EZH2 Inhibitors Enhance CART Cell Quality, Efficacy, In Vivo Homing, Tumor Cell Binding and Killing of Fully Syngeneic Primary B Cell Lymphomas, As Well As Reprogramming Lymphoma Cells to a Highly Immunogenic and T Cell Adherent Phenotype

Although having efficacy in a subset of diffuse large B cell lymphoma (DLBCL) patients, cellular immunotherapies are currently administered without guidance as to whether the underlying biology or immunology are conducive to durable response. Moreover, mechanisms explaining CART resistance in DLBCLs remain largely unknown. One critical missing link to address these unmet needs are syngeneic animal models precisely reflecting the genetics and immunology of human DLBCLs. To overcome these hurdles, we developed a suite of genetically engineered mouse models (GEMM) faithfully recapitulating different types of DLBCL. A retrospective study showed that C3/EZB DLBCLs, featuring EZH2 mutations and BCL2 translocation, experienced inferior outcomes after CART therapy. We reported that Ezh2 mutations cause aberrant repression of genes required for productive immune synapses with T cells. We hypothesized that this phenotype limits CART cell activity in vivo, and that EZH2 inhibitors could overcome CART resistance. We generated EZB GEMMs with conditional expression of Ezh2 ^Y641F and BCL2 in germinal center (GC) B cells. These mice developed T cell depleted lymphomas, histologically similar to human C3/EZB DLBCLs. We generated several EZB cell lines from these mice, which recapitulated the epigenetic, transcriptional, histological, and immune microenvironment of EZB DLBCLs in humans when engrafted in immunocompetent recipients.

Strikingly, treating EZB cells with EZH2 inhibitor tazemetostat (taz) ex vivo reprogrammed them to re-express the full spectrum of T cell engagement genes such as ICOSL, ITGB7, 4-1BBL, and OX40L and rendered them highly immunogenic. For example, mixing endogenous T cells with taz pre-treated EZB cells yielded markedly enhanced killing (p<0.001). Taz pre-treated EZB cells failed to engraft in syngeneic mice due to immunological rejection, but caused fatal DLBCL in Rag1-/- mice. EZB cell lines did not undergo proliferation arrest after taz ex vivo, but showed a significant reduction in tumor burden (p<0.05) when treated in vivo, along with a 3-fold increase in tumor CD4 and CD8 cells (p<0.05), and 50% reduction in Tregs (p<0.01). Taz also increased the proportion of naïve/memory CD8 and reduced effector CD8 cells. EZH2 inhibitors therefore mediate their effects at least in part through immunological mechanisms.

We next examined whether taz had any deleterious effect on CART cells. Pre-treating donor mice with taz prior to harvesting T cells yielded significant increase in memory CD8 CART cells after transduction (p<0.05). Taz pre-treated CART cells displayed superior killing of EZB cells (p<0.001) and proliferative capacity ex vivo (p<0.001). Infusing taz pre-treated CART cells in EZB mice resulted in significantly reduced tumor burden in vivo (p<0.001) and reduction in PD1+CD38+ exhausted CD8 CART cells (p<0.001).

On the other hand, pre-treatment of EZB cells (but not CART cells) with taz ex vivo enhanced CART cell killing by 3-fold (p<0.001). Administering CART cells in vivo in EZB mice after taz treatment significantly prolonged their survival (100% vs 30%, p<0.01) and was associated with significant reduction in tumor burden assessed by IVIS and post-mortem examination. Therefore, EZH2 inhibition enhances CART cell activity through direct effects on CART cells, in addition to making EZB immunogenic.

Finally, to directly demonstrate the functional consequence of rendering EZB lymphomas immunogenic on CART performance in vivo, we performed intravital 2-photon imaging of popliteal lymph nodes using dTomato labeled CART cells and GFP+ EZB lymphomas. Remarkably, pre-treatment of EZB cells with taz, doubled the recruitment of CART cells into lymphomas (p<0.001) and significantly enhanced the CART to EZB lymphoma cell surface engagement index (strength and duration of binding, p<0.001). This was associated with increased killing, and CART-engaged lymphoma cells were observed to be ingested/engulfed or phagocyted by tumor resident macrophages.

Collectively, EZH2 inhibitors yield a potent boost to CART mediated anti-lymphoma effects by enhancing CART cell functions and B cell immunogenicity, which would likely yield a significant clinical benefit for these patients where CART cells are less active. These results prompted us to initiate a clinical trial evaluating the safety and efficacy of this combination in B cell lymphomas (NCT05934838).