangements must be made for the temporary storage of used wire prior to its disposal. Removal of used wire from the hospital site will normally be done by arrangement with the supplier or recognized disposal agency. National regulations will normally limit the activity of used wire that may be stored on site, and the length of time for which it may be stored. In general, sources intended for use in a patient (and here we may include gynecological source trains) should have some manner of coding to ensure that the appropriate source or train of sources is placed in the correct applicator in the patient. For iridium wires, this will usually be done by using a source holder with labeled compartments or slots so that individual sources may be identified. For gynecological source trains, it is preferable to use a mechanical coding system on the applicator which matches one on the source train. Also, source trains of different configurations should be made easily identifiable by a number system or color code. Preferably, the color code chosen should avoid the combinations of colors associated with the more frequent types of color blindness. Procedures for the removal of sources from patients should be laid down and followed by all personnel. In particular, it is important that iridium wires are removed in a manner that avoids the possibility of cutting through the wire, as this may lead to its incomplete removal. Also, a hand-held radiation monitor must be used to check that all sources have been properly removed from the patient. The correct operation of the monitor should be checked before removal of the sources to ensure that it is working properly. Failure to follow these procedures may result in sources inadvertently remaining in the patient [12]. The ward where patients are treated with manual brachytherapy sources should have available at all times
a shielded container which will accept the sources and applicators in the event of their unplanned or unexpected removal. A useful arrangement is for this container to be mobile, so that it may also serve as a transport container for the routine movement of used sources back to the sources dispensary. Clear, written procedures and instructions must be drawn up relating to the movement of sources to and from the dispensary and wards. Sources that have been so returned must be checked as soon as practicable after their return to ensure that they are all accounted for and that they are undamaged. It is recommended that they should be checked, at the latest, on the next working day following removal. They should then be cleaned and returned to storage. It is recommended that, at each stage of the procedure, a named individual be responsible for the 'ownership' of the sources, this person being responsible for their care and custody at their stage. One way of achieving this is to make use of a 'sources receipt' document, which accompanies the sources around the circuit from the dispensary to the operating room, the ward, and eventually back to the dispensary, with signatures being given when the sources move from one area of responsibility to another. In this way, control of the sources is maintained and the movement of the sources is recorded. It is imperative to avoid the possibility of the unintentional removal of a radioactive source from the hospital site. There is a risk that a source, particularly a physically small source such as a short length of iridium wire, might be accidentally removed from (or perhaps by) the patient and find its way into a dressing, bed linen, or similar. In order to guard against this possibility, it is recommended that radiation detectors be placed at exits from patient ward areas to detect any unintentional or accidental removal from the ward area. Where practicable, radiation detectors should also be placed at exits from the hospital building(s), particularly at service exits through which clinical waste and laundry are normally moved.
33.15
TRAINING
The provision of training to all groups of involved staff is essential if safe provision of the brachytherapy service is to be maintained. The number of staff in a facility must therefore be sufficient to allow enough time for this training to be prepared and given. Managers of the facility should be aware of the need for training and continuing education, and staff must be given time for this important contribution to safety. Senior professional staff should assess the training and education needs of each discipline and ensure that the appropriate training is provided. It is recommended
Appendix: Example protocol for an MDR Selectron treatment 441
(indeed, a necessary requirement in some countries) that a written record of training be kept for each person. The training program should be tailored to the requirements and duties of each individual; for example, the training required by a radiographer/radiation technologist will differ from that required by a nurse or junior clinician. The training should include not only a broad view of the subject, for example new developments in brachytherapy equipment, technology, and techniques, but also practical aspects such as patient care and emergency procedures. Reference has been made earlier to the advantages of having a dedicated team of brachytherapy specialists. However, in many situations this is not practical and there will be a rotation of individuals to and from the specialty. Any person of whichever profession moving into the specialty must be provided with adequate training before being permitted to work unsupervised. A regular, ongoing training program is vital when staff are moved from one specialty to another. Also, as brachytherapy is constantly undergoing development with the consequent introduction of new concepts and procedures, training in such new methods must be given before they are implemented. Training is particularly important for those groups of staff whose professional education would not normally include radiation physics. Physicists and radiographers/radiation technologists would normally be expected to have gained a working knowledge of the principles of radiation protection, for example, but this may not necessarily be true for nurses and junior doctors. LDR and PDR machines generally treat over a number of days, so there is a need for nursing and clinical staff to have gained appropriate knowledge and experience relating to these issues in order that they may work in a safe manner, and also operate the equipment when necessary. In particular, they should know how to interrupt and restart treatments. They must also be aware of the limits induced by the fact that they are working in a radiation environment. They should understand there might be restrictions on the time that they can spend giving nursing attention to a patient with radiation sources in situ. They may be involved by assisting in the procedure for the removal of sources, and they need to know what action to take in the event of a radiation or clinical emergency. The added complexity of new equipment puts a further continual training burden on radiotherapy departments. The Medical Devices Agency (MDA) safety bulletin [13] advises that safety is dependent on training and that, for equipment, model-specific training is required. The Clinical Oncology Information Network (COIN) generic radiotherapy guidelines [14] also suggest training in both techniques and equipment new to each individual. This essential training should be part of a risk-management strategy for this potentially damaging but curative modality.
ACKNOWLEDGMENTS The authors gratefully acknowledge Pat Earle for assisting with the MDR Selectron protocol, and David Scrimger for information regarding licensing of installations in North America.
APPENDIX: EXAMPLE PROTOCOL FOR AN MDR SELECTRON TREATMENT Objective To prepare and perform an MDR Selectron treatment for individual patients in accordance with the clinician's clinical requirements. Scope Patients undergoing MDR Selectron treatments. Responsibilities Clinician Book the patient in for the procedure with theater, medical physics, the treatment unit Selectron, and admissions. Insert applicators and complete treatment card. Approve the films and treatment calculation. Anesthetist Ensure patient is suitable for the proposed anesthetic procedure. Perform appropriate anesthesia. Physicist/physics technician Oversee booking arrangements. Deliver test sources to theater. Oversee planning films. Calculate treatment time, and perform any other dosimetry required, liaising with the clinician if necessary. Check and sign calculations and treatment card. Record patient details in section records. Hand over all relevant data to Selectron radiographers. After treatment, check Selectron printout and record relevant data in physics logbook/database. Respond to faults/alarms, as required. Respond to radiation emergencies, as required. Theater nurses Prepare instruments, applicators etc. Identify and reassure patient. Assist clinical medical staff, as required. Assist with anesthetic recovery. Hand over patient to ward nurse after procedure, with documentation. Radiographers Check patient identity against prescription. Program Selectron as required by the prescription. Check program printout. Reassure patient and explain procedure. Assist clinician with connection of patient to Selectron, checking as necessary.
442 Safe practice and prevention of accidents in afterloading brachytherapy
If clinician not available, two radiographers to perform connections. Advise patient and nursing staff of likely time of treatment termination. Initiate treatment. Respond to faults/alarms, as required. Respond to radiation emergencies, as required. On completion, unload all channels and switch off. Complete documentation for recording treatment. Ward nurses Identify and reassure patient. Escort patient to theater, with documentation. Escort patient from theater to simulator, then to treatment room. Prepare patient for connection to Selectron. Attend to patient's needs throughout treatment. Check connections to transfer tubes regularly. Remove applicators at end of treatment. Report faults/alarms, as instructed. Respond to radiation emergencies, as required. Documentation Admissions control chart/calendar. Prescription card. Brachytherapy planning manual. Selectron patient logbook/database. Hospital patient database. Ionising Radiation Regulations/local rules. ICRU 38. ICRU 58. Radiographers' on-call rota. Selectron manual. Quality assurance protocol. Selectron operations logbook. Ward and operating theater records. REFERENCES
3. BS 5724 Section 2.17 (1990) Specification for Remotecontrolled Automatically Driven Gamma-ray Afterloading Equipment. London, British Standards Institution. 4. USNRC Policy and Guidance Directive FC 86-4 (1986) Washington DC, United States Nuclear Regulatory Commission. 5. ICRU Report 38 (1985) Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology. Bethesda, Maryland, International Commission on Radiation Units and Measurements. 6. ICRU Report 58 (1997) Dose and Volume Specification for Reporting Interstitial Therapy. Bethesda, Maryland, International Commission on Radiation Units and Measurements. 7. IPEM B Report No 68 (1996) A Guide to Commissioning and Quality Control of Treatment Planning Systems. York, England, The Institution of Physics and Engineering in Medicine and Biology. 8. Meisberger, LL, Keller, R.J. and Shalek, R.J. (1968) The effective attenuation in water of the gamma rays of gold198, iridium-192, cesium-137, radium-226, and cobalt60. Radiology, 90,953-7. 9. Recommended Minimum Staffing Levels for the Medical Physics Support of Radiotherapy (1989) York, England, Institute of Physical Sciences in Medicine. 10. The Ionising Radiations Regulations 1999 (1999) Statutory Instrument 1999 No. 3232. London, Her Majesty's Stationery Office. 11. Glasgow, G.P. (1996) Radiation control, personnel training, and emergency procedures for remote afterloading units. Endocuriether./Hypertherm. Oncol., 12, 67-79. 12. Arnott, S.J., Law, J., Ash, D. et al. (1985) Problems associated with iridium-192 implants. Clin. Radial., 36, 283-5. 13. Medical Devices Agency (1998) Hazard Circular, Safety Action Bulletin Medical Device and Equipment Management for Hospital and Community Based
1. IPSM Report No. 46 (1986) Radiation Protection in Radiotherapy. London, The Institute of Physical Sciences in Medicine. 2. AAPM Report No. 41 (1993) Remote Afterloading
Organisations, MDA DB 9801. London, Medical Devices Agency. 14. Clinical Oncology Information Network (COIN) (1999) Guidelines for external beam radiotherapy. Report of the
Technology. New York, American Association of Physicists
Royal College of Radiologists Generic Radiotherapy
in Medicine.
