Anastomotic leakage after esophagectomy for cancer: a mortality-free experience.

Leakage is a serious complication of esophagectomy and is historically associated with high mortality. This study aimed to describe the morphology and strategies for clinical management of leakage after esophagectomy. A database prospectively maintained from July 2002 to July 2005 at a referral unit for foregut cancer was used to identify patients with leakage of saliva or gastrointestinal contents after esophagectomy and reconstruction with stomach. Contrast swallow was routinely performed on postoperative day 7. Leakage was diagnosed and classified by well-defined criteria. There were 99 men and 27 women, yielding an institutional volume of 42 esophagectomies per year. There was no in-hospital mortality from any cause. Actual 1-year survival was 87%. An Ivor Lewis operation was performed on 103 patients (82%); 4 patients had leakage within 5 days of operation and all had immediate rethoracotomy. An additional 8 patients with Ivor Lewis operation had leakage after day 5, and this was detected by contrast swallow in only 3 patients; 2 patients had no intervention, 4 patients had radiology-guided drainage, 1 had thoracoscopy, and 1 had rethoracotomy. Leakage was from the actual esophagogastric anastomosis in eight patients, from the linear gastric staple line in three patients, or from gastric necrosis in one patient. Twenty-three patients had a transhiatal or three-stage operation; leakage was from the actual anastomosis in five patients or gastric necrosis in one patient. After Ivor Lewis esophagectomy, leakage was from the actual anastomosis in two-thirds of patients or from the gastric conduit in the remaining one-third. Prompt reoperation is recommended for early postoperative leakage. Most patients with leakage after day 5 can be treated nonoperatively.