Radiotherapy in Practice:

External Beam Therapy

Radiotherapy in Practice

Radiotherapy in Practice: Brachytherapy
Edited by Peter Hoskin and Catherine Coyle

Radiotherapy in Practice: Imaging for Clinical Oncology

Edited by Peter Hoskin and Vicky Goh

Radiotherapy in Practice: Physics for Clinical Oncology

Edited by Amen Sibtain, Andrew Morgan, Niall MacDougall

Radiotherapy in Practice: Radioisotope Therapy

Edited by Peter Hoskin
Radiotherapy in
Practice: External
Beam Therapy
THIRD EDITION

Edited by

Peter Hoskin
Consultant Clinical Oncologist,
Mount Vernon Cancer Centre
Professor in Clinical Oncology,
University of Manchester and
Honorary Consultant in Clinical Oncology,
University College London Hospitals NHS Trust, London, UK
and the Christie Hospital NHS Trust, Manchester UK

1
1
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Oxford University Press 2019
The moral rights of the authors have been asserted
First Edition Published in 2006
Second Edition Published in 2012
Third Edition Published in 2019
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2018962716
ISBN 978–​0–​19–​878675–​7
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Oxford University Press makes no representation, express or implied, that the
drug dosages in this book are correct. Readers must therefore always check
the product information and clinical procedures with the most up-​to-​date
published product information and data sheets provided by the manufacturers
and the most recent codes of conduct and safety regulations. The authors and
the publishers do not accept responsibility or legal liability for any errors in the
text or for the misuse or misapplication of material in this work. Except where
otherwise stated, drug dosages and recommendations are for the non-​pregnant
adult who is not breast-​feeding
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
 Contents

List of contributors  vii

List of abbreviations  ix

1 Introduction  1
Peter Hoskin
2 Basic physics  6
Karen Venables
3 Treatment delivery, intensity-​modulated radiotherapy, and
image-​guided radiotherapy  27
Yat Man Tsang
4 Proton therapy  53
Ranald MacKay, Adam Aitkenhead
5 Breast radiotherapy  70
Charlotte Coles, Murray Brunt, Anna Kirby, Sara Lightowlers,
Nicola Twyman
6 Radiotherapy for thoracic tumours  115
Kevin Franks, Fiona McDonald, Gerard G Hanna
7 Upper gastrointestinal tract  145
Stephen Falk
8 Rectal cancer  165
Rob Glynne-​Jones, Mark Harrison
9 Squamous cell carcinoma of the anus  196
Rob Glynne-​Jones, Mark Harrison
10 Prostate cancer  224
Linus Benjamin, Alison Tree, David Dearnaley
11 Bladder cancer  263
Nicholas James, David Fackrell, Anjali Zarkar
12 Testis  279
Peter Hoskin
13 Penis  282
Peter Hoskin
14 Uterus: Endometrium and cervix  288
Melanie Powell, Alexandra Taylor
vi Contents

15 Vulva and vagina  309

Peter Hoskin
16 Lymphomas  317
Richard W Tsang, Mary K Gospodarowicz, Peter Hoskin
17 Central nervous system tumours  351
Neil G Burnet, Fiona Harris, Mark B Pinkham, Kate E Burton,
Gillian A Whitfield
18 Head and neck cancer  405
Christopher Nutting, Dorothy Gujral
19 Skin cancer  438
Carie Corner, Hannah Tharmalingam, Peter Hoskin
20 Sarcomas of soft tissue and bone  454
James Wylie
21 Principles of paediatric radiation oncology  468
Henry C Mandeville
22 Radiotherapy planning for metastatic disease  508
Peter Hoskin
23 Quality assurance in radiotherapy  520
Patricia Díez, Edwin GA Aird

Index  537
 List of contributors

Edwin GA Aird Patricia Díez

Mount Vernon Cancer Centre, Mount Vernon Cancer Centre,
Northwood, UK Northwood, UK
Adam Aitkenhead David Fackrell
The Christie NHS Foundation Trust & Queen Elizabeth Hospital,
Manchester Cancer Research Centre, Birmingham, UK
Manchester, UK
Stephen Falk
Linus Benjamin Bristol Oncology Centre, Bristol, UK
The Royal Marsden NHS Foundation
Kevin Franks
Trust, London, UK and Mount Vernon
St James’s Institute of Oncology,
Cancer Centre, Northwood, UK
Leeds, UK
Murray Brunt
Rob Glynne-​Jones
University Hospitals of North Midlands
Mount Vernon Cancer Centre,
NHS Trust, Stoke-​on-​Trent, UK & Keele
Northwood, UK
University, Keele, UK
Mary K. Gospodarowicz
Neil G Burnet
Princess Margaret Cancer Centre,
University of Manchester, Manchester
University of Toronto, ON, Canada
Cancer Research Centre and The
Christie NHS Foundation Trust, Dorothy Gujral
Manchester, UK Imperial College Healthcare NHS
Trust and Imperial College London,
Kate E Burton
London UK
Addenbrooke’s Hospital, Cambridge
University Hospitals NHS Foundation Gerard G. Hanna
Trust, Cambridge, UK UKPeter MacCallum Cancer Centre,
Melbourne, Australia
Charlotte Coles
Addenbrooke’s Hospital, Cambridge Fiona Harris
University Hospitals NHS Foundation Addenbrooke’s Hospital, Cambridge
Trust, Cambridge, UK University Hospitals NHS Foundation
Trust, Cambridge, UK
Carie Corner
Mount Vernon Cancer Centre, Mark Harrison
Northwood, UK Mount Vernon Cancer Centre,
Northwood, UK
David Dearnaley
The Royal Marsden NHS Foundation
Trust, UK
viii List of contributors

Peter Hoskin Melanie Powell
Mount Vernon Cancer Centre, Barts Health NHS Trust, Barts Institute
Northwood, and Manchester of Cancer, Barts and The London
Cancer Research Centre, Medical School, London, UK
University of Manchester,
Alexandra Taylor
Manchester, UK
The Royal Marsden NHS Foundation
Nicholas James Trust, London, UK
Institute of Cancer and Genomic
Hannah Tharmalingam
Sciences, University of Birmingham
Mount Vernon Cancer Centre,
and Queen Elizabeth Hospital,
Northwood, UK
Birmingham, UK
Alison Tree
Anna Kirby
The Royal Marsden NHS Foundation
Royal Marsden NHS Foundation
Trust, UK
Trust & Institute of Cancer Research,
Sutton, UK Yat Man Tsang
Mount Vernon Cancer Centre,
Sara Lightowlers
Northwood, UK
Addenbrooke’s Hospital,
Cambridge University Hospitals Richard W Tsang
NHS Foundation Trust, Cambridge, Princess Margaret Hospital, University
UK, and Cambridge University, of Toronto, ON, Canada
Cambridge, UK. Nicola Twyman
Ranald MacKay Addenbrooke's Hospital, Cambridge
The Christie NHS Foundation Trust & University Hospitals NHS Foundation
Manchester Cancer Research Centre, Trust, Cambridge, UK
Manchester, UK Karen Venables
Henry C Mandeville Mount Vernon Cancer Centre,
The Royal Marsden NHS Foundation Northwood, UK
Trust, London, UK Gillian A Whitfield
Fiona McDonald University of Manchester, Manchester
The Royal Marsden NHS Foundation Cancer Research Centre and The
Trust, London, UK Christie NHS Foundation Trust,
Manchester, UK
Christopher Nutting
Royal Marsden Hospital and Institute of James Wylie
Cancer Research, London, UK The Christie NHS Foundation Trust,
Manchester, UK
Mark B Pinkham
Princess Alexandra Hospital, Brisbane, Anjali Zarkar
and University of Queensland, Brisbane, Queen Elizabeth Hospital,
Australia. Birmingham, UK
 List of abbreviations
2D two-​dimensional
3D three-​dimensional
3D-​CRT three-​dimensional conformal
radiotherapy
5FU 5-​fluorouracil
ACTH adrenocorticotrophic hormone
AJCC American Joint Committee
on Cancer
ALARP as low as reasonably practicable
ALL acute lymphoblastic leukaemia
AML acute myeloid leukaemia
ANLL acute non-​lymphoblastic
leukaemia
ART adaptive radiotherapy
ASCT autologous stem-​cell
transplantation
BASO British Association of Surgical
Oncology
BCC basal cell carcinoma
BED biologically effective dose
BEV beam’s eye view
BMI body mass index
BMT bone marrow transplantation
CBCT cone-​beam computed
tomography
CCLG Children’s Cancer and
Leukaemia Group
CCR complete clinical response
CHART continuous hyperfractionated
accelerated radiotherapy
CHOP cyclophosphamide,
doxorubicin, vincristine, and
prednisone
CI confidence interval
cm centimetre
COPD chronic obstructive pulmonary
disease
CRM circumferential
resection margin
CRT chemoradiotherapy
CSA craniospinal axis
CSF cerebrospinal fluid
CSRT craniospinal radiotherapy
CT computed tomography
CTV clinical target volume
DCIS ductal carcinoma in situ
DFS disease-​free survival
DHAP dexamethasone, cytarabine,
cisplatinum
DLCO diffusing capacity of the lung
for carbon monoxide
DRL diagnostic reference level
DRR digitally reconstructed
radiograph
DVH dose–​volume histogram
EBCTCG Early Breast Cancer Trialists’
Collaborative Group
EBT external beam radiotherapy
EBUS endobronchial/​endoluminal
ultrasound
EFRT extended-​field radiotherapy
EIC extensive intraductal
component
EMVI extramural vascular invasion
EORTC European Organisation for
Research and Treatment
of Cancer
EPI electronic portal imaging
EPID electronic portal imaging device
ERUS endoscopic rectal ultrasound
ESR erythrocyte sedimentation rate
ETAR equivalent tissue–​air ratio
EUA examination under anaesthetic
EUD equivalent uniform dose
EUS endoluminal ultrasound
FDG fludeoxyglucose
FEV1 forced expiratory volume in
1 second
x List of abbreviations
FFF flattening filter free
FSD focus-​to-​skin distance
FSRT fractionated stereotactic
radiotherapy
FVC forced vital capacity
GBM glioblastoma
G-​CSF granulocyte-​colony
stimulating factor
GTV gross tumour volume
Gy gray
HDR-​ILBT high dose rate intraluminal
brachytherapy
HFRT hyperfractionated radiotherapy
HIV human immunodeficiency virus
HLA human leukocyte antigen
HPV human papilloma virus
HR hazard ratio
HVL half-​value  layer
IAM internal auditory meatus
IBD inflammatory bowel disease
ICRP International Commission on
Radiological Protection
ICRU International Commission
on Radiation Units and
Measurements
IFRT involved-​field radiation therapy
IGRT image-​guided radiotherapy
IM internal margin
IMAT intensity-​modulated arc
radiotherapy
IMC internal mammary chain
IMRT intensity-​modulated
radiotherapy
IORT intraoperative radiotherapy
IPI International Prognostic Index
IR irradiated volume
ITV internal target volume
IVU intravenous urogram
KCO corrected transfer factor
kV kilovoltage
LDH lactate dehydrogenase
LET linear energy transfer
LGG low-​grade  glioma
LSCLC limited small cell lung cancer
MALT mucosa associated
lymphoid tissue
MCC Merkel cell tumour
MDT multidisciplinary team
MIBG metaiodobenzyl guanidine
MIP maximum intensity projection
MLC multileaf collimator
mm millimetre
MMC mitomycin C
MR magnetic resonance
MRF mesorectal fascia
MRI magnetic resonance imaging
MU monitor unit
MV megavoltage
NB neuroblastoma
NF neurofibromatosis
NTCP normal tissue complication
probability
OAR organ at risk
OEV operator’s eye view
OS overall survival
OTT overall treatment time
Pb lead
PBS pencil beam scanning
PCI prophylactic cranial irradiation
PCNSL primary central nervous system
lymphoma
pCR pathological complete response
PET positron emission tomography
PNET primitive
neuroectodermal tumour
PORT postoperative radiotherapy
PPNET peripheral primitive
neuroectodermal tumour
PRV planning-​risk  volume
PRV planning organ at risk volume
PS performance status
PSA prostate-​specific antigen
PTV planning target volume
QA quality assurance
QC quality control

QUANTEC Quantitative Analyses of
Normal Tissue Effects in
the Clinic
RMS rhabdomyosarcoma
RTCT radiotherapy and chemotherapy
RTOG Radiation Therapy
Oncology Group
RVR remaining risk volume
SABR stereotactic ablative
radiotherapy
SBRT stereotactic body radiotherapy
SCC squamous cell carcinoma
SCF supraclavicular fossa
SCPRT short-​course preoperative
radiotherapy
SIB simultaneous integrated boost
SIOP International Society of
Paediatric Oncology
SLNB sentinel lymph node biopsy
SM set-​up  margin
SMART simultaneous modulated
accelerated radiotherapy
SRS stereotactic radiosurgery
SRT stereotactic radiotherapy
SSD source-​to-​surface distance
List of abbreviations xi
Sv sievert
TBI total body irradiation
TCP tumour control probability
TLD thermoluminescent dosimetry
TLI total lymphoid RT
TLI total nodal RT
TME total mesorectal excision
TMZ temozolomide
TNM tumour, node, metastasis
TP treatment planning
TPR tissue phantom ratio
TPS treatment planning system
TRH thyrotropin-​releasing hormone
TURP transurethral resection of the
prostate
TV treated volume
UICC International Union
Contra Cancer
UK United Kingdom
USA United States of America
VMAT volumetric-​modulated arc
radiotherapy
VNPI Van Nuys Prognostic Index
WHO World Health Organization
 Chapter 1

Introduction
Peter Hoskin

1.1 Introduction
Radiotherapy remains the most important non-​surgical treatment in the manage-
ment of cancer. Over 50% of patients will receive treatment at some time during
the management of their malignant disease. In recent years, rapid advances in the
technology available to radiotherapy have been made and there is a challenge to
the practising clinician to remain abreast of these and harness them to their best
use in the management of patients. For most patients receiving radiotherapy this
will mean treatment delivered with external X-​ray or electron beams. The processes
required for the safe delivery of modern radiotherapy comprise a lengthy pathway
from treatment decision to treatment delivery and verification. For the more com-
plex treatments this will involve sophisticated immobilization devices, high-​preci-
sion computed tomography (CT), magnetic resonance imaging (MRI), and positron
emission tomography (PET) image-​guided volume localization, complex and in-
creasingly accurate physics planning systems with state-​of-​the-​art algorithms to
account for tissue inhomogeneities and beam variables, and, finally, the widespread
use of high-​energy linear accelerators with multileaf collimators (MLCs), the cap-
acity for conformal and intensity-​modulated radiation therapy, and the ability to
provide on-​line image guidance of treatment delivery. Intensity-​modulated radio-
therapy (IMRT) is now the recognized standard of care for radical treatment, and
tomotherapy and stereotactic radiotherapy for more precise high-​dose delivery are
both widely available. Despite this, however, the basic principles of radiotherapy re-
main unchanged. Radiotherapy is a loco-​regional treatment suitable for radical treat-
ment of tumours in their early stages with high success rates where there has been no
metastatic spread. The basic steps of treatment delivery remain: defining the patient
position with a means of reproducing that position day to day with appropriate im-
mobilization, followed by accurate localization and definition of the volume to be
covered by the high-​dose envelope, and then collaboration with medical physicists
to identify the optimal means of doing this, using available beams with appropriate
modifications. The process of daily implementation of the treatment plan is often
neglected but is of vital importance in ensuring accurate and effective radiotherapy
together with verification that treatment delivery is reproducing the expected beam
as defined in the planning process.
2 Introduction

1.2  External beam sources

Linear accelerators are the common source of high-​energy X-​ray beams producing
megavoltage photons of between 4 and 20 million volt energy able to penetrate to the
most deep-​seated tumours in the largest of patients. Clinically, 4–​8 MV beams are the
most useful, providing a balance between penetration and adequate surface dose. The
fundamental property of megavoltage beams to have skin sparing is both beneficial in
terms of reducing skin reaction but also potentially hazardous in reducing the dose to
a surface or superficial tumour. Modern linear accelerators are highly sophisticated
machines working within a high precision of 2–​3%. Recent years have seen widespread
implementation of the MLC to provide complex beam shaping and in IMRT appli-
cations varying beam transmission. Additional beam modification using motorized
wedges has largely replaced the manual wedged-​shape filters which used to be placed
in the beam by the machine operators. The smooth and reliable running of such ma-
chines requires careful maintenance and quality assurance, the importance of which
cannot be emphasized too much; it is often carried out unseen and out of hours by the
dedicated band of bioengineers and physicists who attend to the processes required to
enable safe delivery of radiotherapy to patients.
Particle therapy includes both electron treatment which is widely used for superficial
tumours being produced from the standard linear accelerator and proton therapy. The
main advantage of protons is that their energy deposition follows the Bragg peak with
a high-​intensity, highly localized deposition of energy at a fixed depth. This has advan-
tages in the treatment of certain sites, for example, retinal tumours and tumours of the
brainstem where highly localized energy deposition avoiding surrounding structures
is required. They have also been used in other sites, for example, prostatic carcinoma,
as a means of enabling dose escalation within normal tissue tolerance. There are an
increasing number of facilities available across the world and in the UK two NHS
proton facilities will come on line in the next two years. For this reason proton therapy
is now included in some detail in this book.

1.3 Radiotherapy planning
Planning is a critical step in the delivery of clinical radiotherapy. For any treatment to
be effective it must be delivered accurately to the region of interest. The identification
of the GTV (gross tumour volume), CTV (clinical target volume), and PTV (plan-
ning target volume) representing sequential volume expansions from the macroscopic
identifiable tumour to including areas where there is risk of tumour spread even if not
identified, to a larger volume which takes into account patient movement and other
variations in day-​to-​day set-​up of a radiation beam during a fractionated course of
treatment is now embedded in the practice of modern radiotherapy.
Alongside this, major developments in imaging technology have allowed us to iden-
tify tumours with far more accuracy and certainty than before. It is now routine prac-
tice to identify internal tumours with CT planning. Increasingly, where appropriate,
MRI and PET images are also imported into the planning system and image regis-
tration used to provide greater certainty and clarification of the anatomy. Functional
 Radiation dose prescriptions 3

imaging techniques with MRI can enhance this further to provide ever more sophis-
ticated information on the tumour and its surrounding areas, alongside the equally
critical identification of the organs at risk where dose should be minimized.
Effective treatment planning ultimately depends upon complex computer algo-
rithms to simulate the effect of a beam passing through the designated area and the
amount of radiation energy deposited at any one site. The mathematical accuracy of
such algorithms has increased considerably in recent years and combined with the
use of CT imaging to provide accurate inhomogeneity data across different tissues, far
greater accuracy in dose distribution is now achievable. This is of particular import-
ance where large areas of lung or other air-​filled cavities such as the paranasal sinuses
are present in the treatment volume.

1.4  Treatment implementation and verification

The delivery of high-​dose fractionated radiotherapy requires implementation on the
treatment machine of a complex multistep process which may be repeated on a daily
basis for 30 or 40 fractions. This presents a significant challenge in devising quality
assurance processes to ensure that safe and reproducible radiotherapy is the standard.
The increasing use of computerized data sets, with electronic transfer of treatment
parameters from the planning software to the linear accelerator, minimizes the risk of
human error in transferring data. The use of mechanized wedges and MLCs takes out
the risk of the machine operator placing the wrong wedge in position or when using
lead shielding on a lead tray having this wrongly positioned. The importance of im-
mobilization to reproduce patient set-​up is now widely recognized and facilitated by
the use of immobilization devices ranging from simple head shells and vacuum bags
to stereotactic frames.
The megavoltage beams used for radiotherapy produce images have poor definition
between bone and soft tissue and are therefore more difficult to interpret than kilo-​
voltage (kV) energy beams. The advent of modern linear accelerators with silicone
diode array detectors, electronic portal imaging devices, and on-​board kV imaging
equipment has greatly enhanced the ability to accurately verify beam position. Many
patients are now treated with implanted fiducial markers in the treatment region to
improve identification with radiotherapy imaging. More frequent validation with
electronic portal imaging devices can be achieved and presentation of these on the
computer screen alongside the planning images with software tools to facilitate com-
parison greatly improves the accuracy of treatment and enables systematic errors to be
readily identified and corrected.
A further development is the incorporation of an MRI imaging facility within the
linear accelerator, the MR linac, which will enable on line verification with MRI.

1.5  Radiation dose prescriptions

Radiotherapy prescriptions have long carried an air of mystery and confusion. This
is hardly surprising when three or four different total doses given in a variation of
fractions of treatment over differing times are recommended for exactly the same
4 Introduction

tumour. Much of this relates to history and legend rather than systematic evalu-
ation. The problem is now compounded by new concepts in which acceleration
of fractions is recommended by some, hyperfractionation by others, use of a con-
comitant boost by yet others, whilst in most centres daily fractionation Monday to
Friday remains the standard. For radical treatment, the following schedules may be
encountered:
◆ Conventional fractionation usually refers to daily treatment on a Monday to
Friday basis.
◆ Accelerated fractionation means that the overall total dose is given in a shorter time
than would be achieved with conventional fractionation. This results in greater
toxicity and therefore only limited acceleration is possible without altering frac-
tion size. An example of this is the DAHANCA regimen in which six fractions are
given over 5 days so that a conventional 6-​week treatment schedule is delivered in 5
weeks. This modest acceleration has been shown to improve the results in head and
neck cancer.
◆ Hyperfractionation refers to the practice of reducing the fraction size of a conven-
tional regimen, often delivering treatment twice or even three times a day in the
smaller fraction sizes to enable a higher dose overall to be delivered. This is possible
because the toxicity, in particular the late toxicity, is reduced when the fraction size
is reduced for a given total dose. This approach has been investigated in many sites
including head and neck cancer and non-​small cell lung cancer.
◆ CHART (continuous hyperfractionated accelerated radiotherapy) is a schedule
which encompasses both acceleration and hyperfractionation delivering the total
dose in a shorter overall time (acceleration) and in smaller individual fractions
(hyperfractionation). The original CHART schedule delivered 54 Gy in 36 fractions
of 1.5 Gy over 12 days.
◆ Hypofractionation refers to giving a treatment in a shorter time than conventional
treatment using bigger doses per day and in order to do so safely reducing the total
dose. In the radical setting, examples are the delivery of 55 Gy in 20 daily fractions
or 50 Gy in 16 daily fractions which are considered equivalent to a radical con-
ventional dose of 65 Gy in 6½ weeks. There is increased interest in such schedules
following the observation that some tumours such as prostate cancer may have radi-
ation response characteristics with low alpha/​beta ratios. As a result, large doses per
fraction are biologically more effective. Concerns relating to normal tissue effects
even with highly accurate IMRT remain; hence formal evaluation in randomized
trials are now complete.
◆ Palliative radiotherapy is one area where hypofractionation is indicated. In symptom
control the aim is not to deliver a high dose to eradicate tumour but a sufficient dose
to enable symptom control. It has been widely shown that single doses of 8 Gy or
thereabouts are sufficient to improve bone pain and single doses of 10 Gy will im-
prove symptoms from non-​small cell lung cancer. Other common schedules in use
for palliation are 21 Gy in three fractions, 20 Gy in five fractions, and 30 Gy in 10
fractions.
 Radiation dose prescriptions 5

The third edition of this book continues its aim to provide a practical guide to
the use of external beam radiotherapy incorporating the substantial technological
advances that have been made in recent years. It will provide a firm background
in the physics of external beam radiotherapy and then deals with each anatomical
site in turn with details of the indications and techniques used for radiotherapy
delivery.
 Chapter 2

Basic physics
Karen Venables

2.1 Introduction
The distribution of radiation within the patient will be affected by many factors. These
include the energy and modality of the beam, the density of the tissue, the use of beam
modifiers such as wedges and compensators, and the distance of the patient from the
machine. The apparent distribution will also be affected by the accuracy of the algo-
rithm used on the planning system.

2.2 Interaction processes
The deposition of dose within the patient is dependent on the interaction process or
processes involved. Dose is the energy deposited in the material as a result of inter-
actions of photons and electrons with the material. When photons undergo an inter-
action, energy is transferred to electrons, which will then deposit their energy in the
medium. At low energies, photons interact predominantly by the photoelectric effect
in which a tightly bound electron is ejected from the atom. The dominant interaction
process in tissue for photons produced from linear accelerators (1–​20 MeV) is the
Compton process whereby the photon interacts with a loosely bound electron, re-
sulting in a free electron and a scattered photon of reduced energy. Pair production
also occurs above 1.02 MeV whereby a photon interacts within the nucleus of the
atom producing an electron and a positron; the positron will travel a short distance
and then annihilate, producing two further photons of energy 0.511 MeV. In con-
trast to photons in which the probability of an interaction occurring is governed by a
chance process, electrons deposit energy continuously along the length of their path by
collision with atomic electrons. They also lose energy through Bremsstrahlung: when
electrons pass close to a nucleus (which has a positive charge), it attracts the negatively
charged electron, changing its direction, and an X-​ray photon is emitted. The range
of an electron in a medium is dependent upon its initial energy and the density of
material through which it is traveling. Interactions produced from photon beams are
illustrated in Fig. 2.1. Monte Carlo is a powerful computing tool for determining the
result of irradiating a material with a beam of photons or electrons. It uses statistical
methods to determine the outcome of interactions and can be used to follow the his-
tory of individual particles in a beam. It is used in some treatment planning system
(TPS) algorithms to generate dose at a point. Diagrams illustrating the deposition of
dose are shown in Figs 2.2–​2.7. The deposition of dose within a medium can be de-
scribed by a dose deposition kernel.
 Dose deposition within the patient 7

(a)

Characteristic X-rays
(photons with energy
Incident Photon determined by the
atomic energy levels)

Ejected electron

(b)

Electron
Incident Photon

Scattered photon

(c)

Annihilation
Incident Photon positron photons

Electron

Fig. 2.1  Photon interaction processes. (a) Photoelectric effect: the incident photon

interacts with a bound electron which is ejected from the atom. Other electrons of
higher energy take its place and their excess energy is emitted as characteristic X-​rays.
(b) Compton effect: the incident photon interacts with a loosely bound electron
producing a scattered photon and scattered electron. (c) Pair production: the incident
photon interacts with the nucleus of the atom.

2.3  Dose deposition within the patient

The penetration of X-​rays or electrons will be dependent on the effective accelerating
potential to which the electrons in the waveguide have been subjected, although the
design of the treatment machine head will also affect this and photons or electrons
with a nominal beam energy from one machine may not have the same properties as
those with the same nominal energy from another machine. The photons incident on
the patient will have a spectrum of energies with the maximum possible energy being
that of the accelerating potential. The mean energy will be much lower, usually ¼ to
⅓ of the maximum.
8 Basic physics

3
2

1
Photon
beam

Water surface

Fig. 2.2  Monte Carlo calculation of twenty-​five 100 kV photons incident on a 20-​cm

water slab. The yellow lines (e.g. 1) show the photon and the blue circles (e.g. 2) the
electrons. At this energy, the electron range is below the resolution of the diagram and
their energy is absorbed at the points of interactions. All of the photons interact before
leaving the slab and a number have been scattered back towards the surface (e.g. 3).
Large angle scattering is common (e.g. 4).

Photon
1 beam

Water surface

Fig. 2.3  Monte Carlo calculation of twenty-​five 6-​MV photons incident on a 20-​cm

water slab. The yellow lines show the photons, the blue lines and circles the electrons,
and the red lines and crosses the positrons. Most of the electrons travel a short distance
before losing their energy and being reabsorbed but as shown by the number of blue
circles, they interact many times each time losing a little of their energy, in contrast
to the photons most of which travel a much larger distance between interactions.
Some of the photons do not interact before leaving the slab and the number scattered
back towards the surface is reduced compared to the 100 kV beam. Pair production is
possible at this energy as illustrated by the production of a positron (1) which travels a
short distance before annihilating. Compton interactions are governed by chance and
dependent upon energy; for the photons shown in this diagram no photon interactions
occurred in the first 2.5 cm, and this explains the build-​up in dose that occurs for high
energy accelerators. Before dose can be deposited, the photons must have interacted,
electrons produced in photon interactions gradually deposit dose along their path length.
 Dose deposition within the patient 9

1
Photon
beam

Water surface

Fig. 2.4  Monte Carlo calculation of twenty-​five 20 MV photons incident on a 20-​cm

water slab. The yellow lines show the photons, the blue lines and circles the electrons,
and the red lines and crosses the positrons. Some of the photons do not interact before
leaving the slab and the scattered photons and electrons are predominately in the same
direction as that of the incident photon. There is an increase in pair production and the
positrons travel further before being annihilated. Many of the photons outside of the
beam are the result of this process (e.g. the track traced in green is the result of a single
annihilation process).

2 Photon
3
beam
Fig. 2.5 Monte
Carlo simulation
1
of a positron
from 6-​MV
Water surface
beam, showing
the two 511 keV
photons almost
in opposing
directions (e.g.
1), the positron
(e.g. 2), and
electron (e.g. 3)
10 Basic physics

Fig. 2.6  Monte Carlo

simulation showing
electrons produced from
interactions of 6 MV
photons. Note the tortuous
paths and the increased
number of interactions
(each shown by a blue
cross) as the electron
reaches the end of its range
and has only a low energy.

The deposition of dose within a material is often described in terms of either per-
centage depth dose or tissue phantom ratios (TPRs). Percentage depth dose values
relate the dose at a given depth to that at the depth of maximum dose for the same
distance of the radiation source to the surface. They are dependent on the treatment
machine energy, distance from the source, and irradiated area, as well as the material
in which the dose is deposited. TPRs relate the dose at a reference depth in a phantom
to the dose at a point the same distance from the source but with a different depth of
material above the point. Tissue maximum ratios are a special case of TPR where the
reference depth is taken to be the depth at which maximum dose is deposited. TPRs
are dependent on field size, machine energy, and material in which the dose is depos-
ited but have only a very small dependence on distance from the source of radiation.
TPRs are often used for quick calculation of isocentric treatments, whereas percentage
depth doses are preferred in centres that treat patients at a fixed focus to surface dis-
tance. This is illustrated in Fig. 2.8.

Photon
beam

Fig. 2.7  Monte Carlo simulation of a narrow beam of a hundred 6 MV photons on a

water phantom, illustrating that most of the electrons deposit their energy close to
the position of the track of the incident photon beam. Generation of pencil beams for
planning systems using Monte Carlo calculated pencil beams is performed in this way
and the resultant dose distribution characterized. In the case of planning system pencil
beams, the incident photon beam is not a single energy but a spectrum of energies to
represent those found clinically in linear accelerator beams.
 Sources of high energy X-rays 11

100-Dref 100-D

D
Dref
Dmax
S T
B
D
A

Fig. 2.8  Diagram to illustrate the difference between percentage depth dose and TPR.
The percentage depth dose at point A would be found by dividing the dose at A by the
dose at B and multiplying by 100 to convert to a percentage. In contrast, the TPR for a
depth of D would be found by dividing the dose measured at T and dividing by the dose
measured at S. Note that in both the above cases, the same field size (jaw settings) has
been used throughout. The variation of the machine output with field size must also be
incorporated.

2.4  Sources of high energy X-​rays

Historically patients were treated with orthovoltage and superficial X-​ray units (up to
300 kV). These deliver high dose to the surface whilst still contributing dose at depth.
They are still used to treat some superficial lesions, particularly in the head and neck
region. Cobalt 60 machines were developed in the 1950s and deliver a higher dose at
depth due to the energy of the photons (1.17 MeV and 1.33 MeV). They are usually
reliable machines and still have a place in a few radiotherapy departments for simple
treatments. The photons are produced from the radioactive decay of the source. The
strength of the source (and therefore the intensity of the radiation) decreases with
time. The source must be changed approximately every 5  years to prevent treat-
ment times becoming too long. Depth dose curves for these machines are shown in
Fig. 2.9. Modern high-​energy linear accelerators offer a choice of photon and electron
energies. The production of high-​energy photons can be described briefly as follows.
Electrons are emitted from the heated gun filament, and their energy is gradually in-
creased as they move through the waveguide, transported by high-​power radio waves.
The beam of electrons is focused and steered through an angle of between 90° and
270° (depending on manufacturer’s design) (if necessary) to hit a high atomic number
target. The resultant X-​ray beam is collimated and the intensity of the radiation modu-
lated using a metal cone, known as the flattening filter, which is thickest in the centre
to produce a beam with a near uniform intensity within the treatment machine head.
The beam is collimated using two pairs of diaphragms or one pair of diaphragms and
12 Basic physics

120

100
Percentage depth dose
80

60

40

20

0
0 5 10 15 20 25
Depth (cm) water

100kV 230kV Co 60
Fig. 2.9  Depth–​dose curves for superficial and orthovoltage units. The 100 kV curve is
for a 30-​cm focus-​to-​skin distance (FSD) unit, with half-​value layer (HVL) of 3 mm and
a 10-​cm diameter field, the 230 kV curve is for a 50-​cm FSD unit with an HVL of 2-​mm
Cu and a closed end applicator, field size 10×10 cm. The Cobalt 60 data is for an FSD of
80 cm and a field size of 10×10 cm.
Source: Data from British Institute of Radiology (BJR), Supplement 25, Copyright © 1996 British
Institute of Radiology.

a set of multileaf collimator (MLC) leaves. The components of the conventional linear
accelerator are illustrated in Fig. 2.10.
Speciality linacs vary this basic design, both tomotherapy and cyberknife use
compact linacs with shorter waveguides which can still operate at about 6 MV. In
tomotherapy units, the linac is mounted on a computed tomography-​type gantry
system and the collimation is provided by a binary MLC (leaves are either open or
closed at any point in time). In cyberknife, the linac is fitted to a robotic arm allowing
many degrees of freedom in the direction in which the radiation can enter the patient.
Collimation is provided by a selection of fixed collimators from 0.5 cm to 6 cm radius
or a variable aperture collimator. Both of these linacs operate without a flattening filter
and at higher dose rates than a standard accelerator.
Electron beams used for treatment can be produced either by rapidly scanning
the narrow beam of electrons across the desired area or more commonly the beam is
broadened by the use of a scattering foil in place of the X-​ray target. In normal use, a
series of openings in an electron ‘applicator’ are used to collimate the beam down to or
close to the patient’s skin.
Typical depth dose curves for photons and electrons are shown in Figs 2.11 and
2.12. Dose is not deposited directly by the photons but rather by electrons set in
motion through interaction processes; therefore, for megavoltage photons, the max-
imum dose (dmax) does not occur at the surface but at a depth of 1–​4 cm. The number
of photons in the beam will begin to decrease immediately the beam enters the pa-
tient; however, even those photons that interact in the first millimetre of tissue will
 Sources of high energy X-rays 13

Bending
magnet

Electron Accelerating waveguide

gun Target

Primary
collimators

Flattening
filter
Ionization
Backscatter plate chamber
Mirror
Wedge
Light source

Y jaws

X jaws

MLC

End
plate
Fig. 2.10 Block diagram showing the components of a linear accelerator.

set in motion electrons which will travel a short distance before they have deposited
all of their energy. A  ‘build-​up effect’ occurs with increasing dose deposited with
depth until a condition is met whereby the energy transferred to electrons generated
from interactions is matched by the energy deposited by electrons already set in mo-
tion. The dose at the surface is typically between 10% and 30% of the dose at dmax,

120

100
Percentage depth dose

80

60

40

20 Fig. 2.11 Percentage
depth dose curves
0 for 6 and 15 MV
0 50 100 150 200 250 300 350 400
photon beams at
Depth (cm) water
100-​cm FSD,
6MV 15MV 10×10 cm.
14 Basic physics

120

100
Percentage Depth Dose

80

60

40

20

0
0 2 4 6 8 10 12
Depth (cm) water
Fig. 2.12 Electron
18MeV 15MeV 12MeV 9MeV 6MeV depth dose curves.

dependent on beam energy, field size, linac design, and the presence of scattering
materials such as wedges in the beam. The depth at which the maximum dose occurs
is dependent primarily on the beam energy. After dmax, a gradual decrease in the dose
deposited occurs as the number of photons in the beam is reduced. Two effects con-
tribute to this: the reduction in intensity due to the larger area that the photons cover
as the distance is increased and the decrease due to attenuation. For a very narrow
beam of monoenergetic photons, at a large distance from the source, the decrease
due to attenuation would be exponential. Deviations from exponential decrease
occur for two reasons:  the beams from linear accelerators are not monoenergetic
but comprise a spectrum of radiation and for the majority of cases in radiotherapy a
broad beam is used and therefore scatter from the medium will also affect the beam
intensity. The irradiated area will affect the number of scattered photons generated.
As the field size is increased from zero, there is initially a rapid increase in dose to
a point at the centre of the beam. This rate of increase slows as larger field sizes are
reached.
In contrast to photons, electron beams begin to deposit energy immediately on
entering the patient. There is a small build-​up as the electrons will travel a short dis-
tance before finally being absorbed into the medium. The range of electrons within
tissue will determine the distance into the material which they can penetrate and once
this distance has been reached there is a rapid decrease in the depth dose curve, be-
yond which the only significant dose deposited is that from contaminant photons
within the beam. In contrast to photons, correcting the intensity of a beam of electrons
using the inverse square law is complex, as the source of electrons will not be the ra-
diation target but an effective scatter source within the accelerator head. It is usually
advisable to measure outputs at non-​standard distances.
The radial profile of the beam is dependent primarily on the shape of the
flattening filter. When the machine is purchased, a depth for which the beam
 Radiation distributions within the patient 15

(a) (b)

Fig. 2.13  (a) Isodose distribution for a 10×10 field incident on a water phantom at 100-​
cm FSD. Note the change in shape of the isodose lines as the depth is increased. At
3-​cm deep, the 95% isodose is deeper at the outside of the beam, in contrast to this
at approximately 15-​cm deep, the 50% isodose is deepest at the centre of the beam.
(b) A 25° wedge field also incident on a flat water phantom.

intensity will be uniform is stated and the manufacturer will make any required
adjustments to the flattening filter. This depth is typically 5 or 10  cm. For large
field sizes, at shallower depths, the profile will have ‘horns’ or areas of increased
intensity whereas at greater depths the intensity at the edges of the beam are de-
creased. Two factors contribute to this: a non-​equilibrium of scatter from the edge
of the beam and a small change in mean energy as the distance from the centre
of the field is increased. The energy change is caused by absorption of low energy
photons at the flattening filter. In the centre where the filter is thickest, more low
energy photons will be absorbed in comparison with the edges of the field. This
absorption of low energy photons is often referred to as ‘beam hardening’. This is
adequately accounted for by most planning systems and can be seen on the isodose
distribution shown in Fig. 2.13.

2.5  Radiation distributions within the patient

Some patients may be treated using either a single field or a parallel-​opposed pair.
The appropriateness of each of these is related to the patient size and the location of
the tumour. Distributions for a range of patient sizes for 6 MV are shown in Fig. 2.14.
For separations of 12 cm, two opposing beams produce a uniform distribution of ra-
diation through the patient. As the patient separation is increased, areas of increased
dose relative to the dose at the centre of the volume are seen towards the surface.
For a patient separation of 24 cm, these areas reach 114% of the dose at the centre
and the use of higher energy beams should be considered to give a more even dose
distribution.
16 Basic physics

6MV 15MV
6MV 24cm

6MV
12cm

Fig. 2.14  Radiation distribution for a parallel opposite 6 MV beam for patient separations of
12 cm, 18 cm, and 24 cm, and for a parallel-​opposed 15 MV beam for a patient separation
of 24 cm. For the higher-​energy beam the depth of the high-​dose areas are more interior.

2.6  Machine dependent factors affecting

the dose deposition
2.6.1  Field size effects
As the radiation field size is increased, the amount of radiation to a point on the cen-
tral axis per monitor unit increases. The major cause of this is an increase in the con-
tribution of scattered radiation both from within the treatment machine head and
within the patient. This will affect the depth dose curves; the depth of dmax will decrease
and the percentage depth dose at depth will increase compared to a smaller field size.
Measurements of the relative output at different field sizes and of appropriate isodose
distributions are performed during commissioning. When irregular field shapes such
as those produced by MLCs are used, the contribution of scatter is more difficult to
assess and individual calculations for each patient may be necessary.

2.6.2  Effect of distance

As the distance of the patient from the accelerator is increased, the intensity of the ra-
diation at the surface will decrease due to the inverse square law, the irradiated area will
increase due to the beam divergence, and the penetration of the radiation will increase
due to a reduction in low energy photons in the beam. The increased area is utilized
in treatments such as total body irradiation where the patient is placed at an increased
distance (typically 4 m) from the machine. A slightly increased mean energy may be
noticed for example when changing from a Cobalt machine with an isocentre at 80 cm
to one with an isocentre at 100 cm. The effect for a 6 MV beam is shown in Fig. 2.15.

2.7  Modifications to the radiation beam

2.7.1 Wedges
The intensity of the radiation can be modified by the presence of a wedge. These are
used for multifield plans to compensate for the weighting of other fields, for example
in the treatment of the parotid or the prostate. They can also be used to compensate
 Modifications to the radiation beam 17

120

100
Percentage depth dose

80

60

40

20

Fig. 2.15  Photon depth

0 dose curves for 6 MV
0 5 10 15 20 25 30 35 40
photons at different
Depth (cm) water
focus to surface
80cm FSD 125cm FSD distances.

for missing tissue as in the case of breast radiotherapy, or low-​density tissue in part of
the field as for lung treatments. Examples of typical plans for these sites are given in the
appropriate chapters. Three types of wedge are in current use:
◆ Manually fitted wedges which are usually external to the treatment head.
◆ Steep internal wedges which are driven in and out of the beam by a motor.
◆ Dynamic wedges where the jaw moves across the field partway through the treatment
reducing the beam intensity to give an appropriate profile, as illustrated in Fig. 2.16.
The definition of wedge angle has changed over time as the design of accelerators
has changed. One definition of wedge angle is the angle between the central axis of
the beam and the normal to the 50% isodose. Motorized internal wedges are typically
approximately 60° and are combined with an open field of the same size to produce
different effective wedge angles. Internal and dynamic wedges have the advantage that
they do not have to be manually lifted by the treatment unit staff reducing staff in-
juries and these are now present in the majority of machines. Physical wedges (either
internal or external) will change the penetration of the beam on the central axis as well
as modifying the radiation profile, whereas dynamic wedges do not change the pene-
tration of the beam on the central axis. External wedges can also increase the patient
surface dose slightly due to the production of scattered radiation.

2.7.2  Multi-​leaf collimators
These are used to shape the radiation beam to protect organs at risk or to modify the
intensity of the beam by using a segment field as simple intensity-​modulated radio-
therapy (IMRT). The width of the MLC leaves varies between machines, high reso-
lution MLCs used in stereotactic work have a width of 2–​3 mm projected at isocentre,
whereas other designs of accelerator have leaves which are 10  mm projected at
isocentre.
18 Basic physics

Fig. 2.16  Diagram to illustrate the dynamic wedge. The top row shows the jaw positions,
the second row the beam intensity, and the third a plot of beam intensity across the
field. The deeper shades represent more radiation.

2.7.3  Compensators and

intensity-​modulated radiotherapy
These are discussed further in Chapter 3.

2.8  Geometrical characteristics of radiation beams

2.8.1 Beam divergence
As the radiation beam leaves the treatment head it will diverge. This is illustrated in
Fig. 2.17. The divergence of the beam must be considered when treating adjacent areas
on the patient if overlapping areas of high dose are to be avoided.

2.8.2 Penumbra
The intensity does not drop immediately to zero at the edge of the radiation beam.
The width of the penumbra (defined as distance between the 80% and 20% isodoses
at the edge of the field) is affected by both geometric and dosimetric factors: geometric
factors include the size of the focal spot and the geometry of the treatment head (in
particular the collimators and their distance from the X-​ray source); dosimetric fac-
tors include the width of the dose deposition kernel. The width of the kernel varies
with the energy of the beam and the density of the material in which the dose is de-
posited. In low-​density materials, the electrons will travel further and so the size of the
kernel will be increased and the penumbra will be larger.
 Beam matching and use of asymmetric fields 19

2R
R

2X
Fig. 2.17  The inverse square law. A beam of X-​rays is emitted from a point and passes
through a circular collimator. At a distance X from the source the radius of the beam is
R, the area of the beam cross-​section is A = μR2 and the number of photons per square
centimetre is I. At a distance of 2X from the source, the beam radius is 2R (by similar
triangles).Thus the irradiated area = μ(2R) 2 = 4μR2 = 4A and the intensity of photons
must be I/​4. In general, at a distance x, the intensity is (X/​x)2 × I.

2.8.3 Asymmetry
The majority of modern accelerators have the ability for the jaws to be moved inde-
pendently of each other, producing fields that are asymmetric about the beam centre.
This can be useful for a non-​symmetrical volume and for producing non-​divergent
beam edges for use in beam matching. The width of the penumbra produced by an
asymmetric jaw placed at the centre of the field may be less than that produced by the
corresponding jaw placed at a distance from the central axis.

2.9  Beam matching and use of asymmetric fields

For some patients it is necessary to treat with two adjacent radiation beams either
because of the size of the field or because of the geometry of the radiation volume,
for example in the treatment of lateral and anterior neck fields. In these cases, if no
modification is made, because of divergence the beams can only match at a single
point, areas of over and under dose will be present above and below this. Optimization
will ensure that there is minimum overlap between the beams or areas of potential
underdosing (Fig. 2.18). It is preferable to match the beams in an area in which there
is no residual disease. For patients undergoing fractionated treatment moving the

(a) (b)

Skin surface

Field 1 Field 2 Field 1 Field 2

Fig. 2.18  Adjacent divergent radiation fields matched at the skin surface (a), matched at
depth (b). The dark area indicates field overlap and regions of potential overdose and the
hatched area region potential underdose. (Angles exaggerated for purpose of illustration.)
20 Basic physics

(a) (b)

Skin surface

Field 1 Field 2 Field 1 Field 2

Fig. 2.19  Methods for removing divergence. The centre of the fields are shown by the
dotted line and the cross. In (a), a single isocentre is used and each field uses an asymmetric
jaw to remove the divergence. In (b), a gantry rotation is used to remove the divergence.

position of the match on alternate days may be considered. Three cases are considered
as described as follows.

2.9.1  Photon photon match

Use of a single isocentre is often considered as the gold standard, allowing the
smoothest match between fields. Matching is usually done at the isocentre limiting
the length of the fields to 20 cm in most modern accelerators. It is also possible to use
other match points which increases the field length available. The quality control of
the asymmetric jaws must be excellent otherwise a systematic overlap or underdosing
could occur, and the sharp penumbra obtained at the centre of the field means that if
there is an overlap, the dose would increase rapidly (Fig. 2.19).
An alternative technique is to use one asymmetric jaw with collimator and couch
rotation to remove the divergence from the other beam or beams. The required couch,
collimator, and floor rotation can be calculated using established formula and tables of
these values are often produced. This method allows for longer fields and also broadens
the penumbra of one beam meaning that a slight systematic error in jaw position is less
likely to lead to a hotspot. However, it is much less elegant than the single isocentre
technique and the different widths of the penumbra will lead to a slight overdose in
one field and a slight underdose in the other.
It is also possible to match photon beams without the use of any asymmetric jaws.
Casebow(1) equations can be used to compute the required angles.

2.9.2  Photon electron match

When matching an electron beam to a photon beam the divergence from the electron
beam is usually accounted for by couch collimator and floor rotation. The bulging of
the low dose isodoses means that it is not possible to achieve an exact match and there
is usually a small area of slightly increased dose in the photon field.
 Patient dependent factors affecting the dose deposition 21

2.9.3  Electron electron match

The matching of electron fields is complex particularly on curved surfaces due to
the bulging of the low dose isodoses. On a flat surface, an approximate match can be
obtained by using appropriate gantry and collimator twist to remove divergence; how-
ever, there will be areas of non-​uniform dose in the junction area.

2.10  Patient dependent factors affecting

the dose deposition
2.10.1 Inhomogeneity
The deposition of dose is dependent on the electron density of the material involved.
Within tissues, electron density is directly proportional to the physical density.
Information about the relative electron density of a material can be obtained from
a computed tomography scan which can be calibrated using tissue-​like materials to
relate the Hounsfield units of the scanner to the density of material. Some planning
systems, particularly those using convolution algorithms, require calibration in terms
of physical density. The majority of planning systems are able to correct for density on
a pixel-​by-​pixel basis although correction for large areas of tissue using a bulk density
correction (where a single typical density is chosen for a whole region of the body, for
example setting the relative density of the lung to 0.3) is also applied. Fig. 2.20 illus-
trates the effects of different densities of tissue on the radiation beam. In low-​density
materials, the radiation will travel further before depositing dose, whereas in high-​
density materials it will be attenuated more rapidly. The increased lateral spread of the

110.0 %
105.0 %
100.0 %
95.0 %
90.0 %
80.0 %
70.0 %
60.0 %
50.0 %
40.0 %
20.0 %

Fig. 2.20  12×12 cm 6 MV photon

field incident on a phantom
consisting of a 2-​cm slab of water
followed by inhomogeneities. The
region on the left has a density of
0.3 gcm-​3, whilst that on the right
has a density of 1.6 gcm-​3. The
central column is water. Note the
broader penumbra in the low density
region, the high value isodoses curve
towards the centre of the beam
whilst the low value isodoses spread
away from the beam edge.
22 Basic physics

radiation into low-​density materials such as lung is also illustrated in this figure. Note
that there is also a decrease in dose in the central water column at the junction with
the low-​density material. Many of the simple inhomogeneity correction algorithms are
unable to account for changes in lateral scatter due to the presence of inhomogeneities.
The dose in the region of the interface is complex and an in-​depth study of this is be-
yond the scope of this chapter.

2.11  Treatment planning systems

A wide range of algorithms are used in TPSs. The majority will accurately calculate
doses (to within 1–​2%) for square and rectangular fields incident perpendicular to a
flat homogeneous phantoms at the reference distance. Limitations become apparent
when irregular fields, inhomogeneities, surface curvature, wedges, and different focus
to surface distances are incorporated. Tables 2.1 and 2.2 summarize the calculation

Table 2.1  Summary of main differences between planning systems

Algorithm type Comments
Stored beam Based on the work of Bentley-​Milan(2) further developed by Redpath(3).
Data are stored on an array of fan lines; Intermediate data are
interpolated.
Very accurate for square fields on homogeneous media; however,
less accurate for field shapes and sizes where measured data are
not available such as conformally shaped fields or IMRT. A variety of
algorithms are used to correct for inhomogenities which vary in their
accuracy.

Beam model Dose divided into primary and scatter component, which are considered
systems separately. More accurate than stored beam models for irregular fields;
however, there are often limitations for fields in which the scatter is
uneven (steep wedges and inhomogeneities).
Convolution In these algorithms the kernel (dose deposition from an individual
algorithms photon) is ‘convolved’ with the TERMA (energy removed from the beam).
Convolution When convolution algorithms were first developed(4) the computing
algorithms:  power available made it unrealistic to convolve dose deposition kernels
pencil beams with TERMA at every point in the calculation matrix. In order to reduce
the computing requirements, kernels were defined to represent the
entire dose distribution due to an incident small (‘pencil’) beam. The
intensity of these pencil beams across the field could be varied to allow
for changes in intensity caused by wedges or compensators, and they
could be scaled to allow for inhomogeneities; however, accounting
for scatter from inhomogeneities lateral to the pencil beam was not
possible. In some implementations the pencil beam was calculated using
appropriate Monte Carlo calculated kernels; in others it was derived
from the measured data. Modern versions of these algorithms such as
the anisotropic analytical algorithm (AAA) split the pencil beam kernels
into primary and scatter and allow the distortion of the scatter kernel
to better correct for the distortion of the dose distribution close to
inhomogeneities.
 Treatment planning systems 23

Table 2.1 Continued
Algorithm type Comments
Convolution Individual kernels are retained and convolved with TERMA. To reduce
algorithms: the computational power needed, radiation transport to or from a point,
collapsed cone viewed in polar co-​ordinates, is ‘collapsed’ onto the axes of a set of
cones centred on that point to generate the polar equivalent of large
pixels. The retention of individual kernels makes it possible to account for
the lateral disequilibrium in regions of low or high-​density material; when
computing kernels account must be taken of the polyenergetic nature of
the beam.
Multigrid ‘Multigrid’ or ‘adaptive’ techniques reduce computation time,
superposition and are usually used with superposition algorithms (though they
can be used with any calculation algorithm). In this approach, the
spatial frequency of dose calculation is varied depending on the
dose gradient at each point within the volume being calculated.
Interpolation, usually linear, is used to estimate the dose between
calculated points. This allows the dose calculation to be done at fewer
points (for a given accuracy) than is the case if a uniform rectilinear
grid of calculation points is used.
Monte Carlo True Monte Carlo calculations follow the histories of millions of particles
using sound physical principles combined with statistical methods to
determine the outcome at each point. They are currently regarded as
the gold standard for calculations; however, they have not been widely
clinically implemented due to the intensive computer requirements to
generate a plan in a reasonable amount of time. Methods for reducing
the time required include macro Monte Carlo and Voxel Monte Carlo,
where the energy and fluence distribution for a smaller number of
particles is used to characterize a distribution. This is then used as input
for the next phase of the model.
Boltzman In some situations, it is possible to formulate differential equations for
linear transport radiation transport through the volume to be considered. These may be
equation (BLTE) solved numerically. In theory, this allows a calculation that rivals Monte
Carlo for accuracy, since the basic physics of the transport equations is
the same as that used to compute histories in Monte Carlo techniques.
In practice, the accuracy of both depends on the implementation used,
the effects of pixel sizes, and the treatment of multiple scatter. BLTE
algorithms however may, in some implementations, run many times
faster than Monte Carlo approaches. This approach has yet to be widely
implemented clinically(5).

Source: data from Redpath AT et al., A Comprehensive radiotherapy planning system implemented in

Fortran on a small interactive computer. Br J Radiol. Volume 50, pp. 51–​57, Copyright 1977; Storchi PR
et al., Calculation of a pencil beam kernel from measured photon beam data. Phys Med Biol. Volume
44, Issue 12, pp. 2917–​2819, Copyright © 1999; Failla GA et al., Acuros XB advanced dose calculation
algorithm for the Eclipse treatment planning system, Varian medical systems, Palo Alto, CA, USA, Copyright
© Varian medical systems.
24 Basic physics

Table 2.2  Inhomogeneity corrections used in treatment planning systems and their

accuracy
Correction Brief description Limitation Estimate of Ref
method accuracy
Equivalent Radiological path Corrects in direction Errors up to 10.4% 6,7
path length lengths calculated of ray trace only. No
using relative account taken of
electron density size or position of
inhomogeneity

Batho and Uses TPRs Poor agreement -​0.17 for Co60 8,9,10
modified is obtained within beyond cork(10),
Batho the inhomogeneity, 3% within(13)
especially for large field
sizes, because of lateral
disequilibrium, but
agreement beyond the
inhomogeneity is good.
Equivalent All CT slices are Overestimates dose 0.18% for Co60 11,3,7
tissue–​air combined to give within low density beyond cork(10)
ratio (ETAR) one effective slice for inhomogeneity, Errors up to
the computation of particularly for small 7.2% within
scatter. This reduced field sizes because of inhomogeneity(6).
the calculation time lateral disequilibrium.
compared with
systems that retain
each individual slice
but reduces the
accuracy.
Convolution Dose deposition Allowing spatially 12

techniques kernels are scaled variant kernels

according to the increases calculation
density of the media time as transform
techniques cannot be
used. Penumbra width
in low density regions
will be increased.
Monte No additional
Carlo correction needed

methods and inhomogeneity corrections used in most commercial TPSs. TPS algo-
rithms are classified into two categories depending on their ability to model the lateral
electron transport within the patient. The older type A algorithms do not account for
energy transported laterally by electrons and can only correct for tissue inhomogen-
eity in the direction of the primary radiation; in contrast, type B algorithms correct
for the effects of the inhomogeneity both lateral to and in the primary direction. This
produces noticeable differences, for example, in lung plans
 References 25

Further reading
Cunningham JR. Tissue inhomogeneity corrections in photon-​beam treatment planning.
Progress in Medical Radiation Physics 1982; 1:103–​31.
Green D, Williams PC Linear Accelerators for Radiation Therapy 2nd edition. Institute of
Physics Publishing, Bristol, 1997.
Hurkmans C, Knoos T, Nilsson P, Svahn-​Tapper G, Danielsson H. Limitations of a pencil
beam approach to photon dose calculations in the head and neck region. Radiotherapy and
Oncology 1995; 37:74–​80.
International Electrotechnical Commission (IEC) 60601-​2-​1 Ed 2.0 Medical electrical
equipment—​Part 2-​1: Particular requirements for accelerators in the range 1MeV to 50MeV
IEC 1998.
Johns HE, Cunningham JR The Physics of Radiology CC Thomas, Springfield, IL, 1983.
Metcalfe P, Kron T, Hoban P .The Physics of Radiotherapy X-​Rays from Linear Accelerators
Madison, WI, Medical Physics Publishing, 1997.
Thwaites, DI, Williams JR Radiotherapy Physics in Practice Oxford, Oxford University
Press, 1993.
British Journal of Radiology Supplement 25 Central Axis Depth Dose Data for Use in
Radiotherapy:1996 British Institute of Radiology, London, 1996.

References
1. Casebow MP. Matching of adjacent radiation beams for isocentric radiotherapy British
Journal of Radiology 1984; 57:735–​40.
2. Bently RE, Milan J. An interactive digital computer system for radiotherapy treatment
planning. British Journal of Radiology 1971; 44:826–​33.
3. Redpath AT, Vickery BL, Duncan WA. A Comprehensive radiotherapy planning system
implemented in Fortran on a small interactive computer. British Journal of Radiology 1977;
50:51–​7.
4. Storchi PR, van Battum LJ, Woudstra E. Calculation of a pencil beam kernel from
measured photon beam data. Physics in Medicine and Biology 1999;.44:2917–​28.
5. Failla GA, Wareing, T, Archambault Y, Thompson S; Acuros XB. Advanced dose
calculation algorithm for the Eclipse treatment planning system. Palo Atlo, CA: Varian
medical systems.
6. Butson MJ, Elferink R, Cheung T, Yu PKN, Stokes M, Quach KY et al. Verification of
lung dose in an anthropomorphic phantom calculated by the collapsed cone convolution
method. Physics in Medicine and Biology 2000; 45:143–​9.
7. Parker RP, Hobday PA, Cassell KJ. The direct use of CT numbers in Radiotherapy
dosage calculations for inhomogeneous media. Physics in Medicine and Biology 1979;
24:802–​9.
8. Batho HF. Lung corrections in Cobalt 60 beam therapy. Journal of the Canadian
Association of Radiologists 1964; 15:79–​83.
9. Cassell KJ, Hobday PA, Parker RP. The implementation of a generalised Batho
inhomogeneity correction for radiotherapy planning with direct use of CT numbers.
Physics in Medicine and Biology 1981; 26(4):825–​33.
10. Wong JW, Purdy JA. On methods of inhomogeneity corrections for photon transport.
Medical Physics 1990; 17(5); 807–​14.
26 Basic physics

11. Sontag MR, Cunningham JR The equivalent tissue air ratio method for making absorbed
dose calculations in a heterogeneous medium. Radiology 1978; 129; 787–​94.
12. Mackie TR, Scrimger JW, Battista jJ. A convolution method of calculation dose for 15 MV
x-​rays. Medical Physics 1985; 12(2); 188–​96.
13. Cunningham JR. Tissue inhomogeneity corrections in photon beam treatment planning.
Progress in Medical Radiation Physics 1982; 1:103–​31.
 Chapter 3

Treatment delivery,
intensity-​modulated radiotherapy,
and image-​guided radiotherapy
Yat Man Tsang

3.1  Introduction and background

Every year there are approximately 350,000 newly diagnosed cancer patients with
more than half of them dying from cancer in the UK (1). Radiotherapy plays a key role
in cancer management and it is second only to surgery in terms of its effectiveness
in treating cancer(2). The successful delivery of external beam radiotherapy involves a
number of complex processes beginning with the decision by the clinical oncologist to
use radiotherapy as part of the patient’s cancer management, through the preparation
and planning of the patient’s treatment, to the verification of the patient position and
radiation dose delivered at the time of treatment. Fig. 3.1 highlights the major pro-
cesses involved in external beam radiotherapy.
The aim of radiotherapy is to deliver a homogenous radiation dose to tumour in-
side the patient, whilst minimizing dose to all other parts of the body, in particular
to organs which are especially radiosensitive or in close proximity to the tumour. In
order to begin to achieve these aims, detailed information is required about the tu-
mour position, size, and shape within the patient and the location of radiosensitive
organs at risk (OARs). This is achieved by utilizing three-​dimensional (3D) patient
images for treatment target and OARs localization. With its excellent spatial reprodu-
cibility, kilovoltage X-​ray computed tomography (CT) imaging is the current standard
modality for radiotherapy planning. The various structures are outlined on a series of
axial slices to produce 3D volumes. Often, other imaging modalities with superior tu-
mour imaging such as magnetic resonance imaging (MRI) or positron emission tom-
ography (PET) are incorporated into the planning process with CT for more accurate
target volume definitions.
Once this has been done, a radiotherapy plan can be designed for the individual pa-
tient to meet the treatment goals using a detailed computer model of the way radiation
dose will be deposited within the patient anatomy.
To obtain accurate dose calculations from the patient representation, the relative
electron or physical density of each voxel of the patient is required. Corrections can
28 Treatment delivery, intensity-modulated radiotherapy

Clinical treatment decision

- Radiotherapy required as part of the patient’s overall management
- Intention of radiotherapy treatment: palliative or curative (radical)
- Justification of radiotherapy treatment: risk benefit analysis

Treatment preparation
- Patient preparation: immobilization, dietary advice, etc
- Patient imaging: CT, MRI, MRS, PET
- Patient model generation: registration of multi-modality images, delineation of
tumour and radiosensitive normal tissues
- Treatment planning: selection of technique, dose distribution generation and
optimization
- Quality assurance/dosimetry checks of the treatment plan

Treatment execution
- Patient set-up: verification of patient position & adjustment
- Dosimetric verification: in vivo dosimetry
- Quality assurance/dosimetry checks: record & verify

Fig. 3.1  The radiotherapy process.

be applied from knowledge of the mass attenuation coefficients of different tissue

types in the CT. This is usually achieved by the use of a look-​up table within the treat-
ment planning system that converts the CT numbers in the CT images to electron
or physical density relative to water. The look-​up table is generally derived from the
CT scan of a phantom containing a number of inserts of known density. It can be de-
scribed by two linear fits; the first linear fit for the CT number range −1000 to 50, and
the second linear fit for CT numbers greater than 50. Some planning systems require
that 1000 be added to CT numbers within a CT image so that all numbers become
positive, and therefore reduce the data storage requirements. Fig. 3.2 is a look-​up
table of CT number against physical density with 1000 automatically added to all CT
numbers.

3.1.1  ICRU 50 and 62

Radiotherapy given with the intent to cure the patient, termed radical radiotherapy, is
given by small doses of radiation delivered once or twice a day for several weeks. Each
radiotherapy treatment session is called a fraction of radiotherapy. When considering
the development of the radiotherapy plan it is clear that the initial 3D image or images
used to develop the patient anatomical model is a single snapshot of the patient con-
cerned, whereas the resulting treatment plan has to be designed in a way that delivers
an adequate radiation dose to the tumour when delivered with multiple fractions of
 Introduction and background 29

Fig. 3.2  A look-​up table to convert from CT number to relative physical density used by
a treatment planning system to calculate dose. All CT values have automatically been
increased by 1000 so that air has a CT value of 0 and water a value of 1000.

radiotherapy over a time period of several weeks. To ensure that the treatment plan is
capable of meeting this requirement, treatment margins have to be added to the ini-
tial tumour shape at the treatment planning stage. The International Commission on
Radiation Units and Measurements (ICRU) have produced a number of reports(1–​3)
with the aim of standardizing the practice of prescribing radiotherapy and designing
treatment plans that will give an adequate and reliable dose to the tumour when
treating with fractionated radiotherapy.
ICRU report 50(3), published in 1993, was developed to contain recommendations
on how to report a treatment in external beam radiotherapy. It introduced common
terminology for reporting and prescribing the radiotherapy treatment that could be
followed in all centres worldwide and intended to standardize prescribing practice.
In relation to treatment planning the report gave clear guidance on the need for treat-
ment margins to ensure that the tumour is adequately treated with a fractionated
course of radiotherapy. The volumes described in the report are outlined in Table 3.1.
In addition to the treatment volumes, the report outlined that treatment plans should
be prescribed to a stable point of high dose and locally a low-​dose gradient within the
planning target volume (PTV); this was termed the ICRU reference point. Often the
isocentre of the treatment is used as the appropriate ICRU reference point. The treat-
ment plan should also be designed with a limit on the variation in dose within the PTV
of −5% and + 7% of the ICRU reference point.
A supplement to ICRU 50 was produced in 1999 as ICRU report 62(4) to address the
changes in treatment planning and radiotherapy treatment that had occurred with
the increased use of pretreatment imaging, computerized treatment planning, and
treatment verification. ICRU report 62 introduced the concepts of planning-​risk vol-
umes (PRVs) and the conformity index. It retained the basic concepts of gross tumour
volume (GTV), clinical target volume (CTV), and PTV but refined the construction
of the margin required to create the PTV with new volumes for radiotherapy treat-
ment planning. Table 3.2 outlines the additional margins and volumes described in
ICRU 62.
30 Treatment delivery, intensity-modulated radiotherapy

Table 3.1  The various volumes described by ICRU 50

Volume Description Comments
name
Gross tumour The gross palpable or visible/​demonstrable The GTV is usually defined from
volume (GTV) extent and location of the malignant reconstructed 3D images of the
growth patient such as CT, MR, PET.
Multiple image modalities can
be used to define this volume.

Clinical target The tissue volume that contains a GTV
volume (CTV) and/​or subclinical microscopic malignant
disease, which has to be eliminated
Planning A geometrical concept, defined to
target volume select appropriate beam sizes and beam
(PTV) arrangements, taking into consideration
the net effect of all the possible
geometrical variations and inaccuracies in
order to ensure that the prescribed dose is
delivered to the CTV
Treated The volume enclosed by an isodose
volume (TV) surface, selected and specified by the
radiation oncologist as being appropriate
to achieve the purpose of treatment
Irradiated The tissue volume which receives a dose
volume (IR) that is considered significant in relation to
normal tissue tolerance
Organs at risk Normal tissues whose radiation sensitivity
(OARs) may significantly influence treatment
planning and/​or prescribed dose
This volume has to be treated
adequately in order to achieve
the aim of the therapy: cure or
palliation.
Although it may be described
as a geometric expansion of the
GTV, the CTV is constrained by
anatomical boundaries such as
bone.
The size of the margin from
CTV to PTV can only be reduced
if the likely geometric variations
are reduced.
The 90% or 95% isodose is
often used to define the treated
volume.
The 50% isodose is often used
to define the irradiated volume.

Source: data from International Commission on Radiation Units and Measurements. Prescribing, recording
and reporting photon beam therapy. Report 50. Bethesda, MD: ICRU, 1993.

The report described OARs as having distinct tissue architectures:

◆ Serial OARs, for example, the spinal cord, are the continuous organs in which ra-
diation damage at one point of the organ impairs the complete function of the
entire organ;
◆ Parallel OARs, for example, the lungs, are the organs consisting of several func-
tioning units. The main parameter impairing the organ’s function is the proportion
of the OAR receiving a dose above a specified tolerance. In reality, many organs
have tissue architecture with both serial and parallel components.
 Introduction and background 31

Table 3.2  The volumes described in ICRU 62

Volume name Description
Internal The margin that must
margin (IM) be added to the CTV to
compensate for the expected
physiological movements and
variations in size, shape, and
position of the CTV during
treatment
Comments
The motion occurs when the CTV position
changes on a day-​to-​day level and is mainly
associated with organs that are part of
or adjacent to the digestive or respiratory
system.
Changes in the patient’s condition, such as
weight gain/​loss, can also affect the relative
position of the CTV.

Set-​up margin The margin that must
(SM) be added to the internal
margin to compensate for
the expected motion of the
internal margin due to the
repeated set-​up of the patient
during treatment
Planning risk The OAR volume with a
volume (PRV) margin added to compensate
for internal motion of the
OAR and changes in position
due to set-​up errors
Conformity The quotient of the treated
index (CI) volume and the volume of
the PTV (used to compare
techniques)
The uncertainties depend on different
factors and can include variations in patient
positioning, mechanical uncertainties of the
equipment, dosimetric uncertainties (light-​
radiation field agreement), transfer set-​up
errors, and human-​related uncertainties.
The PRV concept is often used for serial
OARs where the maximum dose to the PRV
is kept below tolerance to ensure that the
OAR is not impaired if a systematic set-​up
variation occurs when treating the patient.
The conformity index can be used to
compare different treatment techniques.
The concept is insensitive to the shape of
the treated volume compared to the PTV
volume, i.e. some parts of the PTV may
not receive adequate dose but the treated
volume may be adequate due to the
irradiation of tissue outside the PTV in other
regions.

Source: data from International Commission on Radiation Units and Measurements. Prescribing,

recording and reporting photon beam therapy. Report 62 (Supplement to ICRU Report 50). Bethesda,
MD: ICRU, 1999.

Fig. 3.3 demonstrates the delineation of structures for a patient with a naso-
pharyngeal tumour and includes the delineation of the GTV, CTVs, and PTVs, as
well as the spinal cord and brainstem as radiosensitive OARs in this region. The
spinal cord and brainstem are both serial-​like structures and so the expansion of
both volumes by 5 mm to create PRVs as described in ICRU report 62 can clearly
be seen.
Margin recipes have been developed to describe the required margin size to en-
sure the CTV is covered by a certain isodose. The margin recipes can be derived from
Monte Carlo methods where a number of systematic and random errors are intro-
duced and the dose to the CTV determined. The most commonly used margin re-
cipe derived from a Monte Carlo study of prostate radiotherapy treatments defines
32 Treatment delivery, intensity-modulated radiotherapy

Fig. 3.3  The delineated volumes for a patient with a nasopharyngeal tumour showing
the GTV (orange), CTV1(sky blue), and CTV2 (green) that will receive different radiation
doses, and the spinal cord (red) and brainstem (yellow green) with their respective PRV
volumes also shown.

the margin size to ensure that the CTV receives at least 95% dose in 90% of patients(5).
This is given as:

( )
M ptv = 2.5 ∑ + 1.64 √ σ2 + σ p 2 − 1.64 σ p

(1)

where Σ refers to the systematic variation, σ to the random variation, and σp to the
penumbra margin.
For the case of soft tissue and megavoltage photon beams where σp ≈3.2 mm the
equation can be simplified to:
M ptv = 2.5 ∑ + 0.7 σ (2)

The recipe highlights that the systematic variation is more than 3.5 times more im-
portant in the resulting margin required than random variations. This is because in
 Forward planning process 33

general a systematic variation results in a shift in the dose distribution leading to an

underdose to some part of the CTV whereas random variations lead to a blurring of
the dose distribution with a less pronounced dosimetric effect.
Similar margin recipes have been developed for OARs. In general, if an organ is
parallel and large, and the risk of complication is known to be acceptable, then the
clinician may decide to ignore the geometric error. However, for dose levels that cause
unacceptable complications, e.g. serial organs, then the geometric errors need to be
taken into account by the application of margins to OARs to create larger PRV vol-
umes. Dose constraints for serial OARs are then applied to the PRV volume. A margin
recipe for small and/​or serial organs at risk in low (−) or high (+) radiation dose re-
gions has been proposed(6) as:

M prv = 1.3 ∑ ± 0.5 σ (3)

where again Σ refers to the systematic variation and σ to the random variation.
Margins applied in practice are often based on legacy protocols or adopted from
clinical trials. In reality, treatment planning margins should be based on an individual
departmental practice and information regarding levels of systematic and random
error. Margins should only be reduced if the technique has been altered or developed
to reduce either the systematic and/​or random errors in the radiotherapy process.
Otherwise reducing the margin to improve the dosimetric quality of the treatment
plan may well result in the underdosing of the CTV during the fractionated course of
radiotherapy.

3.2  Forward planning process

The process of developing an individualized radiotherapy plan has traditionally been
performed using forward planning. With this approach a planner starts the process by
choosing the appropriate number and directions for the treatment beams to be used.
In conventional radiotherapy the shape of each beam is simply square or rectangular.
More commonly, the beam shape is created to conform to that of the PTV from each
treatment beam direction by the application of multileaf collimators (MLCs); this is re-
ferred to as conformal radiotherapy. Fig. 3.4 shows a lateral beam from a conventional
and conformal radiotherapy plan highlighting how the use of the MLC for each treat-
ment beam can reduce the dose to OARs, the bladder and rectum in this case as por-
tions of each OAR are now shielded from high levels of radiation dose by the MLCs.
The treatment planner has a number of options available in developing a clinically
acceptable treatment plan such as the radiation modality used, the energy of the ra-
diation used, the relative beam weights of each treatment beam, and the use of simple
wedged beams for some or all treatment beams. The planner goes through an itera-
tive process to alter the available treatment parameters to produce a plan that meets
the ICRU limits for coverage of the PTV, and meets the dose constraints established
for OARs.
Wedged beams can either be produced by the introduction of a physical wedge
into the beam, or from the dynamic motion of the jaw. When the physical wedge is
34 Treatment delivery, intensity-modulated radiotherapy

(a) (b)

Fig. 3.4  (a) A conventional lateral beam for a patient with prostate cancer showing the
PTV (sky blue), the rectum (brown) and the bladder (yellow-​green). (b) The same beam
with a multileaf collimator (MLC) used to reduce doses to the bladder and rectum, the
main organs-​at-​risk.

integrated in to the treatment head and computer controlled, this is termed a motor-
ized wedge. Fig. 3.5 shows isodose plots for an open, 20°, and 55° wedged beams at 6
MV using a motorized wedge. The wedged beam creates a non-​uniform intensity in
one direction only. The amount of radiation leaving the radiation machine, the linac,
is less at the thick end of the wedge. Similar plots to those seen in Fig. 3.5 can be pro-
duced by the dynamic motion of a jaw across the beam creating an approximately
linear change in intensity. Wedged treatment beams of varying angles can be produced
either by the combination of an open and motorized wedge beam, or by altering the

(a) (b) (c)

100%
90%
80%
70%
60%
50%

Fig. 3.5  Isodose plots for radiation beams at 6 MV for (a) an open beam, (b) a 20°
wedge, and (c) a 55° degree.
 Forward planning process 35

speed of the dynamic jaw across the treatment beam. These simple non-​uniform in-
tensity treatment beams can be used to improve the homogeneity of the dose within
the PTV. In the most simplistic example, the wedged beam compensates for missing
patient tissue.
Fig. 3.6 shows an example of wedged beams used in a simple parallel-​opposed pair
treatment of a larynx. Less radiation is required anteriorly as the patient thickness
is less in this region than posteriorly where the greater thickness of the patient will
increase the attenuation of the beam. The introduction of wedged beams is able to
significantly improve the homogeneity of the radiation dose delivered to the patient.
When using forward planning, the skill of the planner is in the understanding of
how altering the available treatment parameters to individualize the patient’s treat-
ment plan. This involves the refinement of beam weight for different treatment beams,
the addition of a wedged beam, and the changing of a wedge angle for a wedged beam.
The adjustments of all these parameters can affect the resulting dose distribution in
order to achieve the aims of treatment.
Once a treatment plan has been generated, it needs to be evaluated to ensure that
it meets the clinical requirements of the proposed radiotherapy treatment. At a basic
level, this is done by inspecting lines of isodose on each CT slice of the patient repre-
sentation axially, sagittally, and coronally. This approach can determine whether the
PTV receives an adequate radiation dose and the OARs are not receiving too much
radiation.
There is often a significant amount of information to be considered at each treat-
ment plan evaluation. Simple graphical tools and the reduction of the 3D dose dis-
tribution to single numbers via the use of biological models have been developed to
aid in the process of evaluating treatment plans. For example, a plot of the cumulative

Fig. 3.6  A clinical example

of use of wedged beams
to compensate for missing
tissue in the radiotherapy
treatment of the larynx.
36 Treatment delivery, intensity-modulated radiotherapy

dose–​volume frequency for a particular volume, such as the PTV(s) or OARs can be a
useful tool used at the treatment planning stage. The plot graphically summarizes the
radiation dose distribution to a volume of interest of a patient and is commonly known
as the dose–​volume histogram (DVH)(7). DVHs can be used to aid the evaluation of
the treatment plan for a particular patient, or to compare multiple treatment plans.
Fig. 3.7 demonstrates a typical DVH for a patient with prostate cancer and graphically
demonstrates the dose distribution to the PTV as well as the major OARs in this re-
gion, the rectum, the bladder, and the femoral heads. For any given dose on the x-​axis,
the ordinate on the DVH is the percentage volume of the structure receiving that dose
level or more. Therefore each volume has an ordinate value of 100% for zero dose.
As well as looking at the isodoses on the patient anatomical representation and
investigating the DVHs of PTV(s) and OARs, biological models involving tissue re-
sponses to radiation can be used to evaluate treatment plans. The most commonly used
biological models are the tumour control probability (TCP) and normal tissue compli-
cation probability (NTCP) of OARs(8), and the equivalent uniform dose (EUD)(9). As
with all models the validity of their use in radiotherapy relies heavily on the param-
eters used within them. The Quantitative Analyses of Normal Tissue Effects in the
Clinic (QUANTEC) have published evidence-​based guidelines on how 16 different

Fig. 3.7  A dose–​volume histogram (DVH) for a patient with prostate cancer. The dose
distribution to the PTV and major OARs (rectum, bladder, and femoral heads) are
graphically represented by a DVH.
 Inverse planning 37

organs are damaged by radiation(10). One of the main goals of QUANTEC is to provide
practical guidance on identifying toxicity risk based on DVH parameters or NTCP
model results.

3.3 Inverse planning
3.3.1  Principles of inverse planning
Inverse planning is a process where instead of manually adjusting beam weights and
wedges as in forward planning, the planner is required to describe the dose distribu-
tion that they want at the start of the planning process and then use computer opti-
mization to develop the desired treatment plan.
This planning approach can be described on a voxel-​by-​voxel basis but is more often
described as a series of minimum dose, maximum dose, mean dose, dose–​volume
limit, or biologically based objectives for the outlined PTV and OARs. The skill in
developing a treatment plan is evolving from the planner manually changing beam
angles, weights, and wedges to improve the dose distribution, to altering the objectives
to refine the dose distribution produced by the computer optimization to achieve a
clinically acceptable treatment plan. Fig. 3.8 demonstrates the objectives for an inverse
plan for an example patient with a lung tumour.
Inverse planning is used for most intensity-​ modulated radiotherapy (IMRT),
volumetric-​modulated arc radiotherapy (VMAT), and intensity-​modulated arc radio-
therapy (IMAT) planning for treatments such as tomotherapy. In such treatment tech-
niques, the fluence from any given beam direction does not necessarily correspond
to an intuitive shape of the PTV. The combination of a number of non-​uniform treat-
ment beams allows individualized highly conformal dose delivery to the PTV while
sparing the surrounding OARs. These forms of advanced radiotherapy delivery enable
the delivery of heterogeneous dose distribution in a target volume or multiple target

Fig. 3.8  The objectives used to optimize an IMRT plan for a lung cancer patient.
38 Treatment delivery, intensity-modulated radiotherapy

volumes, and bring the potential of dose escalation to the tumour by increasing the
therapeutic ratio.

3.3.2  Applications of inverse planning: IMRT and VMAT

IMRT is a radiotherapy technique where the intensity of the radiation beams is modu-
lated for the treatments. It is achieved by delivering a number of non-​uniform radiation
beam fluences from different beam directions, in order to customize the radiation dose
that closely conforms to the shape of the PTV.
If IMRT is taken as an example, the inverse planning process begins by the tumour
and relevant OARs being outlined to create an anatomical model of the patient. The
PTVs and PRVs are created by the addition of suitable margins. The planner chooses
the number of treatment beams to be used and the placement of those beams around
the patient. Often five to nine equispaced beam arrangements are employed, but in
some clinical cases the application of individually designed beam directions is advan-
tageous. The planner must also describe the final dose distribution that they want the
planning system to achieve at the end of the optimization process. In simple terms this
can be created by describing what the final dose–​volume histogram curves look like
for the PTV and OARs. It is important for the planner to define the relative import-
ance of meeting the different objectives in the optimization for an efficient planning
process. An example set of objectives for a lung IMRT plan is given in Fig. 3.8.
As the first step of the inverse planning process, the planning system applies initial
fluences, such as open conformal fields of equal weight, from a set of radiation beams
at different directions. The open fluence is divided in to a number of squares, called
bixels, and the intensity in each bixel is allowed to vary to alter the dose distribution.
Computer optimization algorithms are used to refine the fluences of each beam by
changing bixel intensities to minimize the difference between the desired and actual
dose distributions. The modern treatment planning systems use a cost function tool
to determine how close the resulting dose distribution is to the desired distribution
described by the planner. The cost function needs to be minimized and a cost func-
tion of zero would mean that the actual and desired dose distributions were identical.
This is achieved by altering the fluence to see if it reduces the cost function. Usually
downhill optimization methods are used in this process; if the cost function of the
new fluences is less than the previous iteration, the new fluences are accepted and the
process is repeated, if the cost function is higher then the new fluences are rejected.
By altering the bixel intensities in an iterative process, the cost function is reduced
and the dose distribution gets closer to the desired dose distribution given by the
planner. Fig. 3.9 shows the fluences for each beam of a five-​field IMRT plan for a pa-
tient with prostate cancer.
In traditional inverse planning for IMRT, the optimal fluence for each beam direc-
tion is developed by the computer optimization. This ‘ideal’ fluence has to be converted
into something that can be delivered by the linac: a two-​step approach to inverse plan-
ning optimization. There are different ways in which IMRT fluences can be produced
by linacs. Each linac type has different physical limits and constraints on its MLCs
and radiation output. These are required to be taken in to account when converting
from the ‘ideal’ fluence to the ‘deliverable’ fluence. In general, IMRT can be delivered
Fig. 3.9  The final fluences for each beam in an example IMRT plan for a patient with prostate cancer.
40 Treatment delivery, intensity-modulated radiotherapy

either as a series of static subfields with the radiation switched off between subfields
(step-​and-​shoot), or dynamic delivery of continuous motion of the MLCs with the ra-
diation on (sliding window). In either case the deliverable fluence is likely to be slightly
different to the ideal fluence produced by the optimization and this needs to be con-
sidered when evaluating the final treatment plan.
VMAT is a rotational IMRT technique where the intensity of the radiation beams
is modulated using simultaneous variation in gantry rotation speed, dose rate, and
MLC leaf positions. The treatment machine continuously reshapes and changes the
intensity of the radiation beam as it rotates around the patient. One or more arcs can
be used for VMAT delivery. The complex nature of VMAT treatment planning can be
due to the interconnectedness of the beam shapes from one beam angle to the next
in each VMAT arc. The interconnectedness of the adjacent beam shapes is defined
by the motion of the MLCs between radiation beams at adjacent gantry angles. The
MLC motion is often limited by the leaf travel speed and gantry rotation speed. For
example, if the gantry speed is 10 degrees per second and the MLCs travel speed is
4 cm per second, the maximum MLCs travel distance between two adjacent beam an-
gles will be 4 cm. Comparing with the fixed gantry IMRT, VMAT has the same ability
to achieve highly conformal dose distributions but with an improvement in treatment
delivery efficiency. This is achieved by the reduction in treatment delivery time and the
reduction in MU usage in VMAT treatment delivery. Details on clinical applications of
VMAT can be found in the other chapters of the book.
A different way of dealing with how the radiation can be delivered in IMRT and
VMAT is to include the delivery constraints directly in the optimization. This is usu-
ally achieved by beginning the optimization by developing the ideal fluence for a few
iterations of the optimization, converting to a deliverable solution, and performing
further optimization by working directly on the beam shapes and beam weights to
improve the dose distribution. This approach to the inverse problem can lead to fewer
segments when using a step-​and-​shoot approach to IMRT delivery than the two-​step
optimization approach.
It is clear from the process outlined here that the planner is required to define a lot
of information before starting the inverse planning process. In general, by considering
a test cohort of patients for a particular treatment site it is often possible to define a
set of starting conditions for the process including the number and orientation of the
treatment beams and the initial set of planning objectives. This initial set of conditions
is termed as a class solution. Despite the fact that the class solution may require modi-
fication for individual patients, it can not only significantly reduce the time taken to
arrive at a clinically acceptable treatment plan, but also minimize the variations in plan
quality between planners.

3.3.3  Special considerations when using inverse planning

Computer optimization used in inverse planning can minimize the cost function re-
sulting in a dose distribution that closely resembles the desired dose distribution as ini-
tially described by the planner. As the dose distribution gets closer to the desired value,
the optimization works less hard to reduce the cost function further at the next iteration.
This often leads to an undershoot issue. For example, when the planner sets the plan
 Inverse planning 41

objective as more than 95% of the PTV volume receiving 95% of the prescription dose,
the optimization will more likely produce an ‘undershoot’ treatment plan with less than
95% of the PTV volume receiving 95% of the prescription dose. In order to overcome
this, the planner can choose to increase the volume limit, the dose limit, or both in the
initial set of plan objectives, e.g. more than 97% of the PTV receiving 97% of the pre-
scription dose, to obtain a clinically acceptable result at the end of the optimization.
In general, computer optimization algorithms undershoot minimum dose objectives
and overshoot maximum dose objectives. An example of overshoot is that if in lung
planning no more than 35% of the normal lung should receive 20 Gy or more then in
setting an appropriate objective the planner will chose to try to limit the volume re-
ceiving at least 20 Gy to 25% or 30%. The distribution returned by the optimization
will be slightly higher than that set in the objectives but probably less than the clinical
limit of 35%. Fig. 3.10 highlights the set of objectives given in Fig. 3.8 on the DVH with
the final curves for some of the regions of interest. The figure clearly shows overshoot
of the ‘Lungs–​PTV’ objective and undershoot of the ‘CTV’ and ‘PTV–​IMRT’ min-
imum dose/​dose-​volume objectives.

Fig. 3.10  Various DVHs for the objectives set in Fig. 3.8. The set of objectives are shown
graphically on the DVH.
42 Treatment delivery, intensity-modulated radiotherapy

Inverse planning by computer optimization is completely dependent on the quality

and quantity of information provided a priori by the planner. The optimization solely
works towards to minimize the difference between the desired and actual dose dis-
tributions under a set of defined plan objectives. It cannot control the dose to areas
that are not defined in the set of objectives. This can lead to dose distributions that
look quite different to those from forward planning, and may result in adverse treat-
ment complications if dose is inadvertently deposited in normal tissues that have not
been outlined and considered in the inverse planning process. Compared to forward
planning, it is a common practice that more OARs and normal tissues are outlined in
inverse planning. Another way to circumvent this undesired dose distribution issue
is by the use of dummy volumes. This approach can control the dose distribution and
make it mimic the traditional conformal radiotherapy distributions. Dummy volumes
are not anatomically based and may sometimes be automatically generated by modern
treatment planning systems. Fig. 3.11 demonstrates the use of dummy volumes in the
treatment of a patient with a lung tumour. The rinds surrounding the PTV are used
to guide the optimization to create a more conformal dose distribution. By setting a
maximum dose constraint to the rind volume the optimization will attempt to push
the maximum dose isodose constraint to the edge of the rind closest to the PTV.
Radiotherapy techniques such as IMRT, IMAT, and VMAT can produce significantly
higher dose gradients than more traditional radiotherapy. In such circumstances it is
possible that small systematic variations in patient position compared to the planning
CT scan can lead to significantly higher doses to OARs than evaluated at the treatment

Fig. 3.11  An example of dummy rinds used in inverse planning to guide the optimization
process. Such rinds can be used to limit doses to normal structures without the need for
time-​consuming delineation of normal structures.
 Inverse planning 43

planning stage. To overcome this problem PRVs as outlined in ICRU report 62 are
often used for serial OARs in inverse planning to limit the dose a safe distance from
the OAR. For example, if it is unlikely that a patient will have a systematic error of
more than 3 mm in their treatment, then adding 3 mm to a serial organ such as the
spinal cord, and limiting the dose that the PRV volume receives would make it highly
unlikely that the spinal cord would receive a dose higher than that given to the PRV
volume even in the presence of systematic set-​up errors. Fig. 3.11 highlights the use of
a PRV for the spinal cord, a serial-​like OAR that needs to be considered when treating
thoracic tumours.

3.3.4 ICRU 83
ICRU report 83(11) refers specifically to the prescribing, recording, and reporting
of IMRT treatments, but the principles outlined in the report are appropriate to all
radiotherapy approaches employing inverse planning. The general principles of the
previous ICRU reports were preserved within the report. The new concept of the re-
maining risk volume (RVR) was introduced by the report. The RVR is defined by the
difference between the volume enclosed by the external contour of the patient and that
of the CTVs and OARs outlined. It is expected that the RVR may be a useful metric in
looking at different treatment modalities and their risk of producing very late effects
such as second malignancies from the radiation exposure of normal tissues during
radiotherapy treatment.
Refinements of the notation of the GTV were recommended to account for changes
in the delineation of the tumour and the increased use of image-​guided radiotherapy
(IGRT). For example, the GTV can be delineated from a CT scan, or a T2-​weighted
MRI scan. It should be clear when reporting the radiotherapy treatment which image
modality was used for delineation, and so should be given as GTV(CT) or GTV(MRI-​
T2) respectively. With the introduction of IGRT, the GTV may be delineated more
than once during the course of the radiotherapy treatment and this should also be
clearly demonstrated in the notation. For instance, if a treatment plan is created by
taking a repeat MRI scan after 30 Gy of dose has already been delivered to the patient,
then the GTV can be described as GTV(MRI-​T2, 30 Gy).
The report also highlighted the need for consistent dose reporting of PTVs by re-
porting the median absorbed dose, and the near maximum and near minimum ab-
sorbed doses. For OARs and PRVs, the report recommended the reporting of the
mean absorbed dose, the near maximum dose, and VD which if exceeded has a known
probability of causing complication, e.g. V20Gy for lung. The report recommends the
use of D2% and D98% for the near maximum and near minimum doses, where D2% is
the highest dose received by at least 2% of the volume of the PTV or OAR. A more
consistent reporting may be achieved by the reporting of the highest dose or lowest
dose received by at least 1 cc (cm3) as this is independent of the volume of the region
of interest.
In terms of prescribing the dose, ICRU report 83 recommends prescribing the re-
quired dose to the median PTV dose rather than a single point in a region of high-​
dose and low-​dose gradient. For IMRT dose distributions, the isocentre is often not a
suitable prescription point as there are more local dose gradients within the PTV than
44 Treatment delivery, intensity-modulated radiotherapy

for conformal dose distributions, particularly when individual treatment beams are
considered rather than the dose distribution produced from the combination of all
treatment beams.

3.4  Image-​guided radiotherapy
3.4.1  Principles of IGRT
The advanced radiotherapy techniques such as IMRT and VMAT generate highly con-
formal dose distributions with steep dose gradients. The benefits of these techniques
are potentially large for patients with cancer where the tumour shape is complex and
its location often close to OARs. However, the dose distribution of an IMRT/​VMAT
plan is totally based on the volumetric data of the planning CT, which represents only
the ‘frozen status’ at CT scanning and differs from the actual treatment position. These
differences which arise from various sources, such as patient positioning and motion,
could not be detected or overcome by body surface markers or external fixation de-
vices. The deviation also varies between fractions and during treatment. For IMRT/​
VMAT planning in which steep dose gradients are required to spare adjacent critical
structures, these daily setup variations imply a theoretically higher risk of underdosing
tumour and overdosing adjacent normal structures.
To account for these geometrical uncertainties safety margins are applied.
Geometrical uncertainties may be reduced by implementation of IGRT. An efficient
and successful IGRT implementation refines the delivery of radiotherapy by using
imaging techniques to visualize and localize target volumes, in order to allow proper
patient repositioning for the purpose of ensuring accurate treatment and minimizing
the volume of normal tissue being irradiated.
Clinically, IGRT is often referred as the ability (i) to quantify the variation in pos-
ition of anatomical target between the planned and initial setup treatment images and
(ii) to correct any patient misalignments by changing the relative geometry of the
treatment machine (couch position) before the treatment is delivered. There is a tech-
nical component that involves the acquisition and registration of images and a profes-
sional component that involves imaging review and decision-​making on the extent of
patient repositioning required.
In the UK, it is recommended that each department should consider training impli-
cations of its own IGRT techniques and develop training packages and clinical guide-
lines accordingly for the initial and ongoing implementation and documentation of
IGRT(12). In particular, considerations should be given in establishing a threshold of
couch positioning changes that requires the radiation oncologist’s involvement before
the treatment is delivered to verify the patient/​tumour positioning and assess whether
any couch adjustments are warranted.

3.4.2  Implications of IGRT in treatment planning

Traditionally the verification of patient position has been based on bony anatomy
seen in planar imaging and there has been little interaction with treatment planning
during the course of the patient’s treatment. However with the introduction of IGRT
 Image-guided radiotherapy 45

using 3D cone-​beam CT (CBCT), there is much more information about the loca-
tion, size, and shape of the tumour and normal tissues, and this leads to a greater
interaction with treatment planning to refine the original treatment plan to ensure
that the patient receives the radiation dose intended whilst minimizing damage to
normal tissues; this is sometimes termed adaptive planning, but may be better de-
scribed as reactive planning. Fig. 3.12 describes the traditional and IGRT-​based
radiotherapy processes.
There are a number of anatomical changes that can be observed from IGRT; they
include changes in patient weight, internal organ motion, and systematic anatom-
ical changes due to medical interventions. Where systematic changes to the patient
anatomy compared to the treatment planning stage have occurred it may be neces-
sary to modify the treatment plan to adequately treat the patient. This technique is
often referred as adaptive radiotherapy (ART). Care needs to be taken when con-
sidering changes to the treatment plan when the anatomical changes to the patient
observed by IGRT are random or transient in nature, e.g. daily changes in the amount
of gas in the rectum. ART utilizing CBCT approaches are most applicable for direct
replanning as they produce CT-​like images with the patient in the treatment pos-
ition. It is likely that the introduction of cone-​beam-​based ART will lead to some-
where in the region of 20% of patients requiring replanning during the course of their
radiotherapy(13).

(a)

CT scan Treatment plan Treatment

(b) Treatment verification

CT scan

Treatment Treatment
plan

Fig. 3.12  (a) The traditional model of radiotherapy where the treatment verification
process is applied at the treatment phase. (b) The IGRT model of radiotherapy where
the treatment verification process feeds back in to treatment planning and may result
in multiple treatment plans for an individual patient during their fractionated course of
treatment.
46 Treatment delivery, intensity-modulated radiotherapy

3.4.3  Plan of the day adaptive strategies

ART has the potential to anticipate anatomical changes from the CBCT from each
treatment. Each radiotherapy treatment can be adapted based on pre-​fraction CBCT
during the whole course of treatment. Due to the internal organ motion and anatomy
deformations on a day-​to-​day basis, the commonest ART approach in external beam
radiotherapy is the plan library based plan-​of-​the-​day strategy for sites such as bladder
and gynaecological cancers. At the treatment planning stage, an individualized plan li-
brary with multiple treatment plans can be generated corresponding to the size, shape,
and location of tumours from a series of planning CT scans. For each fraction, the plan
best fitting to the anatomy as visualized on the per fraction CBCT is selected in order
to accommodate the interfraction anatomical changes. The clinical applications of
plan-​of-​the-​day ART strategy will be discussed further in other chapters of the book.

3.4.4  Replanning strategies using IGRT

The simplest strategy for the replanning of patients is to rescan and restart the treat-
ment planning process if IGRT images highlight significant patient changes compared
to the planning scan. This approach results in a repeat of the treatment planning pro-
cess outlined in previous sections. One difficulty often encountered when replanning
patients part way through their course of treatment is how to evaluate the total dose
to the tumour and OARs from the separate plans with differing anatomy. The ability
to accumulate dose for the same patient from multiple patient anatomical models and
treatment plans is an active area of development for manufacturers of treatment plan-
ning systems and is an essential requirement for the individualization of radiotherapy
treatment with adaptive/​reactive radiotherapy.
IGRT using CBCT provides the potential to use the IGRT information directly in
the treatment planning process, particularly for simple changes in the patient anatomy,
such as weight loss or weight gain, during treatment. However, IGRT usually involves
the use of low-​dose imaging with small or medium fields of view to limit the radi-
ation doses given to the patient from the imaging procedure. Such techniques present
challenges when using the CBCT images in the replanning process as outlined earlier.
CBCT-​based IGRT systems can be calibrated to produce appropriate CT numbers
under certain conditions. However, the use of CBCT geometry leads to inherent inac-
curacies in CT number to electron or physical density conversions due to changes in
scatter conditions from patient to patient, and the use of half rotation scans to reduce
patient imaging doses. These inaccuracies can be significant for CBCT images of the
thorax and abdomen regions.

Replanning strategies
Patient ­example 1—​patient weight loss
A common example of a patient change seen by IGRT is weight loss during the course
of radiotherapy treatment. Fig. 3.13 shows an example of a head and neck patient
losing a significant amount of weight.
If the patient has lost weight and verification of patient set-​up is based on bony
anatomy, the change to the patient can be approximated by altering the external patient
 Image-guided radiotherapy 47

Fig. 3.13  An example of a head and neck patient experiencing significant weight loss
during the course of radiotherapy treatment.

contour on the original planning CT(14). The assumption is that the bony anatomy is ri-
gidly fixed in place and that changes in the proportion of muscle and fat do not greatly
affect the accuracy of the dose calculation. The process involves changing the external
patient outline on the planning scan via image fusion with the CBCT images and the
relative electron or physical density outside the new external patient contour being
set to that of air. A  heterogeneous dose calculation can be performed on the plan-
ning scan with the altered external contour. This assumes that the relative position of
bony anatomy and air spaces do not change significantly during weight loss. Fig. 3.14
highlights this method of replanning for the patient example. The advantage of this
approach is that the heterogeneity correction will be reasonably accurate as it uses the
original CT scan. It is also relatively quick to recalculate and evaluate the dose distri-
bution, particularly the appearance of significant hotspots and any changes to the dose
delivered to visible OARs such as the spinal cord in this example. The disadvantage is
that it is difficult to determine the dosimetric effect on soft tissues, such as the parotid
glands in this example, as they may not be clearly demonstrated on the IGRT scan.

Fig. 3.14  The original dose calculation on the CT scan and the dose calculation on the
CT scan but based on the external contour from the IGRT scan for fraction 26.
48 Treatment delivery, intensity-modulated radiotherapy

(a) (b)

Fig. 3.15  The Hounsfield numbers between the two points marked on the CT slice for
(a) the conventional CT scan and (b) the IGRT cone-​beam scan.

Patient e­ xample 2—​patient anatomy change in a region of significant

density heterogeneity
To dosimetrically evaluate the effect of the anatomical changes seen from IGRT, the
Hounsfield numbers of the CBCT images need to be accurate enough for direct dose
calculations to be made. Fig. 3.15 shows a line scan between two points in the con-
ventional CT and IGRT images for the same patient. The Hounsfield numbers are all
increased by 1000 within the treatment planning system, but it is clear from the figure
that the IGRT images are not calibrated for use in treatment planning. Additionally,
the use of half-​rotation scans to reduce patient doses from IGRT leads to a gradient
across the images and therefore different Hounsfield values for the left and right lung
images in the figure.
A simple strategy for the use of CBCT images for replanning is to outline different
tissue types, such as lung and soft tissue, prior to applying a bulk density override using
the average density from the original treatment plan for the same tissue types. This ap-
proach ensures an accurate dosimetric evaluation, but requires manual outlining of
regions of interest prior to applying the density override which can become a time-​
consuming step in the process. Alternatively the CT-​to-​density look-​up table within
the treatment planning system can be altered to produce a step-​wise table as shown
in Fig. 3.16. The table blocks ranges of Hounsfield numbers in the CBCT images into
average lung, soft tissue, and bone density regions. This approach has the advantage
of not requiring manual contouring of the images prior to bulk density overrides, and
works well if the average densities of the tissues match those within the look-​up table.
However, average lung tissue densities can vary significantly from patient to patient
 Image-guided radiotherapy 49

Fig. 3.16  A step-​wise CT-​to-​density look-​up table providing an automatic bulk density

override for lung, soft tissue, and bone regions.

with average physical densities in the range 0.1–​0.3 g/​cm3. Systematic errors will be
introduced when calculating dose in lung regions if the average density is different to
that provided by the look-​up table.
Fig. 3.17 shows the change in the dose distribution from changes in the patient
anatomy seen from IGRT for an example patient. The dosimetric evaluation has been
performed by outlining soft tissue and lung regions and overriding the density to the
same value as the original planning scan. The figure shows unacceptably high hotspots
of 110% of the prescription dose (5500c Gy in 20 fractions in this case) and therefore
required modification.

(a) (b)

Fig. 3.17  (a) The original dose distribution, and (b) the new dose distribution due to the
patient anatomical change observed with IGRT and a bulk density override approach.
The new distribution shows unacceptable hotspots of 110% of the prescription dose.
50 Treatment delivery, intensity-modulated radiotherapy

Fig. 3.18  An example IGRT image using a small field of view and low-​dose protocol to
reduce the radiation dose received by the patient. The central axis of each treatment
beam is also shown.

3.4.5  Concomitant dose in IGRT

When using ionizing imaging radiation approaches for IGRT it is important to care-
fully consider the concomitant doses given to the patient and to justify any additional
exposure via a risk–​benefit analysis for the patients. The methodology outlined by
Harrison et al.(15) was to consider the imaging dose to OARs and compare them to the
dose received from leakage and scatter during the radiotherapy. For the case of 26 CT
scans during the course of radiotherapy, it was found that the imaging dose was 5–​25%
of the leakage and scatter dose from the radiotherapy. However Perks et al.(16) meas-
ured peripheral doses in an anthropomorphic phantom and found the IGRT dose to
be as high as 50% of the leakage and scatter dose from the radiotherapy. This highlights
the need to perform a risk–​benefit analysis to appropriately justify the use of IGRT and
the additional radiation dose given to the patient from IGRT.
Imaging protocols for IGRT should be designed to limit the radiation dose to the pa-
tient in line with the ‘as low as reasonably practicable’ (ALARP) principle. This results
in imaging protocols using small fields of view and low imaging doses that produce
images of sufficient quality for the purposes of verification of patient set-​up. However,
such imaging protocols often produce IGRT images that are difficult to use for direct
dosimetric evaluation due to incomplete patient outlines, poor soft tissue differenti-
ation, and inaccurate CT numbers. Fig. 3.18 highlights an IGRT image for a sarcoma
of the chest wall where a small field of view and low imaging dose protocol has been
applied. The images are adequate for the purpose of patient set-​up verification, but
cannot be directly used for dosimetric evaluation of anatomical changes to the patient.
 references 51

A solution to the conflicting requirements on the IGRT protocols for patient set-​
up verification and replanning is to establish two-​step imaging protocols. Initially all
patients receive low-​dose IGRT imaging with field of view that are sufficient for the
purposes of patient set-​up verification. If anatomical changes are observed in such im-
ages then higher dose images with a larger field of view can be justified and applied to
enable dosimetric analysis of the patient changes to be evaluated.

3.5 Discussion
The use of advanced radiotherapy techniques such as IMRT, VMAT, and IMAT pro-
vides highly conformal dose distributions with steeper dose gradients than those pro-
duced by simpler forward planned techniques.
The introduction of IGRT provides the ability (i)  to ensure that the tumour re-
ceives the intended radiation dose on every day of treatment (plan-​of-​the-​day adap-
tive strategy) and (ii) to adjust to patient changes during the course of fractionated
radiotherapy (Replanning adaptive strategy). The combination of advanced de-
livery techniques and IGRT approaches has the potential to improve radiotherapy
outcomes.

References
1. Cancer Research UK. (2013). Cancer Statistics for the UK. [online] Available at: http://​
www.cancerresearchuk.org/​health-​professional/​cancer-​statistics [Accessed 22 Jun. 2016].
2. Department of Health. (2012). Radiotherapy Services in England 2012—​Publications—​
GOV.UK. [online] Available at: https://​www.gov.uk/​government/​publications/​
radiotherapy-​services-​in-​england-​2012 [Accessed 22 Jun. 2016].
3. International Commission on Radiation Units and Measurements. Prescribing, recording
and reporting photon beam therapy. Report 50. Bethesda, MD: ICRU, 1993.
4. International Commission on Radiation Units and Measurements. Prescribing, recording
and reporting photon beam therapy. Report 62 (Supplement to ICRU Report 50). Bethesda,
MD: ICRU, 1999.
5. Van Herk M, Remeijer P, Rasch C, Lebesque JV. The probability of correct target
dosage: dose-​population histograms for deriving treatment margins in radiotherapy.
International Journal of Radiation Oncology, Biology, Physics 2000; 47: 1121–​35.
6. McKenzie A, van Herk M, Mijnheer B. Margins for geometric uncertainty around organs.
at risk in radiotherapy. Radiotherapy and Oncology 2002; 62: 299–​307.
7. Drzymala RE, Mohan R, Brewster L, Shu J, Goitein M, Harms W, et al. Dose-​volume
histograms. International Journal of Radiation Oncology, Biology, Physics 1991; 21: 71–​8.
8. Kutcher GJ. Quantitative plan evaluation: TPC/​NTCP models. Frontiers of Radiation
Therapy and Oncology 1996; 29: 67–​80.
9. Niemierko A. Reporting and analyzing dose distributions: a concept of equivalent uniform
dose. Medical Physics 1997; 24: 103–​10.
10. Quantitative Analyses of Normal Tissue Effects in the Clinic (QUANTEC). International
Journal of Radiation Oncology, Biology, Physics 2010; 76(Suppl 3): S1–​S160.
11. International Commission on Radiation Units and Measurements. Prescribing, recording,
and reporting intensity-​modulated photon-​beam therapy (IMRT). Report 83. Bethesda,
MD: ICRU, 2010.
52 Treatment delivery, intensity-modulated radiotherapy

12. National Radiotherapy Implementation Group Stereotactic body radiotherapy: guidance for

clinicians, providers and commissioners. London, UK: NRIG; 2010.
13. Tanyi JA, Fuss MH. Volumetric image-​guidance: Does routine usage prompt adaptive
replanning? An institutional review. Acta Oncologica 2008; 47: 1444–​53.
14. Van Zijtveld M, Dirkx M, Heijman B. Correction of conebeam CT values using a planning
CT for derivation of the ‘dose of the day’. Radiotherapy and Oncology 2007; 85: 195–​200.
15. Harrison RM, Wilkinson M, Shemilt A, Rawlings DJ, Moore M, Lecomber AR. Organ
doses from prostate radiotherapy and associated concomitant exposures. British Journal of
Radiology 2006; 79: 487–​96.
16. Perks JR, Lehmann J, Chen AM, Yang CC, Stern RL, Purdy JA. Comparison of peripheral
dose from image-​guided radiation therapy (IGRT) using kV cone beam CT to intensity-​
modulated radiation therapy (IMRT). Radiotherapy and Oncology 2008; 89(3): 304–​10.
 Chapter 4

Proton therapy
Ranald MacKay and Adam Aitkenhead

4.1 Introduction
It has long been recognized that proton therapy has a potential advantage over radio-
therapy delivered with photons or electrons. The first paper describing the potential
therapeutic application of proton therapy was published by Wilson(1). The paper notes
that the technology developed to accelerate protons to a high energy will produce pro-
tons that can penetrate deeply enough into tissue to treat a range of conditions, and
shows that the shape of the depth dose curve for monoenergetic protons is well suited
to deliver localized regions within the body.
Proton therapy was first implemented in Berkley, USA and Uppsala, Sweden in the
1950s, but gained momentum in 1960s when the Harvard cyclotron in Cambridge,
USA began a programme of proton therapy. The first hospital-​based facility was a
low energy (62 MeV) cyclotron opened in Clatterbridge, UK in 1989 for the specialist
treatment of ocular tumours which require a proton range of only a few millimetres.
This was followed in 1990 by the first high energy (250 MeV), multiple treatment room
centre at Loma Linda, USA.
In general, a proton centre needs several key pieces of technology. These may include:
◆ An accelerator to produce the beam of high-​energy protons.
◆ A beam line to transport the beam to the treatment room.
◆ Gantries to rotate the treatment beam around the patient.
◆ A nozzle to deliver the beam to the patient.
◆ A patient-​positioning system, such as a couch for the patient to lie on.
More detailed technology descriptions can be found elsewhere(2) but a brief overview
of the key components is included here.
An accelerator is essential for any proton therapy centre. It is the device that ac-
celerates protons to the energy required to be therapeutically effective, i.e. to pene-
trate to the required depth of the tumour. For applications other than ocular proton
therapy, that energy is generally ~230–​250 MeV which corresponds to a depth of
approximately 30 cm.
The proton accelerator and the treatment nozzle are connected by a beam line.
Fundamentally this beam line uses magnets to focus and steer the beam to the patient.
Its complexity depends on several factors such as the number of rooms and the accel-
erator used. A multiple room centre requires a much longer beam line than a single
room centre and many more components are used.
54 Proton therapy

In the treatment room the proton beam will either be delivered using a fixed beam
or a gantry that rotates the beam around the patient. Gantries for proton therapy are
considerably bigger than those required by photon radiotherapy. Most gantries are
isocentric, rotating around the patient who is normally in a fixed position. It can be
shown that a normally incident beam that can rotate 180° about the patient can deliver
a full 4π coverage of irradiation angles if used in combination with a patient table that
rotates by 360° in the horizontal plane. However, many gantries allow a full 360° of
gantry rotation which offers maximum flexibility and reduces the need to move pa-
tients between beams.
The cost of gantries and associated building costs mean that many centres use a
combination of gantries and fixed beams in different rooms. Fixed beams are widely
used for simpler treatments, such as for irradiation of the prostate that can be de-
livered using two lateral beams. However, fixed beams have drawbacks for complex
treatments that require careful, and at times impractical, procedures for positioning
the patient such that the fixed beam angle can deliver an effective treatment plan.

4.2 Proton interactions
For proton therapy to be effective, the proton beam must be directed at the tumour
and the correct dose deposited in the defined target. Fundamental to achieving this
is the deposition of dose by protons traversing a medium. The proton experiences a
number of interactions with both the electrons and nuclei of the atoms in the medium
through which it is traversing. Any one proton will experience a series of interactions
that may change its energy and direction. The sum of these interactions for all protons
determines the shape of the dose distribution from the proton beam. Schematic dia-
grams of these interactions are shown in Fig. 4.1.
The most common interaction is Coulomb scattering which is caused by the
Coulomb force on the proton as it passes close to atomic electrons in the medium. The
proton is a positively charged particle and in traversing patient tissue it passes close

Coulomb interactions with atomic electrons Coulomb interactions with atomic nuclei

Non-elastic nuclear interactions Bremsstrahlung

Fig. 4.1  Proton interactions within the medium, showing examples of a proton incident
on a carbon atom.
 The proton advantage 55

to many negatively charged electrons. There is a natural Coulomb attraction between

the positively and negatively charged particles that depends on the distance between
them. The heavier proton travelling at speed effectively pulls the electron from the
atom, which is ionized. The proton loses some of its energy to the electron in this pro-
cess and the energized electron then produces a series of ionizations over a short range
in a similar fashion to the electrons produced by photon interactions (Chapter 2).
Typically, a proton will experience many Coulomb interactions with electrons that
each reduce the energy of the proton by a small amount, progressively slowing down
the proton until it loses all of its energy and comes to rest. Since the proton is more
than 1800 times heavier than the electron, the Coulomb interactions with the atomic
electrons hardly alter the path of the proton.
The proton also undergoes Coulomb interactions with the atomic nuclei. Unlike
electrons, the nucleus is positively charged and heavier than the proton. When the
proton passes close to a nucleus, it is repelled and deflected from its original path. This
interaction slightly reduces the energy of the proton but more significantly, it deflects
the proton, altering its direction of travel. These Coulomb interactions with the nuclei
are responsible for the lateral spread of a proton beam, and thus the increase in the
beam width, with depth.
Coulomb interactions with the atomic nuclei also produce Bremsstrahlung radi-
ation as the proton is deflected. However, the mass of the proton means that the energy
loss by Bremsstrahlung is insignificant in comparison to the energy loss by electron
interactions and has no clinical consequences.
As well as Coulomb interactions, protons undergo elastic and non-​elastic collisions
with the atomic nucleus. Elastic collisions leave the nucleus intact but reduce the en-
ergy of the incident proton and deflect it strongly. In non-​elastic collisions, the nucleus
is broken apart and the incoming proton cannot be distinguished from the other sec-
ondary particles produced in the collision. Non-​elastic nuclear interactions are rela-
tively rare but are nonetheless important as they account for the neutron field scattered
far outside both the clinical target and the patient.
Table 4.1 provides a useful overview of the proton interactions and their dosimetric
manifestation.

4.3  The proton advantage

Inherent to the perceived advantage of proton beams for external beam radiotherapy
is the particular way in which the dose distribution of the proton beam varies with
depth. When the proton enters a medium it has a high initial energy, and the energy
steadily falls as the proton penetrates into the medium due to the interactions de-
scribed in section 4.2.
The energy transferred per unit distance as the proton traverses the medium is
known as the linear energy transfer (LET). The LET of a proton depends on its en-
ergy, with low-​energy protons having a higher LET than high-​energy protons. As the
proton penetrates into the medium, it continually loses energy, and its LET gradually
increases with depth until the proton nears the end of its range and its remaining en-
ergy is low. Its LET then rapidly increases, and the proton loses its remaining energy in
56 Proton therapy

Table 4.1  Summary of proton interaction types, targets, ejectiles, influence on projectile,

and selected dosimetric manifestations
Interaction type Interaction Principal Influence on Dosimetric
target ejectiles projectile manifestation
Inelastic Atomic Primary proton, Quasi-​continuous Energy loss
Coulomb electrons ionization energy loss determines range in
scattering electrons patient

Elastic Coulomb Atomic Primary proton, Change in Determines lateral

scattering nucleus recoil nucleus trajectory penumbral sharpness
Non-​elastic Atomic Secondary Removal of Primary fluence,
nuclear reactions nucleus protons and primary proton generation of stray
neutrons, and from beam neutrons, generation
gamma rays of prompt gammas for
in vivo interrogation
Bremasstrahlung Atomic Primary proton, Energy loss, chage Negligible
nuleus Bremsstrahlung in trajectory
photon

Reproduced from Newhauser WD, Zhang R, 'The physics of proton therapy', Physics in Medicine and
Biology, 2015; 60(8): R155-​R209. doi:10.1088/​0031-​9155/​60/​8/​R155) under CC BY 3.0

a short distance. This rapid deposition of energy causes a spike in the dose distribution
called the Bragg peak at the end of the proton range. The resulting depth dose curve is
shown in Fig. 4.2. By altering the energy of the incident proton beam, the depth of the
Bragg peak can be controlled. Beyond the Bragg peak, a low level of dose is deposited
due to the secondary products of nuclear interactions, but this is <1% of the dose at
the Bragg peak.

100

80 Fig. 4.2  Depth dose

curves in water showing
(black solid line)
Dose/%

60
measurements of a 6
MV photon beam from
40
an Elekta Agility linear
accelerator; (orange
20 dashed line) Monte Carlo
simulation of a 130
0 MeV proton beam; (blue
0 5 10 15 20 dotted line) Monte Carlo
Depth/cm
simulation of a spot-​
Photons: 6 MV scanned uniform dose
Protons: 130 MeV field composed of 115–​
Protons: 115–148 MeV 147 MeV spots.
 Treatment planning 57

Fig. 4.2 also shows the depth dose curve of a photon beam, which shows marked
skin sparing close to the surface due to lack of electronic equilibrium. The depth of the
maximum dose for the photon beam is shallow, ~1.5 cm for a 6 MV beam, and beyond
this, the delivered dose falls off slowly (approximately 4% per cm) due to the inverse
square law and attenuation of the photon beam.
It is the shape of the Bragg peak and the ability to control its position that leads to
the proton ‘advantage’. In contrast to photon radiotherapy where the dose from a single
beam reaches a maximum 1–​2 cm from the surface and falls off steadily through the
patient, the sharp spike of maximum dose at depth that occurs with proton beams and
the negligible dose delivered after this peak offer great potential for optimizing the
dose to the tumour whilst sparing normal tissue.
In modern radiotherapy, single field photon deliveries are rarely used for treat-
ment and photon techniques such as stereotactic ablative radiotherapy (SABR),
intensity-​modulated radiotherapy (IMRT), or volumetric-​ modulated arc radio-
therapy (VMAT) are effective at targeting disease and positioning sharp dose
gradients between disease and normal tissue. In many cases the skin sparing and
penumbra of the photon beam will be better than that of the proton beam. The
proton still has to prove its advantage in comparison to these advanced photon
radiotherapy techniques. However, the shape of the proton depth dose curve pro-
vides it with an inherent advantage, enabling the same dose to be delivered to the
target while delivering less dose to normal tissue.

4.4 Treatment planning
Treatment planning for proton therapy typically follows a similar scan-​plan-​treat pro-
cess that is used for external beam photon radiotherapy.
The first step in the process is to determine a suitable method of setting up and im-
mobilizing the patient for treatment. Immobilization methods are broadly similar to
those used for photon radiotherapy, and a range of immobilization devices is commer-
cially available (e.g. thermoplastic moulds, vacuum bags).
Patients are then imaged in the treatment position on a computed tomography
(CT) scanner using kilovoltage X-​rays, and outlining of the target volume(s) and
relevant organs-​at-​risk (OARs) is performed on this image. While other imaging
modalities (e.g. magnetic resonance imaging (MRI), positron emission tomog-
raphy (PET)) may be used to aid identification of the target volume, the treat-
ment planning process and dose calculations are performed using the X-​ray CT
image alone.
The computation of dose within the patient is crucial to any effective radiotherapy
treatment. In both photon and proton radiotherapy, the X-​ray CT image is used as
the basis for treatment planning dose calculations. Each voxel in this image contains
a numerical value (in Hounsfield units). In photon radiotherapy, the dose calcula-
tion requires that the relative electron density in each voxel is known, and this is
easily obtained from the Hounsfield unit value. In proton therapy, the dose calcula-
tion instead requires that the proton stopping power is known, but the relationship
between proton stopping power and Hounsfield unit is not simple. Proton therapy
58 Proton therapy

(a) Passive Scattering system

Compensator Collimator
Energy Lateral
spreading spreading

(b) Pencil Beam Scanning (PBS) system

Energy Steering
selection magnets
Fig. 4.3  Illustrations of
(a) passive scattering and
(b) pencil beam scanning
delivery systems.

uses conversion schemes such as the stoichiometric CT calibration method, where CT

scans of materials of known elemental composition and density are used to generate a
conversion table for HU to proton stopping power(4). The uncertainty in this conver-
sion presents a source of uncertainty in the range of proton beams(5).
Recently researchers have looked to the potential advantage of dual energy CT to
provide a more accurate estimate of proton stopping power. Many theoretical studies
have sought to quantify the potential advantage of dual energy CT. However, currently
available dual energy CT scanners have not been designed specifically with proton
therapy in mind and many of the theoretical gains have not been realized yet in clin-
ical practice(6). Despite this, in looking for the best CT scanner for proton therapy, dual
energy capabilities are an important consideration.
Following the calculation of stopping powers from the CT image, proton dose can
be computed by treatment planning systems (TPSs). As described earlier, dose is de-
posited by primary protons undergoing elastic Coulomb interactions and secondary
particles produced by non-​elastic nuclear interactions. Many TPSs use pencil beam
models to calculate proton dose. Pencil beam algorithms convolve infinitesimal proton
beamlets with fluence at the position of the beamlet.
The TPS is used to determine how the prescribed dose to the target volume may
be achieved using the delivery system. The process of creating a treatment plan
has several steps, some of which are performed by the planner and others which
are performed by the planning software. For those steps that are performed by the
planner, departmental protocols may be used to give guidance to standardize plan
quality.
Proton therapy treatments can be delivered using either scattering or scanning tech-
nology and Fig. 4.3 shows a schematic diagram of both delivery systems. Scattering
and scanning are distinct delivery methods, the choice of which affects the dose distri-
bution that can be delivered. As such they often require different methods of treatment
planning and are considered here separately.
 Treatment delivery: scattered beams 59

4.5  Treatment delivery: scattered beams

Scattered beams were the first to be used therapeutically, and use a hardware system
in the delivery nozzle to produce a broad beam of protons from the narrowly focused
beam provided by the accelerator. This beam encompasses the whole of the target and
is analogous to the beams delivered in external beam photon radiotherapy where a
broad beam is also employed.
In a single scattering system, the beam is spread laterally using a high-​density scat-
terer, giving a beam with a reasonably flat central portion which can be used to treat
the tumour. In a single scattering system, the field size is limited by the region of the
beam that is ‘near flat’. In a double scattering system, a second scatterer is used to
flatten the single scattered profile, producing a wider beam for treatment and allowing
the use of larger field sizes.
The broad beam that results from a scattering system must be shaped laterally to
conform the dose to the target. This is done using a collimator. In the majority of cases,
the field shape for a passive scattering system is defined by a collimator, which is often
fabricated from brass and is created for each individual field. These collimators are cre-
ated using precision milling machines and a high degree of accuracy can be achieved
in their manufacture.
It is the properties of the collimator and the physics behind proton scattering that
determine the lateral fall-​off, or penumbra, of the proton treatment beam. The air
gap to the patient and the energy range of the beam all affect the penumbra of the
treatment field in the patient. When treating superficial targets a collimated, scattered
beam can achieve sharper fall-​off than that of scanned beam proton therapy.
As well as the lateral fall-​off of the beam, the energy and shape of the distal fall-​off
must be carefully chosen to match the target region. The maximum beam energy re-
quired for treatment is determined by the maximum depth of the target, while the
spread of energies is determined by the depth range required to cover the target.
Energy spreading is typically done using a range modulator, and the distal edge of
the beam is shaped to the target contour using a compensator. A feature of scattering
systems is that the longitudinal length of the spread-​out Bragg peak is the same at all
points across the beam’s-​eye-​view, with the result that the dose to normal tissues prox-
imal to the target may be high. This can be seen in Fig. 4.3a where the high dose does
not conform to the spherical target at the proximal side, whereas it conforms well to
the target at the distal side.
There is no doubt that passive scattering in combination with collimator and com-
pensator can produce high-​precision proton beams. However, the requirement for
custom devices for each field has disadvantages in terms of the flexibility required for
the treatment of a patient over several treatment fractions. One such issue affects the
design of the treatment compensator. If designed only for coverage of the target geom-
etry on the planning CT scan, set-​up uncertainties will inevitably lead to underdosage
of the target over the full course of treatment. To prevent this, the compensator must
be adjusted by smearing to reduce the effect of sharp lateral gradients. While this
smearing is necessary to ensure target coverage, it reduces the conformity of the treat-
ment beam to the planned target.
60 Proton therapy

(a) Single field, with organ-at-risk (b) Multiple patched fields, with
distal to the target. organ-at-risk lateral to each field.

Match-lines

Target
Organ-at-risk

Fig. 4.4  Schematic representation of a target encompassing a normal tissue and possible

proton beams for treatment. (a) A single proton beam is shown covering the target but
directed at the normal tissue. This apparently simple arrangement is not often used due
to the danger of overshoot. The more complex arrangement shown in (b) avoids aiming
a beam at the normal tissue but the volume must be split and patched fields used.

Another complexity of passive scattering treatment planning occurs when the target
closely wraps round a sensitive normal tissue, such as in the schematic case in Fig. 4.4.
In theory, the properties of the Bragg peak would allow the target volume to be effect-
ively treated by a single beam stopping prior to the sensitive normal tissue (Fig. 4.4a).
In practice, the uncertainty in range associated with protons requires a more cautious
solution as overshoot could result in the sensitive normal tissue receiving an unaccept-
able dose. In such cases, treating the target using a single field from another direc-
tion may not be possible without irradiating parts of the OAR to the target dose. This
problem can be solved by splitting the target into sub-​targets that can be covered by
a combination of fields, as in Fig. 4.4b, where individual fields are ‘patched’ together
to ensure that the target volume is fully covered. However, the sharp gradients at the
edge of the collimated patched fields could potentially lead to large under and over-
doses given the normal setup errors of radiotherapy. To mitigate this, the positions of
the patches can be varied on different fractions, effectively moving the locations of the
patch junctions. This requires that multiple plans must be created, which inevitably
increases the cost and complexity of treatment.
The use of field-​specific collimators and compensators are also an issue when treat-
ments require adaption due to anatomical changes during the course of fractionated
treatment. Producing an adapted plan during a course of treatment could require com-
pletely new collimators and compensators which are costly and time consuming to
produce. In practice, for most cases this is dealt with by modification of the original
beam specific devices.
Scattering systems were the first to be employed in clinical proton therapy and have
so far been used to treat the majority of patients with proton therapy. Although they
are an effective method of targeting the tumour, they have several disadvantages com-
pared to scanned beams, as will be described next.

4.6  Treatment delivery: scanned beams

In a scanned proton beam instead of producing a broad beam to encompass the
target, the narrow proton beam from the accelerator is deflected using a magnetic
 Treatment delivery: scanned beams 61

field to place spots at different positions in the target where dose is deposited. The
monoenergetic proton beam is thus ‘scanned’ over the extent of the target at a par-
ticular depth (defined by the energy). The beam energy is then altered to position the
Bragg peak at a different depth, and the next layer is scanned. Scanned beams were
first implemented clinically at Paul Scherrer Institute in Switzerland and were slow
to be developed as part of commercial proton systems. However, the majority of new
proton therapy systems are now scanned beam systems and proton beam scanning is
the future of proton therapy.
There are several advantages of scanned beam systems:
◆ In comparison to passive scattered proton therapy, scanned systems can effectively
paint the tumour with the Bragg peak and achieve better conformality of the high
dose region to the target. This is particularly true in the region proximal to the
target where passive scattering cannot shape the high dose region.
◆ Scanned beams reduce the neutron scatter associated with passive scattering
systems.
◆ Scanning systems remove the need for the collimators and compensators associated
with passive scattering. This reduces the amount of patient specific hardware that is
both time consuming and expensive to produce, presents a manual handling issue
in the treatment room and can slow throughput since individual collimators and
compensators are required for each field.
◆ Scanned beams are often simpler to plan, in particular for complex volumes where
passive scattering requires patched fields.
The disadvantages of scanned beam therapy are related to the safety of the more com-
plex technology. Scanning proton therapy requires more faith in (or quality assurance
of) the delivery technology than the mechanically simpler passive scattering systems.
In particular, the treatment of moving targets with scanning is the subject of much on-​
going research and clinical implementation is still in its infancy.
Most spot scanning plans consist of 2–​5 fields. Beam angles are chosen to avoid
heterogeneities (e.g. bones and air cavities) where possible, particularly where reliable
setup may be difficult, or to avoid OARs either in the entrance path or immediately
distal to the target. The selection of beam angles is performed by the planner. Due to
the greater degree of freedom in proton planning, it is typical to use fewer fields than
would be used for a comparable photon IMRT plan.
After the beam angles have been defined, the planning software performs three key
tasks: spot placement, dose calculation, and spot weight optimization.
Spot placement for each field is done by calculating the beam energies required to
cover the target in the longitudinal (i.e. depth) direction, and then within each energy
layer the beams-​eye-​view of the target is covered by a hexagonal or rectilinear grid of
spots. Spot spacing is dependent on the lateral width of an individual spot and is typ-
ically 1.5–​2.0 sigma.
The dose calculation for each spot is typically performed using a pencil beam algo-
rithm, where the longitudinal dose profile is calculated analytically using a method
such as that described by Bortfeld(7). To a first approximation, the lateral dose profile
of each proton spot is Gaussian, and so the longitudinal dose profile can be spread lat-
erally using a Gaussian to model the spot in three dimensions. However, to calculate
62 Proton therapy

(a) (b)
–15 1 1
–10 0.8 0.8
–5
0.6 0.6

Dose
Dose
x/cm

0
0.4 0.4
5
10 0.2 0.2

15 0 0
0 10 20 30 40 –15 –10 –5 0 5 10 15
z/cm x/cm
(c) (d)
–15 0 0
MC data
–10 –1 –1 Gaussian A
Gaussian B
–5 –2 –2 Gaussians A+B
log (Dose)

log (Dose)
x/cm

0 –3 –3

5 –4 –4

10 –5 –5

15 –6 –6
0 10 20 30 40 –15 –10 –5 0 5 10 15
z/cm x/cm

Fig. 4.5  Monte-​Carlo simulation of a proton beam (with initial energy = 230 MeV,

initial spot sigma = 2.5 mm) in water: (a) 2D beam profile on a normal scale. (b) Lateral
profile through the Bragg peak. (c) 2D beam profile of the same simulation shown on
a logarithmic scale, in order to show the low-​dose halo. (d) Lateral profile through the
Bragg peak, plotted on a logarithmic scale. The black solid line represents the Monte-​
Carlo data; the blue dashed line represent a narrow Gaussian fit to the central peak of
the lateral profile; the blue dotted line represents a wide, lower level Gaussian fit to the
halo; and the orange dashed line demonstrates that the sum of two (or more) Gaussians
can be used to approximate the lateral profile of a physical beam.

dose to the required degree of accuracy requires that the wider ‘nuclear halo’ which
results from non-​elastic nuclear interactions is also taken into account. Neglecting this
effect can result in differences of up to 4% between the calculated and delivered dose.
TPSs commonly do this by spreading the spot profile laterally using a second, wider
Gaussian, as illustrated in Fig. 4.5.
The spot weight optimization calculates the relative number of protons that should
be delivered by each spot to achieve the desired planning aims. To compute the weights
the treatment planning uses an inverse planning approach, where the planner first spe-
cifies dose objectives for targets and OARs. During optimization, the TPS computes a
penalty function which scores how far the current dose distribution is from meeting
each of the objectives. By adjusting the spot weights, the TPS attempts to minimize the
penalty function and find the combination of spot weights which is closest to meeting
the plan objectives.
For multi-​field plans, there are typically many different combinations of spot-​
weights that can be used to achieve comparable dose distributions—​that is, there is
a high degree of degeneracy. There are different planning techniques which aim to
 Treatment delivery: scanned beams 63

control this degeneracy, or to utilize it to improve other characteristics of the plan such
as its robustness to uncertainties.
The two main planning techniques for spot scanning are single-​field optimization
and multi-​field optimization:
◆ Single-​field optimization (SFO): As the name suggests, the TPS optimizes each field
individually, attempting to meet the plan objectives for each field. As a result, the
dose distribution for each field typically provides reasonably uniform coverage of
the target volume.
◆ Multi-​field optimization (MFO): Using this technique, the TPS is permitted to op-
timize all fields simultaneously. This provides greater control of the deposited dose
than is possible with an SFO plan, particularly for complex geometries (e.g. multi-​
dose level targets, or targets lying distal to an OAR). Using this technique, the dose
deposited by each field can be highly modulated, and it is only the combined dose
from all fields which gives uniform coverage of the target. An alternative name for
the MFO technique is intensity-​modulated-​proton-​therapy (IMPT).
Fig. 4.6 shows a demonstration of these two techniques to deliver dose to a water
phantom containing a spherical target of diameter 8 cm and a smaller spherical OAR
within the target region. In the SFO plan, there is good coverage and uniformity of the
target, but sparing of the OAR is limited since each field must pass through the OAR to
deliver dose to the distal part of the target. OAR sparing in the SFO plan can only be im-
proved as a trade-​off against target coverage and uniformity. In the MFO plan, the OAR
sparing is much improved, since only the combined dose from both fields is assessed
against the plan objectives. This allows each individual field to treat a limited part of the
target, avoiding the need for spots to pass through the OAR. As a result, the sparing of
the OAR is much greater in the MFO plan, and each MFO field is highly modulated.
The minimum energy that is required to allow a beam to be transported through
the beam line is typically around 70 MeV, an energy that corresponds to around 4 cm
depth in the patient. In order to deliver dose closer to the surface a range-​shifter posi-
tioned at the exit of the nozzle is used to degrade the beam to provide energies below
70 MeV. The range-​shifter also broadens the beam due to the Coulomb interactions
with nuclei in the range-​shifter. As a result the spot size for very low energy beams is

Fig. 4.6  Comparison of a SFO

and a MFO plan in a 20 × 20
× 20 cm water tank. Each plan
has two fields delivering dose
to a spherical target (red) of
diameter 8 cm. Within the
spherical target is a spherical
OAR (green) of diameter
3 cm, which the plans attempt
to spare.
64 Proton therapy

increased, with the result that the lateral edge of a spot-​scanning field for a superficial
target is less sharp than can be achieved using photon or scattered proton technolo-
gies(8) (both of which employ collimators to define the field edge). Several groups are
currently investigating the use of collimators during delivery of spot scanned fields to
recover a sharp edge for shallow targets(9,10).

4.7 Plan robustness
Treatment planning is performed using a CT image of the patient, and delivery of the
resulting plan is done with the patient set-​up in the same position as in the original
planning CT scan. However, the CT image is only a snapshot of the patient, and over
the course of their treatment (often around 30 fractions), the set-​up or anatomy of the
patient may vary from that of the planning CT scan. This may be systematic, e.g. due to
range uncertainty or uncertainty in the set-​up of the patient, or it may be random such
as from daily changes in the patient position. There may also be anatomical changes
such as result from weight loss, tumour regression, or cavity filling, etc. A robust plan
is one which is tolerant to such changes and continues to meet the plan objectives. For
photon radiotherapy, plans are commonly made robust to anticipated variation (such
as small changes in the patient position that result from daily patient positioning)
by applying margins to the treatment volumes and OARs, and there are well-​known
methods for calculating appropriate margins(11). In proton therapy, the use of similar
margins is currently a common practice. However, proton plans are likely to be much
more sensitive to uncertainties than photon plans(12), and the use of margins alone may
not be sufficient to ensure a plan is robust(13).
Recently, TPSs have provided robust optimization tools which allow the optimizer
to take multiple error scenarios (in terms of range and set-​up) into account during the
optimization process. Robust optimization introduces an inevitable trade-​off between
robustness and plan quality in the nominal (i.e. error-​free) scenario. For example, to
ensure that coverage of the target volume is robust to daily set-​up variation it may be
necessary to deliver some additional dose to normal tissues immediately adjacent to
the target in the nominal scenario. Currently, robust optimization tools in TPSs do
not take fractionation into account, which may mean that plan quality in the nominal
scenario is degraded more than is necessary to achieve a robust plan(14). However, de-
velopment of these tools is ongoing.
In addition to the complex robust optimization features available in TPSs, it should
be remembered that there are several relatively simple things the planner can do to
aid robustness. These include:  the selection of beam angles to avoid placing OARs
proximal and distal to the target; the selection of beam angles to avoid heterogeneities;
careful use of highly modulated fields to avoid very high dose regions close to OARs
(such as those in the MFO example in Fig. 4.6) and; the use of additional fields.

4.8 Adaption
Robust optimization tools in TPSs currently do not attempt to make plans robust to
changes in anatomy. Making a plan robust in advance against potential anatomical
 Clinical case example 65

changes is likely to result in an unacceptable reduction in plan quality. Due to the

potential sensitivity of proton plans to anatomy changes, it is important to monitor
the patient for anatomical changes during treatment. Until recently such moni-
toring was performed with in-​room planar imaging and offline repeat CT scans;
however, the first proton facilities began to implement cone beam CT imaging
in 2016. The use of cone beam images for re-​planning is more difficult in proton
therapy due to the inherent inaccuracies in Hounsfield units making range predic-
tion uncertain.
If changes occur, the impact on the dose distribution should be assessed, and the
plan adapted where necessary using a repeat CT scan.

4.9  Clinical case example

The following clinical case example illustrates some of the differences between pas-
sively scattered proton therapy, spot scanned proton therapy, and contemporary
photon techniques. This case is a 14-​year-​old female with Ewing sarcoma of the right
ilium. The patient was treated on the Euro-​Ewing 99 protocol and received passive
scattered proton therapy. Equivalent treatment plans have been created using spot
scanned proton therapy (both SFO and MFO) as well as 3D-​conformal, IMRT, and
VMAT photon therapy using the same CT scan and target and OAR contours. Two
target volumes (labelled PTV1 and PTV2) were delineated, with the aim of delivering
a higher dose to PTV2.
The clinical passive scattered proton plan consisted of a three-​phase treatment plan,
with the same four beam orientations used in each phase. The use of multiple phases
enabled the dose to PTV2 to be boosted compared to the PTV1 dose, while minim-
izing the dose to anterior OARs such as the bowel. In the proton pencil beam scanning
(PBS) plan and both the photon IMRT and VMAT plans, the ability to modulate the
dose allowed the plan to use only a single phase. The 3D conformal-​radiotherapy (3D-​
CRT) plan was also planned in a single phase, which inevitably limited its capability to
modulate the dose to the target.
The dose distributions achieved for each technique are shown in Fig. 4.7, and illus-
trate some of the key differences between the techniques:
◆ The absence of exit dose in the proton plans provided improved normal tissue and
OAR dose sparing compared to the photon plans. In particular, the proton plans
greatly reduced the volume of normal tissue exposed to low doses, almost entirely
sparing the contralateral ilium and femur in comparison to the photon plans (la-
belled ‘a’ in Fig.  4.7) and reducing the dose to the large bowel, sigmoid, bladder,
rectum, and uterus. The integral dose to normal tissues was therefore substantially
lower in the proton plans than in the photon plans.
◆ PTV coverage in the proton plans was comparable to the photon IMRT and VMAT
plans. Since the target volume was large and irregularly shaped, it was impossible to
achieve adequate coverage using a photon 3D-​CRT technique without comprom-
ising OAR tolerance doses or having high dose hot-​spots in the target volume (‘b’ in
Fig. 4.7).
66 Proton therapy

◆ Conformity of the high dose region around the target volume was improved in the
PBS plans compared to the passive scattered plan. This is a consequence of the im-
proved control of the dose proximal to the target by each PBS field compared to
the equivalent passive scattered field. The increased proximal dose per field in the
passive scattered plan leads to additional high dose regions around the edge of the
target (‘c’ in Fig. 4.7).

c
Passive scattering

Photon VMAT
Proton

a
Proton PBS SFO

Photon IMRT
d b
Proton PBS MFO

Photon 3DCRT
–1000

–500

500

1000

1500

2000

2500

0
10

35

65

85
90
95
105

120

CT number/HU Dose/%
Fig. 4.7  Clinical case example showing transverse dose distributions for treatment
of a Ewing sarcoma of the right ilium, comparing proton and photon techniques.
PTV1 (volume = 2400 cm3) was a uniform 4 mm expansion of CTV1, while PTV2
(volume = 1600 cm3) was a uniform 2 mm expansion of CTV2. The CTV1 (cyan dashed line),
PTV1 (cyan solid line), CTV2 (white dashed line), PTV2 (white solid line), sigmoid (magenta),
large bowel (green) and left ilium (orange) contours are shown. The arrows labelled a, b,
c, d indicate key features which are discussed in the text. The proton DS (double scattered)
plan was the clinical plan used to treat the patient, and is courtesy of Andrew Chang
(Procure Proton Therapy Centre, Oklahoma). The proton PBS plans are representative of the
spot-​scanning system installed in 2018 at the Christie Hospital (UK), while the photon plans
are representative of current photon techniques at the Christie Hospital.
 Radiobiological effect of proton therapy 67

◆ The PBS MFO plan gave improved sparing of proximal OARs such as the sigmoid
and large bowel compared to the PBS SFO plan (‘d’ in Fig. 4.7). This is a consequence
of the ability of the MFO plan to treat limited parts of the target with each field, only
requiring the combined dose to meet the plan objectives, as demonstrated earlier in
Fig. 4.6.
In general, the differences between PBS and passive scattering plans are dependent on
anatomy. In more complex anatomies with non-​mobile targets, PBS techniques may be
able to provide advantageous dose distributions, while passive scattering techniques
may be more suitable for moving targets.

4.10  Radiobiological effect of proton therapy

In radiotherapy dose is used as a surrogate for biological effect. Physicists in particular
are comfortable with the use of dose and are confident in its measurement. The dose,
measured in units of gray (symbol Gy), is typically used in the prescription of radio-
therapy treatments. However, as described in Chapter 2 the relationship between dose
and effect is often not a simple one and, while dose is used clinically, there are many
factors that need to be considered. In proton therapy, one of these factors is the dif-
ference in radiobiological effect of protons compared to the effect of the same dose
delivered by photons in conventional external beam therapy.
The relative biological effectiveness (RBE) is defined as the ratio of the doses required
by two types of radiation to cause the same level of effect. In proton therapy, we are
interested in the RBE of protons compared with the dose delivered by external beam
radiotherapy with photons. The RBE of protons is defined as the ratio of the dose in
photons and protons required to achieve a given biological effect. In proton therapy
practice, an RBE of 1.1 is generally applied which indicates that the physical dose (in
units of Gy) required from protons is less than the dose required from photons(15).
It has long been recognized that the dose delivered by protons and photons is not
equivalent. However, the RBE of the photon is, despite decades of clinical treatment,
the subject of much research. The value of RBE is routinely simplified to a single value
when planning treatments, when in reality there is consensus that the RBE of the
proton is not a single value.
The obvious starting point for any consideration of RBE is to look at why the bio-
logical effect of protons is different and to do this we need to consider the way that
dose translates into cell damage. Any biological effect in radiotherapy is the combin-
ation of the physical interactions of the radiation with matter, the chemical effects that
occur as a result of these interactions, and the consequential damage to cells. This text
is not the place for a detailed examination of the RBE of protons, but in all of the steps
that produce radiation damage the protons act slightly differently from photons. Most
markedly, the physical deposition of energy by the proton is different from that of the
photon. On a microscopic scale, the proton deposits its dose more densely and causes
more irreparable damage to cells.
However, this is a gross approximation of complex biology. In particular, at the end
of the proton range, at the Bragg peak, the LET of the proton increases and as the LET
rises so does the RBE. At this crucial point in the proton dose delivery, the biological
68 Proton therapy

effect of the dose is the most uncertain. This has profound implications for the delivery
of proton therapy. As a consequence, many advocate the incorporation of LET and
RBE into the proton planning process(16).

4.11  Proton range in the patient

One of the key consequences of the shape of the proton depth dose curve and its char-
acteristic Bragg peak is the importance of proton range. The proton advantage is de-
pendent on the high dose delivered in a sharp peak at the end of the range. Beyond this
peak very little dose is delivered. Theoretically the maximum range can be set as the
end of the tumour and sensitive normal tissue beyond the tumour can be completely
spared. In practice, there are several factors that affect the position of the Bragg peak
making precise positioning difficult. In proton therapy, the difficulty in predicting the
position of the Bragg peak is termed ‘range uncertainty’ and it is an important factor
in planning and delivering therapy.
The main sources of proton range uncertainty can be summarized as:
◆ The uncertainty in converting the CT Hounsfield numbers to proton stopping
power, and other issues with CT such as artefacts.
◆ The increase in the RBE at the end of the Bragg peak, which increases the biologic-
ally effective range 1–​2 mm beyond the physical end of the Bragg peak.
◆ Anatomical variations in the patient throughout the treatment course that change
the range of the beam calculated during the treatment planning process.
As with many uncertainties in radiotherapy planning, range uncertainty is generally
dealt with by placing a margin on the treatment beam. For example, Massachusetts
General Hospital add a margin of 3.5% of the proton range + 1 mm to beams(6).
An important consequence of this range uncertainty is despite the capacity for pro-
tons to spare normal tissues beyond the Bragg peak a proton beam is rarely aimed
directly at a sensitive normal tissue leading many to say that the protons are not used
to their full potential.
Reducing range uncertainty is one of the key challenges to improve proton therapy,
and methods of verifying the range in the patient using techniques such as PET or
prompt gamma imaging are in development to address this issue(17).

4.12 Conclusions
The use of proton therapy worldwide is expanding rapidly and consequently there is
considerable investment in proton therapy technology. Recent advances have seen the
development of proton scanning systems in many clinics and improved image guid-
ance technology. Continuing improvements in technology and reductions in the cost
of developing a centre will accelerate this trend.
 References 69

References
1. Wilson RR Radiological use of fast protons Radiology 1946; 47:487–​91.
2. Owen H, MacKay RI, Peach K, Smith S Hadron accelerators for radiotherapy
Contemporary Physics 2014; 55:55–​74.
3. Newhauser WD, Zhang R. The physics of proton therapy. Physics in Medicine and Biology
2015; 60(8):R155–​209. doi:10.1088/​0031-​9155/​60/​8/​R155.
4. Schneider U, Pedroni E, Lomax A. The calibration of CT Hounsfield units for
radiotherapy treatment planning. Physics in Medicine and Biology 1996; 41:111–​24.
5. Paganetti H. Range uncertainties in proton therapy and the role of Monte Carlo simulations.
Physics in Medicine and Biology 2012; 57:R99–​117. doi:10.1088/​0031-​9155/​57/​11/​R99.
6. Landry, G., Seco J, Parodi K, Verhaegen, F. Dual energy CT to reduce range uncertainties
in hadrontherapy. Radiotherapy and Oncology 2014; 110,  S57–​8
7. Bortfeld, T., Schlegel, W. An analytical approximation of depth-​dose distributions for
therapeutic proton beams. Physics in Medicine and Biology 1996; 41:1331–​9.
8. Safai S, Bortfeld T and Engelsman M. Comparison between the lateral penumbra of a
collimated double-​scattered beam and uncollimated scanning beam in proton radiotherapy.
Physics in Medicine and Biology 2008; 53:1729–​50. doi:10.1088/​0031-​9155/​53/​6/​016
9. Charlwood FC, Aitkenhead AH, Mackay RI. A Monte Carlo study on the collimation of
pencil beam scanning proton therapy beams Medical Physics 2016; 43, 1462–​72. doi.org/​
10.1118/​1.4941957
10. Hyer DE, Hill PM, Wang D, Smith BR, Flynn RT. A dynamic collimation system for
penumbra reduction in spot-​scanning proton therapy: proof of concept. Medical Physics
2014; 41:091701. doi: 10.1118/​1.4837155.
11. van Herk M, Remeijer P, Rasch C, Lebesque JV. The probability of correct target
dosage: dose-​population histograms for deriving treatment margins in radiotherapy.
International Journal of Radiation, Oncology, Biology, Physics 2000; 47:1121–​35.
12. Intensity modulated proton therapy and its sensitivity to treatment uncertainties 1: the
potential effects of calculational uncertaintiesPhys. Medical Biology 2008; 53:1027–​42.
doi:10.1088/​0031-​9155/​53/​4/​014
13. Albertini F, Hug EB, Lomax AJ. Is it necessary to plan with safety margins for actively
scanned proton therapy? Physics for Medicine and Biology 2011; 56:4399–​413
14. Lowe M, Albertini F, Aitkenhead A, Lomax AJ, MacKay RI. Incorporating the effect
of fractionation in the evaluation of proton plan robustness to setup errors. Physics for
Medicine and Biology 2016; 61:413–​29. doi: 10.1088/​0031-​9155/​61/​1/​413.
15. Paganetti H, Niemierko A, Ancukiewicz M, Gerweck LE, Goitein M, Loeffler JS. Suit
HD. Relative biological effectiveness (rbe) values for proton beam therapy. International
Journal of Radiation, Oncology, Biology, Physics 2002; 53:407–​21.
16. Grassberger, C, Trofimov A, Lomax A, Paganetti H. Variations in linear energy transfer
within clinical proton therapy fields and the potential for biological treatment planning
International Journal of Radiation, Oncology, Biology, Physics 2011; 80: 1559–​66.
17. Knopf AC, Lomax A. In vivo proton range verification: a review. Physics in Medicine and
Biology, 2013; 58:R131-​60.
 Chapter 5

Breast radiotherapy
Charlotte Coles, Murray Brunt, Anna Kirby,
Sara Lightowlers, and Nicola Twyman

5.1  Evidence based rationale for radiotherapy

5.1.1  Adjuvant
radiotherapy to breast after breast
conserving surgery

Whole breast radiotherapy
Irradiation of the whole breast in women who have undergone breast-​conserving sur-
gery (BCS) has long been a standard of care in the treatment of early breast cancer. This
practice is predominantly based on the local control and survival gains demonstrated
by the Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) meta-​analyses(1).
These are based on data from 10,801 women treated in 17 randomized trials of BCS
plus or minus whole breast irradiation (WBI), and demonstrate that the addition of
WBI to BCS approximately halves the risk of any first recurrence at 10 years (from 35%
to 19%), and reduces the risk of breast cancer death at 15 years by around one-​sixth
(from 25% to 21%). In women with pathologically node-​negative disease (n = 7287),
the risk of local recurrence was reduced from 31% to 16% at 10 years and the risk of
breast cancer death from 21% to 17% at 15 years. In women with node-​positive breast
cancer (n = 1050), radiotherapy reduced the 10-​year risk of local recurrence from 64%
to 43% and the 15-​year risk of breast cancer death from 51% to 43%. Consistent with
previous meta-​analyses, the prevention of four local recurrences at 10 years prevented
one breast cancer death at 15 years. Excision margins are not closely correlated with
local recurrence risk provided no disease is seen at the resection margins. In addition,
adjuvant systemic therapies provide a significant reduction in local recurrence risk
(see Fig. 5.1).
Hypofractionation has been the focus of a number of trials. The UK START-​B study
found no difference in the primary endpoint of local-​regional recurrence at a median
of 9.9 years between the standard regimen of 50 Gy in 25 fractions (5.5% local-​regional
recurrence rate) and the hypofractionated regimen of 40 Gy in 15 fractions (4.3% local-​
regional recurrence rate)(2). Normal tissue effects were less with the hypofractionated
regimen in this study with significant differences noted for breast shrinkage, telangi-
ectasia, and breast oedema. The conclusion from START-​A and START-​B is that
hypofractionation is both safe and effective as long as the total dose is appropriately
 Evidence based rationale for radiotherapy 71

1214 women with BCS and node-positive disease

5-year gain 30.1% (SE 2.8) 15-year gain 7.1% (SE 3.6)
60 60
Logrank 2p = 0.01 55.0%
46.5 BCS
50 50 45.2
Isolated local recurrence (%)

Breast cancer mortality (%)

41.1 BCS 47.9%
40 40 BCS + RT

36.5
30 30 24.3

20 20
BCS + RT 20.9
10 13.1 10
11.0

0 0
0 5 10 15 0 5 10 15
Time (years) Time (years)
Fig. 5.1  EBCTCG graphs.
Reproduced with permission from Clarke, M. et al. Effects of radiotherapy and of differences in the
extent of surgery for early breast cancer on local recurrence and 15-year survival: an overview of the
randomised trials. The Lancet, 366: 2087–106. Copyright © 2006 Elsevier Ltd. All rights reserved.
https://doi.org/10.1016/S0140-6736(05)67887-7.

reduced. The Canadian study used 42.5 Gy in 16 fractions against the standard of 50
Gy in 25 fractions and reported local recurrence rates of 6.2% and 6.7% respectively at
10 years with no significant difference in good or excellent cosmesis rates(3).
It has been suggested that there might be sub-​groups for whom hypofractionation
is less effective but the evidence shows that this is not the case. A meta-​analysis of UK
data from the START trials including the pilot study of 4883 patients, where grade was
known, shows no difference in outcomes based on tumour grade or cancer sub-​type.
A  central histopathological review of the Canadian study designed to address spe-
cifically the issue of tumour grade, found no significant difference across all groups.
The data from these hypofractionation studies is also reassuring from the point of
view of late-​reacting normal tissue, such as the heart. The linear-​quadratic model also
supports this: 40 Gy in 15 fractions is gentler than 50 Gy in 25 fractions if an α/​β of
3.0 (or even 1.5) is used for late-​reacting normal tissue toxicity. The UK breast radio-
therapy consensus meeting in 2016 concluded that 40 Gy in 15 fractions should be the
standard of care with no exceptions. Work is ongoing to test further hypofractionation
(26–​27 Gy in five fractions over 1 week) against 40 Gy in 15 fractions but efficacy data
is not expected until 2019 at the earliest.

Tumour bed boost radiotherapy

Breast boost policies have been strongly influenced by the European Organization
for Research and Treatment of Cancer (EORTC) trial testing 16 Gy in eight fractions
against no boost in women with early breast cancer who underwent complete micro-
scopic excision of their tumours, appropriate systemic therapy, and 50 Gy in 25 frac-
tions of whole breast radiotherapy. The trial randomized 5318 women with stage I/​II
breast cancer to boost (2657) or no boost (2661) and the latest update was reported
72 Breast radiotherapy

in 2015 at a median follow up of 17.2 years(4). Whole breast radiotherapy was 50 Gy

in 25 fractions over 5 weeks and the boost was 16 Gy in eight fractions delivered with
electrons, photons, or Ir192 implant. Adequate excision was defined as no invasive tu-
mour at inked margin. In women randomized to the boost arm, the hazard rate for
local recurrence was 0.65 (99% confidence interval 0.52–​0.81) compared with the no
boost control arm. This corresponds on average, to four fewer local recurrences at
20 years (12% vs 16.4%) for every 100 women treated. In patients <50 years with high
grade tumour and ductal carcinoma in situ (DCIS) a boost reduced 20-​year local re-
lapse from 38% to 9% (HR = 0.21, p = 0.002). The boost did not impact on overall
survival.
Premenopausal status, age <40 years, large tumour size, absence of oestrogen or pro-
gesterone receptors, and lack of systemic adjuvant treatment were shown to be associ-
ated with recurrence in univariate analysis. An EORTC subgroup analysis with central
pathological review shows only high grade as an additional independent risk factor for
recurrence other than young age. High grade was typically associated with early local
recurrence within the first 5 years following completion of treatment. Close surgical
margins were not associated with increased relapse. In a more recent update of this
central pathological review with 17.2 years of median follow up, extensive intraductal
component (EIC) correlates with late recurrences after 5  years (oral presentation
ECCO 2015). Factors reported to predict for local recurrence in other randomized
trials of breast-​conserving surgery and radiotherapy include lymph node positivity,
lymphovascular invasion, high grade, and EIC in younger women.
The risks of moderate and severe fibrosis in the boost group were 25.2% and 5.2%
(total 30.4%) at 20  years respectively, compared to 13.2% and 1.8% (total 15%) in
women treated without boost. In principle, it is reasonable to spare patients the late
adverse effects of boost therapy if risks of local recurrence after whole breast radio-
therapy are less than 5% at 10  years, representing a population in whom the boost
prevents only one or two local recurrences per 100 women treated. On the basis of
the EORTC boost trial, this applies to the majority of patients more than 50 years old.
In summary, a tumour bed boost should be considered for women less than 50 years
old and for those over 50 years with the higher risk pathological features of grade 3
and/​or an extensive intraductal component. The decision over whether or not to offer
a boost should take into account both the risks of local recurrence and of normal tissue
toxicity.

Partial breast radiotherapy
The rationale for partial breast irradiation (PBI) is based on the observation that the
majority of ipsilateral local recurrences occur close to the site of the original tumour,
i.e. the so-​called ‘tumour bed’. PBI gives the potential advantage of matching radio-
therapy dose intensity closely to spatial variation in local recurrence risk within the
breast and may reduce late normal tissue side effects as a smaller breast volume is
irradiated. Many PBI techniques also incorporate hypofractionation, so fewer frac-
tions may be attractive to patients, especially if they need to travel long distances for
treatment.
 Evidence based rationale for radiotherapy 73

PBI is not a new concept. Trials were conducted in the 1980s and 1990s, but these
showed high local recurrence rates compared with whole breast radiotherapy. This was
probably due to a combination of suboptimal patient selection and less sophisticated
radiotherapy planning and treatment techniques. Therefore, PBI fell out of favour
until a renaissance in the 2000s that was triggered, at least in part, by implementation
of more conformal image-​guided radiotherapy methods. As a result, many ‘modern’
randomized controlled trials (RCTs) of PBI were opened internationally. The 2016
Cochrane Review(5) of PBI included seven of these trials and analysed data from 7586
patients. At this time point, local recurrence seemed worse for PBI compared with
whole breast radiotherapy with a hazard ratio of 1.62 (95% CI 1.11–​2.35); however,
there was no statistically significant difference with overall survival (see Figs 5.2 and
5.3). Table 5.1 illustrates the four large RCTs, representing more than 14,000 patients,
that are in follow-​up and as yet unreported for their primary outcome.
IMPORT LOW is a randomized, multi-​centre UK phase III trial testing PBI using
intensity modulated radiotherapy (IMRT), in women with low risk early stage breast
cancer. The 5-​year results were published after the Cochrane review so are not in-
cluded(6). Patients were randomly assigned (1:1:1) to whole breast radiotherapy of 40
Gy delivered in 15 fractions (control); Test 1: 36 Gy to the whole breast and 40 Gy to
the tumour bed (reduced) in 15 fractions; or Test 2: 40 Gy to the tumour bed only
(partial) in 15 fractions. For this study, 2018 women were recruited and, at a median
follow-​up of 71.3 (IQR 60.6–​74.1) months, the 5-​year local recurrence rates were
1.1% (95% CI 0.5–​2.4), 0.2% (0.02–​1.2) and 0.5% (0.2–​1.4) in the whole, reduced,
and partial groups, respectively(6). With estimated absolute treatment differences in
the local relapse rate compared to whole breast radiotherapy of -​0.76 (-​1.03–​0.13)%
for the reduced group and -​0.63 (-​0.98–​0.48)% for the partial group, non-​inferiority
could be claimed for both experimental schedules. Both clinicians and patients re-
ported low levels of moderate/​marked normal tissue effects by 5 years, with a stat-
istically significant improvement in breast appearance and breast hardness in the
PBI group.
The Royal College of Radiologists (RCR) Breast Radiotherapy Consensus (2016)
strongly supported the following statement:
Partial breast radiotherapy can be considered for patients ≥50 years with grade 1–​2,
≤3cm, ER (oestrogen receptor) positive, HER2 (human epidermal growth factor receptor
2) negative and node negative tumours, using either (i) external beam radiotherapy
with 40 Gy in 15 fractions over 3 weeks or (ii) multi-​catheter brachytherapy using
fractionation regimens as per the GEC-​ESTRO trial.