Working Group. Clin. Oncol., 11(4), reprint.
34 Pulsed low dose-rate brachytherapy in clinical practice PATRICK S.SWIFT
34.1 ADVANTAGES AND DISADVANTAGES OF PULSED LOW DOSE-RATE APPROACH
Prior to the development of the pulsed low dose-rate remote afterloading unit, interstitial brachytherapy for implantable lesions required the use of manually afterloaded strings of radioactive sources such as iridium, or permanent implantation of sources such as iodine or palladium, as in the case of prostate cancer. The introduction of a remote afterloading unit adaptable to interstitial approaches marks an important step forward in terms of radiation safety for personnel and the application of optimization formerly restricted to intracavitary procedures. The pulsed low dose-rate (PDR) Selectron
uses a single iridium-192 source of 1.1 mm diameter and 2.6 mm length (Figure 34.la and b), activity generally between 0.5-1.0 Ci, secured at the end of a cable-driven wire. This single source is programmed to move through a series of positions within catheters or needles placed previously in the tumor bed, stopping for lengths of time varying from 0 to 999.9 s per position per pulse ('dwell times') throughout the array. The position and dwell times are selected to deliver an average isodose distribution per pulse that most closely conforms to the geometry of the region to be treated. One complete movement of the source through the entire array constitutes a single pulse. The total duration of the pulse, the dose delivered per pulse, and the interval between pulses are all capable of being manipulated (Table 34.1).
Table 34.1 Pulse variables
0.5 0.5 1.0 1.0 3.0 3.0 3.0
60 60 30 30 10 10 10
Hourly Every 45 min Hourly Every 2 h Hourly Every 3 h Every 6 h
30 30 30 30 30 30 30
60 45 30 60 10 30 60
The total duration of the pulse, the dose delivered per pulse, and the interval between pulses are all capable of being manipulated. In all the examples, a dose of 30 Gy is delivered. The total treatment times and the fraction sizes differ significantly, however.
444 Pulsed low dose-rate brachytherapy in clinical practice
Figure 34.1 (a) Pulsed Selection from Nucletron, with 78 channels; (b) single iridium-192 source of 1.7 mm diameter and 2.6 mm length, with an example of the curvature of radius of a catheter negotiable by the cable-driven source.
The use of a single source of iridium reduces the need for an extensive and expensive inventory of sources for use in various situations. Given a half-life of 74 days, the iridium source is replaced at 3-month intervals. Storage of the source is simplified by its location within the shielded unit. Because the maximum activity of the source is 1.0 Ci, additional shielding is required to meet governmental standards compared to that required for standard manual afterloading. One of the prime advantages of this remote afterloading approach is the elimination of radiation exposure to radiation oncology personnel, nursing staff, physicians from other disciplines, or visiting family members. As the unit is designed to allow treatment for a fraction of each hour, the source is safely isolated for the remainder of each hour, allowing the nursing staff to work more extensively with the patient. This becomes particularly
important for patients with extensive medical problems. The pulse is generally timed to end precisely at the hour to reduce confusion as to when the nursing staff may enter the room. The main therapeutic advantage of pulsed brachytherapy lies in the process of isodose optimization. In a process identical to that used in high dose-rate (HDR) remote afterloading units, the positions at which the source stops and dwell time of the source at each position are carefully manipulated. Average isodose distributions are designed to reduce 'hot spots' secondary to decreased distances between needles or catheters in an implant, or carefully to increase doses to areas with inadequate dosing ('cold spots'). Within certain limits, dose homogeneity throughout an implant volume can be 'optimized' in pulsed brachytherapy in a fashion not possible with static afterloading using fixed sources. It
Radiobiologic rationale 445
must be pointed out, however, that even dose optimization cannot make a poorly implanted array good. There is no substitution for careful attention to uniform implantation of a volume with evenly spaced needles. Optimization also allows improved conformation of isodose distribution to tumor geometry as seen on imaging studies. In a report from Heidelberg, Berns et al. compared 25 consecutive patients undergoing interstitial breast implants as boost therapy [1]. Isodose distributions obtained with geometrical dose optimization were compared to non-optimized distributions produced by iridium wires. In order to compensate for the increased reference volume irradiated due to increased dwell times at the periphery of the implant, the active lengths of each catheter were shortened by 5-10 mm compared to the lengths used for static iridium wires. In all but three of the cases, the geometrical dose optimization resulted in distributions that were similar or superior to those obtained with static wires, with improvement of dose uniformity and an increase in the minimum implant dose. A similar comparison of isodose distributions obtained with geometric distance optimization or static iridium wires for surface moulds showed that the PDR approach resulted in increased dose uniformity in the implant as well as greater conformity to the implanted area. When used for standard intracavitary procedures, the pulsed low dose-rate unit has an additional advantage over standard continuous brachytherapy using other multisource remote afterloading continuous low doserate (LDR) units. With the latter, any entry into the patient's room, for doctors' visits or nursing attention, necessitates a break in the treatment as the sources are removed automatically. The overall duration of the treatment is lengthened to account for these interruptions. In the pulsed setting, such visits ideally are timed to coincide with the breaks in treatment between pulses, keeping the overall duration of the implant constant. Manipulation of dwell times also provides an increased degree of flexibility for determining the average dose per pulse (average dose per hour) compared to that possible with remote afterloaded fixed-position sources. Any LDR system suffers by comparison with HDR approaches because of the former's requirement of several days of hospitalization to deliver the dose. Risks of prolonged bedrest (e.g., deep venous thromboses, pulmonary emboli, stasis ulcers, etc.) exist with PDR but not with HDR. During the hospital stays, there also exists the potential for movement of the instrumentation out of the initial desired position (with tandem and ovoids, for instance). Significant movement of the patient in the bed may result in disruption of the unimpeded transit of the cable-driven source, with resultant treatment interruptions, requiring intervention to untwist the cables and catheters. Such events may occur after standard working hours and, although not generally a danger to the patient, may lead to prolongation of the overall treat-
ment time. These disadvantages are weighed against the radiobiologic benefit (see discussion below) of a low dose-rate equivalent process over a high dose-rate approach in terms of damage to late-reacting tissues. Compared to continuous LDR treatments, the main disadvantages of PDR are limited to the development of technical difficulties that are not seen with manual afterloading. These difficulties are decreasing with greater use of the units, and are far outweighed by the potential benefits.
