Purpose: Responses to therapy with chimeric antigen receptor T cells recognizing CD19 (CART19, CTL019) may vary by histology. Mantle cell lymphoma (MCL) represents a B-cell malignancy that remains incurable despite novel therapies such as the BTK inhibitor ibrutinib, and where data from CTL019 therapy are scant. Using MCL as a model, we sought to build upon the outcomes from CTL019 and from ibrutinib therapy by combining these in a rational manner.
Experimental Design: MCL cell lines and primary MCL samples were combined with autologous or normal donor-derived anti-CD19 CAR T cells along with ibrutinib. The effect of the combination was studied in vitro and in mouse xenograft models.
Results: MCL cells strongly activated multiple CTL019 effector functions, and MCL killing by CTL019 was further enhanced in the presence of ibrutinib. In a xenograft MCL model, we showed superior disease control in the CTL019- as compared with ibrutinib-treated mice (median survival not reached vs. 95 days, P < 0.005) but most mice receiving CTL019 monotherapy eventually relapsed. Therefore, we added ibrutinib to CTL019 and showed that 80% to 100% of mice in the CTL019 + ibrutinib arm and 0% to 20% of mice in the CTL019 arm, respectively, remained in long-term remission (P < 0.05).
Conclusions: Combining CTL019 with ibrutinib represents a rational way to incorporate two of the most recent therapies in MCL. Our findings pave the way to a two-pronged therapeutic strategy in patients with MCL and other types of B-cell lymphoma. Clin Cancer Res; 22(11); 2684–96. ©2016 AACR.
Most patients with relapsed mantle cell lymphoma can now be treated with the BTK inhibitor ibrutinib. However, up to 30% of these patients do not respond to ibrutinib and the majority of responders eventually relapse. Recent reports highlight potent activity of anti-CD19 chimeric antigen receptor T cells (CART19, CTL019) in B-cell malignancies. In this study, we illustrate for the first time that ibrutinib can be added to CTL019 and that only the combined approach leads to profound, durable responses in xenograft models of MCL. These findings set the stage for future clinical trials evaluating this combination in B-cell neoplasms.
Mantle cell lymphoma (MCL) accounts for up to 10% of all lymphomas (1) and typically presents in advanced stage (2). For most patients with MCL, the prognosis is poor with a median survival of 4 years (3). Currently, there is no curative treatment for MCL and, therefore, novel therapies for this type of lymphoma are urgently needed.
The B-cell receptor (BCR) complex is critical for antigen-induced activation of normal B lymphocytes and plays a key role in the pathogenesis of certain types of B-cell lymphoma. BCR engagement activates several kinases including LYN, SYK, and BTK (4,5). BTK recently gained particular attention, as the potent BTK inhibitor ibrutinib demonstrated therapeutic efficacy in several types of B-cell lymphoma including MCL (6–8). However, up to one third of MCL patients do not respond to ibrutinib and among the responders only a third achieve complete remission (CR). Furthermore, the therapy usually leads to drug resistance as the median duration of response is only 17.5 months with a 24-month PFS of 31% (8,9). The mechanisms of resistance are currently poorly understood but are thought to involve mutations in BTK that impair ibrutinib binding, or activating mutations of the enzyme PLCγ2 resulting in constitutive BTK-independent cell signaling (10,11). Furthermore, because blockade of BTK function is not directly cytotoxic, at least in some types of lymphoma (11), it may predispose to clonal evolution by conferring a selection pressure. Rationally designed combinations of ibrutinib with other antilymphoma modalities could potentially overcome this shortcoming and thereby improve patient outcomes.
Infusion of autologous T cells transduced with chimeric antigen receptors (CAR) against the B-cell–specific CD19 antigen (CART19, CTL019) leads to dramatic clinical responses in many patients with various types of B-cell neoplasms but CTL019 efficacy against MCL specifically has not yet been established (12–17). The presence of bulky masses may hinder T-cell infiltration with consequent impairment of antitumor activity (18). Conversely, bulky lymphadenopathy does not appear to impair the response to ibrutinib and the drug actually triggers mobilization of the malignant cells to peripheral blood, potentially making them more accessible to CTL019 cells (8).
