|Home | About | Journals | Submit | Contact Us | Français|
Cellular immunotherapy can be an effective adjuvant treatment for multiple myeloma (MM), as demonstrated by induction of durable remissions after allogeneic stem cell transplantation. However, anti-myeloma immunity is often hampered by suppressive mechanisms in the tumor micro-environment resulting in relapse or disease progression. To overcome this immunosuppression, new cellular immunotherapies have been developed, based on the important effector cells in anti-myeloma immunity, namely T cells and natural killer cells. These effectors can be modulated to improve their functionality, activated by dendritic cell vaccines, or combined with immune stimulating antibodies or immunomodulatory drugs to enhance their efficacy. In this review, we discuss promising pre-clinical and clinical data in the field of cellular immunotherapy in MM. In addition, we address the potential of combining these strategies with other therapies to maximize clinical effects without increasing toxicity. The reviewed therapies might pave the way to effective personalized treatments for MM patients.
Multiple myeloma (MM) is an aggressive plasma cell disease, which accounts for approximately 10% of all hematologic neoplasms.1 In Western countries, the annual incidence is 5.4 cases per 10000 persons.2 MM is characterized by clonal proliferation and accumulation of malignant plasma cells in the bone marrow and a high concentration of monoclonal immunoglobulin light and/or heavy chains in the blood or urine. This results in organ damage and clinical symptoms including anemia, bone pain, renal insufficiency, hypercalcemia, and infections.3 Over the past decade overall survival (OS) has improved significantly, especially in younger patients, because of the introduction of novel therapies like immunomodulatory drugs and proteasome inhibitors before and after autologous hematopoietic stem cell transplantation (SCT). Currently, the average 10-y OS is approximately 17% for all ages, and in patients younger than 60 y the 10-y OS is about 30%.4
Current treatment of young patients, generally defined as 65 y of age or younger, consists of induction therapy followed by autologous SCT. The induction therapy can reduce tumor mass and create a state of minimal residual disease (MRD). Initially, it consisted of conventional chemotherapy, however, nowadays immunomodulatory derivates (IMiDs), like thalidomide and lenalidomide and the proteasome inhibitor bortezomib are combined with chemotherapy. These new combination therapies improved the complete response (CR) rates, which are correlated with improved progression-free survival.5 After induction therapy, patients are treated with high dose melphalan in order to destroy residual tumor cells, followed by autologous stem cell rescue6
Older patients more often have a poor performance status and suffer from co-morbidities. Although it has been shown that autologous SCT is feasible in patients up to 75 y of age with good performance status,7 benefits for older patients have not been demonstrated in clinical trials as these patients are often excluded.8 Therefore, patients above 65 y of age are generally excluded from autologous SCT.7 Traditionally, older patients were treated with melphalan and prednisone (MP). In recent years, phase III randomized trials in patients not eligible for SCT have investigated the combination of MP plus thalidomide or lenalidomide, or proteasome inhibitors, like bortezomib and recently carfilzomib, on outcome. These studies, demonstrated improved progression-free survival (PFS) in patients with combined therapy.9 MP was compared with the combination of MP and thalidomide (MPT) in a meta-analysis using pooled data of 1682 patients treated in 6 different trials. Median OS was 32.7 (95% CI 30.4–36.5) months in the MP arm and 39.3 (95% CI 35.6–39) months in the MPT arm. Median PFS was 14.9 (14.0–16.6) in the MP arm and 20.4 (18.8–21.6) months in the MPT arm.10 The largest MP-based phase III study so far, the VISTA (Velcade as Initial Standard Therapy) trial, investigated combination of MP with bortezomib (VMP) in 682 patients. Time to progression in the VMP group was 24.0 mo, as compared with 16.6 mo in the MP group (P < 0.001). Furthermore CR rates were 30% and 4%, respectively (P < 0.001).11 After 3 y, OS rates were 68.5% in the VMP group vs. 54.0% in the MP group.12 However, despite these improvements in MM treatment, OS is still poor and most patients eventually experience relapse of the disease. Therefore, additional potent therapeutic strategies are urgently needed.
In this review, we will discuss promising novel cellular immunotherapeutic therapies, which could improve outcome in MM patients with reduced side effects. We will first describe how allogeneic SCT, which is the oldest immunotherapeutic strategy in MM, indicated the importance of the immune system in targeting MM. Second, we will explain how MM can progress or relapse by evasion of the immune system. Finally, we will address how different cellular immunotherapeutic strategies, alone or in combination with other therapies, can circumvent immune evasion and thereby improve anti-myeloma immune responses.
Hematopoietic SCT is a well-established treatment for MM patients. In autologous SCT, stem cells are isolated from the patients themselves and may contain residual tumor cells, which can cause relapse of the disease. Additionally, malignant plasma cells that survive the high dose melphalan may cause relapse of the original disease. In allogeneic SCT, stem cells are derived from a Human Leukocyte Antigen (HLA)-matched healthy donor and a potent graft-vs.-myeloma (GVM) response can be induced. This response can eliminate residual tumor cells in the patient, thereby resulting in long-term remission and potentially even cure of the disease. However, allogeneic SCT is curative only in a minority of MM patients, and treatment-related mortality (TRM) is generally high.
