Radiation Oncology: Rationale, Technique, Results, 9th Edition

1600 John F. Kennedy Blvd. Ste 1800 Philadelphia, PA 19103-2899

Radiation Oncology: Rationale, Technique, Results Copyright © 2010 by Mosby, Inc., an affiliate of Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notice Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researches must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. With respect to any drug or pharmaceutical products identified, readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindictions. It is the responsibility of practitioners, relying on their own experience and knowledge of their patients, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take all appropriate safety precautions. To the fullest extent of the law, neither the publisher nor the authors, contributors or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. The Publisher Previous editions copyrighted 2003, 1994, 1989, 1979, 1973, 1969, 1965, 1959

Library of Congress Cataloging-in-Publication Data Radiation oncology: rationale, technique, results. — 9th ed. / [edited by] James D. Cox, K. Kian Ang. P. ; cm. Includes bibliographical references and index. ISBN 978-0-323-04971-9 1. Cancer—Radiotherapy. I. Cox, James D. (James Daniel), 1938- II. Ang, K. K. (K. Kian) [DNLM: 1. Neoplasms—radiotherapy. QZ 269 R12803 2010] RC271.R3R3315 2010 616.99’40642-dc22 2009018138

Acquisitions Editor: Dolores Meloni Publishing Services Manager: Frank Polizzano Senior Project Manager: Pete Faber Design Direction: Ellen Zanolle

Printed in China Last digit is the print number:  9  8  7  6  5  4  3  2  1

ISBN: 978-0-323-04971-9

Contributors

Anesa Ahamad, M.D.

James D. Cox, M.D., F.A.C.R., F.A.S.T.R.O.

K. Kian Ang, M.D., Ph.D.

Christopher H. Crane, M.D.

Associate Lecturer, Faculty of Medical Sciences University of the West Indies at St. Augustine Trinidad and Tobago Professor and Deputy Chair, Department of Radiation ­Oncology Gilbert H. Fletcher Distinguished Memorial Chair The University of Texas M. D. Anderson Cancer Center Houston, Texas Matthew T. Ballo, M.D.

Associate Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Peter A. Balter, Ph.D., D.A.B.R.

Assistant Professor, Department of Radiation Physics The University of Texas M. D. Anderson Cancer Center Houston, Texas James D. Brierley, M.B.B.S., F.R.C.P.(UK), F.R.C.R., F.R.C.P.(C)

Professor, Department of Radiation Oncology University of Toronto Princess Margaret Hospital Toronto, Ontario Canada

Professor and Head, Division of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Bouthaina Dabaja, M.D.

Assistant Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Prajnan Das, M.D., M.S., M.P.H.

Assistant Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Marc E. Delclos, M.D.

Associate Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Lei Dong, Ph.D.

Associate Professor, Department of Radiation Physics The University of Texas M. D. Anderson Cancer Center Houston, Texas

Thomas A. Buchholz, M.D.

Patricia J. Eifel, M.D., F.A.C.R., F.A.S.T.R.O

Roger W. Byhardt, M.D.

Douglas B. Evans, M.D.

Min Rex Cheung, M.D., Ph.D.

Selim Y. Firat, M.D.

Professor and Chair, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Professor, Department of Radiation Oncology Medical College of Wisconsin Milwaukee, Wisconsin Assistant Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas

Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Professor and Chairman, Department of Surgery Medical College of Wisconsin Milwaukee, Wisconsin Assistant Professor, Department of Radiation Oncology Medical College of Wisconsin Milwaukee, Wisconsin Adam S. Garden, M.D.

Jay S. Cooper, M.D., F.A.C.R., F.A.C.R.O., F.A.S.T.R.O.

Director, Maimonides Cancer Center Chair, Department of Radiation Oncology Maimonides Medical Center Brooklyn, New York

Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas



vi

Contributors

Michael Gillin, Ph.D.

Sunil Krishnan, M.D.

Mary K. Gospodarowicz, M.D., F.R.C.P.C., F.R.C.R.(Hon)

Deborah A. Kuban, M.D.

Professor, Department of Radiation Physics The University of Texas M. D. Anderson Cancer Center Houston, Texas Professor and Chair, Department of Radiation Oncology University of Toronto Princess Margaret Hospital Toronto, Ontario Canada B. Ashleigh Guadagnolo, M.D.

Assistant Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Chul S. Ha, M.D.

Distinguished Chair in Radiation Oncology Professor and Chair, Department of Radiation Oncology The University of Texas Health Science Center at San Antonio Associate Director & Associate Physician in Chief Cancer Therapy and Research Center San Antonio, Texas

Associate Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Colleen A. Lawton, M.D., F.A.C.R.

Professor, Department of Radiation Oncology Medical College of Wisconsin Milwaukee, Wisconsin Andrew K. Lee, M.D., M.P.H.

Associate Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Valerae O. Lewis, M.D.

Associate Professor, Department of Orthopedic Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Zhongxing Liao, M.D.

Eric J. Hall D.Sc., F.A.C.R., F.R.C.R.

Higgins Professor Emeritus Columbia University New York, New York

Associate Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Michelle S. Ludwig, M.D., M.P.H.

David H. Hussey, M.D., F.A.C.R.

Professor Emeritus, Department of Radiation Oncology The University of Texas Health Science Center at San Antonio San Antonio, Texas Nora A. Janjan, M.D.

Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas

Resident, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Marvin L. Meistrich, Ph.D.

Ashbel Smith Professor, Department of Experimental Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Raymond E. Meyn, Jr., Ph.D.

Anuja Jhingran, M.D.

Associate Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas

Professor, Department of Experimental Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas

Stella K. Kim, M.D.

Luka Milas, M.D., Ph.D.

Ritsuko Komaki, M.D., F.A.C.R., F.A.S.T.R.O.

Michael F. Milosevic, M.D., F.R.C.P.C.

Associate Professor, Section of Ophthalmology Department of Head and Neck Surgery The University of Texas M. D. Anderson Cancer Center Houston, Texas Professor, Department of Radiation Oncology Gloria Lupton Tennison Endowed Professor in Lung Cancer Research The University of Texas M.D. Anderson Cancer Center Houston, Texas

Professor, Department of Experimental Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Professor, Department of Radiation Oncology University of Toronto Princess Margaret Hospital Toronto, Ontario Canada

Contributors

Radhe Mohan, Ph.D.

Professor and Chair, Department of Radiation Physics The University of Texas M. D. Anderson Cancer Center Houston, Texas

Elizabeth L. Travis, Ph.D.

Professor, Department of Experimental Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas

William H. Morrison, M.D.

Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Firas Mourtada, Ph.D.

Associate Professor, Department of Radiation Physics The University of Texas M. D. Anderson Cancer Center Houston, Texas Michael S. O’Reilly, M.D.

Associate Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas

Richard W. Tsang, M.D., F.R.C.P.(C)

Professor, Department of Radiation Oncology University of Toronto Princess Margaret Hospital Toronto, Ontario Canada Gauri R. Varadhachary, M.D., M.B.B.S.

Associate Professor, Department of Gastrointestinal Medical Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas B-Chen Wen, M.D.

Alan Pollack, M.D., Ph.D.

Chair and Sylvester Professor of Radiation Oncology The University of Miami Miller School of Medicine Miami, Florida

Professor, Department of Radiation Oncology Miller School of Medicine The University of Miami Miami, Florida

David L. Schwartz, M.D.

Robert A. Wolff, M.D.

Assistant Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Almon S. Shiu, Ph.D.

Professor, Department of Radiation Physics The University of Texas M. D. Anderson Cancer Center Houston, Texas George Starkschall, Ph.D.

Professor, Department of Radiation Physics The University of Texas M. D. Anderson Cancer Center Houston, Texas Eric A. Strom, M.D.

Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Gillian M. Thomas, B.Sc., M.D., F.R.C.P.C., F.R.C.R.(Hon)

Department of Radiation Oncology Edmond Odette Cancer Centre at Sunnybrook University of Toronto Toronoto, Ontario Canada

Professor of Medicine Department of Gastrointestinal Medical Oncology Division of Cancer Medicine The University of Texas M. D. Anderson Cancer Center Houston Texas Shiao Y. Woo, M.D.

Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas Gunar K. Zagars, M.D.

Professor, Department of Radiation Oncology The University of Texas M. D. Anderson Cancer Center Houston, Texas

vii

preface

Radiation oncology has undergone a major transformation in the past few years. Results of three-dimensional conformal radiation therapy and intensity-modulated radiation therapy have confirmed the promise that was evident from treatment-planning studies and early results of phase II and phase III trials. Advances in functional and metabolic imaging have augmented the tools for evaluating disease and planning treatments. Refinements of cytotoxic chemotherapy in conjunction with radiation therapy have reduced side effects and improved efficacy. The rapidly developing arena of molecular therapeutic agents and the emerging ability to select some agents based on molecular abnormalities in tumors promise to revolutionize combined modality therapy. Surgical procedures have emphasized preservation of structures, video assistance, and robotic tools such that the indications for adjuvant radiation therapy have been expanded and modified. This ninth edition has updated, replaced, and eliminated chapters from previous editions to seek the most contemporary discussions of principles and specific disease sites. New contributors have been enlisted and new subjects have been added. In recognition of the importance of color displays, most figures in the ninth edition are now in color. This greatly enhances the value of the figures, especially those that show dose distributions for specific diseases.

As in previous editions, we have sought consistency in the presentation of each chapter. Each begins with discussions of effects of treatments on normal tissues. Emphasis has been placed on synthesis of the available literature for clarity and brevity. The thrust of this approach is accessibility of radiation oncology to students and colleagues from other disciplines as well as physicians, physicist, and biologists from various countries. We are thankful to our colleagues who have worked arduously to meet the guidelines we have set for presenting their material in this manner. This edition has been prepared in the context of ongoing care of patients, research, and administration. We are grateful to those who have worked with us to permit its completion. We are indebted beyond words to Christine Wogan, who has edited this edition and the previous one. Her knowledge and advice has served to gather the thoughts and words of the many contributors into a more unified whole. Her patience and goodwill are without compare. James D. Cox, M.D. K. Kian Ang, M.D., Ph.D.

ix

Physical and Biologic Basis of Radiation Therapy

1

Eric J. Hall and James D. Cox HISTORICAL BACKGROUND PHYSICAL BASIS Types of Radiation Production of Radiation for Therapeutic Applications The Measurement and Meaning of Dose Physical Basis for Choice of Treatment Techniques CANCER BIOLOGY Mechanisms of Carcinogenesis The Multistep Nature of Cancer Functions of Oncogenes and Tumor Suppressor Genes Signal Transduction The Cell Cycle Cancer Genetics BIOLOGIC BASIS OF RADIATION THERAPY Radiation Chemistry DNA Breaks Chromosomal Aberrations

Cell Survival Curves Effects on Normal Tissues Effects on Tumors Tissue and Tumor Kinetics Systemic Effects Fractionation Effect of Tumor Volume on Tumor Control by Irradiation Predicting Tumor Sensitivity or Resistance to Radiation Modifiers of Radiation Effects Radiation Quality Use of Neutrons in Radiation Therapy Use of Protons in Radiation Therapy Use of Carbon Ions in Radiation Therapy Adjuncts to Irradiation Mutagenesis and Carcinogenesis

Historical Background Radiation has been an ever-present ingredient in the evolution of life on Earth. It is not something invented by human ingenuity in the technologic age; it has always been there. What is new and manmade is the extra radiation to which we are subjected, largely for medical purposes, but also from journeys in high-flying jet aircraft and from the nuclear reactors that are used to generate electrical power. X-rays were discovered in 1895 by the German physicist Wilhelm Conrad Roentgen. He found that “this new kind of ray” could blacken photographic film sealed in a container and stored in a drawer and could pass through materials opaque to light, including cardboard and wood. During a public demonstration of the production of ­  x-rays, Roentgen asked his colleague, Herr Kölliker, to put his hand in front of the x-ray machine, and with a sheet of photographic film, he made the first public radiograph displaying the bony structure of the hand. With this action, Roentgen became the father of diagnostic radiology and radiation physics. Some controversy exists about who was the first to use x-rays therapeutically. In 1897, Professor Freund demonstrated before the Vienna Medical Society the disappearance of a hairy mole by the use of x-rays, and by the turn of the century, x-rays had been used in Europe and America in primitive therapeutic applications. Parallel to the discovery of x-rays, Becquerel discovered radioactivity in 1898. Three years later, he performed what is arguably the first radiobiologic experiment when

he ­inadvertently left a container with 200 mg of radium in his vest pocket for 6 hours. He subsequently described the erythema of the skin that became evident in 2 weeks and the ulceration that developed and required several weeks to heal. During the early 1900s, radiobiologic experiments were conducted with simple biologic systems in parallel with the development of radiation therapy. One of the most wellknown results, which is still cited, was the so-called law of Bergonié and Tribondeau, which states that radiosensitivity is highest in tissues with the highest mitotic index and lowest in differentiated tissues. From 1912 to 1940, several ­investigators, first in Germany and later in the United Kingdom, demonstrated the dependence of radiation response on oxygen. The magnitude of the oxygen effect was determined by using seedlings of Vicia faba, and the possible implications for radiation therapy were discussed. In Paris in the 1920s and 1930s, famous experiments were performed in which the testes of rams were irradiated with x-rays. It proved to be impossible to sterilize the animals in a single dose without causing a severe reaction in the skin of the scrotum, but if the dose was fractionated over several weeks, sterilization could be achieved with little apparent skin damage. It was argued that the testes were a model for a rapidly growing tumor, whereas the skin represented a normal tissue response. On this basis, fractionation was introduced into clinical radiation therapy. Brachytherapy underwent a similar conceptual ­evolution after its first use in the early years of the 20th century. As





PART 1  Principles

fractionation was recognized as being advantageous in external irradiation, protraction in brachytherapy (i.e., using low-activity sources for longer periods) was thought to improve the therapeutic ratio, although radiobiologic studies of dose-rate effects were still decades away. This concept became the hallmark of the “Paris approach” to intrauterine and intravaginal radium therapy for cancer of the ­cervix. In Manchester, England, optimal arrangements of radium sources were sought to achieve a consistent dose rate and a more nearly homogeneous dose distribution through the tumor-bearing volume while sparing surrounding normal structures; this led to a more systematic application of the Paris concepts. Later, afterloading systems were developed in brachytherapy that used nonradioactive applicators so that the radioactive sources were introduced only after the desired relationships of sources were ensured; the afterloading approach reduces radiation exposure to the administering personnel. Large quantities of radium were gathered in some ­centers to produce a telecurietherapy unit (i.e., one that would permit the treatment of patients with gamma rays at a distance) in contrast to the placement of encapsulated radium in body cavities or directly into tumors (i.e., brachycurietherapy). Clinical experience with these units suggested advantages of high-energy radiation over the widely available 200 kilovolt-peak (kVp) x-ray generators. Between 1930 and 1950, technical advances permitted the development of much higher energy x-ray generators. In the 1950s, cobalt 60 (60Co) teletherapy units became widely available, and the first generation of medical linear accelerators was developed. In the wake of World War II and the use of atomic weapons on Hiroshima and Nagasaki, research in radiobiology developed rapidly. Significant milestones include the development of techniques to culture single mammalian cells in vitro in 1956 and to determine survival curves in vivo in 1959. These developments ushered in a greatly enhanced effort in radiation biology to understand conventional ­radiation therapy and suggested new horizons for improvements in treatment. In national laboratories on both sides of the Atlantic, radiation biology studies not related to ­radiation therapy were developing at the same time, involving basic studies of mutagenesis and carcinogenesis.

Physical Basis Types of Radiation Types of radiation of concern in this book are those with the capacity to produce ionizations and excitations during the absorption of energy in biologic material. The raising of an electron in an atom or molecule to a higher energy level, without the ejection of that electron from the atom or molecule, is called excitation. If the radiation has sufficient energy to eject one or more orbital electrons from the atom or molecule, the process is referred to as ionization, and the radiation is said to be ionizing radiation. The important characteristic of ionizing radiation is the localized release of large amounts of energy. The energy dissipated by an ionizing event is approximately 33 electron volts (eV), which is more than enough to break a strong chemical bond; for example, the energy associated with a carbon-carbon bond is 4.9 eV. Ionizing radiation produces substantial biologic effects for the relatively small total

amounts of energy involved because the energy is released locally in “packets” large enough to break chemical bonds and initiate the chain of events that lead ultimately to a biologic effect.

Electromagnetic Radiation Electromagnetic radiation (x-rays and gamma rays) is indirectly ionizing. These types of radiation do not themselves produce chemical and biologic damage, but when they are absorbed in the medium through which they pass, they give up their energy to produce fast-moving electrons by the Compton, photoelectric, or pair-production processes (Fig. 1-1). X-rays and gamma rays are forms of electromagnetic radiation that do not differ in nature or properties; the designation of x or gamma reflects the way in which they are produced. X-rays are produced extranuclearly, which means that they are generated in an electric device that accelerates electrons to high energy and then stops them abruptly in a target, made usually of tungsten or gold. Part of the kinetic energy, or energy of motion of the electrons, is converted into photons of x-rays. Gamma rays are produced intranuclearly (i.e., emitted by radioactive isotopes); they represent excess energy that is given off as the unstable nucleus breaks up and decays in its efforts to reach a stable form.

Particulate Radiation Other forms of radiation that are used experimentally and used or contemplated for radiation therapy include electrons, protons, alpha particles, neutrons, negative ­­ pi­mesons, and high-energy heavy ions. Electrons are light, negatively charged particles that can be accelerated to high energy and to a speed close to that of light by means of an electrical device such as a betatron or linear accelerator. Protons are positively charged particles with a mass nearly 2000 times that of an electron. Protons require more complex and expensive equipment to accelerate them to useful energies. Protons are increasingly used for radiation therapy because of the favorable dose distributions that can be obtained. For example, 160-megaelectron volt (MeV) protons have a range of about 12 cm in tissue. Alpha particles are nuclei of helium atoms, each consisting of two protons and two neutrons in close association. Because they have a net positive charge, they can be accelerated in large electrical devices similar to those used for protons. Alpha particles are also emitted during the decay of some radioactive isotopes. Neutrons are particles having a mass similar to that of protons, but they carry no electrical charge. Because they are electrically neutral, they cannot be accelerated in an electrical device but rather are produced when a charged particle, such as a deuteron or proton, is accelerated to high energy and then made to impinge on a suitable target material. Neutrons are indirectly ionizing because the first step in their absorption is for them to collide with nuclei of the atoms of the absorbing material and produce recoil protons, alpha particles, or heavier nuclear fragments. These charged particles are responsible for the biologic effects. Neutron beams have been used experimentally for radiation therapy at several centers because of their theoretical advantage over photons in low-oxygen (hypoxic) conditions, but interest in the use of neutron therapy has waned

1  Physical and Biologic Basis of Radiation Therapy COMPTON EFFECT

PHOTOELECTRIC EFFECT

e–

Vacancy in k-shell

e– e–

p +np + np + n p + np + np + n

e–

e–

e

id

c In

n

to

ho

e–

Incident photon

e–

e–

p nt

Fast electron

p +np + np + n p + np + np + n

e–

e–

Sc



Fast electron e–

e–

atte

red

pho

e–

ton

Characteristic x-rays PAIR PRODUCTION e–

Incident photon

e– p +np + np + n p + np + np + n

e–

e– e–

e+

Positron electron pair

e– e–

FIGURE 1-1.  The first step in the absorption of a photon of x-rays or gamma rays is the conversion of the energy of the photon into the kinetic energy of an electron, or electron-positron pair. At higher energies, when the energy of the incident photon greatly exceeds the binding energy of the planetary electrons in the atoms of the absorber, the Compton process dominates. The photon interacts with the electron in a classic “billiard-ball collision.” Part of the photon energy is given to the electron as kinetic energy, while the photon is deflected and has reduced energy. At lower energies, when the binding energy of the planetary electrons of the atoms of the absorber is not small compared with the photon energy, the photoelectric effect is most important. The photon disappears completely as it interacts with a bound electron. The electron is ejected with kinetic energy equal to the photon energy, less the energy required to overcome the electron bond. The vacancy caused by the removal of the electron must be filled by an electron dropping from an outer orbit, giving rise to a photon of characteristic radiation. At sufficiently high photon energies, the photon may interact with the powerful nuclear forces to produce an electron-positron pair. The first 1.02 MeV of photon energy is used to create the rest mass of the pair, and the remainder is distributed equally between them as kinetic energy.

because it has shown no advantages in terms of outcome over irradiation with other types of particles. Negative pi-mesons are negatively charged particles with a mass 273 times that of an electron. They are produced by a complex process that necessitates a huge linear accelerator or synchrocyclotron capable of accelerating protons to energies of 400 to 800 MeV. When pi-mesons are absorbed in biologic material, they behave like overweight electrons as long as they are relativistic (i.e., as long as their velocity is close to that of light). However, when they slow down, they spiral down the energy levels of an absorbing atom and are absorbed by the nucleus of that atom, which then explodes to produce fragments consisting of neutrons, alpha particles, and larger nuclear fragments. Radiation therapy with negative pi-mesons has been attempted experimentally in several countries but has been abandoned. Heavy ions are nuclei of elements such as nitrogen, carbon, neon, argon, or silicon that are positively charged because some or all of their planetary electrons have been stripped from them. To be useful, they must be accelerated to energies of thousands of millions of volts and can therefore be produced in only a very limited number of laboratories in the world. Most of the interest in heavy ions for biologic applications has focused on carbon ions, which combine the dose-distribution advantages of protons with an increase in biologic effectiveness toward the end of the

particle range. This topic is discussed further in the “Use of Carbon Ions in Radiation Therapy” section later in this chapter.

Production of Radiation for Therapeutic Applications When x-rays or gamma rays enter biologic material, energy is converted into chemical damage and heat. At the energy levels of most x-ray and gamma-ray sources in use, the primary events are the interactions of photons with electrons in the outer shells, resulting in scattering of the photons and electrons (i.e., Compton scattering). With higher-­energy photons, secondary electrons scatter more in the forward direction (i.e., in the direction of the primary beam). It takes some distance for the interactions to summate and reach a maximum, after which the energy of the beam dissipates by a constant fraction per unit depth. Figure 1-2 compares depth-dose characteristics of several radiation beams commonly used in radiation therapy. The insert in this figure demonstrates the physical basis for skin sparing; high-energy photons, unlike conventional x-rays, deposit the maximum dose below the skin surface. The most commonly used sources for external irradiation are listed in Table 1-1. Much overlap exists among the teletherapy sources. For example, there is relatively little difference in depth doses between the gamma rays from 60Co and 2- to 6-megavolt (MV) x-rays. However, the



PART 1  Principles 100

100 Percent depth dose

Protons with range shifter

90

Percent depth dose

80 70 Monoenergetic protons

60

80

2 mm Cu

6 MV

18 MV

60 40 20

50

0 0 2 4 6 8 10 12 14 16 18 20

40

Depth (mm)

18 MV

30 20

12 6 MeV MeV

10 0 0

5

10

15 Depth (cm)

6 MV

D-T neutrons

20 MeV

60Co

2 mm Cu 20

25

30

FIGURE 1-2.  Percentage depth-dose curves are shown for a variety of radiation types used in radiation therapy. These include xrays and γ-rays up to 18 MV, electrons of 6 to 20 MeV, and protons. The inset shows the pattern of absorption at shallow depths and provides a rationale for the skin-sparing effect. Cu, copper; MeV, megaelectron volt; MV, megavolt.

TABLE 1-1.    Teletherapy Sources

TABLE 1-2.    Therapeutic Isotopes

Unit

Mean Energy (MeV) of Photons

Isotope

150-440 kVp x-ray

0.06-0.14

137CS

0.66

60Co

teletherapy

teletherapy

1.25

4 MV linear accelerator

1.3

6 MV linear accelerator

1.8

20-24 MV betatron and linear ­accelerator

6.2-7.0

kVp, kilovolt peak; MeV, million electron volts; MV, megavolts.

edge of the beam produced by a linear accelerator is much sharper than that from cobalt units, which may be an important factor when radiation is delivered close to critical structures, such as the lens of the eye. The sources most widely used for intracavitary or interstitial therapy are listed in Table 1-2. The various brachytherapy sources have overlapping capabilities, but some have specific advantages. For example, iridium 192 (192Ir) is the only isotope listed that has been widely and satisfactorily used with an afterloading technique for interstitial therapy. Radium and cesium can be used in afterloading intracavitary applicators. Gold, iodine, and palladium can be used for permanent interstitial implants. When used appropriately and meticulously, teletherapy units and brachytherapy sources have had great success in the permanent control of a variety of malignant diseases. However, there is nothing inherent in any of these x-ray or gamma-ray generators that can ensure success independent of the expertise of the radiation oncologist.

The Measurement and Meaning of Dose Quantification of dose has been an important factor in the development of modern radiation therapy. The ­ability to prescribe and deliver specific amounts of radiation

Half-life

Energy (keV)

226Ra

1620 yr

830*

137Cs

30 yr

662*

198Au

2.7 days

412*

192Ir

73.8 days

370*

125I

60 days

28*

103Pd

16.97 days

21*

Photon

Beta 32P

14.3 days

1710†

90Sr/90Y

28.5 yr/2.7 days

550/2280†

188W/188Re

69.4 days/17 hr

350/2120†

186Re

3.8 days

1070†

62Zn/62Cu

9.3 hr/9.7 min

660/2930†

133Xe

5.2 days

360†

131I

8.0 days

600†

89Sr

50.5 days

1495†

166Ho

26.8 hr

1850†

*Average

energy. energy. keV, kiloelectron volt.

†Maximum

r­ epeatedly with precision and accuracy is essential if a ­given treatment is to be duplicated and if clinical data are to be accumulated and compared. Biologic effects in laboratory research and in clinical practice correlate well with absorbed dose, expressed as energy per unit mass of tissue. This correlation holds true for all types of radiation used in conventional radiation therapy, such as x-rays, gamma rays, fast neutrons, and electrons, but it is probably less satisfactory for very-high-energy heavy ions.

1  Physical and Biologic Basis of Radiation Therapy

For many years, the unit of dose commonly used was the rad, defined as an energy absorption corresponding to 100 erg/g. This unit has been replaced by the gray (Gy), defined as energy absorption of 1 J/kg. One gray is equivalent to 100 rad.

Measurement of Dose Dose can be measured directly, in terms of energy absorbed per unit mass, by measuring the temperature rise in the absorbing material. However, this can be done in only a few research centers because doses of several grays result in a temperature rise of only a few hundredths of a degree Celsius, which is exceedingly difficult to measure. For practical day-to-day use, what is measured is the ability of the x-rays or gamma rays to ionize air because such measurements are sensitive and relatively easy to make. Absorbed dose is then calculated on the basis of the ionization measurements. An ionization chamber consists of a cavity containing a gas that is surrounded by tissue-equivalent material. A central electrode within the cavity is maintained at a high ­­voltage relative to the cavity walls. Because the gas is an insulator, no charge passes across the gas until x-rays or gamma rays pass through the cavity and produce ionizations. The collection of the charges produced leads to the current between the central electrode and cavity walls, which can be detected with a sensitive instrument. In practice, most of the ionizations occur in the cavity wall and result in electrons that pass through the cavity. If the ionization chamber is connected across a capacitor, the dose meter is an integrating dose meter (i.e., it measures the ­total amount of incident radiation) (Fig. 1-3). If the chamber is connected across a



resistor and the current through that resistor is measured, the dose meter measures the instantaneous dose rate.

Patterns of Energy Deposition: Microdosimetry Dose is an average quantity (i.e., the average energy deposited per unit mass of tissue). However, at a microscopic level, energy from ionizing radiation is not deposited uniformly but tends to be localized along the tracks of charged particles. What is critical to the biologic effect produced by a given amount of radiation is the pattern of energy ­deposition. Figure 1-4 shows the tracks of secondary charged particles in cells irradiated with 1 centigray (cGy) of x-rays or 1 cGy of neutrons. In the case of x-rays, the charged particles set in motion are electrons, whereas in the case of neutrons, the secondary charged particles are protons.  A marked difference exists between the two. A dose of  1 cGy of x-rays results in essentially all cells being traversed by several particles. For neutrons, only a few cells are traversed by a single particle. Consequently, when the dose of x-rays is doubled, the average energy per cell is increased; however, in the case of neutrons, the number of cells traversed is increased, but the amount of energy per cell traversed does not change. This is an essential difference ­between the two types of radiation. Biologic effects correlate with dose for a given type of radiation (i.e., a bigger dose leads to a bigger biologic effect), but the same dose of a different type of radiation does not produce equal biologic effects because of the considerations referred to earlier. In other words, the different relative biologic effectiveness (RBE) of different types of radiation is largely a reflection of the different patterns of energy deposition.

Physical Basis for Choice of Treatment Techniques

DOSE-RATE METER Thimble chamber

R

INTEGRATING DOSE METER Thimble chamber

C

FIGURE 1-3.  An ionization chamber consists of a cavity fabricated from tissue-equivalent plastic and filled with an appropriate gas. A voltage is applied between the central electrode and the wall of the cavity. The gas is an insulator so that no current flows unless x-rays pass through the gas and ionize the gas into positive and negative ions that are collected by the voltage across the chamber. The charge generated is proportional to the x-ray intensity. If the charge is passed through a resistor (R), the voltage generated across it is a measure of the instantaneous dose rate. It is then a dose-rate meter. If the charge is passed onto a condenser (C), the change in voltage across the condenser is a measure of the total dose. It is then an integrating dose meter.

The determinants for choosing particular techniques in clinical radiation therapy are the limits of localization of the malignant tumor and especially its subclinical extensions and the availability of particular sources of radiation. Except for the few institutions that have accelerators ­capable of producing beams of heavy charged particles, treatment choices are based on the external beams listed in Table 1-1 or the intracavitary sources listed in Table 1-2. The particular techniques that are suitable for individual malignant tumors are found in chapters elsewhere in this book, but a few generalizations can be made here. The intent when administering ionizing radiation for the purpose of killing all cells in a malignant tumor is to deliver a dose that is sufficient to eradicate these cells but is limited with respect to the surrounding normal tissues. In light of the preceding discussion of dose and its meaning, it is worth considering the appropriate designations of dose in the clinical setting. The term tumor dose is inappropriate because dose distribution is rarely homogeneous throughout the tumor-containing volume. It is more appropriate to refer to minimum tumor dose because this dose determines control of the tumor. Similarly, it is most meaningful to refer to maximum doses in specific normal tissues surrounding the tumor because these doses determine the adverse effects of radiation therapy. In all cases, it is virtually meaningless to express a total dose without qualifying it by the fractionation regimen used (fractionation is discussed further later in this chapter).



PART 1  Principles 10-mGy (1 rad) x-rays

10-mGy (1 rad) neutrons

FIGURE 1-4.  When a population of cells is irradiated with 1 cGy of x-rays, many charged particle tracks traverse every cell. In the case of neutrons, only a small proportion of the cells are traversed by a charged particle track. Consequently, when the x-ray dose is increased, the average energy deposited per cell is increased. When the neutron dose is increased, the number of cells in which energy is deposited is increased. This is a fundamental difference between x-rays and neutrons.

A more detailed discussion of dose distributions and the related subject of selection of single fields of specific radiation beams or the use of combinations of fields is beyond the scope of this chapter, but these topics are addressed in standard texts. One principle of choosing treatment techniques can be discussed here. Given the contemporary availability of radiation beams, computers, and treatment-planning software, the temptation is to seek refinements of dose distribution that are conceptually satisfying but have little or no ­clinical importance. Such temptations are best resisted in the interest of daily reproducibility. To paraphrase a ­concept ­ espoused by the 14th-century philosopher, William of Occam, “complexity is not to be assumed without necessity.”

Cancer Biology Cancer is thought to be a clonal disorder because most types of cancer seem to arise from a single cell that has suffered a disruption in its regulatory mechanisms for proliferation and self-elimination. Malignant cells differ from their normal tissue counterparts in several ways. First, they are often immortal or at least have the ability to divide many more times than do cells from normal somatic tissues. Second, they often grow more rapidly than do cells of the same tissue of origin. Third, they have abnormal cellto-cell interactions, which result in their being able to invade, metastasize, and grow in foreign environments.

Mechanisms of Carcinogenesis Carcinogenesis seems to be a multistep process, with multiple genetic alterations occurring over extended periods. Sometimes, genetic alterations are carried in the germline and form cancer-predisposing syndromes (described later); however, heritable mutations are the exception. Most of

the alterations that lead to cancer are acquired in the form of somatic mutations. Cell proliferation is controlled through a series of positive and negative signals that affect cell division and differentiation. The acquisition of tumorigenicity results from changes that affect these signal-control points. The conversion of a cell to a malignant state can result from gain-of-function mutations that activate oncogenes, which are positive growth regulators; loss of function of a tumor suppressor gene product that normally acts as a negative growth regulator; or mutations that affect death-regulating genes that control apoptosis (i.e., programmed cell death). Oncogenes are found in a mutated or abnormally expressed form in many human cancers.1 Oncogenes can be activated by a point mutation, by a chromosomal rearrangement such as a translocation, or by gene amplification. Some human leukemias and lymphomas seem to be caused by specific chromosomal translocations that lead to oncogene activation in several different ways. Because oncogenes are gain-of-function mutations, only one of the two copies needs to be activated for their effects to be ­expressed; oncogenes act in a dominant (rather than a recessive) fashion. The concept of tumor suppressor genes arose from the observation that the in vitro fusion of a normal human ­fibroblast with a HeLa cervical carcinoma cell suppressed the expression of the malignant phenotype of the HeLa cell. Of the many tumor suppressor genes whose location and function are known, two of the most intensively studied are TP53 (formerly designated p53) and RB1 (formerly designated Rb).2,3 In contrast to oncogenes, tumor suppressor genes involve loss-of-function mutations, and both copies of the gene must be lost for their effects to be expressed (i.e., they act in a recessive fashion). In practice, the loss of both copies of a gene often occurs by somatic ­homozygosity, which is the process by which one chromosome of a pair is

1  Physical and Biologic Basis of Radiation Therapy

lost, a deletion occurs in the remaining chromosome, and the chromosome with the deletion replicates. A third class of genes associated with some forms of cancer affects apoptosis. The most relevant example is the BCL family of genes, which includes BCL2, BCL2L1 (formerly designated BclX), and BAX. BCL2 is deregulated by the chromosomal translocation t(14;18), which blocks apoptosis and contributes to the development of follicular lymphoma.

The Multistep Nature of Cancer The idea that carcinogenesis is a multistage process in which several distinct events occur, separated in time, is almost 70 years old. Skin cancer experiments in mice led to the development of the concepts of initiation, promotion, and progression as being stages in tumor development.  Genetic analysis of cells from solid tumors also suggests that multiple alterations, mutations, or deletions are ­needed in multiple signaling genes for cancer to form. In the case of colorectal cancer, a model has been proposed that correlates a series of 6 to 12 chromosomal and molecular events producing changes in the histopathology of normal epithelium during the formation and metastasis of colorectal cancer.4 These events are thought to involve a “mutator” phenotype in which multiple mutations are likely in cancer-associated genes (i.e., oncogenes and tumor suppressor genes). This leads to progression of the cancer and ultimately to its invasive and metastatic properties.

Functions of Oncogenes and Tumor Suppressor Genes The identification of genes associated with human ­cancer quickly led to studies of the function of these genes. Cells must repeatedly “decide” whether to divide, differentiate, or undergo apoptosis. Because all three outcomes affect cell number, it is not surprising that these decision pathways are the primary targets of oncogenes or tumor ­suppressor genes. Most tumor suppressor genes can be classified as gatekeepers or caretakers. Gatekeepers are genes that directly regulate the growth of tumors by inhibiting cell division or by promoting cell death. The function of these genes is rate limiting for tumor initiation and growth; both  alleles ­(maternal and paternal) must be lost or inactivated for a tumor to develop. Predisposed individuals inherit one damaged copy of such a gene and require only one additional mutation for tumor initiation. The identity of gatekeeper genes varies among tissues, such that inactivation of a given gene leads to predisposition to specific forms of cancer. For example, inherited mutations in the APC gene lead to ­colon cancer, and mutations in the VHL gene predispose an individual to develop kidney cancer. ­Because these gatekeeper genes are rate limiting for tumor ­initiation, they ­often tend to be mutated in sporadic cancer through somatic mutations and in the germlines of predisposed individuals. Inactivation of caretaker genes, in contrast, does not directly promote the growth of tumors but leads instead to genomic instability, thereby indirectly promoting growth by causing an increase in mutation rate. Such an increase in genetic instability can greatly accelerate the development of cancer, especially those types—such as colon cancer—that require numerous mutations for their full ­ development.



The targets of the accelerated mutation rate in cells with defective caretakers are the gatekeeper tumor suppressor genes or oncogenes, or both.

Signal Transduction An elaborate system of biochemical communication networks exists in eukaryotic cells to provide information to the nucleus regarding the environment outside the cell. Signal transduction is the term used to describe the flow of information from the cell’s outer membrane through the cytoplasm and into the nucleus. Multiple signal transduction cascades can operate simultaneously, with variable amounts of cross-talk occurring between them. Even a weak signal can elicit a powerful response because it is amplified by biologic processes in much the same way as a radio signal is amplified by an electronic device. Signals in the form of cytokines, low-molecular-weight hormones, growth factors, and other proteins that arrive at the outer plasma membrane can affect the cell by two mechanisms. In the first, the signaling factor binds to a specific cell surface receptor that spans the plasma membrane. On binding, these receptor molecules transduce a signal from the extracellular domain to the cytoplasmic domain on the inner surface of the plasma membrane; this transduction often leads to the activation of an enzyme such as a protein kinase. The second mechanism involves the migration of the signaling agent itself across the plasma membrane, followed by its binding to specific receptor molecules in the cytoplasm. In either case, a biochemical cascade ultimately ­allows the signal to reach its destination—the nucleus—and ­alter cellular behavior. These signal transduction processes involve signaling molecules that are bound together by ­protein-protein interactions and by the production of ­second-messenger molecules. One important signaling pathway involves GTP-binding proteins such as those produced by the RAS family; the RAS gene is often mutated in human cancer. Another important pathway is that associated with protein kinase C. Ionizing radiation can activate nuclear and membrane-cytoplasmic signal-transduction pathways; damage to deoxyribonucleic acid (DNA) modulates the transcriptional activity of TP53 through the activity of several kinases and by protein-protein interactions. These and other perturbations to signaling pathways can produce cell cycle arrest or apoptosis.

The Cell Cycle The ability of cells to produce exact, accurate copies of themselves is essential to the continuance of life; it is accomplished through highly organized processes that have been well conserved through evolution. Lack of fidelity in cellular reproduction, as manifested by alterations in DNA and chromosomes, is a hallmark of cancer. The division of the mammalian cell cycle into its constituent phases  (M, G1, S, and G2) was accomplished by radiation biologists in the 1950s. As detailed under “Tissue and Tumor Kinetics,” it is now known that progression through the cell cycle is governed by protein kinases, which are activated by cyclins. Each cyclin protein is synthesized at a discrete phase of the cycle—cyclins D and E in G1, cyclin A in S and G2, and cyclin B in G2 and M. Transitions in the cycle occur only when a given kinase activates the proteins required for progression.

10

PART 1  Principles

The arrest of cells at various positions in the cycle by the action of “checkpoint genes” is an important response to DNA damage. The two principal checkpoints are the G1/S and the G2/M boundaries, but a checkpoint also exists in S. G2/M is the best-known checkpoint after radiation damage; cells pause at G2/M to repair damage before initiating the complex process of mitosis. The hallmark of cancer is a lack of the ability to respond to signals that would normally cause the cell to stop progressing through the cycle and dividing. Of the tumor suppressor genes discovered so far in human cancers, only RB1 and TP53 seem to be directly implicated in cell cycle control; clearly, much more remains to be learned on this topic.

Cancer Genetics The link between cancer risk and genetic factors varies among tumor types. Up to 30% of rare childhood cancers occur in individuals who are predisposed to the condition, whereas only about 5% to 10% of common adult cancers are thought to be hereditary. Many instances of cancer predisposition are characterized by a distinctive pattern involving several different types of cancer in several different organs, sometimes accompanied by unusual physical features. In general, a family history of cancer, early age at onset, or the presence of multiple primary tumors strongly suggests the presence of a genetic disorder. Germline mutations can result in a predisposition to cancer in any of four ways. Most cases of hereditary cancer predisposition can be traced to germline mutations in tumor suppressor genes. (Tumor suppressor genes were first discovered through a study of familial predisposition to retinoblastoma.) Germline mutations that activate oncogenes are a second possible mechanism for hereditary predisposition to cancer, but the only disorders suspected of arising through this mechanism are multiple endocrine neoplasias. Medullary thyroid carcinoma and pheochromocytoma are a feature of these autosomal dominant syndromes, and physical abnormalities often accompany the malignancy. Third, any syndrome that includes DNA-repair defects is likely to increase the probability of mutations in genes that lead to cancer. Xeroderma pigmentosum, ataxia telangiectasia, Bloom syndrome, and Fanconi anemia are examples of relatively uncommon autosomal recessive syndromes that involve faulty DNA repair or replication.  A particularly interesting member of this class is hereditary nonpolyposis colon cancer, which is caused by a mutation in one of five mismatch repair genes; a mutation in one of these genes leads to a general instability known as the mutator phenotype. A fourth, less well understood category of hereditary disorders predisposes to cancer through unusual sensitivity to common carcinogens. For example, lung cancer is usually considered to have only a small genetic component, but genetic variations in the metabolism of the carcinogens in tobacco smoke may be important in determining who develops a malignancy. Genetic variations also may influence susceptibility to hepatic cellular carcinoma caused by aflatoxins. Several hereditary disorders are known to predispose to cancer, but understanding of those disorders at the molecular level is at various stages. Retinoblastoma displays the clearest pattern of a familial and sporadic form. Some researchers speculate that when all of the information is in, all common cancers will turn out to have familial and

sporadic components but that the genetics will be more complicated and the pattern far less obvious than those of retinoblastoma.

Biologic Basis of Radiation Therapy Radiation Chemistry Strong circumstantial evidence exists to suggest that DNA is the principal target for radiation damage that leads to the loss of reproductive integrity of a cell, although the discovery of the bystander effect has shown that damage can be passed from a “hit” cell to a “nonhit” neighboring cell through gap junction communications.5-8 When x-rays or gamma rays are absorbed in biologic material, the first step in the absorption process is the conversion of their energy to fast-moving recoil electrons. When neutrons are absorbed, the first step is the conversion of that energy into protons, alpha particles, or heavy nuclear fragments. Radiochemical lesions in the DNA may result from the direct action of these charged particles or from indirect action mediated by highly reactive free radicals (Fig. 1-5). Absorption of energy from direct action requires ionizations to occur within the DNA molecule itself. This is

INDIRECT ACTION n

to

o Ph

H OH• S A TS H P P G SP S C P S T S A P P S G CS

O e– p+

n

to

o Ph e– p+

2 nm 4 nm

DIRECT ACTION

FIGURE 1-5.  Direct and indirect actions of radiation are depicted. The structure of DNA is shown schematically. In direct action, a secondary electron resulting from absorption of an x-ray photon interacts with the DNA to produce an effect. In indirect action, the secondary electron interacts with, for example, a water molecule to produce a hydroxyl radical (OH·), which produces the damage to the DNA. The DNA helix has a diameter of about 20 Å (2 nm). It is estimated that free radicals produced in a cylinder with a diameter double that of the DNA helix can affect the DNA. Indirect action is dominant for ­sparsely ionizing radiation, such as x-rays. A, adenine; C, ­ cytosine; G, guanine; P, phosphorus; S, sugar; T, thymine. Modified from Hall EJ, Giaccia AJ. Physics and chemistry of radiation absorption. In: Radiobiology for the Radiologist. 6th ed. Philadelphia: Lippincott Williams & Wilkins, 2006, pp 5-15.

11

1  Physical and Biologic Basis of Radiation Therapy

the dominant process for radiation of high linear energy transfer (LET), such as neutrons or alpha particles. LET is discussed further under “Radiation Quality.” Alternatively, the radiation may interact with other atoms or molecules in the cell, particularly water, to produce free radicals that can diffuse far enough to reach and damage the critical targets. A free radical is an atom or molecule carrying an unpaired electron in its outer shell. The most important of these is probably the hydroxyl radical (OH•). Free radicals have a lifetime of about 10−5 seconds, which is long relative to the time for the initial interaction of the radiation with the atoms and molecules of the absorber (10−15 seconds). Experiments in which free-radical scavengers are added to biologic material during irradiation suggest that about two thirds of the damage produced by x-rays or gamma rays is mediated by free radicals. In the case of radiation of higher LET, the relative balance moves away from indirect action in favor of direct action. The component of the radiation damage mediated by free radicals is readily amenable to modification by the presence or absence of oxygen or other chemical compounds. In contrast, direct action cannot be modified by chemical means.

DNA Breaks DNA is a large molecule assembled as a double helix. Its two helical strands are held together by hydrogen bonds between the bases. The “backbone” of each strand consists of alternating sugar-phosphate groups. The sugar involved is deoxyribose. Attached to this backbone are four bases, the sequence of which specifies the genetic code. Two of the bases—thymine and cytosine—are single-ring groups (i.e., pyrimidines); the other bases—adenine and guanine— are double-ring groups (i.e., purines). The structure of a single strand of DNA is illustrated in Figure 1-6. Bases on opposite strands must be complementary; adenine pairs with thymine, and guanine pairs with cytosine (Fig. 1-7A). When cells are irradiated with x-rays, many breaks occur in single strands. These breaks can be observed and scored as a function of dose if the DNA is denatured and the supporting structure is stripped away. However, in intact DNA, single-strand breaks are of little biologic

consequence as far as cell death is concerned because  single-strand breaks are readily repaired using the opposite strand as a template (see Fig. 1-7B). If the repair is incorrect (i.e., misrepair), it may result in a mutation. If both strands of the DNA are broken but the breaks are well separated (see Fig. 1-7C), repair again occurs readily because the two breaks are handled separately. If the breaks in the two strands are opposite one another or are separated by only a few base pairs (see Fig. 1-7D), this may lead to a double-strand break (DSB) (i.e., the piece of chromatin snaps into two pieces). When this occurs as the result of the passage of a charged particle, considerable damage can be produced in the region, producing a “locally multiple damaged site.”6 A DSB is thought to be the most important lesion produced in chromosomes by radiation; as described in the next section, the interaction of two DSBs may result in cell death, mutation, or carcinogenesis.

Chromosomal Aberrations The most important of the chromosomal aberrations are exchange-type aberrations, which are formed when a DSB in each of two chromosomes, or chromosome arms, rejoins in an illegitimate way. A detailed discussion of cytogenetics is beyond the scope of this chapter; a variety of complex aberrations can occur, but in this chapter, attention focuses on four types of aberrations that are of special importance (Fig. 1-8). The consequences of DSBs in two separate prereplication chromosomes are shown in Figure 1-8A and B. In Figure  1-8A, the rejoining occurs in such a way as to produce a

G

A

G

B

P

O

CH2

O Sugar (deoxyribose)

O O

Phosphate

C

G

O O

C

P

Adenine

C

O

H O

Cytosine CH2

O O O

C

P

G

O

D

H O

Thymine CH2 O

O

O O

P O

H O

CH2

Guanine O

OH

H

FIGURE 1-6.  Structure of a single strand of DNA.

C

C

T

T

G

A

A

C

T

T

G

A

C

T

G

A

C G

A T

A T

T

A T

T A

T A

C G

C G

C G

A

A

G

T

T

C

A

A

G

T

T

C

T A

T

A T

A

C

T

T

G

A

G

A

A

T

A T

G C

T A

FIGURE 1-7.  Single- and double-strand breaks are caused by radiation. A, Two-dimensional representation of the normal DNA double helix. The based pairs carrying the genetic code are complementary; adenine pairs with thymine, and guanine pairs with cytosine. B, A break in one strand is of little significance because it is readily repaired, using the opposite strand as a template. C, Breaks in both strands, if well separated, are repaired as independent breaks. D, If breaks occur in both strands and are directly opposite each other or are separated by only a few base pairs, a double-strand break may occur in which the chromatid snaps into two pieces.

12

PART 1  Principles Translocation

Chromosome 9

Chromosome 22

A

Dicentric

B FIGURE 1-9.  A symmetric translocation between chromosomes 9 and 22 may bring together the BCR and ABL1 genes, a change associated with chronic myelogenous leukemia. Modified from Hall EJ, Giaccia AJ. Cancer Biology. In: Radiobiology for the Radiologist. 6th ed. Philadelphia: Lippincott Williams & Wilkins, 2006, pp 274-302.

C Ring

Dic

D Trans Deletion

FIGURE 1-8.  Four types of chromosomal aberrations can ­occur. A, If breaks occur in two prereplication chromosomes, a symmetric exchange may occur. This is consistent with cell viability, but the process can lead to activation of an oncogene. B, Breaks in two prereplication chromosomes may reconstitute to form a dicentric and an acentric fragment. This aberration is lethal to the cell at a subsequent division. C, Breaks in the two arms of a prereplication chromosome may rejoin to form a ring and an acentric fragment. This aberration is lethal to the cell. D, Two breaks in the same arm of a prereplication chromosome may result in a deletion. If the deletion is small, it may be consistent with cell viability. If the deleted piece of DNA includes a suppressor gene, it may result in carcinogenesis.

FIGURE 1-10.  Fluorescence in situ hybridization (FISH) uses probes to “paint” a given chromosome with a characteristic color. Multiplex-FISH paints each chromosome with a different color, enabling symmetric translocations to be then readily visualized. (Courtesy of Charles R. Geard, MD, Columbia University, New York, NY.)

symmetric translocation. This is compatible with cell viability but is occasionally associated with carcinogenesis. For example, if breaks occur in just the right places on chromosomes 9 and 22, the symmetric translocation that occurs may bring together the genes BCR and ABL (Fig. 1-9). This translocation is associated with chronic myelogenous leukemia. Symmetric translocations and other chromosomal abnormalities can be visualized in human cells by the technique of multiplex fluorescent in situ hybridization (FISH), in which the chromosomes are “painted” in different colors (Fig. 1-10). In Figure 1-8B, the two broken chromosomes rejoin in such a way as to produce a dicentric fragment (i.e., chromosome with two centromeres) and an acentric fragment. The acentric fragment will be lost at a subsequent mitosis 

because it has no centromere and therefore cannot be segregated ­ correctly to the daughter cells. The formation of a dicentric fragment is usually lethal to the cell. Another lethal aberration is the ring chromosome, the formation of which is illustrated in Figure 1-8C. In this case, a DSB is produced in each arm of the same prereplication chromosome. The rejoining of the “sticky” ends may result in an acentric fragment and a ring. This aberration is usually not consistent with cell viability. ­Figure 1-8D illustrates the formation of an interstitial deletion in which two DSBs occur close together in the same arm of the same chromosome. This may lead to the loss of a section of genetic material, which may not affect cell viability. However, an interstitial deletion may be a mechanism of carcinogenesis if the deleted genetic material happens to be a tumor ­  suppressor gene.

r+f

1  Physical and Biologic Basis of Radiation Therapy

Cell Survival Curves A cell survival curve describes the relationship between the absorbed dose of radiation and the proportion of cells that survive in the sense that they are able to grow into a colony, thereby demonstrating retention of their reproductive integrity.10 With modern tissue culture techniques, it is possible to establish cell lines from many different tissues and tumors. If these cells are seeded as single cells and exposed to radiation or to some other cytotoxic agent, it is then possible to count the proportion of cells that are able to form macroscopic colonies after graded doses of the cytotoxic agent. Cell survival curves for mammalian cells have a characteristic shape. If surviving fraction is plotted on a logarithmic scale against dose on a linear scale, the curve has an initial slope at low doses, at which the surviving fraction seems to be an exponential function of dose (Fig. 1-11). As the dose is increased, the curve bends and becomes progressively steeper. At very high doses, the curve tends to become straight again (i.e., the surviving fraction returns to being an exponential function of dose). Over the first few decades of survival, survival curves for mammalian cells closely approximate that of a linear quadratic form expressed as follows: S  eBD  CD

2

In the equation, S is the fraction of cells surviving a dose (D), and α and β are constants. Over a wider range of doses, a more complex relationship that contains at least three parameters usually proves to be a better fit. The parameters include an initial slope (D1), a final slope (D0), and some quantity that is a measure of the width of the shoulder. This quantity can be the extrapolation number (n) or the quasi-threshold dose (Dq). For densely ionizing radiation, the shoulder of the survival curve tends to disappear, and the surviving fraction then approximates an exponential function of dose over the entire dose range. For a wide range of cells derived from normal tissue or from tumors investigated in vitro, the D0

α Cell kill

Damage

β Cell kill

2

S = e−αD − βD

α /β Dose per fraction

FIGURE 1-11.  Dose-response curves for mammalian cells are adequately fitted by the linear-quadratic relationship, at least over the range of doses of concern in radiation therapy. The form of the equation is S = e−αD  −βD2, in which S is the fraction of cells surviving a dose D, and α and β are constants. Cell killing by the linear and quadratic terms is equal when αD = βD2. This occurs when the dose D = α:β.

13

tends to be within the range of 1 to 2 Gy and the extrapolation number, n, varies from little more than 1 to about 20.

Effects on Normal Tissues Clonogenic Survival End Points Although normal tissue cells can be difficult to maintain in culture for long periods, cell survival curves for normal tissues have occasionally been obtained through the use of ingenious techniques. The first example historically was for the survival of irradiated skin cells in the mouse. In this technique, devised by Withers,11-13 a superficial x-ray machine is used to irradiate a circular annulus to the massive dose of 30 Gy to produce a moat of sterilized cells. In the center of this moat is an isolated island of intact skin, protected during the first exposure by means of a small metal sphere. This small area of intact skin is then given a test dose, D, and subsequently observed for the regrowth of a nodule of skin. If one nodule regrows in one half of the areas exposed, it indicates that a single cell survived. By varying the area of skin irradiated and dose of the radiation, it is possible to calculate the number of cells per square centimeter surviving various doses of radiation and to construct a cell survival curve. When this is done, the D0 turns out to be about 1.35 Gy. Because the quantity scored is not the surviving fraction but rather the number of surviving cells per square centimeter of skin, no extrapolation number can be obtained. Instead, a second dose-response curve is constructed for split doses in which the radiation is administered in 2 equal fractions 24 hours apart. The horizontal separation between the dose-response curve for single and split doses is then Dq. This value is about 3.5 Gy for mouse skin and is very similar for human skin. Another normal tissue for which cell survival curves can be obtained is the “surviving crypts in the mouse jejunum system.” The intestinal epithelium represents a classic selfrenewing tissue. The crypts at the base of the villi contain dividing cells, which pass up the villi as they differentiate and become functional cells. A dose of radiation of several Gy sterilizes a proportion of the dividing cells in the crypts but does not affect the mature, differentiated, functioning cells in the villi. After a dose of radiation, the villi shrink because no replacement cells from the crypts exist to take the place of those sloughed away from the tips of the villi. If the test animal is killed 3 or 4 days after the irradiation and sections are made of the jejunum, the number of regenerating crypts can be counted, giving a measure of surviving cells per circumference. This value can then be plotted as a function of dose. It turns out with this system that the D0 is about 1.3 Gy, but the most interesting fact is that the Dq is very large, indicating the presence of a large amount of accumulation and repair of sublethal damage. A third normal tissue system with which cell survival curves can be obtained was developed by McCullough and Till.14 In this technique, a suspension of nucleated bone marrow cells is prepared from the bone marrow of a donor mouse and injected into a recipient mouse. Some of the circulating cells lodge in the spleen and produce nodules that can be counted after they have grown for 9 or 10 days; in effect, the investigator counts nodules in the spleen instead of colonies in a Petri dish. By comparing the number of spleen colonies resulting from cells taken from irradiated and unirradiated donor animals, the fraction of cells ­surviving a given dose can be calculated. The cell survival

14

PART 1  Principles

curve for these stem cells closely approximates an exponential function of dose (i.e., it has little or no shoulder, and the D0 is about 0.95 Gy). These cells are radiosensitive because they die an apoptotic rather than a mitotic death. Dose-response curves for these three normal tissue systems are shown in Figure 1-12.

Functional End Points Although dose-response curves for clonogenic cell survival can be obtained in a limited number of circumstances, in many other situations, it is more practical to obtain a doseresponse relationship for the loss of function of an organ or tissue. Organ function is related more to the proportion of functional cells remaining in an irradiated organ at a particular time than to the proportion of clonogenic, or stem, cells. The functional cell, not the stem cell, data determine tolerance doses in multifraction radiation therapy. ­Functional dose-response curves can be constructed from multifraction experimental data by making several simple assumptions: (1) the linear-quadratic model is an adequate fit of the dose-response data, at least up to doses of a few Gy; (2) the response of the tissue or organ is the same to each dose in a multifraction regimen; (3) the repair of ­sublethal damage is complete between multiple dose fractions; and (4) no cell division occurs during the ­fractionated regimen. The basis of the method is to expose groups of animals to various numbers of fractions, with the dose per fraction titrated to produce a given functional impairment. The reciprocal of the total dose is then plotted against the dose per fraction, and the values of α and β corresponding to the linear-quadratic form of the dose-response relationships can then be inferred from the intercept on the ­abscissa and from the slope of this plot. An example is shown in ­Figure 1-13.15 The extent to which the experimental data fit this

straight-line plot, which is a transcription of the ­ linear­quadratic curve, is an indication of how well the linear­quadratic relationship fits the basic dose-response data. Acute functional effects for which multifraction techniques have been used to obtain the parameters of the dose-  response relationship include the following: Acute desquamation in mouse, rat, and pig skin Human skin “tolerance” Callus formation after bone breakage in mice Mouse tail necrosis Mouse LD50/30 (i.e., dose that is lethal to 50% of the animals within 30 days) Rat tail necrosis For these acute effects, the ratio of α:β is usually close to  10 Gy, with the ratio being greater for the LD50 end points. Late functional effects for which multifraction experiments have been performed to obtain the parameters of the dose-response curve include the following: Cervical and lumbar spinal cord function in rats Kidney function in the rabbit, pig, or mouse Lung LD50 Lung (i.e., number of breaths/min) Bladder (i.e., urination frequency) Skin of the pig (i.e., late contraction) In these instances, the α:β ratio usually falls in the range of 1.5 to 4 Gy.

Calculations of Tumor Cell Kill Although the survival curve for cells exposed to graded single doses of radiation may be a linear-quadratic formulation, the effective survival curve for a fractionated regimen delivered as daily 2-Gy fractions is an exponential function of dose. The D0, the reciprocal of the slope of the curve

Reciprocal total dose (Gy−1)

Surviving fraction

1

0.04

1

Jejunal crypt cells

0.1

2

0.03

Skin cells

17 64

8

∴α / β = 10.1 Gy

16 32 Intercept =

0

0.01

Slope =

5 0.02

0.01 Colony forming bone marrow cells

β logeS = 0.00129 Gy −2

4

5

α = 0.013 Gy−1 logeS

10

15

20

Dose per fraction (Gy) 0

2

4

6

8

10

Dose (Gy)

FIGURE 1-12.  X-ray survival curves are reconstructed for three mammalian cell types using in vivo systems. Colony-forming units in the bone marrow are cultured according to the technique of J. E. Till and E. A. McCullough; skin colonies, by a ­technique devised by H.  R. Withers; and regenerating crypts in the intestinal epithelium, by a technique used by H. R. Withers and M. M. Elkind. The slopes of the survival curves are different, ­particularly in the width of the shoulder.

FIGURE 1-13.  Reciprocal of the total dose required to produce a given level of injury (i.e., acute skin reaction in mice) is a function of the dose per fraction in multiple equal doses. The overall time of these experiments was sufficiently short that proliferation could be neglected; the numbers of fractions are shown by each point. From the values of the intercept and slope of the best-fit line, the values of α and β and the ratio α:β for the dose-response curve for organ function can be determined. (Modified from Douglas BG, Fowler JF: The effect of multiple small doses of x-rays on skin reactions in the mouse and a basic interpretation. Radiat Res 1976;66:401-426.)

1  Physical and Biologic Basis of Radiation Therapy

and defined as the dose required to reduce the fraction of cells surviving to 37%, has a value of about 3 Gy for cells of human origin. This average value can differ significantly for different tumor types. For calculation purposes, it is often useful to use the D10, the dose required to kill 90% of the population: $  . × $ In the equation, 2.3 is the natural logarithm of 10. Two sample calculations follow. Example 1: A tumor consists of 109 clonogenic cells. The effective dose-response curve, given in daily dose fractions of 2 Gy, has no shoulder and has a D0 of 3 Gy. What total dose is required to give a 90% chance of tumor cure? Answer: To give a 90% probability of tumor control in a tumor containing 109 cells requires a cellular depopulation of 10−10. The dose to result in 1 decade of cell killing (D10) is given by the following: $  . × $  . ×   . 'Y Total dose for 10 decades of cell killing, therefore, is 10 × 6.9 = 69 Gy. Example 2: Suppose that in the previous example the clonogenic cells underwent three cell doublings during treatment. Approximately what total dose would then be required to achieve the same probability of tumor ­control? Answer: Three cell doublings would increase the cell number by 2 × 2 × 2 = 8. Consequently, about an extra ­decade of cell killing would be required, corresponding to an additional dose of 6.9 Gy. Total dose is 69 + 6.9 = 75.9 Gy.

