ABSTRACT
Glioblastoma (GBM), the most common primary brain tumor in adults, is associated with a poor prognosis and
short survival in spite of aggressive management. T cell dysfunction is a crucial component in the
immunosuppressed state observed in the GBM tumor microenvironment and remains a significant barrier for the
development of successful immunotherapies. Recent work suggests modulation with the immune-checkpoint
inhibitors a-PD-1 or a-CTLA-4 do not provide significant clinical benefit in GBM, as opposed to other cancers.
These data suggest that multiple co-inhibitory receptors are involved in the GBM tumor microenvironment driving
T cell dysfunction. While investigating T cell dysfunction in human inflammatory disease, we first identified a key
role for engagement of TIGIT on T cells by CD155 as a major co-inhibitory pathway; we then demonstrated a
role for TIGIT in maintaining functional stability of Tregs, in that TIGIT signaling prevents conversion of Tregs
into proinflammatory cells. Moreover, we and others have found that TIGIT and its ligand CD155 are highly
expressed in GBM. Our new preliminary data using single-cell RNAseq shows that, among GBM tumor-infiltrating
T cells, Tregs express the highest levels of TIGIT. Anti-TIGIT therapies are now entering clinical trials in cancer
and recently, a-TIGIT mAb tiragolumab was given FDA breakthrough designation. In this proposal, we will
elucidated mechanisms of T cell dysfuction in GBM by conducting a phase 0/1 clinical trial of a-TIGIT and a-PD-
1 mAbs as checkpoint inhibitors for treatment of recurrent GBM. The trial design allows for establishing the safety
of this combination, while providing tumor samples and blood specimens for a robust evaluation of in human
effects of a-TIGIT & a-PD-1 either alone or in combination, with transcriptomic single-cell resolution. Our
mechanistic hypothesis is that while anti-PD-1 enhances both effector and regulatory T cell proliferation, the
addition of a-TIGIT to a-PD-1 will destabilize Treg function and optimize the immune response. To investigate
this hypothesis, our trial utilizes a “window-of-opportunity” design, within a bench-to-bedside-to-bench approach.
Prior to surgery for resection of recurrent tumor, patients will be randomized to receive placebo, or a-PD-1, or a-
TIGIT either alone or in combination. Single-cell RNAseq (10x) will be performed on baseline blood, post-
treatment blood and tumor specimens from the trial patients. We will compare clonality and functional traits of
tumor-infiltrating Tregs and effector T cells, and frequency of proinflammatory and suppressive tumor-derived
myeloid cells. Multiplexed immunofluorescence (4i) will be used for spatial validation. We will analyze circulating
T cells clonally related to tumor-infiltrating T cells to identify peripheral correlates and develop a non-invasive
biomarker of emerging immune response. We will then model effects of PD-1 and TIGIT blockade on T regs in
vitro using healthy donors T regs and compare to effects observed in the trial. Through this comprehensive
approach, our study will determine whether PD-1 and TIGIT receptor engagement on T cells in the GBM tumor
microenvironment induces immune dysfunction not allowing for immune survelliance in this CNS tumor.
