Diagnostic accuracy of p16/Ki-67 dual-stain cytology for detecting CIN2+ and CIN3+ in HPV-positive women: A systematic review and meta-analysis.

BACKGROUND: Primary human papillomavirus (HPV) testing has improved the sensitivity of cervical cancer screening but has lower specificity, creating a need for effective triage strategies among HPV-positive women. p16/Ki-67 dual-stained cytology has been proposed as a biomarker-based triage test that identifies HPV-associated transforming activity. OBJECTIVE: To evaluate the diagnostic accuracy of p16/Ki-67 dual-stained cytology for detecting histologically confirmed cervical intraepithelial neoplasia grade 2 or worse (CIN2+) and grade 3 or worse (CIN3+) in HPV-positive women. METHODS: PubMed, Scopus and Web of Science were searched from inception to 1 July 2026. Eligible studies included HPV-positive women, evaluated p16/Ki-67 dual-stained cytology as the index test, used histopathology as the reference standard, and provided extractable 2 × 2 diagnostic accuracy data. Only one non-overlapping dataset per study and endpoint was retained. No missing diagnostic cells were imputed, and published adjusted non-integer estimates were retained as reported. Pooled sensitivity and specificity were estimated using a bivariate random-effects Reitsma model. Certainty of evidence was assessed using the GRADE approach. RESULTS: Twenty-seven studies were included in the main CIN2+ analysis, comprising 24,519 HPV-positive women. Pooled sensitivity for CIN2+ was 0.837 [95% confidence interval (CI) 0.804-0.866], pooled specificity was 0.643 (95% CI 0.585-0.697), and area under the curve (AUC) was 0.826. The CIN3+ analysis included 22 endpoint-specific datasets comprising 22,489 women. Pooled sensitivity for CIN3+ was 0.862 (95% CI 0.825-0.892), pooled specificity was 0.614 (95% CI 0.550-0.675), and AUC was 0.830. Integer-only sensitivity analyses produced similar estimates for CIN2+ and CIN3+, and leave-one-out analyses showed that the pooled results were not driven by any single study. Matched comparative datasets suggested that p16/Ki-67 dual-stained cytology was generally more sensitive than cytology, particularly for CIN3+, although specificity advantages were inconsistent across cohorts. The overall certainty of evidence was very low for both endpoints because of risk of bias, indirectness, heterogeneity, imprecision and limitations in the interpretation of publication bias. CONCLUSION: p16/Ki-67 dual-stained cytology showed high pooled sensitivity and moderate specificity for detecting CIN2+ and CIN3+ among HPV-positive women, which may support its potential role as a reflex triage test in HPV-based cervical screening. However, the very low certainty of evidence, heterogeneous study populations, and variable verification pathways require cautious interpretation.