Devimistat in Combination with Bendamustine Potently Inhibits Tumor Growth and Promotes CD8+ T Cell Tumor Infiltration in a Syngeneic Mouse Model of CD30+ T-Cell Lymphoma

Introduction: Chimeric antigen receptor-based T cell therapy (CAR-T) has shown great promise in B cell malignancies. CD30-targeted therapies like brentuximab vedotin have activity in T cell malignancies expressing CD30. There is interest in developing CD30-targeted CAR-T therapies for the treatment of relapsed T cell malignancies. As the process of generating autologous T cells takes several weeks, there is a need for active bridging therapeutic regimens that can reduce tumor burden to facilitate subsequent treatment with CAR-T cell therapy. We propose the use of devimistat, a TCA cycle inhibitor, in combination with bendamustine as bridging therapy for the treatment of CD30+ T cell lymphoma (TCL) prior to the use of CAR-T cells. We hypothesize that TCA cycle inhibition will increase the sensitivity of TCL cells to bendamustine, negatively impact the function of immunosuppressive cell populations that depend on mitochondrial metabolism, and create a conducive environment for enhanced CAR-T cell efficacy. To test our hypothesis, we developed a model of CD30+ TCL, evaluated the efficacy of the combination, and characterized its effects by transcriptome profiling of immune cell populations in tumors.

Methods: Mouse TCL cell line EL4-LUC2 was engineered to express human CD30 by retroviral transduction. Single and combination in vitro efficacy was evaluated by viability assays. Combination indices (CI) were calculated using Compusyn. CD30+ EL4-LUC2 cells (EL4.CD30) were inoculated subcutaneously into the flanks of C57Bl6 mice, tumor volumes were calculated, and CD30 expression in tumors was evaluated by immunohistochemical staining and flow cytometry. Transcriptome profiling was performed on a representative subset of tumors by single cell RNA sequencing (scRNA-seq).

Results: EL4.CD30 cells displayed comparable CD30 expression to the human TCL cell line Karpas 299. EC ₅₀ values for Devimistat and bendamustine in EL4.CD30 cells were 64.9 mM (95% CI: 62.4-67.4) and 101.2 mM (96.2-106.4), respectively; in Karpas 299, EC ₅₀ values for Devimistat and bendamustine were substantially higher: 117.2 mM (114.7-119.6) and 188.2 mM (180.0 - 196.8), respectively. When dosed in combination, devimistat and bendamustine (D/B) showed synergy (CI = 0.5-0.8) at effect levels > 0.9 in 8 of the 16 dose combinations tested. D/B synergy at an effect level > 0.9 was less overt in Karpas 299 with only 2 of 16 tested combination levels displaying CI values of 0.5-0.9. Efficacy studies with EL4.CD30 tumor-bearing mice revealed potent D/B anti-tumor activity relative to vehicle and single-agent treatment groups. Six complete regressions and 4 partial regressions were observed in the D/B group (Fig. 1A). Moreover, tumor growth rates and median survival values were significantly different (p<0.0001) in D/B treated mice (Fig 1B-C). Transcriptome profiling of harvested tumors by scRNA-seq revealed substantial infiltration of cytotoxic T lymphocytes (CTL) exclusively in a combination-treated mouse that responded to therapy (Fig. 2A-C). Direct interrogation of Treg-associated transcripts revealed very low Treg numbers and no clear association to any treatment.

Conclusions: Our data demonstrate that D/B has potent pharmacological activity in vitro and in vivo and promotes CTL influx into tumors. This suggests that D/B creates an immunopermissive environment fit for CAR-T cell activity. Subsequent studies will investigate the efficacy of CD30.CAR-T cells following D/B treatment. A Phase II pilot study evaluating the feasibility, safety, and tolerability of D/B in patients with relapsed/refractory T-cell Non-Hodgkin Lymphoma is currently ongoing (NCT04217317).

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal