Background: Although intrathoracic leakage is a major complication of oesophagectomy, precise data concerning diagnostic features and results of conservative treatment are lacking.
Methods: From 1986 to 1994, 409 oesophagectomies with stapled oesophagogastrostomy were performed, including 358 Lewis-Tanner and 51 Sweet procedures. A water-soluble contrast swallow was routinely performed on day 7 or later, before oral intake was begun. All patients except one received conservative non-surgical treatment, including nutritional support and perianastomotic drainage.
Results: Leaks were diagnosed in 38 patients (9.3 per cent). The leakage rate was 7.8 per cent after the Lewis-Tanner procedure and 20 per cent after the Sweet procedure (P < 0.01). Eleven patients had no symptoms. Seven of the 27 patients with symptoms had a contrast swallow that was normal, and subsequently developed a confirmed fistula after the onset of oral intake. Five patients had to undergo reoperation. All asymptomatic patients and 18 symptomatic patients recovered. Nine patients died, mainly from multiple organ failure, including three who had reoperation for resection of the gastroplasty.
Conclusion: The potential presence of clinically silent fistula and the deleterious role of oral intake still justify routine detection of leakage after oesophageal resection. Conservative treatment results in survival with preservation of the gastroplasty in most patients, unless multiple organ failure occurs.