34.2
RADIOBIOLOGIC RATIONALE
Pulsed LDR brachytherapy creates a dose-rate condition that is different from both high dose rate and continuous low dose rate. The first assumption which remains to be tested is that a dose delivered to a given volume as a brief pulse of a single stepping source, at very high instantaneous dose rates, is biologically equivalent to the same average dose delivered continuously by a series of static sources at a much lower instantaneous dose rate. The second assumption is that the dose is relatively equivalent in terms of its effect both on early-reacting tissues (including tumor) and on late-reacting tissues. Are these total doses (one continuous, one pulsed) equivalent over the range of half-times of tissue repair that are clinically relevant in the surrounding normal tissues? Brenner and Hall [2] utilized the linear-quadratic formalism of Lea and Catcheside [3] in the analysis of in-vitro dose-response data available for cell lines of human origin in an attempt to answer these questions. Survival (S) at a dose D is given by the following equation:
where a is the portion of cell-killing due to a single hit (the linear component), and (3 is that portion due to multiple hits (the exponential component). G is a function of the repair that occurs between successive hits of radiation and is dependent on the total time of irradiation, the spatial distribution of the radiation-induced events, and the repair capabilities of the particular cell line. Looking at experimental data from 36 human cell lines, with their observed values for a, (3, and t0 (characteristic repair time of the tissue), Brenner and Hall have defined the conditions under which pulsed therapy would be equivalent to continuous LDR therapy [2]. The standard chosen for comparisons was that of a typical continuous LDR implant delivering a total dose of 30 Gy over 60 h. Various combinations of pulse widths and interval durations between pulses were examined to identify conditions under which the cell survival of early-reacting tissues would be comparable to that seen with a continuous regimen. In further analysis, using the limited data available for late effects in humans, as well as animal data on late effects (mouse lung, rat spinal
446 Pulsed low dose-rate brachytherapy in clinical practice
cord), the authors predicted the impact of a variety of pulsed regimens on late effects. Their conclusion was that 10-min pulses, separated by 1-h intervals, with the overall implant duration kept constant at 60 h, would result in a similar cell survival for early-reacting tissues and only a 2% increase in late effects when compared to the continuous regimen. Although increasing dose rates result in increasing biologic effectiveness, they are also accompanied by a decline in the therapeutic ratio [4-7]. Two alterations might result in an expected decrease in the therapeutic ratio: increasing the dose delivered per pulse and increasing the time between successive pulses. Keeping the total duration of an implant static but increasing the interval between pulses, thereby increasing the dose per pulse, increases the biologic effectiveness of a dose, particularly in cell lines with a shorter half-time of repair (ti). Because the amount of repair capacity is believed to be greater in late-reacting normal tissues than in earlyreacting tissues (including most tumors), an increase in relative effectiveness would be expected to be more significant for late-reacting tissues than for early-reacting tissues, resulting in a decrease in the therapeutic ratio. If the t} for late-reacting tissues is significantly longer than that of a particular tumor tissue, the effect of pulsed therapy on the therapeutic ratio would be minimized. Data from recent experiments on rodent kidney, spinal cord, and lung tissue suggest that a component of repair of later-responding damage of approximately 4 h exists [8-10]. If this is the case, and early-responding tissues, including tumor tissue, have half-times of repair considerably shorter than those of late-responding tissues, then the pulsed approach would be expected to result in levels of late tissue damage lower than that seen with the continuous LDR approach (Figure 34.2) [11] In an effort to establish the conditions least likely to result in a decrease in the therapeutic ratio for PDR compared to continuous LDR, Fowler and Mount [12] calculated the expected effect of various pulsed regimens (with dose rates in the pulse varying from 0.5 to 120 Gy h'1 and pulses delivered every 1-4 h) on early-responding and late-responding tissues, using a wide range of possible half-times of repair from 0.1 h to 3 h. Duration and total dose of the implant were kept at 70 Gy in 140 h, and all effects were considered relative to a continuous regimen at 0.5 Gy h-1. Looking first at early-reacting (normal and tumor) tissues, biologic effectiveness would not be expected to increase by more than 3% if dose rates remained in the 0.5-3 Gy h"1 range and pulses were given hourly, regardless of the assumed Ti. As the dose per pulse and interval duration increase, the biologic effectiveness also increases for all Ti. This is true for late-reacting tissues as well. If intervals increase to one pulse per 4 h, biologic effect in late tissue may increase as much as 15% (Figure 34.3). Tissues with the shortestT1/2of repair would be at greatest risk. This would necessitate a decrease in the
Figure 34.2 Calculated fractional change in cell survival for PDR compared with LDR as a function of the assumed 7|. Both treatments consist of 30 Gy delivered in 60 h, either continuously (LDR) or in 60 10-min pulses of 0.5 Gy delivered hourly (PDR). The calculated quantity is (SPDR - SLDJ/SLDR: here the survival, S, = exp(-aD - GbD2J where D is the total dose, a and (3 are the linear-quadratic formalism parameters, and G is the quantity describing sublethal damage repair which depends on the half-time of sublethal damage repair [11].
overall dose to sustain levels of late effects similar to that seen with continuous LDR regimens, a decrease which would result in a less-than-desired effectiveness for tumor control. Fowler and Mount [12], therefore, arrived at a conclusion similar to that of Brenner and Hall [2]: keeping the repetition frequency at one pulse every 1 or 2 h would be comparable to a continuous regimen with a negligible increase in late effects. Due to concerns about the safety of operating a unit with a strong source during the night shifts, many clinics
Figure 34.3 The increase of relative effectiveness (RE) with increasing assumed values for 7| is shown for five different dose rates in the pulses. When the deviation of the curves from the straight line representing the continuous low dose rate of 0.5 Gy h-1 exceeds about 10%, the extra biologic effect (for the fixed total dose) might become clinically significant [4].
Results: in vitro and in vivo 447
have been interested in developing an approach that would allow treatment only during normal or extended daily working hours, with discontinuation of treatment during the evenings. If no change in the dose per pulse were to occur, this would result in an unacceptable prolongation of the overall duration of the implant. Adhering to a single pulse per hour would also limit the use of each unit to only one person at a time. Increasing the length of time between pulses to 2 or 3 h might allow several patients to be treated concurrently in a busy clinic. Further work by Brenner et al.[l3], Sminia et al.[14], Visser et al[!5], and Millar et al. [16], suggests that increasing the interval between pulses to up to 3 h, as well as increasing the dose per pulse and switching to a daytime only treatment schedule may, in fact, be safe. These theoretical considerations are based upon assumptions regarding the a/(3 ratios and repair kinetics of the critical tissues in the implant region. Because these are not generally known with any great degree of certainty for most human tissues, such treatment regimens must be used with caution and with close observation to determine if there are increases in late sequelae in late-reacting tissues [ 16].
343
RESULTS: IN VITRO AND IN VIVO
It is of critical importance that validation of these theoretically designed treatment schemes be evaluated in in-vivo and in-vitro animal models to minimize the risk of severe injury to late-reacting tissues. Great is the uncertainty surrounding the radiation repair times of the wide variety of normal human cell lines likely to be affected by novel and convenient treatment timing schemes. Using rat 9L/SF gliosarcoma cells, Armour et al. [17] compared continuous (0.5 Gy h-1) and pulsed (0.5 Gy h-1 averages using 0.25 Gy per 0.5-h intervals, 1 Gy per 2-h intervals, 3 Gy per 5-h intervals) LDR approaches. No differences were discernible between the pulsed and continuous regimens in the cell survival curves at 37 °C and 41°C. When cells were treated with pulses of 6 Gy every 12 h, an increase in cell killing (a relative decrease in DJ was noted. Cell killing was equivalent between pulsed and continuous techniques as long as the overall dose rate remained constant and the dose per fraction was less than the width of the shoulder of the dose-response curve. Pulsed regimens with constant overall total dose and variable interfraction intervals ranging from 1 h to 12 h were compared for three carcinoma lines (two cervix and one breast) [18]. Schedules with hourly pulses were found to have similar biologic effectiveness when compared to continuous LDR radiation, but decreased survival was noted with increased dose per pulse, as predicted. When cell lines of different radiosensitivity were compared (human bladder cancer and neuroblastoma) using different pulsing regimens, once again little
difference was noted between single hourly pulses and continuous LDR irradiation [19]. However, larger doses per fraction resulted in markedly different levels of increased biologic effectiveness for cell lines of varying radiosensitivity, much greater for the cell line of greater radiosensitivity. In this study, a significant dose-rate effect was also unexpectedly seen. These results reveal that changing the dose per pulse, and possibly the dose rate per pulse, will variably impact the biologic effectiveness of the dose for different tumor lines. This information also reveals the necessity of determining the impact of such changes on normal tissues as well. In the mouse model, Mason et al. [24] delivered totalbody irradiation (19-40 Gy), using two continuous and two pulsed regimens. Mice were sacrificed and the surviving cells per circumference of jejunum counted and plotted against the total dose for each regimen. In the continuous radiation schemas, average dose rates of 4.2 Gy h"1 and 0.7 Gy h-1 were used. For both of the pulsed schedules, average dose rates of 0.7 Gy h-1 were delivered in either 1 or 10 min and resulted in nearly identical cell survival curves. The shorter pulse duration of 1 min (a ten-fold increase in dose rate) resulted in only a 3-4% shift in the cell survival curve to the left of the curve for the 10-min pulse. Experiments evaluating the cataractogenic potential of pulsed versus continuous regimens in the rat lens model showed no major difference in cataract formation between continuous and pulsed regimens with similar doses delivered in 10-min pulses h-1, 10-min pulses every 4 h, or 100-s pulses every hour [20]. Armour et al. [21] evaluated the histologic response of injury to the rat rectum caused by the same total dose delivered via continuous or pulsed strategies. The pulse intervals were 30 min, 1 h, 2 h, 4 h, and 8 h. The therapeutic ratio remained relatively constant for continuous irradiation and pulsed regimens where the dose per pulse remained below 1.5 Gy, or 2-h intervals. Pulse fraction size greater than 1.5 Gy resulted in a steady increase in the damage to rectal tissue using five different assessment scales. Pulsed regimens utilizing intervals greater than 2 h might therefore be expected to result in a decrease in the therapeutic ratio for the rectum, if the rat rectum has characteristics similar to the human and if the dose per pulse is allowed to exceed a certain threshold as one attempts to keep the overall treatment duration constant. Haustermans.et al. [22] reported on the effects of continuous versus pulsed regimens of radiation directed at the rat cervical spinal cord. Continuous LDR irradiation was delivered using a specially designed collar with iridium wires, over a 72-h period. The pulsed doses were delivered over 9-h periods daily with overnight gaps, with radiation delivered using a linear accelerator at a constant dose rate per pulse, but with pulse intervals ranging from 1 to 3 h, and required from 5 to 7 days to deliver the full dose. Total doses ranged up to 68 Gy. No rats treated with continuous LDR irradiation developed
448 Pulsed low dose-rate brachytherapy in clinical practice
myelopathy up to 9 months. Pulsed regimens resulted in a 50% myelopathy incidence at an average of 60 Gy. The reason for this absence of observed effect in the continuous group put forth by the authors is that there are two components of repair in the rat spinal cord, one of which is very rapid. This leads to the suggested need for caution when utilizing pulsed regimens of increasing pulse size when the spinal cord may be affected. Hall and Brenner [23] published a note of caution in the editorial, however, regarding extrapolating the results of this study to clinical experience in humans.
