Induction of Immune Response after Allogeneic Wilms’ Tumor 1 Dendritic Cell Vaccination and Donor Lymphocyte Infusion in Patients with Hematologic Malignancies and Post-Transplantation Relapse

Abstract

Relapse of hematologic malignancies is the primary cause of treatment failure after allogeneic hematopoietic stem cell transplantation (HCT). Treatment for post-HCT relapse using donor lymphocyte infusion (DLI) has limited utility, particularly in the setting of acute leukemia, and can result in the development of graft-versus-host disease (GVHD). The Wilms’ tumor 1 (WT1) gene product is a tumor-associated antigen that is expressed in acute leukemia and other hematologic malignancies, with limited expression in normal tissues. In this pilot trial, we assessed safety and feasibility of a WT1 peptide–loaded donor-derived dendritic cell (DC) vaccine given with DLI designed to enhance and direct the graft-versus-leukemia effect. Secondary objectives were to evaluate immunologic and clinical responses. A total of 5 subjects, median age 17 years (range, 9 to 19 years), with post-HCT relapse were enrolled. Disease subtypes included acute lymphoblastic leukemia (n = 3), acute myelogenous leukemia (n = 1), and Hodgkin lymphoma (n = 1). Successful vaccine production was feasible from all donors. DC vaccination and DLI were well tolerated. One recipient developed grade 1 skin GVHD not requiring systemic therapy. The most common adverse events included grade 1 reversible pain and pruritus at the vaccine injection and delayed-type hypersensitivity (DTH) skin testing sites. There were no grade 3 or higher adverse events related to the research. Immune responses consisted of ELISpot response in 3 recipients and positive DTH tests to WT1 peptide cocktail in 2 subjects. Our study provides 1 of the first attempts to apply tumor-specific vaccine therapy to the allogeneic setting. Preliminary results show the DC-based vaccination is safe and feasible after allogeneic HCT, with a suggestion that this approach can be used to sensitize the repopulated allogeneic-donor immune system to WT1. Future directions may include testing of vaccination strategies in the early post-transplantation setting for relapse prevention.

INTRODUCTION

Allogeneic hematopoietic stem cell transplantation (HCT) can be curative for many hematologic malignancies with efficacy due, in part, to an allogeneic graft-versus-leukemia (GVL) effect. However, relapse remains the primary cause of treatment failure after transplantation. Post-transplantation therapies, including donor lymphocyte infusions (DLI), have variable effectiveness in the treatment of post-transplantation relapse, with especially limited potency in the setting of acute leukemia [1,2]. Novel therapies to address post-transplantation relapse are needed [3,4].

Dendritic cells (DCs) are professional antigen-presenting cells that can be readily generated from peripheral blood monocytes. Tumor vaccine trials using purified populations of DCs as antigen-presenting cells have been reported [5,6]. The Wilms’ tumor 1 (WT1) gene product is a tumor-associated antigen that represents a potential target for immunotherapy in hematologic malignancies [7–11]. WT1 is expressed in most cases of acute leukemia and in many cases of chronic myelogenous leukemia and myelodysplastic syndromes [7,8]. Importantly, WT1 has limited expression in normal tissue beyond embryogenesis [12,13]. Promising clinical results, including evidence for induction of immune responses, using autologous monocyte-derived DC WT1 vaccines have been observed [14,15]. Effective antitumor immune responses after vaccination, however, may be impaired because of host immune depletion associated with standard anticancer therapies, which may be more profound in the post-transplantation setting. This might be overcome by the use of allogeneic approaches. Experience with DC vaccination, particularly in the allogeneic post-transplantation setting, is limited [16–19].

We describe results from a pilot trial that incorporates antigen-specific immunotherapy and allogeneic adoptive cell transfer for pediatric and adult patients with relapsed hematologic malignancies after allogeneic HCT using GMP methods for exvivo generation of monocyte-derived DCs [20]. The primary objective was to assess safety and feasibility of aWT1 peptide-loaded, HLA-A2 restricted, allogeneic, donor-derived DC vaccine designed to enhance the GVL effect of coadministered DLI. A secondary objective was to determine if immunologic and clinical responses to WT1-specific peptides could be generated by this novel allogeneic vaccine strategy for treatment of post-HCT relapse. As 1 of the few experiences reporting on the use of DC vaccination in the post-allogeneic transplantation setting, our results demonstrate feasibility of this platform, which importantly helps in setting the stage for future vaccination strategies in the post-transplantation setting, potentially for relapse prevention.

