High expression of urokinase plasminogen activator receptor (UPA-R) in acute myeloid leukemia (AML) is associated with worse prognosis.

Urokinase-type plasminogen activator receptor (UPA-R; CD87) is a membrane protein responsible for plasmin expression on cells facilitating cellular extravasations and tissue invasions. We studied the expression of the UPA-R on bone marrow (BM) cells of 93 patients with acute myeloid leukemia at first diagnosis and 8 healthy probands as controls by FACS analysis using phycoerythrin (PE)-conjugated antibodies. A case was defined as UPA-R-positive (UPA-R+) if >20% of the gated cells expressed UPA-R. Whereas none of the 8 healthy BM samples was positive for the UPA-R, 32 (34%) of the 93 AML samples were UPA-R+. Expression of UPA-R was heterogeneous in different FAB types, however, with the highest expression rates in monocytic subtypes (FAB M4/M5): 18%/19%/30% of UPA-R+ cases were found in M1/M2 or M3, and 58%/80% of cases with M4 or M5 were UPA-R+. Proportions of UPA-R+ cells varied between 1% and 98% of the mononuclear cell fractions, with the highest proportions in M4/M5 subtypes (on average 27%/40% UPA-R+ cells) and the lowest expression in AML M2 (11% UPA-R+ cells). The density of expressed UPA-R, estimated as mean channel fluorescence activity, was highest in cases with AML M1 (mFI: 124) followed by M4 and M5 (mFI: 78/77) and lowest in AML M2 (mFI: 43). In sAML, higher proportions of UPA-R+ cases (8 of 18; 44%) compared to pAML (24 of 75; 32%) were found as well as higher proportions of UPA-R+ cells (27% vs. 19%). Separating our patients' cohort in cytogenetic risk groups, we could not detect significant differences in the UPA-R expression profiles. For evaluations of the clinical course of AML, only patients treated by the AML-CG protocol (n = 65) were included. In the group of patients who did not respond to AML-CG therapy, significantly higher proportions of UPA-R+ cells (31% vs. 14%, P = 0.0015, t-test) were found. By evaluating a cut-off value for the percentage of positive cells that allows the most significant separation and differentiation between cases with shorter or longer relapse-free survival times, we could show that patients with >26.5% UPA-R-positive cells were characterized by a significantly higher risk for relapse compared to cases with <26.5% positive cells (P = 0.05). In summary, our data show a high expression of the UPA-R in AML, especially in (myelo)monocytoid subtypes. Cases with higher proportions of UPA-R+ cells were characterized by a significant lower remission rate after AML-CG therapy and a higher risk for relapse. Although prospective trials are still lacking, UPA-R is a prognostically relevant factor independent from the karyotype. UPA-R positivity may identify subtypes of AML associated with a more aggressive clinical course. Thus due to lower remission probabilities in UPA-R+ cases, a more intensive induction therapy regimen could be considered. 2005 Wiley-Liss, Inc.