Overcoming the Lack of CD2:CD58 Costimulation in Chimeric Antigen Receptor T-Cell Immunotherapies

Background: CD2 is a key co-stimulatory molecule in T cell activation that enhances immune responses through its interaction with CD58. Evidence suggests the CD2:CD58 axis is crucial for the efficacy of chimeric antigen receptor (CAR) T cells. CD58 loss in cancer cells leads to CAR T cell resistance and CD2-dependent T cell engagement in the tumor microenvironment is essential for sustained cytotoxicity. However, clinical evidence is sparse, and strategies to counteract this resistance are underdeveloped. We assessed the impact of CD2:CD58 axis disruption in CAR T cells and developed a strategy to compensate for the lack of CD58 costimulation.

Methods and Results: We deleted CD2 in anti-CD19 CAR T cells (CD2KO CART19) using CRISPR-Cas9 and demonstrated reduced in vitro cytotoxicity compared to CD2 wild-type CART19 (CD2WT CART19) using a B-cell leukemia (Nalm6) model (p<0.001). These findings were consistent in xenogenic mouse models, with worse median overall survival in CD2KO CART19 treated mice (p<0.05). Mechanistically, we found CD2KO caused poor cytoskeletal organization at the CART19 cell immune synapse. Poor synapse engagement correlated with reduced serum cytokine concentration (IFN-γ, IL-2, sILR2-α, IL-5, and CXCL10/IP) in CD2KO CART19 treated mice. We next performed a similar approach with tumor CD58. We deleted CD58 (CD58KO) in B cell lymphoma cell lines (OCI-Ly18) and Nalm6. We found that CD58KO OCI-Ly18 and Nalm6 cells were less sensitive to CART19 cytotoxicity when compared to CD58WT controls in vitro (p<0.05) and confirmed it in xenograft models (p<0.05).

We next correlated our findings clinically by studying CD58 expression in advanced B cell lymphoma biopsies from patients treated with CART19 at the University of Pennsylvania. Bulk RNA sequencing was performed on diagnostic biopsies from 17 patients with diffuse large B cell lymphoma (DLBCL) prior to CART19 treatment (NCT02030834). Patients in the highest CD58 expression tertile had a significantly longer overall survival (median OS: 26.2 months, n=12) compared to those in the lowest tertile (7.9 months, n=5) (p=0.0083). To gain clinical insight into the role of CD58 in CART19 immunotherapy, we also analyzed CD58 expression via immunohistochemistry in patients’ biopsies before (n=27) and after (n=25) commercial CART19 treatment. We observed a generalized reduction in CD58 H-score post-CART19 (157.0 ± 18.3) compared to pretreatment (184.7 ± 16.4). Moreover, higher pre-CART19 CD58 H-scores were associated with durable (over 1 year) complete responses vs. shorter or no responses (Long CR: 224.5 ± 21.1; Short CR: 183.0 ± 20.0; Refractory/Relapsed: 176.1 ± 28.0).

Lastly, we aimed to address deficiencies in the CD2:CD58 signaling axis to improve CART19 outcomes. We hypothesized that an in-trans switch receptor with an intracellular CD2 domain could recover intracellular CD2 signaling in CAR T cells that lacked extracellular CD2 engagement. We chose PD-1 to serve as the extracellular domain due to its proximity to CD2 in T cell synapses and the analogous location of CD58 and PD-L1 in TCR synapses (Demetriou P et al., Nat Immunol. 2020). The PD-1-CD2 switch receptor engages PD-L1 on tumor cells to trigger CD2 signaling upon CD19:CART19 engagement. Using a P2A lentiviral plasmid, we co-expressed the switch receptor and CAR19 on human T cells. First, we found that the switch receptor did not affect CART19 manufacturing kinetics. We then engrafted PD-L1-expressing Nalm6 in NSG mice and compared leukemic growth following treatment with CD2KO CART19 or CD2KO PD-1:CD2 CART19 cells. CD2KO PD-1:CD2 CART19 showed improved tumor control and overall survival (p=0.001). Repeating the experiment with CD58KO Nalm6, CD2KO PD-1:CD2 CART19 proved superior tumor control and overall survival to both CD2KO and CD2WT CART19, demonstrating the strategy’s potential to overcome CD58 loss in tumor cells (CD2WT CART19 vs CD2WT PD-1:CD2 CART19 p=0.031; CD2KO CART19 vs CD2KO PD-1:CD2 CART19 p=0.022).

Conclusions: In summary, we demonstrated the critical role of the CD2:CD58 axis in CAR T cell immunotherapy using preclinical models and confirmed our findings in lymphoma patients treated with CART19. We developed a novel PD-1-based switch receptor that overcomes the lack of CD2 signaling in preclinical CART19 xenograft models. This modified CAR T cell product may serve as a strategy to enhance clinical responses in patients with CD58-deficient tumors.