Drug interactions with complementary medicines

Health professionals are expected to be familiar with common and clinically significant complementary medicine interactions or at least know where to look them up. Knowing the dynamic and kinetic interactions associated with commonly used complementary medicines helps to identify the risk of drug interactions. Although information on complementary medicine interactions is not readily provided by the manufacturers, evidence is available by way of case reports, independent research and web-based resources, which have increased in recent years. Collectively, these data make interactions with complementary medicines largely predictable and therefore preventable.

The Therapeutic Goods Administration refers to complementary medicines as 'medicinal products containing herbs, vitamins, minerals, nutritional supplements, homoeopathic medicines, traditional medicines and certain aromatherapy products'.¹In Australia, complementary medicines are largely regulated as unscheduled medicines, and are usually self-selected.

Complementary medicines are very popular among Australians, with surveys indicating that up to 60% of people use at least one complementary medicine on a regular basis. However, about 50% of consumers also report using a conventional medicine on the same day as their complementary medicine.^2,3It is not surprising, therefore, that healthcare professionals and consumers alike are concerned about the potential for drug interactions between these medicines.

As so many Australians use complementary medicines, including children, the elderly, patients with chronic disease, mental health disorders and cancer, it is important that prescribers always ask what complementary products their patients are taking in addition to any conventional medicines. Knowing this, and extrapolating reported pharmacodynamic and pharmacokinetic outcomes, can help predict potential drug interactions.

Complementary medicines are frequently used in the context of polypharmacy. A study of 3070 elderly people found that 74.2% took at least one prescription drug and one complementary medicine, with 32.5% of them using three or more prescription medicines with three or more complementary medicines.⁴This translates to an increased risk of drug interactions. In a study of 458 US Veterans' Administration patients, 197 of them reported taking complementary medicines combined with prescription medicines. Of these patients, 45% had potential for interactions, which was rated as serious in 6% of patients.⁵In another study which interviewed 3000 people (aged 57−85) about prescription, over-the-counter and complementary medicine use, 4% of them were potentially at risk of a major drug–drug interaction.⁶It has been suggested that once a patient is on eight or more medicines, regardless of origin, there is a 100% chance of a drug interaction occurring.⁷

As with other drugs, complementary medicine interactions can be broadly classified by their mechanism, that is, pharmacodynamic and pharmacokinetic. The former are due to overlap of pharmacological actions, while the latter result from changes in absorption, distribution, metabolism or excretion.

Risk factors for significant complementary medicine interactions are the same as for conventional medicines. These include patient characteristics (such as extremes of age, frailty, female gender, cognition, comorbities and genetic disposition) and medication factors (such as high medication burden, recent changes in medicines, drugs with a low therapeutic index and limited elimination pathways).

Due to their complex chemical structure, herbal medicines are prone to interactions via the oxidative cytochrome P450 system or the efflux drug transporter P-glycoprotein.^8,9In vitroassays, using human tissue or cell lines, are frequently used to determine whether a herb affects these enzymes.¹⁰However,in vitrofindings do not necessarily correlate with what happens in the human body. As several herbal medicines and many prescription drugs are substrates, inducers or inhibitors of CYP isoenzymes or P-glycoprotein, interactions can ensue when they are used concomitantly.⁹A classic example is St John's wort, which has kinetic interactions with a wide range of drugs via the induction of CYP1A2, CYP3A4, CYP2C9 and P-glycoprotein.¹¹This lowers the concentration of the concomitant drug.

Table 18,11,13,14 shows selected documented interactions which have been chosen based on a composite of:

• the most frequently used complementary medicines in Australia, from survey and sales data¹²
• interactions with serious or clinically significant outcomes.

