Antibody-Dependent Cellular Phagocytosis by Macrophages is a Novel Mechanism of Action of Elotuzumab

Multiple myeloma is a plasma cell dyscrasia characterized by clonal expansion of malignant plasma cells in the bone marrow leading to lytic bone lesions, cytopenias, and immunodeficiency (1). In light of recent advances of therapy in multiple myeloma including proteasome inhibitors and immunomodulatory therapies, multiple myeloma has transformed into a chronic disease with an approximate median overall survival of 7 to 10 years (2). Despite the advances in therapeutic regimens available, multiple myeloma largely remains incurable with the eventual progression to drug-resistant or relapsed disease (2). Recent multiple myeloma treatment strategies have focused on the use of novel therapies and combinations that harness the antitumor potential of the immune system with the aim of improving prognosis of the disease (3). The implementation of mAbs in the treatment of solid and hematologic malignancies has resulted in considerably improved clinical outcomes for several malignancies (4, 5). Elotuzumab is a novel IgG1 mAb that has recently been approved for the treatment of relapsed or refractory multiple myeloma in combination with lenalidomide and dexamethasone (6). This mAb recognizes a cell surface glycoprotein signaling lymphocytic activation molecule F7 (SLAMF7) that is primarily expressed on myeloma and other cells of hematopoietic lineage including natural killer (NK) cells (7, 8). Although shown to induce NK cellular activation, the binding of elotuzumab to SLAMF7 on myeloma cells has not been shown to induce cellular proliferation or promote tumor cell death (9). In addition, the fragment crystallizable (Fc) region of elotuzumab has been shown to activate NK cells by binding to the Fc gamma receptor III (FcγRIII), which induces myeloma directed antibody-dependent cellular cytotoxicity (ADCC; refs. 7, 8).

The role of the innate immune system has not been thoroughly examined in multiple myeloma. Effector functions of the innate immune system (such as ADCP and ADCC) have been shown to be dependent on the predominantly stimulatory role of FcγR (10). The effect of elotuzumab on other effector immune cells expressing SLAMF7 and FcγR remains widely unexplored. However, an understanding of these mechanisms is necessary if we are to fully exploit the potential of this drug in immune-modulating combinations (11) and further define specific subgroups of patients who would benefit the most from this agent. Within the context of multiple myeloma, tumor-associated macrophages (TAMs) have demonstrated a non-redundant and central role in tumor niche homeostasis (12). FcγR expressing macrophages have been shown to play a considerable role that contributes to efficacy of several widely used mAbs used in antitumor therapy (13, 14). Targeting Her2/neu, CD20, and CD38 respectively, trastuzumab, rituximab, and daratumumab are high FcγR affinity IgG1 antibodies that induce ADCP by macrophages (15–17). In this study, we demonstrate via in vivo xenograft and syngeneic tumor models in addition to in vitro assays the potential role of elotuzumab in inducing FcγR-mediated antitumor effects of macrophages.

The MM.1S GFP⁺Luc⁺ cell line was generated via retroviral transduction using the pGC-GFP/Luc vector (gift of A. Kung, Dana Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts) and cultured in RPMI1640 medium supplemented with l-glutamine and antibiotics (penicillin and streptomycin at 100 U/mL and 100 μg/mL, respectively) and 10% FBS. EG7-human SLAMF7 (hSLAMF7) cells (18) were cultured in RPMI with 2 mmol/L l-glutamine, 10% FBS, 1.5 g/L sodium bicarbonate, 4.5 g/L glucose, 10 mmol/L HEPES (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid), 1 mmol/L sodium pyruvate, 0.05 mmol/L 2-mercaptoethanol, and 0.4 mg/mL G418 (Life Technologies). OPM-2 (DSMZ) cells were cultured in RPMI with 10% FBS and 10 mmol/L HEPES.

Severe combined immunodeficient beige (SCID-beige) and NOD scid gamma (NSG) mice were purchased from Taconic and used for as myeloma xenograft. C57Bl/6j and SCID mice were purchased from The Jackson Laboratory and used for subcutaneous injection models. Female mice used from the indicated strains were 6 to 7 weeks old. All animal studies were approved by the Dana Farber Cancer Institute IACUC and Bristol-Myers Squibb Animal Care and Use Committee and accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International.

