Inr. 1.Cancer: 52, 171-173 (1992) 8 1992 Wiley-Liss, Inc.
Publicationof the International Union Against Cancer Publicationde I Union Internationale Contre le Cancer
SECOND 1NTERN.ATIONAL WORKSHOP ON THE EPIDEMIOLOGY OF CERVICAL CANCER AND HUMAN PAPILLOMAVIRUS Meeting held in Bnissels, Belgium, November 25- 28, 1991 F.X. BOSCH,N. MuSOz, K.V. SHAHand A. MEHEUS In November 1991 a second international workshop was convened by the IARC and the W H O Programme of Sexually Transmitted Diseases, to review the most recent epidemiological evidence on the association between cervical cancer and human papillomavirus (HPV) infections. The first workshop, held in Copenhagen in 1988, concluded that well-designed epidemiological studies were lacking, and that the validity of the hybridization methods used to assess HPV exposure was crucial to the assessment of the association (Armstrong et al., 1988; Munoz et al., 1989). The second workshop focused mainly on developments in these areas between 1988 and 1991. The participants recognized that substantial progress had been made on both the epidemiological and the virological fronts, and that the methodological concerns expressed at the 1988 workshop had been largely resolved, despite the short time that had elapsed. Five sessions dealt with (i) the epidemiology of cervical cancer with respect to factors other than HPV and the epidemiology of HPV infection; (ii) the diagnostic procedures for detection of HPV and of other sexually transmitted agents; (iii) hybridization and serological methods for HPV detection; (iv) validation of methods for HPV detection and potential use of other markers in epidemiological studies; and (v) review of epidemiological studies of the association between HPV infection and cervical neoplasia (as well as cancers at other anogenital sites) and implications of these results for screening policy. The papers presented at the meeting and revised following discussion will be published as IARC Scientific Publication 119 later in 1992. Highlights are briefly described below. The best epideniiological evidence linking cervical cancer with HPV has come from case-control and cohort studies (reviewed by D r N. Muiioz and F.X. Bosch). An association with HPV 16, 18 and a few other viral types has been consistently reported by all case-control studies so far conducted. The association is very strong. Relative risk estimates tend to increase as more reliable HPV detection methods are used. In the large IARC case-control study in Colombia and Spain on invasive cervical cancer, HPV detection was based on PCR methods, a comprehensive questionnaire was used, and biological measurements were made for several other sexually transmitted agents. The study found a very strong association between HPV and cervical cancer, and there was also an indication of an interaction between HPV and use of oral contraceptives. Cohort studies on HPV and cervical cancer have not been completed. However, some follow-up studies which used CIN lesions as the end-point found that detection of HPV, particularly of high-risk types, was a good predictor of CIN. Retrospective cohort studies are in progress within the context of screening programmes. Ecological studies on cervical cancer and HPV (reviewed by Dr. S.K. Kjax) have revealed some gaps in the coherence of the etiological evidence. However, this type of study design may not be the best way to investigate the association of cervical cancer with HPV, because at present it is possible to measure only a fraction of total HPV exposure (current presence of HPV DNA) and not cumulative exposure over a lifetime. Moreover, most ecological studies that have been performed used hybridization methods of limited validity. A review of epidemiological studies (by Dr. L.A. Brinton) aimed at identifying non-HPV risk factors for cervical cancer supports the notion of increased susceptibility of the cervix to carcinogenesis at young ages and of possible interaction between HPV and HSV-2 infections. Recent estimates of risk for factors other than HPV tend to show that the lower the validity of the hybridization assay, the higher the role attributed to other factors, notably to surrogates of HPV infection such as the number of sexual partners. Many basic aspects of the epidemiology of HPV infections (reviewed by Dr. A. Schneider and Dr. L. Koutsky) remain poorly understood. The few studies that have used PCR methods for HPV detection have revealed that cervical cancer and HPV infections share key risk factors, notably an increased number of sexual partners. Several HPV-prevalence surveys have shown that the age distribution for HPV D N A detected in the genital tract follows the pattern of other