MATERIALS AND METHODS

Patients

This was a single-institution pilot study for HLA-A2 positive recipients between the ages of 1 and 75 years with WT1-expressing hematologic malignancies who experienced relapse after allogeneic HCT. HLA-A2 positive, healthy, related or unrelated donors who were 5- or 6-antigen (or 8 to 10/10 allele) genotypic HLA-matched (single HLA-A or -B locus mismatch allowed) were eligible. WT1 expression of the hematologic malignancy was confirmed by either having greater than 15% of malignant cells react with anti-WT1 by immunohistochemistry or by having a positive quantitative RT-PCR of WT1 compared with a negative control using approaches that have been previously described [21]. Recipients with rapidly progressive disease, with greater than 25% marrow blasts (in the setting of acute leukemia), with active graft-versus-host disease (GVHD) greater than grade 1 or those on immunosuppression were not eligible. This study was approved by the institutional review boards of the National Cancer Institute and the National Marrow Donor Program. Parental consent with age appropriate assent was employed as needed for pediatric enrollment. This trial is registered at clinicaltrials.gov (NCT00923910).

Study Design

All recipients received a DLI product (dose at 1 × 10⁶ CD3/kg) once every 4 weeks for a total of 3 DLIs and a DC vaccine once every 2 weeks, for a total of 6 vaccines. To manufacture the DC vaccine, a peripheral blood mononuclear cell concentrate was collected from each donor by apheresis. The peripheral blood mononuclear cell concentrates were enriched for monocytes by counterflow elutriation and the monocytes were cryopreserved. An aliquot of monocytes was thawed incubated for 3 days with granulocyte macrophage–colony stimulating factor(10 micrograms/mL) and IL-4 (2000 IU/mL) in RPMI with 10% heat-inactivated AB plasma, followed by maturation for 1 day with LPS (30 nanograms/mL) and IFN-gamma (1000 IU/mL). The mature DCs were loaded for 2 hours with a combination of 3 WT1-derived HLA A2 binding peptides (WT1 37-45, WT1 126-134, WT1 187-195). Each peptide was linked to the 11-mer HIV TAT protein transduction domain known to enhance peptide loading and antigen presentation [22–24]. Additionally, keyhole limpet hemocyanin (KLH), a neoantigen known to induce helper response, was used concurrently as a vaccine adjuvant and control antigen (Figure 1). Vaccines were administered in 2 forms: subcutaneous at a dose of 10 × 10⁶ DCs and intradermal injection at a dose of 2 × 10⁶ DCs.

Figure 1.

Protocol schema.

Immune Surveillance

Immune responses were evaluated every 4 weeks after initiation of protocol therapy and was monitored by use of interferon gamma ELISpot and by delayed-type hypersensitivity (DTH) skin testing. DTH skin testing was performed using KLH and with a cocktail of WT1 peptides as 2 separate injections. ELISpot was performed against each peptide and was considered positive if results were at least 10 spots above background on at least 2 measurements. DTH was considered positive if there was at least .5 cm induration 48 to 72 hours after placement.

RESULTS

Patients

Five recipients with 5 donors were enrolled on the trial. Patient characteristics are shown in Table 1. The median age for recipients was 17 years (range, 9 to 19 years). Four of the recipients had fully 10/10 HLA–matched related donors. One recipient had a 10/10-matched unrelated donor. Three recipients had acute lymphoblastic leukemia (ALL), 1 had acute myelogenous leukemia (AML), and 1 had Hodgkin lymphoma, all with active disease. One patient with leukemia had minimal residual disease (MRD, <5% marrow blasts), whereas the other 3 patients with leukemia had disease ranging from 5% to 10% marrow ALL or AML blasts at the time of enrollment with no interval disease-directed therapy given before initiation of protocol-specified therapy. The patient with Hodgkin lymphoma had multiple lymph nodes as sites of involvement. The median interval from transplantation day 0 to relapse was 11 months (range, 1 to 12 months after transplantation). The median time to enrollment on this protocol from transplantation was 15 months (range, 14 to 32 months), with 4 of 5 patients having received interval therapy after post-transplantation relapse to enrollment on this trial. All patients were off immunosuppression for a median of 11 months (range, 9 to 28 months) before receiving the first DC vaccination.

Table 1.

