Placental Growth Factor: a Novel, Stromal-Derived Target in Human CML.

Abstract 42

Although BCR-ABL kinase inhibitors (TKIs) have revolutionized treatment of CML, they do not eradicate disease and treatment resistance can emerge. We recently identified Placental Growth Factor (PlGF) as an additional pathogenic factor in murine BCR-ABL1+ leukemia. PlGF induces proliferation of bone marrow stromal cells and of BCR-ABL1+ cell lines. PlGF is upregulated in blood (PB) and bone marrow (BM) of mice with BCR-ABL1+ leukemia and it's inhibition by a monoclonal antibody (αPlGF) significantly prolongs survival of mice bearing BCR-ABL1-unmutated and -T315I mutant blast crisis (BC) CML (Loges et al., Blood 2008; 112: abstract 1094). Now, we have further characterized the role of PlGF in disease progression, the mechanism of action of αPlGF and validated PlGF as novel target in human CML. We first quantified PB and BM PlGF protein levels at different time points after inoculation of BCR-ABL1 transduced BM cells and found significantly increasing PlGF levels upon disease progression. In line with those data, PB and BM PlGF protein levels correlated with numbers of GFP-marked leukemia cells (r=0.81; p=<0.001 and r=0.85; p=0.001, respectively). We unraveled that PlGF is almost exclusively expressed by the non-hematopoietic (CD45-), non-leukemic (GFP-) stromal cell compartment. In vitro data indicate specific PlGF upregulation by CD45- BM stromal cells upon co-culture with BCR-ABL1+ leukemia cells, but not in hematopoietic cells or endothelial cells. Hence, PlGF is a stromal target induced by interaction with BCR-ABL+ leukemia cells and significantly correlates with disease burden. Neutralization of PlGF by αPlGF and/or its genetic deletion prolong survival of mice bearing BCR-ABL1+ leukemia. Therefore we analyzed the anti-leukemic efficacy of αPlGF treatment by measuring GFP+ leukemia cells in the PB of mice at an early time-point (d14) and at end-stage disease (d28). We found a reduction of leukemia cells by 42 and 51% in mice treated with αPlGF when compared control mice (p<0.05). Furthermore, we analyzed the effect of PlGF inhibition on the leukemic BM microenvironment. Morphometric analysis of CD31 immunostaining revealed reduced microvessel density (MVD) in αPlGF treated mice compared to controls (236+15 vs. 307+17 vessels/mm2; p=0.006). Investigation of reticulin fibers indicated that αPlGF reduces the prominent fibrosis of the BM occurring upon disease progression by 30% (p=0.001), which complements the finding that PlGF expands BM derived stromal cells in vitro. Hence, αPlGF reduces the leukemic burden and normalizes the abnormal leukemic BM by reducing hypervascularization and BM fibrosis. To investigate the relevance of PlGF as novel target molecule in human CML, we determined PlGF plasma levels in healthy controls (n=10), untreated patients in Chronic Phase (CP) upon primary diagnosis (n=32) and patients with BC under treatment with different TKIs (n=9). These analyses revealed 2.1-fold upregulation of PlGF in newly diagnosed patients in CP (p<0.0001) and 3.7-fold increase of PlGF levels in patients with BC (p<0.0001) compared to healthy controls. We then investigated a potential relation between PlGF protein and BCR-ABL1 transcript numbers as determined by QRT-PCR in a single centre (n=43 CP, n=2 Accelerated Phase; treatment with different TKIs, imatinib+interferon, or homoharringtonine). We found a significant correlation between PlGF levels and BCR-ABL1 transcript numbers (r=0.45; p=0.0016), indicating that PlGF represents a disease specific target in human CML. Subsequently, we isolated CD34+ cells from healthy donors and from CML patients in CP and BC and determined PlGF expression by QRT-PCR. These analyses revealed that PlGF expression is equally low in leukemia cells as in healthy CD34+ cells. Thus, elevated circulating PlGF is most likely not secreted by leukemia cells, but by stromal cells. To investigate this hypothesis, we isolated adherent BM stromal cells from patients with newly diagnosed CML and compared PlGF expression levels to those in CD34+ leukemia cells. We found that stromal cells express >7-fold more PlGF than leukemia cells (p=0.003), which corroborates our preclinical data and extends the concept that PlGF is primarily produced by stromal cells in CML patients. In summary inhibition of PlGF may serve as a new candidate to be targeted in combination therapies or in TKI refractory CML.