Gynaecological Oncology, Second Edition

This page intentionally left blank Gynaecological Oncology Gynaecological Oncology Mahmood I. Shaﬁ Consultant Gyna...

Author: Mahmood I. Shafi MD MRCOG | Helena M. Earl PhD FRCP | Li Tee Tan MRCP FRCR

116 downloads 1299 Views 5MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

This page intentionally left blank

Gynaecological Oncology

Gynaecological Oncology Mahmood I. Shaﬁ

Consultant Gynaecological Surgeon and Oncologist Department of Gynaecological Oncology Addenbrooke’s Hospital Cambridge University Hospitals NHS Foundation Trust Cambridge

Helena M. Earl

Reader in Clinical Cancer Medicine and Honorary Consultant Medical Oncologist University of Cambridge Department of Oncology Addenbrooke’s Hospital Cambridge University Hospitals NHS Foundation Trust Cambridge

Li Tee Tan

Consultant Clinical Oncologist Oncology Centre Addenbrooke’s Hospital Cambridge University Hospitals NHS Trust Cambridge

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo, Delhi, Dubai, Tokyo Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521730242 © Cambridge University Press, 2010 This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2009 ISBN-13

978-0-511-69122-5

eBook (NetLibrary)

ISBN-13

978-0-521-73024-2

Paperback

Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate. To the extent permitted by applicable law, Cambridge University Press is not liable for direct damages or loss of any kind resulting from the use of this product or from errors or faults contained in it, and in every case Cambridge University Press’s liability shall be limited to the amount actually paid by the customer for the product. Every eort has been made in preparing this publication to provide accurate and up-to-date information which is in accord with accepted standards and practice at the time of publication. Although case histories are drawn from actual cases, every eort has been made to disguise the identities of the individuals involved. Nevertheless, the authors, editors and publishers can make no warranties that the information contained herein is totally free from error, not least because clinical standards are constantly changing through research and regulation. The authors, editors and publishers therefore disclaim all liability for direct or consequential damages resulting from the use of material contained in this publication. Readers are strongly advised to pay careful attention to information provided by the manufacturer of any drugs or equipment that they plan to use.

To our respective families for the support and love that we enjoy.

Contents List of contributors page viii Preface xi List of abbreviations xii

1

Epidemiology of gynaecological cancers 1 Anjum Memon

2

Screening and prevention of gynaecological cancers 15 Ranjit Manchanda and Ian J. Jacobs

3

Pathology of gynaecological cancers 33 Mark J. Arends

4

Imaging in gynaecological oncology 51 Nyree Griﬃn and Evis Sala

5 Surgery for gynaecological cancers 69 Alan Farthing 6

Radiotherapy for gynaecological cancers 79 Charlotte Coles

7

Systemic therapies for gynaecological cancers 91 Helena M. Earl

8

Palliative care in gynaecological oncology 103 Sara Booth

9

Ovarian cancer 119 Robin Crawford and Yin-Ling Woo

10

Endometrial cancer Catherine Holland

11

Cervical and vaginal cancer 147 Ahmed Ahmed, Li Tee Tan and Mahmood I. Shaﬁ

12

Vulvar cancer Peter Baldwin

13

Gynaecological sarcomas 175 John Latimer and Helena M. Earl

14

Gestational trophoblastic disease and ovarian germ cell tumours 181 Neil J. Sebire and Michael J. Seckl

15

Decision making in oncology Li Tee Tan

133

163

199

Appendices Appendix 1:

Miscellaneous information Appendix 2: Incidence and mortality data Appendix 3: Staging: FIGO and TNM Appendix 4: Useful websites Index

207 209 213 223 225

Contributors Ahmed Ahmed Department of Obstetrics and Gynaecology and Department of Oncology University of Cambridge Cambridge Mark J. Arends Department of Histospathology Addenbrooke’s Hospital Cambridge University Hospitals NHS Foundation Trust Cambridge Peter Baldwin Department of Gynaecological Oncology Addenbrooke’s Hospital Cambridge University Hospitals NHS Foundation Trust Cambridge Sara Booth Addenbrooke’s Hospital Cambridge University Hospitals NHS Foundation Trust Cambridge Charlotte Coles Addenbrooke’s Hospital Cambridge University Hospitals NHS Foundation Trust Cambridge Robin Crawford Department of Gynaecological Oncology Addenbrooke’s Hospital Cambridge University Hospitals NHS Foundation Trust Cambridge Helena M. Earl Department of Oncology Addenbrooke’s Hospital

Cambridge University Hospitals NHS Foundation Trust Cambridge Alan Farthing West London Gynaecological Cancer Centre Queen Charlotte’s Hospital London Nyree Griﬃn Department of Radiology Guy’s and St Thomas’ Hospitals London Catherine Holland Academic Unit of Obstetrics and Gynaecology University of Manchester School of Cancer and Imaging Science St Mary’s Hospital Manchester Ian J. Jacobs EGA Institute of Women’s Health University College London London John Latimer Department of Gynaecological Oncology Addenbrooke’s Hospital Cambridge University Hospitals NHS Foundation Trust and University of Cambridge Cambridge Ranjit Manchanda Department of Gynaecological Oncology Institute of Women’s Health University College London London

List of contributors

Anjum Memon Brighton and Sussex Medical School Falmer Sussex Evis Sala Department of Radiology University of Cambridge School of Clinical Medicine Cambridge Neil J. Sebire Department of Histopathology Great Ormond Street Hospital and Trophoblastic Disease Unit Charing Cross Hospital London Michael J. Seckl Trophoblastic Disease Unit Department of Medical Oncology Charing Cross Hospital London

Mahmood I. Shaﬁ Department of Gynaecological Oncology Addenbrooke’s Hospital Cambridge University Hospitals NHS Foundation Trust Cambridge Li Tee Tan Oncology Centre Addenbrooke’s Hospital Cambridge University Hospitals NHS Foundation Trust Cambridge Yin-Ling Woo Department of Gynaecological Oncology Addenbrooke’s Hospital Cambridge University Hospitals NHS Foundation Trust Cambridge

ix

Preface The management of gynaecological cancers has seen considerable changes over recent years. We now have a greater understanding of cancer epidemiology and signiﬁcant advances have been made in genetics/screening. The approach to women with cancer has become multidisciplinary, with many professionals contributing to the excellent care and outcome that we wish to see for those individuals we are privileged to look after. Women with gynaecological cancers have general as well as speciﬁc issues and these are addressed by the various experts who have contributed to this book. The overall desire for a holistic approach should be evident in trying to address these issues. This book is aimed at trainees in gynaecology and oncology as well as consultants with an interest in cancer care. Allied medical staﬀ, palliative services and nurse specialists will also ﬁnd it a useful adjunct to getting current information on the whole range of gynaecological oncology. Chapters 1–8 and 15 cover the broad issues in gynaecological oncology, while Chapters 9–14 discuss in some detail the issues in relation to site-speciﬁc gynaecological cancers. The further reading sections at the end of each chapter are intended to guide the interested reader wishing to seek detailed information relating to the issues covered in the chapters. Our task in editing this handbook has been made enjoyable by the contributors who have provided material relating to their area of expertise. It has been a pleasure working with colleagues and friends whom we respect and admire.

Abbreviations ACE α-AFP ASCCP ASCUS ASR ASTEC BCC BEP BMI BRCA BSCCP BSO BT CA125 CA19.9 CEA CGIN CHM CIN CNS CO CSF CT CXR D&C DES DNA DVT EBRT ECX EMA EOC EP ER EUA FBC FDG FIGO FNA(C) FOV

actinomycin D, cyclophosphamide, etoposide alpha fetoprotein – tumour marker for hepatic and germ cell tumours American Society of Colposcopy and Cervical Pathology atypical Squamous Cells of Undetermined Signiﬁcance age-standardized incidence (or mortality) rate A Study in the Treatment of Endometrial Cancer basal cell carcinoma bleomycin, etoposide, cis-platinum body mass index breast cancer-associated antigen British Society for Colposcopy and Cervical Pathology bilateral salpingo-oophorectomy brachytherapy cancer-associated antigen 125 – tumour marker for ovarian cancer cancer-associated antigen 19.9 – tumour marker for pancreatic cancer carcinoembryonic antigen – tumour marker for colon cancer cervical glandular intraepithelial neoplasia complete hydatidiform mole cervical intraepithelial neoplasia central nervous system cyclophosphamide and vincristine cerebrospinal ﬂuid computed tomography chest X-ray dilatation and curettage diethylstilbestrol deoxyribonucleic acid deep vein thrombosis external beam radiotherapy epirubicin, cisplatin, oral capecitabine etoposide, methotrexate and actinomycin D epithelial ovarian cancer etoposide and cisplatin estrogen receptor examination under anaesthesia full blood count ﬂuorine-18-ﬂuoro-2-deoxy-d-glucose International Federation of Gynecology and Obstetrics ﬁne needle aspirate (cytology) ﬁeld of view

List of abbreviations

GA GCT GFR GnRH GOG GTD GTN GTT β-hCG hCG HDR HIV HNPCC HPF HPL HPV HRT IACR IARC ICD ICON IU IVU LA LDH LLETZ MAS MDSCR MDT MMMT MOGCT MR(I) MRC MTD NHSCSP NOS NSAID ONS PE PET PHM PMP POMB PPV PSTT PVB

general anaesthesia granulosa cell tumour glomerular ﬁltration rate gonadotrophin releasing hormone Gynecologic Oncology Group gestational trophoblastic disease gestational trophoblastic neoplasia gestational trophoblastic tumour beta human chorionic gonadotrophin human chorionic gonadotrophin high dose rate human immunodeﬁciency virus hereditary non-polyposis colorectal cancer high power ﬁeld human placental lactogen human papillomavirus hormone replacement therapy International Association of Cancer Registries International Agency for Research on Cancer International Classiﬁcation of Disease International Collaborative Ovarian Neoplasm international units intravenous urogram local anaesthesia lactate dehydrogenase large loop excision of the transformation zone minimal access surgery minimum data set for cancer registration multidisciplinary team malignant mixed Müllerian tumour malignant ovarian germ cell tumour magnetic resonance (imaging) Medical Research Council maximum tolerated dose National Health Service Cervical Screening Programme not otherwise speciﬁed non-steroidal anti-inﬂammatory drug Oﬃce for National Statistics pulmonary embolism positron emission tomography partial hydatidiform mole pseudomyxoma peritonei cisplatin, vincristine, methotrexate, bleomycin positive predictive value placental site trophoblastic tumour cisplatin, vinblastine, bleomycin

xiii

List of abbreviations

RMI RRS SCC SCCA SCJ SIL SLCT SLND STI TAH TNM TP-TE TVS TZ USS VAC VaIN VIA VIAM VILI VIN VTE WHO YST

xiv

risk of malignancy index risk reduction surgery squamous cell carcinoma squamous cell carcinoma antigen squamo-columnar junction squamous intraepithelial lesion Sertoli–Leydig cell tumour sentinel lymph node dissection sexually transmitted infection total abdominal hysterectomy Tumour, Node, Metastases (staging system) paclitaxel, cisplatin alternating with paclitaxel, etoposide transvaginal scan transformation zone ultrasound scan vincristine, doxorubicin (adriamycin), cyclophosphamide vaginal intraepithelial neoplasia visual inspection with acetic acid visual inspection with acetic acid and magniﬁcation visual inspection with Lugol’s iodine vulval intraepithelial neoplasia venous thromboembolism World Health Organization yolk sac tumour

1

Chapter

Epidemiology of gynaecological cancers Anjum Memon

Introduction Gynaecological cancers encompass a diverse group of tumours with diﬀerent epidemiological and pathological features, clinical presentations and treatment strategies. This chapter aims to provide an overview of basic concepts in cancer epidemiology and to describe the global patterns and trends in incidence and mortality, aetiology and prevention of the ﬁve main types of gynaecological cancer.

What is epidemiology? Epidemiology is the basic science underpinning public health and clinical medicine. It describes the occurrence of health-related states or events (incidence, prevalence), quantiﬁes the risk of disease (relative risk) and its outcome (prognosis, survival, mortality), and postulates causal mechanisms for disease in populations (aetiology, prevention). The main function of epidemiology is to provide evidence to guide public heath policy and clinical practice to protect, restore and promote health. The applications of epidemiology can be summarized as follows: *

*

*

*

*

*

To describe the spectrum and extent of disease in the population – e.g. what is the prevalence of human papillomavirus (HPV) infection among young girls? To identify factors that increase or decrease the risk of disease – e.g. what factors increase the risk of, or protect against, endometrial cancer? To study the natural history and prognosis of disease – e.g. does early diagnosis of cervical carcinoma in situ through cytological screening prevent future morbidity and improve survival? To monitor and predict disease trends in the population – e.g. what impact will the increasing prevalence of obesity in women have on future disease trends and healthcare needs? To provide evidence for developing public health policy and making regulatory decisions – e.g. will a smoking ban in public places promote smoking cessation and reduce the incidence of smoking-related disease? To evaluate the eﬃcacy of preventive and therapeutic interventions – e.g. does postmenopausal hormone replacement therapy (HRT) do more harm than good?

Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 1: Epidemiology of gynaecological cancers

*

*

To evaluate public health programmes – e.g. will the mandatory vaccination of school girls against oncogenic HPV prevent vulvar/vaginal/cervical cancers and save lives? To evaluate the eﬀectiveness of health services – e.g. are known contacts of people with sexually transmitted diseases (STDs) followed up and treated?

Classiﬁcation of gynaecological cancers *

*

*

The International Classiﬁcation of Diseases (ICD): this is the global standard diagnostic classiﬁcation for epidemiological, clinical and health service data. It is used by hospital records departments, cancer registries and government agencies responsible for collection of health statistics (e.g. the Oﬃce for National Statistics in the UK) to classify diseases and other health problems recorded on many types of health and vital records. The ICD is essential for compilation of morbidity (e.g. cancer incidence) and mortality statistics (e.g. underlying cause of death) and allows comparison at an international level of health data collected in diﬀerent countries at diﬀerent times. In the Tenth Revision of the ICD (ICD-10), the malignant neoplasms of female genital organs are coded from C51 to C58. The category C57 includes neoplasms of the fallopian tube, broad and round ligaments, uterine adnexa and overlapping lesions (e.g. tubo-ovarian) (Table 1.1). The International Classiﬁcation of Diseases for Oncology (ICD-O-3): this is used principally by cancer registries for coding the site (topography) and the histology (morphology) of neoplasms, usually obtained from a pathology report. The Tumour Node Metastasis Classiﬁcation of Malignant Tumours (TNM-6): this is a cancer staging system used for describing the anatomical extent of cancer (see Appendix 3). It is based on the assessment of three components: T – describes the extent of the primary tumour; N – describes the absence/presence and extent of regional lymph node involvement/metastasis; and M – describes the absence/presence of distant metastasis. The classiﬁcation is used to: (i) aid the clinician in the planning of treatment; (ii) assist in evaluation of the results of treatment; (iii) give some indication of prognosis; and (iv) facilitate the exchange of information between treatment centres. The TNM system is approved by the International Federation of Gynecology and Obstetrics (FIGO), and its categories have been deﬁned to correspond to the FIGO classiﬁcation.

Measuring the risk or burden of gynaecological cancers Incidence Incidence (or incident cases) is a count of new cases of cancer in the population during a speciﬁed time period. The incidence rate is the number of new cases of cancer in a deﬁned population within a speciﬁed time period (usually a calendar year), divided by the total number of people in that population. Cancer incidence rates are typically expressed as per 100 000 population. Incidence rate measures the rapidity (or ‘speed’) at which new cases of cancer are occurring in the population within a speciﬁed time period. Increase in incidence of a cancer in the population can be due to: in-migration of susceptible people, a change in diagnostic criteria, improved case ascertainment, introduction of a new screening/diagnostic test, introduction of new, or changes in exposure to, existing aetiologic agent(s). Incidence rate is used to: predict

2

Chapter 1: Epidemiology of gynaecological cancers

Table 1.1 ICD-10 codes and morphological classiﬁcation of malignant neoplasms of female genital organs

ICD-10 code

Organ

Morphological subtypes

C51

Vulva

Squamous cell carcinoma Extramammary Paget’s disease Malignant melanoma

C52

Vagina

Squamous cell carcinoma Adenocarcinoma, botryoid Rhabdomyosarcoma

C53

Cervix uteri

Squamous cell carcinoma Neuroendocrine carcinoma Adenocarcinoma

C54

Corpus uteri

Endometrial adenocarcinoma Malignant mixed Müllerian tumours Leiomyosarcoma

C55

Uterus (part unspeciﬁed)

C56

Ovary

Surface epithelial tumours Serous adenocarcinoma Endometrioid adenocarcinoma Mucinous adenocarcinoma Clear cell carcinoma Germ cell tumours Dysgerminoma Yolk sac tumour Choriocarcinoma Mature and immature teratoma Sex cord-stromal tumours Granulosa cell tumour Sertoli–Leydig cell tumour

C57

Other and unspeciﬁed female genital organs

C58

Placenta

Hydatidiform mole Placental site trophoblastic tumour Choriocarcinoma

the average risk of developing cancer; research causes and treatment of cancer; describe trends of cancer over time; and evaluate the eﬀectiveness of prevention programmes.

Age-standardized incidence (or mortality) rate (ASR) As the risk of cancer increases exponentially with age, the crude incidence rate, which is inﬂuenced by the population age structure, cannot be used to evaluate whether the risk/

3

Chapter 1: Epidemiology of gynaecological cancers

burden of cancer diﬀers between populations. It is therefore necessary to use ASRs when comparing rates of populations that have diﬀerent age structures. The ASR is obtained by applying the (crude) age-speciﬁc rates in the observed population to the age-speciﬁc population counts (or weights) of a ﬁxed reference (or standard) population. The most commonly used standard population is the world (and also European) standard population of Doll. Age-standardization controls for the confounding eﬀect of age on cancer incidence and allows direct comparison of diﬀerent populations.

Cumulative incidence (or cumulative risk) Cumulative incidence is the probability or risk of developing cancer during a speciﬁed period (e.g. lifetime). It measures the number or proportion of people (out of 100 or 1000) who would be expected to develop a particular cancer by the age of 64 (or 74) years if they had the rates of cancer currently observed. Like the ASR, cumulative incidence permits comparisons among populations of diﬀerent age structures. For example, the cumulative risk of a woman in the UK developing ovarian cancer by age 74 is 14 per 1000, which can be interpreted as a 1.4% (1 in 71) probability or (lifetime) risk of developing ovarian cancer by the time she completes 74 years.

Prevalence Prevalence is the number of existing cases of cancer in a deﬁned population at a notional point in time, divided by the total number of people in the population at that time. It is usually expressed as an absolute number of existing cases or as the proportion of a population that has the disease. For example, the prevalence of cervical cancer can be deﬁned as the number of women in a deﬁned population who have been diagnosed as having the cancer, and who are still alive at a given point in time. Mathematically, prevalence can be deﬁned as follows: let a = the number of individuals in the population with the disease at a given time let b = the number of individuals in the population without the disease at a given time then prevalence ¼

a aþb

Thus prevalence is a measure of the burden of cancer in the population. The prevalence of a cancer in the population can increase due to: in-migration of cases, increase in incidence and/or improved prognosis/survival (e.g. due to better treatment). Prevalence data are used for planning health services, resource allocation and organization of prevention programmes.

4

*

Partial prevalence is the estimation of the number of cases of cancer diagnosed within one, three and ﬁve years to indicate the number of patients undergoing initial treatment (cases within one year of diagnosis), clinical follow-up (within three years) or not considered cured (before ﬁve years). Patients alive ﬁve years after diagnosis are usually considered cured because, for most cancers, the death rates among such patients are similar to those in the general population.

*

Complete prevalence represents the proportion of patients alive on a certain day who previously had a diagnosis of cancer, regardless of how long ago the diagnosis was, or if the patient is still under treatment or is considered cured.

Chapter 1: Epidemiology of gynaecological cancers

Survival Survival is the proportion (%) of people still alive one, three, ﬁve and 10 years after they have been diagnosed as having cancer. This observed survival probability is inﬂuenced by mortality both from the cancer itself and from other causes. For this reason, relative survival (%) is usually calculated (ratio of the observed survival in a particular group of patients to the survival expected in a group of people in the general population).

Quality-adjusted life-years (lost)/disability-adjusted life-years (lost) Quality-adjusted life-years (QALYs) and disability-adjusted life-years (DALYs) quantify the spectrum of morbidity (between the diagnosis and cure/death) due to cancer in terms of its duration and severity. The calculation of these indices requires three elements: the incidence of cancer, its mean duration (survival probability) and a measure of life ‘quality’ between the diagnosis and cure/death. These indices are used to estimate the impact of cancer on the individual and society, and for establishing priorities for healthcare programmes.

Mortality Mortality is the number of deaths occurring, and mortality rate is the number of deaths in a deﬁned population within a speciﬁed time period (usually a calendar year), divided by the total number of persons in that population. Cancer mortality rates are expressed as per 100 000 persons per year. Mortality is the product of the incidence and the fatality of a given cancer, and measures the average risk to the population of dying from a speciﬁc cancer within a speciﬁed period. Fatality, the complement of per cent survival, is the probability (%) that a cancer patient will die from the disease.

Cancer screening Screening is the presumptive identiﬁcation of an unrecognized disease or defect by the application of tests, examinations or other procedures that can be applied rapidly. Cancer screening is the testing of apparently healthy volunteers from the general population for the purpose of separating them into high and low probabilities of having a given cancer. The rationale behind cancer screening is that the disease has a natural history that includes a clearly deﬁned preclinical phase with biological characteristics, which allows for detection of the disease in an early treatable stage that, in turn, will reduce the risk of future morbidity and improve survival (e.g. cytological screening for carcinoma in situ of uterine cervix → intervene with surgery → cure/reduced risk of invasive cervical cancer). Randomized controlled trials and both case–control and cohort observational study designs are used to evaluate cancer screening programmes. The performance of a screening test is based on its sensitivity, speciﬁcity and predictive value (Table 1.2). *

*

Sensitivity – this is the ability of the test to identify correctly those who have the disease (true positives). Speciﬁcity – this is the ability of the test to identify correctly those who do not have the disease (true negatives).

5

Chapter 1: Epidemiology of gynaecological cancers

Table 1.2 Calculation of sensitivity, speciﬁcity and predictive value of a screening test

Disease according to gold standard

Screening test result

Positive

Present

Absent

Total

A (True +)

B (False +)

A+B

Negative

C (False −)

D (True −)

C+D

Total

A+C

B+D

A+B+C+D

Sensitivity = A/(A + C) × 100 (%) Speciﬁcity = D/(B + D) × 100 (%) Positive predictive value = A/(A + B) × 100 (%) Negative predictive value = D/(C + D) × 100 (%) Prevalence of disease = A + C/(A + B + C + D) × 100 (%)

*

*

Positive predictive value (PPV) – this is the proportion of individuals who test positive and actually have the disease. PPV is a function of sensitivity, speciﬁcity and disease prevalence. A high PPV is essential for a successful screening programme, whereas a low PPV implies that resources are being wasted on diagnostic follow-ups of false-positive individuals. Negative predictive value – this is the proportion of individuals who test negative and actually do not have the disease.

Ovarian cancer Incidence and mortality Worldwide, ovarian cancer is the sixth most common cancer among women, with an estimated 204 000 new cases (4% of cancer in women) and 125 000 deaths (4.3% of cancer deaths in women) in the year 2002 and a ﬁve-year prevalence of 538 000 cases. The incidence rates of ovarian cancer vary from a low of about 2 per 100 000 women in Algeria to a high of 15 per 100 000 in Poland. In general, incidence rates are relatively higher in developed countries – the highest rates are observed in European populations – followed by the Philippines and Brazil (11–15/100 000 women), intermediate rates are observed in North America, Australia/New Zealand and some populations in Asia and South America, and the lowest rates (2–7/100 000 women) are observed in Africa, the Caribbean, the Middle East and parts of Asia. Evidence from the USA suggests that ovarian cancer is relatively more common in white than in black women (10.6 vs. 7.2/100 000 women). In the UK, ovarian cancer is the most common gynaecological cancer. It is the fourth most common cancer among women, with 6615 new cases in the year 2004 accounting for around 5% of all cancers in women, with a cumulative risk of 0.8% (1 in 125) by age 64 (Fig. 1.1). In most European populations, the incidence rates of ovarian cancer generally increase exponentially with age, with a sharp increase after about 40 years. It is predominantly a disease of older, postmenopausal women – almost 85% of cases occur in women aged over 50 (Fig. 1.2). In developed countries, the mortality/incidence ratio is 56% and ovarian cancer accounts for more deaths than all the other gynaecological cancers

6

Chapter 1: Epidemiology of gynaecological cancers

Breast

31

Colorectal

12

Lung

11

Ovary

5

Uterine corpus

5

Malignant melanoma

3

Non-Hodgkin’s

3

Type of cancer (%)

Pancreas

3

Stomach

2

Leukaemia

2

Bladder

2

Uterine cervix

2

Oesophagus

2

Kidney

2

Brain, CNS

1

Multiple myeloma

1

Oral

1

Thyroid

1

Liver

1

Vulva

1

0%

5%

10%

15%

20%

25%

30%

35%

Fig. 1.1 Frequency distribution (%) of the 20 most common cancers in women (UK, 2004).

put together. In 2005, 4447 women in the UK died from ovarian cancer, accounting for around 6% of all female deaths from cancer. Ovarian cancer has a relatively poor prognosis as most cases (about 75%) are diagnosed at an advanced stage – with overall survival rates of about 30–40%. When diagnosed at the localized stage, the ﬁve-year relative survival rate is about 75%.

Trends in incidence and mortality In most developed countries there has been little change in the incidence of ovarian cancer over the past 40 or so years. In the UK, the age-standardized (European standard) incidence rates of ovarian cancer increased by 16% during the period 1975–2004 (from 14.7/100 000 to 17.1/100 000 women). Most of this increase was observed in women aged over 65, in whom the rates increased by 48% during that period. The incidence in younger women has

7

Chapter 1: Epidemiology of gynaecological cancers

Uterine cervix

Uterine Corpus

Ovary

Vulva

Vagina

70

60

Rate per 100 000

50

40

30

20

10

0 0–

5– 10– 15– 20– 25– 30– 35– 40– 45– 50– 55– 60– 65– 70– 75– 80– 85+ Age at diagnosis (years)

Fig. 1.2 Age-speciﬁc incidence rates of gynaecological cancers (West Midlands, UK, 1998–2002). (Source: Curado MP, Edwards B, Shin HR, et al., 2007.

remained fairly stable or may be declining – which in part is owing to the protective eﬀects of oral contraceptives widely used by younger women. In most populations, mortality rates have remained fairly stable or declined slightly over the past 40 years. In the UK, the agestandardized (European standard) mortality rates have remained stable over this period at between 10/100 000 and 12/100 000 women.

Aetiology Compared with other gynaecological cancers, little is known about the aetiology of ovarian cancer. In most studies, family history of ovarian cancer has been associated with an increased risk, whereas oral contraceptive use, increased parity, breastfeeding, tubal ligation and hysterectomy have been associated with decreased risk.

Prevention Apart from prophylactic oophorectomy, oral contraception and (possibly) tubal ligation, there are few readily modiﬁable risk factors for ovarian cancer. There is currently inconsistent evidence for a possible increase in risk with consumption of lactose/galactosecontaining foods, intake of saturated or animal fats and use of perineal talcum powder and postmenopausal HRT, and for a decrease in risk with consumption of vegetables. It is unclear whether obesity, body mass index (BMI) and physical activity inﬂuence ovarian cancer risk.

8

Chapter 1: Epidemiology of gynaecological cancers

Endometrial cancer Incidence and mortality Worldwide, endometrial cancer is the seventh most common cancer among women, with an estimated 199 000 new cases (3.9% of cancer in women) and 50 000 deaths (1.7% of cancer deaths in women) in the year 2002 and a ﬁve-year prevalence of 776 000 cases. In contrast with cervical cancer, endometrial cancer is relatively more common in developed countries. About 69% of the cases occur in developed countries where it accounts for 5.9% of all cancers in women, with a cumulative risk of 1.3% (1 in 77) by age 64. In developing countries, endometrial cancer accounts for only 2.3% of cancers in women, with a cumulative risk of 0.2% (1 in 500) by age 64. In the UK, endometrial cancer is the ﬁfth most common cancer among women, with 6438 new cases in the year 2004 accounting for around 5% of all cancers in women, with a cumulative risk of 0.8% (1 in 125) by age 64 (Fig. 1.1). In most European populations, the incidence rates of endometrial cancer begin to rise steadily 5–10 years before the menopause and reach a peak usually around 65–70 years (Fig. 1.2). It is a cancer of postmenopausal women – over 90% of cases occur in women aged 50 or older, with very few cases diagnosed under the age of 35. The incidence rates of endometrial cancer vary from a low of 0.9 per 100 000 women in Oman to a high of 18.8 per 100 000 in white women in the USA . The highest rates are observed among women in North America, Europe, Australia/ New Zealand and Israel. Incidence rates are generally low in countries in Latin America, Asia (including China and Japan) and Africa. Evidence from the USA suggests that endometrial cancer is considerably more common in white than in black women (18.8 vs. 13.5/100 000 women). Mortality rates are substantially lower than the incidence. Worldwide, the mortality/incidence ratio is 25%. In Europe, the cumulative mortality rates are generally three to four times lower than the incidence; in North America, these rates are about eight times lower. Endometrial cancer has a relatively better prognosis than cervical cancer, with ﬁveyear survival rates of 86% in the USA and 78% in European cancer registries.

Trends in incidence and mortality The trends in incidence of endometrial cancer have varied among populations and by age groups over the past 40 or so years. In the UK, the overall incidence (i.e. all ages combined) remained stable between 1975 and 1992, and then increased by 24% between 1993 and 2004. Most of this increase has occurred in older women – since the mid 1980s, there has been a steady increase in incidence among women aged over 60. Generally, similar trends in incidence are observed in most other European countries. The mortality rates, however, have steadily declined in most developed countries over this time period. In the UK, the agestandardized (European standard) mortality rates decreased by 27% between 1971 and 2005 (from 4.8/100 000 to 3.5/100 000).

Aetiology In contrast with cervical cancer, which is a model of viral carcinogenesis, endometrial cancer is a model of hormonal carcinogenesis. The 10-fold variation in age-standardized incidence rates across populations point to the role of modiﬁable factors in the aetiology of endometrial cancer. Among these, oestrogens and progestins are considered to have an important role in malignant transformation. The most compelling evidence has come from studies of HRT

9

Chapter 1: Epidemiology of gynaecological cancers

following the menopause – unopposed oestrogen use for 10 or more years increases the risk about 10-fold. This excess risk can be counteracted substantially by combined use of oestrogens and progestins.

Prevention Obesity (which increases peripheral production of oestrogens) and diabetes mellitus are associated with an increased risk of developing endometrial cancer, whereas past oral contraceptive use, childbearing and physical activity (potentially mediated by hormones) are associated with risk reduction. It is therefore possible to substantially reduce the incidence of endometrial cancer through modiﬁcation of lifestyle, maintenance of normal weight and optimal use of oral contraceptives and postmenopausal HRT.

Cervical cancer Incidence and mortality Worldwide, cervical cancer is the second most common cancer among women, with an estimated 493 000 new cases (or 9.7% of cancer in women) and 274 000 deaths (or 9.3% of cancer deaths in women) in the year 2002 and a ﬁve-year prevalence of 1.4 million cases. About 83% of the cases occur in developing countries, where cervical cancer accounts for 15% of all cancers in women, with a cumulative risk of 1.5% (1 in 67) by age 64. In developed countries, cervical cancer accounts for only 3.6% of cancers in women, with a cumulative risk of 0.8% (1 in 125) by age 64. In the UK, 2726 new cases of cervical cancer were diagnosed in the year 2004, accounting for around 2% of all cancers in women, with a cumulative risk of 0.6% (1 in 167) by age 64 (Fig. 1.1). In most European populations, the incidence of cervical cancer begins to increase at 20–24 years and the risk increases rapidly to reach a peak usually around 35–39 years (Fig. 1.2). The incidence rates of cervical cancer vary substantially between diﬀerent regions, from a low of 2.8 per 100 000 women in China to a high of 47.3 per 100 000 in Zimbabwe. The highest rates are observed among the populations of sub-Saharan Africa, Melanesia, Latin America and the Caribbean, and south-central and south-east Asia. Incidence rates are generally low in developed countries in Europe, North America, Australia/New Zealand, the Middle East, China and Japan. Mortality rates are substantially lower than the incidence. Worldwide, the mortality/ incidence ratio is 55%. Five-year relative survival rates vary between regions and according to the extent of the disease (15–80%), with good prognosis in countries with a low incidence (63% in the European cancer registries).

Trends in incidence and mortality Overall, incidence and mortality have declined considerably during the past 40 years in Western Europe, North America, Australia/New Zealand, China and Japan. The decline has been attributed to a combination of factors, including improved genital hygiene, improved treatment modalities and the beneﬁcial eﬀects of organized population-based cytological screening programmes. In the UK, the age-standardized (European standard) incidence rates of cervical cancer declined by around 46% during the period 1975–2004 (from 14.9/100 000 to 8.0/100 000), whereas, in the same period, the mortality rates declined by around 63% (from 7.5/100 000 to 2.8/100 000).

10

Chapter 1: Epidemiology of gynaecological cancers

Aetiology A persistent infection with an oncogenic HPV type is now recognized as a causal factor for preceding precancerous changes and cervical cancer. However, infection with HPV is extremely common compared with the relatively rare development of cervical cancer. There is compelling evidence that HPV is necessary for cervical carcinogenesis, but infection alone is not suﬃcient for the cancer to develop. A number of co-factors have been identiﬁed as possible modiﬁers of HPV infection during the developmental stages of cervical cancer, including early sexual debut, increasing number of sexual partners, smoking, long-term oral contraceptive use, high parity, dietary factors, certain human leucocyte antigen (HLA) types and co-infection with other sexually transmitted agents such as Chlamydia trachomatis, herpesvirus type 2 and the human immunodeﬁciency virus (HIV).

Prevention Increased public awareness about sexually transmitted infections, early detection of precursor lesions by regular cytological screening, HPV testing and the recently developed vaccine against certain high-risk types of HPV makes cervical cancer one of the most preventable forms of cancer, on a global scale.

Vulvar and vaginal cancers Incidence and mortality Vulvar and vaginal cancers are rare throughout the world, and constitute less than 5% of all gynaecological cancers. The incidence rates of vulvar cancer vary from a low of 0.2 per 100 000 women in Japan to a high of 2.0 per 100 000 in Peru. In general, incidence rates are relatively higher in developed countries – the highest rates are observed in some European populations and in white women in the USA, followed by Peru and Brazil. The lowest rates (<1/100 000 women) are observed in parts of Asia, Africa, the Middle East and South America. In the UK, there were 1022 new cases of vulvar cancer in the year 2004, accounting for around 0.7% of all cancers in women (Fig. 1.1). Vulvar cancer is predominantly a disease of older women, with a steep rise in incidence after 60+ years – from 4.2 per 100 000 to 30.5 per 100 000 women by age 85+ (Fig. 1.2). Each year 300–400 women die from vulvar cancer in the UK. The overall incidence rates of vaginal cancer do not exceed 0.8 per 100 000 women in any world region. In the UK, there were 246 new cases of vaginal cancer in the year 2004, accounting for around 0.2% of all cancers in women. Like vulvar cancer, it is predominantly a disease of older women, with a steep rise in incidence after 60+ years – from 0.7 per 100 000 to 4.1 per 100 000 women by age 85+ (Fig. 1.2). Each year around 100 women die from vaginal cancer in the UK. Five-year relative survival rates for vulvar and vaginal cancers vary signiﬁcantly by stage of disease and age at diagnosis. In most developed countries, the survival rates of vulvar cancer vary between 31% and 98%; vulvar cancer has a relatively better prognosis than vaginal cancer (ﬁve-year survival, 40–72%). In the UK, the overall survival rate for vulvar and vaginal cancers combined has improved from 40% to 58% over the past 40 or so years.

11

Chapter 1: Epidemiology of gynaecological cancers

Trends in incidence and mortality In most developed countries there has been little change in the overall (i.e. all ages combined) incidence of vulvar and vaginal cancers over the past 40 or so years. In the UK, the incidence rate of vulvar cancer has remained steady at around 2 per 100 000 women and for vaginal cancer around 0.6 per 100 000 during this period. However, recently there has been some increase in the rates of vulvar cancer among young women (< 50 years) in many countries. This has been linked to increasing incidence of vulvar intraepithelial neoplasia (VIN) in young women, which is caused by infection with HPV. In most populations, the mortality rates of vulvar and vaginal cancers have declined steadily over the past 40 or so years. In the UK, the age-standardized mortality rates (European standard) for vulvar cancer declined by 54% and for vaginal cancer by 45% during this period.

Aetiology A persistent infection with an oncogenic HPV type is now considered to have a central role in the initiation and promotion of the majority of vulvar and vaginal cancers. HPV is more strongly associated with cancers in younger women, and about 70% of in situ squamous cell carcinomas and 20–50% of invasive vulvar cancers contain HPV DNA. About 80% of in situ and 60% of invasive vaginal squamous cell carcinomas contain HPV DNA. A history of cervical intraepithelial neoplasia (CIN) or cervical cancer is considered a strong risk factor for vulvar and vaginal cancers. A number of HPV co-factors have been identiﬁed, including a history of genital warts (which are caused most commonly by non-oncogenic HPV types), smoking and infection with other sexually transmitted agents such as herpesvirus type 2 and HIV. Iatrogenic immune suppression in transplant recipients has been associated with a 100-fold increased risk of vulvar cancer.

Prevention Increased public awareness about sexually transmitted infections, surveillance for precancerous lesions, self-awareness, smoking cessation, HPV testing and HPV vaccination make vulvar and vaginal cancers one of the most preventable forms of cancer.

Further reading

12

*

Adami H-O, Hunter D, Trichopoulos D, eds. Textbook of Cancer Epidemiology, 2nd edn. Oxford: Oxford University Press, 2008.

*

Cancer Research UK. (http://info. cancerresearchuk.org).

*

Curado MP, Edwards B, Shin HR, et al., eds. Cancer Incidence in Five Continents, Vol. IX. IARC Scientiﬁc Publications no. 160. Lyon: IARC Press, 2007.

*

Ferlay J, Bray F, Pisani P, Parkin DM. GLOBOCAN 2002: Cancer Incidence, Mortality and Prevalence Worldwide. In IARC CancerBase No. 5. Version 2.0. Lyon: IARC Press, 2004. Available at: www-dep.iarc.fr.

*

International Union Against Cancer. Tumour Node Metastasis Classiﬁcation of Malignant Tumours (TNM-6). Available at: www.uicc.org.

Chapter 1: Epidemiology of gynaecological cancers

*

Oﬃce for National Statistics. (www.statistics. gov.uk).

*

Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005; 55: 74–108.

*

Schottenfeld D, Fraumeni JF, eds. Cancer Epidemiology and Prevention, 3rd edn. Oxford: Oxford University Press, 2006.

*

World Health Organization. International Classiﬁcation of Diseases, Tenth Revision (ICD-10). Available at: www.who.int/ classiﬁcations/icd/en.

*

World Health Organization. International Classiﬁcation of Diseases for Oncology (ICD-O-3). Available at: www.who.int/ classiﬁcations/icd/adaptations/oncology/en/ index.html.

13

2

Chapter

Screening and prevention of gynaecological cancers Ranjit Manchanda and Ian J. Jacobs

Introduction A screening test is a test used to detect disease in asymptomatic individuals. A good screening test should have high speciﬁcity (probability of the test being negative in individuals without the disease) and high sensitivity (probability of the test being positive in individuals with the disease). It is also important to consider the problems of over-diagnosis, misdiagnosis and potential adverse eﬀects linked to any screening programme. Screening the asymptomatic population for early-stage disease and premalignant conditions is an established strategy for early detection and prevention of cancer. Population screening strategies are based on the original criteria described by Wilson and Junger (Table 2.1). Mass screening for cervical cancer fulﬁls most of these criteria, and establishment of screening programmes in diﬀerent countries has led to a marked reduction in the incidence of and mortality associated with cervical cancer. Ovarian cancer is the other gynaecological malignancy for which population screening may become justiﬁed. Although eﬀective treatment is available for early-stage disease, the disease is usually diagnosed in advanced stages, when the prognosis for long-term survival is poor. There is preliminary evidence that early detection may increase long-term survival and ongoing research trials are addressing this issue. Mass screening of the low-risk population for endometrial cancer is unlikely to be of beneﬁt, as most women present in early stages with symptomatic disease and have a good prognosis. However, screening of ‘high-risk’ populations is recommended. Screening is not justiﬁed for vaginal and vulval cancers as they are rare diseases, though increasing awareness of these conditions is important.

Ovarian and fallopian tube cancer Ovarian cancer is the commonest gynaecological cancer in England and Wales, with women having a 2.1% lifetime risk of developing the disease. It is also the most lethal, due to the advanced stage at which most women present. Around 85% of ovarian cancers occur over the age of 50 and 80–85% of cancers are epithelial in origin.

Screening for ovarian and tubal cancer Our knowledge of the molecular and biological events in ovarian carcinogenesis is limited and a true precursor lesion for ovarian cancer has not been identiﬁed. This has restricted the ability to screen for this disease and limited the goal of screening to detection of asymptomatic, early-stage disease. Over the years a number of diﬀerent modalities have been used to Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 2: Screening and prevention

Table 2.1 World Health Organization criteria for a screening programme (Wilson and Junger, 1968) 1. The condition sought should be an important health problem. 2. There should be accepted treatment for patients with recognized disease. 3. Facilities for diagnosis and treatment should be available. 4. There should be a recognizable latent or early symptomatic stage. 5. There should be a suitable test or examination. 6. The test should be acceptable to the population. 7. The natural history of the condition, including development from latent to declared disease, should be adequately understood. 8. There should be an agreed policy on whom to screen. 9. The cost of case-ﬁnding (including diagnosis and treatment of patients diagnosed) should be economically balanced in relation to possible expenditure on medical care as a whole. 10. Case-ﬁnding should be a continuing process and not a ‘once and for all’ project.

screen for ovarian cancer. Although screening can detect cancers early and there is evidence of a survival beneﬁt, it has not yet been shown to be eﬀective at reducing mortality.

Gynaecological examination Pelvic/vaginal examination lacks adequate sensitivity and speciﬁcity to be used as a screening tool for ovarian cancer.

Tumour markers A wide variety of macromolecular tumour antigens, including enzymes, hormones, receptors, growth factors, biological response modiﬁers and glycoconjugates, have been investigated as potential tumour markers. Novel biomarkers are constantly being identiﬁed by innovative screening strategies using new micro-array- and mass spectrometry-based technologies exploring DNA, RNA and protein overexpression. A detailed review of these is beyond the scope of this chapter. The most widely investigated tumour marker in ovarian cancer is cancer-associated antigen (CA)125, and it was ﬁrst described by Bast in 1981. Approximately 85% of women with epithelial ovarian cancer (EOC) have CA125 levels of >35 U/ml, with elevated levels found in 50% of women with stage I disease and over 90% of women with stage II–IV disease. CA125 levels are less frequently elevated in mucinous and borderline tumours compared with serous tumours. However, the levels can also be elevated in other malignancies, pregnancy, inﬂammatory disorders, non-gynaecological diseases (such as pancreatitis, hepatitis, cirrhosis, ascites and tuberculosis) and benign gynaecological conditions (such as endometriosis, ﬁbroids, infections and pelvic inﬂammatory disease).

Improving CA125 sensitivity and speciﬁcity A speciﬁcity of 97% and positive predictive value (PPV) of 4.6% was achieved using CA125 (30 U/ml) in a prospective ovarian cancer screening study of 5550 volunteers aged over 50. Speciﬁcity of screening with CA125 was initially improved by the addition of pelvic ultrasound as a second-line test to assess ovarian volume and morphology. Transvaginal ultrasound is preferred to transabdominal ultrasound because of its better resolution, which

16

Chapter 2: Screening and prevention

allows more accurate assessment of ovarian morphology, and the lack of need for a full bladder. Persistence of abnormalities on repeat scanning four to six weeks following initial detection helps reduce false positive rates. A multimodal screening approach incorporating sequential CA125 measurements and pelvic ultrasound achieved a speciﬁcity of 99.9% and PPV of 26.8% (approximately four operations for each cancer) for detection of ovarian and fallopian tube cancer in 22 000 postmenopausal women. Further improvements in screening performance have been possible with the development of the statistical ‘risk of ovarian cancer’ algorithm (ROCA), which incorporates patient age and the rate of change of absolute CA125 values over time. The ROCA increases the sensitivity of CA125 compared with a single cut-oﬀ value because women with normal but rising levels are identiﬁed as being at increased risk. At the same time, speciﬁcity is improved, as women with static but elevated levels are now classiﬁed as low risk. For a target speciﬁcity of 98%, the ROCA calculation achieved a sensitivity of 86% for preclinical detection of ovarian cancer.

Improving discriminatory ability for benign and malignant adnexal masses Accurate ability to distinguish benign from malignant tumours is the key to avoiding unnecessary operations in women with benign lesions and planning appropriate surgery by gynae-oncologists in tertiary care centres for those with cancer. Using an upper limit of 35 U/ml for CA125, a sensitivity of 78%, speciﬁcity of 95% and PPV of 82% can be achieved for ovarian cancer. Using additional markers or artiﬁcial neural networks for interpretation can further increase sensitivity and speciﬁcity over CA125 alone. Various tumour marker panels have been reported to be better than CA125 alone: CA125, CA 72–4, CA 15–3 and lipid-associated sialic acid; leptin, prolactin, osteopontin, macrophage inhibitory factor, insulin growth factor (IGF)-II and CA125; CA125, SMRP, HE4, CA72–4, activin, inhibin, osteopontin, epidermal growth factor (EGFR) and ERBB2 (Her2). HE4 is emerging as a promising marker, with a combination of CA125 and HE4 having been shown to be a better predictor (sensitivity 76.4%, speciﬁcity 95%) than either marker alone. A recent report indicates it may hold promise for early-stage disease. The ‘risk of malignancy index’ (RMI), which combines serum CA125 values with ultrasound-based ovarian morphology and menopausal status, is a widely used discriminatory tool, which has been validated prospectively and retrospectively in gynaecological oncology and general gynaecological units. The risk score has a sensitivity of 85% and a speciﬁcity of 97%. It performs as well as most logistic regression or artiﬁcial neural network models. Women with an elevated RMI score have been shown to have, on average, a 42-fold increase in the background risk of ovarian cancer. The score’s sensitivity can be further increased by increasing the RMI cut-oﬀ and using artiﬁcial neural networks. Addition of specialist ultrasound, assessment for an ‘ovarian crescent sign’ and MRI further improves accuracy. The International Ovarian Tumour Analysis (IOTA) researchers recently reported an ultrasound-based pattern recognition model for discriminating benign from malignant ovarian masses, which they found to be superior to CA125 alone.

Screening trials An initial randomised trial in 21 935 women with ovarian cancer showed a markedly better median survival in women who underwent screening (72.9 months) compared with those who did not (41.8 months). Subsequently, a number of prospective trials have used CA125 alone or in combination with ultrasound to screen for ovarian cancer.

17

Chapter 2: Screening and prevention

Two large randomised controlled trials (RCTs) – the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS) (www.ukctocs.org.uk) and the Prostate Lung Colorectal and Ovarian Cancer Screening Trial (PLCO) – in the USA are evaluating the impact of screening on ovarian cancer mortality in the postmenopausal population. In the UKCTOCS trial over 200 000 postmenopausal women are being randomised to annual screening with CA125 (multimodal) or transvaginal ultrasound (TVS) and control groups. The ROC algorithm is used to triage women into low-, intermediate- and elevated-risk categories based on their CA125 result. Intermediate-risk women have repeat CA125 screening in six to eight weeks, while those with elevated risk are referred for a TVS. An abnormal TVS leads to gynaecological assessment with a view to surgery. As the eﬃcacy of the ROC algorithm in detecting early-stage disease is still a matter of research, the trial is also evaluating TVS alone as a ﬁrstline test for screening. In the PLCO trial, 78 000 postmenopausal women have been randomised to either screening with clinical examination, ultrasound and CA125 or to a control group. Any abnormal test leads to gynaecological assessment and further investigation. The recently published prevalence screen results of another study in 28 816 screened women included 29 cancers, of which 20 were invasive and 9 were borderline. The PPV for invasive cancer was 3.7%, 1% and 23.5% for abnormal CA125, TVS and both CA125 and TVS, respectively. A large Japanese RCT in which an annual ultrasound and absolute CA125based strategy was used in 82 487 low-risk postmenopausal women did not ﬁnd a signiﬁcant diﬀerence in the number of ovarian cancers detected in the screened arm (n = 27) and those occurring in the control arm (n = 32). Eight additional interval cancers occurred in the screened arm. The proportion of early-stage cancers was greater in the screened (63%) than the control group (38%). In the reproductive age group, screening for ovarian cancer is limited to research trials involving women at risk for familial ovarian cancer. These women belong to high-risk cancer families and many carry BRCA or mismatch repair gene (MMR) mutations. The sensitivity and eﬀectiveness of screening in this younger population is still not established. Further, multifocal primary peritoneal cancer is considered to be a phenotypic variant of familial ovarian cancer and both CA125 and ultrasound are not reliable for detecting early-stage disease. Most studies have used a combination of absolute CA125 levels and ultrasound. Annual screening using this modality is not eﬀective in detecting early-stage disease. A modiﬁed premenopausal version of the ROC algorithm is being piloted in phase II of the UK Familial Ovarian Cancer Screening Study (UKFOCSS) and in the Cancer Genetics Network (CGN) and Gynecologic Oncology Group (GOG) trials in the USA. These trials are evaluating more frequent three- to four-monthly screening. Preliminary results are available for the CGN trial, on 2343 high-risk women screened at three-monthly intervals. Thirty-eight women underwent surgery following 6284 screens. Five ovarian cancers were detected, two prevalent (one early, one late stage) and three incident (three early) cases, resulting in a PPV of 13%. The impact of ovarian cancer screening on mortality and the proportion of early cancers it will detect are not yet known. In addition, it is not known whether the screen-detected early-stage disease includes a signiﬁcant number of high-grade serous carcinomas or just the cancers with better histological prognosis. The current recommendation following international consensus is that screening should only be carried out within the context of a clinical trial. The ongoing trials are expected to report after 2012 and should help address the above issues.

18

Chapter 2: Screening and prevention

Prevention A number of factors, such as early menarche, sub-fertility, obesity, hormone replacement therapy, smoking and nulliparity, have been linked to an increase risk of ovarian cancer. Pregnancy, breastfeeding (relative risk [RR] 0.5–0.7), taking the contraceptive pill (odds ratio [OR] 0.4–0.6), being sterilized/having a hysterectomy (OR 0.5–0.7) and a diet rich in carotenoids (found in carrots and tomatoes) can reduce the risk of ovarian cancer. The beneﬁt gained by taking the contraceptive pill is cumulative and lasts well beyond the period of use. Although data on the beneﬁt of physical exercise in preventing ovarian cancer are inconsistent, a number of studies suggest that it may have a modest protective eﬀect.

Familial risk Most cancers are sporadic but up to 10% are associated with an inheritable genetic predisposition. Inherited mutations in the BRCA1, BRCA2 and MMR genes account for most of the known hereditary risk of breast, ovarian and endometrial cancer. MMR gene mutations are dealt with in a later section of this chapter. The prevalence of BRCA mutations in the general population is around 1 in 400 to 1 in 800. These mutations are found to occur more frequently in Ashkenazi Jews, and certain other population groups such as the Icelandic, Norwegian and Swedish communities. Three ‘founder’ mutations, two in BRCA1 (185delAG and 5382insC) and one in BRCA2 (6174delT), account for most of the BRCA-related risk in the Ashkenazi Jew population. These gene mutations are called ‘founder mutations’ because they are present in the population at an increased frequency (1 in 40) by virtue of having existed in a small isolated group of ‘founders’, who were the ancestors of most of the individuals in the present day population. BRCA mutation carriers have an increased risk for developing breast, ovarian, prostate and pancreatic cancer. Male breast cancer, prostate cancer and pancreatic cancer is linked more to BRCA2 than BRCA1 gene mutations. A recently published meta-analysis calculated the cumulative cancer risks for BRCA1 and BRCA2 mutations carriers at age 70 to be 57% and 49% for breast cancer and 40% and 18% for ovarian cancer, respectively. Much higher penetrance – over 80% for breast cancer and over 50% for ovarian cancer – has been documented in mutation carriers ascertained from high-risk families with multiple cancer cases. A number of risk models use family history-based criteria to predict carrier probability and identify these high-risk individuals. The criteria used by the UKFOCSS trial to identify women at high risk for ovarian cancer are given in Table 2.2. Individuals from high-risk families are managed in special carrier clinics by a multidisciplinary team consisting of clinical geneticists, gynae-oncologists, breast surgeons and psychologists with a special interest in familial cancer. They undergo extensive counselling regarding cancer risk, predictive gene testing and preventive options, which include screening/surgery and lifestyle issues. The decision-making process is complex and tailored to each patient.

Risk-reducing salpingo-oophorectomy Surgical removal of the fallopian tubes and ovaries can prevent cancer arising in the ovary or the tubes. However, a 2–4% residual risk of primary peritoneal cancer remains. Riskreducing salpingo-oophorectomy (RRSO) is an option oﬀered to women at high risk (greater than 10% lifetime risk) for ovarian cancer and is usually performed laparoscopically. RRSO also decreases the risk of breast cancer by about 50% in premenopausal women. Preliminary

19

Chapter 2: Screening and prevention

Table 2.2 Criteria to identify high-risk families (used by UKFOCSS)

Volunteer/proband should either have been aﬀected by cancer or be an FDR of an aﬀected family member. Families with ovariana or ovariana and breast cancer Families with only breast cancer 1. ≥2 individuals with ovarian cancer who are FDRs

1. ≥4 breast cancers

2. One ovarian cancer and 1 breast cancer <50 years who are FDRs

2. 3 breast cancers related by FDR:

3. One ovarian cancer and 2 breast cancers <60 years who are FDRs

a. one ≤30 years or b. all ≤40 years or c. one MBC and one bilateral breast cancer

4. Breast cancer in volunteer/proband (≤45 years) and mother with both breast and ovarian cancer (in the same person)

2. Breast cancer in volunteer/proband (≤50 years) and:

5. Breast cancer in volunteer/proband (≤40 years) and sister with both breast and ovarian cancer (in the same person)

b. breast cancer in mother (≤40 years onset) or

6. Criteria 1, 2 and 3 can be modiﬁed where paternal transmission is occurring, i.e. families are eligible where aﬀected relatives are related by second degree through an unaﬀected intervening male relative and there is an aﬀected sister.

Families with a known gene mutation

a. breast cancer in mother (age of onset being ≤30 years in one and ≤50 years in the other) or c. one MBC and one bilateral breast cancer 3. Two MBC (one <40 years) in the family and proband is a FDR of one of them

Families with Ashkenazi Jewish ethnicity (additional criteria)a

Ashkenazi Jewish ethnicity and any one of the following: 1. The family contains an aﬀected individual 1. Breast cancer (<40 years) or bilateral breast cancer with a mutation of one of the known (ﬁrst cancer <50 years) in volunteer/proband, OC-predisposing genes, e.g. BRCA1, BRCA2, MLH1, irrespective of family history of cancer MSH2, MSH6, PMS1 and PMS2 2. Breast cancer in volunteer/proband (<50 years) and one FDR with breast cancer (<50years) or ovarian cancer (any age) or MBC (any age) 3. Breast cancer in volunteer/proband (<60 years) and one FDR with breast cancer (<40 years) or ovarian cancer (any age) or MBC (any age) 4. One FDR with ovarian cancer (<50 years) 5. FDR with breast and ovarian cancer in the same woman (any age) 6. Two FDRs with breast cancer (<40 years) 7. Two MBC (<60 years) in the family and proband is an FDR of one of them

Families with colorectal cancer (HNPCC or Lynch syndrome) 1. The family contains ≥3 individuals with a HNPCC-related cancer,b who are FDR and ≥1 case is diagnosed before 50 years and the cancers aﬀect >1 generation a

History of tubal or primary peritoneal cancers may be considered equivalent to ovarian cancers. HNPCC-related cancers include: colorectal, endometrial, small-bowel, ureteric and renal pelvic cancer. FDR, ﬁrst-degree relative; OC, ovarian cancer; MBC, male breast cancer.

b

20

Chapter 2: Screening and prevention

analyses suggest that it also reduces overall and possibly cancer-speciﬁc mortality in these women. Occult cancer at histological examination has been reported in 4–5% of BRCA carriers undergoing prophylactic surgery. This may subsequently necessitate a full staging procedure. Rigorous operative and pathological sampling (serial sectioning at 2–3-mm intervals) has been shown to markedly increase the rate of detection of occult ovarian and fallopian tube cancers and alter subsequent management. A hysterectomy is not routinely recommended in these high-risk women in view of the small but signiﬁcant increase in the surgical risk. Peritoneal washings and endometrial sampling at the time of RRSO are part of standard protocols. Other issues involved in decision making include loss of subsequent fertility and the onset of menopause in premenopausal women. In addition, women also need to be counselled regarding the risks associated with any surgical procedure. Surgery is usually undertaken after the age of 40, when women have completed their family. It may also be considered up to ﬁve years before the age when the youngest person in the family was diagnosed with ovarian cancer. Post-surgery, these women are advised to take hormone replacement therapy (HRT) up to the age of 50. This helps oﬀset or ameliorate the deleterious eﬀects of surgical menopause, such as hot ﬂushes, sweats, reduced libido, osteoporosis and cardiovascular disease. Short-term HRT has not been found to negate the protective eﬀect of RRSO on subsequent breast cancer risk in BRCA1/2 mutation carriers. Although HRT has a positive impact on surgically induced vasomotor symptoms, it may be less eﬀective than is often assumed. Recent reports suggest that surgical menopause may be associated with increased mortality if performed before the age of 45, primarily in women who do not receive oestrogen up to the age of 45.

Endometrial cancer Endometrial cancer is the second most common cancer of the female genital tract. It accounts for 5% of cancers in women and is mainly seen in postmenopausal women. Just 7% occur in the reproductive age group and are mainly linked to familial predisposition, obesity and polycystic ovarian syndrome (PCOS). Most endometrial cancers (77%) are diagnosed at an early, favourable stage. The consensus is that screening for endometrial cancer is not warranted for women who have no identiﬁable risk factors.

Familial endometrial cancer Women from Lynch syndrome or hereditary non-polyposis colorectal cancer (HNPCC) families have a 40–60% lifetime risk of endometrial cancer and a 10–15% lifetime risk of ovarian cancer. Lynch syndrome is an autosomal dominant (AD) syndrome, characterized by the development of a number of diﬀerent cancers. Although colorectal cancer (CRC) and endometrial cancer are the commonest cancers found in Lynch syndrome families, the tumour spectrum also encompasses gastric, small-bowel, ovarian, hepatobiliary, ureteric and renal pelvic (upper urological tract) cancers and brain tumours. Current evidence does not support the inclusion of breast and prostate cancer in this spectrum. The population incidence of Lynch syndrome has been estimated to lie between 1:660 and 1:2000. Recent data suggest that MMR gene mutations are found in 2% of individuals with colorectal and ovarian cancer, unselected for age or family history, and 9% of women with endometrial cancer under the age of 50, unselected for family history. Stage for stage there is no signiﬁcant diﬀerence between survival rates for HNPCC-associated and sporadic endometrial or

21

Chapter 2: Screening and prevention

Table 2.3 Diagnostic criteria for Lynch syndrome Amsterdam IIa

1. 3 relatives with HNPCC-relatedb cancer, 1 an FDR of other 2 2. HNPCC-relatedb cancer aﬀecting more than one generation 3. At least 1 HNPCC-relatedb cancer <50 years

Revised Bethesdac

1. CRC in an individual diagnosed at <50 years 2. Synchronous, metachronous colorectal or other HNPCC-associated tumoursd, regardless of age 3. CRC with MSI-h histology diagnosed <60 years 4. CRC in ≥1 FDR with an HNPCC-related tumourd, with one of the cancers being diagnosed at <50 years 5. CRC in ≥2 FDRs or SDRs with HNPCC-related tumoursd, regardless of age

a

Each condition fulﬁlled. Includes colorectum, endometrium, small-bowel, ureter and renal pelvis. ≥1 fulﬁlled. d Includes tumours listed in footnote b, stomach, ovary, pancreas, biliary tract, brain tumours, sebaceous gland adenomas and keratoacanthomas in Muir–Torre syndrome. CRC, colorectal cancer; FDR, ﬁrst-degree relative; SDR, second-degree relative. b c

ovarian cancer. However, HNPCC-related CRC has better survival estimates than corresponding sporadic CRC. Historically, HNPCC has been diagnosed on the basis of strict clinical criteria based on family history, called the Amsterdam criteria. The original criteria for ascertaining HNPCC families have been modiﬁed over the years (Table 2.3) to include extracolonic tumours and better reﬂect the spectrum of disease. HNPCC or Lynch syndrome is caused by a mutation in one of the DNA MMR genes: MLH1, MSH2, MSH6, PMS1 and PMS2. Among the MMR genes, mutations in MLH1 and MSH2 are the most frequent, whereas MSH6 mutations account for a minority (15%). Risk models based on family history have been used to predict risk of carrying a gene mutation. These models complement the clinical criteria used to identify high-risk families. The Bethesda criteria (Table 2.3) are less stringent than the strict pedigree-based Amsterdam criteria and have higher sensitivity but lower speciﬁcity. These criteria are used to select individuals for additional molecular studies (of tumour) using microsatellite instability (MSI) and/or immunohistochemistry (IHC) analysis. Presence of MSI and negative staining on IHC is used to identify individuals who may beneﬁt from genetic analysis. This approach is deemed to increase cost-eﬀectiveness of genetic testing. If MSI or IHC tissue testing is not possible or there is a strong clinical suspicion of Lynch syndrome despite negative or uninformative tests, directly proceeding to genetic testing of the proband has also been advised.

Screening High-risk population Although several strategies have been used to screen for endometrial cancer in women with Lynch syndrome, the eﬃcacy of endometrial screening in these women remains unproven.

22

Chapter 2: Screening and prevention

Interval cancers occur despite screening and the impact of screening on morbidity/mortality is unknown. Notwithstanding these shortcomings, screening may have a role in women who wish to delay or avoid preventive surgery. The main screening modalities that have been used for endometrial cancer are TVS and endometrial sampling (with or without hysteroscopy). However, available data are limited and there is a lack of evidence to recommend any particular method of screening. Most at-risk women are likely to be premenopausal and a proportion of them may not present with irregular bleeding. Although TVS has been used as a ﬁrst-line screening tool, there is lack of consensus on an appropriate cut-oﬀ value for endometrial thickness in asymptomatic premenopausal women and interval cancers are known to occur. Pipelle endometrial biopsy is a well-established method for endometrial sampling that can be performed as an outpatient procedure and is well tolerated by women. However, it has a tissue yield and procedure failure rate of around 10%, and inadequate samples are more common in the postmenopausal age group. Cancers have been reported to be missed on pipelle alone. A large meta-analysis reported the sensitivity of pipelle for diagnosing endometrial cancer to be 99% in postmenopausal women and 91% in premenopausal women, with a speciﬁcity of >98%. For atypical hyperplasia the sensitivity fell to 81% while speciﬁcity was maintained at 98%. Hysteroscopy-directed endometrial sampling remains the accepted gold-standard and is now routinely performed as an outpatient procedure. Hysteroscopy can detect polyps and submucous ﬁbroids, which may be missed by both ultrasound and pipelle. It may also permit directed biopsy from a focal lesion that may not otherwise be detected. Overall, ﬁve endometrial cancers have been recently reported in two prospective series using annual hysteroscopy for endometrial surveillance in 119 high-risk women. The rate of failure of outpatient hysteroscopy was 8% and 11% in these two series and women with cancer were found to present with abnormal bleeding up to three months prior to diagnosis. Two interval and 11 screen-detected cancers were found in 175 Finnish women undergoing yearly surveillance with ultrasound and endometrial biopsy. The interval cancers occurred in symptomatic women at 3 and 31 months after a surveillance visit. This emphasizes the importance of counselling women undergoing screening to report any abnormal menstrual symptoms or bleeding patterns, and subsequently investigating them. TVS alone would have missed six of the cancers in this cohort. In addition, complex atypical hyperplasia was found in four women, complex hyperplasia without atypia in eight women, and simple hyperplasia in two women undergoing surveillance. On the basis of available evidence, an annual TVS and hysteroscopy with endometrial biopsy from the age of 30–35 is recommended in women with Lynch syndrome opting for endometrial screening. This is consistent with most current guidelines.

Low-risk population There are well-established guidelines for screening low-risk, symptomatic postmenopausal women for endometrial cancer. However, screening of low-risk asymptomatic women for endometrial cancer is not routinely recommended. Women with postmenopausal bleeding are initially investigated using TVS as a ﬁrst-line test. This is followed by endometrial sampling with or without hysteroscopy in those with increased endometrial thickness (ET). It is conventional to measure double thickness (ET of both layers) at the thickest point in the mid-sagittal view. Any abnormality, such as a polyp, should be investigated irrespective of ET.

23

Chapter 2: Screening and prevention

The Postmenopausal Estrogen/Progestin Interventions (PEPI) trial found that an ET threshold of 5 mm yielded a PPV of 9%, NPV of 99%, sensitivity of 90% and speciﬁcity of 48% for detecting endometrial hyperplasia or cancer. A subsequent meta-analysis suggested an ET of 5 mm as a cut-oﬀ for investigating postmenopausal bleeding. A negative test would reduce the likelihood of endometrial cancer to 2.5%. The ET in women taking sequential HRT is greater than in those using a continuous combined/bleed-free preparation, and, hence, diﬀerent corresponding cut-oﬀs have been proposed for investigating unscheduled bleeding in these groups. The Scottish Intercollegiate Guidelines Network recommends using 3 mm as a cut-oﬀ for women on continuous combined HRT or not on HRT for a year and for those who have never taken HRT, and a cut-oﬀ of 5 mm in women on a sequential preparation. Women taking tamoxifen are more prone to develop endometrial polyps and have a four- to ﬁvefold higher risk of endometrial cancer. The speciﬁcity of TVS as a screening tool is considerably reduced in women on tamoxifen therapy and the ideal cut-oﬀ for women on tamoxifen is not known. This may be partly due to the tamoxifen-induced thickness artefact seen on the scan. Hence, hysteroscopy and biopsy are considered to be the best method of investigating symptomatic women on tamoxifen, in view of their inherent higher-risk status and limitations of TVS.

Other modalities Endometrial hyperplasia or carcinoma has been diagnosed in 13.5% of postmenopausal women with normal endometrial cells on routine cytology, 23% of those with atypical cells, and 77% of those with suspicious cells. Endometrial cytological sampling has also been investigated for surveillance but is not recommended as it has a low PPV and lower diagnostic accuracy than pipelle biopsy. DNA for molecular analysis of endometrial tissue can be obtained from tampons/sanitary towels, cervical cytology and pipelle. Its utility as a potential screening tool is still a matter of research. Other techniques that are under investigation and not part of routine protocols include three-dimensional ultrasonography for the measurement of endometrial volume and power Doppler analysis, and saline infusion sonohysterography. The ability of three-dimensional sonography to distinguish between hyperplasia and cancer is still limited. Saline infusion sonohysterography may be better than standard TVS at evaluating intrauterine pathology such as polyps or ﬁbroids but has limited value in diagnosing hyperplasia or carcinoma.

