Our long-term objective is to bring about prevention of cervix cancer through improved biologic understanding and more cost-effective screening strategies. Although human papilloma virus (HPV) infection is an established cause of cervical cancer, it is incompletely known if viral load of HPV influences progression from cancer in situ (CIS) to invasive cancer and/or interacts with genetic factors. Since clinical intervention precludes direct observation of this progression. unconventional approaches are needed. Our main specific aims are to ; 1) quantify the absolute and relative risks for CIS and invasive cancer as a function of time since detected HPV and HPV 16 high viral load, 2) assess whether persistent HPV 16 high viral load is a determinant for development of CIS and invasive cancer, 3) assess whether the specific HLA DQ6/DR15 haplotype is associated with risks for CIS and invasive cancer, and if the association is mediated via a higher viral load and/or persistence of HP V. and 4) assess whether Chlamydia infection is associated with risks for CIS and invasive cancer. Building on experience from an earlier study of CIS (funded by NCI). we will take advantage of unique prerequisites in Sweden created by extensive population-based PAP smear screening documented in computerised registers. ascertainment of all incident cases of CIS and invasive cancer. and access to archival smears and tissue specimens. Using a nested design in this large study base with up to 25 years of complete follow-up, we will identify 600 women with invasive cancer, 600 women with CIS and 600 individually matched control women to each case-group. Using validated and sensitive PCR assays, the presence of viral DNA - and for HPV 16, also the viral load -will be analyzed in all available smears from each participant (on average four per individual, giving a total of about 9600 smears). HLA and C trachomatis will be analyzed in the first smear from all included women. Relative risks and interactions will be estimated by conditional logistic regression and absolute risk functions by non-parametric methods.
