The present study is highlighted by two principal findings. The first is that single time-point phenotypic ratios can provide a practical method for identifying herb-drug interactions that involve CYP inhibition. Although moderate inhibition of CYP2E1 by garlic oil had been demonstrated earlier,26
the significant reductions in CYP2D6 and CYP3A4/5 phenotype following goldenseal extend the method’s utility to include these two important CYP isoforms. Previously, this approach had been used to document CYP3A4/5 induction as evidenced by significant elevations in 1-hydroxymidazolam/midazolam ratios following St. John’s wort supplementation. The method also demonstrated that an absence of change in mean phenotypic ratios following botanical supplementation could be interpreted as a lack of effect on CYP activity. Such was the case with milk thistle, Citrus aurantium, Ginkgo biloba
, Panax ginseng,
and saw palmetto extracts—the latter three examples being confirmed by other investigators using more conventional area-under-the-curve assessments.24
Thus, a range of herb-mediated effects on CYP activity (e.g. induction, inhibition, or no effect) can be differentiated with single time-point phenotypic ratios. It must be emphasized, however, that single-time point phenotypic ratios simply provide estimates of probe drug clearance. Yet, even with this limitation, the method’s distinct advantage lies in an ability to evaluate multiple CYP enzymes and multiple botanical supplements in vivo
while using a limited blood-sampling scheme.
The second important finding emanating from the study is that goldenseal appears to inhibit CYP2D6 and CYP3A4/5 in vivo
, which implies a significant pharmacokinetic herb-drug interaction potential for this supplement. These findings bolster recent in vitro
investigations demonstrating inhibition of CYP2D6- and CYP3A4-mediated biotransformations by goldenseal extracts. Using human hepatic microsomes, Chatterjee and Franklin50
found that goldenseal extract as well as its two principal isoquinoline alkaloids, berberine and hydrastine, inhibited CYP2D6-mediated bufuralol 1′-hydroxylation. Of the two alkaloids, berberine was more inhibitory toward bufuralol 1′-hydroxylation (IC50
= 45 μM) than hydrastine (IC50
= 350 μM), implying a greater contribution of this phytochemical to CYP2D6 inhibition. When evaluating a series of single-entity herbal tea extracts, Foster et al noted that Hydrastis canadensis
produced the greatest percent inhibition of cDNA expressed human CYP2D6.14
Budzinski et al first noted that commercial extracts of Hydrastis canadensis
were potent in vitro
inhibitors of CYP3A4.12
Chatterjee and Franklin later observed that goldenseal extracts as well as individual isoquinoline alkaloids inhibited CYP3A4-mediated testosterone 6β-hydroxylation.50
In the case of CYP3A4, however, hydrastine was more inhibitory (IC50
= 25 μM) than berberine (IC50
= 400 μM). Inactivation of the enzyme appears to stem from the methylenedioxyphenyl moiety of hydrastine interacting with the heme iron of CYP3A4 to produce a stable heme-adduct. These adducts, termed CYP metabolic-intermediate (MI) complexes provide a mechanistic basis for the inhibition of CYP3A4 by goldenseal, and possibly CYP2D6.50
Little is known about the pharmacokinetics of goldenseal alkaloids in humans, but animal studies indicate that berberine bioavailablity is relatively low.51,52
Although the daily dose of isoquinoline alkaloids ingested in the present study was 142 mg (), plasma concentrations of berberine and hydrastine were not determined; nevertheless, the significant effect observed on CYP2D6 phenotype indicates that phytochemicals present in goldenseal can cross the intestinal mucosa. Whether the effect on CYP3A4/5 are limited to intestinal enterocytes or extends to hepatocytes remains to be determined. Interestingly, the only prospective study to date in which the influence of goldenseal supplementation on the pharmacokinetics of a CYP3A4 substrate (indinavir) has been evaluated failed to register any significant effects.53
This may stem from the relatively high oral bioavailability of indinavir, which renders the drug less effective as a probe for assessing herb-mediated changes in intestinal CYP3A4/5 activity.
The statistically significant reduction in debrisoquine urinary recovery ratios following black cohosh supplementation implies a much weaker inhibitory effect on CYP2D6 for this botanical extract than for goldenseal; however, the magnitude of this result (~7%) may not be clinically relevant. A trend toward CYP3A4/5 inhibition (~14%, ) was also noted for black cohosh, but the magnitude was not statistically significant (p = 0.09). Black cohosh’s complex phytochemical makeup54,55
coupled with an absence of in vitro studies into the extract’s effect on CYP isoforms, render a connection between CYP2D6 (and possibly CYP3A4/5) inhibition and a specific marker compound(s) especially difficult. As with many botanical extracts, the pharmacokinetic profile of black cohosh’s constituent phytochemicals has yet to be investigated. One group has reported their attempt at quantitating mercapturate conjugates of black cohosh constituents (fukinolic acid, fukiic acid, caffeic acid, and cimiracemate B) in the urine of women who consumed up to 256 mg of a standardized black cohosh extract;56
however, none of the target conjugates were detected, which brings into question the bioavailability of these specific phytochemicals. Nonetheless, black cohosh supplements appear to have a good safety profile57
and exhibit modest efficacy when used to alleviate perimenopausal and postmenopausal symptoms;58
however, their pharmacokinetic herb-drug interaction profile remains uncertain. Therefore, until further studies are conducted, caution should be exercised when taking these supplements in conjunction with conventional medications.