Patient Characteristics

ID	Disease	Age/Sex	Transplantation Type	Conditioning Regimen	Donor	Immunosuppression	Time from HCT to Relapse, mo	Time to First Vaccine (from day 0 of HCT), mo	Time Off Immunosuppression before Protocol Enrollment, mo	Pre-Enrollment Disease Status	DLIs Received	Vaccines Received
1	ALL	18/F	NMA	Fludarabine/cyclophosphamide	MSD	Cyclosporine	8	14	10	6%-8% blasts	3/3	6/6
2	Hodgkin	18/M	NMA	Fludarabine/cyclophosphamide	MSD	Cyclosporine	1	32	28	Multiple 1-2 cm lymph nodes	2/3	3/6
3	ALL	15/M	MA	Cyclophosphamide, TBI, thiotepa	MSD	Sirolimus/tacrolimus, methotrexate	12	14	11	5%-10% blasts	3/3	6/6
4	ALL	11/M	MA × 2	HSCT 1: cyclophosphamide, TBI; HSCT 2: busulfan, melphalan, etoposide	MRD*	None after second BMT	11	29	23	2% blasts	3/3	6/6
5	AML	19/F	MA	Cyclophosphamide, TBI	MUD	Cyclosporine	12	15	9	5% blasts	1/3	2/6

F indicates female; NMA, nonmyeloablative; MSD, matched sibling donor; M, male; MA, myeloablative; TBI, total body irradiation; MRD, matched related donor; MUD, matched unrelated donor.

^*MRD was a 10/10 matched related parent for both transplantations. The first used marrow and the second used peripheral blood stem cells.

Primary Objectives

Successful vaccine production was obtained from all donors. DC vaccination and DLI were well tolerated by recipients. One recipient developed grade 1 skin GVHD not requiring systemic therapy. The most common adverse events included grade 1 reversible pain and pruritus at the vaccine at DTH testing sites. There were no grade 3 or higher adverse events with an attribution of possible, probable, or definite relationship to the research. Additionally, there were no complications from the donor procedures. Three recipients, all with ALL, received all 6 scheduled DC vaccines and 3 DLI products. Two recipients were taken off study early for progressive disease, with 1 subject with Hodgkin lymphoma having received 2 DLI products and 3 vaccines, and 1 subject with AML having received 1 DLI product and 2 vaccines.

Secondary Objectives

Immune responses

Immunologic and clinical Responses are displayed in Table 2. Three of 5 recipients had a positive ELISpot response to any of the 3 WT1 peptides (Figure 2). These 3 patients also demonstrated a positive DTH response to KLH control, with 2 of these patients also having a positive response to the WT1 DTH cocktail containing the combination of the 3 peptides used. The 2 patients without any demonstration of an early immune response had Hodgkin lymphoma and AML and did not complete the full series of DLI/vaccines because of early disease progression. One patient with AML did not have any DTH skin testing performed because of early disease progression and initiation of alternative therapy before the testing was performed.

Table 2.

Immunologic and Clinical Responses

ID	Disease	Age/Sex	Pre-Enrollment Disease Status (Marrow Aspirate)*	DLIs Received	Vaccines Received	Time to Immune Response	WT1 ELISpot	WT1 DTH	KLH DTH	Disease Response
1	ALL	18/F	6%-8% blasts	3/3	6/6	Week 12	+	+	+	PD
2	Hodgkin	18/M	Multiple 1-2 cm lymph nodes	2/3	3/6	N/A	−	−	−	PD
3	ALL	15/M	5%-10% blasts	3/3	6/6	Week 8	+	−	+	PD
4	ALL	11/M	2% blasts	3/3	6/6	Week 4	+	+	+	PD
5	AML	19/F	5% blasts	1/3	2/6	N/A	−	N/A	N/A	PD

PD indicates progressive disease; N/A, not available.

^*Unless otherwise noted.

Figure 2.

Immune response seen in subject number 4. (A) Positive response to delayed type hypersensitivity to the WT1 peptide cocktail (right arm) as well as the KLH control (left arm) at week 4 of protocol. (B) ELISpot results by week on study, in correlation with concurrent DTH responses.

Clinical response

All subjects experienced progressive disease by the end of the study, with 2 subjects coming off treatment midtherapy because of early progression. As previously discussed, 1 subject (subject number 4) developed grade 1 skin GVHD that was controlled with topical therapy. No autoimmunity was seen, although late cytopenias and prolonged remission status were observed in subject number 4 after 4-drug reinduction for ALL progression. This patient was previously chemotherapy refractory and had relapsed after 2 allogeneic HCTs. After chemotherapy administered for progressive disease after this protocol, the patient attained a morphologic remission with prolonged pancytopenia. He subsequently received a CD34⁺ stem cell boost and remained in complete remission for 7 months. Week-36 immune results after the stem cell boost in this subject showed an ongoing positive ELIspot response to WT1.

DISCUSSION

Relapse is the primary cause of treatment failure after allogeneic HCT. Treatment options for post-HCT relapse are limited because of the possibility of ongoing GHVD; transplantation-related toxicities and cumulative comorbidities; limited efficacy of donor lymphocyte infusions, particularly in the acute leukemia setting; and rapidly progressive disease. Therefore, relapse prevention strategies are needed improve post-transplantation outcomes; however, such strategies must be feasible, effective, and safe in the post-transplantation setting. Vaccination may represent 1 such model.

