Cervical or thoracic anastomosis for esophagectomy for carcinoma

Abstract

A prospective trial was conducted to compare intrathoracic and cervical anastomoses after esophagectomy for squamous cell carcinoma of the middle or lower one-third of the esophagus. One hundred and twenty-three patients were randomized to have either a cervical or thoracic anastomosis. Thirty-one patients were subsequently excluded either because esophagectomy was performed without thoracotomy or the tumor was unresectable or because the randomization protocol was not complied with. Transfusion requirements and operating time were similar for the 49 patients having esophagectomy by way of the laparotomy and right thoracotomy (TA) and the 43 patients who underwent laparotomy, right thoracotomy and cervicotomy (CA). Forty-three per cent of the CA and 49 per cent of the TA patients had involved lymph nodes. An esophagectomy incorporating a cervical anastomosis resulted in a significantly greater margin of macroscopically normal esophagus above the tumor (median of 4.0 versus 1.5 centimeters for TA). A leak was significantly more frequent after cervical anastomosis (26 per cent) than thoracic (4 per cent) (p less than 0.002). Respiratory complications were more frequent with a thoracic anastomosis, but this was not statistically significant. Thirty day mortality rates were similar for the two groups: 14.3 per cent, TA, and 9.3 per cent, CA (p = N.S.). Postoperative strictures occurred in 14 per cent of TA and 23 per cent of CA patients and were most common after an anastomotic leak. The survival patterns of the two groups were similar. The median survival time for CA patients was 23 months and for TA, 20 months. Excluding hospital mortality, 47 per cent of patients were alive at two years and 30 per cent at 40 months. Survival was related to extent of disease. The greater length of tumor-free esophagus removed with a cervical anastomosis did not result in an improved long term survival period, but was associated with a significantly higher incidence of anastomotic fistula.