sexually transmitted agents, with peak prevalence rates at around 20 :years of age, ranging from 20% to 50%, and a stabilization of the rates at levels below 10% for age groups above 40 years. This pattern is independent of the hybridization method used and is similar for high- and low-risk types of HPV. The hybridization methods for HPV-DNA detection were reviewed by Dr. E.M. d e Villiers; specific discussion on PCR-based methods was led by Dr. M. Manos, and on non-PCR-based methods by Dr. A. Lorincz. Validation studies for hybridization methods were reviewed by Dr. M. Schiffman and the impact of misclassification in estimating the association of HPV with cervical cancer was reviewed by Dr. E. Franco. It was agreed that PCR-based diagnosis offers advantages for epidemiological studies, largely due to the ability of PCR to detect small amounts of viral DNA. PCR-based assays have been used to identify specific types and offer the possibility of amplifying new types. It has been shown that large numbers of PCR-based assays can be done without sample contamination. It was concluded that the different PCR methods currently in use need to be compared. No proper validation studies have been performed for many of the steps of the PCR method (e.g., comparison of amplification by type-specific and consensus primers; accuracy of specific diagnosis). It was concluded that the use of PCR technology would be much more effective if standardized reagents were available and if one or more reference centres were established to assist investigators in monitoring test performance in their laboratories. Hybridization tests not based on amplification have also proved very useful, and ideally specimens should be tcsted by 2 assays, one PCR-based and the other not based on amplification. If a single assay is used, the investigator should use one that has been validated in his or her own laboratory. The filter in situ hybridization test was considered to be inaccurate and therefore unsuitable for use in epidemiological studies. It is expected that new assays which test
172
BOSCH ETAL.
for a larger number of HPV types and which can be performed in the clinical setting or in the field will become available in the near future. Results from serological assays for the diagnosis of HSV-1, HSV-2, Trepunema pallidum, cytomegalovirus, human immunodeficiency virus and Chlamydia (reviewed by Dr. C.J. Lacey), were discussed in conjunction with the available evidence on the association between HPV and cervical cancer. The evidence for a role of other sexually transmitted infections on cervical cancer was found to be inconsistent. A n interaction between HPV and HSV-2 has been suggested by one study, but was not confirmed by the IARC study. Dr. F. Judson reviewed the interaction of HPV and HIV. There is no clear evidence of an impact of HPV on the clinical course or infectiousness of HIV infection. On the other hand, HIV infection increases occurrence, severity and duration of HPV-related anogenital lesions, including intra-epithelial neoplasia. Future studies on the epidemiology, natural history and prevention of cervical cancer and/or HPV infection must control for HIV infection status in areas where HIV is common. Serological assays for HPV that are under development (reviewed by Dr. D. Galloway) make use of peptides or proteins derived from the L1, L2, E6, E7, E4 and E 2 regions of HPV 16 and HPV 6. Rigorous inter- and intra-laboratory comparisons need to be performed, as well as tests for specificity and sensitivity. Several groups have shown an association between the presence of antibodies to HPV 16 E7 and cervical cancer. Although it may be premature, serological assays are beginning to be used in large epidemiological studies. Different classification schemes for pre-neoplastic cervical lesions were reviewed by Dr. N. Kiviat, who suggested that low-grade (LG) and high-grade (HG) squamous intra-epithelial lesions (SIL) represent 2 distinct entities rather than parts of a continuum. HPV types not associated with neoplasia appear to be more common among LGSIL, which may be prone to regress spontaneously. In contrast, HGSIL may be able to appear de nuvu on normal epithelium and carry a high risk for progression to more advanced neoplastic stages. Among possible markers for HPV studies and for studies on cervical cancer (reviewed by Dr. K.V. Shah) are the use of antibodies to HPV proteins and to herpes simplex type 2, detection of mutations in oncogenes and tumor-suppressor genes, and measurements of micronutrients and HLA antigens. Recent studies have suggested that the presence of antibodies to HPV 16 E7 protein is related to disease-stage. Evidence on the role of HPV in other