Chimeric antigen receptor T-cell (CART) immunotherapy led to unprecedented responses in patients with refractory/relapsed B-cell non-Hodgkin lymphoma (NHL); nevertheless, two thirds of patients experience treatment failure. Resistance to apoptosis is a key feature of cancer cells, and it is associated with treatment failure. In 87 patients with NHL treated with anti-CD19 CART, we found that chromosomal alteration of B-cell lymphoma 2 (BCL-2), a critical antiapoptotic regulator, in lymphoma cells was associated with reduced survival. Therefore, we combined CART19 with the FDA-approved BCL-2 inhibitor venetoclax and demonstrated in vivo synergy in venetoclax-sensitive NHL. However, higher venetoclax doses needed for venetoclax-resistant lymphomas resulted in CART toxicity. To overcome this limitation, we developed venetoclax-resistant CART by overexpressing mutated BCL-2(F104L), which is not recognized by venetoclax. Notably, BCL-2(F104L)-CART19 synergized with venetoclax in multiple lymphoma xenograft models. Furthermore, we uncovered that BCL-2 overexpression in T cells intrinsically enhanced CART antitumor activity in preclinical models and in patients by prolonging CART persistence.

Significance:

This study highlights the role of BCL-2 in resistance to CART immunotherapy for cancer and introduces a novel concept for combination therapies—the engineering of CART cells to make them resistant to proapoptotic small molecules, thereby enhancing the therapeutic index of these combination therapies.

This article is highlighted in the In This Issue feature, p. 2221

Chimeric antigen receptor T-cell (CART) immunotherapy has significantly improved the clinical outcomes of patients with relapsed/refractory B-cell lymphomas and leukemias (1–11). Despite the remarkable results of anti-CD19 CART (CART19), more than 60% of patients with lymphoma either do not respond to or eventually relapse after CART19 treatment (10). We demonstrated that, at 5 years of follow-up, 69% of patients with large B-cell lymphoma (LBCL) treated with a 4-1BB costimulated CART19 (CTL019, tisagenlecleucel) have not achieved or maintained a complete remission (CR; ref. 3). Thus, enhancing the antitumor ability of CART19 is critically needed to ensure prolonged responses (12).

Several mechanisms of CART failure have been described, including CART dysfunction (10, 13, 14), an immunosuppressive microenvironment, and tumor-intrinsic alterations (15, 16). Loss of target antigen and persistence, while very common in patients with B-cell acute lymphoblastic leukemia (B-ALL) after CART19 immunotherapy, is less prevalent in patients with B-cell non-Hodgkin lymphoma (B-NHL; ref. 16). Recently, our group identified tumor resistance to apoptosis as a critical mechanism of resistance to CART immunotherapy (17). Indeed, we demonstrated that B-ALL blasts with reduced expression of proapoptotic factors of the extrinsic pathway were resistant to CART19 killing and mediated clinical resistance in patients enrolled in two different clinical trials. Moreover, Upadhyay and colleagues found that the high levels of FAS, a key positive mediator of extrinsic apoptosis, in neoplastic B cells was associated with enhanced survival in lymphoma patients after CART19 treatment (18). However, less is understood about the impact of the intrinsic pathway on CART immunotherapy. We recently demonstrated that the knockout of BH3-interacting domain death agonist (BID), a proapoptotic mediator of both the extrinsic and intrinsic pathways, renders leukemic cells resistant to CART-induced apoptosis (17). Indeed, BID is a proapoptotic initiator protein of the B-cell lymphoma 2 (BCL-2) family that interacts with another BCL-2 family member, BAX, leading to the insertion of BAX into the outer mitochondrial membrane, triggering the release of cytochrome c and the subsequent apoptotic cascade. BID has a role opposite the one of BCL-2, which is a strong antiapoptotic BH-family protein for which there are potent inhibitors in the clinic. Therefore, these findings strongly suggest that inhibition of BCL-2 should enhance CART-mediated apoptosis in cancer cells.

BCL-2 is a key antiapoptotic factor that is overexpressed in several cancers, particularly in B-cell malignancies (19). Overexpression of BCL-2 has been correlated with poor clinical outcomes in multiple neoplasms, including B-NHL, acute myeloid leukemia (AML), melanoma, breast cancer, and prostate cancer (20). BCL-2 is localized on the outer membrane of mitochondria and stabilizes proapoptotic factors such as BAX and BAK that promote cytochrome c and reactive oxygen species release into the cytosol, leading to intrinsic apoptosis (20). Gain of chromosome 18q and translocation (14;18) lead to overexpression of BCL-2 and are frequently found in B-NHL (2123), resulting in reduced apoptosis that is a key step in lymphomagenesis (24). For these reasons, in the last decades, several BCL-2 antagonists have been developed and tested clinically for B-cell malignancies (25). In particular, venetoclax is a BCL-2–selective BH3 mimetic that potently triggers apoptosis in BCL-2–overexpressing cancer cells (26). Venetoclax was approved by the FDA in 2018 as a single agent for relapsed or refractory chronic lymphocytic leukemia (CLL; refs. 27–29) and in 2020 for AML in combination with de­methylating agents and low-dose cytarabine (30). Moreover, several studies demonstrated the clinical activity of venetoclax as a single agent or in combination, even if not FDA approved, in the treatment of mantle cell lymphoma (MCL; refs. 31–33).

In this study, we investigated the role of BCL-2 modulation to enhance CART immunotherapy in several lymphoma and leukemia models. We started with a small-molecule drug screening and then developed a strategy to overcome BCL-2–mediated resistance using in vitro and in vivo modeling. Importantly, although the combination of venetoclax synergistically increased CART-mediated tumor apoptosis in some venetoclax-sensitive lymphoma models, it also resulted in significant dose-dependent toxicity to CART cells. Indeed, BCL-2 plays a critical role during T-cell survival (34, 35). Therefore, to overcome venetoclax-induced CART apoptosis, we developed a strategy to make CART cells resistant to venetoclax by genetically engineering them to express BCL-2 with a venetoclax resistance mutation. We then analyzed clinical cohorts of CART-treated patients and identified BCL2 chromosomal gain or translocation as a poor prognostic factor in CART19-treated patients and also showed that bridging therapy including venetoclax improves the efficacy of CART19 immunotherapy in patients with MCL. Finally, we demonstrated that BCL-2 overexpression intrinsically significantly enhances CART cells’ antitumor activity by promoting their long-term survival. We ensured the safety of BCL-2–overexpressing CART cells by engineering them with a kill switch receptor.

A Proapoptotic Small-Molecule Screening Identifies BCL-2 Inhibitors as Enhancers of CART Cytotoxicity

Acquisition of resistance to apoptosis in cancer cells leads to reduced efficacy of CART cell treatment in the clinical setting (17). We and others previously demonstrated that small molecules that enhance the death receptor pathway, such as inhibitor of apoptosis protein (IAP) antagonists or SMAC mimetics, can also enhance the antitumor efficacy of CART cells in preclinical models (17, 36, 37). In order to screen proapoptotic small molecules that can enhance CART cell killing, we performed a targeted screening assay including a library of 29 proapoptotic drugs (Supplementary Table S1), 11 of which have already been tested in the clinical setting. The library included IAP inhibitors (n = 6), BCL-2 antagonists (n = 13), p53-acting agents (n = 6), caspase activators (n = 2), and ferroptosis activators (n = 2). For this screening, human anti-CD19 CART cells were incubated with CD19+ neoplastic B cells (NALM6) in the presence of two different clinically relevant doses (100 and 1,000 nmol/L) of the drugs or vehicle control [dimethyl sulfoxide (DMSO)]. Tumor killing was measured by luminescence at 48 hours. As shown in Fig. 1A, we identified several proapoptotic small molecules that increased CART cell cytotoxicity, including IAP inhibitors as previously reported (e.g., birinapant, BV-6; refs. 17, 37). Interestingly, in both screenings, the class of BCL-2 inhibitors, particularly the FDA-approved agent venetoclax, demonstrated strong enhancement of CART19 killing (CART alone 47%–63% vs. CART plus BCL-2 inhibitors 75%–88%).

Figure 1.

Venetoclax enhances CART cell–mediated killing of venetoclax-sensitive lymphomas. A, Drug screening of proapoptotic small molecules combined with CART19 against the B-cell leukemia cell line NALM6. Two concentrations of 29 drugs were used (100 nmol/L and 1,000 nmol/L). Killing of NALM6 cells was assessed at 48 hours by luminescence. Two independent screenings, combined data. B, Proposal of CART/venetoclax combination therapy to enhance CART-mediated tumor killing. C, IC50 of venetoclax against several lymphoid malignancy cell lines. Quantification of tumor killing by control untransduced control T cells (UTD) or CART19 in the presence of vehicle (DMSO) or venetoclax (48 hours). Effector-to-target (E:T) ratios = 0.125:1 (OCI-Ly18), 0.06:1 (MINO), 0.125:1 (NALM6), 0.006:1 (primary MCL). Venetoclax concentration = 10 nmol/L (OCI-Ly18 and MINO), 250 nmol/L (NALM6), 3 nmol/L (primary MCL). D, Tumor killing by the combination of venetoclax with CART19 cells that contain either the CD28 or 4–1BB costimulation domain. E:T ratios = 0.1:1. Venetoclax concentration = 20 nmol/L. E, Tumor killing by UTD or CART33 in the presence of vehicle (DMSO) or venetoclax (48 hours). E:T ratios = 0.063:1 (MOLM-14), 0.15:1 (KG-1). Venetoclax concentration = 125 nmol/L (MOLM-14), 50 nmol/L (KG-1). F, Caspase-3/7 activity by flow cytometry. E:T ratio = 0.1:1. Venetoclax concentration = 10 nmol/L. G, Schematic of the xenograft model of venetoclax-sensitive lymphoma (OCI-Ly18). CART cells (2 × 106) were infused via intravenous injection when tumor volume reached ∼150 mm3. Either vehicle or venetoclax (25 mg/kg/daily) was administrated for 3 weeks via oral gavage. Tumor burden over time was measured by caliper. OS was also monitored. All data represent mean ± SD. A two-tailed unpaired Student t test with Welch correction was performed (CF). In G, tumor volume was compared with one-way ANOVA with post hoc Tukey tests, and OS was analyzed using the log-rank (Mantel–Cox) test. All data presented are representative of at least two independent experiments. ns, not significant; *, P < 0.05; **, P < 0.01.

Figure 1.

Venetoclax enhances CART cell–mediated killing of venetoclax-sensitive lymphomas. A, Drug screening of proapoptotic small molecules combined with CART19 against the B-cell leukemia cell line NALM6. Two concentrations of 29 drugs were used (100 nmol/L and 1,000 nmol/L). Killing of NALM6 cells was assessed at 48 hours by luminescence. Two independent screenings, combined data. B, Proposal of CART/venetoclax combination therapy to enhance CART-mediated tumor killing. C, IC50 of venetoclax against several lymphoid malignancy cell lines. Quantification of tumor killing by control untransduced control T cells (UTD) or CART19 in the presence of vehicle (DMSO) or venetoclax (48 hours). Effector-to-target (E:T) ratios = 0.125:1 (OCI-Ly18), 0.06:1 (MINO), 0.125:1 (NALM6), 0.006:1 (primary MCL). Venetoclax concentration = 10 nmol/L (OCI-Ly18 and MINO), 250 nmol/L (NALM6), 3 nmol/L (primary MCL). D, Tumor killing by the combination of venetoclax with CART19 cells that contain either the CD28 or 4–1BB costimulation domain. E:T ratios = 0.1:1. Venetoclax concentration = 20 nmol/L. E, Tumor killing by UTD or CART33 in the presence of vehicle (DMSO) or venetoclax (48 hours). E:T ratios = 0.063:1 (MOLM-14), 0.15:1 (KG-1). Venetoclax concentration = 125 nmol/L (MOLM-14), 50 nmol/L (KG-1). F, Caspase-3/7 activity by flow cytometry. E:T ratio = 0.1:1. Venetoclax concentration = 10 nmol/L. G, Schematic of the xenograft model of venetoclax-sensitive lymphoma (OCI-Ly18). CART cells (2 × 106) were infused via intravenous injection when tumor volume reached ∼150 mm3. Either vehicle or venetoclax (25 mg/kg/daily) was administrated for 3 weeks via oral gavage. Tumor burden over time was measured by caliper. OS was also monitored. All data represent mean ± SD. A two-tailed unpaired Student t test with Welch correction was performed (CF). In G, tumor volume was compared with one-way ANOVA with post hoc Tukey tests, and OS was analyzed using the log-rank (Mantel–Cox) test. All data presented are representative of at least two independent experiments. ns, not significant; *, P < 0.05; **, P < 0.01.

