Transthoracic versus extrathoracic esophagectomy: mortality, morbidity, and long-term survival.

Extrathoracic esophagectomy for carcinoma is an acceptable substitute for transthoracic resection if it can be shown to have comparable or superior safety and no adverse effect on long-term survival. To test this hypothesis, we employed extrathoracic esophagectomy in 30 consecutive patients with carcinoma of the esophagus from January, 1978, to July, 1984. During this period, 65 comparable patients underwent transthoracic resection through a left thoracotomy for lower esophageal lesions or a right thoracotomy and laparotomy for upper thoracic lesions. Only patients with carcinoma limited to the gastric cardia were excluded from the study. Overall morbidity was higher in the extrathoracic than in the transthoracic group (13 of 30 or 43.3% versus 15 of 65 or 23.1%; p = 0.05), but the differences in hospital mortality (4 of 65 or 6.2% for the transthoracic group versus 4 of 30 or 13.3% for the extrathoracic group) and duration of hospital stay (17.4 +/- 11.7 days for the transthoracic group versus 20.5 +/- 13.4 days for the extrathoracic group) were not statistically significant. Considering all patients who either died or sustained a postoperative complication, we found significant differences favoring transthoracic resection in those subgroups of patients who were able to undergo primary reconstruction at the time of resection (12 of 57 or 21.1% versus 15 of 28 or 53.6%; p = 0.004), those with advanced Stage III lesions (11 of 47 or 23.4% versus 12 of 20 or 60%; p = 0.006), those with tumor of the lower esophagus (8 of 35 or 22.9% versus 6 of 10 or 60%; p = 0.04), and those with tumor that could be resected through a left thoracotomy (2 of 18 or 11.1% versus 17 of 30 or 56.7%; p = 0.002). Actuarial survival curves for all transthoracic and extrathoracic resections and separate analysis for Stage I and Stage III tumors revealed no statistically significant differences between these two techniques.