Nonendoscopic Screening for Barrett's Esophagus and Esophageal Adenocarcinoma in At-Risk Veterans.

Abstract

Although rates of esophageal adenocarcinoma (EAC) in the United States continue to rise, many patients at risk of disease are not screened. EsoCheck (EC), a nonendoscopic esophageal balloon sampling device coupled with EsoGuard (EG), a DNA-based screening assay, is an US Food and Drug Administration-approved minimally invasive alternative to the traditional screening method of upper endoscopy. The objective of this study was to prospectively determine the diagnostic accuracy, tolerance, and acceptability of the EC/EG test in a screening population. We recruited veterans who met the American College of Gastroenterology Guideline criteria for endoscopic Barrett's esophagus (BE) and EAC screening at the Louis Stokes Cleveland Veterans Affairs Medical Center. All study participants completed unsedated EC-guided distal esophageal sampling followed by a sedated esophagogastroduodenoscopy (EGD). Diagnostic yield of the EG assay and EGD was recorded and used in calculation of sensitivity and specificity of EC/EG in prospective screening. The abbreviated Spielberger State-Trait Anxiety Inventory questionnaire was administered before and after completion of EC. Overall tolerance of EC sampling was evaluated on a 10-point Likert scale. Esophageal cancer screening was accepted by 130 of 782 eligible veterans (16.6%), and we analyzed results of those who completed both screening tests (N = 124). Prevalence of BE/EAC among studied veterans was 12.9% (16/124), based on EGD. Sensitivity and specificity of EC/EG for EGD-detected BE/EAC were 92.9% (95% confidence interval [CI] 66.1-99.8) and 72.2% (95% CI 62.1-80.8), respectively. Positive and negative predictive values were 32.5% (95% CI 18.6-49.1) and 98.6% (95% CI 92.4-100), respectively. Baseline Spielberger State-Trait Anxiety Inventory-6 scores were reflective of notable levels of anxiety among veterans in the periprocedural setting. The mean postprocedure acceptability score for the EC test was 7.23 (SD 2.45). Our data suggest excellent sensitivity and negative predictive value of EC/EG in a screening population of veterans, making this modality a powerful screening tool for BE and EAC.