Elotuzumab and elotuzumab variants were provided by Bristol-Myers Squibb (BMS). Fc-inert variant of elotuzumab (elotuzumab hIgG1.1): has five mutations L234A, L235E, G237A, A330S, and P331S in Fc region that abrogate binding to human FcγR. The Fc mutant form of the antibody was expressed by the CHO-S cell line cotransfected with vectors pICOFSCneoK (encoding Elo variable regions) and pODpurIgG1.1f (encoding IgG1 heavy chain constant region with L234A-L235E-G237A-A330S-P331S mutations). To generate a mouse IgG2a variant of elotuzumab (elotuzumab-g2a), the heavy chain variable domain (VH) for the elotuzumab parental mAb was cloned into an expression vector containing the mouse IgG2a constant region. The light chain variable region (Vκ) was cloned into an expression vector containing the mouse light chain constant region (18). Human IgG1 (hIgG1) and mg2a isotype controls were obtained from R&D systems (catalog no. 110-HG) and BioXCell (clone C1.18.4), respectively.

SCID-beige and NSG mice were injected with 5 million MM1S GFP⁺ Luc⁺ cells intravenously. For the early tumor model, after 48 hours of tumor injection, mice were treated with intraperitoneal injections of hIgG1, Fc-inert variant of elotuzumab, or elotuzumab at a 10 mg/kg dose twice weekly. The late tumor model comprised treatment of mice posttumor development and randomization, using bioluminescence imaging as described below. SCID mice were injected subcutaneously in their hind flanks with 13 × 10⁶ million OPM2 cells suspended in matrigel diluted in PBS (1:1). Antibody treatment with 1 mg/kg of hIgG1 or 0.5 and 1 mg/kg doses of elotuzumab intraperitoneally was initiated once tumor volume reached an average of 135 mm³ and administered on days 7, 10, and 14.

C57Bl/6j mice were subcutaneously injected in their hind flanks with 5 × 10⁶ million EG7-hSLAMF7 cells. After 5 to 7 days, mice were randomized into treatment groups when tumor volumes reached 80 to 120 mm³. Mice were then injected intraperitoneally with elotuzumab-g2a or mg2a isotype control (10 mg/kg) at the indicated times for a total of three doses. NK cells and macrophages were depleted in vivo using 50 μg of anti-asialo GM1 (ref. 19; eBioscience) or 600 μg anti-CSF1R intraperitoneally (ref. 20; clone AFS98, BioXCell), respectively. Depletion antibodies were administered to separate groups 3 days before initiation of elotuzumab-g2a or mg2a treatment and every 7 days thereafter. Tumor volume was measured biweekly by digital calipers (Fowler) and calculated by the formula: length × width² × 0.52 (21).

MM1S GFP⁺ Luc⁺ xenograft mice were injected with 75 mg/kg luciferin i.p. (Caliper Life Sciences) and imaged with real-time whole-body bioluminescence 5 minutes after the injection. BLI images were obtained via the following settings: Emission Filter Open on IVIS 1327 camera; binning (HR), 4; field of view, 25; flux stop, 1; total flux [photons per second (p/s)] was calculated based on 4 × 8-cm regions of interest.

In SCID-beige xenograft mice, homing of MM1S GFP⁺ Luc⁺ cells to distant bone marrow niches (skull) was traced in vivo, using in vivo confocal microscopy. After 18 hours of MM1S GFP⁺ Luc⁺ intravenous injection and intraperitoneal treatment with elotuzumab or hIgG1 control, myeloma cell homing to the bone marrow was assessed using a Zeiss 710 confocal system (Carl Zeiss Microimaging) with an upright examiner stand and a custom stage. Imagining was done through a skin flap made in the scalp of mice exposing the central and coronary veins on dorsal surface of the skull. GFP was excited using the 488-nm Argon laser while blood vessels were visualized using Evans Blue (100 μL IV; Sigma-Aldrich) excited via a 633-nm laser. Emission signals were quantified by the Zeiss internal confocal Quasar detectors.

Spleen and subcutaneous tumor from OPM2- or EG7-hSLAMF7-bearing mice were resected and single-cell suspensions were prepared by dissociation and passing cells through a 70-μm filter. Cells (5 × 10⁵) were plated in 96-well plates, treated with 2.4G2 (BD Biosciences), and stained with fluorochrome-conjugated antibodies against the following surface markers: NK1.1, CD11b, Gr1, F4/80 (Biolegend). Stained cells were acquired on the FACS Aria II (BD Biosciences) and analyzed using FlowJo software (Tree Star).