Close modal

BCL-2 Inhibition Using Venetoclax Enhances the Antitumor Effect of CART Cells through Enhanced Caspase-3/7 Cleavage

To investigate whether the administration of venetoclax enhances CART cell–mediated tumor killing (Fig. 1B), we initially used two different B-cell lymphoma cell lines and one leukemia cell line—OCI-Ly18 [diffuse large B-cell lymphoma (DLBCL)], MINO (MCL), and NALM6 (B-ALL)—that have different sensitivities to venetoclax: high for OCI-Ly18 [half-maximal inhibitory concentration (IC50): 18.5 nmol/L], medium for MINO (IC50: 68.17 nmol/L), and low for NALM6 (IC50: 1,300 nmol/L; Fig. 1C; Supplementary Fig. S1A–S1F). We cocultured CART19 cells with either vehicle (DMSO) or venetoclax and measured cytotoxicity at 48 hours. In this short-term model, venetoclax combined with CART19 led to a substantial increase in tumor killing compared with single-agent CART19 or venetoclax and CART19 plus vehicle (Fig. 1C). This effect was further confirmed using primary NHL cells (MCL; Fig. 1C). Of note, inhibition of MCL1, a key negative regulator of intrinsic apoptosis, did not lead to synergy with CART, possibly indicating that MCL1's role in CART-driven toxicity in lymphoma cells is minor as compared with BCL-2 (Supplementary Fig. S2). Furthermore, to confirm the importance of the BCL-2 pathway in resistance to CART killing, we induced overexpression of BCL-2 in B-cell lymphoma and leukemia cell lines (MINO, SU-DHL-4, and NALM6) that lack genetic alterations of BCL2 and accordingly found that BCL2 overexpression led to a significant reduction of tumor killing by CART cell in all models, in particular the venetoclax medium sensitive model (MINO; Supplementary Fig. S3A and S3B).

To test whether the synergistic increase of CART tumor killing by venetoclax was observed independently of the CAR costimulatory domain, we cocultured cancer cells with venetoclax in the presence of CART cells that contained either the CD28 or the 4-1BB costimulatory domain. As shown in Fig. 1D, we found that venetoclax enhanced CART cell–mediated tumor killing regardless of costimulatory domains. To assess if the same effect was also demonstrated in different hematologic cancers, we repeated the experiment using an AML model, a disease for which venetoclax has recently received approval by the FDA (38). AML is an aggressive cancer derived from myeloid progenitors and usually displays a dismal overall survival (OS) despite the best available treatments (39). In the last few years, several CART products have been tested in the clinical setting, including anti-CD33 and anti-CD123 CART cells (40–42). We repeated the in vitro models described above using two AML cell lines and a CD33-targeting CART (CART33). As shown in Fig. 1E, we found that tumor killing by anti-CD33 CART significantly improved when venetoclax was coadministrated in both the MOLM-14 and KG-1 AML cell lines.

In order to investigate the mechanism of enhanced tumor cell death, we measured caspase-3/7 activity in cancer cells cocultured with CART19 cells in the presence or absence of venetoclax. Interestingly, we found that venetoclax treatment led to a synergistic increase of caspase-3/7 activity in NHL cells and to a lesser extent in B-ALL cells when combined with CART19 (Fig. 1F; Supplementary Fig. S4A and S4B). More­over, we investigated the key mediators involved in triggering this enhanced apoptosis. Based on our previous work on BID knockout (KO) tumor cells (17), we expect these cancer cells to be resistant to both FASL/TNFα and perforin/granzyme via direct or indirect mechanisms. We therefore generated several CART cell populations that were knocked out for key triggers of apoptosis: FASL, TRAIL, and granzyme B. Interestingly, we found that the synergy of BCL-2 inhibition and CART killing was significantly diminished when CART cells were knocked out for either FASL or TRAIL (Supplementary Fig. S4C). This result implies that BCL-2 in lymphoma plays an important role in blocking FASL- or TRAIL-mediated apoptosis. To further study the effect of venetoclax on lymphoma cells in vivo during CART19 treatment, we performed single-cell RNA sequencing (scRNA-seq) of lymphoma cells (OCI-Ly18) harvested from mice treated with CART19 or CART19 plus venetoclax (Supplementary Fig. S5A). Lymphoma cells with shared gene expression profiles were clustered using uniform manifold and approximation (UMAP) analysis. We identified six clusters characterized by different cell-cycle phases (Supplementary Fig. S5B and S5C), including one G1-dominant (G1-dom) cluster, two S clusters, one G1–G2 cluster, one G2–M cluster, and an M cluster with high Ki-67 expression. First, we observed a substantially lower proportion of cells assigned to G1-dom in the CART19/venetoclax-treated condition (8.4%) than in the CART19-treated condition (24%). These indicated a prevalent depletion of the G1-dom cluster by the addition of venetoclax (Supplementary Fig. S5D). In accordance with recent reports that venetoclax can induce cell-cycle arrest and death in tumor cells in G1 (43), these results suggest that venetoclax treatment also enhances CART antitumor efficacy by hindering the progression of the cell cycle. Interestingly, the G1-dom and the additional “MKI67hi” cluster (high proliferative cells) showed significant enrichment of genes corresponding to IFNγ responsiveness, suggesting that the cells of these two clusters might have been interacting with CART cells (Supplementary Fig. S5E). Of note, by performing gene ontology enrichment analysis with differentially expressed genes (DEG) between CART19 and CART19/venetoclax combination in the MKI67hi cluster that represents a rapidly proliferating tumor subpopulation, we identified several pathways, including enrichment of the negative regulation of the G2–M phase transition in the CART19/venetoclax-treatment condition in the MKI67hi cluster (Supplementary Fig. S5F and S5G). Taken together, these data implicate that venetoclax treatment enhances CART-mediated tumor killing by promoting tumor apoptosis and inhibiting the cell cycle in cancer cells while also enhancing IFN responses in neoplastic B cells engaged with CART cells.

Finally, to further validate this combination, we used an in vivo B-NHL xenograft model using the DLBCL cell line OCI-Ly18, which is highly sensitive to venetoclax (Fig. 1G). We subcutaneously implanted OCI-Ly18 cells into immunodeficient NOD-SCID gamma chain–deficient (NSG) mice. When the tumor volume reached ∼150 mm2, mice were randomized to receive a suboptimal dose of CART19 (2 × 106 CAR+ cells/mouse, i.v.) in the absence or presence of suboptimal doses of venetoclax (25 mg/kg/daily for 3 weeks, oral gavage). The suboptimal dose of venetoclax was determined based on a venetoclax dose-escalation study (Supplementary Fig. S6A and S6B). Although neither single-agent venetoclax nor CART19 at these doses delayed tumor growth, venetoclax synergistically augmented CART-mediated tumor control (CART19 plus vehicle vs. CART19 plus venetoclax, P = 0.0035), resulting in 100% OS as compared with 0% in the control groups (Fig. 1G). In conclusion, these results demonstrate that combining venetoclax with CART cells could be a promising strategy to improve the clinical outcomes of CART19 therapy in venetoclax-sensitive lymphomas.

Venetoclax Treatment Causes CART Cell Toxicity in the Long Term

Given the fact that in the clinical setting, the sensitivity to venetoclax varies considerably among different lymphoma and leukemia subsets (44, 45), it is crucial to investigate whether the beneficial effect shown in venetoclax-sensitive cell lines would apply to malignancies that have moderate to low sensitivity to venetoclax (Fig. 2A). To this end, we used two xenograft models: the B-cell lymphoma MINO model and the B-ALL NALM6 model that respectively showed intermediate and high resistance to venetoclax in vitro (Fig. 1C; Supplementary Fig. S1A–S1F). We injected NSG mice with luciferase-expressing MINO cells, and on day 14, mice were randomized to receive a relatively low dose of CART19 (5 × 104 cells/mouse) or control T cells [untransduced T cells (UTD)] in combination with venetoclax (50 mg/kg daily, oral gavage for 5 weeks) or vehicle. A higher dose of venetoclax was used because the venetoclax IC50 for MINO is 5-fold higher than that for OCI-Ly18 (Supplementary Fig. S1A–S1F). Interestingly, we found that mice treated with CART19 and venetoclax showed slightly better antilymphoma efficacy early after CART infusion (day 7) compared with mice treated with CART19 alone. However, in the long term, this beneficial effect was lost. In fact, overall, there was no statistical benefit despite the addition of venetoclax (Fig. 2B). We then used the venetoclax-resistant model (NALM6), and, due to the higher resistance to venetoclax of NALM6, a B-ALL cell line, we increased the amount of venetoclax (100 mg/kg daily, oral gavage for 5 weeks). We observed that 40% of the mice (2/5 mice) continuously treated with high doses of venetoclax showed tumor relapse, whereas no evidence of tumor relapse was identified in mice treated with CART19 alone (Fig. 2C). These in vivo findings appeared contradictory to our early short-term in vitro results that showed a benefit of the venetoclax/CART19 combination in virtually all cell lines tested and hinted that higher doses of venetoclax may cause CART cell toxicity. We hypothesized that venetoclax induced apoptosis in CART cells, thereby diminishing their long-term ability to control cancer cells.

Figure 2.

Venetoclax treatment induces CART cell toxicity. A, Schematic of the in vivo xenograft model of venetoclax-resistant tumors. For the MINO model, CART cells (5 × 104) were infused 14 days after luciferase+ MINO cells were implanted (intravenous injection). For the NALM6 model, CART cells (5 × 105) were infused 3 to 4 days after luciferase+ NALM6 cells were implanted (intravenous injection). Either vehicle or venetoclax (50 mg/kg/daily) was administrated for 5 weeks via oral gavage. UTD, untransduced T cells.B and C, Tumor progression of mice bearing MINO (B) or NALM6 (C) cells treated with UTD or CART19 plus either vehicle or venetoclax. D,In vivo CART cell expansion. To quantify CART cell expansion in the NALM6 xenograft model, peripheral mouse blood (PB) was harvested on day 10 after CART cell infusion and analyzed by flow cytometry.E, Quantification of venetoclax-induced CART cell toxicity upon treatment of various doses of venetoclax in vitro (110–10,000 nmol/L). Each dot indicates CART cells generated from different healthy donors (n = 8). Effector-to-target ratio = 0.25:1. Venetoclax concentration = 1,100 nmol/L. All data represent mean ± SD. One-way ANOVA with post hoc Tukey tests was performed (B and C). A two-tailed unpaired Student t test with Welch correction was performed (D and E). All data presented are representative of at least two independent experiments. ns, not significant; *, P  < 0.05; **, P  < 0.01.

Figure 2.

Venetoclax treatment induces CART cell toxicity. A, Schematic of the in vivo xenograft model of venetoclax-resistant tumors. For the MINO model, CART cells (5 × 104) were infused 14 days after luciferase+ MINO cells were implanted (intravenous injection). For the NALM6 model, CART cells (5 × 105) were infused 3 to 4 days after luciferase+ NALM6 cells were implanted (intravenous injection). Either vehicle or venetoclax (50 mg/kg/daily) was administrated for 5 weeks via oral gavage. UTD, untransduced T cells.B and C, Tumor progression of mice bearing MINO (B) or NALM6 (C) cells treated with UTD or CART19 plus either vehicle or venetoclax. D,In vivo CART cell expansion. To quantify CART cell expansion in the NALM6 xenograft model, peripheral mouse blood (PB) was harvested on day 10 after CART cell infusion and analyzed by flow cytometry.E, Quantification of venetoclax-induced CART cell toxicity upon treatment of various doses of venetoclax in vitro (110–10,000 nmol/L). Each dot indicates CART cells generated from different healthy donors (n = 8). Effector-to-target ratio = 0.25:1. Venetoclax concentration = 1,100 nmol/L. All data represent mean ± SD. One-way ANOVA with post hoc Tukey tests was performed (B and C). A two-tailed unpaired Student t test with Welch correction was performed (D and E). All data presented are representative of at least two independent experiments. ns, not significant; *, P  < 0.05; **, P  < 0.01.

Close modal

To investigate this hypothesis, we analyzed the expansion and persistence of CART19 cells in the peripheral blood of NSG mice treated with CART19 plus venetoclax or CART19 alone using flow cytometry. As shown in Fig. 2D, we found that the levels of CART19 cells in the blood of mice treated with venetoclax plus CART19 were lower than those of CART19 alone. These results suggested that venetoclax affects the ability of CART cells to survive and/or proliferate, thus hindering their overall antitumor effect. To test whether prolonged exposure to venetoclax leads to CART cell toxicity, we performed an in vitro venetoclax toxicity assay using CART cells manufactured from eight different T-cell donors. As shown in Fig. 2E, venetoclax caused a significant reduction in the survival of CART19 cells by 5 days of coculture with venetoclax. Of note, the level of toxicity to CART cells varied among the different T-cell donors, likely due to different apoptotic priming statuses at baseline. Interestingly, we observed similar CART cell toxicity when other members of the same family were inhibited in CART cells. Indeed, MCL1 inhibition led to reduced CART survival, suggesting that modulation of mitochondrial-mediated apoptosis is important for CART cell fitness (Supplementary Fig. S7). In order to discern whether the reduced survival was due to increased apoptosis or reduction of proliferation, we assessed the caspase-3/7 activity in CART19 in the presence or absence of venetoclax. Indeed, venetoclax-induced significant caspase-3/7 activation in CART cells, promoting apoptosis (Fig. 2E) and, in so doing, reduced proliferation. Overall, these studies indicate that the higher doses of venetoclax required to suppress neoplasms with venetoclax resistance can cause apoptosis in CART19 cells. However, as the BCL-2 pathway is a critical node for cancer resistance to CART immunotherapy and venetoclax can be toxic to CART cells, we sought to overcome the limitation of targeting BCL-2 combined with CART immunotherapy.

A Novel Strategy to Endow CART Cells with Resistance to Venetoclax

We hypothesized that the development of CART cells with the intrinsic ability to resist venetoclax toxicity would permit the successful combination of venetoclax with CART cells. To this end, we exploited the mechanisms known to drive resistance to venetoclax in leukemia and lymphomas to make the CART cells resistant to venetoclax (46). In particular, previous studies have identified various types of BCL-2 mutations in patients with CLL and B-NHL cell lines that are associated with resistance to venetoclax (47, 48). Of note, a mutant BCL-2 that harbors a point mutation at the 104 amino acid residue (Phe104Leu or F104L) showed strong resistance to venetoclax (47, 49). However, the role of these mutations in T cells and, in particular, CART cells is unknown.

