Risk stratification by p16 immunostaining of CIN1 biopsies: a retrospective study of patients from the quadrivalent HPV vaccine trials.

Previous studies of p16 immunohistochemistry (IHC) on CIN1 have suggested the likely utility of p16 in stratification of women at risk for subsequent CIN2/3. But those studies had limitations in statistical power, histologic diagnosis, and disease ascertainment. We conducted a retrospective study of p16 IHC on adjudicated CIN1 tissue diagnosed in young women participating in the placebo arm of the quadrivalent human papillomavirus (HPV) vaccine trials. Tissue sections were stained with p16 IHC and hematoxylin and eosin. p16 IHC was scored using LAST criteria, and hematoxylin and eosin-stained sections were reviewed for concordance with the adjudicated diagnosis. p16 IHC, antecedent high-grade cytology, review diagnosis, and HPV16 detection were assessed as independent risk factors for subsequent CIN2/3. Five hundred twenty-four patients with CIN1 biopsies and complete data were identified, 63 (12.0%) of whom developed CIN2/3 in follow-up. p16 positivity (P=0.06), review diagnosis of CIN2/3 (P=0.04), HPV16 positivity (P=0.01), and antecedent high-grade cytology (P=0.02) were (marginally) associated with CIN2/3. In a logistic regression model, the associations with CIN2/3 (vs. other), expressed as odds ratios (95% confidence intervals), were 1.6 (0.91-2.8) for p16, 2.0 (1.0-3.7) for HPV16, and 2.2 (1.1-2.4) for antecedent high-grade cytology. The mean risks for CIN2/3 estimated from the model ranged from 7.6% for negative for all markers to 36.3% for positive for all 3 markers. p16 IHC does not risk stratify CIN1 patients in a manner that would alter recommended management for CIN1. This reinforces the LAST recommendations that p16 should only be used selectively for problematic scenarios, such as CIN2 because of its inherent lack of reproducibility, cases in which one is struggling between CIN1 and CIN2, and benign mimics of CIN3.