NanoString immune profiling of pediatric brain tumors

With the increasing use of immunotherapy in cancer, understanding the tumor immune microenvironment (TIME) has become increasingly important. There has been great success in treatment of some pediatric high-grade gliomas with immune checkpoint inhibition, however overall, the immune microenvironment in pediatric brain tumors remains poorly understood. Accordingly, we developed a clinical immune-oncology gene expression assay and used it to profile a diverse range of pediatric brain tumors with detailed clinical and molecular annotation, including a total of 1379 cases. Our findings demonstrate the critical importance of understanding the TIME to guide the prognosis and management of pediatric brain tumors. In pediatric low-grade gliomas we identified distinct patterns of immune activation with prognostic significance in BRAF V600E mutant tumors. In pediatric high-grade gliomas, we observed elevated inflammation and T cell infiltrates in tumors that have historically been considered immune cold, as well as genomic correlates of inflammation levels. In our cohort of mismatch repair deficient high-grade gliomas treated with immune checkpoint inhibition, we found that high tumor inflammation signature was a significant predictor of treatment response. Furthermore, we demonstrated the potential for multimodal biomarkers to improve the treatment stratification of these patients. Importantly, while overall patterns of immune activation were observed for histologically and genomically defined tumor types, there was nonetheless significant variability within each diagnostic entity, indicating that the immune response must be evaluated as an independent feature. In sum, in addition to the histology and molecular profile, this work establishes the importance of assessing and reporting on the tumor immune microenvironment as an essential axis of cancer diagnosis in the era of personalized medicine. This study analyzed frozen or formalin-fixed paraffin-embedded (FFPE) tissue from 571 patient samples, including 559 pediatric brain tumors and 12 non-tumor brain control samples. Gene expression profiling was carried out using the NanoString nCounter system with a customized CodeSet composed of a total of 103 genes focused on the tumor-immune microenvironment.