Efficacy and Predictors of Stricture Development Following Radiofrequency Ablation for Barrett's Esophagus at a Tertiary Referral Center

Abstract

BACKGROUND: Radiofrequency ablation (RFA) is an endoscopic ablation modality used to treat Barrett's esophagus (BE) with the goal of eliminating dysplasia and metaplasia. Factors associated with stricture formation or incomplete eradication of intestinal metaplasia (EIM) are poorly understood. AIM: To determine the factors associated with stricture formation or incomplete EIM. METHODS: This was a retrospective study of all patients treated with RFA for BE at a tertiary care referral center between June 2006 and November 2010. Pertinent information was extracted from medical records, including: demographics, history of BE (pre-ablation histology, duration of pre-treatment dysplasia), medication and substance use, indicators of GERD activity (symptoms, presence of erosive esophagitis), upper endoscopy findings (Prague criteria, hiatus hernia), ablation outcomes (elimination of metaplasia and dysplasia), and complications (perforation, stricture, bleeding, and hospitalization). Outcomes related to RFA were described for all patients as well as stratified by pre-ablation histology. Comparative analysis of patients with and without stricture and complete and incomplete elimination of dysplasia were performed with non-parametric tests (Fisher's exact test for categorical data, Wilcoxon rank-sum test for continuous data) to determine associated factors. RESULTS: Among 113 patients who received RFA for BE (22 low-grade dysplasia, 77 high-grade dysplasia, 14 intramucosal carcinoma), 83 (73.5%) completed treatment with 95.2% complete elimination of dysplasia and 85.5% complete EIM. Of the remaining 30 subjects, 22 had ongoing treatment, 6 were lost to follow up, one had treatment delay for antireflux surgery and 1 opted for esophagectomy. Nine patients (8.0%) experienced a treatment-related complication, including8 strictures and 1 post-procedure hemorrhage. Stricture formation was associated with receiving endoscopic mucosal resection (75.0% vs. 36.2%, p=0.05) and number of EMR sessions (mean 1.4 vs. 0.4, p=0.007). A trend toward stricture formation existed with number of focal RFA treatments (mean 3.3 vs. 2.2, p=0.09), active NSAID use (75.0% vs. 44.8%, p=0.14) and prior peptic stricture (25.0% vs. 7.6%, p=0.15). Incomplete EIM was associated with ongoing GERD symptoms (75.0% vs. 33.8%, p=0.01) while increased Prague M length had a trend toward association (mean 6.6 vs. 4.5cm, p=0.11). CONCLUSIONS: RFA at a tertiary referral center is both safe (8.0% with complications, mostly benign strictures) and efficacious (95.2% elimination of dysplasia, 85.5% elimination of intestinal metaplasia). Previous EMR is associated with stricture formation while ongoing GERD symptoms are associated with incomplete elimination of intestinal metaplasia. These findings should be considered in planning treatment protocols for patients with BE.