Combined EZH2 and Bcl-2 inhibitors as precision therapy for genetically defined DLBCL subtypes

Abstract

Molecular alterations in the histone methyltransferase EZH2 and the antiapoptotic protein Bcl-2 frequently co-occur in diffuse large B-cell lymphoma (DLBCL). Because DLBCL tumors with these characteristics are likely dependent on both oncogenes, dual targeting of EZH2 and Bcl-2 is a rational therapeutic approach. We hypothesized that EZH2 and Bcl-2 inhibition would be synergistic in DLBCL. To test this, we evaluated the EZH2 inhibitor tazemetostat and the Bcl-2 inhibitor venetoclax in DLBCL cells, 3-dimensional lymphoma organoids, and patient-derived xenografts (PDXs). We found that tazemetostat and venetoclax are synergistic in DLBCL cells and 3-dimensional lymphoma organoids that harbor an EZH2 mutation and an IGH/BCL2 translocation but not in wild-type cells. Tazemetostat treatment results in upregulation of proapoptotic Bcl-2 family members and priming of mitochondria to BH3-mediated apoptosis, which may sensitize cells to venetoclax. The combination of tazemetostat and venetoclax was also synergistic in vivo. In DLBCL PDXs, short-course combination therapy resulted in complete remissions that were durable over time and associated with superior overall survival compared with either drug alone.

Graphical abstract

Figure 1.

Combination therapy with EZH2 and Bcl-2 inhibitors has in vitro antitumor activity in DLBCL model systems. (A) Cell viability, as measured by CellTiter-Glo, in a panel of DLBCL cell lines treated with vehicle, venetoclax alone (V), tazemetostat alone (T), or venetoclax and tazemetostat in combination (C). Drug dosing for each cell line was optimized to approximate 50% killing (supplemental Table 2). Results represent the mean of 3 biologic replicates, each of which was performed in experimental triplicates. Error bars represent standard error of the mean (SEM). (B) Confidence interval value for DLBCL cells treated with vehicle, venetoclax alone, tazemetostat alone, or venetoclax and tazemetostat in combination over a range of doses at fixed ratios (supplemental Table 3). Cell viability was measured using CellTiter-Glo. Synergy was calculated using CompuSyn. Results represent the mean of 3 biologic replicates, each of which was performed in experimental triplicates. Error bars represent SEM. (C-D) Immunofluorescence of DLBCL organoids treated with vehicle, 50 nM venetoclax, 5 μM tazemetostat, or venetoclax and tazemetostat in combination. Live and dead cells were identified by calcein-AM and ethidium homodimer, respectively, on day 8. Image were obtained on a Nikon TE200U microscope. (E) Percentage of live OCI-Ly1 cells by flow cytometry. Cells were stained for live and dead populations using a LIVE/DEAD Fixable Far Red Dead Cell Stain Kit (Thermo Fisher Scientific; cat. no. L34974). (F) Heat map showing RNA sequencing log2 fold change in BCL2 family members in DLBCL cell lines treated with vehicle vs tazemetostat. The Wilcoxon test was used to determine statistical significance. (G) BH3 profiling of DLBCL PDX cells treated with tazemetostat or vehicle. BIM and PUMA evaluate general priming, BAD evaluates Bcl-2–specific priming, MS1 evaluates MCL-1 priming, and HRK evaluates BCL-XL priming. Error bars represent SEM. *P < .05, **P < .01, ***P < .001, ****P < .0001. Com, combination; ns, not significant; Taz, tazemetostat; Veh, vehicle; Ven, venetoclax.

Figure 2.

Inhibition of EZH2 and Bcl-2 in combination results in tumor reduction and prolonged survival in vivo. (A,D,G) Tumor volume over time in SUDHL-6 xenografts and PDXs, as measured by calipers. Error bars represent standard error of the mean (SEM). Tumor growth in SUDHL-6 (B) and PDX (E) xeongrafts, as measured by area under the curve (AUC), for the duration of treatment. Error bars represent SEM. (C,H) Kaplan-Meier curve for overall survival. P values represent comparison with combination therapy for each cohort. Among PDX mice treated with combination therapy, 2 deaths occurred, both in mice without tumors. (F) MRI 3D renderings of PDX mice taken at day 19 and at day 92. Images were obtained using a 1T M3 compact MRI system (Aspect Imaging Ltd.) with a T2-weighted scan without contrast. The tumor area of interest, as determined by automatic thresholding settings, is shown in red. ***P < .001, ****P < .0001.