Dendritic Cell Therapies for Hematologic Malignancies.

Dendritic cells (DCs) are potent antigen-presenting cells that constitute a major component of the immune system's role in the recognition, elimination, and tolerance of cancer. The unique immunologic capabilities of DCs have recently been harnessed for therapeutic use with the creation of DC-based anti-tumor vaccines, several of which have moved into testing in clinical trials for hematologic malignancies. This review summarizes how treatment strategies using DC-based anti-tumor vaccines are advancing immunotherapeutic options for these diseases.

Main Text

Dendritic cells (DCs) consist of a network of immunomodulatory cells with the ability to present antigen to T lymphocytes and to induce powerful antigen-specific primary immune responses. DCs may exert an anti-neoplastic effect by processing and presenting tumor antigen to autologous effector lymphocytes and thereby stimulating tumor-specific cytotoxicity. The presence of increased numbers of intratumoral DCs has been associated with improved clinical outcomes in human patients with non-small cell lung cancer and colorectal cancer, among other malignancies.2, 3 However, DCs can also exert an opposing, immunosuppressive, and tumor-sustaining effect in the context of malignancy, either by increased surface expression of immune checkpoint proteins such as PD-1 or by the elaboration of tolerogenic substances such as indoleamine 2,3-dioxygenase. In this sense, DCs also participate in the fundamental immune tolerance of many tumors, as they are eventually able to evade the regulatory function of the immune system and grow without restriction. The field of cancer vaccines seeks to reverse such tumor-mediated immune tolerance by fostering the development of clinically meaningful tumor-specific immunity.

Dendritic Cell Tumor Vaccine Production: Rationale and Technical Considerations

The use of DCs as a platform for cancer vaccine development is based on their unique potency as antigen presenting cells with the capacity to induce a primary immune response. Mature DCs constitutively express co-stimulatory molecules, such as CD80 and CD86, which facilitate T lymphocyte immunoreactivity. Additionally, DCs have the ability to participate in “cross-presentation”, in which exogenous antigen is presented to CD8-positive cytotoxic T lymphocytes via the major histocompatibility complex (MHC) class I molecule. By preferentially engaging in cross-presentation, DCs are able to exert a direct cytotoxic effect on exogenous antigens expressed by tumor cells. As a result, DCs represent a potentially critical platform to stimulate tumor specific immunity.

Fundamental issues regarding optimization of the DC model for tumor vaccination include: (1) choosing ideal single tumor antigen targets, (2) selecting the appropriate strategy for loading of single tumor antigens onto DCs, and (3) determining the role for multiple tumor antigens and/or whole-cell approaches. Each of these aspects of DC-tumor vaccine production will be discussed briefly below.

Single Tumor Antigen Identification

The selection of an appropriate target single tumor antigen is critical for the development of a vaccine strategy that preserves tumor specificity and immunologic efficacy. Common shared tumor antigens that have been explored in this setting include proteins/peptides otherwise expressed only during embryonic development (e.g., cancer testis antigens such as NY-ESO-1 and SP17),8, 9 peptides aberrantly or preferentially expressed by malignant cells (e.g., MUC1 in acute myelogenous leukemia and multiple myeloma or BCMA, which is selectively expressed by B-lymphocytes and plasma cells), or antigens truly unique to the tumor cell, such as the idiotype protein arising from the variable region of the immunoglobulin gene. Additionally, intense research has recently focused on the use of tumor neoantigens—those generated by somatic alterations in the genomes of cancer cells as they acquire neoplastic characteristics—as an antigenic source for DC-tumor vaccination.