In addition to BTK, ibrutinib irreversibly inhibits the TEC family kinase ITK (IL2-inducible T-cell kinase). ITK activates PLCγ upon T-cell receptor (TCR) ligation and leads to a signaling cascade that culminates in the activation of T lymphocytes (19). Recent preclinical data suggest that ibrutinib preferentially inhibits Th2-polarized CD4 T cells thus skewing T cells towards Th1 antitumor immune response (20). However, another recent study shows that ibrutinib can antagonize rituximab-dependent NK cell–mediated cytotoxicity and reduce cytokine production, indicating that ITK inhibition may also lead to reduced tumor killing (21). In this context, it is important to discover whether stimulation of the chimeric antigen receptor in CTL019 cells would lead to activation of ITK and, if so, whether inhibition of ITK by ibrutinib would have an advantageous or deleterious effect on CTL019 function.
In principle, the combination of the BTK inhibitor ibrutinib with CTL019 brings together two leading novel approaches to the treatment of B-cell lymphoma and taking advantage of their vastly different mechanisms of action may prove particularly effective. Using in vitro and in vivo models of MCL, including a novel cell line highly sensitive to ibrutinib, we demonstrate here that CTL019 is more effective than ibrutinib as monotherapy, and that the addition of ibrutinib to CTL019 further augments the antitumor effect and leads to prolonged remissions.
Materials and Methods
Cell lines and primary samples
Cell lines were originally obtained from ATCC (K-562, Mino and JEKO-1) or DSMZ (MOLM-14 and NALM-6); cell lines were obtained more than 6 months prior experiments and authentication was performed by cell banks utilizing short tandem repeat profiling. MCL-RL was generated in our laboratory from a pleural effusion of a MCL patient (the presence of the t(11, 14) characteristic of MCL was tested by FISH). All cell lines were tested for the presence of mycoplasma contamination (MycoAlert Mycoplasma Detection Kit, LT07-318, Lonza). For some experiments, MCL-RL and JEKO-1 cells were transduced with firefly luciferase/eGFP and then sorted to obtain >99% positive population. Cell lines MOLM-14, K562, and NALM-6 were used as controls as indicated in the relevant figures. The cell lines were maintained in culture with RPMI1640 (Gibco, 11875–085, LifeTechnologies) supplemented with 10% FBS (Gemini, 100–106) and 50 U/mL penicillin/streptomycin (Gibco, Life Technologies, 15070–063). Deidentified primary human MCL bone marrow and peripheral blood specimens were obtained from the clinical practices of University of Pennsylvania under an Institutional Review Board–approved protocol (UPCC #03409). For all functional studies, primary cells were thawed at least 12 hours before experiment and rested at 37°C.
FISH and IHC
The FISH analysis and IHC were performed according to the standard method and as described previously (22). Specifics of the experiment of this article are detailed in the Supplementary Methods section.
Thin-layer cell preparation was obtained by Cytospin (Thermo Scientific) and stained with Giemsa. For formalin-fixed paraffin-embedded tissues, immunohistochemical staining was performed on a Leica Bond-III instrument (Leica Biosystems) using the Bond Polymer Refine Detection System. Antibodies against CD2, SOX-11, Pax5, and CyclinD1 were used undiluted. Heat-induced epitope retrieval was done for 20 minutes with ER2 solution (Leica Microsystems, AR9640). Images were digitally acquired using the Aperio ScanScope (Leica Biosystems).
Generation of CAR constructs and CAR T cells
Ibrutinib (PCI-32765) was purchased from MedKoo (#202171) or Selleck Biochemicals (#S2680) as a powder or DMSO solution. The products obtained from the two companies were compared and proven to have equivalent activity (data not shown). For in vitro experiments, ibrutinib was dissolved in DMSO and diluted to 2, 10, 100, or 1,000 nmol/L in culture media. For in vivo experiment, ibrutinib powder was dissolved in a 10% HP-β-cyclodextrin solution (1.6 mg/mL) and administered to mice in the drinking water.
