To recapitulate tumor microenvironments of distinct GBM molecular subtypes ex vivo for a preclinical, multi-arm screening of immunotherapy responses.

Programmed cell-death protein-1 (PD-1) checkpoint blockade (Nivolumab) has shown unpredictable efficacy in Glioblastoma (GBM) patients due to the heterogeneity in tumor genetics and microenvironment composition. Improving the clinical use of Nivolumab for each molecular GBM subtype requires a holistic understanding of the diverse immunosuppressive milieus that modulate anti-GBM immunity and a robust microphysiological system to disrupt the evolving niche interactions among GBM and its tumor microenvironment.

Glioma stem cells were isolated from 9 adults with Isocitrate Dehydrogenase-1/2 wildtype GBM, who were molecularly classified by DNA methylation as proneural (PN) (n = 3), classical (CL) (n = 3), and mesenchymal (MES) (n = 3), and treated with Nivolumab (median 3.7 months). In ‘GBM-on-a-Chip’, over a 72-hour window, we infused therapy regimens daily and charted the extravasation events of circulating CD8+T-cells across a micropatterned brain vessel and a 3-D (three-dimensional) hyaluronan hydrogel constituting patient-specific GBM tumorspheres and tumor-associated macrophages (TAMs).

It was found that, under control and Nivolumab monotherapy conditions, CD8+T-cell infiltrate (cell/mm²), migration trajectories, and speed (μm/min) were largely restricted in MES GBM-on-a-chip cultures, relative to CL and PN GBM (P <0.05). This was reasoned by abundant perivascular M2-like (CD163+) TAMs, less activated (CD154+) CD8+T-cells, and elevated checkpoint (PD-1/PD-L1) and anti-inflammatory cytokine (TGFβ, IL-10, CSF-1) expressions (P <0.05). To enhance Nivolumab efficacy, BLZ945, a CSF-1R inhibitor, was co-administered to attenuate TAM-mediated immunosuppression by targeting a CSF-1/PD-1 signaling network. Together, BLZ945 and Nivolumab enhanced CD8+T-cell extravasation and caspase-3/7-mediated GBM apoptosis (P <0.05).

We demonstrate an integrated approach to profile the differences in TAM-mediated immunosuppression for subtype-specific GBM and target signaling milieus to optimize Nivolumab efficacy ex vivo. Further parallel in vivo immunotherapy trial investigations are warranted to validate the prognostic efficacy of GBM-on-a-Chip.