Important immune effectors involved in the GVM response are T cells and Natural Killer (NK) cells. T cells can recognize specific antigens presented by HLA molecules via their T cell receptor (TCR). When T cells encounter their cognate antigens and receive appropriate co-stimulation, they become activated and acquire effector functions. In MM, T cell responses can be induced toward the tumor specific immunoglobulin idiotype (Id) protein and/or tumor-associated antigens (TAAs). These latter are antigens expressed at high levels by the tumor cells, but generally also at low levels by normal cells which limits their immunogenicity.13 Important TAAs in MM are cancer germline antigens (CGAGs), like Mage, Gage, Lage and NY-ESO-1,14 Survivin,15 BCMA,16 and MUC1.17 Moreover, in the allogeneic SCT setting potent immune responses can be generated against recipient-specific allo-antigens, known as minor histocompatibility antigens (MiHAs). MiHAs are polymorphic peptides derived from intracellular proteins that are presented by HLA molecules, and differ between donors and recipients. Numerous MiHAs have been identified in the past decades and T cell responses against these MiHAs have been associated with improved relapse-free survival. While in some studies the induction of MiHA-specific T cell responses was associated with an increase in the incidence of GVHD and improved relapse-free survival,18-21 other studies could not confirm these results.22,23 Importantly, boosting of T cell responses against MiHAs with a hematopoietic-restricted expression pattern, e.g., HA1,24 LRH1,25 ARHGDIB,26 and UTA2–127 has the potential to induce a selective GVM effect with only limited risk of eliciting GVHD. Therefore, these MiHAs are interesting candidates for targeted immunotherapy.
The other important immune effectors are NK cells, which are part of the innate immune system. Their activation is regulated by the balance in expression levels of numerous inhibitory and activating receptors. The most well characterized inhibitory receptors are the killer immunoglobulin-like receptors (KIR) and NKG2A. KIR receptors can bind to HLA-A, -B, and -C molecules, while NKG2A binds to HLA-E. Examples of activating receptors are CD16, which is involved in antibody-dependent cytotoxicity (ADCC), activating KIRs (e.g., KIR2DS, KIR3DS), NKG2D, DNAX accessory molecule-1 (DNAM-1), and the natural cytotoxicity receptors (NCRs). These latter receptors can interact with ligands, like UL16-binding protein (ULBP)1–4, MHC class I chain-related protein A (MIC-A) and Nectin-2, that are expressed during infections or stress. In homeostasis, NK cells are inhibited by their inhibitory receptors recognizing self HLA class I molecules. On the other hand, GVM effect can be induced by upregulation of activating ligands or downregulation of MHC class I molecules. In addition, in the setting of allogeneic SCT, donor NK cells may lack expression of inhibitory KIRs for recipient MHC class I molecules and hence be activated. This phenomenon is called missing-self recognition and can contribute to the GVM effect.28,29 Nevertheless, this effect is usually limited because in allogeneic SCT donor and recipient are matched for their HLA molecules. This is essential to prevent induction of severe alloreactive T cell responses against healthy tissues expressing foreign HLA molecules causing graft-vs.-host disease (GVHD).30
Despite the immune susceptibility of MM, the OS after allogeneic SCT was not improved compared with autologous SCT. This was mainly due to high TRM after allogeneic SCT consisting of conditioning-related toxicity, infections, and GVHD.31,32 Furthermore, relapse rates were still high. In recent years, transplant-related toxicity decreased significantly due to the introduction of reduced intensity conditioning (RIC) chemotherapy regimens, but unfortunately OS did not improve due to higher relapse rates.33 In order to reduce the incidence and severity of GVHD, complete and partial T cell-depleted allogeneic SCT programs have been developed. After six months, when the treatment-related inflammation has resolved, GVM responses can be boosted by giving donor lymphocyte infusions (DLI).34 Furthermore, DLI can also be an effective therapy in relapsed MM.35
RIC allogeneic SCT has been compared with autologous SCT in six large trials.36-41 Two trials reported a survival benefit for patients receiving allogeneic SCT. The first was performed by the Italian GIMEMA group and found a median OS of 80 mo in 80 patients in the allogeneic SCT group compared with 54 mo in 82 patients in the autologous SCT group.37 In addition the EBMT observed improved OS after 6 y in 108 patients who underwent autologous/allogeneic SCT (auto/alloSCT) compared with 249 patients receiving only autologous SCT (49% vs. 36%, P = 0.030).40 Four trials did not show improved OS after allogeneic SCT. The first study was performed by Garban et al.36 who observed a trend toward better OS in the autologous group at 56 mo (OS 48 vs. 34 mo, P = 0.07),42 however high dose anti-thymocyte globulin (ATG) was part of the conditioning regimen. The PETHEMA study included patients not achieving a CR after a first autologous SCT and found no difference in OS after 5 y (autologous vs. allogeneic = 60% [95% CI 48.3–73%] vs. 61.8% [95% CI 40.6–82%]).38 The BMT CTN study compared auto/alloSCT in 226 patients with double