AMN107 (Novartis) is a novel Abl tyrosine kinase inhibitor specifically developed to be more selective for BcrAbl. AMN107 also maintains activity against the most common mutations associated with clinical resistance to imatinib mesylate (IM). In preclinical studies in cell lines and animal models, AMN107 was found to have greater potency than IM. By 3H-thymidine proliferation assays, the IC50 for AMN107 in K562 cells was 30 +/− 10nM compared with 600 +/− 60nM for IM. AMN107 and IM reduced K562 output cell number to 25% of input at 50 and 1000nM respectively, at 72h. These data are in keeping with the reported 20-fold increase in potency of AMN107 over IM. In addition, we have tested AMN107 for in vitro activity against primary CD34+Ph+ CML cells during 72h of culture in 5 growth factors. In CML cells (n=5), AMN107 and IM failed to reduce input cell number although the total cell output was restricted to 50% of PBS treated control at 2 +/− 1μM for AMN107 and to 31 +/− 7% of PBS treated control for 5μM IM suggesting the drugs were equipotent. The ability of the drugs to inhibit BcrAbl activity was then measured indirectly via the phosphorylation status of CrkL using a specific antiphospho-CrkL antibody and flow cytometry. Once again AMN107 and IM appeared equipotent in CML cells with 5μM of each compound leading to equal de-phosphorylation of CrkL. We next tested the efficacy of AMN107 as a single agent and in combination with IM against quiescent CML cells using in vitro dye (CFSE) tracking experiments. We evaluated by flow cytometry the proportion of input cells remaining alive, CD34+ and undivided (CFSEmax) or in first division. Compared to PBS treated control, 1.7, 2.5, 3.8 and 4.7-fold increases were found in the proportion of input CD34+ cells recovered in divisions 0 and 1 after 3 days exposure to 0.005, 0.05, 0.5 and 5μM AMN107, respectively. This was less accumulation than observed in the IM (5μM)-treated cells (11.0-fold). The combination of IM and AMN107, each at 5μM, was more effective in terms of total cell kill (54 and 74% fewer total cells remaining than with IM and AMN107 alone, respectively) and resulted in fewer viable cells recovered in divisions 0 and 1 than with either agent alone (for the combination, 1.9-fold on PBS treated recovery). We finally assessed the role of ABCG2 in modulating AMN107’s access to its intracellular BcrAbl target. We have previously shown ABCG2 to be over-expressed on CML stem cells and to interact with IM (

). We hypothesised that AMN107 and IM may co-operate as ABCG2 substrates or inhibitors to increase the intracellular levels of either or both drugs thus amplifying their efficacy against target protein specifically in CML stem cells. In competition assays with a known fluorescent substrate of ABCG2 (ie BODIPY-prazosin, BP), a specific inhibitor of the ABCG2 pump (fumitremorgin C, FTC) and an ABCG2 stably transfected AML cell line (AML6.2), the sample treated with BP plus FTC is taken to have greatest retention (100%). AMN107 inhibited efflux in a dose dependent manner to a maximum of 88% at 5μM, similarly to IM. Thus, AMN107 was equipotent with IM in primary CML stem cells in terms of restricting cell growth, inhibiting BcrAbl activity and interacting with ABCG2. However, AMN107 alone lead to less accumulation of quiescent CML cells in vitro as compared to IM, with the combination even more effective in this regard. The apparent co-operative effect of AMN107 and IM at the stem cell level would be predicted to improve clinical responses if tolerated in patients.

Author notes

Corresponding author