Reviewed: Teschke R. Kava hepatotoxicity. A clinical review. Ann Hepatol. 2010;9(3):251-265.

Hepatotoxicity (liver injury) caused by ingestion of herbal dietary supplements (or, as they are often regulated in many countries, herbal drugs) is considered comparatively rare. When hepatotoxicity does occur in someone using an herbal product, it is often difficult to determine if the herbal product contributed to or caused the injury. Determining causality can be further complicated by a lack of information about the processing and content of the product, poor ability to remember the duration and doses of the products consumed, lack of information about underlying diseases or other potentially confounding medical conditions, and the concomitant use of other herbs, conventional pharmaceutical drugs, or alcohol.

Many studies, reviews, and meta-analyses have demonstrated the efficacy of kava (Piper methysticum, Piperaceae) as a monotherapy for treating anxiety. Nonetheless, hepatotoxicity concerns have led to its withdrawal or restriction in numerous countries (including at least 3 nations in Europe: Germany, Switzerland, and the United Kingdom), with other countries following suit.* Kava hepatotoxicity is rare but it has been reportedly associated with ingestion of aqueous, ethanolic, and acetonic kava extracts.

The precise mechanism of kava hepatotoxicity is unknown, and many studies have evaluated different theories. These include the following: metabolic interactions with exogenous compounds at the hepatic microsomal cytochrome P450 level, alcohol abuse, genetic enzyme deficiencies, toxic constituents and metabolites derived from the kava extract including impurities and adulterations, cyclooxygenase inhibition, P-glycoprotein alterations, hepatic glutathione depletion, solvents and solubilizers of the extracts, and kava raw material of poor and/or non-officially approved quality (i.e., inclusion of non-rhizomatous [root] material, e.g., stems). The majority of patients with suspected kava hepatotoxicity co-medicated (with up to 5 different drugs). However, while conceivable, there is little clinical evidence for drug interactions with the main active compounds (kavalactones, aka kavapyrones) to be a contributing factor in hepatotoxic effects. Theoretically, if the metabolism of co-medicated drugs is altered by kava’s interaction with liver cytochrome P450 enzymes, then even relatively benign compounds may exert hepatotoxic effects.

Beyond efforts to establish toxicological mechanisms, solvents used for extraction and kava quality have become key topics of investigation. Since toxicity has been demonstrated with kava formulations made with each of the 3 solvents (ethanol, water, acetone), the nature of the solvent does not appear to be the issue. Instead, the quality of the kava raw material probably plays an important role in the observed hepatotoxicity. There are 6 major kavalactones that comprise 96% of the lipophilic resin compounds of the plant: kavain, dihydrokavain, methysticin, dihydromethysticin, yangonin, and desmethoxyyangonin. They are used to define kava chemotypes and serve as markers of quality.

Good quality kava can be expected for kava cultivated varieties (cultivars) consumed traditionally for centuries by the people of the South Pacific Islands. These cultivars are named “noble” cultivars because they have a solid safety record, are commonly used in recreational quantities delivering up to 2500 mg kavalactones per day, and may be consumed on a daily, long-term basis. Aside from the cultivar, the quality of kava is based on the plant part used. The aerial, stump, stem, adventitious roots, bark, and rhizome all have different levels of kavalactones. The peeled rhizome is used traditionally. It has higher levels of kavalactones than other kava parts and contains no pipermethystine, an alkaloid which has come under increasing recent scrutiny as the putative source of the potential kavaassociated hepatotoxicity.

The purpose of this review was to evaluate the clinical data of patients with suspected kava hepatotoxicity and to recommend ways to minimize the risk of hepatotoxicity associated with kava-containing products.

The author, Rolf Teschke, MD—a hepatologist at the Department of Internal Medicine at the Teaching Hospital of the Johann Wolfgang Goethe University of Frankfurt/Main, and author and co-author of several previous and recent reviews and evaluations of kavaassociated hepatotoxicity^1-5—states that all patients with suspected hepatotoxicity from drugs and dietary supplements (including herbs) should be assessed using specific diagnostic criteria. These criteria include elevated concentrations of certain liver enzymes and differentiation of the form of hepatotoxicity (hepatocellular, cholestatic, or mixed type). Causality should be assessed using rigorous scientific methods, as opposed to the ad hoc assessments made by regulatory agencies (including, but not limited to, Germany and Switzerland) following the initial reports of kava hepatotoxicity. The World Health Organization and Naranjo assessment scales, used by regulatory agencies for previous kava assessments, are nonspecific and should not be applied in cases of suspected hepatotoxicity from herbs and other dietary supplements, according to the author. The Council for International Organization of Medical Sciences (CIOMS) scale is a validated, structured, specific method for assessing hepatotoxicity, and the author notes that it is better suited for use by regulatory agencies.

This clinical review describes 31 cases of kava-associated hepatotoxicity for which there is sufficient information to assess causality. Causality for all 31 cases was assessed using the CIOMS scale. The 31 reports came from 26 regulatory cases originating in Germany and Switzerland and 5 case reports originating in Australia, New Caledonia, the United States, and Germany. In 14 of these cases, the causality was assessed as either highly probable, probable, or possible for kava, with or without involvement of other medicines. The following points summarize the findings from these 14 cases:

Patients ranged in age from 14 to 60 years, and 12 of the 14 patients were female.

Aqueous kava extracts were involved in 3 cases, ethanolic extracts in 5 cases, acetone extracts in 4 cases, and mixtures of kava with other herbs in 2 cases.

Daily intake of kavalactones (ranging from 45 to 2571 mg/day) exceeded manufacturers’ and/or prescribers recommendations in 5 patients, the duration of use (ranging from 1 week to 12 months) exceeded recommendations in 2 patients, and daily dosage and duration of use exceeded recommendations in 3 cases.

Kava is contraindicated in depression, yet half of the patients using ethanolic- or acetone-based kava extracts were taking kava for depression.

Hepatotoxicity was reversed in 10 patients following discontinuation of kava. Four patients required a liver transplant, and 1 of those patients died after liver transplant.

A positive re-exposure test was reported in 1 patient; i.e., symptoms of hepatotoxicity re-occurred upon resumption of use of kava.

Causality for kava with or without co-medications (conventional pharmaceutical drugs, herbs, non-herbal dietary supplements) was assessed as highly probable in 1 patient, probable in 4 patients, and possible in 9 patients.

Additional case series, case reports, and spontaneous reports to regulatory agencies are available, but the quality of these reports is low because of lack of detailed clinical data, failure to report testing for hepatitis and other viral infections, failure to consider existing liver disease or co-medications, and other methodological problems.

The author concludes that kava hepatotoxicity was confirmed in a small number of patients, even though causality was assessed as only “possible” in the majority of the 14 cases reviewed. The lack of epidemiologic data makes it difficult to calculate the incidence of kava hepatotoxicity, but the incidence appears to be low. The author explains that risk factors for hepatotoxicity include combining kava with other medicines, exceeding the recommended dose of kava, exceeding the recommended duration of kava use, and using kava for unapproved conditions such as depression. Despite previous speculation, the type of solvent extraction could not be confirmed as a risk factor for hepatotoxicity, but poor quality of raw material may play a role.

The author recommends that certain measures be taken if kava is to return to the market in Europe. Providers and patients must adhere to the dose and duration recommendations, adhere to the recommended indications for use, and avoid the use of drugs, dietary supplements, and other herbs while using kava. In addition, quality standards must be implemented to reduce the risk of hepatotoxicity.