Table 1 8,11,13,14 categorises interactions by their possible outcome, severity, supporting evidence and proposed mechanism. Generic guidance on interaction management is given in the key, within the definitions of severity (major, moderate, minor). Certain therapeutic drug classes appear repeatedly in the table such as antiplatelet drugs, anticoagulants, antidepressants, antihypertensives, hypoglycaemics, immunosuppressants, antiretrovirals and hormones. Health professionals should monitor patients closely when a complementary medicine is taken concomitantly with these drugs.

Table 18,11,13,14

Table 1 Evidence-based complementary medicine interactions8,11,13,14 This table shows complementary medicines with at least one 'major' interaction. Complementary medicine Interacting drug Possible outcome Severity and level of evidence* Proposed mechanisms/comment Black cohosh Cisplatin ↓ cytotoxic effect Moderate, level D Animal data only Hepatotoxic drugs e.g. high-dose paracetamol, alcohol ↑ risk of hepatotoxicity Moderate, level D Pharmacodynamic additive risk CYP2D6 substrates e.g. amitriptyline ↑ substrate levels and drug effect Moderate, level B Moderate inhibitor of CYP2D6 Calcium Quinolone antibiotics, sotalol, tetracycline, thyroxine ↓ drug effect Moderate, level B Formation of insoluble salts and decreased absorption Celery seed Thyroxine ↓ drug level and effect Moderate, level D Mechanism unknown but two case reports Chamomile (German) CYP1A2 and CYP3A4 substrates ↑ drug levels Moderate, level D Theoretical CNS depressants ↑ drug effect Moderate, level D Additive sedative effects Coenzyme Q10 Chemotherapy (e.g. alkylating drugs) ↓ cytotoxic effect Moderate, level B May counteract cytotoxic oxidative stress Antihypertensives ↑ drug effect Moderate, level B Coenzyme Q10 has an antihypertensive effect Warfarin ↓ drug effect Moderate, level D Coenzyme Q10 may have vitamin K-like effects Cranberry Warfarin ↑ drug effect Moderate, level B Cranberry has variable effects on CYP3A4, CYP2C9, CYP1A2 Echinacea CYP1A2 and CYP3A4 substrates e.g. clopidogrel (prodrug), olanzapine, warfarin ↑ substrate levels Moderate, levels B and D respectively Inhibition of CYP1A2 and CYP3A4 Evening primrose oil Antiplatelet drugs, warfarin ↑ drug effect Major, level B Contains gamma-linolenic acid, probable anticoagulant Fenugreek Hypoglycaemic drugs ↑ drug effect Moderate, level B Additive hypoglycaemic effect Antiplatelet drugs, warfarin ↑ bleeding risk Moderate, level D Antiplatelet activity Fish oil Antihypertensive drugs Additive blood pressure lowering Moderate, level B Additive blood pressure-lowering effect found with some antihypertensives. Antiplatelet drugs, warfarin ↑ bleeding risk Minor, level B Antiplatelet activity in high dose Contraceptives, oral ↓ fish oil effects Moderate, level B May decrease triglyceride-lowering effects Garlic Contraceptives, oral ↓ drug effect Moderate, level D Induces CYP3A4 Saquinavir/non-nucleoside reverse transcriptase inhibitors ↓ drug levels and effect Major, level B Induces CYP3A4 Antiplatelet drugs, warfarin ↑ bleeding risk Moderate, level D Theoretical antiplatelet activity Ginger Antiplatelet drugs, warfarin ↑ bleeding risk Moderate, level B Antiplatelet activity Ginseng (Panax) Hypoglycaemic drugs ↓ blood glucose Moderate, level B Hypoglycaemic effect. Conflicting evidence. Immunosuppressants e.g. azathioprine ↓ drug effect Moderate level B Largely theoretical CYP2D6 substrates ↑ substrate levels Moderate, level B CYP2D6 inhibitor but conflicting evidence Stimulants e.g. caffeine ↑ drug effects Moderate, level B Additive pharmacodynamic effect Ginkgo Anticonvulsants ↑ seizure risk Moderate, level D Large amounts of ginkgotoxin can cause neurotoxicity Warfarin, antiplatelet drugs ↑ bleeding risk Major, level D Antiplatelet activity after several weeks CYP2C9 substrates e.g. glipizide, warfarin, celecoxib ↑ substrate levels Moderate, level D Inhibits CYP2C9 activity CYP1A2, CYP2C19, CYP2D6 and CYP3A4 substrates ↑ substrate levels Moderate, level B Potentially inhibits these enzymes Hypoglycaemic drugs ↑ ↓ drug effect Moderate, level B Variably affects blood glucose concentrations Glucosamine Warfarin ↑ bleeding risk Major, level D Several case reports of increased INR Hawthorn Calcium channel