Acute and long-term toxicity profiles and final oncologic outcomes in the phase III ARTFORCE trial

Background and purpose: We previously published the two-year outcomes of a multicenter randomized phase III trial evaluating whether FDG-PET-guided adaptive dose-redistributed radiotherapy (rRT) could improve locoregional control (LRC) in locally advanced head and neck squamous cell carcinoma (LAHNSCC) without increasing toxicity. This paper reports on acute and five-year toxicity as well as five-year oncologic outcomes. Materials and Methods: Patients with T3-4, N0-3, M0 LAHNSCC were randomized (1:1) to rRT (64-84 Gy in 35 fractions with adaptation at fraction 10) or conventional radiotherapy (cRT; 70 Gy in 35 fractions), both with concurrent three-cycle 100 mg/m2 cisplatin. Results: The trial closed at 84% of planned accrual (221 eligible; rRT = 109, cRT = 112). Acute toxicity rates were similar between arms (P >= 0.10). At five years, LRC remained consistent with the two-year results (HR 0.78; 95% CI 0.45-1.36; P = 0.39). Late toxicity results were also consistent with the two-year results, showing higher grade >= 3 pharyngolaryngeal stenosis in the rRT arm (0% vs 4%, P = 0.05). No additional differences in toxicities were found. Most toxicities stabilized after two years, but hypothyroidism and fibrosis increased throughout the five-year follow up. Exploratory subgroup analyses show trends favoring rRT in patients with oropharyngeal cancer and N0-1 disease. No additional subgroups were identified that might be benefitted by rRT. Conclusion: FDG-PET-guided adaptive dose-redistributed radiotherapy resulted in similar acute and long-term toxicity, tumor control and survival as cRT. These findings confirm the previously reported two-year results and support the long-term safety of rRT.