Background: The results of a Phase II study of concurrent chemotherapy and accelerated fractionation radiation therapy followed by surgical resection for patients with both adenocarcinoma and squamous cell carcinoma of the esophagus are presented. Pretreatment and postinduction staging were correlated with pathologic findings at surgery to assess the role of surgical resection and the predictive value of noninvasive staging techniques.
Methods: Patients received 2 induction courses with 4-day continuous intravenous infusions of cisplatin (20 mg/m2/day) and 5-fluorouracil (1000 mg/m2/day) beginning on Day 1 and Day 21, concurrent with a split course of accelerated fractionation radiation (1.5 grays [Gy] twice daily, to a total dose of 45 Gy). All patients were subsequently referred for surgical resection. A single, identical postoperative course of chemotherapy and 24 Gy accelerated fractionation radiation was planned for patients with residual tumor at surgery.
Results: Seventy-four patients were entered on this study; 72 patients were considered eligible and evaluable. Induction toxicity included nausea (85%), increased dysphagia (90%), neutropenia (<1000/mm3) (43%), thrombocytopenia (<20,000/mm3) (10%), and reversible nephrotoxicity (8%). Sixty-seven patients (93%) underwent surgery, and 65 (90%) were found to have resectable tumors. Twelve of these patients (18%) died perioperatively, and 18 (27%) had no residual pathologic evidence of disease. Resolution of symptoms and normalization of radiographic studies, endoscopy, or esophageal ultrasound did not identify pathologic complete responders accurately. No patient completing induction therapy and surgery experienced a locoregional recurrence. The Kaplan-Meier 4-year projected recurrence free and overall survival rates were 49% and 44%, respectively.
Conclusions: Although this regimen is feasible, there was significant preoperative toxicity and perioperative mortality. Nonetheless, the recurrence free and overall survival rates were encouraging. However, no staging tool can predict a pathologic complete response after induction therapy accurately, suggesting a continued need for surgical resection.