Brain dose burden is associated with treatment-related lymphopenia in IDH-wildtype glioblastoma: dosimetric determinants and prognostic impact.

BACKGROUND AND PURPOSE: Treatment-related lymphopenia is common during chemoradiotherapy for glioblastoma and has been associated with adverse outcome. Prior studies have implicated whole-brain V20Gy-V25Gy, mean brain dose, and treatment-volume reduction, but the relative contribution of target size and cerebral dose burden remains uncertain. We evaluated the relationship between brain dose-volume exposure, longitudinal lymphocyte depletion, and landmark outcome in 102 patients with IDH-wildtype glioblastoma. MATERIALS AND METHODS: This retrospective single-center cohort included 102 adults with histologically confirmed IDH-wildtype glioblastoma, paired blood counts, target-volume and brain-DVH data, and clinical follow-up. Absolute lymphocyte count (ALC) was assessed before radiotherapy and at treatment completion. Brain mean dose, V5Gy, V10Gy, V15Gy, V20Gy, V25Gy, integral dose, target volumes, corticosteroid exposure, temozolomide, and clinical-molecular variables were evaluated. V20Gy was used as the principal clinically interpretable dosimetric exposure, with the remaining dose metrics treated as correlated secondary exposures. Multivariable robust linear regression modeled log2 end-of-radiotherapy ALC after adjustment for baseline ALC and major confounders. Survival analyses were anchored at radiotherapy completion. RESULTS: Median ALC decreased from 1810 to 1050 cells/µL, corresponding to a median relative decline of 42.6%. Grade ≥ 2 and grade ≥ 3 lymphopenia occurred in 33 (32.4%) and 14 (13.7%) patients, respectively. Greater cerebral dose burden was associated with lower lymphocyte preservation. In adjusted models, mean brain dose (β per 5 Gy, - 0.251; 95% CI, - 0.430 to - 0.071; p = 0.006), V15Gy (β per 10% points, - 0.198; p = 0.005), V20Gy (β, - 0.241; p = 0.003), V25Gy (β, - 0.262; p = 0.004), and integral dose (β per 5 Gy·L, - 0.197; p = 0.001) predicted lower end-of-radiotherapy ALC; V10Gy was also significant, whereas V5Gy was not. PTV was strongly correlated with V20Gy (ρ = 0.78, p < 0.001) but showed little direct association with lymphocyte preservation (ρ=-0.06, p = 0.55). Grade ≥ 2 lymphopenia was independently associated with inferior landmark overall survival (HR, 2.43; 95% CI, 1.30-4.56; p = 0.005) and progression-free survival (HR, 2.42; 95% CI, 1.37-4.29; p = 0.004). CONCLUSION: Greater cerebral dose burden is associated with treatment-related lymphocyte depletion beyond the information conveyed by target volume alone. The most consistent associations involve mean brain dose, V15Gy-V25Gy, and integral dose, while clinically relevant grade ≥ 2 lymphopenia identifies patients with substantially inferior landmark outcome. These findings support prospective evaluation of immune-sparing radiotherapy strategies that reduce unnecessary cerebral dose exposure without compromising target coverage.