autologous SCT in 484 patients. OS survival was 77% in the auto/alloSCT-group and 80% in the double autologous SCT-group after three years, however this follow-up period is still relatively short, as the EBMT trial only observed improved OS after six years.41 In a study by the Dutch HOVON Group 115 patients who underwent a single autologous SCT followed by maintenance therapy consisting of α-interferon or thalidomide were compared with 99 patients with an auto/alloSCT with a follow-up of 6 y. After 6 y, PFS was prolonged in patients who underwent allogeneic SCT compared with autologous SCT (HR 0.75, 95% CI = 0.55–1.03, P = 0.07), however OS was not different.39 Remarkably, survival rates of the autologous SCT-group in this study were better than those in other studies. An explanation for this could be that one of the novel agents, thalidomide, bortezomib, or lenalidomide could have been given to patients with relapsed disease. These compounds can significantly prolong survival of relapsed patients, therefore the survival benefit for allogeneic SCT will only become clear very long time after transplantation. Therefore, allogeneic SCT is currently not considered first line treatment for MM patients.43
Despite the high immunogenicity of MM, as demonstrated with allogeneic SCT, still too many patients relapse after initial therapy. This might be the result of immune evasion mechanisms exploited by the tumor cells, including intrinsic alterations and the establishment of an immunosuppressive milieu thereby limiting the efficacy of immune effector cells.44 Here, we will discuss which mechanisms can be involved in immune escape of MM (Fig. 1).
Presentation of the tumor antigen is essential for the induction of tumor-reactive T cell responses. In order to activate T cells, the TCR:CD3 complex should interact with the HLA-antigen complex presented on antigen presenting cells (APCs).44 It has been described that antigen expression can be downregulated by tumor cells. Furthermore, defects in components of the antigen presentation machinery can occur, including the transporter associated with antigen processing-1 (TAP-1) and subunits of the immunoproteasome (LMP-2, LMP-7).45 In addition, tumor cells can downregulate or even lose expression of HLA class I proteins, due to deletions in chromosome 6, where HLA alleles are located, or due to mutations in the β2-microglobulin, an essential molecule for stable HLA expression on the cell surface.46 All these modifications contribute to invisibility of the tumor cells, thereby hampering efficient recognition by tumor-reactive T cells.
In addition to presentation of the antigen by an APC, T cells need a second signal provided by co-stimulatory molecules of the B7/CD28 family in order to become activated. Together, these two signals trigger expansion and differentiation of the T cells, and induce acquisition of effector functions. Following T cell activation, expression levels of co-inhibitory molecules like Cytotoxic T lymphocyte associated antigen-4 (CTLA-4), B and T lymphocyte attenuator (BTLA), and Programmed death-1 (PD-1) are upregulated. Ligation of these receptors to their corresponding ligands on APCs results in functional inhibition of the T cells. Via this natural feedback loop sustained T cell activation is prevented and the effector T cell response is resolved.47 However, tumor cells can upregulate co-inhibitory molecules and downregulate co-stimulatory molecules to acquire an immune inhibitory phenotype and thereby prevent productive tumor-reactive T cell responses. In MM increased expression levels of the co-inhibitory ligands of PD-1, BTLA, and CD200R have been observed on MM cells.47-50 Moreover, interference with BTLA signaling using blocking antibodies can augment proliferation of tumor-reactive CD8+ T cells stimulated with peptide-loaded HVEM+ DCs49 and small interfering RNA (siRNA) mediated silencing of PD-L1 and PD-L2 on DCs enhances their potential to stimulate T cell proliferation and cytokine production.51 In addition to the B7/CD28 family members, MM cells can also express other inhibitory molecules, like carcinoembryonic antigen-related cell adhesion molecule-6 (CEACAM-6), an immunoglobulin-like receptor. It has been shown that increased expression of CEACAM-6 in MM resulted in inhibition of anti-myeloma T cell responses.52
Another way to evade the immune system is creation of an immunosuppressive microenvironment by secretion of immune inhibitory factors like soluble MIC-A, interleukins (IL), transforming growth factor (TGF)-β and indoleamine 2,3-dioxygenase (IDO). Increased levels of these factors have been detected in serum of MM patients.53-56 Secretion of soluble MIC-A by MM cells can mediate inhibition of NK cells and CD8+ T cells by downregulation of NKG2D, and is associated with poor survival.54,55 Moreover, IL-6 producing tumors can directly impair NK cell cytotoxicity and can stimulate IL-10 production in MM cells.57,58 This IL-10 can suppress dendritic cell (DC) function and induce the development of Th2 T cells, which are less effective in supporting tumor-reactive cytotoxic T cell formation and function than the Th1 counterparts.44 Furthermore, interferon (IFN)-γ production by NK cells, which is important for their killing capacity, is inhibited by IL-10.59 In addition, TGF-β can downregulate expression of activating NK cell receptors, such as NCRs and NKG2D, and also inhibits IFN-γ production by NK cells, as well as T cells. Additionally, other effector molecules like perforin are downregulated and DC activation can be blocked by TGF-β.60 Finally, IDO is an enzyme that catalyzes metabolization of the essential amino acid tryptophan resulting in tryptophan depletion in the tumor microenvironment. This causes cell cycle arrest and apoptosis of effector T cells.61