Avoidance of radiotherapy for patients at very low risk of recurrence

Local recurrence rates have fallen dramatically over the last 30 years (see Fig. 5.4) and
improvements in diagnosis, multidisciplinary team working, and treatment (surgery,
systemic therapy, and radiotherapy) have undoubtedly contributed to this decrease.
Despite improvements in technical breast radiotherapy, late normal tissue toxicity is
still an issue for some patients (see section 5.2: Adverse effects of breast/​loco-​regional
radiotherapy), with cardiac morbidity and second radiation-​induced malignancy
 PBI/APBI WBRT Hazard Ratio Hazard Ratio
Study or Subgroup Events Total Events Total O-E Variance Weight Exp[(O-E)/V], Fixed, 95% Cl Exp[(O-E)/V], Fixed, 95% Cl
1.1.1 2.4 years’ median follow-up
TARGIT 23 1721 11 1730 5.34 7.44 27.2% 2.05 [1.00, 4.21]
Subtotal (95% CI) 1721 1730 27.2% 2.05 [1.00, 4.21]
Total events 23 11
Heterogeneity: Not applicable
Test for overall effect: Z = 1.96 (P = 0.05)

1.1.2 5 years’ follow-up

ELIOT 21 651 4 654 6.63 3.36 12.3% 7.19 [2.47, 20.96]
GEC-ESTRO 9 633 5 551 1.45 3.21 11.7% 1.57 [0.53, 4.69]
Livi 2015 0 260 3 260 –1.38 0.75 2.7% 1.16 [0.02, 1.53]
Rodriguez 0 51 0 51 1 0 Not estimable
Subtotal (95% CI) 1595 1516 26.7% 2.50 [1.21, 5.15]
Total events 30 12
Heterogeneity: Chiz = 10.14, df = 2 (P = 0.006); Iz = 80%
Test for overall effect: Z = 2.48 (P = 0.01)

1.1.3 10 years’ follow-up

Polgár 2007 7 128 6 130 1.09 12.61 46.1% 1.09 [0.63, 1.89]
Subtotal (95% CI) 128 130 46.1% 1.09 [0.63, 1.89]
Total events 7 6
Heterogeneity: Not applicable
Test for overall effect: Z = 0.31 (P = 0.76)

Total (95% CI) 3444 3376 100.0% 1.62 [1.11, 2.35]

Total events 60 29
Heterogeneity: Chiz = 13.90, df = 4 (P = 0.008); Iz = 71%
Test for overall effect: Z = 2.51 (P = 0.01) 0.001 0.1 1 10 1000
Test for subgroup differences: Chiz = 3.76, df = 2 (P = 0.15); Iz = 46.8% Favours PBI/APBI Favours WBRT

Fig. 5.2  Cochrane review.

Reproduced with permission from Hickey, B. E., Lehman, M., Francis, D. P. & See, A. M. Partial breast irradiation for early breast cancer. Cochrane
database Syst. Rev. 7, CD007077 (2016) Copyright © 2016 The Cochrane Collaboration. Published by John Wiley & Sons Ltd.
 PBI/APBI WBRT Hazard Ratio Hazard Ratio
Study or Subgroup Events Total Events Total O-E Variance Weight Exp[(O-E)/V], Fixed, 95% Cl Exp[(O-E)/V], Fixed, 95% Cl
ELIOT 34 651 31 654 1.47 15.41 15.6% 1.10 [0.67, 1.81]
GEC-ESTRO 27 633 32 551 –6.12 14.64 14.8% 0.66 [0.39, 1.10]
Livi 2015 1 260 7 260 –1.55 0.88 0.9% 0.17 [0.02, 1.39]
Polgár 2007 25 128 23 130 2.29 46.93 47.5% 1.05 [0.79, 1.40]
TARGIT 37 1721 51 1730 –7.04 21.044 21.3% 0.72 [0.47, 1.10]

Total (95% CI) 3393 3325 100.0% 0.90 [0.74, 1.09]

Total events 124 144
Heterogeneity: Chiz = 6.68, df = 4 (P = 0.15); Iz = 40%
Test for overall effect: Z = 1.10 (P = 0.27) 0.01 0.1 1 10 100
Favours PBI/APBI Favours WBRT

Fig. 5.3  Cochrane review.

Reproduced with permission from Hickey, B. E., Lehman, M., Francis, D. P. & See, A. M. Partial breast irradiation for early breast cancer. Cochrane
database Syst. Rev. 7, CD007077 (2016) Copyright © 2016 The Cochrane Collaboration. Published by John Wiley & Sons Ltd.
76 Breast radiotherapy

Table 5.1  Partial breast irradiation trials: awaiting full publication

Trial Patients Status Eligibility Control Intervention
recruited
NSABP B-​39, 4214 2005 –​ in ≥18 years; 50–​50.4 Gy/​ 34 Gy/​10F
RTOG 0413 follow up unifocal 25–​28F ± 10–16 twice daily
tumour; pTis-​ Gy boost (HDR) or 38.5
2a; pN0-​N1 Gy/​10F twice
daily (3D-​CRT)

RAPID 2135 2006 –​ in ≥50 years; 42.5 Gy/​16F ± 38.5 Gy/​10F

follow up pT1-​2a; pN0 10 Gy boost twice daily
IRMA 3302 2007 –​ in 49–​85 years; 45 Gy/​18F or 50 38.5 Gy/​10F
follow up pT1-​2a; Gy/​25F or 50.4 twice daily
pN0-​N1 Gy/​28F with ±
boost 10–​16 Gy
SHARE 2796 2010 –​ in ≥50 years; 50 Gy/​25F + 16 40 Gy/​10F
follow up pT1; pN0 Gy boost twice daily
or 40–​42.5 Gy/​
15–​16F

being rare but serious side effects. Therefore, the absolute benefit of radiotherapy for
some patients may not outweigh the potential risks.
RCTs to date comparing whole breast radiotherapy with no radiotherapy in women
at low-​risk of local recurrence show an increase in local recurrence without radio-
therapy but consistently no increase in breast cancer death. In addition, evidence
shows that it is detrimental to avoid both radiotherapy and endocrine therapy in ER +
patients, such that compliance with endocrine therapy should be strongly encouraged
if radiotherapy is omitted.
Local recurrences %

BCT arm of the BCT -Mastectomy trial

Boost arm of the Boost no Boost trial 41–61 yrs

Young Boost trial <51 years

0
0 2 4 6 8 10 12 14 16 18
Fig. 5.4  Decreasing local relapse.
Reproduced with permission from Bartelink, Harry et al., 'Has partial breast irradiation by IORT or
brachytherapy been prematurely introduced into the clinic?', Radiotherapy and Oncology, Volume 104,
Issue 2 , pp. 139–142, DOI: (https://​doi.org/​10.1016/​j.radonc.2012.07.010) Copyright © 2012 Elsevier Ltd.
 Evidence based rationale for radiotherapy 77

These RCTs have so far not changed clinical practice significantly such that the
majority of patients still receive breast radiotherapy following breast conserving sur-
gery. A  likely explanation for this is that fact that there is still difficulty in clearly
identifying which groups of patients are at the very lowest risk of recurrence. In an
unplanned subgroup analysis of the PRIME II trial, ER-​rich patients receiving radio-
therapy had only a 2.4% absolute gain in local relapse over non-​irradiated patients at
5 years: 0.8% (95% CI 0.3–​1.9) local relapse with radiotherapy vs 3.2% (95% CI 2.1–​
5.2) without radiotherapy(7). This observation suggests that it is possible to identify a
group of patients who are at very low risk of recurrence, and a ‘personalized radiation
therapy’ approach using biomarkers of risk is now being adopted in research studies
worldwide.
PRIMETIME is a UK prospective, biomarker-​directed study. It will use ‘IHC4 + C’
which is a refinement of immunohistochemical phenotyping, combining protein ex-
pression of ER, progesterone receptor (PgR), HER2, and Ki67 with clinico-​pathological
parameters to identify breast cancer patients at very low, low, intermediate, or high risk
of distant disease recurrence. Patients identified as being at very low risk of relapse
will be offered the opportunity to avoid adjuvant radiotherapy. Patients in the other
three risk groups will be offered standard adjuvant therapy. All patients will be recom-
mended adjuvant endocrine therapy as per local policy for a minimum of 5 years (see
Fig. 5.5 for study design).

Radiotherapy for ductal carcinoma in situ

Whole breast radiotherapy reduces the risk of ipsilateral breast recurrence after com-
plete excision of pure DCIS in women who do not opt for mastectomy. Four RCTs
indicate that the risks of both recurrent DCIS and invasive disease are reduced by
approximately 50%. The dose prescription in all trials was 50 Gy in 25 fractions, or
its near equivalent, to the whole breast without a planned boost dose to the tumour
bed. A twofold reduction in local recurrence risk does not necessarily reduce absolute
recurrence risks to very low levels. In women randomized to breast radiation, the rate
of ipsilateral tumour recurrence was 8.9% at 15 years (NSABP B-​17), 18% at a me-
dian of 15 years (EORTC 10853), 7.1% at a median of 12.7 years (UK Co-​ordinating
Committee on Cancer Research, UKCCCR), and 7% at 5 years (SweDCIS). About 50%
of the local recurrences are invasive carcinomas. The overview confirms data from the
individual trials, reporting the 10-​year risk reduction of ipsilateral tumour recurrence
from 28.1% to 12.9%(8).
The risk of recurrence after complete microscopic resection of symptomatic DCIS
is reported by Silverstein to be closely related to patient age, tumour size, grade, and
margins of excision. The Van Nuys Prognostic Index (VNPI) system is shown in
Table 5.2(9). The VNPI has been widely applied as a basis of treatment recommenda-
tions, but it is not independently validated as a guide to therapy. Retrospective
evaluations by the NSABP and EORTC do not appear to lend strong support to Van
Nuys. Data from both the most recent EBCTCG overview and the Cochrane review
suggest that it is correct, in principle, to consider whole breast radiotherapy after
breast conservation surgery to all patients with DCIS regardless of size, grade, or
margins(10).
78 Breast radiotherapy

P R I M E T I M E
Eligible Patient Group (n = 2400)
• ≥60 years
• T1, N0, G1–2
• ER/PR+ve, HER2-ve

Central testing of ki67 Fig. 5.5 Primetime.

Reproduced with permission from
WLE & SLNB Kirwan, C. C., Coles, C. E. & Bliss, J.
‘It’s PRIMETIME. Postoperative
Avoidance of Radiotherapy:
Confirmation of eligibility-
PRIMETIME study registration Biomarker Selection of Women at
Very Low Risk of Local Recurrence.’
Clin. Oncol. (R. Coll. Radiol).
IHC4+C score: IHC4+C score: Volume 28, pp. 594–6 (2016).
very low Low, intermediate, high DOI: https:// doi.org/ 10.1016/
j.clon.2016.06.007, Copyright
No Radiotherapy Radiotherapy © 2016 The Royal College of
(endocrine therapy as per (endocrine therapy as per Radiologists. Published by
standard of care) standard of care)
Elsevier Ltd.

The EBCTCG data showed an absolute reduction of 18% in ipsilateral breast cancer
recurrence in a low-​risk group of women with negative margins and small low-​grade
DCIS. The ECOG 5194 trial enrolled patients with low-​and intermediate-​grade DCIS
less than 25 mm (n = 565) or high-​grade DCIS less than 10 mm (n = 105), and sur-
gical excision margins greater than 3 mm(11). Twelve-​year recurrence rates were 14.4%
for the low and intermediate-​grade group and 24.6% for the high-​grade cohort, with
7.5% and 13.4% respectively, being invasive. The role of multi-​gene assays to identify
a low-​risk group of patients who may not require radiotherapy after breast conserving

Table 5.2  Van Nuys Prognostic Index (VPNI)

Age DCIS size (mm) Histological grade Margin width (mm)
Score 1 More than 60 Less than 16 Grade 1–​2 without 10 or more
necrosis

Score 2 40–​60 16–​40 Grade 1–​2 with necrosis 1–​9

Score 3 Less than 40 More than 40 Grade 3 Less than 1

Overall VPNI score 3–​4 5–​7 8–​9

8 year local recurrence free survival rate 97% 77% 20%
8 year breast cancer specific survival rate 100% 97% 100%
 Evidence based rationale for radiotherapy 79

surgery is being explored. A subset of 327 patients with adequate tissue from ECOG
5194 trial was assessed using a 12-​gene pre-​defined scoring system. A lower risk group
was identified but not low enough to recommend omitting radiotherapy. The concern
with all studies omitting radiotherapy is the failure for recurrences to plateau with
time but the reassurance is failure to find a detriment to survival. Trials such as the UK
LORIS study continue to try to identify a group that do not require surgery or radio-
therapy and it is recommended that these be supported.
Retrospective studies have reported on hypofractionated schedules with no con-
cerns regarding efficacy identified. Hypofractionated regimens such as the UK
standard START trial B(10) schedule of 40 Gy in 15 fractions are proven for invasive
breast cancer and there is no logical reason why they should not be used for DCIS. At
the other end of the prognostic scale, it is reasonable to consider a radiotherapy boost
dose after whole breast radiotherapy in ‘high-​risk’ patients who decline mastectomy,
based on a high-​quality retrospective study, although there are other retrospective re-
ports not all of which support the use of boost. The results of the BIG 3-​07/​TROG
07.01 trial testing both hypofractionation and boost are awaited. Radiotherapy to the
nodal areas is not indicated for pure DCIS, and is rarely indicated for the chest wall
after mastectomy.

5.1.2  Adjuvant loco-​regional radiotherapy

Chest wall radiotherapy
The EBCTCG meta-​analysis of the effect of radiotherapy after mastectomy and axillary
surgery reported outcomes in 8135 women treated within 22 trials of post-​mastectomy
radiotherapy (PMRT) vs no PMRT between 1964 and 1986(12). In women with patho-
logically node-​negative disease on axillary node dissection, PMRT had no significant
effect on loco-​regional recurrence or breast cancer mortality. However, in women with
lymph-​node positive disease, PMRT reduced the risk of both loco-​regional recurrence
and breast cancer mortality irrespective of the number of involved lymph nodes and
the use of adjuvant systemic therapy. In terms of absolute benefits, in women with 4–​9
positive lymph nodes treated with systemic therapy, PMRT reduced the risk of local
recurrence from 35% to 11% but had no significant effect on breast cancer mortality
(68% vs 67%). In women with 1–​3 positive axillary lymph nodes, however, PMRT
significantly reduced the risk of both local recurrence (21% vs 4%) and breast cancer
mortality (49% vs 41%).
Of course, in patients treated with modern systemic therapy, it is expected that the
absolute gains will be much smaller such that, in the 1–​3 lymph node positive group
for example, the reduction in breast cancer mortality at 20 years would fall from 8% to
4%. However, in contrast to the breast conservation surgery literature, the EBCTCG
meta-​analysis suggests a ratio of local recurrences: breast cancer deaths of around 1.5:1
such that, for every three local recurrences prevented at 10 years, two breast cancer
deaths will be prevented at 20 years.
Aside from lymph-​node involvement, other risk factors for loco-​regional recur-
rence after mastectomy have been identified by a review of 5352 women entered into
International Breast Cancer Studies Group (IBCSG) trials between 1978 and 1993.
80 Breast radiotherapy

These include age, tumour size, and the presence of lymphovascular invasion(13).
Microscopic disease at resection margins has also been reported to be a risk factor for
local recurrence.
In terms of current recommendations, the St Gallen consensus statement for
PMRT(14) recommends PMRT as standard in patients with T3 disease, those with a
positive macrometastatic sentinel lymph node biopsy and no axillary node dissection,
and those with 1–​3 lymph nodes positive and other adverse pathological features. The
most recent amendments to the NCCN guidelines(15) continue to recommend PMRT
in women with four or more lymph nodes positive, and also state that regional nodal
irradiation be strongly considered for women with 1–​3 positive lymph nodes.
Sources of ongoing variations in international standard practices are based on a
failure to reach consensus on: (i) the level of local relapse risk that justifies radiotherapy
and (ii) the characteristics of the population of women treated in the current era by
mastectomy and systemic therapies that are exposed to that risk. Whatever level of
local recurrence risk is chosen and whichever criteria are applied to estimate this risk,
local guidelines or algorithms (such as the Cambridge Post-​Mastectomy Radiotherapy
Index(16)) should be in place to ensure treatment decisions are consistent. Regular audit
of populations managed with and without radiotherapy offer a guide to the suitability
of local management guidelines.
In patients who have undergone neoadjuvant systemic therapy prior to mastec-
tomy, international guidelines recommend that PMRT be directed towards the extent
of disease prior to systemic therapy. There is growing debate, however, regarding the
role of local therapies (surgery and radiotherapy) in patients who have had a com-
plete pathological response to systemic therapy. A retrospective analysis of outcomes
in the NSABP B-​18 and B-​27 trials suggests that patients who achieve a complete
pathological response have low loco-​regional relapse rates even in the absence of
radiotherapy. Another retrospective analysis however suggests that, in women with
lymph-​node positive disease at presentation, even where a complete pathological re-
sponse is achieved, loco-​regional relapse rates are considerably reduced in those who
undergo loco-​regional radiotherapy. The US NSABP B-​51 trial of PMRT vs no PMRT
in patients treated with systemic therapy to pathological complete response is cur-
rently recruiting (NCT01872975).

Radiotherapy to the axilla
The axilla refers to levels 1–​4, with level 4 corresponding to the supraclavicular fossa
(SCF). When referring to SCF radiotherapy for many patients, this includes level 3 of
the axilla, which is the infraclavicular fossa, on the basis that level 3 has not usually
been completely cleared. For patients with a pre-​operatively negative axilla by imaging
with or without biopsy, the use of sentinel node biopsy is routine. The presence of
isolated tumour cells or micrometastases requires no further axillary therapy but ad-
equate systemic adjuvant therapy is required. The EORTC AMAROS trial studied axil-
lary lymph node dissection (ALND) against radiotherapy for patients with one or two
macroscopically involved nodes. This trial enrolled 4823 patients and they were pre-​
randomized from 2001 to 2010 with 1425 found to have involved axillary nodes. Of
the 744 patients out of 1425 undergoing ALND, 33% had additional node involvement
 Evidence based rationale for radiotherapy 81

to the sentinel node(s). At a median follow up of 6.1 years axillary recurrence occurred

in 4 (0.43%) surgical patients and 7 of 681 (1.19%) randomized to radiotherapy, with
isolated axillary recurrence in 2 and 4 patients respectively. Twenty-​five (0.72%) of
3131 with negative sentinel nodes also developed an axillary recurrence. Radiotherapy
is an alternative intervention to surgery when axillary therapy is required providing
there is low burden disease (clinically node negative), although the long-​term results
of AMAROS will be awaited with interest.
There is debate about whether intervention is required when one or two sentinel
nodes are involved macroscopically with the Z0011 trial concluding that the use of
sentinel lymph node biopsy (SLNB) did not result in inferior survival compared to
ALND. The Z0011 trial planned to recruit 1900 patients with T1/​2 breast cancer man-
aged with breast conservation and one or two involved sentinel nodes over a 4-​year
time span and to randomize to ALND or SLNB alone. They reported on 445 patients
randomized to ALND and 446 to SLNB (47% of planned number) from 115 centres
but in total only 430 patients had axillary macrometastases. The trial is underpowered
and therefore the results are hypothesis generating. Further information is required
and it is recommended that trials such as POSNOC be supported.
After axillary node clearance, radiotherapy to the dissected axilla is not recom-
mended even if there is extracapsular spread, due to an enhanced risk of treatment-​
related morbidity. In the subgroup with microscopic tumour at the margins of axillary
resection, positive excision margins do not appear to be an independent risk factor
for axillary recurrence. Macroscopic residual disease at the axillary apex is a different
matter: if further surgery is not possible, it justifies cytoreductive chemo-​endocrine
therapy and radiotherapy (with a boost to the axillary apex, assuming the patient is fit
enough) in an attempt to prevent malignant brachial plexopathy. Uncontrolled axillary
recurrence is an endpoint to try to avoid if possible.
Patients with positive axillary nodes have ≥5% lifetime risk of SCF node involvement
and SCF recurrence is associated with significant morbidity. In a large retrospective
study of over 1000 patients, the number of positive nodes and tumour grade were
strong independent risk factors for recurrence. Radiotherapy to the SCF in patients at
more than 5% risk of positive SCF nodes is expected to prevent much more morbidity
than it causes in terms of shoulder stiffness, arm oedema, brachial plexopathy, and
vascular disease (cerebral vascular accident).
A common practice has been to give radiotherapy to the SCF in patients with four
or more metastatic axillary lymph nodes, on the basis that this strikes an appropriate
balance between the morbidity of local tumour recurrence and of late radiotherapy
adverse effects. The EBCTG looked at 1314 women with one to three involved nodes
and concluded ‘After mastectomy and axillary dissection, radiotherapy reduced
both recurrence and breast cancer mortality in the women with one to three posi-
tive lymph nodes in these trials even when systemic therapy was given.’ The EORTC
22922/​10925 study reported that irradiation of all regional nodes, including SCF,
resulted in an improved disease-​free and distant disease-​free survival at a median
follow-​up of 10.9 years. Similarly, the Canadian MA.20 study showed that regional
radiotherapy including to the SCF reduced breast cancer recurrence at a median of
9.5 years(17).
82 Breast radiotherapy

There is increasing use of neoadjuvant chemotherapy (NACT) for biologically ag-

gressive or loco-​regionally advanced breast cancer. Mamounas et  al. examined pre-
dictors of loco-​regional recurrence following NACT with a combined analysis of
NSABP B-​18 and B-​27(18). The patients were recruited between 1988 and 2000 at which
time few patients received taxanes and no patients received adjuvant trastuzumab.
Following breast conserving surgery or mastectomy with ALND, the protocols did not
allow for regional nodal irradiation. The 10-​year cumulative regional recurrence after
breast conserving surgery with radiotherapy and ALND was 0–​2.4% if clinical node
negative pre-​NACT or clinically node positive pre-​NACT but pathologic node nega-
tive post-​NACT. If pathologically node positive post-​NACT, the regional recurrence
rate was 7.5–​8.7%. In the light of this data, prospective studies including NSABP B-​
51 and RAPCHEM are evaluating reducing or omitting loco-​regional radiotherapy in
patients who have achieved complete pathological response after NACT. Whilst these
studies accrue and mature, the majority of UK centres continue to make loco-​regional
radiotherapy recommendations based on pre-​chemo staging.

Radiotherapy to internal mammary chain

The 2014 systematic overview by the EBCTCG reported a 9% absolute reduction in
breast cancer mortality in node-​positive women randomized to PMRT vs no radio-
therapy irrespective of the number of involved nodes and the use of adjuvant systemic
therapies. The majority of patients received chest wall and regional lymph node radio-
therapy (including the SCF (or level IV axilla), level I–​III axilla, and internal mammary
chain (IMC)). These 20-​year analyses are consistent with data from three recently re-
ported randomized trials (n>9000) testing the addition of loco-​regional lymph node
radiotherapy to local radiotherapy and reporting a 3–​5% disease-​free survival benefit
in women with stage I–​III breast cancer, again independent of the number of involved
lymph nodes.
The EORTC IM-​MS study randomized 4004 women with node positive breast
cancer (43%) or node negative central or medial cancers to chest wall/​breast radio-
therapy only plus or minus radiotherapy to the IMC and SCF (with fields equivalent to
level III–​IV axilla). Patients were treated to 50 Gy in 25 fractions with mixed electron
and photon fields, and results were reported at a median follow-​up of 11 years. The
addition of SCF and IMC radiotherapy improved the 10-​year disease-​free survival
from 69% to 72% (p = 0.044)(19).
The MA-​20 study randomized 1832 patients with node positive breast cancer (90%)
or high-​risk node negative breast cancer (predominantly T3N0) to breast radiotherapy
plus or minus radiotherapy to the level III–​IV axilla and IMC. Patients were treated to
50 Gy in 25 fractions with wide tangential fields matched to an anterior level III–​IV
axillary field. The majority of patients in this study were treated with chemotherapy
(including taxanes where indicated). Irradiation of the loco-​regional lymph nodes im-
proved the 10-​year disease-​free survival from 77% to 82% (p = 0.01)(18).
Although the results of the EBCTCG meta-​analysis, EORTC IM-​MS, and MA-​20
studies are supportive of including the IMC in the irradiated volume in patients with
higher-​risk early breast cancer, none of these studies are pure tests of the effect of
IMC radiotherapy. The Danish Breast Cancer Co-​operative Group (DBCG), however,
 Evidence based rationale for radiotherapy 83

undertook a population-​based cohort study in which patients with node-​positive early

breast cancer were naturally randomized according to laterality of their cancer(20). Thus,
patients with node-​positive right breast cancer were treated with loco-​regional radio-
therapy (breast/​chest wall, level I–​IV axilla) and IMC radiotherapy, whilst those with
node-​positive left breast cancer were treated with loco-​regional radiotherapy without
IMC radiotherapy. Similar to the MA-​20 study, the majority of patients were treated
using wide-​tangential fields matched onto anterior axillary fields to a dose of 50 Gy
in 25 fractions; 3071 patients were accrued between 2003 and 2007, 55% of whom re-
ceived chemotherapy, albeit none with taxanes. At a median follow-​up of 10 years, the
inclusion of the IMC improved overall survival from 68% to 72% (p = 0.015), with the
greatest degree of benefit being seen in the N2–​3 patients and in those with N1 disease
together with medial or central primary tumours.
No increase in non-​breast cancer-​related mortality was reported in any of the EORTC
IM-​MS, MA-​20, or DBCG studies, albeit differences in cardiac or second cancer mortality
might not be expected before 15–​20 years of follow-​up has elapsed. The DBCG group has,
however, modelled the expected benefits vs risks of implementing IMC radiotherapy in
women with node-​positive early breast cancer. For their whole trial population (i.e. all
women with node-​positive disease), the number needed to treat (NNT) with IMC radio-
therapy would be 33 patients to prevent one breast cancer death. By limiting IMC only
to those women with N2–​3 or N1 medial/​central disease, the NNT would be 14 patients.
Assuming IMC radiotherapy increases mean heart doses by around 5 Gy, and using the
Darby data relating mean heart dose to risk of radiation-​induced coronary events(21), the
number need to harm (NNH) would be around 10,000 for women with no pre-​existing
cardiac risk factors, and 1000 for women with pre-​existing cardiac risk factors.
The RCR Breast Radiotherapy Consensus (2016) agreed that IMC radiotherapy
should be considered in patients at high risk of loco-​regional recurrence (i.e. T4 or
N2–​3 disease). The consensus meeting also agreed that, in patients with 1–​3 axillary
macrometastases who have been recommended loco-​regional irradiation based on
risk factors, inclusion of the IMC in the target volume should be considered in those
with central or medial disease. It was also unanimously agreed that IMC radiotherapy
should be given using techniques which minimize doses to organs-​at-​risk (OARs) and
that every UK radiotherapy centre should have a breath-​hold technique available for
patients undergoing IMC radiotherapy.
In terms of refining the selection criteria for IMC radiotherapy, the EBCTCG are likely
to undertake an individual patient data meta-​analysis of trials of IMC radiotherapy. It
may also be possible to select patients by testing patterns of lymphatic drainage.

5.1.3  Other indications for radiotherapy

Primary breast radiotherapy
The indication for primary breast radiotherapy has been for locally advanced tumours
to either downstage prior to surgery or to achieve best local control for patients with in-
operable cancer. There is a paucity of evidence in this area, with the majority of reports
consisting of retrospective series or single arm cohort studies. Primary radiotherapy
is usually administered following systemic treatment, or in some series concomitantly
84 Breast radiotherapy

with chemotherapy. A standard adjuvant breast dose is generally used for down-​staging
before surgery, whereas a higher dose may be used when surgery is not possible.
Recently, there has been renewed interest in primary breast radiotherapy. For ex-
ample, there are some series reporting possible benefits of scheduling primary radio-
therapy before planned breast reconstruction. In addition, pre-​operative PBI has
received some attention. A European phase II single arm trial has published on the
possible advantages of pre-​operative PBI in terms of improved tumour/​tumour bed
delineation and reduced breast fibrosis. This will be investigated further within a ran-
domized trial and other studies are underway in North America. The introduction of
more targeted radiotherapy techniques with image guidance could facilitate delivery
of pre-​operative PBI. These newer potential indications for primary breast radio-
therapy, coupled with down-​staging for less chemosensitive tumours will be tested in
on-​going/​future research studies.

Re-​irradiation
There are no randomized trials testing re-​irradiation for breast cancer. A  relatively
large series of patients with unresectable local relapses has demonstrated long-​term
local control with acceptable toxicities following hypofractioned radiotherapy and
hyperthermia. There are some studies suggesting that re-​irradiation to higher doses
can achieve good local control rates (66–​77% at 2–​4 years) with a 5% risk of grade
3–​4 toxicity (principally rib fractures and skin damage). However, these studies used
older conformal radiotherapy techniques and older fractionations (1.6 to 2.5 Gy per
fraction). More recent advances in radiotherapy including volumetric-​modulated arc
therapy allow for better shaping of the dose around the area of disease thereby min-
imizing the volume of overlap with the previously treated area which reduces the risk
of toxicity and, in turn, allows for higher doses to be delivered. The increasing use
of breath-​hold techniques also facilitates reduced re-​irradiation by reducing the vol-
umes of lung and heart tissue within the treated volume. Partial breast or chest wall
radiotherapy, mostly with brachytherapy but also with electrons, has also been used
following further surgery for those patients declining mastectomy or following mast-
ectomy with close or involved margins.

Palliative breast radiotherapy
Breast radiotherapy can be useful in palliating local breast symptoms in the metastatic
setting and also for locally advanced inoperable tumours. Simple radiotherapy tech-
niques with hypofractionated dose regimens are usually indicated in this situation.

5.2  Adverse effects of breast/​loco-​regional

radiotherapy
The potential acute, intermediate and late adverse effects of radiotherapy for breast
cancer are listed in Table 5.3 and will need to be discussed with patients as appropriate
as part of informed consent.
 Adverse effects of breast/loco-regional radiotherapy 85

5.2.1 Acute effects
The most common acute side effects of radiotherapy for breast cancer are skin red-
ness and soreness, and fatigue. The figures in Table 5.3 relate to the 40 Gy in 15
fraction regimen. Using this regimen, moist desquamation is uncommon unless the
patient is large-​breasted or bolus has been used. The skin reaction peaks around 1–​2
weeks after the radiotherapy course has been completed. There is little evidence to
suggest that any particular skin cream is better than any other in reducing the se-
verity or duration of the skin reaction. Patients are however advised to reduce fric-
tion on the skin

Table 5.3  Side effects from breast radiotherapy

Side effect Risk Ref
Acute side effects

Skin redness and irritation 90–​100% 34

Moist desquamation <10% unless patient is larger-​breasted or 34

being treated with bolus

Fatigue 50–​100%
Intermediate side effects
Pneumonitis 4% in patients undergoing IMC 19

radiotherapy
Late side effects (breast/​chest
wall)
Breast shrinkage Moderate or marked 22%
Breast induration Moderate or marked 18%
Breast/​chest wall tenderness 6% at 5 years
Rib fracture <1% at 10 years
Heart disease <1% at 10 years 21

Symptomatic lung fibrosis <1% at 10 years 19

Second malignancy 0.1 to 0.3% at 15 years

Breast implant capsule 30% risk of needing implant replaced
contractions within 5 years
Late side effects (regional LNs)
Arm lymphoedema <10% at 10 years 19

Shoulder stiffness 10% at 10 years (only where humeral head

in field)
Brachial plexopathy <0.1% at 10 years
86 Breast radiotherapy

5.2.2 Late effects
Cardiac
The EBCTCG meta-​analysis demonstrates a 1% increase in the risk of non-​breast-​
cancer-​related death at 15  years in patients undergoing radiotherapy for breast
cancer(1), the majority of which is due to heart disease. The relationship of the mean
dose of radiation to the heart and the risk of major coronary events has been shown
to be linear with no threshold below which the risk is zero(21). The proportional in-
crease in risk was similar in women with and without cardiac risk factors. It is not yet
clear which cardiac substructures, when irradiated, contribute the most to the risk of
cardiovascular disease but evidence from myocardial perfusion and coronary angiog-
raphy studies suggests that the left anterior descending coronary artery is a key struc-
ture in the pathogenesis of radiation-​induced heart disease, such that heart-​sparing
breast radiotherapy techniques should aim to reduce not just mean heart dose but the
dose delivered to the anterior aspect of the heart in particular (see section 5.3.1).

Second radiation-​induced malignancy
Death from second malignancy in lung tissue accounts for <10% of non-​breast-​
cancer-​related deaths after radiotherapy for breast cancer(1). The relative-​risk of death
from a second malignancy in the lung ranges from 1.5 to 2.8 at 15 years, with an odds
ratio of up to 37.6 in smokers(22). Data suggest a dose-​response relationship with an
incremental relative risk of 0.2 per Gy to ipsilateral lung (equating to 9 cases of second
primary lung malignancy/​year/​10,000 women receiving 10 Gy to lung and living to
10 years)(22).
The EBCTCG meta-​analysis also demonstrates an increased incidence of contra-
lateral breast cancer in irradiated women (9.3% vs 7.5% at 15 years, p = 0.02)(1), with
the main excess risk appearing at years 5 to 14 following radiotherapy. A case–​control
study found that radiotherapy increased the risk of second primary contralateral
breast cancer only in those irradiated under 45 years of age(22). Although the majority
of contralateral breast cancers arise in the upper outer quadrant, a higher proportion
was found in the inner quadrants in previously irradiated women. In women aged
<40 years, those who received more than 1 Gy of radiation to the index quadrant had
a 2.5-​fold greater risk of contralateral breast cancer than unexposed women (95% CI
1.4–​4.5)(23). The dose–​response relationship was also significant (excess relative risk
per Gy of 1.0, 95% CI 0.1–​3.0) suggesting that attempts should be made to limit the
mean contralateral breast dose to less than 1.0 Gy in younger women undergoing
breast cancer radiotherapy.
With regards to other tissues, the EBCTCG study(1) reported a 20% increased in-
cidence of second primary malignancies (SPM) in irradiated vs unirradiated women
(standardized incidence ratio (SIR)  =  1.20). This equates to around 60 second ma-
lignancies per 10,000 women at 10 years. Significant excess risks were found for oe-
sophagus (SIR = 2.06), soft-​tissue sarcoma (SIR = 2.34), and leukaemia (SIR = 1.71).
Two small studies have estimated sarcoma incidence following breast radiotherapy to
be 2 per 10,000 women at 10 years, the majority arising in the breast and chest wall.
 Pre-planning procedures 87

The incidence of SPM is markedly increased in women irradiated under 40 years of

age and, with the increasing use of systemic therapies, the incidence of second malig-
nancy may increase further, one study reporting a 4% incidence of second malignancy
at 10 years following treatment with chemo-​and radiotherapy(24).

Lymphoedema
A recent systematic review of breast-​cancer-​related lymphoedema (BCRL) sug-
gests that the incidence is three times higher in those who have undergone axillary
node clearance (20%) than in those who had a sentinel lymph node biopsy (6%)(25).
Radiotherapy is one of a number of factors including number of lymph nodes dis-
sected, mastectomy, obesity, use of chemotherapy, and lack of physical exercise that
have a moderate to strong level of evidence for increasing the incidence of BCRL.
Arm lymphoedema rates in the START trial (in which the majority of women under-
went radiotherapy to the breast or chest wall only) were very low (3.7% at 10 years
in the 40 Gy arm). In the loco-​regional lymph node irradiation arm of MA-​20 the
lymphoedema rate was 8.4% at 10  years albeit that the nodal fields included the
level III and much of level II axilla even in patients who had undergone axillary
node clearance. In the AMAROS study, there was a measured increase in arm cir-
cumference at 5 years in 43 of 328 (13%) patients after axillary lymph node dissec-
tion and 16 of 286 patients (6%) following sentinel lymph node biopsy followed by
radiotherapy.

Other late effects
Medium to long-​term complications in breast and chest wall tissues include breast
pain, swelling, firmness, shrinkage, and skin changes. The START-​B trial reported
a 22% risk of moderate or marked breast shrinkage and 18% risk of moderate of
marked breast firmness at 10 years but used predominantly 2-​dimensionally-​planned
radiotherapy. The Cambridge IMRT study used simple IMRT techniques to reduce
inhomogeneities of dose within the breast. Poor cosmesis as rated by clinicians was
22% in the control arm and 12% in the IMRT arm at 5 years. The overall rate of adverse
outcomes for both arms reported using PROMs was low at 5 years (6% reported breast
pain, 4% skin problems, less than 0.5% breast swelling, 15% change in breast appear-
ance, 13% breast shrinkage, and 8% breast firmness).

5.3  Pre-​planning procedures
5.3.1  Patient position and immobilization
The most common positioning for breast radiotherapy is with the patient lying supine
on a CT compatible breast board with either both or one arm raised above their head.
Specialized breast boards have support systems that are adjusted to suit the size and
shape of the individual. The head, elbow, and wrist supports as well as the board
inclination and bottom stop should all be adjustable with clearly marked labels so
the exact set-​up can be reproduced throughout treatment. The patient and the board
should be positioned on the CT couch such that they are able to fit into the bore of
88 Breast radiotherapy

Fig. 5.6  MT-​350-​N
Carbon Fibre Tilting
Breastboard with arm
supports and indexed
‘bum stop’.
Courtesy of CIVCO and
Oncology Systems Ltd,
Shrewsbury, UK.

the CT scanner. Once on treatment the set-​up should be reproduced accurately using
the information recorded (including photographs) at the time of the planning CT
(Fig. 5.6).
A prone treatment position (see Fig.  4.2, Chapter  4) for breast radiotherapy im-
proves dose homogeneity, reduces wedge requirements with consequent reduction of
scattered dose, and reduces dose to lung, particularly in women with larger breast cup
sizes (≥D), but also in women of average cup size (median C). However, data on the
effect of prone positioning on cardiac dosimetry are more conflicting. A small study
in large-​breasted women treated with conventional tangential fields found that prone
positioning reduced heart V30Gy but larger studies in women unselected by cup size
report no significant difference in mean heart doses between supine and prone posi-
tions. Indeed, given that the heart falls anteriorly in the prone position, the prone
position may even increase mean heart doses in smaller-​breasted women. Coverage
of level I and II axillary lymph nodes by tangential fields is also reduced in the prone
position.
Reproducing the prone position is challenging and successful implementation is
likely to require more complex verification protocols such as cone-​beam CT. Recent
work demonstrating that supine breath-​hold is both superior to prone positioning in
terms of mean heart dose and more reproducible is likely to support implementation
of supine breath-​hold rather than prone positioning as the heart-​sparing technique of
choice for most UK radiotherapy departments (Fig. 5.7).
Breath-​holding techniques
Breath-​holding techniques are predominantly used in combination with supine posi-
tioning. In deep-​inspiratory breath-​hold (DIBH), the diaphragm pulls the heart pos-
teriorly, medially, and inferiorly away from the chest wall such that mean heart doses
are approximately halved regardless of technique used. For example, Wang et  al.(26)
reported a reduction in mean heart dose from 3.2 Gy (free-​breathing) to 1.3 Gy
(DIBH) for forward-​planned breast IMRT, whilst Nissen et  al.(27), in the context of
 Pre-planning procedures 89

Fig. 5.7  Example of immobilization for prone positioning.

Reproduced with permission from Radiotherapy and Oncology, 100(2), Anna M. Kirby et al.
‘A randomised trial of Supine versus Prone breast radiotherapy (SuPrstudy): Comparing set-​up errors
and respiratory motion,’ pp. 221–​6 (2011), DOI: https://​doi.org/​10.1016/​j.radonc.2010.11.005,
Copyright © 2010 Elsevier Ltd.

forward-​planned IMRT to breast and pan-​regional lymph nodes, reported a reduction

in mean heart dose from 5.2 Gy (free-​breathing) to 2.7 Gy (DIBH).
Technical approaches to delivering radiotherapy in breath-​ hold range from
maintaining the breath-​ hold using an external valve (e.g. the Active-​ Breathing-​
Controlled device (Elekta)), to delivering radiotherapy only during the inspiratory
phase of the breathing cycle (for example the Varian Real-​Time Position Monitoring
gating solution (Varian Medical Systems, Palo Alto, USA)). The reproducibility of
gating solutions can be facilitated by the use of goggles providing visual feedback to
the patient on their breathing cycle, and/​or by 3D-​optical surface imaging technolo-
gies such as AlignRT® (Vision RT Ltd, London, UK).
An alternative approach is to more simply ask patients to take a breath in and hold
it for around 20 seconds whilst the radiotherapy is delivered. This ‘voluntary breath-​
hold’ technique has been shown to be as heart-​sparing and reproducible as active
breathing controlled breath-​hold and is simple and cheap to implement given that
it requires little more than a standard linear accelerator and a felt-​tip pen to mark
the edges of the light field in relation to the position of the chest wall in breath-​hold.
The Royal College of Radiologists Breast Radiotherapy Consensus Meeting (March
2016)  recommended that the heart be routinely excluded from the radiotherapy
field in patients undergoing radiotherapy for left breast cancer. Multi-​leaf collimator
(MLC) cardiac shielding can be considered in patients with tumours in the upper
half of the breast, in whom the risk of shielding the tumour bed is low. However, for
patients with disease in the central or lower left breast, breath-​holding techniques are
recommended in order to spare heart without compromising coverage of the target
volume (Fig. 5.8).
90 Breast radiotherapy

(a)

(b)
Fig. 5.8  Axial images of
a patient planned for left
chest wall radiotherapy in
a) free-​breathing and b)
deep inspiratory breath-​
hold (using the voluntary
breath-​hold technique)
demonstrating that, in
breath-​hold, the cardiac
structures are pulled
posteriorly and medially
away from the tangential
fields.

5.3.2  Imaging for radiotherapy planning

Computed tomography (CT) imaging
CT imaging, combined with 3-​dimensional treatment planning software packages,
is standard for delineating structures and planning breast radiotherapy. CT imaging
clearly depicts thoracic anatomy in 3-​dimensions enabling localization of breast tissue,
nodal areas, vessels, and regions of interest, usually without the need for IV contrast. It
also facilitates placement of the tangential fields so as to cover the target volume with
minimal inclusion of underlying lung, ribcage, and heart. Three-​dimensional planning
using CT imaging also facilitates homogenization of dose as well as providing imaging
reference data for on-​treatment verification.
CT images are acquired from mastoid processes to below the diaphragm. The en-
tirety of the lungs should be imaged in order that accurate OAR dose volume statis-
tics be reported. Marking the edge of palpable breast tissue prior to CT imaging with
radio-​opaque wire can be helpful (see section 5.3.3 Contouring). The CT slice spacing
should be chosen to suit the technical requirements of each case, with a maximum
interval of 5 mm. Finer slice spacing may facilitate more accurate volume definition,
for instance of nodal regions or small boost volumes.
 Pre-planning procedures 91

In-​room laser positioning systems, present at localization and later in the linac treat-
ment room, project reference position marks onto the patient’s skin. At localization,
radio opaque markers are temporarily attached prior to the scan enabling this refer-
ence position to be visible on the treatment planning scan. After the scan, the markers
are removed and the laser mark positions are generally made permanent by tattooing
their position onto the patient’s skin. These markers facilitate accurate reproducing of
patient position at the time of radiotherapy treatment.

Other imaging
Limited soft-​tissue contrast on CT can make it a less reliable modality for
detecting small volumes of seroma between clips and for distinguishing surgically
induced densities from normal glandular breast tissue. In the light of this, tumour
bed delineation using CT alone has been compared with tumour bed delinea-
tion based on fused MR plus CT (MRCT) in patients with 6–​12 titanium tumour
bed clips(28). MRCT defined larger TB volumes by identifying additional seroma,
haemorrhage, and haematoma, but satisfactory coverage of the MRCT-​defined
clinical target volume (CTV) was achieved in all cases by tangential external-​
beam PBI fields designed to cover the CT-​defined CTV, suggesting that the add-
ition of MRI to CT/​clip-​b ased tumour bed delineation is unnecessary. The use
of MRI for RT-​planning is likely to become more common, however, in the con-
text of neo-​adjuvant PBI in which the tumour is still in-​situ. In this setting, the
use of MRI leads to reduced target volumes and interobserver variability when
compared to CT.
Ultrasound can be used to help define the tumour bed for electron boosts but relies
on the presence of seroma in the tumour bed which is becoming less common with
full-​thickness closure of the tumour bed at surgery.

5.3.3 Contouring
It is strongly recommended that the ESTRO guidelines are followed for delineation of
the whole breast, chest wall, and regional nodes(29). Full atlases showing slice by slice
contouring are available for download with these guidelines. Therefore, only a brief
summary for contouring is given in Figs 5.9 and 5.10.

Whole breast clinical target volume

This includes all the glandular breast tissue, the peripheries of which can be difficult
to define using CT imaging alone. Placing a radio-​opaque wire around the palpable
breast tissue at the time of CT scanning can be helpful to help distinguish breast from
subcutaneous fat (see Fig. 5.11). A guide to whole breast CTV borders is as follows, but
must be adapted to individual anatomy:
◆ Anterior border: This extends to 5 mm from the skin surface unless there is skin
involvement as a result of the malignant process in which case the anterior border
is at the skin surface.
◆ Posterior border: This is at the deep fascia, anterior to the surface of the pectoral
muscles, unless this is breached by tumour.
92 Breast radiotherapy

(a)

(b)

(c)

(d)

Fig. 5.9  ESTRO consensus guideline on target volume delineation for elective radiation
therapy of early stage breast cancer.
Reproduced with permission from Radiotherapy and Oncology, 114, Offersen, B. V. et al., 'ESTRO
consensus guideline on target volume delineation for elective radiation therapy of early stage breast
cancer', pp. 3-​10 (2015). Copyright © 2014 Elsevier Ireland Ltd. All rights reserved.
 Pre-planning procedures 93

(a)

(b)

Fig. 5.10  ESTRO consensus

guideline on target volume
delineation for elective radiation
therapy of early stage breast
cancer, version 1.1.
Reproduced with permission from
Radiotherapy and Oncology, 114,
Offersen, B. V. et al., 'ESTRO consensus
guideline on target volume delineation
for elective radiation therapy of early
stage breast cancer', pp. 3-​10 (2015).
Copyright © 2014 Elsevier Ireland Ltd.
All rights reserved.

◆ Superior border: This is usually defined as being at the level of the inferior aspect of
the medial head of the clavicle.
◆ Inferior border: This is determined by lowest CT slice where breast tissue can still
be seen.
◆ Medial border: This is usually defined as the ipsilateral lateral edge of the sternum.
◆ Lateral border: This may be difficult to identify, but may be helped by breast contour
or in some cases, the thoracic vessels that lie laterally to the breast. There is rarely
breast tissue posterior to the anterior edge of latissimus dorsi.

Chest wall CTV
The borders of the chest wall CTV are usually determined by the position of the
contralateral breast in combination with the border definitions for whole breast CTV.
The skin including the mastectomy scar should be included in the CTV in patients in
94 Breast radiotherapy

Fig. 5.11  X-​ray CT scan

through centre of breast
with patient lying in
treatment position (both
arms elevated above
head). Radio-​opaque
skin markers are visible
on anterior midline and
right/​left mid-​axillary lines.
The glandular tissue and
surrounding anatomy of
the breast are clearly seen.

whom the skin is involved and sometimes in those in whom the anterior margin is
close or involved. However, it is standard practice to ensure that the entire length of the
mastectomy scar is encompassed in the tangential fields in all patients.

Boost CTV
The boost CTV is based on the tumour bed, which is usually marked with implanted
surgical clips and also includes any associated seroma and other surgical changes. It is
important to also refer to pre-​operative imaging and to surgical annotations and dia-
grams. Occasionally, it can be challenging to identify the tumour bed when oncoplastic
surgery has been carried out, even when the cavity has been marked with clips prior to
manipulation of breast tissue. Contouring with the surgeon present can be helpful in
this situation. A margin is then added to the tumour bed to create the boost CTV. This is
typically 5–​10 mm and is limited to 5 mm from the skin surface and by the deep fascia.

Partial breast CTV
The partial breast CTV is also based on the tumour bed, which is then typically grown
by 10–​15 mm, and again limited to 5 mm from the skin surface and by the deep fascia.
(Fig. 5.12)

Fig. 5.12  This shows the partial breast CTV in axial, sagittal and coronal views as per
IMPORT LOW trial: the red volume is the tumour bed and the blue volume is the partial
breast CTV.
 Radiotherapy planning 95

Regional nodal areas CTV

These should be outlined with reference to the ESTRO breast nodal contouring guide-
lines that were developed by consensus based on anatomy of the vessels. A summary of
CT landmarks to be used in conjunction with the atlases is shown in Table 5.4.

Organs at risk
The contralateral breast, ipsilateral, and contralateral lungs and heart (to the level of
the pulmonary arch) are usually contoured so that dose to these OARs can be reported.

5.3.4  Planning target volumes

Planning target volumes (PTVs) are created by adding margins to the CTV to account
for set-​up error and patient movement (including breast swelling and respiratory mo-
tion). A number of studies have used electronic portal imaging devices to quantify the
extent of positional errors and patient movement. Data from the Royal Marsden show
that changes in breast volume (peaking early in the treatment course at ~105% of the
initial whole breast volume) and set-​up errors are adequately encompassed by a 5 mm
margin(30) (mean set-​up error = 1.2 mm, standard deviation = 2.5 mm, CTV to PTV
margin = (2.5×1.2) + (0.7×2.5) = 4.75 mm). A further margin of 5 mm to allow for
the effects of respiratory motion has been deemed adequate in the majority of patients
generating a total CTV–​PTV margin of 10 mm.
More recent studies demonstrate that implanted tumour bed markers and spe-
cific on-​treatment imaging protocols reduce the set-​up errors A  study specifically
investigating the effect of imaging frequency and soft tissue motion on PTV margin
size for boost or partial breast RT(31) has demonstrated that no imaging correction
requires a CTV–​PTV margin of 10 mm; an extended no action level (eNAL) imaging
protocol with 5 images during 15 fractions requires a margin of 6 mm; and daily on-​
line imaging requires a margin of less than 5 mm.
In practice, a 5 mm PTV margin tends to be used when a steep gradient is required
and the volume needs to be limited and online imaging is recommended, such as in
the context of the simultaneous integrated boost. In most other situations, a 10 mm
CTV–​PTV margin is still recommended. The CTV–​PTV margin should be tailored to
the institution’s measured set-​up errors, the type of radiotherapy technique, the veri-
fication protocol, and individual patient characteristics such as size and compliance.

5.4 Radiotherapy planning
5.4.1 Whole breast
Standard beam geometry for treating the whole breast consists of two opposed tangen-
tial beams covering the entire breast. The posterior beam edges are coincident to avoid
divergence into lung tissue. The collimators should be rotated and/​or MLCs used to
minimize the lung volume irradiated and to shield the heart if it is in the field, whilst
not infringing on the region surrounding the tumour bed. The anterior field border is
extended into air by around 2 cm to allow for set-​up uncertainties and possible breast
swelling during treatment (Fig. 5.13).
Table 5.4  ESTRO delineation guidelines for the CTV of lymph node regions, breast and postmastectomy thoracic wall for elective irradiation in breast cancer
Borders Axilla level 1 Axilla level 2 Axilla level 3 Lymph node level 4 Internal Interpectoral nodes Residual Thoracic wall
per CTVn_​L1 CTVn_​L2 CTVn_​L3 CTVn_​L4 mammary chain CTVn_​ breast CTVp_​thoracic
region CTVn_​IMN interpectoralis CTVp_​breast wall
Cranial Medial: 5 mm Includes the Includes the Includes the cranial Caudal limit of Includes the cranial Upper border of Guided by
cranial to the axillary cranial extent cranial extent of extent of the CTVn_​L4 extent of the axillary palpable/​ visible palpable/​
vein Lateral: max up of the axillary the subclavian subclavian artery artery (i.e. 5 mm breast tissue; visible signs;
to 1 cm below the artery (i.e. 5 mm artery (i.e. 5 mm (i.e. 5 mm cranial of cranial of axillary maximally up to if appropriate
edge of the humeral cranial of axillary cranial of subclavian vein) vein) the inferior edge guided by the
head, 5 mm around vein) subclavian of the sternoclavi­ contralateral
the axillary vein vein) cular joint breast; maximally
up to the inferior
edge of the
sterno-​clavicular
joint

Caudal To the level of rib The caudal 5 mm caudal to Includes the Cranial side of Level 2's caudal Most caudal CT Guided by
4–​5, taking border of the the subclavian subclavian vein with the 4th rib (in limit slice with visible palpable/​
also into account minor pectoral vein. If 5 mm margin, thus selected cases breast visible signs;
the visible effects of muscle. appropriate: connecting to the 5th rib, see text) if appropriate
the sentinel lymph If appropriate: top of surgical cranial border of guided by the
node biopsy top of surgical ALND CTVn_​IMN contralateral
ALND breast
Ventral Pectoralis major & Minor pectoral Major pectoral Sternocleidomastoid Ventral limit of Major pectoral 5 mm under skin 5 mm under skin
minor muscles muscle muscle muscle, dorsal edge the vascular area muscle surface surface
of the clavicle
Dorsal Cranially up to Up to 5 mm Up to 5 mm Pleura Pleura Minor pectoral Major pectoral Major pectoral
the thoraco-​dorsal dorsal of axillary dorsal of muscle muscle or costae muscle or costae
vessels, and more vein or to costae subclavian vein and intercostal and intercostal
caudally up to and intercostal or to costae muscles where no muscles where
an imaginary muscles and intercostal muscle no muscle
line between the muscles
anterior edge of
the latissimus dorsi
muscle and the
intercostal muscles
Medial Level 2, the Medial edge of Junction of Including the 5 mm from Medial edge of Lateral to Guided by
interpectoral level minor pectoral subclavian jugular vein without the internal minor pectoral the medial palpable/​
and the thoracic muscle and internal margin; excluding mammary vein muscle perforating visible signs;
wall jugular veins the thyroid gland (artery in cranial mammarian if appropriate
-​>level 4 and the common part up to and vessels; guided by the
carotid artery including first maximally to contralateral
intercostal the edge of the breast
space) sternal bone
Lateral Cranially up to Lateral edge of Medial side Includes the anterior 5 mm from Lateral edge of Lateral breast Guided by
an imaginary line minor pectoral of the minor scalene muscles the internal minor pectoral fold; anterior palpable/​
between the major muscle pectoral muscle and connects to the mammary vein muscle to the lateral visible signs;
pectoral and deltoid medial border of (artery in cranial thoracic artery if appropriate
muscles, and CTVn_​L3 part up to and guided by the
further caudal up including first contralateral
to a line between intercostal breast. Usually
the major pectoral space) anterior to the
and latissimus dorsi mid-​axillary line
muscles

ALND = axillary lymph node dissection.

Reproduced with permission from Offersen, B.V. et al. ESTRO consensus guideline on target volume delineation for elective radiation therapy of early stage breast cancer. Radiotherapy and Oncology,
114(1): pp.3–10. Copyright © 2014 Elsevier Ireland Ltd. All rights reserved. DOI: https://doi.org/10.1016/j.radonc.2014.11.030.
98 Breast radiotherapy

Fig. 5.13  An example of
tangential field geometry.
The medial field is shown
in red and the lateral field
in green.

The dose reference point, set to receive 100% of the prescribed dose, should be lo-
cated in the centre of the breast. It is typically on a point midway between the skin sur-
face and the underlying chest wall and midway between the upper and lower borders
of the tangential fields. The resulting dose distribution can be homogenized, either
by using wedges or by using simple field-​in-​field IMRT. Occasionally, more complex
IMRT may be used for patients with challenging anatomy such as pectus excavatum.

5.4.2 Chest wall
Chest wall radiotherapy utilizes the same opposed beam field arrangement as for
whole breast radiotherapy. These treatments can be challenging to plan due to shallow
chest wall tissue in unreconstructed patients, which makes it difficult to define an ap-
propriate normalization point. Field-​in-​field IMRT is often used to minimize dose
inhomogeneity. In patients who had skin involvement at presentation or involved an-
terior margins on pathology, bolus (tissue equivalent material) is placed over the chest
wall or reconstructed breast to ensure full dose to the skin surface.

5.4.3 Partial breast
The partial breast beam geometry used in the IMPORT LOW trial was based on the
same opposed beam field arrangement as described for whole breast radiotherapy, but
with a reduced superior-​inferior field length to cover the partial breast PTV with a 5–​
6 mm margin to field edge. As the irradiated volume is reduced, fewer inhomogeneity
corrections are needed. Other more conformal partial breast techniques can be used,
wherein the beam directions are not limited to tangents. These techniques usually re-
sult in smaller irradiated volumes but have the disadvantage of beams exiting through
larger volumes of normal tissue.

5.4.4  Tumour bed boost

Tumour bed boosts are delivered either sequentially (most commonly using either
photons or electrons), or concurrently with whole breast treatment as a simultaneous
integrated boost (SIB). A tumour bed photon boost can be delivered using a conformal
or an IMRT plan with 3–​5 static beams or 1–​2 arcs, or by using mini-​tangential fields
(as described in section 5.4.3 for partial breast irradiation).
 Radiotherapy planning 99

Sequential photon boosts are ‘stand-​alone’ treatments. The dose distribution is

homogenized and the plan is normalized to the standard ICRU reference point in the
middle of the boost volume. It is possible to produce a composite plan of the whole
breast treatment plus boost, although the resulting dosimetry is often not ideal and it
can be challenging to modify the separate plans independently in order to produce a
homogenous composite plan.
Where the tumour bed boost is delivered as part of an SIB, the composite plan is
normalized to deliver the total dose of both the whole breast and the boost prescription to
a point central in the boost volume. The median dose to the remaining whole breast tissue,
outside the SIB region, should be normalized to the prescribed whole breast dose.(Fig. 5.14)
Electron boost planning usually requires information from the planning CT to guide
the beam placement. For example, the electron depth can be estimated by measuring
the distance from the skin to the deepest tumour bed clip/​chest wall. If the patient
cannot be treated in the same position as whole breast radiotherapy, then ultrasound
can sometimes be used to help localization if a tumour bed seroma is present. The
dose should be prescribed to the 100% isodose at dmax (maximum dose) aiming to
cover the tumour bed with the 90% isodose. Where electron boosts are marked on-​set
using information from CT scans, the electron field size will need to be larger (7–​8 cm
circle) in order to minimize the risk of a geographical miss. Where electron boosts are
virtually simulated using CT imaging, smaller treatment volumes can be selected. The
electron energy will depend on the depth of the deepest clip from the skin surface and
is usually somewhere between 6 MeV and 12 MeV.

5.4.5  Regional nodal areas

Regional nodal treatments vary in complexity from field-​based techniques to rota-
tional IMRT. The most appropriate method should be selected to meet the clinical
requirements. For all methods, doses to normal tissue will need to be carefully con-
sidered and assessed.
Historically, in some patients, the breast and nodal areas were positioned, planned,
and treated as entirely separate entities. For a few patients this led to overlap of
doses from breast and nodal fields leading to serious clinical issues such as brachial
plexopathy. In modern practice, the patient’s position is maintained between breast
and nodal treatments. A number of techniques maybe used, all of which avoid overlap
of treatment areas.

Field-​based technique
In this technique, an anterior nodal field is placed superiorly to the whole breast field.
The anterior field may be angled to avoid exiting through OARs (e.g. the spine, tra-
chea, oesophagus). The field is either directly matched to the superior edge of the tan-
gents, or offset superiorly with a gap between fields. The most common method at the
time of writing is the direct match, using a single isocentre for all of the beams. The
isocentre is placed cranio-​caudally at the join between the nodal field and tangen-
tial breast fields. All fields are half-​beam-​blocked along the junction, resulting in a
non-​divergent match-​line. Consideration is needed regarding field lengths and breast
coverage. If longer fields are required, the same match can be achieved with symmetric
beams with separate isocentres for each of the treatment areas. In this case, the floor,
100 Breast radiotherapy

Fig. 5.14  Radiotherapy planning CT image showing a dose distribution in the sagittal

plane for a patient being treated with Whole Breast radiotherapy and an IMRT
Simultaneous Integrated boost. Treatment dose is 48Gy in 15 fractions to the boost area
and 38Gy to the surrounding whole breast.

gantry, and collimator angles need to be adjusted from the standard set-​up to produce
a non-​divergent join (see Fig. 5.15).
Posterior boosts to the axillary nodes can be used for patients with wide anterior to
posterior skin separations in order to increase the dose delivered to the deepest axil-
lary lymph nodes. In order to minimize uncertainty, the resultant dose distributions
should always be calculated and visualized on a treatment planning computer.
The normalization point for the breast treatment remains in the middle of the breast
as standard. The abutting nodal field should be normalized separately, generally to the
point of dose maximum. There are limited options for modifying the basic nodal dose
distribution such as adjusting MLCs, adding wedges, altering the beam energy, or pos-
sibly normalizing to a different depth. Care should be taken to ensure point doses de-
livered do not exceed set limits, e.g. 110% of the prescribed dose. This technique may
have limitations in nodal dose coverage for some patients, which is usually a result of
body habitus (Figs 5.15 and 5.16).
 Radiotherapy planning 101

(a)

(b)

Fig. 5.15  (a, b) Anterior

& lateral surface rendered
treatment planning views,
showing the field placement
for a patient being treated with
(c) whole breast radiotherapy (blue)
and a directly matched nodal
field (red), treated with separate
isocentres (green). For all
beams the treatment unit floor,
collimator and gantry have been
rotated to produce the exact
beam match at the join.
(c) Planning software
generated stylised observer eye
view of the same treatment
showing the nodal beam’s
placement, couch and gantry
rotation.
102 Breast radiotherapy

Fig. 5.16  Radiotherapy planning CT image showing the dose distribution in the sagittal
plane for a patient being treated with whole breast radiotherapy (blue) and a directly
matched 10MV anterior nodal beam (red). Treatment dose = 40Gy in 15 fractions.

Volume-​based technique
Rotational IMRT has the advantage of being able to treat the nodal areas and the breast
as one volume. This avoids field matching and improves nodal coverage compared with
field-​based techniques for some patients. However, this is a more complex technique
with a requirement for longer planning time and greater expertise, advanced dosim-
etry checks, and subsequent image-​guided radiotherapy. Implementation of class so-
lutions and increased experience in planning and treating with IMRT is likely to make
the process more efficient. This technique will, however, produce a larger volume of
normal tissue irradiated to low doses, the so-​called ‘low-​dose bath’ and the extent to
which this will increase the incidence of late RT effects is not yet known.(Fig. 5.17)
The challenge for the treating team is to select the most appropriate technique based
on a combination of risk of recurrence with the need for optimal nodal coverage, vs
risk of late normal tissue side effects. It is likely that the patient’s body habitus will also
play a part in technique selection .
 Radiotherapy planning 103

Fig. 5.17  Example of a dose distribution from a rotational method (tomotherapy)

showing the good high-​dose coverage, and the low-​dose bath.

Internal mammary chain irradiation

The methods already described can also be used to incorporate the IMC in the treated
volume. Simply widening the standard breast tangential beams to cover the IMC is
commonly referred to as ‘wide tangents’. This will increase the doses to the heart and
lungs and therefore heart (and lung)-​sparing techniques, such as breath-​hold, are re-
commended where possible. If used, rotational IMRT should be optimized to spare
the heart, lungs, and contralateral breast and this technique can also be combined
with breath-​hold techniques to reduce OAR doses even further. Helical IMRT, such as
tomotherapy, is not compatible with breath-​hold as the treatment beam times are long
and the unit cannot be interrupted and re-​started efficiently. As with all radiotherapy
treatments, the risks and benefits need to be carefully evaluated on an individual pa-
tient basis. Matched electron-​photon techniques have been used historically but are
not recommended due to unacceptable uncertainty in dose distribution (Fig. 5.18).

5.4.6  Dose constraints and objectives

Ideally, the dose distribution should meet ICRU 50 criteria (95% to 107%) for dose
variation throughout the treatment volumes. The majority of whole breast plans re-
quire some form of compensation to achieve this. Inhomogeneity corrections in the
planning calculations must be used to correctly account for the presence of lung.
Oncologists may need to engage with dosimetrists/​physicists to indicate areas of im-
portance when compromises are needed between PTV coverage and dose to OARs
(see Fig. 5.19 and Tables 5.5 and 5.6).

Calculation algorithms
3D treatment planning calculations should be performed with corrections for tissue
heterogeneity. Treatment planning systems vary in how they model side scatter and
104 Breast radiotherapy

(a)

(b)

Fig. 5.18  Radiotherapy planning CT images showing dose distribution using a) volumetric

modulated arc radiotherapy (VMAT) and b) wide tangents demonstrating increased internal
mammary chain lymph node clinical target volume (cyan volume) coverage with VMAT.

inhomogeneities, especially for higher energy beams. This is particularly relevant to

the calculation of doses in chest wall irradiation, due to the large proportion of the
field including either lung or air. ‘Type-​b’ superposition-​convolution algorithms or
Monte Carlo algorithms should be used as they calculate lung and tumour doses more
accurately than older ‘type-​a’ algorithms.
 Radiotherapy planning 105

1.0

0.9

0.8

0.7
Norm. Volume

0.6

0.5

0.4

0.3

0.2

0.1

0.0
0 1000 2000 3000 4000 5000 6000
Dose (cGy)
Fig. 5.19  This is a typical dose–​volume histogram for the IMPORT HIGH trial test
arm 2. The red, green and yellow volumes show the 3 distinct PTV targets for the
breast: tumour bed, quadrant and whole breast. Heart—​the pink line shows the heart
DVH and illustrates that a very small volume of heart is irradiated. Lung—​the dark blue
shows the ipsilateral lung and the light blue shows the contralateral lung. The purple line
represents the contralateral breast.
Reprinted with permission from the International Journal of Radiation Oncology, Biology,
Physics, Volume 79, Ellen M. Donovan et al. ‘Planning with intensity-​modulated radiotherapy
and tomotherapy to modulate dose across breast to reflect recurrence risk (IMPORT High Trial),’
pp. 1064–​72. DOI: 10.1016/​j.ijrobp.2009.12.052, Copyright © 2011 Elsevier Ltd.

Choice of photon beam energy

Most breast radiotherapy is delivered using 4–​6 MV photons. In some patients, to
achieve better homogeneity, it may be preferable to treat with higher energies, e.g. 10
MV, in some or all beams. However, with increasing energy, the build-​up depth at the
patient’s surface increases. The balance between increasing the relative dose delivered
at depth (thereby reducing hot spots medially and laterally in breast tissue), and pro-
viding sufficient isodose coverage at the patient surface, will depend on the clinical
scenario. Supraclavicular nodal regions often benefit from energies greater than 6 MV
to allow optimal coverage.
There is a variety of methods for optimizing a treatment plan’s dose distribution. The
choice depends partly on the clinical aim (simple whole breast dose homogenization
or producing steep dose gradients to limit dose to critical organs), and partly on the
available technology and expertise. As such, a whole spectrum of techniques can be
used including various forms and combinations of inverse and forward planned IMRT.
These techniques will be described starting with the simplest and ending with the most
complex (Fig. 5.20).
106 Breast radiotherapy

Table 5.5  Example of dose objectives and constraints for breast + boost (as per IMPORT
HIGH trial)
Breast + boost
Organ Objective Constraint

Ipsilateral lung V18 Gy <10% V18 Gy <15%

Mean <6 Gy
Contralateral lung V2.5 Gy <3% V2.5 Gy <15%
Mean <1 Gy
Heart (left-​sided tumour) V13 Gy <2% V13 Gy <10%
Mean <3 Gy
Heart (right-​sided tumour) V5 Gy <6%
Mean <1.7 Gy
Contralateral breast Mean <0.5 Gy Mean <1.5 Gy

Field-​in-​field IMRT is commonly used and it is usually all that is needed to bring
the dose inhomogeneity within acceptable limits for the majority of tangential breast
plans. It is relatively simple to both plan and treat. The number of additional boost
fields is generally limited to a few low-​weighted apertures, sometimes using higher
energies than the main treatment beams. When the planner manually determines the
sub-​fields, this is known as forward planning. When the computer determines the sub-​
fields, this is known as inverse planning. (Fig. 5.21)
Static IMRT, or step and shoot IMRT, is a form of inverse planning. In its simplest
form the planner specifies the gantry angles and the maximum beam aperture, then
the planning computer will carry out an iterative process of adding sub-​sets of field
shapes to meet pre-​determined dosimetric aims. The linear accelerator will deliver
the shapes as unique beamlets, pausing the radiation between the sub-​fields and
then moving the MLCs to the next sub-​field position before re-​starting the radi-
ation. Dynamic IMRT is a more complex form of inverse planned IMRT, moving the

Table 5.6  Example of organ-​at-​risk dose constraints for breast and loco-​

regional lymph node radiotherapy (including the internal mammary chain)
Breast/​chest wall plus loco-​regional lymph nodes
(including internal mammary chain)
Organ Constraint

Ipsilateral lung V17 Gy <35%

Heart V17 Gy <10%
Mean <6 Gy (objective)
Contralateral breast Mean <3.5 Gy
 Radiotherapy planning 107

Typical dose distribution in the

sagittal plane with wedges only:
orange and red >105%

Typical dose distribution with

IMRT in the sagittal plane

Fig. 5.20  Dose distributions for whole breast using two different methods of
compensation based on wedge only (left) and full dose compensation using MLC (right).
Regions coloured red or orange indicate doses > 110% or 105%, respectively, of the
prescribed dose. Regions in blue or green indicate areas where the dose is > 95% but <
105% of prescribed dose.
Reprinted with permission from Radiotherapy and Oncology, Volume 82, Ellen Donovan et al.
Randomised trial of standard 2D radiotherapy (RT) versus intensity modulated radiotherapy (IMRT) in
patients prescribed breast radiotherapy, pp. 254–​64. DOI: 10.1016/​j.radonc.2006.12.008, Copyright
© 2007 Elsevier Ltd.

MLCs and modulating the treatment beam’s dose rate while the radiation is being
delivered.
Rotational IMRT utilizes the most complex level of inverse planned IMRT. Based on
dynamic IMRT, the planning computer calculates the optimal treatment plan allowing
the beam to irradiate the patient from a whole range of gantry angles. During treatment
the linear accelerator is allowed to rotate, the MLCs will change position, and the radi-
ation beam intensity is modulated. The treatment may be delivered in one or two arcs.
The rotation can be through a full 360° rotation, e.g. helical delivery, whereby individual
MLCs block or open as the beam moves across target and non-​target tissue. For standard
linear accelerators, the gantry rotation is generally limited to about 180° approximately
between the angles of standard tangential beams, or maybe further limited to a small
range of angles either side of the standard tangential beam angles in a so-​called bow-​tie
field arrangement. Bow-​tie field arrangements have the advantage of delivering lower
dose to the underlying tissue compared to fuller arcs. Rotational IMRT is used to treat
complex planning target volumes whilst attempting to spare dose to OARs. However,
the methods will add a low dose bath to surrounding and underlying tissues and OARs,
the long-​term effect of which on normal tissues is unknown. Examples include treat-
ment of patients with pectus excavatum and deep-​seated multiple nodal volumes that
are inadequately covered using standard field-​based methods.
108 Breast radiotherapy

Fig. 5.21  Example of a
beam’s eye view (BEV)
isodose contouring
approach to IMRT for the
intact breast. The 105%
isodose surface in a BEV
window is shown in red
with some of the MLC
leaves positioned over part
of this area.

5.4.7 Dose prescription
Dose prescription for breast and loco-​regional radiotherapy
The UK standard regimen is 40 Gy in 15 daily fractions of 2.67 Gy as used in the
START-​B trial(2) and is recommended for whole breast, partial breast, chest wall fol-
lowing mastectomy, and nodal radiotherapy. Standard boost dose regimens after com-
plete tumour excision are strongly influenced by the EORTC trial (n = 5318), which
tested 16 Gy in eight fractions against no boost after complete microscopic excision of
primary tumour(4). It is entirely reasonable to hypofractionate the boost schedule, for
example, a five-​fraction regimen of 2.67 Gy, which is approximately equivalent to 16
Gy in eight fractions assuming an alpha/​beta value of 3.0 Gy. Other hypofractionation
regimens such as 12 Gy in four daily fractions are also acceptable.
Some frail patients struggle with 15 fractions delivered over 3 weeks. In such cases,
it is reasonable to offer a hypofractionated regimen such as that used in the FAST
trial(32). The regimen with equivalent normal tissue effects to 50 Gy in 25 fractions
is 28.5 Gy in 5 fractions of 5.7 Gy once weekly over 5 weeks. If a boost is required, a
single extra fraction in week 6 can be used. The possible risks and benefits of radio-
therapy must be carefully evaluated in patients with active collagen vascular disease,
especially systemic sclerosis and avoidance may be preferable in patients at very low
risk of recurrence.

Dose prescription for palliative radiotherapy

The same principles apply in breast cancer as elsewhere in oncology. Palliative radio-
therapy, by definition, aims to relieve symptoms, not to eradicate cancer. However, the
 Treatment & verification 109

long natural history of disease in a proportion of patients presenting with advanced

loco-​regional disease or skeletal metastases can complicate decision-​making. Lifetime
palliation of fungating primary tumours or regional lymph nodes often requires more
than 20 Gy in five fractions in the absence of life-​threatening metastatic disease.
Without randomized evidence on which to base reliable treatment recommendations,
a reasonable rule is to plan a potentially curative dose to a PTV encompassing all
known disease (GTV) plus a 1-​cm margin. A curative dose is not precisely defined,
but implies a dose in excess of 60 Gy in 2.0 Gy equivalent fractions. Although there are
few published data on hypofractionated external beam therapy in this setting, either a
regimen incorporating the boost is employed or a once-​weekly fraction of 6.0 Gy to a
total dose of 36 Gy in 6 weeks will palliate symptoms arising from well-​circumscribed
lesions in the breast or chest wall if the patient is frail by virtue of comorbidity or ex-
treme age. Where nodal irradiation is required, the brachial plexus dose must be care-
fully considered when deciding on an appropriate regimen.

5.5  Treatment & verification

5.5.1  Set-​up and on treatment imaging
The patient’s set-​up and position obtained at the planning scan must be accurately
reproduced and maintained on the treatment unit. On-​treatment verification must
identify and correct for both geometric set-​up issues and patient specific issues such
as shape changes. Patient set-​up verification is broadly split into two main areas: the
initial manual set-​up and subsequent confirmatory imaging checks.
Patient treatment positions should be stable and reproducible. Department-​wide
protocols and standardized equipment, supported by on-​going staff training, will op-
timize patient set-​up accuracy and reproducibility. Individual patient records should
have clear descriptions of the set-​up used, preferably with photographs of any unusual
positioning. Details of set-​up variables should include:  couch location holes, breast
board inclination, the use and position of any other supports, such as lower limb fix-
ation devices, and patient-​specific details, such as whether the patient was scanned
with breath-​hold. Camera-​based optical tracking systems and even electromagnetic
tracking systems can also be used to aid set-​up reproducibility and to monitor patient
set-​ups, particularly for patient in breath-​hold.
On-​treatment verification imaging is used extensively to confirm the patient’s
treatment geometric accuracy. Following national guidelines (e.g. On Target(33)), it is
possible to limit the frequency of imaging checks, balancing the need to ensure the
accuracy of treatment delivery whilst minimizing the patient’s concomitant imaging
dose and appointment time. A typical departmental verification protocol for standard
tangential breast treatments is to image directly prior to the delivery of the first three
treatment fractions, making gross error corrections each day, following which a more
rigorous offline assessment is made and any systematic correction is applied to sub-
sequent treatment fractions. Should corrections be necessary, the imaging assessment
process should be repeated until the set-​up is deemed stable. A decision is then made
either for on-​going daily imaging or to reduce to once a week imaging to monitor
110 Breast radiotherapy

for shape changes or set-​up trends. Daily imaging should be used for patients with
complex treatment or set-​ups where accuracy is crucial (e.g. simultaneous integrated
tumour bed boosts).
Each department should review their set-​up uncertainties for each technique and
use these to set achievable tolerances. These tolerances should be incorporated in
margin calculations determining and influencing the planning target margins used.
With all methods of on-​treatment imaging, carefully constructed protocols with ap-
propriate personnel training should be established and maintained to determine ap-
propriate imaging protocols, assessing and recording any additional patient doses and
determining appropriate match criteria and action levels.

5.5.2 Imaging methodology
Planar imaging
kV or MV planar images obtained are either manually or automatically compared with
the planning system’s reference images. Planning systems will provide electronic refer-
ence images depicting the expected anatomical projection in the form of DRRs (digi-
tally reconstructed radiographs) or DCRs (digital composite radiograph). DRRs and
DCRs are essentially the same: a composite projection of the expected tissue density
the treatment beam will pass through. DCRs can be windowed to better show soft
tissue anatomy.

3D on set imaging
On set KV CT (e.g. cone-​beam CT) or MV CT verification can also be used. This is
machine dependent using the treatment unit’s on board imaging system to obtain and
reconstruct a series of images from multiple angles around the patient. These can be
used to produce adequate quality CT images, which are then co-​registered manually
or automatically to reference planning CT images.
If the patient has a photon boost and has implanted fiducial markers in the tumour
bed, any of the methods described can be used to verify the position of both the whole
breast and the tumour bed. After the initial treatments in which the whole breast is im-
aged, it may be appropriate to continue with daily imaging to a smaller volume centred
around the tumour bed clips in order to minimize imaging dose.

5.5.3  Dose delivery checking

Patient treatment plans can look acceptable on the planning computer, but they may
not be representative of what is actually delivered to the patient. Treatment plans
made up of a few simple photon or electron fields can be checked by simple com-
parison with independent dosimetry tables or check programs. For the majority of
plans this combined with some form of in vivo dosimetry, such as diodes, is all that
is required.
When field arrangements become complex, i.e. highly modulated, with moving gan-
tries and rapidly changing shapes, the ability of the linear accelerator to accurately deliver
the treatment as planned will need to be carefully assessed. An independent software
 Patient assessment and follow-up 111

program can be used to check the theoretical agreement of the expected dosimetry with
that on the plan but the beam delivery should also be checked on a treatment unit for
at least a selection of patients’ treatments. To achieve this, the patient’s treatment plan is
superimposed and re-​calculated onto a CT image of a phantom into which measurement
devices can be placed or from which the exit dose acquired at an electronic portal imager
can be predicted. A comparison between the expected dose and the delivered dose can
then be assessed. More direct measurements on patients themselves require the use of in
vivo dosimeters, such as diodes. These can be used with fixed gantry deliveries, including
IMRT beams, but not for rotational therapy. Advances in in vivo dosimetry utilizing the
on-​board electronic imaging devices are being introduced into clinical use to compare
actual delivered doses to those planned. In time, these will become more common and
should work with simple, complex, fixed, and rotational treatments.

5.6  Patient assessment and follow-​up

Patients requiring radiotherapy for breast cancer are assessed whilst on treatment and
afterwards. The assessment during therapy and in the following weeks is primarily
about the acute reaction whereas after this it is focused on possibility of recurrence and
late-​reacting normal tissue complications. Acute toxicity is assessed by the CTCAE or
RTOG scales. The acute skin reaction, as an early reacting normal tissue with a high α/​
β ratio, is related to total dose and with hypofractionation schedules will be less than
with the previous standard regimen of 50 Gy in 25 fractions as shown in the FAST
study with trial comparison arms of 30/​28.5 Gy in five fractions over 5 weeks(32). The
FAST-​Forward study documents clearly that the peak acute reaction occurs 1–​2 weeks
after completion of therapy, that the majority of reactions do not exceed grade 1 tox-
icity, and that it is unusual to see more than grade 2 toxicity in non-​boost patients(34)
across all three arms of the study (40 Gy in 15 fractions over 3 weeks, 26 Gy in five
fractions over 1 week and 27 Gy in five fractions over 1 week).
Loco-​regional recurrence rates are low and steadily reducing but remain an ex-
tremely important patient end-​point. Follow-​up aims to pick this up at an early sal-
vageable stage where possible, whether this is by imaging, clinician-​led, or patient-​led
via education and easy access back to the breast cancer multidisciplinary team. Follow-​
up also looks for normal tissue complications. This includes acute pneumonitis, not
often seen nowadays with lung doses assessed at the planning stage, and early onset
breast oedema. Longer follow-​up will assess changes in the irradiated breast such as
shrinkage, induration, telangiectasia, and lymphoedema. Irradiation of OARs can lead
to symptomatic rib fracture, lung fibrosis, ischaemic heart disease, lymphoedema of
the arm, and, most uncommonly, brachial plexopathy. Reduction of the incidence of
the long-​term effects is a focus of modern therapy and achieved by both technical im-
provements and the introduction of hypofractionation regimens. Long-​term follow-​
up is far less common in standard practice due to early discharge strategies, but is still
mandated for 10 years or more after treatment in most of the UK radiotherapy trials.
This data will continue to give us important insight into toxicities that continue to
evolve many years after radiotherapy.
112 Breast radiotherapy

5.6 Conclusion
There has been a dramatic change in the planning and delivery of breast radiotherapy
over the last decade, with 3D CT planning and varying levels of complexity of IMRT
now the standard of care. Breast fractionation research continues, with a 3-​week
schedule now firmly established as standard whilst the results of five-​fraction whole
breast regimens and hypofractionated simultaneous integrated boost techniques are
awaited.
Awareness of late cardiac morbidity has also increased following reports in the
literature and cardiac sparing techniques are now well established. Recently, there
has been a renaissance in use of nodal radiotherapy both for the axilla and IMC as
a result of new evidence and changes in surgical practice. As a result, many depart-
ments are working towards routine nodal contouring and implementing new nodal
techniques.
Falling local recurrence rates have highlighted the need to estimate the absolute
risks and benefits of breast radiotherapy for individual patients and to tailor treatment
accordingly. For example, emerging evidence suggests that selected women at low of
recurrence may benefit from partial breast radiotherapy and some at even lower risk
may be able to avoid radiotherapy completely.
The challenge is to ensure equity of access to the best quality breast radiotherapy
for all women, in order to achieve optimal local control and survival rates whilst
minimizing late normal tissue toxicity. This is best achieved with a multidiscip-
linary team of oncologists, radiographers, physicists, and surgeons, with the patient
being at the centre of decision making. Future progress also depends on implemen-
tation, recruitment, and long-​term follow-​up within well-​designed clinical trials
linked with translational research to advance the concept of ‘personalized’ radi-
ation therapy.

References
1. Darby S, McGale P, Correa C et al. Effect of radiotherapy after breast-​conserving surgery
on 10-​year recurrence and 15-​year breast cancer death: meta-​analysis of individual patient
data for 10,801 women in 17 randomised trials. Lancet 2011; 378:1707–​16.
2. Haviland JS, Owen JR, Dewar JA, et al. The UK Standardisation of Breast Radiotherapy
(START) trials of radiotherapy hypofractionation for treatment of early breast cancer: 10-​
year follow-​up results of two randomised controlled trials. Lancet Oncology 2013;
14:1086–​94.
3. Whelan TJ, Pignol JP, Levine MN, et al. Long-​term results of hypofractionated radiation
therapy for breast cancer. New England Journal of Medicine 2010; 362: 513–​20.
4. Bartelink H, Maingon P, Poortmans P, et al. Whole-​breast irradiation with or without a
boost for patients treated with breast-​conserving surgery for early breast cancer: 20-​year
follow-​up of a randomised phase 3 trial. Lancet Oncology 2015; 16:47–​56.
5. Hickey BE, Lehman M, Francis DP, See, AM. Partial breast irradiation for early breast
cancer. Cochrane Database Systematic Reviews 2016; 7:CD007077.
6. Coles CE, Griffin CL, Kirby AM, et al., IMPORT Trialists. Partial-​breast radiotherapy after
breast conservation surgery for patients with early breast cancer (UK IMPORT LOW trial):
 References 113

5-​year results from a multicentre, randomised, controlled, phase 3, non-​inferiority trial.

Lancet 2017; 390:1048–​60.
7. Kunkler IH, Williams LJ, Jack WJL, et al. Breast-​conserving surgery with or without
irradiation in women aged 65 years or older with early breast cancer (PRIME II): a
randomised controlled trial. Lancet Oncology 2015; 16:266–​73.
8. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG). Overview of the
randomized trials of radiotherapy in ductal carcinoma in situ of the breast. Journal of the
National Cancer Institute. Monographs. 2010: 162–​77.
9. Silverstein MJ. The University of Southern California/​Van Nuys prognostic index for
ductal carcinoma in situ of the breast. American Journal of Surgery 2003; 186:337–​43.
10. Goodwin A, Parker S, Ghersi D., Wilcken N. Post-​operative radiotherapy for ductal
carcinoma in situ of the breast. Cochrane Database Systematic Reviews 2013; 11:CD000563.
doi:10.1002/​14651858.CD000563.pub7
11. Solin LJ, Gray R, Hughes LL, et al. Surgical excision without radiation for ductal
carcinoma in situ of the breast: 12-​year results from the ECOG-​ACRIN E5194 Study.
Journal of Clinical Oncology 2015; 33:3938–​44.
12. EBCTCG (Early Breast Cancer Trialists' Collaborative Group), McGale P, Taylor C, et al.
Effect of radiotherapy after mastectomy and axillary surgery on 10-​year recurrence and 20-​
year breast cancer mortality: meta-​analysis of individual patient data for 8135 women in 22
randomised trials. Lancet 2014; 383:2127–​35.
13. Wallgren A, Bonetti M, Gelber RD, et al. Risk factors for locoregional recurrence among
breast cancer patients: results from International Breast Cancer Study Group Trials
I through VII. Journal of Clinical Oncology 2003; 21:1205–​13.
14. Coates AS, Winer EP, Goldhirsch A, et al. Tailoring therapies—​improving the
management of early breast cancer: St Gallen International Expert Consensus on the
Primary Therapy of Early Breast Cancer 2015. Annals of Oncology 2015; 26:1533–​46.
15. Gradishar W, Salerno, KE. NCCN Guidelines Update: Breast Cancer. Journal of the
National Comprehensive Cancer Network 2016; 14: 641–​4.
16. Mukesh MB, Duke S, Parashar D, et al. The Cambridge post-​mastectomy radiotherapy
(C-​PMRT) index: A practical tool for patient selection. Radiotherapy and Oncology 2014;
110:461–​6.
17. Whelan TJ, Olivotto IA, Parulekar WR, et al. Regional nodal irradiation in early-​stage
breast cancer. New England Journal of Medicine 2015; 373:307–​16.
18. Mamounas EP. Anderson SJ, Dignam JJ, et al. Predictors of locoregional recurrence after
neoadjuvant chemotherapy: results from combined analysis of National Surgical Adjuvant
Breast and Bowel Project B-​18 and B-​27. Journal of Clinical Oncology 2012; 30:3960–​6.
19. Poortmans PM, Collette S, Kirkove C, et al. Internal mammary and medial
supraclavicular irradiation in breast cancer. New England Journal of Medicine 2015;
373:317–​27.
20. Thorsen LB, Offersen BV, Danø H, et al. DBCG-​IMN: a population-​based cohort study
on the effect of internal mammary node irradiation in early node-​positive breast cancer.
Journal of Clinical Oncology 2016; 34:314–​320.
21. Darby SC, Ewertz M, McGale P,et al. Risk of ischemic heart disease in women after
radiotherapy for breast cancer. New England Journal of Medicine 2013; 368:987–​98.
22. Kaufman EL, Jacobson JS, Hershman DL, et al. Effect of breast cancer radiotherapy and
cigarette smoking on risk of second primary lung cancer. 2008; Journal of Clinical Oncology
26: 392–​8.
114 Breast radiotherapy

23. Hill-​Kayser CE, Harris EER, Hwang W-​T, Solin LJ. Twenty-​year incidence and patterns of
contralateral breast cancer after breast conservation treatment with radiation. International
Journal of Radiation Oncology, Biology, Physics 2006; 66: 1313–​19.
24. Kirova YM, De Rycke Y, Gambotti L, et al. Second malignancies after breast cancer: the
impact of different treatment modalities. British Journal of Cancer 2008; 98:870–​4.
25. DiSipio T, Rye S, Newman B, Hayes S. Incidence of unilateral arm lymphoedema after
breast cancer: a systematic review and meta-​analysis. Lancet Oncology 2013; 14:500–​15.
26. Wang, W. Purdie TG, Rahman M, et al. Rapid automated treatment planning process to
select breast cancer patients for active breathing control to achieve cardiac dose reduction.
International Journal of Radiation Oncology, Biology, Physics 2012; 82: 386–​93.
27. Nissen HD, Appelt AL. Improved heart, lung and target dose with deep inspiration breath
hold in a large clinical series of breast cancer patients. Radiotherapy and Oncology 2013;
106:28–​32.
28. Kirby AM, Yarnold JR, Evans PM. et al. Tumor bed delineation for partial breast and
breast boost radiotherapy planned in the prone position: what does MRI add to X-​ray CT
localization of titanium clips placed in the excision cavity wall? International Journal of
Radiation Oncology, Biology, Physics 2009; 74:1276–​82.
29. Offersen BV, Boersma LJ, Kirkove C,et al. ESTRO consensus guideline on target
volume delineation for elective radiation therapy of early stage breast cancer, version 1.1.
Radiotherapy and Oncology 2016;118:205–​8.
30. Hector CL, Webb S, Evans PM. The dosimetric consequences of inter-​fractional patient
movement on conventional and intensity-​modulated breast radiotherapy treatments.
Radiotherapy and Oncology 2000; 54:57–​64.
31. Harris EJ, Donovan EM, Coles CE, et al. How does imaging frequency and soft tissue
motion affect the PTV margin size in partial breast and boost radiotherapy? Radiotherapy
and Oncology 2012; 103:166–​71.
32. Agrawal RK, Alhasso A, Barrett-​Lee PJ, et al. First results of the randomised UK FAST
Trial of radiotherapy hypofractionation for treatment of early breast cancer (CRUKE/​04/​
015). Radiotherapy and Oncology 2011; 100:93–​100.
33. Royal College of Radiologists, Society and College of Radiographers, Institute of Physics
and Engineering in Medicine. On target: ensuring geometric accuracy in radiotherapy.
London, The Royal College of Radiologists, 2008.
34. Brunt AM, Wheatley D, Yarnold J, et al. Acute skin toxicity associated with a 1-​week
schedule of whole breast radiotherapy compared with a standard 3-​week regimen delivered
in the UK FAST-​Forward Trial. Radiotherapy and Oncology 2016; 120:114–​18.
 Chapter 6

Radiotherapy for thoracic tumours
Kevin Franks, Fiona McDonald,
and Gerard G Hanna

6.1 Lung cancer

6.1.1 Introduction
Lung cancer is the most common cause of cancer death in the UK, accounting for one
in five of all deaths. Unfortunately, despite treatment advances survival for all stages
remains poor with < 10% of patients alive at 5 years. More than 40,000 cases are diag-
nosed each year in the UK and though the incidence in males is declining, it has only
stabilized in females. This is because 90% of lung cancers are related to active or passive
smoking and the smoking rate in women peaked in the 1960s, 20 years later than for
men. Lung cancer is, therefore, becoming a disease of the elderly with the incidence
560 per 100,000 for men over the age of 80 (200 per 100,000 for men aged 60–​69) and
273 per 100,000 for women over 80 (140 per 100,000 for women aged 60–​69). Over
87% of patients diagnosed with lung cancer between 2006 and 2008 were over the age
of 60 years (http://​info.cancerresearchuk.org/​cancerstats/​types/​lung).
The increasing age and comorbidity of patients with lung cancer means that many
are medically unfit for surgery. Consequently, radical radiotherapy is playing an
increasing role in the curative management of lung cancer(1).
There are two main subtypes of lung cancer: small cell (SCLC) and non-​small cell
(NSCLC). However, the management is becoming increasingly similar so the general
principles of lung cancer radiotherapy will be discussed first, followed by the specific
features of management of the two pathological entities. Finally, this chapter will con-
sider radiotherapy for rarer thoracic malignancies, namely mesothelioma and thymic
carcinoma.

6.1.2  Assessment of patients with lung cancer

for radical radiotherapy
Two key factors are important in the assessment of the suitability of patients for radical
radiotherapy: comorbid illnesses and the stage/​configuration of the tumour and nodes.

Fitness for treatment
Age per se is not a contraindication for radical radiotherapy. When assessing a pa-
tient, it is important to look at their general well-​being and make an estimate of their
116 Radiotherapy for thoracic tumours

life expectancy: a fit 80-​year-​old may live another 10–​15 years whereas a 60-​year-​old
with significant ischaemic heart disease and chronic obstructive pulmonary disease
(COPD) may have a short life-​expectancy making the radical treatment of their lung
cancer futile. However, even patients with severe COPD have a similar or better prog-
nosis than untreated lung cancer. For example, data from the Californian Cancer
Registry showed that untreated patients with stage I non-​small lung cancer have a very
poor survival with a median survival of 9 months for all stage 1 patients, 13 months for
T1 N0 disease alone, and 14 months for those patients who refused surgical resection
with only a 16% 5-​year survival rate(2). In contrast, in a patient with a FEV1 (Forced
Expiratory Volume in 1 second) of 40%, an exercise tolerance of 150 metres, a body
mass index (BMI) of 23, and who gets dyspnoea on walking on level ground has a pre-
dicted 4-​year overall survival of 57% using the Body mass index, airflow Obstruction,
Dyspnea and Exercise capacity score (BODE)(3). Using the BODE scoring system and
considering the worst predicted prognosis (FEV< 35%, 6 minute walk distance < 50
metres, modified medical research council (MMRC) score 4, and BMI ≤ 21) the 4-​year
overall survival is estimated to be 18% which is a similar prognosis to untreated stage
I NSCLC. Thus the assessment of fitness for treatment should be objective and utilize
validated tools such as those suggested above.
The key comorbid diseases that can preclude the use of radical doses of radiotherapy
are COPD and pulmonary fibrosis. Relative contraindications include systemic scler-
osis and scleroderma though the evidence for increased toxicity is limited and, in
part, anecdotal(4). There appears to be less of a risk for other connective tissue dis-
orders (including systemic erythematosus lupus and discoid lupus). However, given
the absence of robust evidence, each individual case with a relative contraindication
to radiotherapy should be discussed by the multidisciplinary team weighing up all the
potential radical treatment options. Finally, before deciding on radiotherapy as a treat-
ment option, a careful discussion with the patient regarding the potential increased
risk of radiotherapy is essential. Classically FEV1, FVC (Forced Vital Capacity), and
DLCO (diffusing capacity for carbon monoxide or carbon monoxide transfer factor)
have been used to assess suitability for radical treatment but the lower threshold has
not been formally established and studies examining the association of these param-
eters with toxicity have demonstrated conflicting results(5). However, this being said,
the inclusion criteria for most clinical trials have a cut-​off for inclusion of an FEV1 and
DLCO of 40% of predicted values and treating with pulmonary function below this
threshold requires careful discussion of the risks with the patient.
When assessing a patient’s suitability for radical radiotherapy other factors that
should be taken into account include weight loss and decline in performance status,
both of which are associated with an inferior prognosis.

Disease stage and configuration

All patients undergoing radical treatment should be staged with a contrast-​enhanced
computed tomography (CT) scan of chest and upper abdomen, and those with
NSCLC, a positron emission tomography (PET)-​CT scan. For NSCLC, CT scans have
a sensitivity of 60% and specificity of 80% for nodal involvement, compared with
84% and 89%, respectively for PET(6). PET also has a higher sensitivity (93%) and
 Lung cancer 117

specificity (96%) for distant disease(7). However, it performs poorly for the detection of
brain metastases (sensitivity 60%) so a contrast-​enhanced CT or magnetic resonance
imaging (MRI) brain scan should be considered, especially for patients with medias-
tinal lymphadenopathy(7). If the PET-​CT scan demonstrates positive lymph nodes(s),
it is good practice to confirm this pathologically with EBUS/​EUS (endobronchial/​
endoluminal ultrasound) and/​or mediastinoscopy(8). Consideration should be given
to biopsy of solitary metastatic deposits identified on PET-​CT scan.
The location of the primary and involved nodes is often the principal determinant
of the suitability for radical radiotherapy. For example, a right upper lobe lesion with
a right paratracheal (R4) node may be easily encompassed within a suitable volume,
whereas a left lower lobe tumour with an identical node may not be, due to the doses
to the organs at risk, exceeding normal tissue tolerances (particularly the lung). See
Tables 6.1 and 6.2.

6.1.3  Patient
positioning for radical radiotherapy
for lung cancer
◆ The patient should be planned and treated in the same position: supine with arms
above their head unless it is a superior/​apical tumour or the patient is unable to lift
their arms above their head. A variety of immobilization devices are available, but
none has proved superior to a bar for the hands to hold and support under the el-
bows (e.g. T-​bar and wing-​board). For comfort, a knee roll should be used.
◆ When patients are to be treated with stereotactic ablative radiotherapy (SABR, also
referred to as SBRT) the duration of each treatment may be considerably longer
than that for conventional treatment. There is clear evidence that patient comfort
and minimizing treatment times (< 30 minutes) are the most important factors in
minimizing intrafraction movement(9). Some advocate specialized devices, for ex-
ample, a body vacuum mould, whereas others suggest that standard systems with
time taken to ensure patient comfort are equally reproducible(10). If on assessment,
the tumour movements exceed 1 cm then abdominal compression or breath-​hold
techniques, may be considered to reduce diaphragmatic movement. An alternative
is to use respiratory gating and/​or tracking if available.
◆ For superior sulcus tumours, the patient’s arms should be placed by their sides, and
an immobilization shell covering neck and shoulders used to maintain a consistent
shoulder position.
◆ The patient should have a CT scan performed on a flat couch top with ≤3-​mm
slice thickness covering the entire lungs (cricoid to L2 vertebra). IV contrast, unless
contra-​indicated, should be used to aid target delineation in node positive disease
or when the tumour is close to a vessel. For all radical lung cancer patients ideally,
and definitely for peripheral mobile tumours, a respiratory correlated or 4DCT
planning scan should be acquired to enable the tumour motion to be assessed (see
below).
◆ Ideally, the treatment isocentre should be fixed and tattooed at the time of the plan-
ning CT scan. This minimizes systematic errors which result from isocentre shifts.
118 Radiotherapy for thoracic tumours

Table 6.1  TNM Descriptors (8th edition)(26)

TX Primary tumour cannot be assessed, or tumour proven by the presence of
malignant cells in sputum or bronchial washings but not visualized by imaging
or bronchoscopy

T0 No evidence of primary tumour

Tis Carcinoma in situ
T1 Tumour ≤ 3 cm, surrounded by lung or visceral pleura with no evidence of
invasion of main bronchus
T1mi Minimally invasive adenocarcinoma
T1a ≤ 1 cm
T1b > 1 cm ≤ 2 cm
T1c > 2 cm ≤ 3 cm
T2 Tumour > 3 cm but ≤ 5 cm, or tumour with any of the following features:
◆ Involves main bronchus regardless of distance to carina, but without
involving carina
◆ Invades visceral pleura
◆ Associated with atelectasis or obstructive pneumonitis that extends to the
hilar region, either involving part of the lung or the entire lung
T2a > 3 cm ≤ 4 cm
T2b > 4 cm ≤ 5 cm
T3 Tumour > 5 cm ≤ 7 cm, or associated with separate tumour nodule(s) in the
same lobe as the primary, or tumour that directly invades any of the following:
◆ Chest wall (including superor sulcus tumours)
◆ Phrenic nerve
◆ Parietal pericardium
T4 Tumour > 7cm, or associated with separate tumour nodule(s) in a different
ipsilateral lobe to that of the primary, or tumour that directly invades any of
the following:
◆ Diaphragm
◆ Mediastinum
◆ Heart/​Great vessels
◆ Carina/​Trachea
◆ Recurrent laryngeal nerve
◆ Oesophagus
◆ Vertebral body
NX Regional lymph nodes cannot be assessed
N0 No regional lymph node metastases
N1 Metastases in ipsilateral peribronchial and/​or ipsilateral hilar node(s) and intra-​
pulmonary lymph node(s), including involvement by direct invasion
N2 Metastases in ipsilateral mediastinal and/​or subcarinal lymph node(s)
N3 Metastases in contralateral mediastinal, contralateral hilar, ipsilateral or
contralateral scalene or supra-​clavicular lymph node(s)
 Lung cancer 119

Table 6.1 Continued
M0 No distant metastasis
M1
M1a Separate tumour nodule(s) in a contralateral lobe, or tumour with pleural or
pericarial nodule(s) or malignant pleural or pericardial effusion.
M1b Single extra-​thoracic metastasis in a single organ
M1c Multiple extra-​thoracic metastases in one or several organs

Reproduced with permission from Goldstraw P, Chansky K, Crowley J, et al. ‘The IASLC Lung Cancer Staging
Project: Proposals for Revision of the TNM Stage Groupings in the Forthcoming (Eighth) Edition of the TNM
Classification for Lung Cancer.’ Journal of Thoracic Oncology2016 Jan;11(1):39-​51. DOI: https://​doi.org/​
10.1016/​j.jtho.2015.09.009 , Copyright © 2016 Elsevier Ltd.

However, estimating the isocentre position can be difficult if there is disease pro-
gression between the diagnostic and planning scans.
◆ If fixing the isocentre at the time of scanning is not possible, radio-​opaque markers
should be placed on reference tattoos and an isocentre shift calculated at time of
planning.
◆ Lateral tattoos should be placed on stable areas of the lower chest to minimize errors
caused by lateral rotation.

6.1.4 Tumour motion
Tumour movement, resulting in a geographic miss, is a major concern in the treat-
ment of lung cancers, particularly for lesions close to the diaphragm, which can move
as much as 5 cm in the cranio–​caudal direction, though movement over 2 cm is rare.
Several techniques to try and reduce the effect of this have been developed(11):
1. Deep inspiration breath hold (DIBH):  patient is coached to hold their breath in
maximum inspiration during treatment.
2. Active breathing control (ABC): the patient is connected to a breathing apparatus
with a control valve that immobilizes breathing motion at a fixed level.
3. Gating: the beam is turned on and off, synchronized with respiratory movement
to deliver treatment during only part of the respiratory cycle. Gating is challenging
in patients with poor pulmonary function as they have unpredictable breathing
patterns.
4. Slow CT scanning:  the planning CT is acquired at a much slower rotation than
usual (4 seconds per revolution) to produce a gross tumour volume (GTV), which
includes all tumour positions during the respiratory cycle. This produces a blurred
GTV.
5. Respiratory correlated or 4D scanning:  a multislice CT scan is acquired with
respiratory monitoring and the images sorted according to the respiratory cycle
to produce a series of GTVs that demonstrate the tumour position at different
times.
120 Radiotherapy for thoracic tumours

Table 6.2  Stage Grouping (8th edition)(26)

Stage T N M
Occult TX N0 M0
carcinoma

0 Tis N0 M0
IA1 T1mi—​T1a N0 M0
IA2 T1b N0 M0
IA3 T1c N0 M0
IB T2a N0 M0
IIA T2b N0 M0
IIB T1a-​T2b N1 M0
T3 N0 M0
IIIA T1a-​T2B N2 M0
T3-​T4 N1 M0
T4 N0 M0
IIIB T1a-​T2b N3 M0
T3-​T4 N2 M0
IIIC T3-​T4 N3 M0
IVA Any T Any N M1A-​M1B
IVB Any T Any N M1C

Reproduced with permission from Goldstraw P, Chansky K, Crowley J, et al.

‘The IASLC Lung Cancer Staging Project: Proposals for Revision of the TNM
Stage Groupings in the Forthcoming (Eighth) Edition of the TNM Classification
for Lung Cancer.’ Journal of Thoracic Oncology 2016 Jan;11(1):39–51.
DOI: https://​doi.org/​10.1016/​j.jtho.2015.09.009, Copyright © 2016
Elsevier Ltd.

Respiratory correlated or 4DCT is now considered the gold standard for curative
thoracic radiotherapy, particularly for stage I tumours, with many commercial ‘pack-
ages’ available. The patient is scanned during multiple (e.g. 6–​12) breathing cycles and
then the GTV contoured on: (a) the maximum intensity projection (MIP) of the 4D
dataset, (b) the MIP (see Fig. 6.1 a), the maximum inspiratory and expiratory datasets,
or (c) all ten phases of the 4DCT scan(12).

6.1.5 Target delineation
The delineation of the GTV, clinical target volume (CTV), and planning target volume
(PTV) varies with the pathological subtypes and location of the tumour and is de-
scribed in the appropriate later sections.
 Lung cancer 121

(a)

(b)

Fig. 6.1  (a) Four-​dimensional target delineation: ITV drawn on MIP in blue and

PTV (ITV + 5 mm) in red. (b) Stereotactic plan using two intensity modulated arcs.
Image courtesy of Corrine Faivre-​Finn and Gareth Webster.

6.1.6  Organs at risk

Lungs
The lungs should be contoured from apex to bases (usually using automatic contouring
sequence) and then combined.
1. V20: this is the percentage volume of normal lung receiving more than 20 Gy. In
the original publication by Graham et  al. the volume of normal lung within the
PTV was removed from the calculation(13); however, the lung QUANTEC paper
recommends subtracting the GTV from the volume of normal lung for assess-
ment of lung constraints and this is more commonly used in clinical practice. The
differing pneumonitis scoring systems and methods used in studies make drawing
conclusions about the optimal percentage threshold challenging, but most recom-
mend that the V20 should be kept ≤35%(14) to keep the risk of clinically significant
122 Radiotherapy for thoracic tumours

pneumonitis to < 20%. Although these thresholds are frequently also applied to the
UK fraction schedules (e.g. 55 Gy, in 20 fractions), it should be appreciated the data
sources for these recommendations used 2 Gy per fraction, and without chemo-
therapy. Other dose–​volume thresholds have been explored, for example, V13,
V10, and V5, and high values of these may be associated with increased pneumon-
itis and/​or late fibrosis; however, a definitive threshold has not been established.
2. Mean lung dose (MLD):  this may be a better predictor of the risk of symptom-
atic pneumonitis than V20(14). A standard maximum threshold of 20 Gy is recom-
mended(10) and for hypofractionated regimes a lower V20 may be considered (e.g.
18 Gy for patients receiving 55 Gy in 20 fractions).
3. When treating with SABR, the dose recommendations from the UK SABR consor-
tium guidelines are V20 of < 10%(15).
Other factors for pulmonary toxicity:
◆ The tumour location appears important with lower lobe lesions having a greater risk
of pneumonitis.
◆ Concurrent chemotherapy with platinum and vinca alkaloids or etoposide does
not appear to increase the risk of pneumonitis(16). Concurrent chemotherapy with
Gemcitabine is not recommended.
◆ The presence of idiopathic pulmonary fibrosis is anecdotally linked with increased
risk of toxicity but there are no published series from which to calculate the extent
of the risk.

Spinal cord
The spinal cord should be defined throughout the thoracic region and for a few centi-
metres above and below if non-​coplanar techniques are employed. A set up margin,
similar to that applied to the CTV (3–​5 mm), should be added to produce a planning
at risk volume.
Sometimes the location of the tumour, especially if there is invasion of the vertebral
body and into the foramina, can preclude a radical radiotherapy dose. It is estimated
that using 2 Gy per fraction a total dose of 50 Gy is associated with a 0.2% risk of myel-
opathy(17). However, this dose should be reduced with the use of concurrent chemo-
therapy (usually by 10% to 46 Gy) or hypofractionated (44 Gy at 2.75 Gy per fraction)
or accelerated hyperfractionated schedules where repair maybe incomplete(18).
When treating with SABR the maximum acceptable point dose (0.1 cc) to the spinal
cord is 21.9 Gy when giving the primary 54 Gy in three fractions and 30 Gy for the 55
Gy in five-​fraction regimens(12).

Oesophagus
The oesophagus should be contoured, including all mucosal and muscular layers, from
cricoid to oesophago-​gastric junction. For patients receiving radiotherapy alone the
risk of clinically significant oesophagitis is low (< 5%), but with the addition of con-
current chemotherapy this is quadrupled(16). As most cases of oesophagitis settle with
supportive measures it is rarely a dose-​limiting toxicity but establishing the risk is im-
portant for obtaining informed consent and managing patients during treatment (e.g.
 Lung cancer 123

early assessment by dietician and monitoring of the patient’s weight). Some patients
do, however, go on to develop late oesophageal stricture. There is no single dosimetric
parameter on which to base predictions, though a V70 > 20%, V50 > 40%, and a V35 >
50% appear to predict for grade 2 + oesophagitis(19).
When treating with SABR a maximum acceptable cumulative dose to a volume of
0.5 cc of the oesophagus is 25.2 Gy when giving the primary 54 Gy in three fractions
and 34 Gy for 55 Gy in five fractions(15).

Heart
The heart should be contoured from the superior aspect of the left atrium and ex-
tended to the apex where it touches the diaphragm. Radiation-​induced cardiac damage
can either be acute (pericarditis) or long-​term (ischaemic heart disease, myocardial
dysfunction, and valvular disease). The majority of the cardiac tolerance data comes
from fit patients with either Hodgkin lymphoma or breast cancer when radiotherapy
is often given in conjunction with anthracyclines. The challenge for patients with lung
cancer is they often have pre-​existing cardiovascular disease due to smoking and there
is little good data on this specific group of patients. The CONVERT trial protocol re-
commended V66 < 30% and D50 of < 33 Gy.
For SABR the maximum dose to 0.5 cc of the heart is 26 Gy and 29 Gy for the three-​
and five-​fraction schedules, respectively(15).
The following organs at risk (OARs) should be considered when treating patients
with SABR.
Brachial plexus: all major trunks of the ipsilateral plexus should be contoured using the
subclavian and axillary vessels as a surrogate for the position of the plexus, extending
from the bifurcation of the brachiocephalic trunk to where the neurovascular struc-
tures cross the second rib. The maximum dose to a volume of 0.5 cc is 26 Gy for the
three-​fraction regimen and 29 Gy for the five-​fraction regimen(15).
Proximal trachea: contours should begin 10 cm superior to the superior extent of
the tumour or 5 cm above the carina, whichever is more superior. It should continue
to the superior aspect of the proximal bronchial tree. The maximum dose to 0.5 cc of
the proximal trachea is 32 Gy and 35 Gy respectively for the three-​and five-​fraction
regimens(15).
Proximal bronchial tree: this will include the inferior 2 cm of the trachea, the carina,
right and left main bronchi, right and left upper lobe bronchi, right middle lobe bron-
chus, lingular bronchus, and right and left lower lobe bronchi. The dose constraints for
the proximal bronchial tree are identical to those for the trachea.
Others: if non-​coplanar beams are used, other organs such as the liver and small
bowel may be irradiated and, therefore, will need to be taken into consideration. The
body contour should always be outlined and efforts made to ensure beam entry points
on the skin do not overlap resulting in unacceptable skin toxicity.

6.1.7 Planning

◆ The treatment is usually planned using photons of ≤ 10 MV (usually 6 or 8 MV);

beam energies greater than this result in a larger penumbra and, consequently,
124 Radiotherapy for thoracic tumours

greater normal tissue irradiation. There is also a greater secondary build-​up effect
when higher beam energies are used.
◆ As lungs have a density of around 30% of that of normal soft tissue, the treatment
should be planned with a tissue inhomogeneity correction. A planning algorithm
which takes into account lateral electron transport should be used (so-​called type B
models)(10).
◆ To minimize normal tissue irradiation, conformal shaping should be used on all
fields.
◆ When delivering conventionally fractionated radiotherapy, a simple three-​field
technique with an anterior and two ipsilateral oblique fields may often suffice, but
a more complex field arrangement may be required for large, central lesions, where
achieving a dose distribution in the ICRU range of 95–​107% can be difficult. In
these cases, additional coplanar beams, or non-​coplanar beams exiting into medi-
astinal structures, may be required. An alternative for difficult cases is to use a two-​
phase technique starting with parallel-​opposed anterior and posterior fields and
then a second phase with a three-​field technique. However, this technique increases
the dose delivered to the oesophagus and spinal cord, which may increase toxicity,
especially if delivered with concurrent chemotherapy.
◆ Intensity-​modulated radiotherapy (IMRT) has been shown, in a number of plan-
ning studies and single institution series, to reduce the doses to OARs (particularly
lung and spinal cord) enabling large target volumes (e.g. stage III with contralateral
nodes) to be treated within OAR dose constraints and smaller tumours to have the
treatment dose escalated (see Fig. 6.2)(20). Potential concerns exist about the ‘low-​
dose bath’ delivered to the lungs (e.g. V5, V10, and V13) and the impact on late
pulmonary toxicity. At present there are no recommended thresholds for low dose
radiation to the lungs. Concerns regarding tumour motion and the risk of interplay
have not led to inferior tumour control outcomes in any published results of IMRT
treatment to the lung. In a recent analysis of the RTOG 0617 study, the use of IMRT
was associated with lower rates of pneumonitis, lower heart doses, and comparable
tumour control(21).
◆ When planning SABR in order to achieve adequate target coverage with high con-
formity and a rapid isotropic dose fall-​off beyond the PTV, at least seven beams
are required. Alternatively, intensity-​modulated arc-​based therapy can be used (see
Fig. 6.2b)(22). These may be coplanar or non-​coplanar. For SABR it is recommended
that 95% of the PTV should receive the prescribed dose, with 99% of PTV receiving
at least 90% of this dose. Due to the nature of the planning, there is inevitably less
homogeneity in the dose distribution. Therefore, it is recommended that the dose
maximum within the PTV should be a minimum of 110% and a maximum of 140%
of the prescribed dose(12).

6.1.8  Implementation on the treatment machine

If the treatment has not been planned around a fixed isocentre, an isocentre shift must
be performed, either at time of a plan check or on the machine prior to treatment.
 Lung cancer 125

(a) (b)

(c) Dose Volume Histogram

1.0

0.9

0.8

0.7
Norm. Volume

0.6

0.5

0.4

0.3

0.2

0.1
0.0
0 1000 2000 3000 4000 5000 6000 7000
Dose (cGy)
Fig. 6.2  (a) T4N3M0 NSCLC conventional 3D-​CRT V20 = 38.5%. (b) IMRT plan
V20 = 33.3%. (c) Comparison DVH: improved PTV coverage with IMRT. Key: solid
line = IMRT, dotted line = 3D-​CRT, pale blue = PTV, green = lungs.
Image courtesy of Corrine Faivre-​Finn and Gareth Webster.

This shift should be verified by comparison of the portal images with digitally recon-
structed radiographs (DRRs) derived from the planning CT scan. Patients receiving
curative intent lung cancer radiotherapy should not have their treatment prolonged
beyond the intended duration and should be treated as a category 1 patient as per the
RCR Gaps policy(23).
126 Radiotherapy for thoracic tumours

6.1.9  Treatment
verification (as per National
Radiotherapy Implementation Group Report:
Image Guided Radiotherapy Guidance for 
Implementation and Use)
◆ For radical lung treatments, the treatment position should be verified against the
original planning CT scan comparing the CT set directly with volumetric 3-​D
imaging, e.g. cone-​beam CT (CBCT)(24).
◆ For conventionally fractionated schedules, imaging should be acquired on a daily
basis for at least the first 3 days to ensure no systematic errors have occurred and
then at least weekly to ensure no changes in the patient position have occurred(25).
◆ Traditionally imaging comparison has looked at the position of the carina, vertebral
bodies, sternum, and chest wall, but with the advent of CBCT it is feasible to com-
pare the position of the target lesion.
◆ If daily CBCT imaging is performed, the cumulative radiation dose should be
considered.
◆ For SABR, daily pretreatment imaging and on-​line patient position correction
should be performed. CBCT is the most commonly used system, though KV
tracking systems (e.g. EXAC TRAC) are also used when available. The images are
matched initially to bony anatomy to identify major inaccuracies and to assess ro-
tation. A soft tissue match is then performed in which the average tumour image
on the CBCT is manually registered to ensure the tumour localizes well within the
internal target volume (ITV)/​PTV defined on the planning CT. If the match is out-
side a 2–​3 mm tolerance a shift is applied. Further imaging during and at the end
of treatment may also be performed to ensure patient stability and provide further
information on the adequacy of treatment margins.

6.1.10  Radical treatment of non-​small cell lung cancer

Stage I and II NSCLC
Radical radiotherapy is indicated when the patient is medically inoperable or declines
surgery. There is currently no evidence for the use of chemoradiation in patients with
stage I–​II NSCLC. Patients can be treated using conventional techniques or SABR.
Patients with T1N0M0 or T2/​3N0M0 disease where the tumour is < 5  cm may be
candidates for SABR. The lesion must be outside a 2-​cm radius from the main airways
and proximal bronchial tree. This is to minimize the risk of stenosis or perforation of
the main airways. SABR aims to improve clinical outcomes in early stage lung cancer
by delivering higher biologically effective doses (BED > 100 Gy) than are possible
with conventional radiotherapy. It does this by using multiple beams to deliver high
doses to a carefully defined target with steep dose gradients beyond the target. There
is now a considerable body of evidence supporting SABR as superior to conventional
radiotherapy in terms of local control and survival in patients with inoperable disease
in historical cohorts(27). However, there is one small published randomized control
phase II trial, the Stereotactic Precision And Conventional Radiotherapy Evaluation
(SPACE) trial, that compared conventionally fractionated radiotherapy (70 Gy in 35
 Lung cancer 127

fractions) with SABR (45 Gy in 3 fractions)(28). This trial showed no difference in local
control but fewer side effects with SABR.
In addition, the results of the CHISEL trial (NCT01014130) of 60–​66 Gy in 30–​33
fraction +/​-​carboplatin/​taxol vs SABR (54 Gy in 3 fractions) showed improved local
control with SABR. SABR was also associated with an overall survival benefit com-
pared to conventionally fractionated radiotherapy (+/​-​chemotherapy)(29). Whether
SABR could be an alternative to surgery in early stage peripheral stage I NSCLC is a key
question, particularly in patients considered higher-​risk for surgery. The ROSEL and
STARS trials both failed to recruit though a joint analysis of the trials was published in
2015(30).This analysis was limited by the small sample size and short follow-​up schedule
but suggested that SABR could be an alternative for operable stage I NSCLC. Current
trials are trying to address this question and include SABRTooth (NCT02629458),
STABLEMATES (NCT02468024), and VALOR (NCT02984761) trials.

Stage III NSCLC
Radical radiotherapy, often combined with concurrent or sequential chemotherapy,
is the principal potentially curative treatment for stage III NSCLC. A  recent meta-​
analysis has confirmed an improved outcome (absolute survival benefit 4.5% at
5 years) with concurrent chemoradiation, but at the cost of increased toxicity, particu-
larly oesophageal(16).
The routine use of preoperative chemoradiation for stage III NSCLC is controversial;
though advocated by some, it has not been shown to improve overall survival when
compared with chemoradiation alone. Analysis of the data from the International
Association for the Study of Lung Cancer (IASLC) database has suggested that patients
with single-​station N2 disease have similar survival outcomes to patients with mul-
tiple N1 nodal involvement. This has resulted in a growing argument that there may
be a subgroup of patients with single level N2 disease, resectable with lobectomy, that
might benefit from combined modality treatment with surgery plus chemotherapy ±
radiotherapy. There is, however, no randomized data to support this at present.
There are, however, some phase II data, which demonstrated good local control re-
sults from the use of preoperative chemoradiation in superior sulcus lesions(31,32).

Stage IV NSCLC due to a solitary brain metastasis

Some patients with a solitary brain metastasis (confirmed on MRI as solitary) can derive
long-​term benefit if the metastasis is either resected or treated with radiosurgery(33, 34).
For such patients, if the extracranial disease is otherwise treatable in to a curative dose
within normal thoracic tissue constraints, radical radiotherapy or chemoradiotherapy
should be considered. For good performance status, patients with one to three brain me-
tastases and/​or < 20 cc of intracranial disease (confirmed on MRI) and radically treatable
extra-​cranial disease, discussion at a neurosurgical multidisciplinary team is advised.

Treatment volume and definition

◆ The treatment volume consists of the primary lesion and all involved nodes; by
limiting the treatment volume to known disease, a higher radiation dose can be de-
livered with minimal toxicity(10).
128 Radiotherapy for thoracic tumours

◆ Historically elective nodal irradiation (ENI) is no longer recommended as the dose

that can be delivered safely is limited to around 40 Gy and several studies using
involved field radiotherapy have demonstrated that few patients (< 5%) have an iso-
lated mediastinal relapse.
◆ The GTV consists of the primary lesion and all involved nodes based on PET/​CT
and mediastinal pathological staging (EBUS and/​or mediastinoscopy).
◆ The primary lesion should be contoured on lung windows (W1600, L-​600) and the
nodes on mediastinal windows (W400, L20) of the planning CT scan(10).
◆ Where PET/​CT planning scan has been used, the PET images should be used to
identify the location of the tumour and the CT used to identify the tumour edge
(IAEA)(35).
◆ A CTV margin of 5 mm to the GTV is commonly used though larger margins (up
to 9  mm) can be considered for adenocarcimonas(36). The CTV can be manually
edited where microscopic invasion is unlikely (e.g. next to a vertebral body where
there is no radiological evidence of bone invasion).
◆ The definition of the ITV will depend on the scanning technique used. If conven-
tional scanning is used then a margin needs to be added to account for tumour
movement. This is usually of the order 1–​2 cm, more in the cranio-​caudal direction.
If 4D scanning is used then the ITV will be generated using the MIP, end inspiratory
and expiratory positions, or composite of the multiple positions throughout the
breathing cycle(12).
◆ To produce the PTV, a margin should be added to the ITV to account for set-​up
variability and should not be manually edited. This margin will vary between de-
partments. It is usually around 5–​10 mm. If daily pretreatment imaging and on-​
line correction is performed, this margin can be reduced (CBCT or orthogonal
images).
◆ For SABR, the GTV is usually defined on the 4D CT scan using the MIP dataset,
the maximum inspiratory and expiratory datasets, or all phases of the 4D CT scan.
By convention, there is no CTV expansion (i.e. microscopic spread is not taken into
account). This approach is supported by the high levels of local control in the pub-
lished series. The ITV is, therefore, derived from the union of the GTV delineations.
To form the PTV, the ITV is uniformly expanded by 3–​5 mm. If tracking is used
for SABR delivery, e.g. on a Cyberknife platform, the GTV is expanded directly
to a PTV.
Dose prescription
◆ There is evidence from several studies of a dose response in NSCLC, but the optimal
schedule has yet to be established. Though total dose is important in the radical
treatment of NSCLC, so is overall treatment time due to accelerated repopula-
tion and accelerated schedules are associated with improved outcomes compared
to conventional fractionated schedules(37,38). The RTOG 0617 randomized phase
III trial comparing 60 Gy to 74 Gy at 2 Gy per fraction in patients with stage III
 Lung cancer 129

NSCLC reported inferior outcome with the high-​dose arm(39). Further studies of
novel dose intensification strategies in stage III NSCLC are ongoing including use of
individualized isotoxic schedules, dose boosts based on functional imaging, novel
radiosensitizers, and immunotherapy.
◆ For fit patients (WHO PFS 0–​1) with stage III disease concurrent chemoradiation is
considered the standard of care:
o 60–​66 Gy to isocentre in 30–​33 daily over 6–​7 weeks or 55Gy in 20 fractions
over 4 weeks with four cycles of cisplatin-​based chemotherapy (usually with
vinorelbine or etoposide), starting with the first or second cycle (depending pri-
marily on logistics)(40).
o Some patients not quite fit enough for concurrent treatment or large volumes
may benefit from two to four cycles of neo-​adjuvant platinum based chemo-
therapy followed by sequential radiotherapy with 60–​66 Gy to isocentre in 30–​
33 daily over 6–​7 weeks or 55Gy in 20 fractions over 4 weeks(40).
◆ For patients with stage I and II NSCLC not treated with SABR, particularly due to
central location:
o 60–​66 Gy to isocentre in 30–​33 daily fractions over 6–​7 weeks
o 55 Gy to isocentre in 20 daily fractions over 4 weeks
o 54 Gy to isocentre in 36 thrice-​daily fractions (minimum 6-​hour gap) over
12 days (CHART)(38).
◆ Patients with stage I lesions ≤ 5 cm in diameter and outside a 2-​cm radius from the
main airways and proximal bronchial tree and who are medically inoperable, may
benefit from SABR. The optimal schedule has not been established; however, a risk-​
adapted approach is used based on proximity to OARs.
i. 54 Gy in three fractions on alternate days over 5–​8 days (minimum interval
40 hours) with 95% of the PTV receiving this dose and 99% receiving >
48.6 Gy
ii. Lesions which are in contact with the thoracic wall, mediastinum or heart: 55
Gy in five fractions on alternate days over 10–​14 days (minimum interval 40
hours) with 95% of the PTV receiving this dose and 99% receiving > 49.5 Gy
◆ Decision about the number of fractions in the SABR schedule is based on the loca-
tion of the lesion in relation to OAR with an eight-​fraction schedule being recom-
mended for lesions within 2 cm of the mediastinal structures or brachial plexus,
a five-​fraction schedule being recommended for lesions near the chest wall and a
three-​fraction schedule being recommended for lesions away from OAR. The is re-
commended definition of central lesion: a tumour within 2 cm in all directions of
any mediastinal critical structure, including the bronchial tree, oesophagus, heart,
brachial plexus, major vessels, spinal cord, phrenic nerve, and recurrent laryngeal
nerve(41). Treatment of central lung lesions with SABR is still considered experi-
mental by many and it is recommended that treatment of central lesions should
only be considered within the context of a clinical trial.
130 Radiotherapy for thoracic tumours

Postoperative radiotherapy
A number of meta-​analyses of the trials of postoperative radiotherapy (PORT) have
been published demonstrating that the routine use of PORT may have a detrimental
effect on survival, particularly those with early stage (N0 or N1) disease(42). Whether
in the modern 3D conformal radiotherapy era, PORT would be equally harmful is
unknown. Those patients with mediastinal nodes (N2 or N3) had improved local
control but no change in survival. A recent meta-​analysis suggested a small survival
advantage with PORT, however, prospective randomized trials are needed to verify
this(43). Therefore, current recommendations are to consider PORT for patients with
either an incomplete resection and/​or mediastinal nodal involvement when there has
been incomplete surgical staging. The current LUNGART trial (NCT00410683) is re-​
examining the role of PORT in patients with N2 disease following R0 resection and
complete staging of the mediastinum.
Pulmonary function tests should be repeated postoperatively to ensure the patient
has sufficient reserve to tolerate PORT. The OAR constraints must be strictly adhered
too during the planning process.

Treatment volume and definitions

◆ The treatment volume should be defined after discussion with the operating sur-
geon to accurately establish the area at increased risk of recurrence. The option of a
second operation to complete the resection should also be discussed.
◆ Ideally areas of uncertain surgical clearance should be marked with surgical clips.
If a nodal group has been found to be involved, this region should be encompassed
within the target volume.
◆ There should not be, by definition, a GTV unless there is gross residual disease.
The CTV therefore consists of the bronchial stump, the region of the involved
nodes, and the ipsilateral hilar, subcarinal (level 7), and paratracheal (level
4) nodes.
◆ A margin of around 5 mm should be added to account for mediastinal movement
and then a further margin applied for set-​up inaccuracy according to departmental
guidelines, typically 5 mm.

Dose prescription
◆ There is no proven benefit from the use of concurrent chemotherapy with PORT.
◆ The role of adjuvant chemotherapy, and how it should be scheduled with PORT,
remains controversial but usually chemotherapy precedes the radiotherapy.
◆ The radiotherapy dose depends on the nature of the residual disease; macroscopic
disease should be treated with full dose radical radiotherapy, but the dose may be
reduced for microscopic only disease.
◆ Gross residual disease:
o 64–​66 Gy to isocentre in 32–​33 daily fractions over 6–​7 weeks
o 55 Gy to isocentre in 20 fractions over 4 weeks.
 Lung cancer 131

◆ Microscopic residual disease:

o 50–​54 Gy to isocentre in 25–​30 daily fractions over 5 weeks using 6–​8-​MV
photons
o 50 Gy to isocentre in 20 daily fractions over 4 weeks using 6–​8-​MV photons.

6.1.11  Radical treatment for small cell lung cancer

Small cell tumours account for only around 15% of all lung cancers and the propor-
tion is dropping. Though the current recommendation is to use the standard tumour,
node, metastasis (TNM) staging, the terminology ‘limited stage’ (can be encom-
passed with a radical radiotherapy volume) and ‘extensive stage’ is still widely used as
it describes the potential for cure. Only about a third of patients present with limited
stage disease.

Stages I–​III (limited stage) SCLC

High-​dose thoracic radiotherapy (either delivered concurrently or sequentially with
chemotherapy) has been demonstrated to improve the 3-​year survival of limited stage
SCLC by 5%.
◆ Early concurrent radiotherapy delivered with the first or second cycle of cisplatin-​
based chemotherapy (cycle 3 at the latest(44)) appears to have a small survival ad-
vantage when compared to radiotherapy delivered later, provided that both the
chemotherapy and radiotherapy are delivered with the optimal dose intensity.
Modifications of either negate the benefit.
◆ If early chemoradiation, which has enhanced toxicity, is to be used then it is im-
perative that the patient should be fully staged to ensure the disease is indeed
truly ‘limited’. Patients should have CT scan of chest and upper abdomen and a
CT (or MRI) brain scan. The usefulness of PET and bone scans remains unproven.

Treatment volume and definition

◆ The original studies of thoracic radiotherapy in limited stage SCLC treated the
primary, ipsilateral hilum, bilateral mediastinal, and, often, supraclavicular fossa
nodes. This was achieved by using parallel-​opposed anterior-​posterior fields with
the dose limited to 40 Gy in 15 fractions(45).
◆ With the introduction of 3D conformal radiation therapy (3D-​CRT) there has been
a trend to limit the treatment volume to just the primary and involved nodes as
this enables dose escalation. Which technique is superior is unknown, as they have
never been compared in a clinical trial.
◆ If the patient has received chemotherapy, a small-​randomized trial suggested that
the post-​chemotherapy volume should be used(46).
◆ The GTV should be defined using mediastinal and lung windows and consists of the
primary tumour and all enlarged (> 1-​cm short axis) lymph nodes. If prior chemo-
therapy has been delivered evidence from a small randomized clinical trial suggests
that the GTV may be limited to just the residual volume(46).
132 Radiotherapy for thoracic tumours

◆ The CTV consists of a margin for microscopic spread (usually around 5  mm)
around the GTV and the mediastinal nodal groups deemed to be at risk of con-
taining microscopic disease.
◆ The ITV should be individualized for the tumour movement and include 5 mm for
mediastinal movement.
◆ To produce the PTV a margin for setup error should be added according to depart-
mental guidelines, typically 5 mm.

Dose prescription
◆ The optimal dose has not been established though there are data to support both a
dose response and the importance of overall treatment time.
◆ The best published survival rates were achieved with hyperfractioned radio-
therapy though as with CHART, for logistical reasons, this has not been widely
adopted(47).
◆ The CONVERT study compared 66 Gy in 33 fractions with 45 Gy in 30 twice-​daily
fractions and has shown no difference in outcome for either fractionation(48).
Current recommended dose schedules include:
◆ 45 Gy to isocentre in 30 twice-​daily fractions (minimum 6-​hour gap) over 15 days(47).
◆ 45–​50 Gy to isocentre in 20–​25 daily fractions over 4–​5 weeks.
◆ 60–​66 Gy to isocentre in 30–​33 daily fractions of 6–​6½ weeks.
◆ 40 Gy to mid-​plane in 15 daily fractions over 3 weeks.(45)

Prophylactic cranial irradiation (PCI)

◆ PCI has also shown a 5% improvement in 3-​year survival and is recommended for
all patients with LSCLC who have achieved a complete response.
◆ In order to reduce the risk of neurotoxicity, PCI should be delivered after the com-
pletion of all chemotherapy.
◆ There is little data on the use of PCI in patients over 70  years of age. The initial
trials excluded older patients (due to concerns regarding potential increased
neurocognitive toxicity) and the most recent dose finding study(49) had a median
age of only 60 years.
◆ Most centres use thermoplastic shells for PCI to improve localization and avoid skin
marks, which can be easily washed off.
◆ The treatment is planned using two lateral portals to encompass the whole of the
brain, paying particular attention to the temporal lobe reflection.
◆ MLC may be used to reduce skin dose at the skull vertex and, consequently, long-​
term alopecia.
◆ The international EULINT trial compared 25 Gy in 10 fractions with 36 Gy in either
18 daily or 24 twice-​daily fractions. The intracerebral recurrence rate was similar
with 26% of patients developing brain metastases by 2 years and the overall survival
was marginally worse in the higher dose arm (42% 2 years in 25 Gy vs 37% in 36
Gy p  =  0.05)(49). By 3  years, both groups were noted to have mild memory and
 Lung cancer 133

intellectual impairment (5% Grade 2 or more) so patients need to be informed of

these at time of consent(50).
The most widely used fractionation schedule is therefore:
◆ 25 Gy mid-​plane dose in 10 fractions over 2 weeks.

Postoperative radiotherapy
◆ Occasionally patients who have had a lobectomy or pneumonectomy are found,
unexpectedly, to have SCLC.
◆ The outcome may be better as they have more localized disease, but the standard
recommendation, particularly for node positive disease, is that they should receive
postoperative treatment with chemotherapy, mediastinal radiotherapy (unless fully
explored negative mediastinal nodes) and PCI(51).

6.1.12  Palliative thoracic treatment for lung cancer

Palliative radiotherapy should be considered for all patients who have incurable lung
cancer and local symptoms. Radiotherapy is good at palliating haemoptysis (80%
of patients benefit), chest pain (60%), and cough (40%), but is less good at helping
dyspnoea (30%) and fatigue (20%).
For poor performance status patients with NSCLC who are asymptomatic, pal-
liative radiotherapy can be deferred until symptoms develop. Data from a ran-
domized controlled trial demonstrated that 60% of such patients never required
radiotherapy(52). However, patients with locally advanced NSCLC and a good per-
formance status (WHO 0 or 1) may have a survival benefit from a higher dose of
radiotherapy. In an MRC study, patients receiving 39 Gy in 13 fractions had a me-
dian survival of 2  months longer than those receiving 17 Gy in two fractions(53).
A  similar result was also observed in a Canadian study of 20 Gy in five fractions
vs 10 Gy in one fraction(54). Consequently, high-​dose palliative radiotherapy is fre-
quently offered to fitter patients as part of the initial management, regardless of
symptoms, either before or following chemotherapy, or for those unsuitable for, or
who decline chemotherapy(55).
Chemotherapy is the mainstay of treatment for both limited (stages I–​III) and exten-
sive (stage IV) SCLC therefore thoracic palliative radiotherapy is used for patients who:
1. are not fit enough/​decline to receive chemotherapy;
2. have progressed during chemotherapy;
3. have a partial response to chemotherapy with persistent local symptoms;
4. relapse shortly after chemotherapy.
The use of consolidation thoracic radiotherapy in patients with stage IV SCLC
should be considered. One study suggests that patients with low-​volume meta-
static SCLC, who have had a complete response to their extrathoracic disease,
might gain a survival benefit from consolidation thoracic irradiation(56). A phase
III randomized trial in patients with extensive stage SCLC who had completed
chemotherapy compared radiotherapy (30 Gy in 10 fractions) to residual thor-
acic to no radiotherapy(57). A  survival benefit was shown at later time points of
134 Radiotherapy for thoracic tumours

18 and 24  months in patients with encompassable thoracic disease, who were
chemoresponsive but had not had a complete intrathoracic response and who had
no brain metastases. Hence, in addition to PCI, consolidation thoracic radiation
(TCI) should be considered in patients with mediastinal disease who demonstrate
a response to chemotherapy.

Patient positioning for palliative radiotherapy

Today most patients having palliative thoracic radiotherapy are planned using virtual
simulation. This makes assessment of the current disease configuration and design of
the treatment portals much easier. However, if required, the fields can be defined using
a conventional simulator.
On the CT scanner the patient should be scanned in a comfortable position,
usually on a mattress, with their arms by their sides. A  scan with 3–​5-​mm slice
thickness should be obtained whilst the patient breaths gently. The centre of the
treatment field should be tattooed before the patient moves (if necessary an ap-
proximate isocentre can be established swiftly and then the treatment finalized
using asymmetric jaws).

Volume/​field localization
If using virtual simulation software, there are two approaches that can be used
depending on clinician’s preference; some clinicians define the GTV including the
primary disease, enlarged nodes, and any that might be involved, then use this
target to design the treatment portals whereas others directly define the portals
without initial contouring. Either approach is reasonable and depends on one’s
experience.
◆ A margin should be placed around the gross tumour to account for movement and
set-​up inaccuracies. It should be remembered that often these patients are frail so
set-​up maybe less accurate. A margin of 1.5–​2 cm is usually added to the field edge
(50% isodose).
◆ Most patients can be treated using parallel anterior–​posterior beams though some
(especially those receiving higher doses) may benefit from the use of a longitudinal
wedge to compensate for the slope of the sternum. If the patient has very poor lung
function then a formal 3D volume and calculation may be considered to minimize
normal lung irradiation.
◆ Shielding should be used to reduce the risk of lung toxicity.
◆ When using higher dose palliation with a spinal cord BED > 100 (39 Gy in 13 frac-
tions or 17 Gy in two fractions) a radiotherapy plan with lung and cord dose calcu-
lated or a spinal cord block should be considered(58).

Treatment verification
It is considered good practice to verify all treatments on the machine prior to the first
fraction. This ensures no systematic error has occurred.
 Mesothelioma 135

Dose prescriptions
NSCLC
1. High-​dose palliation: 39 Gy or 36 Gy mid-​plane dose in 13 or 12 fractions over 2½
weeks using ≤ 10-​MV photons (block spinal cord on posterior field for first three
or four fractions.
2. Poor performance status patients: 10 Gy mid-​plane dose in a single fraction pro-
vided area of field < 150 cm2 using ≤ 10-​MV photons.
3. If fields are > 150 cm2 or the patient has stridor or superior vena caval obstruction
then: 20 Gy mid-​plane dose in five fractions over 1 week using ≤ 10-​MV photons.
4. An alternative dose is: 17 or 16 Gy in two fractions 1 week apart using ≤ 10-​MV
photons (block spinal cord for first fraction if using 17 Gy). In some cases with a
large variation in separation a superior–​inferior wedge may be beneficial.

SCLC
A variety of schedules are used:
1. For patients with localized SCLC unfit for chemotherapy then: 40 Gy mid-​plane in
15 fractions over 3 weeks using ≤ 10-​MV photons.
2. For fitter patients (PS 0–​1) with stage IV SCLC and a good response to chemo-
therapy: 30 Gy mid-​plane dose in 10 fractions over 2 weeks using ≤ 10-​MV photons.
3. For less fit patients, more hypofractionated schedules can be used such as: 20 Gy
mid-​plane dose in five fractions over 1 week using ≤ 10-​MV photons. Or 16 Gy
mid-​plane dose in two fractions 1 week apart using ≤ 10-​MV photons.
4. Frail (PFS 3) patients with respiratory symptoms may gain some benefit from 10 Gy
mid-​plane dose in a single fraction provided area of field < 150 cm2.

PCI in Stage IV (extensive) SCLC

Slotman et al. demonstrated an improvement in overall survival (27% at 1 year vs 13%
in control) for patients (up to age 75 and WHO performance status 0–​2) with exten-
sive SCLC who have responded to chemotherapy with a halving of the incidence of
brain metastases (15% vs 40%)(59). Because of the shorter duration of survival, a sim-
pler technique is usually employed using two standard whole brain portals without
MLC blocking or immobilization.
Dose prescription
Depending on the patient’s performance status and extent of disease:
◆ 20 Gy mid-​plane dose in five fractions over 1 week.
◆ 25 Gy mid-​plane dose in 10 fractions over 2 weeks.

6.2 Mesothelioma
Mesothelioma is one of the most common occupation-​related malignancies with
the majority of patients having a history of asbestos exposure. Patients usually
136 Radiotherapy for thoracic tumours

present with locally advanced disease causing breathlessness and/​or pain. Though a
staging system exists (IMIG) it is infrequently used as it rarely affects management
decisions.
Pleural malignant mesothelioma is a radiosensitive tumour but the diffuse spread
throughout the pleural cavity precludes radical radiotherapy with the lung in situ.
Consequently, a number of studies have been conducted to look at the role of radio-
therapy combined with chemotherapy following extrapleural pneumonectomy
(EPP)—​so called ‘trimodality therapy’. The MARS study showed a significant det-
riment in survival compared to standard therapy of best supportive care ± chemo-
therapy, primarily due to high number of postoperative deaths(60). This approach
continues to be advocated by individual centres but should be considered investiga-
tional. The radiotherapy modalities used include IMRT and proton therapy; however,
the optimal technique has yet to be defined. Studies have stressed the importance of
minimizing the dose to the remaining lung(61).

6.2.1  Current indications for radiotherapy

Port site radiotherapy
With the decline in the use of EPP the management of pleural disease has changed; the
diagnosis of malignant mesothelioma is confirmed with a video-​assisted thoracoscopic
surgery (VATS) biopsy followed by either talc pleurodesis or insertion of an indwelling
pleural catheter. Likewise pleural decortication via a limited thoracotomy scar is in-
creasingly common. The reported incidence of procedure tract metastases (PTM)
varies widely from 4% to 40%. Many of these series predate the routine use of chemo-
therapy for fit patients. Therefore the exact role of adjuvant radiotherapy to operative
port sites in the current era is controversial. Some small historical randomized trials
suggested a reduction of symptomatic local recurrence(62–​64) but pooled analysis failed
to confirm this benefit(65). In patients, not receiving Pemetrexed-​based chemotherapy,
there may be benefit in considering PORT site irradiation(66). The results of the PIT
study are awaited (NCT01604005).

Palliative radiotherapy
Patients with pleural mesothelioma can benefit from localized palliative radiotherapy
to aid pain control(67).

6.2.2 Patient positioning
The treatment field is clinically marked either with the patient on the treatment table
or using virtual simulation. The advantage of the latter is that the target depth can be
more accurately defined. The patient should be positioned comfortably and reprodu-
cibly with the affected area in a position so it is easily accessible by the machine.

6.2.3 Target delineation
Usually a 2–​3-​cm margin is added to the surgical bed or nodule. Where a pleural cath-
eter has been placed this should include the entire tract.
 Thymic tumours 137

6.2.4 Planning
Treatment can either be delivered using 9–​18-​MeV electrons or 250–​300-​kV or 4–​6-​
MV photons. The electron energy should be selected to treat the entire chest wall to
90%. Appropriate depth bolus should be used to ensure the skin dose is adequate.
When treating with photons, either a single applied field or a tangential parallel-​
opposed pair may be used. The dose to the underlying lung should be considered.

6.2.5 Dose prescription
For procedure tract site radiotherapy: 21 Gy in three fractions over 3 working days.
For palliative radiotherapy: 10 Gy in single fraction or 20 Gy in five fractions over
1 week.

6.3 Thymic tumours
Thymomas and thymic carcinomas are rare lesions originating in the anterior medias-
tinum. Surgery, often requiring extended thymectomy, is the mainstay of treatment for
the majority of lesions and progression-​free and overall survival is strongly associated
with the completeness of resection.
The most commonly used staging system is that described by Masaoka with the Koga
modification, refined by the International Thymic Malignancy Interest Group (ITMIG).
Careful preoperative staging with CT and MRI is required to confirm that resection
is likely to be successful and result in a macroscopic clearance of the disease.

6.3.1  Current indications for radiotherapy

◆ If resection is likely to leave gross residual disease then preoperative treatment with
chemotherapy alone or in combination with radiotherapy should be considered, as
shrinkage may allow a complete resection.
◆ Patients who are medically unfit for surgery or who are surgically inoperable
due to great vessel involvement may be considered for definitive radiotherapy or
chemoradiation.
◆ Postoperative radiotherapy is recommended for patients with an R1 or R2 resection
or who have stage III or IVa disease. Two recent meta-​analyses examining the role
of post-​operative radiotherapy reported an overall survival benefit in patients with
stage III and IV disease(69,70). However, the two reports differed regarding whether or
not there is a survival advantage for the use of PORT in stage II disease. At present
PORT for patients with stage II disease should only be considered in patients who
have histological features predicting an aggressive clinical course (higher grader
and larger tumour size).

6.3.2 Radical treatment

Patient positioning
The treatment position depends on the site of the lesion. If the lesion is located su-
periorly, close to the neck, the patient may be treated with their arms by their sides,
138 Radiotherapy for thoracic tumours

sometimes in a mask. Usually, however, they will be immobilized on a lung board,

with their arms above their heads and neck slightly extended as for NSCLC. Whilst
breathing gently, the patient should, have a contrast-​enhanced CT scan performed
with 3-​mm slice spacing covering the entire lungs.

Treatment volume and definitions

Primary radical/​preoperative
◆ The GTV consists of the primary lesion and any enlarged nodes identified on the
staging CT and MRI scans. Mediastinal windows (W400, L20) should be used.
◆ The CTV includes a margin of 5–​10 mm added for microscopic spread.
◆ The definition of the ITV will depend on the scanning technique used. If conven-
tional scanning is used then a 10–​15-​mm margin needs to be added. It has been
reported that this margin may be reduced to 5–​10 mm if 4D scanning is used.
◆ To produce the PTV, a margin should be added to the ITV to account for setup vari-
ability. This margin will vary between departments, but is usually around 5 mm.

Postoperative
◆ GTV is the surgical bed of the tumour defined using the staging CT and MRI scans.
Any areas of macroscopic residual disease should be marked by clips at time of sur-
gery and should be boosted.
◆ CTV accounts of microscopic spread and areas that might have been affected by the
surgery.
◆ ITV and PTV are the same as for primary lung radiotherapy.

Dose prescription
The treatment should be planned in the same manner as that described for the rad-
ical treatment of NSCLC. A  variety of field arrangements have been advocated.
Traditionally, anterior field and two anterior oblique or posterior oblique fields were
used although increasingly IMRT is utilized to minimize the dose to normal lung(71).
The OAR dose constraints are the same as those used in the treatment of lung cancer.
◆ When radiotherapy is the sole modality of treatment and surgery is not planned
the current recommendation is 54 Gy to isocentre in 28–​30 fractions using 6–​10-​
MV photons with areas of bulky disease escalated to 60–​66 Gy either with a syn-
chronous or sequential boost.
◆ When the treatment is being delivered preoperatively with a view to downstaging
there is an increasing trend to use combined chemoradiation. No single regimen
can be recommended, but with increasing experience in the preoperative setting
in NSCLC, one regimen being increasingly adopted is 45 Gy in 25 fractions over 5
weeks in combination with cisplatin and etoposide. A dose below 40 Gy or above 64
Gy would be deemed inappropriate in this setting.
When delivered postoperatively usually a dose of 45–​54 Gy in 25–​30 fractions
◆

using 6–​10-​MV photons is recommended. Areas of gross residual disease should be

boosted to 64–​66 Gy.
 References 139

Treatment verification
This is identical to the method used for radical treatment of NSCLC.

6.3.3 Palliative radiotherapy
Occasionally, a patient may present with advanced disease who is not fit for either
chemotherapy or radical radiotherapy yet has local symptoms or who develops symp-
tomatic recurrences. In these cases the treatment is similar to that described for the
palliative treatment of NSCLC.

Acknowledgements
The authors would like to acknowledge Sara C Erridge, Elizabeth Toy, and Sorcha
Campbell who are the authors of this chapter from the previous edition and who pro-
vided the material on which this chapter is based. We also would like to thank Corrine
Faivre-​Finn and Gareth Webster from the Christie Hospital Manchester for supplying
the IMRT images.

References
1. Erridge S, Murray B, Price A, et al. Improved treatment and survival for lung cancer
patients in South-​East Scotland. Journal of Thoracic Oncology 2008; 3:491–​8.
2. Raz DJ, Zell JA, Ou SH, et al. Natural history of stage I non-​small cell lung
cancer: implications for early detection. Chest 2007; 132:193–​9.
3. Celli BR, Cote CG, Marin JM, et al. The body-​mass index, airflow obstruction, dyspnea,
and exercise capacity index in chronic obstructive pulmonary disease. New England
Journal of Medicine 2004 4; 350:1005–​12.
4. Giaj-​Levra N, Sciascia S, Fiorentino A, Fersino S, Mazzola R, Ricchetti F, Roccatello D,
Alongi F. Radiotherapy in patients with connective tissue diseases. Lancet Oncology 2016;
17:e109–​17.
5. Brunelli A, Charloux A, Bolliger CT, et al. ERS/​ESTS clinical guidelines on fitness for
radical therapy in lung cancer patients (surgery and chemo-​radiotherapy). European
Respiratory Journal 2009; 34: 17–​41.
6. Birim O, Kappetein AP, Stijnen T, Bogers AJ. Meta-​analysis of positron emission
tomographic and computed tomographic imaging in detecting mediastinal lymph node
metastases in nonsmall cell lung cancer. Annals of Thoracic Surgery 2005; 79: 375–​82.
7. Lim E, Baldwin D, Beckles M, et al. Guidelines on the radical management of patients
with lung cancer. Thorax 2010; 65(Suppl 3): iii1–​27.
8. Peeters ST, Dooms C, Van Baardwijk A, et al. Selective mediastinal node irradiation in
non-​small cell lung cancer in the IMRT/​VMAT era: How to use E(B)US-​NA information
in addition to PET-​CT for delineation? Radiotherapy and Oncology 2016; 120:273–​78.
9. Purdie TG, Bissonnette JP, Franks K, et al. Cone-​beam computed tomography for on-​
line image guidance of lung stereotactic radiotherapy: localization, verification, and
intrafraction tumor position. International Journal of Radiation,Oncology, Biology, Physics
2007 May 1;68:243–​52.
10. De Ruysscher D, Faivre-​Finn C, Nestle U, et al. European Organisation for Research and
Treatment of Cancer recommendations for planning and delivery of high-​dose, high-​
precision radiotherapy for lung cancer. Journal of Clinical Oncology 2010; 28: 5301–​10.
140 Radiotherapy for thoracic tumours

11. Cole AJ, Hanna GG, Jain S, O’Sullivan JM. Motion management for radical radiotherapy
in non-​small cell lung cancer. Clinical oncology (Royal College of Radiologists (Great
Britain)) 2014; 26:67–​80.
12. Hurkmans CW, Cuijpers JP, Lagerwaard FJ, et al. Recommendations for implementing
stereotactic radiotherapy in peripheral stage IA non-​small cell lung cancer: report from
the Quality Assurance Working Party of the randomised phase III ROSEL study. Radiation
Oncology 2009; 4:1.
13. Graham MV, Purdy JA, Emami B, et al. Clinical dose-​volume histogram analysis for
pneumonitis after 3D treatment for non-​small cell lung cancer (NSCLC). International
Journal of Radiation Oncology, Biology, Physics 1999; 45:323–​9.
14. Marks LB, Bentzen SM, Deasy JO, et al. Radiation dose-​volume effects in the lung.
International Journal of Radiation Oncology, Biology, Physics 2010; 76(3 Suppl):S70–​6.
15. Hanna GG, Patel R, Aitken K, et al. EP-​1937: UK stereotactic ablative radiotherapy trials
normal tissue dose constraints tolerance consensus. Radiotherapy and Oncology 2016;
119(Suppl 1):S919.
16. Aupérin A, Le Péchoux C, Rolland E, et al. Meta-​analysis of concomitant versus sequential
radiochemotherapy in locally advanced non-​small-​cell lung cancer. Journal of Clinical
Oncology 2010; 28:2181–​90.
17. Kirkpatrick JP, van der Kogel AJ, Schultheiss TE. Radiation dose-​volume effects in
the spinal cord. International Journal of Radiation Oncology, Biology, Physics 2010;76(3
Suppl): S42–​9.
18. Dische S, Saunders MI. Continuous, hyperfractionated, accelerated radiotherapy
(CHART): an interim report upon late morbidity. Radiotherapy and Oncology 1989;
16:65–​72.
19. Werner-​Wasik M, Yorke E, Deasy J, et al. Radiation dose-​volume effects in the
esophagus. International Journal of Radiation Oncology, Biology, Physics 2010; 76
(3 Suppl): S86–​93.
20. Haslett K, Franks K, Hanna GG, et al. Protocol for the isotoxic intensity modulated
radiotherapy (IMRT) in stage III non-​small cell lung cancer (NSCLC): a feasibility study.
BMJ Open.2016; 6:e010457. doi: 10.1136/​bmjopen-​2015-​010457.
21. Chun SG, Hu C, Choy H, et al. Impact of intensity-​modulated radiation therapy
technique for locally advanced non–​small-​cell lung cancer: A secondary analysis of the
NRG Oncology RTOG 0617 randomized clinical trial. Journal of Clinical Oncology 2017
Jan;35:56–​62.
22. Ong CL, Verbakel WF, Cuijpers JP, et al. Stereotactic radiotherapy for peripheral
lung tumors: a comparison of volumetric modulated arc therapy with 3 other delivery
techniques. Radiotherapy and Oncology 2010; 97: 437–​42.
23. Royal College of Radiologists. The timely delivery of radical radiotherapy: standards
and guidelines for the management of unscheduled treatment interruptions, Third
edition, 2008.
24. National Radiotherapy Implementation Group. National Radiotherapy Implementation
Group Report, Image Guided Radiotherapy (IGRT), Guidance for implementation and use.
National Cancer Action Team 2012.
25. Royal College of Radiologists. ‘On target: ensuring geometric accuracy in radiotherapy’
2008.
26. Goldstraw P, Chansky K, Crowley J, et al. The IASLC Lung Cancer Staging
Project: proposals for revision of the TNM stage groupings in the forthcoming (eighth)
 References 141

edition of the TNM Classification for Lung Cancer. Journal of Thoracic Oncology 2016;
11:39–​51.
27. Palma D, Senan S. Stereotactic radiation therapy: changing treatment paradigms for stage
I nonsmall cell lung cancer. Current Opinion in Oncology 2011; 23:133–​9.
28. Nyman J, Hallqvist A, Lund JÅ, et al. SPACE—​A randomized study of SBRT vs
conventional fractionated radiotherapy in medically inoperable stage I NSCLC.
Radiotherapy and Oncology. 2016; 121:1–​8.
29. Ball D, Mai GT, Vinod S, et al. Stereotactic ablative radiotherapy versus standard
radiotherapy in stage 1 non-small-cell lung cancer (TROG 09.02 CHISEL): a phase 3,
open-label, randomised controlled trial. Lancet Oncol. 2019;pii: S1470–2045(18)30896-9.
doi: 10.1016/S1470-2045(18)30896-9.
30. Chang JY, Senan S, Paul MA, et al. Stereotactic ablative radiotherapy versus lobectomy
for operable stage I non-​small-​cell lung cancer: a pooled analysis of two randomised trials.
Lancet Oncol. 2015 Jun;16:630–​37.
31. Kappers I, Klomp HM, Koolen MG, et al. Concurrent high-​dose radiotherapy with low-​
dose chemotherapy in patients with non-​small cell lung cancer of the superior sulcus.
Radiotherapy and Oncology. 2011; 101:278–​83.
32. Rusch VW, Giroux DJ, Kraut MJ, et al. Induction chemoradiation and surgical resection
for superior sulcus non-​small-​cell lung carcinomas: long-​term results of Southwest
Oncology Group Trial 9416 (Intergroup Trial 0160). Journal of Clinical Oncology 2007;
25:313–​18.
33. Andrews DW, Scott CB, Sperduto PW, et al. Whole brain radiation therapy with
or without stereotactic radiosurgery boost for patients with one to three brain
metastases: phase III results of the RTOG 9508 randomised trial. Lancet 2004;
363(9422):1665–​72.
34. Patchell RA, Tibbs PA, Walsh JW, et al. A randomized trial of surgery in the treatment of
single metastases to the brain. New England Journal of Medicine 1990; 322:494–​500.
35. Konert T, Vogel W, MacManus MP, et al. PET/​CT imaging for target volume delineation
in curative intent radiotherapy of non-​small cell lung cancer: IAEA consensus report 2014.
Radiotherapy and Oncology 2015; 116:27–​34
36. Grills IS, Fitch DL, Goldstein NS, et al. Clinicopathologic analysis of microscopic
extension in lung adenocarcinoma: defining clinical target volume for radiotherapy.
International Journal of Radiation, Oncology, Biology, Physics 2007; 69:334–​41.
37. Mauguen A, Le Péchoux C, Saunders MI, et al. Hyperfractionated or accelerated
radiotherapy in lung cancer: an individual patient data meta-​analysis. Journal of Clinical
Oncology 2012; 30:2788–​97.
38. Saunders M, Dische S, Barrett A, et al. Continuous, hyperfractionated, accelerated
radiotherapy (CHART) versus conventional radiotherapy in non-​small cell lung
cancer: mature data from the randomised multicentre trial. CHART Steering committee.
Radiotherapy and Oncology. 1999; 52:137–​48.
39. Bradley JD, Paulus R, Komaki R, et al. Standard-​dose versus high-​dose conformal
radiotherapy with concurrent and consolidation carboplatin plus paclitaxel with or without
cetuximab for patients with stage IIIA or IIIB non-​small-​cell lung cancer (RTOG 0617): a
randomised, two-​by-​two factorial phase 3 study. Lancet Oncology 2015; 16:187–​99.
40. Maguire J, Khan I, McMenemin R, et al. SOCCAR: A randomised phase II trial
comparing sequential versus concurrent chemotherapy and radical hypofractionated
radiotherapy in patients with inoperable stage III Non-​Small Cell Lung Cancer and good
performance status. European Journal of Cancer 2014 Nov;50:2939–​49.
142 Radiotherapy for thoracic tumours

41. Chang JY, Bezjak A, Mornex F, IASLC Advanced Radiation Technology Committee.
Stereotactic ablative radiotherapy for centrally located early stage non-​small-​cell lung
cancer: what we have learned. Journal of Thoracic Oncology 2015; 10:577–​85.
42. PORT Meta-​analysis Trialists Group. Postoperative radiotherapy for non-​small cell lung
cancer. Cochrane Database of Systematic Reviews 2005; 2: CD002142.
43. Billiet C, Peeters S, Decaluwé H, et al. Postoperative radiotherapy for lung cancer: Is it
worth the controversy? Cancer Treatments Reviews 2016; 51:10–​18.
44. Fried DB, Morris DE, Poole C, et al. Systematic review evaluating the timing of thoracic
radiation therapy in combined modality therapy for limited-​stage small-​cell lung cancer.
[Erratum appears in Journal of Clinical Oncology 2005; 23: 248]. Journal of Clinical
Oncology 2004; 22: 4837–​45.
45. Murray N, Coy P, Pater JL, et al. Importance of timing for thoracic irradiation in the
combined modality treatment of limited-​stage SCLC. Journal of Clinical Oncology 1993;
11: 336–​44.
46. Kies MS, Mira JG, Crowley JJ, et al. Multimodal therapy for limited small-​cell lung
cancer: a randomized study of induction combination chemotherapy with or without
thoracic radiation in complete responders; and with wide-​field versus reduced-​field
radiation in partial responders: a Southwest Oncology Group Study. Journal of Clinical
Oncology 1987; 5: 592–​600.
47. Turrisi AT, 3rd, Kim K, Blum R, et al. Twice-​daily compared with once-​daily thoracic
radiotherapy in limited small-​cell lung cancer treated concurrently with cisplatin and
etoposide. New England Journal of Medicine 1999; 340: 265–​71.
48. Faivre-​Finn C, Falk S, Ashcroft L, et al. Protocol for the CONVERT trial-​Concurrent
ONce-​daily VErsus twice-​daily RadioTherapy: an international 2-​arm randomised
controlled trial of concurrent chemoradiotherapy comparing twice-​daily and once-​daily
radiotherapy schedules in patients with limited stage small cell lung cancer (LS-​SCLC) and
good performance status. BMJ Open 2016 Jan 20;6:e009849.
49. Le Péchoux C, Dunant A, Senan S, et al. Standard-​dose versus higher-​dose prophylactic
cranial irradiation (PCI) in patients with limited-​stage small-​cell lung cancer in complete
remission after chemotherapy and thoracic radiotherapy (PCI 99–​01, EORTC 22003–​
08004, RTOG 0212, and IFCT 99–​01): a randomised clinical trial. Lancet Oncology 2009;
10: 467–​74.
50. Le Péchoux C, Laplanche A, Faivre-​Finn C, et al. Clinical neurological outcome and
quality of life among patients with limited small-​cell cancer treated with two different
doses of prophylactic cranial irradiation in the intergroup phase III trial (PCI99–​
01, EORTC 22003–​08004, RTOG 0212 and IFCT 99–​01). Annals of Oncology 2011;
22:1154–​63.
51. Wong AT, Rineer J, Schwartz D, Schreiber D. Assessing the impact of postoperative
radiation therapy for completely resected limited-​stage small cell lung cancer using the
National Cancer Database. Journal of Thoracic Oncology 2016;11:242–​8.
52. Falk SJ, Girling DJ, White RJ, et al. Immediate versus delayed palliative thoracic
radiotherapy in patients with unresectable locally advanced non-​small cell lung cancer and
minimal thoracic symptoms: randomised controlled trial. British Medical Journal 2002;
325:465.
 References 143

53. Macbeth FR, Bolger JJ, Hopwood P, et al. Randomized trial of palliative two-​fraction
versus more intensive 13-​fraction radiotherapy for patients with inoperable non-​small
cell lung cancer and good performance status. Medical Research Council Lung Cancer
Working Party. Clinical Oncology 1996; 8:167–​75.
54. Bezjak A, Dixon P, Brundage M, et al. Randomized phase III trial of single versus
fractionated thoracic radiation in the palliation of patients with lung cancer (NCIC CTG
SC.15). International Journal of Radiation Oncology, Biology, Physics 2002; 54:719–​28.
55. Stevens R, Macbeth F, Toy E, et al Palliative radiotherapy regimens for patients with
thoracic symptoms from non-​small cell lung cancer. Cochrane Database Systematic Reviews
2015;1:CD002143.
56. Jeremic B, Shibamoto Y, Nikolic N, et al. Role of radiation therapy in the combined-​
modality treatment of patients with extensive disease small-​cell lung cancer: A randomized
study. Journal of Clinical Oncology 1999; 17:2092–​9.
57. Slotman BJ, van Tinteren H, Praag JO, et al. Use of thoracic radiotherapy for extensive
stage small-​cell lung cancer: a phase 3 randomised controlled trial. Lancet. 2015;
385:36–​42.
58. Macbeth FR, Wheldon TE, Girling DJ, et al. Radiation myelopathy: estimates of risk in
1048 patients in three randomized trials of palliative radiotherapy for non-​small cell lung
cancer. The Medical Research Council Lung Cancer Working Party. Clinical Oncology 1996;
8:176–​81.
59. Slotman B, Faivre-​Finn C, Kramer G, et al. Prophylactic cranial irradiation in extensive
small-​cell lung cancer. New England Journal of Medicine 2007; 357:664–​72.
60. Treasure T, Lang-​Lazdunski L, Waller D, et al. Extra-​pleural pneumonectomy versus no
extra-​pleural pneumonectomy for patients with malignant pleural mesothelioma: clinical
outcomes of the Mesothelioma and Radical Surgery (MARS) randomised feasibility study.
Lancet Oncology 2011; 12:763–​72.
61. Chi A, Liao Z, Nguyen NP, et al. Intensity-​modulated radiotherapy after extrapleural
pneumonectomy in the combined-​modality treatment of malignant pleural mesothelioma.
Journal of Thoracic Oncology 2011; 6:1132–​41.
62. Boutin C, Rey F, Viallat J. Prevention of malignant seeding after invasive diagnostic
procedures in patients with pleural mesothelioma. A randomized trial of local
radiotherapy. Chest 1995; 108:754–​8.
63. Bydder S, Phillips M, Joseph D. A randomised trial of single-​dose radiotherapy to prevent
procedure tract metastasis by malignant mesothelioma. British Journal of Cancer 2004;
91:9–​10.
64. O’Rourke N, Garcia J, Paul J. A randomised controlled trial of intervention site
radiotherapy in malignant pleural mesothelioma. Radiotherapy and Oncology 2007;
84:18–​22.
65. Ung YC, Yu E, Falkson C, et al. The role of radiation therapy in malignant pleural
mesothelioma: a systematic review. Radiotherapy and Oncology 2006; 80: 13–​8.
66. Clive AO, Taylor H, Dobson L, et al. Prophylactic radiotherapy for the prevention of
procedure-​tract metastases after surgical and large-​bore pleural procedures in malignant
pleural mesothelioma (SMART): a multicentre, open-​label, phase 3, randomised controlled
trial. Lancet Oncology 2016; 17:1094–​104.
67. MacLeod N, Chalmers A, O’Rourke N, et al. Is radiotherapy useful for treating pain in
mesothelioma? A Phase II trial. Journal of Thoracic Oncology2015; 10:944–​50.
144 Radiotherapy for thoracic tumours

68. Detterbeck F. The Masaoka-​Koga Stage classification for thymic malignancies.

Clarification and definition of terms. Journal of Thoracic Oncology 2011; 6:S1710–​16.
69. Zhou D, Deng XF, Liu QX, et al. The effectiveness of postoperative radiotherapy in
patients with completely resected thymoma: A meta-​analysis. Annals of Thoracic Surgery
2016; 101:305–​10.
70. Lim YJ, Kim E, Kim HJ, et al. Survival impact of adjuvant radiation therapy in Masaoka
Stage II to IV thymomas: A systematic review and meta-​analysis. International Journal of
Radiation, Oncology, Biology, Physics. 2016; 94:1129–​36.
71. Gomez D. Radiation therapy definitions and reporting guidelines for thymic malignancies.
Journal of Thoracic Oncology 2011; 6: S1743–​46.
 Chapter 7

Upper gastrointestinal tract
Stephen Falk

7.1 Introduction
Cancers of the upper gastrointestinal tract (GI) and hepato-​biliary system represent a
challenge for the practicing radiotherapist. The overall outlook for patients with these
diseases is poor, with survival rates generally < 10% at 5 years worldwide. The ma-
jority of patients present with either locally advanced or metastatic disease, typically of
poor functional status, and are unsuitable for aggressive therapies. The technical chal-
lenges of these diseases are considerable related to tumour volumes, anatomical situ-
ation, and poor normal tissue tolerance particularly of the intra-​abdominal contents.
Contemporaneous treatment techniques such as intensity-​modulated radiotherapy
(IMRT) and stereotactic ablative radiotherapy (SABR) have not currently made sig-
nificant impact in the routine treatment of upper GI tumours in the UK.