anogenital cancers (penis, vulva, vagina and anal region) is limited and to some extent indirect (time-trend analysis, clinical series and case reports). However, the same HPV types identified in cervical cancer are often found in cancers at other sites (reviewed by Dr. J. Daling and Dr. K. Sherman). HPV D N A is found in the external genitalia or the urethra of about 50% of the male partners of women with HPV lesions or HPV-related neoplasia (reviewed by Dr. R. Barrasso). Evidence was presented suggesting that a proportion of the anogenital cases may not b e linked to HPV. Cancers unrelated to HPV may be more common among older women and carry a poorer prognosis. The possible impact of HPV testing in screening programmes was evaluated. Women with latent HPV infections are not identified by current cytology-based screening programs. The most common HPV types identified in such infections were shown to be due high-risk types (reviewed by Dr. S. de Sanjose). Other and unknown HPV types, which are also frequent in latent infections, have recently been linked to cervical cancer. Current cervical-cancer screening programmes are of proven efficacy and are based on cytology. Dr. V. Beral proposed that before HPV testing was incorporated into any such screening programme, randomized controlled trials in women with borderline cytology (cytology that is not normal but not indicative of dyskaryosis or CIN) should be conducted, in order to discover whether detection of pre-cancerous lesions was better than by the use of cytology alone. Dr. C.J.L.M. Meijer presented results of on-going studies in The Netherlands that suggest that PCR-based HPV testing in women over 35 years of age could be the first screening step, followed by conventional cytology. Patient management would depend on the combined findings of cytology and HPV typing. In conclusion, the association between HPV infection and cervical cancer was accepted as proven beyond any reasonable doubt and, given all the data available. should be considered as causal. The principal arguments for causality were as follows: (i) the prevalence of HPV DNA in cervical cancer specimens and high-grade pre-neoplastic lesions, detected by PCR methodology and/or high-quality Southern hybridization methods, has been consistently reported to be in the range 85-100%; (ii) the association between HPV DNA detection and cervical cancer is strong and consistent in 5 case-control studies; (iii) the association is specific to a limited number of the HPV types commonly present in the genital tract; (iv) a dose-response relationship has been reported between estimates of the viral load and the risk of cervical cancer. Although adequate follow-up studies have not been completed, the risk for progression to more advanced CIN lesions has been reported as higher among HPV-16il8-positive women than among women with other types of HPV or women without any viral DNA. It was remarked by Dr. H. zur Hausen that the criteria of causality in infectious disease contained in Koch’s postulates might need to be revised in the light of modern advances in molecular and experimental biology. Areas of interest for future studies were identified as: (i) studies on the epidemiology of HPV infection, including aspects of transmission, risk factors for viral persistence, and studies on the prevalence of HPV markers in population groups such as cancer patients, HIV-infected individuals, pregnant women and prostitutes; (ii) identification of factors that may interact with HPV in the establishment of chronic infections and the progression of HPV infections to cervical neoplasia and cancer. Such factors may include hormonal, immunological and dietary factors as well as smoking, other sexually transmitted agents and H L A haplotypes; (iii) validation studies on the different PCR methods available, using material collected in the context of well planned epidemiological studies. Reference centres might need to be established for methodological quality control; (iv) trials to determine the utility of introducing HPV DNA testing into screening programmes for cervical cancer and in the clinical management of patients with HPV infections.
ACKNOWLEDGEMENTS
The workshop was supported by the CEC “Europe Against Cancer Programme”, the Ministry of Public Health and Environment of Belgium, the Danish Cancer Registry and the Danish Cancer Society.
EPIDEMIOLOGY OF CERVICAL CANCER AND HPV
173
REFERENCES 0..Mufioz, N. and BOSCH,F.X., ARMSTRONG, B.. M ~ L L EJENSEN, R Human papillomavirus and cervical cancer. Lancet, 1, 756-758 (1988).
Mufioz, N., BOSCH,F.X. and JENSEN, O.M. (eds.), Hitmanpupillomavims and cervical cancer, IARC Scientific Publication 94, IARC, Lyon (1989).