34.4
CLINICAL RESULTS
A limited number of clinical studies of the use of pulsed LDR irradiation has been published to date. These studies are for the most part retrospective analyses of limited numbers of patients with various tumor types, with relatively short follow-up. From 1992 to 1995, Swift et al [25] reported on 65 patients with pelvic malignancies (54 primary and 11 recurrent) who underwent a total of 77 brachytherapy procedures, 45 intracavitary and 32 interstitial. Isodose distributions were planned using the Nucletron PLATO system, with geometric optimization carried out on all cases. Patients were treated with a combination of externalbeam radiation plus the brachytherapy procedures, with the implant dose being determined by the clinical situation. In adherence to the recommendations of Brenner and Hall [2], all patients were treated at a dose per pulse of 0.4-0.85 Gy, with hourly pulses delivered around the clock. Extreme care was taken during the implantation and planning process to keep rectal dose at a minimum. At a median follow-up of 16 months, local control was 78%, and 2-year survival 67%. An absolute incidence of 6.5% for RTOG acute grade 3 or greater complications was seen, with a 15% incidence of RTOG delayed grade 3 or greater complications. Of four cases with rectovaginal fistulae, three had massive central recurrence contributing to the fistula formation. Omitting these patients as treatment-related complications, the complication rate dropped to 8%. The authors make it clear that longer follow-up is essential to determine the true delayed complication rate seen with this approach, but the results to that point are quite similar to those seen with continuous LDR implants. Rogers et al. [26] reported on 46 patients with cervical carcinoma treated with 28 intracavitary and 18 interstitial implants in conjunction with external-beam radiation. All patients were treated with hourly pulses of 0.4—0.7 Gy, with plans derived from the Nucletron PLATO planning system. Mention of the type of optimization used was not made. Median follow-up was 25 months. Pelvic control at 4 years was 86%, with a 4-year disease-free survival of 65%. Acute complications were
not discussed, but the late complication rate (RTOG grade 3 or greater) was only 7%. One vesico-vaginal fistula and one recto-vaginal fistula were seen, both after biopsies were obtained that revealed benign tissue. The delayed complication rate compares quite favorably with those seen using standard continuous LDR approaches. An additional report from Jensen et al. [27] deals with 34 patients with locally advanced or recurrent gynecologic malignancies, treated with 46 Gy external beam followed by 30 Gy delivered in hourly pulses of 0.6 Gy, using volume optimization. At a median follow-up of 14 months, 17 chronic grade 3 or higher complications were reported in ten patients (five in one patient alone). The median treatment volume for these patients was 177.5 cm3, with volumes ranging from 200 to 650 cm3 for those who developed late complications. DePree et al. [28] reported on 43 patients treated with PDR regimens for tumors of various sites, including 34 pelvic sites. All patients were treated using the hourly pulse of 0.4-1.0 Gy, with a 10% reduction in the overall dose compared to what would have been prescribed in the setting of continuous LDR brachytherapy. No excess of acute or delayed toxicities was reported compared to historical results at a median follow-up of 18 months. Seventeen patients with anal carcinoma were treated with external-beam radiation to a dose of 46 Gy, without chemotherapy, followed by a PDR implant of 25.2 Gy, delivered at 0.6 Gy rr1 over 42 h [29]. Although only three local recurrences were noted, the toxicity level was unacceptable. Thirteen patients developed necrosis, and eight required a colostomy. The volume that received in excess of 25 Gy (71 Gy including the external component) ranged from 21 cm3 to 400 cm3, with higher volumes being associated with a greater risk of colostomy. Fritz et al. [30] reported on 65 patients with breast cancer considered to be at high risk for local recurrence, who underwent external-beam irradiation to a dose of 50 Gy, followed by a PDR boost (usually with a twoplane implant) of 20 Gy, using 1 Gy h-1 intervals. In the initial 35 cases, concurrent plans were run using the PDR with geometric volume optimization and static iridium wire plans for comparison of the quality and heterogeneity indices of these implants. These comparisons showed a clear superiority in the majority of cases favoring the PDR approach with geometric volume optimization over continuous LDR plans in terms of dose homogeneity within the reference volume and increased minimal dose within the volume. At a median follow-up of 30 months, there has been only one local failure, only one acute complication noted (temporary erythema), and 11% with minimal telangiectasis, with no breast retraction, soft-tissue necrosis, or severe fibrosis noted. Care was taken by the authors to point out that the dwell positions were selected with an increased offset from the skin surface compared to the continuous LDR plans to avoid overdosing the skin, resulting in the superior cosmetic results in these high-risk patients.
References 449
Levendag et al. [31] used two different treatment regimens for 38 patients with squamous cell carcinoma of the soft palate or tonsillar fossa, the majority of whom received external-beam radiation first to a dose of 46-50 Gy. These patients were then treated either with a daytime fractionated HDR approach using 3 Gy fractions twice daily at 6-h intervals to a dose of 21-27 Gy, or a PDR approach consisting of a schedule of 3-h intervals between pulses, with either four fractions of 2 Gy during the daytime only or eight fractions of 1-1.5 Gy given over 24-h periods. These patients had identical results, in terms of both local control and observed toxicities. No increase in toxicity was noted compared to historical controls using a continuous LDR approach, with a suggestion of an increased efficacy of this approach compared to the continuous LDR approach. The authors point out that this lack of increased toxicity with the longer pulse intervals was not unexpected, assuming a rather long T1/2 of repair for late effects in the normal tissues of this region of the anatomy, and cannot necessarily be extrapolated to other regions of the body. To our knowledge, this is the only major report to date that utilizes intervals in excess of greater than 2 h.
brachytherapy-is routine geometrical optimization recommendable?inf.y. Radial Oncol. Biol. Phys., 37(5), 1171-80. 2. Brenner, D.J. and Hall, E.J. (1991) Conditions for the equivalence of continuous to pulsed low dose rate brachytherapy. Int.J. Radial Oncol. Biol. Phys., 20(1), 181-90. 3. Lea, D.E. and Catcheside, D.G. (1942) The mechanisms of the induction by radiation of chromosome aberrations in tradescantia.y. Genet., 44,216-45. 4. Fowler, J.F. (1993) Why shorter half-times of repair lead to greater damage in pulsed brachytherapy. Int.J. Radiat. Oncol. Biol. Phys., 26(2), 353-6. 5. Hall, E.J. (1991) Weiss lecture. The dose-rate factor in radiation biology. Int.J. Radiat. Biol., 59(3), 595-610. 6. Hall, E.J. and Brenner, D.J. (1991) The dose-rate effect revisited: radiobiological considerations of importance in radiotherapy. Int.J. Radiat. Oncol. Biol. Phys.,21(6), 1403-14. 7. Hall, E.J. and Brenner, D.J. (1992) The dose-rate effect in interstitial brachytherapy: a controversy resolved. Br.J. Radial., 65,242-7. 8. Ang, K.K., Jiang, G.L., Guttenberger, R. et al. (1992) Impact of spinal cord repair kinetics on the practice of altered fractionation schedules. Radiother. Oncol., 25(4), 287-94.
