Longitudinal methylation analysis for risk stratification of CIN2 lesions by HPV genotype

INTRODUCTION: There is an urgent need for biomarkers that can distinguish regressive from progressive cervical intraepithelial neoplasia (CIN) lesions to improve risk stratification in cervical cancer screening programs and avoid overtreatment. We aimed to evaluate the longitudinal performance of the S5 methylation classifier in predicting the clinical trajectory of untreated CIN grade 2 (CIN2) across different HPV genotype groups. MATERIAL AND METHODS: This study included 122 women (aged 18-30) with histologically confirmed CIN2 enrolled in an active surveillance trial at Helsinki University Hospital, Finland (ISRCTN91953024). Patients were monitored every 6 months for up to 24 months with cytology, HPV genotyping, and colposcopy. S5 methylation levels (pyrosequencing) were compared between women who regressed to normal histology (n = 89) and those who progressed to CIN3+ (n = 22, including one microinvasive cancer) or had persistent CIN2 (n = 11). Analysis was stratified based on baseline HPV genotypes. RESULTS: The baseline S5 classifier significantly predicted progression/persistence with an odds ratio (OR) of 1.15 (95%CI 1.05-1.25). When excluding HPV16/18, the S5 classifier demonstrated a similar, borderline predictive association (OR 1.19; 95%CI 1.00-1.41), whereas cytology and HPV genotyping did not. Longitudinal analysis showed that S5 methylation levels in progressors remained elevated throughout the 24-month period, significantly predicting progression/persistence up to 18 months prior to histological diagnosis of progression. CONCLUSIONS: In this active surveillance cohort, the S5 classifier showed potential as a prognostic triage biomarker for identifying progression in untreated CIN2, including cases with non-HPV16/18 types. Our exploratory findings suggest that elevated methylation levels may precede histological progression by up to 18 months. As HPV16/18 prevalence declines, host-viral methylation panels may offer improved risk stratification compared to standard triage, though larger prospective studies in vaccinated populations are needed to confirm these findings and establish clinical utility.