Peptide Vaccines for Patients With Acute Myeloid Leukemia

Michael Schmitt, MD; Rosaely Casalegno-Garduño, MSc; Xun Xu, MSc; Anita Schmitt, MD


Expert Rev Vaccines. 2009;8(10):1415-1425. 

In This Article

Peptide Vaccination Trials for AML Patients

The basis of peptide vaccination is the observation that leukemia cells can act as antigen-presenting cells (APCs) that cleave endogenous leukemia antigens in their proteasome and present them on the HLA class I and II pathway. Eventually, such LAA-derived peptides are presented towards CD4+ helper and CD8+ cytotoxic T lymphocytes (CTLs). Peptide vaccination now consists of the inoculation of peptides, preferentially in micellar delivery systems such as incomplete Freund's adjuvant (IFA) for the slow release of peptide to APCs, such as Langerhans cells or DCs. APCs migrate to the local-regional lymph node and present a peptide to T cells in a more efficient manner than the AML blasts itself. These lymphocytes become activated and proliferate. Once such primed T cells encounter the same epitope peptide on the surface of an AML blast, the malignant cell can be lysed either through preformed granules or the Fas-Fas ligand pathway (Figure 1).

Figure 1.

Peptide vaccination concept. The vaccine is administered to the patient by a subcutaneous injection (syringe) and consists of a peptide (red bullets in the grey bubble) embedded in an oil-in-water emulsion, such as incomplete Freund's adjuvant (purple symbols). An antigen-presenting cell, such as a Langerhans cell in the subcutis, processes and presents the peptide on its surface through the HLA class I molecules, thus activating a naive CD8+ T lymphocyte. Such an activated T cell might encounter an AML blast presenting the same peptide derived from a LAA. In this case, the T cell docks again on the HLA class I molecule with the epitope peptide, resulting in either a release of granzyme B and perforin or the FasL pathway, therefore lysing the AML blast (blue arrow). This lysis is the definitive goal of an immune therapy for AML by using a defined T-cell epitope peptide derived from a LAA.
AML: Acute myeloid leukemia; CTL: Cytotoxic T lymphocyte; FasL: Fas ligand; LAA: Leukemia-associated antigen; TCR: T-cell receptor.

Different peptides derived from LAAs are under clinical investigation for AML patients in current peptide vaccination trials:

  • The PR-3-derived peptide PR-1

  • Different WT-1-derived peptides

  • RHAMM-derived peptide R3

  • Combinations of the aforementioned

A summary of the published data from these trials is summarized in Table 1.

The PR-3-derived peptide PR-1[33,34,35] has been investigated in a Phase I/II study. The first results have been presented at the Annual Meeting of the American Society of Hematology (ASH), and an update was presented in the 2007 ASH convention.[36,37,38] The HLA-A*0201-restricted CD8+ T-cell epitope peptide PR-1 has been combined with IFA and granulocyte-macrophage colony-stimulating factor (GM-CSF) and administered every 3 weeks. In total, 66 patients with AML, CML and MDS were included in the study (see also Table 1). The first 54 patients received three vaccinations; the last 12 received six vaccinations. Only skin toxicity common toxic criteria grade I in terms of erythema and induration of the subcutis at the site of injection of the vaccine was observed, but no other toxicities occurred. Remarkably, no signs of autoimmune disease were seen. In 25 out of 53 patients with active disease (47%), an immune response to PR-1 could be detected. In nine out of 25 responders (36%), clinical responses in terms of improved peripheral blood counts up to complete remission could be observed, compared with three out of 28 (10%) nonresponders. The authors pointed out that the event-free survival was 8.7 months in comparison with only 2.4 months in the nonresponder group (p = 0.03), combined with a trend of better overall survival.[37] Taking into account the heterogeneity of the different entities and subclassification of the diseases, these results need to be interpreted with caution, and should be tested in bigger and more homogeneous patient cohorts. A subgroup of 11 of 20 patients (55%) after allogeneic stem-cell transplantation showed PR-1-directed immune responses. Quite a disparity of clinical response was observed in this subgroup: while nine out of 11 patients with immunological responses also improved clinically, only one patient out of nine without PR-1-specific T-cell response showed a clinical response. Importantly, the T-cell receptor (TCR) avidity of the vaccine-induced PR-1-specific CTL was higher in the clinical responders than in the nonresponders.[35,36,37,38]