7.2 Oesophageal cancer
Radiotherapy can be used as a neo-​adjuvant prior to surgery; in the adjuvant setting
following oesophagectomy; as sole modality (definitive therapy) usually with con-
comitant chemotherapy; or as palliation. Radiotherapy issues in oesophageal cancer
include tumour localization, radiation volumes, dose, and fractionation, planning
techniques, optimum chemotherapy regimen and scheduling, and whether or not tri-​
modality or bi-​modality therapy is appropriate.

7.2.1  Radical primary treatment

In 1966 Pearson published encouraging results for radiotherapy as sole modality with
a 23% 5-​year survival for middle third tumours and 17% 5-​year survival for lower
third tumours. However, a Cochrane meta-​analysis indicates < 10% 5-​year survival
with radiotherapy alone(1).

Indications
Whether definitive chemo-​radiotherapy or surgery +/​-​neo-​adjuvant chemotherapy
is offered in particular to patients with squamous cell carcinoma (SCC) varies widely
largely dependent on available surgical expertise and treatment philosophy. A recent
feasibility study in SCC showed that due to small numbers, it was not going to be
possible to perform a randomized trial comparing non-​surgical therapies with radical
surgery and current views are that either treatment strategy is acceptable.
146 Upper gastrointestinal tract

Treatment volume and definition

Investigations should include:
◆ Computed tomography (CT) scan of chest, abdomen, and pelvis with intravenous
contrast.
◆ Laparoscopy, biopsy, and peritoneal washings are recommended for full thick-
ness tumours that extend below the diaphragm for evaluation of peritoneal spread,
which is poorly shown on CT scan.
◆ 18-​fludeoxyglucose (FDG) positron emission tomography (PET) scans can iden-
tify occult metastatic disease in an additional 15% of patients when compared with
other imaging modalities notably CT scans.
◆ Endoluminal ultrasound (EUS) is the most reliable indicator of the local stage of
the disease, the extent of the disease superiorly and inferiorly, being able to visualize
disease in the sub-​mucosa, and also of the extent of local lymph node spread (Fig.
7.1). The top of the aortic arch and superior and inferior extent of tumour should be
recorded to allow accurate planning. In particular, the incisor to carina distance is
more variable than previously thought. Classically it is considered to be 25 cm, but
the range in vivo when measured in one clinical study was 20.5–​29 cm(2).
◆ Fibre-​optic bronchoscopy should be performed if CT or EUS suggests invasion of
the trachea or left main bronchus. Should mucosal involvement of the airway or
main bronchi be confirmed histologically then prophylactic placement of a covered
metal stent in the airway is recommended to lessen risks of the severe morbidity
and mortality associated with tracheo-​oesophageal fistula.
The gross tumour volume (GTV) is the gross primary and nodal disease and the
entire circumference of the oesophagus as defined on the planning-​CT scan with
all available diagnostic information. Elective nodal irradiation is not practised, al-
though surgical series show extensive distant local nodal spread particularly in
adenocarcinoma.

Fig. 7.1 Endoluminal
ultrasound
showing a locally
advanced tumour
involving aorta
Source: Bristol Royal
Infirmary
 Oesophageal cancer 147

A Cochrane collaboration meta-​analysis indicates that the addition of chemotherapy

to radiotherapy improves both local control and survival(1). Thirteen randomized trials
were included in the analysis. There were eight concomitant and five sequential radio-
therapy and chemotherapy (RTCT) studies. Concomitant RTCT provided significant
overall reduction in mortality at 1 and 2 years of 7% (95% confidence interval (CI)
1–​15%). The mortality in the control arms was 62% and 83% respectively, and the
local recurrence rate for the control arms was in the order of 68%. Combined RTCT
provided an absolute reduction of local recurrence rate of 12% (95% CI 3–​22%). There
was, however, a significant increase of severe and life-​threatening toxicities.

Clinical target volume definition

Middle /​upper third oesophageal cancers:
◆ The clinical target volume (CTV) should comprise the GTV plus a margin of up to
1cm laterally, anterior, and posterior, 2 cm superiorly and inferiorly (shaped along
the contour of the oesophagus), and then edited to exclude lung, pericardium, large
vessels, trachea, and right/​left main bronchi, and the vertebrae.
Lower third tumours, which involve or come within 2 cm of the gastro-​oesophageal
junction (3D CT target tumour definition):
◆ The CTV should comprise the GTV grown and trimmed as for middle/​upper third
tumours. Inferiorly (this will include the mucosa of the stomach in the direction of
the lymph node stations along the lesser curve including the para-​cardial and left
gastric lymph nodes).

Planning target volume definition

The planning target volume (PTV) is created by the addition of the following
margins:
◆ Superiorly and inferiorly: 1.0 cm
◆ Laterally, anteriorly and posteriorly: 0.5 cm
The maximum treatment field length is 17 cm, i.e. maximum EUS disease length of
primary tumour and lymph nodes is usually 10 cm.
The PTV is, in summary, the EUS-​defined GTV + affected peritumoral nodal sta-
tions with a 1.5 cm lateral, anterior, and posterior margin and 3 cm superior–​inferior
margin.
Lower third tumours, which involve or come within 2 cm of the gastro-​oesophageal
junction (4DCT target tumour definition):
The GTV and CTV should be separately defined on time-​weighted average scans
(e.g. 40% expiratory phase), maximum exhale and maximum inspiration phases.
CTVB combines the CTV volumes to allow for organ motion.
The PTV is created by the addition of 0.5 cm in all directions.
◆ For middle and lower third tumours the patient is immobilized in a vacuum fixed
polystyrene bag and treated supine with arms above head held on an arm pole,
and the shoulders supported for stability by the vacuum bag to allow for the lateral
fields. A device similar to a ‘knee-​fix’ is also encouraged. For the treatment of the
upper third of the oesophagus, a supine cast is made to immobilize the neck and
148 Upper gastrointestinal tract

jaw. The patient is positioned with the cervical spine straight and parallel to the
couch top.
◆ The anatomical landmarks of the oesophagus are arbitrary in nature and include the
cervical oesophagus which begins at the cricopharyngeal muscle at the level of C7
and extends to the thoracic inlet at T3. The mid-​thoracic segment extends from T4
to 8, and the lower thoracic oesophagus from T8 to T10.

Dose distribution and field modifications

Radiation techniques for carcinomas of the mid and lower oesophagus typically com-
prise a conformally planned single phase technique usually requiring a multi-​field
five-​field plan comprising two posterior oblique, two anterior oblique, and an anterior
field (Fig. 7.2). This reduces spinal cord doses and results in relative sparing of the lung
fields. Reducing lung volumes is particularly important if either subsequent planned
or salvage surgery is considered. The greatest morbidity and mortality for tri-​modality
patients is pulmonary toxicity.
The cervical oesophagus can usually be treated by a single plan, comprising two an-
terior placed oblique and wedged fields. The wedge angle is usually 30° when 4–​8 MV
photons are used. Sometimes the distribution is improved by the addition of a straight
anterior field (Fig. 7.3). Given the curvature of the oesophagus within the neck and oe-
sophagus, compensation in the superior–​inferior plane with either a tissue compensator,
or a longitudinal wedge in the superior–​inferior plane is commonly required. The gantry
is angled at 50–​60° for the lateral fields to allow for changes in depth of the oesophagus,
an angle greater than 65° indicates that the beam would pass through the head of the
humerus. The small volumes of lung in the radiation field in the superior portion of the
chest allows for treatment with a single phase two-​or three-​field plan throughout.
Tissue constraints as shown in Table 7.1 should be observed and recorded for each
patient plan.

Fig. 7.2  5-​field plan

for radical therapy
of oesophageal
cancer
 Oesophageal cancer 149

Fig. 7.3  3-​field plan

for treatment of
carcinoma of cervical
oesophagus

Conformal radiotherapy can significantly reduce lung volumes. Dosimetrically,

new techniques especially volumetric-​modulated arc radiotherapy (VMAT), are well
suited to definitive chemo-​radiation of oesophageal cancers in all anatomical locations
(Fig. 7.4)(3). Whether the ability to increase dose as a consequence improves clinical
outcomes remains uncertain.
The patient is aligned longitudinally using anterior and lateral lasers. The anterior
field centre is either marked on the cast for upper third tumours or found in rela-
tion to either anterior or lateral tattoos, and sometimes both. All fields are treated

Table 7.1  Tissue constraints

Region of interest/​organ at risk Dose constraint
PTV V95% > 99.0%

DMAX <107% (53.5 Gy)

GTV GTV min > 100% (50.0 Gy)
Spinal cord PRV Cord Max <80% (40 Gy)
Combined lungs V40% (V20 Gy) <25%
Heart V80% (V40 Gy) < 30%
Liver V60% (V30 Gy) < 60%
Individual kidneys V40% (V20 Gy) <25%
150 Upper gastrointestinal tract

Fig. 7.4  Treatment volume for lower third carcinoma of oesophagus.

 Oesophageal cancer 151

every day. Portal imaging will be performed on the first fraction and then weekly
thereafter.
Appropriate prescription options include:
◆ Neo-​adjuvant chemotherapy is often given such that the chemotherapy prescription
includes cisplatin 60 mg/​m2 day 1 with capecitabine given continuously at a dose
of 625 mg/​m2 twice daily. Continuous infusion 5-​fluorouracil (5FU) 225 mg/​m2/​
day is substituted for the capecitabine when patients cannot swallow capsules. Four
courses of treatment are given repeated every 21 days.
◆ Radiotherapy: 50 Gy in 25 daily treatments 5  days per week concomitant with
chemotherapy, commences at the start of the third chemotherapy cycle.
When radiotherapy alone is used recommended doses include:
◆ 55 Gy in 20 fractions in 4 weeks 5 days per week
◆ 66 Gy in 33 fractions in 6 weeks, 5 days per week.
Treatment is prescribed to the ICRU PTV reference point.
When concomitant chemo-​radiotherapy (CRT) is used, there is currently no good
evidence that radiation doses >50.4 Gy are beneficial. No survival benefit, yet a sig-
nificant increase in toxicity, was seen when a concomitant chemo-​radiation schedule
employing 64 Gy was randomized against 50.4 Gy total dose(4). In spite of this, pat-
terns of failure continue to be predominantly local. Table 7.2 shows patterns of
failure in oesophageal cancer associated with modality of therapy(5). One important
current question is whether dose escalation using techniques such as VMAT, as-
sociated with currently available improved localization techniques including EUS
and PET may improve outcome when compared with clinical studies performed in
the 1980s.

Neo-​adjuvant therapy prior to surgery

Practice varies hugely around the world largely according to custom and practice. The
following remain acceptable therapies for squamous cell and adenocarcinoma of the
oesophagus

Table 7.2  Patterns of failure of oesophageal cancer

Modality No. (%) Local and Neck (%) Mediastinum Local + Distant (%)
marginal (%) (%) distant (%)
RT 517 25–​84 25 10–​43 23–​65

Surgery 266 21–​50 44 33 17–​65

RT/​Surg 2078 22–​87 53 20 17–​43
RT/​Che 254 15–​39 5–​25 6–​25
Tri mod 150 2–​36 5–​38 16–​29

Reproduced with permission from Aisner J. et al. ‘Patterns of recurrence for cancer of the lung and
esophagus.’ In ‘Cancer Treatment Symposia: Proceedings of the Workshop on Patterns of Failure After
Cancer Treatment,’ Vol 2. Washington DC, USA. © US Department of Health and Human Services.
152 Upper gastrointestinal tract

◆ Surgery alone (favoured in the much of the world)

◆ Neo-​adjuvant chemotherapy and surgery (favoured in the UK)
◆ Neo-​adjuvant CRT favoured in Europe and the US.

Pre-​operative radiotherapy
There is no evidence that sole modality pre-​operative radiotherapy improves the
survival of patients with potentially resectable oesophageal cancer. A Cochrane group
quantitative meta-​analysis of pre-​operative radiotherapy using updated data from
1147 patients in five randomized trials has been performed(6). With a median follow-​
up of 9 years, in a group of patients with mostly squamous carcinomas, the hazard
ratio (HR) of 0.89 (95% CI 0.78–​1.01) suggests an overall reduction in the risk of death
of 11% and an absolute survival benefit of 3% at 2 years and 4% at 5 years. This result
is not statistically significant (p = 0.062).
Pre-​operative chemotherapy
British practice has been dominated by the 802 patient MRC OE02 study which
reported a significant prolongation of the 2-​year survival (43 vs 34%) and median
survival (16.8 vs 13.3 months) in favour of the chemotherapy arm(7). In a recent trial
update, the survival benefit was maintained at 5 years with 23% overall survival in the
chemotherapy arm compared to 17% in surgery alone arm, which was independent
of histological subtype (p = 0.03). Gebski et al. (2007) reported a review of eight ran-
domized trials of surgery alone or chemotherapy followed by surgery in patients with
potentially operable oesophageal cancer (n = 1724)(8). The hazard ratio for all-​cause
survival at 2 years favoured the use of chemotherapy followed by surgery (HR = 0.90
(0.81 to 1.0)) with an absolute survival benefit of 7%. However, no survival advantage
was observed in patients with SCC compared to patients with adenocarcinomas.
Pre-​operative chemo-​radiotherapy
Numerous tri-​modality phase II studies utilizing pre-​operative CRT therapy have sug-
gested prolonged and improved survival compared with surgical reports. Treatment-​
related toxicity can, however, be significant. In particular, most centres report an
increase in post-​operative mortality from 2–​3% for surgery alone to up to 8–​10% with
tri-​modality therapy. A  recent systematic overview of 3840 patients from 38 trials
reported an average R0 resection and pathological complete response (pCR) rate of
88.4%, and 25.8%, respectively, with the use of CRT. The post-​operative mortality was
reported as 5.2% and the 5-​year survival rates ranged from 16 to 59% for all patients
and from 34 to 62% in those with pCR(9).
Following the Dutch Trials Group CROSS trial where post-​operative morbidity
and mortality were not increased by tri-​modality therapy, there has been a significant
increase in the numbers of patients treated, particularly for bulky T3/​4 tumours at
risk of a R1 resection with concomitant poor prognosis. In the CROSS trial 363 pa-
tients were randomized to neo-​adjuvant CRT weekly cycles of paclitaxel (50 mg/​m2)
and carboplatin (AUC2) combined with concurrent radiotherapy (RT) (41.4 Gy in
23 fractions) for five weeks or surgery alone(10). Most cancers were adenocarcinomas
(n = 273) and the reported R0 resection rate was 92.3% in the CRT arm vs 64.9% in the
surgery-​alone arm, and the pCR rate was 32.6%. Importantly post-​operative mortality
 Oesophageal cancer 153

was 3.7% in the surgery-​alone arm compared to 3.8% in the CRT arm. The overall
survival was significantly better in the group of patients treated with CRT (HR = 0.67
(0.50–​0.92))and the median survival was 49 months in the CRT arm vs 26 months in
the surgery-​alone arm (10).
The controversy regarding optimal policy of pre-​operative therapy is being ad-
dressed by international collaborative studies e.g. (neoAEGIS).
Indications
Patients with potentially operable adenocarcinoma or squamous cell carcinoma usu-
ally T3-​4 NO or N1 disease of the middle and lower thirds are eligible for tri-​modality
therapy when there is no evidence of disease outside the standard operative field. To
be eligible for therapy, in general patients will:
◆ Be biologically less than 70 years of age without significant co-​morbidities.
◆ Have good performance status (WHO 0 or 1), adequate haematological and renal
function to tolerate all treatments including cisplatin-​based chemotherapy.
◆ Have adequate respiratory function: a resting p02 of > 10 kPa and FEV1> 1.5 litres
is recommended; transfer factor > 50%
◆ have adequate cardiac function as defined by a cardiac ejection fraction of >50%
and normal echocardiography.
Indications for exclusion are:
◆ Patients with distant metastases, including involvement of the supraclavicular lymph
◆ Patients with clear evidence of disease outside a planned radiation volume: In par-
ticular endoscopically visible involvement of the gastric lesser curve, any evidence
of extension to the rest of the stomach (endoscopy or imaging), large-​volume in-
volvement of gastrohepatic ligament/​coeliac axis lymph nodes). In practice more
than 2 cm of sub-​mucosal spread into the stomach shown on EUS is likely to render
this treatment impractical.
Planning techniques are as described for radical primary treatment.
Dose prescription
Treatment includes:
Paclitaxel 50 mg/​m2; Days 1, 8, 15, 22, and 29
Carboplatin Area Under Curve = 2; Days 1, 8, 15, 22, and 29.
The specified radiation treatment dose should be 41.4 Gy in 23 fractions treating
5 days per week with the dose prescribed at the isocentre of a planned target volume.
◆ Constraints should be as for primary therapy.

Postoperative adjuvant therapy
Given the major morbidity of oesophageal cancer surgical procedures, patients are
significantly debilitated following surgery and often unfit for post-​operative therapy
to be given within a reasonable time-​frame following operation. Furthermore,
attempts at radiation therapy are compromised by the need to irradiate the gas-
tric pull-​up or occasionally intestinal interposition. Traditionally post-​operative
154 Upper gastrointestinal tract

radiation has been reserved for patients with microscopic evidence of residual dis-
ease. Post-​operative radiation reduces local recurrence rates but with the rare excep-
tion of lymph-​node negative patients has no clear impact on survival. Techniques
and doses are as described in the neo-​adjuvant section with particular care to re-
duce the length of gastric mucosa irradiated to an absolute minimum (maximum
advised field length 12 cm)

Palliative therapy
The majority of patients with oesophageal cancer are elderly, and often significantly
debilitated following rapid weight loss due to dysphagia. Many have, often asymp-
tomatic, metastatic disease at presentation, discovered during staging procedures
as described above. Such patients are suitable for palliative therapies only. The
options are:
◆ Palliative radiotherapy including brachytherapy.
◆ Stent placement.
◆ Laser treatment or ablation.
◆ Palliative chemotherapy.
Palliative radiation can however improve symptoms of pain and dysphagia in up to
80% of patients. Placement of an expanding metal stent is, however, favoured if there is:
◆ Complete dysphagia or very tight stricturing at endoscopy (e.g. the gastroscope will
not pass through the tumour).
◆ Evidence either clinically or radiologically of fistulation into the airway.
The PTV is usually defined either by barium swallow performed on the simulator or
from endoscopic findings combined with volumes shown on planning CT scans.

Planning technique
The patient lies supine with arms by side. Immobilization devices are not usually
employed.
◆ Typically treatment fields extend 3 cm inferior and superior from the tumour with
a field width of 9  cm, which may be increased as necessary if CT scan suggests
extra-​oesophageal tumour extension. A  simple anterior-​posterior parallel pair is
employed; longitudinal wedges or tissue compensators are not used.
◆ Typical dose prescriptions include 30 Gy in ten daily fractions treating 5 days per
week or 20 Gy in five fractions treating daily over one week.

7.3  Carcinoma of the stomach

7.3.1  Radical primary treatment
Anecdotal evidence suggests that radiotherapy has the ability to salvage a small number
of patients who present with adenocarcinoma of the stomach and locally advanced
unresectable disease. Probably no more than 5–​10% of patients can be cured by such
therapy, and it is not practised in the UK. Standard therapy in the UK is primary
chemotherapy with consideration of surgery when there is a favourable response. The
 Carcinoma of the stomach 155

only commonly used indication for radical radiotherapy to the stomach is as con-
solidation radiotherapy following chemotherapy for high grade B cell non-​Hodgkin’s
lymphoma localized to the stomach, as a substitute for surgery; or as primary therapy
for mucosa associated lymphoid tissue (MALT) lymphomas and low grade B-​cell
lymphomas again localized to the stomach.
Typically the radiation volume encompasses the tumour and major draining
nodal chains. This includes the lesser and greater curvature of the stomach, coeliac
axis including the pancreatico-​ duodenal, splenic, suprapancreatic, and porta
hepatis. Para-​aortic lymph nodes to the level of mid L3 or mid L4 are included
and paraoesophageal nodes for more proximal lesions encompassing the lower
oesophagus.
The treatment volume can be defined by CT scan or less frequently with orthogonal
films following barium contrast.
The patient is treated in the supine position. Since a portion of both kidneys will be
in the treatment volume, care must be taken to not irradiate at least two-​thirds of one
kidney. For proximal gastric lesions, 50% or more of the left kidney is inevitably within
the treatment volume and the right kidney must be shielded. For more distal lesions,
the right kidney is commonly within the treatment volume and the left kidney must
be spared. For lesions at the gastro-​oesophageal junction a CTV margin of 3–​5 cm of
the distal oesophagus is indicated.
The most efficient way to obtain dose distribution through the often large volume
and simultaneously to be able to spare the kidneys remains the use of anterior and
posterior opposed fields.
Due to the poor tolerance of gastric mucosa, the usual dose is 45–​50 Gy in 1.8–​2.0
Gy daily fractions. It is normal practice to reduce the volume to known disease only
after 45 Gy. At this point it may be appropriate and or practical to employ multiple
planned beams. A maximum dose of 55 Gy is recommended.

7.3.2  Post-​operative adjuvant therapy

Loco-​regional recurrence in the gastric or tumour bed, at the anastomosis, or regional
lymph nodes occurs in 40–​65% of patients after gastric resection. The frequency of
such relapses makes regional radiation attractive as adjuvant therapy. Randomized
trials have addressed the extent of surgical resection (D1 vs D2 resection). Whilst in
Japanese hands extended lymph node dissection has improved outcome, these results
have not been replicated in European studies, at least in part due to enhanced mor-
bidity and mortality observed in the groups treated with extended surgery. Radiation
therapy by itself has no significant role following R0 resection. Indeed level A  evi-
dence of benefit of any post-​operative treatment is limited. Sequential British Stomach
Cancer Group (BSCG) trials suggested no survival benefit from post-​ operative
chemotherapy and/​or radiotherapy. However, interpretation of these studies is limited
by 29% of patients randomized having macroscopic residual disease and 18% having
positive resection margins(11). Such patients would not now be considered eligible for
trials of adjuvant therapies. More recent data from MacDonald et  al. using a post-​
operative regimen of 5FU/​leucovorin and radiation demonstrated an improvement in
median overall survival from 27 months, in the surgery-​only group, to 36 months in
156 Upper gastrointestinal tract

the chemo-​radiotherapy group(12). This trial has, however, been criticized in Europe for
inadequate surgical therapy. Of 552 patients, only 54 (10%) had undergone a formal
D2 dissection. A D1 dissection (removal of all invaded (N1) lymph nodes) had been
performed in 199 patients (36%), but most patients (54%) had undergone a D0 dissec-
tion, which is less than a complete dissection of the N1 nodes. Furthermore, survival
outcomes in the combined modality arm of the MacDonald study are equivalent to
European surgery-​only studies where D1 and/​or D2 dissection has been performed.
Post-​operative CRT thus remains controversial in Europe and further studies are
underway.
If post-​operative treatment is planned treatment given should reflect the protocol
published by MacDonald et al.(12). The PTV should encompass the tumour bed, re-
gional nodes, and extend 2  cm beyond the proximal and distal margins of resec-
tion. The presence of proximal T3 lesions will necessitate treatment of the medial
left hemidiaphragm. Perigastric, coeliac, local para-​aortic, splenic, hepatoduodenal
or hepatic-​portal, and pancreaticoduodenal lymph nodes may be included in the ra-
diation fields. In patients with tumours of the gastroesophageal junction, paracardial
and paraoesophageal lymph nodes are included in the radiation fields. Exclusion of
the splenic nodes is recommended in patients with antral lesions to spare the left
kidney.
The CTV is defined by preoperative computed tomographic CT imaging, barium
swallow if necessary, and in some instances, if available surgical clips.
The usual way to obtain dose distribution through the often large volume and sim-
ultaneously to be able to spare the kidneys is by the use of anterior and posterior op-
posed fields with the patient prone in an immobilization device such as a vacuum bag
with arms extended and holding an arm pole.
More recently a ‘split-​field’ technique has been described in which the PTV is div-
ided into two abutting sections, with each section treated using a separate, optimized
field arrangement(13). This conformal technique provides more adequate coverage of
the target volume with 99% of the PTV receiving 95% of the prescribed dose, com-
pared to 93% using AP–​PA fields. Comparative dose–​volume histograms for the
right kidney, left kidney and spinal cord demonstrate lower radiation doses using the
conformal technique, and although the liver dose is higher, it is still well below liver
tolerance. The upper half of the PTV which includes the tumour bed, anastomosis,
and splenic hilar nodes is treated using a 3-​field arrangement comprising anterior
and posterior fields, and a left lateral field angled to avoid the spinal cord. The lower
half of the PTV which includes the subpyloric, pancreatico-​duodenal, and local para-​
aortic nodes is treated using a 3-​field arrangement comprising right and left lateral
fields and an anterior field angled to minimize the dose to the kidneys. The coeliac,
suprapancreatic, and porta hepatis nodes are included in either the upper or lower
half of the PTV, depending upon individual patient anatomy, as well as the level at
which the PTV is split. This technique involves the use of a single isocentre placed at
the level of the split and asymmetric collimator jaws in the superior–​inferior direction
to achieve an effective match-​line.
Check films on the simulator with intravenous contrast to image the kidneys
and portal imaging within the first 3  days of treatment are recommended. Quality
 Carcinoma of the stomach 157

assurance for radiotherapy was an essential component of the Macdonald/​Intergroup

trial. At review, 35% of the treatment plans were found to contain major or minor
deviations from the protocol, most of which were corrected before the start of radio-
therapy. Given the complexity and potential toxicity of therapy it seems appropriate
that there should be at least internal institutional review of all such treated cases when
a centre starts these treatments.
Chemotherapy (5FU, 425 mg/​m2/​day, and folinic acid, 20 mg mg/​m2/​day for 5 days)
is given on day 1 and followed by chemo-​radiotherapy beginning 28 days after the start
of the initial cycle of chemotherapy.
CRT consists of 45 Gy radiation delivered in 25 fractions, 1.8 Gy per day, treating
5 days per week for 5 weeks, with 5FU (400 mg/​m2/​day) and folinic acid (20 mg/​m2/​
day) on the first four and the last three days of radiotherapy.
One month after the completion of radiotherapy, two 5-​day cycles of 5FU (425 mg/​
m2/​day) plus folinic acid (20 mg/​m2/​day) are given 1 month apart. The dose of 5FU is
reduced in patients who have documented grade 3 or 4 toxicity.
Current trials examining the substitution of single agent 5FU with combinations
including cisplatin have not currently demonstrated any survival advantage.
Radiation doses should be restricted so that:
◆ Less than 60% of the hepatic volume is exposed to more than 30 Gy.
◆ The equivalent of at least two-​thirds of one kidney is spared from the field of
radiation.
◆ No portion of the heart representing 30% of the cardiac volume receives more than
40 Gy of radiation.

7.3.3 Palliative treatment
The majority of patients with gastric cancer present with advanced or metastatic dis-
ease. Chemotherapy has an established role in providing palliation for patients with
gastric cancer with evidence of prolongation of quality and quantity of life. There are
no randomized published series employing palliative radiotherapy for gastric cancer.
The literature contains series describing 50–​75% patients obtaining palliation for
symptoms by the use of radiotherapy with or without chemotherapy. The median dur-
ation of palliation varies from 4 to 18  months. Recognized symptoms that may be
palliated include bleeding and pain due to local tumour infiltration. Whilst the use
of radiation in obstruction is described in the literature such therapy has now been
superseded by the radiological placement of stents
The volume is determined by simulation with barium and includes known tumour,
which can be assessed by endoscopic description aided by CT films. Often the whole
stomach requires irradiation.
Palliative radiotherapy is given by anterior and posterior opposing fields with the
patient in the supine position and arms adducted. Field margins are marked on the
patients skin and the central point marked by tattoo.
Bleeding can often be stopped by relatively modest doses of irradiation.
Commonly applied treatments comprise 20 Gy in 5 fractions in 1 week. In very
unwell patients a single fraction of 8–​10 Gy can often stop bleeding within a few
158 Upper gastrointestinal tract

days. Some authors advocate higher doses (such as 40 Gy in 2 Gy daily fractions

for pain relief).

7.4  Carcinoma of pancreas

Pancreatic cancer has a poor prognosis with an overall survival rate of less than 5%.
Surgery is considered the only potentially curative treatment for pancreas cancer but
no more than 10–​20% of patients are able to undergo resection. Fifty to sixty percent
of patients present with metastatic disease and approximately 30–​40% have locally
advanced disease, usually defined in terms of vascular encasement not amenable to
surgical resection and/​or vascular reconstruction. The use of radiotherapy in all stages
of pancreatic cancer is controversial and the data is of generally poor quality.

7.4.1 Indications
Radiotherapy usually with chemotherapy may be considered:
◆ As neo-​adjuvant therapy in particular when the disease is considered borderline
resectable
◆ In the post-​operative setting (adjuvant) particularly when the surgical margins are
involved
◆ As primary palliative therapy for clearly unresectable localized disease with re-​con-
sideration of surgery in good responders.

Neoadjuvant therapy
A meta-​analysis of 111 retrospective studies concluded that approximately one-​
third of initially staged non-​resectable tumour patients could be expected to have
resectable tumours following neoadjuvant therapy, with similar survival but higher
resection-​associated morbidity and mortality rates when compared with initially re-
sectable tumour patients. There is, however, no randomized data to support this
finding(14) and its use remains uncertain. Current trials such as ESPAC 5 address the
issue of neo-​adjuvant chemotherapy and chemo-​radiotherapy in borderline resect-
able disease.

Adjuvant therapy
A meta-​analysis from Stocken et al. (15) included five randomized controlled trials of
adjuvant chemoradiation and adjuvant chemotherapy. Adjuvant chemotherapy im-
proved survival in patients with R0 resections. The 2-​and 5-​year survival rates were
estimated at 38% and 19%, respectively, with chemotherapy and 28% and 12% without.
This benefit was not seen with adjuvant chemoradiation (HR = 1.09, 95% CI 0.89–​1.32,
p = 0.43). The group concluded that adjuvant chemoradiation is only more effective
than chemotherapy alone after R1-​resections.

Palliative therapy
Studies comparing chemotherapy alone or CRT have yielded inconsistent results.
Several studies have now shown that selected patients who attain stable or partially
responsive disease after 3–​4  months of induction chemotherapy may benefit from
 Carcinoma of pancreas 159

consolidation CRT thus excluding the up to 35% patients with initially rapid progres-
sive disease that is treatment refractory.
The LAP07 trial compared chemotherapy alone (gemcitabine with/​without erlotinib)
vs the same chemotherapy for 4 months followed by consolidation capecitabine‐based
CRT(16). The study closed early following a planned interim analysis after 442 patients.
This study showed no improvement in survival applying consolidation CRT over con-
tinuing chemotherapy alone (median overall survival 15.2 vs 16.4 months).
In the UK, the SCALOP trial randomized patients to gemcitabine or capecitabine-​
based CRT following 4 months of induction GEMCAP chemotherapy(17). One hun-
dred and fourteen patients were recruited from 28 centres over 2  years, and 74
non-​progressive patients were randomized to CRT. SCALOP suggested superiority
of capecitabine‐based CRT over gemcitabine based‐CRT in terms of overall survival
(median overall survival 15.2 vs 13.4 months, HR 0.39, p = 0.012).
Current opinion is that both chemotherapy alone and consolidation CRT are op-
tions for this group of patients and studies are investigating intensifying both radio-
therapy and chemotherapy.

7.4.2  Treatment volume and definition

There is a wide variation in the practice of pancreatic RT in international literature
with no consensus on definition of PTV, radiation dose, and acceptable dose to organs
at risk (OARs).
Magnetic resonance imaging scans have not shown clear advantages over CT scan-
ning for volume localization in surgical series(18). FDG PET may also be useful because
an unexpected distant metastasis can be detected by whole-​body PET in about 40% of
the cases. However, active and chronic pancreatitis and autoimmune pancreatitis may
sometimes show high FDG accumulation and may mimic pancreatic cancer(19).
GTV includes the primary tumour as visualized on a contrasted planning CT scan
along with lymph nodes > 1 cm diameter identified on pre-​treatment scans.
The CTV is defined as the GTV either defined on 3D CT or for 4D planning on
using a Boolean union of GTV_​T, GTV_​N, GTV_​T_​inhale, GTV_​N_​inhale, GTV_​
T_​exhale, and GTV_​N_​exhale with 5 mm margin in all directions. The CTV is then
edited to exclude areas of overlap with uninvolved GI tract.
The PTV is defined as the CTV with:
◆ Cranial: 0.5 cm (exhale breath-​hold) or 1.5cm (free breathing)
◆ Caudal: 1.5 cm.
◆ Ant-​post and lateral: 1.0 cm.
◆ For 4DCT planning 0.5 cm in all directions.

7.4.3 Treatment techniques
Patients should be planned and treated in the supine position with their arms above
their heads and immobilized, ideally with the use of a chest-​board and knee-​fix. For
the CT planning scan, three horizontally aligned tattoos are marked at the right, an-
terior, and left-​hand surfaces. Typical 3D conformally planned conventional field
160 Upper gastrointestinal tract

arrangements include an anterior open field and two lateral wedged fields usu-
ally at an angle of 90°. This angle may need to be modified to reduce renal dose
(Fig. 7.5). The lateral fields may be weighted in order to reduce the exit dose to the
spinal cord from the anterior field. If the volume lies to the right of the midline,
anterior and right lateral wedged fields may be used to spare the small bowel and
left kidney. An isocentric technique is employed on a linear accelerator and 6–​8 MV
photons are used.
Verification can take place on the simulator and by using on treatment portal views.
Anterior fields are relatively easy to verify but images of the lateral or oblique fields are
less easy to interpret
The use of IMRT or VMAT is attractive to simultaneously allow dose-​escalation and
further enhance loco-​regional control, but also to limit doses in particular to the small
bowel and kidneys. A  hypothetical planning study from the MD Anderson Cancer
Center group has demonstrated that dose to the pancreas can be escalated to 72 Gy/​
36 fractions based on relationship of GTV to the GI‐OARs, using IMRT or proton
therapy(20). However, it remains unknown if dose escalation with better treatment de-
livery will translate clinically into improved local control.
Intraoperative RT (IORT) allows dose-​limiting normal structures, such as the
bowel, to be physically moved out of the radiation field. A randomized trial by the
National Cancer Institute in just 24 patients ()suggested an improvement in local con-
trol with the use of 20 Gy IORT following surgical resection compared to standard
therapy but no improvement in survival has been shown in any setting of pancreas
cancer(21).

Fig. 7.5  3-​field
plan for treatment
of carcinoma of
pancreas.
 Gall bladder and biliary tree 161

Stereotactic body RT (SBRT). The duodenum is the primary dose-​limiting normal

tissue for radiotherapy and pancreas cancer. One major concern with SBRT is thus
small bowel ulceration, perforation, or obstruction. Doses used include 30 Gy in three
fractions. The Stanford group using a 25 Gy single-​dose SBRT have reported an overall
rate of freedom from local progression at 6 months and 12 months of 91% and 84%,
respectively accompanied by grades ≥ 2 late largely GI toxicity of 11% and 25%(22). The
addition of an SBRT boost to 45 Gy conventional fractionated RT has also been in-
vestigated by the Stanford group, yielding a 94% local control rate and a 12.5% rate of
late duodenal ulcers. Although local control rates have been encouraging (57–​100%),
median survival times in these trials have not been substantially different compared
to historical controls (5.4–​14  months), primarily due to the development of distant
disease(23). Well-​conducted phase III trials are required given the not inconsiderable
late toxicity rates.

7.4.4  Dose prescription

The dose is limited by critical structures (Table 7.3).When the dose exceeds 50 Gy with
chemotherapy there is a 10% risk of small bowel damage most commonly seen in the
duodenal loop and manifested by bleeding.
◆ Typical dose prescription is 50.4 Gy in 1.8Gy daily fractions single phase.
◆ Concomitant chemotherapy comprises:
o Capecitabine 830mg/​m2 bd po, Monday–​Friday.

7.5  Gall bladder and biliary tree

Carcinomas of the gall bladder and extrahepatic biliary tree are uncommon and
account for about 4% of all GI tumours. The principal therapeutic measure for these
tumours remains surgical, whether by excision of the primary tumour, or where neces-
sary palliative stenting and bypass procedures to relieve obstructive jaundice.
Given the relative rarity of these tumours the role of radiotherapy is uncertain with
the literature containing small, often single institution, studies describing patients with
both gallbladder cancer and cholangiocarcinoma and no randomized data. There is

Table 7.3  Dose-​volume constraints

Dose volume constraints Dose constraint
Region of interest/​organ at risk
PTV V95% > 99.0%

DMAX < 107%

Spinal Cord PRV V40 Gy < 0%
Liver V30 Gy < 30%
Kidney receiving the higher dose V20 Gy < 40%
Combined kidney V20 Gy < 30%
162 Upper gastrointestinal tract

currently no primary role for external beam radiotherapy as curative treatment, and its
use is limited to either the post-​operative adjuvant setting or for palliation.

7.5.1  Post-​operative adjuvant treatment

Indications
Even after apparently curative resection loco-​regional failure is common for both gall
bladder and extrahepatic bile duct lesions. In bile duct cancer, proximal and distal
margins are often narrow, as are the circumferential margins if lesions extend through
the entire duct wall. For ductal lesions, local failure is the commonest cause of death.
Overall, after simple or extended cholecystectomy for gall bladder cancer, 75–​85% of
patients with early recurrence will die with or because of local failure. There is there-
fore a rationale for employing post-​operative radiation therapy.

Treatment volume and definition

A major deterrent to the use of radiation is the limited tolerance of surrounding struc-
tures in particular liver, duodenum, jejunum, and stomach. The presence of surgical
clips eases difficulties in identifying residual disease. It is helpful to plan treatment
with the referring surgeon present to clarify both the post-​operative anatomy and
areas at risk of relapse.

Positioning
The patient is treated supine arms above head and holding an arm pole, with the
trunk immobilized in an individualized bean-​bag. Patient alignment is from lateral
tattoos.

Planning technique
The CTV is defined as the tumour bed, any residual tumour, lymph node drainage
along the porta hepatis, pancreatico-​duodenal system, and coeliac axis.
One phase technique uses a three-​field plan using one anterior and two angled lateral
oblique fields. Other techniques employ shrinking field techniques where large initial
volumes need to be reduced at 35 Gy to avoid radiation-​induced hepatitis. VMAT
techniques have demonstrated better normal tissue sparing and dose conformity than
3D conformal therapy but this has not translated as yet into better clinical outcomes
or completed dose escalation studies.

Verification
Verification can take place on the simulator and/​or using on treatment portal views.
Whilst anterior fields are relatively easy to verify, images are also routinely taken of
the lateral fields, which in the usual absence of reliable landmarks are difficult to in-
terpret and probably non-​contributory. Image-​guided radiation techniques with soft
tissue algorithms in development offer the prospect of significantly improving quality
of radiotherapy delivery.

Prescription
Tolerance and potential for morbidity dictates dose.
 further reading/References 163

◆ 50 Gy in 25 fractions in 5 weeks, single phase.

◆ Concomitant chemotherapy where applicable comprises:
• 5FU 500 mg/​m2 days 1–​3 with folinic acid 20 mg/​m2 days 1–​3 weeks 1 and 5
• OR gemcitabine 300mg/​m2 weekly (IV)
• OR capecitabine 830mg/​m2 bd po, Monday–​Friday.
An alternative when large volume treatments are employed is 45 Gy in 25 fractions in
5 weeks to the large volume followed by a boost using multiple field, arc techniques to
a total dose of 55 to 60 Gy over 6–​7 weeks.
Buskirk et al. (24) showed that with total doses of 55 Gy or less, the risk of severe GI
complications was 5–​10%. At doses greater than 55 Gy one-​third of patients developed
severe problems. In rabbits significant biliary fibrosis is seen at doses greater than 30
Gy. Temporary biliary tree fibrosis with consequent secondary biliary cirrhosis is rec-
ognized usually resolving within 18–​24 months of treatment

7.5.2 Palliative treatment
Palliative treatments are rarely employed. Intra-​luminal brachytherapy may be used
for palliation and its use has been facilitated by the development of smaller more flex-
ible catheters for use with high-​dose rate machines.
Significant palliation can be achieved using techniques identical to the post-​operative
adjuvant therapies discussed above. Permanent local control is uncommon and doses
are usually limited to 40–​60 Gy over 4.5–​7 weeks. Fluouropyrimidines or more re-
cently gemcitabine chemotherapy is commonly co-​administered with treatment.
Currently evidence about the use of SABR for cholangiocarcinoma is very limited.
Current reported studies contain fewer than 90 participants. All are uncontrolled and
provide little evidence to support the treatment’s effectiveness

Further reading/​references
1. Wong R, Malthaner R. Combined chemotherapy and radiotherapy (without surgery)
compared with radiotherapy alone in localized carcinoma of the esophagus (Cochrane
Review). The Cochrane Library, Issue 2, Chichester, UK. John Wiley & Sons, Ltd. 2004.
2. Rice PF, Crosby TL, Roberts SA. Variability of the carina-​incisor distance as assessed by
endoscopic ultrasound. Clinical Oncology 2003; 15:383–​5.
3. Yap JC, Malhotra HK, Yang GY. Intensity modulated radiation therapy in the treatment of
esophageal cancer. Thoracic Cancer 2010; 1:62–​9
4. Minsky B, Pajak T, Ginsberg R, et al. INT 0123 (Radiation Therapy Oncology Group 94-​
05) Phase III trial of combined-​modality therapy for esophageal cancer: high-​dose versus
standard-​dose radiation therapy. 2002; Journal of Clinical Oncology, 20 (5), 1167–​74.
5. Aisner J, Forastiere A, Aroney R. Patterns of recurrence for cancer of the lung and
esophagus. In Wittes RE, ed, Cancer Treatment Symposia: Proceedings of the Workshop
on Patterns of Failure After Cancer Treatment, Vol 2, p87. US Department of Health and
Human Services, Washington DC. 1983
6. Arnott SJ, Duncan W, Gignoux M, et al. (Oeosphageal Cancer Collaborative Group).
Preoperative radiotherapy for esophageal carcinoma (Cochrane Review). In: The Cochrane
Library, Issue 2, Chichester, UK: John Wiley & Sons, Ltd. 2004.
164 Upper gastrointestinal tract

7. Allum WH, Stenning SP, Bancewicz J, et al. Long-​term results of a randomized trial of
surgery with or without preoperative chemotherapy in oesophageal cancer. Journal of
Clinical Oncology 2009; 27:5062.
8. Gebski V, Burmeister B, Smithers BM, et al. Australasian Gastro-​Intestinal Trials Group.
Survival benefits from neo-​adjuvant chemoradiotherapy or chemotherapy in esophageal
cancer—​a meta-​analysis. 2007; Lancet Oncology 8:226–​34.
9. Courrech Staal EF, Aleman BM, Boot H, et al. Systematic review of the benefits and risks
of neoadjuvant chemoradiation for oesophageal cancer. British Journal of Surgery 2010;
97:1482–​96.
10. van Hagen P, Hulshof MCCM, van Lanschot JJB, et al. Preoperative chemoradiotherapy
for esophageal or junctional cancer. New England Journal of Medicine 2012; 366:2074–​84.
11. Hallissey MT, Dunn JA, Ward LC, Allum WH. The second British Stomach Cancer Group
trial of adjuvant radiotherapy or chemotherapy in resectable gastric cancer: five-​year
follow-​up. 1994; Lancet 343:1309–​12.
12. Macdonald JS, Smalley SR, Benedetti, J, et al. Chemoradiotherapy after surgery compared
with surgery alone for adenocarcinoma of the stomach or gastroesophageal junction. New
England Journal of Medicine 2001; 345:725–​30.
13. Leong T, Willis D, Joon DL, et al. 3D conformal radiotherapy for gastric cancer—​results of
a comparative planning study. Radiotherapy and Oncology 2005; 74:301–​6.
14. Gillen S, Schuster T, Meyer Zum Büschenfelde C, et al. Preoperative/​neoadjuvant therapy
in pancreatic cancer: a systematic review and meta-​analysis of response and resection
percentages. PLoS Medicine 2010; 7:e1000267
15. Stocken DD, Büchler MW, Dervenis C et al. Meta-​analysis of randomised adjuvant
therapy trials for pancreatic cancer. British Journal of Cancer 2005; 92: 1372–​81.
16. Hammel P, Huguet F, Van Laethem J-​L, et al. Comparison of chemoradiotherapy (CRT) and
chemotherapy (CT) in patients with a locally advanced pancreatic cancer (LAPC) controlled
after 4 months of gemcitabine with or without erlotinib: Final results of the international
phase III LAP 07 study. ASCO Meeting Abstracts, 2013. 31(18_​suppl): p. LBA4003.
17. Mukherjee S, Hurt CN, Bridgewater J, et al. Gemcitabine‐based or capecitabine‐based
chemoradiotherapy for locally advanced pancreatic cancer (SCALOP): a multicentre,
randomised, phase 2 trial. Lancet Oncology 2013; 14:317–​26.
18. Pauls S, Sokiranski R, Schwarz M, et al. Value of spiral CT and MRI (1.5 T) in
preoperative diagnosis of tumors of the head of the pancreas. Rontgenpraxis 2003; 55:3–​15.
19. Higashi T, Saga T, Nakamoto Y, et al. Diagnosis of pancreatic cancer using fluorine-​
18 fluorodeoxyglucose positron emission tomography (FDG PET) -​-​usefulness and
limitations in ‘clinical reality’. Annals of Nuclear Medicine 2003; 17:261–​79.
20. Bouchard M, Amos RA, Briere TM, et al. Dose escalation with proton or photon radiation
treatment for pancreatic cancer. Radiotherapy and Oncology 2009; 92:238–​43.
21. Sindelar WF, Kinsella TJ. Studies of intraoperative radiotherapy in carcinoma of the
pancreas Annals of Oncology 1999; 10 Suppl. 4: S226–​3.
22. Chang DT, Schellenberg D, Shen J. Stereotactic radiotherapy for unresectable
adenocarcinoma of the pancreas. Cancer 2009; 115: 665–​72.
23. Koong AC, Le QT, Ho A, et al. Phase II study to assess the efficacy of conventionally
fractionated radiotherapy followed by a stereotactic radiosurgery boost in patients with
locally advanced pancreatic cancer. International Journal of Radiation, Oncology, Biology,
Physics 2005; 63:320–​3.
24. Buskirk SJ, Gunderson LL, Adson MA et al. Analysis of failure following curative
irradiation of gallbladder and extrahepatic bile duct carcinoma. International Journal of
Radiation, Oncology, Biology, Physics 1984; 10: 2013.
 Chapter 8

Rectal cancer
Rob Glynne-​Jones and Mark Harrison

8.1 Introduction
Approximately 15,000 patients in the UK develop rectal cancer each year. Unlike the
colonic portion of the large bowel, the majority of the rectum lies below the peritoneal
reflection and has no serosa. This anatomical feature allows tumour to penetrate easily
and deeply into perirectal fat. An additional issue is that the rectum lies within the
bony anatomy of the pelvis and offers a very narrow space for dissection in males.
Historically, these factors have previously been associated with a high risk of loco-​
regional failure within the pelvis. In addition, surgery during the 1960s to 1990s did
not use meticulous sharp dissection along mesorectal and levator planes as currently
practised today.
Surgery can be curative as a single modality, but a multimodality approach is now
considered the most effective management, and different strategies have developed
independently in different countries to prevent local recurrence and reduce distant
metastases.
The precise definition of the upper rectum influences the reported local recurrence
rate, since the risks of local failure are much higher for cancers in the lower rectum. In
the UK rectal cancers are categorized as low (up to 5 cm), middle (from >5 to 10 cm)
or upper (from >10 up to the upper limit of 15 cm from the anal verge (as measured
by rigid sigmoidoscopy).
The past 20 years have seen a significant evolution in surgical practice, particularly
with the technique of meticulous mesorectal dissection. The surgeon now removes
all of the surrounding mesorectal fat en bloc in a neat anatomical package, which re-
moves all lymph nodes within the mesorectum and is associated with lower rates of an
involved circumferential resection margin (CRM) and consequently less local recur-
rence, and population data show an improved survival. This technique is termed total
mesorectal excision (TME). As survival has increased and local recurrence rates have
fallen, long-​term outcomes in terms of function, late effects, and quality of life have
become an increasingly important endpoint to evaluate new treatments.
The CRM or radial margin refers to the non‐peritonealized bare area of the rectum
in the surgical specimen. The presence of microscopic tumour cells (primary tumour
or lymph node—​either as continuous or discontinuous extension) within 1 mm of the
CRM is associated with a high rate of local recurrence and poor survival(1).
In the post TME era local recurrence is more frequent after partial mesorectal ex-
cision and as a result of inadequate surgery (2). The quality of surgery is therefore a
166 Rectal cancer

strong quality control measure and histopathological examination should include an

independent photographic record of the surgical specimen, which is then scored for
TME quality (see Table 8.1), and the CRM should be measured in mm in the report (1).
A proforma report is therefore recommended, but may need expansion to clarify CRM
and the residual (R) tumour classification.
Surgery remains the mainstay of treatment, yet the optimal multimodality man-
agement of stages II and III rectal cancer currently remains an increasing challenge.
Short-​course preoperative radiotherapy (SCPRT) and chemoradiotherapy (CRT) are
frequently combined with surgery. Yet, apart from the results of the Swedish Rectal
Cancer Trial(3), achieving better local control has not impacted on overall survival
(OS). Hence intensification or the additional use of systemic chemotherapy is being
explored, despite persistent controversy regarding the utility of postoperative adjuvant
chemotherapy.
Optimal results are achieved by a specialized multidisciplinary team (MDT) of
radiologists, surgeons, radiation oncologists, medical oncologists, and pathologists at-
tending regular meetings. Core members should be available for the discussion of all
relevant cases. The MDT should take clinical guidelines into account in making deci-
sions, but consider the individual circumstances. We describe the current ESMO rectal
cancer guidelines(4) and options for integrating chemotherapy preoperatively in the
total neo-​adjuvant treatment (TNT) approach endorsed by the NCCN guidelines(5).
This chapter assesses the role of radiation therapy in rectal cancer, with an emphasis
on preoperative imaging with magnetic resonance imaging (MRI). We discuss patient
selection for preoperative CRT and SCPRT, and postoperative CRT. Treatment deci-
sions regarding neoadjuvant therapy in the UK are mainly based on this pre-​operative
MRI and computed tomography (CT) imaging, but post-​operative histopathological
tumour, node, metastasis (TNM) staging determines the need for adjuvant chemo-
therapy or CRT.
We describe the various available conformal planning techniques. More complex
techniques such as intensity-​modulated radiotherapy (IMRT), volume-​modulated arc

Table 8.1  Grading of quality and completeness of the mesorectum in a total mesorectal
excision specimen
Mesorectum Defects Coning MRF
Complete Intact, smooth Not deeper than 5 mm None Smooth,
regular

Nearly Moderate bulk, No visible muscularis Moderate Irregular

complete irregular propria
Incomplete Little bulk Down to muscularis Moderate–​ Irregular
propria marked

Both the specimen as a whole (fresh) and cross-​sectional slices (fixed) are examined in order to make an
adequate interpretation.
Reproduced with permission from Glynne-​Jones R. (2014) Do T3 Rectal Cancers Always Need
Radiochemotherapy? In: Otto F., Lutz M. (eds) Early Gastrointestinal Cancers II: Rectal Cancer. Recent Results in
Cancer Research, vol 203. Springer, Cham. Copyright © 2014, Springer International Publishing Switzerland
 Introduction 167

therapy (VMAT), and brachytherapy are also described. In addition, chemoradiation

and radiotherapy as an adjunct to local excision and endoluminal irradiation are also
reviewed.

8.1.1 Staging
Approximately 20% of patients with colorectal cancer present with metastases at
diagnosis, and this figure has remained consistent over the last two decades(6). A his-
tory and physical examination requires a digital rectal examination if the tumour is
low. Additional tests include full blood count, liver and renal function tests, serum
carcinoembryonic antigen (CEA), as well as imaging (see below).These results should
help define functional status and presence of metastases.
Rectal cancer is classified according to the tumour, node, metastasis (TNM) system.
The current version is TNM 8. Further categorization of cT3 cancers into T3-​substages
according to the depth of penetration into the muscularis propria can define the risk
of involved lymph nodes and predict both local recurrence and/​or synchronous and
subsequent metastatic disease (see Table 8.2).

8.1.2 Imaging
Accurate information on primary tumour local extension, the precise location,
clinical nodal-​stage (inside and outside the mesorectal fascia (MRF)), potential
CRM involvement, and presence of extramural venous invasion is essential for
defining the optimum treatment strategy on an individual basis. High-​resolution
pelvic MRI is routine in the UK as a preoperative staging and selection tool for
the use of preoperative radiation and the use of systemic chemotherapy pre-​and
post-​surgery. MRI can predict the extramural depth of invasion, the likelihood of
involvement of the CRM or MRF particularly in the mid-​rectum, and involvement
of the levators in the low rectum. The depth of penetration usually can be accur-
ately determined, but occasionally bulky T3 /​T4 cancers extending beyond MRF
distort the anatomy with a pushing border and make it difficult to capture precise
measurements.

Table 8.2  Subclassification of T3 rectal cancer

Subclassification of T3 rectal cancer
Depth of invasion in mm beyond the
muscularis propria

T3a <1mm
T3b 1-​5mm
T3c 6-​15mm
T3d >15mm

Reproduced with permission from Edge, S.B. & Compton, C.C. Ann Surg
Oncol (2010) 17: 1471. https://​doi.org/​10.1245/​s10434-​010-​0985-​4. ©
Society of Surgical Oncology 2010
168 Rectal cancer

Preoperative assessment can identify patients at risk of an R1 resection and can ac-
curately predict ultimate outcome. The Mercury Study Group used preoperative MRI
to extend the clinical subclassification of T3 into four groups:  ‘a’ (< 1  mm outside
the wall), ‘b’ (1–​5 mm), ‘c’ (5–​15 mm), and ‘d’ (> 15 mm). Several systems are cur-
rently used for classification before(7,8) and after treatment(9,10,11). MRI assists the MDT
in defining the required extent of surgery to ensure an R0 resection. RI can also define
MRI extramural vascular invasion (EMVI), which can predict the risk of synchronous
and metachronous metastatic disease(7,12).
However, MRI, multislice CT, and endoscopic rectal ultrasound are all equally in-
accurate for detecting involved lymph nodes (13,14) despite specific imaging features such
as size/​round/​heterogenous/​irregular in nodal border. In one pathological study 95/​
334 (28%) positive nodes were ≤ 3 mm in diameter (15). Hence, reliance simply on the
size of nodes over 10 mm as a criterion has poor accuracy in predicting node-​positive
disease. More reliable features are shape, irregular border, and heterogeneous signal
within the node, although the nodal status may be less relevant to decision-​making if
the surgeon routinely performs good quality TME and removes the mesorectal nodes
en bloc.
CT scan of thorax and abdomen can determine metastatic disease. Fludeoxyglucose
positron emission tomography (FDG-​PET) has not improved the accuracy, but may
lead to nodal upstaging. PET may provide additional information in terms of disease
outside the pelvis, but there is currently no consensus as to the patient population with
the most to gain from this strategy.

8.1.3 
Choice of preoperative treatment—​short-​course
preoperative radiotherapy (5 × 5 Gy) or chemoradiation
There are two preoperative radiation approaches commonly used in the UK. These
are SCPRT with immediate surgery within days of completion and long-​course pre-
operative CRT with an interval to surgery of 6–​12 weeks. Current trials suggest that
in resectable cancers, where the preoperative MRI suggests the MRF is not potentially
involved, these schedules of SCPRT and CRT are equivalent in terms of outcomes such
as local recurrence, disease-​free survival (DFS) and OS (16,17). A recent meta-​analysis
suggests SCPRT with delayed surgery is as effective as CRT with delayed surgery in
terms of sphincter preservation, local recurrence, grade 3/​4 acute toxicity, R0 resec-
tion rate, and down-​staging (18). For more advanced cases, where the CRM/​MRF is
breached or threatened according to the MRI, the addition of 5-​fluorouracil (5FU)
to radiation has favourable effects on resectability, relapse-​free survival, and cancer-​
specific survival with a trend to improve OS(19).
However, the pattern of care is changing as SCPRT can be delivered with a delay
before surgery to allow response. This strategy is useful in elderly and frail patients
and where 5FU may be relatively contra-​indicated because of cardiac problems. In
the Stockholm III randomized phase III trial, three arms were compared in resectable
cancers: SCPRT and immediate surgery, SCPRT and a delay of 6–​8 weeks, and long-​
course radiotherapy alone. The median time between the start of SCPRT and surgery
was 8 days for immediate surgery and 45 days for the arm with a delay of 4–​8 weeks (20).
 Introduction 169

There was a significant reduction in postoperative complications between these arms

(HR 0.61, p = 0·001) and a non-​significant reduction in inter-​current deaths with 57/​
357 versus 47/​355 for the arm with delay respectively (p = 0.13), but no difference in
local recurrence or OS (HR 0·90 (0·70–​1·15) p = 0·46).
SCPRT followed by systemic chemotherapy has been used as the novel arm of
the Rectal Cancer and Preoperative Induction Therapy Followed by Dedicated
Operation (RAPIDO) trial (NCT01558921). RAPIDO randomly allocated patients
with more locally advanced rectal cancer according to MRI criteria (CRM threat-
ened, t3c/​d, EMVI, etc.) either to SCPRT followed by six cycles of oral capecitabine
with intravenous oxaliplatin, and then TME or the control arm of standard pre-
operative long-​course CRT followed by TME, and the option of postoperative
chemotherapy. The trial has completed recruitment but comparative results are not
yet available.

8.1.4  Short-​course preoperative radiotherapy

SCPRT employs a short intensive course of radiotherapy of 25 Gy over 5 days fol-
lowed by immediate surgery within 2–​5  days, with the aim of reducing the risk
of pelvic recurrence. This approach has gained widespread acceptance in many
Northern European and UK centres after the publication of the Swedish, Dutch,
and UK rectal cancer trials. The landmark Swedish rectal cancer trial randomized
patients to surgery or SCPRT followed by surgery and demonstrated not only a
significant reduction in local recurrence but also a 10% absolute improvement in
survival8.
Subsequently, the Dutch trial(21,22) examined the routine use of SCPRT followed by
TME, against TME alone (and selective postoperative radiotherapy in the event of
histopathological evidence of involvement of the CRM). This trial demonstrated local
recurrence rates at 43 months of 4.1% after SCPRT compared with 11.5% after initial
TME alone. The 10-​year local recurrence cumulative incidence was 5% in the group
assigned to SCPRT versus 11% in the surgery-​alone group (p < 0.001) demonstrating
a sustained long-​term improvement in local control. Post-​hoc analysis of the subset of
patients with cancers in the mid-​rectum with histology showing a negative CRM (>
1 mm) and stage III showed a 10% benefit (from 40% to 50%) in OS(23). However, the
overall results of the Dutch trial do not show a difference in survival which implies that
some groups may be disadvantaged in terms of survival by radiotherapy.
The MRC CR07 trial(1,24) randomized 1350 rectal cancer patients to either SCPRT (5
× 5 Gy) or selective postoperative chemoradiation (25 × 1.8 Gy with concurrent 5FU)
administered only for patients with histologically involved (≤ 1 mm) resection mar-
gins. Overall clinically significant absolute risk reduction in 3-​year local recurrence
rate of 6.2% was observed, corresponding to a relative risk reduction of 61%. Early
results showed DFS was 6% better in the preoperative radiotherapy group, but there
was no improvement in OS. The CR07 trial suggests SCPRT is of some benefit in terms
lowering local recurrence for all tumour locations, all pathological stages, and good,
average, or poor quality surgery—​although the effect is minimal when good quality
mesorectal excision is performed (1).
170 Rectal cancer

8.1.5 Chemoradiation
Since the early 1980s, 5FU alone, and more recently combinations of cytotoxic
chemotherapy using oxaliplatin or irinotecan, have formed the basis of chemo-
therapy treatment for patients with metastatic colorectal cancer. Historical
postoperative studies in the 1980s examined 5FU-​based chemotherapy and radio-
therapy or their combination, and showed a significant benefit for the concurrent
chemoradiation(25,26). In the EORTC 22921 trial, the addition of 5FU to radiation
tripled the rate of complete pathological response (pCR) from 4 to 13% (27), but CRT
did not improve OS.
With the introduction of improved preoperative imaging (CT, transrectal ultra-
sound, and MRI) to stage the patient, the strategy of postoperative chemoradiation
has been extrapolated to the preoperative setting with less acute and late toxicity. In
the German Trial CAO/​ARO/​AIO—​94(28) a total of 823 patients were randomized
between preoperative CRT and postoperative CRT (patients received postoperative
adjuvant chemotherapy in both arms of this trial). Acute and late toxicity were signifi-
cantly reduced with the preoperative approach. Loco-​regional failure was 6% in the
preoperative arm vs 13% in the postoperative arm. There was, however, no difference
in the distant metastases rate or OS. Long-​term follow-​up showed that local recur-
rences were still occurring after 5 years and that preoperative CRT did not reduce the
risk of metastases or improve DFS and/​or OS(28).
Using 5FU-​based chemoradiation downstaging is commonly achieved, and between
10–​15% of patients will attain a pCR at surgery. The converse of this also true in that
that many still fail to respond sufficiently. When compared with radiation alone, 5FU-​
based chemoradiation achieves a higher pCR and has improved loco-​regional con-
trol, but does not improve DFS or OS. However, for more advanced clinically defined
unresectable/​borderline and recurrent cases, or when the preoperative MRI shows a
threatened or breached CRM, 5FU-​based chemoradiation has a statistically significant
effect on resectability and DFS.
The NSABP R03 (29) used a similar design but only recruited 267 of its planned pa-
tient target number (n = 900), so results should be interpreted with caution. In the
preoperative CRT arm 15% of patients achieved a pCR. Five-​year loco-​regional re-
currence was 10.7% in each treatment arm (p  =  0.693). A  significant improvement
of 5-​year DFS (65% vs 53%, p = 0.011), and a non-​significant improvement in 5-​year
OS (75% vs 66%, p = 0.065) were also observed for the preoperative arm. Since local
recurrence remained 10.7% in both arms, improvements in local control are unlikely
to be responsible.
This reduction in local recurrence from radiotherapy is balanced by worse outcomes
in terms of long-​term adverse consequences of surgery and radiotherapy. Symptoms
such as chronic pain, faecal incontinence, and sexual difficulties are reported in both
sexes. The ‘low anterior resection syndrome’ (LARS) is frequently reported by patients
and enhanced by the addition of SCPRT/​CRT. The gains in function achieved by a long
rectal remnant are lost if radiotherapy is added.
In addition to preventing local recurrence, much interest is growing concerning
preservation of the anal sphincter and avoiding a permanent stoma with organ
 Introduction 171

preservation with the increasing popularity of the ‘watch and wait’ non-​operative
strategy if a complete clinical response is achieved after CRT. Many, especially in the
USA, have considered CRT and the consequent downstaging/​downsizing achieved
will facilitate sphincter-​sparing surgery. Enthusiasts cite the results of the German
trial to support this argument, but specifically designed trials(16) and meta-​analyses
have failed to confirm that neoadjuvant chemoradiation facilitates sphincter sparing,
although a proportion will be amenable to local excision or watch and wait.
Interest in intensifying chemotherapy has been stimulated by the success of combin-
ation chemotherapy in advanced disease and the improvement in DFS and OS when
oxaliplatin is added to a fluoropyrimidine in the adjuvant setting in colon cancer.
However, to date these combinations with oxaliplatin in chemoradiation schedules
have not consistently improved either early pathological outcomes in resectable rectal
cancer or enhanced DFS or OS—​apart from the German CAO/​ARO/​AIO-​04 trial. In
this latter trial in the oxaliplatin arm pCR rate increased compared with the control
arm and the 3-​year DFS improved(30), but the design of the trial does not allow us to
distinguish whether oxaliplatin as the preoperative component, the postoperative ad-
juvant component or both impacted on the DFS.
Phase II studies examining the addition of irinotecan to standard fluoropyrimidine
based chemoradiation suggested increased pCR rates, but a randomized phase trial
(RTOG0012) showed no additional benefit from adding weekly irinotecan to con-
tinuous infusional 5FU and concurrent pelvic hyperfractionated radiation (31).
Currently in the UK a phase III randomized trial (ARISTOTLE) is examining the
benefit of incorporating irinotecan into chemoradiation with capecitabine, in MRI de-
fined unresectable or borderline resectable rectal cancers (www.controlled-​trials.com/​
ISRCTN09351447).
Molecularly targeted agents such as cetuximab, panitumumab, and bevacizumab
have also been integrated into standard chemotherapy regimens in colorectal cancer,
and therefore have been incorporated in phase I/​II studies into chemoradiation sched-
ules. However, although the observed increased toxicity is acceptable there is con-
sistent evidence that surgical morbidity is higher when bevacizumab has been used.
There is no current evidence to support additive effects. At present combination CRT
and the addition of a biological agent remain investigational.

8.1.6  Selection
of patients for short-​course preoperative
radiotherapy or chemoradiotherapy
In the UK patients have previously been categorized by MRI criteria into three
groups—​‘The good, the bad and the ugly’(32), which allows definition of three different
settings where preoperative neoadjuvant treatment may or may not be required. The
2011 colorectal guidelines from NICE describe three different risk groups of patients
with rectal cancer, defined by their possibility of local recurrence. These groups are
defined in Table 8.3.
More sophisticated risk adapted stage groupings are suggested in the recent ESMO
guidelines—​see Table 8.4 (4).
The appropriate choice of preoperative regimen is summarized in Fig. 8.1.
172 Rectal cancer

Table 8.3  Risk groups of patients with rectal cancer

Risk of local recurrence Characteristics of rectal tumours predicted by MRI
for rectal tumours as
predicted by MRI
High A threatened (<1 mm) or breached resection margin or Low
tumours encroaching onto the inter-​sphincteric plane or with
levator involvement

Moderate Any cT3b or greater, in which the potential surgical margin is not
threatened or
Any suspicious lymph node not threatening the surgical resection
margin or
The presence of extramural vascular invasion
Low cT1 or cT2 or cT3a and
No lymph node involvement

Reproduced with permission from CG 131 Colorectal cancer: The diagnosis and management of colorectal
cancer. London: NICE. Available from www.nice.org.uk/​guidance/​CG131, Copyright © 2011 National
Institute for Health and Clinical Excellence

8.1.7  Rendering unresectable tumours resectable

The majority of randomized trials in rectal cancer have been performed in patients
with resectable rectal cancer where the aim of treatment has been to lower the risk of
local recurrence, which has often formed the primary endpoint.
A minority of patients (approximately 10–​20% defined by MRI) will present with a
rectal tumour that is unlikely to be completely resected with a standard conventional
TME without leaving residual macroscopic or microscopic disease. It is essential that
these patients are identified before an (unsuccessful) attempt at surgical resection is
made. Once identified preoperative strategies to facilitate tumour shrinkage and an R0
resection can be initiated.
In those patients with locally advanced rectal cancer, the pelvic MRI can demon-
strate the relationship of the primary tumour to the surrounding mesorectal fascia
and the surrounding organs, such as bladder, vagina, and levators, as well as enlarged
lymph nodes outside the mesorectal fascia. Thus patients may be selected for pre-
operative chemoradiation in this category if there is evidence of:
◆ Primary tumour involving or extending beyond the mesorectal fascia
◆ Primary tumour within 1–​2 mm of the mesorectal fascia. Whether this should be 1
or 2 mm is debated
◆ Involved lymph nodes outside the mesorectal fascia (usually pelvic side wall or
iliac nodes)
◆ Invasion of other solid pelvic organs, i.e. prostate/​uterus.
There is general agreement that this group of patients should receive preoperative
CRT, but recent data suggest SCPRT followed by a fluoropyrimidine and oxaliplatin
based chemotherapy is equally effective and can be also applied.
 Introduction 173

Table 8.4  Recommended choice of treatment options within TNM risk category of primary
rectal cancer without distant metastases (from ESMO Guidelines 2017)
Risk group TN substage Possible therapeutic options
Very early cT1 sm1 N0 (on Local excision (TEM)
ERUS and MRI If pT1 and no adverse
features, TEM is sufficient
If adverse histopathology (sm
2, G3, V1, L1), requires radical
resection (TME) as standard
Further considerations
Alternatively, in the case
of adversefeatures on
pathology, TEM plus
salvage (or adjuvant) CRT
in periopera-​tive high-​risk
patients (butunproven
benefit—​with high riskof
local recurrence for pT2

Early cT1-​cT2; cT3a/​b if Surgery (TME) alone is For fragile, high-​risk

middle or high,N0 standard. patients or those
(or also cN1 if high), If unexpectedpoor prognostic rejecting radical surgery
MRFclear, no EMVI signs on histopathology (CRT with evaluation,
(CRM+,and or extranodal/​ local excision or if
N2), consider postoperative achieving cCR, ‘watch-​
CRT/​CT (see post-​operative and-​wait’,organ
recommendations in Table preservation
Intermediate /​ cT3a/​b very low, Surgery (TME) alone is a If CRT is given and cCR
More locally levators clear, standard only if good-​quality is achieved,’watch-​and-​
advanced MRF clear or cT3a/​ mesorectal resection assured wait’ in high-​risk patients
b in mid-​or high (and local recurrence <5% or, for surgery may be
rectum, cN1-​2 (not if not, preopertive SCPRT (5X5 considered
extranodal), no EMVI Gy) or CRT followed by TME
Locally cT3c/​d or very low Preoperative SCPRT (55cGy) If CRT and cCR achieved,
advanced localisation levators or CRT followed by TME, ‘watch-​and-​wait’ in
threatened, MRF clear depending on need for high-​risk patients may be
cT3c/​d mid-​rectum, regression considered
cN1–​N2, (extranodal),
EMVI+, limited
cT4aN0
Advanced (Ugly) cT3 with any MRF Preoperative CRT followed Alternatively, 5X5 Gy
involved, any cT4a/​b, by surgery (TME and more alone with a delay to
lateral node+ extended surgery if needed surgery in fragile/​elderly
due to tumour overgrowth), or in patients with severe
or preoperative SCPRT (5X5 comorbidity who cannot
Gy) plus FOLFOX and delay to tolerate CRT
surgery

Other factors besides T and N stages are relevant, such as EMVI, MRF involvement, distance from the anus and
sphincters, size of mesorectum and patient characteristics. Patient preferences are also important.
cCR, clinical complete response; CRM, circumferential resection margin; CRT, chemoradiotherapy; CT, computed
tomography; EMVI, extramural vascular invasion; ERUS, endoscopic rectal ultrasound; FOLFOX, leucovorin/​fluorouracil/​
oxaliplatin; MRF, mesorectal fascia; MRI, magnetic resonance imaging; SCPRT, short-​course preoperative radiotherapy;
TEM, transanal endoscopic microsurgery; TME, total mesorectal excision; TNM, tumour, node, metastasis.
Reproduced with permission from Glynne-​Jones R, Wyrwicz L, Tiret E et al., Rectal Cancer: ESMO Clinical Practice
Guidelines. Annals of Oncology 2017:28 (suppl 4): iv22–​iv40, Copyright © Oxford University Press 2017.
174 Rectal cancer

Patient with
rectal cancer

MRI to assess local recurrence

determined by anticipated resection
margin, tumour and lymph node staging,
unless contraindicated
Patient information and support

Risk of local recurrence

High risk (locally

Low risk Moderate risk
advanced)1

Consider Consider
Chemoradio-
therapy2

SCPRT

Interval before surgery to allow

shrinkage and response
Proceed immediately to

Surgery

See algorithm on “Post-operative care”

Fig. 8.1  Management of local disease—​patients with rectal cancer.

Current approaches in CRT use a fluoropyrimidine combined with radiation al-

though a considerable number of phase II and phase III studies of the integration of
combination chemotherapy schedules have recently been presented. To date, this ap-
proach remains investigational and preliminary results have been disappointing.

8.1.8 Postoperative chemoradiotherapy
The North American standard of care in the past was to deliver postoperative chemo-
therapy and CRT to all patients post-​operatively with T3/​4 or N + disease (approxi-
mately 80% of all resected cancers). This approach was defined within the NIH
consensus statement in 1990 based on the results of three USA trials. With publication
of the results of the German AIO/​ARO study, Europe and the USA extrapolated CRT
to the preoperative setting for patients who are considered T3/​4 or N + on the basis of
preoperative transrectal ultrasound.
 Introduction 175

Historical studies prior to TME suggest the postoperative histopathological fea-

tures which impact on the risk of local recurrence include: pathological TNM stage,
T-​substage (Table 8.3), CRM status, the number/​proportion of involved lymph nodes,
extracapsular extension, extranodal deposits, tumour differentiation, lymphovascular
invasion (LVI), EMVI, and perineural invasion (PNI). Hence, it is recommended that
pathologists review MRI scan reports when assessing EMVI status.
Histologically involved nodes have in the past been associated with a high risk of
local recurrence. However, the risk of local recurrence is reduced if the quality of
mesorectal excision is good (i.e. with a complete smooth mesorectum with no defects
and no coning) ensuring removal of all mesorectal lymph nodes.
Hence, postoperative CRT is used sparingly and selectively only in patients with
adverse histopathological features after surgery, which were not predicted by the sta-
ging MRI—​i.e. an involved CRM, perforation in the proximity of the tumour, an in-
complete i.e. poor quality mesorectal resection, the presence of extensive extranodal
deposits, or extracapsular nodal spread.

8.1.9  A blanket
approach to short-​course
preoperative radiotherapy
In the UK, there is a huge variation in the use of radiotherapy because some colo-
rectal MDTs have adopted the non-​selective use of routine SCPRT whilst others use
pelvic MRI to determine patients whose primary tumour is predicted to be clear of the
CRM and in whom initial surgery is performed. Studies have demonstrated that radio-
therapy has acute and long-​term detrimental effects on quality of life with significant
risks of permanent morbidity. About 5–​10% of patients will experience such grade 3 or
4 late morbidity. Effects on sexual functioning, urinary incontinence, bowel function,
and an increase in faecal incontinence and of insufficiency fractures in the pelvis have
all been documented after SCPRT
These complications depend on the size of the radiation field, shielding, the overall
treatment time, the fraction size, and total dose. Protocols for the management of late
toxicity including bowel, urinary, and sexual dysfunction are highly recommended
within each department.
IMRT offers better conformality than 3D and generally better organ-​at-​risk sparing,
and hence might be expected to show lower rates of late effects. Studies in CRT suggest
IMRT is associated with a clinically significant reduction in acute lower gastrointes-
tinal toxicity because of small bowel sparing compared with three-​field conventional
radiotherapy, and is generally accepted as a safe and effective radiotherapy method
although some have questioned the clinical relevance of IMRT bowel sparing for most
patients with rectal cancer So IMRT should not be considered mandatory for treat-
ment of rectal cancer.
The most recent studies suggest that 17–​20 patients need to undergo adjuvant radio-
therapy to prevent one local recurrence using a blanket approach. For example, the
Dutch trial demonstrated an absolute reduction in local recurrence of 6%. Thus, if
100 are irradiated, then six local recurrences are prevented, so the number of patients
treated to prevent one local recurrence is 16.7. Current evidence does not therefore
176 Rectal cancer

support the widespread advocacy for routine adjuvant radiotherapy as used in the
treatment arms of recent trials. Most would accept a local recurrence of <5% with
good quality TME without preoperative radiotherapy.
There is increasing evidence that patients with very low tumours that require
abdominoperineal excision are at higher risk of involvement of the CRM, local re-
currence, and inferior survival. Many concur that routine preoperative radiotherapy
is indicated for this group of patients. It is a source of considerable debate whether
SCPRT may be used in some, if CRT is preferred in others, and if a wider ‘cylin-
drical’ or extra-​levator surgical technique should be adopted to reduce the rate of an
involved CRM.

8.2  Radiation treatment planning

To target the tumour accurately, and get the best functional outcome out of the com-
bination of surgery and radiotherapy, good collaboration between surgeons, radiolo-
gists, and radiation oncologists is essential. This approach is facilitated by the MDT
meeting, which discusses every patient both in the preoperative and postoperative
setting. Useful information includes knowledge of the extent and position of the tu-
mour, the likely operation that will be performed, and any relevant previous medical,
surgical, and oncological treatment history. MRI and or PET/​CT or PET MRI can now
also be co-​registered on to the planning CT.

8.2.1  Computed tomography planning techniques

Patients are given instructions to maintain a full bladder as far as possible prior to each
simulation,scanning,treatment session in order to push small bowel cranially out of
the field as far as possible, and are reminded throughout the treatment period. A CT
planning scan with 3-​mm thick slices should be performed with the patient prone in
the treatment position. Intravenous contrast is ideally required, as the vessels cannot
always be easily identified from sequential slices. A radio-​opaque marker is useful to
demonstrate the position of the anal verge.
Patients are scanned from the superior aspect of L5, to 2 cm beyond the anal marker
in order to cover the whole of the pelvis, recto sigmoid, and rectum. The anterior bor-
ders can be defined in female patients with the use of a radio-​opaque tampon placed
in the vagina at CT planning or at simulation. Dilute small bowel contrast may be
used at this stage (gastrograffin 20 mL in 1 L of water given 45–​60 minutes prior to the
planning scan).
The radiation fields used in current practice are based on a number of factors
including the fields used in previous and current Phase III trials, the limited informa-
tion available on the pattern of failure, and differences in the philosophy of field size
according to whether certain lymph node regions should or should not be included.
A number of series prior to the TME era mapped recurrence, including the sem-
inal work from Gunderson at the Mayo clinic, amassed from ‘second-​look’ surgery (33)
and then made recommendations for clinical target volume (CTV) contouring based
on observed sites of local recurrence. Early studies defined the superior border as the
junction of L5/​S1, but some studies treated up to the origin of the inferior mesenteric
 Radiation treatment planning 177

artery at L4 and were associated with significant morbidity. Most current radiation
fields are still based on patterns of failure.
Rectal surgical quality has markedly improved with TME, and more recent studies
examining the site of local recurrence show these are often either at the site of the anas-
tomosis, the posterior pelvis, and lateral pelvic side-​wall, or low down representing
inadequate surgery. The risk of involvement of regional lymph node groups is also
recognized to depend on whether the primary tumour lies in the upper middle-​or
lower-​third of the rectum. All these sites have a slightly different natural history with
different areas of known lymphatic drainage.
With this knowledge several groups redefined planning volumes for rectal cancer
(34,35)
. The Roels guidelines identified five predominant areas of risk for local recurrence
and potential lymph node involvement, but relied on retrospective series predating
TME. In contrast, the RTOG guidelines are a one-​size-​fits-​all consensus of experts,
describing their practice.
Further recent modifications in CTV are derived from a study where the site of
recurrence has been analysed in patients within the Dutch TME study (36)83. No recur-
rences were found cranial to the S2–​S3junction in patients without nodal involvement.
In the absence of nodal involvement and with a negative CRM, only one recurrence
was found cranial to the S2–​S3 junction allowing a major reduction in exposure of
small bowel. Hence, an update of the original Roels guidelines recommend lowering
the superior/​cranial aspect of the CTV(37) based on this published evidence of the sites
of local recurrences. Recent additional consensus guidelines have been developed
which extend tumour and target volumes with major modifications for the lateral
lymph nodes and the ischio-​rectal fossa delineation(38)
Diagnostic MRI is usually performed with the patient supine. Studies are required
to evaluate the role of MRI/​CT co-​registration for the planning process and whether
prone pelvic MRI is required. Clinical studies have looked at MRI CT fusion for delin-
eation of the CTV. Others have examined whether greater accuracy can be achieved
using MRI and FDG-​PET-​CT to give additional information to standard pretreatment
evaluation and whether it would change the shape and the size of the gross tumour
volume (GTV) delineation.
With a PET-​CT simulator fused PET images can be obtained with the treatment
planning CT scan, and the integration of PET scanning in combination with conven-
tional morphological imaging is under investigation. Evidence suggests there may be
more accurate and reliable definition of the tumour volume leading to changes in the
GTV compared to CT and more consistency in planning. FDG-​PET-​CT co-​registered
with a planning CT can be useful, but it remains unclear whether this use of different
volumes impacts on clinical outcome, unless a boost to higher than conventional
doses is planned.
Prior to initiating a Phase III multicentre CRT trial in the UK for rectal cancer
(ARISTOTLE), we reviewed the literature on the definition of target volumes in
rectal cancer, the evidence of the site of lymph node metastases, and potential areas
of subclinical disease, the site of recurrence after TME for more locally advanced dis-
ease and late morbidity. With the collaboration of the trial management group, we
have established recommendations for target volume definition in a simple practical
178 Rectal cancer

guide for 3D planning within the trial (http://​www.ukctg.nihr.ac.uk/​trialdetails/​

ISRCTN09351447).
A summary of this approach is provided as follows.

Gross tumour volume
GTV is defined as all sites of gross tumour seen on the planning CT scan with the help
of the information from diagnostic CT and MRI scans clinical examination, endos-
copy, barium enema, and PET scan.
This concept is more difficult than in some other disease sites. The discontinuous
nature of many rectal cancers with extranodal deposits may require the demar-
cation of more than one GTV area. Also, if there is a large lymph node separate
to the primary tumour then both should be outlined separately. However, loops
of unopacified small bowel and perirectal soft tissue densities may be easily con-
fused with lymph node structures and correlation with the diagnostic pelvic MRI
is important.
It is useful to document the sites of all areas of GTV at the time of MDT discussion.
This includes extra nodal deposits, involved lymph nodes with irregular borders and
mixed signal characteristics, and extra mural vascular invasion.
It is recommended that the normal rectal wall is also included if the GTV is not
circumferential (see Fig. 8.2a, left-​ hand panel). Clearly, following surgery when
postoperative radiotherapy is given, it is not possible to define a GTV but preoperative
imaging can indicate the site of the primary tumour prior to surgery.

Clinical target volume
The CTV will encompass areas of microscopic spread beyond the defined GTV. The
CTV is defined in three separate steps:
1. CTVA:  a 1-​cm margin is applied in all directions to all sites of GTV (Fig.  8.2a,
right-​hand panel).
2. CTVB:  includes sites of potential microscopic disease including the mesorectal,
presacral, and internal iliac node regions. The limits of this volume are defined as
follows:
Superior limit: is the S2/​S3 junction provided that there is a 2-​cm margin above
•
the most superior aspect of the GTV (Fig. 8.2b, left-​hand panel, black line). If
this is not the case, the superior border is defined as 2 cm superior to the most
superior aspect of GTV (Fig. 8.2b, right-​hand panel, white line).
Inferior limit: is at the level of puborectalis which corresponds with the level that
•
the mesorectal fat is no longer seen on the axial planning CT scans (Fig. 8.2c,
left-​hand panel) providing that there is at least a 2-​cm margin below the most in-
ferior aspect of the GTV. If this is not the case (i.e. a very low tumour) the CTV
is placed 2 cm below the inferior aspect of the GTV (Fig. 8.2c, grey line).
Posterior limit: is the anterior surface of the sacrum and the coccyx. In the pres-
•
ence of the symptom of nerve infiltration but in the absence of macroscopic
tumour, the CTVB may be placed 0.5 cm posterior to the anterior border of the
sacrum.
 Radiation treatment planning 179

(a)

(b)

(c)
Mesorectum Mesorectum

Puborectalis Puborectalis

External sphincter External sphincter

Internal sphincter Internal sphincter

Fig. 8.2  (a) Left hand panel: GTV. Right hand panel: CTVA and GTV. (b) Superior border
CTVB. Left hand panel: S2/​3 junction. Right hand panel: 2 cm superior to the most
superior limit of GTV. (c) Inferior border CTVB at level of puborectalis (left-​hand panel)
and 2 cm below most inferior limit of HGTV (right-​hand panel). (d) Anterior CTVB border
upper pelvic. (e) Anterior CTVB border mid pelvis—​this patient has a small mesorectum
and the border is determined by the lateral pelvic nodal compartment.
180 Rectal cancer

(d)

External iliac artery

(e)

Fig 8.2 Continued

Anterior limit: is determined at different levels of the pelvis:

•
◆ Upper pelvis—​a 7-​mm margin is applied to the internal iliac arteries. The
most anterior of the arteries is used to determine the border (Fig. 8.2d).
◆ Mid pelvis—​is 1 cm anterior to the anterior mesorectal fascia or the anterior
limit of the lateral (iliac) pelvic lymph node compartment, whichever is the
more anterior. In the example shown, the patient has a very small mesorectum
and the border is determined by the lateral nodal compartment.
◆ Lower pelvis—​is the outer aspect of the sphincter complex.
Lateral limit: is also determined at three levels:
•
◆ Upper pelvis—​7 mm lateral to the internal iliac arteries.
◆ Mid pelvis—​the medial aspect of obturator internus in the absence of nodal
enlargement outside the mesorectum. In the presence of involved lateral
pelvic nodes the border is the lateral border of obturator internus on the side
of the involved nodes.
 Radiation treatment planning 181

◆ Lower pelvis—​the outer aspect of the sphincter complex unless there is in-
volvement of the sphincter complex when a 1-​cm margin later to the sphincter
complex is used.
The ARISTOTLE trial management group did not find any evidence to justify the in-
clusion of the entire ischio-​rectal fossa or the external iliac nodes in the CTVB.
3. Final clinical target volume (CTVF)—​defined by the union of CTVA and CTVB.

Target delineation following local excision

If post-​operative CRT is recommended after local excision, the whole mesorectum will
still be in situ. For cT1 cancers only 1% were pN2 and only 3/​198 patients had lymph
node metastases beyond the pararectal mesorectum. In contrast, for cT2 tumours 58/​
194 (30%) had N1 and 14 (7%) had N2 nodal disease, but only 8/​194 (4%) had lymph
node metastases beyond the pararectal mesorectum. Precise mapping of sites of lymph
nodes within the mesorectum in more advanced stages are sparse and in few. We treat
the mesorectum 5  cm in the cephalad direction superior to the excision site of the
primary tumour, and 3 cm distal following local excision. It seems unlikely that lat-
eral pelvic side wall nodes will be involved in early cancers. Rates of involvement are
reported to be extremely low <1% in PT1 patients up to approximately 29% in PT4
patients (overall 5%)(39). Hence, wider and larger pelvic fields are unnecessary.

Planning target volume
The PTV is defined by applying a 1-​cm direction in all directions to the CTVF.
This volume ensures coverage of the CTV taking into account the systematic and
random set-​up errors and internal movement that may occur when delivering a rad-
ical course of radiation, including variations in tissue position, size, shape, and also
variations in patient position(40). Few departments have data to determine their sys-
tematic and random set up errors.

8.2.2  Organs at risk

3D treatment planning allows visualization on non-​axial planes with a beam configur-
ation, and will allow preparation of a dose–​volume histogram (DVH) and hence derive
optimal and maximal doses for both tumour and normal tissue. The organs at risk in
radiotherapy for rectal cancer include small bowel, femoral heads, genitalia, and bladder.
Normal critical tissues such as small bowel, femoral heads, ureter, and bladder can
be contoured and doses to these organs kept to a minimum. The small bowel is often
close to the target volume, and the dose is specified such that not more than 250 mL
receives in excess of 45 Gy. Significant small bowel sparing is offered in the future by
the increasing use of IMRT in rectal cancer (either focusing on small bowel volume in
terms of individual small bowel loops or the bowel bag). However, doses to significant
volumes of small bowel in excess of 50.4 Gy are not recommended. There is no con-
sensus on external genitalia delineation for males or females.
182 Rectal cancer

8.2.3  Normal tissue tolerances

The advent of 3D computerized planning has led to the concept of tolerance of an
organ being replaced by DVHs, which focus on the dose delivered to the percentage
of any specific organ. Most normal tissue tolerance algorithms assume a daily dose of
1.8 Gy per fraction. It is also wise to maintain the dose per fraction to small bowel,
ureter, and bladder to 2 Gy or less, i.e. avoiding gross inhomogeneities of dose. When
adjuvant radiation is given the tolerance doses of most organs is not exceeded with the
exception of the small bowel (see section 8.2.4). Although DVHs may be produced
for organs/​structures at risk, it is currently unclear whether the planned treatment ap-
proach should be altered if predefined dose constraints cannot be met, except possibly
for the small bowel. It is not possible to generate a true DVH, because currently, the
assessment of dose to small bowel is usually based on the volume (cm3) of small bowel
within the high-​dose volume in the pelvis rather than the entire small bowel volume
within the whole abdominal cavity (which is difficult to measure).

8.2.4 Small bowel
Small bowel, bladder, and femoral heads have usually been considered the main dose-​
limiting organs when high doses of radiotherapy are delivered to the pelvis. Late
effects in small bowel range from malabsorption to strictures, obstruction, and ulcer-
ation. The volume of the small bowel within the radiation field is crucial. Late small
bowel complications are estimated to occur with an actuarial 5-​year risk of 5–​40%
incidence following pelvic postoperative radiotherapy to a dose of 45–​50 Gy in rectal
cancer. IMRT may substantially reduce the dose of radiation that the small bowel
within the pelvis may receive but at the expense of a larger volume of small bowel re-
ceiving a more modest radiation dose. However, IMRT is technically challenging and
not proven to be cost-​effective because it has not been shown to improve oncological
outcomes particularly OS, and is not considered mandatory in rectal cancer.

8.2.5 Patient position
The prone position is thought to displace small bowel out of the treatment volume
mainly by anterior displacement. It also allows the accurate placement of a radio-​
opaque marker on the anal verge and the use of a belly board. It is, however, uncom-
fortable, particularly when the patient has a stoma, and is often not feasible because of
impaired mobility and comorbidities. High body mass index, psychiatric disturbances,
and having an ileostomy or colostomy are associated with an increased risk of set-​up
instability in the prone position. Interfractional set-​up errors in pelvic tumours can be
minimized by using immobilization devices, and employing regular imaging devices
(pre-​treatment cone beam CT).
Whilst the supine position is considered more reproducible, most clinicians are con-
cerned that this position may increase the small bowel in the treated volume. The
validity of this is unclear with studies that show no significant difference in the ir-
radiated small bowel volume between the two treatment positions. More recent data
suggests although the supine position is more stable and more reproducible, yet prone
 Radiation treatment planning 183

position may still offer significant dose reductions for the small bowel compared with
the supine position(41).

8.2.6  Small bowel exclusion techniques

Most units employ bowel displacement techniques such as a Styrofoam mould or
the belly board (the use of a table cut-​out which allows the upper abdominal small
bowel contents to fall forwards and upwards, and further reduce the volume of
small bowel irradiated(42,43). There is no consensus as to the optimal standardized
position of the belly board opening. However, if the opening is placed near the
lumbosacral junction, the reduction in volume of small bowel irradiated appears
to be maximal(43).

8.2.7  Organmotion set-​up variation and treatment

margins in radical radiotherapy
Studies in organ motion and set-​up uncertainties have usually focused on prostate ir-
radiation because radiotherapy doses are very high. In contrast, radiotherapy in rectal
cancer delivers doses which are relatively modest (45–​54 Gy) and can be boosted with
endoluminal brachytherapy, if required. The rectum does move during a course of
radiotherapy and there are substantial variations in both the rectal volume and rectal
wall displacement during a course of radiation. However, there is little data to quan-
tify rectal motion and set-​up variation. Several studies have reported on rectal motion
based on cone-​beam computed tomography (CBCT)(44,45,46), and conclude that 10–​
15 mm is sufficient as a margin to the rectum during simultaneous integrated boost
(SIB)-​IMRT in the supine position. Anatomical considerations suggest the rectum is
more fixed at the distal end than proximally.

8.2.8  Postoperative radiotherapy clinical target volume

Postoperative radiotherapy is currently delivered for those patients who did not re-
ceive preoperative radiotherapy and proceeded to surgery, but on histology review
are seen to have an involved circumferential margin (≤ 1 mm), large defects in the
mesorectum, evidence of perforation, or widespread extranodal deposits. We do not
recommend the routine use of postoperative chemoradiation for pT3 N0 or N1 pro-
vided there is a good quality mesorectal excision. The indication for postoperative
chemoradiation in patients with pT4 and/​or pN2 histology remains a matter of
debate.
After TME surgery there is no defined GTV. The CTV includes a surrounding
safety margin beyond the mesorectum and the nodal groups at risk of presumed sub-​
clinical involvement. In patients who have had abdominoperineal resection, we do
not include the perineal scar. However, there is no consistent approach or agreement
as to the CTV. The application of ligaclips at the time of surgery to areas at high
risk of local recurrence is useful for defining a CTV in this situation. Displacement
of the omentum and attachment within the pelvis may minimize the late effects of
radiotherapy.
184 Rectal cancer

8.2.9 Total dose
Conventionally, when 1.8 Gy per fraction is used, total doses in the range 45–​50.4
Gy have been delivered in the preoperative setting, and 50.4 Gy with the option of
a 5.4-​Gy boost to the tumour bed in the postoperative setting. It is assumed that the
treatment will be delivered 5 days per week, one fraction per day, 1.8 Gy per fraction.

8.2.10 Preoperative dose
Of the SCPRT studies that delivered a biologically equivalent dose > 30 Gy, the Swedish
Rectal Cancer Study, the Dutch study, and the CR07 study all use 25 Gy in five frac-
tions. Most clinicians accept this dose/​fractionation as a standard.
With the use of 1.8–​2.0 Gy per fraction the total dose in phase III trials has ranged
from 45–​54 Gy in most reported series with a sequential boost administered to the
GTV. This practice is facilitated by IMRT and a SIB, but the optimal dose remains
undefined. In contrast to a SIB, a sequential boost following primary treatment al-
lows for adaptive treatment strategies, e.g. with re-​planning and optimization for
the boost treatment, to the smallest possible boost volume, but tumour may not be
visible on CBCT late in the treatment course. Treatment verification is easier if the
sequential boost is given prior to the primary treatment, but treats a larger boost
volume.
A sequential Phase II from Canada has been reported. Three sequential schedules
combined radiation with infusional 5FU, escalated from 40 Gy in 20 fractions to 46 Gy
in 23 fractions and finally to 50 Gy in 25 fractions. A statistically significant difference
in terms of local control was observed for doses of 46 Gy and above, but no difference
between 46 Gy and 50 Gy(47). The same study also appeared to show a trend to higher
pathological complete response rates with increasing radiation dose of 13%, 21%, and
31% for 40 Gy, 46 Gy, and 50 Gy respectively.
Currently, there are proponents that a much higher dose (in the region of 60 Gy)
can be safely delivered with the high conformality of IMRT or brachytherapy in order
to increase the pCR and achieve more R0 resections. Higher doses of radiotherapy re-
sults in higher pCR rates. Studies by the Danish group demonstrated a dose-​response
relationship for pathological tumour regression(48) and provided a quantitative esti-
mate of this relationship(49). A  meta-​analysis of studies of high-​dose pre-​operative
radiotherapy concluded that radiation doses ≥60Gy result in higher rates of pCR(50).
However, long-​term oncological outcomes have not improved with such dose escal-
ation(51). In addition, a study of dose escalation and simultaneous integrated external
beam boosts cancer using 62.5 Gy in 25 fractions in patients with non-​resectable lo-
cally advanced rectal cancer, suggested such treatment in the largest tumours had an
unacceptable risk of intestinal toxicity(52).

8.2.11  Standard dose prescriptions

◆ Short course preoperatively: 25 Gy in five daily 5-​Gy fractions
◆ Preoperative CRT: 45–​50.4 Gy in 25–​28 daily 1.8-​Gy fractions
◆ Postoperative CRT: 50.4–​54 Gy in 28–​30 daily 1.8-​Gy fractions.
 Radiation treatment planning 185

It is conventional to report the dose to the ICRU reference point, the maximum dose
to the PTV, and the minimum dose to the PTV. The isocentric treatment plan is usu-
ally specified to receive 100% with the 95% isodose line encompassing the PTV. The
minimum dose is defined as 99% of the PTV receiving ≥ 95% of the prescribed dose.
The maximum dose is defined as < 5% and < 2% of the PTV receiving 105% and 110%
of the prescribed dose, respectively.
Boosting to higher doses aims to increase response rates and allow more patients
to undergo minimally invasive surgery or even omit surgery in an organ-​preservation
strategy. There is a particular rationale to dose-​escalate above these doses in resect-
able cancers where the patient is either frail or on the grounds of extreme age or other
co-​morbidity unsuitable for radical surgery. The high risk of radical surgery in some
groups—​not only in terms of 30-​day mortality but also in terms of 6-​month mor-
tality, may recommend an alternative to radical surgery. Some also will refuse radical
surgery. In these circumstances it may possible to avoid radical surgery with the use
of either radiotherapy or CRT and a brachytherapy boost in a proportion of patients.

Organ sparing approaches
Increasingly in the decision-​making process, patients express a strong preference to avoid a
stoma and request alternatives. Patients with distal rectal cancers treated with neoadjuvant
chemoradiation are clearly a relevant group because preoperative CRT results in a pCR in
10–​20% of patients and a complete clinical response (CCR) in 20–​40% (depending on ini-
tial clinical stage). CCR has been defined as the absence of any palpable tumour or irregu-
larity with no visible lesion except a flat scar, telangiectasia, or whitening of the mucosa.
If a CCR is observed after CRT, radical surgery is deferred in a watch and wait pro-
gramme. Hence, organ-​sparing is being adopted in selected patients. Angelita Habr-​
Gama in Brazil has consistently demonstrated the feasibility and relative safety of a
non-​operative approach if meticulous and rigorous surveillance is undertaken(53,54,55).
Hence CRT is being offered up front as an alternative to radical surgery either with a
local excision or there may be a watch and wait strategy if there is a CCR. Subsequent
regrowth of the primary is reported in about 30% of patients, but is normally
endoluminal, and can generally be salvaged pCR is more likely in earlier cT stage.
Some studies suggest small tumours <4 cm in diameter extending less than half the
rectal circumference with a normal CEA level and cN0(56) are more likely to achieve a
CCR. In contrast, few if any clinically staged T4 cancers achieve pCR.
However, how best to select such patients for a non-​operative approach either by
imaging or clinical factors remains elusive. The optimal concurrent chemotherapy,
radiotherapy field sizes and doses remain to be confirmed but the best results are likely
to be achieved with the precise and detailed protocolized strategy of Habr-​Gama. The
strategy remains unproven but the European Registry of Cancer Care (EURECCA)
‘International Watch & Wait Database’ www.iwwd.org may in future provide robust
information from a large number of patients on its safety.

8.2.12  Image-​guided radiotherapy
Systematic errors in terms of set-​up uncertainties result from the fact that the imaging
performed for treatment planning is typically just a snapshot, and the target position
186 Rectal cancer

determined at that moment may differ from the average target position at any subse-
quent treatment time. The random error is the day-​to-​day deviation from the average
target position (which depends on internal organ motion and the repeated treatment
set-​up over 25 treatments). The concept of IGRT achieves tumour and soft tissue
imaging in real-​time to allow correction for both systematic and random errors on a
daily basis. IGRT should utilize ideally 4D assessment of the target volume, efficient
comparisons of images with reference data and fast automated corrections.

8.2.13 Portal imaging
The initial CT planning scan may not continue to represent the internal anatomy for
every treatment fraction, because rectum and bladder may change in size and position
throughout treatment. Portal imaging for verification of isocentre position and treat-
ment fields should be acquired on the first few treatment sessions both for AP and lat-
eral images and compared to simulator films. Electronic portal imaging (EPI) devices
are used to match for bony anatomy and can monitor set-​up displacement on a daily
basis in the initial phase of treatment to detect dose delivery discrepancies. Additional
information on internal motion can be obtained by CBCT or tomotherapy, which are
performed on a weekly basis.
Fields should be moved if they fall outside an agreed tolerance level—​usually 5–​8 mm
for patients who are treated prone. This process also allows radiographers to evaluate
the whole set-​up, and thus to assess and correct systematic errors. Reverification is
recommended on a weekly basis. These EPI images should be audited at the clinician’s
weekly meeting. The multi-​leaf collimator configuration can also be verified for con-
sistency and reproducibility.

8.3  Supportive care during radiotherapy

Expected acute side effects include diarrhoea, proctitis, urinary frequency and dysuria,
erythema, and moist desquamation of the perineum in low rectal cancers. Patients
should be assessed at least once weekly with regard to toxicity and their overall tol-
erance to treatment. Toxicity should be scored according to the Common Toxicity
Criteria (CTC) v 3.0.
It is important to maximize patient tolerance with the use of simple antiemetics,
anti-​diarrhoea agents, analgesia, and nutritional support. For severe haematological
toxicity, treatment is usually omitted and resumed when the blood counts recover to
grade 2 or better. For acute gastrointestinal toxicity, treatment is resumed when the ef-
fects decrease to grade 2 or better and supportive care with intravenous fluids and oral
anti-​diarrhoea agents may be administered.

8.4  Intensity-​modulated radiotherapy
Technical advances such as IMRT and more recently VMAT allow greater preci-
sion and sparing normal surrounding structures such as small bowel compared to
 High dose rate brachytherapy 187

conventional 2D or 3D planning. IMRT can conform radiation dose to the target

structure by partitioning the radiation field into smaller beamlets of different dose
intensities. During IMRT treatment planning, standard radiation dose constraints can
be used to spare critical normal tissues. Hence, IMRT may allow improved compli-
ance, or facilitate dose-​escalation, without increasing late morbidity. The benefit of
reducing the irradiated small bowel volume is greatest when a very large clinical target
is treated. For example, if the CTV extends superiorly to the sacral promontory or the
external iliac nodes are electively irradiated.
In the EORTC 22921 study trial, 522 patients retained sphincters and of these, only
1.4% required surgery for small bowel complications. This low level of late morbidity
calls into question the drive to deliver IMRT to unselected patients with the aim of
reducing small bowel toxicity, unless dose escalation becomes more commonly em-
ployed. Although IMRT is not invariably clinically indicated, it should be available for
complex planning and delivered according to clearly defined protocols.

8.5 Endocavitary radiotherapy
Appropriately selected patients with early rectal cancer can be controlled locally,
and experience long-​term survival with the use of endocavitary or contact radiation
therapy. Treatment guidelines and indications for this technique have been developed
originally by Papillon. This radical technique as a sole modality is appropriate if the tu-
mour is confined to the rectal wall (T1 N0). Very small volumes of tissue are irradiated
in this situation. Doses of 100 Gy can be delivered to such a small volume without risk
of unacceptable late complications. Some clinicians endorse proactive treatment with
combined external beam radiotherapy (EBRT) or CRT and contact therapy for both
cT1No and also cT2 or ‘early’ cT3 N0(57).
A 3-​cm diameter applicator can be introduced into the rectum using local anaes-
thesia, and this can be performed on an outpatient basis. Standard treatment regimen
is four applications of 30 Gy (applied dose) with an interval of at least 2 weeks be-
tween each application. The applicator is a short focal contact X-​ray unit using 50-​
kV rays with a dose rate of approximately 10 Gy per minute. Lack of response after
two applications of radiotherapy is usually considered a sign of radio-​resistance and
most will then proceed to surgery. High rates of local control are possible with this
approach(58,59).

8.6  High dose rate brachytherapy

High dose rate intraluminal brachytherapy (HDR-​BRT) offers an alternative to EBRT
or can be administered as a boost to the primary tumour following CRT prior to sur-
gery or to dose-​escalate after chemoradiation for curative treatment as an alternative
to surgery(60). HDR-​BRT produces a rapid fall-​off of radiation dose at depth, allowing
delivery of high doses to mural tumour while sparing normal structures such as skin,
sphincter muscles, bladder, small bowel, and even the contralateral rectal mucosa.
The volume irradiated appears mainly limited to an approximate 2-​cm radius from
188 Rectal cancer

the primary tumour, and hence provides limited dose to more distant mesorectal
lymph nodes.
A Canadian phase I/​II study evaluated preoperative HDR-​BRT in patients with T2-​3
Nx tumours. A total dose of 26 Gy in four daily fractions of 6.5 Gy was prescribed to
the GTV and any intramesorectal deposits on MRI. TME surgery was performed 6–​8
week later. The pCR rate was 27% and positive nodes were found in only 31%. With a
median follow-​up 63 months, the actuarial local recurrence rate was only 4.8%, DFS
65.5%, and OS 72.8%(61).
HDR-​ILBT is also employed as a boost to GTV following EBRT and may be asso-
ciated with higher rates of ypT0 when CRT up to 10/​17 (58%), but substantial acute
toxicity and proctitis may limit its use. A  matched analysis compared the results
of preoperative EBRT and HDR-​BRT in terms of local recurrences, cancer-​specific
deaths and OS between Dutch and Canadian centres, and found no significant
differences(62).
HDR-​ILBT has been used at Mount Vernon Cancer Centre(63) as a boost along-
side external chemoradiation, as a sole treatment for small localized tumours, and
for short-​term palliation for advanced symptomatic tumours particularly in the very
elderly and frail. It is possible to place clips at the superior inferior extent of the tu-
mour although these can only be relied upon to be maintained for 10–​12 days. When
the tumour is not circumferential, it is possible to use segmental shielding with an
applicator which shields 25% or 50% of the rectal circumference. Treatment prescrip-
tion is defined at 1 cm from the source access. HDR-​ILBT for advanced or inoperable
tumours of the rectum has been used both in the palliative setting and to dose escalate
after chemoradiation for curative treatment(64).

8.7 Intraoperative radiotherapy
Intraoperative radiotherapy (IORT) allows additional irradiation of the tumour bed
(i.e. higher doses) without compromising the surrounding organs at risk. IORT is usu-
ally delivered under anaesthetic using electrons, and dose range of 10–​20 Gy. However,
no randomized trials have evaluated the benefit of IORT in addition to EBRT. Recent
pooled results of multimodality treatment of locally advanced rectal cancer in four
major treatment centres with particular expertise in IORT suggest an advantage to
IORT(65). The limitations of these analyses are the inclusion of patients treated during
a relatively long period, the heterogeneity of external beam treatments, and variations
in adjuvant chemotherapy.

8.7.1  Locally recurrent rectal cancer

Isolated local recurrence after preoperative radiotherapy and TME surgery is infre-
quent and rarely isolated, and has been almost invariably associated with fatal out-
comes. More recent series have documented 50% R0 resections and 40% 5-​year
survival. In contrast, surgery for central/​anastomotic local recurrences resulted in
the most favourable outcomes, with 77% R0 resections and 60% 5-​year survival(66).
Currently in the UK, the widespread use of preoperative radiotherapy, either short
 Treatment of the primary in the presence of metastatic disease 189

course or long course with chemoradiation, has ensured that isolated local pelvic re-
currence is a rarity.
When local recurrence occurs and radiotherapy has not previously been adminis-
tered, radiotherapy or chemoradiation can produce good palliation of symptoms, but
long-​term local control is seldom achieved. The duration of effective palliation is usu-
ally short with further progression of symptoms within 3–​6 months after irradiation,
and complete responses are rarely achieved, even with high radical doses in the region
of 60 Gy.

8.7.2  Re-​irradiation
Further re-​irradiation in the case of local recurrence after previous CRT remains
a controversial issue. Some single-​centre experience suggests this practice may be
safe in the short term, although long-​term evidence is sparse. Others have recom-
mended hyperfractionated accelerated radiotherapy with doses of 39 Gy (1.5 Gy
twice daily) following previous pelvic radiotherapy with a retreatment interval of
> 1 year(67).
If radiotherapy has not already been given, patients should be considered for
standard dose preoperative CRT (45–​50 Gy in 5–​6 weeks) [III, A] prior to an attempt
at resection. Alternatively SCPRT followed by a fluoropyrimidine and oxaliplatin
based chemotherapy as used in Polish-​2 study can be also given(68).
In patients previously irradiated, re-​irradiation to lower doses (with concomitant
chemotherapy) is safe and can be used in selected patients to facilitate a curative
resection or to palliate symptoms. Brachytherapy can also be an effective palliative
option.
To consider re-​irradiation, it is essential to know when and where the pri-
mary CRT treatments were undertaken, along with details of total dose, fraction
size, radiation fields primary and elective, the overall treatment time, and how
radiotherapy was tolerated. Contemporary evidence of late radiation sequelae
(objective—​ telangiectasia, fibrosis—​ and subjective—​ small bowel symptoms)
should be reviewed to enable decisions regarding the feasibility and planning of
further CRT regimens(69).
IORT has also been advocated in these circumstances.

8.8  Treatment of the primary in the presence

of metastatic disease
The lack of randomized studies in this setting, means strategy will depend on the
resectability of both primary and metastatic disease and patient-​related factors (func-
tion, comorbidity, socioeconomic factors, and patient views). Local palliation of rectal
symptoms with radiotherapy may be required. SCPRT is more flexible than CRT since
systemic chemotherapy can start within 2 weeks from the start of treatment, and pal-
liates symptoms in about 80% of patients thereby avoiding a stoma. If metastatic dis-
ease is limited and cure is a possibility, local control with systemic chemotherapy and
appropriate recourse to metastasectomy may be effective. SCPRT has been combined
190 Rectal cancer

with triplet-​chemotherapy (capecitabine, oxaliplatin, and bevacizumab) to facilitate

the resection of borderline resectable liver metastasis and the primary tumour(70) and
can lead to long-​term survival(71).

8.9 The future
We recommend both a more selective approach to use of radiotherapy and exam-
ining dose escalation of radiotherapy to the primary tumour where MRI predicts a
threatened CRM—​either with external beam or brachytherapy. Potential strategies of
neo-​adjuvant, concurrent, consolidation (after chemoradiation and before surgery),
and postoperative adjuvant chemotherapy with cytotoxic agents, are promising, but
consolidation chemotherapy following chemoradiation in locally advanced disease
appears the most attractive way forward. Stereotactic body radiotherapy, whereby a
small number of high-​dose-​per-​fraction treatments are given to a small target volume
also warrants further attention particularly in the setting of local rectal recurrence and
oligometastatic disease.
Traditional preoperative doses and fractionation in rectal cancer, the precise target
definition, and finally the timing of surgery have aimed to maximize tumour cell
kill, minimize normal tissue effects, and surgical morbidity in order to prevent local
recurrence.
The immune environment and the influence of radiotherapy is currently a source of
discussion. The Immunoscore may serve as a both as a prognostic marker in patients
with rectal cancer treated by primary surgery and predict response to preoperative
CRT(72). If radiation is considered as an immunological intervention, current dogma
may be obsolete and the aims of treatment may well require a rethink. Novel para-
digms may be required. For instance, the whole tumour with all possible microscopic
extension in an elective pelvic volume may not require high-​dose radiotherapy, be-
cause simply irradiating the macroscopic primary tumour may prove sufficient to
trigger the immune system. Radiation in appropriate doses and fractionation may be
able to convert some completely or partially non-​immunogenic tumours into highly
immunogenic tumours or conversely result in a partially immunogenic tumour be-
coming completely non-​immunogenic.

8.10 Conclusion
The importance of achieving a R0 (CRM ≥ 1 mm) endpoint, with the emphasis on
the circumferential margin, still remains paramount. Bowel cancer screening is redu-
cing the numbers of patients with locally advanced cancers. In the era of personalized
medicine the need for radiotherapy should be discussed in the MDT and based on
the individual clinical and imaging features and the fitness of the patient rather than
simply on stage. Risk assessment tools such as POLARS may aid decision making by
helping patients understand the likelihood of bowel dysfunction with the use of radio-
therapy and allow more informed treatment choices.
The future hope is of increasingly accurate methods of selecting patients—​
potentially by means of even more sophisticated MRI staging and molecular pre-
dictive markers—​who would benefit most and least from preoperative treatment. If
 References 191

we could more accurately predict the risks of local recurrence, we could spare those
for whom radiotherapy may be unnecessary. The tumour, node, metastasis stage
(obtained either clinically by MRI or histologically at surgery) is the only proven
prognostic marker to aid in the identification of patients with aggressive patterns
of disease. However, clinical assessment of nodal status remains inaccurate. Current
phase III trials therefore have an obligation to collect tissue and perform translational
studies.
The recommendations proposed here are limited by current knowledge and sub-
jective biases and should not be considered definitive. More precise treatment plan-
ning systems using IMRT or VMAT approaches may in future be required for the
treatment of rectal cancer—​particularly if we intend to dose escalate above 45 Gy
routinely.

References
1. Quirke P, Steele R, Monson J, et al. MRC CR07/​NCIC-​CTG CO16 Trial Investigators;
NCRI Colorectal Cancer Study Group. Effect of the plane of surgery achieved on
local recurrence in patients with operable rectal cancer: a prospective study using data
from the MRC CR07 and NCIC-​CTG CO16 randomised clinical trial. Lancet 2009;
373:821–​8.
2. Bondeven P, Hagemann-​Madsen RH, et al. Extent and completeness of mesorectal
excision evaluated by postoperative magnetic resonance imaging. British Journal of Surgery
2013; 100:1357–​67.
3. Anonymous. Improved survival with preoperative radiotherapy in resectable rectal cancer.
Swedish Rectal Cancer Trial. New England Journal of Medicine 1997; 336:980–​7.
4. Glynne-​Jones R, Wyrwicz L, Tiret E, et al. Rectal cancer: ESMO Clinical Practice
Guidelines. Annals of Oncology 2017; 28 (suppl 4):iv22–​iv40.
5. National Clinical Practice Guidelines in Oncology (NCCN Guidelines): Rectal Cancer
Version 2.2017 www.ncrn.org . accessed 4/​7/​2017.
6. Van van der Geest LG, Lam-​Boer J, Koopman M. et al. Nationwide trends in incidence,
treatment and survival of colorectal cancer patients with synchronous metastases. Clinical
& Experimental Metastasis 2015; 32:457–​65.
7. Hunter CJ, Garant A, Vuong T, et al. Adverse features on rectal MRI identify a high-​risk
group that may benefit from more intensive preoperative staging and treatment. Annals of
Surgery and Oncology 2012; 19(4):1199–​205.
8. Glimelius B, Tiret E, Cervantes A, et al. ESMO Guidelines Working Group. Rectal
cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-​up. Annals
of Oncology 2013; 24:Suppl 6:vi81–​88.
9. Shin R, Jeong SY, Yoo HY, et al. Depth of mesorectal extension has prognostic
significance in patients with T3 rectal cancer. Diseases of the Colon and Rectum 2012;
55(12):1220–​8.
10. Merkel S, Weber K, Schellerer V, et al. Prognostic subdivision of ypT3 rectal tumours
according to extension beyond the muscularis propria. British Journal of Surgery 2014;
101(5):566–​72.
11. Lino-​Silva LS, Loaeza-​Belmont R, Gómez Álvarez MA, et al. Mesorectal invasion depth
in rectal carcinoma is associated with low survival. Clinical Colorectal Cancer 2016;
pii: S1533-​0028(16)30074-​3.
192 Rectal cancer

12. Barbaro B, Leccisotti L, Vecchio FM, et al. The potential predictive value of MRI and
PET-​CT in mucinous and nonmucinous rectal cancer to identify patients at high risk of
metastatic disease. British Journal of Radiology 2017; 90(1069):20150836.
13. Al-​Sukhni E, Milot L, Fruitman M, et al. Diagnostic accuracy of MRI for assessment of
T category, lymph node metastases, and circumferential resection margin involvement
in patients with rectal cancer: a systematic review and meta-​analysis. Annals of Surgical
Oncology 2012; 19:2212–​23.
14. Li XT, Zhang XY, Sun YS, et al. Evaluating rectal tumor staging with magnetic resonance
imaging, computed tomography, and endoluminal ultrasound: A meta-​analysis. Medicine
(Baltimore) 2016; 95(44):e5333.
15. Langman G, Patel A, Bowley DM. Size and distribution of lymph nodes in rectal cancer
resection specimens. Diseases of the Colon and Rectum. 2015; 58(4):406–​14.
16. Bujko K, Nowacki MP, Nasierowska-​Guttmejer et al. Long-​term results of a randomised
trial comparing preoperative short-​course radiotherapy with preoperative conventionally
fractionated chemoradiation for rectal cancer. British Journal of Surgery 2006; 93:1215–​23
17. Ngan SY, Burmeister B, Fisher RJ, et al. Randomized trial of short-​course radiotherapy
versus long-​course chemoradiation comparing rates of local recurrence in patients with
T3 rectal cancer: Trans-​Tasman Radiation Oncology Group Trial 01.04. Journal of Clinical
Oncology 2013; 31(3):399.
18. Liu SX, Zhou ZR, Chen LX, et al. Short-​course versus long-​course preoperative
radiotherapy plus delayed surgery in the treatment of rectal cancer: a meta-​analysis. Asian
Pacific Journal of Cancer Prevention 2015; 16:5755–​62.
19. Braendengen M, Tveit KM, Berglund A, et al. Randomized phase III study comparing
preoperative radiotherapy with chemoradiotherapy in nonresectable rectal cancer. Journal
of Clinical Oncology 2008; 26(22):3687–​94.
20. Erlandsson J, Holm T, Pettersson D, et al. Optimal fractionation of preoperative
radiotherapy and timing to surgery for rectal cancer (Stockholm III): a multicentre,
randomised, non-​blinded, phase 3, non-​inferiority trial. Lancet Oncology 2017; 18(3):336–​46.
21. Kapiteijn E, Marijnen CA, Nagtegaal ID, et al. Dutch Colorectal Cancer Group.
Preoperative radiotherapy combined with total mesorectal excision for resectable rectal
cancer. New England Journal of Medicine 2001; 345:638–​46.
22. Peeters KC, Marijnen CA, Nagtegaal ID, et al. The TME trial after a median follow-​up of
6 years: increased local control but no survival benefit in irradiated patients with resectable
rectal carcinoma. Annals of Surgery 2007; 246(5):693–​701.
23. van Gijn W, Marijnen CA, Nagtegaal ID, et al. Dutch Colorectal Cancer Group.
Preoperative radiotherapy combined with total mesorectal excision for resectable rectal
cancer: 12-​year follow-​up of the multicentre, randomised controlled TME trial. Lancet
Oncology 2011; 12(6):575–​82.
24. Sebag-​Montefiore D, Stephens RJ, Steele R, et al. Preoperative radiotherapy versus
selective postoperative chemoradiotherapy in patients with rectal cancer (MRC CR07 and
NCIC-​CTG C016): a multicentre, randomised trial. Lancet 2009; 373(9666):811–​20.
25. Gastrointestinal Tumour Study Group–​GiTSG 7175. Prolongation of disease free interval
in surgical treated rectal carcinoma. New England Journal of Medicine 1985; 312: 1464–​72.
26. Krook JE, Moertel CG, Gunderson LL, et al. Effective surgical adjuvant therapy for high-​
risk rectal carcinoma. New England Journal of Medicine 1991; 324:709–​15.
27. Bosset JF, Calais G, Mineur L, et al. Enhanced tumorocidal effect of chemotherapy with
preoperative radiotherapy for rectal cancer: preliminary results-​-​EORTC 22921. Journal of
Clinical Oncology 2005; 23(24):5620–​7.
 References 193

28. Sauer R, Liersch T, Merkel S et al. Preoperative versus postoperative chemoradiotherapy

for locally advanced rectal cancer: results of the German CAO/​ARO/​AIO-​94 randomized
phase III trial after a median follow-​up of 11 years. Journal of Clinical Oncology 2012;
30(16):1926–​33.
29. Roh MS, Colangelo LH, O’Connell MJ, et al. Preoperative multimodality therapy improves
disease-​free survival in patients with carcinoma of the rectum: NSABP-​R03. Journal of
Clinical Oncology 2009; 27:5124–​30.
30. Rödel C, Graeven U, Fietkau R, et al. German Rectal Cancer Study Group. Oxaliplatin
added to fluorouracil-​based preoperative chemoradiotherapy and postoperative
chemotherapy of locally advanced rectal cancer (the German CAO/​ARO/​AIO-​04
study): final results of the multicentre, open-​label, randomised, phase 3 trial. Lancet
Oncology 2015 Aug;16(8):979–​89.
31. Mohiuddin M, Paulus R, Mitchell E, et al. Neoadjuvant chemoradiation for distal rectal
cancer: 5-​year updated results of a randomized phase 2 study of neoadjuvant combined
modality chemoradiation for distal rectal cancer. International Journal of Radiation
Oncology, Biology, Physics 2013; 86:523–​8.
32. Smith N, Brown G. Preoperative staging of rectal cancer. Acta Oncologica 2008;
47(1):20–​31.
33. Gunderson LL, Sosin H. Areas of failure found at reoperation (second or symptomatic
look) following curative surgery of adenocarcinoma of the rectum. Clinicopathological
correlation and implications for adjuvant therapy. Cancer 1974; 34:1278–​92.
34. Roels S, Duthoy W, Haustermans K, et al. Definition and delineation of the clinical target
volume for rectal cancer. International Journal of Radiation Oncology, Biology, Physics 2006;
65:1129–​42.
35. Myerson RJ, Garofalo Mc, Naqa I, et al. Elective clinical target volumes for conformal
therapy in anorectal cancer: A radiation therapy oncology group consensus panel
contouring atlas. International Journal of Radiation Oncology, Biology, Physics 2009;
74:824–​30.
36. Nijkamp J, Kusters M, Beets-​Tan RG, et al. Three-​dimensional analysis of recurrence
patterns in rectal cancer: the cranial border in hypofractionated preoperative radiotherapy
can be lowered. International Journal of Radiation Oncology, Biology, Physics 2011;
80:103–​10.
37. Joye I, Haustermans K. Clinical target volume delineation for rectal cancer radiation
therapy: time for updated guidelines? International Journal of Radiation Oncology, Biology,
Physics 2014;1:690–​1.
38. Valentini V, Gambacorta MA, Barbaro B, et al. International consensus guidelines on
Clinical Target Volume delineation in rectal cancer. Radiotherapy and Oncology 2016;
20(2):195–​201.
39. Sugihara K, Kobayashi H, Kato T, et al. Indication and benefit of pelvic sidewall dissection
for rectal cancer. Diseases of the Colon and Rectum 2006; 49:1663–​72.
40. International Commission on Radiation Units and Measurements. Prescribing, recording
and reporting photon beam therapy. Report 50. Bethesda, MD: ICRU, 1993.
41. Koeck J, Kromer K, Lohr F, et al. Small bowel protection in IMRT for rectal cancer: A
dosimetric study on supine vs. prone position. Strahlentherapie und Onkologie 2017;
193:578–​88.
42. Rodat V, Flentje M, Engenhart R, et al. The belly-​board technique for the sparing of
small intestine. Studies on positioning accuracy taking into consideration conformational
irradiation techniques. Strahlentherapie und Onkologie 1995; 171:437–​43.
194 Rectal cancer

43. Koelbl O, Vordermark D, Flentje M. The relationship between belly board position
and patient anatomy and its influence on dose-​volume histogram of small bowel for
postoperative radiotherapy of rectal cancer. Radiotherapy and Oncology 2003; 67:345–​9.
44. Nijkamp J, de Jong R, Sonke JJ, et al. Target volume shape variation during hypo-​
fractionated preoperative irradiation of rectal cancer patients. Radiotherapy and Oncology
2009; 92:202–​9.
45. Chong I, Hawkins M, Hansen V, et al. Quantification of organ motion during
chemoradiotherapy of rectal cancer using cone-​beam computed tomography. International
Journal of Radiation Oncology, Biology, Physics. 2011; 81:e431–​38.
46. Yamashita H, Takenaka R, Sakumi A, et al. Analysis of motion of the rectum during
preoperative intensity modulated radiation therapy for rectal cancer using cone-​beam
computed tomography. Radiation Oncology 2015; 10:2.
47. Wiltshire KL, Ward IG, Swallow C, et al. Preoperative radiation with concurrent
chemotherapy for resectable rectal cancer: effect of dose escalation on pathologic complete
response, local recurrence-​free survival, disease-​free survival, and overall survival.
International Journal of Radiation Oncology, Biology, Physics 2006; 64:709–​16
48. Jakobsen A, Ploen J, Vuong T, et al. Dose-​effect relationship in chemoradiotherapy
for locally advanced rectal cancer: a randomized trial comparing two radiation doses.
International Journal of Radiation Oncology, Biology, Physics 2012; 84:949–​54.
49. Appelt AL, Pløen J, Vogelius IR, et al. Radiation dose-​response model for locally advanced
rectal cancer after preoperative chemoradiation therapy. International Journal of Radiation
Oncology, Biology, Physics 2013; 85:74–​80.
50. Burbach JP, den Harder AM, Intven M, et al. Impact of radiotherapy boost on
pathological complete response in patients with locally advanced rectal cancer: a
systematic review and meta-​analysis. Radiotherapy and Oncology 2014; 113:1–​9
51. Appelt AL, Vogelius IR, Pløen J, et al. Long-​term results of a randomized trial in locally
advanced rectal cancer: No benefit from adding a brachytherapy boost. International
Journal of Radiation Oncology, Biology, Physics 2014; 90:110–​8.
52. Radu C, Norrlid O, Brændengen M, et al. Integrated peripheral boost in preoperative
radiotherapy for the locally most advanced non-​resectable rectal cancer patients. Acta
Oncologica 2013; 52:528–​37
53. Habr-​Gama A, Gama-​Rodrigues J, São Julião GP, et al. A local recurrence after
complete clinical response and watch and wait in rectal cancer after neoadjuvant
chemoradiation: impact of salvage therapy on local disease control. International Journal of
Radiation Oncology, Biology, Physics 2014; 88:822–​8.
54. Renehan AG, Malcomson L, Emsley R, et al. Watch-​and-​wait approach versus surgical
resection after chemoradiotherapy for patients with rectal cancer (the OnCoRe project): a
propensity-​score matched cohort analysis. Lancet Oncology 2016; 17:174–​83.
55. Martens MH, Subhani S, Heijnen LA, et al. Can perfusion MRI predict response to
preoperative treatment in rectal cancer? Radiotherapy and Oncology 2015; 114:218–​23.
56. Gérard JP, Chamorey E, Gourgou-​Bourgade S, et al. Clinical complete response (cCR)
after neoadjuvant chemoradiotherapy and conservative treatment in rectal cancer. Findings
from the ACCORD 12/​PRODIGE 2 randomized trial. Radiotherapy and Oncology 2015;
115:246–​52.
57. Frin AC, Evesque L, Gal J, et al. Organ or sphincter preservation for rectal cancer. The role
of contact X-​ray brachytherapy in a monocentric series of 112 patients. European Journal of
Cancer 2017; 72:124–​36.
 References 195

58. Gerard JP, Roy P, Coquard R, et al. Combined curative radiation therapy alone in (T1)
T2–​3 rectal adenocarcinoma: a pilot study of 29 patients. Radiotherapy and Oncology 1996;
38: 131–​7.
59. Rauch P, Bey P, Peiffert D, et al. Factors affecting local control and survival after treatment
of carcinoma of the rectum by endocavitary radiation: a retrospective study of 97 cases.
International Journal of Radiation Oncology, Biology, Physics 2001; 49:117–​24.
60. Appelt AL, Pløen J, Harling H, et al. High-​dose chemoradiotherapy and watchful
waiting for distal rectal cancer: a prospective observational study. Lancet Oncology 2015;
16:919–​27.
61. Vuong T, Devic S, Podgorsak E. High dose rate endorectal brachytherapy as a neoadjuvant
treatment for patients with resectable rectal cancer. Clinical oncology (Royal College of
Radiologists (Great Britain)) 2007; 19:701–​5.
62. Breugom AJ, Vermeer TA, van den Broek CB, et al. Effect of preoperative treatment
strategies on the outcome of patients with clinical T3, non-​metastasized rectal cancer: A
comparison between Dutch and Canadian expert centers. European Journal of Surgical
Oncology 2015; 41:1039–​44.
63. Corner C, Bryant L, Chapman C, et al. High-​dose-​rate afterloading intraluminal
brachytherapy for advanced inoperable rectal carcinoma. Brachytherapy 2010; 9:66–​70.
64. Appelt AL, Pløen J, Harling H, et al. High-​dose chemoradiotherapy and watchful
waiting for distal rectal cancer: a prospective observational study. Lancet Oncology 2015;
16:919–​27.
65. Kusters M, Marijnen CA, van de Velde CJ, et al. Patterns of local recurrence in rectal
cancer: a study of the Dutch TME trial. European Journal of Surgical Oncology 2010;
36:470–​6.
66. Kusters M, Dresen RC, Martijn H, et al. Radicality of resection and survival after
multimodality treatment is influenced by subsite of locally recurrent rectal cancer.
International Journal of Radiation Oncology, Biology, Physics 2009; 75:1444–​9.
67. Das P, Delclos ME, Skibber JM, et al. Hyperfractionated accelerated radiotherapy for
rectal cancer in patients with prior pelvic irradiation. International Journal of Radiation
Oncology, Biology, Physics 2010; 77:60–​5.
68. Bujko K, Wyrwicz L, Rutkowski A, et al. Long-​course oxaliplatin-​based preoperative
chemoradiation versus 5 × 5 Gy and consolidation chemotherapy for cT4 or fixed cT3
rectal cancer: results of a randomized phase III study. Annals of Oncology. 2016; 27:834–​42.
69. Guren MG, Undseth C, Rekstad BL, et al. Reirradiation of locally recurrent rectal cancer: a
systematic review. Radiotherapy and Oncology 2014; 13:151–​7.
70. van Dijk TH, Tamas K, Beukema JC, Beets GL, Gelderblom AJ, de Jong KP, et al.
Evaluation of short-​course radiotherapy followed by neoadjuvant bevacizumab,
capecitabine, and oxaliplatin and subsequent radical surgical treatment in primary stage IV
rectal cancer. Annals of Oncology. 2013; 24:1762–​69.
71. Bisschop C, van Dijk TH, Beukema JC et al. Short-​course radiotherapy followed by
neoadjuvant bevacizumab, capecitabine, and oxaliplatin and subsequent radical treatment
in Primary Stage IV rectal cancer: long-​term results of a Phase II study. Annals of Surgical
Oncology 2017; 24:2632–​38.
72. Anitei MG, Zeitoun G, Mlecnik B, et al. Prognostic and predictive values of the
immunoscore in patients with rectal cancer. Clinical Cancer Research 2014; 20:1891–​9.
 Chapter 9

Squamous cell carcinoma

of the anus
Rob Glynne-​Jones and Mark Harrison

9.1 Introduction
Squamous cell cancer (SCC) of the anus is rare, with an annual incidence of approxi-
mately 1.8 in 100,000 (approximately 1100 cases per year in the UK, and 7000 in the
USA), and has been increasing over the past three decades(1,2). Anal cancers are more
common in women than men(3) and just under half the patients are over the age of 65.
There is usually an indolent natural history and a low rate of distant metastases unless
the primary tumour is uncontrolled or recurrent(4–​9).
Loco-​regional failure is often the first event and leads to subsequent deaths related
to SCC of the anus (SCCA), hence, achieving local control is the over-​riding aim of
treatment. Primary surgery has given way to chemoradiation. If chemoradiation is
successful, then most patients preserve their anal sphincter, but anorectal function is
often moderately impaired(10). About one-​third of patients experience late gastrointes-
tinal toxicity with grade 3 and above (classified as severe).
The randomized trials in SCCA have tested chemoradiotherapy regimens with three
cytotoxic agents as partners with the radiotherapy (5-​fluorouracil (5FU), mitomycin
C (MMC), and cisplatin). Timing of administration has been similar (days 1–​5 and
days 29–​33), but the number of doses and total mg delivered have been different. The
majority of these phase III trials have used chemoradiotherapy (CRT) with 5FU and
MMC, either as a novel or standard arm (5,9,11,12,13). This regimen using 5FU and MMC
remains the current standard of care which is recommended in European and US
guidelines(14,15).
The relative 5-​year survival rate is 60–​75% depending on stage, and has changed
little for patients treated in the last two decades. Despite different eligibility criteria,
5-​year survival appears very similar in phase III trials. The MMC control arm of
the Radiotherapy Therapy Oncology Group RTOG 98-​11 trial(16) reported a 5  year
survival of 78% compared with 71% in the CRT-​alone arms of the Action Clinique
Co-​ordonees en Cancerologie Digestive (ACCORD-​03)(17) and 79% in the MMC arm
of the Anal Cancer Trial (ACT II)(9) respectively. Further updates of the RTOG 98-​11
trial show the tumour node (TN) status significantly impacts on overall survival (OS)
in patients treated with CRT and offers strong prognostic information(18).
Human papilloma virus (HPV) infection has a recognized role in the development
of SCCA(19) and is reported in 80–​90% of cases(20,21) (predominantly HPV16 or HPV18
 Introduction 197

subtypes in Europe). HPV is virtually endemic, with a lifetime risk of acquiring genital
HPV at least once of > 80%(22), but there are no validated interventions to treat active
HPV infection. Immunization in childhood may reduce the risk of HPV-​related ma-
lignancies such as SCAA but this, as yet, is unproven.
Immuno-​suppression is a further important risk factor; particularly in renal and car-
diac transplant recipients, who experience a ten-​fold higher risk compared to the gen-
eral population. Long-​term use of azothiaprin and corticosteroids, and autoimmune
disorders such as systemic lupus are also risk factors. Human immunodeficiency virus
(HIV) confers a 30-​fold higher risk of SCCA compared to the general population(23).
Anal cancer is strongly associated with cigarette smoking,(24) possibly by making it
more difficult to clear the HPV virus infection. In a case-​control study current cigar-
ette smoking was a major risk factor in both sexes, relative risk 7.7 in women and 9.4 in
men(25). Continuing to smoke can also adversely influence outcomes from treatment.
Tumours of the anal canal are often poorly differentiated SCC, in contrast to anal
margin tumours which are usually well differentiated, but grading is subject to con-
siderable inter-​observer variability and considerable heterogeneity in larger tumours.
Generally, males have an excess of well-​differentiated tumours arising in the anal
margin. Although high-​grade tumours are generally considered to have a worse prog-
nosis, this has not been confirmed in multivariate analysis(26).
p16INK4A (p16) is a tumour suppressor protein. Positive immunohistochemical
staining for p16 is a surrogate marker for HPV involvement, based on the high con-
cordance between these two biomarkers. p16 does not distinguish HPV-​16 (the most
common HPV genotype) and more aggressive genotypes.
Small early cancers often cause few symptoms, and are sometimes diagnosed ser-
endipitously with the removal of anal tags. More advanced lesions present as non-​
healing ulcers, perineal pain, sensation of a mass, rectal bleeding, itching, mucous
discharge, tenesmus, and faecal incontinence. Tumours may also be diagnosed con-
comitantly with a benign anal condition such as haemorrhoids, anal fissure, or fistula.
Occasionally patients present with enlarged inguinal lymph nodes in the absence of
anal-​related symptoms.
The anal canal extends approximately 3–​5 cm in length, depending on the sex of the
patient, from the anal verge to the sphincter muscles. The anal margin is considered
as perianal skin with a 5-​cm radius from the anal verge. The superior portion of the
anal canal drains to perirectal nodes in the mesorectum and nodes along the superior
rectal vessels to the inferior mesenteric system, and after to the para-​aortic nodes,
with additional drainage to the internal iliac and obturator nodes. In tumours arising
above the dentate line, lymphatic drainage is again to mesorectum and also via internal
pudendal nodes to the internal iliac system. Inguinal and mesorectal nodes are com-
monly involved in more advanced T stages. Clinically palpable (inguinal) lymph nodes
are found in 16–​25% of cases(13,27,28).
There have been considerable advances in prevention, imaging, and treatment
of SCCA and major developments in terms of our understanding of the molecular
biology and the role of the immune system. This chapter will focus on the treatment of
SCC arising in the anal canal and margin in terms of chemoradiotherapy, the chemo-
therapy agents, the radiation dose, field size, fraction size, brachytherapy, intensity
198 Squamous cell carcinoma of the anus

modulated radiotherapy (IMRT), target delineation, dose-​prescription, verification,

and future modalities such as protons.

9.1.1  Human immunodeficiency virus

Patients with either low CD4 counts or positive HIV status are typically a decade
younger than the average. The results of the AMC045 trial(29) give confidence to
clinicians to treat patients with HIV the same way as immunocompetent patients.
Experience is generally limited to super-​specialized centres, but in the era of highly
active antiretroviral therapy (HAART), HIV-​positive patients can expect similar out-
comes to HIV-​negative patients in terms of local control—​although current reports
still differ as to the tolerability and toxicity of treatment.

9.2 Classification
Staging is based on the TNM (tumour, nodes, metastases) classification developed
by the American Joint Committee on Cancer (AJCC) and the Union Internationale
Contre le Cancer (UICC) (Table 9.1). Because few cancers are resected surgically,

Table 9.1  TNM classification for anal cancer

Stage T* N+ M
0 Tis N0 M0

I T1 N0 M0
II T2 N0 M0
T3 N0 M0
IIIA T1 N1 M0
T2 N1 M0
T3 N1 M0
T4 N0 M0
IIIB T4 N1 M0
Any N2 M0
Any N3 M0
IV Any Any M1
*
Tumour stages: Tis, carcinoma in situ; T1, < 2 cm; T2, 2–​5 cm; T3, > 5 cm; T4, invading adjacent organs
but not anal sphincter.
+
Nodal stages: N0, no regional nodes; N1, perirectal nodes; N2, unilateral internal iliac or inguinal nodes;
N3, perirectal and inguinal, or bilateral internal iliac or bilateral inguinal nodes.
Edge SB, Byrd DR, Compton CC, eds. AJCC Cancer Staging Manual, 7th ed. New York, NY.: Springer 2010.
Reproduced with the permission from the American Joint Committee on Cancer (AJCC), Chicago, Illinois.
The original source for this material is the AJCC Cancer Staging Manual, Seventh Edition (2010) published
by Springer Science and Business Media LLC, www.springer.com. Although the 8th edition has been
published, implementation of the new system has been delayed until January 1, 2018. The AJCC advises
that, ‘All newly diagnosed cases through December 31st 2017 should be staged with the 7th edition.’
© Society of Surgical Oncology 2010.
 Classification 199

this classification is based on clinical factors such as tumour size (assessed by clin-
ical examination and imaging studies). Nodal status is based on distance from the
primary site rather than the number of nodes involved, as this has more prognostic
significance, and it should be noted that change to TNM8 in 2018 has led to substan-
tial changes.
All tumours of anal margin and perianal skin within 5cm of the anal margin are now
classified as carcinomas of the anal canal. T and M categories are unchanged, but N2
and N3 are abolished in the N category and external iliac lymph nodes become re-
gional lymph nodes-​such that N categories are defined as:
◆ N0 No regional lymph node metástasis.
◆ N1 Metastasis in regional lymph node(s).
◆ N1a Metastases in inguinal, mesorectal, and/​or internal iliac nodes.
◆ N1b Metastases in external iliac nodes.
◆ N1c Metastases in external iliac and in inguinal, mesorectal and/​or internal.
Surgical resection was standard treatment in the past, requiring removal of the anal
canal and a permanent stoma. Local control rates were reported between 53% and
73%, and 5-​year OS rates between 40 and 70% (depending on stage and extent of the
disease at presentation)(4,30). There has never been a randomized trial comparing rad-
ical surgery with radiotherapy or CRT
Historically, radiation alone with high total doses and split-​course schedules or
interstitial brachytherapy were used. However, the pioneering work of Nigro(31,32) and
subsequent confirmatory studies in the USA highlighted the efficacy of CRT with high
rates of local control using relatively low doses of fractionated radiotherapy (30–​45
Gy) combined with 5FU and MMC. Interstitial implantation of radioactive sources as
a boost after external beam radiotherapy(33,34) continues to be used in parts of Europe,
but requires considerable expertise.
The evidence for the advantage of CRT over radiation alone is based on a series
of Phase III trials which compared radiotherapy with radiotherapy and concurrent
5FU and MMC. Two initial randomized trials(5,12) compared a radiotherapy schedule
of 45 Gy boosted with a further 15–​25Gy after a gap of 6 weeks against an identical
regimen with concurrent 5FU/​MMC. These trials showed radiation alone could result
in local control in approximately 45–​55% of patients. However, both trials confirmed
chemoradiotherapy significantly improved oncological outcomes over radiation alone.
Substantial acute toxicity was reported for the use of MMC, so concurrent 5FU and
MMC or 5-​FU alone were compared in the CRT arm in the RTOG 87-​04 trial(10).
The addition of MMC significantly improved both disease-​free survival (DFS) and
colostomy-​free survival(CFS) (10). Subsequent randomized trials(12,13) also confirmed
the efficacy of concurrent CRT with 5FU/​MMC and relegated the role of surgery to
salvage of CRT failures. The small ACCORD-​03 trial in contrast used concurrent 5FU/​
cisplatin(17).
The strategies of utilizing cisplatin either as an alternative radiation-​sensitizer or
systemically as induction before CRT or maintenance treatment after CRT have not
improved outcome. The RTOG-​9811(13) and the ACCORD-​03 Phase III trials(17) tested
induction chemotherapy with cisplatin and radiation dose-​escalation with no benefit
200 Squamous cell carcinoma of the anus

either in local control or DFS. In the RTOG 9811 trial both disease-​free survival and the
colostomy rate were significantly inferior in the arm testing induction chemotherapy
with cisplatin. In ACC0RD-​03 induction with cisplatin failed to improve colostomy
free survival (the main endpoint).The colostomy rate appears higher with cisplatin(13).
Results of the ACT II trial(9) failed to confirm an advantage in terms of outcome by
replacing cisplatin in the CRT, or from additional 5FU and cisplatin-​based mainten-
ance chemotherapy. There was no difference between MMC-​based and cisplatin-​based
CRT in response at 6 months or DFS and OS—​although MMC provoked more haem-
atological toxicity.
Standard CRT administered with large parallel-​opposed fields led to substantial
acute toxicity, which includes painful desquamation, of inguinal and perianal skin,
diarrhoea, and a loss of bone marrow reserve. The RTOG 98-​11 reported acute non-​
haematological grade 3 or 4 toxicity of 74% in both MMC/​5FU and cisplatin/​5FU
groups(13). The ACT II trial reported grade 3–​4 adverse events at 71% and 72%, in
the MMC/​5FU and cisplatin/​5FU groups respectively(9). The large randomized CRT
trials have shown that approximately 60–​70% of patients experience grade 3–​4 haem-
atological toxicity(9,13). Investigational radiation studies in various pelvic malignancies
show a relationship between radiation dose to pelvic bone marrow and haematological
toxicity(35,36), which reflects the high proportion of the active bone marrow located in
the pelvis and lumbar spine(37), which can lie within the radiation field
IMRT is a technique that allows conformal dose distribution, limiting unnecessary
dose to nearby normal tissues. Prospective data have demonstrated modest advantages
for IMRT over 3D-​CRT for anal cancer as regards treatment-​related acute toxicity in
terms of grade 3 + gastrointestinal, genitourinary, and dermatologic toxicities(38), so
IMRT is now standard of care .Yet, low-​dose pelvic radiation is still sufficient to cause
significant bone marrow suppression. Severe haematological toxicity increases the risk
of infection, bleeding, and fatigue, which have significant adverse effects and may lead
to treatment breaks, which are associated with adverse outcomes.

9.3  Specific Phase III trials that establish

chemoradiotherapy as the treatment of choice
The UK trial(11) recruited 585 patients(3) and demonstrated at a median follow-​up of
42 months a 46% reduction (95% CI 0.42–​0.69, χ2 = 24.6, p < 0.0001) in the risk of
local treatment-​failure using combined modality treatment (CMT) with 5FU/​MMC
and radiation, over that achieved by radiation alone (61% vs 39% at 3 years). A first
phase of treatment using 45 Gy in 20 or 25 fractions using large parallel-​opposed fields
was followed by a boost of 15 Gy in six fractions (external beam) or 25 Gy to the 85%
isodose (brachytherapy). There was also a reduced risk of death from anal cancer and
a non-​significant OS advantage. Long-​term data from ACT I with a median follow-​up
of 12 years, showed a 25% difference in time to local failure in favour of CRT over RT
alone(4) but no significant difference in survival.
The European Organisation for Research and Treatment of Cancer (EORTC) trial
of 110 patients used the same design as the ACT 1 trial but was restricted to T3/​T4
and T1–​4N + tumours(12). This trial also used a single dose of MMC on day 1 in the
Specific Phase III trials that establish chemoradiotherapy as the treatment of choice 201

combined modality arm and closed when the outcome of the UK trial was known.
Local control was also improved with chemoradiotherapy (68% vs 55% at 3 years).
The US trial run by the Radiation Therapy Oncology Group RTOG-​8704(10) ex-
plored the advantage of adding two courses of MMC at a dose of 10 mg/​m2to 5FU
and concurrent RT. The trial recommended an initial phase of RT, evaluation after
6 weeks and the delivery of a boost of 9 Gy with concurrent 5FU and cisplatin for
histologically confirmed residual disease, With a median dose of 48 Gy the RTOG-​
8704 confirmed the superiority of 5FU and MMC over 5FU alone when combined
with radiotherapy(10).
Studies vary in the proportion of early (T1 and T2) tumours between 16% in the
EORTC 22861 study, 41% in ACT I, and 57% in the RTOG-​8704. The T-​staging clas-
sification are also not the same, since early trials used the 1978 UICC, or the 1985
UICC classification based on anatomical extent, and the proportion of the circumfer-
ence involved by tumour. The populations treated and the endpoints are different with
varying definitions. Hence the outcome from these studies should not be compared.

9.3.1  RTOG  98–​11

The US Intergroup RTOG 98–​11 phase III trial(13) randomly assigned 682 patients with
anal canal tumours (35% with T3/​4 tumours, and 26% with clinically involved lymph
nodes to either neo-​adjuvant 5FU and cisplatin for two cycles prior to concurrent CRT
with 5FU and cisplatin (n = 341), or the standard arm of concurrent CRT with 5FU
and mitomycin (n = 341). The primary endpoint was 5-​year DFS.
The neo-​adjuvant cisplatin-​based chemotherapy arm failed to improve DFS, loco-​
regional control and distant relapse, or OS. In fact, initial results favoured the control
arm of mitomycin. The 5-​year DFS rates were 61% in the mitomycin arm vs 54% for
the cisplatin arm, and 5-​year OS rates were 75% and 70%, respectively (p = 0.1). More
mature data showed the outcome for cisplatin is significantly inferior for 5 year DFS
(67.7 vs 57.6%; p = 0.0044) and 5-​year OS (78.2 vs 70.5% respectively; p = 0.02)(16). The
requirement for a colostomy was also significantly higher in the cisplatin arm com-
pared with the mitomycin arm (19% vs 10%; p = 0.02). Haematological toxicity was
worse with mitomycin. Similar compliance to both chemotherapy and radiotherapy
in each arm suggests that the differences did not relate to excess toxicity. Both neo-​
adjuvant and concurrent cisplatin was used in the same experimental arm of the trial,
however, making analysis of the individual role of each strategy difficult. In total, 10%
of patients developed long-​term toxicity after chemoradiotherapy, with 5% requiring a
colostomy for treatment-​related problems.

9.3.2  ACCORD-​03
The Action Clinique Coordonees en Cancerologie Digestive ACCORD-​03 phase III
trial(17) tested neo-​adjuvant chemotherapy (NACT) with cisplatin and also a radiation
dose-​escalation with concurrent 5FU/​cisplatin in a factorial 2 × 2 trial design. The trial
compared 45 Gy in 25 daily fractions plus a 15-​Gy boost with a higher boost dose, but
found no benefit in CFS at doses above 59 Gy. Event-​free survival for the induction
high dose and high dose-​alone arms were 78% and 68% compared to the reference
202 Squamous cell carcinoma of the anus

arm with 67%. In contrast the colostomy-​free survival was 85% and 80% arms respect-
ively, compared to 86% in the control arm(17). Although not significant (p = 0.06), these
results imply that the boost could improve control—​but at the expense of a higher risk
of a colostomy. This result was used to support the dose-​escalation in the current UK
Plato trial.

9.3.3  Anal cancer trial—​ACT II

The ACT II multi-​centre, randomized trial recruited 940 patients in a 2 × 2 factorial
design and enrolled patients with histologically confirmed SCCA without metastatic
disease from 59 centres in the UK(9). Patients received 5FU (1000 mg/​m2/​day on days
1–​4, 29–​32) and radiotherapy (50.4 Gy in 28 daily fractions), and were randomized
to receive MMC (12 mg/​m2, day 1; (n = 472) or high-​dose cisplatin RG (60 mg/​m2 on
day 1, 29; (n = 468). A second randomization directed two courses of consolidation
therapy (n = 492) 5 and 8 weeks after CRT (5FU/​CDDP, i.e. weeks 11, 14), or no con-
solidation (n = 448). Response was assessed clinically at 11 and 18 weeks, and by com-
puted tomography (CT) at 6 months. Overall 391/​432 (90.5%) patients in the MMC
group vs 386/​431 (89.6%) in the cisplatin group achieved a complete response at 26
weeks (p = 0·64), with similar overall toxicity for each group.
Results show 3-​year recurrence-​free survival rates of 75% in T1/​2 tumours overall,
and 68% for more advanced T3/​T4 tumours(9). Hence neither strategy, i.e. CRT with
cisplatin vs CRT with MMC, nor chemotherapy consolidation with cisplatin was more
effective for achieving complete clinical response (CCR), reducing tumour relapse or
cancer-​specific deaths than the standard of MMC/​CRT. Acute haematological toxicity
was more pronounced for MMC, but non-​haematological toxicity similar.
In summary, the RTOG-​9811, ACCORD-​03, and ACT II phase III trials in anal
cancer showed no benefit for cisplatin-​based induction and maintenance chemo-
therapy, or radiation dose-​escalation above 59 Gy. Neither the RTOG-​9208 trial nor
the ACCORD-​03 trial, which compared radiation dose-​escalation from 60 Gy to
65–​70 Gy support the view that radiation dose-​escalation within a CRT schedule in-
creases local control in anal cancer. Normal tissue effects may be dose-​limiting over
a given threshold, and outweigh any potential tumoricidal advantages from higher
radiation doses.
The present authors feel that 5FU and a single dose of MMC (12 mg/​m2 day 1) is the
recommended standard.

Oral fluoropyrimidines
Capecitabine is an orally administered fluoropyrimidine, which offers an equally ef-
fective alternative to 5FU. An initial phase II study successfully substituted Capecitabine
at a dose of 825 mg/​m2 twice a day on radiation days and MMC on day 1 for infusional
5-​FU(39) . Subsequent reports have confirmed both the safety and efficacy of oral
capecitabine at this same dose (825 mg/​m2 twice a day) and retrospective comparisons
of capecitabine and infusional 5FU show a trend towards lower recurrence rates with
capecitabine (40). A  recent national United Kingdom audit of 242 patients managed
with IMRT and a single dose of MMC with either 5-​FU (5-​FU/​MMC) or capecitabine
(capecitabine/​MMC) showed similar G3/​G4 toxicity as well as oncological outcomes
 Brachytherapy 203

at 1 year(41). Hence, capecitabine is recommended as a suitable alternative to 5-​FU by

NCCN, ESMO-​ESSO-​ESTRO, and recent UK intensity-​modulated radiation therapy
(IMRT) guidelines(14, 15).

