Purpose:

Inhibitors of Bruton's tyrosine kinase (BTKi) and PI3K (PI3Ki) have significantly improved therapy of chronic lymphocytic leukemia (CLL). However, the emergence of resistance to BTKi has introduced an unmet therapeutic need. Hence, we sought evidence for essential roles of PI3K-δi and PI3K-γi in treatment-naïve and BTKi-refractory CLL.

Experimental Design:

Responses to PI3K-δi, PI3K-γi, and the dual-inhibitor duvelisib in each B, T, and myeloid cell compartments of CLL were studied in vitro, and in a xenograft mouse model using primary cells from treatment-naïve and ibrutinib-resistant patients, and finally, in a patient with ibrutinib-resistant CLL treated with duvelisib.

Results:

We demonstrate the essential roles of PI3K-δ for CLL B-cell survival and migration, of PI3K-γ for T-cell migration and macrophage polarization, and of dual inhibition of PI3K-δ,γ for efficacious reduction of leukemia burden. We also show that samples from patients whose disease progressed on ibrutinib were responsive to duvelisib therapy in a xenograft model, irrespective of BTK mutations. In support of this, we report a patient with ibrutinib-resistant CLL, bearing a clone with BTK and PLCγ2 mutations, who responded immediately to single-agent duvelisib with redistribution lymphocytosis followed by a partial clinical remission associated with modulation of T and myeloid cells.

Conclusions:

Our data define the mechanism of action whereby dual inhibition of PI3K-δ,γ affects CLL B-cell numbers and T and myeloid cell pro-leukemia functions and support the use of duvelisib as a valuable approach for therapeutic interventions, including for patients refractory to BTKi.

Translational Relevance

The dual PI3K-δ and PI3K-γ inhibitor duvelisib has been approved for relapsed chronic lymphocytic leukemia (CLL). Here we extensively studied how targeting each subunit of PI3K contributes to the efficacy of duvelisib. We demonstrate the specificity of PI3K-δ for the growth and migration of CLL B cells, and the function of PI3K-γ in CLL patient-derived T cells and myeloid cells. As Bruton's tyrosine kinase mutations have been acquired in patients with CLL relapsed to the front-line ibrutinib therapy, we report the efficacy of duvelisib in a patient with ibrutinib-resistant CLL using clinical samples acquired at the beginning and timepoints after duvelisib treatment. Similar results are shown for another 3 patients with ibrutinib-resistant CLL in a preclinical xenograft model. Our study suggests that the combination treatment of duvelisib may be used to improve the treatment in relapsed disease, including ibrutinib-resistant CLL.

Chronic lymphocytic leukemia (CLL) results from the clonal expansion and accumulation of CD19+CD5+ B lymphocytes (1, 2). CLL B cells shape the microenvironment in which they survive and grow, by acting, directly and indirectly on various cell populations, including T cells, macrophages, mesenchymal stromal cells, follicular dendritic cells, and natural killer cells.

Like normal B cells, CLL cells are activated by signals delivered through multiple receptors, including the B-cell receptor for antigen (BCR) and cytokine and chemokine receptors (1, 2). PI3K, which is essential for BCR-regulated malignant B-cell growth (3, 4), signal by four PI3K isoforms (α, β, δ, γ), each with distinct functional roles. PI3K-δ and PI3K-γ are specifically expressed in leukocytes (5). PI3K-δ is critical throughout the B-cell life cycle, being essential for B lymphocyte survival (6, 7), with some effects in T cells shown in vitro (8); in contrast, PI3K-γ plays a vital role in macrophages that control immune stimulation and suppression by regulating the activation and function of cytolytic T cells (9).

Duvelisib (IPI-145) is an oral, dual PI3K-δ,γ inhibitor (5) that potently inhibits PI3K-δ and PI3K-γ in CLL B cells (5, 10). Dual inhibition offers a novel approach to treat patients with CLL with enhanced efficacy by targeting both CLL cells and CLL-supporting cells in the tumor microenvironment (TME). In vitro, duvelisib sensitizes primary CLL cells for apoptosis and abrogates bone marrow (BM) stromal cell–mediated CLL-cell survival by inhibiting BCR-mediated signaling and chemotaxis (10, 11).

Of the several approved targeted agents in the CLL treatment armamentarium, the most widely used is ibrutinib, a Bruton's tyrosine kinase inhibitor (BTKi) approved for first-line and relapsed-disease treatment in CLL and other B-cell malignancies (12–14). Ibrutinib is taken continuously until progression of disease or intolerance develops. Resistance to ibrutinib can occur through the development of mutations in the BTK-binding site, for example, a BTK C481S, and/or of mutations in mediators of BCR-signaling downstream of BTK, for example, PLCG2 (12–14). Because mutations in BTK and PLCγ2 are found in approximately 80% of patients with CLL with acquired resistance to ibrutinib (15), each can lead to expansion of ibrutinib-resistant subclones. In certain patients, therapy is immediately needed upon drug resistance as the disease can progress rapidly (16). Little is known about the role of dual PI3K-δ,γ inhibition as a salvage approach for patients who develop these mutations.

Here we aimed to understand how duvelisib targets leukemic B cells and the TME and can hence affect CLL, even BTK-resistant disease, as has been shown for the C481S-mutant XLA B cell line (17). We performed in vitro assays and utilized a patient xenograft model of CLL (18) to explore the mechanisms of action of duvelisib and to characterize the distinct functions of PI3K-δ and PI3K-γ in CLL B cells and CLL-supporting cells. This documented that, in vitro and in vivo, the two PI3K isoforms play distinct roles in sustaining the survival of CLL cells and that, in vivo using xenografts, the inhibitory activity of duvelisib is effective against CLL B cells from patients at various diseases stages and from those whose disease progressed while receiving ibrutinib, even in the setting of BTK and/or PLCG2 mutations. To extrapolate these findings to clinical care, we describe a patient who underwent a clinical remission upon dual PI3K-δ,γ inhibition after disease progression on ibrutinib.

Samples from patients with CLL

Studies were approved by the Institutional Review Board of Northwell Health. Patients with treatment-naïve and ibrutinib-resistant CLL were selected using current International Workshop on Chronic Lymphocytic Leukemia (IWCLL) guidelines (19). All patients provided written informed consent to participate. IGHV-IGHD-IGHJ sequences of CLL clones were determined by direct sequencing of PCR products from these samples as described; clones with ≤2% difference from the most similar germline IGHV gene were considered U-CLL, and those with >2% difference from the most similar germline IGHV gene were designated as M-CLL.

All patients used in the study were treatment-naïve patients, except for the 3 patients who exhibited clinical resistance to ibrutinib therapy (CLL1782, CLL1273, and CLL1570; Table 1). CLL1782 and CLL1570 received ibrutinib monotherapy as a participant in the RESONATE trial (20, 21), and CLL1273 was treated with commercially available ibrutinib (standard of care). Targeted DNA sequencing for BTK C481S of the leukemic clone of patients CLL1782 and CLL1273 was performed; genomic profiling was conducted on a BM biopsy from patient CLL1570 within 1 month of discontinuing ibrutinib. Patient CLL1570 immediately received commercially available duvelisib as salvage therapy after discontinuation of ibrutinib treatment.

Table 1.

Mutation characterization of ibrutinib-resistant CLL samples used in this study.

Patient identifierIGHVIbrutinib progressionBTK mutationPLCG2 mutation
CLL1782 Unmutated Yes Wildtype  
CLL1273 Mutated Yes C481S  
CLL1570 Mutated Yes C481S PLCG2 L845F 
Patient identifierIGHVIbrutinib progressionBTK mutationPLCG2 mutation
CLL1782 Unmutated Yes Wildtype  
CLL1273 Mutated Yes C481S  
CLL1570 Mutated Yes C481S PLCG2 L845F 

Therapeutic response was assessed according to IWCLL criteria (19). Response assessments included evaluation of disease-related constitutional symptoms; focused physical examinations; complete blood counts and differential white blood cell counts; BM biopsy; and CT scans, when indicated.

Xenograft model

Alymphoid NSG mice were maintained on a 12:12-hour light/dark schedule. All experiments were conducted in accordance with the guidelines of the Northwell Health Institutional Animal Care and Use Committee.

Engraftment and growth of human CLL cells in NSG mice was carried out as described previously (18). Briefly, CD3+ T cells were isolated from CLL peripheral blood mononuclear cells (PBMC) using EasySep (STEMCELL Technologies) and preactivated for 7 to 10 days with anti–CD3/CD28-coated Dynabeads and IL2 following manufacture's recommendations (Thermo Fisher Scientific). NSG mice received 0.5 × 106 activated T cells and 20 × 106 autologous CLL PBMCs injected retro-orbitally. Two weeks later, mice were treated with duvelisib (70 or 100 mg/kg), a selective PI3K-δi (IPI-3063; 10 mg/kg), and/or a selective PI3K-γi (IPI-5243; 15 mg/kg) 5 days per week for 3 weeks by oral gavage. These doses were determined on the basis of IC50 inhibitory values of the different agents used in these assays. For the four class I PI3k isoforms are listed in Supplementary Table S1.

