S859 Detecting What Matters: A Meta-Analysis of a Minimally Invasive Screening Device for Long-Segment Barrett’s Esophagus in Surveillance Populations

Introduction: Current guidelines recommend endoscopic surveillance every 3 to 5 years for patients diagnosed with Barrett’s esophagus (BE) due to the risk of progression to esophageal adenocarcinoma (EAC). However, long (≥3 cm) segment BE (LSBE) carries a 3-fold higher risk of progression compared to short (<3 cm) segment BE. In this study, we assessed the screening performance of Cytosponge with Trefoil Factor 3 (TFF3) for detecting BE and LSBE in individuals undergoing surveillance. Methods: Following PRISMA guidelines, our study protocol was pre-registered in PROSPERO (CRD420250652091). Eligibility criteria included adults who were undergoing surveillance for BE or had a history of BE diagnosed by endoscopic assessment with biopsies. An extensive literature review of databases including Embase, PubMed, and Cochrane Central was performed from January 2009 through March 29, 2025. Two reviewers independently screened and extracted data using the COVIDENCE tool, then assessed risk of bias using the QUADAS-2 tool. Sensitivity and specificity were estimated in Statistical Analysis Software v9.4 using bivariate mixed models, which were used to estimate likelihood ratios, diagnostic odds ratios, and predictive values, whereas forest plots were created in Stata v. 19.5. Results: We included 9 studies involving 6,697 participants with history of BE, of which 5 studies (N = 6,298 participants) reported results stratified by BE length. The pooled screening performance of Cytosponge TFF3 for detecting any BE was as follows: sensitivity 0.75 (95% confidence interval [CI] 0.56-0.87), specificity 0.83 (95% CI 0.66-0.92) and diagnostic odds ratio (DOR) of 14.63 (95% CI 1.55-27.70). For LSBE, pooled estimates showed higher sensitivity at 0.83 (95% CI 0.72-0.90 lower specificity at 0.59 (95% CI 0.28-0.84), and lower DOR at 7.10 (95% CI 2.09-12.10). Conclusion: This is the first meta-analysis to assess the screening performance of Cytosponge with TFF3 for detecting LSBE in patients with BE. Some heterogeneity existed across studies due to differences in timing of endoscopic BE diagnosis, and most studies were conducted in the U.K. Nonetheless, the overall findings remain informative for evaluating performance in real-world BE populations. Cytosponge demonstrated reliable specificity for ruling out BE and strong sensitivity for detecting LSBE. Overall, Cytosponge combined with TFF3 is a valuable, minimally invasive tool for risk stratifying patients with BE in clinical practice.