Cancer testis antigens serve as attractive platforms for DC-tumor vaccine creation because of their limited expression on normal tissues and high expression by malignant hematologic cells, as well as the ease with which their mRNA can be incorporated into autologous DCs via electroporation. Liggins et al. have demonstrated that mRNA from at least eight cancer testis antigen genes—SP17, PRAME, CSAGE, PASD1, CAGE/DDX53, CTAGE1, HAGE/DDX43, and PLU-1/JARD1B—is expressed across numerous human B- and T cell lymphoma cell lines. A similar observation has been made regarding the expression of cancer testis antigens of the MAGE and SSX families in bone marrow biopsy specimens of human patients with multiple myeloma.15, 16, 17 Expression of such cancer testis antigens may vary across different disease states. For example, expression of the cancer testis antigens NY-ESO-1 and MAGEA3/A6 increases on leukemic blasts following treatment with the hypomethylating agent decitabine in human patients with acute myelogenous leukemia. Likewise, expression of the cancer testis antigen NXF2 is increased on neoplastic lymphocytes after the administration of decitabine in patients with chronic lymphocytic leukemia. Such variability in cancer testis antigen expression over time and following specific therapy has important implications not only for the choice of which of these antigens are best suited for incorporation into a DC-tumor vaccine, but also for the most appropriate timing of such vaccination. Several cancer testis antigen vaccines have already been tested in clinical trials for the hematologic malignancies, with variable degrees of immunologic and clinical success (Table 1). Additional cancer testis antigen-DC vaccines are currently in development (Table 2).

Table 1

Selected Clinical Trials of DC-Based Vaccines for Hematologic Malignancies

Disease	Number of Patients	DC Source	Tumor Antigen	Antigen Loading	Route of Administration	Immunologic Findings	Clinical Findings	Reference
Follicular lymphoma	35	PBMCs	tumor idiotype	ex vivo pulse (co-culture)	intravenous	65% anti-idiotype response	22% regression of residual disease70% without tumor progression at 43 months	50
Follicular lymphoma (relapsed)	18	peripheral blood monocytes	heat-shocked, irradiated tumor cells	ex vivo co-culture	subcutaneous, close to axillary and inguinal lymph nodes	objective clinical responses were associated with reduction in Tregs and increase in NK cells	33% objective clinical response (16.7% complete response, 16.7% partial response)44% stable disease	59
CML	3	PBMCs	NA	ex vivo	intradermal	66% with delayed type hypersensitivity reaction to DCs	33% with measurable anti-leukemic response at 20 months	60
CML	6	peripheral blood monocytes	NA	ex vivo	subcutaneous	increase in T lymphocyte immunogenicity	no clinical responses	61
CML	10	peripheral blood monocytes	NA	ex vivo	subcutaneous	30% with expansion of T cells with specificity for leukemia-specific antigens	40% with cytogenetic/molecular response	62
CLL	12	PBMCs	leukemia cell lysate	ex vivo co-culture	intradermal, close to axillary and inguinal lymph nodes	33% with increased in CD8+ T lymphocytes against leukemia-associated antigens increase in IL-12 and decrease in Tregs noted to patients with clinical response	41.7% with decreased peripheral blood leukemia cells25% with stable disease33% with disease progression	64
CLL	15	PBMCs	apoptotic tumor bodies	ex vivo co-culture	intradermal and intravenous	66% with leukemia-specific immune response	no objective clinical responses	65
CLL	9	PBMCs (NB: allogeneic source)	leukmia cell lysate, tumor apoptotic bodies	ex vivo co-culture	intradermal, close to axillary and inguinal lymph nodes	11.1% with expansion of cytotoxic T lymphocytes against leukemia-associated antigen	decrease in amount of circulating CLL cells in all patients	63
ATLL	3	PBMCs	tax peptide	ex vivo co-culture	subcutaneous	tax-specific cytotoxic T lymphocyte response in all patients	two patients with partial remission in first 8 weeks, one with subsequent complete remission	66
Multiple myeloma	12	PBMCs	tumor idiotype	ex vivo pulse	intravenous, with subcutaneous idiotype-KLH boosters	16.7% anti-idiotype proliferative immune response	16.7% with anti-idiotype proliferative immune response in complete remission at minimum follow-up of 16 months	67
Multiple myeloma	26	PBMCs	tumor idiotype	ex vivo pulse	intravenous	15.4% anti-idiotype proliferative immune response	65% alive at median follow-up of 30 months	73
Multiple Myeloma	27	PBMCs	tumor idiotype	ex vivo pulse	intravenous	NA	median overall survival 5.3 years (compared to 3.4 years in non-randomized control group)	75
Multiple myeloma	12	PBMCs	mRNA from MAGE3, Survivin, and BCMA	ex vivo pulse and electroporation with mRNA	intravenous and intradermal	16.7% vaccine-specific T lymphocytes	83% overall survival at 55 months (50% of those alive with stable disease)	76
Multiple Myeloma	18	PBMCs	whole tumor cell	ex vivo DC-tumor cell fusion	subcutaneous	73.3% with expansion of circulating myeloma-reactive T cells	68.9% with stable disease after vaccination	77
Multiple myeloma	36	PBMCs	whole tumor cell	ex vivo DC-tumor cell fusion	subcutaneous	all evaluable patients with at least 2-fold expansion of myeloma-specific T lymphocytes	47% CR/nCR78% CR/VGPR24% with PR converted to a CR after vaccination	78
AML	10	PBMCs	mRNA from WT1	ex vivo pulse and electroporation with mRNA	intradermal	vaccinated patients showed increased levels of WT1-specific CD8+ T lymphocytes	50% with molecular CR20% with PR to a CR converted to a CR after vaccination	85
AML	17	PBMCs	whole tumor cell	ex vivo DC-tumor cell fusion	subcutaneous	5.4-fold increase in AML-specific CD4+ T cells, and 15.7-fold increase in AML-specific CD8+ T cells	71% alive without AML recurrence at median follow-up of 57 months	89

