NONENDOSCOPIC DETECTION OF BARRETT'S ESOPHAGUS IN PATIENTS WITHOUT GERD SYMPTOMS.

Abstract

Background: Screening efforts for Barrett's Esophagus (BE) predominantly focus on performing upper endoscopy (EGD) on patients with gastroesophageal reflux disease (GERD) symptoms who have additional risk factors for BE. However, cost and invasiveness preclude EGD in those who have no prior GERD symptoms, despite having other risk factors, representing missed opportunities for BE screening in individuals who account for approximately 40% of the patients who eventually develop esophageal adenocarcinoma (EAC).

Aim: The aim of this study was to evaluate if non-endoscopic methods can enable BE detection in an at-risk population without GERD symptoms.

Methods: Patients presenting for colonoscopy who had not undergone previous EGD plus had ≥3 BE risk factors (from among age > 50 years, male sex, white race, smoking history, family history of BE/EAC, or central obesity) without chronic GERD symptoms were prospectively recruited for non-endoscopic screening. Trained nurses administered the EsoCheck (Lucid Diagnostics) encapsulated balloon. Samples were assayed with the EsoGuard BE detection methylated DNA marker panel (Lucid Diagnostics). Patients with a positive result were offered standard-of-care EGD, while patients with a negative EG result were offered free of cost research EGD. Positive predictive value (PPV), negative predictive value (NPV), and BE prevalence were calculated.

Results: The mean age of the 132 study subjects was 60.7 years, 129 (98%) were white, 124 (94%) were male, 71 (54%) had a prior smoking history, 46 (35%) were centrally obese, and 5 (4%) reported a family history. EsoCheck was successfully administered in 124 (94%) and the EsoGuard methylated DNA marker panel could be assayed in 120 (97%) of the samples. Thirty-four assays were positive of which 27 underwent a follow-up EGD and BE was identified in 9, PPV = 33% [17%, 54%] subjects. EGD was also performed in 22 of the 86 subjects whose assays were negative and none of them had BE, NPV = 100% [85%, 100%]. A logistic regression model fitted to impute the presence of BE estimated the PPV as 27% [13%, 44%], NPV as 98% [92%, 100%], and BE prevalence as 8.4% [4.5%, 14.3%].

Conclusion: Patients without chronic GERD who have ≥3 BE risk factors have a moderately high prevalence of BE. Non-endoscopic detection can effectively identify BE, enabling expansion of screening to this larger at-risk population. Those with a negative EG assay have a low likelihood of BE.