Prevention Unlike BRCA carriers, the appropriate preventive surgery for women with Lynch syndrome includes a hysterectomy and bilateral salpingo-oophorectomy (BSO). In a study of 315 MMR gene carriers, 61 underwent prophylactic hysterectomy and 47 of them underwent RRSO along with hysterectomy. No ovarian or endometrial cancer occurred in the preventive surgery group at 10 years’ follow-up, while in the non-surgical group 33% developed endometrial and 5.5% developed ovarian cancer. Given the lack of data on the eﬃcacy of screening and deﬁnite beneﬁt from prophylactic surgery, risk-reducing hysterectomy and BSO is the recommended option for women with Lynch syndrome who have completed their families. Occult endometrial cancer has been reported in up to 5% of women undergoing the procedure. Hence, a thorough preoperative assessment including a TVS, endometrial sampling and CA125 levels is recommended in all those opting for prophylactic surgery.

24

Chapter 2: Screening and prevention

Using the contraceptive pill can halve the risk of endometrial cancer. Progestogens have been found to be eﬀective in the treatment of atypical endometrial hyperplasia and endometrial cancer. A phase II RCT coordinated by the MD Anderson Cancer Center in Texas, USA, is comparing the risk-reducing beneﬁt of injectable medroxyprogesterone acetate and the contraceptive pill (ethinyl estradiol and norgestrel) in women with HNPCC-associated cancers or HNPCC-related gene mutations. The ongoing UK multicentre RCT, POET (Prevention of Endometrial Tumours), is evaluating the eﬃcacy of Mirena-IUS as a chemopreventive agent in women with Lynch syndrome. Women aged 35–65 who have Lynch syndrome are being randomized to Mirena-IUS insertion or no treatment and followed up for 4 years.

Cervical cancer Cervical cancer is the major cause of death from gynaecological cancer worldwide, with most deaths occurring in the developing world. In the UK, there are about 2700 cases and 1000 deaths from cervical cancer every year. Over 90% of the cases of carcinoma in situ (CIN) occur in women under 45, with the peak incidence being in the 25–29-year age group. In contrast the occurrence of invasive cervical cancer is fairly evenly spread across the age groups over 25, with around 42–46% occurring in the reproductive age. Cervical cancer is a preventable disease and is completely curable if detected at an early stage.

Screening for cervical cancer A screening programme based on repetitive exfoliative cytology and colposcopy has led to a marked decrease in the incidence of cervical cancer in the UK and the rest of the developed world. The eﬀect is greatest in those countries with organized screening programmes. The rate of protection against cervical cancer oﬀered by cervical screening ranges from approximately 60% to 85%. The rate is slightly lower in women under 40 than over 40. The coverage of the National Health Service (NHS) Cervical Screening Programme (NHSCSP) in England is around 79.5% (2006 statistics). Of the 4.4 million women invited, 3.6 million underwent screening. It has been suggested that an overall coverage of 80% can potentially decrease the associated mortality rates by 95%. Although cytological screening may be less eﬀective against cervical adenocarcinoma (15% of cervical cancers), it does have a substantial impact even in this subgroup. Sensitivity and speciﬁcity of cervical cytology has been reported to range between 30% and 87% and 86% and 100%, respectively.

Pap smear and liquid-based cytology Following recommendations from the National Institute for Health and Clinical Excellence (NICE), the traditional method of obtaining cytological specimens – the Papanicolaou smear – has been replaced by the liquid-based cytology test (LBC) in the UK. There are several commercially available LBC systems, such as SurePath, ThinPrep, Cytoscreen and Labonard Easy Prep. LBC leads to the preparation of more homogeneous, easier-to-read slides and a more eﬃcient automated laboratory sample-handling process, resulting in increased productivity. However, there are no formal criteria assessing the adequacy of the LBC procedure. Data suggest that compared with pap smears, LBC has increased sensitivity for abnormal cytology and there are fewer inadequate/unsatisfactory cytology reports. An English pilot study reported an increase in sensitivity of 2.8% to 12% and a 87% reduction in inadequate cytology reports (from 9% to 1.6%) with LBC, while maintaining speciﬁcity.

25

Chapter 2: Screening and prevention

It also found a reduction in the glandular neoplasms detected (from 0.08% to 0.04%), though follow-up data showed no change in the number of adenocarcinomas. A number of systematic reviews have advocated a preference for LBC, whereas others have found no diﬀerence between LBC and conventional pap smears. A recent meta-analysis of studies using colposcopy-directed biopsy as the gold standard found no signiﬁcant diﬀerences between the sensitivity and speciﬁcity of LBC and pap smears.

Human papillomavirus DNA testing HPV infection is sexually transmitted, with an established causal link central to the development of CIN and cervical cancer. HPV DNA has been found in 99.7% of cervical cancers, which can develop 5–30 years after the primary infection. Over 100 HPV types have been identiﬁed, of which around 20 are high risk. There is considerable worldwide variation in the distribution of high-risk viral types, but around 70% cancers have been linked to HPV types 16 and 18. Low-risk HPV types 6 and 11 are responsible for genital warts, low-grade cervical lesions and respiratory papillomatosis. NHSCSP data from Manchester (ARTISTIC trial, 2008), suggest that types 16, 18, 31, 51 and 52 account for 60% of all high-risk HPV types and that multiple infections were common below the age of 30, but less so between 30 and 64. Systematic reviews and meta-analysis show that the performance of HPV DNA testing may exceed that of cytological follow-up. High-risk HPV DNA testing has been shown to have higher sensitivity but slightly lower speciﬁcity (of the order of 8–12%) than cytology alone for detecting high-grade disease. A combination of HPV testing and cytology is associated with even higher sensitivity, and may save additional years of life at reasonable costs compared with cytology testing alone. Women who test negative for high-risk HPV DNA and have normal cytology are extremely unlikely to develop CIN/cancer in the next 5–10 years. High-risk HPV DNA-negative women with a borderline/ASCUS (Atypical Squamous Cells of Undetermined Signiﬁcance) cytology have a <2% risk of high-grade CIN/cancer in the next two years, which is similar to women with negative cytology. A positive high-risk HPV DNA test in the presence of severely dyskaryotic cytology and highgrade disease at colposcopy suggests a 60–80% risk of CIN3 or worse in the next two years. HPV DNA testing has been recommended in triaging ASCUS or borderline cytology, for predicting the risk of post-treatment recurrence, and for identiﬁcation of women with abnormal cytology and low-grade histology needing post-colposcopic follow-up. It helps identify the group of women who are more likely to develop high-grade CIN/cancer. Thus, compared with cytology alone, high-risk HPV testing can improve risk stratiﬁcation and increase the eﬃciency of cervical cancer screening. However, it may not be that eﬀective in adolescent girls, in whom there is a high prevalence of HPV but low incidence of precancerous lesions. The recently published 2006 consensus guidelines adopted by the American Society for Colposcopy and Cervical Pathology (ASCCP) recommends high-risk HPV DNA testing as part of routine screening protocols in women over 30.

Other screening methods Various other strategies, such as naked eye visual inspection of the cervix after application of acetic acid (VIA), visual inspection after Lugol’s iodine (VILI), visual inspection with a magnifying glass (VIAM) and HPV DNA testing are emerging as eﬀective screening options, especially in developing countries with limited resources and a lack of established eﬃcient screening programmes. A recent meta-analysis of ﬁve screening strategies, including

26

Chapter 2: Screening and prevention

11 studies involving 58 000 women, showed that VIA, VILI and VIAM had high sensitivity (79%) and speciﬁcity (85%) for high-grade disease. Pap smears had lower sensitivity (57%) but higher speciﬁcity (93%). A large RCT has shown that a VIA-based screening strategy is eﬀective and can lead to decreased mortality in the screened arm (hazard ratio 0.65). However, sensitivity reported with strategies such as VIA may be slightly inﬂated due to a gold standard (colposcopic directed biopsy) misclassiﬁcation error. Other novel techniques being evaluated include automated slide-reading techniques such as DNA image cytometry and hyperspectral imaging, as well as the use of molecular markers such as telomerase and hTRT to detect high-grade/progressive disease.

Screening protocols /guidelines Screening protocols and practices vary between institutions and countries. Some of the salient features mentioned in Tables 2.4 and 2.5 are derived predominantly from the British Society of Colposcopy and Cervical Pathology (BSCCP) guidelines and reﬂect mainly Table 2.4 Recommendations for screening frequency and colposcopy referral (UK)

Screening frequency Recommended screening frequency

25 years – ﬁrst invitation; 25–49 years: 3 yearly; 50–64 years: 5 yearly; >65 years: those not screened since 50 and those with recent abnormal tests

Individuals <25 years

Screening not routinely recommended (low incidence of cervical cancer, high HPV prevalence, mainly self-resolving low-grade disease seen. Screening this age group can lead to increased anxiety and unnecessary overtreatment)

Unscheduled screening (with a normal test in preceding 3–5 years)

Not recommended in cases of: multiple sexual partners, heavy smokers, vaginal discharge, sexually transmitted disease/infection/genital warts, commencing contraceptive pill, intrauterine contraceptive device, pregnancy, hormone replacement therapy

Colposcopic screening

Should be considered in immunosuppressed women, transplant recipients, human immunodeﬁciency virus-positive women (annual screening), recurrent inadequate cytology sample

Colposcopy referral rules Inadequate cytology

Three consecutive

Borderline cytology

Three consecutive borderline squamous; or one borderline endocervical cytology sample

Mild dyskaryosis

Ideally one cytology sample but two consecutive cytologies also acceptable

Moderate dyskaryosis

One cytology sample (74% may have CIN2/3)

Severe dyskaryosis

One cytology sample (80–90% may have CIN2/3)

Invasive cytology

One cytology sample (up to 56% PPV reported for cancer)

Glandular neoplasia

One cytology sample (up to 43% may have cancer and 28% CGIN)

Any grade abnormality

Three tests with any grade abnormality in the past 10 years

Symptomatic women; women with abnormal cervix

Referral for gynaecological examination and then for colposcopy if necessary

Post-CIN treatment

Follow-up cytology mild dyskaryosis or worse and not yet returned to routine recall

CIN, cervical intraepithelial neoplasia; CGIN, cervical glandular intraepithelial neoplasia.

27

Chapter 2: Screening and prevention

Table 2.5 Recommendations for colposcopy practice (UK)

Colposcopy ‘See and treat’ policy

Usually followed for high-grade lesions. Not routinely used for mild dyskaryotic/borderline cytology at ﬁrst visit

Punch biopsy

With high-grade cytology (moderate/severe dyskaryosis) if excisional treatment not performed; with low-grade colposcopic abnormality and conservative management; may not be necessary with low-grade cytology and negative colposcopy; should not be performed in pregnancy

Destructive/ablative treatment

Biopsy(ies) essential prior to local destructive treatment. Only suitable when entire transformation zone is visualized; no glandular abnormality present; no invasive disease; and when low-grade CIN is present. A double freeze–thaw–freeze technique is advised for cryocautery

Excisional biopsy/treatment

High-grade abnormality at colposcopy; low-grade abnormality associated with high-grade cytology; lesion extending into canal (punch biopsy may be inadequate in these situations). Treatment depth should be 7 mm or more

Repeat excision

Should be considered in women over 50 with high-grade lesions and incomplete excision, especially if satisfactory cytology/colposcopy cannot be guaranteed. Is not essential in women <50 if there is no glandular abnormality or invasive disease

CGIN/AIS

Local excision for those wishing to conserve fertility. Clear endocervical margin is essential and if necessary repeat procedure should be performed to obtain this. Additional endometrial biopsy should be considered if symptoms or in older women. Punch biopsy may be inadequate for abnormal high-grade glandular cytology.

High-grade cytology and no treatment done

Punch biopsy should be performed; treatment essential if highgrade cytology persists at follow-up (6 months colposcopy and cytology)

Low-grade cytology and untreated low-grade lesion at colposcopy

6-monthly follow-up in colposcopy clinic. If not resolved in 2 years, treatment performed or at least biopsies warranted

Post-treatment follow-up First follow-up

Ideally at treatment centre, 6 months after treatment. Cytology alone is adequate, though in practice this may often occur in combination with colposcopy. If the cytology is not negative, repeat colposcopy is advised within the year

Subsequent cytology

At 12 months’ post-treatment and then yearly

CIN1 (low-grade disease)

Returned to routine screening at 2 years if cytology negative

CIN2/3 (high-grade disease)

Returned to routine screening at 10 years if cytology negative

Special considerations Post-hysterectomy Post-trachelectomy

28

No follow-up if on routine recall for 10 years before surgery and no CIN at histology; follow-up: 6-month vault cytology if <10-year routine recall with no CIN; follow-up: 6- and 18-month vault cytology if completely excised CIN at hysterectomy; if uncertain or incomplete CIN excision: routine follow-up as per low/high-grade CIN guidelines.

Chapter 2: Screening and prevention

Table 2.5 (cont.)

Colposcopy Pregnancy

Routine cytology can be postponed until after delivery but if a previous cytology was abnormal it should be repeated in pregnancy. Colposcopy in pregnancy should be performed for those women fulﬁlling referral criteria, preferably by clinicians with particular experience due to the diagnostic diﬃculties it presents. The main aim of colposcopy in pregnancy is to rule out invasive disease, to defer treatment until after delivery. An appropriate diagnostic biopsy is performed if invasion is suspected. Colposcopy is repeated 3 months after delivery if CIN1 is suspected and in the second trimester if CIN2/3 is suspected.

Immunocompromised states

Individuals with multifocal disease should have 6-monthly cytology, colposcopy, vulvoscopy and biopsy if needed; human immunodeﬁciency virus (HIV)-positive individuals should have annual cytology

AIS, adenocarcinoma in situ; CIN, cervical intraepithelial neoplasia; CGIN, cervical glandular intraepithelial neoplasia.

UK practice. Readers are advised to refer to guidelines published by the BSCCP and the ASCCP for detailed information.

Prevention of cervical cancer In addition to HPV, other risk factors associated with cervical cancer include use of the contraceptive pill, smoking, multiple sexual partners, early coitarche and low socioeconomic status. Use of the contraceptive pill for ten years from 20 to 30 years of age can increase the risk of cervical cancer by age 50 from 7.3/1000 to 8.3/1000 in less developed countries and from 3.8/1000 to 4.5/1000 in more developed countries. However, the causal role of contraception and smoking is not established because of confounding with sexual behaviour. Delayed ﬁrst sexual intercourse and use of condoms can lead to reduced incidence of HPV infection, and hence, a reduction in its potential consequences.

HPV vaccines Although cervical cancer is a preventable disease and is completely curable if detected at an early stage, cytological screening alone will not lead to the eradication of cancer. Two prophylactic HPV vaccines have been found to be safe, immunogenic and eﬃcacious at providing protection from type-speciﬁc HPV infection. The ﬁrst one (CervarixTM) is a bivalent HPV L1 virus-like-particle vaccine of HPV types 16/18, used at 20/20 μg/dose. It is administered at 0, 1 and 6 months as a 0.5-ml intramuscular injection. The second one (GardasilTM) is a tetravalent HPV L1 virus-like-particle vaccine of HPV types 16/18/6/11, used at 20/40/40/20 μg/dose. It is administered at 0, 2 and 6 months as a 0.5-ml intramuscular injection. The tetravalent vaccine has the added advantage of protecting against low-risk HPV types 6/11, which cause genital warts. The eﬃcacy of these vaccines in preventing persistent HPV infection has been found to range between 90% and 100%, and the immunity provided has been shown to last for in excess of six years. Mathematical modelling suggests immunity persists for 20–30 years. A recent meta-analysis found that following vaccination, the relative risk of HPV 16 and HPV 18 infections was 0.13 and 0.22, respectively. Although there is lack

29

Chapter 2: Screening and prevention

of long-term eﬃcacy data for vaccination in the prevention of cervical cancer, Markov modelling suggests that vaccination will be cost eﬀective. However, there are still unresolved issues surrounding the integration of HPV vaccination into current screening programmes. Modelling of data from the ARTISTIC trial suggests vaccination leads to a 45% reduction in high-grade cytologies and 7% reduction in low-grade cytologies. Polyvalent vaccines developed in the future are likely to increase eﬃcacy. HPV vaccination for girls aged 12–13 was introduced into the UK national immunization programme in September 2008, using CervarixTM. Subsequently, a two-year catch-up drive for vaccinating girls (up to 18 years) is planned to commence from autumn 2009 for girls aged between 16 and 18, and from autumn 2010 for girls between 15 and 17. The NHSCSP does not intend to oﬀer vaccination to women over the age of 18 at present.

Conclusions Well-established eﬀective cytology/colposcopy-based screening programmes for cervical cancer exist in the developed world. Considerable changes have occurred over the past few years, such as the introduction of LBC and HPV DNA testing. Newer methods such as VIA/VILI/VIAM/HPV DNA testing hold potential for the developing world. Availability of HPV vaccines is leading to a phased introduction in the developed world. HPV vaccination holds real promise for the future, with the potential to markedly impact on the global incidence of and mortality associated with cervical cancer. However, longterm data are currently lacking and a number of challenges remain, including the development of polyvalent vaccines and delivering an aﬀordable programme in the developing world. The eﬀectiveness of ovarian cancer screening is being addressed by ongoing research trials, both in high- and low-risk populations. Screening outside the context of research trials is not routinely recommended. RRSO is an eﬀective preventive option for ovarian/tubal cancer in women at high risk for the disease. Although there is a good evidence base to determine protocols for screening low-risk symptomatic women for endometrial cancer, the ideal method of screening high-risk asymptomatic women for endometrial cancer is not yet known. A screening approach using annual ultrasound scan and hysteroscopy-based endometrial sampling is the current accepted standard. Prophylactic hysterectomy along with RRSO is the optimal preventive surgery for women with Lynch syndrome.

Further reading

30

*

Cancer Research UK. Latest UK summary – cancer incidence 2004 and mortality 2005. Cancer Research UK, 2007. Available at: http://info.cancerresearchuk.org/images/ pdfs/2004inc2005mortpdf.

*

National Institute for Clinical Excellence (NICE). Guidance on the Use of Liquid-based Cytology for Cervical Screening. London: National Institute for Clinical Excellence, 2003.

*

Jacobs I, Oram D, Fairbanks J, Turner J, Frost C, Grudzinskas J G. A risk of malignancy index incorporating CA125, ultrasound and menopausal status for the accurate preoperative diagnosis of ovarian cancer. Br J Obstet Gynaecol 1990; 97: 922–9.

*

NHS Cancer Screening Programmes. Colposcopy and Programme Management – Guidelines for the NHS Cervical Screening Programme. London: NHS Cancer Screening Programmes, 2004.

*

Sasieni P, Adams J, Cuzick J. Beneﬁt of cervical screening at diﬀerent ages: evidence

Chapter 2: Screening and prevention

*

from the UK audit of screening histories. Br J Cancer 2003; 89: 88–93. Wilson J, Junger G. Principles and Practice of Screening for Disease. Public Health Papers no. 34. Geneva: World Health Organization, 1968: 66–7.

*

Wright T C, Jr, Massad L S, Dunton C J, Spitzer M, Wilkinson E J, Solomon D. 2006 consensus guidelines for the management of women with cervical intraepithelial neoplasia or adenocarcinoma in situ. Am J Obstet Gynecol 2007; 197: 340–5.

31

3

Chapter

Pathology of gynaecological cancers Mark J. Arends

Introduction There is a wide variety of tumours that may occur in the diﬀerent parts of the female genital tract. The ovarian tumours constitute a wide array of diﬀerent tumour types that may be broadly categorized into three groups: epithelial carcinomas, sex cord/stromal tumours and germ cell tumours. In the uterus, endometrial tumours are mostly of the endometrioid adenocarcinoma type, but there are also other types of adenocarcinoma, including the biphasic (mixed epithelial and mesenchymal) malignant mixed Müllerian tumour of the endometrium; pure sarcomas also occur, including stromal sarcomas arising in the endometrium and smooth muscle sarcomas (leiomyosarcomas) in the myometrium. Although squamous cell carcinomas tend to dominate in the vulva and cervix, glandular adenocarcinomas and other tumours can also occur. As well as tumour type, the tumour grade and stage often have an important inﬂuence on patient management, inﬂuencing decision making about chemotherapy or radiotherapy following biopsy or surgical resection of the tumour. Hence, there is an important requirement for good working relationships between pathologists and clinical colleagues to make best use of pathology services. Over the past decade, such relationships have been built up within the context of the multidisciplinary team meeting, which allows integration of clinical, radiological, pathological and other sources of relevant information, so that decisions may be taken in the best interests of the patient on the basis of detailed knowledge of the behaviour of the cancer in each case.

Ovarian tumours The three main categories of ovarian tumours are: epithelial, sex cord-stromal and germ cell tumours. Only the most important of these ovarian tumour types can be covered in a chapter of this size, and serous tumours of the ovary will be used as the prototype epithelial tumour group.

Epithelial serous tumours *

There are three types of serous tumour: benign, borderline and malignant.

*

The prognosis of borderline serous tumours depends on the nature of any extraovarian implants.

Serous tumours fall into three basic types: benign, borderline (or low malignant potential or atypical proliferating) and malignant. Benign serous tumours are cystadenomas or Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 3: Pathology of gynaecological cancers

Fig. 3.1 Serous papillary carcinoma with several psammoma bodies, which appear as dark purple, heavily calciﬁed, laminated, small spherical structures, often associated with surrounding clusters of malignant serous carcinoma cells (H&E, magniﬁcation × 200).

adenoﬁbromas, with the epithelial elements resembling a single layer of tubal epithelium with little or no nuclear pleomorphism. Small ﬁbrous papillae are common on the inner lining and are composed predominantly of stroma. They are bilateral in between 7% and 20% of cases. Surface serous papillary adenoﬁbromas show multiple warty serosal excrescences. The stromal component of the adenoﬁbromatous variant usually consists of abundant dense ﬁbrous tissue. Borderline serous tumours typically have multiple polypoid excrescences with ﬁne papillae occupying part or all of the lining of one or more cysts and the outer surface of the ovary may also be involved. Borderline tumours are bilateral in 25–40% of cases and around 70% are conﬁned to one or both ovaries. Histologically, the lining epithelium shows mild to moderate nuclear pleomorphism, a relatively low number of mitoses, stratiﬁcation and ﬁne papillation of the epithelium, which is typical, with formation of cellular tufts or buds that appear to ﬂoat into the intracystic ﬂuid, as their point of attachment is not in the plane of section. However, by deﬁnition there is no destructive invasion of the stroma. Psammoma bodies may be present and can be numerous (Fig. 3.1). When marked micropapillation with very extensive multilayered epithelium is present, the term micropapillary serous carcinoma can be used. It is considered to indicate a higher risk of extraovarian invasive implants, but this is controversial. Invasion of the stroma by one or more discrete foci of serous tumour cells within a borderline tumour, where the invasive focus occupies an area <10 mm2, constitutes microinvasion and some studies have suggested that this has an adverse eﬀect on prognosis, which is otherwise excellent overall for serous borderline tumours without either microinvasion or frank destructive invasion (due to serous carcinoma). Some borderline serous tumours are associated with extraovarian deposits known as implants and these may vary from foci of benign-appearing serous epithelium with psammoma bodies (endosalpingiosis), through non-invasive implants with typical features of borderline serous tumour, to invasive implants indistinguishable from serous carcinoma. Non-invasive implants, even with a stromal reaction, seem to be associated with a good outcome, and invasive implants with disorderly inﬁltration of the stroma seem to be associated with a poor prognosis. The so-called implants probably arise in situ in the peritoneum, and this notion has led to the concept of the secondary peritoneal Müllerian system, with its ability to develop primary tumours of the peritoneum, such as primary peritoneal serous papillary carcinoma. Such a diagnosis is merited when an extensive serous tumour involving the peritoneum shows only minor and focal involvement of the ovary, usually on the surface.

34

Chapter 3: Pathology of gynaecological cancers

Serous carcinomas are characterized by complex patterns of papillae with slit-like spaces, extensive cellular budding and more marked nuclear pleomorphism than serous borderline tumours and show destructive invasion of the underlying stroma. These tumours vary from well to poorly diﬀerentiated. Psammoma bodies may be very frequent and a particular variant, so called serous psammocarcinoma, shows extensive formation of psammoma bodies with loss of much of the epithelium. This variant seems to have a good prognosis. Serous carcinomas spread particularly via the peritoneal cavity, and this is the major reason for their very poor overall prognosis, except in stage I (tumour conﬁned to the ovary). Prognosis in serous carcinoma depends most importantly on stage (ﬁve-year survival of around 75% for stage I, but 25% for stage III), and also on tumour grade. Although the criteria for grading are not well deﬁned, mostly the tumours can be divided into low-grade and high-grade carcinomas. BRCA germline mutation carriers may develop serous tumours, whereas women with Lynch/hereditary non-polyposis colorectal cancer (HNPCC) syndrome can develop a wider range of epithelial ovarian tumours.

Mucinous tumours *

Distinguishing borderline from malignant mucinous tumours can be diﬃcult.

*

Prognosis of borderline mucinous tumours depends on extraovarian implants.

Mucinous tumours are characterized by glands and cysts lined by epithelial cells containing abundant intracytoplasmic mucin. The tumour cells may resemble endocervical or intestinal epithelial cells, but typically the benign tumours resemble endocervix and the borderline and malignant resemble intestinal epithelium. Benign mucinous tumours tend to be the largest of all ovarian tumours and are a common constituent of teratomas. About 5% of mucinous tumours develop in association with dermoid cysts. They may also be associated with mucinous tumours of other organs, particularly the appendix. Benign mucinous tumours show regular epithelium lying on a ﬁbrous stroma or stroma resembling ovarian stromal tissue. In contrast with benign tumours, borderline tumours are characterized by nuclear pleomorphism, a complex glandular pattern and epithelial overgrowth (Fig. 3.2), whereas mucinous carcinomas show destructive invasion of the stroma often with marked architectural abnormalities and malignant nuclear features. The endocervical-like mucinous borderline tumour is a variant of borderline mucinous tumour with endocervical-type epithelium, papillation of the eosinophilic epithelium and a marked inﬂammatory cell inﬁltrate in the stroma. These tumours appear to behave in a manner similar to serous borderline tumours Fig. 3.2 Borderline mucinous tumour of the ovary showing the papillary architecture of the neoplastic mucinous epithelium, including intestinal goblet cells, without destructive stromal invasion (H&E, magniﬁcation × 200).

35

Chapter 3: Pathology of gynaecological cancers

and may be associated with endometriosis. Borderline mucinous tumours without pseudomyxoma have an excellent prognosis.

Pseudomyxoma peritonei *

*

Most are closely associated with appendiceal mucinous tumours rather than ovarian primary mucinous tumours. There are four subtypes: mucinous ascites, organizing mucinous ﬂuid, disseminated peritoneal adenomucinosis and mucinous adenocarcinomatosis.

Benign, borderline and malignant mucinous intestinal-type ovarian tumours can be associated with pseudomyxoma peritonei (PMP), which is the presence of jelly-like mucinous material in the pelvis or abdomen. PMP is of four types: mucinous ascites (mucinous material only), organizing mucinous ﬂuid (mucin with granulation tissue and inﬂammation in surrounding peritoneum), disseminated peritoneal adenomucinosis (abundant mucin in pools with scanty neoplastic epithelium showing low-grade atypia) and mucinous adenocarcinomatosis (widespread malignant mucinous epithelium showing high-grade atypia). PMP occurs typically with intestinal-type borderline or malignant mucinous tumours. There has been considerable debate regarding whether it is associated with primary ovarian mucinous tumours or with primary appendiceal mucinous tumours with ovarian metastases (often showing a borderline mucinous tumour pattern). Available evidence suggests that most PMP are associated with primary appendiceal mucinous tumours, but not all. Unfortunately, PMP is not responsive to chemo- or radiotherapy and is usually progressive.

Endometrioid tumours * *

Endometrioid tumours are closely associated with endometriosis. Borderline endometrioid tumours are rare and have a good prognosis.

Endometrioid tumours can also be classiﬁed into benign, borderline (or proliferating) and malignant types. Occasional benign tumours are endometrioid cystadenomas but these are rare, and some argue that such tumours are examples of endometriosis with loss of stroma. The other variant is the endometrioid adenoﬁbroma with glands lying in a dense ﬁbromatous stroma. Borderline endometrioid tumours usually have an adenoﬁbromatous pattern with overgrowth of atypical endometrioid epithelium, but the prognosis is good. Ovarian endometrioid carcinomas resemble endometrioid carcinomas of the uterine corpus. Endometrioid ovarian tumours are often associated with endometriosis, and some foci of endometriosis may be associated with endometrioid hyperplasia – encompassing borderline tumour through to endometrioid carcinoma.

Clear cell tumours *

Clear cell tumours show a close association with endometriosis.

*

Clear cell carcinomas have a worse prognosis than serous tumours.

Clear cell tumours are also commonly associated with endometriosis, perhaps as often or more frequently than endometrioid tumours. Benign variants (mostly adenoﬁbromas) are rare and borderline clear cell tumours account for less than 1% of ovarian borderline tumours. Clear cell carcinomas make up around 5% of epithelial ovarian carcinomas. Most

36

Chapter 3: Pathology of gynaecological cancers

occur in women in the ﬁfth decade and above. The borderline tumours are adenoﬁbromas with pleomorphism in the epithelial element and with epithelial overgrowth compared with the simple benign clear cell adenoﬁbroma. Both tumours appear to have an excellent prognosis. Clear cell carcinomas are composed of tumour cells with eosinophilic or clear cytoplasm lining cysts and acini, often with a tubulopapillary pattern and possibly marked nuclear pleomorphism. They have been suggested to have a poorer prognosis than other ovarian carcinomas, including high-grade serous carcinomas. Grading is according to nuclear pleomorphism, but is not of proven value.

Other ovarian carcinomas *

Transitional cell carcinoma may be more chemosensitive than other carcinomas.

*

Small cell carcinoma has a very poor prognosis.

Transitional and squamous cell carcinomas of the ovary are rare but well described, and Brenner tumours also have benign, borderline and malignant types. There is evidence that a tumour with a major proportion of transitional cell carcinoma may be more chemosensitive than other variants. Even metastases with appreciable amounts of such diﬀerentiation may respond well to therapy. Surprisingly, this does not apply to malignant Brenner tumours even though the only diﬀerence between a transitional cell carcinoma and a malignant Brenner tumour is the presence of coexistent borderline or benign Brenner tumour in the latter. Small cell carcinoma of hypercalcaemic type is an undiﬀerentiated carcinoma made up of small dark cells. It tends to occur in young women between the ages of 1 and 46 years (mean 24 years) and is associated with hypercalcaemia in 60% of cases. Most small cell carcinomas are solid, large tumours and 50% are high stage by the time of presentation. They have a very poor prognosis and only a third of aﬀected women, even in stage Ia, are disease-free on follow-up; most patients with advanced-stage disease die within two years of presentation. There is also an ovarian small cell carcinoma of pulmonary type, which tends to occur in older women and has a very poor prognosis.

Sex cord-stromal, germ cell and other ovarian tumours *

*

*

Granulosa cell tumours are the most frequent malignant sex cord-stromal tumours and are often oestrogenic. Most Sertoli–Leydig cell tumours have a favourable prognosis, although it depends on the degree of diﬀerentiation. Germ cell tumours are chemosensitive.

It is not possible to cover this broad group of tumours in detail here and only key features are described below. In the sex cord-stromal tumour group, granulosa cell tumours are the most frequent malignant tumours. These are divided into two types, adult and juvenile. Juvenile granulosa cell tumours occur predominantly in the ﬁrst three decades, and mostly behave as benign tumours. Adult granulosa cell tumours (AGCTs) commonly secrete oestrogen and present with abnormal endometrial bleeding, often resulting from endometrial hyperplasia or carcinoma; 10% present with abdominal symptoms due to rupture and haemoperitoneum. AGCTs display a variety of histological patterns, and all patterns have malignant potential with spread mainly within the pelvis and lower abdomen. Recurrences may occur

37

Chapter 3: Pathology of gynaecological cancers

(a)

(b)

Fig. 3.3 Ovarian adult granulosa cell tumour displaying: (a) a typical microfollicular pattern with several Call–Exner bodies containing eosinophilic material, with some tumour nuclei showing ‘coﬀee bean’ nuclear grooves (H&E, magniﬁcation × 400); and (b) an immunostain (magniﬁcation × 400) for inhibin, a marker of sex cord tumours.

within 5 years, but are commonly detected even up to three or more decades postoperatively. AGCTs secrete inhibin, which can be a useful immunomarker (Fig. 3.3), and the serum inhibin level may be helpful in detecting recurrent disease. Outcome depends on tumour stage, with ∼90% survival for stage I, reducing to ∼60% for higher-stage tumours at ten years. Rupture is an adverse feature, as is large size. Grade, mitotic activity and histological pattern have prognostic importance in stage I AGCT. Of the stromal ovarian tumours, thecomas are almost always benign, though their oestrogenic eﬀect may lead to an association with endometrioid carcinoma of the endometrium. The great majority of ﬁbromatous tumours are also benign, though some cellular ﬁbromas may recur locally and ﬁbrosarcomas do rarely occur. Sertoli–Leydig cell tumours (SLCTs) show androgenic hormonal manifestations (e.g. virilization) in ∼50% of cases. Prognosis depends on stage, grade and tumour rupture. Welldiﬀerentiated tumours rarely recur, and intermediate diﬀerentiation tumours metastasize in ∼10% and poorly diﬀerentiated tumours in ∼60% of cases. Retiform tumours are more aggressive. The rare SLCTs that present at stage II or higher are almost always fatal. Recurrences of SLCT typically occur earlier (within one year) than those of granulosa cell tumours. Steroid (previously called lipid) cell tumours are uncommon and divided into three groups: stromal luteoma, Leydig cell tumour and steroid cell tumour – not otherwise speciﬁed (NOS). The ﬁrst two types are benign, whereas steroid cell tumours (NOS) are associated with androgenic changes in ∼50% and extraovarian spread in ∼20%, with a third being clinically malignant. Poor prognostic features include higher stage, size (>7 cm indicating ∼80% malignant), two or more mitoses per 10 high power ﬁelds (∼90% malignant), necrosis (86% malignant), haemorrhage (77% malignant) and grade 2 or 3 nuclear atypia (64% malignant). Germ cell tumours account for 30% of primary ovarian tumours, but the majority of these (over 95%) are simple mature cystic teratomas (dermoid cysts). Dysgerminomas are the most common (∼50%) malignant germ cell tumours and most (up to 80%) occur in the 15–30-year age group. Prognosis is excellent with modern therapy (100% ﬁve-year survival for stage I and 80–90% for higher stages); stage seems to be the only important prognostic factor. Immature teratomas are usually seen in young adults and children and may be associated with a mildly raised serum α-fetoprotein (AFP) (or occasionally human chorionic

38

Chapter 3: Pathology of gynaecological cancers

gonadotrophin (hCG)). Approximately a third have extraovarian spread at presentation, but 90–100% of aﬀected women experience sustained remission with chemotherapy, even though metastases may persist as fully matured tissue that may continue to grow and require reoperation. Prognosis for these tumours depends on the nature of extraovarian disease. In women who present with exclusively mature extraovarian disease, usually peritoneal gliomatosis, the disease almost always has a benign clinical course without further therapy, whereas women with immature extraovarian implants require aggressive chemotherapy, as these tumours behave in a malignant fashion. Grade is a prognostic feature for adult immature teratomas, but not for those in children (grade depends on the proportion of immature neural tissue). Monodermal teratomas include: struma ovarii (showing thyroid tissue), which is almost always benign; carcinoid tumour, which tends to behave as a low-grade carcinoma, with only rare deaths due to intra-abdominal recurrence; and strumal carcinoid tumour, which is an admixture of carcinoid tumour and struma ovarii, sometimes with an intermediate tumour pattern and this is usually clinically benign. Mucinous or goblet cell carcinoid, although rare, is mostly benign, although sometimes has a carcinomatous component with spread and a fatal outcome. Yolk sac tumours are now associated with a survival rate of >90% in stage I and 75%, 30% and 25% in stages II, III and IV, respectively, with modern chemotherapy. Serum AFP levels are useful in monitoring therapy and recurrence. Adverse prognostic factors are stage II or higher, gross residual tumour after surgery and liver involvement.

Endometrial tumours Endometrial carcinomas Endometrial carcinomas have been divided into two broad groups. The ﬁrst group, which occurs in pre-, peri- and postmenopausal women and is associated with unopposed oestrogenic stimulation, includes endometrioid adenocarcinoma, usually of grade 1 or 2, and is often preceded by endometrial complex atypical hyperplasia. The second major group of endometrial cancers occurs in older postmenopausal women, is usually not preceded by hyperplasia and includes serous, clear cell, grade 3 endometrioid and undiﬀerentiated carcinomas. These tumours are not usually hormone sensitive, but are more aggressive with a poor prognosis in comparison with the ﬁrst group.

Endometrioid carcinoma *

Most endometrial carcinomas are grade 1/2 endometrioid adenocarcinomas.

Approximately 80% of endometrial carcinomas are of endometrioid type. They are composed of tubular glands, rounded acini with some cribriform structures and are graded 1–3, mostly according to the proportion of architecturally solid, non-acinar tumour (International Federation of Gynecology and Obstetrics (FIGO) grades: G1 <5%, G2 5–50%, G3 >50% solidity) with some room for upgrading, depending on severity of nuclear pleomorphism (Fig. 3.4). Within endometrioid adenocarcinomas, small areas of solid squamous diﬀerentiation are common and do not contribute to grading. Villoglandular endometrioid carcinoma is a variant pattern showing papilla formation with thin ﬁbrovascular cores. These appear to have the same outcome stage for stage as endometrioid adenocarcinomas. Ciliated and secretory variants have also been described. Endometrial cancers may develop in young women with Lynch syndrome (HNPCC) because of DNA mismatch repair deﬁciency and these are usually grade 3 endometrioid

39

Chapter 3: Pathology of gynaecological cancers

(a)

(b)

Fig. 3.4 (a) Endometrioid adenocarcinoma with FIGO grade 1 pattern showing gland formation with some cribriform glandular structures, invading into the smooth muscle wall of the myometrium (H&E, magniﬁcation × 200). (b) Endometrioid adenocarcinoma showing mostly solid areas typical of FIGO grade 3 (left and centre) and a few complex cribriform glandular structures indicative of FIGO grade 2 (right) (H&E, magniﬁcation × 200).

adenocarcinomas with prominent lymphoid cell inﬁltration and peritumoural lymphoid aggregates. Most show abnormal immunohistochemical expression of one of the mismatch repair proteins MSH2, MLH1 or MSH6.

Prognostic factors in endometrial carcinoma *

Stage and grade correlate with prognosis in endometrioid adenocarcinoma.

Apart from tumour type, tumour grade and stage are the most important prognostic factors. Histological grade shows a strong correlation with survival, with approximately 80% 10 years’ survival for grade 1, 60% for grade 2 and 45% for grade 3 endometrioid adenocarcinomas at all stages. The depth of myometrial invasion is also of considerable importance, particularly in endometrioid tumours. Tumours invading the outer half of the myometrium are associated with a death rate of approximately 30%, compared with a death rate of only around 10% for tumours invading the inner half and 1–4% for those conﬁned to the endometrium. The least tumour distance from the uterine serosal surface also has strong independent prognostic signiﬁcance, with tumours extending to within 5.0 mm of the serosa having a poorer prognosis. Lymphovascular space invasion has been shown in several studies to have adverse prognostic signiﬁcance and cervical involvement is important, as reﬂected in the staging system (see Appendix 3), as is lymph node and adnexal involvement. It can be diﬃcult to distinguish metastatic endometrial adenocarcinomas in the ovaries from synchronous primary ovarian tumours. Although many histological features have been proposed to help with this, most of the metastatic ovarian tumours seem to behave as independent primaries. Positive peritoneal cytology as a prognostic factor is controversial and its major value may be, when associated with extrauterine disease, in predicting the likelihood of intra-abdominal recurrence or distant metastases.

Serous papillary and clear cell carcinomas *

Serous and clear cell carcinomas are grade 3 cancers usually seen in older women.

Serous and clear cell carcinomas usually occur in women in their late sixties or older, many are associated with TP53 mutations and p53 immunoreactivity, and 25–60% have pelvic and

40

Chapter 3: Pathology of gynaecological cancers

para-aortic lymph node involvement. Serous carcinomas account for 5–10% of endometrial cancers and are composed of complex papillary arrangements of tumour cells with slit-like spaces, showing grade 3 nuclear pleomorphism and frequent mitoses. Psammoma bodies are present in around a third of cases. They are often locally extensive with deep myometrial invasion present in 40–70% of cases and lymphovascular space invasion is common. They may also show a tendency to spread transperitoneally in a manner similar to ovarian epithelial carcinomas. Stage is the most important prognostic factor. However, around half of women with stage I tumours may develop recurrences and the overall survival rates are low. Some serous carcinomas occur in women with germline BRCA mutations. Endometrial clear cell adenocarcinomas account for 1–3% of endometrial cancers and resemble ovarian clear cell carcinomas with solid and tubulocystic formations of cancer cells with clear cytoplasm, crisp plasma membranes and high-grade nuclear pleomorphism. They are also aggressive tumours with a similar prognosis to serous carcinomas, stage for stage.

Other variants of endometrial carcinoma Mucinous adenocarcinoma has a similar prognosis to endometrioid carcinoma and is characterized by the presence of abundant intracellular mucin. Pure squamous cell carcinoma of the endometrium is rare and may be associated with predisposing features such as cervical stenosis, pyometra, pelvic radiation or extensive endometrial squamous metaplasia. Around a third to half present as stage III or IV cancer, but there is a 70–80% survival for stage I squamous carcinomas, which is poorer than for endometrioid adenocarcinomas.

Malignant mesenchymal tumours of the uterus There are three major groups of malignant mesenchymal tumours of the uterus: leiomyosarcomas, endometrial stromal sarcomas and carcinosarcomas. Other rare sarcomas include angiosarcomas, neurogenic sarcomas, rhabdomyosarcomas and osteosarcomas.

Leiomyosarcomas *

Leiomyosarcomas are the commonest uterine sarcomas and often have a poor prognosis.

Leiomyosarcomas account for up to 80% of uterine sarcomas (excluding carcinosarcomas) and most occur in women over 40. They are typically large, solitary uterine masses with extrauterine extension in a third of cases. Around 5% arise in the cervix. They are less well circumscribed than leiomyomas and show a ﬂeshy, focally necrotic and haemorrhagic cut surface. Most appear to arise de novo rather than in pre-existing leiomyomas. Histologically, they are composed of spindle-shaped smooth muscle cells positive for smooth muscle actin and desmin on immunohistochemistry. Features that distinguish malignant leiomyosarcomas from benign leiomyomas include mitotic rate, tumour necrosis, invasive margins, vascular invasion and nuclear pleomorphism. Leiomyosarcoma survival rates are about 15–30% in most series. Size <5 cm is a favourable prognostic characteristic. Epithelioid, myxoid and other variants have been described, and some epithelioid and myxoid leiomyosarcomas may be diﬃcult to distinguish from benign leiomyomas. Unless small and well circumscribed, most myxoid tumours are probably best regarded as potentially malignant, as they may have a bland histological appearance but still metastasize. The term ‘smooth muscle tumour of uncertain malignant potential’ (STUMP) can be applied to a tumour whose histological appearance is between that of a simple leiomyoma and a leiomyosarcoma.

41

Chapter 3: Pathology of gynaecological cancers

Endometrial stromal sarcomas *

Low-grade stromal sarcomas display invasion, whereas stromal nodules do not.

*

High-grade stromal sarcomas are rare.

Endometrial stromal sarcomas were traditionally divided into three groups: benign endometrial stromal nodule, low-grade endometrial stromal sarcoma and high-grade endometrial stromal sarcoma. Low-grade stromal sarcomas inﬁltrate uterine wall at the margin and typically show extensive worm-like vessel invasion. Stromal nodules are well circumscribed, without invasion. Both tumours are composed of small fusiform and rounded cells that closely resemble endometrial stromal cells and are positive for CD10. High-grade stromal sarcoma is a pleomorphic highly malignant tumour that exhibits evidence of endometrial stromal diﬀerentiation or has arisen from a low-grade stromal sarcoma. However, many high-grade sarcomas do not have these features and are better regarded as undiﬀerentiated uterine sarcomas as they are not deﬁnitely of stromal origin. Most endometrial stromal sarcomas occur in middle-aged women and diﬀerentiation of low-grade sarcoma from stromal nodule on curettage is often impossible as the margin is not available to assess. Around a third of cases of low-grade stromal sarcoma have extrauterine extension at presentation, often in the form of plugs of tumour within vessels. Variant patterns of diﬀerentiation including sex cord-like patterns and foci of smooth muscle diﬀerentiation may complicate the diagnosis, and true endometrioid glands may be seen, leading to a false diagnosis of endometriosis in extrauterine deposits. The only signiﬁcant predictor of recurrence and survival in low-grade stromal sarcoma is stage. Approximately a third of women with stage I disease will have recurrence, whereas 75% of women with stage III and IV disease will have recurrence. Five-year survival rates exceed 90% for stage I disease and are approximately 60–70% for stages III and IV. Some studies show high survival rates even in high-stage disease; this is most likely due to sensitivity to progestagen therapy, as most tumours express progesterone receptors.

Mixed Müllerian tumours of the uterus including carcinosarcoma/malignant mixed Müllerian tumour * *

Carcinosarcoma is the commonest (and most aggressive) variant. Carcinosarcomas are monoclonal with biphasic diﬀerentiation to carcinomatous and sarcomatous components.

Mixed Müllerian tumours of the uterus have mixed epithelial and mesenchymal neoplastic components and can be divided into four groups: adenoﬁbroma, adenosarcoma, carcinoﬁbroma and carcinosarcoma. Adenoﬁbromas consist of benign epithelium, typically an admixture of endometrioid, endocervical and squamous, lying in a benign ﬁbrous stroma, with a frond or leaf-like pattern to the arrangement of the glands and with cystic glandular dilatation. Adenoﬁbromas are rare and benign, but may recur locally within the uterus and treatment is therefore usually hysterectomy. It is not uncommon for a tumour to have a predominantly adenoﬁbromatous pattern, but with focal areas of stromal malignancy that may not be seen without complete removal. Müllerian adenosarcoma shows benign, though sometimes hyperplastic, epithelial glands, usually endometrioid in type or a combination of endometrioid and endocervical, with a cellular cuﬀ of mesenchymal or stromal tissue that is mitotically active and malignant.

42

Chapter 3: Pathology of gynaecological cancers

Heterologous elements may be seen in the stroma, including cartilage and rhabdomyoblasts. Frond-like and papillary structures are typical. Adenosarcomas behave in a lowgrade malignant fashion, with 25–40% undergoing pelvic or vaginal recurrence, and haematogenous spread is rare. Recurrences may be very late, in excess of ﬁve, or even ten, years after hysterectomy. The recurrence is usually pure sarcoma, but may be adenosarcoma in 30%. Predictors of recurrence include invasion of the myometrium, particularly deep invasion, and overgrowth of the sarcomatous component, which worsens prognosis very considerably, close to that of carcinosarcoma. Carcinoﬁbromas are extremely rare tumours, with malignant epithelium and abundant benign neoplastic mesenchymal stroma. They behaviour is between adenosarcoma and carcinosarcoma. Carcinosarcomas (malignant mixed Müllerian tumours, mixed mesodermal tumours) are usually large, soft, polypoid uterine tumours which may ﬁll the cavity and histologically show a mixture of malignant epithelial and malignant mesenchymal elements (Fig. 3.5). Carcinosarcomas are monoclonal tumours that are considered to be carcinomas from which sarcomatous elements have evolved. The grade and type of the carcinomatous component is of most prognostic value. There is an association with prior pelvic irradiation and with (a)

(b)

(c)

(d)

Fig. 3.5 Malignant mixed Müllerian tumour (MMMT). (a) The biphasic nature of MMMT is illustrated with the epithelial or carcinomatous component represented by undiﬀerentiated adenocarcinoma cells with large ovoid pleomorphic nuclei and a large amount of cytoplasm (lower left) and the mesenchymal or sarcomatous component represented by more elongated spindle cells (upper right). (b) Undiﬀerentiated adenocarcinoma cells with prominent nuclear pleomorphism. (c) Sarcomatous component of elongated spindle cells forming bundles also showing abnormal multipolar mitotic ﬁgures. (d) An area of large, bizarre tumour cells with marked nuclear enlargement with striking atypia and an abnormal mitotic ﬁgure (all H&E, magniﬁcation × 400).

43

Chapter 3: Pathology of gynaecological cancers

the predisposing features usually associated with endometrial carcinoma, including unopposed oestrogenic stimulation. The mesenchymal elements may resemble endometrial stroma, ﬁbrosarcoma or other homologous elements, or may include heterologous elements (in 50–70% of cases), including striated muscle, cartilage, osteoid and fat. They are highly aggressive tumours with ﬁve-year survival rates between 5% and 40% (all stages). Recent large series suggest a better survival for women with stages I and II tumours of 40–60%. Prognosis also depends on the depth of myometrial invasion, and women with involvement of the outer half having a poorer prognosis. Vessel or lymphatic invasion has been shown to be a poor prognostic feature, as is the presence of serous or clear cell carcinoma as the malignant epithelial component. Most tumours recur within two years of primary diagnosis, and death is usually due to complications locally rather than at distant metastatic sites.

Cervical cancer Microinvasive squamous cell carcinoma *

Stage IA1 carcinoma has a very low risk of metastasis.

Early stromal invasion often begins as one or more tongues of more eosinophilic squamous carcinoma breaking through the basement membrane underlying high-grade CIN, either grade 2 or 3 (Fig. 3.6). The risk of invasion is higher in CIN 3, with aggressive features that include gland expansion by CIN 3, central cell death or squamous maturation. In the USA, the term low-/high-grade squamous intraepithelial lesion (SIL) is in use. Microinvasive carcinoma of the cervix is deﬁned as less than 5 mm in depth and 7 mm in width and this includes stages IA1 and IA2 disease. Tumours within the 1A1 group showing less than 3.0 mm depth of invasion have a very low risk of nodal metastasis of 0.5–2.0%, whereas the risk is approximately 6–8% for IA2 tumours showing 3.0–5.0 mm depth of invasion. The relevance of lymphovascular space involvement has been the subject of disagreement and it may increase risk of nodal spread both below and above 3.0-mm depth of invasion. The margins of the cone or loop specimen must be free of carcinoma and CIN to make the diagnosis of microinvasive carcinoma, otherwise key information is considered to be incomplete and the term early stromal invasion should be used. Tumour volume has been used as an accurate measure of risk of metastasis, but is time consuming to assess and not widely used. Grading microinvasive carcinomas is very diﬃcult as they are very small tumours. Fig. 3.6 Cervical intraepithelial neoplasia 3 (on the left) with a focus of budding microinvasive squamous cell carcinoma forming small invasive clusters (H&E, magniﬁcation × 200).

44

Chapter 3: Pathology of gynaecological cancers

Invasive squamous cell carcinoma (stage IB and above) *

The most important prognostic factors are stage, depth of invasion and nodal metastasis.

Larger squamous carcinomas are usually classiﬁed according to predominant cell type into large cell keratinizing, large cell non-keratinizing and small cell non-keratinizing carcinomas. Small cell non-keratinizing carcinomas have a poorer ﬁve-year survival than the other two types. Grading depends upon the extent of keratinization and pleomorphism together with other features; however, most studies have not found histological grade to have prognostic value.

Other prognostic factors and human papillomaviruses The most important prognostic factors are stage, size, depth of invasion, presence of lymphovascular space invasion and presence and extent of lymph node metastases. The distance between the deepest point of the tumour and the deep cervical resection plane is important, and parametrial margin positivity predicts local recurrence. Squamous cell carcinoma of the cervix spreads by local invasion of tissues and lymphatics and less frequently to distant sites by haematogenous spread. Local spread is via the parametrial tissues and then extends to the pelvic side wall and adjacent organs, sometimes obstructing the ureters. Fewer than half of stage IB and stage II carcinomas have metastatic tumour involvement of nodes, usually the internal iliac, obturator, external iliac and common iliac nodes, with extension to the para-aortic, inguinal and sacral nodes occurring late in the disease. Around 95–99% of squamous carcinomas contain human papillomavirus (HPV) DNA, with high-risk HPV types 16 and 18 together accounting for about 70–80% of cancers. The existence of HPV DNA-negative tumours is controversial. Low-risk HPV types 6 and 11 are mostly found in warts and CIN 1. There are over 200 HPV types and, after types 16 and 18, the next most prevalent are the intermediate-risk HPV types 45, 31 and 33, accounting for a further 10–12% cancers, with other less frequent HPV types found in the remaining cancers. The DNA of the most common high-risk HPV types 16 and 18 is often found integrated within the host cell genome and this leads to deregulation of expression of the two viral oncogenes E6 (degrades p53) and E7 (inactivates Rb). Vaccines have been produced that are immunogenic and safe with high eﬃcacy, consisting of virus-like particles (VLPs) made from self-assembling L1 capsid proteins. Vaccines currently include combinations of VLPs for either HPV types 16 and 18 or HPV types 6, 11, 16 and 18. In the future, vaccination programmes are predicted to reduce incidence of cervical and vulval cancers, but in the meantime continued cervical screening by modern liquid-based cytology techniques remains essential.

Variants of squamous cell carcinoma Verrucous carcinoma is a rare, large, exophytic variant of very well-diﬀerentiated squamous carcinoma that extends locally but rarely involves lymph nodes. Wide surgical excision usually provides a good outcome. Warty carcinoma is a typical squamous carcinoma with warty changes. These have been suggested to be less aggressive than non-warty squamous carcinomas. Papillary squamous cell carcinoma is a rare variant with similarities to transitional cell carcinoma of the urinary bladder, but prognosis appears similar to invasive squamous carcinoma.

45

Chapter 3: Pathology of gynaecological cancers

Adenocarcinoma * *

Stage IA adenocarcinoma has a very low risk of metastasis. Prognosis depends on stage, size, type, grade and depth of invasion.

The precursor of cervical adenocarcinoma is cervical glandular intraepithelial neoplasia (CGIN), graded as either high or low grade (previously known as adenocarcinoma in situ). Controversy exists regarding whether low-grade CGIN is premalignant. Adenocarcinomas also associate with high-risk HPV, particularly HPV type 18. They account for about 25% of all cervical cancers, representing an increase in frequency over the past 30 years. The same staging system used for squamous carcinoma may be applied to adenocarcinoma, although there are fewer data for stage IA adenocarcinomas that invade <5 mm in depth and <7 mm in width (also known as early invasive adenocarcinomas or microinvasive adenocarcinomas); these tumours have around 98% ﬁve-year survival. This tumour has several histological types. The commonest is the mucinous endocervical type, closely followed by endometrioid carcinoma, with clear cell carcinoma, intestinal-type mucinous, minimal deviation, serous, mesonephric and villoglandular adenocarcinomas being less frequent. Poor prognostic features in the two common types include primarily higher stage, and also tumour size, depth of invasion, grade as well as nodal involvement. It may be diﬃcult to distinguish an endometrioid type of cervical adenocarcinoma from spread to the cervix of an endometrioid adenocarcinoma of the endometrium. Strong, diﬀuse p16 staining on immunohistochemistry or HPV DNA positivity indicate endocervical origin, but the presence of either extensive CGIN in the adjacent endocervical epithelium or complex atypical hyperplasia in the endometrium is more reliable.

Variants of adenocarcinoma Clear cell adenocarcinoma used to be associated with diethylstilbestrol (DES) exposure, but it may occur in its absence. Minimal deviation adenocarcinoma (also known as adenoma malignum) shows little or no pleomorphism and no clear stromal reaction to the invasive tumour, making diagnosis diﬃcult. It is a rare tumour and occasionally associated with Peutz–Jeghers syndrome. Serous carcinoma is very rare and appears similar to serous carcinoma of the ovary or endometrium, with a poorer outcome. Mesonephric carcinoma arises from mesonephric duct remnants, is a very rare tumour and must be diﬀerentiated from mesonephric duct hyperplasia. Villoglandular adenocarcinoma is a well-diﬀerentiated variant of cervical adenocarcinoma, occurring predominantly in young women. Most are only superﬁcial tumours, with an excellent prognosis after local excision; however, deeply invasive villoglandular adenocarcinomas have a poorer prognosis. Adenosquamous carcinomas contain both histologically malignant squamous and glandular elements. They are relatively more frequent in younger women and account for about 40% of cervical adenocarcinomas in women below the age of 50 years. They may be associated with adjacent CIN and CGIN and with HPV. There is uncertainty about their prognosis.

Rare variants of cervical carcinoma Glassy cell, mucoepidermoid, adenoid cystic, adenoid basal, carcinoid-like and small cell carcinomas of neuroendocrine type are all rare tumours arising in the cervix. Of these perhaps the most important is small cell carcinoma of pulmonary type or neuroendocrine carcinoma. This is a highly aggressive tumour with an extremely poor prognosis and occurs in younger

46

Chapter 3: Pathology of gynaecological cancers

women around 40–50 years. It is composed of small dark cells resembling oat cell carcinoma of the lung and is associated with high-risk HPV.

Vaginal cancer Vaginal squamous carcinoma *

Squamous carcinoma of the vagina is uncommon and usually has a poor prognosis.

Squamous carcinoma of the vagina is an uncommon tumour and is mostly found in elderly women, although 10% of patients are below 40. Most are seen in association with extensive vaginal intraepithelial neoplasia (VaIN) and/or CIN with high-risk HPV. There are no agreed criteria to deﬁne microinvasive carcinoma of the vagina. Prognosis depends on stage and amenability to surgery. Most cancers occur in the upper third of the vagina, and these usually spread to internal pelvic nodes whereas those vaginal cancers occurring in the lower third spread to inguinal nodes.

Vaginal clear cell carcinoma This tumour of adolescent girls and young women is rare and associated with intrauterine exposure to DES. It develops within areas of adenosis, where the vagina is lined by glandular epithelium or glands are present beneath the epithelial surface.

Vulval cancer Superﬁcially invasive carcinoma of the vulva *

Stage IA carcinoma of the vulva has a negligible risk of nodal metastasis.

There is no internationally agreed deﬁnition of the term microinvasive carcinoma of the vulva. The current staging system (see Appendix 3) deﬁnes a stage IA cancer as having stromal invasion less than 1.0 mm deep when measured from the adjacent epithelial–stromal junction. Stage IA cancers have a negligible risk of nodal metastasis, but risk of metastasis increases with depth of invasion beyond 1.0 mm. Most superﬁcially invasive carcinomas of the vulva are squamous cell carcinomas associated with usual-type vulval intraepithelial neoplasia (uVIN), mostly VIN 2 or 3 and with high-risk HPV, and are seen in a relatively young age group (Fig. 3.7).

Squamous carcinoma *

There are two groups of vulval squamous carcinoma – those occurring in young women with usual-type VIN and those occurring in elderly women without uVIN.

There are two groups of squamous carcinoma of the vulva: the ﬁrst group is associated with uVIN (graded 1–3) with high-risk HPV types 16 and 18, with a peak incidence in women in their forties and ﬁfties, and these usually have a warty or basaloid histopathological pattern. The second group occurs in association with either no pre-existing epithelial abnormality or squamous hyperplasia, lichen sclerosus or diﬀerentiated VIN (dVIN – not graded), in women in an older age group. These latter tumours are also associated with p53 abnormalities, but not with HPV. Around two-thirds of squamous carcinomas fall into the second

47

Chapter 3: Pathology of gynaecological cancers

(a)

(b)

Fig. 3.7 (a) Vulval biopsy specimen showing vulval intraepithelial neoplasia (VIN) 3 with neoplastic cells occupying almost the full thickness of the epithelium (H&E, magniﬁcation × 200). (b) Specimen immunostained (magniﬁcation × 50) for cytokeratin displaying islands and clusters of invasive squamous carcinoma inﬁltrating through the stroma underlying the surface VIN (on the right).

group and are often well diﬀerentiated with keratin pearls. Prognosis in squamous carcinoma is determined by the size, depth of invasion and degree of histological diﬀerentiation of the tumour and the presence and extent of lymph node metastases. When sentinel lymph node surgery is performed, the node is examined at multiple levels using immunohistochemistry for epithelial cytokeratins to identify isolated tumour cells (<0.2 mm) or micrometastases (>0.2 mm and <2.0 mm).

Paget’s disease of the vulva *

Prognosis of Paget’s disease depends on whether there is dermal invasion or associated carcinoma.

Paget’s disease of the vulva is a condition where mucin-producing adenocarcinoma cells are present within the squamous epithelium. These intraepithelial malignant glandular cells often spread to involve underlying adnexal structures (hair follicles, sebaceous and sweat glands). A small proportion have an underlying adnexal or other contiguous carcinoma. They may arise from intraepidermal pluripotent stem cells or sweat gland duct cells. Prognosis depends on whether the tumour has invaded the underlying dermis through the basement membrane or whether there is an associated invasive carcinoma. Complete local excision is diﬃcult as microscopic spread is often wider than macroscopically obvious erythematous areas of tumour. Hence, local recurrence is a common problem.

Other malignant tumours Other malignant vulval tumours are uncommon. The most important are basal cell carcinomas which may be cured by simple local excision. Vulval malignant melanomas usually present late and have a poor prognosis, depending on the depth of invasion (<0.75 mm deep have a good prognosis, 0.75–1.5 mm deep have an intermediate prognosis and >1.5 mm deep have a poor prognosis). Rarely tumours may arise from Bartholin’s gland of the vulva; there are several types, including squamous carcinoma, adenocarcinoma, adenosquamous carcinoma and adenoid cystic carcinoma.

48

Chapter 3: Pathology of gynaecological cancers

Further reading *

Arends MJ, Buckley H, Wells M. Cervical neoplasia: aetiology, pathogenesis and pathology. J Clin Pathol 1998; 51: 96–103.

*

Clement PB, Young RH, eds. Atlas of Gynaecologic Surgical Pathology, 2nd edn. Philadelphia, PA: Saunders Elsevier, 2008.

*

International Union Against Cancer. TNM Classiﬁcation of Malignant Tumours, 6th edn. New York: John Wiley & Sons, 2002.

*

Smart GE, Arends MJ, Al-Nafussi A. Vulval neoplasia. In Langdon SP, Miller WR, Berchuck A, eds. Biology of Female Cancers. Boca Raton, FL: CRC Press, 1997, pp. 263–70.

*

Stanley M. Prophylactic HPV vaccines: prospects for eliminating ano-genital cancer. Br J Cancer 2007; 96: 1320–3.

49

4

Chapter

Imaging in gynaecological oncology Nyree Griﬃn and Evis Sala

Principles of imaging Imaging is essential in the evaluation of patients with gynaecological malignancies. Imaging modalities used in gynaecological oncology include ultrasound (US), computed tomography (CT), magnetic resonance imaging (MRI), ﬂuorine-18-ﬂuoro-2-deoxy-d-glucose (FDG) positron-emission tomography (PET) and, more recently, FDG-PET/CT. These modalities have a variable role in screening, diagnosis, staging, treatment planning, treatment follow-up, detection of recurrence and in predicting prognosis. Skeletal scintigraphy has traditionally been useful for evaluating bony metastases, though its role as a ﬁrst-line investigation in gynaecological malignancies is limited. Other imaging tools that have recently been introduced include MRI lymphography for the detection of nodal metastases and functional MRI. This chapter is organized in three parts. The ﬁrst part discusses the technique and recommended protocols for US, CT, MRI and FDG-PET/CT, and describes how these modalities have been used in imaging of ovarian, endometrial, cervical and vulval carcinomas. Other rarer malignancies such as uterine sarcomas, gestational trophoblastic neoplasia and vaginal carcinoma will not be speciﬁcally discussed here, although MRI and CT assist in locoregional and distant staging of these tumours, respectively, and in the assessment of their suitability for surgical resectability. The second section provides a recommended imaging algorithm for the more common gynaecological cancers. Finally, the third part discusses the role of other imaging modalities such as skeletal scintigraphy and MR lymphography.

Imaging modalities Ultrasound Technique and general issues Transabdominal ultrasound is performed using a convex abdominal transducer operating between 3.5 MHz and 5.0 MHz. For optimal views of the uterus and ovaries, the patient must have a full bladder to provide a sonic window. For more detailed views of the endometrium and of the adnexa, a transvaginal ultrasound (TVS) is essential as there is closer apposition to the pelvic organs; the patient must have an empty bladder, and a higher-frequency transducer

Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 4: Imaging in gynaecological oncology

(between 5 MHz and 8 MHz) is used. Transrectal ultrasound and saline infusion sonohysterography (US-HSG) are other techniques that provide high spatial resolution, with US-HSG being more accurate than TVS in the detection of endometrial abnormalities and in diﬀerentiating endometrial polyps from submucosal leiomyomas. Colour, power and spectral Doppler assists assessment of abnormal vascularity. Ultrasound is a relatively inexpensive, quick, portable and widely available investigation, which involves no ionizing radiation or contrast medium and no patient preparation. However, it is operator dependent and views may be suboptimal in the presence of excess bowel gas or large body habitus. Ultrasound is the primary imaging modality in the initial assessment of suspected gynaecological pathology. It is used to evaluate a suspected pelvic mass, characterize adnexal lesions and to identify endometrial abnormalities in the postmenopausal patient. Transabdominal and transvaginal ultrasound can assist in image-guided ﬁne needle aspiration cytology (FNAC) or biopsy and can also be used to guide placement of the brachytherapy applicator for cervical and endometrial malignancy.

Role of ultrasound in ovarian carcinoma Screening Ultrasound has a role in both the detection of ovarian cancer and the characterization of adnexal masses. However, it is not useful in staging. Ultrasound fulﬁls several criteria as a screening test, being widely available and easy to perform with a high sensitivity and speciﬁcity. Screening is more likely to detect lower-grade epithelial ovarian tumours at an early stage, such as mucinous or endometrioid, rather than the more aggressive serous papillary carcinoma, which usually presents at stage III or IV at diagnosis. Screening with ultrasound and cancer-associated antigen (CA)125 has been advocated in high-risk patients with one of three gene mutations associated with ovarian carcinoma, such as BRCA1, BRCA2 and hereditary non-polyposis colorectal cancer (HNPCC). The largest randomized trial of screening into ovarian cancer commenced in 2001 and is currently in progress – the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS). In this trial, postmenopausal women aged between 50 and 74 years have been randomized to: no treatment; annual CA125 screening (interpreted using a risk of ovarian cancer algorithm) with transvaginal ultrasound as a second-line test (multimodal screening (MMS)); or annual screening with transvaginal ultrasound only. The preliminary ﬁndings showed that the two screening procedures had similar sensitivity for all primary ovarian and tubal cancers, but speciﬁcity was higher with MMS (99.8% versus 98.2%; p <0.0001). The results of ongoing screening are awaited to determine the eﬀect of screening on mortality. Characterization of ovarian masses Many studies have shown the value of ultrasound in assessing the morphology of ovarian masses, with several scoring systems described. Features suspicious for malignancy include wall irregularity, thick septations (>3 mm), papillary projections, solid components, size (>4 cm), bilaterality and presence of ascites or metastases. Colour Doppler ultrasound assesses tumour neovascularity (Fig. 4.1). New blood vessels are fragile and on duplex have a characteristic waveform with a low resistive index (RI <0.4).

