T Cells from CLL Patients on Venetoclax Mount Potent Cytotoxic Responses in Combination with Epcoritamab, a CD20/CD3 Bispecific Antibody

In patients with chronic lymphocytic leukemia (CLL), venetoclax (ven), achieves deep clinical responses with high rates of undetectable minimal residual disease (uMRD). Epcoritamab (epco), a bispecific CD3xC20 mAb approved for 3rd line therapy of diffuse large B cell lymphoma (DLBCL) and follicular lymphoma (FL), is in clinical development for CLL. However, immune dysfunction common to CLL negatively impacts T-cell dependent effector mechanisms. Treatment with Bruton tyrosine kinase inhibitors (BTKi) can enhance cytotoxic T-cell responses, and the combination of ven with epco in vitro increased killing of CLL cells over single agents. Here, we examined the effect of ven treatment on T-cell differentiation, activation state, and cytotoxicity in combination with epco. Furthermore, we compared effects of ven and BTKi treatment on T-cell function in patients with CLL. Peripheral blood mononuclear cells were collected from 47 CLL patients (18 treatment-naïve(TN)) on study NCT03986034. After ramp-up, ven was continued by the referring community oncologist, and a CD20 mAb was added at their discretion. On ramp-up, CLL cell counts decreased rapidly and after 12 months, the rate of uMRD at 10-4 was 92%. T cell and NK cell counts also decreased; at 12 months, over 1/3 of patients had CD3 counts below reference-range. We observed a relative increase in naive, a decrease in effector-memory T cells. We also observed concurrent significant decreases in expression of activation markers and checkpoint inhibitors (HLA-DR, PD-1, CTLA4). To compare effects of ven and BTKi we performed single cell sequencing (sc-seq; 10x Genomics) of longitudinally collected PB T cells from 7 patients who were treated with BTKi and switched to ven at progression. We analyzed single T-cell transcriptomes and TCR repertoire at BTKi response, BTKi progression, and at 6 months on ven. 43,582 cells passed quality control and were analyzed;13,937 cells were from BTKi response,12,915 from BTKi progression and11,641 from ven response timepoints. Unsupervised clustering identified 10 T-cell clusters; 4 for CD4+, 4 for CD8+ and 2 for regulatory T cells (T-reg). CD4:CD8 ratios were 0.8 on ven compared to 2.4 at BTKi response and progression timepoints. Two CD8+ effector T-cell clusters were expanded on ven, while CD4+ memory-Th2 and 2 T-reg clusters were significantly smaller on ven than on BTKi. Transcriptionally, the largest CD8+ cluster on ven strongly expressed cytotoxic effector and cytokine gene signatures, while being relatively low for T-cell dysfunction gene sets. TCR analysis also mapped hyper-expanded clonotypes (>50 cells) to this same CD8+ effector cluster. In vitro, epco effectively killed CLLs by engaging T-cells. Notably, autologous T cells from patients on ven or allogeneic T cells from healthy donors were equally cytotoxic, and both were strikingly more potent than T cells from TN CLL patients. CD4:CD8 ratios were inversely correlated with the degree of CLL cell lysis and shaped interindividual differences. Using sc-seq of samples (BTKi and ven response) reacted with epco, we identified T cell clusters highly expressing gene sets related to activation, cytotoxicity, immunological synapse formation, interferon, and cytokines. Trajectory analysis of CD8+ clusters revealed that the dominant activated CD8+ T-cell cluster originated from effector and not naïve T cells. Overall, epco-triggered T-cell responses were more robust in samples from patients on ven than on BTKi. We identify beneficial effects of ven on the cytotoxicity of autologous T cells in response to epco, including improved effector:target ratios, transcriptional programs enhancing cytotoxic T-cell responses, expansion of tumor-reactive TCR clonotypes, and reduced or delayed exhaustion of effector T cells. Thus, the combination of epcoritamab with ven is promising and worthy of clinical investigation. The investigators are supported by the intramural program of the NIH.