Aggressive multimodality therapy for stage III esophageal cancer: A phase I/II study

Background. Stage III advanced locoregional esophageal carcinoma is frequen tly unresectable and inconsistently represented in therapeutic trials of es ophageal cancer. Methods. From 1992 to 1998, 34 of 131 total esophageal cancer patients were designated stage III (16 T3N1, 9 T4N0, 9 T4N1) and medically fit to enter a combined modality protocol with continuous infusion 5-fluorouracil (C15-F U, 300 to 600 mg/m(2)/day), high-dose external beam irradiation (60 Gy), an d interval esophagectomy. Staging before and after induction therapy includ ed computed tomography, endoscopy, and endoscopic ultrasound. Results. Significant toxicity from induction therapy included death (5/34; 14.7%), pneumonitis (5/34; 14.7%), mucositis (13/34; 38%), and hand-foot sy ndrome (3/34; 8.8%). In addition to the five deaths, 11 patients did not pr oceed to operation because of development of esophagorespiratory fistula in 3, distant disease in 2, persistence of T4 stage in 2, progression of como rbidities in 2, and patient refusal in 2. There was a discrepancy between c linical complete response (cCR) at restaging 56% (19/34) and pathologic CR (pCR) noted at the time of operation (8/34; 23.5%). Complete resections wer e possible in 16 of 18 patients explored. Complications in 4 patients inclu ded: death (1), airway injury (1), chylothorax requiring reoperation (1), a nastomotic leak (1), recurrent nerve injury with vocal cord paresis (2), an d ascaris infection (1). Actuarial survival analysis using the Kaplan-Meier method and log-rank testing showed a 36-month survival of 20% for the grou p as a whole and 27% for patients restaged cCR (cCR vs PR, p = 0.0046). Tre atment failure is predominantly distant, with good local control in resecte d patients. NO node status was strongly associated with survival (N0 vs N1 p = 0.0024). There is a trend towards improved survival in the resected gro up (resected 22% vs nonresected 10% at 3 years, p = 0.17). Conclusions. Response rates and survival are commensurate with multiple com pleted phase II and III trials. These are attained at a higher treatment-re lated mortality. T4 patients can be successfully resected in selected patie nts. Even in advanced disease, nodal status is a significant predictor of s urvival. (C) 2000 by The Society of Thoracic Surgeons.