Prospective medical and surgical treatment of gastroesophageal reflux in esophageal atresia.

BACKGROUND

Gastroesophageal reflux is a major cause of anastomotic complications after repair of esophageal atresia. For this reason, we evaluated a prospective, postoperative treatment protocol with the emphasis on comparing medical and operative treatment.

STUDY DESIGN

From 1994 to 1995, 26 consecutive patients underwent correction of esophageal atresia in the Sophia Children's Hospital. These patients were enrolled in a decision-making protocol meant to establish the effect of medical treatment on gastroesophageal reflux and stricture formation, the relation between anastomotic tension and reflux, and the question of whether performing a Nissen fundoplication was justified. Patients who showed reflux on the first postoperative x-ray were given medical treatment. Reflux was assessed after 6-12 weeks by x-ray and 48-hour pH-metry (24 hours with and 24 hours without medication). Evaluations were repeated at 18 weeks, 6 months, and 1 year. Twenty-three patients were followed for > or = 1 year.

RESULTS

Twenty-four patients had classic esophageal atresia combined with tracheoesophageal fistula. Two had isolated atresia and underwent a colonic interposition. One of the others died of severe cerebral hemorrhage early after the operation. Twenty-two of the remaining 23 showed reflux on the first postoperative x-ray and were given medical treatment. The mean 3.8% total mild reflux time (range, 0.0-11.0%) decreased to a mean of 1.47% (range, 0.0-6.8%). Medical treatment given according to protocol did not influence severe reflux. Eleven of 23 patients showed stricture formation, requiring a mean of four dilatation procedures (range, 1-9). Defining a real stricture as one needing three or more dilatations, as seen in seven patients, the following results were seen: four nonrefluxing patients (proved by x-ray and pH-metry) needed a mean of 4.2 dilatations (range, 3-7), and three refluxing patients (proved by x-ray and pH-metry) needed a mean of 7.3 dilatations (range, 5-9). Three of seven patients with anastomotic tension had proved gastroesophageal reflux; reflux was also diagnosed in 8 of 15 patients without any tension on the anastomosis. Nine of 23 patients underwent a Nissen fundoplication according to the protocol. In four of them, this was decided because of severe reflux-associated respiratory problems; in one, for resistant stenosis after a Livaditis procedure; and in one with normal pH-metry, the procedure was done on clinical grounds. The latter patient needed an aortopexy at a later stage. A late fundoplication was performed in two patients for persistent gastroesophageal reflux unresponsive to medical treatment, and in one for persistent stenosis and reflux. In all patients, the outcomes were successful, without complications.

CONCLUSIONS

Medical treatment of gastroesophageal reflux after repair of esophageal atresia has a distinct effect on the duration of reflux and could have a positive effect on the occurrence and treatment of stenosis. There is no clear relation between the occurrence of reflux and tension on the anastomosis. Nissen fundoplication according to the protocol was done appropriately in eight of nine patients.