Predictors of treatment failure after radiofrequency ablation for intramucosal adenocarcinoma in barrett esophagus: A multi-institutional retrospective cohort study

Radiofrequency ablation (RFA), with or without endoscopic mucosal resection (EMR), is a safe, effective, and durable treatment option for Barrett esophagus (BE)-associated dysplasia (DYS), but few studies have identified predictors of treatment failure in BE-associated intramucosal adenocarcinoma (IMC). The aim of this study was to determine the rate of IMC eradication when using RFA±EMR and to identify clinical and pathologic predictors of treatment failure. A retrospective review of medical records and a central review of index histologic parameters were performed for 78 patients who underwent RFA±EMR as the primary treatment for biopsy-proven IMC at 4 academic tertiary medical centers. Complete eradication (CE) (absence of IMC/DYS on first follow-up endoscopy) was achieved in 86% of patients, and durable eradication (DE) (CE with no recurrence of IMC/DYS until last follow-up) was achieved in 78% of patients, with significant variation between the 4 study sites (P=0.03 and 0.09 by analysis of variance for DE and CE, respectively). Use of EMR before RFA significantly reduced the risk for treatment failure for IMC/DYS (hazard ratio, 0.15; 95% confidence interval, 0.05-0.48; P=0.001), whereas IMC involving Z50% of the columnar metaplastic area on index examination significantly increased the risk for treatment failure (hazard ratio, 4.24; 95% confidence interval, 1.53-11.7; P=0.005). Endoscopic and pathologic factors associated with treatment failure in BE-associated IMC treated with RFA± EMR may help identify the subset of IMC patients for whom a more aggressive initial approach may be justified.