PD-l1 expression associated with glioma molecular subtypes: Clinical implications
Recommended Citation
Malta T, Snyder J, Kalkanis S, Castro AV, and Noushmehr H. PD-l1 expression associated with glioma molecular subtypes: Clinical implications. Neuro-Oncology 2017; 19:vi126.
Document Type
Conference Proceeding
Publication Date
2017
Publication Title
Neuro-Oncology
Abstract
The development of immune checkpoint blockade therapy has ushered in a new era of cancer treatment. The use of inhibitors targeting programmed cell death protein 1 (PD1) and/or its ligand 1 (PD-L1) has shown activity in several cancer types and is currently being tested in glioma clinical trials. Previous reports suggested that therapeutic responses are more frequent in patients with higher expression of PD-L1; however, this has been disputed. We aim to evaluate PD-L1 expression in glioma samples and its association with clinical and molecular features. We used protein expression, RNA expression, and DNA methylation data from glioma samples profiled by TCGA (n=1,122). PD-L1 protein and CD274(PD-L1) gene expression were positively correlated (p<0.001). PD-L1 expression was upregulated in IDHwildtype gliomas compared to IDH-mutant (p<0.001) and was associated with poor overall survival (HR=2.392, p<0.001). When PD-L1 was evaluated in the context of recently described TCGA glioma molecular subtypes, we found high PD-L1 expression in the mesenchymal-like and classic-like subtypes, which present poor outcome among all gliomas. Interestingly, PD-L1 protein expression is the lowest in the G-CIMP-low subgroup, which has a GBM-like molecular phenotype and overall poor outcome among IDH-mutants. In fact, within IDH-mutant cohort, PD-L1 expression did not correlate with clinical outcome. We found higher promoter methylation of PD-L1 in IDH-mutant tumors compared to IDH-wildtype. PD-L1 promoter methylation was negatively correlated with PD-L1 gene expression suggesting an epigenetic regulation of PD-L1. Our findings suggest that G-CIMPlow, which has low expression of PD-L1, might not benefit from PD1/PD-L1 blockade. Therefore, we suggest that glioma molecular subtypes should be considered in PD-L1 blockade clinical trial design. Moreover, since PD-L1 expression seems to be regulated by promoter methylation, adding demethylating agents to immunotherapy may improve therapeutic outcomes in gliomas. Further exploration into the mechanisms of tumor immunity and its epigenetic regulation is warranted.
Volume
19
First Page
vi126