Prognostic Value of the Rate of BCR-ABL Decline for Patients with Chronic Myelogenous Leukemia in Chronic Phase on Tyrosine Kinase Inhibitors Treatment

Objectives and background: Introduction of tyrosine kinase inhibitor (TKI) for the treatment of chronic myeloid leukemia (CML) led to favorable outcome in the majority of patients.About 70% of patients with early molecular response (BCR-ABL^IS ≤ 10% at 3-months) have 5-year overall survival of 95%. Nonetheless, CML patients remain heterogeneous group and several studies in recent years were aimed to personalize treatment based on individual patients’ characteristics. One of them was the study by B. Hanfstein et al. (2014), which showed good prognostic potential of 0.35 ratio BCR-ABL level at 3 months to absolute transcript level at diagnosis[1]. In this study, GUS was used as control gene, but at present ABL is normalization gene for BCR-ABL quantification worldwide. One of the obstacles to use of baseline BCR-ABL/ABL level is a distortion of the results of its measurement (non-linearity) due to the mixture of BCR-ABL with normal ABLgene. During the first month of therapy there takes place a rapid tumor mass reduction.

The aims of our study were to assess potential of ratio BCR-ABL level at 3 months to baseline and ratio BCR-ABL level at 3 months to 1 month using ABLas control gene to predict optimal response related to individual patient’s tumor characteristic.

Methods: Forty-three patients (median age, 50 years; range 24-84; 17 male and 26 female) with chronic phase CML were included in the study, Sokal risk groups were low-23 / intermediate-10 / high-10; 8 patients had EUTOS high-risk. Thirty-one patients started treatment with Imatinib 400 mg/day, 11 patients started with Nilotinib 600 mg/day and 1 patient started with Dasatinib 100 mg/day. Median BCR-ABL^IS transcript levels was 18.886% at diagnosis, range 3.390-3185.361%. In all patients BCR-ABL levels were monitored at diagnosis and at 3, 6 and 12 months of treatment, additionally 10 patients from this group had BCR-ABL levels evaluation at 1 month. The ratio of BCR-ABL levels at 3 months to baseline for each patient was calculated. In addition, we calculated ratio of BCR-ABL levels at 3 months to BCR-ABLlevels at 1 month for 10 patients. We performed ROC curve analysis to establish the best cut-off value to predict MMR achievement as optimal treatment results at 12 months. Then we compared predictive sensitivity of our ratio cut-off and early molecular response at 3 months (10% by IS). Statistical analysis was conducted with ROC analysis and Fisher exact test.

Results: The ratio BCR-ABL at 3 months to baseline as 0.1 had chosen as best cut-off value (sensitivity 83.33 CI 62.6-95.3; specificity 66.67 CI 34.9-90.1) to predict MMR at 12 months. Nineteen out of 23 patients (82.6%) with ratio below than 0.1 achieved MMR at 12 months, while only 9 of 20 patients (45%) with ratio more than 0.1 had optimal response (hazard ratio = 0.2625; p=0.013). Ratio of BCR-ABL levels at 3 months to BCR-ABL levels at 1 month also showed good results with the same cut-off value – 5 out of 6 patients (83.3%) with ratio BCR-ABL at 3 months to 1 month below than 0.1 achieved MMR, while patients with ratio more than 0.1 none achieved optimal response (p=0.0238). Application of early molecular response at 3 months (10% by IS) yielded worse discrimination results: 25 of 35 (73.9%) patients with BCR-ABL ≤10% at 3 months had achieved MMR at 12 months, whereas 3 of 8 (37.5%) patients with BCR-ABL level >10% had MMR at 1 year (p=0.1036). Moreover application of our cut-off value among patients with BCR-ABL level ≤10% at 3 months allowed us to revealed additional 4 high-risk patients have not reached MMR to a 1 year of therapy.

Conclusions: Our study demonstrated that the individual BCR-ABL decline rate from baseline to 3 months might be useful prognostic marker that allowed detecting more patients at risk who had no MMR at 1 year of treatment and ABL should be used as control gene. Also the study showed that the individual ratio of BCR-ABL level at 3 months to 1 month might be studied as more predictive landmark for change of TKI treatment even among these patients that have BCR-ABLlevels ≤10% at 3 months.

References: 1. B. Hanfstein, V. Shlyakhto, M. Lauseker et al. Velocity of early BCR-ABL transcript elimination as an optimized predictor of outcome in chronic myeloid leukemia (CML) patients in chronic phase on treatment with imatinib. Leukemia. 2014 May 6. doi: 10.1038/leu.2014.153.