34.5
CONCLUSION
9. Moulder, J.E. and Fish, B.L (1992) Repair of sublethal damage in the rat kidney. In Radiation Research: a
The pulsed LDR approach to brachytherapy has a number of technical advantages that relate to improved patient and staff safety. It has a theoretic advantage over HDR approaches in terms of the risk to late-reacting tissues. It also has been shown to have the capacity to decrease the underdosage in areas of the implant as well as increase the conformity of the implant reference volume to the target volume in a way not generally possible with static implants. Early clinical reports, with a few notable exceptions, show no increase in early or acute complication rates, as long as a conservative dosing schedule of < 1.5 Gy is delivered in pulse intervals of 2 h or less. Longer followup is clearly needed to substantiate these early findings. More convenient approaches with longer pulse intervals may be safe in certain settings, but this needs to be tested further. A regimen that is safe and effective in the management of tumors with one type of normal surrounding tissues may not be as safe in other regions. On the other hand, an approach with larger doses per pulse may actually improve the local control rate in certain settings without decreasing the therapeutic ratio. Much work still remains to be done in terms of defining theT1/2of various normal and tumor tissues in humans.
REFERENCES
Twentieth Century Perspective, Vol. 1, ed. J.D. Chapman, W.C. DeweyandG.F. Whitmore. San Diego, Academic Press, 238. 10. van Rongen, E., Thames, H.J. and Travis, E.L (1993) Recovery from radiation damage in mouse lung: interpretation in terms of two rates of repair. Radiat. Res., 133(2), 225-33. 11. Brenner, D.J., Hall, E.J., Huang, Y. and Sachs, R.K. (1995) Potential reduced late effects for pulsed brachytherapy compared with conventional LDR [letter; comment]. Int. J. Radiat. Oncol. Biol. Phys., 31(1), 201-2. 12. Fowler, J. and Mount, M. (1992) Pulsed brachytherapy: the conditions for no significant loss of therapeutic ratio compared with traditional low dose rate brachytherapy. Int.]. Radial Oncol. Biol. Phys.,23(3), 661-9. 13. Brenner, D.J., Schiff, P.B., Huang, Y. and Hall, E.J. (1997) Pulsed-dose-rate brachytherapy: design of convenient (daytime only) schedules. IntJ. Radiat. Oncol. Biol. Phys., 39(4), 809-15. 14. Sminia, P., Schneider, C.J., Koedooder, K., van Tienhoven, G., Blank, L.E.C.M. and Gonzalez, D.G. (1998) Pulse frequency in pulsed brachytherapy based on tissue repair kinetics. Int.J. Radiat. Oncol. Biol. Phys.,41(1), 139-50. 15. Visser, A.G., van den Aardweg, G.J.M.J. and Levendag, P.C. (1996) Pulsed dose rate and fractionated high dose rate brachytherapy: choice of brachytherapy schedules to replace low dose rate treatments. Int.J. Radiat. Oncol. Biol. Phys., 34(2), 497-505.
1. Berns, C., Fritz, P., Hensley, F.W. and Wannenmacher, M. (1997) Consequences of optimization in PDR
16. Millar, W.T., Hendry, J.H. and Canney, P.A. (1996) The influence of the number of fractions and bi-exponential
450 Pulsed low dose-rate brachytherapy in clinical practice
repair kinetics on biological equivalence in pulsed brachytherapy. Br.J. Radial., 69(821), 457-68. 17. Armour, E., Wang, Z.H., Corry, P. and Martinez, A. (1992) Equivalence of continuous and pulse simulated low dose rate irradiation in 9L gliosarcoma cells at 37 degrees and
Janjan, N. (1994) Comparison of continuous and pulsed low dose rate brachytherapy: biological equivalence in vivo. InlJ. Radial Oncol. Biol. Phys., 28(3), 667-71. 25. Swift, P.S., Purser, P., Roberts, L.W., Pickett, B., Powell, C.B. and Phillips, T.L (1997) Pulsed dose rate
41 degrees C. Int.J. Radial Oncol. Biol. Phys., 22(1),
brachytherapy for pelvic malignancies. InlJ. Radial
109-14.
Oncol. Biol. Phys., 37(4), 811-17.
18. Chen.C.Z., Huang, Y., Hall, E.J.and Brenner, D.J. (1997)
26. Rogers, C.L, Freel, J.H. and Speiser, B.L (1998) Pulsed low
Pulsed brachytherapy as a substitute for continuous low dose rate: an in vitro study with human carcinoma lines. Int.). Radial Oncol. Biol. Phys., 37(1), 137-43. 19. Pomp, J., Woudstra, E.G. and Kampinga, H.H. (1999)
dose rate brachytherapy for uterine cervix carcinoma. Int. J. Radial Oncol. Biol. Phys., 43(1), 95-100. 27. Jensen, P.T., Roed, H., Engelholm, S.A. and Rosendal, F. (1998) Pulsed dose rate (PDR) brachytherapy as salvage treatment of locally advanced or recurrent gynecological cancer. InlJ. Radial Oncol. Biol. Phys.,42(5), 1041-7.
Pulsed dose rate and low dose rate brachytherapy: comparison of sparing effects in cells of a radiosensitive and a radioresistant cell line. Radial Res., 151,449-53.
28.
20. Brenner, D.J., Hall, E.J., Randers-Pherson, Q.et al. (1996) Quantitative comparisons of continuous and pulsed low
dose rate interstitial brachytherapy. Int.J. Radial Oncol.
dose rate regimens in a model late-effect system. InlJ.
Biol. Phys., 43(5), 971-6.
Radial Oncol. Biol. Phys., 34(4), 905-10. 21. Armour, E.P., White, J.R.,Armin, A.R. efo/. (1997) Pulsed low dose rate brachytherapy in a rat model: dependence of late rectal injury on radiation pulse size. Int. J. Radial
29.
brachytherapy with flexible implants for interstitial boost after breast-conserving surgery and external beam
Haustermans, K., Fowler, J., Landuyt, W., Lambin, P., van
radiation therapy. Radiother. Oncol., 45,23-32.
der Kogel, A. and vanderScheuren, E. (1997) Is pulsed dose rate more damaging to spinal cord of rats than continuous low dose rate? Radiother. Oncol., 45,39-47. 23.
Roed, H., Ehgelholm, S.A., Svendsen, L.B., Rosendal, F. and Olsen, K.J. (1996) Pulsed dose rate (PDR) brachytherapy of
anal carcinoma. Radiother. Oncol., 41,131-4. 30. Fritz, P., Berns, C., Anton, H.W. et al. (1997) PDR
Oncol. Biol. Phys., 38(4), 825-34. 22.
DePree, C., Popowski, Y., Weber, D., Nouet, P., Rouzad, M. and Kurtz, J.M. (1999) Feasibility and tolerance of pulsed
31.
Levendag, P.C., Scmitz, P.I.M., Jansen, P.P. et al. (1997) Fractionated high dose rate and pulsed dose rate
Hall, E.J. and Brenner, D.J. (1997) Pulsed dose-rate
brachytherapy: first clinical experience in squamous cell
brachytherapy (editorial). Radiother. Oncol., 45,1-2.
carcinoma of the tonsillar fossa and soft palate. InlJ. Radial Oncol. Biol. Phys., 38(3), 497-506.
