CAR T cells for Immunotherapy of Cancer

In recent years, immunotherapy has revolutionized cancer treatment by prolonging survival and even curing patients lacking other available therapies. Besides immune checkpoint inhibitors, one of the major advances is attributed to the success of chimeric antigen receptor (CAR)-T cell therapy in treating patients with B-cell malignancies. Although many patients respond, some are resistant to treatment and others will relapse after an initial response. CAR-T cell efficacy can be enhanced by adding a transgene cassette encoding an immune-stimulatory molecule that induce bystander immunity and a broader anti-cancer response. This is particularly important to improve efficacy in solid tumors where the “CAR-target-antigen” is often heterogeneously expressed. The neutrophil-activating protein (NAP), a virulence factor of Helicobacter Pylori, is an efficient immune stimulator. When secreted from CAR-T cells NAP can induce bystander immunity with killing of “CAR-target-negative” tumor cells. In my thesis, I aim at investigating resistance mechanisms to CAR-T therapy and developing new improved CAR-T therapies. In paper I, clinical response in patients with B-cell lymphoma was evaluated after treatment with two doses of third-generation CD19CAR-T cells. Response was observed in 9 out of 24 patients (37.5%) irrespectively of receiving one or two doses. We also analyzed whether the profile of the individual CAR-T infusion products could predict response. CAR-Ts from responders had high cytotoxic and low dysfunctional profile while CAR-Ts from non-responders had high dysfunction profile. Extended culture time during manufacturing was associated with dysfunction. As CAR-T products of non-responders were on average cultured longer, this may explain their dysfunctionality. In paper II, we developed CD20-targeting CAR-T cells (CAR20-T), which efficiently killed CD20-expressing human lymphoma cell lines and patient-derived lymphoma cells in vitro. By arming CAR20-T cells with NAP we could delay tumor growth in lymphoma-bearing mice compared to conventional, unarmed CAR-T cells. In paper III, we engineered five IL13Rα2-targeting CAR-T cell constructs for glioblastoma therapy. The candidates only differed in the complementary determining regions (CDRs) of the single-chain variable fragment portion of the CAR. We found that CDR-mediated CAR clustering could lead to antigen-independent tonic signaling and subsequent CAR-T cell dysfunction. We also identified one candidate which did not display any significant tonic signaling and that possessed therapeutic effect on mice with orthotopic glioblastoma. IL13Rα2 is heterogeneously expressed in glioblastoma tumors so in paper IV, we armed the best CAR-T cell construct with NAP and found that NAP-armed IL13Rα2 CAR-T cells could prolong survival of glioblastoma-bearing mice