Thus, we engineered a new lentiviral construct that included both CAR19 and the mutated BCL-2(F104L) linked with a 2A self-cleaving peptide (P2A) sequence (Fig. 3A). We first confirmed that the transgenes were correctly expressed in target cells by performing intracellular staining for CAR19 and BCL-2 using flow cytometry (Fig. 3B). We then validated that BCL-2(F104L)–expressing CART19 were indeed functional in killing lymphoma cells and that the short-term synergy with venetoclax was maintained in vitro (Fig. 3C; Supplementary Fig. S8). Most importantly, we performed venetoclax CART cell toxicity assays to evaluate whether the mutant BCL-2 could provide resistance to venetoclax. As shown in Fig. 3D, expression of BCL-2(F104L) successfully rescued CART cells from venetoclax-related toxicity in long-term in vitro assays (i.e., average IC50 value: CART19-BCL-2(F104L): 9027 nmol/L; CART19: 130.7 nmol/L, P = 0.0071). Of note, increased expression of BCL-2 wild-type (WT; used as a control) also provided some degree of CART cell protection from venetoclax toxicity, but the effect was significantly inferior compared with BCL-2(F104L) (i.e., average IC50 value: 997.6 nmol/L). These data suggest that direct inhibition of the attachment of venetoclax to BCL-2 via a point mutation in the binding pocket is an efficient strategy for developing venetoclax-resistant CART cells.

Figure 3.

Expression of mutant BCL-2 prevents venetoclax-mediated CART cell toxicity. A, Strategy to develop venetoclax-resistant CART cells. B, BCL-2 expression in CART cells measured by flow cytometry. C, Quantification of tumor (MINO) killing by untransduced T cells (UTD) or CART19, BCL-2(WT)–expressing CART19 [CART19-BCL-2(WT)], or BCL-2(F104L)–expressing CART19 [CART19-BCL-2(F104L)] in the presence of vehicle (DMSO) or venetoclax. Effector-to-target ratio = 0.06:1. Venetoclax concentration = 10 nmol/L. D, Evaluation of venetoclax-mediated toxicity on either CART19, CART19-BCL-2(WT), or CART19-BCL-2(F104L). CART cell survival (left) and IC50 value (right). Each dot indicates CART cells generated from different healthy donors (n = 3). E, Tumor progression and survival of xenografted mice bearing MINO treated with CART19 or CART19-BCL-2(F104L) plus either vehicle or venetoclax. All data represent mean ± SD. One-way ANOVA with post hoc Tukey tests was performed (C and D). In E, tumor volume was compared with one-way ANOVA with post hoc Tukey tests, and survival was analyzed using the log-rank (Mantel–Cox) test. All data are representative of at least two independent experiments: ns, not significant; *, P  < 0.05; **, P  <  0.01; ***, P  <  0.001.

Figure 3.

Expression of mutant BCL-2 prevents venetoclax-mediated CART cell toxicity. A, Strategy to develop venetoclax-resistant CART cells. B, BCL-2 expression in CART cells measured by flow cytometry. C, Quantification of tumor (MINO) killing by untransduced T cells (UTD) or CART19, BCL-2(WT)–expressing CART19 [CART19-BCL-2(WT)], or BCL-2(F104L)–expressing CART19 [CART19-BCL-2(F104L)] in the presence of vehicle (DMSO) or venetoclax. Effector-to-target ratio = 0.06:1. Venetoclax concentration = 10 nmol/L. D, Evaluation of venetoclax-mediated toxicity on either CART19, CART19-BCL-2(WT), or CART19-BCL-2(F104L). CART cell survival (left) and IC50 value (right). Each dot indicates CART cells generated from different healthy donors (n = 3). E, Tumor progression and survival of xenografted mice bearing MINO treated with CART19 or CART19-BCL-2(F104L) plus either vehicle or venetoclax. All data represent mean ± SD. One-way ANOVA with post hoc Tukey tests was performed (C and D). In E, tumor volume was compared with one-way ANOVA with post hoc Tukey tests, and survival was analyzed using the log-rank (Mantel–Cox) test. All data are representative of at least two independent experiments: ns, not significant; *, P  < 0.05; **, P  <  0.01; ***, P  <  0.001.

Close modal

Next, we sought to confirm the resistance of CART19-BCL-2(F104L) in vivo using MINO (moderate resistance to venetoclax; NHL) and NALM6 (high resistance to venetoclax; B-ALL) xenograft models (Fig. 3E). In the MINO model, we showed that although venetoclax (50 mg/kg) was toxic to CART19, CART19-BCL-2(F104L) showed significant synergy in combination with venetoclax in terms of both tumor control and survival. In particular, the venetoclax combination with CART19-BCL-2(F104L) led to 100% survival, whereas venetoclax combined with control CART19 had no long-term survival (P = 0.0024; Fig. 3E). Of note, the BCL-2(F104L) mutation was confirmed to be protective also in the highly resistant NALM6 model using 100 mg/kg of venetoclax (Supplementary Fig. S9). Indeed, we performed blood flow cytometry on days 10 and 14 after CART infusion and observed no toxicity in CART-BCL-2(F104L). As expected in a B-ALL model not sensitive to venetoclax, the addition of venetoclax to CART19 led to only a minimal enhancement of antitumor activity, drastically lower than the NHL models (Supplementary Fig. S9A and S9B).

These results demonstrate that mutations leading to resistance to venetoclax in cancer cells can be repurposed to induce a similar degree of resistance in CART cells, thereby allowing the development of otherwise toxic CART–drug combinations.

The Clinical Role of BCL2 Chromosomal Alterations in Lymphoma Cells of CART19-Treated Lymphoma Patients

To validate the preclinical discovery that the BCL-2 axis is relevant for response to CART therapy in lymphoma, we analyzed two cohorts of patients with NHL treated with CART19 at the University of Pennsylvania. Based on our preclinical results, we hypothesized that alterations in BCL2 in B-cell lymphomas might contribute to resistance to CART19 immunotherapy in the clinical setting. To test this hypothesis, we retrospectively analyzed the clinical outcomes of a cohort of 87 patients with LBCL treated with FDA-approved CART19 products (tisagenlecleucel and axicabtagene ciloleucel) according to the presence of chromosomal alteration of the BCL2 gene—namely, BCL2 chromosomal translocation t(14;18) (n = 40) or BCL2 chromosomal gain (n = 16)—or its absence (n = 31; Fig. 4A). As shown in Supplementary Table S2, patients from the three groups were balanced for age at infusion, performance status, CAR costimulatory domain used, and disease status at infusion. Importantly, in this group of patients, including DLBCL not otherwise specified (NOS), transformed follicular lymphoma (tFL), double-hit large B-cell lymphoma (BCL), and high-grade BCL (HGBCL) NOS, progression-free survival (PFS) did not change based on the different histologies (P = 0.918; Supplementary Fig. S10A). However, we observed that patients harboring BCL2 translocation t(14;18) and BCL2 gain had an inferior best overall response rate (BORR; 52.5% and 37.5%, respectively) as compared with patients without BCL2 alteration (67.7%; P = 0.195 and P = 0.047, respectively; Fig. 4B). An inferior CR rate was also observed in patients harboring BCL2 gain (31.2%) and BCL2 translocation (40.0%) compared with patients without BCL2 chromosomal alteration (61.3%; Supplementary Fig. S10B). The results were confirmed when looking at the 3-month response rates (Supplementary Fig. S10C). Moreover, at a median follow-up time of 12.6 months, patients with BCL2 chromosomal translocation or gain had lower OS as compared with patients with no alteration of BCL2 (Fig. 4C). Median OS was reached for patients without BCL2 alteration at 15.7 and 15.8 months, respectively, for patients with BCL2 translocation t(14;18) and BCL2 gain (P = 0.009 and P = 0.056, respectively). Patients with BCL2 translocation t(14;18) and BCL2 gain had a significantly shorter OS compared with patients without BCL2 chromosomal alterations, with most patients with BCL2 alteration experiencing treatment failure within 5 months (Supplementary Fig. S10D). The impact of BCL2 chromosomal gain on PFS was validated in a multivariate analysis, including sex, age, disease status at infusion, and presence of BCL2 translocation variables (Supplementary Table S3). Of note, the incidence and entity of other clinical outcomes such as any grade CART-mediated toxicities [e.g., cytokine release syndrome (CRS) and immune effector cell–associated neurotoxicity syndrome (ICANS)] did not correlate with BCL2 alterations (Supplementary Fig. S10E and S10F).

Figure 4.

Chromosomal alterations of BCL2 in lymphoma patients is associated with poor prognosis of CART therapy. A, Schematic description of the strategy to investigate whether genetic alteration of BCL2 affects CART's antitumor clinical response. Pre-CART biopsies from patients with LBCL were analyzed by FISH to search for BCL2 chromosomal aberration. B, BORR of 87 LBCL patients treated with CART19 according to the presence of a BCL2 chromosomal alteration (gain or translocation). PD, progressive disease; PR, partial response; SD, stable disease. C, OS of 87 LBCL patients treated with CART19 according to the presence of a BCL2 chromosomal alteration (gain or translocation). C.I., confidence interval. D, Best overall response of 37 DLBCL patients treated with CART19 according to the presence of a BCL2 chromosomal alteration (gain or translocation). E, OS of 37 DLBCL patients treated with CART19 according to the presence of a BCL2 chromosomal alteration (gain or translocation). F, Schematic description of the strategy to investigate the impact of venetoclax bridging therapy on CART19's clinical response in patients with MCL. G, BORR of 18 MCL patients treated with CART19 according to bridging therapy including venetoclax or not. H, Event-free survival (EFS) of MCL patients treated with CART19 after bridging therapy with (yes) or without (no) venetoclax. Comparisons between the groups were performed with the chi-square test for categorical variables and Student t test for continuous variables, as appropriate. Survival analysis was performed by the Kaplan–Meier estimation and compared with a log-rank test. All statistical tests were two-sided, and statistical significance was defined as P < 0.05. Analysis was performed with the Statistical Package for the Social Sciences software v.22.0.

Figure 4.

Chromosomal alterations of BCL2 in lymphoma patients is associated with poor prognosis of CART therapy. A, Schematic description of the strategy to investigate whether genetic alteration of BCL2 affects CART's antitumor clinical response. Pre-CART biopsies from patients with LBCL were analyzed by FISH to search for BCL2 chromosomal aberration. B, BORR of 87 LBCL patients treated with CART19 according to the presence of a BCL2 chromosomal alteration (gain or translocation). PD, progressive disease; PR, partial response; SD, stable disease. C, OS of 87 LBCL patients treated with CART19 according to the presence of a BCL2 chromosomal alteration (gain or translocation). C.I., confidence interval. D, Best overall response of 37 DLBCL patients treated with CART19 according to the presence of a BCL2 chromosomal alteration (gain or translocation). E, OS of 37 DLBCL patients treated with CART19 according to the presence of a BCL2 chromosomal alteration (gain or translocation). F, Schematic description of the strategy to investigate the impact of venetoclax bridging therapy on CART19's clinical response in patients with MCL. G, BORR of 18 MCL patients treated with CART19 according to bridging therapy including venetoclax or not. H, Event-free survival (EFS) of MCL patients treated with CART19 after bridging therapy with (yes) or without (no) venetoclax. Comparisons between the groups were performed with the chi-square test for categorical variables and Student t test for continuous variables, as appropriate. Survival analysis was performed by the Kaplan–Meier estimation and compared with a log-rank test. All statistical tests were two-sided, and statistical significance was defined as P < 0.05. Analysis was performed with the Statistical Package for the Social Sciences software v.22.0.

Close modal

In order to confirm the role of BCL2 alterations in a more homogeneous cohort of patients, we analyzed a more limited group of patients, focusing only on the DLBCL-NOS histology (n = 37; Supplementary Table S4). In this subpopulation, BORR was inferior among patients harboring BCL2 translocation t(14;18) (50%) and BCL2 gain (18.2%) as compared with patients without BCL2 abnormalities (65.0%; P = 0.508 and P = 0.013, respectively; Fig. 4D). The CR rate of DLBCL patients with BCL2 gain abnormality was significantly inferior compared with patients without BCL2 abnormalities (18.2% vs. 60.0%; P = 0.025; Supplementary Fig. S10G). Moreover, as observed in the parental LBCL cohort, patients with DLBCL-NOS characterized by BCL2 alteration had poorer OS (Fig. 4E). The results were also confirmed by the 3-month response rates (Supplementary Fig. S10H). Of note, no patients with BCL2 gain were in response 1 year after infusion as compared with 50% in the control group without BCL2 chromosomal aberrations. The median PFS was 9.0 months for patients without BCL2 alteration with 2.5 and 3.1 months, respectively, for patients with BCL2 translocation t(14;18) and BCL2 gain (P = 0.936 and P = 0.006, respectively; Supplementary Fig. S10I). The impact of BCL2 chromosomal gain on shorter PFS was validated in a multivariate analysis also in the DLBCL cohort (Supplementary Table S5). As for the previous cohort, the incidence of CRS and ICANS in this group of patients did not correlate with BCL2 alterations (Supplementary Fig. S10J and S10K).

In summary, these two retrospective analyses in a large cohort of lymphoma patients treated with CART19 and preclinical investigation show that chromosomal aberrations of BCL2, in particular BCL2 gain, are associated with reduced response, median PFS, and OS.

