Revealing Pan-Histology Immunomodulatory Targets in Pediatric Central Nervous System Tumors

SIMPLE SUMMARY: Pediatric brain tumors, comprising both benign and malignant forms, rank as the second most prevalent neoplasms in children and confer the highest mortality among pediatric cancer cases. While the immune system exhibits anticancer capabilities under certain circumstances, tumors deploy diverse mechanisms to stifle immune responses. These mechanisms are poorly understood in pediatric brain tumors, which are developmentally unique. In this study, we were motivated to find common mechanisms of immune resistance employed by aggressive pediatric brain tumors, leveraging patient-derived molecular data from the expansive Childhood Brain Tumor Network. Our analyses identify shared immunologic clusters between tumor types that correlate with patient outcomes, immune cell content, and immunosuppressive gene expression. We identify various immunosuppressive genes that contribute to worse outcomes within clusters. We further utilize this dataset to examine genes implicated in preventing immune recognition in malignant pediatric brain tumors and demonstrate how a targeted therapeutic approach is applicable in the central nervous system context to overcome this mode of immunologic resistance. Understanding how pediatric brain tumors evade the immune system can guide the development of immunotherapies, offering new hope for improved outcomes and quality of life for young patients facing these challenging conditions. ABSTRACT: Background: The application of immunotherapy for pediatric CNS malignancies has been limited by the poorly understood immune landscape in this context. The aim of this study was to uncover the mechanisms of immune suppression common among pediatric brain tumors. Methods: We apply an immunologic clustering algorithm validated by The Cancer Genome Atlas Project to an independent pediatric CNS transcriptomic dataset. Within the clusters, the mechanisms of immunosuppression are explored via tumor microenvironment deconvolution and survival analyses to identify relevant immunosuppressive genes with translational relevance. Results: High-grade diseases fall predominantly within an immunosuppressive subtype (C4) that independently lowers overall survival time and where common immune checkpoints (e.g., PDL1, CTLA4) are less relevant. Instead, we identify several alternative immunomodulatory targets with relevance across histologic diseases. Specifically, we show how the mechanism of EZH2 inhibition to enhance tumor immunogenicity in vitro via the upregulation of MHC class 1 is applicable to a pediatric CNS oncologic context. Meanwhile, we identify that the C3 (inflammatory) immune subtype is more common in low-grade diseases and find that immune checkpoint inhibition may be an effective way to curb progression for this subset. Conclusions: Three predominant immunologic clusters are identified across pediatric brain tumors. Among high-risk diseases, the predominant immune cluster is associated with recurrent immunomodulatory genes that influence immune infiltrate, including a subset that impacts survival across histologies.