The present findings suggest that phytochemical components in the milk thistle and black cohosh formulations investigated in this study were not potent modulators of P-gp activity in vivo, and therefore do not pose a significant interaction risk with digoxin. This interpretation is bolstered by the significant changes in digoxin pharmacokinetics observed following the administration of clarithromycin, a known P-gp inhibitor, and rifampin, a recognized inducer of P-gp expression. In addition, our results do not support previous in vitro findings that milk thistle flavanolignans (i.e., silymarin) inhibit P-gp mediated digoxin efflux, at least not in the context of recommended supplementation regimens. The discrepancy may stem from the fact that silymarin is practically insoluble in water and that in vitro studies demonstrating an inhibitory effect of milk thistle flavanolignans on P-gp activity have utilized dimethylsulfoxide as a solubilizing agent(Maitrejean M., 2000
; Zhang and Morris, 2003a
). The product used in the present study was formulated with soybean oil, glycerin, and lecithin in a soft gelatin capsule, and upon disintegration the contents appeared to remain undissolved. Since flavanolignan plasma concentrations were not measured, any indication as to their in vivo solubility and/or bioavailability status remains unknown. Nevertheless, bioavailability and dissolution characteristics for silymarin-containing products have been shown to vary widely. An evaluation of nine separate silymarin-containing products found that the amount of silibinin released over one hour into an aqueous buffered solution (pH 7.5, 37°C) ranged from 0–85%(Schultz et al., 1995); while a comparative bioavailability study of three silibinin-containing dosage forms found that values for AUC and Cmax
varied among products by factors of 3 and 6, respectively(Kim et al., 2003
). Moreover, several studies have demonstrated that silymarin-containing products exhibit especially poor bioavailability and drug-release properties when not formulated with solubility-enhancing agents (i.e., phosphatidyl choline and polyethylene glycol)(Schandalik R, et al., 1992
; Schultz et al., 1995; Savio D, et al., 1998
; Li and Hu, 2004
). From the results of the present study, it would appear that local silymarin concentrations at intestinal enterocyte membrane interfaces were lower than the 50-200μM necessary for in vitro inhibition of P-gp(Maitrejean M., 2000
; Zhang and Morris, 2003a
; Patel et al., 2004
). Nevertheless, other in vivo evidence suggests that milk thistle’s principal flavanolignan, silibinin, is preferentially excreted into bile(Schandalik et al., 1992
), lending further support for this compound as a substrate for P-gp and/or some other apical efflux transporter in the liver.
While the milk thistle supplement used here was not as potent an inducer of P-gp as rifampin, we did observe a reduction in digoxin AUC that approached statistical significance. Others have also noted that milk thistle supplementation produced non-statistically significant reductions in AUC(0–8)
for the dual P-gp/CYP3A4 substrate, indinavir(Piscitelli et al., 2002
; DiCenzo et al., 2003
; Mills et al, 2005
). Somewhat more compelling was the effect of milk thistle on metronidazole, another substrate of both P-gp and CYP3A4(Rajnarayana et al., 2004
). Nine days of silymarin therapy (140 mg per day) produced a statistically significant increase in the apparent oral clearance (Cl/F) of metronidazole, as well as significant reductions in AUC, elimination half-life, and Cmax
, leading the authors to conclude that the flavanolignans induced both intestinal P-gp and CYP3A4. In contrast, 28 days of milk thistle supplementation (317 mg silymarin daily) failed to affect the metabolism of the CYP3A4 substrate, midazolam(Gurley et al. 2004
). Such distinctions may be related to the use of different milk thistle formulations or they may suggest that flavanolignans have a greater affinity for P-gp than for specific CYP isoforms.
From the human studies conducted to date, it would appear that the herb-drug interaction potential for milk thistle is relatively low; however, variability in product formulation, dissolution, and bioavailability may render this interpretation tenuous. Thus, it remains to be determined whether our findings involving digoxin can be extrapolated to other milk thistle dosage forms. Further studies with various milk thistle formulations will be needed in order to better clarify the herb-drug interaction potential of this dietary supplement.