Multiparametric flow cytometry
Flow cytometry was performed as described previously (24,25) and detailed characteristics of the experiments are provided in Supplementary Methods.
MTT enzymatic conversion assay
The assay was performed as described previously (26). Specifications of this experiment are detailed in the Supplementary Methods section.
DNA fragmentation (TUNEL) assay
ApoAlert DNA fragmentation assay kit (Clontech, 630108) was used according to the manufacturer's protocol. In brief, cells were cultured at 0.5 × 106 cells/mL for 72 hours with DMSO (control) or ibrutinib at the listed doses. The cells were then washed, fixed, permeabilized, and incubated for 1 hour at 37°C with or without terminal deoxynucleotidyl transferase (TdT). After exposure to the stopping buffer and washing, the cells were analyzed by flow cytometry using the CellQuest PRO software v. 5 (BD Biosciences).
Western blot analysis
The assay was performed as described previously (27). Specifications of this experiment are detailed in the Supplementary Methods section.
CTL019 cells were screened by RT-PCR analysis for Fas ligand (Applied Biosystems, Life Technologies, Hs00181225_m1), granzyme B (Applied Biosystems, Hs01554355), perforin (Applied Biosystems, Hs00169473_m1), and TRAIL [Applied Biosystems, Hs00921974 mRNA expression at the end of expansion (day 10)]. RNA was extracted with RNAqueos-4PCR Kit (Ambion, Life Technologies, AM-1914) and cDNA was synthesized with iScript Reverse Transcription Supermix for RT-qPCR (Bio-Rad, 170-8841). The relative target cDNA copies were quantified by relative qPCR (qPCR) with ABI TaqMan-specific primers and probe set; TaqMan GUSB primers (AB, Hs00939627) and probe set were used for normalization.
In vitro T-cell effector function assays
CD107a degranulation, carboxyfluorescein diacetate succinimidyl ester proliferation (CFSE), cytotoxicity assays, and cytokine measurements were performed as described previously (24,28). Specifications of this experiment are detailed in the Supplementary Methods section.
In vivo experiments were performed as described previously (24,25,29). Schemas of the utilized xenograft models are discussed in detailed in the relevant figure legends, Results, and the Supplementary Material section.
All statistical analyses were performed as indicated using GraphPad Prism 6 for Windows, version 6.04. Student t test was used to compare two groups; in analysis where multiple groups were compared, one-way ANOVA was performed with Holm–Sidak correction for multiple comparisons. When multiple groups at multiple time points/ratios were compared, the Student t test or ANOVA for each time points/ratios was used. Survival curves were compared using the log-rank test. In the figures, asterisks represent P values (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001) and “ns” means “not significant” (P > 0.05). Further details of the statistics for each experiment are listed in figure legends.
Sensitivity of MCL cell lines to ibrutinib
Most MCL cell lines in existence have been immortalized and propagated for many generations in vitro and are poorly sensitive to ibrutinib (30,31). We harvested MCL cells from the pleural effusion of a patient with an advanced MCL and established a cell line (MCL-RL) that retained the primary cell polymorphic morphology, the characteristic MCL immunophenotype including CD19 and CD5 coexpression, and the classical t(11;14) translocation with 6 to 7 copies per cell of the IgH-Cyclin D1 fusion gene cell (Fig. 1A). Exposure of the MCL-RL cell line to increasing concentrations of ibrutinib led to a dose-dependent inhibition of cell growth with an IC50 of 10 nmol/L (Fig. 1B), including their apoptotic cell death (Supplementary Fig. S1A, top row). In contrast, the commonly used MCL cell lines Mino and JEKO-1 were relatively resistant to ibrutinib, with IC50 of 1 and 10 μmol/L, respectively (Fig. 1B) and showed no evidence of cell death (Supplementary Fig. S1A, bottom row). Of note, ibrutinib inhibited phosphorylation of BTK to a similar degree in both sensitive (MCL-RL) and resistant (JEKO-1) cell lines, indicating that the resistance in JEKO-1 cells is BTK-independent (Supplementary Fig. S1B). Non-MCL cell lines NALM-6 (B-cell acute lymphoid leukemia) and K562 (acute myeloid leukemia) were also tested for ibrutinib sensitivity showing IC50 of >1 and >10 μmol/L, respectively (Supplementary Fig. S1C) and hence served as additional negative controls throughout the study. To determine the suitability of MCL-RL cells for in vivo experiments, we injected immunodeficient NSG mice intravenously with 1 × 106 MCL-RL cells expressing firefly luciferase and monitored the mice for tumor burden by bioluminescence imaging and for survival. The MCL-RL cells engrafted in all mice and localized predominantly to the spleen and liver, followed by dissemination to bone marrow, blood, and other organs (Supplementary Fig. S2A). Histology and IHC of the tumors recapitulated the morphology and immunophenotype of the original MCL-RL cells (Supplementary Fig. S2B). Importantly, MCL-RL also demonstrated a response to ibrutinib treatment in this in vivo setting with a dose-dependent reduction in tumor growth (Fig. 1C, top) and improvement in overall survival (Fig. 1C, bottom).
