EFFECT OF MILK THISTLE ON PHARMACOKINETICS OF INDINAVIR Piscitelli et al
555
before her final pharmacokinetic analysis,
although she was not anemic. Participants
complied with dietary restrictions against
inducers or inhibitors of CYP except for one
participant who consumed broccoli soup during
washout.
At the end of study, all but two participants had
laboratory values within normal limits. One had
a slight elevation in fasting blood glucose level,
and one had hepatic aminotransferase levels just
above the upper limits of normal; in both
participants, test results were normal on repeat.
indinavir. Because all protease inhibitors and
nonnucleoside reverse transcriptase inhibitors are
metabolized at least in part through the CYP3A4
pathway, it is unlikely that milk thistle would
alter significantly the pharmacokinetics of any of
these antiretroviral agents. Although the AUC8
values were similar, the trough level of indinavir
was decreased by approximately 25% in the milk
thistle phase of the study compared with that at
baseline. In the washout phase, Cmax, AUC8, and
C8 were 22-28% decreased from baseline. This
suggests a time-dependent but not a milk
thistle–dependent effect; however, an alternative
explanation might be the delayed generation of a
milk thistle metabolite with long-lived effects on
indinavir concentrations. If this were the case,
milk thistle might have some long-term effect on
concentrations of concomitant indinavir, but the
effects are fairly small and might not be of
clinical importance.
Some individual patients had large decreases in
plasma concentrations. The greatest decrease in
an individual patient was a 60% decrease in the
indinavir trough. Since the trough level has been
correlated with virologic outcome in a number of
studies, a decrease of this magnitude might be
clinically significant. If milk thistle does cause
some modest and prolonged effect on indinavir
levels, the mechanism is unclear and its
occurrence is highly variable.
Discussion
Results of clinical trials suggest that milk
thistle might be more effective than placebo for
viral hepatitis and for hepatitis caused by toxins
and alcohol.7, 8 Milk thistle also has been
postulated to prevent liver damage caused by a
variety of drugs including phenytoin, halothane,
and phenothiazines.7 These studies generally
were performed with small samples, hetero-
geneous populations, and various dosages, and
many were not well controlled for factors such as
discontinuation of alcohol intake. Despite these
limitations on data, milk thistle has become a
common adjunctive therapy in patients with HIV
infection who are at risk of liver disease from
antiretroviral drugs or hepatitis B or C infection.
Milk thistle is reported to be generally well
tolerated, with adverse effects consisting
primarily of loose stools at high dosages and mild
allergic reactions.3
Little is known regarding the drug interaction
potential for milk thistle. Animal studies have
suggested an inhibitory effect on the CYP system,
which is responsible for metabolism of protease
inhibitors and nonnucleoside reverse transcriptase
inhibitors.4, 5 A clinical study examining the effect
of 28 days of therapy with silymarin showed no
effect on the metabolism of aminopyrine and
phenylbutazone, but neither of these drugs is
specific for the CYP3A4 isoform that mediates
protease inhibitor metabolism.9 In addition, the
dosage of silymarin administered in that trial was
lower than that commonly administered in other
clinical studies or for treatment of liver disease.
Thus, we chose to evaluate the effect of standard
dosages of milk thistle on plasma concentrations
of indinavir, a protease inhibitor known to be
significantly affected by CYP induction.
Our study also demonstrated that in vitro
studies of herbal products may give very different
results than those seen in a clinical trial.
Although two in vitro studies predicted milk
thistle would increase indinavir concentrations,4,
5
no such effect was seen in this study. Similarly,
a study in hepatic microsomes demonstrated that
St. John’s wort is an inhibitor of CYP3A4, but the
opposite effect was demonstrated in a number of
studies and case reports.1, 10–12 The reasons for
the discordance are varied and include not
studying the true active ingredient, using
concentrations that are not relevant clinically,
using an in vitro system that can only evaluate
inhibition, and not being able to evaluate the
effects of any metabolites that may be produced.
In general, results of in vitro studies with herbal
products should not be used as a basis for
therapeutic decisions on drug interactions and
require confirmation in a clinical trial.
Our knowledge of herb-drug interactions is
rapidly expanding. Some drugs have potent
effects on antiretroviral agents (St. John’s wort,
garlic); milk thistle does not appear to do so.
Our study assessed the effect of milk thistle on
Our study demonstrated that 3 weeks of dosing
with a commercial milk thistle product did not
alter substantially the pharmacokinetics of