Molecular epidemiology of human papillomavirus and cervical cancer

Abstract

Background: Cervical carcinoma is globally the second most common malignant disease in women accounting for approximately 10 % of all cancers in women worldwide. Human Papillomavirus (HPV) is a sexually transmitted infection (STI) known to be the main causative agent of cervical preinvasive and invasive neoplasia. Therefore, detailed, knowledge regarding the epidemiology and biological properties of HPV is important for the global public health.

Aims: We investigated: I) the determinants of spread of HPV: II) whether interactions between HPV types exist; III) whether HPV is a prognostic marker in cervical cancer and IV) quantified the risk factors for cervical cancer in Latvia. The thesis is based on 3 different studies: A case-control study from Stockholm (Sweden), where 218 cervical cancer cases and 219 healthy controls were enrolled and followed up regarding survival; A cross-sectional survey of healthy Swedish women, where a randomly chosen subsample of 274 women was evaluated; A population-based case-control study from Latvia, containing 223 cervical cancer cases and 239 healthy controls.

Results: Exposure to oncogenic HPV is strongly dependent on sexual history (OR 8,7; CI 95% 3,3-22,6), but exposure to benign HPV types was not significantly associated with number of life-time partners (LTP). All three investigated oncogenic types (16,18, 33) were associated with an excess cancer risk. The highest OR was found for HPV 16: 2,39 (CI 95 % 1,54-3,72). HPV types 6 and 16 appeared to have an antagonistic interference (Relative Excess Risk due to Interaction: -2,35; CI 95 % -0,04 to -4,65). Although some STI tended to occur together in the same women, notably HPV types 6 and 11; HPV 16,18 and 33 and C.trachomatis and Herpes simplex virus 2 (HSV-2), clustering of different STI was commonly absent. Presence of antibodies to HPV did not predict survival among cervical cancer patients (OR 0,64; CI 95 % 0,4-1,1). In Latvia, the most important risk factor for cervical cancer was presence of cervical HPV 16/18 DNA (OR 31,54; CI 95 % 16,41-60,62). A strong protective effect was also found for increasing number of Pap smears during the last 5 years (OR 0,47; CI 95 % 0,35-0,62).

Conclusion: We found marked differences in determinants of exposure to oncogenic and benign HPV and limited clustering of different STIs, arguing that prevention may need to target different risk groups. Several HPV types, in particular HPV 16 and HPV 6 appeared to have an interference in cervical carcinogenesis, a finding of possible relevance to vaccine design and evaluation. Finally, we found that prevention of HPV type 16 infection and introduction of organised Pop smear screening would be the most important cervical cancer preventive measures to introduce in Latvia.