A similar approach was used to evaluate the effects of these reagents on CLL-cell growth in NSG mice bearing ibrutinib-resistant CLL PBMCs. Leukemia-bearing mice were treated with duvelisib (100 mg/kg; 5 days a week via oral gavage) or ibrutinib (25 mg/kg; daily via drinking water). After 3 weeks of treatment, murine spleens were assessed by flow cytometry (FACSymphony, BD Biosciences) for patient-derived CLL B cells (gated on human CD45+CD19+CD5+ cells), proliferating CLL B cells (human CD45+CD19+Ki67+ cells), patient-derived T cells (human CD45+CD3+CD19 cells), or M2 macrophages (mouse F4/80+CD11b+CD11CCD206+ cells or human CD14+CD80CD163+ CD206+ cells).

In vitro CLL-cell viability assessment

For viability assays, patient-derived CLL cells were incubated with PI3K inhibitor (PI3Ki) in an enriched medium containing an insulin/transferrin/selenium supplement plus beta-mercaptoethanol (BioWhittaker; ref. 22), stained with the SYTOX Dead Cell Stains (Thermo Fisher Scientific) at the indicated timepoints, and then analyzed by flow cytometry. Resultant data were calculated using FlowJo software (FlowJo, LLC). For BM-derived monocytes (BMDM)-supported CLL-cell survival assays, BMDMs were polarized with 50 ng/mL MCSF for 6 days. After treatment with duvelisib for 1 hour, BMDMs were cultured with MCSF and IL4 for 12 hours. Finally, BMDM were added to leukemic B cells at a ratio of 4:1 (monocytes to CLL cells) and CLL-cell viability assessed via Guava ViaCount assay (MilliporeSigma) on days 0 and 5 of culture.

In vitro CLL-cell proliferation assay

Human PBMCs obtained from whole blood of CLL donors were purchased from a commercial source (Bioreclamation IVT or ALLCELLS) were resuspended (1 × 106 cells/mL) in RPMI1640 medium containing 50 μmol/L β-mercaptoethanol, 10% heat-inactivated FBS, and 1,000 units/mL of penicillin and streptomycin. Cells were seeded at a density of 1 × 106 cells/mL/well in 24-well tissue culture plates. The PI3Ki IPI-549 was added at varying concentrations together with 1 μg/mL sCD40L (GIBCO), 10 ng/mL IL10 (Gibco), and 10 ng/mL IL2 (PeproTech). Cells were incubated at 37°C for 3 days. Percent inhibition of Ki67+pAKT+ CLL cells was calculated by flow cytometry. These values were used to generate EC50 values using GraphPad Prism Software. This experiment was repeated with additional patients using primary CLL patient PBMCs collected at the Northwell Health; sorted CLL B cells were cultured in the same condition as above for 3 days. Results shown are the absolute number or the percent inhibition of Ki67+ CLL cells calculated by flow cytometry.

In vitro T-cell migration assay

CLL PBMCs collected at Northwell Health were pretreated with DMSO, duvelisib, PI3K-γi (IPI-549), or PI3K-δi (IPI-3063) for 48 hours and then allowed to migrate across a 6.5-mm polycarbonate insert with a 5-μm pore size in a transwell (CoStar/Corning) plate in response to a 3-hour exposure at 37°C to CXCL12 (300 ng/mL; R&D Systems). Cells were harvested from the lower chamber, and CD3+CD4+ or CD3+CD8+ cells were quantified by flow cytometry. The percent inhibition of T-cell migration was calculated in response to each of the three PI3Ki used and compared with the DMSO control. EC50 values were then calculated using GraphPad Prism Software.

In vivo CLL-cell and T-cell migration assay in NSG mice

CLL PBMCs collected at Northwell Health were pretreated with PI3Kis for 48 hours and then injected into NSG recipients. Twenty-four hours later, mice were sacrificed, and the numbers of CLL B cells in spleen and blood were measured. Migration of T cells from patients with CLL was measured by intravenously injecting 5 × 106 T cells preactivated for 2 days by anti-CD3/28 Dynabeads + IL2 in the presence of PI3Ki.

M2 macrophage polarization

BM cells collected from 4-week-old wild-type male C57BL/6 mice (n = 9) were plated in macrophage differentiation culture media (DMEM + 20% FBS + 1% Pen/Strep + 50 ng/mL MCSF) for 6 days (23). BM cells were then transferred to 96-well plates and pretreated for 1 hour with various doses of duvelisib, PI3K-δi (IPI-3063), or PI3K-γi (IPI-549) in triplicate. After 1 hour, MCSF (50 ng/mL) and IL4 (20 ng/mL) were added to the culture. A total of 48 hours later, RNA was extracted from cells per manufacturer's protocol (Qiagen), and Arg1 mRNA levels were determined by qRT-PCR in triplicate. The threshold cycle (CT) values for Arg1 and Actb were obtained using an ABI7300 TaqMan instrument. IC50 calculations were obtained using Graphpad Prism by nonlinear regression of log inhibitor concentration versus percent inhibition of Arg1 expression.

Intracellular staining for cytokine-producing cells

Samples were stimulated with phorbol 12-myristate 13-acetate (10 ng/mL) and ionomycin (1 μmol/L) for 4 hours at 37°C in the presence of monensin (10 μg/mL), exposed to mAbs reactive with IL4, IFNg, IL17A, or IL17F for 30 minutes at room temperature and then analyzed intracellularly by flow cytometry for cytokine-producing live cells.

Statistical analyses

All statistical tests were performed using Prism v8 (GraphPad Software, Inc). Normality was assessed using the D'Agostino-Pearson Omnibus test, and appropriate parametric and nonparametric analyses performed thereafter. Mann–Whitney U tests were used to compare the numbers of CLL cells and T cells and murine macrophages and the percentages of Ki67+ cells. An unpaired parametric t test was used for the comparisons of CLL B-cell viability at day 5. Figures show data represented as means or means ± SEM. P < 0.05 was considered significant for all tests.

Ethics statement

The studies involving human participants were reviewed and approved by Institutional Review Board (approval no. 08-202A), The Feinstein Institutes for Medical Research. The patients/participants provided their written informed consent to participate in this study. Informed consent was collected according to the Declaration of Helsinki, and ethical approval was granted by local review committees.

Data availability

All data generated or analyzed for this study are included in this published article and its Supplementary Data. All the data are also available from the corresponding author upon request.

Inhibition of PI3K-δ impairs CLL-cell proliferation, survival, and migration in vitro and in vivo

We first assessed the distinct contributions of PI3K-δ and PI3K-γ on leukemic B-cell survival, proliferation, and migration in vitro by using PI3K-δ−specific inhibitor (PI3K-δi, IPI-3063) or PI3K-γ−specific inhibitor (PI3K-γi, IPI-549) in addition to duvelisib that inhibits both isoforms (Supplementary Table S1). All in vitro experiments were carried out using concentrations of inhibitors that are clinically achievable and tolerable in patients (range: 1 nmol/L to 1 μmol/L).

To evaluate the direct effects of PI3Ki on CLL-cell survival, purified leukemic B cells from 4 patients were incubated with PI3K-δi, or PI3K-γi, or duvelisib for 48 or 96 hours in an enriched medium that better supports cell survival (22), and the percentage of live CLL cells were quantified. Although at 48 hours, cell viability was not altered (Supplementary Fig. S1A, top), a significant reduction in viable CLL cells was detected at 96 hours after exposure to PI3K-δi and duvelisib with IC50 values at 100 nmol/L and 1 μmol/L (Supplementary Fig. S1A, bottom). In contrast, PI3K-γi did not significantly affect CLL-cell viability (Fig. 1A). Data in the figure were normalized to the values in the control, which was set as 1.

Figure 1.

Inhibition of PI3K-δ impairs CLL-cell proliferation, survival, and migration in vitro and in vivo. Dose-dependent effects of PI3K-δi monotherapy (IPI-3063), PI3K-γi monotherapy (IPI-549), and dual PI3K-δ,γi inhibition (duvelisib) were analyzed in 4 patients with CLL (2 U-CLL and 2 M-CLL) for survival of CLL cells 96 hours after incubation with inhibitors ranged from 0.1 to 1,000 nmol/L (A); CLL-cell proliferation after stimulation by sCD40L, IL2, and IL10 for 72 hours together with PI3Kis ranged from 10−5 to 104 nmol/L (B); and in vivo migration after treating CLL cells with PI3K-γi, PI3K-δi, or duvelisib at various doses (0 to 10 nmol/L) for 48 hours (C). A total of 5 × 106 live cells obtained from each culture condition were then injected into NSG mice. A total of 24 hours later, CLL patient cells moved to spleens were then quantified by flow cytometry. Data shown represent means ± SEM. *, P < 0.05; **, P ≤ 0.01 analyzed by two-way ANOVA.

Figure 1.

