IPH2101, a novel anti-inhibitory KIR antibody, and lenalidomide combine to enhance the natural killer cell versus multiple myeloma effect

Natural killer (NK) cells exert cytotoxicity against multiple myeloma (MM), and some therapies for MM appear to recover or enhance NK-cell function against MM.^1-5  Lenalidomide in particular confers NK-cell expansion and activation associated with tumor cell apoptosis.^4,5  MM cells up-regulate the expression of ligands to NK cell–inhibitory killer cell immunoglobulin–like receptor (KIR)⁶  and KIR-ligand mismatch in T cell–depleted, allogeneic stem cell transplantation may reduce the risk of relapse in MM patients, suggesting that this signaling axis may be particularly important.⁷ 

IPH2101 is a human IgG4 mAb against common inhibitory KIR2DL-1, KIR2DL-2, and KIR2DL-3.⁸  IPH2101 enhances NK-cell function against malignant cells by preventing inhibitory KIR-ligand interaction and subsequent inhibitory signaling.⁸ 

In the present study, we provide novel data characterizing mechanisms by which inhibitory KIR blockade augments NK-cell function against MM, sparing normal cells. In addition, we uncover novel immunomodulatory properties of lenalidomide that likely contribute to enhanced NK-cell activity. We demonstrate that a murine anti-inhibitory NK-cell receptor Ab and lenalidomide further augment NK-cell function against MM compared with either agent alone, leading to in vivo rejection of a lenalidomide-resistant tumor. These data support the initiation of a steroid-sparing, phase 2 trial of IPH2101 and lenalidomide in MM.

Cells were cultured in RPMI 1640 medium (Invitrogen) supplemented with 10% FBS (ICN Biomedicals) at 37° in 5% CO₂. NK cells and PBMCs from healthy donors (American Red Cross, Columbus, OH, and Indiana Blood Center, Indianapolis, IN) and PBMCs and BM aspirates from patients with MM obtained per Institutional Review Board–approved protocols were prepared as described previously.⁹  The MM cell lines U266 and K562 were from the American Type Culture Collection. We were unable to procure sufficient patient blood volume to enrich for NK cells from MM patient donors; therefore, experiments using patient-derived NK cells were conducted in PBMCs at effector:target (E:T) ratios based on the proportion of CD56⁺, CD3⁻ NK cells in patient PBMCs determined by flow cytometry.

IPH2101 (and PE-labeled anti-IPH2101) were provided by Innate Pharma. Lenalidomide was from Toronto Research Chemicals and John C. Byrd (The Ohio State University, Columbus, OH). Flow Abs were from Beckman Coulter, BD Pharmingen, eBioscience, R&D Systems, and Miltenyi Biotec. NKG2D-blocking Ab was from BioLegend. Abs against TRAIL, DNAM-1, and HLA class I (and isotypes) were from BD Biosciences, and 7-amino-actinomycin D was from Sigma-Aldrich.

U266 cells were stained with 7-amino-actinomycin D and PE-Ab, incubated at 4°C for 15 minutes, and washed with MACS buffer. Ten thousand cells and QuantiBRITE PE beads (BD Biosciences) were collected with a FACSCalibur flow cytometer (BD Biosciences) and analyzed using FlowJo Version 7.6.1 software (TreeStar). The median PE-relative fluorescence intensity (RFI) measurements were converted to number of PE Abs bound per cell using the QuantiBRITE PE beads. Expression of activating ligands on CD138⁺ MM cells was assessed at baseline and after 24 hours in lenalidomide (5nM) and on U266 cells after 5 days in lenalidomide (5μm).

Immune complex formation between patient PBMCs and autologous CD138⁺ MM cells or CD138⁻ BM cells was examined using a 2-color flow cytometric technique described previously.⁹  Effector cells (IPH2101- or control-treated) were labeled with CFSE (Sigma-Aldrich) and target cells were labeled with Paul Karl Horan dye (PKH; Sigma-Aldrich). Data were collected using a FACSCalibur flow cytometer (BD Biosciences) and analyzed using CellQuest Version 4.0 software (BD Biosciences).

ELISPOT experiments were conducted using MultiScreen 96-well plates (Millipore) as described previously⁹  to measure NK-cell IFN-γ and granzyme B production.¹⁰  Spots were counted using an Immunospot Imaging Analyzer (Cellular Technology).

Cytotoxicity was analyzed in europium-release assays per the manufacturer's instructions (DELFIA EuTDA Cytotoxicity kit; PerkinElmer). Specific cytotoxicity was calculated as follows: % cytotoxicity = (experimental release − spontaneous release) × 100/(maximal release − spontaneous release). Effectors incubated with or without 30 μg/mL of IPH-2101 for 30 minutes at 4°C and 10μM lenalidomide (or DMSO) were cocultured with DELFIA BATDA–labeled targets. In blocking experiments, NK cells were first incubated with 80 μg/mL of NKG2D, 40 μg/mL of DNAM-1, and 50 μg/mL of TRAIL mAbs or 170 μg/mL of isotype for 30 minutes at 4°C.