CLL, chronic lymphocytic leukemia; KLH, keyhole limpet hemocyanin; NK, natural killer; CR, complete response; nCR, near complete response; VGPR, very good partial response; PR, partial response; AML, acute myelogenous leukemia; PBMCs, peripheral blood mononuclear cells.

Table 2

Currently Enrolling Clinical Trials of Autologous Dendritic Cell-Tumor Vaccines for the Hematologic Malignancies

Disease	Setting	Trial Design	Antigen Loading	Clinical Trials.Gov	Reference
Multiple myeloma	primary therapy, following autoSCT	vaccine and lenalidomide versus lenalidomide alone	whole cell DC-tumor fusion	NCT02728102	79
Multiple myeloma	primary therapy, prior to and following autoSCT	vaccine alone	adenovirus vector to load Survivin	NCT02851056	81
Multiple myeloma	primary therapy, following autoSCT	vaccine and lenalidomide	electroporated mRNA for CT7, MAGE-A3, and WT1	NCT01995708	82
AML	primary therapy, following induction and consolidation chemotherapy	vaccine alone	electroporated mRNA for WT1	NCT01686334	86
AML	primary therapy, following induction chemotherapy	vaccine alone	two leukemia antigens and one CMV antigen	NCT01734304	87
AML	primary therapy, following induction chemotherapy	vaccine alone	electroporated mRNA for WT1 and PRAME	NCT02405338	88

autoSCT, autologous hematopoietic stem cell transplantation; SC, subcutaneous; CMV, cytomegalovirus; WT1, Wilms tumor protein; CT7, cancer testis 7 protein; MAGE-A3, melanoma-associated antigen gene-A3 protein; LMP-2, latent membrane protein 2.

Tumor Antigen Loading

Loading strategies of single tumor antigens for DC-tumor vaccine production in the treatment of hematologic malignancies consist of in vivo and ex vivo methods. In vivo DC antigen loading for tumor-specific vaccination has been accomplished by using nanoparticles coated with antibodies specific to DC surface markers as the vehicle by which to carry tumor antigen and adjuvant immunostimulatory molecules to DCs.20, 21 This approach has several advantages compared to ex vivo manipulation, including decreased cost, improved ease of administration, and applicability in a variety of clinical practice environments, as well as the potential to target more DCs and different DC subsets than those available during ex vivo production.

By contrast, the ex vivo method of DC-tumor vaccine generation allows for more standardization and control of antigen loading onto DCs, as well as the ability to ensure that autologous DCs have fully matured prior to vaccination and thus have reached their complete immunogenic capability. DC maturation can be stimulated ex vivo by exposing CD14-positive peripheral blood monocytes in culture to various cocktails of cytokines, including interleukin (IL)-1β, IL-6, tumor necrosis factor alpha (TNF-α), and PGE2, or, more commonly, granulocyte-macrophage colony-stimulating factor (GM-CSF), IL-4, and TNF-α. For this reason, monocyte derived precursor populations have been the most common source for autologous DC tumor vaccine production.

