Selection for CML cells with BCR-ABL kinase domain mutations represents the predominant molecular mechanism responsible for loss of response to imatinib. Similarly, we have found acquired resistance to dasatinib to be associated with kinase domain mutations in 100% of cases (n=15). However, unlike the multitude of imatinib resistant mutations, which to a large extent occur at non-contact residues and destabilize the inactive confirmation to which imatinib binds, only two mutations appear to be responsible for nearly all cases of dasatinib resistance, T315I and V299L (Shah et al, submitted, ASH 2006). Both of these mutations occur at critical contact residues between the ABL kinase domain and dasatinib. Successful treatment of dasatinib-resistant cases will therefore require strategies to successfully eliminate cells that harbor these mutations.

Use of a combination of kinase inhibitors with the ability to collectively suppress all BCR-ABL kinase domain mutants would be predicted to lead to profoundly minimize disease resistance and relapse on targeted therapy. Although the combination of imatinib and dasatinib may prevent selection of the dasatinib-resistant V299L mutation, these agents share many targets, and their combination may therefore result in substantial toxicity. Moreover, the combination of imatinib and dasatinib is not predicted to effectively inhibit the growth of cells harboring BCR-ABL/T315I.

We previously have shown that the Aurora kinase inhibitor VX-680 can bind to the ABL kinase domain and inhibit the kinase activity of the T315I mutation at low micromolar concentration. VX-680 is showing early signs of efficacy in CML cases associated with the T315I mutation. Interestingly, the co-crystal structure of VX-680 reveals that V299 is one of 14 contact residues within the ABL kinase domain. Substitution of leucine at this residue might therefore be expected to diminish the potential affinity of VX-680 for BCR-ABL/V299L. However, analysis of the amino acid sequence of Aurora-A revealed divergence from native BCR-ABL at this the corresponding amino acid position due to the presence of a leucine in Aurora-A. BCR-ABL/V299L is therefore a mimetic of Aurora-A, and as a result, predicted to retain sensitivity to VX-680.

We therefore assessed VX-680 for its ability to inhibit the kinase activity of BCR-ABL/V299L in Ba/F3 cells and found effective inhibition of the kinase activity at low micromolar concentration. Consistent with predictions based upon structural considerations, the V299L mutation is somewhat more sensitive to VX-680 than BCR-ABL/T315I. We confirmed these results in an analysis of primary human PBMCs obtained from a dasatinib-resistant patient who had evolved the V299L mutation on therapy.

Our findings suggest that early combination therapy with two kinase inhibitors, dasatinib and VX-680, may successfully suppress resistant disease by collectively eliminating BCR-ABL kinase domain mutation as a mechanism of resistance, and thereby achieve effective long-term disease control in the vast majority of patients.

Disclosures: Neil Shah has served as a consultant for Bristol-Myers Squibb Oncology. Ronald Paquette has served as a consultant for Telik.; Timothy Hughes has received research funding from Novartis Pharmaceuticals. Susan Branford has received research funding from Bristol-Myers Squibb Oncology.; Neil Shah has received honoraria from Bristol-Myers Squibb Oncology. Timothy Hughes has received honoraria from Novartis Pharmaceuticals. Ronald Paquette has received honoraria from Telik and Amgen.; Neil Shah and Susan Branford have served on the dasatinib advisory committee for Bristol-Myers Squibb Oncology. Timothy Hughes has served on advisory boards for Novartis Pharmaceuticals and Bristol-Myers Squibb Oncology, and is on the Spearkers Bureau of Novartis Pharmaceuticals.

Author notes

*

Corresponding author

Sign in via your Institution