Differential Effects on Normal Tissues In experimental animals and humans, normal tissues vary considerably in their sensitivity to ionizing radiation. Some of the least sensitive tissues (e.g., vagina, uterus, bile ducts) can tolerate a dose nearly 100 times that of very sensitive tissues (e.g., bone marrow, lens, gonads). The details of the effects that occur and the doses at which various effects can be seen are explored in the individual chapters of this textbook. However, some comparisons and generalizations of the biologic effects of ionizing radiation can be made here. Acute Effects. Certain tissues manifest evidence of radiation effects in a matter of hours to days; they can be classified as acutely responding tissues. In approximate ­order of clinical significance, these tissues are as follows: Bone marrow Ovary Testis Lymph node Salivary gland Small bowel Stomach Colon Oral mucosa Larynx Esophagus Arterioles Skin Bladder Capillaries Vagina

15

Tissues that do not reveal acute effects of clinical significance but show evidence of damage of clinical significance in a matter of weeks to a few months after irradiation are often included with acutely responding tissues. Tissues such as these may be better categorized as subacutely responding tissues. In order of decreasing sensitivity, these tissues include the following: Lung Liver Kidney Heart Spinal cord Brain Late Effects. If given sufficient doses of radiation, all tissues can manifest late effects. However, several tissues are characterized as showing little or no evidence of acute effects but having well-recognized late effects. These ­tissues include the following: Lymph vessels Thyroid Pituitary Breast Bone Cartilage Pancreas (endocrine) Uterus Bile ducts

Effects on Tumors Comparison of Normal and Malignant Cells Malignant cells differ little from normal cells in their responses to ionizing radiation, tending to be most sensitive in the same phases of the cell cycle. Most cells seem to undergo repair of sublethal damage if sufficient time elapses between two doses of radiation. Depending on the interval after irradiation, repopulation may result from the proliferation of clonogenic cells. Redistribution of cells as a function of their differing sensitivities in different phases of the cell cycle, with consequent radiation-induced synchrony, may occur in benign and malignant cells. However, only malignant cells are thought to undergo reoxygenation ­after the administration of ionizing radiation. The radioprotective effect of hypoxia in tumors is an important reason for the failure to control tumors with irradiation. The concepts of reoxygenation and radioprotection are discussed further under “Tissue and Tumor Kinetics.” Malignant tumors differ greatly in radiosensitivity. ­Although cell survival data suggest that different cell types have some inherent differences in radiosensitivity, the differences in sensitivities of tumors are the result of many factors, including inherent cellular sensitivity, hypoxic ­fraction, vascularity, and changes in vascularity between repeated applications of radiation. Normal tissues and malignant tumors have dose­response relationships that tend to be sigmoidal (Fig.  1-14). A certain level of dose must be reached before any response is seen, after which the rate of response increases rapidly, followed by a diminishing rate of response at the highest dosage levels. As expected from the logarithmic ­reduction in the surviving fraction of cells after irradiation, higher doses are required to control larger tumors. Unfortunately, the term radiosensitivity has often been used to describe the rapidity with which a tumor shrinks

Probability of local control or complications

16

PART 1  Principles 1.0 0.90

Local control

Complications

0.05 Dose

FIGURE 1-14.  The probability of local tumor control and the probability of complications are a sigmoid function of dose. If the two curves are well separated, it is possible to achieve a high rate of tumor control with a small complication rate. If the curves are closer together, as may be the case for large, resistant tumors, this favorable situation would not apply.

after irradiation. The rate of regression after irradiation is more appropriately called radioresponsiveness; it has no relationship to radiosensitivity at the cellular level. Shrinkage is more related to the rate of cell removal (i.e., to the cell loss factor, as discussed later under “Tumor Cell ­Kinetics”). Shrinkage, or response of a tumor, is an important element for determination of tumor control. Complete response is the only meaningful indicator of tumor control. Because of the heterogeneity of cells within a tumor, there is no certainty of permanent tumor control after a complete response. A partial response is a radiotherapeutic failure except in the infrequent cases of tumors that may reach a stable state after initial shrinkage because they have a large stromal component (e.g., desmoid tumors) or because they have residual mature elements after the malignant component is eradicated (e.g., ganglioneuroblastomas, teratocarcinomas).

Assays for Transplantable Tumors in Laboratory Animals Several assay systems have been developed for transplantable tumors in small laboratory rodents. The advantage of these systems is that they are reproducible and highly quantitative. The disadvantage is that the tumors involved are highly anaplastic, undifferentiated, and rapidly growing, but they remain encapsulated, so that they show little resemblance to spontaneous tumors in humans. The following paragraphs describe five basic assays that have been developed and used widely over the years. These assays vary in complexity and expense, and they have different advantages and disadvantages. Tumor-Control Dose Assay. The most direct experimental assay, which is parallel to the observation of patients treated with x-rays, is to determine the dose of radiation required to produce tumor control (i.e., tumorcontrol dose [TCD]) in irradiated animals.16 In this assay, tumors are transplanted into a large number of animals and allowed to grow to a certain size, and then groups of animals are treated with graded doses of radiation. The animals are then observed for some period, and the dose is calculated that produces tumor control in one half of the animals treated. This quantity is the TCD50. The disadvantage of this system is that it is costly, inasmuch as perhaps  100 animals are required to produce this single-figure ­estimate of the TCD50. The balancing advantage is that the tumor is treated in situ, with all of the complexities

of the tumor-host response, and the tumor remains undisturbed after treatment. Growth Delay Assay. In the growth delay assay, a large number of identical animals are used in which tumors of a given size are produced by transplantation. After irradiation, the tumors are measured daily so that the patterns of regression and regrowth can be established. The time in which a tumor grows back to its original size after irradiation is a measure of tumor response called the growth delay. For tumors in which shrinkage does not occur, an alternative end point is to express the time for the tumor to grow from size A to size B (perhaps from 2 to 8 mm) in irradiated animals compared with the comparable time in unirradiated controls. Dilution Assay. The dilution assay technique devised by Hewitt and Wilson17 was the method by which the first in vivo dose-response curve was obtained. In this technique, initially devised for lymphocytic leukemia and subsequently adapted for use with solid tumors, tumor-bearing animals are irradiated and then cells are withdrawn and diluted, and various numbers are inoculated into recipient animals, which are then observed to see if they develop a tumor. The TD50 (i.e., number of cells required to produce a tumor in one half of the animals inoculated) is determined for control animals and irradiated animals. The ratio of TD50 values for control and irradiated animals is the surviving fraction of cells for that particular radiation dose. The strength of this system is that the tumor is irradiated in situ. An additional advantage of this assay over the two methods referred to earlier is that the reproductive ­integrity of individual cells is determined rather than the response of the tumor as a whole. The disadvantage of the dilution assay is that after the solid tumors are irradiated, the tumor must be broken up and disaggregated. Although this practice does not seem to produce any significant artifacts in the case of radiation, it may produce spurious results in  evaluations of chemotherapy agents. This assay system is also time-consuming and expensive, inasmuch as many animals are required to produce a dose-response curve. For example, a control group of at least six animals and an irradiated group of six animals produces the same amount of information as that obtained from one control and one irradiated Petri dish in the in vitro assay. Lung Colony Assay. The lung colony assay system was developed and has been used largely in Canada.18 In this system, solid tumors are irradiated in situ and then removed, disaggregated into a single-cell suspension, and inoculated into recipient animals that are killed some weeks later. The number of metastases in the animals’ lungs is counted. This system, too, exploits the realism of in vivo irradiation and assessment of the reproductive integrity of survivors, but it suffers the disadvantages of requiring large numbers of animals for a given amount of data and requiring that the tumor be disaggregated after treatment, before its assay. In Vivo/In Vitro Assay. In the in vivo/in vitro assay, the tumor is irradiated in situ, removed, and disaggregated into a single-cell suspension. Various numbers of cells are then plated onto Petri dishes, and their reproductive integrity is assessed by their ability to produce colonies in vitro. This technique combines the advantages of the in vivo and in vitro assays; the tumors are irradiated in situ with all of the complexities of the in vivo milieu, and they are then assessed with the speed, accuracy, and economy of the in

17

1  Physical and Biologic Basis of Radiation Therapy

vitro cell culture technique. Only a limited number of tumors have been developed that can be transferred back and forth between the animal and the Petri dish; unfortunately, these are all highly undifferentiated tumors. This technique also has the same limitation as the dilution assay and the lung colony assay systems in that it involves the disaggregation of the tumor after treatment. These assay systems, developed initially for the assessment of response to x-rays, neutrons, and other types of radiation, have allowed a vast body of data to be accumulated for chemotherapy agents and hyperthermia. Use of these systems resulted in the development of the important concept that the presence of a proportion of hypoxic cells can limit the curability of a tumor with a single dose of x-rays, leading to the demonstration and explanation of the phenomenon of reoxygenation. Models that involve the use of transplantable tumors in small rodents have ­fallen into disfavor to some extent because the dose-response curves obtained do not give much more information than an in vitro survival curve that requires much less time and a fraction of the cost. Criticism levied against these ­ models is that the tumors are highly undifferentiated and not at all like spontaneous human tumors. Their response to hyperthermia and chemotherapy agents may give spurious results for various reasons that are described later.

Tissue and Tumor Kinetics The Cell Cycle Mammalian cells propagate and increase in number by mitosis. The division cycle for mammalian cells is illustrated in Figure 1-15. The principal phases of the cell cycle (M, G1, S, G2) were described by Howard and Pelc in 1953.19 During a complete cell cycle, the cell must accurately replicate the DNA during S phase and distribute an identical set of chromosomes equally to two progeny cells during M phase. Mitosis

Regulation of the cell cycle in eukaryotic cells ­ occurs by the periodic activation of different members of the ­cyclin-  dependent kinase (CDK) family. In its active form, each CDK is complexed with a particular cyclin. Different CDKcyclin complexes are required to phosphorylate several protein substrates, and that phosphorylation drives cell cycle events such as the initiation of DNA replication or the onset of mitosis. CDK-cyclin complexes are also vital in preventing the initiation of a cell cycle event at the wrong time. Extensive regulation of CDK-cyclin activity by several transcriptional and posttranscriptional mechanisms ensures the perfect timing and coordination of cell cycle events. The CDK catalytic subunit by itself is inactive, requiring association with a cyclin subunit and phosphorylation of a key threonine residue (usually T-160) to become fully active. The CDK-cyclin complex is reversibly inactivated by phosphorylation on a tyrosine residue (usually Y-15) located in the ATP-binding domain or by association with cyclin-­dependent kinase inhibitor (CKI, also known as celldirected inhibitor [CDI]) proteins. After the completion of the cell cycle transition, the complex is irreversibly inactivated by ubiquitin-mediated degradation of the cyclin subunit. The entry of cells into S phase is controlled by CDKs that are themselves sequentially regulated by cyclins D, E, and A. D-type cyclins act as growth-factor sensors, with their expression depending more on the extracellular cues than on the cell’s position in the cycle. Mitogenic stimulation governs their synthesis and the formation of complexes with CDK4 and CDK6, and catalytic activity of the assembled complexes persists through the cycle as long as mitogenic stimulation continues. Cyclin E expression in proliferating cells is normally periodic and maximal at the G1-S transition; throughout this interval, cyclin E enters into active complexes with its catalytic partner, CDK2. This view of the cell cycle and its regulation is illustrated in Figure 1-16.

G0 M

G2

EGF FGF PDGF

M

G1

Transcription factors

Cyclin B/A and CDK1

G2

Cyclin D1 and CDK4 Cyclin A and CDK2 S

Cyclin E and CDK2

G1

p53 Rb p21 inhibits CDKs

Late S

S

FIGURE 1-15.  With a conventional microscope, the only phase of the mammalian cell cycle that can be identified is mitosis, in preparation for which the chromosomes condense and become visible. With the use of autoradiography, the DNA synthetic phase (S) can be identified to occupy a discrete length of time in the cycle.

FIGURE 1-16.  Updated view of the phases of the cell cycle shows how the phases are regulated by the periodic activation of different members of the cyclin-dependent kinase (CDK) family. Various CDK-cyclin complexes are required to phosphorylate several protein substrates that drive key events, including the initiation of DNA replication and the onset of mitosis. EGF, epidermal growth factor; FGF, fibroblast growth factor; PDGF, platelet-derived growth factor.

PART 1  Principles

The only phase of the cycle that can be observed with a light microscope is the process of division itself, in preparation for which the chromosomes condense and become visible. To visualize the DNA-synthesis (S) phase, the technique of autoradiography must be used. In this technique, cells are fed one of the precursors needed to build DNA that is labeled with a radioactive isotope such as tritium. Only cells in the DNA-synthesis phase take up the tritium. If those cells are later fixed, stained, and covered with a layer of nuclear emulsion, the beta particles from the ­tritium pass through the nuclear emulsion and produce an image that is visible when the emulsion is fixed and stained. To avoid using radioactive material, cells are more often labeled by the incorporation of bromodeoxyuridine (BrdU), the presence of which can be detected by a ­fluorescent-­labeled antibody (Fig. 1-17). However, the principle remains the same.

A

B FIGURE 1-17.  A, Autoradiograph of cultured Chinese hamster cells flash-labeled with tritiated thymidine. Black grains indicate cells synthesizing DNA when they were labeled. The labeled mitotic cell (right) was in S phase when the culture was flash-­labeled but in M phase at staining and autoradiography. B, Photomicrograph of cells grown in the presence of bromodeoxyuridine (BrdU) and fixed and stained 20 hours later. First-generation mitotic cells show uniform staining of both chromatids of each chromosome. Inset, A second-generation mitotic cell (which incorporated BrdU during two S phases); the dark chromatids indicate BrdU incorporation by both strands of the DNA double helix, and the light chromatids by only one. (Courtesy of Charles R. Geard, MD, Columbia University, New York, NY.) Modified from Hall EJ, Giaccia AJ. Radiosensitivity and cell age in the mitotic cycle. In: Radiobiology for the Radiologist. 6th ed. Philadelphia: Lippincott Williams & Wilkins, 2006, pp 47-59.

The simplest experiment that can be done with a cell culture, a tissue, or a tumor is to flash-label it by exposing the cells or tissues for 15 to 20 minutes with the labeled DNA precursor and then to look at the samples after they have been autoradiographed. The labeling index (LI), the proportion of cells that are labeled, reflects the ratio of the length of the DNA-synthesis phase (Ts) to that of the total cell cycle (Tc): ,)  4S/4C

From the same sample, it is also possible to observe the mitotic index (MI), defined as the proportion of cells in mitosis. This index represents the ratio of the length of the mitotic time (Tm) to that of the Tc: -)  4M/4C

It is therefore possible in a simple experiment to observe the ratio of the lengths of mitosis and the DNA-­synthesis phase to that of the total cell cycle. However, these experiments involve the measurement of ratios and cannot be used to estimate the length of any phase of the cycle. To determine the length of the various phases of the cell cycle, a much more complex procedure is necessary—the percent-labeled mitoses technique. The cell culture, tissue, or tumor is flash-labeled with tritiated thymidine, after which samples are taken, fixed, stained, and autoradiographed hourly for some period that is longer than the anticipated cell cycle. A typical percent-labeled mitoses curve is shown in Figure 1-18.20 This technique follows the progress of the labeled cohort of cells through DNA synthesis, to the next mitosis, and through the cell cycle to another mitosis. The second wave of labeled mitoses is always smaller than the first because the cells in any population in vivo do not have a single cell cycle time but rather are characterized by a range of cell cycle times. A computer technique can be used to unfold the spectrum of cell cycle times from the decrease in height of the second peak and to obtain an estimate of the range of cell cycles involved. The width of the first wave of labeled mitoses at the 50% height gives an estimate of the Ts, and the time between the first and second peaks is an estimate of the average Tc. In many instances, particularly with normal tissues, the second peak is too indistinct to be recognized; then the cell cycle cannot be estimated from the distance between consecutive peaks. In such cases, the Ts must be obtained Percent labeled mitoses

18

100 80 60 40 20 0 0

2

4

6

8

10 12 14 16 18 20 22 24

Hours after injection of 3HTdR

FIGURE 1-18.  In the percent-labeled mitoses curve for a chemically induced carcinoma in the cheek pouch of a hamster, the squares and circles refer to two separate experiments. (From Brown JM. The effect of acute x-irradiation on the cell proliferation kinetics of induced carcinomas and their normal counterpart. Radiat Res 1970;43:627-653.)

1  Physical and Biologic Basis of Radiation Therapy

from the width of the first peak and the cell cycle determined by the observation of the LI: 4C  4S/,)

Population Kinetics of Tumors When animal tumors are studied with the percent-­labeled mitoses technique, it becomes evident that the cell cycle time is much shorter than the gross volume-doubling time. There are two reasons for this. First, not all cells in a ­growing tumor are actively cycling at any given time. By observing the uptake of tritium in mouse tumors, it is possible to can estimate that the growth fraction, defined as a proportion of cells that are actively cycling, ranges from 30% to 50%. From the cell cycle of individual cells and the growth fraction, it is possible to calculate the potential doubling time (Tpot). The potential doubling time is still much shorter than the actual gross volume-doubling time, and this difference is accounted for by the cell loss fraction (Φ), defined as the proportion of cells produced by mitosis that are lost from the tumor. These cells are lost largely into the necrotic “sinks” within the tumor, areas where cells die because they are remote from a source of oxygen and nutrients. The overall picture of tumor growth that emerges is that of a population of cells, some of which are dividing and others are quiescent, with the progeny of division largely being lost into necrotic areas of the tumor. For this reason, the gross volume-doubling time of the tumor, as a whole, is much longer than the cell cycle time. In humans, a limited number of population kinetic studies have been performed because it is not possible in most countries to give large doses of tritiated thymidine to patients. Generally, human tumors are observed to have gross volume-doubling times on the order of months, judging from repeated measurements of skeletal or pulmonary metastases. In contrast, 80% of human tumors have a cell cycle that lasts between 2 and 5 days. These numbers are consistent with the estimate of Gordon Steel21-23 that, on average, human tumors have a cell loss factor of 70% (i.e., almost three fourths of the cells produced by division are lost into necrotic areas within the tumor).

Clinical Applications Although results from population kinetic studies performed in the past are illuminating and instructive, they cannot be applied to an individual patient because of the long periods required to perform cell cycle analyses. For example, an autoradiograph takes about 6 weeks to be produced, which includes the time necessary for sufficient beta particles to be emitted by tritium and incorporated within the cells and to produce a visible image in the nuclear emulsion above the cells. The breakthrough in this area has been the application of flow cytometry.

Measurement of Potential Tumor Doubling Time The growth rate of a tumor is slower than the growth rate of individual dividing cells for two reasons. First, most ­tumors contain significant numbers of cells that are not cycling at any given time; the fraction of cells in cell cycle is the growth fraction. Second, many cells produced by division are lost, mainly in necrotic areas. The fraction of cells produced by mitosis that are lost from the tumor is known as the cell loss fraction (Φ).

19

Tpot accounts for the proportion of dividing cells but not for cell loss. It is a measure of the rate of increase of cells capable of continued proliferation, and it therefore determines the outcome of a treatment delivered in fractions over an extended period. Tumors with a short Tpot may repopulate if fractionation is extended over too long a period. Tpot can be calculated from the following equation: 4POT  Ȉ4S/,) In the equation, λ is a correction factor for the nonlinear distribution in time of the cells as they pass through the cycle (λ is between 1 and 0.67). The Tpot of human tumors can be estimated by flow cytometry analysis of cells from a biopsy specimen taken 6 hours after an injection of BrdU.24-26

Systemic Effects Systemic effects of ionizing radiation are determined by the proportion of the body irradiated, the specific organs included, and the dose received. At the extreme, the lethal effects of total-body irradiation are well recognized. In addition to results from experiments with animals, especially experiments in which investigators are seeking the dose that results in the death of one half of the animals (lethal dose in 50% [LD50]), limited data are available from accidental exposures to humans, such as that at Chernobyl. Three types of death from total-body exposures have been delineated. Doses on the order of 100 Gy or higher cause death in  a matter of hours from effects on the central nervous system and cardiovascular collapse. Intermediate doses of 10 to 20 Gy result in death in several days caused by elimination of the intestinal epithelium with intractable diarrhea and infection. Lower doses of 1 to 5 Gy result in depopulation of hematopoietic stem cells; death may occur several weeks after the exposure from infection or bleeding. The LD50/60 for healthy young adults without medical intervention is between 3 and 4 Gy. Individuals can survive doses perhaps twice as large with supportive treatment, including antibiotics and transfusions of blood products to help them over the nadir in peripheral blood counts. Single-dose total-body irradiation has been used alone or with systemic cytotoxic drugs as ablative therapy to eradicate malignant cells from the bone marrow (especially in cases of leukemia) or from other parts of the body, followed by autologous, syngeneic, or allogeneic reconstitution of the bone marrow. In clinical experiments in which single total-body doses of 5 to 10 Gy and fractionated doses of 12 to 16 Gy have preceded marrow transplants, it has become apparent that the limiting normal tissue is the lung; deaths from interstitial pneumonia have occurred weeks to months after irradiation. The effects of irradiation on immune responses are complex. When total-body irradiation is given in doses that are potentially lethal, the immune mechanisms throughout the body, both humoral and cell-mediated, are depressed. This is a direct result of the suppression of the progenitors of plasma cells and lymphocytes that produce immunoglobulins or participate in cell-mediated responses. The effects of local irradiation, particularly with regard to the volume irradiated and the amount of bone marrow within that volume, are only partially defined. Total lymphoid ­irradiation can have effects that are qualitatively similar to total-body

20

PART 1  Principles

irradiation but are quantitatively less ­similar. In the clinical setting, the malignant process itself may impair humoral and cell-mediated immunity in humans, and thereby, the contributions of the primary disease and the therapy are difficult to separate. Radiation intended as therapy for many common malignant conditions does not impair cellular immunity, as measured by the delayed hypersensitivity reaction. When very large volumes are irradiated, as takes place in the treatment of Hodgkin disease, cutaneous anergy may temporarily follow. Patients who have anergy, presumably related to the malignant process, often recover their delayed hypersensitivity response after radiotherapeutic elimination of the tumor. Depressions of the peripheral blood count from local irradiation are seen only when the volume irradiated is large and encompasses a significant portion of the active bone marrow. Irradiation of small areas does not significantly influence white blood cell count, platelet count, or hematocrit. The assumption that local irradiation is sufficient to cause profound pancytopenia can lead to serious delay in seeking the actual causes. Nausea and vomiting have been considered by patients and physicians to be a necessary accompaniment of any sort of radiation therapy. However, nausea and ­ vomiting may also accompany irradiation as a function of the suggestibility of the patient. Undoubtedly, the range of ­individual sensitivities to these occasional side effects of irradiation is considerable, but nausea is most likely to occur when the stomach or significant portions of the intestine are ­included within the treatment volume. The irradiation of a very large volume of tissue, even when the gastrointestinal tract is excluded (e.g., treatment with the mantle field), also can result in nausea. Nausea that is truly related to irradiation for one of the previous reasons occurs at the very start of a course of treatment. Serious delays in finding the causative factors can occur if it is assumed that the nausea and vomiting that appear well into or near the end of a course of treatment are related to the irradiation when they are not.

Fractionation The earliest attempts at therapeutic uses of the newly discovered x-rays, especially in Germany, involved single, large, caustic doses with profound effects on the tumor but severe injuries to normal tissues. More than 20 years passed before the advantages of repeated, brief applications of x-rays were appreciated. The most important conceptual development in the history of clinical radiation oncology came from studies of spermatogenesis in the ram, reported by Regaud in the early 1900s.27 Regaud demonstrated that a ram could not be sterilized by exposing its testes to a single dose of radiation without extensive damage to the skin of the scrotum, but if the radiation was given in a series of daily fractions, sterilization was possible without producing unacceptable skin damage. Because the testis could be considered a model for a rapidly growing tumor in which the skin of the scrotum represented a dose-limiting normal tissue, the strategy of multifraction radiation therapy was born. The efficacy of fractionation can be understood in terms of radiobiologic principles established with more relevant test systems. These principles have been described as the four Rs of radiobiology: Repair Reassortment of cells within the cell cycle

Repopulation Reoxygenation

Repair Sublethal damage repair and potentially lethal damage ­repair are defined in operational terms. The molecular ­basis of either is not understood, nor is it known whether the two forms of repair involve the same or different ­molecular processes. Sublethal damage repair is defined as the increase in survival observed when a dose of radiation is split into two (or more) fractions with a time interval between. Figure  1-19A shows a dose-response curve for cells given single doses up to the dose value D (solid line). If a dose D/2 is given and a time interval of several hours is allowed to elapse before the remaining portion of the dose is delivered, survival follows the dotted line as shown in Figure 1-19B (i.e., the shoulder of the curve must be re-expressed). ­ Figure  1-19B shows the result of an experiment in which two doses of D/2 were delivered, separated by various time intervals. Repair takes place with a half-life of about an hour and is essentially complete by about 3 hours. Sub­lethal damage repair seems to be a ubiquitous phenomenon; it has been demonstrated in cells in ­ culture, transplanted ­ tumor systems, and normal tissue assay ­systems. In the example of rapidly dividing cells of rodent origin grown in vitro, the half-time of repair is about an hour. However, the rapidity of repair varies considerably. For cells of human origin cultured in vitro, it may vary from a few minutes to several hours.28 Precise estimates for normal tissues in vivo are more difficult to make, but fractionation experiments suggest that the repair of sublethal damage may be very much slower in late-responding tissues.25 The extent of repair varies widely among different cells and tissues, tending to correlate with the size of the shoulder on the cell survival curve. One manifestation of sublethal damage repair is the effect of dose rate. Radiation delivered continuously at a low dose rate over some period approximates to an infinite number of infinitely small dose fractions. Sublethal damage repair occurs during the protracted exposure; as a consequence, the survival curve becomes shallower (i.e., D0 increases), and the shoulder has a tendency to disappear as the dose rate is progressively reduced (Fig. 1-20). Repair of potentially lethal damage is defined as the increase in cell survival that can be produced by manipulation of the postirradiation conditions. If, after irradiation, cells are kept in an environment that prevents cell division, more cells survive than if they are called on to divide soon after the exposure. This effect can be achieved with cells in culture by replacing the growth medium with saline or depleted medium, by lowering the temperature, or by maintaining the cells in plateau phase in culture conditions where they are confluent and have no room to proliferate. Potentially lethal damage repair occurs in cells of normal and malignant origin and has been demonstrated to occur in transplanted tumors in animals. If a tumor is left in situ for 6 to 24 hours after irradiation before being removed, subcultured, and transplanted, more cells survive than if the tumor is transplanted immediately. The naïve interpretation of these results is that some radiation damage that is potentially lethal can be repaired in a few hours, provided the cells are not called on to complete the complex task of mitosis during that time interval.

1  Physical and Biologic Basis of Radiation Therapy Dose

D/2

21

Time between split doses D

2

4

6

8

10

12

Cell cycle

1

0.1 D/2- + -D/2

0.01

A

Reassortment in fastcycling cells

Proliferation

Single dose

Repair SLD

Surviving fraction

Split dose

B

FIGURE 1-19.  A, Increase in cell survival is observed when a dose of radiation is delivered in 2 fractions separated by a time interval adequate for repair of sublethal damage (SLD). When the dose is split into 2 fractions, the shoulder must be expressed each time. B, The fraction of cells surviving a split dose increases as the time interval between the 2 dose fractions increases. As the time interval increases from 0 to 2 hours, the increase in survival results from the repair of sublethal damage. In cells with a long cell cycle or cells that are out of cycle, cell survival cannot be further increased by separating the dose by more than 2 or 3 hours.

Presumably, repair of potentially lethal damage also occurs in human tumors in vivo. It has been speculated that the ability to repair potentially lethal damage efficiently may be related to the resistance of some tumors, such as melanoma. Whether potentially lethal damage repair occurs in normal tissues in vivo is unknown because suitable assays are unavailable. However, it has been demonstrated in cells of normal tissue origin cultured in vitro.

Reassortment The radiosensitivity of mammalian cells varies according to their position in the cell cycle. In fast-growing cells in culture, such as Chinese hamster cells, which have a cell cycle of only 9 or 10 hours, cells in mitosis (M) or late in G2 usually are the most sensitive. They are characterized by a survival curve that is steep, with little or no initial shoulder. Cells late in the DNA-synthetic phase (late S) are most resistant and tend to have a survival curve with a broad shoulder. A similar pattern of radiosensitivity has been shown for the rapidly dividing crypt cells in the mouse intestinal epithelium, the only situation in vivo in which good synchronization can be achieved by the use of a drug such as hydroxyurea. Cells cultured in vitro that have a longer cell cycle (e.g., 24 to 30 hours) and therefore a long G1 phase seem to have a second period of resistance in G1. When an asynchronous population of cells is exposed to a single large dose of radiation, most of the survivors are in the resistant phases of the cell cycle (i.e., the surviving population is partially synchronized). A second dose of radiation is less effective than the first because the population is resistant. If a time interval is allowed, however, cells tend to reassort themselves and become asynchronous

again as cells move from resistant into sensitive phases of the cycle.

Repopulation Repopulation refers to regrowth of cells after irradiation. It was described earlier under “Tissue and Tumor Kinetics.”

Reoxygenation Reoxygenation is the process whereby cells in a tumor that are hypoxic after a dose of radiation become oxygenated again as the tumor shrinks or as the demand for oxygen is reduced. This process has been demonstrated in many transplanted animal tumors. The extent and rapidity of reoxygenation vary considerably in animal tumors, but little is known about the process in humans. Reoxygenation is detailed under “Modifiers of Radiation Effects.”

Fractionation and the Four Rs Dividing a dose into fractions spares normal tissues because of the repair of sublethal damage between dose fractions and because of the repopulation of cells, if the overall time is sufficiently long. At the same time, dividing a dose into fractions increases damage to the tumor as a result of ­reoxygenation. The effect of repopulation of normal tissue is rather complex. The extra dose that is required (because of ­fractionation) to produce a given level of damage does not increase until several weeks after the start of a multifraction regimen because proliferation does not occur until damage is expressed in the normal tissue. After damage to normal tissue is apparent, compensatory proliferation is rapid. However, all normal tissues are not the same. 

22

PART 1  Principles Dose (Gy) 0

2

4

6

8

10

12

14

100

Reaction Sites Low dose rate

Cell surviving fraction

TABLE 1-3.   Ratio of Linear to Quadratic Terms from Multifraction Experiments

10-1

Acute exposure

Dose-rate effect caused by repair of sublethal damage

10-2

α:β (Gy)

Early Reactions Skin

9-12

Jejunum

6-10

Colon

10-11

Testis

12-13

Callus

9-10

Late Reactions Spinal cord

1.7-4.9

Kidney

1.0-2.4

Lung

2.0-6.3

Bladder

3.1-7.0

From Fowler J. The second Klaas Breur memorial lecture. La Ronde—radiation sciences and medical radiology. Radiother Oncol 1983;1:1-22.

10-3

FIGURE 1-20.  The dose-rate effect is depicted. For sparsely ionizing types of radiation such, as x-rays and γ-rays, the biologic effectiveness of a given radiation dose decreases as the dose rate is lowered. This occurs when the exposure time becomes comparable to or longer than the half-time of repair of sublethal damage, so that sublethal damage is repaired during the protracted exposure. As the dose rate is progressively reduced, the slope of the survival curve becomes shallower (the final slope [D0] increases), and the shoulder of the survival curve disappears. The dose-response curve then approximates an exponential function of dose. Dose

In particular, a clear distinction exists between tissues that are rapidly responding, such as the skin, the intestinal epithelium, or the oral mucosa, and late-responding tissues, such as the spinal cord. Prolonging overall treatment time within the normal radiation therapy range has little ­sparing effect on late reactions, but it has a large sparing effect on early reactions. This axiom is of far-reaching importance in radiation therapy. Early reactions, including those in the skin and the mucosa, can be avoided by the simple expedient of prolonging the overall time. Although such a strategy avoids the problems of the early reactions, it has no effect whatsoever on the late reactions because compensatory proliferation in late-responding tissue does not occur during the time course of conventional ­radiation therapy.

Surviving fraction

Early- and Late-Responding Tissues

Late-responding tissue

Tumor and earlyresponding tissue

FIGURE 1-21.  The dose-response relationship for late­responding tissue is “curvier” than for early-responding tissue. The dose at which cell killing is equal by the linear and quadratic components is α:β. This is about 2 Gy for late-responding and 8 to 10 Gy for early-responding tissue. (Information courtesy of H. R. Withers, MD, UCLA Medical Center, Los Angeles, CA.)

Differences in the repair and efficiency of reassortment of cells within the cell cycle lead to quite different shapes of the dose-response curves for early-responding and late-responding tissues. Dose-response relationships for late-responding tissues tend to be more curved than for early-responding tissues.30 In terms of the linear-quadratic relationship between effect and dose, this translates into a larger ratio of α:β for early effects than for late effects. The difference in the shapes of the dose-response relationships is illustrated in Figure 1-21. Values of α:β for a range of tissues in experimental animals is shown in Table 1-3.31 These data came from multifraction experiments with rats and mice. In general, the value of α:β is on the order of 2 Gy for late-responding tissues but is closer to 8 or 10 Gy for early-responding tissues. Consideration of the shape of the dose-response relationship for early-responding and late-responding tissues leads to the important truism that fraction size is the dominant factor in determining late effects, whereas overall treatment time has little influence. In contrast, fraction size and overall treatment time determine the response of acutely responding tissues.

1  Physical and Biologic Basis of Radiation Therapy

The Strandqvist Plot and Nominal Standard Dose Early attempts to understand and account for fractionation in clinical radiation therapy gave rise to the well-known Strandqvist plot, in which the effective single dose was plotted as a function of the overall treatment time. Because all treatments were given as 3 or 5 fractions per week,  overall time in these plots carried with it the implication of a certain number of fractions. These plots commonly showed that the slope of the isoeffect curve for skin was approximately 0.33. The most important contribution in the area of fractionation, made by Ellis and colleagues32 with the ­introduction of the nominal standard dose system, was the recognition of the importance of separating overall treatment time from the number of fractions. This system was used to ­predict which radiation regimens would produce equivalent results with different numbers of fractions, but it has fallen out of favor.

Tumor Control and Dose-Time Considerations For many years, the basic determinant of total dose in the clinical setting was the tolerance of the surrounding normal tissues, especially the skin. With the advent of megavoltage treatment units and the attendant skin-sparing effect, greater attention has focused on the depth and distribution of dose correlated with permanent eradication of the disease. It became apparent that tumor-control probability for epithelial tumors is profoundly affected by small changes in the time-dose relationships, and the size of the tumor greatly influences the dose necessary to achieve a high probability of control. Although the time-dose relationships are of great importance for controlling epithelial tumors, the time factor seems to be of much less significance in the treatment of highly radiosensitive tumors (e.g., Hodgkin disease, malignant lymphomas, seminoma, Wilms’ tumor, neuroblastoma, retinoblastoma). The total dose is a more important determinant in the control of Hodgkin disease and nodular lymphomas, whereas the number of fractions and the time over which the dose is delivered are less important. ­Because deleterious effects on normal tissues are related to dose and to fractionation and time, radiation that is more fractionated and given over a longer period tends to have a greater margin of safety. Radiation in laboratory animals or humans is used, with rare exceptions, as a locoregional treatment. Although laboratory studies have consistently evaluated tumor control (TCD50) as the end point, clinical results are often ­expressed in terms of survival or cancer-free survival. The only direct measure of the effectiveness of irradiation is control of the tumor within the irradiated volume (i.e., ­ local control). Margin failure, the failure to appreciate the original extensions of the tumor, should be analyzed ­separately. The only measure of failure of local control (i.e., true recurrence) is regrowth of the tumor within the irradiated volume. Identifying failure of local control requires ­careful evaluation of the patient over months and years, and ­ attempts to predict failure of local control by other means have been unsuccessful. The rate at which the tumor ­ regresses and, consequently, the extent of residual tumor at the completion of irradiation have been evaluated in ­ animals and in humans and have not proved to

23

correlate with ultimate control of disease. Suggestions that ­ postirradiation biopsies might help were based on the expectation that tumor cell “viability” could be assessed histologically. This largely was proved incorrect in the ­ immediate postirradiation period. However, if a high growth fraction is present and most cells are actively cycling, eventual control may be predicted by repeat biopsies at the site of the original tumor. Arriagada and colleagues33 documented a high local failure rate of radiation to treat non–small cell carcinoma of the lung by using repeat bronchoscopy and biopsy 3 months after radiation therapy with and without induction chemotherapy. In that study, almost every patient with positive biopsy results developed recurrent disease and eventually died, and all long-term survivors had negative bronchial biopsy results. In contrast, biopsies of the prostate 2 years or more after external beam radiation therapy are much less predictive of clinical local failure,34 although they have been found to correlate with levels of prostatespecific antigen.35

Large-Dose Fractionation The strong influence of fraction size on late effects led most investigators to abandon the use of large fractions in the 1980s.36 Since that time, a resurgence of interest evolved from the ability to achieve highly conformal dose ­distributions by using advanced imaging for the planning and delivery of radiation treatment. Stereotactic radiation therapy for cerebral38 and pulmonary tumors37 uses between 1 and 5 fractions ranging from 12 Gy to 20 Gy each. The use of radiation fractions larger than traditional 1.8 Gy or 2.0 Gy is being studied for the treatment of disease at a variety of anatomic sites.

Multiple Fractions Per Day Most malignant tumors arising from the epithelium require total doses near or even above those tolerated by the surrounding normal tissues. Consideration of the different α:β ratios for early-responding tissues, including tumors, and late-responding tissues suggests an advantage to altered fractionation. Prolongation of treatment has the advantage of avoiding acute reactions and allowing adequate reoxygenation in tumors. However, excessive prolongation can decrease acute reactions without necessarily sparing late injury, and it may allow surviving tumor cells to proliferate during treatment. Two separate strategies use multiple fractions per day:  hyperfractionation and accelerated fractionation. The objectives of these two strategies are quite different. Both fractionation strategies involve multiple fractions per day, and in this context, it is important to ensure that the two fractions are separated by a sufficient time interval for the effects of the two doses to be independent, allowing the repair of sublethal damage from the first dose to be complete before the second dose is delivered. With cells cultured in vitro, the half-time of repair is usually about an hour, but repair may be much slower in late-responding tissues. Clinical trials indicate that for a given total dose delivered in a given number of fractions, the incidence of late effects was worse for average interfraction intervals of less than 4.5 hours.39 Current wisdom dictates an interfraction interval of 6 hours or more when multiple fractions per day are used.

24

PART 1  Principles

Hyperfractionation. The basic aim of hyperfractionation is to further separate early and late effects. “Pure” ­hyperfractionation may be defined as keeping the same ­total dose as in a conventional regimen in the same overall time but delivering it in twice as many fractions by the ­expedient of treating twice each day. However, this strategy would not be satisfactory because the total dose would have to be increased if the dose per fraction were ­decreased. In practice, “impure” hyperfractionation involves an increase in the total dose and sometimes a longer overall time, as well as many more fractions delivered twice daily. The intent of hyperfractionation is to reduce late effects while achieving the same or better tumor control and the same or slightly increased early effects. A large, controlled clinical trial of hyperfractionation for the treatment of head and neck cancer was conducted by the European Organisation for Research and Treatment of Cancer (EORTC).40 A hyperfractionated schedule of 80.5 Gy delivered in 70 fractions (1.15 Gy delivered twice daily) over a period of 7 weeks was compared with a conventional regimen of 70 Gy delivered in 35 fractions of 2 Gy over 7 weeks. Local tumor control at 5 years after the hyperfractionated protocol increased from 40% to 59%, and this increase was reflected in improved survival. No increase in late effects or complications was reported. These results led to the conclusion that hyperfractionation confers an unequivocal advantage in the treatment of oropharyngeal cancer. Two fractions per day may not be the limit of hyperfractionation. A further sparing of late effects by splitting the dose into more and more fractions of smaller and smaller size would occur as long as the dose fraction is still on the curved portion of the dose-response curve. Accelerated Treatment. The alternative strategy to hyperfractionation is accelerated treatment. “Pure” accelerated treatment may be defined as the same total dose delivered in one half of the overall time by the expedient of giving two or more fractions each day. In practice, this can never be achieved because the acute effects become limiting. A rest period must be imposed in the middle of the treatment, or the dose must be reduced slightly, with acute effects being the limiting factor. The intent of this accelerated treatment strategy is to reduce repopulation in rapidly proliferating tumors. Late effects should show little or no change because the number of fractions and the dose per fraction are unaltered. In addition to the hyperfractionation trial described previously, the EORTC conducted a large prospective randomized clinical trial of accelerated treatment for head and neck cancer (except that of the oropharynx).41 The accelerated treatment consisted of 72 Gy given in 45 fractions (1.6 Gy in 3 fractions per day) over a 5-week span, with a 2-week rest in the middle. The conventional control condition involved 35 fractions of 2 Gy for a total dose of 70 Gy in 7 weeks. The results of this trial showed a 15% increase in locoregional control for accelerated fractionation, but this increase did not translate into a survival advantage. As expected, acute effects were significantly increased, but the observed increase in late effects was decidedly not ­expected; moreover, some of those effects involved complications that proved to be lethal. This EORTC trial and several others testing accelerated treatment show that attempting to keep the total

dose as high as 66 to 72 Gy while shortening the overall time by as much as 2 to 3 weeks from a conventional 6 or 7 weeks leads to serious problems with late complications. There are probably two reasons for this. First, the late effects observed are consequential late damage (i.e., late damage developing out of the very severe acute  effects). Second, repair is probably incomplete between dose fractions, particularly for protocols involving 3 fractions per day, because any unrepaired damage in the first interval would accumulate in the second interval in each day. Moreover, the interval between fractions in the early years of the EORTC trial was only 4 hours. Continuous Hyperfractionated Accelerated Radiation Therapy Protocol: The CHART Trial. The strategy underlying the CHART trial was to use a low dose per fraction to minimize late effects and a very short overall time to minimize tumor proliferation.42 In this unique study, conducted at the Mount Vernon Hospital in association with the Gray Laboratory, patients were given a total dose of 50.4 to 54 Gy in 36 fractions over 12 consecutive days, with 3 fractions delivered daily and an interfraction interval of 6 hours. The dose per fraction was 1.4 to 1.5 Gy. By conventional standards, the total dose was very low, but it was delivered in a very short time. This protocol produced good local tumor control; despite the severe acute reactions experienced, patients reportedly favored the protocol because treatment was concluded quickly. The incidence of late effects in general was not increased, and by some ­measures, the incidence was decreased. The notable ­exception was damage to the spinal cord. Several myelopathies were recorded at total doses of 50 Gy, probably because an interfraction interval of 6 hours was not sufficient to allow full repair of sublethal damage in this tissue. CHART is the only fractionation schedule that resulted in a lower incidence of late complications. The severe but tolerable acute reactions associated with this protocol did not translate into late sequelae, probably because the total dose was relatively low. Effectiveness in tumor control was not lost, even at this low dose, because the overall treatment time was extremely short, minimizing tumor cell proliferation. Compliance was also high because the acute reactions did not peak and become uncomfortable until ­after the end of treatment. Radiation Therapy Oncology Group Trial 9003. In the largest fractionation trial conducted (1073 evaluable patients), trial 9003 by the Radiation Therapy Oncology Group (RTOG) compared a standard fractionation protocol (70 Gy delivered at 2.0 Gy per fraction) with three other fractionation protocols in patients with stage III to IV, M0 squamous carcinoma of the oral cavity, oropharynx, or supraglottic larynx or with stage II to IV carcinoma of the base of the tongue or the hypopharynx.43 Two of these protocols, one involving hyperfractionation of a total dose of 81.6 Gy given in 68 fractions twice daily over 7 weeks and the other involving accelerated fractionation by a concomitant boost with 30 fractions of 1.8 Gy in 6 weeks plus a second fraction of 1.5 Gy during the last 12 treatment days (total dose of 72.0 Gy given in 42 fractions over 6 weeks), produced significantly higher local control rates than did standard fractionation. A third accelerated fractionation scheme with a total dose of 67.2 Gy given in 42 fractions over 6 weeks with an interruption after 38.4 Gy produced a local control rate similar to that of

1  Physical and Biologic Basis of Radiation Therapy

25

the standard fractionation protocol. The incidence of acute effects was higher among those treated with the nonstandard fractionation regimens than for those treated with the standard regimen, but the late effects were similar. These results provided immediate opportunities for mathematical modeling of radiation effects, especially the interplay of time and total dose in treatment outcome.44,45 Accelerated Hyperfractionated Radiation ­ Therapy with Carbogen Breathing and Nicotinamide: The ARCON Protocol. Also worthy of mention is a strategy involving accelerated hyperfractionated radiation therapy while the patient breathes carbogen (95% O2 and 5% CO2) with nicotinamide. The goals of this regimen, known as the ARCON protocol, were to accelerate treatment to avoid tumor proliferation, to hyperfractionate the doses (small doses per fraction) to minimize late effects, to add carbogen breathing to overcome chronic hypoxia, and to add nicotinamide to overcome acute hypoxia. Clinical trials of this complex but imaginative protocol are being conducted.  Early results of a trial conducted in The Netherlands for patients with advanced laryngeal cancer were spectacular compared with historical controls.45a Results of a prospective, randomized trial have yet to be published.

of hypoxic cell radiosensitizers (discussed further under “Modifiers of Radiation Effects”). Nevertheless, in the case of relatively rapidly growing tumors such as head and neck cancer, ­overall treatment time can be a dominant factor in determining outcome. Overall Treatment Time and Local Control. One of the most important lessons from fractionation studies is that local control is lost when overall treatment time is prolonged. For head and neck cancer in particular, local control is reduced by 0.4% to 2.5% for each day that the overall treatment time is prolonged.44,45 In other words, after the first 4 weeks of a fractionated schedule, the first 0.61 Gy of each day’s dose fraction is required to overcome proliferation from the previous day. Similar estimates can be made for carcinoma of the cervix, in which a mean of 0.5% local control (range, 0.3% to 1.1%) is lost for each day that the overall time is prolonged. On the other hand, the overall treatment time is not so critical for cancer of the breast or prostate in which tumor cell proliferation is not so rapid. Although estimates of Tpot have not proved useful as predictors for individual patients, mean Tpot values for groups of patients are in accord with the importance—or lack thereof—of overall treatment time.

Lessons Learned from Fractionation Studies

Accelerated Repopulation

Trials involving changes in fractionation patterns have produced several important results. The first of these ­“lessons learned” is that hyperfractionation seems to ­ confer an ­unequivocal benefit in the treatment of head and neck ­cancer in terms of local control and survival, without ­producing any increase in late sequelae. However, accelerated treatment should be used cautiously given the findings from the EORTC trial of an unexpected increase in serious late complications. Particular caution is necessary if the ­spinal cord is in the treatment field for twice-daily treatments because repair of sublethal damage is slow in this tissue. A second lesson is that late effects depend ­primarily on total dose and dose per fraction; the overall treatment time (within the usual therapeutic range) has little influence. Another important lesson is that overall treatment time affects acute effects and tumor control. Gaps in treatment should be avoided because they lead to an increase in overall treatment time and a concomitant decrease in tumor control. The importance of overall treatment time was illustrated dramatically by Overgaard and colleagues46 in their retrospective analysis of three consecutive trials of the Danish Head and Neck Cancer (DAHANCA) study (Fig. 1-22).46 All three trials involved a total dose of 66 to 68 Gy. The first trial, a split-course regimen that extended over a total of 9.5 weeks, produced a 3-year local control rate of 32%. The second trial, involving 5 fractions per week given over 6.5 weeks, had a 3-year local control rate of 52%. The most recent trial, in which 6 fractions were given per week and the overall treatment time was shortened to 5.5 weeks, produced a 3-year local control rate of 62%. No change in late effects was observed among the three treatment groups, but as expected, the acute reactions increased as the overall treatment time decreased. These intriguing findings must be viewed with some caution, however, because the three treatment groups came from different trials conducted over a period of several years; moreover, the initial purpose of the trials was to investigate the usefulness

Radiation or any other cytotoxic agent can trigger surviving clonogens in a tumor to divide faster than they did before treatment. This concept of accelerated repopulation is illustrated for a rat rhabdomyosarcoma in Figure 1-23.47 Figure 1-23A shows the overall growth curve for the tumor and the shrinkage and regrowth that occurs after a single dose of 20 Gy. Figure 1-23B illustrates the repopulation of individual surviving cells, which after treatment are dividing with a cell cycle of only 12 hours—considerably faster than before irradiation. The rapid proliferation of surviving clonogens occurs while the tumor is overtly shrinking and regressing. This same phenomenon occurs in human tumors, ­although evidence to support this must be indirect. For ­example, from a survey of reports in the literature of ­radiation therapy of head and neck cancer, Withers and colleagues48-50 concluded that the dose required to achieve local control in 50% of the cases (TCD50) increased with overall treatment time beyond approximately 28 days. A dose increment of about 0.6 Gy/day is required to compensate for this accelerated repopulation of clonogens. Such a dose increment is consistent with a clonogen doubling time of 4 days, compared with a median value of about 60 days for unperturbed growth. The conclusion to be drawn from these findings is that radiation therapy, at least for head and neck cancer and probably in other instances, should be completed as soon as is practical after it has begun. It may be better to delay initiation of treatment than to introduce delays during treatment. If overall treatment time is too long, the effectiveness of later dose fractions will be adversely affected because the surviving clonogens in the tumor have been triggered into rapid repopulation. The experimental data referred to earlier all relate to radiation therapy. However, similar considerations may apply to chemotherapy or to a combination of radiation therapy and chemotherapy. As discussed in Chapter 4, some evidence exists to suggest that in some human malignancies,

26

PART 1  Principles

Split-course RT 66Gy/33 fractions/9.5 weeks (DAHANCA 2, 1979–1985)

Conventional RT 66Gy/33 fractions/6.5 weeks (DAHANCA 5 and 7, 1986–1996)

Accelerated RT 66Gy/33 fractions/5.5 weeks (DAHANCA 7, 1992–1996)

Tumor control (%)

80

9.5 weeks 6.5 weeks 5.5 weeks

60

77%

50%

65% 50%

55%

40 22% 20 0 Well-/moderately differentiated

Poorly differentiated

FIGURE 1-22.  Top, Overview of the fractionation schedules used in the three Danish Head and Neck Cancer (DAHANCA) trials. Bottom, Relationships of histopathologic grade, overall treatment time, and local tumor control in the three trials. Only the welldifferentiated or moderately differentiated tumors were significantly influenced by overall treatment time. (Modified from Overgaard J, Sand Hansen H, Overgaard M, et al. Importance of overall treatment time for the outcome of radiotherapy in head and neck carcinoma. Experience from the Danish Head and Neck Cancer Study, 6th International Meeting on Progress in Radio-Oncology, Salzburg, Austria, 13-17 May 1998. Bologna, Italy: Monduzzi Editore, 1998, pp 743-752.)

the results of radiation therapy are poorer if it is preceded by a course of chemotherapy. Accelerated repopulation that is triggered by the chemotherapy may be the ­explanation.

Effect of Tumor Volume on Tumor Control by Irradiation Presumably, larger tumors need larger irradiation doses if those tumors are to be controlled. However, this topic cannot be dealt with rigorously by using radiobiologic principles. Ultimately, the most pertinent information must be derived from clinical experience. Information from animal models cannot be extrapolated to the human situation, because no information on scaling factors is available. ­Another problem is the absence of experimental data on volume-­related variation in the tolerance of normal tissue to a given radiation dose because this tolerance ultimately limits the dose delivered and the strategy adopted. Clinical experience with common epithelial tumors sheds some light on the issue of tumor control as a function of tumor volume. Because local control of very radiosensitive tumors (e.g., Hodgkin disease, nodular lymphoma, seminoma) can be achieved readily regardless of tumor size, the most meaningful data come from irradiating less radiosensitive epithelial tumors such as squamous cell carcinomas and adenocarcinomas (Table 1-4). Although most larger tumors require higher doses for control, some exceptions exist, such as squamous cell carcinoma of the glottis or the skin. In the glottis, the tumor volume is so small that sufficiently high doses have been used since the ­earliest years of radiation therapy. In the skin, the wide range of fractionation regimens used and the difficulty of comparing those regimens make these tumors poor models

for other squamous cell carcinomas arising in the mucosal surfaces or viscera.

Predicting Tumor Sensitivity or Resistance to Radiation In day-to-day practice, radiation treatment schedules are designed for the average patient with a given type of malignancy at a given anatomic site. Considerable effort has been devoted to the identification and development of assays to predict the sensitivity of a particular tumor or normal tissue in an individual patient, with the goal of tailoring the treatment to each individual.51 No assay is suitable for routine use in the clinic, but the potential value of such an assay is enormous. Predictive assays ­focus on three factors: the intrinsic radiosensitivity of the ­tumor and normal tissues (which limits the dose that can be given), the tumor’s oxygen status, and proliferative ­potential of the tumor.

Intrinsic Radiosensitivity The radiosensitivity of normal tissues and tumors varies considerably among individuals; for any group of patients given the same treatment, some will experience more severe reactions than others, and a small proportion will experience unacceptable late sequelae. The means of prospectively identifying in advance which individuals are particularly sensitive to radiation would be of tremendous benefit for clinical treatment. Unfortunately, no assay has been identified that would be suitable for routine clinical use. A group at The University of Texas M. D. Anderson Cancer Center found a link between late effects and the radiosensitivity of fibroblasts grown from skin biopsy specimens.52 West and colleagues53-56 in the United Kingdom

1  Physical and Biologic Basis of Radiation Therapy 1

20 Gy of 300-kV x-rays

10

27

2

1

A

2

1

10−1

10−2

1

10−3

10−1

10−4

10−2

10−5

−20 −16 −12 −8 −4

10−3 0

4

8

Fraction of clonogenic cells

Relative tumor volume

2

12 16 20 24 28 32

B

Days after irradiation

FIGURE 1-23.  Growth curves of a rat rhabdomyosarcoma show shrinkage, growth delay, and subsequent recurrence after treatment with a single dose of 20 Gy of x-rays. A, Curve 1 is the growth curve of unirradiated control tumors; curve 2 is the growth curve of tumors irradiated at time t = 0, showing tumor shrinkage and recurrence. B, Variation in the fraction of clonogenic cells as a function of time after irradiation, obtained by removing cells from the tumor and assaying for colony formation in vitro. (From Hermens AF, Barendsen GW: Changes of cell proliferation characteristics in a rat rhabdomyosarcoma before and after irradiation. Eur J Cancer 1969;5:173-189.)

TABLE 1-4.   Relationships of Biologic Dose, Tumor Size, and Control by Irradiation Total Dose (Gy)*

Control (%)

Histology

Size

Squamous Adenocarcinoma

Subclinical (4 cm

90 60

75+

Squamous

>4 cm

90

50

*Approximation

based on a minimum tumor dose of 2 Gy per f­raction and five fractions per week. Data from Shukovsky LJ, Fletcher GH. Time-dose and tumor ­volume relationships in the irradiation of squamous cell carcinoma of the tonsillar fossa. Radiology 1973;107:621-626; Fletcher GH, ­Shukovsky LJ. The interplay of radiocurability and tolerance in the irradiation of human cancers. J Radiol Electrol Med Nucl 1975;56:383-400; Fletcher GH, Shukovsky LJ. Isoeffect exponents for the ­production of dose-response curves in squamous cell carcinomas treated between 4 to 8 weeks. J Radiol Electrol Med Nucl 1976;57:825-827; Perez CA, Brady LW. Overview. In Perez CA, Brady LW (eds): Principles and Practice of Radiation Oncology, 2nd ed. Philadelphia: JB Lippincott, 1992, pp 1-63.

impractical for clinical use. Perhaps a more likely possibility for the future may be to screen patients who are to undergo radiation therapy for the presence of genes that may confer radiosensitivity or radioresistance.

Oxygen Status The oxygen status of a tumor can be assessed by the deposition of labeled nitroimidazoles in the tumor or with polarographic oxygen probes. Compounds labeled with the ­short-lived, gamma-emitting isotope iodine 123 (123I) also can be used in regions of low oxygen tension. The presence of hypoxia can be visualized by single photon emission computed tomography (SPECT).57 Eppendorf probes have a very fast response and can be moved quickly through a tumor under computer control to obtain an oxygen profile directly. A clinical trial in Germany of patients with advanced carcinoma of the cervix treated with radiation therapy has shown lower overall and recurrence-free survival rates among patients whose tumors exhibited median partial oxygen pressure (Po2) values less than or equal to 10 mm Hg as measured with the Eppendorf probe.58 Although this evidence was initially interpreted to show that radioresistance caused by hypoxia was important, later studies indicated that hypoxia leads to more aggressive and malignant tumors.59

Proliferative Potential found that radiation therapy was less likely to ­ produce local control of carcinoma of the cervix when the SF2   (i.e., fraction of tumor cells surviving 2 Gy) was greater than 0.55. However, no other studies have shown such promising results. Attempts to develop more rapid assays (e.g., assessments of chromosomal aberrations, DNA damage, cell growth rather than clonogenicity) have proved

The proliferative potential of a tumor can be expressed in terms of its Tpot, which takes into account cell cycle time and growth fraction but not cell loss. Tpot can be measured by labeling tumor biopsy specimens with BrdU and then using flow cytometry to estimate doubling time. In a trial of head and neck cancer conducted by the EORTC, local control of fast-growing tumors (Tpot < 4 days) was ­better

28

PART 1  Principles Dose (Gy) 0

2

4

6

8

100

10 12 14 16 18 20 22 24

1

Surviving fraction

Surviving fraction

OER = 2 at doses below 2 Gy

0.1

10−1

10−2

0.01

OER = 3 at large doses

FIGURE 1-24.  Cells irradiated in the presence of molecular oxygen are more sensitive to killing by x-rays than cells that are hypoxic (deficient in oxygen). The ratio of doses that produce the same level of biologic damage in the absence of oxygen and in the presence of oxygen is known as the oxygen enhancement ratio (OER). At high doses, the OER has a value of about 3; its value is close to 2 at doses below 2 Gy.

after accelerated fractionation than after conventional therapy, but no difference was found in local control of slow-growing tumors after either type of therapy.60 A later study found no correlation between Tpot and outcome after radiation therapy; although the LI and the length of the DNA-synthesis phase (Ts) correlated weakly with treatment outcome, that correlation was not strong enough for use as a predictor.61 Predictive assays are still in developmental stages. ­Although none of the markers explored has been reliable for predicting the response of a tumor to radiation therapy, some may show promise for identifying groups of patients who may benefit from altered treatment protocols such as accelerated treatment, hyperfractionation, bioreductive drugs, or neutron therapy.

