More than 1,200 detailed monographs cover more than 20,000 brand and generic drugs and more than 70 therapeutic classes. Every monograph summarizes the onset, severity, and documentation of clinically significant interactions, including their effects, mechanism, and management. Significance ratings provide relative rankings on the interactions.
Clinically important interactions between an herbal supplement and a drug typically manifest as pharmacokinetic interactions, which affect a drug's concentration in the blood and pharmacologic action. In many cases, pharmacokinetic interactions can be safely countered by adjusting the drug's dosage. Risk of a pharmacokinetic interaction occurs when an herbal supplement shares the same mechanism of absorption, distribution, metabolism, or excretion (ADME) as a coadministered drug. Competition between an herbal supplement and a drug for a shared ADME mechanism may result in a change in the drug's concentration at the site of action.
Less commonly, herb-drug interactions may manifest as pharmacodynamic interactions, which involve direct pharmacologic actions of an herbal supplement that are unrelated to changes in blood concentrations. Risk of a pharmacodynamic interaction occurs when an herbal supplement has a direct effect on the mechanism of action of a coadministered drug. Direct pharmacologic effects of an herbal supplement may antagonize or exacerbate the drug's clinical effects without changing the drug's concentration. In most cases, a change in drug dosage will not counter a pharmacodynamic herb-drug interaction.
Clinicians who are able to distinguish between pharmacokinetic/ADME interactions and pharmacodynamic interactions will be able to make better clinical decisions about whether to adjust the drug's dosage or discontinue the supplement. A change in drug dosage rarely results in a predictable change in clinical outcomes when a pharmacodynamic interaction occurs (e.g., warfarin [Coumadin] dosage adjustment with changes in daily leafy green vegetable intake).
Drug interactions are initially evaluated through in vitro systems. Although in vitro evaluations have high sensitivity and can be used to rule out potential herb-drug interactions, it is important to follow up positive in vitro findings with a human clinical trial to estimate the potential impact of an interaction on clinical outcomes. Many positive in vitro interactions have not been borne out in human trials, highlighting the importance of confirming potential interactions.
In a human clinical trial, cranberry (Vaccinium spp.) has been shown to have no inhibitory or induction effects on the drug-metabolizing enzymes CYP1A2, CYP2C9, and CYP3A4.16 Additionally, despite anecdotal case reports of cranberry increasing warfarin concentrations and international normalized ratio (INR), two human clinical trials did not show a significant effect on either outcome.17 As a result, the likelihood that cranberry has any clinically important drug interactions is low.
A single study showed that curcumin (Curcuma longa) induces CYP1A2, which could cause decreased levels of many antidepressant and antipsychotic medications.18 It has also been shown to increase sulfasalazine (Azulfidine) levels.19 However, a few human clinical trials have demonstrated no effect on several important enzymes, including CYP2C9, CYP3A4, and UGT.20 Given these differences in enzymatic effects, consultation with appropriate dietary supplement resources is needed to determine the potential for interaction between curcumin and many medications.
Ginkgo (Ginkgo biloba) is known to inhibit platelet aggregation, which could theoretically increase bleeding risk, especially in combination with antiplatelet or anticoagulant drugs. Several population-based and clinical studies, including a meta-analysis of 18 trials, failed to demonstrate that ginkgo increased bleeding risk or had significant effects on hematologic parameters.27 However, analysis of a single medical record database suggested an increased risk of bleeding with concurrent ginkgo and warfarin use.9 Patients taking warfarin should have their INR closely monitored or refrain from ginkgo use. Several human clinical trials have demonstrated no clinically important effects on CYP1A2, CYP2C9, CYP2C19, CYP2D6, CYP2E1, or CYP3A4.24,28
American ginseng (Panax quinquefolius), although similar to its Asian counterpart (Panax ginseng), has a somewhat different profile of ginsenosides; therefore, generalizations about Asian ginseng should not be extended to American ginseng. There are fewer human clinical studies of American ginseng than Asian ginseng, although preclinical studies support a low potential for drug interactions. Two human trials have demonstrated no effect of American ginseng on the human immunodeficiency virus (HIV) agents indinavir (Crixivan) and zidovudine (Retrovir).29,30 A single trial of American ginseng in healthy volunteers taking warfarin demonstrated a 0.2-point drop in INR.10 Patients taking warfarin should have their INR closely monitored or refrain from taking ginseng-containing supplements.