After 5 to 6 weeks of injection of MM1S GFP⁺ Luc⁺ in SCID-beige mice, TAMs were isolated as outlined by Zhang and colleagues (22) from the femurs of mice exhibiting signs of hind limb paralysis. TAMs were cultured in DMEM medium the presence of 40 ng/mL of M-CSF for 6 days. Subsequently, TAMs were polarized into an antitumor M1 phenotype using 100 ng/mL of each IFNγ and LPS. Concurrently, target MM1S GFP⁺ Luc⁺ cells were co-cultured at an effector to target (E:T) ratio of 3:1. Also, tumor opsonizing antibodies (100 μg/mL of hIgG1, Fc-inert variant of elotuzumab, or elotuzumab) were added to the co-culture lasting 18 to 24 hours. Subsequently, cells were analyzed using flow cytometry. Colocalization of the fluorescence of F4/80 (Clone: BM8; Biolegend) stained TAMs and GFP labeled MM1S cells (percentage of double positive events) indicated phagocytosis of the tumor cells.

TAMs were isolated from MM1S GFP⁺ Luc⁺ xenograft SCID-beige mice and cultured with MM1S GFP⁺ Luc⁺ cells using E:T ration of 1:5 and stained for F4/80 using 5 μg/ml of Alexa Fluor-594 anti-mouse F4/80 antibody (Clone: BM8, Biolegend). ADCP was visualized using the Leica Sp8 confocal microscope (HC PL APO CS2 40x/1.3 oil objective) at the Harvard NeuroDiscovery Center. Images were saved as .lif files and analyzed in FIJI (ImageJ2).

MM1S GFP⁺ Luc⁺ cells were seeded in a 96-well plate (10,000 cells/well) and cultured in the presence of hIgG1, Fc-inert variant of elotuzumab, or elotuzumab (100 μg/mL). Proliferation was assessed at 24, 48, and 72 hours by measuring 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; Chemicon International) dye absorbance.

Statistical analysis of total BLI flux, percentage of macrophages and their associated surface markers, cellular proliferation, phagocytosis assay, and tumor volumes were performed using unpaired Student t test or Mann–Whitney U test if the data were not normally distributed. For survival analyses, Kaplan–Meier survival probability curves were used with log-rank analysis was used to test for statistical significance in survival differences. P values <0.05 were considered statistically significant. Prism 6 (GraphPad Software) was used to calculate and analyze the statistical differences between experimental groups.

Although elotuzumab has previously been shown to primarily promote myeloma cytotoxity through the stimulation of NK cells via ADCC, we sought to define the antitumor effect of elotuzumab in the absence of functional NK cells by using a severe combined immunodeficiency xenograft mouse model of multiple myeloma, the SCID-beige mouse model (23–25). Macrophages, hence, remain as one of the primary effector immune cells retaining functional antitumor activity in this mouse model (26).

Using the early xenograft model of myeloma (Supplementary Fig. S1A), elotuzumab significantly inhibited tumor growth in treated mice compared with the hIgG1 control group as shown by serial BLI imaging (Fig. 1A; Supplementary Fig. S2A). Moreover, elotuzumab significantly prolonged survival in treated mice compared with the control P < 0.001 (Fig. 1b). To exclude potential direct effect of elotuzumab on the survival and proliferative ability of injected tumor cells, a methyl tetrazolium (MTT) survival assay was performed in vitro. No notable differences in survival was detected in multiple multiple myeloma cell lines tested (MM1S, RPMI, U266) upon culture with hIgG1 or elotuzumab for up to 72 hours (Supplementary Fig. S3A). Similarly, the elotuzumab did not affect the migration and homing of MM1S-GFP-Luc+ cells to the bone marrow in vivo (Supplementary Fig. S3B). Together, these studies suggest that elotuzumab induces its antitumor properties indirectly through immune cell mediators in the xenograft myeloma model.

To further validate the results obtained with the early xenograft myeloma model, the late tumor model was similarly used whereby mice bearing significant tumor burden were treated with elotuzumab (Supplementary Fig. S1A). MM1S GFP⁺ Luc⁺ xenograft mice with BLI signal indicating established tumors were randomized into two groups which received either elotuzumab or hIgG1. There was a significant reduction in tumor volume as measured by the BLI signal intensity in the elotuzumab-treated group compared with the control group (Fig. 1C; Supplementary Fig. S2B). This translated to improved survival of elotuzumab-treated mice in comparison to the control group P = 0.027 (Fig. 1D).

One of the mainly characterized mechanisms of action of elotuzumab is the activation of NK cells via the FcγRIII. Therefore, we aimed to investigate if the antitumor effect of elotuzumab in the SCID-beige model is due to the activation of Fc receptors on the main effector cells, macrophages. Because elotuzumab is a humanized antibody, its activity is restricted to hSLAMF7 interactions with human multiple myeloma cell lines injected into these mice along with the Fc–FcγR interaction with murine effector immune cells. The use of the humanized antibody in xenograft animal models, therefore, focuses on the effects of engagement of FcγR and does not encompass total effects of the antibody comprising SLAMF7 binding on murine immune cells. To delineate the biologic effects mediated via binding of elotuzumab to SLAMF-7 and Fc receptors, an elotuzumab variant incapable of binding to Fc receptors (Fc-inert variant of elotuzumab) was used. Two days after the injection of SCID-beige mice with MM1S GFP⁺ Luc⁺ cells, mice were treated with either hIgG1, Fc-inert variant of elotuzumab, or elotuzumab. BLI intensity was significantly reduced in the elotuzumab-treated group compared with the hIgG1 and Fc-inert variant of elotuzumab-treated groups, indicating a central role of the Fc–FcγR interaction in inducing the observed antitumor effects (Fig. 2A; Supplementary Fig. S2C). In addition, elotuzumab resulted in a significantly improved survival benefit in the treatment group compared with both hIgG1 and Fc-inert variant elotuzumab-treated groups P < 0.001 (Fig. 2B). We then used NSG mice that have severe immunodeficiency comprising the function of macrophages (27) to confirm the requirement of macrophages in the anti-myeloma effect of elotuzumab (Supplementary Fig. S1A). MM1S GFP⁺ Luc⁺ cells were injected into NSG mice and antibody treatment was administered 48 hours after myeloma cell injection as described previously. Interestingly, the beneficial effect of elotuzumab was abrogated as mice that received the treatment exhibited comparable BLI intensity to mice receiving the Fc-inert variant of the antibody (Fig. 2C). No difference in the survival of mice between groups was observed (Supplementary Fig. S3D). These results demonstrate that elotuzumab enables macrophages to prevent tumor growth and progression through the activation of FcγR.

To further confirm that macrophages are implicated effector cells of anti-myeloma activity, we used a subcutaneous syngeneic tumor model in immunocompetent C57Bl/6j mice (Supplementary Fig. S1A). The use of EG7-hSLAMF7, a mouse cell line expressing hSLAMF7, provided a syngeneic surrogate of human myeloma cells and allowed the ex vivo characterization of effector immune cells mediating the antitumor effects of elotuzumab. Because NK cells are well recognized mediators of the antitumor properties of elotuzumab, we aimed to compare the antitumor potency of macrophages exposed in vivo to elotuzumab to that of NK cells. Consequently, macrophage and NK cells were respectively depleted using anti-CSF1R and anti-asialo GM1 antibodies administered intraperitoneally to separate groups of EG7-hSLAMF7 bearing mice. At day 18 posttreatment, the median tumor growth in control mice receiving mIgG2a (667.17 mm³) was significantly reduced in mice receiving elotuzumab without immune cell depletion (elotuzumab-g2a group) that yielded the lowest tumor burden (median tumor volume = 111.25 mm³; P = 0.028). The benefit of elotuzumab treatment in immunocompetent mice (elotuzumab-g2a group) was comparably abrogated upon the depletion of either macrophages (elotuzumab-g2a + CSF1R) or NK cells (elotuzumab-g2a + asialo GM1) given the increase in median tumor volume (669.51 and 686.56 mm³, respectively, P = 0.037 for both groups; Fig. 3a).