Inhibition of PI3K-δ impairs CLL-cell proliferation, survival, and migration in vitro and in vivo. Dose-dependent effects of PI3K-δi monotherapy (IPI-3063), PI3K-γi monotherapy (IPI-549), and dual PI3K-δ,γi inhibition (duvelisib) were analyzed in 4 patients with CLL (2 U-CLL and 2 M-CLL) for survival of CLL cells 96 hours after incubation with inhibitors ranged from 0.1 to 1,000 nmol/L (A); CLL-cell proliferation after stimulation by sCD40L, IL2, and IL10 for 72 hours together with PI3Kis ranged from 10−5 to 104 nmol/L (B); and in vivo migration after treating CLL cells with PI3K-γi, PI3K-δi, or duvelisib at various doses (0 to 10 nmol/L) for 48 hours (C). A total of 5 × 106 live cells obtained from each culture condition were then injected into NSG mice. A total of 24 hours later, CLL patient cells moved to spleens were then quantified by flow cytometry. Data shown represent means ± SEM. *, P < 0.05; **, P ≤ 0.01 analyzed by two-way ANOVA.

Close modal

We next investigated the direct effects of inhibiting PI3K-δ or PI3K-γ individually or simultaneously on CLL-cell growth using the percentage of Ki67+ CLL cells as an indicator of tumor cell proliferation, as well as the levels of phosphorylated AKT (p-AKT) as a downstream effect of PI3K to document inhibition of the two PI3Ks. To mimic the survival and proliferation signals provided by non-neoplastic cells in the TME, CLL cells were stimulated for 72 hours with soluble(s) CD40L, IL2, and IL10 in the absence or presence of PI3Kis over a broad range of concentrations (10−5 to 104 nmol/L). Results, measured by Ki67 and p-AKT levels, suggest that duvelisib (EC50 = 0.46 nmol/L) and PI3K-δi (EC50 = 0.10 nmol/L) more potently block CLL-cell proliferation than PI3K-γi (EC50 = 256 nmol/L; Fig. 1B).

Finally, we assessed the effects of individual or combined PI3K-γi or PI3K-δi on CLL-cell migration in vivo. CLL cells were preincubated with the PI3K-γi or PI3K-δi or duvelisib for 48 hours at doses at 0.1 to 10 nmol/L, a dose range that did not show significant changes in CLL-cell viability (Fig. 1B). Then equal numbers of viable leukemic B cells were injected intravenously into NOD-scid IL2Rγnull (NSG) mice, and the numbers of CLL cells in the spleens and blood of recipient mice were quantified 24 hours later. While PI3K-γi did not change CLL-cell homing, PI3K-δi and duvelisib significantly reduced the migration of CLL B cells into the spleen relative to control at the doses of 1 and 10 nmol/L (Fig. 1C); in contrast, there were more CLL cells remaining in the blood of treated mice, although the difference was not significant (number of circulating cells in control vs. 10 nmol/L duvelisib treated mice: 230 ± 50 vs. 4,200 ± 2,500, P > 0.05). Overall, these data indicate that PI3K-δi and duvelisib significantly diminish CLL-cell survival, proliferation, and migration.

The results above suggest similar effects between PI3K-δi and duvelisib, but because of the differences in potency (IC50) between PI3K-δi, PI3K-γi, and duvelisib (Supplementary Table S1), it is difficult to understand the additive or synergistic effects of PI3K-δi and PI3K-γi. Thus, we repeated the above experiments with samples from 4 additional patients with CLL. Isolated primary patient B cells were treated with 10 nmol/L PI3K-δi, PI3K-γi, the combination of the two inhibitors, or duvelisib as a way to directly investigate the synergistic effects of targeting both PI3K-δ and PI3K-γ (Supplementary Fig. S1). In all analyses, only PI3Kδi but not PI3K-γi inhibited in vitro CLL-cell viability (Supplementary Fig. S1A), in vitro CLL-cell proliferation upon stimulation (Supplementary Fig. S1B and S1C), and in vivo CLL cell homing to spleens (Supplementary Fig. S1D). Treatment of duvelisib and the combination of PI3K-δi and PI3K-γi achieved levels of inhibition similar to treatment with only PI3K-δi. Overall, these data suggest no additive effects of targeting both PI3K-δ and PI3K-γ in pure CLL B cells for their function of survival, proliferation, and migration. In addition, when comparing the data from the total of 8 patient samples used in Fig. 1 and Supplementary Fig. S1, there is no difference in responsiveness to PI3Ki between patients with IGHV-unmutated (U-CLL) and IGHV-mutated (M-CLL) disease.

Inhibition of PI3K-γ impairs the ability of T cells and macrophages to support CLL-cell growth and survival

Previous in vitro cellular assays, respiratory models, and autoimmune and inflammatory disease models have indicated a role for both PI3K-δ and PI3K-γ in the regulation of T-cell subsets (5, 24). However, how PI3K-δ and PI3K-γ affect the functions of CLL-derived T or myeloid cells in vivo, and how such functional changes affect the crosstalk between these nonmalignant cells and CLL cells has not been elucidated. We therefore investigated the impact of single versus dual inhibitory agents on two key cellular components of the CLL TME, T cells, and myeloid cells.

First, CXCL12-induced T-cell migration was evaluated in vitro. CXCL12 is a potent chemotactic signal for lymphocytes and mesenchymal stem cells, thereby leading to a protumor TME (11). CLL PBMCs treated with various concentrations of DMSO, duvelisib, PI3K-γi (IPI-549), or PI3K-δi (IPI-3063) were allowed to migrate in transwell plates in response to a 3-hour exposure to CXCL12. Afterward, cells were harvested from the lower chamber, and the percentage of migrated T cells was quantified. At nonlethal doses (<1 μmol/L), inhibition of CD3+, CD4+, and CD8+ T-cell migration in response to CXCL12 differed for the three inhibitors, with PI3K-γi (EC50: 17± 17 nmol/L) being more effective than duvelisib (128 ± 39 nmol/L), and PI3Kδi being highly ineffective (EC50 630 ± 71 nmol/L; Fig. 2A).

Figure 2.

Inhibition of PI3K-γ impairs migration of CLL-derived T cells. A, CXCL12-induced migration of CD3+ T cells from 3 patients with CLL after the treatment of PI3Ki ranged from 1 to 104 nmol/L. EC50 are represented as means ± SD. B, Schematic of technique used to evaluate in vivo migration of CLL T cells. Equal numbers of T cells, collected after treatment with PI3K-γi (IPI-549), PI3K-δi (IPI-3063), or duvelisib, were injected into NSG mice retro-orbitally. T-cell migration to the spleen, and the contributions of the PI3K-γi versus PI3K-δi components of duvelisib (ranged from 0.1 to 10 nmol/L) on T-cell migration were evaluated 24 hours later. Samples for 4 patients with CLL were tested. Absolute number of T cells that migrated to spleen were calculated using flow cytometry and shown as means ± SEM. Statistical analyses were done using two-way ANOVA. Duvelisib versus PI3K-δi: P = 0.03; PI3K-γi versus PI3K-δi: P < 0.01.

Figure 2.

Inhibition of PI3K-γ impairs migration of CLL-derived T cells. A, CXCL12-induced migration of CD3+ T cells from 3 patients with CLL after the treatment of PI3Ki ranged from 1 to 104 nmol/L. EC50 are represented as means ± SD. B, Schematic of technique used to evaluate in vivo migration of CLL T cells. Equal numbers of T cells, collected after treatment with PI3K-γi (IPI-549), PI3K-δi (IPI-3063), or duvelisib, were injected into NSG mice retro-orbitally. T-cell migration to the spleen, and the contributions of the PI3K-γi versus PI3K-δi components of duvelisib (ranged from 0.1 to 10 nmol/L) on T-cell migration were evaluated 24 hours later. Samples for 4 patients with CLL were tested. Absolute number of T cells that migrated to spleen were calculated using flow cytometry and shown as means ± SEM. Statistical analyses were done using two-way ANOVA. Duvelisib versus PI3K-δi: P = 0.03; PI3K-γi versus PI3K-δi: P < 0.01.

Close modal

Next, migration of CLL-derived T cells was assessed in vivo, after stimulating them in vitro with anti-CD3/28 Dynabeads + IL2 for 2 days in the presence of PI3K-γi, PI3K-δi, or duvelisib, at doses ranging from 0 to 1 μmol/L. Notably, none of these treatments altered T-cell growth (Supplementary Fig. S2A) nor surface levels of CXCR4, CXCR5, CCR6, and CCR7 (not shown). However, when equal numbers (5 × 106) of T cells from each treatment condition were injected intravenously into NSG mice, migration to the spleen at 24 hours was significantly reduced for T cells that had been incubated with PI3K-γi and duvelisib, but not for those exposed to PI3K-δi pretreatment. Specifically, duvelisib and PI3K-γi alone but not PI3K-δi alone significantly inhibited T-cell homing, while PI3K-γi and duvelisib were not significantly different from each other (Fig. 2B). Thus, PI3K-γ function is critical for the migration of CLL-derived T cells. Notably, duvelisib also effectively reduces T-cell migration, presumably by its action on PI3K-γ.