There is limited experience utilizing DC vaccination in the allogeneic setting [16–19]. In this pilot study, despite the small number of patients, we describe 1 of the first attempts to apply DC-based cancer vaccines to the allogeneic setting. We demonstrate that not only is this approach safe and feasible, but also that vaccination can be used to sensitize the allogeneic donor immune system to WT1. Both in vivo and in vitro, WT1 immune response were demonstrated in 3 of 5 enrolled patients, which included responses seen in all 3 patients with ALL. This study provides proof of concept that such a strategy can be effectively employed in the post-transplantation setting, providing a justification for further study.

Although the allogeneic effect contributes to the success of HCT in ALL, a number of factors are believed to limit the GVL effect in ALL, making the immune response seen in this trial even more notable. B lineage lymphoblasts have very low expression of T cell costimulatory molecules and are poor antigen-presenting cells [25]. In addition, recent data suggest that B cells may induce tolerance during lymphopenic states [26], as may be seen in the post-transplantation setting. To overcome this obstacle, DLI was given concurrently with the DC vaccine, which at the low dose of 1 × 10⁶ CD3/kg was well tolerated, did not lead to severe GVHD, and may have contributed to the immune responses seen in this setting. Together, this suggests the possibility that leukemia-associated antigen-specific DC vaccination might be used to improve the efficacy of DLI for ALL that has relapsed after HCT. Additionally, it also supports incorporation of vaccination strategies earlier in the post-HCT course while there is ongoing, active immune reconstitution.

Although objective responses were not appreciated, it should be noted that the patient population enrolled on this study was skewed towards the majority of patients having overt relapse (2% to 10% marrow leukemia burden and prominent nodal disease, [Table 2]). The time needed to mount an immune response to vaccination may be insufficient to control disease progression in the setting of rapidly proliferating active acute leukemia/lymphoma. Furthermore, at the time of enrollment, all patients were at least 1 year after transplantation, potentially limiting the ability to harness the GVL effect; hence, the coadministration of DLI on this protocol. Subsequent strategies might incorporate DC vaccination for relapse prevention earlier after HCT, which might overcome the challenges of having active disease and the need to incorporate cancer vaccines at a time when there is active immune reconstitution. Specifically, in the setting of ALL, where CD8⁺ T cell responses to WT1 in the early post-transplantation period have been seen [27], a relapse prevention approach should be considered. In this regard, we did have 1 subject with ALL (subject number 4, Figure 2) who was noted to develop cytopenia after reinduction chemotherapy with subsequent detection of an ongoing WT1 response. This included a 7-month sustained complete remission without any additional therapy, suggesting the possibility that an ongoing WT1 response contributed to the maintained remission, particularly relevant in the post-HCT setting. With increasing data supporting the importance of monitoring post-HCT MRD and the high risk of relapse associated with post-HCT MRD [28], vaccination strategies employed when post-HCT MRD is detected could represent a relapse prevention strategy.

There are a number of limitations to this study. Patient accrual on this trial was limited because of the need for subjects to be HLA-A2 expressing for this peptide-based immunotherapeutic approach. Although HLA-A2 is 1 of the most frequent HLA molecules and is found in approximately 30% to 50% of the North American population [29], it was the most frequently cited exclusion criterion for recipients who would have otherwise been eligible for the study. Additionally, the peptide-based strategy employed on this trial is associated with epitope restriction. To overcome the limitations of peptide-based approaches, our lab has recently worked on developing a full-length WT1 mRNA electroporated DC vaccine to overcome HLA and epitope restriction and initial preclinical studies demonstrate the feasibility of this approach (data not shown). Initial experience with this approach in the clinical setting for patients with AML supports this concept as a viable approach [30]. Importantly, given that most patients on this study were in overt relapse at the time of enrollment, there was limited time to observe the possibility for an immune response to develop. This was most notable in 2 subjects who could not complete the intended regimen of 6 vaccines and 3 DLIs because of rapid disease progression and started on the protocol with higher burden disease. Future strategies to evaluate for immune response should be considered in an MRD or complete remission setting.

In summary, WT1 is a promising target for cancer immunotherapy and our study provides 1 of the first attempts to apply tumor-specific vaccine therapy to the allogeneic setting. Although results are limited in scope, we were able to demonstrate that DC-based vaccination is safe and feasible after allogeneic HCT, with a suggestion that this approach can be used to sensitive the repopulated allogeneic donor immune system to WT1. Future plans include expanding this strategy to overcome HLA and/or WT1 epitope restriction by incorporation of a full-length WT1 mRNA–based strategy and potentially moving this approach into the early post-transplantation setting, before overt relapse, for relapse prevention.