Modifiers of Radiation Effects Oxygen and the Tumor Microenvironment The sensitivity of biologic material to sparsely ionizing radiation critically depends on the presence or absence of molecular oxygen. Survival curves for mammalian cells exposed to x-rays in the presence or absence of oxygen are shown in Figure 1-24. The ratio of doses needed under hypoxic and aerated conditions to achieve a given level of biologic effect is the oxygen enhancement ratio (OER). At high radiation doses, the OER has a value of between 2.5 and 3, and some evidence suggests that it may be reduced to about 2 for doses below 2 Gy. The magnitude of the oxygen effect seems to be the same for bacteria, plants, mammalian cells, and even some chemical systems, despite great differences in dose ranges. For oxygen to act as a sensitizer, it must be present during the radiation exposure, or at least during the lifetime of the free radicals that are involved in the indirect action of radiation, which is about 10−5 seconds. According to the oxygen fixation hypothesis, in the absence of oxygen, the chemical changes produced in the target molecule may

10−3 0

10

20

30

Dose (Gy)

FIGURE 1-25.  A biphasic survival curve is characteristically obtained for tumors from air-breathing mice. The steep initial portion of the survival curve reflects the killing of aerated cells; the shallower tail of the curve reflects the death of the more radioresistant hypoxic cells. (Courtesy of Sara Rockwell, MD, Yale University School of Medicine, New Haven, CT.)

be repaired; if oxygen is present, the damage is fixed (i.e., made permanent and irreparable). A question of obvious importance concerns the concentration of oxygen that is required to potentiate the effect of x-rays. The answer is that only a very small amount of ­oxygen is necessary to produce a sensitizing effect. A Po2 of about 3 mm Hg, corresponding to about 0.5%, is enough to take the radiation response halfway from fully hypoxic to fully oxygenated conditions. At concentrations characteristic of most normal tissues (about 30 mm Hg), the radiation response is essentially the same as that under fully aerated conditions or in the presence of 100% oxygen. The Importance of Oxygen in Tumor Response. Oxygen was suspected of influencing the radiosensitivity of tumors in vivo as early as the mid-1930s. However, in 1955, a paper written by Thomlinson and Gray62 led to ­overwhelming interest in the importance of oxygen in clinical radiation therapy. This paper, reporting results of a histologic study of bronchial carcinoma, concluded that tumor cords contained necrotic centers because of the limited distance that oxygen could diffuse in respiring tissue (100 to  180 μm). The authors proposed that close to the end of the diffusion range of oxygen, there would be a region, possibly two cell layers thick, in which the oxygen tension would be low enough to be intransigent to killing by x-rays but high enough to be viable. (This concept was later recognized as the mechanism underlying chronic hypoxia.) They postulated that these cells might constitute a focus for tumor regrowth and represent the limiting factor in the curability of some human tumors by x-rays. With the development of quantitative assay systems for transplanted tumors in experimental animals, by 1963, these tumors were shown to contain a proportion of viable hypoxic cells, and the

1  Physical and Biologic Basis of Radiation Therapy

Anoxic necrotic cell

Chronic hypoxia

Acute hypoxia

Normoxic

O2 70 µm

FIGURE 1-26.  Diagram illustrating the difference between chronic and acute hypoxia in a tumor. Chronic hypoxia results from the limited diffusion distance of oxygen in a respiring ­tissue that is actually metabolizing oxygen. Cells that become ­hypoxic for this reason remain hypoxic for long periods of time and eventually die and become necrotic. Acute hypoxia results from the temporary closing of tumor blood vessels. Some cells are intermittently hypoxic because normoxia is restored each time the blood vessel opens up again. At one moment one area of the tumor may temporarily be hypoxic, while minutes later a different area may be hypoxic. (Courtesy of J. Martin Brown, MD, Stanford University School of Medicine, Stanford, CA.) Modified from Hall EJ, Giaccia AJ. Oxygen effect and reoxygenation. In: Radiobiology for the Radiologist. 6th ed. Philadelphia: Lippincott Williams & Wilkins, 2006, pp 85-105.)

r­ esistance of these cells did constitute a limit for the sterilization of these tumors by single large doses of x-rays. Figure 1-25 shows a typical dose-response curve for the cells of an experimental animal tumor. Most of the cells have a dose response characteristic of fully aerated ­conditions, but some proportion of cells have a sensitivity characteristic of hypoxia. Many tumors of various histologic types have been evaluated in different animal species. The proportion of hypoxic cells in these tumors usually ranges from 10% to 15%. Although comparable data are not available for ­humans, experiments involving radiosensitizers indicate that the proportion of hypoxic cells in some skin nodules in humans may be similar. Chronic and Acute Hypoxia. Hypoxia in tumors can result from two quite different mechanisms (Fig. 1-26). As Thomlinson and Gray62 surmised, chronic hypoxia results from the limited distance that oxygen can diffuse in respiring tissues. Regions of acute hypoxia within a tumor result from the temporary blockage of a blood vessel. If the blockage were permanent, the cells downstream would eventually die and be of no further consequence. However, good evidence exists to suggest that tumor blood vessels open and close randomly and that different regions of the tumor become hypoxic intermittently. Acute hypoxia is the result of blood flow instability and is seldom a result of total ­stasis. At the moment when a radiation dose is delivered, some proportion of the tumor cells may be hypoxic; if the radiation were to be delayed, a different group of cells may be hypoxic. This concept of acute hypoxia was postulated

29

in the late 1970s by Brown63 and was later demonstrated unequivocally in rodent tumors. Areas of acute hypoxia and chronic hypoxia can be present in the same tumor,64 which may affect the choice of treatment for such tumors. One oxygen-sensing mechanism that has evolved to maintain cell and tissue homeostasis under chronic low-oxygen conditions involves the transcription factor hypoxia-inducible factor 1 (HIF-1) and its alpha subunit. The HIF-1α subunit protein is ­ rapidly degraded under aerobic conditions, but its stabilization under hypoxic conditions leads to the upregulation of several genes that promote tumor cell survival under those conditions. Preclinical studies have suggested that HIF-1α levels may be useful indicators of a tumor’s current ­oxygen status. Indicators of longer-lasting oxygenation conditions such as the chemical hypoxia marker pimonidazole have been explored in preclinical studies. The hope is that means of visualizing HIF1α levels, in combination with indicators of longer-lasting oxygenation conditions such as pimonidazole, can be used with clinical samples to indicate which areas of tumors may be experiencing transient changes in oxygen perfusion and to link the results with response to therapy.64 Reoxygenation. When an animal tumor that consists of a mixture of aerated and hypoxic cells is irradiated, the aerated cells are killed preferentially because of their greater sensitivity, and the survivors contain a preponderance of ­hypoxic cells. However, experimentation has shown that this situation is not static, but that over time after irradiation, the proportion of hypoxic cells tends to return to the preirradiation level. This shift of cells from the hypoxic to the aerated compartment is known as reoxygenation. Not surprisingly, the rapidity and extent of reoxygenation vary greatly among the different types of animal tumors that have been studied. In some tumors, such as mammary carcinomas, reoxygenation is rapid and complete. Predictably, these tumors can be readily eradicated, with acceptable normal tissue effects, by using a fractionated radiation schedule. Other animal ­ tumors, such as the transplanted osteosarcoma, reoxygenate inefficiently and slowly; they cannot be eradicated with any fractionated schedule of xrays without incurring substantial normal tissue damage. Reoxygenation cannot be studied in humans, and we are left to speculate on the patterns of reoxygenation in human tumors of different histologic types. The effect of reoxygenation on a fractionated course of radiation is illustrated in Figure 1-27. Each dose of ­radiation reduces the aerated compartment but has comparatively little effect on hypoxic cells. If the interval between radiation doses is long enough to allow reoxygenation to take place, hypoxic cells do not greatly affect the eventual outcome of the irradiation of the tumor. Few hypoxic cells are killed with each dose, but they are killed after they have become aerated as a result of the process of reoxygenation. In other words, tumors that reoxygenate quickly and efficiently can be sterilized with a fractionated course of x-rays. They cannot be sterilized with a single dose, because the resistance of the hypoxic cells would dominate. However, if the radiation is split into fractions delivered over time, with the interval between the fractions sufficient for reoxygenation to be complete, the presence of hypoxic cells has a minimal effect. Hypoxia and Tumor Progression. Evidence that lowoxygen conditions play an important role in malignant

30

PART 1  Principles REOXYGENATION

Ae

ra te d

ce

lls

X-rays

15% hypoxic cells

100% hypoxic cells

15% hypoxic cells

FIGURE 1-27.  Many animal tumors contain a mixture of aerated and hypoxic cells, with about 15% of the cells being deficient in oxygen. When a tumor is exposed to a dose of x-rays, a large proportion of the aerated cells is killed, and most of the surviving cells are hypoxic. If a time interval is allowed before a subsequent dose of radiation is administered, the initial pattern of aerated and hypoxic cells is restored. Consequently, if reoxygenation is rapid and complete between dose fractions, the presence of hypoxic cells does not affect the outcome of a multifraction regimen.

progression comes from studies of the correlation between tumor oxygenation and treatment outcome in patients and in laboratory studies of cells and animals. Findings from these studies are summarized briefly here. In a clinical study of radiation therapy for advanced cervical carcinoma performed in Germany, local control was found to be better for those tumors in which oxygen-probe measurements showed a Po2 of 10 mm Hg than for ­tumors in which the Po2 was less than 10 mm Hg. However, a similar improvement in outcome was observed for betteroxygenated tumors that were treated surgically rather than radiologically.54 This finding suggests that hypoxia may ­indicate tumor aggressiveness rather than conferring radioresistance to some cells. In a clinical study conducted in the United States, ­patients were given radiation therapy for soft tissue ­sarcoma; the extent of tumor oxygenation was found to be associated with the frequency of distant metastasis.66 Among patients with tumors having a Po2 of less than 10 mm Hg, 70% developed distant metastases, compared with 35% of those whose tumors had a Po2 of more than 10 mm Hg. These findings were particularly compelling because the primary tumor was eradicated in all patients regardless of the level of oxygenation; only the incidence of metastasis varied ­between groups. The results from this study argue strongly that tumor oxygenation level influences tumor aggressiveness. With regard to laboratory studies, at least three lines of evidence suggest a link between hypoxia and malignant progression.67 First, hypoxia seems to destabilize the genome by inducing gene amplification and increasing gene mutation. Although the mechanisms by which this takes place are far from clear, genomic instability results in the production of transformed cells with mutations that give them a survival advantage in adverse conditions. A second insight into how low-oxygen conditions may affect the aggressiveness of tumors is the observation that hypoxia ­initiates apoptosis in oncogenically transformed cells, functioning

to enhance the selection of cells with reduced apoptotic sensitivity. Cells with a defective apoptotic pathway from inactivation of the TP53 tumor suppressor gene or from overexpression of antiapoptotic genes such as BCL2 have a survival advantage in a low-oxygen environment. The third consequence of hypoxia is the induction of proangiogenic stimulatory factors. Of the factors described, vascular ­endothelial growth factor is highly specific in stimulating endothelial cell proliferation and migration.

Radiosensitizers of Hypoxic Cells Methods that have been proposed to eliminate foci of hypoxic cells within tumors, foci that are thought to lead to tumor regrowth after irradiation, include treatment in hyperbaric oxygen chambers, the use of chemical sensitizers, the use of hypoxic cytotoxins, and conversely, the use of radioprotectant substances to protect normal tissue. These methods are reviewed briefly in the remainder of this section; another approach, the use of high-LET radiation such as neutrons or heavy ions, is discussed under “Radiation Quality.” Hyperbaric Oxygen Treatment. After hypoxia was identified as being a possible source of tumor resistance, the use of hyperbaric oxygen was explored for cancer treatment, first by Churchill Davidson at St. Thomas Hospital in London and later by others. Clinical trials were performed with patients sealed in a chamber filled with pure oxygen raised to a pressure of 3 atmospheres. Results from these trials were difficult to interpret given the small numbers of subjects and the use of unconventional fractionation schemes involving a few large fractions delivered over short periods. The largest multicenter trials of hyperbaric oxygen, performed by the Medical Research Council in the United Kingdom, showed significant benefits in local control and survival for patients with carcinoma of the uterine cervix or advanced head and neck cancer but not for those with bladder cancer. Hyperbaric oxygen treatment has fallen into disuse, partly because of its cumbersome nature and partly because drugs were thought to be able to achieve the same end by simpler means. Chemical Radiosensitizers. Instead of trying to force oxygen into tissues by the use of high-pressure chambers, the emphasis in the chemical-radiosensitizing approach ­focused on the use of oxygen substitutes that diffused into poorly vascularized areas of tumors and achieved the ­desired effect by chemical means.68-70 The vital difference between these drugs and oxygen, on which their success depends, is that the sensitizers are not rapidly metabolized by the cells in the tumor through which they diffuse. They can penetrate further than oxygen and reach all of the ­hypoxic cells in the tumor, including those that are most remote from a blood supply. Three nitroimidazole-type radiosensitizers have been tested in clinical trials. Misonidazole produces appreciable sensitization of tumor cells in culture and in vivo. Of the 20 or so randomized, prospective, controlled clinical trials performed in the United States by the RTOG, none showed a statistically significant advantage for misonidazole, although several suggested a slight benefit. Moreover, the dose-limiting toxicity of misonidazole in clinical use was found to be peripheral neuropathy, which progressed to central nervous system toxicity if use of the drug was continued. This effect precluded misonidazole being used at adequate levels throughout multifraction regimens.

31

1  Physical and Biologic Basis of Radiation Therapy

Compounds Cytotoxic to Hypoxic Cells An alternative approach to designing drugs that preferentially radiosensitize hypoxic cells is to develop drugs that selectively kill hypoxic cells. It was pointed out at an early stage that the greater reductive environment of tumors might be exploited by developing drugs that are preferentially reduced to cytotoxic species in the hypoxic regions of tumors. The best-studied of these drugs is tirapazamine, which shows highly selective toxicity toward hypoxic cells in vitro and in vivo. Results from studies of this compound in laboratory and clinical trials are described in the ­following paragraphs. Laboratory Studies of Tirapazamine. Survival curves for Chinese hamster cells treated with various doses of tirapazamine are illustrated in Figure 1-28. The ratio of cytotoxicity to hypoxic cells versus aerated cells is about 100. In this system, tirapazamine is thought to be activated by the enzyme cytochrome P450. The smaller cytotoxicity ­ratio for cell lines of human origin (about 20) presumably reflects a different spectrum or different levels of enzymes.64. Results of an experiment in which a transplanted mouse carcinoma was treated with x-rays or tirapazamine, or both, are shown in Figure 1-29.72 The drug was injected  30 ­ minutes before each irradiation, which consisted of 2.5-Gy fractions given twice daily. The combination of ­tirapazamine and radiation was highly effective in ­reducing

101

O N

Surviving fraction

100

N

10−1

N NH2

O

Hypoxic cytotoxicity ratio

−2

10

10−3 10−4

Hypoxic cells Aerobic cells

10−5 100

101

102

103

104

Dose of SR-4233 (µM)

FIGURE 1-28.  Tirapazamine (SR-4233), synthesized by Stanford Research International, is cytotoxic to hypoxic cells. Its structure is shown in the inset. Dose-response curves are shown for Chinese hamster cells exposed for 4 hours to graded concentrations of the drug under aerated and hypoxic conditions. Cells deficient in oxygen are preferentially killed. (Courtesy of J. Martin Brown, MD, Stanford University School of Medicine, Stanford, CA.) 104 Median tumor volume (mm3)

The experience with misonidazole led to the evaluation of etanidazole, another 2-nitroimidazole with equivalent sensitization but much less toxicity than misonidazole. The lower neurotoxicity of etanidazole is a function of its ­shorter half-life in vivo plus its lower partition coefficient, so it penetrates poorly into nerve tissue and does not cross the blood-brain barrier. Controlled clinical trials by the RTOG in the United States and by a multicenter consortium in Europe showed no benefit from adding etanidazole to conventional radiation therapy. The third of the nitroimidazole-type radiosensitizers tested, nimorazole, is less effective as a radiosensitizer than misonidazole or etanidazole, but it is much less toxic, allowing very large doses to be given. In a Danish ­DAHANCA trial, this compound produced significant improvement in locoregional control rates and survival rates among patients with supraglottic or pharyngeal carcinoma relative to radiation therapy alone. Surprisingly, nimorazole has not been used elsewhere. Nicotinamide and Carbogen Breathing. Hypoxic cell radiosensitizers such as the nitroimidazoles were designed primarily to overcome chronic hypoxia, which is diffusionlimited hypoxia resulting from the inability of oxygen to diffuse further than about 100 μm through respiring ­tissue. However, hypoxia also arises through acute mechanisms (i.e., the intermittent blockage of blood ­vessels). ­Nicotinamide, a vitamin B3 analogue, has been shown in mouse tumors to prevent the transient fluctuations in tumor blood flow that lead to the development of acute hypoxia. The combination of nicotinamide to overcome acute ­hypoxia with carbogen breathing to overcome chronic ­hypoxia is the basis of the ARCON trials being ­conducted in several European centers. These trials also involve ­accelerated and hyperfractionated radiation therapy to avoid tumor proliferation and damage to late-responding normal tissues.

TPZ Saline

X-rays alone

103 X-ray + TPZ TPZ + x-ray

102

8 × 2.5 Gy 10

1

0

10

20

30

40

50

Days after first treatment

FIGURE 1-29.  Tumor volume is shown as a function of time after various treatments of a squamous cell carcinoma (SCCVII) transplantable mouse carcinoma. Tirapazamine (TPZ) or radiation alone had little effect. The combination of TPZ and x-rays caused significant growth delay, and giving radiation after the drug produced a slightly greater effect. (Modified from Brown JM, Lemmon MJ. Tumor hypoxia can be exploited to preferentially sensitize tumors to fractionated irradiation. Int J Radiat Oncol Biol Phys 1991;20:457-461.)

tumor ­ volume; giving the drug before the radiation was slightly more effective than giving the treatments in the reverse order. The effect of the combination was greater than additive. In parallel experiments using the same  x-ray and drug protocols, no evidence of radiosensitization or ­ additive cytotoxicity (in terms of skin reactions) was produced by adding tirapazamine to the radiation treatments. This finding substantiates the tumor selectivity of the radiation enhancement. Clinical Trials of Tirapazamine. Despite extensive laboratory findings in vitro and in vivo showing a ­greaterthan-additive effect for the combination of x-rays and

32

PART 1  Principles

t­ irapazamine, clinical trials have shown only modest success. In a phase III trial comparing cisplatin or cisplatin combined with tirapazamine for advanced (stage IIIB or IV) non–small cell lung cancer,73 patients given the combination therapy had twice the response rate and significantly longer survival than patients given cisplatin alone. Although tirapazamine did not increase systemic cisplatin toxicity, some nausea and muscle cramping were reported. The only reported trial showing a benefit from the combination of tirapazamine with radiation comes from Australia. In that trial, Rischin and colleagues74,75 showed tirapazamine, cisplatin, and radiation to be superior to fluorouracil, cisplatin, and radiation for patients with locally advanced head and neck cancer. Meta-analysis of Clinical Trials. Of the dozens of clinical trials performed in attempts to overcome the perceived problem of hypoxic cells in tumors, most have shown inconclusive or borderline results. In 1996, Overgaard and Horsman76 published the results of a meta-analysis of 10,602 patients treated in 82 randomized clinical trials involving hyperbaric oxygen, chemical sensitizers, carbogen breathing, or blood transfusions; the tumor sites studied included bladder, uterine cervix, central nervous system, head and neck, and lung. Overall, antihypoxic cell treatments were found to improve local tumor control by 4.6% and survival rate by 2.8%. The complication rate increased by only 0.6%, which was not statistically significant. The ­largest number of trials involved head and neck tumors, which also showed the greatest benefit. Overgaard and Horsman also concluded from the meta-analysis that the problem of hypoxia may be marginal in most adenocarcinomas and most important in squamous cell carcinomas.

Radioprotective Compounds In 1948, Harvey Patt and colleagues77 discovered that cysteine afforded mice considerable protection from death by whole-body irradiation. The dose required to produce lethality was almost doubled in animals that had been given the drug compared with those that had not. The structure of cysteine is illustrated as follows: NH2 SH – CH2 – CH COOH Almost simultaneously in Europe, Bacq and colleagues78 found that cysteamine was also a protector. The structure of this compound is illustrated as follows: SH – CH2 – CH2 – NH2 In the years that followed, many similar compounds were found to be effective protectors, and they all tended to have the same general structure: a free sulfhydryl (SH) group (or a potential sulfhydryl group) at one end of the molecule and a strong basic function, such as an amine or guanidine, at the other end, separated by a straight chain of two or three carbon atoms. The problem with these simple compounds was their toxicity. In the years ­ immediately

after World War II, in the wake of the first use of atomic weapons, the possibility of a compound that would protect against whole-body irradiation was of great interest to the military. Over the years, the Walter Reed Army Hospital in Washington, DC, synthesized several thousand compounds that were similar in structure to cysteine and cysteamine, with the aim of finding something less toxic. The first breakthrough in this research was the discovery that covering the sulfhydryl group with a phosphate allowed the dose resulting in a given level of toxicity to be doubled. Amifostine and Radiation Therapy. Of the many compounds produced in the Walter Reed program, the one used most often in radiation therapy is amifostine (WR2721). Amifostine is a prodrug; the presence of a terminal phosphorothioic acid group makes it relatively nonreactive, and it does not readily permeate cell membranes. Dephosphorylation by alkaline phosphatase, which is present in high concentrations in normal tissues and capillaries, converts amifostine to its active form, WR-1065. This metabolite readily enters normal cells by facilitated diffusion, where it scavenges free radicals generated by ionizing radiation or by some alkylating chemotherapeutic agents. It is taken up much more slowly by tumor tissues, perhaps because of the superior vascularization of normal tissues or by differences in the membrane structure of the tumor cells that impedes entry of this relatively hydrophilic compound. Whatever the underlying mechanism, the compound quickly floods normal tissues but penetrates tumors more slowly. Experiments with animal tumors have confirmed that concentrations of amifostine are high in several normal tissues shortly after drug administration but rise much more slowly in tumors. The extent to which amifostine protects normal tissues from radiation effects varies considerably among ­ tissue types.79 The hematopoietic system, the gut lining, and the salivary glands are generally well protected. However, because the drug does not cross the blood-brain barrier, it provides no protection to the brain and a disappointingly low level of protection to the lung. Ideally, for the maximum radioprotective effect, radioprotectants such as amifostine should be administered at the maximum tolerable concentration immediately before the radiation dose is delivered, so that the drug can be present in and protect normal tissues—but not tumor cells—when the radiation is delivered. Although radioprotectants such as amifostine have several potential applications in combination with radiation therapy, their clinical use has been slow in coming. Phase I toxicity trials of amifostine conducted in humans in the United States have shown that the dose-limiting toxicity is hypotension; this and other adverse effects such as sneezing and somnolence have tended to limit the amount of drug given to less than the dose needed to achieve maximum protection according to animal experiments. In a randomized clinical trial conducted in mainland China,80 100 patients with inoperable, unresectable, or recurrent adenocarcinoma of the rectum were stratified and randomly assigned to receive amifostine plus radiation therapy or radiation therapy only. The amifostine was administered 15 minutes before the radiation, 4 days each week, for 5 weeks. The amifostine group showed protection of the skin, mucous membranes, bladder, and pelvic structures against late,

1  Physical and Biologic Basis of Radiation Therapy

moderate, and severe reactions; none of the 34 evaluable patients in this group had such reactions, compared with 5 of 37 evaluable patients in the radiation-only group, a statistically significant difference. Moreover, amifostine afforded no apparent protection to the tumors. One of the worrisome factors in the experimental use of radioprotectors in the clinic is that their use is not failsafe. To exploit a benefit, radiation doses must be increased, with the confidence that the normal tissues are protected and that the extra dose will improve tumor response. If radioprotection does not occur, unacceptable normal tissue morbidity results. A more modest but achievable goal is to use a radioprotectant to reduce the troublesome side effects of radiation therapy. The RTOG conducted a phase III randomized clinical trial in which amifostine was found to reduce xerostomia (dry mouth) without affecting early tumor control in patients with head and neck cancer given radiation therapy.81 The drug was administered daily, 30 minutes before each dose fraction in a multifraction regimen. Three months after treatment, the incidence of xerostomia was significantly reduced in patients treated with amifostine; these patients also reported having less difficulty in eating or speaking and in the need for fluids and oral comfort aids. No difference was found in locoregional tumor control rates between patients given amifostine and those who were not. Radioprotectors and Chemotherapy. Although sulfhydryl compounds were initially developed as protectors against ionizing radiation, they also protect against the cytotoxic effects of several chemotherapeutic agents. Amifostine, for example, offers significant protection against the nephrotoxicity, ototoxicity, and neuropathy associated with cisplatin and against the hematologic toxicity associated with cyclophosphamide. In a trial conducted at the M. D. Anderson Cancer Center, investigators tested whether amifostine would reduce the acute toxicity associated with concurrent chemoradiation therapy.82 Amifostine was given twice each week at 20 to 30 minutes before doses of intravenous cisplatin, oral etoposide, or irradiation. Amifostine did reduce the incidence and severity of esophageal, pulmonary, and hematologic toxic effects but did not ­ affect survival time, suggesting that it did not have a

3-MeV proton track

Heavy ion track

33

tumor-­protective effect. This and other studies indicated that ­amifostine has no obvious antitumor activity, suggesting that its uptake by normal tissues may be different from that by malignant tissues. Mechanism of Action. The mechanisms most often implicated in sulfhydryl-mediated cytoprotection include free-radical scavenging, which protects against oxygenbased free radicals generated by ionizing radiation or ­alkylating agents, and hydrogen-atom donation, which facilitates direct chemical repair at sites of DNA damage. The protective effect of sulfhydryl compounds tends to parallel the oxygen effect and is maximal for sparsely ionizing, low-LET radiation such as x-rays or gamma rays and minimal for densely ionizing radiation such as low-energy alpha particles, which tend to damage DNA directly rather than generating free radicals.

Radiation Quality The energy deposited in biologic materials by radiation is in the form of ionizations and excitations that are not distributed at random; they tend to be localized along the tracks of individually charged particles. The pattern of this energy deposition varies substantially with the type of ionizing radiation involved (Fig. 1-30). For example, photons give rise to fast electrons, particles carrying a unit-negative electric charge and having a very small mass. Neutrons give rise to recoil protons or alpha particles, which are particles carrying 1 or 2 units of positive electric charge and having a mass approximately 2000 and 8000 times, respectively, that of an electron. The spatial distribution of the ionizing events they produce differs markedly. The tracks of energetic electrons, produced as a result of the absorption of x-ray photons, result in primary events that are well separated in space, and for this reason, x-rays are described as sparsely ionizing. Alpha particles give rise to individual ionizing events that occur close together, giving rise to tracks that consist of a column of ionizations; they are therefore said to be densely ionizing. The computer simulation of ionizing events from several different types of radiation against the background of a mammalian cell shown in Figure 1-30 is an attempt to illustrate the wide differences in ionization density associated with different types of radiation in common use.

Delta ray

Electron from 250-kV x-rays

FIGURE 1-30.  Computer simulations demonstrate sections of charged-particle tracks produced by different types of radiation passing through a strand of chromatin. Each cross represents a single ionization of the chromatin or the surrounding medium. Right track, Low–linear energy transfer (LET), 100-keV electron is typical of those produced by 250-kVp x-rays. Center track, High LET, high-energy iron ion produces a dense column of ionization; notice the high-energy, secondary delta ray coming out of the track. Left track, Medium LET, 3-MeV proton. The scale bar represents 50 nm. (Electron micrograph of the 30-nm chromatin fiber courtesy of Barbara Hamkalo, MD, University of California, Los Angeles, CA; calculated particle tracks and diagram courtesy of David Brenner, MD, New York, NY.)

34

PART 1  Principles Dose (Gy) 0

2

4

6

8

10

12

14

6

2

4 RBE

2

X-rays

0.1

OER

Surviving fraction

OER

8

1.0 RBE

1.0

0.1

3

10

0 0.1

1

10

100

1 1000

LET ∞, keV/µ RBE = 4.1

Neutrons RBE = 2.7

0.01 0

2

4

6

0.01 8

10

12

14

FIGURE 1-31.  The survival curve for x-rays is characterized by a broad initial shoulder, whereas for neutrons, the survival curve has little or no shoulder. Consequently, the relative biologic effectiveness (RBE) gets larger as the dose gets smaller. When a dose is fractionated, the RBE is larger for a given level of cell killing than if the dose is given in a single exposure because the large shoulder of the x-ray dose-response curve is repeated each time.

FIGURE 1-32.  Variation of the oxygen enhancement ratio (OER) and the relative biologic effectiveness (RBE) is a function of the linear energy transfer (LET) of the radiation involved. The data were obtained by using T1 kidney cells of human origin, which were irradiated with various naturally occurring alpha particles or with deuterons accelerated in the Hammersmith cyclotron. The RBE increases rapidly, and OER falls rapidly at about the same LET (about 100 keV/μ). (Modified from Barendsen GW. Radiobiological dose-effect relations for radiation characterized by a wide spectrum of LET: implications for their application in radiotherapy. In Proceedings of the Conference of Particle ­Accelerators in Radiation Therapy, Los Alamos Science Laboratory, Los Alamos, NM, October 2-5, 1972. LA-5180-C, 120-125. Oak Ridge, TN: U.S. Atomic Energy Commission, Technical Information Center, 1972.)

Linear Energy Transfer The quality of the radiation is described quantitatively in terms of LET, which is defined as the average energy imparted locally to the absorbing medium per unit length of track. The unit in which LET is commonly expressed is keV/μm; the kiloelectron volt is a unit of energy, and the micron (micrometer) is a unit of length. It is evident from Figure 1-30 that LET is very much an average quantity. In some regions along the track, for a considerable distance on the microscopic scale, no energy at all is deposited. In other regions, a cluster of ionizations occurs, and a large amount of energy is deposited in a small distance. The LET of 60Co gamma rays is about 0.2 keV/μm; that of 250-keV x-rays is about 2 keV/μm; that of 14-MeV neutrons is about  75 keV/μm; and that of heavy charged particles is ­anywhere from 100 to 2000 keV/μm. In general, for a given type of radiation, the LET goes down as the energy goes up. For ­example, 250-keV x-rays have a higher LET (and are more effective biologically) than are cobalt gamma rays. Likewise, for a given type of charged particle (e.g., alpha particle), a high-energy particle has a lower LET and less biologic ­effect per unit dose than a low-energy particle.

Linear Energy Transfer and Shape of Dose-Response Curve Dose-response curves from mammalian cells exposed to low-energy (high-LET) neutrons and low-LET x-rays are shown in Figure 1-31. When the LET of the radiation increases, the slope of the survival curve gets progressively steeper, and the shoulder gets smaller. The RBE of a given type of radiation is expressed in terms of standard radiation, usually taken to be low-LET photons having a LET value between 0.2 and 3 keV/μm. The RBE of neutrons is defined as the ratio of doses of x-rays to neutrons required to produce a given biologic effect. What biologic effect? No particular biologic effect is specified. At a surviving fraction of 10−2, the ratio of doses to produce the same effects is 2.7, whereas at a surviving fraction of 0.5,

the RBE is 4.0. The important point to make is that the value of the RBE for a given type of radiation (in this case, neutrons) varies with the level of biologic damage involved or, what amounts to the same thing, the size of dose involved. The effect of fractionation on RBE is illustrated in ­Figure 1-31. When a dose of x-rays or neutrons is split into multiple fractions, separated by sufficient time to allow sublethal damage repair to be completed, the shoulder of the doseresponse curve must be re-expressed with each fraction. For a given level of biologic damage, it is obvious that the RBE is greater for a schedule that involves multiple fractions than for a single exposure. The reason for this is that the large shoulder of the dose-response curves for x-rays must be repeated with each exposure, whereas for neutrons, only a small shoulder is repeated each time. For the fractionated regimen, the value of the RBE involved is that characteristic of the individual-dose fraction.

Relative Biologic Effectiveness and Linear Energy Transfer The variation of RBE as a function of LET has a characteristic shape (Fig. 1-32).83 As LET increases from 1 to 10 keV/μm, RBE increases slowly but steadily and then increases more rapidly beyond 10 keV/μm to reach a peak at about 100 keV/μm, after which RBE tends to fall for higher values of LET. The decrease of RBE at very high values of LET usually is attributed to the phenomenon of overkill. At these very high LETs, the density of ionization is such that if a charged particle track passes through a cell, it deposits far more energy than is required to kill the cell. Consequently, much of this energy is wasted because the cell cannot be killed more than once. Because so much ­energy is deposited in each cell, the radiation is relatively inefficient, and the killing effect per unit dose (which is what RBE amounts to) decreases.

1  Physical and Biologic Basis of Radiation Therapy

Linear Energy Transfer and Oxygen ­ Enhancement Ratio As LET increases, a corresponding decrease occurs in the dependence of cell killing on the presence of molecular oxygen, which is expressed in terms of the OER. This variation with LET is illustrated in Figure 1-32.83 ­­  Low-LET forms of radiation, such as x-rays or gamma rays, are characterized by a large OER of between 2.5 and 3. Heavy charged particles with an LET of several hundred keV/μm have an OER of 1 (i.e., cell killing is independent of the presence or absence of oxygen). ­Between these two extremes, neutrons with an LET of about 70 keV/μm show an intermediate OER of about 1.6. The value of the OER reflects the relative importance of the direct-killing effect of the radiation versus the indirect-killing effect, which is mediated by free radicals; this latter component of radiation damage depends on the presence or absence of oxygen.

Radiation Quality and Biologic Effectiveness RBE varies with radiation quality. In general, the more densely ionizing the radiation, the more biologically effective it is, at least up to a certain point. The biologic effectiveness of a certain type of radiation depends on several factors, including radiation quality, dose per fraction, dose rate, the presence or absence of oxygen, and the biologic system or cell type involved. In this context, radiation quality is measured in terms of the LET. The RBE depends on the dose level and the ­number of dose fractions (or, alternatively, the dose per fraction), because in general, the shapes of the dose­response relationship are different for types of radiation that differ substantially in LET. The RBE can vary with dose rate, too, because the slope of the dose-response curve for sparsely ionizing radiation varies with changing dose rate. In contrast, the biologic response to densely ionizing radiation depends little on the rate at which the radiation is delivered. The biologic system, end point, or cell type chosen to determine RBE has a marked influence on the ­values obtained. RBE values usually are high for tissues that ­ accumulate and repair a large amount of sublethal damage and consequently have large shoulders to their xray dose-response relationships, and RBE values are low for those that do not.

Use of Neutrons in Radiation Therapy Neutrons were first used in radiation therapy in the 1930s at the University of California at Berkeley. Their introduction was not based on any physical or radiobiologic ­rationale, and they were abandoned because of severe late effects produced in many patients. After World War II, the use of neutrons was reintroduced at the Hammersmith Hospital in England because neutrons have a lower OER than x-rays and because of the premise that hypoxic cells limit the curability of tumors by x-rays.84-86 After ­apparent early success, neutron machines were used for therapy in the United States, continental Europe, and ­ Japan. Controlled clinical trials for a wide range of tumors indicated little or no advantage for the new modality, except perhaps for a few disease sites, specifically prostate cancer, salivary gland tumors, and possibly soft tissue ­sarcomas.87,88 The biologic properties of neutrons differ from those of x-rays in several respects, apart from their lesser dependence on molecular oxygen for cell killing. Neutrons tend

35

to be associated with less repair of sublethal and potentially lethal damage, and they are associated with smaller variations in radiosensitivity according to the phase of the cell cycle. Perhaps because of these differences, neutron RBE values tend to be higher for slowly growing tumors. The rationale for the use of neutrons has evolved considerably over the years, but for the most part, the use of neutrons for radiation therapy has been abandoned.

Use of Protons in Radiation Therapy The biologic properties of high-energy protons do not differ significantly from those of x-rays. Protons produce ionizations and excitations in atoms along a track in much the same way as electrons are set in motion when x-rays are absorbed. The clinical use of protons is based on the superior dose distributions that can be obtained. Protons have a place in the treatment of tumors for which dose localization is important, such as tumors that are close to sensitive normal structures. Protons have been widely used to treat choroidal melanoma of the eye, sarcoma of the base of the skull, and chordomas. The increasingly impressive results of conformal radiation therapy, including intensity-modulated radiation therapy, in decreasing effects on normal tissues and allowing much higher total doses than have been used previously, suggest wider applicability of proton therapy. As a consequence, several elaborate proton facilities have been built or are under construction in the United States, Japan, and  Europe.89,90 This issue is discussed further in Chapters  2 and 40.

Use of Carbon Ions in Radiation Therapy Early investigations of the use of heavy ions for radiation therapy were done in the 1950s at the Lawrence Berkeley National Laboratory in the United States. Although interest in this therapy has declined over time in the United States, interest has been rekindled in Europe and Japan, with attention focused largely on high-energy carbon ions.91 The characteristic depth-dose profile of heavy ions, which they share with protons (i.e., dose increases with the depth of penetration), make them an attractive modality for the irradiation of deep-seated tumors. Another advantage shared by high-energy carbon ions (typically 300 MeV/μm 12C  ions) and protons is attractive physical dose distributions. Carbon ions also produce a substantial increase in RBE toward the end of the particle range. In principle, the good news of this latter feature is that the end of the particle’s range, and therefore the greatest antitumor effectiveness, can be located within the tumor. The bad news is that the appropriate RBE level to be used is unknown, which introduces an element of uncertainty. Compared with protons, carbon ions show less lateral scattering, leading to sharper beam edges (which is good). However, carbon ions produce a tail of lighter fragments beyond the Bragg peak, and they do not share the advantage of protons that the dose stops sharply at the end of the range of the primary particle.

Adjuncts to Irradiation Hyperthermia The use of hyperthermia to treat cancer has a history much longer than that of radiation. The first references are in an Egyptian papyrus roll 5000 years old. Just before the discovery of x-rays, it was observed that tumors regressed in

36

PART 1  Principles

patients with high fevers resulting from various bacterial infections.92 The interest in hyperthermia in modern times results from experiments with cells in vitro and with animal ­tumors and normal tissues. Several biologic effects of heat seem favorable for cancer therapy: • Heat kills; survival curves for heat are similar in shape to those for radiation, except that time at the elevated temperature replaces absorbed dose. • The age-response function for heat complements that for x-rays. S-phase cells that are resistant to x-rays are also sensitive to heat. • Cells that are at low pH and nutritionally deprived (more likely in tumors) are more sensitive to heat, although cells can adapt to pH changes in 80 to 100 hours and can lose their sensitivity to heat. • Hypoxia does not protect cells from heat as it does from x-rays. • Heat preferentially damages tumor vasculature. After heating, blood flow goes down in tumors but increases in normal tissues. This may result in an enhanced temperature differential between tumors and normal tissues. Good evidence exists to support this contention in transplantable animal tumors, but it is less clear in spontaneous human tumors. • Mild hyperthermia can promote tumor reoxygenation; this has been shown in animal experiments and in clinical studies. Although the biologic properties of heat seem favorable for its use in cancer therapy, physical devices that can produce uniform heating at a particular depth are not well developed, and adequate thermometry is also difficult. The basic laws of physics make the desired end difficult or impossible to achieve. Microwaves provide good localization at shallow depths, but localization is poor at greater depths, where low frequencies are needed. With ultrasound, deep heating can be achieved with focused arrays, but bones or air cavities cause distortions. Because of these considerations, in practice, primary or recurrent breast tumors and melanomas can be heated adequately with microwaves, and deep-seated tumors below the diaphragm can be ­heated adequately with ultrasound. For tumors at any other site, however, the current methods of heating pose a major problem. In contrast, the combination of interstitial implants of radioactive sources with hyperthermia generated by radiofrequency power applied to the implanted sources seems to be particularly promising because a good radiation dose distribution is combined with a good heat distribution. Many thousands of patients have been treated with ­hyperthermia. The general consensus is that hyperthermia alone has no role in the curative treatment of cancer; its usefulness is limited to palliation or retreatment of recurrent tumors. However, several clinical studies have shown that the combination of hyperthermia plus radiation produces more complete and partial tumor responses than does radiation alone. Several phase III randomized trials have shown a clear and significant benefit from the addition of hyperthermia to standard radiation therapy.93,94 The tumor sites evaluated include those of superficial localized breast cancer, recurrent or metastatic melanoma, and nodal metastases from head and neck cancer. In contrast, trials involving advanced breast cancer and neck nodes with 

full-dose radiation therapy did not show any benefit from the addition of hyperthermia. A novel and easily achievable means of producing modest levels of hyperthermia (41° to 42°C) is to trigger the ­release of drugs encapsulated in liposomes.95 Liposomes are spherical constructs, usually membranes, that can be filled with a cytotoxic chemotherapy agent such as doxorubicin or cisplatin. Injected liposomes tend to become trapped in the chaotic disorganized blood vessels of tumors, and the external application of a modest level of hyperthermia can cause them to extravasate into the tumor ­interstitium, depositing the chemotherapeutic agent. This form of ­hyperthermia-triggered drug release can have substantial therapeutic effects.96 In summary, the clinical value of hyperthermia for the routine treatment of cancer is still not clear, despite its proven efficacy in a few specific instances. The situations in which hyperthermia seemed to be effective involved its combination with external beam radiation therapy for the treatment of superficial tumors, for which adequate heating is not so difficult, or its combination with brachytherapy, for which heating is induced by implanted electrodes, inevitably involving all the complications and limitations of accessibility.

Cytotoxic Drugs Drugs with anticancer activity interact in complex and only partially understood ways with ionizing radiation. ­Although the combination is often used in clinical settings, such use is based on an empiric rationale with little regard for cellular interactions. Possible interactions of drugs with radiation are illustrated graphically in Figure 1-33.97 The drugs under consideration in this section are those with known dose-response curves; the interactions are considered at least potentially subadditive. However, they may be additive with regard to their effects on tumors and normal tissues. The object in combining anticancer drugs with ionizing radiation is to improve the therapeutic ratio relative to the use of either agent alone. It is beyond the scope of this chapter to consider cooperative interactions in which the effect of one agent is local, the other is systemic, and no interaction on tumors or normal tissues is assumed. Little is known about the interactions of cytotoxic drugs and radiation at the subcellular or molecular levels. Although much is known about the mechanisms of action of the various classes of drugs, such as alkylating agents, antimetabolites, plant alkaloids, and antibiotics, little is known about specific drug-radiation interactions. Several mechanisms by which cell death can be ­increased in vitro by the combined use of drugs and ­radiation can be identified: S-phase cytotoxicity, cell cycle ­ redistributions that increase drug or radiation effects, and ­ prevention of repair of potentially lethal damage. ­ Cellular studies have suggested preferred sequencing and timing of drugs and radiation types, as well as interactions with other treatment modalities such as hyperthermia. However, no ­ consistent correlations have been found between interactions that are apparent at the cellular level and in vivo effects. For ­example, hydroxyurea, a cell cycle S-phase–specific agent, enhances radiation effects on cells in tissue culture. ­ However, with few exceptions, the effects of this combination in vivo and in clinical settings have been ­disappointing.

1  Physical and Biologic Basis of Radiation Therapy One agent inactive by itself Sensitization

Fixed doses

One D/E curve

Cooperation Enhancement

37

Two D/E curves Supraadditive

Positive interaction Noninteractive

Noninteractive

Negative interaction Protection

Antagonism

Inhibition

Subadditive

Protection

Conceivably additive response

FIGURE 1-33.  Suggested terminology for drug-radiation interactions is derived from four experimental situations: the drug is inactive by itself, fixed doses of drug and radiation are combined, the dose-response (D/E curve) for one agent is studied with and without a fixed dose of another agent, and dose-response curves for both agents are available. (From Steel GG. Terminology in the description of drug-radiation interactions. Int J Radiat Oncol Biol Phys 1979;5:1145-1150.)

Clinical data suggest that normal tissues are at least as likely as tumors, if not more so, to be affected by drug­radiation interactions. Notable examples of enhanced acute and late effects with radiation therapy include those seen with simultaneous administration of doxorubicin, dactinomycin, fluorouracil, methotrexate, bleomycin, nitrosoureas, and possibly etoposide. In 1980, Tubiana98 stated that “the use of drugs during radiation therapy has never improved control of ­local ­lesions or reduced frequency of metastases from any ­tumor.” Clinical data convincingly show that local control can be enhanced with simultaneous chemotherapy and irradiation. SchaakeKoning and colleagues99 reported statistically significant increases in locoregional control of non–small cell carcinoma of the lung when cisplatin was given, weekly or daily, simultaneously with thoracic irradiation. Concurrent fluorouracil and pelvic irradiation has been shown to decrease local failure rates when given after resection of adenocarcinoma of the rectum.100 Improved local control has resulted from the use of vincristine and cyclophosphamide (with or without dactinomycin and doxorubicin) for Ewing’s sarcoma of bone and for embryonal rhabdomyosarcoma of the upper aerodigestive tract. Small cell carcinoma of the lung has been controlled much more consistently with concurrent ­chemotherapy and radiation therapy than with radiation therapy or chemotherapy alone. A French cooperative group101 led by investigators from Tubiana’s institution showed a ­significant reduction in the frequency of distant metastasis and improved long-term survival with induction chemotherapy (i.e., vindesine, lomustine, cisplatin, and  cyclophosphamide) followed by thoracic radiation therapy for non–small cell carcinoma of the lung, despite local control not being improved. Herskovic and colleagues102   reported significant decreases in rates of local failure and distant metastasis and increases in survival when fluorouracil and cisplatin were given concurrently with moderate-dose radiation therapy (50 Gy given in 25 fractions) ­compared

with higher-dose radiation therapy alone (64.8 Gy in  36 fractions).

Targeted Therapy Achieving tumor response with traditional chemotherapy agents such as the alkylating agents cisplatin or fluorouracil inevitably involves substantial toxicity to self-renewing normal tissues such as the gastrointestinal tract and bloodforming organs that are characterized by large growth ­fractions. Newer targeted therapeutic agents show some promise when given in combination with radiation therapy in terms of enhancing tumor control with little increase in normal-tissue toxicity. Cetuximab (Erbitux) can be considered a model for this class of agent; other examples that target specific cellular signaling pathways include trastuzumab (Herceptin), imatinib mesylate (Gleevec), and rituximab (Rituxan). Cetuximab is a recombinant human/mouse chimeric monoclonal antibody originally produced in mammalian cell culture. Cetuximab specifically binds to the epidermal growth factor receptor, blocking the phosphorylation and activation of receptor-associated kinases and resulting in inhibition of cell growth, induction of apoptosis, and decreased production of matrix metalloproteinases and vascular endothelial growth factor. Studies of cetuximab for the treatment of xenografted tumors in mice indicate that cetuximab given alone produces only a modest tumor-growth delay but that it shows substantial interaction with radiation.103 These promising preclinical results led to a phase III trial that demonstrated that adding cetuximab to high-dose radiation therapy for locoregionally advanced squamous cell carcinoma of the head and neck produced statistically significant extensions in overall survival time with minimal enhancement of toxicity (mostly skin reactions).104 The substantial effects of cetuximab and other forms of targeted therapy in experimental systems have translated into relatively modest gains

38

PART 1  Principles

in the clinic; however, this approach continues to be the subject of intense research.

Biological Response Modifiers Biological agents, once limited to derivatives of CalmetteGuérin bacillus, are being discovered at a rate unimaginable only a few years ago. Such agents have been considered possibly restorative of cell-mediated immunity in patients with immune suppression, whether from cancer or the modalities with which cancer is treated. Very few laboratory studies and even fewer clinical investigations are available describing possible interactions between local irradiation and biological agents.

Gene Therapy The term gene therapy covers several different ap-  proa­­ches,105 but all require some means of introducing a gene into tumor cells. The most common of these means are viral vectors. Among the several types of vectors used, retroviruses are convenient to work with but can infect only ­ dividing cells; adenoviruses can infect dividing and quiescent cells, but they evoke an immune response that can make their repeated use difficult. Derivatives of the herpes simplex virus are attractive because of their large size, which allows more material to be packaged in them, but they can be pathogenic and difficult to control. In most cases, the virus is used as a simple vector to deliver a gene of interest to the cells. For safety’s sake, the virus in these cases is usually engineered so that it cannot replicate. In a few cases, the virus itself is designed to be cytotoxic, and some means is needed to selectively limit its cytotoxicity to tumor cells. Several different forms of gene therapy are described briefly in the following sections. Suicide Gene Therapy. The strategy in suicide gene therapy is to transduce cells with a gene that can convert an inert prodrug (administered separately) into a toxic agent. One system that has been extensively investigated is the combination of the herpes simplex virus (HSV) thymidine kinase gene (HSV-TK) with ganciclovir.100 Several steps are involved in suicide gene therapy. The virus containing the HSV-TK gene, which phosphorylates the ganciclovir, is injected into the tumor, and the ganciclovir is administered systemically. Within the transduced cells, phosphorylation of the ganciclovir by the thymidine kinase converts it into a cytotoxic agent, killing the cells. Significant numbers of nontransduced cells also are killed (i.e., bystander effect) through various other transport and immune mechanisms. An alternative suicide gene system involves cytosine deaminase, which converts 5-fluorocytosine (nontoxic) to fluorouracil (toxic) within the cells.107 A fusion gene that combines the HSV-TK and cytosine deaminase systems in a single viral vector has also shown promise. Suicide gene therapy has led to tumor growth delays and some outright cures in animal models, but it has been less successful in clinical trials. Because of the ­ limited ­efficiency of gene delivery, suicide gene therapy needs to be combined with conventional radiation therapy.106   Phase I/II clinical trials of double-suicide gene therapy in combination with radiation therapy for men with prostate cancer have produced encouraging results.108,109 Cytotoxic Viruses. A variant of gene therapy involves the construction of cytotoxic viruses that are engineered to replicate and kill only cells with mutations in TP53, which

influences a variety of critical cellular functions including transcription, cell cycle checkpoints, DNA repair, ­apoptosis, and angiogenesis. To the extent that mutant TP53 is a hallmark of cancer, this treatment distinguishes between ­normal and cancer cells on that basis. This approach has produced evidence of tumor growth arrest in animal models and in early clinical trials.110-112 Immunomodulatory Gene Therapy. The basis of the immunomodulatory gene therapy approach (e.g., cancer vaccines), also known as molecular immunotherapy, is to provoke or stimulate a cellular immune response to cancer cells by injecting a vaccine consisting of suspensions of irradiated tumor cells that have been genetically engineered to express immune stimulatory molecules or tumor-specific antigens. This approach has been somewhat successful in animal models but generally is effective against only small tumor burdens. Tumor Suppressor Gene Therapy. In its purest form, tumor suppressor gene therapy seeks to replace a gene in which a mutation initiates or otherwise enhances the malignant phenotype. The goal of treatment may be to modify cell growth, invasiveness, or metastatic potential as much as it is to kill the cell. The TP53 gene has received the most attention as a target for such therapy because it is so commonly mutated in human cancer.113-122 Phase I/II clinical trials of tumor suppressor gene therapy have shown some positive results in the treatment of non–small cell lung ­cancer. The chief limitations to tumor suppressor gene therapy are the lack of information on exactly which target genes are essential for inducing or maintaining the malignant phenotype and the apparent necessity of multiple genetic changes for cancer to develop. Radiation-Activated Gene Therapy. The strategy behind radiation-activated gene therapy, also called radiogenetic therapy, combines the physics of radiation-­targeting technology with molecular aspects of gene therapy.  A ­ radiation-inducible promoter sequence is linked to a cyto­toxic agent, which is then “turned on” only in carefully delineated radiation fields.123-125 In one animal model system, the radiation-inducible Egr1 promoter is linked with tumor necrosis factor α and then packaged in an adenovirus. The adenovirus is injected into the tumor, but the cytokine is activated only in the target volume delineated by the radiation field, thereby minimizing total-body toxicity. Released in the tumor, the tumor necrosis factor causes vascular destruction and apoptosis of tumor cells. Efforts are underway to improve this approach by identifying and using more radiation-specific or tumor-specific promoter genes and more effective toxic agents. The strategies described here have all had some measure of success in reducing tumor growth, if not eliminating tumors entirely, in animal model systems, and some have progressed to phase II trials in patients. Some forms of gene therapy claim to evoke bystander effects in which more cells are killed than are initially transduced. Nevertheless, all of the strategies share the difficulty of transducing therapeutic genes into sufficiently large proportions of tumor cells. Growth reduction is common, but cures are rare. For this reason, a logical strategy would be to combine one or more of the gene therapy approaches with radiation therapy or chemotherapy, with the hope of translating a small increase in cell killing from the gene therapy into

1  Physical and Biologic Basis of Radiation Therapy

a large increase in tumor control from the conventional therapies. The guiding principle must be to identify combinations that have additive or synergistic effect with regard to tumor cell kill but have tolerable, non-overlapping toxic effects on normal tissues.

Mutagenesis and Carcinogenesis When radiation is absorbed in biologic material, several consequences can result, including cell death, hereditary effects, and carcinogenesis. Cell death is the end point of principal concern in radiation therapy, in which the object of the treatment is to kill (i.e., remove the reproductive integrity of) as many malignant cells as possible. It is also the end point of concern for some radiation effects on the developing embryo and fetus, which are cell-depletion phenomena. Hereditary effects may be promoted. Radiation may affect the germ cells in a way that does not kill the cells but allows them to survive and carry with them some legacy of the radiation exposure. These effects will not be expressed until a later generation. Carcinogenesis may occur. The radiation may affect a somatic cell, leading to a point mutation or a chromosome rearrangement that results, for example, in the release or expression of an oncogene. This may result in leukemogenesis or carcinogenesis. Hereditary effects and carcinogenesis are potential hazards for people who are occupationally exposed to ionizing radiation and for patients who are long-term survivors of radiation therapy. These effects are discussed in greater detail in the remainder of this chapter.

Hereditary Effects As early as 1927, Müller reported that exposure to x-rays could cause observable hereditary mutations in Drosophila melanogaster, including changes of eye color, the appearance of vestigial wings, and recessive lethal mutations.126 With the advent of the nuclear age after World War II, ­hereditary effects were considered to be the principal hazard of ionizing radiation, in part because solid tumors had not yet appeared in atomic bomb survivors. Hereditary diseases are classified in three principal categories: mendelian, chromosomal, and multifactorial. Mendelian diseases are caused by mutations in a single gene; they can be autosomal dominant, autosomal recessive, or sex linked. Chromosomal diseases are caused by gross abnormalities in the structure or number of chromosomes. Multifactorial diseases have a genetic component, but transmission cannot be described as a simple mendelian pattern. By far the most common type of hereditary diseases, multifactorial diseases include everything from cleft palate to coronary heart disease. Estimating Risks of Radiation-Induced Hereditary Effects. In the 1960s, relative mutation rates were assessed in the megamouse project, in which about 7 million mice were used.127,128 A species was chosen that had seven easily identified specific locus mutations, six of which involved coat color changes and the seventh a shortened ear. Male and female animals were irradiated with graded doses, subsequently bred, and mutations scored in the offspring. Conclusions from this project are as follows: 1. The radiosensitivity of different mutations varies substantially, by a factor of about 20.

39

2. A dose-rate effect exists in the mouse, unlike Droso­ phila, such that spreading radiation over time greatly reduces the number of mutations observed. 3. The male mouse is much more sensitive than the female, and at low dose rates, essentially all of the genetic load is carried by the male. This may be an artifact of the species because murine oocytes are exquisitely sensitive to radiation, and if they are killed by the radiation, mutations produced in them cannot be expressed. 4. The number of mutations produced by a given dose of radiation drops if a time interval is allowed to elapse between irradiation and mating. This finding is already used as a basis for human genetic counseling. If an individual is exposed to a large dose of radiation accidentally or as a consequence of medical procedures, a planned conception should be delayed to minimize the genetic consequences of the radiation. 5. The megamouse project indicates that the doubling dose is about 1 Gy. Estimates of the risk of radiation-induced hereditary ­effects in humans can be made with knowledge of two ­factors—the baseline rate of spontaneous mutations, which is known, and the doubling dose, which is not known for humans but has been estimated from mouse experiments to be about 1 Gy.129 Estimates of the risk of hereditary effects resulting from radiation exposure, published by the International Commission on Radiological Protection (ICRP), are 0.2% per sievert (Sv) for the whole population (including children) and 0.1%/Sv for the working population (those 18 to 65 years old). However, these risks can be considered in light of detriment, a term introduced to account for years of life lost or years of life spent impaired. The detriment from radiation-induced hereditary effects from radiation is far smaller than the detriment from  radiation-induced carcinogenesis or from the effects of  radiation on an embryo or fetus. Some direct information on the hereditary effects of radiation on humans is available from the detailed studies of the survivors of the atomic bomb attacks on Hiroshima and Nagasaki and their children. Among the many factors documented in the children are congenital defects, physical development, the sex of child, ­cytogenetic abnormalities, and electrophoretic variations in blood proteins.130 No statistically significant excess in hereditary effects has been found in these children, but in statistical terms, the numbers are relatively small compared with the number of animals used in the megamouse project.

Carcinogenesis In sharp contradistinction to the situation for the hereditary effects of radiation, an abundance of information exists concerning radiation carcinogenesis directly from humans. The human experience can be summarized as follows: 1. Early workers, largely physicists and engineers working around accelerators, showed an increased incidence of skin cancer and other forms of tumors. 2. Miners in Saxony in the late 1800s and early 1900s and, more recently, uranium workers in the Central Colorado Plateau showed an increased incidence of lung cancer as a result of inhaling radon in poorly ventilated mines.

40

PART 1  Principles

3. Workers who painted luminous dials on clocks and watches and who ingested radium by using their tongues to lick the tip of their brushes into a point subsequently developed bone tumors. 4. Patients in whom the contrast material Thorotrast (which contains radioactive thorium) was used subsequently developed liver tumors. 5. The survivors of Hiroshima and Nagasaki have shown a whole spectrum of malignancies, from leukemia to solid tumors, particularly cancer of the gastrointestinal tract, breast, thyroid, and lung. 6. Children given therapeutic doses of radiation for infected tonsillar or nasopharyngeal lymphoid tissue or what was perceived to be an enlarged thymus subsequently developed benign and malignant thyroid tumors and leukemias. 7. Children epilated by x-rays as part of the treatment for tinea capitis subsequently developed thyroid and skin tumors. 8. Patients with ankylosing spondylitis who received x-rays for the relief of pain subsequently had an increased incidence of leukemia. 9. Women who underwent fluoroscopy many times during the management of tuberculosis had an increased incidence of breast cancer. The best quantitative data come from the Japanese atomic bomb survivors. For solid tumors, risk seems to be a linear function of dose between about 0.2 and 2.5 Gy. For leukemia, a linear-quadratic relationship fits the data more closely. For protection purposes, it is assumed that these risks observed at high doses can be extrapolated linearly to low doses and that no dose threshold exists below which there is no risk. The Latent Period. In leukemia, the latency period (i.e., time between the radiation exposure and the appearance of the disease) is relatively short, with a peak excess incidence occurring by 7 to 10 years after irradiation and with the radiation-induced incidence disappearing by about 15 years after irradiation. In the case of solid tumors, the latent period may be considerably longer; for example, the excess incidence of solid tumors is still evident in the Japanese survivors more than 60 years after the atomic bomb attacks. Part of the latency may have resulted from a form of induction period before the initially transformed cell or cells start to divide to form a tumor or, alternatively, to a period before the tumor assumes the malignant characteristics of growth and spread. For example, at short intervals after irradiation, the thyroid gland contains tumors that are of a benign histologic character, and malignant tumors become detectable only at later stages. It is widely believed from animal studies that tumorigenesis consists of three separate stages: initiation, promotion, and progression. Some chemicals can only initiate, whereas others can only promote. Radiation seems to be capable of both. The perception of the latency period has changed. ­Radiation-induced tumors appear not at a fixed time ­after exposure, but rather at the age at which spontaneous  tumors of the same type are most prevalent. For example, ­radiation-induced breast cancers in women appear in middle age, regardless of whether the radiation exposure occurred when the women were teenagers or in their 30s. This fact suggests that radiation may initiate the process at a young age but that completion requires additional steps, some of which may depend on hormones.

Absolute versus Relative Risk. The preceding information leads to the discussion of relative versus absolute risk models in radiation carcinogenesis. Although they can seem unduly complex, use of one model over another can produce substantial differences in risk estimates of carcinogenesis, and some basic understanding of their principles is warranted. The absolute risk model assumes that radiation produces a discrete crop of malignancies that are unrelated to the natural or spontaneous incidence. This seems to be the case for leukemia; radiation-induced leukemia in Japanese survivors of atomic bomb blasts was observed for some period after the exposure, but the incidence subsequently returned to that characteristic of the control population. The relative risk model assumes that radiation increases the spontaneous incidence by some fixed factor. According to this model, the radiation-induced cancer incidence will increase with the age of the population as the spontaneous incidence increases. This seems to be the case for the Japanese atomic bomb survivors as the population ages and the data mature; a relative risk model would predict a large number of tumors occurring late in the life of the irradiated population. The currently favored model is a relative risk model with allowance for age at exposure and time since exposure. Another point of interest is that risk for three of the four common types of leukemia seems to be increased by radiation, but the risk of chronic lymphocytic leukemia has never been reported to increase from exposure to radiation. The ICRP, working with data published by the Committee on the Biological Effects of Ionizing Radiation (BEIR V)131 and the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR 88),132 has generated the following risk estimates for ­ excess cancer ­ mortality from radiation exposure. For a working population (men and women between 18 and 65 years old), the lifetime risk of death from cancer is 0.08/Sv for high-dose or high-dose-rate radiation and 0.04/Sv for low-dose or low-dose-rate radiation. Comparable values for the general population are slightly higher because of the higher susceptibility of the young; those values are 0.1/Sv for high-dose, high-dose-rate radiation and 0.5/Sv  for low-dose, low-dose-rate radiation. Acknowledging the assumptions inherent in the relative-risk projection model, the ICRP further estimates that, on average, 13 to 15 years of life are lost for each radiation-induced cancer, with death occurring, on average, at 68 to 70 years of age.133 As the data from Japan have matured and more ­detailed information has become available, it has ­ become evident that the risk of radiation-induced cancer varies considerably with age at the time of exposure. In most cases, ­individuals exposed at an early age are much more ­ susceptible than are those exposed at older ages. This ­ difference seems to be the most dramatic for female breast cancer; children exposed before 15 years of age are most susceptible, whereas women 50 years old or older at ­ exposure show little or no excess risk. There are ­ exceptions to this general rule. Susceptibility to radiation-induced leukemia is relatively constant throughout life, and susceptibility to respiratory cancers increases in middle age. ­However, the overall risk drops drastically with age; children and young adults are much more susceptible to radiation-induced cancer than are middle-aged and older persons (Fig. 1-34).131,133

41

1  Physical and Biologic Basis of Radiation Therapy PERCENTAGE INCREASE IN RELATIVE RISK FOR RT vs. SURGERY %

Risk per unit dose (percent per Sv)

16 14 All years 12

All solid tumors

5+ years

10

10+ years

8 ICRP 60

0

BEIR V

6

10

20

30

40

50

60

SECOND CANCERS AFTER PROSTATE RT

4 % contribution to total number of radiationinduced second cancers

2 0 0

10

20

30

40

50

60

70

80

Age of acute exposure

FIGURE 1-34.  The attributable lifetime risk from a single, small dose of radiation at various ages at the time of exposure is described by two curves. Notice the dramatic decrease in radiosensitivity with age. The higher risk for the younger age groups is not expressed until late in life. These estimates are based on a relative risk model and on a dose and dose-rate effectiveness factor (DDREF) of 2. BEIR, Biological Effects of Ionizing Radiation study; ICRP, International Commission on Radiological Protection study. (Modified from ICRP. Recommendations. Annals of the ICRP publication 60. Oxford, England: Pergamon Press, 1990.)

Second Malignancies in Patients Given Radiation Therapy The risk of second malignancies after radiation therapy is a controversial topic for several reasons, among them that patients undergoing radiation therapy are often at high risk for a second cancer because of lifestyle factors, and these factors may be more dominant than the radiation risk. Although many single-institution studies involving radiation therapy for a variety of tumors have concluded that radiation is not associated with an increase in second malignancies, many of the studies were much too small to detect any increased incidence of second malignancies induced by the treatment. Larger studies have consistently indicated that radiation therapy is associated with a small but statistically significant enhancement in the risk of second malignancies, particularly among long-term survivors. To credibly answer the question of whether radiation therapy is associated with a risk of second malignancies, a study must have (1) a sufficiently large number of patients, (2) a suitable control group (i.e., patients with the same cancer treated by some means other than radiation), and (3) a sufficiently long follow-up period for radiationinduced solid tumors to manifest. Studies that satisfy these criteria describing patients treated for prostate cancer, cervical cancer, and Hodgkin disease are discussed briefly in the following sections. Second Malignancies after Treatment for Prostate Cancer. In the largest study of its kind, Brenner and colleagues134 used data from the National Cancer Institute’s Surveillance Epidemiology, and End Results (SEER) Program to evaluate outcomes among 51,584 men with ­prostate cancer treated with radiation and 70,539 treated

Sarcoma (in field) 6% Sarcoma (out of field) 2%

Rectum 12%

Bladder 37%

Lung 34%

Colon 9%

FIGURE 1-35.  Top, Percentage increase in the relative risk for all solid tumors (except prostate cancer) for individuals who underwent radiation therapy (RT) for prostate cancer relative to the risk for individuals who underwent surgery for prostate cancer. Bottom, Distribution of radiation-induced second cancer at 5+ years after the radiation therapy. (Illustration courtesy of David Brenner, MD, New York, NY; data from Brenner DJ, Curtis RE, Hall EJ, Ron E. Second cancers after radiotherapy for prostate cancer. Cancer 2000;88:398-406.) Modified from Hall EJ, Giaccia AJ. Radiation carcinogenesis. In: Radiobiology for the Radiologist. 6th ed. Philadelphia: Lippincott Williams & Wilkins, 2006, pp 135-155.

with surgery. No evidence was found of a difference in the risk of leukemia for radiation-therapy versus surgery patients, but the risk of a second solid tumor at any time after diagnosis was significantly greater after radiation therapy than after surgery (Fig. 1-35).134 This increased relative risk became greater with time, reaching 34% after 10 years or more. The greatest increases in relative risk were for tumors of the bladder (77%) and rectum (105%) at 10 years or more after diagnosis. The increased risk of developing sarcoma in the heavily irradiated tissues within the field amounted to 145% at 5 years or more, compared with surgical patients. The increase in relative risk for carcinoma of the lung that was exposed to relatively low doses (about 0.5 Gy) was close to that for carcinoma of the bladder, rectum, or colon, all of which were subject to much higher doses (typically more than 5 Gy). This pattern may reflect the fact that carcinoma originating in actively dividing cells or cells under hormonal control can be efficiently induced by relatively low doses of radiation, as evidenced by the atomic bomb survivors, but that the cancer risk at high doses decreases because of cell death. In contrast to this pattern for radiation-induced carcinomas, radiation-induced sarcomas usually appeared only in heavily irradiated sites (i.e., close to the treatment volume), where large radiation doses were needed to produce sufficient tissue damage to stimulate cellular renewal in otherwise dormant cells. Second Malignancies after Treatment for Cervical Cancer. In a landmark study, Boice and colleagues135 studied the risk of second malignancies as a consequence

42

PART 1  Principles

of ­radiation treatment for carcinoma of the uterine cervix. This huge international study, involving 42 investigators from 38 institutions, had a sample size of 150,000 patients and had an ideal control group for comparison, because cervical cancer is equally well treated by radiation or surgery. Some of the pertinent findings from this large study are summarized here. Very high doses (on the order of several hundred Gy) were found to increase the risk of developing cancer of the bladder, vagina, bone, uterine corpus, and cecum, as well as non-Hodgkin lymphoma. Risk ratios ranged from a high of 4.0 for bladder cancer to a low of 1.3 for bone cancer. A steep dose-response curve was observed for the risk of developing subsequent genital cancer, with a fivefold excess at doses of more than  150 Gy. Doses of several Gy were found to increase the risk of stomach cancer and leukemia. Somewhat surprisingly, radiation therapy for cervical cancer did not increase the overall risk of subsequent cancer of the small intestine, colon, ovary, vulva, connective tissue, or breast, nor did it increase the risk of Hodgkin disease, multiple myeloma, or chronic lymphocytic leukemia. The overall conclusions of this study were that radiation therapy, unlike surgery, was certainly associated with a risk of excess cancer and that the risks were highest among long-term survivors and concentrated among women who were irradiated at relatively young ages. Second Malignancies after Treatment for Hodgkin Lymphoma. In a 1996 report of the Late Effects Study Group, which followed 1380 children with Hodgkin ­disease, 17 of 483 girls in whom Hodgkin disease was diagnosed before the age of 16 years subsequently developed breast cancer, with radiation therapy implicated in most cases.136 The ratio of observed to expected cases in that study was 75.3. Another study involving 1641 patients treated for Hodgkin disease as children in five Nordic countries reported 16 cases of breast cancer, for a relative risk 17 times that of the general population.137 The largest study of this kind evaluated 3869 women in population-based registries participating in the SEER Program.138 All of these women had been given radiation therapy as initial treatment for Hodgkin disease. Breast cancer developed in a total of 55 patients, which represents a ratio of observed to expected cases of 2.24. However, the risk of subsequent breast cancer in women treated before the age of 16 years was 61%, with most tumors appearing 10 or more years after treatment. This finding agrees with those from previous studies showing the prepubertal and pubertal female breast to be very radiosensitive. The risk of breast cancer decreased as the age at therapy increased and was only slightly elevated in women who were 30 years old or older when they were treated. In one study,139 27 of 202 patients with Hodgkin disease treated with radiation developed a second cancer. Most of the second tumors appeared within or near the treatment field, most commonly in the lung and breast; however, the spectrum was large, and the risk was greatest among patients irradiated at young ages. In summary, studies of large numbers of patients clearly indicate that radiation therapy induces an excess of second cancers. In the young, the breast is especially sensitive to the carcinogenic effects of radiation. The latency period for the development of excess cancers can range from 10 years to decades after the radiation treatment.

