Emerging Issues on HPV Infections From Science to Practice
Emerging Issues on HPV Infections From Science to Practice
Editor
Joseph Monsonego
Paris
36 figures, 14 in color, and 35 tables, 2006
Basel · Freiburg · Paris · London · New York · Bangalore · Bangkok · Singapore · Tokyo · Sydney
Joseph Monsonego, MD Head of Colposcopy Unit Institut Fournier 174 rue de Courcelles 75017 Paris France
Library of Congress Cataloging-in-Publication Data Emerging issues on HPV infections : from science to practice / editor, Joseph Monsonego. p. ; cm. Includes bibliographical references and index. ISBN 3-8055-8120-3 (hard cover : alk. paper) 1. Papillomavirus diseases. 2. Papillomaviruses. I. Monsonégo, Joseph. [DNLM: 1. Papillomavirus Infections–complications. 2. Uterine Cervical Neoplasms–etiology. 3. Mass Screening. 4. Papillomavirus Infections–diagnosis. 5. Uterine Cervical Neoplasms–prevention & control. 6. Viral Vaccines. WP 480 E53 2006] QR201.P26E44 2006 616.9⬘11–dc22 2006007074 Bibliographic Indices. This publication is listed in bibliographic services, including Current Contents® and Index Medicus. Disclaimer. The statements, options and data contained in this publication are solely those of the individual authors and contributors and not of the publisher and the editor(s). The appearance of advertisements in the book is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements. Drug Dosage. The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any change in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug. All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher. © Copyright 2006 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland) www.karger.com Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel ISBN 3–8055–8120–3
Contents
IX Foreword Monsonego, J. (Paris) High-Risk and Low-Risk HPV Infections: The Basic Differences 1 Biomarkers in Screening of Cervical Cancer von Knebel Doeberitz, M. (Heidelberg) 20 Epidemiology of Oncogenic and Nononcogenic HPV Types, and the Evidence for Differences in Their Sexual Transmissibility Burchell, A.N.; Franco, E.L. (Montreal) 34 Immune Responses to Genital HPV Stanley, M. (Cambridge) 44 Bridging the Communication Gap between Practitioners and Their Patients. Role of HPV and Advances in Cervical Cancer Prevention Savard, M. (Wynnewood, Pa.) Methods for HPV Detection 54 HPV Testing by Hybrid Capture Lörincz, A.T. (Gaithersburg, Md.) 63 Methods for HPV Detection: Polymerase Chain Reaction Assays Garland, S.M.; Tabrizi, S. (Melbourne)
V
73 Molecular Markers for Cervical Cancer Steenbergen, R.D.M.; Meijer, C.J.L.M.; Snijders, P.J.F. (Amsterdam) 82 Direct Detection of Cervical Carcinogenesis through mRNA Skomedal, H. (Klokkarstua); Kraus, I. (Oslo/Klokkarstua); Silva, I. (Klokkarstua); Molden, T. (Oslo/Klokkarstua); Hovland, S.; Morland, G.; Morland, E.; Karlsen, F. (Klokkarstua) HPV Testing and Patient Management 103 Human Papillomavirus Testing for Primary Cervical Cancer Screening Dalstein, V.; Bory, J.-P.; Graesslin, O.; Quereux, C.; Birembaut, P.; Clavel, C. (Reims) 120 HPV Testing in Patient Management: Atypical Squamous Cells of Undetermined Significance and Low-Grade Squamous Intraepithelial Lesion Cox, J.T. (Santa Barbara, Calif.) 140 Management of Women with High-Grade Squamous Intraepithelial Lesion and Atypical Glandular Cell Cervical Cytology Wright, T.C. (New York, N.Y.) Morphological Diagnosis and Treatment Decisions 147 Liquid-Based Cytology. The New Pap Test McGoogan, E. (Crawley) 157 Morphological Diagnosis. Histology Syrjänen, K.J. (Turku) 165 Morphological Diagnosis and Treatment Decisions. Colposcopy Singer, A. (London) 178 Basic Treatment Options for Cervical Intraepithelial Neoplasia and Warts Ferenczy, A. (Montreal) Prevention by Vaccines: Current Status, Impact and Prospects 184 Prevention of Cervical Cancer: Challenges and Perspectives of HPV Prophylactic Vaccines Monsonego, J. (Paris)
Contents
VI
206 Rationale for Human Papillomavirus (HPV) Vaccines Bosch, F.X. (Barcelona) 217 Perspectives on HPV Virus-Like Particle Vaccine Efficacy Schiller, J.T.; Lowy, D.R. (Bethesda, Md.) 227 Assessment and Follow-Up of HPV Vaccines Barr, E. (West Point, Pa.) 235 Cost-Effectiveness of HPV Vaccines Myers, E.R. (Durham, N.C.) 242 HPV Vaccination: Unresolved Issues and Future Expectations Barnabas, R.V. (Oxford/London); French, K.M. (London); Laukkanen, P.; Kontula, O. (Helsinki/Oulu); Lehtinen, M. (Helsinki/Oulu/Tampere); Garnett, G.P. (London) 253 Public Health Issues Related to HPV Vaccination Jenkins, D. (Rixensart) 266 How to Implement HPV Vaccines in Practice Harper, D.M. (Hanover, N.H.)
269 Author Index 270 Subject Index
Contents
VII
Foreword
All of us working in the field of cervical cancer obviously feel privileged among medical professionals, because during the past 20 years, we have had the opportunity of witnessing such an incredible breakthrough in our understanding of a major human disease, with significant impact on women’s health on a global scale. Indeed, looking back at the history of medicine, it is rare that such a major progress has been made in such a relatively short time period as has elapsed since the early 1980s, when the basic concepts on cervical cancer and its causes were elaborated. Despite this tremendous progress made along several different lines of research, it does not mean that cervical cancer has been overcome. Instead, the contrary is true. The disease continues to be the leading cause of cancer mortality among women worldwide and responsible for a significant annual morbidity, despite the increased knowledge and the substantial efforts made to prevent the disease by early detection and other measures pursued by national health authorities and international organizations like the World Health Organization, the International Agency for Research on Cancer and the European Research Organization on Genital Infection and Neoplasia (EUROGIN). Indeed, the unsatisfactory state of affairs was the single most important reason why EUROGIN was established in the early 1990s. In alignment with its core mission, EUROGIN continues to pursue a wide variety of educational and training activities, all aiming at the widespread distribution of accurate and state-of-art information in the field of cervical cancer prevention, in the widest sense of the word. The latest outcome of these activities is presented in this new book.
IX
This book is a concise update of the current knowledge on HPV infections and their intimate links to cervical cancer, provided by the basic, epidemiological and clinical research, and provides valuable implications in therapy and, most importantly, in the prevention of cervical cancer by prophylactic HPV vaccination. Thus, the book represents an important contribution to continuing medical education, and as such, aimed to be both convenient and educational in its very nature. It is written by foremost international authorities who share their experience and summarize the current state-of-art knowledge on the highly specialized topics in research, diagnosis and management of these conditions, aiming to reach a multidisciplinary readership consisting of health care professionals and representing divergent specialities coping with HPV-associated diseases at genital and extragenital sites. The main focus of the text is on five different but interrelated topics: (1) high-risk and low-risk HPV infections, (2) methods for HPV detection and use of molecular markers, (3) HPV testing in primary screening and patient management, (4) morphological diagnosis and treatment, as well as (5) prevention of cervical cancer by vaccines: current status, impact and prospects. In the era of HPV vaccine development, a lot of confusion still exists among physicians and patients in regard to HPV infections in general. One of the aims of this book is to help elucidate these controversies by emphasizing the importance of distinguishing the two types of diseases, i.e. those induced by low-risk HPV and those due to high-risk (oncogenic) HPV types. Throughout the book, the authors highlight these differences, while describing the HPV biology and carcinogenesis, as well as issues related to infection, epidemiology, natural history, clinical features, treatment and prevention of the infections. Cervical cancer is caused by infections with a range of high-risk (oncogenic) HPV types which differ from the low-risk HPV types in many of their key biological characteristics. Indeed, molecular markers targeted at specific intracellular pathways hold great promise to become important tools not only in dissecting the molecular pathways involved in HPV-induced carcinogenesis, but also in increasing our understanding of the fundamental differences between the low-risk and high-risk HPV types in causing the human disease. Similarly, infections due to oncogenic HPV types demonstrate a different epidemiology and natural history from those of the low-risk HPV genotypes. Some of the differences between these two categories of viruses are likely to be due to their different immunological recognition by the infected host. Because of the huge morbidity and potential risk of cancer associated with oncogenic HPV infections, it is essential to make a distinction between the low- and high-risk viruses in the communication between the practitioners and their patients. In the same way, proper attention should be paid to providing adequate diagnostic and
Foreword
X
management services to populations at particularly high risk of contracting oncogenic HPV infections and prone to disease progression towards invasive cancer. The principles of the different HPV detection techniques and the recent developments in the field of HPV diagnosis are addressed by several authors. The Hybrid Capture technique has an established position as a cornerstone diagnostic tool, subjected to extensive testing in different settings (both screening and diagnosis) during the past several years. Also PCR-based technology has undergone major technical development towards more user-friendly applications that currently compete with Hybrid Capture 2 assay as routine diagnostic tools in HPV detection and typing. The past few years have witnessed a remarkable explosion of research focused on molecular markers that would accurately predict the outcome of HPV infections and their associated clinical disease at the level of individual patients. Apart from methods based on HPV DNA detection, assays analyzing HPV RNA expression in clinical lesions have become commercially available as well and are currently under vigorous testing in different diagnostic and screening settings. Cervical cancer has an unequal geographic distribution, with the highest global disease burden confined to the developing countries, where the facilities to combat the disease are clearly insufficient. On the other hand, the declining trends in incidence and mortality rates witnessed in many of the developed countries during the past 4 decades are mainly attributable to the implementation of organized screening programs based on the use of cervical Pap smear, e.g., in the Nordic Countries, where an organized screening has resulted in up to 80% reduction in cervical cancer incidence since the early 1960s. Unfortunately, these highly effective organized screening programs exist in few countries only, and the prospects for effective cervical cancer screening based on the Pap test in the majority of the developing countries seem gloomy, if not entirely pessimistic, even in the foreseeable future. This fact has been well appreciated among the scientific community and has led to an extensive search for optional screening tools, currently under testing in different countries. One of these optional screening tools is HPV testing, that enables to target the screening directly to the key etiological agent of cervical cancer, instead of the clinical precursor (CIN) lesions detected by the Pap test. Adequate early detection by screening is the prerequisite for effective management of the cervical lesions, which follow different algorithms for low-grade and high-grade abnormalities. On the other hand, no screening is meaningful unless adequate facilities for proper treatment are available, which is not the case in many of the poorest developing countries. One of the main focuses in the future research is the development of effective prophylactic HPV vaccines and their testing for applicability in general use
Foreword
XI
in populations at high risk of cervical cancer. The current status, impact and prospects for cervical cancer prevention by vaccines are addressed by several authors in this section. The texts following an in-depth review on this subject by the editor are dealing with different aspects, including the discussion about the rationale of HPV vaccines, their immunogenicity and safety, as well as the efficacy and its assessment and follow-up. Cost-effectiveness of the vaccination programs plays an important role, particularly in the developing countries. In addition, HPV vaccination holds great promise to further reduce the disease burden of cervical cancer, even in countries where organized screening programs are effectively implemented. In these countries, the vaccination strategies will most probably be different and have important implications on the execution of the existing screening programs, with a great potential to reach substantial cost savings, e.g., by extending the screening intervals and changing the target age groups. However, there are still several issues to be solved before such programs are ready to be implemented. This applies differently to high- and low-resource settings. These include important public health issues related to HPV vaccines, and several more specific unsolved issues that need to be addressed before the measures for implementation can be undertaken. By compiling this book, EUROGIN intended to bring these topics into general awareness of colleagues who are not able to attend the regular EUROGIN educational and training events. Cervical cancer is our common target, and the global control of this major disease burden necessitates a joint effort of a wide spectrum of medical and paramedical expertise. We sincerely hope that you will find this information helpful in your daily practice and other activities aiming at this goal. According to its mission, EUROGIN continues pursuing these efforts by distributing timely information and by active training of various groups of health professionals in these traditional educational events and scientific meetings, of which more information is available at our website (www. eurogin.com). J. Monsonego Paris
Foreword
XII
High-Risk and Low-Risk HPV Infections: The Basic Differences Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 1–19
Biomarkers in Screening of Cervical Cancer Magnus von Knebel Doeberitz Department of Applied Tumor Biology, Institute of Pathology, University of Heidelberg, Heidelberg, Germany
A decline in the incidence and mortality of cervical cancer has been observed in most of the Western countries since the first third of the past century. This may be partially attributed to improved hygienic and living conditions. However, the introduction of a cytological screening test (Pap test) that was developed by George Papanicolaou in the 1940s [1] had a significant impact on this positive development. Since the 1970s, population-wide screening programs were implemented in many Western countries, and this further accelerated the decline in the incidence and mortality of cervical cancer, despite an increasing incidence of precursor lesions over the same time period [2]. These data impressively demonstrated how successful early detection and prevention programs for cancer may work. However, there are major limitations associated with the Pap test that are primarily attributed to the rather complex infrastructure required to establish population-wide screening programs but also to the extensive training required to perform the test and read the slides appropriately. Thus, many countries, particularly in the developing world, could not afford to establish these programs and are still confronted with an exceedingly high incidence and mortality of this disease. On the technical level, the subjective interpretations of scoring criteria that are used to classify the cytological samples were and still are associated with a significant lack of reproducibility of test results [3]. This causes substantial secondary costs due to repeated testing and further clinical work-up of patients with ambiguous test results and may also lead to psychological as well as physical distress of the affected women [3, 4]. Due to these limitations of an otherwise very successful concept in cancer prevention, substantial research efforts were undertaken to better understand the molecular events involved in cervical
carcinogenesis and to delineate novel screening methods. The purpose of this chapter is to summarize the critical molecular mechanisms that contribute to HPV-mediated cervical carcinogenesis, review the current status of novel diagnostic technologies and biomarkers, and to discuss the available data on clinical applications and future premises of the most promising novel markers and technologies that result from this work.
Role of HPV in Cervical Carcinogenesis
There is now no further doubt that persisting infections with about 13 different high-risk HPV (HR-HPV) types are the initial prerequisite to induce most cervical cancers [5]. However, infections with these viruses are very widespread also among healthy women. The infection is usually self-limited and transient without causing cellular transformation or dysplasia, and therefore, HPV infections represent an important and necessary but by far not sufficient risk factor. It is estimated that the cumulative risk of such infections in 70 years of life of an average women is more than 70% [6]. Only very few of these women will ever develop a persistent infection that may last for longer than 6–12 months and may pose a significantly increased risk of the subsequent development of cervical lesions. To establish an infection, HPVs apparently require access to the basal and parabasal cell layers of the epithelium, or in case of the cervix, to cells located in the transformation zone (fig. 1). To replicate their genomes and successfully produce new infectious virions, the host cells require a certain degree of terminal differentiation. It appears that in the replication-competent basal and parabasal cells, only very little if any gene expression activity of the virus can be observed [7]. This strategy to avoid viral gene expression and replication in epithelial stem cells but to permit it in differentiated cells that are determined to die because of their physiological differentiation processes is a very elaborate strategy to permit maximal production of new infectious viral capsids causing almost no damage to the infected host. This situation changes dramatically if the regulatory intracellular features that control the fine-tuned expression control of the viral genes along with the differentiation processes of the epithelium are disturbed and deregulated uncontrolled expression of genes involved in the replication of the viral genome suddenly occur in epithelial cells that have not yet reached the irreversible status of terminal differentiation. In this situation, interference of viral genes with those pathways that control the replication of the epithelial cell and life cycle may result in the initiation of chromosomal instability (fig. 1). The genetic functions of the virus that contribute to the induction of chromosomal instability have been very well documented by a long series of very
von Knebel Doeberitz
2
Latent infection
Normal
Replicative infection
CIN1
Persistent, deregulated infection
CIN2/3
Fig. 1. Schematic representation of the natural history of HR-HPV infections. The virus infection gets access through scratches, scars or at the transformation zone, even directly to epithelial cells within the basal and parabasal cell layers. Here, it establishes a latent infection, i.e. the viral genome replicates along with the host cell, and only very few genome copies are generated once the host cell divides into the daughter cells. In this stage, there is only marginal viral gene expression and the virus does not harm or destroy the host cells. However, if infected epithelial cells reach terminal differentiation, i.e. the stage where the capacity to proliferate is irreversibly lost, the viral genes may be strongly expressed and the replication cycle of the virus is initiated until finally mature viral capsids are release at the surface of the epithelium. Transformation of the epithelial cell may only occur in persisting infections in that the molecular mechanism that prevents expression of viral genes in the immature basal and parabasal cells is lost. In this condition, viral genes that may induce chromosomal instability are coexpressed with functions that maintain ongoing proliferation of the immature epithelial stem cells. These cells may consequently undergo mitotic defects and, due to the accumulation of further genomic alterations, finally progress to invasive carcinomas. Hence, the initiation of the deregulated type of the viral gene expression pattern can be regarded as the initial event in cervical carcinogenesis. Due to the increasing alteration, viral genomes may also integrate; however, the current evidence suggests that this is a consequence of substantial genomic alterations, and therefore, rather a sign for neoplastic progression.
detailed molecular and biochemical studies. Three major aspects are involved: (1) the E6 protein of the oncogenic HPV types supports premature degradation of the p53 tumor suppressor gene and thus interferes with apoptotic functions [8], (2) the E7 protein induces destabilization of the retinoblastoma protein complex and thus allows the cell to evade cell cycle control through the pRB pathway [9], and (3) both genes induce substantial disturbances of the mitotic functions by interfering with centrosome synthesis and function that results in desegregation of the chromosomes during mitosis and numerical and structural chromosomal aberrations [10].
Biomarkers and Cervical Cancer
3
Taken together, cervical carcinogenesis is induced by the deregulated expression of the viral E6 and E7 oncogenes in the basal and parabasal cell compartment. This is clearly an exceptional molecular accident during a normal viral life cycle and, in view of the many HPV-infected cells within the genital epithelium, an extremely rare event. For still unknown reasons, the cells of the transformation zone of the uterine cervix appear to be particularly sensitive to these events. This is clinically well reflected by the relative incidence of the HPV-associated cancer or the respective precursor lesions in the human genital tract, where lesions of the cervix occur at least ten times more frequently as compared with vaginal, vulvar or penile cancers. Aims and Scope of Novel Cervical Cancer Screening Technologies Cervical cancer emerges through a series of precursor lesions that are classified as cervical intraepithelial neoplasia (CIN) grades 1, 2 and 3. These preneoplastic lesions retain the capacity to spontaneously regress to an extent that depends on the degree of the dysplastic changes (fig. 2) [11]. Classically, the progression of the lesions has been regarded as a linear process. However, clinical data suggest that the CIN1 lesions usually have a relatively small risk and require, on average, a rather long latency time before they progress to invasive cancer. Therefore, they are also referred to as low-grade squamous intraepithelial lesions (LSIL). However, the progression risk of CIN2 and CIN3 lesions is more evident, and they are regarded as high-grade squamous intraepithelial lesions. This dual classification system appeared to be particularly useful in the classification of the cytological changes that reflect the cervical preneoplastic lesions, since it allowed a more simplified communication between cytopathologists and clinicians and became the basis of the Bethesda system for reporting cervical cytology [12]. However, a dual classification system would also fit theoretical aspects, as the molecular events that provoke changes that impress as CIN1 lesions may principally differ from those that impress as CIN2, CIN3 or invasive cancers, since the latter lesions are characterized by the deregulated uncontrolled expression of viral genes in the basal and parabasal cells, whereas in most of the CIN1 lesions, this uncontrolled viral gene expression pattern did not yet emerge. The current concept of cervical cancer screening primarily aims to identify patients who have high-grade squamous intraepithelial lesions or lesions ⱖCIN2. These patients are believed to require immediate medical intervention (i.e. surgical removal of a high-grade lesion by loop electrosurgical excision procedure or conization) due to the significantly elevated risk of the subsequent development of invasive cancers in comparison with the other earlier lesions. Theoretically, one may therefore differentiate three levels of risk of the
von Knebel Doeberitz
4
HPV replication Latent
Productive HPV infection
Acute
Persistent
Type of HPV infection
E6/E7 off/p16INK4a⫺ Regulated
Normal
E6/E7 on/p16INK4a⫹ Deregulated HPV gene expression
CIN1/LSIL
Risk level 1
CIN2/3 HSIL
Risk level 2
CIS/cancer
Risk level 3
Episomal HPV genomes Integrated HPV genomes
Fig. 2. Schematic representation of the progression of HPV-infected epithelial lesions to invasive cervical carcinomas. As outlined in figure 1, the deregulated expression of the viral oncogenes E6 and E7 in basal and parabasal cells initiates chromosomal instability and thus the transformation process. This is associated with the strong overexpression of the p16INK4a protein in the basal and parabasal cell compartment. Increasing genomic instability may finally permit the integration of viral genomes; however, this is observed rather late in the progression cascade. The risk of progression and thus the chance of spontaneous regression increases with increasing severity of the lesion that directly reflects the status of genomic instability which is already reached by the cells within the lesion. CIS ⫽ Carcinoma in situ; HSIL ⫽ high-grade squamous intraepithelial lesions.
development of cancers: (1) infections with oncogenic HPV types, (2) emergence of cell clones with deregulated viral oncogene expression and thus initiated chromosomal instability in the basal and parabasal cells, and (3) progression of these cell clones to cell populations with a significantly increased level of chromosomal instability that goes along with the progression to invasive fullblown cancer cells. The major aim of cancer screening programs is to identify patients who would benefit from medical intervention, without concerning individuals who would not benefit from medical intervention. In most women who are HPV
Biomarkers and Cervical Cancer
5
infected but have not yet developed clinically relevant lesions (risk level 1), information about their HPV infection may cause substantial concern, although nothing can be done about it. Based on these conceptual considerations and given the high number of HPV-infected individuals, the best and presumably also most cost-effective target for cancer early detection assays appears to be the cell population that just acquired chromosomal instability and was thus initiated for transformation (represented by risk level 2). These are the cells that evaded the control of the viral oncogenes in the basal and parabasal cells. Specific and sensitive markers that would allow to exactly identify those cells hold the greatest promise for improving cervical cancer screening techniques. So far, the research to identify useful markers is based on three different approaches: (1) analysis of genomic or epigenomic alterations of the host cell chromosomes in cervical dysplasia and cancer that comes along with the E6/E7induced chromosomal instability, (2) gene expression profiling of replicating cells that express the viral oncogenes, and (3) analysis of protein expression profiles that occur as direct or indirect consequence of the deregulated expression of the viral oncogenes. In the following, we aim to review the current status of this research. We will restrict the discussion to markers that have reached some level of clinical applicability and have been evaluated in clinical studies.
Biomarkers to Identify Epithelial Cells with Deregulated Viral Oncogene Expression
Structural Chromosomal Aberrations As outlined above, the deregulated expression of the viral oncogenes E6 and E7 induces chromosomal instability and results in loss and gains of chromosomal material of the affected cells [13]. Although no absolutely consistent chromosomal aberrations have been so far identified, a systematic search for more prevalent chromosomal aberrations in HPV-transformed cells found a significant gain of material of the chromosome 3q during the transition of highgrade dysplastic lesions into invasive carcinomas [14–17]. Although gain of 3q was the most consistent genomic aberration yet identified by comparative genomic hybridization, its rather late occurrence during the preneoplastic progression cascade does not suggest that monitoring cells sampled from cervical swabs for gain of 3q may be a relevant diagnostic marker for primary cervical cancer screening. The current evidence suggests that the detection of gain of 3q may better serve as a progression marker for cervical lesions. However, the extent of clinical data that allow to define the clinical use of this marker is still rather limited.
von Knebel Doeberitz
6
Integration of HPV Genomes As a consequence of chromosomal instability induced by deregulated viral gene expression, viral genomes or fragments thereof may become integrated into the host cell chromosomes. There is no evidence for particular specific integration loci within the host cell genome, although fragile sites are apparently preferred as integration sites [18]. In contrast, the viral genome reveals few highly characteristic features. Most importantly, in all analyzed cervical carcinoma cells, a cassette consisting of the viral promoter and enhancer element located in the upstream regulatory region, and the genes E6 and E7 are, at least in one copy, retained intact upon integration. This cassette is consistently transcribed, and several lines of evidence show that fusion of the viral sequences with cellular sequences that are cotranscribed at the 3⬘end of the transcript favor the stability of these transcripts, resulting in a higher oncogenic potential [19, 20]. Consequently, it is speculated that integration of the viral genome and expression of these integrated viral genome fragments further enhance the neoplastic progression of the respective cell clones. This hypothesis is in good agreement with the clinical finding that the transition of highgrade dysplasia to invasive cancer is associated with a significant increase in integrated viral genomes in the host cell chromosomes [21]. This is consistent with the observation that the number of chromosomal aberrations appears to increase with progression from preneoplastic lesions to invasive carcinomas and suggests that a significant increase in recombination events occurs during progression. In terms of diagnostic applications, detection of integrated genomes in preneoplastic lesions is clearly a sign for advanced and most likely rapidly progressing dysplasia [21]. However, the detection of integrated viral genomes with whatever assay is clearly not suited for screening purposes if level 2 of the risk cascade is to be targeted by the screening assay. Practically, in situ hybridization (ISH) techniques may be helpful to monitor the physical status of the HPV genomes in cytological samples or biopsies; however, ISH is hampered by several technical flaws, in particular superposition of episomal HPV genomes and cellular chromosomes that may mimic integrated viral genomes. Recently, some novel ISH techniques have been improved and are able to more reproducibly highlight the integration status of cervical lesions [22]. More specific methods, as for example direct amplification of integrated DNA fragments using detection of integrated papillomavirus sequences by ligation-mediated PCR [23], or RNA-based methods, as for example the amplification of papillomavirus oncogene transcript reverse-transcription polymerase chain reaction, appear to be reasonable alternatives [18, 21, 24]. These methods take advantage of the fact that integrated viral genomes are linked to cellular sequences. Coamplification of cellular and viral sequences by these methods points to covalently integrated viral genomes. Since the integration
Biomarkers and Cervical Cancer
7
locus of the HPV genome in a given cancer cell clone creates a unique molecular marker for the tumor, detailed analysis of the integration locus may provide a very sensitive tool to monitor for residual disease in postsurgery surveillance of patients [25]. This may allow for an earlier diagnosis of relapsed disease as compared with the conventional cytological techniques [26]. More recently, epigenetic modifications of the genome of HPV-infected cells are considered as diagnostic marker for cervical dysplasia or cancer. For example, DNA methylation refers to the addition of a methyl group to the cytosine ring of a cytosine that precedes a guanosine (referred to as CpG dinucleotides) to form methylcytosine (5-methylcytosine). Usually, DNA methylation plays a role in maintaining genome stability and in regulating gene expression [27]. Global hypomethylation and hypermethylation of CpG clusters (referred to as CpG islands) present in the promoter region of multiple genes have been associated with malignancy [28]. Hypermethylation in a promoter region is associated with ‘gene silencing’, i.e. inhibiting expression of a gene that is normally expressed in the absence of methylation. Several animal and clinical studies have demonstrated that these epigenetic methylation changes are an early event in carcinogenesis and are often present in the precursor lesions of a variety of cancers. Aberrant methylation of various genes has also been observed in cervical cancer [29–32]. Although the sensitivity of this approach still appears to be unsatisfactory, the specificity to detect histologically confirmed CIN3 seems to be quite high. These data suggest that the analysis of methylation patterns of certain genes may be a valuable tool in forthcoming screening programs; however, here again, they appear more likely to play a role as progression markers for cervical precancer, indicating risk level 3 rather than risk level 2 (i.e. the early initiating events of cervical carcinogenesis), as outlined above. Changes of the Gene Expression Profile To analyze molecular events associated with carcinogenesis, comprehensive analyses of mRNA expression profiles have been extensively used. These studies revealed several genes significantly altered in cervical cancer development, a number of them associated with proliferation functions and DNA metabolism of cells [33]. A different set of genes was found to be particularly associated with progression of CIN lesion to more advanced dysplasia or invasive carcinomas [34]. Among the proliferation-associated genes were mcm5, mcm7, cdc6, claudin 1, topoisomerase II␣, NET1/C4.8, survivin, and several others [34, 35]. A cocktail of antibodies detecting the individual gene products is now proposed to highlight dysplastic cells; however, clinical studies in which such marker cocktails have been evaluated for their clinical applicability have not been reported so far. Taken together, although tremendous efforts have been
von Knebel Doeberitz
8
made to identify differentially expressed genes that may serve as adequate early detection markers, by this approach, no marker that is uniformly expressed could be identified up to now. This suggested that a more evidence-based approach, eventually in combination with array technologies, may be more effective in the identification of suitable biomarkers. Evidence-Based Approaches to Identify Differentially Expressed Genes in Dysplastic Cervical Cells Cervical cells present on a smear prepared from a cervical swab usually predominantly encompass differentiated epithelial cells that have undergone cell cycle arrest and are not very likely to express high levels of proteins involved in active DNA metabolism or the cell cycle, unless proliferating dysplastic cells reach the surface of the area from where the cells for the sample were taken. This suggests that the group of proliferation-associated gene products may serve as a potential source for candidate markers. Accordingly, proliferation-associated antigens like Ki-67 or PCNA have extensively been used in an attempt to identify dysplastic cells with improved sensitivity and specificity. However, these attempts were notoriously hampered by the fact that normal basal cells as well as primarily nonepithelial cells like lymphocytes and other mononuclear cells expressed high levels of these proliferation-associated markers [36]. To render this approach more specific, antibodies against proteins (cdc6 and mcm5) that are sequentially assembled into a prereplicative complex or ‘replication license’, which is essential for the initiation of DNA replication, were analyzed in greater detail. Although this attempt seemed to work successfully [37], clinical data did not confirm the initial enthusiasm due to the fact that too many other cells expressed comparable high levels of these proteins. Several cyclins were considered as candidate markers [38] and in particular cyclin E appeared to be a promising candidate [39, 40]. However, more extensive analyses could not confirm the initially promising results and showed that these markers did not meet sufficient sensitivity and specificity to serve the demands required in cytological screening. A further promising candidate was a novel tumor antigen that appeared to be specifically expressed in the transformed cells of the uterine cervix [41]. This gene encodes a carboanhydrase (MN/CA9) and, apparently, is also expressed in cells experiencing competitive nutrition conditions or hypoxemia [42]. Detection of this antigen in cells on cervical smears of liquid-based cytology samples seemed to improve the detection rate, particularly of glandular lesions, and be of some value in the stratification of the atypical squamous cells of undetermined significance (ASCUS) category [43, 44]. However, further detailed studies revealed that by far not all dysplastic cells express this antigen and its diagnostic value was not confirmed in further clinical trials.
Biomarkers and Cervical Cancer
9
RB
Promoter
E2F
E7
p16INK4a
RB
Promoter
E7
E2F
p16INK4a
Fig. 3. Schematic representation of the mechanism that induces p16INK4a overexpression in proliferating epithelial cells that express the HR-HPV E7 protein. For details, see text.
A further resource for potential useful biomarkers may be represented by genes whose expression is related to the specific interferences of HR-HPV oncogenes E6 and E7 and their host cells. The E6/E7 oncogenes are required to induce and maintain neoplastic growth of cervical cancer cells [45–47]. Thus, genes that are normally not expressed in cervical epithelia but specifically upregulated by the action of these two oncogenes might provide an alternative source for potential candidates. One of these candidates is the gene coding for the cyclin-dependent kinase inhibitor p16INK4a [for a review, see ref. 48]. This gene appeared particularly attractive as candidate marker since the interaction of the E7 oncoprotein with the pRB pathway uniformly results in premature degradation of the pRB complex. pRB inhibits the transcription of p16INK4a [42, 49]. p16INK4a is a very strong negative regulator of the cell cycle; thus, cells that express this antigen under physiological conditions are determined to irreversibly arrest the cell cycle and to subsequently undergo apoptosis [50]. If this mechanism is lost, e.g., the action of a HR-HPV E7 protein, the respective cells start to overexpress p16INK4a in very high levels (fig. 3). p16INK4a was initially described as a potent tumor suppressor in many cancers, as for example lung, breast and colon cancers [51], the expression of which is frequently inhibited by hypermethylation of its promoter or, less frequently, by loss of the gene locus [52]. The initial analysis of the p16INK4a status in cervical carcinomas showed equivocal results. While some groups
von Knebel Doeberitz
10
a
b
c
d Fig. 4. Distinct p16INK4a staining patterns. Focal staining (a–c): metaplasia (a, b), and CIN1 lesions caused by a low-risk HPV infection (c). In contrast, diffuse staining in a CIN1 lesion caused by HPV16 (d). Note the lack of p16INK4a reactivity in the basal and parabasal cell layers (a–c), where some reactivity is observed in the differentiated part of the epithelium. d CIN1 lesion in which the diffuse staining pattern occurred in the basal and parabasal cell layers, indicating activated HR-HPV oncogene expression. Arrows indicate the basal cells of the epithelium.
found hypermethylation or even deletion of the gene [53, 54], others did not find any alterations of the p16INK4a locus or its gene sequence [55]. A comprehensive analysis of the expression pattern in a larger series of dysplastic lesions of the cervix and also invasive carcinomas revealed that p16INK4a is strongly overexpressed in a diffuse staining pattern in the proliferating parts of the dysplastic epithelium, whereas normal epithelium does not show p16INK4a immunoreactivity (fig. 4a) [56, 57]. The diffuse staining pattern was also found in glandular lesions and allowed a significantly better assessment of glandular dysplasia (fig. 2) [56–64].These data suggested that p16INK4a might fulfill most of the criteria that are required to identify cells or lesions that just entered risk level 2 and may thus represent a useful biomarker for cervical cancer screening because it indicates the deregulated expression pattern of the
Biomarkers and Cervical Cancer
11
viral oncogenes in the basal and parabasal cells. In good agreement with this hypothesis, a couple of studies revealed that among the low-grade lesions (CIN1), expression of p16INK4a in basal or parabasal cells indicates those that have an substantially increased risk of progression to high-grade lesions in comparison with the corresponding CIN1 lesions that do not yet display the diffuse staining pattern (fig. 2) [65–67]. In some instances, intermediate or superficial epithelial cells express some p16INK4a if disturbances of the normal differentiation pattern occurs in the cells (fig. 4). This phenomenon is especially observed in some metaplastic lesions, as well as in atrophic lesions or in some cases of condylomatous lesions associated with low-risk HPV types [68]. However, this staining pattern can easily be distinguished from the diffuse staining of the transformed cells in the basal and parabasal cell layers that is highly specific for the HPV-transformed epithelial cells that express the viral oncogenes. A first practical application of this novel concept is that the reproducibility and therefore the reliability of the histopathological diagnosis appears to be significantly improved by the use of p16INK4a as a biomarker [69]. p16INK4a Immunocytochemistry as Screening Tool in Cervical Cytology Stimulated by the promising data on tissue sections, several groups started to develop protocols to stain dysplastic cells in liquid-based cytology samples as well as in conventional cervical smears [70–80]. Further studies showed that the use of p16INK4a cytochemistry in cervical cytology in patients who had cervical biopsies following a cytologic diagnosis of LSIL or ASCUS was associated with an improved positive predictive value for biopsy-confirmed CIN2 and CIN3 if compared with the detection of HR-HPV genomes by the hybrid capture assay (hybrid capture 2). Therefore, the use of p16INK4a immunocytochemistry may be particularly helpful as adjunct test in the triage of patients with LSIL and ASCUS cytology [81]. However, other authors speculate that the use of p16INK4a may even be a sufficient biomarker for primary cervical cancer screening strategies [80]. Despite the very high specificity of p16INK4a for transformed cells, there are special conditions in that also nondysplastic cells stain for p16INK4a in histology as well as cytology specimens. These p16INK4a-expressing cells encompass some desquamated endometrial cells as well as atrophic or metaplastic epithelia. The expression of p16INK4a in these cells is due to irregularities in the normal differentiation pattern, and the cells aim to arrest their cell cycle or may even initiate apoptosis. In contrast to the histology setting, the cellular context is conserved and facilitates the discrimination between dysplastic cells (diffuse staining involving the basal and parabasal cell layers) and nondysplastic cells
von Knebel Doeberitz
12
(focal staining, varying levels of the epithelium, but primarily in the differentiated superficial cell sheets). In contrast, in cytology specimens, the cellular context is lost and makes the assessment of p16INK4a-positive cells more difficult, requiring the indepth analysis of the cellular and nuclear morphology of the cells [68]. It is important to consider that the nontransformed cells that express high levels of p16INK4a are irreversibly cell cycle arrested. Thus, simultaneous staining, e.g. with antibodies directed against proliferation-associated markers like Ki-67 or MIB1, always gives negative results. This is in sharp contrast to the HPV-transformed cells that overexpress p16INK4a. These constitute an actively proliferating cell population that strongly expresses both markers, p16INK4a and proliferation-associated antigens like Ki-67. Therefore, staining cells in a cytological sample with both antibodies unambiguously allows to identify the HPVtransformed cell population [Trunk et al., in preparation]. Similar to the fact that not necessarily all p16INK4a-positive cells need to be HPV-transformed cells, there may also be cytological specimens in that clearly abnormal dysplastic cells may remain p16INK4a negative. Again, this finding is not surprising if one considers the histological staining pattern of cervical lesions [68]. In those lesions that still retain a certain degree of epithelial differentiation, p16INK4a expression is downregulated once the cells loose their proliferation capacity and start to undergo terminal differentiation. If those cells contain abnormal nuclei, they will appear as p16INK4a-negative but abnormal cells in the cytology samples. In practice, this does not seem to be a problem, since virtually all cytological samples that display those p16INK4a-negative but abnormal cells also display plenty of intensively p16INK4a-stained abnormal cells. In practical terms, there are several criteria that may help to correctly interpret the p16INK4a staining pattern in cervical cytology samples. In a recent study, the number of p16INK4a-positive cells per 1,000 cells from cervical swab specimens was used to discriminate dysplastic from nondysplastic changes [78]. Alternatively, p16INK4a-stained cells can be scored according to conventional cytological criteria of nuclear abnormalities, such as nucleocytoplasmic ratio, hyperchromasia, nuclear border abnormalities, anisokaryosis and structural aspects of the nuclear chromatin (coarse or fine granulated chromatin) (fig. 5). To standardize these morphological criteria, a nucleomorphological score was proposed that may facilitate the assessment of p16INK4astained cells [82]. If these criteria were applied to a series of samples including all diagnostic categories from normal to high-grade dysplasia, the p16INK4a staining pattern was compared with the score obtained by conventional Pap stains, and an excellent agreement was achieved.
Biomarkers and Cervical Cancer
13
a
b
c
d Fig. 5. Examples for p16INK4a-positive cells. a None of the nuclear score criteria applies. b Cells display a increased nuclear cytoplasmic ratio. c Cells display an increased nuclear ratio and hyperchromasia. d Cells display an increased nuclear cytoplasmic ratio, hyperchromasia, altered nuclear shape and anisokaryosis.
Advantages Achieved by the p16INK4a Biomarkers in Cervical Cytology Screening for dysplastic cells in cervical smears generally encompasses two different tasks: (1) the screen for few potentially dysplastic cells in a huge excess of normal cells, and (2) the interpretation of the degree of dysplasia of the identified potentially dysplastic cells. These two distinct processes are referred to as location and interpretation function. In the assessment process of conventional Pap stains, these two tasks have to be performed in a parallel process, since it is not feasible to go back to a suspicious cell in a second turn, after the complete slide has been evaluated. Thus, in the observer’s mind, both tasks (location and interpretation function) are performed in parallel. This is a very important source of subjectivity and misinterpretation. The use of a biomarker allows dissecting these two functions. First, all potentially stained cells can be identified, and in a second step, the detailed morphology can be analyzed in detail. This allows a much more efficient scoring
von Knebel Doeberitz
14
process, since the vast majority of slides can be only scored for the presence or absence of stained cells. Second, the few cells that display staining can be subjected to a detailed morphological analysis. This simple dissection of the location and interpretation function is expected to significantly improve the reproducibility of the cytological reading process. Using p16INK4a as a locator tool followed by the nucleomorphic analysis of the stained cells allows locating dysplastic cells on a given slide. A further significant advantage of the use of p16INK4a as a locator tool will become evident if automatic screening of slides using computerized image analysis will be feasible. Here, the automatic screening for stained cells does not have to perform complex interpretation functions based on analysis of nuclear parameters in parallel to the screening process. In contrast, automatic screening can focus on a single parameter process, i.e. searching for stained cells in the slide. The detailed interpretation of these cells can be performed in an independent second process, either by cytometry or by a cytopathologist who may have a direct look through the microscope or at an image of the stained cells on a computer screen. The data that are so far available and that have been summarized here suggest the use of p16INK4a might be more sensitive and more specific than the conventional Pap test to detect high-grade cervical dysplasia. It is expected that the p16INK4a-based approach has a higher intra- and interobserver agreement in assessing the slides, since the reading is less complex than that of Pap-stained samples. Moreover, this approach will overcome many of the conceptual and practical problems associated with the detection of HR-HPV types, either as an adjunct test or as a potential primary screening tool. This already became evident in studies where the p16INK4a was used to triage low-grade cytology samples. The positive predictive value to identify high-grade lesions among samples that were scored as LSIL in the conventional Pap stain was significantly better than that of the HR-HPV test [81, 82]. Taken together, these data demonstrate that p16INK4a has indeed a very high potential to fulfill the criteria for a useful biomarker in primary cervical cancer screening. Although the data collected clearly do not answer all questions about the clinical value of p16INK4a as a potential biomarker, the present evidence strongly suggests that the use of p16INK4a as a biomarker in cervical cancer screening will help overcome many problems associated with the poor reproducibility of the Pap test and still retains a tremendous potential for automation. These aspects suggest that there will be a significant number of technical improvements in the forthcoming years that will ultimately lead to reduced cost, better reliability for the gynecologists or primary care physicians and, last but not least, significantly higher safety for the patients.
Biomarkers and Cervical Cancer
15
References 1 2 3 4 5
6 7 8 9 10 11 12 13
14
15
16
17 18
19 20
21
Papanicolaou GN, Traut HF: The diagnostic value of vaginal smears in carcinoma of the uterus. Arch Pathol Lab Med 1997;121:211–224. Anderson CM, Thornton JG: Screening for cervical cancer. Graphs may mislead. BMJ 1994;309: 953–954. Stoler MH, Schiffman M: Interobserver reproducibility of cervical cytologic and histologic interpretations: realistic estimates from the ASCUS-LSIL Triage Study. JAMA 2001;285:1500–1505. Solomon D: Chapter 14: role of triage testing in cervical cancer screening. J Natl Cancer Inst Monogr 2003;31:97–101. Munoz N, Bosch FX, de Sanjose S, Herrero R, Castellsague X, Shah KV, et al: Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518–527. Syrjänen K, Syrjänen S: Epidemiology of human papilloma virus infections and genital neoplasia. Scand J Infect Dis Suppl 1990;69:7–17. Zur Hausen H: Disrupted dichotomous intracellular control of human papillomavirus infection in cancer of the cervix. Lancet 1994;343:955–957. Mantovani F, Banks L: The human papillomavirus E6 protein and its contribution to malignant progression. Oncogene 2001;20:7874–7887. Munger K, Basile JR, Duensing S, Eichten A, Gonzalez SL, Grace M, et al: Biological activities and molecular targets of the human papillomavirus E7 oncoprotein. Oncogene 2001;20:7888–7898. Duensing S, Munger K: Mechanisms of genomic instability in human cancer: insights from studies with human papillomavirus oncoproteins. Int J Cancer 2004;109:157–162. Ostor AG: Natural history of cervical intraepithelial neoplasia: a critical review. Int J Gynecol Pathol 1993;12:186–192. Solomon D, Nayar R: The Bethesda System for Reporting Cervical Cytology. Definitions, Criteria, and Explanatory Notes. New York, Springer, 2004. Duensing S, Lee LY, Duensing A, Basile J, Piboonniyom S, Gonzalez S, et al: The human papillomavirus type 16 E6 and E7 oncoproteins cooperate to induce mitotic defects and genomic instability by uncoupling centrosome duplication from the cell division cycle. Proc Natl Acad Sci USA 2000;97:10002–10007. Allen DG, White DJ, Hutchins AM, Scurry JP, Tabrizi SN, Garland SM, et al: Progressive genetic aberrations detected by comparative genomic hybridization in squamous cell cervical cancer. Br J Cancer 2000;83:1659–1663. Heselmeyer K, Macville M, Schrock E, Blegen H, Hellstrom AC, Shah K, et al: Advanced-stage cervical carcinomas are defined by a recurrent pattern of chromosomal aberrations revealing high genetic instability and a consistent gain of chromosome arm 3q. Genes Chromosomes Cancer 1997;19:233–240. Kirchhoff M, Rose H, Petersen BL, Maahr J, Gerdes T, Lundsteen C, et al: Comparative genomic hybridization reveals a recurrent pattern of chromosomal aberrations in severe dysplasia/carcinoma in situ of the cervix and in advanced-stage cervical carcinoma. Genes Chromosomes Cancer 1999;24:144–150. Matthews CP, Shera KA, McDougall JK: Genomic changes and HPV type in cervical carcinoma. Proc Soc Exp Biol Med 2000;223:316–321. Wentzensen N, Vinokurova S, von Knebel DM: Systematic review of genomic integration sites of human papillomavirus genomes in epithelial dysplasia and invasive cancer of the female lower genital tract. Cancer Res 2004;64:3878–3884. Jeon S, Allen-Hoffmann BL, Lambert PF: Integration of human papillomavirus type 16 into the human genome correlates with a selective growth advantage of cells. J Virol 1995;69:2989–2997. Jeon S, Lambert PF: Integration of human papillomavirus type 16 DNA into the human genome leads to increased stability of E6 and E7 mRNAs: implications for cervical carcinogenesis. Proc Natl Acad Sci USA 1995;92:1654–1658. Klaes R, Woerner SM, Ridder R, Wentzensen N, Duerst M, Schneider A, et al: Detection of highrisk cervical intraepithelial neoplasia and cervical cancer by amplification of transcripts derived from integrated papillomavirus oncogenes. Cancer Res 1999;59:6132–6136.
von Knebel Doeberitz
16
22
23
24
25
26
27 28 29 30 31
32 33 34 35
36 37
38
39 40 41
42
43
Hopman AH, Kamps MA, Smedts F, Speel EJ, Herrington CS, Ramaekers FC: HPV in situ hybridization: impact of different protocols on the detection of integrated HPV. Int J Cancer 2005;96:345–353. Luft F, Klaes R, Nees M, Durst M, Heilmann V, Melsheimer P, et al: Detection of integrated papillomavirus sequences by ligation-mediated PCR (DIPS-PCR) and molecular characterization in cervical cancer cells. Int J Cancer 2001;92:9–17. Ziegert C, Wentzensen N, Vinokurova S, Kisseljov F, Einenkel J, Hoeckel M, et al: A comprehensive analysis of HPV integration loci in anogenital lesions combining transcript and genome-based amplification techniques. Oncogene 2003;22:3977–3984. Vinokurova S, Wentzensen N, Einenkel J, Klaes R, Ziegert C, Melsheimer P, et al: Clonal history of papillomavirus-induced dysplasia in the female lower genital tract. J Natl Cancer Inst 2005;97:1816–1821. Melsheimer P, Vinokurova S, Wentzensen N, Bastert G, von Knebel DM: DNA aneuploidy and integration of human papillomavirus type 16 E6/E7 oncogenes in intraepithelial neoplasia and invasive squamous cell carcinoma of the cervix uteri. Clin Cancer Res 2004;10:3059–3063. Baylin SB, Herman JG: DNA hypermethylation in tumorigenesis: epigenetics joins genetics. Trends Genet 2000;16:168–174. Ehrlich M: DNA methylation in cancer: too much, but also too little. Oncogene 2002;21:5400–5413. Dong SM, Kim HS, Rha SH, Sidransky D: Promoter hypermethylation of multiple genes in carcinoma of the uterine cervix. Clin Cancer Res 2001;7:1982–1986. Virmani AK, Muller C, Rathi A, Zoechbauer-Mueller S, Mathis M, Gazdar AF: Aberrant methylation during cervical carcinogenesis. Clin Cancer Res 2001;7:584–589. Steenbergen RDM, Kramer D, Braakhuis BJM, Stern PL, Verheijen RHM, Meijer CJLM, et al: TSLC1 gene silencing in cervical cancer cell lines and cervical neoplasia. J Natl Cancer Inst 2004;96:294–305. Feng Q, Balasubramanian A, Hawes SE, Toure P, Sow PS, Dem A, et al: Detection of hypermethylated genes in women with and without cervical neoplasia. J Natl Cancer Inst 2005;97:273–282. Chen Y, Miller C, Mosher R, Zhao X, Deeds J, Morrissey M, et al: Identification of cervical cancer markers by cDNA and tissue microarrays. Cancer Res 2003;63:1927–1935. Sopov I, Sorensen T, Magbagbeolu M, Jansen L, Beer K, Kuhne-Heid R, et al: Detection of cancer-related gene expression profiles in severe cervical neoplasia. Int J Cancer 2004;112:33–43. Santin AD, Zhan F, Bignotti E, Siegel ER, Cane S, Bellone S, et al: Gene expression profiles of primary HPV16- and HPV18-infected early stage cervical cancers and normal cervical epithelium: identification of novel candidate molecular markers for cervical cancer diagnosis and therapy. Virology 2005;331:269–291. Keating JT, Ince T, Crum CP: Surrogate biomarkers of HPV infection in cervical neoplasia screening and diagnosis. Adv Anat Pathol 2001;8:83–92. Williams GH, Romanowski P, Morris L, Madine M, Mills AD, Stoeber K, et al: Improved cervical smear assessment using antibodies against proteins that regulate DNA replication. Proc Natl Acad Sci USA 1998;95:14932–14937. El Ghobashy AA, Shaaban AM, Herod J, Innes J, Prime W, Herrington CS: Overexpression of cyclins A and B as markers of neoplastic glandular lesions of the cervix. Gynecol Oncol 2004;92: 628–634. Quade BJ, Park JJ, Crum CP, Sun D, Dutta A: In vivo cyclin E expression as a marker for early cervical neoplasia. Mod Pathol 1998;11:1238–1246. Weaver EJ, Kovatich AJ, Bibbo M: Cyclin E expression and early cervical neoplasia in ThinPrep specimens. A feasibility study. Acta Cytol 2000;44:301–304. Liao SY, Brewer C, Zavada J, Pastorek J, Pastorekova S, Manetta A, et al: Identification of the MN antigen as a diagnostic biomarker of cervical intraepithelial squamous and glandular neoplasia and cervical carcinomas. Am J Pathol 1994;145:598–609. Lieskovska J, Opavsky R, Zacikova L, Glasova M, Pastorek J, Pastorekova S: Study of in vitro conditions modulating expression of MN/CA IX protein in human cell lines derived from cervical carcinoma. Neoplasma 1999;46:17–24. Liao SY, Stanbridge EJ: Expression of the MN antigen in cervical Papanicolaou smears is an early diagnostic biomarker of cervical dysplasia. Cancer Epidemiol Biomarkers Prev 1996;5:549–557.
Biomarkers and Cervical Cancer
17
44
45
46
47
48 49
50 51 52 53 54 55
56
57
58
59
60 61
62
63 64
Liao SY, Stanbridge EJ: Expression of MN/CA9 protein in Papanicolaou smears containing atypical glandular cells of undetermined significance is a diagnostic biomarker of cervical dysplasia and neoplasia. Cancer 2000;88:1108–1121. von Knebel Doberitz M, Oltersdorf T, Schwarz E, Gissmann L: Correlation of modified human papilloma virus early gene expression with altered growth properties in C4–1 cervical carcinoma cells. Cancer Res 1988;48:3780–3786. von Knebel Doberitz M, Gissmann L, Zur Hausen H: Growth-regulating functions of human papillomavirus early gene products in cervical cancer cells acting dominant over enhanced epidermal growth factor receptor expression. Cancer Res 1990;50:3730–3736. von Knebel Doberitz M, Rittmuller C, Zur Hausen H, Durst M: Inhibition of tumorigenicity of cervical cancer cells in nude mice by HPV E6-E7 anti-sense RNA. Int J Cancer 1992;51: 831–834. von Knebel Doberitz M: New markers for cervical dysplasia to visualise the genomic chaos created by aberrant oncogenic papillomavirus infections. Eur J Cancer 2002;38:2229–2242. Khleif SN, DeGregori J, Yee CL, Otterson GA, Kaye FJ, Nevins JR, et al: Inhibition of cyclin D-CDK4/CDK6 activity is associated with an E2F-mediated induction of cyclin kinase inhibitor activity. Proc Natl Acad Sci USA 1996;93:4350–4354. Beausejour CM, Krtolica A, Galimi F, Narita M, Lowe SW, Yaswen P, et al: Reversal of human cellular senescence: roles of the p53 and p16 pathways. EMBO J 2003;22:4212–4222. Sherr CJ: Principles of tumor suppression. Cell 2004;116:235–246. Esteller M: Cancer epigenetics: DNA methylation and chromatin alterations in human cancer. Adv Exp Med Biol 2003;532:39–49. Nakashima R, Fujita M, Enomoto T, Haba T, Yoshino K, Wada H, et al: Alteration of p16 and p15 genes in human uterine tumours. Br J Cancer 1999;80:458–467. Yang HJ, Liu VW, Wang Y, Chan KY, Tsang PC, Khoo US, et al: Detection of hypermethylated genes in tumor and plasma of cervical cancer patients. Gynecol Oncol 2004;93:435–440. Kim JW, Namkoong SE, Ryu SW, Kim HS, Shin JW, Lee JM, et al: Absence of p15INK4B and p16INK4A gene alterations in primary cervical carcinoma tissues and cell lines with human papillomavirus infection. Gynecol Oncol 1998;70:75–79. Klaes R, Friedrich T, Spitkovsky D, Ridder R, Rudy W, Petry U, et al: Overexpression of p16(INK4A) as a specific marker for dysplastic and neoplastic epithelial cells of the cervix uteri. Int J Cancer 2001;92:276–284. Sano T, Oyama T, Kashiwabara K, Fukuda T, Nakajima T: Immunohistochemical overexpression of p16 protein associated with intact retinoblastoma protein expression in cervical cancer and cervical intraepithelial neoplasia. Pathol Int 1998;48:580–585. Ansari-Lari MA, Staebler A, Zaino RJ, Shah KV, Ronnett BM: Distinction of endocervical and endometrial adenocarcinomas: immunohistochemical p16 expression correlated with human papillomavirus (HPV) DNA detection. Am J Surg Pathol 2004;28:160–167. Cameron RI, Maxwell P, Jenkins D, McCluggage WG: Immunohistochemical staining with MIB1, bcl2 and p16 assists in the distinction of cervical glandular intraepithelial neoplasia from tuboendometrial metaplasia, endometriosis and microglandular hyperplasia. Histopathology 2002;41: 313–321. McCluggage WG, Jenkins D: p16 immunoreactivity may assist in the distinction between endometrial and endocervical adenocarcinoma. Int J Gynecol Pathol 2003;22:231–235. McCluggage WG: Re: distinction of endocervical and endometrial adenocarcinomas: immunohistochemical p16 expression correlated with human papilloma virus (HPV) DNA detection. Am J Surg Pathol 2004;28:974. Negri G, Egarter-Vigl E, Kasal A, Romano F, Haitel A, Mian C: p16INK4a is a useful marker for the diagnosis of adenocarcinoma of the cervix uteri and its precursors: an immunohistochemical study with immunocytochemical correlations. Am J Surg Pathol 2003;27:187–193. Riethdorf L, Riethdorf S, Lee KR, Cviko A, Loning T, Crum CP: Human papillomaviruses, expression of p16, and early endocervical glandular neoplasia. Hum Pathol 2002;33:899–904. Tringler B, Gup CJ, Singh M, Groshong S, Shroyer AL, Heinz DE, et al: Evaluation of p16INK4a and pRb expression in cervical squamous and glandular neoplasia. Hum Pathol 2004;35: 689–696.
von Knebel Doeberitz
18
65 66
67
68 69
70 71 72 73
74
75 76
77
78 79
80
81
82
Negri G, Vittadello F, Romano F, Kasal A, Rivasi F, Girlando S, et al: p16INK4a expression and progression risk of low-grade intraepithelial neoplasia of the cervix uteri. Virchows Arch 2004;445:616–620. Wang JL, Zheng BY, Li XD, Angstrom T, Lindstrom MS, Wallin KL: Predictive significance of the alterations of p16INK4A, p14ARF, p53, and proliferating cell nuclear antigen expression in the progression of cervical cancer. Clin Cancer Res 2004;10:2407–2414. Wang SS, Trunk M, Schiffman M, Herrero R, Sherman ME, Burk RD, et al: Validation of p16INK4a as a marker of oncogenic human papillomavirus infection in cervical biopsies from a populationbased cohort in Costa Rica. Cancer Epidemiol Biomarkers Prev 2004;13:1355–1360. Trunk MJ, Dallenbach-Hellweg G, Ridder R, Petry KU, Ikenberg H, Schneider V, et al: Morphologic characteristics of p16INK4a-positive cells in cervical cytology samples. Acta Cytol 2004;48:771–782. Klaes R, Benner A, Friedrich T, Ridder R, Herrington S, Jenkins D, et al: p16INK4a immunohistochemistry improves interobserver agreement in the diagnosis of cervical intraepithelial neoplasia. Am J Surg Pathol 2002;26:1389–1399. Akpolat I, Smith DA, Ramzy I, Chirala M, Mody DR: The utility of p16INK4a and Ki-67 staining on cell blocks prepared from residual thin-layer cervicovaginal material. Cancer 2004;102:142–149. Bibbo M, Klump WJ, DeCecco J, Kovatich AJ: Procedure for immunocytochemical detection of p16INK4A antigen in thin-layer, liquid-based specimens. Acta Cytol 2002;46:25–29. Bibbo M, DeCecco J, Kovatich AJ: p16INK4A as an adjunct test in liquid-based cytology. Anal Quant Cytol Histol 2003;25:8–11. Murphy N, Ring M, Killalea AG, Uhlmann V, O’Donovan M, Mulcahy F, et al: p16INK4A as a marker for cervical dyskaryosis: CIN and cGIN in cervical biopsies and ThinPrep smears. J Clin Pathol 2003;56:56–63. Nieh S, Chen SF, Chu TY, Lai HC, Fu E: Expression of p16INK4A in Papanicolaou smears containing atypical squamous cells of undetermined significance from the uterine cervix. Gynecol Oncol 2003;91:201–208. Nieh S, Chen SF, Chu TY, Lai HC, Fu E: Expression of p16INK4A in Pap smears containing atypical glandular cells from the uterine cervix. Acta Cytol 2004;48:173–180. Pientong C, Ekalaksananan T, Swadpanich U, Kongyingyoes B, Kritpetcharat O, Yuenyao P, et al: Immunocytochemical detection of p16INK4a protein in scraped cervical cells. Acta Cytol 2003;47:616–623. Pientong C, Ekalaksananan T, Kongyingyoes B, Kritpetcharat O, Swadpanich U, Pengsa P, et al: Immunocytochemical staining of p16INK4a protein from conventional Pap test and its association with human papillomavirus infection. Diagn Cytopathol 2004;31:235–242. Sahebali S, Depuydt CE, Segers K, Moeneclaey LM, Vereecken AJ, Van Marck E, et al: p16INK4a as an adjunct marker in liquid-based cervical cytology. Int J Cancer 2004;108:871–876. Saqi A, Pasha TL, McGrath CM, Yu GH, Zhang P, Gupta P: Overexpression of p16INK4A in liquidbased specimens (SurePath) as marker of cervical dysplasia and neoplasia. Diagn Cytopathol 2002;27:365–370. Yoshida T, Fukuda T, Sano T, Kanuma T, Owada N, Nakajima T: Usefulness of liquid-based cytology specimens for the immunocytochemical study of p16 expression and human papillomavirus testing: a comparative study using simultaneously sampled histology materials. Cancer 2004;102:100–108. Guo M, Hu L, Baliga M, He Z, Hughson MD: The predictive value of p16(INK4a) and hybrid capture 2 human papillomavirus testing for high-grade cervical intraepithelial neoplasia. Am J Clin Pathol 2004;122:894–901. Wentzensen N, Bergeron C, Cas F, Eschenbach D, Vinokurova S, von Knebel DM: Evaluation of a nuclear score for p16INK4a-stained cervical squamous cells in liquid-based cytology samples. Cancer 2005;105:461–467.
Prof. Magnus von Knebel Doeberitz, MD Department of Applied Tumor Biology, Institute of Pathology University of Heidelberg, Im Neuenheimer Feld 220 DE–69120 Heidelberg (Germany) Tel. ⫹49 6221 56 28 76, Fax ⫹49 6221 56 59 81 E-Mail
[email protected], http://www.klinikum.uni-heidelberg.de/atb
Biomarkers and Cervical Cancer
19
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 20–33
Epidemiology of Oncogenic and Nononcogenic HPV Types, and the Evidence for Differences in Their Sexual Transmissibility Ann N. Burchell, Eduardo L. Franco Departments of Oncology and Epidemiology and Biostatistics, McGill University, Montreal, Canada
Human papillomaviruses (HPV) are among the most common sexuallytransmitted pathogens worldwide [1]. Yet the sexually transmitted disease profile of oncogenic (or high-risk, HR) versus nononcogenic (or low-risk, LR) HPV infections has not been uniformly revealed in studies of different populations. A central risk factor for sexually transmitted infections (STI) is the number of sexual partners [2]. Differences in the strength of association between the number of sexual partners and risk for HPV infection have been observed between oncogenic and nononcogenic types [3–9]. It is conceivable that the variability in results seen in different studies might be caused by differences in the sexual transmissibility of type-specific infections. To date, only one study empirically measured transmissibility between sexual partners. Oriel [10] examined the transmission of genital warts before HPV was identified as the causal agent. Sixty percent of sexual partners of index patients subsequently developed warts, suggesting high transmissibility, at least for those HPV types that commonly cause warts. Whether transmissibility is similar or different for oncogenic types is simply unknown, such that inferences must be made from the epidemiological patterns of these infections. This paper will explore differences in the epidemiology of oncogenic versus nononcogenic types, and evidence for differences in their sexual transmissibility. Some important methodological issues should be noted for comparison studies of HR-HPV versus LR-HPV infections. First, the types classified as oncogenic or nononcogenic may vary across studies, as does the sensitivity of laboratory assays and the number of types that may be identified. The HPV
Table 1. HPV types designated as of high oncogenic risk in representative studies and reviews HPV type
Original taxonomic designation
16 18 26 31 33 35 39 45 51 52 53 55 56 58 59 66 68 73 82
Lorincz Bauer Nindl Walboomers et al. [11] et al. [12] et al. [13] et al. [14] 1992 1993 19981 19992
Bosch et al. [15] 19953
Muñoz et al. [16] 2003
IARC 20054
X X
X X
X X
X X
X X X X X X X
X X X X X X X
X X X X X X X
X X Probable X X X X X X X Probable
X X
X X X
X X X X X X X X X X
X X
X X X
X X X
X X X Probable X X X
X X X X
X
X X X X X
13
14
X 18
15–18
13
X X X
X
83 Number of types
Pap238A, MM9 W13B, MM4, IS39 (subtype) Pap291, MM7 9
11
X X X X
1
X X X X X X X
The HR-HPVs listed have become widely used as probes in diagnostic assays used in epidemiological and clinical studies. For instance, the HR-HPVs under Nindl et al. [13] are part of the probe B set in the commercially available Hybrid Capture 2 assay (Digene Co.). 2 As above, for the GP5/6⫹ general primer polymerase chain reaction (PCR) widely used in many international studies of cervical cancer etiology and screening. 3 As above, for the PGMY line blot PCR protocol produced by Roche Diagnostics that is currently under evaluation as a diagnostic tool. 4 Based on a consensus review panel of all published studies by the International Agency for Research on Cancer as part of its Carcinogenicity Evaluation Monograph, vol 90.
types that have been classified as oncogenic have gradually increased over time, as evidence for their propensity to cause pre-cancerous or cancerous lesions accumulates [11–16]. Table 1 shows the different HPV types that have been classified as oncogenic. The population studied is also important; e.g., it is known that HPV infections vary considerably across age groups [7]. The extent of sexual activity in a given study sample can greatly affect the ability to
Epidemiology of Oncogenic and Nononcogenic HPV Infections
21
observe associations between the number of sex partners and infection risk. In studies where few women report multiple partners, precision of disease association measures will be limited. The duration of follow-up and intervals between study visits may affect incidence and disease duration measures. Finally, definitions used by authors for the number of sex partners are variable. The time frame chosen as well as the grouping into categories will all affect comparability between studies.
Prevalence
In a comparison of HR-HPV and LR-HPV infections, an important issue is differences in disease prevalence. Prevalence will affect the probability of encountering an infected partner, which is an essential component in determining the spread of an STI in a population [17]. A large-scale study of HPV prevalence in 15,613 women in four continents was published recently by the IARC HPV Prevalence Surveys Study Group [1]. Using a standardized protocol, a random sample of women was selected in 13 areas of 11 countries (Nigeria, India, Vietnam, Thailand, Korea, Columbia, Argentina, Chile, the Netherlands, Italy, and Spain) [1]. These surveys identified 15 HR-HPV and 16 LR-HPV types. Overall, 6.1% of women were positive for HR-HPV types. The authors did not specifically report prevalence for any LR types, but instead reported the prevalence of harboring LR types only, which was 2.5%. Nevertheless, it was reported that out of the 2,003 type-specific infections observed in 1,429 women, 62.4% were with HR and 33.7% were with LR types. The most prevalent HR-HPV, HPV-16 (19.7%), was present in over twice as many women as the most prevalent nononcogenic infection, with HPV-42, at 9.4%. These results suggest that infections with HR-HPVs are considerably more prevalent among women than those with LR ones. In North America, findings were similar. In a cross-sectional survey of 489 female university students attending a university health clinic in Montreal, Canada, LR-HPV infections were observed in 6.2% of the women, whereas 11.8% had a HR-HPV infection [5]. A longitudinal study of the same population observed baseline prevalence of 21.8% for HR-HPV types and 14.8% for LR-HPV types [18]. Both of these studies identified 13 HR and 14 LR types. Similarly, in a study of 3,863 women aged 18–40 in Albuquerque, New Mexico, HR infections were more prevalent than LR ones, at 26.7% and 14.7%, respectively [9]; in this study, 18 HR and 9 LR types were identified. A recently published study of 8,514 women in Guanacaste, Costa Rica, observed different results [7]. This study used a PCR assay that identified 17 HRHPV and 33 LR-HPV types and considerably more LR types were individually
Burchell/Franco
22
identified than in the above studies. In this population-based sample, 13.7% (95% CI 13.0–14.5) were positive for any HR type, compared to 17.5% (95% CI 16.7–18.3) for LR types [7]. The difference in prevalence for HR-HPV versus LR-HPV infections varied strongly with age. There was only a slight difference in women aged less than 25 years (24.4% versus 22.5%, respectively). The difference then became greater with age so that among women aged 65 years or older, HR infections (13.6%) were nearly half as prevalent as LR infections (24.8%). The age-specific prevalence curves were U-shaped for both HR-HPV and LRHPV types; however, this curve was more distinct for the LR types [7].
Incidence
Differences in the incidence of HPV infections by oncogenic potential grouping may also suggest differential transmissibility. However, the pattern is not consistent across studies. In the Ludwig-McGill cohort study, which enrolled Brazilian women (mean age: 33 years) attending a maternal and child health program, LR infections were acquired at a higher rate than HR infections [19]. The rate of acquisition per 1,000 woman-months was 9.1 (95% CI 7.5–10.9) for LR types and 6.8 (95% CI 5.4–8.4) for HR types. There was an interaction with age, such that rates were virtually identical for women aged less than 35 (9.0 LR and 8.8 HR), but differed substantially for women aged 35 and older (9.2 LR and 4.0 HR). Conversely, incidences of HR infections were higher than in LR infections among women aged 18–35 attending a planned parenthood clinic in Arizona [6]. The cumulative incidence at 12 months was 0.32 for the HR and 0.18 for the LR types. Similar results were observed in a cohort of Colombian women attending cervical cancer screening centers and family-planning clinics [20]. The incidence of HR infections was 5.0 per 100 woman-years (95% CI 4.4–5.6), significantly higher than the rate of LR infections (2.0 per 100 woman-years, 95% CI 1.7–2.4). Differences in the incidence rate by oncogenicity varied with age. Incidence of LR infections gradually decreased from 4% among 20-year-old women to 1% among women aged 50 and older. Rates of HR infections were highest among young women, at over 10%, declined considerably, peaked slightly at age 50, followed by further decline such that incidence of HR and LR infections were low and approximately equal in women over the age of 60. One possible explanation for the difference between the study by Muñoz et al. [20] and Franco et al. [19] may be the classification schemes used to group types by oncogenicity. The earlier study by Franco’s group [19] considered 12 types as oncogenic (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 68) based on an augmented classification by Bauer et al. [12]. The later study by Muñoz et al. [20]
Epidemiology of Oncogenic and Nononcogenic HPV Infections
23
grouped 19 types as oncogenic (16, 18, 26, 31, 33, 34, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 68, 73, 82) based on the classification by Muñoz et al. [16]. Among young female college students in New Jersey, type-specific incidence rates were reported by Ho et al. [21]. The authors did not report incidence rates by risk groups. Nevertheless, they did report that the mean (⫾ SD) 24-month cumulative incidence was 3 ⫾ 2% for the 16 HR types and 2 ⫾ 2% for the other types (combining LR and uncharacterized types), suggesting somewhat higher incidence rates for HR types. These differences did not reach statistical significance. Little difference was observed in a longitudinal study of young female university students in Montreal, where incidence rates were approximately equal for 13 grouped HR and 14 grouped LR infections. These were 14.0 cases/1,000 woman-months (95% CI 11.4–16.3) and 12.4 cases/1,000 woman-months (95% CI 10.4–14.8), respectively [18].
Duration
It is well known that prevalence is a function of incidence and duration. Therefore, differences in the duration of HR and LR infections may explain differences in their prevalence. Disease duration is also an important component in the spread of an STI in a population, with infections of longer duration having a potentially greater impact [17]. A growing body of evidence suggests that HR infections may persist longer than LR ones [19, 22–24]. In a closely followed cohort study of 60 adolescent women, with weekly HPV testing, KaplanMeier estimates of persistence/runs for HR and LR types were 226 and 170 days, respectively [22]; run length was significantly associated with HPV type (p ⫽ 0.03). Similarly, among women aged 18–35 attending a planned parenthood clinic in Arizona, the median time to clearance of prevalent infections at baseline was 9.8 and 4.3 months for HR and LR infections, respectively [23]. Among Brazilian women in the Ludwig-McGill cohort, the mean duration of infections (measured actuarially) detected at enrolment was 8.9 months (95% CI 7.6–10.2) for HR and 7.0 months (95% CI 6.2–7.8) for LR types [19]. These estimates may not accurately reflect the true duration of infections, since the duration of infections prior to enrolment was unknown. To overcome this limitation, unbiased measures of mean duration were estimated based on the relationship between prevalence and incidence estimates; these were 13.5 and 8.2 months for HR and LR types, respectively [19]. Similarly, the duration of incident infections among Colombian women was longer for HR-HPV types (median 14.8 months, 95% CI 13.8–17.0) than for LR types (median 11.1 months, 95% CI 8.2–16.5) [20].
Burchell/Franco
24
Contrary to other results, among female university students in Montreal with incident HPV infections, there was little difference in the duration of HR and LR infections. The median time to clearance was 13.2 months for HR types compared to 12.3 months for LR types, with overlapping confidence intervals. Mean times to clearance were 16 and 13 months, respectively. Differences in duration by oncogenicity were not observed in the studies by Ho et al. [21] or Woodman et al. [25], based on type-specific rather than grouped median durations. An important methodological issue in the estimation of infection duration is whether or not prevalent or incident infections are studied. Most research has generally evaluated clearance of prevalent or mixed prevalent and incident infections [6, 19, 24, 25], and only a few discriminated among more than a substantial number (⬎10) of LR types [19, 21]. The length of study follow-up, definition of clearance, and time interval between study visits may also affect duration estimates [18]. These different design issues could complicate comparisons of the duration of grouped HR-HPV and LR-HPV infections across studies. Nevertheless, the emerging evidence appears indicative of a longer duration of HR versus LR infections.
Sexual Risk Profiles for HR-HPV versus LR-HPV Types
Since the number of sexual partners is a key determinant of STI acquisition, the magnitude of the association between these two variables, and how this varies by oncogenicity, may provide insights into transmissibility. Studies that have stratified HPV outcome according to oncogenicity suggest that are different sexual risk profiles for LR-HPV and HR-HPV infections [3–9]. A cross-sectional study conducted in north eastern Brazil was the first to suggest that HPV types may differ in their sexual transmissibility [3]. Sexual measures included the life-time number of partners, age at first intercourse, years of sexual activity, and lifetime occasions of sexual intercourse. Infection with LR-HPV types was only weakly associated with life-time measures of sexual behavior among women younger than 40. Conversely, life-time sexual activity variables were strong predictors of HR infections, regardless of age [3]. Two explanations for this difference were proposed. The first is that the variability might be caused by differences in the sexual transmissibility, such that HR types are more transmissible than LR types. Second, the LR types may have a lower likelihood of persistence than the HR types. Kjaer et al. [4] also proposed the latter hypothesis in their Danish study of a population-based random sample of women in their twenties. In that study, prevalent HR-HPV infections were strongly correlated with the life-time number of partners, but no such association was observed for LR types [4].
Epidemiology of Oncogenic and Nononcogenic HPV Infections
25
The same pattern was observed among young university students in Montreal [5]. Only prevalent HPV infections with oncogenic types were associated with lifetime measures of sexual activity. Giuliano et al. [6] observed that the number of sexual partners (life-time and new partners in the past 3 months) was associated with both types of infections in a study of women frequenting family planning clinics along the US-Mexican border. However, the strength of the associations was greater and more evident for HR types. In a similar study of women attending clinics for routine gynecologic care in Albuquerque, New Mexico, the odds ratios for prevalent HPV infection comparing women with two or more partners in the past year to women with no or one partner were nearly identical at 1.8 (95% CI 1.4–2.2) and 1.7 (95% CI 1.3–2.2) for HR and LR infections, respectively [9]. In a large-scale, population-based sample of Costa Rican women, prevalent HR and LR types were both associated with the number of recent and lifetime partners, although the relative strengths of the associations differed [7]. Compared to only one life-time partner, four or more sexual partners in one’s life-time was more strongly associated with prevalent HR infections (OR ⫽ 2.5, 95% CI 2.0–3.0) than with LR infections (OR ⫽ 2.1, 95% CI 1.7–2.5). Conversely, the number of sexual partners in the past year correlated more strongly with LR infections. Having two or more recent partners, compared to no partners, was associated with a 2.6-fold increase in risk of prevalent LR-HPV (95% CI 1.7–3.0) whereas the odds ratio for HR infections was 2.2 (95% CI 1.4–3.4). The confidence intervals overlap considerably for these findings. Nevertheless, the differences in the point estimates suggest that HR infections correlate more strongly with life-time partners and LR infections correlate more strongly with recent partners. A limitation in the interpretation of risk factors for prevalent infection is that they may reflect associations with both incidence and duration. Stronger evidence would be obtained from analyses of incident HPV infection. Rousseau et al. [8] documented differences in risk factor profiles for HR versus LR type infections using a cumulative case-control design among Brazilian women in the Ludwig-McGill cohort. Incident and prevalent infections were combined in their analysis. Age at first intercourse was strongly associated with a decreasing risk of HR-HPV, but was unrelated to LR-HPV infections. Two or more sexual partners in the past year was associated with a greater risk of HPV infection compared to zero or one partner, but the magnitude of the age-adjusted odds ratio was greater for HR (OR ⫽ 5.33, 95% CI 2.14–13.28) compared to LR types (OR ⫽ 3.34, 95% CI 1.30–8.55). In their longitudinal study of Colombian women, Muñoz et al. [20] did not observe differences in the strength of the association between the number of partners and incident infections. Compared to women with only one partner at
Burchell/Franco
26
enrolment and no new partners at follow-up, women with one partner at entry and at least one new partner at follow-up were at a 2.5-fold (95% CI 1.5–4.2) greater risk of a HR infection, and a 2.7-fold (95%CI 1.3–5.5) greater risk of a LR infection. Among young women aged ⬍40, the effect of having two or more partners at entry and one or more new partners at follow-up was stronger for HR (OR 3.5, 95% CI 1.6–7.8) than for LR infections (OR 2.3, 95% CI 0.8–6.9). In summary, differing associations were found between HPV infection and the number of sexual partners after stratification of HPV types by oncogenicity potential. There is an emerging pattern that the number of partners in one’s lifetime tends to correlate strongly with prevalent HR infections [3–7]. Conversely, life-time number of partners correlates more weakly with prevalent LR infections [3, 6, 7, 9], or not at all [4, 5]. In studies of incident HPV infections, the magnitude of the association between the number of recent partners, or sex with new partners, was either equivalent for HR and LR types [20], or somewhat stronger for HR types [8].
Expected Relationship between the Number of Sexual Partners and HPV Infection
In order to draw conclusions regarding differences in the sexual transmissibility of HR versus LR infections, one would ideally study incident infections, thereby eliminating effects of duration for prevalent infections. A simple simulation analysis was carried out to determine the expected magnitude of the association between the number of partners and incident HPV infection, and how it may vary with transmissibility. Six hundred sexually active women were simulated in a hypothetical cohort study of 12 months duration. The number of sex partners was a Poisson mean of 2.0 (ranging from 1 to 6). The frequency of sex per month was randomly assigned according to a gamma distribution (median ⫽ 7, mean ⫽ 9.5, SD ⫽ 10). For women randomly assigned more than one partner, the number of sexual acts was assumed to be equally divided across partners. As described by Allard [26], the probability of infection for an individual woman can be calculated as Prob (infection) ⫽ 1 – {1 – P[1 – (1 – )A]}N, where P is the probability that a partner is infected (i.e. male prevalence), is the probability of transmission in a single sexual contact with a partner, A is the number of sexual contacts per partner, and N is the number of sexual partners. (This equation is of course a simplification that ignores re-emergence of previously latent infections, as well as other factors that may affect transmissibility, such as whether or not condoms are used, viral load in the male, and susceptibility in the female, perhaps due to inflammation.) The output of a simulation is the expected odds ratio for incident HPV infection for women with two partners and
Epidemiology of Oncogenic and Nononcogenic HPV Infections
27
women with three or more partners, with one partner being the referent group. The simulation was repeated 10,000 times, and the results averaged to obtain the best estimate of the expectation under those conditions. Figure 1 shows the relationship between the categorized number of partners and incident HPV infection among women, and how this varies with transmissibility. When male prevalence is held constant, the strength of the association between number of partners and HPV infection will increase with higher transmissibility, as might be expected. The above equation also implies that HPV infection is not just a function of the transmission probability, but of the probability of encountering an infected male partner, that is, male prevalence. As shown in figure 2, the strength of the association between number of partners and HPV infection will increase in magnitude with higher prevalence among men, when transmissibility is held constant. In both figures 1 and 2, most of the variation occurs in the odds ratio for three or more compared to one partner. This suggests that studies of women with few a partners, who must compare one partner to two or more, may be limited in their ability to detect effects, and to observe differences by oncogenicity. If HR and LR types differ in both male prevalence and transmissibility, then these effects should act in concert (table 2). If HR types are more prevalent in men and more transmissible (i.e. if scenario 1 represents LR types and scenario 4 represents HR types), then the relationship between the number of partners and HR-HPV infection in women should be far stronger than that for LR-HPV. Conversely, LR types may be more transmissible but less prevalent (i.e. if scenario 2 represents LR types and scenario 3 represents HR types). In this case, the effects may virtually cancel each other out, such that there appears to be no difference. As reviewed above, differences in the prevalence of HR versus LR infections are common among studies, with HR-HPV infections being more prevalent than LR-HPV ones, likely due to the longer duration of the former. HR infections may also be more prevalent and of longer duration in men. In a longitudinal study in a Mexican military zone including Mexico City at baseline, HR infections (34.8%) were considerably more prevalent among 1,030 soldiers than LR infections (23.9%) [27]. Among prevalent infections at baseline, more HR infections tended to be present at 1 year of follow-up (31%) than LR infections (23%), although actuarial methods were not used to calculate persistence. Two other studies observed an only slightly higher male prevalence of HR infections. More young male university students in South Korea had HR infections (4.2%) than LR (2.6%) [28]. Prevalence of HPV infection was nearly equal by group among Mexican males living in Morelos State, with HR infections being slightly more prevalent (19.8%) than LR (17.7%) [29].