LIST OF PARTICIPANTS ARMSTRONG, B.K., International Agency for Research on Cancer, F-69372 Lyon Cedex 08, France. BARRASSO. R., Pasteur Hospital, F-75015 Paris, France. BERAL,V., Cancer Epidemiology Unit. Imperial Cancer Research Fund, GB-OX2 6HE Oxford, UK. BOSCH.F.X., Unit of Field and Intervention Studies, International Agency for Research on Cancer, F-69372 Lyon Cedex 08, France. BRINTON, L.A., Environmental Studies Section, Environmental Epidemiology Branch, National Cancer Institute, Bethesda. MD 20892, USA. CASTELLSAGUE, X., Unit of Field and Intervention Studies, International Agency for Research on Cancer, F-69372 Lyon Cedex 08, France. CHENEY, J.. Editorial and Publications Service, International Agency for Research on Cancer, F-69372 Lyon Cedex 08, France. DALING,J.R., Division of Public Health Sciences, WeissiDaling Studies (MP 381), Fred Hutchinson Cancer Research Center, Seattle, WA 98 104, USA. DE VILLIERS, E.M.. German Cancer Research Center, D-6900 Heidelberg, Germany. FRANCO, E.L.F., Epidemiology and Preventive Medicine Research Centre, lnstitut Armand-Frappier, Laval, Qutbec H7N 423, Canada. FRISCH, M.. Danish Cancer Registry, DK-2100 Copenhagen 0, Denmark. GALLOWAY. D.A., University of Washington, Fred Hutchinson Cancer Research Center, Seattle, WA 98104-2092. USA. GUERRERO. E., Abbott Cientifica s.a., E-28034 Madrid, Spain. HIGGINS, G., Department of Pathology, GB-Cambridge CB2 lQP, UK. HORI)ING, U., Rigshospitalet, DK-2100 Copenhagen, Denmark. JHA,P., Cancer Epidemiology Unit, Imperial Cancer Research Fund, GB-Oxford OX2 hHE, UK. JUDSON, F., Director, Denver Public Health, Denver, CO 80204-4507, USA. KIVIAT, N., Harborview Medical Center, Seattle, WA 98104, USA. Kouisku. L., Center for AIDS and ST'Ds, Seattle, WA 98122, USA. KJA.R, S.K., Danish Cancer Registry, DK-2100 Copenhagen 0, Denmark.
LACEY,C.J.N.. Department of Genito-urinary Medicine, General Infirmary, GB-Leeds LS13EX, UK. LORINCZ, L., Digene Diagnostics. Inc., Silver Spring, MD 20901, USA. MANOS,M., Department of Infectious Diseases, Cetus Corporation, Emeryville, CA 94608, USA. MEHEUS, A,, Sexually Transmitted Diseases. World Health Organization, CH-1211 Geneva 27, Switzerland. MEIJER.C.J.L.M., Department of Pathology, Free University Hospital, NL-1081 HV Amsterdam, The Netherlands. MELBYE,M., Department of Epidemiology, Statens Serum Institut. DK-2300 Copenhagen S, Denmark. Mufioz, N., Unit of Field and Intervention Studies, International Agency for Research on Cancer, F-69372 Lyon Cedex 08, France. PETO,J., Institute of Cancer Research, Clifton Avenue, GB-Sutton Surrey SM2 5PX, UK. ROHAN, T., NCIC Epidemiology Unit, Faculty of Medicine, University of Toronto, Toronto, Ontario M5S lA8, Canada. SANTAMARIA, M., Laboratorio Anatomia Patologica, Hospital Provincial de Navarra, E-31008 Pamplona, Spain. SCHIFFMAN, M., National Institutes of Health, National Cancer Institute, Bethesda, MD 20892, USA. SCHNEIDER, A,, Department of Obstetrics and Gynecology, University of Arizona, College of Medicine, Tucson, AZ 85724, USA. SHAH,K.V., Department of Immunology and Infectious Diseases, The Johns Hopkins University, School of Hygiene and Public Health, Baltimore, MD 21205. USA. SHERMAN, K., Division of Public Health Sciences. WeissiDaling Studies (MP 381), Fred Hutchinson Cancer Research Center, Seattle, WA 98104, USA. SINGER, A., Consultant Gynaecologist, Wittington and Royal Northern Hospitals, GB-London N19 5NF, UK. THIRY.L., Laboratoire de Microbiologie, Faculte de Medecine. Universite Libre de Bruxelles, B-1000 Bruxelles, Belgium. VAN DEN BRULE,A.J.C., Department of Pathology, Free University Hospital, NL-1081 HV Amsterdam, The Netherlands. ZURHAUSEN, H., German Cancer Research Center, D-6900 Heidelberg 1, Germany.
