Interactions between complementary medicines and drugs used in primary care and oral COVID-19 antiviral drugs

Unwanted drug interactions with complementary medicines (CMs) can be avoided by being aware of the risks and discussing them with patients.¹ CMs contain nutritional and herbal ingredients that are complex, multiconstituent compounds with demonstrable pharmacological actions.² As such, the use of CMs is associated with potential benefits and harms, including harm from CM–drug interactions.³

Approximately 50% of adults living in Australia reportedly use CM products. Of these, 80% also use prescription and over-the-counter drugs, and 73% have chronic conditions that further increase their risk of unwanted drug interactions due to polypharmacy.^4,5

Aim

This narrative review provides an overview of selected CM–drug interactions reported in human studies that may occur when CMs are used with drugs that are commonly prescribed in Australian general practice and reviews potential interactions between CMs marketed for respiratory tract infections or immune support in Australia and drugs used to treat COVID-19.

Methods

This review article summarises the findings from systematic reviews and comprehensive narrative reviews and other primary studies known to the authors. Interactions between CMs and medicines typically only prescribed in secondary care settings are not reported. Given the emerging evidence for interactions between CMs and drugs used to treat COVID-19, product information documents for drugs and reputable interaction checkers listed in Table 1 were also used to identify interactions. 

Mechanisms of drug interactions

Drug interactions can occur when two or more compounds taken within a certain period of time alter the pharmacokinetics or pharmacodynamics of either compound. Both are implicated in an increased or decreased clinical effect that is not observed when either compound is used alone.⁶

Pharmacokinetic interactions change the systemic concentration of a compound and/or its active metabolites by altering absorption, distribution, metabolism or excretion.⁶ Most CM–drug pharmacokinetic interactions are associated with the induction or inhibition of cytochrome P450 (CYP) enzymes or transporters such as P-glycoprotein and organic anion transporting polypeptides.⁴ Approximately 80% of drugs used in clinical practice are metabolised by CYP enzymes.⁷

Pharmacodynamic interactions occur when the compounds have negative, additive or synergistic effects on drug targets.⁸

Determining clinically important interactions

Preclinical studies, such as in vitro and animal studies, are useful for identifying potential CM–drug interactions and the mechanisms by which they occur. Human studies are still required to determine the clinical relevance of any such interactions.⁶ As such, for the purpose of the present narrative review, our focus is on human studies.

Many of the CM–drug interactions flagged by interaction checkers or in the drug product information or CM monographs are theoretical and based on knowledge of shared mechanisms of action and/or shared metabolism. Sometimes, they are inferred from the findings of studies evaluating other drugs with similar pharmacological properties or in vitro and animal studies.⁶ However, there are a growing number of case reports, case series and controlled trials reporting clinically important CM–drug interactions.⁹

Collectively, this evidence warrants attention, especially for drugs that have a narrow therapeutic index or when there are potentially serious consequences.⁶ These include drugs used for contraception, mental health, epilepsy, cancer, immunosuppression, infections, diabetes and cardiovascular conditions.⁶

However, not all CM–drug interactions are necessarily unwanted or undesirable, and some could improve clinical outcomes (eg by augmenting the clinical effects of the drug, increasing drug concentrations or reducing side effects).⁹

CM–drug interactions relevant to general practice

Table 2 summarises some of the clinical research on CM–drug interactions relevant to general practice. Notable in some reports, altered plasma or serum concentrations of a drug were not always associated with altered clinical effects. For example, Echinacea purpura was found to lower warfarin levels, yet there was no significant change in INR.¹⁰ 

Given this review is not comprehensive, practitioners should also refer to interaction checker databases to identify other potentially important interactions (Table 1).

