Focus on Alternative and Complementary Therapies
www.pharmpress.com/fact
Focus Alternat Complement Ther©2005 Pharmaceutical Press
Focus Altern Complement Ther 2002; 7: 323–4
More and more cases of liver damage have recently been associated with the use of kava, a herbal anxiolytic medicine. Consequently, several regulatory authorities have taken action. At the time of writing (end of July 2002), these consisted of the following:
In July 2002, the UK ‘Committee on the Safety of Medicines’ (CSM) of the Medicines Control Agency (MCA) advised that ‘the possible therapeutic benefits of medicinal products containing the herbal ingredient kava kava cannot be considered to outweigh the safety risks and that consideration should be given to appropriate regulatory action’.1 This means that, after a consultation exercise, kava will probably be withdrawn from the UK market.
A summary was appended of 68 documented cases of suspected kava hepatotoxicity worldwide. In many of these instances, the exact nature of the extract was not specified. It is clear, however, that all types of extract and synthetic kavain are implicated. In the vast majority of these cases, other drugs (some with known hepatotoxicity) were taken concomitantly, a fact that considerably complicates causal attribution. Similarly, no data for alcohol consumption were provided. The problems typically occurred 2–3 months after kava intake; in some cases the length of kava use was not known. The adverse reactions ranged from mere transient elevations of liver enzymes to severe (often cholestatic) hepatitis and fulminant liver failure. Several cases were bizarre (e.g. ‘gall bladder pain’ or ‘cirrhosis’). In most instances, the patients seemed to have recovered fully after discontinuation of kava. However, six patients required liver transplants and three patients died.
Causality was classified according to standard criteria into four categories: probable, possible, unassessable or unlikely. Fourteen cases were deemed as probably related to kava use; these included three of the severe cases mentioned above. In 30 cases, causality was judged possible; this category includes all three cases from the UK. Five cases were unassessable, and in five cases, causality was thought to be unlikely.
Collectively, these data seem impressive. However, proper incidence or prevalence figures are not currently available. In other words, we do not know the frequency of hepatotoxic effects because the denominator is unknown. In our own systematic review of the data, two drug-monitoring studies of kava were located.2 These included 7078 patients taking 150 mg or 120 mg extract per day, respectively. No cases of hepatotoxicity emerged. The question remains whether the frequency of liver damage in kava users differs significantly from non-kava users.
What about the ‘therapeutic benefit’ mentioned in the CSM statement? The most conclusive evidence on efficacy probably originates from our meta-analysis,3 which the MCA apparently used for their efficacy assessment. It included seven placebo-controlled, double-blind randomised controlled trials (RCTs) of good to excellent methodological quality, with a total of 377 patients using 300–800 mg kava extract (equivalent to 60–240 mg of kavapyrones) per day. All RCTs yielded positive results (i.e. demonstrated efficacy). The weighted mean difference amounted to 9.7 [95% confidence interval (CI) 3.5–15.8] points on the Hamilton Anxiety Scale. Our conclusion was thus positive: ‘compared to placebo, kava extract is an efficacious treatment option for anxiety’.3 A recent trial not included in our meta-analysis demonstrated that kava extract reduced symptoms of patients previously treated with benzodiazepines.4
The benefit of kava has obviously been interpreted in more than one way. It seems that the CSM does not rate it highly and thus their risk–benefit analysis is negative.1 I find a negative risk–benefit balance difficult to accord with the best available evidence regarding efficacy.3 I would therefore desire that the CSM and similar institutions worldwide fully disclose their reasoning for arriving at a negative risk–benefit assessment. Furthermore, the kava case can only be argued in relation to the risks and benefits of other anxiolytic drugs. In other words, one should ask what the risk–benefit profile of other anxiolytic drugs looks like and how it compares with that of kava.5 Is it worse, similar or better than that of kava? Regulatory decisions regarding kava must be rational and transparent – if not, the herbal sector will (perhaps rightly) suspect foul play, an uneven playing field or simple old-fashioned bias.
