Surface marker expression profiles of dendritic cells (DC) generated from blasts in patients with acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS) are valuable tools to characterize and quantify DC in experimental settings

Dendritic cells (DC) play a central role in connecting innate with
specific adoptive immunity resulting in target specific activation
T-cells. As professional antigen presenting cells (APC) DC
specifically stimulate T-effector cells, especially tumor-cytotoxic
T-cells. Therefore they are regarded as interesting candidates for
anti-tumor or anti-leukemic vaccination strategies. The
insufficient expression of costimulatory antigens, MHC molecules
and tumor-associated antigens (TAA) on the surface of cancer cells
and disturbed mechanisms of apoptosis are the main reason for an
ineffective immune response in oncologic diseases. It was shown
that acute myeloid leukemic cells can be differentiated to
leukemia-derived DC (DCleu ), regaining the stimulatory capacity of
professional DC while potentially presenting the whole leukemic
antigen repertoire. Thus, vaccination strategies, using ex vivo or
in vivo generated DC, might induce a highly specific anti-leukemic
T-cell response circumventing the cumbersome identification of
leukemia-associated antigens. In this thesis DC antigen (DCA)
expression profiles of mononuclear cells (MNC) and dendritic cells
(DC) generated from these MNC should be analyzed. The generated MNC
and DC should be compared with respect to their DC antigen (DCA)
expression profiles and the DCAs value to detect and quantify
(leukemia-derived) DC in different AML/MDS subtypes and under
different culture conditions. Therefore MNC and DC were generated
from 137 patients with acute myeloid leukemia (AML) and 49 patients
with myelodysplastic syndromes (MDS) under 6 different serum free
culture conditions. DCA studied were: CD1a/1b/1c, CD206, CD25,
CD137L, CD83, CD86, CD80 and CD40. DC-generating media were chosen
according to their different mechanisms of inducing
DC-differentiation: 1. ‚Basic method‘: TNF/GM-CSF/IL-4, 2.
MCM-Mimic, 3. Ca Ionophore, 4. Picibanil, 5. Poly I:C and 6.
Cytokines. Quality and quantity of generated DC was estimated by
Flow cytometry applying a specified, ‘DC-based’ gating-strategy.
Expression and coexpression profiles of 10 different DCA as well as
various costimulatory molecules, maturation markers and blast
antigens were evaluated. Only those DCA qualified for the
quantification of leukemia-derived DC that were not expressed on
uncultured MNC fractions. AML patients presented with an average of
58 % blasts, MDS patients with 13 % blasts in MNC fractions. DCA
were expressed on average on less than 7% of uncultured MNC,
however some of the markers could be expressed on up to 77% of
uncultured cells in single AML cases. Consequently these DCA did
not qualify for detection of DC in those cases. Highest expression
rates were found for CD86 and CD40 in naïve AML and for CD137L and
CD40 in naïve MDS samples. Other DCA (e.g. CD1a, 1b, 1c) were only
rarely found on naïve blasts. DCA expression on uncultured AML and
MDS MNC varied with FAB types and cytogenetic risk. After culture
in different DC-differentiating media, on average 28% DC could be
generated from AML MNC and 30% from MDS MNC, depending on methods
used, with an average DC viability of more than 60% and an average
DC maturity of 49% (AML) and 56% (MDS). On average 36% of leukemic
blasts could be converted to DC. Proportions of DCleu in the total
DC fraction varied from 40-58% and were on average 49% (AML) and
43% (MDS) after culture. Average results of all culture methods
tested were comparable, however every method failed to create DC in
some individual cases. The most important results of this thesis
are: 1. It could be shown that DCA are expressed on naïve blasts in
AML and MDS in individual patients. That means that the individual
patients’ DCA-profiles have to be evaluated before DC-culture to
find suitable DCA to detect and quantify (leukemia-derived) DC
after culture. 2. Different methods of DC-generation qualify with
varying individual efficiency to generate leukemic, mature,
migratory and viable DC in individual cases. 3. To select the best
DC-generating method the best DC-marker (no expression on naïve
blasts, high expression on DC) has to be chosen to quantify DC in
individual samples. 4. The use of only one method is not sufficient
to create DC in every single AML and MDS sample. However, a
successful, quantitative DC/DCleu -generation is possible in every
case of AML and MDS by the combination of 3 different DC-generating
media, but not every blast is convertible to DC leu . 5. There is a
need for new, specific DC-markers that are not expressed on naïve
blasts.