52

Chapter 4: Imaging in gynaecological oncology

Fig. 4.1 Transabdominal colour Doppler ultrasound of patient with stage III ovarian carcinoma, demonstrating ascites and ovarian mass (arrows) with vascularity noted within the soft tissue component.

Fig. 4.2 Transvaginal colour Doppler ultrasound in a patient with endometrial carcinoma, demonstrating thickened endometrium with abnormal vascularity.

Role of ultrasound in endometrial cancer Ultrasound is usually performed in the context of postmenopausal bleeding. Usually a normal postmenopausal endometrium is atrophic, measuring less than 5 mm thick and appearing as a thin echogenic line. Endometrial thickness must be assessed on TVS by measuring the two opposing layers in the sagittal plane, excluding endometrial ﬂuid in the measurement. Endometrial cancer appears as non-speciﬁc thickening of the endometrium (Fig. 4.2), which is diﬃcult to distinguish from hyperplasia or a polyp. The diagnosis of endometrial cancer is usually established on the basis of endometrial biopsy or curettage, usually undertaken at the time of hysteroscopy. Colour and pulsed Doppler improves diagnostic accuracy as most endometrial carcinomas have abnormal blood ﬂow, with intratumoural or peritumoural blood vessels. Endometrial carcinoma can be associated with low impedance, both in the intramyometrial arteries and in the uterine arteries. Ultrasound has a limited role in the staging of endometrial cancer.

Role of ultrasound in cervical cancer Ultrasound has no role in the diagnosis of cervical cancer. Transvaginal and transrectal ultrasound may be used to assess the local extent of disease but there are limitations in detecting parametrial and pelvic side wall extension, due to poor soft tissue contrast and small ﬁeld of view. A bulky tumour may be visualized as an irregular hypoechoic mass causing enlargement of the cervix. Fluid may be seen in a distended uterine cavity if the tumour obstructs the endocervical canal, resulting in hydro- or haematometra. Hydronephrosis can also be demonstrated on ultrasound.

Role of ultrasound in vulval cancer Ultrasound may be used to detect involvement of inguinal lymph nodes, which may be enlarged and of abnormal morphology. Groin nodes are an important prognostic marker, as

53

Chapter 4: Imaging in gynaecological oncology

women with negative groin nodes have approximately 90% chance of ﬁve-year survival whereas a woman with positive nodes has approximately 50% chance of survival. Ultrasound-guided FNAC and core biopsy can be carried out on suspicious nodes. If cytology/histology is negative, groin dissection may be avoided.

Computed tomography Technique and general issues Faster and higher-resolution images are now possible due to the development of helical (spiral) and multidetector CT (MDCT). Images are acquired in the axial plane but can then be reformatted in the sagittal and coronal planes. Contrast-enhanced CT of the abdomen and pelvis is performed in the portal venous phase, 70 seconds following an injection of intravenous low-osmolar contrast medium; this enhances blood vessels and viscera, allowing easier identiﬁcation of lymphadenopathy and parenchymal lesions, especially in the liver. Contraindications to use of intravenous contrast medium include allergy to iodine and severe asthma. Use of oral (and more rarely rectal) contrast medium is mandatory to opacify the bowel and assist in detection of bowel serosal deposits that may occur in ovarian and endometrial tumours. CT is useful in evaluating the presence of distant metastases and recurrent pelvic tumours but has a limited role in characterizing early-stage pelvic disease. The newer-generation CT equipment provides better delineation of uterine anatomy, although MRI is still the gold standard. Disadvantages of CT include the use of ionizing radiation, the risk of morbidity with iodinated contrast agents, and image degradation with large body habitus and metallic hip prostheses.

Role of CT in ovarian carcinoma The two most important prognostic factors are tumour stage and residual size of tumour deposits after cytoreductive surgery. Staging laparotomy is traditionally the gold standard technique for assessing ovarian cancer, involving total abdominal hysterectomy, bilateral salpingo-oophorectomy, omentectomy and lymphadenectomy. Peritoneal cytological specimens and multiple peritoneal biopsies are also obtained. The International Federation of Gynecology and Obstetrics (FIGO) and TNM classiﬁcations of ovarian cancer are based on intraoperative ﬁndings, including histological and cytological evaluation. Over the past two decades, cross-sectional imaging has become established in the preoperative assessment, with modiﬁed imaging staging criteria, replacing conventional imaging studies such as barium enema and intravenous urography. Diagnosis In advanced ovarian carcinoma, CT can be used to guide percutaneous biopsy of peritoneal or adnexal disease. This has been shown to be a safe technique with high diagnostic accuracy. Staging and assessment of resectability CT is currently the investigation of choice for preoperative assessment and radiological staging of ovarian carcinoma (Fig. 4.3), allowing for individualized treatment. It provides

54

Chapter 4: Imaging in gynaecological oncology

Fig. 4.3 Axial contrast-enhanced CT in the same patient as in Figure 4.1, demonstrating peritoneal metastases (white arrow) and adnexal mass (black arrow).

Box 4.1 CT criteria for surgical non-resectability in ovarian carcinoma Tumour deposit greater than 1 cm: *

Dome of liver

*

Subcapsular implant

*

Intersegmental ﬁssure

*

Porta hepatis

*

Gastrohepatic/gastrocolic/gastrosplenic ligaments

*

Lesser sac

*

Root of mesentery

*

Enlarged lymph nodes above renal hilum

information on the primary tumour, the site and size of peritoneal deposits and the presence of enlarged lymph nodes, identifying those patients with non-resectable disease who would beneﬁt from neo-adjuvant chemotherapy rather than primary cytoreductive surgery. Box 4.1 shows the criteria for non-resectability, although some studies suggest peritoneal deposits should be greater than 2 cm rather than 1 cm to deﬁne surgical inoperability. Ovarian masses may appear as mixed solid/cystic tumours that are often bilateral, or as multilocular cystic lesions with thick internal septations and solid mural or septal components. Calciﬁcation may be seen in the wall and soft tissue components. Pelvic side wall invasion is suggested by tumour encasing the iliac vessels and should be suspected if tumour is within 3 mm of the pelvic side wall. Involvement of the rectosigmoid colon or bladder is suggested by frank invasion or by replacement of the normal fat planes by solid tumour. Peritoneal deposits characteristically involve the greater omentum, the liver, splenic, diaphragmatic and bowel surfaces, the paracolic gutters, the pouch of Douglas, the gastrosplenic, gastrocolic and gastrohepatic ligaments, the porta hepatis, the mesentery and the

55

Chapter 4: Imaging in gynaecological oncology

lesser sac. Liver surface deposits cause scalloping of the underlying liver. Deposits appear as discrete, enhancing soft tissue lesions or as subtle areas of plaque-like thickening involving the parietal or visceral peritoneum. More rarely, peritoneal lesions may be cystic. Foci of calciﬁcation may be seen within deposits from serous tumours, a feature which may become much more prominent following chemotherapy. A large volume of ascites is associated with a positive predictive value of 72–80% for peritoneal spread, indicating stage III ovarian carcinoma. Involvement of para-aortic and pelvic nodes is suspected on CT on the basis of size greater than 1 cm, giving a sensitivity of approximately 40% and speciﬁcity of up to 90%. The presence of superior diaphragmatic lymph nodes is deﬁned as signiﬁcant if greater than 5 mm in short axis, and these are independent predictors for recurrence and death. Haematogenous spread to solid abdominal organs is unusual and indicates stage IV disease, with metastases most likely in the liver, lung and kidneys. It is important to distinguish between subcapsular liver implants and liver parenchymal metastases, as this alters staging and therapy. Dissemination outside the abdominal cavity (such as malignant pleural eﬀusion) also indicates stage IV disease.

Role of CT in endometrial cancer CT with both intravenous and oral contrast medium may be used to stage endometrial carcinoma (Figs. 4.4 and 4.5), although MRI is the imaging modality of choice. Endometrial tumours appear as low-attenuation lesions relative to normally enhancing myometrium. CT is most useful in conﬁrming the presence of enlarged lymph nodes, which indicates stage III disease. There is a tendency to understage bladder/bowel involvement. Cervical extension is diﬃcult to evaluate as the cervix is poorly delineated on CT. A sensitivity of 83% and speciﬁcity of 42% have been reported for the detection of stage IB disease and a sensitivity of 25% and speciﬁcity of 70% for the detection of cervical involvement (stage II).

Fig. 4.4 Axial contrast-enhanced CT in a patient with recurrent endometrial carcinoma presenting with mediastinal adenopathy (arrow).

56

Fig. 4.5 Axial contrast-enhanced CT in the same patient as in Figure 4.4, with recurrent endometrial carcinoma presenting with lung metastases (arrows).

Chapter 4: Imaging in gynaecological oncology

Role of CT in cervical cancer CT has a diagnostic accuracy of approximately 90% in staging advanced cervical carcinoma. Nodal status can also be assessed. Large tumours appear as low-attenuation masses with possibly a rim of high attenuation and associated hydro/haematometra, whereas small tumours are isodense to normal cervical tissue and may only be appreciated as cervical enlargement. Early parametrial extension is poorly demonstrated, with CT having a positive predictive value of approximately 60%.

Role of CT in vulval cancer CT is used for distant staging of vulval cancer, as nodal metastases and haematogenous dissemination can be easily assessed.

Magnetic resonance imaging Technique and general issues MRI is the imaging modality of choice for evaluation of pelvic pathology, oﬀering superb spatial and contrast resolution and multiplanar capability. It is superior to CT in assessment of uterine and cervical cancer and acts as a problem-solving tool in ovarian cancer. No ionizing radiation is involved, and breath-hold techniques allow for faster imaging. MRI is contraindicated in patients with claustrophobia, pacemakers, neural stimulators or cochlear implants, certain vascular clips and metallic implants and intraocular foreign bodies. The patient is imaged supine with a pelvic or cardiac phased-array coil, which increases the signal-to-noise ratio. A partially ﬁlled bladder is helpful in displacing loops of small bowel from the pelvis. Prior to the study, an intramuscular or intravenous injection of hyoscine butylbromide (Buscopan, an antimuscarinic) or glucagon reduces bowel peristalsis, which may reduce the image quality because of motion artefact. Standard sequences include axial T1- and axial and sagittal T2-weighted sequences (small ﬁeld of view, FOV). The T1-weighted sequence delineates pelvic structures from surrounding fat. This sequence also evaluates for the presence of enlarged lymph nodes and bone marrow metastases. In addition, blood, proteinaceous products or fat have high signal intensity on T1 images, with the latter being suppressed on sequences where fat saturation is applied. The T2-weighted sequence is helpful in demonstrating pathology. Normal signal characteristics of the uterus on T2-weighted sequence include a high signal intensity-endometrium (<5 mm in thickness in the postmenopausal patient), a junctional zone of low signal intensity (normally measuring <8 mm in thickness) and a myometrium of intermediate signal intensity. The endocervical canal has a high signal intensity and the cervical stroma has a low signal intensity. The presence of a cervical or uterine tumour causes distortion of the normal anatomy and signal characteristics on MRI. Axial and coronal oblique T2-weighted sequences are also performed when imaging the cervix to assess parametrial extension, with images acquired along a plane either perpendicular or parallel to the cervix, respectively. Intravenous administration of gadolinium is used for characterization of adnexal lesions, or to distinguish recurrent tumour from post-treatment ﬁbrosis. Dynamic contrast-enhanced, T1-weighted images in the sagittal and axial plane are routinely used for staging of endometrial cancer. Axial and coronal fat-suppressed T2-weighted sequences are used for staging of vulval cancer where the tumour and involved lymph nodes appear to be of high signal intensity. An axial

57

Chapter 4: Imaging in gynaecological oncology

T1-weighted sequence through the upper abdomen is routinely performed in patients with suspected gynaecological malignancies to evaluate for the presence of enlarged para-aortic lymph nodes.

Role of MRI in ovarian carcinoma Characterization of ovarian masses MRI remains predominantly a problem-solving modality in evaluating adnexal masses following inconclusive pelvic ultrasound. In a meta-analysis of indeterminate masses in sonography, MRI demonstrated superiority over CT and Doppler sonography in predicting malignancy. However, it is generally accepted that MRI cannot reliably diﬀerentiate borderline from malignant tumours, although tumours that are predominantly solid with areas of necrosis and multiple papillary projections suggest the latter. Ninety-ﬁve per cent of malignant adnexal masses can be correctly characterized on MRI (Fig. 4.6) using primary and ancillary criteria. Primary criteria for malignancy are: size greater than 4 cm; irregular wall thickness greater than 3 mm; cystic lesion with solid component present; septa greater than 3 mm and/or presence of vegetations or nodularity; and solid mass with areas of necrosis. Ancillary criteria are: involvement of pelvic organs or sidewall; presence of peritoneal, mesenteric or omental disease; ascites; and lymph node involvement. Subtypes of ovarian epithelial carcinoma include clear cell and endometrioid tumours, which can be associated with endometriosis; malignancy should be suspected if a nodule is seen within a predominantly cystic endometrioma. Serous cystadenocarcinomas are frequently bilateral, appearing as solid/cystic masses with solid enhancing components and areas of necrosis; large-volume ascites, enlarged lymph nodes and peritoneal deposits are often present. A multilocular, thin-walled cyst with ﬂuid of varying signal intensity on T2-weighted imaging may suggest a mucinous cystadenoma (because of the proteinaceous content), with a greater number of loculi suggesting a mucinous cystadenocarcinoma. The latter may be associated with pseudomyxoma peritonei, which appears as mucinous, lobulated collections of low signal intensity on T1-weighted images and high signal intensity on T2-weighed images, with collections causing scalloping of the liver and spleen. Fig. 4.6 Axial T2-weighted MR image in the same patient as in Figures 4.1 and 4.3, again demonstrating peritoneal metastases (white arrow) and an adnexal mass (black arrow).

58

Chapter 4: Imaging in gynaecological oncology

Staging and assessment of resectability MRI has a staging accuracy comparable with CT (78–88% for MRI versus 53–92% for CT). CT is preferred as a ﬁrst-line investigation, with MRI being recommended if there is an iodine allergy, if the patient is pregnant or if the ﬁndings are inconclusive on CT. However, MRI takes longer to perform and interpret. T1 and gadolinium-enhanced, fat-suppressed T1-weighted sequences should be performed in at least two planes; the coronal plane is useful in assessing peritoneal deposits related to the diaphragm and liver. Peritoneal metastases are often best demonstrated on the delayed (5–10 minute) images. The bowel can be opaciﬁed with dilute barium sulphate suspension to facilitate the detection of peritoneal metastases. Both CT and MRI have a high sensitivity for detection of peritoneal disease, especially when implants are larger than 1 cm. Diﬀusion-weighted imaging is a new technique oﬀering improved detection of peritoneal metastases. Lymph node involvement is assessed on the basis of size, as with CT. However, enlarged lymph nodes may be reactive in nature, and tumour cannot be detected in normal-sized lymph nodes with conventional MRI. MRI is more accurate than CT in evaluating tumour resectability. Assessment of follow-up response and detection of recurrent disease In women treated for ovarian carcinoma, gadolinium-enhanced MRI has a sensitivity, speciﬁcity and accuracy of 90%, 88% and 89%, respectively, for the detection of residual or recurrent peritoneal disease. This is comparable with laparotomy and superior to CA125 (sensitivity, speciﬁcity and accuracy of 65%, 88% and 67%, respectively). Recurrent disease is considered non-resectable on MRI when it lies within 3 mm of the pelvic side wall.

Role of MRI in endometrial cancer Diagnosis MRI does not have a role in the detection of endometrial carcinoma, which is usually diagnosed on transvaginal ultrasound and endometrial biopsy. The role of MRI in staging and treatment planning Clinical staging may understage endometrial carcinoma by up to 22% and hence in 1998, FIGO adopted a surgicopathological staging system, which has been revised in 2009 (see Appendix 3). The depth of myometrial invasion is probably the single most important morphologic prognostic factor, as it correlates with tumour grade, tumour extension into the cervix and pelvic and para-aortic lymph node metastases. MRI is better than clinical assessment, CT and ultrasound at staging endometrial carcinoma, with a staging accuracy of 85–93%. Imaging criteria for staging of endometrial cancer are based on the TNM/FIGO classiﬁcation. MRI can assist in treatment planning, as it determines which patients would beneﬁt from pelvic or para-aortic lymph node sampling or lymphadenectomy in the case of enlarged lymph nodes and which patients might beneﬁt from preoperative radiotherapy when more advanced disease is suspected. On unenhanced T1-weighted images, endometrial carcinoma is isointense with the normal endometrium. On T2-weighted sequences (Figs. 4.7–4.9), the tumour typically shows heterogeneous signal intensity. Dynamic imaging during injection of gadolinium (Fig. 4.10) has a higher accuracy than the T2-weighted sequence alone, as there is early

59

Chapter 4: Imaging in gynaecological oncology

Fig. 4.7 Sagittal T2-weighted MR image demonstrating stage IA endometrial carcinoma (FIGO 2009 classiﬁcation) with extension of tumour into inner half of the myometrium (arrows).

Fig. 4.8 Sagittal T2-weighted MR image in the same patient as in Figure 4.2, demonstrating stage IA endometrial carcinoma (FIGO 2009 classiﬁcation) with extension into the endocervical canal but no stromal involvement (arrows).

Fig. 4.9 Sagittal T2-weighted MR image in the same patient as in Figures 4.4 and 4.5, with recurrent endometrial carcinoma presenting with a large lesion in the urethra/anterior wall of the vagina (arrow).

Fig. 4.10 Sagittal dynamic contrast-enhanced MRI demonstrating stage IA endometrial carcinoma (FIGO 2009 classiﬁcation) with disrupted subendometrial enhancement line (arrows).

enhancement of endometrial cancer relative to the normal endometrium, allowing identiﬁcation of small tumours. In the later phases of enhancement, tumour appears hypointense relative to the myometrium. Early physiological subendometrial enhancement is seen in approximately 50% of patients and excludes deep myometrial invasion. Diagnostic accuracy with MRI may be decreased by the following factors: poor visualization of the junctional zone in postmenopausal women; distortion of the myometrium by tumour, adenomyosis or multiple ﬁbroids; extension of tumour into the uterine cornu, where the myometrium is physiologically thinner; and absence of subendometrial enhancement.

60

Chapter 4: Imaging in gynaecological oncology

Role of MRI in cervical cancer Staging MRI, although not included in the original staging system, is now widely accepted as the best imaging tool in staging of cervical carcinoma, with an overall accuracy of approximately 90%. MRI avoids the need for invasive procedures such as cystoscopy and proctoscopy when there is no indication of bladder or rectal invasion on imaging. MRI can inﬂuence treatment planning in up to 50% of women by identifying crucial prognostic factors such as lesion volume, depth of stromal invasion and presence of metastatic lymph nodes, which indicates a poorer prognosis. MRI has a 95% negative predictive value in excluding presence of parametrial invasion. External-beam radiotherapy and brachytherapy can be optimized by the use of MRI, as it can accurately evaluate the shape and direction of growth of the tumour. In general, cervical tumours are best demonstrated on T2-weighted sequences (Figs. 4.11–4.13), where they appear of intermediate signal intensity and can be exophytic, inﬁltrating or endocervical. Smaller tumours may be better identiﬁed on dynamic contrast imaging, where they show early enhancement. Fistulous tracts into adjacent organs are also more easily identiﬁed following gadolinium administration. Treatment follow-up Tumour volumetry by MRI has been shown to be a sensitive measure of responsiveness of cervical cancer to chemoradiotherapy and helps identify women at high risk for treatment failure. Disease recurrence MRI can be used to detect tumour recurrence and distinguish tumour from radiation changes, using both T2-weighted sequences and dynamic contrast-enhanced imaging.

Fig. 4.11 Sagittal T2-weighted MR image demonstrating a large endocervical tumour (arrows) with full-thickness stromal involvement and extension into the parametria, in keeping with stage IIB cervical adenocarcinoma.

Fig. 4.12 Axial oblique T2-weighted MR image of the same patient as in Figure 4.11, demonstrating left parametrial extension (arrow).

61

Chapter 4: Imaging in gynaecological oncology

Fig. 4.13 Coronal T2-weighted MR image in a diﬀerent patient, demonstrating stage IIIB cervical tumour (white arrows) extending into the lower third of the vagina and to the pelvic side wall, causing left-sided hydroureter (black arrow).

Fig. 4.14 Axial T1-weighted MR image with fat saturation, pre-intravenous gadolinium administration, showing lesion in vulva (arrows).

Fig. 4.15 Axial T1-weighted MR image with fat saturation, post-intravenous gadolinium administration, showing enhancing lesion in vulva (arrows).

Role of MRI in vulval cancer The diagnosis of primary vulval cancer is essentially clinical. However, MRI (Figs. 4.14 and 4.15) has been shown to accurately identify the extent of primary tumour and thus assist assessment of surgical resectability. Vulval tumours demonstrate low signal intensity on T1 and intermediate to high signal intensity on T2. Fat-saturated axial (Fig. 4.16) and coronal (Fig. 4.17) T2-weighted sequences are particularly useful in demonstrating the tumour. MRI is most beneﬁcial in more advanced tumours where the relationship to adjacent structures can be assessed, e.g. anal sphincter and urethra. Recurrent tumour can also be demonstrated on MRI, retaining similar signal intensity to the primary tumour. In addition, involvement of inguinal and pelvic lymph nodes can be assessed using a size criterion for abnormal superﬁcial inguinal lymph nodes of 1 cm or greater in the short axis diameter. Morphological criteria such as loss of fatty hilum, lobular outline and signal intensity paralleling that of the

62

Chapter 4: Imaging in gynaecological oncology

Fig. 4.16 Axial T2-weighted MR image with fat saturation in the same patient as in Figures 4.14 and 4.15. There is a high signal lesion in vulva (arrows) in keeping with early-stage squamous cell carcinoma of the vulva.

Fig. 4.17 Coronal T2-weighted MR image with fat saturation in diﬀerent patient, showing a lesion in left vulva in keeping with squamous cell carcinoma of vulva (short arrow) and necrotic left inguinal lymph node (large arrow).

primary tumour are also used. The identiﬁcation of involved lymph nodes is very important for both surgical and radiotherapy planning.

FDG-PET and FDG-PET/CT FDG-PET and FDG-PET/CT are functional imaging tools which use short-lived radionuclides attached to tracers to image metabolic processes in the body. The most commonly used radiotracer is FDG, which evaluates glucose metabolism. FDG-PET and FDG-PET/CT are useful in various cancers due to the increased glycolytic rate of many malignant tumours. They are developing a role in gynaecological malignancies, especially cervical and ovarian cancer. For FDG-PET, the patient should be fasted for at least four to six hours, following which an intravenous injection of approximately 10 mCi FDG is administered. A wholebody emission scan is performed approximately one hour later with the patient in a supine position, taking approximately ﬁve to eight minutes for ﬁve to eight bed positions. Attenuation correction is achieved using a radionuclide transmission scan. Normal physiological FDG uptake is seen in the digestive tract, thyroid gland, skeletal muscle, myocardium, bone marrow and urinary tract. Uterine physiological FDG uptake can also be appreciated in the ovulatory and menstrual phase in premenopausal women. Physiological uptake can occur in ovarian follicles and cysts in the corpus luteum. FDG uptake in the uterus and ovaries in postmenopausal women should be regarded as abnormal. Intramuscular or intravenous injection of hyoscine butylbromide (Buscopan) may be helpful in reducing bowel uptake. Benign pathological FDG uptake can be seen in certain benign ovarian and uterine tumours, as well as in inﬂammatory and infectious processes. Imaging of pelvic tumours is improved if the bladder is empty prior to the study. The advantage of PET is that it images the whole body and thus can detect unexpected distant sites of disease. Ionizing radiation is still involved with standard dose of around 7 mSv.

63

Chapter 4: Imaging in gynaecological oncology

FDG-PET/CT is a relatively new technique that combines both functional and anatomical information. Novel hybrid systems have been designed integrating FDG-PET with CT. It is replacing FDG-PET in ovarian and cervical cancer imaging. Both studies are performed in the same setting with the same device, thus allowing for fusion of the image and reducing errors caused by mis-registration. FDG-PET/CT is quicker than FDG-PET alone, and CT data allow for attenuation correction. The CT study is usually performed prior to the PET study. Intravenous contrast medium is not routinely given as this may lead to artefacts. FDGPET/CT has the beneﬁt of lesion conspicuity using FDG-PET and anatomic localization on CT. This helps localize pathology and diﬀerentiate from physiological uptake. However, at present, FDG-PET/CT is not as readily available as FDG-PET.

Role of FDG-PET and FDG-PET/CT in ovarian carcinoma Diagnosis FDG-PET/CT has demonstrated a sensitivity of 100% and speciﬁcity of 92.5% in diagnosing malignant pelvic tumour and may be of value in identifying primary ovarian carcinoma prior to surgery in postmenopausal women with a pelvic mass, high CA125 and suspicious ultrasound ﬁndings. Staging and prediction of tumour response The sensitivity of FDG-PET in the detection of small-volume disease is reduced due to lack of anatomical localization, physiological uptake in the bowel and relatively low spatial resolution, which means that it cannot reliably detect disease <0.5 cm in size. FDG-PET is better at identifying metastatic disease in the para-aortic and pelvic nodal regions rather than in peritoneal disease. It has been reported that FDG-PET/CT is superior to CT in staging ovarian carcinoma and overcomes the problems of PET alone. Omental disease, detected on PET/CT prior to surgery for primary ovarian carcinoma, is an independent predictor of incomplete cytoreduction. Evaluation of treatment response FDG-PET/CT is not routinely used to monitor response to treatment due to limited availability and high cost. FDG-PET has been compared with second-look laparotomy in patients receiving primary chemotherapy to assess response, and comparable results have been shown. Detection of recurrent disease Studies have shown that the sensitivity, speciﬁcity and accuracy of FDG-PET/CT in the detection of recurrent disease is between 73% and 94.5%, between 40% and 100% and between 63% and 95%, respectively, when using clinical or histological ﬁndings as a reference standard. FDG-PET/CT evidence of recurrence can precede conventional diagnosis by six months and the technique detects a higher percentage of patients with relapse than FDG-PET or CT alone. FDG-PET/CT ﬁndings can lead to a change in clinical management in up to 60% of patients and help optimize site-speciﬁc treatment, such as radiation treatment planning, as well as assist in selection of patients for secondary cytoreductive surgery. There is, however, no evidence that commencing salvage therapy at an earlier stage improves overall survival.

64

Chapter 4: Imaging in gynaecological oncology

Role of FDG-PET in endometrial cancer At present there is little literature on the use of FDG-PET in endometrial cancer. A sensitivity of 96–100% and speciﬁcity of 78–88% have been reported in detecting recurrence. One study has evaluated the impact of FDG-PET on the management of patients with advanced (stage III/IV) endometrial carcinoma or recurrent carcinoma and shown that FDG-PET or FDGPET combined with MRI and CT has a higher sensitivity in overall lesion detection than MRI and CT alone. FDG-PET had a positive clinical impact in up to 50% of patients.

Role of FDG-PET/CT in cervical cancer Recently, a limited number of studies have reported the use of FDG-PET/CT in both primary staging and restaging of cervical cancer. For example, one study has shown a sensitivity and speciﬁcity of 92.8% for detecting cervical tumours and a sensitivity and speciﬁcity of 95.6% and 94.7%, respectively, in detecting cervical cancer metastases. FDG-PET/CT assists in the identiﬁcation of involved nodes when CT ﬁndings are negative, resulting in a change in management in up to 25% of cases. For recurrent cervical cancer, one study has reported a sensitivity, speciﬁcity and accuracy for FDG-PET/CT of 90.3%, 81.0%, and 86.5%, respectively.

Role of FDG-PET in vulval cancer There have been few studies evaluating the use of whole-body FDG-PET in vulval cancer, with one study reporting a sensitivity of 80%, speciﬁcity of 90%, positive predictive value (PPV) of 80% and negative predictive value (NPV) of 90% on a patient basis, but lower sensitivity on a groin basis.

Recommended imaging algorithms Ovarian carcinoma In the characterization of ovarian masses, ultrasound should be used as a ﬁrst-line investigation, with MRI reserved as a problem-solving tool. MRI is useful in assessing pelvic side wall involvement. CT should be used for staging in cases of suspected advanced ovarian carcinoma to evaluate for peritoneal disease and nodal metastases. CT can assess surgical resectability and complications such as hydronephrosis and bowel obstruction. In patients with advanced ovarian carcinoma deemed not suitable for primary surgery, ultrasound- or CT-guided percutaneous biopsy of omental/peritoneal disease or adnexal mass is used to establish the diagnosis prior to neoadjuvant chemotherapy. Cross-sectional imaging, primarily by CT, is used for treatment follow-up and in the detection of recurrent disease (Figs. 4.18 and 4.19). FDG-PET/CT is of value in cases of suspected recurrence where there is an increase in tumour markers but indeterminate ﬁndings on CT or MRI. Second-look surgery is no longer a routine procedure, and secondary cytoreduction is only justiﬁed if recurrent disease is deemed resectable on imaging, usually as isolated lesions in the absence of ascites.

Endometrial cancer Transvaginal ultrasound is the primary imaging approach for diagnosis of endometrial cancer. MRI is reserved for cases where hysteroscopy is unsuccessful or repeated biopsy results are inconclusive. MRI is the imaging modality of choice for staging of endometrial cancer, with implications for treatment planning and prognosis. CT is of value in evaluating

65

Chapter 4: Imaging in gynaecological oncology

Fig. 4.18 Axial contrast-enhanced CT in a patient with recurrent ovarian carcinoma with a peritoneal implant in the right paracolic gutter (arrow).

Fig. 4.20 Axial contrast-enhanced CT showing a peritoneal implant in recurrent cervical cancer (arrow). This could be mistaken for non-opaciﬁed bowel. (Courtesy of Dr H K Cheung, Addenbrooke’s Hospital.)

Fig. 4.19 Axial contrast-enhanced CT in the same patient as in Figure 4.18 with recurrent ovarian carcinoma and multiple enlarged retroperitoneal lymph nodes (arrows).

Fig. 4.21 FDG-PET/CT in the same patient as in Figure 4.20, showing pathologically increased activity in the peritoneal implant (arrow), conﬁrming recurrent cervical cancer. (Courtesy of Dr H K Cheung, Addenbrooke’s Hospital.)

for distant peritoneal or nodal metastases in patients with clear cell and serous papillary carcinoma. FDG-PET/CT may have a role in the future in post-treatment surveillance of these patients and also in assessing distant disease.

Cervical cancer For locoregional staging, MRI is the investigation of choice, being able to assess size, site (endocervical or exocervical), depth of stromal invasion and presence of enlarged lymph nodes. It is important to distinguish early disease (stages I–IIA), that can be treated with surgery, from advanced disease that must be treated with radiotherapy or radiotherapy combined with chemotherapy. CT is mostly used in the assessment of distant metastases, for radiotherapy planning and detection of recurrence (Fig. 4.20). FDG-PET and FDG-PET/ CT have shown very promising results in primary staging, treatment follow-up and detection of recurrent disease (Fig. 4.21).

66

Chapter 4: Imaging in gynaecological oncology

Vulval cancer Diagnosis of primary tumour and involvement of superﬁcial inguinal lymph nodes is essentially clinical. Ultrasound-guided FNAC is useful in assessing lymph node involvement and, if cytology is negative, may obviate the need for lymph node dissection. MRI assists in treatment planning as it can assess local extent and involvement of adjacent structures and identify involvement of deep inguinal and pelvic lymph nodes. CT is used for distant staging in locally advanced cases.

Other imaging modalities Skeletal scintigraphy Technetium (Tc)-99 m methylene diphosphonate is the most commonly used radiopharmaceutical agent in skeletal scintigraphy. The phosphonate compounds bind to the bone and, as the technetium decays, the gamma rays it releases are detected by a gamma camera. Activity on skeletal scintigraphy usually reﬂects osteoblastic activity in the bone. Images are usually acquired after two to three hours. Skeletal scintigraphy has a relatively high sensitivity but poor speciﬁcity, as other processes such as osteoarthritis, osteomyelitis and fractures can also show activity. However, metastases are suspected if activity is asymmetrical and predominantly axial in location. Correlation with other forms of imaging such as plain ﬁlm, CT or MRI is helpful. Skeletal scintigraphy can be used as a screening tool for oncological patients although bony metastases are not a prominent feature in gynaecological malignancies. Its speciﬁcity is increased if single photon emission-computed tomography is used. Bone marrow scintigraphy using Tc-99 m-marked monoclonal antibody against granulocyte antigen (NCA95) is superior to skeletal scintigraphy in the diagnosis of certain tumour types. It is well recognized that MRI can identify bone marrow metastases at an earlier stage than skeletal scintigraphy. However, MRI has a limited ﬁeld of view unless whole-body imaging is performed. FDG-PET/CT has a higher sensitivity than planar scintigraphy and also a higher spatial resolution.

Lymph node imaging with MR lymphography Nodal status is important in treatment planning and in predicting prognosis in women with gynaecological malignancies. However, cross-sectional imaging has limited sensitivity when using a size criterion of greater than 1 cm as a marker of metastatic involvement. Recently, dextran-coated ultra-small superparamagnetic iron oxide (USPIO) particles have been used with MRI as a lymph node-speciﬁc contrast agent. USPIO is given as an intravenous injection which extravasates into the interstitium and is transported to the lymph nodes. In healthy nodes, it is taken up by macrophages resulting in signal loss. Persistently high signal intensity indicates metastatic involvement. In a study of 44 patients with endometrial or cervical cancer, MRI and MR lymphography were compared with histological examination for 335 pelvic and para-aortic lymph nodes. On a node-by-node basis, the sensitivity, speciﬁcity and positive and negative predicative values by the size criterion were 29%, 99%, 56% and 96%, respectively, and by the USPIO criteria were 93%, 97%, 61% and 100%, respectively. On a patient-by-patient basis, sensitivity, speciﬁcity and positive and negative predicative values by the size criterion were 27%, 94%, 60% and 79%, respectively, and by the USPIO criteria were 100%, 94%, 82% and 100%, respectively. Hence MR lymphography signiﬁcantly improved sensitivity, with no loss in speciﬁcity in the detection of lymph node metastases.

67

Chapter 4: Imaging in gynaecological oncology

Conclusion Various imaging techniques can be used in the diagnosis, treatment planning and follow-up of the main gynaecological cancers, with ultrasound, CT, MRI and FDG-PET/CT having complementary roles. It is important to know the indications for these investigations and also their limitations to optimize patient management. In the future, it is likely that FDGPET/CT will become more readily available and develop a greater role in gynaecological oncology imaging.

Further reading *

68

Forstner R. Radiological staging of ovarian cancer: imaging ﬁndings and contribution of CT and MRI. Eur Radiol 2007; 17: 3223–46.

*

Hricak H. Diagnostic Imaging: Gynaecology. Salt Lake City: Amirsys, Elsevier 2007.

*

Rockall AG, Sohaib SA, Harisinghani MG, Babar SA, Singh N, Jeyarajah AR, et al. Diagnostic performance of nanoparticleenhanced magnetic resonance imaging in the diagnosis of lymph node metastases in

patients with endometrial and cervical cancer. J Clin Oncol 2005; 23: 2813–21. *

Sala E, Wakely S, Senior E, Lomas D. MRI of malignant neoplasms of the uterine corpus and cervix. AJR Am J Roentgenol 2007; 188: 1577–87.

*

Subhas N, Patel PV, Pannu HK, Jacene HA, Fishman EK, Wahl RL. Imaging of pelvic malignancies with in-line FDG PET-CT: case examples and common pitfalls of FDG PET. Radiographics 2005; 25: 1031–43.

5

Chapter

Surgery for gynaecological cancers Alan Farthing

Introduction The practice of surgical gynaecological oncology covers a very disparate group of procedures. This is clearly illustrated through the comparison of a radical vulvectomy and inguinal groin node dissection with a debulking laparotomy for an advanced ovarian tumour. There are, however, clear underlying principles common to all gynaecological oncology procedures. This chapter aims to provide a logical basis that will facilitate the understanding of management patterns for the speciﬁc tumour types detailed in subsequent chapters. It also attempts to deﬁne where minimal access techniques ﬁt into the conventional patterns of open surgical treatment.

Basic principles Before an operation is performed for any particular indication, a few important and basic principles need to be applied.

Choose your patient It is important to understand the potential risks of any operation as they apply to the patient in front of you. Many a surgeon has quoted to their pupils that a good surgeon is one who knows how to operate, an excellent surgeon knows when to operate but the very best surgeon knows when not to operate!

Know your aims Have a clear understanding of what you are trying to achieve with any operation. Do not be afraid to re-examine those aims during an operation if diﬃculties occur or if the risk–beneﬁt balance changes.

Know your anatomy All surgery is ‘applied anatomy’. There are some areas of medicine in which a small amount of anatomical knowledge is suﬃcient for the procedure required, but in the ﬁeld of gynaecological oncology many operations are made more diﬃcult because the anatomy is distorted by the disease process. Therefore thorough knowledge of normal and abnormal anatomy is essential.

Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 5: Surgery for gynaecological cancers

Apply good surgical techniques Many surgeons start their surgical careers with poor knowledge of basic surgical techniques such as knot tying, how to handle instruments, how to treat tissues and how to look for tissue planes. In gynaecological oncology surgery, it is important to be systematic and methodical, which helps to minimize complications. Make sure all your movements have a purpose and that thought is given to the surgical procedure and any diﬃculties encountered during the operation. Many of the studies based on the use of simulators have shown that, although the more skilled surgeons make slower hand movements, they move from point to point with precision and therefore complete any task more quickly.

Treat complications early The early recognition of complications will result in rapid treatment and a decrease in the morbidity resulting from that complication. All surgeons will experience complications, and dealing with these is an important aspect of overall care.

Be a good team member Intraoperative haemorrhage is inevitable in some gynaecological oncology operations. Keeping the anaesthetist informed of any potential haemorrhage and being realistic about the volume can help reduce complications. There are occasions when a surgeon needs to stop and compress a bleeding vessel while the rest of the team catches up with the volume loss or opens required instruments. Accept that everyone has their part to play.

Know your limitations There is a ﬁne distinction between a conﬁdent surgeon who is able to deal with the problem in hand without being inhibited by a previous complication and an individual who is overconﬁdent. Every surgeon will need to call for help from someone with greater experience in a particular area. Within gynaecological oncology procedures this is most likely to be a bowel or urological surgeon. Every surgeon is always learning, and knowing when to ask for help from someone with the appropriate skills is important to achieve good outcomes.

Preoperative assessment The purpose of any preoperative assessment is to identify factors that might make the surgery more complex or increase the risk of postoperative complications. This starts with a good history and meticulous recording in the notes of all the relevant information to achieve the optimal preoperative condition for each patient by anticipating potential problems, and thereby improving the outcome. Eﬀective communication within the team as well as with the patient during counselling will go a long way towards risk management and avoidance of complications. Irrespective of a surgeon’s dexterity, if the wrong operation is performed the outcome is likely to be suboptimal. The information that is gathered from obtaining a full history and performing a complete examination is vital. This process of collecting information needs to be initiated prior to admission. All the information, however, must be collated and reviewed pre-operatively. The use of pre-admission clinics is now the norm so that all this can take place in an organized manner.

70

Chapter 5: Surgery for gynaecological cancers

History Details of previous surgery may be important because of the risk of adhesions or complications may have occurred that increase the risk of a repeated episode, for example thromboembolic disease or postoperative bowel obstruction. Any coexisting pathology for example, insulindependent diabetes, obstructive airways disease or cardiac disease that may lead to an inability to deal with ﬂuid shifts, can signiﬁcantly inﬂuence the complication rate and should be identiﬁed. The surgeon needs to ask about symptoms to be able to identify previously undiagnosed conditions. Speciﬁc questions about medications and allergies should not be forgotten.

Examination In women undergoing major abdominal surgery, the cardiorespiratory system should be examined, together with abdominal palpation. A vaginal or rectal examination would often be part of the assessment of a pelvic mass and may well determine the safest route for an operation to be performed. Obesity may make surgical access diﬃcult and restricted movements, particularly of the hips, should be noted in case lithotomy position is required under anaesthesia. An examination may identify previously undiagnosed conditions.

Special investigations Every gynaecological oncology department, centre or unit should have protocols detailing the special investigations required for women for diﬀerent procedures. However, many preoperative investigations are performed with limited beneﬁt because these protocols assume that, for example, everyone over 50 has the same ﬁtness level and therefore the same requirements for a preoperative electrocardiogram (ECG). Where possible it is better to use the history and examination together with the likely intraoperative risks to determine the investigations. Where blood loss is possible, a preoperative haemoglobin measurement is required, and to have a sample grouped and saved in the laboratory may be helpful. Most transfusion services will have rapid cross-match times of about 15 minutes and therefore blood should only be cross-matched if it is extremely likely to be used. This is not the case for most gynaecological cases with normal anatomy, but may be necessary in cases of grossly distorted anatomy or the extensive debulking procedures that are more commonly undertaken for ovarian malignancy. Examples of appropriate investigations are: *

anaesthesia-related investigations – full blood count (FBC), urea and electrolytes (U&Es), group and save (and cross-match if there is a signiﬁcant chance of >1 l of blood loss), possibly respiratory tests;

*

chest X-ray (CXR), if the patient has suspected heart or lung disease; ECG, if the patient has cardiovascular disease;

*

liver function tests (LFTs) and CXR for all oncology patients;

*

blood glucose, if abnormal glucose tolerance; pregnancy test, in all women of childbearing age.

*

*

Bowel preparation There is some debate among colorectal surgeons as to whether preoperative bowel preparation prevents complications. Certainly, inadequate bowel preparation increases the amount

71

Chapter 5: Surgery for gynaecological cancers

of liquid stool in the gut and makes the spread of organisms more likely. Polyethylene glycolelectrolyte lavage solutions and sodium phosphate are the most commonly used preparations and are considered safe. Meta-analyses of studies on the use of bowel preparation have not found a decrease in complications-related bowel surgery except perhaps when the distal part of the large bowel is resected. Although there is an increasing trend towards primary anastomosis, avoiding ileostomy or colostomy, this will be unavoidable for some patients and preoperative counselling, preparation and marking up by the stoma nurse is sensible.

Drainage of pleural eﬀusion The preoperative drainage of a malignant/symptomatic pleural eﬀusion can improve lung function and increase the respiratory reserve in some patients. This should be undertaken in consultation with anaesthetic advice.

Antibiotic prophylaxis The female genital tract cannot be sterilized and therefore the chance of infection secondary to surgery is increased. Most surgical site infections are caused by organisms such as Escherichia coli, Streptococcus, Staphylococcus aureus, Enterococcus and anaerobes such as Bacteroides. Antibiotic prophylaxis with a single dose of intravenous co-amoxiclav at the time of anaesthesia reduces the risk of infection at hysterectomy from 21% to 9%. An alternative regimen includes metronidazole and a cephalosporin, but the latter’s association with the subsequent risk of Clostridium diﬃcile has resulted in this being abandoned in most units.

Thromboprophylaxis Complex pelvic surgery is a risk factor for developing thromboembolism. There is an exponential increase in risk with age and a sevenfold increased risk in someone with malignancy. Additional risk factors include hormone replacement therapy (HRT), the combined oral contraceptive and high-dose progestogens. All hospitals should have a thromboprophylaxis protocol with input from surgeons, haematologists and nursing staﬀ. An example would be to give low molecular weight heparin subcutaneously, daily, starting at 6 pm the day before the operation together with ﬁtted thigh-length compression stockings. Other regimens include restricting the heparin to post-surgery and using intermittent pneumatic compression stockings until mobile. Some patients present with deep vein thrombosis already established. In these patients the preoperative insertion of an inferior vena cava ﬁlter, which prevents embolic disease, can be a life-saving measure. These ﬁlters can usually be removed a week post-operation, but advice needs to be sought from interventional radiology.

Counselling Since the introduction of the ‘improving outcomes guidelines’ for cancer care in the UK, women have generally been introduced to many more members of the team. They may see the lead clinician at the unit as well as the unit clinical nurse specialist. They are then likely to see the consultant at the centre along with the clinical nurse specialist, and then, preoperatively on the ward they may see various members of the junior staﬀ as well. It is important that each of these members of the team is consistent in their counselling about the underlying diagnosis, need for treatment and potential complications of that treatment.

72

Chapter 5: Surgery for gynaecological cancers

A number of support organizations provide excellent information, leaﬂets and helplines for gynaecological cancer patients (see Appendix 4).

Consent Good practice principles include taking consent. Consent is required before any operation and the process of consenting a patient starts in clinic. A standardized consent form is now used throughout the UK and, besides the patient, it should be signed by a member of the team who understands the operation that is to be performed. This structured form requires the doctor to explain and complete each section of the form. It includes the name of the proposed procedure, its description, type of anaesthesia to be used and intended beneﬁts and serious or frequently occurring risks of the surgery Surgical ethics serve as a guide to help respect the autonomy of the patient, take informed consent from a competent patient and subsequently ensure conﬁdentiality. Discussion should take place in the presence of any relative, friend or other person the woman has asked to attend for support.

Skin preparation Avoid shaving if possible, or shave as close to the time of surgery as possible. Use of epilatory creams or clipping of hair is preferable.

Abdominal incisions The ability to accurately deﬁne anatomical structures is dependent on adequate visualization of the anatomical site in addition to anatomical knowledge. The choice of incision/approach can only be made once the surgical objective and the surgical procedure have been deﬁned. The proposed surgical procedure or the management of frequently encountered surgical complications must not be compromised by the incision/approach. Abdominal incisions in open gynaecological procedures are broadly divided into vertical and transverse incisions (Fig. 5.1). The ultimate choice depends on factors such as presence of abdominal scars, desired exposure, pathology expected at surgery and the risk of wound dehiscence. The following points should be borne in mind while making the choice of incision. * *

Abdominal incisions should be based on anatomical principles. They must allow adequate access to the abdomen.

Paramedian incision 3

Fig. 5.1 Abdominal incisions used in gynaecological oncology surgery.

4 Midline incision 5 Modified Gibson incision

Maylard incision 2 Pfannenstiel incision 1

73

Chapter 5: Surgery for gynaecological cancers

*

They should be capable of being extended if required.

*

Ideally muscle ﬁbres should be split rather than cut.

*

Nerves should not be divided. The rectus abdominis has a segmental nerve supply, and it can be cut transversely without weakening a denervated segment.

*

Above the umbilicus, tendinous intersections prevent retraction of the rectus muscle.

*

The most commonly used incision in gynaecological oncology surgery is the midline incision. This incision is ﬂexible, allowing access to all areas of the peritoneal cavity with the possibility of extending into the upper abdomen if necessary. However, it is also the weakest incision and is more prone to dehiscence, infection and hernia formation. In previous decades, these complications were considered important, and therefore the paramedian incision was used by many surgeons. However, with the modern suture materials the paramedian incision is now unnecessary. A recent large review demonstrated that mass closure of the midline incision with slowly absorbable monoﬁlament sutures minimizes the risks of complications. The length of suture should be four times the length of the wound, with the sutures placed at least 1 cm from the fascial edge and 1 cm apart. The abdominal wall regains its strength slowly, with 50–60% of its original strength being regained by six weeks. At nine months post-surgery the abdominal wall will still be recovering, but it will never regain more than 93% of its original tensile strength. Other incisions include the suprapubic transverse or Pfannenstiel incision, which is commonly used in benign gynaecology and for caesarean sections. This is a strong incision and heals well within the skin creases, as it follows the dermatomes. However, it oﬀers restricted access to the pelvis and abdomen. The Maylard incision, in which the muscle and inferior epigastric vessels are divided, provides better access. Within gynaecological oncology, the operations that used to be performed through a Pfannenstiel incision are now usually performed laparoscopically, making this incision relatively unusual in modern practice. The modiﬁed Gibson incision is useful for an extraperitoneal approach to the pelvic side wall, especially for lymph node dissection.

Intraoperative complications Good surgical technique based on anatomical knowledge, together with experience, can prevent many but not all intraoperative complications. In many complex procedures there is a deﬁning moment where the surgeon has to be conﬁdent enough to perform a manoeuvre, knowing that if they get it right the operation will be much easier to complete but if it is wrong then an unintended visceral injury may arise.

Bowel The large or small bowel may be injured during surgery within the abdominal cavity. The small bowel is usually injured because of adhesions from previous infection or surgery. The bowel may be adherent to the anterior abdominal wall and cause problems during the incision. However, the small-bowel wall has a characteristic colour and muscular appearance, which can usually be recognized, so that despite serosal injury the bowel lumen is not opened. Deep in the pelvis and where inﬂammation coexists this distinction can be more diﬃcult. If a single injury occurs to an otherwise normal loop of small bowel then closure with continuous or interrupted dissolvable sutures is desirable. Care must be taken to ensure the lumen is not

74

Chapter 5: Surgery for gynaecological cancers

compromised; this requires sutures perpendicular to the course of the bowel. For damaged segments, resection and direct anastomosis may be more satisfactory. The large bowel is more commonly damaged during an omentectomy for a large omental tumour mass or in the pelvis posterior to the uterus and cervix because of adhesions or tumour. As the large bowel is full of pathogens, healing of the injured large bowel can be less successful than that of the small bowel. The consequences are more devastating as faecal peritonitis can be fatal, especially if the diagnosis is delayed. Although some large-bowel damage can be sutured directly, a colorectal surgeon may consider a defunctioning colostomy or ileostomy to allow healing to take place over three to four months. These can usually be reversed once the suture line has sealed.

Urinary tract The bladder is often adherent to the lower segment of the uterus after previous caesarean section, and sharp dissection and diathermy are useful techniques to separate this muscular structure. It is often necessary to strip the entire pelvic peritoneum oﬀ the bladder in women with tumour deposits in ovarian cancers. These tumour deposits distort the surface of the peritoneum but rarely do they actually invade through it and involve the bladder wall. A damaged bladder is usually straightforward to repair especially if the damage is away from the ureteric oriﬁces. Closure with a continuous absorbable suture in two layers is optimal. If the defect is close to the ureteric oriﬁces it may be necessary to stent the ureters ﬁrst. Continuous drainage with a Foley catheter for 10 days is suﬃcient time for the bladder wall to repair, and many urologists suggest taking a cystogram prior to removal of the catheter to conﬁrm healing. Long-term function is rarely aﬀected by a bladder repair. Ureteric damage should not occur during routine hysterectomy with normal anatomy. However, the ureter does run close to the ovarian vessels, and then underneath the uterine artery midway along its length, and if the anatomy is distorted because of ﬁbrosis, adhesions or tumour, the ureter can be endangered. There are occasions where the ureter may be obstructed by tumour and ﬁrmly adherent to a mass that needs removal. In addition, complete dissection of the ureters is necessary at radical hysterectomy, risking devascularization. Ureteric stents can be placed cystoscopically to help identify the ureter during surgery and may have been placed preoperatively to drain an obstructed renal tract. When stents have been present for some time preoperatively, the ureteric wall becomes quite enlarged secondary to an inﬂammatory reaction. If a ureter is partially transected, the insertion of a stent and repair with 3.0 or 4.0 absorbable suture is usually suﬃcient. The prophylactic insertion of a stent where devascularization may have occurred can sometimes prevent further problems later. Where the ureter is completely transected a full repair is necessary. If this is close to the bladder, reimplantation of the ureter directly into the bladder is usually the optimal procedure. It may be necessary to mobilize the bladder to allow the ureter to reach without tension, and subsequent ﬁxing of the bladder to the psoas muscle (psoas hitch) or refashioning the bladder shape (Boari ﬂap) may be necessary. The end of the ureter should be spatulated in the bladder to prevent stenosis at the point of its insertion. If the ureter is transected higher in the pelvis it may be necessary to re-anastomose it end to end. Again, spatulation of both ends to ensure the join is not stenosed is considered the optimal technique and suturing with 3.0 or 4.0 interrupted absorbable sutures is usually suﬃcient to hold the ends together over a stent. It is important this anastomosis is not under tension.

75

Chapter 5: Surgery for gynaecological cancers

Haemorrhage Haemorrhage can occur from any of the major vessels in the pelvis. Direct haemorrhage can occur from the pedicles as a result of poor surgical technique or inadequate knot tying. In an inﬂamed pelvis, the tissue is often friable and when tying knots it is important not to pull on the suture or the vessel may avulse. Damage to the larger vessels can occur because of closely adherent tumour or lymph node. Arterial bleeding is usually easily identiﬁed and the vessel occluded and tied oﬀ or repaired. Venous bleeding can be more diﬃcult to identify, and with delicate vessel walls can be more diﬃcult to occlude with a tie. Direct damage to a large vein can occur at lymphadenectomy, especially in the para-aortic region, causing small defects in the inferior vena cava. Direct suturing with 4.0 (or ﬁner) monoﬁlament non-absorbable sutures is necessary; care should be taken not to put any tension on the vessel while tying the knot. Occasionally if the defect is greater than 2–3 mm in width it may be necessary to position a vascular clamp across the defect to allow visualization and direct suturing, and each suture must passes through the full thickness of the vessel wall. Various haemostatic products are available to help with the oozing that may occur around a repair to a large vein where further insertion of sutures may actually cause more bleeding than it controls.

Minimal access surgery Minimal access surgical techniques have transformed surgical practice. Technology has improved such that excellent visualization of organs can occur and operations can be performed with the patients making a more rapid return to normal. Advantages of minimal access surgical techniques are: * * * *

faster recovery; decreased pain, decreased analgesic requirements; improved visualization for the surgeon; smaller scars.

Not everyone is convinced of the value of minimal access surgery. However, surgeons who have learned to perform minimal access surgery, along with the ward nurses who care for these patients, are convinced of its value. There is increasing evidence that laparoscopic hysterectomy is the gold standard treatment for endometrial cancer. Many studies have shown faster recovery with laparoscopic hysterectomy when compared with abdominal hysterectomy. Serious complication rates are low for both, but are slightly higher for laparoscopic hysterectomy. However, studies have invariably compared the experience of a surgeon who has performed hundreds of abdominal hysterectomies with that of a surgeon who has performed only a handful of laparoscopic ones. Within gynaecological oncology, minimal access surgery would be unacceptable if it did not have the same cure rates as open surgery. After initial concerns that port site metastases were increased, the data suggest no compromise in eﬃcacy. However, more studies are needed to convince some surgeons. Equipment required for minimal access surgery is more expensive than that required for open surgery. Many quality instruments are disposable and costs to the theatre department are increased. Most of the money saved by the techniques (e.g. early return to work) are not fed back into the hospital system. Routine gynaecological surgery is very forgiving of poor surgical technique whereas minimal access surgery is not. The techniques required involve

76

Chapter 5: Surgery for gynaecological cancers

practice and a surgical workload that allows for learning at all levels. Anatomical knowledge and the ability to perceive a three-dimensional picture using the two-dimensional images on the screen are essential. Laparoscopy is a technique widely used within all of gynaecological practice. Open (Hasson) or closed (Verres needle) insertion of the pneumoperitoneum can be used in women with normal body mass index (BMI). The open Hasson method is safer in those with high or low BMI. An alternative in these women is to use Palmer’s point for entry. In general, the rates of complications associated with the insertion of the primary trochars are 0.4/1000 for bowel injury and 0.2/ 1000 for large vessel injury. Minimal access surgery utilizes various energy sources for haemostasis and complications related to the use of diathermy can occur. However, these were more common in the earlier days, when surgeons did not fully understand the principles of diathermy, and have now reduced. If everyone realizes the entire length of the instruments is potentially able to conduct electricity and therefore cause tissue damage, most complications would be avoided. Minimal access techniques can be used for either diagnostic or therapeutic purposes.

Diagnostic laparoscopy Minimal access surgery has improved the outcome for many women who present with an ovarian mass of uncertain malignant potential. Although imaging techniques and tumour markers can give a risk estimation of malignancy, the ﬁnal histopathological examination is the only method of being certain. If a laparotomy is required to remove the ovarian mass, there has been a tendency to advise that full pelvic clearance is sensible. After all, recovery from a laparotomy is anyway prolonged, and a second laparotomy soon afterwards if the mass was found to be malignant makes recovery even more prolonged. On the other hand, use of minimal access surgery will allow safe removal of an ovarian mass with rapid recovery, so that only those women who will truly beneﬁt from a pelvic clearance or other staging surgery will need to undergo that procedure. Laparoscopy can be used for taking biopsies in women with known advanced ovarian malignancy, to help identify the correct chemotherapy required. It may also have a use in identifying patients in whom full debulking surgery is possible. Second-look laparoscopy has fallen out of favour as a way of routinely assessing the success of chemotherapy in ovarian cancer. However, laparoscopic staging with omentectomy and pelvic and para-aortic lymphadenectomy, in women who have had an unexpected malignancy diagnosed, can be a useful way of determining if or what type of chemotherapy is required.

Therapeutic laparoscopy Hysterectomy for endometrial cancer requires careful inspection of the peritoneal cavity, washings to be taken, the ovaries removed and possibly removal of lymph nodes. This type of surgery is ideally performed laparoscopically with minimal morbidity. The relative improvement in outcome with a laparoscopic hysterectomy is more marked in obese patients; these patients also have a higher risk of developing endometrial cancer in the ﬁrst place. The only factor preventing this technique being used more widely is a relative lack of surgeons appropriately trained in it; however, this is rapidly changing.

Radical hysterectomy Radical hysterectomy can be performed laparoscopically with decreased hospital stay and smaller wounds. The procedure is more complex than a simple hysterectomy, and with the

77

Chapter 5: Surgery for gynaecological cancers

decrease in incidence of cervical cancer it is being performed less frequently than in the past. These factors have resulted in relatively few practitioners gaining enough experience in this technique in the UK. Those working in less-developed countries where the incidence of cervical cancer is higher are leading the world in the development of this particular technique.

Radical trachelectomy Radical removal of the cervix can be performed either vaginally or abdominally. Removal of the lymph nodes, which is an essential part of this therapeutic option, can be done laparoscopically or extraperitoneally. In expert hands, the entire dissection for the trachelectomy can also be done laparoscopically.

Summary This chapter has outlined the basic principles of gynaecological oncology surgery, many of which are taken for granted. These are the same as the underlying principles of most therapeutic options. The surgeon should have good background knowledge, they should be clear about the aims of the planned surgery and should assess the risk of potential complications prior to the procedure.

Further reading

78

*

Ceydeli A, Rusckinski J, Wise L. Finding the best abdominal closure: an evidence based review of the literature. Cur Surg 2005: 62; 220–5.

*

Kirk RM. Basic Surgical Techniques. London: Elsevier Health Sciences, 2002

*

Antibacterial prophylaxis in surgery: 2 – Urogenital, obstetric and gynaecological surgery. Drug Ther Bull 2004; 42: 9–13.

*

Royal College of Obstetricians and Gynaecologists.Preventing Entry-related Gynaecological Laparoscopic Injuries. Green top guideline 49. London: Royal College of Obstetricians and Gynaecologists, May 2008.

*

Shawski S, Wexner SD. Oral colorectal cleansing preparation in adults. Drugs 2008; 68: 417–37.

6

Chapter

Radiotherapy for gynaecological cancers Charlotte Coles

Introduction Cancer is common and aﬀects one in three people. Radiotherapy is the use of high-energy X-rays to treat cancer, and detailed modelling suggests that 52% of all cancer patients should receive radiotherapy at some point in their pathway. It is estimated that of those cured, radiotherapy either alone or in combination with other treatments contributes to that cure in 40% of cases. This chapter will give an overview of radiotherapy principles and practice, with reference to the treatment of gynaecological cancers.

Indications for radiotherapy Radiotherapy is essentially a local treatment. It has a role across the spectrum of gynaecological cancers (Table 6.1) and includes radical (primary curative treatment), adjuvant (supplement to primary curative) and palliative (symptom relief) treatments (see Chapter 15 for deﬁnitions). The aim of radical radiotherapy is to eradicate macroscopic disease and minimize late toxicity. The treatment of bulky or locally advanced cervix cancer is an example of radical radiotherapy treatment. This treatment is usually provided in combination with concomitant cisplatin chemotherapy (chemoradiotherapy), which acts as a radiosensitizer. The aim of adjuvant radiotherapy is to eradicate microscopic disease and minimize late toxicity. Pelvic irradiation following surgery for endometrial cancer is an example of adjuvant radiotherapy. Palliative radiotherapy may be indicated in all gynaecological cancers. It is more eﬀective in relieving some symptoms than others (see Box 6.1), and should be selected for use accordingly.

Types of radiotherapy Two types of radiotherapy are mainly used for the treatment of gynaecological cancers.

External-beam radiotherapy External-beam radiotherapy (EBRT) is otherwise known as teletherapy, which is Greek for ‘long’. This means that the radiation is delivered at a distance from the patient, most commonly using a linear accelerator or ‘linac’ (see Fig. 6.1). The linear accelerator produces X-rays (otherwise known as photons), by accelerating electrons through a waveguide into a

Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 6: Radiotherapy for gynaecological cancers

metal target, such as a tungsten alloy. These X-rays are then collimated (shaped) to produce treatment beams. Linacs typically operate in the range of 4–25 megavolts (MV) to produce X-rays that are commonly used to treat deep-seated tumours, such as cervix cancers. Superﬁcial and orthovoltage X-ray treatment units (Fig. 6.2) have the same design as diagnostic X-ray machines, and operate in the range of 50–500 kilovolts (kV). They are used to treat more superﬁcial lesions such as skin, e.g. vulval cancer, and bone lesions, e.g. rib metastases. Figure 6.3 shows how the dose delivered is related to the voltage of the X-rays.

Box 6.1 Indications for palliative radiotherapy Local indications: *

Bone pain

*

Pressure – nerve/cord compression; soft tissue

*

Bleeding

Not: *

Discharge

*

Shortness of breath

*

General indications, e.g. weight loss

Table 6.1 Summary of treatment for gynaecological cancers

Radical

Adjuvant

Palliative

Ovary

S

C

C, RT, S

Endometrium

S (RT)

RT

RT, C

Cervix

S, RT

RT

RT, C

Vulva

S, RT

RT

RT, C

S, surgery; C, chemotherapy; RT, (chemo)radiotherapy.

Fig. 6.1 A linear accelerator. The patient is positioned on a treatment couch and the gantry of the ‘linac’ moves around the patient to deliver the radiation to the tumour site.

80

Chapter 6: Radiotherapy for gynaecological cancers

Fig. 6.2 A superﬁcial/orthovoltage X-ray machine. The X-rays are produced as accelerating electrons across a vacuum into a metal target. The X-rays are collimated (shaped) by using an applicator that is placed against the patient’s skin.

Fig. 6.3 X-ray depth and dose characteristics. Megavoltage X-rays are used to treat deep tumours and do not treat the skin surface to 100% of the dose, a characteristic known as ‘skin sparing’. In contrast, superﬁcial/ orthovoltage X-rays do not spare the skin and are used to treat superﬁcial lesions.

Percentage of dose remaining

100

16 MV

6 MV 50 250 kV

5

10

15

20

Depth (cm)

81

Chapter 6: Radiotherapy for gynaecological cancers

Fig. 6.4 Diagrammatic illustration of a typical brachytherapy ring applicator: the tube is inserted into the uterus and two ovoids are positioned in the vaginal fornices.

The process for EBRT requires the following steps: 1. Immobilization – for pelvic EBRT the patient is positioned supine with knees ﬂexed and lower legs slightly apart, using specially designed immobilization supports. The aim is to make the patient comfortable in a reproducible position for daily treatment. The patient is also given small, permanent tattoos on the pelvic skin to ensure accurate set-up, using treatment room lasers just before each treatment. 2. Computed tomography (CT) simulation – the patient has a CT radiotherapy planning scan in the treatment position. 3. Contouring – the clinician outlines the target volume which he or she wishes to treat, such as the cervix plus tumour, uterus and pelvic nodes in the case of cervix cancer. The clinician may also wish to contour surrounding normal tissue (organs at risk), to which he or she wishes to limit the dose. 4. Treatment planning – the planner, usually a physicist or radiographer, uses the dose targets and normal tissue constraints established by the clinician to create a radiotherapy plan (see beam arrangement later). 5. Treatment – after the plan is approved by the clinician, the patient is set up using the room lasers to make the tattoos, and treated with EBRT. 6. Treatment veriﬁcation – X-ray (portal) images are taken at the beginning of treatment and at set times throughout radiotherapy. These are compared with the images acquired at CT simulation, and the patient’s position is corrected if necessary.

Brachytherapy The term ‘brachy’ is derived from the Greek word for ‘short’, meaning short-distance therapy, where the radioactive source is placed inside or next to the region requiring treatment. The most common type of gynaecological brachytherapy (BT) is intracavitary treatment, in which the radioactive source is placed inside a body cavity. For radical cervix BT, a typical applicator used is the tandem and ring applicator, which consists of an intrauterine tube inserted into the uterus and a ring applicator, which is placed against the cervix (see Fig. 6.4). A vaginal applicator is used to treat the vault post-hysterectomy in women with cervix or endometrial cancer. Occasionally, interstitial BT may be used whereby hollow needles are placed directly into the tumour. Interstitial BT can also be used to boost treatment for persisting parametrial disease and facilitate dose modulation in patients with cervix cancer following EBRT.

82

Chapter 6: Radiotherapy for gynaecological cancers

Fig. 6.5 The typical pear-shaped dose distribution for the treatment of cervix cancer. The pear shape is ‘ﬂattened’ in the coronal plane to minimize dose to the bladder and bowel.

Today most gynaecological BT treatments involve high-dose rate (HDR) BT, which uses the radioisotope iridium-192. Iridium-192 has a half-life of 73.83 days and decays by emitting beta particles (electrons) and gamma radiation (photons). The iridium source is driven along the applicators on the end of a wire by an ‘after-loading’ machine while the patient is isolated in the treatment room. This ensures that staﬀ are protected from the radiation. Intracavitary HDR BT is a key component of chemoradiotherapy for cervix cancer, and will be used to illustrate the concepts. Following EBRT, the patient is examined under anaesthesia to assess response to initial treatment and to insert BT applicators. The patient has a CT planning scan with the ring applicator in situ, and the clinician outlines the target for treatment (cervix), to ensure adequate coverage. The clinician also identiﬁes the ‘organs at risk’, such as the rectum, urinary bladder and sigmoid colon, with the aim of minimizing the dose to these normal tissues. A BT plan is then created by adjusting the dwell times and positions of the iridium source within the applicator, to build up the required dose distribution. A typical dose distribution is the ‘ﬂattened pear’, which allows adequate dose to the cervix while reducing dose anteriorly and posteriorly, thus limiting radiation to the bladder and rectum, respectively (Fig. 6.5). The radioisotopes emit the radiation dose according to the inverse square law (Fig. 6.6). Thus, the dose is greatly reduced with only small distances from the source. This physical property is advantageous for dose distribution, as it can create excellent coverage to a small tumour boost volume with far less dose to surrounding normal tissue when compared with EBRT.