Beside apoptosis of T cells, IDO and also IL-10 and TGF-β can recruit or induce suppressive immune cells like M2 macrophages, regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs). Tregs are characterized by expression of CD25 and transcription factor FOXP3 and are capable of inducing expression of co-inhibitory molecules on APCs and IDO production by APCs. In addition, they produce IL-10 and TGF-β themselves, and have the potential to kill APCs and cytotoxic T cells via perforin- and granzyme-dependent pathways.62,63 Increased levels of Treg subsets have been observed in blood of MM patients and were associated with increased disease burden and poor survival.56,64 Other suppressive cell types are MDSCs and M2 macrophages. MDSCs are a heterogeneous subset of immature myeloid progenitor cells, characterized by expression of CD33 and CD11b, and lack of CD14 and HLA-DR expression. M2 macrophages have an IL12loIL-10hiIL-1decoyRhiIL-1RAhi phenotype and express scavenger receptors such as the mannose receptor (CD206) and hemoglobin/haptoglobin scavenger receptor (CD163) on their cell surface.65 MDSCs and M2 macrophages can suppress both T cell and NK cell mediated immune responses by arginine depletion due to catabolism by arginase-1, and nitric oxide production by inducible nitric oxygen synthase.66,67 Increased numbers of MDSCs have been observed in the blood and bone marrow of MM patients. Notably, it has been demonstrated that MM cells can induce development of MDSCs from healthy donor peripheral blood mononuclear cells (PBMCs).68,69 In one study in 68 patients, increased numbers of M2 macrophages, characterized by expression of CD163 and CD68, in the bone marrow of MM patients, were associated with an unfavorable prognostic impact on 6-y OS.70
To overcome immune escape, resulting in improved PFS and OS in MM patients, potent cellular immunotherapies boosting T or NK cell-mediated immunity can be exploited. These highly potent immune effectors can be given either as single therapy, or in combination with immune stimulating antibodies or immunomodulatory drugs (Fig. 2). First, we will address promising T cell based therapies, and in the next part we will discuss interesting NK cell based therapies. An overview of the reported and ongoing clinical trials in MM patients is given in Table 1.
The presence of circulating T cells specific for myeloma-associated antigens has been correlated with increased OS in MM patients, indicating the importance of exploiting myeloma-reactive T cells to further improve outcome.71 However, the isolation and expansion of these low frequent T cells for adoptive transfer is challenging. Therefore, genetic modification of T cells by induction of CARs is a promising alternative. These CARs consist of a single chain variable fragment, derived from a monoclonal antibody specific for a tumor specific protein, fused to native human T cell receptor CD3ζ signaling domains. The 2nd and 3rd generation CARs have one or more additional activating signaling domains derived from co-stimulatory molecules such as CD28, 4–1BB, and OX40. Very promising results have been obtained in pilot studies using CD19-reactive CAR-T cells in CD19 expressing B cell leukemias and lymphomas.72-74 Kochenderfer et al.74 infused 8 patients with either advanced chronic lymphocytic leukemia or B cell lymphoma with 0.3 to 3 × 107 CAR-T cells/kg in combination with IL-2, after conditioning with cyclophosphamide and fludarabine. Importantly, six patients obtained remission showing the potent anti-tumor efficacy of this CAR-T cell therapy. However, 4/8 patients showed long-term depletion of normal polyclonal CD19+ B-lineage cells. Furthermore, 4 patients had prominent elevations in serum levels of IFNγ and TNFα resulting in acute toxicity. Other studies also reported chronic hypogammaglobulinemia72,73 and delayed tumor lysis syndrome.72 In order to improve safety, new methods to ablate CAR-T cells are being developed, like the incorporation of a suicide gene that encodes human caspase-9 fused to a modified human FK-binding protein that confers sensitivity to a synthetic small molecule.75
For the treatment of MM various CARs have been developed. For instance, T cells can be transfected with a chimeric NKG2D receptor. MM cells express the corresponding ligands, which enhance NK cell and CD8+ T cell mediated lysis. Infusion of these NKG2D-expressing CAR-T cells induced tumor reduction in a mouse model.76 Even though most healthy tissues do not express NKG2D ligands, low levels of NKG2D ligands could potentially induce toxicity in patients. Therefore, more specific targets are probably warranted. In this regard, expression of the carbohydrate antigen Lewis(Y) has been detected on 52% of plasma cells in bone marrow samples of MM patients, while it is expressed in very low levels on healthy tissues.77 In a mouse model delayed growth of myeloma xenografts was observed in NOD/SCID mice treated with CAR-T cells directed against Lewis(Y).78 Safety of the Lewis Y CAR is currently investigated in a phase I trial in AML, myelodysplastic syndrome (MDS), and MM patients (NCT01716364). Notably, due to the relatively low expression levels of Lewis(Y) on MM cells there is a potential risk of immune escape. Hence, B cell maturation antigen (BCMA) might be a more suitable target, as it is uniformly and highly expressed on MM cells and expression is absent on normal tissues, except for normal plasma cells. Preclinical studies have shown that CAR-T cells specific for BCMA are able to eradicate human MM cell line tumors in immunodeficient mice. These data indicate that adoptive transfer of CAR-T cells directed against BCMA might be an interesting approach for the treatment of MM.79
