Oesophageal Resection with Gastric Tube Reconstruction - A clinical study of a surgical method and its consequences on the upper gastrointestinal tract and survival

Abstract: In a prospective randomised study following oesophagectomy and gastric tube reconstruction patients with hand-sutured neck anastomoses were compared with patients with circular stapled chest anastomoses. The two groups had equal rates of anastomotic or conduit leaks of 7.3 % (3/41) and 7.1 % (3/42) and of benign anastomotic strictures of 19.5 % (8/41) and 28.6 % (12/42). Another prospective randomised study showed that chest draining in uncomplicated oesophagectomies can be safely managed by a portable lightweight system instead of a stationary one as judged by equal rates of pleural effusions (18 %), atelectases (17 %), and spontaneously resolved pneumothorax (13 %). Following oesophageal resection and replacement with a gastric tube the oesophageal remnant proximal to the anastomoses cleared 70 % of a swallowed bolus within 15 seconds irrespective of the anastomotic site in the neck or in the chest. The transposed stomach had a half-emptying time of one hour, a feature that may suggest backward reflux of the contents of the conduit. Pharyngeal reflux following the gastric reconstruction as determined by pH studies was low (0.2-0.96 %) during the study year, but was more pronounced in patients with neck than with chest anastomoses (p=0.002). The difference may be an effect of the additional surgical trauma and subsequent healing when a cervical approach is used compared with when it is omitted in chest anastomoses. Oesophago-gastric anastomoses dilate during the first postoperative year, however not to the diameters of Roux-en-Y oesophago-jejunal anastomoses after total gastrectomy. Benign anastomotic strictures and narrower anastomoses were more frequent after gastric tube reconstructions (47 dilatations in 107 patients; anastomotic diameter 20-23 mm) than following oesophago-jejunal Roux-en-Y-reconstructions (7 dilatations in 149 patients; anastomotic diameters 26-30.5 mm). With a hospital mortality rate of 1.4 % (2/139) in oesophagectomy patients, the tumour stage was the only significant predictor for long-term survival rates, whereas age above or below 70, or tumour type were not significant predictors. Irrespective of the anastomotic site or of tumour recurrence during the first postoperative year, patients with gastric (n=125), colon (n=10) or jejunal (n=4) reconstructions after oesophagectomy had no dysphagia in approximately 75 %, and severe dysphagia in 5 % or less. Patients with adenocarcinoma at the gastro-oesophageal junction and Barrett's metaplasia presented more frequently in an earlier tumour stage with reflux related symptoms and bleeding than those without Barrett's who had advanced tumour stage and dysphagia.