9.4 Radiotherapy dose
The determination of optimal dose fractionation is limited by a lack of data regarding
the pattern of failure. No randomized phase III study has published the site(s) of local
failure (within, marginal to, or outside of the radiotherapy field). Therefore it remains
uncertain whether the majority of loco-​regional failure is due to inadequate clinical
target volumes (CTVs), insufficient radiation dose, or intrinsic radio-​resistance, but
preliminary data from the ACT II trial suggest the majority of loco-​regional failure is
in-​field(8). The ACT II trial mandated a total dose of 50.4 Gy, which probably delivered
in the region of 53 Gy. Many centres in Europe and other parts of the world routinely
administer higher doses.

9.4.1  Gaps in treatment

The ACT I, EORTC, and RTOG 87-​04 trials scheduled a 4–​6-​week treatment gap after
the initial 45 Gy followed by a boost. Split-​course radiotherapy with high total doses
and the use of interstitial implants were based on the tradition of Papillon(33) allowing
sufficient time for the bulk of the tumour to shrink and hence facilitate an interstitial
implant to the smallest possible volume. This practice minimized the risk of necrosis
in the high dose area, and selected patients who fail to respond to be eligible for sur-
gical salvage. In contrast, ACT II mandated a continuous 50.4 Gy regimen in 28 daily
fractions, which was proposed as the reason for the high complete response rate and
improved DFS rates relative to ACT I(9).
Studies in anal cancer in Europe have continued to advocate a gap but limit to 2
weeks if clinically feasible. However, this practice defies standard radiotherapy prin-
ciples regarding overall treatment time, breaks in treatment, and the concept of re-
population. The use of IMRT reduces unplanned treatment breaks and shortens the
overall treatment time compared with previous techniques. If a repopulation constant
is presumed to have a value of approximately 0.6 Gy/​day, this hypothesis means only
an additional 2.4 Gy would have been delivered. The present authors believe the use of
a planned gap is detrimental to local control, and should be avoided(42, 43).

9.5 Brachytherapy
Brachytherapy may potentially increase dose to the primary tumour in T3/​T4 tumours
but requires skill and expertise to avoid radionecrosis due to an unsatisfactory dose
distribution. A low dose-​rate iridium interstitial implant was originally advocated as a
boost following radiotherapy alone(33) after an interval of 6 weeks, but this technique
does not achieve current standards of conformal treatment. The strategy influenced
the design of the two early European phase III studies(5,12), where a brachytherapy
boost could be delivered using 25 Gy following CRT after an interval of 6 weeks.
Enthusiasm that brachytherapy achieves better outcomes in terms of local control
204 Squamous cell carcinoma of the anus

probably reflects patient selection—​patients with smaller tumours or good responders

to chemoradiation being preferentially selected for brachytherapy boost.

9.6  Radical primary treatment

9.6.1 Indications
Patients who should receive radical CRT
Definitive chemoradiation with curative intent is recommended for patients with an
initial diagnosis of epidermoid (squamous) carcinoma of the anal canal and margin
in whom metastatic disease has been excluded by a staging CT scan of thorax and ab-
domen. It is important to emphasize that patients should be assessed for their fitness to
receive this relatively intensive treatment approach taking into account performance
status and medical comorbidity—​particularly renal function. It is assumed that that
scattered dose of radiation to the testes is very likely to induce permanent sterility and
therefore sperm banking should be discussed with male patients who wish to preserve
fertility prior to the commencement of treatment.

Patients in whom radical CRT may be considered but where

caution is necessary
There are a number of clinical scenarios where definitive chemoradiation may be in-
dicated but where there is an increased risk of treatment-​related toxicity or where
modifications to the standard treatment approach might be necessary. The scenarios
include:
◆ HIV infection
◆ Renal transplant
◆ Inflammatory bowel disease (IBD)
◆ Prior pelvic resectional surgery.
There is an increased incidence of anal cancer in renal transplant patients. This pro-
duces challenges due to the site of the transplant kidney in the pelvis. Depending on its
position this may require significant reduction in field sizes to avoid significant irradi-
ation to the kidney. Surgery may still therefore have a role in selected cases.
Patients with IBD and those who have undergone prior resectional pelvic surgery
are at increased risk of treatment-​related toxicity with radical chemoradiation. After
resectional pelvic surgery, the volume of small bowel in the pelvis is increased and this
may lead to increased risks of early and late small bowel complications. In patients
with IBD either the small or large bowel may be involved by this disease process and
limit the tolerance of these normal tissues to standard doses of radical chemoradiation.
These difficult cases need discussion within the multidisciplinary team. Although
there are increased risks of treatment-​related toxicity with radical chemoradiation,
there are also commonly increased risks of the alternative approach of radical surgery.
In the experience of the authors these difficult cases commonly receive modified CRT
and considerable multidisciplinary expertise is required to minimize the risks of treat-
ment related morbidity and mortality.
 Radical primary treatment 205

Patients who should not receive radical chemoradiation

It is preferable to consider local excision of small anal margin tumours < 2  cm in
diameter if there is no evidence of disease within the anal canal, there is no evidence of
nodal spread clinically or on imaging, and if clear lateral and deep margins are likely
to be obtained.
Sometimes patients are referred with symptoms and in whom biopsy has shown
anal intraepithelial neoplasia, which can be graded from 1 to 3 in severity. The as-
sessment and management of high-​grade anal intraepithelial neoplasia (AIN II/​III) is
beyond the scope of this chapter. The natural history of progression from AIN3 to in-
vasive malignancy is poorly documented but a recent estimate is 10% after 5 years(44).
At present it is recommended that chemoradiation is reserved for the development of
established invasive carcinoma.
Radical chemoradiation should be used with great caution in patients who have re-
ceived prior pelvic radiation. Depending on the treatment previously given, it might be
possible to use modified chemoradiation to a total dose of 30–​40 Gy (see section 9.9.1.

9.6.2  Treatment volume and definition

Clinical examination is an extremely valuable method of assessment. The clinical on-
cologist requires different information to a surgical assessment at the time of examin-
ation under anaesthetic. The clinical oncologist is interested in accurate measurements
of tumour length, degree of circumferential involvement, and the extension of any
macroscopic disease in all directions including below the anal verge. Vaginal exam-
ination is essential in female patients to assess whether there is any extension into the
postvaginal wall or even breaching vaginal mucosa. It is also most important that this
examination is documented in the patient record to assist post-​treatment assessment
of response to CRT. Examination under anaesthetic allows more detailed palpation of
the anorectal and pelvic structures and is used particularly when clinical examination
is limited by discomfort or pain.

9.6.3 General points
Patients should be assessed for performance status, renal function, and other medical
comorbidity prior to treatment. Patients should be tested for relevant infections, and
other malignancies.
Assessment of the cervix, vagina, and vulva is suggested in female patients, and in-
cludes screening for vaginal and cervical cancer (and the penis in men), because of the
common role of HPV in these tumours.
HIV testing is recommended because of its implications of excess toxicity from
radiotherapy and chemotherapy and the development of infections. In HIV-​positive
patients, the CD4 count, measured viral load, and optimization of highly antiretroviral
therapy are all essential to determine the management plan.
Sperm banking should be discussed prior to the commencement of treatment with
male patients who wish to preserve fertility, because of the risk of permanent azoo-
spermia and offered the opportunity for sperm storage. Testosterone levels may be
reduced with impaired physical, psychological, and sexual function after treatment(45).
206 Squamous cell carcinoma of the anus

Erectile dysfunction may be due to low testosterone levels or nerve damage by radi-
ation. Testosterone hormone replacement may be beneficial in some.
Premenopausal women should be informed that fertility will be lost, and hormone
replacement therapy may be appropriate in those in whom an early menopause is in-
duced. A defunctioning colostomy should be considered in patients with transmural
vaginal involvement (at risk of development of an anorectal–​vaginal fistula), or faecal
incontinence.
Staging investigations should include:
◆ Clinical examination
◆ Whole body CT (chest, abdomen, and pelvis)
◆ High-​resolution MRI of the pelvis.
Magnetic resonance imaging (MRI) is more accurate in distinguishing and delineating
primary tumour and lymph nodes and offers the advantage of coronal sagittal and
axial views of the extent of the primary tumour and with the addition of diffusion
weighting offers a considerable aid to radiotherapy treatment planning.
The optimal method to determine inguino-​femoral lymph node status of patients
with anal canal cancer remains controversial. Clinical examination supplemented
by fine-​needle aspiration cytology (FNAC) is the traditional method(46) but may not
be sufficiently accurate. Approximately one-​third of patients have enlarged inguinal
lymph nodes but on biopsy only 50% will confirm metastatic spread. The remainder
are caused by secondary infection. Tumours of the anal margin are more likely to
involve inguinal nodes than anal canal tumours. In retrospective surgical series,
30% of patients will have involvement of their inguinal nodes; however, in early T1/​
T2 tumours the rate of involvement is approximately only 12%(47–​49). Involvement
is usually unilateral and occasionally bilateral but not usually contralateral to the
tumour.
Small shotty nodes may be inflammatory whereas palpable nodes > 1 cm are clearly
at risk of microscopic involvement. Clinically suspicious nodes should be assessed
by biopsy where possible. However, formal biopsy of these nodes significantly delays
CRT and leaves a surgical scar with the potential for seeding microscopic tumour
cells. In contrast a FNAC or core biopsy is only helpful if cancer cells are detected.
A  negative sample is compatible with either a sampling error of an involved node
or a truly uninvolved node. Where there is suspicion the nodes should be treated as
involved.
The pelvic nodes are frequently involved particularly with increasing tumour stage
and in poorly differentiated tumours. If the tumour extends up into the rectum, spread
may occur via the inferior mesenteric lymph nodes. The overall incidence of pelvic
lymph node metastases is in the region of 25–​30%(49).
Fludeoxyglucose positron emission tomography (FDG PET) has high sensitivity in
identifying involved lymph nodes, and a high specificity in immunocompetent pa-
tients, and may have a role to investigate equivocal lesions on CT or MRI not amenable
to biopsy. However, some series show that up to 40% of patients with PET/​CT avid
nodes are false-​positive results. Haematogenous spread at presentation is noted in <
5% of cases and predominantly involves lung or liver.
 Radical primary treatment 207

9.6.4  Sentinel lymph node biopsy

Sentinel lymph node biopsy (SLNB) has not fulfilled the initial hopes of this staging
modality, and in addition management strategies are not in place to stratify treatment
for the findings of macroscopic involvement, microscopic involvement, and the pres-
ence of isolated cells. The presence of inguinal node metastases which are not clinically
palpable is of uncertain relevance, since a worse survival relates to clinically involved
palpable inguinal lymph nodes. Some studies suggest that all patients found negative
for inguinal lymph node metastases at PET-​CT were also negative at SLNB, i.e. few
false-​negative results are observed. Also, neither a negative PET-​CT nor a negative
SLNB excludes involved nodes, as elective bilateral inguinal node dissection is not
normally performed. In addition, the morbidity of irradiating to 30.6 Gy or even 50.4
Gy after + SLNB, is unknown—​particularly as some SLNB will require bilateral nodes
to be removed. SLNB is probably most useful in the setting of loco-​regional recurrence
after CRT in deciding whether a radical inguinal dissection should be performed in
addition to salvage abdominoperineal resection.

9.6.5  Radiotherapy planning technique

The important principles of radical chemoradiation currently used in the UK are
that all microscopic disease at risk receives a minimum of 30–​40 Gy of radiation, and
secondly that all macroscopic disease receives at least 50 Gy. This was achieved by the
use of a two-​phase shrinking field technique in ACT II, but now more complex radio-
therapy techniques such as IMRT or VMAT are used, which allows conformal dose
distribution, limiting unnecessary dose to nearby normal tissues. Prospective data
have demonstrated modest improvement in treatment-​related acute toxicity.
The current approach in the UK is IMRT using the protocol designed for the cur-
rently recruiting UK PLATO trial and is described in the following sections. www.
analimrtguidance.co.uk

ACT 3,4 and 5planning technique

Patient position and immobilization
Patients are usually treated supine because this position is more reproducible and
stable. This position and IMRT do not allow direct visualization of the anal verge, and
hampers the application of perianal bolus (if required).There is no consensus on the
use of immobilization techniques, but IMRT requires immobilization of the patient,
the accurate delineation of the CTV, and precise definition of normal tissue constraints
and dose fractionation with a simultaneous integrated boost (SIB) that does not pro-
long the overall treatment time (OTT).
Target volume and field definition
Gross tumour volume  It is common for macroscopic disease to regress quickly during
chemoradiation. IMRT requires all target areas to be encompassed in a single radio-
therapy plan. In anal cancer, the pelvic nodes, inguinal nodes, and primary tumour
represent three different targets. All palpable macroscopic disease in the inguino-​
femoral and perianal region should be identified by the use of radio-​opaque markers
208 Squamous cell carcinoma of the anus

at initial CT simulation. The extent of gross tumour within the anal canal is diffi-
cult to determine on a CT planning scan, although inguinal nodes are easily imaged.
MRI (especially diffusion weighted MRI) and PET particularly if co-​registered can
be helpful.
All macroscopic disease in the inguino-​femoral and perianal region should be iden-
tified by the use of radio-​opaque markers at initial simulation. The extent of gross
tumour within the anal canal is difficult to determine on a non-​contrast CT planning
scan. A number of approaches are used. Firstly the measurements from clinical exam-
ination are extremely useful with the length of tumour extension up the anal canal
measured from the anal verge. Using a radio-​opaque marker on the anal verge the
superior extent of the gross tumour volume (GTV) can be determined in this way.
High-​resolution MRI also provides this measurement as well as measurements of
gross tumour extension in the lateral anterior and posterior directions. There is a need
to accurately fuse the diagnostic pelvic MRI in the supine position with the planning
CT images performed on a flat couch to improve definition of the GTV. If not available
the clinician must take measurements from the MRI scan and translate this on to the
planning CT scan. The use of rectal contrast, a vaginal tampon, or a radio-​opaque wire
in the anal canal may assist anatomical orientation/​localization on CT.
All areas of macroscopic disease should be identified as GTV. In patients with
clinically or radiologically significant lymphadenopathy these areas are identified as
separate GTVs.
Treatment fields (including the clinical target volume and planning target volume) There
is still little information on the pattern of failure of anal cancer, although evidence
from ACT II suggests that the majority of pelvic failures were reported in the high dose
volume(8) (see Table 9.2). Loco-​regional failure in the elective nodal volume treated
with 30.6 Gy appears rare, suggesting insufficient radiation dose or intrinsic radio-​
resistance rather than inadequate CTV is relevant.
There is also a need for further studies to determine the extent of organ motion and
departmental set up errors for the techniques used in the treatment of anal and rectal
cancer.

Table 9.2  Sites of relapse in ACT II trial

Site of relapse (includes primary site, Number % total
inguinal/​pelvic nodes or metastases) relapses
Pelvic—​no metastases 133 64%

Pelvic—​with metastases 30 14%

Distant metastases only 46 22%
Total crude pelvic failure (with or without 163 78%
metastases)
Total relapses 209

Source: data from Sebag-​Montefiore D, James R, Meadows H, The pattern and timing of

disease recurrence in squamous cancer of the anus: Mature results from the NCRI ACT II
trial. J Clin Oncol 30, 2012 (suppl; abstr 4029).
 Radical primary treatment 209

Basic treatment still acceptable—​but IMRT recommended

Phase 1—​large parallel-​opposed fields  This phase includes all sites of GTV and all
microscopic disease at risk in the inguino-​femoral and pelvic lymph nodes (Fig. 9.1)
and is used in all patients receiving radical chemoradiation. This approach may be
planned using a conventional or virtual simulator. All borders described below are
field. By describing field borders, this technique does not require the specific delinea-
tion of the CTVs or planning target volumes (PTVs). This very simple technique was
considered necessary at the start of the ACT II trial to ensure high protocol compliance.
◆ Superior border: 2 cm above inferior aspect of the sacroiliac joints. The superior
field border is standard unless pelvic lymph nodes are seen on CT scan or the pri-
mary tumour extends to within 3  cm of this border, in which case the border is
recommended to extend 3 cm above the upper limit of macroscopic disease.
◆ Lateral border:  to cover fully both inguinal nodal regions—​in practice this field
border is approximately the midpoint of the femoral neck.
◆ Inferior border: 3 cm below the anal verge (for disease confined to the anal canal
only) or 3 cm below most inferior extent of tumour (for anal margin tumours).

(a)
FIELDS defined as:-
Phase I
30.6 Gy in 17 fractions
Parallel opposed fields
Anal bolus
Phase II (N0 groins)
19.8 Gy in 11 fractions
3cm margin to GTV
Anal bolus

(b)
FIELDS defined as:-
Phase I
30.6 Gy in 17 fractions
Parallel opposed fields
Anal bolus
Phase II (N+ groins)
19.8 Gy in 11 fractions
3cm margin to all GTV
Parallel opposed fields
Anal bolus

Fig. 9.1  (a) Two-​phase radiotherapy technique N0. (b) Two-​phase radiotherapy

technique N+.
210 Squamous cell carcinoma of the anus

Parallel-​opposed fields are used with equal weighting. Wax bolus is used, placed
between the buttocks for all patients with anal margin tumour and in those patients
with anal confined disease that extends down to within 2 cm of the anal verge (the
vast majority of patients). A wedge-​shaped piece of bolus is more comfortable for the
patient than layers of bolus material and also corrects the differences in separation in
this region between the buttocks. A longitudinal wedge may improve the homogeneity
of the dose distribution in some patients. A minimum energy of 8 MV is used.
Dose prescription
◆ Phase 1—​30.6 Gy in 17 fractions of 1.8 Gy per fraction.

Phase 2
The technique used is different depending on the presence or absence of clinically or
radiologically significant lymphadenopathy and the position of the primary tumour.
◆ If there is no evidence of lymphadenopathy then the GTV is treated with a 3-​cm
margin for anal canal tumours.
◆ If the tumour is confined to the anal margin only, a direct photon field may be used
with 3-​cm lateral superior and inferior margins. Electrons are not recommended.
◆ If clinically or radiologically significant lymphadenopathy is present then the nodal
and primary GTVs are defined and reduced parallel opposed fields with 3-​cm lat-
eral superior and inferior margins are used.
This phase may be planned using orthogonal films, CT planning, or virtual simulation.
It is essential that all visible tumour at and around the anal margin is marked using a
radio-​opaque marker. Rectal contrast is also recommended. If the disease is confined
to the anal canal, then a radio-​opaque marker placed on the anal verge is essential.
All significant inguino-​femoral lymphadenopathy is also marked with radio-​opaque
markers. The GTV is determined using clinical and diagnostic MRI measurements.
Primary tumour without  significant lymphadenopathy—​anal canal A 3-​cm margin
is applied superiorly, laterally, anteriorly, and posteriorly to determine the treatment
fields. The inferior 3-​cm margin is applied to the anal verge if tumour is confined to
the anal canal or 3 cm inferior to the inferior extent of GTV if tumour extends inferior
to the anal verge marker.
A three-​or four-​field arrangement is used. This is usually posterior and two
wedged lateral fields. Occasionally an anterior fourth field is required to improve any
inhomogeneity.
Primary tumour without  significant lymphadenopathy—​anal margin only For anal
margin confined tumours a direct photon field (electrons should not be used) is used
and the margins (superior, inferior, and lateral) are 3 cm from the limits of the GTV
to define the treatment fields (see Fig. 9.1). Wax bolus as described for phase 1 is used
for phase 2.
Primary tumour with  significant lymphadenopathy  All macroscopic disease will be
considered as GTV including the primary site and all sites of involved nodal disease.
The field borders are defined as:
 Radical primary treatment 211

◆ Superior field border: 3 cm superior to the most superior extent of GTV

◆ Inferior field border: as phase I—​3cm below the anal margin (for disease confined
to the anal canal only) or 3  cm below most inferior extent of tumour (for anal
margin tumours)
◆ Lateral field border: 3 cm lateral to the most lateral extent of the GTV.
Lead shielding and multileaf collimation may be used to shield normal tissue pro-
viding a minimum of a 3-​cm margin from all delineated GTV is achieved.
Parallel-​opposed fields are used with equal weighting.
Dose prescription for phase 2 
◆ 19.8 Gy in 11 fractions at 1.8 Gy per fraction to the ICRU intersection point (for
anal canal tumours or when parallel-​opposed fields are used. When a direct field is
used for an anal margin tumour the dose is prescribed to the anal margin.

Implementation and quality assurance

The techniques described require electronic portal imaging(50) to ensure the reproduci-
bility of the treatment set up. In vivo dosimetry is not routinely used.
Fields should be moved if they fall outside an agreed tolerance level—​usually 5 mm
for patients who are treated prone. This process also allows radiographers to evaluate
the whole set-​up, and thus to assess and correct systematic errors. Reverification is
recommended on a weekly basis. These electronic portal imaging images should be
audited at the clinician’s weekly meeting.

Implementing complex radiotherapy techniques for anal cancer

The ACT II technique is associated with significant acute toxicity, particularly of the
perineal skin and genitalia. IMRT can reduce mean and threshold doses to genitals,
perineum, small bowel, and bladder and is therefore strongly recommended as the
treatment of choice.
The delineation of pelvic nodes radiologically is described in a recent review(51), and
other relevant non site-​specific pelvic nodal atlases(52), but is, to date, insufficiently
relevant to anal cancer. With routine use of CT simulation and intravenous contrast,
contouring vessels on the CT image can be used as a surrogate for lymph node local-
ization and can offer a more precise, and individualized field delineation compared to
that achieved when using conventional pelvic fields. In addition involved nodes can
be imaged on MRI(53).
On review of the literature we could find no meta-​analyses or systematic reviews
in anal cancer to offer guidance on the optimal total dose or target definition or for
selection and delineation of subclinical lymph nodal areas. The RTOG have made
suggestions based on a consensus of nine experts(54) and developed further when the
Australasian Gastro-​Intestinal Trials Group produced a consensus atlas for target de-
lineation and prophylactic nodal irradiation (55). The RTOG-​Protocol 0529 is useful
and lists dose constraints(38). In addition, in the UK a national initiative has developed
IMRT anal cancer guidance with a view to a national trial(56), which has been supple-
mented by recent genitalia contouring guidelines(57).
212 Squamous cell carcinoma of the anus

Although long-​term survival rates in SCCA are high, permanent radiotherapy-​

related morbidity is not uncommon, and subsets of patients do not respond
to standard CRT and demonstrate poor outcomes. These observations have
prompted the development of both dose de-​escalation and escalation clinical
protocols. Future UK studies plan to reduce dose to early tumours, and dose es-
calate to T3/​T4 tumours, using 3D-​conformal or IMRT, or VMAT techniques.
We have therefore calculated retrospectively the dose received by the PTV in
patients planned according to the ACT II protocol, for clinically involved areas
(primary tumour/​lymph nodes) and determined the dose to relevant uninvolved
lymph nodes (nodal CTVs), and organs at risk (OARs) to design a pilot con-
formal 3D-​regimen.
In future biological imaging based on PET/​CT, MRI, and magnetic resonance spec-
troscopy imaging, in conjunction with radiotherapy, may make dose painting feasible
according to radiobiological principles.

Intensity-​modulated radiotherapy
IMRT allows both precision and sparing normal surrounding structures (perineal
skin, external genitalia, and bladder) compared to conventional 3D planning(58–​
61)
, which may lead to reduced acute toxicity and fewer treatment breaks, which
may in the past have compromised efficacy. IMRT may also allow radical treatment
to be delivered to tumours with extensive nodal involvement without excess tox-
icity(61). However, in obese patients with non-​reproducible external skin contours,
or a major component of tumour outside the anal canal, IMRT may prove more
problematic.
For full optimization of IMRT plans and best outcomes robust, consistent con-
touring techniques for tumour, elective volumes, and OARs are needed. Toxicity to
the perineum skin and genitalia contribute to more difficult acute radiotherapy side
effects in SCCA, because the genitalia are anterior, close to the primary tumour, and
difficult to avoid completely.
Planning studies comparing conventional radiotherapy and IMRT show less gastro-
intestinal toxicity(62) with reduced dose to genitalia with IMRT(60), and less skin tox-
icity(64) although oncological outcomes are similar to 3D-​CRT(65).
IMRT is challenging even for experienced clinicians. In a recent multicentre
study, even after centres had been approved and accredited, 79% of IMRT plans
required field modification of elective nodes after central review(38). The RTOG-​
0529 Phase II study confirms that IMRT can reduce the OTT. When the RTOG-​
9811 and the RTOG-​0529 studies were compared, the median RT duration was
49  days and 42  days respectively. Physician experience may be crucial because
higher volume radiation oncology centres appears to achieve better oncological
outcomes(66).
IMRT in the future offers the promise of less toxicity, shorter OTT, and potentially
higher radiotherapy doses. In order to use IMRT or VMAT, we need to define subclin-
ical areas, which potentially harbour microscopic disease, the optimal dose for macro-
scopic and microscopic disease. We also need to decide a suitable prescribed planning
dose to the PTV and an appropriate PTV margin.
 Radical primary treatment 213

PersonaLising Anal cancer radioTherapy dOse (PLATO)

PLATO is an umbrella study which encompasses three randomized phase 2 studies
with different eligibility criteria, which examine management of various stages
of SCCA.
Anal Cancer 3 (ACT3) randomizes patients with SCC of the anal margin into either
a watch and wait or post-​operative radiotherapy arm. The radiation treatment field is
localized and a dose of 41.4 Gy in 1.8 Gy fractions given with concomitant chemo-
therapy. For the PLATO trial 5FU has been substituted with capecitabine.
Anal Cancer 4 (ACT4) looks at early stage anal cancer—​up to 4 cm and node nega-
tive and asks whether radiation dose de-​escalation is possible. Patients are randomized
to either 50.4 Gy in 28 fractions or 41.4 in 23 fractions with concomitant chemo-
therapy. The radiation volumes encompass the primary and draining pelvic nodes.
Anal Cancer 5 (ACT5) examines radiation dose escalation for more advanced tu-
mours with patients being randomized to higher doses of radiation to the primary
(and nodes if greater than 3 cm in diameter).
To facilitate PLATO, guidelines to help with IMRT planning were developed.
These have had several iterations but are now accepted within the community of
oncologists treating anal cancer. Essential requisites for planning of this nature in-
clude contrast-​enhanced CT, MRI and, if at all possible, PET-​CT. These facilitate
the development of GTVs to the primary and any pathologically involved nodes,
CTVs to the primary and nodes (CTVA and N) as well as an elective nodal volume
(CTVE) which encompasses the draining lymph nodes of the internal and external
iliac vessels.
For each of the trials the planning algorithm is as follows:
ACT3:
GTV_​A = Includes the site of primary tumour and excision scar.
CTV_​A = GTV_​A + 10 mm. Following this, manually enlarge to ensure coverage of
entire anal canal including outer border, from the ano-​rectal junction (approxi-
mately 4 cm superiorly from anal verge identified by the radio-​opaque marker)
to the anal verge including the internal and external anal sphincters. Edit to ex-
clude bone and muscle. (See Fig. 9.1.)
PTV_​A = CTV_​A + 10 mm
ACT4:
GTV_​A = Includes the gross primary anal tumour volume. The volume should be
limited to the gross tumour and not include the whole lumen.
CTV_​A = GTV_​A + 10 mm. Following this, manually enlarge to ensure coverage of
entire anal canal including outer border from the ano-​rectal junction (approxi-
mately 4 cm superiorly from anal verge identified by the radio-​opaque marker)
to the anal verge including the internal and external anal sphincters. Edit to ex-
clude bone and muscle. (See Fig. 9.1.)
CTV_​E = Elective nodal regions (see Appendix 1).
PTV_​A = CTV_​A + 10 mm.
PTV_​E = CTV_​E + 5 mm.
214 Squamous cell carcinoma of the anus

ACT 5:
GTV_​A = Includes the gross primary anal tumour volume. The volume should be
limited to the gross tumour and not include the whole lumen.
GTV_​N = Includes all involved nodes ≤3 cm
GTV_​N3 = Includes all involved nodes >3 cm
GTV_​Boost * = GTV_​A & GTV_​N [If node positive ≤3 cm] & GTV_​N3 [node
positive >3 cm]
CTV_​A = GTV_​A + 15 mm. Following this, manually enlarge to ensure coverage of
entire anal canal including outer border from the ano-​rectal junction (approxi-
mately 4 cm superiorly from anal verge identified by the radio-​opaque marker)
to the anal verge including the internal and external anal sphincters (see Fig.
9.1). If no bone or muscle involvement, edit to exclude bone and muscle; if bone
or muscle involvement, only edit structure free from infiltration.
CTV_​N = GTV_​N + 5 mm.
CTV_​N3 = GTV_​N3 + 5 mm.
CTV_​E = Elective nodal regions (see Appendix 1).
PTV_​Boost * = GTV_​Boost + 5 mm
PTV_​A = CTV_​A + 10 mm.
PTV_​N = CTV_​N + 5 mm.
PTV_​N3 = CTV_​N3 + 5 mm
PTV_​E = CTV_​E + 5mm.
Organs at risk
OARs include the following and should be delineated by the radiographer/​dosimetrist/​
physicist/​consultant:
1. Small bowel: Contouring should include all individual small bowel loops to at least
20 mm above the superior extent of both PTVs. It may be helpful to initially delin-
eate the large bowel +/​-​endometrium to exclude these from subsequent delinea-
tion of small bowel.
2. External genitalia: Delineation of the male genitalia should include the penis and
scrotum out laterally to the inguinal creases. In woman it should include the clit-
oris, labia majora and minora, out to the inguinal creases. Superior border in both
sexes should lie midway through the symphysis pubis.
3. Bladder: The entire bladder including outer bladder wall should be delineated.
4. Right and left femoral heads: The femoral heads should be contoured separately
on each side, to include the ball part of the joint, the trochanters, and proximal
shaft to the level of the bottom of ischial tuberosities.
Planning parameters
Prescription point: 100% to the median dose in PTV (ICRU 83)  PTV dose objectives
and realistic OAR dose constraints are set in the treatment planning system and pri-
oritized to produce the most homogeneous dose to the PTV and minimize the dose
to the OARs. PTVs take priority over any OAR constraints. These dose constraints are
 Postoperative adjuvant treatment 215

usually based upon treatment planning studies and toxicity outcome data for the par-
ticular OAR—​although details for some organs are poorly documented.
For IMRT:
Suggested beam positions if supine: 0°; 310°; 275°; 210°; 150°; 85°; 50°
Suggested beam positions if prone: 180°; 130°; 95°; 30°; 330°; 265°; 230°

Treatment delivery
Verification should be practiced according to standards for the unit, but daily
cone-​beam CT imaging is recommended for the first 5 days and subsequently on a
weekly basis.

9.6.6 Chemotherapy
The cytotoxic drugs are given during the first and fifth week of radiotherapy using:
◆ MMC day 1, 12 mg/​m2 bolus day 1 only (max. 20 mg).
◆ 5FU 750mg/​m2 in 1 litre N saline over 24 hours days 1–​5, 29–​33.
◆ An estimated glomerular filtration rate of >50 mL/​
min is required for this
prescription.
For patients who are elderly and when there is concern about a possible increased risk
of neutropenic sepsis dose modifications are used:
◆ MMC day 1, 8 mg/​m2 bolus day 1 only (max. 10 mg).
◆ 5FU 750 mg/​m2 in 1 litre N saline over 24 hours days 1–​4, 29–​32.
◆ Capecitabine has been shown to be effective and is given at a dose of 825 mg/​m2
twice daily on days of radiotherapy(39).

9.7  Postoperative adjuvant treatment

9.7.1 Indications
A few patients will be referred for consideration of chemoradiation after local ex-
cision of a small tumour either in the anal canal or margin. Well-​differentiated
cancers < 2 cm in diameter, involving the anal margin can be treated by local ex-
cision if it is felt that clear margins (>2  mm) can be obtained. For anal margin
cancers, adverse features after local excision are those relevant to squamous skin
cancers (tumour diameter >2 cm, poor differentiation and perineural invasion(67),
which predict a risk of nodal metastases. Treatment after an excision biopsy with
positive margins is often delayed and/​or compromised by an associated infected
wound cavity. Sometimes re-​excision is feasible if margins are close, but it is dif-
ficult for the surgeon to be sure the increased depth of tissue is taken from the
correct place. Piece-​meal resections render assessment of resection margins in
the specimen impossible and should not be performed, because additional post-​
operative chemoradiation is usually required, which in itself will also have an im-
pact on long-​term anorectal function. There is no evidence to support a ‘debulking’
approach prior to chemoradiation.
216 Squamous cell carcinoma of the anus

A few patients may undergo initial abdominoperineal excision as definitive treat-

ment of their anal cancer if initial biopsy is not categorized as squamous or there
are contraindications to chemoradiation. If previous CRT has not been given,
postoperative chemoradiation should be considered when there is evidence of involve-
ment of the circumferential resection margin.
The indications for postoperative adjuvant treatment are therefore:

Absolute indications
◆ Incomplete local excision of squamous carcinoma of the anus (deep or lateral resec-
tion margins ≤ 1 mm).
◆ Involvement of the circumferential resection margin after initial abdominoperineal
excision.

Relative indications
◆ After local excision:
• Narrow margin > 1–​2 mm, or
• Primary tumour with clear margins but primary large ie > 2–​5 cm (T2).
◆ After APER:
• Node + ve circumferential resection margin–​ve defects in the specimen.
These indications are uncommon, and there is uncertainty as to the best approach
with respect to postoperative chemoradiation. It is important that patients are care-
fully examined, as there may be palpable residual disease after local excision. All pa-
tients should undergo staging investigations as described in previous sections.
If initial surgery, comprising either local excision for a ≤ 2  cm tumour or an
abdominoperineal excision with involvement of the circumferential resection margin
has occurred and there is no evidence of nodal or distant spread, it is recommended
that patients are treated with radical chemoradiation using the technique described for
a primary tumour without lymphadenopathy using a two-​phase technique. However
for the phase 2 component of the planning process the site of the initial (excised) tu-
mour should be used as the ‘presumed GTV’ even though this now should only har-
bour microscopic disease at risk. It is reasonable to assume that this area may harbour
a greater microscopic disease burden and require the full dose of radiation.
If a small (< 2 cm) tumour has been treated by local excision and there is a close or
involved margin with no evidence of nodal or distal spread a number of approaches
have been used including brachytherapy alone, a single phase of involved field irradi-
ation, either alone or combined with concurrent chemotherapy (similar to the phase 2
approach described previously for radical chemoradiation) or the two-​phase shrinking
field approach described earlier for radical chemoradiation.
In both of these situations, there is concern about the risks of increased acute and
late toxicity. This results in discussion about the extent of the radiotherapy fields
(whether smaller fields can be treated) and the total dose (whether a lower total
dose may be used). There is insufficient evidence to make clear recommendations at
this time.
 Palliative treatment 217

9.8  Supportive care during radiotherapy

Patients should be assessed at least once weekly with regard to skin toxicity and their
overall tolerance to treatment. It is advisable to check full blood counts at least once
weekly during radiation therapy particularly if MMC is used. It is also important to
maximize patient tolerance with the use of simple antiemetics, antifungals, analgesia,
skin care, and nutritional support. Expected acute side effects include diarrhoea, proc-
titis, urinary frequency and dysuria, loss of pubic hair, and erythema and moist des-
quamation of the skin in the groins and perineum.
Patients should be informed of the negative effect of smoking before chemoradiation
starts. Smoking may worsen acute toxicity during treatment and lead to a poorer out-
come in terms of DFS and CFS(68).
The post-​treatment use of vaginal dilators in sexually active females has been re-
commended, although the evidence for their effectiveness is weak.
Skin effects rapidly disappear within 2–​3 weeks after treatment is completed. Longer-​
lasting side effects are unusual with external beam alone but within the UKCCR Act
I trial were more common after a brachytherapy boost. Long-​term morbidity includes
ulceration of the anal/​rectal area, small bowel obstruction, urethral obstruction, and
fistula formation.

9.9 Palliative treatment
9.9.1  Retreatment with further radiotherapy for recurrent disease
Following initial radiotherapy, tolerance to further radiation may improve with time
elapsed, with a modest long-​term recovery of radiation DNA damage(69). Patients
with local recurrence of SCCA who are not considered suitable for salvage surgery
to achieve a curative (R0) resection may benefit from further radiotherapy, with or
without chemotherapy to the pelvis. Treatment to a small volume, using a dose and
fractionation defined by the cumulative prior doses in the organs at risk and taking
into account the degree of normal tissue recovery expected over time would be
appropriate(70).
Hence, it is important to access details of when and where the primary CRT
treatment was undertaken. Precise details are required of total doses, fraction size,
radiation fields (primary and elective), OTT, and how the treatment was tolerated par-
ticularly if CRT was performed in a different unit. An assessment of current late ra-
diation sequelae (objective—​telangiectasia, fibrosis as well as subjective—​small bowel
symptoms) is also crucial to enable decisions of feasibility and planning of further
CRT regimens.
Stereotactic radiotherapy may be possible in selected cases, with the intention of
delivering a tumoricidal dose where feasible. A long interval from completion of initial
CRT to recurrence (>2 years) predicts for a good response to further RT. Although this
approach is unlikely to be curative, it can offer medium-​term control of local disease
and palliation of symptoms.
A small minority (< 10%) of patients will have synchronous metastases at presenta-
tion. Clinical experience with cytotoxic drugs (5FU, cisplatin, carboplatin, paclitaxel,
218 Squamous cell carcinoma of the anus

irinotecan) or the biologicals (bevacizumab, cetuximab) is limited, with few publica-

tions regarding the efficacy of chemotherapy alone
If the patient is fit for systemic treatment, we recommend the majority should re-
ceive initial chemotherapy—​usually with a combination of cisplatin and infusional
5FU. Responses are rarely complete and usually of short duration. A  current trial
under the aegis of the International Rare Cancers Initiative (IRCI) is comparing the
international standard of 5FU/​cisplatin vs carboplatin/​taxol.
If the volume of metastatic disease is relatively small, then in order to attempt to
achieve loco-​regional control in the pelvis, patients may benefit from a total dose of
30–​50 Gy combined with systemic chemotherapy as described in the earlier section
on radical chemoradiation. Modification of this approach to an involved field single-​
phase of treatment of GTV + 3 cm may be appropriate for some patients.
There remains a group of patients who are commonly elderly and of poor perform-
ance status where local palliative treatment is required. The most appropriate approach
for such patients may range from a single fraction of 8 Gy to 20 Gy in five fractions,
again using an involved field approach particularly to attempt to control bleeding,
pain, or enlargement of a fungating tumour.
However, if significant tumour regression is to be achieved it is the authors’ experi-
ence that modified chemoradiotherapy is more effective than palliative radiotherapy
alone. The use of 30 Gy in 15 fractions combined with 600 mg/​m2 5FU by continuous
infusion days 1–​4 to an involved field (GTV + 3 cm) is associated with major tumour
regression and is feasible in elderly frail patients(71).

9.10  Therapeutic assessment and follow-​up

Patients who do not respond to CRT are usually treated with abdominoperineal re-
section hence it is important to assess patients following treatment. MRI is performed
most commonly 6–​8 weeks following treatment although there is little data of its ac-
curacy or predictive value. Early (6–​8 week) assessment by MRI or other imaging mo-
dalities is generally unhelpful and is not recommended(72). However, recent data has
also indicated that MRI assessment may be able to identify those at risk of early relapse,
who are amenable to R0 salvage(73). One study has suggested MRI appearances such as
tumour size reduction and signal intensity change are predictive of a good outcome.
Surveillance is typically performed by clinical examination with the aid of add-
itional proctoscopy. Given that metastatic relapse is uncommon, the scheduling of CT
scans for metastatic surveillance outside trials remains controversial.

9.11  Current and future research

The results for T1/​T2 anal cancers appear good, but for T3/​T4 tumours remain poor.
A ‘one size fits all’ stages approach is no longer appropriate. Probably early T1 tumours
are over-​treated; more advanced T3/​T4 merit treatment escalation. IMRT abolishes
the gap and shortens OTT and thus looks promising in this setting.
Newer risk assessment techniques (using p16, SCCAg, PET/​CT, TSIL, PD-​1 expres-
sion) may allow us to stratify SCCA more accurately and help to tailor the intensity of
treatment to the individual.
 References 219

The integration of other potential chemotherapy agents including oxaliplatin,

gemcitabine carboplatin, vinorelbine, paclitaxel, etoposide, topotecan, and the bio-
logical agents are all potential future strategies. However, the treatment with the most
exciting potential is to use immune modulators and the check point inhibitors such as
PD-​1 inhibitors, which have shown spectacular early promise.

References
1. Siegel RL, Miller KD, Jemal A, Cancer statistics, 2016 CA: a Cancer Journal for Clinicians
2016; 66:7–​30.
2. Wilkinson JR, Morris EJ, Downing A, et al. The rising incidence of anal cancer in England
1990-​2010: a population-​based study. Colorectal Disease 2014; 16:O234–​39.
3. Hartwig S, St Guily JL, Dominiak-​Felden G, et al. Estimation of the overall burden of
cancers, precancerous lesions, and genital warts attributable to 9-​valent HPV vaccine types
in women and men in Europe. Infectious Agents and Cancer 2017; 12:19.
4. Boman BM, Moertel CG, O’Connell MJ, et al. Carcinoma of the anal canal. A clinical and
pathologic study of 188 cases. Cancer 1984; 54: 14–​25.
5. UKCCCR Anal Cancer Working Party. Epidermoid anal cancer: Results from the
UKCCCR randomised trial of radiotherapy alone versus radiotherapy, 5-​fluorouracil and
Mitomycin. Lancet 1996; 348:1049–​54.
6. Bilimoria KY, Bentrem DJ, Rock CE,et al., Outcomes and prognostic factors for
squamous-​cell carcinoma of the anal canal: analysis of patients from the National Cancer
Data Base. Diseases of the Colon and Rectum 2009; 52:624–​31
7. Northover J, Glynne-​Jones R, Sebag-​Montefiore D, et al. Chemoradiation for the
treatment of epidermoid anal cancer: 13-​year follow-​up of the first randomised UKCCCR
Anal Cancer Trial (ACT I). British Journal of Cancer 2010; 102:1123–​8.
8. Sebag-​Montefiore D, James R, Meadows H. The pattern and timing of disease recurrence
in squamous cancer of the anus: Mature results from the NCRI ACT II trial. Journal of
Clinical Oncology 2012; 30(Suppl abstr 4029).
9. James RD, Glynne-​Jones R, Meadows H, et al. Mitomycin or cisplatin chemoradiation
with or without maintenance chemotherapy for treatment of squamous-​cell carcinoma of
the anus (ACT II): a randomised, phase 3, open-​label, 2x2 factorial trial. Lancet Oncology
2013; 14:516–​24.
10. Bentzen AG, Guren MG, Vonen B, et al. Faecal incontinence after chemoradiotherapy in
anal cancer survivors: long-​term results of a national cohort. Radiotherapy and Oncology
2013 Jul;108(1):55–​60
11. Flam M, John M, Pajak TF, et al. Role of mitomycin in combination with fluorouracil
and radiotherapy, and of salvage chemoradiation in the definitive nonsurgical treatment
of epidermoid carcinoma of the anal canal: results of a phase III randomized intergroup
study. Journal of Clinical Oncology 1996; 14: 2527–​39
12. Bartelink H, Roelofsen F, Eschwege F, et al. Concomitant radiotherapy and chemotherapy
is superior to radiotherapy alone in the treatment of locally advanced anal cancer: results
of a phase III randomized trial of the European Organization for Research and Treatment
of Cancer Radiotherapy and Gastrointestinal Cooperative Groups. Journal of Clinical
Oncology 1997;15:2040–​9
13. Ajani JA, Winter KA, Gunderson LL, et al. Fluorouracil, mitomycin, and radiotherapy
vs fluorouracil, cisplatin, and radiotherapy for carcinoma of the anal canal: a randomized
controlled trial. JAMA 2008; 299:1914–​21
220 Squamous cell carcinoma of the anus

14. Glynne-​Jones R, Nilsson PJ, Aschele C, et al.Anal Cancer: ESMO-​ESSO-​ESTRO clinical

practice guldelines for diagnosis treatment and follow up. Annals of Oncology 2014;
25(Suppl 3):iii10–​20.
15. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines®). NCCN
Guidelines Version 2.2018 Anal Carcinoma https://​www.nccn.org. NCCN.org.
(Last visited on 16 January 2019).
16. Gunderson LL, Winter KA, Ajani JA, et al. Long-​term update of U.S. GI Intergroup RTOG
98-​11 phase III trial for anal carcinoma: Survival, Relapse and Colostomy failure with
concurrent chemoradiation involving Fluorouracil/​mitomycin versus Fluoruracil/​cisplatin.
Journal of Clinical Oncology 2012; 30:4344–​51.
17. Peiffert D, Tournier-​Rangeard L, Gerald JP, et al. Induction chemotherapy and dose
intensification of the radiation boost in locally advanced anal canal carcinoma: final
analysis of the randomized UNICANCER ACCORD 03 trial. Journal of Clinical Oncology
2012; 30:1941–​44.
18. Gunderson LL, Moughan J, Ajani JA, et al. Anal carcinoma: impact of TN category
of disease on survival, disease relapse, and colostomy failure in US Gastrointestinal
Intergroup RTOG 98-​11 phase 3 trial. International Journal of Radiation Oncology, Biology,
Physics 2013 Nov 15;87(4):638–​45
19. Williams GR, Lu QL, Love SB, et al. Properties of HPV-​positive and HPV-​negative anal
carcinomas. Journal of Pathology 1996; 180: 378–​82.
20. De Vuyst H, Clifford GM, Nascimento MC et al. Prevalence and type distribution of
human papillomavirus in carcinoma and intraepithelial neoplasia of the vulva, vagina and
anus: a meta-​analysis. International Journal of Cancer 2009; 124:1626–​36.
21. Serup-​Hansen, E., Linnemann, D., Skovrider-​Ruminski, W., et al. Human papillomavirus
genotyping and p16 expression as prognostic factors for patients with American Joint
Committee on Cancer stages I to III carcinoma of the anal canal. Journal of Clinical
Oncology 2014; 32, 1812–​17.
22. Baseman JG, Koutsky LA. The epidemiology of human papillomavirus infections. Journal
of Clinical Virology 2005; 32(suppl 1):S16–​24
23. Grulich AE, van Leeuwen MT, Falster MO, et al. Incidence of cancers in people with HIV/​
AIDS compared with immunosupressed transplant recipients: a meta-​analysis. Lancet
2007; 370:59–​67.
24. Tseng HF, Morgenstern H, Mack TM, Peters RK. Risk factors for anal cancer: results of a
population-​based case-​-​control study. Cancer Causes and Control 2003; 14: 837–​46.
25. Daling JR, Madeleine MM, Johnson LG, et al. Penile cancer: importance of circumcision,
human papillomavirus and smoking in in situ and invasive disease. International Journal of
Cancer 2005; 116:606–​16.
26. Shepherd NA, Schofield JH, Love SB et al. Prognostic factors in anal squamous
carcinoma: a multi variant analysis of clinical, pathological and flow cytometric perimeters
in 235 cases. Histopathology 1990; 16:545–​55.
27. Deniaud-​Alexandre E, Touboul E, et al. Results of definitive irradiation in a series of 305
epidermoid carcinomas of the anal canal. International Journal of Radiation Oncology,
Biology, Physics 2003; 56(5):1259–​73.
28. Nilsson PJ, Svensson C, Goldman S, et al. Epidermoid anal cancer: a review of a
population-​based series of 308 consecutive patients treated according to prospective
protocols. International Journal of Radiation Oncology, Biology, Physics 2005; 61(1): 
92–​102.
 References 221

29. Sparano JA, Lee JY, Palefsky J et al., Cetuximab plus chemoradiotherapy for human
immunodeficiency virus-​associated anal carcinoma: A Phase II AIDS Malignancy
Consortium Trial (AMC045). Journal of Clinical Oncology 2017; 35:727–​33.
30. Greenall MJ, Quan SH, Urmacher C, DeCosse JJ. Treatment of epidermoid carcinoma of
the anal canal. Surgery, Gynecology and Obstetrics 1985; 161:509–​17.
31. Nigro ND, Vaitkevicus VK, Considine B Jr. Combined therapy for cancer of the anal
canal: a preliminary report. Diseases of the Colon and Rectum 1974; 17:354–​35.
32. Leichman L, Nigro N, Vaitkevicius VK, et al. Cancer of the anal canal: model for
preoperative adjuvant combined modality therapy. American Journal of Medicine 1985;
78(2):211–​21.
33. Papillon J, Montbarbon JF. Epidermoid carcinoma of the anal canal. A series of 276 cases.
Diseases of the Colon and Rectum 1987; 30:324–​33.
34. Newman G, Calverley DC, Acker BD, et al. Management of carcinoma of the anal canal
by external beam radiotherapy, experience in Vancouver 1971–​1988. Radiotherapy and
Oncology 1992; 25:196–​202.
35. Mell LK, Schomas DA, Salama JK, et al.Association between bone marrow dosimetric
parameters and acute hematologic toxicity in anal cancer patients treated with concurrent
chemotherapy and intensity-​modulated radiotherapy. International Journal of Radiation
Oncology, Biology, Physics 2008; 70: 1431–​43.
36. Liang Y, Messe K, Rose BS, et al. Impact of bone marrow radiation dose on acute
hematologic toxicity in cervical cancer: Principal component analysis on high dimensional
data. International Journal of Radiation Oncology, Biology, Physics 2010; 78:912–​9.
37. Mauch P, Constin L, Greenberger J, et al. Hematopoietic stem cell compartment: acute
and late effects of radiation therapy and chemotherapy. International Journal of Radiation
Oncology, Biology, Physics 1995; 31:1319–​39.
38. Kachnic LA, Winter K, Myerson RJ et al. RTOG 0529: a phase 2 evaluation of dose-​
painted intensity modulated radiation therapy in combination with 5-​fluorouracil
and mitomycin-​C for the reduction of acute morbidity in carcinoma of the anal canal.
International Journal of Radiation Oncology, Biology, Physics 2013; 86:27–​33.
39. Glynne-​Jones R, Meadows H, Wan S, et al. EXTRA-​-​a multicenter phase II study of
chemoradiation using a 5 day per week oral regimen of capecitabine and intravenous
mitomycin C in anal cancer. International Journal of Radiation Oncology, Biology, Physics
2008; 72:119–​26.
40. Goodman KA, Julie D, Cercek A, et al. Capecitabine with mitomycin reduces
acute hematologic toxicity and treatment delays in patients undergoing definitive
chemoradiation using intensity modulated radiation therapy for anal cancer. International
Journal of Radiation Oncology, Biology, Physics 2017; 98:1087–​95.
41. Jones CM, Adams R, Downing A, et al. Toxicity, tolerability, and compliance of concurrent
capecitabine or 5-​fluorouracil in radical management of anal cancer with single-​dose
mitomycin-​c and intensity modulated radiation therapy: evaluation of a national cohort.
International Journal of Radiation Oncology, Biology, Physics 2018; 101:1202–​11
42. Glynne-​Jones R, Sebag-​Montefiore D, Adams R, et al. for the UKCCCR Anal Cancer
Trial Working Party. ‘Mind the Gap’–​the impact of variations in the duration of the
treatment gap and overall treatment time in the first UK Anal Cancer Trial (ACT I).
International Journal of Radiation Oncology, Biology, Physics 2011; 81:1488–​94.
43. Glynne-​Jones R, Meadows AH, Lopes A, et al. Compliance to chemoradiation (CRT) using
mitomycin (MMC) or cisplatin (CisP), with or without maintenance 5FU/​CisP chemotherapy
222 Squamous cell carcinoma of the anus

(CT) in squamous cell carcinoma of the anus (SCCA) according to radiotherapy (RT) dose,
overall treatment time (OTT) and chemotherapy (CT) and their impact on long-​term
outcome: Results of ACT II. Journal of Clinical Oncology 2015; 33: (suppl; abstr 3518).
44. Scholefield JH, Nugent KP. Anal cancer. Position statement of the Association of
Coloproctology of Great Britain and Ireland introduction. Colorectal Disease 2011; 13,
S1: 3–​10.
45. Buchli C, Martling A, Arver S, Holm T. Testicular function after radiotherapy for rectal
cancer-​-​a review. Journal of Sexual Medicine 2011; 8:3220–​26.
46. Gerard JP, Chapet O, Samiei F, et al. Management of inguinal lymph node metastases in
patients with carcinoma of the anal canal: experience in a series of 270 patients treated in
Lyon and review of the literature. Cancer 2001; 92:77–​84.
47. Clark J, Petrelli N, Herrera L, Mittelman A. Epidermoid carcinoma of the anal canal.
Cancer 1986; 57: 400–​6.
48. Pyper PC, Parks TG. The results of surgery for epidermoid carcinoma of the anus. British
Journal of Surgery 1985; 72: 712–​14.
49. Stearns MW, Urmacher C, Sternberg SS. Cancer of the anal canal. Current Problems in
Cancer 1980; 4:1–​44.
50. Tinger A, Michalski JM, Bosch WR, et al. An analysis of intratreatment and intertreatment
displacements in pelvic radiotherapy using electronic portal imaging. International Journal
of Radiation Oncology, Biology, Physics 1996; 34:683–​90.
51. Lengelé B, Scalliet P. Anatomical bases for the radiological delineation of lymph node
areas. Part III: Pelvis and lower limbs. Radiotherapy and Oncology 2009; 92:22–​33.
52. Gay HA, Barthold HJ, O’Meara E, et al. Pelvic normal tissue contouring guidelines
for radiation therapy: a Radiation Therapy Oncology Group consensus panel atlas.
International Journal of Radiation Oncology, Biology, Physics 2012; 83:e353–​62.
53. Roach SC, Hulse PA, Moulding FJ, et al. Magnetic resonance imaging of anal cancer.
Clinical Radiology 2005; 60:1111–​19.
54. Myerson RJ, Garofalo MC, El Naqa I, et al. Elective clinical target volumes for conformal
therapy in anorectal cancer: a radiation therapy oncology group consensus panel
contouring atlas. International Journal of Radiation Oncology, Biology, Physics 2009;
74:824–​30.
55. Ng M, Leong T, Chander S, et al. Australasian Gastrointestinal Trials Group (AGITG)
contouring atlas and planning guidelines for intensity-​modulated radiotherapy in anal
cancer. International Journal of Radiation Oncology, Biology, Physics 2012; 83:1455–​62.
56. Muirhead R, Adams RA, Gilbert DC, et al. Anal cancer: developing an intensity
modulated radiotherapy solution for ACT2 fractionation. Clinical Oncology (Royal College
of Radiology) 2014; 26:720–​1.
57. Brooks C, Hansen VN, Riddell A, et al. Proposed genitalia contouring guidelines in anal
cancer intensity-​modulated radiotherapy. British Journal of Radiology 2015; 88:20150032.
58. Mell LK, Schomas DA, Salama JK, et al. Association between bone marrow dosimetric
parameters and acute haematologic toxicity in anal cancer patients treated with concurrent
chemotherapy and intensity-​modulated radiotherapy. International Journal of Radiation
Oncology, Biology, Physics 2008; 71:1431–​7.
59. Menkarios C, Azria D, Laliberte B, et al. Optimal organ-​sparing intensity-​modulated
radiation therapy (IMRT) regimen for the treatment of locally advanced anal canal
carcinoma: a comparison of conventional and IMRT plans. Radation Oncology
2007; 2:41.
 References 223

60. Milano MT, Jani AB, Farrey KJ, et al. Intensity-​modulated radiation therapy (IMRT)
in the treatment of anal cancer: Toxicity and clinical outcome. International Journal of
Radiation Oncology, Biology, Physics 2005; 63:354–​61.
61. Hodges JC, Das P, Eng C, et al. Intensity-​modulated radiation therapy for the treatment
of squamous cell anal cancer with para-​aortic nodal involvement. International Journal of
Radiation Oncology, Biology, Physics 2009; 75:791–​4
62. Koerber SA, Slynko A, Haefner MF, Krug D, Schoneweg C, Kessel K, et al. Efficacy and
toxicity of chemoradiation in patients with anal cancer—​a retrospective analysis. Radiation
Oncology 2014; 9:113.
63. Lin A, Ben-​Josef E. Intensity-​modulated radiation therapy for the treatment of anal cancer.
Clinical Colorectal Cancer 2007; 6:716–​19.
64. Chuong MD, Freilich JM, Hoffe SE, et al. Intensity-​modulated radiation therapy vs. 3D
conformal radiation therapy for squamous cell carcinoma of the anal canal. Gastrointestal
Cancer Research 2013; 6:39–​45.
65. Dasgupta T, Rothenstein D, Chou JF, et al. Intensity-​modulated radiotherapy vs.
conventional radiotherapy in the treatment of anal squamous cell carcinoma: a propensity
score analysis. Radiotherapy and Oncology 2013; 107:189–​90.
66. Amini A, Jones BL, Ghosh D, et al. Impact of facility volume on outcomes in patients with
squamous cell carcinoma of the anal canal: Analysis of the National Cancer Data Base.
Cancer 2017; 123:228–​36.
67. Karia PS, Jambusaria-​Pahlajani A, Harrington DP, et al. Evaluation of American Joint
Committee on Cancer, International Union Against Cancer, and Brigham and Women’s
Hospital tumor staging for cutaneous squamous cell carcinoma. Journal of Clinical
Oncology. 2014; 32:327–​34.
68. Mai SK, Welzel G, Haegele V, Wenz F. The influence of smoking and other risk factors on
the outcome after chemoradiotherapy for anal cancer. Radation Oncology 2007; 2:30.
69. Stewart FA, van der Kogel AJ. Retreatment tolerance of normal tissue. Seminars in
Radation Oncology 1994; 4:103–​11.
70. Guren MG, Undseth C, Rekstad BL, et al. Reirradiation of locally recurrent rectal cancer: a
systematic review. Radiotherapy and Oncology 2014; 113:151–​57.
71. Charnley N, Choudhury A, Chesser P, et al. Effective treatment of anal cancer in the
elderly with low-​dose chemoradiotherapy. British Journal of Cancer 2005; 92: 1221–​5.
72. Goh V, Gollub FK, Liaw J, et al. Magnetic resonance imaging assessment of squamous cell
carcinoma of the anal canal before and after chemoradiation: an MRI predict for eventual
clinical outcome? International Journal of Radiation Oncology, Biology, Physics 2010;
78:715–​21.
73. Kochhar R, Renehan AG, Mullan D, et al. The assessment of local response using
magnetic resonance imaging at 3-​and 6-​month post chemoradiotherapy in patients with
anal cancer. European Journal of Radiology 2017; 27:607–​17.
 Chapter 10

Prostate cancer
Linus Benjamin, Alison Tree,
and David Dearnaley

10.1 Introduction
Prostate cancer is the most prevalent cancer in men in the United Kingdom with an
estimated lifetime risk of one in eight, and the second most common cause of cancer-​
related death in men. It accounts for 13% of all new diagnoses of cancer in men in UK,
and 13% of all male cancer deaths. The incidence of prostate cancer has increased by
5% in the last decade and more than 155% in the last 40 years. This is principally due
to more prostate-​specific antigen (PSA) testing. Prostate cancer mortality rates have
reduced more rapidly in the USA than UK, presumably from the earlier uptake and
higher prevalence of opportunistic PSA screening.
Histopathological data from postmortem studies demonstrate that prostate cancer
is found in approximately half of all men in their 50s and in 80% of men by age 80 but
only one in 26 men (3.8%) will die from this disease(1). The challenge is to identify
those men who will benefit from radical treatment and in whom it is therefore justifi-
able to risk the side effects associated with such treatment.
Patients with NCCN low-​risk prostate cancer, whose disease has a long natural
history and who have a good life expectancy, may be suitable for active surveillance.
This entails the close monitoring of the patient’s prostate cancer with regular clin-
ical review, digital rectal examinations, PSA monitoring, and prostate magnetic res-
onance imaging (MRI) imaging or repeat prostate biopsy. The decision to proceed
from active surveillance to radical treatment is taken when the survival and quality
of life benefits from active treatment are deemed to be greater than the competing
risks and side effects from treatment. In contrast to active surveillance, watchful
waiting may be suitable in patients who have a poor life-​expectancy owing to co-​
morbidities. In such patients, active treatment, delivered with palliative intent, is
considered in the setting of symptomatic disease progression. PSA monitoring may
still be undertaken in these patients, though less frequently than in those under ac-
tive surveillance.
Radical treatment options for localized disease include prostatectomy, external beam
radiotherapy +/​-​hormonal therapy, and interstitial brachytherapy. Contemporary
series suggest that outcomes of each of these treatment modalities are similar(2,3). There
is a relative paucity of randomized comparisons and long-​term follow-​up data after
 Introduction 225

Table 10.1  NCCN risk stratification for men with localised prostate cancer

PSA Gleason score Clinical stage
Low risk < 10 ng/​mL and ≤6 and T1–​T2a

Intermediate risk 10–​20 ng/​mL or 7 or T2b–​T2c

High risk > 20 ng/​mL or 8–​10 or T3–​T4

brachytherapy. In recommending whether and how to treat a man with localized pros-
tate cancer, one needs to consider the risk grouping (see Table 10.1), the life expect-
ancy of the patient, any comorbidities, and his preference for and between treatment
options, taking into account their expected side effects.
Radiotherapy is the most commonly used curative treatment modality for localized
prostate cancer in the UK and about 15,000 men are treated annually, with intensity-​
modulated radiotherapy (IMRT) now becoming the standard of care. Radical treat-
ment improves prostate cancer-​specific survival. A randomized Scandinavian trial has
shown that, in patients < 65 years, radical prostatectomy for localized prostate cancer
reduces overall mortality, prostate cancer mortality, and risk of metastases compared
to watchful waiting(4). In patients ≥ 65 years, radical prostatectomy leads to a signifi-
cant reduction in the development of distant metastases.
The PIVOT study randomized 731 patients with screen-​detected cancer to radical
prostatectomy or observation. Prostatectomy did not improve overall survival or PCa-​
specific survival. However, in patients with a PSA >10 ng/​ml, prostatectomy was asso-
ciated with an improvement in all-​cause mortality (p = 0.04)(5).
The landmark SPCG-​7/​SFUO-​3 study (6) randomized patients with locally advanced
prostate cancer to 3  months of total androgen blockade followed by continuous
flutamide +/​-​70 Gy of radiotherapy to the prostate and seminal vesicles. The addition
of radiotherapy to androgen blockade improved both prostate cancer-​specific survival
and median overall survival.
Similarly, the NCIC PR.3/​MRC PR07/​Intergroup study(7) randomized patients with
high-​risk, locally advanced prostate cancer, to lifelong androgen blockade with or
without 65–​69 Gy of radiotherapy to the prostate +/​-​pelvis. After a median follow-​up
of 8 years, and despite the lower dose of radiotherapy in comparison to the current
standard of care, there was a significant improvement in overall survival in patients re-
ceiving radiotherapy in addition to androgen blockade, with no significant difference
in patient reported quality of life outcomes at 3 years.
Mottet et al. randomized patients with locally-​advanced, T3-​T4 disease, to 3 years of
hormonal therapy with or without prostate (74–​78 Gy) and pelvis radiotherapy. There
was significant improvement in the combined therapy arm for 5-​year progression-​free
survival, local-​regional control, and metastasis-​free survival(8).
Non-​randomized data from the MRC STAMPEDE trial shows significant advan-
tage in failure-​free survival for the addition of radiotherapy to hormonal therapy in
patients with locally advanced disease, with or without lymph node involvement(9).
226 Prostate cancer

10.2 Indications
External beam radiotherapy is employed in six settings for prostate cancer which will
be discussed in turn:
1. Radical radiotherapy to prostate ± seminal vesicles.
2. Radical radiotherapy to prostate and pelvic lymph nodes.
3. Radical radiotherapy following prostatectomy, adjuvantly or as a salvage therapy
following biochemical failure.
4. Palliative radiotherapy to prostate ± pelvis
5. Palliative radiotherapy to distant metastases.
6. Breast bud radiotherapy.

10.2.1  Radical radiotherapy to prostate ± seminal vesicles

Radical radiotherapy to the prostate is the most commonly used curative treatment in
the UK. The proximal 1 cm of the seminal vesicles is treated as standard in conjunc-
tion with the prostate, unless the patient has low risk disease, when no seminal vesicle
is treated. For patients with high-​risk disease, the distal seminal vesicles will also be
treated. The risk for microscopic disease involvement of the seminal vesicles can be
estimated from the Roach formula(10):

seminal vesicle risk (%) = PSA +[(Gleason score − 6) ×10]

Technological advances over recent years, with improvements in image guidance

for treatment delivery and improved dose distribution with intensity modulated
treatment, have increased the precision of external beam radiotherapy and have led
to improved outcomes. The use of IMRT reduces the dose-​limiting side effects and
has allowed for dose escalation to the whole prostate, which has improved biochem-
ical control. In comparison to 3D-​CRT, the greater conformality of dose distribution
has also resulted in a reduction in toxicity, particularly gastro-​intestinal (GI) toxicity.
IMRT has now become the standard of care for treating localized prostate cancer.
Radiotherapy is usually given in combination with hormonal therapy. The duration
of hormonal therapy depends on the disease risk stratification. External beam radio-
therapy may be given with a short course of neo-​adjuvant androgen suppression for
intermediate risk patients or combined with longer-​term hormonal therapy for pa-
tients with advanced or high risk disease, particularly if the Gleason score is ≥ 8. See
section 10.10.

10.2.2  Radical radiotherapy prostate plus pelvis

Radiotherapy to prostate and pelvis—​node negative disease
Elective irradiation of pelvic lymph nodes in the context of clinically localized inter-
mediate or high-​risk disease remains controversial. Pelvic radiotherapy has been used
as standard by many North American and European centres for men with locally ad-
vanced disease and has been included in phase III trials for these groups of men(11–​14).
 Indications 227

In the UK, pelvic radiotherapy has been infrequently used due to concerns over bowel
toxicity.
While several retrospective studies support pelvic nodal irradiation, phase III trials
of pelvic nodal irradiation have, to date, been negative, but these studies have meth-
odological flaws. Further studies are ongoing (including the UK PIVOTALboost
study), but it remains to be proven that treatment of sub-​clinical regional disease im-
proves outcomes as in breast cancer.
RTOG 77-​06(11,12) was the first prospective study evaluating prostate +/​-​pelvic radio-
therapy. The trial used modest radiotherapy doses, included mostly lower-​risk patients
and showed no benefit from pelvic radiotherapy.
Two further phase III trials evaluating the role of whole pelvic radiotherapy in pa-
tients with intermediate-​and high-​risk prostate cancer have been published.
RTOG trial 94–​13 included 1323 patients with estimated lymph node risk ≥ 15%.
Patients were randomized between prostate only (dose received 70.2 Gy in 1.8-​Gy
fractions) and whole pelvis radiotherapy (dose to whole pelvis 50.4 Gy in 1.8-​Gy
fractions), and in the second randomization between neo-​adjuvant/​concurrent
and adjuvant hormonal therapy. Updated results show no statistically significant
benefit in biochemical control with pelvic radiotherapy compared with prostate
only radiotherapy.
However, a subset analysis of the RTOG 94-​13 neoadjuvant hormonal therapy
arm has demonstrated a statistically significant difference in disease-​free survival at
7 years, for variation in treatment field size, suggesting a relationship between field size
and disease-​free survival.
GETUG-​01, a smaller French phase III trial of 444 patients, with T1b-​T3N0 dis-
ease, failed to show any difference between whole pelvic (46 Gy in 2-​Gy fractions)
and prostate-​only radiotherapy (66–​70 Gy in 2-​Gy fractions, median dose to the
prostate was 68 Gy): the GETUG group used a lower radiotherapy dose to the pros-
tate than RTOG, with a significant cohort being treated to 66 Gy. They used a lower
superior border of the pelvic field than in the RTOG trial and > 50% of patients had
< 15% risk of lymph node involvement. Furthermore, not all patients received hor-
monal therapy.
These factors may have contributed to the lack of an observed effect. As a result
the treatment volume (i.e. prostate or prostate and pelvic lymph nodal regions) in
high-​risk patients remains an unresolved question and further trials are currently
underway(15,16).
Pelvic radiotherapy may be offered to patients with a high predicted risk of micro-
scopic pelvic lymph node involvement (≥ 15–​30% as defined using the equation de-
vised by Roach formula), but low risk of subclinical systemic disease. Extended pelvic
nodal dissection studies have raised the possibility of Roach formula underestimating
the extent of regional nodal involvement(10). Newer nomograms based on extended
lymph node dissection studies are now available for estimating probability of regional
nodal involvement(17). Additionally, high-​risk patients (e.g. Gleason 8–​10, clinical T3–​
T4 tumours, or lymph node risk > 30%), should be offered 2–​3 years of neo-​adjuvant
and adjuvant hormonal therapy(18).
228 Prostate cancer

Radiotherapy to prostate and pelvis—​node positive disease

Increasing burden of regional nodal involvement correlates with poorer outcomes in
terms of overall survival and progression-​free survival. Management options for node-​
positive non-​metastatic disease include observation alone, hormonal therapy alone,
or radiotherapy with 2–​3 years of neoadjuvant/​adjuvant hormonal therapy +/​-​early
docetaxel chemotherapy. The MRC STAMPEDE trial has shown that early docetaxel
chemotherapy confers a disease control advantage in locally-​advanced, high-​grade
cancers, but as yet no overall survival advantage for this group of non-​metastatic pa-
tients has been demonstrated(19).
Prospective randomized trials for high-​risk disease (RTOG 85-​01, RTOG 86-​10,
RTOG 92-​02, EORTC 22961), which included patients with node-​positive disease,
provide the evidence for radiotherapy with hormonal therapy in locally advanced
node-​positive disease. In these studies, treatment of the prostate to a dose of 65–​70 Gy,
the pelvic lymph nodes to a dose of 45–​50 Gy, in conjunction with hormonal therapy,
resulted in good long-​term outcomes.
EORTC 22961(18) has demonstrated the superiority of long-​course hormonal therapy
over short-​course treatment in locally advanced disease. EORTC 30891 evaluated
early versus deferred hormonal therapy alone in patients with non-​metastatic prostate
cancer, and included node-​positive disease. After a median follow-​up of 7.8 years, a
small statistically significant overall survival difference (42% vs 48%) was reported in
favour of immediate hormonal therapy(20).
SEER data reviewing outcomes in patients with node-​positive disease treated with
or without radiotherapy, demonstrates a 10-​year cancer-​specific survival benefit with
radiotherapy (62.5% vs 50.3%; p  =  0.01)(21). The STAMPEDE trial, where initially
radiotherapy was not mandatory for locally advanced or node positive disease sug-
gests a large disease-​free survival benefit at 2 years (89% vs 64%)(19).

10.2.3  Post-​prostatectomy radiotherapy
The number of radical prostatectomies being performed in the UK is increasing, with
laparoscopic techniques including minimally invasive and robot-​assisted procedures
increasing in popularity over recent years. However, there is still considerable uncer-
tainty over the optimal postoperative oncological management. Biochemical recur-
rence is seen in 15–​30% of patients undergoing radical prostatectomy, with recurrence
rates over 50% in patients with unfavourable pathological features(22).
In the setting of an undetectable post-​operative PSA, with pathological features
indicating a high risk of biochemical failure, either immediate adjuvant radiotherapy
or observation followed by salvage radiotherapy may be offered. Retrospective reviews
suggest that results are better for patients with pre-​radiotherapy PSA < 0.5ng/​mL, PSA
doubling time ≥ 9 months, and positive surgical margins(23).
Salvage prostate bed irradiation can be offered to patients post prostatectomy with
biochemical failure, defined as either two consecutive rises in PSA and final PSA > 0.1
ng/​mL, three consecutive rises in PSA, or a PSA > 0.2 ng/​ml.
Adjuvant radiotherapy can be considered in patients with positive surgical margins,
and/​or pT3/​4 disease due to the associated high risk of residual local disease, but the
 Indications 229

evidence supporting the benefit for adjuvant radiotherapy over salvage radiotherapy
is lacking.

Adjuvant radiotherapy
There are three published randomized controlled trials evaluating adjuvant radio-
therapy. The EORTC 22911 trial(24), after a median follow-​up of 10  years, showed a
statistically significant advantage which was maintained with respect to biochemical
progression-​free survival (61% vs 39%; p < 0.001), with a hazard ratio of 0.49 (p <
0.0001). The 10-​year cumulative late-​toxicity incidence was significantly higher in the
adjuvant radiotherapy group (70.8% vs 59.7%; p = 0.001). Evaluating patients with pT2
margin-​positive disease, showed a 5-​year biochemical-​disease free survival benefit in
favour of adjuvant radiotherapy (76.4% vs 52.2%)(25).
The second trial, SWOG 8794 (NCIC CTG PR-​2)(26) had a similar design and has
longer follow-​up. It randomized 425 patients with T2-​T3 disease, with extracapsular
extension, seminal vesicle involvement or a positive surgical margin, to 60–​64 Gy of
adjuvant radiotherapy, using a 2D radiotherapy technique. At the time of randomiza-
tion, 33% of patients had a PSA > 0.2 ng/​ml, with 8% of patients receiving hormonal
therapy prior to their radical prostatectomy.
After a median follow-​up of 12.5  years, adjuvant radiotherapy was associated
with a statistically significant increase in biochemical control and importantly both
metastasis-​free and overall survival were improved with hazard ratio 0.71 and 0.72,
respectively (p = 0.02). Toxicity was more common in the adjuvant group, with stat-
istically significant increases in proctitis, urethral stricture, and urinary incontinence.
Thirdly, the smaller ARO 96–​02 trial(27) randomized men with pT3 disease, irrespective
of surgical margin status, with a postoperative undetectable PSA, to either observation
or adjuvant radiotherapy (60 Gy), with adjuvant radiotherapy being delivered using
a 3D-​CRT technique. Eleven percent of patients received hormonal therapy prior to
their prostatectomy. Again adjuvant radiotherapy was associated with improved bio-
chemical control. At 10 years, the biochemical failure-​free survival was 56% and 35%
in the adjuvant radiotherapy and observation arms respectively (p < 0.0001). A meta-​
analysis of the three randomized studies reported, improvements in biochemical pro-
gression free survival at 5 and 10 years, and a metastasis-​free/​overall survival benefit
at 10 years(28).
There is no prospective data on the role of adjuvant radiotherapy in node-​positive
disease.

Salvage radiotherapy
Biochemical failure following prostatectomy can be defined as a PSA > 0.2ng/​ml(23),
with post-​operative PSA values > 0.4 ng/​ml being associated with a high-​probability
of systemic disease. Several factors have been identified, which predict for sustained
biochemical disease control following salvage radiotherapy. These include a slow PSA
rise (doubling time > 12 months or velocity < 0.75 ng/​ml), positive surgical margin,
negative lymph nodes and favourable Gleason score (< 8). The relapse free survival
following salvage radiotherapy initiated at a PSA of < 0.2 ng/​ml has been estimated to
be 64%, with 2.6% reduction in relapse free survival for every 0.1 ng/​ml incremental
230 Prostate cancer

rise in PSA at the time of salvage radiotherapy(29). Therefore early salvage radiotherapy
is preferred.

Adjuvant radiotherapy vs salvage radiotherapy

It remains unclear whether adjuvant radiotherapy is superior to observation followed
by salvage radiotherapy, in patients with an undetectable PSA and high-​risk features
on surgical pathology.
Ongoing randomized trials are addressing these questions related to adjuvant and
salvage radiotherapy. The threshold PSA for salvage radiotherapy in RAVES and
RADICALS is 0.10–​0.20 ng/​ml(30).

10.2.4  Palliative radiotherapy to prostate ± pelvis

Patients with locally advanced disease, with or without evidence of distant spread,
may be offered radiotherapy to the prostate and pelvic nodes if present, as a means of
palliation.
Treating the prostate to 20 Gy in five daily fractions results in a symptom response
rate of 80–​90% with complete response rate of 50–​60%. Durability of symptom re-
sponse may be improved by escalating the radiation dose. An alternative fractionation
schedule commonly used includes 30–​36Gy in 5–​6 Gy weekly fractions, over 6 weeks.

10.2.5  Palliative radiotherapy to metastases

External beam radiotherapy is commonly used for palliation of bone pain from meta-
static disease. A single fraction of 8 Gy is appropriate for symptom relief(31). This may
be repeated safely if required. A higher, fractionated dose, e.g. 20 Gy in five fractions,
is traditionally used for overt or subclinical spinal cord compression, or if there is con-
cern over risk of pathological fracture. Symptomatic nodal disease and rarely visceral
metastases may also benefit from a short course of palliative radiotherapy.
Widespread symptomatic osseous metastases can be treated with hemibody radio-
therapy, with 6 Gy to the upper hemibody (extending from mastoid to iliac crest, and
8Gy to the lower hemibody (from iliac crest to ankles)(32). The role of hemibody radi-
ation has declined with the availability of better systemic therapy agents.

10.2.6  Breast
bud radiotherapy for the prevention or
treatment of gynaecomastia
Patients treated with long-​term oestrogen or anti-​androgen monotherapy may de-
velop gynaecomastia which can be painful and distressing.Radiotherapy to the breast
buds can reduce the incidence of gynaecomastia if used prophylactically(33). Target
volume is the glandular tissue of the breast, with a typical field size being an 8-​cm
diameter circle centred on the nipple. A single fraction of 8 Gy is used for prophylactic
treatment and 12 Gy in two fractions has been used for established gynaecomastia.
Applied doses are prescribed in the case of orthovoltage treatment and to the 90%
isodose if electrons are used. For orthovoltage treatment 160-​kv photons are employed
if breast thickness is < 2.5 cm, above which 300 kv is used. Electron energy is typically
6 or 9 MeV.
 Radical radiotherapy planning 231

10.3  Radical radiotherapy planning

10.3.1  Patient position and immobilization
Patients are treated in the supine position with arms across the chest. Prostate
movement in the antero-​posterior direction in the supine position is less than in the
prone position. A variety of immobilization devices are available but available evi-
dence, suggests that ankle stocks combined with a foam head-​pad and knee supports
provide a high degree of accuracy in patient positioning, making pelvic immobil-
ization unnecessary(34,35). Immobilization that incorporates an abdominal compres-
sion component, has not been shown to improve inter-​or intra-​fraction prostate
movement(36).
Principal causes for prostate displacement are a very full bladder or a distended
rectum. Many centres attempt to minimize prostate movement by standardizing
bladder and rectal filling. The bladder should be comfortably full prior to scanning.
This also serves to reduce the proportion within the planning target volune (PTV).
In a typical preparation protocol, patients are encouraged to drink approximately 350
ml during the hour prior to scanning. This protocol is modified as needed for patients
with urinary symptoms. The rectum should ideally be empty of both faeces and flatus.
Patients with a distended rectum at planning have been reported to have increased bio-
chemical and local failure after radiotherapy(37,38). In an attempt to achieve a consistent
degree of rectal filling during planning and treatment, some advocate the routine use
of laxatives or enemas, in addition to dietary modifications prior to each radiotherapy
fraction. Studies have failed to show a reduction in inter-​or intra-​fraction prostate
movement with diet or enemas/​laxatives(39–​41); however, microlette enemas may reduce
the need to rescan patients.
Others have investigated the use of endorectal balloons or endorectal obturators
during planning and treatment. Such devices may reduce prostatic movement in
the anterior to posterior direction(42), but issues remain over patient acceptability.
Intraprostatic radio-​opaque markers placed in the prostate before treatment is an ef-
fective method of visualizing the prostate position during treatment and to enable
online positional correction.
In order to reduce the dose to the rectum, several groups have investigated the
use of spacers (e.g. hyaluronic acid, Space OAR, or biodegradable balloon) between
Denonvilliers fascia and the anterior rectal wall. These increase the separation of the
space between the rectum and prostate. Early reports suggest a significant improve-
ment in rectal dose and side effects(43); however, quality-​of-​life assessments suggest
that bowel bother scores do not improve with the use of spacers(44).

10.3.2  Computed tomography (CT) scanning

The patient is positioned on the CT scanner in the treatment position. The skin surface
is marked at points anteriorly in the midline of the pubic symphysis and laterally over
each hip. Radio-​opaque reference markers are placed over the skin marks.
Scout views are performed followed by a short axial scan starting superiorly at the
level of the pubic symphysis. The axial images are reviewed prior to proceeding to
the full helical scan. The rectal diameter is measured at the level of the prostate. If the
232 Prostate cancer

rectal diameter is > 4 cm in the anterior–​posterior plane the scan is usually aborted.
The patient is assessed and re-​scanned after implementing appropriate intervention to
aid rectal voiding (e.g. microenemas). If the rectal diameter is < 4 cm, the full helical
scan is acquired. The planning CT scan is generated using a slice interval of 5  mm
or less (ideally 2.5  mm or less) from the L3/​4 interspace to 2  cm below the ischial
tuberosities.
For those having pelvic radiotherapy, scans are taken from the bottom of the L1/​L2
vertebral space. Inclusion of the whole bladder and rectum is required to ensure that
the dose-​volume histogram (DVH) dose constraints can be calculated. Following the
CT scan, the skin marks will be made with permanent tattoos. The CT data is then
transferred to a radiotherapy planning computer for outlining.
A multiparametric MRI of the prostate is performed in some centres. MRI images
are co-​registered with planning CT images to facilitate definition of target volumes.

10.4  Target volume definition

10.4.1  Radical radiotherapy: prostate ± seminal vesicles
Standard nomenclature for target volume definition does not strictly apply to con-
formal prostate radiotherapy. The gross tumour volume is often difficult to define
precisely by clinical examination and conventional imaging. Current practice is to
define the entire prostate gland and all or part of the seminal vesicles, using the diag-
nostic MRI scan if available. MRI provides better soft tissue contrast compared to CT,
making it particularly useful in identifying the prostatic apex and distinguishing be-
tween bladder base and anterior rectal wall(45). On T2-​weighted MRI images tumour
within the prostate gland can be seen as a focus of low signal intensity. Several studies
have shown that the prostate, as defined on CT, is often overestimated in size(46,47).
MRI for radiotherapy planning is complicated by the geometric distortion, as well as
lack of direct electron density data arising directly from MR. Methods of creating syn-
thetic CT from MR are now available. MR diffusion and dynamic contrast enhanced
MRI (dMRI)(48) are utilized to clarify intraprostatic lesions, differentiating malignant
disease from potential benign prostatic tissue and making it possible to deliver a radio-
therapy boost to the lesion.
The prostate is first outlined on the central slice where the gland is clearly demar-
cated, and then on successive caudal slices (see Fig. 10.1b). Differentiating the prostate
becomes more difficult near the apex on CT. Identification of the penile bulb using
both planning CT and MR images is helpful. The cranial extent of the bulb lies below
the pelvic floor which is about 1 cm thick and defines the caudal extent of the apex in
T1-​T3 cancer. The prostate is then outlined on slices cranial to the central slice. The
pelvic sling muscles should be excluded from the prostate outline, and often a fat plane
exists between these structures and the prostate. Care is needed to define the prostate
in the more cranial slices because the prostatic base may bulge into the bladder. Again,
reference to the MR is valuable.
The base of the seminal vesicles is outlined for all patients, except true low-​risk pa-
tients. The extent of outlining of the remainder of the seminal vesicles depends on the
 Target volume definition 233

(a)
(b)

(c)

(a)

(b)

(c)

Fig. 10.1  Coronal MR image of the prostate and corresponding axial CT images at level
of: (a) the seminal vesicles, (b) the mid prostate, (c) the penile bulb.
234 Prostate cancer

patient’s individual risk of involvement and anatomy. If the seminal vesicles extend
predominantly laterally, then they can be included in their entirety without signifi-
cantly affecting rectal dose. If there is significant posterior extension of the seminal
vesicles and they are closely applied to the rectal wall, then rectal dose constraints may
be exceeded unless the tips of the seminal vesicles are excluded from the volume.
Pathological data(49) has suggested three patterns of seminal vesicle involvement: (1)
tumour spread along ejaculatory ducts (35% of cases); (2) direct extension through
capsule (61% of cases); and (3)  the presence of isolated tumour deposits (12% of
cases). The entire seminal vesicle is included for stage T3 tumours, or if the risk of
seminal vesicle involvement is > 15%, provided that the predicted dose to the rectum
and sigmoid colon is acceptable. If the bowel dose is unacceptable then the tips of the
seminal vesicles are excluded, but the proximal 2 cm is treated. If the predicted risk of
microscopic seminal vesicle involvement is < 15%, the base of the seminal vesicles only
(proximal 1–​2 cm) is outlined.
The target volume is expanded with an adequate margin to allow for microscopic
spread. In a surgical series of patients with clinically localized prostate cancer(50), the
median distance of extra capsular spread measured radially from the capsule was
2 mm (range 0.5–​12 mm). The PTV also includes an additional margin to allow for
patient and prostate movement and variations in treatment set-​up. In practice the
prostate and some or all of the seminal vesicles are outlined. Margins are then grown
to form the PTV. In a small randomized trial(51) comparing a margin of 1.0 cm with
1.5 cm, there was no difference in tumour control but an increase in rectal and bladder
side effects with the larger margin.
Without image-​guided radiotherapy (IGRT) margins will usually be 1.0 cm but may
be non-​uniform, with tighter margins (5 mm) posteriorly to spare the posterior wall
of the rectum(52). Margins can be reduced (e.g. 6 mm/​3 mm posteriorly) if gold seed
image guidance is used.
Table 10.2 shows the treatment protocols for inclusion of the seminal vesicles in
recent and ongoing major clinical trials. Note that two or three phases of treatment
are used in some studies, with reduced margins for higher dose levels. Normal tissue
structures are also contoured in order to generate DVHs to evaluate dose to organs
at risk (OARs) (rectum, bladder, femoral heads, bowel, and urethral bulb). They are
outlined as solid organs by defining the outer wall. The bladder is outlined from the
base to dome. The rectum is outlined from the anus (or 1 cm below the lower margin
of the PTV whichever is more inferior) to the rectosigmoid junction; frequently the
position is best appreciated on the sagittal CT reconstruction as the level at which
the rectum curves anteriorly. Additional bowel within the PTV is defined separately.
Femoral heads are usually contoured as a globe representing the femoral head.
It is very useful to review the target and normal tissue contours in the sagittal and
coronal planes. Inconsistencies and irregularities should be identified and corrected.

10.4.2  Radical radiotherapy: pelvis and prostate

Normal-​sized lymph nodes are not readily identifiable on planning CT scans. Vascular
and bony anatomy is therefore used to define the lymph node CTV. The pelvic nodal
position should be determined with reference to the pelvic vasculature(53,54.
Table 10.2  Summary of contemporary radiotherapy trials in prostate cancer showing details of radiotherapy technique (SIB = simultaneous integrated boost)
Trial Seminal vesicle Target + margin (mm)** Dose*** (Gy)
involvement Phase I Phase II Phase III Ph I Ph II Ph III Total
risk* (%)
Dutch, Peeters et al.(100) Low (< 10) P + 10 P + 5/​0 –​ 68 0,10 –​ 68, 78
Mod (10–​25) P + SV + 10 P + 10 P + 5/​0 50 18 0,10 68, 78
High (> 25) P + SV + 10 P + 5/​0 –​ 68 0,10 –​ 68, 78
involved P + SV + 10 P + SV + 5/​0 –​ 68 0,10 –​ 68, 78

Protect Low (< 15) P + bSV + 10/​5 P + 0 –​ 56 18 –​ 74

High (≥ 15) P + SV P+0 –​ 56 18 –​ 74
RT 01 Low (<15) P + bSV + 10 P + 0 –​ 64 0,10 –​ 64, 74
High (≥ 15) P + SV P+0 –​ 64 0,10 –​ 64, 74
EORTC 22991 All P + SV + 10 P + bSV + 10 P + 5/​0 46 24 0, 4, 8 70, 74, 78 (not
± nodes randomised)
RTOG P–​0126 All P + SV + 10/​5 P + 10/​5 –​ 58 15, 24 –​ 73, 82
RTOG 9406 Low (< 15) P + 5–​10 P + 5–​10 –​ 68 0, 6, 11 –​ 68, 74, 79 68, 74,
High ≥ 15) P + SV + 5–​10 P + 5–​10 –​ 56 12, 8, 23 –​ 79 68, 74, 79
Involved P + SV + 5–​10 –​ –​ 68, 74, 79 –​ –​
MSKCC All P + SV + 10/​6 P + SV + 10/​6 –​ 65, 70, 76, 76 0, 0, 0, –​ 65, 70, 76, 81
Rectal block 5
CHHIP (conventionally Low (< 15) P + bSV + 10 P + 10/​5 P + 5/​0 54 16 4 74 SIB technique
fractionated arm) High(≥ 15) P + SV + 10 P + 10/​5 P + 5/​0 59 12 3 74 SIB technique
HYPRO (conventionally Low (<10) P + 3–​10 -​ -​ 78 -​ 78
fractionated arm) Moderate (10–​25) P + SV + 3–​10 P + 3–​5/​0 68 10 78
High (>25) P + SV + 3–​10 -​ -​ 78 -​ 78
236 Prostate cancer

The Radiation Therapy Oncology Group (RTOG)(55) has published a consensus

document on guidelines for lymph node delineation which has been clarified
and modified for use in the PIVOTAL and PIVOTALboost trials. These docu-
ments should be referred to for a detailed description and atlas to aid lymph node
delineation.
In summary, the pelvic vessels are outlined on each axial CT slice in continuity.
The volume should encompass the iliac vessels which lie in the pelvis on its posterior
wall, encompass the internal iliac/​presciatic and external iliac vessels in continuity,
and cover the upper pre-​sacral area anterior to the upper three sacral segments. The
RTOG guidelines suggest a 7-​mm radial margin around the pelvic blood vessels to
cover the involved lymph nodes (the CTV), whilst excluding adjacent bowel, bladder,
bone, and pelvic muscles. Volumes begin at the L5/​S1 interspace and end just above
the superior aspect of the pubic bone. The PTV is then created by applying a 5-​mm
margin in all directions.

10.4.3  Post-​prostatectomy radiotherapy
For patients undergoing adjuvant or salvage radiotherapy post prostatectomy, the
CTV consists of the prostate bed, i.e. the estimated location of the preoperative pros-
tate volume (including sites of possible microscopic tumour extension), plus the extent
of the surgical bed, and should normally include any surgical clips provided that the
normal-​tissue dose-​constraints are within tolerance. The region of the seminal vesicles
should be included in the CTV. In some high-​risk cases, the pelvic lymph node re-
gions may also be included, particularly if a pelvic lymph node dissection has not
been undertaken. The surgical clips now used in robotic prostatectomy are not readily
seen on CT.
The volume is localized using CT (as previously described). Preoperative imaging
(pelvic CT/​MRI), operative notes, and histopathological details from the prostatec-
tomy specimen, including prostate size and tumour extent to specific boundaries of
the surgical resection, as well as the anatomy seen on the postoperative planning CT
scan, can all help to define the CTV.
Since the prostate is located between the rectum posteriorly, the pubis anteriorly,
and the pelvic sling muscles laterally, the volume of the prostate bed can be defined
in relation to these structures. Superiorly, since the bladder will fill part of the space
previously occupied by the prostate between the pubis and the rectum, the volume will
out of necessity include the inferior part of the bladder. The inferior extent of the CTV
should lie at the level of the pelvic floor above the penile bulb.
Guidelines from the RTOG and RADICALS trial protocol are described in Table
10.3(56,57).
Prostate bed PTV: add 1.0cm in all directions, for day-​to-​day variation in set-​up and
for CTV motion.
In a survey of UK practice, Morris et al. found that most UK oncologists used pre-
operative imaging if available to assist with localization, but the location of surgical
clips and bony landmarks were also commonly used as planning aids. As localiza-
tion is less precise than with the prostate in situ, target volumes may be larger than
 Target volume definition 237

Table 10.3  Comparison of RTOG and RADICALS contouring guidelines for post-​

prostatectomy radiotherapy
RTOG RADICALS
Superior Level of cut end of vas deferens or 3–​4 cm Low risk SV Base of SV
above top of symphysis. involvement

High-​risk of SV Tips of SV
involvement
SV absent Superior border
should be
determined with
reference to
the estimated
position of the
preoperative SV
Anterior Above superior Posterior 1–​2 cm of < 2 cm above Posterior aspect
edge symphysis bladder wall. anastomosis of symphysis
pubis pubis
Below superior Posterior edge of > 2 cm above Posterior 1/​3 of
edge of symphysis pubic bone. anastomosis bladder wall
pubis
Posterior Above superior Mesorectal fascia. Anterior rectal wall
edge symphysis
pubis
Below superior Anterior rectal wall.
edge of symphysis
pubis
Lateral Above superior Sacrorectogenitopubic Medial border of obturator
edge symphysis fascia; extend to internus and levator ani muscles
pubis obturator internus
if concern about
extraprostatic disease.
Below superior Levator ani muscle,
edge of symphysis obturator internus.
pubis
Inferior 8–​12 mm below vesicourethral anastomosis; 5 mm cranial to the superior
more if concern for apical margin; extend border or penile bulb
to slice above penile bulb, if vesicourethral
anastomosis poorly visualised.

Sources: data from C. Parker et al., ‘Radiotherapy and androgen deprivation in combination after local
surgery (RADICALS): a new Medical Research Council/National Cancer Institute of Canada phase III trial of
adjuvant treatment after radical prostatectomy,’ BJU Int., Volume 99, Issue 6, pp. 1376–1379, 2007.
J. Croke et al., ‘Proposal of a post-prostatectomy clinical target volume based on pre-operative MRI:
volumetric and dosimetric comparison to the RTOG guidelines,’ Radiation Oncology, Volume 9, Issue 1,
p. 1, 2014.
238 Prostate cancer

for radical prostate treatments. However, rectal sparing can be achieved and standard
dose of 66 Gy in 2-​Gy fractions is generally well tolerated.

10.4.4  Palliative radiotherapy to the prostate/​pelvis

The same principles for localization apply here as for radical treatment. The size of
the PTV and field arrangement will impact on the dose that can be safely prescribed.

10.5 Dose distribution
10.5.1  3D Conformal radiotherapy
IMRT should now be considered standard of care for prostate irradiation, but where
not available 3D conformal is used. A  treatment planning computer which can ac-
curately account for tissue inhomogeneity is used to generate a 3D conformal plan.
A three-​or four-​coplanar field arrangement is usually chosen for treatment.
A three-​field technique would normally consist of an anterior field and two wedged
opposed lateral fields as shown in Fig. 10.2. In patients with asymmetry in the pos-
terior extent of the PTV, substituting a posterior oblique field for one lateral field may
improve target coverage. An anterior and two wedged lateral fields are also normally
used to treat the pelvis, although a posterior field may be added to improve the dose
distribution in larger individuals. Acceptability of a particular plan is assessed by in-
spection of target and OAR DVHs. Dose is prescribed to the isocentre (100%).
In assessing the acceptability of the plan, the clinician should: ·
◆ Ensure target volume covers prostate + seminal vesicles with adequate margins in
all directions.
◆ Check 95% isodose cover of PTV.
◆ Confirm that no unacceptable ‘hot spots’ occur within or outside the PTV.
◆ Confirm volume of rectum irradiated, particularly avoiding circumferential
exposure.
◆ Assess dose to rectum, bladder, and femoral heads, using DVH data (see Table 10.4).
◆ Verify field sizes (approximately 8 × 8 cm). ·
◆ Check weighting of different fields.
DVHs are used to develop local guidelines to help produce acceptable planning con-
straints. Current constraints are based on treatment using conformal techniques
including IMRT and dose escalation studies. A review of the dose volume dependence
of external beam radiotherapy toxicity has been carried out (58). Those for rectum have
been derived from studies comparing DVH data in patients who have or have not
developed rectal morbidity. The volume of rectum receiving 60 Gy is associated with
the risk of Grade 2 rectal toxicity or rectal bleeding. Conservative constraints for 3D
conformal planning are V50 ≤ 50%, V60 < 35%, V70 < 15%, and V75 <3%. The NTCP
models predict that using these constraints will limit Grade 2 late rectal toxicity to <
15% and the probability of Grade 3 late rectal toxicity to < 10% for prescriptions up
to 79.2 Gy in 1.8Gy per fraction(59). Factors which may be associated with increased
 Dose distribution 239

(a)

(b)

(c)

(d)

Fig. 10.2  Axial dose distribution showing prostate conventional 3D plan (a) and IMRT
plan (b) (PTV magenta; rectum orange); Sagittal dose distribution showing prostate
conventional 3D plan (c) and IMRT plan (d) (PTV magenta; rectum orange).
240 Prostate cancer

Table 10.4  Suggested dose–​volume constraints for organs at risk

Organ at risk Dose (Gy) at 2-​Gy per Dose volume constraint (cc)
fraction to 100% Optimal Mandatory
Rectum 30 80 –​

40 65 –​
50 50 60
60 35 50
70 15 15
75 3 5
Bladder 50 50 –​
60 25 –​
65 –​ 50
70 5 35
Sigmoid, small & large 45 78 158
bowel
50 17 110
55 14 28
60 0.5 6
65 0 0
Urethral bulb 50 50 –​
60 10 –​
Femoral heads 50 5 25

toxicity are diabetes mellitus haemorrhoids, inflammatory bowel disease, and prior
rectal or abdominal surgery. Bladder constraints are derived from acute side effect
data only. There is only limited evidence available in the literature regarding the risk of
small bowel toxicity. When delineating contours of bowel loops, the volume of small
bowel receiving 15 Gy or more should be < 120 cc where possible to minimize severe
acute toxicity(60). Studies have found an association between the dose to the urethral
bulb and erectile dysfunction(61). Suggested optimal and mandatory dose constraints
can be found in Table 10.4.

10.5.2  Intensity-​modulated radiotherapy
In the UK, IMRT is the standard delivery technique for radical external beam prostate
radiotherapy. Inverse planned techniques are commonly used.
Inverse planned IMRT is also used for pelvic lymph node irradiation. The superior
conformality allows better sparing of bowel and bladder, as shown in preclinical studies
(see Fig. 10.3)(62). By reducing the dose to OAR, IMRT reduces treatment-​related com-
plications, allowing dose escalation to high-​risk lymph node areas(63).
 Dose distribution 241

Fig. 10.3  Axial dose distribution with colourwash for a prostate IMRT plan.

There are disadvantages to IMRT, namely the higher number of monitor units
(MUs) required and the resulting longer treatment times per fraction when compared
to unmodulated 3D-​CRT. There are also concerns about a potential increased risk of
secondary cancers with IMRT, due to increased scatter dose and increased spread of a
‘low dose bath’ around the pelvis

10.5.3  Volumetric-​modulated arc therapy

Volumetric-​modulated arc therapy (VMAT) is a form of IMRT optimization that al-
lows the radiation dose to be delivered in a single gantry rotation of up to 360°. This
facilitates a reduced treatment delivery time (which in turn potentially reduces the
risk of intra-​fraction organ motion), reduced number of MUs, and better OAR sparing
with comparable target dose coverage to IMRT.
VMAT achieves this by allowing more degrees of freedom such as variations in
gantry-​speed, dose rate, and collimator angle, in addition to dynamically changing
multi-​leaf collimator shaped fields.
The potential disadvantages include the theoretical risk of increased secondary
malignancy. There is less scatter dose from the reduced MU but a larger proportion
of non-​target tissue volume receives 5–​40 Gy with VMAT techniques which may in-
crease the number of secondary cancers.

10.5.4 Tomotherapy
Treatment delivery is based on a CT scanner where the diagnostic X-​ray tube has been
replaced with a 6-​MV linac. The patient is treated in slices by a narrow photon beam.
As the image acquisition is integrated in to the treatment system (same delivery de-
vice), MV CT images are generated. This has an advantage over kV CT imaging in that
the MV CT is less likely to introduce significant artefacts because of the dominance
242 Prostate cancer

of Compton scatter at 6 MV which has a linear relationship between electron density

and Hounsfield units

10.6 Implementation
Before the first treatment, conventional set up accuracy is assessed by comparing the
portal images with digitally reconstructed radiographs (DRRs).
The patient is treated with a linear accelerator of appropriate energy (normally 6 MV
or greater) using an isocentric technique. The same patient conditions as for planning
are sought, i.e. with the patient supine, using the same immobilization technique and
similar rectal and bladder filling. Lasers are aligned with the reference tattoos such that
the approximate treatment isocentre is positioned at that of the machine. The appro-
priate couch moves are then made according to the treatment plan to position the true
treatment isocentre correctly.
An allowance can also be made for ‘couch sag’ on an individual patient basis, which
is measured by reading the couch height with the lateral (coronal) laser aligned with
the couch top. Skin marks are made corresponding to the shifts from the approximate
to true treatment isocentre which should correspond to the position of the lasers on
the skin. These can be observed remotely during treatment to ensure that the treat-
ment position is maintained.
A measurement of focus-​to-​skin distance (FSD) is made from the treatment ma-
chine, which is compared to the expected measurement calculated from the treatment
plan. Using an isocentric technique, the sum of anterior FSD and the distance from
the anterior skin surface to the isocentre, as per treatment plan, should equal 100 cm.
Day-​to-​day variations in bladder filling may cause variation in FSD and discrepan-
cies of up to 1 cm can be accepted. Larger discrepancies indicate a set-​up error or a
more significant change in the patient outline. All fields are treated daily with the ap-
propriate shielding with multi-​leaf collimators.

10.7 Verification
Accurate positioning of the target volume in relation to shaped radiation fields is es-
sential to maximise the effectiveness of treatment.
◆ Systematic errors can occur as a result of incorrect target outlining and data transfer
between the treatment planning stage and actual treatment set-​up. It may also be
caused by incorrect design, marking or positioning of treatment accessories such as
immobilization devices.
◆ Random errors include daily variations in patient set-​up and anatomical changes
between treatments due to bladder and rectal filling, tumour growth or shrinkage,
respiratory movement, and human error.
The practical effect of a systematic error is to shift a dose distribution with reference
to its PTV, and that of a series of random errors is to ‘blur’ its edges, creating a smaller
high-​dose volume and a broader penumbra. Algorithms have been derived allowing
the calculated treatment margins to account for the effects of both random and sys-
tematic geometrical deviation with known probability
 Verification 243

The verification process is designed to detect all positioning errors. For each daily
treatment without fiducials or volumetric imaging, the position of bony structures
obtained by a linear accelerator portal image is compared with their position in ref-
erence images obtained during treatment planning. The reference images are usually
DRRs (see Fig 10.4).
Gold seeds as fiducial markers within the prostate have been investigated(64). Studies
have suggested that insertion, although invasive, is tolerable and seed migration is
rare(65). These studies confirm motion of the prostate occurs most frequently in the
anterior–​posterior and superior–​inferior direction, largely as a result of rectal filling.
With daily on-​line verification using gold seeds to aid set up, there is evidence that the
CTV to PTV margin can be reduced(66).

(a) Right Anterior Oblique DRR

(b) Left Anterior Oblique KV image

Fig. 10.4  Gold markers as seen on verification imaging (a) anterior DRR (b) anterior

electronic portal image (c) lateral DRR (d) lateral electronic portal image.
 (c) Right anterior Oblique KV image

(d) Coronal electronic portal image verification with fiducials

(e) Sagittal electronic portal image verification with fiducials

Fig 10.4 Continued
 Dose prescription 245

Beacons or transponders, although larger, can also be implanted in a similar way to

gold seeds and their position during treatment is monitored by an in-​room electro-
magnetic Calypso localization system. This allows real time tracking data of prostate
motion to be obtained(66). These transponders can cause artefact with MRI scanning
and, although they are usually inserted after diagnostic imaging has been performed,
they can make subsequent imaging to assess disease progression difficult.
Other, less invasive methods of localizing the prostate during radiotherapy, specif-
ically ultrasound and CT have been evaluated. The Clarity transperineal ultrasound
system has been shown to be accurate for intra-​fraction motion estimation.
Cone beam CT incorporates a diagnostic CT device within the linear accelerator
gantry to produce kV images. These CT images can be used to localize the prostate
during a course of treatment and errors due to organ movement are corrected accord-
ingly (see Fig. 10.5). In addition CT gives volumetric information about filling of the
bladder and rectum.
Stereoscopic kV allows both 3D and 4D information to be obtained using two X-​ray
sources and detectors which takes images simultaneously. They can be used to track
movement of both internal and external markers in real time and are used in treatment
systems such as EXACTRAC and CyberKnife.
If discrepancies are detected they are corrected before treatment begins. Daily image
guidance, particularly employing fiducial markers or high quality cone-​beam CT, is
now preferred, resulting in systematic and random errors of ≤ 2mm. This facilitates a
reduction in target volume margins(66).
If daily IGRT is not available, we recommend taking at least three images in the first
week of treatment in order to judge random and systematic components of set-​up un-
certainty. If accuracy is within 3 mm then subsequent weekly checks are adequate. If
errors ≥ 5 mm are found then accuracy should be ensured before the delivery of the
next treatment fraction. If adjustments are required, then three sequential measure-
ments are again made before moving to weekly measurements.

10.8 Dose prescription
With the introduction of 3D-​CRT techniques, the high dose volume was more closely
matched to the tumour target, whilst reducing the radiation to dose limiting normal
tissues. This resulted in a reduction in radiation-​related side effects, as well as a poten-
tial improvement in tumour control. It also facilitated dose escalation. With the wide-
spread adoption of image-​guided IMRT, radiotherapy dose to the prostate can now
be safely escalated to ≥ 74Gy in 2 Gy fractions, improving tumour control probability
whilst maintaining a good safety profile.

10.8.1  Dose escalation—​conventional fractionation

A meta-​analysis of dose-​escalated radiotherapy in localized prostate cancer, included
12 randomized controlled trials with a total patient population of 6884, demon-
strated a significant improvement in biochemical control with increasing biologically
equivalent dose to the prostate(67). The 10-​year absolute benefit in biochemical con-
trol was 9.6% and 7.2% in low-​and intermediate-​risk disease respectively. There were
no significant improvements in cancer-​specific survival, distant metastasis or overall
246 Prostate cancer

(a)

(b)

(c)

Fig. 10.5  Cone-​beam CT (Elekta Synergy) registered with the planning CT and viewed
using a cut plane view (a) coronal (b) sagittal (c) axial views.
 Dose prescription 247

survival. A significant increase in GI toxicity with dose-​escalation was also noted, par-
ticularly in patients treated with 3D-​CRT. Patients treated with IMRT had fewer late
side-​effects.

10.8.2  Dose escalation—​hypofractionation

The evolution of technology enabling better dose conformality and increasing under-
standing of the radiobiology of prostate cancer has resulted in the development of
several hypofractionated schedules for the treatment of prostate cancer.
The alpha/​beta ratio for prostate cancer was estimated at 1.5 Gy (95% CI 0.8–​2.2)
based on the analysis of 367 patients(68). Subsequent larger analyses corroborated the
alpha/​beta ratio at 1.5 Gy (95% CI 1.25–​1.75). The alpha/​beta ratio of the bladder and
rectum has been estimated as 3–​5 Gy. This has two implications for hypofractionation.
The relatively low alpha/​beta ratio of prostate cancer makes it sensitive to increasing
fraction sizes. Furthermore, the relatively higher alpha/​beta ratio of the surrounding
OARs, is likely to increase the therapeutic ratio, with hypofractionation.
Moderately hypofractionated prostate radiotherapy uses 2.4–​4 Gy per fraction, daily
over 4–​6 weeks. Extreme hypofractionation employs 6–​10 Gy per fraction, daily or on
alternate days, over 1–​2 weeks.
Moderate hypofractionation
Randomized data for moderate hypofractionation is available from four modern su-
periority studies and three non-​inferiority studies. The modern studies were designed
with the assumption that the alpha/​beta ratio for prostate cancer is 1.5 Gy.
In the four modern hypofractionation superiority randomized trials(69–​72), the dose
in the conventionally fractionated arm was escalated to 75.6–​80 Gy at 1.8–​2.0 Gy per
fraction, while the dose in the hypofractionated arm was 62–​72 Gy in 2.4–​3.1 Gy per
fraction. No significant differences in toxicity or efficacy have been reported in these
trials after 5 years. No differences in metastasis-​free, cancer-​specific survival, or overall
survival have been demonstrated in these studies.
The three randomized studies evaluating the non-​inferiority of hypofractionated
radiotherapy to conventionally fractionated radiotherapy include the CHHiP
study(71), PROFIT(72), and RTOG 0415(73). The doses in the hypofractionated arms in
these studies range from 57–​70 Gy in 2.5–​3.4 Gy per fraction. These studies dem-
onstrate the safety and efficacy of hypofractionation to be similar to converntional
fractionation.
The largest randomized non-​inferiority study of moderate hypofractionation is the
CHiPP study. Five-​year clinician and patient-​reported side effects were not signifi-
cantly different. The estimated alpha/​beta ratio for prostate cancer from this data is
1.8 Gy.
This study provides high-​level evidence for 60 Gy in 20 fractions becoming the new
standard of care for the management of localized prostate cancer.
CHHiP study(71), PROFIT(72), and RTOG 0415(73) all give hazard ratios <1.0 for
their primary end-​points, demonstrating that the efficacy of hypofractionation
is non-​inferior to conventional fractionation. However, the studies differ in their
late toxicity outcomes. PROFIT reports a decrease in the late toxicity rates with
248 Prostate cancer

hypofractionation, CHHiP reports no difference in late toxicity, while RTOG 0415

reports a small increase in late toxicity with the hypofractionated schedule used.
These differences may be accounted for by differences in the biological effective dose
delivered to bladder/​rectum. If the alpha/​beta ratio for bladder/​rectum is estimated
at 3.0 Gy, the biological effective dose in the hypofractionated arm is less than the
control arm for PROFIT (120 Gy vs 130 Gy), similar to the control arm in CHHiP
(120 Gy vs 123Gy), and greater than the control arm in RTOG 0415 (128 Gy vs
118 Gy).

Extreme hypofractionation
Evidence for profoundly hypofractionated, image guided radiotherapy (stereotactic
body radiotherapy, SBRT) is emerging. Each fraction is delivered daily, on alter-
nate days, or weekly. The dose delivered is biologically equivalent to high-​dose rate
brachytherapy. Owing to the steep dose gradients and small margins of expansion,
image-​guidance forms an essential component to treatment delivery with extreme
hypofractionation. A typical dose distribution is seen in Fig. 10.6.
An early study(74) reports 10 year survival was > 50% in 232 patients treated with
extreme hypofractionation, delivering a dose of 36 Gy in six fractions over 18 days,
using a field-​based technique. In the last decade, several phase I-​II studies using either
IMRT(75–​78) or non-​coplanar techniques(79–​82) have demonstrated that the toxicity and
efficacy of extreme hypofractionation is comparable to conventional fractionation.
Randomized studies evaluating extreme hypofractionation include the PACE B
study, the HYPO study, and the PATRIOT study.
The PACE B (NCT01584258) has now completed accrual, and randomizes patients
36.25 Gy in five fractions (7.25 Gy per fraction, daily) or conventional fraction (which
may be 62 Gy in 20 fractions or 78 Gy in 39 fractions).
The Swedish extreme hypofractionation trial trial HYPO (ISRCTN45905321) has
closed to accrual and randomized patients with intermediate-​risk disease to 78 Gy
in 39 fractions or 42.7 Gy in 7 fractions (6.1 Gy per fraction, alternate days). It has
been presented at ESTRO 2018, showing almost identical 5-​year biochemical control
(83.8% HF, 83.7% CF). Acute genito-​urinary (GU) toxicity was worse for HF, but no
difference in late toxicity was observed.
In the PATRIOT study (NCT01423474), 40 Gy in five fractions is delivered to the
prostate, with patients being randomized to either alternate day treatment (treatment
duration 11 days) or weekly treatment (treatment duration 29 days).The outcomes of
PATRIOT and PACE B are awaited.

10.8.3 Recommended doses
All doses should be prescribed as per ICRU 83.
Conformal radiotherapy (two-​phase)/​IMRT radical treatment to prostate ± seminal
vesicles (SVs) with short-​course neo-​adjuvant androgen deprivation:
Prostate: 60 Gy in 20 fractions over 4 weeks or 74 Gy in 37 fractions over 7½ weeks.
Uninvolved SVs: (2Gy equivalent) 54–​56Gy.
 Toxicity and care during treatment 249

Radical post prostatectomy radiotherapy:

Prostate bed: 66 Gy in 33 fractions over 6½ weeks.
Radical pelvis and prostate radiotherapy in combination with long-​term androgen
deprivation:
Prostate and pelvic 3D-​CRT:
Prostate and SVs: 74 Gy in 37 fractions over 7½ weeks.
Pelvic lymph nodes: 46 Gy in 23 fractions over 4½ weeks.
Prostate and pelvic IMRT (simultaneous boost technique):
Prostate: 74 Gy in 37 fractions over 7½ weeks.
Pelvic lymph nodes and SVs: 55–​60 Gy in 37 fractions over 7½ weeks.
Palliative treatment:
Prostate ± SVs: 30 Gy in 10 fractions over 2 weeks (three or four-​field plan).
36 Gy in weekly fractions over 6 weeks (three or four-​field plan, or IMRT when
available).

10.9  Toxicity and care during treatment

The RT01 trial using 3D-​conformal techniques (83) and CHHiP trial(71) using IMRT
give contemporary estimates of the expected toxicities using conventional dose escal-
ated (74 Gy at 2 Gy per fraction) and modest hypofractionation schedules (60 Gy at
3 Gy per fraction). Late GI side effects are reduced by about 50% using the more ad-
vanced CHHiP techniques and rigorously applied dose constraints.

10.9.1  Acute side effects

Gastrointestinal toxicity
Symptoms of acute proctitis include tenesmus, rectal bleeding, pain, and mucous dis-
charge. There may be accompanying perianal reaction and soreness. Peak acute RTOG
grade ≥2 bowel toxicity was 33% and 25% for the 74 Gy groups in RT01 and CHHiP
respectively. Peak RTOG grade ≥2 bowel toxicity was significantly earlier (3–​4 weeks)
and higher (38%) for the CHHiP 60 Gy group(71). However the acute reaction fol-
lowing modest hypofractionation settles rapidly within 2–​4 weeks and very few pa-
tients remain with symptoms by week 18.
It is important that stools remain soft, and bulking agents (e.g. bran, Normacol,
Fybogel) may be helpful. If bowel frequency is a problem, loperamide, initially at low
dosage, is effective. Proctitis/​anal soreness are helped by steroid/​local anaesthetic sup-
positories or cream (e.g. cinchocaine/​prednisolone).
Diarrhoea with flatulence can indicate small bowel toxicity, particularly in pa-
tients who are receiving radiotherapy to the whole pelvis. Dietary measures in-
clude reducing fibre and lactose content. Short treatment breaks to allow symptoms
to settle can be useful, as may cautious use of loperamide, taking care to avoid
constipation.
250 Prostate cancer

Genitourinary toxicity
Urinary symptoms result from a combination of obstructive and irritative prostatic
symptoms, often on a background of pre-​existing prostatism. Symptoms include fre-
quency, urgency, poor stream, and rarely haematuria.
Peak acute RTOG grade ≥2 bladder toxicity was 39% and 46% for the 74 Gy groups
in RT01(83) and CHHiP(71) respectively. Peak acute RTOG grade ≥ 2 bladder toxicity
is similar (49%) for the CHHiP 60Gy group, in whom the peak reaction occurred at
weeks 4–​5 rather than weeks 7–​8 with conventional fractionation. Again, the acute
reactions settle rapidly after treatment in all groups. Patients are usually advised to
remain well hydrated to decrease bladder irritation and prevent infection although
increased fluid intake may contribute to increased frequency. Urinary tract infection
should be ruled out by urinalysis. Urinary frequency and poor stream secondary to
worsening prostatic obstruction during treatment may be improved by an alpha ad-
renergic blocking drug, such as tamsulosin 400 micrograms daily. Simple analgesics
and antispasmodics may alleviate dysuria and irritative symptoms.
Rarely acute urinary obstruction occurs, necessitating short-​term catheterization
(2% of men in RT01 trial).

Other acute toxicities
Skin erythema or dry desquamation is rare with modern techniques. This can be
treated with an emollient. Fatigue, lethargy, and pubic hair loss are also possible during
treatment.

10.9.2  Late side effects

Late genitourinary toxicity
This includes cystitis, haematuria, and urethral stricture. Post treatment urinary
symptoms should be evaluated in a multidisciplinary team with access to urodynamic
studies. Obstructive symptoms can be treated with alpha adrenergic antagonist drugs,
or urethral dilatation in the case of stricture formation.
Irritative bladder symptoms can be treated with anticholinergic agents. Overall,
however, bladder symptoms improve after prostate radiotherapy, presumably due to
prostate shrinkage and reduction in obstructive symptoms.
In CHHiP, there was a cumulative incidence of RTOG grade ≥2 GU symptoms of
9.1% in the 74 Gy group and 11.7% for the 60 Gy cohort. At 5 years, the prevalence of
small or worse bladder bother was 16–​17%.

Late gastrointestinal toxicity
In the RT01 trial by 5 years, 33% of patients had reported RTOG grade ≥ 2 GI tox-
icity(84). In CHHiP there was a cumulative incidence of 13.7% in the 74 Gy group and
11.9% for the 60 Gy cohort. The prevalence of ≥ small bowel bother at 5 years was 14–​
15% and for ≥ moderate bother 5% in the CHHiP trial, which was approximately half
that seen in the RT01 study. This substantial improvement is attributed to the change
in planning technique, use of IMRT and mandatory dose constraints.
 Treatment outcome 251

Erectile dysfunction
Erectile dysfunction is prevalent in the first 6 months, mainly due to neo-​adjuvant an-
drogen suppression. However, in many men it continues long term. There appears to
be a dose–​volume effect on the penile bulb and erectile dysfunction(61).
Some men benefit from drug treatment such as an oral phosphodiesterase type-​
5 inhibitor. Of this class of drugs sildenafil (Viagra) is commonly used initially. If
treatment is unsuccessful with sildenafil, vardenafil (Levitra) may still prove helpful.
Tadalafil (Cialis) results in a longer duration of erectile ability after administration
than either sildenafil or vardenafil, and consequently can allow a greater degree of
spontaneity of sexual activity. Other drugs include prostaglandin E1 administered ei-
ther intra-​urethrally (MUSE), or by intra-​cavernosal injection (Caverject). A vacuum
device is an alternative strategy for the management of erectile dysfunction.

10.10  The role of hormonal therapy in combination

with radical radiotherapy
Based on the results of many trials it is common practice to use a short-​course of
androgen suppression prior to and during radiotherapy treatment for intermediate-​
and high-​risk patients, and long-​course hormonal therapy in some high-​risk patients.
There is insufficient data to recommend androgen deprivation therapy in low-​risk
patients.

10.10.1  Adjuvant
androgen suppression
with radiotherapy
Meta-​ analysis of randomized control trials demonstrate that adjuvant hormone
therapy following radiotherapy for localized and locally advanced prostate cancer has
significant clinical benefit, with improved overall survival, disease-​specific survival,
and disease-​free survival up to 10 years with no additional radiotherapy toxicity(85).
Monotherapy with bicalutamide for 2 years improves survival when given with sal-
vage radiotherapy(86).

10.11 Treatment outcome
Prognostic groups can be defined using combinations of PSA level, Gleason score, and
clinical stage. Table 10.5 is derived from the CHHiP study
After completion of radiotherapy and hormonal treatment, testosterone recovery
usually occurs. This can cause some PSA elevation that is related to normal prostate
tissue recovery and not reflective of disease recurrence. Rising PSA after radical radio-
therapy may be a sign of local failure, metastatic disease, or both.
Local failure is more likely in patients with low-​to intermediate-​risk prostate cancer
at diagnosis who have a slow rising PSA, with PSA failure occurring some time after
radiotherapy.
The ‘Phoenix criteria’ of PSA failure (defined as PSA nadir plus 2 ng/​mL) is the ac-
cepted standard
252 Prostate cancer

Table 10.5  Stratification of localized prostate cancer into three NCCN risk groups

with corresponding 5-​year biochemical failure free survival after radiotherapy,
within the 60Gy /​20 fraction arm of the CHHiP study, which included 3–​6 months
of neoadjuvant & concurrent hormonal therapy
NCCN risk Group 5-​year biochemical failure-​free survival (%)
Low risk 96.6%

Intermediate risk 90.2%

High risk 84.2%

Source: data from D. Dearnaley et al. ‘Conventional versus hypofractionated high- dose intensity-
modulated radiotherapy for prostate cancer: 5- year outcomes of the randomised, non- inferiority, phase 3
CHHiP trial’ Lancet Oncology, Volume 17, Issue 8, pp. 1047–1060, Aug. 2016.

Patients who have local failure only could be considered for salvage therapy. Such
therapies include radical prostatectomy, cryotherapy, and high-​intensity focused ultra-
sound (HIFU). The role of salvage radical prostatectomy, cryotherapy and brachy-
therapy for disease recurrence following definitive radiotherapy has been reviewed(87).
Salvage radical prostatectomy was associated with 5-​year biochemical disease free
survival rate of 55–​69%. There was significant incidence of complications including
anastomotic stricture, urinary incontinence, and rectal injury.

10.12 Future developments
Future technical developments in prostate radiotherapy are aimed at improving the
dose delivered to the tumour while reducing irradiation to the OARs. Advances can be
divided into four general categories:
◆ Improved target definition.
◆ Improved radiotherapy delivery.
◆ Image guided therapy.
◆ Individualizing therapy.
These four complementary aspects all contribute to improving the therapeutic ratio of
prostate cancer radiotherapy and allow safe dose escalation.

10.12.1  Improved target definition

The greater degree of dose heterogeneity achievable with IMRT makes it possible to
deliver boost doses of radiotherapy to discrete areas of intraprostatic tumour(88). These
are best seen on multiparametric MRI. Diffusion-​weighted MRI, magnetic resonance
spectroscopy, and dynamic contrast-​enhanced MRI have all shown to be effective in
differentiating between normal tissue and malignant prostate tissue.

10.12.2  Improved radiotherapy delivery

The following techniques enable high doses of conformal radiotherapy to be delivered,
optimizing tumour control while sparing normal tissue.
 Future developments 253

Stereotactic body radiotherapy (SBRT)

As discussed in section 10.8.2, SBRT delivers highly conformal, dose-​sculpting radio-
therapy treatment in large fraction sizes. It can be delivered using a conventional linear
accelerator or CyberKnife. CyberKnife is a linear accelerator mounted on a robotic
arm which allows delivery of 6-​MV photons in multiple non-​coplanar beam directions
without movement of the patient (see Fig. 10.6).
Early data on 4–​5 fractions SBRT for prostate cancer reveal low rates of acute(75) and
late toxicity(80,81). The largest follow-​up data of 1100 prostate cancer patients treated
with SBRT using Cyberknife, shows SBRT to have similar efficacy to conventional
fractionation(89).
Randomized trials currently evaluating SBRT to the prostate include the PACE
(NCT01423474 multicentre, phase III), PATRIOT (NCT01423474, phase II), and
HYPO-​RT-​PC (ISRCTN45905321) studies.

Dominant intraprostatic lesion boost

Data on patterns of failure in prostate cancer shows that the local recurrence occurs in
the dominant intraprosatic lesion in 89–​100% of cases. Dose-​escalation to the dom-
inant intraprostatic lesion, targeted using multiparametric MRI, may improve local
control, while reducing the probability of toxicity associated with escalating the dose
to the entire prostate.
Early studies have shown the feasibility and potential benefit of escalating dose to
the dominant nodule within the prostate(90,91). One phase II study has shown that sim-
ultaneous integrated boost to the dominant intraprostatic nodule did not increase
acute toxicity compared to homogenous dosing of the entire prostate(92). Ongoing clin-
ical studies investigating escalating dose to the dominant intraprostatic lesion include
the FLAME study (NCT01168479), DELINEATE study, and the HYPO-​FLAME study
(NCT02853110) using SBRT at 35 Gy in five fractions with a boost of up to 50 Gy. In
the DELINEATE study, dose escalation to the dominant nodule has been performed
with both conventionally fractionated and moderately hypofractionated radiotherapy.

Proton beam therapy
Charged particles such as protons deposit their energy within a small area known as
the Bragg peak, and the radiation dose beyond this rapidly falls to zero. The advantage
is that the volume of normal tissue receiving low-​dose radiation is reduced. There is
limited data of its effective use in localized prostate cancer(93) and proton therapy may
be associated with more GI side effects than IMRT. The significant financial resources
required to implement proton therapy at present restricts its widespread use.

10.12.3  Imaged-​guided adaptive therapy

With newer radiotherapy techniques providing greater 3D conformity around the
PTV, it is becoming increasingly necessary to identify prostate motion during and
between treatments to avoid geometric uncertainty or miss. A number of methods of
tracking prostate motion are available. Such measures have permitted further dose es-
calation, and allow the use of novel fractionation schedules.
254 Prostate cancer

Fig. 10.6  Axial and sagittal images of a prostate SBRT treatment.

Image courtesy of Dr Nicholas van As and Dr Kirsty Morrison.
 References 255

Currently, image-​guidance for inter-​fraction variability in anatomy and position is

performed by adjusting the patient’s position based on co-​registration of simulation
images with imaging acquired at treatment, using fiducials or soft tissue registration,
or both. This can reduce set-​up errors and improves precision of delivery. It does not,
however, address inter-​fraction organ movement or deformation. To account for inter-​
fraction variability in these parameters, an offline or online adaptive treatment plan
may be selected based on anatomy at treatment.
Adaptive radiotherapy techniques for prostate cancer have been investigated(94,95).
Adaptive techniques for prostate radiotherapy will be particularly important in the
context of extreme hypofractionation, where the implications of a geographic miss
during one fraction are significant.

MR linac
MRI image guidance is being integrated with radiotherapy delivery machines. One
system integrates a 0.35 T MRI with Cobalt-​60 sources(96) or, more recently, a linear
accelerator. The MR-​linac combines a 1.5 T MRI with a 6 MV linear accelerator(97).
MR-​guided radiation systems are capable of simultaneous delivery of radiation and ac-
quisition of diagnostic-​quality MR images, with real-​time monitoring of organ motion,
as well as the ability to replan on the day, to account for anatomical changes(98,99). This
will improve image-​guidance, allow the implementation of real-​time adaptive radio-
therapy delivery techniques, and permit reduction of CTV-​PTV expansion margins.

10.12.4 Individualizing therapy
Localized prostate cancer biology is heterogeneous, particularly perhaps for
intermediate-​risk disease. Tumours may differ for example, in DNA repair capacity,
hypoxia, apoptosis, cell-​proliferation, and androgen sensitivity. Greater individualiza-
tion of treatment might be achieved if cancer biology were more precisely defined.
This could include selection of regions for dose-​escalation, concomitant use of sys-
temic agents, and individualization of fractionation schedules. Genomic predictors of
biochemical relapse and of radiosensitivity have been developed but are not in regular
clinical use.We hope that these may be able to assist in determining the the sensitivity
of individual tumours and specific regions within a tumour so that fraction size sensi-
tivity might be better predicted, and fractionation schedules personalized.

References
1. Burford DC, Kirby M, Austoker J. Prostate cancer risk management programme. NHS
Cancer Screen. Programme, 2009.
2. Kupelian PA, Potters L, Khuntia D. et al. Radical prostatectomy, external beam
radiotherapy< 72 Gy, external beam radiotherapy? 72 Gy, permanent seed implantation, or
combined seeds/​external beam radiotherapy for stage T1–​T2 prostate cancer. International
Journal of Radiation Oncology, Biology, Physics 2004; 58:25–​33.
3. Hamdy FC, Donovan JL, Lane JA, et al., 10-​year outcomes after monitoring, surgery,
or radiotherapy for localized prostate cancer. New England Journal of Medicine 2016;
375:1415–​24.
256 Prostate cancer

4. Bill-​Axelson A, Holmberg L, Garmo H, et al., Radical prostatectomy or watchful waiting

in early prostate cancer. New England Journal of Medicine 2014; 370:932–​42.
5. Wilt TJ, Brawer MK, Jones KM, et al. Radical prostatectomy versus observation for
localised prostate cancer. New England Journal of Medicine 2012;367:203–​13.
6. Fossa SD, Wiklund F, Klepp O, et al. Ten-​and 15-​yr prostate cancer-​specific mortality in
patients with nonmetastatic locally advanced or aggressive intermediate prostate cancer,
randomised to lifelong endocrine treatment alone or combined with radiotherapy: final
results of the Scandinavian Prostate Cancer Group-​7. European Urology 2016: S0302-​2838–​5.
7. Mason MD, Parulekar WR, Sydes MR, et al. Final report of the intergroup randomized
study of combined androgen-​deprivation therapy plus radiotherapy versus androgen-​
deprivation therapy alone in locally advanced prostate cancer. Journal of Clinical Oncology
2015; 33:2143–​50.
8. Mottet N, Peneau M, Mazeron J-​J, et al. Addition of radiotherapy to long-​term androgen
deprivation in locally advanced prostate cancer: an open randomised phase 3 trial.
European Urology 2012; 62:213–​19.
9. James ND, Spears MR, Clarke NW, et al. Failure-​free survival and radiotherapy in patients
with newly diagnosed nonmetastatic prostate cancer: data from patients in the control arm
of the STAMPEDE trial. JAMA Oncology 2016; 2:348–​57.
10. Roach 3rd M. Re: The use of prostate specific antigen, clinical stage and Gleason score
to predict pathological stage in men with localised prostate cancer. Journal of Urology
1993:150: 1923–​4.
11. Asbell S, Krall JM, Pilepich MV, et al. Elective pelvic irradiation in stage A 2, B carcinoma
of the prostate: analysis of RTOG 77-​06. International Journal of Radiation Oncology,
Biology, Physics 1988; 15:1307–​16.
12. Hanks GE, Asbell S, Krall JM, et al. Outcome for lymph node dissection negative T-​1b,
T-​2 (A-​2, B) prostate cancer treated with external beam radiation therapy in RTOG 77-​06.
International Journal of Radiation Oncology, Biology, Physics 1991; 21:1099–​1103.
13. Roach M, Bae K, Speight J, et al. Short-​term neoadjuvant androgen deprivation therapy
and external-​beam radiotherapy for locally advanced prostate cancer: long-​term results of
RTOG 8610. Journal of Clinical Oncology 2008; 26:585–​91.
14. Pilepich, MV Winter K, Lawton CA, et al. Androgen suppression adjuvant to definitive
radiotherapy in prostate carcinoma: long-​term results of phase III RTOG 85–​31.
International Journal of Radiation Oncology, Biology, Physics 2005; 61:1285–​90.
15. Urbano TG, Khoo V, Staffurth J, et al. Intensity-​modulated radiotherapy allows escalation
of the radiation dose to the pelvic lymph nodes in patients with locally advanced prostate
cancer: preliminary results of a phase I dose escalation study. Clinical Oncology 2010;
22:236–​44.
16. ICR/​CRUK, Portfolio NIfHRCRNT. Pivotal. A randomised phase II trial of prostate and
Pelvis versus prostate alone treatment for locally advanced prostate cancer. The Institute of
Cancer Research. CRUK/​10/​022; 2011. .
17. Briganti A, Larcher A, Abdollah F, et al. Updated nomogram predicting lymph node
invasion in patients with prostate cancer undergoing extended pelvic lymph node
dissection: the essential importance of percentage of positive cores. European Urology
2012: 61:480–​7.
18. Bolla M, Van Tienhoven G, Warde P, et al. External irradiation with or without long-​term
androgen suppression for prostate cancer with high metastatic risk: 10-​year results of an
EORTC randomised study. Lancet Oncology 2010; 11:1066–​73.
 References 257

19. James ND, Sydes MR, Clarke NW et al. Addition of docetaxel, zoledronic acid, or both
to first-​line long-​term hormone therapy in prostate cancer (STAMPEDE): survival results
from an adaptive, multiarm, multistage, platform randomised controlled trial. Lancet 2016;
387:1163–​77.
20. Studer UE, Whelan P, Albrecht W, et al. Immediate or deferred androgen deprivation for
patients with prostate cancer not suitable for local treatment with curative intent: European
Organisation for Research and Treatment of Cancer (EORTC) Trial 30891. Journal of
Clinical Oncology 2006; 24:1868–​76.
21. Tward JD, Kokeny KE, Shrieve DC. Radiation therapy for clinically node-​positive prostate
adenocarcinoma is correlated with improved overall and prostate cancer-​specific survival.
Practical Radiation Oncology ; 3:234–​40.
22. Kawamorita N, Saito S, Ishidoya S, et al. Radical prostatectomy for high-​risk prostate
cancer: Biochemical outcome. International Journal of Urology 2009; 16:733–​8.
23. Stephenson AJ, Scardino PT, Kattan MW, et al. Predicting the outcome of salvage
radiation therapy for recurrent prostate cancer after radical prostatectomy. Journal of
Clinical Oncology 2007; 25:2035–​41.
24. Bolla M, van Poppel H, Tombal B, et al. Postoperative radiotherapy after radical
prostatectomy for high-​risk prostate cancer: long-​term results of a randomised controlled
trial (EORTC trial 22911). Lancet 2012; 380:2018–​27.
25. Bolla M, van Poppel H, Collette L, et al. Postoperative radiotherapy after radical
prostatectomy: a randomised controlled trial (EORTC trial 22911). Lancet 2005;
366:572–​8.
26. Thompson IM, Tangen CM, Paradelo J et al. Adjuvant radiotherapy for pathological
T3N0M0 prostate cancer significantly reduces risk of metastases and improves survival: long-​
term followup of a randomised clinical trial. Journal of Urology 2009; 181:956–​62.
27. Wiegel T, Bartkowiak D, Bottke D, et al. Adjuvant radiotherapy versus wait-​and-​see after
radical prostatectomy: 10-​year follow-​up of the ARO 96–​02/​AUO AP 09/​95 trial. European
Urology 2014; 66:243–​50.
28. Daly LM, Hickey BE, Lehman M, et al. Adjuvant radiotherapy following radical
prostatectomy for prostate cancer. Cochrane Database Syst Rev, 2011; 12.
29. King CR. The timing of salvage radiotherapy after radical prostatectomy: a systematic
review. International Journal of Radiation Oncology, Biology, Physics 2012; 84:104–​11.
30. Parker C, Sydes MR, Catton C et al. Radiotherapy and androgen deprivation in
combination after local surgery (RADICALS): a new Medical Research Council/​National
Cancer Institute of Canada phase III trial of adjuvant treatment after radical prostatectomy.
BJU International 2007; 99:1376–​79.
31. Chow E, Harris K, Fan G, et al. Palliative radiotherapy trials for bone metastases: a
systematic review. Journal of Clinical Oncology 2007; 25:1423–​36.
32. Salazar OM, Sandhu T, da Motta NW, et al. Fractionated half–​body irradiation (HBI) for
the rapid palliation of widespread, symptomatic, metastatic bone disease: a randomised
Phase III trial of the International Atomic Energy Agency (IAEA). International Journal of
Radiation Oncology, Biology, Physics 2001; 50:765–​75.
33. Dicker AP. The safety and tolerability of low-​dose irradiation for the management of
gynaecomastia caused by antiandrogen monotherapy. Lancet Oncology 2003; 4:30–​6.
34. Weber DC, Nouet P, Rouzaud M, Miralbell R. Patient positioning in prostate
radiotherapy: is prone better than supine? International Journal of Radiation Oncology,
Biology, Physics 2000; 47:365–​71.
258 Prostate cancer

35. Nutting CM, Khoo VS, Walker V., et al. A randomized study of the use of a customized
immobilization system in the treatment of prostate cancer with conformal radiotherapy.
Radiotherapy and Oncology 2000; 54:1–​9.
36. Rosewall T, Chung P, Bayley A, et al. A randomized comparison of interfraction and
intrafraction prostate motion with and without abdominal compression. Radiotherapy and
Oncology 2008; 88:88–​94.
37. Heemsbergen WD, Hoogeman MS, Witte MG, et al. Increased risk of biochemical and
clinical failure for prostate patients with a large rectum at radiotherapy planning: results
from the Dutch trial of 68 GY versus 78 Gy. International Journal of Radiation Oncology,
Biology, Physics 2007; 67:1418–​24.
38. de Crevoisier R, Tucker SL, Dong L, et al. Increased risk of biochemical and
local failure in patients with distended rectum on the planning CT for prostate
cancer radiotherapy. International Journal of Radiation Oncology, Biology, Physics
2005;62:965–​73.
39. Nichol AM Warde PR, Lockwood GA, et al. A cinematic magnetic resonance imaging
study of milk of magnesia laxative and an antiflatulent diet to reduce intrafraction prostate
motion. International Journal of Radiation Oncology, Biology, Physics 2010; 77:1072–​8.
40 Lips IM, Kotte AN, van Gils CH, et al. Influence of antiflatulent dietary advice on
intrafraction motion for prostate cancer radiotherapy. International Journal of Radiation
Oncology, Biology, Physics 2011; 81:e401–​e406.
41. McNair HA, Wedlake L, McVey GP, et al. Can diet combined with treatment scheduling
achieve consistency of rectal filling in patients receiving radiotherapy to the prostate?
Radiotherapy and Oncology 2011; 101:471–​8.
42. Wang KK-​H Vapiwala N, Deville C, et al. A study to quantify the effectiveness of daily
endorectal balloon for prostate intrafraction motion management. International Journal of
Radiation Oncology, Biology, Physics 2012; 83:1055–​63.
43. Hamstra DA, Mariados N, Sylvester J, et al. Continued benefit to rectal separation
for prostate radiation therapy: final results of a Phase III trial. International Journal of
Radiation Oncology, Biology, Physics 2017;97:976–​85.
44. Wilder RB, Barme GA, Gilbert RF, et al. Cross-​linked hyaluronan gel improves the
quality of life of prostate cancer patients undergoing radiotherapy. Brachytherapy 2011;
10:44–​50.
45. Khoo V, Padhani A, Tanner S, et al. Comparison of MRI with CT for the radiotherapy
planning of prostate cancer: a feasibility study. British Journal of Radiology 1999; 72:590–​7.
46. Rasch C, Barillot I, Remeijer P, et al. Definition of the prostate in CT and MRI: a
multi-​observer study. International Journal of Radiation Oncology, Biology, Physics 1999;
43:57–​66.
47. Roach M, Faillace-​Akazawa P, Malfatti C, et al. Prostate volumes defined by magnetic
resonance imaging and computerized tomographic scans for three-​dimensional
conformal radiotherapy. International Journal of Radiation Oncology, Biology, Physics 1996;
35:1011–​18.
48. Kagawa K, Lee WR, Schultheiss TE, et al. Initial clinical assessment of CT-​MRI image
fusion software in localization of the prostate for 3D conformal radiation therapy.
International Journal of Radiation Oncology, Biology, Physics 1997; 38:319–​25.
49. Ohori M, Scardino PT, Lapin SL, et al. The mechanisms and prognostic significance of
seminal vesicle involvement by prostate cancer. American Journal of Surgical Pathology
1993; 17:1252–​61.
 References 259

50. Teh BS, Bastasch MD, Mai WY, Butler EB, Predictors of extracapsular extension and
its radial distance in prostate cancer: implications for prostate IMRT, brachytherapy, and
surgery. Cancer Journal 2003; 9:454–​60.
51. Dearnaley D, Hall E, Lawrence D, et al. Phase III pilot study of dose escalation using
conformal radiotherapy in prostate cancer: PSA control and side effects. British Journal of
Cancer 2005; 92:488–​98.
52. Pickett B Roach M, 3rd, Verhey L. The value of nonuniform margins for six-​field
conformal irradiation of localised prostate cancer. International Journal of Radiation
Oncology, Biology, Physics 1995; 32:211–​18.
53. Harris VA Staffurth J, Naismith O, et al. Consensus guidelines and contouring atlas for
pelvic node delineation in prostate and pelvic node intensity modulated radiation therapy.
International Journal of Radiation Oncology, Biology, Physics 2015; 92:874–​83.
54. Shih HA, Harisinghani M, Zietman AL, et al. Mapping of nodal disease in locally
advanced prostate cancer: rethinking the clinical target volume for pelvic nodal irradiation
based on vascular rather than bony anatomy. International Journal of Radiation Oncology,
Biology, Physics 2005; 63:1262–​9.
55. Lawton CA, Michalski J, El-​Naqa I, et al., RTOG GU Radiation oncology specialists reach
consensus on pelvic lymph node volumes for high-​risk prostate cancer. International
Journal of Radiation Oncology, Biology, Physics 2009; 74:383–​7.
56. Parker C, Sydes MR, Catton C. et al. Radiotherapy and androgen deprivation in
combination after local surgery (RADICALS): a new Medical Research Council/​National
Cancer Institute of Canada phase III trial of adjuvant treatment after radical prostatectomy.
BJU International, 2007; 99:1376–​9.
57. Croke J, Maclean J, Nyiri B, et al. Proposal of a post-​prostatectomy clinical target volume
based on pre-​operative MRI: volumetric and dosimetric comparison to the RTOG
guidelines. Radiation Oncology 2014; 9:1.
58. Jackson A, Marks LB, Bentzen SM, et al. The lessons of QUANTEC: recommendations
for reporting and gathering data on dose–​volume dependencies of treatment outcome.
International Journal of Radiation Oncology, Biology, Physics 2010;76:S155–​60.
59. Michalski JM, Gay H, Jackson A, et al. Radiation dose–​volume effects in radiation-​
induced rectal injury. International Journal of Radiation Oncology, Biology, Physics 2010;
76:S123–​9.
60. Kavanagh BD, Pan CC, Dawson LA, et al. Radiation dose–​volume effects in the
stomach and small bowel. International Journal of Radiation Oncology, Biology, Physics
2010;76:S101–​7.
61. Mangar SA, Sydes MR, Tucker HL, et al. Evaluating the relationship between erectile
dysfunction and dose received by the penile bulb: using data from a randomised controlled
trial of conformal radiotherapy in prostate cancer (MRC RT01, ISRCTN47772397).
Radiotherapy and Oncology 2006; 80:355–​62.
62. Nutting CM, Convery DJ, Cosgrove VP, et al. Reduction of small and large bowel
irradiation using an optimized intensity-​modulated pelvic radiotherapy technique in
patients with prostate cancer. International Journal of Radiation Oncology, Biology, Physics
2000; 48:649–​56.
63. McVey G, Van As N, Thomas K, et al. Intensity modulated radiotherapy (IMRT) can safely
deliver 60 Gy to the pelvic lymph node regions in patients with prostate cancer: report of a
Phase I dose escalation study. International Journal of Radiation Oncology, Biology, Physics
2009; 75:S48.
260 Prostate cancer

64. Vigneault E, Pouliot J, Laverdière J, et al. Electronic portal imaging device detection
of radioopaque markers for the evaluation of prostate position during megavoltage
irradiation: a clinical study. International Journal of Radiation Oncology, Biology, Physics
1997; 37:205–​12.
65. Moman MR, van der Heide UA, Kotte AN, et al. Long-​term experience with transrectal
and transperineal implantations of fiducial gold markers in the prostate for position
verification in external beam radiotherapy; feasibility, toxicity and quality of life.
Radiotherapy and Oncology 2010; 96:38–​42.
66. McNair HA, Hansen VN, Parker CC, et al. A comparison of the use of bony anatomy
and internal markers for offline verification and an evaluation of the potential benefit of
online and offline verification protocols for prostate radiotherapy. International Journal of
Radiation Oncology, Biology, Physics 2008; 71:41–​50.
67. Zaorsky NG, Keith SW, Shaikh T, et al. Impact of radiation therapy dose escalation on
prostate cancer outcomes and toxicities. American Journal of Clinical Oncology. 2016;
41:409–​15  .
68. Fowler J, Chappell R, Ritter M. Is α/​β for prostate tumors really low? International Journal
of Radiation Oncology, Biology, Physics 2001; 50:1021–​31.
69. Pollack A, Walker G, Horwitz EM, et al. Randomized trial of hypofractionated external-​
beam radiotherapy for prostate cancer. Journal of Clinical Oncology 2013; 31:3860–​8.
70. Incrocci L, Wortel RC, Alemayehu WG, et al. Hypofractionated versus conventionally
fractionated radiotherapy for patients with localised prostate cancer (HYPRO): final
efficacy results from a randomised, multicentre, open-​label, phase 3 trial. Lancet Oncology
2016: 17:1061–​9.
71. Dearnaley D, Syndikus I, Mossop H, et al. Conventional versus hypofractionated high-​
dose intensity-​modulated radiotherapy for prostate cancer: 5-​year outcomes of the
randomised, non-​inferiority, phase 3 CHHiP trial. Lancet Oncology 2016; 17:1047–​60.
72. Catton CN, Lukka H, Julian JA, et al. A randomised trial of a shorter radiation
fractionation schedule for the treatment of localised prostate cancer. Journal of Clinical
Oncology 2016; 34:suppl; abstr 5003.
73. Lee WR, Dignam JJ, Amin MB, et al. Randomized phase III noninferiority study
comparing two radiotherapy fractionation schedules in patients with low-​risk prostate
cancer. Journal of Clinical Oncology 2016; 30:2325–​32.
74 Collins C, Lloyd-​Davies R, Swan A. Radical external beam radiotherapy for localised
carcinoma of the prostate using a hypofractionation technique. Clinical Oncology 1991;
3:127–​32.
75. Madsen BL, His RA, Pham HT, et al. Stereotactic hypofractionated accurate radiotherapy
of the prostate (SHARP), 33.5 Gy in five fractions for localised disease: first clinical trial
results. International Journal of Radiation Oncology, Biology, Physics 2007; 67:1099–​1105.
76. Loblaw A Cheung P, D'Alimonte L, et al. Prostate stereotactic ablative body radiotherapy
using a standard linear accelerator: toxicity, biochemical, and pathological outcomes.
Radiotherapy and Oncology 2013; 107:153–​8.
77. Aluwini S, van Rooij P, Hoogeman M, et al. Stereotactic body radiotherapy with a focal
boost to the MRI-​visible tumor as monotherapy for low-​and intermediate-​risk prostate
cancer: early results. Radiation Oncology 2013; 8:1.
78. Kim DN, Cho LC, Straka C, et al. Predictors of rectal tolerance observed in a dose-​
escalated phase 1-​2 trial of stereotactic body radiation therapy for prostate cancer.
International Journal of Radiation Oncology, Biology, Physics 2014; 89:509–​17.
 References 261

79. Fuller DB, Naitoh J, Mardirossian G. Virtual HDR CyberKnife SBRT for localised
prostatic carcinoma: 5-​year disease-​free survival and toxicity observations. Frontiers in
Oncology 2014; 4:321.
80. King CR, Brooks JD, Gill H, Presti JC. Long-​term outcomes from a prospective trial
of stereotactic body radiotherapy for low-​risk prostate cancer. International Journal of
Radiation Oncology, Biology, Physics 2012; 82:877–​82.
81. Chen LN, Suy S, Uhm S, et al. Stereotactic body radiation therapy (SBRT) for clinically
localised prostate cancer: the Georgetown University experience. Radiation Oncology
2013; 8:1.
82. Oliai C, Lanciano R, Sprandio B, et al. Stereotactic body radiation therapy for the primary
treatment of localised prostate cancer. Journal of Radiation Oncology 2013; 2:63–​70.
83. Dearnaley DP, Sydes MR, Graham JD, et al. Escalated-​dose versus standard-​dose
conformal radiotherapy in prostate cancer: first results from the MRC RT01 randomised
controlled trial. Lancet Oncology 2007; 8:475–​87.
84. Syndikus I, Morgan RC, Sydes MR, et al. Late gastrointestinal toxicity after dose-​escalated
conformal radiotherapy for early prostate cancer: results from the UK Medical Research
Council RT01 trial (ISRCTN47772397). International Journal of Radiation Oncology,
Biology, Physics 2010; 77:773–​83.
85. Bria E, Cuppone F, Giannarelli D, et al. Does hormone treatment added to radiotherapy
improve outcome in locally advanced prostate cancer? Cancer 2009; 115:3446–​56.
86. Shipley WU, Seiferheld W, Lukka HR, et al. Radiation with or without antiandrogen
therapy in recurrent prostate cancer, New England Journal of Medicine 2017; 376:417–​28.
87. Touma NJ, Izawa JI, Chin JL. Current status of local salvage therapies following radiation
failure for prostate cancer. Journal of Urology 2005; 173:373–​9.
88. Nutting C, Corbishley C, Sanchez-​Nieto B, et al. Potential improvements in the
therapeutic ratio of prostate cancer irradiation: dose escalation of pathologically identified
tumour nodules using intensity modulated radiotherapy. British Journal of Radiology 2002;
75:151–​61.
89. King CR, Collins S, Fuller D, et al. Health-​related quality of life after stereotactic
body radiation therapy for localized prostate cancer: results from a multi-​institutional
consortium of prospective trials. International Journal of Radiation Oncology, Biology,
Physics 2013; 87:939–​45.
90. Arrayeh E, Westphalen AC, Kurhanewicz J, et al. Does local recurrence of prostate cancer
after radiation therapy occur at the site of primary tumor? Results of a longitudinal MRI and
MRSI study. International Journal of Radiation Oncology, Biology, Physics 2012; 82:e787–​93.
91. Cellini N, Morganti AG, Mattiucci GC, et al. Analysis of intraprostatic failures in patients
treated with hormonal therapy and radiotherapy: implications for conformal therapy
planning. International Journal of Radiation Oncology, Biology, Physics 2002; 53:595–99​
92. Fonteyne V, Villeirs G, Speleers B, et al. Intensity-​modulated radiotherapy as primary
therapy for prostate cancer: report on acute toxicity after dose escalation with simultaneous
integrated boost to intraprostatic lesion. International Journal of Radiation Oncology,
Biology, Physics 2008; 72:799–​807.
93. Yamoah K, Johnstone PA. Proton beam therapy: clinical utility and current status in
prostate cancer. Onco Targets and Therapy 2016; 9:5721–​7.
94. Ahunbay EE, Peng C, Holmes S, et al. Online adaptive replanning method for prostate
radiotherapy. International Journal of Radiation Oncology, Biology, Physics 2010;
77:1561–​72.
262 Prostate cancer

95. Stanley K, Eade T, Kneebone A, Booth JT. Investigation of an adaptive treatment regime
for prostate radiation therapy. Practical Radiation Oncology 2015; 5:e23–​9.
96. Mutic S, Dempsey JF. The ViewRay system: magnetic resonance–​guided and controlled
radiotherapy. Seminars in Radiation Oncology 2014; 24:196–​9.
97. Lagendijk JJ, Raaymakers BW, van Vulpen M. The magnetic resonance imaging–​linac
system. Seminars in Radiation Oncology 2014; 24:207–​9.
98. McPartlin AJ, Li XA, Kershaw LE, et al. MRI-​guided prostate adaptive radiotherapy -​
A systematic review. Radiotherapy and Oncology 2016; 119:371–​80.
99. Pathmanathan AU, van As NJ, Kerkmeijer LGW, et al. Magnetic resonance imaging-​
guided adaptive radiation therapy: A game changer for prostate treatment? International
Journal of Radiation Oncology 2018; 100:361–​73.
100. Peeters ST, Heemsbergen WD, Koper PC, et al. Dose-​response in radiotherapy for
localized prostate cancer: results of the Dutch multicenter randomized phase III
trial comparing 68 Gy of radiotherapy with 78 Gy. Journal of Clinical Oncology 2006;
24:1990–​96.
 Chapter 11

Bladder cancer
Nicholas James, David Fackrell,
and Anjali Zarkar

11.1 Indications
Bladder cancer is the eighth most common male cancer in the UK and the thirteenth
most common among females. There are around 10,000 cases and 5000 deaths an-
nually in the UK and 429,000 cases worldwide. There remains considerable contro-
versy as to the optimal management of localized, muscle invasive bladder cancer.
Surgical removal of the bladder is considered the ‘gold standard’ in many countries
with series citing very high success rates(1,2). However, when data from registry series
are examined, the 5-​year survival from both surgical and radiotherapy series is similar
at around 45–​50%(3,4). In this context, it is likely that the impressive results from
single centre series are more indicative of case selection than superiority of surgery
over radiotherapy—​for example, the pooled surgical data in the neoadjuvant chemo-
therapy trials shows cystectomy outcomes similar to the registry data(5,6). Evidence for
this is the age distribution in the widely cited paper by Stein et al. from University of
Southern California—​median age in this large series was 66 years(1) with a similar me-
dian in the two largest neoadjuvant chemotherapy trials(5,6), whereas 55% of UK cases
are aged over 75 years at diagnosis (CRUK Cancerstats: http://​info.cancerresearchuk.
org/​cancerstats/​). There are no randomized data comparing surgery with bladder pres-
ervation. The SPARE trial(7), comparing surgery with selective bladder preservation in
patients responding to neoadjuvant chemotherapy, failed to recruit and the question
is unlikely to ever be definitively addressed. As around two out of three bladder pres-
ervation patients in our practice are not suitable for neoadjuvant chemotherapy, so
whatever the results of SPARE (median age of included patients around 65 years) they
would have had limited applicability to older less fit bladder cancer patients.
More relevant than the surgery vs RT debate is how to manage older patients or
those unfit for surgery. Furthermore, reported 5-​year survival rates with radiotherapy
are remarkably similar to surgical series. For example, Stein et al. reported results from
a large series of 1054 surgically treated patients, obtaining 5-​and 10-​year survivals, re-
spectively, of 60% and 43%(1). However, when the surgical results are confined to those
patients with muscle invasive disease, overall 5-​year survival drops to 47%. This is very
similar to the 5-​year survival observed in the surgery-​only arm of the SWOG 8710
trial of chemotherapy + surgery vs surgery alone(5). In contrast, Rodel et al.(8) reported
results with endoscopic resection and radiotherapy. When patients with inoperable
264 Bladder cancer

disease are removed from the radiotherapy series to allow direct comparison, 5-​year
survival is reported as 45%. In the recent update of 10-​year outcomes from BC2001,
mature 5-​year survival for chemoradiation is 50%(9,10). A population-​based study from
Ontario looking at bladder cancer outcomes could find no link between treatment
modality and survival which was solely determined by tumour-​related factors such as
stage and grade(3). This lack of data supporting a survival advantage for surgery does
not stop its proponents presenting it as the gold standard(1,2). It is, however, more likely
that survival in bladder cancer is driven by the presence or absence of distant spread
at the time of local therapy and will not be affected by the means adopted for local
control.
Furthermore, all patients undergoing surgery will need either reconstructive bladder
surgery or an ileal diversion. Thus even if surgery is genuinely better than radiotherapy
for patients fit for both approaches, there are many patients for whom radical surgery is
simply not suitable and hence bladder-​preserving techniques are appropriate. Despite
this, use of radiotherapy varies enormously worldwide with possibly a majority re-
ceiving radiotherapy in the UK(4), around 25% in Scandinavia(11) but only around 10%
in the USA(12). Radiotherapy alone suffers from a relatively high rate of incomplete
response or local recurrence (up to 50% or more) possibly due to the effects of case
selection with many poor-​risk patients unsuitable for surgery being referred. A  re-
port of long-​term follow-​up from our institution reported a salvage cystectomy rate
of around 24% with a median time to cystectomy of 12–​18 months(13,14). However, the
addition of synchronous chemotherapy with 5-​fluorouracil and mitomycin C (5FU/​
MMC) reduces the invasive recurrence rate by 45% with improved bladder cancer spe-
cific survival(9,10). Furthermore long-​term quality of life was excellent, with no penalty
from adding 5FU/​MMC to standard dose radiotherapy(15). Similar results were seen
in the BCON trial using carbogen/​nicotinamide as hypoxic cell sensitizers(16) and in
a non-​randomized trial using gemcitabine(17). The more complicated North American
‘trimodality therapy’ schedules show similar outcomes(18). Radiotherapy should thus
always be given, wherever possible, with a simultaneous radio-​sensitizer, the most ro-
bust data with UK fractionation being with 5FU/​MMC or the BCON schedule.
Patients with node positive disease have not been extensively reported in trials.
A small series from our institution has demonstrated that radical treatment sched-
ules can be delivered to bladder and pelvic nodes with synchronous 5FU/​MMC
using IMRT/​IGRT techniques(19). We have separately shown in BC2001 that syn-
chronous 5FU/​MMC can be given after platinum-​based neo-​adjuvant chemotherapy
and that the benefits of synchronous and neoadjuvant therapy are distinct and
complementary(9,10).

11.1.1  Indications by stage

(See Table 11.1 for TNM staging.)
◆ CIS, Ta, T1: no role for radiotherapy.
◆ T2–​T4a N0 M0: potential role for radiotherapy, combined with synchronous radio-​
sensitizer if patient sufficiently fit.
 Indications 265

Table 11.1a  Staging of bladder cancer (TNM 2009)

Tx Primary tumour cannot be assessed

T0 No evidence of primary tumour

Ta Non-​invasive papillary carcinoma
Tis Carcinoma in situ: ‘flat tumour’
T1 Tumour invades subepithelial connective tissue
T2 Tumour invades muscle
T2a Superficial muscle (inner half)
T2b Deep muscle (outer half)
T3 Tumour invades perivesical tissue
T3a Microscopically
T3b Macroscopically (extravesical mass)
T4 Tumour invades any of the following: prostate stroma, seminal
vesicles, uterus, vagina, pelvic wall, abdominal wall
T4a Tumour invades prostate stroma, seminal vesicles, uterus, or
vagina
T4b Tumour invades pelvis or abdominal wall
N—​regional lymph nodes Defined as nodes of the true pelvis below the bifurcation of
the common iliac arteries. Laterality does not affect the N
classification
Nx Nodes cannot be assessed
N0 No lymph node metastasis
N1 Metastasis in a single lymph node in the true pelvis
(hypogastric, obturator, external iliac, or presacral)
N2 Metastasis in multiple lymph nodes in the true pelvis
(hypogastric, obturator, external iliac, or presacral)
N3 Metastasis in a common iliac lymph node(s)
M—​distant metastasis
Mx Cannot be assessed
M0 No distant metastasis
M1 Distant metastasis present
G—​histopathological grading
Gx Grade of differentiation cannot be assessed
G1 Well differentiated
G2 Moderately differentiated
G3–​4 Poorly differentiated/​undifferentiated
266 Bladder cancer

Table 11.1b  Stage grouping

Stage 0a Ta N0 M0

Stage 0is Tis N0 M0

Stage I T1 N0 M0
Stage II T2a,b N0 M0
Stage III T3a,b N0 M0
T4a N0 M0
Stage IV T4b N0 M0
Any T N1, 2, 3 M0
Any T Any N M1

Reproduced with permission from Edge SB, Byrd DR, Compton CC, eds. AJCC Cancer Staging
Manual. 7th ed. New York, NY: Springer, 2010. Used with the permission of the American Joint
Committee on Cancer (AJCC), Chicago, Illinois. The original source for this material is the AJCC
Cancer Staging Manual, Seventh Edition (2010) published by Springer Science and Business Media
LLC, www.springer.com. © Society of Surgical Oncology 2010.

