Induction chemotherapy in primary resectable head and neck tumors - a prospective randomized trial

Although induction chemotherapy prior to local therapy produces encouraging initial response rates in head and neck cancer, randomized studies have failed to demonstrate an advantage in survival. All randomized studies included only patients with far advanced stage III and IV disease which appears to be the main reason for the low rate of complete responses (max. 18%) in these trials. According to statistical considerations nearly 50% complete responders are necessary before improved survival benefit can be expected. Until now, such complete response rates are only achieved with induction chemotherapy in resectable T2-T3, N0-N2 disease. Therefore, we started a prospective randomized trial including only patients with these stages of disease. Patients were stratified by primary tumor site and neck node status and were randomized to receive either induction chemotherapy with three cycles of carboplatin/5-FU prior to surgery and radiotherapy (arm A, 49 patients) or standard treatment with surgery and radiotherapy (arm B, 47 patients). Patients were stratified by primary tumor site and neck disease. After a follow-up of 12-48 months overall survival was 72% in arm A and 53% in arm B (n.s.). Considering only the patients with cancer of the oral cavity and the tonsil overall survival was 88% in arm A and 44% in arm B (p<0.05). As of today, the number of patients with carcinomas of the hypopharynx and base of tongue is too small for a statistically significant statement, but preliminary data indicate a better overall and disease-free survival without chemotherapy in these patients. Therefore, we recommend controlled trials with induction chemotherapy in patients with primary resectable carcinomas of the oral cavity and the tonsil, stages T2-T3 and N0-N2, prior to surgery but not in patients with cancer of the hypopharynx and base of the tongue.