A POPULATION-BASED SEROEPIDEMIOLOGICAL STUDY OF CERVICAL-CANCER

The epidemiology of cervical cancer indicates the presence of a sexual ly transmitted risk factor, attributable at least in part to infection with human papillomavirus (HPV) type 16 or 18. We performed a seroepi demiological study of HPV and cervical cancer in the counties of Vaste rbotten and Norrbotten in Northern Sweden, a low-risk area for cervica l cancer. Sera from 94 cases of incident cervical cancer were matched against 188 age- and sex-matched controls derived from a population-ba sed blood bank. IgG and IgA antibodies were measured against a panel o f 12 antigens derived from HPV types 6, 11, 16, or 18, as well as agai nst Herpes simplex virus type 1 and 2, Chlamydia trachomatis, cytomega lovirus, Epstein-Barr virus, and bovine papillomavirus. Significantly increased relative risks (RRs) were found for IgG to HPV 16- or 18-der ived antigens from the L1 (RR = 3.1), E2 (RRs = 2.8 and 9.2), and E7 ( RRs = 3.8 and 2.7) open reading frames and for IgA to HPV 16-derived a ntigens from the E2 (RR = 3.3) and E6 (RR = 2.7) open reading frames. The highest RR (9.2, confidence intervals 4.4-19.4) was associated wit h IgG to an HPV 18 E2 antigen. Antibodies against cytomegalovirus, Her pes simplex virus type 2, Epstein-Barr virus, or bovine papillomavirus were, on their own, not significantly associated with cervical cancer , but seropositivity against multiple infections was associated with a successively increased relative risk. An increased risk was also foun d for IgG to Chlamydia trachomatis (RR = 1.7, confidence interval = 1. 0-2.7). The results indicate that several HPV antibodies are strongly associated with cervical cancer, providing further seroepidemiological support for an etiological role of HPV in cervical cancer.