Burchell/Franco
28
4.50 4.00
Odds ratio
3.50 3.00 2.50 2.00 1.50 1.00 0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
Transmissibility (lambda)
Fig. 1. Effect of varying transmissibility on the relationship between the number of sex partners and incident HPV infection among women. Transmissibility (lambda) is the probability of transmission in a single sexual contact with a partner. Male prevalence held constant at 0.15. Dashed line ⫽ Odds ratio of 2 versus 1 partner; solid line ⫽ odds ratio of 3 or more versus 1 partner.
4.00
3.50
Odds ratio
3.00
2.50
2.00
1.50
1.00 0.05
0.10
0.15
0.20
0.25
Male prevalence
Fig. 2. Effect of varying male HPV prevalence on the relationship between the number of sex partners and incident HPV infection among women. Transmissibility held constant at 0.20 (i.e. the probability of transmission in a single sexual contact with a partner). Dashed line ⫽ Odds ratio of 2 versus 1 partner; solid line ⫽ odds ratio of 3 or more versus 1 partner.
Epidemiology of Oncogenic and Nononcogenic HPV Infections
29
Table 2. Expected association between the number of partners in the past year and incident HPV among women, with varying transmissibility per sexual contact with a partner () and male prevalence Scenario
Male prevalence
Transmissibility
Odds ratio 2 versus 1 partner
Odds ratio 3⫹ versus 1 partner
1 2 3 4
Low (0.05) Low (0.05) High (0.25) High (0.25)
Low ( ⫽ 0.05) High ( ⫽ 0.40) Low ( ⫽ 0.05) High ( ⫽ 0.40)
1.84 2.09 2.01 2.47
2.50 2.91 2.84 4.08
However, in a study of men attending an STD clinic in Arizona, slightly fewer men (12.0%) were positive for HR-HPV than for LR-HPV (14.8%) [30]. In the same population, the relative differences in prevalence of HPV by oncogenicity increased with age [31]. Among men aged 18–24, HR-HPVs were slightly more prevalent (18.4%) than LR-HPVs (15.2%). The reverse applied to men aged 40–70, where LR infections (21.7%) were two times more prevalent than HR ones (10.8%). These age effects were similar to those observed in Costa Rican women [7].
Conclusions
Studies of HPV prevalence among women suggest that HR-HPV infections tend to be more prevalent among women than those with LR-HPV infections, although their relative prevalence may vary with age, and with the number of HPV types studied and the classification scheme used to define HRHPV infections (table 1). Differences in incidence are inconsistent in the literature. Far more consistent is the observation that HR infections are of longer duration than LR ones. Prevalent HR infections tend to correlate more strongly with life-time measures of sexual behavior than those with prevalent LR infections, possibly due to the longer duration of the former. There may be a difference in the relationship between the number and nature of recent partners and incident HPV infections by oncogenicity, but more studies in multiple populations, with adequate precision, are needed to confirm this. A finding that the number of sex partners predicts more strongly incident HPV infection of one category over another (i.e. HR versus LR) could be an indication that HR types are more transmissible than LR ones. Nevertheless, the simulation exercise showed that the magnitude of the association between the number of sex partners and incident infections among women relates not just to
Burchell/Franco
30
transmissibility, but also to male prevalence. This review showed that the relative prevalence of HR and LR infections varies across geographic regions and age groups among women and men. Differences in the relationship between the number of sexual partners the women have and HPV infection by oncogenicity may then be due to population differences in male prevalence. The transmissibility of a particular HPV type would likely be less variable across populations if it is predominantly a function of the virus. Yet within each type group, there may still be considerable heterogeneity, such that the transmissibility of HR infections analyzed as a whole may differ across populations. These hypotheses underline the need for more research on HPV infections among men before an in-depth understanding of HPV transmission dynamics can be obtained. Moreover, it is only in studies of male and female couples that questions can be answered regarding the transmissibility of HPV, and how this may vary by HPV type.
Acknowledgments A.N.B. is recipient of a Tomlinson Graduate Fellowship at McGill University. E.L.F. is recipient of a Distinguished Scientist Award from the Canadian Institutes of Health Research.
References 1
2 3 4
5
6
7
Clifford G, Herrero R, Munoz N, Snijders P, Vaccarella S, Anh P, Ferreccio C, Hieu N, Matos E, Molano M, Rajkumar R, Ronco G, de Sanjosé S, Shin H, Sukvirach S, Thomas J, Tunsakul S, Meijer C, Franceschi S, the IARC HPV Prevalence Surveys Study Group: Worldwide distribution of human papillomavirus types in cytologically normal women in the International Agency for Research on Cancer HPV prevalence surveys: a pooled analysis. Lancet 2005;366: 991–998. Anderson RM: Transmission dynamics of sexually transmitted infections; in Holmes KK, Mardh P-A, Sparling PF, et al (eds): Sexually Transmitted Diseases. New York, McGraw-Hill, 1999, pp 25–37. Franco E, Villa L, Ruiz A, Costa M: Transmission of cervical human papillomavirus infection by sexual activity: differences between low and high risk types. J Infect Dis 1995;172:756–763. Kjaer S, Van den Brule A, Bock J, Poll P, Engholm G, Sherman M, Walboomers J, Meijer C: Determinants for genital human papillomavirus (HPV) infection in 1000 randomly chosen young Danish women with normal Pap smear: are there different risk profiles for oncogenic and nononcogenic HPV types? Cancer Epidemiol Biomar Prev 1997;6:799–805. Richardson H, Franco E, Pintos J, Bergeron J, Arella M, Tellier P: Determinants of low-risk and high-risk cervical human papillomavirus infections in Montreal university students. Sex Transm Dis 2000;27:79–86. Giuliano A, Papenfuss M, Abrahamsen M, Inserra P: Differences in factors associated with oncogenic and nononcogenic human papillomavirus infection at the United States-Mexico border. Cancer Epidemiol Biomarker Prev 2002;11:930–934. Herrero R, Castle P, Schiffman M, Concepción B, Hildesheim A, Morales J, Alfaro M, Sherman M, Wacholder S, Chen S, Rodriguez A, Burk R: Epidemiologic profile of type-specific human papillomavirus infection and cervical neoplasia in Guanacaste, Costa Rica. J Infect Dis 2005;191:1796–1807.
Epidemiology of Oncogenic and Nononcogenic HPV Infections
31
8
9
10 11
12
13
14
15
16
17
18
19
20
21 22
23
24
25
Rousseau M, Franco E, Villa L, Sobrinho J, Termini L, Prado J, Rohan T: A cumulative casecontrol study of risk factor profiles for oncogenic and nononcogenic cervical human papillomavirus infections. Cancer Epidemiol Biomarker Prev 2000;9:469–476. Peyton C, Gravitt P, Hunt W, Hundley R, Zhao M, Apple R, Wheeler C. Hunt WC, Hundley RS, Zhao M, Apple RJ, Wheeler CM: Determinants of genital human papillomavirus detection in a US population. J Infect Dis 2001;183:1554–1564. Oriel JD: Natural history of genital warts. Br J Venereol Dis 1971;47:1–13. Lorincz A, Reid R, Jenson A, Greenberg M, Lancaster W, Kurman R: Human papillomavirus infection of the cervix: relative risk associations of 15 common anogenital types. Obstet Gynecol 1992;79:328–337. Bauer H, Hildesheim A, Schiffman M, Glass A, Rush B, Scott D, Cadell D, Kurman R, Manos M: Determinants of genital human papillomavirus infection in low-risk women in Portland, Oregon. Sex Transm Dis 1993;20:274–278. Nindl I, Lorincz A, Mielzynska I, Petry U, Baur S, Kirchmayr R , Michels W, Schneider A: Human papillomavirus detection in cervical intraepithelial neoplasia by the second-generation hybrid capture microplate test, comparing two different cervical specimen collection methods. Clin Diagn Virol 1998;10:49–56. Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, Snijders PJ, Peto J, Meijer CJ, Munoz N: Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189:12–19. Bosch F, Manos M, Munoz N, Sherman M, Jansen A, Peto J, Schiffman M, Moreno V, Kurman R, Shah K, IBSCC Study Group: Prevalence of human papillomavirus in cervical cancer: a worldwide perspective. J Natl Cancer Inst 1995;87:796–802. Muñoz N, Bosch FX, de Sanjosé S, Herrero R, Castellsagué X, Shah KV, Snijders PJ, Meijer CJ, the International Agency for Research on Cancer Multicenter Cervical Cancer Study Group: Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518–527. Aral SO, Holmes KK: Social and behavioral determinants of epidemiology of STDs: industrialized and developing countries; in Holmes KK, Mardh P-A, Sparling PF, et al (eds): Sexually Transmitted Diseases. New York, McGraw-Hill, 1999, pp 39–76. Richardson H, Kelsall G, Tellier P, Voyer H, Abrahamowicz M, Ferenczy A, Coutlée F, Franco EL: The natural history of type-specific human papillomavirus infections in female university students. Cancer Epidemiol Biomarker Prev 2003;12:485–490. Franco EL, Villa LL, Sobrinho JP, Prado JM, Rousseau MC, Desy M, Rohan TE: Epidemiology of acquisition and clearance of cervical human papillomavirus infection in women from a high-risk area for cervical cancer. J Infect Dis 1999;180:1415–1423. Muñoz N, Méndez F, Posso H, Molano M, Van den Brule A, Ronderos M, Meijer C, Munoz A, for the Instituto Nacional de Cancerología HPV Study Group: Incidence, duration, and determinants of cervical human papillomavirus infection in a cohort of Colombian women with normal cytological results. J Infect Dis 2004;190:234–242. Ho GY, Bierman R, Beardsley L, Chang CJ, Burk RD: Natural history of cervicovaginal papillomavirus infection in young women. N Engl J Med 1998;338:423–428. Brown D, Shew M, Qadadri B, Neptune N, Vargas M, Tu W, Juliar B, Breen T, Fortenberry J: A longitudinal study of genital human papillomavirus infection in a cohort of closely followed adolescent women. J Infect Dis 2005;191:182–192. Giuliano A, Harris R, Sedjo R, Baldwin S, Roe D, Papenfuss M, Abrahamsen M, Inserra P, Olvera S, Hatch K: Incidence, prevalence and clearance of type-specific human papillomavirus infections: the young women’s health study. J Infect Dis 2002;186:462–469. Moscicki AB, Shiboski S, Broering J, Powell K, Clayton L, Jay N, Darragh TM, Brescia R, Kanowitz S, Miller SB, Stone J, Hanson E, Palefsky J: The natural history of human papillomavirus infection as measured by repeated DNA testing in adolescent and young women. J Pediatr 1998;132:277–284. Woodman CB, Collins S, Winter H, Bailey A, Ellis J, Prior P, Yates M, Rollason T: Natural history of cervical human papillomavirus infection in young women: a longitudinal cohort study. Lancet 2001;357:1831–1836.
Burchell/Franco
32
26 27
28
29
30
31
Allard R: A family of mathematical models to describe the risk of infection by a sexually transmitted agent. Epidemiology 1990;1:30–33. Lajous M, Mueller N, Cruz-Valdéz A, Aguilar L, Franceschi S, Hernández-Ávila M, LazcanoPonce E: Determinants of prevalence, acquisition, and persistence of human papillomavirus in health Mexican military men. Cancer Epidemiol Biomarker Prev 2005;14:1710–1716. Shin HR, Franceschi S, Vaccarella S, Roh JW, Ju YH, Oh JK, Kong HJ, Rha SH, Jung SI, Kim JI, Jung KY, van Doorn LJ, Quint W: Prevalence and determinants of genital infection with papillomavirus, in female and male university students in Busan, South Korea. J Infect Dis 2004;190: 468–476. Lazcano-Ponce E, Herrero R, Munoz N, Hernandez-Avila M, Salmerón J, Leyva A, Meijer C, Walboomers J: High prevalence of human papillomavirus infection in Mexican males: comparative study of penile-urethral swabs and urine samples. Sex Transm Dis 2001;28:277–280. Baldwin S, Wallace D, Papenfuss M, Abrahamsen M, Vaught L, Kornegay J, Hallum J, Redmond S, Giuliano A: Human papillomavirus infection in men attending a sexually transmitted disease clinic. J Infect Dis 2003;187:1064–1070. Baldwin S, Wallace D, Papenfuss M, Abrahamsen M, Vaught L, Giuliano A: Condom use and other factors affecting penile human papillomavirus detection in men attending a sexually transmitted disease clinic. Sex Transm Dis 2004;31:601–607.
Prof. Eduardo Franco, MPH, DrPH Division of Cancer Epidemiology, McGill University 546 Pine Avenue West Montreal, Quebec H2W 1S6 (Canada) Tel. ⫹1 514 398 6032, Fax ⫹1 514 398 5002, E-Mail
[email protected] Epidemiology of Oncogenic and Nononcogenic HPV Infections
33
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 34–43
Immune Responses to Genital HPV Margaret Stanley Department of Pathology, University of Cambridge, Cambridge, UK
Host Defence to Pathogens
Host immune defences to pathogens are a partnership between innate immunity (phagocytes, soluble proteins, e.g. cytokines, complement and epithelial barriers) and adaptive immunity (antibody, cytotoxic effector cells). Put simply, the innate immune system detects the pathogen and acts as first-line defence with, it is estimated, 90% of microbial assaults cleared by innate responses alone. Innate responses do not have memory, but critically, innate immunity activates the appropriate adaptive immune response that will kill and clear the pathogen and generate specific memory to the insult. Thus, the adaptive, antibody-mediated, humoral immune response clears free virus particles from body fluids and surfaces and can prevent re-infection by virus, whereas cell-mediated immune (CMI) responses are essential for the clearance of virusinfected cells and the generation of immune memory. Innate immunity is alerted and activated by cell injury or cell death and manifested by inflammation (the local vascular response to injury). In the inflammatory process, soluble and cellular innate immune effectors are recruited, and local parenchymal cells and phagocytes (both recruited and local) are activated to secrete inflammatory cytokines and other defence molecules. Crucially, dendritic cells, the only antigen-presenting cells (APC) that can activate naïve T lymphocytes, are activated to kick start the adaptive immune response. The innate immune response generates the signals that identify the nature of the antigen and the type of effector response to be induced via germ line-encoded receptors that recognise conserved molecular targets, usually, but not always, essential products of microbial physiology central to microbial survival. These invariant products known as pathogen-associated molecular patterns are molecular motifs not specific to a particular pathogen but shared by groups of pathogens. Pathogen-associated molecular patterns are ligated by pattern recognition receptors, such as the Toll-like
receptors (TLRs), present in high density on the APC and macrophages of the innate immune system, resulting in the activation of innate immune effectors and induction of an appropriate adaptive immune response [1]. Lymphocytes are the key players in adaptive immunity, and CD4⫹ T lymphocytes have central roles in both arms of the adaptive response. The unique T cell receptor on the naïve CD4⫹ T cell recognises foreign peptide presented as a complex together with major histocompatibility (MHC) class II molecules by a professional APC. CD4⫹ T cell activation results in the secretion of a repertoire of small proteins or cytokines that help and regulate other cells. The pattern of cytokine expression defines two major subsets of CD4⫹ T cells known as Th2 or Th1 cells. Th2 cells secrete interleukin (IL)-4, IL-13 and other cytokines and help antigen-primed B lymphocytes to differentiate into plasma cells, secrete effector antibody molecules of humoral responses or develop as memory B cells. Th1 cells secrete interferon (IFN)-␥ and create a milieu in which key cytotoxic effector macrophages, natural killer cells and cytotoxic CD8⫹ T lymphocytes are activated generating CMI and memory T cells. A third category of T cells, generally described as regulatory T cells, with the phenotype CD4⫹CD25⫹ that express the signature transcription factor Foxp3 and usually secrete IL-10 and transforming growth factor (TGF)- are important in controlling whether there will be a response to antigen. Cells with this phenotype are considered to have their major function in recognising self antigens and preventing auto-immunity. However, they also regulate responses to exogenous antigens and have been implicated in chronic viral infections favouring the persistence of virus and immunopathology [2]. The APC is a critical player in dictating whether the T cell takes the Th2, Th1 or regulatory path, as a consequence of the receptors it expresses and the cytokines secreted in the local milieu. These functions of the APC are in turn activated by signals received from the receptor ligand interactions between the APC and the pathogen, particularly the TLRs, and also by the cytokines released by the APC and other cells in the immediate vicinity. The APC and the cytokines released by it and other cells in the immediate milieu are the bridge between innate and adaptive immunity. In effect, the APC ‘tells’ the T cell if a defence is needed and if so, what sort of defence is required. It is central to the generation and regulation of an effective and appropriate immune response.
Host Defence to Papillomavirus Infections
The Infectious Cycle The papillomaviruses are ubiquitous and successful infectious agents that are characterised by absolute species specificity and tissue tropism. The
Immune Responses to Genital HPV
35
infectious cycle of these viruses is tailored to the differentiation programme of the target cell – the keratinocyte – including different phases of permissive viral growth accompanying the steady maturation of the keratinocyte as it progresses up the epithelium to its destiny as a terminally differentiated squame. Infection and vegetative viral growth are absolutely dependent upon a complete programme of keratinocyte differentiation. The virus infects primitive basal keratinocytes, probably targeting stem cells, but high-level viral expression of viral proteins and viral assembly occur only in the upper layers of the stratum spinosum and granulosum of squamous epithelia [3–5]. Viral gene expression is confined to the keratinocyte, and there is no evidence from clinical infections that viral genes are expressed in any cell other than keratinocytes, or cells with the potential for squamous maturation. This infectious cycle raises several important issues with respect to immune recognition. First, the replication cycle takes a long time. The period between the infection and the appearance of lesions is highly variable, with a minimum period of about 4–6 weeks [6, 7]; however, it can be months or years, indicating that the virus can effectively evade the immune system. The infectious cycle is exclusively intra-epithelial; there is little or no viraemia and viral protein expression in the lower layers of the epithelium that are patrolled by macrophages and immunocytes, is minimal especially for high-risk genital HPVs [3, 8]. Furthermore, there is no cytolysis or cytopathic death accompanying productive viral growth. This occurs in the differentiating keratinocyte, a cell destined for apoptotic death and desquamation far from the sites of immune activity. In consequence, productive HPV infection is not accompanied by inflammation, and there is no obvious ‘danger signal’ to alert the immune system – a viral strategy that results in chronic infections as the host remains ignorant of the pathogen for long periods. Cell-Mediated Immunity in HPV Infections In view of this, one might ask whether there is an immune response to papillomaviruses, but the evidence from natural history studies and HPV infection in immunocompromised individuals is clearly indicating that there is. Large cohort studies have delineated the natural history of genital HPV infections in women. They show that these infections are common in young sexually active women, but that the majority of these resolve [9], indicating that there are effective anti-papillomavirus host defence mechanisms. The increased incidence and progression of HPV infections in immunosuppressed individuals (allograft recipients, inherited immunodeficiencies and HIV-infected patients) illustrates the critical role of CMI responses in the resolution and control of HPV infections [10]. Overall, the evidence from such patients suggests that it is the absolute deficit in CD4⫹ T cells that is important for their florid
Stanley
36
HPV-induced disease and associated neoplastic progression, implying a central role for CD4-mediated mechanisms in the control of HPV infection. Clues to the nature of the cellular immune response to HPV infection have come from immunohistological studies of spontaneously regressing genital warts [11]. Non-regressing genital warts are characterised by a lack of immune cells. The few intra-epithelial lymphocytes are CD8⫹ cells, and mononuclear cells are present mainly in the stroma. Wart regression is characterised by a massive mononuclear cell infiltrate in both stroma and epithelium. The infiltrating lymphocytes are predominantly CD4⫹ cells, but many CD8⫹ T cells are present and are concentrated in the epithelium. The wart keratinocytes have up-regulated MHC class I expression and have induced expression of MHC II and the intracellular adhesion molecule 1, implying that there has been a local release of cytokines such as IFN-␥ [12, 13]. Up-regulation of the adhesion molecules is required for lymphocyte trafficking on the endothelium of the wart capillaries [14], reinforcing the notion of local releases of pro-inflammatory cytokines and chemokines. These appearances are characteristic of a Th1biased lymphocyte response. In these cross-sectional studies, only a snapshot is provided of what is a dynamic process, and detailed longitudinal studies in humans are complicated by ethical and logistic issues, whereas in animal models of mucosal papillomavirus infection such as the canine oral papillomavirus, the immunological events of the entire wart cycle from infection to regression can be followed. In the week just prior to wart regression, there is an intense infiltrate of mononuclear cells into the wart. CD4⫹ T cells arrive first, populating the stroma and, to a lesser extent, the infected epithelium, followed by CD8⫹ T cells that flood the epithelium and stroma. The infected keratinocytes and the endothelial cells of the small vessels of the wart stroma are induced to express MHC class II, and MHC class I expression is massively up-regulated; all features are characteristic of a strong CMI response [15]. The viral protein that is the principal target of this attack appears to be the E2 protein. Systemic T cell responses directed to E2 peptides can be detected at low frequency at distinct time points during the infectious cycle [Jain et al., unpubl. data]. These responses occur in narrow time windows that coincide with periods of viral DNA amplification and are maximal at wart regression, thereafter declining quite rapidly. Dog immunisation with a DNA vaccine encoding canine oral papillomavirus E2, modified to enhance protein expression, results in both prophylactic and therapeutic protection [16, 17], and a recombinant adenovirus E2 vaccine is prophylactically protective [18]. Epidemiological and natural history studies strongly suggest that a successful immune response to HPV infection follows a similar pattern to that seen in animal infections. Genital HPV infection (as determined by detection of HPV DNA in
Immune Responses to Genital HPV
37
cervico-vaginal lavages) is extremely common in young sexually active women with a cumulative prevalence of 60–80% [19]. Most of these HPV infections ‘clear’, i.e. DNA for that specific HPV type can no longer be detected, and most individuals sero-convert [20–22]. The time taken for clearance of the high-risk HPVs (HR-HPVs), particularly HPV16, is on average 8–16 months, considerably longer than the 4–8 months reported for the low-risk HPVs [23, 24]. However, if the immune response fails to clear or control the infection, then a persistent infection (i.e. the production of infectious virus in the face of an immune response) as detected by the continuing presence of HR-HPV DNA, is established. It is this cohort of individuals that have an increased probability of progression to highgrade squamous intra-epithelial lesions (HSIL) and invasive carcinoma. There is increasing evidence that, as in the animal infections, T cell responses to E2 and, in addition, E6 are important at least in HR-HPV infections [25]. A non-intervention follow-up study of women with cytological evidence of low-grade cervical intra-epithelial neoplasia showed that HPV16 E2-specific T cell responses, as measured by specific IL-2 release in vitro, occurred frequently at the time of lesion clearance [26]. Good Th1-type immunity against the E2 and E6 protein has been detected in healthy individuals with no clinical signs of HPV16 infection [27, 28]. Importantly, these Th1-type responses were found only occasionally in high-grade cervical intra-epithelial neoplasia patients and were impaired in cervical cancer patients [29]. These data suggest that a hallmark of effective immune control of HPV16 infection in the cervix is the generation of CD4⫹ cells specific to E2 and E6. Immune Evasion by HPV Why HPV infection remains ignored or undetected by the immune system for so long is a central question. HPV infections are exclusively intra-epithelial and, theoretically, HPV attack should be detected by the professional APC of squamous epithelia, the Langerhans cell (LC). The activated LC should then migrate to the draining lymph node, processing HPV antigens en route, present antigen to naïve T cells in the node that then differentiate into armed effector cells, migrate back to the infected site and destroy the infected keratinocytes. This cycle of events is deflected in a number of ways. The infectious cycle of HPV is in itself an immune evasion mechanism inhibiting the host detection of the virus. HPV replication and release do not cause cell death since the differentiating keratinocyte is already programmed to die, and this ‘death by natural causes’ does not act as a danger signal in the infected site. For most of the duration of the HPV infectious cycle, there is little or no release into the local milieu of the pro-inflammatory cytokines important for dendritic cell activation and migration, and the essential signals to kick start the immune response in squamous epithelia are absent.
Stanley
38
However, even in the absence of viral-induced cytolysis and cell death, HPV-infected keratinocytes should activate type 1 IFN synthesis and secretion. Most DNA viruses have mechanisms for inhibiting IFN synthesis and signalling, and the papillomaviruses are no exception. HR-HPV infection downregulates IFN-␣-inducible gene expression [30, 31], and the HPV16 E6 and E7 oncoproteins directly interact with components of the IFN signalling pathways [32], abrogating these pathways [32, 33]. The interaction of low-risk HPV early proteins and type I IFN responses is not known, but topical therapies such as Aldara™, an agonist of TLR7 that induces strong type I IFN secretion, have been shown to be effective for genital warts [34].
HR-HPV Infection and Immune Failure
One can conclude that HPV efficiently evades the innate immune response and delays the activation of the adaptive immune response, but eventually, the defences are activated, the infection is controlled and immune memory to that specific HPV type is established. There are risks of such a strategy, since the host dendritic cells are exposed to low levels of viral proteins in a non-inflammatory milieu for a protracted time period and local immune non-responsiveness may be established in the infected mucosa. In this milieu which is operationally HPV antigen tolerant, host defences could become irrevocably compromised, and HPV antigen-specific effector cells may either not be recruited to the infected focus or their activity could be down-regulated or both. Thus, if during a persistent HR-HPV infection, there is deregulation of HR-HPV E6 and E7 with increased protein expression, this might not result in an armed effector CMI response, and progression to high-grade squamous intra-epithelial lesions (HSIL) and invasive carcinoma would not be impeded. There is accumulating evidence that this is indeed the scenario in HR-HPV cervical infection. It is important to recognise that there are two biological processes in such an infected epithelium – HR-HPV infection and neoplasia. The local immune environment in low-grade squamous intra-epithelial lesions (LSIL) and HSIL may be influenced by the events of neoplastic progression. In the main, LSIL reflect HPV infection in contrast to HSIL, which are neoplastic, aneuploid, genetically unstable lesions exhibiting heterogeneity in the expression of immunologically relevant molecules. In LSIL, the majority of lesions maintain the virus as an episome, support a complete virus replication cycle, and viral gene expression is tightly regulated. HSIL do not support a complete viral infectious cycle. Late gene expression is either lost or significantly reduced, viral DNA sequences may be integrated into the host genome, and over-expression of the E6 and E7 oncogenes occurs.
Immune Responses to Genital HPV
39
Overall, immunohistochemical studies of cervical LSIL suggest that these lesions are immunologically quiescent with few infiltrating T cells and a decreased number of LC compared with normal ectocervical epithelium [35]. HSIL also show a decreased number of LC, a phenomenon that may be related to neoplasia since there is in vitro evidence that immortalised HPV-infected cervical cells inhibit LC recruitment [36]. In addition, expression of HPV16 E6 in keratinocytes in vitro inhibits E-cadherin expression [37], and since LC/keratinocyte adhesion is mediated via E-cadherin:E-cadherin interactions, this could reduce LC retention in squamous epithelium. Immature CD1a⫹ stromal dendritic cells expressing IL-10 and TGF- have been shown to be resident throughout the normal cervical stroma but increased in HSIL [38]. These dendritic cells have been implicated in inducing immunotolerance [39], and their increased density in HSIL would support the notion that the immune milieu in high-grade lesions is immunosuppressive. Paradoxically, strong pro- and anti-inflammatory responses can both be detected in HSIL [12, 13, 38, 40]. In particular, a robust Th1 response with abundant expression of IFN-␥ by CD4, CD8 and natural killer cells exists, but despite this, HR-HPV-infected HSIL persist. The evidence shows that progression in HR-HPV-induced neoplasia is accompanied by an increasingly immunosuppressive environment with the recruitment of regulatory T cells [41] and a cytokine milieu in which IL-10 [42–44] and TGF- dominate [45, 46]. At the same time, responses to pro-inflammatory anti-viral cytokines such as tumour necrosis factor-␣ [42, 43] and IFN-␣ and IFN- [47, 48] are lost, and the key immune defences inducing apoptosis and death of infected cells are crippled.
References 1 2 3 4 5
6 7 8
Medzhitov R, Janeway CA Jr: Decoding the patterns of self and nonself by the innate immune system. Science 2002;296:298–300. Rouse BT, Suvas S: Regulatory cells and infectious agents: detentes cordiale and contraire. J Immunol 2004;173:2211–2215. Doorbar J: The papillomavirus life cycle. J Clin Virol 2005;32(suppl 1):S7–S15. Stanley MA: Replication of human papillomaviruses in cell culture. Antiviral Res 1994;24:1–15. Middleton K, Peh W, Southern S, Griffin H, Sotlar K, Nakahara T, El Sherif A, Morris L, Seth R, Hibma M, Jenkins D, Lambert P, Coleman N, Doorbar J: Organization of human papillomavirus productive cycle during neoplastic progression provides a basis for selection of diagnostic markers. J Virol 2003;77:10186–10201. Oriel JD: Natural history of genital warts. Brit J Vener Dis 1971;47:1–13. Koutsky L: Role of epidemiology in defining events that influence transmission and natural history of anogenital papillomavirus infections. J Natl Cancer Inst 1991;83:978–979. Iftner T, Böhm S, Oft M, Wilczynski SP, Pfister H: Expression of the E6 and E7 genes of human papillomaviruses in tumors of different dignity; in Stanley M (ed): Immunology of Human Papillomaviruses. New York, Plenum Press, 1994, pp 25–32.
Stanley
40
9
10 11
12 13
14
15
16
17
18
19 20
21
22 23
24
25
26
27
Herrero R, Hildesheim A, Bratti C, Sherman ME, Hutchinson M, Morales J, Balmaceda I, Greenberg MD, Alfaro M, Burk RD, Wacholder S, Plummer M, Schiffman M: Population-based study of human papillomavirus infection and cervical neoplasia in rural Costa Rica. J Natl Cancer Inst 2000;92:464–474. Stanley MA: Imiquimod and the imidazoquinolones: mechanism of action and therapeutic potential. Clin Exp Dermatol 2002;27:571–577. Coleman N, Birley HD, Renton AM, Hanna NF, Ryait BK, Byrne M, Taylor Robinson D, Stanley MA: Immunological events in regressing genital warts. Am J Clin Pathol 1994;102: 768–774. Coleman N, Stanley MA: Analysis of HLA DR expression on keratinocytes in cervical neoplasia. Int J Cancer 1994;56:314–319. Coleman N, Greenfield IM, Hare J, Kruger Gray H, Chain BM, Stanley MA: Characterization and functional analysis of the expression of intercellular adhesion molecule 1 in human papillomavirus related disease of cervical keratinocytes. Am J Pathol 1993;143:355–367. Coleman N, Stanley MA: Characterization and functional analysis of the expression of vascular adhesion molecules in human papillomavirus-related disease of the cervix. Cancer 1994;74: 884–892. Nicholls PK, Moore PF, Anderson DM, Moore RA, Parry NR, Gough GW, Stanley MA: Regression of canine oral papillomas is associated with infiltration of CD4⫹ and CD8⫹ lymphocytes. Virology 2001;283:31–39. Moore RA, Santos EB, Nicholls PK, White KL, Anderson DM, Lloyd A, Topley P, Romanos M, Thomsen L, Parmar V, Walcott S, Gough GW, Stanley MA: Intraepithelial DNA immunisation with a plasmid encoding a codon optimised COPV E1 gene sequence, but not the wild-type gene sequence completely protects against mucosal challenge with infectious COPV in beagles. Virology 2002;304:451–459. Moore RA, Walcott S, White KL, Anderson DM, Jain S, Lloyd A, Topley P, Thomsen L, Gough GW, Stanley MA: Therapeutic immunisation with COPV early genes by epithelial DNA delivery. Virology 2003;314:630–635. Johnston KB, Monteiro JM, Schultz LD, Chen L, Wang F, Ausensi VA, Dell EC, Santos EB, Moore RA, Palker TJ, Stanley MA, Jansen KU: Protection of beagle dogs from mucosal challenge with canine oral papillomavirus by immunization with recombinant adenoviruses expressing codon-optimized early genes. Virology 2005;336:208–218. Schiffman M, Kjaer SK: Chapter 2: natural history of anogenital human papillomavirus infection and neoplasia. J Natl Cancer Inst Monogr 2003;4:14–19. Carter JJ, Wipf GC, Hagensee ME, McKnight B, Habel LA, Lee SK, Kuypers J, Kiviat N, Daling JR, Koutsky LA, et al: Use of human papillomavirus type 6 capsids to detect antibodies in people with genital warts. J Infect Dis 1995;172:11–18. Carter JJ, Koutsky LA, Wipf GC, Christensen ND, Lee SK, Kuypers J, Kiviat N, Galloway DA: The natural history of human papillomavirus type 16 capsid antibodies among a cohort of university women. J Infect Dis 1996;174:927–936. Dillner J: The serological response to papillomaviruses. Semin Cancer Biol 1999;9:423–430. Giuliano AR, Harris R, Sedjo RL, Baldwin S, Roe D, Papenfuss MR, Abrahamsen M, Inserra P, Olvera S, Hatch K: Incidence, prevalence, and clearance of type-specific human papillomavirus infections: The Young Women’s Health Study. J Infect Dis 2002;186:462–469. Brown DR, Shew ML, Qadadri B, Neptune N, Vargas M, Tu W, Juliar BE, Breen TE, Fortenberry JD: A longitudinal study of genital human papillomavirus infection in a cohort of closely followed adolescent women. J Infect Dis 2005;191:182–192. Steele JC, Mann CH, Rookes S, Rollason T, Murphy D, Freeth MG, Gallimore PH, Roberts S: Tcell responses to human papillomavirus type 16 among women with different grades of cervical neoplasia. Br J Cancer 2005;93:248–259. Bontkes HJ, de Gruijl TD, Bijl A, Verheijen RH, Meijer CJ, Scheper RJ, Stern PL, Burns JE, Maitland NJ, Walboomers JM: Human papillomavirus type 16 E2-specific T-helper lymphocyte responses in patients with cervical intraepithelial neoplasia. J Gen Virol 1999;80:2453–2459. de Jong A, van der Burg SH, Kwappenberg KM, van der Hulst JM, Franken KL, Geluk A, van Meijgaarden KE, Drijfhout JW, Kenter G, Vermeij P, Melief CJ, Offringa R: Frequent detection of
Immune Responses to Genital HPV
41
28
29
30 31
32 33 34 35
36
37
38
39
40 41
42
43 44 45 46 47
human papillomavirus 16 E2-specific T-helper immunity in healthy subjects. Cancer Res 2002;62:472–479. Welters MJ, Filippov DV, van den Eeden SJ, Franken KL, Nouta J, Valentijn AR, van der Marel GA, Overkleeft HS, Lipford G, Offringa R, Melief CJ, van Boom JH, van der Burg SH, Drijfhout JW: Chemically synthesized protein as tumour-specific vaccine: immunogenicity and efficacy of synthetic HPV16 E7 in the TC-1 mouse tumour model. Vaccine 2004;23:305–311. de Jong A, van Poelgeest MI, van der Hulst JM, Drijfhout JW, Fleuren GJ, Melief CJ, Kenter G, Offringa R, van der Burg SH: Human papillomavirus type 16-positive cervical cancer is associated with impaired CD4⫹ T-cell immunity against early antigens E2 and E6. Cancer Res 2004; 64:5449–5455. Chang YE, Laimins LA: Microarray analysis identifies interferon-inducible genes and Stat-1 as major transcriptional targets of human papillomavirus type 31. J Virol 2000;74:4174–4182. Nees M, Geoghegan JM, Hyman T, Frank S, Miller L, Woodworth CD: Papillomavirus type 16 oncogenes downregulate expression of interferon-responsive genes and upregulate proliferationassociated and NF-kappaB-responsive genes in cervical keratinocytes. J Virol 2001;75: 4283–4296. Ronco LV, Karpova AY, Vidal M, Howley PM: Human papillomavirus 16 E6 oncoprotein binds to interferon regulatory factor-3 and inhibits its transcriptional activity. Genes Dev 1998;12:2061–2072. Barnard P, McMillan NA: The human papillomavirus E7 oncoprotein abrogates signaling mediated by interferon-alpha. Virology 1999;259:305–313. Beutner KR, Spruance SL, Hougham AJ, Fox TL, Owens ML, Douglas JM Jr: Treatment of genital warts with an immune-response modifier (imiquimod). J Am Acad Dermatol 1998;38: 230–239. McKenzie J, King A, Hare J, Fulford T, Wilson B, Stanley M: Immunocytochemical characterization of large granular lymphocytes in normal cervix and HPV associated disease. J Pathol 1991; 165:75–80. Hubert P, van den Brule F, Giannini SL, Franzen-Detrooz E, Boniver J, Delvenne P: Colonization of in vitro-formed cervical human papillomavirus-associated (pre)neoplastic lesions with dendritic cells: role of granulocyte/macrophage colony-stimulating factor. Am J Pathol 1999;154:775–784. Matthews K, Leong CM, Baxter L, Inglis E, Yun K, Backstrom BT, Doorbar J, Hibma M: Depletion of Langerhans cells in human papillomavirus type 16-infected skin is associated with E6-mediated down regulation of E-cadherin. J Virol 2003;77:8378–8385. Kobayashi A, Greenblatt RM, Anastos K, Minkoff H, Massad LS, Young M, Levine AM, Darragh TM, Weinberg V, Smith-McCune KK: Functional attributes of mucosal immunity in cervical intraepithelial neoplasia and effects of HIV infection. Cancer Res 2004;64:6766–6774. Wakkach A, Fournier N, Brun V, Breittmayer JP, Cottrez F, Groux H: Characterization of dendritic cells that induce tolerance and T regulatory 1 cell differentiation in vivo. Immunity 2003;18: 605–617. Coleman N, Stanley MA: Characterization and functional analysis of the expression of vascular adhesion molecules in human papillomavirus related disease of the cervix. Cancer 1994;74:884–892. Kobayashi A, Greenblatt RM, Anastos K, Minkoff H, Massad LS, Young M, Levine AM, Darragh TM, Weinberg V, Smith McCune KK: Functional attributes of mucosal immunity in cervical intraepithelial neoplasia and effects of HIV infection. Cancer Res 2004;64:6766–6774. Mota F, Rayment N, Chong S, Singer A, Chain B: The antigen-presenting environment in normal and human papillomavirus (HPV)-related premalignant cervical epithelium. Clin Exp Immunol 1999;116:33–40. Stanley MA, Scarpini C, Coleman N: Cell mediated immunity and lower genital tract neoplasia. RCOG Monogr 2003;3:123–134. Offringa R, de Jong A, Toes RE, van der Burg SH, Melief CJ: Interplay between human papillomaviruses and dendritic cells. Curr Top Microbiol Immunol 2003;276:215–240. Sheu BC, Lin RH, Lien HC, Ho HN, Hsu SM, Huang SC: Predominant Th2/Tc2 polarity of tumor infiltrating lymphocytes in human cervical cancer. J Immunol 2001;167:2972–2978. Hazelbag S, Fleuren GJ, Baelde JJ, Schuuring E, Kenter GG, Gorter A: Cytokine profile of cervical cancer cells. Gynecol Oncol 2001;83:235–243. Chang YE, Pena L, Sen GC, Park JK, Laimins LA: Long-term effect of interferon on keratinocytes that maintain human papillomavirus type 31. J Virol 2002;76:8864–8874.
Stanley
42
48
Alazawi W, Pett M, Arch B, Scott L, Freeman T, Stanley MA, Coleman N: Changes in cervical keratinocyte gene expression associated with integration of human papillomavirus 16. Cancer Res 2002;62:6959–6965.
Prof. Margaret Stanley, MD, PhD Department of Pathology, Tennis Court Road Cambridge CB2 1QP (UK) Tel. ⫹44 1223 333736, Fax ⫹44 1223 333346, E-Mail
[email protected] Immune Responses to Genital HPV
43
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 44–53
Bridging the Communication Gap between Practitioners and Their Patients Role of HPV and Advances in Cervical Cancer Prevention
Marie Savard Clinical Associate Professor, University of Pennsylvania, Wynnewood, Pa., USA
Educating women and their practitioners about new advances in the prevention and early detection of cell changes and cervical cancer presents a significant challenge. There are many myths and misconceptions about the purpose and effectiveness of the Pap test, the role of HPV in cervical cancer, and new diagnostic tests, these myths and misconceptions can compromise the possibility of an informed and collaborative practitioner-patient relationship. In this discussion, we will dispel the practitioner myths, such as HPV information causing fear in patients, inadequate reimbursement for HPV testing, insufficient time and resources for counseling patients, and loss of the annual visit; address the patient myths (e.g. the belief that the Pap test is accurate and sufficient, confidence in doctors to provide the most appropriate tests, and trust that ‘no news is good news’ in regard to test results); and, finally, pinpoint the misconceptions surrounding screening guidelines in the new era of HPV testing. These myths and misconceptions will be discussed within the context of a recent survey from the Association of Reproductive Health Professionals, which evaluated patient and provider attitudes and perceptions of HPV and cervical cancer. We recently introduced the ‘take charge’ program of empowering patients to become fully informed about their health, specifically in regard to HPV and cervical cancer, which will be briefly addressed as well. The intention is to offer ways for practitioners and patients to work together towards the goal of providing women with the most up-to-date and best care. A brief address is also given to the role of policymakers and advocacy groups in helping inform and empower women with knowledge about HPV and cervical cancer. It is the conviction of this author that only through a combination of grassroots efforts
of women and advocacy, collaborative practitioner-patient partnerships, and work of policymakers and other experts can the goal of eliminating cervical cancer be realized.
History and Evolution of Patient-Provider Relationships
To date, practitioners have recommended and women have passively and without complete understanding accepted an annual Pap test as an important tool for ‘assuring’ women that they are in good health. Many practitioners have established a routine with patients, offering a standard annual visit, which always includes a gynecological exam and a Pap test. Women have come to expect this test every year, often without knowledge or understanding of the role of the Pap test and what it can and cannot detect. For many years, patients readily accepted diagnostics, treatments and the guidance of their healthcare providers, without taking responsibility for or becoming involved in decisions regarding their own health. With the increase in educational programs and the availability of medical information on the internet, women have become more active participants in their healthcare decisions and more knowledgeable about new technologies available to improve their health. As the environment changes, practitioners must adapt their practices and cooperate with their informed (or misinformed) patients by educating them during their annual visits and involving patients in health decisions. In addition, clinical guidelines, practice and perspectives are beginning to change regarding the role of HPV testing and its ability to improve screening sensitivity. Clinicians are also becoming more aware of the significant failure rate of the Pap smear. The nearly 100% negative predictive value (NPV) of combined Pap and HPV testing, in combination with the near perfect cure rates of cervical disease if caught early, makes every cervical cancer death a failure of the health system.
Educating Practitioners and Dispelling Myths
Practitioners first need to understand HPV and its role in cervical cancer. They must be aware of the effectiveness of new screening tests and the role of vaccine-based prevention options before educating patients and involving them in shared decision-making. Many doctors are not yet aware of the role of HPV in cervical cancer, the natural history of HPV infections, the limitations of the Pap test, and the role of HPV testing. Once they are comfortable with the information, they can begin sharing knowledge with their patients. A survey by
Communication Gap between Doctors and Patients
45
the Henry J. Kaiser Family Foundation found that only half of doctors discuss sexually transmitted diseases (STDs) with all or most of their patients and among those, only 10% specifically mentioned HPV. With greater awareness about the link of HPV to cervical cancer, this omission is no longer an option. There are a number of reasons healthcare providers resist changing their practices to incorporate HPV information and testing. Many doctors are not yet aware that the HPV test has been approved by the FDA for routine cervical cancer screening in women 30 and older. And even those who are aware may not offer the test out of fear that women will be needlessly alarmed if the test shows they carry the virus. Women are hearing about HPV, HPV testing, and more recently the HPV vaccines through the media and through their friends. In fact, women expect to learn about HPV. They resent paternalistic and patronizing providers and will not trust their providers if they learn about HPV from another trusted source. A recent study published in the British Journal of Cancer, ‘Making sense of information about cervical cancer screening: a qualitative study’, found that the way in which information is presented to women is crucial in minimizing the negative psychological impact of testing positive for HPV and ensuring that participation in screening remains high. Women were actually reassured when their doctors discussed HPV openly with them. They were particularly reassured to know that HPV is common, has no symptoms, can lay dormant for many years, can clear up on its own and need not raise concerns about transmission to sexual partners. Women also expect their providers to offer new technologies and discuss important advances with them. Healthcare providers are often concerned about the amount of time it will take them to incorporate new information and messaging into their already tight office visit, not giving them adequate time to truly counsel their patients about the intricacies of HPV and its role in cervical cancer. In fact, experience has proven that patients can be counseled about HPV testing and results effectively in a short amount of time, if the appropriate information is presented. Only a few basic facts are essential for dispelling a woman’s misconceptions and providing her with enough material to be an informed patient. To truly provide an educational office environment, nurses and office staff should be prepared to answer questions and offer educational resources, including the wealth of information available on the Web. Some practitioners may fear that the increased screening interval gained with the use of the HPV test will result in the loss of the annual visit, putting women at risk for other health issues. Women may still see their providers on an annual basis if their cervical screening interval is extended. In reality, interval extension affords time for increased emphasis on chronic disease screening, health risks and health behaviors. In addition, there are great benefits of having a periodic chance for women to talk with their doctors about any concerns or
Savard
46
questions. Screening for cervical cancer, like other cancer screening such as for colon cancer, should now be conducted at intervals appropriate to a woman’s risk for the disease. Practitioners also are unaware of the widespread reimbursement in place for HPV testing. They express concern about problems with reimbursement as a reason for not adopting co-testing. In fact, the majority of all large regional and national payers in the United States, as well as Medicaid programs in 46 states plus the District of Columbia, cover HPV testing for both reflex testing and primary screening.
Educating Patients and Dispelling Myths
Women’s beliefs are widely varied about the etiology of cervical cancer and its relationship with sexual activity. According to a recent survey by the Association of Reproductive Health Professionals (ARHP), just 49% of women say they have heard of HPV and only 23% correctly identified HPV as the primary cause of cervical cancer1. Essential facts regarding HPV and cervical cancer include: (1) cervical cancer is almost always caused by the HPV; (2) HPV is very common; (3) having HPV does not mean a woman will have cervical cancer; (4) cervical cancer is very rare, and (5) HPV status is not a marker of behavior or fidelity. These essential facts are often enough to provide women with the background necessary to ask further questions and open communications, paving the way for a collaborative patient-provider relationship. Most women have been dutifully receiving their annual Pap test, often unaware of the purpose of and the effectiveness of the test. While the Pap test identifies cervical cell abnormalities, many women believe that a normal Pap test means they are free of any gynecological health problems, including sexually transmitted diseases and other gynecological cancers. Women trust their healthcare providers and expect not only the most up-to-date information, but also the most recent technologies available for improved health. Physicians have a responsibility to their patients to fulfill this sense of trust and offer this to women. The ARHP survey found that 88% of women rely on their provider for important information about reproductive/gynecological issues; yet, 81% say that their healthcare provider has never talked to them about the connection between HPV and cervical cancer.
1
HPV Attitudes and Awareness Survey, June 2005, Association of Reproductive Health Professionals.
Communication Gap between Doctors and Patients
47
Another issue with which both patients and practitioners contend is followup. Women often leave clinician offices and assume that unless they hear from their doctor, no problems exist; however, this is not always the case. Women need to understand the importance of keeping track of their records and asking for test results. As practitioners, we need to dispel the myth that ‘no news is good news’, encouraging our patients to follow-up with us for their results and talk to us about their meaning. Women are eager to be given as much information as possible and to take a much more active role in managing their own health decisions. They want to know all there is to know about a specific condition and want to know how best to collaborate with their practitioners to get the best possible care.
Misconceptions Surrounding Screening and the Upcoming Vaccine
News of the development of an HPV vaccine has brought the link between the most common sexually transmitted infection, HPV, and cervical cancer to the forefront. While the HPV vaccine offers women the hope of eliminating cervical cancer, it is still not yet approved. In the meantime, the importance of screening needs to remain highlighted and practitioners must learn how to incorporate advanced screening technologies into their current practice. Even when the vaccine is approved, it will not substitute the need for regular screening. Much like the Pap, if used alone, the vaccine would still leave thousands of women at risk for cervical cancer, either because they were infected with HPV prior to the introduction of the vaccine or because they have a high-risk HPV type that the vaccine cannot prevent. The HPV vaccine will not replace, but rather enhance, the advanced technologies already available today to ensure the best defense against cervical cancer.
Counseling Principles for Practitioners
Physicians often face a number of challenges when educating women about HPV and new technologies, and implementing new screening options within an existing practice. A few basic principles in counseling will help practitioners overcome these issues. • Use a matter-of-fact, nonjudgmental tone when speaking with a patient about HPV and her relationship to cervical cancer. If a doctor is stressed about communicating a message, the patient will be stressed about the content of that message.
Savard
48
•
Emphasis on how common the virus is and how rare cervical cancer is reassures women. Let them know that almost all women will get the virus, but few will develop cervical cancer. • Don’t assume that women understand the Pap test just because they have been getting screened for a number of years. Many women have misunderstandings about what the Pap is and what it is not. For women to understand how they can prevent cervical cancer, they will need to understand HPV, its role in cervical cancer, and the difference between Pap testing and HPV testing. Some basic steps practitioners can follow when educating patients include2: (1) Determine knowledge of HPV and address concerns Like most women, your patient may not know what HPV is, how it is contracted or that there are different types of the virus. It is important to explain that there are over a hundred strains of HPV, and that the virus usually disappears on its own. It also may be necessary to define the difference between high- and low-risk strains and what they can cause. Stress that although HPV is common, cervical cancer is not, and remind patients that it’s important not to be complacent – there are steps you can take to watch for the development of cell changes until the virus clears. (2) Clear up misconceptions Many women who first hear that they have a sexually transmitted infection – especially those in a monogamous relationship – can easily become upset, because they may think their partners have been unfaithful. As a result, testing positive for HPV often comes with a stigma. Patients must be informed that HPV can remain dormant in the body for months or even years before it becomes detectable, so it is important for women to be monitored for HPV even when they and their partners have maintained a long-term, monogamous relationship. Open a dialogue with your patient about her knowledge on the topic and aim to dispel any inaccurate information she has. We need to stop referring to HPV as a sexually transmitted disease – after all – we don’t tell women they are getting a Pap test to diagnose a sexually transmitted disease. If we did, the Pap test would never have gained such widespread acceptance and use – nor would we have saved so many lives. (3) Discuss next steps and screening options Explain to your patient who has a positive HPV test that she will have to be watched more closely to ensure that the HPV infection does not become a problem. Educate her on the use of the Pap and the HPV test to screen for HPV 2
Adapted from Marie Savard’s Website, www.drsavard.com.
Communication Gap between Doctors and Patients
49
and signs of cervical cancer so she is aware of what is being done to her and how it will help ensure her good health for future visits. Women will often ask what steps they can take to boost their immune systems and ‘fight off’ a persistent infection. Some measures such as discontinuing smoking are very important for overall health status, whereas others such as taking a multivitamin or B-complex vitamin or discontinuing oral contraceptives may be reasonable but are not tested. (4) Encourage a collaborative partnership between practitioner and patient Give your patients an original copy of their Pap test and HPV test results, encouraging them to maintain their own medical files. Consider giving your patients a preprinted chart or checklist of important screening and diagnostic tests that they may need so that they too can keep track of results and assume more responsibility for their own health care. Many resources are available to the clinician for self-education, education of their patients, and implementation of new screening and preventive technologies in their offices. Doing the research may take time, but in the end, will help save lives.
Opportunities to Empower Women
There are many things we can teach our patients to do to take a more active role in their health care. (1) Keep track of your own medical records We no longer live in an era where we have just one primary care provider for our entire lives. The more doctors we see, the more information is lost in the shuffle. Come to your doctor’s visit prepared with copies of medical records, recent test results and family history information. Your doctor will have a better picture of your previous medical care and can keep it in mind for future treatment. Also, consider carrying an emergency medication information card with you at all times, which lists up-to-date medical conditions, medications, family history, emergency contacts, allergies, immunization status and information on advance directives. (2) Trust Your Instincts and Ask Questions Medicine is a fast-paced field, with advancements and discoveries happening by the minute. It is impossible for doctors to be aware of everything new. If you see something in the news that might apply to your care, write it down
Savard
50
and ask about it on your next visit. If you experience any strange symptoms throughout the year, make sure you ask about those as well. Don’t be afraid to pull out a list or diary of symptoms, as these will likely affect your treatment and even which tests a doctor will administer. (3) Talk with your doctor about which tests are right for you Along with the basics, don’t be afraid to ask for the most up-to-date tests available. Although the most common screening method for cervical cancer is the Pap test, the liquid-based Pap is more accurate than the traditional Pap ‘smear’. Women 30 and older, those most at risk for cervical cancer, should ask for the HPV test along with their Pap to find out if they have a high-risk form of the virus that could lead to cervical cancer. In fact, the HPV test, when used in combination with the Pap – has been proven more effective than the Pap alone in identifying women at risk for cervical cancer. Unfortunately, not all doctors are aware of this technology, and some don’t believe in its necessity, thus making it essential for women to get educated about the most up-to-date medical information, screening technologies and treatments available. (4) Monitor and manage your healthcare For the 364 days of the year between visits, a woman must be responsible for her own health. So women should use the limited time they have with their doctor to review and set target goals for the coming year. This could include getting more exercise, tracking their menstrual cycle or doing a monthly selfexamination of the breast. (5) Follow-up on test results Test results in even the most organized medical office can be lost, misplaced or even misfiled. The fact is this: no news is not necessarily good news. It’s a good idea to give your doctor’s office a self-addressed stamped envelope so all results can be mailed directly to you. If you don’t receive results within 3 weeks, give your doctor’s office a call to follow-up.
Ways to Work Together
The explosion of national dialogue about HPV is likely to drive women to ask more questions regarding HPV, how it is transmitted and its link to cervical cancer. This forecast stresses the importance of being prepared to discuss the virus with patients in a way that will lessen the stigma, clear up misconceptions and define the differences between high- and low-risk strains. For those who do not ask questions, it is up to us – the practitioners – to engage women in conversations to
Communication Gap between Doctors and Patients
51
educate them. While no doctor intentionally avoids having life-saving conversations with their patients, too often these discussions fall through the cracks because of the concern that the patient might not fully understand or will needlessly worry, or the physician is already working on an overbooked schedule.
Roles of Advocacy Groups and Policymakers
Advocacy groups and policymakers can be of tremendous help to practitioners and patients. The efforts of Women in Government and The Balm in Gilead are two such examples. They have become actively engaged in the process of educating women about HPV and cervical cancer in far-reaching campaigns to eliminate cervical cancer. • Women In Government is a bi-partisan organization of women state legislators in the United States providing leadership opportunities, networking, expert forums and educational resources to address and resolve complex public policy issues. Women In Government launched their Challenge to Eliminate Cervical Cancer campaign in January 2004, encouraging states to become actively involved in taking a community approach to tackling cervical cancer in their states. Since the launch of the campaign, nearly all states throughout the US have introduced and enacted legislation calling for the study of cervical cancer in an individual state, mandated insurance coverage of screening technologies for all women, and increased public awareness about HPV and cervical cancer. Most recently, Women In Government hosted an HPV and Cervical Cancer Summit, the largest cervical cancer gathering to date of US state legislators, advocates, medical experts, and public policy officials. • The Balm In Gilead is a not-for-profit, nongovernmental organization whose mission is to improve the health status of people of the global African community by building the capacity of faith communities to address lifethreatening diseases, initially focusing solely on HIV/AIDS and in 2005, expanding to include cervical cancer. The Balm In Gilead has worked to enable thousands of churches around the world to provide comprehensive educational programs and offer compassionate support to encourage education, screening, and treatment for diseases affecting the African community, thereby reducing health disparities around the world. The goal of the ISIS Project, the Balm’s HPV and cervical cancer initiative, is to educate and empower Black women about HPV and cervical cancer. The ISIS Project tours the United States through the infrastructure of 3 historical Black church denominations, spreading the word about healthy sexuality and prevention and treatment of cervical cancer through advanced technologies.
Savard
52
Conclusion
Doctors need to educate themselves and their patients on the life-saving tests available to women now that will help them prevent cervical cancer in the future. Educating patients is a large part of ensuring they receive the best quality of care. Patients should be encouraged to become actively involved in their health care by empowering themselves with education about cervical cancer and the ways to prevent it. Through improved communication and shared decision-making, patients and their practitioners can assume a more collaborative and effective relationship. Utilizing the advances in technology and treatments along with improved communication and collaborative partnerships with patients, advocacy groups and policymakers, cervical cancer can be eliminated in the foreseeable future. Dr. Marie Savard, MD 421 Owen Road Wynnewood, PA 19096 (USA) Tel. ⫹1 610 580 7240, Fax ⫹1 610 649 0431, E-Mail
[email protected] Communication Gap between Doctors and Patients
53
Methods for HPV Detection Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 54–62
HPV Testing by Hybrid Capture Attila T. Lörincz Digene Corporation, Gaithersburg, Md., USA
HPV Is an Important Pathogen
Cervical cancer in the absence of high-risk HPV (HR-HPV) is rare [1]. The clinical negative predictive value (NPV) of a highly accurate and well-validated HPV DNA test can reach 99.95% or better in a screening population of women over age 30 years [2]. HPV propagates predominantly as a sexually transmitted infection, most commonly in young adults, and regardless of genotype is of minor clinical consequence in the early years. More than 90% of infections are quickly cleared or suppressed below a sub-pathogenic threshold. However, a HR-HPV infection that detectably persists over a span of 5–20 years predisposes the host to a dramatically increased risk of carcinoma of 100-fold or greater [1, 3]. The global incidence of cervical cancer is more than 493,000 cases per year with approximately 274,000 deaths in 2002 [4]. The Papanicolaou (Pap) test has been the basis of cervical cancer prevention for 50 years and where applied in well-designed programs produced an impressive reduction in mortality. Unfortunately, the success of Pap cytology is tarnished by a general lack of quality screening services in resource-poor regions, where the incidence of cervical cancer continues unabated or is increasing. Even in economically advanced nations no cytology tests come close to the 95–100% clinical sensitivity needed to allow a truly effective program [5, 6].
Rationale for HPV Diagnostics
HR-HPV DNA is a highly sensitive marker for women at risk of prevalent and long-term incident high-grade cervical intraepithelial neoplasia (CIN) and cancer [2, 3, 5, 6]; however, the test has a somewhat lower specificity than Pap
cytology [3]. At the moment, HPV DNA testing is widely accepted in the USA as an adjunct to cervical cytology in cancer prevention programs for triage of atypical squamous cells of undetermined significance (ASCUS), and increasingly in primary screening to augment cytology. HR-HPV DNA testing both as a reflex to ASCUS and as a primary screening test has been shown to be costeffective [7]. HPV types are broadly categorized into three groups called the low-risk (LR), intermediate-risk (IR) and HR types [8]. In contrast to LR-HPVs, IRHPVs and HR-HPVs establish a dramatically elevated risk of cervical cancer and high-grade CIN (CIN2/3), in the order of 50- to 300-fold higher than the risk observed for HR-HPV-negative control women. HPV16 and HPV18 (associated with ⬃70% of invasive cervical cancer) cause particularly risky infections strongly predisposing women to cervical squamous carcinomas and adenocarcinomas, respectively; consequently, there is a growing need for HPV genotyping for these and other risk types [9].
HPV Detection Methods
Polymerase chain reaction (PCR) and the Hybrid Capture® 2 test (HC2, Digene Corporation, Gaithersburg, Md., USA) are the two relevant test methods for HPV. Each has its particular strengths and weaknesses. PCR has very high analytical sensitivity, with some expert PCR labs able to detect fewer than 10 copies of HPV genomic DNA, typically in a few microliters of input specimen [10]. Certain PCR procedures also appear to provide accurate HPV genotyping. The majority of available validation data for PCR is analytical and based primarily on plasmid reconstructions in artificial matrix, sometimes in combination with simple studies that compare two or more PCR tests. Few papers show an ability of the PCR test to detect greater than about 95% of high-grade CIN in large, well-designed studies, more typically, the clinical sensitivity is reported in the range of 75 to 95% [6, 11–15] with a median in 16 recent papers of 82%. Of interest is the PCR versus HC2 data from the large ALTS study on 278 cases of CIN3/cancer [11], where a prototype PCR test employing the PGMY09/11 primers attained a clinical sensitivity and specificity of 87.4 and 55.6%, while the corresponding values for HC2 were 92.5 and 51.1%, respectively. HC2 is a relatively simple, high-throughput, semi-automated test [16] approved by the US Food and Drug Administration (FDA) for detecting 13 HRHPVs (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68) at a clinically relevant cutoff of 5,000 HPV genomes per reaction. There are two preferred methods for collecting specimens for an HC2 test. One approach employs co-collection of a standard cytology smear and a small conical shaped nylon bristle brush that is
Papillomavirus Testing
55
rotated three times in the cervical os and then placed into a Specimen Transport Medium (STM) tube (Digene Corporation, Gaithersburg, Md., USA) for HPV DNA detection. The other type of specimen is obtained as a liquid cytology specimen collected into a medium such as PreservCyt® fluid (PC, Cytyc Corporation, Marlborough, Mass., USA). HC2 technology operates on the principle of signal amplification. Up to 176 tests can be completed in approximately 5–6 h under routine manual operation. Alternatively, up to 354 specimens can be run in a similar time on the automated Rapid Capture System® instrument. HC2 utilizes long (⬎1 kb) single-stranded RNA probes to facilitate both capture and detection of target molecules within the specimens. Following denaturation of HPV DNA targets and hybridization to the probes, RNA-DNA hybrids are captured onto microplates containing immobilized antibody specific for RNA-DNA hybrids on a shaking apparatus at room temperature. Then an alkaline phosphatase-labeled anti-RNA-DNA monoclonal antibody is reacted to the captured hybrids. Nonhybridized nucleic acids (including non-target DNA and probes) and conjugates are removed by washing. Incubation of retained alkaline phosphatase with the chemiluminescent substrate CDP Star® (Tropix PE, Bedford, Mass., USA) produces light which is measured by a luminometer. The readings are transferred into a software program where results are analyzed and the number of hybrids immobilized can be expressed as a semi-quantitative value using a relative light unit (RLU) ratio to the 5,000 HPV16 copy positive control. Currently, HC2 seems the most sensitive method to detect high-grade CIN, exhibiting a median sensitivity of 94% [2, 3, 5, 6, 17, 18]. The test exhibits robust performance [19, 20] with repeat test kappa values in the 0.85–0.95 range. Reproducibility is typically better from STM than from PC specimens, and for the former specimen type no retesting is required. For PreservCyt specimens, the FDA has approved a retest zone for initial RLU/cutoff values of 1.0–2.5. In this case, one or two additional tests need to be performed to confirm or contradict the original result. A drawback of HC2 is that the test exhibits some cross-reactivity to other HPV types including both LR- and nontargeted IR- or HR-HPVs not present in the probe cocktail [21]. This cross-reactivity is of concern, yet it has a relatively minor affect on the clinical specificity of the HC2 test, as only about 10–15% of test positives are attributable to nontargeted HPV types and many of these test positives occur in women with CIN [21]. Test Validation Issues The term sensitivity often lacks clarity and, used without further qualifiers, may be misleading. An important error arises when analytical performance is used as a direct indicator of medical utility. For example, analytical sensitivity merely refers to a lower limit of detection of the target. In contrast, clinical
Lörincz
56
sensitivity is the measure related to medical utility because it expresses the detectability of relevant specific disease such as high-grade disease (CIN 2/3 or cancer). A test with high analytical sensitivity may not be sufficiently clinically sensitive if it is compromised by inhibitions and target genome deletions, etc. As another caution, a test that has too much analytical sensitivity may produce poor clinical specificity if it detects low levels of viral DNA not related to clinical disease or risk of cancer [22]. Suppositions about diagnostic test performance predicted from theoretical considerations have little bearing on actual real-world performance. Furthermore, analytical validation studies alone or in combination with small or otherwise inadequate clinical validation studies do not improve this situation in a meaningful way. The complexity of carcinogenesis and HPV natural history make it imperative that tests for HPV intended for routine clinical use undergo extensive clinical validation studies in realistic use settings.
Advances in Hybrid Capture HPV DNA Testing
There is a growing desire for ultrasensitive clinical tests that have multiplex capability to detect many possible pathogens in one sample. The next-generation Hybrid Capture system (fig. 1) is designed to meet these evolving needs. The test includes three main steps: (1) Target purification from crude specimens on magnetic beads. (2) Uniform isothermal amplification of large regions of multiple targets. (3) Identification of individual amplicons by hybridization to targetspecific oligonucleotides in multiplex mode. For the most sensitive and specific detection of amplicons, we combined the Luminex xMAP® bead-microarray technology with the existing HC2 signal amplification method. The prototype test takes approximately 4 h and has demonstrated the ability to detect and distinguish among 17 carcinogenic HPV types simultaneously in a single reaction with an analytical sensitivity of 100 copies and no cross-reactivity to other HPV such as low-risk types. The test is capable of detecting many more than 20 individual targets in a multiplexed mode. In the case of HPV detection, the new test detects both the L and E regions for most HPV types and mitigates problems experienced by L1 primer-based PCR systems, where up to 5% of cancers and CIN 3 may not be detectable because of deletions of the HPV L region. For routine use, the new test will be fully automated and include the capability to detect a broad panel of pathogens of interest to clinicians. Obviously, the ultrasensitive detection of HPV is undesirable for clinical use because it produces unacceptable clinical false-positives. Thus, the cutoff sensitivity for detection of HPV and other targets have been adjusted to optimal clinical thresholds. Initial research validation of the new test for 17 HR-HPV types was conducted and compared to a gold standard of more than 1,000 STM cervical specimens tested extensively
Papillomavirus Testing
57
Sample preparation
Amplification
Detection
Fig. 1. The small spheres in the left and center panel indicate magnetic beads, and the Y-shaped structures on their surfaces represent capture antibodies. The green straight line paired to the red straight line indicates a target DNA hybridized to an RNA capture probe, while the wavy red lines in the left and middle panel represent single-stranded capture RNA probes. The amplified target DNA is shown in the middle panel as several sinuous green lines. The white dotted lines in the right panel represent RNA detection probes bound to the amplified target DNA, the large sphere represents a Luminex bead, the short straight lines represent capture oligonucleotides, and the Y-shaped structures are antibodies conjugated to phycoerythrin (represented by the small yellow spheres).
by both HC2 and GP5⫹/6⫹ PCR. Essentially no cross-reactivity to nontarget HPVs was seen. Positive and negative agreement to an overall HC2 virologic endpoint were both better than 90%. Similarly, HPV genotyping results compared to the GP5⫹/6⫹ PCR test revealed a greater than 90% agreement. Cervical cancer is among the most common causes of cancer mortality among women in resource-constrained regions. Given the pressing need for humanitarian interventions, we developed an HPV test designed to be regionally affordable and robust enough for routine use in developing countries (fig. 2). The new test is based on a magnetic particle technology partly evolved from HC2. It has been designed for long-term robustness without refrigeration and needs to retain stability at up to 40⬚C for several months. Analytical and clinical performance data on hundreds of clinical specimens indicate that we met our design goals and the new test may be suitable for screening use. The test appears accurate, reproducible, simple, rapid, portable, and of moderate throughput in initial evaluations using reference criteria analogous to those employed in HC2 validations. The new assay can be performed in a small batch mode but can also produce over 100 results in 2 h, facilitating same-day follow-up examination and treatment, if
Lörincz
58
Brush or swab in empty tube 45min, 65˚C Add RNA and magnetic beads via dropper bottles; hyb and capture 30min at 65˚C Incubate on heater shaker
Start denaturation
45˚C DR-1, 15min Wash, 10min
Luminometer readout of plate
or
DR-2 and expose to instant film 10min
Magnetic capture
Fig. 2. HPV DNA test for use in resource-constrained regions.
necessary. The format is relatively uncomplicated, allowing minimally trained personnel to perform the assay with simplified liquid transfers. Reagent usage per test has been reduced to lower costs and to minimize waste. Results are delivered in a qualitative format typically employing a photographic film readout although a luminometer may be preferred for larger-scale screening operations. Challenges remaining include validation of test robustness with provider-collected cervical specimens and self-collected vaginal specimens under field conditions, especially as regards suboptimal environmental conditions and sample preparation in the absence of mechanical mixers. Coupled with appropriate screening coverage, dedicated follow-up examination, and appropriate treatment when necessary, this innovative HPV DNA test may contribute to improved outcomes for women at risk of high-grade cervical disease in underserved regions.
Triage Tests for HPV DNA-Positive Women
There has been criticism of the use of HPV testing for routine screening because most infections do not progress to cancer and thus there is the danger of excessive and costly interventions and negative psychological consequence for patients. These criticisms tend to assume that HPV tests will be conducted in an
Papillomavirus Testing
59
indiscriminate way by uninformed clinicians. Not only is this unlikely, but it can also be easily minimized by proper education on appropriate test usage. The best cutoff age for HPV testing continues to be debated but is felt to be in the range of 25–35 years. However, even with this age stratification there is a need for additional improvement in the specificity of HPV tests. Many reflex tests to follow an initial HPV DNA positive result are under investigation including: (1) cytology; (2) HPV genotype; (3) HPV RNA expression; (4) p16 and other marker panels; (5) HPV viral load, and (6) DNA methylation. HPV genotyping is clearly a frontrunner among these potential triage tests. HPV16 and HPV18 in particular appear to warrant extremely careful clinical attention and follow-up regardless of concurrent cytology results. It may be possible to stratify on the basis of E6 and/or E7 expression levels although direct clinical data in support of this strategy are presently lacking. Some have suggested that novel protein markers such as p16, MCM5, EGFR, various cyclins, etc., may replace the need for HPV testing [23]. However, these arguments overlook the inability of such tests to identify the full extent of the at-risk pool, namely those women who are persistently HPV infected without concurrent cytological abnormalities but who are at risk for high-grade disease in the coming years. Preliminary data reveal that none of the markers are as sensitive for detection of high-grade squamous intraepithelial lesions (HSIL) as HPV DNA testing and must be used as panels to reach adequate sensitivity. However, use of panels as stand-alone tests suffers from poor specificity. HPV viral load may be a useful triage test but the data so far are inconsistent and no definitive understanding has emerged. It is clear that HPV viral load is higher in women with prevalent CIN than in HPV-positive women who do not have detectable prevalent CIN; however, it does not appear to be possible to clearly distinguish women with high-grade CIN versus low-grade CIN on the basis of viral load because there is a great overlap between these categories. It may be possible to distinguish a subset of HPV-positive women at lower risk for prevalent high-grade CIN but more data are needed to substantiate this idea. It is also unclear if cytologically normal women with higher viral loads have a greater risk for incident high-grade CIN than women who have lower viral loads and if HPV genotype modulates the relationship [24]. Until these complex questions are satisfactorily answered, it is best to continue use of HPV DNA and Pap cytology tests with follow-up by colposcopy or repeat screening tests.
Acknowledgements The many contributions of Drs Nazarenko, Eder, and our research team members in the creation of new HPV genotyping tests and the HPV test for resource-poor regions are greatly appreciated.
Lörincz
60
References 1 2
3 4 5
6
7 8
9
10
11
12
13 14 15
16
17
18
19
Bosch FX, Lorincz A, Munoz N, Meijer CJLM, Shah KV: The causal relation between human papillomavirus and cervical cancer. J Clin Pathol 2002;55:244–265. Petry K-U, Menton S, Menton M, van Loenen-Frosch F, de Carvalho Gomes H, Holz B, et al: Inclusion of HPV testing in routine cervical cancer screening for women above 29 years in Germany: results for 8468 patients. Br J Cancer 2003;88:1570–1577. Lörincz AT, Richart RM: Human papillomavirus DNA testing as an adjunct to cytology in cervical screening programs. Arch Pathol Lab Med 2003;127:959–968. Parkin DM, Bray F, Ferlay J, Pisani P: Global cancer statistics, 2002. CA Cancer J Clin 2005;55: 74–108. Bigras G, de Marval F: The probability for a Pap test to be abnormal is directly proportional to HPV viral load: results from a Swiss study comparing HPV testing and liquid-based cytology to detect cervical cancer precursors in 12 842 women. Br J Cancer 2005;93:575–581. Kulasingam SL, Hughes JP, Kiviat NB, Mao C, Weiss NS, Kuypers JM, et al: Evaluation of human papillomavirus testing in primary screening for cervical abnormalities: comparison of sensitivity, specificity, and frequency of referral. JAMA 2002;288:1749–1757. Goldie SJ, Kim JJ, Wright TC: Cost-effectiveness of human papillomavirus DNA testing for cervical cancer screening in women aged 30 years or more. Obstet Gynecol 2004;103:619–631. Lörincz AT, Reid R, Jenson AB, Greenberg MD, Lancaster W, Kurman RJ: Human papillomavirus infection of the cervix: relative risk associations of 15 common anogenital types. Obstet Gynecol 1992;79:328–337. Khan MJ, Castle PE, Lorincz AT, Wacholder S, Sherman M, Scott DR, et al: The elevated 10-year risk of cervical precancer and cancer in women with human papillomavirus (HPV) type 16 or 18 and the possible utility of type-specific HPV testing in clinical practice. J Natl Cancer Inst 2005;97:1072–1079. Coutlee F, Gravitt P, Kornegay J, Hankins C, Richardson H, Lapointe N, et al: Use of PGMY primers in L1 consensus PCR improves detection of human papillomavirus DNA in genital samples. J Clin Microbiol 2002;40:902–907. Schiffman M, Wheeler C, Dasgupta A, Solomon D, Castle PE, for the ALTS Group: A comparison of a prototype PCR assay and Hybrid Capture 2 for detection of carcinogenic human papillomavirus DNA in women with equivocal or mildly abnormal Papanicolaou smears. Am J Clin Pathol 2005;124:1–11. Kulmala S-M, Syrjänen S, Shabalova I, Petrovichev N, Kozachenko V, Podistov J, et al: Human papillomavirus testing with the Hybrid Capture 2 assay and PCR as screening tools. J Clin Microbiol 2004;42:2470–2475. Sotlar K, Diemer D, Dethleffs A, Hack Y, Stubner A, Vollmer N, et al: Detection and typing of human papillomavirus by E6 nested multiplex PCR. J Clin Microbiol 2004;42:3176–3184. Cuzick J, Beverley E, Ho L, Terry G, Sapper H, Mielzynska I, et al: HPV testing in primary screening of older women. Br J Cancer 1999;81:554–558. Lörincz AT, Smith JS: Sexually transmissible viral pathogens: the human papillomaviruses and herpes simplex viruses; in Lörincz AT (ed): Nucleic Acid Testing for Human Disease. Boca Raton, Taylor & Francis, 2006. Lörincz AT: Hybrid Capture™ method for detection of human papillomavirus DNA in clinical specimens: a tool for clinical management of equivocal Pap smears and for population screening. J Obstet Gynaecol Res 1996;22:629–636. Belinson JL, Qiao YL, Pretorius RG, Zhang WH, Rong SD, Huang MN, et al: Shanxi Province cervical cancer screening study. II. Self-sampling for high-risk human papillomavirus compared to direct sampling for human papillomavirus and liquid based cervical cytology. Int J Gynecol Cancer 2003;13:819–826. Schiffman M, Herrero R, Hildesheim A, Sherman ME, Bratti M, Wacholder S, et al: HPV DNA testing in cervical cancer screening: results from women in a high-risk province of Costa Rica. JAMA 2000;283:87–93. Carozzi FM, Del Mistro A, Confortini M, Sani C, Puliti D, Trevisan R, et al: Reproducibility of HPV DNA testing by Hybrid Capture 2 in a screening setting: intralaboratory and interlaboratory
Papillomavirus Testing
61
20
21
22 23 24
quality control in seven laboratories participating in the same clinical trial. Am J Clin Pathol 2005;124:1–6. Cubie HA, Moore C, Waller M, Moss S, on behalf of the National Cervical Screening Committee LBC/HPV Pilot Steering Group: The development of a quality assurance programme for HPV testing with the UK NHS cervical screening LBC/HPV studies. J Clin Virol 2005;33:287–292. Castle PE, Schiffman M, Burk RD, Wacholder S, Hildesheim A, Herrero R, et al: Restricted cross-reactivity of Hybrid Capture 2 with non-oncogenic human papillomavirus types. Cancer Epidemiol Biomarkers Prev 2002;11:1394–1399. Snijders PJF, van den Brule AJC, Meijer CJLM: The clinical relevance of human papillomavirus testing: relationship between analytical and clinical sensitivity. J Pathol 2003;201:1–6. von Knebel Doeberitz M: New markers for cervical dysplasia to visualise the genomic chaos created by aberrant oncogenic papillomavirus infections. Eur J Cancer 2002;38:2229–2242. Lörincz AT, Castle PE, Sherman ME, Scott DR, Glass AG, Wacholder S, et al: Viral load of human papillomavirus and risk of CIN3 or cervical cancer. Lancet 2002;360:228–229.
Dr. Attila T. Lörincz, PhD Digene Corporation 1201 Clopper Road Gaithersburg, MD 20878 (USA) Tel. ⫹1 301 944 7350, Fax ⫹1 301 944 7301, E-Mail
[email protected] Lörincz
62
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 63–72
Methods for HPV Detection: Polymerase Chain Reaction Assays Suzanne M. Garland a,b, Sepehr Tabrizia,b a
Department of Microbiology and Infectious Diseases, The Royal Women’s Hospital, and bDepartment of Obstetrics and Gynaecology, University of Melbourne, Melbourne, Australia
HPV and Cervical Cancer
Worldwide, carcinoma of the uterine cervix is a common cancer of women, being second only to breast cancer. Since the initial reports by Harald Zur Hausen in the 1970s, suggesting a role for HPV in the development of cervical cancer, there have been a number of molecular, epidemiological and clinical observational studies clearly implicating HPV as an etiological agent in various anogenital cancers, including the cervix [1]. The lack of ability to utilize conventional viral culture methods initially made detection and diagnosis for HPV difficult until the advent of molecular methods, particularly amplification technology (such as polymerase chain reaction, PCR), which has allowed detection of low-level virus copy numbers in clinical samples. Utilizing these more sophisticated and sensitive diagnostic assays for HPV DNA detection, the consensus conference of the International Agency of Research on Cancer (IARC) in Lyon, France, concluded that certain high-risk (HR)-HPV genotypes (i.e. HPV16 and HPV18), which collectively contribute to around 70% of cervical cancers worldwide, be formally named as human carcinogens [2]. Moreover, these same IARC scientists in a multinational case-control study showed that the strength of the association with less-prevalent HR-HPV genotypes, including 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68, 73, and 82, and cervical cancer was also very strong with odds ratios of 72–347 being the strongest ever observed for human cancer [3]. Analyses of cervical cancers from different regions of the world using PCR methods have shown oncogenic HPV DNA in 95–99.7% and with consistent findings in a large number of investigations in different countries and populations [4].
Despite this role as a necessary cause for cervical cancer, HPV is a common viral infection of squamous epithelial tissues, with genital HPV infections being the commonest viral sexually transmitted infection (STI). This apparent enigma that HPV infection alone is not sufficient as a cause for cancer reflects the fact that most genital HPV infections are transient and asymptomatic (i.e. natural immune surveillance usually clears infectious HPV). It is only in a small number of women, with chronic carriage of oncogenic or HR genotypes that severe dysplasia (CIN2/3) eventuates, over several decades, to cancer. Carcinogenesis requires additional genetic changes such as HPV integration and possibly other cofactors in complex pathways not totally understood [5]. In contradistinction, CIN1/HPV is a transient manifestation of productive cervical HPV infection (median duration 6 months), with natural immune surveillance usually clearing the virus, leading to resolution of the lesion. Genotyping, through comparison of viral sequences and comparison of genetic homology of viral genomes, has shown HPV to be very heterogeneous, with the presence of over 100 HPV genotypes fully sequenced and identified to date, with many more likely in the near future [6]. Biologically, HPVs are divided into two groups: cutaneous and mucosal. A subset of about 30 appears to regularly infect the genital/mucosal epithelium, with the remaining infecting cutaneous areas (skin warts and other lesions). Genital mucosal HPV types such as 6, 11, 42, 43, 44, 53, 54, 55, 62 and 66 are mainly found in low-grade cervical lesions with type 6 and 11 being the primary cause of genital warts. They almost never occur in cervical cancer (apart from the rare malignant change in giant condylomata acuminata of Buschke-Lowenstein tumor) and are designated low-risk (LR) HPV types. HR genotypes are found regularly in high-grade dysplasias or high-grade squamous intraepithelial lesions (HSIL), as well as in cervical cancers. Whilst there is consistency in findings for types 16 and 18 in cervical cancers worldwide, there is some geographical variation in prevalence of types following these two HPV types [7]. Not only do epidemiological studies require reliable and reproducible identification of genital HPV genotypes, but with the imminent licensure of prophylactic HPV vaccines, there will be a need to evaluate the efficacy of the vaccine and potential changes in prevalence of HPV genotypes in the years to come.
