The Utility of Endoscopic Ultrasound (EUS) in Patients with Barrett's Esophagus (BE) and High Grade Dysplasia (HGD): Analysis of the AIM Dysplasia Trial Experience

Abstract

Introduction: EUS is often used in evaluating BE patients with HGD prior to treatment. Previous reports evaluating its utility in HGD have predominantly consisted of single-center studies with limited patient enrollment and have shown conflicting results. Aims: The primary aim was to assess if EUS findings alter management decisions in BE patients with HGD being considered for ablation therapy. A secondary aim was to determine the overall frequency of abnormal findings on EUS. Methods: We analyzed data from the AIM Dysplasia Trial; a randomized, sham-controlled study involving 19 US sites that evaluated radiofrequency ablation (RFA) for the treatment of BE patients with dysplasia. BE patients with HGD confirmed by a centralized expert pathology group (Cleveland Clinic(CC)) that had undergone EUS evaluation were included in the analysis. Abnormal EUS findings included the presence of submucosal invasion, nodules, and lymphadenopathy (LA). The frequency of EUS findings not visualized at endoscopy, the impact of EUS on patient management, and the frequency of complications were also assessed. Results: Of 84 patients with HGD confirmed by CC being evaluated for RFA, 62 patients with available EUS results were included in this analysis. EUS was normal in 53 (86%) patients, while abnormal findings were seen in 9 (14%) patients. These findings included the detection of esophageal nodules in 2 (3%) and LA in 7 (11%). EUS-guided fine needle aspiration (EUS-FNA) was limited to 2 patients with LA suspicious for malignancy. One showed benign cytology while the other was positive for malignant celiac and mediastinal LA from an undiagnosed lung cancer. Only 2 (3%) patients had EUS findings that were not previously identified on standard endoscopy. In one, a 7 × 7 mm esophageal submucosal nodule was seen on EUS and underwent endoscopic mucosal resection of a benign lesion. Patient management regarding ablation therapy was unaffected, as the patient was subsequently enrolled in the trial. The second was the aforementioned EUS-FNA of an undiagnosed lung cancer, resulting in trial exclusion. Finally, an attempted EUS in 1 patient was complicated by a cervical esophageal perforation that precluded further evaluation. Conclusions: In this study, the frequency of EUS findings altering management (1.6%) was equaled by the risk of a major complication. The overall frequency of abnormal EUS findings in patients with BE and HGD without mucosal defects on standard endoscopy was low, rarely altered management, and hence questions the utility of the routine use of EUS in these patients.