◆ Tany N1-​3 M0: Systemic platinum-​based chemotherapy where possible. Consider

radiotherapy to bladder plus pelvic nodes using IMRT/​IGRT techniques as it allows
synchronous administration of 5FU/​MMC chemotherapy.
◆ Tany Nany M1: there is no role for radical radiotherapy as sole treatment for Stage
IV disease. It may be worth considering, however, as part of a package of ‘radical’
palliation in concert with systemic chemotherapy. No randomized data on the use
of radiotherapy in this setting beyond studies of fractionation.

11.2  Radical primary treatment

11.2.1  Treatment volume and definition
The CTV consists of the bladder including the primary lesion.
The question of whether whole bladder irradiation is necessary is partially unre-
solved. Generally, the whole bladder has been included in the radiotherapy fields as it
has been thought that the development of bladder cancer is associated with a mucosal
field change. However, the need to treat the whole bladder rather than the tumour
alone has not been clearly established. The BC2001 trial compared radiotherapy to the
whole bladder to a reduced dose of 85% to uninvolved bladder in a 2 × 2 trial, which
also compared radiotherapy with chemo-​radiotherapy with 5FU/​MMC (see section
11.2.5). The reduced dose group showed no improvement in either acute or late tox-
icity, nor was there any impact on locoregional control. One interpretation of these
results is that dose to uninvolved bladder was not reduced sufficiently and that with
more sophisticated planning and delivery systems, dose to tumour could be escalated
and dose to uninvolved bladder reduced. For the time being, the whole bladder should
be considered as constituting the CTV.
 Radical primary treatment 267

The normal tissues of concern when treating pelvic lesions are the rectum, small
bowel, and to a lesser extent prostatic urethra if there is no involvement of the ur-
ethra with the tumour. Female patients who are sexually active should be counselled
about the risk of vaginal dryness and stenosis. They should be routinely offered treat-
ment with vaginal dilators, as for women receiving treatment for cervical, vaginal, or
endometrial cancers. Male patients are at risk of erectile dysfunction and should be
counselled about this. It should be noted, however, that the long-​term risk of sexual
dysfunction may be higher in those undergoing radical surgery(20–​22).

11.2.2 Planning technique
Patient position and immobilization
◆ The patient should be planned and treated in the same position; supine with arms
on their chest. Knee and ankle immobilization should be used to ensure patient
positioning is reproducible.
◆ The rectum should be empty of flatus and faeces. The use of daily micro-​enemas
may be considered.
◆ Patients will be asked to empty their bladder 15 minutes prior to scan.
◆ Whilst breathing normally, the patient should have a CT scan performed with 3–​5-​
mm slice spacing. Patients are scanned from bottom of ischial tuberosities to 3 cm
above the dome of the bladder or bottom of L5 (whichever is higher). A flat top CT
scanner should be used.
◆ Neither IV nor oral contrast is thought to be of benefit in this instance.
◆ Reference tattoos should be made at the base of the abdomen and over each hip. The
location of the tattoos should be marked on the planning scan by the use of radio-​
opaque markers to allow cross-​referencing of planning scan and set-​up instructions.

Volume/​field localization
◆ The GTV can be difficult to define and should integrate information from the sta-
ging CT or MRI as well as the diagnostic transurethral resection of the tumour
(TURBT). MRI/​CT fusion may be helpful, where available.
◆ The use of fiducial markers or contrast medium such as lipidiol at the time of
TURBT has been explored and may help identify tumour for image-​guided adap-
tive radiotherapy.
◆ There are little in the way of data on the optimal radiotherapy volume (Fig. 11.1).
A standard approach is to define the CTV as the whole bladder identified by its non-​
involved outer bladder wall plus any extravesical extent expanded to a PTV with a
1.5-​cm margin of tumour with a 2-​cm margin on any extravesical tumour (Fig. 11.2).
◆ All planning and treatment should be carried out with the bladder empty to min-
imize the risk of geographical miss and to keep the treated volumes as small as
possible. Patients with significant residual volumes post voiding should be con-
sidered for planning and treatment with a catheter in situ, although this is likely to
increase urinary toxicity.
268 Bladder cancer

Fig. 11.1  CTV has been outlined in red.

Fig. 11.2  CTV has been grown to PTV by addition of 1.5 cm margin all around PTV.
 Radical primary treatment 269

◆ There are no data to support the routine irradiation of radiologically negative lymph
nodes. The nodal relapse rate in the BC2001 trial, with PTV and CTV defined as pre-
viously, was only 4% in the chemoradiotherapy arm and 6% with radiotherapy only.

Dose distribution, fields, and dose constraints

◆ The treatment is planned using photons of > 8 MV: beam energies less than this re-
sult in higher superficial doses and greater normal tissue irradiation (depending of
course on patient size, larger patients will require photon energies in the 10–​15-​MV
range).
◆ PTV should be covered to encompass the PTV in the 95% isodose (Figs 11.3
and 11.4).
◆ Usually conformal technique using either three-​(an anterior and two lateral/​pos-
terior oblique fields) or four-​field plan is created.
◆ Increasingly, IMRT and IGRT techniques will be used for delivery of radiotherapy.
The combined use of these two modalities will allow more precise tailoring of dose
delivery to take account of changes in bladder and rectal filling and reduce late ef-
fects by reducing dose to the small/​large bowel. Adaptive radiotherapy with three
to four radiotherapy plans using different margins may be used to take account of
changes in bladder filling.
Dose constraints (dose 2 Gy/​fraction) used for organs at risk are as follows(14–​20):

Fig. 11.3  Typical beam arrangement to encompass the PTV by 95% isodose.

270 Bladder cancer

Fig. 11.4  Dose–​volume histogram showing dose received by PTV and critical structures.
(Dose/​fractionation used in this particular plan is 55 Gy in 20 fractions, i.e. 2.75 Gy per
fraction.)

◆ Rectum: V66 < 30%, V60 < 50%, V50 < 60%, V40 < 70%, V30 < 80%.

◆ Femoral heads: V50 < 50%.
Pelvic nodal disease—​see Table 11.2.

11.2.3  Implementation on the treatment machine

The patient should use the same bowel and bladder preparation protocol as at the time
of CT scanning. They should be positioned using the same immobilization and tattoos
should be aligned using wall lasers (Figs 11.5 and 11.6).

11.2.4 Treatment verification
◆ The isocentre position should be verified using methods such as electronic portal
imaging (EPI), megavoltage CT (MVCT), cone-​beam (CBCT), or in-​room CT/​
stereo X-​ray. Three-​dimensional imaging such as CBCT has recently become more
widely available and allows much more accurate definition of the bladder and OARs
on a potentially daily basis.
◆ Likewise, with better image guidance systems on linacs, use of fiducial markers (ei-
ther gold seeds or use of lipid-​based contrast media injected submucosally) and
on-​board kV imaging systems may become more widespread.
◆ It is recommended that daily imaging and on-​line corrections are made where facil-
ities allow. At a minimum, patients should be imaged for the first three fractions and
then weekly, with off-​line (systematic error) correction for errors of 5 mm or greater.
 Radical primary treatment 271

Table 11.2  Dose constraints for OAR when treating bladder

and pelvic nodes
OAR Dose (Gy) Mandatory Optimal

Rectum 30 80% vol

40 65% vol
50 60% vol 50% vol
60 50% vol 35% vol
65 30% vol 30% vol
70 15% vol 15% vol
75 5% vol 3% vol
Sigmoid, small and 45 158 cc 78 cc
large bowel
50 110 cc 17 cc
55 28 cc 14 cc
60 6 cc 0.5 cc
65 0 cc 0 cc
Left femoral head 50 25% vol 5% vol
Right femoral head 50 25% vol 5% vol

Fig. 11.5  Anteroposterior digitally constructed radiograph (DRR) for verification with

portal images.
272 Bladder cancer

Fig. 11.6  Lateral DRR for verification with portal images.

11.2.5 Dose prescription
Radiotherapy as sole treatment
Acceptable radical schedules used in the UK are:
◆ 64–​66 Gy to the reference point in 32–​33 fractions over 6½ weeks.
◆ 55 Gy to the reference point in 20 fractions over 4 weeks.
The optimal schedule has yet to be established. In North America, split schedules are
often used with an interval cystoscopy after a dose of 39–​40 Gy in 1.8–​2-​Gy fractions
is reached. Patients with refractory disease proceed to cystectomy; patients with re-
sponding disease proceed to complete a radical course of radiotherapy alone to a total
dose of 64–​66 Gy. A significant risk of cystectomy remains, however, with 22% under-
going immediate cystectomy, 13% delayed cystectomy for local recurrence, and 65%
retaining the bladder(18,23).
In contrast, patients treated with RT alone to a full radical dose have around a 24%
cystectomy rate at 10 years median follow-​up with conventional radiotherapy alone(13).
This rate drops significantly with synchronous 5FU/​MMC(9,10), suggesting the split
course approach does not really offer any advantage over radiotherapy administered
as a continuous block as in the UK.

Chemoradiotherapy
Two trials have compared this approach to radiotherapy alone in bladder cancer(10,15,24).
The Canadian study(24) randomized 99 patients to radiotherapy (40 Gy in 20 fractions
 Node positive disease 273

over 4 weeks) with or without cisplatin (100 mg/​m2 2-​weekly for three cycles) followed
by elective cystectomy or further radiotherapy. The chemoradiotherapy group had
improved pelvic progression-​free survival (adjusted HR  =  0.50; 90% CI:  0.29–​0.86;
logrank p = 0.038) but was too small to provide reliable estimates of overall survival
effects.
Chemotherapy with cisplatin at this dose is not ideal for bladder cancer as many
patients, particularly those referred for radiotherapy, have impaired renal function
or poor performance status. The BC2001 trial(9,10) tested the hypothesis that syn-
chronous chemoradiotherapy with 5FU/​MMC (mitomycin C 12 mg/​m2 on day 1
and 5FU 500 mg/​m2/​day on days 1–​5 and the last 5 days of treatment) is more ef-
ficacious than radiotherapy alone. Adding chemotherapy to full dose radiotherapy
(55 Gy in 20 fractions or 64 Gy in 32 fractions) was associated with a 33% reduction
in the risk of locoregional recurrence with a reduction of almost 50% in invasive
recurrence. This benefit appeared consistent in preplanned subgroup analyses and
was not affected by prior neoadjuvant chemotherapy suggesting that neoadjuvant
and concomitant chemotherapy confer separate benefits on distant and local con-
trol respectively. The improvement in locoregional control was achieved with
modest increases in acute toxicity that did not reach statistical significance with re-
spect to grade 3 or 4 outcomes. Chemoradiotherapy, even when co-​administered
after neoadjuvant chemotherapy, did not result in impaired late bladder function
or a significant reduction in bladder volume. Late toxicity, measured using RTOG
and LENT/​SOM scales, showed no significant increase with combination therapy
compared to radiotherapy alone. Quality of life transiently fell during radiotherapy
before recovering to pre-​treatment levels by 6 months with no QOL penalty from
synchronous therapy(15).
An alternative approach to radiosensitization is to address tumour hypoxia as re-
ported in another Phase III UK trial (BCON) that randomized 333 patients to radio-
therapy or radiotherapy with synchronous nicotinamide and carbogen(16). Analysis of
the primary endpoint of local relapse-​free survival did not meet statistical significance
(3-​year local RFS: 54% radiotherapy plus nicotinamide/​carbogen vs 43% radiotherapy
alone; HR = 0.88, 95% CI: 0.76–​1.01; p = 0.06) although significant improvements in
overall survival were reported (3-​year rates of 59% radiotherapy plus nicotinamide/​
carbogen vs 46% radiotherapy alone; HR = 0.86, CI: 0.74–​0.99; p = 0.04). No increase
in acute toxicity was reported.

11.3  Node positive disease

Despite the significantly increased mortality and potential increased morbidity, a
pelvic lymph node dissection is considered standard when performing radical cyst-
ectomy. Improvements in surgical outcomes have been reported in several retro-
spective studies(1,25,26). There may, therefore, be potential benefit by treating the pelvic
nodes to a prophylactic, or in some cases, radical dose of radiation (see Figs 11.7 and
11.8). The images show an example of the dose distribution achieved when bladder
and pelvic node have been treated. Nodal radiotherapy, as a treatment for detectable
274 Bladder cancer

Fig. 11.7 Bladder and nodes.

Fig. 11.8 Bladder and nodes 2.

 Palliative treatment 275

disease or prophylaxis, is common in other tumour sites. With IMRT, due to improved
conformality, pelvic radiation is now more tolerable.
Patients with node positive disease could be treated with an integrated boost to posi-
tive nodes and high-​risk patients could receive an adjuvant dose to pelvic nodes. Such
extensive radiotherapy is not suitable for all patients but, in a select few, may prove to
be a suitable and successful treatment option as seen for node positive prostate cancer
patients in STAMPEDE(27), which permitted pelvic nodal radiotherapy for N1 disease.
Limited experience from our centre showed good local control with no grade 3 or 4
toxicity. Five patients with node positive bladder cancer were treated with 64Gy in 32
fractions to the bladder and 53Gy in 32 fractions to bilateral pelvic lymph nodes. Four
patients received concurrent chemotherapy with 5FU/​MMC(28). Median follow-​up of
11.8 months reported minimal urological and gastrointestinal toxicity. However, three
of the five patients had thrombocytopenia and all those affected received concomitant
chemotherapy. The relative high incidence of this complication compared to that re-
ported in BC2001(10) may be explained by the effect of radio-​sensitizing agents on a
larger volume.
More work is necessary to assess which patients would benefit and are most suitable
for such treatment.

11.4 Postoperative radiotherapy
11.4.1 Indications
There is no indication for routine postoperative radiotherapy after cystectomy.

11.5 Palliative treatment
11.5.1 Indications
◆ Stage IV disease not suitable for chemotherapy.
◆ Stage II and III disease in elderly patients with significant comorbidity.
In the UK, patients present with a median age of 72 years (males) and 75 years (fe-
males) with 60% aged > 70 years, often with significant comorbidity due to the associ-
ation between smoking and bladder cancer. Only around 4% of patients present with
de novo metastatic disease (source: British Association of Urological Surgeons (BAUS)
Audit 2003, http://​http://​www.baus.org.uk). Palliative treatment must, therefore, take
into account the pattern of disease and the extent of other clinical problems.

11.5.2  Treatmentvolume and

definition—​locoregional palliation
Whenever possible, all the pelvic disease (primary and nodes) should be encom-
passed. If, however, the volume is excessively large the treatment should be concen-
trated on the area causing the main symptoms. The primary lesion plus any enlarged
276 Bladder cancer

nodes within the compass of a reasonable treatment volume should be included with a
margin of 1.5 cm to give the PTV.
For stage II and III disease the CTV and PTV are the same as for radical treatment
(see section 11.2).

11.5.3 Planning technique
Patient position and immobilization
◆ The patient should be simulated in a comfortable supine position with their arms on
their chest.
◆ Patient comfort should be prioritized over rigid immobilization.
◆ Bladder to be emptied before lying on the scanner bed. No bowel preparation is
required.
◆ A  planning CT scan with the bladder empty should be performed as for radical
treatment.

Volume definition
The CTV will encompass the entire bladder and any extravesical extent of tumour.
A PTV is defined by an expansion of 1.5 cm from the CTV.

Dose distribution, fields, and dose constraints

The majority of cases can be treated with an anterior and two lateral fields to minimize
rectal toxicity, unless the rectum is encompassed by disease in which case anterior–​
posterior parallel-​opposed fields should be used.

11.5.4 Treatment verification
As for radical treatment but where only three fractions are to be given, verification is
done during the first fraction only and treatment delivered unless there is a significant
(> 5 mm) displacement.

11.5.5 Dose prescription
For palliative therapy of bladder recurrence or disease unsuitable for radical treat-
ment, a dose of 21 Gy in three fractions has been shown to be as effective as a longer
palliative course(29).

References
1. Stein JP, Lieskovsky G, Cote R, et al. Radical cystectomy in the treatment of invasive
bladder cancer: long-​term results in 1,054 patients. Journal of Clinical Oncology 2001;
19:666–​75.
2. Zehnder P, Studer UE, Skinner EC, et al. Unaltered oncological outcomes of radical
cystectomy with extended lymphadenectomy over three decades. BJU International 2013;
112:E51–​8.
3. Hayter CR, Paszat LF, Groome PA, et al. The management and outcome of bladder
carcinoma in Ontario, 1982–​1994. Cancer 2000; 89:142–​51.
 References 277

4. Munro NP, Sundaram SK, Weston PM, et al. A 10-​year retrospective review of a
nonrandomized cohort of 458 patients undergoing radical radiotherapy or cystectomy
in Yorkshire, UK. International Journal of Radiation Oncology, Biology, Physics 2010;
77:119–​24.
5. Grossman HB, Natale RB, Tangen CM, et al. Neoadjuvant chemotherapy plus cystectomy
compared with cystectomy alone for locally advanced bladder cancer. [see comment]
[erratum appears in New England Journal of Medicine 2003; 349:1880]. New England
Journal of Medicine 2003; 349:859–​66.
6. Griffiths G, Hall R, Sylvester R, et al. International phase III trial assessing neoadjuvant
cisplatin, methotrexate, and vinblastine chemotherapy for muscle-​invasive bladder
cancer: long-​term results of the BA06 30894 trial. Journal of Clinical Oncology 2011;
29:2171–​7.
7. Huddart RA, Birtle A, Maynard L. Clinical and patient‐reported outcomes of SPARE –​a
randomised feasibility study of selective bladder preservation versus radical cystectomy.
BJU International 2017; 120: 639–​50.
8. Rodel C, Grabenbauer GG, Kuhn R, et al. Combined-​modality treatment and selective
organ preservation in invasive bladder cancer: long-​term results.[see comment]. Journal of
Clinical Oncology 2002; 20:3061–​71.
9. Hall E, Hussain S, Porta N, et al. Outcome of BC2001 patients (CRUK/​01/​004) who
received neoadjuvant chemotherapy prior to randomization to chemo-​radiotherapy (cRT)
versus radiotherapy (RT). Journal of Clinical Oncology 2017; 35:298.
10. James ND, Hussain SA, Hall E, et al. Radiotherapy with or without chemotherapy
in muscle-​invasive bladder cancer. New England Journal of Medicine 2012; 366:
1477–​88.
11. Jahnson S, Damm O, Hellsten S, et al. A population-​based study of patterns of care
for muscle-​invasive bladder cancer in Sweden. Scandinavian Journal of Urology and
Nephrology 2009; 43:271–​6.
12. Konety BR, Joslyn SA. Factors influencing aggressive therapy for bladder cancer: an
analysis of data from the SEER program. Journal of Urology 2003; 170:1765–​71.
13. Cooke PW, Dunn JA, Latief T, et al. Long-​term risk of salvage cystectomy after
radiotherapy for muscle-​invasive bladder cancer. European Urology 2000; 38:
279–​86.
14. Cooke PW, Wallace DMA, Dunn J, et al. Long term follow-​up after radiotherapy for
muscle-​invasive bladder cancer. British Journal of Cancer 1998; 78:26 .
15. Huddart R, Hall E, Miranda M, et al. Quality of life of patients treated for muscle invasive
bladder cancer with radiotherapy +/​-​chemotherapy in the BC2001 trial (CRUK/​01/​
004): Analysis of impact of treatment at an individual level, GU ASCO. Orlando, Florida,
USA, 2017.
16. Hoskin P, Rojas A, Bentzen S, et al. Radiotherapy with concurrent carbogen and
nicotinamide in bladder carcinoma. Journal of Clinical Oncology 2010; 28:4912–​18.
17. Choudhury A, Swindell R, Logue JP, et al. Phase II study of conformal hypofractionated
radiotherapy with concurrent gemcitabine in muscle-​invasive bladder cancer. Journal of
Clinical Oncology 2011; 29:733–​8.
278 Bladder cancer

18. Efstathiou JA, Spiegel DY, Shipley WU, et al. Long-​term outcomes of selective bladder
preservation by combined-​modality therapy for invasive bladder cancer: the MGH
experience. European Urology 2012; 61:705–​11.
19. Fackrell D, Ford D, Chetiyawardana S, et al. The delivery of radical radiotherapy to the
bladder and pelvis in node-​positive (N1) bladder cancer: a five patient case series. BJR Case
Reports 2016; 2:1.
20. Henningsohn L, Wijkstrom H, Dickman PW, et al. Distressful symptoms after radical
radiotherapy for urinary bladder cancer. Radiotherapy and Oncology 2002; 62:215–​25.
21. Henningsohn L, Steven K, Kallestrup EB, et al. Distressful symptoms and well-​being after
radical cystectomy and orthotopic bladder substitution compared with a matched control
population. Journal of Urology 2002; 168:168–​74.
22. Henningsohn L, Wijkstrom H, Dickman PW, et al. Distressful symptoms after radical
cystectomy with urinary diversion for urinary bladder cancer: a Swedish population-​based
study. European Urology 2001; 40:151–​62.
23. Kaufman DS, Winter KA, Shipley WU, et al. The initial results in muscle-​invading bladder
cancer of RTOG 95-​06: phase I/​II trial of transurethral surgery plus radiation therapy
with concurrent cisplatin and 5-​fluorouracil followed by selective bladder preservation or
cystectomy depending on the initial response. Oncologist 2000; 5:471–​6.
24. Coppin CM, Gospodarowicz MK, James K, et al. Improved local control of invasive
bladder cancer by concurrent cisplatin and preoperative or definitive radiation. The
National Cancer Institute of Canada Clinical Trials Group. Journal of Clinical Oncology
1996; 14:2901–​7.
25. Vieweg J, Gschwend JE, Herr HW, et al. Pelvic lymph node dissection can be curative in
patients with node positive bladder cancer. Journal of Urology 1999; 161:449–​54.
26 Herr HW, Donat SM. Outcome of patients with grossly node positive bladder cancer after
pelvic lymph node dissection and radical cystectomy. Journal of Urology 2001; 165:62–​4;
discussion 64.
27. James ND, Spears MR, Clarke NW, et al. Failure-​free survival and radiotherapy in patients
with newly diagnosed nonmetastatic prostate cancer: Data from patients in the control
arm of the STAMPEDE Trial. JAMA Oncology 2016; 2:348–​57.
28. Fackrell D, Ford D, Chetiyawardana S, et al. The delivery of radical radiotherapy to the
bladder and pelvis in node-​positive (N1) bladder cancer: a five patient case series. BJR Case
Reports 2016; 1:2.
29. Duchesne GM, Bolger JJ, Griffiths GO, et al. A randomized trial of hypofractionated
schedules of palliative radiotherapy in the management of bladder carcinoma: results of
medical research council trial BA09. International Journal of Radiation Oncology, Biology
and Physics 2000; 47:379–​88.
 Chapter 12

Testis
Peter Hoskin

12.1 Introduction
The role of radiotherapy in testicular cancer is becoming less prominent. The main-
stay of treatment is radical orchidectomy and, where there is a risk of metastatic dis-
ease, combination chemotherapy. Radiotherapy may be indicated in the following
situations:
◆ Stage I or IIA testicular seminoma delivering prophylactic para-​aortic lymph node
irradiation.
◆ Palliative treatment in the management of chemotherapy resistant disease.

12.2  Prophylactic para-​aortic lymph node irradiation

This has in the past been widely used in stage I  seminoma which is marker nega-
tive (normal alpha-​fetoprotein and human chorionic gonadotrophin). However, in-
creasingly radiotherapy has been displaced by either surveillance or single-​agent
carboplatin and is now only occasionally used for patients declining carboplatin or
those with stage IIA disease declining chemotherapy.
Trials by the Medical Research Council in the UK have clarified the treatment
volume (CTV) and dose. TE10 randomized patients to receive treatment with either
para-​aortic lymph node irradiation or a dogleg field incorporating the ipsilateral iliac
lymph nodes, both arms receiving a dose of 30 Gy in 15 fractions as well. There were
four pelvic relapses in the para-​aortic field group compared to none in the dogleg
group but overall nine relapses occurred in each group and there were no significant
differences in disease-​free or overall survival; on the basis of this trial para-​aortic
lymph node irradiation alone is considered the standard of care.
TE18 compared to radiation dose 30 Gy in 15 fractions and 20 Gy in 10 fractions
and again no significant difference has emerged between these doses for relapse-​free
or overall survival. The 20-​Gy arm reports better quality of life scores for acute toxicity
and therefore 20 Gy in 10 fractions is regarded as the standard of care.

12.2.1  Patient position and immobilization

The patient is treated supine, arms by the side with no specific immobilization or ankle
stocks to prevent pelvic rotation.
280 Testis

12.2.2 Field localization
In the past standard fields defined using the orthovoltage X-​ray simulator have been
employed based on bony landmarks as follows:
◆ Superior: bottom of T10.
◆ Inferior: bottom of L5.
◆ Lateral borders to edge of the renal hila.
Using this approach, kidneys will be localized by either reconstructing from computed
tomography (CT) or using an intravenous urogram (IVU) at the time of simulation.
An asymmetric field may be used with right lateral border constrained to the trans-
verse processes of the vertebrae and on the left to the edge of the renal hilum to account
for the different drainage on the left side which feeds into the left renal vein distinct
from the right side feeding directly into the inferior vena cava. In practice, both fields
can be extended laterally to the edge of the renal hilae without significantly increasing
volume or toxicity and this is the recommended approach.
In most modern centres, this will now be undertaken using the CT simulator. The
same borders may be employed or a formal CTV defined based on a node atlas for the
subdiaphragmatic para-​aortic nodes.

12.2.3 Dose distribution
Anterior and posterior opposed fields are used. No additional shielding is
recommended.

12.2.4 Dose prescription
The standard dose is 20 Gy in 10 fractions treating daily Monday to Friday.

12.2.5  Implementation and verification

The field centre is tattooed together with lateral tattoos to identify rotation of the
trunk. The field is then set up to the central tattoo and skin marks. Verification using
kV EPID images and bony match should be undertaken daily for the first 3 days to
identify any systematic error in set-​up and at the beginning of the second week.

12.2.6 Patient care
Treating a significant amount of small bowel and stomach nausea is common in pa-
tients receiving this treatment. In some centres prophylactic antiemetics are offered; if
this is not the case then access to antiemetics to be taken regularly if nausea develops
should be facilitated.
Patients having testicular cancer postorchidectomy are often concerned regarding
future fertility. They can be reassured that the para-​aortic lymph node field, distinct
from the dogleg fields used in the past, results in no significant dose to the testis
and would have no impact on fertility. There is no indication for in vivo on the testis
dosimetry.
 Palliative treatment in chemoresistant disease 281

12.3  Palliative treatment in chemoresistant disease

This may embrace a number of scenarios:
◆ Persistent para-​aortic lymphadenopathy.
◆ Mediastinal or supraclavicular lymphadenopathy.
◆ Bone and cerebral metastasis.

12.3.1  Para-​aortic lymphadenopathy
This should be treated in a similar fashion to that described in section 12.2 for prophy-
lactic treatment. It is, however, important to identify residual tumour masses and such
patients may be better treated with a formal CT plan defining a gross tumour volume
and expanding this by 0.5 to 1 cm to a CTV which will be further expanded to a plan-
ning tumour volume. As in prophylactic treatment it is important to identify the kid-
neys and ensure that the treatment does not exceed renal tolerance.

12.3.2  Mediastinal and supraclavicular lymph nodes

These should be treated in an analogous fashion to any other palliative mediastinal
disease as described in Chapter 6 for lung cancer. Localization based on CT simulator
images for anterior and posterior paired beams will be the usual approach.

12.3.3  Bone and cerebral metastasis

Bone and brain metastasis should be treated in a standard fashion as described in
Chapter 22.

12.3.4 Dose
Standard palliative doses should be used, for example, 20 Gy in five fractions or 30 Gy
in 10 fractions to lymph node masses and bone and brain metastases as described in
Chapter 22.

Further reading
Chung P, Mayhew LA, Warde P, et al. Management of stage I seminomatous testicular
cancer: A systematic review. Clinical Oncology 2010; 22:6–​16.
Fossa SD, Horwich A, Russell JM, et al. Optimal planning target volume for stage I testicular
seminoma: A Medical Research Council randomized trial. Medical Research Council
Testicular Tumor Working Group. Journal of Clinical Oncology 1999; 17:1146.
Jones WG, Fossa SD, Mead GM, et al. Randomized trial of 30 versus 20 Gy in the adjuvant
treatment of stage I testicular seminoma: A Report on Medical Research Council Trial
TE18, European Organization for the Research and Treatment of Cancer Trial 30942
(ISRCTN 18525328). Journal of Clinical Oncology 2005; 23:1200–​8.
Warde P, Specht L, Horwich A, et al. Prognostic factors for relapse in stage I seminoma
managed by surveillance: a pooled analysis. Journal of Clinical Oncology 2002;
20(22): 4448–​52.
 Chapter 13

Penis
Peter Hoskin

13.1 Introduction
Carcinoma of the penis is typically a squamous carcinoma arising on the penile shaft
or glans in an uncircumcised patient. Management is most commonly by primary sur-
gery, either total amputation or partial amputation with reconstruction, but primary
radiotherapy remains an option for selected patients. It may be considered for those
patients with T1 and T2 tumours < 4 cm in diameter, particularly in those unfit for
surgery, those with locally advanced disease and fixed inguinal lymph nodes, and for
patients in whom surgical treatment may require total amputation and where they
choose to have organ preservation by radiotherapy as an alternative. No randomized
trial comparison is available to give accurate figures for the relative efficacy of either
treatment. Brachytherapy is an alternative means of delivering high-​dose radiotherapy
to the penis and may be considered where there is local expertise for this instead of
external beam treatment.
Postoperative radiotherapy may be indicated in some circumstances where there
has been inadequate proximal clearance or where inguinal lymph nodes are found to
contain metastatic tumour with high-​risk criteria. There are no robust criteria upon
which to base recommendations but as in other sites where there is heavy involvement
(more than four nodes), extracapsular extension, or extensive lymphovascular infiltra-
tion then postoperative radiotherapy may be considered. It is important to recognize,
however, that adding radiotherapy to surgery in this region will substantially increase
the risks of long-​term toxicity, in particular pain and lymphoedema.
Full computed tomography (CT) restaging of the internal iliac nodes should be
available and fludeoxyglucose –​positron emission tomography (FDG-​PET) may help
refine status of the higher lymph nodes.

13.2  Technique for radical external beam treatment

13.2.1  Patient position and immobilization
The patient is treated supine.
◆ The common technique is to hold the penis in a wax or Perspex block which acts
both as an immobilization device and also provides build-​up to ensure maximum
dose deposition in the clinical target volume (CTV) from megavoltage beams. The
 Technique for radical external beam treatment 283

(a)

(b)

Fig. 13.1 Demonstrating
(a) customized wax block
or (b) standard Perspex
block on scrotal shield.

block typically rests on a lead shield with a hole through which the penis protrudes
into the block as shown in Fig. 13.1. The penis may be held in a length of tubigrip to
facilitate position within the block.
◆ An alternative approach is required where the penis is short or retracted and cannot
be pulled into the wax block. In this setting lead shielding to the underlying testis
and skin of the lower abdomen and groins is used and surface bolus applied.
◆ Note: the penis may become swollen from the acute radiation reaction and this can
lead to a paraphimosis; all patients should therefore be circumcised or undergo a
dorsal slit of the foreskin prior to treatment.

13.2.2  Volume definition

The CTV should include the gross tumour volume (GTV) with a 2-​cm margin prox-
imally and distally. In practice this will often represent most of the penile shaft. The
entire circumference of the penile skin should be included in the CTV. No further
expansion to PTV is usually employed.
284 Penis

13.2.3 Field localization
This will depend upon the immobilization and set-​up.
◆ Where the block technique is used, which is the common approach, field localiza-
tion using lateral opposing fields to encompass the block shown is readily identified
with the light beam which should be seen to splash outside the edges of the block
shown in Fig. 13.2.
◆ Where a wax block cannot be used then a CT-​planned volume is best using a CT-​
defined volume. Typically a field arrangement using two lateral oblique beams using
6-​MV energy will be best as shown in Fig. 13.3.

13.2.4  Implementation and verification

Where the block is used then this defines the patient set-​up. Often the patient can be
educated to position the block themselves each day on the underlying lead shielding.
Micropore tape or similar may be required to steady the set-​up on the linear acceler-
ator couch. Where no block is used then set-​up will be to a lateral tattoo from which
the isocentre is defined since the field centre beneath the midline is likely to be under
penile tissue or overlying bolus on which it is not possible to define an accurate set-​up.
Verification using kV electronic portal imaging device (EPID) images should be
undertaken. Where the wax block is used then this is simply to ensure adequate coverage
of the block. Cone beam CT imaging may aid verification of a planned volume.

13.2.5 Dose prescription
A total dose of 64 Gy in 32 daily fractions treating Monday to Friday is given. This is
prescribed to the mid plane, that is the centre of the block, where this technique is used
and to the intersection point where the alternative technique is used.

Fig. 13.2  Lateral field set-​

up marks on block.
 Technique for lymph node irradiation 285

Fig. 13.3  Anterior oblique fields to treat CT-​defined PTV.

13.3  Technique for lymph node irradiation

13.3.1  Patient position and immobilization
The patient is treated supine.

13.3.2 Volume definition
CT planning scans, 3-​mm slices with intravenous contrast are used.
The CTV comprises the inguinofemoral canal, and external iliac and internal iliac
lymph nodes. These should be defined based on the standard lymph node atlases and
as described in Chapter 10 for prostate and Chapter 16 for lymphomas.
Where there are palpable or radiologically abnormal lymph nodes then these should
be defined in a separate subvolume on the planning CT images.
The PTV will be derived by a volumetric expansion according to local practice, typ-
ically of the order of 5 mm.

13.3.3 Field arrangement
A parallel-​opposed pair of beams using the multileaf collimator (MLC) or individu-
alized blocks to fit the nodal chains may be used; however, both static field IMRT and
VMAT solutions will reduce the volume of small bowel, bladder, rectum, and bone
marrow irradiated and are therefore preferable. An example is shown in Fig. 13.4.
Where there is a nodal subvolume then this is best treated using a simultaneous
integrated boost with an intensity-​modulated radiotherapy (IMRT) or volumetric-​
modulated arc radiotherapy (VMAT) plan. Alternatively a separate plan will be re-
quired which will be delivered as a second phase treatment. These will typically be
in the inguinal region and best covered by a direct electron field or wedged photon
beam plan.
286 Penis

(a)

(b)

Fig. 13.4 Rapidarc
plan to treat
post operative
nodal volume in
(a) transaxial and
(b) coronal views.

13.3.4  Implementation and verification

The field centre is tattooed together with lateral tattoos to identify pelvic rotation.
Verification will follow standard practice using kV electronic portal imaging veri-
fication with match to bone landmarks against the planning digitally reconstructed
radiograph.
 Further reading 287

13.3.5 Prescription
◆ 50–​50.4 Gy in 25–​28 fractions over 5–​5½ weeks.
◆ Or 40 Gy in 15 fractions over 3 weeks.
The dose to a nodal subvolume will depend upon technique:
◆ SIB: 56–​60Gy in 25–​28 fractions over 5–​5½ weeks.
◆ Phase II boost: 16 Gy in eight fractions over 1½ weeks.

13.4 Palliative treatment
Palliative treatment for locally-​advanced fixed fungating or bleeding tumours may be
indicated. These should be treated pragmatically with derivations of the previously de-
scribed technique. Doses for palliation may include the following:
◆ 21 Gy in three fractions treating three times weekly.
◆ 20 Gy in five daily fractions.
◆ 8–​10 Gy as a single dose.

13.5 Patient care
Acute skin and mucosal reactions are inevitable and should be managed conserva-
tively. The penis will become swollen and dysuria may develop.
There are important late effects to be considered in preparing the patient for
radiotherapy:
◆ Lymphoedema of the lower limbs, more marked where there has been previous
groin surgery.
◆ Sexual dysfunction related to fibrosis and penile shortening.
◆ Urethral stricture.

Further reading
Hakenberg OW, Compérat EM, Minhas S, et al. European Association of Urology. EAU
guidelines on penile cancer: 2014 update. European Urolology 2015; 67(1):142–​50.
Van Poppel H, Watkin NA, Osanto S, et al. ESMO Guidelines Working Group. Penile
cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-​up. Annals
of Oncology 2013; 24 Suppl 6:vi115–​24.
 Chapter 14

Uterus: Endometrium and Cervix
Melanie Powell and Alexandra Taylor

14.1 Introduction
Tumours of the uterus may originate from either the uterine body (corpus) or the
uterine cervix. Radiotherapy has an important role in the management of these can-
cers, as either the primary treatment modality, an adjuvant to surgery or palliation of
symptoms.
Uterine corpus cancer (endometrial) is increasing in incidence and with an inci-
dence of 19 per 100 000 is the fourth commonest malignancy in women in the UK. It
is related to obesity and is therefore likely to become more prevalent.
Following the introduction of a national screening programme in 1989 the inci-
dence of cervical cancer almost halved. In the UK it is now a relatively rare cancer,
ranking at number 12 for women with an incidence of 8 per 100 000. Worldwide, how-
ever, it remains a major problem and is the second commonest malignancy affecting
women. In India, parts of Africa, the Caribbean, and South America it is the leading
cause of cancer in women.

14.2 Anatomy
The uterus consists of the uterine body and the cervix that are separated by the in-
ternal os. Posteriorly, the pouch of Douglas is the region between the posterior va-
ginal fornix and the rectum. The parametrium is a layer of connective tissue that lies
adjacent to the uterine body, cervix, and vagina. It is rich in vascular and lymphatic
vessels and contains the ureters that pass below the broad ligament and lateral to the
cervix. The floor of the parametrium is formed by the cardinal (lateral cervical) liga-
ments, which arise from the lateral margins of the cervix and insert into the pelvic
sidewall, and the uterosacral ligaments that pass from the uterus around the rectum
to the sacrum.
There are several interlinking pathways of lymphatic drainage from the uterus and
cervix to the pelvic lymph nodes. The nodal groups lie in close proximity to the pelvic
blood vessels as shown in Fig. 14.1. The cervix and lower uterine body drain predom-
inantly via the parametrial nodes to the obturator and external iliac nodes, and along
the route of the uterine vein to the internal iliac nodes. Spread can also occur along the
pathway of the uterosacral ligaments to the sacral nodes. The upper uterine body also
drains along the pathway of the ovarian vessels to the common iliac and para-​aortic
nodes. As a result of this extensive network, nodal metastases may occur at any level
and in any combination.
 Cervical cancer 289

(a)

(b)

Fig. 14.1  Position of the

pelvic lymph nodes in
relation to blood vessels
and bony landmarks.
Nodal groups: PA, para-​
aortic; CI, common iliac;
EI, external iliac; ING,
inguinal; OB, obturator;
II, internal iliac; PS,
presacral.

14.3 Cervical cancer
The majority of cervical tumours are human papilloma virus (HPV)-​induced squamous
cell carcinomas arising from the squamo-​columnar junction. There is an increasing inci-
dence of adenocarcinoma (also HPV related) that usually originates in the endocervical
canal and must be differentiated from primary endometrial adenocarcinoma.

14.3.1  Primary radical treatment for cervical cancer

Indications
The selection of the appropriate primary treatment for cervical cancer is essential and
depends on the tumour stage, size, nodal status, local expertise, and patient preference.
290 Uterus: endometrium and cervix

Although imaging plays an increasingly important part in the management of cervix

cancer, FIGO staging is based on clinical findings. This means that a careful exam-
ination under anaesthetic (EUA) assessing the size and extent of the tumour by
evaluating vaginal, parametrial, pelvic sidewall, and uterosacral involvement must be
carried out. Cystoscopy and sigmoidoscopy are used to visualize bladder and rectum.
The combination of both radiotherapy and surgery has a greater risk of morbidity
than either treatment as a single modality and should be avoided. In FIGO stage IB2
and IIA disease, surgery and radiotherapy have equivalent local control and survival
rates. In general, these early stage tumours are treated with surgery as this enables
preservation of ovarian function for younger women. Radiotherapy is the treatment of
choice for bulky stage IIB to IVA tumours, node positive disease, and for those unfit
for surgery. Radiotherapy may also be curative for recurrent pelvic disease in patients
previously treated surgically (salvage therapy).
Radical radiotherapy comprises a combination of external beam radiotherapy to
encompass the primary tumour and regional nodes followed by intrauterine brachy-
therapy to deliver a high central dose to the primary tumour while sparing the sur-
rounding normal tissues. There is strong evidence for a dose–​response relationship
for local control and overall survival, but there is a corresponding increase in the in-
cidence of normal tissue toxicity that limits the total radiation dose that can be safely
delivered.

Treatment volume and definition

Cervical cancer may become very locally advanced without distant spread. It can
infiltrate superiorly to the uterine body, inferiorly to the vagina, or laterally to the
parametrium. Extension to the pelvic sidewall causes hydronephrosis by compression
or infiltration of the distal ureter as it passes lateral to the cervix. Pelvic sidewall fix-
ation may be due to either direct tumour extension or coalescence with paracervical
lymphadenopathy. Tumours may also invade the uterosacral ligaments and rectum
posteriorly or the base of bladder anteriorly.
There is a rich lymphatic drainage of the cervix via several pathways. Lymph node
spread is common and may be non-​contiguous. From surgical series, the most fre-
quently involved nodes are the medial external iliac, internal iliac, and obturator
nodes although primary spread to the common iliac and para-​aortic nodes is not un-
common. Presacral and mesorectal nodal involvement is rare in the absence of pos-
terior tumour extension.
The risk of nodal metastatic spread increases with tumour size and stage. The ap-
proximate incidence of pelvic lymph node involvement is 15–​20% in stage IB2, 30% in
stage IIB, 50% in stage IIIB, and 80% in stage IVA disease. The incidence of para-​aortic
node involvement is 5–​10% in stage IB2, 15% in stage IIB, 30% in stage IIIB, and 50%
in stage IVA disease. Haematogenous spread, which is most frequently to the lungs,
liver, and bone, is usually a late event.
Magnetic resonance imaging (MRI) is the optimal method for imaging local disease
with accurate assessment of stromal invasion, parametrial involvement, and uterine
extension (Fig. 14.2). The pelvic and para-​aortic nodes should be imaged although
the assessment of lymph nodes with computed tomography (CT) and MRI using size
 Cervical cancer 291

Fig. 14.2  Sagittal T2 MRI of

pelvis showing large cervical
tumour. PTV borders outlined in
yellow.

criteria is a relatively insensitive technique for predicting microscopic nodal metas-

tases. Positron emission tomography (PET) imaging has higher sensitivity and specifi-
city for detecting regional nodal involvement and distant metastases, and can alter the
target volume in nearly 20% of cases.

Clinical target volume for cervical cancer

The clinical target volume (CTV) for external beam radiation covers all areas of poten-
tial microscopic spread and comprises:
◆ Gross tumour volume (GTV).
◆ Cervix.
◆ Uterus.
◆ Upper half of the vagina (full length of vagina in stage 3A disease).
◆ Parametria.
◆ Proximal uterosacral ligaments (entire mesorectum if uterosacral ligament
involved).
◆ Ovaries.
◆ Pelvic lymph nodes to include internal iliac, obturator, external iliac, common iliac
(to the level of the aortic bifurcation), and subaortic presacral nodes.
The EUA findings and imaging must be used to modify the target volume.
◆ In stage 3A disease the entire vagina is treated.
◆ The full length of the uterosacral ligaments and presacral nodes are covered when
there is posterior tumour extension.
292 Uterus: endometrium and cervix

◆ The inguinal nodes are irradiated when there is involvement of the lower third of
the vagina.
◆ The para-​aortic region is also treated if there is overt involvement or if there is
common iliac node enlargement due to the high risk of microscopic para-​aortic
nodal metastases (see section 14.5).

Planning technique
Patient position and immobilization
The patient is immobilized supine with knee and ankle supports, and with their arms
on the chest. Skin tattoos are placed anteriorly in the midline and laterally to prevent
lateral rotation.
Target volume localization
CT localization  Three-​dimensional CT planning is the best method for ensuring full
coverage of the target volume. The patient is immobilized in the treatment position
with radio-​opaque markers placed over the skin tattoos. A planning CT scan is taken
with 2.5–​3 mm slices from the top of the third lumbar vertebra to 5 cm below the is-
chial tuberosities. Administration of intravenous contrast improves visualization of
the pelvic blood vessels, the uterus, and the primary tumour and is recommended
provided the patient has good renal function.
The bladder should be comfortably full in order to displace small bowel from the
pelvis and to reduce the volume of bladder irradiated. Bladder and rectal filling affects
the position of the cervix and uterus and therefore a bladder-​filling protocol should
be used both for consistency and to limit the interfractional variation in uterine and
cervix position. Since overfilling the bladder is uncomfortable and likely to be impos-
sible to sustain towards the end of treatment, a moderately full bladder is suggested.
An example of a bladder-​filling protocol is that 1 hour prior to planning or treatment
patients are asked to empty their bladder and then drink 350 mL of fluid. A repeat
CT scan with an empty bladder can be used to construct an individualized range of
uterine motion.
Target volume delineation
The target volume is delineated on each axial CT slice. Image registration of MRI and
CT-​PET images can aid volume localization. If an adaptive approach is to be used then
variable bladder filling CT scans are used to aid CTV-​PTV margin selection and/​or
production of plans.
GTV: This is contoured if an integrated boost technique is used.
GTVprimary: defined with clinical findings and MRI scan.
GTVnodes: Involved nodes identified on diagnostic imaging. A 5–​7 mm margin is used
to define the PTV for the high dose boost (PTVboost).
CTV:  The structures described above (see ‘Clinical target volume for cervical
cancer’) are contoured on each slice. The nodal areas are defined using the blood
vessels as a surrogate target. A 7-​mm margin around vessels extending to the pelvic
sidewall is recommended for the pelvic nodal regions. Any visible nodes must also be
encompassed.
 Cervical cancer 293

PTV: The PTV margin will allow for internal motion and set-​up error. Since the
uterus and cervix are affected by both rectal and bladder filling a larger margin
is required for these structures than for the nodal areas. A margin of 15–​20 mm
should be added to the cervix, tumour, and uterus and co-​registration of the PET,
MRI, or empty bladder CT scans can be used to modify this volume. For all other
structures a margin of 5–​7 mm is sufficient depending on local departmental policy
for verification.
For the minority of patients who cannot receive brachytherapy, a second phase of
treatment with external beam therapy will be delivered to a smaller volume. The PTV
for this second phase is created by adding a 10–​12 mm margin around macroscopic
disease (GTV).
Simulator localization Standard field borders have a high risk of a geographical
miss and cross-​sectional imaging must be used to determine the target volume. If
conventional simulation is necessary, the patient is examined in the treatment pos-
ition and the lower extent of vaginal disease or the introitus is marked with a radio-​
opaque marker. The diagnostic MRI scans are then used to modify the field borders
(Fig. 14.2).
The PTV is defined with anteroposterior and lateral simulator images. The superior
volume margin should cover the common iliac nodes and is usually at the upper
border of the fifth lumbar vertebra. The inferior border is placed at least 3 cm below
the inferior aspect of disease, either clinically or on the MRI, and is usually at the
lower border of the obturator fossae. The lateral volume border is 2 cm lateral to the
pelvic sidewall. The anterior volume border is 1 cm anterior to the uterine body on
MRI, ensuring adequate coverage of the common iliac nodes superiorly, and is usu-
ally at the anterior symphysis pubis. The posterior volume border is 1.5–​2 cm behind
primary disease, ensuring coverage of the proximal uterosacral ligaments and the in-
ternal iliac nodes. Small bowel and femoral heads may be shielded on the anterior and
posterior fields and the posterior sacrum can be shielded on the lateral films as shown
in Fig. 14.3.

Dose distribution
Conformal and conventional radiotherapy  The pelvis is treated with a three-​or four-​
field ‘box’ arrangement consisting of two lateral wedged fields, an anterior field, and a
posterior field if required. There is a low weighting of the posterior field to reduce the
rectal dose. This field arrangement achieves better homogeneity and tissue sparing,
even when there is posterior extension of the volume, compared to a two-​field ar-
rangement with anterior and posterior fields (Fig. 14.4).
The fields are designed to ensure coverage of the planning target volume by the 95%
isodose with a maximum dose of 107%. The fields are shaped to the 3D volume or the
shielding is applied from the simulator images. The use of higher energy photons (10–​
15 MV) improves superficial tissue sparing and achieves a better dose distribution.

Intensity-​
modulated radiotherapy (see Fig. 14.5) IMRT can significantly reduce
the dose to normal structures. Retrospective studies have shown that it can reduce
both early and late toxicity when compared to conventionally planned treatment in
294 Uterus: endometrium and cervix

(a)

PTV

Nodal CTV

Uterus cervix
vagina CTV

MLC shielding

Field edges

(b)

PTV

MLC shielding

Nodal CTV

Uterus cervix
vagina CTV

Field edges

Fig. 14.3  Anterior–​posterior DRR showing treatment fields with CTV (cervix, uterus upper
vagina and pelvic nodes) and PTV lateral DRR showing treatment fields with CTV (cervix,
uterus upper vagina and pelvic nodes) and PTV.
 Cervical cancer 295

Field Energy Weight Gantry Coll Couch Wedge X Y

1 Ant 15MV 0.36 0 0 0 18cm 20cm
2 Post 15MV 0.08 180 0 0 18cm 20cm
3 L lat 15MV 0.28 90 0 0 30 14cm 20cm
4 R lat 15MV 0.28 270 0 0 30 14cm 20cm

Fig. 14.4  Axial CT slice showing isodose distribution for conformal treatment of cervical
cancer. PTV (red) with bladder anteriorly and rectum posteriorly. Isodose curves yellow
100%, cyan 95%, green 75%, bright pink 50%, pale pink 20%.

historical controls. In addition, a simultaneous integrated boost technique can be used

to deliver a higher dose to involved lymph nodes.
If it is to be used in the treatment of primary cervix cancer, great care needs to be
taken in the delineation of the target in order to avoid compromising success with a
geographical miss. This is particularly important for the cervix and uterus which, be-
cause of bladder and rectal filling, can show interfractional movement of up to 30 mm
in the superior–​inferior direction and 15 mm in the anterior–​posterior direction.
Further studies with image guidance are required for ensuring safe implementation
of this technique.
The plan is optimized to cover the PTV with at least 99% volume receiving at least
95% prescription dose. Field arrangement options include a fixed field IMRT technique
using seven equally spaced fields or a VMAT technique which has equivalent OAR
sparing but uses fewer monitor units resulting in shorter machine treatment time.

Implementation on treatment machine
The patient is treated in the same position as for volume localization. The patient is
immobilized with knee or ankle supports, and aligned using laser lights to check the
position of the anterior and two lateral skin tattoos. The field centre is marked in rela-
tion to the anterior tattoo. All fields are treated isocentrically daily.
296 Uterus: endometrium and cervix

Fig. 14.5  Dose distribution of IMRT for primary cervix cancer using a VMAT technique
delivering 55 Gy to involved lymph nodes (PTV-​1) and 45 Gy to the uterus, cervix, and
involved nodal regions (PTV-​2). PTV-​1 (light blue volume) is encompassed by the red 95%
isodose while PTV-​2 (dark blue) is covered by the yellow 77.7% isodose, equivalent to
95% of 45 Gy.

There is a 12% absolute benefit in survival with the use of concomitant cisplatin
compared with radiotherapy alone, although the addition of chemotherapy does in-
crease normal tissue toxicity. Concomitant cisplatin is administered on a weekly basis
with the radiotherapy treatment delivered within 1 hour of completing the cisplatin
infusion.
The treatment course should be completed in the planned total number of days. The
overall treatment time is a further prognostic factor with decreasing survival associ-
ated with protracted treatment. Ideally the treatment course (including the brachy-
therapy component) should be completed within 42–​45 days and not prolonged over
49 days.
Unscheduled gaps are to be avoided and should be compensated for. This can be
done by treating at a weekend, treating two fractions in 1 day with an interfraction
interval of at least 6 hours, or by increasing the dose per fraction for the remaining
doses.

Verification
Electronic portal imaging device (EPID) or cone-​beam CT (CBCT) scans are taken
on the first, second, and third days of treatment and compared to the reference im-
ages. Depending on local policy, if the set-​up error is < 3–​5 mm and, if CBCT imaging
is used, target volume coverage is satisfactory, verification images are taken weekly.
Daily portal imaging before treatment, with subsequent field position adjustment if
 Cervical cancer 297

required, may be necessary for obese patients with poor set-​up due to highly variable
contours. Daily bladder ultrasound may be used to ensure consistent bladder volumes.
Further studies are required using daily soft-​tissue imaging to allow an adaptive or
‘plan of the day’ approach

Dose prescription
Primary disease
External beam radiotherapy 
◆ 45–​50.4 Gy in 25–​28 fractions over 5 to 6 weeks (1.8–​2 Gy per fraction) prescribed
to the isocentre or median PTV dose.
◆ Integrated boost to nodes to 58–​60 Gy (EQD2-​10).
◆ Weekly cisplatin 40 mg/​m2 (maximum 70 mg) for 5 weeks should be administered
concurrently.
◆ When intracavitary treatment is not possible, a second phase of external beam
radiotherapy is delivered to macroscopic disease with a small CT planned volume.
16–​20 Gy in eight fractions prescribed to the isocentre or median PTV dose.
Intrauterine brachytherapy
To be commenced during the last week of external beam radiotherapy or as soon as
possible after completing external beam radiotherapy. Image guided brachytherapy
enables dose escalation to macroscopic disease and cervix, defined as the high-​risk
CTV (HR-​CTV). The GEC-​ESTRO guidelines recommend cumulative organ at risk
dose limits (EQD2) for 2 cc bladder, rectum, and sigmoid of 90 Gy, 70–​75 Gy, and
70–​75 Gy respectively.
◆ High dose rate: 18–​28 Gy in three to four fractions prescribed to Point A or HR-​CTV.
◆ Low or pulsed dose rate: 27 Gy prescribed to Point A or HR-​C TV in a single
application.
Recurrent disease 
◆ Whole pelvis:  45–​50.4 Gy in 25–​28 fractions over 5–​6 weeks prescribed to the
isocentre or median PTV dose.
◆ Phase two: 16–​20 Gy in eight to 10 fractions over 2 weeks prescribed to the isocentre
or median PTV dose.

14.3.2  Postoperative adjuvant treatment for cervical cancer

Indications
The addition of adjuvant radiotherapy improves local control when the histology
shows high-​risk factors for pelvic recurrence, but increases normal tissue toxicity.
Absolute indications for adjuvant pelvic radiotherapy are:
◆ Presence of tumour at the resection margin.
◆ More than one node with metastatic infiltration or extracapsular nodal spread.
◆ Inadequate surgery such as an incidental finding at simple hysterectomy.
298 Uterus: endometrium and cervix

Relative indications that may require radiotherapy, particularly if more than one fea-
ture is present, include:
◆ Lymphovascular invasion.
◆ Deep stromal invasion.
◆ Poorly differentiated tumour.
◆ Invasive tumour < 5mm from the resection margin.
◆ A single lymph node with metastatic involvement.

Target volume and definition

The target volume includes the vaginal cuff and upper half of the vagina, paravaginal
tissue, and parametrium, and the external iliac, obturator, and internal iliac nodes for
all patients. The common iliac nodes are included for node positive patients.

Radiotherapy technique
The radiotherapy technique is the same as adjuvant radiotherapy for uterine tumours
as described in section 14.4.2.
Dose prescription
External beam radiotherapy 
◆ 45–​50.4 Gy in 25–​28 fractions delivered over 5 weeks prescribed to the isocentre or
median PTV dose.
◆ Weekly cisplatin 40 mg/​m2 (maximum 70 mg) for 5 weeks is administered concurrently.
Intracavitary vaginal vault brachytherapy  To be commenced after completing external
beam radiotherapy:
◆ High dose rate: 8–​11 Gy at 0.5 cm from applicator surface in two fractions.
◆ Low or pulsed dose rate: 15 Gy to 0.5 cm from applicator surface in a single application.

14.4  Uterine body tumours

The majority of uterine tumours arise from the endometrium and are predominantly
adenocarcinomas of endometrioid type. The less common subtypes of serous, clear
cell, carcinosarcoma (mixed Mullerian tumours), and squamous cell carcinoma have a
worse prognosis. Uterine sarcomas are rare and include leiomyosarcoma, endometrial
stromal sarcoma, and undifferentiated sarcoma.

14.4.1  Primary radical treatment for uterine tumours

Indications
Uterine tumours are ideally treated surgically with a total abdominal hysterectomy
and bilateral salpingo-​oophorectomy. However, it is possible to gain long-​term con-
trol with primary radiotherapy alone. Radiotherapy should only be used for primary
treatment if the patient is medically unfit for surgery or when the tumour is inoperable
due to local invasion.
 Uterine body tumours 299

Treatment volume and definition

The target volume for external beam radiation comprises the primary tumour, cervix,
uterus, upper vagina, ovaries, parametrium, and the pelvic lymph nodes. The common
iliac, external iliac, obturator, internal iliac, and subaortic presacral nodes should be
included for all patients. Macroscopic disease (GTV) will receive a higher dose with
either intrauterine brachytherapy or a second phase of external beam radiotherapy.

Radiotherapy technique
The radiotherapy technique is similar to the method for primary cervical cancer de-
scribed in section 14.3.1.
Dose prescription
External beam radiotherapy 
◆ 45–​50.4 Gy in 25–​28 fractions prescribed to the isocentre or median PTV dose de-
livered over 5–​5½ weeks.
◆ When intracavitary treatment is not possible, a second phase of external beam radio-
therapy is delivered to macroscopic disease with a small CT planned volume: 16–​20
Gy in 8–​10 fractions.
Intrauterine brachytherapy 
◆ High dose rate: 18–​24 Gy to surface of uterus in three to four fractions.
◆ Low dose rate: 27 Gy to surface of uterus in a single application.

14.4.2  Postoperative adjuvant treatment for uterine tumours

Indications
Over 80% of endometrial cancer presents as stage 1 disease and the overall survival
rates for this group are very high. Subsequent adjuvant treatment is based on the risk
of relapse which is determined by the following:
◆ Depth of myometrial invasion.
◆ Grade of tumour.
◆ Lymphovascular space invasion.
◆ Clear cell, serous or carcinosarcoma subtype.
Table 14.1 outlines adjuvant treatment by stage and risk group. For more advanced
tumours stage 2 and above, although it has not been shown to improve survival,
pelvic radiotherapy is indicated to reduce the incidence of loco-​regional recurrence.
If patients have undergone a systematic lymph node dissection with negative nodes,
vaginal vault brachytherapy can be recommended instead of EBRT. Adjuvant chemo-
therapy may also confer a survival benefit and should be considered for high risk
disease.
For the uterine sarcomas, pelvic radiotherapy has not been shown to provide a
survival benefit. It is, however, recommended for loco-​regional control where there is
extension beyond the uterus.
300 Uterus: endometrium and cervix

Table 14.1  Adjuvant radiotherapy treatment for endometrial cancer

according to risk of relapse
Risk Stage Treatment
Low 1A G1 G2 None

Intermediate 1A G3, 1B G1, 1B G2 Vault brachytherapy

High risk 1B G3, 2, 3, 4A External beam pelvic RT ±
brachytherapy

Treatment volume and definition

Most tumours are confined to the uterus at presentation. Endometrial cancer fre-
quently grows as a polypoidal mass that can ulcerate to cause bleeding. It invades dir-
ectly into the myometrium and can penetrate the serosa from where it can spread
into the peritoneal cavity or adjacent organs. Peritoneal metastases are particularly
common with serous histology. Tumours can spread inferiorly into the cervical canal
with invasion of the cervical glands and stroma.
Lymphatic spread from the middle and lower corpus follows a similar course to the
cervix, draining initially to the parametrial nodes, and then to external, obturator, and
internal iliac nodes. From the upper corpus, metastases occur more commonly to the
common iliac and para-​aortic nodes. The risk of lymph node metastases increases
with tumour grade and depth of invasion. The approximate incidence of pelvic lymph
node involvement is 6% in FIGO stage IA, 10% in stage IA (with myometrial invasion),
20% in stage IB, 20% in stage II, and 57% in stage III.
Haematogenous spread is uncommon at presentation for endometrial cancers al-
though it is a common method of spread for the uterine sarcomas, which frequently
metastasize to lung. A CT scan of the chest and abdomen is performed for high-​grade
tumours to exclude visceral metastases.

Clinical target volume for uterine tumours

◆ Vaginal vault.
◆ Upper half of vagina.
◆ Paravaginal tissues.
◆ External iliac, obturator, and internal iliac and proximal common iliac nodes (2 cm
inferior to aortic bifurcation).
Obviously enlarged nodes and lymphocoeles must be included.
Where pelvic nodes are involved the common iliac nodes should be encompassed
up to the bifurcation of the aorta or 2 cm above known disease whichever is higher.

Planning technique
Patient position and immobilization
The patient is immobilized supine with knee and ankle supports, and with their arms
on the chest. Skin tattoos are placed anteriorly and laterally.
 Uterine body tumours 301

Volume localization
CT localization The patient is immobilized in the treatment position with radio-​
opaque markers placed over the skin tattoos. A  planning CT scan of the pelvis is
taken with 2.5-​mm slices from the top of L3 to 5 cm below the ischial tuberosities.
Administration of intravenous contrast enhances visualization of the pelvic blood ves-
sels. The bladder is comfortably full in order to displace small bowel from the pelvis
and ideally the rectum should be empty.
◆ The CTV is outlined on each axial CT slice with a 7-​mm margin around the ves-
sels used as a surrogate for the nodal regions. The rectum, bladder, and bowel are
contoured.
◆ The PTV is created by adding a 12–​15-​mm margin around the vaginal vault and a
5–​7-​mm margin around the lymph nodes and parametria.
Simulator localization  The target volume is defined on orthogonal simulator images:
◆ The superior volume margin is approximately at the upper border of S1 to cover the
external iliac nodes or at upper border of the fifth lumbar vertebra to include the
common iliac nodes.
◆ The inferior volume border is the lower border of the obturator foramen. The lateral
volume border is 1.5 cm lateral to the pelvic sidewall.
◆ The anterior volume border is at the anterior symphysis pubis.
◆ The posterior volume border set at mid-​S2/​3 interspace. Shielding is applied as
shown in Fig. 14.6.

Isodoses
105 Field One
100
95
50
20

Field Three Field Two

Field Energy Weight Gantry Coll Couch Wedge X Y

1 15 MV 0.4 0 0 0 – 15 cm 16 cm
2 15 MV 0.3 90 0 0 45° 12 cm 16 cm
3 15 MV 0.3 270 0 0 45° 12 cm 16 cm

Fig. 14.6  Dose distribution with a three-​field conformal plan for endometrial cancer.
302 Uterus: endometrium and cervix

Dose distribution
Conventional and conformal radiotherapy  The pelvis is treated with a three-​or four-​
field ‘box’ arrangement consisting of two lateral wedged fields, an anterior field, and,
if required, a posterior field. The fields are shaped to the 3D volume or the shielding
is applied from the simulator images. The PTV is covered by the 95% isodose with a
maximum dose of 107%.
Intensity-​modulated radiotherapy  IMRT can significantly reduce the volume of small
bowel, bladder, and rectum within the PTV and initial clinical studies report a corres-
ponding reduction in both acute and late toxicity. Postoperatively, there is less organ mo-
tion within the target volume compared to primary cervical cancer although the vaginal
vault may still move 5–​12 mm antero-​posteriorly subject to bladder and rectal filling.
The dose distribution is inversely planned with at least 99% of the PTV covered by
the 95% isodose and < 1% of the PTV should receive > 105% (Fig. 14.7). The plan is
optimized to achieve minimal dose to the small bowel, rectum, and bladder. Field ar-
rangement is with either seven equally spaced static beams or with a VMAT technique.

Implementation on treatment machine
The patient is treated in the same position as for volume localization. The patient is
aligned using laser lights to check the position of the anterior and two lateral skin tat-
toos. The field centre is marked in relation to the anterior tattoo. All fields are treated
isocentrically daily.

Verification
EPID or CBCTs are taken on the first, second, and third days of treatment and com-
pared to the reference images. If the set-​up error is < 5 mm on all images, verification

Fig. 14.7  Dose distribution of IMRT for endometrial cancer using a seven-​field IMRT
technique (gantry angles 180°, 235°, 285°, 335°, 25°, 75°, 125°) with the yellow 95%
isodose fitting closely to the PTV.
 Para-aortic nodal irradiation 303

images are taken weekly. Daily portal imaging before treatment, with subsequent field
position adjustment if required, may be necessary for obese patients with poor set-​up
due to highly variable contours.

Dose prescription
External beam radiotherapy
◆ 45 Gy in 25 fractions or 48.6 Gy in 27 fractions prescribed to the isocentre or me-
dian PTV dose delivered over 5 weeks.

Intracavitary brachytherapy
To be commenced after completing external beam radiotherapy.
◆ High dose rate: 8–​11 Gy to 0.5 cm from the applicator surface in two fractions.
◆ Low or pulsed dose rate: 15 Gy to 0.5 cm from the applicator surface in a single
application.

14.5  Para-​aortic nodal irradiation

14.5.1 Indications
Patients with asymptomatic para-​aortic metastases have 5-​year survival rates of
approximately 30–​50% when treated with radical radiotherapy but long-​term con-
trol of symptomatic para-​aortic disease is uncommon. The total dose that can
be safely delivered is limited by small bowel, spinal cord, and kidney radiation
tolerances. There may be a role for debulking enlarged nodes with laparoscopic
resection followed by radiotherapy. In the presence of common iliac node disease,
there is a very high incidence of microscopic involvement of para-​aortic nodes
and in primary cervical cancer the para-​aortic nodes should also be covered in the
treatment field.

14.5.2  Treatment volume and definition

The para-​aortic, aorto-​caval, and retro-​caval nodes are included in the treatment
volume from the level of the aortic bifurcation up to the renal vessels, or 2 cm above
disease if this is higher. In patients with primary cervical cancer, the target volume
includes the pelvic disease as described in ‘Treatment volume and definition’ within
section 14.3.1. For adjuvant radiotherapy in endometrial cancer with para-​aortic
nodal involvement, the target volume also includes the pelvis as described above (see
‘Clinical target volume for uterine tumours’).

14.5.3 Planning technique
Patient position and immobilization
The patient is immobilized supine with knee or ankle supports. Arms are placed
on the chest or above the head to allow entry of lateral beams. Skin tattoos are
placed in midline and laterally at the level of the centre of the target volume to limit
rotation.
304 Uterus: endometrium and cervix

(a) Two-field conventional plan

(b) Four-field conformal plan

(c) VMAT plan

Fig. 14.8  Comparison of
planning techniques for
treating the para-​aortic
nodes. The PTV is shown
in red and the isodoses
are 95% yellow, 80%
green, 50% brown, 30%
black, and 20% blue.
 Para-aortic nodal irradiation 305

Volume localization
A planning CT scan is taken from the level of the top of the diaphragm to the pelvis
with 2.5-​mm slices. Administration of intravenous contrast enhances visualization of
the blood vessels and lymph nodes.
The radiation fields can also be defined by virtual simulation to cover the para-​aortic
nodal region with maximal sparing of the kidneys (Fig. 14.8). The superior field border
is the top of the first lumbar vertebra and the inferior border matches to the top of the
pelvic field.

14.5.4 Dose distribution
A more homogeneous dose distribution is achieved using a conformal technique
compared to virtual simulation. This can be done using a four-​field technique,
i.e. anterior and posterior fields with low weighted lateral fields. This helps to re-
duce dose to bowel and spinal cord compared to using anterior and posterior fields
(see Fig. 14.8). If the pelvis is also to be treated a single isocentre technique is pre-
ferred. The superior para-​aortic section also utilizes a four-​field technique with low
weighted lateral fields to ensure sparing of the kidneys and to avoid hot spots within
small bowel.
IMRT can be used to deliver an integrated boost to involved nodes, and achieves
greater bowel sparing when the pelvis is also treated. However, the kidney doses need
to be carefully reviewed as the beams will result in higher low–​moderate doses (10–​20
Gy) than with conformal radiotherapy (Fig. 14.8).

14.5.5  Implementation on treatment machine

The patient is treated in the same position as for volume localization. The patient is
aligned using laser lights to check the position of the anterior and two lateral skin tat-
toos. The field centre is marked in relation to the anterior tattoo. All fields are treated
isocentrically daily.

14.5.6 Verification
EPID or CBCTs are taken on the first, second, and third days of treatment and com-
pared to the reference images. If the set-​up error is < 5 mm on all images, verification
images are taken weekly.

14.5.7 Dose prescription
Radical treatment
◆ 45 Gy in 25 fractions prescribed to the isocentre or median PTV dose over 5 weeks.
◆ Integrated boost to involved nodes to 54–​58 Gy (EQD2).

Palliative treatment
◆ 30 Gy in 10 fractions over 2 weeks prescribed to mid-​plane dose.
306 Uterus: endometrium and cervix

14.6 Palliative treatment
14.6.1 Indications
A short course of radiotherapy can provide excellent palliation of pelvic pain and
bleeding for patients with disseminated disease and for patients who are medically
unfit for radical treatment.

14.6.2  Treatment volume and definition

In the palliative setting, the treatment volume should be kept as small as possible in
order to reduce toxicity. The volume consists of macroscopic disease only. There is no
indication for including the pelvic nodes.

14.6.3 Planning technique
Patient position and immobilization
The patient is immobilized supine with knee and ankle supports. Skin tattoos are
placed in midline and laterally.

Field localization
The patient is immobilized in the treatment position with radio-​opaque markers
placed. Administration of intravenous contrast may enhance visualization of the tu-
mour, but caution is needed if there is impaired renal function. The bladder is com-
fortably full in order to displace small bowel from the pelvis. Macroscopic disease is
outlined on each CT slice and the PTV created by adding a 10–​15-​mm margin.

14.6.4 Dose distribution
A field arrangement is selected to deliver a homogenous dose to the target volume
while sparing normal tissues. Typically, an anterior and two lateral wedged fields are
shaped to the target volume. If initiation of immediate treatment is necessary, opposing
anterior and posterior fields may be used.

14.6.5  Implementation on treatment machine

The patient is treated supine with a comfortably full bladder, immobilized with knee
or ankle supports and aligned using laser lights to check the position of the skin tat-
toos. The field centre is marked in relation to the anterior tattoo. All fields are treated
isocentrically daily.

14.6.6 Verification
EPID or CBCT should be taken on the first day of treatment and compared to the ref-
erence images, either the digitally reconstructed radiograph (DRRs) or the simulator
images. If the set-​up error is < 5 mm no further EPIDs are required as long as the clin-
ical set-​up is acceptable.
 Recommended reading 307

14.6.7 Dose prescription

Whole pelvis
◆ 30 Gy in 10 fractions to mid-​plane dose over 2 weeks.
◆ 20 Gy in 5 fractions to mid-​plane dose over 1 week.

CT planned volume
◆ 27–​30 Gy in 6 fractions prescribed to the isocentre treating twice weekly over
3 weeks.

Advanced disease for bleeding or pain control to small volume

◆ 15–​21 Gy in 3 fractions treating on alternate days.
◆ 10 Gy single fraction.

14.7  Management of patients undergoing

pelvic radiotherapy
Patients should be reviewed weekly during the course of treatment to assess acute tox-
icity. Acute gastrointestinal side effects include diarrhoea, abdominal cramps, rectal
discomfort, and bleeding. Review by a dietician may be necessary and a low-​residue
diet and loperamide help to control symptoms. Skin care advice is given as erythema
and desquamation may develop in the perineum, natal cleft, and groins. Urinary
frequency and dysuria occur due to radiation cystitis, although infection should be
excluded.
In cervical cancer, anaemia has an adverse effect on outcome with primary radio-
therapy and the average haemoglobin level during treatment is an independent prog-
nostic factor. It is important to maintain the haemoglobin level above 12 g/​dL with
blood transfusions if necessary.
Radiotherapy causes rapid ablation of ovarian function and hormone replacement
therapy should be considered upon completing treatment. Sexual function may be sig-
nificantly impaired and patients should be offered sexual counselling. The long-​term
regular use of vaginal dilators reduces the incidence of vaginal stenosis.

Recommended reading
ASTEC/​EN.5 Study Group, Blake P, Swart AM, et al. Adjuvant external beam radiotherapy
in the treatment of endometrial cancer (MRC ASTEC and NCIC CTG EN.5 randomised
trials): pooled trial results, systematic review, and meta-​analysis. Lancet 2009; 373:137–​46.
Bondar ML, Hoogeman MS, Mens JW, et al. Individualized nonadaptive and online-​adaptive
intensity-​modulated radiotherapy treatment strategies for cervical cancer patients based on
pretreatment acquired variable bladder filling computed tomography scans. International
Journal of Radiation Oncology, Biology, Physics 2012; 83:1617–​23.
308 Uterus: endometrium and cervix

Chemoradiotherapy for Cervical Cancer Meta-​analysis Collaboration (CCCMAC). Reducing

uncertainties about the effect of chemoradiotherapy for cervical cancer: individual data
meta-​analysis (Review). Cochrane Database of Systematic Reviews 2010; 1: CD008285.
Lim K, Small W, Portelance L, et al. Consensus guidelines for delineation of clinical
target volume for intensity modulated radiotherapy for the treatment of cervix cancer.
International Journal of Radiation Oncology, Biology, Physics 2011; 79:348–​55.
Mundt AJ, Mell LK, Roeske JC. Preliminary analysis of chronic gastrointestinal toxicity in
gynecology patients treated with intensity-​modulated whole pelvic radiation therapy.
International Journal of Radiation Oncology, Biology, Physics 2003; 56:1354–​60.
Nout R, Smit V, Putter H, et al. Vaginal brachytherapy versus pelvic external beam
brachytherapy for patients with endometrial cancer of high-​intermediate risk (PORTEC-​2);
an open-​label non-​inferiority randomized trial. Lancet 2010; 375:816–​23.
Pötter R, Haie-​Meder C, Van Limbergen E et al. Recommendations from gynaecological
(GYN) GEC ESTRO working group (II): concepts and terms in 3D image-​based treatment
planning in cervix cancer brachytherapy-​3D dose volume parameters and aspects of
3D image-​based anatomy, radiation physics, radiobiology. Radiotherapy Oncology 2006;
78:67–​77.
Salem A, Salem AF, Al-​Ibraheem A, et al. Evidence for the use PET for radiation therapy
planning in patients with cervical cancer: a systematic review. Hematology/​Oncology and
Stem Cell Therapy 2011; 4:173–​81.
Small W, Mell L, Anderson P, et al. Consensus Guidelines for delineation of clinical target
volume for intensity modulated pelvic radiotherapy in postoperative treatment of
endometrial and cervical cancer. International Journal of Radiation Oncology, Biology,
Physics 2008; 71:428–​34.
Taylor A, Powell M. An atlas of the pelvic lymph node regions to aid radiotherapy target
volume definition. Clinical Oncology 2007; 19:542–​50.
van de Bunt L, Jurgenliemk-​Schulz IM, de Kort GA, et al. Motion and deformation of the
target volumes during IMRT for cervical cancer: what margins do we need? Radiotherapy
and Oncology 2008; 88:233–​40.
 Chapter 15

Vulva and vagina
Peter Hoskin

15.1 Vulva
Carcinoma of the vulva is primarily a surgical disease best treated by wide surgical re-
section, radical vulvectomy, and inguinal lymph node dissection based on presenting
stage(1). Rarely, locally advanced primary disease may be presented for primary radio-
therapy treatment. Postoperative radiotherapy is recommended for tumours invading
> 7 mm in a vertical direction(2) and where surgical clearance cannot be achieved.
The first station regional lymph nodes in the inguinal region are best treated by rad-
ical surgical dissection(3), but fixed inoperable lymph nodes may benefit from primary
radiotherapy which may be followed where appropriate by surgery. Postoperative
radiotherapy should be considered for women having more than one node involved
with metastatic tumour at surgery(2) and where there is extracapsular extension. This
must be balanced against the increased risk of lymphoedema where both surgery and
radiotherapy are delivered to the groins.
Chemoradiation using cisplatin or 5-​FU/​mitomycin C-​based schedules has been
reported(4,5) but no randomized comparison with radiotherapy alone has been under-
taken; whilst high response rates are seen, there is a considerable increase in acute
toxicity.

15.1.1  Radical treatment of inoperable disease

Patient position and immobilization
Patient supine, arms by side, no specific immobilization.

Treatment volume
Small tumours of the clitoris or labia may occasionally be treated as local skin tumours.
The majority are squamous cell carcinomas although rarely basal cell carcinomas and
melanoma may be found. Treatment should follow the same guidelines as those de-
scribed in Chapter 19 using a gross tumour volume (GTV) to clinical target volume
(CTV) margin of 7 mm for squamous carcinoma and 5 mm for basal cell carcinoma
with an expansion from CTV to planning tumour volume (PTV) of 3 mm.
It should, however, be noted that even superficial squamous cell tumours in this
region have an incidence of inguinal lymph node involvement, approaching 10% for
a tumour invading 2  mm and 20% once there is 3-​mm depth of invasion(2). Given
the possible sampling error in estimating depth of invasion from a marginal biopsy,
310 Vulva and vagina

most patients should undergo lymph node surgery or receive prophylactic lymph node
irradiation also.
The primary tumour will be treated en bloc with the lymph nodes, the CTV
encompassing both inguinal lymph node regions and the entire vulva. If there are
positive lymph nodes or extracapsular extension then the volume is extended to in-
clude the internal and external iliac lymph nodes also.

Field localization
Computed tomography (CT) planning scans, 3-​mm slices with intravenous contrast
are used.
The CTV comprises the vulva and lower vagina, the inguinofemoral canal, and ex-
ternal iliac and internal iliac lymph nodes. The node regions should be defined based
on the standard lymph node atlases and as described earlier in Chapter 14 for uterine
tumours and in Chapter 16 for lymphomas.
If IMRT is to be used then the primary, nodal, and any boost volume to involved
nodes should be outlined separately, designated CTVp, CTVn, and CTVb respectively.
The planning target volume (PTV) will be derived by a volumetric expansion ac-
cording to local practice, typically of the order of 5 mm.

Dose distribution
Static five or seven field IMRT or VMAT should be used with a simultaneous inte-
grated boost (SIB) to nodal subvolumes where there is macroscopic involvement. An
example is shown in Fig. 15.1
If 3D-​CRT is the only technique available then anterior and posterior opposed
megavoltage beams are used; typically 6 MV will be sufficient unless the separation is
> 24 cm when higher energy beams may give a better distribution. Care is then needed
to ensure that the primary tumour in the vulva is not within the build-​up region of
higher-​energy megavoltage beams. Bolus may be required and in vivo dosimetry with
thermoluminescent dosimetry or diode measurements may be helpful.

Implementation and verification
Central and lateral tattoos are placed at CT simulation from which the plan set-​up can
be measured.
kV EPID images should be taken using bony anatomy to match to the planning
images. As a minimum this should be performed daily for the first 3 days to identify
systematic error and then weekly. Where there is repeated set up variation then daily
images can be justified.

Dose prescription
IMRT or VMAT
A dose of 45–​50 Gy in 25 fractions daily treating Monday to Friday is delivered for
prophylactic and post-​operative irradiation(6). Alternative dose fractionation sched-
ules in use include:
◆ 50.4 Gy in 28 daily fractions.
◆ 45 Gy in 20 daily fractions.
◆ 40 Gy in 15 daily fractions.
 Vulva 311

Fig. 15.1  IMRT dose distribution for post operative vulval cancer.

Where there is a SIB to the primary this should bring the total dose to an EQD2
(equivalent dose in 2 Gy per fraction) to > 60 Gy(6,7). This may be achieved delivering
58–​60 Gy in 25 fractions to the primary SIB.
Nodal subvolumes should receive a similar dose if this can be achieved within organ
at risk tolerance doses.
Higher doses to the primary volume will be achieved using brachytherapy and this
should be considered for all patients following 50 Gy rather than an SIB.
3D-​CRT
A dose of 45–​50 Gy in 25 fractions treating daily Monday to Friday is delivered for
prophylactic and post-​operative irradiation(6). Alternative dose fractionation sched-
ules in use include:
◆ 50.2 Gy in 28 daily fractions.
◆ 45 Gy in 20 daily fractions.
◆ 40 Gy in 15 daily fractions.
312 Vulva and vagina

A boost will then be required to the primary tumour and to palpable lymph nodes if
present to bring the total dose to > 60Gy EQD2:
◆ The boost CTV will encompass GTV with a 1-​cm margin.
◆ This may be treated by smaller anterior and posterior opposed megavoltage fields
or in the case of a relatively localized primary vulval or clitoral tumour by a direct
electron field provided the applicator can access the region satisfactorily.
◆ A direct electron boost may also be used for palpable node areas.
◆ The boost dose is 14–​16 Gy in seven to eight daily fractions to give a total dose of
64–​66 Gy in 32–​33 fractions.
◆ The same boost dose is recommended where the 15-​or 20-​fraction schedules have
been used for the phase 1 treatment.

15.1.2 Palliative treatment
Palliative treatment for locally advanced fixed fungating or bleeding tumours may be
indicated. These should be treated pragmatically with derivations of the previous tech-
nique. Doses for palliation may include the following:
◆ 21 Gy in three fractions treating three times weekly.
◆ 20 Gy in five daily fractions.
◆ 8–​10 Gy as a single dose.

15.2 Vagina
In contrast to carcinoma of the vulva, radiotherapy has an important role in the radical
treatment of vaginal cancer, particularly where organ preservation is an important consid-
eration for the patient, the alternative surgical approach often requiring total vaginectomy.
The results of treatment with radiotherapy suggest that patients with adenocarcinoma
and those with distal (lower third) vaginal lesions have a worse prognosis independent of
stage(8,9), but it is not clear that surgery necessarily gives better results in this group.
Small tumours localized to the vaginal mucosa (FIGO stage I) are well treated by
brachytherapy alone(10); detailed techniques are outside the scope of this chapter.
More advanced tumours involving submucosal tissues (Stage II) or fixed to the
pelvic side wall (Stage III) will require external beam therapy.
In the past using conventional 3D-​CRT treatment has required two phases, the first
to include the regional lymph nodes and the second to boost the primary site. It is now
usual to use static IMRT or VMAT solutions with a SIB.

15.2.1  Patient position and immobilization

Supine, arms by side.

15.2.2 Treatment volume
CT planning scans, 3-​mm slices with intravenous contrast are used. MRI using image
registration will give the best means of defining an accurate CTV.
The CTV will be different depending upon the position of the tumour, because of
the different lymphatic drainage along the length of the vagina. The lower third drains
 Vagina 313

to inguinal nodes and the middle and upper thirds to internal iliac and higher pelvic
nodes. Hence the CTV will be defined:
◆ Lower-​third tumours to include whole vagina and inguinal nodes.
◆ Middle-​and upper-​third tumours to include whole vagina, and internal, external,
and common iliac nodes.

Dose prescription
IMRT or VMAT
A dose of 45–​50 Gy in 25 fractions or 50.4 Gy in 28 fractions daily treating daily
Monday to Friday is required for prophylactic regions.
Alternative hypofractionated schedules may be considered for lower-​third tu-
mours as described for vulval cancer but for higher tumours where the PTV is a
large pelvic volume, including significant amounts of small bowel, these are not
recommended.
A higher dose is required to the primary and any clinically or radiologically positive
lymph nodes. An SIB to the primary should bring the total dose to an EQD2 (equiva-
lent dose in 2 Gy per fraction) to > 60 Gy(6,7). This may be achieved delivering 58–​60Gy
in 25 fractions to the primary SIB. (See Fig. 15.2.)
Nodal subvolumes should receive a similar dose if this can be achieved within organ
at risk tolerance doses.
Higher doses to the primary volume will be achieved using brachytherapy and this
should be considered for all patients following 50 Gy rather than an SIB. There is some
evidence that where a brachytherapy boost is possible then this results in better tu-
mour control than external beam alone, enabling total doses of > 75 Gy to be achieved
from the total combined treatment schedule(11,12).
3D-​CRT
A dose of 45–​50 Gy in 25 fractions daily treating Monday to Friday is delivered for
prophylactic and post-​ operative irradiation(6). An alternative dose fractionation
schedule in use is:
◆ 50.4 Gy in 28 daily fractions.
For localized tumours a brachytherapy boost will give the best means of delivering a
localized high dose of radiation to the primary site as discussed above.
If brachytherapy is not available then an external beam boost will be required to the
primary tumour and to palpable lymph nodes if present to bring the total dose to >
60 Gy EQD2:
◆ The boost CTV will encompass GTV with a 1-​cm margin.
◆ This may be treated by smaller anterior and posterior opposed megavoltage fields.
A direct electron boost should also be used for palpable node areas.
◆ The boost dose is 14–​16 Gy in seven to eight daily fractions to give a total dose of
64–​66 Gy in 32–​33 fractions.
One study has suggested that overall time is important in the radiotherapy of car-
cinoma of the vagina with a pelvic control rate of 97% when treatment was completed
within 63 days compared to 54% when the overall time was > 63 days(13).
314 Vulva and vagina

Fig. 15.2  IMRT plan to treat mid-​vaginal cancer.

15.2.3 Patient care
Acute skin and mucosal reactions are inevitable and should be managed conserva-
tively. There are two important late effects to be considered in preparing the patient
for radiotherapy:
◆ Sexual dysfunction related to fibrosis and vaginal shortening and narrowing is re-
ported in around one-​third of patients; the use of vaginal dilators may mitigate this
and should be encouraged from early after treatment once the acute reaction has
settled.
◆ The close anatomical relations with the rectum and bladder mean that fistulae are
a serious potential complication; reported series vary in incidence from 4 to 12%
related to tumour stage and dose.
 References 315

15.2.4 Palliative treatment
Palliative treatment for locally advanced fixed fungating or bleeding tumours or for
recurrent disease may be indicated.
Brachytherapy may deliver a localized high-​dose palliative treatment particularly
for intravaginal disease.
External beam treatment may be equally effective using limited anterior and pos-
terior opposed fields to the true pelvis defined on CT simulator.
Dose prescription
The following are in common use and effective; single doses may be most appropriate
for control of bleeding in advanced disease, three-​or five-​fraction schedules are appro-
priate for other local symptoms and in good performance status patients:
◆ 21 Gy in three fractions treating three times weekly.
◆ 20 Gy in five daily fractions.
◆ 8–​10 Gy as a single dose.

References
1. Woelber L1, Kock L, Gieseking F, et al. Clinical management of primary vulvar cancer.
European Journal of Cancer 2011; 47:2315–​21.
2. Sharma DN. Radiation in vulvar cancer. Current Opinion in Obstetrics and Gynecology
2012; 24:24–​30.
3. van der Velden J, Fons G, Lawrie TA. Primary groin irradiation versus primary groin
surgery for early vulvar cancer. Cochrane Database Systematic Reviews 2011; CD002224.
doi: 10.1002/​14651858.CD002224.pub2.
4. Shylasree TS, Bryant A, Howells RE. Chemoradiation for advanced primary vulval cancer.
Cochrane Database Systematic Reviews. 2011;CD003752. doi: 0.1002/​14651858.CD003752.
pub3.
5. Han SC, Kim DH, Higgins SA, et al. Chemoradiation as primary or adjuvant treatment
for locally advanced carcinoma of the vulva. International Journal of Radiation Oncology,
Biology, Physics 2000; 47:1235–​44.
6. Perez CA, Grigsby PW, Chao C, et al. Irradiation in carcinoma of the vulva: factors
affecting outcome. International Journal of Radiation Oncology, Biology, Physics 1998;
42:335–​44.
7. Bush M, Wagener B, Schaffer M, Duhmke E. Long term impact of post operative
radiotherapy in carcinoma of the vulva FIGO I/​II. International Journal of Radiation
Oncology, Biology, Physics 2000; 48:213–​18.
8. Chyle V, Zagars GK, Wheeler JA, et al. Definitive radiotherapy for carcinoma of the
vagina: outcome and prognostic factors. International Journal of Radiation Oncology,
Biology, Physics 1996; 35:891–​905.
9. Ali MM, Huang DT, Gopelrud DR, et al. Radiation alone for carcinoma of the
vagina: variation in response related to the location of the primary tumour. Cancer 1996;
77:1934–​9.
10. Mock U, Kucera H, Fellner C, et al. High-​dose-​rate (HDR) brachytherapy with or without
external beam radiotherapy in the treatment of primary vaginal carcinoma: long term
results and side-​effects. International Journal of Radiation Oncology, Biology, Physics 2003;
56:950–​7.
316 Vulva and vagina

11. Fine BA, Piver MS, McAuley M, Driscoll D. The curative potential of radiation therapy in
the treatment of primary vaginal carcinoma. American Journal of Clinical Oncology 1996;
19:39–​44.
12. Pingley S, Shrivastava SK, Sarin R, et al. Primary carcinoma of the vagina: Tata Memorial
Hospital experience. International Journal of Radiation Oncology, Biology, Physics 2000;
46:101–​8.
13. Lee WR, Marcus RB, Sombeck MD, et al. Radiotherapy alone for carcinoma of the
vagina: the importance of overall treatment time. International Journal of Radiation
Oncology, Biology, Physics 1994; 29:983–​8.
 Chapter 16

Lymphomas
Richard W Tsang, Mary K Gospodarowicz,
and Peter Hoskin

16.1 Introduction
Lymphomas are a heterogeneous group of malignancies with diverse pathologic fea-
tures, clinical course, and outcomes. They represent approximately 3% of all malignan-
cies worldwide with over 385,000 new cases of non-​Hodgkin lymphoma (NHL) being
diagnosed each year (Globocan 2012: http://​www-​dep.iarc.fr/​) and 200,000 lymphoma
deaths annually. In addition, over 65,000 cases of Hodgkin lymphoma (HL) occur each
year and 25,000 die of the disease. Unlike other malignancies, NHL is one of only few
cancers increasing in frequency. The cause for this is unknown.
The current standard for pathology classification of lymphomas is that adopted by
the World Health Organization(1). The WHO classification encompasses B-​cell, T-​cell,
and NK-​cell lymphomas as well as HL and myeloproliferative malignancies. Even
with modern immunohistochemical techniques, lymphoma disease entities are het-
erogeneous in outcome. This heterogeneity can be partly explained by genetic char-
acterization with the use of DNA microarrays that identify variable gene expression
profiles within morphologically similar subtypes. The most important prognostic
factors, other than the histopathologic type, in lymphomas include stage and the
presence of lymphoma-​associated systemic symptoms (B-​symptoms), patient age,
performance status, lactate dehydrogenase (LDH), and extent of extranodal involve-
ment. These factors comprise the International Prognostic Index (IPI) for diffuse large
cell lymphomas. A similar system has been developed for follicular lymphomas, the
follicular lymphoma IPI (FLIPI)(2). The extent of disease is described using the Ann
Arbor classification, based on the distribution and number of involved sites, the pres-
ence or absence of extranodal involvement, and B-​symptoms including unexplained
weight loss greater than 10% over previous 6 months, unexplained fever greater than
38 °C, or drenching night sweats. Tumour bulk is not part of this system despite its
prognostic importance. Modern pre-​treatment evaluation includes history and phys-
ical examination, complete blood count, renal and liver profiles, and bone marrow
biopsy and cardiac evaluation. LDH levels reflect disease bulk. β2-​microglobulin pre-
dicts response to treatment and time to failure. Staging investigations should include
computed tomography (CT) imaging of the neck, thorax, abdomen, and pelvis and
FDG PET (fludeoxyglucose positron emission tomography). For histologies that are
typically FDG avid, PET scan has high sensitivity and specificity for bone marrow
318 Lymphomas

involvement and has therefore replaced bone marrow biopsy in the initial staging
workup. Magnetic resonance imaging (MRI) helps to delineate extent of disease in
central nervous system (CNS) and bone.
A full description of the biology and management of all forms of lymphoma is be-
yond the scope of this chapter. Therefore, emphasis has been placed on the manage-
ment of the most common forms of NHL and HL, particularly those presentations
treated with radiation therapy (RT) or combined modality therapy (CMT).

16.2  Non-​Hodgkin lymphoma

16.2.1 Follicular lymphoma
Follicular lymphoma is one of the more common types of NHLs comprising about
20% of cases in Western countries. A  significant proportion of patients (20–​33%)
present with localized disease (stage I-​II)(3), although with PET scan there is some de-
gree of upstaging. Advanced stage follicular lymphoma is very responsive to therapy
but relapse is common. Although the interval between relapses decreases with time,
the median survival in most patients exceeds 7–​10 years. The optimal treatment of
disseminated or recurrent follicular lymphomas is controversial. The disease often
follows an indolent course with prolonged survival. Therefore, patients with small
disease burden and no symptoms are often managed with observation and deferred
treatment. Fewer than 50% of the patients in a selected series of observation needed
treatment within 6 years from diagnosis(4). To date there is no evidence that follicular
lymphoma can be cured with systemic treatment, although the response rates exceed
60% in advanced stage disease. The treatment options vary from watch and wait
strategy to aggressive chemotherapy followed by stem cell support.
Most patients with stage I-​II disease receive involved field radiation therapy (IFRT) or
more recently involved site radiation therapy (ISRT). Following IFRT alone patients with
stage I-​II disease have a 10-​year disease-​free survival of 40–​64% and median survival of
13.8–​15.3 years(4). In a series of 460 patients with stage I-​II follicular lymphoma treated
with IFRT alone, 98% achieved durable local control and the actuarial freedom from
relapse rate at 25  years was 42%. Therefore, stage I-​II follicular lymphoma is curable
with radiotherapy. The UK trial tested a total dose of 24 Gy vs 40 Gy in patients with
follicular lymphoma. No significant differences in outcomes were seen, but the com-
plete response rates were 82% and 79%(5), far lower than seen in other series. There is
little evidence supporting the routine use of adjuvant chemotherapy in patients with
stage I-​II follicular lymphoma treated with RT, although phase II trials combining ra-
diation and chemotherapy have shown promising outcome. IFRT or ISRT alone is still
considered as standard therapy. However, since over 50% of patients with stage I and II
follicular lymphoma relapse, CMT has been explored in a prospective randomized trial
by the Trans-​Tasman Radiation Oncology Group (TROG) with recent results showing a
reduced relapse rate with use of adjuvant CVP-​rituximab in addition to standard IFRT,
although no significant benefit to overall survival has been demonstrated to date(6).
The majority of patients with follicular lymphoma (70%) present with stage III or IV
disease. For patients with recurrent or advanced presentations, treatment options vary
from observation to aggressive chemotherapy.
 Non-Hodgkin lymphoma 319

◆ Observation alone is appropriate for asymptomatic patients, as active treatment has

not been shown to have an impact on overall prognosis in this group. However
current NICE guidelines recommend four weekly doses of rituximab in these pa-
tients as a more cost effective approach reducing the need for further interventions
(https://​www.nice.org.uk/​guidance/​ng52).
◆ Single-​agent chemotherapy with oral alkylating agents such as cyclophosphamide,
chlorambucil with or without prednisone, or a purine analogue fludarabine pro-
duces a survival rate equivalent to combination regimens.
◆ Monoclonal antibody-​based therapy is useful alone or in combination with chemo-
therapy. Rituximab is a chimeric human-​murine monoclonal IgG1 κ antibody that
binds the transmembrane phosphoprotein CD20 present on benign and malignant B
cells. It has moderate activity when used alone and is often used in combination with
chemotherapy. Bendamustine and rituximab is an effective combination. Rituximab
maintenance therapy (for 2 years) after a favourable response to front line chemo-
therapy has also been shown to improve progression-​free survival (PFS)(6).
◆ Two radiolabelled anti-​CD20 monoclonal antibodies, 131I-​tositumomab (Bexxar)
and 90Y-​ibritumomab tiuxetan (Zevalin) have been developed for the treatment of
indolent B-​cell lymphomas, mostly in the relapsed/​progressive disease setting.
◆ Radiotherapy has an important palliative role in the management of advanced fol-
licular lymphoma where localized enlarging nodes masses are symptomatic in the
context of stable disease elsewhere or chemotherapy resistance. Doses as low as 4 Gy
may produce complete responses in sites of relapsed disease(7).
◆ Follicular lymphoma is associated with a 3% per year risk of transformation to dif-
fuse large B-​cell lymphoma. To date, other than advanced staging, no definitive fac-
tors predictive of future transformation have been identified(8).

16.2.2  Marginal
zone lymphoma of MALT type-​mucosa
associated lymphoid tissue lymphoma
Mucosa associated lymphoid tissue (MALT) lymphoma was first described by Isaacson
and Wright in 1983(9) and is now widely accepted as a distinct disease entity. It com-
prises 7–​9% of all lymphomas(1).
◆ Between 60 and 70% of patients present with stage I or II disease. RT is often used
either initially, or at a later time in the course of the disease; aggressive surgical
management is not indicated, as RT has fewer side effects and preserves cosmesis
and normal tissue function.
◆ Although stage III and IV disease is currently not curable, the progress of disease is
usually extremely indolent. MALT lymphomas may progress locally, spread to re-
gional lymph nodes and/​or distant sites, or rarely transform to a diffuse large B-​cell
lymphoma.
After biopsy or excision of a MALT lymphoma, further treatment is generally re-
commended for residual disease, since untreated low grade MALT lymphoma
may eventually lead to recurrence with possible later transformation to diffuse
large B-​cell lymphoma. The most common presenting site of MALT lymphoma
is in the stomach. Gastric infection with Helicobacter pylori (H.  pylori) plays an
320 Lymphomas

important role in the aetiology of gastric MALT lymphoma. The molecular events
following H. pylori infection lead to the development of low-​grade gastric MALT
lymphoma and transformed MALT lymphoma. Following antibiotic treatment
of H.  pylori infection, complete regression of lymphoma occurs in 50–​80% of
cases(10). Recommended anti-​H. pylori therapy includes ranitidine or omeprazole,
clarithromycin, and amoxicillin for 7–​10 days. Regression of lymphoma generally
takes 5–​8 months, and may take up to 18 months. Patients with no H. pylori infec-
tion, with perigastric lymph node involvement, or with t(11; 18)(q21; q21) trans-
location usually do not respond to antibiotic therapy. In such cases, RT is very
effective in providing local disease control and cure. Local control rates are close
to 95–​100%.
The European Gastrointestinal Lymphoma Study Group (EGILS) has published
consensus-​based guidelines for investigation and treatment of gastric extranodal
marginal B-​cell MALT Lymphoma(11). Although consensus-​based, the evidence of a
number of recommendations is limited. In particular, the recommendations for endo-
scopic surveillance in patients in complete remission (CR) may be questioned.
Other common presenting sites of MALT lymphoma include orbital adnexae,
skin, salivary glands, and thyroid. The less common sites include bladder, cervix,
breast, lung, dura, and rectum. Orbital lymphomas arise in superficial tissues
including the conjunctiva and eyelids, or in deep tissues including the lachrymal
gland and retrobulbar tissues. Treatment is directed to cure, while preserving vi-
sion and the integrity of the orbit. Extensive surgery should be avoided. There have
been reports of an association with Chlamydia psittaci and hence the success of
antibiotic therapy in producing regression of lymphoma in some patients(12). The
overall actuarial 10-​year survival rates reported in the literature with radiotherapy
are 75–​80%. Most deaths are due to causes other than lymphoma. The risk of local
failure is extremely low. Contralateral orbit involvement is common either in syn-
chronous or metachronous fashion. Distant failure rates vary from 20% to 50%, but
as in other cases of indolent lymphoma, prolonged survival is observed. Although
the literature describing the outcomes in less common presentations of MALT
lymphoma is limited, available results indicate excellent local control rates. The
Princess Margaret Hospital experience showed 95% local control rate, 87% overall
survival, and 76% disease free survival at 10 years in 167 patients treated with in-
volved field RT(13).

16.3  Diffuse large B-​cell lymphoma

16.3.1  Stage I and II
◆ Radiotherapy alone:  historically, when patients with stage I-​II diffuse large cell
lymphoma were treated with RT alone the long-​term disease control was obtained
in 40–​50% of patients. Although a significant proportion of patients with stage I and
II aggressive lymphoma were cured with RT alone, failure rates were high and the
addition of chemotherapy resulted in better outcomes and has been the standard of
care for over 30 years.
 Diffuse large B-cell lymphoma 321

◆ CMT and radiotherapy:  the commonly used chemotherapy regimen is RCHOP

(rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone). With
CMT, excellent local control has been obtained with doses of 30–​35 Gy delivered in
1.75–​3-​Gy fractions over 2–​4 weeks. A phase III clinical trial demonstrated the effi-
cacy of a total dose of 30 Gy in 15 fractions with no benefit of a higher dose regimen
40–​45  Gy(5).
The question of the benefit of combined chemotherapy and radiation over chemo-
therapy alone has been addressed in several phase III trials, all completed in the
pre-​rituximab  era.
◆ The Eastern Cooperative Oncology Group (ECOG) trial of eight cycles of CHOP
vs eight cycles of CHOP followed by IFRT included 352 patients(14). The study in-
cluded stage I patients with bulky or extranodal disease, and stage II patients. Those
achieving complete response with CHOP were randomized to consolidation RT (30
Gy) or observation. With an intent-​to-​treat analysis, the 6-​year disease-​free survival
was 53% in the CHOP arm and 69% in the CHOP and RT arm (p  =  0.05). The
overall survival at 6 years was 67% for CHOP alone and 79% for CHOP and RT
(p = 0.23)(14). All patients with a partial response received RT to 40 Gy and their
6-​year failure free survival was 63%, similar to the patients achieving a complete
response to CHOP. This trial confirmed the benefit of IF RT who received CHOP
chemotherapy in terms of disease control. No overall survival benefit was evident
in this trial that was not powered to detect clinically important (10%) survival
differences.
The Southwest Oncology Group (SWOG S8736) study excluded stage II patients
◆

with bulky disease (tumour mass ≥10 cm). SWOG compared three cycles of CHOP
and IFRT to eight cycles of CHOP alone(16). The radiation dose was 40 Gy with a
boost to 50 Gy for partial responders. The 5-​year progression-​free survival rates were
77% for CHOP-​RT vs 64% for CHOP alone (p = 0.03), and 82% and 72% for overall
survival (p = 0.02)(15). The adverse risk factors included stage II disease, age greater
than 60 years, increased LDH and ECOG performance status of greater than 1. The
high rate of systemic failure has raised concern that patients with adverse prognostic
factors might have had inadequate chemotherapy in the CMT arm. Longer follow-​up
showed increased continuous failure rate in both the CHOP×3 + RT (15-​year PFS
40%) and chemotherapy alone (15-​year PFS 41%) arms(16).
A longer course of chemotherapy followed by RT may be optimal in patients pre-
senting with poor features, stage II disease, and rare or unfavourable extranodal sites
(bone, extradural, testes, etc.). Additional phase III trials from the French cooperative
group GELA have been reported(17,18). They did not show a benefit for the CMT arm
when compared to chemotherapy alone. Whether the addition of involved field RT
is of value in patients who obtain complete response with regimens more intensive
than CHOP +/​–​rituximab awaits further testing. A  phase III trial of RT (39.6 Gy)
vs no RT in adult patients below 60 years of age with IPI of 0 or 1 have been com-
pleted in Germany (UNFOLDER trial) and the final results are awaited. In primary
mediastinal large B-​cell lymphoma, phase II study showed that an intensive regimen
with dose-​adjusted EPOCH-​R results in excellent clinical outcomes, without routine
322 Lymphomas

consolidation RT(19). Currently the question of the benefit of adjuvant RT is being

tested in the phase III IELSG 37 trial where patients who achieve complete metabolic
response (FDG PET Deauville score 1–​3) are randomized to observation vs adjuvant
RT (http://​www.ielsg.org/​).
The principles of therapy of localized diffuse large B-​cell lymphoma and aggres-
sive lymphomas of other histologies (anaplastic large cell lymphoma, peripheral T cell
lymphoma) are similar. The choice of brief chemotherapy (3 cycles) followed by RT, a
full course of chemotherapy (6–​8 cycles) followed by adjuvant RT, or chemotherapy
alone is based chiefly on tumour bulk, presence of adverse prognostic factors such as
B-​symptoms and high LDH and the anatomic extent of disease. Rituximab is included
in all B-​cell lymphomas positive for CD20 antigen. CNS prophylaxis with intrathecal
methotrexate or cytosine arabinoside should be given to patients with testis lymphoma
and disease involving parameningeal sites. These principles are most important in
cases of primary extranodal lymphomas involving rare sites, where the available litera-
ture may not reflect the optimal approach.

16.3.2  Stage III and IV

In patients with stage III and IV lymphoma, who present with bulky disease, the
relapse is most frequent in sites of bulky disease at presentation. However, after a
complete response to chemotherapy, the role of consolidative RT for advanced dif-
fuse large B-​cell lymphoma has been controversial. Population database studies have
shown the use of RT is associated with improved clinical outcome(20,21), but should
be interpreted with caution as it is uncertain how well the confounding selection
factors for RT can be accounted for. Aviles et al, in a small randomized trial study
of patients with stage III and IV diffuse large cell lymphoma, showed improved
survival and local control for those who received adjuvant RT to sites of bulk dis-
ease following a complete response to chemotherapy(22). The German High Grade
Lymphoma Study Group had previously incorporated RT in their treatment pro-
grammes, for specific indications such as bulky disease (> 7.5  cm), or extranodal
sites, and found that RT is associated with better tumour control for older patients
with bulky disease(23), or when there was bone involvement(24). Others have found
similar results with better local control for bony involvement(25). Although it appears
tantalizing to use PET scan results to help determine the need for consolidative RT,
use of interim PET scan remains investigational at this time as it has been shown to
provide no additional discriminatory prediction over the standard approach of per-
forming an end-​of-​treatment PET.

16.4 Extranodal presentations
Extranodal involvement of lymphoma occurs in between 24% and 48% of new
lymphoma cases. Often presenting as localized disease, the sites are diverse and many
have unique clinical and pathological characteristics, and distinct biological behaviour
thereby requiring a different therapeutic approach as compared with nodal lymphomas
of similar histology. Frequent presentation with localized disease (stage IE or IIE) is of
special interest to the radiation oncologist. However, the majority of these diseases are
 Extranodal presentations 323

very rare and the literature to guide their treatment is limited. To accumulate the evi-
dence required for practice guidelines, the International Extranodal Lymphoma Study
Group has originated a number of retrospective and prospective trials to clarify the
management issues distinct to extranodal presentations. (http://​www.ielsg.org/​.) In
general, the management follows that recommended for a specific histologic disease
entity. However, several distinct extranodal presentations deserve attention.

16.4.1  Primary central nervous system lymphoma

The most common site of primary central nervous system lymphoma (PCNSL) is
the brain. Primary spinal cord lymphoma is less common, and primary leptomen-
ingeal lymphoma is rare, as is primary ocular lymphoma. Aggressive histology B-​cell
lymphomas are most common; T-​cell lymphomas have been reported, but are exceed-
ingly rare. Disease is characteristically multifocal or diffuse involving periventricular
areas with easy access to the cerebrospinal fluid. Meningeal seeding has been reported
to occur infrequently. The characteristic CT and MRI appearances include diffuse con-
trast enhancement of tumour, but a proportion of patients have non-​enhancing le-
sions. Treatment options include:
◆ Radiotherapy alone: traditional treatment consists of whole brain irradiation and
corticosteroids. Primary brain lymphoma is extremely sensitive to RT and cortico-
steroids, producing a rapid symptomatic response. The recommended doses are 40–​
50 Gy to the whole brain in 1.8–​2.0 Gy fractions. Performance status, age, multiple
lesions, deep location within the brain, high LDH, and elevated cerebrospinal fluid
protein level are the most important prognostic factors. The median survival has
been reported to be 12–​18 months with 2-​year and 5-​year survival rates of 28% and
3–​4% respectively(26,27,28).
◆ Because of the high risk of neurotoxicity with whole brain RT, the preferred ini-
tial treatment approach is high dose methotrexate given intravenously, sometimes
in combination with other drugs. One such regimen uses high dose methotrexate,
procarbazine, and vincristine followed by radiation 45 Gy and resulted in a me-
dian survival of 60  months(29). These and similar reports showed that PCNSL is
a chemosensitive disease. Some trials incorporate rituximab, and also high dose
chemotherapy with autologous stem cell transplant with the use of CNS-​penetrating
chemotherapy agents such as thiotepa and carmustine. However, methotrexate
chemotherapy combined with RT results in long-​term treatment related neuro-
logic toxicity. This is especially a problem in the older patients. In a series from
the Memorial Hospital, 11.5% of 1-​year survivors developed dementia(30). Presently
in those with a good response to chemotherapy, the RT dose is safely reduced to
23.4 Gy(31).
◆ A phase III German trial randomized 551 patients treated with high dose metho-
trexate to chemotherapy alone vs chemotherapy followed by high dose (45 Gy)
whole brain radiotherapy. There was no significant difference in the overall survival
and in median time to progression. In spite of several study limitations, treatment
with chemotherapy alone has become the preferred intervention for primary CNS
lymphoma(32).
324 Lymphomas

◆ Conclusions:
• younger patients with primary brain lymphoma receive chemotherapy alone
or combined chemotherapy including high dose methotrexate and whole
brain RT,
• patients over the age of 60 years receive chemotherapy alone or palliative RT.

16.4.2 Extradural lymphoma
Primary extradural lymphoma presents commonly with pain and progressive neuro-
logical deficit or spinal cord compression. Histological diagnosis is imperative and
biopsy is the first step in management. Historically, patients were treated with surgical
decompression followed by RT to the affected area of the spine. RT alone resulted
in excellent local disease control, but as with other localized aggressive histology
lymphomas, was associated with a 40–​50% distant failure. The use of CMT reduced
failure rate and an improved survival. In the Princess Margaret Hospital’s experience
the 5-​year survival of patients treated with RT alone was 33% compared with 86% for
those treated with CMT(33). Although the traditional approach was to deliver RT be-
fore CT, this may not be the optimal sequence. Eeles et al. documented that the use of
CT followed by RT does not compromise neurological function as compared to that
achieved when RT is followed by CT(34). The RT target volume should be carefully de-
fined using CT or MRI to include paraspinal tumour extension. A controversial aspect
of the management of primary extradural lymphoma relates to the use of CNS prophy-
laxis. In the Princess Margaret Hospital’s experience isolated CNS relapse in patients
treated without CNS prophylaxis was rare.

16.4.3 Cutaneous lymphomas
Numerous distinct lymphomas present with isolated cutaneous involvement, as de-
tailed by the EORTC(35,36,37). Primary lymphomas of the skin may be divided into three
broad categories:
◆ Cutaneous B-​cell lymphomas (25%) of indolent histologies with follicular lymphoma
and marginal-​zone (MALT), and the clinically aggressive:  diffuse large B cell
lymphoma of the legs.
◆ Cutaneous T-​ cell lymphoma of large cells (10%) including pleomorphic,
immunoblastic, anaplastic large cell lymphoma (CD30 + ), and rarely natural killer
(NK) cell lymphoma (CD56 + ).
◆ Cutaneous T-​cell lymphomas with indolent clinical behaviour (65%) including small
lymphocyte type (mycosis fungoides/​Sezary syndrome), lymphomatoid papulosis,
and some CD30 + large cell types.
Cutaneous large B cell lymphomas that occur in legs of elderly patients have aggressive
behaviour with a 5-​year survival of only 58%(35,37,38). In contrast, cutaneous follicular
lymphomas are usually confined to the head and neck region or the trunk, with 5-​
year survival of 97%. Infection with Borrelia burgdorferi has been implicated in the
pathogenesis of indolent cutaneous B-​cell lymphoma(39). RT is a preferred treatment
modality with very high local control rates of 85–​100% and favourable survival(37).
 Extranodal presentations 325

Although many patients eventually relapse, frequently with new skin lesions, death
from cutaneous B-​cell lymphoma is rare.
Primary cutaneous large T-​cell lymphomas are a heterogeneous group(1). However,
primary cutaneous anaplastic large cell lymphoma (C-​ALCL), which is positive for
CD30, is a specific clinical entity. In contrast to the systemic form of ALCL, it does
not express ALK protein and does not possess the t(2;5) translocation. In patients
with a solitary lesion or localized skin disease, RT is the treatment of choice. These
lymphomas relapse frequently in the skin, but generally have a favourable prog-
nosis(37,40). In a Dutch study of 79 patients, only 16% of patients had a systemic relapse
of lymphoma 10 years after initial treatment(40). Some patients demonstrated recurrent
self-​healing skin lesions. Durable spontaneous regression of even large lesions has been
observed in primary cutaneous anaplastic large cell lymphoma and it is reasonable to
observe lesions before embarking on treatment, depending on the clinical context.
Lymphomatoid papulosis is a related condition that has an indolent course, and usu-
ally follows a benign course with spontaneously remitting disease(40). Cytotoxic treat-
ment is usually not necessary and life expectancy is not adversely affected, although
infrequently the disease can progress to other types of T-​cell lymphomas. Patients with
T-​cell lymphoma negative for CD30 have a worse prognosis, with an estimated 5-​year
survival of 15%(35). Recent experience shows that brentuximab vedotin is a very useful
agent in CD30 + ALCL patients producing durable remissions(41,42).

16.4.4 Gastric lymphoma
Prior to the modern era of endoscopy and conservative management of gastric disease,
diffuse large cell lymphoma of the stomach was controlled with surgery and adjuvant
RT or chemotherapy. Currently, only chemotherapy followed by RT is used, while sur-
gical treatment has been largely abandoned. In the past, RT to the stomach exposed
the left kidney to radiation doses in excess of tolerance levels, However, the currently
available CT-​based RT and IMRT/​VMAT planning techniques including breath-​hold
techniques and image guidance allow for excellent protection of kidneys, liver, and
other normal organs.

16.4.5  Radiotherapy in recurrent lymphomas

For patients with relapsed or chemotherapy-​refractory large cell lymphoma, RT
alone is rarely curative. However in selected cases, durable long-​term control
or even cure can be achieved with salvage RT. For patients with chemotherapy-​
sensitive disease, a phase III trial tested the role of high dose therapy in 109 patients.
Those demonstrating a response to two cycles of DHAP (dexamethasone, cytosine
arabinoside, and cisplatin)chemotherapy were randomized to four further courses
of DHAP, vs a conditioning regimen of BEAC (carmustine, etoposide, cytarabine,
cyclophosphamide) followed by ABMT. RT was given as a protocol treatment for
patients with tumour bulk greater than 5  cm at the time of relapse in the trans-
plant arm (26 Gy in 20 fraction, twice daily), and in the conventional chemotherapy
arm (35 Gy in 20 fractions, daily). The 5-​year event-​free and overall survival for
the transplant arm were 46% and 53%, vs 12% (p = 0.001) and 32% (p = 0.038) for
326 Lymphomas

the conventional DHAP arm(43). There was a trend favouring the RT patients with a
lower relapse rate in the transplant group (8/​22 RT patients relapsed vs 18/​33 non-​
radiated patients relapsed, p = 0.19), and no obvious difference in the conventional
chemotherapy group (10/​12 RT patients relapsed, vs 35/​42 non-​radiated patients
relapsed). Although this was not a trial designed to examine the role of radiation
in the salvage setting, it lends support to the use of RT for bulky disease when in-
corporated into a salvage treatment plan that includes high dose therapy. The role
of RT in bulky disease following partial response to salvage chemotherapy deserves
further study in a randomized trial in patients undergoing haematopoietic stem cell
transplant. Until such evidence is available, we recommend routine RT to sites of
bulky disease, in sequence with salvage chemotherapy, and also RT to sites of in-
complete response to chemotherapy, for example those demonstrating persistent
fludeoxyglucose positron emission tomography (FDG-​PET) uptake following sal-
vage chemotherapy. Moderate doses of 30–​35 Gy should be the goal with individu-
alization of the treatment plan in regards to the exact target volume (involved-​site
RT is preferred), and the timing of RT in relation to chemotherapy and transplant to
facilitate the collection and harvesting of stem cells and minimize treatment-​related
toxicity. For example, if large RT fields are required, RT should be given after stem
cell harvesting and preferably pre-​transplant. Thoracic RT that will include signifi-
cant volumes of lung tissue may be better tolerated if given after transplantation.
In general, the principles of RT should be to treat the most likely site of relapse, or
progressive disease. The decision to treat with RT is influenced by the distribution
and location of the disease.

16.5 Hodgkin lymphoma
The management of HL evolved over the past four decades from that based on the
use of RT as the main curative modality to that relying on chemotherapy to cure
the disease. RT was shown to have curative potential in the 1950s. In the 1960s
MOPP chemotherapy was introduced. Staging laparotomy was used in the 1960s
and 1970s to identify patients with localized disease who have high probability of
cure with radiation alone. In the1980s clinical prognostic factors were shown to
be as good as staging laparotomy in identifying patients who required the com-
bined modality approach, i.e. the so-​called ‘risk adapted therapy’. In the late 1980s
ABVD chemotherapy was shown to be more effective than MOPP, with less tox-
icity. The risk of treatment-​induced leukaemia and infertility were substantially
reduced with ABVD. At the same time it became apparent that extended-​f ield RT
was associated with significant delayed morbidity, particularly second cancers and
heart disease.
In stage I-​II HL, the overall risk of death from the late effects of RT based treatment
strategies exceeded the risk of death from HL. To improve the overall survival, two
main treatment directions have been pursued:
◆ Short chemotherapy with low dose involved site radiotherapy (ISRT),
◆ Chemotherapy alone in low risk presentation, or adaptive approach based on re-
sponse assessed with functional imaging.
 Treatment of Stage I-II Hodgkin lymphoma 327

Currently, most patients with early stage HL are treated with ABVD chemotherapy
and ISRT and those with advanced disease are treated with ABVD alone or, if adverse
factors are present, BEACOPP chemotherapy. Many clinical trials are now in progress
and will lead to further improvement in the management of HL.
Treatment decisions are based on the Ann Arbor classification supplemented by
other prognostic factors and pathology. The anatomic extent of disease reflected by
Ann Arbor stage is the most important prognostic factor. Other factors known to influ-
ence the outcome in patients with HL include histological type, tumour bulk, number
of involved nodal regions and extranodal sites, age, gender, erythrocyte sedimentation
rate (ESR), B-​symptoms, anaemia, elevated white cell count, and lymphocytopenia.
Clinical trial groups combine these factors to create prognostic categories, and treat-
ment is limited or intensified accordingly. With individually tailored CMT, some of
these variables have less predictive value, as relapse is uncommon. However, the pres-
ence of a large mediastinal mass, B symptoms, or unexplained anaemia is associated
with poor prognosis. In advanced stage HL, factors identified to have independent
adverse effect on the outcome include:  male sex, age ≥ 45  years, stage IV disease,
haemoglobin < 105 g/​L, serum albumin < 40 g/​L, leukocyte count ≥ 15 ×109/​L, and
lymphocyte count < 0.6 × 109/​L (or < 8% of leukocyte count).

16.6  Treatment of Stage I-​II Hodgkin lymphoma

The management of early stage HL is largely dictated by the disease extent and prog-
nostic factors including histology, age, ESR, and the extent and bulk of disease. The
presence of one or more adverse prognostic factors categorize patients with higher risk
of treatment failure and hence the treatment is intensified. The strategy is to maximize
the therapeutic ratio by giving the minimal therapy necessary to achieve a high cure
rate with initial treatment, yet minimizing the long-​term side effects of therapy.
◆ CMT and radiotherapy vs radiotherapy alone: Several trials have demonstrated
that CMT improves disease-​free survival, compared to extended-​field radiotherapy
(EFRT) alone. This included the EORTC H7-​F trial with EBVP (epirubicin, bleo-
mycin, vinblastine, and prednisone), (44), the EORTC-​GELA H8-​F trial with three
cycles of MOPP/​ABV (mechlorethamine, vincristine, procarbazine, prednisone/​
adriamycin, bleomycin, vinblastine)(45), and the GHSG HD7 trial with two cycles of
ABVD(46). These trials show that in favourable-​risk early-​stage HL, the addition of
chemotherapy to RT improves disease-​free survival, and may also result in a small
improvement in overall survival. Additionally, a reduced RT volume was possible
following chemotherapy and this may reduce the late toxicity of treatment and im-
prove long-​term results.
◆ CMT radiotherapy vs chemotherapy alone: Recently randomized trials compared
chemotherapy alone with treatment that includes RT for patients with early stage
disease. The National Cancer Institute of Canada HD6 trial randomized patients
to ABVD alone or standard treatment including RT with or without two cycles
of ABVD depending on risk factors. Although 12-​year PFS was higher in the
RT containing arm (92% vs 87%), the overall survival was lower (12-​year overall
survival 87%) compared with ABVD alone (12-​year overall survival 94%), which
328 Lymphomas

was statistically significant (p  =  0.04)(47). A  Children’s Cancer Group study ran-
domized patients to risk-​adapted chemotherapy alone or chemotherapy followed
by IFRT and found superior 3-​year event free survival in favourable risk stage I-​II
patients receiving IFRT (97% vs 91%) with no difference in survival(48). Current ap-
proach uses FDG-​PET to select patients in metabolic CR after chemotherapy alone
(or doing an interim FDG-​PET scan after 2–​3 cycles) to decide on omitting RT
(risk-​adapted therapy). The UK RAPID trial results demonstrated the feasibility of
this approach as non-​bulky patients with metabolic CR after three cycles of ABVD
with no radiation had 3-​year PFS of 90.8%, while the addition of consolidative RT
gave a higher PFS of 94.6%(49). The early findings of EORTC H10 study appears
to confirm that interim FDG-​PET scan negative patients do have a small signifi-
cant relapse rate, which can be lowered by routine consolidative RT(50). The goal is
to have equally good tumour control rates as CMT, yet avoiding RT exposure and
hence eliminating the long-​term risks of radiation such as secondary malignancies,
heart disease, and thyroid gland dysfunction. It would be prudent to incorporate
other clinical factors into decision making for the use of chemotherapy alone, e.g.
age of the patient, and also whether the disease location and bulk requires a large
volume of breast tissue (in woman) or lung and heart tissue to be within the radi-
ation volume.
The current treatment recommendations are still based on known prognostic factors.
◆ Favourable risk classical HL: two to three cycles of ABVD are considered standard
when used with IFRT(51), and data from GHSG (HD10 trial) which compared two vs
four cycles of ABVD, and 30 Gy vs 20 Gy indicate that two cycles followed by IFRT
20 Gy is the standard approach[52], for those satisfying the low risk criteria of the
GHSG (no bulky mediastinal mass, ESR < 30 (with B symptoms) or < 50 (without
B symptoms), fewer than three nodal areas involved, and no extranodal disease).
The 5-​year PFS was > 90% in all four treatment arms, with 5-​year overall survival of
96–​97%(52).
◆ Unfavourable classical risk HL: more chemotherapy is required. The GHSG HD11
trial in stage I-​II disease and at least one adverse risk factor (as defined by GHSG
above) randomized patients to four cycles of ABVD, vs four cycles of BEACOPP
followed by IFRT of either 20 Gy or 30 Gy(53). The 5-​year PFS and overall survival
were 87% and 94% respectively, and the trial affirmed ABVD × 4 followed by 30 Gy
as the standard approach.
◆ Stage I/​II nodular lymphocyte-​predominant HL (NLPHL): Accounting for 5% of
HL, NLPHL is characterized by male predominance, and typically a stage I pres-
entation with involvement of a peripheral lymph node region, for example neck,
axilla, or inguinal location. It has distinct pathological features of nodular archi-
tecture, variant RS cells (L & H cells), and in contrast with classical HL, is CD20 +
and CD30-​. RT alone is the standard approach. Large series of patients treated with
RT alone using IFRT showed long term PFS of approximately 90%(54–​57). There is a
tendency for late relapse in this histology. There has not been a randomized com-
parison of RT vs CMT specifically for NLPHL, although one retrospective study
showed that CMT (with 2 cycles of ABVD) may improve the long term results(58). In
 Hodgkin’s lymphoma Stage III and IV 329

rare cases where all disease has been removed by excisional biopsy, observation is a
viable option.

16.7  Radiation therapy volume and dose

Modern chemotherapy provides adequate control of microscopic disease, eliminating
the need to irradiate clinically uninvolved nodal regions. A randomized trial of pa-
tients with stage I-​II disease and one unfavourable risk factor compared four cycles
of ABVD plus either STNI or IFRT. No significant differences in PFS (97% vs 94%),
or overall survival (93% vs 94%) were observed. Similarly, the GHSG HD8 trial ran-
domized patients to EFRT or IFRT after four cycles of chemotherapy and found no
difference in freedom from failure (EFRT 94%, IFRT 92%), or overall survival (97%
for both arms)(59). These excellent results, and the desire to reduce RT volumes, have
made IFRT (and now ISRT) following chemotherapy the standard of treatment in
stage I-​II HL. Recent clinical trials focus on further reducing RT volumes, based on
the concept of not irradiating whole lymph node ‘Kaplan’ regions, but covering the
actual extent of the initially involved lymph nodes. This is termed involved node ra-
diation therapy (INRT), and this approach has been adopted by the EORTC H10
study(50), and the GHSG HD17 study(60) testing the role of FDG-​PET in treatment
adaptation. INRT up to a 5-​cm margin has been shown in one retrospective study
to have similarly low local/​regional nodal disease recurrence rates (< 5%)(61). A more
strictly adhered INRT policy when used in CMT in Copenhagen in 93 patients re-
sulted in 4-​year PFS of 94%, and did not result in excessive relapse outside of radi-
ation fields(62).
The optimal RT dose following ABVD chemotherapy appears to be 20 Gy for fa-
vourable stage I-​II HL based on GHSG HD10 criteria(52), and preliminary results
of the EORTC H9F study (unpublished), and 30 Gy for unfavourable or initially
bulky presentations(53). It remains uncertain if FDG avid disease post chemotherapy
may require a higher dose for optimal local control, although it is common prac-
tice to give a boost dose to 35–​40 Gy to residual masses particularly if it remains
FDG avid.

16.8  Hodgkin’s lymphoma Stage III and IV

Advanced Hodgkin’s lymphoma is a heterogeneous disease. The outcomes vary from a
5-​year survival of 80% in patients without adverse prognostic factors to a 50% 5-​year
survival in those with several adverse prognostic factors. The prognostic factors iden-
tified to have an adverse effect include male sex, age 45 years or more, stage IV disease,
haemoglobin less than 105 g/​L, serum albumin less than 40 g/​L, leukocyte count 15 ×
109 /​L or greater, and lymphocyte count less than 15 × 109 /​L (or less than 8% of leuko-
cyte count). The standard treatment for stage III and IV HD is chemotherapy with the
ABVD regimen. The 5-​year freedom from treatment failure ranged from 69%–​87%
in recent studies, with overall 5-​year survival approximately 85%–​90%. Alternatively
low-​risk patients may be managed with Stanford V regimen combined with involved
field RT. (63). A  more intensive regimen, BEACOPP, has shown an advantage over
330 Lymphomas

COPP-​ABVD or ABVD for both event free survival and overall survival(64). Current
clinical trials focus on risk-​adapted approaches by stratifying patients according to the
number of these prognostic attributes present at diagnosis, and response as assessed
by FDG-​PET scans. Interim PET scan assessment after two cycles of ABVD has been
shown to be helpful in a phase III trial, as rapidly responded patients with a nega-
tive interim PET can have bleomycin deleted from subsequent cycles of chemotherapy
(AVD), and PET positive patients can be considered for escalation of chemotherapy
with BEACOPP(65).
The role of RT in patients with advanced stage HL is limited. A meta-​analysis of
prospective randomized trials has shown that consolidation radiation does not im-
prove overall survival in patients with advanced stages HL(66), despite a 11% improve-
ment in tumour control as compared to the same chemotherapy given alone. In most
of these studies, MOPP or MOPP-​like chemotherapy was used. A study from India
indicated that radiation may improve outcome when added to six cycles of ABVD.
The 8-​year overall survival was 100% with RT, vs 89% without RT(67). The interpret-
ation of this study is hampered by the inclusion of early stage patients and very young
patients. The UK LY09 study comparing ABVD with two other multidrug regimens
used consolidation RT for incompletely responding disease, and those with bulky
disease. Radiated patients had better tumour control: 5-​year PFS was 86% with RT,
and 71% without(68). Other studies have supported the findings of the meta-​analysis
showing no benefit from adding radiation to chemotherapy regimens similar to
ABVD(64,69,70,71). In a EORTC study pre-​PET era, patients obtaining CR after four or
six courses of MOPP/​ABV were randomized to no further treatment or consolidation
with involved field radiation after receiving a total of six and eight cycles, respect-
ively(69). The 5-​year event free survival rates were 84% and 79% (p = 0.35) and the
overall survival rates were 91% and 85% (p = 0.07) in the non-​radiated and the radi-
ated groups, respectively. Phase III trials by the ECOG and the UK NCRI Lymphoma
group compared the use of Stanford V regimen that includes short chemotherapy
followed by radiation to ABVD +/​-​RT for bulk disease(63, 72). There was no survival
advantage for either arm, and at present ABVD alone represents a standard approach
to stage III/​IV HL.
Current approach would attempt to define the remission status more accurately with
FDG-​PET and consider the use of consolidation RT only in those with residual FDG
activity. This has been the approach used by the GHSG HD15 trial, where following
BEACOPP chemotherapy, the use of RT can be restricted to only 11% of all patients,
with indication for its use in those who have a residual mass (> 2.5 cm) which remain
FDG-​PET avid in a localized site.
Summary:
◆ In stage III and IV HL presenting without bulk disease (node >10 cm or mediastinal
mass > 1/​3 transthoracic width) there is no role for RT in patients who achieve com-
plete metabolic response with standard chemotherapy
◆ Consolidation RT following definitive chemotherapy in stage III and IV disease
should be considered for those with initial bulk disease, which show residual FDG
activity.
 Radiation therapy techniques in lymphomas 331

16.9  Treatment for relapsed or refractory

Hodgkin lymphoma
The management of relapsed HL depends on patient age, performance status, and ex-
tent of disease at relapse, and prior treatment. For patients with good performance
status and chemotherapy-​sensitive disease, randomized trials have demonstrated that
high dose chemotherapy with autologous stem-​cell transplantation (ASCT) is more ef-
fective than aggressive conventional chemotherapy alone. This treatment can also pro-
duce durable responses among those not responding to first-​line chemotherapy. The
role of RT in a salvage treatment strategy that includes ASCT is important for nodal
disease, particularly if localized, bulky sites > 5 cm, and residual disease despite sal-
vage chemotherapy. FDG-​PET disease status prior to ASCR has prognostic value(73,74).
For patients with mediastinal recurrences, RT following ASCT produces less lung tox-
icity than pre-​transplant RT. In patients with more advanced disease at relapse, RT
is not likely to influence survival. Novel drug treatment with brentuximab vedotin
(antiCD30 conjugated to microtubule toxin) can result in a high overall response
rate (~ 75%)(75), and the checkpoint inhibitor drugs (PD1 inhibitor, e.g. nivolumab)
have been recently shown to have a high response rate as well. In patients who have
predominately local or regional symptomatic disease, RT may be considered with or
without chemotherapy. Several studies demonstrated durable responses to RT among
patients who relapsed following chemotherapy, although the intensity of the initial
chemotherapy in these studies was not as great as in current practice. The prognosis
in patients with relapsed HL is related to the disease-​free interval (better prognosis if
the disease-​free duration was > 12 months), extent of disease at relapse, performance
status, presence of anaemia, and response to salvage chemotherapy Some patients with
relapsed HL may survive with disease for prolonged periods of time, and this should
be considered when RT is given for palliation.

16.10  Radiation therapy techniques in lymphomas

There is paucity of level I evidence to inform the technical aspects of radiotherapy in
lymphomas. However, there are several decades of carefully documented experience
with excellent local control rates and well documented early and late toxicity data. It
is unlikely that there will be new prospective trials addressing solely the specific issues
of RT techniques. Therefore, the radiation oncology experts in lymphoma formed the
International Lymphoma Radiation Oncology Group (ILROG) to produce consensus-​
based practice guidelines to provide instruction for the uniform application of modern
RT technologies(78,79,80). Very similar guidelines have been published under the auspices
of the UK National Cancer Research Institute Lymphoma Radiotherapy group(81,82).
The technical aspects of treatment planning for lymphomas are highly dependent on
the location and extent of the target volume. In most cases established techniques for
cancer of other anatomic sites can be adapted to lymphomas. In general, planning in-
volves the appropriate use of immobilization devices, simulation, CT-​assisted tumour
localization and planning, localization of adjacent normal tissues, custom-​designed
beam modifying devices, and computerized calculation of isodose distributions. The
332 Lymphomas

goal of these steps is to achieve dose uniformity in the target volume while minimizing
RT dose to normal tissues.

16.10.1  Nodal irradiation: general considerations

The historical literature describes the extent of RT in terms of IFRT, EFRT, and total
lymphoid (or nodal) RT (TLI or TNI). Subtotal nodal irradiation (STNI) is a spe-
cific type of EFRT previously used for supradiaphragmatic HL when RT alone is pre-
scribed. More recently involved node irradiation (INRT) and involved site irradiation
(ISRT) have evolved. Definitions are provided here for comparative and purposes.
◆ INRT defines RT to the clinically involved lymph node(s) only, without intentional
coverage of the uninvolved lymph nodes in the same or adjacent nodal region and
with no safety margin to allow for variations in imaging and image fusion. In the
CMT context where there has been accurate determination of initial extent of dis-
ease with CT scan and PET scan, with adequate image fusion to the post chemo-
therapy planning CT scan, ISRT is the same as INRT with an additional 15-​mm
margin superiorly and inferiorly and 5 mm laterally.
◆ IFRT defines RT to the clinically involved lymph node region(s), with or without
coverage of the first echelon adjacent lymph node region uninvolved by disease.
Lymph node regions are defined by Kaplan and Rosenberg and have been used for
over 30 years. This has been replaced by ISRT practice in the last few years.
◆ EFRT defines RT to include the adjacent first echelon and the second echelon adja-
cent lymph node regions.
◆ TLI implies treatment to all the major lymphoid regions, with the mantle and the
inverted Y fields, with or without Waldeyer’s ring fields. It was used mainly in the
management of HL. The mantle field includes all major lymph node regions above
the diaphragm including cervical, supraclavicular, axillary, hilar, and mediastinal
lymph nodes treated in contiguity.
◆ STNI includes the mantle field, spleen, and abdominal para-​aortic lymph nodes.

16.10.2  Involved site radiation therapy

As the use of IFRT implied the assignment of radiation portals based on anatomic
landmarks, and almost always encompassed whole nodal regions, it is no longer the
standard RT planning process in the majority of modern RT centres. Current practice
utilizes 3D or 4D planning, with CT scans and additional imaging including PET and
MR imaging datasets, where applicable and follow the ‘ “involved site’ radiation therapy
(i.e. ISRT) concept. Guidelines of the ISRT practice have been published for HL(79), and
nodal NHL(78,81), as well as extranodal NHL(80,82). The essence of the concept is that
in the CMT setting, only the initial gross disease involvement (pre-​chemotherapy)
requires RT coverage, as potential microscopic disease in adjacent lymph nodes or
distant sites would be sterilized by chemotherapy. At the time of planning, as gross dis-
ease has regressed substantially, it is important to modify contours of the clinical target
volume taking account of the regression as well as displacement of normal tissues such
that organs at risk can be optimally spared. An example is a bulky mediastinal mass
that did not infiltrate lung tissue and is smaller in size following chemotherapy, where
 Planning techniques 333

Fig. 16.1  Involved site radiotherapy CTV for treatment to left submandibular node; CTV
in blue, PTV in green.

the RT plan need only cover the post-​chemotherapy width of mediastinum and not
include healthy lung. If, however, the disease was infiltrative initially into adjacent
normal tissue, regression of the tumour mass may leave microscopic residual disease
in the infiltrated tissue and consideration must be given to adequately cover initial dis-
ease extent. Use of image fusion using pre-​chemotherapy scans (including FDG PET)
and fusing to the planning CT scans is encouraged. Optimal performance of fusion
requires that patient’s positioning for pre-​chemotherapy scans are similar to that of
the post-​chemotherapy CT simulation scan. It is paramount to have accurate deter-
mination and documentation of the initial anatomic extent of disease, prior to starting
chemotherapy. Terminology for targets should follow ICRU report 83, using termin-
ologies of gross tumour volume (GTV), clinical target volume (CTV), internal target
volume (ITV), and planning tumour volume (PTV). Where appropriate, methods to
deal with motion control for targets/​organs, or respiratory motion should be imple-
mented, for example, active breathing control with inspiratory breath-​hold for me-
diastinal radiation to minimize dose to lungs and heart. An example of ISRT to the
axilla with a significant residuum is shown in Fig. 16.1 and to the groin after complete
response to chemotherapy in Fig. 16.2.
In lymphomas managed with radiation alone, the available data is based on including
the entire nodal region in an involved field (IFRT). There is no consensus currently as
to whether these patients can be safely treated with lesser volumes as in ISRT.

16.11 Planning techniques
16.11.1  Supradiaphragmatic lymph node irradiation
Patient position and immobilization
◆ Supine: hands by side or where axilla is to be irradiated abducted hands on hips or
supported on an arm board or T bar as for breast irradiation
◆ Immobilization is required for neck irradiation using an appropriate shell; for the
mediastinum immobilization is not usual but some centres advocate the use of a
vacuum body bag to avoid lateral rotation. CTV is determined on planning CT.
◆ Deep inspirational breath-​hold and other 4D techniques should be employed at ac-
quisition of planning images.
334 Lymphomas

Fig. 16.2  Involved site radiotherapy CTV for post chemotherapy treatment to the right
groin following partial response.

Field arrangement
Field arrangements should be optimized to minimize organ-​at-​risk (OAR) doses; anterior
and posterior opposed fields shaped to avoid critical structures may still be a good solution
in the neck or axilla as shown in Fig. 16.3, but in the mediastinum more complex arrange-
ments using IMRT or VMAT together with deep inspirational breath-​hold. A particularly
elegant method minimizing heart and lung exposure is ‘butterfly VMAT’ as shown in Fig.
16.4. Extranodal sites will be treated using techniques applicable to that site.

Dosimetry
Prescribed dose of radiation should be delivered to the target volume with no more
than -​5% and +7% dose variation within the target for conventional field arrange-
ments and conforming with ICRU82 for IMRT or VMAT. Normal tissue volumetric
constraints are defined as shown in Table 16.1.
 Planning techniques 335

Fig. 16.3  Dose distribution for ISRT to left submandibular node.

Dose prescription
◆ Non-​Hodgkin lymphoma:
• 30 Gy in 15 fractions is considered adequate for consolidation after complete response
to chemotherapy in diffuse large B-​cell lymphoma and other high grade tumours.
• 24 Gy in 12 fractions is adequate for follicular lymphoma and low grade MALT
lymphoma.
• NK cell lymphoma should receive at least 50 Gy in 25 fractions

Fig. 16.4  Graphical representation of arcs and beams configuration for butterfly VMAT.

Reprinted from Fiandra et al., ‘Different IMRT solutions vs. 3D-Conformal Radiotherapy in early
stage Hodgkin’s lymphoma: dosimetric comparison and clinical considerations’, Figure 1, Radiation
Oncology, 2012 Volume 7, Issue 186, DOI: 10.1186/1748-717X-7-186, Copyright © Fiandra et al.;
licensee BioMed Central Ltd. 2012. Printed under the Creative Commons Attribution License 2.0
(http://creativecommons.org/licenses/by/2.0).
336 Lymphomas

Table 16.1  Normal tissue volumetric constraints

OAR Limiting Dose /​Volume
Breast Minimize volume inside PTV, particularly in young women

Heart Maximum dose of 40 Gy to any part of the volume

Kidney V40 of 40%, if single kidney irradiated V15 of 65–​70%, if both V15 of 20–​
25% for each kidney
Larynx Maximum dose of 20 Gy to any part of the volume
Lens Maximum dose of 10 Gy to any part of the volume
Liver V40 of 30–​35%, D100 of 20 Gy
Lung (whole) V20 of 35–​40%
Ovary Maximum dose of 10 Gy to any part of the volume outside PTV. If inside
PTV discuss individual case with clinician
Parotid Maximum dose of 32 Gy to any part of the volume outside the PTV of the
contra-​lateral parotid
Spinal cord Maximum dose of 40 Gy to any part of the volume
Testis Maximum dose of 2 Gy to any part of the volume

• Palliative radiotherapy in lymphomas often achieves local tumour control and

therefore similar doses are recommended as for radical RT.
• Follicular lymphomas are more sensitive to radiation. For palliation, doses as low
as 4 Gy in two fractions in 3 days have been shown to be effective and result in
local disease control.
◆ Hodgkin lymphoma:
• Favourable disease after chemotherapy: 20 Gy in 10 daily fractions
• In all other situations: 30 Gy in 15 daily fractions.

Implementation and verification
Where large fields to mediastinum and neck are used, careful set-​up is required
to ensure reproducibility from day to day: this is assisted by using more than one
permanent set-​up mark (tattoo), typically two along the central axis of the field,
one in field centre, and two lateral tattoos are defined in addition to usual skin
marks.
Verification with IGRT is mandatory.

16.11.2  Infradiaphragmatic lymph node irradiation

Patient position and immobilization
◆ Supine: hands by side.
◆ Immobilization is not usual but some centres advocate the use of a vacuum body
bag to avoid lateral rotation. CTV is determined on planning CT.
 Planning techniques 337

Localization
CT-​based planning with where appropriate CT PET fusion is essential to ensure not
only accurate CTV definition but also the important organs at risk including the kid-
neys, liver, and spinal cord.

Field arrangement
Field arrangements should be optimized to minimize OAR doses:  anterior and
posterior opposed fields shaped to avoid critical structures may still be a good so-
lution in the para-​aortic region and groin but more complex arrangements using
IMRT or VMAT shaped to avoid critical structures, e.g. kidneys and bowel, should
be used.

Dosimetry
CT-​based planning 3D dose distributions will be generated to optimize dose distribu-
tions. Particular attention to kidney doses is important since unlike most situations
in lymphoma radiotherapy the doses delivered may be ablative to renal parenchyma.
This is a particular issue for splenic fields where the proximity of the left kidney may
require compromises to the renal dose to ensure complete coverage of the spleen. In
this setting, it is essential to have performed an isotope renogram to be certain that
the contralateral kidney has adequate function if the left kidney is sacrificed. Refer to
Table 16.1 for recommended tolerance doses.

16.11.3 Extranodal sites
In stage I extranodal disease, the CTV is limited to a margin around gross disease.
Exceptions to this include MALT lymphomas, known for multifocal involvement
where the whole organ or area at risk is treated:
◆ Gastric MALT, the CTV includes the entire stomach as shown in Fig. 16.5
◆ Orbital MALT lymphoma, the CTV includes the whole orbit, except in limited con-
junctival involvement, where recent data from small series of patients indicate par-
tial orbit coverage is safe(83).

Orbit
Traditionally orbital lesions are invariably controlled with low to moderate doses of
20–​30 Gy in 10–​20 daily fractions, with local control rate in excess of 95% for indolent
histologies such as MALT or occasionally follicular lymphoma.

Head and neck
Waldeyer’s ring: previously, the involvement of any component of Waldeyer’s ring (lin-
gual tonsil, oropharyngeal tonsil, nasopharynx) led to treatment of the entire ring, but,
currently, this is not practiced. Certainly in the combined modality setting only the
involved site is treated. CT imaging and planning should be used for greater accuracy,
and IMRT is preferred to spare the salivary glands to minimize xerostomia. An ex-
ample of an antral lymphoma is shown in Fig. 16.6
338 Lymphomas

Fig. 16.5 Involved
site radiotherapy
CTV for treatment
of gastric MALToma
showing PTV in green
and 95% isodose
in thicker red line.
Kidneys are shown in
yellow.

Stomach
RT is given to the whole stomach. This should be localized using CT planning which
enables not only accurate definition of the stomach but also the critical OARs, in par-
ticular the kidneys. To optimize treatment reproducibility, stomach distension should
be avoided and the patient is best treated in a near-​fasting state in the morning.
Current experience with MALT low-​grade lymphomas indicates that 24 Gy in 12
daily fractions is adequate although older series used 30–​35 Gy over 3–​4 weeks. This
results in local control in almost 100% of cases.
High-​grade or transformed MALT lymphoma is managed as diffuse large B-​cell
lymphoma with chemotherapy and RT delivering 30 Gy in 15 daily fractions.

Central nervous system (primary central nervous system lymphoma)

Patient should be treated supine, immobilized in a customized shell.
Whole brain irradiation using lateral fields with standard baseline from external
auditory meatus to supraorbital ridge adjusted to ensure coverage of middle cranial
fossa and olfactory groove.
Dose of 30–​40 Gy in 15–​20 daily 2-​Gy fractions prescribed to the intersection point.
Verification using an electronic portal imaging device or lateral linear accelerator
films; in vivo dosimetry to measure lens doses also recommended.
If craniospinal radiotherapy indicated a standard technique as described in
Chapter 17 should be used to deliver a dose of 30 Gy in 15 fractions.
 Total body electron treatment 339

Fig. 16.6  IMRT dose distribution for treatment to right antrum showing volume and dose
distribution on left and beam arrangement for a four field IMRT plan on the right.

16.12 Skin
Standard techniques to treat localized skin lymphoma using a direct electron or super-
ficial x-​ray beam should be used as described in Chapter 19.
Prescribed doses of 30 Gy in 15 fractions are usual but for indolent B cell lymphomas
24 Gy in 12 fractions is sufficient.
Palliative treatment of skin plaques in mycosis fungoides can be given with 8 Gy in
two fractions. For widespread mycosis fungoides whole body electron treatment may be
considered. This is usually delivered in a few specialized centres offering this treatment.

16.13  Total body electron treatment

Treatment volume: there is no formal definition of GTV, CTV, or PTV but in practice
the technique aims to treat the entire whole-​body skin surface to a depth of 3 to 5 mm.
Patient position: Most techniques treat the patient in four to six different standing
positions rotating at 60o intervals to enable coverage of the entire body surface. Posture
may be adjusted to facilitate beam access to skin folds particularly under the arms and
between the legs. Treatment is usually delivered in cycles, e.g. anterior and two posterior
oblique fields on day 1 and posterior and two anterior oblique fields on day 2(84). An al-
ternative approach uses a rotational technique in which the patient stands on a rotating
platform. This achieves greater dose homogeneity than multiple field techniques(85).
Shielding of the eyes, nail beds, lips, and hands is used with lead goggles and mittens.
Field arrangement and dosimetry: A  number of individualized techniques have
been described by different centres across the world. These may be broadly categor-
ized as follows:
340 Lymphomas

◆ Matching dual field system which is perhaps the most commonly used today.
A standard linear accelerator beam is used with 6 MeV electrons at a treating dis-
tance of 3.5 to 4 metres. One beam is setup with the central axis focused towards the
head and the other towards the feet; typically each is angled 15 to 20o above or below
the horizontal as shown in Fig. 16.7. A tissue equivalent sheet or ‘scatter screen’ is
placed in front of the patient to reduce the effective beam energy to 4MeV. Some
techniques employed a reflector at the vertex to increase the dose uniformity at the
vertex(86).
◆ Single field technique which requires a much longer source-​to-​surface distance of
6 to 7 metres and is therefore often not practicable in a modern linear accelerator

(a)

Beam axis

117.1 cm

700 cm

(b)
3.2 mm Laxen

287.5°

Beam axis

129 cm 252.5°
27 cm

380 cm
Fig. 16.7  Examples of total body electron techniques using single beam at 7 metre
source-​to-​surface distance (a) or dual-​field technique at 3.8 metre source-​to-​surface
distance (b) as used at Stanford.
Reprinted with permission from Zhe Chen et al. ‘Matching the dosimetry characteristics of a
dual-field Stanford technique to a customized single-field Stanford technique for total skin
electron therapy,’ International Journal of Radiation Oncology, Biology, Physics, Volume 59,
Issue 3, pp. 872–85, Copyright © 2004 Elsevier Ltd.
 Toxicity after radiotherapy for lymphoma 341

room. When used it is simpler than the dual beam system; 6 MeV beams are usually
employed with a screen as described above.
◆ A moving beta particle emitting source, e.g. strontium or a fixed narrow electron
beam below which the patient moves across the beam have also been described.
Surface thermoluminescent dosimetry measurements are taken to measure the dose
distribution; this is often quite inhomogeneous, particularly with fixed beam tech-
niques, with variations between 25% and 140% described(87), typically lowest at tan-
gential surfaces such as the top of the head, shoulders, and perineum. Low dose areas,
in particular the top of the head, inframammary folds, ventral penile surface, peri-
neum, and soles of feet may be topped up with local electron or superficial photon
fields; this may also be required for thicker areas of disease.
X-​ray contamination giving a whole body x-​ray dose is a concern with these tech-
niques. This is low with modern dual field techniques using a modern linear acceler-
ator at around 1%.
Dose prescription: Common prescriptions include:
◆ 12 Gy in 12 fractions treating four times per week
◆ 24 Gy in 8 fractions treating three times per week
◆ 30 Gy in 15 fractions treating four or five times per week.

16.14  Toxicity after radiotherapy for lymphoma

16.14.1 Acute toxicity
◆ Mediastinal RT may produce fatigue, dysphagia, and a gradual decline in blood
counts. RT to the upper neck may produce xerostomia if the parotid glands are in-
cluded in the treatment field.
◆ Radiation pneumonitis, which may occur 1 to 4  months following treatment, is
characterized by dry cough, shortness of breath, and fatigue. A chest x-​ray will re-
veal pulmonary infiltrates in the characteristic distribution of the irradiated lung.
The incidence is generally less than 5%, although it increases with larger volumes of
irradiated lung.
◆ Herpes zoster may occur within 1 year of treatment in approximately 15% of pa-
tients, more commonly in those receiving CMT.
◆ Lhermitte’s sign may occur 1–​3 months following RT and generally resolves without
treatment in weeks to months.

16.14.2 Late toxicity
Radiation to thoracic structures
The mantle field, used to treat the bilateral neck including supraclavicular, axillary,
and mediastinal lymph nodes, is now rarely used. Because of the irregular field ar-
rangement and contour variation, beam modifiers such as compensators or attenu-
ators together with custom shielding blocks are mandatory for the protection of lung
and cardiac tissues. The TD5/​5 for 20 fractions of whole lung radiation was 26.5 Gy
342 Lymphomas

(TD50 30.5 Gy) from paediatric series of patients irradiated for Wilm’s tumour(88).
Partial lung irradiation with fractionated doses have been studied by Mah et al., using
radiographic changes as the endpoint(89). A  steep dose response relationship was
found, where the TD5 was 24.7 Gy, increasing to 33.9 Gy for TD50, and to 43.5 Gy
for TD95%, normalized to 15 fractions. Many chemotherapy drugs may potentiate
the effects of RT on lung tissue, e.g. bleomycin, cyclophosphamide, and doxorubicin.
Careful treatment planning, use of optimal dose-​fractionation parameters, and the
allowance of an interval of 4 weeks between CT and RT should minimize the inci-
dence of symptomatic radiation pneumonitis. Information on the cardiac tolerance to
irradiation has been largely based on patients treated for HL. Pericarditis is the most
commonly reported subacute effect. Late effects may include coronary artery disease,
valvular disease(90), and conduction defects. Following now outdated treatment, the
risk of developing symptomatic coronary artery disease is 6–​10% at 10 years and 10–​
20% at 20 years(91). These rates represent relative risks for surgical intervention or hos-
pitalization of 1.5–​2.5. The actuarial risks of death from cardiac ischaemia appear to be
2–​6% at 10 years and 10–​12% at 15 to 20 years(91). As for lung tissue, careful treatment
planning with maximum cardiac shielding (subcarinal area and the left ventricle), and
the avoidance of dose per fraction over 2 Gy should minimize the risk of cardiac com-
plications(91). The potential cardiotoxic effects of doxorubicin may also be additive to
the long-​term damaging effects of radiation on the heart(92).

Pelvic radiation and effects on the reproductive system

Since radiation for lymphomas frequently involve large fields, treatment planning must
address effects of direct as well as scatter radiation to the genital organs. For example,
doses of 20–​24 Gy will invariably produce ovarian ablation with loss of hormonal
function resulting in menopause and sterility. Therefore, treatment planning for pelvic
lymph node RT should spare one ovary if possible, with ultrasound or CT guidance
to locate the ovary. When bilateral pelvic irradiation is inevitable, the use of surgical
oophoropexy with transposition of one ovary to the lateral abdomen and marking it
with surgical clips to ensure exclusion from the radiation fields should be considered.
However, this procedure alone can produce infertility because of interference with
the vascular supply of the transposed ovary. In males, direct radiation to the testes
generally occurs in the CMT of primary testis lymphoma, or for testicular relapse of
lymphoma where chemotherapy may not have full effect for this sanctuary site. Since
the germinal epithelium of the testes is extremely radiosensitive, where doses as low
as 15 cGy can produce transient oligospermia, and doses of 4–​6 Gy can result in per-
manent azospermia, infertility is always a consequence of scrotal irradiation. However,
Leydig cell function with testosterone production can be preserved after doses of 30–​
35 Gy, although dysfunction as manifested in a rise in luteinizing hormone (LH) and
follicle stimulating hormone levels can be seen even for lower doses of 5–​6 Gy. Based
on a review of the literature, Izard concluded that approximately 50% of males re-
ceiving 14 Gy in fractionated doses will have an abnormal LH, while 33 Gy is required
to see an abnormal testosterone level in 50%(93). Indirect radiation to the testes is also
a special consideration in the planning of fields close to the scrotum, particular if the
field size is large. The scatter dose is mainly a function of distance from the field edge,
 References 343

where gonadal doses of < 5% are usually achieved with a distance from the field edge
of 10 cm or over, for a field size of 25 cm.

Bone marrow
With the more frequent use of CMT and bone marrow transplantation in the treatment
of lymphomas, radiation is frequently required in patients who had extensive previous
chemotherapy, or who had undergone bone marrow transplantation. Haematopoietic
reserve may be significantly compromised in these patients, increasing the risk of
radiation-​induced myelosuppression. This is particularly a problem where the treat-
ment field encompasses a significant proportion of the bone marrow. The use of
granulocyte-​colony stimulating factor (G-​CSF) has been shown to ameliorate neu-
tropenia. However, G-​CSF does not correct thrombocytopenia, which is likely to be
a limiting factor once neutropenia is reversed. While platelet transfusions can be per-
formed for platelet counts of < 10–​20 × 109/​L, or for active bleeding, there is a risk
of rendering patients platelet-​transfusion dependent for prolonged periods of time
particularly when RT is given in the post-​bone marrow transplantation setting. It is
possible that platelet growth factors will become clinically available within the next
few years and its use may avert radiation-​induced thrombocytopenia. Anaemia, if pre-
sent, is usually not a dose-​limiting problem and red cell transfusions may be given as
required.

16.15 Second malignancies
An increased risk of a second malignancy has been described among HL survivors. In
general, there are approximately 55–​65 excess malignancies per 10,000 person-​years
of follow-​up among patients treated for HL, and a 20-​year cumulative incidence of
second malignancy of approximately 15–​20%(94,95). The excess risk of a second solid
tumour is generally 40–​50 per 10,000 person-​years of follow-​up. Secondary breast
cancers are more common among females treated in adolescence; however, the risk
decreases substantially for women treated over the age of 30  years(94, 96,97,98). Recent
data from British Columbia, Canada, suggest that modern treatment protocols for HL
with smaller radiation volumes (ISRT and INRT) are not associated with high risk
of breast cancer compared with historical larger volume RT treatments such as the
mantle technique(99). The risk of lung cancer is very high among smokers, particu-
larly those receiving thoracic RT(100). MOPP chemotherapy was associated with second
lung cancers, but the association between ABVD chemotherapy and second solid can-
cers has not been determined. The incidence of leukaemia following modern chemo-
therapy is 1–​3% and occurs within 10 years of treatment.