Principles of radiotherapy Mechanism of action Radiotherapy works by damaging the deoxyribonucleic acid (DNA) of cells. The X-ray (photon) is the commonest form of radiotherapy, but other particles such as electrons, protons and ions can be used. Radiotherapy can cause damage by directly ionizing the DNA atoms, or indirectly by reacting with water to form free radicals. This indirect eﬀect is commonest with photons. Radiotherapy can clearly damage both tumour cells and healthy normal tissue. However, the tumour cells have a reduced capacity to repair the DNA damage

83

Chapter 6: Radiotherapy for gynaecological cancers

Fig. 6.6 Graph illustrating how the radiation dose using brachytherapy (BT) decreases with distance from the radioactive source according to the inverse square law. The ‘A’ dose is referring to the dose prescription point for cervix BT: point A. This is classically deﬁned as 2 cm lateral to the central uterine canal and superior to the lateral fornix in the plane of the uterus. In practice, it is taken to be 2 cm superior and lateral to the upper surface of the ring applicator at the internal os.

Dose profile at level of point ‘A’ Percentage of ‘A’ dose 300

250

200

150

100

50

0 0

0.5

1

1.5 2 2.5 3 3.5 Distance from line source (cm)

4

4.5

5

compared with normal cells. This damage is inherited through cell division, accumulating damage to the cancer cells, which ultimately results in cell death. The most eﬀective form of radiation-induced DNA damage is the double-strand break, as it is the hardest to repair.

Dose and fractionation The gray (Gy) is the SI unit (international system of units) of absorbed radiation dose. One Gy is the absorption of 1 joule of radiation energy by 1 kg of matter. For radical radiotherapy, doses in excess of 60 Gy are required as the aim is to eradicate macroscopic tumour. For adjuvant radiotherapy, lower doses in the region of 50 Gy are used to eradicate microscopic disease. A palliative radiation dose is not intended to eradicate tumour, but to reduce symptoms. It may alleviate the pain of tumour pressure by partially shrinking the cancer, or reduce bleeding by damaging tumour vasculature. The total radiation dose is fractionated (divided into episodes of treatment delivery spread out over a period time) to allow recovery of normal tissues. The commonest dose per fraction is 2 Gy for radical treatments. Doses of less than 2 Gy can be given, with the aim of minimizing the risk of late permanent normal tissue side eﬀects (see side eﬀects of radiotherapy and therapeutic ratio later). For example, pelvic EBRT is often delivered in 1.8 Gy fractions, ﬁve times a week over a total of ﬁve weeks (total of 45 Gy). In contrast, palliative radiotherapy is hypofractionated, i.e. given in larger but far fewer fractions. The rationale is that the palliative eﬀect of radiotherapy can be gained quickly with

84

Chapter 6: Radiotherapy for gynaecological cancers

minimal disruption for the patient. Late normal tissue side eﬀects are less of an issue, as the total dose is far less than with radical treatment, and the patient’s life expectancy is often short. An example of a hypofractionated palliative treatment is a single 8-Gy treatment to a bone metastasis. Rapidly proliferating cancers such as squamous cell carcinomas of the cervix display the characteristic of accelerated repopulation during radiotherapy. Typically the cancer cells start to divide more rapidly after three to four weeks following the commencement of radiotherapy, often assisted by tumour reoxygenation and re-assortment within the cell cycle. This phenomenon has been shown to result in a detrimental eﬀect on local control if the overall treatment time is prolonged due to unplanned interruptions during radiotherapy. To combat this eﬀect, radiotherapy can be hyperfractionated, i.e. two treatments can be given on the same day to compensate for missed treatments.

Normal tissue tolerance Normal tissues vary in their ability to tolerate radiotherapy. This partly depends on the structure of the organ, for example, the lung can tolerate a relatively high dose to a small volume, as the remaining lung will compensate. In contrast, a relatively high dose to a small volume of spinal cord can cause severe neurological signs. In pelvic radiotherapy, typical normal tissue tolerances to the small bowel and rectum are 50 Gy and 60 Gy, respectively.

Side eﬀects of radiotherapy The side eﬀects of radiotherapy have traditionally been classiﬁed as early and late side eﬀects, and are summarized in Box 6.2.

Box 6.2 Summary of side eﬀects of radiotherapy Acute (early): *

Occur during or shortly after treatment

*

Only occur in tissues within the treatment volume

*

Tissues aﬀected: Fast-proliferating “itis”: dermatitis; mucositis; cystitis; enteritis; proctitis

*

Hair

*

Bone marrow

Late: *

Occur after six months of stopping radiotherapy

*

Pathogenesis:

*

Fibrosis – strictures

*

Vascular damage: Small-vessel dilatation (telangiectasia); bleeding; small-vessel constriction (ischaemia + necrosis leading to perforation + ﬁstula)

*

Infertility

*

Second malignancies

85

Chapter 6: Radiotherapy for gynaecological cancers

(a)

(b)

Fig. 6.7 (a) An acute vulval skin reaction illustrating conﬂuent moist desquamation. (b) Several weeks later, this has healed completely.

Acute side eﬀects Acute side eﬀects usually occur during treatment or shortly after completion and are usually reversible. They are therefore considered less relevant for limiting radiotherapy dose. Typically, acute reactions are the result of inﬂammation (“itis”) in rapidly proliferating tissues and, depending on the part of the body that is irradiated, result in dermatitis, mucositis, cystitis, proctitis, etc. Other examples of acute radiotherapy reactions include hair loss and bone marrow suppression, if such tissues are within the radiotherapy ﬁeld. These early-reacting tissues tend to be relatively insensitive to changes in the radiation dose per fraction. Typical acute side eﬀects of pelvic radiotherapy are diarrhoea, dysuria and urinary frequency. Radiotherapy to the vulva usually produces moist desquamation, which heals within a few weeks following completion of treatment (Fig. 6.7).

Late side eﬀects Late-reacting tissues manifest radiation damage typically six months after completion of treatment; the damage can progress over several years. These late side eﬀects can be permanent, and therefore provide the basis for dose constraints to limit normal tissue toxicity. Late eﬀects typically occur in more slowly proliferating tissues. The pathogenesis includes ﬁbrosis and atrophy, which are the result of speciﬁc responses of ﬁbrocytes to irradiation. Fibrosis represents a proliferative response of surviving ﬁbrocytes to growth

86

Chapter 6: Radiotherapy for gynaecological cancers

Fig. 6.8 Example of a late radiation reaction after pelvic radiotherapy – pubic fractures due to bone necrosis.

factors released by injury, and atrophy reﬂects both loss of ﬁbrocytes and collagen reabsorption. Examples of this process in gynaecological radiotherapy include shrinkage and instability of the irradiated bladder, and strictures and malabsorption of irradiated small bowel. Vascular damage is another important pathological process involved in late normal tissue toxicity. It can be in the form of small-vessel dilatation, which can manifest as telangiectasia in the skin, such as in vulva radiotherapy, or bleeding, such as haematuria in the irradiated bladder. In addition, vascular damage can cause small-vessel constriction, which may manifest as ischaemia and necrosis (Fig. 6.8). Clinical examples of this process include bowel perforation and ﬁstulae. Other important late normal tissue side eﬀects include hormone deﬁciencies, infertility and second malignancies. In contrast with acute side eﬀects, the eﬀects of late toxicity are not limited to the tissues within the radiation ﬁeld, but can be more widespread due to damage to the blood supply of more distant organs. Another diﬀerence is that the late-reacting tissues tend to be more sensitive to changes in the radiation dose per fraction than acute-reacting tissues.

The therapeutic ratio The therapeutic ratio (also known as therapeutic index) is the ratio of the beneﬁt produced by a treatment compared with the toxicity resulting from the treatment. The therapeutic ratio changes with treatment intent. For example, for radical radiotherapy, where the intention is cure, a greater risk of side eﬀects is generally accepted. However, fewer side eﬀects may be acceptable for adjuvant radiotherapy, as it is only possible that the patient has subclinical disease. Clearly, palliative radiotherapy aims to alleviate symptoms, so any treatment-related side eﬀects must be kept to the minimum. In radiotherapy practice, the skill is to manipulate the therapeutic ratio to maximize beneﬁt and minimize toxicity. This may be achieved by exploiting the radiobiological diﬀerences between tumour and normal tissue, and minimizing the volume of normal tissue irradiated.

Exploiting radiobiological diﬀerences As previously discussed, most normal tissues that are susceptible to late radiotherapyinduced side eﬀects are more sensitive to the fraction size than the tumour. This means that by reducing the radiotherapy dose per fraction, late toxicity is minimized but tumour

87

Chapter 6: Radiotherapy for gynaecological cancers

(a)

(b)

Fig. 6.9 (a) The colourwash isodose of a two-ﬁeld beam arrangement for pelvic radiotherapy. The yellow and orange regions represent 95% and >107% of the prescribed dose (isodoses), respectively, and the rectum and bladder are located within this high-dose region. (b) The four-ﬁeld plan shows the 95% isodose covering the pelvis, but the rectum and bladder are likely to be in the 50% isodose region (green). Fig. 6.10 The yellow line is delineating the radiotherapy ‘ﬁeld’ to treat the pelvis and para-aortic nodes. The red regions will be shielded with blocks or MLCs to spare the normal tissue, such as kidneys. Thus the treated region becomes the smaller green ‘spade’.

control is maintained. An example is pelvic radiotherapy for cervical cancer, which is given in 1.8 Gy per fraction (see section on dose and fractionation above).

Minimizing volume of normal tissue The volume of normal tissue can be minimized in the following ways. *

*

88

Beam arrangement – simple radiotherapy treatments use just one to two treatment beams. This is usually adequate for palliative treatment, but may result in larger volume treated with a less homogeneous dose. By using three to four ﬁelds, the volume of normal tissue irradiated to a high dose can be reduced considerably and the dose distribution improved (Fig. 6.9). Shielding – normal tissue can be shielded from the treatment ﬁeld by using lead blocks, or more recently, multileaf collimators (MLCs). These are typically 1 cm-wide lead ‘leaves’ positioned in the head of the linear accelerator, which can be moved into position to shield part of the beam (Fig. 6.10).

Chapter 6: Radiotherapy for gynaecological cancers

Conventional plan

‘2D’ CT plan

3D conformal

Fig. 6.11 Diagrammatic illustration of the reduction in volume of normal tissue irradiated with increasing complexity of radiotherapy. Conventional square-shaped treatment ﬁelds result in a cuboid volume. A two-dimensional (2D) plan produced by shaping the treatment ﬁelds with blocks or MLCs creates a smaller cylindrical volume. Three-dimensional (3D) conformal radiotherapy reduces the volume treated still further, by creating a spherical treatment volume.

(a)

(b)

Fig. 6.12 (a) A radiotherapy planning computed tomography (CT) scan showing the target (red) wrapping around the rectum (green). (b) The intensity-modulated radiotherapy (IMRT) beam arrangement showing the colourwash isodose. Regions of high and low dose are shown in red and blue, respectively. The high-dose region is concave and spares the rectum posteriorly.

*

Conformal therapy – three-dimensional conformal radiotherapy (3DCRT) enables each radiation beam to be shaped to ﬁt the proﬁle of the target from a beam’s eye view, using MLCs and a variable number of beams. This will reduce the volume of normal tissue irradiated (Fig. 6.11).

Intensity-modulated radiotherapy (IMRT) – this is high-precision radiation that is the next generation of 3DCRT. It has the ability to treat concave target volumes while sparing normal tissue (Fig. 6.12). This is achieved by using many beams positioned around the patient, each with a number of intensity levels or ‘beamlets’. Although it greatly reduces high dose to organs at risk, the volume of normal tissue irradiated to a much lower dose may be increased. This may have implications for the development of second malignancies in the future.

89

7

Chapter

Systemic therapies for gynaecological cancers Helena M. Earl

Introduction Chemotherapy consists of many diﬀerent drug classes with a wide variety of diﬀerent modes of action. In gynaecological malignancies a wide range of agents are used, in the adjuvant, neoadjuvant and metastatic disease settings. Endocrine manipulation/treatments are also used in gynaecological cancers when tumours are found to be oestrogen or progesterone receptor (ER/PR) positive. In this chapter, the main chemotherapeutic agents used in gynaecological malignancies are outlined. The principles and philosophies of use of systemic therapies in the adjuvant, neoadjuvant and metastatic disease settings are examined. The use of chemotherapy in epithelial ovarian cancer (EOC) and in endometrial cancer is summarized. When referring to chemotherapy treatment of EOC, we will assume that fallopian tube and primary peritoneal cancer are managed and treated in the same way. As theories of the fallopian tube as the primary site for most serous papillary carcinomas gather increasing credibility, this strategy seems increasingly appropriate. Systemic treatment of cervical cancer, germ cell ovarian tumours, sarcomas of the uterus (leiomyosarcomas and endometrial stromal sarcomas) and malignant mixed Müllerian tumours of the uterus and ovary will be covered in more detail in other, site-speciﬁc, chapters.

Types of chemotherapy and modes of action Platinum compounds Cisplatin and carboplatin are highly eﬀective agents in gynaecological cancers and have been used for many years, having a role in EOC, endometrial cancers and germ cell tumours of the ovary. These agents act by causing DNA adducts, and as such are not cycle speciﬁc, having some activity in non-cycling cells. Essentially they are DNA-damaging agents, which make it diﬃcult for the cells to subsequently divide, and can induce apoptosis – ‘programmed cell death’. Cisplatin is the parent compound, and carboplatin was developed subsequently. The advantage of carboplatin over cisplatin is that it produces a lower incidence of severe side eﬀects (nausea and vomiting, renal impairment, fatigue, high-tone hearing loss, peripheral neuropathy and anorexia), all of which are common with cisplatin. Carboplatin, however, causes more myelosuppression, particularly anaemia and thrombocytopenia. Neither drug causes hair loss, and both are almost exclusively renally excreted. Carboplatin doses are calculated on a formula based on glomerular ﬁltration rate (GFR) called the Calvert formula, Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 7: Systemic therapies for gynaecological cancers

after the oncologist who was instrumental in developing the drug and introducing it into routine clinical practice. Carboplatin is individualized for patients by GFR and an AUC (area under the concentration-time curve) calculation. All other chemotherapy drugs are individualized for patients on a body surface area calculation (BSA), which is dependent on height and weight. Often EOC remains sensitive to carboplatin in relapse and, when this drug is used for multiple successive courses, major allergic reactions develop with increasing frequency. In our own practice in Cambridge, all but one patient developing an allergic reaction to carboplatin have been treated successfully by switching to cisplatin, with premedication with corticosteroids and histamine receptor-1 and -2 blockers. Germ cell tumours of the ovary have a high cure rate ( >90%), and cisplatin is the preferred platinum compound. This is because patients are on average much younger and can tolerate the side eﬀects better. In addition, in testicular germ cell tumours, randomized trials have demonstrated that cisplatin is better than carboplatin in terms of cure rate. In older women with EOC and endometrial cancer, carboplatin is the preferred drug because it is better tolerated. Peripheral neuropathy with cisplatin is age and cumulative dose related, and unfortunately may progress even once the course of treatment has been completed. Since peripheral neuropathy from cisplatin does not usually improve, in our practice we limit the cumulative dose to 450 mg/m2, to avoid serious peripheral neuropathy.

Taxanes Taxanes are used in combination with platinums in EOC and endometrial cancers, and in combination with gemcitabine in relapsed leiomyosarcomas of the uterus. These agents deregulate tubulin synthesis by promoting the assembly of microtubules and stabilizing their formation by inhibiting depolymerization. These microtubules are extremely stable and therefore non-functional. This in turn inhibits the proper functioning of the mitotic spindle and therefore prevents cellular division. Other microtubular functions that are inhibited have to do with modulation of motility and cellular attachment and intracellular transport. In terms of side eﬀects, in addition to the common problems of nausea, vomiting, hair loss, myelosuppression and fatigue, taxanes cause peripheral neuropathy. This is because taxanes also inhibit β-tubulin, which is a neuro-transport protein in peripheral nerve axons. Although damage to peripheral nerves is said to be temporary, and certainly does improve after treatment is completed, most patients who experience damage will have residual loss of feeling in ﬁngers and toes, which is permanent. About 15% of patients experience acute arthralgias and myalgias a few days after treatment with taxanes, which last for a few days. Some patients will be very sensitive to the eﬀects of taxanes on the peripheral and even central nervous system, and will be unable to tolerate the treatment. Whether this intolerance is due to single nucleotide polymorphisms (SNPs) in β-tubulin, or SNPs in drug-metabolizing enzymes, which make some patients highly sensitive to the eﬀects of these drugs, is not known. In the early days of use of paclitaxel there was a high incidence of immediate allergic reactions to the drug, which was due to the cremophor required for formulation of this essentially insoluble compound for intravenous use. Paclitaxel requires high-dose corticosteroids for 24 hours to prevent this. Both paclitaxel and docetaxel can cause major immediate allergic reactions, including anaphylaxis. Although there are now desensitization programmes, these are prolonged and may not be successful. Some patients who are allergic to paclitaxel have been treated with docetaxel successfully, but for the majority developing an allergic reaction to one of the two taxanes, there will be cross-reactivity.

92

Chapter 7: Systemic therapies for gynaecological cancers

Topo-isomerase inhibitors Topo-isomerases broadly speaking maintain the topological integrity of DNA during the duplication of complementary DNA strands, which is an essential part of mitosis. In order for the tumour cell to divide (and any other cell for that matter), the DNA has to unwind, form two new complementary strands at the ‘replication fork’ and then rewind as two new doublestranded DNA sequences. This is a highly complex process and topo-isomerase I starts the process by cleaving DNA, binding covalently to the 3′ end of the broken DNA and taking the form of a ‘cleavable complex’. This allows the process of DNA replication to begin. Topoisomerase I inhibitors (camptothecin, topotecan and irinotecan), interfere with this early phase of DNA replication, and cause DNA fragmentation by stabilizing the ‘cleavable complexes’. Topo-isomerase II works at the ﬁnal stage of this process to complete the re-ligation repair. Topo-isomerase II inhibitors (the epidophyllotoxins – etoposide and teniposide; and the anthracyclines – doxorubicin and epirubicin) subvert the enzyme into a DNA-damaging agent by stabilizing the DNA – topoisomerase II complex, and causing cessation of cell cycle progression, DNA strand breaks and cell death. In gynaecological cancers etoposide and topotecan are both used in relapsed EOC, and etoposide is used at high dose intravenously in the curative treatment of ovarian germ cell tumours. Anthracyclines (doxorubicin and epirubicin) are used in relapsed EOC as part of combination treatment, and also in endometrial cancer and leiomyosarcoma of the uterus.

Oral etoposide Oral etoposide use (twice daily dosing for 10–14 days) in relapsed EOC, although associated with reported good response rates, is fraught with diﬃculty because of unpredictability. The drug has a widely varied bioavailability, which means that diﬀerent patients can absorb anything between 30% and 100% of the ingested dose. Lower bioavailability will make the drug ineﬀective (although non-toxic), and higher bioavailability will cause extreme and prolonged toxicity. Side eﬀects include life-threatening and prolonged myelosuppression, and severe pan-gut mucositis. Etoposide albumin-binding of 94% also causes problems in the relapsed EOC population. The majority of these women will have low serum albumin levels, causing increased levels of free etoposide, which is the active cytotoxic component. Low serum albumin levels are associated with severe toxicity. Oral delivery can also be ineﬀective because many women with relapsed EOC have impending or actual subacute bowel obstruction, and have poor absorption of oral drugs. Therefore despite early promise, oral etoposide is not widely used and should only be used with caution in relapsed EOC.

Anthracyclines There is evidence from meta-analysis of randomized trials in EOC that anthracyclines contribute to response rates in advanced EOC. They are used in combination with other drugs, most notably a combination of epirubicin with cisplatin and oral capecitabine – the ECX regimen, which has one of the highest response rates in relapsed EOC. Single-agent doxorubicin is used as liposomal doxorubicin – a preparation that encapsulates the active drug in liposomes. This converts the drug into a ‘slow-release’ preparation, which increases its single-agent activity, and changes the proﬁle of side eﬀects. Doxorubicin causes myelosuppression, hair loss and mouth ulcers, whereas liposomal doxorubicin can cause palmar plantar erythema (PPE), generalized oral mucositis and low-grade alopecia. Sometimes there

93

Chapter 7: Systemic therapies for gynaecological cancers

is a severe and generalized rash relating to pressure areas, and severe ‘pan-gut’ mucositis, with a very sore mouth and grade 3/4 diarrhoea. Severe reactions may require hospitalization for supportive care and discontinuation of the drug.

Anti-metabolites Anti-metabolite drugs work by inhibiting essential biosynthetic processes, or by being incorporated into DNA and RNA and inhibiting their normal function. The drugs in this group used in gynaecological cancers include 5-ﬂuorouracil (5-FU) and capecitabine, and gemcitabine.

5-Fluorouracil/capecitabine 5-FU/capecitabine’s mechanism of cytotoxicity is through misincorporation of ﬂuoronucleotides into RNA and DNA, and also because of the inhibition of the nucleotide synthetic enzyme, thymidylate synthetase (TS). Oral capecitabine works in the same way but is a prodrug requiring metabolism to its active form.

Gemcitabine Gemcitabine is an anti-cancer nucleoside that is an analogue of deoxycytidine. It is a prodrug, and after transport into cells is phosphorylated by deoxycytidine kinase to its active form. Gemcitabine di- and triphosphate inhibit processes required for DNA synthesis. Incorporation of the active drug into DNA is the major mechanism by which gemcitabine causes cell death. After incorporation of gemcitabine nucleotide on the end of the elongating DNA strand, one more deoxynucleotide is added but after that the DNA polymerases are unable to proceed. Once attached to DNA it is impossible to remove the active form of the drug. In addition, gemcitabine metabolites exert inhibitory eﬀects on cellular regulatory processes, which further inhibits cell growth. 5-FU and capecitabine are used in combination with epirubicin and cisplatin in relapsed EOC (in ECX). It may prove particularly useful in the future in mucinous histotype EOC. Gemcitabine is used in EOC in combination with carboplatin, and in leiomyosarcoma of the uterus in combination with docetaxel.

Neoadjuvant/adjuvant treatment in epithelial ovarian cancer The eﬀect of neoadjuvant chemotherapy in EOC can be evaluated by clinical examination and radiographic studies (clinical and radiological response), by falling serum cancerassociated antigen CA125 (biomarker response) and by interval debulking surgery (histopathological response). Primary surgery followed by platinum-based chemotherapy remains the preferred initial management for women with stage III or IV EOC. However, primary surgery may not be feasible in all patients because of disease bulk and patient performance status and, in one report, maximal cytoreduction was achieved in fewer than half of patients. In addition, many women present with massive ascites or poor performance status that may place them at high risk for perioperative morbidity and even mortality. An accepted alternative is to give neoadjuvant chemotherapy, and then interval debulking surgery can be carried out in responding patients. One advantage of this approach in women with chemoresistant disease is that aggressive surgery can be avoided completely. Chemotherapy before surgery may also increase the proportion of women for whom optimal debulking surgery is possible. Furthermore, less tumour burden after chemotherapy may

94

Chapter 7: Systemic therapies for gynaecological cancers

make the surgery less diﬃcult and may be associated with fewer complications. These beneﬁts are highly relevant in women with poor performance status. Several uncontrolled series document the feasibility of initial chemotherapy followed by cytoreductive surgery, and some suggest that outcomes are at least comparable with those in women who have undergone initial surgery followed by chemotherapy. Furthermore, in one study, patients with extra-abdominal disease who received neoadjuvant chemotherapy with carboplatin and paclitaxel rather than initial debulking surgery had a signiﬁcantly longer median overall survival (31 versus 20 months). Some retrospective data suggest that survival might be inferior using this approach. A systematic review included 26 uncontrolled studies of women with stage III or IV EOC who received neoadjuvant chemotherapy with a platinum drug in lieu of an attempt at initial cytoreductive surgery. Although all women subsequently underwent laparotomy for interval cytoreduction, the ability to achieve optimal cytoreduction at this procedure was highly variable (24–85%). In ten of the studies, neoadjuvant chemotherapy with interval cytoreduction was associated with survival rates that were inferior to those of a historical control group treated with initial cytoreductive surgery followed by chemotherapy. An earlier metaanalysis by the same investigators, which included 22 studies exploring neoadjuvant chemotherapy, noted that a higher number of preoperative chemotherapy cycles correlated negatively with survival: each additional cycle of preoperative chemotherapy was associated with an incremental decrease in median survival time of 4.1 months. However, it must be remembered that all these reports are retrospective and therefore there is a high degree of selection bias. For example, it is fairly obvious that patients receiving more neoadjuvant chemotherapy before surgery is attempted are likely to have a worse prognosis than others, because of relative resistance to chemotherapy. In the only randomized trial to address the impact of neoadjuvant chemotherapy on surgical debulking and survival, 128 women with stage IIIC or IV (pleural eﬀusion only) previously untreated EOC were randomly assigned to upfront debulking surgery, followed by six cycles of carboplatin and paclitaxel versus three cycles of the same chemotherapy followed by debulking surgery and three additional chemotherapy courses. In a preliminary report of 93 evaluable women who had completed all treatment, those receiving neoadjuvant chemotherapy had a signiﬁcantly higher optimum debulking rate, less blood loss during surgery and fewer postoperative infections. At a median follow-up of 41 months, the diﬀerence in median overall survival (42 versus 29 months for upfront surgery and neoadjuvant chemotherapy, respectively) and median disease-free survival (20 versus 25 months, respectively) was not statistically signiﬁcant. The optimal setting for neoadjuvant chemotherapy is unknown and two ongoing trials are addressing this question. The European Organisation for Research and Treatment of Cancer (EORTC) is conducting a randomized trial (EORTC 55971) of neoadjuvant chemotherapy versus primary cytoreductive surgery. The UK CHemotherapy OR Upfront Surgery (CHORUS) trial is also addressing the same question with randomization to initial surgical debulking or neoadjuvant chemotherapy. From the perspective of non-surgical gynaecologic oncology, the use of prolonged and hazardous primary cytoreductive surgery in bulky stage III and IV EOC is in some senses a historical anomaly. Other metastatic ‘solid’ cancers are not treated by primary surgery to ‘de-bulk’ metastatic disease, before chemotherapy. Given that ovarian cancer is also more sensitive to conventional chemotherapy than other epithelial tumours (e.g. lung and gastrointestinal tract), the role of neoadjuvant chemotherapy in bulky stage III/IV ovarian cancer is

95

Chapter 7: Systemic therapies for gynaecological cancers

even more compelling. The primary neoadjuvant treatment of late-stage ovarian cancer also oﬀers valuable opportunities to establish predictive molecular biomarkers of response to established and novel therapies, through translational research. Response indicators include clinical indicators, CA125 and imaging. There are many new exciting imaging modalities, including dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and ﬂuoroL-thymidine positron-emission tomography (FLT-PET), which oﬀer the opportunity of early response indicators to conventional and novel treatments.

Epithelial ovarian cancer and ﬁrst-line chemotherapy Cisplatin gained a deﬁnite place in the primary treatment of ovarian cancer some years ago. More recently single-agent carboplatin has been shown to be equivalent to cisplatin/cyclophosphamide/doxorubicin (Adriamycin) (CAP) in the International Collaborative Ovarian Neoplasm (ICON) 2 study but is much less toxic.

Role of paclitaxel ﬁrst line with carboplatin Two large trials, Gynecologic Oncology Group (GOG) 111 (410 patients) and the International European and Canadian Study OV10 (680 patients), have shown median survival advantages of 14 and 10 months, respectively (highly statistically signiﬁcant) for the addition of paclitaxel to platinum as ﬁrst-line therapy. GOG 111 entry was restricted to women with high-risk stage III and IV disease who had been suboptimally debulked. OV10 did include some high-risk stage II patients, but neither study included the early-stage lower-risk patients who were part of ICON 3. If the analysis of outcome in ICON 3 is restricted to stage III and IV patients with > 2 cm residual bulk disease (i.e. a similar patient group to GOG 111), then there emerges a trend in favour of the paclitaxel/carboplatin combination (although not statistically signiﬁcant). However, combined with GOG 111 and OV10 results, ICON 3 will contribute evidence for the beneﬁt of paclitaxel with carboplatin ﬁrst line in these poor-risk women. The ICON Group reported the ICON 3 study. This study compared paclitaxel plus carboplatin versus standard chemotherapy with either single-agent carboplatin or CAP, in women with ovarian cancer. This study raises some interesting and important questions for the ovarian cancer research community in the UK, and perhaps most importantly for women with advanced ovarian cancer. This study did not conﬁrm an advantage for ovarian cancer patients from the addition of paclitaxel. We feel it is important for readers to have a full understanding of the place of these therapies in the treatment of women with advanced and relapsed ovarian cancer and therefore will take some time to discuss this result further. The GOG 111 study and the OV10 demonstrated a beneﬁt from using ﬁrst-line platinum with paclitaxel versus platinum combinations without paclitaxel. These studies are now in long-term follow-up, and the advantage at 10 years appears to be of the order of 10%. If this is sustained, a 10% improvement in cure rate will be the largest improvement in treatment of advanced EOC to emerge in the past two decades. The Scottish Randomized Trial in Ovarian Cancer (SCOTROC) study demonstrated equivalent eﬃcacy for docetaxel to paclitaxel in combination with carboplatin, and docetaxel showed less peripheral neuropathy. However, docetaxel shows more myelosuppression than paclitaxel, and thus paclitaxel remains the taxane of choice in EOC. In Cambridge we completed a phase I/II randomized translational study, in which we utilized delayed surgery in women with stage IIIC/IV disease, to give monotherapy (either carboplatin or paclitaxel) prior to surgery. This study looked at prospective molecular proﬁling and candidate

96

Chapter 7: Systemic therapies for gynaecological cancers

gene expression as predictors of response and resistance to carboplatin and paclitaxel given as monotherapy. Window studies such as this will increase our ability to individualize therapy by deﬁning robust, predictive molecular proﬁles of response to diﬀerent therapies. Our study led to the interesting and novel observation that expression of transforming growth factor (TGF) β1 stabilized microtubules and thus made EOCs more sensitive to paclitaxel. This interesting observation requires conﬁrmation in larger translational research studies.

Duration of treatment The duration of ﬁrst-line chemotherapy is usually six cycles, although if the patient has residual disease and is continuing to respond, therapy may be extended to nine cycles. In second- and third-line therapy we would usually limit therapy to six cycles. When there is no prospect of cure, therapy should be given for as short a time as possible to induce a useful remission.

Dose intensiﬁcation It is unproven whether dose intensity is more important than total dose delivered in the treatment of ovarian cancer. It must always be borne in mind that for the majority of women with advanced ovarian disease, the treatment is palliative, making quality of life an important issue. There is an ongoing European study of high-dose chemotherapy in women who have been maximally debulked, and have no residual disease at the completion of primary therapy. There are few data as yet on the eﬀectiveness of this in ovarian cancer and it is not a strategy that has proved eﬀective in other common solid tumours, e.g. breast cancer.

Philosophy of treatment of relapsed epithelial ovarian cancer/palliative chemotherapy of gynaecological cancers When a woman experiences EOC relapse there is no prospect of cure and the aims of treatment change to palliation of symptoms and prolongation of life. When cancers remain sensitive to chemotherapy, this may be the best way to palliate symptoms of disease. Some women with ovarian cancer may survive for a long period of time with metastatic disease, receiving intermittent courses of chemotherapy. The management of these patients, who are incurable but remain sensitive to chemotherapy, becomes a complex balancing act. This involves trying to judge likely response rates and side eﬀects of chemotherapy in diﬀerent individuals, with diﬀerent cancers and diﬀerent attitudes to their disease and to life in general. The aims of management can be summarized in a phrase often used in our clinics: ‘We want to keep you as well as possible for as long as possible.’ This conveys to the patient the aspects of caring and ‘aiming to do our best’ that are so important, and the notion that chemotherapy is likely to prolong their life. The comment also communicates the complexity of decision making: ‘We hope to prolong your life but we don’t want you to pay too high a price in terms of toxicity.’ However, we must be clear that the only judge of what is ‘too much’ is the patient herself. Our view is that most patients (although not all) welcome the opportunity to hand over (although not completely) these diﬃcult decisions to someone they ‘trust’. This point is well illustrated by Dr Ingelﬁnger’s narrative of his own diagnosis of oesophageal cancer. As the editor of the New England Journal of Medicine he was not short of advice about how his cancer should be managed, but he longed for a kind, compassionate, experienced doctor to come along and make the best decisions with him and for him. I see one of our most important functions as oncologists involved in the ‘active’ management of ovarian cancer to lead patients

97

Chapter 7: Systemic therapies for gynaecological cancers

to make the best decisions about their active treatment. The ‘active palliative’ management (and this is not an oxymoron) of women with advanced ovarian cancer requires an essentially close relationship between the oncology team and the palliative care team in hospital practice, and the primary care-based and Macmillan-supported team at home. In best ovarian cancer management it is important that we all have a real understanding of what the ‘other’ team has to oﬀer (oncology for the palliative care team and vice versa), and this understanding will make an important diﬀerence to the lives of women with advanced ovarian cancer. Relapsed ovarian cancer often remains conﬁned to the peritoneal cavity. Recent in vitro studies growing ovarian cancer cells in culture have shown that in the environment of the peritoneal cavity (the pH of which is signiﬁcantly lower than the normal physiological pH of 7.4) ovarian cancer cells proliferate by activating growth factors and growth factor receptors. However, at normal body pH the ovarian cancer cells become quiescent. This may be one explanation for the extraordinary clinical pattern of spread of ovarian cancer, with disease remaining largely in the peritoneal and pleural cavities. Although ovarian cancer can spread into the liver, lungs, bones and brain, this happens rarely and often only very late in the disease.

Signiﬁcance of platinum-free interval Treatment planning is based around the concept of platinum sensitivity, or the platinum-free interval. Thus women whose disease progresses while on platinum therapy are classiﬁed as having platinum-refractory disease. If they experience relapse within six months of receiving platinum they have platinum-resistant disease. Patients are classed as having platinumsensitive disease if they experience relapse after six months of receiving platinum therapy. The probability of response to second-line chemotherapy following platinum-based treatments is related to the platinum-free interval. Other factors shown to have some predictive value include tumour burden and histology. Salvage monochemotherapy is generally used, but when the platinum-free interval is longer than 24 months, re-treatment with platinum compounds and/or taxanes is indicated. The response rates seen when the treatment-free interval exceeds 24 months are almost equivalent to that of primary chemotherapy. Women with platinum-refractory or platinum-resistant disease are encouraged to enter clinical trials of therapy, which often include non-platinum agents. A number of new agents have demonstrated activity in ovarian cancer. These drugs include liposomal doxorubicin (Caelyx), gemcitabine and topotecan.

Treatment beyond second line There are many, and increasing numbers of, chemotherapy treatments for ovarian cancer that can be used at relapse. As referred to above, if the treatment-free interval is >12 months then there is nearly a 50% chance that single-agent carboplatin will remain eﬀective. In women with a long treatment-free interval this is therefore the treatment of choice. Van der Burg and her group in 2002 published the results of their study from the Netherlands on the use of intensive cisplatin and etoposide therapy in relapsed ovarian cancer. The study included 170 women who were treated with induction therapy of cisplatin 50–70 mg/m2 on days 1, 8, 15, 29, 36 and 43, in conjunction with oral etoposide 50 mg daily for days 1–15 and 29–43. The women who showed a response or stable disease were then placed on maintenance therapy with oral etoposide for 21 days in 28 days for a further six to nine cycles. Weekly chemotherapy administration has been used in the treatment of other solid tumours, in an attempt to modify the toxicity proﬁle while maintaining eﬃcacy. Weekly

98

Chapter 7: Systemic therapies for gynaecological cancers

epirubicin is used in metastatic breast cancer patients who have bone marrow inﬁltration, or who are frail, as it seems to produce less myelotoxicity and is generally well tolerated. The rationale behind the Van der Burg study, however, was to intensify the dose delivery of cisplatin to induce remission and to follow this with maintenance oral chemotherapy, to keep the established remission for as long as possible. Thirty-eight patients assessed had ‘platinum-sensitive’ disease, of whom 92% showed a response, with median survival of 26 months. In the ‘platinum-intermediate’ disease group (platinum-free interval of four to 12 months) 91% of patients responded, with median survival of 16 months. Of most interest, however. was the response seen in the group termed ‘platinum-refractory’. This group included 28 patients, of whom 46% responded, although this response was short lived, with median duration of ﬁve months; median survival was 13 months. Toxicity reported in the much-publicized paper in January 2002 included myelotoxicity, although only 10% of patients required a delay in treatment of more than one week. Reported non-haematological toxicity was relatively uncommon, with nephrotoxicity in 4%, and grade II sensory neuropathy in only 7% of patients. These excellent second-line response rates and low rates of toxicity persuaded us to give this treatment to some of our patients, and indeed many of them had read about this treatment in newspapers and were keen to try it. Unfortunately, although we have seen some good radiological and serological responses with the treatment, many women have experienced severe side eﬀects, with marked reduction in general performance status during treatment, loss of appetite and taste, severe fatigue, nephrotoxicity resulting in profound and prolonged hypomagnesaemia, and grade II–III peripheral sensory neuropathy. Indeed we have had two treatment-related deaths from neutropenic sepsis. Our view is that patient selection is of prime importance when oﬀering such intensive treatment. For the majority of women with relapsed ovarian cancer this therapy is too toxic, whatever the potential gain, because, after all, it will not ultimately cure the patient.

Oral chemotherapy The oldest oral agent available to EOC patients is chlorambucil, which we still use to good eﬀect in some patients who have already had multiple lines of previous chemotherapy. The development of oral 5-FU equivalents (capecitabine and tegafur-uracil [Uftoral ]) will probably see these oral agents replace continuous intravenous infusion 5-FU, as oral administration avoids the risk of infection and thrombosis, which accompany the insertion of Hickman lines in these patients.

®

Quality-of-life assessments and patients’ expectations Optimizing quality of life is an important part of all cancer therapy, none more than the treatment of ovarian cancer. In women with recurrent or refractory ovarian cancer, the aims of treating with chemotherapy are palliation of disease-related symptoms, with improvement in quality, and to some extent length, of life. In many ovarian cancer trials much information is collected on toxicity of chemotherapy and general quality of life, using the Common Toxicity Criteria (CTC), which are the same across the globe, and the EORTC quality-of-life patientreported questionnaires for cancer patients, including the new ovarian cancer module. Although routine quality of life measures are not usually collected in patients outside clinical trials, there is nevertheless a continual analysis of ‘pain versus gain’, that is, weighing the degree of response occurring against the toxicity that the woman is experiencing.

99

Chapter 7: Systemic therapies for gynaecological cancers

Many studies evaluating palliative chemotherapy in advanced ovarian cancer have relied on surrogate measures of patient beneﬁt using biomedical responses rather than appraising palliative end-points such as quality of life and clinical beneﬁt. This issue has been addressed in a Canadian study that prospectively evaluated patient expectation, palliative outcomes of chemotherapy and resource utilization in women undergoing second- or third-line chemotherapy for recurrent or refractory advanced ovarian cancer. Quality of life was assessed using EORTC Quality of Life Questionnaire C30 (QLQ C30) and Functional Assessment of Cancer Therapy-Ovarian (FACT-O). In this study, objective responses were seen in seven of 27 women, with median survival of 11 months. Interestingly 65% of women expected that chemotherapy would make them live longer and 42% expected that it would cure them, suggesting that patient expectations from treatment are often unrealistic. However, after two cycles of treatment, improvement in quality of life was seen, especially in global function and emotional function with EORTC QLQ C30. This improvement was sustained for a median of two and three months, respectively. Therefore despite the fact that objective responses were low, active palliation with chemotherapy seems to be associated with improvements in patients’ emotional function and global quality of life.

Endometrial cancer Endometrial cancers that have the common histopathological phenotype, i.e. well or moderately diﬀerentiated, stage I/II endometrioid adenocarcinomas, do not require adjuvant chemotherapy. In women who have poorly diﬀerentiated, grade 3 tumours, in particular, those which have evidence of lymphovascular space invasion and lymph node involvement, the disease-free survival is improved when treated with adjuvant chemotherapy after deﬁnitive surgery. Adjuvant chemotherapy has consisted of platinum and anthracycline chemotherapy for a number of years, although non-surgical oncologists are increasingly using platinum and paclitaxel at the present time. Women presenting with serous papillary or clear cell histology of >10% of the primary are treated as for women with EOC – with adjuvant platinum and paclitaxel. These women have a distinctly worse prognosis than similar-stage classic ‘endometrioid’ endometrial cancer. When patients with endometrial cancer present with advanced-stage disease with local or distant metastatic disease, then the situation is incurable. However, some patients will respond well to chemotherapy and the decisions about whether to proceed with treatment will be dependent on an honest appraisal with the specialist of risks and beneﬁts of treatment. The philosophy behind palliative chemotherapy has been discussed in detail above, with regards to relapsed EOC.

Hormonal treatment in gynaecological malignancy Hormonal treatment for gynaecological malignancy has been used for endometrial cancer for many years. In principle, whenever a cancer expresses suﬃcient hormone receptors to be positive for ER and PR by immunohistochemical (IHC) tests, then theoretically there should be some advantage from using hormonal or ‘anti-hormonal’ therapy that interferes with the action of ER and PR in promoting cancer cell growth. The ‘solid’ cancers in which there is the most randomized evidence for hormonal manipulation are breast and prostate cancer. Trials of hormonal manipulation in these cancers have been conducted ever since Beatson in 1896 removed the ovaries from premenopausal women

100

Chapter 7: Systemic therapies for gynaecological cancers

with advanced breast cancer. He reported excellent responses in a number of women, and considerable prolongation of life.

Endometrial cancer and hormonal manipulation Endometrial cancer expresses high levels of both ER and PR, and has been treated with megestrol acetate or medroxyprogesterone acetate for many years. Although this treatment is eﬀective in the medium term, side eﬀects include marked weight gain, hypertension and venous thromboembolism. When pre-existing obesity, hypertension and cardiovascular disease preclude the use of progestogens, the aromatase inhibitors (AIs) are being increasingly used, although to date their use is not supported by large randomized clinical trial evidence of eﬀectiveness. Tamoxifen should not be used in endometrial cancer, as it may stimulate growth of the cancer. The drug is still widely used in breast cancer, but is one of the aetiological factors in the development of endometrial cancer, particularly when used for more than ﬁve years. Tamoxifen stimulates endometrial proliferation and causes high rates of endometrial thickening and dysfunctional/postmenopausal bleeding. There is no doubt that it stimulates the endometrium to proliferate, and is thought to act as an ‘agonist’ (rather than antagonist) on the ER-α present in the endometrium. However, in breast there is overexpression of ER-β, of which tamoxifen is a competitive antagonist. At present, hormonal manipulation has a place in the palliative treatment of advanced endometrial cancer but is not used in the adjuvant setting. This treatment can be used when the patient has co-morbidities that make surgery/radiotherapy/and chemotherapy impossible. Hormonal treatment can be used as eﬀective palliative treatment in this context.

Epithelial ovarian cancer and endocrine manipulation Recently it was established that up to 30% of serous papillary EOCs are markedly ER positive. Early trials of aromatase inhibitors (AIs) using letrozole showed a high rate of tumour response and control in women with relapsed disease. This area of EOC treatment requires further research and clinical trials to establish whether AIs will beneﬁt women with ER-positive disease when treated at an early stage. Trials are in development at present. There is clearly a wealth of randomized trials in breast cancer demonstrating the great beneﬁt from hormonal treatments for ER-positive breast cancer. Whether these very large trials need to be repeated in ER-positive gynaecological malignancies is a matter for further debate in the specialist research community.

Further reading *

*

Ahmed AA, Mills AD, Ibrahim AE, Temple J, Blenkiron C, Vias M, et al. The extracellular matrix protein TGFBI induces microtubule stabilization and sensitizes ovarian cancers to paclitaxel. Cancer Cell 2007; 12: 514–27. Bristow RE, Eisenhauer EL, Santillan A, Chi DS. Delaying the primary surgical eﬀort for advanced ovarian cancer: A systematic

review of neoadjuvant chemotherapy and interval cytoreduction. Gynecol Oncol 2007; 104: 480–90. *

Ingelﬁnger FJ. Arrogance. N Engl J Med 1980; 303: 1507–11.

*

Liu B, Earl HM, Poole CJ, Dunn J, Kerr, DJ. Etoposide protein binding in cancer patients. Cancer Chemother Pharmacol 1995; 36: 506–12.

101

8

Chapter

Palliative care in gynaecological oncology Sara Booth

Palliative care is deﬁned as ‘the active total care of patients when the disease is no longer curable and the prognosis is short’ (World Health Organization (WHO) 1967). Two of the most important concepts that this deﬁnition highlights are the active nature of palliation and its breadth, encompassing every aspect of the impact of illness, pathophysiological, psychosocial and spiritual. Palliation includes holistic assessment and instigation of a range of strategies (medical, surgical and psychosocial) with the focus on enhancing quality of life for those living with lifethreatening or advanced disease. Initial assessment is followed by frequent reassessment, adjustment of initial management strategies and continued review and revision as the disease advances. The gynaecological oncologist provides most of the palliative care that these patients require. Palliative care, when specialist interventions are not required, meshes with supportive care. It includes ensuring that patients receive those aspects of care that are additional to ‘disease management’ but essential if patients are to have the best chance of engaging with their treatment and obtaining the best outcome from it. Some diﬀerences between specialist and generalist palliative care approaches are outlined in Table 8.1. It is often helpful for patients with incurable disease to have contact with specialist palliative care services early even if the patient is asymptomatic: they are then likely to feel more conﬁdent about using them when they are less well.

Symptom control Whatever the symptom, the most eﬀective treatment will be found most quickly if a systematic approach is taken to management. The general principles of symptom management in palliative care are set out below. *

*

Make the best diagnosis possible of the cause(s) and ensure that the patient understands this. Symptoms are often multifactorial in advanced disease, and symptom control often involves incremental improvement in a number of causative factors to bring about signiﬁcant relief. A good history of the problem is essential. The patient’s wishes and her prognosis are central to decisions about appropriate investigation and treatment – it is usually not in the patient’s best interests to treat refractory malignant hypercalcaemia repeatedly with intravenous bisphosphonates or perform

Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 8: Palliative care in gynaecological oncology

Table 8.1 Outline of generalist and specialist palliative care aspects of problems typical of those occurring in women with gynaecological malignancies: the generalist palliative care problems are those which every clinician should be able to tackle

Palliative care aspects

104

Problem

Generalist

Specialist

Pain

Recognizing that the patient has a pain problem that needs treatment, and ability to take full pain history Knowledge of common pain syndromes in gynaecological malignancy Begin pain therapy using the WHO pain ladder for guidance Recognizing when the pain needs immediate specialist help or when it develops into a specialist palliative care problem, e.g. neuropathic features not responding to initial therapy

Ability to take detailed pain history and initiation of treatment on specialist pain drugs such as anticonvulsants, antidepressants, ketamine Recognizing when referral to the specialist pain team for regional pain blockade is appropriate

Distressed patient or family

Ability to listen and elicit worries, fears and concerns about illness or its wider eﬀect on the patient and their family Ability to reassure and comfort honestly, giving appropriate information when the patient desires Recognizing when specialist psychological, psychiatric, psychosexual or palliative care is needed and helping patient to accept this help.

Helping a patient who is persistently depressed or worried despite extra support Recognizing and initiating treatment for depression and/or specialist psychological treatment where available in the multidisciplinary team, or referring for specialist psychological treatment

Bowel obstruction

Skilled in predicting when partial bowel obstruction (PBO) is likely to occur and in diagnosing on clinical grounds. With seniority, ability to recognize when surgery may be a helpful option and the possibility of venting gastrostomy Initiating medical therapy for symptomatic relief of PBO and giving helpful basic dietary advice and information on prognosis

Able to diagnose PBO and instigate treatment when it has not responded to initial therapy Taking part in discussions about when surgery is possible but not appropriate or where patient or gynae-oncology team is concerned Managing refractory obstruction (including feculent vomiting, giving advice regarding venting gastrostomies, frozen pelvis) Helping with community care and diﬃcult end-of-life care

Extensive or malodorous discharge or malodorous or distressing wounds or tumours

Ability to anticipate when diﬃcult wound is likely to develop as complication of advanced disease or treatment Providing initial counselling and advice regarding treatment (e.g. diagnosis and treatment of infection)

Recognizing when specialist palliative care nursing assessment and treatment may be needed. Hospices may oﬀer frequent dressing of complex wounds for community patients, particularly for frail patients living alone

Chapter 8: Palliative care in gynaecological oncology

Table 8.1 (cont.)

Palliative care aspects Problem

Chronic haemorrhage

Generalist

Specialist

Recognizing when extra specialist help is necessary, e.g. specialist nursing, plastic surgery, radiotherapy If experienced or undertaken specialist training, may cross-over with specialist palliative care

Arranging specialist psychological help if needed for patient and for family. Hospices may also have access to specialist treatments that may help disguise/educe odours and may oﬀer symptom control admissions for particularly complex wounds and end-of-life care, when cannot be managed at home Oﬀer help with topical analgesia and management of friable or bleeding wounds

Ability to anticipate when diﬃcult bleeding may be a problem and, with seniority, knowledge and ability to use surgical options, other local treatments (e.g. radiotherapy) and supportive care, e.g. transfusions

Stepping in when local treatments and supportive care (e.g. transfusions) no longer helpful, although earlier referral important for general assessment if cause not treatable Helping with end-of-life issues

major surgery in a patient with widely disseminated disease. Nor is it right to pressurize someone into a course of action which they are unwilling to undertake, even if it is the acknowledged ‘next step’ on a care pathway or protocol. *

*

*

Document your ﬁndings with a quantitative assessment of severity to allow treatment responses to be monitored – complex regimens, built up incrementally, are often necessary to manage symptoms in advanced disease. Recognize and manage any precipitating or exacerbating factors such as fear, fever, underlying metabolic problems and other symptoms (e.g. pain exacerbating dyspnoea). Helping patients and families manage the anxiety associated with life-threatening disease is always essential and will be more successful if anxiety is acknowledged and addressed at diagnosis. Assess and reassess the eﬀectiveness of your symptom control regimen regularly – if the disease is advancing, treatment strategies may need frequent revision.

*

Ensure that you have explained and discussed your assessment and plans with the patient in detail, so that her goals and ideas guide management; if managing symptoms is diﬃcult, this becomes even more important. The patient needs to understand and agree with your reasoning and choice of the next possible line of treatment available if initial approaches fail.

*

Be prepared to re-evaluate your initial diagnoses if your management plan is not working. Be aware of the psychosocial context of the symptom – vulval pain has diﬀerent connotations and meanings from arm pain; smelly, fungating wounds have immense impact on a woman’s social existence and view of herself.

*

105

Chapter 8: Palliative care in gynaecological oncology

*

*

*

Get specialist help earlier, rather than later, when a symptom or situation is not improving – for diﬃcult symptoms or psychosocial distress contact the palliative care team. Make sure everyone involved in the patient’s care keeps communicating so that the patient and her family do not get mixed messages or confusing or contradictory pictures of what is going on. Make sure everyone (patient, family and professionals) understands and agrees the goals of care. Women may ﬁnd it more acceptable to be referred to specialist palliative care services for a diﬃcult symptom than for psychosocial support – if patients ﬁnd palliative care helpful they may be more prepared to ‘use’ them for other needs.

Pain management Pain remains the most feared symptom. It is still under-diagnosed and undertreated for many reasons, including patients’ reluctance to complain and physicians’ (and other clinicians’) failure to elicit pain and treat it systematically. It is estimated that if the WHO analgesic ladder (Fig 8.1) is used to guide therapy, about 80% of pain accompanying advanced cancer can be controlled on simple, oral, pharmacological treatment alone. To optimally control the patient’s pain, the following approach is recommended. 1. A full pain history may have to wait until the patient is comfortable. Obtain enough initial information (from general practitioner (GP), notes and family if the patient is too distressed) to be able to make a ‘good-enough’ diagnosis and start appropriate analgesics – then complete the history as soon as possible so that other necessary investigations may be carried out.

WHO recommendations for standard drugs for each step

Step 3

Standard protocol for drugs in UK

Severe PAIN or PAIN persisting or increasing

e.g. Morphine

Strong Opioid

e.g. Morphine

+Non-opioid ±Adjuvants

Step 2 PAIN persists or increases

Weak Opioid

e.g. Codeine

e.g. Tramadol

+Non-opioid ±Adjuvants

Step1 e.g. Aspirin

Pain

Non-opioid ±Adjuvants

106

e.g. NSAID or paracetamol

Fig. 8.1 The WHO analgesic ladder. NSAID, non-steroidal anti-inﬂammatory drug.

Chapter 8: Palliative care in gynaecological oncology

2. Establish the cause(s) of the pain using special investigations where needed (an increase in pain will often be due to disease progression, requiring both oncological treatment and more analgesia). 3. Assess and record the level of the pain to facilitate continuing estimation of the eﬀect of the therapeutic strategy adopted. 4. Start treatment using the WHO pain ladder at the ‘rung’ nearest the patient’s pain level, e.g. if patient in severe pain, start with strong analgesics and non-steroidal antiinﬂammatory drugs (NSAIDs) in combination. The oral route is preferable with transdermal or subcutaneous routes as ﬁrst alternatives. Morphine is still the strong opioid of ﬁrst choice, but change to transdermal fentanyl (particularly if constipation or sedation are major problems) or oxycodone, hydromorphone or buprenorphine if adverse eﬀects cannot be managed. Discuss with your palliative care service, as switching between opioids can be complex. When initiating pain control it is usual to use ‘short-acting, immediate release, morphine’ (t½ 3–4 hours) and to switch to modiﬁed-release preparations (greater convenience) when pain is controlled. 5. Reassess treatment eﬀectiveness at least daily when titrating analgesics – adjust regimen to control pain by, for example increasing the dose of your chosen drug, adding coanalgesics, changing analgesics or route where necessary. 6. Anticipate, prevent or, where necessary, treat adverse eﬀects of the analgesics prescribed. 7. If pain does not respond to standard treatment or is a recognized ‘diﬃcult pain’ – get help – from specialist palliative care. The patient with uncontrolled pain will lose conﬁdence and may develop ‘diﬃcult pain’. Intervene early to get the best pain relief. 8. Make direct contact with the GP/district nursing team when the patient is discharged to discuss diagnosis, any drugs or other treatments being used, their indications and how they need to be continued.

Diﬃcult pains Some pains can be identiﬁed as diﬃcult to treat at the time of diagnosis. If the patient knows this it will help her maintain her conﬁdence when a number of diﬀerent approaches are needed to help. Otherwise, it will be frightening if the drugs prescribed initially have only limited analgesic eﬀect. Diﬃcult pains are of several types.

Incident pain Incident pain is present only during activity and not when the patient is at rest, and can be as incapacitating as constant, severe pain. Pain control is satisfactory most of the time, but on movement (e.g. cough) the patient has overwhelmingly distressing pain, albeit short lived. Incident pain is diﬃcult to manage because if a patient is prescribed enough analgesia regularly to cover these short-lived but severe exacerbations she will have intolerable side eﬀects, as she will be relatively overdosed most of the time. The diagnosis of incident pain is a reason to refer the patient for specialist palliative care early; a pain block, surgery for a pathological fracture, radiotherapy, rapid-release analgesia such as a fentanyl lozenge or intranasal preparation may be needed for pain control.

107

Chapter 8: Palliative care in gynaecological oncology

Neuropathic pain Neuropathic pain is caused by damage to the central or peripheral nervous system and is only partially responsive to opioids and co-analgesics such as anticonvulsants and antidepressants. Neuropathic pain may be helped by additional non-pharmacological techniques such as a transcutaneous electrical nerve stimulation (TENS) or other specialist interventions such as ketamine infusions. Neuropathic pain can be identiﬁed on history and the clinical diagnosis.

Bone pain Bone pain is very common in cancer, and is often quite resistant to treatment (even with adequate doses of radiotherapy, opioids and NSAIDs). Bisphosphonates are increasingly used in the treatment of this condition and surgery may be needed if pathological fracture is threatened or complete.

Total pain Total pain was ﬁrst described by Cicely Saunders in 1967. All pain has physical, psychological, social and spiritual aspects; a neuropathic lower limb pain causing weakness will cause disability, will stop the patient working (social loss of role and income) and will cause fear and anxiety about the advance of the tumour. The pain and the illness will impact on her family, her self-esteem, her work, her ability to be a parent. It may throw up all kinds of existential spiritual questions about why she has cancer, and psychological distress becomes overwhelming. Total pain can be recognized by bearing in mind the following clues. *

Patients may seem unemotional but have behaviours suggesting distress, e.g. using the buzzer repeatedly, frequent emergency admissions with no discernible cause.

*

New treatments often work well for a short time then severe pain returns, sometimes re-doubled. Teams of clinicians may disagree about how to manage the patient’s condition, each feeling strongly that they understand what the patient needs, but diﬀerent clinicians may understand diﬀerent things from the patient.

*

It is very easy to feel ‘angry’ with patients (although diﬃcult to acknowledge), when pain does not respond to carefully thought out treatment. It is crucial to get help from specialist palliative care, as soon as you perceive that the pain is not going to be managed by ﬁrstline treatments or if there is signiﬁcant underlying psychological distress, apparent or not.

Nausea and vomiting First distinguish whether the patient has: *

nausea or vomiting

or *

nausea and vomiting.

And which is most distressing. Partial obstruction is a very common problem in gynaecological malignancy, particularly carcinoma of the ovary (avoid the term subacute obstruction). This may be caused by tumour occluding or narrowing the bowel, which may be resectable or not, depending on:

108

Chapter 8: Palliative care in gynaecological oncology

*

the patients’ wishes;

*

the stage of the illness;

*

the presence of other metastatic spread; the patient’s general condition.

*

Multiple adhesions or tumours cannot be resected. Partial obstruction is often due to dysmotility, which in turn is the result of inﬁltration of: the nerve plexuses innervating the bowel by small-volume disease (sometimes invisible on scans) leading to gut dysfunction; and the gut wall by the tumour. Features in the history indicating obstruction include: *

vomiting after meals, sometimes hours after, producing undigested food;

*

colicky abdominal pain;

*

infrequent stools or diarrhoea; borborygmi audible by the patient;

* * *

abdominal distension with gas or tumour; constipation with failure to pass ﬂatus.

The general principles of management are: *

stop all prokinetic drugs, e.g. metoclopramide;

*

use only faecal softeners; advise a low-residue diet, i.e. one without much ﬁbre;

* *

consider a surgical option if there is one area of occlusion (suggested clinically by distension; ‘stuck down’ bowel cannot distend) and if the patient is prepared to have surgery.

Palliative care clinicians rarely advise ‘drip and suck’ with nasogastric tube and intravenous ﬂuids, preferring: *

a continuous subcutaneous infusion (CSCI), via syringe driver, of an anti-emetic (e.g. cyclizine or haloperidol) with analgesia (e.g. diamorphine or morphine);

*

adding hyoscine butylbromide (60–120 mg, Buscopan) to CSCI if colicky pain is present; a separate CSCI of octreotide (or other somatostatin analogue) may be needed to reduce the volume of bowel secretions if there is insuﬃcient response to hyoscine butylbromide (indicated by continuing large-volume vomits or nasogastric aspirates).

*

Nasogastric tubes are needed when there is upper gastrointestinal obstruction. Metoclopramide may be prescribed if there is certainty that there is dysmotility and no physical blockage, but it should be stopped immediately if the patient develops colicky pain or an exacerbation of her symptoms. High-dose metoclopramide should only be used on specialist advice and by those who are experienced in this technique. Sometimes patients who recover quickly in hospital on a CSCI are sent home but then are back with obstruction within a few days. It is often better to discharge the patient on the CSCI in the medium to longer term; she will be able to take ﬂuids and light foods and this may protect her from frequent readmissions, which are demoralizing. It is much more important for the patient to be at home with well-controlled symptoms than to be on or oﬀ a CSCI. A few patients require long-term somatostatin analogues by depot injection.

109

Chapter 8: Palliative care in gynaecological oncology

Other causes of nausea and vomiting are: *

adverse eﬀects of drugs e.g. NSAIDs, opioids, antibiotics (particularly metronidazole); hypercalcaemia;

*

‘squashed stomach’ from large intra-abdominal masses;

*

* *

raised intracranial pressure; massive ascites.

This list is not exhaustive, and there may be several coexisting causes.

Anti-emetics (Box 8.1) Do not prescribe anti-emetics randomly; they act in very diﬀerent ways. Prescribe on the basis of the likely diagnoses and manage exacerbating factors such as fear, anxiety and fever. A typical prescription for an opioid-naïve patient with partial bowel obstruction would be: 5 mg haloperidol (anti-emetic), 10 mg diamorphine (if pain) and 120 mg hyoscine butylbromide (reduces abdominal secretions and colic) in 24 hours by CSCI (see Box 8.1). 1. Start with your ‘best guess’ anti-emetic, e.g. haloperidol, given subcutaneously as a prn. 2. Review after an hour to assess impact.

Box 8.1 Commonly used anti-emetics Prokinetic anti-emetic (about 50% of prescriptions) For gastritis, gastric stasis, functional bowel obstruction (peristaltic failure): metoclopramide 10 mg PO stat & q.d.s or 10 mg SC stat & 40–100 mg/24h CSCI & 10 mg p.r.n. up to q.d.s. Anti-emetic acting principally in chemoreceptor trigger zone (about 25% of prescriptions) For most chemical causes of vomiting, e.g. morphine, hypercalcaemia, renal failure: haloperidol 1.5–3.0 mg PO stat & o.n. or 2.5–5.0 mg SC stat & 2.5–10.0 mg/24h CSCI and 2.5– 5.0 mg p.r.n. up to q.d.s. Metoclopramide also has a central action. Antispasmodic and antisecretory anti-emetic If bowel colic and/or need to reduce GI secretions: hyoscine butylbromide 20 mg SC stat, 60–120 mg/24h CSCI (occasionally as high as 300 mg/24h) & 20 mg SC hourly p.r.n. Anti-emetic acting principally in the vomiting centre For raised intracranial pressure (in conjunction with dexamethasone), motion sickness and in organic bowel obstruction: cyclizine 50 mg PO stat & b.d.–t.d.s or 50 mg SC stat & 150 mg/24h CSCI, & 50 mg p.r.n. up to q.d.s. Broad-spectrum anti-emetic For organic bowel obstruction and when other anti-emetics are unsatisfactory: levomepromazine 6.0–12.5 mg PO/SC stat, o.n. & p.r.n. up to q.d.s. (Reproduced with permission from Wilcock A, Twycross R. Palliative Care Formulary, 3rd edn. Nottingham, UK: Palliativedrugs.com, 2007. Available only from www.palliativedrugs.com.) PO, per os (orally); stat, immediately; q.d.s., four times daily; SC, subcutaneously; CSCI, continuous subcutaneous infusion; p.r.n., when required.

110

Chapter 8: Palliative care in gynaecological oncology

3. If the patient is improved start a CSCI with haloperidol ± other appropriate drugs. 4. Prescribe second-line anti-emetic (with a diﬀerent mode of action) on prn side. 5. Move to or add second-line drug if ﬁrst approach ineﬀective. Avoid prescribing a prokinetic such as metoclopramide at the same time as an anticholinergic drug like cyclizine, as the latter will block the prokinetic action. Drugs that are useful in chemotherapy-induced nausea and vomiting are rarely useful in palliative care: 5-hydroxytryptamine (5-HT)3 antagonists (e.g. ondansetron) can cause severe constipation and even precipitate obstruction.

Constipation Bowel habit is very individual so it is a change in this (passage of stools uncomfortable, infrequent or incomplete) rather than an absolute frequency of stool that characterizes constipation. It is still an embarrassing subject for many people, so direct questions on bowel symptoms are very important in this group of women with pelvic disease. Severe constipation is both painful and very disabling (with patients unable to enjoy anything because their bowel problems overwhelm everything else and they spend hours trying to empty them). Constipation can mimic bowel obstruction, cause rectal and abdominal pain as well as nausea and vomiting.

Prevention *

Patients with pelvic and /or abdominal disease and those talking opioids are vulnerable to constipation, so encourage patients to remain mobile (helpful for general health and preventing deep vein thromboses as well) and to keep up an adequate ﬂuid intake.

*

Start laxatives when starting opioids: the dose needed to keep the bowel habit comfortable (which may be diﬀerent from the patient’s previous habit). This is very individual, but patients on opioids needs a bowel stimulant (e.g. senna) and a stool softener (such as Milpar or docusate). Co-danthrusate is a combination of softener and stimulant in one preparation but is only available for short-term use or for patients with life-limiting illness.

*

Mitigate the eﬀects of a number of risk factors: for example if a patient who is taking a 5-HT3 antagonist receives a drug with an anticholinergic eﬀect (e.g. amitriptyline or cyclizine) and an opioid, she will be very constipated, possibly developing partial obstruction, particularly if she is immobile and dehydrated. Think about drug combinations; other drugs which can precipitate constipation are carboplatin, iron, some antacids and calcium channel blockers.

*

Many patients dislike opening their bowels on a commode in a ward; it is better to take them to the lavatory in a wheelchair. Anticipate (to prevent it) constipation in all patients with neurological problems aﬀecting the pelvis, e.g. spinal cord compression, cauda equine syndrome, lumbosacral plexopathy, autonomic neuropathy.

*

*

*

Local pain will inhibit bowel activity, so actively prevent constipation in those with perineal wounds or scars, anal ﬁssures, haemorrhoids or other incidental bowel conditions. Treat metabolic and electrolyte abnormalities such as hypercalcaemia.

111

Chapter 8: Palliative care in gynaecological oncology

Constipation • History • Physical Examination

Complete: Stop Laxatives Consider Bowel Obstruction

Fig. 8.2 Treatment options for constipation PR, per rectal examination.

Incomplete: Softeners Yes Surgery?

No

PR

No

Hard

Glycerol suppositories and PO Laxative

Soft

Bisacodyl suppositories and PO laxative

Full Rectum

Yes

See above

No

PO Laxative

Bowel Obstruction

X-ray

Full Colon

No, reconsider diagnosis

Yes

PO Laxative

Diagnosis Diagnosis is made on the basis of the history – special investigations are rarely necessary but if faecal impaction develops patients may have faecal leakage or spurious diarrhoea (leaking from above the faeces impacting the rectum), and a plain abdominal ﬁlm will reveal faecal loading.

Treatment All the preventive measures outlined above should be continued or instituted and need to be continued once the immediate problem is solved (Fig 8.2). Faecal impaction requires oral pharmacological treatment to prevent continuation of the problem, Immediate rectal measures to manage an established problem include: *

glycerol suppositories (softening, also stimulate local peristaltic action): retention enemas used overnight such as arachis oil (contraindicated in peanut allergy);

*

stimulant suppositories such as bisacodyl; these have a more marked peristaltic action; manual removal of faeces (under sedation and analgesia) and/or high enemas are rarely needed but may be used after specialist advice.

*

Oral treatment If routine laxative treatment (e.g. co-danthrusate) is not working, consider Movicol (a macrogol), one to three sachets daily until bowel starts to work. Patients with life-limiting illness may remain on Movicol in the longer term, provided electrolytes are monitored regularly.

Hypercalcaemia Once diagnosed, this is simply treated but when it occurs at the end of life or becomes refractory it has a very poor prognosis and it may be inappropriate to use intravenous rehydration and bisphosphonate therapy.