However, the ontogeny of human DCs is complex and has yet to be elucidated fully. Myeloid/classic DC, plasmacytoid DC, monocyte-related DC, and Langerhans cell subtypes exert divergent biological effects as antigen presenting cells, including differential capacity for cross presentation, cytokine production, and polarization toward Th1, Th2, Th17, or regulatory T cell (Treg) immunophenotypes under the influence of various microbiological and biochemical stimuli.25, 26, 27, 28 Each of these alternative DC subtypes, including Langerhans cells,29, 30, 31 DCs derived from CD34-positive umbilical cord blood progenitor cells,32, 33, 34, 35 and plasmacytoid DCs36, 37, 38, 39 have been explored for use in DC vaccine development in different solid and hematologic malignancies. Further pre-clinical and clinical studies are necessary to determine precisely how the choice of DC subtype for vaccine production will impact vaccine yield and tumor-specific immunogenicity.

The Role of Multiple Tumor Antigens and Whole-Cell Approaches

A potential limitation of single antigen DC vaccines is their susceptibility to immune evasion due to the downregulation of antigen expression. Alternatively, loading of antigen derived from autologous whole tumor cells allows for the induction of a polyvalent response that can potentially target shared and patient-unique neoantigens. Whole-cell techniques for vaccine production include creation of DC-tumor fusion cells, as well as loading of DCs with cellular lysates,40, 41 apoptotic bodies,42, 43 tumor exosomes,44, 45 or whole-cell DNA or RNA. Galea-Lauri et al. have determined that DC-tumor whole-cell fusions induce a more potent cytotoxic T-lymphocyte response against human acute myelogenous leukemia (AML) cells in the in vitro setting than those DC vaccines created with tumor cell lysates or apoptotic bodies. Nevertheless, further research into the most appropriate whole-cell approach for tumor antigen loading remains necessary.

Our group has focused on using a whole-cell DC-tumor fusion approach for the creation of autologous DC vaccines in AML and multiple myeloma (Figure 1). Using this technique, patient-derived tumor cells are fused to autologous ex vivo-generated DCs by co-culture in the presence of polyethylene glycol. These DC-tumor cell fusions have several distinct immunologic advantages for vaccination in that they present numerous shared antigens and tumor neoantigens to immune effector cells and do so through both the MHC class I/CD8 (cytotoxic T lymphocyte) and MHC class II/CD4 (helper T lymphocyte) pathways. This vaccine platform has proved to be successful in early phase clinical trials for AML and multiple myeloma, with larger phase III clinical trials to be performed in the near future.

Figure 1

Production of Dendritic Cell-Myeloma Cell Fusion Vaccine

Key: The production of our dendritic cell-myeloma cell fusion vaccine begins with leukapheresis of the patient’s peripheral blood mononuclear cells (PMBCs). PBMCs that are adherent to plates are then cultured with the cytokines IL-4, GM-CSF, and TNF-α, which induces DC differentiation and maturation with a characteristic immunophenotype (positive for CD86, CD80, CD40, and CD83). In parallel, bone marrow biopsy yields myeloma tumor cells, which are assessed for expression of myeloma-specific cell surface markers (CD38, CD138, and MUC1). The autologous mature DCs are then fused with the myeloma cells by co-culture in the presence of polyethylene glycol (PEG). The resultant fusion cells are quantified by co-expression of both DC and myeloma surface markers via flow cytometry and immunohistochemisty. The fusion vaccine is then prepared and frozen and is administered subcutaneously to the patient along with GM-CSF at the vaccine site for 4 days.

DC-Tumor Vaccines in Clinical Trials for Hematologic Malignancies

DC-tumor vaccination has been explored in clinical trials for several hematologic malignancies, including indolent non-Hodgkin’s lymphoma, chronic myelogenous leukemia (CML), chronic lymphocytic leukemia (CLL), adult T cell leukemia/lymphoma), multiple myeloma, and AML (Table 1). Several notable examples of these studies are discussed below. While biologic potency has been demonstrated in many instances, the clinical efficacy of the DC-tumor vaccine treatment strategy for the hematologic malignancies continues to be investigated in ongoing studies (Table 2).

Indolent Non-Hodgkin’s Lymphoma

In an initial trial of a DC vaccine pulsed with tumor-specific idiotype for follicular lymphoma, a subset of patients developed a measurable anti-tumor immune response to the vaccine, and one of the patients experienced complete tumor regression. In a subsequent larger study by the same group, 35 patients with follicular lymphoma underwent DC vaccination using the tumor idiotype vaccine platform following standard chemotherapy. Immunologic response and regression of residual disease were noted in 65% and 22% of patients, respectively. Of note, 70% of patients in this study remained without tumor progression at a median follow-up of 43 months.

