Selective Support of CAR-T Cell Therapies By Cis-Targeted IL-2 Enhances Anti-Tumor Activity and Obviates the Need for Lymphodepletion in Non-Human Primates

Limited in vivo persistence and the need for lymphodepleting chemotherapy are ongoing challenges in CAR T cell therapy for hematologic malignancies. In vivo behavior of CAR T cells has shown a strong correlation with improved outcomes, leading to an interest in developing approaches to enhance the expansion and persistence of infused CAR T cells. Exogenous cytokines such as IL-2 can be administered in vivo, but are associated with adverse effects and have the potential to stimulate regulatory T cells, thus impeding the immune response. Lymphodepletion aids CAR T cell function but causes pancytopenia and associated complications. To address these challenges, we developed a CAR T-specific IL-2 fusion molecule that selectively activates CAR T cells by recognizing an extracellular tag. We hypothesized that this method should improve the anti-tumor efficacy and may eliminate the need for lymphodepleting chemotherapy.

We developed an IL-2 molecule that signals selectively on CD19 or CD20-specific CAR T cells by fusing a CAR T cell-targeting antibody to an attenuated IL-2 mutein with diminished binding to IL-2R⍺ and IL-2Rβ. CAR T cell targeting antibodies were against the non-signaling truncated EGFR tag co-expressed with the CAR. While wild-type IL-2 signals on cells expressing its receptor, cis-targeted IL-2 fusion molecules targeting the EGFRt (EGFRt-IL2) enable selective IL-2 signaling by binding in cis to EGFRt on CAR T cells.

Using STAT5 phosphorylation as a readout, in vitro stimulation of human or rhesus macaque CAR T cells that express EGFRt with EGFRt-IL2 led to a >100-fold increase in selectivity compared with wild-type (WT) IL-2. In vivo activity was demonstrated in NSG mice bearing an acute lymphoblastic leukemia (ALL) cell line treated with a “stress” dose of healthy donor anti-CD19-EGFRt CAR T cells with or without EGFRt-IL2. This was subsequently confirmed in a model where the anti-CD19-EGFRt CAR T cells were generated from a lymphoma patient. Response to treatment was accompanied by CAR T cell expansion, resulting in superior survival in the EGFRt-IL2-treated mice with a median overall survival of 61 days compared to 20 days for mice treated with PBS (p = 0.0015).

To test whether IL-2 mutein administration could abrogate the need for lymphodepleting chemotherapy (LD), we treated two rhesus macaques (RM) with a single dose of autologous anti-CD20-EGFRt CAR T cells with (RM#1) or without (RM#2) IL-2 mutein. In the absence of LD, there was no evidence of B cell depletion after CAR T 20 infusion in RM#2, and anti-CAR antibodies were detected from day 11 onwards. In contrast, RM#1 experienced profound B cell depletion in blood and bone marrow lasting approximately 35 days. The animal also experienced cytokine release syndrome (CRS) on day 7 and required dexamethasone and tocilizumab, showing clinical improvement within 12-24 hours. B cell recovery was accompanied by the development of anti-CAR antibodies (which were first detected on day 28). To test whether re-dosing with IL2 mutein at a later time point (day 49) could expand the previously administered CAR T cells and induce B cell aplasia again, we injected RM#1 with the same dose of EGFRt-IL2 on day 49 without re-administering CAR T cells. This time there was no reduction in B cells, and anti-drug antibody (ADA) became detectable on day 62 (14 days post re-dosing).

In summary, EGFRt-IL2 enhances in vivo anti-tumor activity and survival in xenografted mice, enhances the activity of lymphoma patient CAR T cells, and may abrogate the need for lymphodepletion prior to CAR T cell infusion. Temporal control of cis-targeted cytokines directed by anti-tag antibodies represents a promising approach to enhance CAR T cell therapies.