South Pacific islanders have long consumed “traditional” kava beverages prepared from coldwater extracts of powdered kava roots.59
These “traditional” preparations are imbibed in social, recreational, and ceremonial settings to imbue psychotropic, hypnotic, and anxiolytic effects. Since the 1990s, commercial kava extracts, prepared with nonaqueous solvents (acetone, ethanol or methanol) and formulated as tablets and/or capsules, have been marketed as dietary supplements for the alleviation of stress, anxiety, or insomnia.34,62
Recently, a spate of reports linking kava use to liver toxicity has prompted the removal of these products from Europe, Australia, and Canada.36-38,59,60
In the United States the FDA has issued a warning to consumers alerting them to possible hepatotoxic side effects associated with kava supplementation.61
For the majority of case reports documenting possible kava-related hepatotoxicity, comedication with prescription drugs and/or other botanical supplements has been a confounding, although common, variable.58,59
This raises the question as to whether an underlying kava/drug interaction may be responsible.
Kava lactones, which include methysticin, dihydromethysticin, kavain, dihydrokavain, yangonin, and desmethoxyyangonin, are the active principles of kava extracts. Using c-DNA-expressed CYPs, human liver microsomes, and/or cryopreserved hepatocytes, both kava extracts and specific kava lactones have been shown to inhibit a variety of human CYP isoforms in vitro, with individual IC50
values ranging from 1 to 100 μM.18,19,21,34
Of the kava lactones examined, methysticin, dihydromethysticin, and desmethoxyyangonin appear to be the most potent CYP inhibitors, with CYP3A4 being affected by all three.19,21
In contrast, Raucy noted that 100 sg/mL of kava extract (a concentration not likely achievable in vivo) induced CYP3A4 in human hepatocytes and this effect was SXR-mediated.15
Collectively, the in vitro studies suggest that kava supplementation may give rise to significant CYP-mediated herb/drug interactions; however, the in vivo data presented herein suggests otherwise.
Phenotypic metabolic ratio comparisons indicated that 28 days of kava supplementation did not significantly affect CYP1A2, CYP2D6 or CYP3A4/5 activity in healthy human volunteers, although CYP2E1 phenotype was decreased appreciably (~40%). In light of the preponderance of in vitro data, it is unclear why we observed an inhibitory effect of kava only on CYP2E1, and not for the other isoforms. This disconnect might stem from lower concentrations of kava lactones realized in vivo relative to those achieved in vitro. Perhaps the daily administered dose of kava lactones (138 mg, ) was simply too low to elicit any clinically observable effects on CYPs other than 2E1, a factor that could also account for a lack of any observable kava-related liver toxicity in our subjects. Currently, there are no published studies describing the pharmacokinetics of kava lactones in human subjects, therefore, the relationship between in vitro IC50
values and steady state plasma concentrations of kava lactones remains unknown. Interestingly, methysticin and dihydromethysticin, like hydrastine and berberine, contain methylenedioxyphenyl moieties that can form metabolic-intermediate (MI) complexes with CYP isoforms.19
Assuming formation of MI complexes lies at the heart of CYP inhibition by kava and goldenseal, other factors (e.g. differences in membrane permeability,62
carrier-mediated transport, or other physicochemical properties) may contribute to the in vitro/in vivo discrepancies observed for kava lactones and the isoquinoline alkaloids. A methodological explanation for the observed absence of CYP3A4/5 or CYP2D6 inhibition following kava administration seems unlikely since single-time point phenotypic ratios clearly identified an inhibitory effect for goldenseal.
Regardless of the disparity between previous in vitro studies and the clinical results presented here, a considerable body of evidence suggests that kava can modulate human CYP activity. While we observed no appreciable changes in CYP3A4/5, CYP2D6, or CYP1A2 phenotype at a daily kava lactone dose of 138 mg, these results may not extend to doses exceeding this value. Therefore, to guard against possible CYP-mediated interactions, coadministration of kava with conventional medications should be avoided.
An absence of effect on CYP phenotype for the valerian product used in this study could stem from a paucity of valerenic acid and a long disintegration time (42 minutes, ). Wide variability in valerenic acid () content among commercial brands is not uncommon;35,63
however, the presence of only trace amounts of the principal marker compound (valerenic acid) render this product less suitable for evaluating valerian’s influence on CYP activity in vivo. In vitro examinations of valerian extracts on CYP3A4 activity have provided mixed results even though the same general type of assay (fluorometric microtitre plate assay, GENTEST™
) was used.12,16,35
Strandell et al16
and Lefebvre et al35
both found valerian extracts to have moderate to potent CYP3A4/5 inhibitory capabilities, while Budzinski et al ranked them as poor inhibitors.12
Recenty, Donovan et al assessed the effects of a valerian supplement on alprazolam pharmacokinetics and dextromethorphan metabolic ratios in healthy volunteers.33
The valerian product used by Donovan contained a total valerenic acid content (valerenic acid, acteoxyvalerenic acid, and hydroxyvalerenic acid) of 5.5 mg/tablet. No significant changes in dextromethorphan metabolic ratios and only a slight increase in alprazolam Cmax values led the authors to conclude that “typical doses of valerian are unlikely to produce clinically significant effects on the disposition of medications dependent on the CYP2D6 or CYP3A4/5 pathways of metabolism.”33
In conclusion, single-time point CYP phenotypic ratios indicated that goldenseal supplementation significantly inhibited human CYP3A4/5 and CYP2D6 activity in vivo. Therefore, in order to avoid potentially serious pharmacokinetic interactions, goldenseal supplements should not be taken concomitantly with conventional medications. Black cohosh exhibited mild inhibition of CYP2D6, however, the clinical relevancy of this effect remains uncertain. Of the four CYP isoforms investigated, kava supplementation significantly inhibited CYP2E1 but caused no elevations in serum liver enzymes. Although valerian produced no marked changes in CYP phenotypes, this result may not be representative of products containing greater quantities of valerenic acid. Taken together, these results provide further evidence that botanical supplementation can modulate human CYP activity in vivo, which may lead to significant herb-drug interactions.