A Phase I/II peptide vaccination trial has been performed for patients with AML, CML and MDS using WT-1-derived peptides matching HLA-A2 or HLA-A24.[46,52,53] For one case of AML, the induction of a complete remission of a HLA-A2+ patient vaccinated with a WT-1 peptide was described by Mailänder et al..[54] Scheibenbogen and Keilholz reported on the vaccination of 26 HLA-A2+ patients at the American Society of Hematology 2007 meeting.[53] A dose of 200 µg WT-1 peptide was injected subcutaneously until progress using GM-CSF and HLA class II-stimulating keyhole limpet hemocyanine as adjuvants. Overall, 18 of the 26 enrolled patients had active disease; eight patients were in complete remission (CR). At the end of vaccination, 21 patients had a stable disease and five patients were in CR. Immune responses were detected by tetramer and intracellular cytokine staining in flow cytometry.[53] Oka et al. reported on a Phase I study using a HLA-A24-restricted peptide for patients with breast cancer, lung cancer, AML and MDS.[45] Three different doses (0.3, 1.0 or 3.0 mg) of peptide derived from WT-1 were emulsified with the IFA and administered biweekly three times intradermally. A total of 26 patients were enrolled and 12 out of 26 patients responded in terms of shrinkage in solid tumor size or decreased MRD after vaccination with WT-1.[45] Among the 12 AML patients enrolled in the study, five kept their status of hematological CR (continuous CR [cCR]). Four patients were not evaluable, in four patients the WT-1 PCR signal decreased, in one patient the number of leukemic blast decreased, one patient showed a stable disease (SD) and two patients showed a progressive disease (PD). Such responses of MRD could be correlated to the increase of WT-1-specific immune responses.[45]

For RHAMM,[7,8,27,50] we conducted a Phase I/II peptide vaccination trial using the HLA-A2-restricted RHAMM-R3-peptide of patients with AML, MDS and MM. In vitro experiments showed specific T-cell responses from CD8+ early effector T cells against these antigens in AML patients.[27] RHAMM is a naturally processed antigen. High expression of this antigen was found in most patients with AML, CML and MM, but also in different solid tumors.[27] Seven out of ten patients vaccinated with the RHAMM-R3 peptide showed immunological responses, and clinical responses were observed in five out of ten patients. This patient cohort received 300 µg peptide emulsified in IFA (Montanide™) with concomitant administration of GM-CSF subcutaneously. Clinical responses included cessation of erythrocyte substitution therapy, decrease of MRD and even development of a CR from a partial remission.[55] As these results were unexpected and extremely encouraging, our group extended the trial to six patients with early-stage chronic lymphocytic leukemia. Again, we noticed immunological and even clinical responses in terms of more than a 25% decrease in white blood cell counts under the vaccination therapy. Of note, in patients developing higher frequencies of RHAMM-R3-specific CD8+ T cells during the vaccination therapy, we observed a decrease in regulatory T cells defined as CD4+CD25hiCD127dimFoxP3+ T cells, a fact that underlines the role of this immunosuppressive T-cell subpopulation.[56] To further increase the immune response to our RHAMM-R3 peptide vaccine, we augmented the dose per vaccination from 300 to 1000 µg (1 mg). In this cohort of nine patients, we observed four out of nine patients with immunological and two out of nine patients with clinical responses.[57] The frequency of RHAMM-R3-specific CD8+ T cells and regulatory T cells defined as before were inversely correlated. However, by increasing the peptide dose per vaccination we observed a rather distinguished reduced response than with the lower dosage of 300 µg per vaccination.