References
1. Swerdlow SH, Campo E, Harris NL, et al., eds. WHO Classification of Tumours of
Haematopoietic and Lymphoid Tissues. 4th ed. World Health Organization Classification
of Tumours, ed. F.T. Bosman, et al. 2008. IARC Press: Lyon, France.
2. Solal-​Celigny P, Roy P, Colombat P, et al. Follicular lymphoma international prognostic
index. Blood 2004; 104:1258–​65.
344 Lymphomas

3. Armitage JO, Weisenburger DD. New approach to classifying non-​Hodgkin’s lymphomas:

clinical features of the major histologic subtypes. Non-​Hodgkin’s Lymphoma Classification
Project. Journal of Clinical Oncology 1998; 16:2780–​95.
4. Advani R, Rosenberg SA, Horning SJ. Stage I and II follicular non-​Hodgkin’s lymphoma:
long-​term follow-​up of no initial therapy. Journal of Clinical Oncology 2004; 22:
1454–​9.
5. Lowry L, Smith P, Qian W, et al. Reduced dose radiotherapy for local control in non-​
Hodgkin lymphoma: a randomised phase III trial. Radiotherapy and Oncology 2011;
100:86–​92.
6. MacManus M, Fisher R, Roos D, et al. Randomized trial of systemic therapy after
involved-​field radiotherapy in patients with early-​stage follicular lymphoma: 
TROG 99.03. Journal of Clinical Oncology 2018; 36:2918–​25.
7. Salles G, Seymour JF, Offner F, et al. Rituximab maintenance for 2 years in patients with
high tumour burden follicular lymphoma responding to rituximab plus chemotherapy
(PRIMA): a phase 3, randomised controlled trial. Lancet 2011; 377:42–​51.
8. Hoskin J, Kirkwood AA, Popova B, et al. 4 Gy versus 24 Gy radiotherapy for patients with
indolent lymphoma (FORT): a randomised phase 3 non-​inferiority trial. Lancet Oncology
2014; 15:457–​63.
9. Relander T, Johnson NA, Farinha P, et al. Prognostic factors in follicular lymphoma.
Journal of Clinical Oncology 2010; 28:2902–​13.
10. Isaacson DH, Wright P. Malignant lymphoma of mucosa-​associated lymphoid tissue.
A distinctive type of B-​cell lymphoma. Cancer 1983; 52:1410–​16.
11. Ruskone-​Fourmestraux A, Fischbach W, Aleman BM, et al. EGILS consensus
report. Gastric extranodal marginal zone B-​cell lymphoma of MALT. Gut 2011; 60:
747–​58.
12. Ferreri A, Guidoboni M, Ponzoni M, et al. Evidence for association between Chlamydia
psittaci infection and ocular adnexal lymphoma (OAL). Proceedings of the American Society
of Clinical Oncology 2003; 22:565.
13. Sutcliffe SB, Gospodarowicz MK, Bush RS, et al. Role of radiation therapy in localized
non-​Hodgkin’s lymphoma. Radiotherapy and Oncology 1985; 4:211–​23.
14. Horwich A, Catton CN, Quigley M, et al. The management of early-​stage aggressive non-​
Hodgkin’s lymphoma. Hematology and Oncology 1988; 6: 291–​8.
15. Coiffier B, Lepage E, Briere J, et al. CHOP chemotherapy plus rituximab compared with
CHOP alone in elderly patients with diffuse large-​B-​cell lymphoma. New England Journal
of Medicine 2002; 346:235–​42.
14. Horning SJ, Weller E, Kim K, et al. Chemotherapy with or without radiotherapy in
limited-​stage diffuse aggressive non-​Hodgkin’s Lymphoma: Eastern Cooperative Oncology
Group Study 1484. Journal of Clinical Oncology 2004; 22:3032–​8.
15. Miller T, Dahlberg S, Cassady JR, et al. Chemotherapy alone compared with
chemotherapy plus radiotherapy for localized intermediate-​and high-​grade non-​Hodgkin’s
lymphoma. New England Journal of Medicine 1998; 339:21–​6.
16. Stephens DM, Li H, LeBlanc ML, et al. Continued risk of relapse independent of treatment
modality in limited stage diffuse large B-​cell lymphoma: Final and long-​term analysis of
SWOG study S8736. Hematolgy and Oncology 2015; 33(Suppl 1):170.
17. Fillet G, Bonnet N, Mounier N, et al. Radiotherapy is unnecessary in elderly patients with
localized aggressive non-​Hodgkin’s lymphoma: Results of the GELA LNH 93-​4 study.
Blood 2002; 100:92a, abstr 337.
 References 345

18. Reyes F, Lepage E, Ganem G, et al. ACVBP versus CHOP plus radiotherapy for localized
aggressive lymphoma. New England Journal of Medicine 2005; 352:1197–​205.
19. Dunleavy K, Pittaluga S, Maeda LS, et al. Dose-​adjusted EPOCH-​rituximab therapy
in primary mediastinal B-​cell lymphoma. New England Journal of Medicine 2013;
368:1408–​16.
20. Dabaja BS, Vanderplas AM, Crosby-​Thompson AL, et al. Radiation for diffuse large B-​cell
lymphoma in the rituximab era: analysis of the National Comprehensive Cancer Network
lymphoma outcomes project. Cancer 2015; 121:1032–​9.
21. Vargo JA, Gill BS, Balasubramani GK, Beriwal S. Treatment selection and survival
outcomes in early-​stage diffuse large B-​cell lymphoma: do we still need consolidative
radiotherapy? Journal of Clinical Oncology 2015; 33:3710–​17.
22. Aviles A, Delgado S, Nambo MJ, et al. Adjuvant radiotherapy to sites of previous bulky
disease in patients stage IV diffuse large cell lymphoma. International Journal of Radiation
Oncology, Biology, Physics 1994; 30: 799–​803.
23. Held G, Murawski N, Ziepert M, et al. Role of radiotherapy to bulky disease in
elderly patients with aggressive B-​cell lymphoma. Journal of Clinical Oncology 2014;
32:1112–​18.
24. Held G, Zeynalova S, Murawski N, et al. Impact of rituximab and radiotherapy on
outcome of patients with aggressive B-​cell lymphoma and skeletal involvement. Journal of
Clinical Oncology 2013; 31:4115–​22.
25. Tao R, Allen PK, Rodriguez A, et al. Benefit of consolidative radiation therapy for primary
bone diffuse large B-​cell lymphoma. International Journal of Radiation Oncology, Biology,
Physics 2015; 92:122–​9.
26. Laperriere NJ, Cerezo L, Milosevic MF, et al. Primary lymphoma of brain: results of
management of a modern cohort with radiation therapy. Radiotherapy and Oncology 1997;
43:247–​52.
27. Krogh-​Jensen M, d'Amore F, Jensen MK, et al. Incidence, clinicopathological features
and outcome of primary central nervous system lymphomas. Population-​based data from
a Danish lymphoma registry. Danish Lymphoma Study Group, LYFO. Annals of Oncology
1994; 5:349–​54.
28. Nelson DF, Martz KL, Bonner H, et al. Non-​Hodgkin’s lymphoma of the brain: can high
dose, large volume radiation therapy improve survival? Report on a prospective trial by
the Radiation Therapy Oncology Group (RTOG): RTOG 8315. International Journal of
Radiation Oncology, Biology, Physics 1992; 23:9–​17.
29. Abrey LE, Yahalom J, DeAngelis LM. Treatment for primary CNS lymphoma: the next
step. Journal of Clinical Oncology 2000; 18:3144–​50.
30. DeAngelis LM. Primary central nervous system lymphoma. Recent Results in Cancer
Research 1994; 135:155–​69.
31. Shah GD, Yahalom J, Correa DD, et al. Combined immunochemotherapy with reduced
whole-​brain radiotherapy for newly diagnosed primary CNS lymphoma. Journal of Clinical
Oncology 2007; 25:4730–​5.
32. Thiel E, Korfel A, Martus P, et al. High-​dose methotrexate with or without whole brain
radiotherapy for primary CNS lymphoma (G-​PCNSL-​SG-​1): a phase 3, randomised, non-​
inferiority trial. Lancet Oncology 2010; 11:1036–​47.
33. Rathmell AJ, Gospodarowicz MK, Sutcliffe SB, Clark RM. Localized extradural
lymphoma: survival, relapse pattern and functional outcome. The Princess Margaret
Hospital Lymphoma Group. Radiotherapy and Oncology 1992; 24:14–​20.
346 Lymphomas

34. Eeles RA, O'Brien P, Horwich A, Brada M, et al. Non-​Hodgkin’s lymphoma presenting
with extradural spinal cord compression: functional outcome and survival. British Journal
of Cancer 1991; 63:126–​9.
35. Willemze R, Kerl H, Sterry W, et al. EORTC classification for primary cutaneous
lymphomas: a proposal from the Cutaneous Lymphoma Study Group of the European
Organization for Research and Treatment of Cancer. Blood 1997; 90:354–​71.
36. Willemze R, Meijer CJ. EORTC classification for primary cutaneous lymphomas: a
comparison with the R.E.A.L. Classification and the proposed WHO Classification. Annals
of Oncology 2000; 11 Suppl 1(12):11–​15.
37. Willemze R, Hodak E, Zinzani PL, et al. Primary cutaneous lymphomas: ESMO Clinical
Practice Guidelines for diagnosis, treatment and follow-​up. Annals of Oncology 2013; 24
Suppl 6: vi149–​54.
38. Bekkenk MW, Vermeer MH, Geerts ML, et al. Treatment of multifocal primary cutaneous
B-​cell lymphoma: a clinical follow-​up study of 29 patients. Journal of Clinical Oncology
1999; 17: 2471–​8.
39. Roggero E, Zucca E, Mainetti C, et al. Eradication of Borrelia burgdorferi infection
in primary marginal zone B-​cell lymphoma of the skin. Human Pathology 2000; 31:
263–​8.
40. Bekkenk MW, Geelen FA, van Voorst Vader PC, et al. Primary and secondary cutaneous
CD30(+) lymphoproliferative disorders: a report from the Dutch Cutaneous Lymphoma
Group on the long-​term follow-​up data of 219 patients and guidelines for diagnosis and
treatment. Blood 2000; 95:3653–​61.
41. Berger GK, McBride A, Lawson S, et al. Brentuximab vedotin for treatment of non-​
Hodgkin lymphomas: A systematic review. Critical Reviews in Oncology and Hematology
2017; 109: 42–​50.
42. Gandolfi L, Pellegrini C, Casadei B, et al. Long-​term responders after brentuximab
vedotin: single-​center experience on relapsed and refractory Hodgkin lymphoma and
anaplastic large cell lymphoma patients. Oncologist 2016; 21:1436–​41.
43. Philip T, Guglielmi C, Hagenbeek A, et al. Autologous bone marrow transplantation as
compared with salvage chemotherapy in relapses of chemotherapy-​sensitive non-​Hodgkin’s
lymphoma. New England Journal of Medicine 1995; 333: 540–​5.
44. Noordijk EM, Carde P, Dupouy N, et al. Combined-​modality therapy for clinical stage I or
II Hodgkin’s lymphoma: long-​term results of the European Organisation for Research and
Treatment of Cancer H7 randomized controlled trials. Journal of Clinical Oncology 2006;
24:3128–​35.
45. Ferme C, Eghbali H, Meerwaldt JH, et al. Chemotherapy plus involved-​field radiation in
early-​stage Hodgkin’s disease. New England Journal of Medicine 2007; 357:1916–​27.
46. Engert A, Franklin J, Eich HT, et al. Two cycles of doxorubicin, bleomycin, vinblastine,
and dacarbazine plus extended-​field radiotherapy is superior to radiotherapy alone in early
favorable Hodgkin’s lymphoma: final results of the GHSG HD7 trial. Journal of Clinical
Oncology 2007; 25:3495–​502.
47. Meyer RM, Gospodarowicz MK, Connors JM, et al. ABVD alone versus radiation-​based
therapy in limited-​stage Hodgkin’s lymphoma. New England Journal of Medicine 2012;
366:399–​408.
48. Nachman JB, Sposto R, Herzog P, et al. Randomized comparison of low-​dose involved-​
field radiotherapy and no radiotherapy for children with Hodgkin’s disease who achieve a
complete response to chemotherapy. Journal of Clinical Oncology 2002; 20:3765–​71.
 References 347

49. Radford J, Illidge T, Counsell N, et al. Results of a trial of PET-​directed therapy for early-​
stage Hodgkin’s lymphoma. New England Journal of Medicine 2015; 372:1598–​607.
50. Raemaekers JM, André MP, Federico M, et al. Omitting radiotherapy in early positron
emission tomography-​negative stage I/​II Hodgkin lymphoma is associated with an
increased risk of early relapse: clinical results of the preplanned interim analysis of
the randomized EORTC/​LYSA/​FIL H10 Trial. Journal of Clinical Oncology 2014;
32:1188–​94.
51. Bonadonna G, Bonfante V, Viviani S, et al. ABVD plus subtotal nodal versus involved-​
field radiotherapy in early-​stage Hodgkin’s disease: long-​term results. Journal of Clinical
Oncology 2004; 22:2835–​41.
52. Engert A, Plütschow A, Eich HT, et al. Reduced treatment intensity in patients with early-​
stage Hodgkin’s lymphoma. New England Journal of Medicine 2010; 363:640–​52.
53. Eich HT, Diehl V, Görgen H, et al. Intensified chemotherapy and dose-​reduced involved-​
field radiotherapy in patients with early unfavorable Hodgkin’s lymphoma: final analysis
of the German Hodgkin Study Group HD11 trial. Journal of Clinical Oncology 2010;
28:4199–​206.
54. Nogova L, Reineke T, Eich HT, et al. Extended field radiotherapy, combined modality
treatment or involved field radiotherapy for patients with stage IA lymphocyte-​
predominant Hodgkin’s lymphoma: a retrospective analysis from the German Hodgkin
Study Group (GHSG). Annals of Oncology 2005; 16:1683–​7.
55. Wirth A, Yuen K, Barton M, et al. Long-​term outcome after radiotherapy alone for
lymphocyte-​predominant Hodgkin lymphoma. Cancer 2005; 104:1221–​9.
56. Nogova L, Reineke T, Brillant C, et al. Lymphocyte-​predominant and classical Hodgkin’s
lymphoma: a comprehensive analysis from the German Hodgkin Study Group. Journal of
Clinical Oncology 2008; 26:434–​9.
57. Eichenauer DA, Plütschow A, Fuchs M, et al. Long-​term course of patients with stage
IA nodular lymphocyte-​predominant Hodgkin Lymphoma: A report from the German
Hodgkin Study Group. Journal of Clinical Oncology 2015; 33:2857–​62.
58. Savage KJ, Skinnider B, Al-​Mansour M, et al. Treating limited-​stage nodular lymphocyte
predominant Hodgkin lymphoma similarly to classical Hodgkin lymphoma with ABVD
may improve outcome. Blood 2011; 118:4585–​90.
59. Engert A, Schiller P, Josting A, et al. Involved-​field radiotherapy is equally effective and
less toxic compared with extended-​field radiotherapy after four cycles of chemotherapy
in patients with early-​stage unfavorable Hodgkin’s lymphoma: results of the HD8 trial
of the German Hodgkin’s Lymphoma Study Group. Journal of Clinical Oncology 2003;
21:3601–​8.
60. Eich HT, Müller RP, Engenhart-​Cabillic R, et al. Involved-​node radiotherapy in early-​
stage Hodgkin’s lymphoma. Definition and guidelines of the German Hodgkin Study
Group (GHSG). Strahlentherapie und Onkologie 2008; 184:406–​10.
61. Campbell BA, Voss N, Pickles T, et al. Involved-​nodal radiation therapy as a component
of combination therapy for limited-​stage Hodgkin’s lymphoma: a question of field size.
Journal of Clinical Oncology 2008; 26:5170–​4.
62. Maraldo MV, Aznar MC, Vogelius IR, et al. Involved node radiation therapy: an effective
alternative in early-​stage hodgkin lymphoma. International Journal of Radiation Oncology,
Biology, Physics 2013; 85:1057–​65.
63. Hoskin J, Lowry L, Horwich A, et al. Randomized comparison of the Stanford V regimen
and ABVD in the treatment of advanced Hodgkin’s Lymphoma: United Kingdom National
348 Lymphomas

Cancer Research Institute Lymphoma Group Study ISRCTN 64141244. Journal of Clinical
Oncology 2009; 27:5390–​6.
64. Diehl V, Franklin J, Hasenclever D, et al. Results of the Third Interim Analysis of the
HD12 Trial of the GHSG: 8 Courses of Escalated BEACOPP Versus 4 Escalated and 4
Baseline Courses of BEACOPP with or without Additive Radiotherapy for Advanced Stage
Hodgkin s Lymphoma. Blood 2003; 102:Abstract 85.
65. Johnson P, Federico M, Kirkwood A, et al. Adapted treatment guided by interim PET-​
CT scan in advanced Hodgkin’s lymphoma. New England Journal of Medicine 2016;
374:2419–​29.
66. Loeffler M, Brosteanu O, Hasenclever D, et al. Meta-​analysis of chemotherapy versus
combined modality treatment trials in Hodgkin’s disease. International Database on
Hodgkin’s Disease Overview Study Group. Journal of Clinical Oncology 1998; 16:818–​29.
67. Laskar S, Gupta T, Vimal S, et al. Consolidation radiation after complete remission
in Hodgkin’s disease following six cycles of doxorubicin, bleomycin, vinblastine, and
dacarbazine chemotherapy: is there a need? Journal of Clinical Oncology 2004; 22:62–​8.
68. Johnson W, Sydes MR, Hancock BW, et al. Consolidation radiotherapy in patients with
advanced Hodgkin’s lymphoma: survival data from the UKLG LY09 randomized controlled
trial (ISRCTN97144519). Journal of Clinical Oncology 2010; 28:3352–​9.
69. Aleman BM, Raemaekers JM, Tirelli U, et al. Involved-​field radiotherapy for advanced
Hodgkin’s lymphoma. New England Journal of Medicine 2003; 348:2396–​406.
70. Ferme C, Sebban C, Hennequin C, et al. Comparison of chemotherapy to radiotherapy
as consolidation of complete or good partial response after six cycles of chemotherapy for
patients with advanced Hodgkin’s disease: results of the groupe d’etudes des lymphomes de
l’Adulte H89 trial. Blood 2000; 95:2246–​52.
71. Eich HT, Gossmann A, Engert A, et al. A contribution to solve the problem of the need for
consolidative radiotherapy after intensive chemotherapy in advanced stages of Hodgkin’s
lymphoma-​-​analysis of a quality control program initiated by the radiotherapy reference
center of the German Hodgkin Study Group (GHSG). International Journal of Radiation
Oncology, Biology, Physics 2007; 69:1187–​92.
72. Gordon LI, Hong F, Fisher RI, et al. Randomized phase III trial of ABVD versus Stanford
V with or without radiation therapy in locally extensive and advanced-​stage Hodgkin
lymphoma: an intergroup study coordinated by the Eastern Cooperative Oncology Group
(E2496). Journal of Clinical Oncology 2013; 31:684–​91.
73. Hoppe BS, Moskowitz CH, Filippa DA, et al. Involved-​field radiotherapy before high-​dose
therapy and autologous stem-​cell rescue in diffuse large-​cell lymphoma: long-​term disease
control and toxicity. Journal of Clinical Oncology 2008; 26:1858–​64.
74. Moskowitz AJ, Yahalom J, Kewalramani T, et al. Pretransplantation functional imaging
predicts outcome following autologous stem cell transplantation for relapsed and
refractory Hodgkin lymphoma. Blood 2010; 116:4934–​37.
75. Younes A, Gopal AK, Smith SE, et al. Results of a pivotal phase II study of brentuximab
vedotin for patients with relapsed or refractory Hodgkin’s lymphoma. Journal of Clinical
Oncology 2012; 30:2183–​9.
76. O’Brien C, Parnis FX. Salvage radiotherapy following chemotherapy failure in Hodgkin’s
disease–​what is its role? Acta Oncologica 1995; 34:99–​104.
77. Goda JS, Massey C, Kuruvilla J, et al. Role of salvage radiation therapy for patients with
relapsed or refractory hodgkin lymphoma who failed autologous stem cell transplant.
International Journal of Radiation Oncology, Biology, Physics 2012; 84:e329–​35.
 References 349

78. Illidge T, Specht L, Yahalom J, et al. Modern radiation therapy for nodal non-​Hodgkin
lymphoma-​target definition and dose guidelines from the International Lymphoma
Radiation Oncology Group. International Journal of Radiation Oncology, Biology, Physics
2014; 89:49–​58.
79. Specht L, Yahalom J, Illidge T, et al. Modern radiation therapy for Hodgkin
lymphoma: field and dose guidelines from the international lymphoma radiation oncology
group (ILROG). International Journal of Radiation Oncology, Biology, Physics 2014;
89:854–​62.
80. Yahalom J, Illidge T, Specht L, et al. Modern radiation therapy for extranodal
lymphomas: field and dose guidelines from the International Lymphoma Radiation
Oncology Group. International Journal of Radiation Oncology, Biology, Physics 2015;
92:11–​31.
81. Hoskin J, Díez P, Williams M, et al. Recommendations for the use of radiotherapy in
nodal lymphoma. Clinical Oncology (Royal College of Radiologists (Great Britain)) 2013;
25:49–​58.
82. Hoskin J, Díez P, Gallop-​Evans E, et al. Recommendations for radiotherapy technique
and dose in extra-​nodal lymphoma. Clinical Oncology (Royal College of Radiologists (Great
Britain)) 2016; 28:62–​8.
83. Binkley MS, Hiniker SM, Donaldson SS, Hoppe RT, et al. Partial orbit irradiation
achieves excellent outcomes for primary orbital lymphoma. Practical Radiation Oncology
2016; 6:255–​61.
84. Chen Z, Agostinelli AG, Wilson LD, Nath R, et al. Matching the dosimetry characteristics
of a dual-​field Stanford technique to a customized single-​field Stanford technique for total
skin electron therapy. International Journal of Radiation Oncology, Biology, Physics 2004;
59:872–​85.
85. Kumar PP, Patel IS. Comparison of dose distribution with different techniques of total
skin electron beam therapy. Clinical Radiology 1982; 33:495–​7.
86. Peters VG. Use of an electron reflector to improve dose uniformity at the vertex during
total skin electron therapy. International Journal of Radiation Oncology Biology, Physics
2000; 46:1065–​9.
87. Antolak JA, Cundiff JH, Ha CS. Utilization of thermoluminescent dosimetry in total
skin electron beam radiotherapy of mycosis fungoides. International Journal of Radiation
Oncology, Biology, Physics 1998; 40:101–​8.
88. McDonald S, Rubin P, Phillips TL, et al. Injury to the lung from cancer therapy: clinical
syndromes, measurable endpoints, and potential scoring systems. International Journal of
Radiation Oncology, Biology, Physics 1995; 31:1187–​203.
89. Mah K, Van Dyk J, Keane T, et al. Acute radiation-​induced pulmonary damage: a clinical
study on the response to fractionated radiation therapy. International Journal of Radiation
Oncology, Biology, Physics 1987; 13:179–​88.
90. Cutter DJ, Schaapveld M, Darby SC, et al. Risk of valvular heart disease after treatment for
Hodgkin lymphoma. Journal of the National Cancer Institute 2015; 107.
91. Hull MC, Morris CG, Pepine CJ, et al. Valvular dysfunction and carotid, subclavian, and
coronary artery disease in survivors of Hodgkin lymphoma treated with radiation therapy.
JAMA 2003; 290:2831–​7.
92. Myrehaug S, Pintilie M, Tsang R, et al. Cardiac morbidity following modern treatment for
Hodgkin lymphoma: supra-​additive cardiotoxicity of doxorubicin and radiation therapy.
Leukemia & Lymphoma 2008; 49:1486–​93.
350 Lymphomas

93. Izard MA. Leydig cell function and radiation: a review of the literature. Radiotherapy and
Oncology 1995; 34:1–​8.
94. van Leeuwen FE, Klokman WJ, Veer MB, et al. Long-​term risk of second malignancy in
survivors of Hodgkin’s disease treated during adolescence or young adulthood. Journal of
Clinical Oncology 2000; 18:487–​97.
95. Swerdlow AJ, Barber JA, Hudson GV, et al. Risk of second malignancy after Hodgkin’s
disease in a collaborative British cohort: the relation to age at treatment. Journal of
Clinical Oncology 2000; 18:498–​509.
96. Travis LB, Hill D, Dores GM, et al. Cumulative absolute breast cancer risk for young
women treated for Hodgkin lymphoma. Journal of the National Cancer Institute 2005;
97:1428–​37.
97. Aleman BM, van den Belt-​Dusebout AW, Klokman WJ, et al. Long-​term cause-​specific
mortality of patients treated for Hodgkin’s disease. Journal of Clinical Oncology 2003;
21:3431–​9.
98. Hodgson DC, Gilbert ES, Dores GM, et al. Long-​term solid cancer risk among 5-​year
survivors of Hodgkin’s lymphoma. Journal of Clinical Oncology 2007; 25:1489–​97.
99. Conway JL, Connors JM, Tyldesley S, et al. Secondary breast cancer risk by radiation
volume in women with Hodgkin lymphoma. International Journal of Radiation Oncology,
Biology, Physics 2017; 97:35–​41.
100. Travis LB, Gospodarowicz M, Curtis RE, et al. Lung cancer following chemotherapy
and radiotherapy for Hodgkin’s disease. Journal of the National Cancer Institute 2002;
94:182–​92.
 Chapter 17

Central nervous system tumours

Neil G Burnet, Fiona Harris, Mark B Pinkham,
Kate E Burton, and Gillian A Whitfield

17.1 Introduction
17.1.1 General introduction
Primary brain tumours account for about 3% of total cancer cases, representing about
11,000 patients annually in the UK (in 2014), and sadly, the 10-​year survival remains
at only 14% overall. Primary central nervous system (CNS) tumours are responsible
for more loss of life per patient than any other adult cancer, at just over 20 years per
patient, and the gliomas are particularly devastating. They affect patients of all ages,
from childhood to old age, increasing in incidence with age.
Of new referrals in neuro-​oncology, gliomas constitute two-​thirds, but World Health
Organization (WHO) grade IV gliomas or glioblastomas (GBM) alone account for al-
most a half. For this reason, neuro-​oncology appears dominated by GBM. Primary
CNS lymphoma contributes 3–​5% of cases, meningiomas approximately 10%, and pi-
tuitary tumours another 10%. A further 10% of patients have a wide range of rare tu-
mours. These numbers exclude patients with metastases, which have approximately a
three times higher incidence.
Primary brain tumours can be broadly divided into primary glial tumours,
ependymoma, medulloblastoma, germ cell tumours (germinoma and teratoma), men-
ingioma, nerve sheath tumours (such as vestibular schwannoma), and pituitary tu-
mours (including craniopharyngioma). Several of these can also affect the spinal cord.
Although these tumours differ widely in their pathology, treatment, and outcome,
there are common themes which apply to radiotherapy (RT) for all CNS tumours.
Rigorous attention to RT detail, including careful consideration of the dose to organs
at risk (OARs), is necessary if best results are to be achieved.
Patients with primary brain tumours experience some problems which are unique
in oncology: a significant number develop intellectual and personality change; motor
disorders including hemiplegia are especially disabling; side effects of muscle wasting
and weight gain from dexamethasone are problematic; fortunately, seizures and head-
aches are controllable in most cases. Very occasionally a young female patient presents
during pregnancy, adding an extra dimension to the management problems.

17.1.2  Diagnosis—​both radiology and pathology

Accurate diagnosis is central to effective management. Although in some tumour
types, such as vestibular schwannoma, imaging provides a definitive diagnosis, in
352 Central nervous system tumours

other cases, including intrinsic tumours in the brain, imaging offers only a differential
diagnosis. For example, difficulties occur in distinguishing glioma from primary cere-
bral lymphoma, and solitary metastasis from abscess or glioma. For gliomas, the final
histological diagnosis is now incorporating molecular pathology information. Benign
lesions with imaging appearances suggesting malignancy also occur, albeit uncom-
monly (about 0.2% in our practice). Thus, a histological diagnosis is usually desir-
able, if it is possible and safe. Occasional tumours may be hazardous to biopsy, for
example, those arising in the brainstem. Careful consideration must then be given to
the management.
Cases should be discussed in the multidisciplinary forum, so this can include cor-
relation of the pathology with the clinical history and imaging appearances. This is es-
pecially important for gliomas, to decide on the ‘effective’ clinical grade of the tumour
and its treatment.

17.1.3  Performance status in the treatment decision

Performance status (PS) remains a significant predictor of outcome, particularly for
glial tumours, and must be taken in to consideration in determining treatment in-
tent(1). Treatment can be radical, palliative, or active supportive care. For some patients
with severe disabling neurology, especially due to GBM, supportive care alone may be
the most appropriate management option.

17.2  Principles of radiotherapy planning

for CNS tumours
17.2.1. Introduction
The fundamental principles of RT planning and treatment apply to CNS tumours.
These include reproducible immobilization whose accuracy is known, high-​quality
imaging for target volume delineation, the identification of critical normal structures,
and routine use of image guidance and intensity modulated radiotherapy (IMRT).
In general, IMRT provides better conformation of the high-​dose treatment to the
shape of the target, and reduces high dose to normal tissue structures. It is usually
faster to plan and deliver than conformal radiotherapy (CRT). Image guidance im-
proves the accuracy of dose delivery, which is all the more important where steep dose
gradients are present, and should be included in some form in the treatment protocol
for all patients.

17.2.2  Patient position for radiotherapy

The majority of patients are now treated supine, which has the advantage of comfort
for the patient. Although some posterior intracranial lesions have traditionally been
treated prone, if couch extensions, such as an ‘S’ frame or a relocatable stereotactic
radiotherapy (SRT) frame are available, a supine position is possible and preferable.
These devices attach to the end of the treatment couch and allow 360° access to the
target area, without collision between gantry and couch.
 Principles of radiotherapy planning for CNS tumours 353

For palliative RT, the supine position is to be preferred because of patient comfort,
and it is difficult to safely position patients with poor PS in the prone position.
Some patients with spinal tumours may be treated prone. However, the advent of
rotational IMRT combined with on-​line image guidance makes this less attractive.
Although the prone position avoids the loss of skin sparing when beams pass through
the treatment couch, this is less important if the entry dose is spread over several beam
directions. The big disadvantage of a prone position is that respiration moves the target
volume antero-​posteriorly, as much as 1.0 cm in the thoracic region. Lying prone may
also be less comfortable, an important consideration for positional reproducibility and
intra-​fraction movement. Patients who have significant neurological disability may be
unable to easily adopt a prone position, and this affects reproducibility of set-​up. Such
patients are better treated supine.
Cervical tumours can now be managed supine using a beam direction shell, al-
though were traditionally treated prone. The same applies to lumbar tumours. For
thoracic tumours position the patient supine to avoid respiratory motion, using a thor-
acic board with the arms supported above the head.
For craniospinal axis treatment there are advantages in a prone treatment position
when using traditional techniques, which allow palpation of the spine and accurate
visualization of matching field junctions, but at the expense of comfort. However, in-
creasingly a supine position is being used, especially with IMRT, and facilitated by
image guidance. A  position in which the chin is elevated remains useful for either
position to minimize exit dose through the mouth.

17.2.3  Immobilization methods for CNS radiotherapy

The main method of immobilization for fractionated treatment is a thermoplastic beam
direction shell. For stereotactic radiosurgery (SRS), the choice is between an invasive,
skull-​fixed neurosurgical frame and a thermoplastic mask system (see section 17.12).
Thermoplastic materials soften in hot water and can be manipulated to form a mask.
Consideration must also be given to the head rest and base board; a complete system
is required for accurate, reproducible immobilization. The number of fixation points
of the shell to the base board may influence accuracy, particularly for lower target vol-
umes at shoulder level. Some thermoplastic materials shrink slightly, which may be
problematic for treatment over several weeks. Accuracy for all immobilization devices
is patient-​and operator-​dependent.
The necessary clinical target volume-​planning target volume (CTV-​PTV) mar-
gins are also highly dependent on the image guidance protocol. A margin of around
5 mm may be needed simply for set-​up uncertainties (excluding other sources of un-
certainty) with minimum image guidance, whereas with daily image guidance and
positional correction it is possible to reduce this, perhaps to as little as 1.0–​1.5 mm
according to some authors.
For palliative RT a thermoplastic shell is still recommended. It improves the ac-
curacy of relocation of the patient from one treatment to the next, and maintains pos-
ition during treatment. Positioning marks can be shown on the shell rather than the
patient.
354 Central nervous system tumours

17.2.4  Imaging for radiotherapy planning

Imaging for radical radiotherapy
A planning CT is the primary data set, performed in the treatment position, with
fiducial markers attached to the immobilization device (shell or stereotactic frame).
Computed tomography (CT) provides precise information on the shape and position
of the patient, produces density data for the treatment planning system, and permits
localization of tumour and normal tissue structures especially those related to bone.
The whole head must be scanned to allow planning with superior or superior-​oblique
beams if needed, for production of meaningful dose–​volume histograms (DVHs), and
to allow optimum electronic image co-​registration with magnetic resonance imaging
(MRI) (see later in this section). CT slice spacing requirements may vary according to
the precision required for planning; for high-​precision treatments, a spacing of 1 mm
is usual. If a digitally reconstructed radiograph (DRR) is to be used for comparison
with orthogonal X-​ray images for the purpose of image guidance, then this also re-
quires fine slice CT. Where the target comes close to the skull, the planning CT should
also be reviewed with bony window settings, because the thickness of the bone is actu-
ally less than it appears at soft tissue settings.
In most circumstances, the planning CT should be performed with intravenous
contrast; this alters dosimetry calculations by < 1%. High-​grade gliomas (HGGs), and
many benign tumours, enhance with contrast. Most low-​grade gliomas (LGGs) take
up less contrast than normal brain, so a paradoxical increase in discrimination of the
edge of the tumour occurs as a result.
For most CNS tumours additional valuable information is obtained from MRI and
is a prerequisite for high-​precision modern planning. MRI should be electronically
co-​registered with the planning CT.
MRI does not have to be performed in the exact treatment position provided
image co-​registration is available, but the closer it is to the orientation of the plan-
ning CT the better the quality after co-​registration. The appropriate MR sequence
or sequences (discussed later) should be carried out as a continuous scan without
gaps, with as fine a slice spacing as possible. The thinner and greater the number
of slices, the longer the sequence takes. The signal:noise ratio also worsens, but the
reconstruction to match the CT is better, so a balance must be struck. Contiguous
slices of 1–​3  mm are recommended. The whole head must be scanned to provide
the most anatomical information for electronic image co-​registration. Different MR
sequences may be of value for different tumours. While T1 weighting with gado-
linium contrast (T1W + Gd) is optimum for many tumours, LGGs are best dem-
onstrated with a FLAIR or T2-​weighted sequence (Fig. 17.1). In some cases, both
sequences may be useful.
Occasionally, sequences which specifically demonstrate normal tissue structure may
be valuable. For example, for hippocampal sparing 3-​dimensional spoiled gradient
(3D-​SPGR) axial MRI may facilitate contouring.
For gliomas, especially HGGs, an early postoperative MRI performed within 24–​48
(maximum 72) hours of surgery demonstrates the extent of residual tumour and dis-
tinguishes residual tumour from inflammatory post-​operative changes which develop
 Principles of radiotherapy planning for CNS tumours 355

(a)

(b)

(c)

Fig. 17.1  (a) Right temporo-​parietal GBM: CT and co-​registered T1W + gadolinium MR.
Gross tumour shows more clearly on MRI than CT. (b) Right fronto-​parietal LGG imaged
with CT and co-​registered T2W MR sequence. In both cases, the skull and patient
outline have been contoured from the planning CT. (c) Right vestibular schwannoma
demonstrated on CT, on MR using the T1W sequence with gadolinium, and with both
co-​registered.
356 Central nervous system tumours

after this time. Tumour growth between surgery and RT planning can also be evalu-
ated by comparing the planning MR with the postoperative scan. The original pre-
operative MRI should also be inspected.
The planning MRI should be performed close to the time of CT planning and the
start of treatment. If debulking surgery has been performed, then a specific post-​
operative planning scan should be used, which allows for reduction of tumour mass
and secondary mass effect, and assessment of any tumour growth over time (Fig. 17.2).
In some circumstances, information from the preoperative MRI is also valuable.
For example, the area most at risk of recurrence in a complex meningioma is the base
of the tumour. This is easy to identify on a preoperative assessment, but may be in-
visible postoperatively. In such circumstances, co-​registration with both pre-​and
postoperative MRI may be advantageous (see section 17.10).
The first step when beginning planning with co-​registered imaging is to check that
the co-​registration has worked correctly. If not, this needs to be repeated. For target
volume delineation, it is valuable to use multiple (orthogonal) planar views, as well as
the best possible imaging. Both factors have been shown to reduce inter-​operator vari-
ation in target volumes.
Other imaging modalities may have clinical utility in future, though none have been
widely adopted yet, and remain investigational. Magnetic resonance spectroscopy
(MRS) appears to contribute to localization of the edge of HGG. Diffusion-​weighted
and diffusion tensor imaging are being investigated for the same purpose. Positron
emission tomography (PET) scanning has been investigated as an aid to target local-
ization in HGG, and meningioma. It may have a niche role for rare pituitary cases.
There are limits to the spatial resolution, and caution is needed in defining the edge of
the tracer uptake.
Additional imaging considerations for intra-​cranial SRS are given in Section 17.12.

(a) (b) (c)

Fig. 17.2  Pre-​operative diagnostic (a—​axial) and post-​operative RT planning (b—​axial)

and (c—​coronal) MRI scans (T1 with contrast) of a patient with GBM. The pre-​operative
tumour contour has been transferred onto the post-​operative scan for comparison. The
pre-​operative tumour volume was 40 cm3, the postoperative GTV 18 cm3. The mass
effect has also been reduced and brain stem compression has resolved.
 Principles of radiotherapy planning for CNS tumours 357

Imaging for palliative radiotherapy
CT localization for palliative RT, for example for HGG, gives the most accurate local-
ization of the tumour, with minimum discomfort to the patient, allowing reduction in
the volume of normal tissue which will be irradiated. If craniotomy and debulking has
been carried out, then the planning CT with contrast will show the extent of tumour
and mass effect, and planning can allow for this. Even if biopsy alone has been carried
out, there is usually sufficient delay between the presentation CT and palliative RT
planning to warrant re-​imaging. It is preferable to co-​register the available MRI.

17.2.5  Planning volumes for radiotherapy

Standard volumes should be used in planning all CNS tumours, as laid out in the
International Commission on Radiation Units & Measurements (ICRU) Reports 62,
and 83(2,3) and the British Institute of Radiology (BIR) Report 2003(4). The latter con-
tains a ‘recipe’ for calculation of the PTV margin in each dimension separately, which
is especially useful.

Gross tumour volume
The gross tumour volume (GTV) is defined as actual tumour that can be seen on
imaging. If this is outlined as the starting point for planning, then the other volumes
follow easily. MRI provides the mainstay of target volume delineation. CT shows bone
much better, and is therefore useful for anatomical barriers, or for skull base tumours
involving bone (e.g. meningioma, chordoma).
For gliomas following resection, although the surgical cavity does not contain gross
tumour, it is now regarded as part of the GTV. Following temporal lobectomy, where
there is no brain substance antero-​laterally in the middle fossa, the surgical cavity is
incorporated in the GTV by convention.

Clinical target volume
The CTV is formed by a margin around the GTV to account for microscopic spread. In
principle, it is this volume that must be treated adequately to achieve cure. The precise
CTV margin depends on the tumour type, and is addressed in detail in later sections.
ICRU 83(3) introduced the concept of a CTV margin with a certain probability of tu-
mour within it. In general, this margin is based on population data rather than meas-
urements for the individual patient because tumour infiltration cannot, by definition,
be imaged. For gliomas, the margins have been defined from studies correlating post-​
mortem brain sections with ante-​mortem imaging, biopsy with imaging, and recur-
rence patterns. As yet, there is no mechanism for individualizing CTV margins.
As further imaging modalities develop, such as MRS or diffusion tensor imaging, it
may be possible to identify abnormalities resulting from tumour outside the standard
GTV. Such abnormalities should currently lie within the CTV.
The CTV can, and should, be edited according to routes of spread and anatomical
barriers. For example, gliomas do not penetrate the skull, or the meningeal structures
of the falx and tentorium, though they can spread around the edge. Meningiomas,
by contrast, do spread along meningeal surfaces. Consideration of these aspects can
358 Central nervous system tumours

reduce the volume of normal tissue receiving high-​dose irradiation, but knowledge of
the relevant anatomy is essential. Simply growing the GTV isotropically may not be
the optimum method to fashion the CTV.

Planning target volume
The PTV is applied around the CTV to ensure that it is adequately treated, and in a
sense, it is really a volume referenced to 3D space rather than to the patient. It is de-
signed to account for internal (target) organ movement and set-​up inaccuracies (see
ICRU 62)(2). These are most effectively considered as systematic (treatment prepar-
ation) errors and random (treatment execution) errors (see BIR 2003(4) and ICRU 62
and 83(2,3)).
The BIR 2003 report gives examples for several tumour sites including CNS. It is
noteworthy that systematic errors are substantially more important than day-​to-​day
random errors in treatment set-​up. This report also suggests that discrepancies in
outlining between clinicians should be incorporated as part of the PTV margin.
Reduction of both systematic and random errors can be achieved by the use of
image guidance, and the correct PTV margin depends on the immobilization tech-
nique, the image guidance protocol used, and local measurements of positional
accuracy.

Organs at risk and planning organ at risk volumes

OARs are normal tissues whose radiation sensitivity influences treatment planning
or the prescribed radiation dose. Different risks may be considered appropriate under
different clinical circumstances, so that ‘tolerance’, or the OAR ‘constraints’ defining it,
depends on the clinical setting.
CNS tumours are intimately involved with normal tissues, so that dose is deter-
mined by the tolerance of those tissues. This applies both to invasive tumours, like
the gliomas, where brain must be included within the CTV, or extrinsic tumours
lying in contact with normal tissue, such as meningioma, where normal tissue may
be included in the CTV and will lie within the PTV. Outlining of OARs is useful
to exclude ‘hot spots’ and to document dose (e.g. to the hypothalamus and pitu-
itary). They may also help the planner in choice of beam directions, or directions
to avoid.
To help in optimizing the plan, a margin around an OAR is often needed to allow
for uncertainties in set-​up. This generates the Planning Organ at risk Volume (PRV),
which is analogous to the PTV margin around the CTV. The PRV margin can be cal-
culated in the same way as the PTV margin, though it can usually be smaller. If the
PTV and PRV overlap, it is important to establish the priority, and to use a subtracted
volume, ‘PTV-​PRV’, for IMRT planning. The IMRT optimizer can then be set up to
deliver a plan with the specified priorities.
The PRV is useful principally for ‘serial’ architecture neural structures, specifically
the optic nerves, chiasm, optic tracts, brainstem, and spinal cord, when a high dose is
to be given to a nearby target. A PRV is less useful around parallel structures, since it
can change the dose–​volume histogram (DVH) in an unpredictable way, sometimes
suggesting over-​, sometimes under-​dose.
 Principles of radiotherapy planning for CNS tumours 359

In IMRT planning, ‘optimizing structures’ are used to drive the optimization, but
these should not be confused with PRVs; the two are quite distinct and used for dif-
ferent functions.
The normal tissue structures considered most relevant have been described(5), al-
though the hypothalamus should be added to this list of structures.

17.2.6  Beam energy and dose distribution considerations

For most intracranial tumours the CTV extends to the inside of the skull, at least in
part. Therefore 6 MV X-​rays are ideal, so that the CTV is below the build-​up depth.
Exceptions to this principle are pituitary tumours, and other centrally located lesions,
such as para-​sellar meningioma, where higher energy photon beams can be used. The
application of tissue-​equivalent ‘bolus’ material to the surface is not recommended.
In some surgical patients, particularly those with vertex meningioma, the skull flap
may be left out at the time of the operation. It is preferable to have a cranioplasty car-
ried out prior to RT, in order to interpose some tissue so that the build-​up zone moves
out of the CTV, and to lift the hair-​bearing scalp away from the CTV.
IMRT provides the greatest flexibility and is often easier than conformal planning in
achieving appropriate coverage of the target. Beam directions, including directions to
avoid, are still important considerations.
The ICRU recommends a dose variation across a treatment plan which is no greater
than −5% to +7% (i.e. 95–​107% of the reference point dose). However, in most situ-
ations a smaller dose variation can be achieved, and is preferred in benign disease. Try
to ensure that areas of dose > 100% are within the CTV.
The DVH for the PTV may be misleading in CNS tumours because the PTV quite
often extends into the build-​up region. In reality, this does not present a problem with
dosimetry, but makes assessment of the dose plan more difficult. The CTV, however,
should not fall into the build-​up zone, so its DVH should be close to a square shape.
DVHs for adjacent dose-​limiting normal tissues are also very important. The max-
imum dose within all neural normal tissues must be examined, to avoid ‘hot spots’
within them. As well as raising the total dose, such ‘hot spots’ also deliver a higher dose
per fraction—​so called ‘double trouble’(6) in tissues with a high fractionation sensitivity
(low alpha/​beta ratio).

17.2.7  Normal tissue tolerance doses

Introduction
Normal tissue tolerance is an important concept, embodying both the risk of a com-
plication and also the severity of its effect on the patient. It is also contextual, with ac-
ceptance of a higher risk in patients with highly malignant tumours than in those with
benign conditions. This means that stated tolerances may vary from one condition to
another; they may also change over time if the probability of cure increases. Toxicities
and tolerance doses are different in children (see Chapter 21).
The first systematic summary of tolerance dose data was produced by Emami
et  al.(7), who reviewed outcomes in patients treated before the advent of modern
imaging and 3D dose calculation. The dose-​response literature has been reassessed
360 Central nervous system tumours

more recently, in the excellent QUANTEC (Quantitative Analysis of Normal Tissue

Effects in the Clinic) reviews (see Marks et al.(8)). For many normal tissues, but not
all, tolerance doses have been revised upwards. The original Emami data described
the volume dependence of toxicity, including parameters to model the effect, but the
recent QUANTEC reviews do not include this, due to a lack of reliable data; brain tol-
erance doses are assumed to apply to partial brain irradiation.

Brain
In adults, the conventional dose-​limiting toxicity in the brain is necrosis, and dose,
fraction size, and volume are the major factors that influence risk. Although location
does not, per se, influence the risk, it does alter the risk of manifesting a clinical effect.
Other factors have been suggested to alter risk, and these include chemotherapy. This
definitely applies to combination with methotrexate, but seems not to be an issue for
temozolomide (TMZ).
Tolerance of most of the brain substance can be considered to be above 60 Gy, given
in approximately 30 fractions. Smaller volumes of brain may be able to tolerate higher
doses without additional risk. The QUANTEC review(9) suggests a 5% risk of symp-
tomatic necrosis with a dose of 72 Gy for partial brain irradiation, though notes a
range of doses from 60–​84 Gy (at 2 Gy/​fraction). The dose for a 10% risk is said to be
90 Gy (range 84–​102 Gy). These doses are substantially higher than the Emami data
suggested, where a 5% risk of necrosis was suggested with a dose of 60 Gy in 30 frac-
tions, though with an additional volume effect not characterized in the QUANTEC
report.
A separate issue is the possibility of intellectual damage from RT. The QUANTEC
review suggests that the evidence that RT induces neurocognitive injury in adults up to
4 years after RT is ‘weak’, and using planned volumes, intellectual damage in adults is
uncommon or does not occur. It should be noted that other factors may impair intellect
in those receiving RT, including direct effects of the tumour and surgery, and indirect
effects such as hydrocephalus, anticonvulsants, and untreated pituitary dysfunction.
There is limited evidence on neurocognitive effects from small studies of LGG pa-
tients specifically, where both survival and follow-​up have been relatively long. In the
EORTC randomized trial of early vs delayed RT, there were no differences in late tox-
icity between the patient cohorts, but neurocognitive functions were not specifically
assessed. Studies comparing LGG patients to cohorts with other malignancies, or to
controls, have suggested equivocal results. Taphoorn et al.(10) found no difference in
intellectual function in two cohorts of LGG patients treated with either RT or biopsy/​
surgery only. They did, however, perform less well than a cohort with haematological
malignancy, suggesting an effect of the tumour rather than the treatment. In a ran-
domized trial of two dose levels, 45 Gy (25 fractions) vs 59.4 Gy (33 fractions), pa-
tients treated to the higher dose reported lower levels of global functioning and greater
fatigue/​malaise (11), suggesting the potential for neuropsychometric effects from RT.
Gregor et al.(12) presented evidence that localized volume irradiation was substantially
(7 times) less likely to produce neuropsychometric deficit than whole brain RT. Brown
et al.(13) concluded that there is evidence of only sporadic effects on intellect in LGG
patients, provided that treatment was localized using focal RT and conventional doses.
 Principles of radiotherapy planning for CNS tumours 361

Klein et al. (14) compared LGG patients, about half of whom had received RT, to a
cohort with haematological malignancies and a cohort of healthy patients. LGG pa-
tients had lower cognitive ability than low-​grade haematological patients, and lower
still than healthy controls. Use of RT was associated with poorer cognitive function,
but memory was affected only in patients whose RT had used doses > 2 Gy/​fraction.
Antiepileptics were associated with disability in attention and executive function.
Douw et al.(15) reported on longer follow-​up of these patients, though there were only
32 RT patients in the final group. RT patients had worse attention, executive func-
tioning, and information processing speed. Cognitive disability affected about half of
the RT patients, but also a quarter of non-​RT treated patients.
In summary, it is clear that RT may cause some significant effects in some but not
necessarily all patients, and the risks can be minimized by careful planning and choice
of dose. The potential value of hippocampal sparing in reducing neurocognitive effects
of RT is a key area of investigation. As well as the other indirect factors mentioned
above, it should be noted that there can also be significant toxicities from both surgery
and chemotherapy.
For comparing dose-​fractionation schedules an alpha/​beta ratio of 2.9 is recom-
mended(9). Re-​irradiation is possible, with total cumulative doses up to about 100 Gy,
given a reasonable interval (perhaps a minimum of 1 year) between courses(9). Twice
daily schedules may be damaging, and sufficient time must be left between fractions
to allow recovery.

Brainstem
The stated tolerance of the brainstem is approximately 54 Gy in 30 fractions to 55 Gy
in 33 fractions (which are essentially the same). The QUANTEC summary supports
this limit, suggesting that the entire brainstem may be treated to 54 Gy (conventional
fractionation) with ‘limited risk’ (probably < 3%) of severe or permanent neurological
effects, and that smaller volumes (1–​10 mL) may be irradiated to maximum doses of
59 Gy, using dose per fractions of ≤ 2 Gy(16). This risk applies to the use of photons;
tolerance doses may be different for proton beam therapy.
Current consensus is to limit dose to the whole brainstem to 54 Gy, although al-
lowing a small (1–​10 mL) volume to reach a dose up to 59 Gy, even for treatment of
GBM (receiving 60 Gy).

Spinal cord
The QUANTEC review provides the most up-​to-​date summary data (17). For the cer-
vical cord, the alpha/​beta ratio is estimated at 0.87 Gy, much lower than previous litera-
ture values, and indicating a very strong relationship between risk of myelopathy and
dose per fraction. The clinical data fit well to a sigmoid dose response curve, and using
conventional doses of 1.8–​2 Gy/​fraction, and for full thickness irradiation of the cord,
the estimated risk of myelopathy is < 1% at 54 Gy and < 10% at 61 Gy. The thoracic
cord may be slightly less sensitive than the cervical cord, but data for the thoracic cord
tolerance fits poorly to a standard sigmoid dose response curve, so specific estimates of
tolerance doses and alpha/​beta ratio are not possible. As with the brainstem, tolerance
doses may be different for proton beam therapy, but this is conjecture at present.
362 Central nervous system tumours

Re-​irradiation data in animals and humans suggest partial repair of subclinical RT

damage from about 6 months, and increasing over the next 3 years(17).

Optic nerves and chiasm

The optic nerves and chiasm are thought to be slightly more sensitive than brain par-
enchyma. The QUANTEC data suggest that total dose and fraction size are the two
most important determinants of risk of optic nerve and chiasm damage. For a frac-
tionated course using 1.8–​2.0 Gy/​fraction, 50 Gy represents a ‘near zero’ risk. The risk
rises to around 3–​7% for doses in the range 55–​60 Gy and rises further with higher
doses. These doses represent a dose increase, or a risk reduction, compared to the his-
toric data(18).
It has been suggested that tolerance may be lower in patients with pituitary ad-
enoma, and this presumably also applies to craniopharyngioma. Nevertheless, for
these benign tumours where safety is paramount, doses in the region of 45–​50 Gy at
1.8 or 1.67 Gy per fraction should be entirely safe (i.e. 45 Gy in 25 fractions to 50 Gy
in 30 fractions).
The question of whether TMZ chemotherapy affects optic pathway tolerance is not
fully answered. Very few cases are reported of visual damage for combined treatment.
The original GBM ‘Stupp’ trial recommended a dose limit of 55 Gy to the chiasm
(and brain stem)(19), so higher doses have not been tested in the setting of a systematic
modern trial.
Current consensus suggests limiting optic chiasm dose to 54–​55 Gy (in 30
fractions).

Pituitary and hypothalamus
The pituitary gland and hypothalamus have a much lower tolerance for hormonal dys-
function. There is probably little effect in the long term for doses under 20–​24 Gy
but the relationship between dose, volume, and dysfunction is not fully established
in adults; in children, doses in the range 40–​60 Gy are most damaging to hormone
secretion, though growth hormone secretion is particularly sensitive to lower doses,
with an effect seen in many children at doses even below 23.4 Gy. Certainly adult pa-
tients who receive a full prescription dose of 50–​60 Gy have a significant long-​term
risk of hypothalamic–​pituitary axis dysfunction. Replacement of deficient hormones
may include growth hormone (GH). Concerns that this is mitogenic and may lead to
increased recurrence risk appear unfounded.

Ear and cochlea
The middle and inner ears are also sensitive structures. The traditional view was that
hearing would be preserved, or recover, in most patients after doses up to 60 Gy. More
recently the dose response data has been reviewed and a more conservative limit sug-
gested(20). For conventionally fractionated RT, to avoid sensory-​neural hearing loss,
efforts should be made to keep the mean dose to the cochlea to 45 Gy or less. Since a
threshold for hearing loss has not been determined, as low a dose as possible should be
sought. Potential overlapping ototoxicity relating to cytotoxic chemotherapy use, and
cisplatin in particular, is worth considering.
 High-grade glioma 363

Other cranial normal tissue structures

The lacrimal gland may be treated during RT for some intracranial and skull base le-
sions. Reduced tear output occurs at doses over about 20 Gy, matching salivary toler-
ance, though a dose below 10 Gy is safer.
The lens of the eye should have virtually no risk of cataract after receiving a dose
of 5–​6 Gy, delivered over 30 fractions. There is a 50% incidence of cataract formation
after a dose of 15 Gy(21).
The risk of permanent alopecia depends on dose to the hair follicles in the dermis.
The risk is very low or absent with doses below 10–​15 Gy; the dose to produce per-
manent alopecia in 50% of patients is 43 Gy, and the slope of the central part of the
dose response curve (γ50) is 0.9. These doses refer to the total dose given in 30 frac-
tions, and dose equivalence was calculated using a linear quadratic formulation with
an α/​β ratio of 2 Gy(22). Measurements of the actual dermal dose are difficult, but this
data can at least help assessment of risk. Hair washing has no effect on hair loss or scalp
reactions, provided it is carried out carefully.

17.3  High-​grade glioma
HGGs include WHO grade III and IV gliomas, the latter also known as glioblastoma
(GBM) (previously termed glioblastoma multiforme). The first craniotomy for glio-
blastoma was performed in London in 1884, in the same era as the discovery of X-​rays
and the start of their use as a treatment for cancer.
In 2016, the WHO changed the description of gliomas from being based simply
on microscopic appearance to being an integrated diagnosis combining molecular
pathological findings with conventional microscopy. This has arisen as a result of the
variability of outcomes in patients with histologically similar disease. Now, instead of
a diagnosis defined by cell type—​such as oligodendroglioma, astrocytoma or mixed
tumours—​the cell type is named but refined by molecular pathology. Currently, the
main molecular markers are IDH-​1 (isocitrate dehydrogenase; mutant or wild type);
ATRX (the α-​thalassemia/​mental retardation X-​linked protein); tumour protein p53
(TP53) mutation; telomerase reverse transcriptase (TERT; part of the telomerase com-
plex); and the presence (or absence) of 1p/​19q chromosome co-​deletion. ATRX is an
essential chromatin-​remodeling protein, and tumour protein p53, also known as p53,
has a regulatory role in DNA repair, cell cycle arrest and apoptosis.
Generally, oligodendrogliomas are IDH-​1 mutated and 1p/​19q co-​deleted, whereas
astrocytomas may be IDH-​1 mutant or wild-​type, but are 1p/​19q non-​co-​deleted. The
entity of mixed oligoastrocytoma has essentially disappeared from the 2016 classifica-
tion, with tumours explicitly identified as either astrocytoma or oligodendroglioma,
based on their status with respect to IDH-​1, ATRX, TP53 and the presence (or ab-
sence) of 1p/​19q chromosome co-​deletion. Most LGGs which are IDH-​1 wild-​type are
molecularly and clinically similar to glioblastoma.
Gliomatosis cerebri has been deleted from the 2016 CNS WHO classification as a
distinct entity. It is now considered a growth pattern associated with other gliomas.
Originally defined as widespread brain invasion involving three or more cerebral lobes,
frequently with bilateral growth and possible extension to infra-​tentorial structures,
364 Central nervous system tumours

further studies are needed to clarify the biological basis for the unusually widespread
infiltration in these tumours. Neurological deficits often improve with treatment, al-
though volumes are very large indeed, so a smaller dose per fraction may be useful
(e.g. 55 Gy in 33 fractions at 1.67 Gy/​fraction).

17.3.1 Radical radiotherapy

Glioblastoma—​indications
Although surgical resection has been known to be an independent prognostic factor for
some time, with survival advantage seen in patients undergoing resection compared to
biopsy alone, it is only comparatively recently that clear evidence has emerged that the
extent of resection, rather than simply selection bias, is responsible(19). Techniques to
increase the extent of safe resection have shown survival advantage, especially when
combined with chemo-​radiotherapy (chemo-​RT). Maximal safe surgical resection
should therefore be undertaken, both to improve prognosis, and to improve pressure
symptoms by reducing mass effect, and consequently reducing steroid requirements.
Biopsy may be the only appropriate option for very deep-​seated lesions, including in
the brainstem.
In 2004, randomized control trial evidence (the ‘Stupp trial’) showed that the com-
bination of radical RT and chemotherapy with TMZ, given both concurrently and as
an adjuvant following RT, produced a survival advantage compared to RT alone at the
median survival time and at 2 years(23). This result transformed the treatment of GBM.
Follow-​up at 5 years has confirmed the survival advantage (10% vs 2% 5-​year survival
for chemo-​RT vs RT alone). It is important to note that the majority of patients in the
study (84%) had undergone surgical resection, so this outcome is the result of tri-​
modality treatment.
The EORTC study also demonstrated a survival advantage in patients whose tu-
mours have inactivation, by methylation, of the gene coding for O6-​methylguanine–​
DNA methyltransferase (MGMT) which removes the predominant DNA lesion caused
by TMZ. However, a small survival advantage for chemo-​RT exists even for patients
without MGMT methylation. This treatment regimen is generally well tolerated, al-
though in the trial only 47% of patients completed the adjuvant chemotherapy phase,
due either to disease progression or chemotoxicity.
A further management challenge is the phenomenon of pseudoprogression. This is
a treatment effect, where imaging 1–​3 months following completion of the concomi-
tant chemo-​RT phase shows appearances suggestive of disease progression but where
the patient remains clinically stable. Re-​imaging at an appropriate interval shows im-
proved MRI appearances. This phenomenon occurs in around 20% of cases, and is
more common in patients with MGMT methylation.
The gold standard treatment for GBM is now considered to be maximal safe sur-
gical resection, followed by concomitant chemo-​RT, then adjuvant chemotherapy
with TMZ. This approach was endorsed in the UK in 2007 by the National Institute
for Health and Clinical Excellence (NICE), for patients of WHO PS 0 or 1. Survival
in older patients with GBM is significantly less. For patients 65 and older, 40 Gy in
 High-grade glioma 365

15 fractions with concomitant TMZ and followed by adjuvant TMZ if the tumour is
MGMT methylated, is tolerated as well and appears similarly efficacious.
Performance status is one of the most important predictors of survival(1, 23), and to-
gether with age, should guide management. Radical chemo-​RT should only be offered
to patients with excellent performance status, i.e. WHO PS 0 or 1, and therefore with
no significant neurological deficit. Occasional patients have a clinically detectable
deficit which does not impair performance, such as hemianopia from an occipital le-
sion. For those with a significant neurological deficit, palliative treatment or best sup-
portive care are more appropriate options (see section 17.1.3).

Grade III glioma—​indications

These tumours are less common than GBM (about 1:4), and most of the consider-
ations for GBM also apply to patients with WHO grade III tumours.
The EORTC has recently completed recruitment into a randomized trial (EORTC
26053–​22054, BR14 CATNON in the UK) to evaluate the role of adjuvant or con-
comitant TMZ, or both, compared to RT alone, in patients with non-​1p/​19q co-​
deleted grade III gliomas. At present in the UK these tumours are treated with
RT alone. Early data suggests an advantage from adjuvant TMZ, but the benefit
from concomitant treatment is not established yet. Definitive data from the trial
will be forthcoming over the next few years, and is likely to change practice in this
patient group.
Patients with 1p/​19q co-​deleted tumours should already receive adjuvant chemo-
therapy with a combination of procarbazine, lomustine (CCNU), and vincristine
(PCV) after completion of RT, following the publication of the updated results from
the EORTC 26951 and RTOG 94-​02 studies(24,25). These showed a significant survival
advantage in this patient population with the addition of at least two cycles of chemo-
therapy, although the treatment aim should be four to six cycles, before or more com-
monly after RT.

Treatment volume and definition—​GBM

For GBM, outlining should be performed with post-​operative MRI co-​registered with
the planning CT (see section 17.2.4). The GTV is defined as the contrast-​enhancing
abnormality on post-​operative T1 MRI (see sections 17.2.4 and 17.2.5) (Fig. 17.3).
Some patients with a secondary GBM (IDH-​mutated) may have areas of the tumour
which do not enhance; in such cases, consideration should be given to including in
the GTV hyperintensity seen on FLAIR or T2. Although not containing actual gross
tumour, the surgical cavity should be included in the GTV.
CTV margins for GBM are based on work correlating CT findings with postmortem
tumour extent, biopsy findings, and recurrence patterns. The CTV can be defined in
one or two phases; however, the recently published ESTRO-​ACROP guidelines ad-
vocate a single phase treatment with a CTV margin of 2 cm constrained by anatom-
ical boundaries(26). Previously a two-​phase approach was used, although these can be
treated with a single plan using IMRT. In this setting, phase 1 uses a 2.5 cm margin,
phase 2 a 1.5  cm margin, both grown isotropically, also constrained by anatomical
boundaries.
366 Central nervous system tumours

(a) (b)

Fig. 17.3  Planning CT
and co-​registered T1W +
gadolinium planning MR
in a patient with GBM.
Two levels are shown, with
corresponding CT (a) and
(c) and MRI (b) and (d) scans.
(c) (d) Gross tumour shows more
clearly on MRI than CT, but
the skull boundary for the
CTV is better shown on CT.
GTV (skyblue) = contrast-​
enhancing tumour plus surgical
cavity; single phase CTV
(mid blue) = GTV + 2.5 cm,
limited to natural boundaries;
PTV (red) = CTV + 0.5 cm.
Several relevant OARs are also
delineated.

These CTVs should be constrained by the skull or dura, as GBM does not extend
through the dura. Because of this, they can also be edited to midline and off the pos-
terior fossa, but great care must be taken not to omit potential routes of spread through
the corpus callosum, cerebral peduncles, and brainstem.
It is not necessary to include all the oedema around the tumour(27). Oedema sur-
rounding HGG certainly includes tumour but the relationship between level of oedema
and tumour burden is inexact. It appears to be more reliable to use the contrast-​
enhancing (gross) tumour as the starting point before adding the CTV margin. This
generally includes most of the oedema. Although oedema volume does not necessarily
change greatly with steroid administration (in HGG as distinct from metastases), it
does become less distinct and therefore harder to use as the basis for target volume
definition.
The PTV is defined with an appropriate margin, typically 0.3–​0.5 cm, depending on
the individual departmental image guidance protocols. See IMRT plan in Fig. 17.4.

Treatment volume and definition—​Grade III glioma

For Grade III gliomas, a combination of T1 + Gd and FLAIR (or T2) is usually the op-
timal imaging to include enhancing and non-​enhancing components of the tumour.
Planning should be performed on post-​operative imaging. Post-​operative MRI should
be used for planning. The GTV includes the surgical cavity, FLAIR (or T2) abnor-
mality, and areas with contrast uptake on T1 + Gd MRI.
 High-grade glioma 367

(a) (b)

(c)

Fig. 17.4  IMRT plan for a patient with left temporal lobe GBM, the same as Fig. 17.3.
A single arc VMAT technique was used to deliver 60 Gy in 30 fractions, mean dose. GTV,
sky blue; CTV, mid blue; PTV, red. Image (a) shows sparing of the optic nerves, chiasm,
and brain stem; image (b) is at a more cranial level, above these OARs, and shows the
PTV encompassed by the 57 Gy isodose (95%). (c) Sagittal view of the same plan.
For optimization, the neural OARs were expanded into PRVs with a 3 mm isotropic expansion. The
planning objectives were: brainstem: 1/​3 of volume may exceed 54 Gy, brainstem PRV: 1 cm3 may
exceed 57 Gy, Dmax < 60 Gy, left and right optic nerve PRVs D1% < 55 Gy, optic chiasm PRV D1%
< 54 Gy.
In (a), the red circle (sphere) lying within the most anterior part of the PTV is centred on the point
max dose (Dmax = 63.24 Gy = 105.4%). In (b) the small red circle (sphere) is centred on the origin
to which the patient is set up with the room lasers and marks on shell. Isodoses shown are: 105%
maroon, 100% red, 95% orange, 90% light orange, 80% yellow green, 60% green, 40% sky blue,
20% blue.
368 Central nervous system tumours

CTV margins vary from 1.5–​2.5 cm. The exact margin depends in part on the exact
definition of the GTV, and on use of single or two-​phase approach. The inclusion of
FLAIR/​T2 abnormal regions has led to bigger GTVs than historically, allowing a con-
sequential reduction in CTV margins. PTV margins are exactly as for GBM.

Dose prescription—​GBM
◆ Grade IV glioma: total dose 60 Gy in 30 fractions over 6 weeks, in either a single
phase or two phases, both with concomitant TMZ (75 mg/​m2 once daily):
• Single phase:
• Phase 1: 60 Gy in 30 fractions
• 2 phase:
• Phase 1: 50 Gy in 25 fractions (or 46 Gy in 23 fractions)
• Phase 2: 10 Gy in 5 fractions (or 14 Gy in 7 fractions).

Dose prescription—​Grade III glioma

◆ Grade III glioma: total dose 54 Gy in 30 fractions over 6 weeks, in one or two phases:
• Phase 1: 45 Gy in 25 fractions (or 39.6 Gy in 22 fractions).
• Phase 2: 9 Gy in 5 fractions (or 14.4 Gy in 8 fractions).
• Alternative based on the EORTC and RTOG trials(24,25):
• 59.4 Gy in 33 fractions (of 1.8 Gy/​fraction), in a single phase or two-​phase
treatment.

Patient care during radiotherapy

Steroid dose can be titrated down during RT, though it is unwise to stop during the
course if the patient was on dexamethasone at the start.
Lens doses should be estimated with thermoluminescent dosimetry (TLD) in pa-
tients treated radically.

17.3.2  Palliative radiotherapy (HGG)

Indications
Palliative RT is appropriate for patients over the age of 70 years using a short treatment
schedule (e.g. 2 weeks), or those under 70 with a significant neurological deficit, but
who have PS 2 or better. For patients with poor PS (3–​4), there is unlikely to be any
survival advantage with treatment, and these patients should be referred to the pallia-
tive care team, for supportive management in the community.
Palliative RT should improve or stabilize neurological function in the majority of
patients, and may prolong survival; the worse the neurological status at treatment, the
shorter the survival time.

Treatment volume and definition

CT planning should be used, and contrast is useful to visualize the tumour. This
should be outlined as the GTV. There is usually a diagnostic MRI, which can be co-​
registered to confirm localization and extent of the gross tumour. The CTV margin is
 Low-grade glioma 369

grown isotropically by 2.0–​2.5 cm but limited by the skull, and a 0.5-​cm margin used
for the PTV.

Planning technique
Best palliation of symptoms, with a minimum of side effects, is achieved with CRT or
IMRT. Position can be confirmed with on-​board imaging.
Use of parallel-​opposed lateral fields should usually be avoided, with the rare excep-
tion of a tumour spreading through the corpus callosum to affect both hemispheres,
when the collimators should be angled to avoid treating the eyes and the pharynx.
Orthovoltage X-​ray techniques are totally outmoded.

Dose prescription
For palliation of all HGGs, 30 Gy in six fractions treating three times per week is a
common schedule which is well tolerated. Alternatives are 34 Gy in 10 fractions(28) or
35 Gy in 10 fractions over 2 weeks, treating daily. In Continental Europe and the USA,
40 Gy in 15 fractions is popular, though these different doses have never been tested
against each other.
◆ 30 Gy in 6 fractions over 2 weeks, treating three times per week.

Patient care during radiotherapy

Steroid dose can be titrated during RT, though it is unwise to drop them much. More
important is a post-​RT plan to reduce the dose, to minimize muscle wasting and other
side effects. Care should be taken to not decrease the dose too rapidly as most RT tox-
icity will occur in the 2 weeks after completion.

17.4  Low-​grade glioma
17.4.1 Radical radiotherapy
Indications
Patients with LGGs are normally referred because of tumour progression, detected
either clinically or on imaging. As well as neurological deterioration, worsening of
seizure control may indicate progression. Pathological confirmation is recommended
before any decision to treat is made, and adverse features on histology may prompt
early treatment. There is developing evidence that surgical resection may be of value,
though risks and benefits must be carefully balanced. Occasional patients may seek an
oncology opinion at first presentation, e.g. following a scan performed as the result of
a seizure.
If a patient has worsening neurology, worsening seizures which are not adequately
controlled, or imaging evidence of disease progression they should be treated. If the
disease is stable, then an active surveillance programme can be initiated. The EORTC
randomized trial of the timing of RT(29) indicated that there is no overall survival
difference according to the timing of RT, although early RT often produces an im-
provement in disease-​free survival. This is valuable because it shows that an active
surveillance policy may be an excellent choice for some patients.
370 Central nervous system tumours

Neurological deficits caused by LGG may improve as a result of RT and this may be
an additional indication for treatment, even in patients with relatively poor PS. This is
quite different from HGG. RT improves epilepsy in about 50% of patients with LGG.
In rare circumstances, it can be used to try to improve intractable epilepsy in patients
with LGG.
Currently, the optimum first-​line treatment, whether RT or chemotherapy, or the
combination, is not clear. Both are certainly effective in some patients. Chemotherapy
has demonstrated particular efficacy in patients with oligodendrogliomas
demonstrating 1p/​19q co-​deletion. In a good prognosis group of patients, the RTOG
has reported results of a trial (RTOG 9802) testing RT plus PCV chemotherapy vs RT
alone, finding a better median overall survival of 13 years with RT + PCV vs 8 years
with RT alone(30).
A separate international randomized trial (BR13 in the UK, EORTC 22033–​26033)
recently closed, which compared TMZ chemotherapy vs RT as first-​line treatment in
patients with high risk LGG(31). Differences in median progression-​free survival were
not significant (39 months for TMZ, 46 months for RT) and median survival had not
been reached in 2016. At present the results do not support choosing TMZ in place
of RT for first line treatment in patients with high-​risk LGG, but mature results are
needed.
Late toxicity is an important component of overall outcome for LGG, and consider-
ation simply of progression-​free or overall survival is no longer considered sufficient
to evaluate new treatment strategies.

Treatment volume and definition

Planning should be based on MRI, co-​registered with the planning CT (see 17.2.4).
Usually the optimum sequence is FLAIR or T2W, there being no contrast enhance-
ment in LGG normally. The GTV can be considered as the edge of the high signal
region on FLAIR or T2W (Fig. 17.5). Normally this can be seen to correlate with the
low-​density region on planning CT.
The GTV is enlarged by 1.5 cm to form the CTV. This margin is grown isotropically
but edited at the skull and skull base (see 17.2.5). The CTV margin is smaller than for
HGG, partly because these tumours usually infiltrate less widely than HGG, and also
because the zone of low density on CT and high signal on FLAIR (and T2W) MRI
already indicates some infiltration. The PTV should be grown isotropically from the
edited CTV with a margin of 0.3–​0.5 cm, depending on departmental image guidance
protocols. Example plans are shown in Figs 17.6 and 17.7.
The outcome of the majority of patients with LGG is relatively good, and survival
times may be long, so long-​term complications need to be considered carefully. It is
worth outlining the pituitary and hypothalamus, as well as other relevant normal tis-
sues, so that the treatment planning system returns dose–​volume data on the crit-
ical structures. There is increasing interest in reducing dose to structures such as the
hippocampus which are involved in information processing and memory, in an at-
tempt to reduce long-​term cognitive toxicity. This may become increasingly relevant
as the cohort of long-​term survivors increases, but may carry a risk of undertreating
the tumour.
 Low-grade glioma 371

(a) (b)

Fig. 17.5  Target volume delineation for a patient with LGG (astrocytoma), shown on CT
(a) and MRI (b). The GTV (light blue) was localized from a FLAIR sequence MRI, and is
defined as the edge of the high signal region. Note that the edge of the GTV is difficult
to define, even using optimal MR imaging, which is characteristic of this type of LGG.
CTV (mid blue) = GTV + 1.5 cm; PTV (red) = CTV + 0.5 cm. Same case as 17.6.

(a) (b)
Dose-Gy
55.6 Gy
54.0 Gy
51.3 Gy
48.6 Gy
43.2 Gy
37.8 Gy
27.0 Gy
16.2 Gy

Fig. 17.6  IMRT plan for a patient with a left-​sided LGG, (a) axial and (b) para-​sagittal
slices. Same case as 17.5. The PTV is shown in red. The GTV was localized from a FLAIR
sequence MRI. IMRT was used in order to achieve dose homogeneity in the target. The
optic pathway ran through the PTV, but is not shown in these sections. The individual
components of the pathway were contoured, with their PRVs, to be sure of avoiding a
hot spot. In panel (a), a scalp volume has been added (purple) to aid the optimizer in
reducing contra-​lateral scalp dose. The globe of the eye and lens can be seen in panel
(b). Dose: 54 Gy in 30 fractions. Doses are shown for 103%, 100%, 95%, 90%, 80%,
70%, 50%, and 30%.
372 Central nervous system tumours

Dose prescription
◆ Standard:
• Total dose 54 Gy in 30 fractions over 6 weeks, in a single phase.
• 54Gy in 30 fractions over 6 weeks in a single phase is a well-​tolerated schedule,
and strikes a good balance between efficacy and toxicity. It has been used in
formal clinical studies, especially in North America (including RTOG 9802)(30).
• 55 Gy in 33 fractions (1.67 Gy/​fraction) may be useful for very large
volumes (Fig. 17.7).

(a)

Dose-Gy
56.7 Gy
55.0 Gy
52.3 Gy
49.5 Gy
44.0 Gy
40.0 Gy
27.5 Gy
11.0 Gy

(b) (c) (d)

Fig. 17.7  IMRT plan for a patient with LGG affecting the brain stem and upper cervical
spinal cord; (a) sagittal section, showing the length of the volume and the excellent
longitudinal homogeneity; (b–​d) axial sections at 3 levels. Tumour was localized using
T2W MRI. Dose homogeneity would have been virtually impossible without IMRT
because of the long volume with variable body contour. In (c) contours for the optic
chiasm and optic tract PRVs are shown. The hypothalamus (blue-​grey) and the lacrimal
glands (pink and purple) have been contoured; in (d) the parotids are shown. Note
sparing of the orbital structures and parotid glands.
In this case the diagnosis was radiological; biopsy was not performed for fear of causing neurological
damage. Dose: 55 Gy in 33 fractions. This was chosen for the smaller dose per fraction (1.67 Gy)
because of the large volume of brain stem irradiated. Neurological symptoms and signs which were
moderate at the start of treatment improved gradually over the next year. Doses are shown for
103%, 100%, 95%, 90%, 80%, 70%, 50%, and here 20% isodoses.
 Ependymoma (intracranial) 373

◆ Alternative:
• 50.4 Gy in 28 fractions of 1.8 Gy/​fraction.
• The most recent EORTC LGG studies, including BR13/​EORTC 22033–​26033,
used this dose(31).
• The EORTC undertook a randomized trial of two dose levels (45 Gy vs 59.4 Gy).
They found no tumour control advantage with the higher dose but there was
worse toxicity(29), and other data is consistent with this(32).

Patient care during radiotherapy

Steroid dose can be titrated during RT, though it is unwise to drop them much. More
important is a post-​RT plan to reduce the dose, to minimize muscle wasting and other
side effects.

17.4.2  Palliative radiotherapy (LGG)

Indications
Most patients with LGG, if treated, are treated with radical intent and the need for pal-
liative RT is relatively uncommon. It may be used in patients with a significant neuro-
logical deficit where histology suggests LGG, but radiology implies HGG. Effectively,
such patients should be managed as if they have HGG. Rarely, an elderly patient with
a poor PS and a definitive diagnosis of LGG can be treated with standard short course
fractionation.

Treatment volume and definition

As per HGG (section 17.3.2)

Planning technique
As per HGG (section 17.3.2)

Dose prescription
As per HGG (section 17.3.2)
◆ 30 Gy in 6 fractions over 2 weeks, treating 3 times per week.

Patient care during radiotherapy

As for HGG, steroid dose can be titrated during RT, and a post-​RT plan to reduce the
dose is needed.

17.5  Ependymoma (intracranial)

A rare tumour occurring most often in the posterior fossa, though it can arise in a
supra-​tentorial location, this is curable in approximately 50% of cases. Surgical resec-
tion is important; as with children complete resection probably confers an advantage.
There is some evidence of a dose response, albeit within a limited dose range. Supra-​
tentorial tumours have a worse prognosis than infra-​tentorial ones. Grade is also prog-
nostic. The role of adjuvant chemotherapy in adults is unclear(33). Spinal ependymoma
is addressed later (see section 17.11).
374 Central nervous system tumours

17.5.1 Radical radiotherapy
Indications
Radiotherapy is used as a postoperative adjuvant treatment.

Treatment volume and definition

A major development in RT management was the finding that craniospinal axis (CSA)
RT adds nothing to the local control of ependymoma, compared to localized RT.
Treatment volume and localization should be along the lines of HGG, in a single
phase. Localize using CT and MRI, to treat the postoperative volume. The GTV is any
residual contrast-​enhancing tumour and surgical cavity; if a complete resection has
been achieved, consider the surgical cavity as the GTV. Always appraise the preopera-
tive imaging as well, and obtain the operation note to determine the completeness of
resection. Add a CTV margin of 1.0 cm, grown isotropically, and edited along the skull
if necessary. The PTV should be grown from the CTV.

Dose prescription
◆ Supra-​tentorial, especially if anaplastic
• 60 Gy in 30 fractions over 6 weeks.
• Alternative: 59.4 Gy in 33 fractions (of 1.8 Gy/​fraction)(34).
◆ Infra-​tentorial
• 54–​59.4 Gy in 30–​33 fractions (of 1.8 Gy/​fraction)(34).

Patient care during treatment

Most patients do not require steroids, but those that do can have dose titrated
during RT.

17.6 Medulloblastoma
This is rare in adults, but does occur. Up to a quarter of patients are diagnosed at
21–​40 years of age. Surgical resection should be as complete as possible, as for chil-
dren. Postoperative MRI is needed to confirm the extent of resection. Patients should
be staged, preferably preoperatively, with MRI of the whole CSA. The spinal MRI will
also define the lower border of the thecal sac. In women, pelvic MRI can be used to
identify the position of the ovaries, in order to estimate dose, which can be minimized
by careful technique.
Attention to the details of RT is essential to avoid local recurrence. Prolongation of
the course is an adverse prognostic factor for recurrence in children, so interruptions
to the RT course should be avoided if possible.
Adults are said to tolerate CSA radiotherapy better than children, but caution is needed
in monitoring blood counts. Obsessional attention to RT detail is required, exactly as for
children with this condition (see Chapter 21). There is less need to consider reducing
doses in adult patients than in children, and full standard doses can be used.
There is increasing interest in the use of IMRT for craniospinal RT, especially in com-
bination with image guidance and positional correction. IMRT improves target volume
 Medulloblastoma 375

coverage; it also reduces the volume of many normal tissues receiving high dose, albeit
at the expense of larger volumes receiving low dose. Proton beam therapy has been
shown to reduce acute toxicity in adults compared to conventional RT but the theoret-
ical reductions in late effects including second malignancies are as yet unproven.
The use of helical IMRT allows treatment of the CSA in a single field, thus entirely
avoiding field junctions. At the present time this is only possible using TomoTherapy™
(Fig. 17.8). IGRT appears to make field junctioning safer; it can also potentially reduce
the PTV margin used, or increase the certainty of target volume coverage.
Adjuvant chemotherapy may be considered in adults. Although there is compelling
randomized data in children, the value in adults is not clear. Bone marrow toxicity may
be worse; vincristine neuropathy in adults may also be problematic. A formal clinical
trial is still needed to address the issue of chemotherapy in adults.
Supra-​tentorial embryonal tumours (formerly called primitive neuroectodermal tu-
mours or PNET) and pineoblastoma are probably best managed according to the same
concepts, though have a worse prognosis.

17.6.1 Radical radiotherapy
Indications
Radiotherapy is used as a postoperative adjuvant treatment.

17.6.2  Treatment volume and definition

The first phase CTV is the CSA. As noted earlier, scrupulous attention to RT detail is
required (see Chapter 21 for details). This includes care to achieve adequate coverage
of the cribriform plate and optic nerves, the lower end of the thecal sac, and the full
width of the spinal canal including lateral extensions of the theca. In the adult it is not
necessary to treat the whole of the vertebral body. An adequate PTV margin is essen-
tial, even with good immobilization and image guidance (Fig. 17.8). This margin may
vary along the length of the CSA.
For medulloblastoma, phase 2 can be based on developments in paediatric man-
agement. Current practice is to include in the GTV any residual tumour and the edge
of the resection cavity. The GTV to CTV margin is 1 cm, and this can be constrained
by bone or the tentorium. A PTV margin of 0.3–​0.5 cm is typical, depending on de-
partmental image guidance protocols. Use of non-​coplanar CRT or IMRT is essential.
Older practice was to treat the whole posterior fossa, and some important large series
of results for adults with medulloblastoma present results using this boost volume.
Nevertheless, paediatric practice suggests no loss of local control with a smaller CTV
margin.
For supratentorial embryonal tumours, phase 2 treats the demonstrable disease plus
a margin of 1.0 –​1.5 cm.
Spinal or other metastases should receive a boost, along the lines of the
medulloblastoma phase 2 above.
IMRT appears to achieve better target volume coverage than conventional con-
formal RT, and combined with image guidance is ideal for treating patients comfort-
ably in a supine position.
376 Central nervous system tumours

(a) Dose %
103.0 %
100.0 %
95.0 %
90.0 %
80.0 %
60.0 %
40.0 %
20.0 %

(b) (c) Dose %

103.0 %
100.0 %
95.0 %
90.0 %
80.0 %
60.0 %
40.0 %
20.0 %

Fig. 17.8  Medulloblastoma. IMRT plan for adult craniospinal radiotherapy, delivering 35

Gy in 21 fractions. The patient is supine, and was managed with daily image guidance
and positional correction, focused on accurate positioning of the head. Coverage of the
target volume is typically better with IMRT than with conventional approaches, and it
also reduces the volume of high dose received by the normal tissues and organs. The
patient also received cranial and spinal boosts (not shown).
(a) Sagittal view showing entire craniospinal target. The PTV has been divided into cranial, cervical
thoracic, and lumbo-​sacral components. The PTV margin has been increased from cervical to
the thoracic and then to the lumbo-​sacral spine. The high dose volume follows the curvature of
the spinal target very well, as the result of use of IMRT. In this case treatment was delivered with
TomoTherapy™, which can treat the whole volume without field junctions. Treatment time for the
whole craniospinal axis was 12.3 minutes. (b) Axial section through the level of the cribriform plate
(specifically outlined in dark blue) showing complete coverage by the 95% isodose. (c) Axial upper
lumbar view showing coverage of the spine with sparing of the kidneys on either side. The isodoses
shown are 103%, 95%, 90%, 80%, 60%, 40%, and 20%. The 100% isodose is turned off for
clarity. There was no tissue receiving ≥ 103%.

Dose prescription
Adult doses are different from those used in children. Treatment is usually given at
1.67 -​1.8 Gy/​fraction:
◆ Total dose to the posterior fossa: 54 Gy in 30 fractions in 6 weeks or 55 Gy in 33
fractions in 6½ weeks:
• Phase 1 CSA: 36 Gy in 20 fractions in 4 weeks or 35 Gy in 21 fractions in just over
4 weeks.
 Germ cell tumours—germinoma and teratoma 377

• Phase 2 tumour bed boost: 18 Gy in 10 fractions in 2 weeks or 20 Gy in 12 frac-

tions in just over 2 weeks.
• Spinal metastases boost: 14.4 Gy in 8 fractions in 1½ weeks or 15 Gy in nine frac-
tions in 2 weeks.
• Cranial metastases boost: 18 Gy in 10 fractions in 2 weeks or 20 Gy in 12 frac-
tions in 2½ weeks.
54 Gy in 30 fractions (or 55 Gy in 33 fractions) is the approximate tolerance of the
brainstem.

Patient care during treatment

Patients require routine anti-​emetics. Blood counts should be done once or twice per
week during the craniospinal phase. If a serious drop in counts is observed, treatment
can usually be continued with platelet support although changing to phase 2 has been
used in the past to avoid interruptions at the primary site .
Estimate of lens dose using TLD should be carried out. Typically, doses are high, and
often sufficient to give a moderate risk of cataract. However, there is a lower overall
hazard to the patient from cataract than from under dose to the cribiform plate with
an associated risk of recurrence.
See Chapter 21 for further details.

17.7  Germ cell tumours—​germinoma and teratoma

In adults, the standard treatment for intracranial germinoma should be RT, and results
are exceptionally good, with near 100% cure. There is no added advantage in consid-
ering dose reduction together with chemotherapy at the present time: the toxicity pro-
file of RT in adults is excellent, because growth is complete, so the therapeutic ratio is
extremely high. Nevertheless, results from paediatric trials will be of interest.
Germinoma spreads easily through the cerebrospinal fluid (CSF), and has a predi-
lection for the optic chiasm and supra-​sellar cistern. It may also progress extremely
fast. Thus, patients with an entirely curable tumour may go blind or develop cranial
nerve palsies within the space of a few days if not handled appropriately. The tumour
also responds extremely fast to RT, but not necessarily to steroids. Occasionally, a few
fractions of emergency RT must be given to reverse neurological deterioration, while
the full CSA plan is prepared.
The management of intracranial teratoma is entirely opposite, since this tumour is
much less radiosensitive than germinoma. In adults, these tumours are extremely rare,
and the management should be based on paediatric protocols; see Chapter 21. Broadly,
the diagnosis may be made on raised markers alone or on biopsy in non-​secreting
tumours, and treatment commences with chemotherapy. Residual tumours may be
offered surgical resection. RT may be given to the primary site, to a dose of approxi-
mately 54 Gy in 30 fractions. See Chapter 21 for full details.

17.7.1  Radical radiotherapy—​germinoma

Indications
Following biopsy, radiotherapy is used as the definitive treatment modality.
378 Central nervous system tumours

Treatment volume and definition

Because of the high risk of CSF spread, the first phase is CSA radiotherapy. As with
other craniospinal radiotherapy, scrupulous attention to detail is required (see above
and Chapter 21 for details).
Phase 2 treats the primary site, and any sites of metastatic disease. The imaging on
which this is to be based needs to be performed before starting treatment, because the
tumour is so radiosensitive. Localize using CT and MRI. The GTV is the contrast-​
enhancing tumour. The CTV margin can be small, e.g. 1–​2 cm, grown isotropically.
The PTV should be grown from the CTV, with a standard margin.
IMRT provides an excellent solution for craniospinal treatment, as for
medulloblastoma.

Dose prescription
◆ Total dose: 40 Gy in 24 fractions, at 1.67 Gy/​fraction:
• Phase 1 CSA: 25 Gy in 15 fractions in 3 weeks.
• Phase 2 boost: 15 Gy in 9 fractions in 2 weeks.
For cranial or spinal metastatic deposits, aim for a total of 40 Gy. Where there is ex-
tensive spinal disease, it is possible to treat the whole CSA axis to 40 Gy, with careful
observation of full blood count.
If the disease is rapidly progressive, then RT may need to be started urgently, occa-
sionally within hours. Although this precludes sophisticated planning, it is possible
to give a few fractions (2–​3) to the site of progressive disease, and then wait while
the craniospinal treatment is planned. In principle, the emergency dose could be sub-
tracted from the boost, but with only relatively modest doses required, and in the con-
text of rapid tumour growth, it is probably safer (in adults) to give the full 40 Gy total
dose in the planned phases.

Patient care during treatment

As for medulloblastoma, patients require anti-​emetics, blood counts should be done
once or twice per week, and lens doses should be estimated with TLDs.

17.8 Vestibular schwannoma
These tumours are correctly known as vestibular schwannomas (VSs), rather than
acoustic neuromas, because they arise from one of the divisions of the vestibular
nerve. A multidisciplinary approach is required for their management. Some are asso-
ciated with neurofibromatosis (NF); these patients may also have other tumours and
their care may be extremely complex, with outcomes generally inferior.
Patients usually present with unilateral sensorineural hearing loss, tinnitus, and ver-
tigo. Treatment should be considered for disease progression. It can also be justified
for those with progressive hearing loss in the absence of growth on imaging, since
treatment may arrest the decline in hearing, and for younger patients who present
with large tumours even if asymptomatic. Management options include surgery, RT,
and continued observation. Treatment selection is not always straightforward. Patient
preference should be taken into account, and this group of patients is typically very
well informed.
 Vestibular schwannoma 379

Surgery is an excellent treatment, provided complete excision is achieved without

undue morbidity, and should be curative. It is preferred for large (≥ 3cm) tumours.
Surgery is the only modality which can relieve brainstem compression, but patients
with very large tumours may not be fit enough for major surgery. Hearing loss may
occur as an inevitable part of the procedure, but this depends on the surgical ap-
proach. Brainstem (cochlear) implants are possible and MRI-​compatible implants
are preferred, especially in patients with NF2. Surgery is required for salvage after RT
failure(35).
For smaller tumours, RT is an effective alternative to surgery, with either fractionated
stereotactic radiotherapy (FSRT) or single fraction SRS. For a review, see Mendenhall
et al. (35). It avoids the upfront general risks of major surgery, can often preserve hearing
at least initially, and permits a prompt return to normal activities.
FSRT may be favoured for bulkier, inoperable lesions that make broad contact with
the brainstem. For small lesions, SRS is an excellent choice. Some also consider a de-
cline in serviceable hearing as an indication for RT, irrespective of evidence of growth.
This is because early SRS might improve hearing preservation in patients with normal
hearing at diagnosis.
Randomized data is lacking but single-​institution comparative series suggest SRS
and FSRT achieve similar outcomes in terms of excellent tumour control and a low
risk of cranial neuropathy. Hearing preservation has been reported to be superior with
FSRT but such series may not reflect modern SRS practice where greater emphasis is
now placed on target conformity and cochlear dose.
For FSRT control rates at 5 years have been reported in the region of 97–​98%, and
useful hearing preservation in around 85%. A 4% risk of trigeminal dysaesthesia has
been reported, and patients with NF2 had a lower rate of hearing preservation (60%)(36).
With SRS, 5-​year local control rates of 92%, and discernible hearing preservation of
75%, have been reported for small tumours(37). Persistent impairment of function of
the facial nerve occurred in 1%, and trigeminal in about 1.5%, though transient symp-
toms occurred in a slightly higher proportion. Results were less good for larger tu-
mours, e.g. 35–​45 mm, both for tumour control and toxicity. This group also reported
transient vestibulo-​cochlear symptoms in 13%. This certainly can occur with FSRT as
well, though the rates of this temporary side effect are not clear.
Another group has reported results of both SRS and FSRT(38). Although outcomes
were similar, there was a slight advantage in favour of FSRT, as a result of lower rates of
trigeminal nerve impairment, 2% vs 8% with FSRT and SRS respectively. However, in
one of the largest vestibular schwannoma SRS series reported, of 829 patients, using a
median marginal dose of 13 Gy, local control was 97% at 10 years, facial nerve injury
of any degree occurred in less than 1%, and the risk of trigeminal sensory loss was
3.1%(39). Useful hearing preservation was possible in 50–​77% of patients and up to 90%
in those with intra-​canalicular tumours.

17.8.1 Radical radiotherapy
Indications
A progressive vestibular schwannoma requires some form of treatment, and RT is an
excellent modality, with a high therapeutic ratio. Reducing hearing may also be an
380 Central nervous system tumours

indication. Here, the technique for FSRT will be described. SRS for these tumours is
briefly discussed in Section 17.12; it is a highly specialized technique, which should be
studied in more specific texts.

Treatment volume and definition

The target is well shown on MRI, but not CT. A T1W + Gd sequence (≤ 2mm slices)
should be co-​registered; in addition, a high resolution T2 weighted MRI sequence
(e.g. CISS, FIESTA 0.8 mm slices) can be very helpful. For high precision planning, the
registration must be achieved to better than 1 mm accuracy.
The GTV is defined as the contrast-​enhancing abnormality on MRI, including ex-
tension into the internal auditory meatus (IAM). No CTV margin is needed, provided
the imaging and co-​registration are accurate. High precision immobilization is re-
quired, to minimize the PTV margin.
Adjacent critical normal structures, such as the brainstem, should be outlined. The
cochlea should also be outlined on the CT bone windows but usually lies within the
PTV. The position of the pituitary gland should be reviewed. It can be helpful to out-
line this, to clarify for the future the dose received, but it is normally far enough away
from the target to receive only minimal dose. The ipsilateral parotid gland should be
noted, and can be outlined if desired. The lenses should be outlined to provide esti-
mates of dose.
Excellent dose distributions can be obtained using coplanar fields (see Fig. 17.9), or
with IMRT. In this benign setting, careful consideration should be given to beam entry
and exit corridors.

Dose prescription
◆ 50 Gy in 30 fractions over 6 weeks (1.67 Gy/​fraction), or 50.4 Gy in 28 fractions (1.8
Gy/​fraction)
◆ Some centres use higher doses, in the region of 54 Gy. Some use shorter fraction-
ation schedules for small tumours, but these are untested at present.
◆ Although the hearing preservation rate is lower in NF2 patients, a lower dose is not
recommended lest the tumour control rates are reduced.

17.9  Pituitary tumours (including craniopharyngioma)

The management of pituitary tumours, including craniopharyngioma, requires a multi-
disciplinary approach(40). Broadly, RT should be seen as an adjunct to neurosurgical
and medical management. For pituitary adenomas, surgery achieves decompression,
particularly of the optic chiasm, histological confirmation of the diagnosis and early re-
duction in hormone levels in secretory cases. For craniopharyngioma, surgery achieves
decompression and histological confirmation (usually). Patients with prolactinoma
have primary medical therapy, which is also an important component in treating acro-
megaly. Patients require dedicated endocrine and ophthalmological review.
RT following surgery produces very high rates of control. Overall, control rates
for pituitary adenoma should be around 95% at 10  years, and 90% at 20  years(41).
 Pituitary tumours (including craniopharyngioma) 381

(a) (b)

(c) (d)

(e)

Fig. 17.9  Vestibular schwannoma (FSRT), in a patient with preserved hearing. Panels (a) and
(b) show contouring on MRI (T1W + gadolinium) and CT, with GTV (light blue) and PTV
(red). The brain stem and both cochleas have been contoured. Immobilization: thermoplastic
mask; PTV = GTV + 5 mm (image guidance first 3 days, then once weekly).
Four field coplanar 3D conformal plan: (c) axial dose distribution showing field directions and
wedging; (d) enlarged view with fields removed; (e) sagittal view to show head position and
treatment plane below the level of the eyes. Dose: 50 Gy in 30 fractions to isocentre. Doses are
shown for 105%, 100%, 95%, 90%, 80%, 60%, 40%, and 20%.

Hormone-​secreting tumours have slightly lower control rates than non-​functioning

adenomas (NFAs): e.g. 89% vs 97% at 10 years, although worse outcomes than this have
also been reported (62% vs 96% at 10 years ). Control rates for craniopharyngioma are
around 85%–​92% at 10 years, and 80%–​88% at 20 years(42).
The risk of second tumour after pituitary RT has been estimated at about 1% per
decade (2.4% at 20 years) of which about half are malignant. However, not all series
382 Central nervous system tumours

have found an increased incidence. This most likely indicates that the risk is low rather
than completely absent.
Patients with pituitary adenoma and craniopharyngioma have an increased risk of
stroke and cardiovascular disease. However, risks after RT for pituitary adenoma are
the same as for surgery alone(43), suggesting that the underlying disease and the endo-
crine disturbance, and not the RT, are responsible.
Although not yet proven, the hope is that achieving optimal control of hypersecretion
and replacement of deficiencies will help to abrogate this risk but the potential mor-
bidity of hypopituitarism after RT should also be considered.
Most cases of deteriorating vision in patients with adenoma are not due to RT(41),
and an alternative, potentially treatable, cause should be sought. However, caution is
needed with craniopharyngioma, as discussed below (see 17.9.1 Radical radiotherapy).
With fixed fields there is a small advantage in using higher energy X-​rays, of 10–​
15 MV, compared to 6 MV, but this difference disappears using rotational techniques
(or IMRT).
Reports of the use of radiosurgery (SRS) for pituitary adenoma are increasing. There
is an obvious attraction of a single treatment episode, compared to 25 fractions. SRS is
typically used for smaller tumours, provided there is separation from the optic nerves
and chiasm. Serious toxicity, e.g. blindness, which has been reported with SRS, is now
uncommon. However, there is still a paucity of evidence that the results are truly com-
parable, since SRS series often have comparatively short follow-​up and FSRT series
with long follow-​up describe outcomes for patients treated many decades ago.

17.9.1 Radical radiotherapy
Indications—​pituitary adenoma
For pituitary adenomas, the indications for RT are relative. Not all cases need to be
treated. Indications in favour of RT include extensive residual tumour, invasion of the
cavernous sinuses, uncontrolled elevated hormone levels, and progression of tumour
after surgery alone. Factors militating against RT include normal pituitary function,
and young age because of the risk of second tumour. None of these factors is absolute,
and patient preference should also be considered. Tumours with raised proliferation
markers (Ki67 >3%) may have a higher risk of relapse after surgery alone. Rare aden-
omas with very high proliferation rates (e.g. 10%) should be treated early.

Indications—​craniopharyngioma
For craniopharyngioma, RT is recommended for almost all patients. Recurrence rates
following surgery alone, even after apparently complete removal, are high. The re-
sulting recurrence can increase neurological deficits, which are often present in pa-
tients with this disease. The consequences of recurrence usually far outweigh the risks
from RT. However, modern neurosurgical techniques do allow complete resection in
some cases, and it may be reasonable to manage such patients with careful MRI and
ophthalmological surveillance. Most patients have hypopituitarism already. Typically,
conservative surgery and RT is far less morbid than ‘radical’ surgery attempting to re-
move all tumour, which may cause neurological traction injury.
 Pituitary tumours (including craniopharyngioma) 383

Treatment volume and definition—​pituitary adenoma

The target is based on residual tumour bulk. This may include supra-​sellar extension,
spread into the cavernous sinuses or even the middle cranial fossa, and occasionally
into the sphenoid sinus or bones of the skull base.
MRI planning provides greater accuracy of localization, especially of the lateral mar-
gins, and any supra-​sellar extension. A current MRI (T1W + Gd) should be obtained,
and co-​registered. Dynamic contrast-​enhanced MRI may help to better discriminate
enhancing tumour from other structures. The combination of CT, which shows the
bony anatomy around the sella, with MRI, to show the tumour and surrounding soft
tissues, provides for optimum delineation. CT can also show bone invasion in the oc-
casional tumours which invade the skull base, and an additional high resolution diag-
nostic bone window CT may be helpful. Imaging with methionine PET may assist in
localizing residual or recurrent tumour for surgery, and may have a role in defining the
GTV in occasional tumours.
The GTV is defined as the post-​operative contrast-​enhancing abnormality, best
shown on MRI. Cavernous sinus involvement can be seen on MRI. Thus, the lateral
and superior limits are typically defined from MRI. CT shows the anterior, posterior,
and inferior limits, determined by bony structures of the skull base. Inferiorly, it may
be difficult on MRI to distinguish tumour from bone marrow in the skull base, but CT
shows the bone detail well. It is not necessary to treat surgical packing in the sphenoid,
but occasional tumours do invade the sphenoid. The cavernous sinuses can also be
localized from CT, provided that contrast has been given and the anatomy is under-
stood. Review the drawn volumes on coronal section to confirm coverage, especially
superiorly and inferiorly (Fig. 17.10).
Adenomas do not normally infiltrate at a microscopic level, so little or no CTV
margin is normally required for infiltration beyond macroscopic tumour. However,
with fractionated RT at least, it is prudent to include the whole surface within the sella
which was in contact with the tumour, since occasional tumours relapse in multiple
intra-​sellar locations. A small margin may be needed if there is uncertainty about the
localization from the imaging. The PTV margin should be grown isotropically, appro-
priate to the patient immobilization device used.
The optic nerves, chiasm, and tracts, should be outlined. Be certain to avoid ‘double
trouble’ hot spots. The brain stem, lenses, and lacrimal glands should also be con-
toured. For very inferior tumours, add the cochleas; contouring the Eustachian tubes
allows a very rapid estimate of whether secretory otitis media may occur.
Occasionally the DVH for the PTV has a lower dose tail, resulting from the PTV
including air in the sphenoid sinus. This is analogous to the ‘build-​up’ effect at the
skin. Clinically this is not a problem, provided there is build-​up tissue outside the CTV
which would move with any inferior positional displacement.

Treatment volume and definition—​craniopharyngioma

Craniopharyngiomas are very adherent to adjacent structures at both a macroscopic
and microscopic level, and all parts of the brain with which the original tumour was
in contact need to be treated. Resection, which decompresses the tumour, reduces
the mass effect, and shifts the anatomy. Neither pre-​op nor post-​op imaging exactly
384 Central nervous system tumours

(a) (b) (c)

(d) (e) (f)

Fig. 17.10  Case of pituitary adenoma causing acromegaly. Previous endoscopic trans-​

sphenoidal surgery. Recurrence with rising growth hormone levels resulted from left-​
sided tumour within the cavernous sinus, considered not surgically resectable. The
location of the hormonally active tumour (GTV) was confirmed using methionine PET
imaging. Some anterior hypothalamic pituitary axis hormone function remained.
Panel (a) planning CT, (b) planning T1w+Gd MRI, (c) methionine PET. Note that some tracer seen
antero-​medially to the GTV has been taken up by mucosa in the sphenoid sinus. Panel (d) shows
the original pre-​operative MRI in which the tumour can be seen occupying the whole pituitary
fossa, which is therefore enclosed within the CTV. The coronal and sagittal views (e) and (f) confirm
that the outlined GTV covers the tumour, and show the value of CT to confirm the inferior extent,
especially in the coronal plane. The GTV (light blue) to CTV (mid blue) margin was 1 mm, the
CTV to PTV (red) margin 3 mm, and daily CT imaging with positional correction was used. The
hypothalamus (pink) has been contoured and dose minimized here to minimize the risk of further
hormonal dysfunction.
Rotational IMRT plan. Panels (g), (i), and (j) are at the same positions as (a–​d), (e), and (f). Panel (h) is
at a more cranial level with 2 different levels shown. In the more caudal, the optic nerves and chiasm
are shown as PRVs and lie within the PTV. Plan review can confirm that no ‘hot spots’ lie within these
critical structures. Dose: 45 Gy in 25 fractions. Doses represent 103%, 100%, 95%, 90%, 80%,
70%, 50%, and 30%.
 Pituitary tumours (including craniopharyngioma) 385

(g) (h)
Dose-Gy
46.4 Gy

45.0 Gy

42.8 Gy
40.5 Gy

36.0 Gy

31.5 Gy

22.5 Gy
13.5 Gy

(i) (j)

Fig. 17.10 Continued

represents the location of the target, so that the internal anatomy of the brain must be
considered in the planning.
There may be residual solid or cystic remnants on post-​operative MRI. Both com-
ponents represent GTV. Occasionally it is worth requesting a T2W as well as the usual
T1W + Gd sequence for planning. It is important to treat all areas of the brain with
which the cyst wall has been in contact. For this, co-​registration of the preoperative
MRI may be helpful, even though it does not represent the current position of the
target accurately. The CTV is therefore the surface of the pre-​op GTV, plus a margin
for any localization uncertainty at doubtful margins, and including any postoperative
GTV. In practice, this is usually in the range 0.2–​0.5  cm, and can vary around the
perimeter of the GTV. If displaced brain has fallen back into place after surgical de-
compression, the CTV can follow this. If a substantial amount of tumour has been
removed, then the CTV is determined by following the anatomy of the brain. The PTV
margin is grown isotropically, depending on the immobilization used.
Since craniopharyngioma cysts can recur during RT (as well as before and after),
some centres re-​scan with MRI during the RT course. Recurrence is probably less
likely after more complete surgery, but this cannot be taken for granted. In children,
386 Central nervous system tumours

cyst increases requiring a change to the treatment plan have been reported in 20–​25%
of patients(44), and regular MRI imaging during the RT course is advocated by some
centres, such as at weeks 3 and 5. The rate may be half this in adults, but it is still an im-
portant consideration. It is more important with more conformal techniques as CTV
and PTV margins are reduced, which includes proton beam therapy.

Dose prescription
Doses around 45–​50 Gy are typical for adenoma. Doses may vary slightly from one
centre to another, according to what is considered as the tolerance dose of the chiasm.
Some centres favour higher doses for secretory adenomas. Craniopharyngioma doses
are slightly higher.
◆ Pituitary adenoma:
• Total dose 45 Gy in 25 fractions over 5 weeks (at 1.8 Gy/​fraction).
◆ Pituitary adenoma—​extremely large (e.g. tumour > 5 cm):
• Total dose 50 Gy in 30 fractions over 6 weeks (at 1.67 Gy/​fraction).
◆ Craniopharyngioma:
• Total dose 50 Gy in 30 fractions over 6 weeks (at 1.67 Gy/​fraction).
• Alternative: 54 Gy in 30 fractions over 6 weeks (at 1.8 Gy/​fraction).

Patient care during radiotherapy

Many patients are already taking corticosteroid replacement, especially those with
craniopharyngioma. Some will experience RT as a stressful event and will conse-
quently require an increase in steroid dose by 50–​100%. However, some seem unper-
turbed by RT and can be managed without a dose increase. Hydrocortisone dose can
be titrated against tiredness, which should never be worse than ‘mild’.
Consideration should be given to imaging craniopharyngioma patients during the
course to exclude cyst enlargement. These patients should be asked to report any change
in vision immediately, because of the risk of re-​accumulation of a craniopharyngioma
cyst causing compression of the optic chiasm. Urgent MRI is needed, and if cyst re-
currence is confirmed, repeat operation is usually needed, to try to preserve vision;
this is a neurosurgical emergency. RT does not cause the re-​accumulation, but may be
coincidentally associated, since it is given postoperatively.
Measurement of lens dose using TLD shows that the lens dose is as low as expected
from the plan. This may be useful later if there are questions about cataract.

17.10 Meningioma
17.10.1  Radical radiotherapy, primary and adjuvant
Meningiomas are usually highly localized, but some have a tendency to spread, espe-
cially along meningeal surfaces. Careful consideration must therefore be given to any
‘tail’ of tumour, seen on MRI, spreading along meningeal surfaces in contact with the
tumour mass (Fig. 17.11). The extent of macroscopic infiltration is not necessarily re-
lated to grade.
Skull base tumours may invade through the foramina in the skull base to enter the
infratemporal and pterygopalatine fossae and through the orbital fissures into the orbit.
 Meningioma 387

Fig. 17.11  Cavernous sinus

meningioma: MRI co-​registered
to the planning CT, showing
tumour extending into the
free edge of the tentorium
(arrowed). This must be
included as part of the GTV,
though it is not easy to know
where the gross tumour
stops and how much CTV
margin should be allowed
for microscopic growth along
the meningeal surfaces. The
tumour has also spread across
the top of the pituitary fossa,
in the diaphragma sellae. The
occipital pad for the relocatable
stereotactic frame can be seen.

A detailed knowledge of the anatomy of the skull base and meningeal structures is de-
sirable. Meningiomas may invade bone. Some show microscopic brain invasion and
in the WHO 2016 revision this designates a meningioma as WHO grade II (atypical).
If the bone flap is removed at operation, it is normally an advantage for the recon-
struction cranioplasty to be performed before RT. This improves the surface dosimetry
and moves the skin out of the target volume.
IMRT normally achieves better high dose conformation, though dose can still be
limited by the proximity to critical normal tissues.

17.10.2 Indications
Radical primary treatment
RT is a useful definitive treatment for progressive meningioma at inoperable locations.
The two commonest sites in this category are cavernous sinus (parasellar) meningiomas
and optic nerve meningiomas. Meningiomas involving the cavernous sinus cannot be
fully resected because of the critical structures within the sinus. Discussion with neuro-
surgical colleagues should be undertaken to assess whether subtotal surgical removal is
appropriate and safe. Biopsy is not required, provided the radiological diagnosis is definite.
For optic nerve meningiomas, the diagnosis is radiological. Vision often improves
after RT unless the patient is already blind, which is an argument for early interven-
tion. Surgery, even biopsy, may render the patient blind on the affected side.

Radical postoperative adjuvant radiotherapy

The exact indications for RT following surgical resection are not completely established.
However, where it is indicated, RT is probably better given early; non-​randomized
388 Central nervous system tumours

data suggest a survival advantage with early RT, though follow-​up must be long to
demonstrate this. The consequences of a recurrence and the ease of reoperation should
be considered.
Completely resected grade I  meningiomas are usually observed with imaging
follow-​up. For an incompletely resected grade I meningioma, RT may be considered.
However, if progression would not result in significant neurological injury and the
surgeon feels that further resection would be possible, RT can be deferred and close
imaging follow-​up undertaken. Early RT might, however, be favoured for a histological
grade I meningioma with a very high labelling index, identified using Ki-​67 or mini-​
chromosome maintenance proteins (MCM2). With a grade II meningioma, the same
general approach can be used, although incomplete resection may push the decision
more towards early RT. For malignant (i.e. grade III) meningiomas, post-​operative RT
is definitely indicated. Outcome is dependent on grade.
The ROAM/​EORTC-​1308 Phase III randomized controlled trial which opened in
2016 is comparing post-​operative adjuvant RT to a dose of 60 Gy in 30 fractions with
observation after complete resection of grade II meningioma(45). A smaller Phase II
and observational trial (EORTC 22042–​26042) which examined dose escalation in
grade II and III meningiomas, where the dose depended on the degree of surgical re-
section (Simpson grade), closed in 2013 and is now in follow-​up.

17.10.3  Treatment volume and definition

For radical primary treatment, the objective is to treat the tumour bulk, and any exten-
sion along meningeal surfaces. In the postoperative setting, RT must treat any residual
tumour bulk, any spread along meningeal surfaces, and the meninges close to the tu-
mour bed, where there is potential for recurrence. Discrepancies between planning CT
and co-​registered MRI have been reported, but modern high-​resolution scanning and
electronic coregistration have reduced or abolished this problem.
Co-​registered MRI (T1W + Gd, preferably fat suppressed for skull base locations)
is essential for planning. Where residual tumour is present, the GTV is defined as the
contrast-​enhancing abnormality, including visible meningeal extension. Meningeal
surfaces in contact with the tumour mass should be evaluated carefully. It is usual to
see a ‘tail’ of enhancement around the tumour, which may be tumour or inflammatory
reactive meningeal response (which does not require treatment). There are no com-
pletely reliable methods for distinguishing extension of tumour from inflammatory
response, but thickening of the meningeal surfaces is suspicious of tumour. In some
cases, tumour infiltration can be very extensive, for example extending from the re-
gion of the cavernous sinus, along the free edge of the tentorium cerebelli, as far as the
attachment of the tentorium to the falx (Fig. 17.11). Meningeal infiltration inferiorly
and anteriorly from the middle cranial fossa can also occur, penetrating through the
skull base. Infiltration is common across the roof of the sella.
Following successful macroscopic clearance, gross tumour is not evident on
postoperative imaging. The areas most at risk from tumour recurrence are those men-
ingeal surfaces to which the tumour was attached. Therefore, coregistration of the pre-
operative MRI with the RT planning CT is helpful to accurately localize the tumour
base (Fig. 17.12). Identifying and outlining this volume gives rise to a ‘pre-​op GTV’. In
 Meningioma 389

(a) (b) (c)

Fig. 17.12  Meningioma. (a) Postoperative MRI co-​registered with the planning CT,

showing contrast-​enhancing residual tumour, the GTV, although the contour is not
shown. The possible meningeal involvement around the original tumour base on the
falx is visible, but its extent is difficult to assess. (b) The preoperative MRI was also co-​
registered, to demonstrate and accurately localize the tumour base (green). The central
area outlined represents the falx, which had been slightly displaced by long-​standing
tumour pressure, and which was relieved at surgery. The surgical operation note was also
helpful in assessing this volume. (c) The definitive CTV (blue) and PTV (red) have been
added, guided by the combination of pre-​and postoperative imaging information (and
the operation note).

defining this volume, the operation note and pathology report should be scrutinized.
A  particular difficulty is the superior sagittal sinus whose meningeal layers can be
extensively infiltrated, without necessarily occluding the sinus itself. A  patent sinus
may be dangerous to excise, so this is frequently a site for microscopic residuum.
Postoperative imaging will show surgical changes as well as tumour, but these areas
need not be included if they are away from the bed of the tumour.
The CT should be reviewed as well as the MRI. The CT shows bony erosion, which
is related to tumour, and areas of hyperostosis which may be infiltrated. Include the
former, and the latter unless there is good evidence that it is not involved. CT also de-
lineates well the fissures and foramina at the skull base, especially in the middle fossa.
This may be helpful in outlining extension into and through the skull base. It is also
helpful in outlining the full length of the optic nerve, since the optic canal is seen more
readily on CT than the nerve passing through it is on MRI.
For cases requiring radical primary treatment, the CTV includes a margin
around gross tumour and areas of extension along the meninges, and in the case of
meningiomas involving bone, a margin into bone. It does not need to include a large
margin of adjacent brain. These cases are normally treated with a CTV margin which
varies from zero (i.e. CTV = GTV) to 0.5 cm (Fig. 17.13).
In postoperative cases, the CTV margin may be informed by the operation note
and histology. Where complete resection has been possible, the pre-​op GTV should
be defined and an appropriate margin for possible tumour spread should be allowed.
390 Central nervous system tumours

(a) (b)
Dose - Gy
51.5 Gy
50.0 Gy
47.5 Gy
45.0 Gy
40.0 Gy
35.0 Gy
25.0 Gy
10 Gy

Fig. 17.13  Target volume delineation (a) and IMRT plan (b) for a cavernous sinus
meningioma, similar to that shown in Fig. 17.11. The PTV is entirely covered by the 95%
isodose, which conforms to the complex target shape more tightly than an equivalent
conformal plan could. Note the (median) prescription dose was 50 Gy in 30 fractions, within
tolerance of the brainstem and optic pathway. The PTV is enclosed by the 95% (47.5 Gy)
isodose. The isodoses represent 103%, 100%, 95%, 90%, 80%, 60%, 40%, and 20%.

This should be extended in the directions of possible spread along meningeal surfaces.
Only in rare circumstances is it necessary to extend the volume into the brain, al-
though in the presence of brain invasion or with malignant tumours the surface of the
brain in contact with the tumour should be included in the CTV.
More aggressive grades should be treated with a larger CTV margin. The exact margin
to use for a CTV is not clear from the literature. For grade II and III meningiomas the
CTV margin along the meningeal surfaces should probably be in the range of 1–​2 cm.
A smaller margin can be used where the GTV abuts the brain.
Thus, the CTV margin is normally not isotropic, but reflects differential growth
along the meningeal surfaces compared to other directions. The PTV margin should
be added, appropriate to the patient immobilization device and image guidance used.
Critical normal structures adjacent to the tumour should be outlined, so that the dose
distribution and the DVHs for these can be reviewed. IMRT offers an obvious advan-
tage over standard treatment planning if the dose is to be escalated, as it allows for dose
limitation to normal tissues. Dose-​painting can also be utilized, particularly if there is
residual disease, which can be treated to a higher dose whilst keeping OARs below tissue
tolerance. Avoid a ‘hot spot’ lying within a critical normal tissue, to avoid ‘double trouble’.

17.10.4 Dose prescription
Doses in the range of 50–​60 Gy are typical. Lower doses, and lower doses per fraction,
can be used for tumours adjacent to the optic nerves and chiasm, determined by their
 Spinal cord tumours (primary) 391

tolerance, and higher doses can be used for vertex lesions, and for grade II and III tu-
mours. Doses may vary slightly according to what is deemed an appropriate tolerance
dose or acceptable risk.
For tumours close to the optic nerves and chiasm or brainstem, a dose per fraction
in the range 1.67–​1.8 Gy is usual. IMRT offers the possibility of treating the majority of
the PTV to a higher dose, e.g. for a grade II or III tumour, while respecting OAR toler-
ances. The ROAM/​EORTC-​1308 trial specifies 60 Gy in 30 fractions, but with conven-
tional constraints on the OARs. The dose constraints also reduce the dose per fraction.
◆ Tumours adjacent to the optic nerves and chiasm:
• Total dose 50 Gy in 30 fractions over 6 weeks (at 1.67 Gy/​fraction)
◆ Tumours at other sites, and higher grade tumours:
• 55 Gy in 33 or 54 Gy in 30 fractions (1.67–​1.8 Gy/​fraction)
• 59.4 Gy in 33 (1.8 Gy/​fraction)
• 60 Gy in 36 fractions in 6½ to 7 weeks (1.67 Gy/​fraction)
◆ ROAM study—​grade II (atypical) meningioma:
• 60 Gy in 30 fractions over 6 weeks (2.0 Gy/​fraction)

17.10.5  Patient care during radiotherapy

For tumours near the orbits, measuring lens doses using TLD shows that the lens dose
is as low as expected from the plan. This may be useful later if there is visual deteri-
oration, since this is more likely related to other factors, which need to be diagnosed,
than to the RT.

17.11  Spinal cord tumours (primary)

17.11.1 Radical radiotherapy
The majority of these tumours are intrinsic, predominantly astrocytoma or
ependymoma. Glioblastoma is rare, with an appalling prognosis, and recurrence
often includes CSF spread. There is no evidence on the potential role of TMZ chemo-
therapy (concurrent or adjuvant) though it is being used in some centres. Occasionally
meningiomas and schwannomas occur. RT is essentially localized, and dose is limited
by cord tolerance.
It is always valuable to have a postoperative, pre-​radiotherapy MRI to use as a base-
line for future comparison.
IMRT achieves greater dose homogeneity than conformal planning, especially by
reducing inhomogeneity in the longitudinal direction. Avoid any ‘double trouble’ hot
spots within the cord and consider the potential clinical significance of low dose wash
to structures such as the lungs.

Indications
Patients should have at least a biopsy carried out, but cautious neurosurgical debulking
is almost certainly an advantage. Surgical resection is often incomplete, except in
some low-​grade ependymomas, arising in the lumbar region. RT is then used as a
postoperative adjuvant treatment.
392 Central nervous system tumours

In (grade I) lumbar myxopapillary ependymoma specifically, total excision surgery

is curative in about half of patients. If the surgeon is confident that a plane of cleavage
was found and removal was complete, it may be reasonable to withhold RT. If recur-
rence occurs, radiotherapy should be given after further surgery. However, adjuvant
RT after surgery appears to significantly reduce the rate of recurrence, which may in-
fluence the patient’s choice of management(46). Failures occur exclusively in the neural
axis, mainly at the primary site.

Treatment volume and definition

The preoperative MRI is the basis for localizing the target. However, the postoperative,
pre-​radiotherapy MRI should also be reviewed for planning.
Precise MRI co-​registration is often not easy, but location of the tumour can be judged
from vertebral levels, and accurately transferred to the planning CT in this way. If the
neurosurgical procedure has included spinal stabilization with metal, this may produce
significant artefact, especially on MRI but also on CT. In practice, the spinal canal can be
reliably defined on the planning CT, and approximates to the cord with a small margin.
For ependymoma, the GTV is also the spinal cord axially. For the CTV the whole
CSF space around the cord should be included. Longitudinally, a margin is needed,
typically 2–​3  cm proximally and distally (Fig. 17.14). Localized volumes are as ef-
fective as larger volumes, and most recurrences occur locally(47). Some tumours are
associated with a proximal syrinx, which does not have to be included, but this must
be carefully distinguished from tumour cyst, which does.
For lumbar ependymoma, extend the CTV laterally to include the nerve roots up
to the limit of the meningeal covering. Traditionally, the CTV was extended inferiorly
to include the CSF down to bottom of the thecal sac (usually in the range S2 to S4).
The rationale is that some tumours present with ‘sugar coating’ of tumour nodules
on the nerve roots here. However, in the absence of tumour nodules on MRI, this is
probably not necessary. It was the basis for the two-​phase treatment of lumbro-​sacral
ependymoma, which can be found in the literature.
For astrocytoma or GBM, the GTV is defined as the whole width of the cord, over a
length equal to the extent of the preoperative GTV. The CTV is the same axially; this
approximates to the spinal canal with a small margin. Longitudinally, the CTV requires
an additional margin beyond the GTV, usually 2–​3 cm proximally and distally. The
PTV margin should be grown isotropically; a margin of 0.5–​1.0 cm is recommended.
The occasional spinal meningioma or schwannoma which is inoperable or recurrent
needs to be treated. The GTV is the imageable tumour, and only a small CTV margin
(1–​2 mm) is needed for these tumours. The PTV should be the same as for other spinal
tumours.
Contour the critical normal structures close by. In the cervical region, this should in-
clude the thyroid gland to note dose in relation to risk of late primary hypothyroidism.
In the thoracic region, the whole lungs should be contoured so that dose-​volume con-
straints can be considered; the oesophagus normally falls close to the PTV and should
be outlined; cardiac doses are normally low but may be worth documenting; re-
member that the kidneys lie close to the lower thoracic spine, so their position should
be reviewed, and outlined if near the PTV. In the lumbo-​sacral region, the kidneys are
one of the most important normal tissues to contour (Fig. 17.14); in premenopausal
 Spinal cord tumours (primary) 393

Dose - Gy
(a) (b) 51.5 Gy
50.0 Gy
47.5 Gy
45.0 Gy
40.0 Gy
30.0 Gy
20.0 Gy
10 Gy

(c) STANDARD Cumulative DVH Relative Options

100
95
90 PTV
85
80
75
Relative Volume (% Normalized)

70
65
60
55
50
45
40
35
30
25
20
15 Left & right kidneys
10
5
0
0 5 10 15 20 25 30 35 40 45 50 55
20% 40% Dose (Gy) 80% 100%

Fig. 17.14  Rotational IMRT plan for an upper lumbar ependymoma, with target volume
extending from T10–​L2 inclusive; (a) axial view, (b) sagittal view. Note the supine
position, which is possible using daily image guidance and reduces respiratory motion of
the target compared to a prone position. The PTV (red) is partially covered by the 100%
isodose, and entirely enclosed within the 95% isodose. Dose has been pulled medially
away from the kidneys as much as possible in order to minimize the dose they receive.
The isodoses for 103%, 100%, 95%, 90%, 80%, 60%, 40%, and 20% are shown.
(c) DVH confirming low kidney doses and excellent homogeneous coverage of the PTV.

women the ovaries may be visible on the planning CT, and if so can be outlined and
dose minimized.
Dose prescription
Doses around 50 Gy, with a dose per fraction of 1.67 Gy, are recommended, this
being a reasonable safe estimate of cord tolerance. For patients with highly malignant
394 Central nervous system tumours

tumours such as anaplastic (grade III) tumours or glioblastoma, a higher dose may be
reasonable.
◆ Intrinsic tumours (astrocytoma and ependymoma):
• Total dose 50 Gy in 30 fractions over 6 weeks.
◆ Glioblastoma:
• Total dose 54 Gy in 30 fractions over 6 weeks, single phase (the role of TMZ
chemotherapy is unknown).
◆ Extrinsic tumours (meningioma, schwannoma):
• Total dose 50 Gy in 30 fractions over 6 weeks, single phase.
Patient care during radiotherapy
Anti-​emetics may be needed occasionally. Minimize steroid dose.

17.12 Stereotactic radiosurgery
17.12.1 General introduction
SRS involves the precise delivery of a single high dose of ionizing radiation. It
differs fundamentally from conventional RT, which is used to treat targets with
margins using multiple fractions, leading to a cumulative injury and mitotic cell
death or apoptosis. SRS requires a clearly defined target usually less than about
3–​4 cm in diameter and the objective is tissue destruction, probably through endo-
thelial cell injury and the ceramide pathway. A steep dose gradient and high level
of conformity is required to minimize dose and potential injury to normal tissues
(Fig. 17.15). Hypofractionated SRT is based on SRS principles and may be con-
sidered for larger targets or those in eloquent locations, aiming to improve the
therapeutic ratio.
SRS is used to treat a range of benign and malignant CNS tumours as well as vas-
cular and functional conditions. SRS may be considered an alternative to open surgery
in appropriately selected individuals and complements, rather than replaces, other
treatment modalities. Consequently, management decisions are best made within a
multidisciplinary setting.

17.12.2  Stereotactic radiosurgery technologies

SRS was initially developed localizing intracranial targets according to a fixed reference
point using an invasive neurosurgical frame. Technology permitting SRS without true
stereotaxy is now possible but the terminology has remained. Multiple non-​coplanar
beams are required to achieve high dose conformity to the target and dose gradi-
ents typically in the range 2–​5 Gy/​mm. Various machines can deliver SRS including
a multi-​source cobalt unit (Gamma Knife®, Elekta, Stockholm), a small linear accel-
erator mounted on a robotic arm (Cyberknife®, Accuray, Sunnyvale, CA, USA) and
adapted conventional linear accelerators utilizing fixed-​field 3D-​conformal or IMRT
beams or arcs. Although there are dosimetric differences between technologies their
clinical significance is not well defined. Minimum treatable target size, number of tar-
gets, total body scattered dose, and patient access may be relevant considerations in
certain cases.
 Stereotactic radiosurgery 395

(a) (b)

(c)

(d)

Fig. 17.15  SRS plan for a 58-​year-​old male with melanoma and a small right low frontal
brain metastasis, in close relation to the optic pathway. Planning has been performed
using the gadolinium-​enhanced T1-​weighted MRI. Using Gamma Knife, 20 Gy was
delivered to the 50% isodose whilst limiting the right optic nerve to 8.2 Gy maximum
point dose. (a) Axial, (b) coronal, and (c) sagittal orthogonal views; (d) enlarged axial
view, showing steep dose gradient and limit of dose to the optic nerve. Structures: GTV,
red; optic nerve; sky blue; optic chiasm, pink. Doses: 20 Gy yellow (50% isodose), 12 Gy
inner green, 8 Gy outer green.
396 Central nervous system tumours

17.12.3  SRS planning process and treatment pathway

Generally patients are treated supine which is the most stable and comfortable pos-
ition. Immobilization depends on the technology used and has important implications.
For non-​invasive relocatable systems, target localization with on-​table pre-​treatment
image guidance and intra-​fraction motion monitoring is required. Cone beam com-
puted tomography (CBCT) and/​or kV orthogonal imaging should be used, sometimes
supplemented with infrared guidance. If possible, all beams or arcs should be verified,
not just co-​planar angles.
Target volume delineation on thin-​slice MRI with accurate co-​registration to plan-
ning CT is required, the MRI sequences depending on the indication. Additional com-
plementary imaging modalities, with or without stereotaxy, may also be used, such
as angiography for arteriovenous malformation. Typically the target volume or GTV
is the visible tumour. No expansion to CTV is performed except in cases of SRS or
SRT to the surgical cavity following resection of brain metastases, where it may be
considered. PTV expansion varies with technology and indications, usually in the
range 0–​2 mm. Consideration should be given to mechanical and physical uncertain-
ties inherent in the treatment process as well as the potential for growth of malignant
tumours between planning and treatment. Unnecessary margin expansions are dis-
couraged because toxicity directly correlates with volume of normal tissue irradiated.
Corticosteroids may be added or adjusted at the commencement of treatment to re-
duce the risk of acute symptomatic treatment-​related oedema.

17.12.4  Indications for stereotactic radiosurgery

Brain metastases
Up to 20% of patients with cancer develop brain metastases at some stage during their
illness. Non-​small cell lung cancer, breast cancer, melanoma, renal cell carcinoma, and
colorectal carcinoma are the most common primaries implicated. Although the prog-
nosis is poor overall, outcome varies according to patient age, performance status,
number of brain metastases, primary histology, and extracranial disease control. In
better-​prognosis patients, the addition of SRS to whole brain radiotherapy (WBRT)
improves survival in patients with a single brain metastasis, and functional independ-
ence and steroid dependency in patients with up to three brain metastases. Although
randomized evidence of benefit for SRS to more brain metastases is lacking, total
volume rather than absolute number correlates better with survival in selected patients
with low volume intracranial disease and up to ten lesions.
Disease is defined on the gadolinium-​enhanced T1-​weighted MRI. For SRS the dose
is usually in the range 15–​24 Gy prescribed to the isodose encompassing the PTV and
depends on target volume, proximity to OARs and total number of targets. For SRT,
24–​30 Gy in 3–​5 fractions has been used. Typically, the dose has been prescribed to
the 50–​80% isodose for Gamma Knife (the range is because of the fixed size circular
collimators) (Fig. 17.15) and to the 80–​90% isodose for linear accelerators, although
prescription to a lower isodose may be considered in order to achieve steeper dose
fall-​off outside the target volume.
 Stereotactic radiosurgery 397

The routine addition of WBRT after SRS is no longer recommended(48,49) because

survival is equivalent and it is associated with inferior cognitive function and quality
of life(50). However, the risk of intracranial failure after SRS alone is greater and sur-
veillance MRI is common. Given that the risk of recurrence after surgery is greatest
locally, SRS or SRT to the cavity is an increasingly utilized approach to avoid or defer
WBRT in this setting, although randomized data to support this is lacking. Caution is
required due to difficulties including cavity shift and target delineation after surgery.
For review of SRS see(51) and for neurocognitive effects of cranial irradiation for brain
metastases see(52).
Vestibular schwannoma
For smaller tumours, RT is an effective alternative to surgery, with either SRS or FSRT
(see Section 17.8). In patients with diminishing hearing who had normal hearing at
diagnosis, early SRS might improve hearing preservation. With SRS, 5-​year local con-
trol rates of 92–​97%, and discernible hearing preservation of 75%, have been reported
for small tumours. Persistent impairment of function of the facial nerve occurred in
1%, and trigeminal in 1.5–​3%, or in one series, 8%. Useful hearing preservation is
possible in 50–​77% of patients and up to 90% in those with intra-​canalicular tumours.
Outcomes are slightly less good in larger tumours.
Disease is defined on the gadolinium-​enhanced T1-​weighted sequence. A  high-​
resolution T2-​weighted MRI sequence (e.g. FIESTA, CISS with 0.8 mm slices) can also
be very helpful for SRS planning. Using SRS, 12–​13 Gy is delivered to the margin of the
tumour (Fig. 17.16). Historically higher doses in the range 16–​20 Gy were delivered
but are associated with greater toxicity and similar local control.
SRS reports have clearly demonstrated that normal tissue effects are related to tech-
nique and dose, and outcomes from SRS have improved in the last decade as this has
been better understood.
Meningioma
Definitive SRS to smaller lesions or those at the skull base is effective with high rates
of local control and low morbidity, comparable to conventional fractionated RT. The
prescribed marginal dose is typically in the range 12–​16 Gy. Consideration should be
given to target volume, location, presence of mass effect or brain oedema, and prox-
imity to OARs. Parasagittal meningiomas are associated with a higher rate of symp-
tomatic treatment-​related oedema after SRS than those at other sites. Large inoperable
lesions abutting or encompassing OARs are best treated with fractionated RT but the
transition from one modality to the other depends on departmental experience, avail-
able technology and the clinical condition of the patient. Debulking or decompressive
surgery followed by SRS (or FSRT) to the residuum may be considered in select cases.
WHO grade II and III meningiomas have less discrete borders which warrant treat-
ment with margins and therefore a fractionated approach. However, salvage SRS after
initial radiotherapy may be offered for focally recurrent disease, regardless of grade.

Pituitary
SRS is mainly considered after trans-​sphenoidal surgery for small residual or recurrent
tumours, of < 2.5 cm. For non-​functioning adenomas, doses of 12–​16Gy give tumour
398 Central nervous system tumours

control in 95% of patients at 5 years. For hypersecreting adenomas, the preferred dose
is at least 20 Gy but biochemical remission occurs in only about 50% over a latency of
several years. Distance to optic pathway is the main determinant of SRS feasibility and
at least 2 mm is usually required to respect usual tolerances. The incidence of hypo-
pituitarism (though excluding evaluation of growth hormone) is very low at 5 years
post SRS when the mean dose to the pituitary is less than 15 Gy, but long-​term data
are lacking.
Although SRS may provide a convenient solution for small tumours away from
the chiasm, there is currently no evidence that it is superior to FSRT for pituitary
tumours.

(a)

Fig. 17.16  SRS plan for small vestibular schwannoma, using 9 static 3D conformal fields.
Immobilization: stereotactic relocatable mask (PTV = GTV + 1mm). Planning performed
with CT planning scan at 1-​mm axial slices, co-​registered with T1W + gadolimium at 1-​
mm axial slices (3D volumetric scan of whole head acquired in sagittal plane with 1-​mm
slice thickness, axial reconstruction) together with CISS/​FIESTA at 0.8-​mm slices (limited
skull base block).
(a) Full axial slice MRI (T1W + gadolinium). Enlarged views of CT (b), T1W + gadolimium MRI (d) and
CISS/​FIESTA MRI (e). Structures: GTV—​sky blue, PTV—​red, cochlea—​pink, brainstem—​green. The
trigeminal nerve is not seen in this view. (e) Full axial slice with dose shown on CT, for plan using 9
static 3D conformal fields; (f) enlarged view. Dose: 12 Gy to 80% isodose (15 Gy to isocentre) in 1
fraction. Doses: 100% red = 15 Gy, 95% orange, 90% light orange, 80% yellow/​green, 60% green,
40% sky blue, 20% blue.
 Stereotactic radiosurgery 399

(b)

(c) (d)

(e) (f)

Fig. 17.16 continued
400 Central nervous system tumours

17.13  Proton beam therapy for skull base and

spinal tumours
17.13.1 Introduction
The dosimetric characteristics of proton beam therapy (PBT) can be exploited in the
treatment of CNS and skull base tumours, particularly to deliver escalated doses to
target volumes close to critical OARs. However, uncertainties in the exact position of
the Bragg peak (‘distal edge uncertainty’) and of the relative biological effectiveness
(RBE), especially in the distal Bragg peak, mean that it is desirable to avoid beams
stopping at the edge of radiosensitive OARs, such as brainstem or spinal cord. The
direction of proton beams must be carefully chosen to avoid traversing tissues with
variable density, so that minor variation in the patient’s position does not result in a
major variation in the density of the tissue passed through. Proton therapy needs to be
carefully planned to produce ‘robust’ plans, and it is more likely that proton compared
to photon therapy will need to be replanned during a course of treatment (e.g. if air
sinuses fill with fluid).
A disadvantage of protons is their relative lack of skin sparing, so that the skin dose
is often higher than with photons, increasing the risk of permanent hair loss and
dermal toxicity. For deeper lying tumours this can be avoided by using several beams,
but for more superficial tumours high skin dose may be unavoidable.

17.13.2 Established indications
The most established indications for PBT in adult CNS practice are radioresistant
tumours of the skull base and spine, specifically chordomas and chondrosarcomas,
which require high dose RT above brainstem or spinal cord tolerance. PBT is also
indicated for ocular tumours (especially uveal melanomas), which can be treated at
facilities with only low energy beams.
Chordomas and chondrosarcomas are rare sarcomas which have been treated with
protons over several decades; one of the earliest programmes was based at the Harvard
Cyclotron Laboratory in collaboration with Massachusetts General Hospital in Boston.
These tumours have a clear dose-​response relationship and require high doses for local
control, ideally at least 68–​70 Gy for chondrosarcoma and 70–​78 Gy for chordoma, the
latter having an estimated TD50 of 65 Gy. Delivery of these doses is complicated by the
proximity of the brainstem and often the optic nerves and chiasm.
Surgery is needed for histological diagnosis and to achieve maximal safe debulking,
aiming particularly to clear tumour at least 2–​3  mm away from the brainstem and
ideally at least 5 mm away from optic chiasm and nerves, due to their lower tolerance.
Proton therapy has enabled dose escalation and higher rates of local control com-
pared to photon therapy. Massachusetts General reported a large series using passively
scattered protons (and a proportion of treatments with photons), with 10-​year local
control of 94% for 229 skull base chondrosarcomas, 65% for 159 male chordomas,
and 42% for 131 female chordomas(53). The Paul Scherrer Institute in Switzerland,
which developed spot scanning, has reported results using intensity modulated proton
therapy (IMPT) with 7-​year local control of 94% for 71 chondrosarcomas and 71% for
151 chordomas(54); residual tumour volume < 25 cm3 was prognostic.
 References 401

Massachusetts General has also reported on patients treated with PBT for spinal
chondrosarcomas and chordomas; for both tumour types, en bloc resection and nega-
tive margins were associated with better survival. The Paul Scherrer Institute has re-
ported excellent 5-​year outcomes for spinal chordoma even in the presence of residual
tumour, except in patients with titanium-​based spinal stabilization in whom results
were much inferior. This result has had an important influence on practice.

17.13.3  Assessment of new indications

The question of which additional indications might benefit which patients is an area
of considerable interest(55). Currently, PBT is being used in a wide range of adult CNS
tumours. However, the majority are single arm observational studies. Direct compari-
sons with photon treatment are undoubtedly difficult. Duration of follow-​up must also
be sufficient to define relevant late endpoints. Randomized studies are desirable, where
they are feasible, and novel trial designs and methodologies are needed.
The Particle Therapy Cooperative Group (PTCOG) publishes a summary of clinical
trials in proton therapy [https://​www.ptcog.ch/​index.php/​clinical-​protocols].

Acknowledgements
We would particularly like to thank Mr Tony Geater, Mrs Karen Wildschut, and Dr
Andrew Hoole for help in the preparation of the diagrams.

References
1. Medical Research Council Brain Tumour Working Party. Prognostic factors for high-​
grade malignant glioma: development of a prognostic index. A Report of the Medical
Research Council Brain Tumour Working Party. Journal of Neuro-​oncology 1990; 9:47–​55.
2. ICRU. ICRU Report 62: Prescribing, Recording and Reporting Photon Beam Therapy
(Supplement to ICRU Report 50). Bethesda, MD: International Commission on Radiation
Units and Measurements, 1999.
3. ICRU. ICRU Report 83: Prescribing, Recording, and Reporting Intensity-​Modulated
Photon-​Beam Therapy (IMRT). International Commission on Radiation Units and
Measurements. Journal of the ICRU 2010; 10(1).
4. British Institute of Radiology. Geometric Uncertainties in Radiotherapy—​defining the
planning target volume. London: British Institute of Radiology, 2003.
5. Scoccianti S, Detti B, Gadda D, et al. Organs at risk in the brain and their dose-​constraints
in adults and in children: a radiation oncologist’s guide for delineation in everyday practice.
Radiotherapy and Oncology 2015; 114:230–​38.
6. Withers HR Biologic basis of radiation therapy. In: Perez C.A. and Brady L.W., Editors.
Principles and Practice of Radiation Oncology. 2nd edition. Philadelphia, PA: J.B. Lippincott,
1992, pp. 64–​98.
7. Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to therapeutic irradiation.
International Journal of Radiation Oncology, Biology, Physics 1991; 21:109–​22.
8. Marks LB, Ten Haken RK, Martel MK. Guest editor’s introduction to QUANTEC: a user’s
guide. International Journal of Radiation Oncology, Biology, Physics 2010; 76(3 Suppl):S1–​2.
9. Lawrence YR, Li XA, el Naqa I, et al. Radiation dose-​volume effects in the brain.
International Journal of Radiation Oncology, Biology, Physics 2010; 76(3 Suppl):S20–​7.
402 Central nervous system tumours

10. Taphoorn MJ, Schiphorst AK, Snoek FJ, et al. Cognitive functions and quality of life in
patients with low-​grade gliomas: the impact of radiotherapy. Annals of Neurology 1994;
36:48–​54.
11. Kiebert GM, Curran D, Aaronson NK, et al. Quality of life after radiation therapy of
cerebral low-​grade gliomas of the adult: results of a randomised phase III trial on dose
response (EORTC trial 22844). EORTC Radiotherapy Co-​operative Group. European
Journal of Cancer 1998; 34:1902–​9.
12. Gregor A, Cull A, Traynor E, et al. Neuropsychometric evaluation of long-​term survivors
of adult brain tumours: relationship with tumour and treatment parameters. Radiotherapy
and Oncology 1996; 41:55–​9.
13. Brown PD, Buckner JC, Uhm JH, Shaw EG. The neurocognitive effects of radiation in
adult low-​grade glioma patients. Neurooncology 2003; 5:161–​7.
14. Klein M, Heimans JJ, Aaronson NK, et al. Effect of radiotherapy and other treatment-​
related factors on mid-​term to long-​term cognitive sequelae in low-​grade gliomas: a
comparative study. Lancet 2002; 360: 1361–​8. Erratum in: Lancet 2011; 377: 1654.
15. Douw L, Klein M, Fagel SS, et al. Cognitive and radiological effects of radiotherapy in
patients with low-​grade glioma: long-​term follow-​up. Lancet Neurology 2009; 8:810–​18.
16. Mayo C, Yorke E, Merchant TE. Radiation associated brainstem injury. International
Journal of Radiation Oncology, Biology, Physics 2010; 76(3 Suppl): S36–​41.
17. Kirkpatrick JP, van der Kogel AJ, Schultheiss TE. Radiation dose-​volume effects in
the spinal cord. International Journal of Radiation Oncology, Biology, Physics 2010; 76(3
Suppl): S42–​9.
18. Mayo C, Martel MK, Marks LB, et al Radiation dose-​volume effects of optic nerves
and chiasm. International Journal of Radiation Oncology, Biology, Physics 2010; 76(3
Suppl): S28–​35.
19. Stupp R, Tonn JC, Brada M, Pentheroudakis G. ESMO Guidelines Working Group.
High-​grade malignant glioma: ESMO Clinical Practice Guidelines for diagnosis, treatment
and follow-​up. Annals of Oncology 2010; 21(Suppl 5):v190–​3.
20. Bhandare N, Jackson A, Eisbruch A, et al. Radiation therapy and hearing loss.
International Journal of Radiation Oncology, Biology, Physics 2010; 76(3 Suppl): S5-​–​7.
21. Henk JM, Whitelocke RA, Warrington AP, Bessell EM. Radiation dose to the lens and
cataract formation. International Journal of Radiation Oncology, Biology, Physics 1993;
25: 815–​20.
22. Lawenda BD, Gagne HM, Gierga DP, et al. Permanent alopecia after cranial
irradiation: Dose-​response relationship. International Journal of Radiation Oncology,
Biology, Physics 2004; 60(3): 879–​87.
23. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus concomitant and adjuvant
temozolomide for glioblastoma. New England Journal of Medicine 2005; 352:987–​96.
24. van den Bent MJ, Brandes AA, Taphoorn MJ, et al. Adjuvant procarbazine, lomustine, and
vincristine chemotherapy in newly diagnosed anaplastic oligodendroglioma: long-​term
follow-​up of EORTC brain tumor group study 26951. Journal of Clinical Oncology 2013;
31:344–​50.
25. Cairncross G, Wang M, Shaw E, et al. Phase III trial of chemoradiotherapy for
anaplastic17. oligodendroglioma: long-​term results of RTOG 9402. Journal of Clinical
Oncology 2013; 31:337–​43
 References 403

26. Niyazi M, Brada M, Chalmers AJ, et al. ESTRO-​ACROP guideline ‘target delineation of
glioblastomas’. Radiotherapy and Oncology 2016; 118:35–​42.
27. Minniti G, Amelio D, Amichetti M, et al. Patterns of failure and comparison of different
target volume delineations in patients with glioblastoma treated with conformal
radiotherapy plus concomitant and adjuvant temozolomide. Radiotherapy and Oncology
2010; 97: 377–​8.
28. Malmström A, Grønberg BH, Marosi C, et al.; Nordic Clinical Brain Tumour Study Group
(NCBTSG). Temozolomide versus standard 6-​week radiotherapy versus hypofractionated
radiotherapy in patients older than 60 years with glioblastoma: the Nordic randomised,
phase 3 trial. Lancet Oncology 2012; 13:916–​26.
29. Karim AB, Afra D, Cornu P, et al. Randomized trial on the efficacy of radiotherapy for
cerebral low-​grade glioma in the adult: European Organization for Research and Treatment
of Cancer Study 22845 with the Medical Research Council study BRO4: an interim
analysis. International Journal of Radiation Oncology, Biology, Physics 2002; 52:316–​24.
30. Buckner JC, Shaw EG, Pugh SL et al. Radiation plus Procarbazine, CCNU, and Vincristine
in Low-​Grade Glioma. New England Journal of Medicine 2016; 374:1344–​55.
31. Baumert BG, Hegi ME, van den Bent MJ, et al. Temozolomide chemotherapy versus
radiotherapy in high-​risk low-​grade glioma (EORTC 22033-​26033): a randomised, open-​
label, phase 3 intergroup study. Lancet Oncology 2016; 17:1521–​32.
32. Shaw E, Arusell R, Scheithauer B, et al. Prospective randomized trial of low-​versus high-​
dose radiation therapy in adults with supratentorial low-​grade glioma: initial report of
a North Central Cancer Treatment Group/​Radiation Therapy Oncology Group/​Eastern
Cooperative Oncology Group study. Journal of Clinical Oncology 2002; 20:2267–​76.
33. Wu J, Armstrong TS, Gilbert MR. Biology and management of ependymomas. Neuro-​
oncology 2016; 18:902–​13
34. Merchant TE, Li C, Xiong X, et al. Conformal radiotherapy after surgery for paediatric
ependymoma: a prospective study. Lancet Oncology 2009; 18: 258–​66.
35. Mendenhall WM, Friedman WA, Amdur RJ, Antonelli PJ. Management of acoustic
schwannoma. American Journal of Otolaryngology 2004; 25:38–​47.
36. Fuss M, Debus J, Lohr F, et al. Conventionally fractionated stereotactic radiotherapy
(FSRT) for acoustic neuromas. International Journal of Radiation Oncology, Biology, Physics
2000; 48: 1381–​7.
37. Rowe JG, Radatz MW, Walton L, et al. Gamma knife stereotactic radiosurgery for
unilateral acoustic neuromas. Journal of Neurology, Neurosurgery and Psychiatry 2003;
74: 1536–​42.
38. Meijer OW, Vandertop WP, Baayen JC, Slotman BJ. Single-​fraction vs. fractionated
linac-​based stereotactic radiosurgery for vestibular schwannoma: a single-​institution study.
International Journal of Radiation Oncology, Biology, Physics 2003; 56:1390–​6.
39. Lunsford LD, Niranjan A, Flickinger JC, et al. Radiosurgery of vestibular
schwannomas: summary of experience in 829 cases. Journal of Neurosurgery 2013; 119
Suppl:195–​9.
40. Royal College of Physicians Working Party. Pituitary tumours: recommendations for
service provision and guidelines for management of patients. London, Royal College of
Physicians of London, 1997.
41. Brada M, Rajan B, Traish D, et al. The long term efficacy of conservative surgery and
radiotherapy in the control of pituitary adenomas. Clinical Endocrinology 1993; 38: 571–​8.
404 Central nervous system tumours

42. Harrabi SB, Adeberg S, Welzel T, et al. Long term results after fractionated stereotactic
radiotherapy (FSRT) in patients with craniopharyngioma: maximal tumor control with
minimal side effects. Radiation Oncology 2014; 9:203.
43. Sattler MG, Vroomen PC, Sluiter WJ, et al. Incidence, causative mechanisms, and
anatomic localization of stroke in pituitary adenoma patients treated with postoperative
radiation therapy versus surgery alone. International Journal of Radiation Oncology,
Biology, Physics 2013; 87:53–​9.
44. Merchant TE, Kun LE, Hua CH, et al. Disease control after reduced volume conformal and
intensity modulated radiation therapy for childhood craniopharyngioma. International
Journal of Radiation Oncology, Biology, Physics. 2013; 85: e187–​92.
45. Jenkinson MD, Javadpour M, Haylock BJ, et al. The ROAM/​EORTC-​1308 trial: radiation
versus observation following surgical resection of Atypical Meningioma: study protocol for
a randomised controlled trial. Trials 2015; 16:519.
46. Akyurek S, Chang EL, Yu TK, et al. Spinal myxopapillary ependymoma outcomes in
patients treated with surgery and radiotherapy at M.D. Anderson Cancer Center. Journal of
Neuro-​Oncology 2013; 80: 177–​83.
47. Waldron JN, Laperriere NJ, Jaakkimainen L, et al. Spinal cord ependymomas: a
retrospective analysis of 59 cases. International Journal of Radiation Oncology, Biology,
Physics 1993; 27: 223–​9.
48. Aoyama H, Shirato H, Tago M, et al. Stereotactic radiosurgery plus whole-​brain radiation
therapy vs stereotactic radiosurgery alone for treatment of brain metastases: a randomized
controlled trial. Journal of the American Medical Association 2006; 295:2483–​91.
49. Kocher M, Soffietti R, Abacioglu U, et al. Adjuvant whole-​brain radiotherapy
versus observation after radiosurgery or surgical resection of one to three cerebral
metastases: results of the EORTC 22952-​26001 study. Journal of Clinical Oncology 2011;
29:134–​41.
50. Brown PD, Jaeckle K, Ballman KV, et al. Effect of radiosurgery alone vs radiosurgery
with whole brain radiation therapy on cognitive function in patients with 1 to 3 brain
metastases: A randomized clinical trial. Journal of the American Medical Association 2016;
316:401–​9.
51. Pinkham MB, Whitfield GA, Brada M. New developments in intracranial stereotactic
radiotherapy for metastases. Clinical Oncology (Royal College of Radiologists) 2015;
27:316–​23.
52. Pinkham MB, Sanghera P, Wall GK, et al. Neurocognitive effects following cranial
irradiation for brain metastases. Clinical Oncology (Royal College of Radiologists) 2015;
27:630–​39.
53. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the skull base. Strahlentherapie
und Onkologie 1999; 175 Suppl 2:57–​63.
54. Weber DC, Malyapa R, Albertini F, et al. Long term outcomes of patients with skull-​base
low-​grade chondrosarcoma and chordoma patients treated with pencil beam scanning
proton therapy. Radiotherapy and Oncology 2016; 120:169–​74.
55. Mishra MV, Aggarwal S, Bentzen SM, et al. Establishing evidence-​based indications for
proton therapy: an overview of current clinical trials. International Journal of Radiation
Oncology, Biology, Physics 2017; 97:228–​35
 Chapter 18

Head and neck cancer

Christopher Nutting and Dorothy Gujral

18.1 Introduction
Cancer of the head and neck is a relatively rare cancer accounting for 3% of all cancer
deaths. The incidence in men is 17.2 per 100,000 and in women, 5.6 per 100,000. This
equates to about 7750 cases in the UK(1). The main aetiological factors are excessive
alcohol intake, infection with human papilloma virus (HPV), and smoking.
The tumour, node, metastasis (TNM) staging system is used for staging head and
neck cancer. A current generic staging system is given in Table 18.1, but more detail on
staging for individualized tumour subsites is given in the American Joint Committee
on Cancer (AJCC) TNM Classification(2).

18.2  Radical primary treatment

Patients with head and neck cancer should be discussed by a multidisciplinary team
comprising specialist surgeons, oncologists, pathologists, radiologists, and palliative
care doctors, together with dieticians, speech and language therapists, and clinical
nurse specialists. At this time, decisions as to the modality of treatment(s) to be used
should be made, including surgery, radiotherapy, and chemotherapy.

18.2.1 Indications
Surgery and radiotherapy with or without chemotherapy are the most frequently used
therapeutic modalities in head and neck cancer. For early stage tumours surgical ex-
cision or radiotherapy alone have similar cure rates, but have different adverse effect
profiles. Radiotherapy with or without chemotherapy offers higher rates of organ pres-
ervation, and for some cancers, where function is important, it is the treatment of
choice. For example, in carcinoma of the tongue base or larynx, radiotherapy preserves
swallowing and natural speech respectively. At other sites (e.g. carcinoma of the floor
of mouth), surgical excision alone may be curative and be associated with a very sat-
isfactory functional outcome. The choice of treatment modality therefore depends on
individual factors including patient preference.
For advanced squamous cell carcinoma of the head and neck the combined use
of surgery and postoperative radiotherapy frequently offers the highest chance of
achieving cure. In the light of international trials in the postoperative setting(3,4), con-
comitant chemoradiotherapy has become the standard of care for high-​risk patients
with positive margins or extracapsular spread. Similarly, for inoperable advanced
406 Head and neck cancer

Table 18.1  (a) Generic staging for head and neck cancer (for more detail
on individual subsites, refer to AJCC TNM Classification(2))
Tis Carcinoma in situ

T1 Tumour < 2 cm

T2 Tumour > 2–​4 cm
T3 Tumour > 4 cm
T4 Tumour involves adjacent structures
T4a Operable disease
T4b Inoperable disease
(b) Nodal staging for all head and neck sites (this can be applied to all sites
except nasopharynx carcinoma and thyroid)
N0 No lymphadenopathy
N1 Ipsilateral single node < 3 cm
N2a Ipsilateral single node > 3–​6 cm
N2b Ipsilateral multiple nodes < 6 cm
N2c Bilateral or contralateral nodes < 6 cm
N3 Nodes > 6 cm

Reproduced with permisson from Christian Wittekind, Hisao Asamura, Leslie H. Sobin, 'Head
and Neck Tumours' in TNM Online. https://​doi.org/​10.1002/​9780471420194.tnma01
Copyright © 2017 UICC, published in 2017 by John Wiley & Sons, Inc.

tumours, concomitant chemoradiation schedules offer the highest chance of local con-
trol and survival(5).

18.2.2  General principles of treatment volume

and definition
Patients with head and neck cancer are initially seen in a multidisciplinary outpatient
clinic when a full history and examination of the head and neck area will be carried
out; this will include nasendoscopy if indicated. Following this the pretreatment as-
sessment for radiotherapy planning should include an examination under anaesthesia
(EUA) and tumour biopsy, thoracic computed tomography (CT) scan, CT and/​or
magnetic resonance imaging (MRI) of head and neck, specialist histology review, full
blood count, urea and electrolytes, liver function tests, dental assessment, nutritional
assessment, and written informed consent. Positron emission tomography (PET)-​CT
scanning is indicated for patients with locally advanced disease. Patients should be ad-
vised to stop smoking as this increases toxicity and reduces cure rates, and to reduce
alcohol intake. Chemoradiation protocols are associated with enhanced acute toxicity
and the patient’s performance status, medical co-​morbidities, and social factors should
be taken into account during patient selection. An accurate assessment of dietary in-
take should be undertaken, and, if necessary, a nasogastric tube or percutaneous gas-
trostomy may be required for elective feeding.
 Radical primary treatment 407

Most patients with advanced tumours have their radiotherapy delivered using highly
conformal intensity-​modulated radiotherapy (IMRT) to cover the primary tumour/​
subsite, involved nodal levels, and nodal levels of likely microscopic lymphatic spread.
All sites are treated to different dose levels at the same time using a simultaneous in-
tegrated boost technique. The elective nodal sites depend on the site of the primary
tumour. Table 18.2 details the incidence of occult micrometastases to lymph nodes
which have been documented from surgicopathological series(6,7). Elective irradiation
of lymph node regions is indicated when the risk of harbouring micrometastatic dis-
ease exceeds 15–​20%, and therefore a detailed understanding of the natural history
of each subsite of head and neck cancer is required during radiotherapy planning.
Lymph node levels are defined as: level Ia, submental; level Ib, submandibular; level
II, upper jugular; level III, middle jugular; level IVa, lower jugular; level IVb, medial
supraclavicular; level Va + b, posterior triangle; level Vc, lateral supraclavicular; level
Via, anterior jugular; level VIb, pre-​laryngeal, pre-​tracheal, para-​tracheal, recurrent
laryngeal nerve nodes; level VIIa, retropharyngeal; level VIIb, retrostyloid nodes;
level VIII, parotid node group; level IX, bucco-​facial; level Xa, retroauricular, level Xb,
occipital(8)(Fig. 18.1). For IMRT, a planning target volume (PTV) should be generated
for the tissues to receive radical radiation dose (PTV1) and a second volume is defined

Table 18.2  The risk of nodal metastasis based

on tumour site and location
High (> 60%) Nasopharynx

Oropharynx
Hypopharynx
Supraglottis
Medium (20–​60%) Oral cavity
Advanced larynx
Parotid
Low (< 20%) Skin
Early stage glottic larynx
Nasal cavity
Paranasal sinuses
Risk predominately unilateral Parotid
Early stage tonsil
Lateralized oral cavity
Risk bilateral Base of tongue
Nasopharynx
Advanced larynx/​
hypopharynx
408 Head and neck cancer

Ib II
Ia

III
Fig. 18.1  The lymph
node levels in the neck.
V I: submental (Ia) and
IV
submandibular (Ib),
II: upper deep cervical,
SCF III: middle deep cervical,
IV: lower deep cervical,
V: posterior triangle,
SCF: supraclavicular fossa.

(PTV2) to contain the tissues to be irradiated to an elective dose. Some oncologists

opt to treat the rest of the uninvolved subsite to an intermediate dose; thus, an inter-
mediate dose level is defined (PTV3).
The recommendations for specific node groups to be included in the field of treat-
ment are meant as a guide (Table 18.3). The responsible clinical oncologist must make
the final decision on the individual features of the case.

18.2.3 Planning technique
Patient position and immobilization
The anatomy of the head and neck region is very complex, with bony structures, soft
tissues, and air cavities all present in complicated arrangements in a relatively small
volume. The organs at risk (OARs) include spinal cord, brainstem, optic nerves, retina,
lens, brain, skin, mucosa, and salivary glands. All may lie very close to, or within, the
target volume, making irradiation of tumour within normal tissue tolerance difficult.
By contrast, internal organ motion is relatively limited, and even physiological laryn-
geal motion has little impact on treatment planning. The head and neck region can
be readily immobilized using a custom-​made thermoplastic shell (Fig. 18.2) and this
should ensure reproducible patient set-​up to within 2–​3 mm. Head and neck stabil-
ization systems involving carbon fibre base plates to enable IMRT to be delivered are
standard, replacing the older style Cabulite shells.
It is important to specify neck and head position, shell extent, requirements for
mouth bite, and full planning details prior to manufacturing the immobilization shell.
For most tumour sites, parotid gland-​sparing IMRT is now standard to avoid radiation-​
induced xerostomia(9). An extended neck position is required for all patients treated with
IMRT to avoid the anterior radiation beams irradiating the mucosa of the oral cavity.
 Radical primary treatment 409

Table 18.3  Recommendations for elective lymph node irradiation in radical treatment

with radiotherapy for head and neck cancer
Nasopharynx

Squamous carcinoma T1–​T4 N0 Level VIIa/​Level II to V bilaterally

All undifferentiated carcinoma or Level VIIa/​Level Ib to V bilaterally
node-​positive squamous carcinoma
Paranasal sinuses
Squamous carcinoma N0 Level VIIa
Squamous carcinoma N+ Level VIIa/​Levels II to V bilaterally (add level Ib on
N+ side)
Oropharynx
T1/​2 N0 tonsil (well-​lateralized) Levels Ib to IV ipsilateral
T1/​T2 N1 tonsil Levels Ib to V ipsilateral
T2 N0 tonsil approaching midline + Levels Ib to IV bilaterally
other sites
All others Levels Ib to V bilaterally
Larynx
T1/​T2 N0 glottic No elective nodal irradiation
T3–​T4 N0 glottic Levels II,to IV bilaterally
T1/​T2 N0 supraglottic Levels II and III bilaterally
All others Levels II to V bilaterally (add level Ib on N+ side)
Hypopharynx
All Levels II to V bilaterally (add level Ib on N+ side)

For patients receiving treatment with conformal 3-​dimensional radiotherapy, the

head is positioned in the neutral position and the spinal cord straight, except for
(i) parotid tumours, where the neck is extended to elevate the eyes above the radiation
fields and (ii) irradiation of the unilateral neck only.
Mouth bites are used in patients with carcinoma of the oral cavity, nasal cavity, and
maxillary sinuses. The mouth bite opens the jaw, and depresses the tongue. It may be
possible to exclude either the upper or lower half of the mouth from the treatment field.
Care should be taken to avoid the mouth bite pushing the mobile tongue into or out of the
field posteriorly, and to ensure that the tongue position is reproducible. A custom-​made
mouthpiece to move the tongue out of the field in the lateral position should be con-
sidered in unilateral carcinomas of the oral cavity, e.g. floor of mouth or buccal mucosa.