112

Chapter 8: Palliative care in gynaecological oncology

Diagnosis Although hypercalcaemia is not particularly common in gynaecology malignancies, it should be suspected in any patient with new-onset, refractory or progressive: *

nausea;

*

vomiting; fatigue;

*

*

constipation; confusion;

*

sedation.

*

Have a high index of suspicion in those with bone metastases. Hypercalcaemia is simply diagnosed by measuring serum calcium; ‘corrected serum calcium’ is related to the level of serum albumin and is measured routinely by most pathology laboratories in the UK.

Treatment The treatment of choice is intravenous bisphosphonate therapy and intravenous rehydration, the dose given being dependent on the calcium level. Symptomatic hypercalcaemia (particularly with levels of Ca2+ above 3.0 mmol/l) requires very active rehydration with normal saline and potassium. The most extensively studied and widely used bisphosphonate is pamidronate. Third-generation bisphosphonates include zoledronate and ibandronate, which are more potent and need to be given less frequently. There is good evidence from studies in patients with breast cancer and myeloma that bisphosphonates reduce skeletal complications of malignancy and bone pain, and there is emerging evidence that they have an antitumour eﬀect of their own.

Cautions *

Reduce the dose of bisphosphonate (see manufacturer’s recommendations and seek advice) in patients with renal impairment or those on nephrotoxic drugs.

*

Do not treat hypercalcaemia automatically, if the patient is dying; it is unlikely to be the best palliative treatment in this situation.

Complex, fungating and malodorous wounds Complex, fungating and malodorous wounds occur when cancer erodes through the skin or other epithelial layer, thus preventing healing. The odour is produced by infection with anaerobic organisms. This problem can be devastating for a woman and her family and is often diﬃcult to manage. It is important to make every eﬀort to mitigate the eﬀects of an unhealed, particularly a smelly, wound, as the impact on the patient’s psychological state and her relationships with friends and family can overwhelm the last few months of her life with shame and fear. In gynaecological oncology the wound may be deep, inaccessible and aﬀecting the sexual organs or structures associated with excretion. The problem may seem so personal that the patient may even conceal its extent until it becomes obvious to everyone because of the smell. The general principles of management are listed below.

113

Chapter 8: Palliative care in gynaecological oncology

*

*

*

*

*

*

*

*

Support the woman, ﬁnd out her goals for care, approach the question of smell in a matter of fact way; being ‘very sympathetic’ in an emotional way without oﬀering practical help may make her feel that the problem is even more obvious than it seems to her already. Use imaging to investigate the extent of the wound; biopsy may be necessary to distinguish between tumour, radiation necrosis or infection. Local treatment for cancer, such as radiotherapy, may be needed to palliate the wound (reducing exudate or bleeding) or slow disease progression. If infection is present treat it after making the best attempt possible to isolate the organism; topical metronidazole may be helpful. Careful wound toilet is essential (irrigation and desloughing) – remember the patient may need pain relief to do this successfully. Use opioids before dressing changes or seek specialist advice. Short-acting opioids (such as fentanyl lozenges) which act as rapidly as intravenous opioids may be helpful. Frequent dressings (several times a day) are required and patients dread them (inhibiting those carrying out wound toilet) if they are very painful. Seek expert help in wound assessment/dressing – most hospitals now have specialist nurses in tissue viability and wound healing, a plastic surgery opinion may be helpful or consultation with an interventional radiologist where there is frequent haemorrhage. A huge range of dressings is now available with constantly evolving regimens – charcoal dressings absorb odour, stoma bags can be used to contain exudates, tampons can be used to soak up exudates from internal wounds. If the patient is at home but needs frequent changing of or complex dressings this may be possible at a hospice, thereby building trust for end-of-life care.

Breathlessness Breathlessness, an uncomfortable and distressing awareness of the need to breathe, is one of the most common symptoms of advanced cancer, even when intrathoracic malignancy is absent – its prevalence increases as cancer advances. It is usually multifactorial and reversal of possible precipitating causes in a patient with advanced disease rarely gives helpful results. Intractable breathlessness develops late in the cancer and is associated with a poor prognosis. A systematic approach using pharmacological and non-pharmacological treatments (Table 8.2) is essential, with frequent reviews of treatment. Ensure that reversible conditions are adequately and appropriately (for the stage of illness) treated, for example:

114

*

Pleural eﬀusions: treat these deﬁnitively early on rather than repeatedly draining them so that loculation occurs. A thoracic surgery opinion may be invaluable – obtain this early before multiple thoracenteses.

*

Pulmonary embolus: all patients with cancer, particularly pelvic disease, are at increased risk of thromboembolic disease; initiate prevention and treat problems early where indicated.

*

Weight loss, cachexia and deconditioning: advise patients to stay physically active as far as possible from the time of diagnosis to reduce deconditioning; ensure they are eating well and do not let weight loss become cachexia. Early dietetic intervention is essential. Exercise also helps reduce the incidence of fatigue and may help reduce depression.

Chapter 8: Palliative care in gynaecological oncology

Table 8.2 Management of intractable breathlessness Non-pharmacological treatments, suitable for all breathless patients

Facial cooling with a hand-held fan (aim at area supplied by the ﬁfth cranial nerve around the mouth, nose and surrounding cheeks) Keeping physically active, exercise programmes for suitable patients, tailored to that person Breathing retraining Relaxation and anxiety control techniques Psychological support for patient and carer (very frightening to watch someone being breathless) Addressing fears about breathlessness Education about the symptom

Pharmacological treatment: more important in the severely breathless

Palliative treatment: oral opioids reduce breathlessness by about 20% in many patients, start low and increase slowly to reduce incidence of adverse eﬀects. More useful in those patients with severe breathlessness Opioid-naïve patients: 2.5 mg oral morphine 4-hourly prn Already taking opioids: increase current dose by 25% and encourage use of prn opioids for breathlessness At the end of life: prescribe CSCI of morphine/diamorphine and midazolam. If opioid-naïve, typical regimen would be: 2.5–5.0 mg diamorphine with 5–10 mg midazolam over 24 hours. If already on opioids add midazolam (5–10 mg) or levomepromazine (6.25–12.5 mg, more if anxiety very prominent) Oral benzodiazepines (e.g. lorazepam) or phenothiazines May be helpful if patients cannot use anxiety-reducing strategies Oxygen: no evidence that oxygen is better than a fan in chronic, intractable breathlessness in advanced disease, unless patient desaturates below 90% on exercise

Entering the palliative phase When patients enter the stage where everyone acknowledges cure is no longer possible, even if they have months, or possibly years, to live it is important to ﬁnd out the sorts of choices they would like to make about their future care. Research shows that patients expect healthcare professionals to start discussions about treatment options. Introducing community palliative care services early (perhaps by an initial consultation with the hospital support team) may help women feel they are more in control of their contact with these services and not feel they are associated with certain death but with improving quality of life. Younger women may be embarking on chemotherapy knowing they have a small chance of cure (e.g. for advanced clear cell carcinoma of the ovary). If the tumour does not respond to treatment, you will need to raise the possibility of treatment not working (gently) and ﬁnding out what she would want if this were to happen. Would she like to continue intensive treatment, however small the chance of a cure, or would her main focus be improving quality of life? What is her experience of treatment so far? Focusing on excellent symptom control is likely to extend rather than shorten life, and introducing the concept of palliation makes discussions about stopping chemotherapy or surgery or where the woman would like to be cared for when she is dying a gradually evolving one rather than a sudden occurrence, which

115

Chapter 8: Palliative care in gynaecological oncology

frightens her and shuts the conversation down. The patient may then die in hospital when she would have preferred to be at home or having treatment that the oncologist does not think will help. Support communication in the family by seeing the patient and her ‘signiﬁcant other’ together (with her consent), particularly when the illness enters a palliative or end-oflife phase, putting the emphasis on quality of life rather than disease management.

Anxiety and depression A supportive approach to cancer management from the time of diagnosis will both help to reduce unnecessary anxiety and ensure that contacts with the hospital will not exacerbate the fear inevitable with the diagnosis. If left unaddressed, one anxiety may compound others and patients can spiral into a high sense of arousal: it is important to build and support the natural resilience that most women possess rather than unwittingly exacerbating fear. Most patients stress that a lack of communication about their disease and its treatment is one of the major diﬃculties they face. The circumstances around diagnosis (before you have met the patient) can inﬂuence a woman’s attitude to treatment and it is important to understand these: delays in diagnosis (especially common with carcinoma of the ovary) can lead to anger and distress. A diagnosis of gynaecological malignancy may be made very suddenly, e.g. at ultrasound for an apparently minor problem or investigation for rapidly developing ascites. Remember as you hear the history that someone’s life and expectations have changed over a short space of time and they need time and a supportive, easily accessible oncology team to replenish their personal resources and ﬁnd a new balance.

Practical issues When you take a history, as well as accumulating the necessary facts, demonstrate empathy with the feelings in the story and acknowledge what the patient is saying. For example, the patient may say. ‘I kept going back to the doctor but they could not ﬁnd anything wrong’. So you might say: * * *

‘That must have been frustrating for you.’ ‘I imagine you became concerned about this.’ ‘It was probably initially reassuring for you, but then it must have got a bit frightening.’

These are all empathetic statements and, depending on the patient’s response, will help you form an idea of their mental state without letting feelings escalate out of control. A major cause of anxiety for patients is not being heard; this can be helped by summarizing what you hear as you take the history. For example: *

‘Can I just recap: you have had some nausea throughout your treatment in spite of a number of diﬀerent drugs being tried.’

*

‘Metoclopramide has been the most helpful drug.’

*

‘Milky foods and fatty food smells seem to make your nausea worse.’

Ideally, when it feels that a clinical situation is escalating out of control try to make a space to listen, rather than talk to, the patient (plus anyone else they choose for support) to reach an understanding of what she feels are the major problems. Before such a meeting in clinic or on the ward:

116

Chapter 8: Palliative care in gynaecological oncology

*

*

make sure you are as up to date as possible on what has been going on (and the multidisciplinary team’s decision on treatment is clear), read the notes and talk to the GP beforehand so you understand any longstanding psychosocial issues; start with an open-ended question such as ‘Can you tell me what the worst thing for you at the moment is?’ or ‘What is the most diﬃcult thing you are having to contend with at the moment?’ or ‘What most needs our attention at the moment?’; then be quiet and let the patient tell you how things are for them. The diﬃculties which emerge may not be ones that you are qualiﬁed to manage (e.g. diﬃcult pain, family strife and ﬁnancial problems), but the patient will have been heard and you can start referrals or discussions which may help.

Do not forget to screen for depression at diﬀerent points in the disease course; about 45% of patients with cancer have signiﬁcant anxiety or depression at some point in the illness. The most common points are at diagnosis and when metastatic disease is ﬁrst diagnosed. * *

Look out for low mood, self-blame and someone who always looks sad, and burdened. Ask ‘What gives you pleasure?’; if nothing does the patient is likely to have anhedonia and may well be clinically depressed.

Depression is treatable and anxiety can be helped: talk to the GP, the liaison psychiatry team and the palliative care service to organize a treatment plan. Medication, social support and psychological treatments (as well as exercise and activity) can all improve psychological problems.

Key points *

Palliative care is a multidisciplinary approach to improving quality of life in patients with cancer, particularly those with life-limiting disease.

*

It addresses the medical, psychological, social and spiritual aspects of illness with equal vigour.

*

Symptom control is an important part of helping patients with gynaecological malignancy achieve the best outcomes. Psychological status is becoming recognised as a predictor of recovery from physical illness.

* *

* *

Gynaecological malignancy presents particularly complex psychosocial and psychosexual issues for patients: these need to be addressed actively. Specialist palliative care services are there to help: telephone advice is always available. Hospice Directory. A directory of all palliative care services and palliative care organizations can be obtained from St Christopher’s Hospice Information Service on 0207 7789252 for the cost of the stamps. Please do get a copy of this for your unit or ward.

Further reading *

Booth S, Bruera E, Tate T. Palliative Care Consultations in Gynae-oncology. Oxford: Oxford University Press, 2004.

*

Stannard C, Booth S. Churchill’s Pocketbook of Pain, 2nd edn. Edinburgh: Churchill Livingstone, 2004.

*

Twycross, R. Wilcock, A. Palliative Care Formulary, 3rd edn. Nottingham:

Palliativedrugs.com, 2007. Available at: www.palliativedrugs.com. *

Walker LG, Walker MB, Sharp DM. The organisation of psychosocial support within palliative Care. In: Lloyd Williams, M (ed.) Psychosocial Issues in Palliative Care. Oxford: Oxford University Press, 2003.

117

9

Chapter

Ovarian cancer Robin Crawford and Yin-Ling Woo

Introduction Ovarian cancer, 90% of which is epithelial in origin, is the ﬁfth commonest cancer in women, with over 6000 new cases occurring each year in the UK. Despite the availability of newer chemotherapeutic agents, mortality rates associated with ovarian cancer have increased since the beginning of the century for the group aged over 40. The UK has the highest incidence rate in Europe (155 cases per million per year) with a mortality rate of 150 per million (Fig. 9.1). The death rate associated with ovarian cancer is higher than the rate for all other gynaecological malignancies combined, although the survival rates for ovarian cancer are similar, stage for stage, to rates for other gynaecological cancers. The relative lack of speciﬁc signs and symptoms of this disease, coupled with the lack of reliable screening strategies, contributes to a condition that is diagnosed at advanced stages in most women, resulting in the poor overall outcome. Once the disease is suspected, a tissue diagnosis is sought – usually obtained at laparotomy; women with advanced or poor-prognosis disease are then treated with platinum-based chemotherapy. Although up to 80% of women respond to a combination of surgery and chemotherapy, most eventually will have recurrence of chemoresistant disease. Overall, the ﬁve-year survival rate in the UK is approaching 40%.

Pathology Primary tumours originate from the serosal surface of the ovary (epithelial), the germ cell and the ovarian stroma. Serous carcinoma accounts for 60–75% of epithelial tumours, endometrioid tumours for 15–30%, mucinous type for 5–15% and the clear cell variant for up to 10%. Clear cell tumours occur in a younger age group and about 50% have associated endometriosis. Although most clear cell cancers are diagnosed at an early stage, these tumours tend to behave aggressively. Approximately 60% of serous tumours are bilateral compared with 30% endometrioid and 20% mucinous. Metastases from ovarian tumours spread by direct extension (transcoelomic), exfoliation into the peritoneal ﬂuid and lymphatic invasion. Metastatic invasion of the ovary by other cancers is surprisingly common – the Krukenberg tumour is usually associated with a mucinous cancer originating from the stomach, but it can also occur secondary to tumours in the breast, colon or gall bladder. Another subset of ovarian cancer based on the histopathology is called borderline or ‘lowmalignant potential’. Nearly 15% of serous and 20% of mucinous tumours have borderline histology. The borderline tumour appears malignant but there is no invasion of the stroma. Although these tumours can occur in women of all ages, they are more typically seen in a Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 9: Ovarian cancer

Fig. 9.1 Agestandardized (European) incidence and mortality rates for ovarian cancer (Great Britain, 1975–2006).

Rate per 100 000 women

20.0 15.0 10.0 5.0 Incidence

Mortality 2005

2003

2001

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

1977

1975

0.0

Year of diagnosis/death

younger age group. At diagnosis, approximately 20% of borderline tumours have spread beyond the ovary via peritoneal implants, which are non-invasive. However, in less than 5% of borderline tumours, so-called invasive implants are found. A borderline tumour with invasive implants behaves more aggressively. Surgery is the mainstay of treatment for women with borderline tumours of the ovary. The prognosis is excellent despite its relative resistance to chemotherapy. Limited follow-up is required for women with borderline disease who have had a pelvic clearance.

Risk factors Ovarian cancer is more common in industrialized countries. Dietary (diet rich in meat and animal fat, proteins, ﬁbre, vitamin A and C) and environmental causes have been inconsistently implicated. ‘Incessant ovulation’ is associated with an increased risk of ovarian cancer. Early menarche and late menopause, nulliparity, including involuntary infertility and unsuccessful fertility treatment, are risk factors for epithelial ovarian cancer. The combined oral contraceptive, which suppresses ovulation, reduces the risk of developing the disease by 10% after one year and 50% after ﬁve years. Similarly, there is a reduced incidence in multiparous women, and for those women who have had sterilization or hysterectomy with ovarian conservation. New views on the pathogenesis of epithelial ovarian cancer suggest the origin from within the ﬁmbrial end of the tube.

Familial ovarian cancer Familial ovarian cancer accounts for up to 10% of cases. The lifetime risk of ovarian cancer for the woman with one ﬁrst-degree relative increases to 5%, and to 7% with two aﬀected relatives. Hereditary ovarian cancer is seen most commonly within the breast–ovarian cancer family syndrome because of mutations in BRCA1 (located on chromosome 17q) or BRCA2 (located on 13q). Features suggestive of these gene mutations are early-onset breast cancer (age <50), the presence of ovarian cancer, male breast cancer and Ashkenazi Jewish ancestry (increases risk by over tenfold). The lifetime risk of ovarian cancer for a woman with a BRCA1 mutation approaches 40% and is less than 20% for a woman with a BRCA2 mutation. BRCA1 mutations are associated with early-onset cancer, and therefore in women with these mutations there is a beneﬁt of risk-reducing surgery (RRS) around 40 years of age. In women with BRCA2 mutations, the age of cancer onset is closer to normal menopausal age, and hence

120

Chapter 9: Ovarian cancer

RRS is recommended in the late forties. RRS for BRCA mutations involves removal of the ovaries and the fallopian tubes. The risk reduction for ovarian cancer is over 90%, but women must be advised that they may develop a peritoneal cancer subsequently, which behaves similarly to ovarian cancer (risk similar to the normal woman, i.e. 0.5–1.0%). This surgery will lead to a reduction in breast cancer by up to 50% if carried out at the age of 40. Screening for these high-risk families, including ultrasound and biochemistry, has been suggested but little beneﬁt has been seen to date. The decision to have RRS is personal. Beneﬁts of the surgery are reduction in the risk of cancer, cessation of periods and complete contraception. The risks of surgery include the impact of premature menopause and a possible adverse emotional and psychological impact of the procedure. The use of hormone replacement therapy after RRS is acceptable if the woman has no personal history of breast cancer or has undergone complete bilateral mastectomy. It does not negate the value of the RRS. For women with mutations, we always include discussion about breast RRS in addition to the above. The beneﬁt of chemoprophylaxis with the oral contraceptive in mutation carriers is unknown. Epithelial ovarian cancer is also seen in about 12% of women with hereditary nonpolyposis colon cancer (HNPCC) syndrome (also known as Lynch syndrome) in addition to their increased risk of endometrial, colorectal and gastric cancers. The HNPCC syndrome results from mutations in DNA mismatch repair genes. Screening for this syndrome is possible using immunohistochemistry of the incident tumour. RRS for this syndrome includes hysterectomy as well as removal of the fallopian tubes and ovaries.

Screening programmes Early-stage ovarian cancer is associated with an excellent prognosis after optimal therapy but there is no reliable screening programme for early detection. High speciﬁcity is vital in screening strategies for ovarian cancer because a positive test result requires a surgical assessment. Small-scale screening studies have suggested a survival advantage. The ongoing UK Trial of Ovarian Cancer Screening Study (UKTOCSS) is comparing 100 000 postmenopausal women as controls with 50 000 women having biochemical screening with subsequent ultrasound, if the cancer-associated antigen (CA125) is rising, and 50 000 women having primary ultrasound screening. Initial data have shown that ultrasound screening results in many more operations than the biochemical screen. A smaller American study is also ongoing, with 75 000 postmenopausal women assigned to either annual combination screening or none. One of the spin-oﬀs from these organized population screening trials may be the ﬁnding of better biochemical markers.

Diagnosis and clinical presentation The symptoms of ovarian cancer are non-speciﬁc and often occur when the disease is advanced. Abdominal discomfort and fullness, bloating (often mistaken for irritable bowel syndrome), bowel habit changes and indigestion are frequent presenting symptoms. Occasionally, women may present with bowel obstruction or shortness of breath due to a pleural eﬀusion. Early-stage disease is often diagnosed incidentally on ultrasound. The presence of a pelvic mass at physical examination is an important sign of ovarian cancer. In advanced stages, abdominal distension due to ascites, abdominal masses (omental disease tumour) and nodularity in the pouch of Douglas can also be felt and a pleural eﬀusion will suggest stage IV disease. Rarely, paraneoplastic syndromes may be present, such as cerebellar degeneration, superﬁcial thrombophlebitis, dermatomyositis and polyarthritis.

121

Chapter 9: Ovarian cancer

In England, women with suspected ovarian cancer are referred to a rapid-access clinic by their general practitioner and usually have an abdominal and pelvic ultrasound scan at their initial visit. Features highly suggestive of ovarian cancer include the presence of a complex ovarian mass (having both solid and cystic components), bilateral masses, the presence of ascites or evidence of peritoneal metastases. Once suspected, a chest X-ray and abdominalpelvic computed tomography (CT) scan with intravenous and bowel contrast are usually performed to aid in the evaluation of a pelvic mass.

Markers Tumour markers are widely used in the diagnosis and management of women with ovarian cancer. CA125, discovered in the early 1980s, is an antigenic determinant on a highmolecular weight glycoprotein expressed by epithelial ovarian tumours and by other tissues of Müllerian origin, and which is recognized by the monoclonal antibody OC125. A serum level of CA125 greater than 35 U/ml is considered to be raised and is noted preoperatively in 80% of women who have epithelial ovarian cancer. CA125 serum levels are also raised in a number of benign conditions such as with menstruation, uterine ﬁbroids, endometriosis, peritoneal inﬂammation (e.g. pelvic inﬂammatory disease), pregnancy and liver disease. This marker is not so useful for mucinous tumours and only 50% of women with stage I ovarian cancer have elevated levels. Many groups have assessed the role of a panel of markers to determine presence of disease but, apart from CA125, none are used in common practice. In women suspected of having an ovarian cancer, management by a gynaecological oncologist working within a multidisciplinary team leads to the best outcome. To allow appropriate triage of women with suspicious masses, CA125 measurement in conjunction with ultrasound features (give three points for the presence of two or more of the following criteria: solid areas in the cyst, the presence of septae, bilateral cysts, ascites or distant metastases) and the patient’s menopausal status (give one point for premenopausal status and three points for postmenopausal status) allows the calculation of the ‘risk of malignancy index’ (RMI) by multiplication (Box 9.1). By using the RMI, clinicians can appropriately refer women at high risk of ovarian malignancy to a gynaecological oncologist working in a cancer centre (Table 9.1). In cases of widespread abdominal malignancy, additional tumour markers (carcinoembryonic antigen (CEA), CA19.9, CA15.3) may be of value and a ratio of 25:1 for CA125:CEA is used as an entry criteria for some trials. There is certainly value in using CA125 in follow-up for epithelial tumours. If the woman has an early-stage tumour and she Box 9.1 Calculating the risk of malignancy index (RMI); this is a modiﬁcation of the original RMI, using modiﬁed scores RMI ¼ U M CA125 where: U = 0 (for ultrasound score of 0); 1 (for ultrasound score of 1); 3 (for ultrasound score of 2–5). Ultrasound scans are scored one point for each of the following characteristics: multilocular cyst; evidence of solid areas; evidence of metastases; presence of ascites; and bilateral lesions. M = 3 for all postmenopausal women. CA125 is serum CA125 measurement in units/ml

122

Chapter 9: Ovarian cancer

Table 9.1 An example protocol for triaging women using the risk of malignancy index (RMI)

Risk

RMI

Women (%)

Risk of cancer (%)

Low

<25

40

<3

Moderate

25–250

30

20

High

>250

30

75

is treated with surgery alone, the rise in CA125 gives a lead time of up to three months prior to clinical detection of disease, allowing for the early use of adjuvant chemotherapy. In advanced disease, the CA125 gives a measure of the response to chemotherapy as well as a marker for recurrence. Although there is a lead time prior to clinically obvious disease in these women with advanced disease who have had a complete response to chemotherapy, second-line regimens are not oﬀered until the woman is symptomatic because the treatment of recurrence is only palliative. We await the results of a UK trial which has addressed the survival advantage of early treatment versus delayed until symptomatic. In situations where there is uncertainty about the tissue origin of a complex pelvic mass or widespread metastases, the use of diﬀerent panels of immunohistochemistry markers on biopsy/ascitic cytology specimens can be informative. For example, a combination of cytokeratin, CA125 and CEA markers may help diﬀerentiate a gastrointestinal tumour from a female genital tract cancer.

Treatment The management of patients with epithelial ovarian cancer starts with accurate diagnosis and staging. A combination of surgery, aiming for optimal cytoreduction (see below) of metastatic disease, and platinum-based chemotherapy is used for all but low-grade early-stage disease.

Staging Staging of ovarian cancer (Fig. 9.2) is important for prognosis and treatment. Like other gynaecological cancers, ovarian cancer is staged using the International Federation of Gynecology and Obstetrics (FIGO) system. The woman with suspected ovarian cancer is typically staged with a laparotomy through a midline incision. CT scanning may be helpful in staging women receiving primary chemotherapy.

Surgery Surgery plays a major role in the management of ovarian cancer. At operation, the diagnosis is conﬁrmed (histology of the tumour removed), the disease can be appropriately staged and therapy started. At this staging operation, the extent of the disease is documented. Peritoneal washings or ascites are sent for cytological examination. Biopsies are taken from any adhesion or abnormality detected. Some surgeons take multiple biopsies of the peritoneum, including from the diaphragm. A pelvic and para-aortic lymphadenectomy may be useful in apparently early-stage disease, as micrometastases would up-stage the woman to stage IIIC. In the UK, there is less emphasis on a complete lymphadenectomy and we are often guided by the preoperative imaging. Usually a bilateral oophorectomy, total abdominal hysterectomy including the fallopian tubes and cervix and an omentectomy are performed at the primary surgery. Other tumour masses are removed if possible. When a mucinous

123

Chapter 9: Ovarian cancer

IA

IB

IC

T1a

T1b

T1c

Malignant cells in ascites

IIA

IIB

IIC

T2a

T2b

T2c

Rectum Aorta

Aorta

Malignant cells in ascites

III

III

IV

T3

T3

M1

IIIC/3c Peritoneal metastases >2 cm

IIIC/3c Microscopic only IIIB/3b Microscopic peritoneal metastases ≤ 2 cm

Parenchymal

Liver capsule

Fig. 9.2 Diagrammatic representation of staging for ovarian cancer.

tumour is suspected, it is important to remove the appendix. At the completion of surgery, it is vital to record the extent of residual disease and a presumed stage which will aid the multidisciplinary team. Decisions about adjuvant therapy and overall prognosis depend on accurate staging (Fig 9.3). In early-stage disease, surgery can be curative. This surgery includes removing the aﬀected ovary, sampling of the contralateral ovary and completing a full staging procedure.

124

Grade 3

Observe or carboplatin ± taxane for 3–6 cycles

Grade 2

Carboplatin ± taxane for 3-6 cycles

Stage IC, any grade Stage II

Intraperitoneal chemotherapy may be considered in lowvolume optimally debulked stage III

Carboplatin/cisplatin ± taxane for 6 cycles

Stage III/IV

Cytoreductive surgery if clinical stage II, III or IV

Further 3 cycles of chemotherapy (total six)

Interval debulking after three cycles Chemotherapy as indicated by tumour response and potential resectability

Consider neoadjuvant chemotherapy in women with bulky stage III/IV disease who are not surgical candidates up front (diagnosis by biopsy/paracentesis)

Fig. 9.3 Algorithm for investigation and management of ovarian cancer. TAH, total abdominal hysterectomy; BSO, bilateral salpingo-oophorectomy.

Observe

Grade 1

Stage IA/B

Laparotomy – TAH, BSO with comprehensive staging or unilateral salpingo-oophorectomy (stage IA/B if fertility preservation)

Ultrasound or abdominal-pelvic CT scan CA125 Chest X-ray Full blood count, biochemistry Gastrointestinal investigations if indicated

Suspicious pelvic mass and/or ascites, abdominal distension

Chapter 9: Ovarian cancer

In the woman who has completed her family, it is sensible to remove the contralateral ovary and perform a hysterectomy. It is appropriate to excise the malignant cyst without spillage to reduce upstaging the patient. This is important when laparoscopy is used in the initial diagnosis. We aim to remove the cyst or ovary intact laparoscopically by placing it in a waterproof bag prior to deﬂation and removal from the abdomen. Since the early 1970s, gynaecologists have been trying to reduce or remove the tumour bulk completely in advanced disease; this is called cytoreduction. Primary surgery is that performed at the initial staging. Interval surgery refers to an operation in those women who have unresected disease at presentation and start with chemotherapy (three cycles) with a planned attempt at radical surgery followed by chemotherapy again. Secondary surgery includes all the other aspects such as the second-look procedure (an operation performed when the woman has had primary surgery, adjuvant chemotherapy and then achieved a clinical complete response) and secondary cytoreductive and palliative surgery. The impact of primary cytoreduction on survival seems to be signiﬁcant, although this is based on retrospective data. We feel that radical surgery should be attempted in the treatment of ovarian cancer, either at the initial diagnosis or as an interval procedure. In a small proportion of women receiving neoadjuvant therapy who progress through the chemotherapy (primary chemoresistance) we would then oﬀer palliative care and no surgery. Reducing the size of tumour to either no residual disease or disease less than 0.5 cm is now considered the aim of ovarian cancer surgery, and the reported success rate varies from under 30% to over 90%. In order to achieve optimal cytoreduction (less than 0.5 cm), surgeons have tried peritoneal stripping, bowel resection and the removal of other organs aﬀected by the cancer. It is unclear whether this extended surgery is beneﬁcial, and it may be that aspects of the tumour biology allow the surgeon to remove all the cancer in one woman and not in the next. Other beneﬁts of this surgery are possible relief of symptoms and improvement in the quality of life. It appears that there is some role for interval debulking surgery after initial chemotherapy following inadequate initial surgery or when surgery has been delayed. The second-look procedure is currently not used in the UK as there is no survival advantage, but it may allow triaging patients for maintenance therapy if such therapy becomes available.

Early-stage disease In women with disease apparently conﬁned to the pelvis, thorough staging is essential to deﬁne the correct extent of the disease at time of diagnosis. In patients with stage IA/B and grade 1/2 disease, studies have shown the same outcome whether they have complete staging or adjuvant chemotherapy. However, it is this group with stage IA/B where a woman has had incomplete surgery that we feel will beneﬁt from an appropriate restaging procedure including omentectomy, pelvic and para-aortic lymphadenectomy. If the biopsies were positive, the patient is up-staged and given chemotherapy. Several studies have shown that staging without due assessment of all areas where ovarian cancer can spread will lead to under-staging and inadequate treatment. The surgeon has to balance the extent of surgery and knowledge of the accurate stage against the complications for the patient as well as her general health, and we prefer to use a second laparoscopic staging procedure if required.

Advanced-stage disease Cytoreductive surgery is one of the main components in the treatment of advanced ovarian cancer. Since the mid 1970s, the amount of residual disease is considered the most important modiﬁable prognostic factor that inﬂuences survival in women with advanced disease.

126

Chapter 9: Ovarian cancer

Fig. 9.4 Widespread miliary deposits in a case of advanced ovarian cancer.

The usual sites of disease most frequently precluding optimal cytoreduction are the diaphragm, bowel (Fig. 9.4) and portal triad. The gynaecological oncologist can reduce disease from the bowel and diaphragm with resection, but disease in the upper abdomen in the porta hepatis will not be resectable. Radical surgical procedures, including intestinal resection, splenectomy, diaphragmatic resection and hepatic resection, have been described with acceptable morbidity. At our institution, we undertake bowel resection with peritoneal stripping if we can achieve optimal debulking. However, one must consider the patient’s general state of health and acceptance of a possible stoma before undertaking this type of surgery. In addition, unless one can achieve acceptable cytoreduction of all the cancer there is no point in a radical pelvic resection leaving bulk upper abdominal disease. We favour exploring the upper abdomen and ensuring complete clearance before starting on the pelvic resection. When neoadjuvant chemotherapy is considered, we use CT scanning to determine the extent of disease and histological examination, in the form of biopsies or cytology, to determine cell type to attribute a stage for those women unﬁt for surgery or in whom we do not feel that we will obtain a complete tumour debulk. The rationale for neoadjuvant chemotherapy and interval debulking is for the patient with a high risk of major perioperative morbidity at initial diagnosis, such as with a poor performance status or with radiologically detected unresectable disease. Survival rates are similar among those women undergoing primary surgery with adjuvant chemotherapy and those having neoadjuvant chemotherapy followed by interval debulking surgery. However, the perioperative morbidity appears to be lower in those having initial chemotherapy. Interval debulking surgery by a trained gynaecological oncologist improves survival in women who have undergone initial suboptimal primary debulking surgery, although we await more data relating to the overall eﬃcacy of this management. There is no consensus about the preoperative tests (radiological and biochemical) that can be used to determine surgical respectability. The reﬁnement of such tests would allow us to operate on those women who would have a beneﬁt and oﬀer chemotherapy to the remainder. The success of the cytoreduction is multifactorial but is also surgeon dependent. It is clear, however, that women with widespread disease and poor performance status are less likely to undergo optimal debulking. Although primary cytoreduction is considered the cornerstone of initial management for women with advanced disease, the role of aggressive surgery in those with recurrent disease is debated and appropriate selection is needed. Ideal candidates for secondary surgery would have three or more of the following factors at recurrence: disease-free interval longer than 12 months, no liver metastasis, a solitary tumour and tumour size smaller than 6 cm.

127

Chapter 9: Ovarian cancer

Fertility-sparing procedures Two groups of young women with epithelial ovarian cancer may be considered for a conservative surgical approach: borderline tumours and those with stage IA/B with a wellor moderately diﬀerentiated cancer. Borderline ovarian tumours are often diagnosed incidentally after oophorectomy or cystectomy in young women. As the rate of recurrence after a conservative treatment is very low (~5%) and there is no survival advantage for radical surgery, it would seem sensible for these women not to have a restaging procedure if fertility was still required. A conservative approach should be considered for women requesting fertility preservation, even with invasive disease. However, it is vital for these women to undergo a complete staging including unilateral pelvic and para-aortic lymphadenectomy.

Chemotherapy Early-stage disease Usually, for early-stage disease we consider six cycles of adjuvant carboplatin monotherapy. The addition of a taxane in stage 1C disease is not justiﬁed as there is little added beneﬁt but considerable increase in toxicity. Two large randomized trials have compared chemotherapy with no immediate therapy for early-stage ovarian cancer: the International Collaborative Ovarian Neoplasm (ICON) trial and the Adjuvant ChemoTherapy In Ovarian Neoplasm Trial (ACTION). The results showed survival rates of 82% versus 74%, an 8% increase in favour of chemotherapy. Current recommendations are for no postoperative treatment in women with early-stage (stage IA/B) ovarian cancer with well- or moderately diﬀerentiated tumours – this is in women who have undergone a full staging procedure including lymphadenectomy.

Advanced-stage disease Since the advent of cancer chemotherapy, many treatments and routes of administration have been tried and despite this, most women with stage III and IV ovarian cancer are not cured of their disease. With the widespread adoption of adjuvant platinum-based combination chemotherapy for advanced-stage disease, the relapse-free interval after completion of that therapy (or the platinum-free interval) has been recognized as a predictor of the likelihood of subsequent response to chemotherapy. Patients with a relapse-free interval of more than 6 months (platinum sensitive) have a higher chance of responding to platinum again. Platinum-resistant disease includes disease that relapses within six months of adjuvant therapy or disease that progresses while the patient is taking platinum in the salvage setting. In 1996, the Gynecologic Oncology Group (GOG) 111 trial, which compared cyclophosphamide and cisplatin with paclitaxel and cisplatin in patients with stages III and IV ovarian cancer, showed response rates of 60% versus 73%, progression-free survival times of 13 versus 18 months and median survival times of 24 versus 38 months, all favouring the paclitaxel and cisplatin combination. Several studies, including GOG 158, have shown that cisplatin and carboplatin have equivalent beneﬁt when used with paclitaxel. The ICON group reported a randomized study (ICON 3) of women with stages I to IV ovarian cancer who received paclitaxel plus carboplatin or either single-agent carboplatin or cyclophosphamide, doxorubicin and cisplatin (CAP). No diﬀerence was found in overall survival between the group taking paclitaxel plus carboplatin

128

Chapter 9: Ovarian cancer

and the controls (36.1 versus 35.4 months at follow-up of 51 months). The investigators concluded that single-agent carboplatin and CAP were as eﬀective as paclitaxel plus carboplatin as ﬁrst-line treatment, and the more favourable toxicity proﬁle of single-agent carboplatin favoured it as ﬁrst-line chemotherapy for ovarian cancer. These combined results have led to carboplatin and paclitaxel being used as ﬁrst line in the USA whereas in the UK, the option of single-agent carboplatin is considered. There is no evidence that adding a third cytotoxic agent prolonged progression-free survival.

Intraperitoneal chemotherapy As advanced ovarian cancer is often limited to the peritoneal cavity, intraperitoneal chemotherapy was proposed as a strategy to increase drug concentrations in the abdominal cavity. Following several small studies showing an advantage, GOG 172 in 2006 compared intravenous versus intraperitoneal cisplatin with paclitaxel following optimal debulking surgery. Only 42% of the women in the intraperitoneal group completed six cycles of the assigned therapy, but there was an increase in 15 months in the median overall survival time to over 65 months, favouring the intraperitoneal regimen. Quality of life was markedly worse in the intraperitoneal therapy group before cycle four and at three to six weeks after treatment but, by one year after treatment, quality of life was similar in both arms. As a result of these intraperitoneal studies showing a survival beneﬁt, the National Cancer Institute issued clinical advice recommending an intraperitoneal regimen. However, not all agree and we await further discussion and trials as to the relevance of intraperitoneal therapy.

Consolidation and maintenance therapy Although more than 60% of women with ovarian cancer achieve a clinical complete remission with ﬁrst-line chemotherapy, disease will recur in most. To improve survival, consolidation therapy, maintenance therapy, neoadjuvant therapy, novel cytotoxic agents and novel biological agents are being studied, although no results are forthcoming as yet. Consolidation and maintenance therapies are not used routinely in the UK as there is no evidence that the intervention improves overall survival. There is a trial looking at hormonal treatment (tamoxifen) in women with raised CA125 and no radiological evidence of recurrent disease, but this is not true maintenance therapy.

Chemotherapy for recurrent disease Unfortunately, most women with advanced-stage ovarian cancer will experience disease relapse after primary chemotherapy and will develop symptomatic recurrent disease. Median survival after recurrence is two years. As cure is not possible, the primary goal of therapy is management of symptoms. Women with asymptomatic recurrence do not require immediate therapy because no evidence is available to indicate that survival is improved by earlier administration of chemotherapy. Women who did not receive taxanes in their initial chemotherapy may beneﬁt from its use at this stage. The addition of other drugs such as topotecan, liposomal doxorubicin, gemcitabine and oral etoposide can be considered. High-dose chemotherapy with autologous stem cell transplantation has not been shown to produce better disease-free or overall survival than standard therapy. In vitro chemotherapy sensitivity and resistance assays do not oﬀer any survival advantage. Hormonal therapies have been investigated in women with recurrent ovarian cancer but have not been found to be particularly helpful. Oestrogen receptor status of the primary

129

Chapter 9: Ovarian cancer

tumour is being used by some to guide hormonal therapy, but it is unclear what the receptor status of the recurrent tumour is as it is unusual to take biopsies of the tumour at recurrence. The issue of how many treatment regimens to use in women with advanced ovarian cancer is an area of controversy. With low response rates with subsequent chemotherapies, patients need to decide whether to continue chemotherapy or receive best supportive care only.

Palliative care Most women will die of their cancer and have to contend with pronounced morbidity on this journey. Bowel obstruction is the most common end-of-life complication and gives rise to many hospital admissions. It is important to take a CT scan early in the diagnosis of bowel obstruction, as on occasions surgery may be helpful, especially if there appears to be only one site of obstruction. Typically conservative measures are used ﬁrst. If conservative measures are unsuccessful and there is limited scope for further chemotherapy, a multidisciplinary discussion about the woman’s further management is required as the surgical intervention can be challenging. Intervention may not prolong life. Thus, the patient’s overall health, the aggressiveness of the cancer, the extent of disease within the peritoneal cavity and the availability of future chemotherapy options for a speciﬁc patient must all be carefully considered before taking the decision to operate. The most common reason against surgery is the extent of involvement of the intestinal tract. Palliative care can be subdivided into nonmedical interventions, such as stent placement and gastric venting, and medical interventions, such as octreotide and other antisecretory agents, corticosteroids, pain medications and anti-emetics. A holistic approach for women with this terminal illness is important and requires considerable psychological support. There seems to be no value for parenteral nutrition for these women in their terminal phase.

New therapeutic strategies Ovarian cancer provides a unique challenge in oncology. Aggressive surgery is used in the initial phase, even in the face of extensive metastatic disease, to reduce disease extent to a low level. However, due to drug resistance, the cancer grows again despite the initial excellent chemotherapy responses, and the development of new therapeutic approaches to address drug resistance is a high priority. Numerous strategies are being pursued. High-throughput genomic analysis of thousands of genes may help in identifying a proﬁle that predicts chemosensitivity. Such studies may in turn provide information about intelligent drug design. Technologic advances that allow examination of the molecular machinery driving cancer cells have helped to identify numerous mediators within ovarian cancer cells that can be targeted with new molecular strategies. Examples include agents that block stimulatory growth factor receptors at the cell surface (trastuzumab (Herceptin)), small-molecule inhibitors of signal transduction pathways, approaches that harness speciﬁc elements of the immune system that ovarian cancers have successfully suppressed, gene and viral therapy strategies and antiangiogenesis strategies (bevacizumab (Avastin)).

Conclusion Symptoms of ovarian cancer are often non-speciﬁc and usually occur after the disease has spread. The presence of a pelvic mass in a postmenopausal woman is the most important sign of ovarian cancer. Serum biochemistry (CA125) and ultrasound level are widely used

130

Chapter 9: Ovarian cancer

to triage for a possible epithelial ovarian cancer in the primary assessment of a pelvic mass. The important step in the management of women with epithelial ovarian cancer is treatment at a cancer centre by a multidisciplinary team that can oﬀer accurate diagnosis, thorough staging and appropriate surgery, together with platinum/taxane combination chemotherapy. This achieves response in most patients. CA125 measurement is used to assess response to treatment and for detection of early recurrence. An increasing understanding of the biology of ovarian cancer and its response to treatment is aiding development of new treatment strategies.

131

10 Chapter

Endometrial cancer Catherine Holland

Aetiology and risk factors In the UK, endometrial cancer is the second most common gynaecological cancer and the ﬁfth most common female cancer overall, with approximately 6400 new cases per year. Most women with endometrial carcinoma are postmenopausal, with the majority of cases occurring between 50 and 59 years of age. Up to 25% of aﬀected women are pre menopausal and a minority (<5%) are diagnosed before the age of 40. Endometrial cancer is associated with a number of reproductive risk factors (see Box 10.1), including early menarche and late menopause, continuous anovulation (e.g. polycystic ovarian syndrome), nulliparity and infertility. Excessive unopposed endogenous oestrogen is common to all of these clinical scenarios. Obese women are at increased risk of endometrial cancer. This is probably a result of increased conversion of adrenal androgens to oestrogens in peripheral fat. Levels of sex hormone-binding globulin are also reduced in obese women, further increasing the amount of bioavailable oestrogen. Occasionally, endometrial cancer may arise as a result of excessive endogenous oestrogen production by an ovarian granulosa cell tumour. Endometrial cancer is also associated with diabetes and hypertension, although it is not clear whether these are simply markers for obesity or whether they have an independent association. It has long been accepted that the use of unopposed systemic oestrogen replacement therapy increases the risk of endometrial cancer. Women with a uterus in situ should be given combined oestrogen and progestogen therapy containing at least 12 days of progestogen. Women taking continuous combined hormone replacement therapy (HRT) have a lower risk of developing endometrial cancer than women who have never used HRT. Use of the combined contraceptive pill is also protective. Tamoxifen, widely used in the treatment of breast cancer, is an antioestrogenic drug that has weak oestrogenic eﬀects on the female genital tract. Women taking tamoxifen are at increased risk of developing endometrial abnormalities including polyps, hyperplasia and endometrial cancers. There is no proven role for routine endometrial sampling or surveillance in these women, but abnormal bleeding in these women should be investigated promptly and thoroughly. Hereditary endometrial cancer accounts for <5% of all endometrial cancers and is usually seen in association with hereditary non-polyposis colon cancer syndrome (HNPCC). HNPCC is an autosomal dominant inherited disorder arising from germline mutations in DNA repair genes. It is associated with a 40–60% lifetime risk of developing endometrial cancer and should be suspected in all women below the age of 45 presenting with endometrial cancer, or where there is a strong family history of bowel, endometrial, gastric or other associated cancers. Currently prophylactic hysterectomy is the only proven method of preventing Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 10: Endometrial cancer

Box 10.1 Risk factors for endometrial cancer *

Obesity

*

Early menarche/late menopause

*

Prolonged anovulation

*

Polycystic ovarian syndrome

*

Nulliparity

*

Infertility

*

Diabetes mellitus

*

Hypertension

*

Unopposed systemic oestrogen therapy

*

Tamoxifen

*

Oestrogen-secreting ovarian tumours

*

Hereditary causes, e.g. HNPCC

Box 10.2 Modiﬁed classiﬁcation system for endometrial carcinomas *

Endometrioid adenocarcinoma

*

Mucinous adenocarcinomaa

*

Serous adenocarcinoma

*

Clear cell adenocarcinoma

*

Undiﬀerentiated carcinoma (large and small cell types)

*

Carcinosarcoma

*

Squamous adenocarcinoma

*

Mixed carcinoma

*

Metastatic carcinoma

a

An uncommon form of endometrioid adenocarcinoma with similar behaviour.

endometrial cancer in aﬀected women. Women suspected of having HNPCC should be referred for genetic counselling and subsequent genetic testing if appropriate.

Pathology of endometrial cancers Endometrial carcinomas arise from the epithelial glandular component of the endometrium. A number of distinct histological subtypes of endometrial carcinoma are recognized and classiﬁed according to histological cell type and pattern (Box 10.2). The most common type is endometrioid endometrial adenocarcinoma, which accounts for approximately 75% of cases of endometrial cancer. Endometrioid cancers arise from a precursor lesion, atypical hyperplasia, and are typically associated with unopposed endogenous and exogenous oestrogens, obesity, diabetes mellitus and hypertension. Invasive endometrial carcinoma is seen in up to 50% of hysterectomy specimens removed for a diagnosis of atypical hyperplasia. Other types of

134

Chapter 10: Endometrial cancer

Fig. 10.1 FIGO grade 2 endometrioid carcinoma (shown in sagittal section). The tumour is expanding the uterine isthmus and inﬁltrating the cervical stroma. There are surface deposits on the ovaries. The lymph nodes (not shown) were disease-free (stage IIIA). The patient was treated with adjuvant chemotherapy followed by pelvic radiotherapy.

endometrial hyperplasia, i.e. simple hyperplasia and complex hyperplasia, have a low risk of progression to cancer (approx 1–3%). Non-endometrioid endometrial carcinomas are less common and include serous carcinoma and clear cell carcinoma. Molecular diﬀerences suggest that these tumours develop via diﬀerent pathways from endometrioid tumours. Endometrioid carcinomas are commonly referred to as type 1 tumours and non-endometrioid variants as type 2 tumours. Each tumour type has characteristic histopathological appearances. In contrast with type 1 (endometrioid) tumours, type 2 (non-endometrioid) carcinomas are not associated with hyper-oestrogenism or obesity. Most type 2 cancers are high-grade, aggressive tumours and have a high rate of recurrence even where there is no extrauterine disease. Serous carcinoma in particular is known to have a high risk of metastasis even when conﬁned to the endometrium and exhibits a similar pattern of spread to ovarian serous carcinoma. In contrast with endometrial carcinomas, the endometrial stromal sarcomas are uncommon cancers that originate in the non-glandular connective tissue of the endometrium. Uterine carcinosarcoma (malignant mixed Müllerian tumour (MMMT)), is a rare uterine cancer that contains malignant cells of both epithelial and sarcomatous appearance. This is now considered to be a metaplastic carcinoma rather than a sarcoma. Rarely, other cancers can metastasize to the endometrium either from other pelvic organs or from more distant sites, most commonly breast. Endometrial cancers generally spread via direct extension into the myometrium and cervix (Fig. 10.1). Transtubal metastasis may lead to deposits on the peritoneum and ovaries, and lymphatic spread occurs to pelvic and para-aortic lymph nodes. Spread may occur directly to the para-aortic nodes from the lymphatic channels draining the upper part of the uterus, but is more usually seen when the pelvic nodes are involved. Vaginal metastases arise by direct extension or via lymphatic or vascular channels. Haematogenous spread leads to metastases in distant, extra-abdominal sites, most commonly to the lungs. Carcinosarcoma behaves in a particularly aggressive fashion, with a tendency to early lymphatic and vascular spread. Endometrial tumours are further subclassiﬁed according to grade (grades 1–3), with lowgrade tumours (grade 1) more closely resembling normal endometrium than high-grade tumours (grade 3). Tumour grade is related to the risk of metastasis and outcome, with grade 3 tumours metastasizing more readily than grade 1 tumours. Grade is therefore important in guiding management. Non-endometrioid tumours are all considered high grade. The depth of myometrial invasion is another important prognostic indicator in early endometrial

135

Chapter 10: Endometrial cancer

IA

IA

IB

II

IIIA

IIIB

IVA

I A/B

IIIC 1/2

IVB

Fig. 10.2 Staging for endometrial cancer.

136

cancer and this is reﬂected in the International Federation of Gynecology and Obstetrics (FIGO) staging criteria (2009 classiﬁcation – see Appendix 3; Fig. 10.2). The staging system is surgicopathological, relying on accurate documentation of the extent of disease at surgery as well as histopathological ﬁndings. Disease stage and tumour grade together determine the need for postoperative adjuvant treatment.

Chapter 10: Endometrial cancer

Clinical presentation Postmenopausal bleeding is the most common presenting symptom in women with endometrial cancer. Premenopausal women may present with postcoital or intermenstrual bleeding and, in some cases, change in menstrual pattern, e.g. increasingly heavy, erratic menses. Persistent vaginal discharge in a postmenopausal woman may indicate a pyometra, sometimes as a result of postmenopausal bleeding with a stenosed cervix. The clinician should be aware that endometrial pathology is a possibility and investigate appropriately.

Assessment of the woman with postmenopausal bleeding (Fig. 10.3) Ultrasound scan Following a clinical examination to exclude vulval, vaginal or cervical tumours, the initial investigation of women with postmenopausal bleeding should be a transvaginal ultrasound scan. Using ultrasound, the endometrial thickness can be measured and any focal or global thickening documented. In addition, coexisting ovarian cysts may be identiﬁed. The upper limit of normal for endometrial thickness in the postmenopausal woman is typically quoted as 5 mm. Some clinics use a cut-oﬀ of 4 mm. This not only may increase sensitivity but also increase the number of false positives. The negative predictive value of a normal ultrasound scan is very high, and endometrial cancer is rare when the endometrium is less than 4–5 mm thick. Therefore women with a thin endometrium do not require endometrial biopsy unless they have further bleeding. Where the endometrium is 5 mm thick or greater, further evaluation is necessary. Women taking tamoxifen frequently have appearances of a thickened Fig. 10.3 Algorithm for assessment of the woman with postmenopausal bleeding. Transvaginal ultrasound is an essential investigation in the triage of women with postmenopausal bleeding.

Exclude vulval, cervical and vaginal malignancy by examination

Transvaginal ultrasound scan

Smooth endometrium <5 mm

Discharge with treatment for atrophy if required

Endometrium ≥ 5 mm

Endometrium ≥ 5 mm

Generalised thickening

Focal thickening

Endometrial biopsy

Hysteroscopy and biopsy

137

Chapter 10: Endometrial cancer

endometrium as a consequence of oestrogenic eﬀects on subendometrial tissues. Ultrasound scan is therefore less discriminatory in the investigation of women taking tamoxifen. These women are better investigated with outpatient hysteroscopy as a ﬁrst-line investigation.

Outpatient endometrial assessment Approximately 80% of women with postmenopausal bleeding can be eﬀectively investigated as outpatients. This is an important consideration, as many women with postmenopausal bleeding have signiﬁcant co-morbidities that increase the risks associated with general anaesthesia. Outpatient endometrial biopsy can be undertaken using a number of diﬀerent endometrial aspiration devices. Blind endometrial biopsy is successful in diagnosing endometrial cancer in approximately 80% of cases and is more sensitive where at least 50% of the endometrial cavity is involved with tumour. Small tumours and cancers arising within endometrial polyps are likely to be missed by blind endometrial biopsy, and therefore outpatient hysteroscopy is advantageous where focal endometrial thickening is seen on ultrasound scan. Formal endometrial curettage is necessary in a minority of cases, but may be required where there is severe cervical stenosis.

Role of imaging in treatment planning X-ray Once endometrial carcinoma is conﬁrmed further imaging may be used to aid management decisions. A chest X-ray is required prior to surgery to detect metastases, although these are rare in early-stage endometrioid disease.

Magnetic resonance imaging The current national model of endometrial cancer management mandates that women likely to need additional (adjuvant) treatment following surgery should be referred to a cancer centre for surgery. These are women with disease likely to be stage 1B or greater (FIGO 2009 classiﬁcation) and all those with grade 3 disease or non-endometrioid histology. Women with low-grade, probable stage IA disease may have surgery performed by a lead clinician for gynaecological cancer at a unit hospital. Within the context of this national framework, magnetic resonance imaging (MRI) may be used to conﬁrm suitability for local surgery by assessing depth of myometrial invasion and identifying extrauterine extension of tumour. MRI is more sensitive than either ultrasound scan or computed tomography (CT) in the assessment of myometrial invasion. MRI is also helpful in identifying possible cervical stromal involvement, in which case more radical surgery can be considered (Fig. 10.4).

Computed tomography CT is used as an alternative imaging modality in a minority of patients with endometrial cancer. CT is used to evaluate the upper abdomen in women with serous endometrial tumours in view of the tendency to transabdominal spread. Similarly, CT of the thorax is required in cases of carcinosarcoma to identify pulmonary metastases which are more frequent with this subtype. Postsurgical CT imaging is required for planning adjuvant radiotherapy treatment in those women at high risk of recurrent disease.

138

Chapter 10: Endometrial cancer

Fig. 10.4 MRI scan of the pelvis in a woman with endometrial carcinosarcoma. The tumour is expanding the cervix. The hysterectomy specimen showed inﬁltration of the cervical stroma.

Positron-emission tomography (PET) Fluorine-18-ﬂuoro-2-deoxy-d-glucose positron-emission tomography (FDG-PET) is a specialised nuclear medicine technique that uses uptake of a metabolic marker to identify metabolically active sites. The technique is used in combination with CT scan (PET/CT) to provide anatomical localization of the metabolically active site(s). It is most commonly used in the oncology setting where it can help to identify and localize recurrence. Preliminary studies in endometrial cancer show that PET may have a role in accurately determining the location(s) of disease recurrence although further evaluation in larger studies is awaited.

Surgery for endometrial cancer Total abdominal hysterectomy and bilateral salpingo-oophorectomy Despite the use of preoperative imaging as an aid to treatment planning, staging of endometrial cancer remains surgicopathological, relying both on surgical assessment of intraabdominal disease and, in most cases, where disease is conﬁned to the uterine corpus, depth of myometrial invasion. Full surgical staging using the FIGO system comprises total ‘simple’ hysterectomy, bilateral salpingo-oophorectomy (total abdominal hysterectomy/bilateral salpingo-oophorectomy (TAH/BSO)), peritoneal cytology and pelvic and para-aortic lymphadenectomy. Many women do not, however, undergo full surgical staging. Lymph node metastases are present in less than 5% of cases of grade 1 and 2 endometrioid tumours with less than 50% myometrial invasion. Most gynaecological oncologists therefore accept that the risks and side eﬀects of full surgical staging are not justiﬁed, given the low risk of nodal metastases. These women are suitable for surgery with ‘simple’ extra-fascial TAH/BSO and peritoneal cytology. When the clinical examination and/or preoperative imaging suggest cervical stromal involvement (stage II disease), radical hysterectomy, bilateral salpingo-oophorectomy and full pelvic lymphadenectomy are appropriate. Radical hysterectomy requires extensive pelvic dissection to remove paracervical tissue so that the tumour is completely removed with clear surgical margins. If adequate surgical margins are achieved and the nodes are free of disease,

139

Chapter 10: Endometrial cancer

Table 10.1 Approximate ﬁve-year survival according to the FIGO stage

Stagea

Approximate ﬁve-year survival (%)

I

85

II

75

III

45

IV

25

a

It should be noted that survival varies within substages, e.g. ﬁve-year survival for stage IIIA disease diﬀers from that for stage IIIC disease.

the procedure is considered curative and adjuvant radiotherapy is not indicated. Radical hysterectomy has a higher morbidity than ‘simple’ hysterectomy, and some women with endometrial cancer will not be ﬁt enough for this type of procedure. These women will require ‘simple’ TAH/BSO lymph node sampling followed by adjuvant radiotherapy.

Role of extended surgical staging The overall ﬁve-year survival rate following treatment for endometrial carcinoma is approximately 75–80%. This reﬂects early presentation in the majority of cases, although ﬁve-year survival falls rapidly as stage increases (Table 10.1). Approximately 8–10% of women with disease apparently conﬁned to the uterus have occult nodal disease. It has therefore been surgical practice to selectively perform lymphadenectomy in addition to TAH/BSO for those women with a high risk of lymph node metastasis based on preoperative assessment of depth of myometrial invasion and/or grade. There has been considerable debate regarding the role of lymphadenectomy in endometrial cancer surgery. A large prospective randomized trial, A Study in the Treatment of Endometrial Cancer (ASTEC), has addressed the value of lymphadenectomy in women with disease clinically conﬁned to the uterus. Women were randomized to lymphadenectomy or no lymphadenectomy and 704 patients were randomized to each arm of the trial. The initial results of ASTEC have shown no signiﬁcant diﬀerence in overall survival between those women who had lymphadenectomy and those who did not, although the number of treatment-related deaths was higher in those women who had lymphadenectomy. These results clarify the role of lymphadenectomy and conﬁrm that there is no therapeutic beneﬁt for routine lymphadenectomy in women with endometrioid endometrial carcinoma and moderate to high risk of lymph node involvement. The anticipated ﬁnal results of ASTEC are likely to have a signiﬁcant impact on surgical practice in endometrial cancer. Although some women with other histological subtypes, e.g. serous and clear cell carcinoma, were included in ASTEC, the role of lymphadenectomy in non-endometrioid disease has yet to be established in a large, adequately powered prospective study. The risk of occult extrauterine metastasis is higher for non-endometrioid tumours, and extended surgical staging is usually undertaken. An omental biopsy is taken in cases of serous endometrial carcinoma in view of the risk of omental metastasis. Retrospective studies of fully surgically staged serous uterine cancers suggest that lymphadenectomy may be suﬃcient where the nodes are negative and that adjuvant radiotherapy may not be required in all cases. Prospective randomized trials are thus needed to determine best practice in nonendometrioid tumours.

140

Chapter 10: Endometrial cancer

Fig. 10.5 Laparoscopic lymph node sampling in a woman with a suspicious lymph node on MRI scan. The patient also had laparoscopically assisted vaginal hysterectomy and bilateral salpingo-oophorectomy. EIA, external iliac artery; LN, lymph node.

Minimal access surgery Laparoscopic surgery is increasingly used in the management of endometrial cancer. Both total laparoscopic hysterectomy and laparoscopically assisted vaginal hysterectomy (both with bilateral salpingo-oophorectomy) are feasible procedures in women with endometrial cancer and can be successfully performed with low complication rates in elderly and obese women. Both pelvic (Fig. 10.5) and para-aortic lymphadenectomy and omental sampling can be safely and adequately performed if deemed necessary for the management of high-grade disease. Laparoscopic surgery for endometrial cancer should only be undertaken by appropriately trained surgeons and the surgical aims should not be compromised by the approach taken. In retrospective studies of women with endometrial cancer, recurrence rates have been found to be similar whether women were treated with a laparoscopic approach or an open approach, although treatment-related morbidity, e.g. wound breakdown, blood loss, infection, appears to be lower when laparoscopic techniques are used. Currently, prospective randomized trials are underway to determine overall survival in women treated laparoscopically compared with those treated by an open approach.

Vaginal hysterectomy Occasionally an abdominal hysterectomy is not possible due to severe co-morbidities that prevent safe administration of general anaesthesia. Morbid obesity can also preclude TAH/ BSO in some women. A considerable proportion of deaths in women treated for endometrial cancer, at ﬁve years, are related to these other co-morbidities rather than cancer recurrence. Vaginal hysterectomy alone can be performed with spinal anaesthesia under these exceptional circumstances, with acceptable survival rates and morbidity.

Complications of surgery Complications related to surgery for endometrial cancer may be speciﬁcally related to the surgical procedure itself or may be due to other aspects of the surgical process such as anaesthesia (anaphylaxis, failed intubation) or immobility (thromboembolism, lower respiratory tract infection). Death related to surgery and its complications is rare. At surgery direct damage to intra-abdominal structures may occur. Generally, the risk of intra-abdominal complications is low, although the risk of damage to the bladder

141

Chapter 10: Endometrial cancer

and bowel is higher in women who have had previous surgery. At hysterectomy the ureters are at risk as they cross the pelvic brim in the region of the ovarian vessels, and again as they pass beneath the uterine vessels. Ureteric damage rarely occurs at ‘simple’ TAH/BSO. There is a higher risk of ureteric damage at radical hysterectomy due to the extensive dissection needed. Ureteric injuries are also more common with advanced laparoscopic procedures, although these can be minimized by appropriate training and proper visualization of the ureters. Direct damage to major blood vessels leads to intraoperative haemorrhage and is more likely when dissection occurs around the major vessels, e.g. during lymphadenectomy. General postoperative complications include pulmonary atelectasis, lower respiratory tract infection, ileus, urinary infection and thromboembolism. Infection may occur in the abdominal wound or at the vaginal vault secondary to haematoma. Infection increases the risk of abdominal wound breakdown and occasionally dehiscence may follow. A large number of women with endometrial cancer are at higher risk of intraoperative and postoperative complications by virtue of obesity, diabetes and cardiopulmonary problems such as ischaemic heart disease. The risk of postoperative complications can be reduced by the use of eﬀective thromboprophylaxis and early mobilization. Severe late complications of surgery for endometrial cancer are uncommon. Ureteric ﬁstulae can occur as late complications after radical hysterectomy and are usually secondary to ischaemia. Treatment is by ureteric diversion or re-implantation. Lymphoedema can occur following pelvic lymphadenectomy and the risk is increased if radiotherapy is given postoperatively. The severity of lymphoedema varies but it is diﬃcult to treat and can be physically disabling in severe cases. Massage, compression hosiery and bandaging are used to reduce symptoms.

Radiotherapy in the treatment of endometrial cancer Adjuvant radiotherapy Radiotherapy is used most commonly as adjuvant treatment following surgery. The aim is to treat occult metastases in order to prevent recurrence and improve survival. Radiotherapy can be given as external-beam radiotherapy (teletherapy) or as localized treatment to the vaginal vault (brachytherapy). External-beam radiotherapy (EBRT) is delivered to a pelvic ﬁeld that includes the regional pelvic lymph nodes, the upper vagina and the parametrial tissues. This is achieved by administering radiotherapy to a pelvic ‘box’ in which the sides of the box correspond to bony landmarks. Improvements in radiotherapeutic technology now allow the volume of treated tissue to be modiﬁed according to the clinical circumstances and individual patient anatomy. This is important in maintaining treatment eﬀects while reducing the volume of normal tissue that is exposed to radiation. Treatment is usually given as a sequence of short daily treatments (fractions) over several weeks. A typical dose in the adjuvant setting would be a total dose of 45 Gy given as individual daily fractions of 1.8 Gy over ﬁve weeks. Adjuvant EBRT is given to women with stage IB, grade 3 disease (FIGO 2009 classiﬁcation) as it improves survival in this group. EBRT is given to these women even if they have no lymph node metastases, and therefore lymphadenectomy is not required to inform the need for further treatment in this group (Table 10.2). Adjuvant EBRT reduces the risk of locoregional recurrence but does not improve overall survival in women with low-risk (IA G1/2)

142

Chapter 10: Endometrial cancer

Table 10.2 Evidence-based adjuvant treatment for endometrial carcinoma (FIGO 2009 classiﬁcation)

Disease and patient characteristics IA/grade 1–2 endometrioid carcinoma Age under 60 with:

Recommended adjuvant treatment No adjuvant treatment required

IA/grade 3 endometrioid carcinoma IB/grade 1–2 endometrioid carcinoma Aged 60 or older with:

IA/grade 3 endometrioid carcinoma IB/grade 1–2 endometrioid carcinoma

Discuss risks and beneﬁts of radiotherapy with patient Reduces local recurrence but no improvement in survival

IB/grade 3 endometrioid carcinoma

Adjuvant radiotherapy

Stage II endometrioid carcinoma completely resected with radical hysterectomy (negative nodes)

No adjuvant treatment required

Stage I–II non-endometrioid carcinomaa Incompletely excised stage II disease

Adjuvant radiotherapy

Stage III endometrioid and non-endometrioid carcinoma

Adjuvant chemotherapy followed by adjuvant pelvic radiotherapy

a Women with non-endometrioid cancers and other patients at high risk of distant relapse are being oﬀered entry into the PORTEC-3 clinical trial of radiotherapy chemotherapy.

or intermediate-risk (G1 IB or IA G3) disease. The majority of pelvic recurrences can be cured with radiotherapy in radiation-naive patients. The beneﬁts of adjuvant EBRT radiotherapy in this group of women need to be weighed against the risk of toxicity. EBRT is therefore not used routinely in women at low risk of recurrence. Despite lack of beneﬁt on overall survival, many women at intermediate risk of recurrence still receive adjuvant EBRT in order to reduce local recurrence. Studies are currently under way to determine whether adjuvant vaginal vault brachytherapy alone can reduce the risk of recurrence while reducing the side eﬀects of EBRT. Intracavitary (brachytherapy) treatment delivered directly to the vaginal vault may be used in addition to EBRT where the risk of central pelvic recurrence is high, e.g. stage II disease. A vaginal application system is placed under general or spinal anaesthesia and the radiotherapy delivered over 48 hours using a remote after-loading system with the patient housed in a specially designed lead-lined hospital bedroom. Doses of 20–25 Gy are typically administered to the vaginal vault. Newer high dose rate (HDR) treatment systems are used in some centres.

Radical primary radiotherapy Radical primary radiotherapy is usually given to treat stage III disease where there is vaginal involvement (stage IIIB). Radiotherapy may also be used as primary treatment in the small number of women with endometrial cancer who are unﬁt for any surgical management as a result of severe co-morbidity. Radical radiotherapy may be used to good eﬀect in these women. Combined external-beam radiotherapy and intrauterine brachytherapy can achieve

143

Chapter 10: Endometrial cancer

survival rates of >65% in medically unﬁt patients with low complication rates, although the rate of pelvic recurrence with radiotherapy alone is approximately 20%. Options for management of recurrence are very limited in this situation. Hysterectomy should therefore be performed if at all possible for early-stage disease, in preference to primary radiotherapy.