24. Mason, K.A., Thames, H.D., Ochran, T.G., Ruifrok, A.C. and
Index
Page references in italics refer to figures; those in bold refer to tables AAPM recommendations 15-17 general AAPM formalism 15-16 low energy emitters 15-16 aberration yield 174 accident prevention 433-42 afterloading systems 103-10 advantages 103 costs 414-15 high dose-rate systems 108-10 low dose-rate remote systems 107-8 manual 105-6,439-40 pulsed dose-rate systems 110 air kerma strength 13, 16, 20 americium-241 4,10 Amersham crimping tool 105 Amersham Gynecological System 106 Amersham Iridium Wire Loader crimping tool 105 Amersham Manual Afterloading System 5,5,6 anaplastic astrocytoma 373 anisotropy constant 17 anisotropy factor 17 anisotropy function 17 anorectal carcinoma, interstitial brachytherapy in 387-92 aftercare 391 equipment 388 pretreatment assessment/investigations 387-8 technique 388-90 treatment protocol 388 apoptosis 171 applicator orifice method 143 applicator tip localization method 143 arm, skin tumors 397-8 astrocytoma, anaplastic 373 ataxia telangiectasia 208-9 attenuation in irradiated medium 14-15 autoradiograph device 125, 126 autoradiography _149 average dimension method 42-3 basal cell carcinomas 393 basal dose rate (BD) 84 becquerel 13 bile duct carcinoma 325-30 brachytherapy 326-7 chemotherapy 325-6 dose-fractionation schemes 328 dose specification 327-8 general management 325-9 high dose-rate therapy 329 intraoperative radiation therapy 326 low dose-rate therapy 328-9 multimodality 326 planning and technique 326 pretreatment assessment 325 radiation therapy 326 surgery 325
target volume 327 bioeffect dose model 77-8 biologically effective dose (BED) 190 biophysical modeling in radiotherapy 189-90 bismuth-214 (radium C) 214 bladder, high dose-rate brachytherapy in cervical cancer, effects on 368 bladder reference point 96 brain tumours, brachytherapy for 373-7 low-grade tumors 376-7 newly diagnosed malignant gliomas 375-6 recurrent malignant gliomas 374—5 treatment technique 373-4 breast cancer, low dose-rate brachytherapy 266-79 costs 419 early-stage 267-74 boost treatment 267-73 after breast-conserving therapy 274 as sole radiation treatment 273-4 historical perspective 266-7 locally advanced breast cancer 274 technique 274-7 treatment planning 277-9 British National Radium Standard 20 buccal mucosa 289 Buchler system 108 calibration methods 21-4 calibration of sources in remote afterloading systems 24—32 low dose-rate preloaded cesium-137 source trains 25 low dose-rate sources in form of wires or ribbons 25 multiple high dose-rate sources 27-8 multiple low dose-rate sources 25—7 single high dose-rate sources 28-32 'in-air' calibrations 28-31,29 re-entrant ionization chamber measurements 31-2 solid phantom (or water phantom) measurements 32 pulse dose-rate sources 32 capital costs 411 catheter image points 52 catheter image tracking 52 cell adhesive matrix (CAM) assay 206-7 central plane 87 cervical cancer costs of brachytherapy 419 Fletcher applicators, use of 47 Manchester system 44-7, 45-6, 46 see also cervical cancer, high dose-rate brachytherapy in; cervical cancer, low dose-rate brachytherapy in; cervical carcinoma, interstitial brachytherapy in
cervical cancer, high dose-rate brachytherapy in 354-69 age groups affected 355 combined intracavitary/external-beam irradiation 356, 360 dose distribution 358 dose-volume specification 359 early-stage disease 356 combined intracavitary/external-beam irradiation 356 intracavitary irradiation alone 356 intracavitary irradiation/Wertheim hysterectomy 356 historical background 354—5 intracavitary treatment 357-8 late-stage disease 356—7 external-beam, irradiation 356-7 lymph node status 355-6 normal-tissue effects 365-8 bladder 368 delayed tissue damage 366 early tissue damage 366 rectum and sigmoid colon 366-7 small bowel 368 source position relationships 358 special considerations 359-60 target volume 359 tumor size 355 treatment planning 360-5 bladder dose 363 combined intracavitary/external-beam irradiation 360 homogeneous external-beam/high dose-rate intracavitary boost irradiation 360 intracavitary high dose-rate/shielded external-beam boost irradiation 360-1 optimal dose regimen 365 pelvic disease control, factors affecting 363-5 pelvic failure and distant metastasis 365 rectal dose 362 survival rates 365 treatment procedure 361-2 wedge position 361 cervical cancer, low dose-rate brachytherapy in 343-53 clinical trials 350-2 external-beam (unwedged)/ intracavitary therapy 351-2 external-beam (wedged)/intracavitary therapy 351 intracavitary therapy alone 350—1 historical background 343-4 international experience 352-3 pretreatment assessment and investigations 344-6
452 Index
cervical cancer, low dose-rate brachytherapy in (contd) pretreatment assessment and investigations (contd) biochemical profile 345 biopsy 345 cystoscopy 346 full blood count 345 imaging 345 treatment methods 346-50 external-beam techniques 346—7 intracavitary technique 347 low dose-rate remote afterloading 347 treatment planning 347 treatment results 349-50 cervical carcinoma, interstitial brachytherapy in 379-84 loading and unloading of radioactive sources 381-2 localization films 381 pretreatment work-up 379-80 technique 380-1 treatment protocol 382 cesium-137 4, 5-6,12 for Buchler afterloading 6 for curietron 6 forms for annual afterloading 5 properties 5 remote afterloading 6 sources 5-6 spherical sources 6, 6 checkpoint 165, 184 clinical target volume (CTV) 83 closed radiation source 7 cobalt-60 4 production 4 properties 6 sources 6 complex aberrations 172 computer-assisted dose calculation 49—50 Conformal Index (COIN) 75 contact-inhibited monolayers 166 continuous low dose-rate (CLDR) 105, 219-20 cost benefit 417-18 cost-effectiveness 417 cost minimization 415-17 brachytherapy and electron therapy 416 brachytherapy and external-beam therapy 416 brachytherapy and surgery 416-17 LDR vs. ]HDR 415-16 manual vs. remote afterloading 415 cost utility 417 costs of brachytherapy 410-20 charging for healthcare 413 cost areas 414 costs of afterloading 414-15 costs of manual techniques 414 economic evaluation 415-18 failure 412-13 healthcare resource groups 413-14 principles 411-13 radiotherapy costs 410-11 variables 411 UK funding 413 USA funding 413 Courtenay-Mills soft agar assay 207—8 coverage index (CI) 74, 75 Creteil System 47 CT-based brachytherapy treatment planning 75 CT scanner with a gantry tilt option 77 CTV 91 curie 12-13
Curietron 107 Detex paper 139 direct operating costs 413 discounting 411 dose anisotropy function 51-2 dose calculation, computer-assisted 49-50 dose distribution in interstitial therapy 86-92 Dose Homogeneity Index (HI) 74 dose—non-uniformity ratio (DNR) 74 dose rate 91 dose-rate constant 16 dose-rate corrections 12 dose-rate effects 180-7 bias of tumor size and dose rate 217-18 cell killing around an implanted radiation source 185-7 differences in repair rates 219-20 early-responding and later-responding tissues 218-19 in human tumor cells 182-3 implications for clinical brachytherapy 187 iridium wire implants and 216-17 irradiation on cell-cycle progression 184-5 mechanisms 181-2 models 183-4 radium needle implants and 215-16 sublethal damage repair rates 219 time-scale of radiation action 180—1 dose specification (ICRU recommendations) 81-100 dose uniformity parameters 91 dose-volume histograms 43-4, 65-75 cumulative 72 differential, of a single point source 67-8 differential, of multiple point sources 68-9 evaluation of dose distributions with 72—3 natural 69-72 prescription dose in nonoptimized/optimized volume implants 73-4 dwell position 57 dwell time 57 dwell time gradient 59-60 dwell time gradient restriction 60
ear 288 skin tumors 398 endobronchial brachytherapy (EBBT) 225-40 background 225-6 complications 238 endobronchial catheters 226 management 238-9 pros and cons 239-40 protocol 233-7 eligibility 233-4 indications 234 protocol 1.0 curative intent 234 protocol 2.0 palliative intent 234 protocol 3.0 recurrent patients 234-5 in stages I, II, III or recurrent lung cancer 229-33 treatment prescription 226-7 treatment strategies 227—9 endometrial cancer, brachytherapy for 333-9 clinical aspects 333-4 dose rates and choice of nuclide 334 Freiburg study of therapy and recurrent disease 337-9 management and clinical practice 334-9 applicators 335-6