blockers, nitrates, phosphodiesterase inhibitors ↑ drug effect Major, level D Additive vasodilator effects Digoxin, beta blockers ↑ drug effect Major, level D Additive effects on heart rate and/or blood pressure. Hawthorn has cardiotonic effects. Kava CNS depressants ↑ drug effect Major, level A Additive somnolence CYP1A2, CYP2D6, CYP2C9, CYP2E1, CYP3A4 substrates ↑ substrate levels Moderate, level B Kava potentially inhibits these enzymes P-glycoprotein substrates ↑ substrate levels Moderate, level D Lactobacillus acidophilus Immunosuppressants ↑ risk of infection Moderate, level D Antibiotics ↓ drug effect Moderate, level D Milk thistle CYP2C9 substrates e.g. amitriptyline, phenytoin, warfarin ↑ drug effect Moderate, level B Inhibits CYP2C9, glucuronidase and organic anion transporting polypeptide 1B1. Conflicting evidence. Noni juice Warfarin ↓ drug effect Moderate, level D Contains vitamin K Olive leaf Antihypertensive drugs ↑ drug effect Moderate, level B Additive antihypertensive effects Hypoglycaemic drugs ↑ drug effect Moderate, level B Additive hypoglycaemic effects Psyllium Carbamazepine, lithium ↓ drug effect Moderate, level D Decreases gastrointestinal absorption of other drugs Hypoglycaemic drugs ↑ drug effect Moderate, level B Additive hypoglycaemic effects Selenium Antiplatelet drugs, warfarin ↑ drug effect Moderate, level D Selenium dose of 10 microgram/kg/day can increase bleeding time Statins, niacin ↓ drug effect Moderate, level A Selenium plus beta carotene, vitamins C and E decreased the lipid-lowering effect Senna Digoxin, diuretics ↑ drug effect Moderate, level D Additive potassium loss with long-term use or high doses of senna St John's wort Alprazolam ↓ drug levels & effect Major, level B Increased clearance; half-life reduced by 50% Amitriptyline ↑ drug effect Major, level B Increased risk of serotonin syndrome Antidepressants, tramadol ↑ drug effect Major, level D Pethidine ↑ drug effect Major, level D Triptans ↑ drug effect Moderate, level D Clopidogrel ↑ bleeding risk Moderate, level B Increased conversion to active metabolite CYP1A2, CYP2C9, CYP3A4 substrates e.g. imatinib, indinavir, tacrolimus, carbamazepine, phenytoin ↓ drug levels & effect CYP3A4 =Major, level B CYP1A2, CYP2C9 = Moderate, level B Induces CYP enzymes Non-nucleoside reverse transcriptase inhibitors, protease inhibitors ↓ drug levels & effect Major, level B Induces CYP3A4 Oral contraceptives ↓ drug levels Major, level B Risk of breakthrough bleeding/contraceptive failure P-glycoprotein substrates e.g. digoxin, fexofenadine, irinotecan ↓ drug levels & effect Major, level B Induces intestinal P-glycoprotein Simvastatin ↓ drug levels Moderate, level B Statin levels reduced by up to 28% Warfarin ↓ drug effect Major, level B Induces CYP1A2, CYP2C9 and CYP3A4 Valerian Alprazolam ↑ drug levels Major, level B CYP3A4 inhibitor. Alprazolam increased by 19% in one study. CNS depressants ↑ drug effect Major, level D Pharmacodynamic effect CYP3A4 substrates ↑ substrate effect Moderate, level D Vitamin E Antiplatelet ,warfarin ↑ bleeding risk Moderate, level B Antiplatelet activity Chemotherapy ↓ drug effect Moderate, level D Possible antagonism of oxidative stress CYP INR CNS cytochrome P450 international normalised ratio central nervous system * Interaction rating adapted from Natural Medicines Comprehensive Database.11 The level of severity (major, moderate, minor) has been calculated using the evidence and probability of harm. This rating is linked with a generic recommendation for management. Major Strongly discourage patients from using this combination as a serious adverse outcome could occur. If used, patient should be monitored closely for potential adverse outcomes. Moderate Use cautiously or avoid combination as a significant adverse outcome could occur. If used, monitor for potentialadverse outcomes. Minor Be aware that there is a chance of an interaction. Advise patients of symptoms that may occur and an action planto follow. Level of evidence ratings: AHigh-quality randomised controlled trial or meta-analysis BNon-randomised clinical trial, literature review, clinical cohort or case-control study, historical control or epidemiologic study CConsensus or expert opinion DAnecdotal evidence; in vitro or animal study or theoretical based on pharmacology

