HGG-11. TRANS-SPECIES STUDY OF IDH-MUTANT REPLICATION-REPAIR DEFICIENT HIGH-GRADE GLIOMAS (RRD-HGG) AND RESPONSE TO COMBINED TARGETED AND IMMUNOTHERAPY: AN IRRDC STUDY

Abstract

Abstract BACKGROUND IDH-mutant gliomas comprise <10% of HGG in children. RRD-HGG comprise 5-10% of childhood HGG, demonstrate high mutation burden (TMB) and respond to immune-checkpoint inhibition (ICI). The impact of RRD with childhood IDH-mutant gliomas, and effective therapeutic options for these patients are not well-established. METHODS Clinical and multi-omic analyses were performed on IDHmut-RRD-HGG registered to the IRRDC and the glioma taskforce. Tumor development and genomic data were studied from a novel IDHmut-RRD-HGG immunocompetent mouse model. Outcome following ICI and targeted therapy were evaluated. RESULTS IDH mut-RRD-HGG accounted for >60% of childhood IDH-mutant-HGG. Conversely, IDH1-mutations were detected in 20% of RRD-HGG. All patients (n=48) harboured germline mutations in MMR genes (CMMRD: 62%, Lynch syndrome: 38%). Contrary to sporadic IDH-mutant-gliomas, the majority (>90%) were high-grade. Diffuse/multifocal involvement, predominantly involving the frontal lobe, was frequent. TMB was lower than IDH-wildtype RRD-HGG (median: 28 mutations/Mb, p<0.05). Secondary POLE/POLD1 mutations were absent. TP53 and ATRX were frequent somatic hits. Copy number changes, particularly loss of CDKN2A/2B, were common. CD8-T-cell infiltration (immunohistochemistry) and tumor inflammation score (transcriptome) were lower than IDH-wildtype RRD-HGG (p<0.05). A novel mouse model (Olig2Cre+/Msh2LoxP/LSL-Idh1R132H) revealed similar lower TMB, diffuse cerebral involvement, slower growth, and low immune infiltrates as compared with IDH-wildtype RRD-HGG models. IDHmut-RRD-HGG demonstrated worse survival as compared to sporadic IDH-mutant-HGG (p<0.001). Within RRD-HGG, ICI monotherapy resulted in inferior survival in IDHmut-RRD-HGG vs IDH-wildtype RRD-HGG (p<0.05). Five patients developed metachronous IDHmut-RRD-HGG while on anti-PD1 treatment for IDH-wildtype RRD-HGG, suggesting intrinsic resistance. Interestingly, an IDH-inhibitor demonstrated objective response in 4/8 IDHmut-RRD-HGG. Furthermore, for IDHmut-RRD-HGG on ICI treatment, the addition of an IDH-inhibitor prolonged survival at 12-months in comparison to those without (p=0.01). CONCLUSION Hypermutant RRD-HGG with IDH1;p.R132H harbour unique immuno-biology and do not respond to anti-PD1 monotherapy. The addition of IDH-inhibition demonstrated favourable responses, supporting need for evaluation of the combination in clinical trials.