Physiologic Hypoxia Protects Chronic Myelogenous Leukemia Progenitors From Elimination by Imatinib Mesylate.

Abstract 2181

Poster Board II-158

The use of Abl tyrosine kinase inhibitors (TKI), such as imatinib mesylate (IM), has seen a major advance in the control of chronic myelogenous leukemia (CML). However, since TKIs do not seem to eliminate CML stem cells, the use of TKIs may not represent a curative approach. Recent reports have shown that the bone marrow (BM) microenvironment, which facilitates the proliferation and renewal of resident hematopoietic stem cells (HSC), is hypoxic. The average BM O2 tension has been measured at 6-7% O2 in humans (physiologic hypoxia), including in patients with leukemia (Fiegl et al. Blood, 2009), while levels in the HSC niche are estimated to be even lower (<1% O2). In light of the above, it is likely that CML stem cells also reside in the HSC niche, and are similarly adapted for survival and self-renewal in this environment. Furthermore, although hypoxia is known to be associated with both radio- and chemo-resistance in solid tumors, its role in leukemia has has not been thoroughly investigated, particularly in the context of resistance to targeted therapies. Accordingly, we hypothesized that hypoxia protects CML progenitor cells from elimination by IM, a phenomenon which may contribute to disease persistence. To test the above, we examined if hypoxia modified the response of CML progenitor cells toward IM. Primary chronic phase (CP) CML samples “from either BM or peripheral blood (PB)” were obtained at the time of diagnosis from seven individuals with clinically-defined IM-sensitive (IM-S) or IM-resistant (IM-R) disease. Cells were then incubated with IM (0, 0.25, 1 or 5μM) under hypoxic (0.5% O2) or normoxic (21% O2) conditions for 96 hours. The treated cells were then harvested and plated in methylcellulose for 14 days (under 21% O2), after which the number of colony-forming cells (CFCs) were counted by two independent observers. In the absence of IM, four of the seven samples had a significant increase in CFCs (1.5-2.4 fold) when cultured in 0.5% O2 vs 21% O2, suggesting that primary CML CFCs may be better maintained in hypoxia. Furthermore, when treated with IM, six out of seven samples cultured under 0.5% O2 demonstrated dramatic increases in CFCs compared to those treated under 21% O2: a 2.3 to 9.0-fold increase at 1μM, and a 4.4 to 35.0-fold increase at 5μM IM. We also found that the protective effect of hypoxia was independent of the original source of the CFCs (BM or PB). In addition, we found that progenitors from both IM-S and IM-R patients were protected from IM under 0.5% O2, suggesting that hypoxia-induced protection is a general feature of CML progenitors. In order to establish a model for further study, we also tested if hypoxia elicits similar responses in four human CML cell lines (K562, AR230, LAMA84, and BV173). In contrast to primary CML cells, we found that hypoxia actually impaired the ability of all four cell lines to form colonies, and also did not confer protection from IM. These results show that CML cell lines have adapted to conditions of normoxia, and are thus inappropriate models to study the hypoxic response in CML. Additional data will be presented describing the mechanisms that may underlie the protective effect of hypoxia, as well as compounds that can counteract such effects. Ongoing experiments using the LT-CIC assay will also determine the response of primitive CML progenitors to growth under conditions of physiologic hypoxia vs normoxia. In conclusion, our results show that physiologic hypoxia protects CML progenitors from IM, and suggest that blocking of hypoxia-induced survival pathway(s) in CML progenitor cells may facilitate the elimination of residual CML progenitors.