Comparison of midazolam and simvastatin as cytochrome P450 3A probes

OBJECTIVE: Our objective was to compare simvastatin with the validated probe midazolam in the assessment of cytochrome P450 (CYP) 3A activity. METHODS: This study used an open-label, fixed-sequential, 3-way crossover study design. Nineteen subjects received oral doses of 0.075 mg/kg midazolam and 40 mg simvastatin during 3 phases (baseline, after inhibition with 400 mg ketoconazole for 10 days, and after induction with 600 mg rifampin [INN, rifampicin] for 9 days). Serial plasma concentrations of midazolam and simvastatin were obtained. Oral clearances of midazolam and simvastatin were compared. RESULTS: Oral midazolam clearance decreased after pretreatment with ketoconazole (from a geometric mean of 25 mL x min(-1) x kg(-1) [range, 12-57 mL x min(-1) x kg(-1)] to 2.7 mL x min(-1) x kg(-1) [range, 1.2-8.5 mL x min(-1) x kg(-1)], P < .001) and increased after pretreatment with rifampin (to a geometric mean of 203 mL x min(-1) x kg(-1) [range, 125-371 mL x min(-1) x kg(-1)], P < .001). Oral simvastatin clearance decreased after ketoconazole (from a geometric mean of 312 mL x min(-1) x kg(-1) [range, 151-1478 mL x min(-1) x kg(-1)] to 25 mL x min(-1) x kg(-1) [range, 8.0-147 mL x min(-1) x kg(-1)], P < .001) and increased after rifampin (to a geometric mean of 3536 mL x min(-1) x kg(-1) [range, 413-10,329 mL x min(-1) x kg(-1)], P < .001). The change in simvastatin clearance was highly variable from baseline to inhibition (6- to 33-fold decrease) and from baseline to induction (2- to 39-fold increase) compared with midazolam (7- to 18-fold decrease during inhibition and 4- to 12-fold increase during induction). Midazolam and simvastatin oral clearances were correlated for all study phases (r = 0.5 and P = .03 for baseline and r = 0.53 and P = .02 for inhibition) but were weakest for induction (r = -0.031, P = .22). The area under the concentration-time curve inhibitory ratio for midazolam was 9.4 versus 12.4 for simvastatin (r = 0.3, P = .03). CONCLUSIONS: Compared with midazolam, simvastatin is a nonvalidated, suboptimal probe for studying CYP3A drug interactions because of its lack of CYP3A specificity.