Reduced Chemosensitivity of FLT3/ITD Mutated Cells to Cytarabine and Quizartinib Under Hypoxic Conditions

Background: FLT3/ITD mutated AML is characterized by short remission duration and high relapse rates due to the survival of a small fraction of leukemic cells (LCs) that outlive initial therapy. There is compelling evidence that the hypoxic niche in the bone marrow (BM) provides a sanctuary where subpopulations of LCs evade cytotoxic therapy and acquire drug resistance. In order to define the mechanisms involved in this process, multiple studies have focused on the interactions between LCs and the BM microenvironment but less is known about to the role of hypoxia as a modulator of drug resistance. Hence, the effects of standard and investigational therapies under hypoxic conditions are largely unknown. Here, we investigated the cytotoxic response of FLT3/ITD-mutated cells to conventional and targeted therapy under normoxic (21% O2) and hypoxic (1% O2) conditions. Methods: Molm14 (M14) cells and primary cells from relapsed/refractory FLT3/ITD mutated AML patients were incubated in culture medium under normoxic and hypoxic conditions. Cytarabine (Cy) or Quizartinib (Quiz) were added as single agents at the indicated concentrations. After 48 hours proliferation and apoptosis assays were performed using MTT assays, annexin V/PI staining and FACS analysis (M14). Cells were also assessed for FLT3 protein and Hypoxia inducible factor (HIF) target gene expression by western blotting and PCR arrays (RT₂ Profiler^TM PCR Array, Human Hypoxia Signaling Pathway Plus, Qiagen) (M14 and primary cells), respectively. Results: We found that M14 cells were significantly less susceptible to treatment under hypoxic conditions, both when exposed to Cy and targeted FLT3 inhibition with Quiz (Growth inhibition, 0.25nM Quiz: 8.5±4.2% (1% O2) vs. 22±2.5% (21% O2), p<.05; 49.6±2.5% (1% O2) vs. 58.1±2.3% (21% O2), n=8, p<.05; 0.25µM Cy: 3.7±2.5% (1% O2) vs. 14.4±1.5% (21% O2), p<.01; 1 µM Cy: 26.3±4.1% (1% O2) vs. 53.3±3.8% (21% O2), n=10-12, p<.001). In line with these findings, the apoptotic response of M14 cells to Cy treatment was significantly blunted under hypoxic conditions (Fold increase in apoptotic cells vs. untreated control, 1 µM: 5.7±0.8 fold [21% O2] vs. 2.7±0.5 fold [1% O2], n=4, p<.05). While not reaching the significance threshold, the trend was similar for the Quiz-treated cells. This effect was extended to primary, relapsed/refractory FLT3/ITD-mutated cells derived from patients (n=2) treated at our institution, where growth inhibition was consistently decreased in hypoxia at all doses of Cy and Quiz tested. Not unexpectedly, these primary cells required significantly higher doses compared to the established M14 cell line (1-10µM for Cy, and 50-200nM for Quiz). Importantly, Quiz appeared to be similarly effective in inactivating FLT3 activity under low and high oxygen tension. Therefore identifying tractable pro-survival genes induced by the low oxygen microenvironment should represent a viable strategy to increase the effectiveness of anti-leukemic agents. In order to identify such target candidates, we surveyed the effects of Quiz and Cy on a panel of 84 hypoxia-regulated genes using PCR arrays. Our preliminary studies revealed that while HIF pathway remains largely functional in the presence of Quiz or Cy, these agents severely blunt induction of several hypoxia genes (including Glucose transporter 1 [Glut1] and Carbonic Anhydrase 9 [CA9]) in both primary cells and M14 cells. Conclusions: 1.Hypoxia limits the cytotoxic effects of conventional and targeted therapeutic agents in primary and established FLT3/ITD mutated AML cells. 2. Quizartinib effectively abrogates FLT3 signaling under both normoxic and hypoxic conditions, indicating that other signaling pathways are critical for leukemic cell survival under reduced oxygen tension. 3. The induction of several potentially druggable HIF targets is disrupted by anti-AML agents, warranting the further investigation of combinational approaches between standard anti-AML agents and HIF targeted strategies. 4. Such therapeutic combinations are likely to be particularly effective undermicro-environmental stress conditions (e.g. severe hypoxia) thus addressing an unmet need in AML therapy.