REFERENCES   1. Bishop JM. Cellular oncogene retroviruses. Annu Rev Biochem 1983;52:301-354.   2. Kundson AG. Mutation and cancer: statistical study of ­retinoblastoma. Proc Natl Acad Sci U S A 1971;68:820-823.   3. Knudson A. Anti-oncogenes and human cancer. Proc Natl Acad Sci U S A 1993;90:10914-10921.   4. Fearon ER, Vogelstein B. A genetic model for colorectal ­tumorigenesis. Cell 1990;61:759-767.   5. Nagasawa H, Little JB. Induction of sister chromatid ­exchanges by extremely low doses of alpha-particles. Cancer Res 1992;52:6394-6396.   6. Zhou H, Randers-Pehrson G, Waldren CA, et al. Induction of a bystander mutagenic effect of alpha particles in mammalian cells. Proc Natl Acad Sci U S A 2000;97:2099-2104.   7. Zhou H, Suzuki M, Randers-Pehrson G, et al. Radiation risk to low fluences of alpha particles may be greater than we thought. Proc Natl Acad Sci U S A 2001;98:14410-14415.   8. Mothersill C, Seymour C. Radiation-induced bystander ­effects: past history and future directions. Radiat Res 2001;155:  759-767.   9. Ward JF. DNA lesions produced by ionizing radiation: locally multiplying damaged sites. In Wallace SS, Painter RB, (eds): I­onizing Radiation Damage to DNA: Molecular Aspects. New York: Wiley-Liss, 1990. 10. Puck TT, Markus PI. Action of x-rays on mammalian cells. J Exp Med 1956;103:656-666. 11. Withers HR. The dose-survival relationship for irradiation of epithelial cells of mouse skin. Br J Radiol 1967;40:187-194. 12. Withers HR. Recovery and repopulation in vivo by mouse skin epithelium cells during fractionated irradiation. Radiat Res 1967;32:227-239. 13. Withers HR, Elkind MM. Microcolony survival assay for cells of mouse intestinal mucosa exposed to radiation. Int J Radiat Biol 1970;17:261-267. 14. McCullough EA, Till JE. The sensitivity of cells from normal mouse bone marrow to gamma radiation in vitro and in vivo. Radiat Res 1962;16:822-832. 15. Douglas BG, Fowler JF: The effect of multiple small doses of x-rays on skin reactions in the mouse and a basic interpretation. Radiat Res 1976;66:401-426. 16. Suit H, Wette R. Radiation dose fractionation and tumor control probability. Radiat Res 1966;29:267-281. 17. Hewitt HB, Wilson CW. A survival curve for mammalian ­leukaemia cells irradiated in vivo. Br J Cancer 1959;13:69-75. 18. Hill RP, Bush RS. A lung colony assay to determine the radiosensitivity of the cells of a solid tumor. Int J Radiat Biol 1969;15:435-444. 19. Howard A, Pelc SR. The synthesis of deoxyribose nucleic acid and its relationship to the chromosome damage induced by  x-rays. Heredity 1953;6;(Suppl):261-273. 20. Brown JM. The effect of acute x-irradiation on the cell proliferation kinetics of induced carcinomas and their normal counterpart. Radiat Res 1970;43:627-653. 21. Steel GG. Cell loss as a factor in the growth rate of human ­tumors. Eur J Cancer 1967;3:381-387. 22. Steel GG. Cell loss from experimental tumors. Cell Tissue Kinet 1968;1:193-207. 23. Steel GG. The kinetics of cell proliferation in tumors. In Bond V (ed): Proceedings of the Carmel Conference on Time and Dose Relationships in Radiation Biology as Applied to Radiotherapy. BNL Report 50203 (C-57). Upton, NY: Brookhaven National Laboratory, 1969, pp 130–149. 24. Begg AC, McNally NJ, Shrieve D, et al. A method to measure the duration of DNA synthesis and the potential doubling time from a single sample. Cytometry 1985;6:620-625. 25. Begg AC, Moonen I, Hofland I, et al. Human tumor cell kinetics using a monoclonal antibody against iododeoxyuridine: intratumoral sampling variations. Radiother Oncol 1988;11:337-347. 26. Gray JW, Dolbeare F, Pallavicini MG, et al. Cell cycle analysis using flow cytometry. Int J Radiat Biol 1986;49:237-255.

1  Physical and Biologic Basis of Radiation Therapy 27. Regaud C, Nogier T. Sterilization roentgenienne totale et definitive, sans radiodermite, des testicules du Belier adulte: corditions de sa realisation. Compt Rend Soc Biol 1911;70:202-203. 28. Brenner DJ, Hall E. Conditions for the equivalence of continuous to pulsed low dose rate brachytherapy. Int J Radiat Oncol Biol Phys 1991;20:181-190. 29. Thames HD, Ang KK, Stewart FA, et al. Does incomplete repair explain the apparent failure of the basic LQ model to predict spinal cord and kidney responses to low doses per fraction. Int J Radiat Biol 1988;554:13-19. 30. Withers HR, Peters LJ, Kogelnik HC. The pathobiology of late effects of irradiation. In Meyn RE, Withers HR, (eds): Radiation Biology in Cancer Research. New York: Raven Press, 1980, pp 439-448. 31. Fowler JF. The second Klaas Breur memorial lecture: La Ronde—radiation sciences and medical radiology. Radiother Oncol 1983;1:1-22. 32. Ellis F. Dose time and fractionation: a clinical hypothesis. Clin Radiol 1969;20:1-7. 33. Arriagada R, LeChevalier T, Quoix E, et al. Effect of chemotherapy on locally advanced non–small cell lung carcinoma: a randomized study of 3,553 patients. Int J Radiat Oncol Biol Phys 1991;20:1183-1190. 34. Cox JD, Stoffel TJ. The significance of needle biopsy after irradiation for stage C adenocarcinoma of the prostate. Cancer 1977;40:156-160. 35. Dugan TC, Shipley WU, Young RH, et al. Biopsy after external beam radiation therapy for adenocarcinoma of the prostate: correlation with original histological grade and current prostate specific antigen levels. J Urol 1991;146:1313-1316. 36. Cox JD. Large dose fractionation (hypofractionation). Cancer 1985;55;(Suppl):2105-2111. 37. Flickinger JC, Kondziolka D, Lunsford LD, et al. A multi­institutional experience with stereotactic radiosurgery for  solitary brain metastasis. Int J Radiat Oncol Biol Phys 1994;28:797-802. 38. Onishi H, Arak, T, Shirato H, et al. Stereotactic hypofractionated high-dose irradiation for stage I nonsmall cell lung carcinoma: clinical outcomes in 245 subjects in a Japanese ­multiinstitutional study. Cancer 2004;101:1623-1631. 39. Cox JD, Pajak TF, Marcial VA, et al. Interfractional interval is a major determinant of late effects, with ­hyperfractionated radiation therapy of carcinomas of upper respiratory and digestive tracts: results from Radiation Therapy Oncology Group protocol 8313. Int J Radiat Oncol Biol Phys 1991;20:1191-1195. 40. Horiot J-C., Le Fur R, N’Guyen T, et al. Hyperfractionation versus conventional fractionation in oropharyngeal carcinoma: final analysis of a randomized trial of the EORTC cooperative group of radiotherapy. Radiother Oncol 1992;25:231-241. 41. Horiot J-C., Bontemps P, van den Bogaert W, et al. Accelerated fractionation (AF) compared to conventional fractionation (CV) improves loco-regional control in the radiotherapy of advanced head and neck cancers: results of the EORTC 22851 randomised trial. Radiother Oncol 1997;44:111-121. 42. Dische S, Saunders M, Barrett A, et al. A randomized multicentre trial of CHART versus conventional radiotherapy in head and neck cancer. Radiother Oncol 1997;44:123-136. 43. Fu KK, Pajak TF, Trotti A, et al. A Radiation Therapy ­Oncology Group (RTOG) phase III randomized study to compare ­hyperfractionation and two variants of accelerated ­fractionation to standard fractionation radiotherapy for head and neck ­squamous cell carcinomas: first report of RTOG 9003. Int J Radiat Oncol Biol Phys 2000;48:7-16. 44. Fowler JF, Harari P. Confirmation of improved local-regional control with altered fractionation in head and neck cancer. Int J Radiat Oncol Biol Phys 2000;48:3-6. 45. Withers HR, Peters LJ. Transmutability of dose and time. ­Commentary on the first report of RTOG 9003. Int J Radiat Oncol Biol Phys 2000;48:1-2. 45a. Kaanders JH, Pop LA, Maores HA, et al. ARCON: experience with 215 patients with advanced head-and-neck cancer. Int J Radiat Oncol Biol Phys 2002;52:769-778.

43

46. Overgaard J, Sand Hansen H, Overgaard M, et al. Importance of overall treatment time for the outcome of radiotherapy in head and neck carcinoma. Experience from the Danish Head and Neck Cancer Study, 6th International Meeting on Progress in Radio-Oncology, Salzburg, Austria, 13-17 May 1998. Bologna, Italy: Monduzzi Editore, 1998, pp 743-752. 47. Hermens AF, Barendsen GW. Changes of cell proliferation characteristics in a rat rhabdomyosarcoma before and after ­irradiation. Eur J Cancer 1969;5:173-189. 48. Withers HR, Peters LJ, Thames HD, et al. Hyperfractionation. Int J Radiat Oncol Biol Phys 1982;8:1807-1809. 49. Withers HR, Mason K, Reid BO, et al. Response of mouse ­intestine to neutrons and gamma rays in relation to dose fractionation and division cycle. Cancer 1974;34:3947. 50. Withers HR, Taylor MMG, Maciejewski B, et al. The hazard of accelerated tumor clonogen repopulation during radiotherapy. Acta Oncol 1988;27:131-146. 51. Chapman JD, Peters LJ, Withers HR. Prediction of Tumor ­Treatment Response. New York: Pergamon Press, 1989. 52. Geara FB, Peters LJ, Ang KK, et al. Prospective comparison of in vitro normal cell radiosensitivity and normal tissue reactions in radiotherapy patients. Int J Radiat Oncol Biol Phys 1993;27:1171-1179. 53. West CM. Intrinsic radiosensitivity as a predictor of patient ­response to radiotherapy. Br J Radiol 1995;68:827-837. 54. West CM, Davidson SE, Burt PA, et al. The intrinsic radiosensitivity of cervical carcinoma: correlations with clinical data. Int J Radiat Oncol Biol Phys 1995;31:841-846. 55. West CML, Davidson SE, Hendry JH, et al. Prediction of cervical carcinoma response to radiotherapy. Lancet 1991;338:818. 56. West CM, Davidson SE, Roberts SA, et al. Intrinsic radiosensitivity and prediction of patient response to radiotherapy for carcinoma of the cervix. Br J Cancer 1993;68:819-823. 57. Parliament MB, Chapman JD, Urtasun RC, et al. Noninvasive assessment of human tumour hypoxia with 123I-iodoazomycin arabinoside: preliminary report of a clinical study. Br J Cancer 1992;65:90-95. 58. Höckel M, Knoop C, Schlenger K, et al. Intratumoral pO2 ­predicts survival in advanced cancer of the uterine cervix. ­Radiother Oncol 1993;26:45-50. 59. Höckel M, Schlenger K, Aral B, et al. Association between tumor hypoxia and malignant progression in advanced cancer of the uterine cervix. Cancer Res 1996;56:4509-4515. 60. Begg AC, Hofland I, Moonen L, et al. The predictive value of cell kinetic measurements in a European trial of accelerated ­fractionation in advanced head and neck tumors: an interim ­report. Int J Radiat Oncol Biol Phys 1990;19:1449-1453. 61. Begg AC, Haustermans K, Hart AAM, et al. The value of ­pretreatment cell kinetic parameters as predictors for radiotherapy outcome in head and neck cancer: a multicenter analysis. Radiother Oncol 1999;50:13-23. 62. Thomlinson RH, Gray LH. The histological structure of some ­ human lung cancers and the possible implications for ­radiotherapy. Br J Cancer 1955;9:539-549. 63. Brown JM. Evidence for acutely hypoxic cells in mouse ­tumours, and a possible mechanism of reoxygenation. Br J ­ Radiol 1979;52:650-656. 64. Sobhanifar S, Aquino-Parsons C, Stanbridge EJ, et al. Reduced expression of hypoxia-inducible factor-1alpha in perinecrotic regions of solid tumors. Cancer Res 2005;65:7259-7266. 65. Kwon SJ, Lee YJ. Effect of low glutamine/glucose on hypoxiainduced elevation of hypoxia-inducible factor-1alpha in ­human pancreatic cancer MiaPaCa-2 and human prostatic cancer  DU-145 cells. Clin Cancer Res 2005;11:4694-4700. 66. Brizel DM, Scully SP, Harrelson JM, et al. Tumor oxygenation predicts for the likelihood of distant metastases in human soft tissue sarcoma. Cancer Res 1996;56:941-943. 67. Giaccia AJ, Brown JM, Wouters B, et al. Cancer therapy and tumor physiology. Science 1998;279:12-13. 68. Adams GE. Chemical radiosensitization of hypoxic cells. Br Med Bull 1973;29:48-53.

44

PART 1  Principles

69. Adams GE. Hypoxic cell sensitizers for radiotherapy. In Becker FF (ed): Cancer, vol 6. New York: Plenum Press, 1977,  pp 181-223. 70. Adams GE, Bremner J, Stratford IJ, et al. Nitroheterocyclic compounds as radiosensitizers and bioreductive drugs. ­ Radiother Oncol 1991;20:85-91. 71. Brown JM, Siim BG. Hypoxia-specific cytotoxins in cancer ­therapy. Semin Radiat Oncol 1996;6:22-36. 72. Brown JM, Lemmon MJ. Tumor hypoxia can be exploited to preferentially sensitize tumors to fractionated irradiation. Int J Radiat Oncol Biol Phys 1991;20:457-461. 73. von Pawel J, von Roemeling R, Gatzemeier U, et al. Tirapazamine plus cisplatin in advanced non–small cell lung cancer: a report of the international CATAPULT I study group. Cisplatin and ­ Tirapazamine in Subjects with Advanced Previously Untreated Non–Small Cell Lung Tumors. J Clin Oncol 2000;18:  1351-1359. 74. Rischin D, Peters L, Hicks R, et al. Phase I trial of concurrent tirapazamine, cisplatin, and radiotherapy in patients with ­advanced head and neck cancer. J Clin Oncol 2001;19:535-542. 75. Rischin D, Peters L, Fisher R, et al. Tirapazamine, cisplatin, and radiation versus fluorouracil, cisplatin, and radiation in patients with locally advanced head and neck cancer: a randomized phase II trial of the Trans-Tasman Radiation Oncology Group (TROG 98.02). J Clin Oncol 2005;23:79-87. 76. Overgaard J, Horsman MR. Modification of hypoxia-induced radioresistance in tumours by the use of oxygen and sensitizers. Semin Radiat Oncol 1996;6:10-21. 77. Patt HM, Tyree EB, Straube RL, et al. Cysteine protection against x-irradiation. Science 1949;110:213-214. 78. Bacq ZM, Herve A, Lecompte J, et al. Protection contre le rayonnement x par la β-mercato-éthylamine. Arch Intern Physiol 1951;S9:442-447. 79. Yuhas JM, Spellman JM, Culo F. The role of WR2721 in ­radiotherapy and/or chemotherapy. In Brady L (ed): Radiation Sensitizers. New York: Masson Publishing, 1980. 80. Kligerman MM, Liu T, Liu Y, et al. Interim analysis of a ­randomized trial of radiation therapy of rectal cancer with/without WR-2721. Int J Radiat Oncol Biol Phys 1992;22:799-802. 81. Brizel D, Sander R, Wannenmacher M, et al. Randomized phase III trial of radiation ± amifostine in patients with head and neck cancer [abstract]. Proc Am Soc Clin Oncol 1998;17:1487. 82. Komaki R, Lee JS, Milas L, et al. Effects of amifostine on acute toxicity from concurrent chemotherapy and radiotherapy for inoperable non–small cell lung cancer: report of a randomized comparative trial. Int J Radiat Oncol Biol Phys 2004;58:  1369-1377. 83. Barendsen GW. Radiobiological dose-effect relations for radiation characterized by a wide spectrum of LET: implications for their application in radiotherapy. In Proceedings of the Conference of ­Particle Accelerators in Radiation Therapy, Los Alamos Science Laboratory, Los Alamos, NM, October 2-5, 1972. LA5180-C, 120-125. Oak Ridge, TN: U.S. Atomic Energy Commission, Technical Information Center, 1972. 84. Catterall M, Vonberg DD. Treatment of advanced tumors of head and neck with fast neutrons. Br Med J 1974;3:137-143. 85. Catterall M, Sutherland I, Bewley DK. First results of a ­clinical trial of fast neutrons compared with x or gamma rays in ­treatment of advanced tumors of the head and neck. Br Med J 1975;2:653-656. 86. Catterall M. Results of neutron therapy: differences, correlations, and improvements. Int J Radiat Oncol Biol Phys 1982;8:  2141-2144. 87. Batterman JJ. Clinical Application of Fast Neutrons: The ­Amsterdam Experience. Amsterdam: Rodipi, 1981. 88. Laramore GE, Krall JM, Thomas FJ, et al. Fast neutron radiotherapy for locally advanced prostate cancer: results of an RTOG randomized study. Int J Radiat Oncol Biol Phys 1985;11:1621-1627. 89. Weber DC, Trofimov AV, Delaney TF, et al. A treatment planning comparison of intensity modulated photon and proton therapy for paraspinal sarcomas. Int J Radiat Oncol Biol Phys 2004;58:1596-1606.

90. Sisterson J. Proposed new facilities for proton and ion beam therapy, July 2005. Particles Newslett 2005;36:1-11. Available at http://ptcog.web.psi.ch/Archive/ptles36.pdf (accessed ­January 2009). 91. Kraft G. Tumor therapy with heavy charged particles. Progress in Particle and Nuclear Physics 2000;45;(Suppl 2):S473-S544. 92. Coley WB. The treatment of malignant tumors by repeated ­inoculation of erysipelas: with a report of ten original cases. Am J Med Sci 1893;105:487-511. 93. Overgaard J, Gonzalez Gonzalez D, Hulschof MC, et al. Hyperthermia as an adjuvant to radiation therapy of recurrent or metastatic malignant melanoma. A multicentre randomized trial by the European Society of Hyperthermic Oncology. Int J Hyperthermia 1996;12:3-20. 94. Vernon CC, Hand JW, Field SB, et al. Radiotherapy with or without hyperthermia in the treatment of superficial localized breast cancer: results from five randomized controlled trials. International Collaborative Hyperthermia Group. Int J Radiat Oncol Biol Phys 1996;35:731-744. 95. Kong G, Anyarambhatla G, Petros WP, et al. Efficacy of liposomes and hyperthermia in a human tumor xenograft model: importance of triggered drug release. Cancer Res 2000;60:6950-6957. 96. Kong G, Braun RD, Dewhirst MW. Characterization of the effect of hyperthermia on nanoparticle extravasation from tumor vasculature. Cancer Res 2001;61:3027-3032. 97. Steel GG. Terminology in the description of drug-radiation interactions. Int J Radiat Oncol Biol Phys 1979;5:1145-1150. 98. Tubiana M. Les associations radiotherapie-chimeotherapie. J Eur Radiother 1980;1:107-114. 99. Schaake-Koning C, vanden Bogaert W, Dalesio O, et al. Effects of concomitant cisplatin and radiotherapy on inoperable nonsmall cell lung cancer. N Engl J Med 1992;326:524-530. 100. Krook JE, Moertel C, Gunderson L, et al. Effective surgical adjuvant therapy for high-risk rectal carcinoma. N Engl J Med 1991;324:709-715. 101. Le Chevalier T, Arriagada R, Tarayre M, et al. Significant effect of adjuvant chemotherapy on survival in locally advanced ­nonsmall cell lung carcinoma. J Natl Cancer Inst 1992;84:58. 102. Herskovic A, Martz K, Al-Sarraf M, et al. Combined ­chemotherapy and radiotherapy compared with radiotherapy alone in patients with cancer of the esophagus. N Engl J Med 1992;326:1593-1598. 103. Milas L, Mason K, Hunter N, et al. In vivo enhancement of ­tumor radioresponse by C225 antiepidermal growth factor ­receptor antibody. Clin Cancer Res 2000;6:701-708. 104. Bonner JA, Giralt J, Harari PM, et al. Cetuximab prolongs ­survival in patients with locoregionally advanced squamous cell carcinoma of head and neck: a phase III study of high dose radiation therapy with or without cetuximab [abstract]. J Clin Oncol 2004;22;(14S):5507. 105. Hall SJ. Chen S-H, Woo SLC. Gene therapy 97: the promise and reality of cancer gene therapy. Am J Hum Genet 1997;61:785-789. 106. Kim JH, Kim SH, Brown SL, et al. Selective enhancement by an antiviral agent of the radiation-induced cell killing of human glioma cells transduced with HSV-tk gene. Cancer Res 1994;54:6053-6056. 107. Khil MS, Kim JH, Mullen CA, et al. Radiosensitization by 5-fluorocytosine of human colorectal carcinoma cells in culture transduced with cytosine deaminase gene. Clin Cancer Res 1996;2:53-57. 108. Freytag SO, Stricker H, Pegg J, et al. Phase I study of ­replicationcompetent adenovirus-mediated double-suicide gene therapy in combination with conventional-dose three-dimensional conformal radiation therapy for the treatment of newly diagnosed, intermediate- to high-risk prostate cancer. Cancer Res 2003;63:7497-7506. 109. Freytag SO, Stricker H, Peabody J, et al. Five-year follow-up of trial of replication-competent adenovirus-mediated suicide gene therapy for treatment of prostate cancer. Mol Ther 2007;15:636-642. 110. Ganly I, Kirn D, Eckhardt G, et al. A phase I study of Onyx-015, an E1B attenuated adenovirus, administered intratumorally to patients with recurrent head and neck cancer. Clin Cancer Res 2000;6:798-806.

1  Physical and Biologic Basis of Radiation Therapy 111. Nemunaitis J, Khuri F, Ganly I, et al. Phase II trial of intratumoral administration of ONYX-015, a replication-selective ­adenovirus, in patients with refractory head and neck cancer.  J Clin Oncol 2001;19:289-298. 112. Vasey PA, Shulman LN, Campos S, et al. Phase I trial of intraperitoneal injection of the E1B-55-kd-gene-deleted adenovirus ONYX-015 (dl1520) given on days 1 through 5 every 3 weeks in patients with recurrent/refractory epithelial ovarian cancer.  J Clin Oncol 2002;20:1562-1569. 113. Gallardo D, Drazan KE, McBride WH. Adenovirus-based transfer of wild-type p53 gene increases ovarian tumor radiosensitivity. Cancer Res 1996;56:4891-4893. 114. Spitz FR, Nguyen D, Skibber JM, et al. Adenoviral-mediated wild-type p53 gene expression sensitizes colorectal cancer cells to ionizing radiation. Clin Cancer Res 1996;2:1665-1671. 115. Swisher SG, Roth JA, Nemunaitis J, et al. Adenovirus-mediated p53 gene transfer in advanced non–small cell lung cancer. J Natl Cancer Inst 1999;91:763-771. 116. Clayman GL, el-Naggar AK, Lippman SM, et al. Adenovirusmediated p53 gene transfer in patients with advanced recurrent head and neck squamous cell carcinoma. J Clin Oncol 1998;16:2221-2232. 117. Weill D, Mack M, Roth J, et al. Adenoviral-mediated p53 gene transfer to non–small cell lung cancer through endobronchial injection. Chest 2000;118:966-970. 118. Nemunaitis J, Swisher SG, Timmons T, et al. Adenovirus­mediated p53 gene transfer in sequence with cisplatin to ­tumors of patients with non–small cell lung cancer. J Clin Oncol 2000;18:609-622. 119. Modesitt SC, Ramirez P, Zu Z, et al. In vitro and in vivo ­adenovirus-mediated p53 and p16 tumor suppressor therapy in ­ovarian cancer. Clin Cancer Res 2001;7:1765-1772. 120. Lang FF, Bruner JM, Fuller GN, et al. Phase I trial of adenovirusmediated p53 gene therapy for recurrent glioma: biological and clinical results. J Clin Oncol 2003;21:2508-2518. 121. Pagliaro LC, Keyhani A, Williams D, et al. Repeated intravesical instillations of an adenoviral vector in patients with locally advanced bladder cancer: a phase I study of p53 gene therapy. J Clin Oncol 2003;21:2247-2253. 122. Clayman GL, Frank DK, Bruso PA, et al. Adenovirus-mediated wild-type p53 gene transfer as a surgical adjuvant in advanced head and neck cancers. Clin Cancer Res 1999;5:1715-1722. 123. Weichselbaum RR, Hallahan DE, Beckett MA, et al. Gene therapy targeted by radiation preferentially radiosensitizes tumor cells. Cancer Res 1994;54:4266-4269. 124. Hallahan DE, Mauceri HJ, Seung LP, et al. Spatial and ­temporal control of gene therapy using ionizing radiation. Nat Med 1995;1:786-791. 125. Staba MJ, Mauceri HJ, Kufe DW, et al. Adenoviral TNF-­alpha gene therapy and radiation damage tumor vasculature in a human malignant glioma xenograft. Gene Ther 1998;5:293-300. 126. Müller HJ. Artificial transmutation of the gene. Science 1927;66:84-87. 127. Russell WL. The effect of radiation dose rate and fractionation on mutation in mice. In Sobels FH (ed): Repair of ­Genetic Radiation Damage. New York: Pergamon Press, 1963, pp 205-217. 128. Russell WL. Effect of the interval between irradiation and ­conception on mutation frequency in female mice. Proc Natl Acad Sci U S A 1965;54:1552-1557. 129. United Nations Scientific Committee on the Effects of Atomic Radiation. Hereditary Effects of Radiation: The UNSCEAR 2001 Report to the General Assembly with Scientific Annex. New York: United Nations, 2001. 130. Neel JV, Schull WJ, Awa AA, et al. The children of parents ­exposed to atomic bombs: estimates of genetic doubling dose of radiation for humans. Am J Hum Genet 1990;46:1053-1072. 131. Committee on the Biological Effects of Ionizing Radiation. BEIR V: Health Effects of Exposure to Low Levels of Ionizing ­Radiation. Washington, DC: National Academy Press, 1990. 132. United Nations Scientific Committee on the Effects of Atomic Radiation. Sources and Effects of Ionizing Radiation. New York: United Nations, 1988.

45

133. International Commission on Radiological Protection. ICRP 60: 1990 Recommendations of the International Commission on ­ Radiological Protection. Oxford, England: Pergamon Press, 1990. 134. Brenner DJ, Curtis RE, Hall EJ, et al. Second cancers after ­radiotherapy for prostate cancer. Cancer 2000;15:398-406. 135. Boice JD Jr, Engholm G, Kleinman RA, et al. Radiation dose and second cancer risk in patients treated for cancer of the ­cervix. Radiat Res 1988;116:3-55. 136. Bhatia S, Robison LL, Oberlin O, et al. Breast cancer and other second neoplasms after childhood Hodgkin’s disease. N Engl J Med 1996;334:745-751. 137. Sankila R, Garwicz S, Olsen JH, et al. Risk of subsequent malig­nant neoplasms among 1,641 Hodgkin’s disease patients ­diagnosed in childhood and adolescence: a population-based  cohort study in the five Nordic countries. J Clin Oncol 1996;14:  1442-1446. 138. Travis LB, Curtis RE, Boice JD Jr, et al. Late effects of treatment for childhood Hodgkin’s disease. N Engl J Med 1996;335:  352-353. 139. Nyandoto P, Muhonen T, Joensuu H. Second cancers among long-term survivors from Hodgkin’s disease. Int J Radiat Oncol Biol Phys 1998;42:373-378.

ADDITIONAL READING Historical Background del Regato JA.Radiological Physicists. New York: American Institute of Physics, 1985. Hall EJ. Radiation and Life. New York: Pergamon Press, 1984. Hall EJ, Giaccia AJ. Radiobiology for the Radiologist. 6th ed. ­Philadelphia: Lippincott Williams & Wilkins, 2005.

Types of Radiation Goodwin PN, Quimby EH, Morgan RH. Physical Foundations of Radiology. New York: Harper & Row, 1970. Johns H, Cunningham J (eds): The Physics of Radiology, 4th ed. Springfield, IL: Charles C Thomas, 1983. Smith VP (ed): Radiation Particle Therapy. Philadelphia: American College of Radiology, 1976.

Production of Radiation for Therapeutic Applications  Hendee W (ed) Radiation Therapy Physics. St Louis: Mosby; 1981. Johns H, Cunningham J (eds). The Physics of Radiology, 4th ed. Springfield, IL: Charles C Thomas, 1983.

Dose: Its Measurement and Meaning Rossi HH. Neutron and heavy particle dosimetry. In Reed GW (ed): Radiation Dosimetry. Proceedings of the International School of Physics. New York: Academic Press, 1964.

Physical Basis for Choice of Treatment Techniques Khan F (ed): The Physics of Radiation Therapy. Baltimore: ­Williams & Wilkins, 1984. Levitt SH, Khan FM, Potish RA. Levitt and Tapley’s Technological Basis of Radiation Therapy: Practical Clinical Applications, 2nd ed. Philadelphia: Lea & Febiger, 1992.

Signal Transduction Amundson SA, Myers TG, Fornace AJ. Roles of p53 in growth arrest and apoptosis: putting on the brakes after genotoxic stress. Oncogene 1998;17:3287-3299.

The Cell Cycle Hartwell LH, Weinert TA. Checkpoints: controls that ensure the order of cell cycle events. Science 1989;246:629-634.

46

PART 1  Principles

Kastan MB, Zhan Q, El-Deiry WS, et al. A mammalian cell cycle checkpoint pathway utilizing p53 and GADD45 is defective in ataxia-telangiectasia. Cell 1992;71:587-597. McKenna W, Iliakis G, Weiss MC, et al. Increased G2 delay in ­radiation-resistant cells obtained by transformation of primary rat embryo cells with the oncogenes H-ras and v-myc. Radiat Res 1991;125:283-287.

Oncogenes and Tumor Suppressor Genes Bishop JM, Varmus HE. Functions and origins of retroviral transforming genes. In Weiss R, Teich N, Varmus H, Coffinm J (eds): RNA Tumor Viruses: Molecular Biology of Tumor Viruses, 2nd ed. New York: Cold Spring Harbor Laboratory, 1984, pp 990-1108. Cavanee WK. Tumor progression stage: specific losses of heterozygosity. Int Symp Princess Takamatsu Cancer Res Found 1989;20:33-42. Cole M. The myc oncogene: its role in transformation and differentiation. Annu Rev Genet 1986;20:361-384. Hartwell LH, Kastan MB. Cell cycle control and cancer. Science 1994;266:1821-1828. Rous P. A sarcoma of fowl transmissible by an agent separable from the tumor cells. J Exp Med 1911;13:397. Stanbridge EJ. Suppression of malignancy in human cells. Nature 1976;260:17-20.

Radiation Chemistry Boag JW. The action of ionizing radiation on dilute aqueous ­solutions. Phys Med Biol 1965;10:457-476. Ward JF. Biochemistry of DNA lesions. Radiat Res 1985;104:  S103-S111.

Cell Survival Curves Elkind MM, Sutton H. Radiation response of mammalian cells grown in culture. I. Repair of x-ray damage in surviving Chinese hamster cells. Radiat Res 1960;13:556-593. Puck TT, Markus PI. Action of x-rays on mammalian cells. J Exp Med 1956;103:653-666.

Effects on Normal Tissues Cox JD, Byhardt RW, Wilson JF, et al. Complications of radiation therapy and factors in their prevention. World J Surg 1986;10:  171-188. Denekamp J. Cell kinetics and radiation biology. Int J Radiat Biol 1986;49:357-380. Fowler JF, Kragt K, Ellis RE, et al. The effect of divided doses of 15 MeV electrons in the skin response of mice. Int J Radiat Biol 1965;9:241-252. Fowler JF, Morgan RL, Silvester JA, et al. Experiments with fractionated x-ray treatment of the skin of pigs. I. Fractionation up to 28 days. Br J Radiol 1963;36:188-196. Hendry JH, Thames HD. The tissue-rescuing unit. Br J Radiol 1986;59:628-630. Hopewell JW. Late radiation damage to the central nervous system: a radiobiological interpretation. Neuropathol Appl Neurobiol 1979;5:329-343. Hopewell JW. Mechanisms of the action of radiation on skin and underlying tissues. Br J Radiol 1986;19 (Suppl):39-51. Hopewell JW, Campling D, Calvo W, et al. Vascular irradiation damage: its cellular basis and likely consequences. Br J Cancer 1986;53:181-191. Hopewell JW, Morris AD, Dixon-Brown A. The influence of field size on the late tolerance of the rat spinal cord to single doses of  x-rays. Br J Radiol 1987;60:1099-1108. Kerr JRF, Searle J. Apoptosis: its nature and kinetic role. In Meyn RE, Withers HR (eds): Radiation Biology in Cancer Research. New York: Raven Press, 1980, pp 367-384. Michalowsli A. A critical appraisal of clonogenic survival assays in the evaluation of radiation damage to normal tissues. Radiother Oncol 1984;1:241-246. Phillips TL, Margolis L. Radiation pathology and the clinical response of lung and esophagus. Front Radiat Ther Oncol 1972;6:254-273.

Phillips TL, Ross G. A quantitative technique for measuring renal damage after irradiation. Radiology 1973;109:457-462. Phillips TL, Ross G. Time-dose relationships in the mouse esophagus. Radiology 1974;113:435-440. Rubin P. Late effects of chemotherapy and radiation therapy: a new hypothesis. Int J Radiat Oncol Biol Phys 1984;10:5-34. Rubin P, Casarett GW. Clinical Radiation Pathology, vol 1. ­Philadelphia: WB Saunders, 1968. Stewart FA. Mechanism of bladder damage and repair after treatment with radiation and cytostatic drugs. Br J Cancer 1986;7, (Suppl):280-291. Travis EL. Liao Z-X, Tucker SL. Spatial heterogeneity of the volume effect for radiation pneumonitis in mouse lung. Int J Radiat Oncol Biol Phys 1997;38:1045-1054. Travis EL, Tucker SL. Relationship between functional assays of radiation response in the lung and target cell depletion. Br J Cancer Suppl 1986;7:304-319. van der Kogel AJ. Mechanisms of late radiation injury in the spinal cord. In Meyn RE, Withers HR (eds): Radiation Biology in Cancer Research. New York: Raven Press, 1980, pp 461-470. van der Kogel AJ. Central nervous system radiation injury in small animal models. In Gutin PH, Leibel SA, Sheline GE (eds): Radiation Injury to the Nervous System. New York: Raven Press, 1991,  pp 91-112. van der Schueren E, Landuyt W, Ang KK, et al. From 2 Gy to 1 Gy per fraction: sparing effect in rat spinal cord? Int J Radiat Oncol Biol Phys 1988;14:297-300. Withers HR, Taylor JMG, Maciejewski B. Treatment volume and ­tissue tolerance. Int J Radiat Oncol Biol Phys 1988;14:751-759. Withers HR, Thames HD, Peters LJ, et al. Keynote address: Normal tissue radioresistance in clinical radiotherapy. In Fletcher GH, Nervi C, Withers HR (eds): Biological Bases and Clinical Implications of Tumor Radioresistance. New York: Masson, 1983,  pp 139-152.

Effects on Tumors Barendsen GW, Broerse JJ: Experimental radiotherapy of a rat ­rhabdomyosarcoma with 15 MeV neutrons and 300 kV x-rays. I. Effects of single exposures. Eur J Cancer 1969;5:373-391. Hewitt HB. Studies in the quantitative transplantation of mouse sarcoma. Br J Cancer 1953;7:367-383. Hewitt HB, Wilson CW. A survival curve for mammalian cells ­irradiated in vivo. Nature 1959;183:1060-1061. Hewitt HB, Chan DPS, Blake ER. Survival curves for clonogenic cells of a murine keratinizing squamous carcinoma irradiated in vivo or under hypoxic conditions. Int J Radiat Biol 1967;12:535-549. Hill EP, Bush RS. A lung colony assay to determine the radiosensitivity of the cells of a solid tumor. Int J Radiat Biol 1969;15:  435-444. Powers WE, Tolmach LJ. A multicomponent x-ray survival curve for mouse lymphosarcoma cells irradiated in vivo. Nature 1963;197:710-711.

Tissue and Tumor Kinetics Begg AC, McNally NJ, Schrieve DC, et al. A method to measure the duration of DNA synthesis and the potential doubling time from a single sample, Cytometry 1985;6:620-626. Denekamp J. The cellular proliferation kinetics of animal tumors. Cancer Res 1970;30:393-400. Dolbeare F, Beisker W, Pallavicini M, et al. Cytochemistry for BrdUrd/DNA analysis: stoichiometry and sensitivity. Cytometry 1985;6:521-530. Dolbeare F, Gratzner H, Pallavicini M, et al. Flow ­ cytometric ­measurement of total DNA content and incorporated ­bromodeoxyuridine. Proc Natl Acad Sci U S A 1983;80:55735577. Frindel E, Malaise EP, Tubiana M. Cell proliferation kinetics in five human solid tumors. Cancer 1968;22:661-662. Gray JW. Cell-cycle analysis of perturbed cell populations. ­Computer simulation of sequential DNA distributions. Cell Tissue Kinet 1976;9:499-516.

1  Physical and Biologic Basis of Radiation Therapy Mendelsohn ML. The growth fraction: a new concept applied to  tumors. Science 1960;132:1496. Mendelsohn ML. Principles, relative merits, and limitations of current cytokinetic methods. In Drewinko B, Humphrey RM (eds): Growth Kinetics and Biochemical Regulation of Normal and Malignant Cells.. Baltimore: Williams & Wilkins, 1977. Tubiana M, Malaise EP. Growth rate and cell kinetics in human ­tumors: some prognostic and therapeutic implications. In Symington T, Carter RL (eds): Scientific Foundations of Oncology.  St Louis: Mosby, 1976.

Systemic Effects Bond VP, Fliedner TM, Archambeau JO: Mammalian Radiation Lethality: A Disturbance in Cellular Kinetics. New York: Academic Press, 1965. Hemplemann LH, Lisco H, Hoffman JG. The acute radiation ­syndrome: a study of nine cases and a review of the problem. Ann Intern Med 1952;36:279-510. Lushbaugh CC. Reflections on some recent progress in human radiobiology. In Augenstein LG, Mason R, Zelle M (eds): Advances in Radiation Biology. New York: Academic Press, 1969. Shipman TL, Lushbaugh CC, Peterson D, et al. Acute radiation death resulting from an accidental nuclear critical excursion.  J ­Occup Med 1961;3:145-192. Vriesendorp HM, van Bekkum DW. Role of total body irradiation in conditioning for bone marrow transplantation. In Thierfelder S, Rodt H, Kolb HJ (eds): Immunobiology of Bone Marrow Transplantation. Berlin: Springer Verlag, 1980. Vriesendorp HM, van Bekkum DW. Total body irradiation. In Broerse JJ, MacVittie T (eds): Response to Total Body Irradiation in Different Species. Amsterdam: Martinus Nijhoff, 1984.

Fractionation Bedford JS, Hall EJ. Survival of HeLa cells cultured in vitro and ­exposed to protracted gamma irradiation. Int J Radiat Biol 1963;7:377-383. Bedford JS, Mitchell JB. Dose-rate effects in synchronous mammalian cells in culture. Radiat Res 1973;54:316-327. Begg AC, Hofland I, Van Glabekke M, et al. Predictive value of potential doubling time for radiotherapy of head and neck tumor patients: results from the EORTC Cooperative Trial 22851. Semin Radiat ­Oncol 1992;2:22-25. Belli JA, Shelton M. Potentially lethal radiation damage: repair by mammalian cells in culture. Science 1969;165:490-492. Ben-Hur E, Elkind MM, Bronx BV. Thermally enhanced radiosensitivity of cultured Chinese hamster cells: inhibition of repair of sublethal damage and enhancement of lethal damage. Radiat Res 1974;55:38-51. Dolbeare F, Beisker W, Pallavicini M, et al. Cytochemistry for BrdUrd/DNA analysis: stoichiometry and sensitivity. Cytometry 1985;6:521-530. Dolbeare F, Gratzner H, Pallavicini M, et al. Flow cytometric ­measurements of total DNA content and incorporated bromodeoxyuridine. Proc Natl Acad Sci U S A 1983;80:5573-5577. Elkind MM, Sutton H. Radiation response of mammalian cells grown in culture. I. Repair of x-ray damage in surviving Chinese hamster cells. Radiat Res 1960;13:556-593. Fowler JF. Dose response curves for organ function or cell survival. Br J Radiol 1983;56:497-500. Fowler JF. What next in fractionated radiotherapy? Br J Cancer 1984;49:285-300. Fowler JF. Potential for increasing the differential response between tumors and normal tissues: can proliferation rate be used? Int J Radiat Oncol Biol Phys 12:641-645 1986. Fowler JF, Tepper JE. Fractionation in radiation therapy. Semin Radiat Oncol 1992;2:1-72. Hall EJ. Radiation dose-rate: a factor of importance in radiobiology and radiotherapy. Br J Radiol 1972;45:81-97. Hall EJ. The biological basis of endocurietherapy. The Henschke Memorial  Lecture, 1984. Endocuriether Hypertherm Oncol 1985;1:141-151. Hall EJ, Bedford JS. Dose rate: its effect on the survival of HeLa cells irradiated with gamma rays. Radiat Res 1964;22:305-315.

47

Hansen O, Overgaard J, Hansen HS, et al. Importance of overall ­treatment time for the outcome of radiotherapy of advanced head and neck carcinoma: dependency on tumor differentiation. Radiother ­Oncol 1997;43:47-51. Hendry JH, Bentzen SM, Dale RG, et al. A modeled comparison of the effects of using different ways to compensate for missed treatment days in radiotherapy. Clin Oncol 1996;8:297-307. Little JB, Hahn GM, Frindel E, et al. Repair of potentially lethal radiation damage in vitro and in vivo. Radiology 1973;106:689-694. Mitchell JB, Bedford JS, Bailey SM. Dose-rate effects on the cell cycle and survival of S3 HeLa and V79 cells. Radiat Res 1979;79:520-536. Nakatsugawa S, Kumar A, Ono K, et al. Increased tumor curability by radiotherapy combined with PLDR inhibitors in murine ­cancers. In Prospective Methods of Radiation Therapy in Developing Countries, IAEA-TECDOC 266. Vienna: International Atomic Energy Agency, 1982, pp 77-86. Nakatsugawa S, Sugahara T, Kumar A. Purine nucleoside analogues inhibit the repair of radiation induced potentially lethal damage in mammalian cells in culture. Int J Radiat Biol 1982;41:343-346. Overgaard J, Hjelm-Hansen M, Johnsen LV, et al. Comparison of conventional and split-course radiotherapy as primary treatment in carcinoma of the larynx. Acta Oncol 1988;27:147-152. Overgaard J, Sand Hansen H, Sapru W, et al. Conventional radiotherapy as the primary treatment of squamous cell carcinoma of the head and neck: a randomized multicenter study of 5 versus 6 fractions per week—preliminary report from the DAHANCA 6 and 7 trial. ­Radiother Oncol 1996;40:S31. Peters LJ, Withers HR, Thames HD. Radiobiological bases for ­multiple daily fractionation. In Kaercher KH, Kogelnik HD, Reinartz G (eds): Progress in Radio-oncology, vol 2. New York: Raven Press, 1982. Phillips RA, Tolmach LJ. Repair of potentially lethal damage in  x-­irradiated HeLa cells. Radiat Res 1966;29:413-432. Saunders MI, Dische S, Barrett A, et al. Randomised multicentre trials of CHART vs. conventional radiotherapy in head and neck cancer and non–small cell lung cancer: an interim report. Br J Cancer 1996;73:1455-1462. Saunders MI, Dische S, Hong A, et al. Continuous hyperfractionated accelerated radiotherapy in locally advanced carcinoma of the head and neck region. Int J Radiat Oncol Biol Phys 1989;17:  1287-1293. Sinclair WK. Cyclic x-ray responses in mammalian cells in vitro. Radiat Res 1968;33:620-643. Sinclair WK. Radiation survival in synchronous and asynchronous Chinese hamster cells in vitro. In Biophysical Aspects of ­Radiation Quality. Proceedings of the Second IAEA Panel, Vienna, 1967. Vienna: IAEA, 1968. Sinclair WK. Dependence of radiosensitivity upon cell age. In Proceedings of the Carmel Conference on Time and Dose ­Relationships in Radiation Biology as Applied to Radiotherapy, Brookhaven National Laboratory Report 50203 (C-57). Upton, NY: Brookhaven National Laboratory, 1969, pp 97-107. Sinclair WK, Morton RA. X-ray sensitivity during the cell generation cycle of cultured Chinese hamster cells. Radiat Res 1966;29:  450-474. Suit H, Urano M. Repair of sublethal radiation injury in hypoxic cells of a C3H mouse mammary carcinoma. Radiat Res 1969;37:  423-434. Terasima R, Tolmach LJ. X-ray sensitivity and DNA synthesis in synchronous populations of HeLa cells. Science 1963;140:490-492. Tubiana N, Malaise E. Growth rate and cell kinetics in human tumors: some prognostic and therapeutic implications. In Symington T, Carter RL (eds): Scientific Foundations of Oncology. St Louis: Mosby, 1976. Weichselbaum RR, Little JB, Nove J. Response of human osteosarcoma in vitro to irradiation: evidence for unusual cellular repair activity. Int J Radiat Biol 1977;31:295-299. Weichselbaum RR, Schmitt A, Little JB. Cellular repair factors influencing radiocurability of human malignant tumors. Br J Cancer 1982;45:10-16. Withers HR. Isoeffect curves for various proliferative tissues in experimental animals. In Proceedings of Conference on Time-dose ­Relationships in Clinical Radiotherapy. Madison, WI: Madison ­Printing and Publishing, 1975.

48

PART 1  Principles

Withers HR. Response of tissues to multiple small dose fractions. Radiat Res 1977;71:24-33. Withers HR, Peters LJ, Taylor JMG, et al. Late normal tissue sequelae from radiation therapy for carcinoma of the tonsil: patterns of fractionation study of radiobiology. Int J Radiat Oncol Biol Phys 1995;33:563-568.

Modifiers of Radiation Effects Adams GE. Chemical radiosensitization of hypoxic cells. Br Med Bull 1973;29:48-53. Adams GE, Clarke ED, Gray P, et al. Structure-activity relationships in the development of hypoxic cell radiosensitizers. II. Cytotoxicity and therapeutic ratio. Int J Radiat Biol 1979;35:151-160. Bagshaw MA,Doggett RL,Smith KC.Intra-arterial 5-bromodeoxyuridine  and x-ray therapy. AJR Am J Roentgenol 1967;99:889-894. Barendsen GW. Impairment of the proliferative capacity of human cells in culture by alpha particles with differing linear energy transfer. Int J Radiat Biol 1964;8:453-466. Batterman JJ. Clinical Application of Fast Neutrons: the Amsterdam Experience. Amsterdam: Rodipi, 1981. Brown JM, Goffinet DR, Cleaver JE, et al. Preferential r­adiosensitization of mouse sarcoma relative to normal skin by chronic intra-­arterial infusion of halogenated pyrimidine analogs. J Natl Cancer Inst 1971;47:75-89. Brown JM, Lemmon MJ. Potentiation by the hypoxic cytotoxin SR4233 of cell killing produced by fractionated irradiation of mouse tumors. Cancer Res 1990;50:7745-7749. Catterall M. The treatment of advanced cancer by fast neutrons from the Medical Research Council’s cyclotron at Hammersmith Hospital, London. Eur J Cancer 1974;10:343-347. Catterall M. Results of neutron therapy: differences, correlations, and improvements. Int J Radiat Oncol Biol Phys 1982;8:2141-2144. Chapman JD, Whitmore GF. Chemical modifiers of cancer ­treatment. Int J Radiat Oncol Biol Phys 1984;10:1161-1813. Coleman CN. Hypoxic cell radiosensitizers: expectations and progress in drug development. Int J Radiat Oncol Biol Phys 1985;11:  323-329. Deschner EE, Gray LH. Influence of oxygen tension on x-ray-­induced chromosomal damage in Ehrlich ascites tumor cells ­irradiated in vitro and in vivo. Radiat Res 1959;11:115-146. Elkind MM, Swain RW, Alescio T, et al. Oxygen, nitrogen, recovery and radiation therapy. In Shalek R (ed): Cellular Radiation Biology. Baltimore: Williams & Wilkins, 1965. Field SB, Jones T, Thomlinson RH. The relative effect of fast neutrons and x-rays on tumor and normal tissue in the rat. II. Fractionation recovery and reoxygenation. Br J Radiol 1968;41:597-607. Fowler JF, Morgan RL. Pretherapeutic experiments with the fast neutron beam from the Medical Research Council cyclotron. VIII. General review. Br J Radiol 1963;36:115-121. Fowler JF, Morgan RL, Wood CAP. Pretherapeutic experiments with fast neutron beam from the Medical Research Council cyclotron. I. The biological and physical advantages and problems of neutron therapy. Br J Radiol 1963;36:77-80. Hall EJ. Radiobiology of heavy particle radiation therapy: cellular studies. Radiology 1973;108:119-129. Hall EJ, Astor M, Geard C, et al. Cytotoxicity of Ro-07-0582; enhancement by hyperthermia and protection by cysteamine. Br J Cancer 1977;35:809-815. Hall EJ, Graves RG, Phillips TL, et al. Particle accelerators in ­radiation therapy. Int J Radiat Oncol Biol Phys 1982;8:2041-2207. Hall EJ, Roizin-Towle L. Hypoxic sensitizers: radiobiological studies at the cellular level. Radiology 1975;117:453. Howes AE. An estimation of changes in the proportion and absolute numbers of hypoxic cells after irradiation of transplanted C3H mouse mammary tumors. Br J Radiol 1969;42:441-447. Kallman RF, Jardine LJ, Johnson CW. Effects of different schedules of dose fractionation on the oxygenation status of a transplantable mouse sarcoma. J Natl Cancer Inst 1970;44:369-377. Kinsella T, Mitchell J, Russo A, et al. Continuous intravenous ­infusions of bromodeoxyuridine as a clinical radiosensitizer. J Clin Oncol 1984;2:1144-1150.

Kinsella T, Mitchell J, Russo A, et al. The use of halogenated thymidine analogs as clinical radiosensitizers: rationale, current status, and future prospects: non-hypoxic cell sensitizers. Int J Radiat Oncol Biol Phys 1984;10:1399-1406. Mitchell J, Kinsella T, Russo A, et al. Radiosensitization of hematopoietic precursor cells (CFUc) in glioblastoma patients receiving ­intermittent intravenous infusions of bromodeoxyuridine (BUdR). Int J Radiat Oncol Biol Phys 1983;9:457-463. Mitchell J, Russo A, Kinsella T, et al. The use of non-hypoxic cell sensitizers in radiobiology and radiotherapy. Int J Radiat Oncol Biol Phys 1986;12:1513-1518. Moulder JE, Rockwell S. Hypoxic fractions of solid tumors: ­experimental techniques, methods of analysis and a survey of ­existing data. Int J Radiat Oncol Biol Phys 1984;10:695-712. Nordsmark M, Overgaard M, Overgaard J. Pretreatment oxygenation predicts radiation response in advanced squamous cell carcinoma of the head and neck. Radiother Oncol 1996;41:31-40. Overgaard J, Horsman MR. Modification of hypoxia induced ­radioresistance in tumours by the use of oxygen and sensitizers. Semin Radiat Oncol 1996;6:10-21. Overgaard J, Sand-Hansen H, Lindeløv B, et al. Nimorazole as a hypoxic radiosensitizer in the treatment of superglottic larynx and pharynx carcinoma. First report from the Danish Head and Neck Cancer Study (DAHANCA) protocol 5-85. Radiother Oncol 1991;20:143-150. Overgaard J, Sand-Hansen H, Overgaard M, et al. The Danish Head and Neck Cancer Study Group (DAHANCA) randomized trials with radiosensitizers in carcinoma of the larynx and pharynx. In Dewey WC, Eddington M, Fry RJM, et al (eds): Radiation Research.  A Twentieth Century Perspective, vol �������������������������������� II������������������������������ . New York: Academic Press, 1992, pp 573-577. Powers WE, Tolmach LJ. A multicomponent x-ray survival curve for mouse lymphosarcoma cells irradiated in vivo. Nature 1963;197:710-711. Report of the RBE Committee to the International ­ Commissions on Radiological Protection and on Radiological Units and ­Measurements. Health Phys 1963;9:357. Roizin-Towle LA, Hall EJ, Flynn M, et al. Enhanced cytotoxicity of melphalan by prolonged exposure to nitroimidazoles: the role of endogenous thiols. Int J Radiat Oncol Biol Phys 1982;8:757-760. Rose CM, Millar JL, Peacock JH, et al. Differential enhancement of melphalan cytotoxicity in tumor and normal tissue by ­misonidazole. In Brady LW, ed: Radiation Sensitizers. New York: Masson ­Publishing, 1980. Sutherland RM. Selective chemotherapy of non-cycling cells in an in vitro tumor model. Cancer Res 1974;34:3501. Sutherland RM (ed). Conference on chemical modification: radiation and cytotoxic drugs, Key Biscayne, Florida, 17-20 September 1981. Int J Radiat Oncol Biol Phys 1982;8:323-815. Thomlinson RH. Changes of oxygenation in tumors in relation to irradiation. Front Radiat Ther Oncol 1968;3:109-121. Thomlinson RH. Reoxygenation as a function of tumor size and histopathological type. In Proceedings of the Carmel Conference on Time and Dose Relationships in Radiation Biology as Applied to Radiotherapy. BNL Report 50203 (C-57). Upton, NY: Brookhaven National Laboratory, 1969, pp 242-254. Thomlinson RH, Dische S, Gray AJ, et al. Clinical testing of the radiosensitizers Ro-07-0582, III: response of tumors. Clin Radiol 1976;27:167-174. Utley JF, Marlowe C, Waddell WJ. Distribution of 35S-labeled WR2721 in normal and malignant tissues of the mouse. Radiat Res 1976;68:284-291. van Putten LM. Tumor reoxygenation during fractionated radiotherapy: studies with a transplantable osteosarcoma. Eur J Cancer 1968;4:173-182. Withers HR, Thames HD, Peters LJ. Biological bases for high RBE values for late effects of neutron irradiation. Int J Radiat Oncol Biol Phys 1982;8:2071-2076. Wong TW, Whitmore GF, Gulyas S. Studies on the toxicity and ­radiosensitizing ability of misonidazole under conditions of prolonged incubation. Radiat Res 1978;75:541-555.

1  Physical and Biologic Basis of Radiation Therapy Yuhas J. Active versus passive absorption kinetics as the basis for selective protection of normal tissues by S-2-(3-aminopropylamino)ethyl-phosphorothioic acid. Cancer Res 1980;40:1519-1524.

Adjuncts to Radiation Cox JD, Byhardt RW, Komaki R, et al. Interaction of thoracic irradiation and chemotherapy on local control and survival in small cell carcinoma of the lung. Cancer Treat Rep 1979;63:1251-1255. Dewey WC, Hopwood LE, Sapareto SA, et al. Cellular responses to combinations of hyperthermia and radiation. Radiology 1977;123:463-474. Eddy HA. Alterations in tumor microvasculature during ­hyperthermia. Radiology 1980;137:515-521. Elkind MM. Fundamental questions in the combined use of radiation and chemicals in the treatment of cancer. Int J Radiat Oncol Biol Phys 1979;5:1711-1720. Gerner EW, Boone R, Conner WG, et al. A transient thermotolerant survival response produced by single thermal doses in HeLa cells. Cancer Res 1976;36:1035-1040. Gerweck LE, Gillette EL, Dewey WC. Killing of Chinese hamster cells in vitro by heating under hypoxic or aerobic conditions. Eur J Cancer 1974;10:691-693. Gerweck LE, Gillette EL, Dewey WC. Effect of heat and radiation on synchronous Chinese hamster cells: killing and repair. Radiat Res 1975;64:611-623. Hahn GM, Braun J, Har-Kedar I. Thermochemotherapy: synergism between hyperthermia (42°-43°) and Adriamycin (or bleomycin) in mammalian cell inactivation. Proc Natl Acad Sci U S A 1975;72:937-940. Hall EJ, Roizin-Towle L. Biological effects of heat. Cancer Res 1984;44:4708-4713S. Perez C, Brady L, Cox J, et al. Randomized study to evaluate efficacy of levamisole in patients with unresectable non–oat cell carcinoma of the lung treated with radiation therapy. Int J Radiat Oncol Biol Phys 1984;10:97-98. Rubin P. The Franz Buschke Lecture. Late effects of chemotherapy and radiation therapy: a new hypothesis. Int J Radiat Oncol Biol Phys 1984;10:5-34. Stewart JR. Past clinical studies and future directions. Cancer Res 1984;44:4902-4904S. Tefft M, Chabora BMcC, Rosen G. Radiation in bone sarcomas. A reevaluation in the era of intensive systemic chemotherapy. Cancer 1977;39:806-816. Tubiana M. Les associations radiotherapiechemotherapie. J Eur Radiotherapy 1980;1:107-114. Westra A, Dewey WC. Variation in sensitivity to heat shock during the cell cycle of Chinese hamster cells in vitro. Int J Radiat Biol 1971;19:467-477.

Mutagenesis and Carcinogenesis Boice JD Jr, Land CE, Shore RE, et al. Risk of breast cancer ­following low-dose exposure. Radiology 1979;131:589-597. Bond VP, Thiessens JW (eds). Re-evaluation of Dosimetric Factors: Hiroshima and Nagasaki. Proceedings Symposium. Springfield, VA: US Department of Energy, US Department of Commerce, 1982. Cardis E, Gilbert ES, Carpenter L, et al. Effects of low doses and low dose rates of external ionizing radiation: cancer ­ mortality among ­ nuclear industry workers in three countries. Radiat Res 1995;142:117-132. Committee on the Biological Effects of Ionizing Radiation. BEIR III Committee Report: Effects on Populations of Exposure to Low Levels of Ionizing Radiation. Washington, DC: National Academy of Sciences, National Research Council, 1980. Court Brown WM, Doll R. Mortality from cancer and other causes after radiotherapy for ankylosing spondylitis. Br Med J 1965;2:  1327-1332.

49

Court- Brown WM, Doll R, Hill AB. The incidence of leukemia following exposure to diagnostic radiation in utero. Br Med J 1960;2:1539-1545. Doll R, Smith PG. The long-term effects of x-irradiation in patients treated for metropathia haemorrhagic. Br J Radiol 1968;41:  362-368. Doll R, Wakeford R. Risk of childhood cancer from fetal irradiation.  Br J Radiol 1997;70:130-139. Finkel AL, Miller CE, Hasterlik RJ: Radium-induced tumors in man. In Mays CW (ed): Delayed Effects of Bone-Seeking Radionuclides. Salt Lake City: University of Utah Press, 1969. Jablon S, Belsky JL, Tachikawa K, et al. Cancer in Japanese exposed as children to atomic bombs. Lancet 1971;1:927-932. Jablon S, Kato H. Childhood cancer in relation to prenatal exposure to atomic bomb radiation. Lancet 1970;2:1000-1003. Hempelmann LH. Risk of thyroid neoplasms after irradiation in childhood. Science 1968;160:159-163. Hempelmann LH, Pifer JW, Burke GJ, et al. Neoplasms in persons treated with x-rays in infancy for thymic enlargement: a report of third follow-up survey. J Natl Cancer Inst 1967;38:317-341. IARC Study Group on Cancer Risk Among Nuclear Industry ­Workers. Direct estimates of cancer mortality due to low doses of ionizing radiation: an international study. Lancet 1994;344:1039-1043. International Commission on Radiological Protection. ICRP 60: 1990 Recommendations of the International Commission on ­Radiological Protection. Oxford, England: Pergamon Press, 1990. MacKenzie I. Breast cancer following multiple fluoroscopies. Br J Cancer 1965;19:1-8. MacMahon B. Prenatal x-ray exposure and childhood cancer. J Natl Cancer Inst 1962;28:1173-1191. Modan B, Baidatz D, Mart H, et al. Radiation induced head and neck tumors. Lancet 1974;1:277. Muller HJ. The nature of the genetic effects produced by radiation. In Hollaender A (ed): Radiation Biology. New York: McGraw-Hill, 1954. Pierce DA, Shimizu Y, Preston DK, et al. Studies of the mortality of atomic bomb survivors, report 12. Part I. Cancer: 1950-1990. Radiat Res 1996;146:1-27. Rowland RE, Stehney AF, Lucas HF. Dose-response relationships for female radium dial workers. Radiat Res 1978;76:308. Russell WL. X-ray-induced mutations in mice. Cold Spring Harbor Symp Quant Biol 1951;16:327-336. Russell WL. The nature of the dose-rate effect of radiation on mutation in mice. Jpn J Genet 1965;40:128-140. Schull WL, Otake M, Neal JV. Genetic effects of the atomic bomb:  a reappraisal. Science 1981;213:1220-1227. Selby PB. Induced skeletal mutations. Genetics 92:127-133, 1979. Selby PB. The doubling dose for radiation, or for any other mutagen, is actually several times larger than has been previously thought if it is based on specific-locus mutation frequencies in mice. Environ Mol Mutagen 1996;27:61. Stewart A, Kneale GW. Changes in the cancer risk associated with obstetric radiography. Lancet 1968;1:104-107. Stewart A, Webb J, Hewitt D. A survey of childhood malignancies.  Br Med J 1958;1:1495-1508. United Nations Scientific Committee on the Effects of Atomic ­Radiation. Sources and Effects of Ionizing Radiation. New York: United Nations, 1977. United Nations Scientific Committee on the Effects of Atomic Radiation. Ionizing Radiation Sources and Biological Effects. New  York: United Nations, 2001. Upton AC. The dose response relation in radiation induced cancer. Cancer Res 1961;21:717-729. Wolff S. Radiation genetics. Annu Rev Genet 1967;1:221-244.

