Risk-Based Triage of Women with Persistent HR-HPV Positivity at 12-Month Recall in Cervical Cancer Screening: A Multicenter Retrospective Evaluation of Cytology and Partial HPV Genotyping

Background: Under the current Italian HPV-based cervical screening protocol, women with persistent high-risk human papillomavirus (HR-HPV) positivity at the 12-month recall are referred to colposcopy regardless of cytology results. Although this strategy maximizes lesion detection, it may result in a high number of colposcopies, most of which do not identify CIN2+. We evaluated whether partial HPV genotyping combined with cytology can improve risk stratification in women with persistent HPV positivity at the 12-month recall. Methods: A multicenter retrospective analysis of cervical screening data collected from four Italian screening programs (Florence, Padua, Umbria and Emilia Romagna) was performed. Among the 270,714 women screened between 2015 and 2019 (2016–2018 in Florence), 4589 women with persistent HR-HPV positivity at 12-month recall after baseline HPV-positive/NILM (negative for intraepithelial lesion or malignancy) results were included. Four alternative triage strategies based on cytology and HPV 16/18 genotyping were retrospectively simulated and compared with the current protocol. Outcomes included CIN2 (cervical intraepithelial neoplasia grade 2) or worse (CIN2+) detection, relative sensitivity for CIN2+ and CIN3 (cervical intraepithelial neoplasia grade 3) or worse (CIN3+), positive predictive value (PPV) and reduction in colposcopy referrals. Results: Under the current protocol, 4589 colposcopies identified 327 CIN2+ lesions, including 113 CIN3+ lesions, yielding a PPV for CIN2+ of 7.1%. Overall, 36.7% of CIN2+ lesions were associated with NILM cytology at recall. HPV 16/18 genotyping identified 43.3% of these cytologically occult lesions. In this retrospective simulation, among the evaluated strategies, referral of HPV 16/18-positive women regardless of cytology and referral of women with low-grade or worse cytology regardless of genotype would have achieved the most favorable balance between sensitivity and efficiency. This approach would have reduced colposcopy referrals by 57.7% (95% CI, 56.3–59.2), nearly doubled the PPV for CIN2+ to 13.4% (95% CI, 11.9–14.9) and maintained relative sensitivities of 79.2% (95% CI, 74.5–83.3) for CIN2+ and 85.0% (95% CI, 77.2–90.4) for CIN3+. Performance varied across centers, with imprecise center-specific estimates (relative CIN3+ sensitivity of Strategy 2 from 62.5% in Emilia Romagna to 91.1% in Umbria), so generalizability should be assessed cautiously. Conclusions: Combining partial HPV genotyping with cytology would have substantially improved risk stratification in women with persistent HR-HPV positivity at 12-month recall. In this retrospective simulation, a risk-based approach would have substantially reduced immediate colposcopies while maintaining a relative sensitivity for CIN3+ of 85.0%. These findings are hypothesis-generating and support prospective evaluation of integrating molecular and cytological information into 12-month recall management.