Generation of Leukemia-Derived Dendritic Cells (DCleu) as a Basis for Specific Immunotherapy in Acute Myeloid Leukemia (AML) and Myelodysplastic Syndrome (MDS).

Abstract

The presentation of leukemic antigens can be improved in AML and MDS by in vitro generation of dendritic cells of leukemic origin (DCleu), thereby creating a platform for the generation of leukemia-specific cytotoxic lymphocytes (CTL). To further investigate this approach, we developed a serum-free culture system using MCM-Mimic medium (X-vivo + GM-GSF + IL-4 + FL + IL1β + IL-6 + TNFα + PGE2) for differentiation of malignant myeloid blasts to DCleu. Periperal blood mononuclear cells (PB-MNC) were obtained from 100 AML and 55 MDS-patients. Samples contained a mean of 59%/and 6% of blasts, respectively. After 14 days, cultures contained on average 34% (AML) and 20%(MDS) DC. DC yields were best in monocytic subtypes (AML M4, M5 and CMML). Cytogenetic aberrations of the leukemia had no influence. The leukemic origin of cultured DC was demonstrated using a acombined FISH/immune phenotyping assay (FISH/IPA): In cells showing a characteristic DC morphology and immune phenotype, FISH was used to detect specific cytogenetic aberrations identified in the leukemic population at time of diagnosis. Alternativly, DCleu were identified by detecting coexpression of DC markers and a leukemia-specific immune phenotype (as defined by aberrant expression of lineage markers or a characteristic CD34+/CD117+ phenotype in MDS). However, in most cases not all leukemic blasts in a given sample could be differentiated, since on average 47% of clonal cells did not acquire a DC-like immunophenotype. Vice versa, not all DC identified at the end of the culture period were DCleu. The capacity of DCleu to elicit a specific T-cell response was demonstrated by upregulation of contact-molecules, responsible for DC/T-cell contact, by DC-activated T-cell proliferation and by the capacity of DC-activated T-cells to specifically lyse naive blasts. In 6%/31% of AML/MDS-cases, <10% DC could be generated. Therefore, other DC-culture-assays were compared with respect to DC harvest, to identify the best method for DC-generation in each individual patient. Compared to `MCM-Mimic`, harvest of DC could be improved by `CA-Ionophore-media`(A23187 + Il4) in 4 of 7 cases, whereas in 3 cases, MCM led to higher DC-yields. In conclusion, DCleu can be generated in the majority of patients with AML and MDS. To optimize DC harvest for in vitro and in vivo use, different culture assays should be compared in each individual case. DCleu are able to elicit a specific T-cell response in vitro. Nevertheless, cultures containing DCleu also contain relevant numbers of undifferentiated blasts and DC of non-leukemic origin. These cells may represent an obstacle for the clinical use of DCleu, since they may cause specific anergy or unspecific stimulation of effector T-cells. Improvement of culture conditions for generation of DCleu, and methods to separate DCleu before stimulation of effector cells will be required, before clinical trials are feasible.