Mantle cell lymphoma cells are sensitive to CTL019 effector functions
To examine sensitivity of the MCL cells to killing by CTL019 cells, we transduced healthy donor T cells with the same anti-CD19 CAR construct that has been used in the clinical trials of our group (16) and used for the following experiments. The design of this CAR and the T-cell production schema are shown in Supplementary Fig. S3A and S3B. To test whether CTL019 cells could be manufactured also from the blood of patients with leukemic MCL, we expanded and transduced patient-derived T cells (Fig. 2A, top) and then performed a CD107a degranulation and cytokine production assay to demonstrate reactivity against that patient's own MCL (Fig. 2A, bottom). Given the recent interest (32) in tumor-infiltrating and marrow-infiltrating lymphocytes, we also performed a similar study using marrow-derived T cells from a patient with stage IV MCL (Fig. 2A and Supplementary Fig. S4A and S4B). A series of in vitro experiments showed that both the ibrutinib-sensitive MCL-RL and the ibrutinib-resistant JEKO-1 cell line induced comparably strong activation of CTL019 cells as determined by their degranulation, cytokine production, cytotoxic activity, and proliferation (Fig. 2B and C and Supplementary Fig. S4C). As shown in Fig. 2C, the MCL-RL cell line was less sensitive to CTL019 cytotoxicity as compared with JEKO-1. This was likely due to the increased activation-induced apoptosis of CTL in the presence of MCL-RL (Supplementary Fig. S4D). The CTL019 activation was strictly CAR-dependent, as the untransduced cells (UTD) from the same donors tested in parallel showed no, or very limited, activity in these assays. We next evaluated in vivo different doses of CTL019 cells and demonstrated a dose-dependent antitumor efficacy, with 2 × 106 CTL019 cells per mouse proving to be the most effective. Higher doses of T cells were associated with nonspecific alloreactivity (data not shown). Notably, the antilymphoma activity of CTL019 was observed in NSG mice engrafted with both ibrutinib-resistant (JEKO-1; Fig. 2D, top) and ibrutinib-sensitive (MCL-RL) MCL cell lines (Fig. 2D, bottom). These results indicate that MCL is sensitive to the effector functions of CTL019 cells.
Impact of ibrutinib on CTL019 function in vitro
Ibrutinib was originally thought not to impact T cells based on short-term activity assays (33). However, a comprehensive analysis of the impact of ibrutinib on the T-cell kinase ITK subsequently supported an overall immunomodulatory role of ibrutinib in CD4 T cells as a suppressor of the Th2-type polarization (20). Cytokine expression pattern analysis of patients treated with anti-CD19 CAR T cells performed by several groups indicates that this therapy is associated with both Th1-type (IL2, IFNγ, TNF) and Th2-type (IL4, IL5, IL10), as well as other cytokine-secretion patterns. (13,28) Therefore, we evaluated the effect of ibrutinib on CTL019 function at, above, and below the concentrations that would be expected in patients (mean peak concentration in patient serum is 100–150 ng/mL; ref. 6). We found that CTL019 cells express ITK and that stimulation of CTL019 cells, whether through the TCR complex or through the CAR, led to the phosphorylation of ITK. The presence of ibrutinib resulted in a modest reduction in ITK phosphorylation that was only evident at the highest concentration of ibrutinib (Fig. 3A).
