Treatment with endoscopic submucosal dissection (ESD) is associated with higher rates of complete remission of dysplasia at 2 years compared with cap-assisted endoscopic mucosal resection (cEMR) in patients with Barrett’s esophagus with dysplasia or early-stage intramucosal esophageal adenocarcinoma (EAC), according to study findings.
Despite the seeming advantage of ESD over cEMR, the study found similar rates of complete remission of intestinal metaplasia (CRIM) between the treatment groups at 2 years.
The study authors explained that ESD, a recent development in endoscopic resection, allows for en bloc resection of larger lesions in dysplastic Barrett’s and EAC and features less diagnostic uncertainty compared with cEMR. Findings from the study highlight the importance of this newer technique but also emphasize the utility of both treatments. “In expert hands both sets of procedures appear to be safe and well tolerated,” wrote study authors Don Codipilly, MD, of the Mayo Clinic in Rochester, Minn., and colleagues in Clinical Gastroenterology and Hepatology.
Given the lack of comparative data on the long-term outcomes of cEMR versus ESD in patients with neoplasia associated with Barrett’s esophagus, Dr. Codipilly and colleagues examined histologic outcomes in a prospectively maintained database of 537 patients who underwent endoscopic eradication therapy for Barrett’s esophagus or EAC at the Mayo Clinic between 2006 and 2020. Only patients who had undergone either cEMR (n = 456) or ESD (n = 81) followed by endoscopic ablation were included in the analysis.
The primary endpoint of the study was the rate and time to complete remission of dysplasia (CRD), which was defined by the absence of dysplasia on biopsy from the gastroesophageal junction and tubular esophagus during ≥1 surveillance endoscopy. Researchers also examined the rates of complications, such as clinically significant intraprocedural or post-procedural bleeding that required hospitalization, perforation, receipt of red blood cells within 30 days of the initial procedure, and stricture formation that required dilation within 120 days of the index procedure.
Patients in the ESD group had a longer mean length of resected specimens (23.9 vs. 10.9 mm; P <.01) as well as higher rates of en bloc (97.5% vs. 41.9%; P <.01) and R0 resection (58% vs. 20.2%; P <.01). Patients were generally balanced on other basic baseline demographics, including age, sex distribution, and smoking status.
Over a median 11.2-year follow-up period, a total of 420 patients in the cEMR group achieved CRD. In the ESD group, 48 patients achieved CRD over a median 1.4-year follow-up period. The 2-year cumulative probability of CRD was lower in patients who received cEMR versus those who received ESD (75.8% vs. 85.6%, respectively). In a univariate analysis, the odds of achieving CRD were lower in cEMR versus ESD (hazard ratio, 0.41; 95% CI, 0.31-0.54; P <.01).
According to multivariate analysis, 2 independent predictors of CRD included ESD (hazard ratio, 2.38; P <.01) and shorter Barrett’s segment length (HR, 1.11; P <.01).
The investigators also assessed whether advancements made in cEMR technique have contributed to the findings in an analysis of patients who underwent cEMR (n = 48) with ESD (n = 80) from 2015 to 2019. In this analysis, the researchers found that the odds of CRD were lower than that of ESD (HR, 0.67; 95% CI, 0.45-0.99). Additionally, higher odds of achieving CRD in the cEMR group were observed in years between 2013 and 2019 (n = 129) compared with years 2006 and 2012 (n = 112) (HR, 2.09; 95% CI, 1.59-2.75; P <.01).
Demographic and clinical variables were incorporated into a Cox proportional hazard model to identify factors associated with decreased odds of CRD. This analysis found that decreased odds of CRD were associated with longer Barrett’s esophagus segment length (HR, 0.90; P <.01) and treatment with cEMR versus ESD (HR, 0.42; P <.01).
Over median follow-up periods of 7.8 years in the cEMR group and 1.1 years in the ESD group, approximately 78.5% and 40.7% of patients, respectively, achieved CRIM. While those in the ESD group achieved CRIM earlier, the cumulative probabilities of CRIM were similar by 2 years (59.3% vs. 50.6%; HR, 0.74; 95% CI, 0.52-1.07; P =.11). Shorter Barrett’s esophagus segment was the only independent predictor of CRIM (HR, 1.16; P <.01).
The researchers noted that the study population may have included patients with more severe disease than that in the general population, which may limit the generalizability of the findings. Additionally, the lack of a randomized design was cited as an additional study limitation.
In spite of their findings, the researchers explained that “continued monitoring for additional outcomes such as recurrence are required for further elucidation of the optimal role of these procedures in the management of” neoplasia associated with Barrett’s esophagus.”
The study was funded by the National Cancer Institute and the Freeman Foundation. The researchers reported no conflicts of interest with any pharmaceutical companies.