Abstract
OBJECTIVES: Neo-oesophageal-airway fistula (NEAF) between gastric conduit and airway is a rare but life-threatening complication of oesophagectomy for oesophageal cancer. Optimal treatment remains unknown. A meta-analysis of good-quality case series may help determine whether nonoperative management (NOM) only, upfront surgery (S), or NOM followed by surgery is associated with better 1-year post-treatment mortality, resumption of oral diet and fistula recurrence. METHODS: We systematically searched PubMed, EMBASE and Web of Science for publications in English reporting case series of management and survival in patients with NEAF. Of the 177 identified studies, 62 were duplicates and 95 were not relevant to our topic. Three studies were excluded after a full-text review, due to absence of reporting of 1-year survival. Exclusion criteria to identified publications were: abstract only, malignant NEAF, absence of oesophagectomy and esogastric anastomosis, fewer than 5 patients and NEAF not the main focus of the study. Data-extraction was conducted in accordance with MOOSE guidelines. Data were pooled using random-effects model. RESULTS: Seventeen studies (302 patients) were included. One-year post-treatment mortality was considerably lower with NOM followed by surgery [33%; 95% confidence interval (CI), 0.17-0.48] than with NOM (68%; 95% CI, 0.39-0.97) or S (67%; 95% CI, 0.36-0.98). Fistula location was not associated with 1-year mortality. Neither resumption of an oral diet nor fistula recurrence differed significantly across treatment strategies. CONCLUSIONS: NOM to prepare patients for surgery followed by surgical repair may provide the highest 1-year survival of patients with NEAF. However, patient selection criteria to each of 3 treatment strategies may have affected our findings.