Transthoracic esophagectomy with radical mediastinal and abdominal lymph node dissection and cervical esophagogastrostomy for esophageal carcinoma

Background. Several techniques for esophageal resection have been reported. This study examines the morbidity, mortality, and early survival of patien ts after transthoracic esophagectomy for esophageal carcinoma using current staging techniques and neoadjuvant therapy. The technique includes right t horacotomy, laparotomy, and cervical esophagogastrostomy (total thoracic es ophagectomy) with radical mediastinal and abdominal lymph node dissection. Methods. Three hundred forty-two patients had surgery for esophageal carcin oma between 1989 and 2000 at our institution. Two hundred fifty consecutive patients had esophagectomy using this technique. Kaplan-Meier curves and u nivariate and multivariate analyses were performed by postsurgical patholog ic stage. Results. Median age was 62.7 years (31 to 86 years). Fifty-nine were female . Eighty-one percent (202) had induction chemotherapy (all patients with cl inical T3/4 or N1). Early postoperative complications included recurrent la ryngeal nerve injury (14% [35]), chylothorax (9%, [22]), and leak (8%, [19] ). Median length of stay was 13 days (5 to 330 days). In-hospital or 30-day mortality was 3.6% (9). Overall survival at 3 years was 44%; median surviv al was 25 months, and 3-year survival by posttreatment pathologic stage was : stage 0 (complete response) (n = 60), 56%; stage I (n = 32), 65%; stage I IA (n = 67), 41%; stage IIB (n = 30), 46%; and stage III (n = 49), 17%. Mea n follow-up was 24 months (SEM 1.6, 0 to 138 months). Five patients with tu mor in situ, 6 patients with stage IV disease, and 1 patient who could not be staged (12 pts) were excluded from survival and multivariate calculation s. In univariate and different models of multivariate analysis, age more th an 65 years, posttreatment T3, and nodal involvement were predictive of poo r survival. For univariate analysis, p = 0.002, p = 0.004, p = 0.02, respec tively; for multivariate analysis, p = 0.001, p = 0.003, p = 0.02, respecti vely. Conclusions. Total thoracic esophagectomy with node dissection for esophage al cancer appears to have acceptable morbidity and mortality with encouragi ng survival results in the setting of neoadjuvant therapy. Patients who sho w complete response after induction chemoradiotherapy appear to have improv ed long-term survival. (C) 2001 by The Society of Thoracic Surgeons.