2

Clinical Radiation Oncology Physics George Starkschall, Lei Dong, Peter A. Balter, Almon S. Shiu, Firas Mourtada, Michael Gillin, And Radhe Mohan PLANNING FOR PHOTON BEAM RADIATION THERAPY Target Volume Delineation Imaging for Treatment Planning Immobilization Dose-Calculation Algorithms Inverse Planning Field-in-Field Planning DELIVERY OF PHOTON BEAM RADIATION THERAPY Multileaf Collimation Fixed-Beam Attenuation Helical Tomotherapy Quality Assurance of Intensity-Modulated Radiation Therapy Delivery STEREOTACTIC RADIATION THERAPY Image-Guided Stereotactic Radiosurgery or Radiation Therapy for Cranial Tumors Image-Guided Stereotactic Radiation Therapy for Extracranial Spinal Tumors Image-Guided Stereotactic Body Radiation Therapy for Thoracic Tumors ACCOUNTING FOR MOTION IN RADIATION THERAPY Intrafractional Motion: RespiratoryCorrelated Radiation Therapy Interfractional Motion: Image-Guided Radiation Therapy

The similarly named chapter in the eighth edition of this textbook provided insight into what was perceived to be the frontier of technologic advances in the planning and delivery of radiation in 2002. Probably the major ­advance at that time was the increasing use of ­intensity­modulated radiation therapy (IMRT) to complement three-­dimensional conformal radiation therapy. Cranial stereotactic radiation therapy was also becoming a standard mode of treatment, as was high-dose-rate brachytherapy. Intravascular brachytherapy showed significant promise, and respiratory gating was being proposed as a method to account for intrafractional respiratory-induced tumor ­motion. Major changes have taken place in radiation therapy since the previous edition. IMRT planning and delivery are now performed routinely for the treatment of tumors at many anatomic sites, although the definition of an appropriate objective function and constraints still remains more of an art than a science. As for frontier technologies,

50

EVOLUTION OF IN-ROOM IMAGING FOR RADIATION THERAPY On-Board Kilovoltage X-Rays Sonographically Guided Prostate Localization Three-Dimensional In-Room Volumetric Imaging Image-Guided Adaptive Radiation Therapy DEVELOPMENTS IN BRACHYTHERAPY Pulsed-Dose-Rate Brachytherapy for Cervical Cancer Liquid Source Brachytherapy for Brain Tumors Partial Breast Brachytherapy Beta Particle Plaques for Ocular Melanoma PROTON RADIATION THERAPY Production of Protons Passively Scattered Proton Beams Shaping Proton Dose Distributions Treatment Planning for Passively Scattered Protons Evaluating Proton Treatment Plans Discrete Spot Scanned Proton Beams Proton Radiobiologic Effectiveness FUTURE TECHNOLOGIC DEVELOPMENTS IN RADIATION THERAPY Image-Guided Radiation Therapy Biologic and Functional Imaging and Dose-Response Assessment and Modeling Proton Therapy

cranial stereotactic radiation therapy and high-dose-rate brachytherapy were replaced by extracranial stereotactic radiation therapy and pulsed-dose-rate (PDR) brachytherapy. Intravascular brachytherapy has not become a significant component of brachytherapy practice, having been replaced by medical intervention. Although gated beam delivery is still considered a promising method for reducing the size of treatment portals, it has not achieved widespread acceptance; instead, explicit determination of respiratory motion is used to customize the delineation of the treatment portal to the particular patient. The current buzzword is image-guided radiation therapy, in which imaging immediately before radiation delivery is used to modify the position of the patient or the parameters of the beam to reflect intrafractional motion and interfractional variations in the patient. Several institutions have opened or are soon to open proton radiation therapy facilities, with the promise of still more precise delivery of radiation to the target while sparing uninvolved tissue.

2  Clinical Radiation Oncology Physics

With these developments in mind, this chapter provides a current snapshot of clinical radiation physics, with particular emphasis on procedures in place or nearly so in the Radiation Oncology Clinic at The University of Texas M. D. Anderson Cancer Center. The first part of the chapter addresses the planning and delivery of high-precision photon beam radiation therapy. A section on stereotactic radiation therapy follows, with emphasis on extracranial hypofractionated treatment of the central nervous system and thorax. The handling of intrafractional motion and interfractional variations is the topic of the next section, ­focusing primarily on methods for imaging these variations. New approaches to brachytherapy are discussed, including PDR brachytherapy, liquid-source brachytherapy, and the use of novel beta-emitting isotopes. Some aspects of proton radiation therapy are presented (see Chapter 40). This chapter closes with a brief review of future technologic ­developments that are likely to take place in clinical radiation ­oncology physics, setting the stage for what may be the topics of interest in the next edition of this book.

Planning for Photon Beam Radiation Therapy The current practice of radiation therapy mandates great precision in the planning and the delivery of radiation. As doses and doses per fraction are escalated with the goal of increasing tumor control rates, the margin for uncertainty needs to be made smaller. Consequently, the need for more precise targeting of tumor volumes has increased, as has the need for accurate accounting of uncertainties in patient setup and tumor motion and for sparing critical uninvolved anatomic structures. Newer technologies have facilitated the achievement of these goals. The following section focuses on methods for high-precision planning for photon beam radiation therapy; this section is followed by aspects of highprecision delivery of photon beam radiation therapy.

A

51

Target Volume Delineation Planning and delivery of high-precision radiation therapy rely on the accurate definition and delineation of target volumes and critical anatomic structures. Although the definitions originally presented by the International Commission on Radiation Units and Measurements (ICRU) were designed to facilitate communication of information about radiation dose and volume, these definitions also provide a rational method for understanding the various factors involved in determining what regions of a patient need to be treated.1,2 Although the ICRU dose and volume definitions were specifically designed for photon radiation therapy, some of the concepts presented in these definitions can also be applied to therapy with other forms of ionizing radiation. Several of these definitions are applicable to oncology in general, independent of the therapeutic modality used. The ICRU1 defines gross tumor volume (GTV) as “the gross demonstrable extent and location of the malignant growth.” The basic rule of thumb for defining the GTV is that if it can be imaged and it is tumor, it is part of the GTV. What is not specified in the ICRU definition is the mode of imaging or the imaging parameters. The GTV may be different when it is determined by different ­ imaging ­modalities; the GTV may even be different for the same modality when different image viewing parameters are used (e.g., window width, window level settings). Consequently, consistency in assessing tumor location and ­extent for each tumor site is highly desirable. As an example, Figure 2-1 depicts two transverse computed tomography (CT) images of a patient’s thorax, one with the window and level settings normally used for displaying mediastinal structures (see Fig. 2-1A) and the other with the window and level settings normally used for displaying lung structures (see Fig. 2-1B). Comparing the two figures reveals that a lung tumor depicted with a mediastinal window may appear quite different from that depicted with a lung window. The GTV alone does not constitute the extent of tumor spread, because it does not include microscopic disease.

B

FIGURE 2-1.  Transverse CT images of a thorax using a mediastinal window and level (A) and a lung window and level (B) illustrate differences in the visualization of a lung tumor.

52

PART 1  Principles

According to the ICRU,1 the clinical target volume (CTV) is defined as “tissue volume that contains a demonstrable GTV and/or subclinical malignant disease that must be eliminated.” The CTV constitutes the volume of tumorbearing cells that must be irradiated to an appropriate dose to control the tumor. Failure to adequately irradiate the CTV to an appropriate dose may result in failure to achieve tumor control. It is possible to have a CTV without a GTV, as would be the case for a tumor that has been surgically resected. It is also possible to have multiple CTVs in a treatment scheme, such as when one dose is prescribed for the primary tumor and another for nodal disease. Exactly how the GTV should be expanded to generate the CTV is based partly on clinical experience and partly on specific pathology studies. For example, Giraud and colleagues3 assessed pathologic records from a series of patients with non–small cell lung tumors and established expansion margins of 8 mm for adenocarcinoma and 6 mm for squamous cell carcinoma. However, these expansion margins were based on a particular method for delineating the GTV. As imaging modalities are developed that can more precisely define the extent of tumor-bearing tissue, these expansion margins may need to be modified. For ­example, if some sort of in vivo pathologic assessment were possible by using extremely high-resolution CT, enabling larger amounts of tumor to be visualized, it might be possible to reduce the margins that generate the CTV from the demonstrable GTV. Because of the improvement in imaging, some portion of tumor that was originally considered CTV and not GTV would become part of the GTV. Most radiation treatment planning systems support software tools that generate the CTV from the GTV by means of uniform expansion of the GTV. After this expansion has been generated, however, the contours of the CTV must be reviewed and edited based on clinical experience. For example, if microscopic disease is limited to the tumorbearing organ, the CTV can be edited to avoid adjacent structures, provided the radiologist is confident that microscopic extension has not occurred. Radiation treatment has long been based on building a patient model based on acquisition of a CT scan several days before treatment and on the assumption that this model is an accurate reflection of the patient for the duration of the radiation treatment. Recognizing that tumors move, the ICRU2 defined another volume that accounts for tumor motion, the internal target volume (ITV). The ITV consists of the CTV plus an internal margin that ­accounts for intrafractional and interfractional motion. Margins ­ accounting for motion can be population-based or specifically measured for each patient. Methods to measure and mitigate the effects of various forms of motion that allow minimization of the internal margin are discussed under ­“Accounting for Motion in Radiation Therapy” in this chapter. The final uncertainty specified by the ICRU is the planning target volume (PTV), which accounts for variations in patient setup. The PTV is the ITV plus a setup margin and is defined as “a geometrical concept used for treatment planning, defined to select appropriate beam sizes and beam arrangements to ensure that the prescribed dose is actually delivered to the CTV.”1 The PTV is the entity that must be irradiated to the desired prescription dose to ensure that the entire CTV (i.e., tumor-bearing tissue) ­receives the prescribed radiation dose. Each CTV has a corresponding PTV. The setup

margin is determined by clinical ­experience and depends on the disease site and the type of patient ­immobilization used. Invasive forms of immobilization, such as stereotactic head frames, allow millimeter-level setup margins, whereas other forms of ­ immobilization may result in setup margins of 5 to 10 mm or more. A transverse CT image of a thorax from a treatment plan used to treat a lung tumor, illustrating a GTV, a CTV, and a PTV, is shown in Figure 2-2. The projection of the PTV in a beam’s eye view does not by itself define the radiation treatment portal; instead, an additional 5- to 7-mm margin must be added to the beam’s eye view projection of the PTV to account for the beam penumbra. The ICRU also addresses organs at risk, which are by definition site-specific. For example, the organs at risk in the treatment of lung tumors typically include the spinal cord, lungs, esophagus, and heart. Other organs, such as the brachial plexus or liver, may also be included, depending on the location of the tumor. In planning the radiation treatment of a lung tumor, the planner must remain cognizant of the dose received by these anatomic structures. ­Accounting for setup uncertainty and motion of the organs at risk ­requires placing a margin around those organs and introduces the concept of the planning organ-at-risk volume. The ostensible goal in planning for radiation treatments is to deliver the prescribed dose to the PTV while keeping the dose to specified volumes of the planning organs at risk within appropriate limits. In reality, compromises are often necessary. Sometimes, setup uncertainty and organ motion lead to overlap between portions of the PTV and the planning organ-at-risk volume such that delivery of high doses to the former may result in overdosing the latter. Because the CTV does not spend all of its time in all of the PTV, a typical criterion for PTV coverage is that 95% of the PTV receives the prescribed dose, and the entire CTV would receive the prescribed dose. However, if coverage of the CTV must be reduced to protect an organ at risk, the reduction is done with the recognition that doing so increases the probability that a portion of tumor cells may not receive a radiation dose considered adequate for tumor control. One of the goals of improving the quality of radiation therapy is to reduce the margin between the GTV and PTV. In an ideal situation, the imaging modality or modalities would

FIGURE 2-2.  Transverse CT image of a thorax from a treatment plan used to treat a lung tumor illustrating the gross tumor  volume (GTV; yellow line), clinical target volume (CTV; red line), and planning target volume (PTV; blue line).

2  Clinical Radiation Oncology Physics

53

1.8 1.6 1.4 Density

1.2 1.0 0.8 0.6 0.4 0.2 0.0 0

500

1000 1500 2000 2500 3000 3500 4000 4500 CT number

FIGURE 2-3.  CT number-to-density conversion table extracted from the radiation treatment planning system.

explicitly display all of the tumor-bearing tissue, and motion and setup uncertainties would be completely eliminated; what is imaged would be what is irradiated. ­Although this is often not the case in practice, it does demonstrate where improvements can be made in targeting. For example, better means of imaging the tumor with various imaging modalities and image-enhancing (contrast) agents may enhance the imaging of microscopic disease and ­reduce the margins used between the GTV and the CTV. (Some of these modalities are described under ­“Imaging for Treatment Planning.”) The magnitude of the expansion of the CTV to generate the ITV can be reduced in two ways: by explicitly imaging the tumor motion and by reducing the amount of intrafractional and interfractional motion. (These issues are addressed under “Accounting for Motion in Radiation Therapy.”) The magnitude of setup ­uncertainty can be ­reduced, allowing reductions in the margin used to expand the ITV to form the PTV. Setup uncertainty can be reduced by more effective methods of patient immobilization and by some of the techniques being used for ­ image-guided ­ radiation ­ therapy, which are discussed further in subsequent sections. The pretreatment GTV in a particular patient is assumed in the sense that the extent or location of the tumor cannot be controlled. The CTV is also assumed, but it could be delineated more precisely by improving the ability to detect microscopic disease and eliminate some uncertainty about its extent. The quantities that can be controlled are the ITV and the PTV. Logically, the potential for dose escalation would be based on the ability to safely reduce the size of the PTV by using better imaging and immobilization methods, as described in the sections that follow.

Imaging for Treatment Planning Imaging has a major role in radiation treatment planning. Effectively reducing the size of the region being irradiated requires methods for accurately determining the location and

extent of tumor. This can be achieved by improving existing imaging technologies and by the use of multimodality imaging. CT imaging, currently the standard for radiation treatment planning, is the most desirable imaging modality for this purpose because it most closely simulates the effects of ionizing radiation passing through the patient. However, CT images are typically acquired by using x-rays in the ­ kilovoltage energy range (120 to 140 kVp), whereas the ­radiation delivered as treatment is in the megavoltage ­energy range. In the kilovoltage energy range, a significant fraction of the photons interact by means of the photoelectric effect. In this energy range, the deposition of ­ energy and the attenuation of radiation depend greatly on the atomic number of the absorber. In the megavoltage energy range, almost all interactions are those of Compton scatter. Energy deposition and radiation dose depend almost entirely on electron density. Consequently, the CT voxel values (which reflect x-ray attenuation) do not exactly ­reproduce the extent to which megavoltage radiation interacts with tissue. Before radiation doses based on CT images can be calculated in a radiation treatment planning system, a table must be generated to convert CT numbers to electron densities. In these tables, each region of CT densities is assumed to represent a different type of tissue, with a ­different atomic number, and the CT-to-density relationship is slightly different for each of these types of tissue (Fig. 2-3). High-Z materials may also cause artifacts on CT ­images because of inadequate accounting for beam hardening due to the increased attenuation of the kilovoltage x-rays in the CT scanner.4 The technology of CT image acquisition has progressed substantially in recent years. The current form of CT technology, helical or spiral CT, is characterized by continuous table motion synchronized with continuous gantry ­rotation. Multislice helical CT scanners (Fig. 2-4) acquire

54

PART 1  Principles

A

B

FIGURE 2-4.  Two multislice, helical CT scanners are used to acquire images for radiation treatment planning in the Radiation Oncology Clinic at M. D. Anderson Cancer Center: the Philips Brilliance 64 CT scanner (A) and the GE Discovery ST CT scanner (B), which is a combined CT scanner with a PET scanner and uses the same patient support system.

2 to 64 image slices, corresponding to a region thickness of 2 to 4 cm, during a single gantry rotation. The ability to acquire patient information over a 2- to 4-cm-thick ­region coupled with gantry rotation times of less than 0.5 seconds has resulted in very short image acquisition times. For ­example, a CT image data set of the thorax extending from the mandible to the base of the liver can be acquired in 5 to 10 seconds. Although short CT image acquisition times have the ­advantage of reducing the magnitude of motion during ­image acquisition, they also introduce motion artifacts. For example, the diaphragm moves in an approximately supero­ inferior direction, parallel with that of the patient during CT image acquisition. Mismatches in the synchronization of the motion of the diaphragm with that of the patient during image acquisition can lead to artifacts in the diaphragm images (Fig. 2-5). Because the time scale for image acquisition is significantly shorter than that for radiation treatment, the CT images acquired on a multislice helical CT scanner may not reflect patient motion that is averaged during radiation treatment. An extension of the multislice CT concept is the cone-beam CT scanner. This type of scanner has no ­table ­translation; a flat-panel detector is placed opposite a ­kilovolt-energy radiation source, and an image is acquired during a single gantry rotation. Many cone-beam CT scanners are manufactured as components of linear accelerators (Fig. 2-6). The image quality of the cone-beam CT scanner may not be as high as that of a conventional CT scanner (Fig. 2-7), and cone-beam scanners therefore are used more for image-guided setup for therapy than for ­treatment ­planning. Although CT imaging provides the most appropriate representation of the interactions of radiation with the ­patient and for imaging anatomic regions of different mass densities, various aspects of tumors may be ­visualized ­better with other imaging modalities. Tumor metabolic activity, for example, is better visualized with positron emission tomography (PET) than with CT. PET is based on the annihilation of positrons with electrons, generating two 511-keV gamma rays emitted in opposite directions. Positron-emitting radionuclides such as fluorine 18 (18F)

FIGURE 2-5.  An artifact in a coronal CT image of a the thorax is caused by diaphragm motion during the acquisition of a fast CT scan. The red line delineates the gross tumor volume at end inspiration; the green line delineates the gross tumor volume at end expiration.

tend to be small atoms and typically bind to physiologically active molecules. Significant interest has developed during the past few years in the use of the compound 18F-fluorodeoxyglucose (FDG), the concentration of which is related to the degree of metabolic activity. Because ­ tumors tend to be more metabolically active than uninvolved tissue, the concentration of FDG, indicated by bright areas on PET scans, can be useful for identifying areas of tumor and nodal activity. A major problem in the use of PET imaging is registering the PET image with the CT image. Although PET may indicate areas of tumor activity better than CT, the radiation treatment planning still must be done with CT

2  Clinical Radiation Oncology Physics

images. The lack of commonality between the two types of images has made registration a significant problem. One solution has been to register the CT image with a transmission ­image, which is normally acquired along with the PET image to correct the PET image for the attenuation of the emitted gamma ray photons. The transmission image is usually ­obtained with a gamma emitter and depicts anatomy similarly to that on a CT image. Because the transmission image is acquired with the same apparatus as the PET image, it is automatically registered to the PET image, enabling the PET image to be registered to the CT image. Development of PET/CT scanner combinations has made the acquisition of the transmission image easier and the registration of the PET image with the planning CT

(d) (a)

(b)

(e)

(c)

FIGURE 2-6.  A kilovoltage source (a) and flat-panel imager (b) are attached to a linear accelerator gantry to produce cone-beam CT image data sets. The treatment table (c), linear ­accelerator source (d), and electronic portal imaging device (e) are shown.

A

55

i­mage automatic. These combined scanners consist of a dual gantry, a PET imager, and a CT imager (see Fig. 2-4B), with the patient table translated between the two ­imagers. Registration is no longer an issue, because the amount of table translation between the two imagers is a known quantity. When the CT image data from a PET/CT combination are used as a transmission correction, the difference in ­acquisition time can be problematic. Because the CT image is acquired over a much shorter time than the PET image, motion artifacts on the CT image data set may ­adversely affect the quality of the attenuation correction of the PET image. Pan and colleagues5 proposed that ­attenuation correction data for PET images be based on a respiration-­averaged CT scan rather than on a free-­breathing CT scan. In the example shown in Figure 2-8, use of the ­respiration-averaged CT data set (see Fig. 2-8B) led to a twofold ­increase in the specific uptake value of the tumor on the PET image (3.8 versus 1.9 for the free-breathing CT) (see Fig. 2-8A). Another imaging modality that can aid in treatment planning is single photon emission computed tomography (SPECT), which involves imaging single photons emitted from an internally applied gamma ray source. SPECT can be particularly useful for assessing lung perfusion and thereby the amount of functional lung tissue in a patient. Radiation treatment planning can then be designed to avoid irradiating well-functioning lung regions, with the idea that if a fraction of lung has to be sacrificed to achieve the prescription goals of the treatment plan, that region should be poorly functioning lung. Figure 2-9 illustrates a SPECT image combined with a CT image used for treatment planning for such a case. SPECT scanners can be combined with CT scanners to improve SPECT/CT image registration. Combined SPECT/CT scanners may not have the same advantages for applications involving only imaging as a combined PET/ CT scanner does, because the CT component of the PET scanner is also needed to obtain the transmission image for ­attenuation correction. Some of the issues related to the differences between photon interactions in CT scanners and those in linear ­accelerators are being addressed by the introduction of megavoltage CT imaging.6 High-Z materials, which produce significant artifacts in kilovoltage CT scans, have no such artifacts on megavoltage CT scans. Megavoltage CT scanners are constructed as cone-beam scanners that use the radiation beam from a linear accelerator or in conjunction with a helical tomotherapy unit. Megavoltage cone-beam

B

FIGURE 2-7.  Images of the same transverse slice of a thorax were obtained by conventional CT (A) and cone-beam CT (B).

56

PART 1  Principles

CT scanners typically are used for patient alignment and dose verification rather than for initial treatment planning.

Immobilization A key approach in reducing setup uncertainty and limiting some intrafractional motion is the use of appropriate immobilization techniques. The specific technique to be

A

used is based on the tolerance for setup uncertainty and patient motion for a given type of treatment (Table 2-1). Cranial stereotactic radiosurgery, for example, requires submillimeter tolerances, and invasive immobilization in the form of stereotactic head frames is used. For treatment of most head and neck tumors, the acceptable setup uncertainty is slightly greater than that for stereotactic

B

FIGURE 2-8.  PET images show an abdominal tumor with the attenuation correction based on a short-acquisition free-­breathing CT image (A) or a time-averaged CT image (B). The cursor is pointing to the tumor. (Courtesy of T. Pan, M. D. ­Anderson Cancer Center, Houston, TX.)

FIGURE 2-9.  Transverse, sagittal, and coronal views of SPECT images are combined with CT images to identify regions of high lung perfusion as critical avoidance structures for radiation treatment planning. The solid white and colored lines outline the various ­isodoses. Lighter areas in the lungs indicate regions of higher perfusion. (Courtesy of H. Liu, M. D. Anderson Cancer Center, Houston, TX.)

2  Clinical Radiation Oncology Physics

57

TABLE 2-1.   Immobilization Devices Used in Radiation Oncology Clinics and Their Associated Setup Uncertainties Immobilization Device

Application

Setup Uncertainty

Invasive head frame

Cranial stereotactic

0.5 mm

Thermoplastic mask and vacuum bag

Head and neck, CNS

3 mm

T bar and vacuum bag

Thorax

7 mm

T bar, extended vacuum bag (indexed to table), CT guidance

Thorax (hypofractionated)

3 mm

Vacuum bag, breast board indexed to table

Breast

2-5 mm

Knee and foot immobilization cushions indexed to table

Prostate

3 mm

CNS, central nervous system, CT, computed tomography.

FIGURE 2-10.  A customized thermoplastic mask is used to immobilize a patient being treated for a tumor of the head and neck region. The red lines are the localizing lasers in the simulation and treatment rooms.

FIGURE 2-11.  A vacuum bag and T-bar assembly is used to immobilize patients being treated for thoracic lesions. The red lines are the localizing lasers in the simulation and treatment rooms.

radiosurgery, and a thermoplastic mask (Fig. 2-10) is used. Thoracic treatments are delivered after the patient is set up with a T bar and vacuum bag system (Fig. 2-11). Patients are positioned on the vacuum bag, with their arms ­extended over their heads and their hands holding the T bar. Air is then drawn out of the vacuum bag, which ­resembles a beanbag filled with very small pellets that harden as the air is withdrawn, forming a template on which the patient can lie so that the patient’s position is reproducible from treatment to treatment. Setup marks are made on the ­patient. When greater precision in setup is required, such as for hypofractionated treatments of the thorax, a larger vacuum bag is used in combination with a table index to allow the bag to be placed in the same position on the table for each treatment. Markings in this case are made on the bag rather than on the patient.

use some form of the convolution algorithm,7,8 the current state of practice, for photon beam dose calculations. The concept behind the convolution algorithm is the existence of a dose-spread array that describes how unit photon fluence deposits dose in the vicinity of an interaction. The dose-spread array depends on only the positional relationship between the calculation point and the site of interaction, as well as the energy of the photons, and it needs to be calculated only once and then stored in the computer. Calculating photon fluence is relatively straightforward; as the photon beam passes through the patient, the fluence is determined by means of exponential attenuation. By modeling the photon fluence as a spectrum of energies and attenuating each spectral component of the photon beam based on its own energy-specific attenuation coefficient, hardening of the radiation beam as it passes through beam modifiers or the patient can be accurately modeled. The dose distribution is then calculated by convolving the dose-spread array with the photon fluence ­distribution. Commissioning the convolution model for a radiation treatment planning system involves modeling the photon fluence that is incident on the patient surface. Although this calculation needs to be done only once for a beam, the beam model may consist of many spectral components

Dose-Calculation Algorithms An important aspect of radiation treatment planning is the accurate simulation of the radiation dose distribution on a computer model of the patient. Various dose-calculation algorithms have been developed to balance the need for computational accuracy with time constraints. The most advanced radiation treatment planning systems typically

58

PART 1  Principles

and several empiric parameters describing the incident fluence distribution, and the calculations can be complex and time-consuming.9 Accuracy of computed dose distributions must then be verified against measured beam data. Still more accurate dose calculations can be achieved by using Monte Carlo models, which explicitly model specific photon interactions. These calculations require significantly more computational power than do convolution calculations, but they are also capable of accurately modeling regions of electronic disequilibrium, such as in the buildup region or at tissue interfaces. Monte Carlo calculations are used more for research than for clinical treatment planning, in part because their computational intensity requires high-speed parallel processors. This issue should be overcome in the near future, given the significant increases in computational power during the past few years. Increases in the accuracy and speed of dose calculations have raised interest in incorporating heterogeneity corrections into treatment planning. Although these corrections can more accurately depict the dose received in the delivery of radiation treatment, many radiation oncology practices have been reluctant to incorporate them into treatment planning, perhaps because the doses calculated by using heterogeneity-corrected calculations might be different from what radiation oncologists were accustomed to. For example, the clinical response to a dose of 60 Gy calculated for a “homogeneous” patient (i.e., the traditional technique, still in widespread use) may be different from the clinical response to a dose of 60 Gy calculated with the true heterogeneities present. Consequently, the ­ transition from calculations in homogeneous medium to calculations in a heterogeneous patient must be made carefully, ­because it is important not to modify the radiation dose that is ­delivered to the patient.10

Inverse Planning In traditional treatment planning, radiation was delivered with spatially uniform (“flat”) fields. Occasionally, wedges were used to modify the shape of the dose distribution to account for oblique beam configurations or for sloping surfaces, and compensating filters were used to modify the intensity of the incident beam to account for major surface irregularities, but the ultimate goal was a uniform dose distribution at some specified depth. Multiple fields with specified orientations were required to ensure a uniform dose distribution in the target volume and an acceptable dose distribution to uninvolved tissue outside of the target. The discovery that the conformality of the dose to desired volumes could be improved by using radiation beams of spatially varying intensities gave rise to IMRT and inverse planning, a process to determine the beam intensities needed to deliver the dose distribution that would achieve the desired clinical objectives. The inverse planning process involves identifying and optimizing a metric of plan quality called a cost function or objective function. (For the sake of generality, we ­assume that the objective function must be minimized; the ­inverse planning problem can be easily reformulated if the objective function has to be maximized.) This metric is a measure of the match between a desired dose distribution and the calculated dose distribution. Several issues are involved in selecting an objective function. The function

must be realistic, achievable, and clinically relevant. Early inverse planning efforts used the square of the difference ­between the calculated doses and the desired doses to target ­volumes and each of the normal critical structures as the objective function. Although mathematically this quantity is relatively easy to optimize, it does not realistically depict the true clinical goals. For example, the dose to a normal anatomic structure does not have to be equal to a desired dose for a plan to be optimal; instead, it should be less than some maximum allowable dose. Such inequalities can be incorporated into an optimization scheme as dose constraints; however, this condition is still not necessarily realistic because doses to individual points in a structure can often be exceeded as long as the dose to a specified volume of that structure is not exceeded. This condition gives rise to the specification of dose-volume constraints, stated as, for example, “No more than 35% of the total lung volume may receive more than 20 Gy.” The objective function can also be stated in biologic terms,11 such as a combination of the probability of tumor control and the probability of normal tissue complications. These objectives can be combined with the goal of maximizing the probability of uncomplicated local control.12 Alternatively, the objective can be stated in terms of equivalent uniform dose, which can be maximized to the target volume and minimized to uninvolved normal structures.13 Stating the objective function in biologic terms has several potential problems, however. For example, models for these biologic factors may accurately calculate these quantities but may not have a mechanistic basis or be valid for inverse planning. Verifying the validity of these models can be problematic. The models also increase the mathematical complexity of the objective function, requiring more sophisticated methods to optimize that function. Inverse planning is further complicated by the need to incorporate relative weighting of various components of the objective function. For example, the relative weighting of overdosing the spinal cord versus underdosing the target volume is a clinical decision that may be different among individual clinicians. In addition to uncertainties in weighting, specific values used in dose-volume constraints may be based on specific treatment techniques and may not be necessarily be generalized to other techniques. For ­example, Graham and colleagues14 established V20, the ­total volume of lung receiving 20 Gy, as the most significant risk factor for pneumonitis in the radiation treatment of non–small cell lung cancer. Because fixed beams were used for the treatment, it is possible that the V20 values correlated with the V10, V5, and V30 values, and one of these other dose-volume values may be a more accurate metric of potential lung toxicity. The use of intensitymodulated beams can significantly reduce the amount of lung receiving 20 Gy, but it does so at the expense of lung receiving 5 Gy, breaking the correlation in dose-volume levels. The objectives and constraints in an inverse planning problem may render the problem infeasible in that the dose-volume constraints cannot be met by any combination of beams and beam intensity patterns. In such cases, the constraints must be relaxed before the objective function can be optimized. Another possibility is that even ­after optimization, the optimal dose distribution may not

2  Clinical Radiation Oncology Physics

be clinically acceptable; in that case, it may be necessary to reformulate the constraints and objectives and reoptimize. After the constraints and objectives have been determined, the inverse planning problem must be solved by using one of a large number of optimization algorithms. Typically, the choice of algorithm is not up to the treatment planner but rather is a function of the particular treatment planning system. In summary, the process of setting up a treatment plan for inverse planning includes the explicit identification of target volumes and critical structures, followed by establishment of constraints and objectives for these regions of interest along with relative weightings for each constraint and objective. Because inverse planning is still somewhat of an art, it is often necessary to develop various tricks to aid in achieving inverse planning goals. For example, ­ensuring adequate coverage of target volumes or appropriate avoidance of critical structures may require adding various margins around those volumes and structures. Rings are sometimes defined around target volumes to ensure a high dose gradient outside the target volume. Ensuring that hot spots do not occur outside the specified critical structures may require that otherwise unspecified uninvolved tissue be considered a critical structure. Evaluating an IMRT plan requires distinguishing between structures that have been explicitly defined for the purpose of the inverse planning and the actual critical structures and target volumes.

Field-in-Field Planning Field-in-field planning, another technique used to generate the effect of intensity-modulated fields but based on forward treatment planning, is used extensively at M. D. Anderson Cancer Center for planning radiation treatments of the breast.15 The basic principle behind field-in-field planning is that of starting with a simple beam configuration, computing the dose distribution from this simple configuration, adding smaller fields to fill in cold spots to make the dose distribution more uniform. In the specific example of field-in-field planning of the breast, the planning begins with a pair of opposed tangential open fields. After the doses are calculated, the beam weights are adjusted until any hot spots are balanced ­between medial and lateral fields. The doses are displayed as a percentage of a point dose (typically the isocenter dose). The planner then decides on an isodose curve, corresponding to a hot spot, to block. Figure 2-12A displays an isodose distribution of a treatment plan of the breast on a transverse slice after the beam weights have been adjusted to balance the hot spots. In this case, the 114% isodose is selected for blocking. The medial field is then copied, and the treatment portal is adjusted to cover the identified isodose cloud (see Fig. 2-12B). The dose is calculated for this new field, and the weighting of this field is increased until the isodose cloud disappears. Figure 2-12C illustrates the new dose distribution with the added field and the disappearance of the 114% isodose. The planner then selects a new isodose curve to be blocked, in this case the 109% isodose. The lateral field is copied and the treatment portal is adjusted to cover the selected isodose cloud (Fig. 2-12D). The dose is calculated, and the weighting of the lateral field is increased until the isodose cloud disappears. The selected fields are alternated, modified to block a selected hot spot, the dose

59

r­ ecalculated, and the beam weighted until the dose distribution is sufficiently uniform. Figure 2-12E compares the original dose distribution with the more uniform dose distribution obtained by this field-in-field approach.

Delivery of Photon Beam Radiation Therapy Achieving the goals of high-precision radiation therapy requires that the shape of the radiation field conform precisely to the target volume such that the target volume is irradiated but the surrounding uninvolved anatomy is shielded. The previous section examined how multiple photon beams can be combined to generate a high-dose volume that is shaped to track the outline of the tumor and how modifying the fluence of the incident beams can render the shaping of the high-dose region even more precise. The following section addresses the delivery of these high-precision photon fields.

Multileaf Collimation Collimation, or shaping, of photon beams has progressed from the use of hand-placed blocks to customized blocks cast from low-melting-point alloy to multileaf collimators (MLC), which represent the current state of practice for shaping static radiation treatment fields. The advantage of using MLCs for shaping fields is that no manual intervention is necessary to change the shape of the fields. Field shapes, as determined by MLC locations, are generated by the treatment planning system and transferred directly to the record-and-verify system. The appropriate settings are then delivered to the linear accelerator for beam delivery. Perhaps even more important, the MLC settings can also be verified and recorded. The ability to generate the MLC settings in a treatment plan and transfer the settings directly to the linear accelerator has made feasible programming a sequence of MLC settings that results in the delivery of a spatially dependent fluence pattern, which has given rise to the practical implementation of IMRT delivery. Earlier forms of treatment planning for IMRT gave rise to an optimal fluence map calculated as a distribution of weights of beamlets, which was the solution of the inverse planning problem. One disadvantage of this solution was that the fluence map had to be converted into a series of MLC settings that could deliver the radiation fluence in a practical amount of time. Several algorithms were ­developed that converted the fluence map to a deliverable MLC pattern, but rendering the fluence pattern practical often required degrading the quality of the IMRT plan, ­occasionally to unacceptable levels. Many current radiation treatment planning systems provide algorithms for direct aperture optimization, determining the MLC settings as the specific output of the inverse planning algorithm.

Fixed-Beam Attenuation IMRT delivered by modifying MLC settings during beam delivery is not modulation of intensity but rather modulation of fluence. True modulation of the radiation beam intensity can be achieved by the use of compensating filters placed in the beam path.16 Calculating the various thicknesses of a compensating filter to generate a desired fluence distribution is relatively straightforward, and the

60

PART 1  Principles

A

C

B

D

E FIGURE 2-12.  There are several steps in the development of a field-in-field plan for treatment of a breast tumor. A, Isodose distribution on a transverse CT slice after beam weights have been adjusted to balance hot spots. B, Treatment portal for the additional medial field, blocking the 114% isodose. C, Isodose distribution with additional medial field added with sufficient weight to eliminate the 114% isodose. D, Treatment portal for the additional lateral field, blocking the 109% isodose. E, Comparison of the original dose distribution (left) and dose distribution for field-in-field treatment (right).

2  Clinical Radiation Oncology Physics

61

thickness information can easily be transferred to a milling machine that can mill that shape of a compensating filter out of an appropriate molding material such as Styrofoam blocks. Fine shot or a low-melting-point alloy can be used to fill the mold. Because the mold itself does not attenuate the radiation beam significantly, it can remain in place during treatment delivery. Although relatively straightforward, this technique has several drawbacks. In addition to the time required to fabricate the compensating filters, a reliable quality assurance procedure is needed to ensure that the filters have been fabricated correctly. Moreover, therapist intervention is required to change filters after each treatment field delivery. Consequently, this technique has been superseded in most institutions by the use of multileaf collimation.

Helical Tomotherapy Helical tomotherapy, a technique developed by Mackie and colleagues,17 combines inverse planning with the delivery of a helical CT–like megavoltage beam. Temporally modulated slits open and close across the beam opening to generate an intensity-modulated treatment beam. Using this beam delivery technique in combination with inverse planning allows the high-dose region to conform to highly irregular arbitrary target volumes. The need to support continuous rotation of the beam delivery gantry required the development of novel slip-ring–based technologies. The tomotherapy beam can also be used to generate megavoltage CT images. As discussed further under “Three-Dimensional In-Room Volumetric Imaging,” images acquired just before treatment can be used as the basis for adjustments in the patient setup and modifications in the treatment plan based on interfractional changes in the patient anatomy, making the possibility of adaptive beam delivery feasible.

Quality Assurance of IntensityModulated Radiation Therapy Delivery Quality assurance of beam delivery is even more important in IMRT than in conventional radiation beam delivery. Because of the need for spatially varying fluence patterns, dose distributions generated from intensity-modulated radiation beam delivery are significantly more complex than those obtained by conventional beam delivery. However, with the increased need for accuracy, the dose-calculation models for IMRT, based on small beamlets, may not be quite as accurate as those for conventional beam delivery. Moreover, because radiation fields are small, the accuracy of dose measurement of these fields may not be as good as those for large fields. Quality assurance of the delivery of IMRT treatments is required in the United States for third-party reimbursement for this procedure. Explicitly stated, the “… plan distribution must be verified for positional accuracy based on dosimetric verification of the intensity map with verification of treatment setup and interpretation of verification methodology.”18 Two components of patient-specific quality assurance of IMRT delivery go beyond the quality assurance needed for conventional conformal radiation therapy. The first is the need to verify that the MLC settings that are designed to deliver the beam fluence that solves the inverse planning problem will deliver the planned dose distribution. The

FIGURE 2-13.  A test phantom is used for quality assurance in intensity-modulated radiation therapy. The yellow sheet is the film on which the dose distribution is recorded for comparison with the planned dose distribution.

second is the need to verify that the calculation of monitor units as obtained from the treatment planning system will deliver the correct dose to the patient. In quality assurance of conventional conformal radiation therapy, commissioning of the beam model and the treatment planning system is sufficient to provide the user with the confidence that the calculated dose distribution is correct,19 and the comparison of the monitor unit calculation with a point dose calculation is routinely used as a quality assurance check on the monitor unit calculation. For IMRT delivery, however, relative dose distribution and absolute dose delivery must be verified with measurement. At M. D. Anderson, the quality assurance procedure for IMRT delivery involves delivery of the planned irradiation to a special IMRT quality assurance test phantom (Fig. 2-13) that contains an ion chamber and film. The film is placed in a transverse plane in the phantom, the ionization chamber is placed in an orifice at the center of the phantom, and the quality assurance procedure is carried out as follows. The fluence distributions and monitor units that comprise the IMRT plan are transferred from the treatment plan to a CT model of the test phantom on the treatment planning computer, and the dose distribution is calculated on the test phantom model. The MLC settings and monitor unit settings, which have been transferred to the recordand-verify system, are used to deliver the IMRT plan to the test phantom. The relative dose distribution indicated on the film is compared with the planned dose distribution, and percent dose difference and distances to agreement are determined (Fig. 2-14). The dose as recorded on the ionization chamber is compared with the calculated dose to the point on the treatment plan corresponding to the location of the ionization chamber and reported on a form. The agreement between calculated and measured dose distributions is evaluated by using the gamma index proposed by Low and colleagues,20 in which the goal is 5% dose

62

PART 1  Principles

Measured 0

5

5

10

10 A/P distance (cm)

A/P distance (cm)

Calculated 0

15 20

15 20

25

25

30

30

35

35

A

0

5

10

15 20 25 Lateral distance (cm)

30

35

B

0

5

10

15 20 25 Lateral distance (cm)

30

35

Calculated (solid) and measured (normalized data, dashed)contours. 0 5

A/P distance (cm)

10 15 20 25 30 35

C

0

5

10

15 20 25 Lateral distance (cm)

30

35

FIGURE 2-14.  Dose distributions using the beam configuration for an intensity-modulated radiation therapy plan are calculated from a test phantom (A) or measured from the same fluence pattern and monitor units on the test phantom (B). In the comparison of the two dose distributions (C), solid lines indicate the calculated dose distribution, and dashed lines indicate the measured dose distribution.

agreement in the regions of low dose gradient and 3 mm distance to agreement in the regions of high dose ­gradient. Another technique for evaluating agreement between calculated and measured dose distributions for IMRT quality assurance has been proposed by Palta and colleagues,21 who divide the dose comparison into four ­regions, each of which has its own separate criteria. Since January 2004, the gamma index method of quality assurance has been used for more than 1000 patients at M. D. Anderson, and acceptable gamma values (Median (n = 77) 0

0 1

0

3

4

5

2

3

4

5

Years from randomization 100

100 80 Radiotherapy plus cetuximab

60 40 20

Radiotherapy

80 Radiotherapy plus cetuximab

60 40

Radiotherapy

20 0

0 0

E

1

0

D

Years from randomization

Overall survival (%)

Locoregional control (%)

C

2

10

20

30

40

50

60

70

Months

0

F

10

20

30

40

50

60

70

Months

FIGURE 5-2.  A and B, Immunohistochemical stains used for visualizing epidermal growth factor receptor (EGFR) in human head and neck squamous cell carcinoma specimens. (A, below median; B, above median). C and D, The relationship is shown for EGFR expression and locoregional relapse (C) or overall survival (D) rates for patients with locally advanced head and neck squamous cell carcinoma (HNSCC) treated with conventionally fractionated radiation therapy. E and F, Addition of cetuximab to radiation therapy results in significant improvements in locoregional control (E) and overall survival (F) rates for patients with locally advanced HNSCC. (A – D, From Ang KK, Berkey BA, Tu X, et al. Impact of epidermal growth factor receptor expression on survival and pattern of relapse in patients with advanced head and neck carcinoma. Cancer Res 2002;62:7350-7356; E and F, from Bonner JA, Harari PM, Giralt J, et al. Radiotherapy plus cetuximab for squamous-cell carcinoma of the head and neck. N Engl J Med 2006;354:567-578.)

Challenges for Anti-EGFR Therapy The findings from coordinated laboratory and clinical studies such as those described here generated considerable excitement; in addition to producing a new, less toxic option for therapy, these findings established the proof of principle that modulating a cellular signaling pathway that is expressed differently in tumors and in normal tissues can lead to selective sensitization of tumors to radiation therapy and hence improve the therapeutic index. Selective enhancement of tumor response has not been accomplished by combining radiation with traditional chemotherapeutic agents such as cisplatin, taxanes, or others.

The nonsurgical standard of care for locally advanced HNSCC was shifting from radiation therapy only to irradiation with concurrent chemotherapy (e.g., cisplatin) while the findings from the phase III cetuximab trial were maturing. Since then, additional trials have been ­undertaken to assess the feasibility of combining cetuximab with radiation therapy given with chemotherapy. In a phase II study led by the Memorial Sloan-Kettering Cancer Center, the combination of radiation therapy, cetuximab, and cisplatin was found to result in an overall survival rate of 76%, a progression-free survival rate of 56%, and a locoregional control rate of 71% at 3 years. Two early deaths were

122

PART 1  Principles

­attributed to pneumonia and unknown causes.57 This regimen is now being evaluated in a phase III trial led by the Radiation Therapy Oncology Group. Another challenge is that cetuximab has increased locoregional control and survival rates of patients with locally advanced HNSCC by only 10% to 15%, and more than 50% of patients given the combination of radiation therapy and cetuximab still developed locoregional relapse. Assays that can identify resistant tumors are urgently needed so that this rather expensive therapy can be given to those patients likely to benefit from it. Understanding the biologic basis of tumor resistance to EGFR antagonists is essential to extend this type of treatment to more patients and to develop alternative molecular-level treatment strategies for HNSCC. The search for predictive biomarkers for tumor response to EGFR antagonists has been rather disappointing. Contrary to expectations, the magnitude of tumor EGFR expression before treatment did not predict tumor response to EGFR antagonists in two clinical trials.58,59 Moreover, findings from colorectal cancer trials suggest that some patients whose tumors did not express EGFR (indicated by immunohistochemical staining) still benefited from cetuximab therapy.60,61 Notably, the assay methods used for correlative studies vary widely among investigators. In a standard immunohistochemical assay, for example, a cell is considered positive for EGFR if more than 30,000 receptors are present.62 Unfortunately, the number and perhaps the density of receptors required to confer a given biologic effect are poorly understood. This deficiency may account for some of the discrepancies among study findings. Several strategies for overcoming resistance to radiation sensitization by EGFR antagonists are being investigated. Cetuximab or EGFR tyrosine kinase inhibitors given as single-agent therapy in current dose regimens may not suppress EGFR-mediated signaling sufficiently to affect tumor survival. Other dose regimens, other antibodies and tyrosine kinase inhibitors, antisense nucleotides,63 and various combinations of these agents are being studied. Findings from a preclinical study, for example, showed that administering three additional doses of cetuximab after concurrent irradiation and cetuximab improved local tumor control compared with concurrent irradiation and cetuximab alone.64 Other antibodies being studied include hR3, which has a longer half-life and larger area under the curve than cetuximab, and ABX-EGF, which has greater affinity for EGFR. Another possibility being explored is that mutations in EGFR may result in aberrant signaling and diminished response to EGFR antagonists or ionizing radiation. For example, a form of EGFR with a truncated extracellular domain (EGFRvIII)27 detected in several tumors was different from wild-type EGFR (EGFRwt) in its preferential activation of signaling pathways.65,66 In one series, EGFRvIII expression was found in 14 (42%) of 33 tumors, and all 14 also expressed EGFRwt.67 The same study showed that transfecting HNSCC cell lines with EGFRvIII can ­reduce apoptosis in response to cisplatin or reduce growth inhibition by cetuximab, observations that form the basis for investigating the efficacy of an EGFRvIII-specific monoclonal antibody. Another possible explanation for resistance to EGFR inhibitors is that constitutive activation of signaling ­pathways

downstream of EGFR by mutation or upregulation of other ERBB family receptors or other receptor tyrosine kinase classes can promote cell survival (reviewed by ­Kalyankrishna and Grandis68). For example, high levels of activated AKT1 can occur downstream of EGFR inhibition through upstream activation of SRC, RAS, or mutated PTEN69; ­ amplification of the catalytic subunit of PI3K70; or loss of the phosphatase and tensin homologue tumor suppressor protein.71 Overexpression of six major ERBB family ligands can activate ERBB receptors such as ERBB2 and ERBB372 and the insulin-like growth factor receptor 1 (IGF1R),73 resulting in resistance to EGFR inhibition. Upregulation of IGF1R, for instance, was found to result in sustained signaling through the PI3K pathway, leading to potent antiapoptotic and proinvasion effects and resistance to an EGFR tyrosine kinase inhibitor in a glioblastoma model.74 In summary, discovery of the contribution of EGFR signaling in governing several key cellular regulatory processes, its perturbation in epithelial neoplasms, and its impact on tumor response to some therapeutic modalities led to the conception and completion of a clinical research program that produced a new front-line therapy option for locally advanced HNSCC. Although this pivotal phase III trial validated the concept that tumor cells could be selectively sensitized to radiation therapy through modulation of a perturbed signaling pathway, the clinical benefit resulting from this strategy has been rather modest, and many questions remained to be addressed. For example, the lack of correlation between the magnitude of EGFR expression and tumor response to EGFR antagonists is counterintuitive. We also need ways of identifying predictive biomarkers for response to EGFR inhibitors for clinical use and clarification of the mechanisms underlying the lack of cetuximab-mediated sensitization of most HNSCC tumors to radiation so that approaches to circumvent this obstacle can be formulated. This successful experience in translational research has led to many investigations of various combinations of radiation therapy with other novel agents and therapeutic strategies for HNSCC and other types of neoplasms. Although a comprehensive review of all such strategies is beyond the scope of this chapter, a few examples of other molecular-therapeutic strategies, including those based on tumor angiogenesis, are presented here.

Angiogenesis as a Target in Cancer Therapy Folkman first proposed that the growth and spread of malignancy depends on angiogenesis more than 3 decades ago.75 Since then, much work has been done in the field of angiogenesis, and it has become clear that the induction of adequate vasculature is critical for the growth of tumors, for the development of metastasis, and for several physiologic and pathophysiologic processes.76 Under physiologic con­ ditions that require neovascularization, angiogenesis is a tightly regulated process77 that begins with degradation of the basement membrane surrounding capillaries followed by stromal invasion by endothelial cells. Proliferation and migration of endothelial cells eventually lead to the formation of a network of new blood vessels as the

5  Principles of Combining Radiation Therapy and Molecular Therapeutics

invading endothelial cells organize into three-dimensional structures.78 During tumorigenesis, however, the process is sustained and requires the continued production of stimulators by tumor cells and stromal cells in excess of inhibitors. Several lines of evidence confirm the dependence of tumor growth on angiogenesis.76 A tumor mass that is smaller than about 0.4 mm in diameter can receive oxygen and nutrients by diffusion, but any further increase in tumor mass requires neovascularization.79 In malignant angiogenesis, new vessels are anomalous,80 characterized by a relative lack of pericytes and other support cells, impaired flow,81 the presence of plasma proteins that function as a neostroma,82 and increased vascular permeability. In addition to forming the new vessels that provide nutrients and oxygen and remove catabolites, ­ proliferating endothelial cells found in and around tumors produce several growth factors that can promote tumor cell growth, invasion, and survival.83-85 Angiogenesis therefore provides a perfusion effect and a paracrine effect for growing tumors,76 and tumor cells and endothelial cells can drive each other and perpetuate and amplify the malignant phenotype. These observations led Folkman to propose a two-compartment model of malignancy in which one compartment consists of endothelial cells and the other of tumor cells.86 Both compartments offer potential targets for anticancer strategies, and prudence dictates that the cancer cells and the angiogenic endothelium be targeted in such strategies.

Antivascular versus Antiangiogenic Therapy Vascular-targeted therapies consist of two broad types. They can prevent the formation of new blood vessels or revascularization, or they can disrupt the existing vasculature of a tumor. Antiangiogenic therapy is directed against the formation of neovasculature and does not significantly affect the resting vasculature. Antivascular therapy targets existing vasculature and can disrupt normal vessels. In practice, many agents, such as those that target vascular endothelial growth factor (VEGF), have both antiangiogenic and antivascular effects. One strategy to selectively target tumor vasculature is to use toxins conjugated to antibodies directed against antigens that are selectively expressed by tumor endothelial cells.87 Another approach involves using small-molecule agents that destabilize microtubulin, such as combretastatin A4 and ZD6126.88 Angiogenic endothelial cells can also be targeted by specific peptide sequences that selectively bind to proliferating endothelial cells but not to quiescent cells.89 Whether organ-specific peptides can home to sites of angiogenesis in selected tissues remains to be determined, but targeting the tumor vasculature is an attractive approach. Several studies have demonstrated enhanced antitumor effects from combining vascular-targeting agents with radiation therapy. In a study of KHT sarcoma xenografts, for example, ZD6126 administered 24 hours before, immediately after, or up to 4 hours after radiation therapy resulted in enhanced tumor cell killing compared with that achieved with irradiation alone, and that enhancement was optimal when the vascular-targeting agent was given within 1 hour of the radiation.90 ZD6126 was also effective when administered after fractionated radiation

123

therapy for A431 lung cancer xenografts,91 particularly when given with the anti-EGFR agent ZD1839. However, for U87 human glioblastoma xenografts in mice, ZD6126 given 1 hour before irradiation was associated with an acute increase in tumor-tissue hypoxia and faster tumor growth rates compared with irradiation alone. Giving this drug after irradiation did not affect hypoxia but did not enhance antitumor activity compared with radiation therapy alone.92 These findings seem to demonstrate the potential for antivascular agents to enhance tumor response to radiation therapy, but further studies are needed to optimize this approach. The acute disruption of tumor vasculature must be timed precisely to minimize the potential impairment of radiation effects by any ensuing hypoxia. Optimizing the timing and delivery of each modality may allow eradication of viable peripheral tumor tissue that may be resistant to the effects of vascular targeting agents. However, an antiangiogenic agent may also be required to maximize the endothelial apoptosis that can occur within hours of radiation therapy and to prevent the revascularization that can occur after radiation therapy.93

Radiation Effects on the Vasculature Radiation therapy can have systemic and local effects on angiogenesis, possibly through increasing the expression of proangiogenic factors such as VEGF.94 In the late 1940s, Kaplan and Murphy95 observed that some patients with epidermoid carcinoma of the lip developed widely metastatic disease after radiation therapy. Because that type of cancer rarely metastasizes in this way, they speculated that the irradiation might have been a contributing factor. To test this idea, they implanted mice subcutaneously with a mammary tumor cell line that rarely metastasizes to the lung, irradiated the mice, and found that more than 44% of the irradiated animals developed lung metastases compared with only 10% of nonirradiated controls. More than 20 years later, a report from Suit and colleagues96 observed that animals that developed disease recurrence at the primary tumor site after radiation therapy were more likely to develop metastatic disease. Later work with nude mice showed that the formation of recurrent tumors after radiation therapy was preceded by angiogenesis, most of which occurred within 20 days of the irradiation.97 Investigations of the mechanisms underlying these obser­ vations indicated that radiation therapy to a primary tumor can decrease the mobilization of angiogenesis inhibitors such as angiostatin; administration of an angiogenesis inhibitor can slow the formation of metastases in the lungs of mice after irradiation of primary Lewis lung carcinomas.98 These findings suggest that adding antiangiogenic therapy to radiation treatment may help to restore the balance between endogenous angiogenic and antiangiogenic factors so as to inhibit locoregional and metastatic tumor growth. Under some circumstances, radiation therapy can also impair angiogenesis. In studies of tumor growth delays caused by the irradiation of stroma, a phenomenon known as the tumor bed effect, Milas and colleagues99 confirmed that tumors grew more slowly in mice that received radiation to the tumor bed before tumor cell implantation; however, they later found an increased incidence of ­metastasis among those mice and concluded that a tumor’s angiogenic

124

PART 1  Principles

potential was an important factor in the development of the tumor bed effect.100 Evidence that suppression of angiogenesis by radiation therapy is not restricted to the primary tumor site comes from another study101 showing that irradiation of primary tumors in rodents was associated with inhibition of angiogenesis at a distant site (the cranial window). Moreover, plasma levels of the angiogenic agent endostatin in the irradiated mice were twice those found in the nonirradiated mice. Further study is needed to better define the dynamic interactions between the vasculature and irradiation.

Angiogenesis Inhibitors Several antiangiogenic agents have been described,76,102,103 and some have been approved or are awaiting approval for clinical use in the treatment of cancer and other angiogenesis-dependent diseases. However, given the complexity of the angiogenic process in various disease states, a better understanding of the process of angiogenesis and the interactions between endothelial cells and their microenvironment is necessary before antiangiogenic agents can reach their full potential for the treatment of such diseases in humans. Early experience with antiangiogenic agents in the treatment of human cancer, particularly as monotherapy, revealed some important limitations. An example is the delayed onset of activity of some antiangiogenic agents,104 which can allow significant tumor progression to occur before they become effective. Murine tumors tend to proliferate several times faster than tumors in humans, and this delay in onset of activity may translate to several months in patients. In a veterinary study of the use of the antiangiogenic thrombospondin peptide mimetics ABT-526 and ABT-510 to treat naturally occurring canine malignancies, objective tumor responses were observed only after at least 60 days of therapy.105 Therapy with the antiangiogenic cytokine interferon-α for life-­threatening pediatric hemangiomas106 or giant cell tumors107 ­required several months of continued therapy before significant clinical responses were observed. The delays in the onset of activity may make the use of these agents impractical for patients with locally advanced or metastatic disease. Perhaps a more significant limitation of antiangiogenic monotherapy is the inability of these agents to completely eradicate disease even after prolonged administration at high doses. For most angiogenesis inhibitors, tumor growth generally resumes within a short time after cessation of therapy.104 Although prolonged cyclical therapy with endostatin could induce sustained dormancy in a wide variety of preclinical tumor models that persisted after therapy had been stopped, residual microscopic disease was still found in all of the treated mice.108 Immunohistochemical staining of these residual microscopic tumors revealed little or no evidence of angiogenesis, but it did show a high proliferative index of the tumor cells that was offset by a high apoptotic index. These findings suggest that it may not be feasible or practical to produce the amounts of antiangiogenic agents that would be required for widespread prolonged use. As an alternative approach, combinations of angiogenesis inhibitors that target different angiogenic pathways may be more effective than monotherapy; a related strategy that

may be especially beneficial is to combine direct inhibitors of angiogenesis with agents that target stimulators. Experience with angiogenesis, like that with other molecularly targeted therapies, suggest that strategies that target each of the interdependent steps involved in the angiogenic cascade, perhaps in combination with other treatment modalities, will be necessary to effectively inhibit pathogenic angiogenesis. Strategies that target growth factor receptor tyrosine kinase signaling pathways involved in tumor angiogenesis, growth, and survival offer great potential for antitumor therapy. The most widely studied pathways are those involving VEGF and its receptors. VEGF induces vascular permeability and plays a pivotal role in vasculogenesis and angiogenesis.109,110 Although the effects of VEGFR signaling were initially thought to be specific to the vasculature, VEGF also participates in several other processes.112-115 In pregnant women, the presence of endogenous regulators of VEGF in the circulation has been linked with the risk of hypertension, proteinuria, and preeclampsia, effects thought to result from endothelial dysfunction.111 Hypertension and proteinuria have also been reported in patients with cancer who were treated with VEGF-signaling inhibitors.116,117 Other findings suggest that VEGFR1 expression by colon cancer cells contributes to the motility and invasiveness of those cells but has little direct effect on their proliferation.118 Several approaches to block VEGF activity are being evaluated, including anti-VEGF or VEGFR antibodies and VEGFR tyrosine kinase inhibitors.110 Bevacizumab (Avastin), a humanized antibody against VEGF, has shown great promise for extending survival among patients with metastatic colorectal or lung cancer when used in combination with chemotherapy.119,120 Nevertheless, there is ample room for improving the efficacy of this and other anti-VEGF agents beyond merely extending survival. Side effects are also an issue; an early clinical trial of multimodality therapy that included SU5416 (semaxanib), another VEGF inhibitor, linked its use with thromboembolic and hemorrhagic complications.121 A phase III clinical trial of bevacizumab with chemotherapy for women with previously treated breast cancer showed that the combination of bevacizumab and capecitabine increased response rates but did not improve progression-free or overall survival.122 The addition of PTK787 (vatalanib), a small-molecule inhibitor of VEGFR tyrosine kinase activation, to the chemotherapy regimen FOLFOX4 (5-fluorouracil, leucovorin, and oxaliplatin) did not improve overall survival among previously treated patients with metastatic colorectal cancer compared with chemotherapy only,123 although a ­ progression-free survival benefit might have been observed in a subgroup of patients with elevated lactate dehydrogenase levels.124 Therapy with these agents may have to be optimized for each patient. A further complication in anti-VEGF/VEGFR therapy is that the expression of receptor tyrosine kinases varies markedly among different tumor types, and most advanced malignant tumors produce several angiogenic factors.125 The ability to screen tumor specimens for the expression of angiogenesis ligands and receptors will be critical for appropriately tailoring therapy with specific protein tyrosine kinase inhibitors to individual patients. Many receptor tyrosine kinases have direct and indirect effects on the

5  Principles of Combining Radiation Therapy and Molecular Therapeutics

viability of tumor cells and tumor-associated endothelial cells. For example, in a mouse model of pancreatic cancer, blockade of EGFR led to apoptosis and tumor regression by targeting tumor cells and tumor-associated endothelial cells.126 The blockade of EGFRs works primarily by downregulating the expression of proangiogenic factors by tumor and stromal cells. However, antagonists of EGF and its receptor may also have direct effects on endothelial cells, because the expression of EGF or TGF-α by tumor cells can lead to increased expression of EGFR by endothelial cells in vivo.127 Another proangiogenic growth factor, platelet-derived growth factor (PDGF), may also have indirect effects on the tumor vasculature by increasing the expression of proangiogenic molecules such as VEGF and basic fibroblast growth factor (bFGF)128 and direct effects on endothelial cells by increasing their expression of PDGF receptor (PDGFR) in response to PDGF production by tumor cells.129 These agents offer the potential for indirect and direct effects on the tumor cell and endothelial cell compartments. An alternative to using multiple angiogenesis inhibitors to target multiple growth factors would be to design small molecules that can target several factors. Agents that target VEGFR and other receptor tyrosine kinases (e.g., EGFR, FGFR, PDGFR) are being developed to more broadly, but perhaps less selectively, target cancer growth.130 These agents are being studied as a component of combined­modality therapy for a wide variety of malignancies. One such agent is ZD6474 (vandetanib), an orally available receptor tyrosine kinase inhibitor that selectively targets VEGFR2 and, to a lesser degree, EGFR, with resultant prevention of tumor growth and angiogenesis.131 Another agent, SU11248 (sunitinib), targets VEGFRs, PDGFR, KIT, and FLT3 and has shown significant efficacy in preclinical models of cancer and in clinical trials of renal cancer,132 but it was also associated with significant toxicity that prevented continuous administration. Sorafenib (Nexavan), an orally available inhibitor of RAF1 kinase, PDGFR,VEGFR2, VEGFR3, and KIT, has improved progression-free survival for patients with renal cell carcinoma,133 but like sunitinib, it has been associated with significant toxic effects, including diarrhea, rash, fatigue, hypertension, and cardiac ­ischemia.

