Fig. 2. The glioblastoma tumor microenvironment rather than antigen availability restrains RT response.

a, TMB in the PDG-Ink4a, PDG-p53 and GL261 glioblastoma mouse models and patients with glioblastoma³⁰ as determined by WES analyses (Methods; MB = mutational burden; PDG-Ink4a n = 4 mice, PDG-p53 n = 3 mice, GL261 n = 3 mice, patients with glioblastoma n = 14 patients). b, Donut charts of the variant type (outer circle) and functional class (inner circle) distribution of mutations in each glioblastoma mouse model and from glioblastoma patient datasets³⁰ (MNP = multiple nucleotide polymorphism; PDG-Ink4a control n = 3 mice, tumor n = 3 mice; PDG-p53 control n = 3 mice, tumor n = 3 mice; GL261 control n = 3 mice, tumor n = 3 mice; patients with glioblastoma n = 14 patients). c, Grading of key histopathological features observed in the PDG-Ink4a, PDG-p53, PDG-Ink4a-OVA and GL261 glioblastoma mouse models (pseudop. necr., pseudopallisading necrosis; cystic degen., cystic degeneration; PDG-Ink4a n = 7 mice; PDG-p53 n = 9 mice; PDG-Ink4a-OVA n = 6 mice; GL261 n = 9 mice). d–f, Flow cytometry quantification of CD24⁺CD11b⁻ dendritic cells (cDC1s, gated from CD45⁺Ly6C⁻CD64⁻MHCII⁺CD11c⁺ cells) (d), CD103⁺ cDC1s (e) and CD3⁺ T cells (gated from CD45⁺CD11b⁻cells) (f) in end-stage, treatment-naive PDG-Ink4a, PDG-p53, PDG-Ink4a-OVA and GL261 tumors (in d and e, PDG-Ink4a n = 6 mice, PDG-p53 n = 6 mice, PDG-Ink4a-OVA n = 5 mice, GL261 n = 5 mice; in f, PDG-Ink4a n = 5 mice, PDG-p53 n = 7 mice, PDG-Ink4a-OVA n = 7 mice, GL261 n = 5 mice). g,h, Kaplan–Meier survival curves of GL261 (g) and PDG-Ink4a-OVA (h) tumor-bearing mice treated with rIgG2a isotype control (Cont), anti-PD-1 (IT), 5x2Gy RT or adjuvant combination treatment (RT + Adj.IT). Statistics: one-way ANOVA with Benjamini, Krieger and Yekutieli correction for multiple testing (a and d–f), log-rank test (g and h). Data are represented as mean ± s.e.m. (a and d–f). Median survival and significance depicted in Supplementary Table 1 (g and h).

Source data