Abstract Purpose Boron neutron capture therapy (BNCT) dose planning conventionally uses a tumor-to-blood (T/B) ratio derived from the tumor maximum standardized uptake value (SUVmax) of fluorine-18-labeled 4-borono-L-phenylalanine ( 1 ⁸F-BPA) positron emission tomography (PET) (individual model). It remains unclear whether tumor SUVmax or individual model doses predict outcomes in accelerator-based BNCT for head and neck cancer (HNC). We investigated the correlation of 1 ⁸F-BPA PET SUVmax and individual model doses with treatment response. Materials and methods We retrospectively analyzed 30 patients with HNC treated with accelerator-based BNCT (2020–2021). The Kaplan–Meier method was used for survival analysis. Dose parameters (Dmax, Dmin, and D80%) were calculated using the uniform model (T/B = 2.5) and the individual model (T/B = Tumor SUVmax/Blood-pool SUV). Correlations between tumor SUVmax or dose parameters and best treatment response complete response (CR) vs. non-CR were evaluated using the Mann–Whitney U test. Results The objective response rate was 90%, and the CR rate was 50%. CR status strongly predicted superior 2 years overall survival, locoregional control, and progression-free survival (all p 0.05). Conclusion 1 ⁸F-BPA PET SUVmax and the conventional individual model did not demonstrate a statistically significant association with CR in patients with HNC after accelerator-based BNCT. These findings question the utility of tumor SUVmax-based dosimetry, suggesting that current models require refinement.
Yoshino et al. (Tue,) studied this question.