Publicationof the International Union Against Cancer Publicationde I Union Internationale Contre le Cancer
SECOND 1NTERN.ATIONAL WORKSHOP ON THE EPIDEMIOLOGY OF CERVICAL CANCER AND HUMAN PAPILLOMAVIRUS Meeting held in Bnissels, Belgium, November 25- 28, 1991 F.X. BOSCH,N. MuSOz, K.V. SHAHand A. MEHEUS In November 1991 a second international workshop was convened by the IARC and the W H O Programme of Sexually Transmitted Diseases, to review the most recent epidemiological evidence on the association between cervical cancer and human papillomavirus (HPV) infections. The first workshop, held in Copenhagen in 1988, concluded that well-designed epidemiological studies were lacking, and that the validity of the hybridization methods used to assess HPV exposure was crucial to the assessment of the association (Armstrong et al., 1988; Munoz et al., 1989). The second workshop focused mainly on developments in these areas between 1988 and 1991. The participants recognized that substantial progress had been made on both the epidemiological and the virological fronts, and that the methodological concerns expressed at the 1988 workshop had been largely resolved, despite the short time that had elapsed. Five sessions dealt with (i) the epidemiology of cervical cancer with respect to factors other than HPV and the epidemiology of HPV infection; (ii) the diagnostic procedures for detection of HPV and of other sexually transmitted agents; (iii) hybridization and serological methods for HPV detection; (iv) validation of methods for HPV detection and potential use of other markers in epidemiological studies; and (v) review of epidemiological studies of the association between HPV infection and cervical neoplasia (as well as cancers at other anogenital sites) and implications of these results for screening policy. The papers presented at the meeting and revised following discussion will be published as IARC Scientific Publication 119 later in 1992. Highlights are briefly described below. The best epideniiological evidence linking cervical cancer with HPV has come from case-control and cohort studies (reviewed by D r N. Muiioz and F.X. Bosch). An association with HPV 16, 18 and a few other viral types has been consistently reported by all case-control studies so far conducted. The association is very strong. Relative risk estimates tend to increase as more reliable HPV detection methods are used. In the large IARC case-control study in Colombia and Spain on invasive cervical cancer, HPV detection was based on PCR methods, a comprehensive questionnaire was used, and biological measurements were made for several other sexually transmitted agents. The study found a very strong association between HPV and cervical cancer, and there was also an indication of an interaction between HPV and use of oral contraceptives. Cohort studies on HPV and cervical cancer have not been completed. However, some follow-up studies which used CIN lesions as the end-point found that detection of HPV, particularly of high-risk types, was a good predictor of CIN. Retrospective cohort studies are in progress within the context of screening programmes. Ecological studies on cervical cancer and HPV (reviewed by Dr. S.K. Kjax) have revealed some gaps in the coherence of the etiological evidence. However, this type of study design may not be the best way to investigate the association of cervical cancer with HPV, because at present it is possible to measure only a fraction of total HPV exposure (current presence of HPV DNA) and not cumulative exposure over a lifetime. Moreover, most ecological studies that have been performed used hybridization methods of limited validity. A review of epidemiological studies (by Dr. L.A. Brinton) aimed at identifying non-HPV risk factors for cervical cancer supports the notion of increased susceptibility of the cervix to carcinogenesis at young ages and of possible interaction between HPV and HSV-2 infections. Recent estimates of risk for factors other than HPV tend to show that the lower the validity of the hybridization assay, the higher the role attributed to other factors, notably to surrogates of HPV infection such as the number of sexual partners. Many basic aspects of the epidemiology of HPV infections (reviewed by Dr. A. Schneider and Dr. L. Koutsky) remain poorly understood. The few studies that have used PCR methods for HPV detection have revealed that cervical cancer and HPV infections share key risk factors, notably an increased number of sexual partners. Several HPV-prevalence surveys have shown that the age distribution for HPV D N A detected in the genital tract follows the pattern of other sexually transmitted agents, with peak