To prove that this presumption was the case, we next investigated the role of PI3K-δi or PI3K-γi or the combination had on CLL patient-derived T cells, by repeating the experiments above with samples from an additional 4 patients with CLL. Using 10 nmol/L of each inhibitor, we found similar levels of inhibition in T-cell migration toward CXCL12 with the treatments of PI3K-γi alone or with the combination of PI3K-δi (Supplementary Fig. S2B). Similar results were also seen in vivo. PI3K-γi, or PI3K-γi + PI3K-δi, or duvelisib reached similar level of inhibition in T-cell homing (Supplementary Fig. S2C). Thus, these data are consistent with duvelisib inhibiting T-cell trafficking by inhibition of PI3K-γ, and also indicate that targeting both PI3K-δ and PI3K-γ does not provide increased inhibition of T-cell migration.

Tumor-associated macrophages with an “M2 phenotype” contribute to a protumor TME by preventing the induction of T cell–mediated antitumor immunity that retards the growth and survival of tumor cells (25). PI3K-γ can regulate immune suppression by polarizing myeloid cells to the M2 phenotype (9), which is reflected by increased Arg1 (arginase 1) mRNA expression (26) or the development of an F4/80+CD11b+CD11c+CD206+ surface phenotype (27). Thus, to analyze the effects of PI3Ki on macrophages in CLL, we expanded murine BMDM in vitro with MCSF and then polarized them to M2 macrophages using MCSF plus IL4 (23), in the absence or presence of duvelisib (Fig. 3A). A total of 36-hour treatment with duvelisib significantly reduced Arg1 expression at doses equal to or above 10 nmol/L, indicating dual inhibition of PI3K-γ and PI3K-δ blocks M2 polarization (Fig. 3B).

Figure 3.

Inhibition of PI3K-γ reduces M2 macrophage polarization and M2 macrophage support of CLL-cell survival. A, Schematic of in vitro CLL and M2 macrophage coculture system and assays. Experiment was carried out with samples from 9 patients with CLL. B,Arg1 expression levels determined by qRT-PCR in M2 macrophage–polarized BMDMs treated with duvelisib. C, Effect of the addition of duvelisib to cocultures of CLL cells with M2-polarized macrophages. All the tests were performed using nonlethal doses of duvelisib. D and E,Arg1 expression levels determined by qRT-PCR in M2 macrophage–polarized BMDMs treated with PI3K-γi or PI3K-δi in the presence of MCSF (50 ng/mL) and IL4 (20 ng/mL). Data shown are means of relative expression levels of Arg1 using control DMSO as 1. *, P < 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

Figure 3.

Inhibition of PI3K-γ reduces M2 macrophage polarization and M2 macrophage support of CLL-cell survival. A, Schematic of in vitro CLL and M2 macrophage coculture system and assays. Experiment was carried out with samples from 9 patients with CLL. B,Arg1 expression levels determined by qRT-PCR in M2 macrophage–polarized BMDMs treated with duvelisib. C, Effect of the addition of duvelisib to cocultures of CLL cells with M2-polarized macrophages. All the tests were performed using nonlethal doses of duvelisib. D and E,Arg1 expression levels determined by qRT-PCR in M2 macrophage–polarized BMDMs treated with PI3K-γi or PI3K-δi in the presence of MCSF (50 ng/mL) and IL4 (20 ng/mL). Data shown are means of relative expression levels of Arg1 using control DMSO as 1. *, P < 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

Close modal

In CLL, macrophages provide survival signals for leukemic B cells through PI3K-dependent AKT activation (28, 29), and increasing the M1 to M2 ratio alters the ability of a CLL-cell line to engraft and grow in NSG mice (30). Therefore, we investigated whether PI3K-γ inhibition impeded macrophage-facilitated CLL-cell survival (Fig. 3C). When the same experiment was performed with PI3K-δi or PI3K-γi monotherapy, PI3K-γi (100 nmol/L), but not PI3K-δi, significantly inhibited M2 polarization, while at 1 μmol/L dose, both inhibitors blocked polarization significantly (Fig. 3D and E). Coculturing leukemic B cells with M2-polarized murine macrophages for 5 days increased CLL-cell survival compared with cultures of CLL cells alone, consistent with M2 macrophages playing a supportive role (Fig. 3C). Notably, addition of duvelisib to leukemic B cell/M2 macrophage cocultures significantly reduced the number of M2 macrophages (Supplementary Fig. S3A), as well as the survival of CLL cells at doses > 500 nmol/L at the end of 5-day culture (Fig. 3C). Because duvelisib at 100 nmol/L inhibited CLL-cell survival when the leukemic cells were cultured without macrophages (Fig. 1A), the higher concentration of duvelisib required in the M2 macrophage coculture system to achieve similar results also suggests a protective function of murine M2 macrophages for CLL cells.

In summary, these experiments indicate that inhibition of PI3K-γ alone decreases the trafficking and homing abilities of CLL-derived T cells. Moreover, PI3K-γ modulates macrophage function, decreasing their ability to support leukemic B-cell survival and growth.

Dual inhibition of PI3K-δ,γ reduces the numbers of CLL B cells, CLL-supporting T cells, and myeloid cells in vivo

To assess the impact of duvelisib on CLL cells and CLL-supporting T cells as well as murine myeloid cells located in lymphoid tissues in vivo, we used a CLL xenograft model that allows CLL B- and T-cell engraftment and growth in recipient alymphoid mice. Briefly, CLL patient PBMCs and patient-derived T cells, the latter previously activated in vitro, were injected retro-orbitally into NSG mice and allowed to engraft for 2 weeks prior to administering 70 or 100 mg/kg duvelisib via oral gavage once daily for 3 weeks (Fig. 4A). At 100 mg/kg, duvelisib significantly reduced splenic CLL-cell numbers relative to vehicle control (Fig. 4B), and significantly diminished the percentage of proliferating (Ki67+) CLL cells in the spleen in a dose-dependent manner (Fig. 4C). The efficacy of duvelisib in blocking the growth of tissue-resident CLL cells was similar for samples from patients with U-CLL and M-CLL who differ in disease outcome (Supplementary Fig. S3B). Importantly, for all patient samples, duvelisib significantly reduced the number of patient-derived T cells in the spleen (Fig. 4D) as well as decreasing the number of endogenous murine splenic M2 macrophages defined by F4/80+CD11b+CD11c+CD206+ surface phenotype (Fig. 4E).

Figure 4.

Duvelisib treatment alters the TME that supports CLL-cell growth. A, Schematic of xenograft model using primary CLL cells. B–E, Quantification of human CLL cells and T cells and murine myeloid cells by flow cytometry in mice receiving cells from patients with CLL. A total of 3–5 mice per patient per treatment condition (control vehicle, 70 or 100 mg/kg duvelisib). B, Number of CLL B cells in the spleens after duvelisib treatment. C, Number of proliferating CLL B cells in the cell cycle after duvelisib treatment. D, Number of T cells in mice treated with duvelisib. E, Number of murine GRlowLY-6ClowF4/80+CD11b+CD11c+CD206+ macrophages in recipient mouse spleens. *, P < 0.05; **, P ≤ 0.01. F–I, Quantification of human CLL cells by flow cytometry in mice receiving cells from patients with CLL. A total of 3–5 mice per patient per treatment condition. Numbers of CLL cells (F), CLL-derived T cells (G), and murine macrophages (H) in the spleens of mice treated 5 days each week for 3 weeks with PI3K-γi (15 mg/kg, IPI-5243), PI3K-δi (10 mg/kg; IPI-3063), or duvelisib (100 mg/kg, Duv). I, Number of CLL cells in the spleen after exposure to control vehicle, or the combination of PI3K-δ and PI3K-γ inhibitors, or PI3K-δ inhibitor alone using samples from 6 patients (left of dotted line). Results from another experiment using the same set of patient samples exposed to vehicle control, PI3K-δ inhibitor alone, or PI3K-γ inhibitor alone (right of the dotted line). *, P < 0.05; **, P ≤ 0.01.

Figure 4.

Duvelisib treatment alters the TME that supports CLL-cell growth. A, Schematic of xenograft model using primary CLL cells. B–E, Quantification of human CLL cells and T cells and murine myeloid cells by flow cytometry in mice receiving cells from patients with CLL. A total of 3–5 mice per patient per treatment condition (control vehicle, 70 or 100 mg/kg duvelisib). B, Number of CLL B cells in the spleens after duvelisib treatment. C, Number of proliferating CLL B cells in the cell cycle after duvelisib treatment. D, Number of T cells in mice treated with duvelisib. E, Number of murine GRlowLY-6ClowF4/80+CD11b+CD11c+CD206+ macrophages in recipient mouse spleens. *, P < 0.05; **, P ≤ 0.01. F–I, Quantification of human CLL cells by flow cytometry in mice receiving cells from patients with CLL. A total of 3–5 mice per patient per treatment condition. Numbers of CLL cells (F), CLL-derived T cells (G), and murine macrophages (H) in the spleens of mice treated 5 days each week for 3 weeks with PI3K-γi (15 mg/kg, IPI-5243), PI3K-δi (10 mg/kg; IPI-3063), or duvelisib (100 mg/kg, Duv). I, Number of CLL cells in the spleen after exposure to control vehicle, or the combination of PI3K-δ and PI3K-γ inhibitors, or PI3K-δ inhibitor alone using samples from 6 patients (left of dotted line). Results from another experiment using the same set of patient samples exposed to vehicle control, PI3K-δ inhibitor alone, or PI3K-γ inhibitor alone (right of the dotted line). *, P < 0.05; **, P ≤ 0.01.

