Effect of saquinavir on the pharmacokinetics and pharmacodynamics of oral and intravenous midazolam

Objective: To assess the effect of human immunodeficiency virus protease in hibitor saquinavir on the pharmacokinetics and pharmacodynamics of oral and intravenous midazolam. Methods: In a double-blind, randomized, two-phase crossover study, 12 healt hy volunteers (six men and six women; age range, 21 to 32 years) received o ral doses of either 1200 mg saquinavir (Fortovase soft-gel capsule formulat ion) or placebo three times a day for 5 days. On day 3, six subjects were g iven 7.5 mg oral midazolam and the other six subjects received 0.05 mg/kg i ntravenous midazolam. On day 5, the subjects who had received oral midazola m on day 3 received intravenously midazolam and vice versa. Plasma concentr ations of midazolam, alpha-hydroxymidazolam, and saquinavir were determined for 18 hours after midazolam administration, and midazolam effects were me asured up to 7 hours by six psychomotor tests. Results: Saquinavir increased the bioavailability of oral midazolam from 41 % to 90% (P < .005), the peak midazolam plasma concentration more than twof old, and the area under plasma concentration-time curve more than fivefold (P < .001). During saquinavir treatment, five of the six psychomotor tests revealed impaired skills and increased sedative effects after midazolam ing estion (P < .05). Saquinavir decreased the clearance of intravenous midazol am by 56% (P < .001) and increased its elimination half-life from 4.1 to 9. 5 hours (P < .01). After intravenous midazolam, only the subjective feeling of drug effect was increased significantly (P < .05) by saquinavir. Conclusion: The dose of oral midazolam should be greatly reduced or avoided with saquinavir, but bolus doses of intravenous midazolam can probably be used quite safely. During a prolonged midazolam infusion, an initial dose r eduction of 50% followed by careful titration is recommended to counteract the reduced clearance caused by saquinavir.