Venetoclax Bridging Therapy Is Associated with Better Outcomes after CART19 in MCL

So far we have described a preclinical role for BCL-2 inhibition to enhance CART therapy against lymphoma and presented clinical correlates of BCL2 chromosomal aberrations and CART outcomes. Given that in a specific subset of patients—that is, patients with relapsed or refractory MCL—both venetoclax (50) and CART19 (6) are routinely used in clinical practice, we asked whether bridging therapy with venetoclax would improve CART outcomes. As the concurrent administration of venetoclax during CART19 treatment is not yet approved in the clinical setting, we evaluated the impact of venetoclax exposure as bridging therapy—that is, any antilymphoma therapy that is given between the apheresis and lymphodepletion to control disease progression during CART manufacturing. The hypothesis was that venetoclax would prime tumor cells to CART-mediated apoptosis. We studied 18 patients with MCL who received bridging therapy before the FDA-approved CD28-costimulated CART19 product brexucabtagene autoleucel (brex-cel; Fig. 4F; ref. 6). Of these 18 patients, 8 received bridging therapy, including venetoclax (Supplementary Tables S6 and S7). These patients did not differ in sex, age at infusion, previous treatment with autologous stem cell transplantation, and number of previous lines of therapies as compared with the control group of patients receiving non–venetoclax-based bridging. However, they had higher response rates at infusion (Supplementary Table S6). Of note, we observed that most of the patients treated with venetoclax as bridging therapy achieved a complete response (7/8, 87.5%) after brex-cel, whereas patients receiving non–venetoclax-based bridging therapy displayed a rate of 50% CR (5/10; P = 0.094; Fig. 4G). Moreover, the event-free survival of patients receiving venetoclax bridging therapy was longer as compared with patients not receiving venetoclax (P = 0.018), with 100% of patients taking venetoclax achieving CR at 1 year as compared with ∼60% in the control group (Fig. 4H). Taken together, these results validate the BCL-2 pathway as a critical node in patients with lymphoma receiving CART19 immunotherapy.

BCL-2 Overexpression in CART Cells Enhances Their Antitumor Effect

In the previously described studies, we observed that, in addition to reducing venetoclax toxicity, BCL-2(F104L) overexpression in CART cells inherently increased their ability to control tumors even in the absence of venetoclax (Fig. 3E, red dotted line). Therefore, we speculated that BCL-2(WT) expression in CART cells enhances their survival and long-term persistence, leading to a higher therapeutic index. To study the mechanism by which constitutive BCL-2 expression might improve CART cell antitumor activity, we performed in vitro and in vivo CART cell functional studies (Fig. 5A). As shown in Fig. 5B and 5C, CART19-BCL-2(WT) cells showed substantial enhancement of their antitumor activity against both MINO (MCL) and NALM6 (B-ALL) in vivo in mouse xenograft models. Furthermore, we observed remarkable expansion of CART19-BCL-2(WT) in the blood of mice as compared with CART19 (Fig. 5D). Notably, upon tumor clearance, the levels of CART19-BCL-2(WT) in the blood decreased, indicating the absence of uncontrolled proliferation in this model (Fig. 5E). Mechanistically, we observed no apparent dysfunction related to BCL-2 expression of the in vitro antitumor activity of CART cells; cytotoxicity and cytokine production were not different (Supplementary Fig. S11A–S11C). Given that BCL-2 expression is higher in memory T cells than in effector T cells (51), we monitored whether constant expression of BCL-2 affected the differentiation status of CART cells after stimulation. As shown in Supplementary Fig. S12, there was no significant difference in the frequency of CART cell differentiation over time upon CART activation. In contrast, we observed that BCL-2 overexpression provided CART cells with a substantial advantage in long-term survival in vitro (Fig. 5F). Of note, these long-survived CART cells still showed substantial antitumor activity, as evidenced by the fact that in the long term, they can still respond to phorbol- 12-myristate-13-acetate (PMA)/ionomycin by secreting multiple cytokines (Supplementary Fig. S13).

Figure 5.

Overexpression of BCL-2(WT) in CART cells enhances their antitumor efficacy. A, Schematic of the in vivo xenograft model to study the effect of BCL-2 overexpression on CART antitumor activity. UTD, untransduced T cells. For the MINO model (B), CART cells (5 × 104) were infused 14 days after luciferase+ MINO cell intravenous injection. For the NALM6 model (C), CART cells (5 × 105) were infused 3 to 4 days after luciferase+ NALM6 cell intravenous injection. B and C, Tumor progression and OS over time in mice bearing MINO (B) and NALM6 (C) treated with CART19 or CART19-BCL-2(WT) (representative of two replicate experiments, n = 5). D, Quantification of CART cell peak expansion in mouse blood collected from CART-treated mice bearing NALM6 on day 10 after CART cell infusion by flow cytometry. E, CART cell persistence in CART-treated mouse blood over time by flow cytometry (NALM6 model).F, Fold change of CART cell upon stimulation with irradiated MINO (representative of two replicate experiments). G, Volcano plot showing DEGs in CART19-BCL-2(WT) compared with CART19 on day 18 after stimulation with irradiated MINO. H, GSEA of DEGs in CART19-BCL-2(WT) compared with CART19 on day 18 after stimulation with irradiated MINO. NES, normalized enrichment score. I, Survival of CART cells after withdrawal of cytokines. CART cells were stimulated with irradiated MINO for 48 hours, and culture media were replaced with fresh media to withdraw cytokines. Survival of CART cells was monitored by flow cytometry 48 hours after adding fresh media. All data represent mean ± SD. One-way ANOVA with post hoc Tukey tests was performed for all comparisons. OS was analyzed using the log-rank (Mantel–Cox) test (B and C). All data presented are representative of at least two independent experiments except bulk RNA-seq (performed once with two biological replicates). *, P  < 0.05; **, P  < 0.01.

Figure 5.

Overexpression of BCL-2(WT) in CART cells enhances their antitumor efficacy. A, Schematic of the in vivo xenograft model to study the effect of BCL-2 overexpression on CART antitumor activity. UTD, untransduced T cells. For the MINO model (B), CART cells (5 × 104) were infused 14 days after luciferase+ MINO cell intravenous injection. For the NALM6 model (C), CART cells (5 × 105) were infused 3 to 4 days after luciferase+ NALM6 cell intravenous injection. B and C, Tumor progression and OS over time in mice bearing MINO (B) and NALM6 (C) treated with CART19 or CART19-BCL-2(WT) (representative of two replicate experiments, n = 5). D, Quantification of CART cell peak expansion in mouse blood collected from CART-treated mice bearing NALM6 on day 10 after CART cell infusion by flow cytometry. E, CART cell persistence in CART-treated mouse blood over time by flow cytometry (NALM6 model).F, Fold change of CART cell upon stimulation with irradiated MINO (representative of two replicate experiments). G, Volcano plot showing DEGs in CART19-BCL-2(WT) compared with CART19 on day 18 after stimulation with irradiated MINO. H, GSEA of DEGs in CART19-BCL-2(WT) compared with CART19 on day 18 after stimulation with irradiated MINO. NES, normalized enrichment score. I, Survival of CART cells after withdrawal of cytokines. CART cells were stimulated with irradiated MINO for 48 hours, and culture media were replaced with fresh media to withdraw cytokines. Survival of CART cells was monitored by flow cytometry 48 hours after adding fresh media. All data represent mean ± SD. One-way ANOVA with post hoc Tukey tests was performed for all comparisons. OS was analyzed using the log-rank (Mantel–Cox) test (B and C). All data presented are representative of at least two independent experiments except bulk RNA-seq (performed once with two biological replicates). *, P  < 0.05; **, P  < 0.01.

Close modal

To understand the mechanism for this enhanced antitumor activity, we isolated RNA from CART19 or CART19-BCL-2(WT) 18 days after in vitro activation and performed bulk RNA-seq analysis. As shown in Fig. 5G and Supplementary Table S8, we identified a total of 304 genes that were differentially expressed in CART19-BCL-2(WT) compared with CART19 (upregulated: 117 genes, downregulated: 187 genes; Fig. 5G). Using these DEGs, we performed gene set enrichment analysis (GSEA) and found that CART19-BCL-2(WT) showed downregulation of genes strongly correlated with pathways related to apoptosis, which might explain the enhanced survival (Fig. 5H, left). Moreover, we observed increased expression of genes in the JAK–STAT pathway and IFNα response (Fig. 5H, right), which might indicate a prosurvival phenotype and higher gamma-receptor cytokine-mediated signaling despite the reduced availability of cytokines in this long-term coculture (day 18). Indeed, previous studies in murine T cells showed that overexpression of BCL-2 allowed T cells to survive without gamma-chain receptor cytokines such as IL2 (52) or IL7 (53). To functionally test this hypothesis, we tested whether BCL-2 overexpression rescued CART cells in the absence of cytokines, which are essential for their survival/expansion. Remarkably, in line with previous reports, we observed that CART-BCL-2(WT) cells survived better compared with CART19 cells when cytokines were withdrawn from their culture media, suggesting that BCL-2 overexpression can protect CART cells in the absence of survival signals derived from cytokines, likely through the enhanced JAK–STAT survival signaling pathway (Fig. 5I).

Higher BCL2 RNA Levels in Apheresis Products Correlate with Improved Outcomes after CART19, with Prolonged CART Persistence

Given these results, we then hypothesized that the expression of BCL2 in patient T cells might be associated with improved outcomes after CART19 immunotherapy due to enhanced CART fitness. To address this hypothesis, we analyzed gene expression in T cells from the apheresis products of 38 patients with B-NHL who received CART19 immunotherapy (CTL019, now known as tisagenlecleucel) in the pilot clinical trial NCT02030834 (54) and correlated with long-term outcomes (over 5 years; Fig. 6A). As shown in Fig. 6B, using the NanoString nCounter platform, we observed that BCL2 was among the top genes that were significantly enriched in patients who achieved a CR after CART compared with patients with no response (NR). In addition, absolute BCL2 levels were higher in patients with either CR or PR as compared with NR (Fig. 6C). Moreover, we also identified that BCL2 expression in T cells correlated with prolonged CART persistence (Fig. 6D, P = 0.0005) but not CART peak expansion (Supplementary Fig. S14A), as observed in the preclinical models (Fig. 5F). Finally, we found that the expression level of BCL2 in T cells was significantly correlated with prolonged OS of patients (Fig. 6E, P < 0.0001) but not PFS (Supplementary Fig. S14B). These results suggest that higher levels of BCL2 in the T cells from apheresis products are associated with improved clinical results of CART19.

Figure 6.

Increased BCL-2 expression in T cells from CART apheresis products is associated with positive clinical outcomes in patients with lymphoma in the long term. A, Schematic description of the approach taken to investigate the relationship between the level of BCL-2 and CART clinical response. RNA was extracted from T cells from apheresis products of 38 patients with lymphoma who received CART19 immunotherapy (CTL019, now known as tisagenlecleucel) in the clinical trial NCT02030834. Next, BCL2 mRNA expression was quantified via the nCounter analysis system (NanoString). B, Volcano plot showing differential gene expression in T cells based on best overall response (CR or NR). C, Comparison of BCL2 expression in T-cell apheresis products of CART19-treated patients in CR/partial response (PR) versus NR. D, Correlation of BCL2 expression in T-cell apheresis products with CART persistence. E, Correlation of BCL2 expression in T-cell apheresis products with OS. F, Monitoring of abnormal CART expansion mediated by constant overexpression of BCL-2 [left: CART expansion (fold change), right: frequency of CART (%)]. UTD, untransduced T cell. G, Cytotoxicity on CART19 and CART19-BCL-2(WT) 24 hours after treatment of chemotherapy (doxorubicin, 300 and 1,000 nmol/L). H, Cytotoxicity of anti-CD19 CART cells expressing truncated EGFR (CART19-tEGFR) and CART19 expressing BCL-2(WT) and truncated EGFR [CART19-BCL-2(WT)-tEGFR] after 24-hour treatment with either isotype control or anti-EGFR antibody (cetuximab). Ab, antibody. All data represent mean ± SD. Linear regression analysis was performed (D and E). A two-tailed unpaired Student t test with Welch correction was performed (C, F, G, and H). ns, not significant; *, P  < 0.05; **, P  < 0.01.

Figure 6.

Increased BCL-2 expression in T cells from CART apheresis products is associated with positive clinical outcomes in patients with lymphoma in the long term. A, Schematic description of the approach taken to investigate the relationship between the level of BCL-2 and CART clinical response. RNA was extracted from T cells from apheresis products of 38 patients with lymphoma who received CART19 immunotherapy (CTL019, now known as tisagenlecleucel) in the clinical trial NCT02030834. Next, BCL2 mRNA expression was quantified via the nCounter analysis system (NanoString). B, Volcano plot showing differential gene expression in T cells based on best overall response (CR or NR). C, Comparison of BCL2 expression in T-cell apheresis products of CART19-treated patients in CR/partial response (PR) versus NR. D, Correlation of BCL2 expression in T-cell apheresis products with CART persistence. E, Correlation of BCL2 expression in T-cell apheresis products with OS. F, Monitoring of abnormal CART expansion mediated by constant overexpression of BCL-2 [left: CART expansion (fold change), right: frequency of CART (%)]. UTD, untransduced T cell. G, Cytotoxicity on CART19 and CART19-BCL-2(WT) 24 hours after treatment of chemotherapy (doxorubicin, 300 and 1,000 nmol/L). H, Cytotoxicity of anti-CD19 CART cells expressing truncated EGFR (CART19-tEGFR) and CART19 expressing BCL-2(WT) and truncated EGFR [CART19-BCL-2(WT)-tEGFR] after 24-hour treatment with either isotype control or anti-EGFR antibody (cetuximab). Ab, antibody. All data represent mean ± SD. Linear regression analysis was performed (D and E). A two-tailed unpaired Student t test with Welch correction was performed (C, F, G, and H). ns, not significant; *, P  < 0.05; **, P  < 0.01.

