Abstract Background and aims: Clonal evolution and the immune microenvironment in glioblastoma multiforme (GBM) harbor critical clues to its well-known but poorly understood treatment resistance, and to assessing the potential benefit of checkpoint inhibition. We created a tumor-immune interaction map by computing the mutational and neo-epitope tumoral burden, along with the magnitude and clonality of infiltrating lymphocytes, from initial and recurrent tumor of a single patient who benefited from PD-1 blockade. The question of how this particular GBM evolution trajectory compares to a typical GBM is addressed by projection of mutational and other signatures onto a causal gene-gene interaction network we derived from the TCGA GBM cohort. Methods: We compiled samples from a 57- year-old patient who presented with a seizure and had a resection of a large temporal GBM followed by radiotherapy with concurrent temozolomide, with subsequent re-resection for recurrent disease. Starting two weeks after the final resection the patient received 26 cycles of a PD-1 checkpoint inhibitor simultaneous with 20 cycles of bevacizumab, a VEGF-A inhibitor, for a total survival of 25 months after diagnosis. Samples from the primary and three spatial sectors of the recurrence were submitted for RNA sequencing. Data analyses included intra-tumoral gene expression including MHC-I and MHC-II allele-specific expression, expressed relative mutation and (class I and class II) neo-epitope prediction, T and B cell receptor sequencing, and survival and molecular analysis on the GBM TCGA dataset. Results and Conclusions: We quantified the evolving relative mutational and neo-epitope burden of a GBM, emphasizing the differences between class I and class II neo-epitope spectra and resulting adaptive immune recruitment between the original and recurrent tumor regions. We find no evidence of hypermutation in the recurrence, a relative absence of immune infiltration in all recurrences despite the expression of a number of neo-epitope generating mutations (including a novel clonal mutation of EGFR), and compelling evidence of epigenetically driven aberrations in one recurrence sector. Hypothesizing significant molecular heterogeneity as a result of immune editing and treatment, we derived a tumor-recurrence molecular evolution trajectory and projected it onto a Bayesian inference network we derived from the TCGA GBM cohort. We quantified intra-tumoral heterogeneity, derived a patient survival signature, and also identified immunogenic mutated gene regulatory subnetworks that could be potentially targeted by any induced robust immune response. Citation Format: Bojan Losic, Raymund Yong, Nicholas Akers, Ilaria Laface, Nadejda Tsankova, Robert Sebra, Sacha Gnjatic, Adilia Hormigo. Mapping tumoral and immune heterogeneity in PD-1 responsive glioblastoma multiforme [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1318.