HPV Detection
Initial methods of HPV detection used were direct probe hybridization such as dot blot and Southern blot. Besides being labor-intensive and timeconsuming they had low sensitivity, required large amounts of DNA in clinical
Garland/Tabrizi
64
samples and have largely been superseded by amplification technology, which has allowed detection of low-level virus copy numbers in clinical samples. Two such methods, currently used diagnostically include PCR and Hybrid Capture 2 (HC2) (Digene Corporation, Gaithersburg, Md., USA). In their infancy are newer assays detecting HPV RNA, which await large-scale clinical trials to assess their clinical diagnostic value. This review specifically describes PCR assays in detecting HPV DNA.
Polymerase Chain Reaction
PCR is a selective target amplification assay capable of exponential and reproducible increase in the HPV sequences present in biological specimens. The amplification process can theoretically produce one billion copies from a single double stranded DNA molecule after 30 cycles of amplification. When performing PCR, care must be taken to avoid false-positive results, which may be derived from cross-contaminating specimens or reagents with PCR products of previous rounds. Although this was a serious problem in laboratories when PCR was first utilized, most laboratories now implement procedures to overcome this. The sensitivity and specificity of PCR-based methods can vary, depending on the DNA extraction procedures, site and type of clinical sample, sample transport and storage, primer sets, the size of the PCR product, reaction conditions and performance of the DNA polymerase used in the reaction, the spectrum of HPV DNA amplified and ability to detect multiple types. Generally a sensitivity of 1–10 copies per PCR reaction is achieved by most methods utilized. Most laboratories use PCR assays, which utilize consensus primers, directed to a conserved L1 gene, and hence able to detect all mucosal HPV types. Consensus primers described include GP5/6 and modified GP5⫹/6⫹, MY09/11 and modified PGMY09/11, and SPF primer set. Amplification with each of these primers will result in different size amplification products (fig. 1) and this can result in varying sensitivity for detection of certain HPV genotypes [8]. This is particularly an issue when samples contain multiple HPV types. In a recent study, 120 of 11 (9.2%) specimens had multiple infections. The PGMY09/11 method detected most of them (9/11, 81.8%), MY09/11 detected 2/11 (18.2%), whereas the GP5⫹/6⫹ method detected none. The inability of the GP5⫹/6⫹ method to detect multiple infections was at least partly due to subsequent typing by sequencing that had difficulties in revealing multiple types [9]. Subsequent to PCR, analysis of the amplified products and distinction of HPV types can be achieved by sequencing [10] or hybridization with type-specific oligonucleotide probes, using various methods. The latter can be achieved using
Methods for HPV Detection
65
HPV L1 gene
5⬘ SPF1
3⬘
SPF2 65 bp
GP 5⫹
GP 6⫹ 150 bp
PG (MY11)
PG (MY09) 450 bp
Fig. 1. L1 consensus PCR assays.
various hybridization formats, dot blot (fig. 2a), Southern blot (fig. 2b), microtiter ELISA plate (fig. 2c), reverse line blot strip assays (fig. 2d), and microchip format assays (fig. 2e). The sensitivity and reproducibility of results are different depending on the methods used by the laboratory. PCR assays can also be greatly affected by various extraneous substances in a clinical sample and which can inhibit the amplification reaction. Most laboratories have incorporated the amplification of an internal control, such as beta-globin gene (present at one copy per human cell) in each PCR reaction as a measure to detect potential inhibition and/or sample integrity. This is especially important in paraffin-embedded archival tissue where there may be degradation of DNA; then it is important to assess sample integrity. Comparing results of PCR assays from various studies is difficult as there have not been standardized methodologies described and utilized. However, a number of commercial HPV detection and typing assays have recently been released in order to address this need. The commercial assay by Roche Diagnostics (Amplicor, Indianapolis, Ind., USA) has recently been released, although not yet FDA approved although this is anticipated in 2007. This assay, similar to HC2 is able to detect 13 HR-HPV types and will allow diagnostic laboratories to detect HPV-DNA by PCR, although not to discriminate genotypes specifically. Linear array assay and microarray systems are, however, two methods offering rapid detection and simultaneous typing of multiple HPV types. HPV oligonucleotide microarray has been developed by Biomedlab Company (Seoul, Korea), which allows for detection of 22 HPV types using an aldehydederivatized slide glass. PCR products are generated in the presence of fluorescein-tagged nucleotides and hybridized onto the chip and scanned by laser fluorescence, thus able to detect multiple infections with one hybridization step. Ideally, a larger number of HPV type-specific oligonucleotides could be spotted on the chip, although this method requires expensive equipment and may not be suitable for many (fig. 2e).
Garland/Tabrizi
66
A 450bp
B C D E F G
a
450bp H
b
c
Reference line HPV18 HPV31 HPV35 HPV45 HPV52 HPV56 HPV59 HPVMM4 HPVMM9 HPV6 HPV40 HPV53 HPV57 HPVMM8
HPV16 HPV26 HPV33 HPV39 HPV51 HPV55 HPV58 HPV68 HPVMM7 High globin Low globin HPV11 HPV42 HPV54 HPV66
d
e
Fig. 2. Various hybridization assays for detection of PCR products. a Dot blot showing hybridization to HPV16 probe. b Southern blot showing hybridization to HPV sequences at the 450-bp band corresponding to HPV-specific amplification products. c Microtiter ELISA plate format showing hybridization of PCR products with the HPV16 probe. d Reverse line blot strip assays showing 9 LR- HPV and 18 HR-HPV types on each strip. e Microchip format assays showing hybridization to multiple probe types on surface of a slide.
The Linear array® assay also developed by Roche allows detection of 37 different HPV genotypes. It detects amplified type specific HPV DNA with oligonucleotides immobilized onto a nylon membrane, which is an easy to perform detection method (fig. 3). The advantages of this method are speed, with
Methods for HPV Detection
67
7 ME 6 ME 5 ME 4 ME 3 ME 2 ME Fig. 3. Linear array assay capable of detecting 37 types simultaneously.
a turnaround time of 1 day, plus ease in interpreting individual and/or multiple genotypes, yet without requiring expensive, specialized instrumentation. Utility of real-time and quantitative PCR in detection of HPV has also been investigated in a number of studies and can provide tools for quantitation of various HPV genotypes in specimens [11]. Standardization of Assays Laboratories using molecular assays for detection of infectious organisms should use standardized tools when performing such assays. Although such standards are not available for HPV DNA assays yet, the World Health Organization (WHO) has initiated an International Collaborative Study enrolling several laboratories worldwide [12]. The aim of developing HPV International standard reagents as well as proficiency panels, and which should become available in the very near future for HPV detection and typing [Sonia Pagliusi, Geneva, WHO, pers. commun., 2005], means that clinical diagnostic laboratories will be able to validate their own assays and determine their analytical sensitivity. Moreover, for epidemiological prevalence studies and surveillance studies, this will allow comparisons of HPV DNA detection and typing results over time
Garland/Tabrizi
68
between different geographic locations, populations and anatomical sites. This is, in particular, important as the licensure of prophylactic HPV vaccine studies is imminent, and these standards will allow for accurate documentation and comparison of various methods in determination of the prevalence of HPV in trial pre- and postvaccine populations responses across various studies and geographic areas.
Clinical Indications
There are three areas in clinical care where HPV DNA testing has been endorsed or considered, and these will be briefly discussed here. Primary Screening in Conjunction with the Papanicolaou (Pap) Test or as a Stand-Alone Test for Women over 30 Years of Age Well-organized, high-quality Pap cytology screening programmes, which adequately reach a high proportion of those at high risk have markedly reduced cervical cancer incidence and mortality rates. However, there are limitations to cytology: in studies with the least biased estimates of sensitivity, it ranged from 30 to 87% with a mean of 51% [13]. As persistent infection with oncogenic HPVs precedes virtually all HSIL or neoplasias, HPV DNA detection can be used as a marker for current or subsequent development of precursor lesions. Longitudinal studies show HPV DNA testing has a higher sensitivity for predicting prevalent high-grade dysplasias than cytology. Further, the negative predictive value approaches 100%; superior to Pap smears alone. Ultimately, combined HPV DNA and Pap cytology could result in increasing the screening interval for those with a normal Pap and negative HPV DNA, making the combination cost-effective. Most of the published studies to date have been conducted utilizing HC2, with studies based on PCR detection of HPV DNA only just coming to light. In the recently published ALTS study where HPV genotyping was evaluated, authors concluded that ASCUS and LSIL patients who have a positive HPV16 diagnosis by genotyping are at significantly greater risk for detection of high-grade abnormalities during a 2-year follow-up period when compared to those who test positive for another oncogenic type of HPV or who are HPV-negative [14]. Using a mathematical model to evaluate clinical and economic outcomes, Goldie et al. [15] concluded from evaluation of several different screening strategies that for those ⱖ30 years, using HPV DNA plus cytology were more effective in reducing cancer incidence, as well as being more cost effective than conventional cytology. Thus, less-frequent screening with more sensitive tests is likely to provide a reasonable balance between benefits and costs.
Methods for HPV Detection
69
Furthermore, HPV DNA testing, in particular PCR, can be performed on self-collected samples; these have similar sensitivity to clinician-collected samples. Therefore this approach could provide a simple, less expensive and highly feasible tool for primary screening in low-resource settings many of whom have no Pap programs [16]. Triage of Women with Minimally Abnormal (Borderline, ASCUS) Pap Smears In the USA, HPV DNA testing has an accepted role in the management of women with minor cytological abnormalities [17], the rationale being that a high proportion of the ASCUS group, on consensus cytological review are normal when HPV DNA is negative (around 50%), and hence at extremely low risk for HSIL. The role of HPV DNA testing is to focus on those women diagnosed with ASCUS that are HPV-positive where colposcopy assessment is justified. These recommendations largely came from the large ALTS (ASCUS /LSIL triage study), whereby various options for management of ASCUS and LSIL were evaluated. For ASCUS smears, HPV DNA testing predicted abnormalities sooner, with greater sensitivity and resulted in around 50% requiring referral to colposcopy, as compared to two thirds by repeat Pap [18]. In the recent followup of the ALTS study, in evaluating various test sensitivities for detection of cumulative CIN3⫹ (n ⫽ 306) over 24 months, a single HPV DNA identified all those found subsequently to have HSIL and more effectively than a single colposcopy or two Paps [19]. To Predict Cure and Residual Disease after Ablation for Cervical Dysplasia Following treatment with cryosurgery, laser ablation or LEEP (loop electrosurgical excision procedure) for HSIL up to 10% may have residual disease. Standard of care has been close cytological and colposcopic follow-up at 6, 12 and 24 months post-procedure. However, as follow-up Pap has a low specificity in detecting residual HSIL, and HPV DNA is cleared from the cervix following adequate treatment, DNA testing has been evaluated to predict the presence of residual dysplasia [20]. In a recent meta-analysis of 11 studies evaluating HR-HPV DNA testing and monitoring of women after treatment of CIN 3, the NPV for residual disease was 98%, resection margins 91%, cytology 93% with combined HPV and Pap 99%. Therefore, in combining HR-HPV DNA testing with cytology, this allowed women double-negative for HR-HPV and Pap (70% of this population) less intensive follow-up. Adopting such a clinical algorithm would mean resources could be focused on those most at risk (i.e. women positive for Pap
Garland/Tabrizi
70
and HPV DNA for colposcopy) [20, 21]. We await larger studies with costbenefit analyses.
Conclusions
With the development of highly sensitive molecular assays such as PCR for HPV DNA and high-quality epidemiology, the natural history of HPV and various grades of dysplasia are being unraveled. Furthermore, evidence from clinical trials evaluating the use of assays of HPV in various clinical scenarios, is forming a basis for their use in the management of patient care. In the recent IARC consensus statements, the use of high-risk HPV types in screening and patient management is endorsed as a justified practice [22]. Concurrent with this, is a necessity to appropriately educate the health profession and the general public, particularly women and their partners, about HPV infection. Many women would not be aware that Pap cytological abnormalities relate to a viral infection, nor for that matter that the virus is transmitted sexually. There is a need to convey clear and consistent information about HPV to the general community; and to destigmatize and demystify the whole area of HPV. This will be particularly relevant to ensure appropriate use of HPV DNA assays in clinical care, as well as in the introduction of the HPV prophylactic vaccine in the very near future.
References 1
2 3
4
5 6 7
8
zur Hausen H: Molecular pathogenesis of cancer of the cervix and its causation by specific human papillomavirus types; in zur Hausen H (ed): Human Pathogenic Papillomaviruses. Heidelberg, Springer, 1994, pp 133–516. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans: Human Papillomaviruses. Lyon, International Agency for Research on Cancer, 1995, vol 64. Muñoz N, Bosch FX, de Sanjosé S, Herrero R, Castellsague X, Shah KV, Snijders PJ, Meijer CJ: Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518–527. Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, Snijders PJ, Peto J, Meijer CJ, Muñoz N: Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189:12–19. Stoler MH: The pathology of cervical neoplasia; in Rohan TE, Shah KV (eds): Cervical Cancer: From Etiology to Prevention. Norwell, Kluwer Academic Publishers, 2004, pp 3–59. de Villiers M, Fauquet C, Broker T, Bernard H, zur Hausen H: Classification of papillomaviruses: a minireview. Virology 2004;324:17–27. Muñoz N, Bosch FX, Castellsague X, Diaz M, de Sanjose S, Hammouda D, Shah KV, Meijer CJ: Against which human papillomavirus types shall we vaccinate and screen? The international perspective. Int J Cancer 2004;111:278–285. Kornegay JR, Shepard AP, Hankins C, Franco E, Lapointe N, Richardson H, Coutlee F, Canadian Women’s HIV Study Group: Nonisotopic detection of human papillomavirus DNA in clinical
Methods for HPV Detection
71
9
10 11 12
13
14
15 16 17
18
19
20
21
22
specimens using a consensus PCR and a generic probe mix in an enzyme-linked immunosorbent assay format. J Clin Microbiol 2001;39:3530–3536. Chan PK, Cheung TH, Tam AO, Lo KW, Yim SF, Yu MM, To KF, Wong YF, Cheung JL, Chan DP, Hui M, Ip M: Biases in human papillomavirus genotype prevalence assessment associated with commonly used consensus primers. Int J Cancer 2006;118:243–345. Vernon SD, Unger ER, Williams D: Comparison of human papillomavirus detection and typing by cycle sequencing, line blotting, and hybrid capture. J Clin Microbiol 2000;38:651–655. Josefsson A, Livak K, Gyllensten U: Detection and quantitation of human papillomavirus by using the fluorescent 5⬘ exonuclease assay. J Clin Microbiol 1999;37:490–496. Quint W, Pagliusi SR., Lelie N, de Villiers EM, Wheeler CM, the WHO HPV DNA International Collaborative Study Group (2006): Results of the first WHO international collaborative study on the detection of human papillomavirus DNA. J Clin Microbiol 2006;in press. Nanda K, McCrory D, Myers E, Bastian L, Hasselblad V, Hickey J, Matchar D: Accuracy of the Papanicolaou test in screening for and follow-up of cervical cytologic abnormalities: a systematic review. Ann Intern Med 2000;132:810–819. Castle PE, Solomon D, Schiffman M, Wheeler CM for the ALTS Group: Human papillomavirus type 16 infections and 2-year absolute risk of cervical precancer in women with equivocal or mild cytologic abnormalities. J Natl Cancer Inst 2005;97:1066–1071. Goldie SJ, Kim J, Wright T: Cost-effectiveness of human papillomavirus DNA testing for cervical cancer screening in women aged 30 years or more. Obstet Gynecol 2004;103:619–631. Wright TC, Denny L, Kuhn L, Pollack A, Lorincz A: HPV DNA testing of self-collected vaginal samples compared with cytologic screening to detect cervical cancer. JAMA 2000;283:81–86. Wright TC, Cox JT, Massad LS, Twiggs LB, Wilkinson EJ, for the 2001 ASCCP-Sponsored Consensus Conference 2001: Consensus guidelines for the management of women with cervical cytological abnormalities. JAMA 2002;287:2120. Solomon D, Schiffman M, Tarone R, ALTS Study Group: Comparison of three management strategies for patients with atypical squamous cells of undetermined significance: baseline results from a randomized trial. J Nat Cancer Inst 2001;93:293–299. Cox JT, Schiffman M, Solomon D: ASCUS-LSIL Triage Study (ALTS) Group. Prospective followup suggests similar risk of subsequent cervical intraepithelial neoplasia grade 2 or 3 among women with cervical intraepithelial neoplasia grade 1 or negative colposcopy and directed biopsy. Am J Obstet Gynec 2003;188:1406–1412. Costa S, Simone PD, Venturoli S, Cricca M, Zerbini ML, Musiani M, Terzano P, Santini D, Cristiani P, Syrjänen S, Syrjänen K: Factors predicting Human papillomavirus (HPV) clearance in cervical intraepithelial neoplasia (CIN) lesions treated by conization. Gynecol Oncol 2003;90: 358–365. Zielinksi GD, Bais AG, Helmerorst TJ, Verheijen RH, De Schipper FA, Snijders PJ, et al: HPV testing and monitoring of women after treatment of CIN3: review of the literature and meta-analysis. Obstet Gynecol Surv 2004;59:543–553. IARC: The IARC Commitment to Cancer Prevention: The Example of Papillomavirus and Cervical Cancer. Lyon, International Agency for Research on Cancer, RRCR, 2005, vol 166, pp 277–297.
Prof. Suzanne M. Garland, MD Department of Microbiological Research, Department of Microbiology and Infectious Diseases The Royal Women’s Hospital, 132 Grattan Street Carlton, Vic. 3053 (Australia) Tel. ⫹61 3 9344 2476, Fax ⫹61 3 9344 3173, E-Mail
[email protected] Garland/Tabrizi
72
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 73–81
Molecular Markers for Cervical Cancer Renske D.M. Steenbergen, Chris J.L.M. Meijer, Peter J.F. Snijders Department of Pathology, VU University Medical Center, Amsterdam, The Netherlands
HPV and Cervical Cancer
Cervical cancer is a multistep process initiated by a high-risk HPV (HRHPV) infection. Although population-based cervical screening programs by cytomorphological examinations of cervical smears (i.e. Pap smears) have greatly reduced the incidence and mortality of cervical cancer in developed countries, cytology is still suboptimal because of a considerable degree of false negativity and false positivity. Therefore, there is a strong demand for additional markers to improve the quality and accuracy of cervical screening and the triage of women with equivocal mildly abnormal Pap smears. Since infection with HR-HPV is a prerequisite for the pathogenesis of cervical cancer, many studies have focused on the value of HR-HPV DNA testing in cervical screening and clinical policies to overcome the drawbacks of current cytological examination of cervical smears. It is expected that additional testing for HRHPV will markedly increase the sensitivity for high-grade cervical precancer lesions, i.e. cervical intraepithelial neoplasia (CIN), and cervical cancer [1, 2]. In addition, HR-HPV testing improves the clinical management of women with equivocal smears and the monitoring of women for post-treatment CIN3, as the negative predictive value of HR-HPV testing for CIN3 is extremely high [for a review, see ref. 2]. Despite the promising performance of cytology combined with HR-HPV testing, the positive predictive value of the HR-HPV test for ⱖCIN3 needs improvement. The population-based cervical screening program in the Netherlands includes about 1% of women with ⱖCIN3 and about 7.5% of women who are HR-HPV DNA positive. The latter group mainly involves women with normal cytology (about 4–5% of the screening population), and only 8% of these HR-HPV-positive women will have or acquire ⱖCIN3 within 4 years [3]. Consequently, the current HR-HPV DNA detection
techniques ask for a more precise individual risk assessment of women who will develop CIN3 and invasive cervical cancer. This emphasizes the need for additive markers to reduce the number of redundant follow-up smears of HRHPV-positive women with normal cytology, who will not have or develop ⱖCIN3. In addition, markers reflecting later stages in the multistep process of cervical carcinogenesis will allow identification of women having high-grade CIN lesions with invasive potential who need immediate referral to the gynecologist for colposcopy-directed biopsy. Such molecular markers can either be HPV or host cell specific. HPVspecific markers include viral load, viral typing, viral oncogene expression and viral integration, the potential value of which has been discussed previously [4, 5]. In addition, recent studies indicate that the pattern of viral methylation may be of clinical significance as well [6–9]. In the current paper, we will discuss nonviral (candidate) markers with particular emphasis on those that we encountered during the course of several research lines involving both clinical material and HPV-transfected epithelial cells that underwent progression from mortal to immortal stages.
Markers Reflecting Deregulated E6/E7 Expression and Immortalization
Whereas productive HPV infections are characterized by expression of the viral oncogenes E6/E7 in the differentiated cell layers, deregulated expression of E6/E7 in the dividing basal cells is suggested to represent the first step in the multistep process of HPV-mediated transformation [4]. Interference of the viral oncogenes E6 and E7 with the apoptosis and cell cycle regulators p53 and pRb in the proliferating cells results in the induction of genomic instability. The genomically instable environment provides the driving force for the acquisition of crucial alterations in oncogenes and tumor suppressor genes that are additive requirements for malignant transformation. The inactivation of Rb as a result of deregulated HR-HPV E7 expression is characterized by a permanent upregulation of the cyclin-dependent kinase inhibitor p16INK4A owing to a disruption of the Rb-dependent negative feed-back loop regulating p16INK4A expression. In line with this, diffuse p16INK4A immunohistochemical staining patterns can be found in HR-HPV-containing high-grade CIN lesions, cervical carcinomas [10, 11] and carcinomas of other anogenital sites that are HR-HPV related, such as penile carcinomas [12] and HR-HPVpositive oropharyngeal carcinomas [13]. Recent data indicate that p16INK4A immunostaining may provide a useful adjunctive test for a reliable histological examination of cervical biopsies [14, 15]. In addition, several studies have shown
Steenbergen/Meijer/Snijders
74
that the detection of increased p16INK4A expression by immunocytochemistry in liquid-based cytology specimens may provide a promising marker for the detection of HR-HPV-containing CIN lesions that display deregulated E7 expression [16, 17]. However, larger studies on cytological specimens are needed before the significance of p16INK4A expression in cytological specimens is clarified. The uncontrolled cell proliferation resulting from the inactivation of Rb and concomitant induction of E2F is also reflected by an altered expression of proliferation markers, such as minichromosome maintenance proteins (MCMs), PCNA and Ki-67 in CIN lesions and cervical carcinomas [for a review, see [18]]. As a consequence, these markers may aid in an improvement in histopathological assessment of cervical biopsies. Moreover, preliminary data have shown that staining for MCM5 in liquid-based cytology smears may have potential as a marker of exfoliated cells exhibiting moderate and severe dyskaryosis [15]. In vitro immortalization of human epithelial cells transfected by full-length HR-HPV genomes has been recognized as the first altered phenotype that, next to HPV functions, requires additive alterations in host cell genes [19]. This phenotype is characterized by an increased activity of the telomere-lengthening enzyme telomerase, resulting from elevated expression of the gene, encoding its catalytic subunit human telomerase reverse transcriptase (hTERT) [20, 21]. The telomerase enzyme interferes with the normal erosion of telomeres in dividing cells, thereby disturbing the intrinsic division count system that determines the lifespan of normal somatic cells. Increased activity of the telomerase enzyme and elevated mRNA levels of its catalytic subunit hTERT has been shown in almost all cervical squamous cell carcinomas (SCCs) and a subset of CIN3 lesions, whereas these features were absent or extremely rare in normal cervix, CIN1 and CIN2 lesions [22, 23]. It can be speculated that telomerase-positive CIN lesions have gained an immortal phenotype and as such have reached a point of no return in terms of malignant potential. However, the detection of telomerase activity in fresh cervical scrapes and also the detection of hTERT by immunohistochemistry in thinlayer cervical cytology preparations did not reflect the status of these markers in the underlying lesions, suggesting that these markers lack specificity and sensitivity for clinically relevant cervical lesions [24–27]. By means of microarray expression analysis of mortal and immortal stages of HPV-transfected keratinocytes, we recently identified a number of genespecific markers that were strongly associated with immortalization and deregulated hTERT expression [de Wilde, and Steenbergen et al., unpubl. data]. The potential value of these markers to predict high-grade CIN and cervical cancer is currently being studied using clinical specimens. In addition, immortalization of HPV-transfected primary keratinocytes was found to be associated with specific chromosomal alterations, i.e. clonal allelic losses at chromosomes 3p, 6q, 10p, 11p, 11q, 13q and/or 18q, nonrandom gains at chromosomes 5, 7q, 8q,
Markers for Cervical Cancer
75
9q and 20q, and structural alterations at, amongst others, chromosomes 10, 18 and 20 [for a review, see [28]]. Interestingly, many chromosomal alterations identified in HPV-immortalized cells are also common in cervical carcinomas. These include allele losses at 3p and 6q, which are often detected in CIN3 lesions as well [28, 29]. Allelic loss at 6q14–22 was significantly more frequent in CIN3 lesions that displayed telomerase activity and elevated hTERT mRNA levels [29]. Together with the fact that chromosome 6 can induce downregulation of telomerase in HPV-immortalized keratinocytes and a cervical carcinoma (i.e. SiHa) cell line [30], this suggests that yet unknown gene(s) located at 6q may provide promising future marker(s) for immortality and telomerase activity. A recent genome-wide microarray-based comparative genomic hybridization (CGH) analysis has revealed that gains at 3q12.1–28 and 20q11.21–13.33 are highly frequent in both HPV-immortalized cells and cervical SCCs [31]. Candidate oncogenes at 3q are PIK3CA, the catalytic subunit alpha of phosphatidylinositol 3-kinase, and hTERC, coding for the structural RNA component of telomerase that serves as a template during telomere elongation. An increased expression of PIK3CA was not only observed in cervical carcinomas [32], but also in liquid-based cytology samples of women with CIN3 lesions [33]. Studies using fluorescence in situ hybridization (FISH) have shown that copy number increases of the hTERC gene are highly frequent in high-grade CIN lesions, whereas only a few normal controls and low-grade CIN lesions revealed cells with increased hTERC copy numbers [34]. In a retrospective study on archival smears, hTERC copy number increases could be detected by FISH in smears of women with CIN1/2 who later developed CIN3, and not in smears of women with CIN1/2 lesions that regressed spontaneously. Moreover, hTERC gains were detected in normal Pap smears of women who were later diagnosed with CIN3 or invasive cervical cancer [35]. Further fine mapping of the 20q region by multiplex ligation-dependent probe amplification analysis revealed frequent copy number increases of REM1, DNMT3B, E2F1, TOP1 and ADA in HPV-immortalized cells and cervical carcinomas [31]. Future FISH analysis and immunohistochemical staining will reveal their potential value as molecular markers for CIN3 lesions and cervical cancer.
Markers Reflecting Malignant Transformation and Tumor Invasion
By array CGH analysis, a gain at 1q21.1–31.1 and losses of 11q22.3–25 and 13q14.3–21.33 were found to be rather common in invasive cervical carcinomas, but infrequent in nontumorigenic HPV-immortalized epithelial cells [31]. Therefore, genes at these regions may yield novel candidate biomarkers for tumor invasion. The genomic alterations identified by array CGH are generally
Steenbergen/Meijer/Snijders
76
in concordance with those that were previously identified by conventional CGH studies, as summarized previously [28]. In addition to DNA markers derived from CGH array studies, novel candidate RNA or protein markers may arise from high-resolution microarray expression analyses. To date, a limited number of microarray expression analysis studies have been performed on cervical carcinomas, which have resulted in various lists of cancer-specific over- and underexpressed genes [36–40]. These lists include genes regulating cellular proliferation as well as genes encoding adhesion molecules, matrix proteins and matrix metalloproteinases.
DNA Methylation Markers
Recently, methylation of CpG-rich sequences, so-called CpG islands, located within promoter regions, has been shown to be involved in the silencing of tumor suppressor genes, thereby contributing to carcinogenesis [41]. DNA methylation involves the covalent addition of a methyl (CH3) group at the carbon 5 position of cytosine. Many different studies have shown altered methylation profiles in tumors in which the bulk of the genome is unmethylated, whereas promoters contained within CpG islands are hypermethylated [41]. These include promoters of tumor suppressor genes and mismatch repair genes. In our own recent studies, we showed that silencing of the tumor suppressor in lung cancer 1 (TSLC1) gene by promoter hypermethylation may be a valuable marker for cervical lesions with invasive potential [42]. TSLC1 was found to be silenced in 91% (10/11) of cervical cancer cell lines, primarily resulting from promoter hypermethylation. Ectopic TSLC1 expression suppressed both anchorage-independent growth and tumor growth in nude mice of SiHa cervical cancer cells, whereas the immortal phenotype was unaffected. Furthermore, TSLC1 promoter hypermethylation was detected in 58% of cervical carcinomas and 35% of high-grade CIN lesions, but not in low-grade CIN lesions and normal cervix. Interestingly, TSLC1 promoter hypermethylation could be detected in archival cervical smears of women with cervical cancer taken up to 7 years before cancer diagnosis. Thus, detection of TSLC1 silencing in cervical smears may allow the identification of women having nonregressing CIN lesions with invasive potential. The high frequency of TSLC1 methylation in cervical SCC was confirmed by studies of Li et al. [43] and Gustafson et al. [44], showing TSLC1 methylation in 65% of cervical cancer and 63% of high-grade dysplasias, respectively. Amongst 15 tumor suppressor genes (i.e. DAPK1, TSLC1, CDKN2A, HIC1, TP73, P14ARF, MGMT, RASSF1A, MLH1, RARb, PTEN, CDKN2B, CDH1, CCND2 and GSTP1) analyzed for methylation, TSLC1 appeared the best individual gene in distinguishing high-grade
Markers for Cervical Cancer
77
dysplasia from combined low-grade dysplasia and normal smears. Another study involving gene methylation analysis of a panel of 20 genes in which TSLC1 was not included showed that in combination hypermethylation of DAPK1, RARb and TWIST provided the most informative panel of markers for CIN3/carcinoma in situ and cervical carcinoma [45].
Validation of Molecular Markers
The studies summarized above have resulted in long lists of candidate biomarkers that may improve the early detection of cervical cancer. The number of candidate biomarkers is expected to expand further, as more genomic and proteomic studies will appear in the near future. However, it should be realized that before these markers can be integrated into clinical practice, clinical validation on prospectively collected cervical specimens is needed. Recently, guidelines have been proposed for the development of biomarkerbased screening tools for early detection of cancer, which can be categorized into five phases: (1) preclinical exploratory studies for marker discovery, (2) clinical assay development and validation, (3) retrospective longitudinal repository studies, (4) prospective screening, and (5) cancer control studies, including costbenefit analysis [46]. Most of the markers listed above have not yet passed the first phases. Therefore, at the moment, it is difficult to predict which of the markers or marker panels are ultimately the most promising candidates for the early detection of CIN lesions and invasive cervical carcinomas. Nevertheless, p16INK4A immunohistochemical studies on histopathological specimens show that a diffuse staining pattern is a valuable marker not only in the recognition of HR-HPV-positive CIN lesions but may also help to distinguish between highgrade CIN and repairing/reactive lesions. Furthermore, based on the presently available data, it might be expected that the detection p16INK4A overexpression or other early markers reflecting deregulated E6/E7 expression, e.g. MCM5, in combination with HR-HPV testing will improve the early detection of clinically relevant CIN lesions in cervical scrapes. Provided that these markers can be reliably applied to cervical scrapes with optimal clinical sensitivity and specificity, they will allow the stratification of women with clinically relevant HR-HPV infections that need follow-up. On the other hand, markers reflecting immortalization and tumor invasion, such as the detection of hTERC or PIK3CA copy number increases and TSLC1 promoter hypermethylation, are more likely to identify women with high-grade premalignant lesions who need immediate referral for colposcopy because of the risk of having nonregressing lesions with invasive potential.
Steenbergen/Meijer/Snijders
78
References 1 2 3
4 5 6
7
8
9
10
11
12
13
14
15
16 17 18 19
Lorincz AT, Richart RM: Human papillomavirus DNA testing as an adjunct to cytology in cervical screening programs. Arch Pathol Lab Med 2003;127:959–968. Brink AA, Zielinski GD, Steenbergen RD, Snijders PJ, Meijer CJ: Clinical relevance of human papillomavirus testing in cytopathology. Cytopathology 2005;16:7–12. Rozendaal L, Walboomers JM, van der Linden JC, Voorhorst FJ, Kenemans P, Helmerhorst TJ, van Ballegooijen M, Meijer CJ: PCR-based high-risk HPV test in cervical cancer screening gives objective risk assessment of women with cytomorphologically normal cervical smears. Int J Cancer 1996;68:766–769. Snijders PJ, Steenbergen RD, Heideman DA, Meijer CJ: HPV-mediated cervical carcinogenesis: concepts and clinical implications. J Pathol 2006;208:152–164. Brink AA, Snijders PJ, Meijer CJ, Berkhof J, Verheijen RH: HPV testing in cervical screening. Best Pract Res Clin Obstet Gynaecol 2006, in press. Badal S, Badal V, Calleja-Macias IE, Kalantari M, Chuang LS, Li BF, Bernard HU: The human papillomavirus-18 genome is efficiently targeted by cellular DNA methylation. Virology 2004; 324:483–492. Badal V, Chuang LS, Tan EH, Badal S, Villa LL, Wheeler CM, Li BF, Bernard HU: CpG methylation of human papillomavirus type 16 DNA in cervical cancer cell lines and in clinical specimens: genomic hypomethylation correlates with carcinogenic progression. J Virol 2003;77:6227–6234. Kim K, Garner-Hamrick PA, Fisher C, Lee D, Lambert PF: Methylation patterns of papillomavirus DNA, its influence on E2 function, and implications in viral infection. J Virol 2003;77: 12450–12459. Kalantari M, Calleja-Macias IE, Tewari D, Hagmar B, Lie K, Barrera-Saldana HA, Wiley DJ, Bernard HU: Conserved methylation patterns of human papillomavirus type 16 DNA in asymptomatic infection and cervical neoplasia. J Virol 2004;78:12762–12772. Klaes R, Friedrich T, Spitkovsky D, Ridder R, Rudy W, Petry U, Dallenbach-Hellweg G, Schmidt D, von Knebel DM: Overexpression of p16(INK4A) as a specific marker for dysplastic and neoplastic epithelial cells of the cervix uteri. Int J Cancer 2001;92:276–284. Zielinski GD, Snijders PJ, Rozendaal L, Daalmeijer NF, Risse EK, Voorhorst FJ, Jiwa NM, van der Linden HC, de Schipper FA, Runsink AP, Meijer CJ: The presence of high-risk HPV combined with specific p53 and p16INK4a expression patterns points to high-risk HPV as the main causative agent for adenocarcinoma in situ and adenocarcinoma of the cervix. J Pathol 2003;201:535–543. Ferreux E, Lont AP, Horenblas S, Gallee MP, Raaphorst FM, von Knebel DM, Meijer CJ, Snijders PJ: Evidence for at least three alternative mechanisms targeting the p16INK4A/cyclin D/Rb pathway in penile carcinoma, one of which is mediated by high-risk human papillomavirus. J Pathol 2003;201: 109–118. Klussmann JP, Gultekin E, Weissenborn SJ, Wieland U, Dries V, Dienes HP, Eckel HE, Pfister HJ, Fuchs PG: Expression of p16 protein identifies a distinct entity of tonsillar carcinomas associated with human papillomavirus. Am J Pathol 2003;162:747–753. Dray M, Russell P, Dalrymple C, Wallman N, Angus G, Leong A, Carter J, Cheerala B: p16(INK4a) as a complementary marker of high-grade intraepithelial lesions of the uterine cervix. 1. Experience with squamous lesions in 189 consecutive cervical biopsies. Pathology 2005;37:112–124. Murphy N, Ring M, Heffron CC, King B, Killalea AG, Hughes C, Martin CM, McGuinness E, Sheils O, O’Leary JJ: p16INK4A, CDC6, and MCM5: predictive biomarkers in cervical preinvasive neoplasia and cervical cancer. J Clin Pathol 2005;58:525–534. Bibbo M, DeCecco J, Kovatich AJ: p16INK4A as an adjunct test in liquid-based cytology. Anal Quant Cytol Histol 2003;25:8–11. Sahebali S, Depuydt CE, Segers K, Moeneclaey LM, Vereecken AJ, Van Marck E, Bogers JJ: p16INK4a as an adjunct marker in liquid-based cervical cytology. Int J Cancer 2004;108:871–876. Baldwin P, Laskey R, Coleman N: Translational approaches to improving cervical screening. Nat Rev Cancer 2003;3:217–226. Chen TM, Pecoraro G, Defendi V: Genetic analysis of in vitro progression of human papillomavirus-transfected human cervical cells. Cancer Res 1993;53:1167–1171.
Markers for Cervical Cancer
79
20
21 22
23 24
25 26
27
28 29
30
31
32 33
34
35
36
Steenbergen RDM, Walboomers JMM, Meijer CJLM, van der Raaij-Helmer EM, Parker JN, Chow LT, Broker TR, Snijders PJF: Transition of human papillomavirus type 16 and 18 transfected human foreskin keratinocytes towards immortality: activation of telomerase and allele losses at 3p, 10p, 11q and/or 18q. Oncogene 1996;13:1249–1257. Klingelhutz AJ, Barber SA, Smith PP, Dyer K, McDougall JK: Restoration of telomeres in human papillomavirus-immortalized human anogenital epithelial cells. Molec Cell Biol 1994;14:961–969. Snijders PJ, van Duin M, Walboomers JMM, Steenbergen RDM, Risse EKJ, Helmerhorst TJM, Verheijen RHM, Meijer CJLM: Telomerase activity exclusively in cervical carcinomas and a subset of cervical intraepithelial neoplasia grade III lesions: strong association with elevated messenger RNA levels of its catalytic subunit and high-risk human papillomavirus DNA. Cancer Res 1998;58:3812–3818. Wisman GB, de Jong S, Meersma GJ, Helder MN, Hollema H, de Vries EG, Keith WN, van der Zee AG: Telomerase in (pre)neoplastic cervical disease. Hum Pathol 2000;31:1304–1312. Wisman GBA, Hollema H, de Jong S, ter Schegget J, Tjongahung SP, Ruiters MHJ, Krans M, de Vries EGE, van der Zee AGJ: Telomerase activity as a biomarker for (pre)neoplastic cervical disease in scrapings and frozen sections from patients with abnormal cervical smear. J Clin Oncol 1998;16:2238–2245. Ngan HY, Cheung AN, Liu SS, Liu KL, Tsao SW: Telomerase assay and HPV 16/18 typing as adjunct to conventional cytological cervical cancer screening. Tumor Biol 2002;23:87–92. Cheung AN, Chiu PM, Tsun KL, Khoo US, Leung BS, Ngan HY: Chromosome in situ hybridisation, Ki-67, and telomerase immunocytochemistry in liquid based cervical cytology. J Clin Pathol 2004;57:721–727. Reesink-Peters N, Helder MN, Wisman GB, Knol AJ, Koopmans S, Boezen HM, Schuuring E, Hollema H, de Vries EG, de Jong S, van der Zee AG: Detection of telomerase, its components, and human papillomavirus in cervical scrapings as a tool for triage in women with cervical dysplasia. J Clin Pathol 2003;56:31–35. Steenbergen RD, de Wilde J, Wilting SM, Brink AA, Snijders PJ, Meijer CJ: HPV-mediated transformation of the anogenital tract. J Clin Virol 2005;32(suppl 1):S25–S33. van Duin M, Steenbergen RD, de Wilde J, Helmerhorst TJ, Verheijen RH, Risse EK, Meijer CJ, Snijders PJ: Telomerase activity in high-grade cervical lesions is associated with allelic imbalance at 6Q14–22. Int J Cancer 2003;105:577–582. Steenbergen RD, Kramer D, Meijer CJ, Walboomers JM, Trott DA, Cuthbert AP, Newbold RF, Overkamp WJ, Zdzienicka MZ, Snijders PJ: Telomerase suppression by chromosome 6 in a human papillomavirus type 16-immortalized keratinocyte cell line and in a cervical cancer cell line. J Natl Cancer Inst 2001;93:865–872. Wilting SM, Snijders PJF, Meijer GA, Ylstra B, van den Ijssel PRLA, Snijders AM, Albertson DG, Coffa J, Schouten JP, van de Wiel MA, Meijer CJLM, Steenbergen RDM: Frequent gene copy number increases at chromosome 20q in both squamous cell carcinomas and adenocarcinomas of the cervix. J Pathol 2006;209:in press. Ma YY, Wei SJ, Lin YC, Lung JC, Chang TC, Whang-Peng J, Liu JM, Yang DM, Yang WK, Shen CY: PIK3CA as an oncogene in cervical cancer. Oncogene 2000;19:2739–2744. Goto T, Takano M, Sasa H, Tsuda H, Yamauchi K, Kikuchi Y: Clinical significance of immunocytochemistry for PIK3CA as a carcinogenesis-related marker on liquid-based cytology in cervical intraepithelial neoplasia. Oncol Rep 2006;15:387–391. Heselmeyer-Haddad K, Janz V, Castle PE, Chaudhri N, White N, Wilber K, Morrison LE, Auer G, Burroughs FH, Sherman ME, Ried T: Detection of genomic amplification of the human telomerase gene (TERC) in cytologic specimens as a genetic test for the diagnosis of cervical dysplasia. Am J Pathol 2003;163:1405–1416. Heselmeyer-Haddad K, Sommerfeld K, White NM, Chaudhri N, Morrison LE, Palanisamy N, Wang ZY, Auer G, Steinberg W, Ried T: Genomic amplification of the human telomerase gene (TERC) in pap smears predicts the development of cervical cancer. Am J Pathol 2005;166: 1229–1238. Chen Y, Miller C, Mosher R, Zhao X, Deeds J, Morrissey M, Bryant B, Yang D, Meyer R, Cronin F, Gostout BS, Smith-McCune K, Schlegel R: Identification of cervical cancer markers by cDNA and tissue microarrays. Cancer Res 2003;63:1927–1935.
Steenbergen/Meijer/Snijders
80
37
38
39
40
41 42
43 44
45
46
Sopov I, Sorensen T, Magbagbeolu M, Jansen L, Beer K, Kuhne-Heid R, Kirchmayr R, Schneider A, Durst M: Detection of cancer-related gene expression profiles in severe cervical neoplasia. Int J Cancer 2004;112:33–43. Rosty C, Sheffer M, Tsafrir D, Stransky N, Tsafrir I, Peter M, de Cremoux P, de La RA, Salmon R, Dorval T, Thiery JP, Couturier J, Radvanyi F, Domany E, Sastre-Garau X: Identification of a proliferation gene cluster associated with HPV E6/E7 expression level and viral DNA load in invasive cervical carcinoma. Oncogene 2005;24:7094–7104. Santin AD, Zhan F, Bignotti E, Siegel ER, Cane S, Bellone S, Palmieri M, Anfossi S, Thomas M, Burnett A, Kay HH, Roman JJ, O’Brien TJ, Tian E, Cannon MJ, Shaughnessy J Jr, Pecorelli S: Gene expression profiles of primary HPV16- and HPV18-infected early stage cervical cancers and normal cervical epithelium: identification of novel candidate molecular markers for cervical diagnosis and therapy. Virology 2005;331:269–291. Vazquez-Ortiz G, Ciudad CJ, Pina P, Vazquez K, Hidalgo A, Alatorre B, Garcia JA, Salamanca F, Peralta-Rodriguez R, Rangel A, Salcedo M: Gene identification by cDNA arrays in HPV-positive cervical cancer. Arch Med Res 2005;36:448–458. Esteller M, Herman JG: Cancer as an epigenetic disease: DNA methylation and chromatin alterations in human tumours. J Pathol 2002;196:1–7. Steenbergen RD, Kramer D, Braakhuis BJ, Stern PL, Verheijen RH, Meijer CJ, Snijders PJ: TSLC1 gene silencing in cervical cancer cell lines and cervical neoplasia. J Natl Cancer Inst 2004;96:294–305. Li J, Zhang Z, Bidder M, Funk MC, Nguyen L, Goodfellow PJ, Rader JS: IGSF4 promoter methylation and expression silencing in human cervical cancer. Gynecol Oncol 2005;96:150–158. Gustafson KS, Furth EE, Heitjan DF, Fansler ZB, Clark DP: DNA methylation profiling of cervical squamous intraepithelial lesions using liquid-based cytology specimens: an approach that utilizes receiver-operating characteristic analysis. Cancer 2004;102:259–268. Feng Q, Balasubramanian A, Hawes SE, Toure P, Sow PS, Dem A, Dembele B, Critchlow CW, Xi L, Lu H, McIntosh MW, Young AM, Kiviat NB: Detection of hypermethylated genes in women with and without cervical neoplasia. J Natl Cancer Inst 2005;97:273–282. Pepe MS, Etzioni R, Feng Z, Potter JD, Thompson ML, Thornquist M, Winget M, Yasui Y: Phases of biomarker development for early detection of cancer. J Natl Cancer Inst 2001;93:1054–1061.
Dr. R.D.M. Steenbergen, PhD Department of Pathology, Unit of Molecular Pathology VU Universiteit Medical Center, PO Box 7057 NL–1007 MB Amsterdam (The Netherlands) Tel. ⫹31 20 4440503, Fax ⫹31 20 4442964, E-Mail
[email protected] Markers for Cervical Cancer
81
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 82–102
Direct Detection of Cervical Carcinogenesis through mRNA H. Skomedal b, I. Kraus a, I. Silvab, T. Moldena, S. Hovland b, G. Morland b, E. Morland b, F. Karlsen b a
Institute of Pathology, National University Hospital, Oslo, and NorChip AS, Klokkarstua, Norway
b
Challenges with Current Screening Methods
The first screening programs were based on using the conventional Papanicolaou smear (Pap) which dates back to over 50 years ago. This technology has been complemented with the use of improved cytological testing which has allowed for a way to detect the onset of invasive cervical cancer incidence. However, the performance has been called into question. In particular, the specificity and sensitivity of the test have been lower than originally expected, as shown by different meta-analysis studies [1]. Furthermore, a wide variety of performances from different laboratories have been shown, ranging from high to low specificity and sensitivity [2, 3]. Another facet of consideration is the lack of a standardized sampling site (specimen collection) which is largely left to the individual samplers to decide. Liquid-Based Cytology Liquid-based cytology was introduced in the 1990s as a way to improve the current method of testing. However, there have been significant challenges related to technical problems, as for example residual cellular debris. Therefore, it has been recommended that all labs should compare any cytologically abnormal samples with a histopathological review. Other issues to be addressed using this technique include the extensive training needed for the screeners, and the mandatory need for workload limits as the number of high-grade squamous intraepithelial lesion (HSIL) cells on the slide can be extremely small. Due to this very small percentage of cells, there is a need to rescreen negative samples as well as for a further review of these slides by a pathologist.
Visual Inspection using Acetic Acid This technique involves the use of visual inspection of the cervix after the application of a 3–5% solution of acetic acid. This is a very simple test in which positive samples are graded according to the appearance of acetowhite areas in the transformation zone. While this test is very easy to use by various providers, the actual diagnostic value is biased due to substantial variation which has been reported in the positivity rates which range from 3 to 27%, a sensitivity range from 37 to 92% [4, 5] and a specificity range from 49 to 79% [6, 7]. These numbers only reflect studies which have minimal verification bias, and therefore, they should be considered with caution. Furthermore, the acetowhite effect is not specific to cervical neoplasia and may also occur in immature squamous metaplasia and in inflamed, regenerating cervical epithelium. Colposcopy Colposcopy is a procedure which allows for magnified viewing of the cervix and the vagina using various solutions (normal saline to Lugol’s iodine) which are applied to the cervical epithelium. The aim of this test is to examine the transformation zone and grade abnormal areas according to morphology, acetowhiteness, lesion margin and surface configuration, blood vessels and iodine uptake. Although useful, there are several constraints associated with this technique such as cost and the high level of anxiety in women. The specificity, sensitivity and the acquisition of the sample as well as the accuracy and reproducibility can all be biased, the latter due to the disease status being confirmed by other methods such as cytology prior to testing. Other factors negatively influencing this technique include the type of definition of the abnormality being used and the poor agreement of early disease onset by the colposcopists. DNA-Based Testing for Presence of HPV Hybrid Capture 2 (hc2; Digene, Gaithersburg, Md., USA) is a commercially available and commonly used HPV DNA test which is based on nucleic acid hybridization with signal amplification for the qualitative detection of high-risk HPV types associated with cervical cancer. The main advantage of the hc2 is a high sensitivity and the abundant clinical data which have allowed it to be Food and Drug Administration approved. However, there are many limitations to this technique which have to be taken into account. The test is a timeconsuming and labor-intensive procedure which cannot determine the specific HPV type present. In addition, it does not include internal controls of quality or performance. While the detection is based on 13 high-risk types, it suffers from
mRNA Detection in Carcinogenesis
83
a cross-reactivity with low-risk types [8] which is a considerable problem in populations with a high predominance of low-risk types. Due to the technique targeting the structural region (L1) of the virus as opposed to the viral oncogenes, the test may fail to detect any active infection of the virus should the L1 be deleted or mutated. Furthermore, most of the studies have not included histopathological negative cases in their calculations and the interval cancer incidence has not been accounted for in calculations of the sensitivity and specificity of this test.
Potential Screening Methods
PCR and Reverse Transcriptional PCR Recently, there has been a growing trend of using homebrew setups for HPV testing. Currently, both nonquantitative and quantitative tests are used for HPV DNA detection. The nonquantitative techniques commonly amplify the L1 region using consensus PCR primers [9–11]. This type of setup allows for multiplexing, which is helpful for genotyping purposes. Some tests have been further developed to give new modified versions, with detection and typing being based on line blot or dot blot instead of gel electrophoresis. The main limitation of the different consensus PCR methods is the relative high number of false-positive cases detected, resulting in a rather low clinical specificity. In addition, the consensus primers have been shown to be biased in their amplification [12] which further compounds the obstacles behind standardization of the technique for diagnostic purposes. Reverse transcriptional (RT)-PCR is a powerful, reproducible method. Realtime RT-PCR has been used as a detection technique which requires great care, e.g., loss of standardization in the setup and in the reporting procedure. Factors involved in standardization include normalization, type and number of reference genes, absolute or relative quantification of viral load, standard curves and how to best generate them, justification and validation of the chemistry chosen, as well as the primers, probes and the detection machines to be used. Other facets to consider include any biases encountered in samples that have coinfections of HPV, as well as the difficulty in multiplexing for detection of these coinfections. However, the main scientific problem with RT-PCR is that any contamination with DNA may cause a false-positive result. In order to be sure that mRNA is amplified, RTPCR is dependent upon poly A oligos for the downstream primer. With an upstream primer located far from the poly A primer, it gets difficult to amplify long mRNA without drastically reducing the analytical sensitivity.
Skomedal/Kraus/Silva/Molden/Hovland/Morland/Morland/Karlsen
84
Table 1. Comparison between different commercial HPV detection methods Methods
hc2 Amplicore HPV test Linear Array HPV genotyping PreTect HPV-Proofer
Patient
Principle
Detects persistence
Shows active infection
Avoid unnecessary follow-up
Detect DNA/ mRNA
Genotyping HPV
No No Yes Yes
No No No Yes
No No No Yes
DNA DNA DNA mRNA
No No Yes Yes
Reproduced with permission from Oestfold Hospital Trust, Fredrikstad, Norway.
PreTect HPV-Proofer, a New Direct Detector of Cervical Carcinogenesis A compelling test based on nucleic acid sequence-based amplification (NASBA) has been developed by NorChip AS (Klokkarstua, Norway). The main mode of detection is by the surveillance of oncogenic activity encoded by the E6 and E7 regions of HPV. The literature, clearly and abundantly, describes these proteins to be absolutely necessary for the onset of cervical cancer. Further information on the technical background of this test, including clinical data, is described below (table 1). Biomarkers A relatively new set of biomarkers has come onto the market with the purpose of identifying the risk at each neoplastic stage. Ideally, the candidate markers should give a predictive value of the progression and regression of HPV infection. These biomarkers should define the different stages of cellular changes associated with HPV clearance, HPV persistence and progression to precancer. Commonly used markers for cell cycle deregulation include cdc2, cdc6, mcm5, mcm2, ki-67 and survivin. Survivin, in particular, has been linked to having the same strong predictors of a high-risk HPV infection as p16INK4a, being upregulated due to p53 suppression by the HPV E6 oncoprotein [13]. However, as the cell normally undergoes cellular changes such as regeneration, the clinical usefulness of these markers for cervical cancer is questioned. Although several biomarkers have been used for the detection of neoplastic changes in cervical tissues, these biomarkers must not be considered a replacement for the detection of oncogenic HPV genes.
mRNA Detection in Carcinogenesis
85
New Possibilities with mRNA Technology: Chip Technology
The challenge in the future of diagnostics hinges on the development of a fully automated and disposable diagnostic microsystem with integrated sample preparation and detection modules for virus and bacteria identification. Shorter handling time, combined with reduced reagent and sample consumption, will be the benefits of this system as compared with conventional methods. To date, experimental evidence for real-time NASBA detection in cyclic olefin copolymer microchips, with 80-nl detection volumes, has been carried out [14]. The detection limits are comparable with those obtained for experiments performed in conventional routine-based laboratory systems, demonstrating that the microchip and its detection system have a potential for diagnostic use in a point-of-care setting. Future microchips could contain more reaction channels and be combined with multiplexing of several different targets in each of the channels. Finally, an integration of this microchip, including a sample preparation microchip, would constitute a fully automatic, laboratory-independent diagnostic system, resulting in an overall time and cost reduction of the whole analysis.
DNA Compared with mRNA as Targets for Routine Diagnostics
A main problem with HPV DNA testing is the high prevalence of HPV among women with a cytologically normal, atypical squamous cells of undetermined significance (ASCUS) or low-grade squamous intraepithelial lesion (LSIL) Pap smear, compared with the number of women actually developing severe dysplasia. Commercially available DNA HPV testing assays have exclusively focused on the use of detecting the L1 structural region of the HPV virus. The L1 capsid protein is expressed late in the viral life cycle in highly differentiated suprabasal cells [15] and has been a primary target of choice for both DNA tests and newly developed HPV vaccines. While the L1 region of the virus does play a role in the structural makeup of the virus, it is clearly not involved in the oncogenic mechanism. Therefore, the DNA tests do not detect the activity of the HPV oncoproteins which is the only true measure for the development of carcinogenic changes in the cell. In case of a loss of the L1 region, due to viral integration into the human genome, DNA tests will incorrectly define the sample as being negative, even though there could be aggressive activity, which could lead to cellular changes by the oncoproteins E6/E7.
Skomedal/Kraus/Silva/Molden/Hovland/Morland/Morland/Karlsen
86
E4
L1 E4
CIN2 LSIL
CIN3 HSIL
E7
Viral DNA
Viral DNA
E7
E7
L1 E4 Viral DNA
E7
CIN1
CIN3 Cervical cancer
Fig. 1. Diagrammatic representation of the skin and changes in the HPV16 gene expression during the development of cervical cancer. During progression from CIN1 to CIN3, normal regulation of the papillomavirus life cycle is lost. CIN1 lesions generally resemble productive lesions caused by other supergroup A HPV types and express virus coat proteins at the epithelial surface. In CIN2 and CIN3 lesions, the order of life cycle events is unchanged, but the extent of E7 expression is increased. (Reproduced with permission from John Doorbar.)
HPV Expression during Cervical Carcinogenesis
The oncogenic potential of the high-risk HPV E6 and E7 genes is confirmed by numerous studies: the manifestation of E6 and E7 expression in tumor material [Kraus et al., accepted for publication], the demonstration of their transforming properties and interaction with growth-regulating host cell proteins, and the requirement for E6 and E7 expression to maintain the malignant phenotype of cervical carcinoma [16]. Recent data also suggest that E6 and E7 play an important role in the inhibition of the host cell immune response [17, 18]. The pattern of HPV gene expression changes in different layers of the epithelium and varies along with the grade of the lesion (fig. 1). Consequently, by looking at the transcription pattern, it may be possible to predict the severity of the underlying disease and the outcome of the infection. Additionally, the usage of multiple promoters, mRNA splice sites and polyadenylation sites results in different forms of transcripts. In general, the viral life cycle is tightly linked to the differentiation of the human epithelium. In the undifferentiated basal cells, the viral promoter and enhancer sequences are repressed both by viral and host transcription factors. In the case of a benign HPV infection, the viral genome is present as an episome, and E4 and E5 are typically the most highly expressed proteins. E6 and E7 proteins are necessary for S-phase entry and will be expressed at a very low level in the basal cells [19], sufficient for the start of viral replication. In the upper layers of the epithelium, transcription of the early viral replication proteins E1 and E2 will be upregulated. E2 works as
mRNA Detection in Carcinogenesis
87
an activator of replication by permitting E1 to specifically bind to the viral origin of replication. In addition, E2 also acts as a repressor on the transcription of E6 and E7. At later stages of differentiation, the late promoter is activated and the capsid proteins L1 and L2 are produced; production of viral particles is only detected in highly differentiated keratinocytes. In cancers, viral DNA is often found integrated into the host genome, occurring within the E1/E2 region. The viral transcriptional control by E2 will be lost and, as a consequence, the expression of the viral oncoproteins E6 and E7 will increase to a level causing transformation. Expression of the E6 and E7 proteins will block the process of differentiation and, in the case of highgrade lesions, E6 and E7 are expressed throughout the epithelium (fig. 1). Therefore, detection of E6/E7 transcripts is a more precise indicator for progression towards malignancy, as compared with e.g. the detection of HPV DNA. Function and Importance of the Different E6/E7 mRNA Transcripts HPV E6 and E7 expression is primarily regulated at the transcriptional or posttranscriptional level. For example, for HPV 16, the E6 open reading frame encodes at least three distinct variants of the E6 protein, which may all have different roles in the viral life cycle. These transcripts are either unspliced (fulllength E6/E7 transcript) or spliced transcripts: E6*I is spliced from nucleotide 226 to 409 and E6*II from nucleotide 226 to 526, all being transcribed from the promoter p97 located just upstream of the second ATG of the E6 open reading frame. The full-length E6 protein has been reported to be translated from the E6 unspliced transcript. The E7 protein is likely encoded by the E6*I or E6*II and, for some time, it was thought that this splicing event was a means of obtaining high levels of E7 expression [20]. However, no evidence has been published related to the exact and defined transcript encoding the E6 or E7 proteins. In fact, a report by Stacey et al. [21] states that the HPV16 E7 protein is also translated from full-length E6/E7 mRNA structures, demonstrating that splicing is not required for E7 synthesis. Additionally, only the full-length E6 protein, not the spliced E6 variants, is found to efficiently bind to and promote the degradation of p53 [22], and it is further suggested that spliced transcripts of the HPV18 E6 gene may encode an E6-modified protein that inhibits the fulllength E6-mediated degradation of p53 [23–25]. Moreover, unspliced E6 mRNA is found to be more closely associated with tumorigenicity as compared with the spliced transcripts [26], and studies including cervical cancer samples show that the full-length transcript is always present, either alone or together with the spliced transcript [27; Molden et al., unpubl. data]. Taken together, these studies point to the full-length transcript as being the transcript which is important for the carcinogenic process.
Skomedal/Kraus/Silva/Molden/Hovland/Morland/Morland/Karlsen
88
PreTect HPV-Proofer
The PreTect HPV-Proofer assay (NorChip AS) is a commercially available kit for use in HPV diagnostics. Molecular markers directed for typing and detection of full-length E6/E7 mRNA from the 5 carcinogenic HPV types 16, 18, 31, 33 and 45 are included. In addition, the assay includes a marker for the human U1A housekeeping gene, used as mRNA performance control. Multiplex reactions are performed in three tubes. The PreTect HPV-Proofer assay utilizes the NASBA assay, a technique that amplifies RNA equivalents. By detecting mRNA, it is possible to monitor biological activities such as gene expression in a background of double-stranded DNA. Hence, contamination of DNA is not a problem. The NASBA reaction utilizes the activity of three enzymes, avian myeloblastosis virus reverse transcriptase, RNase H and T7 RNA polymerase, together with two primers to amplify single-stranded RNA at 41⬚C [28]. Realtime detection is performed for the amplified products by the use of molecular beacons [29], which are single-stranded oligonucleotide probes with a stem-loop structure that fluoresce only upon hybridization with their target. One arm of the stem is labeled with a fluorescent dye and the other with a nonfluorescent quencher. In the nonhybridized state, the fluorescent signal is ‘captured’ by the quencher and released as heat [30], but upon hybridization, the fluorescent dye and the quencher are separated and a fluorescent signal is transmitted. As RNA amplification and detection can be carried out in unopened vessels, it minimizes the risk of carryover contaminations. The PreTect HPV-Proofer assay has proved to be a highly stable and reproducible tool in monitoring HPV induced initiation and development of cervical cell dysplasias.
Coverage of Cervical Lesions by HPV Detection
Importance of Executing the Perfect Blend of HPV Target and Types The prevalence of HPV in cervical lesions has mostly been mapped by the use of DNA detection methods. However, unless oncogene expression of E6/E7 proteins is present, a given HPV may possibly not have contributed to the carcinogenic process. Moreover, optimizing the number of types included in a test panel, providing maximum sensitivity and specificity, is important for coverage of cervical lesions by HPV detection and has to be tailor made according to geographical region [31–33]. A difference in the spectrum of HPV types is observed when comparing low-grade and high-grade neoplasia and cancers [31–34; Kraus et al., accepted for publication]. This difference could be due to a diversity in the carcinogenic potential of the HPV types and also seen as presence of HPV ‘passengers’ or
mRNA Detection in Carcinogenesis
89
DNA 16, 18, 31, 33, 45 DNA 58, 52, 35, 59, 56, 51, 68, 39, 66 DNA 13 types RNA 16, 18, 31, 33, 45 (PreTect HPV-Proofer)
Prevalence of HPV (%)
100 90 80 70 60 50 40 30 20 10 0 Normal
LSIL
HSIL
SCC
Diagnosis
Fig. 2. Shift in spectrum of HPV types during carcinogenesis. RNA results are based on different studies representing different populations, as are the DNA results [8, 31–33, 34, 43, 44]. SCC ⫽ Squamous cell carcinoma.
nontransforming HPV infections that have not been involved in the carcinogenic process. Presence of passengers may cause a target competition that reduces the specificity and the overall sensitivity of the detection method, in particular of the L1, E2 or E1 based consensus and multiplex detection systems. Therefore, when the consensus PCR products are analyzed by typing methods, the HPV passengers may be the ones that are detected in addition to, or instead of, the cancer-causing HPV type, which again will give rise to incorrect test results. More than 15 HPV types have been identified in invasive cervical cancer samples by DNA-based detection methods. Schiffman et al. [35] have found that for both screening and triage, testing for more than about 10 HPV types decreased specificity more than it increased sensitivity. The minimal increases in sensitivity and in negative predictive value achieved by adding HPV types to DNA tests must be weighed against the projected burden to thousands of women falsely labeled as being at high risk of developing cervical cancer. Most probably, this number can be further reduced; in a recently performed study, maximum sensitivity and specificity were found when including only 5 HPV types. HPV DNA tests using consensus primers [36] and RNA tests were found to have almost similar accuracy for cervical intraepithelial neoplasia (CIN)2⫹. The specificity of the RNA-based method PreTect HPV-Proofer was found to be higher than the DNA-based method [Arbyn and Hovland, unpubl. data].
Skomedal/Kraus/Silva/Molden/Hovland/Morland/Morland/Karlsen
90
When comparing the 5 most prevalent HPV types 16, 18, 31, 33 and 45 to all 13 types found in squamous cervical carcinoma [31–33; Kraus and Molden et al., accepted for publication], a clear shift in prevalence can be seen during the carcinogenic process (fig. 2). HPV types found to have decreasing prevalence with increasing malignancy could have a lower potential of inducing fully malignant transformation or, preferentially, be cleared by the immune system. Especially in LSIL, expression of RNA is found in only a subset of HPV DNApositive cases, suggesting a lower carcinogenic potential for these HPV types. Therefore, it is important to consider whether a more limited number of HPV types (e.g., 16, 18, 31, 33, 45 and possibly 52 and 58 in Europe) actually causes the development of cervical cancer, as this will have implications for screening and vaccination programs.
Sample Collection and Preparation for HPV Nucleic Acid Detection
Sampling In HPV screening, it is important that the sample is collected in an optimal way. Different practices by gynecologist, physicians, nurses or midwives are observed, and different sample collection devices are used. Failure to include cells from the entire cervical area results in a great number of patients having to undergo a repeat investigation, or worse, that a number of samples will be incorrectly evaluated and a number of cases of early cell proliferation or cancer remain undetected. Furthermore, it is important that a sufficient amount of cell material is collected. When collecting cells for molecular diagnostic methods, subsequent use of assays that includes a performance control will ensure both that a sufficient number of cells have been collected and that the preparation of the sample has been conducted in a satisfactory way. Media for Fixation, Collection and Transportation For preservation of RNA, it is important to fix the cells immediately after collection. Methanol-based media like Preserv Cyt (Cytyc Corp., Boston, Mass., USA) and lysis buffers like NucliSens Lysis buffer (5M Untidiness Isocyanate, Bio Mérieux, Lyon, France) are commonly used. However, formaldehydebased media should be avoided due to dramatic reduction in amplification efficacy of RNA by NASBA (e.g., PreTect HPV-Proofer) [Skomedal et al., unpubl. results]. Sample Preparation Even though DNA is less prone to degradation than RNA, the success of obtaining material for amplification of RNA by the PreTect HPV-Proofer is
mRNA Detection in Carcinogenesis
91
high, ranging from 95–100%, depending on sampling, collection medium and extraction method. There are currently many commercially available methods for RNA and DNA extraction. Some of these methods are not suitable for routine diagnosis but may be used for research and low-throughput assays. Systems that up to now have been tested and found to be suited for extraction of highquality RNA are the RNAquous-Micro (Ambion, Austin, Tex., USA), M48 Biorobot MagAttract, RNeasy mini kit (Qiagen, Hilden, Germany), Perfect RNA (Eppendorf AG, Hamburg, Germany), RNA-spin (Intron Biotechnology, Seoul, South Korea), Versagene, RNA cell kit (Gentra, Minn., USA), Mini Mag, and NucliSens manual and automated (NucliSens, Bio Mérieux, Marcy l’Etoile, France) systems.
Accuracy of HPV mRNA Detection
Performance of the mRNA-Based Assay PreTect HPV-Proofer In a screening program, a method intended for use as an adjunct to cytology has to demonstrate high clinical accuracy. Accuracy can be defined as the degree of conformity of a measured quantity as compared with a gold standard. The gold standard for diagnosis of precancer disease has mainly been histopathology, with the precancer diagnosis being CIN2⫹. Furthermore, persistent HPV infection has proven to be important in predicting cell abnormalities. Hence, a supplementary method should, in addition to correlation with the histopathological diagnosis, reveal HPV persistence and preferentially give additional information about the outcome of the disease. Accuracy includes optimal analytical sensitivity and specificity, possessing a high detection rate of final endpoint disease. Moreover, it is important to reveal underlying abnormalities in cytologically and colposcopically normal samples. Both first-time cytology and colposcopy may miss dyskaryosis or neoplasia in more than 50% of the samples [37]. Therefore, in order to create more reliable statistics, a collection of representative punch biopsies could have been included in more studies and clinical trials independent of the cytological or colposcopic results. To evaluate the diagnostic value of the PreTect HPV-Proofer, more than 50,000 clinical samples have been tested, and several screening studies are currently under progress. The accuracy of the assay has been calculated based on different studies, listed in tables 2 and 3. Accuracy of Molecular Tests as Compared with Conventional Screening Methods In developing countries, it seems unrealistic to introduce cytological screening and histopathological follow-up because of financial, technical and
Skomedal/Kraus/Silva/Molden/Hovland/Morland/Morland/Karlsen
92
Table 2. A summary of six clinical studies performed on outpatient samples mRNA Detection in Carcinogenesis
Publication
Total number
Index HSIL Molden et al. [43], 2005 Index ASCUS and LSIL Molden et al. [44], 2005 Implications for Cervical Disease Progression. Cuschieri et al. [45], 2004 Cross-sectional study Lie et al. [8], and abstract 21th IPC Conference in Mexico, 2005 Cross-sectional study with histology on all Arbyn and Hovland, IPH/EPI Reports2 Cross-section study in Germany, Steinberg et al., 2006
N⫽4136 HSIL⫽ 25 N⫽4136 LSIL and ASCUS ⫽ 73 N⫽3444 54 HPV positive cases N ⫽ 628
Number of CIN2⫹ 14 7
4
281 102
N⫽342
16
N ⫽ 151 N ⫽ 150
24 23
Methods
Test prevalence (%)
Consensus PCR HPV-Proofer Consensus PCR HPV-Proofer
10 3 10 3
Consensus PCR HPV-Proofer
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
93 86 86 86
67 93 50 85
4 10 15 38
100 100 97 98
63 30
100 100
401 761
12 25
100 100
hc2 ⱖ 30 HPV-Proofer ⱖ 30 hcII ⬍ 30 HPV-Proofer ⬍ 30 Consensus-PCR HPV-Proofer
NV NV NV NV 30 7
93 76 98 82 100 81
40 81 20 70 73 96
NV NV NV NV 16 52
NV NV NV NV 100 99
hc2 HPV-Proofer
38 23
100 96
73 90
41 63
100 99
93
NV ⫽ Not valid; NPV ⫽ negative predictive value; consensus PCR ⫽ Gp5⫹/6⫹ primers, except by Cuschieri et al. [45] who use My09/11 primers; PPV ⫽ positive predictive value. 1 According to histological CIN2⫹. 2 Scientific Institute of Public Health, Department of Epidemiology, Brussels, January 7, 2006.
Table 3. HPV detection rate by HPV-Proofer and different HPV DNA methods in ASCUS and LSIL samples Study
Total number
Prevalence HPV-Proofer ASCUS (%)
Prevalence DNA ASCUS (%)
Prevalence Prevalence HPV-Proofer DNA LSIL LSIL (%) (%)
Index ASCUS and LSIL Molden et al. [44], 2005 Routine testing of 19, 152 women Haaheim et al., Vancouver, 2005 Routine testing of 760 women TDL, London, 2005 Cross-sectional study with histology on all Arbyn and Hovland, IPH/EPI Reports Cross-section study in Germany, Steinberg et al., 2006 German Multisenter Study Moeckel et al., Vancouver, 2005 Follow up of low-grade lesions Dillner et al., Mexico, 2004
ASCUS ⫽ 54 LSIL⫽ 19 ASCUS ⫽ 491 LSIL ⫽ 735 ASCUS ⫽ 55 LSIL⫽ 52 ASCUS ⫽ 4 LSIL ⫽ 9
20
46
26
74
26
NA
29
NA
15
51
23
85
25
50
22
89
ASCUS ⫽ 12 0a LSIL⫽ 25 ASCUS ⫽ 12 8a LSIL ⫽ 75 ASCUS ⫽ 234 20 LSIL ⫽74
42a
32b
84b
25a
37b
72b
49
28
74
Average
ASCUS ⫽ 862 23 LSIL ⫽ 989
47
29
77
NA = Not assessed. According to German definition, Pap IIw is an intermediate diagnosis between cytologically normal and ASCUS. b According to German definition, Pap IIId is an intermediate diagnosis between LSIL and moderate dyskaryosis. a
human constraints. Alternative screening tests are needed which accurately predict the presence of cervical cancer precursors. Therefore, a cross-sectional survey was set up in the city of Bukavu, situated in the east of the Democratic Republic of Congo [Arbyn and Hovland, IPH/EPI Scientific Institute of Public Health, Department of Epidemiology, Brussels, January 7, 2006]. Due to expectancy of a high frequency of disease and difficulties in recalling each patient for confirmation of diagnosis and subsequent treatment, it was decided to obtain a histological sample from all patients at the initial visit. This study is also an important contribution to the assessment of test accuracy due to the fact that using a standard evaluation protocol for evaluating test systems [38] is extremely rare. Three hundred and forty-three women between 25 and 60 years of age (median 37 years) participated in the study after having obtained informed consent. During a gynecological examination, cervical cellular material and punch
Skomedal/Kraus/Silva/Molden/Hovland/Morland/Morland/Karlsen
94
biopsies were collected from the squamocolumnar junction of the cervix and stored in a methanol-based medium (PreservCyt; Cytyc). The importance of obtaining histological material from either punch biopsies or with the use of a endocervical curettage was given by the high number of colposcopic positive women and women without visible squamocolumnar junction (in total 52%). A colposcopic examination was performed by at least two clinicians before and after application of 5% acetic acid. Pap smears and vials with cellular and histological material were shipped in PreservCyt medium to Norway, Holland or Sweden for cytological, histological and virological examination by wellexperienced pathologists. Presence of HPV DNA was assessed by PCR using GP5⫹/6⫹ consensus primers, and the amplified viral DNA was identified using cocktails of oligonucleotides hybridizing with 14 high-risk and a number of lowrisk HPV types, respectively [36]. The PreTect HPV-Proofer assay was used to identify presence of transcripts of the viral E6/E7 gene of HPV 16, 18, 31, 33 or 45. All the molecular assays showed significantly higher sensitivity for highgrade CIN than conventional cytological examination or liquid-based cytology. The sensitivity of the PreTect HPV-Proofer for CIN3⫹ was 30–75% higher than for cytology (88% compared with 50% for liquid-based and 67% for conventional cytology at cutoff HSIL⫹). The differences for the HPV-Proofer (87%) were still high at cutoff ASCUS⫹ and was calculated to be 20–25% higher than both cytology methods (63% for liquid-based and 67% for conventional cytology). The specificity of the PreTect HPV-Proofer for CIN3⫹ was similar at the cytological diagnosis ASCUS⫹, around 95%. The detection prevalence was also similar between conventional and liquid-based cytology (cutoff ASCUS⫹) and for the PreTect HPV-Proofer, around 7–8%. The positive predictive value with CIN3⫹ outcome of cytological ASCUS⫹ (Pap 21% and liquid-based cytology 24%) was lower than for the PreTect HPV-Proofer (28%). The positive predictive value with CIN3⫹ outcome for the consensus PCR method was approximately 8% and similar to what has been discovered before [39]. Therefore, in many aspects, the PreTect HPV-Proofer showed similar accuracy to both conventional and liquid-based cytology but had higher odds ratio (3–4 times higher than ASCUS⫹) and significantly higher clinical sensitivity. Detection Accuracy as Compared with Endpoint CIN3 and Cancer To evaluate the presence of HPV E6/E7 mRNA in clinical samples, several studies have been carried out. In one study, 204 cervical cancer cases were screened by the PreTect HPV-Proofer [Kraus and Molden et al., accepted for publication], and to assess the overall prevalence of various HPV types, samples were also tested for HPV DNA by a range of previously published techniques. The overall prevalence of HPV was 97%. For the HPV-negative cases, cervical cancer may have been caused by a different mechanism, an HPV type
mRNA Detection in Carcinogenesis
95
not detected by the methods used or the HPV causing the cervical carcinogenesis may have been lost. This may also be an explanation for samples where only the ‘low-risk’ types 6 or 26 were detected; some cancerous samples may have HPV types presented only as passengers. The 5 HPV types included in the PreTect HPV-Proofer were detected in as much as 89% of the cases. This confirms that these 5 types are the most common types found in cervical cancers. In order to evaluate HPV oncogenic activity in different grades of dysplasia, a study on 190 cervical biopsies was performed [34]. The analysis was done on a biopsy taken from the ectocervix/transformation zone at the time of conization, and the pathological diagnosis of this particular biopsy was not always representative for the existing lesion which was later revealed in the cone. This strategy gave access to a uniformly preserved material with varying degrees of squamous cell dysplasia. All but two PCR-positive cases also showed E6/E7 transcription, confirming that, at the level of severe dysplasia, HPV exerts an oncogenic activity [40]. Lack of high-risk HPV oncogenic expression in CIN1 cases suggests that histologically defined CIN1 may not be strictly related to cancer precursors [41]. However, further studies are needed in order to confirm this assumption. Moreover, current management protocols may in certain cases lead to overtreatment [42]. This problem is emphasized in this same study and implies that unnecessary conizations can possibly be reduced by introducing HPV testing into the routine assessment. Added Value of Molecular Methods in Triage A cross-sectional study with 2–3 years follow-up of 4,136 women older than 30 years has been carried out, including detection of HPV DNA with GP5⫹/6⫹ consensus PCR and detection of E6/E7 mRNA by the PreTect HPVProofer. Cytological and histological data from the Norwegian Cancer Registry were included, presenting the follow-up diagnosis of women with a HSIL [43], ASCUS/LSIL [44] and normal [Molden et al., unpubl. data] index Pap smear. Of 25 HSIL cases, the histological results at 2-year follow-up showed that the PreTect HPV-Proofer missed 2 CIN3 cases, while the consensus PCR method missed 1 case. On the other hand, with CIN2⫹ as endpoint, the PreTect HPVProofer had a specificity of 89% as opposed to PCR, having a specificity of 67%. At the follow-up of ASCUS and LSIL (77 cases), the sensitivity for CIN2⫹ was equal between PCR and PreTect HPV-Proofer. However, the specificity was 85% for the PreTect HPV-Proofer, as compared with 50% for PCR. Further, women with normal Pap smear and a positive HPV test had an increased risk of being diagnosed with CIN2⫹ within 3 years. In conclusion, HPV testing by the PreTect HPV-Proofer shows a higher predictive clinical value than HPV DNA testing which can be confirmed by histology.
Skomedal/Kraus/Silva/Molden/Hovland/Morland/Morland/Karlsen
96
Based on several evaluation studies of the PreTect HPV-Proofer assay, the detection of E6/E7 mRNA by the PreTect HPV-Proofer in ASCUS and LSIL has been shown to be below 30% (table 2). In Norway, two population-based studies have been performed on ASCUS giving a test prevalence of 20 and 26%, and on LSIL, with a test prevalence of 26 and 29% [44; unpubl. data]. In the UK, the Doctors Laboratory, London, observed the detection rate of the PreTect HPV-Proofer in borderline or mild dyskaryosis samples to be 15 and 23%, respectively [unpubl. data]. To conclude, the added value of using a combination of the PreTect HPV-Proofer and cytology is shown by the low detection rate combined with a high specificity for CIN2⫹ in ASCUS and LSIL [44]. Thus, the number of samples having to be followed up after first-time cytological examination can be dramatically reduced compared with using hc2 or consensus PCR. One of the current challenges with cytology is the number of false-positive cases. In a follow-up study of ASCUS and LSIL cases (n ⫽ 240), it was observed that all cases (n ⫽ 18) that were histologically normal in loop electrosurgical excision procedure biopsies were also negative for HPV E6/E7 by the PreTect HPV-Proofer. However, 73% of these histologically normal cases were cytologically positive and hc2 positive [Bjerre et al., unpubl. data]. This study again confirms that the PreTect HPV-Proofer has a potential of giving high clinical specificity. Persistent Transforming Infection Cuschieri et al. [45] have carried out a follow-up study on 54 HPV DNApositive samples obtained from 3,444 cytologically normal women. Samples positive by the PreTect HPV-Proofer, consistently over 9 months (persistence), were proven to have CIN3 in most cases. This shows that identification of mRNA expression correlates with the continuing presence of HPV DNA. The study also showed that the PreTect HPV-Proofer had a significantly higher specificity (81%, compared with cytology) than the DNA-based PCR method (44%); both methods detected all CIN2⫹ cases (n ⫽ 4). To conclude, the PCR consensus method used would have three times as many ‘false positives’ to be followed up than the PreTect HPV-Proofer method: PreTect HPV-Proofer find HPV infections that will most likely persist, and therefore, will reduce the need for follow-up or repeat tests, something that will be necessary if DNA-based technology is used. Comparative Studies between HPV DNA and HPV mRNA-Based Methods The Norwegian Radium Hospital has carried out a study including 383 women with a median age of 35 years (range 19–85) [8]. Colposcopically directed biopsies were taken from all women with high-grade cytology
mRNA Detection in Carcinogenesis
97
(CIN2⫹) or hc2/PreTect HPV-Proofer positive. Histological CIN2 or CIN3 were discovered in 47% of the samples being cytologically normal and HPV positive. For benign and preinvasive lesions, the PreTect HPV-Proofer gave significantly less positive results as compared with the hc2 test. Among samples positive by hc2, some were found to be false positive due to crossreactivity with unrelated HPV types. The specificity for HSIL⫹ in this study was calculated as 20% for hc2 and 70% for the PreTect HPV-Proofer for the population below 30 years of age and as 40% for hc2 and 81% for the PreTect HPV-Proofer for the population above 30 years of age [Lie et al., unpubl. data] (table 2). All of the 20 biopsy-confirmed invasive cancers included in this study were positive by the PreTect HPV-Proofer, while 90% were positive by hc2. The main reason that the PreTect HPV-Proofer was positive in all the tested cancer cases may be a high analytical sensitivity and location of target sequence not affected by integration. The hc2 assay detects the L1 region which may be deleted by integration of the viral genome into the host cell. Approximately 50% of those that had normal histological tissue samples were HPV positive using hc2. Taken together, this shows that the PreTect HPV-Proofer is better suited for risk evaluation compared with hc2. Moreover, most DNA-based methods show a specificity of less than 80% but are yet currently being used for diagnostic purposes. For diagnostic screening, this will create several false positives, providing a test prevalence not comparable with the actual disease prevalence. The test prevalence of hc2 and consensus PCR has been shown in many studies to be between 6 and 62%, dependent on the age and risk level of the selected population [46–48]. Comparative studies are the only way that the true relative prevalence of different methods is determined. In two different studies including an outpatient population, the test prevalence of the consensus PCR has been shown to be 3–4 times higher than that by the PreTect HPV-Proofer [43; unpubl. results]. This would affect 3–4 times more women than necessary.