Another strategy to boost anti-myeloma T cell immunity in MM patients is DC vaccination. DCs are the most potent professional APCs, and therefore they an attractive way to expand myeloma-reactive T cells in vivo. Reichardt et al. were the first to explore DC vaccination in MM patients.80 In this pioneering study, 12 patients were treated with blood-derived precursor DCs, cultured for 36–48 h in the presence of the Id protein. Two out of 12 patients developed an Id-specific proliferative T cell response and remained in complete remission. Moreover, one patient developed a transient Id-specific cytotoxic T cell response. Following this first report, multiple clinical studies have investigated Id-pulsed DCs and showed that these DCs are capable of inducing humoral and cellular immune responses against the Id protein with limited toxicity.81-83 However, clinical efficacy was disappointing. A possible explanation could be that most of these studies were performed in patients with advanced disease. Nevertheless, one study in stage-I myeloma also showed a decrease in M-protein in only 3 out of 9 patients.84 Furthermore, most studies have been performed using relatively immature DCs, while it has been shown that mature DCs are superior in the induction of anti-tumor immunity.85 Another explanation could be the lack of Id-specific T cell precursors in MM patients, due to tolerance and deletion as a result of high amounts of secreted free protein.86 Consequently, other TAAs can be targeted to improve DC vaccination.
Recently, we performed a phase 1 clinical trial in which we evaluated safety and immunological effects of mature DCs pulsed with TAA mRNA, specific for MAGE3, Survivin, and BCMA, in 12 MM patients with a CR or partial response (PR) after high-dose chemotherapy and autologous SCT.87 Loading with TAA RNA results in processing of both class I and II epitopes, and hence has the potential to induce a broad TAA-reactive T cell repertoire.88 In all patients vaccination was well tolerated with limited toxicity. Importantly, in two patients vaccine-specific T cells were detected, illustrating that TAA-mRNA-loaded mature DCs are capable of inducing TAA-reactive T cell responses in MM patients after autologous SCT. Another strategy to achieve presentation of a broad panel of MM antigens is the use of autologous DC/MM cell fusion vaccines. These vaccines expanded circulating CD4+ and CD8+ cells reactive to autologous MM cells, and were demonstrated to stabilize disease in 11 out of 16 MM patients with advanced disease in a phase I study.89 In a phase II study, where patients were vaccinated with DC/MM cell fusion vaccines after autologous SCT, 24% of the patients who achieved a PR upon transplant converted to complete remission/near complete remission (CR/nCR) following vaccination.90
In order to further augment the activation and expansion of anti-myeloma T cells, DC vaccination may be combined with strategies that interfere with immunosuppressive mechanisms exploited by MM cells. We recently showed that siRNA mediated silencing of programmed death ligand-1 (PD-L1) and PD-L2 on monocyte-derived DCs resulted in generation of DCs with superior potential to stimulate tumor-reactive T cell responses ex vivo.51,91
Besides using PD-L siRNAs, DC vaccine efficacy can also be improved by combination with blocking antibodies targeting the PD-1/PD-L co-inhibition pathway. Incubation of DC/MM cell fusions with anti-PD-1 antibody (CT-011) and subsequent co-culture with autologous T cells, resulted in reduced numbers of regulatory T cells, increased levels of Th1 T cells and augmented tumor killing in ex vivo cytotoxicity assays.92 Currently, the combination therapy of DC/MM cell fusions with CT-011 after autologous SCT is being investigated in MM patients (NCT01067287). Recently, a phase I clinical trial investigating CT-011 in patients with varying hematological malignancies, including one patient with MM, has been finished. CT-011 proved to be safe, and a clinical benefit was observed in 33% of 17 patients. Patients showed increased CD4+ T cell counts without further evidence of T cell activation.93 In addition to improved T cell immunity, Benson et al.94 showed that NK cell-mediated killing of MM cells can also be enhanced by CT-011, suggesting a role for CT-011 in NK cell-based therapies. Another potent anti-PD-1 antibody, nivolumab, has shown durable tumor regression in a phase 1 trial in patients with melanoma, renal-cell cancer, and non-small-cell lung cancer, and is currently investigated in patients with hematological malignancies in a phase I trial (NCT01592370).
