Joint effects of infections with different types of human papillomavirus on risk of cervical neoplasia
Institution: | University of Helsinki |
---|---|
Department: | Hjelt Institute, Department of Public Health; Finnish Cancer Registry, Institute for Statistical and Epidemiological Cancer Research |
Year: | 2015 |
Keywords: | epidemiologia |
Record ID: | 1141138 |
Full text PDF: | http://hdl.handle.net/10138/154088 |
Cervical cancer is the 4th most common cancer in women. In its carcinogenesis human papillomavirus (HPV) 16/18 are most important. HPV6/11 cause benign lesions. A small proportion of HPV infection(s) develop into cancer. Therefore, joint effects between HPVs and putative co-factors, Chlamydia trachomatis and smoking, are of interest but largely open. The aims of this work were to understand joint effects of infections with 1) HPV types, 2) HPVs and C. trachomatis, and 3) order of these infections in the carcinogenesis. For the 1st two aims, two case-control studies were nested within cohorts of Nordic biobanks. 1st linkage to cancer registers identified 182 cases of invasive cervical cancer (ICC, 148 squamous cell carcinomas, SCC) with prediagnostic sera until 1994. 2nd linkage with a longer follow-up until 2002 comprised 604 new ICC cases. Incidence density sampled controls were individually matched for age at serum sampling, sample storage time and region. For the 3rd aim, a case-control study in a serial setting was nested within a cohort of Swedish women participating in a cervical cancer screening programme in 1969-1995, and 118 ICCs with age and sampling-time-matched controls were identified. Finally, a case-cohort study in the Finnish Maternity Cohort was based on women with two pregnancies within 5 years. The women were followed on average for 4.8 years, from the 2nd pregnancy sample until the end of 2004. During follow-up, 490 women were diagnosed with cervical high-grade precancer. A comparison subcohort of 2796 women was randomly sampled from age and calendar time strata. IgG antibodies to HPV 6/11/16/18/31/33/45 capsids, and C. trachomatis were determined by ELISA. Serum cotinine, a marker for recent smoking, was measured by immunoassays. HPV and C. trachomatis DNA in smears and biopsy specimen were examined by PCR. HPV DNA-positive specimens were typed. Rate ratios were estimated by conditional logistic or proportional hazards regression. Misclassification of HPV serology was corrected for. In the 1st study, we found no excess risk of cervical carcinoma among women seropositive for both HPV16 and HPV6/11. In the 2nd study, there was excess risk, but the joint effect was again significantly smaller than the expected joint effects. Finally, if infection with HPV6 preceded infection with high-risk HPV31, there was no material excess risk of in situ cervical carcinoma. The smaller than expected joint effect between HPV types was probably due to a cell-mediated immune response to past, natural HPV6/11 infection, of which the serum antibodies were a surrogate. The risk of ICC was highly increased not only among women whose 1st smear was HPV DNA-positive but also among C. trachomatis DNA-positive women. The risk was even higher among HPV or C. trachomatis DNA positives both at the start and end of follow-up. The risk of in situ cervical carcinoma was highly increased among women whose HPV18/45 and C. trachomatis infections were virtually concomitant. The risk of SCC was increased related to C.…