Accelerated regression rates of cervical high-risk human papillomavirus infections by punch biopsies of the uterine cervix

Introduction: The natural course of HPV-infections of the uterine cervix is characterized by a high rate of clearance in the first year and lower regression rates in subsequent follow-up years, while persistent infections may lead to CIN3+. Observational studies often rely on punch biopsies to assess the infection status, ignoring potential effects of this procedure itself on the natural course of HPV-infection. Patients/methods: To assess the effect of biopsies on the rate and time to HPV-regression, we analyzed follow-up periods among 30 – 70 years old women with positive HPV-test and normal cytology from a primary HPV screening program in Germany. Per protocol these women were followed for one year and referred to colposcopy in case of HPV persistency. When biopsies were taken, women were moved to the biopsy cohort with subsequent follow-up periods. We used an interval censored accelerated failure time model assuming exponentially distributed time to HPV-regression. Results: Among 1079 follow-up periods in the baseline non-interventional cohort, we observed spontaneous HPV-clearance in 499 cases while 475 women were transferred for colposcopy with biopsies because of HPV-persistency. 201 out of 475 were not treated and had at least one HPV-test after biopsy and were included in the biopsy cohort. Of those 68.7% experienced an HPV-regression. In the biopsy cohort, time to HPV-regression was about halved compared to the baseline cohort. This result was robust in sensitivity analyses. Conclusions: Taking biopsies substantially accelerates HPV-regression in women with HPV-infection. This has implications for the interpretation of studies on the clinical course of HPV-infection.