Diagnostic testing accuracy of DNA methylation tests for detection of high-grade cervical intraepithelial neoplasia and cervical cancer: A systematic review and meta-analysis

Methods We conducted a meta-analysis of accuracy of human and HPV methylation markers to detect CIN2 or worse (CIN2+), CIN3+, and cervical cancer (CRD42022299760). We systematically searched three databases to March 2025. We used a bivariate random-effects model to produce pooled sensitivity, specificity, diagnostic odds ratios (DORs), positive and negative likelihood ratios (LR+, LR-) for each marker at each available endpoint. Our primary analysis was in HPV+ women, and secondary analysis included all data regardless of HPV positivity

Results We included 51 studies exploring 13 methylation markers in 28,977 HPV+ women, and in total 125 studies assessing 32 markers in 49,242 women. For HPV+, the markers with the highest DORs were JAM3 (20.82 [95%CI 14.41�C30.08]), C13ORF18 (19.50 [95%CI 11.65�C32.65]), and PAX1 (17.13 [95%CI 9.20�C31.89]). EPB41L3 had the lowest LR- (0.26 [95%CI 0.19�C0.35]). In all women, SOX1 had the highest DOR (29.72 [95%CI 12.84�C68.81]), although JAM3 and PAX1 maintained good accuracy (DORs 26.73 [95%CI 19.97�C35.77] and 23.96 [95%CI 17.32�C33.15])

Conclusion Several methylation markers reported good sensitivity in the detection of CIN3+ whilst maintaining high specificity. These have potential to reduce unnecessary referrals to colposcopy. Methylation markers may permit efficient triage on self-samples which are inappropriate for cytology, although this requires further study ideally in screening cohorts.