prevalence rates at around 20 :years of age, ranging from 20% to 50%, and a stabilization of the rates at levels below 10% for age groups above 40 years. This pattern is independent of the hybridization method used and is similar for high- and low-risk types of HPV. The hybridization methods for HPV-DNA detection were reviewed by Dr. E.M. d e Villiers; specific discussion on PCR-based methods was led by Dr. M. Manos, and on non-PCR-based methods by Dr. A. Lorincz. Validation studies for hybridization methods were reviewed by Dr. M. Schiffman and the impact of misclassification in estimating the association of HPV with cervical cancer was reviewed by Dr. E. Franco. It was agreed that PCR-based diagnosis offers advantages for epidemiological studies, largely due to the ability of PCR to detect small amounts of viral DNA. PCR-based assays have been used to identify specific types and offer the possibility of amplifying new types. It has been shown that large numbers of PCR-based assays can be done without sample contamination. It was concluded that the different PCR methods currently in use need to be compared. No proper validation studies have been performed for many of the steps of the PCR method (e.g., comparison of amplification by type-specific and consensus primers; accuracy of specific diagnosis). It was concluded that the use of PCR technology would be much more effective if standardized reagents were available and if one or more reference centres were established to assist investigators in monitoring test performance in their laboratories. Hybridization tests not based on amplification have also proved very useful, and ideally specimens should be tcsted by 2 assays, one PCR-based and the other not based on amplification. If a single assay is used, the investigator should use one that has been validated in his or her own laboratory. The filter in situ hybridization test was considered to be inaccurate and therefore unsuitable for use in epidemiological studies. It is expected that new assays which test
172
BOSCH ETAL.
for a larger number of HPV types and which can be performed in the clinical setting or in the field will become available in the near future. Results from serological assays for the diagnosis of HSV-1, HSV-2, Trepunema pallidum, cytomegalovirus, human immunodeficiency virus and Chlamydia (reviewed by Dr. C.J. Lacey), were discussed in conjunction with the available evidence on the association between HPV and cervical cancer. The evidence for a role of other sexually transmitted infections on cervical cancer was found to be inconsistent. A n interaction between HPV and HSV-2 has been suggested by one study, but was not confirmed by the IARC study. Dr. F. Judson reviewed the interaction of HPV and HIV. There is no clear evidence of an impact of HPV on the clinical course or infectiousness of HIV infection. On the other hand, HIV infection increases occurrence, severity and duration of HPV-related anogenital lesions, including intra-epithelial neoplasia. Future studies on the epidemiology, natural history and prevention of cervical cancer and/or HPV infection must control for HIV infection status in areas where HIV is common. Serological assays for HPV that are under development (reviewed by Dr. D. Galloway) make use of peptides or proteins derived from the L1, L2, E6, E7, E4 and E 2 regions of HPV 16 and HPV 6. Rigorous inter- and intra-laboratory comparisons need to be performed, as well as tests for specificity and sensitivity. Several groups have shown an association between the presence of antibodies to HPV 16 E7 and cervical cancer. Although it may be premature, serological assays are beginning to be used in large epidemiological studies. Different classification schemes for pre-neoplastic cervical lesions were reviewed by Dr. N. Kiviat, who suggested that low-grade (LG) and high-grade (HG) squamous intra-epithelial lesions (SIL) represent 2 distinct entities rather than parts of a continuum. HPV types not associated with neoplasia appear to be more common among LGSIL, which may be prone to regress spontaneously. In contrast, HGSIL may be able to appear de nuvu on normal epithelium and carry a high risk for progression to more advanced neoplastic stages. Among possible markers for HPV studies and for studies on cervical cancer (reviewed by Dr. K.V. Shah) are the use of antibodies to HPV proteins and to herpes simplex type 2, detection of mutations in oncogenes and tumor-suppressor genes, and measurements of micronutrients and HLA antigens. Recent studies have suggested that the presence of antibodies to HPV 16 E7 protein is related to disease-stage. Evidence on the role of HPV in other anogenital cancers (penis, vulva, vagina and anal region) is limited and to some extent indirect (time-trend analysis, clinical series and case reports). However, the same HPV types identified in cervical cancer are often found in cancers at other sites (reviewed by Dr. J. Daling and Dr. K. Sherman). HPV D N A is found in the external genitalia or the urethra of about 50% of the male partners of women with HPV lesions or HPV-related neoplasia (reviewed by Dr. R. Barrasso). Evidence was presented suggesting that a proportion of the anogenital cases may not b e linked to HPV. Cancers unrelated to HPV may be more common among older women and carry a poorer prognosis. The possible impact of HPV testing in screening programmes was evaluated. Women with latent HPV infections are not identified by current cytology-based screening programs. The most common HPV types identified in such infections were shown to be due high-risk types (reviewed by Dr. S. de Sanjose). Other and unknown HPV types, which are also frequent in latent infections, have recently been linked to cervical cancer. Current cervical-cancer screening programmes are of proven efficacy and are based on cytology. Dr. V. Beral proposed that before HPV testing was incorporated into any such screening programme, randomized controlled trials in women with borderline cytology (cytology that is not normal but not indicative of dyskaryosis or CIN) should be conducted, in order to discover whether detection of pre-cancerous lesions was better than by the use of cytology alone. Dr. C.J.L.M. Meijer presented results of on-going studies in The Netherlands that suggest that PCR-based HPV testing in women over 35 years of age could be the first screening step, followed by conventional cytology. Patient management would depend on the combined findings of cytology and HPV typing. In conclusion, the association between HPV infection and cervical cancer was accepted as proven beyond any reasonable doubt and, given all the data available. should be considered as causal. The principal arguments for causality were as follows: (i) the prevalence of HPV DNA in cervical cancer specimens and high-grade pre-neoplastic lesions, detected by PCR methodology and/or high-quality Southern hybridization methods, has been consistently reported to be in the range 85-100%; (ii) the association between HPV DNA detection and cervical cancer is strong and consistent in 5 case-control studies; (iii) the association is specific to a limited number of the HPV types commonly present in the genital tract; (iv) a dose-response relationship has been reported between estimates of the viral load and the risk of cervical cancer. Although adequate follow-up studies have not been completed, the risk for progression to more advanced CIN lesions has been reported as higher among HPV-16il8-positive women than among women with other types of HPV or women without any viral DNA. It was remarked by Dr. H. zur Hausen that the criteria of causality in infectious disease contained in Koch’s postulates might need to be revised in the light of modern advances in molecular and experimental biology. Areas of interest for future studies were identified as: (i) studies on the epidemiology of HPV infection, including aspects of transmission, risk factors for viral persistence, and studies on the prevalence of HPV markers in population groups such as cancer patients, HIV-infected individuals, pregnant women and prostitutes; (ii) identification of factors that may interact with HPV in the establishment of chronic infections and the progression of HPV infections to cervical neoplasia and cancer. Such factors may include hormonal, immunological and dietary factors as well as smoking, other sexually transmitted agents and H L A haplotypes; (iii) validation studies on the different PCR methods available, using material collected in the context of well planned epidemiological studies. Reference centres might need to be established for methodological quality control; (iv) trials to determine the utility of introducing HPV DNA testing into screening programmes for cervical cancer and in the clinical management of patients with HPV infections.
ACKNOWLEDGEMENTS
The workshop was supported by the CEC “Europe Against Cancer Programme”, the Ministry of Public Health and Environment of Belgium, the Danish Cancer Registry and the Danish Cancer Society.
EPIDEMIOLOGY OF CERVICAL CANCER AND HPV
173
REFERENCES 0..Mufioz, N. and BOSCH,F.X., ARMSTRONG, B.. M ~ L L EJENSEN, R Human papillomavirus and cervical cancer. Lancet, 1, 756-758 (1988).
Mufioz, N., BOSCH,F.X. and JENSEN, O.M. (eds.), Hitmanpupillomavims and cervical cancer, IARC Scientific Publication 94, IARC, Lyon (1989).
