Anastomotic stricture prediction in patients with esophageal atresia with distal fistula

PURPOSE: To investigate potential early risk factors for anastomotic stricture formation and assess the predictive role of post-operative esophagrams. METHODS: A retrospective study of patients with esophageal atresia with distal fistula (EA/TEF) operated between 2011 and 2020. Fourteen predictive factors were tested for stricture development. Esophagrams were used to calculate early (SI1) and late (SI2) stricture index (SI = anastomosis diameter/upper pouch diameter). RESULTS: Of 185 patients operated for EA/TEF in the 10-year period, 169 patients met the inclusion criteria. Primary anastomosis was performed in 130 patients and delayed anastomosis in 39 patients. Stricture formed in 55 patients (33%) within 1 year from anastomosis. Four risk factors showed strong association with stricture formation in unadjusted models: long gap (p = 0.007), delayed anastomosis (p = 0.042), SI1 (p = 0.013) and SI2 (p < 0.001). A multivariate analysis showed SI1 as significantly predictive of stricture formation (p = 0.035). Cut-off values using a receiver operating characteristic (ROC) curve were 0.275 for SI1 and 0.390 for SI2. The area under the ROC curve demonstrated increasing predictiveness from SI1 (AUC 0.641) to SI2 (AUC 0.877). CONCLUSIONS: This study identified an association between long gap and delayed anastomosis with stricture formation. Early and late stricture indices were predictive of stricture formation.