Mature survival results with preoperative cisplatin, protracted infusion 5-fluorouracil, and 44-Gy radiotherapy for esophageal cancer.

PURPOSE: To assess the long-term survival results after cisplatin, protracted infusion 5-fluorouracil, and concurrent radiotherapy (RT) followed by surgical resection of esophageal cancer. METHODS AND MATERIALS: Ninety-two patients with esophageal cancer (65 with adenocarcinoma and 27 with squamous cell carcinoma) were treated in two sequential protocols of preoperative chemoradiotherapy. The patients had tumor confined to the esophagus and regional nodes, including celiac nodes for middle and distal lesions. In trial A (1989-1994), 50 patients were treated with 44 Gy RT (2 Gy/d) along with concurrent 5-fluorouracil 300 mg/m(2)/d given by protracted venous infusion on Days 1-30 and cisplatin 26 mg/m(2) on Days 1-5 and 26-30. In trial B (1995-1997, 42 patients), the chemotherapy dosages during RT were reduced to 5-fluorouracil 225 mg/m(2)/d protracted venous infusion and cisplatin 20 mg/m(2)/d on Days 1-5 and 16-30; three cycles of paclitaxel 135 mg/m(2)and cisplatin 75 mg/m(2) were given postoperatively. Surgery generally occurred 4-6 weeks after completion of the planned preoperative therapy. Transhiatal resection was performed whenever possible.
RESULTS: Of the 92 patients, 86 (93%) underwent surgery (1 refused, 2 died preoperatively, and 3 developed evidence of metastatic disease). Of the 92 patients, 80 (87%) had complete resections with negative margins (3 had positive margins and 3 had distant metastases discovered at surgery). The pathologic complete response rate was 33% (30 of 92). The median follow-up was 63.5 months. The median survival and disease-specific survival for all enrolled patients was 35 and 59 months, respectively. The 5-year survival and disease-specific survival rate was 40% and 49%, respectively. Patients with a pathologic complete response had a 67% survival rate at 5 years (median not reached), and the remainder of patients had a 5-year survival rate of 27% (median 21 months; p <0.001). For 21 patients alive after 5 years (60-121 months), 2 died of their disease and all others were disease free. Eight patients with pathologic Stage I tumor at the time of surgery had survival similar to those with a complete response to preoperative therapy. The median survival for patients with pathologic Stage IIA, IIB, III, and IV disease at the time of surgery was 22, 13.5, 18, and 4.9 months, respectively. The pattern of initial failure was local/regional alone in 6% (5 of 90), local/regional plus distant in 3% (3 of 90), and distant alone in 47% (42 of 90). No differences were noted in survival or response rate between those with adenocarcinoma or squamous cell carcinoma.
CONCLUSION: The promising 5-year survival results and low rate of late cancer-related deaths suggest that these regimens of intensive neoadjuvant therapy may improve the overall cure rate. The pathologic stage after neoadjuvant therapy is an important predictor of survival and may be useful in selecting patients for novel adjuvant therapies. Isolated local failure is uncommon, indicating that efforts to improve the therapeutic outcome should focus on optimizing systemic therapy rather than intensifying the RT. Additional randomized data are needed to assess the benefits of this therapeutic approach fully.