Grapefruit juice does not enhance the effects of midazolam and triazolam in man

Objectives: Since grapefruit juice (Gra) inhibits hepatic P450 (CYP3A4), we studied its potential to enhance the effects of midazolam (Mid) and triazolam (Trz), which are metabolized by the CYP3A4 isoenzyme. Methods: In Study I parallel groups of healthy students were given orally Mid 10 mg with water or grapefruit juice (GraMid), two placebo groups receiving water or Gra. The effects of Mid were measured by psychomotor tests and by self-rating on visual analogue scales before and 30 and 90 min after intake. Study II was similar, but the post-treatment tests were at 45 and 90 min, and the active drugs used were 0.250 mg Trz, GraTrz, and Mid 10 mg. In the crossover Study III, 6 subjects took Mid 10 mg alone and with Gra (GraMid) and 750 mg erythromycin (EryMid). Performance tests were made and blood was sampled before and 30, 60 and 90 min after intake. Midazolam and its active metabolite α-OH-midazolam were assayed by gas chromatography (GC) and radioreceptor assay (RRA). Results: In Study I, both Mid and GraMid impaired digit symbol substitution (DSS), letter cancellation (LC) and flicker fusion (CFF) at 90 min. GraMid had more effect (P < 0.05) than Mid on the DSS performance. Mid caused drowsiness at 30 and 90 min. Both Mid and GraMid caused clumsiness and a feeling of impaired performance at 90 min. In Study II, the active drugs impaired objective test performances (DSS, LC, CFF) at 90 min, without having a clear subjective effect. In Study III, Mid, EryMid and GraMid impaired performance in the DSS, LC and CFF tests. EryMid proved stronger than Mid and GraMid on DSS and LC tests at 30 min. Mean values of plasma midazolam (and α-OH-midazolam) at 30, 60, 90 and 120 min after Mid 10 mg were 68 (19), 61 (19), 43 (14) and 42 (12) μg⋅l⁻¹. The corresponding values after EryMid were 164 (14), 137 (13), 104(10) and 89(10) μg ⋅l⁻¹, and after GraMid 60 (12), 69 (16), 61 (15) and 57 (14) μg⋅l⁻¹. Conclusions: The grapefruit juice used did have any particular interaction with oral doses of 10 mg midazolam and 0.25 mg triazolam in healthy young subjects. Received: 4 September 1995/Accepted in revised form: 5 January 1996