dose rate 336 fractionation schemes and different dose rate 336 postoperative brachytherapy of vagina 336 preoperative and postoperative irradiation 336 recurrence rates after surgery and irradiation 337 uterine packing in 99 energy response curve method 30-1 equipment checks 437 equipment design 433-4 equivalent activity 13 equivalent annual costs (EAC) 411 esophageal carcinoma, brachytherapy of 243-54 clinical staging and pretreatment investigations 244-5 complications 253 costs 419 history and treatment rationale 244 intraluminal brachytherapy technique 246-7 natural history 243—4 radiological and clinical considerations 248-50 therapy decision process 245-6 TNM classification 244 treatment results 250-3 exposure rate constant 13 eyelid 288 face, skin tumors 398 Farmer ionization chamber 26 Filmless planning 77 financial accounting 413 fixed costs 411 Fletcher-Suit afterloading 113 Fletcher-Suit ovoids 121 Fletcher-Suit rigid, applicators 88, 96,106 Fletcher-Suit technique 384 Fletcher-type applicators 47 use in cervical cancer 47 fractionated stereotactic radiotherapy 199 fundholding 413 Gafchromic film 139 Gammamed 109 geometric optimization 57 geometric optimization on distance 63 geometric optimization on volume 62-3 geometry factor 16, 51 glioblastoma multiforme 373 gold-198 4,12, 15 gross tumour volume (GTV) 82-3 Gustave-Roussy technique 47 gynecological applicators 107 hand, skin tumors 397-8 head and neck cancer, brachytherapy in 284-93 clinical brachytherapy 302-10 base of tongue/mobile tongue: interstitial volume implant 302-3 nasal vestibule brachytherapy 304 nasal vestibule: interstitial single-plane implant 303-4 nasal vestibule: mould techniques 304 nasopharynx: endocavitary brachytherapy 307-9 tonsil and soft palate: single plane interstitial implant 304-6 tumor control of tonsillar fossa and soft palate 306-7 costs 419
Index 453
fixture directions 291-3 high dose-rate interstitial/endocavitary brachytherapy 296-315 choice of type of remote-controlled afterloaders 297 different repair half-times/alpha/beta values 299-301 dose fractionation 302 fractionated HDR and PDR schedules 297-301 overall effect of external radiotherapy/ brachytherapy boost 301 radiobiological model 298-9 schedules in Rotterdam 297-302 historical background 284 Institut Gustave-Roussy results 289-91 integrated brachytherapy unit (ITU) 310-15 neck: flexible intraoperative template 310-15 re-irradiation 311 pretreatment assessment and investigations 284-5 dental assessment 285 neck assessment 285 primary tumor 284-5 Rotterdam viewpoint 315 treatment 285-7, 288-9 hypodermic needle technique 286-7 iridium-192 285 iridium-192 hairpins 285-6 Paris System 287 plastic tube technique 286 silk suture technique 287 treatment planning 287-8 general assessment 287 geometric configuration 288 tumor evaluation 287 radioactive sources 288 Heyman capsules 47 high dose regions 90-1 high dose-rate (HDR) 49, 104-5, 108-10 high dose-rate remote afterloading, quality assurance 133-45 documentation 143-5 facility design 134-6 machine function tests 136-7 regulatory requirements 134 treatment precision tests 137-43 positional reproducibility 138-42 positional reproducibility in interstitial brachytherapy 142-4 high dose-rate vs. low-dose rate brachytherapy 196-9 in cervical cancer 196-9 general principles 196 hyperthermia 400-7 biological factors 400-2 simultaneous, brachytherapy and 403—5 systems considerations 403 thermometry requirements 403 hyperthermic universal perineal implant template (HUPIT) 404 'in-air' calibration jigs 29 'in-air' measurement using ionization chambers 23—4 Incomplete Repair (IR) model 183-4 indirect operating costs 413 inpatient/outpatient costs 412 input budget 413 installation 435 treatment protocols 435 internal market 413 interstitial therapy, recording and reporting 92-4
dose distribution 93—4 sources 93 technique and source pattern 93 time-dose pattern 93 TRAK93 volumes 93 intracavitary therapy in gynecology applicator 95 reporting, recommendations for absorbed dose at reference points 95—8 close to sources and related to sources 95 related to bony structures 97—8 relatively close to sources but related to organs at risk 96-7 calculation of dose distribution 98—9 description of reference volume 95 dose level 95 pear-shaped volume 95 60-Gy reference volume in special situations 99 TRAK95 simulation of linear sources 94-5 sources 94 intraoperative radiation therapy (IORT) 319 iodine-125 4 forms of seed 8-9, 9 production 4 properties 9 source 8-9 iodine seeds 106 Ionising Radiation (Outside Workers) Regulations 1993 148 Ionising Radiation Regulations (IRR) 147 iridium hairpins 105—6 iridium ribbons 106 iridium wire implants 216-17 iridium wires 105 iridium-191, production of 4-5 iridium-192 4, 15 hairpins 7-8, 8 miniature, for high dose rate 8 production 4 properties 7 seeds 8 source strength 7, 8 sources 6-8 wire 6-7 isocentric reconstruction method 53-4, 85 Joslin-Flynn applicator 109 Kaposi sarcoma 398 kerma rate calculations in water 27 kerma rate measurements in water 26 Kodak X-Omat Verification film 139 lead-214 (radium B) 11 legs, skin tumors 398 Leksell stereotactic system 374 Lethal-Potentially Lethal damage (LPL) model 183-4 linear energy transfer (LET) 175-6 linear-quadratic (LQ) model 76, 189, 191-6 calculations for cervical cancer 198-9 extensions to 192—6 moving isodose surfaces 195—6 redistribution and reoxygenation 193 repopulation 192-3 tumour shrinkage 193-5 mechanistic basis 191-2 parameters 192 lip 288, 290 local minimum doses 86 local rules 439 low dose regions 91
low dose-rate (LDR) 49, 104-5, 107-8 vs. pulsed dose-rate brachytherapy 199-201 low dose-rate and fractionated irradiation 161-77 aberration yield and survival change 174 apoptosis and signal transduction 176 changes with linear energy transfer 175-6 multiple dose fractionation 168-70 potentially lethal damage 166-8 potentially lethal damage repair 174-5 repair of radiation damage cell-cycle complication 163-5 cell-cycle progression, effect on 165 history of 162 split-dose recovery from sublethal damage 162-3 variation in recovery 163 selection of mutants 176-7 subcellular changes 170-1 sublethal damage 171-4 low dose-rate remote afterloading, quality assurance 112-31 acceptance testing 116-20 acceptance test report 120 brachytherapy source testing 116-17 functional tests 119-20 radiation surveys 117 safety features 117-18 source transport and applicator tests 118-19 equipment selection 112-13 ongoing quality assurance program 124-31 documentation 129 machine-testing schedules 126-9 practical operational considerations 130-1 shield placement and surveys 125-6 training personnel for remote afterloading 129-30 treatment verification 124-5 preparation 112—16 quality assurance procedures 121-3 dose computations 123 iridium seed ribbon preparation 122 localization radiographs 122—3 motor-driven source—cable assemblies 121-2 source-spacer pellet configurations 121 treatment planning 122-3 treatment prescription 123 site preparation 113-16 authorization for remote afterloading 115 facility design 114-15 installation 115-16 location 113 shielding considerations 113-14 lung cancer costs of brachytherapy 418-19 endobronchial brachytherapy in 225—40 lymphatic trapezoid 97, 98 machine quality assurance 438 maintenance 438-9 management accounting 413 Manchester System 12, 35, 38-41, 82, 218 carcinoma of body of the uterus 47 cervical cancer and 44—7, 45-6, 46, 343-53 dosage 40, 84 interstitial planar implants 40-1, 40 moulds 39-40, 39, 39 volume implants 41,41 manual afterloading 105-6, 439-40
454
Index