Besides T cells, NK cells are important effector cells involved in anti-myeloma immunity. They are activated upon increased expression of activating ligands or decreased expression of inhibitory ligands. In vitro studies indicated that cytokines, like IL-2 and IL-15 can enhance the killing capacity of NK cells.95 Therefore, the effect of in vivo administration of these cytokines on NK expansion and activation has been investigated in patients.96 It has been shown that IL-2 administration mediated expansion of the NK cells. In addition, IFN-α increased NK cell activity, however the therapy induced toxicity and clinical benefits in MM patients were limited.97-99
Importantly, various studies indicate an anti-tumor effect mediated by NK cells. For example, after allogeneic SCT early NK cell repopulation has been associated with decreased relapse rates, without increasing GVHD incidence.100 In haplo-identical SCT, where patient and donor are not fully matched for HLA-type, improved disease-free survival and lower relapse rates, without increased incidence of GVHD, have been observed if the patient and donor were mismatched for their KIR-ligands.29 Those results have been confirmed in different studies,101,102 though other groups could not reproduce these findings.103,104 These differences can probably be attributed to variance in conditioning and transplantation regimens.96 The KIR gene-gene model, which uses KIR genotype, is another method used to predict NK reactivity and clinical outcome after allogeneic SCT. KIR genes are located apart from HLA genes on chromosome 16. Two KIR genotypes can be distinguished; genotype A, which contains mostly inhibitory KIR receptors, and group B, which comprises more activating receptors. Consequently, donors and recipients can be categorized as having one of two KIR genotypes: AA which is homozygous for group A KIR haplotypes, or Bx, which contains either one (AB) or two (BB) group B haplotypes.105 After non-myeloablative HLA-haploidentical SCT, it was observed that patients with KIR group A haplotypes showed improved overall survival, event-free survival, and non-relapse mortality in case the donor had a KIR group B haplotype. For patients with KIR group B haplotypes no benefit was demonstrated.106 Furthermore, improved PFS and OS was observed in acute myeloid leukemia (AML) patients and relapsed MM patients after HLA-matched SCT when the donor had a group B haplotype. This was independent of the patients haplotype. In AML patients no effect on the incidence of GVHD was found, however in MM patients a donor with a group B haplotype was associated with an increased incidence of chronic GVHD, but not acute GVHD.107,108
Nevertheless, allogeneic SCT is not standard treatment for MM, because of the high TRM.43 Therefore, novel strategies exploiting the anti-myeloma activity of NK cells in MM with less side effects are being explored. One approach is the adoptive transfer of NK cells. This provides the opportunity to infuse alloreactive NK cells, that are optimally mismatched, without the toxicity associated with allogeneic SCT. Shi et al.109 investigated infusion of haplo-identical KIR mismatched NK cells in relapsed MM patients following conditioning with melphalan and fludarabine. Furthermore, patients were treated daily with IL-2 injections for 11 d, and received a delayed autograft at day 14. Although donor NK cells could be detected, they disappeared by day 9 – 14. Moreover, the response rate was 50% compared with 40% in the control group not receiving NK cells (P = 0.32). An explanation for the quick disappearance of NK cells could be the early repopulation of T cells. Furthermore, the number of infused NK cells was low (1x106/kg). In a study by Miller et al.110 in poor risk AML patients a complete hematologic remission was induced in 5 out of 19 patients by infusion of ex vivo expanded alloreactive NK cells combined with a high dose immunosuppressive regimen and low dose IL-2. However, in this study low levels of T cells in the graft could have contributed to the observed effect.110,111 In order to infuse higher numbers of NK cells without T cell contamination a new GMP-grade culture protocol has been developed by our group. In this procedure CD34+ hematopoietic stem cells can be expanded ex vivo 2000-fold, and a NK cell product with a purity >90% and without B or T cell-contamination can be generated.112 Currently, these NK cells are investigated in a phase I clinical trial in older AML patients who are not eligible for allogeneic SCT (NTR2818). In the future these NK cells, combined with autologous SCT, could also be a promising therapy in relapsed MM patients.
Beside adoptive transfer of NK cells, several other therapies exist that aim to improve NK cell mediated anti-myeloma immunity. New anti-myeloma drugs introduced into the clinic have shown to influence NK cell mediated killing, therefore combination therapy with these drugs can be a rational therapeutic strategy. Furthermore, in the past decade increasing numbers of clinical grade antibodies have been developed for the treatment of cancer. These antibodies can mediate tumor clearance via various mechanisms.113 Upon binding to the target molecule, the Fc-tail of the antibody can be recognized by Fc-receptors on NK cells resulting in NK cell degranulation and lysis of the target cell. Besides this ADCC, cytolysis can be mediated by activation of the complement system upon binding of complement components to the Fc-tail of the antibody. In addition, the antibody itself may directly affect signal transduction of the targeted molecule, or induce apoptosis. Furthermore, cytotoxic drugs or radionuclides can be coupled to the antibodies, thereby the targeted cells are specifically damaged and patients will encounter less side effects than with regular treatment. At the moment, several monoclonal antibodies are being developed for MM.114 Some of these antibodies are promising candidates for combination therapy, together with adoptive transfer of NK cells, to enhance NK cell mediated killing.
