Prevention Of Relapse In Acute Myeloid Leukemia By Dendritic Cell Vaccination: Report on a Phase II Study With 29 Patients

Relapse remains a major problem in acute myeloid leukemia (AML) and has an important impact on survival, especially in the majority of older patients who cannot undergo allogeneic hematopoietic stem cell transplantation. In a phase II study, we investigated the impact on relapse of dendritic cell (DC) vaccination as a post-remission treatment in 29 AML patients; 26 patients were in complete remission and 3 in partial remission following chemotherapy. DC were derived from blood monocytes, electroporated with mRNA encoding the Wilms' tumor 1 protein (WT1) and administered via the intradermal route. Three different WT1 constructs were used to generate mRNA by in vitro transcription: 'construct 1' encoding full-length WT1, 'construct 2' with a Sig-DC-LAMP major histocompatibility complex (MHC) class II-skewing signal and a deletion of the WT1 nuclear localization signal and 'construct 3', a codon-optimized version of construct 2. WT1 mRNA levels in blood and marrow were followed as a measure of minimal residual disease.

DC electroporated with mRNA derived from constructs 1, 2 and 3 were used to vaccinate respectively 13, 6 and 10 AML patients at very high risk of relapse. In those 3 groups, clinical and molecular response, as determined by normalization of WT1 transcript levels in blood and/or marrow, occurred in respectively 7/13, 1/4 and 0/6 patients who had increased levels of that marker at the start of DC vaccination. Of these 8 responding patients, 3 relapsed and died; the other 5 are still in complete remission, 4 of them now more than 5 years after diagnosis and most probably cured; 1 of those 4 patients was in partial remission following chemotherapy and was brought into complete and molecular remission by the DC vaccination only. All 15 patients who did not normalize WT1 mRNA levels following DC vaccination relapsed and/or died. There was a possible effect of DC vaccination in 6 additional patients: 3 with stable disease, some of it late; and 3 at high risk of relapse but without increased WT1 mRNA levels before DC vaccination: 1 patient with erythroleukemia and 2 patients with initial leucocytosis > 20,000/µL have remained in complete remission, now at respectively 46, 42 and 35 months post-diagnosis. Overall 8/29 patients have not relapsed yet, with a median follow-up from diagnosis and start of DC vaccination of respectively 60 months (range 35 - 84 months) and 53 months (range 27-78 months).

Delayed type hypersensitivity (DTH) testing showed immunoreactivity to the DC vaccine components in all patients tested. There was also evidence of natural killer (NK) cell activation following DC vaccination. WT1 epitope tetramer+ CD8+ T-cells were evaluated in 13 HLA-A2+ patients: an increase following DC vaccination in tetramer+ T-cells for at least 2/4 epitopes tested was only observed in patients with long-standing complete remission.

In conclusion, DC vaccination had a demonstrable antileukemic effect in 8/29 AML patients and a possible effect in another 6. Contrary to expectations, the WT1 constructs with MHC class II-skewing signal did not seem to have superior activity over the full-length WT1 construct without that signal. Vaccination with WT1 mRNA-transfected DC is emerging as a non-toxic strategy to prevent or delay relapse in AML.