Flow cytometric cytotoxicity assays were conducted as described previously.^9,11  Patient-derived PBMCs were incubated for 72 hours in control conditions or with lenalidomide and/or IPH2101. Data were collected on an LSRII flow cytometer (BD Biosciences) and analyzed with FlowJo Version 7.6.1 software.

The in vivo tumor cell rejection method was described previously.¹²  RMA (H-2^b) is resistant to lenalidomide-induced apoptosis. GFP⁺-RMA cells were generated by lentivirus transduction with TRIpdU3FE1aEGFP. Spleen cells stained with 5μM PKH26 (Sigma-Aldrich) and mixed with GFP⁺-RMA cells (ratio 1:1) were injected intravenously into C57BL/6J mice (Charles River Laboratories). Forty hours later, liver mononuclear cells were separated by Ficoll gradient 80%/40%. Survival of engrafted cell populations was analyzed by flow cytometry. Results are presented as ratio of GFP⁺-RMA to spleen cells.

Freshly isolated, whole BM aspirates from patients with MM (n = 5) were incubated in control conditions or with IPH2101 (30 μg/mL) and/or lenalidomide (5μM) for 24 hours. The percentage of NK cells identified as CD56⁺CD16⁺CD3⁻C107a⁺ in the lymphocyte gate by flow cytometry was determined for each condition.

Means and SDs from experiments with 2 conditions were analyzed with the Student t test. ANOVA was used to analyze experiments with multiple conditions with planned comparisons conducted using the method of Bonferroni. Statistical significance was determined by P < .05.

IPH2101 increased NK-cell cytotoxicity against U266 cells (Figure 1A all pairwise comparisons, P < .05) and patient-derived NK cells against primary MM targets (Figure 1B right panel, P = .008). Interestingly, blocking either inhibitory KIR on NK cells (with IPH2101) or blocking HLA class I on U266 MM tumor cells enhanced killing, but there was no additive effect when both inhibitory KIR and HLA class I were blocked (supplemental Figure 1, available on the Blood Web site; see the Supplemental Materials link at the top of the online article).

We then studied these effects in the autologous setting using patient-derived, effector NK cells against autologous MM tumor cell and normal cell targets. We verified IPH2101 binding on NK cells and HLA class I expression on MM tumor cells (supplemental Figure 2). By preventing inhibitory KIR-ligand interaction, IPH2101 enhanced immune complex formation against autologous MM cells (Figure 1C; E:T = 1:1, control-treated mean 751 ± SD 181 vs IPH2101-treated 1225 ± 207, P = .041) but not against residual CD138⁻ normal BM cells. Similarly, IPH2101 preferentially increased NK-cell lysis of autologous MM targets (Figure 1D; E:T = 20:1, control-treated cytotoxicity = 42% ± 8% vs IPH2101-treated = 63% ± 12%, P = .0135), but no increase in cytotoxicity was observed against CD138⁻ normal BM elements.

Lenalidomide and IPH2101 also increased NK-cell IFN-γ production (Figure 1E left panel shows enriched healthy donor NK cells vs primary MM-cell targets, right panel shows patient-derived NK cells vs autologous MM-cell targets), PBMC cytotoxicity against U266 MM cell line targets (Figure 1F), and patient-derived NK-cell cytotoxicity against autologous MM targets (Figure 1G). Furthermore, in freshly isolated, whole BM aspirates from patients with MM, IPH2101 and lenalidomide altered expression of CD107a on NK cells in BM (P = .0214). Planned comparisons revealed that the percentage of NK cells expressing CD107a increased in BM incubated in the combination of IPH2101 and lenalidomide by 1.81 ± 0.5-fold (P < .05) relative to control conditions (Figure 1H). In these experiments, NK cells comprised 4.1% (± SD 1.7) of BM cells, and this proportion was consistent across conditions.

The expression of MICA (increased 73.6% ± 26.6%), DR4 (increased 95.6% ± 9.5%), ULBP-2 (increased 86.1% ± 16.1%), and CD112 (increased 114% ± 6.9%) were enhanced on U266 cells by 5 days of incubation in lenalidomide (5μM) as shown in Figure 2A left panel (Figure 2A right panel shows that the expression of ULBP-1, CD155, ULBP-3, and ULBP-4 was not affected). In primary MM cells, ULBP-1 expression was consistently increased (153% ± 24%, P = .016 compared with control-treated cells, Figure 2B), whereas the expression of other activating ligands was not reproducibly enhanced in any of the cases examined.