One of these interesting antibodies, is the anti-KIR antibody IPH2101. NK cells initiate cytotoxicity against target cells through a positive balance of signals received via activating and inhibitory receptors. Binding of inhibitory KIRs to MHC class I molecules on target cells prevents NK cell activation.96 IPH2101 is a human IgG4 monoclonal antibody against common inhibitory KIR2DL-1, KIR2DL-2, and KIR2DL-3, which blocks the KIR-ligand interaction and thereby augments killing of autologous tumor cells.115,116 In a phase I clinical trial in 32 relapsed/refractory MM patients, IPH2101 was well tolerated and no evidence of autoimmunity was observed. In the highest dose > 90% KIR occupancy was achieved and enhanced ex vivo patient-derived NK cell cytoxicity against MM was observed.117 IPH2101 is currently investigated in a phase II trial in smoldering MM patients (NCT01248455, NCT01222286). Besides, it is investigated in patients in stable partial response after first line therapy (NCT00999830) and in combination with lenalidomide in relapsed MM. (NCT01217203).
Another interesting antibody which can augment NK cell-mediated immunity against MM is elotuzumab. Elotuzomab (i.e. HuLuc63) is a humanized antibody against CS1 (i.e. CD2 subset-1, CRACC, SLAMF7, or CD319), a cell surface glycoprotein that is highly expressed on MM cells and normal plasma cells.118 In addition to expression on MM cells, CS1 is also expressed on NK cells, NKT cells, CD8+ T cells, activated monocytes and DCs, but to a lesser extent. Notably, CS1 is not expressed by healthy hematopoietic stem cells, other hematological malignancies or non-lymphoid tissues. Although the function of CS1 is not fully known, on MM cell is seems to interact with cell adhesion molecules on bone marrow stromal cells.118,119 On NK cells, CS1 appears to serve as an activator via the adaptor protein Ewing’s sarcoma-activated transcript-2 (EAT-2).120 Elotuzumab performs its anti-MM effects mainly via induction of NK cell-mediated ADCC.118 In ex vivo experiments ADCC against primary MM cells, resistant to conventional or novel therapies like bortezomib, could be enhanced using elotuzumab.119 A phase I study in relapsed/refractory MM patients showed acceptable toxicity and disease stabilization in 26.5% of patients.121 Furthermore, two phase I studies investigating elotuzumab, in combination with either lenalidomide and low-dose dexamethason, or bortezomib, in relapsed or refractory MM patients showed objective response rates of 82% and 48%, including a response in 2 out of 3 bortezomib resistant patients in the study combining elotuzumab with bortezomib.122,123 Currently, phase I/II trials combining elotuzumab with bortezomib/dexamethasone in newly diagnosed and relapsed MM patients are being performed (NCT01668719, NCT01478048). Additionally, phase III trials are performed investigating elotuzumab in combination with lenalidomide and dexamethasone in newly diagnosed patients (ELOQUENT-1, NCT018916430) and relapsed/refractory patients (ELOQUENT-2, NCT01239797).
As described earlier novel anti-myeloma drugs including the IMiDs and proteasome inhibitors have improved outcome of MM patients in the last decade.5 Interestingly, these drugs seem to confer their effects, at least partly, trough stimulation of NK cell-mediated killing.
Thalidomide was the first drug of the IMiD group that was used, though later more potent and less toxic IMiDs like lenalidomide have been introduced.124 IMiDs mediate their anti-myeloma effects via several mechanisms. They can directly kill MM cells by induction of cell cycle arrest and caspase-dependent apoptosis,125 decrease binding of MM cells to bone marrow stromal cells,126 block angiogenesis127 and inhibit production of cytokines (IL-6, TNF-α).128 Furthermore NK cell-mediated cytotoxicity can be augmented by these drugs.129 Hayashi et al.130 reported that IMiDs stimulate NK cells via increased IL-2 production by T cells. Additionally in vitro studies have shown that pretreatment of NK cells with lenalidomide can enhance ADCC against CD40-expressing MM cells if NK cells are combined with SGN-40, an anti-CD40 monoclonal antibody.131 Pomalidomide, a novel IMiD, which has been approved by the FDA recently, significantly increases serum IL-2 receptor and IL-12 levels.132 These cytokines are important for NK cell activation and could contribute to increased NK cell-mediated killing.
