Between 1990 and 2000, we examined the effect of timing of non-platinum chemotherapy when combined with radiotherapy. We aimed to determine whether giving chemotherapy concurrently with radiotherapy or as maintenance therapy, or both, affected clinical outcome. Here we report survival and recurrence after 10 years of follow-up.

Between Jan 15, 1990, and June 20, 2000, 966 patients were recruited from 34 centres in the UK and two centres from Malta and Turkey. Patients with locally advanced head and neck cancer, and who had not previously undergone surgery, were randomly assigned to one of four groups in a 3:2:2:2 ratio, stratified by centre and chemotherapy regimen: radical radiotherapy alone (n=233); radiotherapy with two courses of chemotherapy given simultaneously on days 1 and 14 of radiotherapy (SIM alone; n=166); or 14 and 28 days after completing radiotherapy (SUB alone, n=160); or both (SIM+SUB; n=154). Chemotherapy was either methotrexate alone, or vincristine, bleomycin, methotrexate, and fluorouracil. Patients who had previously undergone radical surgery to remove their tumour were only randomised to radiotherapy alone (n=135) or SIM alone (n=118), in a 3:2 ratio. The primary endpoints were overall survival (from randomisation), and event-free survival (EFS; recurrence, new tumour, or death; whichever occurred first) among patients who were disease-free 6 months after randomisation. Analyses were by intention to treat. This trial is registered at www.Clinicaltrials.gov, number NCT00002476.

All 966 patients were included in the analyses. Among patients who did not undergo surgery, the median overall survival was 2.6 years (99% CI 1.9-4.2) in the radiotherapy alone group, 4.7 (2.6-7.8) years in the SIM alone group, 2.3 (1.6-3.5) years in the SUB alone group, and 2.7 (1.6-4.7) years in the SIM+SUB group (p=0.10). The corresponding median EFS were 1.0 (0.7-1.4), 2.2 (1.1-6.0), 1.0 (0.6-1.5), and 1.0 (0.6-2.0) years (p=0.005), respectively. For every 100 patients given SIM alone, there are 11 fewer EFS events (99% CI 1-21), compared with 100 given radiotherapy, 10 years after treatment. Among the patients who had previously undergone surgery, median overall survival was 5.0 (99% CI 1.8-8.0) and 4.6 (2.2-7.6) years in the radiotherapy alone and SIM alone groups (p=0.70), respectively, with corresponding median EFS of 3.7 (99% CI 1.1-5.9) and 3.0 (1.2-5.6) years (p=0.85), respectively. The percentage of patients who had a significant toxicity during treatment were: 11% (radiotherapy alone, n=25), 28% (SIM alone, n=47), 12% (SUB alone, n=19), and 36% (SIM+SUB, n=55) among patients without previous surgery; and 9% (radiotherapy alone, n=12) and 20% (SIM alone, n=24) among those who had undergone previous surgery. The most common toxicity during treatment was mucositis. The percentage of patients who had a significant toxicity at least 6 months after randomisation were: 6% (radiotherapy alone, n=13), 6% (SIM alone, n=10), 4% (SUB alone, n=7), and 6% (SIM+SUB, n=9) among patients who had no previous surgery; and 7% (radiotherapy alone, n=10) and 11% (SIM alone, n=13) among those who had undergone previous surgery. The most common toxicity 6 months after treatment was xerostomia, but this occurred in 3% or less of patients in each group.

Concurrent non-platinum chemoradiotherapy reduces recurrences, new tumours, and deaths in patients who have not undergone previous surgery, even 10 years after starting treatment. Chemotherapy given after radiotherapy (with or without concurrent chemotherapy) is ineffective. Patients who have undergone previous surgery for head and neck cancer do not benefit from non-platinum chemotherapy.

Cancer Research UK, with support from University College London and University College London Hospital Comprehensive Biomedical Research Centre.