marginal costs 412 MARS Regulations 148 Martinez Universal Perineal Implant Template (MUPIT) 383, 404 mean central dose 82, 87-90 Meisberger polynomials 15 microSelectron-HDR 103,109 microSelectron-LDR 107-8,112 microSelectron-PDR 110 milligram-hour concept 12 minimum peripheral dose (MPD) 67, 74, 90 minimum target dose 82, 84, 90 Monte Carlo techniques 11,14-15,16, 17, 31,76 mouth, floor of 289, 290 MRI 75-6 multiple damaged sites 181 multiple dose fractionation 168-70 nasal vestibule brachytherapy 304 interstitial single-plane implant 303—1 mould techniques 304 nasopharynx 289, 290 endocavitary brachytherapy 307—9 National Institute of Standards and Technology (NIST) (US) 20 natural dose ratio (NDR) 74 natural prescription dose (NPD) 74 natural volume dose histogram (NVDH) 43-4,46, 92 Newcastle System 47 'no system' method 43 Nominal Standard Dose (NSD) approach 189 normal tissue complication probability (NTCP) 190,191-2,298 normal-tissue cellular radiosensitivity 208—9 Norman-Simon capsules 118-19 nose 288, 290, 303-4 operating costs 411 optimization on distance 57 optimization on volume 57 optimization techniques in stepping source brachytherapy 57-64 distance and volume implants 57 geometric optimization on American volume implants 62-3 on European distance implants 63 on European volume implants 63 least square minimization 58-61 dwell time gradient 59-60 polynomial optimization 60-1 linear programming 61 polynomial optimization on volume 64 rules of optimization 57-8 simulated annealing 61-2 volume implants 62-4 oropharynx 290 orthogonal reconstruction method 52-3 output budget 413 paediatric head and neck malignancies 291 palladium-103 4, 9,10 pancreatic cancer, brachytherapy in 317-24 chemotherapy 318 intraoperative brachytherapy 319—20 available isotopes 320 indications 319 intraoperative radiation therapy (IORT) 319 pretreatment assessment 317-18 radiation therapy 318-19 surgery 318 treatment planning and technique 320—4
Paris Dosimetry System 35-8, 47, 61, 64, 82, 92 basic principles 35-6 dose calculation 37, 37, 38 dose specification 84-5 positioning the sources 36—7 problems 38 Paris System 287 Paterson-Parker System, see Manchester System pelvic-wall reference point 97 perspex phantom 27 pillar and soft palate 289 pinna, skin tumors 398 planning target volume (PTV) 83—4 polonium-214 (radium C') 214 polynomial optimization on volume 64 POPUMET Regulations 147 potentially lethal damage 166-8 potentially lethal damage repair (PLDR) 166, 174-5 predictive assays 205-20 cell-cycle analysis and tumor response 211 oxygen measurements and tumor response 209-11 requirements 206 survival assays 206-9 cell adhesive matrix (CAM) assay 206-7 Courtenay-Mills soft agar assay 207-8 normal-tissue cellular radiosensitivity 208-9 tumor-cell radiosensitivity 206 prescription dose 67 pretreatment checks 437 primary care groups 413 procedural checks 437 prostate cancer 75, 257-64 catheter insertion and fixation 259-60 complications and toxicity 263-4 computed tomography-based 3D planning 261 costs of brachytherapy 419-20 dose prescription 261-3 fluoroscopic implantation procedure 258 implant reconstruction 260-1 implant techniques 258 Mount Vernon applicator and template technique 260 procedure 258 transrectal ultrasound implantation technique 258-9 treatment results 263 provider units 413 PTW-Freiburg re-entrant chamber 22 pulsed dose rate (PDR) 105, 110,199-200 'daytime' 200 equivalent regimens 200-1 pulsed low dose-rate brachytherapy 443-9 advantages and disadvantages 443-5 clinical results 448-9 in vitro and in vivo 447-8 radiobiologic rationale 445—7 purchases units 413 quality assurance 424 high dose-rate remote afterloading 133-15 low dose-rate remote afterloading 112-31 Quality Index (QI) 69, 72 quality management 423 administrative requirements 429-31 communications 431 motivation and training 429-31 resource requirements 421 audit 424-5
clinical aspects 424-9 confirmation of delivery of treatment 428 effectiveness of treatment 427-8 local tumor control 428 normal-tissue effects 428-9 time scale for follow-up 429 patient, the 426 prescription and treatment procedure 427-8 pretreatment 426 pretreatment assessment 426 target volume 427 treatment intent 426 treatment plan 426-7 tumor volume 427 treatment optimization 427 common elements in programs 133 quality assurance 424 quality control 424 Quimby System 12, 35, 41-2, 47, 82 dose specification 84 radial dose function 16 radiation protection 147-56 afterloading 152 quality assurance in 148—9 autoradiography and radiography 149 source identification and description 148 source integrity checks 148-9 source strength measurements 149 source handling 149-52 insertion of sources into patients 150-1 preparation of sources/applicators 149-50 removal of sources from patients 151 source and applicator cleaning 151-2 storage of sealed sources 149 transportation of sources 150 treatment delivery 151 treatment rooms 152-6 radiation protection, cost of 418 radionuclides, production of 4-5 radium 3 radium mass equivalent 12 radium substitutes 3-4 radium-226 11 radon 3 radon-22211,12 Rapid Strand 8, 9 recharging 413 reconstruction of source localization 52-7 localization using film imaging techniques 52-4 isocentric reconstruction method 53-4 orthogonal reconstruction method 52-3 reconstruction methods using correspondence lines 54 semi-orthogonal reconstruction method 53 stereo-shift reconstruction method 54 variable angle reconstruction method 54 reconstruction accuracy 55-6 using CT or MRI slices 56 using catheter describing points 56 using catheter image tracking 56 specification of coordinates 52 tracking of catheter images 54-5 rectal dose, reference point for 96-7, 98 rectum absorbed dose rate 97 high dose-rate brachytherapy in cervical cancer, effects on 366-7
Index 455
re-entrant ionization chambers 21-3 reference air kerma rate 13-14 cylindrical line sources and 15 equivalent activity and 14 radium mass equivalent and 14 spherical sources with isotropic emission and 14 reference dose 90 reference standards 19-21 reference volume 94 retroactive synchronization 165 revenue costs 411-12 rodent ulcers 393 room, treatment, see treatment room safe practice 433-42 samarium-145 4,10 scalp, skin tumors 398 scattering in irradiated medium 14—15 sealed radiation source 7 secondary traceability 19 seed sources 15 Selectron-HDR 109 Selectron-LDR112 Selectron-LDR/MDR 107 Selectron Source Dosimetry System 22 semi-fixed costs 411 semi-orthogonal reconstruction method 53 semi-variable costs 411 servicing 438-9 Sievert Integral 11-12,12, 15 sigmoid colon, high dose-rate brachytherapy in cervical cancer, effects on 366—7 singular value decomposition (SVD) 59 skin, brachytherapy 288 skin tumors, high dose-rate brachytherapy in 393-9 absorbed dose distribution 394 afterloading systems 394 clinical practice 397-8 dose fractionation schedules 397 mould production 395-7 applicator supports 396 casts 395 disposition of sources 396
dosimetry measurement 397 provisional treatment times 396-7 selection of treatment distance 395 small bowel, high dose-rate brachytherapy in cervical cancer, effects on 368 soft palate 289, 304-7 source measurements in solid phantoms 24 source strength 51 specific dose-rate constant 51 specific gamma ray constant 11, 13 specification by activity content 12-13 specification emission 13-14 squamous cell carcinoma 393 staff (labour) 439 costs 412 step costs 411 Stepping Source Dosimetry System 64-5 stereo-shift reconstruction method 54, 85 Stockholm System, carcinoma of body of the uterus and 47 sublethal damage repair (SLDR) 162-3 sunk cost 411 Syed-Neblett applicator 380, 383, 384 Syed-Neblett rectal template 389 systems of work 439 tantalum wire 106 TEM Cathetron 108-9 thermoluminescent dosimetry (TLD) 25 three-dimensional imaging techniques 75-6 time-dose factor (TDF) 189 tissue attenuation factors 52 tongue 289, 290 interstitial volume implant 302-3 tonsil 289, 306-7 top-slicing 413 total reference air kerma (TRAK) 93, 94, 95 traceability 19-21 training 440-1 treated volume 84 treatment delivery 437-8 treatment planning 435-7 information transfer to staff 436 information transfer to treatment unit 436-7
production of plan 436 treatment prescription 435 treatment rooms 152—6 design 153, 434-5 for gynecological intracavitary treatments 155 high dose-rate remote afterloading 156 intended use 153 location 153 low dose-rate/medium dose-rate remote afterloading 155 radiation and protection requirements 154 size and layout 153 types 152-3 treatment room scatter correction factors 29-30,30 trust hospitals 413 tumor-cell radiosensitivity 206 tumor control probability (TCP) 189-90, 191-2, 298 ultrasound imaging 75 three-dimensional 77 Uniformity Index (UI) 69, 72 uterus, body of, carcinoma of Manchester System and 47 Stockholm System 47 vaginal irradiation 47 variable angle reconstruction method 54 variable costs 411 VariSource 103, 109, 140, 141 volume-dose calculations 91 volume specification 82 volumes, definition of 82-4 Walstam-type sources of cesium-137 5-6, 6 water phantom 27 whole breast external-beam irradiation (WBRT) 266-79 X-ray transition point 165 ytterbium-169 4,10