Conventional radiotherapy (virtual simulation)

Conventional radiotherapy planning is usually reserved for palliative treatment and early
(T1/​2N0) larynx cancer. Virtual simulation using a CT simulator is now standard for pal-
liative radiotherapy and improves definition of tumour and OARs in these indications.
410 Head and neck cancer

Fig. 18.2  A custom-​made
thermoplastic shell.

Most commonly lateral parallel-​opposed beams are used to irradiate the target
volume. Typically field borders are placed in relation to standard bony anatomical
landmarks that define the extent of tumour subsites, and may be modified in indi-
vidual patients. This planning method may use the GTV, CTV, and PTV definitions
outlined in ICRU 50 and 62(10,11), the field borders representing the PTV plus a physical
margin for penumbra. The clinical results obtained with these techniques are known,
as are the expected toxicities of treatment.

Computed tomography planning
CT planning is the standard of care for head and neck cancer radiotherapy. For CT
planning the recommendations of ICRU reports 50 and 62(9,10) should be followed.
Clinicians should define the GTV, CTV, PTV, and OARs. Outlining of the GTV and
CTV should be done with the aid of diagnostic MRI, and CT or PET-​CT scans, oper-
ation notes, clinical examination, and nasendoscopy.
A planning CT scan with intravenous contrast will provide better definition of the
primary tumour and involved nodes and should always be performed. Joint review
with a radiologist is suggested. The addition of CT information into the treatment
planning process improves localization of both tumour and OARs. It can also lead to
more accurate dose calculation, and allows better estimation of dose received by the
target volumes and OARs. In addition, the use of CT planning allows optimization of
radiotherapy beam direction, beam weight, accurate generation of conformal beam
shaping, and is essential for inverse planning and IMRT. The following sections discuss
the application of cross-​sectional imaging to target volume definition.

Gross tumour volume localization

Head and neck squamous cell carcinoma often spreads through the mucosa, and such
spread is usually best assessed clinically or endoscopically as this type of tumour ex-
tension is not reliably identified from cross-​sectional imaging (CT, MRI, or PET-​CT).
 Radical primary treatment 411

If a tumour is deeply invading tissues, this is better identified from CT or MRI(12). The
sensitivity of such imaging modalities depends on the difference in signal between
tumour and normal tissue, which is variable depending on tumour site (e.g. poor for
tongue tumours, but good for paranasal sinus tumours). For tumours invading the
skull base (e.g. nasopharyngeal carcinoma or paranasal carcinoma), both CT and MRI
may be optimal for detection of bone invasion and soft tissue extension respectively.
For adjuvant radiotherapy, postoperative imaging may be useful to define the tumour
bed especially if radio-​opaque markers have been left by the surgeon to delineate the
tumour bed and guide radiotherapy planning. If a patient has received neo-​adjuvant
chemotherapy, the pre-​chemotherapy tumour volume should be used for planning.
CT scanning is the most frequently used modality for target volume definition.
Postscanning adjustment of CT window levels is of value in defining both tumour
and OARs.
MRI treatment planning for head and neck cancer is under evaluation. Although
this may allow better differentiation of tumour from normal tissue (typically with T1-​
weighted, gadolinium enhanced images), MRI/​CT fusion is required to correct dis-
tortion before treatment planning or dose calculations can be performed. Volumes
outlined on MRI are smaller compared to CT, and there is less interobserver vari-
ability(12). Image fusion using bony anatomy is accurate for tumours in, or attached to,
the skull base, but is less satisfactory for images of the neck that are relatively mobile
due to flexion and extension of the cervical spine. PET-​CT is also increasingly used for
treatment planning with PET-​CT fusion. Patients can be scanned in their treatment
shell to allow for better co-​registration with the planning CT.

Clinical target volume localization

CTV definition for radical radiotherapy is controversial. A margin needs to be added
to the GTV to take into account patterns of local spread of the tumour. This requires
knowledge of the patterns of local tumour extension, anatomical compartments, and
barriers. Where the tumour may spread, e.g. submucosally, the margin needs to be
1–​2 cm, but where there are anatomical boundaries to tumour extension such as air
cavities, periosteum, and compartmental fascia, then margins can be reduced.
Adjuvant (postoperative) radiotherapy to the primary tumour site represents
an even more difficult localization problem. By definition there is no GTV, and so
the localization of the tumour bed is based on preoperative descriptions and cross-​
sectional imaging. Preoperative assessment by the radiation oncologist is invaluable.
Preoperative and postoperative imaging may be fused to improve CTV definition.

Nodal target volume definition

Considerable advances have been made to produce guidelines for reproducible elective
nodal delineation. The use of intravenous contrast is recommended as it helps to de-
fine the carotid arteries, an important anatomical landmark in the definition of the
deep cervical nodes(13).
Frequently there are locoregional lymph node metastases in addition to the pri-
mary tumour, and in this situation individual nodal masses require a separate GTV.
A margin of 1–​2 cm should be added to each nodal GTV to produce the CTV. As for
412 Head and neck cancer

primary tumours, the GTV–​CTV margin can be reduced where there is an anatomical
barrier to tumour spread, e.g. prevertebral fascia, air cavity, or bony wall. For situations
in which the target volume cannot be well defined, an approach of conformal avoid-
ance may be useful when tissues at risk of containing disease are outlined, and normal
tissue structures are removed from the volume.
In the case of adjuvant radiotherapy, radiation is delivered to a CTV where there
is a risk of residual microscopic disease following radical surgical excision. Several
CTVs will usually be identified. A  different CTV should be defined for regional
nodal groups at risk of containing microscopic metastases. For example, in the treat-
ment of early (e.g. T1 N1 M0) carcinoma of the tonsil there will be a GTV1/​CTV1
for the primary tumour, and a different GTV2/​CTV2 for the enlarged lymph node
mass. The nodal CTV2 will also include the adjacent lymph nodes at high risk of in-
volvement (the first lymph node station—​the ipsilateral upper deep cervical lymph
node group). These would be prescribed dose equivalent to 70 Gy. The ipsilateral
clinically uninvolved levels Ib, III, IV, and V, which are at less risk of lymph node
spread, would be defined as a separate nodal CTV3, and may be prescribed a dose
equivalent to 50 Gy to sterilize microscopic foci of metastatic carcinoma. The risk
of nodal metastases varies between different tumour sites and has been extensively
investigated in retrospective surgical studies. These clinicopathological studies(6,7,9)
examined elective neck lymph node dissection specimens for the presence of occult
metastases for individual tumour sites and documented their frequency and distri-
bution. These studies have been collated and recommendations given for elective
nodal irradiation(14).
The identification of nodal volumes varies depending on the technique used.
From 70 to 75% of lymph nodes involved by tumour are enlarged, and can be iden-
tified by clinical palpation. The use of cross-​sectional imaging increases the sen-
sitivity to 85% by the identification of impalpable retropharyngeal nodes, deeply
sited nodes, and normal sized nodes with a necrotic centre (necrosis has low signal
intensity on CT scan and is highly specific for metastatic carcinoma). 18F-​PET-​CT
scanning may further improve diagnostic sensitivity. The number of CTVs outlined
should be representative of the estimated clonogenic cell density and, in practice,
contiguous groups of nodes that are to receive the same dose can be outlined as
one CTV.

Planning target volume localization

A further margin is added to the CTV to produce the PTV. The size of this margin de-
pends on the type of immobilization used. A customized thermoplastic mask system
has a day-​to-​day set-​up accuracy of 2–​5 mm, and to encompass 95% of the errors a
CTV–​PTV margin of 3–​5 mm should be added. For tumours attached to the bones of
the skull base (nasopharynx, paranasal sinuses, oropharynx, and nasal cavity) a tight
shell provides a random error of < 2 mm. For oral cavity tumours, movement of the
mandible may affect the reproducibility of patient set-​up, and this will be greatest if
the mouth is stented open with a mouth bite. For tumours of the hypopharynx and
larynx, physiological organ motion occurs with swallowing and breathing which must
be taken into account in the CTV–​PTV internal movement margin.
 Radical primary treatment 413

CTV–​PTV margin should also be added to elective lymph node irradiation volumes
to account for set-​up inaccuracy. The size of these margins will be dependent on the
type of immobilization used but should be in the region of 3–​5 mm.

Organs at risk
Brain, spinal cord, brainstem, parotid glands, mandible, eyes, optic nerves, and chiasm,
are all close to the target volume for some head and neck tumours. Tolerance doses are
given in Table 18.4(15). These structures should all be outlined in their entirety on CT
planning scans.
OARs that occur close to the PTV should be outlined. The particular organs will
vary from one tumour site to another.

Intensity-​modulated radiotherapy
IMRT is now the standard of care for most patients with head and neck cancer.
Typically five or seven equispaced fixed intensity-​modulated beams, or a 360° dynamic
arc are used. IMRT can generate a distribution with a concave shape to reduce dose
to OARs lying close to the target volume. This has been shown to reduce long-​term
treatment-​related toxicity. Parotid-​gland sparing IMRT has been shown to reduce
long-​term xerostomia in patients with tumours of the oropharynx or nasopharynx(8)
(Fig. 18.3). It has also been shown to reduce the dose to the optic apparatus in treat-
ment of paranasal sinus tumours, and cochlea sparing in parotid tumours.
Studies using IMRT to escalate radiation dose and improve local tumour control are
currently under evaluation.

Table 18.4  Acceptable doses to specific organs at risk. These

doses should not be exceeded unless the extent of tumour makes
higher doses inevitable(14)
Organ Normal tissue tolerance (2 Gy/​fraction)
Lens 6–​10 Gy

Cornea 40 Gy
Retina 50 Gy
Optic nerve 50 Gy TD 5/​50, 60 Gy TD 20/​50
Optic chiasm 50–​55 Gy
Spinal cord 44–​48 Gy
Brainstem 54 Gy
Hypothalamus 44 Gy
Parotid 24 Gy
Mandible 70 Gy

Reproduced with permisson from Levendag P, Braaksma M, Coche E, et al.,

‘Rotterdam and Brussels CT-based neck nodal delineation compared with the
surgical levels as defined by the American Academy of Otolaryngology-Head
and Neck Surgery’, International Journal of Radiation Oncology, Biology, Physics,
2004; 58: pp.113–23. Copyright © 2004 Elsevier Inc. All rights reserved.
414 Head and neck cancer

Fig. 18.3  Parotid gland sparing IMRT: a

dose distribution to deliver a high dose
to the target volume (blue contour and
red colour wash) whilst sparing the
parotid gland (pink contours) can be
achieved with IMRT.

For locally advanced head and neck cancer, it is most efficient to treat with simul-
taneous integrated boost (SIB) or simultaneous modulated accelerated radiotherapy
(SMART) techniques. These are characterized by the delivery of different dose-​
per-​fraction to different targets within the head and neck region. For example, in
PARSPORT, the Cancer Research UK Parotid Sparing IMRT trial(9), a dose of 2.17 Gy
per fraction was delivered to the primary tumour site, and involved lymph nodes, and
1.8 Gy per fraction to elective lymph node groups. After 30 fractions the primary tu-
mour and involved lymph nodes had received a total of 65 Gy, and the elective lymph
nodes 54 Gy (Fig. 18.4). The advantage of the SIB or SMART techniques is that the
whole treatment course is planned only once, with savings in simulation, planning,
delivery, and verification time compared to multi-​phase plans. Radiobiologically, SIB
and SMART techniques represent accelerated fractionation schedules that have been
shown to improve tumour control.
Inverse planning for head and neck IMRT requires the clinician to generate con-
straints in the form of dose volume points to drive the inverse planning algorithm.
These constraints will vary from one planning system to another, but target volume
constraints should be the prescription dose ± 5% and for OAR the tolerance dose to a
small volume of that organ.

18.2.4 Dose prescription
Plans should be normalized to the ICRU reference point(10). The plan should be checked
to ensure adequate target coverage, homogeneous dose, and that doses to OARs are
acceptable. Plans are prescribed to the 100% isodose—​usually to the isocentre, or a
similar representative point in the target volume for 3-​D conformal plans and to the
mean or median for IMRT plans.
In the treatment of squamous cell carcinoma of the head and neck, tumour and
normal tissue will usually be treated close to tolerance to achieve cure. Conventional
 Radical primary treatment 415

Fig. 18.4  SMART boost technique used in PARSPORT trial showing a higher total dose (65
Gy) and dose per fraction (2.17 Gy) delivered to the primary tumour and involved nodes
(red in 3D reconstruction and green colour wash) and lower total (54 Gy) dose and dose per
fraction (1.8 Gy) to the elective nodes (purple in 3D reconstruction and orange colour wash).

radiotherapy involves daily radiotherapy, Monday to Friday, giving a dose of 1.8–​2 Gy

per fraction to a total dose of 66–​70 Gy. Some centres, however, use a 4-​week schedule
of 55 Gy in 20 fractions daily, Monday to Friday, over 4 weeks and this shorter schedule
can be adopted for small volume disease.

Treatment of small volume disease

This encompasses most early stage disease, i.e. T1/​T2 N0 where the target volume in-
cludes the primary site and in some cases the first echelon of nodes only. Many centres
use a 4-​week schedule (e.g. 55 Gy in 20 fractions once daily for 4 weeks) for smaller
volume irradiation, i.e. glottic tumours with field size 5–​6 cm, on the basis that smaller
volumes will tolerate the larger doses per fraction better. Most US studies do not give
doses per fraction greater than 1.8–​2 Gy, however, irrespective of treatment volume.

Treatment of large volume disease

This encompasses T3/​T4 N0–​N3 disease where surgical treatment is either not ap-
propriate or not possible. The target volume includes the primary site and the lymph
nodes containing metastatic carcinoma and at-​risk (elective) nodes. Macroscopic dis-
ease should receive 66–​70 Gy in 2-​Gy fractions or equivalent. Microscopic disease
should receive 50 Gy in 2-​Gy fractions or equivalent.
It is generally accepted that 50 Gy is required to sterilize microscopic disease.
There is some evidence from randomized controlled trials, including the continuous
hyperfractionated accelerated radiotherapy (CHART) trials(16) and other phase II
trials(14,17) that 44 Gy may well be adequate. Treatment planning to achieve an elective
dose of 50 Gy is detailed next and will be referred to, where applicable, throughout the
chapter to avoid repetition.

18.2.5 Radiation dose
Intended dose prescription
◆ Macroscopic disease: 66–​70 Gy in 2-​Gy fractions treating five times per week.
◆ Microscopic disease: 50 Gy in 2-​Gy fractions treating five times per week.
416 Head and neck cancer

Treatment technique
A five to seven field IMRT technique or a volumetric modulated arc therapy (VMAT)/​
rapidarc technique may be used to obtain uniform coverage of the PTV and satisfy the
dose constraints to the OARs.
Planning aims should be prioritized in the following order:
◆ Meeting critical organ constraints (spinal cord and brainstem).
◆ Radical PTV coverage.
◆ Intermediate PTV coverage (if using an intermediate dose level).
◆ Elective PTV coverage.
◆ Non-​critical organ constraints (e.g. parotids).
◆ Other non-​specified normal tissue.

Altered fractionated schedules
CHART was evaluated in a UK-​wide study(15). This intensive regimen requires out-​
of-​hours treatment and failed to show an overall survival advantage in head and neck
cancer.
Large randomized controlled trials of altered fractionation in head and neck cancer
have indicated that maintaining a total dose of 70 Gy gives benefit over conventional
radiotherapy whether a hyperfractionated or concomitant boost technique is used(18,19).
There is also evidence from a meta-​analysis that a modest acceleration of radiotherapy
maintaining the dose at 66 Gy is advantageous(20,21).

18.2.6  Implementation of treatment

Modern thermoplastic shells provide little surface dose build-​up and so cutting out of
the shell (which was commonly performed with the old-​style Cabulite shells) is not
necessary. Indeed, if the tumour extends close to skin, such as a fungating tumour
or laryngeal carcinoma involving the anterior commissure, the application of bolus
may be required to ensure adequate dose to tumour tissue. During treatment, radio-
graphers should check for loosening of the shell, particularly if patients experience
significant weight loss.

18.2.7 Verification
Electronic portal imaging should be used to verify beam position and correct shielding
during the initial 3 days of treatment and weekly thereafter. Typically a tolerance of ±
3 mm is acceptable, although this may need to be more stringent if critical OARs are
very close to the target volume. Volumetric kV or MV imaging may provide more in-
formation about patient position during therapy.

18.3 Postoperative radiotherapy
18.3.1  Indications and treatment volume
Fifteen factors have been identified as important for prediction of recurrence. The first
two, either alone or in combination, are definitive indications for chemoradiotherapy.
 Postoperative radiotherapy 417

1. Positive resection margins.

2. Extracapsular lymph node spread.
Two of the remainder suggest postoperative radiotherapy should be recommended.
3. Close margins < 5 mm.
4. Invasion of soft tissues.
5. Two or more lymph nodes positive.
6. More than one positive nodal group.
7. Involved node > 3 cm in diameter.
8. Multicentre primary.
9. Perineural invasion.
10. Vascular invasion.
11. Poor differentiation.
12. Stage  T3/​4.
13. Oral cancer.
14. Carcinoma in situ dysplasia at edge of resection margin.
15. Uncertainties concerning surgical/​pathological findings.
16. HPV negativity.
High risk of recurrence is associated with either factor 1 or 2 alone or presence of two
or more of factors 3–​9.
Intermediate risk is associated with the presence of any one of the factors 3–​9.
Factors 10–​16 may be of some importance in predicting recurrence and should be
borne in mind.

18.3.2 Planning technique
Treatment volume should encompass the area of surgical resection and will include all
areas considered at risk of recurrence. A margin of at least 10 mm around the maximal
extent of surgery should be included. Where nodal involvement has been proven, the
field should be extended to include the lymphatic drainage down to the clavicle. Areas
of positive margin or other high-​risk features may receive with a boost to the area up
to radical dose.

18.3.3 Dose prescription

◆ 60 Gy in 30 fractions treating daily Monday to Friday; 2 Gy per fraction (boost high

risk areas to 65–​66 Gy e.g. positive margin).
◆ 50 Gy to uninvolved non-​operated sites (54 Gy in 30 fractions if using IMRT), e.g.
contralateral neck.

Intended dose prescription
◆ Definitive dose to postoperative tumour bed: 60–​65 Gy in 30 fractions.
◆ Microscopic non-​operated field dose: 50 Gy in 25 fractions or 54 Gy in 30 fractions.
418 Head and neck cancer

Treatment can be planned as 3-​D conformal fields or IMRT if treating midline and/​
or bilateral neck.

18.4 Palliative radiotherapy
Palliative radiotherapy is only given to a small proportion of patients with head and
neck cancer. This is due to the relatively high doses required to achieve symptom relief.
It is reserved for patients who have metastatic disease at presentation, or those with
performance status ≥ 2.
No standard dose schedule for palliative irradiation exists. Typically large doses per
fraction are used with careful consideration of tolerance doses to OARs such as spinal
cord, brain, and cranial nerves. Regimens that can be considered on an individual
basis include: 50 Gy in 20 daily fractions; 36 Gy in 12 fractions over 2.5 weeks; 20 Gy
in 5 fractions over 1 week.

18.5  Site-​specific treatment planning

IMRT is now the standard of care for locally advanced oropharyngeal and
laryngohypopharyngeal cancers. It is also the standard planning technique for nasopha-
ryngeal cancers. Conformal 3-​D planning is used for treatment of paranasal sinus, orbital
and nasal tumours but IMRT is being used increasingly in this setting. Lateralized tonsil
tumours and parotid tumours are treated with 3-​D conformal planning using a wedged
pair and matched neck technique. Conventional radiotherapy is still the standard of
care for treatment of early glottis cancers using lateral parallel-​opposed beams.

18.5.1 Larynx
The larynx is divided into three distinct anatomical regions: the supraglottis (laryngeal
epiglottis, false cords, ventricles, aryepiglottic folds, and arytenoids), the glottic larynx
(true vocal cords, anterior and posterior commissures), and subglottis (10 mm below
the free edge of the vocal cords to the inferior edge of the cricoid cartilage). Each has
its own natural history, patterns of spread, and treatment protocols.
Immobilization
For all larynx tumours the patient should be immobilized in the supine position with
the cervical spine straight. Patients treated with IMRT should be immobilized with the
neck extended.
Glottic tumours
Patients with carcinoma in situ (Tis) or dysplasia should be treated surgically by laser
excision or cord striping. Radiotherapy is best reserved for recurrent lesions.
Stage T1–​2, N0 tumours can be treated with radiotherapy or surgical excision
with laser cordectomy. There is some controversy as to which modality provides the
best voice quality. In the UK, radiotherapy remains the standard of care. Typically
a parallel-​opposed lateral beam arrangement is used with 5-​cm (T1) or 6-​cm (T2)
square fields centred on vocal cord (1 cm below thyroid promontory and anterior to
the lower border of the C5 vertebrae). The superior border should be at the lower
 Site-specific treatment planning 419

Hyoid bone

Cricoid cartilage
Fig. 18.5  Radiotherapy for
a T1/​2 N0 carcinoma of the
glottic larynx.

edge of the hyoid bone, and inferiorly the field should encompass the cricoid car-
tilage covering the width of the thyroid cartilage. Anteriorly the field border should
be in air at the field centre, and posteriorly should be through the anterior part of the
vertebral body (Fig. 18.5). Usually 10–​20° wedges are used as missing tissue compen-
sators. No prophylactic nodal radiotherapy is given, although the anterior part of the
mid-​cervical lymph nodes (level III) is within the irradiated volume if standard lateral
parallel-​opposed fields are used.
For T2 tumours the superior and/​or inferior borders are individually expanded
based on the supraglottic and/​or subglottic extension. The para-​oesophageal and
paratracheal lymph nodes are included for extensive subglottic extension.
For tumours involving the anterior commissure within a few millimetres of the skin
surface, the skin-​sparing effects of a megavoltage beam risk tumour under dosage, and in
these cases the anterior part of the shell should not be cut out. If the calculated dose to the
anterior commissure is still low, then this can be improved by reducing or removing the
wedge from each lateral field or adding bolus to increase the dose in the superficial tissues.
In patients with short necks, or high shoulder position then lateral fields to the larynx
may not be deliverable, and in this situation an anterior-​oblique wedged pair arrangement
is more appropriate (Fig. 18.6). This will require a PTV to be localized by CT planning.

PTV

Fig. 18.6  A field arrangement for patient

Spinal cord inappropriate for lateral field irradiation for
early larynx cancer due to high shoulder
position.
420 Head and neck cancer

Dose prescription
This is dependent on field size:
◆ Less than 42 cm2: 55 Gy in 20 fractions over 4 weeks(22).
◆ Larger field sizes: 65 Gy in 30 fractions over 6 weeks.

Supraglottic tumours
T1 and T2, N0 tumours are associated with a high incidence of occult positive nodes in
level II and III because of the dense lymphatic supply in this area. All patients therefore
require elective nodal irradiation of these levels. A two-​phase technique is used. Phase
1 should include the primary tumour, the whole larynx, pre-​epiglottic space, and the
cervical nodes levels Ib, II, and III bilaterally anterior to spinal cord. Phase 2 should
encompass the primary tumour only to a dose of 66–​70 Gy. Parallel-​opposed wedged
fields are used for both phases (Fig. 18.7).
Dose prescription
◆ Total dose: 66–​70 Gy in 33–​35 fractions to macroscopic disease; 50 Gy to micro-
scopic disease.
◆ Treatment technique as section 18.2.5.

Node-​positive supraglottic tumours
Supraglottic tumours with nodal involvement are considered for surgery and
postoperative radiotherapy. They may, however, be treated primarily by radiotherapy
or chemoradiation, reserving surgery for treatment failure.
These patients should be treated with IMRT. A CTV1 should be outlined from 1 cm
above the tip of the epiglottis (or a 2-​cm margin above or below the superior and or
inferior extent of tumour, whichever was larger) with a minimum margin of 1  cm
surrounding any involved nodes. This CTV will also include neighboring anatomical
structures at risk of microscopic spread (the margins will be reduced at the borders
of an uninvolved anatomic space or structure, i.e. bone). If the GTV extends into an-
other head and neck subsite (e.g. extension into oropharynx), then this entire subsite
volume is included in the CTV. Some oncologists include the at-​risk subsite (that is
not included in the GTV to CTV expansion) in a separate intermediate-​risk dose level

Phase 1
Superior: Mastoid
process
Inferior: Cricoid
cartilage
Posterior: Mid-
vertebral body
Anterior: Anterior
border of level lb

Phase 2
Primary tumour with
1–2 cm margin

Fig. 18.7  Radiotherapy technique for a node-​negative supraglottic carcinoma.

 Site-specific treatment planning 421

and will only treat the GTV plus a margin to the radical dose. CTV2 will include the
uninvolved nodal neck levels bilaterally (levels II-​IV on node negative side and II-​V on
node positive side; level Ib is included if level II nodes involved) (Fig. 18.8).
Dose prescription
◆ Total dose: 65–​70 Gy in 30–​35 fractions to macroscopic disease (65 Gy in 30 frac-
tions for IMRT); 50 Gy equivalent to microscopic disease (54 Gy in 30 fractions
for IMRT).
◆ Treatment technique as section 18.2.5.

Subglottic tumours
Tumours of the subglottis are rare, and most present with locally advanced disease. In
operable patients surgery with laryngectomy and postoperative radiotherapy may be
employed. For patients with early stage disease primary radiotherapy offers the chance
of larynx preservation. The rate of cervical node metastasis is rare, but involvement of
paratracheal nodes is estimated to be 50% mandating elective treatment. These patients
should be treated with IMRT using the same technique as node positive supraglottic
tumours in section 18.5.1 (similar to that shown in Fig. 18.8).
Dose prescription
◆ 65–​66 Gy in 30 fractions to PTV1 and 54 Gy in 30 fractions to PTV2 treating daily,
five fractions/​week.

Fig. 18.8  Node-​positive supraglottic carcinoma. CTV1 (radical CTV) in red and CTV2
(elective nodal CTV) in cyan for larynx and bilateral neck nodes to be treated with IMRT.
422 Head and neck cancer

Advanced larynx carcinoma
Advanced tumours of the larynx are treated in a similar way, and the exact site of
origin has less effect on treatment technique. Stage T3 represents a very inhomo-
geneous group, which can range from a small tumour with vocal cord fixation to
a large trans-​glottic tumour. Most patients with good performance status who are
medically fit for chemotherapy should receive chemoradiation therapy that maxi-
mizes the chance of larynx preservation. Neo-​adjuvant chemotherapy should also be
considered(23). If tumour invades the thyroid cartilage (T4), then radical surgery and
postoperative radiotherapy represents the treatment of choice. Synchronous chemo-
therapy and radiation offer an alternative in those patients who are medically unfit.
The treatment of advanced larynx cancers in complex and usually involves two to
three phases if conventional radiotherapy is used. IMRT is now used routinely in the
treatment of these patients and avoids the need for electrons and field–​field matching.
If IMRT is used, PTV1 is typically prescribed 65–​66 Gy in 30 fractions and PTV2 54
Gy in 30 fractions.
Radical radiotherapy technique
The target volume includes the larynx, and pre-​epiglottic space, and lymph node areas
at risk of harbouring metastatic disease. This should include levels Ib (if level II nodes
involved), II, III, and IV in all patients, and level V in node-​positive patients. If there
is a tracheostomy then this must be also included.
◆ Total dose: 65–​66 Gy in 30 fractions to macroscopic disease; 540 Gy in 30 fractions
to microscopic disease.
◆ Planning as detailed previously in section 18.2.5.

Postoperative radiotherapy
Post-​operative radiotherapy is delivered using IMRT to cover the tumour bed and
nodal levels (60 Gy in 30 fractions and 54 Gy in 30 fractions to unoperated neck). The
same planning technique in section 18.2.5 is used.
Dose prescription
◆ Macroscopic tumour: 65–​66 Gy in 30 fractions treating daily, five fractions/​week
(see section 18.2.5).
◆ Postoperative: 60 Gy in 30 fractions treating daily, five fractions/​week (see section
18.3.3).

18.5.2 Oropharyngeal tumours
The oropharynx is split into four main subsites: tongue base, tonsil, soft palate, and
pharyngeal wall. They all have a relatively high risk of nodal metastasis. Tumours
occurring in the midline (base of tongue, soft palate, and posterior pharyngeal wall)
can metastasize to either side of the neck and therefore require irradiation to the pri-
mary tumour site and the neck bilaterally. By contrast, lateralized tumours of the tonsil
or lateral pharyngeal wall metastasize unilaterally (24) and therefore can be treated with
less extensive fields allowing sparing of the contralateral structures, most importantly
the contralateral parotid gland.
 Site-specific treatment planning 423

Small lateralized tumours of tonsil (T1/​T2, node negative

or positive)
The following lateral radiation technique is only suitable for tumours confined to
the tonsillar fossa. Patients should be positioned with the cervical spine straight. For
node-​negative patients, the CTV includes the tonsillar fossa, and ipsilateral level Ib–​
IV nodes. The tonsil is irradiated using a wedged pair technique with anterior and
posterior oblique fields (Fig. 18.9). This should be matched at the level of the hyoid to
a CT-​planned ipsilateral anterior neck field treating levels III and IV. For node-​positive
patients the CTV also includes level V nodes, and involved nodes require treatment to
radical dose. IMRT can also be used for this indication. It produces a more homoge-
neous dose distribution within the PTV but at the expense of unecessaary irradiation
of the contralateral tissue of the head and neck. Postoperative irradiation to the tu-
mour bed and/​or neck is usually required after surgical excision.
Dose prescription
◆ Macroscopic disease: 65–​66 Gy in 30 fractions treating daily, five fractions/​week.
◆ Microscopic disease: 50 Gy in 25 fractions treating daily, five fractions/​week

Advanced oropharyngeal carcinoma (tonsil or base of tongue)

When a tonsil tumour approaches the midline at the soft palate, or involves the base of
tongue, then the lateralized radiation technique is not appropriate. Both sides of the neck
may harbour occult metastases and therefore the CTV includes the oropharynx, level VIIa
and level II–​V nodes bilaterally (include level Ib on node positive side). This technique is
the same for tonsil and base of tongue tumours. Patients should be treated with parotid-​
sparing IMRT to reduce the risk of long-​term xerostomia. (Fig. 18.10—​oropharynx IMRT).
Dose prescription
◆ Macroscopic disease: 65–​66 Gy in 30 fractions treating daily, five fractions/​week.
◆ Microscopic disease: 50 Gy in 25 fractions for matched neck field and 54 Gy in 30
fractions for IMRT treating daily, five fractions/​week.
◆ Treatment technique: see section 18.2.5.

Fig. 18.9 Radiotherapy
technique and dose
distribution for a T1/​2
tumour of the right
tonsillar fossa showing
the use of anterior
and posterior oblique
beams with sparing
of the contralateral
parotid gland.
424 Head and neck cancer

Soft palate
The patient should be immobilized with the neck extended. For early stage (T1 or T2)
node-​negative disease, elective nodal irradiation is not necessary. The target volume is
therefore the GTV with a 2-​cm margin only, and can be irradiated with small lateral
opposed fields to radical or postoperative dose.
◆ 65–​70 Gy in 30–​35 fractions treating daily, five fractions/​week.
In patients with advanced T stage (T3 or T4) or node-​positive disease, then bilateral
cervical node irradiation is required using IMRT using the SIB technique (see section
18.2.3), and the technique used is the same as that described in section 18.2.5.
◆ Macroscopic tumour dose:  65–​66 Gy in 30 fractions treating daily, five times
per week.
◆ Microscopic tumour dose: 54 Gy in 30 fractions treating five times per week.
◆ See section 18.2.5.

18.5.3 Hypopharynx
Tumours of the hypopharynx are characterized by a high risk of lymph node me-
tastases. Therefore for any stage tumour elective irradiation of locoregional lymph
nodes is required. The hypopharynx has three recognized subsites:  pyriform fossa,

(a) (b)

(c) (d)

Fig. 18.10  (a + b) Technique for irradiation of lateralized oropharyngeal tumours

using a wedged pair and matched neck technique (c) IMRT plan for non-​lateralised
oropharyngeal tumour and bilateral neck. (d) A dose–​volume histogram (DVH) for the
dose distribution given. Showing spinal cord maximum dose 42 Gy, and prescription
dose to the PTV of 65 Gy.
 Site-specific treatment planning 425

postcricoid, and posterior pharyngeal wall. Most tumours of the hypopharynx are
suitable for organ-​preserving schedules with radiation or chemoradiation(23).

Pyriform fossa
Patients should be positioned with the neck extended. Target volume includes the pri-
mary tumour site and level I–​V lymph nodes bilaterally, and typically treated with
IMRT (Fig 18.11).
Dose prescription
◆ Macroscopic tumour dose: 65–​66 Gy in 30 fractions treating daily, five times per week.
◆ Microscopic tumour dose: 54 Gy in 30 fractions treating five times per week.
◆ See section 18.2.5.

Posterior pharyngeal wall
As with tumours of the pyriform fossa, the target volume includes the primary tu-
mour site and level I–​V lymph nodes bilaterally, and is typically treated with IMRT.
Postcricoid tumours are treated in the same way and can generally achieve the plan-
ning goals while keeping spinal cord dose within tolerance.
Dose prescription
◆ Macroscopic tumour dose: 65–​66 Gy in 30 fractions treating daily, five times per week.
◆ Microscopic tumour dose: 54 Gy in 30 fractions treating five times per week.
◆ See section 18.2.5.

(a) (b)

Fig. 18.11  Technique for irradiation of hypopharyngeal carcinoma a + b) Primary and

nodal GTV in cyan, + 1 cm margin to both represented in red, CTV1 in pink
426 Head and neck cancer

18.5.4 Nasopharynx
Nasopharyngeal carcinoma has a high risk of lymph node metastases initially to
the retropharyngeal and parapharyngeal lymph nodes, and also to the deep cervical
nodes bilaterally. The majority of patients present with locally advanced disease and
are treated with chemoradiation(25). IMRT is the standard technique for patients with
nasopharyngeal carcinoma because of the very high risk of long-​term xerostomia with
conventional radiotherapy.
Patients are positioned with the neck extended with the chin up as far as possible.
This optimizes shielding of the orbit and oral cavity in early phases of treatment. Data
from MRI and CT are useful to accurately delineate the disease volume. The CTV in-
cludes base of skull (middle temporal fossa and cavernous sinus), posterior half of the
orbit, posterior half of the nasal cavity, parapharyngeal space, lateral pharyngeal, and
posterior and upper deep cervical nodes (Fig 18.12).

Intensity-​modulated radiotherapy
PTV1 includes the primary tumour, retropharyngeal nodes, bilateral parapharyngeal
spaces, and any lymph node groups harbouring metastases. PTV1 is treated with 65
Gy in 30 fractions. PTV2 includes all elective nodal groups Ib–​V and is treated to 54
Gy in 30 fractions using SIB technique.

(a) (c)

(b) (d)

Figure 18.12  IMRT for nasopharyngeal tumours a + b) CTV1 includes nasopharynx, posterior

part of nasal cavity, and inferior half of sphenoid sinus (if sphenoid not involved; c) sagittal
view of CTV1 (note inferior half of sphenoid sinus is in CTV1 (red) and superior half in CTV2
(cyan)); d) composite CTVs—​CTV1 in red and CTV2 (uninvolved nodal levels) in cyan.
 Site-specific treatment planning 427

Lymph node-​negative, early stage, well-​differentiated nasopharyngeal carcinoma

may be treated with localized radiotherapy to the nasopharynx, without full elective
nodal irradiation.

18.5.5 Oral cavity
Early oral cancer including superficial (< 5 mm thickness), T1 and T2 lesions should
be considered for brachytherapy. External beam radiotherapy is usually given
postoperatively in patients with high-​risk features or those unsuitable for radical surgery.
The treatment position is with the cervical spine straight. A  mouth bite may be
used to position the tongue. The oral cavity contains a number of individual subsites
including oral tongue, floor of mouth, buccal mucosa, alveolus, and hard palate.

Tongue
CTV should be the tumour bed with a 2-​cm margin. For tumours on the lateral tongue
border this typically constitutes a hemioral cavity irradiation using anterior and pos-
terior oblique fields wedged to produce a homogeneous dose distribution. For deeply
infiltrative tumours approaching or invading the midline then IMRT may be required
to treat the CTV. Irradiation of the neck is indicated electively for infiltrative tumours
and may be unilateral or bilateral depending on the relationship of the tumour to the
midline or postoperatively for patients with high-​risk features.

Buccal mucosa and alveolus

Usually a lateralized CTV is treated in the postoperative setting (Fig. 18.13).

Floor of mouth
Floor-​of-​mouth tumours commonly occur in the midline and therefore irradiation
requires IMRT to cover the target volume that includes the primary tumour site and
locoregional lymph nodes. It is important for the mouth to be stented open using a
mouth bite that reduces irradiation of the hard palate mucosa.

Fig. 18.13 A dose
distribution for
irradiation of a buccal
mucosa tumour.
428 Head and neck cancer

Dose prescription
◆ Postoperative dose: 60 Gy in 30 fractions treating daily, five times per week.

18.5.6 Parotid gland
Tumours of the parotid are treated with surgery and postoperative radiotherapy.
Radiotherapy is indicated in tumours of high grade and those low-​grade tumours that
are recurrent or at very high risk of recurrence (i.e. macroscopic residual disease). For
high-​grade tumours (squamous, adenocarcinoma, and high grade mucoepidermoid),
the target volume should include the parotid bed and the ipsilateral level II nodes. For
low-​grade tumours the risk of lymph node metastases is so low that low neck irradi-
ation is not required.
Patients should be immobilized with the neck extended and mandible perpen-
dicular to the couch top such that the orbit is above the superior border of the radio-
therapy fields. The entire parotid gland should be included in the CTV as these
tumours may spread through the gland along salivary ducts. Anteriorly the CTV
should extend to the anterior border of the masseter to include the parotid duct, medi-
ally the parapharyngeal space and laterally the scar. Adenoid cystic carcinomas spread
along nerves and therefore this should be considered in planning. The facial nerve
and the parasympathetics should be included back to their exit from the skull base by
extending the posterior field border. This CTV is irradiated using anterior and pos-
terior oblique beams in order to avoid irradiating the contralateral parotid gland, and
reducing the risk of xerostomia. IMRT can also be used for this indication. It pro-
duces a more homogeneous dose distribution within the PTV but at the expense of
unecessaary irradiation of the contralateral tissue of the head and neck.
Dose prescription
◆ Postoperatively: 60 Gy in 30 fractions treating daily, five times per week.
◆ If positive resection margins 65–​66 Gy in 30 fractions treating daily, five times per week.
Pleomorphic adenoma
This benign condition should be considered separately from other parotid tumours.
They are usually encapsulated and radiotherapy should only be considered if tumours
are recurrent or incompletely resected.
Typically they occur in the superficial lobe and can be treated with a direct electron
field covering the postoperative tumour bed. Occasionally they occur in the deep lobe
in which case they can be treated with the photon technique described previously.
Dose prescription
◆ 50 Gy in 25 fractions treating daily, five times per week.
◆ Or 45 Gy in 15 fractions treating daily, five times per week.

18.5.7  Irradiation of the neck only and unknown primary

The neck is irradiated alone in the postoperative setting (when primary tumour has
been adequately resected, but with high risk of neck recurrence), and in patients with
unknown primary tumour.
 Site-specific treatment planning 429

The patient should be immobilized supine, with the cervical spine straight and chin
up as high as possible. The upper border is though the mastoid process and the lower
border is below the lower border of the clavicle. The lateral border is the outer two-​
thirds of the clavicle and medial border is lateral to the spinal cord.
If the target volume includes level V then in order to adequately irradiate these
nodes parallel-​opposed anterior and posterior fields are required. Usually the neck
is irradiated postoperatively and for this indication the dose is 60 Gy in 30 daily frac-
tions. The prescription point for neck irradiation is not defined and can be applied, at
dmax or at a specific depth.

Irradiation of the mucosal surface of the head and neck in the

treatment of an unknown primary presenting with nodal disease
Investigation of a patient presenting with nodal metastases includes fine needle as-
piration of the node, EUA, and biopsy of any suspicious lesions. Ipsilateral tonsillec-
tomy should be performed in the absence of an identifiable lesion, and biopsies of
the postnasal space and tongue base may be undertaken. CT or MRI may identify an
occult primary site, but 18-​FDG-​PET scanning probably has a higher sensitivity than
either of these investigations. PET remains positive up to 6 weeks after biopsy and so
needs to be timed and interpreted carefully.
For squamous cell carcinoma metastatic to the deep cervical nodes, the most
common primary tumour sites are the hypopharynx, oropharynx, or nasopharynx.
The treatment of these patients is highly controversial and is the subject of ongoing
research.
Total mucosal irradiation of all the possible primary tumour sites and bilateral neck
nodal levels Ib–​V is possible with IMRT and allows parotid gland sparing. An alter-
native approach is irradiation of the involved hemineck, with an observation policy
for the mucosal surfaces. In approximately 30% of cases, no primary tumour appears
during follow-​up. If a primary tumour does become apparent subsequently, then fur-
ther treatment with surgery and radiotherapy is possible at that time. This approach is
associated with less morbidity, but higher rates of loco-​regional recurrence that can be
salvaged by further therapy.
If the histology of the node is undifferentiated carcinoma of nasopharyngeal type
(UCNT), especially in a posterior triangle node, then there is a high chance that the
tumour arose in the postnasal space, and radiotherapy should given as for nasopha-
ryngeal carcinoma. If it seems highly likely that the primary tumour lies in oropharynx
or hypopharynx, then radiotherapy is given as for these sites.

18.5.8 Orbit
Tumours of the orbit are rare. Most commonly metastases from distant sites are seen
in the context of widely disseminated malignant disease. Palliative radiotherapy using
a single lateral photon field is appropriate for most patients. For bilateral deposits, op-
posed lateral fields may be used. Most metastases are seen in the retina or posterior
orbit, and the use of a non-​divergent anterior field border can avoid irradiation of the
lenses of both eyes (Fig. 18.14). Half beam blocking, or a 5–​10% gantry rotation from
the direct lateral position, should be used to produce the non-​divergent anterior border.
430 Head and neck cancer

Superior border at superior

orbital rim
Inferior border at orbital floor
Anterior border behind lens
Posterior border at orbital apex

Fig. 18.14  Use of a direct

lateral field for orbit
irradiation.

Dose prescription
◆ 20 Gy in five fractions treating daily, five times per week.
◆ Or 30 Gy in 10 fractions treating daily, five times per week.
Lymphoma, rhabdomyosarcoma, and lacrimal gland tumours usually occur outside
the muscle cone posterior to the globe. These extraconal tumours may be treated with
radical radiotherapy leaving the eye intact. The use of CT planning allows accurate
localization of the target volume, and also the critical OARs: the lens, lacrimal gland,
optic nerve, brainstem, and brain.
The technique employs a direct anterior and anterior oblique wedged fields to cover
the PTV (Fig. 18.15).
The fields should be weighted anteriorly, and the anterior oblique field should
come in behind the lenses if possible. Corneal and lens doses can be minimized by
instructing the patient to stare directly into the beam with the eye open. The cornea
and the anterior part of the lens will lie within the build-​up region of the megavoltage
beam. If there is no tumour within the muscle cone, then a pencil lead shield can be
used to further reduce lens dose. One must be certain that no tumour lies within the
shielded tissue. If the tumour extends superiorly or inferiorly, the anterior field can be
angled superiorly or inferiorly to ensure that the corneal shadow falls outside the PTV.

Lenses

PTV: whole orbit

including globe,
rectus muscles, and
other intra-occular
tissues
Fig. 18.15  Irradiation of an
intact eye using anterior and
lateral fields.
 Site-specific treatment planning 431

PTV

Fig. 18.16  Non-​coplanar technique

for irradiation of a proptosed eye.

The patient must be clearly instructed to stare directly into the beam to immobilize
the eye during treatment. The use of lens shielding is unsuitable if there is intraconal
disease.
For patients with proptosis, the use of an anterior oblique beam or lateral field is
precluded because of the risk of irradiation of the contralateral eye. In these patients
a technique using superior and inferior non-​coplanar anterior fields should be used
(Fig. 18.16). The eye should be kept open during treatment and the patient should
stare into the beam from each gantry angle.

Thyroid eye disease
Lateral opposed fields should be used to irradiate the posterior orbit to a dose of 20
Gy in 10 fractions over 2 weeks. CT planning is recommended to assess lens dose that
should not exceed 6–​10 Gy.
If higher doses are required to the orbit, for example, to deliver 68–​70 Gy for a sar-
coma, a wedged pair arrangement in the coronal plane may be used with lateral su-
perior and inferior oblique fields.

Post-​operative radiotherapy
Frequently locally advanced orbital tumours require exenteration, and postoperative
radiotherapy is given. In this circumstance there are no intraorbital OARs, and so a
technique using anterior and anterior oblique fields can be used, similar to Fig. 18.15,
without the need to consider shielding structures within the eye. For carcinoma, doses
of 60–​64 Gy in 30–​32 daily fractions can be used.

18.5.9  Ear and temporal lobe

Tumours described in this section include tumours of the pinna, external auditory
canal, middle ear, and temporal bone.

Pinna
Tumours of the pinna should be considered as cutaneous malignancies. Primary
surgery or radiotherapy is the treatment of choice. Radiotherapy generally gives the
better cosmetic and functional result than pinnectomy. Radiotherapy with kilovoltage
photons is contraindicated by cartilage invasion, (fixation, pain, or infection) be-
cause of a high risk of necrosis secondary to increase in absorbed radiation dose
432 Head and neck cancer

(predominant photoelectric effect). Absorbed dose with electrons is less dependent

on atomic weight and is the modality of choice where possible. Radiation technique
should include definition of GTV, and addition of a margin of 1.5 cm to account for
microscopic spread (5 mm) as well as the penumbra of the electron beam (10 mm).
A wax-​backed lead shield is placed behind the pinna to shield the skin from exit dose
and bolus in the form of wax or wet gauze is used to ensure that the surface dose is
90–​100%; a lead cut-​out may be required.

Dose prescription
◆ 55 Gy in 20 fractions treating daily, five times per week.
◆ Or 45 Gy in 10 fractions treating daily, five times per week.

External auditory canal
Radiotherapy is indicated in early tumours of the external auditory meatus. In more
advanced disease, surgical resection followed by radiotherapy is the treatment of
choice. Patient position is with the neck extended so that exit of the posterior oblique
beam avoids eye and lens. CT planning should be used if possible. The technique uses
anterior and posterior oblique wedged beams (Fig. 18.17).

Dose prescription
◆ 66–​70 Gy in 33–​35 fractions treating daily, five times per week.

Middle ear and temporal bone

Standard treatment is extended total petrosectomy with postoperative radiotherapy.
Radiotherapy technique uses an extended neck position, and CT planning is recom-
mended. The CTV includes the GTV plus the pre and post-​auricular lymph nodes.
The brainstem and orbital contents should be localized as OARs. Anterior and pos-
terior wedged fields are used to irradiate the PTV in an arrangement similar to that
shown in Fig. 18.17.
Dose prescription
◆ 66–​70 Gy in 33–​35 fractions treating daily, five times per week.

Contralateral PTV
parotid gland

Spinal cord Fig. 18.17 Treatment

technique for a carcinoma
of the middle ear.
 Site-specific treatment planning 433

18.5.10  Nose and paranasal sinuses

Surgery alone or in combination with radiotherapy and chemotherapy is required in
the majority of patients with squamous cell carcinoma of the nasal cavity or paranasal
sinuses. Radiotherapy may be given before or after surgery. Chemotherapy and radio-
therapy are also indicated for lymphoma or embryonal rhabdomyosarcoma.

Maxillary antrum radiotherapy technique

The position is supine with the cervical spine straight and a mouth bite to keep the
lower oral cavity and tongue out of the fields. The PTV is often in close proximity to
the optic nerves, chiasm, orbit, temporal lobes, and brainstem. For this reason CT
planning is recommended, as it allows accurate target volume definition, and im-
proved dosimetry around bone and air cavities. The improvements in the dose calcu-
lation provide better estimates of doses to OARs, especially for tumours close to the
optic apparatus. This helps to inform clinical decision-​making and individualization
of radiotherapy dose to keep within optic nerve and chiasm tolerances.
Paranasal sinus tumours can spread mucosally into other adjacent sinuses, but
lymph node metastases to the neck are rare. The CTV for a maxillary antrum tumour
therefore includes the maxillary and ethmoid sinuses, nasal cavity, pterygoid fossa, and
lateral pharyngeal node. The target volume is determined on the basis of CT and MR
images and clinical and surgical assessment. Care should be taken to shield the brain-
stem, optic pathways, eyeball, lacrimal gland, and orbit wherever possible. Typically
a field arrangement using a heavily weighted anterior, and one or two lateral fields is
used. The anterior border of the lateral fields should use a non-​divergent field border
to avoid exit through contralateral lens and is achieved by angling the gantry by 5–​10°
behind the eyes, or using half beam blocking. The lacrimal gland, orbit, anterior part of
the optic nerve, and lens are shielded from the anterior field (Fig. 18.18); the brainstem
and optic chiasm from the lateral field (Fig. 18.19). IMRT may be required to achieve
target volume coverage of the most posterior and medial part of the PTV (Fig 18.20).

Fig. 18.18  Anterior field for a left maxillary

antrum.
434 Head and neck cancer

Fig. 18.19  Lateral field for a

carcinoma of the maxillary antrum.

A dose of 65–​70 Gy in 30–​35 daily fractions is appropriate for squamous carcinoma

or adenocarcinoma but careful attention to optic nerve and chiasm doses is required,
and the superior border of the lateral field(s) may need to be reduced at about 50 Gy
to avoid exceeding optic nerve tolerance (doses > 60 Gy carry a 20% risk of optic neur-
opathy). For lymphoma a dose of 35–​40 Gy is adequate, well below the tolerance of
the optic apparatus.

(a) (b)

Fig. 18.20  IMRT for irradiation of a maxillary antrum tumour.

 Site-specific treatment planning 435

Dose prescription
◆ 65–​70 Gy in 30–​35 fractions treating daily, five times per week.

Ethmoid sinus
The ethmoid sinus is very difficult to irradiate to high dose as it lies between the
optic nerves. A combination of surgery and postoperative radiotherapy gives the best
chances of local control. CT planning is recommended. The CTV includes the medial
half of the maxilla on the involved side, pterygoid fossa, both ethmoid sinuses, and
nasal fossa. As with other paranasal sinus tumours, care should be taken to shield
OARs wherever possible. Typically a three-​field plan is used with an anterior and two
lateral fields, although for the occasional T1 tumours the field can be restricted to
the ethmoid sinus and nasal cavity using superior and inferior non-​coplanar anterior
fields coming between the eyes. IMRT offers the best chance of achieving dose con-
straints to OARs and adequately cover the PTV.

Dose prescription
◆ 66–​70 Gy in 33–​35 fractions treating daily, five times per week.

Nasal cavity
Patient is positioned with the cervical spine straight. Clinical target volume includes
the lesion and 1-​cm margin. Field arrangement is an anterior wedged pair of photon
fields or in cases where the whole of the nasal cavity has to be included, an anterior
and lateral field are then employed to achieve coverage of the target volume at depth.

Columella
Careful assessment is needed, as the deep margins of columella lesions may be diffi-
cult to assess. For extensive lesions, a two-​or three-​field photon technique is used as
for nasal cavity tumours. For superficial lesions, a direct anterior electron field can be
used with a wax block and wax nostril plugs to produce a homogeneous tissue density
for dose deposition.
Dose prescription
◆ For lesions confined to the columella: 55 Gy in 20 fractions treating daily, five times
per week.
◆ For lesions extending up the nasal cavity:  65–​70 Gy in 30–​35 fractions treating
daily, five times per week.

Olfactory neuroblastoma (or esthesioneuroblastoma)

This is a rare tumour arising from the olfactory receptors in the cribriform plate. It
is best treated with surgery and adjuvant radiotherapy. Chemoradiation is often em-
ployed due to the locally advanced nature of the disease. CTV is the preoperative GTV
with a margin—​usually of 1 cm though this may need to be individualized on the basis
of critical structure tolerance. Include the bilateral nasal cavities, ethmoid sinuses,
cribriform plate, and olfactory bulb (so top of CTV is at least 1 cm superior to cribri-
form plate). A three-​field beam arrangement or IMRT is used similar to an ethmoid
436 Head and neck cancer

sinus lesion. The close proximity of the optic nerves and optic chiasm to the tumour
volume is a severe limitation to the dose that can be prescribed.
Dose prescription
◆ 60 Gy in 30 fractions treating daily, five times per week.

18.5.11  Recurrent disease and palliation

Some patients present with very advanced local or metastatic disease such that radical
treatment is not appropriate. In other patients, concomitant medical conditions may
preclude radical treatment. Surgery, radiotherapy, and chemotherapy all have a potential
role in palliation and this is best discussed in the context of the multidisciplinary team.
Palliative radiotherapy requires high doses, and short fractionation regimens are
associated with marked acute toxicity. Elective nodal irradiation is inappropriate in
patients being treated with palliative intent, and therefore the PTV should include
macroscopic disease only.

References
1. Toms JR (ed). Cancer Stats Monograph 2004. London: Cancer Research UK, 2004.
2. Sobin LH, Gospodarowicz MK, Wittekind C (eds). TNM Classification of Malignant
Tumours (7th edn.). New York: John Wiley and Sons, 2009.
3. Bernier J, Domenge C, Ozsahin M, et al. Postoperative irradiation with or without
concomitant chemotherapy for locally advanced head and neck cancer. New England
Journal of Medicine 2004; 350: 945–​52.
4. Cooper JS, Pajak TF, Fporastiere AA, et al. Post-​operative concurrent radiotherapy and
chemotherapy in high-​risk squamous-​cell carcinoma of the head and neck. New England
Journal of Medicine 2004; 350:1937–​44.
5. Pignon JP, Bourhis F, Domenge C, et al. on behalf of the MACH-​NC Collaborative
Group. Chemotherapy added to locoregional treatment for head and neck squamous-​cell
carcinoma: three meta-​analyses of updated individual data. Lancet 2000; 355:949–​55.
6. Candela FC, Kothari K, Shar JP. Patterns of cervical node metastases from squamous cell
carcinoma of the oropharynx and hypopharynx. Head & Neck 1990; 12:197–​203.
7. Shah JP, Candela FC, Poddar AK. Patterns of cervical lymph node metastases from
squamous cell carcinoma of the oral cavity. Cancer 1990; 66:109–​13.
8. Gregoire V, Ang K, Budach W, et al. Delineation of the neck node levels for head and neck
tumors: A 2013 update. DAHANCA, EORTC, HKNPCSG, NCIC CTG, NCRI, RTOG,
TROG consensus guidelines. Radiotherapy and Oncology 2014; 110(1):172–​81.
9. Nutting CM, Morden JP, Harrington JK, et al. Parotid-​sparing intensity modulated versus
conventional radiotherapy in head and neck cancer (PARSPORT): a phase 3 multicenter
randomised controlled trial. Lancet Oncology 2011; 12(2):127–​36.
10. International Commission on Radiation Units and Measurement. Prescribing, recording
and reporting photon beam therapy. ICRU Report 50. Bethesda, MD: ICRU, 1993.
11. International Commission on Radiation Units and Measurement. Prescribing, recording
and reporting photon beam therapy (supplement to ICRU report 50). ICRU Report 62.
Bethesda, MD: ICRU, 1999.
12. Nutting C, Bidmead M, Harrington KJ, Henk JM. BIR Geometric uncertainties in
radiotherapy: Head and neck cancer. In McKenzie A, Bidmead M (eds) Geometric
 References 437

uncertainties in radiotherapy—​defining the planning target volume. London: British Institute

of Radiology, 2003.
13. Gregoire V, Ang K, Budach W, et al. Delineation of the neck node levels or head and neck
tumors: a 2013 update. DAHANCA, EORTC, HKNPCSG, NCIC CTG, NCRI, RTOG,
TROG consensus guidelines. Radiotherapy and Oncology 2014; 110:172–​81.
14. Levendag P, Braaksma M, Coche E, et al. Rotterdam and Brussels CT-​based neck nodal
delineation compared with the surgical levels as defined by the American Academy of
Otolaryngology-​Head and Neck Surgery. International Journal of Radiation Oncology,
Biology, Physics 2004; 58:113–​23.
15. Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to therapeutic irradiation.
International Journal of Radiation Oncology, Biology, Physics 1991; 21:109–​22.
16. Dische S, Saunders M, Barrett A, et al. Randomised multicentre trial of CHART versus
conventional radiotherapy in head and neck cancer. Radiotherapy and Oncology 1997;
44:123–​36.
17. 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. International Journal of Radiation Oncology,
Biology, Physics 2000; 48:7–​16.
18. Horiot JC, Bontemps P, van den Bogaert W, et al. Accelerated fractionation compared to
conventional fractionation improves locoregional control in the radiotherapy of advanced
head and neck cancers: results of the EORTC 22851 randomised trial. Radiotherapy and
Oncology 1997; 44:123–​37.
19. Bourhis J, Etessami A, Wilbault P, et al. Altered fractionated radiotherapy in the
management of head and neck carcinomas: advantages and limitations. Current Opinion in
Oncology 2004; 16:215–​19.
20. Overgaard J, Hansen HS, Specht L, et al. Five compared with six fractions per week of
conventional radiotherapy of squamous-​cell carcinoma of head and neck: DAHANCA 6
and 7 randomised controlled trial. Lancet 2003; 362:933–​40.
21. Wiernik G, Alcock CJ, Bates TD, et al. Final report on the second British institute of
radiology fractionation study: short versus long overall treatment times for radiotherapy of
the laryngo-​pharynx. British Journal of Radiology 1991; 64:232–​41.
22. Spaulding MB, Fischer SG, Wolf GT. Tumor response, toxicity, and survival after
neoadjuvant organ-​preserving chemotherapy for advanced laryngeal carcinoma. The
Department of Veterans Affairs Cooperative Laryngeal Cancer Study Group. Journal of
Clinical Oncology 1994; 8:1592–​9.
23. Lefebvre JL, Chevalier D, Luboinski B, Kirkpatrick A, Collette L, Sahmoud T. Larynx
preservation in pyriform sinus cancer: preliminary results of a European Organization
for Research and Treatment of Cancer phase III trial. EORTC Head and Neck Cancer
Cooperative Group. Journal of the National Cancer Institute 1996; 88(13):890–​9.
24. O’Sullivan B, Warde P, Grice B, et al. The benefits and pitfalls of ipsilateral radiotherapy
in carcinoma of the tonsillar region. International Journal of Radiation Oncology, Biology,
Physics 2001; 51(2):332–​43.
25. Al-​Sarraf M, LeBlanc M, Giri PG, et al. Chemoradiotherapy versus radiotherapy in
patients with advanced nasopharyngeal cancer: phase III randomised Intergroup study
0099. Journal of Clinical Oncology 1998; 16:1310–​17.
 Chapter 19

Skin cancer
Carie Corner, Hannah Tharmalingam,
and Peter Hoskin

19.1  Squamous cell carcinoma and basal

cell carcinoma
Most squamous cell carcinomas (SCCs) can be equally well treated by surgical resec-
tion or local radiotherapy. Local control rates after radiotherapy for T1 and T2 lesions
(up to 5 cm diameter) range from 85 to 95%; maximum tumour diameter; tumour
pathology and total dose equivalent to at least 60 Gy are independent predictors of
local relapse in non melanoma skin cancer(1, 2).
Control rates for basal cell carcinomas (BCCs) tend to be even higher than squa-
mous carcinomas with overall rates in excess of 90%(1,2,4). The superficial and sclerosing
(morpheaform) variants have higher relapse rates than the nodular type(5). Relapse in
BCCs may occur later than in squamous carcinoma, median time to recurrence in
being 20 to 40 months compared to 5 months in squamous carcinoma(2,5).

19.1.1  Indications for radiotherapy

◆ Lesions where radiotherapy would produce better cosmesis and functional outcome
than surgery, e.g. nose, lower eyelid, ear, lower lip.
◆ Lesions with potential for deep tumour infiltration, e.g. inner canthus, nasolabial
fold, ala nasi, tragus, post auricular area.
◆ Large superficial lesions.
◆ Elderly/​frail patients.
◆ Patients who refuse or are unfit for surgery, e.g. due to anaesthetic risk, etc.
◆ Adjuvant treatment in cases at high risk of recurrence.

19.1.2  Indications for surgery

◆ Patients age < 50 years (as radiation scars tend to deteriorate with time).
◆ Sites of previous burns.
◆ Sites of previous radiotherapy.
◆ Areas of vascular insufficiency, e.g. shin, dorsum of hand (where radiotherapy may
cause problems with healing or function).
◆ Lesions overlying the lacrimal gland (upper outer eyelid).
 Squamous cell carcinoma and basal cell carcinoma 439

◆ Conditions with an inherent defect in the DNA repair mechanism, e.g. ataxia tel-
angiectasia, xeroderma pigmentosa resulting in a predisposition to extreme radi-
ation reactions.
Most patients who are treated with radiotherapy will be treated using electrons or
superficial (kV) treatment. A combination of surgical resection and adjuvant radio-
therapy is used for larger tumours; those with deep invasion of bone or cartilage;
extensive skeletal muscle involvement or macroscopic perineural invasion. These
squamous cell T4 tumours may have first station lymph nodes treated electively.

19.1.3  Post operative adjuvant treatment

Superficial SCC of the skin treated surgically will usually be completely excised with
an adequate surgical margin of 4 to 6 mm, or with Mohs micrographic surgery. Closer
histological margins may be dealt with by re-​excision or post-​operative radiotherapy
to the site of excision
Adjuvant radiotherapy may also be indicated for completely excised (> 1 mm) le-
sions with high risk features:  aggressive tumour subtypes; poorly differentiated; tu-
mours > 2 cm in diameter and > 4 mm in depth, and those with perineural invasion or
lymphovascular invasion detected pathologically(6,7).

19.1.4 Radiotherapy planning
At the time of radiotherapy planning the clinic notes, histology, clinical photographs,
and computed tomography (CT)/​magnetic resonance imaging (MRI) images if appro-
priate should be reviewed.

Immobilization and position
For treatments around the head and neck then plastic immobilization shells should be
used, with the shell cut-​out around the treatment area.
If a lead (Pb) mask is used this is placed over the plastic shell. It is important that
the position is stable with appropriate support to the body using a suitable headrest,
pillow, or sandbags. The patients should be as comfortable as possible to aid reproduci-
bility. Access using superficial X-​ray or electron applicators will have to be considered
in defining the optimal patient position.

Target definition
This should be defined on clinical examination using a bright light and magnifying
glass. A  fine indelible marker is used to define the tumour extent with appropriate
margins. For tumours which are fixed or those in the inner canthi, nasal vestibule,
or posterior auricular area, CT or MR scanning is valuable to identify the depth of
invasion.

Treatment volume
The clinician will delineate a gross tumour volume (GTV) and a field size. A margin
of 3 mm is used for the clinical target volume (CTV); a further 2 mm is added for
440 Skin cancer

the planning target volume (PTV). The resultant expansion of GTV to PTV is 0.5 cm
circumferential margins, and 0.5 cm for the deep margin. For morphoeic BCCs and
SCCs the circumferential margins are extended to 1 cm and the deep margin 0.5–​1cm.
An additional margin for field size is required due to the following characteristics of
the electron beam:
• The edge of the electron applicator represents the 50% isodose; the 90% isodose
is typically 3 to 5mm inside this depending on field size and therefore a larger
applicator than the defined PTV will be required.
• As electron energy increases, there is a bowing inwards of the isodoses close to
the surface where the tumour will be as shown in Fig. 19.1. An additional al-
lowance for this should be made so that overall an applicator diameter 10 mm
larger than the defined PTV should be chosen.

(a)

(b)

Fig. 19.1  Electron isodoses demonstrating bowing in effect of higher energy 15-​MeV

beams (a) compared to lower energy 6-​MeV beams (b).
 Squamous cell carcinoma and basal cell carcinoma 441

GTV = 2cm

+3 mm

CTV = 2.6cm

+2 mm

PTV = 3cm

+10mm

Field size = 5cm

Fig. 19.2  Diagrammatic illustration of the margins required for the treatment of a basal
cell carcinoma using electrons.

Examples of margins required for a BCC are shown in Fig. 19.2. A 2 cm diameter tu-
mour (GTV) gives a 3 cm diameter PTV. This has adequate 90% coverage if the field
size diameter is 5cm.

Dose distribution
Treatment will typically require the 90% isodose to cover a depth of 20 to 25 mm. This
may be achieved by a superficial X-​ray beam or an electron beam of 6 to 10 MeV en-
ergy. Early reports suggested that electron treatment was less effective than superficial
X-​rays but subsequent analyses have shown that provided adequate margins are used
taking into account the characteristics of the electron beam dose distribution then
equivalent results are obtained(1)[3).

Planning aims/​prescription
The prescription should be to the ICRU reference point, which is the 100% (i.e. dmax) of
the percentage depth dose. The ICRU reference point (i.e. dmax) should always be at the
centre (or in the central part of the PTV). One should aim to cover the PTV with the
90% isodose. Organs at risk should be defined and doses kept to defined constraints.

Beam energy determination
Energy should be chosen to give the best conformation to the defined PTV.
Determination depends on:
◆ Surface dose required (minimum 90%).
◆ Depth to be treated, i.e. covered with the 90%. This depth should be equal to the
depth of measurable tumour plus 5–​10 mm (GTV to PTV).
◆ Dose to surrounding critical structures.
442 Skin cancer

Important considerations when choosing an electron beam are:

Surface dose
As electron beam energy increases the surface dose also increases (Fig.19.1).
Build-​up material (bolus) may be required depending upon the characteristics of the
beam.

Bolus
Bolus is tissue equivalent material which can be placed over the treatment field to
increase the surface dose. It also acts to spare dose to normal tissues by reducing the
depth in the patient of the high dose volume. Bolus can be used to fill air cavities; e.g.
nostrils and external auditory canal and can correct for inhomogeneities in surface
anatomy; providing a more uniform dose inside the target volume.

Oblique incidence
It is unusual for a tumour to be situated on a flat piece of anatomy; typical sites are
around the face on sun-​exposed areas and there is often a problem in achieving close
applicator apposition to the area to be treated. This results in the problem of ‘stand-​off ’
which may be dealt with in one of two ways:
(i) The area may be made into a flat incident surface using bolus; the disadvan-
tage of this approach is that set up may be less accurate when the underlying
tumour cannot be seen at the time of applicator positioning.
(ii) The machine monitor units can be modified by a simple calculation based on
the inverse square law. The disadvantage of this approach is that the stand-​off
is unlikely to be constant across the treatment area and therefore the calcu-
lation may be based on the maximum stand-​off or mean stand-​off, both rep-
resenting a compromise. However in practice the effect is small, altering the
applied dose across the area by < 5%.
Ideally the electron beam should be perpendicular to the skin (or bolus) surface to en-
sure maximum penetration of therapeutic depth and the most uniform penumbra. As
the angle from perpendicular increases the following is noted (Fig 19.3):
◆ Surface dose increases.
◆ Depth of maximum dose decreases.
◆ Maximum dose increases.
◆ Therapeutic depth decreases.

Influence of dose inhomogenieties
Two main factors that account for the effect of inhomogenities on the dose
distribution are:
◆ Different absorption of various tissues, which depends on density of the tissues.
◆ Alteration in electron scatter pattern (scatter perturbation), which depends on
atomic number of the tissues.
 Squamous cell carcinoma and basal cell carcinoma 443

9 MeV
e
30º
ur fac
n ts
tie
Pa

Depth

120

100
0º

30º
80
Percent depth dose

45º

60º
60

Fig. 19.3  Effect of oblique

40 incidence on electron dose
distribution.
Reproduced with permission from
20 Hogstrom K.R. Electron beam
therapy: dosimetry, planning, and
techniques. In: C. Perez, l. Brady,
E. Halperin, R Schmidt-​Ullrich (eds).
Principles and Practice of Radiation
1 2 3 4 5 6 7 Oncology 5e, pp. 252-​282, Baltimore;
Depth (cm) Lippinkott, Williams, and Wilkins, 2003

In general the denser the material, the greater its absorption and scattering property
with regards electron beams. More electrons are scattered away from higher-​density
materials towards lower-​density material giving rise to hot spots under the low-​density
material. There are corresponding low-​dose areas under the high-​density region, re-
flecting the loss of electrons.
For small inhomogenities (small air cavities, small bony structures), the local scat-
tering of electrons at the edges is the predominant effect. Larger air cavities can result
in more significant hot or cold spots (up to 20%) with a deeper dose fall-​off region
(Figs 19.4 and 19.5).
The effect of patient anatomy (heterogeneity) must be accounted for in electron
beam planning to ensure:
◆ Adequate electron energy, i.e. no geographical miss of PTV in depth.
◆ Adequate dose homogeneity in PTV, i.e. minimal hot/​cold spots.
◆ Minimal dose to critical structures underlying PTV.
444 Skin cancer

Electron field matching
It is more difficult to match electron fields as the isodoses do not follow the geometric
edge of the beam. The high dose isodoses become narrower with depth while the lower
isodoses bow out with depth.
If no gap is used between fields then a hotspot is created at the matched edges of up
to 140%. If a gap of between 0.5 cm and 1 cm is used between fields this hotspot can
be reduced considerably.
Field shaping is achieved with standard electron endplates with preference for cir-
cular or elliptical fields. Straight-​edged fields are preferred if it is likely that further
treatment of nearby tissues will necessitate field matching. Customized endplates are
useful for irregular shapes or treatment near to a critical structure. These are made
from lead of an appropriate thickness for superficial-​X rays or an end frame cut-​out for
an electron applicator as shown in Fig. 19.3. Lead masks are used for areas near the eye
and nose only. The thickness of the Pb should be approximately half the beam energy
in millimetres. It is often valuable to extend the cut-​out to include locating anatom-
ical structures; e.g. the nose and superior orbital ridges on the face to facilitate correct
placement of the cut-​out for each treatment as shown in Fig. 19.5.

(a) 0°
12 MeV

0°
8 MeV

90 80
60 40
–35°
20
8 MeV 10
120

110
100

10 cm

Fig. 19.4a  Electron isodoses assuming lung equivalent to water.

 Squamous cell carcinoma and basal cell carcinoma 445

(b) 0°
12 MeV

0°
8 MeV

90
–35°
80
8 MeV
127
120 * 60

110
100 40

20

10

10 cm

Fig. 19.4b  Electron isodoses accouting for lung tissue.

Reproduced with permission from Hogstrom K.R. Electron beam therapy: dosimetry, planning, and
techniques. In: C. Perez, l. Brady, E. Halperin, R Schmidt-​Ullrich (eds). Principles and Practice of
Radiation Oncology 5e, pp. 252-​282, Baltimore; Lippinkott, Williams, and Wilkins, 2003.

Internal shielding prevents the radiation beam penetrating past a specific boundary;
e.g. the eye or eyelid, lip and areas of the mouth, and behind the ear. Corneal shielding
is used for tumours of the lower eyelid, and inner and outer canthi. Superficial X-​ray
treatments may use a patch or contact lens shield. Electron treatments use a similar
contact lens shield as shown in Fig. 19.4; an important feature of these shields is wax
coating to absorb scattered radiation from the electron beam. Internal mouth shields
or shielding behind the ear should also be covered with a layer of wax to absorb back-
scatter which is dependent on the energy and type of shield.

19.1.5 Verification
Verification is by clinical observation of the applicator set-​up and light beam on the
patient at each treatment. In vivo dosimetry may be used where treatment is close to
a critical structure, e.g. around the eye when lithium fluoride or diode measures are
recommended.
446 Skin cancer

17 MeV e–
10 × 10 cm2
100 cm SSD

Air

100 100
100
95 103
80 90
60
40
20
10

With Air
Without Air

Fig. 19.5  Effect of air on electron dose distribution.

Reproduced with permission from Hogstrom K.R. Dosimetry of electron heterogeneities. In: A. Wright
and A. Boyer (eds.), Medical Physics Monograph No. 9: Advances in Radiation Therapy Treatment
Planning, pp. 223-​243, New York; American Institute of Physics, 1983. © 1983 AIP Publishing LLC.

Indelible skin marks and margin tattoos together with photographs of the treatment
position are valuable to confirm the correct set-​up and applicator position.

19.1.6 Special considerations
◆ Tumours overlying bone are best treated with electrons because of the enhanced
absorbed dose in bone from superficial X-​rays.
◆ Electrons are preferred to superficial X-​rays for treatment of deep tumours and
those overlying cartilage e.g. nose, pinna
◆ Superficial X-​rays are favoured for tumours around the eye due to electron scatter
and the requirement for lead shielding.
◆ With superficial X-​rays crenellation of the margin of a round cut-​out which is then
rotated each day may give better cosmesis, blurring the edge of radiation reaction.
This should be allowed for in the PTV expansion.

19.1.7  Dose prescription for SCC

A range of dose prescriptions are in use. The nominal gold standard is 60 to 66 Gy in 30 to
33 fractions over 6 to 6½ weeks. More pragmatic alternatives are included in Table 19.1.
Doses are defined at the 100% isodose for both superficial X-​rays and electrons.
 Squamous cell carcinoma and basal cell carcinoma 447

Table 19.1  Dose prescriptions in use for SCC skin

18Gy/​1# Small tumours <3cm where cosmesis is relatively unimportant

24Gy/​2# Both fractions given five weeks apart, in elderly frail patients
where cosmesis is less important and target <5cm
35Gy/​5#/​1week For small lesions under 2 cm diameter ideally not when overlying
cartilage.
45Gy/​10#/​2 weeks: Target size up to 5 to 6cm diameter; this can be given as alternate
day fractions over a 3 week period if patient access to hospital is
limited, or in the elderly
55Gy/​20#/​4 weeks Target size up to 5 to 6cm diameter; this can be given as alternate
day fractions
60-​66Gy in 30-​33# Target < 6cm diameter, in areas of poor radiation tolerance
over 6-​6 ½ weeks
Target > 6cm at sites of
poor radiation

Reproduced with permission from the London Cancer skin cancer radiotherapy guidelines 2014. London
Cancer is part of the University College London Hospitals Cancer Collaborative, the Cancer Alliance for
north central and east London.

19.1.8  Dose prescription for BCC

In general BCCs are considered more radiosensitive and a nominal dose of 60 Gy in 30
fractions over six weeks is given. Equivalent doses are included in Table 19.2.

19.1.9 Radiation toxicities
Radiation toxicities after radiotherapy to skin sites are shown in Table 19.3.

Table 19.2  Dose prescriptions in use for BCC skin

18Gy/​1# Small tumours (<3cm)where cosmesis is relatively unimportant

24Gy/​2# Both fractions given five weeks apart, where cosmesis is less important
and target <5cm
35Gy/​5#/​1 week For small lesions under 2 cm diameter ideally not when overly cartilage
40.5Gy/​9#/​2 weeks Target size up to 5 to 6cm diameter; this can be given as alternate day
fractions over a 3 week period if patient access to hospital is limited, or
in the elderly
50Gy/​15#/​3 weeks Target is between 4 and 6cm diameter and not in an area of poor
radiation tolerance
55Gy/​20#/​4 weeks Target < 6cm, in areas of poor radiation tolerance
60Gy/​30#/​6 weeks Target > 6cm in areas of poor radiation tolerance

Reproduced with permission from the London Cancer skin cancer radiotherapy guidelines 2014. London
Cancer is part of the University College London Hospitals Cancer Collaborative, the Cancer Alliance for
north central and east London.
448 Skin cancer

Table 19.3  Radiation toxicity associated with irradiation of skin tumours

Acute toxicity (within 90 days) Late toxicity
Fatigue Hypopigmentation

Skin erythema Atrophy

Skin desquamation Telangiectasiae
Scabbing Alopecia
Alopecia Nasolacrimal duct stenosis (treatment near
medial canthi and nasal bridge)
Hyperpigmentation Failure of skin graft
Cartilage necrosis (treatments of nose
and ear)
Second malignancy

19.2 Melanoma
The primary management of malignant melanoma is by surgical excision. Close surgical
margins should be dealt with by re-​excision rather than radiotherapy. However, radio-
therapy may be used in the adjuvant treatment of the primary if adequate excision mar-
gins cannot be obtained, e.g. some head and neck sites. Desmoplastic melanomas have a
high local recurrence rate and a lower risk of regional and distant metastases. Adjuvant
radiotherapy for desmoplastic melanoma may offer superior local control compared to
surgery alone(8,9). Radiotherapy is rarely used in the adjuvant treatment of the nodal basin
as recent results of a Phase 3 randomized trial suggest high rates of grade 3 and 4 toxicity
and no overall survival benefit(10). Adjuvant radiotherapy may be best reserved for a fur-
ther local recurrence after a period of observation, if there has been no systemic relapse.
Radiotherapy can also be used in the palliation of inoperable or advanced disease.
Lentigo maligna is a non-​invasive melanoma typically occurring on the face of
elderly patients. Surgical resection is the treatment of choice due to the risk of bi-
opsy missing an invasive component and the risk of progression to an invasive lesion.
However, in patients unsuitable for resection, superficial radiotherapy can result in
high rates of local control in excess of 90%(11). A recent published review of the lit-
erature recommends wide treatment margins from GTV to field edge of 10–​20 mm
treating to a minimum depth of 5 mm; with the treatment volume defined using in
vivo reflectance confocal microscopy (RCM)(12.).

19.2.1 Dose prescription for melanoma

Adjuvant treatment of primary
In this setting techniques and doses as for squamous carcinoma are used.
Adjuvant treatment of the nodal basin
Appropriate dose fractionation schedules for adjuvant nodal irradiation are shown in
Table 19.4. Shorter hypofractionated schedules to the lymph node areas are likely to
increase late morbidity in terms of fibrosis and lymphoedema.
 Lymphoedema 449

Table 19.4  Melanoma adjuvant nodal radiotherapy

dose fractionation schedules
Suitable dose/​fractionation 40.05 Gy/​15 fractions
schedules:

50 Gy/​20 fractions
50 Gy/​25 fractions

19.3 Lymphoedema
19.3.1  Merkel cell tumours
Merkel cell tumours (MCC) are rare neuroendocrine tumours with biological fea-
tures analogous to small cell lung cancer. Given the propensity of MCC to recur
locally (sometimes with satellite lesions and/​or in transit metastases), wide local ex-
cision (WLE) to reduce the risk of local recurrence has been recommended. Optimal
minimum margin width and depth of excision around the primary tumour vary
among reports but most advocate 2–​5 cm margins. Primary radiotherapy after initial
biopsy of MCC may also be considered as an alternative to WLE, particularly in areas
where tissue preservation is preferred. Because of the aggressive nature of MCC, its
apparent radiosensitivity, and the high incidence of local and regional recurrences
(including in transit metastases after surgery alone to the primary tumour), adju-
vant radiation to the primary site and nodal basin has been recommended(13,14,15).
Those groups thought to benefit most from adjuvant radiotherapy include those
patients with:
◆ Large tumours.
◆ Locally unresectable tumours.
◆ Close or positive excision margins.
◆ Positive regional nodes (especially after sentinel lymph node dissection).

Primary/​adjuvant radiotherapy to the primary site

Techniques are as described above for squamous carcinoma using direct electron
beams with a margin to define the PTV of 2–​5 cm circumferentially and a minimum
1 cm deep margin.
Possible dose/​fractionation schedules include:
◆ 35 Gy/​5 fractions over 1 week.
◆ 45 Gy/​10 fractions over 2 weeks.
◆ 55 Gy/​20 fractions over 4 weeks.
◆ 60 Gy/​30 fractions over 6 weeks.

Prophylactic nodal irradiation
Megavoltage X-​ray treatment is given to the node areas as described in Chapter 16.
Dose/​fractionation schedules are as described for melanoma.
450 Skin cancer

19.4 Adnexal tumours
The treatment of adnexal tumours will be surgical excision and radiotherapy has no
recognized role in their primary treatment.
Close surgical margins should be dealt with by re-​excision rather than radiotherapy;
only where this is not possible because of co-​morbidity or technical considerations
should post operative radiotherapy be used. In this setting techniques and doses as
for squamous carcinoma are used. A 10 mm margin from CTV to define the PTVis
recommended.

19.5 Palliative treatment
19.5.1 Indications
Palliative treatment of skin tumours may be indicated for large fungating primary
sites, particularly in the elderly or medically frail, and for regional lymph nodes which
are inoperable.

Locally advanced and metastatic squamous cell carcinoma

Radiotherapy of the involved nodal basin is routinely used as adjuvant treatment after
a lymph node dissection has been shown to improve outcomes(16). For locally advanced
primary disease unsuitable for radical treatment, or for inoperable nodal metastases,
radiotherapy can be used in palliation.
Techniques for treatment will be the same as for radical primary treatment.
Exceptions may include the following:
• Deeper tumours may require higher energy electrons or on rare occasions be so
extensive as to require planned photon beam treatments.
• Fixed regional lymph nodes will be treated with photon beams in most in-
stances. Primary tumours in the head and neck region will drain to cervical
lymph nodes, and from the upper trunk and arm to axillary lymph nodes.
Lower limb lesions will involve inguinal lymph nodes. Treatment to these sites
will follow standard techniques as described in Chapter 16.
Dose prescription in this setting can present a difficult balance between the temptation
to give a short pragmatic palliative schedule such as 20 Gy in 5 fractions or 30 Gy in 10
fractions and the knowledge that if inadequate, further local relapse can be very diffi-
cult to manage and therefore a full radical dose should be attempted. Depending upon
the site, hypofractionated schedules such as 55 Gy in 20 fractions, 39 Gy in 13 fractions
or 30 Gy in 6 fractions may be effective for durable control. Larger fraction schedules
should be avoided in lymph node areas.
Systemic chemotherapy may be appropriate in patients with good performance
status. Common protocols include platinum and 5-​fluorouracil. The oral retinoid
acitretin has been shown to be useful in preventing disease recurrence and progres-
sion, in immunosuppressed patients and patients with early onset aggressive tu-
mours(17). More recently, nicotinamide (vitamin B3) has been shown to significantly
reduce the risk of non-​melanomatous skin cancers in high-​risk, immunocompetent
Special scenarios utilizing multiple electron fields or special electron procedures 451

individuals, offering a new chemo-​prophylactic measure easily transferrable into clin-

ical practice[18).

Basal cell carcinoma
Indications for palliative treatment will be the same as for squamous carcinoma except
that lymph node metastases are rarely seen with this tumour. Treatment of the primary
site will follow the same procedures and techniques as for primary radical treatment.
Systemic treatment with the oral hedgehog pathway inhibitor vismodegib is an ef-
fective palliative treatment for locally advanced or metastatic BCC and is generally
well tolerated. It may also be used in the neoadjuvant setting to reduce tumour bulk
prior to radical radiotherapy although this remains experimental.

Melanoma
Palliative treatment may be indicated for fixed inoperable lesions in the skin.
Satellite nodules may also become symptomatic and benefit from local radiotherapy.
Techniques will be the same as those described for primary radical treatment. Where
there are satellite nodules these should, as far as possible, all be included in the CTV.
Dose prescription for melanoma may include those quoted above for SCC. An al-
ternative is to use weekly or twice weekly doses of 6 Gy to a total dose of 30–​36 Gy.

Merkel cell tumours
Locally advanced Merkel cell tumours will respond readily to radiotherapy but are
typically associated with widespread metastatic disease in liver, bone, lungs, and brain.
Chemotherapy will often be more appropriate in this setting, favouring protocols used
to treat small cell lung cancer, e.g. platinum and etoposide, but local control can be
usefully obtained using palliative radiotherapy. Techniques will be as described above
for primary treatment and may have to include node areas also as regional lymph
node disease is common in advanced cases. For durable control doses such as those
discussed above for squamous carcinoma should be used.

Adnexal tumours
Adnexal tumours rarely present as a palliative problem but when they do so the tech-
niques above should be followed.

19.6  Special scenarios utilizing multiple electron

fields or special electron procedures
19.6.1  Total limb electron irradiation
Total limb electron irradiation (TLEI) is indicated in diffuse, symptomatic cutaneous
disease of the limbs in patients deemed unfit for chemotherapy or in those whose dis-
ease has not responded to previous systemic treatment. Common indications include
Kaposi’s sarcoma, recurrent malignant melanoma, and diffuse large B-​cell lymphoma.
The limb is exposed and immobilized. Fig. 19.6 demonstrates a six-​field limb ex-
tremity technique using six equally spaced 5 MeV electron beams. Each beam is wide
enough to cover the entire limb and the resulting distribution is superior to that of a
452 Skin cancer

Fig. 19.6  Total limb electron

irradiation (TLEI) using a
six-​field electron technique.
Reproduced with permission from
KR Hogstrom.

single field giving a homogenous surface dose of 90% circumferentially with a corres-
ponding isodepth dose of 8–​10 mm (Fig 19.6).

19.6.2  Total scalp irradiation

Total scalp irradiation (TSI) is indicated in cases of cutaneous tumours such as
lymphomas, angiosarcomas, or melanoma where there is widespread involvement of
the scalp and forehead. The main objective of treatment is to provide a uniform dose
to the target volume whilst minimizing dose to the brain. This can be challenging
due to the topology of the head and variation in depth of the target volume, as well as
the close proximity of the brain to the scalp and the effects of bone on dose distribu-
tion. Solutions include multiple electron beam techniques with field shifts(19) or com-
bined electron and photon fields(20). Helical tomotherapy is increasingly being used as
a means to deliver TSI, providing more homogenous dose distributions and avoiding
the need for field matching and the use of combined modalities(21). Mould brachy-
therapy is also an alternative approach to delivering total scalp irradiation.

References
1. Locke J, Karimpour S, Young G, et al. Radiotherapy for epithelial skin cancer. International
Journal of Radiation Oncology, Biology, Physics 2001; 51: 748–​55
2. Kwan W, Wilson D, Moravan V. Radiotherapy for locally advanced basal cell and
squamous cell carcinomas of the skin. International Journal of Radiation Oncology, Biology,
Physics 2004; 60: 406–​11
3. Griep C, Davelaar J, Scholten AN, et al.Electron beam therapy is not inferior to superficial
x-​ray therapy in the treatment of skin carcinoma. International Journal of Radiation
Oncology, Biology, Physics 1995; 1347–​50
 References 453

4. Telfer NR, Colver GB, Bowers PW. Guidelines for the management of basal cell
carcinoma. British Journal of Dermatology 1999; 141: 415–​23
5. Zagrodnik B, Kempf W, Seifert B, et al. Superficial radiotherapy for patients with basal
cell carcinoma: recurrence rates histologic subtypes and expression of p53 and Bcl-​2.
Cancer 2003; 98: 2708–​14
6. Jambusaria-​Pahlajani A, Miller CJ,Quon H, et al. Surgical monotherapy versus surgery
plus adjuvant radiotherapy in high-​risk cutaneous squamous cell carcinoma: a systematic
review of outcomes. Dermatologic Surgery 2009; 35:574–​85.
7. Han A, Ratner D. What is the role of adjuvant radiotherapy in the treatment of cuatneous
squamous cell carcinoma with perineural invasion? Cancer 2007; 109:1053–​9.
8. Guadagnolo BA, Prieto V, Weber R, et al. The role of adjuvant radiotherapy in the local
management of desmoplastic melanoma. Cancer 2014; 120:1361–​8.
9. Foote MC, Burmeister B, Burmeister E, et al.Desmoplastic melanoma: the role of
radiotherapy in improving local control. ANZ Journal of Surgery 2008; 78: 273–​6.
10. Henderson MA, Burmeister BH, Ainslie J, et al. Adjuvant lymph node field radiotherapy
versus observation only in patients with melanoma at high risk of further lymph node filed
relapse after lymphadenectomy (ANZMTG 01.01/​TROG 02.01). Lancet Oncology 2015;
16:1049–​60.
11. Farshad A, Burg G, Panizzon R, Dummer R. A retrospective study of 150 patients with
lentigo maligna or lentigo maligna melanoma and the efficacy of radiotherapy using Grenz
or soft x-​rays. British Journal of Dermatology 2002; 146:1042–​6.
12. Fogarty G, Hong G, Scolyer RA et al. Radiotherapy for lentigo maligna.A literature review
and recommendations for treatment. British Journal of Dermatology 2014; 170:52–​8.
13. Poulsen M. Merkel cell carcinoma of the skin. Lancet Oncology 2004; 5:593–​9.
14. Gollard R, Weber R, Kosty MP, et al. Merkel cell carcinoma: review of 22 cases with
surgical, pathologic and therapeutic considerations. Cancer 2000; 88:1842–​51.
15. Eng TY, Boersma MG, Fuller CD, et al. A comprehensive review of the treatment of
Merkel cell carcinoma. American Journal of Clinical Oncology 2007; 30: 624–​36.
16. Veness MJ. Treatment recommendations in patients diagnosed with high risk cutaneous
squamous cell carcinoma. Australsian Radiology 2005; 49:365–​76.
17. Niles RM. Recent advances in the use of vitamin A (retinoids) in the prevention and
treatment of cancer. Nutrition 2000; 16:1084–​9.
18. Chen A, Martin A Choy B, et al. A phase 3 randomized trial of nicotinamide for skin
cancer chemoprevention. New England Journal of Medicine 2015; 373:1618–​26.
19. Able CM, Mills MD, McNeese MD, et al. Evaluation of a total scalp electron irradiation
technique. International Journal of Radiation Oncology, Biology, Physics 1991; 21:1063–​72.
20. Tung SS, Shiu AS, Starkschall G, et al. Dosimetric evaluation of total scalp irradiation
using a lateral electron-​photon technique. International Journal of Radiation Oncology,
Biology, Physics 1993; 27:153–​60.
21. Orton N, Jaradat H, Welsh J, Tome W. Total scalp irradiation using helical tomotherapy.
Medical Dosimetry 2005; 30:162–​68.
 Chapter 20

Sarcomas of soft tissue and bone

James Wylie

20.1 Introduction
Sarcoma is an uncommon malignancy representing less than 1% of cancers. Both soft
tissue and bone sarcomas occur across all age ranges but bone tumours occur most
commonly in the paediatric and young adult range. This chapter will discuss both soft
tissue and bone tumours except for rhabdomyosarcomas which will be covered within
paediatric tumours (Chapter 21).

20.1.1  Soft tissue sarcoma

Soft tissue sarcoma in adults comprises some 50 different subtypes with leiomyosarcoma,
liposarcoma, and pleomorphic sarcoma being the commonest. The principles of man-
agement with radiotherapy are, however, broadly similar. Adjuvant radiotherapy is
routinely advised for intermediate and high grade sarcoma and randomized trials have
shown its effectiveness in reducing local recurrence.
Traditionally postoperative radiotherapy has been employed but there are persua-
sive arguments including a randomized study to support preoperative radiotherapy
which, while increasing acute post surgical toxicity, results in less late morbidity. This
approach may be particularly suited to myxoid liposarcomas where a substantial and
rapid reduction in size may be seen.

20.1.2 Bone sarcoma
The management of sarcomas of bone is much more varied and a complete review
outside the scope of this chapter. The most common malignant tumour which forms
bone is osteosarcoma; that forming cartilage is chondrosarcoma. Other sarcomas
arising from bone include Ewing’s sarcoma, chordoma, and pleomorphic sarcoma.
Chondrosarcoma and chordoma will usually be treated surgically in the first instance.
Chemosensitive tumours such as Ewing’s and osteosarcoma will typically be treated
with initial neo-​adjuvant chemotherapy followed by surgery. For Ewing’s sarcoma
radiotherapy is used as an adjuvant treatment or as an alternative to surgery when
this is not possible or its consequences unacceptable. In osteosarcoma, treatment of
the primary will be surgical unless the tumour is inoperable. Wide resection of skull
base and mobile spine chordoma and chondrosarcoma is rarely possible and because
these are relatively radioresistant tumours high dose adjuvant radiotherapy is often
indicated. Postoperative radiotherapy is normally indicated after marginal en bloc
excision of sacral chordomas. Proton radiotherapy is usually the preferred means of
 Radiotherapy for soft tissue sarcomas 455

dose escalation in these cases due to the steep dose gradient achievable next to critical
structures.

20.2  Radiotherapy for soft tissue sarcomas

(excluding rhabdomyosarcoma of young people type
or soft tissue Ewing’s sarcoma)
The use of adjuvant radiotherapy for the treatment of soft tissue sarcoma over the last
30 years has reduced the indications for major ablative surgery and wide local excision
is often sufficient. Low grade tumours, small tumours superficial to fascia and small
intramuscular tumours (< 5 cm) excised with a 1–​2cm margin may not require adju-
vant radiotherapy. In most other situations surgery alone is associated with a poorer
local control than combined modality treatment. Conventional treatment plans have
used a wide-​field technique, usually with a shrinking field over two phases. A brachy-
therapy trial at Memorial Sloane-​Kettering has suggested that smaller radiotherapy
volumes may be associated with similar outcomes in terms of local control and better
limb function. VORTEX, a UK trial, has shown that a modest reduction in external
beam field lengths is associated with similar local control but demonstrated no im-
provement in function.
Soft tissue sarcomas spread in a longitudinal direction within muscle groups in
the extremity and tend not to breach axial barriers such as major fascial, bone, and
interosseous membranes. In other sites the situation may be more complex. Surgery
should include resection of the biopsy track, and aim for an R0 resection (i.e. margin
negative). Assessment of the margin may be technically difficult and the nature of the
margin is as important as its size. The relapse rate is higher if the margin is positive—​a
failure rate which is reduced, but not eliminated, by radiotherapy. A  margin of 1–​
2 mm is probably adequate if postoperative radiotherapy is to be given, though smaller
margins are adequate if they include a fascial boundary. A ‘planned-​positive’ margin
by an experienced surgeon (e.g. on a neurovascular bundle) does not seem to greatly
increase local relapse.
Whilst, in the UK, it is common for radiotherapy to be administered postoperatively,
there is evidence that preoperative radiotherapy results in an equivalent control rate
and better long-​term function. This is given with a lower dose, and to a smaller target
volume as it is based on the undisturbed tumour. This approach is associated with
an increased postoperative morbidity, particularly in thigh lesions but improved
long-​term function with no deterioration in local control. The appropriateness of the
pre-​versus postoperative approach needs to be made on an individual patient basis
following detailed consultation between surgeon and clinical oncologist considering
any adverse factors relating to healing. It should not be expected that preoperative
radiotherapy will significantly reduce tumour size, but instead reduce the viability of
peripheral cells, R0 resections may be facilitated, and the potential for contamination
by tumour rupture reduced. Myxoid liposarcomas, however, may shrink rapidly after
radiotherapy, presenting the surgeon with the challenge of removing a tumour that
may have virtually disappeared. Where preoperative radiotherapy is used, an interval
of 4–​8 weeks between radiotherapy and surgery is optimal.
456 Sarcomas of soft tissue and bone

20.2.1  Indications for radiotherapy

◆ Preoperative:
• Where the resection is likely to be marginal or where positive margins are likely and,
especially, where tumours are close to but not surrounding neurovascular bundle.
• In peri-​articular locations to minimize dose to the joint.
• In locations close to critical structures where the reduced dose/​volume delivered
will be advantageous.
• In particularly radiosensitive tumours (e.g. myxoid liposarcoma).
◆ Postoperative:
• High grade and intermediate grade sarcomas unless small (< 5cm) and/​or super-
ficial and excised with a 1–​2 cm margin and acceptable salvage options exist if
recurrence develops.
• Low grade sarcomas with a close or positive margin where radical surgery at re-
lapse is likely to be difficult.

20.2.2  Essential investigations for planning radiotherapy

◆ Biopsy results if no resection.
◆ Staging computed tomography (CT) thorax.
◆ Operation note and resection histology.
◆ Preoperative or pre-​chemotherapy contrast-​enhanced magnetic resonance imaging
(MRI) scans of the primary affected region, including the entire involved bone and
adjacent joints.
◆ CT may add information in bone tumours.

20.2.3 Patient preparation

◆ Patients with limb, shoulder girdle, or head and neck tumours should be immo-
bilized in individually made casts. These should be comfortable and stable to aid a
reproducible set-​up.
◆ When the pelvis or proximal thigh is irradiated, sperm storage and oophoropexy
should be considered to spare fertility. Intensity-​modulated radiotherapy (IMRT)/​
volumetric-​modulated arc radiotherapy (VMAT) and protons may allow better
sparing of female reproductive organs and reduce the need for oophoropexy.
◆ A spacer may be needed before radiotherapy to displace bowel away from the high-​
dose areas but the improved conformality achieved with IMRT/​VMAT often make
this unnecessary.
The position for planning should be individualized depending on the precise pos-
ition of the tumour within the limb, with the cast and a point of fixation preferably
around the distal limb and away from the tumour. Vacuum bean bags may be of
value in truncal tumours. Devices should provide reproducibility of 3 mm in the
head and neck and 5–​7 mm elsewhere, although at sites such as the shoulder this
may be difficult. Local audits should determine the individual institutional repro-
ducibility of the treatment set-​up. Treatment position will usually be supine as this
is considered more stable, but in patients with posterior pelvic tumours, prone may
 Radiotherapy for soft tissue sarcomas 457

be preferable in thin patients as this displaces bowel, reduces skin dose, and aids
field positioning.

20.2.4  Planning imaging required for target definition

◆ 3–​5 mm CT scan slices of the tumour region, covering 5 cm cranial and caudal to
the entire extent of the surgical bed and scar in the postoperative setting and 5 cm
around the wired tumour in preoperative cases. The scar and any drain sites should
be marked with wire.
◆ Intravenous contrast use at the clinician’s discretion if there are no contraindications.

20.2.5 Target definition

Gross tumour volume
◆ Preoperative radiotherapy:
• Visible extent of tumour on planning CT scan with reference to the diagnostic
gadolinium enhanced T1/​T2-​MRI. Consider image fusion.
◆ Postoperative radiotherapy:
• Virtual gross tumour volume (GTV) reconstructed using the preoperative
imaging, operation note, and pathology report.
• This principle needs to be applied with care as the anatomy of the region may
have changed.

Clinical target volume
The clinical target volume (CTV) includes all tissues potentially involved in micro-
scopic spread. Small studies have detected tumours cells several centimetres from the
tumour (and mostly within peri-​tumoural oedema). Therefore, any peri-​tumoural oe-
dema should be included in the CTV. In postoperative cases it is usual to include the
whole operative bed for the majority of the treatment (CTV1), as this is considered
at risk of contamination, and includes the scar, drain sites, and manipulated tissues.
It normally includes any postoperative seroma unless this would result in excessive
field sizes. Where a prosthesis has been placed, it is normal to include its full length.
However, there is no good evidence base for this and, where this would compromise
function, it is reasonable not to attempt full inclusion especially if joints or, in growing
children, epiphyses would be included.
The geometrically grown margin may be modified by barriers to spread, such as
deep fascia or bone, or by areas of weakness produced by the neurovascular bundle
passing through fascia.
The CTV should be modified to ensure a ‘corridor’ of tissue is spared from the high
dose volume and to spare joints or other critical structures as shown in Fig. 20.1.
◆ Preoperative radiotherapy:
• CTV length = GTV + 3–​4 cm.
• CTV width = GTV + 1.5–​2 cm depending on site.
• CTV can be edited at interface with uninvolved fascia or bone or joint space.
◆ Postoperative radiotherapy:
• CTV phase 1:
458 Sarcomas of soft tissue and bone

Fig. 20.1  Postoperative radiotherapy for soft tissue sarcoma overlying the right hip joint: 6
MV treatment using inverse planned IMRT. Improved conformality allows reduced dose
to underlying bone/​joint without compromise to PTV. (Virtual GTV orange, CTV dark blue
including seroma, PTV light blue). Note editing of CTV around bone/​soft tissue interface.

• CTV1 length = GTV + 4–​5 cm, or scar + 1 cm, whichever is longer.

• CTV1 width = as per preoperative margins with additional editing to include
surgical clips and drain sites.
• CTV phase 2:
• CTV2 length = GTV + 2 cm.
• CTV2 width = unchanged. For non-​limb sites consider a volume reduction if
possible to spare normal tissues.

Planning target volume (PTV)

o PTV = CTV + 5–​10 mm depending on site and local audit of set-​up.
o Consider larger CTV-​PTV margin in shoulder girdle tumours and trunk sites.

20.2.6  Field arrangement and dose distribution

IMRT/​VMAT/​tomotherapy are now commonly employed in sarcoma planning util-
izing a simultaneous integrated boost (Fig. 20.1). Care should be taken with limb posi-
tioning and beam arrangement with IMRT to limit dose to the contra-​lateral limb.
If the CTV is superficial, thought should be given to the possible underdosing of
superficial tissues. This will be reduced with tangential rather than direct incident
 Radiotherapy for soft tissue sarcomas 459

fields. Use of bolus should be avoided where possible due to the significant adverse
effects on skin and immediate subcutaneous tissue. Occasionally in the wrist or ankle,
due to small separation, or where the skin is involved, the use of bolus for part of the
treatment may be necessary to ensure the tumour bed itself is adequately treated.
Dose specification is according to the ICRU 50, 62, and 83 reports

20.2.7 Dose fractionation
There are no good randomized trials addressing this issue but a dose of at least 60
Gy is required postoperatively. North American practice tends to use doses of 66 Gy,
whereas 60 Gy has been usual in UK sarcoma centres, with consideration of a further
6 Gy in margin-​positive disease.
◆ Preoperative radiotherapy:
• 50–​50.4 Gy in 25–​28 daily fractions over 5–​5½ weeks.
◆ Postoperative radiotherapy:
• Resection margin-​negative:
• Phase 1: 50 Gy in 25 daily fractions over 5 weeks.
• Phase 2: 10 Gy in 5 daily fractions over 1 week.
Alternatively 52.2 Gy to CTV1 and 60 Gy to CTV2 in 30 fractions SIB.
• Resection margin-​positive:
• Phase 1: 50 Gy in 25 daily fractions over 5 weeks.
• Phase 2: 16 Gy in 8 daily fractions over 1½ weeks.

20.2.8  Critical organs and tolerance doses

Limb
◆ A  strip of unirradiated skin and subcutaneous tissues sufficient to maintain the
lymphatic drainage of the distal limb (the ‘corridor’) is essential for acceptable long-​
term function. Data on the amount are not evidence based but, in principle, between
25 and 33% of the circumference including subcutaneous tissues is kept below 40
Gy. It should be borne in mind that the adverse effect on lymphatic draining will
depend on the most narrow part of the ‘corridor’, such that a relationship to a dose–​
volume histogram is difficult to define.
◆ Radiation induced fractures are reported in 3–​4% of cases and are dependent on
dose, sex (F > M) and degree of bone resection. Reducing the volume of bone re-
ceiving > 40 Gy to < 64% and mean dose to < 37 Gy is reported to reduce fracture
risk. Achieving such dose constraints can be challenging, especially around the fem-
oral necks, and IMRT/​VMAT may assist (Fig. 20.1).
◆ The skin and subcutaneous tissues of the anterior lower limb tolerate high dose
radiotherapy relatively poorly. Soft tissue tumours at this site in the very elderly
or those with pre-​existing poor tissue perfusion may not be suitable for high dose
radiotherapy and the management plan may require modification.

Neurological
◆ When treatment of the tumour itself is, or was, adjacent to the spinal cord, the cord
dose is limited to 50–​55 Gy in 1.8 Gy fractions but more generally, the maximum
cord dose is kept to < 50 Gy.
460 Sarcomas of soft tissue and bone

◆ When brachial plexus is in the field, the dose should be limited to 60 Gy in 1.8 Gy
fractions, with a maximum of < 65 Gy to 0.1 cc.
◆ Cauda equina mean dose should be < 60 Gy.
◆ Peripheral nerves tolerate doses of 60–​66 Gy in 2 Gy fractions.

Lung
◆ Lung V20 Gy should be < 30% or mean dose < 23 Gy.
Usual tolerances for other organs at risk are followed but with special caution in
view of the possible enhancement due to chemotherapy given prior to, or concurrent
with, radiotherapy.

20.3  Radiotherapy for bone sarcomas

20.3.1  Ewing’s sarcoma (including extra-​skeletal Ewing’s)
Patients receive several cycles of chemotherapy preoperatively or prior to rad-
ical radiotherapy and continue with chemotherapy concurrent with radiotherapy,
with omission of doxorubicin and actinomycin D within 2 weeks of radiotherapy
and of ifosfamide when there is a significant volume of bladder in the target
volume.
Patients with axial tumours who will require radiotherapy are excluded from treat-
ment with high-​dose busulphan due to enhanced toxicity.

Radiotherapy indications
◆ Inoperable tumours (or operable tumours where morbidity of surgery is not con-
sidered justified).
◆ Preoperative radiotherapy may be considered when a marginal excision is pre-
dicted, where resection of the whole pre-​chemotherapy volume is unlikely or where
there are technical advantages to preoperative radiotherapy such as reduced volume
or less dose to sensitive structures.
◆ Postoperative when one or more of the following applies:
• Margin positive or < 1mm at histology.
• Poor histological response to preoperative chemotherapy (< 90% necrosis).
• Concern that the pre-​chemotherapy tumour extent has not been adequately
excised.
• When patient presents with a displaced fracture.

Gross tumour volume
In all cases the GTV is the visible tumour on the planning CT scan extended to re-
flect the tumour at its greatest extent prior to chemotherapy or surgery, with refer-
ence to the initial diagnostic imaging. Protrusion into body cavities that regresses with
chemotherapy is not included.

Clinical and planning target volume

‘Grown margins’ are edited to take into account patterns of spread and intact fascial
planes:
 Radiotherapy for bone sarcomas 461

◆ CTV 1 (phase 1 postoperative only) = GTV + 1.5–​2 cm in the axial plane and 2–​
3 cm in the long axis of the limb modified to cover scar/​drain site/​prosthesis plus
1cm, whichever is longer.
◆ CTV 2 (all other indications including phase 2 postoperative) = GTV + 1.5–​2 cm
with no need to cover prosthesis/​scar.
◆ PTV = CTV + 0.5–​1 cm.

Current European dose fractionations

◆ Definitive: 54 Gy in 30 daily fractions over 6 weeks, +/​-​5.4 Gy boost in 3 daily frac-
tions if tumour >8 cm at diagnosis.
◆ Postoperative: 45 Gy in 25 daily fractions over 5 weeks to PTV1 and 9 Gy in 5 daily
fractions to PTV2 (consider further boost of 5.4 Gy in 3 daily fractions for tumours
>8 cm at diagnosis).
◆ Preoperative: 50.4 Gy in 28 daily fractions over 5½ weeks.

Radical whole lung radiotherapy for Ewing’s sarcoma

As consolidation when there were initial lung metastases at diagnosis or pleural effu-
sion. It should not be given following high-​dose chemotherapy and is only delivered
after standard chemotherapy is completed.
◆ CTV is the whole pleural cavity.
◆ Radical treatment is CT planned with the prescription point within the lungs rather
than the mediastinum.
◆ < 14 years: dose 15 Gy in 10 daily fractions over 2 weeks.
◆ > 14 years: dose 18 Gy in 12 daily fractions over 2½ weeks.

Treatment of bony secondaries
◆ As consolidation of chemotherapy response to a limited number of sites of bony dis-
ease. There is evidence that this may be associated with improved progression-​free
survival.
◆ 40–​45 Gy in 15–​20 daily fractions. Modified according to number of metastases
being treated.

20.3.2  Radiotherapy for osteosarcoma

Indications
◆ Tumour unresectable for technical and/​or medical reasons. High-​dose radiotherapy
delivered with protons or heavy ions may confer additional benefits.
◆ Resection margins positive and further surgery not possible.
◆ Inadequate resection as defined by intralesional resection or where margins are
contaminated.

Chemotherapy
◆ All patients treated radically receive induction combination chemotherapy and sur-
gery if primary site resectable.
462 Sarcomas of soft tissue and bone

◆ Radiotherapy, if indicated, is usually given after chemotherapy is completed—​

interval 4–​6 weeks.
◆ In inoperable tumours, concurrent cisplatin 60–​100 mg/​m2 may be given 3-​weekly
starting on day 1 or 2, radiotherapy starting 4–​6 weeks after completion of combin-
ation chemotherapy.

Dose fractionation
◆ Adjuvant to surgery:
• 60 Gy in 1.8–​2 Gy daily fractions over 6–​6½ weeks.
◆ Inoperable or macroscopic residual disease:
• 70-​75 Gy in 1.8–​2 Gy daily fractions over 7–​8½ weeks.

Gross tumour volume
◆ In unresected disease, the GTV is the visible extent of tumour on planning CT scan
with reference to the diagnostic imaging.
◆ In postoperative radiotherapy, the GTV is reconstructed using the preoperative
imaging, operation note, and pathology report.

Clinical target volume
CTV = GTV + 2–​3 cm, (this may be amended axially to spare a soft tissue corridor, but
should be at least 2 cm) or full length of scar + 0.5–​1 cm whichever is longer.
◆ Inoperable tumours:
• CTV = GTV + 2–​3 cm.
In all cases CTV is to be edited to take into account patterns of spread and intact fa-
scial planes.

20.3.3  Chondrosarcoma and chordoma

These tumours have not generally been considered sufficiently radiosensitive to benefit
from radiotherapy except, occasionally, for palliation. However, there is evidence that
doses of around 70Gy delivered peri-​operatively or 75–​79 Gy as definitive treatment
can result in long-​term control. Such dose escalation generally requires referral to a
proton centre Fig. 20.2.

20.4  Particular anatomical sites

20.4.1 Forearm
A particular challenge is obtaining a reproducible position for the arm. A slight pro-
nation/​supination of the hand will result in rotation along the axis and lead to set
up errors. Other difficulties can be creating an unirradiated corridor near the narrow
wrist. The major fascia boundary here is the interosseous ligament and as far as possible
the distal forearm the radius and ulna should be parallel. This will facilitate sparing of
one compartment. Full arm immobilization, grab handles, and vacuum bean bags may
all assist in this.
 Particular anatomical sites 463

% of
VMAT IMRT 3D CRT prescribed
(2 arcs) (7 fields) (7 fields) dose
120

105
95
90
85

65

35

10
0
Proton pencil beam scanning
(4 fields)
Fig. 20.2  Inoperable osteosarcoma right iliac crest in 17-​year-​old female. Treated with
protons to 70.2 Gy in 39 fractions following induction chemotherapy. Comparison
dose wash achieved with optimally planned VMAT, IMRT, 3DCRT, and protons showing
improved conformality in isodoses < 65%. Beam direction arrowed.
Reproduced with kind permission from Adam Aitkenhead.

20.4.2  Hands and feet

There has been concern about radiotherapy to these areas because of the perceived
poor tolerance to radiotherapy, and poorly defined fascial planes making target defin-
ition more difficult. The small separation creates difficulties in ensuring the superficial
parts of the target volume receive an adequate dose without the excessive dose to the
skin that results from full thickness build-​up. The tissues on the sole of the foot may
experience significant acute toxicity. The heel pad may experience marked swelling,
with a collection of haemorrhagic fluid under the thick keratin layer, and the toes may
be subject to acute and chronic trauma. The Achilles tendon is at risk of rupture and, as
far as possible, at least part of its width should be spared from doses > 50 Gy. However,
with care in planning and careful monitoring of the patient for excessive acute and
marked late toxicity, full or near full dose radiotherapy is possible with good local con-
trol and acceptable morbidity.

20.4.3 Retroperitoneum
This is a difficult site to irradiate postoperatively to adequate doses. Pre-​operative
radiotherapy is made easier by the tumour acting as its own spacer to displace nearby
critical structures and the lower dose (50 Gy in 25 daily fractions) required. There is
some evidence that local control may be improved by such strategies. This may require
complex inverse planned techniques. The benefit of preoperative radiotherapy is being
investigated in the EORTC 62092-​22092 STRASS trial.
464 Sarcomas of soft tissue and bone

Table 20.1  Indications for radiotherapy in adult soft tissue sarcomas

Stage Surgical margin/​location Plan
I A/​B Marginal Re-​resection or radiotherapy depending on impact of
surgery on functiona

Wide Watch policy

II Superficial Surgery alone
Deep Surgery + XRT unless intramuscular tumour with
1–​2 cm margin
III A/​B Surgery + XRT (XRT may be avoidable if superficial
and excised with 2 cm margin)
a
Unless atypical lipomatous tumour/​well differentiated liposarcoma where simple excision alone will suffice.

20.4.4 Chest wall
The indications for radiotherapy are the same, in principle, as elsewhere (see Table
20.1). Defining the GTV can be more difficult and wider CTV margins are often
employed. Chest wall respiratory movements may necessitate larger PTV margins.
IMRT or VMAT may permit improved sparing of underlying lung tissue. Where
there is cytology positive or haemorrhagic pleural effusion in the chest wall of a
patient with Ewing’s, irradiation of the whole pleural cavity (to the same doses as
adjuvant lung radiotherapy), may be advocated. This needs to be planned bearing
in mind the indications for adjuvant radiotherapy to the tumour bed and it may
not be possible to add the pleural cavity without exceeding recommended lung
tolerance doses.

20.4.5  Spinal/​para-​spinal
Radiotherapy to these locations can be particularly challenging due to the proximity of
the spinal cord. IMRT or VMAT may permit adequate conformality to keep the spinal
cord within tolerance. More challenging cases may be better treated with proton radio-
therapy which permits steeper dose gradients adjacent to the spinal cord but can be
more severely affected by metalwork compared to photon treatments. In some cases it
is only possible to treat parts of the CTV to cord tolerance.

20.5 Particular histologies
20.5.1 Angiosarcoma
Most commonly seen after previous radiotherapy these tumours are usually treated
with radical surgery. However, de novo angiosarcomas do occur, most often in the
scalp, which may be too extensive, or recur after, extensive surgery. These can
prove to be very radiosensitive and good local control can be achieved with mod-
erate dose radiotherapy either using external beam radiotherapy or brachytherapy
Fig. 20.3.
 Palliative treatment 465

Fig. 20.3  74-​year-​old man with extensive angiosarcoma involving left cheek, nasal ala,
and upper lip. Treated with 50 Gy in 25 fractions using simple anterior 6 MV field with
bolus. On the right complete clinical response at 12 months.

20.5.2 Fibromatosis
An EORTC trial has shown moderate dose radiotherapy can be effective in controlling
progressive and symptomatic disease. Common regimens include 50.4–​56 Gy in 28
daily fractions.

20.6  Particular radiotherapy techniques

20.6.1 Protons
The rapid fall-​off in dose at depth achievable with protons can be advantageous in
treating certain sarcomas. In paediatric and young adults the reduced integral dose
to normal tissues should reduce second malignancy risk and late effects, as shown
in Fig. 20.4. In older patients the steep dose gradient may permit improved target
volume coverage when tumours lie adjacent to critical structures, may spare exit dose
to sensitive adjacent organs (e.g. kidney, reproductive organs), and may produce
better coverage of particularly complex target volumes. However, there is normally
less sparing of superficial tissues leading to more severe acute skin reactions.

20.7 Palliative treatment
Depending on the site and comorbidities of the patient, soft tissue and bony-​based
limb masses may be palliated with one of the following regimens:
◆ 6 Gy × 5 or 6 treating weekly.
◆ 40–​45 Gy in 15–​20 fractions.
◆ 30 Gy in 10 fractions.
The dose for cord compression due to secondaries will depend on the anticipated
survival and tumour types but dose of 20 Gy in 5 fractions or 30 Gy in 10 fractions are
usually appropriate.
466 Sarcomas of soft tissue and bone

% of
VMAT IMRT 3D CRT prescribed
(1 arc) (5 fields) (3 fields) dose
120

105
95
90
85

65

35

10
Proton pencil beam scanning 0
(3 fields)
Fig. 20.4.  4-​year-​old with Ewing’s sarcoma of thoracic spine. Treated using protons to a
dose 50.4 Gy in 28 fractions. Comparison dose colour wash between optimally planned
VMAT, IMRT, 3DCRT, and protons. Protons demonstrate considerable reduction in dose
to heart, lung and liver.
Reproduced with kind permission from Adam Aitkenhead.

For pain without significant mass effect a single dose of 8 Gy may produce relief.

20.7.1 Cerebral secondaries
Treatment follows the general principles for cerebral secondaries in other tumour
types. Osteosarcoma and soft tissue tumours may occasionally present with isolated
secondaries which, especially if at an interval after initial treatment are best treated
surgically or with radiosurgery. Ewing’s more often presents with multiple lesions and
whole brain radiotherapy to 20 or 30 Gy may be used.

20.7.2 Lung secondaries
Lung resection remains the standard of care for selected patients who develop lung
metastases. However, if patient fitness or wishes preclude this consideration can be
given to stereotactic ablative radiotherapy which offers an effective alternative with
low morbidity. Whole lung radiotherapy may occasionally have a palliative role in
Ewing’s in the same dose schedule used for radical treatment.

Further reading
Casali PG, Blay JY. ESMO/​CONTICANET/​EUROBONET Consensus Panel of experts Soft
tissue sarcomas: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-​up.
Annals of Oncology 2010; 21(Suppl 5):v198–​203.
Davis AM, O’Sullivan B, Turcotte R, et al. Canadian Sarcoma Group; NCI Canada Clinical
Trial Group Randomized Trial. Late radiation morbidity following randomization
 further reading 467

to preoperative versus postoperative radiotherapy in extremity soft tissue sarcoma.

Radiotherapy and Oncology 2005; 75:48–​53.
ESMO/​European Sarcoma Network Working Group. Bone sarcomas:ESMO Clinical Practise
Guidelines for diagnosis, treatment and follow-​up. Annals of Oncology 2012; 23:100–​109.
Haas RLM, DeLaney TF, O’Sullivan B, et al. Radiotherapy for management of extremity
soft tissue sarcomas: why, when, and where? International Journal of Radiation Oncology,
Biology, Physics 2012; 84:572–​80.
Hogendoorn PC. ESMO/​EUROBONET Working Group. Bone sarcomas: ESMO Clinical
Practice Guidelines for diagnosis, treatment and follow-​up. Annals of Oncology 2010;
21(Suppl 5):v204–​13.
Jebsen NL, Trovik CS, Bauer HC, et al. Radiotherapy to improve local control regardless of
surgical margin and malignancy grade in extremity and trunk wall soft tissue sarcoma: a
Scandinavian sarcoma group study. International Journal of Radiation Oncology, Biology,
Physics 2008; 71:1196–​203.
Kaushal A, Citrin D. The role of radiation therapy in the management of sarcomas. Surgical
Clinics of North America 2008; 88:629–​46,  viii.
Patel S, DeLaney TF. Advanced-​technology radiation therapy for bone sarcomas. Cancer
Control 2008; 15:21–​37.
Tseng WH, Martinez SR, Do L, et al. Lack of survival benefit following adjuvant radiation in
patients with retroperitoneal sarcoma: a SEER analysis. Journal of Surgical Research 2011;
168:e173–​80.
Zagars GK, Ballo MT, Pisters PW, et al. Prognostic factors for patients with localized
soft-​tissue sarcoma treated with conservation surgery and radiation therapy: an analysis of
225 patients. Cancer 2003; 97:2530–​43.
 Chapter 21

Principles of paediatric
radiation oncology
Henry C Mandeville

21.1 Introduction

21.1.1 Incidence
The biology and types of tumour seen in childhood is diverse and very different from
that found in adults. In children, under the age of 16 years, malignancy remains a rare
occurrence, with between 1500 and 1700 cases each year in the UK, although there
has been an 11% increase in the incidence since the early 1990s. Data on the relative
incidence of different paediatric tumour types from Cancer Research UK is given in
Table 21.1. Overall the outcomes for the majority of paediatric cancers are better than
those seen in adult oncology, with approximately 70–​80% long-​term survivors. This
improvement over the last 40 years has been brought about largely through the in-
corporation and refinement of chemotherapy as part of a multimodality treatment
approach, combined with improvements in both surgery and radiotherapy techniques,
and aided by more advanced imaging techniques, and more accurate staging and bio-
logical classification.
The management of paediatric cancers is highly specialized and truly multidiscip-
linary. In the UK, treatment is coordinated through the network of Children’s Cancer
and Leukaemia Group (CCLG) paediatric oncology centres. The CCLG hosts regular
national tumour specific interest groups that cover the wide range of paediatric tu-
mour types, and also hosts treatment modality specific discipline groups, including
one for paediatric radiotherapy which engages interested clinicians, radiographers,
and other allied health professionals. It provides national guidelines for the manage-
ment of paediatric tumours, where trial protocols are not available, and also supportive
information for children and their families. In 2014 the CCLG, in collaboration with
the Royal College of Radiologists and the College of Radiographers, produced the first
edition of ‘Good Practice Guide for Paediatric Radiotherapy’, defining best practice in
the UK(1).
Given the rarity of these types of tumours, international collaboration is essential
to achieve robust and meaningful clinical trials; therefore the majority of paediatric
oncology studies open in the UK are undertaken by European collaborative study
groups aligned to the International Society of Paediatric Oncology in Europe (SIOPe).
Previously the CCLG was the body in the UK responsible for paediatric oncology
 Introduction 469

Table 21.1  Cancer Research UK Annual Incidence of paediatric malignancies in UK

Children’s Cancers (the 12 ICCC Diagnostic Groups): 2006–2008
Average Number of New Cases Diagnosed per Year, Children (0–14), Great Britain
Cancer Type Boys Girls Childrens
Leukaemia 262 205 468

Brain Other CNS and Intracranial Tumours 219 193 412

Lymphomas 115 51 167
Soft Tissue Sarcoma 57 41 98
Sympathetic Nervous System (SNS) Tumours 43 40 83
Renal Tumours 38 44 82
Bone Sarcoma 35 30 65
Carcinomas and Malignant Melanoma 25 30 55
Germ Cell and Gonadal Tumours 24 30 55
Retinoblastoma 22 22 44
Hepatic Tumours 9 10 18
Other and Unspecified Cancers 5 3 8

Leukaemia
Brain Other CNS and Intracranial Tumours
Lymphomas
Soft Tissue Sarcoma
Sympathetic Nervous System (SNS) Tumours
Cancer Type

Renel Tumours
Bone Sarcoma
Carcinomas and Malignant Melanoma
Germ Cell and Gonadal Tumours
Retinoblastoma
Hepatic Tumours
Other and Unspecified Cancers

0 80 160 240 320

Average Number of cases per Year

Boys Girls

Reproduced from Cancer Research UK, https://​www.cancerresearchuk.org/​sites/​default/​files/​cstream-​node/​

child_​inc_​iccc_​groups_​0.pdf, Accessed July 2018.

research, but in recent years this responsibility has devolved to the National Cancer
Research Institute (NCRI) Children’s Cancer and Leukaemia (CCL), and the Teenage
& Young Adult (TYA) clinical studies groups (CSG). In the USA, the main collab-
orative study group is the Children’s Oncology Group (COG), which leads paediatric
470 Principles of paediatric radiation oncology

oncology research in the USA and Canada, and incorporates some other countries
including Australasia; the COG was formed in 2000 from the amalgamation of the
previous Paediatric Oncology Group (POG), Children’s Cancer Group (CCG),
Intergroup Rhabdomyosarcoma Study Group (IRSG), and National Wilms’ Tumour
Study Group (NWTS).
Radiotherapy remains an important curative therapeutic modality for paediatric
malignancies, and the majority of paediatric tumours are radiosensitive. However, ap-
proximately 73% of long-​term survivors will be affected by long-​term consequences
from their treatment, including from radiotherapy(2). For radiotherapy these effects
relate both to the dose given and to the site that was treated, and include soft tissue
hypoplasia, impaired bone growth, neuropsychological/​cognitive effects, vascular dis-
ease, and radiation-​induced malignancy. The majority of toxicities relate to the tissues
receiving the highest dose and for children even minor improvements in radiotherapy
delivery have the potential to result in major long-​term benefits. In an effort to re-
duce the long-​term consequences of radiotherapy, there has been increased adoption
of proton beam therapy, for a variety of paediatric malignancies, and brachytherapy
for selected soft tissue sarcomas. Potential late effects resulting from chemotherapy
need to also be considered including late myocardial damage due to anthracyclines,
nephrotoxicity due to cisplatin or ifosfamide, and secondary leukaemias related to
drugs, such as etoposide and alkylating agents(3).
Currently approximately 40–​50% of children with cancer receive radiotherapy as
part of their initial treatment. Careful planning and delivery are essential in order to
maximize local tumour control whilst minimizing the late effects resulting from ir-
radiation of normal tissues. It is extremely important that, as for the administration
of chemotherapy, radiotherapy for children should be undertaken only in specialized
paediatric radiotherapy centres, associated with the CCLG paediatric oncology
centres or in approved proton therapy centres. The benefits of these centres, where
relatively large numbers of children are treated with radiotherapy, include being
able to establish expert multi-​professional support teams with specialist paediatric
therapy radiographers, specialist paediatric nurses, and particularly play specialists.
Young children frequently find it challenging to lie still for radiotherapy planning
and delivery, and can also be distressed by the radiotherapy tattoos and the making of
thermoplastic masks in the mould room. The support of an experienced radiotherapy
play specialist is essential when preparing children for radiotherapy and has been
shown to reduce the requirement for anaesthesia in children over the age of 3 years(1).
However, short-​acting intravenous general anaesthesia, such as propofol, is routinely
utilized for the immobilization of infants and older children with learning difficul-
ties or behavioural problems. The daily fasting for this results in surprisingly little
disruption to nutrition, although it can cause some challenges if hyperfractionated
schedules are employed.
When designing treatment strategies for paediatric malignancies, it is essential to
consider the likely long-​term consequences of treatment and the quality of life of sur-
vivors, in addition to maximizing the chance of cure. Tailored long-​term surveillance
of survivors to monitor the late effects of treatment is considered best practice and
 Proton therapy for paediatric tumours 471

in most centres this is carried out in dedicated late effects follow-​up clinics. National
treatment-​related guidelines for long-​term follow-​up have been produced by collab-
orative groups including the CCLG.

21.2  Proton therapy for paediatric tumours

Proton therapy is being increasingly used internationally for the treatment of paedi-
atric tumours and has been routinely funded for specific indications in the UK
since 2010. Since this time the age range and number of indications have steadily
increased, with teenagers and young adults (TYA) now also included (see Table
21.2). Numerous planning studies have demonstrated clear dosimetric advantages
for protons, these benefits mainly relating to the Bragg Peak, the critical point at
depth in tissue where the proton releases its energy then comes to rest, and beyond
which there is almost no deposition of dose. This differs from high energy photons
where there is an exit dose of radiation, the beam being gradually attenuated as it
passes through tissue beyond the target. It is possible to achieve high target volume
conformity of dose with protons, whilst irradiating less normal tissue compared to
photon techniques, especially intensity modulated radiotherapy (IMRT) where a
much greater proportion of normal tissues are exposed to low doses of radiation,
described as the ‘low-​dose bath’.
Primarily protons are used for paediatric malignancies for their high conformity,
with excellent sparing of adjacent organs at risk (OAR), and the reduced exposure of
tissues to irradiation, with the aim of reducing second malignancies and other late
effects. The main disease sites where the potential benefits are greatest include the
brain, skull base, pelvis, and spine, and tumours requiring craniospinal irradiation
(see Fig 21.1).
The benefits for extremity tumours remain more uncertain, and for sites where there
is significant motion affecting the target, and/​or adjacent OAR, as this can prove chal-
lenging for safe and precise delivery of proton therapy, due to the range uncertainties
resulting from changing tissue density that impact on proton dose deposition.
For the majority of tumour types an equivalent dose of radiation is delivered with
protons, using a relative biological effectiveness (RBE) of 1.1 with relation to high en-
ergy photons, aiming to achieve similar levels of tumour control in patients. Despite
the encouraging data from planning studies, and the widespread adoption of protons,
there remains a lack of long-​term outcomes for patients treated with protons, and no
prospect of randomized data emerging in the future. Some single institution studies
with up to 7 years median follow-​up have published encouraging outcomes for chil-
dren with common central nervous system (CNS) tumours, such as medulloblastoma
and low grade glioma, but longer-​term data is still needed(4,5). This is particularly true
with regards to the toxicities of protons, as there have been some concerns regarding a
possible increase in radiation necrosis observed in children with brain tumours treated
with proton therapy(6). Some acute toxicities, such as severe radiation dermatitis and
alopecia, are more common in patients treated with proton therapy, as there is a lack
of the skin sparing effect seen with high energy photons. However, overall the current
472 Principles of paediatric radiation oncology

Table 21.2  NHS England Proton Therapy indications for children and young people(9)
NHS England Proton Beam Radiotherapy (High Eenergy) Commisioning Criteria for
Paediatric (<16y) and Teenage & Young Adult (21-​24y) Cancer Treatment
1. General Criteria

1.1 A
  clear indication for radiotherapy and defined as curable and with cancer survival
expectation of 40% 5 year survival and no comorbidities likely to limit life expectancy
to <5 years.
1.2 There should be no evidence of distant metastasis with the exception of:
Rhabdomyosarcoma and Ewing’s Tumours where limited and only lung disease
2. Specific Diagnostic Criteria
2.1 Base of Skull & Spinal Chordoma
2.2 Base of Skull Chondrosarcoma
2.3 ‘Adult type’ Bone and Soft Tissue Sarcomas (excluding extremities)
2.4 Rhabdomyosarcoma (excluding extremitites)
2.5 Ependymoma Ewing’s Sarcoma (excluding extremities)
2.6 Ewing’s Sarcoma (excluding extremities)
2.7 Retinoblastoma
2.8 Pelvic Sarcoma
2.9 Optic Pathway and other selected Low Grade Glioma
2.10 Craniopharyngioma
2.11 Pineal Parenchymal Tumours (not Pineoblastoma)
2.12 Non-​metastatic intracranial germ cell tumours
2.13 Pituitary Adenoma
2.14 Juvenile Angiofibroma
2.15 Meningioma (Excluding Grade 3)
2.16 Nasopharyngeal Carcinoma
2.17 Esthesioneuroblastoma
2.18 Salivary Gland Tumours

Source: NHS, https://​www.england.nhs.uk/​commissioning/​spec-​services/​npc-​crg/​group-​b/​b01/​

published literature supports proton therapy as likely reducing the levels of significant
late toxicities (e.g. hearing loss, neurocognitive effects) and producing better quality
of life for survivors(4,7).
Proton therapy can also be used to achieve dose escalation, with the aim of improving
tumour control compared to high energy photon radiotherapy, and this has now be-
come the standard of care for base of skull and spinal chordoma and chondrosarcoma,
in both children and adults. However, these are rare in children and account for a very
 Intensity-modulated radiotherapy 473

Fig. 21.1  Craniopharyngioma photon (IMRT, top) and proton (bottom) comparison plans
in 3 planes.
Reproduced with permission from Indelicato, Daniel J. et al. ‘Consensus Report From the Stockholm
Pediatric Proton Therapy Conference,’ International Journal of Radiation Oncology, Volume 96, Issue 2,
pp. 387–392 (DOI: https://​doi.org/​10.1016/​j.ijrobp.2016.06.2446) Copyright © 2016 Elsevier Ltd

small proportion of cases. For base of skull chordoma it is possible to deliver doses
of 72–​74 Gy (RBE), whilst limiting dose to adjacent critical OAR, and this has been
shown to improve local control, although with grade 3 + toxicities in approximately
8% of patients(8).

21.3  Intensity-​modulated radiotherapy
Rotational IMRT techniques, including VMAT, RapidArc®, and TomoTherapy®, are
now frequently used for the treatment of paediatric malignancies, to harness the im-
proved conformity of the high doses of radiotherapy to the target. In addition to re-
ducing the acute and late toxicities of radiotherapy, they can have further benefits for
children, such as shorter duration of treatment with VMAT. The main concerns, as
stated previously, relate to the ‘low-​dose bath’ effect and the risk of second malignan-
cies, but the true magnitude of this effect will not be determined for many years.
Rotational IMRT is routinely used for paediatric brain tumours, where proton
therapy is not available, such as the boost to the posterior fossa tumour bed for
medulloblastoma looking to reduce the dose to the adjacent cochleas. The benefits
of these techniques for abdominal neuroblastoma are being investigated in the open
randomized phase 2 IMAT study in the UK, which is looking at the feasibility of dose
escalation up to a dose of 36 Gy with rotational IMRT for these tumours.
474 Principles of paediatric radiation oncology

21.4  Radiotherapy quality assurance

Radiotherapy quality assurance (RTQA) is an essential component of modern
paediatric radiotherapy, particularly with the large proportion of surviving patients
to experience the late effects of treatment. Treatment should ideally be delivered
within clinical trials, or according to current national CCLG treatment guidelines
(http://​www.cclg.org.uk). All radiotherapy departments treating children should
also ideally have an externally validated and audited quality management system
(e.g. ISO 9001).

21.4.1  Quality assurance in multicentre studies

The accuracy of delivery of radiotherapy is known to be an important factor contrib-
uting to improved tumour control, particularly for complex and highly conformal
techniques including proton therapy, craniospinal irradiation, and stereotactic radio-
therapy(10). In recent years international standards for RTQA have been established by
the Global Harmonisation of Clinical Trials RTQA Group, covering the entire radio-
therapy process from target localization to planning, and finally to treatment delivery
and verification.
Clinical trial RTQA has been successfully implemented in the UK for adult studies,
through the UK RTTQA and European EORTC groups, whereas paediatric studies
have so far usually not had uniform trial specific prospective RTQA in Europe. By
comparison, in the USA, the COG has utilized the centralized RTOG Imaging and
Radiation Oncology Core (IROC) facility for its paediatric multicentre studies, en-
suring high quality trial RTQA. In May 2017 the SIOPe ‘Quality and Excellence in
Radiotherapy and Imaging for Children and Adolescents with Cancer across Europe
in Clinical Trials (QUARTET)’ initiative was formally launched looking to address this
deficiency, with initial plans to provide a prospective RTQA platform for upcoming
European studies in rhabdomyosarcoma, neuroblastoma, and some paediatric brain
tumours.

21.5  Toxicity of radiotherapy for children

21.5.1 Acute morbidity
The acute side effects seen relate to the site being treated, dose, technique, and use of
concomitant chemotherapies. They include skin erythema, mucositis, nausea, diar-
rhoea, hair loss, and myelosuppression, occurring in children as in adults and being
managed similarly.

21.5.2 Subacute effects
Liver
A large proportion of the liver can require to be irradiated during radiotherapy treat-
ment for paediatric upper abdominal malignancies. Veno-​occlusive disease (radiation
hepatopathy) is a very rare occurrence, observed at 2–​4  months following radio-
therapy, and evidenced by hepatomegaly, jaundice, ascites, thrombocytopenia, and
 Toxicity of radiotherapy for children 475

elevated transaminases. Long-​term dysfunction is also rare with this risk being dose
related. A risk factor is the administration of actinomycin-​D with hepatic irradiation
therefore this is routinely omitted or dose reduced during radiotherapy.

Lung
For a number of paediatric malignancies whole lung radiotherapy is utilized for the
treatment of pulmonary metastases, including Wilms’ tumour, rhabdomyosarcoma,
and Ewing sarcoma. The risk of interstitial pneumonitis and pulmonary fibrosis are
dose and radiation volume related. It is essential to consider potential interactions
between chemotherapeutic drugs and lung irradiation, in particular with busulfan or
bleomycin. Mild radiation pneumonitis consists of a dry cough and mild dyspnoea,
whereas more severe symptomatic radiation pneumonitis is rare. Historically pneu-
monitis was the dose limiting toxicity for total body irradiation, with severe symptoms
requiring intervention reported in up to 5% of cases.

Central nervous system
After 1–​2 weeks following the initiation of radiotherapy for brain tumours, children
may experience a transient deterioration of neurological symptoms. Post irradiation
somnolence, with varying degrees of fatigue and irritability, frequently occurs in
children 4–​6 weeks after cranial irradiation, and is thought to be related to tran-
sient demyelination. LHermitte’s sign, consisting of electric shock-​like sensations
radiating down the spine into the limbs, can occur following radiation to the upper
spinal cord.

21.5.3  Long-​term effects
Bone growth
Impairment of bone growth and associated soft tissue hypoplasia can be one of the
most obvious and distressing long-​term effects. Abnormalities and asymmetry of
craniofacial growth can result in significant cosmetic and functional deformities,
such as micrognathia leading to problems with dentition. The epiphyseal growth
plates are very sensitive to radiation, and should be excluded from the radiotherapy
field whenever possible. Age at time of treatment, radiation dose, and volume are
factors which have an impact on the severity of these orthopaedic long-​term ef-
fects. There is evidence of a dose response effect, with more growth observed in
average risk medulloblastoma survivors, who received 23.4 Gy craniospinal ir-
radiation compared to high risk patients who received 36 Gy(11). Slipped femoral
epiphysis and avascular necrosis have also been reported following irradiation of
the hip. Laboratory evidence suggests a dose response effect between doses as low as
5–​10 Gy up to 35–​40 Gy, and an effect of dose per fraction. Careful consideration of
the late orthopaedic effects of radiation is extremely important whenever planning
radiotherapy for children. When irradiating the spine, the full width of the vertebra
should receive homogeneous irradiation in order to minimize long-​term kyphosis
or scoliosis.
476 Principles of paediatric radiation oncology

Central nervous system
The largest cohort of children and young people requiring radiotherapy are those
with CNS tumours, and careful consideration needs to be given to the toxicity of
therapy.

Radionecrosis
Radionecrosis results from a direct effect on glial tissue and generally occurs with
a latency of 6 months to 2 years. It is rarely seen at doses < 60 Gy, and for the rad-
ical treatment of children with brain tumours it is very uncommon to exceed doses
of 50–​59.4 Gy. The clinical effects of radionecrosis vary according to the site within
the CNS, and are most devastating in the brainstem and spinal cord. In the current
era, with the increasing vogue for using stereotactic ablative radiotherapy treatments
or fractionated reirradiation treatments, and as a result of better magnetic resonance
imaging (MRI), radionecrosis is being identified more frequently. Radionecrosis of
the spinal cord in children has previously been reported as a consequence of the inter-
action between radiation and cytosine arabinoside given intrathecally for metastatic
rhabdomyosarcoma.

Necrotizing leucoencephalopathy
This has been observed when cranial irradiation was followed by high-​dose metho-
trexate for the treatment of leukaemia. The clinical features include ataxia, lethargy,
epilepsy, spasticity, and paresis.

Neuropsychological effects
The cognitive effects resulting from cranial radiotherapy are well documented, with
the dose, volume of irradiated brain, younger age, and the number of neurosurgical
interventions all important factors. When compared with siblings, children receiving
24 Gy prophylactic cranial irradiation demonstrate an approximate fall in IQ of 12
points. Following higher radiation doses given for brain tumours an increased risk of
learning and behaviour difficulties is observed(12).

Kidney
Long-​term effects on renal function are rarely seen before 2–​3 years after radio-
therapy. The risk increases following a dose of > 15 Gy to both kidneys. The se-
verity is related to the irradiated volume of the kidneys and the dose. When mild
it may be asymptomatic or cause hypertension. Provided renal dose constraints
are adhered to then the more severe problems, including renal failure, are very
unlikely.

Endocrine
Endocrine deficiencies following radiotherapy are common, with the risk and timing
related to dose. Growth hormone and other pituitary hormone deficiencies should
be screened for from 1 year following irradiation to the hypothalamic-​pituitary axis.
 Leukaemia 477

After 25–​40 Gy of radiotherapy to the thyroid the incidence of elevated thyroid stimu-
lating hormone (TSH) is 75%.

Reproductive
In boys the germinal epithelium is very sensitive to the effects of low-​dose irradiation.
In adult males, transient oligospermia is seen after 2 Gy, but slow recovery can occur
after doses up to 5 Gy. In girls the oocytes are also sensitive; subsequent pregnancy is
rare after 12 Gy total body irradiation, but has been reported.

21.6  Chemotherapy and radiotherapy interactions

Interactions between radiation and chemotherapy are complex and poorly under-
stood, although can potentially be exploited in order to attempt to improve outcomes.
The ‘spatial cooperation’ of combined chemotherapy and radiotherapy is the most fre-
quently employed mechanism in paediatric oncology, such as the use of radiation for
local control of the primary tumour, and with chemotherapy for subclinical metastatic
disease for the treatment of rhabdomyosarcoma or Ewing sarcoma.
Chemotherapy and radiotherapy may be combined with the aim of increasing tumour
cell kill without excess toxicity, i.e. improving the therapeutic ratio. An example is the use
of combined chemotherapy and radiotherapy for children with Hodgkin’s lymphoma,
where it has been possible to reduce the intensity of both treatment modalities and
therefore hopefully reducing long-​term morbidity. Many protocols for children involve
the use of concurrent chemotherapy and radiotherapy, including rhabdomyosarcoma
and high grade glioma, and potential toxicities should be considered. Actinomycin D
and cisplatin increase the slope of the radiation dose–​response curve, and actinomycin
D inhibits the repair of sub-​lethal damage. Such clinical interactions include enhanced
skin and mucosal toxicity when radiation is followed by actinomycin-​D (the ‘recall phe-
nomenon’), enhanced bladder toxicity when chemotherapy is combined with cyclo-
phosphamide, enhanced CNS toxicity from combined radiation and methotrexate,
cytosine arabinoside or busulfan, and the enhanced marrow toxicity from spinal or
pelvic irradiation when combined with myelotoxic chemotherapeutic agents. In the
case of the effect of combined radiation and anthracyclines such as doxorubicin on
the heart, doxorubicin has its effects on the myocytes and radiation on the vasculature.

21.7 Leukaemia
The leukaemias account for the largest group (Table 21.1) of paediatric malignancies
with approximately 80% having acute lymphoblastic leukaemia (ALL). The other leu-
kaemia that commonly occurs in paediatrics is acute myeloid leukaemia (AML), with
chronic myeloid leukaemia (CML) and other types more rarely seen. The improve-
ment in survival of children with ALL was one of the early successes of paediatric
oncology, and outcomes continue to improve with the recent UKALL 2003 reporting
5-​year overall survival rates of in excess of 90%(13,14). Patients are categorized into
478 Principles of paediatric radiation oncology

clinical standard, intermediate, and high-​risk groups on the basis of a combination of

National Cancer Institute (NCI) criteria, cytogenetics, and early response to induction
therapy, assessed by bone marrow blast counts taken at days 8 (NCI high-​risk patients)
and 15 (NCI standard-​risk patients) after induction begins.
Further stratification for intensity of therapy is based on analysis of postinduction
bone marrow for minimal residual disease. The four phases of treatment are:

1. Risk-​
adapted multiagent remission induction, based on drug combinations
including dexamethasone, vincristine, asparaginase +/​-​daunorubicin.
2. Intensification with multidrug combinations.
3. CNS prophylaxis, generally with intrathecal methotrexate.
4. Maintenance usually based on a continuous low-​dose antimetabolite drug such as
mercaptopurine, with a total duration of therapy of approximately 2 years.
For AML the basis of therapy is intensive multidrug chemotherapy. As for adults,
haemopoietic stem cell transplantation (HSCT) is frequently employed for children
who have an HLA-​matched sibling; survival rates of 65–​70% have been reported by
international collaborative groups. Allogeneic HSCT is not used for children with fa-
vourable risk factors, and the benefit for intermediate and high risk patients in first
complete remission (CR1) needs to be balanced against toxicity(15). The current inter-
national MYECHILD study is investigating intensification of induction chemotherapy,
fludarabine for consolidation, and reduced intensity conditioning for allogeneic HSCT
in CR1.

21.7.1  Total body irradiation

As in the treatment of adults with haematological malignancies, total body irradiation
(TBI) is an important technique used together usually with high-​dose cyclophospha-
mide (Cyclo-​TBI), or other chemotherapy combinations, as the main conditioning
regimen to facilitate HSCT. Bone marrow donors include HLA-​matched siblings, vol-
unteer matched unrelated donors (available through international donor registries),
donated umbilical cords, and parental (haplo) donors; these options have resulted in
increasing numbers of children for whom HSCT can be considered.
Individual techniques for TBI have evolved in different departments, often
depending on availability of treatment machines. TBI dosimetry has been traditionally
based on in vivo measurements, although increasingly more conformal techniques are
being adopted. For such a large and complex target volume, it is not feasible to adhere
to the ICRU 50 guidelines of a range of -​5% to + 7%, and a range of -​10% to + 10% is
more commonly used. Modern linear accelerator and bunker design can be utilized to
achieve large field sizes at an extended focus-​to-​skin distance; alternatively fixed field
or rotational IMRT techniques are used.

21.7.2  Indications for total body irradiation in children

TBI conditioning to facilitate HSCT is routinely used for ALL in the setting of re-
lapse and for those presenting with higher-​risk features. Other conditions con-
sidered for HSCT where TBI is sometimes employed include severe aplastic anaemia,
 Leukaemia 479

thalassaemia, and immunodeficiency syndromes. Long-​term effects of TBI include

impaired growth due to growth hormone deficiency, and a direct effect from irradi-
ation of epiphyses. There is also a possibility of cataract, hypothyroidism, and in some
studies the possibility of renal impairment. TBI is generally not considered for infants
and children with AML, where a conditioning regimen with two drugs busulfan and
cyclophosphamide (Bu-​Cy) is more commonly used.

21.7.3 Technique
The TBI, cranial, and testicular irradiation techniques described are those in use at
the Royal Marsden Hospital. TBI is delivered using forward planned IMRT, with large
lateral diamond fields and incorporating multiple multi-​leaf collimator defined fields
individualized from a plan template. The distance from the origin of the radiation to
the patient’s midline is 355 cm. The dose rate is set at approximately 14–​19 cGy/​min
using 10 MV photons.

Clinical target volume
Whole body.

Patient position and immobilization

Supine position with arms folded across chest. A  customized whole body vacuum
immobilization bag is used. Additional smaller vacuum immobilization bags may be
used to immobilize the head and foam pads/​vacuum bags may be used to position the
knees as required.

Dosimetry
Dose specified at 100%. The predicted in-​vivo dosimetry is checked via thermolumin-
escent dosimetry , ideally the planned dose and measured dose should agree to within
12% at all sites.

Dose prescription
Full intensity HSCT: Treated twice daily; minimum interfraction interval of 6 hours.
◆ Dose per fraction 1.65 Gy to 2 Gy.
◆ Standard fractionation 14.4 Gy in 8 fractions over 4 days
◆ Alternative fractionations: 13.2 Gy in 8 fractions (double cord); 12 Gy in 6 fractions.
Reduced intensity HSCT:  Single fraction, usually 2 Gy (doses up to 4 Gy can
be used).

21.7.4 Cranial radiotherapy
Previously prophylactic whole brain radiotherapy and intrathecal methotrexate
were used routinely to reduce the risk of CNS relapse to < 10%; however, concerns
over neurotoxicity and secondary brain tumours have led all of the major treatment
consortia to limit the use of cranial irradiation. Whole brain radiotherapy is now usu-
ally reserved for patients who present or relapse with confirmed CNS involvement,
either as a boost prior to TBI, or as a component of primary treatment.
480 Principles of paediatric radiation oncology

Technique
Lateral opposed megavoltage fields, generally 4–​6 MV. Conformal technique with
shielding of face, dentition, nasal structures, and lenses.

Clinical target volume
Intracranial meninges, with care taken to include the cribriform plate, temporal lobes,
optic nerves, and base of skull.

Patient position and immobilization

Supine, immobilized in a thermoplastic head mask.

Localization of clinical target volume and planning target volume

CT virtual simulation is current standard of care, as shown in Fig. 21.2. Typical in-
ferior field border at the inferior aspect of the C2 cervical vertebra, ensuring adequate
coverage of the intracranial meninges; this also limits potential future growth issues
arising with the atlanto-​axial pivot joint of the C1 and C2 vertebrae.

Dosimetry
Dose specified at the midplane (MPD), or at 100% where isodose intervention
undertaken.

Dose prescription
24 Gy in 15 fractions; dose per fraction 1.6 Gy.
Boost prior to TBI conditioning for full intensity allogeneic transplant: 6 Gy in 4
fractions; dose per fraction of 1.5 Gy.

Fig. 21.2  Typical field and axial virtual simulation isodoses for cranial irradiation, or
cranial boost in conjunction with total body irradiation.
 Leukaemia 481

21.7.5 Testicular irradiation
Boys who presenting, or relapsing, with testicular involvement are treated with tes-
ticular radiotherapy, either as a boost prior to TBI or as primary treatment.

Technique
Usually anterior electron field (orthovoltage 200–​300 kV used in some centres).
Shielding of non-​target skin and perineum (see Fig. 21.3).

Clinical target volume
Both testes, scrotum, and inguinal canal supero-​laterally as far as the deep inguinal ring.

Patient position and immobilization

Supine.

Localization of clinical and planning target volumes

Clinical localization, as per Fig. 21.2. A margin of 1.5 cm added to field from CTV for
electrons.

Dosimetry
Dose specified as per individual departmental guidelines for electrons or orthovoltage
techniques, e.g. electrons of appropriate energy to encompass the PTV by the 90%
isodose.

Pb Tape
Anterior Superior
lliac Spine
Deep Inguinal Ring
Clinical Target
Volume
Field Edge
Pb

Fig. 21.3  Typical electron field for testicular irradiation.

482 Principles of paediatric radiation oncology

Dose prescription
Definitive treatment: 24 Gy in 12 fractions of 2.0 Gy daily.
Boost prior to TBI conditioning for full intensity allogeneic transplant: 4 Gy single
fraction.

21.8 Hodgkin lymphoma
Hodgkin lymphoma has its peak age incidence in the adolescent age range. The
survival rate for children with Hodgkin lymphoma is over 90%. In the recent and fu-
ture Inter-​Group Euronet pHL study protocols the aims are to maintain these good
overall survival rates whilst reducing late effects. Important late effects including
impaired bone growth, and also infertility from alkylating agents and procarbazine.
These European protocols employ a chemotherapy-​based approach and utilize the
international Deauville consensus criteria for interim fludeoxyglucose (FDG)-​PET
CT response with the aim of reducing the proportion of patients receiving consoli-
dation involved site radiotherapy (ISRT). In the standard arm of the future pHL C2
study only those with an inadequate FDG response (Deauville Score 4 or 5), on the
Early Response Assessment PET after two cycles of OEPA chemotherapy (vincristine,
etoposide, prednisolone, doxorubicin), will receive ISRT(16).

21.8.1 Technique
IMRT, forward-​planned with anterior and posterior parallel-​opposed fields, or
inverse-​planned using either fixed-​field or rotational IMRT may be used, particu-
larly when there is a large residual mediastinal mass in order to meet lung dose
constraints.

Clinical target volume
◆ CTV1: Involved nodes at diagnosis with a margin of 0.5 cm in all directions.
◆ CTV2: Residual FDG-​avid nodes on late response assessment FDG-​PET imaging
after completion of chemotherapy with 0.5 cm margin.

Patient position and immobilization

Supine with head and neck shell immobilization where necessary.

Localization of clinical target volume/​planning target volume

Contrast-​enhanced CT simulation, incorporating information from FDG-​PET scans.

Dosimetry
Dose specified at 100%, or median dose for inverse planned IMRT (see Fig 21.4).

Dose prescription
◆ PTV 1: 19.8 Gy in 11 fractions of 1.8 Gy.
◆ PTV 2: 10 Gy in 5 fractions of 2.0 Gy.
 Neuroblastoma 483

Fig. 21.4 Bilateral neck modified involved field radiotherapy for Hodgkin lymphoma,

utilizing single 6 MV VMAT arc and without midline sparing to ensure adequate vertebral
coverage.

21.9  Non-​Hodgkin lymphoma
Burkitt, T-​cell lymphoblastic, or anaplastic large cell lymphomas are the commonest
Non-​Hodgkin lymphoma (NHL) in children, whereas diffuse large B-​cell and follicular
lymphomas are rarely seen. Survival rates have improved in recent years, with cur-
rently over 80% long-​term survivors. Therapy is based on intensive multiagent chemo-
therapy including CNS prophylaxis with intrathecal chemotherapy, and there is no
routine role for radiotherapy. However, children with T-​cell lymphoblastic lymphoma,
which is managed according to the same principles as ALL, may be considered for full
intensity allogeneic bone marrow transplantation BMT with TBI conditioning.

21.10 Neuroblastoma
Neuroblastoma (NBL) is the commonest extracranial solid tumour of childhood, and
generally a disease of infants with approximately one-​third presenting aged < 1 year.
NBL arise from the neural crest tissue in the sympathetic nervous system, usually in
the adrenal gland but can arise anywhere from the neck to the pelvis. Children fre-
quently present with Stage 4 disease; common sites of metastases include bones, bone
marrow, lymph nodes, and liver. Despite advances in treatments the current survival
rates remain poor, approximately 45% taking all stages and prognostic groups into
account.
In 2009 the International Neuroblastoma Risk Group (INRG) established a new
international consensus staging system (INRGSS) based on imaging defined risk fac-
tors at diagnosis, and a new classification based on key prognostic factors: stage, age,
histologic category, grade of tumour differentiation, the status of the MYCN onco-
gene, chromosome 11q status, and DNA ploidy(17,18). Management is now stratified
according to risk grouping. Patients in the best risk group with a survival rate of
> 90% can be managed with surgery alone. By comparison patients in the European
SIOPEN High Risk Neuroblastoma 1 study are currently treated with multimodality
therapy:  dose dense induction chemotherapy, using drugs such as vincristine, cis-
platin, carboplatin, etoposide, and cyclophosphamide (e.g. Rapid COJEC), surgical re-
section of the primary tumour (aiming for ≥ 90% resection), high dose chemotherapy,
radiotherapy to the primary site, and finally differentiation therapy (13 cis-​retinoic
acid) and immunotherapy (anti-​GD2 antibody +/– interleukin-​2).
484 Principles of paediatric radiation oncology

Success for this current approach relies on chemotherapy effectively treating the
metastatic disease; the importance of local tumour control is based on series from
North America, which pointed to improvement in local control with the use of tumour
bed radiotherapy in patients undergoing resection of primary tumour. The current
dose for postoperative radiotherapy to the tumour bed and residual tumour is 21 Gy
in 14 daily fractions, although dose escalation up to 36 Gy utilizing rotational IMRT is
being investigated currently in trials including the IMAT study in the UK. Given that
the majority of these tumours arise in the suprarenal region the main OAR limiting
dose to the target are the kidneys and the liver dose also has to be considered. At the
time of radiotherapy the function of these organs may have already been compromised
by high-​dose chemotherapy or surgery.

21.10.1  mIBG therapy for neuroblastoma

The majority of NBL take up the guanethidine analogue meta-​iodobenzyl guanidine
(mIBG). 123I-​mIBG is the radionuclide used for diagnostic imaging and which is now
incorporated in both standard staging and response assessment. 131I-​mIBG can be
used as a form of molecular radiotherapy, and is an important therapeutic option for
patients with relapsed or refractory disease. The challenges of this approach relate
particularly to the logistics of radiation protection for very young children, and so it
is usually delivered by specialized multidisciplinary teams in centres with paediatric-​
specific radioisotope facilities. To date there has been a lack of randomized trials
comparing the efficacy of mIBG with systemic chemotherapies; however, SIOPEN
have proposed the VERITAS study which will compare tandem high-​dose therapies,
randomizing refractory patients to either 131I-​mIBG and topotecan or alternatively
thiotepa, in addition to the standard high-​dose chemotherapy of busulphan and
melphalan.

21.11 Wilms’ tumour
Wilms’ tumour (nephroblastoma) is an embryonic renal tumour with a median age at
diagnosis of between 3–​3.5 years. Patients are staged according to histopathological
findings following nephrectomy. Table 21.3 shows the National Wilms’ Tumour Study
Group (NWTS) staging system for Wilms’ tumour.
In 4–​8% of cases, Wilms’ tumour is bilateral (Stage V). Wilms’ tumour may be gen-
etically associated with aniridia (congenital absence of the iris) and other inherited
syndromes such as the Beckwith–​Wiedemann syndrome (variable features including
macrosomia or hemihypertrophy, macroglossia, omphalocoel). The WT1 gene is lo-
cated on chromosome 11, and is a tumour suppressor gene. If both copies of the gene
are lost by mutation, then Wilms’ tumour may arise.
The long-​term survival rate for Wilms’ tumour is in excess of 80%; the current SIOP
Renal Tumour Study Group (RTSG) Umbrella Protocol is looking to harmonize the
diagnostic procedures, undertake imaging studies, and prospectively collect biological
material, with the aims of improving the current stratification of patients and providing
novel molecular targets. There is a longstanding disparity between approaches adopted
in Europe compared with North America. In the SIOP RTSG Umbrella study even
 Wilms’ tumour 485

Table 21.3  National Wilms Tumour Study Group (NWTS)

Staging System
Stage Clinicopathological features
I Confined to within renal capsule, completely excised

II Invading outside renal capsule, completely excised

III Residual abdominal disease—​positive margins,
tumour rupture, involved nodes
IV Haematogenous metastases
V Bilateral disease

Adapted from Chapter 1, Wilms’ Tumour–Histology and Differential

Diagnosis, National Wilms Tumor Study Group Staging System.
© The Authors. Reproduced under the terms of the Creative
Commons Attribution 4.0 International license (CC BY 4.0).
https://creativecommons.org/licenses/by/4.0/.

initial biopsy is now not recommended for renal tumours where the imaging appear-
ances are consistent with Wilms’ tumour; instead nephrectomy is usually performed
after 6 weeks of neoadjuvant vincristine and actinomycin-​based chemotherapy. By
comparison in the COG studies a policy of immediate nephrectomy is recommended.
Postoperative chemotherapy is given using vincristine, actinomycin D, and doxo-
rubicin, the number of drugs and duration depending upon the staging. Postoperative
flank radiotherapy is employed for local stage III (incompletely resected primary tu-
mours, pre-​or perioperative tumour rupture, or histologically involved lymph nodes),
or high-​risk stage II disease.

