Brain Dose-Volume Thresholds and Survival in Dogs With Intracranial Tumours Treated With the 10 × 4 Gy Radiotherapy...

Dogs with intracranial tumours routinely receive radiotherapy, yet species-specific dose-volume constraints for normal brain tissue remain undefined. In human radiation oncology, exceeding certain brain dose-volume thresholds markedly increases the risk of radiation-induced injury (e.g., radionecrosis). Current veterinary practice often extrapolates human guidelines without validation in discrete species, creating a gap in evidence-based planning. This study aimed to identify brain dose-volume thresholds associated with overall survival (OS) in canine brain-tumour patients. We pooled data from two prospective randomised trials (n = 105 dogs) treated with 10 daily fractions of 4 Gy (total 40 Gy) for intracranial tumours at a single institution. Semi-automated scripting extracted multiple dose-volume metrics, including generalised equivalent uniform dose (gEUD), for the whole brain and brain minus gross tumour volume (Brain-GTV). An iterative Kaplan-Meier and Cox proportional hazards approach identified optimal dosimetric cutoffs, which were then adjusted for tumour volume and body weight via a regression residual method. A brain-volume-adjusted gEUD threshold was also derived to account for variation in brain size. Exposure to normal brain to doses around 30-40 Gy emerged as the strongest predictor of OS. Brain-GTV V32 Gy ≤ 13 cm³ was associated with longer OS (covariate-adjusted cutoff 13.4 cm³, HR = 1.74; p = 0.022, unadjusted optimal split 11.5 cm³, HR = 2.08; p = 0.001). Whole-brain gEUD > 30 Gy similarly predicted poorer survival (HR = 1.72; p = 0.034). Implementing a personalised gEUD ceiling increased 2-year sensitivity from 31% to 38% with only a three-point drop in specificity. In a 10 × 4 Gy canine intracranial radiotherapy model, limiting Brain-GTV V32 Gy to ≤ 13 cm³ and whole-brain gEUD to ≤ 30 Gy was associated with longer overall survival. A brain-volume-adjusted gEUD ceiling further refined risk prediction. These evidence-based thresholds provide actionable guidance for veterinary treatment planning, with the potential to improve outcomes in canine brain tumour therapy.