When compared to clarithromycin and rifampin, our findings also indicate that the dose of black cohosh triterpene glycosides used in this study (1.5 mg) did not affect digoxin disposition and therefore are not potent modulators of P-gp activity. Currently, no in vitro studies have examined black cohosh extracts or individual triterpene glycosides for their effect on P-gp or other xenobiotic transporters; however, a recent in vivo study found that 28 days of black cohosh supplementation (10.8 mg triterpene glycosides daily) did not affect human CYP activity(Gurley et al., 2005
). Taken together these results suggest that black cohosh supplementation poses a minimal risk for engendering clinically relevant herb-drug interactions.
With regard to the effects of rifampin on digoxin pharmacokinetics, our findings are in agreement with those of Greiner et al., who observed significant reductions in AUC and Cmax
for orally administered digoxin, no change in terminal elimination half-life, and an increase in duodenal P-gp expression, suggesting that rifampin-mediated upregulation of ABCB1
is more prevalent in the intestine than in the kidney(Greiner et al., 1999
). As for clarithromycin, significant increases that occurred in digoxin AUC, Cmax
, elimination half-life, and a reduction in systemic clearance are in accordance with previous findings(Rngelshausen et al., 2002), confirming clarithromycin as a potent inhibitor of both intestinal and renal P-gp. Together, rifampin and clarithromycin appear to be acceptable benchmarks in which to gauge the clinical magnitude of herb-mediated changes in P-gp activity.
Consideration must be given to the fact that the ABCB1
gene is polymorphic and several SNPs have been associated with altered digoxin disposition(Johne et al., 2002
; Kurata et al., 2002
); therefore, an absence of herb-mediated changes in digoxin disposition could be related to specific subject ABCB1
haplotypes. All subjects were genotyped for SNPs at exons 21 (G2677T) and 26 (C3435T) and while no statistically significant haplotype-associated changes in digoxin AUC were observed, subjects exhibiting the GC-GC haplotype had lower mean AUC values at baseline (each subject had four separate baseline assessments of digoxin AUC), and those possessing the GC-TT haplotype had greater mean AUC values. (). These findings are in general agreement with those of Kurata et al., who noted that digoxin AUCs were correlated with haplotype and could be ranked as follows: GC-GC < GC-TT < TT-TT(Kurata et al., 2002
). Kurata et al. also found that individuals nullizygous at both loci (i.e., TT-TT) appeared to have a lower expression of functional P-gp and were less susceptible to increases in digoxin bioavailability following clarithromycin administration. Like Kurata et al., we observed that clarithromycin administration appeared to have less of an effect on digoxin bioavailability for TT-TT subjects (see AUC(0–3)clarith.
, ). Unlike Kurata et al., however, we found that clarithromycin gave rise to greater increases in digoxin AUC for the GC-TT haplotype as opposed to GC-GC. Interestingly, rifampin’s effect on digoxin AUC appeared greatest in our two nullizygous subjects (see AUC(0–3)Rif.
, ). These outcomes, however, did not reach statistical significance, and are likely due to the small number of subjects exhibiting each haplotype (). Taken together, our findings hint that nullizygous individuals may be less vulnerable to P-gp inhibition, and more susceptible to induction. Future studies incorporating greater numbers of individuals with each haplotype will be needed to verify this result.
In conclusion, when compared to the effects of rifampin and clarithromycin, the botanical supplements milk thistle and black cohosh produced no significant changes in the disposition of digoxin, a clinically recognized P-gp substrate with a narrow therapeutic index. Accordingly, these two supplements appear to pose no clinically significant risk for P-gp-mediated herb-drug interactions. However, given the inter-product variability in phytochemical content, potency, and formulation among botanical supplements, these results may not extend to regimens utilizing higher dosages, longer supplementation periods, or products with improved dissolution and/or bioavailability characteristics.