Risk‐Based Triage Strategy by Extended HPV Genotyping for Women With LSIL Cytology: A Real‐World Study

TL;DR

This study proposes a new triage strategy for women with LSIL cytology using extended HPV genotyping to reduce unnecessary colposcopies while maintaining detection accuracy.

Contribution

A novel risk-based triage strategy using extended HPV genotyping for LSIL women is developed and validated in a real-world setting.

Findings

HPV16, 73, and 33 are associated with the highest immediate CIN2+/3+ risk.

The new strategy significantly increases specificity while reducing colposcopy referrals by 19.82%.

Group A genotypes require immediate colposcopy, while Groups B and C suggest alternative follow-up methods.

Abstract

To evaluate the immediate risk of (pre)cancer for cytology low‐grade squamous intraepithelial lesion (LSIL) women infected with or without specific HPV genotype and develop a risk‐based management strategy. A total of 4567 LSIL women with extended HPV genotyping and colposcopy results were enrolled according to the inclusive and exclusive criteria. The distribution and immediate cervical intraepithelial neoplasia grade 2 or worse and 3+ or worse (CIN2+/3+) risks of specific HPV genotypes were assessed using Minimum Estimate, Any Type Estimate, and Hierarchical Attribution Estimate. A risk‐based strategy was further established and evaluated. CIN2+/3+ were 729/328 cases, including 691/317 in 3398 HPV‐positive and 38/11 in 1169 HPV‐negative women. HPV16, 52, 58, and 18 were the most prevalent genotypes in both HPV‐positive and CIN2+/3+ cases. HPV16, 73, and 33 carried the highest immediate CIN2+/3+ risk. A risk‐based strategy was established, which suggested Group A (HPV 16, 33, 45, 31, 18, 58, 52, 35, 73, 82; with immediate CIN3+ risk of 4.08%–22.12%) for immediate colposcopy, Group B (HPV 59, 66, 56, 53) for 6‐month follow‐up or p16/Ki‐67 dual stain or DNA methylation triage, while Group C (HPV 51, 68, 39, 26) for 1‐year HPV repeat testing. Compared with conventional strategy, this new strategy showed significantly higher specificity (CIN2+: 52.16% vs. 29.47%, χ
2 = 409.136, p < 0.001; CIN3+: 48.45% vs. 27.32%, χ
2 = 402.395, p < 0.001) but similar sensitivity, which could reduce immediate colposcopy referrals by 19.82%. A risk‐based triage strategy for LSIL women with extended HPV genotyping could effectively reduce unnecessary colposcopies and maintain high efficacy for CIN2+/3+ detection.

Figures