Intended for healthcare professionals


Prospective seroepidemiological study of role of human papillomavirus in non-cervical anogenital cancers

BMJ 1997; 315 doi: (Published 13 September 1997) Cite this as: BMJ 1997;315:646
  1. Tone Bj⊘rge, research fellow (tb{at},
  2. Joakim Dillner, principal investigatorb,
  3. Tarja Anttila, researcherc,
  4. Anders Engeland, statisticiana,
  5. Timo Hakulinen, professord,
  6. Egil Jellum, professore,
  7. Matti Lehtinen, senior research fellowf,
  8. Tapio Luostarinen, statisticiang,
  9. Jorma Paavonen, associate professorh,
  10. Eero Pukkala, researcheri,
  11. Martin Sapp, assistant professorj,
  12. John Schiller, senior investigatork,
  13. Linda Youngman, senior research fellowl,
  14. Steinar Thoresen, project manager, Norwegian mass screening programmesa
  1. a Cancer Registry of Norway, Institute for Epidemiological Cancer Research, Montebello, N-0310 Oslo, Norway
  2. b Microbiology and Tumourbiology Centre, Karolinska Institute, S-17177 Stockholm, Sweden
  3. c National Public Health Institute, FIN-90101 Oulu, Finland
  4. d Department of Cancer Epidemiology, Karolinska Institute, S-17177 Stockholm
  5. e Janus Committee, Norwegian Cancer Society, N-0369 Oslo
  6. f Department of Infectious Disease Epidemiology, National Public Health Institute, FIN-00300 Helsinki, Finland
  7. g Finnish Cancer Registry, Institute for Statistical and Epidemiological Cancer Research, FIN-00170 Helsinki
  8. h Department of Obstetrics and Gynecology, Helsinki University Hospital, FIN-00290 Helsinki
  9. i Tampere School of Public Health, University of Tampere, FIN-33101 Tampere, Finland
  10. j Department of Medical Microbiology, University of Mainz, 55101 Mainz, Germany
  11. k Laboratory of Cellular Oncology, National Cancer Institute, Bethesda, MA 20892 USA
  12. l Clinical Trials Service Unit, University of Oxford, Oxford OX2 6HE
  1. Correspondence to: Dr Bj⊘rge
  • Accepted 14 March 1997


Objective: To evaluate the association between infection with the major oncogenic types of human papillomavirus and the risk of developing non-cervical anogenital cancers in a cohort followed up prospectively.

Design: Data from two large serum banks to which about 700 000 people had donated serum samples were followed up for a mean of 8 years. People who developed non-cervical anogenital cancers during follow up were identified by registry linkage with the nationwide cancer registries in Finland and Norway. Within this cohort a nested case-control study was conducted based on the serological diagnosis of infection with human papillomavirus types 16, 18, and 33.

Subjects: 81 cases and 240 controls matched for sex, age, and storage time of serum samples.

Main outcome measures: Odds ratios of developing non-cervical anogenital cancers in presence of IgG antibodies to specific micro-organisms.

Results: Subjects seropositive for human papillomavirus type 16 had an increased risk of developing non-cervical anogenital cancers (odds ratio 3.1 (95% confidence interval 1.4 to 6.9)). Subjects seropositive for type 33 also had an increased risk (odds ratio 2.8 (1.0 to 8.3)) but not significantly after adjustment for infection with type 16. Seropositivity for human papillomavirus type 16 was associated with an increased risk of developing vulvar and vaginal cancers (odds ratio 4.5 (1.1 to 22)) and a strongly increased risk of developing preinvasive vulvar and vaginal lesions (odds ratio ∞ (3.8 to ∞)). Seropositivity for human papillomavirus type 18 increased the risk of developing preinvasive lesions (odds ratio 12 (1.2 to 590)). High, but non-significant odds ratios for types 16 and 33 were seen for penile cancers.

Conclusions: This study provides prospective seroepidemiological evidence that infection with human papillomavirus type 16 confers an increased risk of developing non-cervical genital cancers, particularly vulvar and vaginal cancers.

Key messages

  • Human papillomavirus has emerged as a leading infectious cause of human cancer, notably cervical and other anogenital cancers, but prospective epidemiological evidence of causality is lacking

  • This study used six million person years of follow up to investigate the relation between seropositivity for human papillomavirus and the development of non-cervical anogenital cancers

  • Infection with human papillomavirus type 16 increases the risk of developing non-cervical genital cancers, particularly vulvar and vaginal cancers

  • Infection with human papillomavirus type 16 should be considered in future intervention strategies for cervical and other genital cancers


    View Full Text