TMIC-10. IDENTIFICATION, VALIDATION AND BIOLOGICAL CHARACTERIZATION OF NOVEL GLIOBLASTOMA TUMOUR MICROENVIRONMENT SUBTYPES: IMPLICATIONS FOR PRECISION IMMUNOTHERAPY

New precision medicine therapies are urgently required for glioblastoma (GBM). However, to date, efforts to subtype patients based on molecular profiles, have failed to direct treatment strategies. We hypothesized that interrogation of the GBM tumor microenvironment (TME) and identification of novel TME-specific subtypes could inform new precision treatment strategies. To this end, a refined and validated microenvironment cell population (MCP)-counter method was applied to > 800 GBM patient tumours and validated by multiplex-immunohistochemistry. The MCP-counter deconvolution method interrogates the TME composition from transcriptomic data. Using this refined method, we classified the GLIOTRAIN(www.gliotrain.eu) IDHwt GBM cohort (n=123) into 3 novel clusters characterised by differences in TME composition and subsequently validated findings in the TCGA (n=69), CGGA (n=72) and DUKE (unpublished)(n=162) cohorts. TME^High tumours (30%) displayed elevated immune populations, functional orientation markers, immune checkpoint genes, and upregulated immunoregulatory pathways. Moreover, tertiary lymphoid structures were a feature of TME^High/mesenchymal⁺ patients. TME^Med (46%) tumours displayed heterogeneous immune populations and upregulated neuronal signalling pathways. TME^Low (24%) tumours represented an ‘immune-desert’ group, high EGFR mutation frequency and upregulated EGFR signalling pathways. Longitudinal analysis of the GLASS cohort revealed TME-subtype transitions upon recurrence, influenced by TME composition changes. Finally, assessment of three GBM immunotherapy clinical trial cohorts revealed that TME^High patients treated with neo-adjuvant anti-PD1 have a significantly improved survival (P=0.04). Moreover, TME^High patients treated with anti-PD1 and an oncolytic virus (PVSRIPO) in the adjuvant setting, showed a trend towards improved survival (P=0.15 and P=0.056 respectively). Overall, we have established a novel TME-based classification system for application in intracranial malignancies. This system may be used to better inform a precision targeting approach in the brain tumour setting. For example, we hypothesise that patients bearing TME^Low tumours may be amenable to neoadjuvant anti-TIM3 + EGFR inhibitor, TME^Med to anti-angiogenic immunotherapy, and TME^High patients to neoadjuvant anti-PD1 + anti-CTLA4.