We next probed the short- and long-term in vitro function of CTL019 cells in the presence of ibrutinib. Following 4 to 6 hours of incubation with MCL cell lines, clinically relevant concentrations of ibrutinib did not influence CTL019 degranulation and cytokine production (Fig. 3B). In a 5-day proliferation assay, we observed a dose-dependent reduction in T-cell proliferation and total T-cell numbers, but this reduction occurred predominantly at supraphysiologic concentrations of ibrutinib (1 μmol/L and above) and, more frequently, upon the CTL019 cell exposure to JEKO-1 as compared with MCL-RL cells. (Fig. 3C and Supplementary Fig. S5A). Similarly, the cell-culture supernatant analysis for 30 different cytokines demonstrated that ibrutinib did not impact cytokine production except in the presence of supraphysiologic drug concentrations (Fig. 3D). We did not find differences in Th1/Th2 polarization between ibrutinib exposed and nonexposed CTL019 using two different techniques (Fig. 3D and Supplementary Fig. S5B). The intrinsic cytotoxic machinery of CTL019 was not significantly impacted in the presence of ibrutinib (Fig. 3E and Supplementary Fig. S5C) and there was no apparent difference in the expression of CD19 or of inhibitory ligands on MCL exposed to ibrutinib (data not shown). Notably, killing of MCL cells by CTL019 cells was significantly augmented in the presence of ibrutinib, suggesting an additive cytotoxic effect of the combination in both ibrutinib-sensitive (MCL-RL) and -resistant (JEKO-1) MCL cells (Fig. 3F). Collectively, these results indicate that ibrutinib has no adverse effect on CART cell function at physiologically relevant concentrations, and that the combination of two agents active against MCL is additive in vitro.
Impact of ibrutinib on circulating CTL019 cells
In our in vitro models, combination with ibrutinib clearly enhanced the already potent antitumor effect of CTL019 and hence it was important to evaluate the nature of the interaction of CTL019 with ibrutinib also in vivo.
Inhibition of ITK has been reported to antagonize Th2 polarization and promote a Th1 phenotype (20). However, in mice treated with CTL019 and ibrutinib we did not find an increase in Th1 cells when compared with CTL019 monotherapy (Supplementary Fig. S6A). Of note, exposure of tumor-bearing mice to ibrutinib led to an increase in peripheral blood T cells, regardless of antigen specificity, as ibrutinib augmented circulating T-cell numbers of both CTL019 and control untransduced cells (Fig. 4A and data not shown). This increase was not due to increased proliferation, as there was no difference in the proliferation marker Ki67 between the treatment groups (Fig. 4B, left). Similarly, we did not find any difference in the antiapoptotic marker Bcl2, suggesting that the difference in the number of circulating CTL019 cells was not related to an impairment of apoptosis (Fig. 4B, right). To differentiate whether the increased number of circulating T cells in ibrutinib-treated mice were due to accumulation in, or mobilization into, the peripheral blood compartment, we engrafted NSG mice with unlabeled MCL-RL cells followed by injection with luciferase-expressing T cells, wherein the bioluminescent signal (BLI) from the whole animal would correlate with total T-cell load. Ibrutinib treatment did not enhance BLI in either CTL019 or control T-cell–treated animals, suggesting that ibrutinib did not increase the total T-cell number but rather triggered T-cell mobilization to the blood (Fig. 4C). We then investigated the frequency of different T-cell subsets among the circulating T cells and could not detect any difference in the T-cell subset distribution between the CTL019- and CTL019/ibrutinib- engrafted mice (Fig Supplementary Fig. S6B and S6C). Because CXCR4 is involved in ibrutinib-driven B-cell mobilization in humans, we measured the expression of CXCR4 in vivo in the circulating T cells of mice treated with CTL019 or CTL019 and ibrutinib and found similar CXCR4 levels in the two groups indicating that the increased mobilization was not due to decreased CXCR4 expression (Supplementary Fig. S7A). Finally, we analyzed the expression of inhibitory/costimulatory receptors in the peripheral blood T cells of mice treated with CTL019 and CTL019 plus ibrutinib. There was a trend to reduced PD-1 expression when ibrutinib was added to CTL019 or untransduced T cell controls, but no differences in expression of TIM3, LAG3, CD137, or CTLA4 were found. (Supplementary Fig. S7B and S7C).
