887 Clinical Efficacy of Endoscopic Vacuum-Assisted Closure Therapy (EVAC) in Gastroesophageal Defects: A Systematic Review and Comparative Meta-Analysis With Endoscopic Stent Therapy (EST)

Abstract

INTRODUCTION: Treatment of gastroesophageal leaks and perforations after upper gastrointestinal surgery remains clinically challenging. Advanced endoscopic interventions have traditionally included placement of self-expanding metal or plastic stents i.e. endoscopic stent therapy (EST) and endoscopic vacuum-assisted closure therapy (EVAC). The aim of our study was to evaluate the overall clinical efficacy of EVAC when compared to EST in the treatment of gastroesophageal defects. METHODS: Our primary goals were to analyze the pooled rates of clinical success and overall mortality with EVAC therapy in gastroesophageal defects. Using EST as a comparator group, we analyzed clinical outcomes and predictors of clinical success. RESULTS: 6 cohort studies compared EVAC (109 patients) to EST (197 patients). These included 313 cases of esophageal anastomotic leaks. The pooled odds ratio of clinical success with EVAC was 2.9 (95% CI 1.4-5.8, P = 0.003, I2 = 13) when compared to EST in the treatment of esophageal leaks secondary to surgical anastomosis. The pooled odds ratio of mortality was 0.43 (95% CI 0.19-0.95, P = 0.04, I2 = 0). When all studies that evaluated EVAC alone were analyzed (14 studies - 89 perforations and 136 anastomotic leaks), the pooled rate of clinical success was 87.5% (95% CI 81.6-91.7, I2 = 0, P = 0.001). For studies evaluating EST (12 studies, 349 patients), the pooled rate of clinical success was 75.4% (95% CI 75.6-84.1, P = 0.001, I2 = 30). EVAC performed significantly better than EST (P = 0.004). The pooled mortality rate with EVAC was 9.1% (95% CI 4.5-17.5, P = 0.01, I2 = 0) and with EST was 17.2% (95% CI 11.2-20.2, P = 0.001, I2 = 38). Based on univariate meta-regression analysis, anastomotic leaks demonstrated statistical significance as a predictive factor of clinical success with EVAC (2-sided P-value 0.05, R2 analog = 1), whereas perforation almost reached statistical significance (2-sided P-value 0.06, R2 analog = 0). On multivariate meta-regression, neither of the etiologies showed statistical significance as a possible predictor of clinical success. We performed meta-regression analysis on mortality data and found no significant predictive effect of etiology in EVAC and EST groups, except for anastomotic leaks that almost reached statistical significance as a predictor of mortality with EST (2-sided P-value 0.06, R2 analog = 1). CONCLUSION: EST performed favorably when compared to EST in the treatment of gastroesophageal defects with higher clinical success and lesser mortality.