Close modal

Next, we examined the distinct actions of PI3K-γi or PI3K-δi, or duvelisib on CLL-cell engraftment and growth in the xenograft model mentioned above by treating recipient mice once a day for 3 weeks (Fig. 4FH). Both PI3K-δi and duvelisib significantly decreased the number of leukemic B cells in murine spleens from patients with CLL as a group (Fig. 4F) or when divided into the two prognostically distinct subgroups, U-CLL and M-CLL (Supplementary Fig. S3C–S3E).

We next analyzed the growth of CLL-derived T cells in spleens from CLL-bearing mice treated as illustrated in Fig. 4A. Three-week treatment with PI3K-γi or PI3K-δi, or duvelisib substantially decreased the numbers of spleen-residing T cells in mice treated with PI3K-γi or duvelisib, but not PI3K-δi (Fig. 4G). This was the case for all samples, regardless of IGHV-mutation status (Supplementary Fig. S3C–S3E).

As CLL patient–derived myeloid cells do not grow in NSG mice, so one could not test this population in the model (18), we examined the distinct functions of PI3K-δ and PI3K-γ on murine splenic macrophage levels in the xenograft model. Notably, in another xenograft system that employs an immortalized CLL cell line, murine macrophages have an essential role on the growth of CLL cells (31). Consistent with the in vitro results, duvelisib and PI3K-γi monotherapy significantly reduced the number of murine M2 macrophages in the spleens after 3 weeks of treatment; also consistent with our in vitro findings PI3K-δi did not lead to a significant change (Fig. 4H). Finally, serum levels of human and mouse cytokines or chemokines involved in T-cell–mediated and myeloid cell–mediated immunity were measured by ELISA. None of the human cytokines/chemokines measured showed significant differences between the treatment groups (Supplementary Fig. S4). Only the murine CXCL5 (LIX) cytokine/chemokine was significantly decreased in PI3K-γi but not PI3K-δi treated mice (Supplementary Fig. S5). CXCL5 (LIX) binds CXCR2 and recruits immune cells to promote cancer cell survival, and CXCL5 can be overexpressed in patients with CLL (32). Thus, decreasing CXCL5 levels could enhance the therapeutic effects of PI3K-γi in CLL.

Finally, because of the differences in IC50 for duvelisib compared with the individual isoform-specific agents (Supplementary Table S1), we used the same patient's cells to test the effects of combination of the PI3K-γi–specific and PI3K-δi–specific agents. In vivo treatment with the combination of PI3K-δ and PI3K-γ inhibitors led to a significant additional decrease in CLL B cells in the spleen compared with the PI3K-δi agent alone (Fig. 4I, left). While this suggests the additive effect of PI3K-γi, when we repeated the experiment using the same six CLL patient samples, PI3K-γi alone was not sufficient to decrease the numbers of CLL cells in recipient spleens (Fig. 4I, right). Overall, these in vivo results support the advantage of dual PI3K-δ,γ inhibition compared with PI3K-δi treatment alone.

Case study of a patient with CLL salvaged with duvelisib after progression on ibrutinib

Intolerance and resistance to ibrutinib are emerging therapeutic problems, with the former developing in 50% to 63% of patients (33) and the latter in 13% to 30% (13, 34, 35). Ibrutinib resistance can be due to the selection for intraclonal variants harboring BTK (C481S) or PLCG2 mutations (13).

In this context, we report a patient (CLL1570), who had failed treatment with FCR (fludarabine, cyclophosphamide, rituximab), tonsillar radiation, and BR (bendamustine and rituximab), before achieving a partial response to ibrutinib. Six years later, the patient became refractory to ibrutinib, and near the time of ibrutinib relapse, the leukemic clone was found to harbor BTK C481S and PLCG2 L845F mutations, along with mutations in BRCA2 (truncation in intron 26); MAP2K1 (MEK1; Q56P, subclonal); SF3B1 (K700E); TP53 (F113C and L330R, splice site 559+1G>A). Immediately after discontinuing ibrutinib (day 0; Fig. 5), the patient was started on duvelisib (25 mg orally, twice daily). Therapy was continued for >266 days, with only two short (13 and 3 day) treatment holds; the first, at day 50, due to grade 2 elevations of alanine aminotransferase/aspartate aminotransferase and low-level cytomegalovirus reactivation, and the second, at day 179, because of a temporary elevation of serum creatinine.

Figure 5.

Analyses of a patient with CLL whose leukemic clone exhibited multiple genomic mutations who failed ibrutinib and several other therapies but achieved a partial response on duvelisib. A, Lymphocyte, platelet, and hemoglobin counts for patient CLL1570 after receiving duvelisib (day 0). Bars indicate times when duvelisib was temporarily held. Patient received once-daily, oral ibrutinib (420 mg) for 1,344 days. Upon relapse, treatment was changed to twice-daily oral duvelisib (250 mg) for approximately 400 days. B, Characterization of CD19+ CLL cells, CD3+ T cells including CD4+Th1 and CD4+Th2 cells, and M1- and M2-macrophages in BM, before and over time after initiation of duvelisib treatment. C, Characterization of CD19+ CLL cells, CD3+ T cells including CD4+Th1, CD4+Th2 cells, CD4+Th17 cells in blood, before and over time after initiation of duvelisib treatment.

Figure 5.

Analyses of a patient with CLL whose leukemic clone exhibited multiple genomic mutations who failed ibrutinib and several other therapies but achieved a partial response on duvelisib. A, Lymphocyte, platelet, and hemoglobin counts for patient CLL1570 after receiving duvelisib (day 0). Bars indicate times when duvelisib was temporarily held. Patient received once-daily, oral ibrutinib (420 mg) for 1,344 days. Upon relapse, treatment was changed to twice-daily oral duvelisib (250 mg) for approximately 400 days. B, Characterization of CD19+ CLL cells, CD3+ T cells including CD4+Th1 and CD4+Th2 cells, and M1- and M2-macrophages in BM, before and over time after initiation of duvelisib treatment. C, Characterization of CD19+ CLL cells, CD3+ T cells including CD4+Th1, CD4+Th2 cells, CD4+Th17 cells in blood, before and over time after initiation of duvelisib treatment.

Close modal

At day 27 on duvelisib therapy, a favorable response, designated as a partial response according to IWCLL criteria (19), was achieved. The thrombocytopenia and elevated absolute lymphocyte count that had developed at the time of ibrutinib resistance normalized on duvelisib treatment, with absolute values of lymphocytes decreasing by ≥50% on days 62–76, 168–236, 266, and beyond (Fig. 5A). Because the patient rapidly developed a lymphocytosis (>5 × 109 cells/L), the initial rise in circulating CLL cells presumably developed because leukemic B cells exited tissue niches, including the BM which was the site of the most dramatic changes upon disease relapse. CLL cells in the BM, which had reached 60% before initiating duvelisib, fell to 10% after 7 months on treatment (Fig. 5B). In addition, other immune populations in the BM involved in leukemic B-cell control were altered by duvelisib treatment. Specifically, duvelisib reduced M2-like macrophages (5.9%–0.2%), increased M1-like macrophages (0.4%–1.2%) and led to changes in the ratio of IFNγ-producing Th1 and IL4-producing Th2 T cells (Th1/2: 3.6–11), compared with baseline values.

Seven months after initiating duvelisib treatment, CLL cells in the blood fell from 74% to 38% although no differences in Th1/Th2 T-cell subset ratios were found (Th1/2: 1.86–2.28). There was, however, a small increase in Th17A-producing T cells (0.03%–1%) and a reduction in regulatory T cells (Treg; 0.17%–0.03%; Fig. 5C).

CLL B cells from patients relapsing with ibrutinib are efficiently eliminated by duvelisib in a patient-derived xenograft model

Because we have documented that our xenograft model can recapitulate the response to ibrutinib in both sensitive and resistant patients (36), we then attempted to model in vivo the beneficial effects of duvelisib in this patient and in two others who also developed ibrutinib-resistant disease using the xenograft approach. Like CLL1570 above, leukemic B cells from patient CLL1273 had at least a C481S BTK mutation; in contrast, neither BTK nor PLCG2 mutations were found in the leukemic cells from patient CLL1782.

NSG mice, engrafted with leukemic B and T cells from each patient, were treated with ibrutinib or duvelisib daily for 3 weeks. As expected from the patients’ clinical histories, in vivo ibrutinib treatment of recipient mice bearing ibrutinib-resistant clones did not significantly change CLL-cell numbers in the spleens (Fig. 6A and B). However, duvelisib treatment of mice engrafted with B and T cells from each ibrutinib-resistant patient led to a significant reduction in leukemic B-cell numbers in the spleen. This was the case regardless of BTK mutation status. In addition, duvelisib, but not ibrutinib, significantly reduced the numbers of Ki67+ CLL B cells (Fig. 6C) and murine M2 macrophages (Fig. 6D) in recipient spleens. Neither duvelisib nor ibrutinib treatment altered the numbers of human T, Th1, or Th2 cells in recipient animals (Supplementary Fig. S6). To our knowledge, this is the first in vivo demonstration of ibrutinib-refractory clones from patients with CLL being successfully treated with duvelisib in clinical practice and a xenograft model.

