The In-Vivo Level of Imatinib Induced Kinase Inhibition Achieved in De-Novo CML Patients during the First 28 Days of Therapy Is a Powerful Predictor of Molecular Outcome.

Crkl is an immediate downstream partner of BCR-ABL, and phosphorylation of Crkl (p-Crkl) occurs as a direct consequence of this interaction. Utilising western blot for the in-vitro imatinib induced reduction in %p-Crkl, we have previously demonstrated that de-novo CML patients with a high intrinsic sensitivity to imatinib (low IC50^imatinib) have an increased probability of achieving a major molecular response by 12 months. In a related assay of in-vivo imatinib activity, we now demonstrate that the percentage reduction of p-Crkl measured in blood collected weekly, over the first 28 days of imatinib therapy is highly predictive of molecular outcome. There was correlation between the % of p-Crkl at day 21–28 and the IC50^imatinib (R²=0.4 p=0.02. n=32) indicating that higher IC50^imatinib predicts decreased in-vivo imatinib sensitivity. Further to this a low IC50 is necessary to achieve good in-vivo kinase inhibition. There was also good correlation between the %p-Crkl at this time and the %BCR-ABL/BCR (R²=0.42 p=0.004) as determined by RQ-PCR at 1 month. Data was grouped into the first and second 14 days of treatment, and the percentage reduction in p-Crkl from baseline calculated. The % reduction in p-Crkl over the first 14 days was not predictive of molecular outcome at any timepoint (p>0.05). However response over days 21–28 was highly predictive of molecular response to 24 months of therapy. Further to this, a significantly higher proportion of patients with low in-vivo kinase inhibition have a sub-optimal response.

These data demonstrate that patients with high in-vivo imatinib-induced kinase inhibition over the first 28 days of therapy have far superior molecular responses to two years compared to those with a low kinase inhibition. In addressing the potential confounding influence of dose, the observation held true even when analysis is limited to those 37 patients who received 600mg daily over the first 28 days of therapy (p<0.05 at all timepoints). Furthermore, a substantial proportion of patients with low imatinib-induced kinase inhibition exhibit a sub-optimal response, which we have previously shown to be predictive of subsequent clinical failure. In conclusion in-vivo imatinib-induced kinase inhibition, as measured by the percentage reduction in p-Crkl provides a powerful early predictor of long term molecular response. This demonstrates that the actual level of kinase inhibition achieved may be the key determinant of molecular outcome and suggests that second generation kinase inhibitors, with increased potency, may result in more uniform molecular responses.