MRP1 and P-glycoprotein expression assays would be useful in the additional detection of treatment non-responders in CML patients without ABL1 mutation

• MRP1 and Pgp expression assays can additionally detect TKI non-responders in CML.
• High specificity and considerable accuracy was observed from these two assays.
• The rhodamine-123 efflux assay is not useful for discriminating TKI poor responders.

We evaluated the ability of the rhodamine-123 efflux assay, multidrug resistance-associated protein-1 (MRP1) expression assay and P-glycoprotein (Pgp) expression assay to discriminate chronic myelogenous leukemia (CML) patients who had failed treatment or were at risk of failure. Each assay was performed in blood samples from CML patients (n = 224) treated with tyrosine kinase inhibitors, taken at diagnosis (n = 14) and follow-up (n = 210). Patient samples were categorized as optimal response (n = 120), suboptimal response (n = 54), and treatment failure (n = 36). Treatment-failed patients had a significantly higher MRP1 expression (5.24% vs. 3.54%, P = 0.006) and Pgp expression (5.25% vs. 3.48%, P = 0.005) than responders. Both MRP1 (%) and Pgp (%) were highly specific (95.2% and 94.5%) and relatively accurate (83.0% and 82.5%) in the detection of treatment non-responders. Of treatment-failed patients, 41.2% had a positive result in at least one assay and of these patients without ABL1 kinase domain mutation, 51.9% were positive in at least one assay. However, the rhodamine-123 efflux assay failed to discriminate two patient groups. Thus, both MRP1 and Pgp expression assays could be useful for additional identification of treatment non-responders in CML patients without ABL1 mutation.