Figure 6.

CLL B cells from patients unresponsive to ibrutinib are efficiently eliminated by duvelisib in a murine xenograft model. A, Representative flow cytometry tracings showing primary CLL cells and T cells from mice that received treatment with duvelisib (Duv, 100 mg/kg), duvelisib vehicle control (Duv-vehicle), ibrutinib (IBR, 25 mg/kg), or ibrutinib vehicle control (IBR-vehicle). B, Number of CLL cells in spleens obtained from mice engrafted with ibrutinib-resistant samples carrying wild type (CLL1782) or at least a C481S mutation in BTK (CLL1273 and CLL1570). C, Measurement of leukemic B-cell proliferation (Ki67+) from patients with CLL whose clones exhibited wildtype or mutant BTK. D, Effects of duvelisib treatment on the number of endogenous murine macrophages in the spleens of mice engrafted with samples from patients with ibrutinib-resistant CLL. *, P < 0.05; **, P ≤ 0.01.

Figure 6.

CLL B cells from patients unresponsive to ibrutinib are efficiently eliminated by duvelisib in a murine xenograft model. A, Representative flow cytometry tracings showing primary CLL cells and T cells from mice that received treatment with duvelisib (Duv, 100 mg/kg), duvelisib vehicle control (Duv-vehicle), ibrutinib (IBR, 25 mg/kg), or ibrutinib vehicle control (IBR-vehicle). B, Number of CLL cells in spleens obtained from mice engrafted with ibrutinib-resistant samples carrying wild type (CLL1782) or at least a C481S mutation in BTK (CLL1273 and CLL1570). C, Measurement of leukemic B-cell proliferation (Ki67+) from patients with CLL whose clones exhibited wildtype or mutant BTK. D, Effects of duvelisib treatment on the number of endogenous murine macrophages in the spleens of mice engrafted with samples from patients with ibrutinib-resistant CLL. *, P < 0.05; **, P ≤ 0.01.

Close modal

PI3Kis target BCR signaling, which is a key pathogenetic pathway in several B-cell lymphoid malignancies. However, in some clinical studies, especially those where idelalisib was given to treatment-naïve patients (37), PI3Ki leads to the emergence of autoimmune hepatitis, pneumonitis, and colitis (38). Thus, new generations of PI3Kis that can achieve therapeutic effects with less toxicity are needed. Dual inhibition of PI3K-δ and PI3K-γ by duvelisib leads to slightly better or similar overall response rate in CLL (39).

Here, we studied, in vitro and in vivo using xenografts, the different actions of PI3K-δ and PI3K-γ in CLL, providing evidence for enhanced efficacy with the addition of PI3K-γ inhibition that blocks recruitment of T cells and macrophages which can support the survival and growth of leukemic B cells. We further show a possible, promising therapeutic advantage of duvelisib in patients with CLL, including those who developed resistance to ibrutinib.

Previous in vitro studies using B-cell line cells or primary CLL patient cells have shown duvelisib-induced CLL-cell cytotoxicity at concentrations from 0.25 to 5 μmol/L; this range corresponds to plasma concentrations achieved in patients who received duvelisib 25 mg twice daily. In our in vitro studies with primary CLL patient B cells, nonlethal doses (0.1–1 μmol/L) of duvelisib and PI3K-δi (IPI-3063) but not PI3K-γi (IPI-549) reduced leukemic B-cell proliferation and survival in vitro (Fig. 1). In vivo, CLL cells pretreated with duvelisib or PI3K-δi, but not PI3K-γi, had a diminished ability to home to the spleen of recipient mice (Fig. 1C). Thus, both PI3K-γ inhibitors used in these studies showed similar effects; and the effects of PI3K-δi (IPI-3063) are consistent with the previous reported results from other PI3K-δi including idelalisib (40). These suggest the on-target inhibition of the PI3Kis used in this study.

Our data support the distinct roles of PI3K-δ and PI3K-γ in different immune cells previously reported. Interestingly, in vitro, we found no synergistic function of PI3K-δi and PI3K-γi for the growth of CLL cells (Supplementary Fig. S1), or the migration of both CLL and T cells (Supplementary Figs. S1 and S2). In vivo, when these drugs were applied in our xenograft model, a model that is highly dependent on T cells for B-cell growth, treatment of PI3K-γi alone 2 weeks after tumor cell injection inhibits the number of T cells, but not to the level to have an impact on CLL cells (Fig. 4F and I). Treatment of PI3K-δi alone is sufficient to reduce the number of CLL cell, importantly, there is an addictive effect with the combination of PI3K-δi + PI3K-γi (Fig. 4I), highlighting the important contribution of PI3K-γi altered TME in CLL cell growth in the tissue niches.

PI3K inhibition by idelalisib (40) and duvelisib (41) can be associated with autoimmune toxicity manifested by T-cell inflammatory phenotypes. In the patients with treatment-naïve CLL treated with duvelisib and FCR, colitis and hepatitis were associated with increased CD8+ T cells, Tregs, and Th17 T cells (41). Here, after in vitro stimulation with anti-CD3/28 beads and IL2, we did not find a reduction in activated T-cell numbers nor Th1/Th2 cytokine production with PI3K-δi nor PI3K-γi. In addition, while our patient-derived xenograft (PDX) model is not ideal for the engraftment of Tregs and Th17 T cells, in the ibrutinib-resistant CLL patient 1570 that we studied, we observed a slight increase in Th17A-producing T cells and a small change in Tregs, as well as more sizable changes in Th1/Th2 T cells (Fig. 5B and C). Interestingly, duvelisib also did no change T-cell counts in the primary ibrutinib-resistant CLL patient 1570 or the PDX model of all 3 patients with ibrutinib-resistant CLL (Supplementary Fig. S6). This result is different than the data obtained from nonresistant patients shown in Fig. 4D, suggesting the altered T-cell immunity in patients with ibrutinib-resistant CLL.

Finally, macrophages, which can dampen immune responses by producing cytokines that suppress cytotoxic T-cell activation and function, were altered by PI3K-γi, and not PI3K-δι. For example, the PI3K-γ inhibitor, IPI-549 (Supplementary Table S1), can reduce the immune suppressive function of M2 macrophages (9, 42), whereas the PI3K-δi idelalisib has minimal impact on macrophage functions in vitro (43). Consistent with this, when we assessed the role of PI3K-δ and PI3K-γ in the endogenous macrophages of NSG mice in vitro and in vivo, duvelisib decreased polarization of murine macrophages to an M2 phenotype and diminished the support that such macrophages provided for CLL survival. These findings are consistent with previous observations in CLL (30) and T-cell lymphoma (44). In addition, because T-cell subset distribution was changed in the TCL-1 murine model of CLL (30) after targeted reduction in macrophages, the decreased T-cell growth that we observed in the xenograft experiments might be influenced by the altered macrophage subsets emerging after PI3K-γi and duvelisib treatment. We similarly found that M2 macrophages were reduced in a patient with ibrutinib resistance (CLL1570) upon initiation of duvelisib treatment, consistent with loss of support for CLL cell survival and loss of a TME that activates effector T cells (45). Thus, the reduced numbers of CLL cells after duvelisib treatment could be a direct effect of PI3K-δi or an indirect one, achieved by altering the supporting effects of T cells and macrophages via PI3K-γi.

The effect of PI3K-γ inhibition (through genetic means or using selective PI3K-γ inhibitors like IPI-549) on the TME in solid tumors has been studied previously in preclinical cancer models in mice (9). Consistent with the effects observed in the CLL xenograft model, PI3K-γ inhibition results in a shift from protumor M2-like macrophages to antitumor M1-like macrophages (9). However, in contrast to the effects on T-cell migration observed in this CLL xenograft model, loss of PI3K-γ in solid tumor models shows enhanced T-cell recruitment and proliferation, particularly within the cytotoxic CD8+ subpopulation of T cells (9). The effects on T cells in this context are thought to be the mediated a reduction in the immune suppressive function of myeloid cells, rather than a direct effect on T-cell proliferation or activation (15). This underscores the importance of context and may reveal differences between the TME of hematologic malignancies and solid tumors.

Our demonstration that duvelisib inhibited the growth of CLL B cells from patients whose disease had progressed during ibrutinib therapy, regardless of the detection of BTK and/or PLCG2 mutations, in NSG mice could have substantial clinical relevance. Consistent with this, patient CLL1570, who had progressed on ibrutinib, achieved a partial clinical response on duvelisib. Conspicuously, because in this patient, the majority of ibrutinib-resistant cells resided in the BM, the duvelisib-induced redistribution lymphocytosis likely affected these cells, potentially by inhibiting the adhesion of CLL cells to BM mesenchymal stromal cells which requires the action of PI3K-δ  (46, 47). Thus, this patient's beneficial response to duvelisib supports the potential value to dual BTK and PI3K inhibition. Finally, the patient's response to duvelisib was manifested by an alteration of the myeloid-cell phenotype (M2 to M1) that could have abrogated support for CLL cell growth.

