Phase III trial of initial chemotherapy in stage III or IV head and neck cancers: a study by the Gruppo di Studio sui Tumori della Testa e del Collo.

BACKGROUND: The standard treatment for advanced (stage III and IV) head and neck squamous cell carcinoma (i.e., surgery with postoperative radiotherapy in operable patients and radiotherapy alone in inoperable patients) has had poor results. A series of randomized trials of induction chemotherapy have up to now failed to demonstrate an improvement in survival.

PURPOSE: This trial was designed to determine whether intensive induction chemotherapy administered before loco-regional treatment would improve survival of patients with advanced disease.

METHODS: Patients had previously untreated, advanced nonmetastatic (stages III and IV) squamous cell carcinoma of the oral cavity, oropharynx, hypopharynx, and paranasal sinuses. The study design was a randomized, multi-institutional, phase III trial. Eligible patients (n = 237) were randomly assigned to receive either initial chemotherapy (cisplatin and infusional fluorouracil) followed by loco-regional treatment (group A, n = 118) or loco-regional treatment alone (group B, n = 119). For operable patients (group A, n = 34; group B, n = 32), loco-regional treatment included resection followed by adjuvant radiotherapy. For inoperable patients, radical irradiation was performed with a planned dose of 65-70 Gy to involved areas. A dose of 45-50 Gy was also planned to the uninvolved neck or postoperatively. The statistical (log-rank) test was performed no earlier than 2 years after the randomization of the last patient.

RESULTS: Seventy-one patients (60%) in group A and 67 patients (56%) in group B were considered free of disease after they completed the treatment sequence. The analysis of time to distant metastases showed an advantage for group A patients. (Respective 2- and 3-year values for inoperable patients were 15% and 24% for group A versus 36% and 42% for group B, P = .04; only one operable group A patient had distant metastases after 49 months versus 26% [2 years] and 31% [3 years] for operable group B patients, P = .01.) For inoperable patients, the combined treatment was significantly associated with an increase in complete remission rate (group A, 44%) as compared with radiotherapy alone (group B, 30%) (P = .037). Inoperable patients also benefitted from induction chemotherapy in terms of disease-free survival (49% and 34% for group A versus 28% and 26% for group B; P = .06) and of overall survival (30% and 24% for group A versus 19% and 10% for group B; P = .04).

CONCLUSIONS: When all 237 randomly assigned patients were analyzed, there were no significant differences in the two treatment strategies in loco-regional failure or in disease-free or overall survival, although the development of distant metastases was reduced. For operable patients, the only benefit from neoadjuvant chemotherapy was a significant reduction in the incidence of distant metastases. For inoperable patients, neoadjuvant chemotherapy improved local control, decreased the incidence of distant metastases, and improved the complete remission rate and overall survival.

IMPLICATIONS: Confirmatory studies with effective chemotherapy regimens delivered for an adequate number of cycles are required.