Introduction/Background Epigenetic alterations are essential in the development of many cancers and can occur years in advance of histopathological change. Interest in methylation testing for enhanced cervical screening is increasing but epigenome-wide exploration has been limited and the best methylation markers are yet unknown. In this epigenome-wide association study (EWAS) of >850,000 methylation sites, we are the first to explore DNA methylation signatures associated to CIN3 and cervical cancer, to better understand potential drivers of cervical carcinogenesis and estimate performance of methylation as a diagnostic test.

Methodology 247 women (119 normal, 74 CIN3/CGIN and 54 cervical cancer) attending gynaecological appointments or oncological treatment between 2014–2020 were recruited. Methylation signatures were obtained following bisulphite conversion of DNA extracted from exfoliated cervical cells and sequenced using the Illumina 850k array. After data QC, logistic regression and a conditional analysis were used to test for independent associations between methylation CpG sites and CIN3 or Cancer. P-values were Bonferroni corrected and adjusted for batch, chip, age and HPV status. Diagnostic test accuracy was estimated and tested in an independent cohort.

Results 409 CpG methylation sites were strongly associated with CIN3 or cancer (P <5x10–8). Two CpG sites located in PAX1 and NREP-AS1 genes were found to be independently associated with CIN3/Cancer. Combining sites into a bigenic marker led to an AUC of 0.92 in a separate subset of the discovery cohort and 0.77 in an independent cohort of CIN3 cases.

Conclusion This is potentially the first study to highlight epigenome-wide epigenetic signatures associated to CIN3 and cervical cancer. Functional annotation highlighted several genes related to tumour suppressor function and apoptosis are increasingly methylated in CIN3 and cervical cancer. PAX1 and NREP-AS1 as a methylation test has the potential to increase the accuracy of cervical screening through enhanced detection of women with CIN3 or cancer and future self-sampling.

• Download figure
• Open in new tab
• Download powerpoint

Abstract #999 Figure 1

ROC curves for logistic models including cg16767801 (in blue), or cg23642047 (in green) or both (in red) as predictors and all cases (A), cancer (B) or CIN3 (C, D) as outcome. Models were fitted on 1,000 training sets with 80% of the cases and 80% of the controls and performances were evaluated on a test set with the remaining 20% of participants. The pointwise average, 5th and 95th percentiles of the True and False Positive Rates and Area Under the Curve (AUC) are reported. Results presented in panels A, B and C were obtained using our study population (N=114 controls, N=73 CIN3 cases and N=54 cancer cases). An independent validation set (accession number GSE14375s, with N=54 controls and N=42 CIN3 cases) was used in panel D.

Disclosures This work was funded by NIHR and the Wellcome Trust. The authors have no conflicts of interest to declare.