Search

Improvements in the quality and fitness of chimeric antigen receptor (CAR)-engineered T cells, through CAR design or manufacturing optimizations, could enhance the therapeutic potential of CAR-T cells. One parameter influencing the effectiveness of CAR-T cell therapy is the differentiation status of the final product: CAR-T cells that are less-differentiated and less exhausted are more therapeutically effective. In the current study, we demonstrate that CAR-T cells expanded in IL15 (CAR-T/IL15) preserve a less-differentiated stem cell memory (Tscm) phenotype, defined by expression of CD62L+CD45RA+ CCR7+, as compared with cells cultured in IL2 (CAR-T/IL2). CAR-T/IL15 cells exhibited reduced expression of exhaustion markers, higher antiapoptotic properties, and increased proliferative capacity upon antigen challenge. Furthermore, CAR-T/IL15 cells exhibited decreased mTORC1 activity, reduced expression of glycolytic enzymes and improved mitochondrial fitness. CAR-T/IL2 cells cultured in rapamycin (mTORC1 inhibitor) shared phenotypic features with CAR-T/IL15 cells, suggesting that IL15-mediated reduction of mTORC1 activity is responsible for preserving the Tscm phenotype. CAR-T/IL15 cells promoted superior antitumor responses in vivo in comparison with CAR-T/IL2 cells. Inclusion of cytokines IL7 and/or IL21 in addition to IL15 reduced the beneficial effects of IL15 on CAR-T phenotype and antitumor potency. Our findings show that IL15 preserves the CAR-T cell Tscm phenotype and improves their metabolic fitness, which results in superior in vivo antitumor activity, thus opening an avenue that may improve future adoptive T-cell therapies.

Immune checkpoint inhibitors have shown promising clinical efficacy in many types of cancers, including some with metastasis into the central neural system (CNS). However, little is known about the CNS penetration of immune checkpoint inhibitors and their bioactivity on the locoregional T cells. Here, we analyzed the CNS and systemic concentration of pembrolizumab (Pembro) in high-grade glioma patients, who were concurrently treated with intravenously administered Pembro and locoregionally delivered chimeric antigen receptor (CAR) T cells. A total of 99 paired cerebrospinal fluid (CSF) and serum samples were collected from 10 patients during multicycle treatment of CAR T cell and Pembro. We discovered that after the first Pembro infusion, CSF antibody concentrations were increased slower than serum concentrations. However, CSF antibody concentrations reached steady-state 24 hours after Pembro infusion and were maintained stable over the 21-day infusion cycles. Although average CSF Pembro concentrations (3.2nM) were only 1% of serum levels (329nM), we observed effective PD-1 blockade on the T cells in CSF samples of all analyzed patients. The blocking effect was found on both endogenous T cells and locoregionally administered CAR T cells. Further, PD-1 on CD3/CD28 beads-activated T cells was completely blocked after incubation with cell-depleted CSF samples, suggesting the bioactivity of Pembro that penetrated into the CSF. Using a dose-titration assay, we also confirmed that CSF Pembro concentrations were sufficient to ameliorate the cytotoxic activity of CAR T cells against repetitive tumor challenges. Together, for the first time, we showed the dynamic Pembro levels in the CSF throughout multiple treatment cycles and confirmed that systemically administered immune checkpoint inhibitors can penetrate into the CNS and elicit blocking effect on T cells. Our findings add strength to the rationale for combining immune checkpoint inhibitors with locoregionally delivered CAR T cells and other types of immunotherapy for the treatment of brain tumors. Citation Format: Dongrui Wang, Jana Portnow, Vivi Tran, Vivian Chiu, Alfonso Brito, Darya Alizadeh, Renate Starr, Julie Kilpatrick, Paige McNamara, Stephen J. Forman, Behnam Badie, Timothy W. Synold, Christine E. Brown. Systemic anti-PD-1 immunotherapy results in PD-1 blockade on T cells in the cerebrospinal fluid [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2019 Nov 17-20; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2020;8(3 Suppl):Abstract nr A52.

