Identification of multivariable microRNA and clinical biomarker panels to predict imatinib response in chronic myeloid leukemia at diagnosis

Imatinib Mesylate (imatinib) was once hailed as the magic bullet for chronic myeloid leukemia (CML) and remains a front-line therapy for CML to this day alongside other tyrosine kinase inhibitors (TKIs). However, TKI treatments are rarely curative and patients are often required to receive life-long treatment or otherwise risk relapse. Thus, there is a growing interest in identifying biomarkers in patients which can predict TKI response upon diagnosis. In this study, we analyze clinical data and differentially expressed miRNAs in CD34+ CML cells from 80 patients at diagnosis who were later classified as imatinib-responders or imatinib-nonresponders. A Cox Proportional Hazard (CoxPH) analysis identified 16 miRNAs that were associated with imatinib nonresponse and differentially expressed in these patients. We also trained a machine learning model with different combinations of the 16 miRNAs with and without clinical parameters and identified a panel with high predictive performance based on area-under-curve values of receiver-operating-characteristic and precision-recall curves. Interestingly, the multivariable panel consisting of both miRNAs and clinical features performed better than either miRNA or clinical panels alone. Thus, our findings may inform future studies on predictive biomarkers and serve as a tool to develop more optimized treatment plans for CML patients in the clinic.