Palliative radiotherapy Radiotherapy is used for palliation of symptoms in cases of advanced inoperable endometrial cancer, where it may be used in conjunction with chemotherapy. Radiotherapy has a beneﬁcial impact on problematic bleeding and is used for treating isolated symptomatic metastases, e.g. bony metastases.

Complications of radiotherapy Acute side eﬀects with radiotherapy are common, aﬀecting approximately 25% women. These eﬀects include tiredness, frequent bowel movements, diarrhoea and frequency of micturition. These symptoms generally reﬂect the eﬀects of radiotherapy on rapidly dividing cells of normal tissues within the radiation ﬁeld. Radiation proctitis varies in severity from frequent passage of loose stools with abdominal cramps and ﬂatus to the passage of large volumes of bloody diarrhoea. Mild to moderate bowel symptoms can be controlled with dietary modiﬁcation and loperamide or codeine phosphate in most cases, but may require treatment modiﬁcation in severe cases. Leucopenia and thrombocytopenia may occur, particularly where the radiation ﬁeld includes a large area of bone marrow. Radiation cystitis presents with irritative urinary symptoms and, in some cases, frank haematuria. Infection must be considered and treated where present and women are advised to maintain a high ﬂuid intake. Radiation eﬀects on the vagina can lead to stenosis. Following completion of treatment women are encouraged to use vaginal dilators with lubricant to maintain vaginal patency until they feel conﬁdent and comfortable enough to resume intercourse. Most acute side eﬀects of radiotherapy are self-limiting and resolve within 6–8 weeks of completing treatment. Severe late toxicity occurs in a small minority of cases. Approximately 3% of women experience severe late eﬀects that require surgery. These include small bowel obstruction due to stenosis and ﬁstula formation. Chronic eﬀects on bone may lead to insuﬃciency fractures of the pelvic bones, and avascular necrosis of the femoral heads may lead to femoral neck fractures, although this is uncommon. Rarely, malignancy can develop within a previously irradiated ﬁeld. Typically, tumours secondary to radiotherapy are sarcomas that present many years after initial treatment.

Chemotherapy Chemotherapy is used as adjuvant treatment in women with more advanced disease, for example those with involved para-aortic nodes, positive peritoneal cytology or involvement of the uterine serosa. The drugs most commonly used are platinum-based drugs, taxanes and anthracyclines. A combination of cisplatin and doxorubicin is most commonly used, although carboplatin and paclitaxel in combination is also used with good eﬀect and relatively low toxicity. Adjuvant chemotherapy is usually given in addition to adjuvant radiotherapy, as chemotherapy does not reduce the risk of local recurrence. Intensive, combination chemotherapy oﬀers a modest improvement in survival in more advanced disease. Use of intensive drug regimens is limited by signiﬁcant toxicity such as myelosuppression, gastric toxicity and, in the case of the anthracycline doxorubicin, cardiac toxicity. In

144

Chapter 10: Endometrial cancer

more frail, elderly women and those being treated with palliative intent, the toxicity of treatment must be carefully balanced against the likely beneﬁt. There is increasing interest in the use of systemic chemotherapy to treat women with early-stage non-endometrioid tumours, e.g. serous carcinoma. A large randomised trial of adjuvant chemotherapy and radiotherapy in women at high risk of distant relapse has begun recruitment (PORTEC-3). At present however, there is insuﬃcient evidence to support the routine use of adjuvant chemotherapy in this setting.

Hormonal treatment Adjuvant progestogens Although most endometrial carcinomas are associated with excessive endogenous or exogenous oestrogens, adjuvant progestogens oﬀer no survival beneﬁt. They should not therefore be used in this setting.

Hormones in the treatment of recurrent and metastatic disease Systemic endocrine hormonal treatment with high-dose oral progestogens can provide useful palliation in women with metastatic disease and where the patient is unﬁt for other types of treatment. Medroxyprogesterone acetate 200–400 mg daily or megestrol acetate 40–320 mg daily provide good symptomatic relief, and tumour deposits often regress. High-grade tumours often lack progesterone receptors and therefore the beneﬁcial eﬀect may be limited in some cases. Gonadotrophin-releasing hormone (GnRH) analogues are another hormonal treatment option for progestogen-refractory disease. GnRH analogues downregulate receptors in the pituitary gland and the resulting decrease in gonadotrophin levels leads to a fall in circulating oestrogen levels. Both complete and partial responses to treatment are seen with monthly injections of goserelin acetate 3.6 mg, and progression-free survival of several months has been reported.

Recurrent endometrial cancer Although recurrence rates vary according to stage, the overall rate of recurrence for endometrial cancer is reported at 7–18%. Recurrences may occur locally, in the pelvis or at distant sites. The pattern of recurrence varies according to the histological subtype and previous treatment modalities used. Serous carcinomas, for example, tend to relapse in the abdomen while the upper vagina is a more common site for recurrence in women with endometrioid adenocarcinomas. A higher proportion of women experiencing disease relapse following adjuvant pelvic radiotherapy have recurrence at distant sites. Conversely, pelvic recurrence appears to aﬀect a higher proportion of women who receive adjuvant chemotherapy alone for high-risk disease. Symptoms of vaginal recurrence include vaginal bleeding and vaginal discharge. A large, central pelvic recurrence may give rise to pressure symptoms. Recurrence in pelvic lymphatics may present as unilateral leg swelling or neuropathic pain. Disease relapse in the para-aortic region may give rise to new-onset central abdominal pain or back pain. Symptoms from distant metastases are dependent on the aﬀected organ(s). The management of recurrent endometrial cancer depends on the site(s) of recurrence and previous treatments used. Generally, women with pelvic recurrence who have not

145

Chapter 10: Endometrial cancer

previously been exposed to pelvic radiation are treated with radiotherapy. Isolated pelvic recurrence is curable and approximately 65% of women survive for at least ﬁve years after treatment with combined EBRT and vaginal brachytherapy. The cure rate is lower for women with pelvic side wall disease. Surgery is a curative option for previously irradiated women with a single recurrence at the vaginal vault. It is vital that other sites of disease are excluded by appropriate imaging before embarking on surgery. Complete removal of the recurrence with a margin of normal tissue is required to eﬀect a curative procedure. This may require exenteration if the bladder or bowel are involved with the tumour. Exenteration is associated with high morbidity and mortality, a major consideration in this group of women, who frequently have major co-morbidities. Few women will be suitable for this approach. Hormonal treatment with progestogens is the mainstay of treatment for women with multiple or distant recurrences and can provide good palliation with acceptable side eﬀects. Chemotherapy may also be used with palliative intent in women ﬁt enough to tolerate such treatment.

Further reading *

*

146

Acharya S, Hensley ML, Montag AC, Fleming GF. Rare uterine cancers. Lancet Oncology 2005; 6 (12): 961–71. Johnson N, Cornes P. Survival and recurrent disease after postoperative radiotherapy for early endometrial cancer: systematic review

and meta-analysis. BJOG 2007; 114(11): 1313–20. *

NHS Executive. National Cancer Guidance Steering Group. Guidance on Commissioning Cancer Services; Improving Outcomes Guidance in Gynaecological Cancers. London: NHS Executive, 1999.

11 Chapter

Cervical and vaginal cancer Ahmed Ahmed, Li Tee Tan and Mahmood I. Shaﬁ

Cervical Cancer Epidemiology Cervical cancer is the ﬁfth commonest cause of death from cancer in women worldwide (following breast, lung, stomach and colorectal cancers). It is the most common cancer in women from southern Africa and Central America. In the UK, the incidence of cervical cancer has declined by almost one half since 1975 (Fig. 11.1). Over 2700 women are diagnosed with, and over 1000 deaths occur due to, cervical cancer each year. The decline in the incidence of cervical cancer in the UK has been largely attributed to the increased coverage rate by the cervical screening programme, from 40% in 1989 to the current rate of around 80%. It is worth noting, however, that cervical cancer remains the second commonest cancer in women under the age of 35 in the UK. In addition, recent data suggest that the absolute and relative incidence of adenocarcinoma of the cervix is increasing in spite of the fall in the incidence of squamous cell carcinoma. Squamous cell carcinoma and adenocarcinoma of the cervix are both eﬀectively detectable by screening methods. However, the risk reduction associated with screening is considerably less for adenocarcinoma compared with squamous cell carcinoma. There are two recognized peaks in the incidence rate of cervical cancer, one in the 30–39 age group and the second one in the over-80 age group. Multiple sexual partners, early age of ﬁrst intercourse, higher parity, younger age of ﬁrst full-term pregnancy and long-term use of oral contraceptives independently increase the risk of adenocarcinoma and squamous cell carcinoma of the cervix. In contrast, smoking increases the risk of only squamous cell carcinoma. Certain types of human papillomaviruses (HPVs) such as HPV 16 and HPV 18 have been found to be associated with a strong risk for cervical cancer. In the late 1980s, the causal relationship between HPV infection and cervical cancer was established. On entry into cervical cells, HPVs release proteins (called E6 and E7) that are essential for maintenance of HPV viruses inside human cells. These viral products directly bind to, and inactivate, important human proteins (the tumour suppressor proteins p53 and Rb), which prevent cancer formation by monitoring deleterious genetic changes. This inactivation leaves cervical cells vulnerable to unchecked genetic alterations, leading to cancer formation. The identiﬁcation of this mechanism represented a breakthrough in the understanding of causes of cancer in general. Since then, cervical cancer has been recognized as a sexually transmitted disease that arises as a direct consequence of chronic, dormant infection of high-risk HPVs.

Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 11: Cervical and vaginal cancer

Fig. 11.1 Agestandardized (European) incidence rates, cervical cancer, Great Britain, 1975–2005

Rate per 100 000 population

18 16 14 12 10 8 6 4 2

2005

2002

1999

1996

1993

1990

1987

1984

1981

1978

1975

0

Year of diagnosis

Diagnosis Irregular vaginal bleeding, in the form of intermenstrual or postcoital bleeding, represents the commonest mode of presentation of cervical cancer. Some women present with a sero-sanguineous oﬀensive discharge or postmenopausal bleeding. In advanced cases, a typical ﬁrst presentation is of obstructive renal failure. Asymptomatic cervical cancer may be diagnosed by cervical cytology/colposcopy or following incidental hysterectomy for benign disease. If cervical cancer is suspected a complete general examination is performed, including palpation of supraclavicular lymph nodes. Speculum examination may reveal a visible cervical mass, which can have a necrotic and hemorrhagic surface. Bimanual and rectovaginal examinations are helpful in the assessment of local spread. Those cervical cancers detected by cytology and colposcopy in the absence of symptoms tend to be diagnosed at an earlier stage and have better prognosis than those presenting with symptoms. Cervical cytology may show tumour diathesis within the sample, indicating the possibility of malignancy. At colposcopy, cervical cancer may be obvious or may display more subtle changes. These include atypical blood vasculature, raised lesions, true erosions and large lesions that have features of high-grade intraepithelial neoplasia (coarse punctuation and mosaicism). In these larger lesions, the chance of ﬁnding malignancy when biopsy/excision is undertaken increases considerably. Multiple biopsies of suspicious cervical lesions, including the tumour margins, should be undertaken. Samples should be of suﬃcient depth to include the cervical stroma for assessment of invasion. Obtaining samples for cervical cytology in the presence of a cervical mass is inappropriate and could be misleading. In the presence of a cervical mass, a recent negative cytology result does not exclude cancer as unrepresentative sampling of necrotic tissue on the surface of the tumour may have occurred; a biopsy is therefore mandatory when there is a clinical suspicion of malignancy. Another typical diagnostic error occurs when a premenopausal woman attends for cervical cytologic examination, and sampling is deferred because of bleeding which is attributed to menstruation. Such women should be referred for gynaecological assessment if the cervix cannot be visualized satisfactorily on the second attempt because of bleeding.

148

Chapter 11: Cervical and vaginal cancer

Staging Clinical examination and routine investigations The current International Federation of Gynecology and Obstetrics (FIGO) staging (Fig. 11.2) for cervical cancer is based on clinical examination. For optimal assessment of the extent of disease, examination under anaesthesia (EUA) with bimanual rectovaginal examination is performed. The tumour size and the extent of vaginal disease should be carefully assessed and documented. Lateral spread into the parametrium and pelvic side wall is best assessed on rectovaginal examination. Joint assessment with a clinical oncologist is recommended to discuss choice of therapy. Cystoscopy and/or proctoscopy may be indicated, depending on the size and stage of the disease. In addition, patients should have a full blood count, urea and electrolytes and liver function tests.

Radiological examination Although radiological assessment of cervical cancers is now routine practice, the ﬁndings do not alter the FIGO stage of the tumour, which is based on clinical ﬁndings. Magnetic resonance imaging (MRI) is the imaging of choice to evaluate tumour size, parametrial invasion and uterine body involvement. Both computed tomography (CT) and MRI can be used to detect pelvic and para-aortic lymphadenopathy. CT-positron-emission tomography (PET) has been shown to be more sensitive at detecting nodal disease but at present it remains an experimental tool. Chest radiography is most commonly used to exclude lung metastases. Although CT is more sensitive at detecting lung metastases, its use is not cost-eﬀective as the incidence of lung metastases is low, particularly in squamous cell carcinomas.

Pathology Squamous cell carcinoma The classiﬁcation of cervical cancer is best understood in the context of the normal histology of the cervix. The cervix is a ﬁbromuscular organ that connects the vagina inferiorly with the uterine body superiorly. The inner surface of the cervix (cervical canal) is lined by columnar epithelium that is continuous with the squamous epithelium lining the outer vaginal part of the cervix. The zone between the two types of epithelia is called the transformation zone, in which the columnar epithelium is in a dynamic state of diﬀerentiation to mature squamous epithelium. The transformation zone is characterized by an active molecular programme that governs the safe transformation of cells from columnar to squamous types without cancer formation. This physiological process is called metaplasia. HPV infection can interfere with these molecular mechanisms, resulting in cervical intraepithelial neoplasia (CIN) and squamous cell carcinoma. Squamous cell carcinoma is the commonest form of malignant neoplasm of the cervix. It is characterized by cytological abnormalities that occupy the full thickness of the epithelium and breach the basement membrane. Macroscopically, these tumours may appear as ﬁrm irregular masses or as exophytic tumours with necrotic and haemorrhagic surface. Verrucous carcinoma of the cervix represents a rare variant of squamous cell carcinoma that only invades locally and may represent a diagnostic challenge. Macroscopically, these tumours present as large polypoid lesions. The diagnosis may be missed if biopsies are not done or if the biopsy samples are not of suﬃcient depth to diagnose invasion.

149

Chapter 11: Cervical and vaginal cancer

I B1

IA

I B2

I A2

I A1

≤4 cm

IIA 1/2

IIIA

IVA

>4 cm

IIB

IIIB

IVB

Fig. 11.2 Diagrammatic representation of cervical cancer staging.

Adenocarcinoma Adenocarcinomas of the cervix arise from the columnar epithelium lining the endocervical canal. It is estimated that 80% of adenocarcinomas of the cervix are characterized by an eosinophilic cytoplasm and mitotic abnormalities. The commonest type of adenocarcinoma

150

Chapter 11: Cervical and vaginal cancer

is called mucinous adenocarcinoma and is characterized by branching glandular epithelium. Although called mucinous, this tumour type often has no mucin. In contrast, a variant of this type, adenoma malignum, is characterized by high amounts of cytoplasmic mucin. Other rare forms of adenocarcinoma include endometrioid, clear cell and serous carcinomas. Because of the rarity of these tumours, it is important to rule out metastasis from uterine, ovarian or peritoneal carcinomas before the diagnosis of primary cervical cancer is made.

Neuroendocrine tumours These are uncommon tumours representing less than 2% of cervical cancers. The commonest form of these tumours is the small cell carcinoma type, which is histologically indistinguishable from small cell carcinoma of the lung. Positive markers for neuroendocrine diﬀerentiation such as chromogranin and synaptophysin help in establishing the diagnosis of such tumours. These tumours often show widespread local and distant metastasis and have poor prognosis in spite of multimodal therapy. Mixed epithelial and stromal tumours or pure sarcomas of the cervix are extremely rare.

Prognostic variables FIGO stage is one of the most important prognostic variables. Stage for stage, the ﬁve-year survival in women with squamous cell carcinomas of the cervix is approximately 10% greater than for other common epithelial cancers due to the low incidence of distant metastases.

Overall survival by FIGO stage (Table 11.1) Besides the FIGO stage, tumour bulk is another important prognostic factor – stage IB2 tumours >4 cm in diameter have a markedly worse prognosis compared with IB1 tumours. In the latest modiﬁcation of the FIGO staging system, stage IIA disease is now divided into substages based on tumour size as for stage I disease (FIGO 2009 classiﬁcation – see Appendix 3). The nodal status is also an important prognostic variable, although this is not currently included in the FIGO staging system as this is a clinical staging system. As a general rule, the presence of positive pelvic nodes reduces the survival by 50% stage for stage, whereas involvement of para-aortic nodes reduces the survival by 75%. Anaemia is an independent prognostic factor in determining recurrence, progressionfree survival and overall survival in cervical cancer treated with radiotherapy. Anaemia is Table 11.1 Overall survival for cervical cancer by FIGO stage

Stage

Overall survival at ﬁve years (%)

Stage IA

95

Stage IB

80–90

Stage IIA

70–90

Stage IIB

60–70

Stage III

30–50

Stage IV

20

151

Chapter 11: Cervical and vaginal cancer

Table 11.2 Incidence of nodal metastasis by FIGO stage

Stage

Positive pelvic nodes (%)

Positive para-aortic nodes (%)

IA1

0

0

IA1 LVSI +ve

4

0

IA2

5

<1

IA2 LVSI +ve

11

<1

IB

15–20

<5

II

25–30

10–20

III

40–50

20–30

IVA

50–60

30–40

LVSI, lymph-vascular space invasion.

thought to decrease tumour oxygen levels, resulting in resistance to radiotherapy. However, the relationship between anaemia and tumour hypoxia as a cause for resistance to radiotherapy is less deﬁned. It remains unclear whether correction of anaemia prior to radiotherapy can improve clinical outcome. The evidence for other prognostic factors is less well established. Bilateral parametrial disease has a worse prognosis than unilateral disease. Some studies have shown that young age (premenopausal) is a poor prognostic factor, particularly for advanced disease. Adenocarcinomas have a higher propensity for metastatic spread than squamous cell carcinomas, which has shown to aﬀect prognosis in some studies.

Lymph node metastasis and histopathological variables Lymph node metastasis is a signiﬁcant determinant of prognosis independent of tumour stage (Table 11.2). In addition, lymph-vascular space invasion (LVSI) and parametrial extension have been associated with higher incidence of local and distant recurrence and lower overall survival.

Radiological characteristics Tumour volume as identiﬁed by pretreatment MRI has been recognized as an independent prognostic factor in cervical cancer. Dynamic contrast-enhanced MRI oﬀers an opportunity to study physical and functional aspects of cervical tumours that are important determinants of prognosis.

Management Cervical cancer is a highly curable disease. The choice of treatment should therefore not only take into account the extent of the disease and the prognostic variables but also the fertility wishes of the patient and the long-term impact on quality of life. Close liaison with the pathologist and clinical oncologist and detailed counselling of the patient are crucial for determining the appropriate management.

152

Chapter 11: Cervical and vaginal cancer

Early-stage disease Stage IA Women with stage IA1 and IA2 disease have an excellent prognosis. The management of stage IA1 and IA2 is usually surgical. Large loop excision of the transformation zone (LLETZ) or cervical conization are adequate treatments for stage IA1 provided that a complete excision is undertaken and clear margins are demonstrated. In those with stage IA2 disease, more radical surgery (hysterectomy or trachelectomy) may be necessary, particularly if LVSI is present. Recent studies suggest that a more conservative approach may be appropriate in those women believed to have no residual disease following conization. The percentage of tumour histological sections showing LVSI and the total number of LVSI foci (>5) strongly predict time to recurrence in women with early-stage disease independent of the depth of invasion. LVSI also predicts lymph node metastasis in early-stage cervical cancer (stages IA to IIA) independent of the depth of stromal invasion. The presence of involved margins with CIN in cases with early invasive focus necessitates a repeat LLETZ or cervical cone procedure as the risk of recurrence is high, and there may be further residual invasive disease.

Stages IB and IIA Stage IB disease includes any lesions with more than 5 mm depth of invasion. In addition, any visible invasive cervical cancer should be classiﬁed as stage IB regardless of the depth of invasion. The choice of treatment for women with stage IB and stage IIA is surgery or radiotherapy, as there is no diﬀerence in survival or recurrence rate. The side eﬀect proﬁle for the two types of treatment should be discussed with the patient to decide on the most appropriate management choice. Women treated with surgery alone tend to have equivalent quality of life scores compared with women without cancer. In contrast, studies have shown that cervical cancer survivors who were treated with radiotherapy have markedly poorer quality of life scores, and psychological and sexual diﬃculties compared with those treated with surgery. In addition, radiotherapy will cause premature ovarian failure. These factors, combined with the woman’s age and sexual function, should be taken in consideration when deciding the best management. The combination of surgery and radiotherapy does not improve outcome and is associated with higher risk of complications and should, therefore, be avoided whenever possible. Postoperative radiotherapy is indicated if the resection margins are involved with tumour or if the pelvic lymph nodes are involved. Careful planning of management is needed to avoid unnecessary combination treatment. Bulky tumours, more than 4 cm in diameter, may represent a surgical challenge. Careful assessment of operability should be performed and the body mass index of the patient should be taken into account before considering surgery (Fig. 11.3). These tumours are generally treated with chemoradiotherapy. Recent advances in image-guided brachytherapy has allowed higher doses to be given with signiﬁcant improvement in local control while minimising the risk of late toxicity. The classical concept of radical surgery for squamous epithelial tumours necessitates the fulﬁlment of three principles:

153

Chapter 11: Cervical and vaginal cancer

Fig. 11.3 Bulky cervical cancer managed by radical hysterectomy – note the enlarged left paracervical lymph node excised with the specimen.

*

excision of all possible tumour (visible tumour and surrounding safety margins of apparent normal tissue); excision of all tissue in continuity (en bloc);

*

complete excision of regional lymph nodes.

*

Surgical management of cervical cancer follows the same principles. Therefore, the cervix is removed with safety margins that include the uterus superiorly, a cuﬀ of vagina inferiorly and around the cervix, utero-vesical fascia anteriorly, uterosacral ligaments posteriorly and the lateral cervical ligament and parametrial tissue laterally. Note that the fallopian tubes and ovaries do not form part of the cervical tumour margins and their excision is, therefore, not an essential requirement for radical hysterectomy. In addition to radical excision of the tumour, complete pelvic lymphadenectomy including the obturator nodes, external iliac nodes and lower common iliac nodes is performed. Although the principles of surgery have remained the same, the extent of surgical margins and surgical radicality has extensively evolved over the years to minimize morbidity while preserving maximum safety and best outcome. The challenge of radical surgery is the close proximity of important structures such as the ureters, bladder and rectum, to the cervix and the relative lack of surgical space in the depth of the pelvis. A clear understanding of the anatomy of the pelvis is essential for the successful completion of surgery. The classical radical hysterectomy necessitates excision of the lateral cervical ligaments at the lateral pelvic side wall and the uterosacral ligament at its origin from the sacrum. As the ureter runs posteroanteriorly across the lateral cervical ligament, its complete mobilization is necessary to gain access to the lateral cervical ligaments. Extensive dissection can result in devascularization of the ureter and consequent ureteric strictures. This dissection can result in interference with the nerve supply to the bladder, which may lead to long-term voiding diﬃculties and the possibility of lifelong self-catheterization. In addition, to achieve excision of the uterosacral ligament, the rectum should be completely mobilized, resulting in possible damage to its neurovascular supply. This type of surgery has been largely replaced by the modiﬁed radical hysterectomy in which the lateral cervical and the uterosacral ligaments are only partially excised. For this type of surgery, de-rooﬁng of the ureter is suﬃcient to enable access to the medial 2 cm of the lateral cervical ligaments. In addition, opening of the rectovaginal space medial to the ureter is suﬃcient for the excision of the terminal 2 cm of the uterosacral ligaments. This type of surgery considerably decreases postsurgical ureteric, bladder and rectal complications while complying

154

Chapter 11: Cervical and vaginal cancer

Table 11.3 Piver–Rutledge classiﬁcation of hysterectomy for cervical cancer

Type

Description

1

Extrafascial hysterectomy; removal of all cervical tissue

2

Modiﬁed radical hysterectomy; removal of medial 50% of cardinal and uterosacral ligaments; uterine vessels are divided medial to ureter

3

Equivalent to the classical Wertheim–Meigs operation; wide radical resection of the parametrium and paravaginal tissues; ureter dissected completely to bladder entry; uterosacral ligaments divided at origin; cardinals divided at pelvic side wall

4

Ureter divided from pubovesical ligament; superior vesical artery ligated and upper two-thirds of vagina excised

5

More radical procedure with possible bowel, bladder or ureteric dissection

Table 11.4 Querleu–Morrow classiﬁcation of radical hysterectomy for cervical cancer

Type

Description

Type A

Minimum resection of paracervix – this is an extrafascial hysterectomy. The paracervix is transected medial to the ureter but lateral to the cervix. The uterosacral and vesicouterine ligaments are not transected at a distance from the uterus. Vaginal resection is generally at a minimum, routinely less than 10 mm, without removal of the vaginal part of the paracervix (paracolpos)

Type B

Transection of the paracervix at the ureter – partial resection of the uterosacral and vesicouterine ligaments, ureter is unroofed and rolled laterally, permitting transection of the paracervix at the level of the ureteral tunnel. At least 10 mm of the vagina from the cervix or tumour is resected

Type C

Transection of paracervix at junction with internal iliac vasculature system – transection of the uterosacral ligament at the rectum and vesicouterine ligament at the bladder. The ureter is mobilized completely; 15–20 mm of vagina from the tumour or cervix and the corresponding paracolpos is resected routinely, depending on vaginal and paracervical extent

Type D

Laterally extended resection – rare operations feature additional ultra-radical procedures. The most radical corresponds to the laterally extended endopelvic resection (LEER) procedure

Table 11.5 Querleu–Morrow classiﬁcation of extent of lymph node dissection

Lymph node dissection Level 1

External and internal iliac

Level 2

Common iliac (including presacral)

Level 3

Aortic inframesenteric

Level 4

Aortic infrarenal

with the basic principles of radical surgery. The Piver–Rutledge classiﬁcation (Table 11.3) is widely used, although newer classiﬁcations have been proposed (Querleu–Morrow). The Querleu–Morrow classiﬁcation (Tables 11.4 and 11.5) can be applied to fertilitysparing surgery and can be adapted to open, vaginal, laparoscopic or robotic surgery. Extrafascial hysterectomy could be performed in women with stage IA1 and IA2 disease. The technique involves ligation and cutting of the horizontal part of the uterine artery lateral to the cervix after crossing the ureter. This is in contrast to clamping and cutting the vertical

155

Chapter 11: Cervical and vaginal cancer

(a)

(b)

Fig. 11.4a, b Typical radiation ﬁelds for the treatment of cervical cancer. A four-ﬁeld brick, consisting of opposed anterior and posterior ﬁelds and two lateral ﬁelds, is used to treat the pelvis while excluding the anterior bladder and posterior rectum.

part that runs parallel to the cervix, as undertaken in hysterectomy for benign disease. Excision of the uterine arteries outside the cervical fascia (extrafascial) ensures the complete excision of the cervix and paracervical tissue. The ﬁnding of positive lymph nodes in histology is an indication for radiotherapy. As combining radiotherapy with radical surgery increases the complication rate, every eﬀort should be taken to exclude positive nodes prior to proceeding to radical hysterectomy. Pelvic lymphadenectomy could be done at a preliminary surgical session, either laparoscopically or via open surgery, to exclude involvement of lymph nodes prior to radical surgery. If positive nodes are identiﬁed, chemoradiotherapy rather than radical surgery should be the treatment of choice. An alternative to the two-session approach is to start with lymphadenectomy and frozen section prior to proceeding to radical surgery in the same surgical session. A similar approach using sentinel lymph node assessment can be used to guide the appropriateness of radical surgery. These approaches do not allow for thorough examination of the lymph nodes prior to radical hysterectomy. Complications of this pelvic lymphadenectomy include vascular injury to the external iliac vessels, obturator nerve damage and lymphocyst formation in the lateral pelvic side wall.

Advanced stages (IIB–IVA) For locally advanced disease, surgery is no longer an option as clear resection margins cannot be achieved. The treatment of choice is therefore radical radiotherapy, consisting of a combination of external-beam radiotherapy and intracavitary brachytherapy. Initial externalbeam radiotherapy is given to the whole pelvis to encompass the cervix, uterus, upper vagina, parametria and regional lymph nodes (Fig. 11.4). Considerable tumour shrinkage can be seen during this initial phase of treatment (Fig. 11.5). Brachytherapy is then used to boost the dose to primary tumour while limiting the dose to the bladder and rectum (Fig. 11.6). The addition of cisplatin-based chemotherapy to

156

Chapter 11: Cervical and vaginal cancer

(a)

(b)

Fig. 11.5 Serial magnetic resonance image showing tumour regression during external-beam radiotherapy. (a) Start of external-beam radiotherapy. (b) During week ﬁve of external-beam radiotherapy.

(a)

(b)

Fig. 11.6a, b Intracavitary brachytherapy to boost the dose to the primary tumour. The dose distribution is optimized to cover the target volume (cervix in red and part of the uterus in blue) while minimizing the dose to the normal tissue (yellow = bladder, green = rectum).

157

Chapter 11: Cervical and vaginal cancer

Fig. 11.7 X-ray showing large bowel obstruction due to a radiation-induced stricture.

external-beam radiotherapy has been shown to improve the survival by 10–17%. The most common regimen is single-agent cisplatin 40 mg/m2 given on a weekly basis. A number of factors have been shown to inﬂuence the response of squamous cell carcinomas to radiotherapy. The presence of hypoxia results in increased radio-resistance and patients therefore actively receive transfusion to maintain the haemoglobin above 12 g/dl during radiotherapy. Squamous cell carcinomas also exhibit a phenomenon known as accelerated repopulation, in which the proliferation rate of the tumour cells is increased as a result of exposure to radiation. This has been shown to result in a 1% loss in tumour control for every additional day by which overall treatment time is prolonged. It is therefore important to avoid unplanned gaps during radiotherapy; if such gaps occur, patients are treated twice a day or at the weekends to compensate for the missed treatments. The most common acute side eﬀects during radiotherapy are diarrhoea, cystitis, proctitis and lethargy. The onset of acute side eﬀects is largely determined by the turnover rate of epithelial cells. Typically, acute side eﬀects appear two weeks after commencing radiotherapy and begin to resolve two weeks after completion of radiotherapy. Late side eﬀects tend to manifest after six months after completion of radiotherapy. Typical late side eﬀects include diarrhoea, abdominal pain, rectal bleeding, urinary frequency and urgency, and haematuria. More serious side eﬀects such as strictures (Fig. 11.7) and ﬁstulae occur in approximately 10% of long-term survivors. Radiation insuﬃciency fractures aﬀecting the sacrum and pubic bones can result in severe pain requiring opiate analgesia; fortunately the pain tends to be self-limiting although it can take up to 18 months to resolve. Vaginal stenosis is common after radiotherapy, and active prevention involving the use of vaginal dilators is important. Premenopausal women undergoing pelvic radiotherapy will be rendered menopausal as well as infertile, and hormone replacement therapy is necessary. Postoperative chemoradiotherapy is indicated in women with positive lymph nodes or close resection margins. The combination of surgery with postoperative chemoradiotherapy

158

Chapter 11: Cervical and vaginal cancer

Squamous cell carcinoma or adenocarcinoma

Stage IB1, node negative on MRI

Stage IB1, node positive on MRI

Stage IB2

Stage IIA/IVA

Staging lymph node dissection

Node negative

Node positive

Radical hysterectomy

Chemoradiotherapy

Fig. 11.8 Algorithm for the treatment of cervical carcinoma.

has been shown to increase the incidence of late side eﬀects without improving survival. Careful patient selection is therefore necessary to avoid unnecessary combination treatment. A typical treatment algorithm is shown in Figure 11.8.

Follow-up and recurrent disease Following treatment of cervical cancer, patients are oﬀered clinical follow-up every three months for the ﬁrst two years. Complete clinical examination including speculum and bimanual examination should be performed. The aims of follow-up are to detect locally recurrent disease which is amenable to curative salvage treatment, and to assess treatmentrelated toxicity. In asymptomatic women, the presence of blood on the examining glove should raise the suspicion of early tumour recurrence. However, radiation necrosis typically also causes vaginal bleeding and discharge and further investigation is necessary to diﬀerentiate between the two. Initial assessment involves MRI with EUA and biopsies as indicated. The management of isolated pelvic recurrences is largely dependent on whether the tumour is central or lateral. Exenterative procedures should be considered in selected cases of centrally recurrent disease with the intention of cure. The reported ﬁve-year survival following exenteration is 50%. A thorough evaluation of tumour spread, including CT of the chest and abdomen to exclude lung metastases and para-aortic lymphadenopathy, should be performed. If available, whole-body PET scanning should be combined with CT to exclude

159

Chapter 11: Cervical and vaginal cancer

occult metastases. In the absence of extrapelvic disease the patient should be counselled regarding the associated morbidity/mortality of surgery. Anterior exenteration involves excision of the tumour with the bladder anteriorly and ureteric diversion to an ileal conduit. Posterior exenteration involves removal of the tumour with the pelvic colon and upper part of the rectum and either colostomy or end-to-end anastomosis. Total exenteration is the combination of cystectomy and colectomy along with resection of the tumour. Exenterative procedures are often accompanied by excision of a considerable proportion of the vagina. Reconstruction of the vagina should be considered if sexual function is desirable. The long-term complications of this procedure include ureteric leakage (15%), wound complications (20%), parastomal hernia and sexual dysfunction. Women who have recurrence after primary surgery are usually oﬀered chemoradiotherapy, if this has not been used during primary treatment phase. The reported cure rates are again approximately 50%. Women with recurrent disease in the para-aortic nodes can also be potentially cured by radical chemoradiotherapy, although the chances of cure are only in the region of 20% as the dose is limited by the spinal cord tolerance. Women with other extrapelvic recurrence are incurable and can be oﬀered palliative chemotherapy or local radiotherapy as appropriate.

Special clinical situations Cervical cancer in pregnancy Cervical cancer in pregnancy is an uncommon occurrence. Women usually present with vaginal bleeding. Speculum examination is needed for all women with bleeding in pregnancy. It is important to note that cervical biopsies in pregnancy could be accompanied by severe bleeding. Women with suspicious lesions should be reviewed by a gynaecological oncologist or a trained colposcopist prior to performing a biopsy. If a biopsy is indicated, this is best undertaken in theatre (without the need for general anaesthesia) as there may be considerable associated bleeding. MRI can be used to further assess any conﬁrmed malignancy. Pregnancy has been shown not adversely to aﬀect the outcome of cervical cancer. Treatment decision should follow the same treatment guidelines for managing non-pregnant women. Radical hysterectomy can be performed at the time of caesarean section if the fetus has reached the age of viability. Before 20 weeks, radical hysterectomy could be performed without prior induction of abortion. This needs careful discussion with the woman and her partner, and their wishes should be taken into consideration. Trachelectomy Trachelectomy refers to surgical removal of the cervix, preserving the uterine body. This is termed fertility-preserving surgery and is an alternative to radical hysterectomy in selected cases. The operation is considered in those women under the age of 40 who wish to preserve fertility with a stage IA2 or small (less than 2 cm in maximal diameter) IB1 squamous cell cancers. A trachelectomy may be simple (equivalent to a very large conization procedure, amputating the cervix) or radical (Dargent’s operation), where the parametrial tissue and vaginal cuﬀ are also excised. A cervical cerclage suture is placed at the cervical internal os. A pelvic lymphadenectomy is also undertaken at the same time, which can be performed laparoscopically or extraperitoneally.

160

Chapter 11: Cervical and vaginal cancer

Recurrence and mortality rates following radical trachelectomy compared with radical hysterectomy in comparable circumstances suggest similar outcomes. Approximately, 70% of women wishing to conceive are successful. There is a considerable risk of pregnancy loss and preterm delivery, which women need to be counselled about. Delivery is by caesarean section.

Vaginal cancer Vaginal cancer accounts for 1–2% of all gynaecological malignancies. The majority of these are extensions of tumours arising in other parts of the genital tract, particularly the cervix. Vaginal cancer occurs mainly in the elderly age group with a peak incidence in the sixth decade. The majority of tumours are squamous (80–90%), with adenocarcinoma, adenosquamous and clear cell types accounting for the remainder. The upper vagina is the commonest site of disease. Vaginal germ cell tumours (teratoma or endodermal sinus tumours) and sarcoma botryoides (rhabdomyosarcoma) are rare and found most commonly in infants or children. Staging of vaginal cancer is detailed in Appendix 3. The cancer should not involve the cervix and should not represent metastatic or recurrent disease. Following thorough assessment (similar to cervical cancer), treatment is primarily using radiotherapy, although surgical excision is appropriate in selected cases. Surgery should be considered in the following situations: *

stage I tumours of the upper third of the vagina, particularly posterior wall, can be treated with radical hysterectomy (if uterus in situ), pelvic lymphadenectomy and vaginectomy;

*

small, mobile stage I tumours low down in the vagina can be treated by vulvectomy with inguinal lymphadenectomy;

*

selected bulky lesions which are unlikely to be cured by radiotherapy may be considered for exenteration.

Overall, ﬁve-year survival rates are in the order of 50%. FIGO stage and tumour size are important prognostic indicators. Lower vaginal tumours tend to have a poorer prognosis than those in the upper vagina.

Further reading *

Cancer Research UK. UK cervical cancer incidence statistics. Available at: http://info. cancerresearchuk.org/cancerstats/types/ cervix/incidence/.

*

World Health Organization. WHO fact sheets, fact sheet number 297. Available at: www.who.int/mediacentre/factsheets/fs297/ en/index.html.

161

12 Chapter

Vulvar cancer Peter Baldwin

Pathology The vast majority (≥85%) of malignant vulval neoplasms are squamous cell carcinomas (SCCs). Other malignant skin tumours of the vulva are less common. Melanomas have the second highest frequency, accounting for ~5% of malignant vulval neoplasms. Basal cell carcinomas, tumours of adnexal structures such as the Bartholin’s glands and sarcomas are rare (1–2% for each). Paget’s disease may aﬀect the vulva and can also be associated with an underlying adenocarcinoma. SCCs can be further divided into several histological subtypes, the major ones being ‘keratinizing’, ‘basaloid’ and ‘warty’. These subtypes show preponderance for diﬀerent age groups and have varying associations with human papillomavirus (HPV) infection. Keratinizing tumours are most common, whereas the ‘basaloid’ and ‘warty’ subtypes are classically associated with HPV-positive tumours in younger women. Metastasis occurs by lymphatic embolization, and may occur at an early stage of tumour development (see Fig. 12.2 and Table 12.2 below). An unusual but clinically important histological variant of SCC is the verrucous carcinoma. These tumours are classically exophytic, ‘cauliﬂower-like’ lesions. Microscopically the lesions show acanthosis with a well-diﬀerentiated squamous epithelium and often little in the way of cellular atypia. The margins of these tumours are often described as ‘pushing’. Although invasive, they behave in an indolent fashion. Lymphovascular space involvement is rare and metastatic disease is uncommon.

Aetiology The aetiology of vulval SCC remains an area of debate. Possible aetiological factors include vulval intraepithelial neoplasia (VIN), HPV infection, squamous hyperplasia, lichen sclerosus, smoking and immunosuppression. In broad terms, HPV-dependent and HPVindependent pathways are recognized. Vulval SCC may be preceded by a period of intraepithelial disease or vulval intraepithelial neoplasia (VIN). This relationship is not as well deﬁned as in SCC of the uterine cervix. ‘Basaloid’ and ‘warty’ SCC are typically found in women <60 years and are associated with adjacent undiﬀerentiated (or ‘classical’) VIN in over three-quarters of cases. In common with undiﬀerentiated VIN, the invasive lesions are often HPV positive. However, in contrast to cervical SCCs, infection is often with a solitary HPV type, typically type 16. Although the risk of progression of VIN in women who have undergone treatment for their disease appears low (5–10%), the limited data available suggest that this ﬁgure may be much higher for untreated individuals. Furthermore, detailed histological assessment reveals that approximately one-ﬁfth of excisional specimens of Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 12: Vulvar cancer

VIN III contain foci of occult microinvasion. Women with established VIN II/III therefore require careful follow-up, and surgical excision remains the gold standard treatment, at least for unifocal disease. The impact of concurrent HPV infection on prognosis for women with SCC of the vulva is not clear. The possibility of multicentric HPV-related genital intraepithelial neoplasia should prompt a thorough examination of the lower genital tract and perianal region in any woman with vulval neoplasia. The more common keratinizing SCCs are associated with adjacent diﬀerentiated VIN (usually HPV negative), squamous hyperplasia and epithelial dystrophies, such as lichen sclerosus. The risk of malignancy associated with lichen sclerosus is estimated at 1–5%. HPV infection and lichen sclerosus can coexist and could therefore act synergistically in the development of vulval SCC in some women. There is evidence from case-control studies of an increased risk of vulval SCC for smokers, and this risk is greatly increased in smokers who exhibit HPV 16 seropositivity. Univariate analysis has suggested that survival is signiﬁcantly worse for smokers with vulval SCC. The increasing incidence of immunosuppression following allogenic organ transplantation has highlighted the inﬂuence of host immune responses in the genesis of intraepithelial neoplasia of the lower genital tract. Such patients are often younger and are prone to multifocal, multicentric intraepithelial disease with the attendant risk of carcinoma. More recently, women with the human immunodeﬁciency virus (HIV-1) have been shown to have a considerably increased risk of vulval SCC.

Presentation: symptoms/signs The vast majority of women with vulval carcinoma are symptomatic. The presence of a mass or ulcerated lesion usually provokes rapid referral to the gynaecology department, but many of the other symptoms and signs are relatively non-speciﬁc. Pruritus or a mass is found in at least 50% of cases. Other common symptoms include pain, burning, ulceration, bleeding and discharge. Incidental lesions may be detected during the examination of women with VIN or concurrent anogenital SCCs. Lesions can occur anywhere on the vulva but the majority arise primarily on the labia and more often the labium majora. Despite the high incidence of symptomatic lesions, delay in diagnosis and referral (typically of four to six months) is common in women with vulval SCC. Both patient and clinician may feel awkward about examination of the lower genital tract. Many women will delay presentation due to embarrassment and primary care physicians may delay an examination for similar reasons or because of the practical limitations of examination in the community. Furthermore, the nonspeciﬁc symptomatology described above frequently results in a trial of some form of topical therapy, e.g. antifungals in the ﬁrst instance. Although this is often appropriate, the management plan should include a short-term review following the introduction of the medication to ensure a therapeutic eﬀect has occurred. Lesions that do not respond as expected or are progressive in nature should be biopsied. It must be remembered that genital ‘warts’ in postmenopausal women are often SCCs and should be biopsied.

Investigations Primary tumour Clinically suspicious lesions should be biopsied to provide histological conﬁrmation of the diagnosis and permit the planning of appropriate intervention. The majority of specimens

164

Chapter 12: Vulvar cancer

(a)

(b)

Fig. 12.1 (a) Keyes punch used to obtain vulval biopsy. This is a circular cutting punch available in a variety of sizes but usually one with a diameter of 4–5 mm is used. After inﬁltration with local anaesthetic, the punch is inserted perpendicular to lesion and with rotating motion. (b) Haematoxylin and eosin section from typical invasive squamous lesion.

are ‘incisional’ rather than ‘excisional’ biopsies, and most are obtained under local anaesthesia using a Keyes punch (Fig. 12.1). Local anaesthesia can be achieved by inﬁltration, often using a mixture of local anaesthetic and vasoconstrictive agents. The use of topical anaesthetics (such as EMLA cream) alone is rarely suﬃcient to allow the biopsy to be performed, but makes inﬁltration of this well-innervated area far more acceptable to the patient. Although this technique has largely replaced excision biopsy in the initial diagnosis of vulval neoplasia, care must still be taken when clinically suspicious lesions are subsequently found to be histologically negative on punch biopsy. False negative biopsies of invasive lesions have been reported, and excisional biopsy should therefore be considered where there is a marked discrepancy between clinical and histological specimens obtained by punch biopsy. Wedge biopsy may be preferable in the case of larger vulval tumours. Biopsy samples should be taken from the edge of such lesions to include the interface between normal skin and tumour in order to facilitate pathological interpretation (biopsy samples from the centre of the lesion may show edge-to-edge dysplasia, but evidence of invasion of the adjacent normal epithelium is necessary to diagnose cancer). Women with extensive disease may require examination under anaesthesia to establish the extent of direct invasion into local structures, e.g. urethra/bladder, anus/rectum. Such examinations are usually performed in a multidisciplinary setting by gynaecological oncologists, radiation oncologists and plastic surgeons.

Locoregional lymph nodes In addition to assessment of the primary tumour, patients should be assessed for metastatic disease in the locoregional lymph nodes. Unfortunately clinical assessment of the lymph nodes is inaccurate, missing up to a third of patients with metastatic disease. A variety of imaging techniques have been assessed in this role. Ultrasound is already used during staging of head and neck tumours and this experience may facilitate its introduction for vulval cancer. Normal lymph nodes are typically small (<1 cm), oval in shape and demonstrate an echogenic hilar region with a ‘halo’ of hypoechoic cortex at the periphery. Involved malignant nodes become more spherical and show a long:short axis ratio of <2. The cortical region may become more pronounced, whereas the echogenic hilum may disappear. Doppler ultrasound may show a variation in the normal hilar blood supply. Some expert centres have reported both high sensitivity and high speciﬁcity for the detection of lymph node

165

Chapter 12: Vulvar cancer

metastases with ultrasound. Combining ultrasound with guided ﬁne needle aspiration (FNA) of the largest or most abnormal node may improve the sensitivity of the technique further, and speciﬁcities of up to 100% have been reported. Ultrasound-based techniques are notoriously operator dependent and it remains to be seen whether this technique will transfer well to other units. Both morphological and size criteria for the assessment of lymph nodes have been described for magnetic resonance imaging (MRI). The published data for its use in vulval cancer are sparse. Compared with ultrasound, MRI is more reproducible and may provide topographical information of use in subsequent surgery or radiotherapy. However, ultrasound is cheaper, quicker and more practical when FNA is required, and MRI is contraindicated in certain patients. The use of iron oxide nano-particles for MRI lymphography may further increase the sensitivity and speciﬁcity of MRI in the detection of lymph node metastases. It has been suggested that these particles may even facilitate the detection of micrometastases in otherwise normal nodes. MRI scanning in patients with extensive disease or recurrent carcinoma may be helpful both in demonstrating the involvement of adjacent structures and in detecting pelvic lymph node involvement. Clinically suspicious nodes should be sampled in the outpatient setting using either FNA or ultrasound-guided core biopsy. The accurate identiﬁcation of nodal metastases has a direct inﬂuence on the choice of initial surgery.

Co-morbidity Given the elderly age of most women with vulval SCC, coexisting medical conditions are frequent. All such patients will need the appropriate preoperative investigations and some may beneﬁt from assessment by the anaesthetist attached to the cancer team. The assessment of performance status may provide useful preoperative prognostic information and has been shown to be an independent predictor of survival in women over the age of 80. Early involvement of social services, physiotherapists and occupational therapists will facilitate a more rapid return to premorbid activity for the older woman.

Staging Vulval SCC is staged surgically. In the UK, the International Federation of Gynecology and Obstetrics (FIGO) classiﬁcation system is in routine clinical use – this was revised Table 12.1 FIGO staging for vulval carcinoma (2009) with ﬁve-year survival

FIGO stage

Comments

Five-year survival (%)

Stage I

Tumour conﬁned to vulva

>90

Stage II

Tumour of any size with extension to adjacent perineal structures with negative nodes

81

Stage III

Tumour of any size with or without extension to adjacent perineal structures with positive inguino-femoral lymph nodes

48

Stage IV

Extensive local or metastatic disease

15 Overall 75%

166

Chapter 12: Vulvar cancer

in 2009 (see Appendix 3). SCC of the vulva may also be staged according to the TNM classiﬁcation system. Depth of invasion is measured from the deepest point of tumour to the epithelial–stromal junction of the adjacent most superﬁcial dermal papilla.

Prognosis Five-year survival for patients is directly related to the stage of disease (Table 12.1). Nodal status is also critical in predicting survival. In lymph node-negative disease ﬁve-year survival is typically excellent at 80–90%. However, metastasis to lymph nodes is associated with a substantial fall in this ﬁgure to 40–50%. Survival seems to be directly related to both the size and number of lymph node metastases. Small (<5 mm) intracapsular metastases are associated with an excellent outcome whereas extracapsular disease appears to be an important poor prognostic indicator. Women with fewer than three nodes involved have a low risk of pelvic node disease and a good prognosis (>80% ﬁve-year survival). The rate of contralateral and pelvic lymph node metastasis appears to rise once three or more nodes are involved and consequently the prognosis of this group is poor. Only 20% of women with deep pelvic nodes survive for ﬁve years.

Principles of surgical management Vulval cancer is rare and, in common with other rare diseases, patients beneﬁt from being managed by teams with expertise. Referral to gynaecological cancer centres allows expert pathology review and high-quality decision making to avoid overtreatment as well as undertreatment. Individualization of care is available, with experienced gynaecological surgeons working in conjunction with other subspecialties, in particular plastic surgeons and radiation oncologists. Concentration of this disease in centres allows recruitment of women into important clinical trials and the training of the next generation of gynaecological oncologists. The low incidence of vulval cancer means that further supraregional centralization of services may be appropriate. The treatment of vulval cancer is essentially surgical, although there is a small role for radiotherapy and chemoradiotherapy, which is discussed later. Surgery is usually curative in intent and has two major aims: *

local disease control;

*

the detection and treatment of locoregional lymph node metastases.

Local disease In the treatment of early vulval cancer it is usually possible to perform a wide local excision of the primary cancer rather than a disﬁguring hemi- or total vulvectomy. However, it is important that the primary cancer is excised with adequate margins as this directly aﬀects the risk of local recurrence. A histological excision margin of less than 8 mm is associated with a local recurrence rate of 50%, whereas a histological margin of 10 mm or greater is associated with a very low risk of recurrence. To achieve a histological margin of 10 mm it is usually necessary to take a measured margin of at least 15 mm at wide local excision. This applies in all planes – medial, lateral and deep. If possible the important structures of the vulva (such as the clitoris, urethra) and the anus should be avoided, but not if this compromises the adequacy of margins. A tumour 4 cm in diameter will need excision of an area of skin up to 7–8 cm in diameter. The gynaecological oncologist can mobilize simple rotational

167

Chapter 12: Vulvar cancer

Table 12.2 Lymph node involvement according to depth of invasiona

Depth of invasion (mm)

Positive lymph nodes (%)

<1

0

1–2

7.7

2–3

8.3

3–5

26.7

>5

34.2

a

Pooled data from ten studies (n = 578). Adapted from NHS Executive. Improving Outcomes in Gynaecological Cancer. Available at: www.dh.gov.uk.

cutaneous or fasciocutaneous ﬂaps for small excisions, but will usually be supported by the plastic surgeons for larger myocutaneous ﬂaps. Similarly, if excision of the anus is being considered for a large perineal tumour, then collaboration with colorectal surgery may be needed for abdominoperineal resection. However, in the presence of advanced local disease it must be remembered that distant spread may already have occurred, which may mean that the disease is incurable even with ultra-radical surgery. Detailed imaging of the pelvic nodes with MRI and chest X-ray will help in this assessment. Some centres are now considering chemoradiation either to reduce tumour volume to allow sphincter-sparing surgery or instead of surgery altogether.

Locoregional lymph node disease One of the commonest causes of treatment failure in vulval cancer is under-staging. The risk of lymph node metastasis is related to tumour size and depth of invasion. For all vulval cancers other than stage IA it is essential that regional lymph node metastasis is considered (Table 12.2). The recognition that tumour spread is embolic prompted a move away from en bloc resection of the lymph nodes with the vulval specimen. A ‘triple incision’ approach to radical vulvectomy is now standard, with separate incisions for the node dissection and vulval excision and a subsequent reduction in complications. If the primary tumour is more than 1 cm from the midline, then it is highly unlikely that there will be metastatic spread to the contralateral inguinofemoral nodes, and only the ipsilateral nodes must be assessed. Therefore the standard management of a vulval tumour cancer that invades by more than 1 mm and is laterally positioned is wide local excision of the primary cancer, together with block dissection of the ipsilateral inguinofemoral nodes. If the tumour crosses the midline or is within 1 cm of the midline, both groins need block dissection (Fig. 12.2). Limiting dissection to the superﬁcial inguinal nodes is inappropriate due to high recurrence rates, and block dissection should therefore include the medial deep femoral nodes. If unilateral dissection reveals involved nodes then the standard practice is to return to dissect the contralateral groin (Fig. 12.2). Contralateral metastasis may occur in as many as a quarter of node-positive patients; however, some authors feel contralateral metastases are rare in true lateral stage I and II lesions. Radiotherapy is not routinely used to ‘treat’ the groins in patients with clinically negative nodes, as early series reported high local recurrence rates. However, subsequent evaluation of

168

Chapter 12: Vulvar cancer

Pelvic nodes Contralateral spread possible Bilateral inguinofemoral nodes

Ipsilateral inguinofemoral nodes

Midline tumour (anterior or posterior)

Lateralized tumour

Fig. 12.2 Lymphatic drainage of midline and lateralized lesions. (Vulval illustration courtesy of The Sourcebook of Medical Illustration, Parthenon, 1991.)

the radiotherapy regimens used has shown many of them to be suboptimal by current standards. Although primary surgery remains the most appropriate modality in the treatment of clinically uninvolved nodes, there may be a role for radiotherapy as an alternative to lymphadenectomy in selected cases, e.g. advanced age/co-morbidity. Perhaps unsurprisingly, dual treatment with radical surgery followed by radical radiotherapy for positive nodes has been associated with higher complication rates. Where nodal disease has been identiﬁed clinically and conﬁrmed pre- or intraoperatively, formal lymphadenectomy may be omitted in favour of debulking followed by radiotherapy to the aﬀected groin/hemipelvis. Pelvic lymph node metastases have been reported in ~5–15% of all vulval tumours. However, even for clitoral tumours, pelvic lymph node involvement is unlikely in women without involvement of the inguinofemoral lymph nodes. The rate of pelvic lymph node recurrence increases with the number and size of groin node metastases. If more than three groin nodes are involved then the rate of pelvic lymph node involvement may be as high as 66%. Pelvic lymphadenectomy has been abandoned in favour of the use of radiotherapy to the deep pelvic nodes in appropriate cases.

Stage IA disease – superﬁcially invasive carcinoma of the vulva Patients with tumour with a depth of invasion of <1 mm have a very low risk of lymph node metastasis. Management in these circumstances may safely be limited to excision of the primary lesion with an adequate margin. This circumstance is rarely identiﬁed before excision of the initial vulval lesion.

Complications of surgical treatment The development of the radical surgical approach to treatment by such pioneers as Antoine Basset, Fred Taussig, Walter Stoeckel and Stanley Way vastly improved survival for women with vulval cancer. Since the 1970s, eﬀorts have concentrated on reducing the impact of surgery without compromising survival. En bloc excision of the vulva and inguinal lymph nodes has largely been replaced by the ‘triple incision’ approach to radical vulvectomy. Pelvic lymphadenectomy has been abandoned in favour of radiotherapy for patients with nodepositive disease. Nevertheless, radical vulval surgery still carries a mortality of 1–2% and is associated with considerable morbidity (see Table 12.3).

169

Chapter 12: Vulvar cancer

Table 12.3 Complications of radical vulval surgery. For early-stage disease, lymphadenectomy accounts for the most signiﬁcant short- and long-term complications (shown in bold)

Speciﬁc to procedure Related to vulval excision

Related to lymphadenectomy

General

Wound infection

Wound infection

Urinary tract infection Thromboembolism

Wound breakdown

Wound breakdown (40%)

Cystocele

Lymphocysts (up to 30%)

Rectocele

Lymphoedema (up to 70%)

Introital narrowing

Cellulitis

Sexual dysfunction

Nerve damage

Osteomyelitis Psychological and psychosexual sequelae

Reducing the extent of the vulval excision for early-stage disease has reduced the incidence of vulval wound complications. However, these are still experienced by up to a third of women. The wound should be closed without tension and, where this is impossible using primary closure, simple rotational ﬂaps usually suﬃce. Provided such principles are adhered to, the minor degree of breakdown that may occur can usually be managed conservatively. Surgical morbidity in stage I/II tumours is predominantly related to the lymphadenectomy. Lymphocysts may develop, and breakdown of the groin incisions occurs in up to 40% of cases. Skin ﬂaps are prone to infection and necrosis, which often becomes apparent several days after the procedure. Minimizing the amount of excised skin and avoiding excessive undermining of skin edges will help to reduce wound problems, and closed, vacuum drainage systems are favoured by many surgeons. Although these appear to reduce the risk of lymphocyst formation, the length of time for which the wounds should be drained remains a subject of debate. When breakdown has occurred, negative pressure wound therapy facilitates rapid improvement and promotes granulation. Delayed secondary suturing is often possible after a relatively short interval of 10–14 days following the use of such dressings. The chronic complication of lymphoedema is the most common and problematic of the long-term complications. The incidence of lymphoedema varies from series to series, but may be as high as 70% if all degrees of this condition are considered. Preservation of the great saphenous vein at the time of lymphadenectomy may reduce the risk of lymphoedema. Patients are encouraged to wear support stockings, to sit with the leg elevated whenever possible and to avoid trauma to the aﬀected leg. Early involvement of specialist lymphoedema clinics is to be encouraged. Manual lymphatic drainage by an appropriately trained therapist may be helpful and pneumatic compression devices may be used in established cases not responding to simple measures. Many surgical procedures in gynaecological oncology are associated with sexual dysfunction. The physical eﬀects of radical vulvectomy are obvious, but many women also develop secondary psychological problems after radical surgery. Both patients and staﬀ may ﬁnd diﬃculty in discussing such issues. It is important to acknowledge the eﬀects on the

170

Chapter 12: Vulvar cancer

Fig. 12.3 Scintigram obtained following administration of technetium nanocolloid around primary tumour. Twohour static image (5-minute capture) with gamma ﬂood source delineating gross anatomy. The sentinel lymph node is ringed in red. A secondary node also visible.

patient’s sexuality and to explore concerns and correct misconceptions that she (or her partner) may have. The specialist nurse may be ideally placed to provide the support that the patient with vulval cancer will undoubtedly require. Some women will require more formal psychosexual counselling from appropriately trained specialists.

Sentinel lymph node dissection Given the excellent survival for early-stage disease, eﬀorts have been made to reduce surgical morbidity. The sentinel node has been deﬁned as ‘any lymph node that receives lymphatic drainage directly from the primary tumour’. The histology of the sentinel node (or nodes) should therefore be reﬂective of the nodal status of all other nodes within the lymphatic basin. Since its introduction for penile cancer in the 1970s, sentinel lymph node dissection (SLND) has become commonplace in the management of other tumours, e.g. breast, melanoma. Initial attempts to locate the sentinel lymph node were intraoperative, with peritumoral injection of vital dyes such as methylene blue or patent blue V. Although the false negative rate was low, the detection of an SLN (or SLNs) was not possible in up to a third of cases. The use of technetium-labelled nanocolloid and hand-held gamma detection cameras has increased the sensitivity of SLND considerably and, when combined with the use of patent blue V dye, both the sensitivity and speciﬁcity of the technique have been consistently reported at around 100%. Dynamic preoperative lymphoscintigraphy (Fig. 12.3) aids in the localization of the true SLN(s) and may direct the surgeon to bilateral drainage in medial tumours. Until recently most case series were small, but a large, prospective, multicentre, international study (GROningen INternational Study on Sentinel nodes in Vulvar cancer, GROINSS-V) has provided evidence that the technique is both reliable and safe in early disease. This study involved SLND in 623 groins from 403 patients. The false negative rate was low (2.3%) and the major complications of wound infection/breakdown, lymphoedema and recurrent cellulitis were all notably reduced compared with formal inguinofemoral lymphadenectomy. The technique requires specialist equipment, a functional multidisciplinary team (including medical physicists and nuclear medicine physicians) and there is a learning curve for the surgeon. SLND is most appropriately provided at a regional or supraregional level.

171

Chapter 12: Vulvar cancer

Radiotherapy in advanced vulval cancer In common with other squamous cell carcinomas, vulval SCCs are very sensitive to radiation. External-beam radiotherapy (EBRT) is used to treat deeper structures such as inguinofemoral and pelvic nodes. Brachytherapy can be used to treat the primary lesion, either in isolation or to boost the local dose following EBRT. Unfortunately the vulval skin tolerates radiotherapy extremely poorly and toxicity limits its use. Desquamation, both dry and moist, is an almost universal side eﬀect, but skin necrosis is rare with modern equipment and techniques. The success of chemoradiation in treatment of SCC of the anus has unfortunately not been replicated in vulval cancer, although some centres report beneﬁt from combination therapy. Radiotherapy is most frequently used in the management of involved groin nodes. If there is micrometastasis aﬀecting just one lymph node then adjuvant radiotherapy is not usually recommended, as groin recurrence is unlikely. However, if there is replacement of one or more nodes with tumour, and particularly if there is extracapsular spread, then adjuvant radiotherapy is usually given to reduce the risk of recurrence. Radiotherapy is typically given as EBRT to the inguinal and lower pelvic nodes (45 Gy/25 fractions), although some clinicians have questioned the additional beneﬁt of treating the pelvic nodes. Neoadjuvant radiotherapy (EBRT brachytherapy) may be useful in the treatment of women with ﬁxed inguinal nodes and for those in whom surgical treatment would involve exenteration, where the use of neoadjuvant radiotherapy may facilitate more conservative local surgery.

Palliative treatment Although surgery is usually curative in intent, some women may present with advanced bulky tumours that are unlikely to be curable, but which are causing major symptoms and/or problems with nursing care. These tumours are very diﬃcult to treat with radiotherapy, particularly if the woman is elderly and confused. ‘Palliative’ vulvectomy may be necessary to remove the tumour, although adequate margins may not be attainable. Very large or symptomatic groin nodes may be removed at the same time. For the unﬁt, but cooperative, patient, radiotherapy for possible subsequent groin node disease may be more acceptable rather than the immediate risks and morbidity of block dissection. Individualized treatment plans should be made in discussion with the woman, her family and her carers. Palliative care services should be involved at an early stage. With uncontrolled locally advanced disease, thought should be given to strategies to manage distressing terminal events such as catastrophic haemorrhage.

Recurrent disease The risk of recurrent disease is dependent on the depth of invasion, margin status and the number of lymph node metastases at the time of initial diagnosis. Recurrent disease may be identiﬁed in the vulva or locoregional nodes and is most typically diagnosed within the ﬁrst two years after primary treatment. Institutional data suggest that the trend towards more conservative primary surgery may be associated with higher rates of local recurrence. Repeat excision is the preferred treatment for recurrence on the vulva. Such procedures frequently involve the use of reconstructive techniques to preserve function. In the case of larger lesions a combination of surgery and radiotherapy may be required. In contrast with those with local recurrence, the outlook for women with either lymph node recurrence or distant metastatic disease is extremely poor. Groin node or skin-bridge recurrence may be managed with

172

Chapter 12: Vulvar cancer

radiation and surgery, whereas chemotherapy regimens often including cisplatin may be employed for distant disease. Long-term survival in such cases is rare.

Unusual types of vulval malignancy There are other rare cancers of the vulva, and as with SCCs these need to be discussed with the gynaecological oncology multidisciplinary team.

Verrucous carcinoma Verrucous carcinoma typically occurs in postmenopausal women and is a locally invasive, somewhat indolent tumour. Treatment is by radical surgical excision. Lymphovascular space involvement and lymph node metastasis are rare, and formal lymphadenectomy is typically omitted. Enlarged reactive nodes may coexist with these large tumours and should be appropriately evaluated preoperatively.

Melanoma Vulval melanoma is usually managed by wide local excision alone. The assessment of the groin nodes is predominantly of prognostic rather than therapeutic beneﬁt. If there are clinically suspicious nodes, then these should be biopsied and a block dissection considered if there is metastatic spread. Sentinel node dissection may be considered to provide additional prognostic information for those women with clinically negative groin nodes and tumours of Breslow thickness >2 mm. Five-year survival is typically between 20% and 50%.

Basal cell carcinoma Basal cell carcinoma accounts for ≤2% of vulval cancers, but fortunately carries the same excellent prognosis as basal cell carcinomas at other sites, as this tumour almost never metastasizes. Wide local excision is typically all that is needed to ensure a cure.

Bartholin’s gland carcinoma Carcinoma of the Bartholin’s gland is thought to account for most cases of adenocarcinoma of the vulva. Treatment is essentially the same as for SCC of the vulva.

Further reading *

*

*

Andrews SJ, Williams BT, DePriest PD, Gallion HH, Hunter JE, Buckley SL, et al. Therapeutic implications of lymph nodal spread in lateral T1 and T2 squamous cell carcinoma of the vulva. Gynecol Oncol 1994; 55: 41–6. Hacker NF, Berek JS, Lagasse LD, Leuchter RS, Moore JG. Management of regional lymph nodes and their prognostic inﬂuence in vulvar cancer. Obstet Gynecol 1983; 61: 408–12. Hall TB, Barton DP, Trott PA, Nasiri N, Shepherd JH, Thomas JM, et al. The role of ultrasound-guided cytology of groin lymph

nodes in the management of squamous cell carcinoma of the vulva: 5-year experience in 44 patients. Clin Radiol 2003; 58: 367–71. *

Homesley HD, Bundy BN, Sedlis A, Yordan E, Berek JS, Jahshan A, et al. Prognostic factors for groin node metastasis in squamous cell carcinoma of the vulva (a Gynecologic Oncology Group study). Gynecol Oncol 1993; 49: 279–83.

*

Van der Zee AG, Oonk MH, De Hullu JA, Ansink AC, Vergote I, Verheijen RH, et al. Sentinel node dissection is safe in the treatment of early-stage vulvar cancer. J Clin Oncol 2008; 26: 884–9.

173

13 Chapter

Gynaecological sarcomas John Latimer and Helena M. Earl

Introduction Gynaecological sarcomas are rare, accounting for less than 1% of all gynaecological cancers. However, gynaecological sarcomas are important as their management is often diﬀerent from that of commoner malignancies of the female genital tract. Gynaecological sarcomas are soft tissue sarcomas that arise from mesenchymal tissues such as smooth muscle, connective tissue, ﬁbrous tissue, fat and endothelial cells. Tumours arising from neuroectodermal tissue are also considered in this category.

Low- and high-grade tumours The histological assessment of gynaecological sarcomas requires expert review by specialised pathologists at cancer centres. Immunohistochemistry may be needed to conﬁrm the diagnosis, e.g. CD10 for endometrial stromal sarcoma. It is likely that in the future microarray analysis will move out of the research laboratory into routine clinical use. In addition to identifying the tumour type it is also important to grade the tumour, usually low (G1, relatively well diﬀerentiated) or high grade (G3, relatively poorly diﬀerentiated). Lowgrade tumours tend to recur locally, and can become high grade over time. High-grade tumours also recur locally, but tend to metastasize to distant sites, often soon after initial treatment. Grading of tumours is based on the degree of cytological atypia, evidence of tumour necrosis and the mitotic index (number of mitoses seen per ten high-power ﬁeld).

