Worldwide Performance of a Methylation Biomarker for Cervical Precancer and Cancer Diagnosis

Abstract

The shift towards primary HPV-based screening has roused the search for a secondary triage test that provides a sufficient sensitivity to detect high grade cervical intraepithelial neoplasia (CIN) and cancer, but also brings a high specificity to avoid unnecessary clinical work and colposcopy referrals. The S5 DNA-methylation classifier based on target CpG sites of the host gene EPB41L3, and viral gene regions of HPV16, 18, 31 and 33 has demonstrated an improved performance for detecting CIN2+ compared to either HPV16/18 genotyping, cytology or combination.In this study, we tested for the first time the performance of the S5 DNA-methylation classifier in detecting CIN3 and invasive cervical cancer from diverse geographic settings using the cut-off of 0·80 and the exploratory cut-offs of 2·62 and 3·70. Assays were performed using exfoliated cervical specimens (n=808) and formalin-fixed biopsies (n=160) from women diagnosed with cytology negative results (n=220), CIN3 (n=204) and cervical cancer (n=544). Methylation increased proportionally with disease severity in all geographical regions tested (Cuzick test for trend, p<0·0001). S5 separated women with negative histology from those with CIN3 or cervical cancer (p<0·0001). At the 0.80 cut-off, a total of 543/544 cervical cancers were correctly identified as S5 positive (99·81%, 95%CI 98·34-99·96) while at 3·70 the classifier showed a positivity of 95·77% (95%CI 92·39- 97·40). The S5 odds ratios when negative for cervical disease versus CIN3+ were significantly higher than for HPV16/18 genotyping at all cut-offs. Notably, at the 0·80 cut-off, cervical cancers which were consistently hrHPV-negative when tested with multiple hrHPV genotyping assays were 96·15% (95%CI 94·38-98·25) identified by the S5 classifier. These cancers would have been missed in the current primary hrHPV screening programme. EPB41L3 methylation had a stronger association with disease severity F=367·50 than age F=81·00 (p<0·0001), indicating that methylation on EPB41L3 might have a strong potential to predict CIN progression, independent of natural epigenetic methylation that occurs with age. The S5-classifier can accurately detect high-grade CIN and malignancy irrespective of geographic context and setting. The classifier has the potential to be used in clinics as a screening and triage tool. Adjustment of the S5 cut-off can be performed taking into account the relative importance given to sensitivity versus specificity, thus reflecting local triage modality needed. Funding Statement: Cancer research UK who funded part of this study (Cancer Prevention programme Grant C569/A10404/CRUK_/Cancer Research UK/United Kingdom Declaration of Interests: The authors declare no conflict of Interest. Ethics Approval Statement: The samples used for this study have received ethical and research approvals from: i) The International Agency for Research on Cancer (IARC) of the World Health Organization (WHO) for all samples from Bhutan, Colombia, Georgia, India and Philippines, South Africa and Spain; ii) The Scottish HPV Archive, MRC Centre for Reproductive Health, The University of Edinburgh for all samples from the United Kingdom; iii) Department of Microbial, Cellular and Molecular Biology, Addis Ababa University for all samples from Ethiopia; iv) University of New Mexico Human Research Review Committee for all samples from USA – New Mexico. The study protocol has been approved by Queen Mary University of London’s Ethical Committee (MTA-2018-ICE-0426, MTA-2018-ICE-1353, 11/06/2018).