Multidrug resistance (MDR) in AML: Predicting P-glycoprotein (Pgp) function by direct measurement of drug efflux is more effective than use of surrogate efflux indicators

3238 Assessment of Pgp function in leukemic blasts provides useful prognostic information regarding rates and treatment response and may be used for therapy selection in acute myeloid leukemia (AML) patients. Traditional assessment of Pgp function has utilized fluorescent indicators, such as DiOC2(3)or rhodamine 123 (Rh123), as surrogates for drug uptake and efflux. The objectives of this study were 1) to compare the performance of these two commonly used probes and representative therapeutic drugs in bidirectional in vitro permeability models, and 2) to validate the methodology for measuring Pgp+ status in patient cells when exposed to actual drugs susceptible to MDR [eg, daunorubicin (DNR), epirubicin (EPI), etoposide (VP-16), idarubicin (IDA), and mitoxantrone (MIT)] versus the surrogate probes and amonafide, a topoisomerase II inhibitor, which is not an MDR substrate (O’Loughlin et al. Blood 2007). Pgp-mediated differential transport was assessed in Caco-2 cells (Pgplow) or MDR1-MDCK (Pgphigh) monolayer cells as a bidirectional efflux/influx ratio (BD) calculated from separate apical-to-basolateral (A-B) and basolateral-to-apical (B-A) drug permeability measurements or as a monodirectional efflux/influx ratio (MD) based on the change in B-A permeability, both performed before and after co-administration of cyclosporin A (CSA), an inhibitor of Pgp. Our results indicated a significant correlation between BD and MD. The two methods afforded the same rank order in extent of efflux for the cohort of test compounds in both Pgplow and Pgphigh. The net efflux of amonafide was