Table 2 lists numerous examples when a CM altered the plasma or serum concentrations of a drug, yet it is unknown whether these changes are clinically important. Clinical importance cannot be assumed; for example, Echinacea purpurea (with or without Echinacea angustifolia) lowered warfarin levels, yet this did not appear to affect the international normalised ratio of prothrombin time (INR).^9,10,14 Similarly, it was unclear whether the small increase in digoxin levels from coadministration with Hydrastis canadensis (goldenseal) was clinically important.^14,18

There were quite a few instances where the clinical outcomes from a CM–drug interaction could be favourable. Many of these interactions were due to the additive hypoglycaemic effects of the CM, which improved glycaemic control when coadministered with a hypoglycaemic drug.^12.13 Depending on the clinical circumstances, close monitoring may still be advisable to reduce the risk of a hypoglycaemic event. Other examples include specific species and strains of probiotic bacteria that may improve vaccine efficacy and duration of protection.²⁹

Coincidentally, except for Hypericum perforatum (St John’s wort), the common name for many of the CMs commonly implicated in clinically important drug interactions begins with the letter ‘G’: Allium sativum (garlic), Zingiber officinale (ginger), Ginkgo biloba (ginkgo), Hydrastis canadensis (goldenseal), Camellia sinensis (green tea), Gymnema sylvestre (gymnema), Panax quinquefolius (American panax ginseng), Panax notoginseng (Korean/Chinese ginseng) and Eleutherococcus senticosus (Siberian ginseng) and glucosamine. Clinically important drug interactions are summarised below.

Pharmacokinetic interactions

Interactions of P-glycoprotein and CYP enzymes with several herbs are relatively common. It is well established that Hypericum perforatum (St John’s Wort) and one of its bioactive constituents, hyperforin, are potent inducers of intestinal P-glycoprotein and several CYP enzymes, specifically cytochrome P450 family 3 subfamily A member 4 (CYP3A4), which is involved in the metabolism of over 50% of commonly prescribed drugs.³⁰ Extracts with hyperforin doses over 0.3 mg have a substantially increased risk of a drug interaction.^30,31 Consequently, some Hypericum perforatum (St John’s wort) products standardise both the hypericin and hyperforin content.

Clinical studies have confirmed that Hypericum perforatum (St John’s Wort) alters the expected pharmacokinetics of drugs commonly used for the prevention and treatment of cardiovascular disease, diabetes and contraception (Table 2).³² In addition, the product information for numerous other drugs often includes an interaction warning for Hypericum perforatum (St John’s wort) when it has been shown to dramatically alter the pharmacokinetics of another drug with similar pharmacokinetic properties.³³

Allium sativum (garlic) is known to substantially increase the intestinal expression of P-glycoprotein and inhibits several CYP enzymes. Thus, there is a risk of pharmacokinetic interactions between garlic and numerous drugs (Table 2).³⁴ The risks may be higher with Allium sativum (garlic) products that have a high allicin content (Table 2).³⁵

Hydrastis canadensis (goldenseal) and its main bioactive constituents berberine and hydrastine inhibit cytochrome P450 family 2 subfamily D member 6 (CYP2D6), CYP3A and P-glycoprotein.^14,36 The clinical importance of this is uncertain given the results of one clinical trial in which digoxin concentrations were only slightly increased by goldenseal (Table 2).¹⁸

Interactions between Camellia sinensis (green tea/green tea supplements) and cardiovascular drugs that are metabolised by CYP3A or transported by P-glycoprotein, organic anion transporting polypeptide (OATP) 1A1 or OATP1A2 are another potential risk, although the clinical significance of some of these interactions is yet to be determined (Table 2).¹⁷

Zinc is a common ingredient in CM used to support immune function and reduce the symptoms and duration of the common cold and flu.³⁷ Of clinical importance to general practitioners prescribing tetracycline and quinolone antibiotics is the potential for zinc to reduce the intestinal absorption of these antibiotics.^27,28

Pharmacodynamic interactions

Along with pharmacokinetic interactions, there is also the potential risk of serotonin syndrome when Hypercium perforatum (St John’s wort) is combined with another drug that has a serotonergic effect.³⁸ It is therefore recommended that this combination is avoided.³⁹

