The unique potential efficacy of cellular immunotherapy for myeloma is highlighted by the observation that allogeneic transplantation induces durable remissions in a subset of patients due to the graft-versus-myeloma effect [50–53]. A summary of early data of myeloablative transplantation from the European Bone Marrow Transplant Registry demonstrated that 28 % of patients remained in remission 7 years posttransplant, suggesting that durable responses were potentially achievable [54]. However, the median survival was only 10 months due to extremely high treatment-related mortality, raising a difficult choice for physicians and patients as to the applicability of this strategy. In a more recent report of 158 patients undergoing autologous or allogeneic transplantation based on donor availability, the event-free survival (EFS) following allogeneic transplantation was 33 and 31 % at 5 and 10 years, respectively, consistent with the presence of a subgroup that appear to have sustained disease response [55]. The role of the graft-versus-myeloma effect in preventing disease recurrence was further supported by a retrospective report from the European registry, in which patients with limited or extensive chronic graft-versus-host (cGVHD) disease demonstrated markedly improved 3-year survival (84 and 58 %, respectively) as compared to those without cGVHD (29 %) [56].

Donor lymphocyte infusion (DLI) as a treatment for posttransplant relapse has been shown to induce disease response, achievement of molecular remission, reconstitution of TCR Vβ repertoire, and long-term disease control in a subset of patients [57–60]. However, DLI therapy is complicated by GVHD due to the lack of myeloma specificity of the alloreactive lymphocytes [61]. Efforts to limit toxicity through the use of reduced intensity conditioning regimens have resulted in a decrease in treatment-related mortality but a concomitant increase in the risk of relapse. As such, immune-based targeting of myeloma cells by alloreactive lymphocytes may carry the unique potential for curative outcomes; nonetheless, the lack of specificity and toxicity significantly limits its use.

Investigators have examined strategies to induce autologous cellular immune responses that selectively target myeloma-associated antigens while minimizing toxicity to normal tissue.

One such strategy is the use of cancer vaccines to foster the expansion of tumor-specific lymphocytes. Myeloma-associated antigens that have been explored as targets for immunotherapy include the idiotype protein, MUC1, WT1, PRAME, CYP1B1, and HSP96 [62–69]. Vaccine strategies have included the introduction of myeloma-specific antigens in the context of immune adjuvants and the loading of individual or whole-cell-derived antigens onto antigen-presenting cells such as DCs.

DCs represent a diverse network of antigen-presenting cells that play a prominent role in mediating immune responsiveness [70]. Circulating DC populations have been identified as myeloid and plasmacytoid in origin with the capacity to elicit Th1 and Th2 responses, respectively. Plasmacytoid DCs have been shown to contribute to the stromal environment in myeloma and may contribute to tumor-mediated tolerance [4]. Myeloma antigens administered in the context of immune adjuvants may recruit and activate native DC populations that subsequently internalize and present tumor antigens [71–73]. However, functional deficiencies have been demonstrated in DCs derived from myeloma patients which may impact their ability to elicit immunologic responses [5]. Alternatively, myeloid DCs with strong expression of costimulatory molecules and stimulatory cytokines may be generated ex vivo through cytokine stimulation of precursor populations [74]. DCs generated ex vivo and loaded with myeloma-associated antigens may act as a platform for cancer vaccines [75]. Strategies to introduce tumor antigens include pulsing with peptides, proteins, or lysates [76], electroporation with tumor-derived RNA or DNA [8, 77, 78], loading of tumor-derived apoptotic bodies [79], transduction with viral vectors expressing tumor antigens potentially enhanced by costimulatory molecules [80–83], and the use of whole-cell fusion between DCs and myeloma cells [84–86].

The idiotype protein represents a truly tumor-specific antigen created by the unique immunoglobulin gene arrangement intrinsic to the malignant clone [87]. Vaccination with the idiotype protein in conjunction with granulocyte-macrophage colony-stimulating factor (GM-CSF) or IL-12 was associated with antigen-specific T-cell responses. Prolonged disease-free progression was observed in patients exhibiting an immunologic response [88]. Responses have also been observed following vaccination with antigen-presenting cells pulsed with M protein or with DCs loaded with idiotype and exposed to CD40L to induce maturation [89–93]. Vaccination with idiotype-pulsed antigen-presenting cells posttransplant was associated with improved progression-free survival as compared to a historical control cohort.

A peptide-based vaccine for WT1 administered with immune adjuvant has been shown to elicit immunologic response in patients with hematological malignancies and a decrease in measures of disease [94, 95]. In a recent study, WT1-specific immunity following allogeneic transplantation for myeloma was associated with long-term disease control. Peptide-based vaccine for MUC1 is currently being explored in patients with myeloma (NCT01232712). Expression of several cancer-testis antigens has been demonstrated and has been shown to be targeted by donor-derived humoral responses following allogeneic transplantation, confirming their potential immunogenicity. The cancer-testis antigen, NY-ESO, demonstrates increased expression by plasma cells in the setting of advanced disease, creating an appealing target for immune-based therapy [96]. Repetitive stimulation with DCs pulsed with an NY-ESO-derived peptide elicits a strong CTL response in vitro, demonstrating an activated phenotype capable of lysing primary myeloma cells [97]. Recent studies have identified a series of antigens recognized by T cells in patients following syngeneic transplantation.