Idiotype-based vaccines subsequently proved successful in a number of phase I and II clinical trials for follicular lymphoma,51, 52, 53, 54, 55 but did not reach pre-specified clinical efficacy endpoints in three separate large phase III studies.56, 57, 58 Importantly, however, all of these larger trials administered tumor idiotype alone as their method of vaccination, without autologous DCs. Therefore, further development and clinical study of idiotype vaccines against follicular lymphoma using a DC platform may still be warranted.

Whole tumor cell techniques of antigen loading have also been studied in clinical trials of follicular lymphoma. For example, a pilot study of 18 patients with relapsed indolent follicular lymphoma demonstrated that vaccination with DCs loaded with autologous tumor cells that were heat-shocked and irradiated was associated with objective clinical response and stable disease in 6 (33%) and 8 (44%) patients, respectively. Clinical responses were significantly associated with a reduction in Tregs and an increase in natural killer (NK) cells. This whole-cell approach to DC vaccine antigen loading against follicular lymphoma, therefore, also merits further clinical study.

Chronic Myelogenous Leukemia

Because of the success of immunologic approaches to the treatment of CML, namely the striking effectiveness of donor lymphocyte infusion following allogeneic hematopoietic stem cell transplantation, it was hypothesized that additional immunotherapeutic techniques may prove to be beneficial in inducing disease control for patients with CML. In 2003, Ossenkoppele et al. reported the results of a pilot study of three patients with CML who were administered DCs that were obtained from autologous peripheral blood mononuclear cells. There were two of these patients that developed delayed type hypersensitivity reactions to CML-derived DCs, and one patient (33%) was found to have a sustained, leukemia-specific response after 20 months of follow up. There were six patients that were subsequently studied in a larger phase I trial of a DC-based vaccine for CML. Although no clinical responses were noted in this study, an increase in T cell immunogenicity to CML was observed. This spurred a further trial of DC vaccination, in which ten patients with chronic-phase CML were treated with DCs that had been harvested from autologous peripheral blood monocytes. In this series, three patients (30%) were found to have an expansion of T lymphocytes with specificity for leukemia-specific antigens and cytogenetic/molecular response was noted in four patients (40%). DC-based vaccination strategies therefore may represent a promising treatment option for CML in the future, particularly for patients with minimal residual disease following treatment with a potent BCR-ABL tyrosine kinase inhibitor.

Chronic Lymphocytic Leukemia

Because the clonal lymphocyte population in CLL represents the overwhelming preponderance of circulating nucleated cells, harvesting of DCs via leukapheresis of peripheral blood mononuclear cells can be technically challenging in patients with this disease. It is for this reason that allogeneic DCs—as opposed to autologous DCs, which have been used in the vast majority of other studies of DC-tumor vaccines for hematologic malignancies—were first used for the creation of a DC-based tumor vaccine against CLL. In a small pilot study of nine patients with early stage CLL (Rai 0-1), allogeneic DCs were obtained from healthy, unrelated donors and were then exposed to the study subjects’ CLL tumor lysate and tumor apoptotic bodies. Administration of the resulting allogeneic DC-tumor vaccine led to a decrease in the amount of circulating CLL cells across all of the patients. Furthermore, one patient (11.1%) developed expansion of cytotoxic T lymphocytes directed against a leukemia-associated antigen.

Autologous DC-based vaccination against chronic lymphocytic leukemia was subsequently reported by Hus et al. in 2008. In this study of 12 patients with early stage CLL (Rai stage 0–2), DC vaccine was produced by isolation of peripheral blood mononuclear cells, which were then co-cultured with leukemia cell lysates. Following vaccination, five patients (41.7%) demonstrated a decrease in the number of peripheral blood leukemia cells, three patients (25%) demonstrated stable disease, and four patients (33%) experienced disease progression. There were four patients (33%) that were found to have increased numbers of CD8-positive T lymphocytes directed against leukemia-specific antigens, and patients with clinical response were noted to have increased levels of the immunostimulatory cytokine IL-12, as well as decreased numbers of immunosuppressive Tregs. Palma et al. subsequently reported a cohort of 15 patients with CLL who were treated with a DC vaccine that was produced with tumor apoptotic bodies. Although no objective clinical responses were noted in this study, ten patients (66%) developed leukemia-specific immune responses. The encouraging results of these early-phase trials indicate that DC-based tumor vaccination strategies may prove to have a role in the therapeutic armamentarium against CLL in the future.

