A multicenter prospective randomized controlled trial was performed to investigate whether dose reduction to the elective nodal volume (PTVelect) in head and neck carcinoma reduces radiation-induced dysphagia, primary endpoint, without compromising tumor control, secondary endpoint. Here, we report on the long-term follow-up of the secondary endpoint (NCT01812486).

Two hundred patients treated with primary (chemo)radiotherapy (RT) were randomized (1:1) between the standard arm, irradiation to PTVelect up to an equivalent dose (EQD2) of 50 Gy and the experimental arm, irradiation to PTVelect up to EQD2 of 40 Gy. The primary tumor and involved nodes were treated according to the standard of care, EQD2 70 Gy (PTVhigh). Regional recurrences (RR) were projected on the initial RT planning-CT to identify the recurrence localization.

The 5-year (5Y) RR was 14.0% (CI95% 7.9; 21.8) in the 40 Gy arm versus 7.5% (CI95% 3.3; 14.0) in the 50 Gy arm (p = 0.10). Majority of RR in the 40 Gy arm (9/13) were projected in PTVhigh and 2 RR were seen outside the treated RT volume. Only 2 RR occurred in PTVelect irradiated up to 40 Gy which was the same number as RR occurring in the 50 Gy PTVelect. The 5Y-overall survival (OS) was 56.5% (CI95% 45.7; 65.9) in the 40 Gy arm versus 49.6% (CI95% 39.0; 59.2) in the 50 Gy arm (p = 0.56).

At 5-years, no statistically significant differences regarding OS, local recurrence, RR nor distant metastases were observed between both treatment arms. This study is underpowered to undoubtedly demonstrate non-inferiority. However, since in both arms only two RR in the PTVelect were observed, reducing the dose to PTVelect appears safe and should be further investigated.