Several other peptides which are highly expressed on myeloma cells and are important in the pathogenesis of the disease have been identified as potential immunogenic targets. Heteroclitic XBP1 (X-box-binding protein 1) (unspliced 184–192 and spliced 367–375), CD138 (syndecan-1)260–268, and CS1239-247 were shown each alone and in a cocktail combination of the four to generate specific CTLs enriched for effector and activated T cells, Ag-specific cytotoxicity against MM cell lines, as well as increased degranulation, proliferation, and INF-γ secretion [98–101].

The use of whole-cell-derived antigens for vaccination may elicit a broad polyclonal response that is better able to target the heterogeneity of the myeloma cell population [86]. Consistent with this hypothesis, a murine model demonstrated the emergence of idiotype-negative variants following idiotype-based vaccinations, while whole myeloma cell-based vaccines did not induce resistance [102]. DCs pulsed with tumor lysates have been shown to induce myeloma-associated immunity, although the clinical efficacy was uncertain [76].

The authors have developed a vaccine model in which patient-derived myeloma cells are fused with autologous DCs, creating a hybridoma which expresses a broad array of myeloma antigens in the context of enhanced costimulation [86]. In a murine model, DC/MM fusions were shown to be protective against lethal challenge with syngeneic myeloma cells, and therapeutic efficacy was further enhanced by coadministration of IL-12 [103]. In preclinical human studies, fusion of DCs and MM cells elicited the expansion of activated T cells that potently lysed autologous myeloma cells in vitro.

A phase I clinical trial was completed in which successive cohorts of patients with advanced myeloma underwent vaccination with escalating doses of autologous DC/MM fusions [104]. Patients had undergone a median of four prior treatment regimens. Myeloma cells were derived from bone marrow aspirates, and DCs were generated from adherent mononuclear cells cultured with GM-CSF and IL-4 and matured with TNF-α. Patients underwent serial vaccination in conjunction with GM-CSF. Vaccine-associated toxicity consisted of transient grade 1–2 vaccine site reactions most commonly, while clinically significant autoimmunity was not observed. Biopsy of the vaccine bed demonstrated a dense infiltrate of CD8⁺ T cells consistent with T-cell expansion occurring at the site of vaccination. Vaccination resulted in the expansion of myeloma-specific T cells in the majority of patients as manifested by the percent of CD4⁺ and/or CD8⁺ T cells expressing IFN-γ following ex vivo exposure to autologous tumor lysate. On SEREX analysis, humoral responses against novel proteins were noted after vaccination. These findings were consistent with the induction of myeloma-specific immunity in patients with advanced disease. Of note, 66 % of patients demonstrated a period of disease stability ranging from several months to greater than 2 years after vaccination.

The authors have completed a phase II clinical trial in which patients underwent vaccination with DC/MM fusions in conjunction with autologous stem cell transplantation. It was postulated that vaccine response would be augmented following transplant-mediated cytoreduction and in the context of lymphopoietic reconstitution with the associated depletion of regulatory T cells. It was demonstrated that the posttransplant period was associated with the expansion of myeloma-reactive T cells which were further boosted by vaccination with DC/MM fusions. Vaccination was associated with the conversion of partial to complete responses greater than 100 days posttransplant in a subset of patients. A clinical trial is now underway examining the efficacy of PD-1 blockade in conjunction with the DC/MM fusion vaccine following autologous transplantation, and a national cooperative group study for the assessment of DC/MM fusion vaccine with lenalidomide versus lenalidomide maintenance alone is being planned.

Augmentation of NK cell-mediated immunity has been explored as therapy for MM. Preclinical models have demonstrated that thalidomide and lenalidomide increase the production of IL-2 by T cells, which stimulates NK cell activation and function against MM [105]. Lenalidomide has also been shown to increase CD16 and LFA-1 expression on NK cells, which facilitates an ADCC response against MM [106]. Lenalidomide also modulates the balance of NK cell-activating and inhibitory ligand expression on MM cells. It decreases expression of PD-L1 and enhances expression of ULBP-1(NKG2D ligand) on MM cells, which both result in improved NK cell immune response, as well as recognition and lysis of MM tumor targets [20, 107]. Bortezomib decreases MM expression of MHC class I and enhances the sensitivity of myeloma to NK cell-mediated lysis [108].

The importance of NK cell-mediated immunity in modulating disease outcome was highlighted by the observation that levels of autologous NK cells reinfused with autologous transplantation correlates with absolute lymphocyte recovery after ASCT for MM and non-hodgkin lymphoma [109]. Lymphocyte subset analyses revealed that an absolute NK cell count of 80/μL or more on day +15 post-SCT correlated significantly with improved progression-free survival [110]. In patients with MM undergoing allogeneic SCT, killer-cell immunoglobulin-like receptor (KIR)-ligand mismatch predicting for NK activation was protective against relapse [111]. In addition, the infusion of T-cell-depleted, haploidentical, KIR-mismatched NK cells, followed by delayed autograft stem cell rescue, has been shown to induce a near-complete/complete response rate of nearly 50 % [112]. Improved disease-free and overall survival was observed in myeloma patients who received grafts from donors with KIR haplotype B, which is associated with more activating receptor genes than KIR haplotype A [113]. Lenalidomide therapy for patients with progressive MM following allogeneic SCT has been associated with an overall response rate of 66 %, and immunomonitoring data show that lenalidomide augments NK cell expression of the activating receptor NKp44 [114]. Moreover, in a recent phase I/II study of lenalidomide given early after allogeneic SCT for MM, lenalidomide treatment resulted in an increase of activating receptors NKp30 and NKp44 on NK cells, as well as an increase in NK cell-mediated cytotoxicity directed against myeloma associated with an increase in the rate of complete remission [115].