21.11.1 Technique
3D conformal planning, commonly using anterior and posterior opposed fields. IMRT
may be considered.

Clinical target volume
Preoperative extent of tumour and kidney, following preoperative chemotherapy, with
a margin of 0.5 to 1.0 cm. Expansion to PTV is based on departmental specific data
but usually is 1 cm. Vertebrae adjacent to PTV need to be homogeneously irradiated
to minimize the long term risk of kyphosis or scoliosis (Fig. 21.5).

Patient position and immobilization

Supine.

Localization of clinical target volume/​planning target volume

CT simulation utilizing preoperative CT/​MRI, operative note, and position of clips
inserted at surgery marking the extent of resection. Intravenous contrast may be used
to aid delineation of adjacent para-​aortic nodal regions.

Dosimetry
Dose specified at 100%, or median dose for IMRT.
486 Principles of paediatric radiation oncology

(a)

(b)

Fig. 21.5  Postoperative Flank (a) and Whole abdomen conformal radiotherapy (b) for

Wilms’ tumour.

Dose prescription
◆ Intermediate risk: 14.4 Gy in eight fractions; additional boost 10.8 Gy in six frac-
tions if residual macroscopic disease, both at a dose per fraction of 1.8 Gy daily.
◆ High risk: 25.2 Gy in 14 fractions of 1.8 Gy.

21.11.2  Whole lung radiotherapy for Wilms tumour

Whole lung radiotherapy is employed for patients with high-​risk histology and lung
metastases or low/​intermediate-​risk histology with residual pulmonary metastases
following induction chemotherapy.
 Rhabdomyosarcoma 487

Technique
3D conformal planning, commonly using anterior and posterior opposed fields; alter-
natively IMRT can be considered with the aim of reducing the dose to the heart.

Clinical target volume
Whole lungs, from lung apices to the base of both lungs. PTV expansion based on de-
partmental data, usually 1 cm.

Patient position and immobilization

Supine.

Localization of clinical target volume/​planning target volume

CT simulation.

Dosimetry
Dose specified at 100%.

Dose prescription
12 Gy in eight fractions of 1.5 Gy.

21.11.3  Whole abdominal radiotherapy for Wilms tumour

Whole abdominal radiotherapy (Fig. 21.5) has considerable acute and long-​term mor-
bidity and should be reserved for those who present with extensive intra-​abdominal
tumour spread, or gross preoperative or perioperative tumour rupture. For inter-
mediate risk disease a dose of 15 Gy can be given, but for high-​risk disease 21 Gy is
recommended; both using a dose per fraction of 1.5 Gy, and limiting the dose to the
remaining kidney to 12 Gy. To minimize long-​term morbidity the femoral heads are
shielded.

21.12 Rhabdomyosarcoma
Rhabdomyosarcoma (RMS) is the commonest paediatric malignant soft tissue tu-
mour, with up to 60 new cases in the UK every year. RMS may arise at any site, com-
monly occurring in the head and neck region, including the orbit and nasopharynx,
and in the genito-​urinary tract, such as bladder, prostate, perineum, and vagina. The
site of the primary tumour is known to be a significant prognostic factor, and those
defined as parameningeal (e.g. nasopharyngeal, infratemporal fossa), often associated
with base of skull invasion and intracranial extension, are particularly challenging to
eradicate. Favourable sites are currently defined as orbit, genito-​urinary (non-​bladder/​
prostate), or head & neck (non parameningeal), with all other sites designated as un-
favourable. The biology of the tumour has important prognostic significance, with the
presence of one of the PAX3/​7-​FOX01 fusion genes conferring worse outcome; previ-
ously the histological appearance, defining RMS as either embryonal (favourable) or
alveolar (unfavourable; 80% of which are PAX3/​7-​FOX01 positive), was used(19). Other
488 Principles of paediatric radiation oncology

important prognostic factors include initial IRS post-​surgical stage (I, II, or III), age
over 10 years and tumour size greater than or equal to 5 cm.
The treatment strategy employed for the majority RMS in Europe is defined in
the recently closed European Paediatric Soft tissue Sarcoma Study Group (EpSSG)
RMS2005 and MTS2008 studies. These have incorporated a stratification of treat-
ment based on the overall risk grouping for each patient:  Low, Standard, High,
Very High or Metastatic. Low-​risk disease is treated with eight cycles of vincris-
tine and actinomycin D (VA), all other RMS receive nine cycles of chemotherapy;
Standard risk (Subgroup B and C) are treated with initial ifosfamide, vincristine,
and actinomycin D (IVA) then VA; high risk (and Standard risk subgroup D) re-
ceive IVA alone. The two randomizations in RMS2005 were both undertaken in the
High Risk RMS grouping, which contains the largest number of patients. Whilst
an improvement was seen in 3-​year event-​free survival, up to 65% compared to a
predicted 50–​55% from prior studies, there was no benefit from the addition of
doxorubicin (IVADo/​IVA) to standard nine cycles of IVA(20). Four cycles of IVADo,
then five cycles IVA, (IVADo/​IVA) remains the current standard for very high-​risk
(node positive alveolar RMS) and metastatic patients. The RMS2005 study was also
investigating in High Risk RMS whether the addition of maintenance vinorelbine
and cyclophosphamide chemotherapy for 6 months improves outcome; this result
is still pending.
Local tumour control is of paramount importance in RMS, with local failure ob-
served in 85% of relapsed cases, and when relapse occurs following radiotherapy
very few will go on to achieve cure(21). The role of radiotherapy in the various risk
categories from the RMS2005 study, together with protocol doses are summarized
in Table 21.4. This study included risk-​adapted radiotherapy related to histology, tu-
mour site and size, extent of resection, and nodal status. Proposed concepts for the
future EpSSG Frontline and Relapsed RMS (FaR-​RMS) study include preoperative
radiotherapy, radiotherapy dose escalation for patients with a high local failure risk,
and the omission of radiotherapy to metastatic sites for those with extensive meta-
static disease.
In addition to maximizing local control and the chance of cure, the sequelae fol-
lowing radiotherapy need to be taken into consideration, particularly for infants,
and for head and neck primaries where 63% of long-​term survivors report one or
more severe or disabling events(22). Highly conformal techniques should be used
for RMS to mitigate the potential late effects, and children with localized disease
are routinely referred for proton therapy. Brachytherapy may be considered for
highly selected children with limited tumours arising in genito-​urinary (vagina,
uterus, bladder/​prostate, and perineum) or head and neck sites. The majority of
brachytherapy treatments for RMS are now undertaken following complete or par-
tial tumour resection, utilizing modern image guidance and afterloading pulsed
(PDR) or high dose rate (HDR) systems. A  number of published single centre
series have reported encouraging levels of late effects and quality of life in sur-
vivors treated with brachytherapy(22–​24). Given the rarity of suitable tumours and
the complexity of the delivery of these highly specialized treatments in children,
brachytherapy for RMS should only be done at specialist national or international
referral centres.
 Ewing sarcoma 489

Table 21.4  Dose fractionation regimens for rhabdomyosarcoma in EpSSG RMS

2005 Study
IRS* Group Histology
Embryonal RMS Alveolar RMS
I—​size < 5 cm, completely No radiotherapy 41.4 Gy
resected with negative margins

II—​macroscopic resection but 41.4 Gy 41.4 Gy

positive margins
III—​macroscopic residual disease
at start of chemotherapy.
Followed by:
Secondary complete resection 36.0—​41.4 Gy depending on 41.4 Gy
(after chemotherapy) response
Incomplete secondary resection 50.4 Gy 50.4 Gy
(after chemotherapy)
Clinical/​radiological complete 41.4 Gy 50.4 Gy
response, no second surgery
Partial response to 50.4 Gy (with boost of 5.4 Gy for 50.4 Gy (with boost
chemotherapy selected patients) Orbit primary—​ of 5.4 Gy for selected
and > 2/​3 partial response—​45 Gy patients)
Nodal involvement—​dose to 41.4 Gy 41.4 Gy
nodal regions

IRS—​Intergroup Rhabdomyosarcoma Study Group

*

Daily fraction size: 1.8 Gy

Adapted with permission from ‘EpSSG, RMS 2005: a protocol for non metastatic rhabdomyosarcoma.’
Version 1.3 International, May 2012. Copyright © 2012 EpSSG. https://www.skion.nl/workspace/uploads/
Protocol-EpSSG-RMS-2005-1-3-May-2012_1.pdf.

The clinical target volume (CTV1) for RMS is defined as the extent of tumour at diag-
nosis (GTV) plus a 1-​cm margin, except at pushing boundaries into body cavities (e.g.
thorax) where CTV is trimmed back. For extremity tumours a 2-​cm CTV margin is used
superiorly and inferiorly and 1-​cm circumferentially. GTV2 is defined as the residual tu-
mour following induction chemotherapy, and a margin of 0.5–​1cm is added to produce
CTV2. All initially involved nodes and metastatic sites should be irradiated where feas-
ible, although this can be very challenging for patients with extensive metastatic disease.

21.13 Ewing sarcoma
The Ewing sarcoma (ES) family of tumours is a group of malignant sarcoma of bone and
soft tissues that have their peak incidence in adolescence. As with RMS, the management
of ES incorporates multimodality treatment, and for localized disease this has achieved
overall survival (OS) rates of 65–​75%; however, for those presenting with metastatic dis-
ease OS remains less than 30%(25). The diagnosis of ES has been refined since the dis-
covery of the underlying translocation which involves the EWS gene on chromosome 22
and an ETS-​type gene, producing the typical EWSR1 fusion oncogene. There is a group
490 Principles of paediatric radiation oncology

of Ewing-​like tumours that are treated similarly to ES, but involve different genetic fu-
sions, such as a non-​ETS gene with the EWSR1 gene. ES include the group of soft tissue
tumours formerly defined as peripheral primitive neuroectodermal tumours (pPNET)
and typified by Askin tumours of the chest wall. Approximately 60% of ES occur in the
long bones of the limbs, and 40% in the flat bones of the ribs, vertebrae, or spine; soft
tissue extension is commonly seen. Initial tumour size > 200 ml and histologic response
to chemotherapy (> 10% viable tumour) remain the most important prognostic factors.
Currently the majority of children, teenagers, and young adults with ES are treated
in the open Euro Ewing 2012 (EE2012) study, which is comparing induction VIDE
(vincristine, ifosfamide, doxorubicin, etoposide) chemotherapy with a compressed
VDC-​IE schedule used by the COG, which contains both ifosfamide and cyclophos-
phamide. It is also investigating the potential benefits of zoledronic acid for ES pa-
tients. Consolidation chemotherapy with either VA plus either cyclophosphamide or
ifosfamide (VAC or VAI) is given after VIDE induction chemotherapy, or VC-​IE con-
solidation for those receiving induction VDC-​IE, in conjunction with planned local
therapy. High-​dose chemotherapy with peripheral blood stem cell rescue was exam-
ined in the previous EE99 study but is no longer being used routinely for ES.
Following induction chemotherapy response assessment, there should be consider-
ation given to the optimal strategy to achieve local tumour control by surgical resec-
tion, radiotherapy, or a combined approach using both. For ES patients where surgical
resection is feasible then adjuvant (postoperative or preoperative) radiotherapy should
be considered for all patients, with the exception of those where all tissues involved
prior to the commencement of chemotherapy are excised, with clear margins (≥
1 mm), and with good histological response (> 90% necrosis) to induction chemo-
therapy. Adjuvant radiotherapy may occasionally be omitted, accepting a higher risk
of local failure, where the risks are felt to be too great, for example persistent wound
problems, concerns regarding infection of prosthesis, or in infants. The intention
of surgery should be to achieve complete resection where possible, as debulking, or
intralesional, surgery has not been shown to improve local control over radiotherapy
alone, and is therefore not recommended; amputation is rarely used for patients with
extremity tumours where there is a high risk of radiotherapy-​related significant long-​
term morbidity(25). Definitive radiotherapy is reserved for those patients deemed inop-
erable following multidisciplinary team review and discussion.
The planning of radiotherapy for ES is highly complex, and for the majority of pa-
tients with localized disease, with the exception of extremity tumours, proton therapy
is considered; otherwise IMRT is commonly used. Effective delivery for ES requires a
truly multidisciplinary approach involving radiologists, physicists, specialist therapy
radiographers, and mould room technicians from the outset.

21.13.1 Technique
Technique will depend upon tumour site and anatomy. Individualized, generally mul-
tiple fields in order to deliver homogeneous radiotherapy to the PTV and to minimize
dose to non-​target tissues and OARs.
 Ewing sarcoma 491

Clinical target volume
◆ GTV includes extent of tumour at diagnosis, with modifications to account for
‘pushing’ margins into body cavities and post-​surgical changes in anatomy for
operative cases.
◆ CTV1 includes GTV plus a margin of 1.5–​2  cm, including metallic prostheses,
drain sites, and surgical scars where feasible, and respecting anatomical boundaries
to spread, such as fascial planes and bones.
◆ CTV2 includes GTV plus a margin of 1–​2 cm (depending on anatomical location)
but not including scars or prostheses, and respecting anatomical boundaries to
spread.

Patient position and immobilization

Depending on site and anatomy. For limbs, a rigid immobilization device is recom-
mended, such as a thermoplastic shell; for lower limbs ideally the contralateral lower
limb should also be immobilized.

Localization of clinical target volume/​planning target volume

Planning CT scan with fusion of diagnostic +/​-​volumetric planning T1w contrast en-
hanced MRIs where feasible.

Dosimetry
Dose specified at 100%, or median dose for IMRT.

Dose prescription
◆ Preoperative radiotherapy:
• PTV1: 50.4 Gy in 28 fractions of 1.8 Gy.
• Can be reduced to 45 Gy in 25 fractions of 1.8 Gy where there are concerns about
future wound healing or adjacent OARs.
◆ Postoperative radiotherapy:
• PTV1: 45 Gy in 25 fractions of 1.8 Gy.
• PTV2: 9 Gy in 5 fractions of 1.8 Gy.
◆ Definitive radiotherapy:
• PTV1: 54 Gy in 30 fractions of 1.8 Gy.
• A further boost of 5.4 Gy in 3 fractions of 1.8 Gy can be considered.
◆ Whole lung radiotherapy:
• < 14 years: 15 Gy in 10 fractions of 1.5 Gy.
• ≥ 14 years: 18 Gy in 12 fractions of 1.5 Gy.
Care must be taken when combining chemotherapy and radiotherapy to avoid
excessive morbidity from enhanced radiation reactions. Actinomycin-​D and doxo-
rubicin are usually avoided during radiotherapy. Patients requiring radical radio-
therapy involving the spinal cord or significant radiotherapy to the lungs should not
receive busulfan.
492 Principles of paediatric radiation oncology

21.14 Osteosarcoma
Osteosarcoma is the most common malignant primary bone tumour in chil-
dren, teenagers, and young adults. It frequently arises in the metaphysis of long
bones:  43% in the distal femur, 23% in the proximal tibia, and 10% in the hu-
merus(26). Around 15–​20% of cases present with metastases, with the lung the most
common metastatic site. Prior to the advent of effective chemotherapy 80% of pa-
tients died with pulmonary metastases. Following the results of the recently pub-
lished EURAMOS 1 study, six cycles of methotrexate, doxorubicin, and cisplatin
(MAP) chemotherapy, given pre-​and postoperatively, remains the standard of care,
with the addition of ifosfamide and etoposide for poor responders, and interferon-​
α for good responders, demonstrating increased toxicity and a lack of superiority(27,
28)
. From this study the 3-​year event-​free survival for good responders was 76% and
for poor responders was 55%, but for metastatic disease was only 24%. Following
preoperative MAP chemotherapy, the majority of osteosarcoma can be resected,
with the affected bone replaced by titanium endoprostheses reducing the need for
an amputation.
Comparatively radiotherapy has a relatively minor role in the treatment of osteo-
sarcoma. The indications for radiotherapy include unresectable primary, intralesional
resection, or unresectable positive margins with poor histological response to chemo-
therapy. When delivered adjuvantly following surgery a dose of 60 Gy in 30 fractions
of 2 Gy is recommended, whereas for macroscopic disease doses of up to 66 Gy in
33 fractions of 2 Gy are used, in both instances respecting the tolerance of adjacent
OAR. For metastatic disease palliative radiotherapy may be utilized, with relatively
high doses used for patients of good performance status, e.g. 40 Gy in 15 daily frac-
tions of 2.67 Gy, 30 Gy in 10 daily fractions of 3 Gy, or 36 Gy in 6 weekly fractions
of 6 Gy.

21.15  Central nervous system tumours

With over 400 children diagnosed every year, CNS tumours account for over one-​fifth
of malignant childhood tumours. Paediatric CNS tumours encompasses a wide range
of tumour types with variable outcomes, from medulloblastoma where the 5-​year OS
is approximately 75–​80% for standard risk disease, to paediatric high grade glioma
with a 5-​year OS of 20% and brainstem diffuse midline glioma where the median
survival is only 9.5 months. Many of the long-​term survivors experience sequelae as a
consequence of the tumour, therapy or both. Whilst chemotherapy and other systemic
therapies are being increasingly used for paediatric CNS tumours, so far they have
only produced incremental gains and not the significant improvements in survival
observed in other paediatric malignancies.
Radiotherapy for children with CNS tumours is technically challenging. Proton
beam therapy is being increasingly used, as previously discussed, in an effort to re-
duce the late effects of radiation, especially neurocognitive sequelae, vascular effects,
and second malignancies. Many children require craniospinal radiotherapy, which is
one of the most technically challenging radiotherapy treatments to deliver safely and
effectively.
 Central nervous system tumours 493

In addition to advances in radiotherapy technology, there have been other im-

portant advances in the management of these patients. Modern surgical techniques
incorporate an operating microscope, which has improved precision and endoscopic
capability, improved the treatment of hydrocephalus, and enabled minimally inva-
sive approaches for pineal, intraventricular, and pituitary tumours(29). Other com-
ponents which have improved the ability to achieve maximal safe resections include
neuro-​navigation techniques, intraoperative/​perioperative MRI, and other imaging
techniques, fluorescing agents to help define tumour margins, and intra-​operative
neuro-​physiological monitoring to limit damage to the eloquent brain. Through these
techniques, and the availability of stereotactic biopsy, histological material is obtained
for a greater number of cases and so is available to help guide management.
Another key area where there have been significant changes is in the understanding
of tumour biology. This is exemplified by medulloblastoma where molecular ana-
lysis has redefined the previous histological-​based classification currently into four
main molecular subgroups (see Fig. 21.6). This work is ongoing, across tumour types,
internationally; it is being incorporated increasingly into trial protocols and into the
new WHO classification(30). In parallel with this has been an evolution of functional
imaging techniques such as diffusion weighted MRI, diffusion tensor imaging, and
MR spectroscopy which are now being used routinely in many units.

Molecular Subgroups of Medulloblastoma

CONSENSUS WNT SHH Group 3 Group 4
Cho (2010) C6 C3 C1/C5 C2/C4
Northcott (2010) WNT SHH Group C Group D
Kool (2008) A B E C/D
Thompson (2006) B C;D E, A A, C

DEMOGRAPHICS
Age Group:

Gender:

CLINICAL FEATURES
Histology
Metastasis
Prognosis

GENETICS

GENE EXPRESSION

Fig 21.6  Molecular subgroups of medulloblastoma(31).

Reproduced with permission from Taylor MD, Northcott PA, Korshunov A, et al. Molecular subgroups
of medulloblastoma: the current consensus. Acta Neuropathologica 2012; 123: pp. 465–​472,
© Springer Nature 2012.
494 Principles of paediatric radiation oncology

21.15.1  Long-​term effects of radiotherapy

for CNS tumours
Endocrine deficiencies are a commonly observed late effect in survivors of child-
hood CNS tumours, usually secondary to irradiation of the hypothalamic–​pituitary
axis, although they can also arise as a result of thyroid irradiation during craniospinal
radiotherapy. Growth hormone deficiency is the most commonly observed but other
endocrine deficiencies such as thyrotropin-​releasing hormone (TRH) or adrenocor-
ticotrophic hormone (ACTH) deficiency may occur.
Due to the inhibitory effects of radiotherapy on the growth of irradiated vertebrae
and soft tissue, the growth of children receiving craniospinal irradiation will inevit-
ably be restricted. The potential late vascular effects of radiotherapy in children are
becoming better understood, and include the long-​term risk of stroke and intracranial
haemorrhage. The long-​term cognitive effects of radiotherapy to the CNS are a sig-
nificant concern, particularly for infants, although other factors affecting this include
the presence of hydrocephalus, the tumour itself, and also the impact of intracranial
surgeries.

21.15.2  Chemotherapy for central nervous

system tumours
The most important factor which predicts for long-​term neuropsychological outcomes
is the age at the time of radiotherapy; therefore, radiotherapy is delayed if possible
for children aged < 3 years with alternative chemotherapy-​based strategies used, for
example low grade glioma (SIOP-​LGG 2004), ependymoma (SIOPe Ependymoma
II), and medulloblastoma, and other CNS embryonal tumours (Head Start II).
Chemotherapy and novel targeted therapies are being used more frequently for CNS
tumours, particularly in the clinical trial setting for relapsed disease and subgroups
with poor prognosis. Commonly used agents include platinum-​based drugs, vinca
alkaloids, lomustine, ifosfamide, cyclophosphamide, methotrexate, and etoposide. In
the current SIOP CNS GCT2 study for intracranial germ cell tumours (GCT), pri-
mary chemotherapy is being used for all non-​germinomatous GCT and localized
germinoma. For medulloblastoma adjuvant Packer (PCV; cisplatin, lomustine, vincris-
tine)-​based chemotherapy has been demonstrated to improve outcomes, and current
strategies aim to reduce the associated toxicities, such as hearing loss through reducing
the amount of cisplatin received, using alternate cycles of PCV and cycles of cyclo-
phosphamide and vincristine, an approach adopted in the SIOP PNET5 MB study.
More intensive schedules are being explored for metastatic and high-​risk disease.

21.15.3  Low-​grade glioma
Paediatric low-​grade glioma (WHO Grade I  and II) are the commonest paediatric
CNS tumour, contributing 40% of such diagnoses. The most frequent histologies in-
clude pilocytic astrocytoma (Grade I), which make up more than two-​thirds of all
cases, glioneuronal tumours (Grade I), diffuse astrocytoma (Grade II), and pilomyxoid
astrocytoma (Grade II). Neurofibromatosis type I (NF1) characteristically predisposes
 Central nervous system tumours 495

to the development of these tumours. The majority are associated with a favourable
prognosis, the 20 year cumulative incidence of death due to glioma being 12%, and
OS 87%(32).
Initial management often involves maximal safe surgical resection, limiting the risk
of related morbidity, particularly as pilocytic astrocytoma may undergo long periods
of ‘quiescence’ even when incompletely resected, and can be effectively treated with
both chemotherapy and radiotherapy. Those involving the hypothalamus and/​or the
optic tracts with typical imaging features on MRI are usually not biopsied, as the risk
of the procedure (including visual deterioration or hypothalamic dysfunction) out-
weighs the very low risk of an incorrect diagnosis. Adjuvant treatment is then reserved
for those with clinical or radiological evidence of progression and for those with severe
symptoms or where there is a threat of visual loss.
For infants and young children (< 8  years) the mainstay of treatment is chemo-
therapy, with carboplatin and vincristine delivered over an 18-​month period, as per
the SIOP-​LGG 2004 study. With this approach a 5-​year progression free survival of
46% (OS 89%) was achieved. This randomized study showed the addition of etoposide
conferred no benefit; age > 8 years, the presence of diencephalic syndrome, and early
disease progression by week 24 were associated with poorer overall survival(33). In
this study radiotherapy was the standard adjuvant treatment for those aged 8 years or
older, and was also used for those with evidence of progressive disease, delivered at a
median age of 8 years. The routine use of chemotherapy for those < 8years achieved a
median delay to the commencement of radiotherapy of 2.3 years, and has the potential
to reduce the impact of radiotherapy on long-​term cognitive and neuropsychological
function, although the magnitude of benefit remains to be established.

21.15.4 Technique
Inverse-​planned rotational or fixed field IMRT is routinely used, for cases where
proton therapy is not feasible, to maximize conformity and homogeneity of dose to
the target and minimize dose to critical OARs (Fig. 21.8).

Clinical target volume
GTV includes the extent of tumour on T2w MRI, including postoperative tumour
bed for where prior surgical resection has been undertaken. GTV expanded by 0.5 cm
margin to CTV.

Patient position and immobilization

Supine, with a thermoplastic mask to immobilize the head (+/​-​ neck).

Localization of clinical target volume/​planning target volume

MRI fusion, with T1w contrast enhanced and T2w (FLAIR) sequences to aid GTV
definition (Fig. 21.7).

Dosimetry
Technique dependent; for IMRT dose specified at the median dose to the PTV.
496 Principles of paediatric radiation oncology

Fig. 21.7  Contrast enhanced planning CT and volumetric MR fusion images for diffuse
astrocytoma.

Dose prescription
54 Gy in 30 fractions of 1.8 Gy daily.
For spinal primaries: dose limited to 50.4 Gy in 28 fractions of 1.8 Gy daily.

21.16  High-​grade glioma
High-​ grade glioma, incorporating anaplastic astrocytoma (Grade III) and glio-
blastoma (Grade IV), are uncommon in childhood making up only 8–​12% of paedi-
atric CNS tumours. They are associated with a poor prognosis, with a 5-​year overall
survival of approximately 20%. Management is based on maximal surgical resection
and postoperative (or primary) radiotherapy to a dose of 54 Gy in 30 fractions, util-
izing a CTV margin of 2 cm from GTV. Despite initial studies suggesting poor sensi-
tivity of paediatric high grade glioma to temozolamide, this has been widely adopted,
and recent phase 2 studies have investigated whether combining concomitant and
adjuvant temozolamide with agents such as lomustine or bevacizumab confers any
benefit(34, 35).

Fig. 21.8  Rotational IMRT for diffuse astrocytoma using single 6 MV VMAT arc.

 High-grade glioma 497

21.16.1  Diffuse midline glioma

In the new WHO 2016 classification there has been a further refinement in the defin-
ition of paediatric diffuse midline glioma incorporating the typical molecular charac-
teristics with the K27M mutations in Histone 3 gene H3F3A, or the related HIST1H3B;
these tumours occurring in the midbrain (thalamus), brainstem, or spinal cord. In the
brainstem diffuse midline glioma were formerly labelled as diffuse intrinsic pontine
glioma (DIPG).
Brainstem diffuse midline glioma are usually diagnosed on the basis of their typical
MR appearance, as biopsy can be associated with significant risk of additional mor-
bidity; however, in the context of the recently opened ‘Biological Medicine for DIPG
Eradication (BIOMEDE)’ study, biopsies are being introduced to guide treatment with
targeted therapies, in combination with radiotherapy. The prognosis for these patients
remains very poor with tumour progression usually within 9–​12 months, and almost
no long-​term survivors. The management of these children is challenging, and age-​
appropriate palliative care should be introduced at an early stage. There is a cohort of
these tumours with rapidly progressing neurological symptoms where radiotherapy
needs to be commenced urgently.
To date there has been no trial of systemic chemotherapy that has demonstrated
any meaningful benefit, and novel treatment approaches including the BIOMEDE
study are being investigated. Radiotherapy remains the standard of care and is able
to achieve symptomatic improvement and disease stabilization in approximately 80%
of children. With such a bleak prognosis alternative radiotherapy strategies have been
explored, demonstrating equivalent outcomes with hypofractionated radiotherapy,
delivered over only 2.5–​3 weeks(36). For patients achieving a meaningful response to
initial radiotherapy, with a sustained symptomatic response for greater than 3 months,
there is emerging data showing improved survival with low dose re-​irradiation at a
dose of 20 Gy(37).

21.16.2 Technique
3D conformal radiotherapy, although lateral opposed fields often used, particularly
as an initial phase of radiotherapy when treatment is required urgently. Alternatively
IMRT, fixed field or rotational, can be used, and IMRT is recommended when under-
taking re-​irradiation for selected cases.

Clinical target volume
GTV as defined on diagnostic MR scan with a margin to CTV of 1.5–​2 cm along po-
tential areas of spread superiorly, inferiorly, and posteriorly along brainstem. For re-​
irradiation a reduced CTV margin of 5 mm should be used.

Patient position and immobilization

Supine in thermoplastic head mask.

Localization of clinical target volume/​planned target volume

CT simulation +/​-​iv contrast. MRI fusion with volumetric T2FLAIR sequences can
aid GTV definition.
498 Principles of paediatric radiation oncology

Dosimetry
Dose specified at 100%, or at the median dose for IMRT.

Dose prescription
Primary radiotherapy:
54 Gy in 30 fractions of 1.8 Gy daily, or alternatively 39 Gy in 13 fractions of 3
Gy daily.
Re-​irradiation:
20 Gy in 10 fractions of 2 Gy daily.

21.16.3 Ependymoma
Ependymomas are thought to arise from the lining of the ventricles of the brain and the
central canal of the spine, and comprise approximately 9% of paediatric CNS tumours.
Prognosis remains poor with different studies reporting 5-​year overall survivals in the
range of 50–​85% for completely resected tumours. In the new WHO 2016 classifica-
tion, they are subdivided into five subtypes: subependymoma (Grade I), myxopapillary
ependyomoma (Grade I), ependymoma (Grade II), ependymoma RELA fusion posi-
tive (Grade II or III), and anaplastic ependymoma (Grade III). Infratentorial primary
site is found in approximately two-​thirds of cases, with supratentorial and spinal
locations less common. For the majority initial management is focused on surgical
excision, with the extent of resection, grade, tumour location, and age all important
prognostic factors which help guide management decisions.
Adjuvant radiotherapy is utilized for the majority of cases except for infants under
the age of 12–​18 months where chemotherapy-​based approaches are preferred.
Current research efforts are aimed at improving the prospect of local tumour con-
trol and to this end the recently opened SIOPe BTG Ependymoma II study has in-
corporated both surgical and radiotherapy QA. As the predominant site of relapse
is within the local tumour bed, there is no evidence to support the routine use of
craniospinal radiotherapy for localized non-​metastatic disease, which had been used
previously. Based on data from St Jude’s showing improved outcomes, 59.4 Gy in 1.8
Gy daily fractions has now been adopted as the standard fractionation for intracranial
ependymoma, although 54 Gy is still used for children less than 18 months old or with
risk factors as multiple surgeries (more than 2) or poor neurological status(38). Proton
therapy or IMRT are used for the majority of cases, with a 1.0 cm expansion from GTV
(including tumour bed and any residual tumour) to CTV. In the Ependymoma II study
there are three investigational strata, including Stratum 2 which is exploring strategies
for patients with gross residual disease through the addition of chemotherapy, and
utilizing a stereotactic boost for persistent residual tumour of 8 Gy in two fractions, in
4 Gy daily fractions.

21.16.4  Medulloblastoma and other childhood CNS

embryonal tumours
Medulloblastoma is a childhood CNS embryonal tumour which arises in the cere-
bellum, and is the most common paediatric malignant brain tumour, accounting
 High-grade glioma 499

for up to 20% of childhood CNS tumours. Like medulloblastoma, the majority of

the other childhood CNS embryonal tumours are notable for their propensity for
metastatic spread via the cerebrospinal fluid; previously they were defined as primi-
tive neuroectodermal tumours (PNET), but they have been newly reclassified as
medulloblastoma and CNS embryonal tumours(30). Medulloblastoma are subdivided
based either on genetics (WNT, SHH, Group  3  & Group  4) and/​or based on hist-
ology (classic, desmoplastic/​nodular, those with extensive nodularity, and large cell/​
anaplastic)(31).
At the present time treatment stratification, in the open SIOP PNET5 MB study,
is still mainly designated based on the risk grouping, Standard Risk (localized dis-
ease, non-​large cell/​anaplastic, non-​MYC/​ MYCN amplified, and no residual tumour
> 1.5 cm2). De-​escalation of treatment for WNT positive (β catenin IHC positivity/​
mutation positive) medulloblastoma in children (< 16 years) is under investigation, as
published molecular data supports this to be a new Low Risk category(30, 39). High Risk
is defined as follows: metastatic disease and/​or localized large cell/​anaplastic and/​or
MYC/​ MYCN amplified tumours, and/​or residual tumour > 1.5 cm2.
In the new WHO 2016 classification, the category of CNS embryonal tumours
incorporates tumours formerly classified as supratentorial PNET. These tumours
typically arise elsewhere in the CNS, often in the supratentorial cerebral cortex,
and are subclassified as follows:  embryonal tumours with multilayered rosettes,
medulloepitheliomas, CNS neuroblastomas, CNS ganglioneuroblastomas, and atyp-
ical teratoid/​rhabdoid tumours (ATRT)(30). Pineoblastomas arising in the pineal
gland are not classified as a CNS embryonal tumour, but instead are in the category
of ‘Tumours of the Pineal Region’, although they are treated in a similar fashion to
medulloblastoma.
The standard therapeutic approach for medulloblastoma is to achieve complete sur-
gical resection of the primary tumour followed by craniospinal irradiation (CSI) and
a boost to the primary tumour site, and sites of metastases where present. Historically
a uniform approach was used for all patients irrespective of their risk grouping, exem-
plified by the European SIOP/​UKCCSG PNET-​3 study where CSI 35 Gy in 21 daily
fractions and a boost to the entire posterior fossa of 20 Gy in 12 daily fractions was
used; analysis from this study also demonstrated the negative impact of gaps in treat-
ment and extended overall treatment time(40, 41). Through further collaborative group
studies in Europe and North America it has been shown that a lower dose of CSI, 23.4
Gy, and boosting the posterior fossa to a total dose of 54-​55.8 Gy, can be used without
compromising cure for standard risk medulloblastoma(44). Recent reports at ASTRO
2016 from the COG ACNS0331 study have shown that further dose reduction, with
18 Gy CSI instead of 23.4 Gy, for standard risk disease compromises survival, but also
importantly confirmed that a reduced volume boost to the posterior fossa tumour bed
produced equivalent outcomes to boosting of the whole posterior fossa(43). For high-​
risk medulloblastoma, including metastatic disease, and for CNS embryonal tumours,
including metastatic ATRT), a higher CSI dose of 35–​39.6 Gy, using a dose per fraction
of 1.67–​1.8 Gy, remains the standard.
Concurrent chemotherapy has been routinely used for medulloblastoma, with
weekly vincristine the standard of care. However, there are associated toxicities,
500 Principles of paediatric radiation oncology

and the benefits of this remain uncertain, which have led to it not being included in
any of the treatment arms for the open SIOP PNET5 standard risk medulloblastoma
study. Various treatment strategies are in use for high-​risk disease, as reflected in
the CCLG high-​risk medulloblastoma guidelines published in 2015; these vary
from the more routine use of high-​dose chemotherapy to the adoption of concomi-
tant daily carboplatin during radiotherapy(44, 45). These treatment paradigms are also
employed for the majority of the childhood CNS embryonal tumours, with the ex-
ception of localized ATRT, where a combined approach of intensive chemotherapy
and focal radiotherapy to 54 Gy is used, as per the current EU-​RHAB consensus
guidelines (46).
Several European and North American collaborative groups have investigated the
role of hyperfractionated radiotherapy (HFRT) for both standard-​risk and high-​risk
medulloblastoma. In the European HIT-​SIOP PNET 4 study for standard-​risk disease
showed no advantage for HFRT in terms of disease control, although HFRT did im-
prove long-​term verbal IQ and executive function, it caused greater impact on growth
delay, and quality of life was not improved, therefore conventional fractionation re-
mains the standard of care(47, 48). There are logistical challenges involved in the delivery
of HFRT or hyperfractionated accelerated radiotherapy for CSI in children, particu-
larly for those requiring general anaesthesia and for centres treating with proton
therapy; however, it remains an attractive concept for patients with metastatic disease
and continues to be investigated by collaborative groups, being one of the randomized
arms in the proposed SIOP-​E HR Medulloblastoma study.

21.16.5 Craniospinal irradiation
The safe and effective delivery of CSI is highly complex and requires a specialized
multi-​disciplinary team, and should only be delivered in paediatric radiotherapy
centres routinely using such treatments. Different radiotherapy techniques can be
used including rotational IMRT and protons. The technique described below is in use
at the Royal Marsden Hospital.

Technique
3D conformal radiotherapy, incorporating lateral cranial fields, with usually a single
posterior spinal field, delivered at extended focus-​to-​skin distance when required to
achieve larger field sizes. The inferior cranial field borders are matched to the superior
field border of the spinal field, using moving junctions with three match points at ap-
proximately 1-​cm intervals; all three junctions treated daily.

Clinical target volume
CTV includes the entire subarchnoid space, including the extensions along the nerve
roots to the level of the intervertebral foramina. For the brain CTV, it is essential to
include accurate delineation of the cribriform fossa, temporal lobes, and base of skull.
For the spinal CTV, care should be taken to define the inferior aspect of the thecal sac,
as the inferior field border should be at least 1 cm inferior to the end of the thecal sac.
Typically the lateral field borders of the posterior spinal field are at least 1 cm beyond
the lateral edge of the pedicles.
 High-grade glioma 501

Patient position and immobilization

Supine with thermoplastic shell to immobilize the head, neck, and shoulders. Midline
and lateral tattoos for alignment of torso.

Localization of clinical target volume/​planning target volume

CT simulation +/​-​iv contrast. MRI fusion with volumetric T1w contrast enhanced
sequences to aid GTV definition, particularly for the boost to the site of the primary
tumour.

Dosimetry
For 3D conformal CSI, dose specified at 100%.

Dose prescription
Standard Risk: 23.4 Gy in 13 fractions of 1.8 Gy daily.
High Risk: 36 Gy in 20 fractions of 1.8 Gy daily.
Meticulous attention to detail in the planning and delivery of CSI is essential and con-
tributes to the cure of medulloblastoma/​PNET. It is essential to avoid areas of underdose
at field junctions and partial shielding of any area of meninges (see Fig 21.10).
The ‘moving junction’ between abutting fields is a ‘safety measure’, which reduces
the risk of underdose or overdose in the cervical spinal cord if a systematic error de-
velops during CSI. In children, the cribriform fossa frequently lies between the lenses
(Fig. 21.9). In many series of patients treated for medulloblastoma, the cribriform
fossa has been the site for isolated recurrence in a significant minority of patients. It
may not always be possible to treat the CTV and adequately shield the lenses, in which
case priority is given to treating the CTV.
CSI techniques continue to evolve, and are incorporating technical developments
in radiotherapy immobilization, imaging, planning, and treatment delivery. For the

Fig. 21.9  Typical cranial field with axial isodoses for craniospinal irradiation, showing
coverage of cribriform plate, and 3 different inferior field borders for feathering of dose
for every fraction.
502 Principles of paediatric radiation oncology

Fig. 21.10  Spinal dosimetry for

craniospinal irradiation, with a single
posterior spinal field matched to the
inferior cranial field border.

posterior fossa tumour bed boost using photons, rotational IMRT should ideally be
used to achieve high conformity to the PTV whilst minimizing dose to important
OARs such as the cochlea, hypothalamus, and hippocampus.

21.16.6  Intracranial germ cell tumours

Intracranial germ cell tumours account for approximately 30% of paediatric germ
cell tumours. Germinomas are the histological equivalent of testicular seminoma.
They generally arise in the suprasellar region and/​or in the pineal region. Non-​
germinomatous germ cell tumours are the histological equivalent of testicular non-​
seminomatous germ cell tumours, i.e. embryonal carcinomas, yolk sac tumours, or
choriocarcinomas. They generally arise in the pineal region and may secrete alpha
fetoprotein (AFP) and/​or human chorionic gonadotrophin (HCG).
More than 90% of germinomas are cured with radiotherapy alone, with historically
CSI employed for all germinoma because of the risk of leptomeningeal metastases.
Given the excellent prognosis, and ability to salvage relapse, strategies to reduce the
risk of long-​term toxicities have been implemented, by reducing both the CSI and the
 Conclusions 503

boost dose, and the adoption of whole ventricular cranial irradiation (WVRT) for lo-
calized (including bifocal) disease. For WVRT the WV-​CTV is defined as the lateral,
third, and fourth ventricles plus a margin of 0.5cm. For tumour bed (TB) irradiation,
or boost, the TB-​GTV is defined as the initial anatomically involved part of the brain,
and any residual tumour, with a margin of 0.5  cm to TB-​CTV for suprasellar and
pineal tumours. However, for atypical primary sites, with a higher risk of infiltration
of adjacent normal tissue a 1-​cm margin should be used to TB-​CTV.
In the current SIOP CNS-​ GCT2 study the dose for microscopic disease in
germinomas, i.e. WVRT or CSI, is 24 Gy in 15 fractions of 1.6 Gy, and the boost dose
to the site of the primary tumour(s) is a further 16 Gy in 10 daily fractions of 1.6
Gy. Neoadjuvant chemotherapy for localized germinoma is being investigated in the
CNS-​GCT2 study, but for metastatic germinoma the standard of care remains CSI
and boost without chemotherapy. For non-​germinomatous tumours the prognosis
is worse; initial treatment for all is with cisplatin, etoposide, and ifosfamide chemo-
therapy followed by radiotherapy. For focal disease irradiation is delivered, treating
the TB-​PTV (as above), to a dose of 54 Gy in 30 fractions, at 1.8 Gy per fraction. For
metastatic disease, the initial phase of radiotherapy is CSI to a dose of 30 Gy in 20
fractions, at 1.5 Gy per fraction, then boosting initial macroscopic intracranial dis-
ease to 54 Gy and spinal disease to 50.8 Gy (both boost phases delivered using a dose
of 1.6 Gy per fraction).

21.17 Conclusions
Paediatric tumours are extremely rare and include a wide variety of different tumour
types, which present in many different ways. Whilst the majority of children receiving
treatment for tumours are cured, especially those with tumour types such as acute
lymphoblastic leukaemia, intracranial germinoma, and Wilms’ tumour, there remain
others where cure is less likely and for the majority treatment is delivered with the
intention of achieving prolongation of life, including metastatic neuroblastoma, meta-
static alveolar rhabdomyosarcoma, and diffuse midline and high-​grade gliomas. Given
the challenges faced by children, and their families, treatment should be delivered by
specialist multiprofessional paediatric oncology teams in highly specialized centres in
order that the best outcomes can be achieved. Best practice for patient-​and family-​
centred management needs be undertaken in accordance with national and inter-
national trial protocols or approved guidelines, including the ‘Good Practice Guide
for Paediatric Radiotherapy’(1).
Radiotherapy plays an important role in the management of many of these children.
They require the highest standard of radiotherapy planning and delivery, incorporating
modern technical developments and quality assurance, including protons, IMRT,
and brachytherapy where dosimetric benefits in the short and/​or long term can be
achieved. With the increasing use of concurrent combined modality therapy, and
novel radiotherapy techniques, constant vigilance for interactions is required. As ap-
proximately 80% of children treated for a paediatric tumour are long-​term survivors,
the study and reporting of the late consequences of treatment is of paramount import-
ance, in order to continue to improve outcomes in the future.
504 Principles of paediatric radiation oncology

Further reading
The Royal College of Radiologists, Society and College of Radiographers, Children’s Cancer
and Leukaeamia Group. Good practice guide for paediatric radiotherapy (2nd edn).
London: The Royal College of Radiologists, 2018.
Constine LS, Tarbell NJ, Halperin EC. Pediatric Radiation Oncology (6th edn). Philadelphia,
PA: Lippincott Williams & Wilkins, 2016.
Pizzo PA, Poplack DG. Principles and Practice of Pediatric Oncology (7th edn). Philadelphia,
PA: Lippincott, Williams and Wilkins, 2015.
Merchant TE, Kortmann RD. Pediatric Radiation Oncology (Pediatric Oncology). Springer; 1st
ed. 2018 edition (6 Nov. 2017).
Schwartz CL, Hobbie WL, Constine LS, Ruccione KS. Survivors of Childhood and Adolescent
Cancer: A Multidisciplinary Approach (Pediatric Oncology). Springer; 3rd ed. 2015 edition
(23 Sept. 2015).
Yock T, De Laney TF, Esty B, Tarbell NJ. Pediatric tumors. In De Laney TF, Kooy HM (eds)
Proton and Charged Particle Radiotherapy. Philadelphia, PA: Lippincott Williams &
Wilkins, 2007, pp. 125–​39.

References
1. The Royal College of Radiologists, Society and College of Radiographers, Children’s
Cancer and Leukaemia Group. Good practice guide for paediatric radiotherapy (2nd edn).
London: The Royal College of Radiologists, 2018.
2. Newhauser WD, Durante M. Assessing the risk of second malignancies after modern
radiotherapy. Nature Reviews Cancer 2011; 11:438–​48.
3. Lipshultz SE, Lipsitz SR, Mone SM, et al. Female sex and higher drug dose are risk factors
for late cardiotoxic effects of doxorubicin therapy for childhood cancer. New England
Journal of Medicine 1995; 332:1738–​43.
4. Yock TI, Yeap BY, Ebb DH, et al. Long-​term toxic effects of proton radiotherapy for
paediatric medulloblastoma: a phase 2 single-​arm study. Lancet Oncology 2016; 17:287–​98.
5. Greenberger BA, Pulsifer MB, Ebb DH, et al. Clinical outcomes and late endocrine,
neurocognitive, and visual profiles of proton radiation for pediatric low-​grade gliomas.
International Journal of Radiation Oncology, Biology, Physics 2014; 89:1060–​68.
6. Gunther JR, Sato M, Chintagumpala M, et al. Imaging changes in pediatric intracranial
ependymoma patients treated with proton beam radiation therapy compared to intensity
modulated radiation therapy. International Journal of Radiation Oncology, Biology, Physics
2015; 93:54–​63.
7. Yock TI, Bhat S, Szymonifka J, et al. Quality of life outcomes in proton and photon treated
pediatric brain tumor survivors. Radiotherapy & Oncology. 2014; 113:89–​94.
8. Weber DC, Malyapa R, Albertini F, et al. Long term outcomes of patients with skull-​base
low-​grade chondrosarcoma and chordoma patients treated with pencil beam scanning
proton therapy. Radiotherapy & Oncology 2016; 120:169–​74.
9. https://​www.england.nhs.uk/​commissioning/​spec-​services/​npc-​crg/​group-​b/​b01/​
10. Carrie C, Hoffstetter S, Gomez F, et al. Impact of targeting deviations on outcome in
medulloblastoma: Study of the French Society of Pediatric Oncology (SFOP). International
Journal of Radiation Oncology, Biology, Physics 1999; 45:435–​9.
 References 505

11. Hartley KA, Li C, Laningham FH, et al. Vertebral body growth after craniospinal
irradiation. International Journal of Radiation Oncology, Biology, Physics 2008; 70:1343–​49.
12. Jannoun L, Bloom HJG. Long-​term psychological effects in children treated for
intracranial tumors. International Journal of Radiation Oncology, Biology, Physics 1990;
18:747–​53.
13. Vora A, Goulden N, Wade R, et al. Treatment reduction for children and young adults
with low-​risk acute lymphoblastic leukaemia defined by minimal residual disease (UKALL
2003): a randomised controlled trial. Lancet Oncology 2013; 14:199–​209.
14. Vora A, Goulden N, Mitchell C, et al. Augmented post-​remission therapy for a minimal
residual disease-​defined high-​risk subgroup of children and young people with clinical
standard-​risk and intermediate-​risk acute lymphoblastic leukaemia (UKALL 2003): a
randomised controlled trial. Lancet Oncology 2014; 15:809–​18.
15. Creutzig U, van den Heuvel-​Eibrink MM, Gibson B, et al. Diagnosis and management
of acute myeloid leukemia in children and adolescents: recommendations from an
international expert panel. Blood 2012; 120:3187–​205.
16. Meignan M, Gallamini A, Haioun C, et al. Report on the Second International Workshop
on interim positron emission tomography in lymphoma held in Menton, France, 8-​9 April
2010. Leukemia & Lymphoma 2010; 51:2171–​80.
17. Monclair T, Brodeur GM, Ambros PF, et al. The International Neuroblastoma Risk Group
(INRG) staging system: an INRG Task Force report. Journal of Clinical Oncology 2009;
27:298–​303.
18. Cohn SL, Pearson AD, London WB, et al. The International Neuroblastoma Risk Group
(INRG) classification system: an INRG Task Force report. Journal of Clinical Oncology
2009; 27:289–​97.
19. Missiaglia E, Williamson D, Chisholm J, et al. PAX3/​FOXO1 fusion gene status is the key
prognostic molecular marker in rhabdomyosarcoma and significantly improves current
risk stratification. Journal of Clinical Oncology 2012; 30:1670–​77.
20. Bisogno G, De Salvo GL, Bergeron C, et al. The role of doxorubicin in the treatment of
rhabomyosarcoma: Preliminary results from the EpSSG RMS2005 Randomized Trial.
Pediatric Blood & Cancer 2014; 61:S133–​34.
21. Chisholm JC, Marandet J, Rey A, et al. Prognostic factors after relapse in nonmetastatic
rhabdomyosarcoma: a nomogram to better define patients who can be salvaged with
further therapy. Journal of Clinical Oncology 2011; 29:1319–​25.
22. Schoot RA, Slater O, Ronckers CM, et al. Adverse events of local treatment in long-​term
head and neck rhabdomyosarcoma survivors after external beam radiotherapy or AMORE
treatment. European Journal of Cancer 2015; 51:1424–​34.
23. Levy A, Martelli H, Fayech C, et al. Late toxicity of brachytherapy after female genital
tract tumors treated during childhood: Prospective evaluation with a long-​term follow-​up.
Radiotherapy & Oncology 2015; 117:206–​12.
24. Martelli H, Borrego P, Guérin F, et al. Quality of life and functional outcome of male
patients with bladder-​prostate rhabdomyosarcoma treated with conservative surgery and
brachytherapy during childhood. Brachytherapy 2016; 15:306–​11.
25. Gaspar N, Hawkins DS, Dirksen U, et al. Ewing Sarcoma: Current Management and
Future Approaches Through Collaboration. Journal of Clinical Oncology 2015; 33:3036–​46.
26. Isakoff MS, Bielack SS, Meltzer P, et al. Osteosarcoma: current treatment and a
collaborative pathway to success. Journal of Clinical Oncology 2015; 33:3029–​35.
506 Principles of paediatric radiation oncology

27. Marina NM, Smeland S, Bielack SS, et al. Comparison of MAPIE versus MAP in patients
with a poor response to preoperative chemotherapy for newly diagnosed high-​grade
osteosarcoma (EURAMOS-​1): an open-​label, international, randomised controlled trial.
Lancet Oncology 2016; 17:1396–​408.
28. Bielack SS, Smeland S, Whelan JS, et al. Methotrexate, doxorubicin, and cisplatin
(MAP) plus maintenance pegylated interferon alfa-​2b versus MAP alone in patients
with resectable high-​grade osteosarcoma and good histologic response to preoperative
MAP: First results of the EURAMOS-​1 good response randomized controlled trial. Journal
of Clinical Oncology 2015; 33:2279–​87.
29. Zebian B, Vergani F, Lavrador JP, et al. Recent technological advances in pediatric brain
tumor surgery. CNS Oncology 2017; 6:71–​82.
30. Louis DN, Perry A, Reifenberger G, et al. The 2016 World Health Organization
Classification of Tumors of the Central Nervous System: a summary. Acta Neuropathologica
2016; 131:803–​820.
31. Taylor MD, Northcott PA, Korshunov A, et al. Molecular subgroups of
medulloblastoma: the current consensus. Acta Neuropathologica 2012; 123:465–​472.
32. Bandopadhayay P, Bergthold G, London WB, et al. Long-​term outcome of 4,040 children
diagnosed with pediatric low-​grade gliomas: an analysis of the Surveillance Epidemiology
and End Results (SEER) database. Pediatric Blood & Cancer. 2014; 61: 1173–​79.
33. Gnekow AK, Walker DA, Kandels D, et al. A European randomised controlled trial of the
addition of etoposide to standard vincristine and carboplatin induction as part of an 18-​
month treatment programme for childhood (≤16 years) low grade glioma -​A final report.
European Journal of Cancer. 2017; 81:206–​25.
34. Lashford LS, Thiesse P, Jouvet A, et al. Temozolomide in malignant gliomas of
childhood: a United Kingdom Children’s Cancer Study Group and French Society for
Pediatric Oncology Intergroup Study. Journal of Clinical Oncology 2002; 20:4684–​89.
35. Jakacki RI, Cohen KJ, Buxton A, et al. Phase 2 study of concurrent radiotherapy and
temozolomide followed by temozolomide and lomustine in the treatment of children with
high-​grade glioma: a report of the Children’s Oncology Group ACNS0423 study. Neuro-​
Oncology 2016; 18:1442–​50.
36. Janssens GO, Jansen MH, Lauwers SJ, et al. Hypofractionation vs conventional radiation
therapy for newly diagnosed diffuse intrinsic pontine glioma: a matched-​cohort analysis.
International Journal of Radiation Oncology, Biology, Physics 2013; 85:315–​20.
37. Janssens GO, Gandola L, Bolle S, et al. Survival benefit for patients with diffuse intrinsic
pontine glioma (DIPG) undergoing re-​irradiation at first progression: A matched-​cohort
analysis on behalf of the SIOP-​E-​HGG/​DIPG working group. European Journal of Cancer.
2017; 73:38–​47.
38. Merchant TE, Li C, Xiong X, et al. Conformal radiotherapy after surgery for paediatric
ependymoma: a prospective study. Lancet Oncology 2009; 10:258–​66.
39. Schwalbe EC, Lindsey JC, Nakjang S, et al. Novel molecular subgroups for clinical
classification and outcome prediction in childhood medulloblastoma: a cohort study.
Lancet Oncology 2017; 18:958–​71.
40. Taylor RE, Bailey CC, Robinson K, et al. Results of a randomised study of pre-​radiation
chemotherapy vs radiotherapy alone for non-​metastatic (M0–​1) medulloblastoma the
SIOP/​UKCCSG PNET-​3 study. Journal of Clinical Oncology 2003; 21:1581–​91.
41. Taylor RE, Bailey CC, Robinson KJ, et al. Impact of radiotherapy parameters on outcome
in the International Society of Paediatric Oncology (SIOP)/​United Kingdom Children’s
 References 507

Cancer Study Group (UKCCSG) PNET-​3 study of pre-​radiotherapy chemotherapy for

M0–​1. International Journal of Radiation Oncology, Biology, Physics 2004; 58:1184–​93.
42. Packer RJ, Gajjar A, Vezina G, et al. Phase III study of craniospinal radiation therapy
followed by adjuvant chemotherapy for newly diagnosed average-​risk medulloblastoma.
Journal of Clinical Oncology 2006; 24:4202–​8.
43. Michalski JM, Janss, Vezina G, et al. Results of COG ACNS0331: A phase III trial of
involved-​field radiotherapy (IFRT) and low dose craniospinal irradiation (LD-​CSI) with
chemotherapy in average-​risk medulloblastoma: A report from the Children’s Oncology
Group. International Journal of Radiation Oncology, Biology, Physics 2016; 96:937–​38.
44. Gajjar A, Chintagumpala M, Ashley D, et al. Risk-​adapted craniospinal radiotherapy
followed by high-​dose chemotherapy and stem-​cell rescue in children with newly
diagnosed medulloblastoma (St Jude Medulloblastoma-​96): long-​term results from a
prospective, multicentre trial. Lancet Oncology 2006; 7:813–​20.
45. Jakacki RI, Burger PC, Zhou T, et al. Outcome of children with metastatic
medulloblastoma treated with carboplatin during craniospinal radiotherapy: a Children’s
Oncology Group Phase I/​II study. Journal of Clinical Oncology 2012; 30:2648–​53.
46. Bartelheim K, Nemes K, Seeringer A, et al. Improved 6-​year overall survival in AT/​RT -​
results of the registry study Rhabdoid 2007. Cancer Medicine 2016; 5:1765–​75.
47. Lannering B, Rutkowski S, Doz F, et al. Hyperfractionated versus conventional
radiotherapy followed by chemotherapy in standard-​risk medulloblastoma: results from
the randomized multicenter HIT-​SIOP PNET 4 trial. Journal of Clinical Oncology 2012;
30:3187–​93.
48. Kennedy C, Bull K, Chevignard M, et al. Quality of survival and growth in children
and young adults in the PNET4 European controlled trial of hyperfractionated versus
conventional radiation therapy for standard-​risk medulloblastoma. International Journal of
Radiation Oncology, Biology, Physics 2014; 88:292–​300.
 Chapter 22

Radiotherapy planning
for metastatic disease
Peter Hoskin

22.1 Introduction
Palliative treatments account for a large proportion of the workload of any department
estimated at between 40% and 50% of new patient treatments per year in an average
cancer centre in the UK. These will be predominantly patients with bone metastasis,
spinal cord compression, and brain metastasis. The techniques for such treatments
are specific to this indication rather than the primary tumour site and are therefore
included here in a separate chapter.
More recently the concept of oligometastases referring to the situation in which a
patient may present with one to three metastases only for which radical local ablative
treatment can be considered has altered the approach to the metastatic patient. This
will usually be delivered with stereotactic techniques as described in Chapter 3.
For those with advanced metastatic disease in multiple sites or with poor perform-
ance status then symptom relief is the goal of treatment and such palliative radio-
therapy should be simple, pragmatic, and quick. This does not, however, mean that
careful planning should be ignored. Wherever possible, simulator verification of
volume and beam positions should be used.
Whilst conventional gross tumour volume (GTV), clinical target volume (CTV),
and planning tumour volume (PTV) definitions are not often used, it is important to
consider the principles in defining palliative fields allowing adequate margins around
symptomatic sites to allow for internal organ movement and set-​up variation. When
defining treatment areas in terms of fields rather than volumes, it is important to re-
member that the field edge defined by the simulator light beam or crosswires in virtual
simulation represents the 50% isodose.

22.2 Bone metastasis
22.2.1 Oligometastases
Most experience with oligometastatic bone disease relates to treatment of solitary
spinal metastases.

Patient position and immobilization

Using a stereotactic platform immobilization is critical and a vacuum bag will usually
give the best solution.
 Bone metastasis 509

Table 22.1  Summary of CTV definitions for stereotactic radiotherapy to vertebra

Site of metastasis Clinical Target Volume

Vertebral body Entire body and if lateralised include ipsilateral pedicle, transverse and
spinous process and if diffuse include both pedicles, transverse and
spinous process

Vertebral body, Entire body, bilateral pedicles, transverse processes and spinous
pedicles and process
transverse processes
bilaterally
Unilateral pedicle Ipsilateral pedicle, transverse process and lamina
Unilateral lamina Ipsilateral lamina, pedicle, transverse process and spinous process
Unilateral spinous Spinous process and both pedicles
process

Source: data from Brett W. Cox, MD, Daniel E. Spratt, MD, Michael Lovelock et al. ‘International Spine
Radiosurgery Consortium Consensus Guidelines for Target Volume Definition in Spinal Stereotactic
Radiosurgery’. Int J Radiation Oncologyl Biology, Physics, Volume 83, No. 5, pp. e597–e605, 2012.

Volume and localization
Detailed imaging including CT, MR, and CT PET is required to accurately define the site of
metastasis. Complex guidelines with regard to the CTV definition for different scenarios
have been published(1) and should be consulted. They are summarized in Table 22.1.

Dose distribution
This will depend upon the stereotactic platform to be used (see Chapter 3). An ex-
ample of a lumbar spine oligometastasis plan using cyberknife is shown in Fig. 22.1.

Dose prescription
There is no consensus on optimal dose fractionation in this setting(2). Options include:
◆ Single dose: 15–​24Gy.
◆ 21–​30Gy in three fractions.

22.2.2  Local bone pain

Patient position and immobilization
In general, most bone metastasis with modern radiotherapy equipment can be treated
with the patient supine. Ankle stocks and a headrest will aid immobilization.
Where electron treatments are used (see later in this section) a semisupine position
to access the site to be treated, e.g. ribs, may be more appropriate.
If orthovoltage is used then direct applicator apposition is required for ribs or spine
and the patient may therefore need to be prone or semisupine.

Volume and localization
Volume definition should be based on symptoms rather than radiological changes al-
though confirmation of bone metastasis as the cause of local bone pain is essential.
Specific considerations include the following:
510 Radiotherapy planning for metastatic disease

Fig. 22.1  Dose distribution to treat solitary lumbar spine metastasis using Cyberknife
platform. Note pane top left shows multiple beamlets used by this system.

1. Wherever possible include a whole bone or recognized anatomical structure, for

example, treating the pelvis to the midline including the entire unilateral pubic
rami, as shown in Fig. 22.2, unless, as shown in Fig. 22.3, there is involvement of
the pubic symphysis when the field is seen to extend 1–​2 cm beyond the midline to
ensure that the involved bone is encompassed in the 90% isodose.
2. Patients with multiple bone metastases will often require several treatments and
treatment fields should be used with an eye to subsequent matching of adjacent
fields, for example, in the spine placing fields in the intravertebral spaces and in the
pelvis including the entire hip joint bringing the medial border to the midline.

Fig. 22.2  Radiotherapy field

to treat left hip, including
ipsilateral pubic bone also
involved with metastatic
disease, to midline.
 Bone metastasis 511

Fig. 22.3 Radiotherapy
field to treat right hip
crossing midline to ensure
coverage of metastasis in
the pubic symphysis.

Dose distribution
◆ The majority of bone metastases are best treated with photon beams of 4–​6 MV.
Single applied fields or parallel opposed anterior and posterior fields will enable
treatment to most of the skeleton.
◆ Direct electron beams are appropriate for superficial bones, in particular the ribs
and clavicles.
◆ Orthovoltage beams (250–​500 kV) will typically reach their 80% isodose at a depth
of 3–​3.5 cm which is adequate for treatment of ribs, clavicles, and sacrum.

Dose prescription
The most common dose for bone pain is a single dose of 8 Gy.
1. This should be prescribed as a midplane dose with parallel opposed fields.
2. The spine should be treated at the depth of the vertebral body. This varies along the
length of the spine. If it is not appropriate or possible to screen the spine laterally
in the simulator then a depth of 4–​5 cm will be adequate for the prescription defin-
ition. This will mean an applied dose of around 10 Gy.
3. Electron prescriptions should be to the 100% isodose taking account of the
isodose distribution and effective depth of the beam which will be chosen
for the appropriate situation (see Chapter  2). For ribs electron energies of
8–​10 MeV will give more than adequate depth penetration. Build-​up is not
required.
Orthovoltage beams should be prescribed to an applied 100% isodose. Whilst it is ac-
knowledged that there is a more prominent photoelectric absorption effect with these
beams which will relatively increase the dose to bone, it is not usual to reduce the pre-
scribed dose from that above, i.e. 8 Gy.
Although there is extensive Level 1 evidence to support the use of single doses of
8–​10 Gy for bone pain(3,4) alternative prescriptions are in use for bone pain including
20 Gy in five fractions and 30 Gy in 10 fractions.
512 Radiotherapy planning for metastatic disease

After a single dose of 8 Gy retreatment with a further 8 Gy or 20 Gy in five fractions

is safe and effective(5).

22.2.3  Wide field (hemibody) irradiation for multiple

bone pains
Where there are scattered bone pains from multiple bone metastases external beam
radiotherapy still has an important and effective role using wide field treatment to en-
compass large areas of the body. Conventionally these are often referred to as upper or
lower hemibody treatment although in practice they need not be constrained to these
precise definitions.

Patient position and immobilization

The patient should be supine and immobilization with leg stocks and headrest may aid
reproducibility.

Treatment volume
This will be defined by the areas of pain and limited by the field size available.
Conventionally the hemibody fields are as follows:
1. Upper hemibody: from top of scalp to umbilicus.
2. Lower hemibody: from umbilicus to soles of feet.
3. Whilst these volumes are relevant to patients who may be effectively treated for
end-​stage chemotherapy-​resistant myeloma or lymphoma, for bone metastasis
such precise and rigid definitions are not necessary. Furthermore upper hemibody
irradiation to incorporate the scalp results in alopecia which is undesirable if
unnecessary.
The volume should then be defined by the sites of pain. Where these are predomin-
antly in the ribs and thoracic spine this area can be encompassed in a large field and if
lumbo-​sacral spine, pelvis, and lower limbs, again a large field covering all the painful
areas should be used.

Dose distribution
These volumes are treated with anterior and posterior parallel opposed fields. The
area to be treated is usually confined by the available field size. At 100 cm focus-​
to-​skin distance (FSD), a modern linear accelerator, will be able to provide a field
of 35–​40 cm in length. If a larger area is treated then extended FSD techniques will
be required.

Dose prescription
◆ 8-​Gy midplane dose for the lower hemibody.
◆ 6-​Gy midplane dose where the lungs are included in the volume as this represents
lung tolerance when delivered in a single dose at linear accelerator dose rate. At
higher doses pneumonitis may be encountered.
Lower doses of 4-​Gy midplane dose may be equally effective although they have not
been compared in a randomized trial. One study has evaluated 8 Gy in two fractions
 Bone metastasis 513

as an effective dose. There is, however, no clear advantage of this over a single dose of
8 Gy except where the lungs are in the field(6,7).

22.2.4  Postoperative radiotherapy for bone metastasis

The preferred treatment for a long bone fracture from bone metastasis is internal fix-
ation. The role of postoperative radiotherapy remains uncertain but most patients
with a prognosis of more than 3 months will be offered and treated with postoperative
radiotherapy.

Patient position and immobilization

This will depend upon the bone to be treated.
◆ Hip and femur are treated with the patient supine using ankle stocks to immobilize
the lower limb.
◆ Humerus: arm abducted, elbows flexed, hands on hip.

Volume definition
There are two views as to the volume definition in this setting.
1. The entire bone should be covered on the basis that the marrow cavity may be con-
taminated peroperatively.
2. Only the entire prosthesis need be covered as the area most at risk of residual tu-
mour and regrowth.
In the absence of any data to support either of these two approaches it is clear that at
least the entire prosthesis should be covered by the treatment volume and a margin
of at least 3  cm of normal bone beyond the prosthesis is recommended. Where an
intramedullary nail has been used then this should be covered completely with the
treatment volume which will usually be the entire bone.
Where large fields are used along the length of a bone, even for palliative doses an
attempt should be made to avoid joint spaces and preserve a corridor of normal tissue
for lymphatic drainage.

Dose distribution
Anterior and posterior parallel opposed fields are used.

Dose prescription
There is no consensus but the following are acceptable:
◆ Single doses of 8–​10-​Gy midplane dose.
◆ 20-​Gy midplane dose in five fractions treating daily.
◆ 30-​Gy midplane dose in 10 fractions treating daily.

22.2.5  Spinal canal compression

Spinal canal compression typically presents as an emergency. Wherever possible, how-
ever, patients should have fields set up with a treatment simulator to ensure accuracy
and there should be access to diagnostic MR scans to define the levels of involvement.
514 Radiotherapy planning for metastatic disease

Because of the emergency nature of this condition stereotactic techniques, which

require more extensive planning procedures, have not been widely advocated even in
the oligometastatic setting.

Treatment volume
This should include the site of spinal canal compression and one vertebral body in
craniao-​caudal direction above and below the site of compression. If a patient is being
treated on clinical diagnosis and plain X-​ray evidence of bone metastasis then two
vertebral bodies above and below the anticipated site of compression should be used
to define the volume pending accurate diagnosis with an MR scan. Attention to the
transverse axial imaging is important to ensure that any lateral or paravertebral exten-
sion is covered in the volume width.
It should be remembered that 25–​30% of patients will have multiple sites of com-
pression and a full spine MRI is recommended(8). All areas should be treated at the
same time by using more than one field if necessary.

Dose distribution
Treatment is given with a direct photon beam of 4–​6 MV.

Dose prescription
Standard doses include the following:
◆ 8–​10-​Gy single  dose.
◆ 20 Gy in five fractions.
◆ 30 Gy in 10 fractions.
◆ Special circumstances may warrant other fractionation schedules, for example:
o Solitary plasmacytoma: 40–​50 Gy in 20–​25 fractions(9).
o Lymphoma: primary radiotherapy for chemoresistant or low-​grade lymphoma
and post-​chemotherapy radiotherapy for high-​grade lymphoma and Hodgkin
lymphoma: 30 Gy in 15 fractions. (See Chapter 16.)
o Recurrent spinal canal compression after previous treatment: the risk of myel-
opathy must be balanced against the need for retreatment and the patient’s likely
prognosis. A cumulative biologically effective dose (BED) less than 120 using
an α/​β ratio of 2 is acceptable with a gap between treatments of several months.
Thus after 20 Gy in five fractions (BED2  =  60), a further 20 Gy in fractions
may be considered or a single dose of 8 Gy (BED2 = 40)or10 Gy (BED2 = 60).
After 30 Gy in 10 fractions (BED2 = 75) a further dose of 20 Gy in 10 fractions
(BED2 = 40) would be acceptable. Higher doses may be feasible with stereotactic
radiotherapy (see later).

Dose prescription point
This should be defined at the depth of the anterior spinal canal; this will vary along the
length of the cord and ideally should be measured for each patient from the available
MR imaging or on lateral X-​ray screening. An alternative is to use Table 22.2 based on
published data(10).
 Brain metastasis 515

Table 22.2  Dose description point

Vertebral level Depth of anterior canal
C6 7 cm

T2 7 cm
T5 6 cm
T8 5 cm
T11 6 cm
L1 7 cm
L3 8 cm
L5 8 cm

Source: data from Barton R, Robinson G, Gutierrez E, Kirkbride

P, McLean M. ‘Palliative radiation for vertebral metastases: the
effect of variation in prescription parameters on the dose received
at depth.’ International Journal of Radiation Oncology, Biology,
Physics 2002; volume 52: pp.1083–​91.

22.3 Brain metastasis
Brain metastases require whole brain radiotherapy. In selected cases with a localized
solitary metastasis and good performance status where surgery is not possible then
radiosurgery should be considered. This section will describe the technique for pallia-
tive whole brain radiotherapy.

22.3.1 Patient position
The patient should be supine with the neck straight on a standard headrest.
For palliative whole brain radiotherapy a head shell may be used, otherwise im-
mobilization using a band across the forehead or sandbags to help support the head
laterally can be used.

22.3.2 Treatment volume
Volume should include the whole brain including the olfactory groove and middle
cranial fossa. Conventionally the treatment field is defined as follows:
◆ Inferior border by a line drawn from the supra-​orbital ridge through the external
auditory meatus, resulting in the baseline as shown in Fig. 22.4, achieved by using
appropriate head twist.
◆ Other borders to cover the scalp and a small margin of 5–​10 mm outside the con-
tours of the scalp to allow for patient movement.
In practice the patient should, wherever possible, be simulated to ensure that this base-
line covers the full extent of the middle cranial fossa and adjustment may be required.

22.3.3 Dose distribution
Lateral opposed fields are used. An isocentric technique is most convenient and quick
for treatment delivery although fixed FSD techniques are acceptable.
516 Radiotherapy planning for metastatic disease

Fig. 22.4  Typical field to

treat whole brain for brain
metastases.

22.3.4 Dose prescription
The following doses are in common use(11):
◆ 12 Gy in two fractions.
◆ 20 Gy in five fractions.
◆ 30 Gy in 10 fractions.
The dose is prescribed to the midplane.
Doses for stereotactic radiotherapy (radiosurgery) range from a single dose of 20–​24
Gy for volumes < 2 cm diameter to three fractions of 8–​10 Gy for larger volumes(12).

22.4 Liver metastasis
Liver metastases are rarely treated with external beam radiotherapy but it may have a
useful palliative action where the liver is large and painful from expanding metastasis
not controlled by systemic treatment(13).
There is also increasing interest in the treatment of solitary or oligometastases in the
liver with stereotactic radiotherapy when surgical resection is not possible.

22.4.1  Treatment volume and definition

The extent of metastasis will be best defined on MR which is mandatory if ablative
treatment for oligometastases is considered. Ultrasound or CT can be used in the set-
ting of multiple metastases scan and the clinical size of the liver is a useful guide to
treatment planning.

22.4.2  Patient position and mobilization

The patient should be treated supine with immobilization using ankle stocks or a
vacuum bag if complex stereotactic techniques are to be used.

22.4.3 Volume localization
For stereotactic treatment then image registration of MR and CT is essential to accur-
ately define the extent of the metastasis. The CTV will be defined by the lesion and a
 Choroidal metastasis 517

2–​3 mm CTV to PTV expansion is usual depending upon the stereotactic platform
and in particular the technique used for respiratory compensation.
For palliative situations CT simulation should be used where the aim is to exclude as
much uninvolved liver as possible to minimize toxicity.

22.4.4 Dose distribution
For oligometastases this will depend upon the stereotactic platform to be used (see
Chapter 3).
In palliative treatment typically anterior and posterior parallel-​opposed fields will
be adequate using an isocentric technique. It is important to note where fractionated
treatment is given that such patients may have ascites and variable abdominal girths
from day to day for which adjustments may be needed.

22.4.5 Dose prescription
The following doses may be used:
Ablative oligometastases
◆ 45 Gy in three fractions
Palliative doses (prescribed to the midplane)
◆ 8-​Gy single dose.
◆ 20–​30 Gy in 10–​20 fractions.

22.5 Choroidal metastasis
Metastases to the choroid and retina are relatively rare but can cause catastrophic loss
of sight. They are seen most commonly in breast and lung cancer and 20% may be bi-
lateral(14). Early diagnosis by clinical examination of the eye with a fundoscope or slit
lamp in any patient presenting with visual disturbance against a background of estab-
lished malignancy is essential to retain vision. Urgent radiotherapy is indicated once
the diagnosis is confirmed.

22.5.1  Treatment position and immobilization

Supine, neck straight.
Although treated as an emergency, immobilization is recommended ideally with an
Orfit-​type shell which can be made rapidly and will not delay treatment.

22.5.2  Volume localization and definition

The CTV includes the entire choroid and retina of the involved eye. A CT or MRI scan
is of value to identify bulky disease or coincident brain metastases. A 4 × 4 cm 4–​6-​MV
photon field with the anterior border at the external canthus of the affected eye should
be used as shown in Fig. 22.5. To prevent irradiation of the contralateral lens by the
exit beam one of the following should be employed:
◆ The beam is angled posteriorly 3° to ‘take off ’ the divergence. Using CT simulation
the precise position of the lens can be identified and the beam placed behind it.
518 Radiotherapy planning for metastatic disease

Fig. 22.5 Beam arrangement to treat choroidal metastases in the left eye, using

a single lateral field with the anterior border at the ipsilateral outer canthus. Note
that if divergence is not removed as in-​-​-​the beam will exit through the lens of the
contralateral eye; an asymmetric beam placing the central axis at the outer canthus,
shown as—​or a beam angled posteriorly 3°, should be employed to exit behind the lens
and cornea of the contralateral eye.

◆ An asymmetric field is used so that the field centre is at the outer canthus, the an-
terior half of the field is then closed so that there is no divergence at the canthus and
the posterior half of the field is exposed.

22.5.3 Dose prescription
The following doses may be used:
◆ 20 Gy in five fractions.
◆ 30 Gy in 10 fractions.
The dose is prescribed to a depth of 2.5 cm.

22.5.4 Verification
A kV image should be taken and in vivo dosimetry with thermoluminescent dosim-
etry or diodes should be undertaken for the first or second fraction to ensure that the
exit beam is behind the contralateral eye.

References
1. Cox BW, Spratt DE, Lovelock M, et al. International Spine Radiosurgery Consortium
Consensus Guidelines for target volume definition in spinal stereotactic radiosurgery.
International Journal of Radiation Oncology, Biology, Physics, 2012; 83:e597ee605.
2. Bhattacharya IS, Hoskin PJ. Stereotactic body radiotherapy for spinal and bone metastases.
Clinical Oncology (Royal College of Radiologists) 2015; 27:298–​306.
3. Chow E, Harris K, Fan G, et al. Palliative radiotherapy trials for bone metastases: a
systematic review. Journal of Clinical Oncology 2007; 25:1423–​36.
 References 519

4. Lutz S, Berk L, Chang E, et al. American Society for Radiation Oncology (ASTRO).
Palliative radiotherapy for bone metastases: an ASTRO evidence-​based guideline.
International Journal of Radiation Oncology, Biology, Physics 2011; 79: 965–​76.
5. Chow E, van der Linden YM, Roos D, et al. Single versus multiple fractions of repeat
radiation for painful bone metastases: a randomised, controlled, non-​inferiority trial.
Lancet Oncology 2014; 15:164–​71.
6. Salazar OM, Rubin P, Hendricksen F, et al. Single dose hemibody irradiation in palliation
of multiple bone metastases from solid tumours. Cancer 1986; 58:29–​36.
7. Salazar OM, Sandhu T, da Motta NW, et al. Fractionated half-​body irradiation (HBI) for
the rapid palliation of widespread, symptomatic, metastatic bone disease: a randomized
Phase III trial of the International Atomic Energy Agency (IAEA). International Journal of
Radiation Oncology, Biology, Physics 2001; 50:765–​75.
8. Prewett S, Venkitaraman R. Metastatic spinal cord compression: Review of the evidence
for a radiotherapy dose fractionation schedule. Clinical Oncology 2010; 22:222–​30.
9. Soutar R, Lucraft H, Jackson G, et al. Guidelines on the diagnosis and management
of solitary plasmacytoma of bone and solitary extramedullary plasmacytoma. Clinical
Oncology 2004; 16:405–​13.
10. Barton R, Robinson G, Gutierrez E, et al. Palliative radiation for vertebral metastases: the
effect of variation in prescription parameters on the dose received at depth. International
Journal of Radiation Oncology, Biology, Physics 2002; 52:1083–​91.
11. Gaspar LE, Mehta MP, Patchell RA, et al. The role of whole brain radiation therapy in the
management of newly diagnosed brain metastases: a systematic review and evidence-​based
clinical practice guideline. Journal of Neuro-​Oncology 2010; 96:17–​32.
12. Wiggenraad R, Verbeek-​de-​kanter A, Kal HB, et al. Dose-​effect relation in stereotactic
radiotherapy for brain metastases. Radiotherapy and Oncology 2011; 98:292–​7.
13. Hoskin PJ. Radiotherapy in symptom management. In Hanks GW, Cherny N, Christakis
N, Fallon M, Kaasa S, Portenoy R (eds). Oxford Textbook of Palliative Care (4th edn).
Oxford: Oxford University Press, 2010, pp. 526–​47.
14. Wiegel T, Bottke D, Kreusel K-​M, et al. External beam radiotherapy of choroidal
metastases—​final results of a prospective study of the German Cancer Society (ARO 95–​
08). Radiotherapy and Oncology 2002; 64:13–​18.
 Chapter 23

Quality assurance in radiotherapy

Patricia Díez and Edwin GA Aird

23.1 Introduction
The World Health Organization defines quality assurance (QA) for radiotherapy as
‘all those procedures that ensure consistency of the medical prescription and the safe
fulfilment of that prescription as regards dose to the target volume, together with
minimal dose to normal tissue, minimal exposure of personnel and adequate patient
monitoring aimed at determining the end result of treatment’. General standards for
QA have been set by the quality management system accredited by the International
Organization for Standardization ISO 9001: 2015(1). A standard of QA specifically for
radiotherapy, that has been used in the UK, is QART(2).
QA in radiotherapy is essential for setting standards of safety and accuracy to ensure
the outcome for the patient is optimal. QA should cover all aspects of the process of
planning and delivery of the treatment. It may form part of a quality system, which
will also encompass quality control (QC) and audit. This chapter will mainly cover QA
of the patient pathway, concentrating on QC of treatment planning and delivery using
linear accelerator (linac) technology. QA for other techniques such as robotic radio-
therapy, tomotherapy, and proton beam therapy will be mentioned but not discussed
in detail. There will also be a brief outline on relevant legislation associated with the
radiotherapy process as well as a section on QA for clinical trials.

23.2  The quality system

In the UK, any cancer centre must comply with the Department of Health Cancer
Standards(2). These are very extensive standards against which centres can be tested
covering all aspects of cancer care and treatment and may be defined in terms such
as accuracy and waiting times. Although these criteria will be defined locally, best
practice is also set by professional colleges (such as the Royal College of Radiologists,
RCR), national standards (such as Cancer Standards in the UK), and clinical trials QA.
Legislation also plays a part in setting these standards for quality.
Most of these objectives may be fulfilled by developing a robust quality system and
are described in McKenzie et al(3). QA operates within the quality system using a set
of procedures written and maintained by a multidisciplinary quality committee within
the cancer centre. This group states the standard that it expects to achieve within its
quality policy. In this document management state their commitment to quality and
the aims which their particular centre hopes to achieve. These aims will incorporate
 Quality control for treatment planning 521

the following principles: safe and accurate treatment using proven and validated tech-
niques; clear communication to the patient about the treatment options available on
a continuing basis, including up-​to-​date information leaflets; a system for patients
to comment on the care they have received. Procedures, including management
procedures and work instructions, will then be put in place to carry out the quality
policy’s aims.
The quality committee will monitor and assess incidents (including near-​misses),
non-​conformances, complaints, and reports from various subgroups which may be
working on specific problems or in specific areas.
Finally, to ensure that any system is operating effectively, it is essential to set up
audit, both internal and external. The distinct addition to this in the UK is the Peer
Review system.

23.3 Legislation
Although not usually part of a quality system, it is important that the reader is aware
of ionizing radiation legislation and how it applies to radiotherapy. The current
European Directives(4), produced by the Council of the European Union, outline the
general principles ‘on health protection of individuals against the dangers of ionizing
radiation’.
In the UK, all medical exposures to ionizing radiation are currently governed by the
Ionising Radiations Regulations (IRR) of 2017(5), and the Approved Code of Practice(6).
These regulations are the basis for radiation protection in the UK and are enforceable
under the Health and Safety at Work Act (1974)(7). Those based upon Directive 2013/​
59/​Euratom(4) are the Ionising Radiations Medical Exposure Regulations (IRMER)(8),
for which the RCR has written a very useful guide(9).
The purpose, scope, and definitions outlined in both the European and the UK le-
gislation are almost identical and identify the following key issues: justification of the
medical exposure for the patient; optimization of all imaging and radiotherapy expos-
ures; responsibilities (of the employer, referrer, practitioner, and operator); procedures
to be undertaken (written protocols, work instructions, clinical guidelines, and QA
programmes and their implementation); staff training; equipment safety (radiotherapy
equipment should comply with all sections of BS EN 60601(10)); special practices for
exposures of children, comforters and carers, volunteers, pregnant or breastfeeding
staff, and for health screening programmes.

23.4  Quality control for treatment planning

Treatment planning (TP) encompasses the whole journey from patient immobiliza-
tion and positioning, through treatment prescription to plan verification. As a result,
the process involves staff from various disciplines and requires extensive and complex
QC. The physicist designs and implements the QA programme, which involves several
steps, e.g. generating treatment machine data for input into the treatment planning
system (TPS), determining the QC tests to be performed, their tolerances and fre-
quency, understanding and responding to any discrepancies, and designing plans that
can be used regularly to check the complete system.
522 Quality assurance in radiotherapy

Target localization is carried out by the radiation oncologist and should follow the
standards defined in ICRU Reports 50, 62, and 83 (11-​13). Close collaboration with a
radiologist specialized in clinical oncology for interpretation of computed tomography
(CT), positron emission tomography-​CT (PET-​CT), and magnetic resonance (MR) im-
ages may prove useful for clinicians to delineate the gross tumour volume accurately.
The radiographer/​radiation therapist (RTT) is involved in various aspects of the TP pro-
cess: immobilization, simulation/​localization, plan design, plan verification, detecting
equipment deviations and malfunctions, understanding the safe operating limits of the
equipment, judging when errors in TP have occurred, and helping with QC. Finally, the
dosimetrist has the role of patient data acquisition, plan design, and assistance with QC.
The TPS itself is subject to rigorous commissioning and ongoing QC checks.
TPSs are becoming more complex and sophisticated, particularly with the advent of
intensity-​modulated radiotherapy (IMRT) and linac-​based rotational therapy (other-
wise known as VMAT: volumetric modulated arc therapy). For this reason, a compre-
hensive set of QA guidelines are necessary and have been provided(14-​16). A patient will
be planned at commissioning and stored as the gold standard. Monthly or quarterly
QC will require re-​planning the same patient and comparing to the gold standard plan
for any changes in absolute dose and dose distribution. However, all the functions,
algorithms, and pathways embedded in the TPS need testing. QC of TP also needs to
examine other equipment and processes besides just the TPS. Patient planning will
involve the gathering of patient data from various sources (e.g. CT, PET-​CT, MRI);
all these, as well as methods of data transfer, need to be tested at different frequencies,
depending on the specific centre (see also section 23.7). The individual patient’s plan
also needs checking by examining volumes, field sizes, and dose calculations during
routine clinical practice. The monitor unit (MU) calculation must be independently
verified by a completely separate system.
The QA programme must focus on the planning process as a whole and assess the
cumulative effects of any uncertainties. ICRU Reports 50(11) and 62(12) require a max-
imum variation in dose of −5% to +7% from the prescription point. To be able to
achieve this, uncertainties in all steps that make up planning must be much smaller
than this. TP can introduce systematic errors that are then carried through to treat-
ment delivery.

23.4.1 Acceptance testing
Acceptance testing should be carried out once a system has been installed but be-
fore it is used clinically. The first tests in QC are performed during acceptance of the
equipment to ascertain that the specifications stated by the manufacturer are satisfied.
These must be reasonable constraints and, where appropriate, specifications should be
measurable, with a stated tolerance. Tests include checks for CT input, anatomical de-
scription, photon and electron beam dose calculations, dose display, evaluation tools,
and hardcopy output.

23.4.2 Commissioning
Both dosimetric and non-​dosimetric commissioning will be described here. For the
latter, QA starts with evaluation of immobilization techniques and equipment. This
 Quality control for treatment planning 523

is required for patient positioning reproducibility and to help the patient remain still
throughout treatment. Positioning and simulation, used to localize the tumour volume
and critical structures, follows. These are generally defined from CT images and refer-
ence marks tattooed on the patient are used to aid localization. Accuracy is crucial at
this point, as all further planning will depend upon it. Hence, CT scanners and virtual
simulators must be subject to rigorous QC, including geometrical accuracy of beam
and couch parameters, as well as laser alignment.
Generally, cross-​sectional CT images are used to define patient anatomy, however
they may also be acquired from other systems too, such as MRI or PET-​CT. QC must
ensure that image acquisition is optimal, and that their transfer to the TPS is accurate.
Some parameters that should be checked during the planning procedure are pixel size,
slice thickness, CT numbers, partial volume effects, artefacts, distortion, use of radio-​
opaque markers, coordinate system of reference, breathing instructions, use of con-
trast agents, and use of immobilization devices. These should match the requirements
for the individual patient. Work instructions should be in place to guarantee correct
working practices.
Even with such advanced imaging systems, correct identification of tumour, target,
or critical organs still remains a critical part of the process. Part of QA is to ensure
that image conversion, input, and registration are accurate. These can be assessed by
testing for image geometry, geometric localization, and orientation of the scan, text
information, transfer imaging data, and image warping. Phantoms can be used to test
CT image transfer and 3D reconstruction. Anatomical structures need checking for
electron density definition and representation, display characteristics, auto-​contouring
parameters, use of structures created from contours, volumes constructed by expan-
sion/​contraction algorithms, structures constructed from non-​axial contours, bolus,
and editing the 3D density distribution, image use and display, and dataset regis-
tration. The AAPM have formed Task Group 132(17) to address and report on image
registration QA. Non-​dosimetric checks of the dose calculation algorithm and density
corrections must also be performed, including testing for regions to be calculated, cal-
culation grid definition, accuracy of density corrections, and appropriate calculation
algorithm selection. Plan evaluation tools including dose display, dose–​volume histo-
grams (DVHs), radiobiological tools, and composite plan dose distributions should
also be assessed. The hardcopy output of all these features must also be investigated.
Finally, checks for plan implementation and verification after planning has been com-
pleted and approved will be carried out, assessing correct use of coordinate systems,
scale conventions, and data transfer.
Dose calculation commissioning involves measurement of a self-​consistent dataset,
input data checks, algorithm verification, applicability and limits of the dose calculation
algorithm, and dose verification applied to the clinical use of the system. Measurement of
a self-​consistent dataset is fundamental to any TPS. This will require information of depth
doses and beam profiles for different field sizes (plain and wedged) and energies to be in-
tegrated into a self-​consistent dataset to be appropriately analysed, handled, and stored.
The data will then need transferring into the TPS. Data transfer will require verification.
Beam model parameters will directly affect the accuracy of dose calculations, as
they will be used to fit the measured data. Determination of these parameters has
524 Quality assurance in radiotherapy

to be precise. Dose calculation verification tests need to be performed to compare

calculated with measured dose distributions. This entails comparisons of 1D lines,
2D isodose lines and colourwash dose displays, DVH analysis, and distance maps.
Required and/​or achievable accuracy needs determining and testing before photon
and electron calculation verification is performed. Calculation verification is then
made in terms of algorithm performance and clinical acceptability. Plan normal-
ization is one of the key elements of the TPS as it determines how MUs are calcu-
lated. This will, in turn, decide the absolute dose output. Verification is, therefore,
essential and should check for beam weights, isodose levels or points used for dose
prescription, correct dosage and fractionation, and MU calculation. Clinical plan
verification on phantoms is the final system check. Commissioning data and/​or
use of phantoms can be employed to confirm dose and MU results for a variety of
clinical cases.

23.4.3  Periodic quality control

Periodic testing of the planning process is necessary to confirm the integrity and se-
curity of the data files, verify correct and accurate functioning of peripheral devices,
check the reliability of the TPS software, and confirm adequate and accurate func-
tioning of any output devices.
Central to QC is the correct design and implementation of plans for individual pa-
tients and that all aspects of treatment planning and delivery undergo comprehensive
checking. To ensure this, periodic staff training is very important, as well as formally
reviewing clinical plans and the QA programme itself.

23.4.4  System management and security

Adequate system management must be part of the QA programme. Data security
is imperative, so responsibility must be assigned to both a TPS manager (an expert
physicist) and a computer systems manager. Computer management comprises
regular back-​up and storage of all files, use and maintenance of computer networks,
and system security. There should be limited access to all areas of the TPS, including
software and data.

23.5  Quality control for treatment delivery

As with treatment planning, acceptance testing, commissioning, periodic and ongoing
QC and system management and security need to be addressed.
In radiotherapy, any small deviation in any of the beam parameters, mechanical sys-
tems, or optical indicators can lead to an increased risk of geographical miss, therefore
an extensive and reliable QC programme is crucial. This programme is set out in a
series of detailed protocols that stipulate the procedures to be performed in thorough
periodic inspections, testing, and calibrations of all the therapy equipment. The proto-
cols also incorporate acceptable tolerances and guidelines for the action that must be
taken if equipment or a beam parameter should fail one of the tests outlined within
them. In addition to the protocols, there is also a set of work instructions that dictate
how and when the ion chambers and dosimeters used in the QC tests are to be tested
 Quality control for treatment delivery 525

and calibrated. The more rigorous QC routines, such as the post-​service checks, often
require independent measurements of output, flatness, and symmetry.

23.5.1  Acceptance testing for linear accelerators

The basic principle behind acceptance testing for treatment delivery is the same as for
planning: to verify that the system satisfies all specifications. Careful attention must go
into the planning and setting out of specifications that meet both the current and likely
future needs of the specific centre for the lifetime of the equipment. Treatment room
design (both for logistic and radiation protection purposes) must also be considered
at this stage.
Some machine requirements that need to be discussed and selected are energy,
electron beam capabilities, use of asymmetric jaws and dynamic wedges, multileaf col-
limators (MLCs) and on-​board imaging (including kV and cone-​beam CT, CBCT),
as well as computer control and information management systems. Ancillary equip-
ment is also available to the user for more specialized treatment delivery, such as
IMRT/​VMAT, stereotactic radiosurgery, image guided radiotherapy (IGRT), and
intraoperative linear accelerators. Any equipment used for treatment delivery must not
only perform with a high degree of mechanical accuracy and precision, but radiation
parameters (beam flatness, dose output, etc) must also meet the correct specifications.
Upon delivery of the equipment acceptance testing is performed to ensure it meets
the specifications outlined to the manufacturer. This includes mechanical tests, radi-
ation performance tests, safety checks, general tests, and radiation protection surveys.
Details on these tests can be found on IPEM Report 81(14) and Report 94(18), AAPM
TG 40(19) and TG 142(20), and the IAEA Radiation Oncology Physics Handbook for
Teachers and Students (21).

23.5.2 Commissioning
Once all aspects of the working of the unit are tested against the specifications given
to the manufacturer and the equipment is accepted, the physicist must acquire a self-​
consistent dataset for radiation performance characterization. These measurements
will be input into the TPS (as discussed in section 23.4), and form the baseline for
subsequent QC testing when the unit is regularly inspected to ensure it complies with
the geometric and dosimetric standards set at acceptance. The equipment used for
obtaining dosimetric data must also be subject to stringent acceptance tests and peri-
odic QC checks. This is beyond the scope of this chapter.
Both photon and electron beam dosimetry should be performed during commis-
sioning. Some measurements are essential, as specified by the basic dataset for the
given TPS; others will be for verification of isodose curves produced by the TPS.
Detailed accounts of commissioning can be found in Almond and Horton(22) and
Johansson et al.(23). A more complete dataset will be required if the equipment is to be
used for IMRT/​VMAT.

23.5.3  Periodic quality control

The frequency and extent of routine testing varies between centres, depending upon
the particular use of the equipment, local conditions, and methods and knowledge
526 Quality assurance in radiotherapy

of what is technically achievable. Basic requirements have been identified and de-
scribed(14,20). These should ensure that machine performance at commissioning is
maintained to guarantee accurate and safe treatment delivery. In summary:  output
and optical systems are checked daily, and beam flatness, energy, and field size are
checked weekly or monthly. MLCs, dynamic wedges, electronic portal imaging devices
(EPIDs), kV, and CBCT imaging require further checks, in particular as tighter mar-
gins on target volumes are now used for IMRT/​VMAT treatments.
There are separate QC protocols for daily, monthly, post-​service, annual, and bi-
ennial QA checks. These are all supplementary to the original acceptance testing and
commissioning performed on each linear accelerator.
Daily QC testing is generally performed by engineers and/​or radiographers/​RTTs
during the early morning run-​ups of the machines, when output, mechanical, and
optical systems are examined. This is a crucial time at which to check the safe oper-
ation of the linear accelerator. An interdisciplinary team made up of an engineer and
a physicist generally performs other, more comprehensive and time-​consuming tests,
usually on a monthly basis. These comprise independent, quantitative measurements
of the beam parameters in addition to further examinations of the mechanical and
optical systems.
Extended safety testing must also be performed: examining safety interlocks, elec-
trical safety, safety of auxiliary devices, and radiation protection, to be carried out after
corrective maintenance, as part of post-​service testing (three-​monthly). In addition
to the regular monthly and post-​service checks, further QC checks are performed on
an annual basis similar to those performed during commissioning of each machine.

23.5.4  In vivo dosimetry
The accuracy of the TP process and final dose delivery can be verified using in vivo
dosimetry performed using detectors such as thermoluminescent dosimeters, diodes
or EPIDs. Any differences that cannot be reconciled with an entrance or exit dose
requires further investigation. Measurements will highlight any inaccuracies in the
TPS dose-​calculation algorithm, treatment machine calibration, mechanical align-
ment and settings, and patient set-​up, movement, and internal anatomy variability.
Dosimetric verification is particularly important for critical sites such as lenses or
testes. It is vital to calibrate any of these devices regularly at the beam energy used for
patient treatments.

23.6  Quality control for intensity-​modulated

radiotherapy and rotational therapy techniques
There are several aspects to this which are similar to those for any complex treatment,
including detailed commissioning of the TPS, network testing, and linear accelerator
delivery checks. These are more critical for IMRT, particularly for dynamically de-
livered IMRT and VMAT, when small changes in performance of MLCs can radically
change the dose or dose distribution. In addition, VMAT has a continuously varying
beam aperture, gantry speed, and dose rate, which need to be tested. VMAT tests need
to verify that beam flatness and symmetry remain stable during gantry arcing and at
 Quality control for imrT and rotational therapy techniques 527

lower dose rates; to ensure leaf travel and gantry rotation are well synchronized; and to
check gantry position, leaf position, leaf speed, and cumulative dose are also correctly
synchronized.
Where the system is all provided by one manufacturer it is essential to follow the
advice given by the company. If the TPS and linear accelerator are from different com-
panies, it is important to make use of any consortium that may work with this equip-
ment combination and draw on the most recent publications as sources of advice for
its correct implementation and operation. This section focuses on conventional linac-​
based QC and will briefly discuss tomotherapy and robotic techniques in a separate
sub-​section (23.6.4).

23.6.1 Commissioning
As mentioned earlier, certain aspects of IMRT/​VMAT require more rigorous com-
missioning. These include penumbra modelling (which is very important as most
fields are a summation of many segments), small field dosimetry, heterogeneity cor-
rections, extended MLC modelling, and modelling of off-​axis fields. Additional TPS
requirements will therefore include leaf and jaw transmission measurements, leaf-​end
shape measurements, minimum MU per segment, minimum field size, and small field
output factors. A  more detailed document on commissioning tests can be seen in
AAPM TG 119(24).
MLC and jaw positioning tests are also very important. IMRT/​VMAT often uses
leaf over-​travel and there is a wider range of leaf positions used than in conformal
planning techniques. There is also a greater dependence of the delivered dose on the
calibration accuracy, especially for sliding window (SW) IMRT and VMAT, where the
dose is determined by the gap between leaves. Even in step-​and-​shoot (SS) small errors
in field size will have a large effect on the output factor for small fields. Details of these
tests can be found in IPEM Report 96(25). It is important to note that these tests will
be different for SS and SW/​VMAT. SS IMRT consists of multiple segments, many of
which have very low MUs, so it is important that dose linearity and beam profiles at
low MUs are also checked. For SW/​VMAT it is critical that the leaf gap is precisely
maintained and so leaf position reproducibility and leaf speed stability are to be tested,
as well as leaf acceleration and deceleration effects.
TPS capabilities can be tested through creating and assessing delivery of both simple
and complex test cases, then investigating the effects of leaf sequencing (from an ideal
fluence map generated by the TPS to a deliverable plan).

23.6.2  Periodic
quality control and intensity-​modulated
radiotherapy verification
Pre-​treatment verification
Pre-​treatment dose prescription verification for the individual patient is important
due to increased effects of penumbra and transmission, and because delivery is
more dependent on MLC calibration for IMRT than conformal techniques. The
main tests carried out are point doses in an appropriate phantom (absolute dose
measurement), fluence maps for individual fields (e.g. using EPIDs and 2-​D arrays),
528 Quality assurance in radiotherapy

and combined dose distributions of the whole plan on film or 3-​D arrays. Details of
measurement and analysis of these can be found in IPEM Reports 81 and 96(14,25),
ESTRO Booklet 9(26) and AAPM TG 218(27). After a centre has assessed their pro-
cedures through extensive individual patient QA, some of the checks can be re-
stricted to a subset of patients per machine. This will depend on each department’s
experience.

Treatment verification and in vivo dosimetry

Positioning and dose verification during treatment should be carried out on all pa-
tients following published recommendations(28); making use of EPID dosimetry(14,29)
and/​or CBCT verification.

Linear accelerator quality control

A total IMRT/​VMAT system check is required which should be carried out at least
annually and compared to the baseline set at commissioning. More frequent tests
include leaf position accuracy and a repeat of the technique-​specific commissioning
tests, on a monthly/​quarterly basis. For SS these will include dose linearity and beam
profile flatness and symmetry for low dose segments. For SW and VMAT, leaf pre-
cision and speed, and position reproducibility should be checked. A  detailed set
can be found in IPEM Report 81(14). EPIDs and linear arrays are now used exten-
sively for these tests. Other tools for QC in IMRT and rotational therapy have been
described(30,31).

23.6.3 External audit
External audit is extremely important for confirmation that the established IMRT/​
VMAT procedures are working as required. This can be done after commissioning
before going clinical to ensure the process is running smoothly. One of the easiest and
more complete ways of doing this is by taking part in an IMRT/​VMAT clinical trial
(see section 23.11) or, in the UK, through the Inter-​departmental Audit Group.

23.6.4  Non-​linac based rotational therapy

Helical tomotherapy shares similar QA to IMRT/​VMAT; however, the couch con-
stantly translates through a continuously rotating fan beam mounted on a slip-​ring
gantry. QC(14,32) therefore requires verification that the couch moves perpendicular to
the gantry plane and that it translates accurately and continuously; beam rotational
stability is maintained; rotational output, integral dose output, and beam profile
shapes for each field width are accurate; and, finally, couch and gantry synchrony are
preserved.
Robotic radiosurgery (CyberKnife®) delivers non-​coplanar, unmodulated circular
fields from a large number of source angles and positions. The nature of the equip-
ment and delivery system requires some differences in the QA programme. This has
been described in detail (14,33) and introduces new checks for collision avoidance and
verification that the radiation field centroid matches the central axis laser. The import-
ance of positional accuracy, imaging geometry and function, and QC of the tracking
system are highlighted and tests described. Dose calibrations also differ from standard
 Stereotactic radiosurgery and stereotactic body radiotherapy 529

linear accelerators due to the non-​isocentric circular fields used. End-​to-​end testing is
particularly important since this technology relies on automatic repositioning of the
patient using frequent X-​ray imaging.

23.7  Quality control for image-​guided radiotherapy

Radiotherapy techniques cannot be successfully delivered without the use of image
guidance. IGRT ranges from simple kV/​MV portal dosimetry for set-​up and CBCT,
to organ motion tracking; automatic patient positioning, and use of EPID for in-
dividual patient plan verification. Appropriate QA must be carried out on both
pre-​treatment, as described in section 23.4, and on-​treatment imaging. The range
of IGRT systems available is extensive therefore the reader should refer to IPEM
Report 81(14) for guidance in producing a good QA programme specific to each
department.

23.8  Stereotactic radiosurgery and stereotactic

body radiotherapy
Stereotactic radiosurgery (SRS) and stereotactic body radiation therapy (SBRT), also
known as stereotactic ablative radiation therapy (SABR), are rapidly becoming ac-
cepted practice for the radiation therapy of certain tumours. The difference between
these therapies and conventional radiotherapy is that they require geometrical ac-
curacy of about 1 mm and use very large doses per fraction (of up to 30 Gy). Intensive
imaging is vital when delivering radiotherapy with either of these modalities(34). Please
refer to Solberg et al.(35) for quality and safety considerations.
The report of AAPM TG 101(36) outlines best practice guidelines for SBRT. This in-
cludes:  simulation and treatment planning; immobilization, target localization, and
delivery. The report deals with the problems associated with measurements within
small fields. Other useful references include IPEM Reports 94 and 103(18,37), and
Williamson et al(38). Rowshanfarzad(39) emphasizes the need to determine the isocentre
(defined with a tolerance of 1 mm) for the delivery system accurately and to ensure
that this accuracy is maintained using routine QC; a specific test to check this should
be performed before each treatment(40). QA of immobilization and target localization
systems(41) is also a very important part of the process.
In the UK, SABR has been adopted through the ‘Commissioning through Evaluation’
(CtE) programme, which completed in 2018. In addition, the ‘SABR UK Consortium’
Guidelines(42) provide guidance and standards for radiotherapy centres setting up pro-
cedures for SBRT. In particular, they recommend external audit through clinical trials
(see section 23.11) and that individual patient specific QA measurements are per-
formed for at least the first ten patients. For specific QA recommendations for intra-
cranial stereotactic radiotherapy, see NCS 2015(43). For recommended frequencies of
QC see AAPM TG 135(33) and TG 142(20).
Most SRS and SBRT treatments on linear accelerators are now performed with
flattening filter free (FFF) fields(14). These need a slightly different QA approach to
conventional fields because of their high dose rate and bell-​shaped profiles. High dose
530 Quality assurance in radiotherapy

rates require that ion chamber recombination corrections are well known throughout
the dose/​pulse range used; profiles require checks against a baseline at several critical
points along the profile (with a tolerance of 1%). It is also important to understand
which detectors to use for small fields and phantoms for end-​to-​end studies. End-​to-​
end QA is essential for characterizing cumulative system accuracy for any given pro-
cedure, as it can be significant(40).

23.9  MR linear accelerator QA

The QA for radiotherapy treatment using an MR linear accelerator covers many
functions, but it is image distortion and dose perturbation that need integrating
particularly for the effects of MR during commissioning of the linear accelerator
and imaging system. The magnetic field produces distortion in the image, which
needs to be assessed and corrected for where necessary. When considering radi-
ation field measurements, it is vital to recognize that the magnetic field also changes
the pattern of dose within the body. Dose perturbation effects include reduced
build-​up; laterally shifted asymmetric penumbra; and the electron return effect.
These can produce hot and cold spots in the dose distribution, which must be mod-
elled correctly in the TPS, but can be mitigated by using multiple fields, such as
IMRT/​VMAT techniques. A comprehensive document on QA is not yet available
but an international consortium(44) has been formed that will be addressing this and
many other aspects of MR-​linac technology.

23.10  Proton beam therapy QA

Many aspects of QA for proton beam therapy (PBT) are similar to those for photons
(and many authors suggest using TG-​142(20) and TG-​54(45) as a guide to setting
standards); however, there are important differences, in particular the length of the
spread-​out Bragg peak needs to be controlled very accurately and therefore must
be included in routine QA. Also, many of the tools used for photon measurements
may not work accurately—​or need more correction factors—​when used for proton
measurements.
The readers can refer to PowerPoint presentations(46,47) from AAPM meetings for
further general information on PBT and its QA. Other useful publications include
the ACR-​AAPM Technical Standard for the performance of proton beam radiation
therapy 2018(48) and Arjomandy et al.(49).

23.11  Quality assurance in clinical trials

It is highly desirable that a cancer centre takes part in clinical trials. This involvement
ensures that each department remains up to date with best practice.
There is now good evidence from studies around the world demonstrating the value
of QA in multi-​centre clinical trials. They not only improve protocol compliance but
ensure outcomes show differences in the treatment arms rather than deviations from
trial protocol. This is particularly true for multi-​centre international trials where lo-
cally developed treatment planning and delivery protocols differ from those outlined
 Quality assurance in clinical trials 531

in the trial protocol. Meta-​analysis(50) has demonstrated worsened patient clinical out-
comes after departures from trial protocol were identified through the QA programme.
The TROG 02.02 trial(51) for advanced head and neck cancer also showed correlation
between non-​compliance and clinical outcome, where there was a 20% decrease in
survival, in both arms of the trial, in those patients planned and/​or treated outside
the trial protocol. This recent evidence has mainly been associated with outlining and
planning of both the target and critical structures.
QA, as part of dosimetric and other comparisons between centres, was first proposed
by Johansson in 1988(52). The main aims initially were dosimetry intercomparison and
resources for planning and delivery of radiotherapy generally.
In the UK, dosimetric inter-​comparisons were started in the late 1980s(53) and
the wider concept of QA in clinical trials was taken up with QA(54) in the CHART
(Continuous Hyperfractionated Accelerated Radiotherapy) clinical trial for bronchus
and head and neck cancers. Trials QA then progressed through several stages with the
START (Standardisation of Breast Radiotherapy) trial and RT01 (a conformal prostate
trial). Protocols for trial QA were developed along different lines, with START pla-
cing more emphasis on planning(55) and in vivo dosimetry(56) and RT01 introducing
a process document to be written by each centre. Both programmes included the re-
quirement to measure in anthropomorphic phantoms(57,58), with much less emphasis
on checking equipment performance.
The National Radiotherapy Trials QA, or RTTQA, Group formed in 2002 under the
auspices of the National Cancer Research Institute (NCRI), UK. Currently the RTTQA
Group carries out QA for all NIHR (National Institute for Health Research) Clinical
Research Network (CRN) portfolio trials that include a radiotherapy component. The
RTTQA team was set up to ensure all patients within a radiotherapy trial are treated
according to a trial protocol and to nationally accepted standards. This serves to min-
imize variations ensuring clinical trial outcomes reflect differences in randomization
schedules rather than departures from the trial protocol. In doing so, best clinical prac-
tice is adopted by all participating centres, raising radiotherapy standards across the
country. The RTTQA Group also provide QA support to the NCRI CTRad (Clinical
and Translational Radiotherapy), established in 2009 to promote and support radio-
therapy and radiobiology research through clinical trials.
All information associated with the QA programmes currently in place and what they
involve can be found on rttrialsqa.org.uk. All UK centres now have well-​established
QA procedures for all equipment and there is a voluntary system of dosimetry audit,
so QA in clinical trials now consists of the following elements which are tailored to fit
each individual trial:

◆ Verification of electronic data transfer.

◆ A process document(59) to be written by the participating centre, using a template
provided by the chief investigator or QA centre, which describes the procedure that
the centre will follow for planning and delivery of radiotherapy according to the
trial protocol.
◆ A  facility questionnaire(60,61) to be completed by participating centres to demon-
strate they have the appropriate resources and have developed a process to deliver
532 Quality assurance in radiotherapy

the radiotherapy prescription required by the trial protocol. These include trial-​
specific techniques and procedures as well as information on planning and treat-
ment equipment and audits at the centre.
◆ Outlining benchmark cases(62,63):  one or more test cases are sent to participating
centres for outlining to check that the clinical oncologists understand the trial
protocol and meet standards of volume outlining defined within it before entering
patients into the trial.
◆ Case evaluations: clinical case scenarios are sent to participating centres to identify
what and how clinicians would treat them for protocol compliance as part of the pre-​
trial QA process. This is introduced for multi-​site malignancies where an outlining
benchmark case for each anatomical site that may be involved is unfeasible
◆ Planning benchmark cases(64):  one or more pre-​outlined test cases are sent to
participating centres for planning to check that the team understand the trial
protocol.
◆ Audit(65,66): the QA team visits participating centres to audit records of planning and
delivery of radiotherapy for individual patients; check verification images; perform
QC on treatment machines, as well as dosimetry checks in anthropomorphic and
semi-​anthropomorphic phantoms, including end-​to-​end audits.
◆ Central review of trial patient outlines and plans can be carried out by the QA
centre either prospectively or retrospectively.
As techniques for delivery of radiotherapy become more advanced so the QA in
clinical trials becomes more complex. In particular, IMRT/​VMAT and SABR require
greater input from the QA team.
As trials became more complex, it has become evident that the target (and critical
structure) volume outlining for clinicians is very important for meeting the standard
set in the trial protocol. This has resulted in extra training and checking which in turn
has not only improved standards for trials but has also led to changes in practice for
patients treated outside clinical trials. Studies of variation in volumes in benchmark
cases have been made. The iterative training procedures have shown improved con-
formity of volumes (67). Careful review of plans has been shown to be particularly im-
portant in dose escalation studies and, following pre-​accrual QA, a reduction in the
number of deviations has been demonstrated.
In Europe, it is the EORTC (European Organisation for the Research and Treatment
of Cancer) that has established a QA programme for clinical trials. TROG Cancer
Research (Trans-​Tasman Radiation Oncology Group) operates in Australia and New
Zealand. Similarly, QA programmes are organized in the USA by the Imaging and
Radiation Oncology Core (IROC) at MD Anderson, in Japan by JCOG (Japan Clinical
Oncology Group) and in Canada by the Canadian Cancer Trials Group (CCTG). All
the above are members of the Global Harmonisation Group, who have developed
standardization of QA naming conventions for QA processes and a guide for defining
minor and major trial deviations. Work is underway to agree minimum standards that
should be included in dosimetry audit reports to enable intergroup acceptance of such
reports and hence reduce repetition of QA for those centres that recruit to both UK and
international trials.
 References 533

References
1. International Organisation for Standardization. Quality Systems: Model for Quality
Assurance in Design, Development, Production and Servicing. ISO 9001:2015.
Geneva: ISO, 2015.
2. Department of Health. Quality Assurance in Radiotherapy (QART) Standard.
London: Department of Health, 1991.
3. McKenzie AL, Kehoe TM, Thwaites DI. Quality assurance in radiotherapy physics. In
Williams JR, Thwaites DI (eds) Radiotherapy Physics in Practice (2nd edn.). Oxford: Oxford
University Press, 2000, pp. 316–​27.
4. European Council Directive. Council Directive BSS 96/​29/​2013/​59/​EURATOM 2013 on
health protection of individuals against the dangers of ionizing radiations in relation to
medical exposure.
5. HSE (Health and Safety Commission). The Ionising Radiations Regulations 2017.
London: HMSO, 2017
6. HSE (Health and Safety Commission). The Ionising Radiations Regulations 2017.
Approved Code of Practice and Supporting Guidance. London: HMSO, 2018.
7. HSWA. Health and Safety at Work etc. Act 1974. London: HMSO, 1974.
8. HSE (Health and Safety Commission). The Ionising Radiation (Medical Exposure)
Regulations. London: HMSO, 2018.
9. RCR, SCR, IPEM. A guide to understanding the implications of the ionising radiation
(medical exposure) regulations in radiotherapy. London: The Royal College of
Radiologists, 2008.
10. British Standards Institute. BS EN 60601-​1-​4: Medical electrical equipment. General
requirements for safety. Collateral standard. General requirements for programmable
electrical medical systems. Milton Keynes: BSI, 1997.
11. ICRU (International Commission on Radiological Units and Measurements).
Prescribing, Recording, Reporting Photon Beam Therapy. ICRU Report 50 Bethesda,
MD: ICRU 1993.
12. ICRU (International Commission on Radiological Units and Measurements).
Prescribing, Recording, Reporting Photon Beam Therapy. ICRU Report 62 Bethesda ,
MD: ICRU 1999.
13. ICRU (International Commission on Radiological Units and Measurements).
Prescribing, Recording, Reporting Intensity Modulated Photon Beam Therapy. ICRU 83
Bethesda, MD: ICRU 2010.
14. IPEM (Institute of Physics and Engineering in Medicine). Physics Aspects of Quality
Control in Radiotherapy. IPEM Report 81, 2nd Edition. IPEM, York, UK. 2018.
15. Fraas B, Doppke K, Hunt M, et al. AAPM Task Group 53. Quality assurance for clinical
radiotherapy treatment planning. Medical Physics 1998; 25:1773–​829.
16. International Atomic Energy Agency. IAEA-​TECDOC-​1583. Commissioning of
Radiotherapy Treatment Planning Systems: testing for Typical External Beam treatment
techniques. Vienna: IAEA, 2008.
17. Brock KK, Mutic S, McNutt TR, et al. Report No 132 Use of image registration and fusion
algorithms and techniques in radiotherapy: Report of the AAPM Radiation Therapy
Committee Task Group No 132. Medical Physics 2017; 44; e43—​76.
18. IPEM (Institute of Physics and Engineering in Medicine). Acceptance testing and
commissioning of linear accelerators. IPEM Report 94. IPEM, York, UK. 2006
534 Quality assurance in radiotherapy

19. Kutcher, GJ, Coia L, Gillin M, et al. Comprehensive QA for radiation oncology. Report of
AAPM Therapy Committee Task Group 40. Medical Physics 1994; 21:581–​618.
20. Klein EE, Hanley J, Bayouth J, et al. AAPM Task Group 142 report: Quality assurance of
medical accelerators. Medical Physics 2009; 36:4197–​216.
21. Thwaites DI, Mijnheer BJ, Mills JA. Quality assurance of external beam radiotherapy. In
IAEA, Radiation Oncology Physics: A Handbook for Teachers and Students. Vienna: IAEA,
2006, pp. 407–​50.
22. Almond PR, Horton JL. Planning and acceptance testing of megavoltage therapy
installations. In Williams JR, Thwaites DI (eds) Radiotherapy Physics in Practice (2nd edn.).
Oxford: Oxford University Press, 2000, pp. 6–​30.
23 Johansson K-​A, Sernbo G, Van Dam J. Quality control of megavoltage therapy units. In
Williams JR, Thwaites DI (eds) Radiotherapy Physics in Practice (2nd edn.). Oxford: Oxford
University Press, 2000, pp. 77–​98.
24. Ezzell GA, Burmeister JW, Dogan N, et al. IMRT commissioning: Multiple institution
planning and dosimetry comparisons, a report from AAPM Task Group 119. Medical
Physics 2009; 36:5359–​73.
25. James H, Beavis A, Budgell G, et al. Guidance for the implementation of intensity
modulated radiotherapy: IPEM Report 96. York: IPEM, 2008.
26. Mijnheer B, Georg D (eds.) Guidelines for the verification of IMRT. Brussels: ESTRO, 2008.
27. Miften M, Olch A, Mihailidis D, et al. Tolerance limits and methodologies for IMRT
measurement-​based verification QA: Recommendations of AAPM Task Group No. 218.
Medical Physics 2018; 45:e53–​83.
28. NRIG: NHS National Cancer Action Team. National Radiotherapy Implementation Group
Report: Image Guided Radiotherapy (IGRT) Guidance for implementation and use.
August 2012.
29. Mijnheer B, Olaciregui-​Ruiz I, Rozendaal R, et al. 3D EPID-​based in vivo dosimetry
for IMRT and VMAT. Journal of Physics: Conference Series 444 012011 2013 (7th
International Conference on 3D Radiation Dosimetry)
30. Low D, Moran JM, Dempsey JF, et al. Dosimetry tools and techniques for IMRT. Medical
Physics 2011; 38:1313–​26.
31. Hussein M, Adams EJ, Jordan TJ, et al. Critical evaluation of the PTW 2D-​ARRAY
seven29 and OCTAVIUS II phantom for IMRT and VMAT verification. Journal of Applied
Clinical Medical Physics 2013; 14(6):274-​92.
32. Langen KM, Papanikolaou N, Balog J, et al. Report of AAPM TG 148: Quality assurance
for helical tomotherapy. Medical Physics 2010; 37:4817–​53.
33. Dieterich S, Cavedon C, Chuang CF, et al. Report of AAPM TG 135: Quality assurance for
robotic radiosurgery. Medical Physics 2011; 38:2914–​36.
34. Wang H, Shui A, Wang C, et al. Dosimetric effect of translational and rotational errors
for patients undergoing image-​guided sterotactic body radiotherapy fro spinal metasteses.
International Journal of Radiation Oncology, Biology, Physics 2008; 71:1261–​71.
35. Solberg TD, Balter JM, Benedict SH, et al. Quality and safety considerations in
stereotactic radiosurgery and stereotactic body radiation therapy: Executive summary.
Practical Radiation Oncology 2012; 2: 2–​9.
36. Benedict SH, Yenice KM, Followill D, et al. Stereotactic body radiation therapy: The
report of AAPM Task Group 101. Medical Physics 2010; 37:4078–​101.
37. Aspradakis M, Byrne JP, Palmans H, et al. Report 103: Small Field MV Photon Dosimetry.
Institute of Physics and Engineering in Medicine (IPEM), York 2010.
 References 535

38. Williamson JF, Dunscombe PB, Sharpe MB, et al. Quality assurance of radiation
therapy: the challenges of advanced technologies. International Journal of Radiation
Oncology, Biology, Physics 2008; 71:Supplement: S1–​214.
39. Rowshanfarzad P, Sabet M, O’Connor DJ, Greer PB. Isocenter verification for linac-​
based stereotactic radiation therapy: review of principles and techniques Journal of Applied
Clinical Medical Physics 2011; 12:185–​95.
40. Galvin JM, Bednarz G. Quality assurance procedures for stereotactic body radiation
therapy. International Journal of Radiation Oncology, Biology, Physics 2008; 71:S122–​5.
41. Solberg TD, Medin PM, Mullins J, Li S. Quality assurance of immobilization and target
localization systems for frameless stereotactic cranial and extracranial hypofractionated
radiotherapy. International Journal of Radiation Oncology, Biology, Physics 2008;
71:S131–​S5.
42. SABR UK Consortium. Stereotactic ablative body radiation therapy (SABR): a
resource v. 5.1. Jan 2016. https://​sabr.org.uk/​wp-​content/​uploads/​2016/​04/​
UKSABRConsortiumGuidelinesv51.pdf.
43. NCS 2015 Process Management and QA for Intracranial Stereotactic radiotherapy
(Netherlands Commission on Radiotherapy Dosimetry).
44. Kerkmeijer LG, Fuller CD, Verkooijen HM, et al. The MRI-​Linear Accelerator
Consortium: Evidence-​Based Clinical Introduction of an Innovation in Radiation
Oncology Connecting Researchers, Methodology, Data Collection, Quality Assurance, and
Technical Development. Frontiers in Oncology 2016; 6:215.
45. Schell MC, Bova FJ, Larson, DA, et al. AAPM Report 54 Report of Task Group 42 1995;
AAPM New York.
46. Park SY. ppt2012 http://​chapter.aapm.org/​GLC/​media/​2012/​Park.pdf
47. Arjomandy B ppt 2015 http://​amos3.aapm.org/​abstracts/​pdf/​99-​28447-​359478-​110128-​
305773996.pdf
48. ACR-​AAPM Technical Standard for the performance of proton beam radiation therapy.
Revised 2018 (CSC/​BOC). https://​www.acr.org/​-​/​media/​ACR/​Files/​Practice-​Parameters/​
proton-​therapy-​ts.pdf?la=en
49. Arjomandy B, Sahoo N, Zhu XR, et al. An overview of the comprehensive proton therapy
machine quality assurance procedures implemented at The University of Texas M.D.
Anderson Cancer Center Proton Therapy Center-​Houston. Medical Physics 2009; 36:2269.
50. Ohri N, Shen X, Dicker AF, et al. Radiotherapy protocol deviations and clinical
outcomes: a meta-​analysis of cooperative group clinical trials. Journal of the National
Cancer Institute 2013; 105:387–​93.
51. Peters LJ, O’Sullivan B, Giralt J, et al. Critical impact of radiotherapy protocol compliance
and quality in the treatment of advanced head and neck cancer: results from TROG 02.02.
Journal of Clinical Oncology 2010; 28:2996–​3001.
52. Johansson KA, Hanson WF, Horiot JC. Meeting Report Workshop of the EORTC
Radiotherapy Group on quality assurance in co-​operative trials of radiotherapy: a
recommendation for EORTC Co-​operative Groups. Radiotherapy and Oncology 1988;
11:201–​3.
53. Thwaites DI, Williams JR, Aird EGA, et al. A dosimetric intercomparison of megavoltage
photon beams in UK radiotherapy centres. Physics in Medicine and Biology 1992;
37:445–​61.
54. Aird EGA, Williams C, Mott GTM, et al. Quality assurance in the CHART clinical trial.
Radiotherapy and Oncology 1995; 36: 235–​45.
536 Quality assurance in radiotherapy

55. Venables K, Winfield EA, Aird EG, Hoskin PJ. Three-​dimensional distribution of
radiation within the breast: an intercomparison of departments participating in the START
trial of breast radiotherapy fractionation. International Journal of Radiation Oncology,
Biology, Physics 2003; 55: 271–​9.
56. Venables K, Miles EA, Aird EGA, Hoskin PJ. The use of in vivo thermoluminescent
dosimeters in the quality assurance programme for the START breast fractionation trial.
Radiotherapy and Oncology 2004; 71:303–​10.
57. Moore AR, Warrington AP, Aird EG, et al. A versatile phantom for quality assurance
in the UK Medical Research Council (MRC) RT01 trial in conformal radiotherapy for
prostate cancer. Radiotherapy and Oncology 2006; 80: 82–​5.
58. Venables K, Winfield E, Deighton A, et al. The START trial-​measurements in semi-​
anatomical breast and chest wall phantoms. Physics in Medicine and Biology 2001;
46:1937–​48.
59. Clark CH, Miles EA, Urbano MT, et al. Pre-​trial quality assurance processes for an
intensity-​modulated radiation therapy (IMRT) trial: PARSPORT, a UK multicentre Phase
III trial comparing conventional radiotherapy and parotid-​sparing IMRT for locally
advanced head and neck cancer. British Journal of Radiology 2009; 82:585–​94.
60. Díez P, Hoskin PJ, Aird EG. Treatment planning and delivery of involved-​field
radiotherapy in advanced Hodgkin’s disease; results from a questionnaire-​based audit for
the UK Stanford V vs. ABVD clinical trial quality assurance programme. British Journal of
Radiology 2007; 80:816–​21.
61. Mayles WPM, Moore AR, et al., on behalf of the RT-​01 management group.
Questionnaire based quality assurance for the RT01 trial of dose escalation in
conformal radiotherapy for prostate cancer. Radiotherapy and Oncology 2004;
73:199–​207.
62. Guerrero Urbano MT, Clark CH, Kong C, et al. Target volume definition for head
and neck intensity modulated radiotherapy: pre-​clinical evaluation of PARSPORT trial
guidelines. Clinical Oncology (Royal College Radiologists) 2007; 19:604–​13.
63. Gwynne S, Spezi E, Wills L, et al. Toward semi-​automated assessment of target volume
delineation in radiotherapy trials: the SCOPE 1 pretrial test case. International Journal of
Radiation Oncology, Biology, Physics 2012; 84:1037–​42.
64. Wills L, Maggs R, Lewis G, et al. Quality assurance of the SCOPE 1 trial in oesophageal
radiotherapy. Radiotherapy and Oncology 2017; 12:179
65. Clark CH, Hussein M, Tsang Y, et al. A multi-​institutional dosimetry audit of rotational
intensity-​modulated radiotherapy. Radiotherapy and Oncology 2014; 113:272–​8.
66. Eaton DJ, Tyler J, Backshall A, et al. An external dosimetry audit programme to credential
static and rotational IMRT delivery for clinical trials quality assurance. Physica Medica
2017; 35:25–​30.
67. Conibear J, Spezi E, Gujral D, Nutting CM. Impact of quality assurance on contour
conformity within two UK head and neck radiotherapy trials. Radiotherapy Oncology 2013;
106 (Suppl 2):S359.
 Index

Tables and figures are indicated by t and f following the page number.
RT has been used an abbreviation for radiotherapy.
Many main entries can also be found as sub-​entries under the specific cancer headings.
Active-​Breathing Control  89
adaptive RT (ART)  45–​46
adnexal tumours  450
anal cancer  see squamous cell carcinoma of the
anus (SCCA)
Askin tumours of the chest wall  490
astrocytoma  391, 392, 394
anaplastic 496
diffuse  494–​95,  496f
high-​grade glioma (HGG)  363
pilocytic  494–​95
pilomyxoid  494–​95
basal cell carcinoma see squamous cell
carcinoma (SCC) and basal cell
carcinoma (BCC)
B-​cell lymphomas  323, 330
cutaneous  324–​25
non-​Hodgkin  154–​55
skin 339
and stomach cancer  154–​55
see also diffuse large B-​cell lymphomas
beam angles  37, 64
beam divergence  16, 18
beam hardening  14–​15
beam line  53
beam matching and asymmetric fields  19–​21
electron electron match  21
photon electron match  20
photon photon match  20
beam model systems  22t
beam weights  35, 37
bile duct cancer  162
biliary tree cancer see gall bladder and biliary
tree cancer
bladder cancer
chemotherapy  263–​64,  275
chemotherapy-​RT (CRT)  266
clinical target volume (CTV)  266, 267, 269, 276
dose prescription  272–​73
chemotherapy-​RT (CRT)  272–​73
RT as sole treatment  272
gross tumour volume (GTV)  267
image-​guided RT (IGRT)  46, 264, 266, 269
indications  263–​66,  265t
intensity-​modulated RT (IMRT)  264, 266,
269,  273–​75
node positive disease  273–​75, 274f
palliative treatment  275–​76
dose prescription  276
indications 275
planning technique  276
treatment verification  276
treatment volume and definition -​
locoregional palliation  275–​76
planning technique  267–​70, 276
dose distribution, fields and dose
constraints  269–​70, 269f, 270f, 271t, 276
patient position and immobilization  267, 276
volume definition  276
volume/​field localization  267–​69,  268f
postoperative RT and indications  275
Boltzman linear transport equation (BLTE)  22t
bone metastasis  470–​514
local bone pain  509–​10
dose distribution  511
dose prescription  511–​12
patient position and immobilization  509
volume and localization  509–​10, 510f
lumbar spine metastasis  510f
multiple bone pain -​wide field (hemibody)
irradiation  512–​13
dose distribution  512
dose prescription  512–​13
patient position and immobilization  512
treatment volume  512
myeloma 512
oligometastases  508–​9
dose distribution  509, 510f
patient position and immobilization  508
volume and localization  509, 509t
plasmacytoma, solitary  514
postoperative RT  513
dose distribution  513
dose prescription  513
patient position and immobilization  513
volume definition  513
spinal cord compression  508, 513–​14
biologically effective dose (BED)  514
dose distribution  514
dose prescription  514, 515t
treatment volume  514
bone sarcoma  454–​55,  460–​62
chondrosarcoma  454–​55,  462
chordoma  454–​55,  462
pleomorphic sarcoma  454–​55
538 Index

bone sarcoma (cont.) lymphoedema 87

see also Ewing’s sarcoma; osteosarcoma; soft second radiation-​induced malignancy  86–​87
tissue and bone sarcomas skin problems  87
brachial plexopathy  99 mastectomy  79–​80, 81–​82,  84
Bragg peak  2 nodal contouring  112
proton advantage  55–​56, 57 post-​mastectomy RT (PMRT)  79, 80, 82
proton beam therapy for skull base and spinal pre-​planning procedures  87–​95
tumours 400 breath-​holding techniques  88–​89,  90f
proton range in patient  68 contouring  91–​95
proton therapy: radiobiological imaging for RT planning  90–​91
effectiveness  67–​68 patient position and
proton therapy: scanned beams  60–​61, 62f immobilization  87–​89,  89f
proton therapy: scattered beams  59, 60 RT planning  95–​109
quality assurance  530 chest wall  98
brain metastasis  396–​97, 508, 515–​16 dose constraints and objectives  103–​7,
and breast cancer  396 103f, 106t
dose distribution  515 dose prescription  108–​9
dose prescription  516 ESTRO delineation guidelines for clinical
patient position  515 target volume (CTV)  96t
treatment volume  511f, 515 partial breast  98
breast cancer regional nodal areas  99–​103
adjuvant loco-​regional RT  79–​83 tumour bed boost  98–​99
adjuvant RT after breast conserving whole breast  95–​98
surgery  70–​79 second primary malignancies (SPM)  86–​87
ductal carcinoma in situ RT  77–​79 simultaneous integrated boost (SIB)  98, 99
partial breast RT  72–​73, 76t standardized incidence ratio (SIR)  86–​87
RT avoidance for low recurrence supraclavicular fossa (SCF)  80–​81, 82
risk  73–​77,  76f treatment and verification  109–​11
tumour bed boost RT  71–​72 dose delivery checking  110–​11
adverse effects of loco-​regional RT  84–​87, 85t imaging methodology  110
acute effects  85, 85t set-​up and on treatment imaging  109–​10
intermediate effects  85t volumetric-​modulated arc RT (VMAT)  84
late effects  85t,  86–​87 Bremsstrahlung  6, 54f, 55, 56t
axilla 112 bronchial cancer  531
axillary lymph node dissection Burkitt lymphoma  483
(ALND)  80–​81,  82 central nervous system (CNS) tumours
brachial plexopathy  99 brain tumours, primary  351
breath-​hold techniques  84, 88–​89, 90f, 103 cervical tumours  353
chest wall  82–​83 chordoma 357
dose constraints and objectives  103–​7, craniopharyngioma  351, 362
103f, 106t craniospinal irradiation (CSI)  492
calculation algorithms  103–​4 glial tumours  352
photon beam energy  104f, 105–​7, 105f intensity-​modulated proton therapy
dose prescription  108–​9 (IMPT) 400
breast and loco-​regional RT  108 lumbar tumours  353
palliative RT  108–​9 medulloblastoma 351
dose reference point  98 nerve sheath tumours  351
ESTRO consensus guideline  92f, 93f, 95 see also vestibular schwannoma
hypofractionation  70–​71, 72, 79 normal tissue tolerance doses  359–​63
intensity-​modulated RT (IMRT)  73, 87, 88–​ alopecia, permanent  363
89, 98, 99, 102, 105–​7, 110–​11, 112 brain  360–​61
internal mammary chain (IMC) brainstem 361
irradiation  82–​83, 102f, 103, 112 cervical cord  361
late effects  85t,  86–​87 children  359, 362
breast appearance, changes in  87 ear and cochlea  362
breast firmness  87 haematological malignancy  360–​61
breast pain  87 lacrimal gland  363
breast shrinkage  87 lens of eye  363
breast swelling  87 optic nerves and chiasm  362
cardiac 86 pituitary and hypothalamus  362
 Index 539

spinal cord  361–​62 dose prescription  335

thoracic cord  361 primary mediastinal  321–​22
primary cerebral lymphoma  351–​52 Stage I and II  320–​22
solitary metastasis  351–​52 Stage III and IV  322
stereotactic radiosurgery (SRS)  353, 356, stomach 338
379–​80, 394–​98, 395f, 398f
brain metastases  396–​97 elastic Coulomb scattering  56t
meningioma 397 elastic nuclear collisions  55
organs at risk (OAR)  396, 397 electron electron match  21
pituitary  397–​98 embryonal rhabdomyosarcoma  433
vestibular schwannoma  397, 398f embryonal tumours  494, 499
whole brain RT (WBRT)  396, 397 embryonic renal tumour see Wilms’ tumour
teratoma  351, 377 endocrine therapy  76
thoracic tumours  353 endometrial cancer  301f, 302f
ependymoma  351, 373–​74, 391, 392, 394, 498
cervical cancer  307, 353 anaplastic 498
adenocarcinoma, endocervical  289 dose prescription  374
conformal treatment  295f intensity-​modulated RT (IMRT)  498
dose distribution  293–​95 intracranial 498
conformal and conventional RT  293, 295f lumbar 392
intensity-​modulated RT (IMRT)  293–​95,  296f myxopapillary  392, 498
dose prescription  297, 298 proton beam therapy  498
external beam RT  298 equivalent uniform dose (EUD)  36–​37
intracavitary vaginal vault brachytherapy  298 Ewing’s sarcoma  454–​55,  460–​61
primary disease  297 bony secondaries  461
planning technique  292–​95 proton beam therapy  65–​67, 66f
dose distribution  293–​95 radical whole lung RT  461
patient position and immobilization  292
target volume localization  292–​93 follicular lymphoma  317–​19, 324, 483
postoperative adjuvant treatment  297–​98 dose prescription  335
indications  297–​98 extranodal sites  337
RT technique  298 palliative RT  336
target volume definition  298
primary disease  297 gall bladder and biliary tree cancer  161–​63
external beam RT  297 palliative treatment  163
recurrent disease  297 post-​operative adjuvant treatment  162–​63
children see paediatrics oncology indications 162
cholangiocarcinoma  161–​62,  163 planning technique  162
chondrosarcoma  400–​1, 454–​55, 462,  472–​73 positioning 162
chordoma  357, 400–​1, 454–​55, 462, 472–​73 prescription  162–​63
choroidal metastasis  517–​18 treatment volume and definition  162
breast cancer  517 verification 162
dose prescription  518 stereotactic ablative RT (SABR)  163
lung cancer  517 volumetric-​modulated arc RT (VMAT)  162
colorectal carcinoma  396 ganglioneuroblastomas 499
Compton process  6, 7f, 8f gastric cancer  157
conformity index (CI)  29, 31t gastric lymphoma  325
convolution algorithms  21–​22, 22t, 24t germ cell tumours (GCT)  494
cost function tool  38, 40–​41 germinoma  351,  377–​78
Coulomb interactions  54–​55, 54f, 58, 63–​64 children  377,  502–​3
craniopharyngioma  351, 362, 473f intracranial  377, 503
dose prescription  386 glial tumours  351, 352
indications 382 glioblastoma (GBM)  363, 391, 392, 393–​94, 496
treatment volume and definition  383–​86 dose prescription  368
indications  364–​65
diffuse large B-​cell lymphoma  317–​18, 319–​22, gliomas  351–​52, 355f, 356f,  357–​58
324, 451, 483 diffuse midline  492, 503
anaplastic 321 see also high-​grade glioma (HGG); low-​grade
chemotherapy  320–​22 glioma (LGG); medulloblastoma
CMT 321 gliomatosis cerebri  363–​64
540 Index
glioneuronal tumours  494–​95
glottic tumours  415, 418–​20, 419f
head and neck cancer
3D conformal RT (3D-​CRT)  409, 418
dose prescription
large volume disease  415
small volume disease  415
floor of mouth carcinoma  405
glottic tumours  415, 418–​20, 419f
human papilloma virus (HPV)  405
hypopharynx  412, 424–​25, 425f
posterior pharyngeal wall  425
pyriform fossa  425
laryngohypopharyngeal cancer  418
lymph node levels  408f
palliative RT  418
postcricoid tumours  425
postoperative RT  416–​18
dose prescription  417–​18
indications and treatment volume  416–​17
intended dose prescription  417–​18
planning technique  417
recurrent disease and palliation  436
simultaneous integrated boost (SIB)  414
simultaneous modulated accelerated RT
(SMART)  414, 415f
soft palate cancer  424
subglottic tumours  421, 421f
supraglottic tumours  420, 420f
tongue base carcinoma  405
tonsil carcinoma  412, 418, 423, 423f
volumetric-​modulated arc RT (VMAT)  416
high-​grade glioma (HGG)  354–​56, 357, 363–​69
astrocytomas 363
glioblastoma (GBM)  363
gliomatosis cerebri  363–​64
oligoastrocytomas 363
oligodendrogliomas 363
palliative RT  369
dose prescription  369
indications 368
planning technique  369
treatment volume and definition  368–​69
radical RT  364–​68
Hodgkin lymphoma (HL)  317–​18, 326–​31,
482, 483f
dose prescription  336
extended-​field RT (EFRT)  327
intensity-​modulated RT (IMRT)  482
planning target volume (PTV)  482
radiation therapy volume and dose  329
extended-​field RT (EFRT)  329
involved-​field RT (IFRT)  329
involved node RT (INRT)  329
relapsed or refractory: treatment  331
risk adapted therapy  326
nodular lymphocyte-​predominant HL
(NLPHL)  328–​29
human immunodeficiency virus (HIV)  197,
198, 204, 205
human papilloma virus (HPV)
head and neck cancer  405
-​induced squamous cell carcinomas  289
and squamous cell carcinoma of the anus
(SCCA)  196–​97,  205
hypopharynx  412, 424–​25, 425f
posterior pharyngeal wall  425
pyriform fossa  425
ICRU 50 and ICRU 62  28–​33, 31t,  42–​43
ICRU 83  43–​44
ICRU limits  33
inelastic Coulomb scattering  56t
internal mammary chain (IMC)
irradiation  82–​83, 102f, 103, 112
intracranial germ cell tumours  502–​3
chemotherapy 503
germinomas  502–​3
non-​germinomatous germ cell tumours  502
intraoperative RT (IORT)  160
intraventricular tumours  493
Kaposi’s sarcoma  451
lacrimal gland tumours  430
large bowel  65, 67
laryngohypopharyngeal cancer  418
larynx cancer  35, 35f, 405, 412, 418–​22
advanced 422
clinical target volume (CTV)  420–​21
glottic larynx  418
glottic tumours  418–​20, 419f
gross tumour volume (GTV)  420–​21
immobilization 418
intensity-​modulated RT (IMRT)  418,
420–​21,  422
planning target volume (PTV)  419, 421–​22
subglottic tumours  421, 421f
supraglottic tumours  420f,  420–​21
leiomyosarcoma 454
leukaemia  476,  477–​82
acute lymphoblastic leukaemia
(ALL)  477–​79,  503
acute myeloid leukaemia (AML)  477–​78
chronic myeloid leukaemia (CML)  477–​78
clinical target volume (CTV)  481
cranial RT  479–​80, 480f
clinical target volume (CTV)  480
dose prescription  480
dosimetry 480
patient position and immobilization  480
technique 480
craniospinal irradiation (CSI)  479
testicular irradiation  481–​82, 481f
clinical target volume (CTV)  481
dose prescription  482
dosimetry 481
patient position and immobilization  481
technique 481
total body irradiation (TBI)  478–​79, 480f,
480, 481

liposarcomas, myxoid  455, 456
liver metastasis  516–​17
dose distribution  517
dose prescription  517
oligometastases  516, 517
palliative treatment  517
patient position and mobilization  516
treatment volume and definition  516
volume localization  516–​17
low-​grade glioma (LGG)  354, 355f, 363,
369–​73,  494–​95
clinical target volume (CTV)  370
oligodendrogliomas 370
palliative RT  373
dose prescription  373
indications 373
planning technique  373
treatment volume and definition  373
radical RT  369–​73
dose prescription  372–​73, 372f
indications  369–​70
treatment volume and definition  370,
371f, 372f
lumbar tumours  353
lung cancer  115–​35
active breathing control (ABC) P  119
breath-​hold techniques  117, 119
clinical target volume (CTV)  120, 122, 128,
130, 132
concurrent chemotherapy  122–​23
gating 119
internal target volume (ITV)  126, 128, 132
organs at risk (OAR)  117, 121–​23, 124
brachial plexus  123
heart 123
lungs  121–​22
oesophagus  122–​23
proximal bronchial tree  123
proximal trachea  123
spinal cord  122
palliative thoracic treatment  133–​35
dose prescriptions  135
patient positioning  134
treatment verification  134
volume/​field localization  134
planning target volume (PTV)  120, 121–​22,
124, 126, 128, 129, 132
radical RT assessment  115–​17
disease stage and configuration  116–​17
fitness for treatment  115–​16
respiratory correlated or 4D
scanning  119–​524
slow CT scanning  119
stereotactic ablative RT (SABR)  117, 122,
123, 124, 126–​27, 128, 129
see also non-​small cell lung cancer (NSCLC);
small cell lung cancer (SCLC)
lymphomas  317–​43
anaplastic large cell  324
Ann Arbor classification  317–​18
Burkitt 483
Index 541
cutaneous  324–​25
cutaneous anaplastic large cell
(C-​ALC)  325
cutaneous follicular  324–​25
extradural 324
extranodal sites/​presentations  322–​26,
337–​38,  338f
central nervous system  338
head and neck  337, 339f
orbit 337
stomach 338
follicular lymphoma International Prognostic
Index (FLIPI)  317–​18
gastric 325
immunoblastic large cell  324
infradiaphragmatic lymph node
irradiation  336–​37
dosimetry 337
field arrangement  337
localization 337
patient position and immobilization  336
International Prognostic Index
(IPI)  317–​18
natural killer (NK) cell  317–​18, 324
nose and paranasal sinuses  433, 434
orbital  320, 430
pleomorphic large cell  324
primary central nervous system lymphoma
(PCNSL)  323–​24,  351
primary cerebral  351–​52
primary extranodal  322
RT techniques  331–​33
extended-​field RT (EFRT)  332
involved field RT (IFRT)  332–​33
involved node RT (INRT)  332
involved site RT ((ISRT)  332–​33,
333f, 334f
nodal irradiation  332
recurrent lymphomas  325–​26
subtotal nodal irradiation (STNI)  332
TLI 332
supradiaphragmatic lymph node
irradiation  333–​36
dose prescription  335
dosimetry  334, 336t
field arrangement  334, 335f
implementation and verification  336
patient position and immobilization  333
T-​cell see T-​cell lymphomas
testes  322,  342–​43
total body electron treatment  339–​41, 340f
dose prescription  341
field arrangement and dosimetry  339–​41
patient position  339
treatment volume  339
see also diffuse large B-​cell lymphoma;
follicular lymphoma; Hodgkin
lymphoma (HL); mucosa associated
lymphoid tissue (MALT) lymphomas;
non-​Hodgkin lymphoma (NHL)
lymphomatoid papulosis  325
542 Index
medulloblastoma  351,  374–​77
clinical target volume (CTV)  375
craniospinal irradiation (CSI)  499
spinal or other metastases  375
supra-​tentorial embryonal tumours
(previously neuroectodermal tumours or
PNET) 375
treatment volume and
definition  375–​77,  376f
cranial metastases boost  377
dose prescription  376–​77
spinal metastases boost  377
medulloepitheliomas 499
melanoma 395f, 396, 443–​51, 452
adjuvant treatment of nodal basin  448, 449t
adjuvant treatment of primary  448
clinical target volume (CTV)  451
desmoplastic 448
lentigo maligna  448
recurrent malignant  451
uveal 400
meningioma  351, 357–​58, 386–​91, 392,
394, 397
cavernous sinus  359, 387f, 387, 390f
clinical target volume (CTV)  389–​90
dose prescription  390–​91
indications  387–​88
optic nerve  387
radical postoperative adjuvant RT  387–​88
radical RT, primary and
adjuvant  386–​87,  387f
Merkel cell tumours (MCC)  449, 451
prophylactic nodal irradiation  449
primary/​adjuvant RT to primary site  449
mesothelioma  135–​37
palliative RT  136
postoperative RT (PORT)  136
proton beam therapy  136
mucoepidermoid tumour  428
mucosa associated lymphoid tissue (MALT)
lymphomas  319–​20,  324
dose prescription  335
extranodal sites  337
gastric  337, 338f
orbital 337
stomach  154–​55,  338
myeloma 512
myxoid liposarcomas  455, 456
nasopharyngeal cancer  31, 32f
natural killer (NK) cell lymphoma  317–​18,
324, 335
necrotizing leucoencephalopathy  476
nephroblastoma see Wilms’ tumour
nerve sheath tumours  351
neuroblastoma (NBL)  483–​84
abdominal 473
metastatic 503
olfactory (or esthesioneuroblastoma)  435–​36
neurofibromatosis type I (NFI)  494–​95
non-​elastic nuclear interactions  54f, 55,
56t, 58
non-​Hodgkin lymphoma (NHL)
see lymphomas
non-​small cell lung cancer (NSCLC)
see lung cancer
nose and paranasal sinuses  410–​11, 413,
418,  433–​36
adenocarcinoma 434
columella 435
embryonal rhabdomyosarcoma  433
ethmoid sinus  435
lymphoma  433, 434
maxillary antrum RT technique  433–​35,
433f, 434f
nasal tumours  418–​35
nasopharyngeal carcinoma  410–​11, 413, 418,
426–​27, 426f, 429
olfactory neuroblastoma (or
esthesioneuroblastoma)  435–​36
ocular tumours  53, 400
oesophageal cancer  145–​54
adenocarcinomas  152–​53
postoperative adjuvant therapy  153–​54
radical primary treatment  145–​54
squamous cell carcinoma (SCC)  145, 153
treatment volume and definition  146–​48
clinical target volume (CTV)
definition 147
endoluminal ultrasound (EUS)  146, 146f
fibre-​optic bronchoscopy  146
laparoscopy, biopsy and peritoneal
washings 146
olfactory neuroblastoma (or
esthesioneuroblastoma)  435–​36
oligoastrocytomas 363
oligodendrogliomas  363, 370
oligometastases  508–​9
oral cavity tumours  412, 427–​28
alveolus 427
buccal mucosa  427, 427f
floor of mouth  427–​28
tongue 427
orbital tumours  418, 429–​31, 430f
dose prescription  430–​31, 430f, 431f
lacrimal gland tumours  430
lymphomas  320, 430
rhabdomyosarcoma 430
sarcoma 431
thyroid eye disease  431
oropharyngeal tumours  413, 418, 422–​24
clinical target volume (CTV)  423
planning target volume (PTV)  423
soft palate  424
tonsil or base of tongue  423, 423f, 424f
osteosarcoma  454–​55, 461–​62, 463f, 466, 492
clinical target volume (CTV)  462
dose fractionation  462
gross tumour volume (GTV)  462
 Index 543

paediatrics oncology planning target volume (PTV)  159

chondrosarcoma  472–​73 stereotactic body RT (SBRT)  161
chordoma, base of skull and spinal  472–​73 volumetric-​modulated arc RT
craniopharyngioma 473f (VMAT) 160
craniospinal irradiation (CSI)  500–​2, 501f, 502f parotid gland tumours  413, 414f, 418, 428
3D conformal RT (3D-​CRT)  500 adenocarcinoma 428
clinical target volume (CTV)  500, 501 adenoid cystic carcinoma  428
dose prescription  501–​2 clinical target volume (CTV)  428
dosimetry 501 dose prescription  428
intensity-​modulated RT (IMRT)  490, 500 intensity-​modulated RT (IMRT)  428
patient position and immobilization  501 mucoepidermoid tumour  428
planning target volume (PTV)  501–​2 pleomorphic adenoma  428
proton beam therapy  500 squamous tumours  428
embryonal rhabdomyosarcoma  433 passive scattering system  58f, 67
embryonal tumours  494, 499 pencil beams  10f, 22t, 58f, 58, 65, 66–​67
ependymoma  373, 494 penile cancer
Ewing’s sarcoma  373, 475, 477, 489–​91, 494 clinical target volume (CTV)  282–​83, 285
proton beam therapy  490 intensity-​modulated RT (IMRT)  285
whole lung RT  491 lymph node irradiation technique  285–​87
germinoma  377,  502–​3 planning target volume (PTV)  283, 285
glioma, diffuse midline  503 radical external beam treatment
high-​grade glioma (HGG)  477, 496–​98, 503 technique  282–​84
astrocytoma, anaplastic  496 photoelectric effect  7f
diffuse midline glioma  497 photon depth dose curves  16f
embryonal tumours with multilayered photon electron match  20
rosettes 499 photon photon match  20
ependymoma 498 pineal tumours  493
glioblastoma 496 pineoblastomas  375, 499
medulloblastoma  498–​500 pituitary tumours  351, 359, 362, 380–​86
Hodgkin lymphoma  477, 482 adenoma
intracranial germ cell tumours  502–​3 dose prescription  386
chemotherapy 503 indications 382
germinomas  502–​3 treatment volume and definition  383, 384f
non-​germinomatous germ cell tumours  502 children  385–​86,  493
leukaemia see leukaemia clinical target volume (CTV)  383, 385–​86
medulloblastoma  374, 375, 471–​72, 473, 475, craniopharyngioma
492, 493, 493f, 494, 498–​500 dose prescription  386
medulloepitheliomas 499 indications 382
neuroblastoma (NBL)  483–​84 treatment volume and definition  383–​86
neuroblastoma (NBL), abdominal  473 dose prescription  386
neuroblastoma (NBL), metastatic  503 planning target volume (PTV)  353,
non-​Hodgkin lymphoma  483 383,  385–​86
normal tissue tolerance doses  359, 362 stereotactic radiosurgery (SRS)  382
osteosarcoma 492 pleomorphic adenoma  428
pineoblastomas 499 pleomorphic sarcoma  454–​55
pulmonary metastases  475 postcricoid tumours  425
rhabdomyosarcoma (RMS)  475, 476, 487–​89 primitive neuroectodermal tumours (PNET)
chemotherapy 488 see central nervous system (CNS)
dose fractionation  489t embryonal tumours; medulloblastomas
metastatic alveolar  503 prophylactic cranial irradiation  476
soft tissue and bone sarcomas  454, 465 prostate cancer  224–​55
pair production  6, 7f, 8f, 9f clinical target volume (CTV)  234, 236,
pancreatic cancer  158–​61 243, 255
clinical target volume (CTV)  159 dose distribution  238–​42
dose prescription  161, 161t 3D conformal RT  238–​40, 239f, 240t
gross tumour volume (GTV)  159, 160 intensity-​modulated RT
indications  158–​59 (IMRT)  240–​41,  241f
adjuvant therapy  158 tomotherapy  241–​42
neo-​adjuvant therapy  158 volumetric-​modulated arc RT
palliative therapy  147–​59 (VMAT) 241
544 Index

prostate cancer (cont.) treatment planning system (TPS)  521–​24,

dose prescription  245–​49 525, 527, 530
dose escalation -​conventional acceptance testing  522
fractionation  245–​47 commissioning  522–​24
dose escalation -​hypofractionation  247–​48 periodic quality control  524
indications  226–​30 system management and security  524
breast bud RT and gynaecomastia  230
palliative RT to metastases  230 rectal cancer  165–​91
palliative RT to prostate and pelvis  230 clinical target volume (CTV)  176–​77, 178–​81,
post-​prostatectomy RT  228–​30 181f, 182, 186–​87
radical RT to prostate plus pelvis  226–​28 endocavity RT  187
radical RT to prostate and seminal intraoperative RT (IORT)  188–​89
vesicles 226 locally recurrent rectal cancer  188–​89
late side effects  250–​51 re-​irradiation  189
erectile dysfunction  251 proton therapy  65
gastrointestinal toxicity  250 postoperative RT clinical target volume
genitourinary toxicity  250 (CTV) 183
post-​prostatectomy RT  228–​30 preoperative dose  184
adjuvant RT  229 small bowel exclusion techniques  183
adjuvant RT vs salvage RT  230 stereotactic body RT (SBRT)  190
salvage RT  229–​30 renal cell carcinoma  396
radical RT to prostate plus pelvis  226–​28 retinal metastases  517
node negative disease  226–​27 rhabdomyosarcoma (RMS)  475, 476, 
node positive disease  228 487–​89
stereotactic body RT (SBRT)  248 clinical target volume (CTV)  489
proton accelerator  53 dose fractionation  489t
proton beam therapy  53–​68 embryonal 433
interactions  54–​55, 54f, 56t metastatic alveolar  503
plan robustness  64 orbit 430
proton advantage  55–​57
proton range in patient  68 sarcomas
radiobiological effect  67–​68 chest wall  50
scanned beams orbit 431
advantages 61 pleomorphic  454–​55
disadvantages 61 see also Ewing’s sarcoma; osteosarcoma
scattered beams skin cancer
double scattering  59 adnexal tumours  450
passive scattering  59–​60 angiosarcoma 452
single scattering  59, 60 diffuse large B-​cell lymphoma  451
Kaposi’s sarcoma  451
quality assurance lymphoedema see Merkel cell tumours
daily QC testing  526 (MCC) below
dose calculation commissioning  523 Merkel cell tumours (MCC)  449
dose calculation verification tests  523–​24 prophylactic nodal irradiation  449
dosimetric commissioning  522–​23 primary/​adjuvant RT to primary
extended safety testing  526 site 449
immobilization techniques and see also squamous cell carcinoma (SCC) and
equipment  522–​23 basal cell carcinoma (BCC)
legislation  520, 521 small cell lung cancer (SCLC)  115, 131–​34
photon and electron beam dosimetry  525 chemotherapy-​RT (CRT)  131
plan evaluation tools  523 dose prescription  132, 135
process document  531 postoperative RT  133–​38
QART 520 prophylactic cranial irradiation
quality system  520–​21 (PCI)  132–​33,  135
treatment delivery  524–​26 soft palate cancer  424
acceptance testing for linear soft tissue and bone sarcomas
accelerators 525 anatomical sites  462–​64
commissioning 525 chest wall  464
in vivo dosimetry  526 forearm 462
periodic quality control  525–​26 hands and feet  463

retroperitoneum 463
spinal/​para-​spinal  464
angiosarcoma  464, 465f
children  454, 465
Ewing’s sarcoma  466, 466f
fibromatosis 465
osteosarcoma 466
palliative treatment  465–​66
cerebral secondaries  466
lung secondaries  466
see also bone sarcoma; soft tissue sarcoma
soft tissue sarcoma  454
leiomyosarcoma 454
liposarcoma 454
myxoid liposarcomas  455, 456
pleomorphic sarcoma  454
spinal cord tumours  30, 31, 42–​43, 353, 391–​94
anaplastic (grade III) tumours  393–​94
astrocytoma  391, 392, 394
clinical target volume (CTV)  392
dose prescription  393–​94
ependymoma  391, 392, 394
ependymoma, lumbar  392
glioblastoma  391, 392, 393–​94
meningioma  391, 392, 394
planning target volume (PTV)  392–​93
vestibular schwannomas  391, 392, 394
squamous cell carcinoma of the anus
(SCCA)  196–​219
inflammatory bowel disease (IBD)  204
palliative treatment  217–​18
radical primary treatment  204–​15
chemotherapy 215
general points  205–​6
indications  204–​5
RT planning technique  207–​15
sentinel lymph node biopsy (SLNB)  207
stomach cancer  154–​58
palliative treatment  157–​58
post-​operative adjuvant therapy  155–​57
radical primary treatment  154–​55
subglottic tumours  421, 421f
supraglottic tumours  420f,  420–​21
supra-​tentorial embryonal tumours  375
T-​cell lymphomas
lymphoblastic 483
peripheral 322
primary cutaneous  325
teratoma  351, 377
testicular cancer
clinical target volume (CTV)  279, 280, 281
and leukaemia  479, 481–​82
palliative treatment  281
bone and cerebral metastasis  281
para-​aortic lymph node
irradiation  279–​80
dose prescription  280
patient position and immobilization  279
Index 545
thymic tumours  137–​39
chemotherapy-​RT  137
clinical target volume (CTV)  138
internal target volume (ITV)  138
palliative RT  139
planning target volume (PTV)  138
postoperative RT (PORT)  137
thyroid eye disease  431
tongue carcinoma  405, 423, 424f
tonsillar carcinoma  412, 418, 423, 423f, 424f
total body irradiation (TBI)  16
total limb electron irradiation
(TLEI) 446f,  451–​52
total scalp irradiation (TSI)  452
upper gastrointestinal tract see gall bladder
and biliary tree cancer; oesophageal
cancer; pancreatic cancer; stomach
cancer
uterine cancer  298–​303
clinical target volume (CTV)  301
dose distribution  301f, 302, 302f
conventional and conformal RT  302
intensity-​modulated RT (IMRT)  302, 302f
dose prescription  299, 303
external beam RT  299, 303
intracavitary brachytherapy  303
intrauterine brachytherapy  299
palliative treatment  306
planning target volume (PTV)  301, 302
postoperative adjuvant treatment  299–​303
primary radical treatment  298–​99
indications 298
RT technique  299
treatment volume and definition  299
vaginal cancer
adenocarcinoma 312
dose prescription  313, 315
palliative treatment  315
vestibular schwannoma  351–​52, 355f, 378–​80,
381f, 391, 392, 394, 397, 398f
dose prescription  380
neurofibromatosis (NF)  378
treatment volume and definition  380
vulval cancer
clinical target volume (CTV)  309,
310,  312–​13
planning target volume (PTV)  309, 310, 313
Wilms’ tumour  341–​42, 475, 477, 484–​87, 485t,
486f, 503
clinical target volume (CTV)  485, 487
dose prescription  486, 487
NWTS staging system  484
planning target volume (PTV)  485, 487
technique  485–​86,  487
whole abdominal RT  487
whole lung RT  486–​87