Esophageal anastomosis medial to preserved azygos vein in esophageal atresia with tracheoesophageal fistula: restoration of normal mediastinal anatomy.

Abstract

We intended to prospectively study the technical feasibility and advantages of esophageal anastomosis medial to the preserved azygos vein in neonates diagnosed with esophageal atresia with tracheoesophageal fistula (EA/TEF). The results were compared to the cases where azygos vein was either not preserved, or the anastomosis was done lateral to the arch of preserved azygos vein. A total of 134 patients with EA/TEF were admitted between January 2007 and July 2008 of which 116 underwent primary repair. Eleven patients with long gap esophageal atresia with or without tracheoesophageal fistula and 7 patients who expired before surgery were excluded. Patients were randomly divided in three groups comparable with respect to the gestational age, age at presentation, sex, birth weight, associated anomalies and the gap between the pouches after mobilization: Group A (azygos vein ligated and divided), Group B (azygos vein preserved with esophageal anastomosis lateral to the vein), and Group C azygos vein preserved with esophageal anastomosis medial to the vein). All the patients were operated by extra-pleural approach. The three groups were compared with respect to operative time and early postoperative complications like pneumonitis, anastomotic leaks and mortality. Odds ratio and Chi square test were used for the statistical analysis. Group A, B and C had 35, 43 and 38 patients respectively. No significant difference was observed in average operative time in the 3 groups. Though incidence of postoperative pneumonitis was higher in group A (28%) as compared to group B (13.95%) and group C (11.62%), it was not statistically significant (p > 0.005). Anastomotic leak occurred in 7 patients in group A (20%), 6 patients in group B (13.95%) and 4 patients (10.52%) in group C (p > 0.005). Group A had 3 major and 4 minor anastomotic leaks; group B had 2 major and 4 minor leaks and group C had 1 major and 3 minor leaks. There were10 deaths in the series- 5 in group A, 3 in group B and 2 in group C (p > 0.005). Patients with major anastomotic leaks in all 3 groups expired after re-exploration. The minor leaks were managed conservatively and all of them healed spontaneously. Severe pneumonitis and septicemia in patients having major associated anomalies also contributed to the mortality. Although esophageal anastomosis medial to the preserved azygos vein restores the normal mediastinal anatomy without technical difficulty or increased operative time, the study could not prove a statistically significant advantage in terms of mortality and postoperative complications.