Chronic Myeloid Leukemia Stem Cells and Their Differentiated Progeny Display Divergent Drug Sensitivities to Imatinib Mesylate and Interferon-Alpha.

Imatinib has largely replaced interferon-alpha (IFN) as front-line therapy in the treatment of chronic myeloid leukemia (CML) because of its favorable toxicity profile and superior initial response rate. However, recent laboratory data demonstrating that CML stem cells may have limited susceptibility to imatinib brings into question the potential durability of these responses. In contrast, responses to IFN often take place over months or years, but extensive long-term clinical data indicate that they are often long lasting. In order to determine if the differences in clinical response kinetics could be explained by the activity of imatinib and IFN against CML stem cells or their differentiated progeny, we examined the effects of each agent on distinct cellular subsets in vitro. CD34⁺ CML progenitors were isolated from 4 patients with newly diagnosed chronic phase CML and incubated with IFN (1000U/ml) or imatinib (10μM) for 96 hours followed by long-term liquid culture over 1-3 weeks to assay more primitive CML stem cells. The effects of each drug on CML progenitors was determined by calculating the number of CFU-GM positive for BCR-ABL by FISH and comparing these values to the bcr-abl⁺ CFU-GM formed by untreated control cells. Initially, the CML CFU-GM recovery immediately following 96 hours of drug treatment was 62.7 ± 8% and 9.7 ± 2% from IFN and imatinib-treated groups, respectively. After 3 weeks of long-term culture, CML CFU-GM recovery was 23.4 ± 2.6% following IFN treatment and 57.3 ± 22% after imatinib. Thus, imatinib was significantly more toxic to committed CML progenitors than primitive CML progenitors responsible for the maintenance of long-term liquid cultures (p = 0.04). Conversely, IFN had significantly greater activity against primitive CML progenitors than committed progenitors (p = 0.05). Although IFN has diverse biological properties, the mechanisms responsible for its antileukemic activity are unknown. Treatment of the human CML cell line KT-1 with IFN resulted in increased expression of the myeloid antigens CD33 and myeloperoxidase as well as the inhibition of clonogenic growth demonstrating that IFN induced terminal differentiation. Furthermore, several groups have shown that myeloid growth factors also induce differentiation of CML cells in vitro and clinically enhance the activity of IFN; accordingly, the addition of GM-CSF (200U/ml) augmented the differentiation of KT-1 cells. Moreover, myeloid growth factors were required for differentiation as neutralizing antibodies against GM-CSF and IL-3 inhibited the activity of IFN. The addition of GM-CSF to IFN produced similar effects on clinically derived CD34⁺ CML cells. Although imatinib is a potent inhibitor of committed CML progenitors, it is relatively ineffective against more primitive CML stem cells. In contrast, IFN appears to act primarily against CML stem cells by inducing terminal differentiation that is dependent on the activity of myeloid growth factors. Imatinib and IFN have divergent effects on CML progenitors at different stages of maturation that may correlate with clinical response kinetics and durability.