In vivo antitumor activity of ibrutinib, CTL019, and their combination
Our in vivo MCL model provided a unique opportunity to perform a direct comparison of two novel therapies that are currently used clinically as single agents. A schema of the treatment protocol is provided in Fig. 5A. Mice treated with CTL019 showed a statistically significant improvement in lymphoma control compared with ibrutinib-treated mice (Fig. 5B). As depicted in Fig. 5C, all mice treated with ibrutinib monotherapy died before day 100, whereas CTL019 fostered long-term survival of the recipient mice, suggesting that CTL019 is therapeutically more effective than ibrutinib in this model.
We next tested the combination of CTL019 and ibrutinib in vivo (Fig. 6A). Because we found no difference in the antitumor effect when comparing untransduced T cells plus ibrutinib with ibrutinib alone (Supplementary Fig. S8A), in all subsequent experiments the control groups were vehicle and ibrutinib alone. Ibrutinib monotherapy led to modestly delayed disease growth at early time points, whereas CTL019 monotherapy led to a profound reduction in tumor burden that was followed by the disease progression beginning at 6–7 weeks. In striking contrast, 80% to 100% of mice treated with the combination of CTL019 and ibrutinib experienced complete, long-term disease control (Fig. 6B and C).
Histopathology of organs harvested at the conclusion of the experiment revealed MCL infiltrates in all untreated and ibrutinib-treated mice with the extent of involvement being relatively diminished in the ibrutinib-treated group. Most of the mice treated with CTL019 alone displayed persistent MCL and some CTL019 cells, while mice treated with CTL019–ibrutinib showed clearance of the tumor and disappearance of CTL019 (Fig. 6D).
Having shown that ibrutinib treatment was associated with a nonsignificant trend to lower PD-1 expression on CTL019 in the blood compartment, we next analyzed the expression of PD-1 on CTL019 in tumor-involved organs. We confirmed the presence of T cells in the livers of mice treated with CART19 and, to a lesser extent, in CART19 + ibrutinib–treated mice (Fig. 6E). Interestingly T cells from mice receiving CTL019 monotherapy had significantly higher levels of PD-1 as compared with mice receiving CTL019 + ibrutinib (Fig. 6F). We then evaluated the expression of inhibitory receptors on CTL019 cells exposed to increasing doses of ibrutinib in vitro as a possible mechanism of improved antitumor activity. Interestingly, we found that CTL019 cells cocultured with MCL-RL for 6 days markedly upregulated inhibitory receptors such as PD-1, LAG-3, TIM-3, CTLA-4 (Fig. 6G). Notably, the addition of ibrutinib to the coculture led to a significant reduction in all inhibitory receptors (Fig. 6G). This mechanism may illuminate the observation of better antitumor activity of the combination in vitro and in vivo.
