Save the child's esophagus: Management of major disruption after repair of esophageal atresia

Purpose: Given the bias that the native esophagus is the best conduit between the oropharynx and the stomach, the authors report a ''conservative'' approach to massive esophageal leak, which may be considered ''radical'' by others. Major disruption of the anastomosis after primary repair of esophageal atresia is a recognized and feared complication. Historically, management has been the performance of cervical esophagostomy and gastrostomy. The aim of this report is to describe the authors' approach to this difficult and serious complication. Methods: A 15-year retrospective analysis was performed of all patients having esophageal atresia. Data collection focused on the management of ail patients with clinically significant esophageal disruption. Radiographically detected (clinically asymptomatic) leaks were managed by continuation of drainage by thoracostomy tubes already in place and are not included. Reoperative thoracotomies were performed, which included primary repair (2), placement of pleural patch alone (2), pleural patch with intercostal muscle flap buttress (2), and operative debridement and drainage alone (1). Results: It was noted that seven patients had clinically significant esophageal disruption requiring reoperation, with circumferential disruptions ranging from 15% to 85%. Presentation included persistent pleural collection (4) and pneumothorax (3). Both patients who underwent primary repair had no evidence of leakage on follow-up esophagograms, neither did one with a pleural patch alone and one with an intercostal muscle flap. Five of the seven patients were tolerating oral feedings at the time of follow-up (range, 6 months to 8 years). One of the two others (both currently inpatients), has a recurrent leak associated with mediastinitis, and the other (who had primary repair) has a presumed neurological impairment of eating. Conclusion: Clinically significant disruption of primary esophageal repair should not warrant a cervical esophagostomy and placement of a gastrostomy tube, thus precluding eventual use of the native esophagus. The authors have shown that management by reoperation with primary repair, intercostal muscle flap with or without pleural patch, and/or drainage allows the patient to maintain the native esophagus and yields a generally good outcome after a prolonged healing time. Copyright (C) 1996 by W.B. Saunders Company