Close modal

Preclinical Safety of CART19(F104L) and Mitigation Strategies

Although BCL-2 overexpression led to dramatic improvement of CART cell antilymphoma activity, a critical issue for this approach could be long-term safety. Indeed, although it is not considered an independent driving factor in lymphomagenesis, BCL-2 overexpression might lead to uncontrolled CART cell proliferation and potentially T-cell transformation (55). Of note, in our studies, the increased CART19 proliferation did not result in an abnormal expansion of these cells in mice; importantly, CART19-BCL-2(WT) do contract in the absence of the target antigen (Fig. 5E). In addition, we performed a proliferation study aimed at assessing the ability of CART19-BCL-2(WT) to proliferate without antigen stimulation in the presence or absence of cytokines such as IL7 and IL15. As shown in Fig. 6F and Supplementary Fig. S15A and S15B, we identified that CART19-BCL-2(WT) showed similar expansion capability to unmodified CART cells. In addition, we observed a decrease of CART expansion in the peripheral blood upon tumor eradication (Fig. 5E).

Nevertheless, to investigate the safety profile of this approach, we investigated whether CART19-BCL-2(WT) cells are still sensitive to conventional cytotoxic drugs such as chemotherapy (e.g., doxorubicin). As shown in Fig. 6G, regardless of the constitutive expression of BCL-2, clinical doses of doxorubicin resulted in fast and effective elimination of CART19-BCL-2(WT). Furthermore, to enhance the safety profile of CART19-BCL-2(WT) cells, we inserted a CART suicide system by expressing truncated EGFR into CART19-BCL-2(WT) (56). We tested whether anti-EGFR antibodies can mediate antibody-dependent cellular cytotoxicity, thereby eradicating CART19-BCL-2(WT), and we demonstrated that CART19-BCL-2(WT) cells were successfully eliminated using anti-EGFR antibodies (Fig. 6H).

In summary, we demonstrated that the constitutive expression of BCL-2 provides significant enhancement of CART cell survival and expansion, which in turn improves their overall antitumor activity in several combination models. In addition, our studies revealed that both conventional lymphocyte-depleting agents and targeted antibody-mediated depletion may be used as a clinical regimen to deplete BCL-2–expressing CART cells in patients if ever necessary.

CART cells achieve their antitumor activity by triggering apoptosis in cancer cells through multiple mechanisms, including engaging death receptors with FASL, TRAIL, TNFα (“extrinsic” apoptosis) and releasing perforin and granzymes (“intrinsic” apoptosis). Not surprisingly, reduced apoptosis plays an essential role in cancer's resistance to treatments, including CART cell immunotherapy (57–60). Indeed, we previously demonstrated that lower sensitivity of B-ALL cells to extrinsic apoptosis in patients treated with CART correlated with reduced antitumor activity (17). Other groups have also established the relevance of this pathway in the setting of adoptive cell therapies (18, 36, 37). We have also identified that reduced BID—a key modulator of apoptosis involved in both the extrinsic and extrinsic pathways—leads to resistance to CART killing in vitro and in vivo (17). As BID plays a role that is opposite of BCL-2, we hypothesized that inhibiting BCL-2 could lead to enhanced CART killing. In this project, we sought to investigate the role of BCL-2 and intrinsic apoptosis in resistance to CART immunotherapy. CART immunotherapy involves both CART cells and tumors, and both of these players can drive resistance; therefore, we studied the role of BCL-2 in both cancer cells and CART cells.

We first investigated the role of BCL-2 inhibition in lymphoma as a strategy to enhance CART cell cytotoxicity. We first used a small-molecule screening and identified BCL-2 inhibitors and, in particular, venetoclax as agents capable of enhancing CART cell killing. We then tested the combination of venetoclax with CART therapy and found that venetoclax significantly increased the antitumor efficacy of CART cells in several B-cell leukemia and lymphoma models. Moreover, we showed that venetoclax enhances the antitumor efficacy of anti-CD33 CART immunotherapy against AML. Interestingly, Karlsson and colleagues previously demonstrated that the addition of another BCL-2 family protein inhibitor, ABT-737, led to a significant increase of CART-mediated tumor killing as compared with CART treatment alone, but that study showed only one lymphoma model and was limited to the in vitro setting (61); furthermore, ABT-737 is not a clinically viable agent, as it is not bioavailable upon oral administration. Recently, Yang and colleagues showed that presensitization of tumors by venetoclax enhanced CART19-mediated tumor killing (62). However, the study was again focused only on a single model, including a non–FDA-approved CART product (CART20), and the concurrent administration or the long-term effects of BCL-2 inhibition on CART were not studied. In this project, we took a rigorous, comprehensive, and translational approach to study the effects of venetoclax in CART immunotherapy in multiple models, looking at the effect on both tumor cells and CART cells. Moreover, we described the effect of venetoclax in different CAR costimulatory domains, using both lymphoid and myeloid neoplasm models. Lastly, we studied patients receiving CART19 and found that BCL2 gain was associated with poor outcomes, and pre-exposure to venetoclax in patients with MCL enhances the complete response rates of CART19.

Given that BCL-2 plays an important role in the survival and proliferation of lymphoid cells, we then sought to study the effect of BCL-2 antagonism not only on cancer cells but also on CART cells. We found that prolonged exposure to venetoclax at mid- and high doses caused CART cell apoptosis, which eventually resulted in the failure of the CART–venetoclax combination in vivo. In line with our results, Kohlhapp and colleagues demonstrated that venetoclax treatment significantly reduces the survival of T cells in the setting of anti–PD-1 antibody treatment (63). Early studies showed that Bcl2−/− mice exhibit rapid loss of peripheral T cells (36, 37), and it was shown that activated T cells that exhibit increased susceptibility to apoptosis in vitro express lower levels of BCL2 (64). Based on our results and the published literature, it became clear that prolonged BCL-2 inhibition would drive reduced CART cell functionality. This finding is of particular clinical relevance, as it warns against combining high doses of venetoclax with CART cells. For instance, a dose of 400 mg/day of venetoclax leads to a tissue concentration of about 2.3 μmol/L (65), which, based on our data, is expected to be toxic to T cells long term. Instead, lower doses (e.g., 50 mg) might be safer. Of course, these results need to be tested in patients, and a clinical trial includes one experimental arm testing the combination of CART19 (liso-cel) and venetoclax (50 mg) has been used in CLL at our and other institutions (NCT03331198), with the potential effect on T cells being actively monitored.

In order to overcome the risk of CART toxicity by venetoclax and increase the therapeutic index of this combination, we evaluated novel strategies to endow CART cells with resistance to venetoclax. To achieve this goal, we developed a venetoclax-resistant CART cell product by expressing in CART cells a mutation found in cancer cells resistant to venetoclax (F104L) and validated the results in multiple in vivo models. This approach introduces a novel concept for combination therapies: engineering CART cells to make them resistant to proapoptotic small molecules and thereby enhancing their therapeutic index. In the future, this strategy could potentially be used for other combinations in which small molecules display toxicity on CART cells or in general T cell–based immunotherapies. For example, CART cells could be made resistant to small-molecule inhibitors such as vemurafenib (66), midostaurin (67), or PI3K inhibitors (68) that are known to affect T-cell functions.

Lastly, we sought to study clinical correlates for the role of BCL-2 in CART treatment. We found that BCL2 gain in lymphoma biopsies before CART treatment was associated with poor prognosis. Interestingly, BCL2 translocation, while showing reduced best overall response and median PFS, was less strikingly associated with poor outcomes compared with no BCL-2 alteration. This is likely due to the fact that despite the enhanced expression of BCL2 found in lymphomas carrying the t(14;18) translocation, they usually do not respond to venetoclax inhibition, indicating that other factors besides BCL2 (e.g., BIM) and oncogenic drivers are critical for these diseases for lymphomagenesis and lymphoma survival. Furthermore, we showed that venetoclax treatment as a bridging therapy before CART infusion led to higher response rates to CART19 in MCL. This effect is likely mediated by a “priming” effect of venetoclax on MCL cells leading to enhanced CART killing.

We then discovered that constant expression of BCL-2 in CART cells can improve CART cells’ long-term survival and antitumor activity independently of venetoclax exposure. Interestingly, other groups (52, 69) have shown in other models that BCL-2 expression can enhance adoptive T-cell immunotherapy. However, in this study, we approached this hypothesis in clinically relevant models and identified the specific mechanism of this effect. When T cells respond to an antigen in vivo, they become activated and divide, and then most of them rapidly die. BCL-2 expression in T cells protects them from death and prolongs their survival (70). Indeed, our results showed that BCL-2 overexpression in CART cells enhances their antilymphoma effect by reducing apoptosis. In addition, we observed that constant expression of BCL-2 upregulated genes associated with the JAK–STAT/type I IFN pathway. Higher activation of these pathways might be related to the higher capability of BCL-2–expressing CART to survive in the absence of cytokines that signal through the gamma-chain receptor. This finding could have important implications in the clinical translation of this approach against cancer, as there is a lack of T-cell supportive cytokines found in the cancer tumor microenvironment (71). Moreover, CART-BCL-2(WT) might require less intensive lymphodepleting regimens mainly by increasing the available T-cell cytokine pool (72, 73). We then observed that the levels of BCL-2 in the T cells from the apheresis products from NHL patients treated with CART19 correlated with clinical outcomes; in particular, BCL-2 levels were higher in responders. This is a new finding that might suggest that basal levels of BCL-2 in T cells provide CART19 through higher resistance to apoptosis and likely longer persistence.

Lastly, as BCL-2 overexpression is a key feature of lymphomagenesis and could potentially cause T-cell lymphomas (55), we aimed to study and enhance the safety of this approach. First, we demonstrated that BCL-2–modified CART cells do not proliferate uncontrollably in vitro and in mice, then we showed that they are effectively eliminated by conventional chemotherapy (i.e., doxorubicin), and finally we generated CART cells carrying a suicide system by expressing truncated EGFR that can be recognized by a clinical-grade anti-EGFR antibody (56). This approach would provide this product with a safety switch that could be triggered if clinically needed. Despite the fact that we can control CART expansion via a safety switch, it is important to monitor alteration of CART expansion mediated by overexpression of BCL-2. Therefore, we monitored xenogeneic graft-versus-host disease (xenoGVHD) clinical scores, weight, and survival in all our animal models. Although we observed GVHD in some animals, especially the ones with high CART proliferation, including in both CART19 and CART19-BCL-2(WT) mice, it is common to see xenoGVHD in NSG xenograft models. In order to avoid this consideration, the initial clinical application of this new technology will be in the autologous setting; therefore, we expect very low chances of GVHD. Nevertheless, in the future, controlled expression of BCL-2 or TCR KO would be needed for allogeneic use.

In conclusion, these studies indicate that the therapeutic index of CART cells can be significantly increased by modulation of BCL-2–mediated apoptosis in cancer cells (enhancement of apoptosis) and CART cells (prevention of apoptosis) and paves the way for future clinical studies, including next-generation CART products that are resistant to small molecules to overcome cancer resistance.

Cell Lines and General Cell Culture

Six B-cell malignant cell lines were used (B-ALL: NALM6; MCL: MINO, Z-138, and MAVER; and DLBCL: OCI-Ly18 and SU-DHL-4). Two AML cell lines were used (MOLM-14 and KG-1). Unless otherwise specified, cells were grown and cultured at a concentration of 1 × 106 cells/mL of standard culture media (RPMI 1640 + 10% FBS, 1% penicillin/streptomycin, 1% HEPES, 1% GlutaMAX) at 37°C in 5% ambient CO2. All cell lines were originally obtained from ATCC or DSMZ, authenticated (University of Arizona Genetics Core, 2019), and tested for Mycoplasma contamination (Lonza). Primary MCL samples were obtained from the clinical practices of the Hospital of the University of Pennsylvania (UPCC55418).

Lentiviral Vector Production and Transduction of CAR-Engineered Human T Cells

Replication-defective, third-generation lentiviral vectors were produced using HEK293T cells (ATCC ACS-4500). Approximately 7 × 106 to 9 × 106 cells were plated in T150 culture vessels in standard culture media and incubated overnight at 37°C. The next day, cells were transfected using a combination of Lipofectamine 2000 (116 μL, Invitrogen); pMDG.1 (7 μg), pRSV.rev (18 μg), and pMDLg/p.RRE (18 μg) packaging plasmids; and 15 μg of expression plasmid (CAR). Lipofectamine and plasmid DNA were diluted in 4 mL Opti-MEM media prior to transfer into lentiviral production flasks. At both 24 and 48 hours following transfection, culture media were isolated and concentrated using high-speed ultracentrifugation (8,000 × g overnight). Human T cells were procured through the University of Pennsylvania Human Immunology Core. CD4+ and CD8+ cells were combined at a 1:1 ratio and activated using CD3/CD28 stimulatory beads (Thermo Fisher) at a ratio of 3 beads/cell and incubated at 37°C overnight. The following day, CAR lentiviral vectors were added to stimulatory cultures at a multiplicity of infection between 1 and 3. Beads were removed on day 6 of stimulation, and cells were counted every other day until growth kinetics and cell size demonstrated they had rested from stimulation (cell volume: ∼350 fL). All experiments used a CAR19 encoding the CTL019 CAR, composed of the FMC63 scFv, 4-1BB, and CD3ζ domains, unless otherwise noted (4). To validate the combination of venetoclax with different CAR constructs, anti-CD19 CART cells with CD28/CD3ζ domains (74) and anti-CD33 CART cells with 4-1BB/CD3ζ domains were generated (41). To develop venetoclax-resistant CART cells, antiapoptotic genes [BCL-2(WT) and BCL-2(F104L)] were cloned into CAR19 followed by a P2A self-cleavage sequence. To generate BCL-2–overexpressing B-cell malignant cell lines, a lentiviral vector encoding BCL-2(WT) was obtained from Addgene.

