In Imatinib Treated CML Patients with Low Baseline IC50, Early Molecular Response Is Highly Predictive of Longer Term Molecular Outcome. In Patients with High IC50, Other Prognostic Indicators Are Needed.

The achievement of a major molecular response (MMR, ≥3 log reduction in BCR-ABL) in newly diagnosed imatinib treated patients is associated with a high progression free survival. In most patients with MMR, the response is achieved by 24 months. We previously demonstrated that the log reduction of BCR-ABL after 3 months of imatinib therapy was predictive of the 24 month molecular response. We sought to improve the predictive power of the early molecular response by measuring IC50 values prior to commencing imatinib. The IC50 indicates the in-vitro sensitivity of cells to imatinib and we have demonstrated that the IC50 is predictive of molecular response to 12 months. The current analysis comprises 60 de-novo CML patients enrolled to a study in which patients received imatinib 600mg initially, increasing to 800mg if specified response criteria were not met. These included major cytogenetic response at 6 months; complete cytogenetic response at 9 months; and >4 log reduction of BCR-ABL at 12 months. The IC50 was determined by measuring the in-vitro imatinib-induced reduction in the phosphorylated form of the adaptor protein Crkl in mononuclear cells derived from blood taken prior to imatinib therapy. Dividing patients into low and high IC50 groups by the median, the probability of MMR at 6 months (33% low vs 8% high, p=0.01) and 12 months (47% low vs 23% high, p=0.03) was significantly different. However equivalent molecular responses were seen at 24 months (64% low vs 69% high p=0.56). Therefore, patients with a high IC50 have a more gradual response but are equally likely to achieve MMR by 24 months as low IC50 patients. This suggests that the IC50 assay predicts the initial slope of the leukemic cell reduction, perhaps because it reflects the sensitivity of the differentiated leukemic cells. The 3 month molecular response was not predictive of a MMR by 24 months in the patients with a high IC50. However, the molecular response at 3 months was highly predictive of a MMR in the patients with a low IC50 and identified a group with suboptimal response (Table). Similar results were seen when we analysed the predictive power of log reduction at 6 months.

The group failing to achieve MMR by 24 months had a median log reduction at 3 months of 0.75 compared to 1.58 in the group achieving MMR by 24 months, p<0.001. Confining this analysis to patients with low IC50 (n=36), the median log reduction at 3 months was significantly different at 0.76 vs 1.96 in the no-MMR vs MMR groups, p<0.001. However in analysing patients with high IC50 (n=24) the median log reduction at 3 months was no different, 1.25 v 1.37 in the no-MMR v MMR groups (p>0.5). In conclusion, for patients in this study with low pre-treatment IC50, log reduction at 3 months is highly predictive of subsequent achievement of MMR. Conversely, patients with a high IC50 have a more gradual molecular response compared to those with a low IC50 and their 3 and 6 month response is not predictive of MMR by 24 months. For patients with high IC50 predictive tests other than log reduction at 3 months may be needed.