Novel therapies for B-cell malignancies include small-molecule inhibitors of BCR signaling and CD19-directed T-cell–based therapies. The BTK inhibitor ibrutinib was recently approved by the FDA for the treatment of therapy-resistant MCL and engenders responses in most (68%) patients. However, these responses are typically partial and relatively short-lived: the median progression-free survival is 17.5 months (8). Anti-CD19 CAR T cell therapy leads to durable responses in subsets of patients with high-risk B-ALL (12–14), DLBCL (16), and to a lesser degree, CLL (15). Combination of chemotherapeutic agents with non–cross-resistant mechanisms of action has a long history in the treatment of cancer (33) and provides the rationale for the current study. Here we evaluated the combined effect of signal transduction (kinase) inhibition and cellular immunotherapy; these two novel therapeutic approaches are poised to revolutionize treatment of patients with lymphoma and cancer in general. Specifically, we investigated the impact of adding the BTK inhibitor ibrutinib to CTL019 using MCL as a model of a currently incurable disease responsive to both these modalities. Although ibrutinib exerted in vitro a profound detrimental effect on the sensitive MCL cells, we found that at all but high supraphysiologic doses of the drug, CTL019 cell function remains unimpaired, with intact proliferative capacity, tumor recognition and cytotoxicity, and cytokine synthesis. This observation was not a foregone conclusion, given that at least a subset of CAR T cells expresses a tyrosine kinase that is inhibited by ibrutinib (ITK). We also demonstrated an additive effect of combining BTK signaling inhibition with the direct cytotoxicity delivered by CTL019. This finding indicates that the combined ibrutinib and CART19 anti-MCL cell activity stems from their direct effect on the malignant B lymphocytes.
The in vitro studies were followed by a clear demonstration of superiority of CTL019 over ibrutinib in the MCL xenotransplant mouse model when each was used as monotherapy at clinically relevant doses and schedules of administration (single dose for CTL019, continuous administration for ibrutinib) and despite the fact that we used a higher dose of ibrutinib than that employed by most groups (20). This approach is supported by our dose-titration experiments and by the fact that the dose of ibrutinib that is used in MCL therapy is higher than that the one to treat CLL.
When combining ibrutinib with CTL019 in vivo, we observed complete and long-lasting tumor responses. We also noted higher numbers of circulating CTL019 cells; ibrutinib is known to lead to a peripheral blood lymphocytosis, predominantly thought to be due to mobilization of malignant B lymphocytes from lymph nodes through inhibition of CXCR4 pathway (34–36). To our knowledge, T-cell lymphocytosis has not been formally demonstrated in patients treated with ibrutinib. Our results indicate that the T-cell lymphocytosis is not specific to antigen-specific cells, as untransduced control T cells were also shown to increase in the peripheral blood. The observed lymphocytosis does not appear to be related to increased proliferation or enhanced T-cell survival, and may be related to differential T-cell trafficking. Current data implicate CXCR4 in malignant lymphocyte trafficking in some models (35,37) and although we did not find CXCR4 to be differentially expressed in ibrutinib-treated mice, our data do not exclude functional involvement of the CXCR4–SDF1 pathway.
Most preclinical work showing the efficacy of CTL019 has been performed using B-ALL cell lines, which are not sensitive to ibrutinib (23). Furthermore, the strongest clinical responses to date have been obtained in patients with B-ALL, whereas patients with diffuse large B-cell lymphoma and indolent B-cell lymphomas have somewhat lower response rates (15). The reasons for this seemingly tumor type–specific heterogeneous responses to CTL019 remain to be elucidated.
The kinetics of the tumor response and subsequent progression suggest that ibrutinib either deepens the initial response achieved by CTL019 alone, or enhances the long-term immunosurveillance capacity of CTL019 cells. In an infectious model, Dubovsky and colleagues (20) showed that ibrutinib enhances the percentage of antigen-specific CD8 T cells and increases the percentage of both CD4 and CD8 T cells that bear CD62L, a marker of memory T-cell differentiation. However, we did not see changes in T-cell polarization, effector function, or memory subsets in the combination therapy in our model; if found, these would have pointed toward immunologic memory as a potential mechanism of action. The most stringent test for initiation of memory is by tumor rechallenge in animals that have cleared disease. However, in this model, the only animals that successfully cleared tumor long-term are those who received the combination therapy and therefore there is not a suitable control group with which to compare. Therefore, the exact mechanism(s) of the strong antilymphoma effect of the CTL019/ibrutinib combination remains to be elucidated but most likely reflects the advantage of simultaneous direct targeting of malignant cells with two therapeutic modalities with vastly different modes of action. The observation that T cells, including CTL019 cells, are mobilized into the peripheral blood may also help to explain the augmented antitumor effect that we observed.