Clinical Specimens

For the LBCL cohort, we collected clinical data from patients diagnosed with DLBCL NOS, HGBCL NOS, HGBCL with MYC and BCL-2 and/or BCL-6 rearrangements, and tFL treated at the University of Pennsylvania using two commercial CART19 products (tisagenlecleucel or axicabtagene ciloleucel, UPCC44420) or enrolled in the CTL019 clinical trial, NCT02030834 (75, 76). Only patients evaluated for chromosomal alterations involving the BCL2 locus by interim FISH analysis were included in the current study. For the MCL cohort, we collected clinical data from patients diagnosed with MCL treated with commercial brex-cel in the commercial setting (UPCC44420). Disease response was determined according to Lugano classification (77). PFS time was defined as the time between CART19 infusion to date of progression (event), death of any cause (event), or last follow-up up to 24 months after infusion (censoring). Relapse-free survival was defined as the time between CART19 infusion to date of progression (event) or last follow-up up to 24 months after infusion (censoring). OS time was defined as the time between CART19 infusion to date of death (event) or last follow-up up to 24 months after infusion (censoring). CRS and ICANS were graded according to the consensus grading criteria defined by Common Terminology Criteria for Adverse Events (for NCT02030834) and the American Society of Transplantation and Cellular Therapy classification (for patients receiving commercial CART; ref. 78). The gene expression profile study using NanoString nCounter was performed on 38 patients enrolled in the CTL019 clinical trial, NCT02030834 (75, 76). All patients provided written informed consent to participate in the study. The study was approved by the Institutional Review Board and was conducted in accordance with the ethical standards of the 1964 Declaration of Helsinki and its later amendments.

Targeted Small-Molecule Screening

We seeded CART19 and NALM6 (luciferase+) cells at an effector-to-target (E:T) ratio of 0.08:1 (i.e., 600 CART:7,000 NALM6) per well in 25 μL of growth medium (RPMI 1640 + 10% FBS + 1% penicillin and streptomycin + 1% glutamine) of a 384-well Corning 3570 microplate using a Multidrop Combi Reagent Dispenser (Thermo Scientific). Following cell seeding, drugs (50 nL) were transferred to assay plates using a 50-nL slotted pin tool (V&P Scientific) and a JANUS Automated Workstation (PerkinElmer). Compounds/drugs were added to a final concentration of 1 μmol/L in 0.2% DMSO. Columns 1 and 23 were treated with 0.2% DMSO (negative control). Columns 2 and 24 were treated with 50 nmol/L of bortezomib (positive control). Cells were incubated for 48 hours at 37°C in 5% CO2 in a humidified chamber. Assay plates were removed from the incubator for 1 hour to equilibrate to room temperature prior to adding 25 μL of 0.25× Britelite (PerkinElmer). Luminescence was measured on an EnVision Xcite Multilabel Plate Reader (PerkinElmer) using the ultrasensitive luminescence measurement technology.

Bioluminescence-based cytotoxicity assays

Cell lines (MINO, Z-138, MAVER, OCI-Ly18, SU-DHL-4, NALM6, MOLM-14, and KG-1) were engineered to express click beetle green, and cell survival was measured using bioluminescence quantification. D-luciferin potassium salt (PerkinElmer) was added to cell cultures (final concentration 15 μg/mL) and incubated at 37°C for 10 minutes. Bioluminescence signal was detected using a BioTek Synergy H4 imager, and the signal was analyzed using BioTek Gen5 software. Percent specific lysis was calculated using a control of target cells without effectors. Cytotoxicity assays were established as previously described (17) with the addition of vehicle or venetoclax.

Flow Cytometry Assays

Cells were resuspended in FACS staining buffer (PBS + 2% FBS) using the following antibodies: human CD3 (clone OKT3, BioLegend), anti–BCL-2 (clone 100, BioLegend), human CD45 (clone 2D1, BioLegend), and mouse CD45 (clone 30-F11, BioLegend). CART19 was detected using PE-conjugated anti-CAR19 idiotype antibody (Novartis). To monitor caspase-3/7 activity, CellEvent Caspase-3/7 Green Read Flow reagent was used following the manufacturer's protocol. To determine the absolute cell numbers (tumor or T cells) acquired during flow cytometry, CountBright absolute counting beads (Thermo Fisher) were used. Cell viability was established using Live/Dead Aqua or violet fixable staining kit (Thermo Fisher), propidium iodide, and 7-aminoacctinomycine D (7-AAD), and data were acquired on an LSRII Fortessa Cytometer (BD). Intracellular staining was performed by using fixation/permeabilization buffer and following the manufacturer's protocol. All data analyses were performed using FlowJo 9.0 or 10 software (FlowJo, LLC).

Long-term Coculture Assays

CART cells were combined with target cancer cells at an E:T ratio of 0.25:1, and cocultures were evaluated for an absolute count of T cells and cancer cells by flow cytometry using CountBright absolute counting beads (Thermo Fisher) every 3 days. Cultures were maintained at a concentration of 1 × 106 total cells/mL. To monitor their differentiation status, CART cells were harvested on days 0, 9, and 18. Next, CART cells were stained with anti-CCR7 and anti-CD45RA antibodies for flow-cytometric analysis. CART cells were restimulated by PMA/ionomycin on day 18 after initial stimulation in order to evaluate the antitumor activity of long-survived CART cells.

Xenograft Mouse Models

Six- to 10-week-old NSG mice were obtained from the Stem Cell and Xenograft Core at the University of Pennsylvania and maintained in pathogen-free conditions. To establish the OCI-Ly18 subcutaneous xenograft mouse model, 5 × 106 of OCI-Ly18 were prepared in 200 μL of PBS containing 50% Matrigel (Corning) and implanted into the flank of NSG mice via subcutaneous injection. Suboptimal doses of CARTs (2 × 106 CAR+ cells) were then introduced via intravenous injection when tumor volumes reached ∼150 mm3. For the systemic tumor model, 1 × 106 of either NALM6 or MINO were administrated to NSG mice by tail-vein injection. When bioluminescence intensity in NSG mice reached ∼107 [total flux (p/s)], either 5 × 104 CAR19+ cells or 5 × 105 CAR19+ cells were injected into MINO-bearing mice or NALM6-bearing mice, respectively. OCI-Ly18 tumors were measured every week by caliper, and tumor volume was calculated according to the equation: tumor volume  =  ½ (L  ×  W2), where L is the longest axis of the tumor and W is the axis perpendicular to L. NALM6 and MINO were monitored over time using the Xenogen IVIS bioluminescence imaging system. In the venetoclax combination studies, venetoclax was prepared in a solution containing 5% DMSO, 40% PEG300, 5% Tween 80, and 50% PBS. Different doses of venetoclax were used as indicated in each figure. Animals were monitored for signs of disease progression and overt toxicity, such as xenoGVHD, as evidenced by >10% loss in body weight, fur loss, diarrhea, conjunctivitis, and disease-related hind limb paralysis. All animal care and use were followed by NIH guidelines, and all experimental protocols were approved by the University of Pennsylvania Animal Care and Use Committee.

nCounter Gene Expression Assays

CART cells were combined with irradiated MINO cells for 48 hours at an E:T ratio of 0.25:1. For the clinical samples, frozen mononuclear cells from apheresis were thawed and T cells were isolated using the Pan T-Cell Isolation Kit (Miltenyi Biotec). RNA from T cells was then isolated using the RNeasy plus mini kit (Qiagen) following the manufacturer protocol. The nCounter gene expression assay (NanoString) was performed with a CART characterization panel following the manufacturer's protocol. Custom probes to CAR19 and woodchuck hepatitis virus posttranscriptional regulatory element (WPRE) were added. Data were analyzed by Rosalind NanoString analysis methods (https://rosalind.onramp.bio/).

RNA-seq and Analysis

Total RNA was extracted from CART19-BCL-2(WT) compared with CART19 on day 18 after stimulation with irradiated MINO stored in PAX gene tubes according to the manufacturer's instructions (Qiagen). Integrity was checked on the Agilent TapeStation (RIN), followed by preparation for sequencing using TruSeq RNA v2 prep (Illumina). High-throughput sequencing was performed on an Illumina HiSeq 2500 platform to a target depth of 50 million paired-end reads per sample. Fastq files were processed for data quality control, read mapping, transcript assembly, and transcript abundance estimation. A number of quality control metrics were assessed, including data quality and guanine and cytosine content on per base and sequence levels, sequence length distribution and duplication levels, and insert size distribution. Finally, HTSeq was used to count the number of reads mapping to each gene. Raw read quality was evaluated using FastaQC (v0.11.2), and low-quality bases were removed using Trimmomatic (v0.36). The remaining reads were then mapped to the human genome (hg38) using STAR (v2.6.0c) with default parameters. Gene count was calculated using featureCounts (v1.6.1), and nonexpressed and lowly expressed genes with a total count of 10 across all samples were removed prior to differential expression analysis. DESeq2 was used for differential expression analysis followed by P-value correction using fdrtools (v1.2.16). DEGs were defined as genes with a log2 fold change of 1 and fdrtools adjusted P value of 0.05. The DESeq2 normalized gene matrix was used for GSEA (v4.1.0) and was conducted, and nonexpressed genes, defined as genes with zero read counts across all samples, were removed prior to analysis.

scRNA-seq and Analysis

Subcutaneous tumor xenografts (OCI-Ly18) were resected from two mice on day 7 of treatment with either CART19 (sample 1) or CART19 + venetoclax (sample 2). Resected tumors were minced and dissociated into single-cell suspensions using a 0.45-μm filter. Libraries for scRNA-seq were prepared using the Chromium Single Cell 5′ Reagent Kit with v1.1 Chemistry (10X Genomics) according to the manufacturer's instructions. After library construction, both libraries were sequenced together on the Illumina NovaSeq 6000. The raw scRNA-seq data were preprocessed using the Cell Ranger software (version 5.0.1; 10X Genomics). Feature-barcode matrices were obtained after aligning reads to the prebuilt GRCh38 human reference genome. Filtered gene expression was processed using the Seurat package (version 4.0.1; ref. 79). For additional quality control, the median absolute deviation (MAD)–based definition of outliers was used to remove putative low-quality cells from the data set. Here, any cells with fewer than 200 expressed genes, with an unusually high number of unique molecular identifier counts (above three MADs), or with high mitochondrial RNA expression (above three MADs) were discarded from downstream analysis. To compare the OCI-Ly18 cells between the two treatment conditions, the two libraries from the CART19 and CART19 + venetoclax samples were first merged and batch-corrected using the IntregrateData function in Seurat. The data were normalized and scaled using the NormalizeData and ScaleData functions. Variable features were identified using the FindVariableGenes function, and the principal components (PC) were calculated using the RunPCA function. An elbow plot, generated from the ElbowPlot function, determined the number of significant PCs required for cell clustering. The top 15 PCs were used to drive unsupervised clustering analysis via UMAP using the RunUMAP function (resolution = 0.4). To determine DEGs between the two treatment conditions for each cluster, the FindMarkers function was used with threshold values of min.pct  =  0.1 and log fold change  =  0.25. Gene ontology gene sets were downloaded from the Molecular Signatures Database (MSigDB), and pathway analysis was performed in R using the gseGO function under default parameters (80). GSEA was performed using the clusterProfiler interface (81). The CellCycleScoring function was also used to confirm a phase of the cell cycle to each cluster in the UMAP.

General Statistical Analysis

All in vitro data presented are representative of at least two independent experiments except for bulk RNA-seq and scRNA-seq (performed once with two biological replicates). All comparisons between two groups were performed using a two-tailed unpaired Student t test with Welch correction unless otherwise specified. All results are represented as mean ± SD unless otherwise noted. Survival data were analyzed using the log-rank (Mantel–Cox) test. Data analysis was performed using GraphPad Prism v9.0.

Data Availability

All requests for raw and analyzed preclinical data and materials are promptly reviewed by the University of Pennsylvania to determine if they are subject to intellectual property or confidentiality obligations. Patient-related data not included in the paper were generated as part of clinical trials and may be subject to patient confidentiality. Any data and materials that can be shared will be released via a material transfer agreement. Bulk RNA-seq and scRNA-seq are available from the Gene-Expression Omnibus using the accession number GSE195814 for all data sets. RNA-seq expression data are included in Supplementary Table S8. Other data generated from this study are available from the corresponding author upon reasonable request.

