[Changes in genotype prevalence of human papillomavirus over 10-year follow-up of a cervical cancer screening cohort].

Objective: To evaluate the dynamic variation of genotypes distribution of human papillomavirus (HPV) over 10-year follow-up in a cervical cancer screening cohort. Methods: Based on the Shanxi Province Cervical Cancer Screening Study Ⅰ cohort, we detected HPV genotypes on the well-preserved exfoliated cervical cells from women who were tested HPV positive from year 2005 to year 2014 using reverse linear probe hybridization assay. The changes of prevalence of type-specific HPV over time among the overall population were estimated using linear mixed models. The association between the type-specific HPV and cervical intraepithelial neoplasia grade 2 or worse (CIN2 +) was calculated by linear Chi-square test. Finally, the trends of multiple infections of HPV with the increase of the age were analyzed. Results: During the cervical cancer screening of the overall population from 2005 to 2014, the most common genotypes among the population were HPV16 and 52. The prevalence of HPV16 decreased over time from 4.6% in 2005 to 2.2% in 2010 and 2014 (F=8.125, P<0.001). The prevalence of HPV52 remained pretty stable and HPV33, 51 and 58 slightly decreased then apparently increased. Further stratification analysis by pathological lesions showed the same trend of the HPV prevalence for the histology normal women with the overall population. Of note, for those women with the cervical intraepithelial neoplasia (CIN2 +), the detection rate of HPV16 decreased from 65.22% in 2005 to 41.03% in 2010 and finally to 31.58% in 2014 (χ(2)=4.420, P=0.036) and that of HPV33 substantially increased. No significant variation was found for other types of HPV. Multiple infection rate varied with the growing age of the women. Conclusions: The genotypes prevalence of HPV tended to vary over time during cervical cancer screening in the context of regular screening combining with immediate treatment for those CIN2 + women. HPV16 prevalence significantly decreased over time, which indicated that the variation of type-specific HPV prevalence should be considered when regular cervical cancer screening was organized using HPV technique.