Conclusions
The main cause of invasive cervical cancer is the production of full-length E6 and E7 proteins following the production of stabilized E6/E7 full-length mRNA from high-risk HPV types. Although the inclusion of screening programs has contributed to reducing the number of cancer cases, combining cytology with the testing for HPV oncogene activity would further improve the screening program. The PreTect HPV-Proofer is based on the detection of oncogene expression in the form of mRNA from the 5 most common high-risk HPV types 16, 18, 31, 33 and 45. The HPV types included in the PreTect
Skomedal/Kraus/Silva/Molden/Hovland/Morland/Morland/Karlsen
98
HPV-Proofer have a coverage rate of 96–98% of cancer cases in Europe and the US. The consequences of increasing the number of HPV types to more than 5 types are unchanged clinical sensitivity, increased number of false positives and reduced clinical specificity. It has been shown that the PreTect HPV-Proofer may identify less than 2.5% of a normal screening population, presenting high potential in reduction in the number of false-positive cases compared with cytology and HPV DNA testing. Moreover, more than 15 extensive studies of HPV mRNA expression using this method have already been carried out in many countries, and the conclusion is that the combination of cytology and the PreTect HPV-Proofer should give optimal sensitivity and specificity. In a triage setting, it has been demonstrated that the PreTect HPV-Proofer may reduce the number of ASCUS and LSIL cases that have to be followed up by colposcopydirected biopsies by more than 70% (table 3). The HPV E6 and E7 oncogene products are the transforming factors – a selectivity factor which is built into the PreTect HPV-Proofer assay for detection of transformation events only. The 5 types included in the assay were detected in as much as 89% of biopsies with squamous cell carcinoma. Moreover, when cytological smears from women with cervical carcinoma were tested, the PreTect HPV-Proofer yielded similar or higher sensitivity as compared with DNA-based testing. This, together with the finding that there is a lack of E6 and E7 mRNA expression in most cytological normal cases, histological CIN1 and histological normal loop electrosurgical excision procedure biopsies, demonstrates the strength of the PreTect HPV-Proofer as a triage method and as a future screening method.
References 1 2 3 4
5
6
7
Fahey MT, Irwig L, Macaskill P: Meta-analysis of Pap test accuracy. Am J Epidemiol 1995;141: 680–689. Cuzick J, Beverley E, Ho L, Terry G, Sapper H, Mielzynska I, Lorincz A, Chan WK, Krausz T, Soutter P: HPV testing in primary screening of older women. Br J Cancer 1999;81:554–558. Ratnam S, Franco EL, Ferenczy A: Human papillomavirus testing for primary screening of cervical cancer precursors. Cancer Epidemiol Biomarkers Prev 2000;9:945–951. Ngelangel CA, Limson GM, Cordero CP, Abelardo AD, Avila JM, Festin MR: Acetic-acid guided visual inspection vs cytology-based screening for cervical cancer in the Philippines. Int J Gynaecol Obstet 2003;83:141–150. Rodriguez-Reyes ER, Cerda-Flores RM, Quinonez-Perez JM, Velasco-Rodriguez V, CortesGutierrez EI: Acetic acid test: a promising screening test for early detection of cervical cancer. Anal Quant Cytol Histol 2002;24:134–136. Cronje HS, Parham GP, Cooreman BF, de Beer A, Divall P, Bam RH: A comparison of four screening methods for cervical neoplasia in a developing country. Am J Obstet Gynecol 2003;188: 395–400. Denny L, Kuhn L, Pollack A, Wright TC Jr: Direct visual inspection for cervical cancer screening: an analysis of factors influencing test performance. Cancer 2002;94:1699–1707.
mRNA Detection in Carcinogenesis
99
8
9 10 11
12
13
14 15 16 17
18 19
20 21
22
23 24 25 26
27
28
Lie AK, Risberg B, Borge B, Sandstad B, Delabie J, Rimala R, Onsrud M, Thoresen S: DNAversus RNA-based methods for human papillomavirus detection in cervical neoplasia. Gynecol Oncol 2005;97:908–915. Gregoire L, Arella M, Campione-Piccardo J, Lancaster WD: Amplification of human papillomavirus DNA sequences by using conserved primers. J Clin Microbiol 1989;27:2660–2665. Gravitt PE, Manos MM: Polymerase chain reaction-based methods for the detection of human papillomavirus DNA. IARC Sci Publ 1992;121–133. Snijders PJ, Schulten EA, Mullink H, ten Kate RW, Jiwa M, van der Waal I, Meijer CJ, Walboomers JM: Detection of human papillomavirus and Epstein-Barr virus DNA sequences in oral mucosa of HIV-infected patients by the polymerase chain reaction. Am J Pathol 1990;137: 659–666. Chan PK, Cheung TH, Tam AO, Lo KW, Yim SF, Yu MM, To KF, Wong YF, Cheung JL, Chan DP, Hui M, Ip M: Biases in human papillomavirus genotype prevalence assessment associated with commonly used consensus primers. Int J Cancer 2006;118:243–245. Branca M, Giorgi C, Santini D, Di Bonito L, Ciotti M, Costa S, Benedetto A, Casolati EA, Favalli C, Paba P, Di Bonito P, Mariani L, Syrjänen S, Bonifacio D, Accardi L, Zanconati F, Syrjänen K: Survivin as a marker of cervical intraepithelial neoplasia and high-risk human papillomavirus and a predictor of virus clearance and prognosis in cervical cancer. Am J Clin Pathol 2005;124:113–121. Gulliksen A, Solli LA, Drese KS, Sorensen O, Karlsen F, Rogne H, Hovig E, Sirevag R: Parallel nanoliter detection of cancer markers using polymer microchips. Lab Chip 2005;5:416–420. Ozbun MA, Meyers C: Characterization of late gene transcripts expressed during vegetative replication of human papillomavirus type 31b. J Virol 1997;71:5161–5172. zur Hausen H: Papillomaviruses causing cancer: evasion from host-cell control in early events in carcinogenesis. J Natl Cancer Inst 2000;92:690–698. Ronco LV, Karpova AY, Vidal M, Howley PM: Human papillomavirus 16 E6 oncoprotein binds to interferon regulatory factor-3 and inhibits its transcriptional activity. Genes Dev 1998;12: 2061–2072. Um SJ, Rhyu JW, Kim EJ, Jeon KC, Hwang ES, Park JS: Abrogation of IRF-1 response by highrisk HPV E7 protein in vivo. Cancer Lett 2002;179:205–212. Iftner T, Oft M, Bohm S, Wilczynski SP, Pfister H: Transcription of the E6 and E7 genes of human papillomavirus type 6 in anogenital condylomata is restricted to undifferentiated cell layers of the epithelium. J Virol 1992;66:4639–4646. Smotkin D, Prokoph H, Wettstein FO: Oncogenic and nononcogenic human genital papillomaviruses generate the E7 mRNA by different mechanisms. J Virol 1989;63:1441–1447. Stacey SN, Jordan D, Snijders PJ, Mackett M, Walboomers JM, Arrand JR: Translation of the human papillomavirus type 16 E7 oncoprotein from bicistronic mRNA is independent of splicing events within the E6 open reading frame. J Virol 1995;69:7023–7031. Shally M, Alloul N, Jackman A, Muller M, Gissmann L, Sherman L: The E6 variant proteins E6IE6IV of human papillomavirus 16: expression in cell free systems and bacteria and study of their interaction with p53. Virus Res 1996;42:81–96. Guccione E, Pim D, Banks L: HPV-18 E6*I modulates HPV-18 full-length E6 functions in a cell cycle dependent manner. Int J Cancer 2004;110:928–933. Pim D, Massimi P, Banks L: Alternatively spliced HPV-18 E6* protein inhibits E6 mediated degradation of p53 and suppresses transformed cell growth. Oncogene 1997;15:257–264. Pim D, Banks L: HPV-18 E6*I protein modulates the E6-directed degradation of p53 by binding to full-length HPV-18 E6. Oncogene 1999;18:7403–7408. Inoue T, Kyo S, Kiyono T, Ishibashi M, Ishiwatari H, Hwang YI, Yutsudo M, Hakura A: Correlation between tumorigenicity and expression levels or splicing patterns of transcripts of the human papillomavirus type 16 E6 gene. Jpn J Cancer Res 1994;85:357–363. Rose BR, Thompson CH, Tattersall MH, Elliott PM, Dalrymple C, Cossart YE: Identification of E6/E7 transcription patterns in HPV 16-positive cervical cancers using the reverse transcription/polymerase chain reaction. Gynecol Oncol 1995;56:239–244. Kievits T, van Gemen B, van Strijp D, Schukkink R, Dircks M, Adriaanse H, Malek L, Sooknanan R, Lens P: NASBA isothermal enzymatic in vitro nucleic acid amplification optimized for the diagnosis of HIV-1 infection. J Virol Methods 1991;35:273–286.
Skomedal/Kraus/Silva/Molden/Hovland/Morland/Morland/Karlsen
100
29 30 31 32
33
34
35
36
37 38
39
40
41
42
43
44
45
46
Tyagi S, Kramer FR: Molecular beacons: probes that fluoresce upon hybridization. Nat Biotechnol 1996;14:303–308. Stryer L: Fluorescence energy transfer as a spectroscopic ruler. Annu Rev Biochem 1978;47:819–846. Clifford GM, Smith JS, Plummer M, Munoz N, Franceschi S: Human papillomavirus types in invasive cervical cancer worldwide: a meta-analysis. Br J Cancer 2003;88:63–73. Clifford GM, Rana RK, Franceschi S, Smith JS, Gough G, Pimenta JM: Human papillomavirus genotype distribution in low-grade cervical lesions: comparison by geographic region and with cervical cancer. Cancer Epidemiol Biomarkers Prev 2005;14:1157–1164. Clifford GM, Gallus S, Herrero R, Munoz N, Snijders PJ, Vaccarella S, Anh PT, Ferreccio C, Hieu NT, Matos E, Molano M, Rajkumar R, Ronco G, de Sanjose S, Shin HR, Sukvirach S, Thomas JO, Tunsakul S, Meijer CJ, Franceschi S: Worldwide distribution of human papillomavirus types in cytologically normal women in the International Agency for Research on Cancer HPV prevalence surveys: a pooled analysis. Lancet 2005;366:991–998. Kraus I, Molden T, Erno LE, Skomedal H, Karlsen F, Hagmar B: Human papillomavirus oncogenic expression in the dysplastic portio; an investigation of biopsies from 190 cervical cones. Br J Cancer 2004;90:1407–1413. Schiffman M, Khan MJ, Solomon D, Herrero R, Wacholder S, Hildesheim A, Rodriguez AC, Bratti MC, Wheeler CM, Burk RD: A study of the impact of adding HPV types to cervical cancer screening and triage tests. J Natl Cancer Inst 2005;97:147–150. van den Brule AJ, Pol R, Fransen-Daalmeijer N, Schouls LM, Meijer CJ, Snijders PJ: GP5⫹/6⫹ PCR followed by reverse line blot analysis enables rapid and high-throughput identification of human papillomavirus genotypes. J Clin Microbiol 2002;40:779–787. Massad LS, Collins YC, Meyer PM: Biopsy correlates of abnormal cervical cytology classified using the Bethesda system. Gynecol Oncol 2001;82:516–522. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, Lijmer JG, Moher D, Rennie D, de Vet HC: Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative. BMJ 2003;326:41–44. Salmeron J, Lazcano-Ponce E, Lorincz A, Hernandez M, Hernandez P, Leyva A, Uribe M, Manzanares H, Antunez A, Carmona E, Ronnett BM, Sherman ME, Bishai D, Ferris D, Flores Y, Yunes E, Shah KV: Comparison of HPV-based assays with Papanicolaou smears for cervical cancer screening in Morelos State, Mexico. Cancer Causes Control 2003;14:505–512. Nakagawa S, Yoshikawa H, Yasugi T, Kimura M, Kawana K, Matsumoto K, Yamada M, Onda T, Taketani Y: Ubiquitous presence of E6 and E7 transcripts in human papillomavirus-positive cervical carcinomas regardless of its type. J Med Virol 2000;62:251–258. Crum CP, Genest DR, Krane JF, Hogan C, Sun D, Bellerose B, Kostopoulou E, Lee KR: Subclassifying atypical squamous cells in Thin-Prep cervical cytology correlates with detection of high-risk human papillomavirus DNA. Am J Clin Pathol 1999;112:384–390. Meijer CJ, Helmerhorst TJ, Rozendaal L, van der Linden JC, Voorhorst FJ, Walboomers JM: HPV typing and testing in gynaecological pathology: has the time come? Histopathology 1998;33: 83–86. Molden T, Kraus I, Karlsen F, Skomedal H, Nygard JF, Hagmar B: Comparison of human papillomavirus messenger RNA and DNA detection: a cross-sectional study of 4,136 women ⬎30 years of age with a 2-year follow-up of high-grade squamous intraepithelial lesion. Cancer Epidemiol Biomarkers Prev 2005;14:367–372. Molden T, Nygard JF, Kraus I, Karlsen F, Nygard M, Skare GB, Skomedal H, Thoresen SO, Hagmar B: Predicting CIN2⫹ when detecting HPV mRNA and DNA by PreTect HPV-proofer and consensus PCR: a 2-year follow-up of women with ASCUS or LSIL Pap smear. Int J Cancer 2005;114:973–976. Cuschieri KS, Whitley MJ, Cubie HA: Human papillomavirus type specific DNA and RNA persistence – implications for cervical disease progression and monitoring. J Med Virol 2004;73: 65–70. Belinson JL, Qiao YL, Pretorius RG, Zhang WH, Rong SD, Huang MN, Zhao FH, Wu LY, Ren SD, Huang RD, Washington MF, Pan QJ, Li L, Fife D: Shanxi Province cervical cancer screening study II: self-sampling for high-risk human papillomavirus compared to direct sampling for human papillomavirus and liquid based cervical cytology. Int J Gynecol Cancer 2003;13:819–826.
mRNA Detection in Carcinogenesis
101
47
48
Petry KU, Menton S, Menton M, Loenen-Frosch F, de Carvalho Gomes H, Holz B, Schopp B, Garbrecht-Buettner S, Davies P, Boehmer G, van den Akker E, Iftner T: Inclusion of HPV testing in routine cervical cancer screening for women above 29 years in Germany: results for 8466 patients. Br J Cancer 2003;88:1570–1577. Winer RL, Kiviat NB, Hughes JP, Adam DE, Lee SK, Kuypers JM, Koutsky LA: Development and duration of human papillomavirus lesions, after initial infection. J Infect Dis 2005;191:731–738.
Dr. Hanne Skomedal, PhD NorChip AS, Industriveien 8 NO–3490 Klokkarstua (Norway) Tel. ⫹47 32798806, Fax ⫹47 32798801, E-Mail
[email protected] Skomedal/Kraus/Silva/Molden/Hovland/Morland/Morland/Karlsen
102
HPV Testing and Patient Management Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 103–119
Human Papillomavirus Testing for Primary Cervical Cancer Screening Véronique Dalsteina, Jean-Paul Boryb, Olivier Graesslinb, Christian Quereuxb, Philippe Birembauta, Christine Clavela a
Laboratory Pol Bouin, and bDepartment of Obstetrics and Gynecology, CHU de Reims, Reims, France
Cervical cancer is an infectious disease and it is now well established that oncogenic, or high-risk, HPVs (HR-HPV) are the main and necessary causal factor in the development of cervical intraepithelial and invasive neoplasia. A limited spectrum of 13–18 HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68, 73 and 82 and probably 26, 53 and 66 are reported from various studies to be linked with cervical cancer and its precursors [1–3]. HPV infections are the most common sexually transmitted viral infection at present with latent or clinical infections including single HPV infection or HPV co-infections. In general, latent genital HPV infection can be detected in 5–40% of sexually active women of reproductive age and, in most cases, genital HPV infection is transient or intermittent. High-risk HPV (HR-HPV) DNA prevalence is between 4 and 20% [1] but is age-dependent and gradually decreases with age and specially in normal smears [2, 4, 5]. The highest peak of prevalence is present at the third decade (24–30%) [6–9]. 10% of young women remain infected at 5 years and, fortunately, the majority of these HPV infections are transient [10, 11]. For women 30 years, there is a decreased prevalence of 5–10% around 45–50 years [12]. Some studies showed a second peak at older ages which is not clear [13, 14] and other cohort studies are needed. In total, only persistent HR-HPV infections are at risk and the disease prevalence is rare in developed countries. However, this cancer still kills a large number of women in developing countries (80% of cervical cancer cases) with undetected and untreated cervical high-grade lesions. Results from numerous epidemiological and screening studies lead to new approaches of triage, screening and prevention of preinvasive and invasive cervical cancer, and also follow-up of women
Both negative
Proposed combined screening algorithm
Proposed new screening algorithm
Use of HPV test as an adjunct to cytology (adapted with permission from Wright et al. [45])
Use of HPV alone as the primary test (adapted with permission from Cuzick [12], Copyright 2002 Elsevier)
Cyto and HPV
Women aged 25–64 years HPV test
ASC-US HPV ()
Cyto () HPV ()
Repeat Cyto and HPV at 3 years
at 12 months
ASC-US HPV ()
Cyto LSIL
Negative
Positive
Routine 5 year recall
Cytology
Colposcopy
Normal or borderline
Mild
at 6–12 months HPV and Cyto at 6–12 months
Any cyto HPV () Colposcopy
Cyto-negative HPV-negative
Routine 5 year recall
Colposcopy
Cyto Mild HPV-positive and Cyto Mild HPV-negative and Cyto borderline
Colposcopy
HPV and Cyto at 6–12 months
Fig. 1. Optional algorithms for screening.
post-treatment for precancerous and cancerous lesions. We discuss the current concepts on primary screening by HPV testing, by summarizing the experience from large international studies and statements of international agencies (‘interim guidance’) taking position on the potential place of HPV testing in primary cervical screening (fig. 1). Programs based only on cytology screening have been effective but have limitations with mostly sampling problems: reproducibility is poor, interpretation of results is subjective and there is a high probability of misclassification [12]. Even if we know that some women never received any Pap test or did not receive it often enough [15], an audit of the UK program found that 47% of the fully invasive cancers in women under 70 years occurred in individuals with an apparently adequate screening history [12]. Then, most cases of cervical cancer occur because of a false-negative result of a Pap test (15–45% in detecting cancer precursors) and false-negative diagnoses have important medical, financial and legal implications [16–18]. To minimize false-negative cytology results, there is a need to improve the quality of sampling, slide processing and diagnostic performance [12, 15, 19]. Some programs propose to repeat the Pap test every year to balance this relative limited sensitivity of conventional cytology, but compromising its cost efficacy and the possibility to use an algorithm with greater intervals and similar safety [19]. Thus, liquid-based cytology
Dalstein/Bory/Graesslin/Quereux/Birembaut/Clavel
104
(LBC) was recently developed, allowing an automated preparation of thin-layer slides, and an automated computer-assisted lecture of the smears improved. LBC has a greater sensitivity, about 10% more than conventional cytology. In addition, the LBC technique affords the possibility to perform other molecular diagnostic procedures on cervical cells collected in a transport medium. On the other hand, HR-HPV being recognized as the central causal agent in cervical cancer, HPV testing has been proposed for years as a potential adjunctive screening tool, based on the detection of the viral DNA. In developed countries, sensitive, reproducible and standardized HPV DNA assays were developed and largely improved for clinical use since 1995. Basically, in the published literature, two main reproducible and sensitive techniques are accepted and recommended: liquid-based hybridization (Hybrid Capture 2®, Digene) and PCR-based assays, both largely used in the literature with possible automation in a large scale. Transport medium, from LBC or others, have to be validated for molecular biology. HC2 (available since 1997) can detect one or more of 13 HR-HPV (coverage of 95% of HR-HPV), at a level of 1 pg/ml (5,000 genomes per test well), showing a high sensitivity 90% to detect HSIL lesions, close to the sensitivity of PCR. Thus, in 2003 the FDA (Food and Drug Administration) only approved HC2® for primary screening used with a Pap test to screen women over 30 years for the presence of HR-HPV infection. For PCR tests, the most common PCR techniques use various consensus primers in the L1 region to detect a large spectrum of HPV with possible genotyping, based on the MY09/11 primers (the Roche tests based on PGMY09/11: Amplicor® detects one or more of 13 HR-HPV types and Linear Array® is used for genotyping), GP5/6 primers (future Hybrid Capture 4 for genotyping) and SPF10 primers (InnoLipa®, Innogenetics, also to genotype) [20–22].
Primary Screening
HPV Background The presence and persistence or the absence of the HR-HPV infection define new possible approaches to screen and prevent cervical cancer. Table 1 presents key data from 14 relevant recent primary screening studies, 7 in Europe, 2 in North America, 1 in Asia, 1 in Africa, 1 in Central America and 2 in South America [9, 23–35]. These studies evaluate the performance of cytology (conventional or LBC) and HPV testing (HC2 or PCR) in detecting CIN2, CIN3 or cervical cancer in several populations. Despite various designs of the studies, various age ranges, various HPV detection techniques, various cytological methods and classifications, various HR-HPV positivity, cytological abnormalities and precancerous lesions prevalence rates, the global survey of the results leads to 3 main
HPV Primary Screening
105
Table 1. Key data from primary screening studies comparing cytology and HPV testing for the detection of histologically proven highgrade cervical lesions (CIN2) Dalstein/Bory/Graesslin/Quereux/Birembaut/Clavel
Author, year
Country
n
Age
106
HRCIN2 HPV- % positive
Sensitivity
Schneider, 2000
Germany, East Thuringia
4,761 18–70 bias median adjusted; 35 crosssectional 8 months FU
Conv.a GP5/6 0.9* PCR
7.8*
2.4
20
89.4 NS
99
93.9 NS
NS
Clavel, 2001
France, Reims
7,932 15–76 crossmedian sectional 34 with FU; HPV not a criterion for immediate colposcopy
Conv. HC2 LBC
8.1*
15.3
1.45*
68 88
100 100
95 93
86
38
100
Cuzick, 2003
UK, national program
10,358 30–60 mean 42
5% random colposcopy in cyto/HPVwomen; 12 months FU
Conv. HC2
2.1
8.1*
0.9*
76.6
97.1 100
95.8
93.3 94
NS
crosssectional
Conv. HC2 LBC
NS
NS
NS
51 96 51 (HSIL)b
Coste, 2003 France, Paris 1,757 mean 33
Comments
Cyto- HPV logy testing method
Abnormal Pap %
Specificity
NPV (comCytology HPV Com- Cytology HPV Com- bined) (ASCUS) bined (ASCUS) bined
NS
99 85 99 (HSIL)b
NS
NS
HPV Primary Screening
Petry, 2003
Germany, Hannover and Tübingen
8,083 30–60 5% mean random 43 colposcopy in cyto/HPVwomen
Conv.a HC2
3.1
6.4
0.6*
43.5
97.8 100
98.0
95.3 93.8
100
Dalstein, 2004
France, Besançon
3,574 16–88 mean 37
Conv. HC2
14.8c
23.5c
7.4*c
86.8
94.3 100
91.8
83.4 80.2
100
Agorastos, 2005
Greece, Northern region
1,296 17–67 crossmean 43 sectional
Conv. LBA
1.7
2.9
0.3*d
50d
75d
98.4
97.4 96.2
100
Ratnam, 2000
Canada, Newfoundland
2,098 18–69 mean 30
10% random colposcopy in cyto-/ HPVwomen
Conv. HC1/HC2 9.2
10.8
1.4* 53.3e (CIN1)f
90.0e 100 e 51.0 e
51.0e 22.4 e 100 e
4,075 18–50
41% random colposcopy in cyto-/ HPVwomen; bias adjusted
LBC
27.4
3.2 61f (CIN3)f
91f 88f
82f
73f 79f
NS
NS
crosssectional
Conv. HC2
20.5*
3.4
83.9 NS
87.7
84.5 NS
NS
Kulasingam, USA, 2002 Washington State
107
Wright, 2000
South Africa, 1,365 35–65 Cape Town
HPV not a criterion for immediate colposcopy referral; 29 mo. FU
HC2 16.6 PCR MY09/11
15.7*
61
100
NS
Table 1. (continued) Dalstein/Bory/Graesslin/Quereux/Birembaut/Clavel
Author, year
Country
n
Age
Comments
Cyto- HPVlogy testing method
Abnormal Pap %
HRCIN2 HPV % positive
Sensitivity
Belinson, 2000
China, Shanxi
1,997 35–45 mean 39
Schiffman, 2000
Specificity
NPV (comCytology HPV Com- Cytology HPV Com- bined) (ASCUS) bined (ASCUS) bined
all women underwent colposcopy
LBC
HC2
25
18
4.3
94
95
Costa Rica, Guanacaste
1,119g 18–90 HPV not Conv. HC2 median a criterion 37 for immediate colposcopy referral
NS
NS
1.6*
77.7
Salmeron, 2003
Mexico, Morelos
7,732 15–85 crossmedian sectional 41
Conv. HC2
2.4
9.3
1.3*
59.4
Ferreccio, 2003
Costa Rica, Guanacaste
8,551 18–70 CIN3 found at enrolment and during the first 2 years of FU
Conv. PCR NS LBC MY09/11
NS
1.1* 63.0 f (CIN3)f 85.7 f
NS
78
85
NS
NS
88.4 NS
94.2
89.0 NS
NS
93.1 NS
98.3
91.8 NS
NS
85.2f 91.7 f 93.7 f 94.3 f 87.8 f
88.2f 84.2 f 99.9 f 80.2 f 99.9 f
108
CIN Cervical intraepithelial neoplasia; Conv. conventional cytology; FU follow-up; HC2 Hybrid Capture 2; LBA line blot assay; NPV negative predictive value; NS not stated; PCR polymerase chain reaction; *calculated by the authors. Commentaries: aMunich classification used for cytology; bcytology considered as positive at the HSIL cut-off; cstudied population attending for both primary and secondary screening, with high-level prevalence of CIN2; dvery low prevalence of CIN2 in the studied population, only 4 cases diagnosed, 3/4 HPVpositive and 2/4 with abnormal cytology; eincludes CIN1 lesions: sensitivity and specificity calculated for CIN1 detection; fonly CIN3 taken into account: sensitivity and specificity calculated for the detection of CIN3; gonly the cohort tested with HC2 (not HC1) presented in the table.
conclusions: (1) sensitivity of HPV testing (88–98%) is superior to that of cytology (51–86%); (2) specificity of HPV testing (83–94%) is lower than that of cytology (92–99%); (3) the sensitivity and negative predictive value of combined testing is near to 100% (percentages cited in parentheses correspond to: the 10 most representative values among those presented in the 14 selected studies). Persistence of HPV Infection Numerous HPV infections are known to regress spontaneously and the mean HPV infection duration is between 8 and 14 months [36]; 12 months after incident infection, 70% of the women were no longer infected, this number increased to 91% at 24 months [7, 37]. In total, only a small fraction of women will get persistent HR-HPV infections and will be at greater risk to develop cervical intra-epithelial neoplasia (CIN) 2, compared to women without any HR-HPV infection. Rozendaal et al. [38] showed that women with a normal smear and a positive HR-HPV have a 116 times higher risk of developing CIN3 than women with a normal smear and negative HR-HPV. Women with persistent cervical HR-HPV infections have a relative risk of developing cervical cancer of 100- to 300-fold increased [39, 40]. Persistence of the HPV infection is generally defined as 2 consecutive HR-HPV DNA tests positive in 1 year and screening should focus only on women with persistent HR-HPV infection, either in cytological normal and abnormal smear. Persistence of HPV infection in normal smears is about 30% [40] so present recommendations consist of performing a more intensive follow-up of HPV-positive women with normal smears, using an algorithm of 6–12 months, to detect the recurrence (or persistence with genotyping) of HPV infection and/or the occurrence of precancerous lesion. Colposcopy is not performed at the first normal smear HPV-positive but at the second combined testing because of the frequent natural clearance of the HPV infection. Moreover, sensitivity and specificity of colposcopy is improved when it is employed with an HPV test [19]. Such women at risk are detected earlier by HR-HPV testing with more CIN2 [40]. Negative Predictive Value Nevertheless, another important finding of most primary screening studies is that the majority of screened women has both normal Pap smears with a negative HR-HPV DNA test (at least 80–90%) [4, 20]. The negative predictive value (NPV) of these combined tests is very high: 99–100% and this indicates that these women negative on both tests are at very low risk for CIN2/3 or cancer and may receive a high level of reassurance [15, 38, 41]. The interval between initial HPV infection and the development of an invasive lesion being classically long, up to 15 years, this has also important implications for screening programs and could safely permit longer screening intervals in HR-HPV-DNA-negative
HPV Primary Screening
109
women with normal smears with decreased costs [4, 19]. Data from Dutch population showed that women double-negative tests were rescreened every 5 years and this interval could be prolonged to at last 8 years, without increasing the number of interval carcinomas, but 3–5 years may be more acceptable [19, 42]. Transient Infections and Improvement of Specificity of HPV Testing Therefore, if HPV DNA testing is proven useful for primary cervical cancer screening, the challenge becomes to manage large numbers of women with normal smears and transient HR-HPV infections (frequent co-infections and/or sequential infections at low risk for cervical cancer) without alarming or overtreating these women [10]. About 80% of infected women have transient infection [11]. These infections are called false positives in relation to morbidity. This notion of ‘positivity’, false or not, may be qualitative (presence of one or several HR-HPV types) and quantitative (viral load). The best number of HRHPV to detect has to be defined soon, to expand or not the number of 13 types of HR-HPV included in HC2 or in Amplicor tests. New evaluations would be necessary to assess between the benefit of HR-HPV DNA testing and the management of more women with more transient infections [10]. Thus, for women with normal smears and positive HR-HPV, the aim would be to offer the best costeffective test including only HR-HPV types that are most likely to progress to cancer [2], and new studies with genotyping and follow-up of patients in terms of clearance, persistence of the infection, with and without lesions, are needed. Otherwise, how to decrease the detection of transient infections, at least for primary screening and how to improve specificity? Guidelines from the ACS and adopted by the FDA restricted screening to women aged 30 years and older, who are the most at risk for HSIL development and no more frequently than every 3 years [43–45]. Moreover, Goldie et al. [46] concluded that such primary screening strategy with combined tests (or cytology with reflex HPV DNA testing for equivocal results) will provide a greater reduction in cancer and be less costly than annual conventional cytology [47]. An additional approach is also the screening of older women 50 years. When to stop the screening, knowing that postmenopausal women without previous HPV infection have very low risk of developing invasive cancer. If up to 25% of cases of cervical cancer occur in women 65 years, primary screening would precede the development of cervical lesions with the detection of the persistence of the HPV infection [5]. So such screening could be improved rationally with safely increased intervals for women who test HPV negative at least 2 consecutive times in 2 years for example, but we need more studies in elderly women with follow-up data on clearance, re-infection, persistence of their HPV infection with the prevalence of specific types with age. In conclusion, Baay et al. [5] estimates that the addition of HPV testing to cytology will
Dalstein/Bory/Graesslin/Quereux/Birembaut/Clavel
110
provide optimal information to distinguish between low- and high-risk women and that nearly 94% of the women 50 years could potentially be withdrawn from screening. Another challenge to improve the low specificity and PPV of HPV DNA test is to study new viral markers like viral load, detection of expression of E6/E7 oncogenes, detection of variants of HPV 16, integration of oncogenic viruses, analysis of ploidy. From previous and recent results, feasible now on a large scale, viral load and expression of HR-HPV oncogenic proteins are promising. A high viral load may be considered as a risk factor and preferentially observed in potentially progressive lesions and in HGSIL. Data for viral load using semiquantitative methods based on HC2 are less conclusive and even confusing, reflecting multiple infections of which many will eventually regress [9, 22]. New real-time PCRs to study viral load offer better specificity for high grade lesion when the thresholds for positivity has a clinical significance, of about 105 HPV copies, which gives better specificity and little loss of sensitivity for high-grade CIN [12, 48]. Differently, expression of E6/E7 viral oncogenes is essential for the lesion development and these mRNAs might more effectively predict lesion presence, with potential implications in triage. Thus, detection of E6/E7 transcripts from HR-HPV types16, 18, 31, 33 and 45 (HPV PreTect Proofer®, NorChip, Norway) represents a new promising technology in HPV diagnostics as an adjunct to cytology [49–51]. Primary Screening Using HPV Testing Alone The latest proposition is to start the primary screening by the most sensitive and automated test which is HPV testing and in a second time to use the best specificity of cytology for diagnosis and triage [9, 12]. Women with abnormal smears would be immediately recalled for colposcopy. This policy has been proposed in low-resource settings, where HPV DNA testing programs may be easier to implement than cytologic screening [16, 33, 52]. Cuzick et al. [12] proposed an algorithm involving 5-yearly screening by HPV from 25 to 64 years, but largescale studies are urgently needed to control the validity of these concepts (fig. 1). Developing Countries In many developing countries, cytology-based screening presents more constraints. Evaluation with alternative simple and cost-effective methods is proposed such as to first screen with visual inspection of the cervix after application of acetic acid (VIA) or Lugol’s iodine (VILI) and HPV testing [4, 19]. HPV prevalence is higher, reaching rates of 20% and more [1, 4, 32]. Planning and screen-and-treat approach are possible during the same visit with logistic advantages, despite the low specificity of VIA and VILI. The benefit is maximal in such settings, where compliance is poor [16].
HPV Primary Screening
111
The self-sampling method using self-collected vaginal samples is another alternative approach for primary screening. The aims would be to try to increase the coverage of a population when women do not undergo a gynecological examination and when cytology screening is not available and would allow also to screen HR-HPV at least once in women aged more than 30–35 years. Prevalence of HR-HPV of self-sampled vaginal material is about 5–10% lower than for cervical smears with a decreased sensitivity for detecting CIN [11, 39]. But for women who do not participate in programs of screening, vaginal self-sampling could be a good alternative and could reduce the risk of cervical cancer [4].
Present Recommendations and Conclusions for Primary Screening
Methods Valuable for Screening Methods have to take in account sensitivity and intervals [19]. In 2003, the FDA approved the use of HC2 for primary screening with HPV testing combined with cytology for women aged 30 and above [53]. With this new screening option, recommendations from an expert panel of representatives from the American Cancer Society (ACS), the American College of Obstetricians and Gynecologist (ACOG), the National Cancer Institute (NCI), the American Society for Colposcopy and Cervical Pathology (ASCCP), and the Center for Disease Control and Prevention (CDC) [45] guide US clinicians with the combined use of HPV testing and the Pap test. Consistent scientific evidence of level A concerns the high sensitivity of HPV testing and its NPV. ‘Because HPV testing is more sensitive than cervical cytology in detecting CIN2 and CIN3, women with negative concurrent test results can be reassured that their risk of unidentified CIN2 and CIN3 or cervical cancer is approximately 1/1,000. Studies using combined HPV testing with cervical cytology have reported a negative predictive value for CIN2 and CIN3 of 99–100%.’ At present, American agencies are the only ones to propose the use of combined cytology and HPV testing in primary screening, on the basis of recommendations of grade B. Other European agencies generally state that further investigation is needed or that insufficient evidence is available to support the use of HPV testing in a primary screening setting [54]. Conventional cytology or liquid-based cytology are both valuable for screening. Initiation of Screening When to initiate screening is controversial. Both the ACOG and ACS American guidelines agree that cytological screening should be initiated 3 years
Dalstein/Bory/Graesslin/Quereux/Birembaut/Clavel
112
after the onset of sexual activity, but no later than 21 years. In France, the ANAES recommends to start cytological screening at 25 years of age and EUROGIN recommends 8 years after the first intercourse [19]. For HPV testing, American societies recommend not to start before the age of 30 years. However, Cuzick’s group proposed a new algorithm to use of HPV alone as the primary test for women 25 years [8, 12], 25 years being certainly more realistic, if we consider the young age at first sexual intercourse. Screening Interval Recommendations on screening interval differ slightly between agencies. For the first screening cycles, the majority agrees that an annual cytological screening is recommended, but expensive [19]. The ACS guidelines recommend also expanding the screening interval to every 2 years when using LBC, in place of yearly testing with the conventional method. After age 30 and at least 3 years with technically satisfactory negative cytological results, women may switch to testing every 2–3 years using either conventional cytology or LBC in case of cytology alone, and no more frequently than every 3 years if combined cytology and HPV testing is used (limited scientific evidence of grade B) [19, 43, 44]. At least, women 30 years with a negative cytology result with positive HR-HPV DNA test results should have both tests repeated every 6–12 months. Those with persistent HR-HPV (on repeat testing) should undergo colposcopy regardless of the cytological result (consensus and expert opinion of level C) [44]. When to Stop Screening When to discontinue screening is another controversial issue. Women older than 65 (French agency) or 70 years (ACS) who are not at high risk may safely stop screening. More precisely, the ACS guidelines recommend that they have had 3 or more documented, technically satisfactory normal cytology (and negative HPV testing) results and have had no abnormal results within the past 10 years. In contrast, the ACOG states that ‘evidence is inconclusive to establish an upper age limit for cervical cancer screening’. Particular Recommendations Depending on the agencies, several groups are identified as women at ‘high risk’, and, as a consequence, more frequent testing (annual or biannual) is recommended in the following cases: immunocompromized patients, HIVpositive women, women with a former history of CIN2/3 or cervical cancer, women with in utero diethylstilbestrol exposure. For ACS and ACOG, women who have undergone a total hysterectomy may safely discontinue cervical screening [53, 55].
HPV Primary Screening
113
Table 2. Forthcoming randomized controlled trials comparing screening strategies for cervical cancer, including cytology and HPV testing Dalstein/Bory/Graesslin/Quereux/Birembaut/Clavel
Study
Country
Total Age HPV test recruit- range ment years
Cytology
Design
Comparison
Main study outcomes
Initiated Anticipated completion
CCCAST
Canada
12,000
30–69 HC2
conventional randomized Pap smear controlled trial
HPV first, then Pap versus Pap first, then HPV
comparative detection of histologically confirmed CIN2
2002
2005
Swedescreen
Sweden
12,527
32–38 PCR conventional randomized (GP5/6) Pap smear controlled trial
HPV Pap versus Pap alone
comparative prevalence of histologically confirmed CIN2 at the exit screen
2001
2006
POBASCAM
The 44,102 Netherlands
30–60 PCR conventional randomized (GP5/6) Pap smear controlled trial
HPV Pap versus Pap alone
proportion of 1999 histologically confirmed CIN3 found at any time during the trial from recruitment to exit screen
2007
114
HPV Primary Screening
ARTISTIC
UK
25,000
20–64 HC2
LBC (ThinPrep, Cytyc)
NTCC
Italy
95,000
25–60 HC2
LBC randomized (ThinPrep, controlled Cytyc) or trial conventional Pap smear
Finnish Finland Randomized Public Health Trial
2,00,000 25–65 HC2
randomized controlled trial
HPV LBC comparative versus LBC prevalence alone of histologically confirmed CIN3 at the exit screen
2001
2007
phase 1: HPV LBC versus Pap phase 2: HPV alone versus Pap alone
comparative detection of histologically confirmed CIN2 from the recruitment screen up to and including the exit screen
2002
2007
cumulative incidence of CIN2, CIN3 and cancer after initial screening
2003
2009
conventional randomized HPV alone Pap smear intervention versus Pap evaluation alone
Adapted from Davies et al., Int J Cancer 2006;118:791–796.
115
Conclusions
HPV testing alone or in combination with cytology has a greater sensitivity and negative predictive values and is able to detect almost all cases of CIN2. Interim guidance with international consensus is evolving but approaches for managing women with normal smears with positive HPV testing have not yet been agreed upon. The challenges for primary screening are to develop strategies for reducing the number of women with transient infections, considering the age of the screened women and the development of new markers of HPV oncogenicity to improve the specificity of HPV testing and to help clinicians distinguish women at real risk of developing high-grade cervical lesions. HPV-DNA-positive women should receive counseling to be followed-up in order to identify those with persistent infection. So, education programs for both the physicians and the women are still key points to manage questions and potential psychosocial difficulties. Self-sampling could be a solution to increase the coverage of the population and is still being studied. Definitive evaluation of efficacy and costeffectiveness is still needed from long-term follow-up studies with invasive cancer as an outcome and from ongoing randomized controlled trials [15] (table 2). HPV testing would logically permit a more rational approach for cervical cancer screening especially with the increasing development of prophylactic and therapeutic vaccines. New guidelines are necessary and become an emergency to define screening and follow-up of vaccinated women or not, whatever their age or country, with acceptable technologies and costs.
References 1 2
3 4 5 6
7 8
Bosch FX, Lorincz A, Munoz N, Meijer CJ, Shah KV: The causal relation between human papillomavirus and cervical cancer. J Clin Pathol 2002;55:244–265. Clifford GM, Gallus S, Herrero R, et al: Worldwide distribution of human papillomavirus types in cytologically normal women in the International Agency for Research on Cancer HPV prevalence surveys: a pooled analysis. Lancet 2005;366:991–998. Munoz N, Bosch FX, de Sanhose S, et al: Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518–527. Denny LA, Wright TC Jr: Human papillomavirus testing and screening. Best Pract Res Clin Obstet Gynaecol 2005;19:501–515. Baay MF, Smits E, Tjalma WA, et al: Can cervical cancer screening be stopped at 50? The prevalence of HPV in elderly women. Int J Cancer 2004;108:258–261. Kjaer SK, van den Brule AJ, Paull G, et al: Type specific persistence of high risk human papillomavirus (HPV) as indicator of high grade cervical squamous intraepithelial lesions in young women: population based prospective follow up study. BMJ 2002;325:572. Ho GY, Bierman R, Beardsley L, Chang CJ, Burk RD: Natural history of cervicovaginal papillomavirus infection in young women. N Engl J Med 1998;338:423–428. Cuzick J, Sasieni P, Davies P, et al: A systematic review of the role of human papillomavirus testing within a cervical screening programme. Health Technol Assess 1999;3:1–196.
Dalstein/Bory/Graesslin/Quereux/Birembaut/Clavel
116
9 10 11 12 13 14 15 16 17
18 19 20 21 22 23 24 25
26
27 28
29 30
31 32
33
Clavel C, Masure M, Bory JP, et al: Human papillomavirus testing in primary screening for the detection of high-grade cervical lesions: a study of 7932 women. Br J Cancer 2001;84:1616–1623. Wright TC Jr, Schiffman M: Adding a test for human papillomavirus DNA to cervical-cancer screening. N Engl J Med 2003;348:489–490. Meijer CJ, Snijders PJ, van den Brule AJ: Screening for cervical cancer: should we test for infection with high-risk HPV? CMAJ 2000;163:535–538. Cuzick J: Role of HPV testing in clinical practice. Virus Res 2002;89:263–269. Herrero R, Hildesheim A, Bratti C, et al: Population-based study of human papillomavirus infection and cervical neoplasia in rural Costa Rica. J Natl Cancer Inst 2000;92:464–474. Sellors JW, Karwalajtys TL, Kaczorowski JA, et al: Prevalence of infection with carcinogenic human papillomavirus among older women. CMAJ 2002;167:871–873. Duarte-Franco E, Franco EL: Cancer of the uterine cervix. BMC Womens Health 2004;4(suppl 1): S13. Franco EL: Primary screening of cervical cancer with human papillomavirus tests. J Natl Cancer Inst Monogr 2003;31:89–96. Nanda K, McCrory DC, Myers ER, et al: Accuracy of the Papanicolaou test in screening for and follow-up of cervical cytologic abnormalities: a systematic review. Ann Intern Med 2000;132: 810–819. Fahey MT, Irwig L, Macaskill P: Meta-analysis of Pap test accuracy. Am J Epidemiol 1995;141: 680–689. Monsonego J, Bosch FX, Coursaget P, et al: Cervical cancer control, priorities and new directions. Int J Cancer 2004;108:329–333. Lorincz AT, Richart RM: Human papillomavirus DNA testing as an adjunct to cytology in cervical screening programs. Arch Pathol Lab Med 2003;127:959–968. Iftner T, Villa LL: Human papillomavirus technologies. J Natl Cancer Inst Monogr 2003;31:80–88. Kulasingam SL, Koutsky LA: Will new human papillomavirus diagnostics improve cervical cancer control efforts? Curr Infect Dis Rep 2001;3:169–182. Schneider A, Hoyer H, Lotz B, et al: Screening for high-grade cervical intra-epithelial neoplasia and cancer by testing for high-risk HPV, routine cytology or colposcopy. Int J Cancer 2000;89:529–534. Cuzick J, Szarewski A, Cubie H, et al: Management of women who test positive for high-risk types of human papillomavirus: the HART study. Lancet 2003;362:1871–1876. Coste J, Cochand-Priollet B, de Cremoux P, et al: Cross sectional study of conventional cervical smear, monolayer cytology, and human papillomavirus DNA testing for cervical cancer screening. BMJ 2003;326:733. Petry KU, Menton S, Menton M, et al: Inclusion of HPV testing in routine cervical cancer screening for women above 29 years in Germany: results for 8466 patients. Br J Cancer 2003;88: 1570–1577. Dalstein V, Riethmuller D, Sautiere JL, et al: Detection of cervical precancer and cancer in a hospital population: benefits of testing for human papillomavirus. Eur J Cancer 2004;40:1225–1232. Agorastos T, Dinas K, Lloveras B, et al: Human papillomavirus testing for primary screening in women at low risk of developing cervical cancer: the Greek experience. Gynecol Oncol 2005;96:714–720. Ratnam S, Franco EL, Ferenczy A: Human papillomavirus testing for primary screening of cervical cancer precursors. Cancer Epidemiol Biomarkers Prev 2000;9:945–951. Kulasingam SL, Hughes JP, Kiviat NB, et al: Evaluation of human papillomavirus testing in primary screening for cervical abnormalities: comparison of sensitivity, specificity, and frequency of referral. JAMA 2002;288:1749–1757. Wright TC Jr, Denny L, Kuhn L, Pollack A, Lorincz A: HPV DNA testing of self-collected vaginal samples compared with cytologic screening to detect cervical cancer. JAMA 2000;283:81–86. Belinson J, Qiao YL, Pretorius R, et al: Shanxi Province Cervical Cancer Screening Study: a cross-sectional comparative trial of multiple techniques to detect cervical neoplasia. Gynecol Oncol 2001;83:439–444. Schiffman M, Herrero R, Hildesheim A, et al: HPV DNA testing in cervical cancer screening: results from women in a high-risk province of Costa Rica. JAMA 2000;283:87–93.
HPV Primary Screening
117
34
35
36
37
38
39
40
41 42
43 44 45
46 47 48 49
50 51 52
53
Salmeron J, Lazcano-Ponce E, Lorincz A, et al: Comparison of HPV-based assays with Papanicolaou smears for cervical cancer screening in Morelos State, Mexico. Cancer Causes Control 2003;14:505–512. Ferreccio C, Bratti MC, Sherman ME, et al: A comparison of single and combined visual, cytologic, and virologic tests as screening strategies in a region at high risk of cervical cancer. Cancer Epidemiol Biomarkers Prev 2003;12:815–823. Franco EL, Villa LL, Sobrinho JP, et al: Epidemiology of acquisition and clearance of cervical human papillomavirus infection in women from a high-risk area for cervical cancer. J Infect Dis 1999;180:1415–1423. Baay MF, Tjalma WA, Lambrechts HA, et al: Combined Pap and HPV testing in primary screening for cervical abnormalities: should HPV detection be delayed until age 35? Eur J Cancer 2005;41:2704–2708. Meijer CJ, Helmerhorst TJ, Rozendaal L, van der Linden JC, Voorhorst FJ, Walboomers JM: HPV typing and testing in gynaecological pathology: has the time come? Histopathology 1998;33: 83–86. Bekkers RL, Meijer CJ, Massuger LF, Snijders PJ, Melchers WJ: Effects of HPV detection in population-based screening programmes for cervical cancer: a Dutch moment. Gynecol Oncol 2006:100:451–454. Bory JP, Cucherousset J, Lorenzato M, et al: Recurrent human papillomavirus infection detected with the hybrid capture II assay selects women with normal cervical smears at risk for developing high grade cervical lesions: a longitudinal study of 3,091 women. Int J Cancer 2002;102: 519–525. Sherman ME, Lorincz AT, Scott DR, et al: Baseline cytology, human papillomavirus testing, and risk for cervical neoplasia: a 10-year cohort analysis. J Natl Cancer Inst 2003;95:46–52. Clavel C, Cucherousset J, Lorenzato M, et al: Negative human papillomavirus testing in normal smears selects a population at low risk for developing high-grade cervical lesions. Br J Cancer 2004;90:1803–1808. Smith RA, Cokkinides V, Eyre HJ: American Cancer Society Guidelines for the Early Detection of Cancer, 2005. CA Cancer J Clin 2005;55:31–44. ACOG: ACOG Practice Bulletin. Clinical Management Guidelines for Obstetrician-Gynecologists. Number 61, April 2005. Human papillomavirus. Obstet Gynecol 2005;105:905–918. Wright TC Jr, Schiffman M, Solomon D, et al: Interim guidance for the use of human papillomavirus DNA testing as an adjunct to cervical cytology for screening. Obstet Gynecol 2004;103: 304–309. Goldie SJ, Kim JJ, Wright TC: Cost-effectiveness of human papillomavirus DNA testing for cervical cancer screening in women aged 30 years or more. Obstet Gynecol 2004;103:619–631. Holmes J, Hemmett L, Garfield S: The cost-effectiveness of human papillomavirus screening for cervical cancer: a review of recent modelling studies. Eur J Health Econ 2005;6:30–37. Snijders PJ, van den Brule AJ, Meijer CJ: The clinical relevance of human papillomavirus testing: relationship between analytical and clinical sensitivity. J Pathol 2003;201:1–6. Molden T, Kraus I, Karlsen F, Skomedal H, Nygard JF, Hagmar B: Comparison of human papillomavirus messenger RNA and DNA detection: a cross-sectional study of 4,136 women 30 years of age with a 2-year follow-up of high-grade squamous intraepithelial lesion. Cancer Epidemiol Biomarkers Prev 2005;14:367–372. Stoler MH: Human papillomavirus biology and cervical neoplasia: implications for diagnostic criteria and testing. Arch Pathol Lab Med 2003;127:935–939. Doorbar J, Cubie H: Molecular basis for advances in cervical screening. Mol Diagn 2005;9:129–142. Kuhn L, Denny L, Pollack A, Lorincz A, Richart RM, Wright TC: Human papillomavirus DNA testing for cervical cancer screening in low-resource settings. J Natl Cancer Inst 2000;92: 818–825. ACOG: ACOG Practice Bulletin: clinical management guidelines for obstetrician-gynecologists. Number 45, August 2003. Cervical cytology screening (replaces committee opinion 152, March 1995). Obstet Gynecol 2003;102:417–427.
Dalstein/Bory/Graesslin/Quereux/Birembaut/Clavel
118
54 55
Scheurer ME, Tortolero-Luna G, dler-Storthz K: Human papillomavirus infection: biology, epidemiology, and prevention. Int J Gynecol Cancer 2005;15:727–746. Saslow D, Runowicz CD, Solomon D, et al: American Cancer Society guideline for the early detection of cervical neoplasia and cancer. CA Cancer J Clin 2002;52:342–362.
Prof. Christine Clavel, MD, PhD Unité de Biologie Cellulaire. Laboratoire Pol Bouin CHU de Reims, 45, rue Cognacq-Jay FR–51100 Reims (France) Tel. 33 326 78 75 52, Fax 33 326 78 77 39, E-Mail
[email protected] HPV Primary Screening
119
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 120–139
HPV Testing in Patient Management: Atypical Squamous Cells of Undetermined Significance and LowGrade Squamous Intraepithelial Lesion J. Thomas Cox University of California, Santa Barbara, Calif., USA
High-grade cervical intraepithelial neoplasia (CIN2/3) and invasive cervical cancer do not occur in the absence of persistent high-risk HPV [1–3]. This simple, but revolutionary, revelation led to the development and evaluation of sensitive, clinically relevant molecular tests for HPV [4–7]. Increasing clinical experience with HPV DNA testing delineated five areas where HPV DNA testing is clinically useful. Four of these areas are in the management of abnormal cervical cytology or documented CIN (table 1). The fifth area is in primary cervical screening as an adjunct to cytology. For some of these uses, HPV DNA testing simply provides a further option to other highly acceptable approaches. In other instances, such as the management of women with atypical squamous cells of undetermined significance (ASCUS), the use of HPV DNA testing is preferred because of cost and relative ease of use. This chapter will review the use of HPV testing in the management of women with cervical cytology interpretations of ASCUS and low-grade squamous intraepithelial lesion (LSIL).
ASCUS and LSIL Cytology: Nature’s Continuum
The nature of humans is to strive to succinctly divide all life into crisply defined categories. However, in truth, much in life is a continuum that cannot be clearly categorized. The 1988 Bethesda system (TBS) that created the equivocal category of ASCUS and the more crisply defined category of LSIL acknowledged this continuum [8]. ASCUS is the border between clearly normal
Table 1. Clinical applications of HPV testing The four areas where HPV testing has been proven to be clinically useful in the management of women with abnormal cervical cytology: Triage of women of any age with ASC-US results Triage of postmenopausal women with LSIL results Postcolposcopy follow-up in selected instances Follow-up of women posttreatment for CIN2/3
and clearly abnormal, as evidenced by TBS terminology ‘of undetermined significance’ attached to the category of ‘atypical cells’. In contrast, LSIL encompasses cytological changes associated with the more specific cytopathic effects of HPV known as koilocytotic atypia and mild dysplasia/CIN1 [8]. LSIL is less common than ASCUS. In the majority of US labs, the ratio between ASCUS and LSIL is in the range of 2:1 [9]. Until the 2001 Bethesda Workshop, atypical cell interpretations that ‘leaned toward’ HPV effect but were not clearly abnormal enough to be rated as LSIL were often read as ASCUS ‘cannot rule out LSIL’. TBS 2001 folded this subcategory and the subcategories of ASCUS ‘unqualified’ and ASCUS ‘favor reactive’ that could not be relegated to normal into ASC-US [10]. A second subcategory of ASCUS also created by TBS 2001 is atypical squamous cells ‘cannot rule out high grade’ (ASC-H). ASC-H is about one tenth as common as ASC-US and is defined as atypical cells, difficult to distinguish from but not definitive for high-grade cells [10, 11]. Since its creation, it has been popular to derisively label the cytologic interpretation of ASC-US as ‘Don’t ASK-US!’, for this has easily been the most perplexing of the abnormal Papanicolaou (Pap) interpretations. The difficulty with this cytologic interpretation has been secondary to a number of factors. While the risk of CIN2/3⫹ for each individual woman with ASC-US is relatively low, the commonness of this Pap interpretation ensures that more CIN2/3⫹ is detected in the follow-up of ASC-US than in the follow-up of any other Pap category [12]. This dichotomy established the need for vigilance in the follow-up of women with ASC-US, while at the same time avoiding the potential for overmanagement of the majority not destined to have serious disease [13]. The difficulty in finding the best path to navigate between these two inherently conflicting goals resulted in the 5,000-patient National Cancer Institute-sponsored randomized ASCUS LSIL Triage Study (ALTS) [7, 14, 15]. Results from this study have provided much of the basis for the US guidelines on the management of ASCUS and LSIL. These results will be discussed along with more recent European data and meta-analyses of worldwide data from multiple studies evaluating ASCUS and LSIL triage.
Management of ASCUS and LSIL
121
ASC-US, ASC-H and LSIL: What Is the Impact? Each year in the United States, approximately 2.5 million of the 55–60 million Pap test results are interpreted as having equivocal cytological changes and another 1.25 million as having low-grade changes. Because ASCUS is the gray zone between normal and abnormal and is not a sharply defined category, the interpretation of ASCUS varies considerably from one observer to another [16–19]. For example, in ALTS, out of 1,473 ASCUS cases originally read by good clinical center pathologists, less than half were reread as ASCUS in a blinded panel review by expert pathologists [16]. Most of the 57% categorized as something other than ASCUS were reread as normal. Such variability in interpretation has resulted in significant differences amongst clinicians in both the perception of risk with ASCUS and in the subsequent clinical response [20]. Additionally, rates of ASC-US are at least partially secondary to the medicolegal climate, with increasing rates often reflecting concern over ‘missing a lesion’ that has varying risks of subsequent litigation depending on the locale [13, 20, 21]. Hence, countries with a traditionally lower risk of legal pursuit following unfavorable outcomes have enjoyed lower ASC-US rates [22]. These international differences in the reading of borderline or equivocal Pap results are quite marked. For example, in one study on interobserver variability in the reading of ASCUS in three countries, the British cytopathologist tended to rate the Pap tests as more abnormal than the American cytopathologists, whereas the Scandinavians tended to rate the same equivocal Pap tests as more normal than the Americans. Oncogenic HPV DNA detection was significantly associated with increasingly abnormal interpretations for each reader and, in reference to the HPV DNA standard, clarified the tendency of readers to render systematically more or less severe interpretations [18]. The study confirmed the subjectivity of cytology, particularly at the lower end of cytological abnormality, and that HPV testing provides the objective measure that can clarify the nature of cell changes that are less clear. ASCUS is more common in young populations with higher rates of infection with HPV. Up to 60–80% of women under the age of 30 have an ASCUS test positive for HPV, whereas the proportion of women with ASCUS testing positive for HPV declines with age [23]. This decline is not secondary to increased difficulty detecting HPV in older women, but rather results from nonimportant but difficult to evaluate minor cytology changes due to the epithelial effects of aging and estrogen decline [24]. ASC-H is of greater concern for high-grade disease than ASC-US. The cytology result of ASC-H interpreted by expert cytopathologists in ALTS was found to be high-risk HPV positive in 86% when the ASC-H interpretation was derived from liquid-based Pap tests and in 69.8% when derived from a conventional Pap smear [25]. This higher association with high-risk HPV was also
Cox
122
Table 2. Cumulative diagnosis of two disease endpoints (CIN2/3 by clinical center pathology diagnosis and CIN3 by pathology quality control consensus diagnosis) by referral Pap interpretations, ASC HPV (⫹) and LSIL
Clinical center CIN2 or CIN3 Pathology quality control group CIN3
HPV ⫹ ASCUS, % (n ⫽ 1,193)
LSIL, % (n ⫽ 897)
26.7 (24.2–29.3) 14.5 (12.6–16.6)
27.6 (24.7–30.7) 15.9 (13.6–18.5)
Women diagnosed with the disease endpoints at any time during ALTS enrollment, 2-year follow-up, or exit. With permission from Cox et al. [28].
demonstrated to carry a higher risk of CIN2/3⫹, which was found in 40% of women with ASC-H on ThinPrep Pap tests (CYTYC, Boxsborough, Mass., USA) and in 27.2% of women with ASC-H interpreted on conventional Pap smears [25]. A 3-year retrospective review of women referred for the evaluation of ASC-H Pap tests found that 51% had CIN, with CIN2/3⫹ comprising about half of these [26]. The risk of CIN for women with ASC-H has been reported to be somewhat lower for those over the age of 40 than for younger women [27]. Typically, between 15 and 30% of women with LSIL, Pap interpretations are found to have CIN2/3 [15, 28]. Women with ASC-US cytology have about half this risk, but when ASC-US is triaged by HPV testing, the risk of HPVpositive ASC-US for CIN2/3 is virtually identical to the risk attendant with an LSIL interpretation [14, 28]. This is clearly demonstrated by the ALTS 2-year follow-up of women with either HPV-positive ASC-US or LSIL, which documented a virtually identical cumulative risk of high-grade disease (26.7% for HPV-positive ASC-US and 27.6% for LSIL) (table 2) [28].
HPV Testing in the Management of Women with ASCUS
It should be understood that the majority of studies on the management of ASCUS was done prior to the separation of ASC-H from ASCUS [7, 14, 23, 25, 28–31]. However, ASC-H is an equivocal subcategory, and although the risk of CIN2/3 for women with ASC-H is higher than for women with ASC-US, separation of the small number given this cytologic interpretation from the ASC-US category does not significantly alter the conclusions derived from these studies. ALTS contributed substantially to understanding the major strengths and weaknesses of all the follow-up options for women with ASCUS and LSIL [7, 14, 15].
Management of ASCUS and LSIL
123
ALTS also clarified the management of women after colposcopy and after treatment, and has contributed to a greater understanding of the natural history of HPV as it relates to abnormal cervical cytology, CIN and cervical cancer [28, 32, 33]. Subsequent meta-analysis of the most validated studies on the management of ASCUS and LSIL that included data from outside the US substantiated the preponderance of the ALTS findings [34]. The sensitivity of detection of CIN2/3 in ALTS (92–95%) of a reflex HPV test was similar to the sensitivity of two repeat liquid-based Pap tests, but only at the threshold for referral to colposcopy of repeat cytology of ⱖASC-US [7, 14, 15]. Reflex HPV testing is cost effective in the management of ASC-US from liquid-based cytology because the same level of detection of high-grade precursor lesions is obtained directly from HPV testing of the residual Pap medium at the lab while eliminating the two office visits required for cytology follow-up. Additionally, at the ASC-US threshold for referral to colposcopy, two thirds had another abnormal Pap on one of the two repeat visits and required colposcopy. In contrast, only 53% were referred to colposcopy on the basis of a positive HPV test [14]. Increasing the threshold for referral to colposcopy to ⱖLSIL would significantly reduce the referral rate to colposcopy but was too insensitive for CIN2/3 (74%) to be considered safe (table 3). Increasing the number of repeat cytology visits to three at this threshold continued to miss 18% of the CIN2/3 [14]. In 2001, the American Society for Colposcopy and Cervical Pathology (ASCCP)-sponsored consensus conference developed the first comprehensive evidence-based management recommendations in the US for women with cervical cytological abnormalities [35]. These guidelines incorporated the changes in terminology that took place as a result of the 2001 Bethesda Workshop and the findings of ALTS, as well as from other important studies on the management of ASCUS [10, 14, 30]. In 2005, the American College of Obstetricians and Gynecologists (ACOG) ratified abnormal Pap management guidelines that are identical in most aspects with the 2001 ASCCP guidelines [35, 36]. Although both organizations consider repeat cervical cytology or immediate colposcopy to be safe and effective options for the management of women with ASC-US, both also highlight testing for HPV as the most effective option when the Pap test is from a liquid-based sample. This ‘preferred’ status is secondary to being able to eliminate the return visit for the follow-up triage tests. Additionally, HPV triage decreases colposcopic referrals by nearly 50%, yet detects at least as many CIN2/3 as immediate colposcopy of all women with ASCUS [7, 14]. This is likely to be secondary to increased colposcopic diligence when the patient appears to be at higher risk (i.e. high-risk HPV-positive ASCUS) than for patients referred with ASCUS without knowledge of HPV status [20].
Cox
124
Table 3. Estimated triage test performance in the ALTS trial for the detection of cumulative CIN3 over the 2 years of the trial Sensitivity for CIN3 (%)
Referral (%)
Enrollment HPV DNA test
92.4 (88.7–95.2)
53.2 (51.5–54.9)
HSIL cytology threshold One Two Three
35.5 (30.0–41.3) 48.3 (38.8–57.7) 60.2 (50.8–69.6)
7.1 (6.2–8.0) 10.2 (8.5–12.0) 11.7 (9.8–13.6)
LSIL cytology threshold One Two Three
59.3 (53.4–65.0) 74.1 (65.8–82.3) 82.0 (74.7–89.4)
25.1 (23.6–26.6) 31.7 (29.0–34.4) 37.2 (34.4–40.1)
ASCUS cytology threshold One Two Three
83.4 (78.7–87.5) 95.4 (91.4–99.3) 97.2 (94.1–100)
58.1 (56.4–59.8) 67.1 (64.4–69.8) 72.7 (70.1–75.4)
With permission from the ASCUS-LSIL Triage Study (ALTS) Group [14]. CIN3 diagnosis is that given by the ALTS pathology quality control group. Sensitivity for CIN3 and percent referral to colposcopy is given for the HC2 HPV test and for repeat Pap at one of three potential thresholds for referral to colposcopy: Paps interpreted as HSIL only, ⱖLSIL or ⱖASCUS. One, two and three represent Paps repeated at ASCCP guideline recommendations of every 4–6 months. One ⫽ First Pap usually repeated at 4–6 months; two ⫽ second Pap repeated at 4–12 months; three ⫽ third Pap, if three Paps were to be done, repeated at 12–18 months.
Arbyn et al. [34] reported on a meta-analysis of all relevant articles published between 1992 and 2002 on the management of ASCUS that included virologic and cytological testing followed by colposcopy. The fifteen articles that met all criteria (8 from the US, 2 from Israel, 1 from each of France, the Netherlands, Taiwan, Canada and the UK) gave the meta-analysis an international perspective [7, 22, 29, 30, 37–47]. A wide range of sensitivities and specificities was reported in the 15 studies for cytology, whereas HPV testing was quite consistent, particularly for the 8 studies using Hybrid Capture 2 (HC2; Digene, Gaithersburg, Md., USA) (tables 4, 5). Sensitivity of both cytology and HPV testing was essentially constant in the range of ages, whereas both became more specific with increasing age. The sensitivity of HPV testing depended upon the type of HPV test used. It was very low for ViraPap and ViraType (38.1%; 95% CI 7.1–69.1), better for HC1 (78.8%; 95% CI 67–90.5) and highest for HC2 (94.8; 95% CI 92.7–96.9). The pooled sensitivity of two repeat cytologies at the ⱖASCUS threshold was 81.8% (95% CI 73.5–84.3)
Management of ASCUS and LSIL
125
Table 4. Triage of atypical squamous cells of undetermined significance: accuracy parameters for the detection of CIN2/3⫹, test positivity rates and prevalence of high-grade CIN derived from different published studies Study
Sensitivity
Specificity
Positive predictive
Negative predictive
Test positivity rate
Prevalence CIN2/3
Manos et al. [30], 1999 Bergeron et al. [22], 2000 Fait et al. [43], 2000 Lin et al. [44], 2000 Shlay et al. [40], 2000 Morin et al. [45], 2001 Rebello et al. [46], 2001 Solomon et al. [7], 2001 Zielinski et al. [47], 2001
0.892 0.833 0.857 1.000 0.933 0.895 0.857 0.959 0.917
0.641 0.616 0.971 0.745 0.739 0.742 0.759 0.484 0.687
0.151 0.208 0.906 0.692 0.230 0.162 0.581 0.196 0.149
0.998 0.968 0.954 1.000 0.993 0.992 0.932 0.989 0.993
0.396 0.432 0.235 0.527 0.313 0.292 0.413 0.568 0.347
0.067 0.108 0.208 0.365 0.077 0.053 0.280 0.166 0.056
Modified with permission from Arbyn et al. [34].
Table 5. Accuracy parameters for repeat cytology at the threshold of ⱖASCUS or ⱖLSIL for the triage of women with ASCUS from the studies listed in table 4, directly comparing the two triage options of HPV testing by HC2 and repeat cytology Study
Type of Pap
Sensitivity
Specificity
Positive predictive value
Negative predictive value
Test positivity rate
ⱖASCUS Manos et al. [30], 1999 Bergeron et al. [22], 2000 Morin et al. [45], 2001 Solomon et al. [7], 2001
TP CP CP TP
0.762 0.667 0.737 0.850
0.638 0.717 0.629 0.447
0.129 0.222 0.100 0.167
0.974 0.947 0.977 0.958
0.389 0.324 0.390 0.588
ⱖLSIL Manos et al. [30], 1999 Morin et al. [45], 2001 Solomon et al. [7], 2001
TP CP TP
0.397 0.421 0.592
0.942 0.935 0.779
0.325 0.267 0.259
0.957 0.967 0.936
0.080 0.084 0.264
CP ⫽ Conventional Pap; TP ⫽ thinPrep. Modified with permission from Arbyn et al. [34].
Cox
126
Table 6. ASCCP and ACOG guidelines for the management of HPV-positive ASC-US
• • •
• •
DNA testing for high-risk types of HPV should be performed using a sensitive molecular test, and all women who are HPV DNA positive should be referred for colposcopic evaluation Women with ASC-US who are high-risk HPV DNA negative can be followed-up with repeat cytology at 12 months Acceptable management options for women who are positive for high-risk types of HPV, but who do not have biopsy-confirmed CIN, include follow-up with repeat cytology at 6 and 12 months with referral back to colposcopy if a result of ⱖASC-US is obtained, or HPV DNA testing at 12 months with referral back to colposcopy of all HPV DNA-positive women Women with HPV-positive ASC-US who are referred to colposcopy and found to have biopsy-confirmed CIN should be managed according to the appropriate ASCCP or ACOG guidelines for the management of histological abnormalities Because of the potential for overtreatment, diagnostic excisional procedures such as loop electrosurgical excision should not be routinely used to treat women with HPV-positive ASC in the absence of biopsy-confirmed CIN
and specificity was 57.6% (95% CI 49.5–65.7). Although specificity (89%) improved considerably at the colposcopy referral threshold of ⱖLSIL, the sensitivity was too low (45.7%; 95% CI 34–57.4) to be recommended for triage. The older commercially available HPV tests (ViraPap and HC1) were both considered too insensitive to be recommended for ASC-US management and although PCR was as sensitive as HC2, it was less specific and considered insufficiently evaluated to be recommended in ASC-US triage. The authors performed a separate analysis excluding the ALTS results (which dominated the other studies in terms of numbers), determining that the inclusion of ALTS did not affect the pooled relative accuracy values. The authors concluded that the HC2 assay is a better triage test for ASC-US than repeat cervical cytology [34]. The international reach of this meta-analysis extends the relevance of these conclusions to the management of women with ASCUS throughout the world. Several evaluations of the economies of different management strategies for ASCUS have demonstrated that reflex HPV DNA testing of ASCUS from liquid-based cytology is the most cost-effective strategy [48, 49]. The ASCCP and ACOG management guidelines for women with HPV-positive ASC-US are shown in table 6. The recommendation for all women with ASCUS testing positive for HPV is immediate referral to colposcopy (fig. 1) due to the 12.5- to 23-time increase in risk of CIN2/3⫹ over that for women with HPV-negative ASCUS, and a risk of invasive cervical cancer of about 1/500 (table 7) [7, 14, 29, 30]. The overall risk of CIN2/3⫹ for HPV-positive ASCUS in ALTS was 26.9% (from 20.1% at initial colposcopy) when women initially not found to have high-grade CIN were followed over 2 years [28]. Women found to have CIN2/3 at colposcopy should be treated according to local protocols. The management of
Management of ASCUS and LSIL
127
Fig. 1. ASCCP algorithm for the management of women with cervical cytology interpreted as ASC-US. Reprinted from Wright TC Jr, Cox JT, Massad LS, Twiggs LB, Wilkinson EJ: 2001 consensus guidelines for the management of women with cervical cytological abnormalities. J Low Gen Tract Dis 2002;6:127–143. With the permission of ASCCP © American Society for Colposcopy and Cervical Pathology, 2002.
women with HPV-positive ASCUS not found to have CIN2/3 requiring treatment at the initial colposcopy will be discussed subsequently. HPV Testing in the Management of ASC-H
Approximately 5–10% of atypical squamous cell interpretations are found to have cells that are of concern for a possible high-grade lesion but which are not conclusively high grade [11, 20]. ASC-H, defined as subcategory by the Bethesda 2001 Workshop, is reported in 0.27–0.6% of all Pap results [20, 26, 50]. ALTS provided the majority of the data providing justification for the creation of this subcategory and for the subsequent management recommendations. This increased the risk of CIN2/3, and the high rate of detection of HPV diminished the value of
Cox
128
Table 7. Three studies indicating the relative risk of initial detection of CIN2/3 for women referred for the evaluation of ASCUS, depending on HPV results Risk for detection of CIN2⫹ at initial colposcopy Study
Cox et al. [29], 19951 Manos et al. [30], 19992 Solomon et al. [7], 20012
ASCUS
Total risk for all ASCUS
HPV positive
HPV negative
17% 14/81 15% 45/300 20% 195/1,087
0.74% 1/136 1.2% 6/498 1.1% 13/1,175
6.9% 15/217 6.4% 51/801 11.9 % 208/2,210
With permission from Ferris et al. [66]. HC1 (expanded first-generation HC test). 2 HC2 (second-generation HC test). 1
HPV testing as a triage test and resulted in the ASCCP and ACOG recommendations that ASC-H should be managed by immediate colposcopy (fig. 2) [35, 36]. However, the cytologic interpretation of ASC-H is known to be very poorly reproducible, with one study reporting agreement between 3 expert cytologists of only 14% [51]. Poor reproducibility results in highly variable rates of detection of both HPV and high-grade lesions [52]. When CIN is not found at colposcopy, benign proliferations such as those in immature metaplasia and microglandular hyperplasia may be the source [53]. Other studies have shown far lower HPV-positive rates for women with ASC-H, confirming that the reading of all equivocal categories varies significantly from one observer to another [54]. HC2 was positive in ASCUS ‘not otherwise specified’ interpretations from one national reference laboratory in 40% and in ASC-H in only 37.5% [54]; in another review of 464 patients with ASC-H, only 44.5% were high-risk HPV positive [55]. Lyman et al. [52] demonstrated in their study that nearly half of the women with ASC-H had no CIN and that HC2 detected all 22 (100%) of the women with CIN2/3, similar to the 93.5% detection of CIN2/3 for women with ASC-H reported by Prince et al. [55]. In the latter study, highgrade lesions were detected in 29.8% of the women with HPV-positive ASC-H in comparison with only 0.016% of the women with HPV-negative ASC-H. In all of these studies evaluating HPV testing in the management of women with ASC-H, the risk of CIN2/3⫹ in women testing HPV negative is close to zero [51, 52, 54, 55]. These results question whether management recommendations
Management of ASCUS and LSIL
129
Fig. 2. ASCCP algorithm for the management of women with cervical cytology interpreted as ASC-H. Reprinted from Wright TC Jr, Cox JT, Massad LS, Twiggs LB, Wilkinson EJ: 2001 consensus guidelines for the management of women with cervical cytological abnormalities. J Low Gen Tract Dis 2002;6:127–143. With the permission of ASCCP © American Society for Colposcopy and Cervical Pathology, 2002.
on ASC-H made by ALTS and other studies primarily from academic sites may not be as universally applicable as initially presumed. Additionally, different geographic regions need to explore the meaning of their local cytological terminology and its correlation with HPV positivity in order to determine which equivocal cytological interpretations can be clarified by triage [56]. Several authors now suggest that high-risk HPV testing for women with ASC-H Pap results may serve as a better means for selecting patients with ASC-H who should undergo immediate colposcopic examination [51, 52, 57–59]. HPV Testing in the Management of LSIL
Over 80% of premenopausal women with a cytological interpretation of LSIL test positive for high-risk HPV, and many of the remainder have low-risk HPV types [15]. Therefore, the cytological changes of koilocytotic atypia and mild dysplasia quite accurately identify women with usually transient HPV infections [15, 17, 54, 60]. The cost effectiveness of a secondary triage test is dependent upon removing a large enough number of women (i.e. HPV negative)
Cox
130
Fig. 3. ASCCP algorithm for the management of women with cervical cytology interpreted as LSIL. Reprinted from Wright TC Jr, Cox JT, Massad LS, Twiggs LB, Wilkinson EJ: 2001 consensus guidelines for the management of women with cervical cytological abnormalities. J Low Gen Tract Dis 2002;6:127–143. With the permission of ASCCP © American Society for Colposcopy and Cervical Pathology, 2002.
from the requirement of having an expensive follow-up diagnostic procedure (i.e. colposcopy) to more than pay for the cost of the triage (HPV) test [61]. In contrast to the clearly cost-effective HPV triage of ASC-US, the high rate of HPV DNA-positive LSIL ensures that the savings accrued from avoided colposcopies would not cover the additional cost of HPV testing. Multiple evaluations of HPV in LSIL cytology have shown rates of high-risk HPV similar to the 83% detection in ALTS [31, 60]. As a result, HPV testing is not included as a triage option in any of the US guidelines for the management of premenopausal women with LSIL despite the high sensitivity of HPV testing for CIN2/3⫹ demonstrated in these studies (fig. 3) [35, 36, 62]. Misclassification of HPV-negative cellular changes that mimic koilocytotic atypia occurs with increasing frequency as women age. This phenomenon is most likely due to the effect of declining estrogen on the cervico-vaginal epithelium and
Management of ASCUS and LSIL
131
Fig. 4. ASCCP algorithm for the management of women with cervical cytology interpreted as LSIL in special situations (post-menopause). Reprinted from Wright TC Jr, Cox JT, Massad LS, Twiggs LB, Wilkinson EJ: 2001 consensus guidelines for the management of women with cervical cytological abnormalities. J Low Gen Tract Dis 2002;6:127–143. With the permission of ASCCP © American Society for Colposcopy and Cervical Pathology, 2002.
to cellular changes associated with aging [35, 63]. This results in only 30–50% of women with LSIL aged ⱖ40 testing positive for HPV [35, 64]. Additionally, fewer postmenopausal women with LSIL have CIN2/3⫹ [31, 64, 65]. Because HPV triage of postmenopausal women with LSIL identifies many women who are not at risk for CIN2/3⫹ and therefore do not need colposcopy, the ASCCP guidelines provide the option of using HPV testing in the management of postmenopausal women with LSIL (fig. 4) [35]. If colposcopy is not selected in the initial management of postmenopausal women with LSIL, two alternative ASCCP management options are given for women with adequate previous screening histories and no previous history of CIN. These include either an HPV test in 12 months or repeat cytology at 6 and 12 months, with referral to colposcopy at a threshold of ⱖASCUS or a positive HPV test. In contrast, the ACOG guidelines do not mention the use of HPV testing for the management of LSIL for women of any age. Both guidelines also provide the option of repeating the Pap test following a course of
Cox
132
vaginal estrogen cream for women with evidence of vaginal atrophy [35, 36]. Although the ASCCP guideline for management of postmenopausal women with LSIL does not provide the option of immediate reflex HPV testing, this option is as valid as it is for the management of ASC-US and therefore may be considered [66].