J. Svoboda reports personal fees from Incyte, AstraZeneca, Pharmacyclics, Bristol Myers Squibb, Seagen, Atara, and Adaptive outside the submitted work. E.A. Chong reports personal fees from Novartis, Bristol Myers Squibb, KITE, Beigene, and Tessa outside the submitted work. K. North reports being employed by NanoString Technologies Inc. at the time of the writing of this publication. J.A. Fraietta reports personal fees from Cartography Bio., other support from Shennon Bio, and grants from Tmunity outside the submitted work. S.F. Lacey reports a patent for Kymriah and related biomarkers licensed to Novartis. J. Gerson reports other support from LOXO, Genentech, and AbbVie outside the submitted work. S.J. Schuster reports personal fees from AstraZeneca, BeiGene, Celgene/Bristol Myers Squibb/Juno, Regeneron, and Takeda, and grants and personal fees from Genentech/Roche, Genmab, Incyte, Janssen, Legend Biotech, Fate Therapeutics, Morphosys, Mustang Biotech, Nordic Nanovector, and Novartis, as well as a patent for combination therapies of chimeric antigen receptors and PD-1 inhibitors pending. M. Ruella reports grants from the Mark Foundation, Upenn TAPITMAT, the Lymphoma Research Foundation, the NCI (1K99CA212302 and R00CA212302), the Parker Institute for Cancer Immunotherapy, and the Berman and Maguire Funds for Lymphoma Research at the University of Pennsylvania during the conduct of the study; grants from Novartis outside the submitted work; a patent for BCL-2 and CART pending; is listed as an inventor of CART technologies, University of Pennsylvania, partly licensed to Novartis, Tmunity, and viTToria Biotherapeutics; research funding from AbClon, Beckman Coulter, Lumicks, and ONI; consultancy for/honoraria from NanoString Technologies Inc. and GLG; advisory boards for AbClon, Bayer, Sana, Bristol Myers Squibb, GSK, and viTToria Biotherapeutics; and is a scientific founder of viTToria Biotherapeutics. No disclosures were reported by the other authors.

Y.G. Lee: Conceptualization, data curation, formal analysis, investigation, methodology, writing–original draft, writing–review and editing. P. Guruprasad: Formal analysis, investigation, writing–review and editing. G. Ghilardi: Formal analysis, investigation, writing–review and editing. R. Pajarillo: Investigation. C.T. Sauter: Investigation. R. Patel: Investigation, writing–review and editing. H.J. Ballard: Investigation. S.J. Hong: Investigation. I. Chun: Investigation. N. Yang: Investigation. K.V. Amelsberg: Investigation. K.D. Cummins: Investigation. J. Svoboda: Investigation. S. Gill: Investigation. E.A. Chong: Investigation, writing–review and editing. K. North: Investigation. S.E. Church: Investigation. J.A. Fraietta: Investigation. W.-J. Chang: Investigation. S.F. Lacey: Investigation. X.M. Lu: Investigation. Y. Zhang: Investigation. K. Whig: Investigation. D.C. Schultz: Investigation, writing–review and editing. S. Cherry: Investigation. J. Gerson: Investigation. S.J. Schuster: Investigation, writing–review and editing. P. Porazzi: Investigation. M. Ruella: Conceptualization, resources, formal analysis, supervision, funding acquisition, validation, investigation, methodology, writing–original draft, project administration, writing–review and editing.

The authors acknowledge the Human Immunology Core, the Stem Cell Xenograft Core, the Flow Cytometry Core, and the Next-Generation Sequencing Core at the University of Pennsylvania for their services. The authors thank Jonathan Schug (University of Pennsylvania Next-Generation Sequencing Core) for the assistance with single-cell RNA sequencing. This research was supported by the Mark Foundation (M. Ruella), a Upenn TAPITMAT grant (M. Ruella), the Lymphoma Research Foundation (Y.G. Lee), NCI 1K99CA212302 and NCI R00CA212302 (M. Ruella), the Laffey McHugh Foundation (M. Ruella), the Parker Institute for Cancer Immunotherapy (M. Ruella), and the Berman and Maguire Funds for Lymphoma Research at the University of Pennsylvania (M. Ruella).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: Supplementary data for this article are available at Cancer Discovery Online (http://cancerdiscovery.aacrjournals.org/).

