Methylation Biomarker Panel Performance in Esophacap Cytology Samples for Diagnosing Barrett's Esophagus: A Double-Blind Prospective Validation Study

Background: Barrett's esophagus (BE) is the only known precursor of esophageal adenocarcinoma (EAC). Current guidelines recommend endoscopic screening of patients with two or more potential BE risk factors but fail to capture two-thirds of prevalent BE cases. Given the lack of an acceptable, accurate and cost-effective tool for BE screening, new methods are needed to allow population-based screening. Here, we sought to develop a BE screening method based on methylation status in cytology samples captured by EsophaCap, a retrievable sponge sampling device, using a streamlined sensitive technique, methylation on beads (MOB). Methods: We conducted a double-blinded, prospective cohort study. 173 participants, meeting inclusion criteria, were enrolled, 80 of whom completed the study and were included in statistical analyses. We employed MOB to extract and bisulfite-convert DNA, then qMSP to assess methylation of 8 candidate genes. Lasso regression was applied to establish a prediction model for BE diagnosis. Findings: 85% (147/173) of participants were able to swallow the EsophaCap. In the training set, 5 of 8 candidate methylation biomarkers (p16, HPP1, NELL1, TAC1, and AKAP12) were significantly higher in the BE group than the control group. In a training set of EsophaCap samples, a 4-biomarker-plus-age lasso regression model showed high accuracy of BE diagnosis: AUC was 0.894, with sensitivity 94.4% (95%CI 0.71%~99%) and specificity 62.2% (95%CI 44.6%~77.3%). The lasso model also performed well in an independent test set: AUC=0.857 (p=0.001), with sensitivity=85.7% (95%CI 56.2%~97.5%) and specificity=92.8% (95%CI 64.1%~99.6%). Interpretation: EsophaCap, in combination with an epigenetic biomarker panel and MOB, is a promising, well-tolerated, low-cost esophageal sampling strategy for BE diagnosis. This approach merits further prospective studies in larger populations. Funding Statement: National Institutes of Health (Grants: CA211457, DK118250) and a Discovery Award from The Johns Hopkins University School of Medicine. Declaration of Interests: All authors state that there are no conflicts of interest. Ethics Approval Statement: This study was approved by the JHM IRB.