Combining Antiangiogenic Agents with Radiation Therapy or Radiation and Chemotherapy The difference in their toxicity profiles makes the concept of combining antiangiogenic agents with irradiation or chemotherapy appealing. However, additional work is needed to identify feasible combinations, to optimize them, and to identify potential problems that may arise from their use. Information available from preclinical and clinical studies is presented in the following paragraphs.

Preclinical Findings Several clinical trials have been designed to study the interaction of chemotherapy and antiangiogenic therapy, but relatively few trials have tested radiation therapy with

125

antiangiogenic therapy. The long-standing assumption was that angiogenesis inhibitors would impair the effect of ionizing radiation by inducing tumor hypoxia. However, in the mid-1990s, Teicher and colleagues134 observed that antiangiogenic therapy given in combination with the angiogenesis inhibitors TNP-470 and minocycline, a weak inhibitor of metalloproteinase, improved tumor oxygenation and enhanced the antitumor effect of radia­ tion therapy in mice. Subjecting mice with primary tumors known to produce proangiogenic factors to curative radiation therapy can lead to the rapid growth of previously dormant metastases that can kill the animals within 18 days after the completion of therapy.98 However, in the same study, the systemic administration of recombinant angiostatin, an angiogenesis inhibitor, prevented the growth of the metastases after irradiation of the primary tumor.98 The findings from these preclinical studies suggest that at least some patients can benefit from the addition of antiangiogenic therapy during or after radiation therapy to prevent the growth of distant metastases. However, the timing and duration of antiangiogenic therapy required to yield such an effect remain unknown. To clarify the possible involvement of tumor oxygenation in the observed benefit from combining antiangiogenic agents with radiation therapy, the Jain laboratory135 assessed tumor hypoxia by measuring the partial pressures of oxygen within hormone-dependent male mouse mammary carcinomas (intratumoral Po2) in mice during therapy with anti-VEGFR-2 antibodies. They found that ­intratumoral Po2 levels initially decreased, indicating hypoxia and vessel regression, but it then subsequently increased, suggesting a second wave of oxygenation and angiogenesis that would enhance the effectiveness of oxygen-dependent treatments such as radiation therapy.135 However, results of another study by Murata and colleagues136 indicated that concurrent treatment of mouse breast carcinoma xenografts with TNP-470 and fractionated radiation therapy led to decreases in tumor oxygenation and corresponding decreases in tumor control. Whether antiangiogenic therapy enhances or impedes the effect of radiation therapy may depend on the tumor microenvironment. To explore this idea, Lund and others137 implanted glioblastoma multiforme xenografts in the thighs or the craniums of mice and then treated the mice with TNP-470 or radiation, or both. In the mice with the thigh tumors, the combined-modality treatment significantly enhanced the antitumor effect of each modality; however, no such enhancement was observed in the mice with intracranial tumors. The reason for this difference was not determined, but it is tempting to speculate that differences in the capillary beds and microenvironment of the brain compared with those in the subcutaneous tissues of the thigh might have contributed to the differences in response. Further evidence of potential antitumor benefit from using antiangiogenic agents with radiation therapy comes from members of the Weichselbaum laboratory,138 who demonstrated synergistic effects from the addition of the specific antiangiogenic factor angiostatin to irradiation for a variety of transplantable tumors in mice; that study involved low doses of both modalities and large, established tumors. Other agents with antiangiogenic activity have been tested in combination with concurrent radiation therapy in preclinical models. Milas and colleagues139 found that a

126

PART 1  Principles

s­ elective inhibitor of cyclooxygenase 2 (COX-2) that also has antiangiogenic properties enhanced the antitumor effects of radiation therapy; those findings are ­discussed further under “Cyclooxygenases as Targets in Cancer Therapy.” Anti-VEGF antibodies also have enhanced the response to radiation therapy in mice bearing a variety of human tumor cell types.94 Investigators at Vanderbilt University found that blocking the VEGFR2 receptor on tumor-associated endothelium eliminated the radioresistance phenotype of glioblastoma multiforme and melanoma xenografts in vivo.140 That same group later found that AZD2171, a potent inhibitor of several receptor tyrosine kinases (including all three VEGFRs), enhanced the antitumor effects of radiation therapy and blocked the proliferation of H460 lung cancer xenografts in mice.141 The first published comparison of different sequencing of radiation therapy and antiangiogenic therapy142 showed that concurrent administration of radiation and angiostatin was more effective in a mouse model of Lewis lung carcinoma than was radiation therapy followed by antiangiogenic therapy. Later studies of another agent showed that the VEGFR antagonist PTK787 enhanced the radiation response in a human squamous cell carcinoma mouse model, but only when it was administered after radiation therapy.143 A similar effect was observed when the VEGFR and EGFR antagonist ZD6474 (vandetanib) was used in combination with radiation therapy; the most profound enhancement of radiation response was observed when the drug was given immediately after fractionated radiation therapy.144 Results of a study of an orthotopic xenograft model of human lung adenocarcinoma (H441) showed that vandetanib was better than paclitaxel in suppressing angiogenesis, diminishing metastatic tumor spread through the pleural cavity, and enhancing the effects of radiation on tumor vasculature and growth.145 Winkler and colleagues146 systematically evaluated several treatment schedules for the combination of DC101 (a VEGFR2-specific monoclonal antibody) and radiation therapy for orthotopic glioblastoma xenografts in mice. Findings from that study indicated that VEGFR2 blockade temporarily normalized tumor blood vessels (defined in terms of changes in vessel coverage by pericytes and basement membrane, found to be regulated by angiopoietin 1), resulting in improved tumor oxygenation and an enhanced response to radiation therapy. However, the period of normalization and the window of opportunity to exploit it were relatively brief, suggesting that these findings may be difficult to translate into clinical practice.

Clinical Findings Clinical studies testing the combination of radiation therapy with antiangiogenic agents have not gained momentum until recently because of the toxicity of some of the available agents and concerns that this type of treatment may induce tumor hypoxia and tumor resistance. Side effects of VEGF inhibitors observed in clinical trials have included hemorrhage, which has occasionally been fatal in patients with lung cancer,147 and thrombosis.121,148 Willett and colleagues149 designed a phase I study in which patients with locally advanced rectal adenocarcinoma were treated with bevacizumab for 2 weeks, followed by bevacizumab with 5-fluorouracil, external beam radiation therapy to the pelvis, and surgery 7 weeks later. The study design included the analysis of several noninvasive and

invasive surrogates for response to the combined therapy. Although the investigators concluded that the combination of bevacizumab with chemoradiation therapy was safe and potentially effective, subsequent studies have demons­ trated increased toxicity on escalation of the bevacizumab dose.150 In another study of patients with pancreatic cancer, concurrent bevacizumab did not significantly increase the acute toxicity of capecitabine-based chemoradiation therapy, and it might have enhanced antitumor efficacy.151 However, ulceration and bleeding occurred in the radiation field when the tumor involved the duodenal mucosa. The encouraging preclinical results for vandetanib summarized in the previous section have led to this agent being studied in several clinical protocols to evaluate its safety, efficacy, and optimal sequencing with other treatment modalities.131,144 In summary, the early clinical experience with antiangiogenic therapy combined with radiation therapy or chemoradiation therapy is sufficiently encouraging to warrant additional clinical trials, although they face several challenges. Antiangiogenic therapy needs to be integrated with existing and emerging anticancer therapies, and it may need to be optimized on a patient-specific basis. However, conventional strategies for monitoring the effectiveness of anticancer therapies may not apply to antiangiogenic agents, and novel surrogates of response must be developed and validated. Angiogenesis profiles based on circulating levels of proangiogenic and antiangiogenic factors; analysis of tumors before and after therapy for evidence of tumor and endothelial cell proliferation, apoptosis, and microvessel density; and means of imaging tumor blood flow, blood volume, and metabolism need to be developed further. The design of future clinical trials should include measures of safety and efficacy and a means of validating invasive and noninvasive surrogates of response.

Cyclooxygenases as Targets in Cancer Therapy Expression and Function The COX-1 and COX-2 enzymes mediate the production of prostaglandins from arachidonic acid (Fig. 5-3).152-154 COX-1 is a ubiquitous enzyme expressed constitutively in normal tissues; COX-2 is an inducible enzyme mainly expressed in pathologic states, primarily inflammatory processes and tumors. COX-2 is induced by several factors, including proinflammatory cytokines such as TNF-α, interleukin-1β, and platelet activity factors; growth factors such as EGF, PDGF, bFGF, and TGF-β; mitogenic substances; hypoxia; and oncogenes. In normal tissues, prostaglandins regulate various homeostatic physiologic functions such as vasomotility, platelet aggregation, and protection from different types of injury, including that inflicted by radiation. In tumors, the presence of COX-2, whether induced by prostaglandins or by mechanisms independent of prostaglandins, is conducive to progressive growth, invasion, and metastasis, and it has been linked with resistance to cytotoxic therapy. COX-2 is overexpressed in premalignant lesions and established tumors (reviewed by Liao and others154). A causal link between COX-2 and carcinogenesis has been demonstrated experimentally in studies of transgenic

5  Principles of Combining Radiation Therapy and Molecular Therapeutics Membrane phospholipids Phospholipase A2 Arachidonic acid

PGG2 NSAIDs

COX-1

NSAIDs COX-2 Cox-2 inhibitors

PGH2

Tissue specific isomerases

PGE2

PGF2α

PGD2

COX-1 • Ubiquitous • Normal tissue • Various physiological functions: – GI tract integrity – Vasodilation, vasoconstriction – Platelet aggregation – Kidney function

PGI2

TXA2

COX-2 • Inducible • Inflammatory states • Cancer

FIGURE 5-3.  Prostanoids are synthesized from arachidonic acid with the involvement of cyclooxygenase 1 and 2 (COX-1 and COX-2). Phospholipase A2 generates free arachidonic acid from membrane phospholipids. COX-1 or COX-2 activity converts arachidonic acid to prostaglandin (PG) G2, followed by conversion to PGH2 into other prostaglandin isoforms or thromboxane (TX) A2. Nonsteroidal anti-inflammatory drugs (NSAIDs) inhibit COX-1 and COX-2, whereas specific COX-2 inhibitors inhibit COX-2 only. (From Liao Z, Milas L. COX-2 and its inhibition as a molecular target in the prevention and treatment of lung cancer. Expert Rev Anticancer Ther 2004;4:543-560.)

mice lacking one or both alleles of COX-2. For example, the presence of COX-2 was found to be associated with higher numbers of premalignant colon cancer lesions in intestines155 and a higher incidence of premalignant breast cancer lesions in mammary glands.156 Conversely, treating rats157 or mice158 with selective COX-2 inhibitors, notably celecoxib, reduced the development of malignant lesions in those animals. COX-2 has been detected in a broad range of tumor types, including head and neck, lung, colon, breast, and pancreatic cancer.154,159,160 Upregulation of COX-2 also has been found in some stromal cells that infiltrate tumors, in endothelial cells of tumor vasculature, and even in nonneoplastic epithelium adjacent to the tumors. Increasing evidence suggests that the presence of COX-2 is associated with an increased propensity for metastatic spread and poor prognosis,149,159,160 as well as a poor response to chemotherapy or radiation therapy.154 In one study involving analysis of tissue samples of cervical carcinoma from women given radiation therapy, the absence of COX-2

127

e­ xpression was associated with better 5-year overall survival rates (75% versus 35% for those whose tumors expressed COX-2).161 COX-2 influences carcinogenesis, tumor growth, and resistance to therapy through several mechanisms, including stimulation of cell growth, angiogenesis, invasion, ­metastasis, resistance to apoptosis, and suppression of the immunosurveillance system.152,154 The radioprotective effect of prostaglandin E2, a major metabolite of COX-2, may be an additional mechanism by which COX-2–expressing tumors resist radiation therapy.153

COX-2 Inhibitors Nonsteroidal anti-inflammatory drugs (NSAIDs) have broad inhibitory effects on COX enzymes. Epidemiologic findings have suggested that nonspecific NSAIDs and more selective COX-2 inhibitors such as celecoxib can prevent the development of colorectal, esophageal, and lung cancer.154,162-167 Preclinical data have shown that selective COX-2 inhibitors can enhance tumor response to radiation and chemotherapeutic agents.152,154,168,169 The ability of the selective COX-2 inhibitor SC-236 to enhance the radiosensitivity of a murine NFSa fibrosarcoma, for example, is illustrated in Figure 5-4. Although this agent by itself had only minor effects on tumor growth, it greatly enhanced the response of the tumors to irradiation, as assessed by regrowth delay and tumor control assays.170 COX-2 inhibitors seem to enhance tumor radioresponse through several mechanisms, including direct ­interaction with tumor cells, thereby enhancing their ­radiosensitivity,171-174 and indirect actions by modulating the tumor cell microenvironment, primarily by suppressing angiogenesis and stimulating immune functions.152,154 In vitro studies showed that COX-2 inhibitors increased cellular radiosensitivity by inhibiting DNA repair,173,174 by increasing susceptibility to radiation-induced apoptosis,172 and by arresting cells in the radiosensitive phases of the cell cycle.173 Inspired by these preclinical findings, investigators have undertaken several clinical trials in which COX-2 inhibitors are combined with radiation therapy or chemotherapy, or both. One such study, a phase I trial done at The University of Texas M. D. Anderson Cancer Center, was designed to determine the maximum tolerated dose of celecoxib when combined with radiation therapy for patients with inoperable non–small cell lung cancer (NSCLC) and poor performance status.175 This trial included patients with locally advanced cancer with obstructive pneumonia or minimal metastatic disease; those with early-stage tumors who could not undergo surgery for medical reasons; and those who had been given induction chemotherapy before radiation therapy. Fortyseven patients received 45 to 66 Gy in 15 to 35 fractions combined with daily celecoxib doses of 200 to 800 mg. The local progression-free survival rate at 20 months was highly encouraging (66%) and similar to that observed after concurrent radiation therapy plus chemotherapy. Toxic effects included grade 1 or 2 nausea and esophagitis in 23 (49%) of 47 patients, grade 3 pneumonitis in two patients, and hypertension that did not resolve after discontinuation of celecoxib in one patient (on the 400-mg twice-daily regimen). Other trials have also shown encouraging results.154 Unfortunately, the discovery176 of

128

Growth delay (days)

PART 1  Principles

Tumor size (mm)

12

10

60 50 40 30 20 10 0

Cyclins and Cyclin-Dependent Kinases as Targets in Cancer Therapy Expression and Function

30 40 50 Radiation dose (Gy)

8

6

4 0

5

10

A

15

20

25

30

35

40

45

50

Days after irradiation 100

Tumor cure (%)

80 60 40 20 0 20

B

30

40

50

60

70

80

Radiation dose (Gy)

FIGURE 5-4.  The effects of the cyclooxygenase 2 (COX-2) inhibitor SC-236 on murine NFSa cells are shown for tumor growth delay (A) and tumor cure (B) after irradiation. Mice with NSFa fibrosarcoma cells implanted in the legs were treated with SC-236 (5 mg/kg) in their drinking water for 10 days, beginning when the tumors were 6 mm in diameter, and they were irradiated (30 Gy) when the tumors reached 8 mm in diameter. A, The graph shows the curves for the vehicle (red circles), SC-236 (red triangles), 30 Gy (blue circles), and SC-236 plus 30 Gy (blue triangles). Vertical bars represent the standard error. Inset, Tumor growth delay is plotted as a function of radiation dose, as shown by the curves for radiation only (red circles) and SC-236 plus radiation (blue circles). B, Tumor cure is plotted as a function of radiation dose, as shown by the curves for radiation only (red circles) and SC236 plus radiation (blue triangles). Horizontal bars represent 95% confidence intervals at the radiation dose level yielding 50% tumor cure (TCD50). (From Milas L, Kishi K, Hunter N, et al. Milas L, Kishi K, Hunter N J Natl Cancer Inst 1999;91:15011504.)

increased cardiovascular toxicity after long-term use of COX-2 inhibitors for chemoprevention of colorectal adenoma in 2005 prematurely dampened the enthusiasm of the industrial sector for further developing celecoxib as anticancer therapy. However, it is worth stressing that the benefit-risk ratio for COX-2 inhibitors may be considered favorable for patients with cancer.

Cyclin-dependent kinases (CDKs) are serine/threonine kinases that become active when associated with regulatory proteins, primarily cyclins. Nine CDKs have been identified in humans. Cyclin D in complex with CDK4 or CDK6 facilitates phosphorylation of the retinoblastoma protein (RB1) and enhances the transcription of E2F, which is important for the progression of cells through the G1/S checkpoint of the cell cycle. This signaling pathway is often dysregulated in neoplastic cells through overexpression of cyclins, downregulation of CDK inhibitors (e.g., CDKN2A [formerly designated p16], CDKN1A [formerly p21], IFI27 [formerly p27], CDKN1C [formerly p57]), or mutations in CDKs.177-179 Upregulation and overexpression of cyclin D1, for example, are common in Barrett’s esophagus and overt esophageal adenocarcinoma; moreover, cyclin D1 overexpression is associated with increased cancer risk from Barrett’s esophagus and with higher cancer stage.180,181 Alterations in the cell cycle cascade involving cyclin D, CDK4, CDK6, INK4, RB1, or E2F have been observed in more than 80% of malignant tumors in humans.178 Abnormalities in other cyclins (e.g., A, E), CDKs (e.g., 1, 2, 7), and the CDK inhibitors CDKN1A (p21) and IFI27 (p27) have been found in many types of cancer.182,183 RB1 can become inactivated through gene mutation, overexpression of cyclins or CDKs, or decreases in CDK inhibitors.183 Loss of the IFI27 (p27) protein has been associated with poor outcome in patients with lung, breast, colon, stomach, and prostate carcinomas, and loss of the CDKN2A (p16) protein with poor outcome in patients with lung carcinoma or melanoma.184 Increased expression of cyclins D and E has been associated with poor prognosis for several types of tumors.181,185 Several lines of evidence have implicated increased cyclin D1 expression with decreased tumor response to cytotoxic agents. Milas and colleagues33 demonstrated strong correlations between the level of cyclin D1 expression in tumors and the proliferation (according to labeling of proliferating cell nuclear antigen) and radioresistance of those tumor cells (according to a TCD50 assay). Reducing cyclin D expression by means of an antisense construct has been shown to inhibit tumor growth, reduce tumorigenicity, or induce cell death in various types of cancer.186-189 Suppression of cyclin D1 expression has been shown to enhance the sensitivity of several human gastrointestinal cancer cell lines to cisplatin and 5-­fluorouracil.186,187

CDK Inhibitors in Cancer Therapy Several CDK inhibitors have been investigated during the past decade; some modulate CDK activity directly and oth­ ers perturb CDK function indirectly by affecting the CDK regulatory proteins.178,183 One such inhibitor, flavopiridol [5,7-dihydroxy-8-(4-N-methyl-2-hydroxypyridyl)-6′chloroflavone hydrochloride], is a synthetic flavone that inhibits several phosphokinases in a dose-dependent manner. At lower doses, it inhibits primarily CDKs (CDK1 to CDK9), and at higher concentrations, it inhibits receptor tyrosine kinases, receptor-associated tyrosine kinases, and

5  Principles of Combining Radiation Therapy and Molecular Therapeutics

signal-transducing kinases.190 Flavopiridol can also repress transcriptional activity, possibly by inhibiting the activi­ ty of CDK9-dependent transcription elongation factor b (P-TEFb), which led to downregulation of several relevant genes, including cyclin D1, VEGF, and the DNA-repair protein KU70 (now designated XRCC6).190,191 At the cellular level, flavopiridol has shown cytostatic (cell cycle arrest) and cytotoxic (inducing apoptosis) effects.178,183,190,191 In various preclinical models, flavopiridol by itself led to modest tumor growth delays,192,193 but it interacted synergistically with several chemotherapeutic agents (e.g., paclitaxel, gemcitabine) in inducing tumor growth delay and apoptosis.192,193 Investigations at the M. D. Anderson Cancer Center demonstrated that flavopiridol potently enhanced radioresponse in cell culture systems191 and in vivo in mouse and human tumor models.192,194 For example, flavopiridol in combination with fractionated irradiation substantially suppressed the growth of MCa-29 murine adenocarcinoma tumor, as measured by assays of tumor growth delay and tumor curability (Fig. 5-5). Clinical studies of flavopiridol based on these preclini­ cal findings have yielded mixed results. An early phase I study showed only modest activity when flavopiridol alone

MCa 29 16

Tumor size (mm)

14 12 10 8 6 4 0

10

20

30

40

50

60

70

Days after initial radiation

FIGURE 5-5.  Effects are plotted for the response of MCa-29 adenocarcinoma tumors in mice treated with flavopiridol or fractionated radiation, or both. Mice bearing 8-mm tumors in the right hind legs were untreated (red circles) or treated with flavopiridol (blue circles), local tumor irradiation (red squares), or a combination of flavopiridol and irradiation (blue squares). Radiation was delivered in 4-Gy fractions daily for 20 consecutive days; flavopiridol was given at a dose of 2.5 mg/kg intraperitoneally twice daily for 10 days. When the agents were combined, flavopiridol was given immediately after each dose of fractionated radiation and given 6 to 8 h later for a total of  20 days. Each data point represents the mean size of eight or nine tumors; bars represent the standard error. Three of eight tumors treated with flavopiridol plus fractionated radiation permanently regressed and were excluded from the regrowth analysis. The dotted horizontal line represents leg thickness of 5 mm, below which it is not possible to measure tumor size accurately. (From Mason KA, Hunter NR, Raju U, et al. Flavopiridol increases therapeutic ratio of radiotherapy by preferentially enhancing tumor ­radioresponse. Int J Radiat Oncol Biol Phys 2004;59:1181-1189.)

129

was administered in a 72-hour continuous infusion every 14 days at a dose of 50 mg/m2/day,195 but more encouraging results were subsequently obtained with higher bolus doses.196 No results are available regarding the effects of combinations of flavopiridol and radiation, with or without chemotherapy.

RAS Signaling Pathway as a Target in Cancer Therapy The RAS oncogene is abnormally activated, usually by point mutations, in many types of cancer in humans. Much preclinical evidence linking activation of RAS with a radioresistant phenotype has led to its identification as a molecular target for restoring radiosensitivity in human tumors. In cultured cell models, for example, transformation of normal rodent fibroblasts with a mutated RAS construct confers radioresistance. One approach to circumventing the radioresistance conferred by RAS activation has involved inhibitors of farnesyltransferase, an enzyme that blocks the posttranslational farnesylation of RAS normally needed to mediate its downstream signaling. Consequently, inhibitors of the farnesyltransferase enzyme can block RAS function and, as shown in preclinical studies, can restore radiosensitivity to cells with mutated RAS genes.197 Two farnesyltransferase inhibitors have been tested in combination with radiation therapy in clinical trials. Investigators at the University of Pennsylvania have conducted phase I trials of the farnesyltransferase inhibitor L-778123 and radiation therapy for locally advanced NSCLC, for head and neck cancer,198 and for pancreatic cancer.199 Although both studies identified dose levels associated with acceptable toxicity and some responses, further clinical development of L-778123 has not been pursued by the manufacturer. The Radiation Therapy Oncology Group is conducting a phase II trial of another farnesyltransferase inhibitor, R115777 (tipifarnib), in combination with radiation therapy plus chemotherapy for the treatment of pancreatic cancer200; this compound is also being examined in combination with radiation therapy for glioblastoma multiforme.201

Gene Therapy Gene therapy continues to be an attractive strategy for treating cancer, and several methods are being tested in clinical trials. Although many studies have shown that viral vectors can effectively deliver therapeutic genes to some proportion of the cells in a solid tumor with mini­ mal systemic toxicity, the overall effectiveness of these approaches when used alone has been disappointing. Strategies in which gene therapy is combined with con­ ventional irradiation or chemotherapy are hoped to be more effective and are being investigated. Combinations of gene therapy with radiation therapy are usually in one of four categories: delivery of radiosensitizing drugs or agents, introduction of one or more replacement genes, use of oncolytic viruses, or suppression of the activity of genes that confer radioresistance.202 These approaches, the first of which has been explored to the greatest extent, are discussed further subsequently.

130

PART 1  Principles

Delivery of Radiosensitizing Drugs

Gene Replacement

The most thoroughly studied strategy for using gene therapy to deliver radiosensitizing drugs has been the use of replication-incompetent adenoviral vectors engineered to express the herpes simplex virus thymidine kinase (HSV-TK) gene.203 Preclinical studies have confirmed the antitumor activity of this approach in several tumor models. In its clinical application, patients are given the vector (in this case, Ad-TK), followed by systemic treatment with ganciclovir, a prodrug that when phosphorylated by the HSV-TK gene product, ultimately becomes a potent inhibitor of mammalian DNA polymerase and kills in­ fected cells by preventing DNA synthesis. This effect may also radiosensitize affected tumor cells through com­ promising their ability to repair radiation-induced DNA double-strand breaks. The safety of this approach has been demonstrated in a phase I/II clinical trial for men with prostate cancer,204 but additional trials are needed to demonstrate its efficacy. A similar strategy involves a more complex, ­replicationcompetent vector that encodes genes for HSV-TK and a prokaryotic cytosine deaminase.205 Cytosine deaminase, when expressed in mammalian cells, deaminates the relatively nontoxic prodrug 5-fluorocytosine, thereby creating the more toxic drug 5-fluorouracil. The advantages of this approach are that the toxic effects of the 5-­fluorouracil are confined to the infected tumor cells and that the 5­fluorouracil has its own radiosensitizing effects. In this “double suicide” gene therapy approach, patients are treated with the vector Ad-CD/TKrep in combination with 5fluorocytosine and valganciclovir. Because Ad-CD/TKrep is a replication-­competent vector, it also has oncolytic properties, an attribute that is discussed later under “Oncolytic Viruses.” The safety of this double-suicide gene therapy in combination with radiation therapy has been demonstrated in phase I testing for men with prostate cancer.206 Radiosensitizing agents other than drugs can also be delivered by gene therapy. The best example is the adenoviral delivery of the gene encoding human TNF-α, a soluble cytokine with potent antitumor and antiangiogenic properties that also mediates the cellular immune response. TNF-α has shown some antitumor activity as a single agent in clinical trials; however, it also has substantial toxicity on systemic administration. TNF-α has synergistic effects with radiation therapy, presumably because it can ­ damage DNA through reactive oxygen species.207 Weichselbaum and colleagues developed and tested a ­replication-­incompetent adenoviral vector that encodes the human TNF-α gene (TNF, formerly designated TNFA), the expression of which is under the control of the radiation-inducible promoter for the immediate early growth response 1 gene (EGR1), which is inserted in the vector upstream of the start site of the TNF cDNA.208 This vector, referred to as TNFerade, is then injected into the tumor, where it prompts infected tumor cells to express TNF-α on radiation. Intratumoral injection also avoids the toxicity associated with systemic administration of this cytokine. TNFerade has demonstrated synergistic effects in combination with radiation in preclinical models of human cancer (reviewed by Mezhir and colleagues209) and was well tolerated in a phase I trial.210 Several phase II trials of this strategy are being conducted.

The second gene therapy strategy that has received significant attention involves replacing a gene that is miss­ ing or abnormally deactivated in tumor cells. The beststudied example of such a gene is the tumor suppressor TP53 (formerly designated p53), which is mutated or otherwise functionally abnormal in most types of cancer in humans. In normal cells, TP53 participates in many cellular functions, including apoptosis, genomic stability, and cell cycle checkpoints. The use of gene therapy to restore TP53 in tumor cells in which it is deleted would be expected to enhance the cells’ response to cancer therapies that rely on the apoptosis pathway to kill tumor cells. Although no consensus has been reached about the importance of apoptosis as a cell death pathway in the clinical response of human tumors to radiation, numerous preclinical tests of adenoviral-mediated TP53 (Ad-p53) have shown restoration of radiation-induced apoptosis and an overall enhancement of tumor response in various tumor models.211,212 Based on these reports, a prospective, single-arm, phase II study was undertaken to evaluate the feasibility of Ad-p53 gene transfer and radiation therapy for patients with NSCLC.213 In that study, 19 patients with nonmetastatic NSCLC who were not eligible for chemotherapy or surgery were treated on an outpatient basis with 60 Gy over 6 weeks combined with three intratumoral injections of Ad-p53 on days 1, 18, and 32. The results showed that this combination was well tolerated and that some proportion of patients showed evidence of tumor regression. Although apoptosis was not assessed, serial biopsies of the tumors suggested that upregulation of the expression of the proapoptotic gene BAK correlated with Ad-p53 injection. These promising results are being followed up in larger trials.

Oncolytic Viruses Most adenoviral vectors that have been tested for gene therapy purposes have been engineered to disable their ability to replicate within the infected cells. This process normally involves splicing out regions of the viral genome required for its replication, the so-called early genes (E1-E4). However, some novel vectors have been developed in which the replication competency of the virus is left intact. When tumor cells are infected with these vectors, the ensuing viral replication eventually leads to oncolysis, which kills the infected cells. The lysed cells release viral progeny, which can infect neighboring, previously uninfected cells. This attribute of replication-competent vectors may overcome the major limitation of replication-incompetent vectors—their poor distribution throughout the injected tumor. Although most such vectors are designed for use primarily as single agents because of their cytotoxicity, some have been used in combination with irradiation. The best example is the combination of radiation therapy with administration of the replication-competent Ad-CD/TK vector for prostate cancer206 (discussed under “Delivery of Radiosensitizing Drugs”). Another type of oncolytic vector is the conditionally replicative vector, the best studied of which has been ONYX-15 (dl1520). This adenoviralbased vector lacks the E1B gene, the product of which binds to and inactivates wild-type TP53 in infected cells, a step that is required for viral replication. ONYX-15 can replicate only in tumor cells with mutant or deleted TP53

5  Principles of Combining Radiation Therapy and Molecular Therapeutics

and presumably not in normal cells, providing it with some tumor cell specificity. The safety of ONYX-15 as singleagent therapy has been demonstrated in several clinical trials (reviewed by Vecil and Lang214), and preclinical studies have shown at least additive effects when ONYX-15 is combined with irradiation.215,216 At least one other conditionally replicative vector is under development. Ad-Delta24-RGD has a deletion in the E1A region of the viral genome, which is responsible for binding to the cellular RB1 protein.214 In normal cells, the vector cannot bind to RB1 and therefore cannot replicate; however, in tumor cells with nonfunctional RB1, replication of this vector is supported. Preclinical studies have shown enhancement of xenograft tumor response when Ad-Delta24-RGD is combined with radiation.217,218 This strategy is expected to be tested in clinical trials in the near future.

Silencing Gene Expression Most of the gene therapy strategies discussed have been designed to directly or indirectly sensitize the targeted tumor cells to radiation by inducing the expression of exogenous genes. An alternative approach is to introduce agents that specifically suppress the expression of cellular genes whose protein products participate in mechanisms that mediate radioresistance. The two technologic means of silencing gene expression are antisense oligonucleotides and RNA interference (RNAi). Antisense oligonucleotides are short, single-stranded deoxynucleotide sequences designed to bind to mRNA that is complementary to the gene of interest in a sequence-specific manner.219 The resulting mRNA/oligonucleotide complex causes the targeted mRNA to be enzymatically destroyed by nucleases, leading to lack of transcription of the gene of interest. Antisense oligonucleotides have been widely tes­ted for the treatment of various diseases, including cancer. In preclinical studies, antisense oligonucleotides have been shown to radiosensitize human tumor cells by targeting different genes involved in the radiation response, especially genes whose protein products mediate blocks to apoptosis (e.g., BCL2) or are involved in the repair of radiation-induced DNA double-strand breaks (e.g., XRCC6 [formerly KU70]). Finding ways of effectively delivering antisense oligonucleotides to patients has been problematic and has required extensive development. However, encapsulating antisense oligonucleotides in liposomes for systemic delivery to patients seems to be a promising approach. A phase I study in which a liposomeencapsulated RAF antisense oligonucleotide was used in combination with radiation therapy for patients with advanced malignancies showed the combination to be well tolerated.220 Gene silencing by RNAi is accomplished through the use of small interfering RNAs (siRNAs) consisting of doublestranded RNA sequences complementary to the gene of interest.221 After being taken up by the recipient cell, the siRNA is processed by cellular enzymes to release the antisense strand, which then targets the respective gene’s mRNA for destruction, much like the process for antisense oligonucleotides described earlier. RNAi technology has become quite efficient and essentially results in a functional knockout of the targeted gene. Preclinical studies have described several examples in which siRNA directed

131

to genes whose products are critical to mediating a radioresistant phenotype has induced radiosensitization of tumor cells. Although no clinical studies of siRNA in combination with radiation therapy have been reported, such trials are expected in the near future.

Immunotherapy Mechanisms It has long been recognized that malignant tumors may express tumor-specific antigens that can elicit immuno­ logic responses that can restrict tumor growth. An effici­ ently functioning immune system may improve the efficacy of conventional treatment modalities, whereas immune deficiency may result in the opposite outcome.222 Tumor antigens evoke innate and adaptive immune res­ponses. Innate immunity involves the activation of phagocytic, natural killer, and antigen-presenting cells, and it is trig­ gered within hours to days after antigen presentation. Adaptive immunity involves functions of T and B lym­ phocytes and begins within days to months after antigen presentation.222 Both active immunization with heavily irradiated tumor cells and nonspecific immunotherapy with bacterial adjuvants to boost the immune response have been extensively investigated. One of the most thoroughly studied adjuvants is complete Freund’s adjuvant, which consists of a suspension of killed bacillus Calmette-Guérin (BCG) bacteria in mineral oil. Several types of bacteria or bacterial extracts, primarily BCG and Corynebacterium parvum, have been used to stimulate antitumor immunologic reactions as early as the 1970s.223,224 Although these agents are potent stimulators of many facets of immunologic reactions and possess substantial activity against murine tumors when given alone or in combination with chemotherapy or irradiation, they have shown little therapeutic benefit in human clinical trials.223,224 An important limiting factor in the clinical application of these agents is the toxicity ­associated with their repeated administration.223,224

Molecular Immunotherapy In the mid-1990s, it was discovered that the key immunostimulatory activity of bacteria resides in their DNA225—specifically in unmethylated cytosine-guanine (CpG) motifs, which are prevalent in bacteria but not in vertebrates.226 This discovery led to the synthesis of oligodeoxynucleotides containing unmethylated CpG motifs that showed strong immunostimulatory activity but little toxicity.226-228 CpG oligodeoxynucleotides are recognized by the innate immune cells of vertebrates, primarily plasmacytoid dendritic cells, through the toll-like receptor 9. These CpG oligodeoxynucleotides have been shown to have antitumor activity on their own229-232 and to enhance the antitumor activity of chemotherapeutic agents230 and that of radiation given as a single exposure232 or in fractionated regimens.231 Substantial increases in tumor cure rates from the addition of CpG oligodeoxynucleotides to fractionated irradiation for the treatment of a murine sarcoma model are illustrated in Figure 5-6. Tumors in mice treated with the CpG oligodeoxynucleotides and radiation were heavily infiltrated by host inflammatory cells, primarily lymphocytes232; moreover, the mice that were cured by the combined treatment remained resistant to re-inoculation

132

PART 1  Principles

100

CpG ODN 1826

Control

Tumor cure (%)

80 60 40 20 0 0

10

20

30

40

50

60

70

80

90

100

Total radiation dose (Gy)

FIGURE 5-6.  Effects on tumor curability are shown for the addition of the CpG motif–based oligodeoxynucleotide (ODN) 1826 to radiation therapy. The percentage of tumor cures is plotted as a function of radiation dose. Mice bearing FSa fibrosarcoma tumors in the legs were exposed to a range of fractionated doses when the tumors reached 8 mm in diameter. Mice were also treated with seven doses of active CpG ODN 1826 (blue circles) or inactive ODN 2138 (red circles) at a dose of 100 μg per mouse given subcutaneously peritumorally. The first dose was given when tumor diameters were 6 mm, the second dose at 8 mm, and then weekly for 5 additional weeks. The radiation dose yielding 50% tumor cure (TCD50) was determined at 100 days after irradiation. Horizontal bars represent 95% confidence intervals. (From Mason KA, Ariga H, Neal R, et al. Targeting toll-like receptor 9 with CpG oligodeoxynucleotides enhances tumor response to fractionated radiotherapy. Clin Cancer Res 2005;11:361-369.)

of tumor cells for many months after tumor regression.231 These results warrant testing of this therapeutic approach in carefully designed clinical trials, as was observed by Koski and Czerniecki233 and Sutherland.234

Summary Research on EGFR signaling exemplifies the development of novel cancer therapy strategies that are based on better understanding of tumor biology. The discovery of how perturbation of EGFR signaling affects malignant transformation and tumor response to therapy led to the conception and completion of a clinical research program that produced a new front-line therapy modality that improved the survival of patients with locally advanced HNSCC. Although this pivotal work established the proof of principle that tumors can be selectively sensitized to radiation therapy through modulation of a signaling pathway, the clinical benefit resulting from this strategy remains rather modest, and many questions remain to be addressed. Validation of this concept, however, has provided a robust foundation for expanding investigations aiming at augmenting tumor response to therapy by modulating other signaling pathways, individually or in combination. These efforts are anticipated to yield novel therapies for various types of cancer in the near future.

REFERENCES   1. Cohen S, Carpenter G, King L Jr. Epidermal growth factorreceptor-protein kinase interactions. Co-purification of receptor and epidermal growth factor-enhanced phosphorylation activity. J Biol Chem 1980;255:4834-4842.

  2. Cohen LF, Balow JE, Magrath IT. Acute tumor lysis syndrome: a review of 37 patients with Burkitt’s lymphoma. Am J Med 1980;68:486-491.   3. Downward J, Yarden Y, Mayes E, et al. Close similarity of epidermal growth factor receptor and v-erb-B oncogene protein sequences. Nature 1984;307:521-527.   4. Davies RL, Grosse VA, Kucherlapati R, et al. Genetic analysis of epidermal growth factor action: assignment of human epidermal growth factor receptor gene to chromosome 7. Proc Natl Acad Sci U S A 1980;77:4188-4192.   5. Haley J, Whittle N, Bennet P, et al. The human EGF receptor gene: structure of the 110 kb locus and identification of sequences regulating its transcription. Oncogene Res 1987;1:375-396.   6. Burgess AW, Cho HS, Eigenbrot C, et al. An open-and-shut case? Recent insights into the activation of EGF/ErbB receptors. Mol Cell 2003;12:541-552.   7. Nishinaka T, Yabe-Nishimura C. EGF receptor-ERK pathway is the major signaling pathway that mediates upregulation of aldose reductase expression under oxidative stress. Free Radic Biol Med 2001;31:205-216.   8. Darnell JE Jr, Kerr IM, Stark GR. Jak-STAT pathways and transcriptional activation in response to IFNs and other extracellular signaling proteins. Science 1994;264:1415-1421.   9. Harari PM, Huang S. Radiation combined with EGFR signal inhibitors: head and neck cancer focus. Semin Radiat Oncol 2006;16:38-44. 10. Kijima T, Niwa H, Steinman RA, et al. STAT3 activation abrogates growth factor dependence and contributes to head and neck squamous cell carcinoma tumor growth in vivo. Cell Growth Differ 2002;13:355-362. 11. Xi S, Zhang Q, Gooding WE, et al. Constitutive activation of Stat5b contributes to carcinogenesis in vivo. Cancer Res 2003;63:6763-6771. 12. Hynes NE, Lane HA. ERBB receptors and cancer: the complexity of targeted inhibitors. Nat Rev Cancer 2005;5:341-354. 13. Thomas SM, Coppelli FM, Wells A, et al. Epidermal growth factor receptor-stimulated activation of phospholipase Cgamma-1 promotes invasion of head and neck squamous cell carcinoma. Cancer Res 2003;63:5629-5635. 14. Dittmann K, Mayer C, Fehrenbacher B, et al. Radiationinduced epidermal growth factor receptor nuclear import is linked to activation of DNA-dependent protein kinase. J Biol Chem 2005;280:31182-31189. 15. Lin SY, Makino K, Xia W, et al. Nuclear localization of EGF receptor and its potential new role as a transcription factor. Nat Cell Biol 2001;3:802-808. 16. Hanada N, Lo HW, Day CP, et al. Co-regulation of B-Myb expression by E2F1 and EGF receptor. Mol Carcinog 2006;45:10-17. 17. Khanna KK, Jackson SP. DNA double-strand breaks: signaling, repair and the cancer connection. Nat Genet 2001;27:247-254. 18. Bao J, Alroy I, Waterman H, et al. Threonine phosphorylation diverts internalized epidermal growth factor receptors from a degradative pathway to the recycling endosome. J Biol Chem 2000;275:26178-26186. 19. Hanover JA, Beguinot L, Willingham MC, et al. Transit of receptors for epidermal growth factor and transferrin through clathrincoated pits. Analysis of the kinetics of receptor entry. J Biol Chem 1985;260:15938-15945. 20. Herbst JJ, Opresko LK, Walsh BJ, et al. Regulation of postendocytic trafficking of the epidermal growth factor receptor through endosomal retention. J Biol Chem 1994;269:12865-12873. 21. Warren CM, Landgraf R. Signaling through ERBB receptors: multiple layers of diversity and control. Cell Signal 2006;18:923933. 22. Pike LJ, Miller JM. Cholesterol depletion delocalizes phosphatidylinositol bisphosphate and inhibits hormone-stimulated phosphatidylinositol turnover. J Biol Chem 1998;273:22298-22304. 23. Grandis J, Tweardy D. Elevated levels of transforming growth factor α and epidermal growth factor receptor messenger RNA are early markers of carcinogenesis in head and neck cancer. Cancer Res 1993;53:3579-3584.

5  Principles of Combining Radiation Therapy and Molecular Therapeutics 24. Grandis JR, Melhem MF, Barnes EL, et al. Quantitative immunohistochemical analysis of transforming growth factor- a and epidermal growth factor receptor in patients with squamous cell carcinoma of the head and neck. Cancer 1996;78:1284-1292. 25. Schmidt-Ullrich RK, Mikkelsen RB, Dent P, et al. Radiationinduced proliferation of the human A431 squamous carcinoma cells is dependent on EGFR tyrosine phosphorylation. Oncogene 1997;15:1191-1197. 26. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non–small-cell lung cancer to gefitinib. N Engl J Med 2004;350:2129-2139. 27. Wikstrand CJ, McLendon RE, Friedman AH, et al. Cell surface localization and density of the tumor-associated variant of the epidermal growth factor receptor. EGFRvIII. Cancer Res 1997;57:4130-4140. 28. Etienne-Grimaldi MC, Pereira S, Magne N, et al. Analysis of the dinucleotide repeat polymorphism in the epidermal growth factor receptor (EGFR) gene in head and neck cancer patients. Ann Oncol 2005;16:934-941. 29. Cappuzzo F, Hirsch FR, Rossi E, et al. Epidermal growth factor receptor gene and protein and gefitinib sensitivity in non–smallcell lung cancer. J Natl Cancer Inst 2005;97:643-655. 30. Psyrri A, Yu Z, Weinberger PM, et al. Quantitative determination of nuclear and cytoplasmic epidermal growth factor receptor ­expression in oropharyngeal squamous cell cancer by using automated quantitative analysis. Clin Cancer Res 2005;11:5856-5862. 31. Lo HW, Hung MC. Nuclear EGFR signalling network in cancers: linking EGFR pathway to cell cycle progression, nitric oxide pathway and patient survival. Br J Cancer 2006;94:184-188. 32. Merlano M, Vitale V, Rosso R, et al. Treatment of advanced ­squamous-cell carcinoma of the head and neck with alternating chemotherapy and radiotherapy. N Engl J Med 1992;327: 1115-1121. 33. Milas L, Akimoto T, Hunter NR, et al. Relationship between cyclin D1 expression and poor radioresponse of murine carcinomas. Int J Radiat Oncol Biol Phys 2002;52:514-521. 34. Withers HR, Taylor JM, Maciejewski B. The hazard of accelerated tumor clonogen repopulation during radiotherapy. Acta Oncol 1988;27:131-146. 35. Schmidt-Ullrich RK, Johnson CR, Khandelwal S, et al. Concomitant boost-accelerated, superfractionated irradiation for the treatment of advanced carcinomas of the oral cavity and oropharynx. Recent Results Cancer Res 1994;134:69-79. 36. Schmidt-Ullrich RK, Valerie K, Chan W, et al. Expression of oestrogen receptor and transforming growth factor-alpha in MCF-7 cells after exposure to fractionated irradiation. Int J Radiat Biol 1992;61:405-415. 37. Kavanagh BD, Lin PS, Chen P, et al. Radiation-induced enhanced proliferation of human squamous cancer cells in vitro: a release from inhibition by epidermal growth factor. Clin Cancer Res 1995;1:1557-1562. 38. Schmidt-Ullrich RK, Valerie K, Fogleman PB, et al. Radiationinduced autophosphorylation of epidermal growth factor receptor in human malignant mammary and squamous epithelial cells. Radiat Res 1996;145:81-85. 39. Schmidt-Ullrich RK, Contessa JN, Lammering G, et al. ERBB receptor tyrosine kinases and cellular radiation responses. Oncogene 2003;22:5855-5865. 40. Akimoto T, Hunter NR, Buchmiller L, et al. Inverse relationship between epidermal growth factor receptor expression and radiocurability of murine carcinomas. Clin Cancer Res 1999;5:2884-2890. 41. Ang KK, Berkey BA, Tu X, et al. Impact of epidermal growth factor receptor expression on survival and pattern of relapse in patients with advanced head and neck carcinoma. Cancer Res 2002;62:7350-7356. 42. Liang K, Ang KK, Milas L, et al. The epidermal growth factor receptor mediates radioresistance. Int J Radiat Oncol Biol Phys 2003;57:246-254. 43. Mendelsohn J, Masui H, Goldenberg A. Anti-epidermal growth factor receptor monoclonal antibodies may inhibit A431 tumor cell proliferation by blocking an autocrine pathway. Trans Assoc Am Physicians 1987;100:173-178.

133

44. Sato JD, Kawamoto T, Le AD, et al. Biological effects in vitro of monoclonal antibodies to human epidermal growth factor receptors. Mol Biol Med 1983;1:511-529. 45. Huang SM, Bock JM, Harari PM. Epidermal growth factor receptor blockade with C225 modulates proliferation, apoptosis, and radiosensitivity in squamous cell carcinomas of the head and neck. Cancer Res 1999;59:1935-1940. 46. Huang SM, Harari PM. Epidermal growth factor receptor inhibition in cancer therapy: biology, rationale and preliminary clinical results. Invest New Drugs 1999;17:259-269. 47. Milas L. Chemoradiation interactions: potential of newer chemotherapeutic agents. In Perry MC (ed): American Society of Clinical Oncology Education Book, 2000, pp 207-213. 48. Saleh MN, Raisch KP, Stackhouse MA, et al. Combined modality therapy of A431 human epidermoid cancer using anti-EGFR antibody C225 and radiation. Cancer Biother Radiopharm 1999;14:451-463. 49. Nasu S, Ang KK, Fan Z, et al. C225 antiepidermal growth factor receptor antibody enhances tumor radiocurability. Int J Radiat Oncol Biol Phys 2001;51:474-477. 50. Krause M, Schutze C, Petersen C, et al. Different classes of EGFR inhibitors may have different potential to improve local tumour control after fractionated irradiation: a study on C225 in FaDu hSCC. Radiother Oncol 2005;74:109-115. 51. Bianco C, Tortora G, Bianco R, et al. Enhancement of antitumor activity of ionizing radiation by combined treatment with the selective epidermal growth factor receptor-tyrosine kinase inhibitor ZD1839 (Iressa). Clin Cancer Res 2002;8:3250-3258. 52. Chinnaiyan P, Huang S, Vallabhaneni G, et al. Mechanisms of enhanced radiation response following epidermal growth factor receptor signaling inhibition by erlotinib (Tarceva). Cancer Res 2005;65:3328-3335. 53. She Y, Lee F, Chen J, et al. The epidermal growth factor receptor tyrosine kinase inhibitor ZD1839 selectively potentiates radiation response of human tumors in nude mice, with a marked improvement in therapeutic index. Clin Cancer Res 2003;9:3773-3778. 54. Solomon B, Hagekyriakou J, Trivett MK, et al. EGFR blockade with ZD1839 (“Iressa”) potentiates the antitumor effects of single and multiple fractions of ionizing radiation in human A431 squamous cell carcinoma. Epidermal growth factor receptor. Int J Radiat Oncol Biol Phys 2003;55:713-723. 55. Robert F, Ezekiel MP, Spencer SA, et al. Phase I study of antiepidermal growth factor receptor antibody cetuximab in combination with radiation therapy in patients with advanced head and neck cancer. J Clin Oncol 2001;19:3234-3243. 56. Bonner JA, Harari PM, Giralt J, et al. Radiotherapy plus cetuximab for squamous-cell carcinoma of the head and neck. N Engl J Med 2006;354:567-578. 57. Pfister DG, Su YB, Kraus DH, et al. Concurrent cetuximab, cisplatin, and concomitant boost radiotherapy for locoregionally advanced, squamous cell head and neck cancer: a pilot phase II study of a new combined-modality paradigm. J Clin Oncol 2006;24:1072-1078. 58. Burtness B, Goldwasser MA, Flood W, et al. Phase III randomized trial of cisplatin plus placebo compared with cisplatin plus ­ cetuximab in metastatic/recurrent head and neck cancer: an Eastern Cooperative Oncology Group study J Clin Oncol 2005;23:8646-8654. 59. Soulieres D, Senzer NN, Vokes EE, et al. Multicenter phase II study of erlotinib, an oral epidermal growth factor receptor tyrosine ­kinase inhibitor, in patients with recurrent or metastatic squamous cell cancer of the head and neck. J Clin Oncol 2004;22:77-85. 60. Chung KY, Shia J, Kemeny NE, et al. Cetuximab shows activity in colorectal cancer patients with tumors that do not express the epidermal growth factor receptor by immunohistochemistry. J Clin Oncol 2005;23:1803-1810. 61. Saltz L. Epidermal growth factor receptor-negative colorectal cancer: is there truly such an entity? Clin Colorectal Cancer 2005;5;(Suppl 2):S98-S100. 62. Pal SK, Pegram M. Epidermal growth factor receptor and signal transduction: potential targets for anti-cancer therapy. Anticancer Drugs 2005;16:483-494.

134

PART 1  Principles

63. Choong NW, Cohen EE. Epidermal growth factor receptor directed therapy in head and neck cancer. Crit Rev Oncol Hematol 2006;57:25-43. 64. Milas L, Fang F, Mason KA, et al. Importance of maintenance therapy in C225-induced enhancement of tumor control by fractionated radiation. Int J Radiat Oncol Biol Phys 2007;67: 568-572. 65. Lammering G, Hewit TH, Valerie K, et al. EGFRvIII-mediated radioresistance through a strong cytoprotective response. Oncogene 2003;22:5545-5553. 66. Pedersen MW, Meltorn M, Damstrup L, et al. The type III epidermal growth factor receptor mutation. Biological significance and potential target for anti-cancer therapy. Ann Oncol 2001;12: 745-760. 67. Sok JC, Coppelli FM, Thomas SM, et al. Mutant epidermal growth factor receptor (EGFRvIII) contributes to head and neck cancer growth and resistance to EGFR targeting Clin Cancer Res 2006;12:5064-5073. 68. Kalyankrishna S, Grandis JR. Epidermal growth factor receptor biology in head and neck cancer. J Clin Oncol 2006;24:26662672. 69. Zhan M, Han ZC. Phosphatidylinositide 3-kinase/AKT in radiation responses. Histol Histopathol 2004;19:915-923. 70. Pedrero JM, Carracedo DG, Pinto CM, et al. Frequent genetic and biochemical alterations of the PI 3-K/AKT/PTEN pathway in head and neck squamous cell carcinoma. Int J Cancer 2005;114:242-248. 71. Okami K, Wu L, Riggins G, et al. Analysis of PTEN/MMAC1 ­alterations in aerodigestive tract tumors. Cancer Res 1998;58: 509-511. 72. Motoyama A, Hynes NE, Lane HA. The efficacy of ErbB ­receptortargeted anticancer therapeutics is influenced by the availability of epidermal growth factor-related peptides. Cancer Res 2002;62:3151-3158. 73. Hurbin A, Dubrez L, Coll JL, et al. Inhibition of apoptosis by amphiregulin via an insulin-like growth factor-1 receptor-­dependent pathway in non–small cell lung cancer cell lines. J Biol Chem 2002;277:49127-49133. 74. Chakravarti A, Loeffler JS, Dyson NJ. Insulin-like growth factor receptor I mediates resistance to anti-epidermal growth factor receptor therapy in primary human glioblastoma cells through continued activation of phosphoinositide 3-kinase signaling. Cancer Res 2002;62:200-207. 75. Folkman J. Tumor angiogenesis: therapeutic implications. N Engl J Med 1971;285:1182-1186. 76. O’Reilly MS. Antiangiogenesis: basic principles. In Rosenberg SA (ed): Principles and Practice of the Biologic Therapy of Cancer, 3rd ed. Philadelphia: Lippincott Williams & Wilkins, 2000, pp 827-843. 77. Folkman J. Angiogenesis in cancer, vascular, rheumatoid and other diseases. Nat Med 1995;1:27-31. 78. Fidler IJ, Langley RR, Kerbel RS, Ellis LM. Angiogenesis. In ­DeVita VT, Hellman S, Rosenberg SA (eds): Cancer Principles and Practice of Oncology, 7th ed. Philadelphia: Lippincott Williams & Wilkins, 2005. 79. Dark GG, Hill SA, Prise VE, et al. Combretastatin A-4, an agent that displays potent and selective toxicity toward tumor vasculature. Cancer Res 1997;57:1829-1834. 80. Jain RK. Normalizing tumor vasculature with anti-angiogenic therapy: a new paradigm for combination therapy. Nat Med 2001;7:987-989. 81. Konerding MA, Malkusch W, Klapthor B, et al. Evidence for characteristic vascular patterns in solid tumours: quantitative studies using corrosion casts. Br J Cancer 1999;80:724-732. 82. Dvorak HF. Tumors: wounds that do not heal. N Engl J Med 1986;315:1650-1659. 83. Hamada J, Cavanaugh PG, Lotan O, et al. Separable growth and migration factors for large-cell lymphoma cells secreted by microvascular endothelial cells derived from target organs for metastasis. Br J Cancer 1992;66:349-354. 84. Nicosia RF, Tchao R, Leighton J. Interactions between newly formed endothelial channels and carcinoma cells in plasma clot culture. Clin Exp Metastasis 1986;4:91-104.

  85. Rak J, Filmus J, Kerbel RS. Reciprocal paracrine interactions between tumour cells and endothelial cells: the ‘angiogenesis progression’ hypothesis. Eur J Cancer 1996;32A:2438-2450.   86. Folkman J. Tumor angiogenesis and tissue factor. Nat Med 1996;2:167-168.   87. Huang X, Molema G, King S, et al. Tumor infarction in mice by antibody-directed targeting of tissue factor to tumor vasculature. Science 1997;275:547-550.   88. Thorpe PE. Vascular targeting agents as cancer therapeutics. Clin Cancer Res 2004;10:415-427.   89. Arap W, Pasqualini R, Ruoslahti E. Cancer treatment by targeted drug delivery to tumor vasculature in a mouse model. Science 1998;279:377-380.   90. Siemann DW, Rojiani AM. Enhancement of radiation therapy by the novel vascular targeting agent ZD6126. Int J Radiat Oncol Biol Phys 2002;53:164-171.   91. Raben D, Bianco C, Damiano V, et al. Antitumor activity of ZD6126, a novel vascular-targeting agent, is enhanced when combined with ZD1839, an epidermal growth factor receptor tyrosine kinase inhibitor, and potentiates the effects of radiation in a human non–small cell lung cancer xenograft model. Mol Cancer Ther 2004;3:977-983.   92. Wachsberger PR, Burd R, Marero N, et al. Effect of the tumor vascular-damaging agent, ZD6126, on the radioresponse of U87 glioblastoma. Clin Cancer Res 2005;11:835-842.   93. Garcia-Barros M, Paris F, Cordon-Cardo C, et al. Tumor response to radiotherapy regulated by endothelial cell apoptosis. Science 2003;300:1155-1159.   94. Gorski DH, Beckett MA, Jaskowiak NT, et al. Blockade of the vascular endothelial growth factor stress response increases the antitumor effects of ionizing radiation. Cancer Res 1999;59:3374-3378.   95. Kaplan H, Murphy E. The effect of local roentgen irradiation on the biological behavior of a transplantable mouse carcinoma. Increased frequency of pulmonary metastasis. J Natl Cancer Inst 1949;9:407-413.   96. Suit H, Sedlacek R, Gillette E. Examination for a correlation between probabilities of development of distant metastasis and local recurrence. Radiology 1970;95:189-194.   97. Hast J, Schiffer IB, Neugebauer B, et al. Angiogenesis and ­fibroblast proliferation precede formation of recurrent tumors after radiation therapy in nude mice. Anticancer Res 2002;22: 677-688.   98. Camphausen K, Moses MA, Beecken WD, et al. Radiation therapy to a primary tumor accelerates metastatic growth in mice. Cancer Res 2001;61:2207-2211.   99. Milas L, Hirata H, Hunter N, et al. Effect of radiation-induced injury of tumor bed stroma on metastatic spread of murine sarcomas and carcinomas. Cancer Res 1988;48:2116-2210. 100. Milas L. Tumor bed effect in murine tumors: relationship to tumor take and tumor macrophage content. Radiat Res 1990;123:232-236. 101. Hartford AC, Gohongi T, Fukumura D, et al. Irradiation of a primary tumor, unlike surgical removal, enhances angiogenesis suppression at a distal site: potential role of host-tumor interaction. Cancer Res 2000;60:2128-2131. 102. Folkman J. Clinical applications of angiogenesis research. N Engl J Med 1995;333:1757-1763. 103. Libutti SK, Pluda JM. Antiangiogenesis: clinical applications. In Rosenberg SA (ed): Principles and Practice of the Biologic Therapy of Cancer, 3rd ed. Philadelphia: Lippincott Williams & Wilkins, 2000, pp 844-864. 104. O’Reilly MS, Holmgren L, Chen C, et al. Angiostatin induces and sustains dormancy of human primary tumors in mice. Nat Med 1996;2:689-692. 105. Rusk A, McKeegan E, Haviv F, et al. Preclinical evaluation of antiangiogenic thrombospondin-1 peptide mimetics, ABT-526 and ABT-510, in companion dogs with naturally occurring cancers. Clin Cancer Res 2006;12:7444-7455. 106. Ezekowitz RAB, Mulliken JB, Folkman J. Interferon alfa-2a therapy for life-threatening hemangiomas of infancy. N Engl J Med 1992;326:1456-1463.

5  Principles of Combining Radiation Therapy and Molecular Therapeutics 107. Kaban LB, Mulliken JB, Ezekowitz RA, et al. Antiangiogenic therapy of a recurrent giant cell tumor of the mandible with interferon alfa-2a. Pediatrics 1999;103:1145-1149. 108. Boehm T, Folkman J, Browder T, et al. Antiangiogenic therapy of experimental cancer does not induce acquired drug resistance. Nature 1997;390:404-407. 109. Alon T, Hemo I, Itin A, et al. VEGF acts as a survival factor for newly formed retinal vessels and has implications for retinopathy of prematurity. Nat Med 1995;1:1024-1028. 110. Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its receptors. Nat Med 2003;9:669-676. 111. Levine RJ, Maynard SE, Qian C, et al. Circulating angiogenic factors and the risk of preeclampsia. N Engl J Med 2004;350: 672-683. 112. Carmeliet P, Storkebaum E. Vascular and neuronal effects of VEGF in the nervous system: implications for neurological disorders. Semin Cell Dev Biol 2002;13:39-53. 113. Gerber HP, Malik AK, Solar GP, et al. VEGF regulates haematopoietic stem cell survival by an internal autocrine loop mechanism. Nature 2002;417:954-958. 114. LeCouter J, Moritz DR, Li B, et al. Angiogenesis-­independent endothelial protection of liver: role of VEGFR-1. Science 2003;299:890-893. 115. Ohm JE, Gabrilovich DI, Sempowski GD, et al. VEGF inhibits T-cell development and may contribute to tumor-induced immune suppression. Blood 2003;101:4878-4886. 116. Kabbinavar F, Hurwitz HI, Fehrenbacher L, et al. Phase II, randomized trial comparing bevacizumab plus fluorouracil (FU)/ leucovorin (LV) and FU/LV alone in patients with metastatic colorectal cancer. J Clin Oncol 2003;21:60-65. 117. Yang JC, Haworth L, Sherry RM, et al. A randomized trial of bevacizumab, an anti-vascular endothelial growth factor antibody, for metastatic renal cancer. N Engl J Med 2003;349: 427-434. 118. Fan F, Wey JS, McCarty MF, et al. Expression and function of vascular endothelial growth factor receptor-1 on human colorectal cancer cells. Oncogene 2005;24:2647-2653. 119. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004;350:2335-2342. 120. Sandler A, Gray R, Perry MC, et al. Paclitaxel-carboplatin alone or with bevacizumab for non–small-cell lung cancer. N Engl J Med 2006;355:2542-2550. 121. Kuenen BC, Rosen L, Smit EF, et al. Dose-finding and pharmacokinetic study of cisplatin, gemcitabine, and SU5416 in patients with solid tumors. J Clin Oncol 2002;20:1657-1667. 122. Miller KD, Chap LI, Holmes FA, et al. Randomized phase III trial of capecitabine compared with bevacizumab plus capecitabine in patients with previously treated metastatic breast cancer. J Clin Oncol 2005;23:792-799. 123. Koehne C, Bajetta E, Lin E, et al. Results of an interim analysis of a multinational randomized, double-blind, phase III study in patients (pts) with previously treated metastatic colorectal cancer (mCRC) receiving FOLFOX4 and PTK787/ZK 222584 (PTK/ZK) or placebo (CONFIRM 2) [abstract]. J Clin Oncol 2006;24;(18S):3508. 124. Major P, Trarbach T, Lenz H, et al. A meta-analysis of two randomized, double-blind, placebo-controlled, phase III studies in patients (pts) with metastatic colorectal cancer (mCRC) receiving FOLFOX4 and PTK/ZK to determine clinical benefit on progression-free survival (PFS) in high LDH pts [abstract]. J Clin Oncol (Meeting Abstracts) 2006;24:3529. 125. Relf M, LeJeune S, Scott PA, et al. Expression of the angiogenic factors vascular endothelial cell growth factor, acidic and basic fibroblast growth factor, tumor growth factor beta-1, plateletderived endothelial cell growth factor, placenta growth factor, and pleiotrophin in human primary breast cancer and its relation to angiogenesis. Cancer Res 1997;57:963-969. 126. Solorzano CC, Baker CH, Tsan R, et al. Optimization for the blockade of epidermal growth factor receptor signaling for therapy of human pancreatic carcinoma. Clin Cancer Res 2001;7:2563-2572.