To test the functional consequence of these effects, the cytotoxicity of NK cells incubated in isotype control mAbs or a mixture of anti-TRAIL, anti–DNAM-1, and anti-NKG2D blocking mAbs was measured against MM-cell targets pretreated in DMSO or lenalidomide. NK-cell cytotoxicity was enhanced 1.22 ± 0.05-fold (Figure 2C left panel, P < .05) against U266 cells pretreated with lenalidomide; however, this effect was lost when NK cells were preincubated in a mixture of anti-TRAIL, anti–DNAM-1, and anti-NKG2D blocking Abs. A similar effect was observed in patient-derived NK cells against autologous MM targets (Figure 2C right panel shows representative finding of n = 2 independent experiments).

Mice received lenalidomide (25 mg/kg/d) for 14, 21, or 28 days. One day before inoculation with RMA tumor cells, mice were treated with a murine Ab [anti-Ly49C/I F(ab′)2 or “5E6”], equivalent to the anti-KIR Ab in humans at suboptimal doses of 5 or 10 μg. After 40 hours, the combination of 5E6 and lenalidomide led to significantly reduced tumor burden (Figure 2D-E). These results suggest at least an additive in vivo effect of mAb-mediated inhibitory NK-cell receptor blockade and lenalidomide in combination. Because the RMA cell line is resistant to direct effects of lenalidomide, these findings suggest the immunomodulatory properties of the agent combined with 5E6 to enhance NK cell–mediated tumor cell in vivo clearance.

The therapeutic potential of KIR-ligand interruption was first demonstrated in haplo-identical, T cell–depleted, allogeneic stem cell transplantation in which donor-derived, alloreactive NK cells could lyse recipient tumor cells.^13-15  Infusion of haploidentical, KIR-ligand–mismatched NK cells to relapsed/refractory MM patients is associated with a 50% response rate after reduced-intensity conditioning in the autograft setting, and KIR-ligand mismatch may contribute to the reduced incidence of relapse in the setting of T cell–depleted allografting as well.^7,16 

Whether an NK cell lyses a target cell depends on signaling via activating and inhibitory receptors interacting with cognate ligands on the surface of candidate targets.¹⁷  NK cells must express DNAM-1, NKG2D, or NKp46 to kill MM cells.¹⁸  However, every cytolytic NK cell must express at least one inhibitory KIR ligand, which are present in greater numbers and appear to bind with greater avidity than activating KIR ligands.^19,20  Therefore, even in the presence of activating receptor-ligand interaction, NK cells may be prevented from initiating cytotoxicity if a concomitant inhibitory KIR-ligand interaction is present.^19,20  Therefore, IPH2101 provides release from inhibition, whereas lenalidomide simultaneously promotes NK-cell expansion and activation and appears to induce favorable modulation of activating ligands on MM targets. That a murine anti-inhibitory NK-cell receptor Ab and lenalidomide mediate in vivo rejection of a lenalidomide-resistant tumor suggests that this effect was due primarily to favorable immunomodulatory effects.

A recent phase 3 trial demonstrated superior survival in MM patients receiving lenalidomide with “low-dose” dexamethasone (40 mg weekly) compared with “high-dose” dexamethasone (40 mg/d on days 1-4, 9-12, and 17-20 of a 28-day cycle).²¹  Recent studies demonstrate that dexamethasone also profoundly impairs the immunomodulatory profile of lenalidomide.^22,23  These findings have fostered interest in developing steroid-limiting or steroid-sparing treatments to maximize the immunomodulating properties of lenalidomide.²⁴  Our data provide a preclinical justification for a phase 2 trial of IPH2101 in combination with lenalidomide as the first “dual immunotherapy” for patients with MM.

This study was supported by a Conquer Cancer Foundation Career Development Award (60019229 to D.M.B.), a Multiple Myeloma Opportunities for Research and Education grant (to D.M.B.), and a Pelotonia Idea grant (to D.M.B.).

Contribution: D.M.B. designed and performed the research and wrote the manuscript; C.E.B., S.Z., S.M.C., S.S., M.B., C.B., J.L., and D.C. designed and performed the research; P.A. and F.R. designed the research and contributed vital reagents; J.Z. analyzed the data; C.C.H. and Y.E. designed the research; M.A.C. designed the research and assisted in writing the manuscript; and S.S.F. designed the research and assisted in writing the manuscript.

Conflict-of-interest disclosure: D.M.B. and S.S.F. have received research funding from Innate Pharma. P.A., F.R., M.B., and C.B. are employees of Innate Pharma. The remaining authors declare no competing financial interests.

Correspondence: Don M. Benson Jr, MD, PhD, B-310 Starling Loving Hall, 320 W 10th Ave, Columbus OH 43210-1240; e-mail: don.benson@osumc.edu.