In addition, the difference in the tumor volume among NK-cell and macrophage depleted groups not treated with elotuzumab (groups asialo GM1 and CSF1R) to similar groups that were treated further suggests the dependency of antitumor effector functions of these cells on elotuzumab. Upon depletion of macrophages, the median tumor volume for mice concurrently receiving elotuzumab treatment (elotuzumab-g2a + CSF1R group: 669.51 mm³) was decreased compared with those subjected only to macrophage depletion (CSF1R group: 956.54 mm³, not statistically significant) and is mainly attributed to the antitumor effector role of NK cells (Fig. 3A). However, compared to the tumor volume of NK-cell–depleted mice not treated with elotuzumab (asialo GM1 group: 1,264.07 mm³), mice that were subjected to NK-cell depletion and received elotuzumab treatment (elotuzumab-g2a + asialo GM1 group) demonstrated a significant reduction in median tumor volume (686.54 mm³; P = 0.005; Fig. 3A). The benefit in tumor volume reduction obtained upon elotuzumab treatment despite the depletion of NK cells strongly suggests the involvement of other effector cells, primarily macrophages, in mediating the antitumor effects of elotuzumab in immunocompetent syngeneic tumor bearing mice.

The effect of elotuzumab treatment of EG7-hSLAMF7 bearing mice on the accumulation of macrophages at the tumor site was assessed 11 days after the initiation of elotuzumab treatment (total of three administered doses). F4/80 expressing macrophages showed a 1.59-fold increase in percentage upon elotuzumab treatment compared with groups subjected to isotype control (P = 0.042). The efficacy of macrophage depletion is also illustrated among groups that received anti-CSF1R antibody compared with isotype control (Fig. 3b).

We aimed to investigate how the function and phenotype of macrophages is modulated by elotuzumab. A subcutaneous tumor model was used where OPM2 myeloma cells were injected in SCID mice (Supplementary Fig. S1A). Mice received intraperitoneal injections of 1 mg/kg hIgG1 or either 0.5 or 1 mg/kg of elotuzumab. As observed in the other mouse models, a significant reduction in tumor progression was noted among mice treated with elotuzumab 0.5 mg/kg (tumor volume at day 21:643.7 ± 703.3 mm³) compared with the control group (1,952 ± 773.3 mm³, P < 0.001). Similarly, the tumor volume of mice treated with 1 mg/kg elotuzumab (488 ± 524 mm³) demonstrated a significant decrease in tumor volume compared with the hIgG1-treated group (1,952 ± 773.3 mm³, P < 0.001). In addition, among the elotuzumab-treated groups, the lower tumor volume obtained upon treatment with 1 mg/kg elotuzumab suggests a dose response (Fig. 4A).

Phenotypically, we investigated if TAM from elotuzumab-treated mice had an activated phenotype. Classically activated macrophages through lipopolysaccharide (LPS) and IFNγ are characterized by an increase in expression of CD86 and PD-L1 (28, 29). Tumors and spleens from these mice were isolated on days 1 and 8 posttreatment and were analyzed by flow cytometry. A significantly higher percentage of F4/80⁺ TAMs was detected in the 1mg/kg elotuzumab-treated group as early as day 1 posttreatment compared with the control group (P = 0.004; Fig. 4B). Also, the higher dose of elotuzumab resulted in a higher percentage of TAMs as compared with group receiving the lower dose of 0.5 mg/kg (P = 0.024). On day 8, however, the percentage of recruited macrophages among elotuzumab-treated mice was comparable to the control group both in the tumor and the spleen. Notably, among the 1 mg/kg elotuzumab-treated group, TAMs exhibited an early and significantly higher expression of activation markers CD86 (P = 0.0063) and PD-L1 (P = 0.0107) compared with the hIgG1 control. Higher TAM expression levels of CD86 and PD-L1 were also noted in the group receiving a lower dose of elotuzumab, indicating a dose-dependent response. Therefore, although the recruitment of TAMs was decreased by day 8, elotuzumab treatment at both dosages showed sustained phenotypic changes among the macrophages with CD86 and PD-L1 maintaining a higher expression compared with the control group (Fig. 4B). Conversely, macrophages analyzed from spleens did not demonstrate similar phenotypic changes to those noted among TAMs (Fig. 4C). Collectively, these results demonstrate the role of elotuzumab in promoting the accumulation of macrophages at the tumor site and inducing phenotypic changes that likely reflect effector cell activation.

