ABSTRACT:The role of P-glycoprotein (P-gp) on the distribution of the benzodiazepine triazolam (TRZ) and the azole antifungal agent ketoconazole (KET), and on the TRZ-KET interaction, was studied using mdr1a(؊) or mdr1a/b(؊/؊) mice (P-gp-deficient mice) and matched controls. TRZ and KET also were studied in Caco-2 cells in Transwell culture. After single i.p. injections of TRZ or KET in separate groups of control mice, brain concentrations of TRZ exceeded those in serum [brain/serum area under the concentration curve (AUC) ratio, 5.0], whereas brain/serum AUC ratios for KET were approximately 0.5. On the basis of single time points, brain concentrations of TRZ, or brain/serum ratios, were similar in P-gp-deficient animals compared with controls, whereas P-gpdeficient animals had significantly higher KET brain concentrations and brain/serum ratios. Coadministration of KET with TRZ increased TRZ concentrations in serum, liver, and brain, both in controls and in P-gp-deficient animals, probably attributable to impairment by KET of CYP3A-mediated clearance of TRZ. However, KET did not increase brain/serum ratios of TRZ in either group. In Caco-2 cells, basal-to-apical flux of TRZ was higher than apical-to-basal flux. However, verapamil (100 M) did not alter flux in either direction. KET inhibited basal-to-apical transport of rhodamine-123, with a 50% inhibitory concentration of 2.7 M. Thus, TRZ does not appear to undergo measurable blood-brain barrier efflux transport by P-gp in this animal model. KET impairs clearance of TRZ but does not increase tissue uptake. However, KET itself may be a substrate for efflux transport at the blood-brain barrier.