Retrospective analysis of P-glycoprotein-mediated drug-drug interactions at the blood-brain barrier in humans.

To date, the in vitro-in vivo correlation (IVIVC) of P-glycoprotein (P-gp)-mediated drug-drug interaction (DDI) at the blood-brain barrier (BBB) in rats indicated that the cutoff value to significantly affect the brain penetration of digoxin was [I,unbound/Ki] of 1, where I,unbound is the unbound plasma concentration of P-gp inhibitors. On the basis of the IVIVC in rats, we speculated that clinically used P-gp inhibitors do not cause DDI at the human BBB, because none of the compounds studied was [I,unbound/Ki]>1 at therapeutic doses. Recently, positron emission tomography studies with P-gp substrates, such as [(11)C]verapamil, [(11)C]N-desmethyl loperamide, and [(11)C]loperamide, together with potent P-gp inhibitors, have indicated that increases in the influx rate constant for brain entry were observed in humans. Therefore, we aimed to retrospectively analyze the results of P-gp-mediated DDIs with in vitro P-gp inhibition assays and to confirm the appropriate cutoff value. In vitro P-gp inhibition assays using verapamil, N-desmethyl loperamide, and loperamide as P-gp probe substrates were performed in human multidrug resistance protein 1-expressing LLC-PK1 cells. The efflux ratios decreased in the presence of P-gp inhibitors, and the Ki of tariquidar was 10 nmol/L, regardless of probe substrates. Taking the in vitro Ki and unbound plasma concentrations in clinical DDI studies together, the criterion [I,unbound/Ki] of 1 was an appropriate cutoff limit to observe significant P-gp-mediated DDI at the BBB in humans. On the other hand, no significant DDI was observed in cases in which [I,unbound/Ki] was less than 0.1. This criterion was comparable to the previous IVIVC result in rats.