Given the role of elotuzumab in inducing macrophage activation in mouse tumor models, we investigated the effect of the mAb on the antitumor effector function of macrophages. TAMs isolated from the femurs of SCID-beige mice that were injected with MM1S GFP⁺ Luc⁺ cells were polarized in vitro into the antitumor M1 phenotype using IFNγ and LPS. Consequently, the effector M1 macrophages were co-cultured with MM1S GFP⁺ Luc⁺ target cells opsonized with hIgG1, Fc-inert variant of elotuzumab, or elotuzumab. ADCP was analyzed using flow cytometry and assessed through the F4/80 and GFP double-positive cells. Elotuzumab opsonized tumor cells resulted in a significantly higher ADCP activity compared with hIgG1 (P = 0.019). In addition, the use of Fc-inert variant of elotuzumab resulted in a two-fold reduction in ADCP compared with elotuzumab-treated cells (P = 0.0077). Also, the classical activation and polarization of TAMs into the M1 phenotype was crucial in inducing ADCP as evidenced by the loss of in elotuzumab-induced ADCP efficacy among myeloma cells treated with the mAb and co-cultured with undifferentiated TAMs (P = 0.0059; Fig. 5A and B). These results show the necessity of an intact Fc region for enhanced ADCP activity along with the specific Fc–FcγR interactions of elotuzumab with classically activated macrophages that promote myeloma cell phagocytosis.

Furthermore, confocal microscopy confirmed cytometry findings as a more robust phagocytic activity was found among elotuzumab opsonized tumor cells compared with hIgG1 or the mutant Fc-inert variant of elotuzumab (Fig. 5C). Notably, macrophages that were activated with elotuzumab demonstrated the ability to phagocytose multiple cells under confocal microscopy.

Among the most abundant innate immune cells in the tumor microenvironment, TAMs are well recognized as key regulators of tumor progression and metastasis (30). Clinically, TAMs and their associated markers have been correlated with multiple myeloma disease progression and proposed as prognostic biomarkers (31, 32). However, regarded as dysregulated immunosuppressive cells, several immunomodulatory therapies focus on the reprogramming of TAMs to harness their antitumor potential (33–35). Within the realm of multiple myeloma, immunotherapy has demonstrated its ability to re-educate TAMs and stimulate antitumor activity of macrophages in a refractory/relapsed mouse model of myeloma (36).

Monoclonal antibody immunotherapy has been shown capable of not only targeting specific surface cellular antigens but also engaging effector immune cells through the Fc–FcγR interaction. Engaging the FcγR results in the induction of effector immune functions such as ADCC and ADCP among NK cells and macrophages (10, 37). Previously underestimated, this mechanism of action is gaining wide interest given the increased dependence on immune modulating mAbs in the therapeutic armamentarium against cancer.

Elotuzumab is among the first monoclonal therapies approved for the treatment of multiple myeloma. Aside from NK-cell activation via SLAMF7, the other aspect of elotuzumab's dual effect on NK cells comprises inducing FcγRIII-mediated ADCC against myeloma cells (6). Although well described for NK cells, the FcγR-mediated mechanisms of action of elotuzumab remain to be explored among other effector cells in the tumor microenvironment. Our work presented herein proposes a new mechanism of action of elotuzumab whereby FcγR activation of TAMs promotes tumor clearance by ADCP. Using the myeloma xenograft model in SCID-beige mice, we showed the central role of intact Fc region of elotuzumab in inhibiting tumor progression and promoting mouse survival. The use of similar model among NSG mice that lack functional macrophages abrogated these beneficial effects. Furthermore, using syngeneic tumor models expressing hSLAMF7 in immunocompetent mice where macrophages and NK cells were selectively depleted, we show that elotuzumab exerts its antitumor effect comparably through NK cells and macrophages. Because elotuzumab specifically binds to human SLAMF-7, the effects seen in the in vivo mouse models were isolated to Fc–FcγR interactions of the mAb with mouse effector cells. Therefore, contributory effects of SLAMF7 stimulation of effector cells were not feasible to assess in this study.

Macrophage accumulation at the tumor sites was also augmented through elotuzumab treatment in a dose-dependent manner. This can be due to the involvement of macrophage chemoattractants such as MIP-1-α, MCP-1, and MIP-2 (38, 39). Phenotypically, these macrophages exhibited an activated phenotype evidenced by increase in CD86 expression, a hallmark of classically activated M1 macrophages with LPS and IFNγ (28). Known to be increased in the presence of IFNγ (40), PD-L1 expression was also upregulated upon treatment with elotuzumab. Hence, given the coexpression of CD86 and PD-L1, elotuzumab potentially primes the tumor microenvironment with an IFNγ predominant inflammatory milieu. Although PD-L1 has been associated with an immunosuppressive function of TAMs, its expression corresponds primarily to a negative feedback mechanism aiming to decrease the inflammatory response (41). Notably, the upregulation of PD-L1 expression on TAMs has been shown to be correspond to higher B and T lymphocyte tumor infiltration (42–46). Therefore, this points to potential future benefit of elotuzumab and adjunctive PD1/PD-L1 axis blockade in the treatment of myeloma.