LIST OF PARTICIPANTS ARMSTRONG, B.K., International Agency for Research on Cancer, F-69372 Lyon Cedex 08, France. BARRASSO. R., Pasteur Hospital, F-75015 Paris, France. BERAL,V., Cancer Epidemiology Unit. Imperial Cancer Research Fund, GB-OX2 6HE Oxford, UK. BOSCH.F.X., Unit of Field and Intervention Studies, International Agency for Research on Cancer, F-69372 Lyon Cedex 08, France. BRINTON, L.A., Environmental Studies Section, Environmental Epidemiology Branch, National Cancer Institute, Bethesda. MD 20892, USA. CASTELLSAGUE, X., Unit of Field and Intervention Studies, International Agency for Research on Cancer, F-69372 Lyon Cedex 08, France. CHENEY, J.. Editorial and Publications Service, International Agency for Research on Cancer, F-69372 Lyon Cedex 08, France. DALING,J.R., Division of Public Health Sciences, WeissiDaling Studies (MP 381), Fred Hutchinson Cancer Research Center, Seattle, WA 98 104, USA. DE VILLIERS, E.M.. German Cancer Research Center, D-6900 Heidelberg, Germany. FRANCO, E.L.F., Epidemiology and Preventive Medicine Research Centre, lnstitut Armand-Frappier, Laval, Qutbec H7N 423, Canada. FRISCH, M.. Danish Cancer Registry, DK-2100 Copenhagen 0, Denmark. GALLOWAY. D.A., University of Washington, Fred Hutchinson Cancer Research Center, Seattle, WA 98104-2092. USA. GUERRERO. E., Abbott Cientifica s.a., E-28034 Madrid, Spain. HIGGINS, G., Department of Pathology, GB-Cambridge CB2 lQP, UK. HORI)ING, U., Rigshospitalet, DK-2100 Copenhagen, Denmark. JHA,P., Cancer Epidemiology Unit, Imperial Cancer Research Fund, GB-Oxford OX2 hHE, UK. JUDSON, F., Director, Denver Public Health, Denver, CO 80204-4507, USA. KIVIAT, N., Harborview Medical Center, Seattle, WA 98104, USA. Kouisku. L., Center for AIDS and ST'Ds, Seattle, WA 98122, USA. KJA.R, S.K., Danish Cancer Registry, DK-2100 Copenhagen 0, Denmark.
LACEY,C.J.N.. Department of Genito-urinary Medicine, General Infirmary, GB-Leeds LS13EX, UK. LORINCZ, L., Digene Diagnostics. Inc., Silver Spring, MD 20901, USA. MANOS,M., Department of Infectious Diseases, Cetus Corporation, Emeryville, CA 94608, USA. MEHEUS, A,, Sexually Transmitted Diseases. World Health Organization, CH-1211 Geneva 27, Switzerland. MEIJER.C.J.L.M., Department of Pathology, Free University Hospital, NL-1081 HV Amsterdam, The Netherlands. MELBYE,M., Department of Epidemiology, Statens Serum Institut. DK-2300 Copenhagen S, Denmark. Mufioz, N., Unit of Field and Intervention Studies, International Agency for Research on Cancer, F-69372 Lyon Cedex 08, France. PETO,J., Institute of Cancer Research, Clifton Avenue, GB-Sutton Surrey SM2 5PX, UK. ROHAN, T., NCIC Epidemiology Unit, Faculty of Medicine, University of Toronto, Toronto, Ontario M5S lA8, Canada. SANTAMARIA, M., Laboratorio Anatomia Patologica, Hospital Provincial de Navarra, E-31008 Pamplona, Spain. SCHIFFMAN, M., National Institutes of Health, National Cancer Institute, Bethesda, MD 20892, USA. SCHNEIDER, A,, Department of Obstetrics and Gynecology, University of Arizona, College of Medicine, Tucson, AZ 85724, USA. SHAH,K.V., Department of Immunology and Infectious Diseases, The Johns Hopkins University, School of Hygiene and Public Health, Baltimore, MD 21205. USA. SHERMAN, K., Division of Public Health Sciences. WeissiDaling Studies (MP 381), Fred Hutchinson Cancer Research Center, Seattle, WA 98104, USA. SINGER, A., Consultant Gynaecologist, Wittington and Royal Northern Hospitals, GB-London N19 5NF, UK. THIRY.L., Laboratoire de Microbiologie, Faculte de Medecine. Universite Libre de Bruxelles, B-1000 Bruxelles, Belgium. VAN DEN BRULE,A.J.C., Department of Pathology, Free University Hospital, NL-1081 HV Amsterdam, The Netherlands. ZURHAUSEN, H., German Cancer Research Center, D-6900 Heidelberg 1, Germany.