Adult T Cell Leukemia/Lymphoma

In 2015, Suehiro et al. reported an early phase clinical trial of a DC-tumor vaccine for patients with ATLL. For this study, a DC-tumor vaccine was created by pulsing autologous DCs with Tax, a peptide product of human T cell leukemia virus type-I (HTLV-1), the causative viral agent of ATLL. Tax-specific cytotoxic T lymphocyte responses were identified in all three patients in this trial. Two of three patients experienced a partial remission within the first 8 weeks, and one of these patients converted to a complete remission subsequently. These two patients remained in remission 24 and 19 months after vaccination, with no need for further chemotherapy. The encouraging clinical outcome in this trial indicates the need for further study of this promising immunotherapy for ATLL.

Multiple Myeloma

The initial feasibility study of an idiotype-pulsed DC vaccine for the treatment of patients with multiple myeloma following autologous peripheral blood stem cell transplantation was reported by Reichardt et al. from Stanford University in 1999. In this series, 12 patients were administered two intravenous infusions of idiotype-pulsed autologous DCs, with 5 monthly subcutaneous “boosters” of idiotype, keyhole limpet hemocyanin (KLH), and immune adjuvant. There were two patients (16.7%) that developed an idiotype-specific cellular immune response, and these patients remained in complete remission at a minimum follow-up of 16 months. Similar results were subsequently reported by groups in Italy, Wales, Australia, Germany, and Arkansas, as well as within a larger cohort of patients from Stanford. An idiotype-pulsed DC vaccine has also been demonstrated to lead to idiotype-specific T lymphocyte expansion in patients with earlier stages of myeloma.

In a subsequent phase II study performed at the Mayo Clinic, 27 patients with multiple myeloma who underwent consolidation therapy with autologous hematopoietic stem cell transplantation received idiotype-pulsed DCs (APC8020, Mylovenge) as post-transplantation adjunctive therapy between July 1998 and June 2001. These patients were compared to a non-randomized, parallel group of control patients who underwent autologous hematopoietic stem cell transplantation for myeloma during that time, but were not vaccinated with APC8020. Comparison of these two groups revealed a statistically and clinically significant improvement in overall survival in the vaccine group (5.3 years) compared to the unvaccinated group (3.4 years, p = 0.02).

Others have reported a cancer testis antigen mRNA antigen-loading strategy for DC-myeloma vaccination. In a phase I study, Hobo et al. obtained autologous monocyte-derived DCs from 12 patients with multiple myeloma who were treated with induction chemotherapy and high-dose melphalan with autologous hematopoietic stem cell transplantation. These DCs were pulsed with KLH and were electroporated with mRNA from the cancer testis antigen MAGE3, as well as SURVIVIN and B cell maturation antigen (BCMA). The DC-mRNA-loaded vaccines were then re-administered to the patients intravenously and intradermally. Two patients (16.7%) in this study developed vaccine-specific T lymphocyte responses.

Our group has developed a DC-tumor fusion vaccine model, whereby patient-derived myeloma cells are fused to ex vivo-generated DCs. In a phase I study of 18 patients with advanced disease, vaccination resulted in the expansion of CD4-positive and CD8-positive myeloma-reactive T lymphocytes in 11 of 15 evaluable patients, with the majority of patients experiencing disease stabilization. In a phase II trial of 36 subjects evaluating the DC-myeloma vaccine following autologous hematopoietic stem cell transplantation, 47% and 78% of patients experienced a complete response/near complete response and complete response/very good partial response, respectively. Importantly, 24% of patients with a partial response following autologous hematopoietic stem cell transplantation were converted to complete response after DC-tumor fusion cell administration in the absence of any other therapy, consistent with the possibility that the vaccine targeted post-transplant residual disease. Based on these data, a randomized trial of DC-tumor fusion vaccination for multiple myeloma with lenalidomide maintenance versus lenalidomide maintenance alone is being conducted under the auspices of the CTN cooperative group (clinicaltrials.gov identifier: NCT02728102). DC-myeloma fusion vaccination is also being studied in conjunction with blockade of PD-1, in an effort to augment vaccine effectiveness by immune checkpoint inhibition.

