Transhiatal esophagectomy without thoracotomy

Abstract

Since 1976, my associates and I have performed approximately 1,950 transhiatal esophagectomies without thoracotomy at the University of Michigan, nearly 80% for carcinoma and 20% for benign disease. The patients have ranged in age from 14 to 92 years (average 62 years), one quarter of the patients being 71 years of age or older. The stomach was used as the esophageal substitute in 97% of these patients. Colon was required to replace the esophagus only in those with a history of a prior gastric resection for peptic ulcer disease or a caustic gastric injury resulting in scarring and a contracted stomach. Categorically, the properly mobilized stomach will virtually always reach to the neck for a cervical esophagogastric anastomosis. There have been 4 intraoperative deaths (<1%) due to uncontrollable mediastinal bleeding during esophageal mobilization. Additional intraoperative complications included the need for a chest tube(s) because of entry into one or both pleural cavities in approximately 75% of patients; a splenectomy (3%); membranous tracheal laceration (<1%); and violation of the gastric or duodenal mucosa during performance of the pyloromyotomy (<2%), managed successfully in all cases by repair with interrupted 5-0 polypropylene sutures and a buttress of the repair with adjacent omentum. Postoperatively, 6 patients (<1%) have required a thoracotomy for control of mediastinal bleeding occurring within 24 hours of THE. Recurrent laryngeal nerve injury occurred in <5% of patients overall, in less than 2% during the past 5 years. The importance of avoiding the placement of a metal retractor against the tracheoesophageal groove during any part of the cervical dissection and construction of theanastomosis is clear. Chylothorax has occurred in <1% of patients and has been successfully managed with an aggressive policy of early ligation of the injured thoracic duct. In this large series of esophagectomies, <2% of patients have experienced clinically significant atelectasis or pneumonia, a testimony to improved preoperative preparation (no smoking, walking, and use of an incentive inspirometer), less impairment of pulmonary function with an upper midline abdominal incision than with a combined thoracic and abdominal approach, and use of epidural anesthesia to manage postoperative pain. The overall cervical esophagogastric anastomotic leak rate of 13% in our first 1,000 patients treated with THE has now been reduced to <4% since initiation of the side-to-side stapled CEGA, which is illustrated in this article. Gastric tip necrosis necessitating takedown of the CEGA and construction of a cervical esophagostomy has occurred in <1% of patients. The overall hospital mortality rate has been <4%, and 82% have experienced no postoperative complications. In general, the functional results of esophageal replacement with stomach and CEGA have been good, approximately 80% reporting absolutely no symptoms related to eating, or minimal occasional dysphagia, regurgitation, or postvagotomy "dumping" symptoms requiring no treatment; 17%, a "fair" result, requiring an occasional dilatation of a cervical esophagogastric anastomotic stricture or periodic need for medication to control reflux symptoms or dumping; and 3%, a "poor" result, requiring regular dilatation for a severe CEGA stricture or chronic use of medication to control reflux or intractable dumping symptoms. For patients with esophageal cancer treated with a THE and CEGA, the overall 2-year survival rate has been approximately 48% and the 5-year survival rate 24%. Survival rates for a group of 49 of our patients receiving neoadjuvant chemotherapy and radiation therapy and being complete responders (T0N0) on final pathology were 86% at 2 years and 48% at 5 years. We continue to advocate neoadjuvant chemoradiation therapy before THE in patients with stage II and III esophageal cancer who are less than 75 years of age and physically able to tolerate this treatment. THE and a CEGA is an effective means of managing the patient with both benign and malignant esophageal disease requiring resection. The advantages of THE over transthoracic esophagectomy have been demonstrated in a recent meta-analysis of 7,527 patients. A successful THE requires rather rigid adherence to a series of technical steps outlined in this article. Our patients typically have epidural catheters for postoperative analgesia and are extubated in the operating room after a 3 to 5-hour operation and a chest radiograph that confirms there is no unrecognized hemo- or pneumothorax. They do not go to the intensive care unit but rather directly to our thoracic surgery general care floor, where immediate use of their incentive inspirometer used preoperatively is resumed, and ambulation the next day is begun. The nasogastric tube is discontinued on the third postoperative day, oral liquids are begun on day 4, and diet is progressively advanced to a soft diet by day 7. Jejunostomy tube feeding is begun on postoperative day 3 and is gradually tapered as oral intake improves. A routine barium swallow examination is obtained on day 7 to document (1) integrity of the anastomosis, (2) adequacy of gastric emptying, and (3) any evidence of partial small bowel obstruction at the jejunostomy tube site. Patients are typically discharged from the hospital on the seventh or eighth postoperative day. © 2005 Elsevier Inc. All rights reserved.