Overall, our study provides a mechanistic understanding of the effects of inhibition of distinct PI3K isoforms as well as an experimental rationale for using duvelisib as a treatment for patients with CLL whom are refractory to BTKi. In this regard, although duvelisib was FDA approved on the basis of the phase III DUO trial, patients who were previously exposed to ibrutinib or any other BTKis were excluded from that trial (48). Hence, data on post-BTKis are limited. Although a larger number of patients will need to be studied to determine the scope of the clinical relevance of our initial findings, our studies suggest that a clinical trial to address this possibility warrants consideration.

J.L. Kutok reports personal fees and other support from Infinity Pharmaceuticals during the conduct of the study. D.T. Weaver reports personal fees from Secura Bio and Yingli Pharma outside the submitted work. J.A. Pachter reports other support from Verastem outside the submitted work. No disclosures were reported by the other authors.

S.-S. Chen: Conceptualization, resources, data curation, formal analysis, validation, investigation, visualization, methodology, writing–original draft, writing–review and editing. J.C. Barrientos: Conceptualization, resources, data curation, formal analysis, validation, investigation, methodology, writing–original draft. G. Ferrer: Methodology. M. King-Richards: Validation, investigation, methodology. Y.-J. Chen: Validation, investigation, methodology. P. Ravichandran: Data curation, formal analysis. M. Ibrahim: Data curation, formal analysis. Y. Kieso: Resources. S. Waters: Writing–review and editing. J.L. Kutok: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. M. Peluso: Formal analysis, validation. S. Sharma: Formal analysis. D.T. Weaver: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration. J.A. Pachter: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. K.R. Rai: Supervision. N. Chiorazzi: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing.

We thank the patients who donated blood for these studies. We also acknowledge The Feinstein Institutes Core facilities for technical and analytical support. This study was funded in part through a Sponsored Research Agreement between the Feinstein Institutes with Infinity LLC and Verastem LLC and supported by NIH NCI grant NIH/NCI R01 CA238523 (to N. Chiorazzi and S.-S. Chen); CLL Global Research Foundation (to N. Chiorazzi); The Nash Family Foundation (to N. Chiorazzi and K.R. Rai); The Marks Foundation (to N. Chiorazzi and K.R. Rai); The Karches Family (to N. Chiorazzi and K.R. Rai); The Jean Walton Fund for Leukemia, Lymphoma, & Myeloma Research (to N. Chiorazzi and S.-S. Chen); Swim Across America (to S.-S. Chen); AWSM awards from the Feinstein Institutes for Medical Research (to S.-S. Chen); Swedish Cancer Society (to K.R. Rai); Alice Wallenberg Foundation (to K.R. Rai); Karolinska Institute (to K.R. Rai); and Cancer Research Foundations of Radiumhemmet (to K.R. Rai).

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).

