European Journal of Clinical Pharmacology

, Volume 61, Issue 1, pp 1–7

Milk thistle and indinavir: a randomized controlled pharmacokinetics study and meta-analysis


    • Faculty of Health Sciences, Clinical Epidemiology and BiostatisticsMcMaster University
  • Kumanan Wilson
    • Department of MedicineUniversity of Toronto
  • Mike Clarke
    • UK Cochrane Centre
  • Brian Foster
    • Therapeutic Products DirectorateHealth Canada
  • Scott Walker
    • Sunnybrooke and Women’s College Hospital
  • Beth Rachlis
    • Canadian College of Naturopathic Medicine
  • Nick DeGroot
    • Canadian College of Naturopathic Medicine
  • Victor M. Montori
    • Department of MedicineMayo Clinic
  • Wayne Gold
    • Department of MedicineUniversity of Toronto
  • Elizabeth Phillips
    • Department of MedicineUniversity of Toronto
    • Sunnybrooke and Women’s College Hospital
  • Stephen Myers
    • Department of MedicineUniversity of Queensland
  • Keith Gallicano
    • Watson Laboratories
Clinical Trials

DOI: 10.1007/s00228-004-0843-z

Cite this article as:
Mills, E., Wilson, K., Clarke, M. et al. Eur J Clin Pharmacol (2005) 61: 1. doi:10.1007/s00228-004-0843-z



To determine whether ingestion of milk thistle affects the pharmacokinetics of indinavir.


We conducted a three-period, randomized controlled trial with 16 healthy participants. We randomized participants to milk thistle or control. All participants received initial dosing of indinavir, and baseline indinavir levels were obtained (AUC0-8) (phase I). The active group were then given 450 mg milk-thistle extract capsules to be taken t.i.d. from day 2 to day 30. The control group received no plant extract. On day 29 and day 30, indinavir dosing and sampling was repeated in both groups as before (phase II). After a wash-out period of 7 days, indinavir dosing and sampling were repeated as before (phase III).


All participants completed the trial, but two were excluded from analysis due to protocol violation. There were no significant between-group differences. Active group mean AUC0-8 indinavir decreased by 4.4% (90% CI, −27.5% to −26%, P=0.78) from phase I to phase II in the active group, and by 17.3% (90% CI, −37.3% to +9%, P=0.25) in phase III. Control group mean AUC0-8 decreased by 21.5% (90% CI, −43% to +8%, P=0.2) from phase I to phase II and by 38.5% (90% CI, −55.3% to −15.3%, P=0.01) of baseline at phase III. To place our findings in context, milk thistle–indinavir trials were identified through systematic searches of the literature. A meta-analysis of three milk thistle–indinavir trials revealed a non-significant pooled mean difference of 1% in AUC0-8 (95% CI, −53% to 55%, P=0.97).


Indinavir levels were not reduced significantly in the presence of milk thistle.


Milk thistleIndinavirRandomized controlled trialPharmacokineticsDrug interactions

Copyright information

© Springer-Verlag 2005