Diagnostic testing accuracy of DNA methylation tests for detection of high-grade cervical intraepithelial neoplasia and cervical cancer: a systematic review and meta-analysis
File(s)
Published version
Author(s)
Type
Journal Article
Abstract
Background
Human papillomavirus (HPV)-based cervical screening offers stronger protection against cervical intraepithelial neoplasia (CIN) than screening with cytology. DNA methylation tests have been proposed as an alternative triage test to cytology, although there is currently no consensus on the most accurate gene or gene panels (“markers”).
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 14.41–30.08]), C13ORF18 (19.50 [95%CI 11.65–32.65]), and PAX1 (17.13 [95%CI 9.20–31.89]). EPB41L3 had the lowest LR- (0.26 [95%CI 0.19–0.35]). In all women, SOX1 had the highest DOR (29.72 [95%CI 12.84–68.81]), although JAM3 and PAX1 maintained good accuracy (DORs 26.73 [95%CI 19.97–35.77] and 23.96 [95%CI 17.32–33.15]).
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.
Human papillomavirus (HPV)-based cervical screening offers stronger protection against cervical intraepithelial neoplasia (CIN) than screening with cytology. DNA methylation tests have been proposed as an alternative triage test to cytology, although there is currently no consensus on the most accurate gene or gene panels (“markers”).
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 14.41–30.08]), C13ORF18 (19.50 [95%CI 11.65–32.65]), and PAX1 (17.13 [95%CI 9.20–31.89]). EPB41L3 had the lowest LR- (0.26 [95%CI 0.19–0.35]). In all women, SOX1 had the highest DOR (29.72 [95%CI 12.84–68.81]), although JAM3 and PAX1 maintained good accuracy (DORs 26.73 [95%CI 19.97–35.77] and 23.96 [95%CI 17.32–33.15]).
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.
Date Issued
2026-07-26
Date Acceptance
2026-05-16
Citation
European Journal of Cancer, 2026, 243
ISSN
0959-8049
Publisher
Elsevier BV
Journal / Book Title
European Journal of Cancer
Volume
243
Copyright Statement
© 2026 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
License URL
Identifier
https://www.ncbi.nlm.nih.gov/pubmed/42235467
PII: S0959-8049(26)00604-0
Subjects
‘Cervical intraepithelial neoplasia’
‘Cervical screening’
‘DNA methylation’
‘Diagnostic testing’
‘Epigenetics’
‘Human papillomavirus’
‘Meta-analysis’
Humans
Female
Uterine Cervical Dysplasia
DNA Methylation
Uterine Cervical Neoplasms
Biomarkers, Tumor
Early Detection of Cancer
Papillomavirus Infections
Human Papillomavirus Viruses
Sensitivity and Specificity
Publication Status
Published
Coverage Spatial
England
Article Number
116823
Date Publish Online
2026-05-27