Predictive biomarkers for regression in women undergoing active surveillance for cervical intraepithelial neoplasia grade 2: A prospective multicenter study in Italy.

Cervical intraepithelial neoplasia grade 2 (CIN2) can spontaneously regress in a sizable proportion of cases. The aim of this prospective multicenter cohort study was to identify the biomarkers associated with a high probability of regression. A total of 319 women aged 25-45 years fulfilling predefined inclusion and exclusion criteria (full visibility of transformation zone and lesion; no previous history of CIN2+ or immune impairment) were enrolled and subjected to active surveillance for 24 months. HPV genotyping, p16/ki67 expression, methylation status of FAM19A4/miR124-2 genes, and cytology were evaluated at baseline. The probability of CIN2 regression according to the different biomarkers was evaluated through binomial logistic regression. At follow-up, regression, persistence, and progression (evaluated on 294 women) were recorded in 165 (56%), 68 (23%), and 61 (21%) cases, respectively; no association with age was observed. Overall, 110 women underwent excisional treatment during follow-up; 53 CIN2 and 50 CIN3+ were diagnosed. The probability of CIN2 regression significantly increased with early HPV negativity (odds ratio [OR] 6.45, 95% confidence intervals [CI] 1.68-42.6), no p16/ki67 expression (OR 2.49, 95%CI 1.38-4.52), and unmethylated status (OR 2.12, 95%CI 1.09-4.20). Our results indicate that active CIN2 surveillance could be implemented for women up to 45 years, after selection according to anatomo-clinical criteria and biomarker status. To improve feasibility, the biomarkers can be used sequentially, taking advantage of the HPV genotyping available in primary screening tests, and eventually refining the selection by using the other biomarkers in selected subgroups.