Encouraging clinical experience with chimeric antigen receptor (CAR) T cells supports the notion that even immune-privileged sites such as the brain may be amenable to CAR T therapy. In the context of hematologic B cell malignancies, CD19-CARs have been shown to accumulate in the cerebrospinal fluid (CSF) and reduce the incidence of metastatic disease in the central nervous system. However, treatment of solid tumors, including brain tumors, has proven quite challenging due to heterogeneous antigen expression, suboptimal trafficking, and immunosuppressive networks in the tumor microenvironment that limit CAR T cell function and persistence. This presentation will describe our efforts to overcome these challenges for the treatment of glioblastoma (GBM), one of the most common and aggressive primary malignant brain tumors. Tumor antigen selection is a key challenge for CAR T therapy of brain tumors, as on-target/off-tumor toxicities could be life-threatening. Our preclinical studies demonstrate that a glioma-associated protein, interleukin 13 receptor alpha 2 (IL13Rα2), is an attractive CAR T cell target for glioblastoma in part because this tumor-restricted antigen is expressed by the majority of GBM. Initial clinical findings demonstrate that local delivery of IL13Rα2-CAR T cells is feasible and safe, with evidence for clinical activity in patients. In addition, as part of our on-going efforts to expand the repertoire of immunologic targets for GBM, we have optimized a HER2-CAR and developed a novel toxin-based CAR harnessing the selective GBM-binding properties of chlorotoxin (CLTX). To reduce the potential for tumor antigen escape, we are exploring approaches that leverage multiple antigens, including combining CAR T cells specific to different antigens and developing tandem CARs. Another challenge is achieving efficient tumor trafficking and infiltration of CAR T cells to brain tumors. While different routes of delivery present advantages and disadvantages, our preclinical studies using several tumor models suggest that locoregional delivery, into either the tumor or cerebral spinal fluid (CSF), is more effective than systemic delivery for treatment of brain tumors. These studies underlie our clinical approach of infusing T cells into both the resected tumor cavity and CSF. Indeed, we have demonstrated that CAR T cells administered into the CSF induced a complete response in a patient with recurrent multifocal GBM, including metastatic lesions in the spine.Additional efforts are addressing the highly immunosuppressive GBM microenvironment that limits the effectiveness of adoptively transferred T cells. Our studies highlight the impact of the PD-1/PDL-1 immune checkpoint axis in reducing the antitumor activity of CAR T cells, an inhibition that can be overcome by anti-PD-1 checkpoint blockade. In another approach designed to interrogate the interplay between the tumor microenvironment (TME), host immune system, and CAR T cells, we have established a syngeneic mouse model recapitulating the invasive nature of GBM. These studies are elucidating the reciprocal interactions of CAR T cells, the immune landscape, and the inflammatory TME. In summary, this presentation will highlight on-going studies focusing on the biology of GBM, its microenvironment, and strategies to enhance CAR T cell approaches to more effectively treat brain tumors. Citation Format: Darya Alizadeh, Dongrui Wang, Renate Starr, Brenda Aguilar, Vanessa D. Jonsson , Saul J. Priceman, Michael E. Barish, Behnam Badie, Stephen J. Forman, Christine E. Brown. Advancing CAR T cell therapy for the treatment of brain tumors [abstract]. In: Proceedings of the Fourth CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; Sept 30-Oct 3, 2018; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2019;7(2 Suppl):Abstract nr IA15.

Glioblastoma (GBM) is the most common type of primary brain tumor, with the standard therapy only modestly improving the prognosis, highlighting the necessity to develop advanced treatments. We and others have established the platform to potentiate immune response against GBMs using chimeric antigen receptor (CAR) engineered T-cells. Specifically, we have shown that intracranial administration of CAR T-cells can be well tolerated in patients with recurrent GBMs, together with some early clinical evidence of antitumor response. However, CAR T-cell therapy against GBMs is complicated by the inter- and intratumoral heterogeneity, while the single-targeting therapies only respond to a subset of tumor cells. The development of new CAR therapy would thus aim for targeting a wider range of tumor cells and bypassing antigen escape. Here, we took an approach different from conventional strategies of tumor antigen discovery, exploiting the tumor-binding potential of a natural peptide to develop CAR T-cells that broadly target GBMs. Chlorotoxin (CLTX) is a 36-amino acid peptide with demonstrated GBM-binding capability. Inspired by the utilization of CLTX in GBM tumor imaging, we used a fluorescence-conjugated CLTX to screen the freshly-dispersed primary GBM cells and patient-derived GBM neurospheres, and found that CLTX binding was more homogeneous than the expression of other GBM-associated antigens including EGFR, HER2 and IL13Rα2. Although CLTX has limited inhibitory effect on GBM growth, its broad binding to GBM cells illustrates the potential to be conjugated with a cytotoxic agent. Therefore, we generated CAR T-cells bearing CLTX as the antigen targeting domain. CLTX-CAR T-cells were able to get activated after stimulating with GBM cells, as indicated by their degranulation, cytokine production and immuno-synapse formation. Modification of CAR constructs revealed that CLTX-CAR T-cells with CD28 costimulatory signal exhibited potent effector activity, while the 4-1BB costimulation resulted in inadequate CAR activation. In both in vitro and in vivo models, CLTX-CAR T-cells effectively eliminated GBM cells and tumors, including the ones with no/low expression of EGFR, HER2 and IL13Rα2. Importantly, CLTX peptide exhibited negligible binding to a panel of normal cells from neural and other tissues, and CLTX-CAR T-cells showed no off-target effect against normal organs in tumor-bearing mouse models. Screening on patient-derived GBM neurospheres, we discovered that the expression of metalloproteinase (MMP)-2 on targeT-cells was correlated with the effector function of CLTX-CAR T-cells. Further, the antitumor function of CLTX-CAR T-cells was severely diminished against GBMs with MMP-2 knockdown. Consistent with the cytotoxicity of CLTX-CAR T-cells, MMP-2 expression was also present in a subgroup of GBM cells with undetectable levels of EGFR, HER2 and IL13Rα2 expression. Our results demonstrate for the first time that a peptide toxin can be successfully used as the tumor targeting domain of a CAR, which eliminates GBMs with high efficiency and selectivity. The CLTX-CAR has the potential to limit GBM heterogeneity and compensate current CAR T-cell therapies against solid tumors. Citation Format: Dongrui Wang, Vanessa Jonsson, Sarah Wright, Wen-Chung Chang, Xin Yang, Renate Starr, Alfonso Brito, Brenda Aguilar, Aniee Sarkissian, Lihong Weng, Stephen J Forman, Michael E Barish, Christine E. Brown. Chlorotoxin redirects T-cells for specific and effective targeting against glioblastomas [abstract]. In: Proceedings of the Fourth CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; Sept 30-Oct 3, 2018; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2019;7(2 Suppl):Abstract nr A045.