The engagement of human SLAMF7 by elotuzumab holds promise in augmenting the potential Fc–FcγR interactions among human effector cells. Transcriptome analysis of the human macrophage surfaceome by Beyer and colleagues has shown differentially higher expression of SLAMF7 among antitumor M1 macrophages as compared with M2 macrophages (47). Also, the expression of SLAMF7 among human M1 macrophages has also been demonstrated in intestinal allografts undergoing rejection (48). Recently, upon SIRPα–CD47 blockade, SLAMF7 has been shown to play a critical role in the phagocytosis of hematologic tumor cells by macrophages (49). Given the overexpression of CD47 in several tumor types including myeloma, synergistic blockade of the SIRPα–CD47 axis along with a tumor-directed mAb treatment is a potentially promising strategy to promote phagocytosis effector functions (13, 16). In this study, the biologic effects of elotuzumab treatment in the mouse models were due to Fc–FcγR interaction given the binding specificity of the mAb to human SLAMF7. Therefore, engaging the SLAMF7 signaling pathway is anticipated to potentially augment the FcγR pro-phagocytic properties of elotuzumab. The dual phagocytosis promoting mechanisms of elotuzumab through engagement of SLAMF7 and the FcγR might also be further improved by blockade of the SIRPα–CD47 blockade and warrant future studies. Given that elotuzumab is most clinically efficacious in combination with lenalidomide and dexamethasone, further work is needed to characterize the role of the FcγR pro-phagocytic properties of elotuzumab in combination with immunomodulatory therapies.

In summary, our study demonstrates the ability of elotuzumab to promote macrophage mediated anti-myeloma phagocytic activity through engaging the FcγR. In addition to increasing macrophage accumulation at the tumor site, we show that elotuzumab increases macrophage activation. This novel mechanism of immune mediated anti-myeloma activity may help to further define useful combination strategies for this immunotherapy in multiple myeloma and precursor multiple myeloma states.

N.A. Bezman has ownership interest (including patents) in Bristol Myers Squibb. A. Roccaro is a consultant/advisory board member for Amgen. M.D. Robbins is a Myeloma Medical Lead at Bristol-Myers Squibb. I.M. Ghobrial is a consultant/advisory board member for elotuzumab. No potential conflicts of interest were disclosed by the other authors.

Conception and design: A.T. Kurdi, S.V. Glavey, N.A. Bezman, A. Jhatakia, J.L. Guerriero, S. Manier, A. Roccaro, M.D. Robbins, J. Azzi, I.M. Ghobrial

Development of methodology: A.T. Kurdi, S.V. Glavey, N.A. Bezman, A. Jhatakia, J.L. Guerriero, D. Huynh, I.M. Ghobrial

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A.T. Kurdi, S.V. Glavey, A. Jhatakia, S. Manier, M. Moschetta, A. Sacco, D. Huynh, Y.-T. Tai, I.M. Ghobrial

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A.T. Kurdi, S.V. Glavey, N.A. Bezman, A. Jhatakia, J.L. Guerriero, S. Manier, M. Moschetta, Y. Mishima, A. Roccaro, A. Detappe, C.-J. Liu, M.D. Robbins

Writing, review, and/or revision of the manuscript: A.T. Kurdi, S.V. Glavey, N.A. Bezman, J.L. Guerriero, S. Manier, M. Moschetta, A. Detappe, C.-J. Liu, M.D. Robbins, J. Azzi

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A.T. Kurdi

Study supervision: N.A. Bezman, J. Azzi

The project was supported by grants from the National Cancer Institute Grant No. R01 CA181683-01A1 and the Leukemia and Lymphoma Society granted to I.M. Ghobrial. The project was partially supported by Bristol-Myers Squibb with a grant supporting A.T. Kurdi and S.V. Glavey. The authors thank the Dana Farber Cancer Institute and Bristol Myers Squibb Animal Facilities where the tumor studies were conducted. We also would like to thank Lai Ding and Daniel Tom at the Harvard Neurodiscovery Group for their assistance with image acquisition using the confocal microscopy.

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.