135

127. Baker CH, McCarty MF, Tsan R, et al. Phenotypic diversity of organ-specific endothelial cells [abstract]. Proc Am Assoc Cancer Res 2001;42:2186. 128. Brogi E, Wu T, Namiki A, et al. Indirect angiogenic cytokines upregulate VEGF and bFGF gene expression in vascular smooth muscle cells, whereas hypoxia upregulates VEGF expression only. Circulation 1994;90:649-652. 129. Uehara H, Kim SJ, Karashima T, et al. Blockade of the PDGF-R signaling by STI571 inhibits angiogenesis and growth of human prostate cancer cells in the bone of nude mice [abstract]. Proc Am Assoc Cancer Res 2001;42:2192. 130. Neeltje S, Johan WRN, Hans G. Small molecule tyrosine kinase inhibitors in the treatment of solid tumors: an update of recent developments. Ann Surg Oncol 2007;14:942-953. 131. Herbst RS, Heymach JV, O’Reilly MS, et al. Vandetanib (ZD6474): an orally available receptor tyrosine kinase inhibitor that selectively targets pathways critical for tumor growth and angiogenesis. Exp Opin Investig Drugs 2007;16:239-249. 132. Motzer RJ, Hutson TE, Tomczak P, et al. Sunitinib versus interferon alfa in metastatic renal-cell carcinoma. N Engl J Med 2007;356:115-124. 133. Escudier B, Eisen T, Stadler WM, et al. Sorafenib in advanced clear-cell renal-cell carcinoma. N Engl J Med 2007;356: 125-134. 134. Teicher B, Dupuis NP, Kusumoto T, et al. Anti-angiogenic agents can increase tumor oxygenation and response to radiation therapy. Radiat Oncol Investig 1995;2:269-276. 135. Hansen-Algenstaedt N, Stoll BR, Padera TP, et al. Tumor oxygenation in hormone-dependent tumors during vascular endothelial growth factor receptor-2 blockade, hormone ablation, and chemotherapy. Cancer Res 2000;60:4556-4560. 136. Murata R, Nishimura Y, Hiraoka M. An antiangiogenic agent (TNP-470) inhibited reoxygenation during fractionated radiotherapy of murine mammary carcinoma. Int J Radiat Oncol Biol Phys 1997;37:1107-1113. 137. Lund EL, Bastholm L, Kristjansen PE. Therapeutic synergy of TNP-470 and ionizing radiation: effects on tumor growth, vessel morphology, and angiogenesis in human glioblastoma multiforme xenografts. Clin Cancer Res 2000;6:971-978. 138. Mauceri HJ, Hanna NN, Beckett MA, et al. Combined effects of angiostatin and ionizing radiation in antitumour therapy. Nature 1998;394:287-291. 139. Milas L, Mason K, Hunter N, et al. In vivo enhancement of tumor radioresponse by C225 antiepidermal growth factor receptor antibody. Clin Cancer Res 2000;6:701-708. 140. Geng L, Donnelly E, McMahon G, et al. Inhibition of vascular endothelial growth factor receptor signaling leads to reversal of tumor resistance to radiotherapy. Cancer Res 2001;61:24132419. 141. Cao C, Albert JM, Geng L, et al. Vascular endothelial growth factor tyrosine kinase inhibitor AZD2171 and fractionated radiotherapy in mouse models of lung cancer. Cancer Res 2006;66:11409-11415. 142. Gorski DH, Mauceri HJ, Salloum RM, et al. Potentiation of the antitumor effect of ionizing radiation by brief concomitant exposures to angiostatin. Cancer Res 1998;58:5686-5689. 143. Zips D, Krause M, Hessel F, et al. Experimental study on different combination schedules of VEGF-receptor inhibitor PTK787/ZK222584 and fractionated irradiation. Anticancer Res 2003;23:3869-3876. 144. Williams KJ, Telfer BA, Brave S, et al. ZD6474, a potent inhibitor of vascular endothelial growth factor signaling, combined with radiotherapy: schedule-dependent enhancement of antitumor activity. Clin Cancer Res 2004;10:8587-8593. 145. Shibuya K, Komaki R, Shintani T, et al. Targeted therapy against VEGFR and EGFR with ZD6474 enhances the therapeutic efficacy of irradiation in an orthotopic model of human non–small-cell lung cancer. Int J Radiat Oncol Biol Phys 2007;69:1534-1543. 146. Winkler F, Kozin SV, Tong RT, et al. Kinetics of vascular normalization by VEGFR2 blockade governs brain tumor response to radiation: role of oxygenation, angiopoietin-1, and matrix metalloproteinases. Cancer Cell 2004;6:553-563.

136

PART 1  Principles

147. Johnson DH, Fehrenbacher L, Novotny WF, et al. Randomized phase II trial comparing bevacizumab plus carboplatin and paclitaxel with carboplatin and paclitaxel alone in previously untreated locally advanced or metastatic non–small-cell lung ­cancer. J Clin Oncol 2004;22:2184-2191. 148. Gordon MS, Margolin K, Talpaz M, et al. Phase I safety and pharmacokinetic study of recombinant human anti-vascular endothelial growth factor in patients with advanced cancer. J Clin Oncol 2001;19:843-850. 149. Willett CG, Boucher Y, di Tomaso E, et al. Direct evidence that the VEGF-specific antibody bevacizumab has antivascular effects in human rectal cancer. Nat Med 2004;10:145-147. 150. Willett CG, Boucher Y, Duda DG, et al. Surrogate markers for antiangiogenic therapy and dose-limiting toxicities for bevacizumab with radiation and chemotherapy: continued experience of a phase I trial in rectal cancer patients. J Clin Oncol 2005;23:8136-8139. 151. Crane CH, Varadhachary G, Wolff RA, et al. The argument for pre-operative chemoradiation for localized, radiographically resectable pancreatic cancer. Best Pract Res Clin Gastroenterol 2006;20:365-382. 152. Milas L. Cyclooxygenase-2 (COX-2) enzyme inhibitors as potential enhancers of tumor radioresponse. Semin Radiat Oncol 2001;11:290-299. 153. Milas L, Hanson WR. Eicosanoids and radiation [review]. Eur J Cancer 1995;31A:1580-1585. 154. Liao Z, Mason KA, Milas L. Cyclooxygenase-2 and its inhibition in cancer: is there a role? Drugs 2007;67:821-845. 155. Oshima M, Dinchuk JE, Kargman SL, et al. Suppression of intestinal polyposis in Apc716 knockout mice by inhibition of cyclooxygenase 2 (COX-2). Cell 1996;87:803-809. 156. Liu CH, Chang SH, Narko K, et al. Overexpression of ­cyclooxygenase-2 is sufficient to induce tumorigenesis in transgenic mice. J Biol Chem 2001;276:18563-18569. 157. Harris RE, Alshafie GA, Abou-Issa H, et al. Chemoprevention of breast cancer in rats by celecoxib, a cyclooxygenase 2 inhibitor. Cancer Res 2000;60:2101-2103. 158. Jacoby RF, Siebert K, Cole CE, et al. The cyclooxygenase-2 inhibitor celecoxib is a potent preventive and therapeutic agent in the min mouse model of adenomatous polyposis. Cancer Res 2000;60:5040-5044. 159. Khuri FR, Wu H, Lee JJ, et al. Cyclooxygenase-2 overexpression is a marker of poor prognosis in stage I non–small cell lung cancer. Clin Cancer Res 2001;7:861-867. 160. Sheehan KM, Sheahan K, O’Donoghue DP, et al. The relationship between cyclooxygenase-2 expression and colorectal cancer. JAMA 1999;282:1254-1257. 161. Gaffney DK, Holden J, Davis M, et al. Elevated cyclooxygenase-2 expression correlates with diminished survival in carcinoma of the cervix treated with radiotherapy. Int J Radiat Oncol Biol Phys 2001;49:1213-1217. 162. Arber N, Eagle CJ, Spicak J, et al. Celecoxib for the prevention of colorectal adenomatous polyps. N Engl J Med 2006;355: 885-895. 163. Stoner GD, Wang LS, Chen T. Chemoprevention of esophageal squamous cell carcinoma. Toxicol Appl Pharmacol 2007;224:337-349. 164. Sjödahl R. Nonsteroidal anti-inflammatory drugs and the gastrointestinal tract. Extent, mode, and dose dependence of anticancer effects [review]. Am J Med 2001;110:66S-69S. 165. Sandler AB, Dubinett SM. COX-2 inhibition and lung cancer. Semin Oncol 2004;31;(Suppl 7):45-52. 166. Steinbach G, Lynch PM, Phillips RK, et al. The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med 2000;342:1946-1952. 167. Thun MJ, Namboodiri MM, Calle EE, et al. Aspirin use and risk of fatal cancer. Cancer Res 1993;53:1322-1327. 168. Koki A, Leahy KM, Masferrer JL. Potential utility of COX-2 inhibitors in chemoprevention and chemotherapy. Exp Opin Investig Drugs 1999;8:1623-1638. 169. Milas L, Fujii T, Hunter N, et al. Enhancement of tumor radioresponse in vivo by gemcitabine. Cancer Res 1999;59:107-114.

170. Milas L, Kishi K, Hunter N, et al. Enhancement of tumor ­response to gamma-radiation by an inhibitor of cyclooxygenase2 enzyme. J Natl Cancer Inst 1999;91:1501-1504. 171. Petersen C, Petersen S, Milas L, et al. Human glioma cell radiosensitization by a selective COX-2 inhibitor. Clin Cancer Res 2000;6:2513-2520. 172. Pyo H, Choy H, Amorino GP, et al. A selective cyclooxygenase2 inhibitor, NS-398, enhances the effect of radiation in vitro and in vivo preferentially on the cells that express cyclooxygenase-2. Clin Cancer Res 2001;7:2998-3005. 173. Raju U, Aroga H, Dittmann K. Inhibition of DNA repair as a mechanism of enhanced radioresponse of head and neck carcinoma cells by a selective cyclooxygenase-2 inhibitor, celecoxib. Int J Radiat Oncol Biol Phys 2005;63:520-528. 174. Raju U, Nakata E, Yang P, et al. In vitro enhancement of tumor cell radiosensitivity by a selective inhibitor of cyclooxygenase2 enzyme: mechanistic considerations. Int J Radiat Oncol Biol Phys 2002;54:886-894. 175. Liao Z, Komaki R, Milas L, et al. A phase I clinical trial of ­thoracic radiotherapy and concurrent celecoxib for patients with unfa­ vorable performance status inoperable/unresectable non–small cell lung cancer. Clin Cancer Res 2005;11:3342-3348. 176. Bresalier RS, Sandler RS, Quan H, et al. Cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial. N Engl J Med 2005;352:1092-1102. 177. MacLachlan TK, Sang N, Giordano A. Cyclins, cyclin-dependent kinases and cdk inhibitors: implications in cell cycle control and cancer. Crit Rev Eukaryot Gene Expr 1995;5:127-156. 178. Senderowicz AM. Targeting cell cycle and apoptosis for the treatment of human malignancies. Curr Opin Cell Biol 2004;16: 670-678. 179. Zafonte BT, Hulit J, Amanatullah DF, et al. Cell-cycle dysregulation in breast cancer therapies targeting the cell cycle. Front Biosci 2000;5:D938-D961. 180. Arber N, Lightdale C, Rotterdam H, et al. Increased expression of the cyclin D1 gene in Barrett’s esophagus. Cancer Epidemiol Biomarkers Prev 1996;5:457-459. 181. Bani-Hani K, Martin IG, Hardie LJ, et al. Prospective study of cyclin D1 overexpression in Barrett’s esophagus: association with increased risk of adenocarcinoma. J Natl Cancer Inst 2000;92:1316-1321. 182. Grant WI, Woo SY. Clinical and financial issues for intensitymodulated radiation therapy delivery. Semin Radiat Oncol 1999;9:99-107. 183. Senderowicz AM. Small-molecule cyclin-dependent kinase modulators. Oncogene 2003;22:6609-6620. 184. Tsihlias J, Kapusta L, Slingerland J. The prognostic significance of altered cyclin-dependent kinase inhibitors in human cancer. Annu Rev Med 1999;50:401-423. 185. Keyomarsi K, Tucker SL, Buchholz TA, et al. Cyclin E and survival in patients with breast cancer. N Engl J Med 2002;347:15661575. 186. Arber N, Doki Y, Han EKH, et al. Antisense to cyclin D1 inhibits the growth and tumorigenicity of human colon cancer cells. Cancer Res 1997;57:1569-1574. 187. Kornmann M, Arber N, Korc M. Inhibition of basal and mitogenstimulated pancreatic cancer cell growth by cyclin D1 antisense is associated with loss of tumorigenicity and potentiation of cytotoxicity to cisplatinum. J Clin Invest 1998;101:344-352. 188. Sauter ER, Nesbit M, Litwin S, et al. Antisense cyclin D1 induces apoptosis and tumor shrinkage in human squamous carcinomas. Cancer Res 1999;59:4876-4881. 189. Zhou P, Jiang W, Zhang YJ, et al. Antisense to cyclin D1 inhibits growth and reverses the transformed phenotype of human esophageal cancer cells. Oncogene 1995;11:571-580. 190. Sedlacek HH. Mechanisms of action of flavopiridol. Crit Rev Oncol Hematol 2001;38:139-170. 191. Raju U, Nakata E, Mason KA, et al. Flavopiridol, a cyclin­dependent kinase inhibitor, enhances radiosensitivity of ovarian carcinoma cells. Cancer Res 2003;63:3263-3267. 192. Mason KA, Hunter NR, Raju U, et al. Flavopiridol increases therapeutic ratio of radiotherapy by preferentially enhancing ­tumor

5  Principles of Combining Radiation Therapy and Molecular Therapeutics radioresponse. Int J Radiat Oncol Biol Phys 2004;59:11811189. 193. Senderowicz AM, Sausville EA. Preclinical and clinical development of cyclin-dependent kinase modulators. J Natl Cancer Inst 2000;92:376-387. 194. Ariga H, Raju U, Liao Z, et al. Flavopiridol enhances radiosensitivity of human head and neck, and esophageal carcinoma cells [abstract 1357]. Proc Am Assoc Cancer Res 2004;45:311. 195. Senderowicz AM, Headlee D, Stinson SF, et al. Phase I trial of continuous infusion flavopiridol, a novel cyclin-dependent kinase inhibitor, in patients with refractory neoplasms. J Clin Oncol 1998;16:2986-2999. 196. Shapiro GI. Preclinical and clinical development of the ­cyclindependent kinase inhibitor flavopiridol. Clin Cancer Res 2004;10:4270S-4275. 197. Gupta AK, Bakanauskas VJ, Cerniglia GJ, et al. The Ras radiation resistance pathway. Cancer Res 2001;61:4278-4282. 198. Hahn SM, Bernhard EJ, Regine W, et al. A phase I trial of the farnesyltransferase inhibitor L-778,123 and radiotherapy for locally advanced lung and head and neck cancer. Clin Cancer Res 2002;8:1065-1072. 199. Martin NE, Brunner TB, Kiel KD, et al. A phase I trial of the dual farnesyltransferase and geranylgeranyltransferase inhibitor L-778,123 and radiotherapy for locally advanced pancreatic cancer. Clin Cancer Res 2004;10:5447-5454. 200. Willett CG, Safran H, Abrams RA, et al. Clinical research in pancreatic cancer: the Radiation Therapy Oncology Group trials. Int J Radiat Oncol Biol Phys 2003;56:31-37. 201. Moyal EC, Laprie A, Delannes M, et al. Phase I trial of tipifarnib (R115777) concurrent with radiotherapy in patients with glioblastoma multiforme. Int J Radiat Oncol Biol Phys 2007;68:13961401. 202. Munshi A, Meyn R. Combination of gene therapy with radiation. In Hunt K, Vorburger S, Swisher S (eds): Cancer Drug Discovery and Development: Gene Therapy for Cancer. Totowa, NJ: Humana Press, 2007, pp 243-256. 203. Collis SJ, Khater KDeWeese TL. Novel therapeutic strategies in prostate cancer management using gene therapy in combination with radiation therapy. World J Urol 2003;21:275-289. 204. Teh BS, Ayala G, Aguilar L, et al. Phase I-II trial evaluating combined intensity-modulated radiotherapy and in situ gene ­therapy with or without hormonal therapy in treatment of prostate ­cancer-interim report on PSA response and biopsy data. Int J Radiat Oncol Biol Phys 2004;58:1520-1529. 205. Freytag SO, Paielli D, Wing M, et al. Efficacy and toxicity of replication-competent adenovirus-mediated double suicide gene therapy in combination with radiation therapy in an orthotopic mouse prostate cancer model. Int J Radiat Oncol Biol Phys 2002;54:873-885. 206. Freytag SO, Stricker H, Pegg J, et al. Phase I study of ­replicationcompetent adenovirus-mediated double-suicide gene therapy in combination with conventional-dose three-dimensional conformal radiation therapy for the treatment of newly diagnosed, intermediate- to high-risk prostate cancer. Cancer Res 2003;63:7497-7506. 207. Yan B, Wang H, Rabbani ZN, et al. Tumor necrosis factor-alpha is a potent endogenous mutagen that promotes cellular transformation. Cancer Res 2006;66:11565-11570. 208. Weichselbaum RR, Kufe DW, Hellman S, et al. Radiation­induced tumour necrosis factor-alpha expression: clinical application of transcriptional and physical targeting of gene therapy. Lancet Oncol 2002;3:665-671. 209. Mezhir JJ, Smith KD, Posner MC, et al. Ionizing radiation: a genetic switch for cancer therapy. Cancer Gene Ther 2006;13:1-6. 210. Senzer N, Mani S, Rosemurgy A, et al. TNFerade biologic, an adenovector with a radiation-inducible promoter, carrying the human tumor necrosis factor alpha gene: a phase I study in patients with solid tumors. J Clin Oncol 2004;22:592-601. 211. Lang FF, Yung WK, Raju U, et al. Enhancement of radiosensitivity of wild-type p53 human glioma cells by adenovirus-­mediated delivery of the p53 gene. J Neurosurg 1998;89:125-132.

137

212. Sasaki R, Shirakawa T, Zhang ZJ, et al. Additional gene therapy with Ad5CMV-p53 enhanced the efficacy of radiotherapy in human prostate cancer cells. Int J Radiat Oncol Biol Phys 2001;51:1336-1345. 213. Swisher SG, Roth JA, Komaki R, et al. Induction of p53­regulated genes and tumor regression in lung cancer patients after intratumoral delivery of adenoviral p53 (INGN 201) and radiation therapy. Clin Cancer Res 2003;9:93-101. 214. Vecil GG, Lang FF. Clinical trials of adenoviruses in brain tumors: a review of Ad-p53 and oncolytic adenoviruses. J Neurooncol 2003;65:237-246. 215. Rogulski KR, Freytag SO, Zhang K, et al. In vivo antitumor activity of ONYX-015 is influenced by p53 status and is augmented by radiotherapy. Cancer Res 2000;60:1193-1196. 216. Geoerger B, Grill J, Opolon P, et al. Potentiation of radiation therapy by the oncolytic adenovirus dl1520 (ONYX-015) in human malignant glioma xenografts. Br J Cancer 2003;89:577-584. 217. Lamfers ML, Grill J, Dirven CM, et al. Potential of the conditionally replicative adenovirus Ad5-Delta24RGD in the ­ treatment of malignant gliomas and its enhanced effect with radiotherapy. Cancer Res 2002;62:5736-5742. 218. Lamfers ML, Idema S, Bosscher L, et al. Differential effects of combined Ad5-{Delta}24RGD and radiation therapy in in vitro versus in vivo models of malignant glioma. Clin Cancer Res 2007;13:7451-7458. 219. Vidal L, Blagden S, Attard G, et al. Making sense of antisense. Eur J Cancer 2005;41:2812-2818. 220. Dritschilo A, Huang CH, Rudin CM, et al. Phase I study of ­liposome-encapsulated c-raf antisense oligodeoxyribonucleotide infusion in combination with radiation therapy in patients with advanced malignancies. Clin Cancer Res 2006;12:1251-1259. 221. Tebes SJ, Kruk PA. The genesis of RNA interference, its potential clinical applications, and implications in gynecologic cancer. Gynecol Oncol 2005;99:736-741. 222. Dunn GP, Bruce AT, Ikeda H, et al. Cancer immunoediting: from immunosurveillance to tumor escape. Nat Immunol 2002;3:991-998. 223. Milas L, Scott MT. Antitumor activity of Corynebacterium parvum. Adv Cancer Res 1978;26:257-306. 224. Mihich E, Fefer A (eds): Biological Response Modifiers: Subcommittee Report. Monograph no. 63. Washington, DC: National Cancer Institute, 1983. 225. Tokunaga T, Yamamoto T, Yamamoto S. How BCG led to the discovery of immunostimulatory DNA. Jpn J Infect Dis 1999;52:1-11. 226. Krieg AM, Yi A, Matson S, et al. CpG motifs in bacterial DNA trigger direct B-cell activation. Nature 1995;374:546-549. 227. Sparwasser T, Vabulas RM, Villmow B, et al. Bacterial CpGDNA activates dendritic cells in vivo: T helper cell-independent cytotoxic T cell responses to soluble proteins. Eur J Immunol 2000;30:3591-3597. 228. Uhlmann E, Vollmer J. Recent advances in the development of immunostimulatory oligonucleotides. Curr Opin Drug Discov Devel 2003;6:204-217. 229. Baines J, Celis E. Immune-mediated tumor regression induced by CpG-containing oligodeoxynucleotides. Clin Cancer Res 2003;9:2693-2700. 230. Krieg AM. Antitumor applications of stimulating toll-like receptor 9 with CpG oligodeoxynucleotides. Curr Oncol Rep 2004;6:88-95. 231. Mason KA, Ariga H, Neal R, et al. Targeting toll-like receptor 9 with CpG oligodeoxynucleotides enhances tumor response to fractionated radiotherapy. Clin Cancer Res 2005;11:361-369. 232. Milas L, Mason KA, Ariga H, et al. CpG oligodeoxynucleotide enhances tumor response to radiation. Cancer Res 2004;64: 5074-5077. 233. Koski GK, Czerniecki BJ. Combining innate immunity with radiation therapy for cancer treatment.Clin Cancer Res 2005;11:7-11. 234. Sutherland RM. Radiation-enhanced immune response to cancer: Workshop, Anaheim, CA, April 17, 2005. Int J Radiat Oncol Biol Phys 2006;64:3-5.

The Skin

6

B. Ashleigh Guadagnolo, K. Kian Ang, and Matthew T. Ballo ANATOMY OF THE SKIN RESPONSE OF THE SKIN TO IRRADIATION Acute Reactions Chronic Reactions ETIOLOGY AND EPIDEMIOLOGY OF SKIN CANCER PREVENTION AND SCREENING CLINICAL MANIFESTATION, PATHOBIOLOGY, AND PATHWAYS OF SPREAD Basal Cell Carcinoma Squamous Cell Carcinoma Melanoma Merkel Cell Carcinoma Adnexal Carcinomas PATIENT EVALUATION AND STAGING Carcinoma Melanoma

PRIMARY THERAPY AND RESULTS Carcinoma Melanoma Merkel Cell Carcinoma Adnexal Carcinomas TREATMENT OF ADVANCED DISEASE AND PALLIATIVE THERAPY PRINCIPLES OF RADIATION THERAPY Carcinoma Melanoma Merkel Cell and Adnexal Carcinomas Patient Care during and after Radiation Therapy

Basal and squamous cell carcinomas of the skin are the most common types of cancer in the United States, with an estimated annual incidence of about 1 million cases.1 These carcinomas most often affect people with fair complexions or outdoor occupations, or both. Skin carcinomas may be disfiguring in certain locations but are rarely lethal. Although the exact figures are difficult to ascertain, the number of people who will die of cutaneous carcinomas in a given year is estimated to be between 2500 and 5000. Cutaneous melanoma is largely a disease of individuals with fair complexions who have a tendency to sunburn even after relatively brief exposure to sunlight.3,4 Approximately 59,940 cases of cutaneous melanoma were diagnosed in the United States in 2007.1 Despite progress in early detection, the incidence rates of malignant melanoma more than ­doubled between 1980 and 1990 and continue to increase rapidly.5,6 At present, 1 in 49 men and 1 in 73 women can be expected to develop malignant melanoma during their lifetime.1 Increasing exposure to sunlight during ­recreational activities and additional ultraviolet (UV) radiation reaching  Earth’s surface are considered strong contributing factors to the epidemic. Fortunately, because of better public ­awareness of the disease, most new cases of melanoma are diagnosed at early stages. Approximately 8110 people in the United States died of cutaneous melanoma in 2007.1 Less common types of skin cancer include Merkel cell carcinoma and adnexal carcinomas (i.e., sebaceous gland, eccrine, and apocrine carcinomas). Merkel cell carcinoma, also known as primary cutaneous neuroendocrine carcinoma or trabecular carcinoma of the skin, is a rare cancer originally described by Toker in 1972.7 Sebaceous gland carcinoma accounts for less than 1% of all cases of skin ­cancer.8 Eccrine (sweat gland) carcinoma is rare, ­accounting

for less than 0.01% of malignant epithelial neoplasms of the skin.9 Apocrine carcinoma is equally rare. Etiologic and epidemiologic data on Merkel cell and adnexal carcinomas are scarce. Likewise, information on prevention and early detection of these lesions is scanty.

Anatomy of the Skin The skin is composed of three layers: epidermis, dermis, and subcutaneous tissue. The epidermis is made up of stratified squamous epithelium that in most regions is only 0.05 to 0.15 mm thick. Cells in the basal layer of the epidermis regularly undergo mitotic division to replace epidermal cells lost from the surface. A thin basement membrane separates the epidermis from the dermis. The dermis, which is 1 to 2 mm thick, is composed of lymphatics, nerves, blood vessels, and sweat and sebaceous glands in a connective tissue stroma. The portion of the dermis in contact with the epidermis is referred to as the papillary layer. The deeper portion of the dermis, which contains bundles of collagenous and elastic fibers, is called the reticular layer. Subcutaneous tissue lies underneath the dermis and supports the lymphatics, blood vessels, and nerves in loose connective tissue containing variable amounts of fat.

Response of the Skin to Irradiation The extent of acute and chronic reactions of the skin to ­irradiation depends on several variables, including radiation dose and energy, number of days over which the dose is delivered, field size, and anatomic location.

141

142

PART 2  Skin

Acute Reactions Mitotic inhibition appears 30 minutes after a single dose of 4 to 5 Gy and is followed by a decrease in the proliferation of the germinal layer of the endothelium. The ­pathologic correlate of the erythema that begins to appear in the ­radiation therapy field within a few hours after irradiation is capillary dilatation and increased capillary ­permeability.10 Higher-energy radiation requires a larger total dose than lower-energy radiation to produce a given degree of  erythema.11 If radiation is given in several fractions (as ­opposed to a single fraction), a higher total dose is required to cause erythema because of repair of injury between fractions. As the radiation therapy continues, swelling and proliferation of capillary endothelial cells appear. In the arterioles, the tunica intima and tunica media are ­disrupted after 2 to 3 weeks of radiation therapy. The basal cells enlarge and show nuclear pyknosis.11 Maturing keratinocytes also show swelling, nuclear pyknosis, and cytoplasmic vacuolization. The moist desquamation phase of acute radiodermatitis begins 3 to 4 weeks after treatment begins and manifests as vascular dilatation and hyperemia, edema, and extravasation of erythrocytes and leukocytes. Approximately 4 weeks after treatment begins, small superficial blisters may form, coalesce, and rupture. Similar acute changes are seen in the adnexal structures of the skin and are caused by death of the rapidly dividing germinal layer of the epithelium. In hair follicles, germinal epithelium dies after a single-fraction dose of 4 to 5 Gy, but hair loss does not occur until approximately 3 weeks after treatment because the hairs adhere to the follicles. The more active the hair growth, the more sensitive the hair follicles; hair follicles in the chin in men and the scalp in both sexes are more sensitive to radiation than are hair follicles at other locations. The epithelial changes that develop in sebaceous and eccrine glands are similar to those in hair follicles, with sebaceous glands being about as sensitive as hair follicles and eccrine glands being slightly less so; these changes are responsible for the skin dryness that develops during treatment. Epidermal necrosis and ulceration can occur as early as 2 months after the start of radiation therapy. Ulcers rarely develop when standard radiation therapy techniques are used; when soft tissue necrosis does occur, it is typically associated with trauma or infection. If irradiation produces mild necrosis of the subepidermis, the new epidermis that forms will be thinner than normal. If the necrosis is sufficient to cause disappearance of the basement membrane beneath the basal cells, the new epidermis will be highly susceptible to trauma, infection, and other stresses. With more extensive necrosis associated with greater damage to blood vessels and connective tissue, repair may be largely by secondary intention (i.e., replacement fibrosis). When severe and deep necrosis of subepidermal supporting tissues and larger blood vessels occurs, an ulcer may take months or even years to heal.

Chronic Reactions After radiation therapy, the decrease in the number of small blood vessels and the increase in the density and amount of fibrous tissue in irradiated regions gradually progress over time.10 The damaged vasculature and connective tissue continue to deteriorate, partly because of the treatment

and partly because of aging and other acquired insults. The irradiated epidermis remains chronically dry because of permanent damage to the sebaceous and eccrine glands. Microscopic atrophy of the epidermis occurs, with loss of epidermal rete (interpapillary) ridges and dermal papillae, a condition known as poikiloderma.11 The basal layer shows few mitoses. Capillaries, venules, and lymphatic vessels are reduced in number and often show telangiectases. Telangiectases result from loss of microvascular endothelial cells and damage to the basement membrane, leading to contractions of multiple capillary loops into a single distorted, sinusoidal channel. Total blood flow in the irradiated skin is reduced.

Etiology and Epidemiology of Skin Cancer The main cause of cutaneous carcinoma is exposure to sunlight.1 Other contributory factors include exposure to chemical carcinogens (e.g., arsenic, tar, anthracene, crude paraffin oil), chronic irritation or inflammation, and ionizing radiation. Several genetic syndromes are also associated with a higher incidence of cutaneous carcinoma. Individuals with xeroderma pigmentosum are prone to the development of basal cell and squamous cell carcinomas because of defective repair of UV light–induced deoxyribonucleic acid (DNA) damage.12 The basal cell nevus syndrome is a genetic form of basal cell carcinoma inherited through an autosomal dominant gene.13 People with epidermodysplasia verruciformis tend to develop squamous cell carcinomas within large verrucous plaques in the third and fourth decades of life.14 Like cutaneous carcinomas, cutaneous melanomas tend to develop in sun-exposed areas in fair-skinned people who burn rather than tan when outdoors.3 ­Blistering sunburns during childhood increase the risk of melanoma later in life.15 A patient with melanoma has a 3% to 5% lifetime risk of developing a second lesion, which is 900 times the risk of the general population.16-18 A patient with multiple dysplastic nevi or the uncommon familial form of melanoma has an even higher lifetime risk (4% to 10%) of developing melanoma.3,16 Individuals with ­occupational exposure to polyvinyl chlorides19 or chemicals used in brewing and leather processing20 are at increased risk for developing melanoma. The incidence and aggressiveness of skin cancer are greatly increased in immunosuppressed individuals (e.g., organ transplant recipients, individuals with acquired immunodeficiency syndrome [AIDS]).

Prevention and Screening Skin cancer is an important target for preventive and screening measures because most cancer is caused by exposure to UV light and the impact of early detection and treatment on mortality and quality of life, particularly in melanoma, has been well documented. In the late 1990s, a panel of the American College of Preventive Medicine (ACPM) thoroughly reviewed the relevant medical literature and issued practice policy statements with regard to skin cancer prevention.21 Recommended preventive measures include avoiding exposure to sunlight, particularly by limiting time spent outdoors between 10 am and 4 pm, and wearing protective physical barriers such as hats and

6  The Skin

c­ lothing. If exposure to sunlight cannot be limited because of occupational, cultural, or other factors, the use of sunscreen that is opaque or blocks UVA and UVB is recommended. The ACPM recommends discussing sun avoidance and sun protection measures with children and teenagers. For high-risk individuals, such as those with fair skin, multiple nevi, a family history of skin cancer, or a personal history of skin cancer, the ACPM also recommends periodic screening, including 2- to 3-minute visual inspection of the entire integument, by an adequately trained physician.

Clinical Manifestations, Pathobiology, and Pathways of Spread Basal Cell Carcinoma Clinical Presentation and Pathology Basal cell carcinoma most commonly arises in the head and neck region, where it may manifest as an asymptomatic nodule, a pruritic plaque, or a bleeding sore that waxes and wanes.22,23 Multiple synchronous lesions may be present. Approximately one third of patients treated for basal cell carcinoma develop at least one additional basal cell carcinoma.24 Careful evaluation of sun-exposed skin should be part of the follow-up examination. Each of the several subtypes of basal cell carcinoma has distinctive histologic and clinical features. The main subtypes are summarized in the following sections. Nodulo-ulcerative Subtype. Nodulo-ulcerative basal cell carcinoma, also referred to as a rodent ulcer because of its deeply infiltrating or burrowing nature, is the most commonly observed subtype. It begins as a papule. Central ulceration develops as the lesion grows. The margins of the lesion appear waxy and contain enlarged capillaries. Histologically, large islands of monomorphous basaloid cells, varying in size and shape, are embedded in a fibroblastic stroma in the dermis. The basaloid cells have large, oval, hyperchromatic nuclei, scant cytoplasm, and no intercellular bridges. The peripheral cell layer of the aggregate often demonstrates palisading, whereas the central cells are less organized. Depending on the number of melanocytes in the lesion, nodulo-ulcerative basal cell carcinomas may be blue, brown, or black. Consequently, they may be difficult to differentiate clinically from malignant melanomas.25 ­Biopsy may be required to establish the correct diagnosis. Superficial Subtype. Superficial basal cell carcinomas present as red, scaly macules with an indistinct margin and are usually located on the trunk. They enlarge into a ­crusted, erythematous patch without induration and may be ­difficult to differentiate clinically from solar keratosis, psoriasis, squamous cell carcinoma in situ, or ­extramammary  Paget’s disease. Histologic examination shows multiple foci of neoplastic cells with peripheral palisading originating from the undersurface of the epidermis.25 Morphea-like Subtype. Morphea-like basal cell carcinomas, also referred to as sclerosing basal cell carcinomas, appear as flat, indurated, ill-defined macules. The lesions generally have a smooth, shiny surface that becomes depressed as the lesions grow. Histologically, morphea-like basal cell carcinomas are characterized by the presence of small groups of narrow strands of basaloid cells embedded

143

in a dense, fibrous connective tissue. Little peripheral palisading occurs.25 Infiltrative Subtype. Infiltrative basal cell carcinomas have an opaque, yellowish appearance and blend subtly into the surrounding skin.26 Histologically, the lesions are characterized by poorly circumscribed cell aggregates near the skin surface. Neoplastic cells infiltrate the reticular ­dermis and subcutis. Basosquamous Subtype. Basosquamous carcinomas, also known as metatypical or keratotic basal cell carcinomas, occur almost exclusively on the face and are uncommon. Histologically, they exhibit features of basal and squamous cell carcinomas. Terebrant Subtype. Terebrant carcinomas manifest as small, ulcerated lesions and typically arise in the triangular region of the cheeks extending from the lower eyelids to the nasolabial folds.27 Histologic examination usually reveals extensive infiltration of deeper structures by small nests of undifferentiated basal cells. This type of lesion may invade the orbit, nose, or maxilla, resulting in significant deformity. It should be treated at an early stage whenever possible.

Biology and Pattern of Spread Between weeks 4 and 8 of embryologic development, anatomic regions grow together, and the covering epithelium subsequently migrates across junctional sites known as embryologic fusion planes. It has commonly been accepted that epithelial carcinomas spread along these embryologic fusion planes such as the nasolabial groove or the infranasal depression.28,29 However, controversy has arisen as to whether embryologic fusion planes persist in adults and influence the spread of skin tumors.30 The perineural space, the cleavage plane between a nerve and its sheath, provides a route of spread for skin cancer.31 When perineural invasion is present at diagnosis,32,33 basal cell carcinomas are often advanced, and they tend to recur. Peripheral branches of the trigeminal and facial nerves are most commonly involved and provide a route for contiguous spread to the central nervous system.34-36 In such cases, skip areas, areas where no carcinoma is identified in the pathologic specimen, may be present along the nerve between the primary tumor and the site of central nervous system involvement.35 About two thirds of patients with cutaneous carcinoma with pathologic evidence of perineural invasion have no symptoms of this invasion.35,37-39   Compression of a nerve may develop over a period of years, causing numbness, formication, burning, stinging, facial weakness, ptosis, or blurred or double vision.35-38,40 Intraneural invasion is rare.31,37,41 The biologic behavior of basal cell carcinomas varies with the histologic subtype. Nodulo-ulcerative basal cell carcinomas grow slowly over many years in people with normal immunity. They spread by peripheral and deep invasion, particularly along the perineural space, perichondrium, and periosteum. Superficial basal cell carcinomas may remain stable for a long period or enlarge gradually over a number of years. In contrast, morphea-like, infiltrative, basosquamous, and terebrant carcinomas behave more aggressively. Because the epidermis lacks lymphatics and blood vessels, basal cell carcinomas rarely spread to regional or distant sites. The overall incidence of metastasis is less than

144

PART 2  Skin

0.01%.42 Two thirds of metastases are detected in the ­regional lymphatics, and up to 20% of metastases involve the bones, lungs, or liver.43 Metastases usually arise from large, ulcerated lesions in the head and neck region that have recurred after repeated surgery and radiation therapy.44

Squamous Cell Carcinoma Clinical Presentation and Pathology Squamous cell carcinomas arise from malignant transformation of keratinocytes in the epidermis. This type of skin cancer most commonly develops as a result of damage from sunlight.45 Much less often, it arises in an area of trauma or chronic irritation, such as a thermal burn scar (i.e., Marjolin’s ulcer)46 or a draining osteomyelitic sinus.47 The three main types of squamous cell carcinoma are squamous cell carcinoma in situ (i.e., Bowen’s disease), superficial squamous cell carcinoma, and infiltrating squamous cell carcinoma. Pathologically, squamous cell carcinoma in situ exhibits a verrucous or papillated border. Atypical  keratinocytes are confined to the epidermis. Clinically, squamous cell carcinoma in situ manifests as a soft, erythematous, scaly, circumscribed patch. Superficial squamous cell carcinoma represents atypical keratinocyte proliferation confined to the upper reticular dermis, whereas infiltrating squamous cell carcinoma denotes extension into and beyond the lower reticular epidermis. Clinically, infiltrating squamous cell carcinoma manifests as a firm, ulcerated mass with an elevated, nodular border. Histologically, squamous cell carcinoma is described as well, moderately, or poorly differentiated on the basis of the magnitude of cellular polymorphism, keratinization, and mitosis. Diagnosis of the spindle cell subtype may ­require immunoperoxidase studies.

Biology and Pattern of Spread Regional and distant metastases are uncommon in squamous cell carcinoma and are rare in basal cell carcinoma. At the time of diagnosis, metastatic spread is present in approximately 2% of patients with squamous cell carcinoma.48 Factors determining the likelihood of metastasis include degree of differentiation, prior treatment, greatest dimension of the lesion, depth of invasion, and anatomic location.49,50 Poorly differentiated or recurrent squamous cell carcinomas and those less than 3 cm in diameter, less than 4 mm thick, or arising on the lips are associated with lymph node metastases at diagnosis in approximately 10% of cases.49,51-56 Reports conflict about whether perineural spread by itself is associated with an increased risk of lymph node metastases.38,55,57-59 Squamous cell carcinomas that arise in burn scars or osteomyelitic foci are associated with lymph node metastases at diagnosis in 10% to 30% of cases.54

Melanoma Clinical Presentation and Pathology Few diseases have the extraordinary metastatic potential of malignant melanoma. Beginning as a superficial pigmented skin lesion, melanoma grows horizontally within the epidermal and superficial dermal layers. This rather innocuous phase of growth, referred to as the radial growth phase, is followed after some time by a vertical growth

phase, ­ characterized by extension to deeper dermal tissues and vessels. When the disease is caught during early growth, surgery is curative. With every millimeter of vertical growth, however, the rate of nodal spread increases, as does the likelihood of distant metastatic spread—a stage of the disease that is almost uniformly fatal. Primary cutaneous melanoma may develop from preinvasive lesions such as lentigo maligna melanomas or dysplastic nevi, or it may arise de novo in normal skin. The four major subtypes are described in the following sections.60 Superficial Spreading Melanoma. The most prevalent subtype is superficial spreading melanoma, which comprises  about 70% of cases.61,62 This subtype can occur in adults of any age and affects women more often than men. ­Superficial spreading melanoma often arises in a junctional nevus, where it first appears as a deeply pigmented lesion that ­ progresses gradually to a flat, indurated macule over a ­ period of years. Patches of amelanotic areas within the lesion are evident, apparently representing focal regression. As the lesion grows, the border may become irregular. Histologically, superficial spreading melanomas are characterized by a prominent intraepidermal proliferation of malignant melanocytes. The malignant cells resemble those of Paget’s disease, and this pattern is called pagetoid melanoma.63 The malignant cells may be confined to the lower portion of the epidermis or may spread up into the granular cell layer of the epidermis, which is often hyperplastic. As the lesion enlarges, clusters of malignant cells invade the dermis and subcutaneous tissue. Nodular Melanoma. Nodular melanoma is the second most common subtype, comprising 15% to 25% of ­cases.61,62 Nodular melanoma most often arises on the trunk or in the head and neck region and most often ­occurs in middle-aged individuals. In contrast to superficial spreading melanoma, nodular melanoma affects men more often than women. It manifests as a raised or dome-shaped, blue or black lesion. In general, nodular melanoma is darker than superficial spreading melanoma. About 5% of nodular melanomas manifest as nonpigmented, fleshy nodules and therefore are referred to as amelanotic melanomas. Histologically, nodular melanoma appears as expanding lesions involving the papillary dermis with little or no epidermal component. Lesions are composed of epithelioid cells. Spindle cells and small epithelioid cells may also be present. Deeper invasion of the dermis and subcutaneous tissue occurs as lesions grow. Lentigo Maligna Melanoma. The lentigo maligna subtype of melanoma comprises less than 10% of cases.62,64 This subtype occurs mainly on the face or neck of fairskinned women older than 50 years of age.65 It typically manifests as a relatively large (>3 cm), flat, tan-colored (with different shades of brown) lesion that has been present for longer than 5 years. The border becomes irregular as the lesion enlarges. Histologically, lentigo maligna lesions are usually composed of spindle-like cells, which may be embedded in a connective tissue stroma with a desmoplastic pattern or may form fascicles displaying neural features and infiltrating the perineural space.65,66 Acral Lentiginous Melanoma. Acral lentiginous melanoma arises characteristically on the palms or soles or ­beneath the nail beds. However, some volar and plantar lesions are superficial spreading or nodular melanomas.67-71 The relative frequency of acral lentiginous melanoma varies

10-year mortality rate (%)

100 80 60 40

Observed Predicted

20 0 0

1

2

3

4

5

6

7

8

9

10 11 12 13

Tumor thickness (mm)

FIGURE 6-1.  Observed (blue circles) and predicted (blue line) 10-year mortality rates of 15,320 patients with clinically localized melanoma are based on a mathematical model f(t)= 1�������� –������� 0.988e (–211t + 0.009t2) derived from the American Joint Committee on Cancer (AJCC) melanoma database. T is the measured thickness (mm) and f(t) is the 10-year melanoma-specific mortality rates (P < .0001). (From Balch CM, Soong S, Gershenwald JE, et al. Prognostic factors analysis of 17,600 melanoma patients: validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol 2001;19:3622-3634.)

substantially with race. This subtype comprises only about 5% of melanomas in fair-skinned people but 35% to 60% in dark-skinned individuals.72-74 Most acral lentiginous melanomas are brown and occur on the soles in individuals older than 60 years of age. Lesions generally evolve over a period of years and reach an average diameter of 3 cm before the diagnosis is established.68,70 Histologically, early-stage acral lentiginous melanoma is composed of large, highly atypical, pigmented cells along the dermoepidermal junction in an area of hyperplasia. At the invasive stage, infiltrating cells may be epithelioid or spindle-shaped.66 Infiltration may be present at diagnosis; if so, it occurs predominantly through the eccrine ducts.63

Biology and Pattern of Spread Although no single model of tumor growth and dissemination applies in all cases of malignant melanoma, growth tends to be stepwise from primary to lymphatic to distant site. Superficial spreading and lentigo maligna melanomas generally grow slowly over a period of many years (i.e., radial growth phase). Left untreated, these tumors gradually invade the dermis and subcutaneous tissue (i.e., vertical growth phase). In contrast, acral lentiginous and nodular melanomas progress more rapidly to a vertical growth phase. The lymphatic system is particularly important in the dissemination of malignant melanoma. The rate of regional lymph node metastasis and the rate of distant metastasis can be predicted based on the thickness of the primary ­lesion. The risk of nodal spread is 4% for lesions up to 1 mm thick, 14% for lesions 1.01 to 2.0 mm thick, 29% for lesions 2.01 to 4.0 mm thick, and 45% for lesions more than 4.0 mm thick. The risk of distant metastases also increases with the thickness of the primary lesion (Fig. 6-1)75 and with the number of involved lymph nodes (Fig. 6-2).76 What thickness cannot predict is the exact site or extent of future spread. Melanoma has an almost unique ability to disseminate early to combinations of dermal, ­subcutaneous,

Probability of distant metastasis (PDM)

6  The Skin 1.0

145

1 PDM = –(0.450+1.232log(n)) 1+e

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

5

10

15

20

25

30

35

Number of positive nodes (n)

FIGURE 6-2.  The observed (blue circles) and predicted (blue line) probabilities of developing distant metastases increase according to the number of involved lymph nodes. The best-fit logistic regression equation is shown. (From Ballo MT, Ross MI, Cormier JN, et al. Combined-modality therapy for patients with regional nodal metastases from melanoma. Int J Radiat Oncol Biol Phys 2006;64:106-113.)

lymphatic, and visceral sites that are not amenable to most forms of therapy.

Merkel Cell Carcinoma Clinical Presentation and Pathology Merkel cell carcinoma, which is also known as trabecular carcinoma or neuroendocrine carcinoma of the skin, is a rare, aggressive neoplasm that was first described by Toker7 in 1972. It typically presents as a firm, painless, red, pink, or blue cutaneous nodule involving the head and neck or extremities in people older than 60 years.77,78 The median size of primary lesions at diagnosis is less than 2 cm, and the incidence of nodal involvement at presentation is approximately 20%.79 The differential diagnosis includes poorly differentiated squamous cell carcinoma, melanoma, eccrine carcinoma, lymphoma, Langerhans’ cell histiocytosis, and metastatic cancer, including islet cell carcinoma, medullary carcinoma of the thyroid, and small cell carcinoma of the lung. The cell of origin is believed to be the dermal neurotactile Merkel cell, arising from the neural crest.80 The lesion usually involves the reticular dermis and subcutaneous tissue, with minimal extension to the papillary dermis. The neoplastic cells possess cytoplasmic extensions and small, uniform, membrane-bound neurosecretory granules.81 Several immunohistochemical markers have been identified in Merkel cell carcinoma. The neoplastic cells stain most consistently with polyclonal antisera to neuron-specific enolase and calcitonin; stains based on monoclonal antibodies bind to neurofilament and cytokeratin.82-84

Biology and Pattern of Spread Merkel cell carcinoma has an aggressive course characterized by a propensity for contiguous spread and a high incidence of lymphatic and hematogenous dissemination.78,79,85-87 The traditional treatment, wide local excision, results in locoregional recurrence in almost three fourths of patients88;

146

PART 2  Skin

in contrast, with surgery plus radiation therapy, the locoregional recurrence rate is only 15%.77,79,84,89-91 Assessment of patterns of failure in a series of 54 patients with Merkel cell carcinoma treated at The University of Texas M. D. Anderson Cancer Center revealed local and regional relapse rates of 35% and 67%, respectively.79 Distant metastases occurred in 32 patients (59%). The median time to recurrence after surgical resection was 5 months, and 91% of relapses occurred within the first year. One patient died of uncontrolled primary disease, and 32 patients died of distant metastases. The 5-year disease-free survival rate was 11% for patients treated with surgery alone and 56% for patients treated with surgery and postoperative radiation therapy (P = .002). Similarly, Meeuwissen and colleagues89 reported that the 3-year disease-free survival rate was 0% for 38 patients treated with surgery alone and 68% for 34 patients treated with surgery and postoperative radiation therapy.

Adnexal Carcinomas Sebaceous Carcinoma Clinical Presentation and Pathology. Sebaceous carcinomas arise most commonly from the sebaceous glands of the ocular adnexa, the meibomian glands, and the glands of Zeis in the lid margin and the caruncle. Sites of origin, in decreasing frequency, are the head and neck, trunk and extremities, and external genitalia.92-95 Histologically, sebaceous carcinomas are composed of dermal-based lobules and cords of tumor cells with various degrees of sebaceous differentiation and infiltration into the surrounding tissues. Differentiated cells contain foamy or vacuolated, slightly basophilic cytoplasm, whereas less differentiated cells contain more deeply basophilic cytoplasm. Clinical and pathologic features indicative of a poor prognosis include size greater than 1 cm (associated with a 5-year mortality rate of 50%), vascular and lymphatic invasion, poor sebaceous differentiation, highly infiltrative growth pattern, and carcinomatous changes in the overlying epithelium.96 Among the malignant neoplasms of the eyelids, sebaceous carcinomas are second only to basal cell ­carcinomas in frequency and second only to malignant ­ melanomas in lethality.96 The most common location of ocular ­sebaceous carcinomas is the upper eyelid; the second most common location is the lower eyelid and caruncle.97 ­Sebaceous ­carcinomas predominantly affect people older than 60 years, with a slight preponderance for women. Sebaceous carcinomas manifest as slowly growing, nontender, deep-seated, firm nodules. Often, they produce a chalazion or manifestations of inflammatory processes such as conjunctivitis, blepharitis, or tarsitis. Small primary tumors may be overlooked until they invade the orbit or spread to preauricular or cervical nodes. The differential diagnosis includes sebaceous adenoma and basal cell carcinoma with sebaceous differentiation (i.e., basosebaceous  epithelioma).97,98 Distinction between basosebaceous epithelioma and sebaceous carcinoma is important because of the difference in the biologic behavior of the two tumors. Biology and Pattern of Spread. Sebaceous carcinoma behaves more aggressively than squamous cell carcinoma, with a propensity for nodal and hematogenous spread. Of 104 patients with sebaceous carcinoma registered by the Ophthalmic Pathology Branch of the Armed Forces

I­ nstitute of Pathology, 23 patients (22%) experienced one local recurrence and 11 patients experienced two or more local recurrences. Twenty-three patients died of metastatic disease.96

Eccrine Carcinoma Clinical Presentation and Pathology. Eccrine (sweat gland) carcinomas arise most often in the skin of the head and neck or the extremities and usually manifest as slowgrowing, painless papules or nodules.99 Eccrine carcinomas are usually diagnosed in people between 50 and 70 years of age.100 Histologically, eccrine carcinomas resemble carcinomas of the breast, bronchus, and kidney and may be difficult to distinguish from cutaneous metastases.99,101 Histologic subtypes include ductal eccrine carcinoma, mucinous eccrine carcinoma, porocarcinoma, syringoid eccrine carcinoma, clear  cell carcinoma, and microcystic adnexal carcinoma.99-106 Biology and Pattern of Spread. Eccrine carcinomas behave more aggressively than do basal cell and squamous cell carcinomas. In a series of 14 patients with eccrine carcinoma who underwent surgical excision with negative margins, 11 had at least one local recurrence, and 5 had recurrence in regional nodes or distant sites. One patient died of an uncontrolled local relapse, and four patients died of distant metastases 2 months to 10 years after diagnosis.99

Apocrine Carcinoma Apocrine carcinoma arises most commonly from apocrine glands in the axilla. It can also originate from apocrine glands in the vulva or eyelids and from ceruminal glands in the external auditory canal.107-109 Apocrine carcinoma usually manifests as a reddish purple nodule in elderly individuals. Because of the rarity of this lesion, its natural history is not well understood. Nodal and distant spread have been reported.110

Patient Evaluation and Staging Clinical evaluation of skin cancer consists of inspection and palpation of the involved area and the regional lymph nodes. Imaging studies, such as chest radiographs, computed tomography scans, and magnetic resonance imaging scans, should be obtained when clinically indicated.

Carcinoma The 2002 American Joint Committee on Cancer (AJCC) tumor-nodes-metastasis (TNM) staging system for carcinoma of the skin (Box 6-1) is based primarily on clinical findings.111

Melanoma Two surgical microstaging systems are used for melanoma: the Breslow system and the Clark method. Both systems are based on the recognition that for patients with melanomas, the prognosis depends on the depth of invasion rather than the greatest dimension of the tumor. The Breslow system classifies lesions by vertical thickness between the granular layer of the epidermis and the deepest part of the lesion as measured with an ocular  micrometer. For ulcerated lesions, measurements are made from the surface of the skin to the deepest part of the primary tumor.112

6  The Skin

BOX 6-1.  American Joint Committee on Cancer Staging System for Carcinoma of the Skin (Excluding Eyelid, Vulva, and Penis) Primary Tumor (T) TX Primary tumor cannot be assessed T0 No evidence of primary tumor Tis Carcinoma in situ T1 Tumor 2 cm or less in greatest dimension T2 Tumor more than 2 cm, but not more than 5 cm, in greatest dimension T3 Tumor more than 5 cm in greatest ­dimension T4 Tumor invades deep extradermal structures (i.e., cartilage, skeletal muscle, or bone) Regional Lymph Nodes (N) NX Regional lymph nodes cannot be assessed N0 No regional lymph node metastasis N1 Regional lymph node metastasis Distant Metastasis (M) MX Distant metastasis cannot be assessed M0 No distant metastasis M1 Distant metastasis Stage Grouping Stage 0 Tis Stage I T1 Stage II T2 T3 Stage III T4 Any T Stage IV Any T

N0 N0 N0 N0 N0 N1 Any N

M0 M0 M0 M0 M0 M0 M1

Note: In case of multiple simultaneous tumors, the tumor with the highest T category is classified, and the number of separate tumors is indicated in parentheses, such as T2 (5). From Greene FL, Page DL, Fleming ID, et al (eds): AJCC Cancer Staging Manual, 6th ed. New York: Springer, 2002, pp 203-208.

The Clark method divides melanomas into five groups on the basis of the depth of invasion relative to normal tissue structures. Level I is confined to the epidermis, level II invades the papillary dermis, level III invades the papillaryreticular dermal interface, level IV invades the reticular dermis, and level V invades the subcutaneous tissue.61 The 2002 AJCC staging system for melanoma  (Box 6-2) includes a four-stage grouping based on the Breslow thickness and the presence or absence of primary ­tumor ulceration, nodal involvement, and satellitosis or distant metastasis to the lung or other viscera. The Clark level of invasion is significant only for lesions less than 1 mm thick. A fifth group—stage 0, melanoma in situ—was ­ added in 1997.113

Primary Therapy and Results Carcinoma Primary Tumor The general strategy for treatment of the primary tumor is the same for basal cell and squamous cell carcinomas. A variety of treatment approaches are available, including curettage and electrodesiccation (not recommended for

147

r­ ecurrent lesions), Mohs’ surgery, cryotherapy, surgical resection, and primary radiation therapy.114-117 Nearly all treatment approaches result in 5-year recurrence rates of 1% to 10% for previously untreated cutaneous carcinomas.118 The choice of treatment for individual patients is determined on the basis of factors such as the location and size of the lesion; expected functional and cosmetic results; treatment time and cost; and patient age, occupation, and general physical condition.114 Surgical modalities are preferred for most patients, particularly those with small lesions, for whom simple surgical procedures yield high cure rates. Primary radiation therapy is most often indicated for lesions on or around the nose, lower eyelids, and ears, where radiation therapy yields better functional and cosmetic results than surgery. Similarly, primary radiation therapy is indicated for extensive lesions of the cheek, lip, or oral commissure that would require full-thickness resection. In a series of more than 1000 patients with previously untreated and recurrent basal and squamous carcinomas of the eyelid who underwent primary radiation therapy, 5-year local control rates were more than 90% (Table 6-1).119 The size of the primary lesion is the major determinant of local control after radiation therapy. In a series of 646 patients with carcinomas of the eyelid, pinna, nose, and lip treated with ­radiation, the 10-year local control rates were 98% for tumors measuring 2 cm or smaller, 79% for 2- to 5-cm lesions, and 53% (at 8 years) for carcinomas larger than 5 cm.120 When results are adjusted for tumor size, local control rates seem to be slightly higher for basal cell carcinoma than for squamous cell carcinoma: 97% versus 91% for lesions smaller than 1 cm and 87% versus 76% for 1- to 5-cm lesions.121 Recurrent ­basal and squamous cell carcinomas are more difficult to control than are previously untreated lesions.89,117,122,123 Radiation therapy usually is not recommended for ­patients younger than 50 years because the late effects after radiation therapy increase over time, as does the risk of a second cancer developing within the radiation portal.124,125 Postoperative radiation therapy may be beneficial in several clinical situations, including positive surgical margins, perineural involvement, invasion of bone or cartilage, extensive skeletal muscle infiltration, a positive node measuring more than 3 cm in the greatest dimension, extranodal spread, and multiple positive nodes. Series published in the 1950s and 1960s reporting the use of orthovoltage x-rays and large daily doses of radiation126-128 suggested that radiation therapy for carcinomas involving bone or cartilage resulted in an excessive risk of osteoradionecrosis or radiochondritis. However, data from more recent series using more fractionated, higher-energy radiation indicate that radiation therapy may be the treatment of choice for selected patients with bone or cartilage involvement because the risk of complications is low when proper techniques are used.129,130

Regional Nodes Elective nodal treatment is not indicated for cutaneous basal cell carcinomas because of the low incidence of lymphatic spread. However, elective nodal treatment is ­indicated for large, infiltrative, ulcerative squamous cell carcinomas or recurrent squamous cell carcinomas.131 The choice of treatment for patients who present with nodal disease depends on the type of therapy selected for

148

PART 2  Skin

BOX 6-2.  American Joint Committee on Cancer Staging System for Melanoma of the Skin Primary Tumor (T) TX T0 Tis T1 T1a T1b T2 T2a T2b T3 T3a T3b T4 T4a T4b Regional Lymph Nodes (N) NX N0 N1 N1a N1b N2 N2a N2b N2c N3 Distant Metastasis (M) MX M0 M1 M1a M1b M1c Clinical Stage Grouping Stage 0 Stage IA Stage IB Stage IIA Stage IIB Stage IIC Stage III

Stage IV Pathologic Stage Grouping Stage 0 Stage IA Stage IB Stage IIA Stage IIB Stage IIC Stage IIIA

Primary tumor cannot be assessed (e.g., shave biopsy or regressed melanoma) No evidence of primary tumor Melanoma in situ Melanoma ≤ 1.0 mm in thickness with or without ulceration Melanoma ≤ 1.0 mm in thickness and level II or III, no ulceration Melanoma ≤ 1.0 mm in thickness and level IV or V or with ulceration Melanoma 1.0-2.0 mm in thickness with or without ulceration Melanoma 1.0-2.0 mm in thickness, no ulceration Melanoma 1.0-2.0 mm in thickness, with ulceration Melanoma 2.01-4.0 mm in thickness with or without ulceration Melanoma 2.01-4.0 mm in thickness, no ulceration Melanoma 2.01-4.0 mm in thickness, with ulceration Melanoma greater than 4.0 mm in thickness with or without ulceration Melanoma > 4.0 mm in thickness, no ulceration Melanoma > 4.0 mm in thickness, with ulceration Regional lymph nodes cannot be assessed No regional lymph node metastasis Metastasis in one lymph node Clinically occult (microscopic) metastasis Clinically apparent (macroscopic) metastasis Metastasis in two to three regional nodes or intralymphatic regional metastasis without nodal metastasis Clinically occult (microscopic) metastasis Clinically apparent (macroscopic) metastasis Satellite or in-transit metastasis without nodal metastasis Metastasis in four or more regional nodes, or matted metastatic nodes, or in-transit metastasis or satellites with metastasis in regional nodes Distant metastasis cannot be assessed No distant metastasis Distant metastasis Metastasis to skin, subcutaneous tissues or distant lymph nodes Metastasis to lung Metastasis to all other visceral sites or distant metastasis at any site associated with an ­elevated serum lactic dehydrogenase (LDH) Tis T1a T1b T2a T2b T3 T3b T4 T4b Any T Any T Any T Any T

N0 N0 N0 N0 N0 N0 N0 N0 N0 N1 N2 N3 Any N

M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M1

Tis T1a T1b T2a T2b T3a T3b T4a T4b T1-4a T1-4a

N0 N0 N0 N0 N0 N0 N0 N0 N0 N1a N2a

M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0

6  The Skin

149

BOX 6-2.  American Joint Committee on Cancer Staging System for Melanoma of the Skin—Cont’d Pathologic Stage Grouping—Cont’d Stage IIIB T1-4b T1-4b T1-4a T1-4a T1-4a/b Stage IIIC T1-4b T1-4b Any T Stage IV Any T

N1a N2a N1b N2b N2c N1b N2b N3 Any N

M0 M0 M0 M0 M0 M0 M0 M0 M1

From Greene FL, Page DL, Fleming ID, et al (eds): AJCC Cancer Staging Manual, 6th ed. New York: Springer, 2002, pp 209-220.

