S3142 Comparative Efficacy of Endoscopic and Surgical Interventions for Gastroparesis: A Systematic Review and Network Meta-Analysis

Introduction: Gastroparesis is a chronic gastrointestinal motility disorder with limited response to pharmacologic therapy. Procedural interventions—including endoscopic and surgical approaches—have emerged as alternatives, but comparative data remain sparse. This network meta-analysis evaluated the relative efficacy of these procedural options for gastroparesis. Methods: We systematically searched PubMed, Embase, and CENTRAL from January 2000 to present for randomized and observational studies assessing procedural interventions in adults with gastroparesis. Interventions included gastric peroral endoscopic myotomy (G-POEM), pyloroplasty/pyloromyotomy, gastric neuronal stimulation (GNS), botulinum toxin injection (BTI), and combination therapies. Primary outcomes were clinical success and gastric emptying study (GES) improvement. A Bayesian network meta-analysis with a common-effects model was used. Surface Under the Cumulative Ranking curve (SUCRA) values ranked treatment efficacy. Results: Forty-nine studies comprising 2,722 patients were included in the study. For clinical success, 12 studies (k = 12) across 6 treatments (n = 6) and 6 unique study designs (d = 6) contributed 12 pairwise comparisons (m = 12). Compared to placebo, G-POEM showed the highest odds of clinical success (odds ratio [OR] 5.72; 95% confidence interval [CI] 2.02-16.18; P = 0.0010; SUCRA = 0.85), followed closely by G-POEM with GES (OR 5.42; 95% CI 1.47-19.99; P = 0.0111; SUCRA = 0.79) and pyloroplasty (OR 4.64; 95% CI 1.46-14.75; P = 0.0093; SUCRA = 0.71). BTI and GNS did not demonstrate a statistically significant benefit. For GES improvement, 8 studies (k = 8) evaluated 7 treatments (n = 7) with 6 designs (d = 6), contributing 8 pairwise comparisons (m = 8). In comparisons against sham procedures for gastric emptying improvement, pyloroplasty with GES ranked the highest (OR 100.0; 95% CI 5.79-1,728.15; P = 0.0015; SUCRA = 0.97), followed by G-POEM (OR 13.33; 95% CI 2.82-63.11; P = 0.0011; SUCRA = 0.69), and G-POEM with GES (OR 11.93; 95% CI 1.47-96.76; P = 0.0203; SUCRA = 0.63). BTI and GES remained the lowest-performing interventions across both primary outcomes. Conclusion: G-POEM, either alone or combined with GNS, offers the strongest evidence for symptom relief and significant improvement in gastric emptying. Pyloroplasty-based therapies also demonstrate high efficacy, particularly for GES outcomes. In contrast, BTI and GNS rank lowest, suggesting limited utility. These findings support G-POEM and pyloroplasty as preferred procedural options in refractory gastroparesis. GES improvement didn’t consistently match symptom relief, limiting its surrogate value.