Background/objectives: The dose painting by number (DPBN) technique is a radiotherapy approach guided by imaging biomarkers. This paper investigates the physical/technical aspects addressed in a phase 1-2 study (GLIORAD, NCT04610229) employing the DPBN technique on patients affected by recurrent glioblastoma.
Methods: The DPBN approach was employed on 12 patients, converting the pixel-by-pixel Apparent Diffusion Coefficient map to a prescription dose ranging from 30 to 50 Gy in 5 fractions. The DPBN maps were discretized in a patient-specific number of levels (DPBC), prescribing a different dose to each level. The agreement between the calculated dose distribution (TDD) and DPBC map was assessed. Plan quality and robustness against geometric uncertainties were evaluated. Patient-specific quality assurance was performed before treatment, and the agreement between measured and calculated doses was assessed with the γ metrics.
Results: A good agreement between DPBN and DPBC maps was observed with a median[range] of the quality factor introduced by Vanderstraeten of 0.067[0.038; 0.137]. In addition, the agreement between TDD and DPBC map was of 0.77[0.44; 0.93], indicating satisfactory conversion in deliverable dose distribution. All plans resulted to be deliverable with a median [range] of γ(10%) of 98.55% [88.3; 99.9] and robust against geometric uncertainties with a dose difference to the 99% of the volume < ±5% in 11/12 patients.
Conclusion: The workflow of DPBN approach for recurrent GBM was investigated. The plan quality, robustness, and deliverability were reported and discussed, and the technical feasibility of the voxel-based DPBN approach in a standard clinical setting was demonstrated.
Keywords: Dose painting by numbers; Recurrent glioblastoma multiforme; Voxel-based dose escalation.
Copyright © 2026. Published by Elsevier Ltd.