Leiomyosarcomas of the uterus Leiomyosarcomas are the commonest group of soft tissue sarcomas found in the uterus, and arise in the smooth muscle of the uterine wall. The incidence is 0.64 per 100 000 women each year. Since benign leiomyomas (uterine ﬁbroids) are so common, diagnosing this malignancy can be challenging. Some of these tumours are diagnosed by the histopathologists without pre- or intraoperative warning of malignancy. Some are diagnosed preoperatively on suspicious radiology (magnetic resonance imaging (MRI)/computed tomography (CT)). These tumours are frequently high-grade, aggressive sarcomas, with spread to local lymph nodes, and also via the bloodstream to give rise to lung, liver and bone metastases. Some leiomyosarcomas (LMS) of the uterus are oestrogen/progesterone-receptor positive (ER/PR+), and these are usually the lower-grade tumours. Hormone replacement therapy should be avoided in these patients, and adjuvant trials of hormone manipulation are awaited. Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 13: Gynaecological sarcomas

Endometrial stromal sarcomas Endometrial stromal sarcomas (ESS) are rare and arise in the stromal layer supporting the endometrium, with an incidence of 0.19 per 100 000 women per year. Spread is similar to that of leiomyosarcomas, i.e. to local lymph nodes and in the bloodstream to lung, liver and bone. These tumours have recently been classiﬁed into two categories. The low-grade tumours are frequently ER/PR positive and even when metastatic can be treated eﬀectively by hormone manipulation. These low-grade tumours tend to have a long natural history, and even with metastatic disease women can live for many years with controlled disease. High-grade ESS is usually very aggressive, ER/PR negative and has a poor prognosis.

Ewing’s family of tumours – primitive neuroectodermal tumours Ewing’s family of tumours are morphologically small, blue round-cell tumours (SBRCTs) and include soft tissue Ewing’s sarcoma, primitive neuroectodermal tumours and embryonal rhabdomyosarcomas. SBRCTs, including embryonal rhabdomyosarcomas, account for 4–6% of all malignancies in childhood and young adults but are extremely rare as gynaecological malignancies. They are highly aggressive, grow and spread rapidly, and importantly are extremely sensitive to chemotherapy.

Malignant mixed Müllerian tumours – gynaecological Until recently malignant mixed Müllerian tumours (MMMTs), or carcinosarcomas, were considered as non-gastrointestinal stromal tumour soft tissue sarcomas. However, most uterine carcinosarcomas are monoclonal and are actually metaplastic carcinomas: there is now compelling evidence for this from molecular studies investigating patterns of X chromosome inactivation. The sarcomatous component is derived from the carcinomatous element, which can be considered as the ‘driving force’. However, carcinosarcomas are aggressive tumours, and behave more like ‘sarcomas’ than poorly diﬀerentiated endometrioid-type endometrial adenocarcinoma. MMMTs tend to spread via the lymphatics and also haematogenously. Carcinosarcomas tend to be sensitive to the chemotherapy drugs carboplatin and paclitaxel used in endometrioid ovarian cancer and endometrial adenocarcinoma, and are also sensitive to chemotherapy drugs used more commonly in sarcomas (ifosfamide and doxorubicin). MMMTs, since they contain both malignant epithelial elements as well as components of mesenchymal malignancy, need to be considered as a separate entity.

Other gynaecological sarcomas Germ cell tumours of the ovary and uterus, whether mature or immature teratomas, can undergo sarcomatous transformation in the mesenchymal elements. Treatment of these depends on the predominant cell type. If a mature teratoma (dermoid) contains a malignant sarcoma, then treatment will be dependent on the particular histological type. A malignant germ cell tumour (immature teratoma) is treated using appropriate chemotherapy (see Chapter 14), unless the malignant sarcoma component is predominant in the tumours, when sarcoma chemotherapy should be considered.

176

Chapter 13: Gynaecological sarcomas

Presentation and staging Gynaecological sarcomas present with abnormal vaginal bleeding, or with pain, or pressure symptoms from an enlarging mass in the uterus. The diagnosis should be suspected in the rapidly enlarging ﬁbroid uterus, especially in the postmenopausal woman; approximately 0.5% of ﬁbroids will be an unexpected ﬁnding of sarcoma after hysterectomy for ﬁbroids. Staging of uterine sarcomas is as for adenocarcinomas of the uterus. Approximately half of sarcomas will appear to present at stage I (conﬁned to the uterus), but nearly half of these will be up-staged surgically. Uterine sarcomas tend to spread locally (stage II and III) before spreading via the bloodstream, particularly to the lungs (stage IV).

Preoperative assessment Women who may have gynaecological sarcomas may present via various routes: they may come through reproductive medicine because of intermenstrual bleeding, or because of presumed ﬁbroids; the majority, however, will present to the rapid access gynaecological oncology clinic with postmenopausal bleeding or a pelvic mass. In any case, the initial assessment will include a thorough history, a careful clinical examination and appropriate special investigations. The history should include details of the current complaint, last menstrual period, date and result of the last cervical cytology sample, previous gynaecological surgery, family history of cancer, personal history of cancer, diabetes and other serious illnesses that might aﬀect treatment options, tamoxifen exposure, past obstetric history and social and smoking history. The examination involves a general assessment, including body mass index, blood pressure, urinalysis, cardiovascular system, evidence of metastasis including pleural eﬀusion and Virchow’s node in the left supraclavicular fossa and scars from previous abdominal surgery; the gynaecological examination will include visualization of the vulva, vagina and cervix to exclude a lower genital tract source of any bleeding and a bimanual examination of the uterus and adnexa, as well as a combined rectovaginal examination to assess the presence of disease in the parametria and pouch of Douglas. The initial special investigation is a pelvic ultrasound scan, including the use of a vaginal probe to give clear images of the uterus and adnexa, which cannot be clearly seen on transabdominal imaging alone. In the postmenopausal woman if the endometrium is thickened above 5 mm, is irregular or heterogeneous, or if there is ﬂuid in the cavity then outpatient hysteroscopy and endometrial biopsy will be necessary to exclude a uterine cancer. Similarly if an abnormality of the myometrium is identiﬁed on ultrasound, a hysteroscopy will probably also be necessary, together with MRI (see below). In women thought to have a sarcoma a chest X-ray is necessary to exclude pulmonary metastases. An electrocardiogram (ECG) and routine blood tests (including full blood count, urea and electrolytes, liver function tests and clotting screen) will usually be necessary as part of the preoperative preparation of the patient. As many patients are obese, unﬁt and elderly, an anaesthetic review may also be necessary before surgery. The most important part of the preoperative assessment is careful review of all the clinical information, including histology and imaging, at the gynaecological oncology multidisciplinary team (MDT) meeting at the cancer centre so that an individualized treatment plan can be made. The MDT should consist of, as a minimum, specialist pathologist, surgical gynaecological oncologist, specialist radiologist, medical and clinical (radiation) oncologist,

177

Chapter 13: Gynaecological sarcomas

together with the related specialist nurses and therapeutic radiographers, geneticists, a coordinator and other members of the extended team.

Imaging modalities in assessment Once a sarcoma is suspected on clinical, hysteroscopic or histological grounds, the deﬁnitive imaging modality is MRI. MRI allows accurate measurement of the size of the tumour and degree of involvement; however, the features are non-speciﬁc and may not discriminate between sarcoma and endometrial adenocarcinoma, although sarcomas are more likely to be large and heterogeneous, with areas of haemorrhage and cystic necrosis. MRI may also show evidence of local extension into the parametria and spread to the pelvic peritoneum, as well as to pelvic and para-aortic lymph nodes. MRI cannot reliably diﬀerentiate between a leiomyosarcoma and a ﬁbroid undergoing degeneration. CT may on occasion be needed to image the upper abdomen and chest.

Surgical management Gynaecological sarcomas should be treated as sarcomas and not as carcinomas. The management will depend on whether disease is conﬁned to the uterus or metastases are present.

Disease conﬁned to uterus For disease conﬁned to the uterus, at least a total hysterectomy is required, frequently removing the ovaries and lymph nodes; however, unlike carcinomas, if there is no obvious extrauterine disease, metastasis to the ovaries or lymph nodes is relatively rare, so routine oophorectomy and lymphadenectomy may not be necessary, depending on the clinical situation.

Metastatic disease present A total hysterectomy is required, together with excision of the metastatic disease, if feasible, followed by adjuvant chemotherapy (see below). Excision of metastases is only likely to be appropriate if there are only one or two sites and the disease appears to be relatively nonaggressive. The prognosis of even stage I leiomyosarcoma is poor, with a ﬁve-year survival of just 50%; the survival of higher-stage disease is even worse. The main risk factors for relapse, whether local or at a distant site, are tumour stage, grade, size and mitotic count. Radiotherapy has been used to try and reduce the risk of local recurrence. There is conﬂicting evidence that this may be eﬀective, but published series include carcinosarcomas, which are more sensitive to radiotherapy than true sarcomas. In any case no overall survival beneﬁt has been shown from radiotherapy.

Role of adjuvant treatment – radiotherapy The use of radiotherapy in the adjuvant setting, particularly for early-stage disease, is controversial. Unfortunately there are no large randomized trials of adjuvant radiotherapy for gynaecological sarcomas. There are some retrospective studies which show beneﬁt in terms of local control. Doses of 40–50 cGy are used to the pelvis for adjuvant treatment. In leiomyosarcoma of the uterus there are some small studies showing beneﬁt and some showing lack of it for adjuvant pelvic radiotherapy in terms of disease-free and overall

178

Chapter 13: Gynaecological sarcomas

survival. An ongoing European Organisation for Research and Treatment of Cancer (EORTC) trial is looking at adjuvant radiotherapy in stage I and II high-grade uterine sarcomas, as these are the most controversial.

Role of adjuvant treatment – chemotherapy Since gynaecological sarcomas are rare, evidence for adjuvant chemotherapy mostly comes from the larger studies of soft tissue sarcomas, which include gynaecological sarcomas, and from the meta-analysis of these studies. The meta-analysis shows an advantage of adjuvant anthracycline-based chemotherapy in terms of disease-free survival (10%, which was statistically signiﬁcant), but overall survival beneﬁt was only 4%, and which was not signiﬁcant. An EORTC trial examining the role of adjuvant doxorubicin and ifosfamide in soft tissue sarcomas, presented at the American Society of Clinical Oncology (ASCO) meeting in 2007, did not show a survival beneﬁt. Retrospective studies have shown no signiﬁcant beneﬁt with chemotherapy in early-stage resected uterine sarcomas. Several studies in carcinosarcoma have shown a beneﬁt to ifosfamide- and cisplatin-containing chemotherapy, although there have been no comparisons without treatment.

International clinical trials in the adjuvant setting for rare diseases Gynaecological sarcomas and MMMTs of the uterus and ovary are ‘rare’ diseases and therefore, in order to carry out adjuvant trials of chemotherapy or radiotherapy, international collaborative trials need to be developed to give women with these tumours the beneﬁt of good evidence for adjuvant treatment after eﬀective surgery. The major gynaecological oncology trial groups internationally should collaborate on research on these soft tissue sarcomas, with enough patient numbers to provide much-needed answers to these important questions.

Role of chemotherapy for metastatic gynaecological sarcomas In the advanced setting, chemotherapy has an established role in the treatment of most uterine sarcomas. Originally the agents used were those that had the highest response rates in soft tissue sarcomas (ifosfamide and doxorubicin). In leiomyosarcomas doxorubicin has provided response rates of 16–25%, but responses are short lived. Gemcitabine has shown activity in leiomyosarcomas, both as a single agent (response rate (RR) 21% in a Gynecologic Oncology Group (GOG) study of 44 women) and combined with docetaxel. The combination of gemcitabine/docetaxel has shown response rates of up to 53% in pretreated leiomyosarcomas, with improvement in overall survival of up to 18 months. Trials are under way to assess the combination as ﬁrst-line treatment. In MMMT trials, the most active single agents have been ifosfamide (39%), cisplatin (19%), paclitaxel (18%) and doxorubicin (10%). The most eﬀective agents used in combination have been cisplatin with ifosfamide (RR 45%), although there are uncontrolled series which show a 35% response rate to carboplatin and paclitaxel.

Uterine sarcomas and endocrine manipulation Low-grade endometrial stromal sarcomas are usually strongly positive for ER and PR, and progestogens can be highly eﬀective both in reducing metastatic disease burden and controlling disease for many years. Since disease control is so good, long-term side eﬀects of

179

Chapter 13: Gynaecological sarcomas

these drugs are seen, with weight gain, risk of venous thromboembolism and hypertension as the more serious side eﬀects. Tamoxifen is contraindicated, as there is a risk of stimulating growth of a previously well-controlled tumour. Women with this diagnosis can live for many years with controlled disease. When progestogens give rise to serious side eﬀects, aromatase inhibitors (AIs) can be used for treatment. The preferred AI for gynaecological cancers is letrozole. Low- and intermediate-grade leiomyosarcomas may be ER/PR positive, and endocrine manipulation can eﬀectively control locally recurrent and metastatic disease when ER/PR positive. ER/PR status should now be requested on all uterine sarcomas (including MMMT) to establish whether hormone manipulation might play a part in management.

Further reading *

Chu M, Mor G, Lim C, Zheng W, Parkash V, Schwartz PE. Low-grade endometrial stromal sarcoma: hormonal aspects. Gynecol Oncol 2003; 90: 170–6.

*

Kapp D, Shin JY, Chan JK. Prognostic factors and survival in 1396 patients with uterine leiomyosarcomas: emphasis on impact of lymphadenectomy and oopherectomy. Cancer 2008; 112: 820–30.

*

180

Krasner C. Aromatase inhibitors in gynaecologic cancers. J Steroid Biochem Mol Biol 2007; 106: 76–80.

*

Leitao M, Soslow RA, Nonaka D, Olshen AB, Aghajanian C, Sabbatini P, et al. Tissue microarray immunohistochemical expression of estrogen, progesterone, and androgen receptors in uterine leiomyomata and leiomyosarcoma. Cancer 2004; 101: 1455–62.

*

Sarcoma Meta-analysis Collaboration. Adjuvant chemotherapy for localised resectable soft tissue sarcoma of adults: meta-analysis of individual data. Lancet 1997; 350: 1647–54.

14 Chapter

Gestational trophoblastic disease and ovarian germ cell tumours Neil J. Sebire and Michael J. Seckl

Introduction Gestational trophoblastic disease (GTD) consists of a spectrum of disorders including two premalignant diseases, termed complete and partial hydatidiform mole (CHM and PHM), and three malignant disorders, invasive mole, gestational choriocarcinoma and placental site trophoblastic tumour (PSTT). The last three conditions are also known as gestational trophoblastic tumours (GTTs) or neoplasia (GTN). Both CHM and PHM can develop into invasive moles, choriocarcinoma and PSTT, but the latter two cancers can also develop after any type of pregnancy including term delivery, miscarriage and an ectopic implantation. GTD remains an important disorder for the gynaecologist and other clinicians to recognize, because it is nearly always curable, with preservation of fertility, if treated appropriately. Germ cell tumours of the ovary are also very curable and will be dealt with in a separate section at the end of the chapter.

Hydatidiform moles Epidemiology, origin and pathology The commonest form of GTD is the hydatidiform mole, which occurs in approximately 1–3:1000 pregnancies in the UK. Both CHM and PHM are more common in women who become pregnant when aged <16 or >40 years. Hydatidiform moles are also more common in women who have had previous molar pregnancies; the incidence rises to 1:76 with one previous mole, and to 1: 6.5 with two previous moles. The Far East had a higher incidence of molar pregnancies, in the past, which has now fallen close to that seen in Europe and North America. Interestingly, this parallels the introduction of the Western diet to countries such as Japan. CHM arises when an ovum devoid of maternal nuclear DNA is fertilized either by two sperm or by a single sperm which duplicates its chromosomes to give a diploid complement of DNA. In contrast, PHMs arise when two sperm fertilize an ovum that has retained its nuclear DNA, forming a triploid conceptus. These proliferate into abnormal trophoblast with variable amounts of fetal tissue. The trophoblast forms hydropic villi, which are most obvious in complete moles and macroscopically resemble a bunch of grapes. Microscopically, the dilated villi consist of hyperplastic syncitio- and cytotrophoblast that initially lines oedematous mesenchyme. This breaks down with increasing gestational age to form cisterns. In PHMs these changes are milder and focal, such that macroscopically in Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 14: GTD and ovarian germ cell tumours

spontaneous abortions the diagnosis may not be apparent. Consequently the true incidence of PHM has been underestimated in some series. A vast increase in maternal blood vessel formation is also triggered by molar tissue, resulting in large arteriovenous shunts that facilitate the metastatic spread of the disease. The ability to synthesize human chorionic gonadotrophin (hCG) is preserved in molar tissue, enabling it to be used as an extremely sensitive tumour marker.

Human chorionic gonadotrophin tests in GTD β-hCG assays The family of pituitary/placental glycoprotein hormones includes hCG, follicle-stimulating hormone (FSH), luteinizing hormone (LH) and thyroid-stimulating hormone (TSH). Each hormone comprises an α-subunit, which is common between the family members, and a distinct β-subunit. Consequently, assays to measure hCG are directed against the β-subunit. Many diﬀerent β-hCG assays are available. Some detect intact β-hCG and others are either selective for individual fragments or detect various combinations of fragments. In pregnancy, hCG is usually intact and fragments of β-hCG are not produced, although the β chain is hyperglycosylated during the ﬁrst trimester. However, in cancer, β-hCG may circulate in many diﬀerent forms that can vary in their glycosylation status. Therefore, assays used in patients with GTD and cancer need to be able to detect all forms of β-hCG. The ideal assay should be able to recognize all forms equally well and be suﬃciently sensitive to limit the risk of false negative results. Moreover, the assay should not produce false positive results, as this is well recognized to be associated with unnecessary medical interventions and potentially life-threatening complications. So how good are existing β-hCG assays? Commercially available β-hCG tests are based on the sandwich assay principle and rely on two antibodies that generally target diﬀerent regions of the molecule. These assays are primarily licensed for use in pregnancy detection. However, they are frequently used for monitoring patients with cancer. The assays can produce both false positive and negative results. The false positives occur because there is another molecule (often a heterophile antibody) that sticks both the capture and detection antibodies together. This can usually be avoided by measuring the hCG in urine, as cross-reacting antibodies are large and cannot pass through the renal glomerulus into the urine. Alternatively, if a false positive is suspected then remeasuring the hCG on an alternative assay or serially diluting the serum sample usually resolves the issue. Real hCG will be seen in another assay and will appropriately dilute, whereas a cross-reacting molecule will be negative in another assay and will not serially dilute away. Recent work from our laboratory shows that commercial assays may be particularly prone to false negatives either because they completely fail to detect, or as a consequence of poor sensitivity for, a particular β-hCG isoform. False negatives can potentially result in failure to diagnose disease or early termination of treatment, and so higher relapse rates. In addition to the commercial assays there are also several in-house assays in various centres around the world, which are usually based on a single antibody to capture the hormone on a competitive basis with labelled (often radioactively) β-hCG. These assays may also produce false positive and negative results. Indeed, all types of assays are only as good as the antibodies used. At Charing Cross, we have been very fortunate to have a rabbit polyclonal antibody which is used in a radioimmune assay (RIA) and recognizes all forms of β-hCG equally well and with suﬃcient sensitivity, so that false negatives appear to be

182

Chapter 14: GTD and ovarian germ cell tumours

relatively rare. Moreover, since this RIA is performed on both serum and urine samples that are serially diluted, the risk of false positive results is very low. Currently, we therefore believe that the Charing Cross RIA for β-hCG remains the gold standard assay for use in the management of GTD and other cancers. The assay is sensitive to 1 IU/l in serum and 20 IU/l in urine. New assays designed to detect speciﬁc hCG variants are in development and are leading to further increases in the speciﬁcity and sensitivity of hCG monitoring. Elevation of hyperglycosylated hCG (hCG-H) levels may be speciﬁc to cases of hydatidiform mole that subsequently require chemotherapy. This rise occurs earlier than in conventional hCG and before clinically apparent GTT develops. However, further work is required and, since there are no commercial hCG-H assays available, progress in this area is likely to be slow. Recent data have also shown that proportionately higher levels of free β-hCG fragments are produced in PSTT and that this is a highly sensitive test for discriminating cases of PSTT from choriocarcinoma. However, it may also be elevated in some non-trophoblastic malignancies, so its clinical role is not yet fully determined.

β-hCG as a tumour marker hCG has a half-life of 24–36 hours and is the most sensitive and speciﬁc marker for trophoblastic tissue. However, hCG production is not conﬁned to pregnancy and GTD. Indeed, hCG is produced by any trophoblastic tissue found, for example, in germ cell tumours and in up to 15% of epithelial malignancies. The hCG levels in such cases can be just as high as those seen in GTD or in pregnancy. Therefore, measurements of hCG do not reliably discriminate between pregnancy, GTD or non-gestational trophoblastic tumours. However, serial measurements of hCG have revolutionized the management of GTD for several reasons. The amount of hCG produced correlates with tumour volume so that a serum hCG of 5 IU/l corresponds to approximately 104–105 viable tumour cells. Consequently, these assays are several orders of magnitude more sensitive than the best imaging modalities available today. In addition, hCG levels can be used to determine prognosis. Serial measurements allow progress of the disease or response to therapy to be monitored. Development of drug resistance can be detected at an early stage, which facilitates appropriate changes in management. Estimates may be made of the time for which chemotherapy should be continued after hCG levels are undetectable in serum to reduce the tumour volume to zero. For these reasons hCG is the best tumour marker known.

Presentation, registration and management The commonest presentations of CHM and PHM are threatened or missed abortions, usually at the end of the ﬁrst trimester, with CHM usually presenting earlier than PHM. Intraperitoneal bleeding may also occur, and occasionally this is catastrophic. Other presentations are those related to very high hCG levels such as: toxaemia, hyperemesis, hyperthyroidism and theca luteal cysts, or the ﬁnding of an excessively large uterus for dates. Symptoms related to pulmonary, vaginal and cervical metastasis can also occur, although these may spontaneously disappear following evacuation of the mole. Trophoblastic embolization and disseminated intravascular coagulation are now rarely seen. Patients with a suspected molar pregnancy should undergo pelvic ultrasound, hCG estimation in serum and/or urine and chest radiography. In CHM, the classical ultrasound appearance is described as a snowstorm, although this is not often seen until the second

183

Chapter 14: GTD and ovarian germ cell tumours

trimester of pregnancy. In PHM, occasionally fetal parts may be identiﬁed on ultrasound, in association with an abnormal placenta. In both CHM and PHM vascular ﬂow abnormalities indicative of arteriovenous malformations, with or without endometrial encroachment, are frequently seen with modern colour power Doppler. The chest radiograph may be normal, or show any known pattern of abnormality associated with tumour metastases. The primary treatment for these women is prompt evacuation of the uterine contents using gentle suction dilation and curettage (D&C). Prostanoids to ripen a nulliparous cervix are not recommended as this may induce uterine contractions and trigger trophoblastic embolization. The latter maybe fatal and is probably also the explanation for why there is a higher risk of needing subsequent chemotherapy if medical or other surgical methods are used to evacuate the uterus. In the UK, all women diagnosed as having GTD are registered with one of three centres located in Dundee, Sheﬃeld and the Charing Cross Hospital in London, which together form the UK GTD national follow-up service. These patients and their gynaecologists then receive an information pack and the histological slides are requested for central review. All patients then send two-weekly blood and urine samples to one of the three centres for serial hCG estimations. In the majority of cases any residual molar tissue regresses and the hCG levels return to normal (≤4 IU/l). The rate of fall of hCG after D&C can predict the likelihood of subsequently developing a trophoblastic tumour. If the hCG has fallen to normal within eight weeks of evacuation then marker follow-up can be safely reduced to six months from the date of evacuation. However, in women whose hCG levels are still elevated beyond eight weeks from the date of evacuation, follow-up is currently continued for six months of normal values. This protocol results in only 1:2000 women being missed with malignant change in their mole. So are there any factors that increase the risk of malignant change? In our experience, the use of the oral contraceptive while the hCG is still elevated does increase the proportion of women developing GTN, but curiously, delayed evacuation for example in the second or even third trimester does not appear to increase the malignancy risk. As women who have had a previous mole or GTN are at an increased risk of having a second molar pregnancy, or of reactivating previous disease, hCG levels are measured at six and ten weeks following the completion of each subsequent pregnancy. Failure of the hCG to normalize heralds the development of GTN. A second D&C should only be performed after consultation with a GTD centre as in most cases it will not prevent the need for chemotherapy. Moreover, each D&C carries a risk of uterine perforation, introducing infection and triggering major haemorrhage. The indications for chemotherapy in patients with a CHM or PHM are shown in Box 14.1. HCG values >20 000 IU/l four weeks after evacuation of a mole or rising values in this range at an earlier stage indicate that the patient is at increased risk of uterine perforation or severe haemorrhage. These complications can be life-threatening and the risk can be reduced by starting chemotherapy. About 16% of CHMs and 0.5% of PHMs ultimately require chemotherapy. A full blood count, coagulation screen, group and save, urea, creatinine and electrolytes and liver function tests are performed prior to starting chemotherapy. In addition, women require a Doppler ultrasound examination of the pelvis (Fig. 14.1) to determine the uterine volume and vascularity, and a chest radiograph to look for pulmonary metastases. Computed tomography (CT) – or preferably magnetic resonance imaging (MRI) – of the brain is only required if one or more of the following are present: clinical suspicion of central nervous system (CNS) involvement; pulmonary lesions on chest radiograph; or the patient’s

184

Chapter 14: GTD and ovarian germ cell tumours

Box 14.1 Indications for chemotherapy *

Evidence of metastases in brain, liver or gastrointestinal tract, or radiological opacities >2 cm on chest radiograph

*

Histological evidence of choriocarcinoma

*

Heavy vaginal bleeding or evidence of gastrointestinal or intraperitoneal haemorrhage

*

Pulmonary, vulval or vaginal metastases unless hCG falling

*

Rising hCG after evacuation

*

Serum hCG >20 000 IU/l more than four weeks after evacuation, because of the risk of uterine perforation

*

Raised hCG six months after evacuation even if still falling

Any of the above are indications to treat following the diagnosis of GTD.

Pre-treatment

Post-treatment

Fig 14.1 A pelvic ultrasound showing a large vascular gestational trophoblastic tumour before (left panel) and after chemotherapy (right panel).

prognostic score predicts high-risk disease (see below). Subsequently, provided no intracranial metastases are identiﬁed, cerebrospinal ﬂuid (CSF) is obtained for hCG estimation (a hCG ratio of >1:60 CSF: blood indicates CNS involvement). Other imaging such as CT of the chest or abdomen is rarely required. The information from these staging investigations is then used in the Charing Cross scoring system (Table 14.1) to determine the risk of developing drug resistance to singleagent chemotherapy. To try to improve data comparisons between centres across the world a new International Federation of Gynecology and Obstetrics (FIGO) scoring system has been developed (Table 14.2), which is currently being used in parallel to the Charing Cross system. The two systems are very similar and thus far it is rare for patients to be classiﬁed diﬀerently in terms of their risk of resistance to single-agent therapy. About two-thirds of women who score low risk (<9 on the Charing Cross and <7 on the new FIGO system) can expect to be cured with methotrexate alone. Methotrexate is given intramuscularly, alternating daily with oral folinic acid for one week followed by a week of

185

Chapter 14: GTD and ovarian germ cell tumours

Table 14.1 Charing Cross Scoring system for gestational trophoblastic tumours

Scorea Prognostic factor

0

1

Age (years)

<39

>39

Antecedent pregnancy (AP)

Mole

Abortion or unknown

Term

Interval (end of AP to chemo at CXH in months)

<4

4–7

7–12

>12

hCG IU/L

<103

103–104

104–105

>105

No. of metastases

Nil

1–4

4–8

>8

Site of metastases

Not detected Lungs, vagina

Spleen, kidney

Gastrointestinal tract

Brain, liver

3–5

>5

Largest tumour mass (cm)

2

Prior chemotherapy

Single drug

6

≥2 drugs

a

The total score for a patient is obtained by adding the individual scores for each prognostic factor. Low risk, 0–8; high risk, ≥9. Patients scoring 0–8 receive methotrexate and folinic acid therapy and those scoring ≥9 are treated with etoposide, methotrexate and actinomycin D (EMA)/cyclophosphamide and vincristine (CO). hCG, human chorionic gonadotrophin; CXH, Charing Cross Hospital.

Table 14.2 FIGO prognostic scoring system for gestational trophoblastic tumours

Scorea Prognostic factor

0

1

2

4

Age (years)

<40

≥40

–

–

Antecedent pregnancy (AP)

Mole

Abortion

Term

–

Interval (end of AP to chemotherapy in months)

<4

4–6

7–13

>13

hCG (IU/l)

<103

103–104

104–105

>105

No. of metastases

0

1–4

5–8

>8

Site of metastases

Lung

Spleen, kidney

Gastrointestinal tract

Brain, liver

3–5

>5

Largest tumour mass (cm) Prior chemotherapy

Single drug

≥2 drugs

a

The total score for a patient is obtained by adding the individual scores for each prognostic factor. Low risk, 0–6; high risk, ≥7.

rest prior to recommencing treatment. Chemotherapy shrinks the disease very rapidly, and this predisposes the patient to an increased risk of tumour haemorrhage. The other principal side eﬀects, aﬀecting about 2% of patients, are mucosal ulceration, conjunctivitis and occasionally serositis. For these reasons, patients remain in hospital for the ﬁrst one to three weeks of treatment. Overall, methotrexate therapy is very well tolerated, does not induce alopecia and can be completed at a local health centre.

186

Chapter 14: GTD and ovarian germ cell tumours

Fig 14.2 A graph of human chorionic gonadotrophin (hCG) in a patient with lowrisk disease who was initially treated with methotrexate (MTX) and folinic acid (FA) but was subsequently changed to actinomycin D (ACTD; arrow) due the development of drug resistance, as indicated by the plateau in hCG levels. RX, treatment.

10 000

hCG (mlU/ml)

1000

100

10

Dec 2007

Jan 2008

Feb

Mar

ACT D## STOP RX

ACT D

ACT D

ACT D

MTXFA

MTXFA

MTXFA

MTXFA

Treatment

1

Apr

Date

In about one third of women, the tumour becomes resistant to methotrexate and, if this occurs when the hCG is <300 IU/l, then most can be salvaged by switching to single-agent therapy with actinomycin D given intravenously daily for ﬁve days every two weeks. This treatment is sometimes used as a ﬁrst-line agent in other countries and from nonrandomized data appears to be equally eﬀective to methotrexate. However, it is more toxic, inducing alopecia, oral ulceration, nausea, vomiting and myelosuppression. If the hCG is >300 IU/l when resistance develops then patients are switched to our multi-agent chemotherapy regimen used for high-risk disease (see below). Patients scoring ≥9 (≥7 on the new FIGO score) receive ‘high-risk’ intravenous combination chemotherapy comprising etoposide, methotrexate and actinomycin D (EMA) alternating weekly with cyclophosphamide and vincristine (CO). Acute side eﬀects include myelosuppression, alopecia, peripheral neuropathy and those associated with single-agent methotrexate therapy. This treatment requires an overnight stay in hospital every two weeks and 70% of the patients require granulocyte colony-stimulating factor injections to help maintain a suﬃcient neutrophil count and treatment intensity. Treatment with either low- or high-risk regimens is continued until the hCG has been normal for six weeks. During this time the serum hCG is measured twice a week, so that the tumour response can be closely monitored and appropriate treatment changes made promptly (Fig. 14.2).

Choriocarcinoma Epidemiology, origin and pathology The incidence of choriocarcinoma following term delivery without a history of CHM is approximately 1:50 000. Although choriocarcinoma can arise following any type of pregnancy, CHM is probably the most common antecedent, with an estimated 3% of CHMs

187

Chapter 14: GTD and ovarian germ cell tumours

Pre-treatment

Post-treatment

Fig 14.3 A magnetic resonance scan showing multiple brain metastases with haemorrhage before (left panel) and after central nervous system (CNS) modiﬁed etoposide, methotrexate and actinomycin D/cyclophosphamide and vincristine (EMA/CO) chemotherapy (right panel) in an 18-year-old patient with choriocarcinoma.

developing into choriocarcinomas. In contrast to CHM there are no clear geographical trends in the incidence of choriocarcinoma, but the eﬀect of age remains important. Choriocarcinoma is a highly malignant tumour, which appears as a soft, purple, largely haemorrhagic mass. Microscopically it mimics the appearances of an early implanting blastocyst, with central cores of mononuclear cytotrophoblast surrounded by a rim of multinucleated syncytiotrophoblast and a distinct absence of chorionic villi. The surrounding areas are usually necrotic and haemorrhagic and tumour is frequently seen within venous sinuses. Genetic analysis frequently demonstrates multiple karyotype anomalies, but none as yet are speciﬁc for choriocarcinoma.

Presentation and management Choriocarcinoma following an apparently normal pregnancy or non-molar abortion usually presents within a year of delivery, but in the Charing Cross series the longest interval to date has been 17 years. The presenting features may be similar to hydatidiform moles, with vaginal bleeding, abdominal pain, pelvic mass and symptoms due to a high serum hCG. However, one third of all patients with choriocarcinomas present without gynaecological features, and instead manifest symptoms and signs associated with metastases. Pulmonary, cerebral (Fig. 14.3) and hepatic deposits are most frequent but any site may be involved, including the cauda equina and skin. In these cases lives can be saved by remembering to include choriocarcinoma in the diﬀerential diagnosis of metastatic malignancy presenting in a woman of childbearing age. Although these tumours are highly vascular, excision biopsy of a metastasis should be considered where it can be safely achieved. This not only enables histological conﬁrmation

188

Chapter 14: GTD and ovarian germ cell tumours

of the diagnosis but also permits genetic analysis to prove the gestational nature of the tumour. Thus, if there are only maternal genes and no paternal genes present, the patient has a non-gestational tumour, e.g. an ovarian choriocarcinoma or more rarely an epithelial tumour which has diﬀerentiated into choriocarcinoma. Frequently, however, biopsy is not possible and the diagnosis is made on the clinical history and other investigation ﬁndings. All women with choriocarcinoma should have a CT whole body, MRI brain and pelvis and Doppler ultrasound pelvis. The patients are then scored and treated as described for molar disease.

Placental-site trophoblastic tumour PSTT can develop following a term delivery, non-molar abortion, CHM or PHM. There are currently about 200 recorded cases of PSTT in the literature and so estimates of its incidence are inaccurate. However, a UK-wide analysis of all PSTTs diagnosed up to 2007 has demonstrated that PSTTs comprise about 0.2% of all GTD in the UK. PSTTs are usually diploid, and may be biparental or androgenetic in origin but, if they follow a PHM, they are triploid. They are slow-growing malignant tumours composed mainly of cytotrophoblast with very little syncytiotrophoblast and so produce little hCG. However, they often stain strongly for human placental lactogen (HPL), which helps to distinguish this tumour from carcinomas, sarcomas, exaggerated placental-site reaction and placental nodule. The raised HPL may cause hyperprolactinaemia, which can result in amenorrhoea and/or galactorrhoea. PSTT usually spreads by local inﬁltration, with distant metastasis occurring late via the lymphatics and blood. The behaviour of PSTT is thus quite diﬀerent from other forms of GTD and it is often relatively chemoresistant. Consequently, the mainstay of treatment is hysterectomy when the disease is localized to the uterus. When metastatic disease is present, individual patients can respond and be apparently cured by multi-agent chemotherapy either alone or in combination with surgery. The key prognostic factor predicting outcome on multivariate analysis appears to be the interval between the antecedent-causative pregnancy and starting therapy. Thus, in our 62-women series from the UK, 100% of the 13 women treated more than four years ago have died. In contrast, 98% (48/49) of those treated within four years survived.

Management of speciﬁc complications of GTD Haemorrhage This usually responds to bed rest and chemotherapy, along with adequate blood transfusion. However on occasion uterine packs, selective embolization or laparotomy may be necessary; hysterectomy is rarely required.

Respiratory failure This can be multifactorial due to parenchymal tumour deposits and/ or tumour or clot embolization. In the majority of patients respiratory failure can be adequately managed by increasing the inspired oxygen concentration and early chemotherapy. When embolism is suspected, the patient should also be anticoagulated with heparin. The administration of dexamethasone (8 mg eight-hourly) may prevent deterioration associated with tumour necrosis and oedema. If possible, mechanical ventilation should be avoided as the high airway pressures can trigger fatal pulmonary haemorrhage from metastases.

189

Chapter 14: GTD and ovarian germ cell tumours

Cerebral metastases Chemotherapy in the presence of cerebral metastases is associated with a signiﬁcant risk of intracranial haemorrhage from the tumour. Where an accessible solitary metastasis is present it should be resected, prior to chemotherapy. Occasionally stereotactic radiotherapy is employed for the treatment of isolated, deep-seated lesions. With multiple deposits prophylactic use of dexamethasone may be useful in preventing cerebral oedema due to chemotherapy. In addition, women with overt cerebral metastases receive intrathecal methotrexate fortnightly with each CO, and the dose of methotrexate in EMA is increased, to optimize CNS penetration of the cytotoxic agents. Patients with lung metastases are at high risk of brain metastases, and therefore receive as prophylaxis three doses of intrathecal methotrexate. This is given at the start of courses of low-risk treatment, or with the CO element of highrisk therapy.

Drug-resistant disease In the case of low-risk patients treated with single-agent methotrexate, resistance can be overcome in all cases by changing to the high-risk EMA-CO regimen, or single-agent actinomycin D in some patients. Patients who become resistant to EMA-CO may be salvaged with EP-EMA, where EP stands for etoposide and cis-platinum, or with paclitaxel and cisplatin alternating every two weeks with paclitaxel and etoposide (TP-TE). The latter regimen is much less toxic than EP-EMA, but a randomized trial is required to determine whether it is as eﬀective. Selected patients may beneﬁt from high-dose chemotherapy with autologous bone marrow or peripheral stem cell rescue. New anti-cancer agents such as pemetrexed and bevacizumab (Avastin) are also under evaluation in GTN. A strategy of debulking surgery in an attempt to excise drug resistant tissue, followed by chemotherapy for residual disease, may also salvage some patients. In this context tumour localization may be aided by whole-body CT or MRI, and by novel imaging techniques such as ﬂuorine-18-ﬂuoro-2-deoxy-d-glucose positron-emission tomography (PET).

Twin pregnancies Although uncommon, more than 70 cases have been observed at our centre of a normal pregnancy in association with a separate CHM. While some elected to terminate their pregnancy early, 56 decided to continue. Although the majority ended in spontaneous abortion, 40% of cases resulted in a live birth, and only 20% of these women subsequently required chemotherapy, with no deaths. Thus, with appropriate counselling these pregnancies can be safely permitted to continue in the absence of any other complications.

Infantile choriocarcinoma This is an extremely rare condition with about 40 reported cases worldwide. Usually the disease presents in the baby within weeks of delivery and is not always coexistent in the mother. The outcome in these infants is often very poor, possibly due to delayed diagnosis, although some have been saved with chemotherapy. We currently advise checking the urine hCG on at least one occasion in all infants born to mothers with choriocarcinoma and conversely to check the mother’s hCG in infants presenting with choriocarcinoma.

190

Chapter 14: GTD and ovarian germ cell tumours

Patient follow-up and prognosis On completion of their chemotherapy, women are advised to: *

avoid pregnancy for one year; use adequate sun-block to minimize the eﬀect of therapy-induced skin photosensitivity;

*

remain on hCG follow-up for life to conﬁrm that their disease is in remission.

*

In addition, women with PSTT may be reviewed regularly in clinic for at least ﬁve years after the completion of treatment, as hCG is not as reliable a marker in PSTT. About 2% of low-risk and 4% of high-risk patients relapse. In our experience all low-risk patients have been salvaged with further chemotherapy (EMA/CO or alternative regimens) and thus this group has a close to 100% cure rate. High-risk patients have an overall 86% ﬁveyear survival rate. Deaths in this group usually occur due to delays in making the diagnosis. Drug resistance is the other main cause of death in high-risk patients, but salvage rates for relapse following EMA/CO therapy can be in excess of 70%. In the long term, EMA/CO therapy increases the risk of second tumours by 1.6-fold compared with the general population, and may expedite the menopause by an average of three years. Low-risk chemotherapy with methotrexate, however, is not associated with any long-term toxicity. Importantly neither treatment aﬀects fertility or rates of fetal abnormality in subsequent pregnancies.

Summary GTD forms a spectrum of illness, from the borderline malignancy of hydatidiform moles to the highly aggressive behaviour of choriocarcinoma, which were fatal in the past. However, in the past 55 years we have learnt much about the biology, pathology and natural history of GTD. Furthermore, accurate diagnostic and monitoring methods have been developed together with eﬀective treatment regimens. As a result, the management of GTD is one of the modern success stories in oncology. Today, with an integrated approach to management, nearly all women are cured from their trophoblastic tumours, with their fertility intact.

Malignant ovarian germ cell tumours Benign and malignant ovarian germ cell tumours (MOGCTs) comprise about 24% and <1% of all ovarian GCTs, respectively. This contrasts with testicular GCT, where 95% of all tumours are GCTs and most of them are malignant. Benign OGCTs, also termed dermoid cysts and mature ovarian teratomas, present at any age but with a peak in the fourth and ﬁfth decades of life, are cured surgically and are not considered further here. MOGCT peaks at around 19 years of age, and in women under 20 there is a more than 25% chance that an ovarian mass will be a MOGCT. Therefore this diagnosis should be high on the diﬀerential of any young woman presenting with a pelvic mass. Indeed, when the diagnosis is considered, fertility-sparing surgery and/or chemotherapy can be instituted thereby sparing fertility while preserving long-term survival.

Pathology GCTs originate from pluripotential germ cells that arise shortly after fertilization and lie mainly in the gonads and the midline structures of the body. The classiﬁcation of GCTs varies but most simply can be considered according to their lines of diﬀerentiation (Box 14.2).

191

Chapter 14: GTD and ovarian germ cell tumours

Box 14.2 Classiﬁcation of GCTs Teratomas (embryonic or somatic tissue): *

Mature teratoma (solid or cystic with mature tissues only such as skin, cartilage, teeth)

*

Immature teratomas (solid with immature tissues present)

Dysgerminomas (germinal epithelium), which are identical to seminomas in the testis (rarely hCG producing) Non-dysgerminomatous or anaplastic GCTs (primitive and extraembryonic tissues). Examples include: *

Embryonal carcinoma (undiﬀerentiated alpha-fetoprotein [AFP] producing)

*

Endodermal sinus or yolk sac tumour (AFP producing)

*

Choriocarcinoma (trophoblast, so hCG producing)

The histological features of MOGCT essentially represent degrees of germ cell diﬀerentiation, ranging from sheets of undiﬀerentiated primitive germ cells (dysgerminoma) to well-formed tissues (malignant teratoma), and many recapitulate various aspects of developing embryonal tissues with some showing variable mixtures of phenotypes in a given lesion (mixed MOGCTs). Lesions are classiﬁed based on the immunohistochemical and morphological features present, with speciﬁc characteristics as follows. The histological classiﬁcation and stage are associated with prognosis and therefore management. Most MOGCTs have no speciﬁc predisposing factors, but it should be noted that phenotypic females with presence of a Y chromosome may develop both gonadoblastoma and MOGCTs in their intra-abdominal gonads.

Dysgerminoma This tumour usually presents as a unilateral solid mass lesion, being predominantly composed of sheets of polygonal monomorphous primitive germ cells with characteristic intervening septae and a prominent associated lymphocytic inﬁltrate. Tumour cells show strong expression of PLAP and CD117 (c-kit). Fifteen per cent are bilateral.

Yolk sac tumour Usually unilateral solid and cystic mass lesions with areas of haemorrhage and necrosis. Yolk sac tumours (YSTs) are histologically composed of malignant epithelial structures with features recapitulating the developing yolk sac, and a wide range of speciﬁc histological patterns are described, many of which coexist in the same tumour. Tumours demonstrate areas of epithelial structures with clear cytoplasm, nuclear atypia and cytoplasmic PASpositive globules and distinctive Schiller–Duval bodies. Speciﬁc patterns include polyvesicular vitelline, solid, glandular, reticular, hepatoid based on morphology. All variants essentially express cytokeratin and variable α-fetoprotein (AFP) immunostaining.

Embryonal carcinoma This tumour is composed of sheets of large, carcinoma-like epithelial cells, which express PLAP and CD30 with scattered areas showing AFP positivity. In addition, multinucleate cells expressing hCG are often seen.

192

Chapter 14: GTD and ovarian germ cell tumours

Non-gestational choriocarcinoma This is a rare variant and is almost always present in association with other malignant elements as part of a mixed MOGCT. Tumours are haemorrhagic and histologically show areas of cytotrophoblast- and syncytiotrophoblast-like cells, which strongly express hCG. Apparently ‘pure’ ovarian choriocarcinomas must be distinguished from gestational choriocarcinoma using molecular techniques.

Malignant mixed germ cell tumours This group represents tumours with more than one germ cell element, as least one of which is malignant. Commonly, this represents dysgerminoma with yolk sac elements or germ cell tumour of any type with choriocarcinomatous areas. In such cases, metastatic disease may be of any type and therapy should be based on the most malignant elements.

Immature teratoma This group represents any teratoma which contains areas of immature neuroepithelial tissue. Such tumours may also contain small foci or frankly malignant elements, most commonly yolk sac tumour, but even in the absence of any additional malignant component, clinical behaviour and prognosis is related to grade and stage, with some stage I tumours requiring no further treatment post-resection, while other cases may spread within the abdomen and/ or metastasize. Ovarian immature teratomas are graded according to the amount of immature elements present.

Other rare subtypes Very rarely, other malignancies may develop within a GCT. These include superimposed somatic tumours such as squamous cell carcinoma within a teratoma and rare monodermal teratoma subtypes such as carcinoid tumour and ovarian neuroectodermal tumours. In clinical practice, any mixture of the above entities can occur and only mature teratomas are considered benign. Immature teratomas, particularly when grade 2 or 3 and especially when >stage I, should be regarded as malignant or at the very least of having malignant potential.

Presentation of MOGCT Most MOGCTs present as non-speciﬁc abdominal pain or an abdominal/pelvic mass in a young woman. On imaging assessment, most appear as mixed solid/cystic ovarian lesions, but such investigations cannot reliably distinguish between benign and malignant lesions. Tumours producing hormones, such as embryonal carcinoma or choriocarcinoma, may also show additional features such as precocious puberty in younger girls and irregular vaginal bleeding or amenorrhea and breast tenderness in older patients. The tumours may also present with symptoms and signs related to distant metastasis in the chest and elsewhere. Rarely, dysgerminomas can present with hypercalcaemia.

Investigations and initial surgery In any girl or young woman presenting with a pelvic mass, it is sensible to request tumour markers including hCG, AFP, cancer-associated antigen (CA)125 and lactate dehydrogenase

193

Chapter 14: GTD and ovarian germ cell tumours

(LDH) while requesting other routine blood tests. Most non-dysgerminomatous MOGCTs produce hCG and/or AFP and a smaller proportion of dysgerminomas secrete hCG. The tumours are also variably associated with elevated serum levels of LDH and CA125, which are non-speciﬁc tumour markers that if elevated in the absence of hCG and AFP can still provide value in monitoring the response to therapy or in surveillance for relapse. Imaging with Doppler ultrasound and MRI of the pelvis will often reveal a complex ovarian mass with an abnormal vascular ﬂow, but imaging features cannot be relied on to distinguish benign disease from malignancy. Before embarking on surgery, a chest and abdomen CT is very helpful to exclude metastatic disease. In the presence of lung metastasis an MRI brain should be performed to exclude CNS disease. The role of PET imaging is still unclear in this disease setting. For patients with disease apparently conﬁned to the ovary, surgery to remove the aﬀected ovary provides the opportunity for histological examination and may be curative. Laparoscopic approaches, while less invasive, are not suitable for large tumours because these often require morcellation in situ before they can be removed through the small portholes. This process, even if carried out within bags that are tightly sealed, cannot exclude the possibility of tumour seeding, rendering stage IA tumours as stage IC. The latter require toxic adjuvant chemotherapy whereas the former can be placed on surveillance (see below). We therefore recommend open surgery for larger tumours. The aim is to conserve fertility and generate accurate pathological staging, so removing the ovary intact is essential, together with pelvic and para-aortic lymph node sampling, biopsy of the omentum and any other suspicious lesions and obtaining peritoneal washings. We no longer recommend biopsy of the contralateral ovary unless it appears abnormal, as the pick-up rate for bilateral tumours is small and the subsequent scaring may impair future fertility. Patients with metastatic disease should be discussed with the specialist centre and are usually transferred urgently for further care. Percutaneous biopsies or surgery to conﬁrm the diagnosis in such cases may not be required and can delay the commencement of life-saving chemotherapy.

Management of stage IA/B We have for many years advocated surveillance for both men and women with stage IA malignant GCTs following their initial surgery. This approach avoids exposing young individuals of childbearing age to the short- and long-term toxic eﬀects of adjuvant chemotherapy. Recent work from several centres, including our own, has demonstrated that in boys this strategy is a safe option, with relapse rates around 20% and 25% for seminomas and nonseminomatous GCTs, respectively. Chemotherapy cures virtually all patients who experience relapse, and overall long-term survival runs at around 99%. However, data for females with stage IA MOGCT managed on a surveillance programme are comparatively scarce. Our own results looking at 32 women with stage IA MOGCT over 25 years with a minimum follow-up of 4 years showed a 22% and 36% relapse rate for dysgerminomas and non-dysgerminomas, respectively. Of the 11 patients who experienced relapse nine were cured, with one death not disease related, so the overall disease-speciﬁc cure rate is approximately 97%. Obviously the numbers are small but it would appear that a surveillance programme for stage IA MOGCT is reasonable. Indeed, even very large tumours appear to do well (Fig. 14.4). So how should these patients be followed up? Thus far, all relapses occurred within the ﬁrst one to two years being detected by one or more of the following, including routine tumour markers, laparoscopy and imaging. Visible relapses all occurred within the pelvis or abdomen. Therefore,

194

Chapter 14: GTD and ovarian germ cell tumours

Fig 14.4 A graph showing the fall in α-fetoprotein (AFP) for a patient with a 17-cm mixed yolk sac endodermal sinus tumour, which was stage IA. On surveillance, the human chorionic gonadotrophin (hCG) rose one year later (arrow), subsequently resulting in a healthy term baby.

1000 000

hCG

100 000 10 000 1000 100 10 10 000

AFP

1000 100

Treatment

10

Oct 1995

Jan 1996

Apr Date

Jul

Oct

our current surveillance protocol for stage IA patients involves monthly clinic visits in the ﬁrst year, two-monthly in the second, three-monthly in the third and four-monthly in the fourth, six-monthly in the ﬁfth and sixth years and annually thereafter. Imaging includes a repeat CT chest and abdomen, Doppler ultrasound and MRI pelvis at three to six weeks and then three months after surgery. If this is clear then a repeat laparoscopy is performed. Subsequently, an MRI pelvis and Doppler ultrasound pelvis, para-aortic area and abdomen are done at 6, 12, 18 and 24 months after surgery. Where imaging is diﬃcult with ultrasound then an abdomen MRI or CT is requested. A chest radiograph is performed on alternate visits except when a chest CT has been done. Serum tumour markers (hCG, AFP, LDH and CA125) are assessed twice weekly to six months and then with each clinic visit. Follow-up is currently for life, until we have suﬃcient data to inform when it is safe to discontinue surveillance. Women who present with disease in both ovaries (stage IB) are managed in the same way and if the uterus is conserved then future egg-donated pregnancies are possible.

Management of stage IC/M–IV Patients with stages IC/M–IV all require chemotherapy. Stage IM patients are identiﬁed by tumour markers that fail to return to normal at an appropriate rate following resection of the primary disease, thereby indicating the presence of extraovarian disease. It has been assumed that MOGCTs behave very much like testicular GCTs and, even in cases which relapse after initial therapy, the outlook with salvage therapy is thought to be good. However, our own analysis of 20 MOGCT cases that relapsed following initial chemotherapy for stage IC/M–IV disease shows that only 10% were subsequently cured. This points to a fundamental diﬀerence in the biology of GCTs between men and women and suggests that more aggressive therapy should be used in the initial treatment of patients with metastatic disease. For this reason, we do not recommend using bleomycin, etoposide and cisplatin (BEP) chemotherapy except in early-stage disease. Indeed, at Charing Cross, we only use BEP chemotherapy given

195

Chapter 14: GTD and ovarian germ cell tumours

Pre-treatment

Post-treatment

Fig 14.5 A computed tomography (CT) scan demonstrating a large pelvic mass from an advanced mixed yolk sac endodermal sinus tumour before (left panel) and after (right panel) cisplatin, vincristine, methotrexate and bleomycin (POMB)/actinomycin D, cyclophosphamide and etoposide (ACE) chemotherapy.

three-weekly for three cycles in women with stages IC/M and II disease. Those with more advanced presentations (stages III and IV) have all received a seven-drug regimen comprising cisplatin, vincristine, methotrexate and bleomycin (POMB) for cycles one, two and four, and actinomycin D, cyclophosphamide and etoposide (ACE) for cycles three and ﬁve, with each cycle starting every two weeks. Prior to commencing this, those patients who are very sick are given induction therapy with low-dose etoposide and cisplatin for two days as per the very advanced choriocarcinomas, to limit excessive tumour oedema and deterioration in symptoms. In patients who have not achieved a complete marker response after three cycles but who are still responding, a further POMB and ACE treatment is added. Although large tumours may completely resolve on therapy (Fig. 14.5), some women have persisting masses which may be very cystic. Residual lesions should be removed at the end of treatment as these may still contain active tumour or diﬀerentiated teratoma. The latter can continue to grow (cystic teratoma-growing syndrome) or dediﬀerentiate back into active cancer. The overall cure rate for our patients is 81% at ten years. So are there any prognostic factors? Our recent analysis revealed that elevated hCG and/or AFP at presentation and stage were independent variables predicting long-term outcome. Thus, the ﬁve-year survival rates for patients with normal as opposed to elevated markers were 90% and 50%, respectively, and for stage IC/M compared with IV were 100% and 71%, respectively.

What to do with relapse If there is an isolated lesion then surgical resection is well worth attempting. Other drugs not used in the initial therapy which are active include paclitaxel, gemcitabine, ifosfamide and pemetrexed. Various regimens have been tried by ourselves and other investigators, including gemcitabine, paclitaxel, ifosfamide and cisplatin (gem-TIP), and paclitaxel and etoposide alternating with paclitaxel and cisplatin (TE/TP). While transient beneﬁt is frequently seen with salvage therapies they are usually short-lived. We have attempted high-dose chemotherapy in ﬁve patients and so far only one has achieved a durable remission. The fact that only 10% of our 20 relapses have been salvaged suggests that it would be best to identify poorprognosis patients upfront and give these individuals high-dose chemotherapy after their initial treatment, analogous to the way lymphoma patients are sometimes treated. However,

196

Chapter 14: GTD and ovarian germ cell tumours

there are no data yet to support such an aggressive approach in MOGCT and our ability to subcategorize the worst cases is not yet suﬃciently reﬁned.

Long-term outlook and summary Despite the fears that chemotherapy might induce infertility in patients who have been treated for MOGCT, 75% of these women who attempt to become pregnant are successful. The success rate is not signiﬁcantly higher in the stage IA patients not given any chemotherapy, indicating only a small risk of chemotherapy-induced infertility. The question immediately arises as to why these women should be less successful with subsequent pregnancies than those who have received combination chemotherapy for GTN (83% pregnancy rate). There are several potential explanations: women with GTN have already proved their fertility as they could not have contracted their illness without getting pregnant, while women with MOGCT are usually nulliparous and therefore of unproven fertility; women with MOGCT have all had pelvic surgery whereas most GTN cases have simply had a D&C; women with MOGCT have had one ovary removed whereas women with GTN have retained both ovaries. Other long-term problems include a small increased risk for second tumours and cisplatininduced vascular damage resulting in hypertension. So in summary, the outlook for patients with MOGCT is generally very good, although not quite in the same league as GTN. It is now clear that it does not behave in exactly the same way as testicular cancer and so cannot be managed in an identical fashion. As MOGCTs are rare, we will probably only optimize therapies by centralizing the care in a few centres. This approach has already proved to be very successful in the management of GTD.

Further reading *

Mitchell H, Seckl MJ. Discrepancies between commercially available immunoassays in the detection of tumour-derived hCG. Mol Cell Endocrinol 2007; 260: 310–13.

*

Murugaesu N, Schmid P, Dancey G, Agarwal R, Holden L, McNeish I, Savage PM, et al. Malignant ovarian germ cell tumors: identiﬁcation of novel prognostic markers and long-term outcome after multimodality treatment. J Clin Oncol 2006; 24: 4862–6.

*

*

Ngan S, Seckl MJ. Gestational trophoblastic neoplasia management: an update. Curr Opin Oncol 2007; 19(5): 486–91. Nogales F, Talerman A, Kubik-Huch RA, Tavassoli FA, Devouassoux M. Germ cell

tumours of the ovary and peritoneum. In: Tavassoli FA, Devliee P. (eds) Tumours of the Breast and Female Genital Organs, World Health Organization Classiﬁcation of Tumours. Lyon: IARC Press, Lyon 2003; 163–75. *

Patterson DM, Murugaesu N, Holden L, Seckl MJ, Rustin GJ. A review of the close surveillance policy for stage I female germ cell tumors of the ovary and other sites. Int J Gynecol Cancer 2008; 18: 43–50.

*

Sebire NJ, Seckl MJ. Gestational trophoblastic disease: current management of hydatidiform mole. BMJ 2008; 337: 454–8.

197

15 Chapter

Decision making in oncology Li Tee Tan

Decision making is one of the most complex skills required of an oncologist. However, the oncologist today is fortunate in having a vast array of treatments available to combat cancer. It is incumbent upon him or her to use their specialist knowledge, experience and clinical judgement to identify the treatment, or combination of treatments, that is likely to result in the greatest overall beneﬁt for the individual patient. The overall beneﬁt of treatment is a delicate balance between the successful eradication of tumour for cure or symptom relief, versus the acute and/or long-term side eﬀects of treatment. Patients vary considerably, not only in their disease and performance status, but also in their psychosocial make-up and life priorities. Thus, some patients may be prepared to accept a treatment with signiﬁcant morbidity for a small chance of cure whereas for others the avoidance of permanent side eﬀects is a priority. It is important that the oncologist does not make any assumptions about the patient’s preferences based on generic factors such as age, social status or cultural background. The investigations or treatment you provide or arrange must be based on the assessment you and the patient make of their needs and priorities, and on your clinical judgement about the likely eﬀectiveness of the treatment options. You must not unfairly discriminate against patients by allowing your personal views to aﬀect the treatment you provide or arrange. This includes your views about a patient’s age, colour, culture, disability, ethnic or national origin, gender, lifestyle, marital or parental status, race, religion or beliefs, sex, sexual orientation, or social or economic status. (General Medical Council. Good Medical Practice. London, General Medical Council, 2006.)

Cancer is an emotive diagnosis that is often associated with fear in the minds of many. Faced with a life-threatening illness, patients may be desperate for any treatment that alleviates the prospect of death, even if the chance of success is remote. Oncologists have to acknowledge the fears of the patient with compassion and yet, not compromise their advice by unrealistic emotionalism. The inﬂuence of family and friends can result in conﬂict if it is at variance with the wishes of the patient. Doctors and other healthcare workers also vary in their beliefs, priorities and psychosocial backgrounds and their opinions often have a marked, and sometimes unknowing, inﬂuence on the patient’s decision. It is important that the oncologist is aware of the factors inﬂuencing the decision-making process in the individual patient, including those subtle inﬂuences that may not be explicitly acknowledged. In a study comparing the attitudes to chemotherapy of cancer patients with those of a matched control group from the general public, cancer specialists, general practitioners Gynaecological Oncology, eds. Mahmood I. Shaﬁ, Helena M. Earl and Li Tee Tan. Published by Cambridge University Press. © Cambridge University Press 2010.

Chapter 15: Decision making in oncology

Table 15.1 Respondents (%) accepting intensive and mild treatment by subject groupa

General public

Cancer nurses

General practitioners

Clinical oncologists

Medical oncologists

Patients

Intensive treatment

19.0

13.5

12.4

4.5

20.0

53.1

Mild treatment

35.0

38.9

44.3

27.3

51.7

67.0

1% chance of cure

Prolong life by three months Intensive treatment

10.0

6.0

3.3

0.0

10.2

42.1

Mild treatment

25.0

25.4

27.3

12.6

45.0

53.0

1% chance of symptom relief Intensive treatment

10.0

5.9

2.4

0.0

6.8

42.6

Mild treatment

19.0

26.4

21.2

2.3

11.7

58.7

a

Data source: Slevin et al. BMJ 1990; 300:1458–1460.

and cancer nurses, the participants were asked to assess their willingness to have two hypothetical treatments, with diﬀerent severity of side eﬀects, and to rate the possible beneﬁt that would make acceptance of these treatments worthwhile. The results showed that most cancer patients were willing to accept intensive chemotherapy for a very small chance of beneﬁt compared with subjects without cancer (Table 15.1). Given the complexity of the decision-making process in oncology, it is helpful for oncologists to have frameworks for making decisions that enable them to be consistent in their advice and yet take into account the individual patient.

Intention of treatment The ﬁrst step in the decision-making process is to decide on the aim of treatment. There are three main aims of treatment in oncology: * *

*

radical – this is the primary treatment for achieving cure; adjuvant – this is additional treatment to improve the success of the primary treatment. The beneﬁt is either an improvement in the chance of cure and/or a reduction in the risk of local recurrence. Traditionally, adjuvant treatments are given after the primary treatment but they can now be given before the primary treatment (neoadjuvant) to down-stage the disease or, at the same time as the primary treatment (concomitant), to increase its eﬃcacy; palliative – in this situation, there is no prospect of cure, and treatment is being oﬀered for symptom relief to improve quality of life.

Most oncological treatments are associated with considerable acute and/or late side eﬀects. The therapeutic ratio (beneﬁt versus side eﬀects; Table 15.2) of treatment is therefore highly dependent on the aim of treatment. For radical treatment, the beneﬁt is indisputable (i.e. potential cure versus inevitable death) and most patients would be prepared to accept some degree of permanent side eﬀects to achieve this. However, they will vary in how much

200

Chapter 15: Decision making in oncology

Table 15.2 Therapeutic ratio = Beneﬁt/toxicity

Radical

Beneﬁt

Cost of toxicity

Cure

Subjective, depends on patient

Adjuvant

Improved chance of cure

Majority of patients treated may not beneﬁt

Palliative

Improved quality of life

Toxicity may reduce quality of life

toxicity they will accept for which magnitude of chance of cure. For palliative treatment, the beneﬁt is improved quality of life through reduction of symptoms, but any side eﬀects from treatment will reduce quality of life. In oﬀering palliative treatment, the oncologist has to have evidence that the beneﬁt of treatment outweighs the side eﬀects. This is not always an easy decision, particularly when faced with a patient desperate for any treatment to delay inevitable death. Decision making for adjuvant treatment is probably the most complex, as the beneﬁt is based on population statistics from randomized clinical trials rather than individual outcomes. In this situation, the patient has already undergone radical treatment for cure. He or she then undergoes additional treatment to improve the chance of cure, but they will never know whether they have beneﬁted from this as they could have been cured by the primary treatment alone. However, their risk of developing toxicity is the same as someone who is cured by the adjuvant treatment. In this situation, the way information is presented has a huge inﬂuence on whether the patient chooses to have treatment. The oncologist therefore has the responsibility of communicating these diﬃcult concepts to the patient in a language that they can understand so that they can reach an informed decision about whether to proceed with treatment. It has been shown that the way results are presented can inﬂuence doctors’ judgements about the treatment of patients. One study presented the results from two previously published studies in two diﬀerent ways, once as the relative improvement in outcome (24% and 20.3%) and once as the absolute change in the outcome rate (0.4% and 1.5%) to doctors and asked them how the information presented would inﬂuence their decisions about treatment. Forty-six per cent of respondents gave diﬀerent responses to the same results presented in diﬀerent ways. Of these, 89.8% indicated a stronger inclination to treat patients after reading about the relative change in the outcome.

Which treatment to give? Oncological treatments can be classiﬁed into two groups – local or systemic. The two main local treatments are surgery and radiotherapy. The standard systemic treatments are chemotherapy and hormone treatment, and now also include newer treatments such as monoclonal antibodies, vascular epidermal growth factor inhibitors and tyrosine kinase inhibitors. In general, local disease should be primarily treated with local treatments and systemic disease with systemic treatments (Fig 15.1). For adjuvant treatments, the TNM (Tumour Node Metastases) staging system provides a useful basis for deciding whether local or systemic treatment is indicated. Thus, adverse prognostic factors associated with local spread (e.g. large tumours or close surgical margins) predict the risk of local recurrence, and adjuvant radiotherapy may be indicated. In contrast, adverse prognostic factors associated with haematogenous spread (e.g. high-grade tumours or lymphovascular invasion) predict the risk of metastatic disease, and adjuvant systemic

201

Chapter 15: Decision making in oncology

Box 15.1 Pitfalls of evidence-based medicine *

Conﬂicting trial results

*

Publication bias for positive trials

*

Validity of trial data (small numbers)

*

Application of population statistics to individual patient

*

Statistically signiﬁcant results may not be clinically signiﬁcant

*

Clinical signiﬁcance depends on individual perspective

T

Surgery Radiotherapy

N

M

Fig. 15.1 ‘Local problem, local treatment.’ T, extent of the tumour; N, extent of spread to lymph nodes; M, presence of metastasis.

Chemotherapy Hormones Biological modifiers

therapy may be indicated. Nodal involvement can be a risk factor for both local recurrence and metastatic disease. In some tumours, lymphatic spread proceeds from the ﬁrst-echelon nodes to subsequent nodal regions in a predictable sequence (e.g. external iliac to common iliac to para-aortic nodes for cervical cancer). In this circumstance, radiotherapy may be given to eradicate subclinical disease in the next echelon of nodes beyond known disease. In other tumours, lymphatic spread is unpredictable (e.g. the ﬁrst-echelon nodes for endometrial cancer can be the pelvic, inguinal or para-aortic nodes) and systemic therapy is the more logical adjuvant treatment. In trying to decide on the most appropriate treatment for the individual patient, the oncologist has to take into account tumour factors, treatment factors as well as patient factors. The impact of tumour factors on prognosis is usually well documented in the published literature and is rarely the subject of debate among clinicians. Published evidence for treatment eﬃcacy is also widely available; however, the evidence is often conﬂicting (see Box 15.1) and can be open to clinician interpretation. In contrast, the inﬂuence of patient factors is rarely well documented and assessment of patient factors is highly subjective. A consistent decision-making process should therefore begin with an assessment of the objective factors before moving onto the more subjective factors (Fig. 15.2). For curative treatments, the ﬁrst tumour factor to consider is histology as it can have a major impact on treatment and prognosis. Thus for tumours arising in the cervix, squamous cell carcinomas and adenocarcinomas have a similar prognosis and are treated similarly (surgery or radiotherapy), but melanomas and sarcomas have a poorer prognosis and are primarily treated with surgery as they are relatively radio-resistant. The next tumour factor to consider is disease stage. Thus a stage IIIB squamous cell carcinoma of the cervix has a cure rate of 50% whereas a smaller stage II small cell neuroendocrine carcinoma is unlikely to be cured because of its unfavourable histology. Having assessed the chance of cure of a

202

Chapter 15: Decision making in oncology

TUMOUR HISTOLOGY

Fig. 15.2 Framework for decision making.

Prognosis? TUMOUR STAGE Prognosis?

Curable?

TREATMENT Therapeutic ratio of standard treatment?

Therapeutic ratio of alternative treatment?

PATIENT

particular tumour type of a particular disease stage, the oncologist has to decide on the treatment, or combination of treatments, with the highest therapeutic ratio. This is not always straightforward, as a treatment with a lower chance of cure but less morbidity may have a similar therapeutic ratio to another treatment with a higher chance of cure but more side eﬀects. Finally, the oncologist has to assess whether there are any treatment options that the patient is unable to tolerate, as well as take into account patient preferences.

Clinical examples Example 1 A 28-year-old woman was diagnosed with a 5-cm FIGO Stage IB2 adenocarcinoma of the cervix. She is nulliparous and wishes to preserve her fertility if possible. The decision making for this patient is relatively straightforward. The standard treatment for bulky cervical carcinomas is chemoradiation, as cure is unlikely with surgery alone. The clinician had to counsel the patient that it was not possible to cure her disease while preserving her childbearing potential, although methods of maintaining the possibility of biological oﬀspring are available, e.g. ovarian mobilization or embryo storage. Given this information, the patient decided to proceed with chemoradiation.

Example 2 A 32-year-old nurse was pregnant with her ﬁrst child after successful fertility treatment. During week 14 of her pregnancy, she was found to have an abnormal smear. Large loop excision of the transformation zone (LLETZ) biopsy showed a 10 mm × 7 mm poorly diﬀerentiated squamous cell carcinoma extending to the margins of excision, with lymphovascular invasion. A magnetic resonance image showed no radiological evidence of the primary tumour and no lymphadenopathy. Her disease was staged as FIGO IB1 (occult). The decision making for this patient is complicated as there are conﬂicting issues to consider. The standard treatment for premenopausal patients with early-stage cervical carcinoma is radical hysterectomy with ovarian conservation. There is some evidence that radical trachelectomy is a safe alternative option for women wishing to preserve fertility, although the technique remains under evaluation. Therefore options for this patient were:

203

Chapter 15: Decision making in oncology

1. to proceed with the standard treatment of radical hysterectomy and permanently forego her childbearing potential; 2. to proceed to trachelectomy and risk terminating this pregnancy, bearing in mind that future fertility treatment may not be successful; 3. to delay deﬁnitive treatment until the fetus is viable but there was a risk that her disease would progress in the intervening period. The multidisciplinary team favoured options 2 or 3 but the patient decided to proceed with option 1. In this case, the clinicians had assumed that maintaining fertility would be the patient’s priority as she was undergoing fertility treatment, but the patient decided that optimal tumour cure was paramount.

Example 3 A 78-year-old woman was diagnosed as having FIGO stage II clear cell carcinoma of the vagina. The standard curative treatment for stage II vaginal carcinomas is chemoradiation, as the vast majority of vaginal cancers are squamous cell carcinomas, which are radiosensitive, and surgery involves radical vaginectomy, which is a relatively complex procedure. Clear cell carcinomas are more commonly seen in the ovary, where they are known to be relatively chemo- and radio-resistant. The decision was therefore taken to treat this patient with radical vaginectomy, with postoperative radiotherapy, despite her advanced age. In this case, the clinicians used basic principles to work out the best treatment for this patient with a rare tumour.

Example 4 A 62-year-old woman was diagnosed as having FIGO stage IBG2 adenocarcinoma of the endometrium (FIGO 2009 classiﬁcation). She has longstanding problems with irritable bowel syndrome. The standard postoperative adjuvant treatment for stage IBG2 endometrial cancer in women aged >60 years is external-beam radiotherapy to the pelvis. Randomized studies have shown that radiotherapy reduces the risk of local recurrence (estimated to be 15–20% in this patient) but does not improve survival. The risk of serious permanent bowel or urinary toxicity after pelvic radiotherapy is 10%. This patient’s risk of serious permanent bowel toxicity is likely to be much higher in view of her pre-existing bowel problems. The clinician therefore advised her that the therapeutic ratio of radiotherapy is too small in her case to be worthwhile. However, the patient decided to proceed with radiotherapy despite the high risk of toxicity. She subsequently developed permanent troublesome bowel toxicity with grade 2 diarrhoea, faecal incontinence and recurrent bouts of colicky abdominal pain. Despite this, she maintained that radiotherapy has been worthwhile as she was disease-free at ﬁve years, even though her disease would probably have remained in remission without radiotherapy.

Example 5 A 66-year-old woman is diagnosed as having FIGO stage IIG3 adenocarcinoma of the endometrium (FIGO 2009 classiﬁcation). She lives 50 miles away from the radiotherapy centre. In this patient, the risk of local recurrence is estimated to be 40–50% without adjuvant pelvic radiotherapy. However, the patient was not willing to undertake daily journeys for ﬁve weeks to the radiotherapy centre, even with hospital transport, and also declined inpatient

204

Chapter 15: Decision making in oncology

admission. She was therefore given vault brachytherapy alone, as there is some nonrandomized evidence that it is nearly as eﬀective as whole-pelvis radiotherapy in reducing the risk of vault recurrence. In this case, the clinicians found an alternative strategy for the patient even though her reasons for declining standard treatment may seem ‘trivial’.

Summary Decision making in oncology can be complex and the oncologist is required to take into account a number of factors regarding the tumour, the treatment and the patient. Systematic decision-making frameworks are useful in helping the oncologist make consistent decisions when faced with conﬂicting issues in trying to decide on the best treatment for an individual patient.

205

Appendix 1: Miscellaneous information WHO performance scale 0 Able to carry out all normal activity without restriction. 1 Restricted in physically strenuous activity, but ambulatory and able to carry out light work. 2 Ambulatory and capable of all self-care, but unable to carry out work; up and about more than 50% of waking hours. 3 Capable only of limited self-care; conﬁned to bed more than 50% of waking hours. 4 Completely disabled; cannot carry out any self-care; totally conﬁned to bed or chair.

WHO criteria of response * *

*

*

Complete response (CR) – resolution of all measureable or evaluable disease. Partial response (PR) – greater than or equal to 50% reduction in measureable or evaluable disease in the absence of progression in any particular disease site. Stable disease (SD) – less than 50% decrease or less than 25% increase in measureable or evaluable disease. Progressive disease (PD) – greater than 25% increase in measureable or evaluable disease or development of a new lesion.

Clinical trials and drug development Phase I: Aim is to establish human toxicity of new drug; to establish the safe dose at which to start further trials and to measure the pharmacokinetics of the drug. Phase II: Aim is to establish the antitumour activity of the drug against a particular tumour in patients in whom no curative therapy is possible, further information on toxicity is obtained. Phase II studies are usually non-randomized. Phase III: Aim is to compare the new drug with the conventional best therapy. Trials are prospective, randomized and controlled. Large number of patients are required to show eﬀect. Phase IV: Aim is to establish eﬃcacy of drugs in adjuvant trials; determine long-term toxicity of drug.

Appendix 2: Incidence and mortality data Breast Lung Colorectal Prostate Non-Hodgkin’s lymphoma Bladder Melanoma Stomach Oesophagus Pancreas Kidney Leukaemias Uterus Ovary Oral Brain with CNS Multiple myeloma Liver Cervix Mesothelioma Other

Male

0

Female

10 000 20 000 30 000 40 000 Number of new cases

50 000

The ﬁgure shows the 20 most commonly diagnosed cancers in the UK in 2005. Breast cancer is by far the commonest cancer for women in the UK. Of the gynaecological cancers, the uterus has become the most common and the incidence continues to increase in the UK. Cervical cancer incidence continues to decrease in the UK. (Source: Cancer Research, UK.)

Appendix 2: Incidence and mortality data

Lung 20%

Other 29%

Uterus 2% All leukaemias 2% Stomach 3% Non-Hodgkin’s lymphoma 3% Oesophagus 3% Pancreas 5% Ovary 6%

Breast 17%

Colorectal 10%

All malignant neoplasms – 73 621

The ten most common causes of cancer death, females, UK, 2006. (Source: Cancer Research, UK.) All malignant neoplasms Malignant melanoma Prostate Mesothelioma Uterus Liver Oral Kidney Breast Non-Hodgkin’s lymphoma Multiple myeloma Oesophagus Pancreas Leukaemia Colorectal Brain + CNS Ovary Lung Cervix Stomach Bladder –50.0

Female Male

–30.0 –10.0 +10.0 +30.0 +50.0 % change in incidence rates

Percentage change in the age-standardized (European) incidence rates, by sex, major cancers, UK (1996–2005). There has been a marked increase in the incidence of uterine cancer, while the incidence of ovarian and, particularly, cervical cancer has decreased. (Source: Cancer Research, UK.)

210

Appendix 2: Incidence and mortality data

–60

–40

–20

0

+20

+40

All cancers Bladder Bone and connective tissue Brain + CNS Breast Cervix Colorectal Kidney Leukaemia Liver Lung Malignant melanoma

Multiple myeloma Non-Hodgkin’s lymphoma Oesophagus Oral Ovary

Male Female

Pancreas Prostate Stomach Uterus

Percentage change in the European age-standardized mortality rates by sex, major cancers, UK (1997–2006). There has been an increase in age-standardized mortality for uterine cancer and ovarian cancer. However, ovary cancer and cervical cancer especially has seen a decrease in mortality rates. (Source: Cancer Research, UK.)

211

Appendix 3: Staging: FIGO and TNM Rules for FIGO clinical staging Clinical staging should be based on careful examination and should be done before any deﬁnitive therapy. It is desirable that the examination be performed by an experienced examiner and under anaesthesia when appropriate. Surgicopathological ﬁndings are adapted for staging of corpus, ovary, fallopian tube, vulvar cancer and trophoblastic disease. The clinical stage, under no circumstances, must be changed on the basis of subsequent ﬁndings. When it is doubtful as to which stage a particular case should be allotted, the case must be referred to the earlier stage. For clinical staging purposes the following examination methods are permitted: palpation, inspection, colposcopy, endocervical curettage, hysteroscopy, cystoscopy, proctoscopy, intravenous urography and X-ray examination of the lungs and skeleton. Suspected bladder or rectal involvement should be conﬁrmed by biopsy and histological evidence. Findings by examinations such as computed tomography, magnetic resonance imaging, lymphangiography, arteriography, venography, laparoscopy, etc. are of value for the planning of therapy, but, because these are not generally available and also because the interpretation of results is variable, the ﬁndings of such studies should not be the basis for changing the clinical staging. The staging for endometrial, cervical and vulval carcinoma was revised by FIGO in 2009 and is included in this appendix. Uterine sarcomas are staged separately in this 2009 revision and are included.

TNM staging system The TNM system is one of the most commonly used staging systems. This system has been accepted by the International Union Against Cancer (UICC) and the American Joint Committee on Cancer (AJCC). Most medical facilities use the TNM system as well as the FIGO staging for cancer reporting. The TNM system is based on the extent of the tumour (T), the extent of spread to the lymph nodes (N) and the presence of metastasis (M). A number is added to each letter to indicate the size or extent of the tumour and the extent of spread. Primary tumour (T) TX

Primary tumour cannot be evaluated

T0

No evidence of primary tumour

Tis

Carcinoma in situ (early cancer that has not spread to neighbouring tissue)

T1, T2, T3, T4

Size and/or extent of the primary tumour

Appendix 3: Staging: FIGO and TNM

Regional lymph nodes (N) NX

Regional lymph nodes cannot be evaluated

N0

No regional lymph node involvement (no cancer found in the lymph nodes)

N1, N2, N3

Involvement of regional lymph nodes (number and/or extent of spread)

Distant metastasis (M) MX

Distant metastasis cannot be evaluated

M0

No distant metastasis (cancer has not spread to other parts of the body)

M1

Distant metastasis (cancer has spread to distant parts of the body)

FIGO staging for carcinoma of the ovary Staging of ovarian carcinoma is based on ﬁndings at clinical examination and by surgical exploration. The histological ﬁndings are to be considered in the staging, as are the cytological ﬁndings as far as eﬀusions are concerned. It is desirable that a biopsy specimen be taken from suspicious areas outside of the pelvis.

Stage 0

214

No evidence of primary tumour

Stage I

Growth limited to the ovaries

Stage IA T1a N0 M0

Growth limited to one ovary; no ascites present containing malignant cells; no tumour on the external surface; capsule intact

Stage IB T1b N0 M0

Growth limited to both ovaries; no ascites present containing malignant cells; no tumour on the external surfaces; capsules intact

Stage ICa T1c N0 M0

Tumour classiﬁed as either stage IA or IB but with tumour on the surface of one or both ovaries; or with ruptured capsule(s); or with ascites containing malignant cells or with positive peritoneal washings

Stage II

Growth involving one or both ovaries, with pelvic extension

Stage IIA T2a N0 M0

Extension and/or metastases to the uterus and/or tubes

Stage IIB T2b N0 M0

Extension to other pelvic tissues

Stage IICa T2c N0 M0

Tumour either stage IIA or IIB but with tumour on the surface of one of both ovaries; or with capsule(s) ruptured; or with ascites containing malignant cells present or with positive peritoneal washings

Stage III

Tumour involving one or both ovaries with peritoneal implants outside the pelvis and/or positive retroperitoneal or inguinal nodes. Superﬁcial liver metastasis equals stage III. Tumour is limited to the true pelvis but with histologically proved malignant extension to small bowel or omentum

Appendix 3: Staging: FIGO and TNM

Stage IIIA T3a N0 M0

Tumour grossly limited to the true pelvis with negative nodes but with histologically conﬁrmed microscopic seeding of abdominal peritoneal surfaces

Stage IIIB T3b N0 M0

Tumour of one of both ovaries with histologically conﬁrmed implants of abdominal peritoneal surfaces, none exceeding 2 cm in diameter; nodes are negative

Stage IIIC T3c N0 M0 Any T N1 M0

Abdominal implants >2 cm in diameter and/or positive retroperitoneal or inguinal nodes

Stage IV Any T Any N M1

Growth involving one or both ovaries, with distant metastases. If pleural eﬀusion is present, there must be positive cytological ﬁndings to allot a case to stage IV. Parenchymal liver metastasis equals stage IV.

Notes on the staging: To evaluate the impact on prognosis of the diﬀerent criteria for allotting cases to stage IC or IIC, it would be of value to know whether the rupture of the capsule was spontaneous or caused by the surgeon and if the source of malignant cells detected was peritoneal washings or ascites. a

FIGO staging for carcinoma of the fallopian tubea Stage 0 TIS

Carcinoma in situ (limited to tubal mucosa)

Stage I

Growth limited to fallopian tubes

Stage IA T1a N0 M0

Growth limited to one tube with extension into submucosa and/or muscularis but not penetrating serosal surface; no ascites

Stage IB T1b N0 M0

Growth limited to both tubes with extension through or onto tubal serosa or with ascites containing malignant cells or with positive peritoneal washings

Stage IC T1c N0 M0

Tumour either stage IA or stage IB with extension through or onto tubal serosa or with ascites containing malignant cells or with positive peritoneal washings

Stage II

Growth involving one or both fallopian tubes with pelvic extension

Stage IIA T2a N0 M0

Extension and/or metastasis to uterus and/or ovaries

Stage IIB T2b N0 M0

Extension to other pelvic tissues

Stage IIC T2c N0 M0

Tumour either stage IIA or IIB and with ascites containing malignant cells or with positive peritoneal washings

Stage III

Tumour involving one or both fallopian tubes with peritoneal implants outside pelvis and/or positive retroperitoneal or inguinal nodes. Superﬁcial liver metastasis equals Stage III. Tumour appears limited to true pelvis but with histologically proved malignant extension to small bowel or omentum

Stage IIIA T3a N0 M0

Tumour grossly limited to true pelvis with negative nodes, but with histologically conﬁrmed microscopic seeding of abdominal peritoneal surfaces

Stage IIIB T3b N0 M0

Tumour involving one or both tubes with histologically conﬁrmed implants of abdominal peritoneal surfaces, none exceeding 2 cm in diameter. Lymph nodes are negative

Stage IIIC T3c N0 M0 Any T N1 M0

Abdominal implants >2 cm in diameter and/or positive retroperitoneal or inguinal nodes

215

Appendix 3: Staging: FIGO and TNM

Stage IV Any T Any N M1

Growth involving one of both fallopian tubes with distant metastases. If pleural eﬀusion is present, cytological ﬂuid must be positive for malignant cells to be stage IV. Parenchymal liver metastasis equals stage IV

Notes on staging: Staging for fallopian tube carcinoma is by the surgicopathological system. Operative ﬁndings designating stage are determined before tumour debulking.

a

FIGO staging for carcinoma of the corpus uteri Stage 0 TIS

Carcinoma in situ (pre-invasive carcinoma).

Stage I

Tumour conﬁned to the corpus uteri

Stage IA T1a N0 M0

No or less than half myometrial invasion

Stage IB T1b N0 M0

Invasion equal to or more than one half of the myometrium.

Stage II T2 N0 M0

Tumour invades cervical stroma, but does not extend beyond the uterus

Stage III

Local and/or regional spread of the tumour.

Stage IIIA T3a N0 M0

Tumour invades the serosa of the corpus uteri and/or adnexaea

Stage IIIB T3b N0 M0

Vaginal and/or parametrial involvementa

Stage IIIC

Metastases to pelvic and/or para-aortic lymph nodesa

Stage IIIC1 T1-3b N1 M0

Positive pelvic lymph nodes

Stage IIIC2 T1-3b N2 M0

Positive para-aortic lymph nodes with or without positive pelvic lymph nodes

Stage IV

Tumour invades bladder and/or bowel mucosa, and/or distant metastases

Stage IVA T4 Any N M0

Tumour invasion of bladder and/or bowel mucosa

Stage IVB Any T Any N M1

Distant metastases including intra-abdominal and/or inguinal lymph nodes.

Notes on endometrial staging: a Positive cytology has to be reported separately without changing the stage. Any stage can be G1, G2, G3. Endocervical glandular involvement only should be considered as stage I and no longer as stage II.

Histopathology – degree of diﬀerentiation Cases of carcinoma of the corpus should be classiﬁed (or graded) according to the degree of histological diﬀerentiation, as follows: grade 1 = 5% or less of a non-squamous or non-morular solid growth pattern; grade 2 = 6–50% of a non-squamous or non-morular solid growth pattern; grade 3 = more than 50% of a non-squamous or non-morular solid growth pattern.

216

Appendix 3: Staging: FIGO and TNM

Notes on pathological grading 1. Notable nuclear atypia, inappropriate for the architectural grade, raises the grade of a grade 1 or grade 2 tumour by 1. 2. In serous adenocarcinoma, clear cell adenocarcinomas and squamous cell carcinomas nuclear grading takes precedence. 3. Adenocarcinomas with benign squamous diﬀerentiation are graded according to the nuclear grade of the glandular component.

Rules related to staging 1. Because corpus cancer is now staged surgically, procedures previously used for determination of stages are no longer applicable (e.g. ﬁndings from fractional dilation and curettage to diﬀerentiate between stage I and stage II). 2. It is appreciated that there may be a small number of patients with corpus cancer who will be treated primarily with radiotherapy. If that is the case, the clinical staging adopted by FIGO in 1971 still would apply, but designation of that staging system would be noted. 3. Ideally, width of the myometrium should be measured along with the width of tumour invasion.

Clinical staging for carcinoma of the corpus uteri The previously accepted FIGO rules for clinical staging are still relevant and should be used for cases not primarily operated upon and cases treated with radiation only. Stage 0

Atypical endometrial hyperplasia, carcinoma in situ. Histological ﬁndings are suspicious of malignancy. Cases of stage 0 should not be included in any therapeutic statistics

Stage I

The carcinoma is conﬁned to the corpus

Stage IA

The length of the uterine cavity is ≤8 cm

Stage IB

The length of the uterine cavity is >8 cm

Stage II

The carcinoma has involved the corpus and the cervix, but not outside the true pelvis

Stage III

The carcinoma has extended outside the uterus, but not outside the true pelvis

Stage IV

The carcinoma has extended outside the true pelvis or has obviously involved the mucosa of the bladder or rectum. A bullous oedema as such does not permit a case to be allotted to stage IV

Stage IVA

Spread of the growth to adjacent organs such as urinary bladder, rectum, sigmoid or small bowel

Stage IVB

Spread to distant organs

217

Appendix 3: Staging: FIGO and TNM

FIGO staging for uterine sarcomas (leiomyosarcomas, endometrial stromal sarcomas, adenosarcomas and carcinosarcomas) (1) Leiomyosarcomas Stage I

Tumour limited to uterus

Stage IA

<5 cm

Stage IB

>5 cm

Stage II

Tumour extends to the pelvis

Stage IIA

Adnexal involvement

Stage IIB

Tumour extends to extrauterine pelvic tissue

Stage III

Tumour invades abdominal tissues (not just protruding into the abdomen)

Stage IIIA

1 site

Stage IIIB

>1 site.

Stage IIIC

Metastases to pelvic and/or para-aortic lymph nodes

Stage IV

Tumour invades bladder and/or rectum, and/or distant metastases

Stage IVA

Tumour invades bladder and/or rectum

Stage IVB

Distant metastasis

(2) Endometrial stromal sarcomas (ESS) and adenosarcomasa Stage I

Tumour limited to uterus

Stage IA

Tumour limited to endometrium/endocervix with no myometrial invasion

Stage IB

Less than or equal to half myometrial invasion

Stage IC

More than half myometrial invasion

Stage II

Tumour extends to the pelvis

Stage IIA

Adnexal involvement

Stage IIB

Tumour extends to extrauterine pelvic tissue

Stage III

Tumour invades abdominal tissues (not just protruding into the abdomen)

Stage IIIA

1 site

Stage IIIB

>1 site

Stage IIIC

Metastases to pelvic and/or para-aortic lymph nodes

Stage IV

Tumour invades bladder and/or rectum, and/or distant metastases

Stage IVA

Tumour invades bladder and/or rectum

Stage IVB

Distant metastasis

(3) Carcinosarcomas Carcinosarcomas should be staged as carcinomas of the endometrium a Simultaneous tumours of the uterine corpus and ovary/pelvis in association with ovarian/pelvic endometriosis should be classiﬁed as independent primary tumours.

218

Appendix 3: Staging: FIGO and TNM

FIGO staging for carcinoma of the cervix uteri Stage 0 TIS

Carcinoma in situ, intra-epithelial carcinoma

Stage I

The carcinoma is strictly conﬁned to the cervix (extension to the corpus should be disregarded)

Stage IA

Invasive carcinoma which can be identiﬁed only by microscopy, with deepest invasion ≤5 mm and largest extension ≤7 mm.

Stage IA1 T1a1 N0 M0

Measured stromal invasion of ≤3.0 mm in depth and extension of ≤7.0 mm.

Stage IA2 T1a2 N0 M0

Measured stromal invasion of >3.0 mm and not >5.0 mm, with an extension of not >7.0 mm.

Stage IB

Clinically visible lesions limited to the cervix uteri or preclinical cancers greater than IAa

Stage IB1 T1b1 N0 M0

Clinically visible lesion ≤4.0 cm in greatest dimension

Stage IB2 T1b2 N0 M0

Clinically visible lesion >4.0 cm in greatest dimension

Stage II

Cervical carcinoma invades beyond the uterus, but not to the pelvic wall or to the lower third of the vagina

Stage IIA

Without parametrial invasion

Stage IIA1 T2a1 N0 M0

Clinically visible lesion ≤4.0 cm in greatest dimension

Stage IIA2 T2a2 N0 M0

Clinically visible lesion >4.0 cm in greatest dimension

Stage IIB T2b N0 M0

With obvious parametrial invasion.

Stage III

The tumour extends to the pelvic wall and/or involves the lower third of the vagina and/or causes hydronephrosis or non-functioning kidneyb

Stage IIIA T3a N0 M0

Tumour involves lower third of the vagina, with no extension to the pelvic wall

Stage IIIB T1 N1 M0 T2 N1 M0 T3a N1 M0 T3b Any N M0

Extension to the pelvic wall and/or hydronephrosis or non-functioning kidney

Stage IV

The carcinoma has extended beyond the true pelvis or has involved (biopsy proved) the mucosa of the bladder or rectum. A bullous oedema, as such, does not permit a case to be allotted to stage IV

Stage IVA T4 Any N M0

Spread of the growth to adjacent organs

219

Appendix 3: Staging: FIGO and TNM

Stage IVB Any T Any N M1

Spread to distant organs

Notes on staging for cervical carcinoma: a All macroscopically visible lesions – even with superﬁcial invasion – are allotted to stage IB carcinomas. Invasion is limited to a measured stromal invasion with a maximal depth of 5.0 mm and a horizontal extension of not >7.0 mm. Depth of invasion should not be >5.0 mm and a horizontal extension of not >7.0 mm. Depth of invasion should not be >5.0 mm taken from the base of the epithelium of the original tissue – superﬁcial or glandular. The depth of invasion should always be reported in millimetres, even in those cases with ‘early (minimal) stromal invasion’ (~1 mm). The involvement of vascular/lymphatic spaces should not change the stage allotment. b On rectal examination, there is no cancer-free space between the tumour and the pelvic wall. All cases with hydronephrosis or non-functioning kidney are included, unless they are known to be due to another cause.

FIGO staging for carcinoma of the vagina Stage 0 TIS

Carcinoma in situ, intra-epithelial carcinoma

Stage I T1 N0 M0

The carcinoma is limited to the vaginal wall

Stage II T2 N0 M0

The carcinoma has involved the subvaginal tissue but has not extended onto the pelvic wall

Stage III T1 N1 M0 T2 N1 M0 T3 N0 M0 T3 N1 M0

The carcinoma has extended onto the pelvic wall

Stage IV

The carcinoma has extended beyond the true pelvis or has clinically involved the mucosa of the bladder or rectum. Bullous oedema as such does not permit a case to be allotted to Stage IV

Stage IVA T4 Any N M0

Spread to adjacent organs and/or direct extension beyond the true pelvis

Stage IVB Any T Any N M1

Spread to distant organs

FIGO staging for carcinoma of the vulva Stage 0

Carcinoma in situ; intraepithelial carcinoma

Stage I

Tumour conﬁned to the vulva.

Stage IA

Lesions ≤2 cm in size, conﬁned to the vulva or perineum and with stromal invasion ≤1.0 mma, no nodal metastasis

Stage IB

Lesions >2 cm in size or with stromal invasion >1.0 mma, conﬁned to the vulva or perineum, with negative nodes

Stage II

Tumour on any size with extension to adjacent perineal structures (⅓ lower urethra, ⅓ lower vagina, anus) with negative nodes

Stage III

Tumour of any size with or without extension to adjacent perineal structures (⅓ lower urethra, ⅓ lower vagina, anus) with positive inguinofemoral lymph nodes

Stage IIIA

(i)

With 1 lymph node metastasis (≥5 mm), or

(ii) 1–2 lymph node metastasis(es) (<5 mm)

220

Appendix 3: Staging: FIGO and TNM

Stage IIIB

(i)

With 2 or more lymph node metastases (≥5 mm), or

(ii) 3 or more lymph node metastases (<5 mm) Stage IIIC

With positive nodes with extracapsular spread

Stage IV

Tumour invades other regional (⅔ upper urethra, ⅔ upper vagina), or distant structures

Stage IVA

Tumour invades any of the following: (i)

Upper urethral and/or vaginal mucosa, bladder mucosa, rectal mucosa, or ﬁxed to pelvic bone, or

(ii) Fixed or ulcerated inguinofemoral lymph nodes Stage IVB

Any distant metastasis including pelvic lymph nodes

Notes about the staging for vulval carcinoma: a The depth of invasion is deﬁned as the measurement of the tumour from the epithelial-stromal junction of the adjacent most superﬁcial dermal papilla to the deepest point of invasion.

FIGO staging for trophoblastic tumour Stage I

Disease conﬁned to uterus

Stage IA

Disease conﬁned to uterus with no risk factors

Stage IB

Disease conﬁned to uterus with one risk factor

Stage IC

Disease conﬁned to uterus with two risk factors

Stage II

Gestational trophoblastic tumour extending outside uterus but limited to genital structures (adnexa, vagina, broad ligament)

Stage IIA

Gestational trophoblastic tumour involving genital structures without risk factors

Stage IIB

Gestational trophoblastic tumour extending outside uterus but limited to genital structures with one risk factor

Stage IIC

Gestational trophoblastic tumour extending outside uterus but limited to genital structures with two risk factors

Stage III

Gestational trophoblastic tumour extending to lungs with or without known genital tract involvement

Stage IIIA

Gestational trophoblastic tumour extending to lungs with or without genital tract involvement and with no risk factors

Stage IIIB

Gestational trophoblastic tumour extending to lungs with or without genital tract involvement and with one risk factor

Stage IIIC

Gestational trophoblastic tumour extending to lungs with or without genital tract involvement and with two risk factors

Stage IV

All other metastatic sites

221

Appendix 3: Staging: FIGO and TNM

Stage IVB

All other metastatic sites with one risk factor

Stage IVC

All other metastatic sites with two risk factors

Notes on staging: Risk factors aﬀecting staging include the following: serum human chorionic gonadotrophin >100 000 mIU/ml, and duration of disease >6 months from termination of antecedent pregnancy. The following factors should be considered and noted in reporting: Prior chemotherapy has been given for known gestational trophoblastic tumour. Placental site tumours should be reported separately. Histological veriﬁcation of disease is not required.

222

Appendix 4: Useful websites ASCCP: www.asccp.org American Society for Colposcopy and Cervical Pathology BAAF: www.baaf.org.uk British Association of Adoption and Fostering BACKUP: www.cancerbackup.org.uk/Home Cancer Backup, MacMillan Cancer Support BASRT: www.basrt.org.uk British Association for Sexual and Relationship Therapy BGCS: www.bgcs.org.uk British Gynaecological Cancer Society BSCCP: www.bsccp.org.uk British Society for Colposcopy and Cervical Pathology Cancer Couselling: www.cctrust.org.uk Cancer Counselling Trust Carers UK: www.carers.org Advice and support for carers COTS: www.surrogacy.org.uk Childlessness Overcome Through Surrogacy CRUK: www.cancerresearchuk.org Cancer Research UK CRUSE: www.crusebereavementcare.org.uk CRUSE bereavement care Daisy Chain: www.daisynetwork.org.uk Information related to premature menopause Dial UK: www.dialuk.org.uk Network of disability information and advice services FIGO: www.ﬁgo.org International Federation of Gynecology and Obstetrics GOG: www.gog.org Gynecologic Oncology Group IGCS: www.igcs.org International Gynecologic Cancer Society Infertility Network UK: www.infertilitynetworkuk.com Infertility support Jo’s Trust: www.jotrust.co.uk National support group and patient information related to both cervical cancer and precancer Look Good … Feel Better: www.lookgoodfeelbetter.co.uk Cancer support charity that helps women manage the visible side eﬀects of cancer treatments Lymphoedema: www.lymphoedema.org Lymphoedema support network Macmillan Cancer Support: www.macmillan.org.uk Macmillan cancer support

Appendix 4: Useful websites

Maggie’s cancer caring centres: www.maggiescentres.org Helping people live with cancer Marie Curie Cancer Care: www.mariecurie.org.uk National Cancer Institute (NCI): www.cancer.gov National Cancer Research Institute (NCRI): www.ncri.org.uk National Osteoporosis Society: www.nos.org.uk Ovacome: www.ovacome.org.uk Support group for anyone aﬀected by ovarian cancer Ovarian Cancer Action: www.ovarian.org.uk Ovarian cancer research and patient information RCOG: www.rcog.org.uk Royal College of Obstetricians and Gynaecologists SGO: www.sgo.org Society of Gynaecologic Oncologists Vulva Awareness Campaign Organisation (VACO): www.vaco.co.uk Vulval awareness organization

224

Index 3DCRT, see Three-dimensional conformal radiotherapy abdominal hysterectomy, 139 abdominal incisions, 73 abortions, 183 actinomycin D, 187, 190, 196 acute side eﬀects of radiotherapy, 86 addresses, 223 adenocarcinomas, 39–40, 46, 147, 150 adenoﬁbroma tumours, 42 adenoma malignum, 46 adenosarcoma tumours, 42 adenosquamous carcinomas, 46 adjuvant treatments aims, 200 chemotherapy, 94–6, 128, 144, 179 decision making, 200, 201 progestogens, 145 radiotherapy, 79, 142–3, 178 adnexal masses, 17, 52 Adriamycin, see doxorubicin advanced-stage disease factors, 126, 128–9, 156–9 aetiology cervical cancer, 11 endometrial cancer, 9, 133–4 ovarian cancer, 8 vaginal cancer, 12 vulval cancer, 12, 163–4 age-standardized incidence rates (ASR), 3 aggressive surgery, 130 alpha-fetoproteins, 193 American Society of Colposcopy and Cervical Pathology (ASCCP), 29 Amsterdam criteria, 22 anaemia, 151 anaesthesia, 165 analgesics, 106 anthracyclines, 93 antibiotic prophylaxis, 72 anti-emetics, 110

anti-metabolite drugs, 94 anxiety, 116–17 appendiceal mucinous tumours, 36 appendices, 207 ASCCP, see American Society of Colposcopy and Cervical Pathology Ashkenazi Jewish ethnicity, 20 ASR, see age-standardized incidence rates assays β-human chorionic gonadotrophin, 182–3 ASTEC (A Study in the Treatment of Endometrial Cancer), 140 autosomal dominant syndrome, 21 Bartholin’s glands, 48, 163, 173 basal cell carcinomas, 48, 163, 173 beam arrangement, 88 benign tumours, 17, 33, 36 Bethesda criteria, 22 bilateral salpingo-oophorectomy (BSO), 24, 139 bisphosphonate, 113 bladder surgery, 75 bleomycin, 195, 196 blue round-cell tumours, 176 bone pain management, 108 borderline tumours, 34, 36 bowel constipation, 104, 111–12 injuries/damage, 74 irritable bowel syndrome, 204 obstruction, 130 palliative care, 104 preparation for surgery, 71 toxicity, 204 brachytherapy (BT), 52, 82, 143, 172 BRCA gene mutations, 19, 120 breathlessness, 114 Brenner tumours, 37

British Society of Colposcopy and Cervical Pathology (BSCCP), 27 BSO, see bilateral salpingooophorectomy BT, see brachytherapy burden measurements, 2 CA125 tumour marker, 16–17, 52, 122–3, 193 cachexia, 114 caesarean section, 160 Cancer Genetics Network (CGN), 18 capecitabine, 94 carboplatin, 91–2, 96–7, 98, 128 carcinoﬁbroma tumours, 42 carcinoma in situ (CIN), 25 catheters, 75 CEA tumour marker, 122 cerebral metastases, 190 cervical cancer, 147–61 adenocarcinoma, 150 advanced-stage disease, 156–9 chemotherapy, 153, 156, 160 computed tomography, 57, 66, 149, 159 decision making, 203 early-stage disease, 153–6 epidemiology, 10–11, 147 FDG-PET, 66 FDG-PET/CT, 65, 66 FIGO staging, 149, 151, 219–20 follow-up treatments, 159–60 HPV, 11 hysterectomy, 154–6 imaging, 61, 65, 66, 149, 152, 159 lymph nodes, 156 magnetic resonance imaging, 61, 66, 149, 152, 159 management, 152–61 pathology, 44–7, 148, 149–51

Index

cervical cancer (cont.) positron emission tomography, 159 pregnancy, 160 prevention, 29–30 prognostic variables, 151–2 radical surgery, 153–6 radiotherapy, 80, 153, 156–9 rare variants, 46 risk factors, 147 screening, 25–9, 147 special clinical situations, 160 staging, 149 surgery, 153–6 trachelectomy, 160 ultrasound imaging, 53 X-rays, 149 cervical intraepithelial neoplasia (CIN), 12 CGN, see Cancer Genetics Network (CGN) Charing Cross scoring, 186 chemosensitivity, 37 chemotherapy, 91–101 anti-metabolite drugs, 94 cervical cancer, 153, 156, 160 dose intensity, 97 endometrial cancer, 91, 100–1, 144 epithelial ovarian cancer, 91, 94–7 gestational trophoblastic disease, 190 hydatidiform moles, 184, 185 malignant ovarian germ cell tumours, 194–7 metastatic sarcomas, 179 modes of action, 91–101 oral chemotherapy, 93, 99 ovarian cancer, 91, 127, 128–9 patients’ expectations, 99 platinum-free intervals, 98 quality of life, 99 sarcomas, 179 topo-isomerase inhibitors, 93 toxicity, 99 types, 91–101 see also individual drugs chest radiography, 138, 183, 184 Chlamydia trachomatis, 11

226

chlorambucil, 99 choriocarcinomas, 187, 188, 190, 192, 193 CIN, see cervical intraepithelial neoplasia cisplatin, 79, 91–2, 96, 98 gestational trophoblastic disease, 190 ovarian cancer, 128 ovarian germ cell tumours, 195, 196 classiﬁcation systems, 2, 134 clear cell carcinomas, 36, 40, 46, 47 clinical examinations, 149 clinical presentation choriocarcinoma, 188 endometrial cancer, 137 hydatidiform moles, 183–7 malignant ovarian germ cell tumours, 193 ovarian cancer, 121 sarcomas, 177 vulval cancer, 164–6 clinical trial phases, 207 colour Doppler analysis, 52 colposcopy, 25, 27, 28 Common Toxicity Criteria (CTC), 99 co-morbidity, 166 complete prevalence, 4 complex wounds, 113–14 computed tomography (CT), 51, 54–7 cervical cancer, 57, 66, 149, 159 endometrial cancer, 56, 65, 138 hydatidiform moles, 184 malignant ovarian germ cell tumours, 195 ovarian cancer, 54–6, 65, 127 sarcomas, 178 technique, 54 vaginal cancer, 67 vulval cancer, 57, 63, 67 see also ﬂourine-18-ﬂuoro-2deoxy-d-glucose positron emission tomography concomitant chemotherapy, 79 conformal radiotherapy, 89 conisation, 153 consent to surgery, 73 consolidation therapy, 129–30

constipation, 104, 111–12 continuous subcutaneous infusion (CSCI), 109 contraception, 25 corpus uteri cancer (uterine corpus cancer), 216–17 counselling, 72, 116–17 criteria of response, 207 CSCI, see continuous subcutaneous infusion CT, see computed tomography CTC, see Common Toxicity Criteria cumulative incidence, 4 cumulative risk, 4 cyclophosphamide, 96, 128, 196 cytological sampling, 24 cytoreduction, 126 D&C, see dilatation and curettage DALYs, see disability-adjusted life-years (lost) decision making, 199–205 clinical examples, 203–5 patient factors, 202–5 treatment aims, 200–1 treatment selection, 201–5 deconditioning, 114 degree of diﬀerentiation, 216 deoxyribonucleic acid (DNA), 24, 83, 93 depression, 116–17 DES, see diethylstilbestrol diagnosis constipation, 112 diagnostic laparoscopy, 77 hypercalcaemia, 113 ovarian cancer, 121 see also pathology diethylstilbestrol (DES), 47 diﬃcult pain management, 107 dilatation and curettage (D&C), 184 disability-adjusted life-years (lost) (DALYs), 5 disseminated peritoneal adenomucinosis, 36 distress management, 104 DNA, see deoxyribonucleic acid docetaxel, 179 Doppler analysis, 24, 52, 184, 195

Index

dose factors, 84, 97 doxorubicin, 96, 128, 179 drip and suck, 109 drug development phases, 207 drug resistance, 185, 190 duration of chemotherapy, 97 dysgerminomas, 192 early-stage disease, 126, 128, 153–6 EBRT, see external beam radiotherapy ectopic implantation, 181 embolism, 189 embryonal carcinoma, 192 embryonal rhabdomyosarcomas, 176 en bloc excision, 169 endocrines, 101, 145, 179 endodermal sinus, 192 endometrial cancer, 133–46 adjuvant radiotherapy, 142–3 aetiology, 133–4 chemotherapy, 91, 100–1, 144 classiﬁcation, 134 clinical presentation, 137 computed tomography, 56, 138 decision making, 204 endometrial sampling, 23–4 epidemiology, 9–10 FDG-PET, 65, 139 FIGO staging, 140 hormonal treatments, 101, 145 hysterectomy, 139–42 hysteroscopy, 23–4 imaging, 53, 59–60, 65, 137, 138–9 laparoscopy, 141 magnetic resonance imaging, 59–60, 65, 138 minimal access surgery, 141 palliative radiotherapy, 144 pathology, 39–42, 134–6 positron-emission tomography, 139 prevention, 24–5 radiotherapy, 80, 142–4 risk factors, 133–4 sampling, 23–4 screening, 21–5

staging, 140 surgery, 139–42 tamoxifen, 101 transvaginal ultrasound, 23–4 treatment planning, 138–9 ultrasound imaging, 53, 65, 137 X-rays, 138 endometrial stromal sarcomas (ESS), 41, 42, 135, 176 endometrial thickness (ET), 23 endometrial volume, 24 endometrioid cancer, 36, 39–40, 41 endometriosis, 36 epidemiology, 1–12 cervical cancer, 10–11, 147 choriocarcinoma, 187 deﬁnition, 1–2 endometrial cancer, 9–10 hydatidiform moles, 181 ovarian cancer, 6–8 vaginal cancer, 11–12 vulval cancer, 11–12 epithelial ovarian cancer (EOC), 91, 119 adjuvant chemotherapy, 94–6 endocrine manipulation, 101 ﬁrst-line chemotherapy, 96–7 neoadjuvant chemotherapy, 94–6 palliative chemotherapy, 97–100 epithelial tumour pathology, 33–5 ESS, see endometrial stromal sarcomas etoposide, 93, 98, 190, 195, 196 European standard population of Doll, 4 Ewing’s sarcoma, 176 examinations, 16, 71, 149 excision, 169 exfoliative cytology, 25 extended surgical staging, 140 external beam radiotherapy (EBRT), 79, 142, 156, 172 extrafascial hysterectomy, 155

fallopian tube cancer, 15–21, 215, 216 see also ovarian cancer familial endometrial cancer, 21–5 familial ovarian cancer, 18–21, 120–6 fatality, 5 FDG, see ﬂuorine-18ﬂuoro-2-deoxy-d-glucose imaging FDG-PET, see ﬂuorine-18ﬂuoro-2-deoxy-d-glucose positron-emission tomography FDG-PET/CT, see ﬂuorine-18ﬂuoro-2-deoxy-d-glucose positron-emission tomography computed tomography Federation of Gynecology and Obstetrics (FIGO) ovarian cancer, 54 Federation of Gynecology and Obstetrics (FIGO) staging cervical cancer, 149, 151, 219–20 corpus uteri cancer, 216–17 endometrial cancer, 140 fallopian tube cancer, 215, 216 gestational trophoblastic tumours, 186, 221–2 hydatidiform moles, 185 ovarian cancer, 214 rules, 213–22 uteri cancer, 216–17 vaginal cancer, 161, 220 vulval cancer, 166, 220–1 fertility factors, 128, 203 ﬁne needle aspiration (FNA), 166 ﬁrst-line chemotherapy, 96–7 ﬂuorine-18-ﬂuoro-2-deoxy-dglucose (FDG) imaging, 51, 63–5 ﬂuorine-18-ﬂuoro-2-deoxy-dglucose positron-emission tomography (FDG-PET), 63–5, 66, 139 ﬂuorine-18-ﬂuoro-2-deoxy-dglucose positron-emission tomography computed tomography (FDG-PET/ CT), 64, 65, 66

227

Index

ﬂuorouracil, 94 FNA, see ﬁne needle aspiration follow-up factors, 159–60, 191 fractionation, 84 frequency of screening fungating wounds, 113–14 gastrointestinal obstruction, 109 gemcitabine, 94, 179, 196 gene mutations, 19, 120 germ cell tumours, 37, 38, 176, 191–7 gestational choriocarcinoma, 181 gestational trophoblastic disease (GTD), 181–91 gestational trophoblastic tumours (GTT), 181–91, 221–2 GnRH, see gonadotrophinreleasing hormone GOG, see Gynecologic Oncology Group gonadotrophin-releasing hormone (GnRH), 145 granulosa cell tumours, 37 gray unit, 84 GTD, see gestational trophoblastic disease GTT, see gestational trophoblastic tumours Gynecologic Oncology Group (GOG), 18, 96 haematogenous spread, 56 haemorrhage, 76, 105, 189 Hasson insertion, 77 hCG, see human chorionic gonadotrophin HDR, see high dose rate hereditary non-polyposis colon cancer (HNPCC), 21, 25, 39, 121, 133 herpes virus type 2, 11 high dose rate (HDR) brachytherapy, 83 high-grade sarcoma tumours, 175 high-risk population screening, 22 histology, 202 histopathology, 152, 216 history factors, 71 HIV, see human immunodeﬁciency virus

228

HNPCC, see hereditary non-polyposis colon cancer hormonal treatments, 100, 101, 129, 145 hormone replacement therapy (HRT), 24, 133 HPV, see human papilloma virus HRT, see hormone replacement therapy human chorionic gonadotrophin (hCG), 182–7, 194 human immunodeﬁciency virus (HIV), 11 human papillomavirus (HPV) cervical cancer, 11, 45, 147 DNA testing, 26 vaccines, 29 vaginal cancer, 12 vulval cancer, 12, 47, 163 hydatidiform moles, 181–91 chemotherapy, 184, 185 clinical presentation, 183–7 computed tomography, 184 Doppler analysis, 184 epidemiology, 181 follow-up, 191 imaging, 183, 184 magnetic resonance imaging, 184 management, 183–7 origins, 181 pathology, 181 prognosis, 191 radiotherapy, 183 registration, 183–7 staging, 185 ultrasound, 183 hypercalcaemia, 37, 112–13 hyperplasia, 24 hypofractionation, 84 hysterectomy cervical cancer, 154–6 decision making, 203 endometrial cancer, 24, 139–42 minimal access surgery, 77 sarcomas, 178 hysteroscopy, 23–4

ICD-10, see International Classiﬁcation of Diseases ICD-O-3, see International Classiﬁcation of Diseases for Oncology ICON, see International Collaborative Ovarian Neoplasm ifosfamide, 179, 196 imaging, 51–68 cervical cancer, 66, 149, 152, 159 endometrial cancer, 53, 59–60, 65, 137, 138–9 hydatidiform moles, 183, 184 ovarian cancer, 65 ovarian germ cell tumours, 194 recommended algorithms, 65–7 sarcomas, 178 vulval cancer, 51, 67 see also individual modalities immunohistochemistry, 22, 175 immunosuppression, 163, 164 IMRT, see intensity-modulated radiotherapy incidence data, 209 endometrial cancer, 9 ovarian cancer, 6 rates, 2 vulvar and vaginal cancers, 11 incident cases, 2 incident pain management, 107 indications for radiotherapy, 79–80 infantile choriocarcinoma, 190 infertility, 194 inheritable genetic predisposition, 19 intensity-modulated radiotherapy (IMRT), 89 Intergroup Study OV10, 96 International Classiﬁcation of Diseases (ICD-10), 2 International Classiﬁcation of Diseases for Oncology (ICD-O-3), 2 International Collaborative Ovarian Neoplasm (ICON), 96, 128

Index

International Ovarian Tumour Analysis (IOTA), 17 international trials for rare diseases, 179 intracavitary brachytherapy, 143, 156 intraoperative complications, 74 intraperitoneal bleeding, 183 intraperitoneal chemotherapy, 129 intravenous ﬂuids, 109 invasion depth, 45, 167 invasive carcinoma of the vulva, 47 invasive moles, 181 invasive squamous cell carcinomas, 45 investigation factors, 71, 164–6, 193 IOTA, see International Ovarian Tumour Analysis iridium-192, 83 irritable bowel syndrome, 204 Keyes punch, 165 lactate dehydrogenase, 193 laparoscopy, 77–8, 141, 194 large bowel surgery, 75 large loop excision of the transformation zone (LLETZ), 153 late side eﬀects of radiography, 86 LBC, see liquid-based cytology leiomyosarcomas, 41, 175, 178, 180 Leydig cell tumours, 38 lichen sclerosus, 163, 164 linear accelerators, 79 lipid cell tumours, 38 liquid-based cytology (LBC), 25 LLETZ, see large loop excision of the transformation zone local anaesthesia, 165 local treatments decision making, 201 vulval cancer, 167 locoregional lymph nodes, 165, 168–9 low-grade sarcoma tumours, 175 low-risk population screening, 23

Lugol’s iodine, 26 LVSI, see lymph-vascular space invasion lymph nodes cervical cancer, 45, 152, 156 dissection, 171 magnetic resonance lymphography, 67 vulval cancer, 168–9, 171 lymphadenectomy, 156 lymphatic spread, 202 lymphoscintigraphy, 171 lymph-vascular space invasion (LVSI), 45, 152, 153 Lynch syndrome, 21, 39, 121, 133 magnetic resonance imaging (MRI), 51, 57–63 cervical cancer, 61, 66, 149, 152, 159 endometrial cancer, 59–60, 65, 138 hydatidiform moles, 184 magnetic lymphography, 67 malignant ovarian germ cell tumours, 195 ovarian cancer, 58–9, 65 sarcomas, 178 technique, 57–8 vulval cancer, 62, 67, 166 magnetic resonance lymphography, 67 maintenance therapies, 129–30 malignant adnexal masses, 17 malignant melanomas, 48 malignant mesenchymal tumours, 41–4 malignant mixed germ cell tumours, 193 malignant mixed Müllerian tumours (MMMT), 42, 176 malignant mucinous tumours, 35, 36 malignant neoplasm classiﬁcation, 2 malignant ovarian germ cell tumours (MOGCT), 191–7 chemotherapy, 194–7 computed tomography, 195 investigations, 193 long-term outlook, 197

magnetic resonance imaging, 195 management, 194–7 pathology, 191–3 presentation, 193 relapse, 194–7 surgery, 193 ultrasound, 195 malodorous wounds, 113–14 management cervical cancer, 152–61 choriocarcinoma, 188 hydatidiform moles, 183–7 malignant ovarian germ cell tumours, 194–7 MAS, see minimal access surgery measuring risk or burden, 2 medroxyprogesterone acetate, 145 melanomas, 163, 173 mesonephric carcinomas, 46 metastasis cervical cancer, 44, 45, 46 microinvasive squamous cells, 44 vulval cancer, 47 metastatic sarcomas, 178, 179 methotrexate, 185, 187, 196 metoclopramide, 109 microinvasive carcinoma, 44, 47 microinvasive squamous cells, 44 microsatellite instability (MSI), 22 minimal access surgery (MAS), 76–8 advantages, 76 complications, 76 endometrial cancer, 141 laparoscopy, 77–8 minimal deviation adenocarcinoma, 46 Mirena-IUS insertion, 25 mismatch repair (MMR) genes, 19, 21 mixed germ cell tumours, 193 mixed Müllerian tumours of the uterus, 42 MMMT, see malignant mixed Müllerian tumours MMR, see mismatch repair

229

Index

modes of action chemotherapy, 91–101 radiotherapy, 83 MOGCT, see malignant ovarian germ cell tumours moles, 181–91 morbidity, 166 morphological classiﬁcation, 3 mortality, 5 cervical cancer, 10 data, 209 endometrial cancer, 9 ovarian cancer, 6 rates, 3, 5 vulvar and vaginal cancers, 11 MRI, see magnetic resonance imaging MSI, see microsatellite instability mucinous adenocarcinomas, 36, 41 mucinous ascites, 36 mucinous tumours, 35 Müllerian tumours, 42, 176 myometrial invasion, 40 nasogastric tubes, 109 nausea, 108–11 neoadjuvant chemotherapy, 94–6 neoplasm classiﬁcation, 2 neuroendocrine tumours, 151 neuropathic pain management, 108 NHS cervical screening programme (NHSCSP), 25 nodal metastasis, 44, 45, 151, 152 non-dysgerminomatous tissue, 192 non-gestational choriocarcinoma, 193 normal tissue tolerance, 85 normal tissue volume, 88 oestrogenic tumours, 37 open Hasson insertion, 77 oral chemotherapy, 93, 99 oral constipation treatments, 112 organizing mucinous ﬂuid, 36 origins of choriocarcinoma, 187 origins of hydatidiform moles, 181 orthovoltage X-ray machines, 81

230

outpatient endometrial assessment, 138 OV10, see Intergroup Study ovarian cancer, 119–31 adjuvant chemotherapy, 94–6 advanced-stage disease, 126, 128–9 chemotherapy, 91, 97–100, 127, 128–9 clinical presentation, 121 computed tomography, 54–6, 65 consolidation, 129–30 cytoreduction, 126 diagnosis, 121 early-stage disease, 126, 128 endocrine manipulation, 101 epidemiology, 6–8 examinations, 16 FDG-PET, 64 FDG-PET/CT, 64, 65 fertility-sparing procedures, 128 FIGO staging, 214 ﬁrst-line chemotherapy, 96–7 hormonal therapies, 129 imaging, 54–6, 64, 65 intraperitoneal chemotherapy, 129 magnetic resonance imaging, 58–9, 65 maintenance therapies, 129–30 neoadjuvant chemotherapy, 94–6 new therapeutic strategies, 130 palliative care, 97–100, 130 pathology, 33–9, 119–20 radiotherapy, 80 recurrent disease chemotherapy, 129 risk factors, 120 screening, 15–21, 121 staging, 123, 126–9 surgery, 123–6, 130 treatment, 123–6 tumour markers, 16–18, 122–3 ultrasound imaging, 52 ovarian germ cell tumours, 191–7 ovarian masses, 52, 55, 58

paclitaxel, 96–7, 128, 190, 196 Paget’s disease, 48, 163 pain management, 104, 106–8 palliative care, 103–17, 201 analgesics, 106 anxiety, 116–17 bowel habit, 104 breathlessness, 114, 115 chemotherapy, 97–100 complex wounds, 113–14 constipation, 104, 111–12 counselling, 116–17 decision making, 201 depression, 116–17 diﬃcult pains, 107 distress, 104 entering the palliative phase, 115 fungating wounds, 113–14 haemorrhage, 105 hypercalcaemia, 112–13 malodorous wounds, 113–14 nausea, 108–11 ovarian cancer, 97–100, 130 pain management, 104, 106–8 radiotherapy, 79, 144 symptom management, 103–17 vomiting, 108–11 vulval cancer, 172 wounds, 104, 113–14 papillary squamous cell carcinomas, 45 pap smears, 25 partial obstruction, 108 partial prevalence, 4 pathology, 33–48 cervical cancer, 44–7, 148, 149–51 choriocarcinoma, 187 endometrial cancer, 39–42, 134–6 epithelial tumours, 33–5 grading, 217 hydatidiform moles, 181 malignant mesenchymal tumours, 41–4 malignant ovarian germ cell tumours, 191–3 neuroendocrine tumours, 151

Index

ovarian cancer, 33–9, 119–20 serous tumours, 33–5 squamous cell carcinoma, 149 uterus malignant mesenchymal tumours, 41–4 vaginal cancer, 47 vulval cancer, 47–8, 163 patient factors chemotherapy expectations, 99 treatment selection, 202–5 pelvic lymph nodes, 169 pelvic lymphadenectomy, 169 PEPI trial, 24 performance scales, 207 peritoneal deposits, 55 PET, see positron-emission tomography pipelle sampling, 24 Piver–Rutledge classiﬁcation, 155 placental site trophoblastic tumours (PSTT), 181, 189 platinum compounds, 91–2, 96–7, 99, 128 platinum-free intervals, 98 platinum sensitivity, 98 PLCO, see Prostate Lung Colorectal and Ovarian Cancer Screening Trial pleural eﬀusions, 72, 114 PMP, see pseudomyxoma peritonei positron-emission tomography (PET), 51, 63–5, 139, 159 postmenopausal bleeding, 23, 137–8, 148 power Doppler analysis, 24 predictive values of screening tests, 5 pregnancy cervical cancer, 160 decision making, 203 gestational trophoblastic tumours, 181–91 ovarian germ cell tumours, 191–7 preoperative assessments, 70, 171, 177 presentation, see clinical presentation prevalence, 4

prevention, 15–30 cervical cancer, 11, 29–30 endometrial cancer, 10, 24–5 ovarian cancer, 8 vulvar and vaginal cancers, 12 primary curative radiotherapy, 79 primary tumours, 164 primitive neuroectodermal tumours, 176 principles of radiotherapy, 83–5 progestogens, 25, 145 prognosis cervical cancer, 151–2 endometrial carcinomas, 40 hydatidiform moles, 191 sarcomas, 178 vulval cancer, 167 prophylactic hysterectomy, 24 Prostate Lung Colorectal and Ovarian Cancer Screening Trial (PLCO), 18 protocols for screening, 27 pseudomyxoma peritonei (PMP), 36 PSTT, see placental site trophoblastic tumours pulmonary embolus, 114 quality-adjusted life-years (lost) (QALYs), 5 quality of life, 99 Querleu–Morrow classiﬁcation, 155 radical treatments aims, 200 cervical cancer, 153–6 decision making, 200, 201 hysterectomy, 77 minimal access surgery, 77, 78 primary radiotherapy, 143 radiotherapy, 79, 143 surgery, 77, 78, 153–6, 168, 169 trachelectomy, 78 vulval cancer, 168, 169 radiobiological diﬀerences, 87 radiotherapy, 79–80 acute side eﬀects, 86 cervical cancer, 80, 153, 156–9

decision making, 204 dose, 84 endometrial cancer, 80, 142–4 fractionation, 84 hydatidiform moles, 183 indications, 79–80 late side eﬀects, 86 modes of action, 83 normal tissue tolerance, 85 normal tissue volume, 88 ovarian cancer, 80 principles, 83–5 radiobiological diﬀerences, 87 sarcomas, 178, 179 side eﬀects, 85–7, 144, 158 therapeutic ratio/index, 87 tolerance, 85 vaginal cancer, 161 vulval cancer, 80, 169, 172 recurrent disease cervical cancer, 61, 159–60 chemotherapy, 129 endometrial cancer, 145–6 ovarian cancer, 59, 64, 129 vulval cancer, 172 registration factors, 183–7 relapse ovarian cancer, 97–100, 129 ovarian germ cell tumours, 194–7 relative survival, 5 repetitive exfoliative cytology, 25 resectability, 54, 59 respiratory failure, 189 response criteria, 207 risk factors cervical cancer, 147 endometrial cancer, 133–4 measuring risk, 2 ovarian cancer, 17, 18, 120 risk of malignancy index (RMI), 17, 122 risk-reducing salpingooophorectomy (RRSO), 19–21 RMI, see risk of malignancy index routine investigations, 149 RRSO, see risk-reducing salpingo-oophorectomy rules for FIGO staging, 213–22

231

Index

saline infusion sonohysterography, 24, 52 sarcomas, 175–80 blue round-cell tumours, 176 carcinosarcomas, 41, 42, 176 chemotherapy, 179 clinical presentation, 177 computed tomography, 178 endometrial stromal sarcomas, 176 germ cell tumours, 176 imaging, 178 leiomyosarcomas of the uterus, 175 magnetic resonance imaging, 178 malignant mixed Müllerian tumours, 176 Müllerian tumours, 176 preoperative assessments, 177 prognosis, 178 radiotherapy, 178, 179 staging, 177 surgery, 178 tamoxifen, 180 screening, 15–30 cervical cancer, 25–9, 147 colposcopy, 27, 28 deﬁnition, 5–6 endometrial cancer, 21–5 fallopian tube cancer, 15–21 familial endometrial cancer, 21–5 familial ovarian cancer, 18–21 frequency, 27 ovarian cancer, 15–21, 121 tests, 5, 15 trials, 17 tubal cancer, 15–21 World Health Organization Criteria, 16 second-line chemotherapy, 98 second-look laparoscopy, 77 selecting treatments, 201–2 decision making, 202–5 patient factors, 202–5 sensitivity CA125, 16 ovarian cancer, 16 platinum, 98 screening tests, 5 sentinel lymph node dissection (SLND), 171

232

serous carcinomas, 40, 46 serous tumours, 33–5, 36 Sertoli–Leydig cell tumours (SLCT), 37, 38 sex-cord stromal tumours, 37 shaving, 73 shielding, 88 side eﬀects radiotherapy, 85–7, 144, 158 surgery, 141 single photon emission computed tomography (SPECT), 67 skeletal scintigraphy, 51, 67 skin preparation, 73 SLCT, see Sertoli–Leydig cell tumours slide-reading techniques, 27 SLND, see sentinel lymph node dissection small bowel surgery, 74 small cell carcinoma, 37 smoking, 163, 164 soft tissue sarcomas, 175–80 sonography, 24 speciﬁcity CA125, 16 ovarian cancer, 16 screening tests, 5 SPECT, see single photon emission computed tomography squamous cell carcinomas cervical cancer, 44–5, 147 endometrial cancer, 41 ovarian cancer, 37 ovarian germ cell tumours, 193 pathology, 149 vaginal cancer, 47 vulval cancer, 47, 163–73 squamous epithelial tumours, 153 staging cervical cancer, 61, 149 decision making, 202 endometrial cancer, 59, 140 hydatidiform moles, 185 ovarian cancer, 54, 59, 64, 123, 126–9 sarcomas, 177

vaginal cancer, 161 vulval cancer, 54, 166 see also Federation of Gynecology and Obstetrics staging, Tumour, Node, Metastases (staging system) steroid cell tumours, 38 stromal luteoma, 38 stromal sarcomas, 41, 42 superﬁcial/orthovoltage X-ray machines, 81 superﬁcially invasive carcinoma of the vulva, 47, 169 surgery, 69–78 abdominal incisions, 73 antibiotic prophylaxis, 72 bladder, 75 bowel injuries/damage, 74 bowel preparation, 71 cervical cancer, 153–6 complications, 141, 169 consent, 73 counselling, 72 endometrial cancer, 139–42 examinations, 71 haemorrhage, 76 history factors, 71 intraoperative complications, 74 malignant ovarian germ cell tumours (MOGCT), 193 minimal access surgery, 76–8 ovarian cancer, 123–6, 130 pleural eﬀusion drainage, 72 preoperative assessments, 70, 171, 177 sarcomas, 178 skin preparation, 73 special investigations, 71 thromboprophylaxis, 72 urinary tract, 75 vaginal cancer, 161 vulval cancer, 167–71 survival, 5, 151 symptom management, 103–17 symptoms of vulval cancer, 164 systemic therapies, 91–101, 145, 201 tamoxifen, 101, 133, 180 taxanes, 92

Index

teletherapy, see external-beam radiotherapy teratomas, 38, 176, 192, 193 term delivery, 181, 187 therapeutic laparoscopy, 77 therapeutic ratio/index, 87, 200, 201, 203 three-dimensional conformal radiotherapy (3DCRT), 89 three-dimensional ultrasonography, 24 thromboprophylaxis, 72 tissue tolerance, 85 tissue volume minimisation, 88 TNM, see Tumour, Node, Metastases (staging system) tolerance factors, 85 topo-isomerase inhibitors, 93 total abdominal hysterectomy, 139 total pain management, 108 toxicity, 99, 201, 204 trachelectomy, 78, 160, 203 transabdominal ultrasound imaging, 51 transitional cell carcinomas, 37 transrectal ultrasound imaging, 52 transvaginal ultrasound (TVS), 18, 23–4, 51 treatment aims of, 200–1 constipation, 112 follow-up, 61 hormonal, 100, 101, 129, 145 hypercalcaemia, 113 ovarian cancer, 123–6 patient factors, 202–5 planning, 59, 138–9 recurrent cancers, 145–6, 159–60 response, 64 selection, 201–5 triple incision, 168, 169 trophoblastic tumour, see gestational trophoblastic tumours tubal cancer, 15–21, 215, 216 tumour factors and decision making, 202 tumour markers CA125, 16–17, 52, 122–3, 193

human chorionic gonadotrophin, 183 malignant ovarian germ cell tumours, 195 ovarian cancer, 16–18, 122–3 ovarian germ cell tumours, 193, 194 tumour response, 64 tumour volumetry, 61 Tumour, Node, Metastases TNM (staging system), 2, 54, 167, 201, 213 TVS, see transvaginal ultrasound twin pregnancies, 190 UKCTOCS, see United Kingdom Collaborative Trial of Ovarian Cancer Screening UKFOCSS, see United Kingdom Familial Ovarian Cancer Screening Study UKTOCS, see United Kingdom Trial of Ovarian Cancer Screening ultrasonography, 24 ultrasound imaging, 51–4 cervical cancer, 53 endometrial cancer, 53, 65, 137 hydatidiform moles, 183 ovarian cancer, 52 ovarian germ cell tumours, 195 technique, 51–2 vulval cancer, 53, 67, 165 United Kingdom Collaborative Trial of Ovarian Cancer Screening (UKCTOCS), 18 United Kingdom Familial Ovarian Cancer Screening Study (UKFOCSS), 18 United Kingdom Trial of Ovarian Cancer Screening (UKTOCS), 52, 121 ureters, 75 urinary tract surgery, 75 uterus corpus uteri cancer, 216–17 FIGO staging, 216–17

leiomyosarcomas of the uterus, 175 malignant mesenchymal tumours, 41–4 sarcoma surgery, 178 uterine sarcomas, 179 vaginal bleeding, 148 vaginal cancer, 161 decision making, 204 epidemiology, 11–12 FIGO staging, 161, 220 pathology, 47 radiotherapy, 161 staging, 161 surgery, 161 vaginal hysterectomy, 141 vaginal intraepithelial neoplasia (VaIN), 47 Verres needle insertion, 77 verrucous carcinomas, 45, 149, 163, 173 villoglandular adenocarcinoma, 46 villoglandular endometrioid carcinomas, 39 VIN, see vulval intraepithelial neoplasia vincristine, 196 visual inspections, 26 vomiting, 108–11 vulval cancer, 163–73 aetiology, 163–4 clinical presentation, 164–6 computed tomography, 57 epidemiology, 11–12 FDG-PET, 65 FIGO staging, 166, 220–1 imaging, 53, 57, 62, 65, 67, 165, 166 investigations, 164–6 local treatments, 167 magnetic resonance imaging, 62, 67, 166 palliative care, 172 pathology, 47–8, 163 prognosis, 167 radiotherapy, 80, 169, 172 recurrent disease, 172 sentinel lymph node dissection, 171

233

Index

vulval cancer (cont.) signs, 164 staging, 166 superﬁcially invasive carcinoma, 169 surgery, 167–71 symptoms, 164 ultrasound imaging, 53, 67, 165

234

vulval intraepithelial neoplasia (VIN), 12, 47, 163 warty carcinomas, 45 weight loss, 114 World Health Organization (WHO) criteria of response, 207 performance scale, 207

screening criteria, 16 world standard population of Doll, 4 wounds palliative care, 104, 113–14 X-rays, 79, 138, 149 yolk sac tumours, 39, 192