Session Item

The multicentre randomized phase III Adaptive Radiation Treatment for Head and Neck Cancer (ARTFORCE) trial (NCT01504815) aimed to improve the locoregional control (LRC) in patients with locoregionally advanced head and neck cancer (LAHNSCC) by increasing dose to radioresistant tumor subvolumes using the fluorodeoxyglucose positron emission tomography (FDG-PET). Secondly, toxicity rates were aimed to be similar to conventional radiotherapy by decreasing dose to the primary tumor planning target volume (PTVp) borders and improving treatment accuracy using adaptive radiotherapy.

A sample of 268 was determined to find a LRC benefit of 15%, given a historic LRC of 65% (two-sided α=0.05, β=0.80). After including 221 eligible patients with LAHNSCC from 9 institutes since 2012, the trial closed in 2019 due to slow accrual. Patients were randomized (1:1) to receive either conventional radiotherapy with a homogeneous dose of 70Gy (cRT, n=112) or adaptive and FDG-PET guided dose redistributed radiotherapy (rRT, n=109). In rRT, dose was redistributed by escalating dose (min 70Gy, max 84Gy, mean 77Gy) in the PTVFDG (>50%SUVmax), and decreasing (min 64Gy) dose at PTVp borders. Both schedules were delivered in 35 fractions by simultaneous integrated boost and combined with concurrent three cycle high dose cisplatinum (100mg/m2). Anatomical adaptation was scheduled after the 10th fraction for patients treated by rRT.



LRC, progression-free (PFS) and overall survival (OS) analyses were performed by the intention-to-treat principle using the Kaplan Meier (KM) method. Unadjusted hazard ratios were obtained using univariable Cox proportional hazards regression (CPH). Toxicities were scored using the CTCAE (v4.0). 2-year toxicity prevalence was evaluated by Fisher’s exact test for patients with complete and locoregional failure-free 2-year follow up. KM and CPH methods were used for 2-year late toxicity incidence analysis. The software R (version 4.1.1.) was used for analysis.

Median follow up time was 4 years (IQR 2.4–5.1). 2 patients in both arms prematurely ended radiotherapy. 2-year LRC was not significantly different between trial arms, with control rates of 76.8% and 82.7% (log-rank p=0.426) for the cRT and rRT arm, respectively (fig. 1A). Neither was a significant difference found for 5-year PFS (p=0.626, fig. 1B) or OS (p=0.964, fig. 1C). 2-year toxicity prevalence was evaluable for 78 cRT and 79 rRT patients. Grade ≥2 or grade ≥3 toxicity prevalence (p≥0.120, table 1) or incidence (p=0.898)

Focal dose escalation was successfully delivered without significantly increasing toxicity rates using adaptive and redistribution strategies. However, FDG-PET guided dose redistribution did not significantly improve LRC, PFS and OS in LAHNSCC.