SUMMARY
Whereas randomized trials for PD1 blockade in glioblastoma (GBM) are negative,2 a subset of these patients
do respond to immunotherapy.3 Determining the basis of response will result in immunotherapy being effectively
used for some patients. We recently reported the analysis of 66 recurrent GBM patients that were treated with
PD1 blockade.1 Whereas response was defined based on imaging criteria, responsive patients exhibited
significant increased survival that was independent of clinical characteristics or additional treatments. 30% of
tumors of responsive patients had enrichment of BRAF and PTPN11 activating mutations, which stimulate MAPK
pathway signaling.1 Therefore, molecular indicators of response to PD1 blockade for the majority of patients
remain unidentified. We performed immunohistochemistry for phosphorylation of p38 and ERK, two effectors of
this pathway. Extent of positivity for these markers was predictive of overall survival following PD1 blockade. Of
relevance to our patient tumor studies, we modeled the effect of CD8 T-cell depletion on tumor progression on
mouse transgenic gliomas that develop de novo. In this setting, CD8 T-cell depletion resulted in immunogenic
tumors that upon transplantation, preferentially engrafted in recipients with CD8 T-cell depletion as opposed to
immunocompetent hosts. Gliomas generated in the absence of CD8 T-cells showed upregulation of Braf and
Ptpn11, and associated MAPK activation indicated by phosphorylation of p38 and Erk. CD8 T-cell depletion
during glioma development also led to robust increase in tumor associated macrophages/microglia (TAM), and
gene expression of pro-inflammatory TAM. MAPK activity correlated with TAM markers in mouse transgenic
gliomas and human GBM, and with several chemokines that promote pro-inflammatory TAM. A CRISPR knock-
out screen for the kinome on intracranial gliomas also implicated MAPK in T-cell recognition of tumor cells. These
findings suggest that MAPK activity in tumor cells (that is suppressed by CD8 T-cells during glioma progression),
promotes pro-inflammatory TAM. We hypothesize that GBM with elevated MAPK signaling, are more susceptible
to CD8 T-cell recognition, as this pathway promotes anti-tumoral TAM. Thus, the subset of GBM that have MAPK
signaling are susceptible to PD1 blockade and CAR T-cell immunotherapy. To investigate this, we will SA1)
Determine the effects of CD8 T-cell depletion during glioma development on TAM phenotype. SA 2) Determine
whether MAPK signaling promotes anti-tumoral glioma TAM. SA 3) Determine the effect of modulating glioma
MAPK signaling on response to PD1 blockade and CAR T-cell therapy. We will use transgenic models in which
tumors develop de novo, cell lines with varying levels of MAPK activation, genetic MAPK manipulation as well
as clinically available drugs inhibit and/or promote MAPK signaling that cross the blood-brain barrier. Successful
execution of these studies will establish a novel immune-related role of MAPK signaling in glioma. We will
determine whether MAPK signaling by tumor cells influences response to PD1 blockade, and CAR T-cell therapy.