Several other trials of DC-myeloma vaccines are currently open to enrollment (Table 2), including those using an adenovirus vector to load the SURVIVIN antigen onto autologous DCs (clinicaltrials.gov identifier: NCT02851056), and those using Langerhans cells electroporated with the mRNA of the cancer testis antigens CT7 and MAGE-A3, as well as WT1, for antigen loading (clinicaltrials.gov identifier: NCT01995708).

AML

Various methods of antigen selection and loading have been employed for the creation of DC vaccines for therapeutic use in AML. These have included whole tumor cell-DC fusions, mRNA coding for particular tumor antigens, apoptotic tumor bodies, and differentiation of leukemic blasts into autologous DCs. Van Tendeloo et al. developed a DC vaccine against AML by electroporating autologous DCs with the mRNA of WT1, an oncogene that is expressed in most cases of AML. In a phase I/II study of ten patients with AML, administration of this vaccine led to increases in WT1-specific T lymphocyte proliferation and WT1-specific interferon-γ-producing CD8-positive T lymphocytes. Furthermore, molecular remission was induced in five of these patients (50%), and two patients who were in a partial remission were converted to complete remission after vaccination. Several clinical trials using electroporation of specific mRNAs as the method of DC vaccine construction against AML are currently ongoing (Table 2).86, 87, 88

In a study of 17 patients with AML who were not candidates for allogeneic hematopoietic stem cell transplantation and achieved first complete remission after induction chemotherapy, our group demonstrated that vaccination with autologous DC/AML whole fusion cells induced the expansion of leukemia specific CD4-positive and CD8-positive T lymphocytes. Remarkably, 12 of the patients (71%) remained alive without AML recurrence after a median follow-up of 57 months. The encouraging results of this trial will lead to a randomized, phase III trial of this DC-AML whole-cell fusion vaccine in the near future.

Challenges and Future Directions: Augmenting and Standardizing the DC-Tumor Vaccine Response

In spite of the therapeutic excitement of DC-tumor vaccines against the hematologic malignancies, significant challenges to their widespread adoption in clinical practice remain. From a technical perspective, standardization of DC-tumor vaccine preparation is required for studies evaluating the relative potency of different approaches. While increasingly sophisticated strategies to identify patient specific mutations/antigens for DC-tumor vaccine production have evolved, determining the immunologic relevance of such individual shared and neoantigens will be crucial. Furthermore, these newly identified antigens may show distinctive patterns of expression in different disease settings, with important implications for the use and timing of DC-tumor vaccines directed against those antigens.

Improvements in elucidating the complex interactions of the tumor microenvironment that induce effector T cell dysfunction and compromise migration and cytotoxic T lymphocyte mediated killing of tumor cells at the tumor bed are critical. While an individual disease setting may be dominated by a particular perturbation of immunity, hematologic malignancies typically manifest multiple areas of dysfunction that necessitate repair. DC vaccines may play a critical role in stimulating expansion of tumor-reactive lymphocytes, but other areas of immune dysregulation will need to be addressed in order for these cells to be maximally effective at the site of disease. Efforts to augment DC vaccine potency are now focusing on reversing these critical elements of the immunosuppressive milieu, including blockade of immunosuppressive accessory cells and reversal of alterations of biologic pathways that foster immune tolerance. Combining vaccination with immunomodulatory drugs such as lenalidomide or immune checkpoint blockade is being explored.91, 92 Future research efforts in this realm are also likely to include exploration of combinations of DC-tumor vaccination with various novel immunotherapy strategies, including chimeric antigen receptor (CAR) T lymphocyte therapy, myeloid-derived suppressor cell (MDSC) inhibitors, and therapies that deplete Tregs, as well as lentviral or retroviral gene therapy techniques for improved induction of anti-tumor immune response.96, 97

While the current generation of DC-based vaccinations has demonstrated promise against the hematologic malignancies, further refinements in vaccine strategies are clearly needed to develop this promising area of investigation into a clinically meaningful therapy for patients with blood cancers.