Objective: To investigate human papillomavirus (HPV) prevalence in the high incidence areas of cervical cancer in Shanxi and to study the association between HPV infection and cervical cancer.
Methods: Cells exfoliated from cervix (collected by clinician) and from vagina (collected by subject herself) of 1997 women aged 35-45 from Xiangyuan county, Shanxi province were analyzed blindly by Hybrid Capture Assay (HC-II), which could detect 13 HPV types of high-risk. Unconditional logistics regression model was used to analyze the relation between HPV and cervical cancer/cervical intraepithelial neoplasia (CIN). Kappa coefficient was used to compare the agreement of the two tests.
Results: HPV DNA detection rate in the population was 20.8% while HPV infection rates increased with the seriousness of cervical lesions (chi 2 = 444.04, P = 0.000). Rates of the two groups (35-39 and 40-45) had no significant difference (20.9%: 20.6%, chi 2 = 0.03, P = 0.86). Unconditional logistic regression analysis showed, when comparing with the normal subjects, the risk odds ratio of HPV infection with cervical cancer/high grade CIN and low grade CIN were 254.2 and 26.4 respectively, with attributable risk proportions (ARP) 98.1% and 83.6% respectively. The sensitivity of self-collected vaginal swabs assayed for HPV DNA was lower than that of clinician-collected cervical samples assayed for HPV DNA (84%: 98%, chi 2 = 5.92, P = 0.015). No significant difference in specificity was seen between them (86%: 85%, chi 2 = 0.00, P = 0.997) and there was fair agreement between the two tests (kappa = 0.74).
Conclusion: High-risk HPV infection in female genital tract was the major risk factor of cervical cancer and CIN in the areas. Prevention and control of cervical cancer should be focused on avoiding HPV infection, screening women for HPV infection and monitoring population infected by high-risk HPV. Self-collected vaginal swabs assayed for HPV DNA seemed a more practical method in primary screening for cervical cancer.