Proteasome inhibitors are a second group of anti-myeloma drugs. These drugs interact with the ubiquitin-proteasome pathway, which is responsible for the degradation of the majority of regulatory proteins in eukaryotic cells, including proteins that control cell-cycle progression, apoptosis, and DNA repair.133 Inhibition of proteasome activity results in growth arrest and cell death due to the induction of an apoptotic cascade as a result of rapid accumulation of incompatible regulatory proteins within the cell.134 Malignant cells show increased sensitivity to proteasome inhibition, probably as a result of higher proteasome activity. Therefore, the proteasome is one of the most interesting therapeutic targets in oncology, and it is currently an important component of MM treatment.135 Beside direct apoptotic effects on MM cells, treatment of MM cells with bortezomib can induce sensitization to NK cell-mediated lysis via several mechanisms. By downregulation of cell-surface expression of HLA class I molecules on MM cells, NK cell activation is no longer inhibited, since HLA class I is an important inhibitory ligand for NK cells.136 In addition, bortezomib can induce expression of NK cell activating ligands like Tumor Necrosis Factor-related apoptosis-inducing ligand (TRIAL)-R2,137,138 DNAM-1 ligands (poliovirus receptor, Nectin-2, CD112, CD155) and NKG2D ligands (MICA/B, ULBPs).139,140 On the downside a dose-dependent suppression of NKp46-mediated and TRAIL-mediated cytotoxicity, as well as induction of apoptosis in NK cells treated with bortezomib has been described.141,142 These findings indicate that low dose bortezomib may support NK cell-based immunotherapy, while high dose bortezomib may disturb NK cell-mediated cytotoxicity.
A third group of anti-myeloma drugs are the histone deacetylase (HDAC) inhibitors. HDACs play an important role in epigenetic modulation of gene expression and alterations in HDAC expression have been found in many types of cancers, making these enzymes a attractive targets for cancer therapy.143 Clinical phase II and phase III studies using the HDAC inhibitor vorinostat in relapsed MM patients have gained promising results.144,145 Another HDAC inhibitor, panobinostat, is currently investigated in a large phase III clinical trial (NCT01023308). In vitro and in vivo studies have reported NK cell-activating effects of HDAC inhibitors. Upon exposure to HDAC inhibitors, increased expression of MICA/B on lymphoma, leukemia, and hepatocellular carcinoma cells was observed.146-148 Another in vitro study reported increased DNAM-1- and NKG2D-dependent NK cytotoxicity.149 However, decreased expression of the activating NKp30 ligand B7-H6 has been observed in cell lines treated with HDAC inhibitors, reducing NKp30-dependent tumor cell recognition by NK cells.150 Additionally, in mice treated with HDAC inhibitors, diminished expression of NKG2D and NKp46 on, and decreased IFN-γ production by NK cells was observed. Therefore, it is essential to obtain data on the influence of HDAC inhibitors on NK cell-mediated killing in clinical studies in order to draw definite conclusions on the benefit of these drugs in NK cell-mediated myeloma therapy.
Survival of MM patients has improved significantly due to the introduction of new therapies like bortezomib and lenalidomide, however OS survival is still poor. Therefore additional potent therapeutic strategies are urgently needed. Currently, several cellular immunotherapeutic strategies are being developed to improve anti-myeloma immunity with reduced side effects. NK cell-based adoptive immunotherapy has shown promising clinical effects, without induction of severe adverse effects. Notably combination of NK cell adoptive transfer with antibodies like elotuzumab and KIR-blocking antibodies, anti-myeloma drugs, or even anti-PD1 antibodies may further enhance anti-myeloma immunity. Furthermore, new strategies like genetic engineering hold great potential for the development of NK cells with superior tumor-reactive functionality. For instance, inhibitory KIR receptors might be specifically downregulated in NK cells by the use of siRNAs. Another strategy would be the introduction of CARs on NK cells. Recently, Chu et al.151 generated a CS1-specific CAR which could be expressed by a human NK cell line. These NK cells showed improved killing of CS1-expressing MM cells in vitro and in vivo. Therefore, it would be highly interesting to explore such MM-specific CAR expressing NK cells in a clinical trial.
Another interesting strategy, which can improve T cell-mediated killing of MM is the application of bispecific antibodies. Blinatumomab, a CD3/CD19 bispecific single chain antibody, can redirect T cells to CD19 expressing B-lineage acute lymphoblastic leukemia (ALL) cells. In a phase II study, 16 out of 21 patients with MRD or relapse after therapy were MRD negative after blinatumomab treatment.152 Based on the successes in B-ALL, development of bispecific antibodies against MM-specific molecules, like CS1 or CD138, would be an interesting strategy to explore for MM treatment.
Taken together, ongoing efforts to improve therapy for MM have resulted in the development of several promising therapies, these therapies will potentially reach the clinic in the coming years. An important question is how to combine and sequence these new therapies in the initial treatment plan of patients with MM. The combination of cellular immunotherapy with other therapies has the potential to maximize the anti-myeloma effect without significant increased toxicity. Further research will have to elucidate prognostic factors predicting therapeutic responses and side effects in patients, enabling more effective personalized treatments and better survival and quality of life for future MM patients.
No potential conflicts of interest were disclosed.