Most complementary medicines are listed (AUST L) medicines, which are not subjected to the same rigorous premarketing safety and efficacy trials as registered (AUST R) medicines. Thus evidence of their interaction potential is often not available. In addition, manufacturers are not obliged to provide a consumer medicine information leaflet with advice or warnings regarding complementary medicine interactions.

Despite the lack of hard data, health professionals still need to make reasonable recommendations to patients about potential interactions. With a view to helping Australians make more informed decisions about using complementary medicines, an independent consortium from Mater Health Services Brisbane, Bond University and University of Queensland, with funding from the National Prescribing Service, evaluated complementary medicines information resources in 2008.¹²Specific criteria were used to identify 52 resources – 26 of these were shortlisted and assessed for technical quality, content and clinical utility. The quality of drug interaction information was also assessed in the review, specifically whether mechanisms were outlined, degree of severity was stated, and whether the absence of known drug interactions was disclosed. While many resources (free or subscription) had technical strengths, few had comprehensive interaction coverage. Those with some detail are included for further reading. Two of the highest ranked resources were online subscription databases, both of which contained reasonably comprehensive complementary medicine–drug interaction checkers. These were:

• Natural Standard which provides detailed monographs and brief summaries ('bottom line')
• Natural Medicines Comprehensive Database 

Consumers frequently use complementary medicines in combination with conventional medicines. For this reason, health professionals should always consider the potential for pharmacodynamic and pharmacokinetic interactions between them. High quality evidence is increasingly available for identification and prevention of these interactions.

Barnes J, Anderson LA, Phillipson JD. Herbal medicines. 3rd ed. London: Pharmaceutical Press; 2007. (also part of MedicinesComplete.

Stockley's Drug Interactions. Baxter K, editor. 9th ed. London: Pharmaceutical Press; 2010. (also part of MedicinesComplete, www.medicinescomplete.com)

... L, Cohen M. Herbs and natural supplements: an evidence-based guide. 3rd ed. Sydney: Elsevier Australia; 2010.

The review of natural products. Facts & Comparisons.

Natural & alternative treatments. EBSCO.

Kuhn MA, Winston D. Winston and Kuhn's herbal therapy and supplements: a scientific and traditional approach. 2nd ed. Philadelphia: Lippincott Williams and Wilkins; 2008.

Therapeutic Goods Administration. New labelling requirements and consumer information for medicines containing Black cohosh. Update 29 May 2007.

Therapeutic Goods Administration. Kava – safety alerts & advisory statements.

Conflict of interest: none declared