Interactions that increase the risk of bleeding from antithrombotic or non-steroidal anti-inflammatory drugs are predominantly due to the additive anticoagulant and/or platelet aggregation inhibitory effects of CMs and/or an increase in the plasma concentrations of the drugs. The commonly used herbs, Allium sativum (garlic), Zingiber officinalis (ginger) and Ginkgo biloba (ginkgo) all have antiplatelet and anticoagulant properties.⁴⁰ Consequently, despite the inconsistent clinical evidence of an increased risk (Table 1), coadministration of these herbs with antithrombotic drugs is generally not recommended.¹⁰ Glucosamine supplements should also be avoided in patients taking warfarin. A case series reported an increased INR associated with the concurrent use of glucosamine supplements and warfarin, in which 17 of 21 cases resolved after discontinuation of the glucosamine supplements.⁴¹ Coadministration of P. quinquefolius (American ginseng) and warfarin can lower INR levels. Changes in the dose, batch or brand of ginseng may destabilise a patient’s INR and increase the risk of bleeding or clotting.^9,10

Interactions can also occur between antihypertensive drugs and several CMs, most notably with Allium sativum (garlic). Garlic formulations containing higher amounts of the bioactive sulfur compound allicin have demonstrated additive hypotensive effects.⁴²

Interactions between glycaemic drugs and Allium sativum (garlic),⁴³ Ginkgo biloba (ginkgo),⁴⁴ P. quinquefolius (American ginseng),⁴⁵ Hydrastis canadensis (goldenseal)⁴⁶ and Gymnema sylvestre (gymnema)¹² could also occur. Clinical monitoring is recommended because of potential additive hypoglycaemic effects or unstable glycaemic control due to pharmacodynamic and/or pharmacokinetic interactions.^{12,13,43–45,47}

Interactions between CMs and oral COVID-19 antiviral drugs

Of the three oral COVID-19 antiviral drugs provisionally registered by the Therapeutic Goods Administration in Australia, only ritonavir plus nirmatrelvir has an interaction warning for a CM (St John’s wort). However, there are other CMs that could potentially interact with drugs that patients might be taking to prevent or treat COVID-19 infections (Table 3).

The gastrointestinal absorption of the three oral antiviral COVID-19 drugs may be reduced by zinc compounds or Althaea officinalis (Marshmallow root). Zinc sulfate has been shown to lower plasma concentrations of ritonavir through chelation in the gut.⁴⁸ It is possible that other zinc compounds could also have chelating properties. However, it should be noted that at the time of writing this review, the University of Liverpool’s COVID-19 drug interaction checker (www.covid19-druginteractions.org/checker) does not mention the study of Moyle et al⁴⁸ and only considers whether zinc is likely to affect the metabolism of ritonavir. Coadministration of marshmallow (A. officinalis) root may also reduce the absorption of any three oral antiviral COVID-19 drugs due to the mucilage content of marshmallow root. As a precaution, patients should avoid CMs with either ingredient, or take the antiviral drugs at least two hours before or six hours after ingestion of the CM.⁴⁸

CMs that induce or inhibit CYP3A4 and/or P-glycoprotein can potentially interact with remdesivir, ritonavir or nirmatrelvir. Pharmacokinetic interactions with molnupiravir are unlikely because it is not metabolised by CYP enzymes or renally excreted.⁵²

Taking Hypericum perforatum (St John’s wort) with ritonavir plus nirmatrelvir should be avoided because there is a theoretical risk that the CM may reduce the plasma concentrations of either drug because of its induction of CYP3A4 and P-glycoprotein, an effect that may take up to two weeks to wear off. Based on this theoretical risk, there is also the potential for a weak interaction between Hypericum perforatum (St John’s wort) and remdesivir. According to the University of Liverpool’s COVID-19 drug interaction checker, strong inducers of CYP3A4 are likely to have a clinically unimportant effect on remdesivir concentrations (~15–30% reduction) and, as such, dose adjustments are not recommended. However, other interaction checkers such as Natural Medicines (https://naturalmedicines.therapeuticresearch.com) take a more cautious approach and recommend against coadministration. Given the limited evidence for molnupiravir,⁵³ it may still be preferable to prescribe remdesivir despite the theoretical risk of reduced efficacy and, of course, stop Hypericum perforatum (St John’s wort), at least temporarily.

