Open Menu

Principal Investigators

Nicholas A Butowski, University of California, San Francisco
Roger Stupp, Northwestern University
Susan Chang (Co-I), University of California, San Francisco

Title

Advancing treatment and understanding of immunotherapy in glioblastoma

Project Information

Project 1 (PIs Roger Stupp, Adam Sonabend) - This project has 3 specific aims: Aim 1: Characterize & optimize the effect of immune modulation with doxorubicin (DOX) and second generation immune checkpoint inhibition strategies (DOX + aPD1)on anti-tumoral immunity in GBM. Aim2: Determine the effect of ultrasound-mediated opening of the blood-brain barrier on the patient’s brain tissue concentrations of drugs administered (DOX + aPD1) , and evaluate the anti-tumoral immune response in glioblastoma patients. Aim3: Investigate whether tumor pERK/MAPK activation predicts a response to DOX + aPD1 in glioblastoma patients. Successful completion will determine the i) value of immune modulation by DOX and aPD1; ii) the value of induction (i.e. pre-resection) treatment and in vivo evaluation of anti-tumor immune response; iii) the value (or absence thereof) of BBB-opening for aPD1 efficacy; and iv) whether pERK predicts response and identifies patients that will benefit from treatment.

Despite preclinical promise and occasional long(er) term tumor growth control in subsets of patients, immune checkpoint blockade (aPD1) has failed to improve outcome in glioblastoma (GBM). The failure to demonstrate a benefit may be due to 1) an immunosuppressive microenvironment with few immunogenic tumor cells; 2) insufficient aPD1 drug penetration across the blood-brain barrier (BBB), thus failing to reach the infiltrating tumor cells in the peritumoral brain; and 3) molecular heterogeneity of GBM. Here we systematically address each of these challenges. We aim to elicit robust anti-tumoral response to novel 2nd generation combinatorial immune checkpoint and immune modulatory treatment. Modulating the immune response by low doses of doxorubicin as has been shown previously in other cancer types. To allow for adequate drug tissue penetration, we will transiently disrupt the blood-brain barrier by a skull implantable ultrasound device (SonoCloud-9) in association with intravenous microbubbles (US/MB). In our ongoing research and clinical trials we have demonstrated the feasibility and safety of this approach including targeted biopsies of enhancing and non-enhancing tissue where we measure drug tissue distribution. Lastly, we aim to identify molecular characteristics that will predict a benefit from aPD1 immunotherapy based on our prior work where we have identified MAPK activation (pERK staining) as a putative biomarker. Steps include firstly refining the sequence of DOX & aPD1 administration in preclinical models. Second, for verification and optimization we will treat a few pilot patients in the lead-in phase of our trial. Third, the optimized regimen will be tested in a multi-cohort study in patients with recurrent GBM who will be treated by these novel drug combinaton (DOX and the aPD1). Primary endpoints are tumor immune response in the resection specimen and/or peripheral immune response, drug concentrations in enhancing and non-enhancing brain tissue (targeted biopsies at the time of resection) and safety. Clinical outcome and biomarkers are secondary endpoints. Four cohorts of patient will be examined where DOX/aPD1 treatment starts at different time points: Induction (neoadjuvant, prior to resection); or intraoperatively upon resection, each with and without US/MB. A non-interventional standard of care cohort will serve as control.

Project 2 (PIs Hideho Okada, Jennifer Clarke)– This project has 3 specific aims: In Aim 1, we will generate GMP-grade lentiviral vector and establish standard procedures for GMP-grade manufacturing of the EGFRvIII-primed IL-13Rα2/ EphA2 CAR T cells. In Aim 2, we will complete both in vitro and in vivo studies that are required for submission of an IND. In vitro studies include confirmation of antigen-specific priming, specific cytotoxicity, and absence of replication competent lentivirus. In vivo, we will confirm preclinical efficacy, toxicology, tissue biodistribution, and engraftment as well as persistence of synNotch CART cells. In Aim 3, in a first-in-human phase 1 clinical trial, we will determine safety and toxicity of the synNotch CART cells in patients with recurrent EGFRvIII+ GBM. In a subsequent expansion cohort, in patients with recurrent EGFRvIII+ GBM, we will infuse synNotch CART cells intravenously prior to surgery, and then evaluate infiltration, priming and function of the infused cells in the resected GBM. These studies will allow us to determine optimal dosing based not only on tolerability but also on pharmacodynamic assessments.

Development of safe and effective (CAR)-transduced T cell (CART) therapy for glioblastoma (GBM) needs to overcome multiple challenges, including on-target off-tumor toxicity, heterogeneity of antigen expression, and exhaustion of CART cells. There are no GBM-specific surface antigens that are uniformly present in tumor tissue. Mutant epidermal growth factor receptor (EGFRvIII), for example, is GBM-specific but its expression is heterogenous within the tumor. On the other hand, while non-mutant GBM-associated antigens (GAAs), such as EphA2 and IL-13Rα2, are more uniformly expressed in GBM, their expression in other organs outside the central nervous system (CNS) raises concern for off-tumor toxicity. As a way to safely and effectively target GAAs, we have adopted a novel synthetic Notch “synNotch” receptor system and developed innovative T cells. In this system, the first antigen, which is expressed exclusively on brain or GBM cells (e.g. EGFRvIII), primes the T cells to induce expression of a CAR that recognizes IL-13Rα2 and EphA2, thereby eradicating GBM cells expressing either EphA2 or IL-13α2. Our data show that EGFRvIII-synNotch primed IL-13Rα2/EphA2 CAR are effectively but restrictedly activated by EGFRvIII as the GBM-specific signal, leading to complete eradication of patient-derived xenografts with heterogeneous EGFRvIII expression but without attacking IL-13Rα2/EphA2- positive cells outside of CNS. Furthermore, these synNotch-CAR T cells were significantly more efficacious than conventional, constitutively expressed IL-13Rα2/EphA2 CAR T cells, and were associated with excellent persistence (>100 days in vivo). Taken together, our data indicate that synNotch CAR T cells can revolutionize the CAR T therapy for solid cancers by overcoming the off-tumor toxicity, antigen heterogeneity, and lack of persistence. Our studies will represent the development of a possible breakthrough CAR T therapy that allows us to overcome the major challenges in immunotherapy of solid cancer.