1.
Kipps
TJ
,
Stevenson
FK
,
Wu
CJ
,
Croce
CM
,
Packham
G
,
Wierda
WG
, et al
.
Chronic lymphocytic leukaemia
.
Nat Rev Dis Primers
2017
;
3
:
16096
.
2.
Fabbri
G
,
Dalla-Favera
R
.
The molecular pathogenesis of chronic lymphocytic leukaemia
.
Nat Rev Cancer
2016
;
16
:
145
62
.
3.
Patel
K
,
Pagel
JM
.
Exploring a future for PI3K inhibitors in chronic lymphocytic leukemia
.
Curr Hematol Malig Rep
2019
;
14
:
292
301
.
4.
Frustaci
AM
,
Tedeschi
A
,
Deodato
M
,
Zamprogna
G
,
Cairoli
R
,
Montillo
M
.
Duvelisib for the treatment of chronic lymphocytic leukemia
.
Expert Opin Pharmacother
2020
;
21
:
1299
309
.
5.
Winkler
DG
,
Faia
KL
,
DiNitto
JP
,
Ali
JA
,
White
KF
,
Brophy
EE
, et al
.
PI3K-delta and PI3K-gamma inhibition by IPI-145 abrogates immune responses and suppresses activity in autoimmune and inflammatory disease models
.
Chem Biol
2013
;
20
:
1364
74
.
6.
Lampson
BL
,
Brown
JR
.
PI3Kdelta-selective and PI3Kalpha/delta-combinatorial inhibitors in clinical development for B-cell non-Hodgkin lymphoma
.
Expert Opin Investig Drugs
2017
;
26
:
1267
79
.
7.
Hoellenriegel
J
,
Meadows
SA
,
Sivina
M
,
Wierda
WG
,
Kantarjian
H
,
Keating
MJ
, et al
.
The phosphoinositide 3'-kinase delta inhibitor, CAL-101, inhibits B-cell receptor signaling and chemokine networks in chronic lymphocytic leukemia
.
Blood
2011
;
118
:
3603
12
.
8.
Faehling
S
,
Coelho
M
,
Floerchinger
A
,
Schneider
C
,
Stilgenbauer
S
,
Lichter
P
, et al
.
Simultaneous inhibition of PI3Kgamma and PI3Kdelta deteriorates T-cell function with implications for chronic lymphocytic leukemia
.
Hemasphere
2023
;
7
:
e840
.
9.
Kaneda
MM
,
Messer
KS
,
Ralainirina
N
,
Li
H
,
Leem
CJ
,
Gorjestani
S
, et al
.
PI3Kgamma is a molecular switch that controls immune suppression
.
Nature
2016
;
539
:
437
42
.
10.
Flinn
IW
,
O'Brien
S
,
Kahl
B
,
Patel
M
,
Oki
Y
,
Foss
FF
, et al
.
Duvelisib, a novel oral dual inhibitor of PI3K-delta, gamma, is clinically active in advanced hematologic malignancies
.
Blood
2018
;
131
:
877
87
.
11.
Balakrishnan
K
,
Peluso
M
,
Fu
M
,
Rosin
NY
,
Burger
JA
,
Wierda
WG
, et al
.
The phosphoinositide-3-kinase (PI3K)-delta and gamma inhibitor, IPI-145 (Duvelisib), overcomes signals from the PI3K/AKT/S6 pathway and promotes apoptosis in CLL
.
Leukemia
2015
;
29
:
1811
22
.
12.
Gango
A
,
Alpar
D
,
Galik
B
,
Marosvari
D
,
Kiss
R
,
Fesus
V
, et al
.
Dissection of subclonal evolution by temporal mutation profiling in chronic lymphocytic leukemia patients treated with ibrutinib
.
Int J Cancer
2020
;
146
:
85
93
.
13.
Woyach
JA
,
Furman
RR
,
Liu
TM
,
Ozer
HG
,
Zapatka
M
,
Ruppert
AS
, et al
.
Resistance mechanisms for the Bruton's tyrosine kinase inhibitor ibrutinib
.
N Engl J Med
2014
;
370
:
2286
94
.
14.
Zhang
SQ
,
Smith
SM
,
Zhang
SY
,
Lynn Wang
Y
.
Mechanisms of ibrutinib resistance in chronic lymphocytic leukaemia and non-Hodgkin lymphoma
.
Br J Haematol
2015
;
170
:
445
56
.
15.
Lampson
BL
,
Brown
JR
.
Are BTK and PLCG2 mutations necessary and sufficient for ibrutinib resistance in chronic lymphocytic leukemia
.
Expert Rev Hematol
2018
;
11
:
185
94
.
16.
Hampel
PJ
,
Ding
W
,
Call
TG
,
Rabe
KG
,
Kenderian
SS
,
Witzig
TE
, et al
.
Rapid disease progression following discontinuation of ibrutinib in patients with chronic lymphocytic leukemia treated in routine clinical practice
.
Leuk Lymphoma
2019
;
60
:
2712
9
.
17.
Dong
S
,
Guinn
D
,
Dubovsky
JA
,
Zhong
Y
,
Lehman
A
,
Kutok
J
, et al
.
IPI-145 antagonizes intrinsic and extrinsic survival signals in chronic lymphocytic leukemia cells
.
Blood
2014
;
124
:
3583
6
.
18.
Patten
PEM
,
Ferrer
G
,
Chen
SS
,
Kolitz
JE
,
Rai
KR
,
Allen
SL
, et al
.
A detailed analysis of parameters supporting the engraftment and growth of chronic lymphocytic leukemia cells in immune-deficient mice
.
Front Immunol
2021
;
12
:
627020
.
19.
Hallek
M
,
Cheson
BD
,
Catovsky
D
,
Caligaris-Cappio
F
,
Dighiero
G
,
Dohner
H
, et al
.
Guidelines for the diagnosis and treatment of chronic lymphocytic leukemia: a report from the international workshop on chronic lymphocytic leukemia updating the national cancer institute-working group 1996 guidelines
.
Blood
2008
;
111
:
5446
56
.
20.
Byrd
JC
,
Brown
JR
,
O'Brien
S
,
Barrientos
JC
,
Kay
NE
,
Reddy
NM
, et al
.
Ibrutinib versus ofatumumab in previously treated chronic lymphoid leukemia
.
N Engl J Med
2014
;
371
:
213
23
.
21.
Munir
T
,
Brown
JR
,
O'Brien
S
,
Barrientos
JC
,
Barr
PM
,
Reddy
NM
, et al
.
Final analysis from RESONATE: up to six years of follow-up on ibrutinib in patients with previously treated chronic lymphocytic leukemia or small lymphocytic lymphoma
.
Am J Hematol
2019
;
94
:
1353
63
.
22.
Mongini
PK
,
Gupta
R
,
Boyle
E
,
Nieto
J
,
Lee
H
,
Stein
J
, et al
.
TLR-9 and IL-15 synergy promotes the in vitro clonal expansion of chronic lymphocytic leukemia B cells
.
J Immunol
2015
;
195
:
901
23
.
23.
Ying
W
,
Cheruku
PS
,
Bazer
FW
,
Safe
SH
,
Zhou
B
.
Investigation of macrophage polarization using bone marrow derived macrophages
.
J Vis Exp
2013
;
23
:
50323
.
24.
McLeod
RL
,
Gil
MA
,
Chen
D
,
Cabal
A
,
Katz
J
,
Methot
J
, et al
.
Characterizing pharmacokinetic-pharmacodynamic relationships and efficacy of PI3Kdelta inhibitors in respiratory models of TH2 and TH1 inflammation
.
J Pharmacol Exp Ther
2019
;
369
:
223
33
.
25.
De Palma
M
,
Lewis
CE
.
Macrophage regulation of tumor responses to anticancer therapies
.
Cancer Cell
2013
;
23
:
277
86
.
26.
Arlauckas
SP
,
Garren
SB
,
Garris
CS
,
Kohler
RH
,
Oh
J
,
Pittet
MJ
, et al
.
Arg1 expression defines immunosuppressive subsets of tumor-associated macrophages
.
Theranostics
2018
;
8
:
5842
54
.
27.
Ono
Y
,
Nagai
M
,
Yoshino
O
,
Koga
K
,
Nawaz
A
,
Hatta
H
, et al
.
CD11c+ M1-like macrophages (MPhis) but not CD206+ M2-like MPhi are involved in folliculogenesis in mice ovary
.
Sci Rep
2018
;
8
:
8171
.
28.
Ten Hacken
E
,
Burger
JA
.
Microenvironment interactions and B-cell receptor signaling in chronic lymphocytic leukemia: implications for disease pathogenesis and treatment
.
Biochim Biophys Acta
2016
;
1863
:
401
13
.
29.
van Attekum
MHA
,
Terpstra
S
,
Slinger
E
,
von Lindern
M
,
Moerland
PD
,
Jongejan
A
, et al
.
Macrophages confer survival signals via CCR1-dependent translational MCL-1 induction in chronic lymphocytic leukemia
.
Oncogene
2017
;
36
:
3651
60
.
30.
Galletti
G
,
Scielzo
C
,
Barbaglio
F
,
Rodriguez
TV
,
Riba
M
,
Lazarevic
D
, et al
.
Targeting macrophages sensitizes chronic lymphocytic leukemia to apoptosis and inhibits disease progression
.
Cell Rep
2016
;
14
:
1748
60
.
31.
Hanna
BS
,
Ozturk
S
,
Seiffert
M
.
Beyond bystanders: myeloid cells in chronic lymphocytic leukemia
.
Mol Immunol
2019
;
110
:
77
87
.
32.
Mahadevan
D
,
Choi
J
,
Cooke
L
,
Simons
B
,
Riley
C
,
Klinkhammer
T
, et al
.
Gene expression and serum cytokine profiling of low stage CLL identify WNT/PCP, Flt-3L/Flt-3 and CXCL9/CXCR3 as regulators of cell proliferation, survival and migration
.
Hum Genomics Proteomics
2009
;
2009
:
453634
.
33.
Mato
AR
,
Thompson
M
,
Allan
JN
,
Brander
DM
,
Pagel
JM
,
Ujjani
CS
, et al
.
Real-world outcomes and management strategies for venetoclax-treated chronic lymphocytic leukemia patients in the United States
.
Haematologica
2018
;
103
:
1511
7
.
34.
Byrd
JC
,
Furman
RR
,
Coutre
SE
,
Flinn
IW
,
Burger
JA
,
Blum
KA
, et al
.
Targeting BTK with ibrutinib in relapsed chronic lymphocytic leukemia
.
N Engl J Med
2013
;
369
:
32
42
.
35.
Furman
RR
,
Cheng
S
,
Lu
P
,
Setty
M
,
Perez
AR
,
Guo
A
, et al
.
Ibrutinib resistance in chronic lymphocytic leukemia
.
N Eng J Med
2014
;
370
:
2352
4
.
36.
Chen
S-S
,
Tam
CS
,
Ramsay
AG
,
Ravichandran
P
,
Couto-Francisco
NC
,
Ibrahim
M
, et al
.
CLL B cells develop resistance to ibrutinib by reinvigorating the IL-4R - IL-4 axis blocked by bruton's tyrosine kinase inhibitors including acalabrutinib and zanubrutinib
.
Blood
2019
;
134
:
477
.
37.
Wagner-Johnston
ND
,
Sharman
J
,
Furman
RR
,
Salles
G
,
Brown
JR
,
Robak
T
, et al
.
Idelalisib immune-related toxicity is associated with improved treatment response
.
Leuk Lymphoma
2021
;
62
:
2915
20
.
38.
Furman
RR
,
Sharman
JP
,
Coutre
SE
,
Cheson
BD
,
Pagel
JM
,
Hillmen
P
, et al
.
Idelalisib and rituximab in relapsed chronic lymphocytic leukemia
.
N Eng J Med
2014
;
370
:
997
1007
.
39.
O'Brien
S
,
Patel
M
,
Kahl
BS
,
Horwitz
SM
,
Foss
FM
,
Porcu
P
, et al
.
Duvelisib, an oral dual PI3K-delta, gamma inhibitor, shows clinical and pharmacodynamic activity in chronic lymphocytic leukemia and small lymphocytic lymphoma in a phase 1 study
.
Am J Hematol
2018
;
93
:
1318
26
.
40.
Dong
S
,
Harrington
BK
,
Hu
EY
,
Greene
JT
,
Lehman
AM
,
Tran
M
, et al
.
PI3K p110δ inactivation antagonizes chronic lymphocytic leukemia and reverses T cell immune suppression
.
J Clin Invest
2019
;
129
:
122
36
.
41.
Gadi
D
,
Griffith
A
,
Tyekucheva
S
,
Wang
Z
,
Rai
V
,
Vartanov
A
, et al
.
A T cell inflammatory phenotype is associated with autoimmune toxicity of the PI3K inhibitor duvelisib in chronic lymphocytic leukemia
.
Leukemia
2022
;
36
:
723
32
.
42.
De Henau
O
,
Rausch
M
,
Winkler
D
,
Campesato
LF
,
Liu
C
,
Cymerman
DH
, et al
.
Overcoming resistance to checkpoint blockade therapy by targeting PI3Kgamma in myeloid cells
.
Nature
2016
;
539
:
443
7
.
43.
Palazzo
A
,
Herter
S
,
Grosmaire
L
,
Jones
R
,
Frey
CR
,
Limani
F
, et al
.
The PI3Kdelta-selective inhibitor idelalisib minimally interferes with immune effector function mediated by rituximab or obinutuzumab and significantly augments B cell depletion in vivo
.
J Immunol
2018
;
200
:
2304
12
.
44.
Horwitz
SM
,
Koch
R
,
Porcu
P
,
Oki
Y
,
Moskowitz
A
,
Perez
M
, et al
.
Activity of the PI3K-delta, gamma inhibitor duvelisib in a phase 1 trial and preclinical models of T-cell lymphoma
.
Blood
2018
;
131
:
888
98
.
45.
Sica
A
,
Mantovani
A
.
Macrophage plasticity and polarization: in vivo veritas
.
J Clin Invest
2012
;
122
:
787
95
.
46.
Ali
AY
,
Guan
Q
,
Wu
X
,
Hou
S
,
Banerji
V
,
Johnston
JB
, et al
.
Expression and function of phosphoinositide 3-kinase delta in mesenchymal stromal cells from normal and leukaemic bone marrow
.
Br J Haematol
2019
;
185
:
883
7
.
47.
Bilancio
A
,
Okkenhaug
K
,
Camps
M
,
Emery
JL
,
Ruckle
T
,
Rommel
C
, et al
.
Key role of the p110delta isoform of PI3K in B-cell antigen and IL-4 receptor signaling: comparative analysis of genetic and pharmacologic interference with p110delta function in B cells
.
Blood
2006
;
107
:
642
50
.
48.
Flinn
IW
,
Hillmen
P
,
Montillo
M
,
Nagy
Z
,
Illes
A
,
Etienne
G
, et al
.
The phase 3 DUO trial: duvelisib vs ofatumumab in relapsed and refractory CLL/SLL
.
Blood
2018
;
132
:
2446
55
.