A combination of PR-1 and WT-1 peptides has been tested by Rezvani et al. in a Phase I safety clinical trial for eight HLA-A*0201-positive patients with low-risk myeloid malignancies.[39,58] Five patients with AML were in a CR, one of them after a second allogeneic stem-cell transplantation from an unrelated donor. One patient suffered from MDS-refractory anemia (MDS-RA) with excess of ring sideroblasts, and one suffered from CML in chronic phase (molecular remission under medication with imatinib. For adjuvants and scheduling see Table 1). Five out of the eight patients mounted CD8+ T-cell responses to WT-1 and seven to PR-1. After vaccination, T-cell responses to the two antigens were associated with a significant, but transient reduction in MRD as assessed by quantitative PCR for WT-1 expression. No severe side effects were observed, particularly no signs of autoimmune disease. Several facts might explain the absence of autoimmunity, despite the identity of PR-3 with the antigen of Wegener's granulomatosis:

  • ANCAs recognize other epitopes than PR-1;

  • T-cell lines established from PR-3/ANCA-positive patients with Wegener's disease do not recognize T2 cells pulsed with PR-1;

  • Normal myeloid cells expressing PR-3 were not lysed by PR-1-specific T cell lines, which may be due to a lower protein expression of PR-3.[39]

Vaccination Approaches in AML using DCs

Dendritic cells are professional APCs with the unique property to prime naive T cells.[59,60,61] Hence, DCs are also considered to be important elements in the induction of specific antitumor immune responses.[62,63,64] Several groups, including ours, showed that DCs can be generated from the leukemic blasts of patients with AML or CML.[28,65,66,67,68] These leukemic-derived DCs were shown in some cases to have a potent capacity to induce T-cell proliferation, while still retaining the leukemic chromosomal abnormality of the original blasts.[28] These observations constitute a unique model where APCs present tumor antigens and the malignant cells themselves exhibit the same epitope. Therefore, DCs generated from AML patients can facilitate an immune response that might help in the induction of effective antileukemic T-cell responses.[69,70,71] Vaccination with DCs as professional APCs generated from autologous leukemic blasts might elicit antileukemic T-cell responses in patients with AML. To test this hypothesis, we initiated a Phase I/II trial using autologous DCs derived from AML blasts from elderly patients with a refractory AML or AML patients with a relapse of the disease. Autologous AML DCs were generated under good manufacturing practice conditions and injected subcutaneously to five AML patients up to four times at a biweekly interval. No severe adverse side effects were observed. Three patients remained in a stable condition for 5.5-13 months, and two patients died from rapidly progressive AML. Compared with the initial T-cell frequency, ELISPOT assays revealed a significant increase of granzyme B releasing by CD8+ T cells specifically recognizing the preferentially expressed antigen of melanoma-derived peptide (ALYVDSLFFL), a LAA expressed by AML blasts. The cytokine levels in the serum of the vaccinated AML patients as assessed by cytokine bead assay changed over the period of vaccination to an elevated Th1 cell pattern. IFN-g production by CD4+ T-helper cells increased during vaccination. In summary, we demonstrated that autologous AML DC vaccination is well tolerated and can result in an enhanced specific response of CTLs in AML patients.[71]

Berneman et al. were able to produce a DC vaccine for all ten patients involved their study.[72] No serious adverse events or toxicities were observed and the vaccinations were well tolerated. A decrease in WT-1 RNA expression was observed during the course of the vaccination in four out of seven patients who had an increased WT-1 mRNA level in peripheral blood. DCs elicited vaccine-specific and WT-1-specific CD8+ T-cell responses. The correlation between a reduction of circulating WT-1 mRNA and the administration of the DC vaccines strongly suggests that these DC vaccines elicit an antileukemic activity.[72,73]

Taken together, these DC vaccination studies demonstrated that DC vaccination in AML patients is safe and feasible. Immunological and, to some extent, hematological responses could be noted; however, these responses are not as pronounced as in patients who received peptide vaccination. This might be at least partly due to insufficient expression and presentation of LAAs by AML-DCs. Moreover, a part of AML-DC might exert tolerogenic effects on the immune system. Further investigation on this immunotherapeutical approach is necessary to define the best antigen format for this therapeutical approach.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: