RESUMO
P-glycoprotein (P-gp) may limit oral drug absorption of substrate drugs due to intestinal efflux. Therefore, regulatory agencies require investigation of new chemical entities as possible inhibitors of P-gp in vitro. Unfortunately, inter-laboratory and inter-assay variability have hindered the translatability of in vitro P-gp inhibition data to predict clinical drug interaction risk. The current study was designed to evaluate the impact of potential IC50 discrepancies between two commonly utilized assays, i.e., bi-directional Madin-Darby Canine Kidney-MDR1 cell-based and MDR1 membrane vesicle-based assays. When comparing vesicle- to cell-based IC50 values (n = 28 inhibitors), non-P-gp substrates presented good correlation between assay formats, whereas IC50s of P-gp substrates were similar or lower in the vesicle assays. The IC50s obtained with a cell line expressing relatively low P-gp aligned more closely to those obtained from the vesicle assay, but passive permeability of the inhibitors did not appear to influence the correlation of IC50s, suggesting that efflux activity reduces intracellular inhibitor concentrations. IC50s obtained between two independent laboratories using the same assay type showed good correlation. Using the G-value (i.e., ratio of estimated gut concentration-to-inhibition potency) >10 cutoff recommended by regulatory agencies resulted in minimal differences in predictive performance, suggesting this cutoff is appropriate for either assay format.
