In vivo inhibition of midazolam disposition by ketoconazole and fluoxetine, and comparison to in vitro prediction

The effect of ketoconazole (K) and fluoxetine (F) on oral midazolam (M) disposition was evaluated in 24 healthy subjects in a parallel study design. M clearance was decreased by 770 ±620% after K (200 mg/d × 11 days), in contrast to a negligible increase (17.6 ±15.9%) after F (60 mg × 5d, then 20 mg/d). The corresponding changes in Cmax of M were 280.9 ±231.4% increase with K, and 23.2 ±51.6% increase with F. To evaluate how these in vivo results related to in vitro inhibition findings, the M AUC ratio (AUC in the presence of K or F/AUC in their absence) in the subjects were compared to the predicted decrement in M clearance calculated using literature values of Ki and Km, the Cmax of M, and the achievable in vivo plasma concentrations of K, F, and norfluoxetine (NF). The AUC ratio in the presence of K was 8.7 ±6.2; and 0.82 ±0.16 in the presence of F. Using achievable Cmax of M (9× difference) and inhibitor concentrations (40-fold range for K, 4-fold range for F and NF), and literature values of Km and Ki, the predicted % inhibition of M clearance was 57.7 ±23.5% for K, 0.4 ±0.1% for F, and 1.8 ±0.5% for NF.