Concomitant chemoradiation or RT/cetuximab versus induction TPF followed by chemoradiation or RT/cetuximab in locally advanced head and neck squamous cell carcinoma: A randomized phase III factorial study.

Abstract

TPS196 Background: Concomitant chemoradiation (CRT) is the standard treatment for locally advanced squamous cell carcinoma (LASCC). RT plus cetuximab is superior to RT alone and it is an alternative treatment to CRT (Bonner, NEJM 2006). The efficacy of induction chemotherapy when added to standard treatment has never been demonstrated, however, induction cisplatin/fluorouracil [PF] plus docetaxel combination (TPF) is superior to PF when added to CRT or RT alone (Posner, NEJM 2005; Vermorken, NEJM 2005). Methods: This ongoing factorial phase III study is the implementation of a previous phase II randomized trial evaluating the efficacy of CRT with or without induction TPF. At the end of the phase II study a Monitoring Committee decided to continue the study as a phase III, as initially planned, based on the efficacy and safety data of the phase II study. This phase III four arm factorial study has now randomized 352 patients (target 420) with LASCC unresectable stage III-IVM0 of oral cavity, oropharynx, hypopharynx, age > 18 years, ECOG PS 0-1, measurable disease by RECIST, adequate organ function. One step randomization is adopted to assign patients to concomitant treatment alone CRT (arm A1) or RT/cetuximab (arm A2) or to 3 cycles of induction TPF followed by CRT (arm B1) or RT/cetuximab (arm B2). Two concomitant PF cycles are administered in the CRT arm. Primary endpoints are the comparison of 3-year overall survival (OS) between induction vs no induction and the incidence of G3-4 in-field toxicity between CRT and RT/cetuximab. During the planned ad-interim analysis, the Independent Data and Safety Monitoring Committee decided to increase the sample size from the initially planned 350 patients to 420 to detect an absolute difference of 12% in 3 years OS in favour of the induction arm. Assuming an hazard ratio of 0.675 this study will have 80% statistical power with a two-sided type I error of 0.05. The total number of 420 patients will also be able to detect a difference of 10%, in grade 3-4 in-field toxicity between CRT vs RT/cetuximab with a power of 80%.