HPV Testing in the Postcolposcopy Management of Women with ASC-US,ASC-H or LSIL
Management of women referred to colposcopy and found to have CIN2/3 is not complicated, as most are treated [36, 67]. In contrast, the management of women with either no evidence of CIN or of only CIN1 has been less clear and often controversial. There is no question that women referred to colposcopy for evaluation of cytology interpreted as HSIL, adenocarcinoma in situ or atypical glandular cells ‘favor high-grade’ continue to be at considerable risk when the colposcopy and biopsy are normal [35, 36]. However, the postcolposcopy risk for women referred for the evaluation of a HPV-positive ASC-US or LSIL and not found to have CIN2/3 has only recently been clarified by the 2-year ALTS follow-up [28]. In this trial, colposcopy was found to be fairly insensitive, as about one third of the total CIN2/3 detected over the course of the trial was not detected at the initial colposcopy [14, 28]. Alternatively, many new incident CIN2/3 lesions arose during the 2-year postcolposcopy follow-up. It is likely that both explanations played a part. Postcolposcopy risk of CIN2/3 was nearly identical for all women irrespective of whether the initial colpobiopsy finding was CIN1 (13% were subsequently found to have CIN2/3), an abnormal transformation zone was biopsied and not found to have CIN (11.7%), or a completely normal colposcopy and no biopsy (11.3%) [28]. LSIL and HPV-positive ASC-US were of nearly equal risk of the subsequent detection of CIN2/3 (27.6 versus 26.7%, respectively) indicating that these two Pap interpretations require similar diligent follow-up. ASCCP and ACOG both recommended that women referred for the evaluation of HPV-positive ASC-US, ASC-H or LSIL and either found to be normal at initial colposcopy or to have CIN1 that is managed expectantly should have either repeat cytology at 6 and 12 months with referral to colposcopy if either Pap is ⱖASC-US, or a single HPV test at 12 months with referral to colposcopy if the HPV test is positive (figs. 1–3) [35, 36]. These recommendations derived primarily from the ALTS finding that 92% of all the CIN2/3 detected over the longitudinal follow-up was predicted by a single HPV test done 12 months following the initial colposcopy. This high rate of detection was achieved by only
Management of ASCUS and LSIL
133
Table 8. The performance of both HPV testing and repeat Pap in the 2-year follow-up of women originally referred for the evaluation of ASCUS or LSIL in the ALTS trial and either not found to have CIN at the original colposcopy or found to have CIN1 and not treated Management strategy
Follow-up by HPV testing At 6 months HPV DNA testing At 12 months HPV DNA testing Follow-up by repeat Pap Repeat cytology at LSIL threshold One Two Three Repeat cytology at ASCUS threshold One Two Three
Sensitivity of detection of subsequent CIN2 or CIN3 (%)
Women that would be positive and referred to colposcopy (%)
90.9 (85.0–95.1) 92.2 (85.7–96.4)
62.4 (59.6–65.1) 55.0 (51.9–57.9)
49.1 (41.4–56.8) 70.5 (63.3–77.7) 77.2 (70.4–83.9)
25.2 (22.9–27.4) 34.6 (32.1–37.1) 38.3 (35.7–40.9)
76.7 (70.2–83.2) 88.0 (82.9–93.1) 95.1 (91.6–98.6)
51.7 (49.1–54.3) 63.6 (61.1–66.1) 70.0 (67.5–72.5)
Figures in parentheses indicate 95% confidence intervals. The sensitivity for CIN2 or CIN3 and percent re-referral to colposcopy of HPV testing at 6 and 12 months is compared with that of one, two and three repeat Paps at approximately 6-month intervals. With permission from Guido et al. [32].
referring the 55% still persistently HPV-positive patients to colposcopy (table 8) [32]. In contrast, the alternative follow-up option of repeating liquid-based cytology at 6 and 12 months referred more (64%) to colposcopy having any abnormal Pap result on repeat (ⱖASC-US) but detected slightly less (88%) of the CIN2/3. Adding a Pap test to the HPV test at the 12-month visit significantly increased referral to colposcopy, while not significantly increasing sensitivity, bolstering the author’s conclusion that the most efficient test for the detection of CIN2/3 postcolposcopy might be an HPV test alone at 12 months. Additional support for HPV testing as the primary follow-up option is supported by evidence that only persistent HPV progresses to CIN3 and that testing for high-risk HPV detects most CIN3 [4, 5, 7, 14, 15]. The risk of detection of CIN2/3 for ASC-H is not nearly as high as it is for HSIL. In ALTS, only 27% of patients referred on the basis of a conventional Pap smear reading of ASC-H were found to have CIN2/3 [25]. Although the 40% detection of high-grade CIN noted from liquid-based cytology was higher, it still does not meet the criteria required for an excisional procedure in the
Cox
134
absence of a CIN2/3 lesion. Therefore, the management of women with ASC-H postcolposcopy who are not found to have CIN2/3 is identical to that recommended for HPV-positive ASC-US and LSIL (fig. 2) [35, 36]. Women negative on two repeat Paps or on a single HPV test are at low risk of having a missed CIN2/3⫹ and may return to routine screening.
Conclusions
Testing for high-risk HPV in the management of women with ASC-US and in the follow-up of women after colposcopy with HPV-positive ASC-US, ASCH and LSIL has significantly reduced problems of subjectivity, misclassification and poor sensitivity of cytology. New cervical screening and triage options that utilize HPV testing are revolutionizing cervical cancer prevention efforts, hopefully decreasing the burden of HPV-induced cervical neoplasia and improving its efficiency [35, 36, 67–69]. HPV testing is objective and highly sensitive for CIN2/3⫹. These positives have clear advantages over repeat cytology in the management of women with equivocal Pap test interpretations. In addition, as HPV testing increasingly assumes the role as an adjunct to cytology in the primary screening of women ⱖ30 years of age, the impact of the subjective nature of cytology will lessen significantly. HPV vaccines targeting HPV16 and HPV18 are likely to become available to the public during 2006 and will have a significant effect on abnormal Pap test rates within a few years after introduction, as cellular changes due to these HPV types will become increasingly uncommon within the vaccinated population. While screening will not change in the near future, colposcopy clinics are likely to see declining referrals as all levels of cytological abnormality become less common. The challenge will be to continue to remain expert in colposcopy and the other aspects of abnormal Pap management in an era of declining abnormalities.
References 1
2 3
4
Bosch FX, Manos MM, Munoz N, Sherman M, Jansen AM, Peto J, Schiffman MH, Moreno V, Kurman R, Shah KV: Prevalence of human papillomavirus in cervical cancer: a worldwide perspective. J Natl Cancer Inst 1995;87:796–802. Bosch FX, Lorincz A, Munoz N, Meijer CJLM, Shah KV: The causal relation between human papillomavirus and cervical cancer. J Clin Pathol 2002;55:244–265. Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, Snijders PJ, Peto J, Meijer CJ, Munoz N: Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189:1–3. Nobbenhuis M, Walboomers JM, Helmerhorst TJ, Rozendaal L, et al: Relation of human papillomavirus status to cervical lesions and consequences for cervical-cancer screening: a prospective study. Lancet 1999;354:20.
Management of ASCUS and LSIL
135
5
6
7
8 9 10
11 12 13 14
15
16
17
18
19
20 21
22 23
24
Jacobs MV, Zielinski D, Meijer CJ, Pol RP, Voorhorst FJ, de Schipper FA, Runsink AP, Snijders PJ, Walboomers JM: A simplified and reliable HPV testing or archival Papanicolaou-stained cervical smears: application to cervical smears from cancer. Br J Cancer 2000;82:1421–1426. Castle PE, Wacholder S, Lorincz AT, Scott DR, Sherman ME, Glass AG, Rush BB, Schussler JE, Schiffman M: A prospective study of high-grade cervical neoplasia risk among human papillomavirus-infected women. J Natl Cancer Inst 2002;94:1406–1414. Solomon D, Schiffman MH, Tarone R: Comparison of three management strategies for patients with atypical squamous cells of undetermined significance: baseline results from a randomized trial. J Natl Cancer Inst 2001;93:293–299. National Cancer Institute Workshop: The 1988 Bethesda system for reporting cervical/vaginal cytologic diagnosis. JAMA 1989;262:931–934. The Bethesda system for reporting cervical/vaginal cytologic diagnoses. Report of the 1991 Bethesda Workshop. Am J Surg Pathol 1992;16:914–916. Solomon D, Davey D, Kurman R, Moriarty A, O’Connor D, Prey M, Raab S, Sherman M, Wilbur D, Wright TC, Young N, The Forum Group Members, The Bethesda 2001 Workshop: The 2001 Bethesda system: terminology for reporting results of cervical cytology. JAMA 2002;287:2114–2119. Selvaggi SM: Reporting of atypical squamous cells cannot exclude a high-grade squamous intraepithelial lesion (ASC-H) on cervical samples: is it significant? Diagn Cytopathol 2003;29:38–41. Kinney WK, Manos MM, Hurley LB, Ransley JE: Where’s the high grade cervical neoplasia? The importance of the minimally abnormal Papanicolaou diagnoses. Obstet Gynecol 1998;91:973–976. Raffle AE, Alden B, MacKenzie EFD: Detection rates for abnormal cervical smears: what are we screening for? Lancet 1995;345:1469–1473. ASCUS-LSIL Triage Study (ALTS) Group: Results of a randomized trial on the management of cytology interpretations of atypical squamous cells of undetermined significance. Am J Obstet Gynecol 2003;188:1383–1392. ASCUS-LSIL Triage Study (ALTS) Group: A randomized trial on the management of low-grade squamous intraepithelial lesion cytology interpretations. Am J Obstet Gynecol 2003;188: 1393–1400. Stoler MH, Schiffman M: Toward optimal laboratory use. Interobserver reproducibility of cervical cytology and histologic interpretations. Realistic estimates from the ASCUS-LSIL Triage Study. JAMA 2001;285:1500–1505. Renshaw AA, Davey DD, Birdsong GG, Walsh M, Styer PE, Mody DR, Solomon D: Precision in gynecologic cytologic interpretation: a study from the College of American Pathologists Interlaboratory Comparison Program in Cervicovaginal Cytology. Arch Pathol Lab Med 2003;127:1413–1420. Scott DR, Hagmar B, Maddox P, Hjerpe A, Dillner J, Cuzick J, Sherman ME, Stoler MH, Kurman RJ, Kiviat NB, Manos MM, Schiffman M: Use of human papillomavirus DNA testing to compare equivocal cervical cytologic interpretations in the United States, Scandinavia, and the United Kingdom. Cancer 2002;96:14–20. Confortini M, Carozzi F, Dalla Palma P, Ghiringhello B, Parisio F, Prandi S, Ronco G, Ciatto S, Montanari G, GISCI Working Group for Cervical Cytology: Interlaboratory reproducibility of atypical squamous cells of undetermined significance report: a national survey. Cytopathology 2003;14:263–268. Cox JT: Management of women with cervical cytology interpreted as ASC-US or ASC-H. Clin Obstet Gynecol 2005;48:160–170. Scheiden R, Wagener C, Knolle U, Dippel W, Capesius C: Atypical squamous cells of undetermined significance: audit and the impact of potential litigation. Retrospective review of 682 cases. Cytopathology 2003;14:257–262. Bergeron C, Jeannel D, Poveda J, Cassonet P, Orth G: Human papillomavirus testing in women with mild cytologic atypia. Obstet Gynecol 2000;95:821–827. Sherman ME, Schiffman M, Cox JT: Effects of age and human papilloma viral load on colposcopy triage: data from the randomized Atypical Squamous Cells of Undetermined Significance/LowGrade Squamous Intraepithelial Lesion Triage Study (ALTS). J Natl Cancer Inst 2002;94:102–107. Flynn K, Rimm DL: Diagnosis of ASCUS in women over age 50 is less likely to be due to dysplasia. Diagn Cytopathol 2001;24:132–136.
Cox
136
25
26
27
28
29
30
31
32
33
34
35
36 37 38
39
40
41 42
Sherman ME, Solomon D, Schiffman M, for the ALTS Group: Qualification of ASCUS. A comparison of equivocal LSIL and equivocal HSIL cervical cytology in the ASCUS LSIL Triage Study. Am J Clin Pathol 2001;116:386–394. Alli PM, Ali SZ: Atypical squamous cells of undetermined significance rule out high-grade squamous intraepithelial lesion: cytopathologic characteristics and clinical correlates. Diagn Cytopathol 2003;28:308–312. Louro AP, Roberson J, Eltoum I, Chhieng DC: Atypical squamous cells cannot exclude high-grade squamous intraepithelial lesion. A follow-up study of conventional and liquid-based preparations in a high-risk population. Am J Clin Pathol 2003;120:392–397. Cox JT, Schiffman M, Solomon D, ASCUS-LSIL Triage Study (ALTS) Group: Prospective follow-up suggests similar risk of subsequent cervical intraepithelial neoplasia grade 2 or 3 among women with cervical intraepithelial neoplasia grade 1 or negative colposcopy and directed biopsy. Am J Obstet Gynecol 2003;188:1406–1412. Cox JT, Lorincz AT, Schiffman MH, Sherman ME, Kurman RJ: HPV testing by hybrid capture appears to be useful in triaging women with a cytologic diagnosis of ASCUS. Am J Obstet Gynecol 1995;172:946–954. Manos MM, Kinney WK, Hurley LB, Sherman ME, Shieh-Ngai J, Kurman RJ, Ransley JE, Fetterman BJ, Hartinger JS, McIntosh KM, Pawlick GF, Hiatt RA: Identifying women with cervical neoplasia: using human papillomavirus DNA testing for equivocal Papanicolaou results. JAMA 1999;281:1605–1610. Wright TC, Lorincz AT, Ferris DG, Richart RM, Ferenczy A, Mielzynska I, Borgatta L: Reflex human papillomavirus deoxyribonucleic acid testing in women with abnormal Pap smears. Am J Obstet Gynecol 1998;178:962–966. Guido R, Schiffman M, Solomon D, Burke L, ASCUS LSIL Triage Study (ALTS) Group: Postcolposcopy management strategies for women referred with low-grade squamous intraepithelial lesions or human papillomavirus DNA-positive atypical squamous cells of undetermined significance: a two-year prospective study. Am J Obstet Gynecol 2003;188:1401–1405. Sherman ME, Wang SS, Wheeler CM, Rich L, Gravitt PE, Tarone R, Schiffman M: Determinants of human papillomavirus load among women with histological cervical intraepithelial neoplasia 3: dominant impact of surrounding low-grade lesions. Cancer Epidemiol Biomarkers Prev 2003;12: 1038–1044. Arbyn M, Buntinx F, Van Ranst M, Paraskevides E, Martin-Hirsch P, Dillner J: Virologic versus cytologic triage of women with equivocal Pap smears: a meta-analysis of the accuracy to detect high-grade intraepithelial neoplasia. J Natl Cancer Inst 2004;96:280–293. Wright TC Jr, Cox JT, Massad LS, Twiggs LB, Wilkinson EJ, ASCCP-Sponsored Consensus Conference: 2001 Consensus Guidelines for the management of women with cervical cytological abnormalities. JAMA 2002;287:2120–2129. ACOG Practice Bulletin. Management of abnormal cervical cytology and histology. Clinical management guidelines for the Obstetrician and Gynecologist, 2005, vol 66. Wright TC, Sun XW, Koulos J: Comparison of management algorithms for the evaluation of women with low grade cytologic abnormalities. Obstet Gynecol 1995;85:202–210. Ferris DG, Wright TC, Litaker MS, Richart RM, Lorincz AT, Sun SW: Triage of women with ASCUS and LSIL on Pap smear reports: management by repeat Pap smear, HPV testing or colposcopy? J Fam Pract 1998;46:125–134. Ferris DG, Wright TC, Litaker MS, Richart RM, Lorincz AT, Sun SW: Comparison of two tests for detection carcinogenic HPV in women with Papanicolaou smear reports of ASCUS or LSIL. J Fam Pract 1998;46:136–141. Shlay JC, Dunn T, Byers T, Baron AE, Douglas J: Prediction of cervical intraepithelial neoplasia grade 2–3 using risk assessment and human papillomavirus testing in women with atypia on Papanicolaou smears. Obstet Gynecol 2000;96:410–416. Goff BA, Muntz HG, Bell DA, Wertheim I, Rice LW: Human papillomavirus typing in patients with Papanicolaou smears showing squamous atypia. Gynecol Oncol 1993;48:384–388. Fait G, Daniel Y, Kupferminc MJ, Lessing JB, Niv J, Bar-Am A: Does typing of human papillomavirus assist in the triage of women with repeated low-grade cervical cytologic abnormalities? Gynecol Oncol 1998;70:319–322.
Management of ASCUS and LSIL
137
43
44
45 46
47
48
49 50
51
52
53
54
55 56 57
58
59
60
61
Fait G, Daniel Y, Kupferminc MJ, Geva E, Ron IG, Lessing JB, Niv J, Bar-Am A: Contribution of human papillomavirus testing by hybrid capture in the triage of women with repeated abnormal Pap smears before colposcopy referral. Gynecol Oncol 2000;79:177–180. Lin CT, Tseng CJ, Lai CH, Hsueh S, Huang HJ, Lay KS: High-risk HPV DNA detection by Hybrid Capture II. An adjunctive test for mildly abnormal cytologic smears in women ⬎ or ⫽ 50 years of age. J Reprod Med 2000;45:345–350. Morin C, Bairati I, Bouchard C, Fortier M, Roy M, Moore L: Managing atypical squamous cells of undetermined significance in Papanicolaou smears. J Reprod Med 2001;46:799–805. Rebello G, Hallam N, Smart G, Farquharson D, McCafferty J: Human papillomavirus testing and the management of women with mildly abnormal cervical smears: an observational study. BMJ 2001;322:893–894. Zielinski DJ, Rozendaal L, Voorhorst FJ, Runsink AP, de Schipper FA, Meijer CJ: High-risk HPV testing in women with borderline and mild dyskaryosis: long-term follow-up data and clinical relevance. J Pathol 2001;195:300–306. Hughes AA, Glazner J, Barton P, Shlay JC: A cost-effectiveness analysis of four management strategies in the determination and follow-up of women with atypical squamous cells of undetermined significance. Diagn Cytopathol 2005;32:125–132. Kim JJ, Wright TC, Goldie SJ: Cost effectiveness of alternative triage strategies for atypical squamous cells of undetermined significance. JAMA 2002;287:2382–2390. Simsir A, Ioffe O, Sun P, Elgert P, Cangiarella J, Levine PH: Effect of Bethesda 2001 on reporting of atypical squamous cells (ASC) with special emphasis on atypical squamous cells-cannot rule out high grade (ASC-H). Diagn Cytopathol 2006;34:62–66. Saad R, Kanborr A, Mauser N, Modery J, Dauds DJ: Atypical squamous cells cannot exclude high grade squamous intraepithelial lesion (ASC-H): diagnostic reproducibility, HPV positivity rates, and clinicial implications (abstract). Mod Pathol 2005;18:68A. Lyman AK, Giampoli EJ, Bonfiglio TA: Should women with atypical squamous cells, cannot exclude high grade squamous intraepithelial lesion receive reflex human papillomavirus HPV testing? Cancer 2005;105:457–460. Shidham VB, Rao RN, Macchi J, Chavan A: Microglandular hyperplasia has a cytomorphological spectrum overlapping with atypical squamous cells-cannot rule out high grade squamous intraepithelial lesion (ASC-H). Diagn Cytopathol 2004;30:57–61. Rowe LR, Aldeen W, Benz JS: Prevalence and typing of high risk HPV DNA by Hybrid Capture II in women with ASCUS, ASC-H, LSIL and AGC in ThinPrep Pap tests. Diagn Cytopathol 2004;30:426–432. Prince TE, Dabbs D, Guido R: The ASC-H dilemma: a new role for human papillomavirus testing (abstract). Obstet Gynecol 2005;105:3S. Solomon D, Schiffman MH: Have we resolved how to triage equivocal cytology? J Natl Cancer Inst 2004;96:250–251. Wu HH, Allan SL, Kirkpatrick J, Elsheik TM: High risk HPV DNA testing by HC II is a useful triage method for women with ASC-H on ThinPrep Pap tests (abstract). Mod Pathol 2005;18: 80A. Srodon M, Dilworth HP, Ronnett RM: Atypical squamous cells cannot exclude high-grade squamous intraepithelial lesion (ASC-H): HPV testing and histologic results (abstract). Mod Pathol 2005:18:78A. Hoschar A, Kink I, Dawson AE: Atypical squamous cells cannot exclude high-grade squamous intraepithelial lesion (ASC-H): correlation of morphologic features, reflex HPV DNA testing and biopsy follow-up. Acta Cytol 2004;48:271–272. Moscicki AB, Shiboski S, Hills NK, Powell KJ, Jay N, Hanson EN, Miller S, Canjura-Clayton KL, Farhat S, Broering JM, Darragh TM: Regression of low-grade squamous intra-epithelial lesions in young women. Lancet 2004;364:1678–1683. Cox JT: Management of atypical squamous cells of undetermined significance and low-grade squamous intra-epithelial lesion by human papillomavirus testing. Best Pract Res Clin Obstet Gynaecol 2001;15:715–741.
Cox
138
62
63 64 65 66 67
68
69
Cox JT, Massad LS, Lonky N, Tosh R, Waxman A, Wilkinson E: ASSCP Practice Guideline: management guidelines for follow-up of low grade squamous intraepithelial lesion (LSIL). J Low Gen Tract Dis 2000;4:83–92. Acs G, Gupta PK, Baloch ZW: Glandular and squamous atypia and intraepithelial lesions in atrophic cervicovaginal smears. One institution’s experience. Acta Cytol 2000;44:611–617. Kaufman RH, Adam E, Icenogle J, Reeves WC: Relevance of human papillomavirus screening in management of cervical intraepithelial neoplasia. Am J Obstet Gynecol 1997;176:87–92. Giles JA, Deery A, Crow J, Walker P: The accuracy of repeat cytology in women with mildly dyskaryotic smears. Brit J Obstet Gynecol 1989;96:1067–1070. Ferris DF, Cox JT, O’Connor D, Wright C, Foerster J: Modern Colposcopy. Dubuque, KendallHunt, 2004, pp 525–578. Wright TC Jr, Cox JT, Massad LS, Carlson J, Twiggs LB, Wilkinson EJ, American Society for Colposcopy and Cervical Pathology: 2001 consensus guidelines for the management of women with cervical intraepithelial neoplasia. Am J Obstet Gynecol 2003;189:295–304. Saslow D, Runowicz CD, Solomon D, Moscicki AB, Smith RA, Eyre HJ, Cohen C, American Cancer Society: American Cancer Society guideline for the early detection of cervical neoplasia and cancer. CA Cancer J Clin 2002;52:342–362. ACOG Practice Bulletin. Cervical Cytology Screening. Washington, American College of Obstetricians and Gynecologists, 2003.
J. Thomas Cox, MD Health Services, University of California Santa Barbara Santa Barbara, CA 93460 (USA) Tel. ⫹1 805 893 4481, Fax ⫹1 805 893 2758 E-Mail
[email protected] Management of ASCUS and LSIL
139
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 140–146
Management of Women with High-Grade Squamous Intraepithelial Lesion and Atypical Glandular Cell Cervical Cytology Thomas C. Wright College of Physicians and Surgeons of Columbia University, New York, N.Y., USA
The management of women with high-grade squamous intraepithelial lesion (HSIL) and atypical glandular cells (AGC) cytological results is less controversial than the management of women with lower grade cytological abnormalities. In large part, this stems from the fact that women with both of these cytological results have a considerable risk of harboring a high-grade cervical intraepithelial neoplasia (CIN2,3) or even an invasive cancer. This chapter outlines the recommendations made by the 2001 Consensus Conference which was sponsored by the American Society of Colposcopy and Cervical Pathology together with 28 other organizations, federal and international agencies, and professional organizations [1].
High-Grade Squamous Intraepithelial Lesions
The 2001 Bethesda System continued the two-tiered classification of squamous cervical neoplasia initially introduced by the 1988 Bethesda System [2, 3]. In this classification, cytological changes corresponding to CIN1 are referred to as low-grade squamous intraepithelial lesions (LSIL), and cytological changes corresponding to CIN2 and CIN3 are combined into a single cytological category referred to as high-grade squamous intraepithelial lesions (HSIL). Women with a cytological result of HSIL have a 70–75% risk of having a high-grade cervical intraepithelial neoplasia (CIN2,3) lesion and a 1–2% chance of having an invasive cervical cancer [4, 5]. HSIL cytology results are
relatively uncommon and according to 2003 data from the College of American Pathologists (CAP) the median HSIL rate of laboratories in the US is only 0.5% [6]. Because HSIL results are relatively uncommon and there is a considerable risk that women with a HSIL result have a high-grade cervical neoplasia, women with HSIL should be referred for colposcopic evaluation with endocervical assessment (if not pregnant). In most instances, high-grade neoplasia is identified on either the cervix (CIN2,3 or AIS) or the vagina (VAIN2,3). If not, the woman must still be considered to be a significant risk for having an unrecognized CIN2,3 lesion since it is now recognized that a single colposcopic examination can miss approximately one-third of CIN2,3 lesions [7, 8]. Therefore women with a referral cytology of HSIL but in whom high-grade neoplasia is not identified require additional evaluation. The type of evaluation that is appropriate depends on whether or not the colposcopic examination is satisfactory, whether the patient is pregnant, and whether or not immediate excisional treatment is considered an option.
Synopsis of 2001 Consensus Guidelines for HSIL Cytology
Initial Evaluation Colposcopy with endocervical sampling is recommended for women with HSIL. Subsequent management depends on whether a lesion is identified, whether the colposcopic examination is satisfactory, whether the patient is pregnant, whether immediate excision is considered acceptable, and the age of the patient. Omission of endocervical sampling is acceptable if a diagnostic excisional procedure is planned. Subsequent Evaluation For women with satisfactory colposcopic examinations it is recommended that when either no lesion or only CIN1 is identified that all cytologic, histologic, and colposcopic results be reviewed. If the diagnosis is not revised or a review is not possible, then a diagnostic excisional procedure is recommended in non-pregnant patients. In young women of reproductive age it is acceptable to follow-up at 4- to 6-month intervals for 1 year using cytology and colposcopy, provided the colposcopic examination is satisfactory. For women with an unsatisfactory colposcopic examination in whom no lesion is identified at colposcopy, it is recommended that all the material be reviewed. If the review does not result in a changed diagnosis, a diagnostic excisional procedure is recommended in all nonpregnant women. Ablative methods should not be utilized in this situation.
Management of HSIL and AGC
141
Pregnant Patients The 2001 Consensus Guidelines recommend that colposcopy for HSIL in pregnant patients be conducted by clinicians who are experienced in the colposcopic changes induced by pregnancy. The biopsy of lesions colposcopically considered to be high-grade or suspicious for cancer is preferred and the biopsy of other lesions is considered acceptable. Endocervical curettage is unacceptable in pregnant women. It should be noted that unsatisfactory coloposcopic examinations frequently become satisfactory as the pregnancy progresses. Therefore, pregnant patients with unsatisfactory colposcopic examinations should undergo repeat colposcopy in 6–12 weeks. Pregnant patients with HSIL who do not have invasive disease identified at the initial colposcopic evaluation should have additional colposcopic and cytologic examinations with biopsy only if the appearance of the lesion worsens or there is a cytological suggestion of invasive cancer. Reevaluation using colposcopy and cytology is recommended no sooner than 6 weeks postpartum [1].
Atypical Glandular Cells
According to data from the College of American Pathologists (CAP), in 2003 the average atypical glandular cell (AGC) rate for laboratories in the US was 0.2% or 2 cases per 1,000 Pap tests [6]. This is considerably lower than the average ASC-US rate which was 3.9%. The 2001 Bethesda System classifies atypical glandular cells (AGC) into three categories: atypical glandular cells, either endocervical, endometrial, or ‘glandular cells’ not otherwise specified (AGC-NOS); atypical endocervical cells or ‘glandular cells’, favor neoplasia (AGC, favor neoplasia); and endocervical adenocarcinoma in situ (AIS). Women with a cytological result of AGC have a somewhat lower risk for having a CIN2,3 lesion than do women with HSIL, but have a considerably increased risk of having invasive cancer. These risks appear to vary as a function of the type of AGC which is diagnosed and the age of the patient. Various studies have documented rates of biopsy-confirmed CIN (of all grades) of 9–54% in women with AGC. The rate of biopsy-confirmed adenocarcinoma in situ ranges from 0 to 8% and the rate of cancer from 1 to 9% in the various studies. The separation of AGC into AGC-NOS and AGC, favor neoplasia was included in the 2001 Bethesda System since these categories appear to be at different risk for having high-grade cervical neoplasia or cancer (table 1) [4, 9–12]. Women with an AGC-NOS result have a 9–41% risk of having a high-grade neoplasia (either glandular or squamous) and a 0–15% risk of having a high-grade glandular lesion (either adenocarcinoma in situ or invasive adenocarcinoma) [4, 9–15]. In contrast, women with an AGC, favor neoplasia result have a 27–96% risk
Wright
142
Table 1. Rates of biopsy-confirmed neoplasia in women with AGC Cytology
Any high-grade lesion
High-grade glandular
AGC-NOS (%) AGC, favor neoplasia (%)
9–41 27–96
0–15 10–93
From [4, 9–15].
Table 2. Findings at work-up of women with AGC Histological findings
n
%
No lesion identified CIN1 CIN2,3 Adenocarcinoma in situ Invasive adenocarcinoma
109 9 12 5 1
80 7 9 4 1
Ronnett et al. [9].
of having high-grade neoplasia and a 10–93% risk of having a high-grade glandular lesion. A cytological result of AIS is associated with a 48–69% risk of biopsyconfirmed AIS and a 38% risk of invasive adenocarcinoma. Because of this differing risk, women with AGC-NOS are managed less aggressively than are women with an AGC, favor neoplasia result or a cytological diagnosis of AIS. As table 1 demonstrates, there is a wide variability between different reports in the prevalence of high-grade neoplasia in women with AGC cytology. This reflects differences in the patient populations that have been studied. For example, reports from centers with busy gynecological oncology services typically report much higher rates of glandular neoplasia and invasive adenocarcinomas than do series produced from more routine screening populations [9, 16]. One of the most representative series is a study in which 46,009 women undergoing routine screening were enrolled from the Kaiser healthcare system (table 2) [9]. Of the 46,009 women a total of 137 (0.3%) had AGC cytology results. When these women were evaluated it was found that 9% had biopsyconfirmed CIN2,3 lesions, 4% had AIS, and one (1%) had an invasive endometrial adenocarcinoma. Almost all series of women with AGC results have reported, somewhat paradoxically, that the single most frequent histopathological abnormality
Management of HSIL and AGC
143
Table 3. Impact of age on rates of neoplasia in women with AGC
Premenopausal (%) Postmenopausal (%)
CIN2,3 or AIS
Endometrial hyperplasia or neoplasia
22–30 6–7
3 19
From [12, 15, 17, 18].
identified in these women is CIN2,3 [9, 16]. Based on the high prevalence of CIN2,3 in women with AGC and the fact that repeat cytological testing is less sensitive for identifying CIN2,3 or AIS in women with AGC, the 2001 Consensus Guidelines recommend that all women with an AGC cytological result be referred for a colposcopic evaluation [1]. Moreover, since it can be quite difficult to identify AIS based on colposcopic appearance alone and AIS lesions can be confined to the endocervical canal, the 2001 Consensus Guidelines recommend that the colposcopic evaluation of women with AGC be accompanied by endocervical sampling. Age is another key indicator for determining both the type and frequency of neoplasia identified in women with AGC. Table 3 provides the impact of menopausal status in women with AGC [12, 15, 17, 18]. Premenopausal patients with AGC are much more likely to have a CIN2,3 lesion or AIS than to have endometrial disease. In contrast, postmenopausal patients are more likely to have endometrial disease including endometrial adenocarcinoma than CIN2,3 or AIS. Based on the relationship between patient’s age or menopausal status and risk for endometrial neoplasia, the 2001 Consensus Guidelines recommend that initial evaluation for all nonpregnant women with ACG be colposcopy with endocervical sampling and that this be accompanied by sampling of the endometrium in women greater than 35 years of age and in all women with AGC with abnormal uterine bleeding. The one exception is postmenopausal women with a cytological result of abnormal endometrial cells who do not require colposcopy. The 2001 Consensus Conference concluded that there was insufficient data to allow an assessment of the role of HPV DNA testing in the management of AGC and AIS [1].
Synopsis of 2001 Consensus Guidelines for AGC Cytology
Initial Evaluation Colposcopy with endocervical sampling is recommended for women with all subcategories of AGC (including AIS) with exception that women with
Wright
144
atypical endometrial cells should initially be evaluated with endometrial sampling. Endometrial sampling should also be performed if a woman is older than 35 years or has unexplained vaginal bleeding. The presence of a co-existing squamous abnormality does not change management. It should also be noted that the 2001 Consensus Guidelines consider initial management of a woman with an AGC or AIS using a program of repeat cytological examination to be unacceptable. Subsequent Evaluation If CIN or AIS is identified at colposcopy in a woman with a cytological result of AGC-NOS, then the patient should be treated according to the appropriate guideline. If no lesion is identified the patient should be followed up with repeat cytological examinations at 6 monthly intervals for 2 years. Patients referred for evaluation of AGC, favor neoplasia or an AIS cytology result who are not found to have an invasive lesion identified at colposcopy should undergo a diagnostic excisional procedure. In these instances a cold-knife conization procedure is preferred because it is more likely to provide diagnostic information on the state of the margins of excision. References 1 2 3 4
5 6
7
8
9
10
Wright TC Jr, et al: 2001 consensus guidelines for the management of women with cervical cytological abnormalities. J Am Med Assoc 2002;287:2120–2129. Luff RD: The Bethesda System for reporting cervical/vaginal cytologic diagnoses: report of the 1991 Bethesda Workshop. Hum Pathol 1992;23:719–721. Solomon D, et al: The 2001 Bethesda System: terminology for reporting results of cervical cytology. JAMA 2002;287:2114–2119. Jones BA, Novis DA: Follow-up of abnormal gynecologic cytology: a college of American pathologists Q-probes study of 16132 cases from 306 laboratories. Arch Pathol Lab Med 2000;124: 665–671. Kinney WK, et al: Where’s the high-grade cervical neoplasia? The importance of minimally abnormal Papanicolaou diagnoses. Obstet Gynecol 1998;91:973–976. Davey DD, et al: Bethesda 2001 implementation and reporting rates: 2003 practices of participants in the College of American Pathologists Interlaboratory Comparison Program in Cervicovaginal Cytology. Arch Pathol Lab Med 2004;128:1224–1229. Guido R, et al: Postcolposcopy management strategies for women referred with low-grade squamous intraepithelial lesions or human papillomavirus DNA-positive atypical squamous cells of undetermined significance: a two-year prospective study. Am J Obstet Gynecol 2003;188: 1401–1405. Cox JT, Schiffman M, Solomon D: Prospective follow-up suggests similar risk of subsequent cervical intraepithelial neoplasia grade 2 or 3 among women with cervical intraepithelial neoplasia grade 1 or negative colposcopy and directed biopsy. Am J Obstet Gynecol 2003;188: 1406–1412. Ronnett BM, et al: Atypical glandular cells of undetermined significance (AGUS): cytopathologic features, histopathologic results, and human papillomavirus DNA detection. Hum Pathol 1999;30: 816–825. Kennedy AW, et al: Results of the clinical evaluation of atypical glandular cells of undetermined significance (AGCUS) detected on cervical cytology screening. Gynecol Oncol 1996;63:14–18.
Management of HSIL and AGC
145
11 12 13 14 15 16 17 18
Duska LR, et al: Clinical evaluation of atypical glandular cells of undetermined significance on cervical cytology. Obstet Gynecol 1998;91:278–282. Zweizig S, et al: Neoplasia associated with atypical glandular cells of undetermined significance on cervical cytology. Gynecol Oncol 1997;65:314–318. Soofer SB, Sidawy MK: Atypical glandular cells of undetermined significance: clinically significant lesions and means of patient follow-up. Cancer 2000;90:207–214. Eddy GL, et al: Papanicolaou smears by the Bethesda system in endometrial malignancy: utility and prognostic importance. Obstet Gynecol 1997;90:999–1003. Chhieng DC, et al: Clinical significance of atypical glandular cells of undetermined significance in postmenopausal women. Cancer 2001;93:1–7. Goff BA, et al: Endocervical glandular atypia in Papanicolaou smears. Obstet Gynecol 1992;79: 101–104. Geier CS, Wilson M, Creasman W: Clinical evaluation of atypical glandular cells of undetermined significance. Am J Obstet Gynecol 2001;184:64–69. Chin AB, et al: The significance of atypical glandular cells on routine cervical cytologic testing in a community-based population. Am J Obstet Gynecol 2000;182:1278–1282.
Dr. Thomas C. Wright, MD College of Physicians and Surgeons of Columbia University 630 West 168th Street New York, NY 10032 (USA) Tel. ⫹1 212 305 3991, Fax ⫹1 212 305 1295, E-Mail
[email protected] Wright
146
Morphological Diagnosis and Treatment Decisions Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 147–156
Liquid-Based Cytology The New Pap Test
Euphemia McGoogan Cytyc, Crawley, UK
It is well accepted that the conventional Pap test (cervical smear test) leads to a marked reduction in the invasive cervical cancer rate and mortality in the population at risk when used within an organised cervical cancer screening programme with high population coverage and regular screening at appropriate intervals. Since the introduction of an organised call/recall programme in the United Kingdom in 1988 there has been a sustained increase in population coverage of the women in the screening age group (20–65 years) to over 80% with a concomitant halving of the invasive cancer rate by the mid-1990s and an accelerated decrease in mortality at a rate of 7% per annum [1]. But even in well-organized, quality-assured programmes with all areas of the screening programme working together as a co-ordinated team, the incidence of and mortality from cervical cancer in the screened population is not eradicated completely. In order to achieve further reductions in the incidence and mortality of cervical cancer, attention has focused on the screening test itself. In the 1990s, many people began to question the accuracy of the Pap smear aware that its efficacy had never been tested in a randomised control trial. Since then, there has been a growing acceptance that the sensitivity of a single Pap test is low (around 50–60%) and it identifies just over half the women who have cervical intraepithelial neoplasia (CIN) at the time of screening [2–4]. The conventional Pap test lacks sensitivity, specificity, reliability and repeatability [5]. However, even today, the general public and most health care professionals fail to recognize the fundamental limitations of the Pap smear as a screening test and expect standards that even most diagnostic tests cannot achieve. The low sensitivity of a single cervical smear is due to a variety of factors including: incorrect or inadequate sampling of the cervix; poor transfer of cells to the glass slide; a non-representative sample placed on the slide; sub-optimal
preparation with uneven cell distribution [6, 7]; poor fixation and only to a lesser extent the microscopic assessment in the laboratory. Only a small proportion of all false-negative tests are due to human error in microscopic assessment in the laboratory [8]. Sampling and preparation together are responsible for about two thirds of false negative tests [3, 9]. Fortunately, it takes on average over 10 years for CIN3 to develop into an invasive cancer during which time a woman may be routinely screened several times overcoming the low sensitivity of a single Pap test. The implementation of new methods is driven by the need to improve these deficiencies in the current screening test. It is recognised that improvements in diagnostic accuracy of the Pap smear need to begin in the doctor’s clinic with improved techniques of specimen collection to allow better quality samples and better slide preparations. If a representative sample is not removed and sent to the laboratory, there is little that laboratory staff can do to improve the sensitivity of the Pap test. Thus, the quality of the sample sent to the laboratory, the presentation for microscopic assessment and the ability of the observer to find abnormal cells all need to be addressed. In addition, serious consideration must be given to the time taken to read a slide since there is a worldwide shortage of trained cytologists.
Liquid-Based Cytology Preparations
Historically, the development of better slide preparations was driven by the limitations of the early automated scanning devices which required a monolayer with little or no cellular overlap. The liquid-based cytology (LBC) technique involves transferring all the material collected on the sampling device into a preservative fluid creating a cell suspension. It is this cell suspension that is sent to the laboratory rather than a glass slide smeared with the cellular material. All of the cellular material collected from the cervix is present in the cell suspension and the cells remain well preserved for several weeks at room temperature [6, 7]. Excess blood and mucus can be removed from the cell suspension and the numbers of leukocytes are diminished. A small randomised aliquot of epithelial cells is deposited in a thin layer on a glass slide. This randomised deposit contains a proportional representation of all the cells removed from the cervix. These preparations result in fewer unsatisfactory test results since the cells are well preserved and clearly visualised. Cytological evaluation and interpretation is facilitated by the thin layer of evenly distributed cells [10]. Abnormal cells are not hidden in thick areas of the slide even when few in number or very small. LBC slides are quicker and easier to screen than conventional Pap slides and can increase the laboratory throughput by up to 40% [5, 11]. To date, three devices have been approved by the USA Food and Drug Administration (FDA) as suitable for preparing LBC cervical scrape samples:
McGoogan
148
The ThinPrep® 2000 Processor (T2000), the more fully automated ThinPrep® 3000 Processor (T3000) both manufactured by Cytyc Corporation (Mass., USA) and the SurePath™ system manufactured by TriPath Imaging Inc. (N.C., USA). The main difference for the smear taker is that ThinPrep requires the sampling device(s) to be rinsed in the preservative collection fluid while the head of the sampling device is avulsed into the vial for the SurePath system. Clinicians must clearly understand which system their laboratory is using since the wrong preparation method invalidates the sample. The ThinPrep T2000 and T3000 collect samples into a methanol-based preservative, PreservCyt, and utilise a simple preparation process of controlled membrane transfer technology with no other pre-processing steps. The slides can be stained using a conventional Pap stain. The T3000 is capable of batch processing 80 samples with walk-away capability. Both produce slides with a 1.9-cm diameter cell deposition area on the glass slide containing around 100,000 cells from an average cellular sample. ThinPrep is FDA labelled for up to 10 equivalent slides per vial. The SurePath system collects samples into an ethanol-based preservative fluid, CytoRich, which completely lyses red blood cells. SurePath utilises a density gradient separation technology with several separate pieces of equipment. It has lengthy pre-processing steps and the PrepStain device is bound to a proprietary Pap stain. The SurePath processes 48 slides per hour and produces a 1.3-cm diameter circle of deposition. It is has a similar number of cells to ThinPrep since these are more densely layered. SurePath is FDA labelled for 1 slide per vial. The multiple manual steps are open to user variability and potential chain of custody error. The general microscopic appearance of a LBC is that of an evenly distributed circular deposit of cells. The margins of the deposit are usually well demarcated and all the cellular material is contained well within the coverslip. Cellular preservation is enhanced and good fixation allows more consistent staining resulting in improved cellular detail under the microscope. Some small sheets and aggregates of cells are retained but, even within groups, most of the nuclei can be assessed microscopically. In ThinPrep slides, the cellular deposit is of a fairly uniform thickness and thus the need for frequent focus changes during microscopic assessment is reduced. SurePath preparations are rather thicker and require constant monitoring through several focus levels. The manufacturer recommends that each SurePath slide is primary screened microscopically two times (i.e. both in north-south and east-west directions). Preparation artefacts that are common in conventional Pap slides such as partial air drying, air bubbles in the mountant or ‘cornflake’ artefact are diminished. Background material such as menstrual or inflammatory exudate, cytolysis, micro-organisms and tumour diathesis can still be identified but it does not obscure the epithelial cells. There
Liquid-Based Cytology
149
is a steep learning curve for assessment of LBC slides even for very experienced laboratory staff and additional training is required. However, there is much anecdotal evidence that LBC is very popular with laboratory staff, most of whom refuse to return to screening conventional Pap slides once trained in LBC. The ThinPrep and SurePath systems are the most widely studied technologies in the literature. There are a number of other devices available which are designed to prepare LBC samples but there is very little or no information about these in the peer-reviewed literature. These include CytoScreen (Seroa), EasyPrep (Labonord), CellPrint (3t), Cyto-Tek (Bayer), PapSpin (Shandon), LiquiPrep (LGM International, Inc.) and MonoPrep (Monogen). While these systems offer alternative methods of processing cell suspensions, good evidence of their ability to consistently deliver an appropriate representative cell sample onto the glass slide is not yet available in the peer-reviewed literature. Pathologists and clinicians must understand that all liquid-based methods are not necessarily equal in their performance, simply because they are ‘liquid based’. This requires potential users to be familiar with the published literature, labelling claims from regulatory bodies, and other studies.
Evidence for Improved Performance of LBC
There are over 100 primary LBC studies in the literature but one must pay attention to the name of the system used and the year of the research study to identify whether it is the current model or a previous model of ThinPrep ( ThinPrep, T2000, T3000) or SurePath (AutoCyte-Prep,  CytoRich) and whether the full system was utilised or just the collection fluid. Prior to 2000, the most frequent study design was a split sample method rather than the intended use which is direct to vial. With this study design, the cervical scrape is first used to make a conventional Pap smear. Then the residual cellular material on the sampler is placed in a vial for LBC. Thus, two specimens are produced for each patient screened – a conventional Pap smear and a LBC and the agreement or difference between the two methods can be compared. These studies are mainly on  models of ThinPrep and SurePath. Since they use only the ‘left over’ material, split sample studies are likely to underestimate any improved performance of the liquid-based method [12]. More recent direct to vial studies [4, 5, 13–16] offer a better evaluation of the methodology. Again these have been almost exclusively on the ThinPrep and SurePath systems. Unsatisfactory Samples The rate of unsatisfactory samples is mentioned in most primary LBC studies. Unfortunately, the criteria used to define adequacy vary between the
McGoogan
150
studies. To date, there is no international definition of adequacy for liquidbased cytology samples and it is inappropriate to use the criteria established for conventional smears. The Bethesda System 2001 provides a definition for adequacy for LBC of more than 5,000 epithelial cells per slide. Even where the Bethesda System was used, it is likely that it was interpreted in different ways since the criteria applied are either not stated or are very subjective and variable. This makes comparative evaluation difficult. However, the majority of studies, both split sample and direct to vial, report that LBC had a larger proportion of specimens classed as totally satisfactory compared to the conventional Pap [4, 5, 13–16]. Sensitivity, Specificity and Accuracy It is difficult to draw clear conclusions from many published articles due to deficiencies in study design. Many studies lack verification of the diagnosis, negative results are not verified and total study numbers are small, often highly selected and carried out on populations with a high incidence of disease. Nonetheless, a fairly consistent finding is an increased detection of abnormal cells in LBC using both ThinPrep and SurePath compared to the conventional Pap [6, 11, 15–17] with ThinPrep tending to deliver a higher sensitivity for HSIL than SurePath. One recent study, however, found no improved detection of HSIL [18]. The English NHSCSP LBC evaluation [14] showed an 80% reduction in the inadequate rate and a reduction from 5.4 to 4.6% in the overall rate of borderline smears. While there was no significant increase in the rates of moderate and severe dyskaryosis when averaged across the three sites, those sites using ThinPrep showed an overall increase in severe dyskaryosis while that using SurePath saw a drop in the detection of severe dyskaryosis. This finding has never been fully explained. The positive predictive value increased indicating improved accuracy in the diagnosis of high-grade cytological abnormalities at all sites. Comparing the running costs of LBC with those of conventional cervical smears was a complex calculation since they utilise differing amounts of laboratory resources. However, Moss et al. [14] concluded there was robust evidence that LBC was a cost-effective alternative to conventional smears especially in terms of life years saved. The Scottish Cervical Screening Programme took a different approach, partconverting four very different laboratories to LBC. The pilots ran for 6 months, included 30,000 routine LBC samples and the results were published by the Scottish Executive Health Department in 2002 [15]. The results showed that the technique was preferred by smear takers and laboratory staff alike. The Scottish pilot report also showed a sharp reduction in the unsatisfactory smear rate. There was a significant improvement in the detection of high-grade lesions between 3 and 9 women per 1,000 tested (almost double) while the borderline (ASCUS) rate
Liquid-Based Cytology
151
was not increased. Reduced workload and increased productivity were also demonstrated in laboratories. Subsequently, the Scottish Cervical Screening Programme has decided to convert to LBC for its routine screening test. There have been many literature reviews and meta-analyses published in an attempt to clarify the position on the sensitivity and accuracy of LBC [4, 12, 19–22]. These authors found that study design was generally poor, with few studies that used adequate methods that allowed comparison of study groups. In particular, there were no randomised controlled trials of LBC. The inclusion criteria, statistical methods and conclusions of these have been contradictory and unhelpful. Unfortunately, they often combine data from  models, split sample and direct to vial studies, screening and high-risk populations and calculations are usually based on a very small number of selected papers thereby lacking appropriate power to show significance of outcome. It is well known that meta-analyses are by no means perfect or represent the final and accurate assessment of a technology or treatment. A detailed editorial published in the New England Journal of Medicine [23] observed that ‘major problems with the implementation of meta-analyses have been common … perhaps most often, overstatements of the strength and precision of the results’. The authors inevitably fall back on demands for large-scale randomised controlled trials (RCT) to obtain a definitive answer. They demand that the RCTs should incorporate colposcopy and biopsy on women with positive results and the histology should be read without knowledge of cytological results as a reference standard. Ideally, they would wish that the colposcopy should also be done without knowledge of the cytology results. However, this might prove extremely difficult to achieve, since it is considered good clinical practice to have the cytology result available to the colposcopists at the time of colposcopy since this greatly increases the accuracy of the colposcopy examination. Furthermore, it is also considered good clinical practice for the cytology result and the colposcopy evaluation to be available to the histopathologist at the time of reporting the biopsy so that deeper levels can be examined if a highgrade squamous lesion or a glandular lesion indicated by the cytology result is not visible on the first histological levels examined. Furthermore, obtaining patient compliance for such RCTs at this stage might be challenging. ThinPrep was approved by the FDA as superior to conventional smears for lower unsatisfactory rates and higher detection rates of LSIL and HSIL in comparison with conventional Pap test in May 1996 for the T2000 and May 2000 for the T3000. The SurePath FDA approval in June 1999 was for improved preparation quality but only equivalent detection of cytological abnormalities. However, in May 2003, the FDA allowed an amendment of increased HSIL detection for the SurePath label. Subsequently, ThinPrep has received FDA approval for improved detection of glandular lesions detection in 2005 and for
McGoogan
152
PreQuot (removal of up to 4 ml from the vial for molecular testing prior to cytology processing) also in 2005. Currently, LBC constitutes over 80% of cervical screening tests in the US. In 2003, the UK National Institute for Clinical Excellence (NICE) recommended to the NHS in England and Wales that LBC should be introduced as the primary means of processing samples in the NHS Cervical Screening Programme [14]. At present, only ThinPrep and SurePath are approved. This followed the Scottish decision in 2002 that LBC should be implemented as the routine screening test throughout Scotland SCSP [15]. All Scottish laboratories opted to use the ThinPrep system and are now fully converted. LBC is also widely used in the private sector in other parts of the world including Western Europe, Australia and New Zealand even although not currently reimbursed in many of these countries. Whatever the sensitivity and specificity of LBC, there are other undoubted advantages, such as reduced reading times, the potential for concurrent HPV DNA testing and the opportunity for computer-assisted screening. Improved distinction between low- and high-grade lesions may be achieved on morphology alone using liquid-based cytology and thus overtreatment of lesions that would regress spontaneously could be avoided. The nature of the evenly dispersed discreet cell deposit on the slide makes LBC eminently suitable for immunocytochemical or immunobiological processes such as in situ hybridisation, e.g. for p16INK4a evaluation, and they are also ideal for automated scanning systems. Since only an aliquot of cell suspension is removed from the vial, the residual cell suspension can be used for additional investigations such as reflex HPV testing that may further improve the sensitivity and specificity of the screening test. Additional investigations for sexually transmitted infections such as Chlamydia and Gonorrhoea can also be carried out on the cell suspensions as can molecular biological techniques and cytogenetic studies. This would prove invaluable should other more specific molecular tests become available in the future to identify women already harbouring or likely to progress to a high-grade lesion that requires treatment. Only the ThinPrep system is currently approved by FDA for HPV testing for using the Digene Hybrid Capture II direct from the vial and for Chlamydia and Gonorrhoea testing using PCR. The ThinPrep vials can be stored at room temperature prior to these tests. If off-label molecular testing from the SurePath vial is planned, the vial and centrifuge tube must be kept refrigerated.
Computer-Assisted Scanning Devices
Screening of Pap smears is a very monotonous, highly skilled, timeconsuming and taxing occupation that places great demands on trained staff
Liquid-Based Cytology
153
who are expected to find even a few abnormal cells lying in a cell population of anywhere from 5,000 to 300,000 cells. It is recognised that the monotony of screening large numbers of Pap smears from normal women promotes periods of lack of attention during which rare abnormal cells may be overlooked. It is therefore not surprising that, for several decades, major efforts have been made to automate this aspect of screening. Computerised scanning devices capable of identifying abnormal slides that are now available, mean that machines can assume some of the burden of primary screening thus reducing the human effort required to perform this onerous task. Initial progress in the development of computer-assisted primary screening devices was slow. Many attempts were made to build systems that would be of practical value in cytology laboratories but without success until, in the mid1990s, two systems were approved by the FDA: AutoPap (Neopath, Seattle, Wash., USA) and PAPNET (Neuromedical Systems, Inc., New York, N.Y., USA). The PAPNET system showed great promise. It analysed conventionally prepared smears with a device that used both traditional computer image technology and neural network software. The system selected and displayed up to 128 images of potentially abnormal cells for each slide. Images were displayed on a monitor with 16 images per screen at low magnification with the option for viewing tiles at a higher power. The slide was referred for full manual screening if the reviewer identified abnormal cells in any tile. Many laboratories throughout Europe and North America were poised to implement it when the company went into receivership in 1999. The AutoPap first received FDA approval in September 1995 as the AutoPap300QC for use in selecting the 10% of slides for quality control by full manual re-screening. This was replaced by the AutoPap Primary Screener which received FDA approval in 1998 for both primary screening and quality control. This system, now manufactured by TriPath, Inc., and called FocalPoint®, classifies up to 25% of batches of routine conventional Pap tests into ‘no further review’ (i.e. within normal limits requiring no manual microscopic assessment) or ‘review’ (requiring full microscopic assessment). The review slides are ranked into quintiles according to the probability that they contain abnormal cells and the cytotechnologist is aware of the ranking when screening the slide. FocalPoint was modified for use with SurePath slides and acquired FDA approval in 2000 for a statistically significant improvement in ASCUS⫹ for routine conventional Pap smears only [24]. FDA approval for high-risk slides and SurePath preparations has still to be granted. Tripath Imaging, Inc., has developed a new version of their system, the FocalPoint GS® but this is not yet FDA approved. The FocalPoint GS uses software to guide the cytotechnologist to the most suspicious areas on the glass slide using a
McGoogan
154
motorised microscope stage. A slide gallery presented on a monitor adjacent to the microscope supports the cytotechnologist. Cytyc Corporation has also developed a computer-assisted system, the ThinPrep Imager®, and this received FDA approval in 2003 for use with all ThinPrep slides including high-risk and screening tests and delivers a statistically significant improvement in ASCUS⫹ sensitivity and HSIL⫹ specificity while doubling the laboratory throughput. A bench-top image processor analyses ThinPrep slides which the cytotechnologist reviews on a microscope with a motorised stage either networked to the image processor or by remote connectivity. Special software takes the reviewer to the 22 most abnormal fields on the slide containing cells selected on the basis of integrated optical density and image analysis data. Full manual review is required only if any of the 22 fields contain a suspicious or abnormal cell. Thus, this is image-directed cytology with dual screening. Computer-assisted screening has the potential to revolutionise cervical screening, because it decreases the fatigue of the user, allows many more slides to be reviewed per day, decreases the screening error rate, and, with appropriate decision support, facilitates identification of morphologic features that may not be not apparent in routine manual review.
References 1
2 3
4 5
6
7 8 9 10
Quinn M, Babb P, Jones J, Allen E: Effect of screening on incidence and mortality from cancer of the cervix in England: evaluation based on routinely collected statistics Br Med J 1999;318: 904–908. Fahey MT, Irwig L, Macaskill P: Meta-analysis of Pap test accuracy. Am J Epidemiol 1995;141: 680–689. McCrory DC, Mather DB, Bastian L, et al: Evaluation of cervical cytology: evidence Report/Technology Assessment, No 5. Rockville, Agency for Health Care Policy and Research, 1999. Payne N, Chilcott J, McGoogan E: Liquid based cytology in cervical screening: a rapid and systematic review Health Technol Assess 2000;4:18. McGoogan E, Reith A: Would monolayers provide more representative samples and improved preparations for cervical screening? Overview and evaluation of systems available. Acta Cytol 1996;40:107–119. Hutchinson ML, Isenstein LM, Goodman A, Hurley AA, Douglass KL, Jui KK, Patten FW, Zahniser DJ: Homogenous sampling accounts for the increased diagnostic accuracy using the ThinPrep processor. Am J Clin Pathol 1994;101:215–219. Laverty CR, Farnsworth A, Thurloe JK, Bowditch RC: The importance of the cell sample in cervical cytology: a controlled trial of a new sampling device. Med J Austr 1989;150:432–436. McGoogan E, Colgan TJ, Ramzy I, et al: Cell preparation methods and criteria for adequacy: IAC Task Force. Acta Cytol 1998;42:25–32. Bergeron C, Debaque H, Ayivi J, Amaizo S, Fagnani F: Cervical smear histories of 585 women with biopsy proven carcinoma in situ. Acta Cytol 1997;41:1676–1680. Austin RM, Ramzy I: Increased detection of epithelial cell abnormalities by liquid-based gynecologic cytology preparations. Acta Cytol 1998;42:178–184.
Liquid-Based Cytology
155
11 12
13 14 15 16 17 18
19
20
21
22
23 24
25
Ferenczy A, Robitaille J, Franco E, et al: Conventional cervical cytologic smears vs. ThinPrep smears: a paired comparison study on cervical cytology. Acta Cytol 1996;40:1136–1142. Arbyn M, Abarca M: Is liquid based cytology an effective alternative for the conventional pap smear to detect cervical cancer precursors? A systematic review and meta-analysis. Brussels, Scientific Institute of Public Health IPH/EPI-REPORTS, 2003, vol 10, pp 1–100. Bergeron C, Bishop J, Lemarie A, Cas F, Ayavi J, Huynh B, Barrasso R: Accuracy of thin-layer cytology in patients undergoing cervical cone biopsy. Acta Cytol 2001;45:519–524. Moss SM, Gray A, Legood R, Henstock E: First report to the Department of Health on evaluation of LBC. 2002. www.cancerscreening.nhs.uk/cervical/lbc-pilot-evaluation.pdf Scottish Cervical Screening Programme (SCSP): Steering group report on the feasibility of introducing liquid based cytology. 2002. www.show.scot.nhs.uk/sehd/pulications McGoogan E: Improved adequacy rates using ThinPrep pap test for routine cytopathology. Cytopathol 1999;10:2. Papillo JL, Zarka MA, St John TL: Evaluation of the ThinPrep Pap test in clinical practice: a seven-month, 16,314-case experience in northern Vermont. Acta Cytol 1998;42:203–208. Coste J, Cochand-Priollet B, de Cremoux P, Le Gales C, Cartier I, Molinie V, Labbe S, VacherLavenu MC, Vielh P: Cross-sectional study of conventional cervical smear, monolayer cytology, and human papillo-mavirus DNA testing for cervical cancer screening. Br Med J 2003;326: 733–737. Abulafia O, Pezzullo JC, Sherer DM: Performance of ThinPrep liquid-based cervical cytology in comparison with conventionally prepared Papanicolaou smears: a quantitative survey. Gynecol Oncol 2003;90:137–144. Bernstein SJ, Sanchez-Ramos L, Ndubisi B: Liquid-based cervical cytologic smear study and conventional Papanicolaou smears: a meta-analysis of prospective studies comparing cytologic diagnosis and sample adequacy. Am J Obstet Gynecol 2001;185:308–317. Klinkhamer PJJM, Meerding WJ, Rosier PFWM, Hanselaar AGJM: Liquid based cervical cytology: a review of the literature with methods of evidence based medicine. Cancer 2003;99: 263–271. Davey E, Barratt A, Irwing L, Chan SF, Macaskill P, Mannes P, Saville AM: Effect of study design and quality on unsatisfactory rates, cytological classifications, and accuracy in liquid-based versus conventional cervical cytology: a systematic review. Lancet 2006;367:122–132. Bailar JC III: The promise and problems of meta-analysis. NEJM 1997;337:559–561. Parker EM, Foti JA, Wilbur DC: FocalPoint slide classification algorithms show robust performance in classification of high-grade lesions on SurePath liquid-based cervical cytology slides. Diagn Cytopathol 2004;30:107–110. Biscotti C, Dawson A, Dziura B, Galup L, Darragh T, Rahemtulla A, Wills-Frank L: Assisted primary screening using the automated ThinPrep imaging system. Am J Clin Pathol 2005;123:281–287.
Dr. Euphemia McGoogan Cytyc, Link 10, Napier Way, Crawley, West Sussex, RH10 9RA (UK) Tel. ⫹44 0 1293 522080, Fax ⫹44 0 1293 528010, E-Mail
[email protected] McGoogan
156
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 157–164
Morphological Diagnosis Histology
Kari J. Syrjänen Department of Oncology and Radiotherapy, Turku University Hospital, Turku, Finland
Because of the fact that any meaningful classification should bear a close correlation with the biological behavior of the lesions classified, the usefulness of all classifications of cervical precancer lesions can only be established by well-controlled prospective follow-up studies. Such studies have a major impact on the understanding of the natural history of cervical cancer and its precursors, as well as important implications in the early detection, diagnosis and treatment of this disease. Indeed, during the past decades, a plethora of such natural history studies has been completed for cervical cancer precursor lesions [1, 2]. The data from all these studies clearly substantiate the view that cervical cancer develops from well-defined precursor lesions; however, during the years, they were named differently. The three nomenclatures include (1) the dysplasia-carcinoma in situ (CIS) terminology, (2) the cervical intraepithelial neoplasia (CIN) classification, and (3) the Bethesda system (TBS), with various modifications. Since 1976, it has been well recognized that HPV infections are associated with precancer lesions and invasive squamous cell carcinomas of the cervix, and high-risk HPV types were recently declared as definitely carcinogenic agents in humans as to the development of cervical squamous cell cancer [3]. However, highly unfortunately, this intimate association of HPV with cervical precancer lesions has complicated the assessment of the natural history of this disease by the traditional follow-up (cohort) studies [1, 2]. Together with the recently introduced TBS, simplifying their grading into two categories only – low-grade squamous intraepithelial lesions (LSIL) and high-grade squamous intraepithelial lesions (HSIL) – the widespread use of screening for the oncogenic HPV types has blurred the traditional concepts of grading cervical cancer precursor lesions into clearly defined categories of mild, moderate and severe.
At extreme, this has raised the question whether or not there is any need for maintaining the intermediate category (CIN2, moderate dysplasia) or just classify the lesions as low-grade (mild dysplasia or CIN1) and high-grade (severe dysplasia, CIN3) categories. As in most cases, there are arguments in favor of both these views [2, 4]. The author (K.J.S.), coming from a European country, is in favor of maintaining the widely accepted CIN terminology, where three grades of lesions are distinguished: CIN1, CIN2 and CIN3. The following arguments are based on the author’s personal experience as a gynecological pathologist examining the morphology of these lesions, as well as on the evidence accumulated from the author’s long-term natural history study on cervical HPV infections (Kuopio Cohort Study 1981–1998) [2]. This evidence implicates two important issues: (1) CIN2 lesions as a morphological entity do exist, and (2) this morphological entity has meaningful biological correlates, i.e. shows a natural history intermediate between that of CIN1 and that of CIN3. A brief account of this evidence will follow, which will be discussed in more detail during the Congress.
CIN2 Is an Independent Morphological Entity
K.J.S. was one of the first in the early 1990s to express a serious concern about the newly introduced TBS, by emphasizing the failure of the new 2-tier classification to reflect the biological behavior of cervical cancer precursors [5]. The key message of that early critical paper was that ‘simplification of the classification into two grades instead of three in CIN terminology inevitably leads to significant loss of diagnostically and prognostically valuable information. The present results firmly advocate preservation of the current descriptive terminology, which has proven its effectiveness for many years’ [5]. The most authoritative support for these views comes from the UK, where the National Health Service Cervical Cancer Screening Programme clearly adopted the policy to classify all cervical pathologies using the CIN terminology. The arguments are presented in detail in the recent document of the Working Party of the Royal College of Pathologists and the National Health Service Cervical Screening Programme (chaired by Prof. Harold Fox) [6]. Apart from giving well-justified arguments favoring the use of CIN terminology in reporting the precursor lesions of the squamous cell carcinoma of the uterine cervix, this valuable document illustrates the morphological criteria for all grades of CIN. As an essential part of the CIN terminology, these arguments are equally valid for substantiating the maintenance of CIN2 as an independent morphological entity, which is also readily favored by the author (K.J.S.) [2, 5].
Syrjänen
158
Model of cervical carcinogenesis Clonal selection Promotion of transformed cells Pap smear normal Reg. 0% Reg. 32%
Reg. 43%
Reg. 57%
Reg. 32%
Latent and subclinical HPV Normal
Mild dysplasia
BIV Moderate dysplasia
CIN1
Severe dysplasia
CIS
CIN2 CIN3
Invasive cancer
LSIL HSIL
Fig. 1. Development of cervical squamous cell carcinoma from its precursor lesions.
According to Fox and Buckley [6] and Fox and Wells [7], the key arguments for the use of CIN are the following: • CIN is a descriptive term and does not involve a value judgment; • the use of three grades underlines the concept of continuity of CIN more than does a two-grade system; • the terms are easy to use and the system is well established and allows for correlation with the cytological grades of dyskaryosis; • the terms allow for continuity of the database and ensure that historical reviews can be undertaken. As further emphasized by Fox and Buckley [6] and Fox and Wells [7], the advantages of the CIN terminology over the previous dysplasia-CIS classification and the subsequent SIL terminology (TBS) [4] are obvious. This is well appreciated by viewing the schematic illustration on the development of squamous cell cancer precursors in figure 1, adapted from the HPV Textbook of K.J.S. [2]. In the CIN classification, the CIN3 category is lumping together
Morphological Diagnosis and Treatment Decisions
159
severe dysplasia and CIS lesions, which is one of its definite advantages; it is certainly easier to differentiate CIN3 from CIN2 than severe dysplasia and CIS. Usually, no reproducibility problems arise in making the distinction between CIN1 and CIN3, but greater difficulty is experienced in differentiating CIN1 and the mildest forms of CIN2, on the one hand, and the most severe forms of CIN2 and CIN3, on the other hand, as can easily be seen in figure 1. These reproducibility problems inherent to CIN classification are not adequately solved; however, any of the subsequent attempts to classify these precursor lesions into two categories only (LSIL and HSIL), e.g., by TBS and its modifications [4], is also depicted in figure 1. Indeed, the major problems of the 2-tier system are clearly understood by viewing this simple illustration. Most importantly, setting the cutoff line between LSIL and HSIL, with the latter comprising CIN2 and CIN3 lesions, leads to inevitable failure to correlate the lesion grade with their biological behavior [2, 7]. Similarly, no distinction can be made on light microscopy between the lesions at the highest boundary of the LSIL category and the mildest forms (low boundary) of HSIL lesions. As will be discussed later, this morphological overlapping is the feasible explanation why the spontaneous regression rates of LSIL and HSIL in our experiment did not deviate from each other more than 20% [5]. If true high-grade lesions should be distinguished, the cutoff line should be moved far more to the right (fig. 1), i.e. between the CIN2 and CIN3. However, unfortunately, even such a trick would not solve the inherent problem of this 2-tier system in adequately separating the ‘high-grade’ and ‘low-grade’ lesions, either morphologically or on the basis of their biological behavior [2, 5]. There is always a morphological overlap between the most severe forms of the lowgrade lesions (CIN2 on the right) and the mildest spectrum of high-grade lesions (CIN3 on the left). In this case, the progression rate of these low-grade lesions would approach that of high-grade lesions, i.e. by increasing from 8.5 towards 40% [5], as discussed later.
CIN2 Has a Biological Behavior Intermediate between CIN1 and CIN3
Despite the constraints associated with this type of study design, a large number of prospective cohort studies have been conducted long before the association of HPV and cervical precancer lesions was appreciated. Such studies have been of crucial importance in our understanding of the premalignant character of these precursor lesions, with true progressive potential if not properly detected and treated [1]. In turn, these data have had a tremendous impact on the practices of early detection by organized screening programs designed in
Syrjänen
160
Table 1. Comparison of the natural history of CIN and cervical HPV lesions Lesion
Regression
Literature (Östör review) CIN1 CIN2 CIN3
57 43 32
Kuopio Cohort Study 1981–1998 HPV NCIN HPV CIN1 HPV CIN2 HPV CIN3
79.9 65.1 58.6 11.6
Persistence
32 35 ⬍56 14.6 20.8 18.6 9.3
Progression to CIS
Progression to invasive cancer
11 22 –
1 5 ⬎12
5.2 14.2 21.4 79.11
0 0 0 0.52
Data are indicated as percentages. HPV NCIN ⫽ HPV lesion without concomitant CIN. Table compiled from the data reviewed by Östör [1] and from the data of the Kuopio Prospective Cohort Study (1981–1998). 1 Progression based on colposcopy (lesion severity and extent) and histology (progress from severe dysplasia to CIS in two subsequent biopsies). 2 One lesion progressed to invasive cancer in less than 3 years [2].
different countries [8], resulting in a significant reduction in the incidence and mortality of cervical cancer. The prospective cohort studies reported until 1993 were subjected to an extensive critical review by Östör [1]. This impressive paper is an excellent source of information, subsequently complemented by a number of more recent reviews [2]. Apart from a purely morphological assessment, these natural history studies are clearly another line of evidence substantiating the maintenance of CIN2 as an independent entity separate from CIN1 and CIN3. As emphasized before, any meaningful classification should bear a close correlation to the biological behavior of the respective lesions [1, 2, 5]. Thus, the value of any new classification (including TBS) can only be established by carefully conducted prospective follow-up studies, capable of establishing the natural history of individual categories of lesions, irrespective which names are used. Indeed, this has been clearly demonstrated for all categories of CIN lesions, including CIN2, as shown by the data published in the literature [1]. By that time, 12 studies from 1955 to 1990 were published, in which CIN2 was the baseline histology at entry to prospective follow-up, covering a total of 2,247 patients. The rates for regression, persistence and progression for CIN2 lesions are 43, 35 and 22%, respectively [1]. As compared with the natural history data of CIN1 and CIN3 lesions (as summarized in table 1), there is no doubt that the biological behavior of CIN2 lesions is intermediate between that of CIN1 and that of CIN3. As such, this is
Morphological Diagnosis and Treatment Decisions
161
probably the single most important argument favoring the concept that CIN2 is a biological entity distinct from CIN1 and CIN3 [2, 6, 7]. Abandonment of CIN2 in the classification inevitably leads to a significant loss of biologically meaningful information, as clearly demonstrated in our study of 1992 [5]. Indeed, we made an experiment to test the biological validity of the 2-tier grading by reclassifying the cervical lesions of our follow-up patients into two categories (LSIL and HSIL) instead of the CIN classification normally used. Using this approach, over 40% of our HSIL did regress spontaneously during the follow-up, as did 64% of the LSIL. We concluded that omitting the CIN2 category leads to a significant loss of prognostically valuable information [5]. The author (K.J.S.) feels that this conclusion is still equally valid today as it was in 1992.
Cervical HPV Infections and CIN Share a Common Natural History
The recognition of the association between HPV and CIN has further complicated the assessment of the natural history of cervical precancer lesions, and the scene is rendered more complex by the recognition of subclinical and latent HPV infections. However, it is evident that the natural history of cervical HPV infections can be adequately approached only by prospective follow-up studies, and a number of such studies have been completed since the association of HPV and CIN was established [2, 7]. One of the first, and until today the most extensive of such prospective follow-up studies on cervical HPV infections was run by K.J.S. in Finland between 1981 and 1998 [2]. Table 1 gives the summary of that long-term natural history data stratified according to HPV lesions with different grades of CIN, as compared with the data from the literature reviewed by Östör [1]. The single most important message of these data is that the natural history of CIN and cervical HPV infections is identical. Indeed, the figures among individual categories are in many respects amazingly similar, almost identical. This applies particularly to the figures of the lower-grade lesions CIN1 and CIN2, which are practically identical for both regression and progression.
If Abandoned, CIN2 Should Be Combined with CIN1 in the Low-Grade Category
In table 1, the progression rate of CIN2 (from the literature) and that of our HPV CIN2 is 22 and 21.4%, respectively. Clearly, CIN2 is an intermediate category with the progression rate approximately two times higher than that of
Syrjänen
162
CIN1, and a regression clearly less than in CIN1. However, in our cohort, HPV CIN2 lesions are much closer to HPV CIN1 in their progression rate, and particularly in their regression rate (65.1 and 58.6%, respectively) (table 1). This clearly implicates the low-grade character of HPV CIN2 lesions, rather than advocating a categorization among the high-grade lesions, as done in the TBS [4]. As discussed above, clumping CIN2 with CIN3 into HSIL leads to a category among which 40% of the lesions regress spontaneously [5]. For the author (K.J.S.), it is difficult to accept any category of lesions being called high-grade, if the spontaneous regression rate is ⬎40%. If such a combination of CIN2 and CIN1 is considered, the histological criteria need to be revised as well, because of the difficulties discussed above in making a distinction between the upper boundary of CIN2 and the lower boundary of CIN3 (also evident from figure 1). Most probably, the best distinction between low-grade and high-grade lesions could be obtained by setting the cutoff in the middle of the CIN2 lesions, i.e. in cases where 50% of the epithelial thickness is occupied by the transformed (basaloid-type) cells, and the upper half still showing normal differentiation. This division would morphologically split the CIN2 category equally between CIN1 and CIN3 of the current classification. It might also effectively distinguish between two categories of lesions with a clearly different spontaneous regression rate (possibly 55–60% versus 25–30% for low-grade and high-grade lesions, respectively) and progression rate (possibly 15% versus 40–50%), provided that the most benign lesions of the spectrum (HPV without CIN) are included among the low-grade category as currently done by the TBS [4]. Whether this concept holds true could be adequately tested in a carefully designed prospective cohort study, testing different optional criteria of dividing CIN into low-grade and high-grade categories.
Conclusions
Because there is a well established morphological and biological category of intermediate-grade lesions (CIN2), the currently used two-grade classifications cannot (1) adequately describe the morphology, and even more importantly, (2) accurately predict the natural history of cervical precancer lesions. It is clearly inaccurate to define any lesions as high-grade if ⬎40% of these regress spontaneously. Two obvious solutions are (1) to be satisfied with the CIN classification as it now stands, which, despite its known shortages, remains to be the most satisfactory descriptive terminology available [6, 7], or (2) to make an attempt to design a two-grade classification, based on newly defined histological criteria. In the latter case, testing the biological validity of such a new classification
Morphological Diagnosis and Treatment Decisions
163
would be of utmost importance. However, this does not mean using an adjunct test (like HPV typing) to modify the histological diagnosis (even if imperfect), but as always, a histological (and cytological) classification must be based exclusively on morphological criteria also in the future. Under current circumstances, to obtain the maximum clinically relevant information from the cervical biopsy, the use of the CIN2 category is justified, because it denotes an abundant category of cervical cancer precursors, with progressive potential intermediate between the low-grade (CIN1) and highgrade (CIN3) lesions.
References 1 2 3 4 5
6 7 8
Östor AG: Natural history of cervical intraepithelial neoplasia – a critical review. Int J Gynecol Pathol 1993;12:186–192. Syrjänen K: Natural history of cervical HPV infections and CIN; in Syrjänen K, Syrjänen S: Papillomavirus Infections in Human Pathology. New York, Wiley and Sons, 2000, pp 142–166. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. IARC, Lyon, 1995, vol 64: Papillomavirus, pp 1–409. Rubin SC, Hoskins WJ (eds): Cervical Cancer and Preinvasive Neoplasia. Philadelphia, Lippincott-Raven, 1996, pp 77–92. Syrjänen K, Kataja V, Yliskoski M, Chang F, Syrjänen S, Saarikoski S: Natural history of cervical HPV lesions does not substantiate the biologic relevance of the Bethesda system. Obstet Gynecol 1992;79:675–682. Fox H, Buckley CH (eds): Histopathology Reporting in Cervical Screening. Sheffield, NHSCSP Publications No 10, 1999, pp 1–56. Fox H, Wells M (eds): Haines and Taylor Obstetrical and Gynaecological Pathology, ed 5. Edinburgh, Churchill Livingstone, 2003, vol 1, pp 297–338. Syrjänen K: Early detection of CIN and HPV: prevention of cervical cancer; in Syrjänen K, Syrjänen S: Papillomavirus Infections in Human Pathology. New York, Wiley and Sons, 2000, pp 167–188.
Prof. Kari Syrjänen, MD, PhD, FIAC Department of Oncology and Radiotherapy, Turku University Hospital Savitehtaankatu 1, Room B 151 FIN–20521 Turku (Finland) Tel. ⫹358 2 3131834, Fax ⫹358 2 3132809, E-Mail
[email protected] Syrjänen
164
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 165–177
Morphological Diagnosis and Treatment Decisions Colposcopy
Albert Singer Department of Obstetrics and Gynaecology, Whittington and Royal Northern Hospital, London, UK
Hinselmann published his first paper about colposcopy in 1925 – he designed the first colposcope and described the colposcopic features of almost every benign and malignant lesion seen on the cervix. Although his work was not fully recognised during his lifetime, in the late 1940s and 1950s, clinicians and investigators in Germany, Central and Eastern Europe and South America gave him due credit for his contribution to the understanding of the pattern and the morphogenesis of cervical premalignant disease and early malignancy. It was not until the 1960s and 1970s that colposcopy became established in the English-speaking countries.
The Colposcope
The binocular colposcopes in use today give a stereoscopic magnification of between 6 and 40 times and provide a transition from macroscopic to microscopic vision. The different epithelial abnormalities (benign, premalignant and malignant) can be studied in vivo, and with experience, it is possible to forecast the histological diagnosis with reasonable accuracy. The most common magnifications used are ⫻10 and ⫻16. The patient should be placed in a modified lithotomy position, and for the comfort of the patient and the colposcopist, it is important to have an examination table which has been designed for colposcopic use.
Fig. 1. Acetic acid has been applied to an atypical (abnormal) transformation zone containing white epithelium. Biopsy shows this tissue to be composed of CIN3 epithelium. Reproduced with permission from J. Jordan/A. Singer (eds.): The Cervix, ed 2. Oxford, Blackwell Publishing Company, 2006.
Methods of Tissue Recognition
The Acetic Acid Technique The traditional method of colposcopy relies on the application of 3–5% acetic acid, following which premalignant disease appears white (referred to as acetowhite) and the sub-epithelial angioarchitecture becomes more prominent. Unfortunately, everything which is acetowhite is not premalignant. Acetic acid causes the tissue, especially columnar and abnormal epithelia, to become oedematous, with the former adopting a white or opaque appearance which is then quite easily distinguishable from normal epithelium which appears pink. The acetic acid acts by causing coagulation of epithelial and stromal cytokeratins in a reversible event. Why the atypical transformation zone epithelium appears white (figs. 1, 2) after the application of 3–5% aqueous acetic acid is still debatable. One explanation could be that the acetic acid is a frequent component of tissue fixatives, and thereby, will rapidly penetrate through the tissue with an effect upon the nucleus of a cell. In this situation, it may precipitate nucleoproteins. The application of acetic acid to areas of atypical epithelium containing cervical intraepithelial
Singer
166
Fig. 2. The white epithelium on the lower or posterior lip is associated with a severe form of premalignancy (CIN3). The upper extent of this epithelium within the endocervix is not seen and the colposcopy in this situation is deemed to be unsatisfactory. Reproduced with permission from J. Jordan/A. Singer (eds.): The Cervix, ed 2. Oxford, Blackwell Publishing Company, 2006.
neoplasia (CIN) causes the precipitated nucleoproteins within the neoplastic cells to be affected, and therefore, obscure the underlying vessels. Consequently, the light is reflected and the epithelium appears white/acetowhite. With low-grade CIN, the acetic acid must penetrate into the lower half of the epithelium and the onset of white is delayed, but with high-grade or full-thickness CIN, it will give an almost instant response and will appear markedly white. The effect is slowly reversed because the acid is buffered and nucleoprotein is no longer precipitated. The appearances are not unique to neoplasia but will be seen on other occasions where increased nucleoprotein is present, as during the process of immature metaplasia formation, healing of the epithelium and with the presence of virus or viral products, such as in clinical HPV changes in the cervical epithelium and/or the presence of condyloma. The Saline Technique The use of saline rather than acetic acid was popularised by Koller [1] and Kolstad and Stafl [2]. This technique relies on studying the sub-epithelial angioarchitecture, as a result of which both authors showed that the degree of premalignant disease could be diagnosed with much more accuracy than with the technique using acetic acid (fig. 3). However, the saline technique is more
Colposcopic Diagnosis
167
Fig. 3. The application of saline to a severe premalignant lesion (CIN3 – microinvasive carcinoma) accentuates the abnormalities within the shape of blood vessels which present as black structures against a white background. Reproduced with permission from J. Jordan/ A. Singer (eds.): The Cervix, ed 2. Oxford, Blackwell Publishing Company, 2006.
difficult to learn and the beginning colposcopist should always use the acetic acid technique. Schiller’s Iodine or Lugol’s Solution Application of Schiller’s or Lugol’s solution as advocated by the German school is largely unnecessary, but on the other hand, does occasionally show areas of abnormality, particularly low-grade abnormality, which would otherwise not be apparent – for this reason, beginning colposcopists may find it valuable. The technique relies on Schiller’s observation that normal squamous epithelium is rich in glycogen, whereas squamous intraepithelial neoplasia is deficient in glycogen; glycogen absorbs iodine and so normal glycogen-containing squamous epithelium takes up the glycogen and stains dark brown/black, whereas the glycogen-free abnormal epithelium does not stain. The Schiller test is designed to detect CIN; therefore, a Schiller positive test identifies an area of the cervix which is non-staining with iodine and a Schiller negative test identifies an area of normal squamous epithelium (fig. 4). Unfortunately, while the test is sensitive, its specificity is low since some non-premalignant disease, especially metaplasia, may be Schiller positive (non-staining).
Singer
168
Fig. 4. The application of Lugol’s or Schiller’s solution has highlighted the presence of a central area of abnormal epithelium which has not taken up the stain because of its glycogenfree characteristic. Reproduced with permission from J. Jordan/A. Singer (eds.): The Cervix, ed 2. Oxford, Blackwell Publishing Company, 2006.
