CSIG-17. SIGNALING PATHWAY-BASED CLUSTERING OF MASS CYTOMETRY DATA UNCOVERS CELLS ASSOCIATED WITH CLINICAL PROGRESSION IN GLIOBLASTOMA

Abstract

Abstract Recent advances in single cell expression technology have allowed us to appreciate the complexity in tumoral heterogeneity of solid tumors such as glioblastoma (GBM), which remains uniformly fatal despite aggressive therapies. Currently, there is no knowledge about cell identities and functional characteristics of the different tumoral populations. To this end, we created a 28-marker mass cytometry panel composed of signaling markers, including phospho-proteins, and lineage markers that have been associated with GBM or are considered putative cancer stem cell markers. We evaluated lines established from core and edge (periphery) areas from the same patient and validated to be functionally different with RNA expression and in vivo experiments. We first clustered the cells from all samples using the signaling markers in order to establish differences based on pathway utilization, and then we evaluated the expression of all markers. We found clusters shared by cells from both regions as well as region-specific ones. In order to better characterize these, we used marker enrichment modeling (MEM) and generated machine-readable labels. These were compared to MEM labels generated after using the same signaling markers to cluster primary samples from newly diagnosed or recurrent patients. Similarly, MEM labels were obtained from patients in a clinical trial treated with capecitabine that had positive or negative responses. Clusters from edge cells were found in higher proportion in recurrent tumors and were correlated with negative responses to the chemotherapeutic agent. Initial analyses have also been done correlating immune clusters with tumor clusters in the same samples. We are currently conducting in vitro studies and analyzing additional lines from other patients in order to validate these findings. Our studies exploit single cell cytometry to underscore pathways that are preferentially use by different cell populations with the ultimate goal of establishing therapeutic efforts to extend patient survival.