Predictors of anti-reflux procedure failure in complex esophageal atresia patients

Background: Anti-reflux procedures (ARP) in esophageal atresia (EA) patients can be challenging and prone to failure. These challenges become more evident with increasing complexity of EA. We sought to determine predictors of ARP failure in complex EA patients. Methods: Single-institution retrospective review of complex EA patients (e.g. long-gap EA, esophageal strictures, hiatal hernia, and reoperative ARP) who underwent an ARP from 2002 to 2019. ARP failure was defined as hiatal hernia recurrence, wrap migration/loosening, or need for reoperation. Predictors of failure were evaluated using univariate and multivariable time-to-event analysis. Results: 121 patients underwent 140 ARP at a median age of 13.5 months (IQR 7, 26.5). Nissen fundoplication (89%) was the most common ARP. Mesh (bovine pericardium) reinforcement was used in 41% of the patients. Median follow-up was 3.2 years (IQR 0.9, 5.8); 44 instances of ARP failure occurred (31%), though only 20 (14%) required reoperation. Median time to failure was 8.7 months (IQR 3.2, 25). Though fewer mesh-reinforced ARP failed (21% with vs 39% without, p = 0.02), on multivariable analysis only partial fundoplication (aHR 2.22 [95% CI 1.01–4.78]) and minimally invasive repair (aHR 2.57 [95% CI 1.12–6.01]) were significant predictors of ARP failure. Conclusion: In our practice of complex EA patients, where ARP fail in nearly one third of cases, a Nissen fundoplication performed via laparotomy provided the lowest risk of ARP failure.