Asian ginseng (P. ginseng) has been shown in one study to induce CYP3A4, which could decrease the effectiveness of many drugs, including calcium channel blockers, many chemotherapy and HIV agents, certain antihypertensive and statin medications, and some antidepressants.31 For this reason, it is suggested to avoid use of Asian ginseng, including products containing Chinese, Japanese, and Korean ginseng, with most medications. However, several human trials have demonstrated that Asian ginseng has no effect on CYP1A2, CYP2D6, CYP2E1, or P-gp, so drugs metabolized by these enzymes may be safe to take concurrently with Asian ginseng.24,31,32 Studies of the effect of Asian ginseng on warfarin metabolism have yielded mixed results, alternately demonstrating no effect or small effects.33 In patients having difficulty maintaining adequate anticoagulation with warfarin, clinicians may suggest avoiding Asian ginseng products.
Goldenseal (Hydrastis canadensis) has been shown to inhibit two major metabolic enzymes, CYP2D6 and CYP3A4, which are responsible for metabolism of more than one-half of currently used pharmaceutical agents.13,15 Although some drug combinations with goldenseal may be safe, until data from further human clinical trials are available, clinicians should recommend against the use of goldenseal in combination with most other medications.
Green tea (Camellia sinensis) extract has been investigated for potential drug interactions in human and in vitro studies with conflicting results. In vitro study results have suggested potential interactions, whereas human clinical trials have not found any effects on the major metabolic enzymes CYP2D6 and CYP3A4.34 However, green tea extract has been shown to increase simvastatin (Zocor) concentrations,35 which may be due to P-gp inhibition. Two additional studies have demonstrated that green tea extract may inhibit the drug transporters OATP1A1 and OATP1A2, which are involved in the transport of many medications, including statins, fluoroquinolones, some beta blockers, imatinib (Gleevec), and antiretrovirals.36,37 Therefore, green tea extract should be avoided in combination with drugs that are transported by P-gp, OATP1A1, or OATP1A2.
Kava kava (Piper methysticum) has been shown in multiple human studies to have no effect on CYP1A2, CYP2D6, CYP3A4, or P-gp.13,15,38,39 In one study of human volunteers, kava inhibited CYP2E1, which is involved in the metabolism of several anesthetic agents, as well as acetaminophen.13 Additionally, the results of two in vitro studies suggest the potential to inhibit CYP2C9 and CYP2C19, which are involved in the metabolism of many nonsteroidal anti-inflammatory drugs, angiotensin receptor blockers, glipizide (Glucotrol), glyburide, rosiglitazone (Avandia), valproic acid (Depakene), warfarin, proton pump inhibitors, phenytoin (Dilantin), and clopidogrel (Plavix).40 Patients taking kava should be counseled to stop at least five days before surgery with general anesthesia. Patients taking medications metabolized by CYP2C9 or CYP2C19 should be closely monitored for clinical adverse effects and laboratory abnormalities (e.g., glucose level, A1C level, INR) or instructed not to use kava-containing supplements. Caution should be exercised in patients using central nervous system depressants, such as benzodiazepines or alcohol, because of the increased risk of drowsiness and motor reflex depression.
St. John's wort (Hypericum perforatum) has been shown in multiple human studies to be a potent inducer of CYP3A4 and P-gp.12,23,44 Clinical studies have shown reductions in cyclosporine (Sandimmune), tacrolimus, warfarin, protease inhibitors, irinotecan (Camptosar), theophylline, digoxin, venlafaxine, and oral contraceptives. It is strongly recommended to avoid concurrent use of St. John's wort with over-the-counter and prescription medications.
It is paramount that clinicians have a continuing, open dialogue with patients about their use of dietary supplements. Several studies have shown that although more than 17% of Americans take dietary supplements, only one-third of them inform their physician.5,46 Although patients do not expect their physician to be an expert about dietary supplements, they prefer to have him or her initiate the conversation about supplement use.47 It is important for clinicians to use current resources because new data from in vitro and human herb-drug interaction studies are being published regularly. Older resources and those without periodic updates may present misleading or incorrect information and recommendations. Examples of reliable sources include PubMed, Natural Medicines database, the Allied and Complementary Medicine Database, Lexi-Natural Products, and the National Institutes of Health's Office of Dietary Supplements (Table 3).
b37509886e