1.
Spiegel
JY
,
Patel
S
,
Muffly
L
,
Hossain
NM
,
Oak
J
,
Baird
JH
, et al
.
CAR T cells with dual targeting of CD19 and CD22 in adult patients with recurrent or refractory B-cell malignancies: a phase 1 trial
.
Nat Med
2021
;
27
:
1419
31
.
2.
Melenhorst
JJ
,
Chen
GM
,
Wang
M
,
Porter
DL
,
Chen
C
,
Collins
MA
, et al
.
Decade-long leukaemia remissions with persistence of CD4+ CAR T cells
.
Nature
2022
;
602
:
503
9
.
3.
Chong
EA
,
Ruella
M
,
Schuster
SJ
,
Lymphoma Program Investigators at the University of P
.
Five-year outcomes for refractory B-cell lymphomas with CAR T-cell therapy
.
N Engl J Med
2021
;
384
:
673
4
.
4.
Ruella
M
,
Gill
S
.
How to train your T cell: genetically engineered chimeric antigen receptor T cells versus bispecific T-cell engagers to target CD19 in B acute lymphoblastic leukemia
.
Expert Opin Biol Ther
2015
;
15
:
761
6
.
5.
Turtle
CJ
,
Hanafi
LA
,
Berger
C
,
Hudecek
M
,
Pender
B
,
Robinson
E
, et al
.
Immunotherapy of non-Hodgkin's lymphoma with a defined ratio of CD8+ and CD4+ CD19-specific chimeric antigen receptor–modified T cells
.
Sci Transl Med
2016
;
8
:
355ra116
.
6.
Wang
M
,
Munoz
J
,
Goy
A
,
Locke
FL
,
Jacobson
CA
,
Hill
BT
, et al
.
KTE-X19 CAR T-cell therapy in relapsed or refractory mantle-cell lymphoma
.
N Engl J Med
2020
;
382
:
1331
42
.
7.
Braendstrup
P
,
Levine
BL
,
Ruella
M
.
The long road to the first FDA-approved gene therapy: chimeric antigen receptor T cells targeting CD19
.
Cytotherapy
2020
;
22
:
57
69
.
8.
Park
JH
,
Rivière
I
,
Gonen
M
,
Wang
X
,
Sénéchal
B
,
Curran
KJ
, et al
.
Long-term follow-up of CD19 CAR therapy in acute lymphoblastic leukemia
.
N Engl J Med
2018
;
378
:
449
59
.
9.
Lee
DW
,
Kochenderfer
JN
,
Stetler-Stevenson
M
,
Cui
YK
,
Delbrook
C
,
Feldman
SA
, et al
.
T cells expressing CD19 chimeric antigen receptors for acute lymphoblastic leukaemia in children and young adults: a phase 1 dose-escalation trial
.
Lancet North Am Ed
2015
;
385
:
517
28
.
10.
Schuster
SJ
,
Bishop
MR
,
Tam
CS
,
Waller
EK
,
Borchmann
P
,
McGuirk
JP
, et al
.
Tisagenlecleucel in adult relapsed or refractory diffuse large B-cell lymphoma
.
N Engl J Med
2019
;
380
:
45
56
.
11.
Maude
SL
,
Laetsch
TW
,
Buechner
J
,
Rives
S
,
Boyer
M
,
Bittencourt
H
, et al
.
Tisagenlecleucel in children and young adults with B-cell lymphoblastic leukemia
.
N Engl J Med
2018
;
378
:
439
48
.
12.
Cerrano
M
,
Ruella
M
,
Perales
MA
,
Vitale
C
,
Faraci
DG
,
Giaccone
L
, et al
.
The advent of CAR T-cell therapy for lymphoproliferative neoplasms: integrating research into clinical practice
.
Front Immunol
2020
;
11
:
888
.
13.
Galon
J
,
Rossi
J
,
Turcan
S
,
Danan
C
,
Locke
FL
,
Neelapu
SS
, et al
.
Characterization of anti-CD19 chimeric antigen receptor (CAR) T cell-mediated tumor microenvironment immune gene profile in a multicenter trial (ZUMA-1) with axicabtagene ciloleucel (axi-cel, KTE-C19)
.
J Clin Oncol
35:15s
,
2017
(suppl; abstr 3025)
.
14.
Fraietta
JA
,
Lacey
SF
,
Orlando
EJ
,
Pruteanu-Malinici
I
,
Gohil
M
,
Lundh
S
, et al
.
Determinants of response and resistance to CD19 chimeric antigen receptor (CAR) T cell therapy of chronic lymphocytic leukemia
.
Nat Med
2018
;
24
:
563
71
.
15.
Neelapu
SS
,
Rossi
JM
,
Jacobson
CA
,
Locke
FL
,
Miklos
DB
,
Reagan
PM
, et al
.
CD19-loss with preservation of other B cell lineage features in patients with large B cell lymphoma who relapsed post-Axi-Cel
.
Blood
2019
;
134
:
203
.
16.
Ghilardi
G
,
Braendstrup
P
,
Chong
EA
,
Schuster
SJ
,
Svoboda
J
,
Ruella
M
.
CAR-T TREK through the lymphoma universe, to boldly go where no other therapy has gone before
.
Br J Haematol
2021
;
193
:
449
65
.
17.
Singh
N
,
Lee
YG
,
Shestova
O
,
Ravikumar
P
,
Hayer
KE
,
Hong
SJ
, et al
.
Impaired death receptor signaling in leukemia causes antigen-independent resistance by inducing CAR T-cell dysfunction
.
Cancer Discov
2020
;
10
:
552
67
.
18.
Upadhyay
R
,
Boiarsky
JA
,
Pantsulaia
G
,
Svensson-Arvelund
J
,
Lin
MJ
,
Wroblewska
A
, et al
.
A critical role for fas-mediated off-target tumor killing in T-cell immunotherapy
.
Cancer Discov
2021
;
11
:
599
613
.
19.
Delbridge
AR
,
Strasser
A
.
The BCL-2 protein family, BH3-mimetics and cancer therapy
.
Cell Death Differ
2015
;
22
:
1071
80
.
20.
Maji
S
,
Panda
S
,
Samal
SK
,
Shriwas
O
,
Rath
R
,
Pellecchia
M
, et al
.
Bcl-2 antiapoptotic family proteins and chemoresistance in cancer
.
Adv Cancer Res
2018
;
137
:
37
75
.
21.
Tsujimoto
Y
,
Finger
LR
,
Yunis
J
,
Nowell
PC
,
Croce
CM
.
Cloning of the chromosome breakpoint of neoplastic B cells with the t(14;18) chromosome translocation
.
Science
1984
;
226
:
1097
9
.
22.
Yunis
JJ
,
Frizzera
G
,
Oken
MM
,
McKenna
J
,
Theologides
A
,
Arnesen
M
.
Multiple recurrent genomic defects in follicular lymphoma: a possible model for cancer
.
N Engl J Med
1987
;
316
:
79
84
.
23.
Weiss
LM
,
Warnke
RA
,
Sklar
J
,
Cleary
ML
.
Molecular analysis of the t(14;18) chromosomal translocation in malignant lymphomas
.
N Engl J Med
1987
;
317
:
1185
9
.
24.
Gaidano
G
,
Dalla-Favera
R
.
Biologic and molecular characterization of non-Hodgkin's lymphoma
.
Curr Opin Oncol
1993
;
5
:
776
84
.
25.
Roberts
AW
.
Therapeutic development and current uses of BCL-2 inhibition
.
Hematology Am Soc Hematol Educ Program
2020
;
2020
:
1
9
.
26.
Souers
AJ
,
Leverson
JD
,
Boghaert
ER
,
Ackler
SL
,
Catron
ND
,
Chen
J
, et al
.
ABT-199, a potent and selective BCL-2 inhibitor, achieves antitumor activity while sparing platelets
.
Nat Med
2013
;
19
:
202
8
.
27.
Jones
JA
,
Mato
AR
,
Wierda
WG
,
Davids
MS
,
Choi
M
,
Cheson
BD
, et al
.
Venetoclax for chronic lymphocytic leukaemia progressing after ibrutinib: an interim analysis of a multicentre, open-label, phase 2 trial
.
Lancet Oncol
2018
;
19
:
65
75
.
28.
Stilgenbauer
S
,
Eichhorst
B
,
Schetelig
J
,
Coutre
S
,
Seymour
JF
,
Munir
T
, et al
.
Venetoclax in relapsed or refractory chronic lymphocytic leukaemia with 17p deletion: a multicentre, open-label, phase 2 study
.
Lancet Oncol
2016
;
17
:
768
78
.
29.
DiNardo
CD
,
Pratz
K
,
Pullarkat
V
,
Jonas
BA
,
Arellano
M
,
Becker
PS
, et al
.
Venetoclax combined with decitabine or azacitidine in treatment-naive, elderly patients with acute myeloid leukemia
.
Blood
2019
;
133
:
7
17
.
30.
Pollyea
DA
,
Stevens
BM
,
Jones
CL
,
Winters
A
,
Pei
S
,
Minhajuddin
M
, et al
.
Venetoclax with azacitidine disrupts energy metabolism and targets leukemia stem cells in patients with acute myeloid leukemia
.
Nat Med
2018
;
24
:
1859
66
.
31.
Tam
CS
,
Anderson
MA
,
Pott
C
,
Agarwal
R
,
Handunnetti
S
,
Hicks
RJ
, et al
.
Ibrutinib plus venetoclax for the treatment of mantle-cell lymphoma
.
N Engl J Med
2018
;
378
:
1211
23
.
32.
Eyre
TA
,
Walter
HS
,
Iyengar
S
,
Follows
G
,
Cross
M
,
Fox
CP
, et al
.
Efficacy of venetoclax monotherapy in patients with relapsed, refractory mantle cell lymphoma after Bruton tyrosine kinase inhibitor therapy
.
Haematologica
2019
;
104
:
e68
e71
.
33.
Davids
MS
,
Roberts
AW
,
Seymour
JF
,
Pagel
JM
,
Kahl
BS
,
Wierda
WG
, et al
.
Phase I first-in-human study of venetoclax in patients with relapsed or refractory non-Hodgkin lymphoma
.
J Clin Oncol
2017
;
35
:
826
33
.
34.
Nakayama
K
,
Nakayama
K
,
Negishi
I
,
Kuida
K
,
Sawa
H
,
Loh
DY
.
Targeted disruption of Bcl-2 alpha beta in mice: occurrence of gray hair, polycystic kidney disease, and lymphocytopenia
.
Proc Natl Acad Sci U S A
1994
;
91
:
3700
4
.
35.
Veis
DJ
,
Sorenson
CM
,
Shutter
JR
,
Korsmeyer
SJ
.
Bcl-2-deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair
.
Cell
1993
;
75
:
229
40
.
36.
Dufva
O
,
Koski
J
,
Maliniemi
P
,
Ianevski
A
,
Klievink
J
,
Leitner
J
, et al
.
Integrated drug profiling and CRISPR screening identify essential pathways for CAR T-cell cytotoxicity
.
Blood
2020
;
135
:
597
609
.
37.
Michie
J
,
Beavis
PA
,
Freeman
AJ
,
Vervoort
SJ
,
Ramsbottom
KM
,
Narasimhan
V
, et al
.
Antagonism of IAPs enhances CAR T-cell efficacy
.
Cancer Immunol Res
2019
;
7
:
183
92
.
38.
Rowe
JM
.
Progress and predictions: AML in 2018
.
Best Pract Res Clin Haematol
2018
;
31
:
337
40
.
39.
Short
NJ
,
Konopleva
M
,
Kadia
TM
,
Borthakur
G
,
Ravandi
F
,
DiNardo
CD
, et al
.
Advances in the treatment of acute myeloid leukemia: new drugs and new challenges
.
Cancer Discov
2020
;
10
:
506
25
.
40.
Mardiana
S
,
Gill
S
.
CAR T cells for acute myeloid leukemia: state of the art and future directions
.
Front Oncol
2020
;
10
:
697
.
41.
Kenderian
SS
,
Ruella
M
,
Shestova
O
,
Klichinsky
M
,
Scholler
J
,
Song
D
, et al
.
CD33 directed chimeric antigen receptor T cell therapy as a novel preparative regimen prior to allogeneic stem cell transplantation in acute myeloid leukemia
.
Biol Blood Marrow Transplant
2015
;
21
:
S25
S6
.
42.
Gill
S
,
Tasian
SK
,
Ruella
M
,
Shestova
O
,
Li
Y
,
Porter
DL
, et al
.
Preclinical targeting of human acute myeloid leukemia and myeloablation using chimeric antigen receptor-modified T cells
.
Blood
2014
;
123
:
2343
54
.
43.
Korashy
HM
,
Alatwi
FO
,
Alhoshani
A
,
El Gamal
HM
,
Shamoon
LS
,
Ghanim
SS
, et al
.
Venetoclax, a novel BCL-2 inhibitor, induces cell growth suppression, apoptosis, cell cycle arrest, and autophagy in triple negative breast cancer MDA-MB-231 cells
.
FASEB J
2019
;
33
:
674.16
.16
.
44.
Juárez-Salcedo
LM
,
Desai
V
,
Dalia
S
.
Venetoclax: evidence to date and clinical potential
.
Drugs Context
2019
;
8
:
212574
.
45.
Klanova
M
,
Klener
P
.
BCL-2 proteins in pathogenesis and therapy of B-cell non-Hodgkin lymphomas
.
Cancers
2020
;
12
:
938
.
46.
Bose
P
,
Gandhi
V
,
Konopleva
M
.
Pathways and mechanisms of venetoclax resistance
.
Leuk Lymphoma
2017
;
58
:
2026
39
.
47.
Birkinshaw
RW
,
Gong
JN
,
Luo
CS
,
Lio
D
,
White
CA
,
Anderson
MA
, et al
.
Structures of BCL-2 in complex with venetoclax reveal the molecular basis of resistance mutations
.
Nat Commun
2019
;
10
:
1
10
.
48.
Tausch
E
,
Close
W
,
Dolnik
A
,
Bloehdorn
J
,
Chyla
B
,
Bullinger
L
, et al
.
Venetoclax resistance and acquired BCL2 mutations in chronic lymphocytic leukemia
.
Haematologica
2019
;
104
:
e434
.
49.
Tahir
SK
,
Smith
ML
,
Hessler
P
,
Rapp
LR
,
Idler
KB
,
Park
CH
, et al
.
Potential mechanisms of resistance to venetoclax and strategies to circumvent it
.
BMC Cancer
2017
;
17
:
1
10
.
50.
Eyre
TA
,
Walter
HS
,
Iyengar
S
,
Follows
G
,
Cross
M
,
Fox
CP
, et al
.
Efficacy of venetoclax monotherapy in patients with relapsed, refractory mantle cell lymphoma after Bruton tyrosine kinase inhibitor therapy
.
Haematologica
2019
;
104
:
e68
71
.
51.
Grayson
JM
,
Zajac
AJ
,
Altman
JD
,
Ahmed
R
.
Cutting edge: increased expression of Bcl-2 in antigen-specific memory CD8+ T cells
.
J Immunol
2000
;
164
:
3950
4
.
52.
Charo
J
,
Finkelstein
SE
,
Grewal
N
,
Restifo
NP
,
Robbins
PF
,
Rosenberg
SA
.
Bcl-2 overexpression enhances tumor-specific T-cell survival
.
Cancer Res
2005
;
65
:
2001
8
.
53.
Maraskovsky
E
,
O'Reilly
LA
,
Teepe
M
,
Corcoran
LM
,
Peschon
JJ
,
Strasser
A
.
Bcl-2 can rescue T lymphocyte development in interleukin-7 receptor–deficient mice but not in mutant rag-1−/− mice
.
Cell
1997
;
89
:
1011
9
.
54.
Schuster
SJ
,
Svoboda
J
,
Chong
EA
,
Nasta
SD
,
Mato
AR
,
Anak
Ö
, et al
.
Chimeric antigen receptor T cells in refractory B-cell lymphomas
.
N Engl J Med
2017
;
377
:
2545
54
.
55.
Linette
GP
,
Hess
JL
,
Sentman
CL
,
Korsmeyer
SJ
.
Peripheral T-cell lymphoma in lckpr-bcl-2 transgenic mice
.
Blood
1995
;
86
:
1255
60
.
56.
Paszkiewicz
PJ
,
Fräßle
SP
,
Srivastava
S
,
Sommermeyer
D
,
Hudecek
M
,
Drexler
I
, et al
.
Targeted antibody-mediated depletion of murine CD19 CAR T cells permanently reverses B cell aplasia
.
J Clin Invest
2016
;
126
:
4262
72
.
57.
Fulda
S
.
Tumor resistance to apoptosis
.
Int J Cancer
2009
;
124
:
511
5
.
58.
Fulda
S
,
Meyer
E
,
Debatin
K-M
.
Inhibition of TRAIL-induced apoptosis by Bcl-2 overexpression
.
Oncogene
2002
;
21
:
2283
94
.
59.
Jiang
Z
,
Liu
Z
,
Li
M
,
Chen
C
,
Wang
X
.
Immunogenomics analysis reveals that TP53 mutations inhibit tumor immunity in gastric cancer
.
Translational oncology
2018
;
11
:
1171
87
.
60.
Maruyama
R
,
Yamana
K
,
Itoi
T
,
Hara
N
,
Bilim
V
,
Nishiyama
T
, et al
.
Absence of Bcl-2 and Fas/CD95/APO-1 predicts the response to immunotherapy in metastatic renal cell carcinoma
.
Br J Cancer
2006
;
95
:
1244
9
.
61.
Karlsson
S
,
Lindqvist
A
,
Fransson
M
,
Paul-Wetterberg
G
,
Nilsson
B
,
Essand
M
, et al
.
Combining CAR T cells and the Bcl-2 family apoptosis inhibitor ABT-737 for treating B-cell malignancy
.
Cancer Gene Ther
2013
;
20
:
386
93
.
62.
Yang
M
,
Wang
L
,
Ni
M
,
Neuber
B
,
Wang
S
,
Gong
W
, et al
.
Pre-sensitization of malignant B cells through venetoclax significantly improves the cytotoxic efficacy of CD19.CAR-T cells
.
Front Immunol
2020
;
11
:
608167
.
63.
Kohlhapp
FJ
,
Haribhai
D
,
Mathew
R
,
Duggan
R
,
Ellis
PA
,
Wang
R
, et al
.
Venetoclax increases intratumoral effector T cells and antitumor efficacy in combination with immune checkpoint blockade
.
Cancer Discov
2021
;
11
:
68
79
.
64.
Broome
HE
,
Dargan
CM
,
Krajewski
S
,
Reed
JC
.
Expression of Bcl-2, Bcl-x, and Bax after T cell activation and IL-2 withdrawal
.
J Immunol
1995
;
155
:
2311
7
.
65.
Jones
AK
,
Freise
KJ
,
Agarwal
SK
,
Humerickhouse
RA
,
Wong
SL
,
Salem
AH
.
Clinical predictors of venetoclax pharmacokinetics in chronic lymphocytic leukemia and non-Hodgkin's lymphoma patients: a pooled population pharmacokinetic analysis
.
AAPS J
2016
;
18
:
1192
202
.
66.
Bollag
G
,
Tsai
J
,
Zhang
J
,
Zhang
C
,
Ibrahim
P
,
Nolop
K
, et al
.
Vemurafenib: the first drug approved for BRAF-mutant cancer
.
Nat Rev Drug Discovery
2012
;
11
:
873
86
.
67.
Stone
RM
,
Mandrekar
SJ
,
Sanford
BL
,
Laumann
K
,
Geyer
S
,
Bloomfield
CD
, et al
.
Midostaurin plus chemotherapy for acute myeloid leukemia with a FLT3 mutation
.
N Engl J Med
2017
;
377
:
454
64
.
68.
Yang
J
,
Nie
J
,
Ma
X
,
Wei
Y
,
Peng
Y
,
Wei
X
.
Targeting PI3K in cancer: mechanisms and advances in clinical trials
.
Mol Cancer
2019
;
18
:
1
28
.
69.
Wang
H
,
Han
P
,
Qi
X
,
Li
F
,
Li
M
,
Fan
L
, et al
.
Bcl-2 enhances chimeric antigen receptor T cell persistence by reducing activation-induced apoptosis
.
Cancers
2021
;
13
:
197
.
70.
Hildeman
DA
,
Zhu
Y
,
Mitchell
TC
,
Bouillet
P
,
Strasser
A
,
Kappler
J
, et al
.
Activated T cell death in vivo mediated by proapoptotic Bcl-2 family member bim
.
Immunity
2002
;
16
:
759
67
.
71.
Zhang
Y
,
Guan
XY
,
Jiang
P
.
Cytokine and chemokine signals of T-cell exclusion in tumors
.
Front Immunol
2020
;
11
:
594609
.
72.
Muranski
P
,
Boni
A
,
Wrzesinski
C
,
Citrin
DE
,
Rosenberg
SA
,
Childs
R
, et al
.
Increased intensity lymphodepletion and adoptive immunotherapy—how far can we go?
Nat Clin Pract Oncol
2006
;
3
:
668
81
.
73.
Gattinoni
L
,
Finkelstein
SE
,
Klebanoff
CA
,
Antony
PA
,
Palmer
DC
,
Spiess
PJ
, et al
.
Removal of homeostatic cytokine sinks by lymphodepletion enhances the efficacy of adoptively transferred tumor-specific CD8+ T cells
.
J Exp Med
2005
;
202
:
907
12
.
74.
Milone
MC
,
Fish
JD
,
Carpenito
C
,
Carroll
RG
,
Binder
GK
,
Teachey
D
, et al
.
Chimeric receptors containing CD137 signal transduction domains mediate enhanced survival of T cells and increased antileukemic efficacy in vivo
.
Mol Ther
2009
;
17
:
1453
64
.
75.
Horn
H
,
Ziepert
M
,
Becher
C
,
Barth
TF
,
Bernd
HW
,
Feller
AC
, et al
.
MYC status in concert with BCL-2 and BCL6 expression predicts outcome in diffuse large B-cell lymphoma
.
Blood
2013
;
121
:
2253
63
.
76.
Kramer
M
,
Hermans
J
,
Wijburg
E
,
Philippo
K
,
Geelen
E
,
Van Krieken
J
, et al
.
Clinical relevance of BCL-2, BCL6, and MYC rearrangements in diffuse large B-cell lymphoma
.
Blood
1998
;
92
:
3152
62
.
77.
Cheson
BD
,
Fisher
RI
,
Barrington
SF
,
Cavalli
F
,
Schwartz
LH
,
Zucca
E
, et al
.
Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: the Lugano classification
.
J Clin Oncol
2014
;
32
:
3059–68
.
78.
Lee
DW
,
Santomasso
BD
,
Locke
FL
,
Ghobadi
A
,
Turtle
CJ
,
Brudno
JN
, et al
.
ASTCT consensus grading for cytokine release syndrome and neurologic toxicity associated with immune effector cells
.
Biol Blood Marrow Transplant
2019
;
25
:
625
38
.
79.
Stuart
T
,
Butler
A
,
Hoffman
P
,
Hafemeister
C
,
Papalexi
E
,
Mauck
WM
3rd
, et al
.
Comprehensive integration of single-cell data
.
Cell
2019
;
177
:
1888
902
.
80.
Yu
G
,
Wang
LG
,
Yan
GR
,
He
QY
.
DOSE: an R/Bioconductor package for disease ontology semantic and enrichment analysis
.
Bioinformatics
2015
;
31
:
608
9
.
81.
Yu
G
,
Wang
LG
,
Han
Y
,
He
QY
.
clusterProfiler: an R package for comparing biological themes among gene clusters
.
OMICS
2012
;
16
:
284
7
.

Supplementary data