DC Vaccination Induces Antigen Specific Immune Responses in AML Patients: A 1-Year Interim Assessment

Acute myeloid leukemia (AML) is a lethal hematological malignancy characterized by high rates of relapse (50%). Studies of immune responses in AML patients led to identification of two leukemia-associated antigens (Wilms' tumor protein 1 (WT1) and preferentially expressed antigen of melanoma (PRAME) for which cellular and humoral immune responses were detected.

Based on expression profiles in AML and normal tissues, WT1 and PRAME were selected as target antigens for a DC vaccine, aiming to elicit adaptive and innate immune responses in AML patients.

The primary objective of the trial (EudraCT No.: 2014-003520-44) is to evaluate safety and feasibility of active immunotherapy with antigen-loaded autologous DCs. The secondary objectives are to assess DC vaccination-induced T cell responses against WT1 and PRAME and to evaluate the clinical outcome over a period of 2 years or until disease progression/relapse.

WT1-positive AML patients, with or without PRAME positivity were included, if they had a morphological remission with or without hematological recovery (CRi) after induction chemotherapy and were not eligible for allogeneic hematopoietic stem cell transplantation.

DCs were administered intradermally (2.5x10⁶ WT1 plus 2.5x10⁶ PRAME DCs) on a weekly basis for 4 consecutive weeks, followed by administration on week 6 and then every 4 weeks until 2 years, tumor progression or patient withdrawal. This report focusses on the immune-monitoring data collected during the first year of treatment.

Multiple parameters were evaluated, including WT1- and PRAME specific T cell responses in peripheral blood (IFN-g ELISPOT), the recurrence of WT1- and/or PRAME-positive AML blasts in bone marrow (qRT-PCR) and peripheral blood (flow cytometry), and occurrence of AML-specific mutations in bone marrow (NGS). DCs were assessed for expression of WT1 and PRAME and costimulatory molecules by flow cytometry, while secretion of IL10 and IL12 was evaluated by double-color ELISPOT.

Twenty CR/CRi AML patients were vaccinated, and vaccination was well tolerated, and no signs of autoimmunity were observed. 12/20 patients remained stable throughout the first year of treatment, while 8/20 relapsed. 2/8 relapsed patients died due to progression of disease during the first year.

In patients in stable remission, 7/12 patients had no detectable IFN-γ ELISPOT response in peripheral blood, 3/12 mounted a response and 2/12 were not evaluable due to high background production of IFN-γ. Positive IFN-γ ELISPOT responses were observed within 1-30 weeks of treatment and associated with a decrease over time in levels of WT1 (3/3) and/or PRAME mRNA (1/3) expression. Except for 2 patients who remained MRD-negative throughout the first year of treatment, ELISPOT-negative patients in remission (5/7) exhibited low, relatively stable levels of target antigens. Stable remissions were associated with one (3/12) or no (8/12) AML-related mutations while one patient was not evaluable.

AML relapses were observed within 50 days in 4 patients, within 51-100 days in 1 patient, and after more than 150 days in 3 patients. Overall, 6/8 relapsing patients exhibited positive IFN-γ ELISPOT responses, while 2 patients relapsing in <50 days failed to respond and one patient relapsing on day 77 exhibited an IFN-γ response at baseline. In all other relapsing patients, IFN-γ ELISPOT responses were detected within 6-30 weeks after vaccination, in association with rapid and progressive increases in levels of WT1 and/or PRAME. AML-related gene mutations were detected in 6/8 relapses, while one patient was not evaluable, and one had no mutation; In 4 of the relapsing patients, 3 to 6 mutations were detected, while in 2/8 a single mutation was observed.

In conclusion, IFN-γ responses to WT1 and/or PRAME were detected in peripheral blood of 75% (6/8) of relapsing patients, but only 25% (3/12) of the patients in remission. It is tempting to speculate that detection of IFN-γ response in peripheral blood is linked to the concomitant presence of AML blasts in the periphery of relapsing patients. Stable or decreasing levels of WT1 and/or PRAME mRNA in bone marrow of most patients in remission is compatible with the hypothesis that a local response against AML antigens may be ongoing, despite a negative IFN-γ response in peripheral blood. Monitoring of the patients currently in remission may shed some light on the role of DC vaccination on the prevention of AML recurrence.