Diagnostic Procedures
Colposcopic Biopsy Colposcopic vision allows biopsies to be taken from the location within the transformation zone with the most severe changes in order that histological confirmation of the degree of severity of the neoplastic process can be obtained to aid the diagnosis. Since colposcopy is a procedure during which cancer must be ruled out, it is a standard practice by most colposcopists to obtain a histological sample, not only from the ectocervix but also from tissue existing within the endocervical canal (if the new squamo-columnar junction cannot be seen), and thus, the entire transformation zone cannot be examined. In such a situation, the colposcopy is deemed to be unsatisfactory. Value of Endocervical Curettage A number of authors, especially those in Europe, are averse to the use of the sharp spoon-shaped or curved curette for endocervical curettage, arguing that in many studies, unsatisfactory material has been obtained from the canal, making an accurate diagnosis impossible. Therefore, no indication of the depth of involvement of the neoplastic tissue within the stroma, such as occurs with microinvasive cancer, can be given. A second objection is that it is a painful procedure being carried out without anaesthetic.
Colposcopic Diagnosis
169
Diagnostic Criteria
There is a widely held perception that colposcopy is nothing more than the recognition of acetowhite epithelium or epithelium which is non-staining following the application of Schiller’s iodine. However, this misconception leads to poor colposcopy, poor assessment of women with abnormal cervical cytology and both over- and under-treatment. There is more to colposcopy than this – the colposcopist should learn to recognize and identify easily observable features [2]. These are: (1) vascular patterns (fig. 3), (2) intercapillary distance, (3) colour tone at the junction of normal and abnormal tissue, (4) surface contour, and (5) a sharp line of demarcation between different types of epithelium. The terminology used in describing the various morphological changes within the cervical epithelium has evolved over many years. Many of these qualitative descriptions have been quantified as to the degree of abnormality, and such a scoring system as described by Reid and Scalzi [3] is used by many colposcopists to grade abnormal squamous epithelial areas. Recently, the Federation of Gynecology and Obstetrics and the International Federation for Cervical Pathology and Colposcopy recommended a revised nomenclature for colposcopic findings.
Vascular Patterns
The vascular pattern is best observed at magnifications of ⫻10 to ⫻25. Following the use of either normal saline or acetic acid, a green filter will allow a very accurate assessment and evaluation of the vascular changes and of the colour tone between normal and abnormal tissue. The arrangement and the distance between the terminal vessels observed within an abnormal colposcopic area can best be assessed by comparing these vessels with those of the adjacent normal squamous epithelium. Intercapillary Distance The capillaries of the original squamous epithelium are characterised by a regular and dense pattern, while those of preinvasive and frank invasive lesions show striking irregularities and variation in their spatial distribution, frequently with an increased distance between adjacent terminal vessels. The more the lesion moves from low- to high-grade premalignant disease to early invasive or frank invasive carcinoma, the more widely spaced the terminal vessels become.
Singer
170
Colour Tone Different lesions may show different colours varying from white, light yellow, yellow red to deep red. For a correct interpretation of the colposcopic image, the most important single factor is the contrast in tone of the abnormal epithelium relative to the adjacent normal epithelium, best seen when the tissues are viewed with a green filter following application of normal saline. The contrast is also clearly seen following the application of acetic acid. Following the application of saline, the CIN, particularly high-grade CIN, appears darker than the original squamous epithelium. Metaplastic epithelium, on the other hand, is whiter and somewhat opaque. Invasive cancer is also whitish, often with a glazed or gelatinous appearance. Following the application of acetic acid, CIN appears white, and although the vessels are more prominent, they are not as prominent or as easy to assess as they are following the use of saline and a green filter. Surface Contour The stereoscopic view provided by the colposcope makes it easier to study the surface contours of the different lesions. Original squamous epithelium has a smooth surface while columnar epithelium is easier to recognise having the appearance of grape-like villi. Preinvasive lesions often have an uneven slightly elevated surface while invasive cancer is characterised by a nodular or polypoid surface with an exophytic or ulcerated growth pattern. Epithelial Borders The last criterion which can easily be studied with the colposcope is the border between lesions and the adjacent normal tissue. The demarcation between CIN and the original squamous epithelium is usually sharp. In contrast, the borderline between normal (original squamous epithelium) and inflammatory lesions or CIN1 may be diffuse.
Colposcopic Terminology
The International Federation for Cervical Pathology and Colposcopy approved a revised colposcopic classification and basic colposcopic terminology in 2002 [4]. The new terminology has the following features: (1) it is descriptive, thereby allowing colposcopists throughout the world to be able to describe lesions to each other and to undertake important collaborative research;
Colposcopic Diagnosis
171
(2) the nomenclature was written in such a way that it can guide a colposcopist in training but can also aid the established colposcopist during the diagnostic process; (3) the terminology is pragmatic and includes a description of the three types of transformation zone – it was felt that this would lead to a more rational triage for the most appropriate treatment for women with abnormal transformation zones.
Unsatisfactory Colposcopy
An unsatisfactory colposcopy examination occurs when the squamocolumnar junction cannot be visualized (figs. 5, 6). It may also occur if associated trauma, inflammation or atrophy preclude a full colposcopic assessment, or when the cervix is not visible.
Uses of Colposcopy
Colposcopy can be used in two circumstances, namely for primary screening and for the diagnosis of women with abnormal cytology and/or clinical suspicion of malignancy. Primary Screening Many clinicians in Latin American countries and Europe use colposcopy routinely as part of the standard gynaecological examination [5, 6]. In these circumstances, cytology usually accompanies colposcopy, and the argument behind this is that the combined testing procedure will decrease the false-negative and positive cytology, thereby reducing the number of women recalled for cytology. Collection of a cytology specimen is also helped by colposcopy. However, in a number of countries, e.g. Germany, there does not seem to be any advantage in using the colposcope to increase the accuracy of cytology. This has been shown to be the case in the recent HART study [7, 8] in which the sensitivity of cytology taken by German clinicians is only 39%. Additionally, recent studies have shown the inability to detect with colposcopy endocervical lesions used in this screening capacity [9, 10]. Abnormal Cytology and/or Suspected Malignancy Colposcopy is used mainly in the English-speaking world to diagnose those women who have been referred with abnormal cytology or those in whom there is a clinical suspicion of malignancy [5, 6].
Singer
172
Fig. 5. Very thick acetowhite epithelium extends high into the endocervix, and its upper extent cannot be visualised. This colposcopic examination is deemed unsatisfactory, as the observer cannot exclude the presence of invasive cancer in the area that is invisible to examination. Reproduced with permission from J. Jordan/A. Singer (eds.): The Cervix, ed 2. Oxford, Blackwell Publishing Company, 2006.
Fig. 6. An area of very atypical epithelium on the upper or anterior cervical lip extends into the endocervix. This is an unsatisfactory colposcopic examination, as the likely presence of early invasive cancer high in the endocervix cannot be excluded because the upper extent of the atypical (abnormal) epithelium cannot be visualised. An excisional procedure is mandatory. Reproduced with permission from J. Jordan/A. Singer (eds.): The Cervix, ed 2. Oxford, Blackwell Publishing Company, 2006. Colposcopic Diagnosis
173
How Accurate Is Colposcopy and How Reproducible Are Its Findings?
Assessing the sensitivity and specificity of colposcopy and directed biopsies is extremely susceptible to bias. The major problem is that the colposcopist’s impression is usually verified by using the reference standards which are essentially based on histology. It is obvious that the colposcopic impression dictates where the biopsy and subsequent histology are obtained. This leads to an inflated estimate of the accuracy of colposcopy. Compared with studies in which colposcopy is employed for primary screening (with or without cytology), studies that assess colposcopy done as a diagnostic procedure are on women who are referred with abnormal cytology (as found at screening) and thus have a higher probability of and possibly a more severe spectrum of cervical pathology. Additionally, women with more pronounced findings and disease are detected more readily by screening, and the performance of colposcopy in a diagnostic capacity may exceed its accuracy and reproducibility when it is used as a screening tool. Studies of Diagnostic Colposcopy Mitchell et al. [11] undertook a meta-analysis of the accuracy of diagnostic colposcopy in those women referred with abnormal cytology in a review of 86 articles published over a 26-year period, beginning in 1960. Strict inclusion criteria were invoked, and only 8 articles were eligible for consideration in this meta-analysis. Compared with histology (more than CIN2), the cut-off of normal neoplasia versus neoplasia on colposcopy had an average weighted sensitivity, specificity and area under the receiver operating characteristic curve of 96, 48 and 80%, respectively. The cut-off of normal and low-grade versus highgrade lesions and cancer on colposcopy is associated with values of 85, 69 and 82%, respectively. Mitchell et al. [11] suggested that independent of prevalence and compared with low-grade lesions, high-grade lesions and cancer were more accurately diagnosed. A second study by Olaniyan [12] looked at 8 studies, 7 of which had been included in the previous meta-analysis of Mitchell et al. [11]. Results were similar in both studies. Studies of Screening Colposcopy In the Belinson study [10] described above, a cross-sectional analysis of nearly 2,000 unscreened Chinese women between 35 and 45 years of age was undertaken, all of whom were assessed first by visual inspection with acetic acid performed by a gynaecologist, and then a second gynaecologist who was blinded to the results obtained by the visual inspection with acetic acid performed colposcopy with biopsy in all abnormal areas. All women had a biopsy from each of the four quadrants with an endocervical curettage in order to
Singer
174
estimate the performance of colposcopy in a screening setting. The sensitivity and specificity of colposcopy and directed biopsy for high-grade CIN or cancer were 81 and 77% in comparison with the combined histological findings from the directed four quadrant and endocervical specimens. In a study undertaken by Schneider et al. [13] in Germany, 4,700 women between 18 and 70 years old were screened by conventional cytology obtained under colposcopic vision with associated HPV DNA testing. Endocervical curettage and biopsy were done where appropriate. The sensitivity and specificity of screening colposcopy for the detection of CIN2 or worse, which was also corrected for verification bias, was only 13.3 and 99.3%, respectively. Reproducibility of Colposcopy Problems have always existed in respect of observer agreement using colposcopy. Three expert colposcopists described in a study by Li et al. [14] showed poor to good intra-observer and inter-observer agreement at assessing two major characteristics, namely borders and colour of acetowhitening. They showed that in respect of border characteristics, the range of inter-observer kappa values was 0.13–0.41, with the intra-observer kappa being 0.26–0.58. In relation to the colour of acetowhitening, inter-observer kappa range was from 0.21 to 0.47 and the intra-observer kappa from 0.34 to 0.75. There was an excellent agreement as to the site of the lesion from where a biopsy should have been obtained (raw agreement 95.3%, 143–150). Assuming that colposcopists use the same definitions, reproducibility of colposcopic assessment depends in part on colposcopists using similar thresholds for categorising findings as to normality versus abnormality and grade.
Potential Side Effects of Colposcopy
The actual colposcopic examination is associated with some discomfort during the insertion of a vaginal speculum and this can be increased if a punch biopsy is performed, especially if local anaesthetic is not used. Psychological morbidity has to be considered. It would seem as though anxiety is elevated, especially prior to surgery and also at the immediate post-operative visit [15–17]. Educational booklets and counselling are highly effective in reducing anxiety [18–20].
How Can Colposcopy Be Improved?
Mention has already been made of quality control aspects of the colposcopy programme. Likewise, adherence to guidelines laid down by various
Colposcopic Diagnosis
175
societies [21, 22] can also lead to improvement in service delivered. However, it would seem as though teaching of colposcopy is also important and this is likewise discussed by Redman et al. [23]. Additionally, a recently described dynamic spectral imaging system has been developed which may dramatically improve the accuracy of colposcopy. Improved Training and Teaching Many centres and countries have their own defined training programmes, but these vary enormously. To help standardise the quality of colposcopy training, the European Federation for Colposcopy has agreed with all 26 member countries on fifty-one core components which form the basis of colposcopy. Conclusions
Colposcopy is simply a means of examining the cervix and upper vagina with low-power magnification (⫻6 to ⫻40.) In most countries, it is used to assess women who have been found to have abnormal cervical cytology, while in other countries, it is used as part of a routine gynaecological examination. In current times, its main use is in the assessment of abnormal cervical cytology, thereby allowing the colposcopist to identify and confirm the extent and characteristic of the abnormal (atypical) epithelium. Once the abnormality has been identified, then the possibility exists for its removal which is usually quite simple, using a diathermy loop. In other cases, particularly in young women with low-grade cytological abnormality, the colposcope will allow the colposcopist to identify the source of the low-grade cytological abnormality and will be able to recognise that in many instances the abnormality is due to ‘nothing more’ than immature squamous metaplasia, usually in association with HPV change; under these circumstances, the colposcopist will be able to advise as to whether conservative management without surgery is suitable. The main purpose of colposcopy is to reduce deaths from cervical cancer by identifying premalignant disease detected during the cervical screening programme, either by cytology or by primary HPV DNA testing. Quality of training is important, and all colposcopists should be trained adequately. They should regularly audit their work to confirm that they are delivering a recognised high-quality standard of service, and finally, they should be willing to undergo re-accreditation of their competence to practice colposcopy. References 1 2
Koller O: The Vascular Patterns of the Uterine Cervix. Oslo, Universitets forlaget, 1963. Kolstad P, Stafl A: Atlas of Colposcopy. Oslo, Universitets forlaget, 1972.
Singer
176
3
4
5 6 7 8
9
10 11 12 13
14
15 16 17 18 19 20 21 22 23
Reid R, Scalzi P: Genital warts and cervical cancer: an improved colposcopic index for differentiating benign papillomaviral infections from high-grade cervical intraepithelial neoplasia. Obstet Gynecol 1985;15:611–618. Walker P, Dexeus S, De Palo G, Barrasso R, Campion M, Girardi F, Jakob C, Roy M: International terminology of colposcopy: an updated report from the International Federation for Cervical Pathology and Colposcopy. Obstet Gynaecol 2003;101:175–177. Jordan JA: Colposcopy in the diagnosis of cervical cancer and precancer. Clin Obstet Gynecol 1985;12:67–76. Dexeus S, Cararach M, Dexeus D: The role of colposcopy in modern gynecology. Eur J Gynaec Oncol 2002;23:269–277. Hilgarth M, Menton M: The colposcopic screening. Eur J Obstet Gynecol 1966;65:65–69. Cuzick J, Szarewski A, Cubie H, Hulman G, Kitchener H, Luesley D, et al: Management of women who test positive for high risk types of human papilloma virus: the HART study. Lancet 2003;362:1871–1876. Van Niekerk WA, Dunton CJ, Richart RM, Hilgarth M, Kato H, Kaufman RH, Mango LJ, Nozawa S, Robinowitz M: Colposcopy, cervicography, speculoscopy and endoscopy. IAC Task Force Summary. Acta Cytol 1998;42:33–49. Belinson JL, Pretorius RG, Zhang WH, Wu LY, Qiao YL, Elson P: Cervical cancer screening by simple visual inspection after acetic acid. Obstet Gynecol 2001;98:441–444. Mitchell MF, Schottenfeld D, Tortolero-Luna G, Cantor SB, Richards-Kortum R: Colposcopy for the diagnosis of squamous intraepithelial lesions: a meta-analysis. Obstet Gynecol 1998;91:626–631. Olaniyan OB: Validity of colposcopy in the diagnosis of early cervical neoplasia – a review. Am J Reprod Health 2002;6:59–69. Schneider A, Hoyer H, Lotz B, Leistritza S, Kuhne-Heid R, Nindl I, Muller B, Haerting J, Durst M: Screening for high-grade cervical intra-epithelial neoplasia and cancer by testing for high risk HPV, routine cytology or colposcopy. Int J Cancer 2000;89:529–534. Li J, Rousseau MC, Franco EL, Ferenczy A: Is colposcopy warranted in women with external anogenital warts? American Society for Colposcopy and Cervical Pathology. J Lower Gen Tract Dis 2003;7:22–28. Campion MJ, Brown JR, McCance DJ, Atia W, Edwards R, Cuzick J, Singer A: Psychosexual trauma of an abnormal cervical smear. Br J Obstet Gynaecol 1988;95:175–181. Jones MH, Singer A, Jenkins D: The mildly abnormal cervical smears: patient anxiety and choice of management associated with colposcopic examination. J R Soc Med 1996;89:257–260. Wilkinson C, Jones JM, McBride J: Anxiety caused by abnormal result of cervical smear test: a controlled trial. BMJ 1990;300:440–444. Ferris DG, Litaker MS, Macfee MS, Miller JA: Remote diagnosis of cervical neoplasia: 2 types of telecolposcopy compared with cervicography. J Fam Pract 2003;52:298–304. Howard M, Sellors J, Lytwyn A: Cervical intraepithelial neoplasia in women presenting with external genital warts. CMAJ 2002;166:598–599. Rogstad KE: The psychological impact of abnormal cytology and colposcopy. Br J Obstet Gynaecol 2002;109:364–368. Luesley D: Standards and Quality in Colposcopy. Sheffield, NHS Cervical Screening Programme (NHSCSP) Publication No 2, January 1996. Luesley D: Standards and Quality in Colposcopy. Sheffield, NHS Cervical Screening Programme (NHSCSP) Publication No 3, April 2004. Redman C, Dollery E, Jordan JA: Development of the European Colposcopy Core Curriculum: use of the Delphi technique. J Obstet Gynecol 2004;24:780–784.
Prof. Albert Singer, MD, PhD Department of Obstetrics and Gynaecology Whittington and Royal Northern Hospital, Highgate Hill London, N19 5NF (UK) Tel. ⫹44 171 288 5409, Fax ⫹44 181 458 0168, E-Mail
[email protected] Colposcopic Diagnosis
177
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 178–183
Basic Treatment Options for Cervical Intraepithelial Neoplasia and Warts Alex Ferenczy Pathology and Obstetrics and Gynecology, McGill University and Sir Mortimer B. Davis-Jewish General Hospital, Montreal, Canada
Cervical Intraepithelial Neoplasia
There are two basic treatments for cervical intraepithelial neoplasia (CIN), namely ablative and excisional procedures. Neither of these procedures is recommended during pregnancy. Other contraindications include active, severe cervicitis, a history of cryoglobulinemia (for cryocoagulation) and active menstruation. Whether ablative or excisional, the treatment field should ideally encompass the entire lesional epithelium and that of the transformation zone. Ablative procedures destroy lesional tissue and provide no material for histologic ascertainment. They include cryocoagulation, electrofulguration (EF) and carbon dioxide (CO2) laser vaporization. The choice of procedure is determined by the availability and cost of the equipment, the technical skills needed and the prevailing concept in a given institution. Table 1 shows the treatment results and table 2 the postoperative adverse effects with the various ablative techniques used for CIN. The most favored ablative technique for CIN has been cryocoagulation. It was introduced in North America in the early 1970s and soon proved to be the simplest and cheapest management approach to CIN, provided that the patients were reliable for long-term follow-up. Among all therapeutic means, cryotherapy has the lowest rates of complications including bleeding, infection or stenosis. Small-sized CINs (⬍2.5 cm2) respond favorably to cryotherapy, but larger lesions (⬎2.5 cm2) have a higher failure rate, unless multiple applications of the cryoprobe tips are carried out in the same session [1]. However,
Table 1. Results of treatment for CIN by ablative techniques1 Method
Cryotherapy CO2 laser vaporization EF
Failure rates (%) small-sized CIN (⬍2.5 cm)
large-sized CIN2 (⬎2.5 cm)
10 6 10
40 8 60
Adapted from Ferenczy A: Management of patients with high-grade squamous intraepithelial lesions. Cancer 1995;76:1923–1928. 1 With endocervical gland involvement. 2 Low- and high-grade lesions involving 3 or 4 quadrants.
Table 2. Adverse events associated with ablative techniques for CIN Adverse events
Cryotherapy (%)
EF (%)
LV (%)
Bleeding Os stenosis Hematometra Infection
ⱕ1 ⱕ5 ⱕ0.1 ⬍2
ⱕ6 ⱕ0 0 2
⬍5 0 2
LV ⫽ Laser vaporization. Adapted from Ferenczy A: Management of patients with highgrade squamous intraepithelial lesions. Cancer 1995;76:1923–1928.
multicryotherapy is associated with high rates of side effects, including severe uterine cramps, vasovagal symptoms (i.e. fainting), profuse vaginal discharge and, occasionally, unsatisfactory colposcopies [1]. EF with a ball-shaped electrode for ablating CIN was used as early as the late 1940s. However, the technique failed to gain general acceptance because it often produced cervical stenosis, the therapeutic results were poor in cases of endocervical gland involvement, and lack of fume evacuation systems prevented adequate visualization of the cervix. Recently, the technique and equipment have been refined, and excellent results with low rates of complications have been reported in patients electrofulgurated for low-grade CIN [2]. The advantage of EF over cryotherapy is that it allows treatment of lesional epithelium regardless of size, and it is not associated with uterine cramps and abundant vaginal discharge during and following the procedure. On the other hand,
Treatment of CIN and Warts
179
EF, unlike cryotherapy, requires local, injectable anesthesia and accessory materials such as return pad electrodes, coated speculum and a fume evacuator system. Overall, cryotherapy is the most cost-effective treatment option for small CINs occupying one to two quadrants, regardless of histological grade, and is the least skill dependent [1]. In the 1980s, CO2 laser vaporization became widely popular in North America. Because the laser beam could destroy tissues with great precision, early studies suggested more favorable treatment results than those obtained with cryocoagulation. However, subsequent experience found no significant difference in treatment results between CO2 laser therapy and cryotherapy, particularly for small (less than two quadrants) lesions [1]. Patients with large lesions responded comparatively better to CO2 laser therapy than to cryotherapy [3]. However, because large lesions are seldom seen in most general gynecological practices (in which patients are regularly screened with Pap tests) and the CO2 laser machine is comparatively more expensive to purchase and maintain than cryosurgical equipment, its use has been largely limited to centers or clinics with high patient referrals. Recurrence rates of CIN (defined as discovery of lesional epithelium after 1 year of postcryotherpy and laser ablation disease-free period) are generally low to very low, in the order of 4 and 0.4%, and largely due to incomplete treatment of lesional epithelium not readily recognized by the operator(s) [4]. In the author’s experience as well as that of others, most invasive cancers that are diagnosed after ablative therapies are a reflection of inadequate preablative diagnostic work-up (pre-existent lesions) rather than genuine new disease [5]. In current practice, the most frequently used and cost-effective excisional treatment method is the loop electrosurgical excision procedure (LEEP) or large loop excision of the transformation zone [6]. It is an office conization procedure done under local anesthesia with the use of loop-shaped electrodes. By choosing the appropriate type of electrode, physicians can excise virtually all CINs regardless of grade, size and distribution. Large clinical series published both in the UK and North America have shown excellent cure rates and less than 10% complication rates [6, 7]. The combined cure rate after one LEEP for both low- and high-grade CIN in a study of 883 women was 95%, and after repeat LEEP, 98% [1]. The rate of unexpected findings (an additional good feature of LEEP), such as adenocarcinoma in situ, early invasive squamous cell carcinoma and adenocarcinoma, ranges from 0.8 to 4%. Only one third to one half of all of these lesions were suspected before LEEP. In most series, complications such as perioperative and postoperative bleeding, infections and stenosis are in the order of 10% or less [3, 6, 7]. With the increasing use of LEEP, two problems with clinical potentials have emerged. One was the removal of an excessive amount of tissue, and the second,
Ferenczy
180
removal of disease-free tissue [6]. The first issue relates to the use of inappropriately designed loop electrodes, some of which have a depth of more than 2 cm. In current practice, the recommended standard loop measures 8 mm in depth. With this type of loop, only about 5- to 6-mm-deep tissue is removed with a single pass, yet more than 90% of lesions are treated successfully. In addition, performing LEEP under colposcopic guidance enhances depth control. In doing so, the incidence of pregnancy among women treated with LEEP is similar (8.1/100 womenyears) to that among untreated women (7.4/100 women-years), as are the rates of premature deliveries in treated (9.9%) versus untreated (9.4%) women [8]. On the other hand, excisional depths greater than 1.5 cm (seldom needed) are associated with relatively high preterm deliveries [9]. The overtreatment issue relates to a series of possible causes, namely (1) classification of benign cellular changes on cytology and/or colposcopy as intraepithelial lesions (false positives), (2) performance of multiple cervical biopsies before LEEP, possibly resulting in the removal of many small lesions, (3) the failure to discover lesions in poorly processed excisional specimens, and (4) destruction of lesional tissue by thermocoagulation injury. Most experts have found the two former factors to be the most likely causes of LEEP negative specimens. Because of the potential risk of overtreating patients with LEEP, the physician must be skilled in colposcopic prediction and have access to a laboratory with a record of low false-positive cytologic results. Although some overtreatment with LEEP is unavoidable (the same is true for any other excisional technique), the risk for doing so can be minimized through application of traditional diagnostic triage to cytologically and colposcopically equivocal lesions. In contrast, for patients whose high-grade lesions are unequivocally documented by positive cytologic and colposcopic results, the ‘see-and-treat’ protocol is appropriate [6, 7]. Among such cases, if the specimen contains no invasion and the patient is disease free on follow-up, then she has been diagnosed and treated in the same visit. Cold-knife conization (CKC) is preferred by most investigators for endocervical canal lesions, particularly endocervical adenocarcinoma in situ. By doing so, ‘clean’ endocervical margins are obtained, facilitating the evaluation of surgical lines by the pathologist. The therapeutic success with CKC is similar to that of LEEP, but postoperative complications are not negligible (table 3). Alternative excisional techniques to CKC are laser and microneedle electroconization procedures [10]. Both techniques require technical skills and general anesthesia. The advantage of these two methods over CKC is the possibility to remove less disease-free tissue, for both techniques allow for excising a square or rectangular, cork-shaped specimen along the long axis of the endocervical canal as opposed to the triangular shape of CKC. Hysterectomy is seldom required for CIN per se, for extension close to the lower uterine segment or higher in the endometrial cavity is a rare to very rare
Treatment of CIN and Warts
181
Table 3. Complications of excision for high-grade CIN Method
CKC CO2 laser cone LEEP
Patients with complications (%) bleeding1
os stenosis
14–22 5–10 8
17 2 4
Adapted from Ferenczy A: Management of patients with high-grade squamous intraepithelial lesions. Cancer 1995;76:1923–1928. 1 Perioperative and postoperative complications combined.
phenomenon. Most patients scheduled for hysterectomy for ‘CIN’ have, in fact, symptomatic endomyometrial disease and no longer desire to conceive. In such cases, CIN is a co-existent ‘passenger’ condition rather than a genuine indication for hysterectomy. Long-term follow-up is important for all women who have been treated for CIN irrespective of its grade and therapeutic methods used. Several large, recent epidemiological studies and meta-analyses indicate that the incidence of invasive cancer of the cervix in the first 10–20 years after treatment of CIN1–3 is higher (23 to 85/100,000 woman-years) than in the average population [11, 12]. However, in others’ experiences, only 3 of 1,696 (0.1%) patients treated with LEEP for high-grade CIN developed invasive cervical cancer. The average follow-up was 6.5 years [13].
Cervical Warts (Condylomata Acuminata)
Most lesions on the cervix are not keratinized and are easy to treat by either ablative or excisional procedures. Therefore, biopsy-proven cervical warts may be treated with cryotherapy using nitrous oxide with appropriate cryoprobe tips or EF with ball-shaped electrodes under local anesthesia. Alternative therapies are excisional procedures using biopsy punches or weekly topical application of 80% trichloro- or bichloroacetic acid solutions.
References 1
Ferenczy A: Comparison of cryo- and carbon dioxide laser therapy for cervical intraepithelial neoplasia. Obstet Gynecol 1985;66:793–798.
Ferenczy
182
2 3
4
5 6
7
8
9 10 11 12 13
Roy MC, Mayrand MH, Franco E, Arseneau J, Ferenczy A: Electrofulguration for low-grade squamous intraepithelial lesions of the cervix. J Low Genit Tract Dis 2004;8:10–15. Mitchell MF, Tortolero-Luna G, Cook E, Whittaker L, Rhodes-Morris H, Silva E: A randomized clinical trial of cryotherapy, laser vaporization and loop electrosurgical excision for treatment of squamous intraepithelial lesions of the cervix. Obstet Gynecol 1998;92:737–744. Richart RM, Townsend DE, Crisp W, DePetrillo A, Ferenczy A, Johnson G, Lickrish G, Roy M, Villa Santa V: An analysis of long term follow-up results in patients with cervical intraepithelial neoplasia treated by cryotherapy. Am J Obstet Gynecol 1980;137:823–826. Townsend DE, Richart RM, Marks E, Nielsen J: Invasive cancer following outpatient evaluation and therapy for cervical disease. Obstet Gynecol 1981;57:145–149. Ferenczy A, Choukroun D, Arseneau J: Loop electrosurgical excision procedure for squamous intraepithelial lesions of the cervix: advantages and potential pitfalls. Obstet Gynecol 1996;87: 332–337. Wright TC, Richart RM, Ferenczy A: Electrosurgery for HPV-Related Diseases of the Lower Genital Tract: A Practical Handbook for Diagnosis and Treatment by Electroexcision and Fulguration Procedures. Montreal, BioVision, 1992. Ferenczy A, Choukroun D, Falcone T, Franco E: The effect of cervical loop electrosurgical excision on subsequent pregnancy outcome: North American experience. Am J Obstet Gynecol 1995;172:1246–1250. Sadler L, Saftlas A, Wang W, Exeter M, Whittaker J, McCowan L: Treatment for cervical intraepithelial neoplasia and risk of preterm delivery. JAMA 2005;291:2100–2106. Ferenczy A: Electroconization of the cervix with a fine-needle electrode. Obstet Gynecol 1994;84:152–159. Kalliala I, Anttila A, Pukkala E, Nieminen P: Risk of cervical and other cancers after treatment of cervical intraepithelial neoplasia: retrospective cohort study. Br Med J 2005;331:1183–1185. Soutter WP, Sasieni P, Panoskaltsis T: Long-term risk of invasive cervical cancer after treatment of squamous cervical intraepithelial neoplasia. Int J Cancer 2005;118:2048–2055. van Hamont D, van Ham MA, Struik-v/d Zanden PH, Keijser KG, Bulten J, Melchers WJ, Massuger LF: Long-term follow-up after LLETZ: evaluation of 22 years treatment of high-grade CIN (abstract). Int J Gynecol Cancer 2005;15:77.
Alex Ferenczy, MD Department of Pathology, SMBD-Jewish General Hospital 3755 Cote Sainte-Catherine Road Montreal, Quebec H3T 1E2 (Canada) Tel. ⫹1 514 340 7526, Fax ⫹1 514 340 7542, E-Mail
[email protected] Treatment of CIN and Warts
183
Prevention by Vaccines: Current Status, Impact and Prospects Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 184–205
Prevention of Cervical Cancer: Challenges and Perspectives of HPV Prophylactic Vaccines Joseph Monsonego Eurogin, Paris, France
Cervical cancer is one of the few avoidable human cancers. Its prevention is currently based on early diagnosis and treatment of benign and precancer lesions. Initial morphological analysis of cells contained in cervical smears is followed by a more precise examination, colposcopy, in which abnormalities are located by examining the surface of the cervical epithelium. In theory, treatment of early lesions prevents the development of invasive cancer [1]. This approach, from screening to effective prevention, is unique to cervical cancer. However, this process is more complex than it appears, as it requires the coverage of a large proportion of the target population, regular screening from the age of 20 to 65 years, adequate sampling, strict morphological analysis, effective colposcopy and biopsy, and appropriate management of any precancer lesions [2]. With the knowledge of the causal relationship between some types of HPV and cervical cancer [3], new possibilities have emerged for screening and preventing cervical intraepithelial neoplasia (CIN) and cervical cancer. It is generally agreed that combining HPV testing with the Pap test improves the performance of early detection (secondary prevention). HPV testing can also be useful when Pap test results are ambiguous (ASCUS, atypical squamous cells of undetermined significance) [4]. In primary prevention, research is currently focusing on candidate vaccines that will prevent HPV infection, and thus, cervical cancer. This review examines the impact of secondary prevention, epidemiological data and the public health implications of primary prevention of cervical cancer based on HPV vaccination.
Per 100,000
15
14.2 11.9 10.1
10 5
4.5
8.8
8.0
3.5
2.8
2.3
1.9
1985
1990
1995
2000
3,400 cases 1,000 deaths
0 1980
Incidence
Mortality
Fig. 1. Incidence of and mortality due to invasive cervical cancer in France (per 100,000); standardized for world population.
Cervical Cancer Worldwide
Cervical cancer is the second most frequent female cancer worldwide, representing about 10% of all female cancers. In 2002, there were an estimated 493,000 new cases of invasive cervical cancer, of which 83% were diagnosed in developing countries [5]. The geographic regions at highest risk of cervical cancer include southern and eastern Africa, the Caribbean and central America, where the annual incidence exceeds 30 per 100,000 [6]. Each year, there are an estimated 273,000 deaths from cervical cancer, over three quarters of which occur in developing countries [6]. Incidence and mortality rates generally correlate with each other, but some regions, such as Africa, have abnormally high mortality rates. Less than 50% of women diagnosed with cervical cancer in developing countries survive beyond 5 years, and most victims are multiparous women at child-bearing age. In contrast, the 5-year survival rate in industrialized countries is about 66%. In France, about 3,400 new cases of invasive cervical cancer were diagnosed in the year 2000 (incidence rate 8/100,000) and there were about 1,000 deaths (4/100,000) [7–9]. In this country, cervical cancer is the eighth most frequent female cancer but the fifth cause of cancer deaths. Between the 1970s and 1990s, the number of cases fell by about 2.5% per year (fig. 1) [10], but it has remained stable during the last decade. In Europe, the estimated average incidence is 15.7/100,000, and 80 women die from cervical cancer each day [11]. Role of HPV
Oncogenic HPV Infection: A Prerequisite for Cervical Cancer It is now clear that high-risk papillomavirus types are responsible for precancer lesions and cervical cancer [3]. Chronic HPV infection is
Cervical Cancer Prevention by HPV Vaccination
185
considered to be an intermediate stage on the way towards invasive cervical cancer [12, 13]. This is a unique situation in oncology: no other human cancer has such a well-established relationship with a virus. Compared with the other known risk factors of human cancer, such as smoking (lung cancer) and hepatitis B virus (HBV) infection (cancer of the liver), the risk associated with HPV is even stronger. The relative risks are about 10 for smoking and lung cancer, 50 for cancer of the liver and HBV, but around 300–400 for cervical cancer and HPV [13]. The link between HPV and cervical cancer has led to two forms of prevention: (1) screening for HPV as a marker of early precancer (CIN) lesions, and (2) HPV vaccination to prevent the onset of these lesions. Natural History of HPV Infection During the late 1960s and early 1970s, epidemiological studies provided data suggesting that cervical cancer was transmitted through sexual intercourse, pointing to the role of an infectious agent in the etiology of cervical neoplasia. First, the available evidence implicated that genital infection by herpes simplex virus would be such an etiological factor. However, it became evident that although herpes simplex virus was oncogenic in vitro and in vivo, the link to cervical cancer was indirect. Since the early 1980s, attention gradually shifted to a new candidate, HPV, based on rapidly accumulating evidence from molecular biological studies. The relative risk of the association of HPV and cervical cancer was two or three times higher than that of other potent risk factors for cancer. In 1995, the International Agency for Research on Cancer finally pronounced HPV16 and HPV18 as carcinogenic in human beings [14]. However, HPV infections are known to be very common among the general population; it is estimated that about 7 out of 10 women are exposed at least once to HPV during their lifetime [15]. Without treatment, about 1 in 5 women exposed to HPV can develop cervical cancer. Exposure to these viruses occurs during sexual intercourse, often with the first partner [16, 17]. The estimated prevalence of HPV infection is about 30% before the age of 30 years. It falls gradually to about 10% between 30 and 50 years, and to 5% beyond 50 years of age (fig. 2) [18, 19]. In Europe, HPV16 and HPV18 are the two most prevalent types in women with a normal Pap test, while geographic variations in the type distribution are observed in Asia and Sub-Saharan Africa [20]. While type 16 accounts for only 26.3% of low-grade squamous intraepithelial lesions (LSIL) and 45% of high-grade squamous intraepithelial lesions (HSIL), a recent meta-analysis showed an HPV16 cancer/LSIL ratio of about 2 [21], compared with 1.21 for cancer/HSIL [22]. More recently, it was reported that the risk of developing CIN3 or cancer 10 years after HPV infection was
Monsonego
186
Cases per 100,000
35
30
30
25
25
20
20
15
15
10
10
5
5
Frequency of HPV infection (all types)2 (%)
%
35
0 0
5
Cervical cancer in EU (1998)1 (per 100,000) Incidence Mortality
10 15 20 25 30 35 40 45 50 55 60 65 70 75 Age (years)
• In Europe, almost 33,500 cases and 15,000 deaths per year3 • Second cause of cancer death among young women (15–44 years)1
Fig. 2. HPV infections in adolescents and adults. 1Ferlay et al., Eucan 1999. Hypothetical distribution of HPV prevalence in EU modelled after data available in the US (Portland) and Canada (Sellors). 3Ferlay et al. [5]. 2
17.2% for HPV16 and 13.6% for HPV18 [23]. Exposure and persistence are more frequent with HPV16 than with the other oncogenic HPV types [24]. All these data substantiate the development and implementation of preventive vaccination against HPV16 and HPV18. Most women exposed to HPV develop a limited immunity, generally clearing HPV within 9–12 months [17]. In contrast, HPV remains persistent for months or even years in 20% of women who may go on to develop CIN and cancer, unless diagnosed and promptly treated (table 1) [25–30]. The onset of CIN lesions thus reflects the failure of the immune system to control HPV, a situation most common with HPV types 16 and 18 [24]. The simplified model of cervical carcinogenesis is shown in figure 3. HPV infection is usually transient in women under 30, but tends to be more persistent and problematic after this age [31]. Types 16 and 18 are more persistent than the other types [24]. Thus, the presence of HPV within the cervix may or may not be associated with clinical lesions. In contrast, the persistence of viral DNA beyond 12 to 18 months is predictive of current or future lesions, especially in case of HPV16 or HPV18 infection [24]. The relative risk of developing neoplasia in subsequent years is between 11 and 350 (table 1). HPV persistence is associated with the expression of certain viral genes, especially E6 and E7
Cervical Cancer Prevention by HPV Vaccination
187
Table 1. Role of persistent infection in lesion onset Author, year
• • • • • •
OR-RR
SIL/CIN
OR for invasive cancer
Dalstein, 2003/Bory, 2002 237 CIN2–3 Meijer/Nobbenhuis, 1999 327 CIN3 Ho, 1998 37.2 SIL Franco, 1998 20.6 SIL Koutsky, 1992 11 CIN2–3 Wallin/Dillner, 1999 Persistence increases with age Development of SIL/Cx in women with normal smears
In 1 year
Normal cervical, no infection
Normal cervical, initial HPV infection
>5 years
Persistent infection (⫾CIN)
16.4
>10 years
CIN 2/3
Cervical cancer
CIN1
HPV detected Cleared HPV infection
HPV not detected
Fig. 3. Natural history of high risk HPV infection.
(high-risk HPV types only), whose role in host cell immortalization involves an action on proteins that regulate the normal cell cycle [32, 33]. Specific binding of E7 protein to products of the cell cycle inhibitory gene pRb is responsible for cell proliferation. E6 protein binding degrades p53 protein, thereby disrupting apoptotic mechanisms [34]. HPV Types with Different Natural History More than 100 HPV genotypes have been characterized. Forty types show tropism for the epithelium in the respiratory, intestinal and lower genital tract.
Monsonego
188
Among the latter, types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68 are considered to be oncogenic, owing to their frequent association with cervical cancer and CIN. Other types with genital tropism (types 6, 11, 42 and 44) and other more rare types are considered as nononcogenic and generally cause subclinical infections or benign clinical lesions such as condyloma acuminatum or flat warts.
HPV Vaccination in Primary Prevention of Cervical Cancer
The fact that cervical cancer is caused by a viral infection raises the possibility of preventing the disease by vaccination against this known etiological agent, similarly as hepatitis B (HBV) vaccination prevents cancer of the liver. Campaigns against sexually transmitted diseases had a limited impact on the incidence of HPV infection, particularly those due to high-risk HPV types. While the treatment of clinical lesions (e.g., condyloma acuminatum) is necessary to limit the disease transmission, the treatment of asymptomatic HPV infection is currently not recommended as measures of controlling cervical cancer. Principles of HPV Vaccination Two types of HPV vaccine are under development: (1) prophylactic vaccines to prevent HPV infection and associated lesions, and (2) therapeutic vaccines to induce regression or remission of established precancer lesions or cervical cancer. In this discussion, the author will only focus on describing the issues related to prophylactic vaccination. Current prophylactic HPV vaccines are based on purified virus-like particles (VLPs), i.e. the viral capsid, without the viral DNA, composed of the main envelope protein L1 of the oncogenic HPV types. VLPs are indistinguishable from the complete virions under electron microscope. At present, two pharmaceutical companies, GlaxoSmithKline (GSK) and Merck and Co. are actively developing prophylactic vaccines. VLPs are produced in baculovirus-infected insect cells (Cervarix®, GSK) or yeast cells (Gardasil®, Merck). They are neither infectious nor oncogenic. They induce a high titer of specific neutralizing antibodies, which protect the cervix by transudating into the cervical mucus. The antibodies bind to the viral capsids and thereby prevent the host cell infection. Vaccination with VLPs of nonhuman papillomavirus types with mucosal tropism confers type-specific protection, but no regression of the existing lesions. This suggests that prophylactic vaccination will not act on clinically established lesions.
Cervical Cancer Prevention by HPV Vaccination
189
Table 2. Characteristics and results of the three major randomized trials of HPV prophylactic vaccines Characteristics of trials
Merck
Merck
GSK
Reference
Koutsky et al. [35], 2002 (phase 2) Monovalent HPV16 VLP L1
Villa et al. [36], 2005 (phase 3) Tetravalent HPV16, HPV18, HPV6 and HPV11 VLP L1 Yeast (Saccharomyces cerevisiae) 20 g HPV6/40 g HPV11/40 g HPV16/ 40 g HPV18 Hydroxy aluminium, phosphosulfate 0.5
Harper et al. [37], 2004 (phase 2) Bivalent HPV16 and HPV18 VLP L1
0.2 and 6 239 vaccinees versus 242 placebos Brazil, Europe, USA 16–23 HPV DNA⫹/⫺, serology⫹/⫺
0.1 and 6 560 vaccinees versus 553 placebos USA, Brazil, Canada 15–25 HPV DNA and serology negative; no cervical lesions, few sexual partners ⬎27, 18 (Ab)
Type of vaccine
Expression system Yeast (Saccharomyces cerevisiae) Concentration 40 g HPV16
Adjuvant Dose (intramuscular), ml Program, months Size of trial Site Age group, years Main inclusion criteria Duration, months
Hydroxy aluminium, phosphosulfate 0.5
0.2 and 6 768 vaccinees versus 765 placebos USA 16–23 HPV DNA and serology negative; no cervical lesions or few sexual partners 48, 7 (Ab)
Efficacy on incident 91 (80–97) and transient infections, % Efficacy on 100 (90–100) persistent infections, % Efficacy on cytologic lesions, % Efficacy on preinvasive lesions, % Tolerability Seroconversion, %
⬎48, 36 (Ab)
Insect cells (Baculovirus) 20 g HPV16/20 g HPV18 ASO4 0.5
92 (65–98)
89% ᭙ HPV
HPV6: 100 HPV11: unknown HPV16: 86 HPV18: 89
{
Not assessed
100 (51–100)
93 (70–98)
100 (24–100)
100
100 (51–100)
Acceptable 100
Acceptable 100
Acceptable 100
Ab ⫽ Neutralizing antibodies; Po ⫽ placebo.
Monsonego
190
Three randomized trials of candidate HPV vaccines have been published so far (table 2), with promising results in terms of efficacy and immunogenicity [35–37]. These studies show that HPV vaccination can prevent not only precancer lesions associated with these viral types but also both persistent and incident infection, with almost total efficacy. Vaccination prevents clinical onset and also eliminates the virus from the genital tract, thus avoiding infection of new partners. In the three studies, vaccination was most effective in preventing the acquisition and persistence of the relevant HPV types and CIN related types. HPV Types in the Vaccines The number of viral types used as immunogens is an important issue. Epidemiological studies have determined the proportion of cervical cancers attributable to each high risk HPV type. A recent review of case-control studies indicates that a pentavalent vaccine composed of VLPs of HPV16, HPV18, HPV45, HPV31 and HPV33 could potentially prevent 83% of all cases of cervical cancer. A heptavalent vaccine, also including types 52 and 58, could prevent 87% of the cases [38]. However, the maximum gain is obtained with four viral types, beyond which one comes up against the law of diminishing returns. The potential impact of HPV vaccination can only be studied by using complex models that incorporate variables which affect the natural history of cervical cancer in different countries and in different scenarios. Markov models are interesting tools with which to predict vaccine efficacy, taking into account the level of immunity and the natural progression of HPV infection to cervical cancer. These same models can be adjusted to evaluate cost effectiveness relative to other preventive measures [39]. Requirements for an HPV Vaccine The ongoing and published randomized studies have focused on a monovalent vaccine (HPV16), a bivalent vaccine (HPV16 and HPV18), and a quadrivalent vaccine (HPV16, HPV18, HPV6 and HPV11). Phase II and III trials have assessed safety and immunogenicity and have a maximum follow-up of about 4 years. Vaccine efficacy is assessed by using viral and lesional markers such as persistent HPV infection, CIN (especially high-grade CIN) and cancer, although follow-up is currently too short for this latter endpoint. The analyses focus on specific efficacy against the vaccine types, but also on potential cross-immunity to other HPV types and their associated lesions. It will also be necessary to determine the duration of protection and the possible need for
Cervical Cancer Prevention by HPV Vaccination
191
booster injections. Ongoing studies are examining the correlation between antibody titers and clinical protection. Two candidate vaccines are currently being tested in phase III trials. The GSK vaccine (Cervarix®) is bivalent, containing HPV16 and HPV18 VLP L1, while the Merck vaccine (Gardasil®) is quadrivalent and contains HPV16, HPV18, HPV6 and HPV11 VLP L1. The GSK vaccine is designed to prevent CIN and cervical cancer due to HPV16 and HPV18, while the Merck vaccine is designed to prevent not only these lesions but also condyloma acuminatum, a frequent disease in external genitalia of young women.
Results of Randomized Phase II and III Trials
The main characteristics and results of the three principal studies of HPV vaccines are summarized in table 2 and briefly discussed here. Results of the GSK Bivalent (HPV16 and HPV18) Vaccine A phase II multi-center, randomized, double-blind, placebo-controlled trial was conducted in Brazil, Canada and the United States [37]. Each vaccine shot dispensed 40 g of HPV16 and HPV18 VLP L1. The adjuvant was ASO4, an original adjuvant composed of aluminum and monophosphoryl lipid A (a mucopolysaccharide as immunogen). ASO4 induces a stronger immune response than the aluminum adjuvant used in most vaccines. Three injections were given, at 0, 1 and 6 months. About 1,100 women aged 15–25 years have been enrolled. The inclusion criteria were HPV16 and HPV18 seronegativity (Elisa), HPV16 and HPV18 PCR negativity and a normal Pap smear (liquidbased cytology). Compared with placebo, efficacy against incident infections ranged from 91.5 (types 16 and 18) to 100% (type 16). Efficacy on persistent infections was 100% (types 16 and 18). Efficacy on cytologic abnormalities related to these two viral types was 93%. The vaccine induced strong seroconversion: 7 months after vaccination, titers of neutralizing antibodies to HPV16 and HPV18 reached levels 1,000 times higher than at baseline and 80–100 times higher than after natural infection. The bivalent vaccine was well tolerated (local reactions in 93.5% of patients, versus 87% with controls; general reactions in 78.6 and 78.5% of patients, respectively). Interesting results were recently obtained [G. Dubin, pers. commun., Vancouver 2005] regarding possible cross-immunization by the bivalent vaccine against virus types phylogenetically related to HPV16 and HPV18. These
Monsonego
192
results need to be confirmed in larger studies. A phase III trial started in 2003 involves 30,000 female volunteers aged from 10 to 25 years and over 25 years. The aims are to determine the best age for vaccination, as well as safety and long-term efficacy.
Merck Trials Phase I Trial. This study involved 300 subjects and assessed the immunogenicity and tolerability of a range of monovalent VLP L1 vaccine doses. The active vaccination group consisted of 82 subjects who received the monovalent HPV16 vaccine, while the controls consisted of 167 subjects, some of whom were vaccinated with an HPV11 vaccine while others received a placebo. No cases of persistent HPV16 infection were detected by PCR in the active vaccination group, while 15 cases occurred among the controls. This study indicated that there was no cross-immunization between the HPV11 vaccine and HPV16. In the same study, the titer of neutralizing antibodies was 1,000 times higher than at baseline 8 months after the first injection. The antibody titer fell slightly with time but remained 100 times higher than at baseline after 36 months. The antibody titer did not vary significantly with the vaccine dose used, and it was decided to use the lowest doses for phase II trials. Phase II Trials. A total of 1,193 subjects received an HPV16 VLP L1 vaccine and were compared with 1,198 controls. The vaccine was injected intramuscularly at 0, 2 and 6 months. General and local tolerability was similar in the vaccine and control groups. An interim assessment was done after 17 months. The HPV16 vaccine was 100% effective against persistent HPV16 infection and against HPV16-associated CIN and CIN3 [35]. Results at 3.5 years [C. Mao, pers. commun., ICAAC 2004] showed 100% efficacy on persistent HPV16 infection and on HPV16-associated CIN3. Whatever the viral type, efficacy was 30% on CIN1, 40% on CIN2, 52% on high-grade lesions and 73% on CIN3. At 36 months, the neutralizing antibody titer was 100 times higher than at baseline and 10 times higher than after natural infection. Phase IIb Trial. This study by Villa et al. [36] examined the performance of a quadrivalent vaccine against HPV16, HPV18, HPV6 and HPV11 (Gardasil®) during a 36-month period. General and local tolerability was excellent. Efficacy reached 90% on persistent infections associated with all four viral types, 90% on associated lesions, 100% on HPV6 and HPV11 condylomata (although the observed number of cases was low), 86% on lesions associated with HPV16 and 89% on lesions associated with HPV18. The titers of neutralizing antibodies against HPV18, HPV6 and HPV11 fell significantly at 36 months;
Cervical Cancer Prevention by HPV Vaccination
193
however, correlation between antibody titers and clinical protection must be further studied. Phase III Trial. A cohort of 25,000 women are participating in this study of vaccine tolerability, immunogenicity, public health benefit and efficacy on high-grade CIN, CIN1 and condyloma acuminatum. Preliminary results of the FUTURE II study, including data from phases II and III, were recently presented to the US Advisory Committee on Immunization Practices, which is charged with making recommendations on HPV vaccination [E. Barr, ACIP October 2005]. The results for more than 10,000 patients at 17 months postvaccination follow-up showed 100% efficacy on HPV16 and HPV18 infection and at 26 months follow-up associated CIN2–3. The FUTURE I study of the quadrivalent vaccine showed 100% efficacy on HPV6, HPV11, HPV16 and HPV18 infection and on associated CIN in situ adenocarcinoma of all stages. Efficacy on HPV6, HPV11, HPV16 and HPV18 infection and on associated vaginal and external genital lesions was also 100%. The full results of these randomized studies should become available before the end of 2006.
Key Questions Involved in HPV Vaccination
Impact of HPV Vaccination Compared with vaccination against HBV, another oncogenic virus, the results of HPV vaccination could be even more spectacular. In Africa and southwest Asia, 10% of children are infected by HBV. An intensive vaccination program was started in some of these countries in 1984, targeting newborn babies. In 1992, the prevalence of hepatitis B in children had fallen from 10.5 to 1.7%, and the frequency of hepatocellular carcinoma had fallen by a factor of four. It has been calculated that by 2010, the prevalence of HBV infection in children should fall to 0.1% (a 99% reduction in the carriage rate) and this will be accompanied by a major reduction in the incidence of hepatitis B and its complications [40]. HPV vaccination will thus have a major impact in developing countries, where 80% of cases of cervical cancer occur and where cytological screening is nonexistent or ineffective. Vaccination against HPV16 and HPV18 would prevent 70% of cervical cancer cases. However, given the natural history of HPV infection, the impact on cervical cancer would only be measurable some 20 years after the beginning of a vaccination campaign. In industrialized countries, the impact on screening results would be observed more rapidly, with likely reductions of 90% in the incidence of HPV16 and HPV18 infection and with a
Monsonego
194
estimated reduction of 50% in cytological abnormalities, 50% in CIN1 and 70% in CIN3. A significant reduction in the prevalence of cytological abnormalities would be perceptible after 3–5 years [36, 37]. Because most cases of cervical cancer are associated with HPV16 and HPV18, more than 95% of deaths due to this malignancy could be prevented with an effective prophylactic HPV vaccination. The younger the target population, the longer the interval before an impact on HPV infection and cancer becomes noticeable. This must be taken into account when considering the optimal age for vaccination, given that the peak incidence of cervical cancer occurs at 45 and 65 years of age in most of the countries. Cost Effectiveness of HPV Vaccination Mathematical models indicate that a combination of HPV vaccination and Pap testing would be cost effective in industrialized countries. Prevention based on HPV vaccination alone would reduce but not eliminate cervical cancer [41], and therefore, cannot altogether replace screening. It is more plausible to foresee a program of cervical cancer prevention based on both primary prevention (vaccination) and secondary prevention (screening) [42–44]. Strategies combining vaccination with cytological screening should be more cost effective than strategies based on screening alone. The best strategy seems to be to combine three 5-yearly screenings starting at the age of 30, with vaccination targeted in preadoloscents. HPV vaccination would have a greater impact on HSIL than on LSIL smears related types, and it would thus be possible to design less aggressive treatments for LSIL related to other high risk types.
Questions Related to Implementation of HPV Vaccination
A series of questions need to be adequately solved before widespread implementation of HPV vaccines is feasible. Most important, the results of ongoing phase III trials will be necessary to answer most of these outstanding questions. The Ideal Target Population As shown in figure 2, HPV vaccination must be performed before the age of 20 if it is to prevent cervical cancer. It is unclear whether vaccination would be beneficial for adults who have already been exposed to the virus. The mean age of the first sexual intercourse is falling in industrialized countries [44] and
Cervical Cancer Prevention by HPV Vaccination
195
Table 3. Impact of different vaccination strategies
Opportunity for vaccination Expected coverage Rapid impact on cervical cancer Rapid impact on early lesions Rapid impact on genital warts Total impact on HPV diseases
Preadolescents (1 age group)
All preadolescents/ adolescents
Young women 18–25 years
⫹⫹⫹ ⫹⫹⫹ ⫹ ⫹⫹ ⫹⫹ ⫹⫹⫹
⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹
⫹⫹ ⫹/⫹⫹ ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹
To optimize the effect on cervical cancer and all HPV infection: • Start vaccination of preadolescents if feasible. • Extend to older adolescents and young women.
is currently around 17 in Europe. HPV vaccination of young adolescents would have to be preceded and accompanied by an education campaign both for girls and their parents. Vaccination would be more acceptable at 18 years of age during a consultation for contraception, e.g., but this population is not very easy to target. Table 3 shows the expected impact of different vaccination strategies targeted to different age groups of the vaccinees. It is likely that HPV vaccination will also be proposed to adults who have already been exposed to HPV, and the benefits of this measure should be determined by ongoing phase III trials. One possible benefit is an action on incident and latent infections, thereby reducing the risk of clinical lesions and transmission to contacts. Gender Should girls alone be vaccinated, or both boys and girls? At present, there is no clear answer to this question. Although vaccination of boys would limit the spread of the disease to women, the risk of precancerous lesions and cervical cancer is linked to the immune status of each individual woman. Cervical cancer is a rare complication of HPV infection. The cost effectiveness of vaccinating both sexes would therefore need to be established. In principle, it would only be necessary to vaccinate women in order to prevent cervical cancer. It would appear more sensible to focus the available (limited) resources on widespread vaccination of girls. In addition, efficacy in boys remains to be demonstrated. However, the introduction of a quadrivalent vaccine including HPV6 and HPV11, that could prevent condyloma acuminatum, might warrant vaccination of young boys as well.
Monsonego
196
Duration of Protection Postvaccination follow-up currently stands at about 4 years. Randomized trials of the monovalent vaccine against HPV16 (Merck) [36] and the bivalent vaccine against HPV16 and HPV18 (GSK) [37] show high antibody titers after three injections. Neutralizing antibody titers are 10–50 times higher than after natural infection by HPV16, 18 months after immunization with the GSK vaccine and 10 times higher 7 months after vaccination with the Merck monovalent vaccine. The phase III trial of the Merck vaccine showed that titers of neutralizing antibodies against HPV16, HPV18, HPV6 and HPV11 were still measurable at 36 months. Antibody titers remained high for HPV16 at 36 months, whereas they fell significantly for HPV18 and HPV6 and remained at the same level as those obtained in the placebo group for HPV11. Vaccination against HPV6 appears to yield a far higher antibody titer when given between the ages of 9 and 15 than in older subjects. The issue of immunological competition among viral types included in the candidate vaccines will have to be examined. We do not currently know the precise correlation between neutralizing antibody titers and the degree of clinical protection. However, neutralizing antibodies remain at a high plateau 4 years after vaccination, suggesting durable protection, especially against HPV16. Natural exposure might itself have a booster effect. Long-term studies are needed to determine whether booster injections are necessary. Groups at Risk? The main group at risk is the population of immunocompromised patients such as HIV infection, autoimmune disorders and immunosuppressive therapy. Highly active antiretroviral therapy has reduced the risk of cervical cancer in HIV-infected patients, close to the levels of the general population. However, recurrent HPV infections of the genital tract, resistance to conventional treatments and multifocal lesions are still problematic in this setting. It remains to be shown whether HPV vaccination will be effective in immunodeficient subjects. HPV vaccination could also be beneficial before the outset of immunosuppressive treatment for transplant recipients or patients with autoimmune disorders. As cervical cancer is a potential risk for all sexually active women, and as the cause is HPV immune escape, which cannot be predicted in a given subject, HPV vaccination cannot be designed or proposed only for groups at risk. Indeed, this approach would have no effect in developing countries and only a minor benefit in industrialized countries. Vaccination: Optional or Recommended? Although cervical cancer is a public health problem, national health authorities may or may not recommend HPV vaccination. The acceptability of
Cervical Cancer Prevention by HPV Vaccination
197
HPV vaccination for healthcare professionals and the public will depend on the campaign message: vaccination to prevent a sexually transmitted infection and/or to prevent cervical cancer? If clinical studies confirm that vaccination protects against CIN, would it be acceptable to offer vaccination mainly to avoid HPV infection? The message of an ‘anticancer vaccine’ would clearly have a stronger impact on women over 30, compared with ‘prevention of a sexually transmitted infection’ message. Promotion and Education One barrier to HPV vaccination is the poor awareness of HPV infection and its relationship with cervical cancer among the public. Surveys show that more than 80% of women have no idea what causes cervical cancer [45, 46]. The arrival of an HPV vaccine will clearly strengthen information campaigns on the causes of this malignancy and the means available for preventing it. A large-scale education campaign targeting the public and healthcare professionals will be necessary. The messages must be clear and unambiguous. It will be necessary to clearly distinguish between HPV infection, which is relatively frequent and generally asymptomatic in the general population, and its relatively infrequent complications, i.e. cervical precancer lesions and cancer. Vaccination Campaigns: The Importance of the Coverage Rate In terms of public health, the participation rate in a vaccination campaign is crucial (as in a screening campaign). Mathematical models indicate that, with a vaccine against HPV16 and HPV18 that prevents more than 90% of cancers associated with these virus types, only 25% of cases of cervical cancer would be prevented by 40% population coverage, compared with 38 and 51% with 60 and 80% coverage [43]. Therefore, the impact on cervical cancer is dependent on the virus types included in the vaccine and on the proportion of the population that is vaccinated. Mammographic screening prevents more than one third of deaths from breast cancer [47, 48], while Pap testing prevents 70% of cervical cancer deaths in industrialized countries. A program combining HPV vaccination and screening would likely reduce mortality from cervical cancer by about 90%. If one accepts that screening prevents 70% of cervical cancers and HPV16/18 vaccination prevents 90% of deaths from this cancer, then a combination of HPV vaccination and cytological screening would avoid 90% of deaths due to cervical cancer. Screening in the Vaccine Era One of the most hotly debated topics is the future role of the primary prevention (vaccination) and secondary prevention (screening) in different
Monsonego
198
Table 4. Screening versus prophylactic HPV vaccination – Strengths and weaknesses HPV (16/18) vaccination
Developed countries 1. Coverage: organized vesus voluntary 2. Sensitivity: false negative 35% K g survival 3. Frequency⫹⫹⫹: every 3 years from 20–65 years 4. Compliance: individual/collective responsibility 5. Cost⫹⫹ 6. False positive (cyto-HPV)/reproducibility Stress-anxiety Developing countries: No screening 1. Proven efficacy 2. Widely used in industrialized countries
1. Incomplete protection (HPV16/18 ⬇ 70%) 2. Vaccine coverage 3. Naive (⫹) versus exposed/infected (?) 4. Duration of protection 5. Only protects vaccinated population 6. Requires mass education and training
1. Lack of immunity more than HPV itself causes disease (persistent HPV infection) 2. Vaccination overcomes all weaknesses of screening 3. Good tolerability, immunogicity, efficacy on persistent HPV infection and high-grade CIN (⬎95%) 4. Cost effective
Table 5. Screening and prophylactic HPV vaccination – Strengths and weaknesses
Weakness
Screening ⫹ HPV vaccination
Strengths
Strengths
Weaknesses
Screening
Cost
• • • •
Vaccination ⬎13 years Screening ⬎30 years Longer screening interval (⬎5–8 years) Risk-based screening Optimal protection
countries. As the success of screening campaigns is dependent on strict implementation and on regular screening (every 3 years between the ages of 25 and 65) [49–51], combined vaccination would help guarantee optimal protection (tables 4, 5). In the unvaccinated population, Pap screening, with or without HPV testing, will continue to be the main way of preventing cervical cancer. In the
Cervical Cancer Prevention by HPV Vaccination
199
vaccinated population, various prevention strategies could be envisaged. On the one hand, these include HPV genotyping (HPV16 and HPV18) at the first screening test. Subjects negative for HPV16 and HPV18 could have a smear and a new HPV test after 3–5 years. If the test is negative, screening could be repeated every 5–10 years. On the other hand, if the subject is positive for HPV16 or HPV18 at the first screening test, a smear (and treatment if necessary) or follow-up could be proposed. It is likely that vaccine uptake will increase gradually over the next 40 years and that screening will decline accordingly. Cancer Prevention by HPV Vaccination In countries where screening exists, vaccination will likely lead to a significant reduction in the frequency of abnormal smears, biopsies, treatments, follow-up and costs in the short term. A reduction in cervical cancer among populations at risk who do not participate in screening, as well as among screened populations (30% of cases of invasive cervical cancer), would be one major effect of this new strategy. It will be necessary to compare the cost effectiveness of this new program with that of conventional screening strategies and that of HPV vaccination followed by less frequent screening that starts later in life. It will also be necessary to bear in mind the possibility that women will no longer seek screening because they are reassured by the protective effect of vaccination. Therefore, it will be necessary to avoid confusion between screening and vaccination. This is why vaccination campaigns will have to be accompanied by education and screening campaigns. Vaccination and screening will be intimately linked, in order to optimize prevention. In developing countries, where screening is nonexistent or poorly effective, the impact on cancer will occur later than in industrialized countries. Promotion of mass screening must continue in the meantime. Even if vaccination programs have proven effective in these countries, many obstacles remain, such as the cost and the use of vaccines poorly suited to mass vaccination (need for three injections, refrigeration). Administration by the nasal or oral route would be preferable in these settings. Vaccination of Adult Women, Including Those with Prevalent HPV Infection It has not yet been demonstrated that current vaccines are less effective in women already exposed to HPV. It will be difficult to refuse to vaccinate these women, who are generally strongly motivated. However, immunization could protect against other HPV types not included in the vaccine (cross-protection), reduce viral persistence by preventing self-infection and help control virus
Monsonego
200
transmission to new contacts. The economic impact of these measures will need to be assessed. Outstanding Questions Ongoing phase III trials are expected to provide new information on the natural history of HPV infections [52]. More than 30% of the population seems to have multiple HPV infections, and the natural history of such infections is not known. It remains to be shown whether cervical cancer is due to latent or persistent infection or rather to infection acquired at a more advanced age? Similarly, we do not yet know whether current vaccines might lead to the selection of viral genotypes that are currently rarely associated with cervical cancer. Many other questions are still open. What is the best age for HPV vaccination? Should men be vaccinated? How long does the protection last? Does cross-protection occur? How to vaccinate people in developing countries? How to incorporate vaccination programs into current screening strategies for cervical cancer? It will be necessary to ensure that screening and prevention of other sexually transmitted diseases are not neglected. All these questions will require further research, taking into account the economic situation of individual countries, but they should not delay vaccine approval.
Conclusions
HPV vaccines are well tolerated, immunogenic and effective on the most common HPV infection and their associated diseases. The immunization is robust, but follow-up is currently limited to 4 years and the minimum protective antibody titer is not known. The most effective strategy for preventing cervical cancer by HPV vaccination is to ensure a high coverage of the program. The supplementary benefit of male vaccination is unknown. Screening for cervical cancer should continue in the meantime. Strategies based on both screening and vaccination are being studied. HPV vaccination would probably have a different impact on cervical cancer in different countries (fig. 4). Industrialized countries would probably see a rapid reduction in the number of precancer lesions detected by screening, while developing countries would have to wait longer to be able to see an impact on the incidence and mortality of cervical cancer. Thus, the link between HPV and cervical cancer, discovered only 30 years ago, is shortly to be broken, to the benefit of individual women and public health alike.
Cervical Cancer Prevention by HPV Vaccination
201
HPV vaccination: 2 emerging worlds
Industrial countries Pap screening (CIN⫹⫹)
HPV vaccine
Developing countries Cervical cancer
Fig. 4. Impact of prophylactic HPV vaccination.
References 1 2 3 4 5 6 7 8 9
10 11 12
13 14
Syrjänen K, Syrjänen S: Papillomavirus Infections in Human Pathology. Chichester, Wiley and Sons, 2000, pp 1–615. Patnick J: Screening that failed to work; in Franco E, Monsonego J (eds): New Developments in Cervical Cancer Screening and Prevention. Oxford, Blackwell Science, 1997, pp 200–202. Bosch FX, Lorincz A, Muñoz N, Meijer CJLM, Shah KV: The causal relation between human papillomavirus and cervical cancer. J Clin Pathol 2002;55:244–265. Solomon D, Schiffman M, Tarrone R: Comparison of three management strategies for patients with atypical squamous cells of undetermined significance. J Natl Cancer Inst 2001;93:293–299. Ferlay J, Bray F, Pisani P, Parkin DM: Globocan 2002: Cancer Incidence, Mortality and Prevalence Worldwide. IARC Cancer Base No 5, version 2.0. Lyon, IARC Press, 2004. Pisani P, Parkin DM, Bray F, Ferlay J: Estimates of the worldwide mortality from 25 cancers in 1990. Int J Cancer 1999;83:18–29. Aubin F, Pretet JL, Mougin C, et al: Papillomavirus Humains. Biologie et Pathologie Tumorale. Paris, Editions TEC et DOC, 2003, pp 1–759. Exbrayat C: Col de l’utérus; in Remontet L, Buemi A, Velten M (eds): Evolution de l’Incidence et de la Mortalité par Cancer en France de 1978 à 2000. Paris, INVS, 2003, pp 107–112. Rousseau A, Bohet P, Merlière J, Treppoz H, Heules-Bernin B, Ancelle-Park R: Evaluation du dépistage organisé et du dépistage individuel du cancer du col de l’utérus: utilité des données de l’assurance maladie. Bull Epidémiol Hebdo 2002;19:81–83. Weidmann C, Schaffer P, Hedelin G, Arveux P, Chaplain G, Exbrayat C, et al: L’incidence du cancer du col de l’utérus régresse régulièrement en France. Bull Epidémiol Hebdo 1998;5:17–19. Council of the European Union: Council recommendation of 2 December 2003 on cancer screening. Off J Eur Union 2003;327:34–38. Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, Snijders PJ, Peto J, Meijer CJ, Munoz N: Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189:12–19. Franco EL, Rohan TE, Villa LL: Epidemiologic evidence and human papillomavirus infection as a necessary cause of cervical cancer. J Natl Cancer Inst 1999;91:506–511. IARC Working Group: Human Papillomavirus. IARC Monographs on the Evaluation of Carcinogenics Risk to Humans, vol 64. Lyon, International Agency for Research on Cancer, 1995.
Monsonego
202
15 16 17
18
19
20
21
22 23
24
25
26
27
28
29
30
31 32
Koutsky L: Epidemiology of genital human papillomavirus infection. Am J Med 1997;102:3–8. Schiffman M, Krüger-Kjaer S: Natural history of anogenital human papillomavirus infection and neoplasia. J Natl Cancer Instit Monogr 2003;31:14–19. Franco EL, Villa LL, Sobrinho JP, Prado JM, Rousseau MC, Désy M, et al: Epidemiology of acquisition and clearance of cervical human papillomavirus infection in women from a high-risk area for cervical cancer. J Infect Dis 1999;180:1415–1423. Rozendaal L, Westerga J, van der Linden JC, Walboomers JM, Voorhorst FJ, Risse EK, Boon ME, Meijer CJ: PCR based high risk HPV testing is superior to neural network based screening for predicting incident CIN III in women with normal cytology and borderline changes. J Clin Pathol 2000;53:606–611. Melkert PW, Hopman E, van den Brule AJ, Risse EK, van Diest PJ, Bleker OP, Helmerhorst T, Schipper ME, Meijer CJ, Walboomers JM: Prevalence of HPV in cytomorphologically normal cervical smears, as determined by the polymerase chain reaction, is age-dependent. Int J Cancer 1993;53:919–923. Clifford GM, Gallus S, Herrero R, Munoz N, Snijders PJ, Vaccarella S, Anh PT, Ferreccio C, Hieu NT, Matos E, Molano M, Rajkumar R, Ronco G, de Sanjose S, Shin HR, Sukvirach S, Thomas JO, Tunsakul S, Meijer CJ, Franceschi S, IARC HPV Prevalence Surveys Study Group: Worldwide distribution of human papillomavirus types in cytologically normal women in the International Agency for Research on Cancer HPV prevalence surveys: a pooled analysis. Lancet 2005;366:991–998. Clifford GM, Rana RK, Franceschi S, Smith JS, Gough G, Pimenta JM: Human papillomavirus genotype distribution in low-grade cervical lesions: comparison by geographic region and with cervical cancer. Cancer Epidemiol Biomarkers Prev 2005;14:1157–1164. GM Clifford, Smith JS, Aguado T, Franceschi S: Comparison of HPV type distribution in highgrade cervical lesions and cervical cancer: a meta-analysis. Br J Cancer 2003;89:101–105. Khan MJ, Castle PE, Lorincz AT, Wacholder S, Sherman M, Scott DR, Rush BB, Glass AG, Schiffman M: The elevated 10-year risk of cervical precancer and cancer in women with human papillomavirus (HPV) type 16 or 18 and the possible utility of type-specific HPV testing in clinical practice. J Natl Cancer Inst 2005;97:1072–1079. Castle PE, Solomon D, Schiffman M, Wheeler CM, The ALTS Group: Human papillomavirus type 16 infections and 2-year absolute risk of cervical precancer in women with equivocal or mild cytologic abnormalities. J Natl Cancer Inst 2005;97:1066–1071. Koutsky LA, Holmes KK, Critchlow CW, Stevens CE, Paavonen J, Beckmann AM, DeRouen TA, Galloway DA, Vernon D, Kiviat NB: A cohort study of the risk of cervical intraepithelial neoplasia grade 2 or 3 in relation to papillomavirus infection. N Engl J Med 1992;327:1272–1278. Ho GY, Burk RD, Klein S, Kadish AS, Chang CJ, Palan P, et al: Persistent genital human papillomavirus infection as a risk factor for persistent cervical dysplasia. J Natl Cancer Inst 1995;87: 1365–1371. Dalstein V, Riethmuller D, Pretet JL, Le Bail Carval K, Sautiere JL, Carbillet JP, Kantelip B, Schaal JP, Mougin C: Persistence and load of high-risk HPV are predictors for development of high-grade cervical lesions: a longitudinal French cohort study. Int J Cancer 2003;106: 396–403. Bory JP, Cucherousset J, Lorenzato M, Gabriel R, Quereux C, Birembaut P, Clavel C: Recurrent human papillomavirus infection detected with the hybrid capture II assay selects women with normal cervical smears at risk for developing high grade cervical lesions: a longitudinal study of 3,091 women. Int J Cancer 2002;102:519–525. Schlecht NF, Kulaga S, Robitaille J, Ferreira S, Santos M, Miyamura RA, et al: Persistent human papillomavirus infection as a predictor of cervical intraepithelial neoplasia. JAMA 2001;286:3106–3114. Wallin KL, Wiklund F, Angstrom T, Bergman F, Stendahl U, Wadell G, Hallmans G, Dillner J: Type-specific persistence of human papillomavirus DNA before the development of invasive cervical cancer. N Engl J Med 1999;341:1633–1628. Wang SS, Hildesheim A: Viral and host factors in human papillomavirus persistence and progression. J Natl Cancer Inst Monogr 2003;31:35–40. Monsonego J: Papillomavirus et cancer du col de l’utérus. Médecine/Sciences 1996;12:733–744.
Cervical Cancer Prevention by HPV Vaccination
203
33
34
35
36
37
38
39
40 41
42 43 44 45 46 47
48 49
50
Chellappan S, Kraus VB, Kroger B, Munger K, Howley PM, Phelps WC, Nevins JR: Adenovirus E1A, simian virus 40 tumor antigen, and human papillomavirus E7 protein share the capacity to disrupt the interaction between transcription factor E2F and the retinoblastoma gene product. Proc Natl Acad Sci USA 1992;89:4549–4553. Thomas M, Matlashewski G, Pim D, Banks L: Induction of apoptosis by p53 is independent of its oligomeric state and can be abolished by HPV-18 E6 through ubiquitin mediated degradation. Oncogene 1996;13:265–273. Koutsky LA, Ault KA, Wheeler CM, Brown DR, Barr E, Alvarez FB, Chiacchierini LM, Jansen KU, Proof of Principle Study Investigators: A controlled trial of a human papillomavirus type 16 vaccine. N Engl J Med 2002;347:1645–1651. Villa LL, Costa RL, Petta CA, Andrade RP, Ault KA, Giuliano AR, Wheeler CM, Koutsky LA, Malm C, Lehtinen M, Skjeldestad FE, Olsson SE, Steinwall M, Brown DR, Kurman RJ, Ronnett BM, Stoler MH, Ferenczy A, Harper DM, Tamms GM, Yu J, Lupinacci L, Railkar R, Taddeo FJ, Jansen KU, Esser MT, Sings HL, Saah AJ, Barr E: Prophylactic quadrivalent human papillomavirus (types 6, 11, 16, and 18) L1 virus-like particle vaccine in young women: a randomised double-blind placebo-controlled multicentre phase II efficacy trial. Lancet Oncol 2005;6: 271–278. Harper DM, Franco EL, Wheeler C, Ferris DG, Jenkins D, Schuind A, Zahaf T, Innis B, Naud P, De Carvalho NS, Roteli-Martins CM, Teixeira J, Blatter MM, Korn AP, Quint W, Dubin G, GlaxoSmithKline HPV Vaccine Study Group: Efficacy of a bivalent L1 virus-like particle vaccine in prevention of infection with human papillomavirus types 16 and 18 in young women: a randomised controlled trial. Lancet 2004;364:1757–1765. Munoz N, Bosch FX, Castellsague X, Diaz M, de Sanjose S, Hammouda D, Shah KV, Meijer CJ: Against which human papillomavirus types shall we vaccinate and screen? The international perspective. Int J Cancer 2004;111:278–285. Goldie SJ, Grima D, Kohli M, Wright TC, Weinstein M, Franco E: A comprehensive natural history model of HPV infection and cervical cancer to estimate the clinical impact of a prophylactic HPV-16/18 vaccine. Int J Cancer 2003;106:896–904. Huang K, Lin S: Nationwide vaccination: a success story in Taiwan. Vaccine 2000;18(suppl 1): S35–S38. Goldie SJ, Kohli M, Grima D, Weinstein MC, Wright TC, Bosch FX, Franco E: Projected clinical benefits and cost-effectiveness of a human papillomavirus 16/18 vaccine. J Natl Cancer Inst 2004;96:604–615. Sanders GD, Taira AV: Cost-effectiveness of a potential vaccine for human papillomavirus. Emerg Infect Dis 2003;9:37–48. Kulasingam SL, Myers ER: Potential health and economic impact of adding a human papillomavirus vaccine to screening programs. JAMA 2003;290:781–789. Bozon M: At what age do women and men have their first sexual intercourse? World comparisons and recent trends. Popul Soc 2003;391:1–4. Anhang R, Wright TC Jr, Smock L, Goldie SJ: Women’s desired information about human papillomavirus. Cancer 2004;100:315–320. Anhang R, Stryker JE, Wright TC Jr, Goldie SJ: News media coverage of human papillomavirus. Cancer 2004;100:308–314. Duffy SW, Tabar L, Chen HH, Holmqvist M, Yen MF, Abdsalah S, Epstein B, Frodis E, Ljungberg E, Hedborg-Melander C, Sundbom A, Tholin M, Wiege M, Akerlund A, Wu HM, Tung TS, Chiu YH, Chiu CP, Huang CC, Smith RA, Rosen M, Stenbeck M, Holmberg L: The impact of organized mammography service screening on breast carcinoma mortality in seven Swedish counties. Cancer 2002;95:458–469. Feig SA: Effect of service screening mammography on population mortality from breast carcinoma. Cancer 2002;95:451–457. Monsonego J: Spontaneous screening: benefits and limitations; in Franco E, Monsonego J (eds): New Developments in Cervical Cancer Screening and Prevention. Oxford, Blackwell Science, 1997, pp 220–240. Monsonego J: Enquête nationale sur le dépistage du cancer du col auprès des gynécologues. Gynécol Obstét Prat 1996;81:1–5.
Monsonego
204
51
52
Camatte S, Morice P, Pautier P, Lhommé C, Haie-Maider C, Duvillard P, Castaigne D: Incidence et mortalité du cancer du col en France. Quelle relation avec le dépistage; in Blanc B (ed): Le dépistage du cancer du col de l’utérus. Paris, Springer, 2005, pp 35–45. Garnett GP, Waddell HC: Public health paradoxes and the epidemiological impact of an HPV vaccine. J Clin Virol 2000;19:101–111.
