New concepts on risk factors of HPV and novel screening strategies for cervical cancer precursors.

During the past several years, this author has been engaged in coordinating two major multicentre trials testing optional screening tools for cervical cancer (CC) in low-resource settings both in East Europe and in Latin America. These international trials include the NIS (New Independent States of the former Soviet Union) cohort (n = 3,187 women) and the LAMS (Latin American Screening) study (n = 12,114 women). In both studies, a sizeable cohort of women (887 and 1,011, respectively) have been prospectively followed-up to assess the natural history of high-risk human papillomavirus (HR-HPV) infections and the role of implicated risk factors as potential predictors of disease outcome (acquisition, persistence and clearance). In this communication some of the key observations recently reported from the NIS and LAMS studies will be discussed, with special emphasis on i) risk factors that are still controversial (i.e., oral contraception; OC, and smoking) or not previously studied (drug addiction), on ii) reproductive factors as potential cofactors of HPV infections in cervical carcinogenesis (i.e., age at menarche, menopause), and finally on iii) the performance of different screening strategies among young and older women. Although closely related to these topics, a detailed discussion on the dynamics of HPV infections (acquisition, persistence, clearance) and their predictive factors falls outside the scope of this communication, because they have been extensively discussed in a series of original reports and in a recent review of the author in this journal. The NIS cohort failed to establish OC as a risk factor of CC. In all future studies, the strong confounding effects from the lifestyle behavioural factors must be taken into account, while interpreting the data on OCs as potential risk factors of CC. Similarly, it now seems that the increased risk (if any) of CC among smokers seems to be attributed to the increased acquisition of HR-HPV infections, of which the smoking status is an independent predictor in a multivariate model. The same seems to apply to drug addiction as a risk factor of CC as well. The recent LAMS data show that drug abuse itself is not a risk factor of i) contracting HR-HPV infection or ii) developing high-grade CIN. Instead, drug abuse seems to be closely associated with several of the indicators of risky sexual behaviour, which predisposes the women to oncogenic HPV infections and thus indirectly contributes to the development of CIN2+ lesions. Data from the NIS cohort clearly implicate that menarche age is not associated with increased risk of HR-HPV infection, or development of high-grade CIN, feasibly explained by the fact that menarche age does not have any effect on the outcome of CIN lesions or HR-HPV infections in a longitudinal setting. Another special group are postmenopausal women, recently shown to have a second peak of HR-HPV prevalence in many populations. The NIS cohort data suggest that among women who fail to eradicate their HR-HPV infection by menopause, there is i) a transition from multiple infections to single-type infections, and ii) selection of an integrated viral clone has already taken place, driving the process towards an aggressively progressing cervical disease. Finally, these special features of HR-HPV infections among younger and older women lead us to consider, whether different screening strategies are needed for younger and older women. Consonant with other recent reports, data from the LAMS study show that conventional Pap and HC2, but not LBC and VIA, perform significantly differently among younger and older women. However, the choice of an optimal screening test for young and older women depends on whether the highest positive predictive value (PPV) (Pap test) or the best balance between sensitivity and specificity (SE/SP) (HC2) is used as the selection criteria. Both the NIS cohort and LAMS study have significantly contributed to solving several of the open issues in the natural history of HR-HPV infections, including their risk factors, covariates necessary in cervical carcinogenesis as well as in sorting out the optional screening strategies in low-resource settings and for women in different age groups. In the long run, it is most likely that the cost-effectiveness will be the decisive factor for which screening tests will be selected. Needless to reiterate that screening for cervical cancer precursors will be mandatory until the foreseeable future, even in this emerging era of prophylactic HPV vaccination.