First Description of Redo Gastric Per-Oral Endoscopic Myotomy for Refractory Gastroparesis: Technical Conduct and Outcomes.

Introduction: Gastroparesis is a progressive disease that may require endoscopic or surgical intervention, such as gastric per oral endoscopic myotomy (G-POEM), when conservative measures are ineffective or not tolerated. Limited data exist on redo G-POEM safety and effectiveness. We describe outcomes of patients undergoing repeat G-POEM for refractory gastroparesis. Methods: A retrospective review of patients who underwent redo G-POEM at our center from 2008 to 2024. Data included demographics, gastroparesis etiology, previous treatments, Gastroparesis Cardinal Symptom Index (GCSI), gastric emptying studies (GES), and clinical outcomes. Results: Three patients (mean age at first G-POEM 52.3 ± 19 years) were identified. All procedures were technically successful, and all 3 patients survived to discharge without major complications. Patient A (sarcoidosis-induced gastroparesis) had transient symptomatic improvement after the first G-POEM, but symptoms deteriorated the following year. GES showed initial improvement (26% to 14% 4-hour retention) but returned to baseline (26%) after the second G-POEM. She ultimately required jejunostomy tube placement. Patient B (idiopathic gastroparesis) with a history of renal transplant showed no objective improvement in GES after either procedure (40% to 41% 4-hour retention) and remained noncompliant with dietary recommendations. He died at age 35 from cardiac arrhythmia 2 years after the second G-POEM. Patient C (postsurgical gastroparesis) had improvement in GES after the first G-POEM (88% to 53% 4-hour retention) but then deteriorated to 73% despite symptomatic improvement. After symptom recurrence, redo G-POEM provided an excellent symptomatic response, but she remained total parenteral nutrition-dependent until death 4 years later. Conclusion: While redo G-POEM is technically feasible and safe, our case series demonstrates poor long-term clinical outcomes across different gastroparesis etiologies. All 3 patients experienced treatment failure, with 2 requiring permanent nutritional support and 1 showing a lack of symptomatic response. These findings may suggest limited utility of redo G-POEM and highlight the need for careful patient selection.