Adverse Events After Radiofrequency Ablation in Patients With Barrett's Esophagus: A Systematic Review and Meta-analysis

Background & Aims Radiofrequency ablation (RFA) with or without endoscopic mucosal resection (EMR) is routinely used for treatment of Barrett's esophagus with dysplasia. Despite the relative safety of this method, there have been imprecise estimates of the rate of adverse events. We performed a systematic review and meta-analysis to assess the rate of adverse events associated with RFA with and without EMR. Methods We searched MEDLINE, Embase, Web of Science, and Cochrane Central through October 22, 2014. The primary outcome of interest was the overall rate of adverse events after RFA with or without EMR. We used forest plots to contrast effect sizes among studies. Results Of 1521 articles assessed, 37 met our inclusion criteria (comprising 9200 patients). The pooled rate of all adverse events from RFA with or without EMR was 8.8% (95% confidence interval [CI], 6.5%–11.9%); 5.6% of patients developed strictures (95% CI, 4.2%–7.4%), 1% had bleeding (95% CI, 0.8%–1.3%), and 0.6% developed a perforation (95% CI, 0.4%–0.9%). In studies that compared RFA with vs without EMR, the relative risk for adverse events was significantly higher for RFA with EMR (4.4) (P = .015). There was a trend toward higher proportions of adverse events in prospective studies compared with retrospective studies (11.3% vs 7.8%, P = .20). Other factors associated with adverse events included Barrett's esophagus and length and baseline histology. Conclusions In a systematic review and meta-analysis, we found the relative risk for adverse events from RFA to be about 4-fold higher with EMR than without; we identified factors associated with these events. Endoscopists should discuss these risks with patients before endoscopic eradication therapy.