Purpose: A randomized trial was designed to compare cisplatin (CDDP) and fluorouracil (FU) versus carboplatin (CBDCA) and FU as neoadjuvant treatment in stage IV-M0 head and neck cancer to assess whether CBDCA-FU is better than CDDP-FU with regard to response and toxicity.
Patients and methods: Patients were randomized to receive CDDP 100 mg/m2 intravenously on day 1 and FU 5,000 mg/m2 over a 120-hour continuous infusion, or CBDCA 400 mg/m2 over a 24-hour continuous infusion on day 1 and FU with the same schedule. Both regimens were repeated every 21 days. The patients received three courses of chemotherapy, excluding those who failed to achieve a partial response (PR) after the second course. Complete responders were treated with radiotherapy. The remaining patients underwent surgery if the tumor was resectable.
Results: Interim analysis was performed when 95 patients were included. The trial was stopped due to significantly better results in the control arm. Differences in response (P = .04) were favorable to CDDP-FU. Hematologic toxicity predominated in the CBDCA-FU arm (P < .001). Mucositis and vomiting predominated in the CDDP-FU arm (P = .03, P < .001, respectively). Favorable outcomes (complete response [CR] plus any grade of toxicity and PR plus grade 0 to 3 toxicity) predominated in the CDDP-FU arm (P = .02). Only the treatment assigned was associated with response (P = .02) and favorable outcomes (P = .009) in the logistic regression analysis. In the CDDP-FU arm, disease-free and overall survival were significantly better. Cox regression analysis showed that only treatment association with disease-free survival remains significant.
Conclusion: Our results indicate that CDDP-FU is more effective than CBDCA-FU as neoadjuvant treatment in stage IV-M0 head and neck cancer.