LAUSR

BACKGROUND In Europe, induction chemotherapy (ICT) followed by radiotherapy is preferred to conventional chemoradiotherapy to avoid total laryngectomy in patients with laryngeal/hypopharyngeal cancer. In comparison with conventional radiotherapy, bioradiotherapy with cetuximab significantly improves locoregional control rates (LCRs) and overall survival (OS) without any increase in unmanageable toxicity. METHODS Patients included had untreated non-metastatic stage III-IV laryngeal/hypopharyngeal invasive squamous cell carcinoma. Good responders after three cycles of docetaxel-cisplatin-5-fluorouracil (TPF)-ICT (docetaxel and cisplatin, 75 mg/m2 each on day 1, and 5-fluorouracil, 750 mg/m2/day on days 1-5) every 3 weeks were randomised to receive radiotherapy (70 Gy) with concurrent cisplatin (100 mg/m2/day on days 1, 22 and 43 of radiotherapy) or cetuximab (400 mg/m2 of loading dose, 250 mg/m2/week during radiotherapy). The primary end-point was larynx preservation. The secondary end-points were laryngo-oesophageal dysfunction-free survival (LEDFS), LCR and OS. RESULTS A total of 153 patients were enrolled. Among 126 TPF-ICT responders, 116 were randomised to receive either cisplatin (n = 60) or cetuximab (n = 56). The median follow-up was 77.5 months. Five-year OS rates were 66.6% (95% confidence interval [CI]: 0.54-0.79) versus 66.9% (95% CI: 0.54-0.79) (p = 0.9), respectively. Five-year LCRs were 79.8% (95% CI: 69.5-90.0) versus 67.8% (95% CI: 55.1-80.5%) (p = 0.18). Five-year LEDFS was 62.2% (95% CI: 49.7-74.8%) versus 56.2% (95% CI: 43.0-69.4) (p = 0.38). Late grade III/IV salivary gland and laryngeal toxicity occurred in 10.3% versus 9.8% and 6.8% versus 11.8% of patients receiving cisplatin-radiotherapy versus cetuximab, respectively. CONCLUSIONS No significant difference in LEDFS was observed between the two arms. TPF-ICT followed by conventional chemoradiotherapy or cetuximab was feasible, and long-term toxicity was not statistically different between the two arms. LEDFS appears as a relevant end-point.