A controlled trial evaluated the effects of two weeks of Echinacea purpurea⁵¹ on ritonavir and another study evaluated the effects of four days of a Allium sativum (garlic) product with a low allicin content.⁵⁰ Neither of the CMs reduced ritonavir plasma concentrations.^35,36 This might be attributed to ritonavir’s potent inhibition of CYP3A4 and P-glycoprotein, which potentially counteracted any induction effects from the CMs.⁵¹

A pharmacokinetic study of a Allium sativum (garlic) product with a high allicin content revealed reduced plasma concentrations of saquinavir, another protease inhibitor,³⁵ and another case study reported that approximately six stir-fried garlic cloves three times per week reduced serum concentrations of atazanavir, as well as its clinical effectiveness.⁵⁴

Most patients using prescription drugs for COVID-19 are at a high risk of complications. Ensuring optimum drug concentrations is likely to be a priority for these patients, including those who are strong proponents of CM. Therefore, the potential interactions between CMs and drugs prescribed for COVID-19 listed in Table 3 should be discussed with all patients.

Reducing the risk of CM–drug interactions

There is a real opportunity for healthcare professionals to engage in informed conversations that promote the safe and appropriate use of CMs (Table 4).

A CM–drug interaction should be suspected when the clinical presentation correlates with the drug’s clinical effects or a known adverse effect. Polypharmacy (five or more medicines) also increases the risk of drug interactions.⁵ People living with chronic health conditions are more likely to use CMs alongside multiple over-the-counter and prescription drugs.⁴

Pharmacodynamic interactions are easier to predict because they correlate with the anticipated clinical effects of the compounds. Although the scientific evidence of clinical efficacy is limited for many CMs, these types of interactions should still be considered, even though some may be welcomed, such as improved glycaemic control (Table 3), which may still require the monitoring of blood glucose.

Predicting pharmacokinetic interactions is more difficult. Clinically important interactions with drugs are more likely when there is a narrow therapeutic index and with CMs that are known perpetrators. The risks are also higher with larger doses of constituents known to cause CM–drug interactions. However, the formulation or minimum dose required for an interaction to occur are not always clear from the available research.

Identifying CM–drug interactions is further complicated by the wide variations in CM quality (especially for raw herbs and products purchased outside of Australia) and their different constituent profiles, formulations and doses.⁶ CM products, especially herbs when used in a traditional context, often have multiple ingredients that may also interact with each other.⁵⁵ Therefore, a CM–drug interaction may be triggered by changing the CM product batch or brand. Unclear labelling and low health literacy can exacerbate these risks; for example, the American, Korean/Chinese and Siberian ginsengs vary in their pharmacological effects and interaction risks. Patients may not realise this and unwittingly switch between CM products containing ‘ginseng’, and therefore tell their healthcare practitioner that there have been no changes in their CM use.

Close monitoring of CM use or avoidance is warranted when the drug has a narrow therapeutic index or the consequences of an interaction are serious.⁵⁶ However, there will be other instances when the benefits of a CM warrant its continuation and other management options should be implemented, such as monitoring clinical outcomes or prescribing another drug with equivalent effectiveness that is unlikely to interact with the CM. Decisions about medicine use, dose adjustments and cessation must also consider patient preferences, including any observed or perceived benefits that are motivators for ongoing use.³⁷ Shared decision making may improve treatment compliance (Table 4).⁵⁷

Conclusion

Vigilance, an understanding of clinical pharmacology and the nuances of CMs, and effective communication skills are required to prevent and identify unwanted CM–drug interactions.

Key points

• Most Australians who use CMs are also using prescription and over-the-counter medicines.
• Pharmacokinetic CM–drug interactions can alter the absorption, distribution, metabolism and elimination of a drug.
• Pharmacodynamic CM–drug interactions can negate or add to a drug’s clinical effects.
• Clinically important CM–drug interactions are more likely when there is a narrow therapeutic index, polypharmacy or changes in CM brands and batches.
• The quality use of medicines includes asking patients about CM use and engaging in informed discussions about CM–drug interactions.