TABLE 6-1.    Control Rates for Carcinomas of the Eyelid Treated with Radiation Therapy Number of Cases Treated Histologic Subtype

Total

Previously Untreated

Recurrent

5-Year Local Control Rate

Basal cell carcinoma

1062

686

376

1009 of 1062 (95%)

Squamous cell ­carcinoma Total

104

62

42

97 of 104 (92%)

1166

748

418

1106 of 1166 (95%)

From Fitzpatrick PJ, Thompson GA, Easterbrook WM, et al. Basal and squamous cell carcinoma of the eyelids and their treatment by radiotherapy. Int J Radiat Oncol Biol Phys 1984;10:449-454.

the primary lesion and the size of the involved lymph nodes. In general, single-modality therapy (surgery or ­ radiation therapy) is sufficient for nodal disease up to 3 cm without evidence of extracapsular spread.131 Taylor and colleagues132 reported results from 36 ­patients who underwent surgery and postoperative radiation therapy for lymph node metastases to the parotid ­region from cutaneous carcinomas. The overall regional control rate was 89%. All four recurrences occurred in ­ patients with positive surgical margins and clinical evidence of facial nerve involvement. For a later series of 37 patients with parotid-area metastases treated in the same way, Audet and associates133 reported a locoregional control rate of 73% and a 3-year disease-specific survival rate of 72%. In that analysis, patients with facial nerve involvement and tumors larger than 6 cm experienced poorer outcomes. Veness and colleagues,56 reporting on 74 patients with nonparotid cervical metastases, found that surgery combined with postoperative radiation therapy led to lower relapse rates and better disease-free survival times than did single-modality treatment for cervical metastases.

Melanoma For patients with localized cutaneous melanoma (stage  I or II), wide local excision of the primary tumor is the standard treatment. For patients with known lymph node metastases at presentation (stage III), wide local excision plus therapeutic lymph node dissection is the accepted surgical approach. Use of adjuvant radiation therapy is based on the clinical and pathologic features of the disease.

Primary Tumor The standard treatment for localized cutaneous melanoma (stages I and II) is wide local excision. Pathologic examination of the specimen is essential for establishment of a ­tissue diagnosis, assessment of the margins, and ­microstaging. The

recommended skin margins are 1 cm for invasive lesions less than 1 mm thick and for lentigo maligna melanomas and 2 to 3 cm for melanomas at least 1 mm thick. Melanoma thickness is the most powerful determinant of local control after wide local excision. In a large multi­institutional trial, long-term local recurrence rates were 2.3% for melanomas 1 to 2 mm thick, 4.2% for melanomas 2 to 3 mm thick, and 11.7% for melanomas 3 to 4 mm thick.134 Local recurrence is a poor prognostic sign; most patients with local recurrence ultimately die of the disease. Radiation therapy is rarely indicated as the definitive treatment for primary malignant melanomas, except in the case of large facial lentigo maligna melanomas, for which wide local excision may necessitate extensive reconstruction. In a series of patients with lentigo maligna melanomas treated with primary radiation therapy at the Princess Margaret Hospital and followed for a period of 6 months to 8 years (median, 2 years), local control was achieved in 23 (92%) of 25 patients.135 The median time to complete regression of lesions was 8 months, although some lesions took 2 years to disappear. Radiation therapy was delivered with orthovoltage x-rays (100 to 250 keV) and consisted of 35 Gy in 5 fractions over 1 week for lesions smaller than  3 cm, 45 Gy in 10 fractions over 2 weeks for lesions ­measuring 3 to 4.9 cm, and 50 Gy in 15 to 20 fractions over 3 to 4 weeks for lesions measuring 5.0 cm or more.

Regional Nodes Elective Lymph Node Dissection. For many years, the most controversial aspect of melanoma management was the role of elective lymph node dissection (ELND), particularly the management of the clinically uninvolved nodal basin. After wide local excision alone, the risk of nodal failure for patients with intermediate-thickness melanoma was up to 50%, and it was believed that early removal of microscopic nodal disease would avoid nodal recurrence

150

PART 2  Skin

and prevent distant spread. In 1982, results of a prospective, nonrandomized study at the Sydney Melanoma Unit involving 1319 patients suggested that ELND improved the survival of patients with intermediate-thickness (0.76 to 4 mm) melanomas.136 Subsequent trials, however, have not confirmed these results. The World Health Organization (WHO) Melanoma Program conducted a prospective phase III trial in which 553 patients with stage I disease were randomly assigned to receive wide local excision and ELND or wide local excision and delayed therapeutic lymphadenectomy (i.e., lymph node dissection only if regional nodes became clinically detectable).137 No difference in the survival rates was found between the two groups. Two other multicenter trials, one sponsored by the National Cancer Institute of the United States (Intergroup Melanoma Surgical Program)138 and the other sponsored by the WHO Melanoma Program,139 were launched in an effort to resolve the controversy. The Intergroup Surgical trial included 740 patients with clinically localized melanomas 1.0 to 4.0 mm thick. No significant difference in 5-year overall survival was found between patients assigned to ELND and those assigned to observation (86% versus 82%, P = .25). The WHO Melanoma Program trial included 240 eligible patients with truncal melanomas at least 1.5 mm thick with no clinical evidence of regional or distant metastases. Multivariate analysis revealed that routine use of immediate ELND had no effect on 5-year overall survival (hazard ratio, 0.72; 95% confidence interval [CI], 0.5 to 1.02). Since that time, Morton and colleagues140 developed a minimally invasive intraoperative lymphatic mapping technique that allows identification of the sentinel node, defined as the first node in the regional nodal basin into which the primary tumor drains. With this procedure, patients with clinically occult nodal disease who may benefit from early lymph node dissection can be identified.141 The results of an international randomized study to assess the efficacy of selective lymphadenectomy became available in 2006.142 In that study, 1269 patients with cutaneous melanomas between 1.2 and 3.5 mm thick were randomly assigned to undergo wide local excision followed by nodal observation and delayed dissection if necessary or sentinel lymph node biopsy, with an immediate completion lymph node dissection performed if micrometastases were found in the sentinel lymph node. The 5-year survival rate was 72% for patients presenting with an involved node, compared with 90% in the patients presenting with an uninvolved node (P < .001). Although the melanoma-specific survival rates were similar for the two groups, the diseasefree survival rate was better in the sentinel lymph node group (78% versus 73%, P = .009). Some therapeutic value was evident for completion lymph node biopsy in patients with a positive sentinel node. Among the patients with nodal disease, the 5-year survival rate was higher in the group who underwent immediate lymphadenectomy than in the group who were observed but required subsequent lymphadenectomy for nodal recurrence (72% versus 52%, P = .004). The group requiring salvage dissection also had higher numbers of involved lymph nodes at the time of dissection (3.3 versus 1.4, P < .001). Sentinel lymph node biopsy is now the standard of care for patients with 1.2- to 3.5-mm thick melanoma. For patients with thinner lesions,

the procedure is performed if the primary tumor is ulcerated, classified as Clark level IV or V, or associated with elevated mitotic counts. Therapeutic Lymph Node Dissection. Therapeutic nodal dissection with wide local excision is the accepted surgical approach for patients with clinically apparent nodal metastases from melanoma.143 Surgery alone is sufficient for more than 85% of patients with stage III disease. However, for patients with nodal extracapsular extension (including matted lymphadenopathy), four or more involved lymph nodes, lymph nodes larger than 3 cm, or recurrent nodal disease after previous dissection, the risk of regional recurrence can range from 30% to 50%.144,145 Although most such patients die as a consequence of distant spread, which usually manifests less than 2 years after treatment,146 locoregional relapses are significant because they often do not respond to therapy and can confer substantial morbidity through pain, infection, and disfigurement. Locoregional relapse from cutaneous melanoma of the head and neck can be particularly distressing because of compression of vital organs. Prevention of locoregional failure is desirable from a quality-of-life standpoint, even though overall survival may not be significantly improved.

Radiation Therapy Melanoma has long had a reputation for being radioresistant, although the origins of this supposed characteristic remain somewhat obscure. Laboratory studies have confirmed that melanoma cell lines have an enhanced ability to repair sublethal damage from radiation. The labeling of melanoma as radioresistant, however, predates such studies. Most likely, early clinical experience with treating visibly and palpably advanced dermal or nodal metastases resulted in visibly and palpably progressive disease inside and outside the radiation therapy field. Whereas for most diseases, radiologic studies are the mainstay of surveillance for detecting progressing disease, melanoma tends to involve superficial sites easily observed on physical examination. Several retrospective studies have revealed response rates of about 80% when radiation is delivered in large doses per fraction (i.e., 4 Gy or more), suggesting that if treated appropriately, melanoma is responsive to radiation therapy.147-149 The Radiation Therapy Oncology Group launched a study in 1983 to compare the effects of four fractions of  8 Gy given at weekly intervals with those of 20 fractions of 2.5 Gy given daily.150 A total of 137 patients were enrolled; although metastatic disease at any site other than the abdomen or brain was allowed, most was in the soft tissue, skin, or lymph nodes. Approximately one half of the lesions in this trial were larger than 5 cm in the greatest diameter, and response was measured clinically and radiographically. Although no difference was found in complete or partial response rates according to fraction size, the overall response rate in both treatment groups was 60%. Investigators at M.  D. Anderson Cancer Center have long questioned the reputed radioresistance of melanoma and recommended radiation therapy for appropriately selected patients. Before the advent of routine sentinel lymph node ­biopsy, elective radiation therapy was the preferred approach for treating primary melanoma of the head and neck ­ region that was more than 1.5 mm thick or Clark level IV or V because such therapy effectively suppressed

6  The Skin

regional ­ recurrence and was associated with less morbidity than elective lymph node dissection. In a retrospective analysis from M.  D. Anderson, Bonnen and colleagues151 reviewed 157 patients with stage I or II cutaneous melanoma of the head and neck who received elective ­regional radiation instead of lymph node dissection after wide  local excision of the primary site. Indications for regional radiation included primary tumor thickness of 1.5 mm or more or Clark level IV or V disease. The results revealed 15 regional failures (for a regional control rate of 89% at 10 years), despite estimates that 33 to 40 patients actually had microscopically involved regional nodes. Six percent of patients required medical care for a clinically significant complication, with moderate hearing loss being the most common complaint (six patients). Although elective nodal radiation suffers from the same limitations as ELND, it can effectively provide regional control for patients at risk of regional recurrence while avoiding surgical dissection. We reserve elective radiation therapy for patients with significant medical comorbid conditions for whom detailed prognostic information is of little relevance and enrollment on clinical trials is unlikely. For patients with nodal extracapsular extension, four or more involved lymph nodes, lymph nodes measuring over 3 cm, or recurrent nodal disease after previous dissection, therapeutic radiation results in excellent regional control. In a review of 466 patients with nodal metastases treated with lymphadenectomy and radiation (with or without systemic therapy), the actuarial 5-year regional (in-basin) control rate for all patients was 89%.76 On multivariate analysis, no patient or disease characteristics were found to be associated with inferior regional control. For this same group of patients, the 5-year disease-specific, disease-free, and distant metastasis-free survival rates were 49%, 42%, and 44%, respectively. Multivariate analysis indicated that increasing numbers of involved lymph nodes and the presence of ulceration in the primary tumor were associated with inferior 5-year actuarial disease-specific and distant metastasis-free survival rates. The number of involved lymph nodes was also associated with the development of brain metastases. The risk of lymphedema was highest for those patients with groin lymph node metastases.76

Systemic Adjuvant Therapy Because distant metastasis is the main pattern of relapse in melanoma, numerous attempts have been made to ­develop effective systemic therapy for this disease. However, none of the regimens tested has been found to improve overall survival in a phase III setting, with the exception of ­interferonalpha (IFN-α). An Eastern Cooperative ­ Oncology Group trial enrolled 280 patients with the following disease characteristics: primary melanoma more than 4 mm thick without palpable nodes, metastasis detected at ELND, primary melanoma of any stage with clinically ­ palpable regional lymph nodes, or nodal recurrence at any interval after surgery for primary melanoma of any depth.152 Patients received IFN intravenously five times each week for 4 weeks and then subcutaneously three times each week for 48 weeks. This treatment significantly increased 5-year actuarial relapse-free survival rates (37% versus 26%, P = .0023) and 5-year actuarial overall survival rates  (46% ­versus 37%, P = .024). The overall benefit of treatment in this trial correlated with tumor burden and the presence

151

of palpable regional lymph node metastases. The benefit of IFN therapy was greatest among patients with palpable nodal metastases at presentation or nodal recurrence. A French multi-institutional trial accrued 489 eligible patients with no palpable lymph nodes and no evidence of distant metastasis who underwent wide local excision of melanomas more than 1.5 mm thick.153 This phase III trial showed that patients treated with subcutaneous IFN three times per week for 18 months had a significantly ­ longer ­relapse-free interval than that of patients in the control group (P = .035). The hazard ratio for relapse in treated patients was 0.75 (95% CI, 0.57 to 0.98), and the 5-year relapse rate was 43% in the IFN group and 51% in the group that did not receive IFN. There was a trend toward improved overall survival in IFN-treated patients compared with the control group (P = .059). IFN remains the only systemic therapy approved by the U.S. Food and Drug Administration for patients with high-risk disease.154,155 Cisplatin-based chemotherapy has resulted in response rates of 12% to 30% in randomized trials, with no improvement in the overall survival, among patients with ­metastatic melanoma.156 Numerous phase II trials have combined tamoxifen with different chemotherapeutic agents. Because results have been conflicting, the benefit of tamoxifen remains unclear.156 A large number of melanoma vaccines are in develop­ ment or clinical testing.157 Autologous and allogeneic ­tumor cell vaccines yielded interesting results in initial studies. However, subsequent randomized trials (e.g., trials of vaccinia melanoma oncolysate vaccine) did not show improved survival. Progress in immunology and molecular biology has led to a newer generation of vaccines that are undergoing clinical testing, including cell-, ganglioside-, peptide-, and DNA-based vaccines.157

Merkel Cell Carcinoma The initial treatment for Merkel cell carcinoma is usually surgery to establish a tissue diagnosis. The role of radiation therapy in the management of Merkel cell carcinoma has been evaluated since the early 1980s because of the poor locoregional control rates obtained with surgery alone. The radiosensitivity of Merkel cell carcinoma has been demonstrated in several studies.77,79,84,89,90,158 Pacella and colleagues84 reported that local control was achieved in 35 of 36 fields irradiated in 19 patients (gross disease was present in 23 of the fields). Similarly, in-field control was obtained in 30 of 31 patients treated in a series at M.  D. Anderson.79 However, marginal recurrences developed in three ­patients,79 indicating that the target volume should include a wide margin around gross disease. Between 1982 and 1987, as the natural history of Merkel cell carcinoma became more apparent, elective postoperative locoregional radiation therapy was given to five of six patients referred to M.  D. Anderson after local excision.79 The primary tumor bed and draining lymphatics were irradiated with an electron beam to cumulative doses of 50 to 55 Gy and 45 to 55 Gy, respectively, in daily 2.0- to 2.5-Gy fractions. At a minimum follow-up time of 3 years, none of these patients had experienced a local or regional recurrence. Although patient numbers in this study were small because of the rarity of Merkel cell carcinoma, many ­retrospective series support the use of surgery and postoperative radiation therapy for previously untreated or

152

PART 2  Skin

recurrent primary tumors.57,77,79,84,88,123,158-161 The available data are insufficient to permit detailed dose-response analysis. However, in-field recurrence seems to be rare after a total dose of 50 Gy to microscopic disease given in daily 2.0- to 2.5-Gy fractions.79,161 Evidence of the responsiveness of Merkel cell carcinoma to cytotoxic agents continues to emerge.162-164 The most active drugs are doxorubicin, cyclophosphamide, and vincristine. The Trans-Tasman Radiation Oncology Group reported the results of a phase II study of radiation therapy with concurrent carboplatin and etoposide for patients with localized, high-risk Merkel cell carcinoma. Three-year outcomes were promising, with an overall survival rate of 76%, a locoregional control rate of 75%, and a distant control rate of 76%.165 However, a comparison of the study cohort with a cohort of similar historical controls showed that chemotherapy did not significantly affect overall survival or disease-specific survival rates.166 Responses of ­metastatic deposits to multiagent chemotherapy can be dramatic, but unfortunately, such responses are often of short duration.164,167 The treatment policy for localized Merkel cell carcinoma at M.  D. Anderson can be summarized as follows. For patients presenting with no evidence of regional or distant metastases, surgical excision of the primary tumor should be followed in all cases by adjuvant radiation therapy to generous margins around the primary site. Although elective irradiation of the draining lymphatics had been recommended previously,117 current guidelines call for routine use of sentinel lymph node biopsy.168 If the sentinel node is involved, radiation therapy to the regional basin is preferred to surgical dissection. Given the radiosensitivity of Merkel cell carcinoma,79,84,158 patients who for medical reasons cannot undergo surgery are treated with primary radiation therapy to higher doses with elective nodal irradiation. For gross, unresected disease, 56 to 66 Gy is recommended.79 For patients presenting with regional lymph node metastases, our current practice is local excision of the primary tumor, therapeutic nodal dissection, and postoperative irradiation of the surgical bed, including the draining lymphatics. Chemotherapy is also considered for those patients with large primary tumors or adenopathy.

Adnexal Carcinomas The standard treatment for sebaceous carcinoma is wide local excision and therapeutic nodal dissection if lymphadenopathy is present. The role of adjuvant therapeutic modalities has not been explored systematically. As is the case for other types of skin cancer, indications for postoperative irradiation include positive surgical margins, perineural spread, invasion of bone or cartilage, extensive skeletal muscle infiltration, a positive node measuring more than  3 cm in the greatest dimension, extranodal spread, and multiple positive nodes. For eccrine and apocrine carcinomas, wide excision of the primary lesion is accepted as the standard treatment for patients with no lymphadenopathy at diagnosis. Postoperative radiation therapy to the surgical bed with 2- to 3-cm margins is indicated, particularly for eccrine carcinoma, because of the high incidence of local recurrence after surgery alone. The role of elective treatment of regional lymphatics is controversial. Treatment for patients with palpable lymphadenopathy consists of wide local excision,

lymph node dissection, and postoperative irradiation of the surgical bed.

Treatment of Advanced Disease and Palliative Therapy Radiation therapy administered to total doses of 60 to  70 Gy (depending on tumor size, location, and anticipated toxicity) can result in local control in some patients with advanced skin carcinomas. Lee and colleagues169 ­reported a 5-year local control rate of 67% in a group of 33 patients with previously untreated T4 skin carcinomas who underwent radiation therapy to doses of 60 to 75 Gy over 6 to 8 weeks. Bone invasion or perineural spread reduced the likelihood of local control in this small series of patients. Radiation therapy can reduce symptoms for most patients with locally advanced inoperable or metastatic skin cancer. The recommended dose-fractionation schedules depend on the tumor location and the patient’s life expectancy. The most commonly used regimens are 30 Gy in 10 fractions over 2 weeks, 45 to 50 Gy in 18 to 25 fractions over 4 to 5 weeks, and 36 Gy in 6 fractions over 3 weeks. The third of these regimens is used for the treatment of melanoma. For patients with one or two brain metastases, particularly from melanoma, and good performance status with stable extracranial disease, stereotactic radiosurgery (15 to 20 Gy in a single fraction) is delivered to treat the brain metastases.170

Principles of Radiation Therapy The general principles of radiation therapy for skin cancer are discussed in this section. The technical details of radiation treatment for skin cancer are described elsewhere.171

Carcinoma In radiation therapy for skin carcinoma, the target volume is determined mainly by the size of the lesion and the type of growth. Infiltrative lesions and those with ill-defined borders (e.g., morphea-like carcinomas) require more generous margins. Orthovoltage x-rays or an electron beam characteristically is used. With the electron beam, the field size at the skin surface is 0.5 cm larger because of constriction of the isodose lines in the depth.117 Typically, the portal is designed to encompass the visible or palpable tumor, along with 0.5 to 1.0 cm of surrounding normal-appearing skin in the case of carcinomas smaller than 1 cm and up to 2.0 cm of normal-appearing skin for larger or ill-defined carcinomas.172 The field size may be reduced after the delivery of a dose sufficient to sterilize microscopic disease. Margins may be smaller when areas close to the eye are treated. In some patients—almost exclusively those with squamous cell carcinoma who present with small (≤3 cm) involved lymph nodes or advanced primary tumors—the initial field is designed to encompass the primary tumor and regional lymphatics. The target volume is subsequently reduced to cover the areas of gross disease with adequate margins. Perineural invasion typically extends 1 to 2 cm along the nerve sheath173 but may rarely extend up to 14 cm.37

6  The Skin

Consequently, in the presence of major nerve invasion, the target volume should encompass the involved nerve to the skull base.57,58,174,175 In patients undergoing postoperative irradiation for pathologically proven perineural invasion or lymph node involvement, the initial target volume includes the primary tumor bed plus the potential route of perineural spread or the regional lymphatics. The target volume is then reduced to include the primary tumor and nodal beds. Several fractionation schedules have been used to treat skin carcinomas. In general, the more protracted schedules produce the best cosmetic results.120,176 The recommended radiation regimen for individual patients depends on the size and location of the carcinoma. For most cases of skin cancer, 40 Gy in 10 fractions, 45 Gy in 15 fractions, or 50 Gy in 20 fractions are appropriate and effective regimens.120 If the patient is elderly and in generally poor health, 32 Gy in 4 fractions or a single dose of 20 Gy may be used. For large lesions close to the eye, 60 to 70 Gy in 30 to 35 fractions is recommended.120 Patients with positive nodes usually receive 2-Gy fractions because of the large irradiated volume. Similarly, postoperative radiation therapy is delivered using 2-Gy fractions. The dose for orthovoltage x-rays is prescribed to Dmax (the maximum point dose), whereas the dose for electrons is prescribed to the 90% isodose line to take into account the difference in relative biologic effectiveness between the beams.177,178

Melanoma The target volume for patients with stage I to III melanoma depends on the site of disease. For adjuvant radiation therapy to the primary tumor site, the surgical bed is treated with margins of at least 2 cm. For cervical nodal metastases, the ipsilateral neck nodes down to the clavicle are included within an appositional electron beam field with the patient in an open-neck position (Fig. 6-3). For axillary disease, anterior and posterior photon beams should include level I through III axillary nodes. Typically, the cervical nodes are included only if they are involved or if more than 10 axillary nodes are involved. For nodal metastases to the groin, anteriorly weighted anterior and posterior photon beams are used to treat the groin, with electron fields matched above and below as needed to cover the full extent of the surgical scar. Elective adjuvant radiation therapy is administered in 6-Gy fractions prescribed to Dmax twice per week to a total dose of 30 Gy. If microscopic residual disease is present, an additional fraction may be given through a smaller portal for a cumulative dose of 36 Gy. Care is taken to ensure that the dose to the spinal cord does not exceed 24 Gy in  4 fractions over 2 weeks.

Merkel Cell and Adnexal Carcinomas In radiation therapy for Merkel cell carcinoma, the target volume consists of the primary tumor bed with 4- to  5-cm margins, except when the lesion is situated at or close to critical structures (e.g., optic apparatus). In the case of Merkel cell carcinoma of the head and neck region, the whole ipsilateral neck is irradiated. In radiation therapy for adnexal carcinoma, the target volume consists of the ­primary tumor bed with 2- to 3-cm margins and the draining lymphatics.

153

FIGURE 6-3.  The typical electron field arrangement is shown for a patient with nodal metastases from melanoma. The patient presented with a nodal recurrence of melanoma several years after a superficial spreading primary lesion on the temporal scalp had been removed. Limited neck dissection was followed by adjuvant radiation therapy. The treatment was simulated with the patient in the indicated open-neck position, and the field encompassed the low parotid down to the supraclavicular lymph nodes. A beveled tissue equivalent bolus was used over the larynx and within the ear canal. A beveled bolus may also be applied over the temporal lobe, represented by a line drawn between the lateral canthus and the mastoid tip (not shown). The radiation dose was prescribed to Dmax.

Radiation is administered, preferably with electrons, in 2-Gy daily fractions. For Merkel cell carcinoma, a dose of 46 to 50 Gy is delivered to the region of elective radiation therapy, followed by a boost consisting of 10 Gy in 5 fractions to the gross disease. At M.  D. Anderson, a cumulative dose of 66 Gy is given in the case of bulky primary tumors. For adnexal carcinoma, a dose of 60 Gy usually is administered in the postoperative setting.

Patient Care during and after Radiation Therapy A tumoricidal dose of radiation usually produces moist desquamation. The skin reaction usually reaches a peak 3 to 6 weeks after the start of radiation therapy and resolves spontaneously within 6 weeks. Irradiated skin should be pro­ tected from heat, cold, sunlight, friction, and other sources of irritation to avoid additional tissue injury. Patients are instructed to use sun blocks with a sun protection factor of at least 15 over the irradiated area and, in the case of facial lesions, to wear a hat when outdoors after the completion of treatment. After moist desquamation occurs, the area should be cleaned with aluminum subacetate (Domeboro) soaks or 1% hydrogen peroxide to prevent secondary infection. Patients should also be informed about the specific acute and late side effects of treatment, such as mucositis of the nasal passages and lips, nasal dryness, synechiae (when nostril lesions are treated), conjunctivitis, and loss of eyelashes (when eyelid lesions are treated). When the lacrimal canaliculi are treated, irrigation of the ducts during the healing phase after treatment helps to prevent synechiae. Late side effects include hypopigmentation, hyperpigmentation, telangiectasia, and skin atrophy, which may

154

PART 2  Skin

develop gradually over a period of years. Hair loss and loss of sweat gland function are usually permanent. Irradiated skin is more sensitive to trauma. Blood flow in the irradiated tissues may decrease. As a result, healing may be delayed after surgery in an irradiated region. Skin retraction at the lower eyelid may result in ectropion. Fibrosis of the lacrimal duct may result in epiphora. These side effects are rarely serious after properly administered radiation therapy. Cataracts can occur after irradiation of lower-eyelid lesions, but this side effect is uncommon when appropriate shielding is used. The long-term cosmetic results after radiation therapy depend on the fraction size, total dose, and volume of tissue irradiated. When appropriate techniques, such as those described here, are used, the risk of soft tissue necrosis is less than 3%, and the risk of osteoradionecrosis is approximately 1%. Radiochondritis is rare and is usually a sign of recurrent disease.179 Regular follow-up examinations include evaluation of the treated area for late complications and—because one third of treated patients will develop metachronous skin cancers—inspection for new lesions.

REFERENCES   1. American Cancer Society. Cancer Facts & Figures—2007.  Atlanta: American Cancer Society, 2007.   2. Brown RO, Osguthorpe JD. Management of the neck in nonmelanocytic cutaneous carcinomas. Otolaryngol Clin North Am 1998;31:841-856.   3. Wang SQ, Halpern AC. Management of cutaneous melanoma: a public health and individual patient care perspective. Adv  Dermatol 2007;23:81-98.   4. Goldberg MS, Doucette JT, Lim HW, et al. Risk factors for presumptive melanoma in skin cancer screening: American Academy of Dermatology National Melanoma/Skin Cancer Screening Program experience 2001-2005. J Am Acad Dermatol 2007;57:60-66.   5. Rigel DS, Kopf AW, Friedman RJ.The rate of malignant melanoma  in the United States: are we making an impact? J Am Acad Dermatol 1987;17:1050-1053.   6. Rigel DS, Friedman RJ, Kopf AW. The incidence of malignant melanoma in the United States: issues as we approach the 21st century. J Am Acad Dermatol 1996;34:839-847.   7. Toker C. Trabecular carcinoma of the skin. Arch Dermatol 1972;105:107-110.   8. Urban FH, Winkelmann RK. Sebaceous malignancy. Arch Dermatol 1961;84:63-72.   9. Tulenko JF, Conway H. An analysis of sweat gland tumors. Surg Gynecol Obstet 1965;121:343-348. 10. Rubin P, Casarett G. Clinical Radiation Pathology; vol 2. ­Philadelphia: WB Saunders, 1968, p 1057. 11. Fajardo LF. ������������������������������������ Skin�������������������������������� . In Fajardo LF, Berthrong M, Anderson RE (eds): Radiation Pathology. New York: Oxford University Press, 2001, pp 411-420. 12. Cleaver JE. Defective repair replication of DNA in xeroderma pigmentosum. Nature 1968;218:652-656. 13. Anderson DE, Taylor WB, Falls HF, et al. The nevoid basal cell carcinoma syndrome. Am J Hum Genet 1967;19:12-22. 14. Lutzner MA. Epidermodysplasia verruciformis: an autosomal recessive disease characterized by viral warts and skin cancer. A model for viral oncogenesis. Bull Cancer 1978;65:169-182. 15. Lew RA, Sober AJ, Cook N, et al. Sun exposure habits in patients with cutaneous melanoma: a case control study. J Dermatol Surg Oncol 1983;9:981-986. 16. Bellet RE, Vaisman I, Mastrangelo MJ, et al. Multiple primary malignancies in patients with cutaneous melanoma. Cancer 1977;40:1974-1981.

17. Scheibner A, Milton GW, McCarthy WH, et al. Multiple primary melanoma—a review of 90 cases. Australas J Dermatol 1982;23:1-8. 18. Veronesi U, Cascinelli N, Bufalino R. Evaluation of the risk of multiple primaries in malignant cutaneous melanoma. Tumori 1976;62:127-130. 19. Lundberg I, Gustavsson A, Holmberg B, et al. Mortality and cancer incidence among PVC-processing workers in Sweden. Am J Ind Med 1993;23:313-319. 20. Linet MS, Malker HS, Chow WH, et al. Occupational risks for cutaneous melanoma among men in Sweden. J Occup Environ Med 1995;37:1127-1135. 21. Ferrini RL, Perlman M, Hill L. American College of Preventive Medicine practice policy statement: skin protection from ultraviolet light exposure. Am J Prev Med 1998;14:83-86. 22. Kopf AW. Computer analysis of 3531 basal cell carcinomas of the skin. J Dermatol 1979;6:267-281. 23. Rahbari H, Mehregan AH. Basal cell epitheliomas in usual and unusual sites. J Cutan Pathol 1979;6:425-431. 24. Robinson JK. Risk of developing another basal cell carcinoma: a 5-year prospective study. Cancer 1987;60:118-120. 25. Aasi SZ, Leffell DJ. Cancer of the skin. In DeVita VT, Hellman S,  Rosenberg SA (eds): Cancer: Principles and Practice of Oncology, 5th ed. Philadelphia: Lippincott Williams & Wilkins, 2005, pp 1717-1744. 26. Siegle RJ, MacMillan J, Pollack S. Infiltrative basal cell carcinoma: a nonsclerosing subtype. J Dermatol Surg Oncol 1986;12:830836. 27. Battle RJ, Patterson TJ. The surgical treatment of basal-celled carcinoma. Br J Plast Surg 1960;13:118-135. 28. Mohs FE, Lathrop TG. Modes of spread of cancer of the skin. AMA Arch Derm Syphilol 1952;66:427-439. 29. Panje WR, Ceilley RI. The influence of embryology of the midface on the spread of epithelial malignancies. Laryngoscope 1979;89:1914-1920. 30. Wentzell JM, Robinson JK. Embryologic fusion planes and the spread of cutaneous carcinoma: a review and reassessment.  J Dermatol Surg Oncol 1990;16:1000-1006. 31. Dodd GD, Dolan PA, Ballantyne AJ, et al. The dissemination of tumors of the head and neck via the cranial nerves. Radiol Clin North Am 1970;8:445-461. 32. Binkley GW, Rauschkolb RR. Basal-cell epithelioma metastasizing to lymph nodes. Arch Dermatol 1962;86:332-342. 33. von Domarus H, Stevens PJ. Metastatic basal cell carcinoma: report of five cases and review of 170 cases in the literature.  J Am Acad Dermatol 1984;10:1043-1060. 34. Carlson KC, Roenigk RK. Know your anatomy: perineural involvement of basal and squamous cell carcinoma on the face.  J Dermatol Surg Oncol 1990;16:827-833. 35. Hanke CW, Wolf RL, Hochman SA, et al. Chemosurgical ­reports: perineural spread of basal-cell carcinoma. J Dermatol Surg Oncol 1983;9:742-747. 36. Smith JB, Bishop VL, Francis IC, et al. Ophthalmic manifestations of perineural spread of facial skin malignancy. Aust NZ J Ophthalmol 1990;18:197-205. 37. Ballantyne AJ, McCarten AB, Ibanez M. The extension of cancer of the head and neck through peripheral nerves. Am J Surg 1963;106:651-657. 38. Goepfert H, Dichtel WJ, Medina JE, et al. Perineural invasion in squamous cell skin carcinoma of the head and neck. Am J Surg 1984;148:542-547. 39. Mendenhall WM, Parsons JT, Mendenhall NP, et al. Carcinoma of the skin of the head and neck with perineural invasion. Head Neck 1989;11:301-308. 40. Clouston PD, Sharpe DM, Corbett AJ, et al. Perineural spread of cutaneous head and neck cancer. Its orbital and central neurologic complications. Arch Neurol 1990;47:73-77. 41. Mark GJ. Basal cell carcinoma with intraneural invasion. Cancer 1977;40:2181-2187. 42. Paver K, Poyzer K, Burry N, et al. The incidence of basal cell carcinomas and their metastases in Australia and New Zealand [letter]. Australas J Dermatol 1973;14:53.

6  The Skin 43. Cannon JR, Schneidman DW. Recent developments in adnexal pathology. In Moschella SL (ed): Dermatology Update: Reviews for Physicians. New York: Elsevier, 1982, pp 217-252. 44. Soffer D, Kaplan H, Weshler Z. Meningeal carcinomatosis due to basal cell carcinoma. Hum Pathol 1985;16:530-532. 45. Epstein JH. Photocarcinogenesis, skin cancer, and aging. J Am Acad Dermatol 1983;9:487-502. 46. Akguner M, Barutcu A, Yilmaz M, et al. Marjolin’s ulcer and chronic burn scarring. J Wound Care 1998;7:121-122. 47. Hejna WF. Squamous-cell carcinoma developing in the chronic draining sinuses of osteomyelitis. Cancer 1965;18:128-132. 48. Epstein E, Epstein NN, Bragg K, et al. Metastases from squamous cell carcinomas of the skin. Arch Dermatol 1968;97:245-251. 49. Friedman HI, Cooper PH, Wanebo HJ. Prognostic and therapeutic use of microstaging of cutaneous squamous cell carcinoma of the trunk and extremities. Cancer 1985;56:1099-1105. 50. Immerman SC, Scanlon EF, Christ M, et al. Recurrent squamous cell carcinoma of the skin. Cancer 1983;51:1537-1540. 51. Binder SC, Cady B, Catlin D. Epidermoid carcinoma of the skin of the nose. Am J Surg 1968;116:506-512. 52. Dinehart SM, Pollack SV. Metastases from squamous cell carcinoma of the skin and lip. An analysis of twenty-seven cases. J Am Acad Dermatol 1989;21:241-248. 53. Modlin JJ. Cancer of the skin: surgical treatment. Mol Med 1954;51:364-367. 54. Møller R, Reymann F, Hou-Jensen K. Metastases in dermatological patients with squamous cell carcinoma. Arch Dermatol 1979;115:703-705. 55. Moore BA, Weber RS, Prieto V, et al. Lymph node metastases from cutaneous squamous cell carcinoma of the head and neck. Laryngoscope 2005;115:1561-1567. 56. Veness MJ, Palme CE, Morgan GJ. High-risk cutaneous squamous cell carcinoma of the head and neck: results from 266 treated patients with metastatic lymph node disease. Cancer 2006;106:2389-2396. 57. Bourne RG. The Costello Memorial Lecture. The spread of squamous cell carcinoma of the skin via the cranial nerves. Australas Radiol 1980;24:106-114. 58. Loeffler JS, Larson DA, Clark JR, et al. Treatment of perineural metastasis from squamous carcinoma of the skin with aggressive combination chemotherapy and irradiation. J Surg Oncol 1985;29:181-183. 59. Mohs FE. Chemosurgery: Microscopically Controlled ­ Surgery for Skin Cancer. Springfield, IL: Charles C Thomas, 1978,  pp 261-267. 60. Gruber SB, Barnhill RL, Stenn KS, et al. Nevomelanocytic proliferations in association with cutaneous melanoma: a multivariate analysis. J Am Acad Dermatol 1989;21:773-780. 61. Clark WH, Ainsworth AM, Bernardino EA. The developmental biology of primary human malignant melanomas. Semin Oncol 1975;2:83-103. 62. McGovern VJ, Murad TM. Pathology of melanoma: an overview. In Balch CM, Milton GW (eds): Cutaneous Melanoma: Clinical Management and Treatment Results Worldwide. Philadelphia: JB Lippincott, 1985, pp 29-53. 63. Barnhill RL, Mihm MC. Histopathology of malignant melanoma and its precursor lesions. In Balch CM, Houghton AN, Milton GW, et al (eds): Cutaneous Melanoma, 2nd ed. Philadelphia: JB  Lippincott, 1992, pp 234-263. 64. Urist MM, Balch CM, Soong SJ. Head and neck melanoma in 534 clinical stage I patients: a prognostic factor analysis and results of surgical treatment. Ann Surg 1984;200:769-775. 65. Clark WH, Mihm MC. Lentigo maligna and lentigo-maligna melanoma. Am J Pathol 1969;55:39-67. 66. Reed RJ. The pathology of human cutaneous melanoma. In Costanzi JJ (ed): Malignant Melanoma I. The Hague: Martinus Nijhoff, 1983, pp 85-116. 67. Arrington JH, Reed RJ, Ichinose H, et al. Plantar lentiginous melanoma: a distinctive variant of human cutaneous malignant melanoma. Am J Surg Pathol 1977;1:131-143. 68. Coleman WP, Loria PR, Reed RJ, et al. Acral lentiginous melanoma. Arch Dermatol 1980;116:773-776.

155

69. Feibleman CE, Stoll H, Maize JC. Melanomas of the palm, sole, and nailbed: a clinicopathologic study. Cancer 1980;46:2492-2504. 70. Krementz ET, Feed RJ, Coleman WP, et al. Acral lentiginous melanoma. A clinicopathologic entity. Ann Surg 1982;195:  632-645. 71. Patterson RH, Helwig EB. Subungual malignant melanoma: a clinical-pathologic study. Cancer 1980;46:2074-2087. 72. Balch CM, Urist MM, Maddox WA, et al. Melanoma in the southern United States: experience at the University of Alabama in Birmingham. In Balch CM, Milton GW (eds): Cutaneous Melanoma: Clinical Management and Treatment Results Worldwide, 2nd ed. Philadelphia: JB Lippincott, 1985, pp 397-418. 73. Reintgen DS, McCarty KM, Cox E, et al. Malignant melanoma in black American and white American populations: a comparative review. JAMA 1982;248:1856-1859. 74. Seiji M, Takahashi M. Acral melanoma in Japan. Hum Pathol 1982;13:607-609. 75. Balch CM, Soong SJ, Gershenwald JE, et al. Prognostic factors analysis of 17,600 melanoma patients: validation of the American Joint Committee on Cancer melanoma staging system.  J Clin Oncol 2001;19:3622-3634. 76. Ballo MT, Ross MI, Cormier JN, et al. Combined-modality therapy for patients with regional nodal metastases from melanoma. Int J Radiat Oncol Biol Phys 2006;64:106-113. 77. Cotlar AM, Gates JO, Gibbs FA. Merkel cell carcinoma: combined surgery and radiation therapy. Am Surg 1986;52:  159-164. 78. Sibley RK, Dehner LP, Rosai J. Primary neuroendocrine (Merkel cell?) carcinoma of the skin. I.  A clinicopathologic and ultrastructural study of 43 cases. Am J Surg Pathol 1985;9:  95-108. 79. Morrison WH, Peters LJ, Silva EG, et al. The essential role of radiation therapy in securing locoregional control of Merkel cell carcinoma. Int J Radiat Oncol Biol Phys 1990;19:583-591. 80. Merkel F. Tastzellen und tastkoerperchen bei den haushieren und beim menschen. Arch Mikr Anat 1875;11:636-652. 81. Silva EG, Mackay B, Goepfert H, et al. Endocrine carcinoma of the skin (Merkel cell carcinoma). Pathol Annu 1984;19:1-30. 82. Bonfiglio TA. Neuroendocrine carcinoma of the skin: diagnostic and management considerations. Int J Radiat Oncol Biol Phys 1988;14:1321-1322. 83. Drijkoningen M, De Wolf-Peeters C, van Limbergen E, et al. Merkel cell tumor of the skin: an immunohistochemical study. Hum Pathol 1986;17:301-307. 84. Pacella J, Ashby M, Ainslie J, et al. The role of radiotherapy in the management of primary cutaneous neuroendocrine tumors (Merkel cell or trabecular carcinoma): experience at the Peter MacCallum Cancer Institute (Melbourne, Australia). Int J Radiat Oncol Biol Phys 1988;14:1077-1084. 85. Bourne RG, O’Rourke MG. Management of Merkel cell tumour. Aust NZ J Surg 1988;58:971-974. 86. Hitchcock CL, Bland KI, Laney RG, et al. Neuroendocrine (Merkel cell) carcinoma of the skin. Its natural history, diagnosis and treatment. Ann Surg 1988;207:201-207. 87. Meland NB, Jackson IT. Merkel cell tumor: diagnosis, prognosis, and management. Plast Reconstr Surg 1986;77:632-638. 88. Rice RD Jr, Chonkich GD, Thompson KS, et al. Merkel cell tumor of the head and neck. Five new cases with literature review. Arch Otolaryngol Head Neck Surg 1993;119:782-786. 89. Meeuwissen JA, Bourne RG, Kearsley JH. The importance of postoperative radiation therapy in the treatment of Merkel cell carcinoma. Int J Radiat Oncol Biol Phys 1995;31:325-331. 90. O’Brien PC, Denham JW, Leong AS. Merkel cell carcinoma: a review of behaviour patterns and management strategies.  Austral NZ J Surg 1987;57:847-850. 91. Wilder RB, Kittelson JM, Shimm DS. Basal cell carcinoma ­treated with radiation therapy. Cancer 1991;68:2134-2137. 92. Batsakis JG, Littler ER, Leahy MS. Sebaceous cell lesions of the head and neck. Arch Otolaryngol 1972;95:151-157. 93. Civatte J, Tsoitis G. Adnexal skin carcinomas. In Andrade R,  Gumport SL, Popkin GL, ��������������������������������������������� et al (�������������������������������������� eds): Cancer of the Skin,  vol 2. Philadelphia: WB Saunders, 1976, pp 1045-1068.

156

PART 2  Skin

94. Oppenheim AR. Sebaceous carcinoma of the penis. Arch Dermatol 1981;117:306-307. 95. Rulon DB, Helwig EB. Cutaneous sebaceous neoplasms. Cancer 1974;33:82-102. 96. Rao NA, Hidayat AA, McLean IW, et al. Sebaceous carcinomas of the ocular adnexa: a clinicopathologic study of 104 cases, with five-year follow-up data. Hum Pathol 1982;13:113-122. 97. Boniuk M, Zimmerman LE. Sebaceous carcinoma of the eyelid, eyebrow, caruncle and orbit. Trans Am Acad Ophthalmol Otolaryngol 1968;72:619-642. 98. Wick MR, Goellner JR, Wolfe JT, et al. Adnexal carcinomas of the skin. II. Extraocular sebaceous carcinomas. Cancer 1985;56:1163-1172. 99. Wick MR, Goellner JR, Wolfe JT III, et al. Adnexal carcinomas of the skin. I. Eccrine carcinomas. Cancer 1985;56:1147-1162. 100. Snow SN, Reizner GT, Dudley CM, et al. Eccrine mucinous carcinoma. In Miller SJ, Maloney ME (eds): Cutaneous ­Oncology: Pathophysiology, Diagnosis, and Management. Malden, MA: Blackwell Science, 1998, pp 758-762. 101. Berg JW, McDivitt RW. Pathology of sweat gland carcinomas. Pathol Annu 1968;3:123-144. 102. Lipper S, Peiper SC. Sweat gland carcinoma with syringomatous features: a light microscopic and ultrastructural study. Cancer 1979;44:157-163. 103. Liu Y. The histogenesis of clear cell papillary carcinoma of the skin. Am J Pathol 1949;25:93-99. 104. Mendoza S, Helwig EB. Mucinous (adenocystic) carcinoma of the skin. Arch Dermatol 1971;103:68-78. 105. Mishima Y, Morioka S. Oncogenic differentiation of the intraepidermal eccrine sweat duct: eccrine poroma, poroepithelioma and porocarcinoma. Dermatologica 1969;138:238-250. 106. Nickoloff BJ, Fleischmann HE, Carmel J, et al. Microcystic adnexal carcinoma: immunohistologic observations suggesting dual (pilar and eccrine) differentiation. Arch Dermatol 1986;122:290-294. 107. Ota Y, Fukasawa I, Shimizu K, et al. A case of T1N0M0 vulvar apocrine gland carcinoma with a positive outcome. Gynecol Oncol 2003;90:601-604. 108. Shintaku M, Tsuta K, Yoshida H, et al. Apocrine ­adenocarcinoma of the eyelid with aggressive biological behavior: report of a case. Pathol Int 2002;52:169-173. 109. Mansour P, George MK, Pahor AL, et al. Ceruminous gland ­tumours: a reappraisal. J Laryngol Otol 1992;106:727-732. 110. Roy S, Shafi NQ, Rose MG. Locally recurrent and metastatic apocrine-gland carcinoma in an elderly man. Nat Clin Pract ­Oncol 2007;4:56-59. 111. Greene FL, Page DL, Fleming ID, et al (eds): AJCC Cancer Staging Manual, 6th ed. New York: Springer, 2002. 112. Breslow A. Thickness, cross-sectional areas and depth of invasion in the prognosis of cutaneous melanoma. Ann Surg 1970;172:902-908. 113. Fleming I, Cooper JS, Henson DE, et al (eds). AJCC Cancer Staging Manual, 5th ed. Philadelphia: Lippincott Williams & Wilkins, 1997, pp 163-170. 114. Neville JA, Welch E, Leffell DJ. Management of nonmelanoma skin cancer in 2007. Nat Clin Pract Oncol 2007;4:462-469. 115. Rowe DE, Carroll RJ, Day CL Jr. Mohs surgery is the treatment of choice for recurrent (previously treated) basal cell carcinoma. J Dermatol Surg Oncol 1989;15:424-431. 116. Morrison WH, Garden AS, Ang KK. Radiation therapy for nonmelanoma skin carcinomas. Clin Plast Surg 1997;24:719-729. 117. Wilder RB, Margolis LW. Cancer of the skin. In Leibel SA, Phillips TL (eds): Textbook of Radiation Oncology, 2nd ed. ­Philadelphia: Lippincott Williams & Wilkins, 2004, pp ­1483-1501. 118. Rowe DE, Carroll RJ, Day CL Jr. Long-term recurrence rates in previously untreated (primary) basal cell carcinoma: implications for patient follow-up. J Dermatol Surg Oncol 1989;15:315-328. 119. Fitzpatrick PJ, Thompson GA, Easterbrook WM, et al. Basal and squamous cell carcinoma of the eyelids and their treatment by radiotherapy. Int J Radiat Oncol Biol Phys 1984;10:449-454. 120. Petrovich Z, Parker RG, Luxton G, et al. Carcinoma of the lip and selected sites of head and neck skin. A clinical study of 896 patients. Radiother Oncol 1987;8:11-17.

121. Lovett RD, Perez CA, Shapiro SJ, et al. External irradiation of epithelial skin cancer. Int J Radiat Oncol Biol Phys 1990;19:235-242. 122. Shimm DS, Wilder RB. Radiation therapy for squamous cell carcinoma of the skin. Am J Clin Oncol 191;14:383-386. 123. Wilder RB, Shimm DS, Kittelson JM, et al. Recurrent basal cell carcinoma treated with radiation therapy. Arch Dermatol 1991;127:1668-1672. 124. Janse GT, Westbrook KC. Cancer of the skin. In Suen JY, Myers EN (eds): Cancer of the Head and Neck. New York: Churchill Livingstone, 1981, pp 212-241. 125. Preston DS, Stern RS. Nonmelanoma cancers of the skin. N Engl J Med 1992;327:1649-1662. 126. Fredricks S. External ear malignancy. Br J Plast Surg 1956;9:  136-160. 127. Metcalf PB Jr. Carcinoma of the pinna. N Engl J Med 1954;  251:91-95. 128. Schewe EJ Jr, Pappalardo C. Cancer of the external ear. Am J Surg 1962;104:753-755. 129. Avila J, Bosch A, Aristizabal S, et al. Carcinoma of the pinna. Cancer 1977;40:2891-2895. 130. Petrovich Z, Kuisk H, Langholz B, et al. Treatment of carcinoma of the skin with bone and/or cartilage involvement. Am J Clin Oncol 1988;11:110-113. 131. Ang KK. Altered fractionation trials in head and neck cancer. Semin Radiat Oncol 1998;8:1-15. 132. Taylor BW Jr, Brant TA, Mendenhall NP, et al. Carcinoma of the skin metastatic to parotid area lymph nodes. Head Neck 1991;13:427-433. 133. Audet N, Palme CE, Gullane PJ, et al. Cutaneous metastatic squamous cell carcinoma to the parotid gland: analysis and outcome. Head Neck 2004;26:727-732. 134. Karakousis CP, Balch CM, Urist MM, et al. Local recurrence in malignant melanoma: long-term results of the multi-­institutional randomized surgical trial. Ann Surg Oncol 1996;3:446-452. 135. Harwood AR. Conventional fractionated radiotherapy for  51 patients with lentigo maligna and lentigo maligna melanoma. Int J Radiat Oncol Biol Phys 1983;9:1019-1021. 136. Milton GW, Shaw HM, McCarthy WH, et al. Prophylactic lymph node dissection in clinical stage I cutaneous malignant melanoma: results of surgical treatment in 1319 patients. Br J Surg 1982;69:108-111. 137. Veronesi U, Adamus J, Bandiera DC, et al. Stage I melanoma of the limbs: immediate versus delayed node dissection. Tumori 1980;66:373-396. 138. Balch CM, Soong SJ, Bartolucci AA, et al. Efficacy of an elective regional lymph node dissection of 1 to 4 mm thick melanomas for patients 60 years of age and younger. Ann Surg 1996;224:  255-266. 139. Cascinelli N, Morabito A, Santinami M, et al. Immediate or delayed dissection of regional nodes in patients with melanoma of the trunk: a randomised trial. Lancet 1998;351:793-796. 140. Morton DL, Wen DR, Wong JH, et al. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 1992;127:392-399. 141. Lang PG Jr. Malignant melanoma. Med Clin North Am 1998;­  ­82:1325-1358. 142. Morton DL, Thompson JF, Cochran AJ, et al. Sentinel-node biopsy or nodal observation in melanoma. N Engl J Med 2006;355:1307-1317. 143. Singletary SE, Balch CM, Urist MM, et al. Surgical treatment of primary melanoma. In Balch CM, Houghton AN, Milton GW, et al (eds): Cutaneous Melanoma, 2nd ed. Philadelphia: JB Lippincott, 1992, pp 269-274. 144. Byers RM. The role of modified neck dissection in the treatment of cutaneous melanoma of the head and neck. Arch Surg 1986;12:1338-1341. 145. Cooper JS, Chang WS, Oratz R, et al. Elective radiation therapy for high-risk malignant melanomas. Cancer J 2001;7:498-502. 146. Balch CM, Soong SJ, Murad TM, et al. A multifactorial analysis of melanoma. IV. Prognostic factors in 200 melanoma patients with distant metastases (stage III). J Clin Oncol 1983;1:126-134. 147. Overgaard J. Radiation therapy of malignant melanoma. Int J Radiat Oncol Biol Phys 1980;6:41-44.

6  The Skin 148. Overgaard J, Overgaard M, Vejby Hansen P, et al. Some factors of importance in the radiation treatment of malignant melanoma.  Radiother Oncol 1986;5:183-192. 149. Strauss A, Dritschilo A, Nathanson L, et al. Radiation therapy of malignant melanoma: an evaluation of clinically used fractionation schemes. Cancer 1981;47:1262-1266. 150. Sause WT, Cooper JS, Rush S, et al. Fraction size in external beam radiation therapy in the treatment of melanoma. Int J Radiat Oncol Biol Phys 1991;20:429-432. 151. Bonnen MD, Ballo MT, Myers JN, et al. Elective radiotherapy provides regional control for patients with cutaneous melanoma of the head and neck. Cancer 2004;100:383-389. 152. Kirkwood JM, Strawderman MH, Ernstoff MS, et al. Interferon alfa-2b adjuvant therapy of high-risk resected cutaneous melanoma: the Eastern Cooperative Oncology Group trial EST 1684. J Clin Oncol 1996;14:7-17. 153. Grob JJ, Dreno B, de la Salmoniere P, et al. Randomised trial of interferon alpha-2a as adjuvant therapy in resected primary melanoma thicker than 1-5 mm without clinically detectable node metastases. Lancet 1998;351:1905-1910. 154. Mays SR, Nelson BR. Current therapy of cutaneous melanoma. Cutis 1999;63:293-298. 155. Tsao H, Atkins MB, Sober AJ. Management of cutaneous melanoma. N Engl J Med 2004;351:998-1012. 156. Rusthoven JJ. The evidence for tamoxifen and chemotherapy as treatment for metastatic melanoma. Eur J Cancer 1998;34:  S31-S36. 157. Wolchok JD, Weber JS, Houghton AN, et al. Melanoma ­vaccines. In Balch CM, Houghton AN, Sober AJ, eds: Cutaneous ­ Melanoma. 4th ed, St Louis: Quality Medical, 2003,  pp 645-656. 158. Wilder RB, Harari PM, Graham AR, et al. Merkel cell ­carcinoma. Improved locoregional control with postoperative radiation therapy. Cancer 1991;68:1004-1048. 159. Pilotti S, Rilke F, Bartoli C, et al. Clinicopathologic correlations of cutaneous neuroendocrine Merkel cell carcinoma. J Clin  Oncol 1988;6:1863-1873. 160. Kokoska GR, Kokoska MS, Collins BT, et al. Early aggressive treatment for Merkel cell carcinoma improves outcome. Am J Clin Oncol 1997;174:688-693. 161. Ott MJ, Tanabe KK, Gadd MA, et al. Multimodality management of Merkel cell carcinoma. Arch Surg 1999;134:388-392. 162. Feun LG, Savaraj N, Legha SS, et al. Chemotherapy for metastatic Merkel cell carcinoma. Review of the M.D. Anderson Hospital’s experience. Cancer 1988;62:683-685. 163. Wynne CJ, Kearsley JH. Merkel cell tumor: a chemosensitive skin cancer. Cancer 1988;62:28-31.

157

164. Grosh WW, Giannone L, Hande KR, et al. Disseminated Merkel cell tumor. Treatment with systemic chemotherapy. Am J Clin Oncol 1987;10:227-230. 165. Poulsen M, Rischin D, Walpole E, et al. High-risk Merkel cell carcinoma of the skin treated with synchronous carboplatin/ etoposide and radiation: Trans-Tasman Radiation Oncology Group Study TROG 96: 07. J Clin Oncol 2003;21:4371-4376. 166. Poulsen MG, Rischin D, Porter I, et al. Does chemotherapy improve survival in high-risk stage I and II Merkel cell carcinoma of the skin? Int J Radiat Oncol Biol Phys 2006;64:114-119. 167. Fenig E, Lurie H, Klein B, et al. The treatment of advanced Merkel cell carcinoma. A multimodality chemotherapy and radiation therapy treatment approach. J Dermatol Surg Oncol 1993;19:860-864. 168. Bichakjian CK, Lowe L, Lao CD, et al. Merkel cell carcinoma: critical review with guidelines for multidisciplinary management. Cancer 2007;110:1-12. 169. Lee WR, Mendelhall WM, Parsons JT, et al. Radical radiotherapy for T4 carcinoma of the skin of the head and neck: a multivariate analysis. Head Neck 1993;15:320-324. 170. Joseph J, Adler JR, Cox RS, et al. Linear accelerator-based stereotactic radiosurgery for brain metastases: the influence of number of lesions on survival. J Clin Oncol 1996;14:1085-1092. 171. Ang KK, Garden AS. Radiotherapy for Head and Neck Cancers: Indications and Techniques, 3rd ed. Philadelphia: Lippincott Williams & Wilkins, 2006. 172. Ang K. Altered fractionation in head and neck cancer. Semin Radiat Oncol 1998;8:230-236. 173. Carter RL, Foster CS, Dinsdale EA, et al. Perineural spread by squamous carcinomas of the head and neck: a morphological study using antiaxonal and antimyelin monoclonal antibodies.  J Clin Pathol 1983;36:269-275. 174. Barrett TL, Greenway HT Jr, Massullo V, et al. Treatment of ­basal cell carcinoma and squamous cell carcinoma with perineural ­invasion. In Callen JP, Dahl MV, Golitz LE, et al (eds): Advances in Dermatology. St Louis: CV Mosby, 1993, pp 277-306. 175. Cottel WI. Perineural invasion by squamous-cell carcinoma.  J Dermatol Surg Oncol 1982;8:589-600. 176. Traenkle HL, Mulay D. Further observations on late radiation necrosis following therapy of skin cancer.Arch Dermatol 1960;81:  908-913. 177. Williams PC, Hendry JH. The RBE of megavoltage photon and electron beams. Br J Radiol 1978;51:220. 178. Boland J. Human skin reaction. Br J Radiol 1957;30:351-352. 179. Wang CC. Cancer of the skin. In Wang CC (ed): Radiation Therapy for Head and Neck Neoplasms: Indications, Techniques, and Results. Chicago: Year Book Medical, 1990, p 85.

Advances in the Treatment of Head and Neck Cancer

7

K. Kian Ang ALTERED FRACTIONATION REGIMENS HIGH-PRECISION RADIATION THERAPY COMBINATION OF RADIATION THERAPY WITH CHEMOTHERAPY Primary Therapy for Locally Advanced Head and Neck Squamous Cell Cancer Adjuvant Treatment of Locally Advanced Head and Neck Squamous Cell Cancer

MOLECULAR BIOMARKERS AND THERAPEUTICS SUMMARY

A variety of neoplasms with widely different natural histories can arise in the head and neck area, a relatively small body region essential for basic physiologic functions and social interactions. Depending on the location, size, and pattern of spread, head and neck tumors can cause various degrees of structural deformities and functional handicaps. Treatment of head and neck tumors can also induce additional disfigurements and malfunctions, worsening quality of life. Consequently, head and neck cancers are challenging to manage. Refinement in surgical resection and reconstructive techniques and advances in radiation therapy planning and delivery technology yield good outcomes for most patients with early head and neck cancer. Unfortunately, for more locally advanced tumors, the results of radiation therapy alone or combinations of surgical resection and preoperative or postoperative radiation therapy are rather poor in terms of disease control or preservation of organ function, or both. Consequently, the search continues for better treatment approaches. This quest along with the simplicity of clinical evaluation and well-characterized patterns of relapse make head and neck carcinomas ideal models for testing the relative efficacy of novel therapy concepts and modalities. Scores of clinical trials have been conducted to test the relative efficacy of altered radiation fractionation regimens, combinations of irradiation with chemotherapy, and the combination of irradiation with molecular therapeutics (see Chapter 5). Most of the randomized trials conducted focus on patients with squamous cell carcinoma (SCC) of the oral cavity, oropharynx, larynx, and hypopharynx. Several treatment regimens have been shown to improve the success rate in achieving locoregional tumor control, survival, or organ preservation. Consequently, the standard of care for intermediate-stage to locally advanced head and neck squamous cell carcinoma (HNSCC) has gradually changed over the past decade. Because most randomized clinical trials have had rather broad eligibility criteria, their results may be applicable to a variety of primary tumor sites. To facilitate cross-referencing and minimize repetition in subsequent chapters in this section, some key findings are summarized in this introductory chapter.

Altered Fractionation Regimens Radiobiologic concepts derived from more than 2 decades of integrated laboratory and clinical investigations led to the conception of two classes of new radiation therapy ­regimens for the treatment of cancer. These altered fractionation regimens are referred to as hyperfractionated and accelerated fractionation schedules. The biologic rationale for modifying fractionation schedules is presented in ­Chapter 1. Briefly, in hyperfractionated schedules, the total dose is increased, the dose per fraction is significantly reduced, the number of dose fractions is increased, and overall time is relatively unchanged. Hyperfractionation exploits the differences in fractionation sensitivity between tumors and normal tissues manifesting late morbidity. Decreasing the fraction size reduces damage to late-responding normal tissues relative to the damage to tumors. In accelerated fractionation schedules, the overall ­treatment duration is reduced, and the number of dose fractions, total dose, and dose per fraction are unchanged or reduced, depending on the extent of overall time reduction. Although radiation therapy can be accelerated in a variety of permutations, for simplicity, the existing schedules can be conceptually grouped into two categories—accelerated fractionation with or without total dose reduction. The goal of accelerated fractionation is to reduce the impact of tumor proliferation as a cause of radiation therapy failure. Altered fractionation regimens have been extensively ­tested in patients with intermediate to locally advanced ­HNSCC. Bourhis and colleagues1 conducted a thorough meta-analysis that included updates of individual patient data. A total of 6515 patients enrolled in 15 phase III ­trials were included in the analysis. The main primary tumor sites were the oropharynx (3079 patients [44%]) and larynx (2377 patients [34%]), and most patients (5221 [74%]) had stage III or IV tumors. The length of follow-up ranged from 4 to 10 years, with a median of 6 years. The treatment regimens tested were considered in three categories: hyperfractionated irradiation and accelerated fractionation with or without dose reduction. Table 7-1 summarizes the benefit of altered fractionation compared with conventional fractionation schedules on different outcome end points.1 On the whole, altered fractionation yielded a significant survival

161

162

PART 3  Head and Neck

TABLE 7-1.   Absolute Benefit and Hazard Ratios for Altered Fractionation versus Conventional Fractionation for Intermediate-Stage to Locally Advanced Head and Neck Squamous Cell Cancer

End Point

Hyperfractionation

Accelerated ­ rac­tionation without F Dose Reduction

Accelerated ­Fractionation with Dose Reduction

Overall

P Value

Overall survival

+3.4%*

0.003

+8.2%

+2.0%

+1.7%

Locoregional control

+6.4%