Esophageal fistula sealing: choice of stent, practical management, and cost.

BACKGROUND: Three models of covered metal stents are available to seal esophageal fistulas. METHODS: Stainless steel covered stents were inserted in 5 patients (group I); nitinol covered stents were inserted in 12 patients (group II) with malignant (n = 14) or benign (n = 3) esophageal fistulas. RESULTS: Stent positioning was satisfactory in all cases. Fistula sealing was complete in 1 of 5 (20%) and 12 of 12 (100%) patients of groups I and II, respectively (p < 0.005). Continued esophageal leakage was initially related to the passage of fluids alongside the stent covering (n = 3) and to early stent migration (n = 1). Complications related to stent placement were observed in 2 of 17 (12%) patients and were fatal. During follow-up (mean 153 +/- 143 days), esophageal fistulas relapsed after initial sealing in 5 of 13 (38%) patients. Further treatment (glue or fibrin sealant injection, additional stent insertion) was attempted in 7 cases of persistent or relapsing esophageal fistula, with sealing obtained in 5 of them. The costs per patient and per day free from symptoms due to the esophageal fistula were $106 and $57 in groups I and II, respectively. CONCLUSION: Nitinol covered stents more frequently provided complete esophageal fistula sealing, as compared with stainless steel covered stents. Further treatments tailored to the mechanisms of fistula persistence or relapse often provided sealing.