Dr. Joseph Monsonego, MD, Medical Director Eurogin, 174 rue de Courcelles FR–75017 Paris (France) Tel. ⫹33 1 47 66 05 29, Fax ⫹33 1 47 66 74 70, E-Mail
[email protected] Cervical Cancer Prevention by HPV Vaccination
205
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 206–216
Rationale for Human Papillomavirus (HPV) Vaccines F. Xavier Bosch IDIBELL, Institut Català d’Oncologia, Barcelona, Spain
All Cervical Cancer Cases Are Related to the Persistent Presence of HPV DNA
The evidence relating HPV infections to cervical cancer includes a large and consistent body of studies indicating a strong and specific role of the viral infection in all settings where investigations have taken place. The association has been recognized as causal in nature by a number of international review parties since the early 1990s [1, 2]. HPV DNA Prevalence in Cervical Cancer Specimens State of the art amplification techniques used in case-control studies, case series and prevalence surveys have unequivocally shown that, in adequate specimens of cervical cancer, HPV DNA can be detected in 90–100% of the cases, compared with a prevalence of some 5–20% from cervical specimens of women identified as suitable epidemiological controls. In most series, detailed investigations of the few cervical cancer specimens that appear as HPV DNA negatives have occasionally been conducted, and the results strongly suggest that these are largely false negatives. As a consequence, the claim has been made that this is the first necessary cause of a human cancer ever identified, providing a strong rationale for the use of HPV tests in screening programs and the development of HPV vaccines [3, 4]. Risk Estimates from Case-Control Studies Observations from preinvasive disease and cohort studies on the HPV natural history have intrinsic limitations for making inferences on cervical cancer causality, because in trials or research contexts, biological progression is not allowed to continue beyond the stage of high-grade squamous intraepithelial
Carcinoma
OR (95% CI)
Control 97.0
Brazil
177 (65.5–478.3)
17.3 96.9
Mali
109.2 (10.6–1,119.0)
33.3 97.1
Morocco
113.7 (42.3–305.3)
21.6 98.1
Paraguay
208.1 (46.4–932.8)
19.8 96.4
Philippines
276.8 (139.7–548.3)
9.2 96.5
Thailand
163.5 (82.0–325.9)
15.7 95.3
Peru
115.9 (48.6–276.4)
17.7 82.4
Spain
75.7 (32.9–174.2)
5.9 78.4
Colombia
17.7 ( 9.1–34.3)
17.5 94.0
Overall
91.4 (71.2–117.4)
14.9 0
10
20
30
40
50
60
70
80
*OR adjusted by country and age-group
90 100 0.1
1
10
100
1,200
OR* (logarithm)
Fig. 1. HPV DNA prevalence by case-control status and OR for the association between papillomavirus DNA detection and risk of squamous cell carcinoma of the cervix in 9 cases. OR ⫽ Odds ratio; CI ⫽ confidence interval.
lesions (HSIL)/cervical intraepithelial neoplasia (CIN)3 or carcinoma in situ. It is of importance to notice that the information applicable to cervical cancer primarily comes from case-control studies in which the target disease is directly investigated. In an effort to simplify the vast literature, the results of the International Agency for Research on Cancer (IARC) multicenter case-control study on invasive cervical cancer will be used as example. In brief, this project included 9 case-control studies performed in different parts of the world, mostly in highrisk countries. A common protocol and questionnaire were used, and HPV DNA testing was done in two central research laboratories using the MYO9/11 and the general primer GP5⫹/6⫹ polymerase chain reaction (PCR) testing systems. Figure 1 shows the summary results of the HPV DNA prevalence in case and control specimens, the risk estimates and their confidence intervals (CIs) for squamous cell carcinomas [5]. The figure shows very high odds ratios (ORs), with estimates in the range of 50–150 and various estimates in the several-hundred range. These risk estimates lead to calculations of attributable fractions of greater than 95% for the entire study [1].
Rationale for HPV Vaccines
207
In the literature, results are strikingly consistent for preinvasive lesions, for squamous cell carcinomas and adenocarcinomas, as well as for studies that tested for HPV DNA as a group, or studies that restricted the analyses to highrisk HPV types. Studies that compared risk factors for CIN3 and invasive cancer have not reported any significant differences in their associations with HPV or the epidemiological profile. The pool of IARC studies was large enough to provide type-specific risk estimates for 18 types. Restricting the analyses to the studies that used the GP5⫹/6⫹ HPV detection system and to squamous cell carcinomas, the adjusted ORs for HPV DNA detection (the factor by which the reference risk of cervical cancer is multiplied if HPV DNA is detected) was 158.2 (95% CI 113.2–220.6). The risk estimates for adenocarcinomas was OR ⫽ 76.3 (95% CI 40.7–143.2). Type-specific risk estimates for squamous cell cancer were as follows: HPV16: OR ⫽ 435; HPV18: OR ⫽ 248; HPV45: OR ⫽ 198; HPV31: OR ⫽ 124; HPV52: OR ⫽ 200; HPV33: OR ⫽ 374; HPV58: OR ⫽ 115; HPV35: OR ⫽ 74; HPV59: OR ⫽ 419; HPV51: OR ⫽ 67; HPV56: OR ⫽ 45; HPV39: infinity; HPV68: OR ⫽ 54. The risk of any given high-risk type was not statistically different from the risk reported for HPV16. Likewise, the risk related to the presence of multiple HPV types in the specimen is no different from the risk linked to a single HPV type. The proportion of multiple types in a given specimen varies across studies and particularly in relation to the HPV detection method used. However, in all studies of invasive carcinoma, the risk linked to multiple HPV types does not vary significantly from the risk linked to single HPV types. These studies and a recent international review concluded that the evidence is now sufficient to consider types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68, 73 and 82 as high-risk carcinogenic HPVs [2]. A second group of HPVs are rarely found and have been classified as low-risk HPVs, including types 6, 11, 40, 42, 43, 44, 54, 61, 70, 72, 81 and CP6108. From the IARC and other studies, a small group of HPV types remains in the category of uncertain risk, currently including HPV26, HPV53, HPV66 and perhaps others [5].
The Role of Other Risk Factors Is Conditional to the Presence of HPV DNA, Probably Modulating Progression Rates
Most of the sexual behavior parameters that were linked to cervical cancer in the past are being re-evaluated in studies that considered the strong influence of the presence of HPV. Soon after the introduction of HPV testing in research protocols, it became clear that the key risk factors that reflected sexual behavior, such as the number of sexual partners, merely reflected the probability of HPV exposure. Because of the growing evidence that HPV is a necessary factor in cervical
Bosch
208
Table 1. Enviromental risk factors for cervical cancer among HPV-positive women Risk factor
Risk exposure
Reference
HPV DNA in cervical exfoliates Use of OCs Smoking Parity Chlamydia trachomatis HSV2
Positive for high-risk types 5 or more years of use Ever 5 or more pregnancies Antibody positive Antibody positive
Negative Never Never None or 1–2 Antibody negative Antibody negative
HSV2 ⫽ Herpes simplex virus 2; OC = oral contraceptives. Table from Bosch et al. [1]. Reproduced with permission from the BMJ Publishing Group.
cancer, it soon became a standard procedure in the reports of case-control studies to include analyses restricted to HPV-positive cases and controls to properly assess the contribution of additional factors to the risk of cervical cancer. In relation to invasive cervical cancer, the pooled HPV-positive restricted analyses of the IARC included 1,768 cases and 262 controls, and the key findings concerning environmental risk factors are summarized in table 1 [for a review, see [6]]. Long-Term Use of Oral Contraceptives Long-term use of oral contraceptives (OCs) was associated with a significant increase in risk of cervical cancer (OR ⫽ 1.47, 95% CI 1.02–2.12). The use of OCs for less than 5 years was not related to cervical cancer (OR ⫽ 0.77, 95% CI 0.46–1.29), but the risk increased significantly for 5–9 years (OR ⫽ 2.72, 95% CI 1.36–5.46) and for ⱖ10 years (OR ⫽ 4.48, 95% CI 2.24–9.36). Evidence for an association of cervical cancer with the use of oral or other hormonal contraceptives is not entirely consistent. A number of studies that investigated HPV-positive women found no associations or only weak associations with HSIL/CIN3 in subgroup analyses. These apparently conflicting results may reflect the increased cytological surveillance of women that are taking OCs in developed countries and the use of different case definitions in cohort studies, i.e. from atypical cells of undetermined significance (ASCUS) up to HSIL/CIN3, as opposed to cervical cancer [for a review, see [6]]. Because of the potential public health importance of an interaction between long-term use of OCs and HPV infections in the development of cervical cancer, efforts are now being devoted to verify the results in different populations. A recent meta-analysis on the association between hormonal contraceptives and cervical cancer concludes that there is a linear dose-response relationship and that the effect tends to return to average with time after OC cessation. The relevant units at the World Health Organization examined the evidence with an international working
Rationale for HPV Vaccines
209
group that recognized the significance of the association. However, the group also concluded that the benefits currently achieved by the use of OCs in developing countries (i.e. unwanted pregnancy avoidance) outbalance the increase in risk and should not force a change in the current family planning strategies. High Parity HPV-positive women who reported 7 or more full-term pregnancies had a 4-fold increased risk of cervical cancer as compared with similar HPV-positive women that were nulliparous (OR ⫽ 3.8, 95% CI 2.7–5.5). There was still a 2-fold increased risk when women reporting 7 or more pregnancies were compared with HPV-positive women who reported 1–2 full-term pregnancies. Similar results were obtained in Costa Rica and Thailand, as well as among women with preinvasive disease in the Portland cohort study. In Denmark and in the Manchester cohort study, two populations with low parity, effects are less visible for preneoplastic lesions [for a review, see [6]]. It has been speculated that the general reduction in the average number of births in developed countries over the last decades may have contributed to the reduction in cervical cancer incidence, but formal proof of the hypothesis has not yet been produced. Cigarette Smoking The pooled results of the IARC studies found that ‘ever smoking’ was associated with a 2-fold, statistically significant, increased risk of cervical cancer with a significant dose response [7]. These findings are consistent with those for ‘current versus never smoking’ among HPV-positive women for preneoplastic cervical lesions in the Costa Rica study (OR ⫽ 2.3), the Portland study (OR ⫽ 2.7 for CIN2–3), the Copenhagen study (OR ⫽ 1.9) and the Manchester study (OR ⫽ 2.2). These recent studies are providing growing evidence on the carcinogenic effect of cigarette smoking in women with persistent HPV infection [for a review, see [6]]. The monograph program at the IARC reviewed the evidence in 2002 and concluded that smoking was an independent risk factor for cervical cancer [8]. However, the mechanisms by which cigarette smoking may affect cervical cancer (i.e. a direct effect of the tobacco metabolites, an indirect effect related to tobacco-induced immunosuppression or to reduced dietary antioxidants) remain elusive. Coinfection with the Human Immunodeficiency Virus Evidence for a putative interaction between HPV and human immunodeficiency virus (HIV) in the origin of cervical cancer was formally recognized when cervical cancer was considered as one of the criteria of the acquired immune deficiency syndrome (AIDS) among HIV-positive women. The subsequent literature
Bosch
210
largely confirmed the evidence, although some major confounders of the epidemiological association tend to obscure the results. In brief, the confounders refer to the powerful impact of screening in some populations, the medical surveillance of HIV carriers in developed countries and the short survival time of HIV/AIDS patients in many populations at high risk of cervical cancer in relation to the time intervals between HPV infection and cervical cancer. The risk of invasive cervical cancer in HIV carriers has been intensively investigated. In the year 2000, the International Collaboration on HIV and Cancer group [9] published cancer data from 23 prospective studies that included 47,936 HIV-infected subjects from North America, Europe and Australia for the period 1992–1999. The study concluded that there had not been a significant change in the incidence of invasive cancer (OR ⫽ 1.87, 95% CI 0.77–4.56) during this period. An overview of early African studies concluded that invasive cancer was not related to exposure to HIV with a summary OR of 0.8 (95% CI 0.5–1.4) from studies carried out in Rwanda, South Africa and Uganda. However, recent data from a hospital-based case-control study in South-Africa identified an increased risk of cervical cancer (OR ⫽ 1.6, 95% CI 1.1–2.3) and cancer of the vulva (OR ⫽ 4.8, 95% CI 1.9–12.2) among HIV-infected patients [10]. In the United States or Europe, reports are generally consistent in detecting an increased risk of cervical cancer among HIV-infected women. Selik and Rabkin [11] from the USA reported a relative risk of cervical cancer of 5.5 among HIV-positive women. Frisch et al. [12] used data from the US Cancer Match Registry for the period 1978–1996 and showed a relative risk of invasive cervical cancer of 5.4 among HIV-positive women compared with the general population in the United States. Similar increases in magnitude were observed for in situ cervical cancer (OR ⫽ 4.6), cancer of the vagina and vulva (OR ⫽ 5.8) and anal cancer (OR ⫽ 6.8). Rates were evaluated at the time prior to AIDS diagnosis, around the period of diagnosis and up to 60 moths after the AIDS diagnosis. The authors identified no major changes in risk before and after diagnosis. Data are also available from the population of New York, with similar results. In southern Europe, a strong association between cervical cancer and AIDS has consistently been found. In Italy, the linkage of the National AIDS Registry and the population cancer registries showed a 15-fold increased risk of invasive cervical cancer for women with AIDS [13]. The joint Italian-French follow-up study of HIV-positive women also showed a 13-fold increased rate of cervical cancer [14]. In Spain, the Catalonian AIDS surveillance system detected 58 cases of invasive cervical cancer among 823 HIV-positive women, an 18-fold increased risk compared with the general population [15]. In summary, HPV and HIV share some behavioral traits that define a particularly vulnerable high-risk group. Progression of HPV infections to CIN lesions and cervical cancer in the context of limited/absent screening seems to
Rationale for HPV Vaccines
211
be increased among HIV carriers and AIDS patients. Furthermore, evidence is growing suggesting that progression is related to the severity of the immunosuppression as indicated by CD4⫹ counts. Coinfection with Other Sexually Transmitted Agents Markers of exposure to other sexually transmitted agents have repeatedly been associated with cervical cancer. Results from the multicenter study of the IARC found a 2-fold increased risk in cervical cancer of the presence of antibodies to Chlamydia trachomatis (OR ⫽ 2.1, 95% CI 1.1–4.0) and herpes simplex type 2 viruses [for a review, see [6]]. Nonspecific inflammatory changes have also been related to modest increases in risk of preneoplastic cervical lesions among HPV-positive women. The difficulties with the evaluations of such factors lies in the strong colinearity observed among all sexually transmitted disease and the limitations of some of the biomarkers currently used to assess ever exposure or persistent exposure.
HPV Types Involved in Cervical Cancer and Precancerous Lesions: Relevance of HPV16 and HPV18
In most countries, of the more than 35 HPV types found in the genital tract, HPV16 accounts for some 50–60% of the cervical cancer cases, followed by HPV18 (10–20%), HPV45 (4–8%) and HPV31 (1–5%). Figure 2 shows the cumulative distribution of the 15 HPV types in a series of close to 3,000 cervical cancer cases [16]. Of these, the 5 most common types (HPV16, HPV18, HPV45, HPV31 and HPV33) account for 80% of the distribution in squamous cell cancers and for 94% in adenocarcinomas. In most studies, HPV18 predominates in adenocarcinomas in absolute or relative terms. The reasons for such specificity are unknown. The selection of HPV16 in cervical cancer cases translates the biological advantage of this type in fully expressing its oncogenic capacity. An interesting study among women with different degrees of HIV-induced immunosuppression pointed at the likely increased ability of HPV16 to escape immunesurveillance as compared with other HPV types as one possible mechanism of such advantage [17].
Association between HPV and Cervical Cancer Is Consistent Worldwide
The distribution of HPV types in cervical cancer in five different regions in the world is described in table 2. It is clear that HPV16 and HPV18 consistently
Bosch
212
Cumulative attributable faction (%) 53.5
53.5
16
17.2
18
70.7 77.4
6.7
45
80.3
2.9
31
82.9
2.6
33
85.2
2.3
52
87.4
2.2
58
88.8
1.4
35
90.1
1.3
59 56
1.2
91.3
51
1.0
92.3
39
0.7
93.0
6 _8 _
0.6
93.6
73
0.5
94.1
82
0.3
94.4 95.6
1.2
Other
4.4
X 0
20
40
60
80
100.0
100
Fig. 2. Percentage of cervical cancer cases attributed to the most frequent HPV genotypes in women 15 years of age and older, all world regions combined. Figure adapted with permission from Muñoz et al. [16], Copyright 2004 John Wiley & Sons.
Table 2. Prevalence of the most common HPV types in cervical cancer by region Sub-Saharan Africa
Northern Africa
Central-South America
South Asia
Europe and North America
HPV type
%
HPV type
%
HPV type
%
HPV type
%
HPV type
%
HPV16 HPV18 HPV45 HPV33 HPV58
47.7 19.1 15.0 3.2 3.2
HPV16 HPV18 HPV45 HPV33 HPV31
67.6 17.0 5.6 4.0 3.4
HPV16 HPV18 HPV31 HPV45 HPV33
57.0 12.6 7.4 6.8 4.3
HPV16 HPV18 HPV45 HPV52 HPV58
52.5 25.7 7.9 3.1 3.0
HPV16 HPV18 HPV45 HPV31 HPV56
69.7 14.6 9.0 4.5 2.2
are the first two types involved in all regions so far explored. HPV45 seems to be a natural candidate for the third ranking place, and some additional variability exists thereafter for the relatively rarer HPV types. The essential traits of the epidemiology of the viral infection and the association for cervical cancer have
Rationale for HPV Vaccines
213
been described and are essentially universal. Some behavioral and cultural traits that vary across cultural environments (i.e. in relation to age at sexual debut) may have an impact in relation to the introduction and acceptability of the vaccine and the ages at which to vaccinate. However, the bulk of the evidence strongly suggests that the current vaccines under evaluation are expected to be recognized as a local priority worldwide.
Current Vaccines under Evaluation
The identification of HPV as a necessary cause of cervical cancer has prompted the development of HPV prophylactic vaccines based upon the properties of one of the most conserved regions in the HPV genome (the L1 region) which is capable of self-assembly as a three-dimensional structure when expressed in vitro (viral-like particles without the oncogenic fragments of the genome) and induces strong immunological responses. Preliminary results are very encouraging. An HPV16 virus-like particlebased vaccine showed immune responses in virtually all vaccinated women: antibody titers were up to 40 fold the level acquired after natural infection and vaccinated women were fully protected against HPV16 infections, persistent HPV16 infections and HPV16-related CIN [18]. A more developed vaccine based upon similar principles and including HPV16 and HPV18 antigens showed equivalent results for both HPV types [19]. Further, one of these vaccines also includes HPV types 6 and 11, responsible for over 90% of the cases of genital warts and the rare disease recurrent respiratory papillomatosis. The protective efficacy of these HPV vaccines for CIN2/3 lesions is currently under evaluation with promising preliminary results [20]. Second-generation vaccines are being developed with the objective of including additional HPV types related to cervical cancer, facilitating administration and incorporating other traits that would facilitate introduction into the countries at higher risk. It is thus likely that the future of cervical cancer prevention will include prophylactic HPV vaccines in the adolescent women followed by a novel type of screening, likely HPV based, with cervical cytology colposcopy and pathology as the triage and diagnostic tools. In developing countries, where regular screening has proven to be largely unrealistic, perhaps a booster vaccination step may be advisable. Therapeutic vaccines may offer interesting alternatives in populations where a relevant fraction of young adult women are already permanent carriers of HPV DNA. These products incorporate modified fragments of the E6 and/or E7 genes, the viral products consistently expressed in persistent infections and in cervical cancer. Chimeric virus-like particles have been shown to induce antigen-specific protection in
Bosch
214
mice from challenge with E7-expressing tumor cells. However, human studies are still in their early phases. Conclusions
In the last 2 decades, the etiology of cervical cancer has become a coherent description that includes the identification of a limited group of HPV types as the necessary, etiologic agents, with a few additional cofactors intervening as such in the presence of viral DNA. The association is universal and the HPV type variability is limited. The clinical implications of these findings have resulted in novel screening and vaccination strategies for the prevention of cervical cancer. Current vaccines may prompt a change in the paradigm of cervical cancer prevention.
Acknowledgements We acknowledge Meritxell Nomen and Cristina Rajo who were responsible for the secretarial workload. Partial support has been received from the Fondo de Investigaciones Sanitarias, Spain (FIS PI030240 and FIS 01/1237), from the European Commission (QLG4CT-2000-01238 and QLG4-CT-2001-30142), from the Agència de Gestió d’Ajuts Universitaris I de Recerca (2005SGR00695), and from the Instituto de Salud Carlos III (Red de Cancer RCESP C03/09 and Red de Salut Pública RTICCC C03/10).
References 1 2 3
4
5
6 7
8
Bosch FX, Lorincz A, Muñoz N, Meijer CJLM, Shah KV: The causal relation between human papillomavirus and cervical cancer. J Clin Pathol 2002;55:244–265. International Agency for Research on Cancer: IARC Handbooks of Cancer Prevention. Cervix Cancer Screening. Lyon, IARC Press, 2005. Bosch FX, Manos MM, Muñoz N, Sherman M, Jansen AM, Peto J, Schiffman MH, Moreno V, Kurman R, Shah KV: Prevalence of human papillomavirus in cervical cancer: a worldwide perspective. International Biological Study on Cervical Cancer (IBSCC) Study Group. J Natl Cancer Inst 1995;87:796–802. Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, Snijders PJ, Peto J, Meijer CJ, Muñoz N: Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189:12–19. Muñoz N, Bosch FX, de Sanjose S, Herrero R, Castellsague X, Shah KV, Snijders PJ, Meijer CJ: Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518–527. Castellsague X, Muñoz N: Chapter 3: cofactors in human papillomavirus carcinogenesis – role of parity, oral contraceptives, and tobacco smoking. J Natl Cancer Inst Monogr 2003;20–28. Plummer M, Herrero R, Franceschi S, Meijer CJ, Snijders P, Bosch FX, de Sanjose S, Muñoz N: Smoking and cervical cancer: pooled analysis of the IARC multi-centric case-control study. Cancer Causes Control 2003;14:805–814. IARC monographs series. Vol 83. Tobacco smoke and involuntary smoking. Lyon, IARC Press, 2004.
Rationale for HPV Vaccines
215
9 10
11 12
13
14
15
16
17
18 19
20
Highly active antiretroviral therapy and incidence of cancer in human immunodeficiency virusinfected adults. J Natl Cancer Inst 2000;92:1823–1830. Sitas F, Pacella-Norman R, Carrara H, Patel M, Ruff P, Sur R, Jentsch U, Hale M, Rowji P, Saffer D, Connor M, Bull D, Newton R, Beral V: The spectrum of HIV-1 related cancers in South Africa. Int J Cancer 2000;88:489–492. Selik RM, Rabkin CS: Cancer death rates associated with human immunodeficiency virus infection in the United States. J Natl Cancer Inst 1998;90:1300–1302. Frisch M, Biggar RJ, Goedert JJ: Human papillomavirus-associated cancers in patients with human immunodeficiency virus infection and acquired immunodeficiency syndrome. J Natl Cancer Inst 2000;92:1500–1510. Franceschi S, dal Maso L, Arniani S, Crosignani P, Vercelli M, Simonato L, Falcini F, Zanetti R, Barchielli A, Serraino D, Rezza G: Risk of cancer other than Kaposi’s sarcoma and non-Hodgkin’s lymphoma in persons with AIDS in Italy. Cancer and AIDS Registry Linkage Study. Br J Cancer 1998;78:966–970. Serraino D, Carrieri P, Pradier C, Bidoli E, Dorrucci M, Ghetti E, Schiesari A, Zucconi R, Pezzotti P, Dellamonica P, Franceschi S, Rezza G: Risk of invasive cervical cancer among women with, or at risk for, HIV infection. Int J Cancer 1999;82:334–337. Vall Mayans M, Maguire A, Miret M, Casabona J: Disproportionate high incidence of invasive cervical cancer as an AIDS-indicative disease among young women in Catalonia, Spain. Sex Trans Dis 1999;26:500–503. Muñoz N, Bosch FX, Castellsague X, Diaz M, de Sanjose S, Hammouda D, Shah KV, Meijer CJ: Against which human papillomavirus types shall we vaccinate and screen? The international perspective. Int J Cancer 2004;111:278–285. Strickler HD, Palefsky JM, Shah KV, Anastos K, Klein RS, Minkoff H, Duerr A, Massad LS, Celentano DD, Hall C, Fazzari M, Cu-Uvin S, Bacon M, Schuman P, Levine AM, Durante AJ, Gange S, Melnick S, Burk RD: Human papillomavirus type 16 and immune status in human immunodeficiency virus-seropositive women. J Natl Cancer Inst 2003;95:1062–1071. Koutsky LA, Ault KA, Wheeler CM, Brown DR, Barr E, Alvarez FB, Chiacchierini LM, Jansen KU: A controlled trial of a human papillomavirus type 16 vaccine. N Engl J Med 2002;347:1645–1651. Harper DM, Franco EL, Wheeler C, Ferris DG, Jenkins D, Schuind A, Zahaf T, Innis B, Naud P, De Carvalho NS, Roteli-Martins CM, Teixeira J, Blatter MM, Korn AP, Quint W, Dubin G: Efficacy of a bivalent L1 virus-like particle vaccine in prevention of infection with human papillomavirus types 16 and 18 in young women: a randomised controlled trial. Lancet 2004;364: 1757–1765. Villa LL, Costa RL, Petta CA, Andrade RP, Ault KA, Giuliano AR, Wheeler CM, Koutsky LA, Malm C, Lehtinen M, Skjeldestad FE, Olsson SE, Steinwall M, Brown DR, Kurman RJ, Ronnett BM, Stoler MH, Ferenczy A, Harper DM, Tamms GM, Yu J, Lupinacci L, Railkar R, Taddeo FJ, Jansen KU, Esser MT, Sings HL, Saah AJ, Barr E: Prophylactic quadrivalent human papillomavirus (types 6, 11, 16, and 18) L1 virus-like particle vaccine in young women: a randomised double-blind placebo-controlled multicentre phase II efficacy trial. Lancet Oncol 2005;6:271–278.
F. Xavier Bosch, MD, PhD, MPH IDIBELL, Institut Català d’Oncologia Epidemiology and Cancer Registration Unit Avda. Gran Via s/n Km. 2,7 ES–08907 L’Hospitalet de Llobregat/Barcelona (Spain) Tel. ⫹34 93 2607812, Fax ⫹34 93 2607787, E-Mail
[email protected] Bosch
216
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 217–226
Perspectives on HPV Virus-Like Particle Vaccine Efficacy John T. Schiller, Douglas R. Lowy Laboratory of Cellular Oncology, National Cancer Institute, National Institutes of Health, Bethesda, Md., USA
In the last 15 years, vaccines to prevent papillomavirus infection have moved from proof of concept immunogenicity studies in laboratory animals to large phase 3 clinical efficacy trials. There is wide-spread optimism that vaccines to prevent the sexually transmitted infections by the HPV types that are the major cause of cervical cancer will receive regulatory approval for commercial sales within the next year or so. This chapter will provide a brief overview of the issues surrounding the preclinical development, clinical efficacy trials and potential effectiveness of the vaccines in general use.
The Historical Perspective
Vaccines to prevent many viral infections, including small pox, measles, mumps, rubella and polio, have been enormously successful [1]. These vaccines are thought to be effective mainly because they consistently elicit high titers of neutralizing antibodies to the surface of the killed or attenuated virions that constitute the vaccine [2]. It was difficult to develop papillomavirus vaccines based on this concept, because the virus could not be efficiently grown in culture and the virions contain oncogenes [3]. Thus, subunit vaccines based on isolated expression of the individual capsid proteins were pursued. Early attempts to develop vaccines based on the L1 major capsid proteins were unsuccessful because L1 was isolated in a denatured form, and neutralizing antibodies to L1 recognize conformation-dependent virion surface epitopes [4]. Significant clinical and commercial development of prophylactic papillomavirus vaccines began only with the discovery that L1 has the intrinsic capacity to self-assemble
into virus-like particles (VLPs) that were morphologically indistinguishable from authentic virion and able to induce high titers of neutralizing antibodies after low-dose injection [5]. Assembly into VLPs did not necessarily imply the ability to efficiently induce neutralizing antibodies. For instance, the major capsid protein of human B19 parvovirus was known to self-assemble into VLPs, but induction of neutralizing antibodies depended on coassembly of a minor capsid protein [6]. Induction of neutralizing antibodies by a papillomavirus VLP was first shown for bovine papillomavirus (BPV) type 1 VLPs, for which authentic virions and a quantitative in vitro neutralization assay were available [5]. Initial studies with HPV16 were hampered by the unappreciated fact that the L1 gene of the prototype HPV16 clone contained a point mutation that inhibited proper assembly into VLPs, leading to the erroneous conclusion that L1 alone could not self-assemble into VLPs [7]. Nevertheless, the efficient assembly of L1 into VLPs and the type-specific neutralizing activity of sera raised against them were subsequently demonstrated for wild-type HPV16 L1 [8, 9] and also HPV11 [10, 11]. Antibodies raised against the mutant L1 of the prototype HPV16 were not neutralizing [9]. Because infectious HPVs do not induce morphological changes when exposed to animal tissues, virus challenge studies of VLP vaccines in animals were conducted using animal papillomavirus, specifically cottontail rabbit papillomavirus in domestic rabbits, canine oral papillomavirus in dogs, and BPV4 in calves [12–14]. In each study, low-dose L1 VLP vaccination, even without adjuvant, induced excellent protection from challenge with the homologous virus. However, none of these models involved sexual transmission or cervicovaginal challenge, and thus, it was unclear how relevant their findings would be to the HPV vaccines in the clinical trials discussed below.
VLP Vaccine Formulations
HPV VLP-based vaccines are under commercial development by two pharmaceutical companies, GlaxoSmithKline (GSK) and Merck. Although both vaccines are based on L1 VLPs, their production and composition have some notable differences. The GSK vaccine is bivalent containing the VLPs of HPV16 and HPV18, the types that cause approximately 70% of cervical cancer, and are produced in L1 recombinant baculovirus-infected insect cells. It also contains their proprietary adjuvant ASO4, which is composed of aluminum salts plus monophosphoryl lipid A, a detoxified form of LPS [15]. The Merck vaccine is tetravalent, composed of HPV6, HPV11, HPV16 and HPV18 VLPs and is produced in L1 recombinant Saccharomyces cerevisiae [16]. Therefore, it is designed to combat two distinct diseases, cervical cancer and genital warts, of
Schiller/Lowy
218
Table 1. Proof of concept HPV VLP prophylactic efficacy trials1 Study
Koutsky et al. [18], 2002
Mao et al. [19], 2006
Villa et al. [16], 2005
Harper et al. [15], 2004
VLP types Adjuvant Sponsor Trial sites Age, years ATP, n Vaccination schedule months Follow-up, years Persistent infections Cont/Vac Efficacy, % CIN1⫹ Cont/Vac Efficacy, %
16 Alum Merck US 16–23 1,533 0, 2, 6 1.5
16 Alum Merck US 16–23 1,505 0, 2, 6 3.5
6, 11, 16, 18 Alum Merck US, EU, BR 16–23 468 0, 2, 6 2.5
16, 18 ASO4 GSK US, CA, BR 15–25 721 0, 1, 6 1.5
42/0 100
111/72 94
36/43 90
7/0 100
9/0 100
24/0 100
3/0 100
6/0 100
ATP ⫽ According to protocol; Con/Vac ⫽ controls/VLP vaccines. 1 According to protocol, analysis for types included in the vaccines. 2 Nineteen of 111 controls and 7 of 7 vaccinees were DNA positive only at the last visit. 3 Ten of 36 controls and 3 of 4 vaccinees were DNA positive only at the last visit.
which approximately 90% are caused by types 6 and 11. The Merck vaccine contains a standard aluminum salts adjuvant. Early clinical trials indicated that HPV VLPs are highly immunogenic even in the absence of adjuvant [17]. Addition of adjuvants to the GSK and Merck vaccines probably serves two functions, stabilization of the particles during storage and induction of peak antibody titers at somewhat lower doses. In the clinical trials described below, the vaccines were injected intramuscularly in three doses at 0, 1 or 2, and 6 months. The doses usually contained 20–40 g per VLP type.
Clinical Efficacy Trials
Four publications have reported the results of randomized and placebocontrolled phase 2b clinical efficacy studies of the VLP vaccines of both companies in young women (table 1) [15, 16, 18, 19]. All are uniformly encouraging in demonstrating excellent safety, exceptional immunogenicity and outstanding efficacy in preventing type-specific cervical infection and cervical dysplasia
Prophylactic Vaccines
219
(CIN) in women who were HPV DNA and seronegative for the corresponding VLP types in the vaccines. The reactogenicities of the vaccines were only slightly greater than that of the adjuvant alone placebo, they did not adversely influence compliance in the vaccine arm, and there were no vaccine-related serious adverse events reported. Seroconversion was greater than 99% in all studies. Type-specific protection from confirmed persistent infection and development of CIN approached 100% in women receiving the full course of three injections (table 1). In modified intention to treat analyses, there were also indications of protection, even if women did not complete the full course of vaccinations. Ongoing phase 3 trials have two common primary endpoints, type-specific intermediate and high-grade cervical dysplasia (CIN2⫹) and type-specific persistent cervical infection, operationally defined as detection of the same HPV DNA type in successive cervical samples at least 6 months apart. Merck has completed an interim analysis of a 17,000-woman multicentric phase 3 trial in young women. The results from this larger cohort, presented at recent conferences, confirm the safety, immunogenicity and efficacy results of the phase 2b studies. In addition, it was reported that the tetravalent vaccine also provides 100% protection against genital warts in women. These results indicate that the vaccine can protect against genital cutaneous as well as genital mucosal infections. There are two ongoing phase 3 trials of the GSK vaccine. One is an 18,000woman multicentric trial, sponsored by GSK, and the other is a population-based study in 7,000 Costa Rican women, sponsored by the US National Cancer Institute and the Costa Rican government. Initial results from the latter trials are expected in the next year or two.
Outstanding Vaccine Efficacy Questions
To date, the results from the VLP vaccine trials are uniformly encouraging with regard to the safety and efficacy of the vaccine. However, there are several questions that will need to be answered before the overall potential for public health impact can be fully appreciated. The first, and perhaps most critical, unknown factor is the duration of protection. Will the vaccines provide longterm protection or will periodic boosting be needed? The results of the first three publications in table 1 and the Merck phase 3 study are for follow-up of only 1.5–2.5 years. The study by Mao et al. [19] best addresses the question of the duration of protection. Encouragingly, it reports 100% protection against type-specific CIN2⫹ even 3.5 years after the last vaccination. As expected, vaccine-induced antibody titers decreased 10–20 fold in the first year or two, but had essentially reached a plateau by years three to four. Since an acceleration in
Schiller/Lowy
220
the rate of decrease in antibody titers would not be expected after 4 years, these findings support the possibility of long-term protection. It is important to note that although this trial employed an HPV16 monovalent vaccine, the results are probably relevant for the current tetravalent vaccine, in that the antibody responses to the HPV16 VLPs were at least as strong with the tetravalent vaccine as they were with the HPV16 only vaccine. Duration of protection could in part be influenced by whether the virus-neutralizing antibodies are primarily the result of transudation of antibodies from the serum into the intact cervicovaginal tract or due to direct exudation of serum antibodies at sites of infection promoting trauma. Duration of protection may be shorter if transudation is the primary mechanism, since levels of vaccine-induced antibodies in cervical mucus are about 10-fold lower than they are in the serum [20]. A second question is the effect of vaccination on women with prevalent infection. If the establishment of a persistently infected cervical lesion with a potential for malignant progression normally takes several rounds of autoinoculation, as suggested by prospective data from Winer et al. [21], then it is possible that vaccination of women with newly acquired infection will decrease the probability that their infections will persist and progress. However, one would not expect that cell-mediated immune responses to VLPs, which are known to be induced after vaccination [22], would induce regression of wellestablished or progressed lesions. This is because L1 is not detectably expressed in the basal cells of productively infected epithelia or anywhere in high-grade cervical dysplasias [3]. Consistent with a possible effect on prevalent infection, Mao et al. [19] reported that there was a trend for HPV16 DNApositive/seronegative vaccinees to clear their infection more readily than controls, but this was not the case for HPV16 DNA-positive/seropositive vaccinees. Seropositivity may serve as a proxy for persistent infection, since seroconversion normally occurs at least 6 months after first detection of HPV DNA [23]. Although the number of prevalent infection in this study was too low to reach a firm conclusion, all of the phase 3 trials have enrolled considerable numbers of women with prevalent infections, and thus, a clear answer to this question should be forthcoming. A third question is whether the VLP vaccines induce crossprotection against other genital HPV types. Animal challenge studies did not detect protection against distantly related types, such as BPV1 and cottontail rabbit papillomavirus [12]. Similarly, HPV6, HPV16 and HPV18 VLPs did not induce antibodies that could crossneutralize the other two types in in vitro neutralization assays [24]. However, these results do not exclude the possibility that VLP vaccines could induce some crossprotection against closely related genital HPV types. Interestingly, this is one of the few cases in which the results from the clinical trials are notably discordant. For the Merck HPV16 vaccine, there
Prophylactic Vaccines
221
was no significant reduction in the number of CINs associated with types other than HPV16 in HPV16 VLP vaccinees compared with controls after 1.5 or 3.5 years follow-up [18, 19], suggesting no crossprotection. In contrast, GSK reported at a recent meeting that their bivalent vaccine induced approximately 50% protection against persistent infection and cytological abnormalities by HPV16/18-related high-risk types during a 1-year follow-up. It is unclear at present whether the differences reflect differences in study design, endpoints evaluated or the ability of the Merck and GSK vaccines to induce crossneutralizing antibodies. Analyses of crossprotection in the large data sets of the phase 3 studies should clarify this apparent discrepancy. It will be important to evaluate the durability of crosstype versus type-specific protection. When detected, crossneutralizing titers are often 100-fold or more lower than type-specific neutralizing titers in in vitro assays [our unpubl. results], raising the possibility that crossprotection will wane more rapidly than type-specific protection. If this were the case, it might raise the question of whether it would be worthwhile to boost periodically to maintain crossprotection against the less common high-risk types, even if boosting were not required for type-specific protection. A fourth outstanding question is whether the vaccine will protect men from genital HPV infections. Protection of women from cervical infection does not necessarily predict protection in men. This is because the external genitalia in men are not bathed in mucus containing transudated serum immunoglobulin G, as is the case for the cervicovaginal tract [25]. Of potential relevance, an HSV glycoprotein-D-based vaccine proved to offer significant protection from genital HSV infection in seronegative women but not in men [26]. However, the excellent protection which the Merck vaccine appears to produce against genital warts in women may be more relevant to predictions of protection against both high-risk and low-risk virus infection in men, since these results indicate that HPV VLP vaccine-induced protection extends to cornified genital epithelia. Trials to determine the safety, immunogenicity and efficacy of prophylactic VLP vaccines in men have been initiated.
Effectiveness Considerations: Who to Vaccinate?
Merck applied for regulatory approval with the US Food and Drug Administration in December 2005. It is therefore possible that an HPV VLP vaccine will be commercially available next summer. Consequently, it becomes critical to consider to which groups of potential vaccinees resources, particularly those of the public sector, should be directed. The ultimate preferred target
Schiller/Lowy
222
population for this vaccine would seem to be pre- or early adolescent girls, before they become sexually active. This is because HPV infection is very common and often acquired soon after initiating sexual activity [27]. It is noteworthy that all of the efficacy trials have been in young women aged 15 years and above. However, bridging studies are underway in 10- to 14-year olds to determine if the vaccines have similar immunogenicity in the younger age group. Safety and immunogenicity data in this age group will likely be required by regulatory agencies before the vaccine will be approved for preadolescent girls. In settings with sufficient resources, it might also be worthwhile to have a catch-up phase, in which sexually active women are also immunized. Vaccination of ‘older’ women could have several benefits. It could protect a woman from the vaccine types to which she has not yet been exposed (and perhaps provide some crossprotection against nonvaccine types), as seen in the efficacy trials described above. Vaccination might also reduce the risk of persistence and progression in women with prevalent infections, as discussed above. Since virions shed from a productive infection of the cervicovaginal tract would presumably contact vaccine-induced neutralizing antibodies in the mucus, it might also inhibit transmission to a new partner and thereby contribute to herd immunity. However, it is important to note that there is as yet no clear evidence to support the latter two conjectures, and thus, it is premature to base pubic health policies on these possibilities. Nevertheless, it seems likely that most vaccine doses will initially be sold to sexually active women, because it will be the largest group of potential consumers and it is the group that is likely to have the highest intrinsic demand for the vaccine. It is difficult to advocate vaccination of boys or men at present, given the unknown effectiveness of the vaccines in males. If inclusion of HPV6 and HPV11 VLPs does induce protection against a substantial percentage of genital warts in men, it would likely prove to be an incentive for men to use the Merck vaccine. Whether male vaccination would contribute to herd immunity is uncertain. In some models of sexually transmitted infections, a relatively small increase in herd immunity is achieved by vaccinating both sexes, if a high proportion of one sex is vaccinated and the vaccine is effective at preventing transmission [28]. In contrast to the opportunity for infectious virus shed from the cervicovaginal tract to contact neutralizing antibodies in the mucus of vaccinated women, virus shed from the external genitalia in men would not be expected to normally contact neutralizing antibodies. Therefore, vaccination of women might have more impact on herd immunity, in addition to having more impact on the overall rates of HPV-associated cancers, of which 80–90% occur in women [29]. Therefore, it would be difficult to advocate male vaccination, even if it were effective, in situations where there were insufficient resources to vaccinate all young women.
Prophylactic Vaccines
223
Expectations for Vaccine Effectiveness
The effectiveness of the vaccines in general use will depend on many variables that are difficult to assess at present, including coverage rates, duration of protection, extent of herd immunity and degree of crossprotection. It is clear from our knowledge of the temporal relationship between initial HPV infection and detection of cervical cancer that these prophylactic vaccines will not have an impact on cervical cancer incidences for more than a decade. However, they would then have the potential to reduce cervical cancer rates in vaccinees by at least 70%. Since the majority of high-grade cervical dysplasias (CIN3) is also caused by HPV16 and HPV18, a substantial reduction in the number of treatments for premaligant cervical disease would be an expected earlier benefit in countries with established screening programs. In contrast, only a moderate reduction in the number of Pap smear abnormalities would be expected, if longterm protection is largely type specific, since only a minority of these abnormalities is caused by HPV16 and HPV18 (or HPV6, HPV11, HPV16 and HPV18) [30]. Therefore, there is a need to educate women and health care providers to have realistic expectations regarding the potential short-term benefits of the vaccine. If this information is not adequately conveyed, vaccinated women may be underscreened. It is also possible that continuing Pap abnormalities seen in vaccinated women could discourage these women, and their health care providers, from promoting vaccination of the next generation of young women, even though it might be very effective at reducing their cervical cancer risk.
References 1 2 3 4
5
6
7
8
Ehreth J: The value of vaccination: a global perspective. Vaccine 2003;21:4105–4117. Zinkernagel RM: On natural and artificial vaccinations. Annu Rev Immunol 2003;21:515–546. Doorbar J: The papillomavirus life cycle. J Clin Virol 2005;32(suppl 1):S7–S15. Pilacinski WP, Glassman DL, Glassman KF, Reed DE, Lum MA, Marshall RF, Muscoplat CC, Faras AJ: Immunization against bovine papillomavirus infection; in Ciba Foundation Symposium: Papillomaviruses – Symposium 120. Chichester, Wiley, 1986, pp 136–156. Kirnbauer R, Booy F, Cheng N, Lowy DR, Schiller JT: Papillomavirus L1 major capsid protein self-assembles into virus-like particles that are highly immunogenic. Proc Natl Acad Sci USA 1992;89:12180–12184. Kajigaya S, Fujii H, Field A, Anderson S, Rosenfeld S, Anderson LJ, Shimada T, Young NS: Selfassembled B19 parvovirus capsids, produced in a baculovirus system, are antigenically and immunogenically similar to nature virions. Proc Natl Acad Sci USA 1991;88:4646–4650. Zhou J, Sun XY, Stenzel DJ, Frazer IH: Expression of vaccinia recombinant HPV 16 L1 and L2 ORF proteins in epithelial cells is sufficient for assembly of HPV virion-like particles. Virology 1991;185:251–257. Kirnbauer R, Taub J, Greenstone H, Roden RBS, Durst M, Gissmann L, Lowy DR, Schiller JT: Efficient self-assembly of human papillomavirus type 16 L1 and L1-L2 into virus-like particles. J Virol 1993;67:6929–6936.
Schiller/Lowy
224
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Roden RBS, Greenstone HL, Kirnbauer R, Booy FP, Jessie J, Lowy DR, Schiller JT: In vitro generation and type-specific neutralization of a human papillomavirus type 16 virion pseudotype. J Virol 1996;70:5875–5883. Rose RC, Bonnez W, Reichman RC, Garcea RL: Expression of human papillomavirus type 11 L1 protein in insect cells: in vivo and in vitro assembly of viruslike particles. J Virol 1993;67:1936–1944. Rose RC, Reichman RC, Bonnez W: Human papillomavirus (HPV) type 11 recombinant viruslike particles induce the formation of neutralizing antibodies and detect HPV-specific antibodies in human sera. J Gen Virol 1994;75:2075–2079. Breitburd F, Kirnbauer R, Hubbert NL, Nonnenmacher B, Trin-Dinh-Desmarquet C, Orth G, Schiller JT, Lowy DR: Immunization with virus-like particles from cottontail rabbit papillomavirus (CRPV) can protect against experimental CRPV infection. J Virol 1995;69:3959–3963. Suzich JA, Ghim S, Palmer-Hill FJ, White WI, Tamura JK, Bell J, Newsome JA, Jenson AB, Schlegel R: Systemic immunization with papillomavirus L1 protein completely prevents the development of viral mucosal papillomas. Proc Natl Acad Sci USA 1995;92:11553–11557. Kirnbauer R, Chandrachud L, O’Neil B, Wagner E, Grindlay G, Armstrong A, McGarvie G, Schiller J, Lowy D, Campo M: Virus-like particles of bovine papillomavirus type 4 in prophylactic and therapeutic immunization. Virology 1996;219:37–44. Harper DM, Franco EL, Wheeler C, Ferris DG, Jenkins D, Schuind A, Zahaf T, Innis B, Naud P, De Carvalho NS, Roteli-Martins CM, Teixeira J, Blatter MM, Korn AP, Quint W, Dubin G: Efficacy of a bivalent L1 virus-like particle vaccine in prevention of infection with human papillomavirus types 16 and 18 in young women: a randomised controlled trial. Lancet 2004;364:1757–1765. Villa LL, Costa RL, Petta CA, Andrade RP, Ault KA, Giuliano AR, Wheeler CM, Koutsky LA, Malm C, Lehtinen M, Skjeldestad FE, Olsson SE, Steinwall M, Brown DR, Kurman RJ, Ronnett BM, Stoler MH, Ferenczy A, Harper DM, Tamms GM, Yu J, Lupinacci L, Railkar R, Taddeo FJ, Jansen KU, Esser MT, Sings HL, Saah AJ, Barr E: Prophylactic quadrivalent human papillomavirus (types 6, 11, 16, and 18) L1 virus-like particle vaccine in young women: a randomised double-blind placebo-controlled multicentre phase II efficacy trial. Lancet Oncol 2005;6: 271–278. Harro CD, Pang YY, Roden RB, Hildesheim A, Wang Z, Reynolds MJ, Mast TC, Robinson R, Murphy BR, Karron RA, Dillner J, Schiller JT, Lowy DR: Safety and immunogenicity trial in adult volunteers of a human papillomavirus 16 L1 virus-like particle vaccine. J Natl Cancer Inst 2001;93:284–292. Koutsky LA, Ault KA, Wheeler CM, Brown DR, Barr E, Alvarez FB, Chiacchierini LM, Jansen KU: A controlled trial of a human papillomavirus type 16 vaccine. N Engl J Med 2002;347: 1645–1651. Mao C, Koutsky LA, Ault KA, Wheeler CM, Brown DR, Wiley DJ, Alvarez FB, Bautista OM, Jansen KU, Barr E: Efficacy of human papillomavirus-16 vaccine to prevent cervical intraepithelial neoplasia: a randomized controlled trial. Obstet Gynecol 2006;107:18–27. Nardelli-Haefliger D, Wirthner D, Schiller JT, Lowy DR, Hildesheim A, Ponci F, De Grandi P: Specific antibody levels at the cervix during the menstrual cycle of women vaccinated with human papillomavirus 16 virus-like particles. J Natl Cancer Inst 2003;95:1128–1137. Winer RL, Lee SK, Hughes JP, Adam DE, Kiviat NB, Koutsky LA: Genital human papillomavirus infection: incidence and risk factors in a cohort of female university students. Am J Epidemiol 2003;157:218–226. Pinto LA, Edwards J, Castle PE, Harro CD, Lowy DR, Schiller JT, Wallace D, Kopp W, Adelsberger JW, Baseler MW, Berzofsky JA, Hildesheim A: Cellular immune responses to human papillomavirus (HPV)-16 L1 in healthy volunteers immunized with recombinant HPV-16 L1 virus-like particles. J Infect Dis 2003;188:327–338. Carter JJ, Koutsky LA, Hughes JP, Lee SK, Kuypers J, Kiviat N, Galloway DA: Comparison of human papillomavirus types 16, 18, and 6 capsid antibody responses following incident infection. J Infect Dis 2000;181:1911–1919. Roden RB, Yutzy WI, Fallon R, Inglis S, Lowy DR, Schiller JT: Minor capsid protein of human genital papillomaviruses contains subdominant, cross-neutralizing epitopes. Virology 2000;270: 254–257.
Prophylactic Vaccines
225
25 26
27 28 29 30
Mestecky J, Russell MW: Induction of mucosal immune responses in the human genital tract. FEMS Immunol Med Microbiol 2000;27:351–355. Stanberry LR, Spruance SL, Cunningham AL, Bernstein DI, Mindel A, Sacks S, Tyring S, Aoki FY, Slaoui M, Denis M, Vandepapeliere P, Dubin G, GlaxoSmithKline Herpes Vaccine Efficacy Study Group: Glycoprotein-D-adjuvant vaccine to prevent genital herpes. N Engl J Med 2002;347: 1652–1661. Baseman JG, Koutsky LA: The epidemiology of human papillomavirus infections. J Clin Virol 2005;32(suppl 1):S16–S24. Garnett GP: Role of herd immunity in determining the effect of vaccines against sexually transmitted disease. J Infect Dis 2005;191(suppl 1):S97–S106. Parkin DM, Bray F, Ferlay J, Pisani P: Global cancer statistics, 2002. CA Cancer J Clin 2005;55: 74–108. Clifford GM, Rana RK, Franceschi S, Smith JS, Gough G, Pimenta JM: Human papillomavirus genotype distribution in low-grade cervical lesions: comparison by geographic region and with cervical cancer. Cancer Epidemiol Biomarkers Prev 2005;14:1157–1164.
Dr. John T. Schiller, PhD Laboratory of Cellular Oncology, National Institutes of Health Building 37, Room 4106, 9000 Rockville Pike Bethesda, MD 20892 (USA) Tel. ⫹1 301 594 2715, Fax ⫹1 301 480 5322, E-Mail
[email protected] Schiller/Lowy
226
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 227–234
Assessment and Follow-Up of HPV Vaccines Eliav Barr Merck Research Laboratories, West Point, Pa., USA
The HPV family consists of more than 90 related epitheliotropic DNA viruses [1]. The primary target of HPV infection is the basal cell of the epithelium. Certain HPV proteins, such as HPV E6 and E7, have been shown to disrupt the cell cycle regulatory apparatus of infected cells [2, 3]. Thus, all HPV types can cause aberrant proliferation of infected cells [3]. HPVs can be divided into two categories based on their association with cancer in humans. High-risk HPV types (e.g., HPV16 and HPV18) are associated with the development of anogenital cancers. Low-risk HPV types (e.g., HPV6 and HPV11) cause benign proliferative lesions [3]. Infection with HPV is associated with the development of cervical/vulvar/vaginal or anal intraepithelial neoplasia, cervical/ vaginal/vulvar or anal cancer, genital warts and recurrent respiratory papillomatosis [4, 5]. Since cervical cancer, cervical dysplasia and genital warts are essentially sequelae of HPV infection, a prophylactic vaccine that prevents infection with the most common pathogenic HPV types is the most efficient means to reduce the incidence of these diseases. The optimal HPV vaccine should substantially reduce the morbidity and mortality associated with HPV disease. Thus, the optimal HPV vaccine should reduce (1) a woman’s risk of development of cervical cancer, by reducing the incidence of cervical cancer and moderate- to high-grade cervical dysplasia, i.e. cervical intraepithelial neoplasia (CIN) grades 2 and 3, caused by common HPV types, and (2) the public health and economic burden of HPV infection, by reducing the incidence of the most common clinical manifestations of HPV infection (genital warts and low-grade cervical dysplasia, or CIN1) caused by common HPV types.
Within 1 year
Initial HPV infection
Up to 5 years
Continuing infection
Up to decades
CIN 2/3
Cervical cancer
CIN 1
Cleared HPV infection
Fig. 1. Natural history of infection with high-risk HPV types (such as HPV16 and HPV18).
Endpoint Assessment for HPV Vaccines
Phase III clinical studies to demonstrate that a prophylactic HPV vaccine reduces the incidence of cervical cancer are neither ethically nor logistically feasible. Thus, the impact of prophylactic HPV vaccines on cervical cancer rates must be extrapolated from the impact of the vaccine on the histopathologic conditions that precede cervical cancer. The selection of one of these conditions as the critical efficacy endpoint for licensure of a prophylactic HPV vaccine requires careful consideration of the natural history of HPV infection and clinical trial logistics. Figure 1 provides schematic representation of the natural history of cervical HPV infection. Cervical cancer is caused by infection with high-risk oncogenic HPV types [6–11]. However, most HPV infections are subclinical and self-limited. Early HPV infection often manifests itself as low-grade dysplasia (CIN1) [7, 12, 13]. In most cases, these lesions resolve without clinical sequelae [7, 12, 13]. In a small fraction of women, HPV infection persists, leading to moderate/high-grade cervical dysplasia (CIN2/3) and cancer [12–14]. The incidence of HPV infection peaks in the early 20s, the incidence of moderate- and high-grade dysplastic lesions (CIN2 and CIN3, respectively) begins to increase in the mid 20s and peaks in the late 20s, and the incidence of cervical cancer peaks at the age of 35–55 [6, 7, 11, 15, 16]. The long interval between the acquisition of HPV infection and the development of cervical cancer has allowed for the institution of screening programs using the Papanicolaou (Pap) test. In industrialized countries with advanced health care infrastructure, such Pap screening reduces the incidence of cervical cancer by detecting and excising CIN2/3 lesions prior to the development of cancer [16–19].
Barr
228
Table 1. Lifetime risk of developing cervical cancer or its precursors HPV-related pathologic state
Lifetime risk
HPV infection CIN1 CIN2/3 Cervical cancer (under no Pap screening conditions)
1:2 (50%) 1:6 (17%) 1:25 (4%) 1:31 (3%)
Table 1 compares the lifetime risk of the development of HPV infection, CIN1, CIN2/3 and cervical cancer (with and without cervical cancer screening). These risk data were calculated from US natural history studies and cancer statistics [12, 14, 17–21]. Overall, 1 in 60 women who develop HPV infection will progress to cervical cancer, even with intensive Pap screening. Without detection and excision, CIN2/3 will progress to cervical cancer (compare the lifetime risk of CIN2/3 with the lifetime risk of cervical cancer in the absence of screening programs) [16]. In countries with organized cervical cancer screening programs, cancer rates have decreased by approximately 75% [17–19]. The licensure of a new vaccine should be predicated on a robust demonstration of the efficacy of the vaccine against the most important clinical disease caused by the pathogen targeted by the vaccine. In the case of high-risk oncogenic HPV types, that disease is cervical cancer. However, a phase III clinical trial to demonstrate the efficacy of a prophylactic HPV vaccine using a cervical cancer endpoint is not feasible because (1) the median time from acquisition of infection to the development of cervical cancer is ⬎20 years, and (2) the standard of care worldwide is to screen women for CIN2/3 and to excise these lesions prior to the development of cancer [7, 22, 23]. In the context of a clinical trial, women must receive the highest level of care. Thus, it is unethical to allow placebo recipients to remain unscreened and untreated solely for the purposes of a cancer efficacy evaluation. Since an efficacy trial of a prophylactic HPV vaccine using a cervical cancer endpoint is not feasible, efficacy studies must use an HPV-related pathologic state that precedes cervical cancer as the primary efficacy endpoint. Such an intermediate pathologic state must be sufficiently robust to ensure that a demonstration of efficacy with regard to this state will be highly likely to translate into a reduction in risk of cervical cancer. Candidate pathologic states must meet the following criteria: (1) the pathologic state must be a necessary step in the development of cervical cancer; (2) the pathologic state must be closely associated in pathophysiologic sequence to the development of cervical cancer; (3) the pathologic state must confer a high risk of development of cervical
HPV Vaccines
229
Table 2. Comparison of candidate intermediate endpoints for efficacy trials of vaccines targeting high risk HPV types with traditional criteria for surrogate markers Criterion
Incident HPV infection
HPV-related CIN1
HPV-related CIN2/3
Obligate precursor of cervical cancer Closely associated in temporal sequence to the development of cervical cancer Risk of development of cervical cancer Reductions in incidence or treatment shown to result in a reduction in the risk of cervical cancer
Yes No
No No
Yes Yes
Low No
Low/moderate No
High Yes
cancer; (4) the pathologic state should represent a clinically important syndrome requiring therapy, and (5) a reduction in the incidence or treatment of the pathologic state must have been already shown to result in a reduction in the risk of cervical cancer. HPV infection, CIN1 and CIN2/3 are pathologic states that may serve as candidate endpoints for demonstrations of the efficacy of a prophylactic HPV vaccine against cervical cancer. The strengths and weaknesses of each of these states as the key proof of efficacy for public health implementation of prophylactic HPV vaccines are listed in table 2. From a practical perspective, CIN2/3 is the most advanced preneoplastic lesion that can be studied in phase III prophylactic HPV vaccine studies. The standard of care for the management of CIN2 and CIN3 lesions in Europe is wide excision. The interval between initial HPV infection and development of CIN2/3 (between 6 months and 5 years) is manageable within a clinical trial. CIN3 represents the final stage prior to invasive cancer. CIN2 may represent a heterogeneous group of lesions, including those that progress to cancer if untreated, or those that may regress. However, by limiting pathology readings for endpoint purposes to a validated panel of expert pathologists and by requiring detection of high-risk HPV DNA within a lesion as part of the endpoint definition, the predictive value of a CIN2 endpoint for cervical cancer risk is substantially enhanced. The Center for Biologics Evaluation and Research (CBER) of the Food and Drug Administration has provided guidance on the efficacy demonstrations required for licensure of a prophylactic HPV vaccine in the US. CBER convened a special meeting of its Vaccine and Related Biological Products Advisory Committee (VRBPAC) in 2001 in response to sponsor requests for guidance regarding phase III study designs and endpoints that would support licensure of a prophylactic HPV vaccine in the US. The VRBPAC considered
Barr
230
the following candidate efficacy endpoints: HPV infection, persistent infection, CIN1, CIN2/3 and cervical cancer. VRBPAC recommended that a definitive demonstration of vaccine efficacy with regards to vaccine HPV-type-related CIN2/3 should be required for licensure of an HPV vaccine. CBER concurred with the committee’s recommendation [24].
Immunology of HPV Infection – Rationale for L1 VLP Vaccines
Studies using live HPV11 and HPV16 virions propagated in infected primary human keratinocytes grown as xenografts in immunodeficient mice have demonstrated that the infectivity of HPV is dependent on the L1 capsid protein. Since there are no animal models of HPV disease, vaccine development has been guided by use of preclinical models of nonhuman papillomavirus infection and disease. Recombinantly expressed L1 proteins self-assemble into virus-like particles (VLPs) [25]. Vaccination with L1 VLPs derived from species-specific papillomaviruses protects against acquisition of infection and disease in the appropriate animal models [25, 26]. This protection is due to the induction of type-specific, conformation-dependent antibody responses that are virus neutralizing [25, 26]. Protection against disease has been shown in all papillomavirus disease models tested [25–27]. These results support the hypothesis that generation of robust systemic anti-HPV responses by vaccination with type-specific HPV L1 VLPs will result in protective immunity against typespecific HPV infection and disease.
Prophylactic HPV L1 VLP Vaccine: Results to Date
Initial studies of monovalent HPV11, HPV16 or HPV18 L1 VLP vaccine showed them to be well tolerated and immunogenic [28–31]. A randomized, placebo-controlled study of a prototype HPV16 L1 VLP vaccine included 2,391 women, 16–23 years of age [32]. The primary efficacy endpoints were: (1) persistent HPV16 infection, defined by positive HPV16 DNA results on two or more consecutive visits at least 4 months apart; (2) HPV16-related CIN, defined as a consensus pathology panel diagnosis of CIN1, CIN2, CIN3, adenocarcinoma in situ or cervical cancer, plus detection of HPV16 DNA in biopsy tissue obtained from the same lesion and detection of HPV16 on the routine visit immediately prior to or following the colposcopy visit in which the biopsy showing CIN was obtained, or (3) detection of HPV16 DNA in cervicovaginal specimens obtained on the last visit on record, without confirmed persistent infection.
HPV Vaccines
231
After a median follow-up of 40 months after completion of the vaccination regimen, the overall vaccine efficacy in the per-protocol population (subjects received three doses, had no major protocol violations, were vaccine-type HPV seronegative at day 1 and vaccine-type HPV DNA negative at day 1 through month 7) was 100% with regard to confirmed persistent HPV16 infections, defined as two consecutive HPV16 DNA-positive tests collected 4 or more months apart. Vaccine efficacy was 100% for preventing HPV16-related lowand high-grade cervical lesions (CIN1–3). In the per-protocol efficacy cohort, administration of the HPV16 L1 VLP vaccine resulted in a 52% reduction in the overall incidence of CIN2/3 and a 30% reduction in the overall incidence of CIN1 due to any HPV type, though these reductions did not reach statistical significance. These results were the first to confirm that an HPV16 L1 VLP vaccine was not only effective for preventing persistent HPV16 infection, but also for preventing HPV16-related CIN2/3. The prototype monovalent vaccine study was followed by a phase II study which evaluated the efficacy of a quadrivalent HPV (types 6, 11, 16 and 18) L1 VLP vaccine (Gardasil®) against vaccine-related infection and disease [33]. This randomized, placebo-controlled study was conducted in 552 women aged 16–23. The infection component of the primary efficacy endpoint was defined as one of the following: (1) persistent HPV6, HPV11, HPV16 or HPV18 infection, defined as detection of HPV6, HPV11, HPV16 or HPV18 DNA in cervicovaginal or biopsy swabs collected on at least two consecutive visits spaced at least 4 months apart; (2) detection of HPV6, HPV11, HPV16 or HPV18 DNA in a biopsy specimen from the same lesion in which CIN or cervicovaginal disease was detected by the pathology panel of the program, together with detection of the same vaccine HPV-type DNA immediately prior to or after the disease diagnosis, or (3) detection of HPV6, HPV11, HPV16 or HPV18 DNA on the last visit on record. The disease component of the primary efficacy endpoint was defined as a pathology panel diagnosis of condylomata acuminata, vulvar intraepithelial neoplasia, vaginal intraepithelial neoplasia, CIN, adenocarcinoma in situ, or cervical, vulvar or vaginal cancer, plus detection of HPV6, HPV11, HPV16 or HPV18 DNA in the same tissue. Compared with placebo recipients, the combined incidence of HPV6-, HPV11-, HPV16- or HPV18-related persistent infection or disease was reduced in vaccinees by 90% (95% CI 71–97; p ⬍ 0.001). Over the 2.5-year postvaccination period, 6 placebo subjects developed HPV6-, HPV11-, HPV16- or HPV18-related genital diseases. There were no cases in the quadrivalent HPV vaccine group. Thus, the point estimate of vaccine efficacy was 100%, with a 95% lower confidence bound of 15.9%. The quadrivalent vaccine was highly immunogenic. All vaccine recipients in the per-protocol immunogenicity cohorts developed detectable anti-HPV6,
Barr
232
HPV11, HPV16 and HPV18 responses after completion of the vaccine regimen. Vaccine-induced anti-HPV responses were substantially higher than those observed in placebo recipients with a previous history of natural HPV infection. Administration of the quadrivalent HPV vaccine was generally well tolerated though a higher proportion of quadrivalent vaccine recipients reported one or more injection site adverse experiences, compared with placebo. The most common injection site and systemic adverse experiences were pain and headache, respectively.
References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
19 20
Alani RM, Münger K: Human papillomaviruses and associated malignancies. J Clin Oncol 1998;16:330–337. Zur Hausen H: Papillomavirus causing cancer: evasion from host-cell control in early events in carcinogenesis. J Natl Cancer Inst 2000;92:690–698. Southern SA, Herrington CS: Molecular events in uterine cervical cancer. Sex Transm Inf 1998;74:101–109. Bernard H, Bosch X, et al: Studies of cancer in humans; in IARC: Human Papillomaviruses. Lyon, International Agency for Research on Cancer Monograph, 1995, vol 64, pp 1–87. Derkay CS, Darrow DH: Recurrent respiratory papillomatosis of the larynx: current diagnosis and treatment. Otolaryngol Clin North Am 2000;33:1127–1141. Bosch FX, de Sanjose S: Human papillomavirus and cervical cancer – burden and assessment of causality. J Natl Cancer Inst Monogr 2003;31:3–13. Schiffman M, Kruger-Kjaer S: Natural history of anogenital human papillomavirus infection and neoplasia. J Natl Cancer Inst Monogr 2003;31:14–19. Bosch FX, Munoz N: The viral etiology of cervical cancer. Virus Res 2002;89:183–190. Munoz N, Bosch FX, de Sanjose S, et al: Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518–527. Clifford GM, Smith JS, Plummer M, et al: Human papillomavirus types in invasive cervical cancer worldwide: a meta-analysis. Br J Cancer 2003;88:63–73. Bosch FX, Lorincz A, Munoz N, et al: The causal relation between human papillomavirus and cervical cancer. J Clin Pathol 2002;55:244–265. Ho GYF, Bierman R, Beardsley L, Chang CJ, Burk RD: Natural history of cervicovaginal papillomavirus infection in young women. N Engl J Med 1998;338:423–428. Woodman CB, Collins S, Winter H, et al: Natural history of cervical human papillomavirus infection in young women: a longitudinal cohort study. Lancet 2001;357:1831–1836. Koutsky LA, Holmes KK, Critchlow CW, et al: A cohort study of the risk of cervical intraepithelial neoplasia grade 2 or 3 in relation to papillomavirus infection. N Engl J Med 1992;327:1272–1278. Melbye M, Frisch M: The role of human papillomaviruses in anogenital cancers. Semin Cancer Biol 1998;8:307–313. Andersson-Ellstrom A, Seidal T, Grannas M, et al: The Pap smear history of women with invasive cervical squamous carcinoma. Acta Obstet Gynecol Scand 2000;79:221–226. Ponten J, Adami HO, Bergstrom R, et al: Strategies for global control of cervical cancer. Int J Cancer 1995;60:1–26. Miller AB, Nazeer S, Fonn S, Brandup-Lukanow A, Rehman R, Cronje H, Sankaranarayanan R, Koroltchouk V, Syrjänen K, Singer A, Onsrud M: Report on consensus conference on cervical cancer screening and management. Int J Cancer 2000;86:440–447. Dillner J: Cervical cancer screening in Sweden. Eur J Cancer 2000;36:2255–2259. Winer RL, Lee SK, Hughes JP, et al: Genital human papillomavirus infection: incidence and risk factors in a cohort of female university students. Am J Epidemiol 2003;157:218–226.
HPV Vaccines
233
21 22 23 24
25
26
27
28
29
30 31
32
33
American Cancer Society: Cancer facts and figures 2003. http://www.cancer.org Lowy DR, Frazer IH: Prophylactic human papillomavirus vaccines. J Natl Cancer Inst Monogr 2003;31:111–116. Wright TC, Cox JT, Massad S, et al: 2001 consensus guidelines for the management of women with cervical intraepithelial neoplasia. Am J Obstet Gynecol 2003;189:295–304. Centers for Biologics Evaluations and Research of the Food and Drug Administration: Minutes of the November, 2001 Vaccine and Related Biologic Products Advisory Committee (VRBPAC). www.fda.gov Jansen KU, Rosolowsky M, Schultz LD, Markus HZ, Cook JC, Donnelly JJ, et al: Vaccination with yeast-expressed cottontail rabbit papillomavirus (CRPV) virus-like particles protects rabbits from CRPV-induced papilloma formation. Vaccine 1995;13:1509–1514. Lowe RS, Brown DR, Bryan JT, Cook JC, George HA, Hofmann KJ, et al: Human papillomavirus type 11 (HPV-11) neutralizing antibodies in the serum and genital mucosal secretions of African green monkeys immunized with HPV-11 virus-like particles expressed in yeast. J Infect Dis 1997;176:1141–1145. Suzich JA, Ghim SJ, Palmer-Hill FJ, White WI, Tamura JK, Bell JA, et al: Systemic immunization with papillomavirus L1 protein completely prevents the development of viral mucosal papillomas. Proc Natl Acad Sci USA 1995;92:11553–11557. Brown DR, Fife KH, Wheeler CM, Koutsky LA, Lupinacci LM, Railkar R, et al: Early assessment of the efficacy of a human papillomavirus type 16 L1 virus-like particle vaccine. Vaccine 2004;22:2936–2942. Fife KH, Wheeler CM, Koutsky LA, Barr E, Brown DR, Schiff MA, et al: Dose-ranging studies of the safety and immunogenicity of human papillomavirus type 11 and type 16 virus-like particle candidate vaccines in young healthy women. Vaccine 2004;22:2943–2952. Ault K, Giuliano AR, Edwards R, Tamms G, Kim LL, Smith JF, et al: A phase I study to evaluate a human papillomavirus (HPV) type 18 L1 VLP vaccine. Vaccine 2004;22:3004–3007. Brown DR, Bryan JT, Schroeder JM, Robinson TS, Fife KH, Wheeler CM, et al: Neutralization of human papillomarvirus type 11 (HPV-11) by serum from women vaccinated with yeast-derived HPV-11 L1 virus-like particles: correlation with competitive radioimmunoassay titer. J Infect Dis 2001;184:1183–1186. Mao C, Koutsky LA, Ault KA, Wheeler CM, Brown DR, Wiley DJ, et al: Effectiveness of human papillomavirus 16 virus-like particle vaccine in preventing human papillomavirus 16-related cervical intraepithelial neoplasia 2–3: a randomized controlled trial. Obstet Gynecol 2006;107:18–27. Villa LL, Costa RLR, Petta CA, Andrade RP, Ault KA, Giuliano AR, et al: Prophylactic quadrivalent human papillomavirus (types 6, 11, 16, and 18) L1 virus-like particle vaccine in young women: a randomised double-blind placebo-controlled multicentre phase II efficacy trial. Lancet Oncol 2005;6:271–278.
Dr. Eliav Barr, MD Merck Research Laboratories, UNC-141 785 Jolly Road Blue Bell, PA 19422 (USA) Tel. ⫹1 484 344 7822, Fax ⫹1 484 344 3357, E-Mail
[email protected] Barr
234
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 235–241
Cost-Effectiveness of HPV Vaccines Evan R. Myers Department of Obstetrics and Gynecology, Duke University Medical Center, Durham, N.C., USA
Estimating the combined clinical and economic impact of HPV vaccines is challenging, since so much of the data needed for generating these estimates will not be available until years after vaccines are introduced into practice. However, it is possible to identify the key factors which will have the largest impact on health outcomes and costs and use this information to focus on data collection after vaccine use becomes common. This chapter will briefly outline (1) the basic principles of economic analysis in health care, (2) the use of models in evaluating cervical cancer prevention strategies, (3) lessons learned about the cost-effectiveness of screening, (4) initial results of modeling studies of HPV vaccines, and (5) key issues for the future.
Definitions of Terms Used in Economic Analyses
Economic analyses of health interventions are necessary because the resources available for improving health are limited and most useful when there are two or more possible strategies for addressing a given health problem. If one compares the relative effectiveness and associated costs of two potential strategies, there are four possible outcomes: Strategy A may be more expensive and less effective than Strategy B, Strategy A may be more expensive and more effective than Strategy B, Strategy A may be less expensive and less effective, or Strategy A may be less expensive and more effective. If a new strategy costs less and results in better outcomes than an existing one, or costs more and results in worse outcomes, it is not worth considering. If the new strategy costs less and results in worse outcomes, or costs more and results in better outcomes, then
the question is whether the improvement in outcomes associated with the more expensive strategy is ‘worth’ the extra expense. The extra expense per health benefit gained is called the incremental cost-effectiveness ratio (ICER) and is calculated by dividing the difference in costs associated with two strategies by the difference in a measure of health outcome: Costs of Strategy A ⫺ Costs of Strategy B __________________________________________ Outcomes of Strategy A ⫺ Outcomes of Strategy B
Costs included in the analysis typically include all costs resulting from the disease in question, as well as costs related to interventions designed to prevent or detect the disease. For example, costs used in an economic analysis of a screening program for cervical cancer would include the costs of the screening test, additional diagnostic tests performed for abnormal screening test results, treatments based on those test results, complications of treatment, as well as the costs of diagnosis and treatment (including complications) of invasive cervical cancer. Ideally, indirect costs associated with the interventions, such as time lost from work, transportation to and from the site where care is received, and child care costs should be included, although these are often excluded because of the difficulty of obtaining accurate measurements. Outcomes counted in the analysis include all of the relevant outcomes for which costs are obtained and are usually summed using some aggregate measure. The most common aggregate outcome used in economic analyses is the quality-adjusted life year (QALY) or disability-adjusted life year, which incorporates both life expectancy and quality of life measures. For potentially fatal diseases, the most obvious health outcome of interest is life expectancy. The impact of other conditions which, while not fatal, may affect quality of life is incorporated by using some measure of quality of life to adjust life expectancy. For example, if the relative value of quality of life for a woman treated for invasive cervical cancer with radiation therapy is 25% less than the quality of life for a woman who never had cervical cancer, than each year of life following treatment would be calculated as 0.75 ⫻ 1 year, or 0.75 QALY. Because quality of life measures are not available for many important health conditions relevant to cervical cancer screening (such as the anxiety induced by abnormal screening tests), life expectancy alone (expressed as years of life saved) is often used for conditions which can result in early mortality, such as cervical cancer. An intervention is considered ‘cost-effective’ compared with the next least expensive alternative if the ICER is less than some threshold value; e.g., in the US, an intervention with an ICER less than USD 50,000–75,000 per QALY has generally been considered ‘cost-effective’.
Myers
236
Modeling the Impact of Cervical Cancer Prevention Strategies
Because of practical and ethical issues, evaluating the relative effectiveness and cost-effectiveness of different strategies for preventing cervical cancer can never be done using traditional methods of study design. For example, a randomized trial comparing the effectiveness of screening every 3 years to every 1 year in preventing cervical cancer would take many years to perform, and, because of ethical reasons, including an arm where no women were screened to establish baseline risks could never be ethically done. Because of this, researchers and policy makers are dependent on the use of computer simulation models [1–9]. The majority of currently used models are similar in structure. The model simulates a cohort of women. During each period of the simulation, women may acquire HPV; the probability of this event is the age-specific incidence of HPV. Depending on the type and possibly age, the HPV infection may regress, stay the same or progress to a more advanced cervical intraepithelial neoplasia (CIN). Each subsequent CIN stage may regress, stay the same or progress within a given time period. Once CIN progresses to invasive cancer, the cancer progresses until detected by symptoms. At each step of the simulation, competing risks such as death from other causes or hysterectomy with removal of the cervix are also included. At the completion of the simulation, life expectancy (or QALYs) for the cohort are summed. Prevention strategies are modeled by incorporating them into the basic natural history model. For example, HPV vaccination is modeled by reducing the incidence of HPV acquisition by an appropriate amount. Screening is modeled by incorporating test sensitivity and specificity for a given underlying normal, CIN or cancer histology. The values used for each of the probabilities in the model are derived from the literature. In some cases (HPV incidence in young women, HPV vaccine efficacy), there are excellent data from prospective cohorts or randomized trials. In other cases (sensitivity and specificity of screening tests, duration of vaccine efficacy), the quality of the data may be somewhat less robust. In still other cases (progression rates between different stages of invasive cancer, the probability that a woman with a given stage of cancer will present with symptoms), data are impossible to obtain, and the modeler must impute the values based on the closeness of fit of the model output for cancer incidence or mortality and observed data. These uncertainties, as well as methodological issues such as the difficulty of correcting for cohort effects in available age-specific incidence and mortality data, require both modelers and users of models to view the quantitative results of modeling studies with appropriate caution. Although there are many ways to address the impact of uncertainty, there will always be a fairly high degree of imprecision in the quantitative estimates derived from models.
HPV Vaccines
237
Often, the most interesting and valuable results of modeling studies are the effects of varying parameter values on the resulting estimates of costs and outcomes; by illustrating which parameters are most important, priorities for future research can be identified. Fortunately, the natural history of cervical cancer is relatively consistent across both time and geography [10]. This means that the qualitative results of these models should be fairly robust, and, indeed, the overwhelming majority of modeling studies have come to very similar conclusions regarding the costeffectiveness of screening interventions; these conclusions have subsequent implications for the potential cost-effectiveness of HPV vaccines.
Findings of Cost-Effectiveness Analyses of Screening
Although quantitative results may differ between studies, the qualitative results of most cost-effectiveness analyses of cervical cancer screening programs are quite similar and can be easily understood in terms of the natural history of HPV and cervical cancer. • If a limited number of screens are to be done over a woman’s lifetime, the most efficient time to screen is approximately age 35. This is because (1) cervical cancer incidence reaches its peak approximately 10 years later, (2) symptomatic cervical cancer is relatively rare in women in their 20s, and (3) transient HPV infections and CIN are very common in women in their 20s – screening in these women results in many abnormal tests for conditions which would never go on to become cancer. • For any given level of test sensitivity and specificity, costs go up faster than effectiveness as screening intervals become more frequent, making ICERs very high. In most analyses in developed countries, screening intervals of 3–5 years usually fall within acceptable limits of cost-effectiveness. Biennial and annual screening have very high cost-effectiveness ratios and are generally considered as not cost-effective from a policy perspective. The high costs of frequent screening are due not only to the increased number of tests, but also because more frequent screening detects more lesions which, in the absence of screening, would go away on their own without intervention. • For any given test, ICERs increase exponentially as specificity decreases (since false-positive test results lead to increased costs with no gain in life expectancy). ICERs also increase, although less dramatically, as test sensitivity increases (since more sensitive tests detect both lesions that would become cancer and lesions that would not become cancer).
Myers
238
•
In a well-screened cohort, screening beyond age 65 is usually not costeffective, because cervical cancer incidence declines due to several factors, including the natural history of the disease, increasing risk of death from other causes, and a reduction in prevalence resulting from screening and treatment at younger ages.
Potential Cost-Effectiveness of HPV Vaccines
To date, only three cost-effectiveness analyses of the impact of HPV vaccines have been published [11–13]. All three conclude that HPV vaccination is potentially cost-effective, even in settings with existing screening programs. Besides the ultimate price of the vaccine, parameters that have the largest impact on cost-effectiveness include the following. • Age at vaccination – current assumptions are that vaccination is most effective if given prior to the onset of sexual activity and subsequent exposure to HPV. As the age of vaccination increases, the proportion of the population already exposed to HPV increases. • Duration of vaccine efficacy – a vaccine which does not provide lifetime protection and subsequently required a booster might still be cost-effective; extending vaccine efficacy beyond peak HPV incidence provides relatively little population benefit. • Proportion of the population vaccinated. • Screening policy – vaccination should result in changes in the epidemiology of HPV and CIN which lead to changes in optimal screening programs. First, vaccination should reduce the incidence of cervical cancer in younger women, further reducing the efficiency of screening in these women and allowing screening to begin at a later age. Second, since HPV vaccination should reduce the prevalence of CIN3 and cervical cancer, the positive predictive value of any screening test will decrease (since the positive predictive value decreases as the disease becomes more rare) and the negative value will increase (since the negative predictive value increases as the disease becomes more rare). This in turn means that screening intervals can be lengthened with minimal impact on effectiveness, but substantial reduction in costs. Another consistent finding is that the reduction in cervical cancer incidence is somewhat less than predicted by the proportion of oncogenic HPV types covered by a given vaccine. In other words, models which estimate the effect of a vaccine which is 100% effective against types which cause 70% of cancer do not result in 70% reductions in cancer incidence. This finding is likely the result of existing model structures and assumptions – by increasing
HPV Vaccines
239
the number of women who are not infected with a given HPV type, the pool of women who may be infected with other types is larger. Whether this replacement effect is real, or is an artifact of current lack of knowledge about the dynamics of multiple HPV types within both populations and individuals, will be addressed by both modeling and empirical research.
Issues for the Future
In addition to questions surrounding the optimal age for vaccination, the duration of vaccine efficacy and type replacement, there are a number of issues which are currently being incorporated into newer models, including the following. • Transmission – classic cohort models of HPV and cervical cancer do not incorporate the fact that HPV is an infectious disease and that vaccination, by reducing the pool of infected individuals, should result in reduction in incidence greater than that predicted by accounting for its protective effect within individuals alone. There is much ongoing work on modeling these effects, which have implications for policy decisions about whether to vaccinate women alone, or men and women. Although combining dynamic transmission models and cohort models of disease progression is complex, modeling transmission is limited even more by a lack of empiric data on transmission rates for HPV, which should be a high priority for future research. One point to bear in mind is that models which do not include transmission are likely to underestimate vaccine efficacy, and thus, underestimate cost-effectiveness; in other words, if a vaccine is cost-effective without considering transmission, it will also be cost-effective when transmission is accounted for in the model. • Impact of vaccine against HPV types 6 and 11 – one of the vaccines soon to be available on the market includes coverage against HPV6 and HPV11, which account for 90% of genital warts. Data on genital warts are less readily available than data on CIN and cervical cancer, and data on the quality of life impact are also relatively lacking. Clearly, to the extent that reduction in the incidence of HPV6 and HPV11 will reduce the incidence of genital warts, as well as less common conditions such as head and neck cancers and recurrent laryngeal papillomatosis, inclusion of HPV6 and HPV11 in vaccines will result in further reductions in expenses associated with HPVrelated disease. If vaccines including HPV6 and HPV11 are more expensive than vaccines which do not include these types, then economic analyses will have to determine whether these reductions in expenses (and gains in quality of life) offset the increased price of the vaccine.