Recently, ibrutinib has been found to enhance the antitumor effect of blockade of the PD1/PD-L1 system in mouse models (38), a phenomenon that was accompanied by enhanced antitumor immune responses. These authors did not show reduction of PD1 or PD-L1 molecules upon exposure to ibrutinib. In contrast, here we found that tumor-infiltrating CTL019 cells had lower PD-1 expression if the animals were also treated with ibrutinib and these results were further corroborated by in vitro studies showing that exposure to MCL cells led to a marked increase in inhibitory receptors (“immune checkpoint molecules”) on CTL019 that was partially abrogated by cotreatment with ibrutinib. These observations may suggest that this two-pronged antitumor approach derives additional synergy from ibrutinib-mediated T-cell mobilization and from ibrutinib-mediated reduction in inhibitory receptor expression on CAR T cells.
Regardless of the above uncertainties, this is the first preclinical study that combines signal transduction inhibition with adoptive T-cell immunotherapy by targeting BTK and CD19, respectively. Our findings document a potent additive therapeutic effect of this novel and highly promising combination acting by enhanced killing of the MCL cells. They also pave the way for clinical trials of this and similar non–cross-resistant combinations in patients with MCL and other types of B-cell lymphoma.
Disclosure of Potential Conflicts of Interest
M.V. Maus and M. Kalos have ownership interest (including patents) in Novartis. S.F. Lacey, M. Milone, C.H. June, S. Gill and M.A. Wasik report receiving commercial research grants from Novartis. S.J. Schuster is a consultant/advisory board member for Pharmacyclics, and reports receiving commercial research support from Janssen, Novartis, and Pharmacyclics. M. Ruella, S.S. Kenderian, J.A. Fraietta, M.V. Maus, M. Milone, M. Kalos, C.H. June, S. Gill, and M.A. Wasik are listed as co-inventors on patents in the area of CAR T cells that are owned by the University of Pennsylvania and licensed to Novartis. All authors work under a research alliance involving the University of Pennsylvania and Novartis pharmaceuticals. No potential conflicts of interest were disclosed by the other authors.
Conception and design: M. Ruella, S.S. Kenderian, J.A. Fraietta, Q. Zhang, S.J. Schuster, M. Kalos, S. Gill, M.A. Wasik
Development of methodology: M. Ruella, S.S. Kenderian, J.A. Fraietta, Q. Zhang, M.V. Maus, M. Milone, S. Gill
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M. Ruella, S.S. Kenderian, O. Shestova, J.A. Fraietta, M.V. Maus, X. Liu, S. Nunez-Cruz, M. Klichinsky, O.U. Kawalekar, S.F. Lacey, M. Milone, A.R. Mato, S.J. Schuster, S. Gill
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M. Ruella, S.S. Kenderian, J.A. Fraietta, S. Qayyum, Q. Zhang, X. Liu, O.U. Kawalekar, S. Gill, M.A. Wasik
Writing, review, and/or revision of the manuscript: M. Ruella, S.S. Kenderian, S. Nunez-Cruz, A. Mato, S.J. Schuster, M. Kalos, C.H. June, S. Gill, M.A. Wasik
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M. Ruella, O. Shestova, Q. Zhang
Study supervision: M. Ruella, Q. Zhang, S.F. Lacey, C.H. June, S. Gill, M.A. Wasik
The authors thank Fang Chen and Natalka Kengle for performing Luminex assay.
This was supported by grants from the University of Pennsylvania-Novartis Research Alliance, Lymphoma Research Foundation (2013–2015 MCL E/D), and JB Cox Charitable Lead Trust. S. Gill is an American Society of Hematology Scholar. M.V. Maus is supported by NCI K08-166039. M. Klichinsky is funded by a NIH T32-GM008076. Imaging was performed at the University of Pennsylvania Small Animal Imaging Facility (SAIF) Optical/Bioluminescence Core, supported by NIH grant CA016520. S.J. Schuster received Philanthropic support from the Jim and Frannie Maguire Lymphoma Research Fund.
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.