Myers
240
References 1 2
3 4 5
6
7
8
9
10 11
12 13
Holmes J, Hemmett L, Garfield S: The cost-effectiveness of human papillomavirus screening for cervical cancer. A review of recent modelling studies. Eur J Health Econ 2005;6:30–37. Goldie SJ, Gaffikin L, Goldhaber-Fiebert JD, Gordillo-Tobar A, Levin C, Mahe C, Wright TC, Alliance for Cervical Cancer Prevention Cost Working Group: Cost-effectiveness of cervicalcancer screening in five developing countries. N Engl J Med 2005;353:2158–2168. Kim JJ, Wright TC, Goldie SJ: Cost-effectiveness of human papillomavirus DNA testing in the United Kingdom, the Netherlands, France, and Italy. J Natl Cancer Inst 2005;97:888–895. Fahs MC, Plichta SB, Mandelblatt JS: Cost-effective policies for cervical cancer screening: an international review. Pharmacoeconomics 1996;9:211–230. Dewilde S, Anderson R: The cost-effectiveness of screening programs using single and multiple birth cohort simulations: a comparison using a model of cervical cancer. Med Decis Making 2004;24:486–492. Mandelblatt JS, Lawrence WF, Gaffikin L, Limpahayom KK, Lumbiganon P, Warakamin S, King J, Yi B, Ringers P, Blumenthal PD: Costs and benefits of different strategies to screen for cervical cancer in less-developed countries. J Natl Cancer Inst 2002;94:1469–1483. Van den Akker-van Marle ME, van Ballegooijen M, van Oortmarssen GJ, Boer R, Habbema JD: Cost-effectiveness of cervical cancer screening: comparison of screening policies. J Natl Cancer Inst 2002;94:193–204. Goldie SJ, Kuhn L, Denny L, Pollack A, Wright TC: Policy analysis of cervical cancer screening strategies in low-resource settings: clinical benefits and cost-effectiveness. JAMA 2001;285: 3107–3115. Myers ER, McCrory DC, Subramanian S, McCall N, Nanda K, Datta S, Matchar DB: Setting the target for a better cervical screening test: characteristics of a cost-effective test for cervical neoplasia screening. Obstet Gynecol 2000;96:645–652. Gustafsson L, Ponten J, Bergstrom R, Adami HO: International incidence rates of invasive cervical cancer before cytological screening. Int J Cancer 1995;71:159–165. Goldie SJ, Kohli M, Grima D, Weinstein MC, Wright TC, Bosch FX, Franco E: Projected clinical benefits and cost-effectiveness of a human papillomavirus 16/18 vaccine. J Natl Cancer Inst 2004;96:604–615. Kulasingam SL, Myers ER: Potential health and economic impact of adding a human papillomavirus vaccine to screening programs. JAMA 2003;290:781–789. Sanders GD, Taira AV: Cost-effectiveness of a potential vaccine for human papillomavirus. Emerg Infect Dis 2003;9:37–48.
Dr. Evan R. Myers, MD, MPH Division of Clinical and Epidemiological Research Department of Obstetrics and Gynecology DUMC 3279, 244 Baker House, Duke University Medical Center Durham, NC 27710 (USA) Tel. ⫹1 919 668 0296, Fax ⫹1 919 668 0295, E-Mail
[email protected] HPV Vaccines
241
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 242–252
HPV Vaccination: Unresolved Issues and Future Expectations Ruanne V. Barnabasa,b, Katherine M. Frenchb, Päivi Laukkanenc,d, Osmo Kontulae, Matti Lehtinenc,f, Geoff P. Garnettb a
Cancer Epidemiology Unit, University of Oxford, Oxford, bDepartment of Infectious Disease Epidemiology, Imperial College, London, UK; cNational Public Health Institute, Helsinki and Oulu, dFinnish Cancer Registry, Institute for Statistical and Epidemiological Cancer Research, eFamily Federation of Finland, Helsinki, f School of Public Health, University of Tampere, Tampere, Finland
HPV vaccines are promising, having prevented 90–100% of persistent HPV infections and cervical dysplasia in clinical trials [1–5]. This has potential consequences for the sequelae of HPV infection, a necessary but not sufficient cause of cervical cancer [6–8] and a possible cause of other anogenital and oropharyngeal neoplasms [9, 10]. Cervical cancer is the second most common cancer in women worldwide with almost 500,000 cases diagnosed in 2002 [11]. Calculating the attributable fraction of HPV as a causative agent for cancer, Parkin [10] estimated that HPV is responsible for 5.2% of the world cancer burden, making it one of the most important infectious causes of cancer. HPV is a common sexually transmitted infection with at least 30 HPV types that infect the genital area, of which 15 are classified as high risk, i.e. having oncogenic potential [6]. HPV type 16 is the most common high-risk type, globally accounting for more than half (approximately 55%) of all cervical cancers [12]. Persistent infection with high-risk types is the most important risk factor for cervical cancer [13]. While most HPV infections are transient and resolve spontaneously, persistent infections over years can result in dysplastic precursor lesions, cervical intraepithelial neoplasia (CIN), which can progress to invasive cervical cancer (ICC) [14, 15]. The long pre-malignant course of HPV infection allows cytological screening programmes to detect and treat early disease and prevent progression to cervical cancer. Low-risk HPV types, of which types 6 and 11 are the most common, cause genital warts and lowgrade squamous intraepithelial lesions.
Trial results for candidate HPV virus-like particle vaccines are encouraging. A randomized, double-blind, phase II trial of an HPV16 vaccine found that the vaccine had an estimated efficacy of 100% (95% CI 90–100) and 91% (95% CI 80–97) in preventing persistent and any HPV16 infection over a median of 17 months [1]. A randomized, double-blind phase II trial of a bivalent HPV16/18 vaccine had, according to protocol analysis, an efficacy of 100% (95% CI 47–100) and 92% (95% CI 65–98) against persistent and all HPV16/18 infections over 18 months [2]. A randomized, double-blind placebocontrolled phase II trial of a quadrivalent vaccine targeting high-risk HPV types 16 and 18 and low-risk HPV types 6 and 11 found that combined incidence of persistent infection or disease with HPV6, HPV11, HPV16 or HPV18 fell by 90% over 36 months (95% CI 71–97) in those assigned vaccine compared with those assigned placebo [3]. In a phase III multinational study evaluating the same quadrivalent HPV6, HPV11, HPV16 and HPV18 vaccine, immunisation prevented 100% (95% CI 76–100) of HPV16/18-related moderate to severe CIN, adenocarcinoma in situ and squamous carcinoma according to the per protocol analysis at 24 months [4]. These results have created many future expectations of reducing cervical cancer incidence and altering cancer screening programmes. However, once the efficacy in protecting against infection and disease has been adequately demonstrated, many questions remain about how to integrate HPV vaccination into cervical cancer control strategies, who should be vaccinated and when to optimize the impact of the vaccine at the population level. Such currently unresolved questions require a debate that can be informed by analyses of mathematical models. Mathematical models provide a framework within which our understanding of HPV epidemiology can be explored and the theoretical potential of different vaccination programmes observed. In this paper we explore the potential impact on disease incidence of vaccinating men and women compared with vaccinating women only, vaccination before and after sexual debut, duration of vaccine-conferred protection and vaccination combined with screening programmes.
Methods Model Definition and Parameter Values A compartmental, deterministic model of HPV16 infection and progression to cervical cancer was developed and used to explore the epidemiology of the virus in Finland. Details of this approach have been published elsewhere [16]. For women, the model describes the flow of incident cases from the acquisition of asymptomatic HPV infection through premalignant disease to ICC. This progression is limited to a fraction of women, because most HPV infections regress spontaneously (fig. 1). A simple structure is included for men who
Future Issues in HPV Vaccination
243
Susceptible (a)
Vaccinated Deaths from ICC
Progression HPV
LSIL
HSIL
ICC
Regression Immune
Screening and treatment
Cancer survivors
Fig. 1. Model schematic of HPV16 natural history in women. Susceptible women acquire HPV as determined by the force of infection, (a), which is the susceptible age-specific risk of acquiring infection. Asymptomatic HPV infection can progress to low-grade squamous intraepithelial lesions (LSIL), HSIL and ICC, but most infections regress spontaneously to an immune state. Ten percent of asymptomatic HPV progresses rapidly to HSIL. Screening and treatment can prevent progress from HSIL to ICC. The model allows for benign hysterectomy at any stage and accounts for loss of detectable antibody over time.
can be susceptible, infected or immune. The spontaneous clearance of virus in most acute cases, the infrequent re-infection with a particular type and the age pattern of infection indicate the natural acquisition of type-specific immunity. We assume that this acquired protection is lifelong. The model incorporates screening and treatment, benign hysterectomy, Finnish demographics and patterns of sexual behaviour. Sero-prevalence data were used to estimate the per partnership transmission probability of HPV from an infectious individual to a susceptible individual. Other model parameters were derived from the literature. The model was calibrated to Finnish age-specific cervical cancer incidence. Estimating the Potential Impact of HPV16 Vaccines on Cervical Cancer Incidence The model, based on patterns of sexual behaviour in Finland in 1992 [17] and the current screening programme (every 5 years between 30 and 60 years of age) [18], was used to explore the potential impact of vaccination, varying vaccine coverage, age at vaccination and the duration of vaccine-conferred protection. For the vaccination ‘base case’, we assumed that 90% of successive cohorts of women were routinely vaccinated before sexual debut and that the vaccine had 100% efficacy with a lifelong duration of protection. The results for the duration of vaccine-conferred protection were compared with a model developed for lowincome countries, described in detail elsewhere [19]. Estimating the Potential Impact of HPV16 Vaccines on Dysplastic Lesions To look more closely at the impact of the age at vaccination on dysplastic lesions, the previously published model was modified to include sexual behaviour before age 15, starting at age 12; the proportion of the population sexually active at each age was set according to data from the Family Federation of Finland and is shown in table 1.
Barnabas/French/Laukkanen/Kontula/Lehtinen/Garnett
244
Table 1. Percentage of sexually active women at ages 12–20 in the model Age, years Sexually active, %
12 0.7
13 2.4
14 4.5
15 10.1
16 45.2
17 80.0
18 92.7
19 97.3
ⱖ20 99–100
Table 2. The potential steady state impact of vaccinating women and men compared with vaccinating women alone Strategy (all strategies assume vaccination before sexual debut and lifelong duration of vaccine-conferred protection)
Vaccination of 10% of women only Vaccination of 10% of women and men Vaccination of 70% of women only Vaccination of 70% of women and men Vaccination of 90% of women only Vaccination of 90% of women and men
Percentage of HPV16-associated cervical cancer cases prevented by age 85 13 16 71 83 91 98
Vaccination of 10% of women only Vaccination of 10% of women and men Vaccination of 70% of women only Vaccination of 70% of women and men Vaccination of 90% of women only Vaccination of 90% of women and men
Percentage of HSIL prevented by age 44 9 10 68 76 91 97
Results
The impact of varying vaccine coverage on cervical cancer incidence for vaccinating women alone or women and men is presented in table 2. At low (10%), moderate (70%) and high (90%) coverage, vaccinating women and men had a benefit over vaccinating women only (3, 12 and 7%, respectively). Additionally, we looked at the potential impact of vaccination on high-grade squamous intraepithelial lesions (HSIL) with different vaccination strategies. Up to age 44, at 90% coverage, vaccination of women and men prevented 6% more cases of HSIL compared with vaccination of women alone. To consider the impact of age at vaccination, the model assumes that 70% of women alone are vaccinated and that the vaccine confers lifelong protection.
Future Issues in HPV Vaccination
245
Proportion of cases prevented
0.8 0.7 0.6
Age 21 Age 18 Age 15 Age 12
0.5 0.4 0.3 0.2 0.1 0 2000
2010
2020
2030
2040
2050
Year
Fig. 2. Proportion of annual incident HPV16-associated HSIL prevented with different ages at vaccination. The potential impact over time of vaccination of 70% of women at ages 12, 15, 18 and 21. As the vaccinated cohort ages and becomes sexually active, a vaccine impact on incidence is observed.
Vaccinating young adolescents (12-year olds) has a delayed impact on incidence compared with older individuals (fig. 2). Peak HSIL incidence in the model is reached at age 19 (table 1). At ages when more of the population is sexually active, the impact of vaccination declines as the age of vaccination increases because of prior infection. There is relatively little difference in the long-term effect between vaccinating 12-year olds and 15-year olds, because in the modelled population, only a small proportion is sexually active at ages 12, 13 and 14 (table 1). When the full impact of vaccination is reached, the annual proportion of cases of HPV16-associated HSIL prevented is progressively higher as age of vaccination is reduced and is 35% if vaccination occurs at age 21, 54% if vaccination occurs at age 18, 68% if vaccination occurs at age 15, and 69% if vaccination occurs at age 12. The duration of vaccine-conferred protection has an important impact on ICC incidence, which depends upon what we assume about the relationship between age and the likelihood of an infection becoming persistent (fig. 3). If we assume that older women are more likely to have persistent HPV infections and progress to cervical cancer, then an average duration of vaccine-conferred protection of 10 years allows women to return to a susceptible state on loss of vaccine protection at a later age, when they are more likely to acquire persistent HPV infection. This means that there is no overall decrease in cancer incidence associated with the vaccine. The impact of this older susceptible cohort, which could be prevented by adequate booster vaccination, is not observed when we
Barnabas/French/Laukkanen/Kontula/Lehtinen/Garnett
246
Decrease in annual cervical cancer incidence (%)
0 20 40 60 80 100 0
20
40
60
80
Duration of protection (years)
Finnish model with older women more likely to have persistent lesions Low-income region model with older women more likely to have persistent lesions Finnish model with equal persistence rate across all ages
Fig. 3. The impact of varying the duration of vaccine-conferred protection. The impact of varying duration of vaccine efficacy on the incidence of HPV16-associated ICC for vaccination of 90% of women before sexual debut for parameters associated with Finland and for low-income regions is illustrated. Screening is not included. Because older women are assumed to be more likely to have persistent infections (a precursor to cancer) than younger women, a vaccine with a duration of 10 years or less shifts incident infections to older women (who are more likely to progress to cancer) and there is no reduction in the incidence of ICC. If women at all ages are likely to have transient infections, then ICC decreases with increasing vaccine duration, and vaccine duration of 15 years reduces HPV16-associated ICC incidence by 70%.
assume no age dependency in the likelihood of persistent infections. We compared this finding with a transmission model we developed for low-income regions [19] and found that under the assumption that older women are more likely to have persistent infections, immunity for at least 40 years was needed in both contexts to realise substantial benefits from the vaccine. Table 3 explores the effect of adding vaccination to screening programmes. Vaccinating 90% of women, without screening, reduces modelled HPV16-associated ICC incidence by 91%, whereas vaccination and screening every 10 years (rather than the current Finnish protocol of 5 years) is predicted to reduce incidence by 94%. Combined vaccination with the current Finnish
Future Issues in HPV Vaccination
247
Table 3. The predicted impact of intervention strategies to prevent cervical cancer Intervention strategy
Cervical cancer cases prevented, (%)
Vaccination only1 Vaccination1 and screening every 10 years Vaccination1 and screening every 5 years2
91 94 97
1 Vaccination strategy assumes vaccination of 90% of women alone, before sexual debut with lifelong duration of vaccine-conferred protection. 2 Screening assumes current Finnish guidelines: screening every 5 years between 30 and 60 years of age.
screening programme is predicted to reduce type-16-associated cancer incidence by 97%.
Discussion
The Finnish model presented here is limited in that it describes HPV16 infection only and assumes lifelong immunity to HPV16 following recovery from infection [16]. However, the transmission model provides a framework to explore some of the unresolved issues regarding HPV vaccine implementation. Incorporating patterns of sexual behaviour allows the model to consider whether women and men or women alone should be vaccinated, the impact of varying the age at vaccination and the effect of varying the duration of vaccineconferred protection. The impact of vaccination in our model is comparable with findings from previous modelling exercises [20–24], where high coverage of women, vaccination before sexual debut and long-term protection (or boosters) providing 3–4 decades of protection are required to substantially reduce ICC incidence. Vaccinating men as well as women prevents more cases of cancer and high-grade dysplasia than vaccinating women alone. Depending on the population coverage, this benefit can be small in terms of incident cancer cases prevented (3–12% additional cases prevented as shown in table 2). Taira et al. [24] found in their modelling evaluation of HPV vaccination programmes that including men further reduced cancer cases by 2.2%, but that it was not cost effective compared with women only vaccination. Thus, the benefits in terms of cervical cancer of vaccinating men may be small. However, the benefits have yet to be evaluated in terms of preventing other anogenital tumours.
Barnabas/French/Laukkanen/Kontula/Lehtinen/Garnett
248
The model predicts that vaccination before sexual debut provides the greatest benefit in terms of cases prevented (fig. 2). Delaying the age at vaccination to age 18 prevents 22% fewer annual cases of HSIL than could be prevented with vaccination at age 12. Catch-up vaccination at the start of a vaccination programme would presumably increase the speed with which a decrease in incidence is observed. The behaviour of young adolescents, their numbers of sexual partners and age of onset of sexual activity are obviously important in determining the age at which vaccination would be most beneficial. If activity is higher in young people, the benefits of vaccinating at younger ages would be greater than observed here. The converse is also true, and reliable data on adolescent behaviour would help identify the optimum age for vaccination. It is possible that a vaccine providing protection of less than 10 years could generate perverse outcomes by shifting susceptibility in women to an older age group. This perverse outcome would result if older women have increased risk of persistence, as was found by Goldie et al. [22] in their cohort natural history model of HPV infection and cervical cancer. In assessing the cost-effectiveness of HPV vaccination, Kulasingam et al. [21], assuming that older women were more likely to have persistent infections, found that a vaccine with 10 years duration of protection was not cost-effective alone compared with current screening protocols in the United States. However, there is a danger of misinterpreting cohort studies of viral clearance. If women already infected are recruited, then older women will more likely already have chronic infections, creating a biased sample of incident HPV infections. HPV vaccine boosters could extend the protection conferred by vaccination if immune memory does not produce an effective, timely response to HPV challenge. Candidate vaccines currently in phase III/IV trials protect against 2 (HPV16 and HPV18) of the 15 identified oncogenic HPV types [6]. Muñoz et al. [25] estimated that an HPV16/18 vaccine could prevent 71% of cervical cancers worldwide. However, regional variation would mean that more cases in Asia and Europe/North America would be prevented. A vaccine containing the 7 most common HPV types (16, 18, 45, 31, 33, 52 and 58) would prevent about 87% of cervical cancers worldwide regardless of regional variation. Screening programmes have the potential to detect precancerous lesions associated with non-vaccine HPV types. The model presented here assumes that HPV types are independent and vaccination against one type would not change the prevalence of other oncogenic types. It is possible that antagonistic or synergistic interactions between HPV vaccine and non-vaccine types may impact on vaccine effectiveness. Observational studies have found that 20–30% of HPV infections occur with multiple HPV types and acquisition of multiple types occurs more frequently
Future Issues in HPV Vaccination
249
than would be expected by chance [26–29], suggesting a synergistic interaction where infection with one type facilitates concurrent or subsequent infection with another HPV type. Elbasha and Galvani [30], in modelling vaccination against multiple types, showed that if interactions among HPV types are synergistic, vaccination may decrease the prevalence of types not included in the vaccine. Conversely, antagonistic interactions would decrease potential vaccine gains. The further development of mathematical models for HPV will help consider the effect of type-specific vaccination, under different assumptions for antagonistic or synergistic interactions between HPV types. Additionally, estimation of the impact of vaccination on the global burden of infection and disease caused by HPV will contribute to the health economic analysis of vaccination programmes. An effective HPV vaccine has the potential to reduce the burden of cervical cancer, the second most common cancer of women worldwide. However, from our modelling analyses, vaccine implementation needs to ensure high coverage, in terms of both the target population and oncogenic HPV types, sustained over many decades, to substantially reduce cervical cancer incidence. The role of both vaccination and cytological screening in cervical cancer prevention needs careful assessment. Ongoing trials will not only yield data on the vaccine but can also improve our understanding of the natural history of HPV infection and the interactions, if any, between HPV types.
Acknowledgements The authors would like to thank Elina Haavio-Mannila for the use of the Finnish sexual behaviour data.
References 1
2
3
Koutsky LA, Ault KA, Wheeler CM, Brown DR, Barr E, Alvarez FB, Chiacchierini LM, Jansen KU: A controlled trial of a human papillomavirus type 16 vaccine. N Engl J Med 2002;347: 1645–1651. Harper DM, Franco EL, Wheeler C, Ferris DG, Jenkins D, Schuind A, Zahaf T, Innis B, Naud P, De Carvalho NS, Roteli-Martins CM, Teixeira J, Blatter MM, Korn AP, Quint W, Dubin G: Efficacy of a bivalent L1 virus-like particle vaccine in prevention of infection with human papillomavirus types 16 and 18 in young women: a randomized controlled trial. Lancet 2004;364:1757–1765. Villa LL, Costa RL, Petta CA, Andrade RP, Ault KA, Giuliano AR, Wheeler CM, Koutsky LA, Malm C, Lehtinen M, Skjeldestad FE, Olsson SE, Steinwall M, Brown DR, Kurman RJ, Ronnett BM, Stoler MH, Ferenczy A, Harper DM, Tamms GM, Yu J, Lupinacci L, Railkar R, Taddeo FJ, Jansen KU, Esser MT, Sings HL, Saah AJ, Barr E: Prophylactic quadrivalent human papillomavirus (types 6, 11, 16, and 18) L1 virus-like particle vaccine in young women: a randomized double-blind placebo-controlled multicentre phase II efficacy trial. Lancet Oncol 2005;6:271–278.
Barnabas/French/Laukkanen/Kontula/Lehtinen/Garnett
250
4
5
6
7
8 9 10 11 12 13 14
15
16
17 18
19
20 21 22
23
24
Skjeldestad FE, for the Future II Steering Committee: Prophylactic quadrivalent human papillomavirus (HPV) (types 6, 11, 16, 18) L1 virus-like particle (VLP) vaccine (Gardasil™) reduces cervical intraepithelial neoplasia (CIN) 2/3 risk. 43rd Annu Meet IDSA, San Francisco, 2005. Mao C, Koutsky LA, Ault KA, Wheeler CM, Brown DR, Wiley DJ, Alvarez FB, Bautista OM, Jansen KU, Barr E: Efficacy of human papillomavirus-16 vaccine to prevent cervical intraepithelial neoplasia: a randomized controlled trial. Obstet Gynecol 2006;107:18–27. Munoz N, Bosch FX, de Sanjose S, Herrero R, Castellsague X, Shah KV, Snijders PJ, Meijer CJ: Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518–527. Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, Snijders PJ, Peto J, Meijer CJ, Munoz N: Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189:12–19. Zur Hausen H: Papillomaviruses causing cancer: evasion from host-cell control in early events in carcinogenesis. J Natl Cancer Inst 2000;92:690–698. Melbye M, Frisch M: The role of human papillomaviruses in anogenital cancers. Semin Cancer Biol 1998;8:307–313. Parkin DM: The global health burden of infection-associated cancers in the year 2002. Int J Cancer DOI: 10.1002/ijc.21731. Ferlay J, Bray F, Pisani P, Parkin DM: GLOBOCAN 2002: Cancer Incidence, Mortality and Prevalence Worldwide. IARC CancerBase. Lyon, IARCPress, 2004, vol 5. Clifford GM, Smith JS, Plummer M, Munoz N, Franceschi S: Human papillomavirus types in invasive cervical cancer worldwide: a meta-analysis. Br J Cancer 2003;88:63–73. Ho GY, Bierman R, Beardsley L, Chang CJ, Burk RD: Natural history of cervicovaginal papillomavirus infection in young women. N Engl J Med 1998;338:423–428. Ho GY, Burk RD, Klein S, Kadish AS, Chang CJ, Palan P, Basu J, Tachezy R, Lewis R, Romney S: Persistent genital human papillomavirus infection as a risk factor for persistent cervical dysplasia. J Natl Cancer Inst 1995;87:1365–1371. Kiviat N, Koutsky L, Paavonen J: Cervical neoplasia and other STD-related genital tract neoplasias; in Holmes KK (ed): Sexually Transmitted Diseases. New York, McGraw-Hill Health Professions Division, 1999. Barnabas RV, Laukkanen P, Koskela P, Kontula O, Lehtinen M, Garnett GP: The epidemiology of HPV 16 and cervical cancer in Finland and the potential impact of vaccination: mathematical modelling analyses. PLoS M 2006;3:e138. Haavio-Mannila E, Kontula O, Kussi E: Trends in Sexual Life. Helsinki, Population Research Institute, 2001. Anttila A, Pukkala E, Soderman B, Kallio M, Nieminen P, Hakama M: Effect of organized screening on cervical cancer incidence and mortality in Finland, 1963–1995: recent increase in cervical cancer incidence. Int J Cancer 1999;83:59–65. Barnabas RV, Garnett GP: The potential public health impact of vaccines against human papillomavirus; in Prendiville W, Davies P (eds): The Clinical Handbook of Human Papillomavirus. Lancaster, Parthenon Publishing/Parthenon Medical Communications, 2004, vol 2. Hughes JP, Garnett GP, Koutsky LA: The theoretical population level impact of a prophylactic human papillomavirus vaccine. Epidemiology 2002;13:631–639. Kulasingam SL, Myers ER: Potential health and economic impact of adding a human papillomavirus vaccine to screening programs. JAMA 2003;290:781–789. Goldie SJ, Grima D, Kohli M, Wright TC, Weinstein M, Franco E: A comprehensive natural history model of HPV infection and cervical cancer to estimate the clinical impact of a prophylactic HPV-16/18 vaccine. Int J Cancer 2003;106:896–904. Goldie SJ, Kohli M, Grima D, Weinstein MC, Wright TC, Bosch FX, Franco E: Projected clinical benefits and cost-effectiveness of a human papillomavirus 16/18 vaccine. J Natl Cancer Inst 2004;96:604–615. Taira AV, Neukermans CP, Sanders GD: Evaluating human papillomavirus vaccination programs. Emerg Infect Dis 2004;10:1915–1923.
Future Issues in HPV Vaccination
251
25
26
27
28
29
30
Munoz N, Bosch FX, Castellsague X, Diaz M, De Sanjose S, Hammouda D, Shah KV, Meijer CJ: Against which human papillomavirus types shall we vaccinate and screen? The international perspective. Int J Cancer 2004;111:278–285. Thomas KK, Hughes JP, Kuypers JM, Kiviat NB, Lee SK, Adam DE, Koutsky LA: Concurrent and sequential acquisition of different genital human papillomavirus types. J Infect Dis 2000;182: 1097–1102. Liaw KL, Hildesheim A, Burk RD, Gravitt P, Wacholder S, Manos MM, Scott DR, Sherman ME, Kurman RJ, Glass AG, Anderson SM, Schiffman M: A prospective study of human papillomavirus (HPV) type 16 DNA detection by polymerase chain reaction and its association with acquisition and persistence of other HPV types. J Infect Dis 2001;183:8–15. Rousseau MC, Pereira JS, Prado JC, Villa LL, Rohan TE, Franco EL: Cervical coinfection with human papillomavirus (HPV) types as a predictor of acquisition and persistence of HPV infection. J Infect Dis 2001;184:1508–1517. Mendez F, Munoz N, Posso H, Molano M, Moreno V, van den Brule AJ, Ronderos M, Meijer C, Munoz A: Cervical coinfection with human papillomavirus (HPV) types and possible implications for the prevention of cervical cancer by HPV vaccines. J Infect Dis 2005;192:1158–1165. Elbasha EH, Galvani AP: Vaccination against multiple HPV types. Math Biosci 2005;197:88–117.
Ruanne Barnabas, MBChB, DPhil Cancer Epidemiology Unit, Richard Doll Building Old Road Campus, Headington Oxford, OX3 7LF (UK) Tel. ⫹44 01865 289620, Fax ⫹44 01865 289610, E-Mail
[email protected] Barnabas/French/Laukkanen/Kontula/Lehtinen/Garnett
252
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 253–265
Public Health Issues Related to HPV Vaccination David Jenkins GlaxoSmithKline Biologicals, Rixensart, Belgium
Each year nearly half a million new cases of cervical cancer are diagnosed and over 250,000 women die from the disease [1], with approximately 8 out of 10 of these cases in developing countries [2]. In Europe, there is considerable variability in cervical cancer incidence and mortality, which in general correlates well with the cervical cancer screening program of each country. Countries with well-organized comprehensive national screening programs have the lowest cervical cancer incidence and mortality, whereas in European countries without such schemes, cervical cancer incidence and mortality figures may approach those in the developing world. It has been estimated that in 2002, 493,000 women had cervical cancer worldwide, which represents 10% of all female cancers [1]. In Northern and Western Europe, where comprehensive cervical screening programs exist, cervical cancer is partly controlled. However, in Eastern Europe, screening programs are generally limited and cervical cancer incidence rates are the highest in the developed world. In the 25 states of the EU, overall mortality associated with cervical cancer is approximately 14,400 per annum. Age-standardized incidence rates for women aged 15–65 years are as high as 24 per 100,000 in the Czech Republic, 27 per 100,000 in Slovakia, 27 per 100,000 in Poland, and over 30 per 100,000 in Romania. Current Prevention Programs through Cervical Screening
In some European countries, cervical screening programs have reduced cervical cancer rates by approximately 80%, but this has not been achieved in many countries [3]. European guidelines for cervical cancer screening programs were produced in 1993 [4] and further recommendations were added
in 1999 [5]. These recommend Pap smear tests, beginning no later than at 30 years of age but not before 20 years of age, with an ideal upper age limit of 60 years of age and an interval between tests of 3–5 years. In 2004, the International Agency for Research on Cancer, a WHO initiative, endorsed the cervical cancer screening program of the UK, which is now recognized as best practice [6]. This system, which was introduced in 1988, has been estimated to prevent approximately 4,500–6,000 deaths each year, with deaths from cervical cancer falling from 7.51 per 100,000 women aged 30–34 in the period 1983–1987 to 3.97 for the period 1998–2002 [7]. Similar schemes operate in the Netherlands and Finland and have been equally effective. The effectiveness of screening programs depends on comprehensive coverage and regular screening of the whole ‘at-risk’ female population, i.e. women aged 25–65 years. However, successful cervical cancer screening programs are expensive. On the other hand, early detection of precancerous cervical lesions can reduce overall costs associated with the management of invasive cervical cancer. In the UK, e.g., the Cervical Screening Program of the National Health Service costs approximately GBP 135 million (EUR 197 million) per annum [8]. In one region of the UK, the average mean cost of management across all grades of preinvasive cancer was found to be GBP 386 (EUR 564), compared with the mean 5-year total costs for managing invasive cancer of GBP 6,623 (EUR 9,670) for stage 1 and GBP 11,000 (EUR 16,060) for stages 2–4 [9]. Cytology-based screening programs optimally require well-organized, nation-wide appointment and recall systems to ensure that the target population of women can be reached and followed up. Cytology-based screening can be highly effective; however, the performance of these tests is not perfect. There remains a lack of consensus on the sensitivity of this testing method. Although the sensitivity of cytology-based testing can be 75–80%, in some programs, it has been reported to be only 55–65% on average across multiple countries, with a considerable number of samples rated as ‘inadequate’ and ‘unable to evaluate’ [10]. Several population-based studies are currently evaluating the value of an approach based upon the detection of viral DNA from HPV [11–13]. This is more sensitive than cytology-based testing but not more specific [14]. It is likely that the introduction of HPV DNA testing in both triage and primary screening will prove to be more effective and cost-effective than cytology testing through its increased sensitivity and reproducibility [15, 16]. However, issues remain for primary HPV DNA screening concerning the diagnosis of relatively large numbers of women who are HPV DNA positive but lack any morphological abnormality or other clinically significant manifestations of HPV infection.
Jenkins
254
The Role of HPV in Cervical Cancer
The two types of HPV most commonly associated with cervical cancer are HPV16 and HPV18; it has been estimated that these two types are responsible for up to 71.5% of cervical cancers in Europe [17]. HPV is the most common sexually transmitted infection with more than 30 types known to infect the anogenital tract. It is now recognized that the necessary cause of cervical cancer is persistent infection with certain types of HPV [17–22], and the WHO recognizes cervical cancer as the only cancer that can be fully attributed to a viral infection [22].
HPV Infection in Precancer Lesions and Cancer Infection with HPV is rare in children, but common in adolescent girls and young adult women who acquire the infection through sexual activity [23]. For example, in one study of women attending college in the USA, more than 50% who were HPV negative at enrolment and who reported never having had sexual intercourse became HPV positive within 4 years [24]. It has also been estimated that over 50% of sexually active women will be infected with HPV at some time during their lives [25, 26]. Data from population surveys and case-control studies suggest that the worldwide prevalence of HPV among young women is 5–40% [25, 27]. The prevalence is highest in women aged less than 25 years and then declining after 30 years of age, with a secondary peak in older women in some populations [7, 28–31]. Natural infection with HPV does not have a blood-borne phase and viral replication takes place within cervical epithelial cells. As a result, antigen-presenting cells have minimal exposure to viral capsid proteins, and consequently, there is only a weak humoral immune response, which does not provide consistent protection [32]. Women at all ages can be at risk for incident HPV infection [32]. Nevertheless, the immune system is usually able to combat infection with HPV. It typically takes 8–14 months to clear HPV, with clearance of HPV16 taking longer than low-risk HPV types, which are normally cleared within 5–6 months [33]. The infection becomes persistent in a small proportion of infected women, and the changes in the cervical epithelium can progress to high-grade cancer precursors. Untreated, 15–50% of these may progress to invasive cancer over a median period of 30 years [34]. Based on a Markov model, the lifetime risk of invasive cervical cancer for a cohort of women aged 15–85 years has been calculated as being 3–4%, with a lifetime cancer mortality risk of 1.26% and a peak incidence of 81 per 100,000 at age 50 years [35].
HPV Vaccination and Public Health
255
Vaccination to Prevent Cervical Cancer
The HPV vaccines in late-stage clinical development are made from DNAfree virus-like particles comprising the major capsid proteins L1 and adjuvants to improve the immune response. These are morphologically and antigenically similar to authentic papillomavirus virions [36–38]. These virus-like particles are capable of inducing high titers of neutralizing antibodies to L1 epitopes [39, 40]. Two different vaccines have been developed covering oncogenic HPV16 and HPV18 types [31, 41, 42]. Efficacy and Effectiveness Clinical trials of vaccines against HPV16/18 have been extremely successful so far. Vaccines were 95–100% effective against persistent infection and against cytological and histological events associated with HPV16/18 in women aged 15–25 years [41, 42]. Approximately 71.5% of cervical cancers in Europe are attributable to HPV16 and HPV18 [17], and HPV16 and HPV18 are involved in a higher proportion of adenocarcinoma, which is less effectively prevented by screening, than squamous cell carcinoma [17, 43, 44]. Therefore, it can be expected that an effective vaccine against these two types of the virus would have a major beneficial impact on the burden caused by cervical cancer, future screening practices and management of precancer lesions. The best protection will come from the vaccine with the broadest and longest protection against oncogenic HPV types. Target Population for HPV Vaccination Programs A number of risk factors have been identified for HPV infection and cervical cancer: the number of lifetime and recent sexual partners, age at first sexual intercourse, smoking, use of oral contraceptives, other sexually transmitted infections, chronic cervical inflammation, immunosuppressive conditions (including infection with HIV) and parity [24–26, 45–49]. The most consistent determinants are sexual activity and age at first sexual activity [50]. In order to effectively prevent oncogenic HPV infection, vaccination should ideally occur before the onset of sexual debut [51–55]. According to a WHO survey in Europe, approximately 20% of 15-year-old girls have already experienced sexual intercourse [56]. Recent studies also suggest that in Europe the number of sexual partners is increasing and that the age of sexual debut is falling [57]. For example, in the UK, the median age of sexual debut for women was 21 years in the 1960s but had fallen to 17 years by the 1990s [58]. Consequently, when the vaccines become available, vaccination will need to be offered before girls will be exposed to infection through sexual activity, i.e. between the ages of 10 and 14 years [59]. Infection with HPV is
Jenkins
256
spread to adolescent girls and women mainly from infected male sexual partners who are often somewhat older and more sexually experienced than their female partners. Although HPV infection can lead to cancers in men, these are much less frequent in the general population than is cervical cancer in women. It might also be argued that breaking the chain of infection could best be addressed by vaccinating both boys and girls; however, efficacy in men has not been proven [60, 61]. Thus, male vaccination, against a background of high vaccination coverage within the female population, is not predicted to be costeffective in mathematical models of transmission and progression of oncogenic HPV infection. In addition, male vaccination could double the associated cost of vaccination at only a minimal gain in terms of reduction in the rate of infection and consequently cancer [62]. The results of efficacy studies, early age of sexual debut, frequency of incident HPV infection in adolescents and the well-established link between oncogenic HPV16 and HPV18 infection and cervical cancer all combine to suggest that the greatest and most cost-effective health benefits in preventing precancerous cervical lesions and invasive cancer will come from well-organized, well-funded, comprehensive, high-coverage mass HPV vaccination of adolescent girls and adult women. Challenges to Vaccination Programs Sexually active women of all ages remain exposed to new oncogenic HPV infection [63] which may progress to precancer and cancer, although the frequency may be less than in younger women. Such women are likely to turn to their gynecologist or primary health care practitioner for advice and vaccination. Achieving high coverage of HPV16/18 vaccination in adolescent girls and adult women is a challenge to vaccination programs. The optimal age to vaccinate can be defined by age of exposure as a consequence of starting sexual activity, but decisions about vaccination are also likely to be influenced by issues related to delivery of immunization and by levels of knowledge and acceptability among the girls and their parents. Vaccination of single-age cohorts of females may be supplemented by catch-up strategies involving a broader age range of young women to more quickly achieve a high population impact. It is also likely that many adult women will choose to be vaccinated, even if vaccination for them is not provided as a mass public health service. One route for delivering vaccination to preadolescent and adolescent girls is through a school-based vaccine program such as with hepatitis B. One advantage of a school-based program is that girls in the target age group generally have minimal contact with healthcare systems outside of school. Special provisions would need to be made to reach girls in the target group who for one reason or another do not attend school. A successful vaccination program may
HPV Vaccination and Public Health
257
need to provide several opportunities to attend for vaccination, and this can be supported by active outreach to contact and follow up individuals at home. Another challenge for healthcare systems implementing HPV vaccination could come from a low perception of the need for such vaccination and this will affect the acceptability of the vaccine. Several studies have highlighted the poor state of knowledge about HPV and its role in the etiology of cervical cancer in both Europe and North America [64–68]. However, there is a general desire among adult and adolescent women for more information [69]. Therefore, good, properly targeted informative materials must be developed and made available to adolescents and their families if they are to be able to make properly informed choices about HPV vaccination. A recent study of the acceptability of HPV vaccination among young men and women [70] revealed that both sexes view HPV vaccination positively. The majority of their parents were also supportive, although a significant minority was opposed to the introduction of the vaccine. Some of these concerns were due to doubts about vaccines in general, but the provision of appropriate educational material could increase the acceptability of HPV vaccination among parents. One common worry among parents is that acceptance of the vaccine will encourage inappropriately early participation in sexual activity, especially unprotected sex [71]. Opposition can also be expected from individuals with strong moral objections to vaccination of any sort or against vaccinating young girls against a sexually transmitted disease [70, 72–74]. However, research suggests that attitudes to the vaccine among adolescents and young women will not be affected by the potential stigma of their view of infection with HPV as any sexually transmitted infection or that acceptance of the vaccine would be equivalent to indulging early in sexual activity [75–78]. Currently, most national vaccination programs are primarily aimed at infants. The healthcare professionals involved are only beginning to understand the role of HPV and the consequences of infection and may need further training and discussion of key issues with adolescents [79]. Other research suggests that educating vaccination clinicians about the prevalence of HPV infection and the associated risk of cervical cancer will increase the likelihood of their recommending anti-HPV vaccination to preadolescents [73].
Integrating HPV Vaccination and Screening Programs
It is important that initiating vaccination does not encourage women to withdraw from the effective protection provided by screening. However, in the longer term, HPV16/18 vaccination has the capacity to reduce the medical, personal and economic burden of cervical screening. Based on the efficacy of the
Jenkins
258
vaccines shown in clinical trials, mathematical models indicate that introduction of vaccination against the two most important oncogenic HPV types, HPV16 and HPV18, would reduce the incidence of cervical cancer in proportion to the fraction of cervical cancer (70%) attributable to these two types of HPV [35, 80]. This figure should be adjusted for coverage and the additional benefit to nonvaccinated women from herd immunity [31, 62]. Achieving maximum benefit will involve the introduction of comprehensive vaccination programs for preadolescent and adolescent girls integrated with existing vaccination programs, personal choice for vaccination by adult women and close integration with cervical cancer screening programs. In countries with less well-developed screening programs, vaccination alone could potentially reduce the burden of cervical cancer substantially and may well be more feasible than establishing cytological screening [81]. Introducing a vaccine against HPV16/18 cannot completely prevent cervical cancer, even when considering additional benefit from some crossprotection against related oncogenic HPV types. Therefore, it will be important to ensure that women are aware of the importance of continued screening, when this is available. However, HPV16/18 vaccination will decrease the number of abnormal smears, with an impact of around 20–30% on low-grade squamous intraepithelial lesion/atypical squamous cells of undetermined significance and 50% on high-grade squamous intraepithelial lesion smears or possibly more [42, 43] and will consequently reduce the number of women who will need further investigation and/or treatment for cervical intraepithelial neoplasia lesions. The number of referrals would fall further if oncogenic HPV DNA triage were added to regular cytology screening programs as recommended in the US guidelines developed by the American Society for Colposcopy and Cervical Pathology [82]. DNA testing is also being evaluated in several countries as a primary screening test with follow-up cytology testing [11]. Once vaccination is in place, it is likely that there will be a greater impetus to introduce HPV DNA testing and typing into screening programs. Introducing HPV DNA testing will have an additional epidemiological and public health value in relation to HPV vaccination by identifying the prevalence of the different types of HPV after vaccination, by reporting any vaccine breakthroughs and by making it possible for investigators to track the changing natural history of HPV infection after vaccination in actual clinical practice. It is also probable that with time, vaccination can lead to changes in guidelines for screening and the management of women with abnormal screening results. For example, there is evidence that HPV16 is more persistent and more likely to cause progression to cervical cancer than other oncogenic HPV types [83], and HPV18 has been identified as being more prone to cause
HPV Vaccination and Public Health
259
adenocarcinoma and progress differently [84]. Thus, the natural history of residual precancer as experienced in routine clinical practice after vaccination may change if infections by these two important HPV types are prevented. Some years after introduction, vaccination programs will begin to reduce the frequency of true disease detected by screening programs. At this point, the positive predictive value of cervical screening tests will fall, as there will be fewer abnormalities and more false positives. It is also possible that evaluation of smears will become more difficult and potentially less accurate as the proportion of normal cytology increases and abnormal cytology becomes less frequent. Introducing HPV DNA testing could help to offset any such trend. Finally, mathematical modeling based upon current practice in the USA suggests that in a vaccinated female population, it may well be cost-effective to increase the time between screening visits and that the cost saving will offset the costs of introducing vaccination [80].
Conclusions
The available evidence strongly suggests that introduction of comprehensive vaccination programs against HPV16/18 will have a major positive impact on reducing the burden of cervical cancer in screened and unscreened populations. The best protection will come from the vaccine with the broadest and longest protection against cervical cancer. However, considerable care will need to be taken in the implementation of programs to achieve high coverage in adolescent women for optimal benefit. Ensuring that the promise of these vaccines will be maximized will probably involve mass vaccination programs for one or more age cohorts, the development of appropriate information materials, and training. In addition, based on market research, the possibility that the vaccine will be generally well accepted by adolescent girls and their families is promising. In time, comprehensive mass vaccination programs will have a major impact on the types of HPV infections and the course of cervical disease associated with HPV. The time this takes will depend on the age of the cohort selected for vaccination, catch-up strategies by public health authorities and the individual choice of mature women to opt for vaccination. It may ultimately be necessary to modify screening programs and management protocols to take into account the impact of vaccination. However, it is important that HPV vaccination is not perceived by either women or healthcare professionals as an immediate alternative to cervical cancer screening. Integrating HPV vaccination into screening programs will also be necessary to maximize the benefits offered by the vaccine.
Jenkins
260
References 1
2 3
4
5
6 7 8
9 10 11
12
13
14
15
16
17
18
Ferlay J, Bray F, Pisani P, Parkin DM: GLOBOCAN 2002. Cancer Incidence, Mortality and Prevalence Worldwide. IARC CancerBase No 5, version 2.0. Lyon, IARCPress, 2004. http:// www-dep.iarc.fr/ Parkin DM, Bray F, Ferlay J, Pisani P: Global cancer statistics, 2002. CA Cancer J Clin 2005;55: 74–108. Anttila A, Nieminen P, Hakama M: Cervical cytology as a screening test; in Prendiville W, Davies P (eds): The Health Professional’s HPV Handbook. Oxon, Taylor and Francis, 2005, vol 3, pp 11–24. Coleman D, Day N, Douglas G, Farmery E, Lynge E, Philip J, Segnan N: European Guidelines for Quality Assurance in Cervical Cancer Screening. Europe against cancer programme. Eur J Cancer 1993;29A(suppl 4):S1–S38. ACCP: Recommendations on cancer screening in the European union; prepared by the Advisory Committee on Cancer Prevention (ACCP) after the Conference on Screening and Early Detection of Cancer, Vienna 18th–19th November 1999. Vienna, European commission, 1999. http:// europa.eu.int/comm/health/ph_determinants/genetics/cancer_screening_en.pdf IARC: IARC Handbooks of Cancer Prevention, vol 10: Cervix Cancer Screening. Lyon, WHO, 2005. Peto J, Gilham C, Fletcher O, Matthews FE: The cervical cancer epidemic that screening has prevented in the UK. Lancet 2004;364:249–256. Karnon J, Peters J, Platt J, Chilcott J, McGoogan E, Brewer N: Liquid-based cytology in cervical screening: an updated rapid and systematic review and economic analysis. Health Technol Assess 2004;8:1–78. Wolstenholme JL, Whynes DK: Stage-specific treatment costs for cervical cancer in the United Kingdom. Eur J Cancer 1998;34:1889–1893. Fahey MT, Irwig L, Macaskill P: Meta-analysis of Pap test accuracy. Am J Epidemiol 1995;141: 680–689. Cuzick J, Szarewski A, Cubie H, Hulman G, Kitchener H, Luesley D, McGoogan E, Menon U, Terry G, Edwards R, Brooks C, Desai M, Gie C, Ho L, Jacobs I, Pickles C, Sasieni P: Management of women who test positive for high-risk types of human papillomavirus: the HART study. Lancet 2003;362:1871–1876. Kitchener HC, Burns S, Nelson L, Myers AJ, Fletcher I, Desai M, Dunn G, Maguire P: A randomised controlled trial of cytological surveillance versus patient choice between surveillance and colposcopy in managing mildly abnormal cervical smears. BJOG 2004;111:63–70. Bulkmans NW, Rozendaal L, Snijders PJ, Voorhorst FJ, Boeke AJ, Zandwijken GR, van Kemenade FJ, Verheijen RH, van Groningen K, Boon ME, Keuning HJ, van Ballegooijen M, van den Brule AJ, Meijer CJ: POBASCAM, a population-based randomized controlled trial for implementation of high-risk HPV testing in cervical screening: design, methods and baseline data of 44,102 women. Int J Cancer 2004;110:94–101. Arbyn M, Buntinx F, Van Ranst M, Paraskevaidis E, Martin-Hirsch P, Dillner J: Virologic versus cytologic triage of women with equivocal Pap smears: a meta-analysis of the accuracy to detect high-grade intraepithelial neoplasia. J Natl Cancer Inst 2004;96:280–293. Legood R, Gray A, Wolstenholme J, Moss S: Lifetime effects, costs, and cost effectiveness of testing for human papillomavirus to manage low grade cytological abnormalities: modelling study 10.1136/bmj.38698.458866.7C. BMJ 2006;bmj.38698.458866.7C. Moss S, Gray A, Legood R, Vessey M, Patnick J, Kitchener H, the Liquid Based Cytology/Human Papillomavirus Cervical Pilot Studies Group: Effect of testing for human papillomavirus as a triage during screening for cervical cancer: observational before and after study 10.1136/bmj.38701.440961.7C. BMJ 2006;bmj.38701.440961.7C. Munoz N, Bosch FX, Castellsague X, Diaz M, de Sanjose S, Hammouda D, Shah KV, Meijer CJ: Against which human papillomavirus types shall we vaccinate and screen? The international perspective. Int J Cancer 2004;111:278–285. Zur Hausen H: Human genital cancer: synergism between two virus infections or synergism between a virus infection and initiating events? Lancet 1982;ii:1370–1372.
HPV Vaccination and Public Health
261
19 20
21
22 23 24
25 26 27
28
29 30
31
32
33
34 35
36
37 38
39
Burd EM: Human papillomavirus and cervical cancer. Clin Microbiol Rev 2003;16:1–17. Schiffman MH, Bauer HM, Hoover RN, Glass AG, Cadell DM, Rush BB, Scott DR, Sherman ME, Kurman RJ, Wacholder S, et al: Epidemiologic evidence showing that human papillomavirus infection causes most cervical intraepithelial neoplasia. J Natl Cancer Inst 1993;85:958–964. Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, Snijders PJ, Peto J, Meijer CJ, Munoz N: Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189:12–19. IARC: Human Papilloma Viruses, vol 64: Summary of Data Reported and Evaluation. Lyon, WHO, 1995. http://www-cie.iarc.fr/monoeval/allmonos.html Moscicki AB: Impact of HPV infection in adolescent populations. J Adolesc Health 2005;37: S3–S9. Winer RL, Lee SK, Hughes JP, Adam DE, Kiviat NB, Koutsky LA: Genital human papillomavirus infection: incidence and risk factors in a cohort of female university students. Am J Epidemiol 2003;157:218–226. Baseman JG, Koutsky LA: The epidemiology of human papillomavirus infections. J Clin Virol 2005;32(suppl 1):S16–S24. Koutsky L: Epidemiology of genital human papillomavirus infection. Am J Med 1997;102:3–8. Franco EL, Villa LL, Sobrinho JP, Prado JM, Rousseau MC, Desy M, Rohan TE: Epidemiology of acquisition and clearance of cervical human papillomavirus infection in women from a high-risk area for cervical cancer. J Infect Dis 1999;180:1415–1423. Burk RD, Ho GY, Beardsley L, Lempa M, Peters M, Bierman R: Sexual behavior and partner characteristics are the predominant risk factors for genital human papillomavirus infection in young women. J Infect Dis 1996;174:679–689. Crum CP: Contemporary theories of cervical carcinogenesis: the virus, the host, and the stem cell. Mod Pathol 2000;13:243–251. Dalstein V, Riethmuller D, Pretet JL, Le Bail Carval K, Sautiere JL, Carbillet JP, Kantelip B, Schaal JP, Mougin C: Persistence and load of high-risk HPV are predictors for development of high-grade cervical lesions: a longitudinal French cohort study. Int J Cancer 2003;106:396–403. Munoz N, Bosch FX, de Sanjose S, Herrero R, Castellsague X, Shah KV, Snijders PJ, Meijer CJ: Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518–527. Tjiong MY, Out TA, Ter Schegget J, Burger MP, Van Der Vange N: Epidemiologic and mucosal immunologic aspects of HPV infection and HPV-related cervical neoplasia in the lower female genital tract: a review. Int J Gynecol Cancer 2001;11:9–17. Brown DR, Shew ML, Qadadri B, Neptune N, Vargas M, Tu W, Juliar BE, Breen TE, Fortenberry JD: A longitudinal study of genital human papillomavirus infection in a cohort of closely followed adolescent women. J Infect Dis 2005;191:182–192. Ostor AG: Natural history of cervical intraepithelial neoplasia: a critical review. Int J Gynecol Pathol 1993;12:186–192. Myers ER, McCrory DC, Nanda K, Bastian L, Matchar DB: Mathematical model for the natural history of human papillomavirus infection and cervical carcinogenesis. Am J Epidemiol 2000;151:1158–1171. Christensen ND, Hopfl R, DiAngelo SL, Cladel NM, Patrick SD, Welsh PA, Budgeon LR, Reed CA, Kreider JW: Assembled baculovirus-expressed human papillomavirus type 11 L1 capsid protein virus-like particles are recognized by neutralizing monoclonal antibodies and induce high titres of neutralizing antibodies. J Gen Virol 1994;75:2271–2276. Hagensee ME, Olson NH, Baker TS, Galloway DA: Three-dimensional structure of vaccinia virus-produced human papillomavirus type 1 capsids. J Virol 1994;68:4503–4505. Rose RC, Reichman RC, Bonnez W: Human papillomavirus (HPV) type 11 recombinant viruslike particles induce the formation of neutralizing antibodies and detect HPV-specific antibodies in human sera. J Gen Virol 1994;75:2075–2079. Harro CD, Pang YY, Roden RB, Hildesheim A, Wang Z, Reynolds MJ, Mast TC, Robinson R, Murphy BR, Karron RA, Dillner J, Schiller JT, Lowy DR: Safety and immunogenicity trial in adult volunteers of a human papillomavirus 16 L1 virus-like particle vaccine. J Natl Cancer Inst 2001;93:284–292.
Jenkins
262
40
41
42
43 44 45 46 47
48
49
50 51
52
53
54 55 56
57 58 59
Kirnbauer R, Chandrachud LM, O’Neil BW, Wagner ER, Grindlay GJ, Armstrong A, McGarvie GM, Schiller JT, Lowy DR, Campo MS: Virus-like particles of bovine papillomavirus type 4 in prophylactic and therapeutic immunization. Virology 1996;219:37–44. Koutsky LA, Ault KA, Wheeler CM, Brown DR, Barr E, Alvarez FB, Chiacchierini LM, Jansen KU: A controlled trial of a human papillomavirus type 16 vaccine. N Engl J Med 2002;347: 1645–1651. Harper DM, Franco EL, Wheeler C, Ferris DG, Jenkins D, Schuind A, Zahaf T, Innis B, Naud P, De Carvalho NS, Roteli-Martins CM, Teixeira J, Blatter MM, Korn AP, Quint W, Dubin G: Efficacy of a bivalent L1 virus-like particle vaccine in prevention of infection with human papillomavirus types 16 and 18 in young women: a randomised controlled trial. Lancet 2004;364: 1757–1765. Clifford GM, Smith JS, Plummer M, Munoz N, Franceschi S: Human papillomavirus types in invasive cervical cancer worldwide: a meta-analysis. Br J Cancer 2003;88:63–73. Krane JF, Granter SR, Trask CE, Hogan CL, Lee KR: Papanicolaou smear sensitivity for the detection of adenocarcinoma of the cervix: a study of 49 cases. Cancer 2001;93:8–15. Schiffman M, Castle PE: Human papillomavirus: epidemiology and public health. Arch Pathol Lab Med 2003;127:930–934. Bosch FX, Munoz N: The viral etiology of cervical cancer. Virus Res 2002;89:183–190. Moscicki AB, Hills N, Shiboski S, Powell K, Jay N, Hanson E, Miller S, Clayton L, Farhat S, Broering J, Darragh T, Palefsky J: Risks for incident human papillomavirus infection and lowgrade squamous intraepithelial lesion development in young females. JAMA 2001;285: 2995–3002. Richardson H, Kelsall G, Tellier P, Voyer H, Abrahamowicz M, Ferenczy A, Coutlee F, Franco EL: The natural history of type-specific human papillomavirus infections in female university students. Cancer Epidemiol Biomarkers Prev 2003;12:485–490. Sellors JW, Karwalajtys TL, Kaczorowski J, Mahony JB, Lytwyn A, Chong S, Sparrow J, Lorincz A: Incidence, clearance and predictors of human papillomavirus infection in women. CMAJ 2003;168: 421–425. Ho GY, Bierman R, Beardsley L, Chang CJ, Burk RD: Natural history of cervicovaginal papillomavirus infection in young women. N Engl J Med 1998;338:423–428. Andersson-Ellstrom A, Dillner J, Hagmar B, Schiller J, Sapp M, Forssman L, Milsom I: Comparison of development of serum antibodies to HPV16 and HPV33 and acquisition of cervical HPV DNA among sexually experienced and virginal young girls. A longitudinal cohort study. Sex Transm Dis 1996;23:234–238. Kahn JA, Rosenthal SL, Succop PA, Ho GY, Burk RD: The interval between menarche and age of first sexual intercourse as a risk factor for subsequent HPV infection in adolescent and young adult women. J Pediatr 2002;141:718–723. Kahn JA, Rosenthal SL, Succop PA, Ho GY, Burk RD: Mediators of the association between age of first sexual intercourse and subsequent human papillomavirus infection. Pediatrics 2002; 109:E5. Scheurer ME, Tortolero-Luna G, Adler-Storthz K: Human papillomavirus infection: biology, epidemiology, and prevention. Int J Gynecol Cancer 2005;15:727–746. Shew ML, Fortenberry JD: HPV infection in adolescents: natural history, complications, and indicators for viral typing. Semin Pediatr Infect Dis 2005;16:168–174. Ross J, Godeau E, Dias S: Chapter 3: young peoples health and health-related behaviour – sexual health; in Currie C, Morgan A, Barnekow-Rassmussen V, Roberts C, Samdal O, Settertobulte W, Smith R (eds): WHO Young People’s Health in Context Report; Health Behavior in School-Aged Children (HBSC) Study: International Report from the 2001/2002 Survey. Copenhagen, WHO Regional Office for Europe, 2004, pp 153–160. Lehtinen M, Paavonen J: Effectiveness of preventive human papillomavirus vaccination. Int J STD AIDS 2003;14:787–792. Wellings K, Field B: Sexual behaviour in young people. Baillieres Clin Obstet Gynaecol 1996;10:139–160. Rosenthal SL: Protecting their adolescents from harm: parental views on STI vaccination. J Adolesc Health 2005;37:177–178.
HPV Vaccination and Public Health
263
60 61 62 63 64 65 66
67 68 69 70 71 72 73
74 75 76 77 78 79 80
81
82
Garnett GP, Dubin G, Slaoui M, Darcis T: The potential epidemiological impact of a genital herpes vaccine for women. Sex Transm Infect 2004;80:24–29. Hughes JP, Garnett GP, Koutsky L: The theoretical population-level impact of a prophylactic human papilloma virus vaccine. Epidemiology 2002;13:631–639. Taira AV, Neukermans CP, Sanders GD: Evaluating human papillomavirus vaccination programs. Emerg Infect Dis 2004;10:1915–1923. Grainge MJ, Seth R, Guo L, Neal KR, Coupland C, Vryenhoef P, Johnson J, Jenkins D: Cervical human papillomavirus screening among older women. Emerg Infect Dis 2005;11:1680–1685. Dell DL, Chen H, Ahmad F, Stewart DE: Knowledge about human papillomavirus among adolescents. Obstet Gynecol 2000;96:653–656. Holcomb B, Bailey JM, Crawford K, Ruffin MT: Adults’ knowledge and behaviors related to human papillomavirus infection. J Am Board Fam Pract 2004;17:26–31. Mays RM, Zimet GD, Winston Y, Kee R, Dickes J, Su L: Human papillomavirus, genital warts, Pap smears, and cervical cancer: knowledge and beliefs of adolescent and adult women. Health Care Women Int 2000;21:361–374. Waller J, McCaffery K, Forrest S, Szarewski A, Cadman L, Wardle J: Awareness of human papillomavirus among women attending a well woman clinic. Sex Transm Infect 2003;79:320–322. Yacobi E, Tennant C, Ferrante J, Pal N, Roetzheim R: University students’ knowledge and awareness of HPV. Prev Med 1999;28:535–541. Anhang R, Wright TC Jr, Smock L, Goldie SJ: Women’s desired information about human papillomavirus. Cancer 2004;100:315–320. Zimet GD: Improving adolescent health: focus on HPV vaccine acceptance. J Adolesc Health 2005;37:S17–S23. Kahn JA: Vaccination as a prevention strategy for human papillomavirus-related diseases. J Adolesc Health 2005;37:S10–S16. Olshen E, Woods ER, Austin SB, Luskin M, Bauchner H: Parental acceptance of the human papillomavirus vaccine. J Adolesc Health 2005;37:248–251. Kahn JA, Zimet GD, Bernstein DI, Riedesel JM, Lan D, Bin Huang B, Rosenthal SL: Pediatricians’ intention to administer human papillomavirus vaccine: the role of practice characteristics, knowledge, and attitudes. J Adolesc Health 2005;37:502–510. Mays RM, Sturm LA, Zimet GD: Parental perspectives on vaccinating children against sexually transmitted infections. Soc Sci Med 2004;58:1405–1413. Kahn JA, Rosenthal SL, Hamann T, Bernstein DI: Attitudes about human papillomavirus vaccine in young women. Int J STD AIDS 2003;14:300–306. Zimet GD, Mays RM, Fortenberry JD: Vaccines against sexually transmitted infections: promise and problems of the magic bullets for prevention and control. Sex Transm Dis 2000;27:49–52. Boehner CW, Howe SR, Bernstein DI, Rosenthal SL: Viral sexually transmitted disease vaccine acceptability among college students. Sex Transm Dis 2003;30:774–778. Hoover DR, Carfioli B, Moench EA: Attitudes of adolescent/young adult women toward human papillomavirus vaccination and clinical trials. Health Care Women Int 2000;21:375–391. Schuster P, Bell RM, Petersen LP, Kanouse DE: Communication between adolescents and physicians about sexual behavior and risk prevention. Arch Pediatr Adolesc Med 1996;150:906–913. Goldie SJ, Kohli M, Grima D, Weinstein MC, Wright TC, Bosch FX, Franco E: Projected clinical benefits and cost-effectiveness of a human papillomavirus 16/18 vaccine. J Natl Cancer Inst 2004;96:604–615. Goldie SJ, Grima D, Kohli M, Wright TC, Weinstein M, Franco E: A comprehensive natural history model of HPV infection and cervical cancer to estimate the clinical impact of a prophylactic HPV-16/18 vaccine. Int J Cancer 2003;106:896–904. Saslow D, Runowicz CD, Solomon D, Moscicki AB, Smith RA, Eyre HJ, Cohen C: American Cancer Society guideline for the early detection of cervical neoplasia and cancer. CA Cancer J Clin 2002;52:342–362.
Jenkins
264
83 84
Schiffman M, Kjaer SK: Chapter 2: natural history of anogenital human papillomavirus infection and neoplasia. J Natl Cancer Inst Monogr 2003;31:14–19. Woodman CB, Collins S, Rollason TP, Winter H, Bailey A, Yates M, Young LS: Human papillomavirus type 18 and rapidly progressing cervical intraepithelial neoplasia. Lancet 2003;361: 40–43.
Dr. David Jenkins, MD, PhD Clinical Development HPV vaccines GlaxoSmithKline Biologicals, Rue de l’Institut, 89 BE–1330 Rixensart (Belgium) Tel. ⫹32 2 656 2111, Fax ⫹32 2 656 2099, E-Mail
[email protected] HPV Vaccination and Public Health
265
Monsonego J (ed): Emerging Issues on HPV Infections: From Science to Practice. Basel, Karger, 2006, pp 266–268
How to Implement HPV Vaccines in Practice Diane M. Harper Departments of Community and Family Medicine and Obstetrics and Gynecology, Dartmouth Medical School, Hanover, N.H., USA
Implementation of the HPV vaccines is the true test of success of this science. An effective, safe, immunogenic, cost-effective agent to prevent one of the most common viruses from infecting human epithelium will only impact individual- and population-based health when it is available, accepted and used. Many logistical events are necessary at the detail level to ensure vaccine delivery, cold chain storage, actual injection and archived regulatory medical documentation. However, before these critical elements are pertinent, there are the abstract concepts of the understanding of the infection and human perspective that must be acknowledged prior to HPV vaccine acceptance.
Educating Physicians and Patients
Using a three-point educational format for HPV infection provides a simple skeleton for both physicians and patients to accrue information necessary to guide their decision making about HPV prophylaxis. Terminology that emphasizes impartiality removes associations similar to the latency/herpes simplex 2 link. Residential Infection This basal cell nuclear infection is a protein coat-less HPV virion that can persist in this lowermost cell layer undetected from weeks to decades after infection. Only one HPV type per cell can exist, but it can be a benign or oncogenic type. This HPV format is the necessary basis for the development of the remaining two formats and is prevented with the current HPV L1 virus-like particle vaccines.
Episomal Infection The HPV replicates in the basal cell nucleus separate from the human genome. As the basal cell differentiates into stratified squamous epithelium, the protein coat-less HPV continues to replicate episomally until it can reproduce its viral coat in the superficial layer where it is released as an infective virus to another basal stem skin cell. Both benign and oncogenic HPV types replicate episomally, causing identical abnormal cytology and colposcopically visible low-grade intraepithelial lesions. Integrated Infection Integration ensues in one of two ways. The first can occur after the residential infection with an oncogenic HPV type is established. The oncogenic HPV genome dissembles in the basal cell nucleus integrating with the human DNA preventing maturation of the basal cell into stratified layers, eventually either regressing, if small enough, or progressing into cancer. This is cytologically and colposcopically detected as a high-grade intraepithelial lesion (cervical intraepithelial neoplasia 2/3). The second occurs after the residential infection with an oncogenic HPV type has become a persistent cytologically and colposcopically visible episomal infection. The ordered viral replication is interrupted by chromosomal breaks, allowing viral integration with the human genome to occur, prohibiting HPV virion production and infectivity, prohibiting squamous cell maturation, and eventually, progressing into invasive cancer. Discussing HPV-Related Sexuality Issues Neutrally Personal perspective and interpretation color scientific understanding, especially when scientific implications bear on personal beliefs and morals. Trigger words that describe HPV infection as a sexually transmitted infection have raised similar issues that oral contraceptives highlighted in public discussions over 70 years ago. Messaging the scientific facts about HPV infection in neutral language is necessary before vaccination can be popularly accepted. Despite oral, anal and genital skin to skin contact, we have a responsibility to communicate that HPV transmission differs significantly from other infections classified as sexually transmitted. HPV infection lives in the 400 nm of epithelial surface that covers the mucosal organs; it cannot live internally. Bodily fluids are not pertinent to HPV transmission. The Treponema spirochete of syphilis is transmitted through abraded anogenital skin, but immediately invades the lymphatic and circulatory systems for multiorgan destruction. Likewise, viremic expression commonly seen in hepatitis B and HIV/AIDS, and cytolysis common with hepatitis simplex virus 2, do not occur with HPV infection. Finally, unlike chlamydia or gonorrhea whose primary site of
Implementing HPV Vaccines
267
infection is the endocervical cell, HPV infections occur in squamous epithelial stem cells, such as the basal and reserve cells. Educational messages that neither ignore nor promote sexuality facilitate a neutral discussion of HPV. To date, the psychological cost of HPV infections has been primarily borne by women. The anxiety, depression, worry and horror that transitions from potentially having a cancer to having a lower genital tract infection can be alleviated or exacerbated by the manner in which the woman receives her HPV DNA or cytology test results. Relationships and families can be destroyed or strengthened by words of condemnation or comfort. The quality of life with an HPV-associated disease can fluctuate daily for her and her partner, making protection against this virus a genderless goal. Infection Affects Both Genders This particularly common virus must be described in terms of a human infection in both men and women. Benign mucosal HPV types replace normally functioning epithelium in the multiple forms of warts in both genders, limiting the external barrier protection that the epithelium is designed to provide. Cancers in the mouth, anus, vagina, cervix, penis, vulva and oro-pharynx can develop from the oncogenic HPV infections requiring surgical treatment to avoid death in both men and women. The L1 virus-like particle vaccines designed to prevent the initial residential HPV infection in the basal stem cell could eventually provide protection for up to 5% of all the cancers worldwide for both genders, numerically having the largest impact for women and cervical cancer. Communicating the science, the shared HPV risks and the benefits of vaccination, potentially for both genders, becomes germane to HPV vaccination implementation. Dr. Diane M. Harper, MD, MPH, MS Norris Cotton Cancer Center One Medical Center Drive Lebanon, NH 03756 (USA) Tel. ⫹1 603 653 3692, Fax ⫹1 603 653 9003, E-Mail
[email protected] Harper
268
Author Index
Barnabas, R.V. 242 Barr, E. 227 Birembaut, P. 103 Bory, J.-P. 103 Bosch, F.X. 206 Burchell, A.N. 20 Clavel, C. 103 Cox, J.T. 120 Dalstein, V. 103 Ferenczy, A. 178 Franco, E.L. 20 French, K.M. 242 Garland, S.M. 63 Garnett, G.P. 242 Graesslin, O. 103
Harper, D.M. 266 Hovland, S. 82
Morland, G. 82 Myers, E.R. 235
Jenkins, D. 253
Quereux, C. 103
Karlsen, F. 82 Kontula, O. 242 Kraus, I. 82 Laukkanen, P. 242 Lehtinen, M. 242 Lörincz, A.T. 54 Lowy, D.R. 217 McGoogan, E. 147 Meijer, C.J.L.M. 73 Molden, T. 82 Monsonego, J. IX, 184 Morland, E. 82
Savard, M. 44 Schiller, J.T. 217 Silva, I. 82 Singer, A. 165 Skomedal, H. 82 Snijders, P.J.F. 73 Stanley, M. 34 Steenbergen, R.D.M. 73 Syrjänen, K.J. 157 Tabrizi, S. 63 von Knebel Doeberitz, M. 1 Wright, T.C. 140
269
Subject Index
Acetic acid test, colposcopy 83, 166, 167 Aldara, genital wart treatment 39 Antigen-presenting cell (APC), host defense 34, 35 Atypical granular cells (AGC) cancer risks 142–144 classification 142 cytology guidelines initial evaluation 144, 145 subsequent evaluation 145 epidemiology 142 Atypical squamous cells of undetermined significance (ASCUS) ASC-H definition 121 epidemiology 122, 123 HPV testing in management 128–130, 135 definition 120, 121 epidemiology 122 HPV testing management 123–128, 135 postcolposcopy management 133–135 Pap test interpretation 122 AutoPap, computer-assisted screening 154 Bethesda system, see High-grade squamous intraepithelial lesion; Low-grade squamous intraepithelial lesion Biomarkers, cervical cancer chromosomal aberrations 6, 76 classification 74
deregulated E6/E7 expression and immortalization 74–76 DNA methylation 77, 78 gene expression profiling 8–12, 86–88 identification 6 malignant transformation and tumor invasion 76, 77 p16Ink4a 10–15, 74, 75, 85 screening aims 5, 6 validation 78 viral genome integration 7, 8 Cervical cancer biomarkers, see Biomarkers, cervical cancer diagnosis and screening, see Colposcopy; Diagnostics, HPV; Liquid-based cytology; Morphological diagnosis incidence and mortality 54, 184, 185, 253 risk levels 4, 5, 229, 242, 255 Cervical intraepithelial neoplasia (CIN) cancer progression risks 64 CIN2 biological behavior 160–162 independent morphological entity 158–160 placement within Bethesda system 162, 163 progression rate 162, 163 follow-up 182 grades 4, 158
270
HPV gene expression by grade 87 natural history 162 treatment 178–182 Cervical warts, see Condylomata acuminata Chlamydia trachomatis, cervical cancer risks 212 CIN2, see Cervical intraepithelial neoplasia Cold-knife conization (CKC), cervical intraepithelial neoplasia management 181 Colposcopy accuracy diagnosis 174 screening 174, 175 acetic acid and visual inspection of cervix 83, 166, 167 biopsy 168 colposcope 165 diagnostic criteria 170 diagnostic limitations 83 endocervical curettage 169 historical perspective 165 HPV testing in postcolposcopy management 133–135 indications abnormal cytology and suspected malignancy 172 primary screening 172 Lugol’s solution test 168 prospects 175, 176 reproducibility 175 saline technique 167, 168 Schiller’s iodine test 168 side effects 175 terminology 171, 172 unsatisfactory colposcopy 172 vascular patterns color tone 171 epithelial borders 171 intercapillary distance 170 surface contour 171 Condylomata acuminata, treatment 182 Cost-effectiveness, see Economic analysis Cryocoagulation, cervical intraepithelial neoplasia management 178, 179
Subject Index
Diagnostics, HPV cervical cancer impact 254 chip technology 86 Hybrid Capture 2 test 55, 56, 83 indications 69–71 limitations acetic acid and visual inspection of cervix 83 colposcopy 83 liquid-based cytology 82, 104, 105 persistent transforming infection 97 polymerase chain reaction 55, 65–69, 84 PreTect HPV-Proofer, see PreTect HPV-Proofer primary screening approaches 112 clinical trials in progress and prospects 114–116 developing countries 111, 112 HPV testing alone 111 initiation 112, 113 interval between tests 113 negative predictive value 109, 110 overview of studies 105–109 persistence of infection 109 professional guidelines 113 termination 113 transient infection and improvement of testing specificity 110, 111 prospects 71 rationale 54, 55 samples for nucleic acid detection media 91 preparation 91, 92 sampling 91 screening algorithms 103, 104 triage tests 59, 60, 96, 97, 135 validation of tests 57 DNA methylation, cervical cancer biomarker studies 77, 78 E1, expression during cervical carcinogenesis 87, 88 E2 cell-mediated immunity 38 expression during cervical carcinogenesis 87, 88
271
E6 cell-mediated immunity 38 chromosomal instability induction 3, 4 expression during cervical carcinogenesis 86, 87 transcript splice variants 88 E7 chromosomal instability induction 3, 4 expression during cervical carcinogenesis 86, 87 transcript splice variants 88 Economic analysis cervical cancer prevention strategies 237, 238 cost-effectiveness analysis cervical cancer screening 238, 239 HPV vaccines 239, 240 incremental cost-effectiveness ratio 236 quality-adjusted life year 236 Electrofulguration (EF), cervical intraepithelial neoplasia management 178–180 Epidemiology, HPV infection duration of infection 24, 25, 255 geographic distribution with cervical cancer 212–214 high-risk versus low-risk infections 20, 21, 103 incidence 23, 24 prevalence 22, 23, 103 prospects for study 30, 31 sexual partner number and infection risk 26, 27, 29, 30 sexual risk profiles 24–26 FocalPoint, computer-assisted screening 154 Genital warts, see Condylomata acuminata Herpes simplex virus type 2, cervical cancer risks 212 High-grade squamous intraepithelial lesion (HSIL) cancer risks 140 cytology guidelines
Subject Index
initial evaluation 141 pregnant patients 142 subsequent evaluation 141 definition 140, 141 epidemiology 141 Human immunodeficiency virus (HIV), cervical cancer risks 210–212 Human papillomavirus (HPV) diagnosis, see Diagnostics, HPV epidemiology, see Epidemiology, HPV infection episomal infection 267 genotyping 64 immune response, see Immune response, HPV integrated infection 267 natural history, see Natural history, human papillomavirus infection residual infection 266 risk classification 55, 63, 64, 89–91, 103, 206–208, 212, 227, 242 vaccination, see Vaccination, HPV Hybrid Capture 2 test (HC2) advances 57–59 advantages 83, 105 HPV serotype specificity and crossreactivity 55, 56, 83 principles 56 sensitivity 56 specimen collection 55, 56 Hysterectomy, cervical intraepithelial neoplasia management 181, 182 Immune response, HPV cell-mediated immunity 36–38 evasion 38, 39 high-risk infection and immune failure 39, 40 host defense to pathogens 34, 35 infectious cycle 35, 36 L1 protein vaccine rationale 231 Incremental cost-effectiveness ratio (ICER), definition 236 Ki-67, cervical cancer biomarker studies 9, 75
272
Laser ablation, cervical intraepithelial neoplasia management 178, 180 Liquid-based cytology (LBC) computer-assisted scanning devices 153–155 limitations 82, 104, 105 microscopic appearance of samples 149 performance studies sensitivity, specificity, and accuracy 151–153 unsatisfactory sample rate 150, 151 principles 148 sample preparation 148–150 Loop electrosurgical excision procedure (LEEP), cervical intraepithelial neoplasia management 180, 181 Low-grade squamous intraepithelial lesion (LSIL) definition 120, 121 epidemiology 123 HPV testing management 130–132 postcolposcopy management 133–135 Lugol’s solution test, colposcopy 168 MCM5, cervical cancer biomarker studies 75 Morphological diagnosis, see also Colposcopy CIN2 as independent morphological entity 158–160 classification schemes 157, 158 Natural history, HPV infection cervical cancer risks and mechanisms 186–188, 228 immunity 186–188 overview 2, 3, 87, 88, 242 virus type effects 188, 189 Nucleic acid sequence based amplification (NASBA), cervical cancer detection 85, 86 Oral contraceptives, cervical cancer risks with long-term use 209, 210
Subject Index
PAPNET, computer-assisted screening 154 Pap test, see also Liquid-based cytology computer-assisted scanning devices 153–155 impact on cancer mortality 47, 228, 253, 254 interpretation 122 limitations 1, 45, 54, 104, 147, 148 Patient-provider relationship advocacy groups and policymakers 52 conversation engagement 51, 52, 267, 268 counseling principles 48–50 empowerment of patients 44, 50, 51 history and evolution 45 patient education 47, 48, 53 practitioner education 45–47 vaccination education 258–260 Polymerase chain reaction (PCR), HPV diagnostics accuracy 95 consensus primers 65 interference 66 Linear array assay 66–68 overview 55, 65, 84 samples for nucleic acid detection media 91 preparation 91, 92 sampling 91 standardization 68, 69 transcript detection 84 Practitioner-patient relationship, see Patient-provider relationship Pregnancy, high parity and cervical cancer risks 210 PreTect HPV-Proofer, cervical cancer detection accuracy 92, 95, 96 comparison with DNA-based tests 97, 98 HPV type specificity 98, 99 markers 89 persistent transforming infection 97 principles 85, 89 sample preparation 91, 92 triage testing 96, 97
273
Proliferating cell nuclear antigen (PCNA), cervical cancer biomarker studies 9, 75 Quality-adjusted life year (QALY), definition 236 Risk cervical cancer risk levels 4, 5, 229, 242, 255 cervical intraepithelial neoplasia progression 64 epidemiology of HPV infection 55, 63, 64, 89–91, 103, 206–208, 212, 227, 242 high-risk vs. low-risk HPV infections 20, 21, 103 sexual partner number and infection risk 26, 27, 29, 30 sexual risk profiles 24–26 Saline technique, colposcopy 167, 168 Schiller’s iodine test, colposcopy 168 Screening, see Biomarkers, cervical cancer; Diagnostics, HPV Smoking, cervical cancer risks 210 SurePath, cytology sample collection 149 Survivin, cervical cancer biomarker studies 85 T cells HPV infection and cell-mediated immunity 36–40 types and host defense 35 Telomerase, cervical cancer biomarker studies 75, 76 ThinPrep Imager, computer-assisted screening 155 ThinPrep Processors, cytology sample collection 149 TSLC1 gene silencing in cervical neoplasia 77, 78 Vaccination, HPV cancer prevention 200 challenges 257 clinical trials endpoint assessment 228–231
Subject Index
GlaxoSmithKline Cervarix® 192, 193 Merck Gardasil® 193, 194 overview 189–191 cost-effectiveness cervical cancer prevention strategy modeling 237, 238 economic analysis of health interventions 235, 236 overview 195, 235 prospects for study 240 studies 239, 240 duration of protection 197 education in implementation 266–268 historical perspective 217, 218 impact potential duration of protection 246, 247, 249 HPV types in modeling 249, 250 overview 194, 195, 201 screening program combination 247–249 study design 243–245, 248 timing with respect to sexual debut 245, 246, 249 women vs. total population vaccination 245, 248 misconceptions 48 principles 189, 191 promotion and education 198 prospects for study 201 rationale geographic distribution of HPV and cervical cancer 212–214 HPV DNA in cervical cancer specimens 206 risk estimates 206–208 screening impact 198–200 screening program combination 247–249, 258–260 second-generation vaccines 214, 215 target population acceptability of vaccine 198, 260 adult women 200, 201, 223 age groups 195, 196, 223, 256 boys vs. girls 196, 197, 223, 257 immunocompromised patients 197
274
vaccine requirements 191, 192, 227 virus-like particle vaccines clinical efficacy trials 219, 220, 243, 256 effectiveness expectations 224 formulations 218, 219 L1 protein vaccine rationale and efficacy 231–233 prospects for study 220–222
Subject Index
target populations 222, 223 virus types in vaccines 191 Vascular patterns, colposcopy color tone 171 epithelial borders 171 intercapillary distance 170 surface contour 171 Warts, see Condylomata acuminata
275