During the past few years, pharmaceutical and dietary supplement products made from kava (Piper methysticum Forst., Piperaceae) have been banned from sale in many countries based on the suspicion of severe hepatic adverse effects. In 1999/2000, 8 spontaneous case reports of liver adverse events reached the Swiss regulatory authorities within just 6 months, all with an acetonic kava extract,¹ followed by a number of rather poorly documented case reports on other kava preparations, mostly from Germany.² This sudden appearance of a potential problem was not backed by traditional experience of kava drinking in the South Pacific, nor with the broad clinical experience with kava products in Europe.

Despite a large number of pharmacological and toxicological studies, and even new clinical trials published since the ban,^3-10 no convincing proof of an inherent toxicity of kava exists.¹¹ Recent toxicological results apparently pointing to a potential toxicity were in fact not obtained with the typical “noble” kava cultivars regularly used for daily kava drinking in the South Pacific. The “noble” cultivar is used for extract production in Europe, but the toxic reactions are traceable to unusual kava qualities such as the cultivar “Isa” (for which the history in Hawaii can be traced back to a relatively recent introduction from Papua New Guinea planted in Hawaii for research purposes). Isa would have to be considered an unacceptable quality for daily drinking, at least according to results from recent in vitro testing.¹² Another recent study on in vitro effects of kava described some potential mechanisms of kava toxicity, but again, the kavalactone signature of the study material points to kava material unacceptable for daily kava drinking.¹³ In addition, some of the models of this study have been applied to the kava material used in European extract production, with no hints to toxic effects.^14,15

Further, in a recently published study to evaluate potential repeat-dose toxicity, no deviation in histological, biochemical, and hematological parameters was found in feeding rats with up to 73 mg/kg of body weight of ethanolic kava extract daily for 3 months.¹⁶ Until 1999, kava was considered an efficacious and exceptionally safe herb for the treatment of stress disorders and mild states of anxiety.¹⁷ Meta-analyses of randomized, double-blind, placebo-controlled clinical trials conducted on the leading European acetonic and ethanolic extract support the phytomedicine’s safe and effective use for anxiety.^17-20 The relatively sudden development of reports of alleged hepatotoxicity several years ago might point to a problem with raw material quality associated with some of the implicated kava preparations.

Quality issues with the raw material from the point of view of the South Pacific producers are presented in HerbalGram 71, “The Quality of Kava Consumed in the South Pacific” by V. Lebot.²¹ In the text below, differences in kava quality shall be discussed as seen from the point of view of Western extract production. Indeed, there are striking differences in the quality of the raw material used for extraction, which have not been taken into account when the case reports of alleged liver toxicity were discussed. The potential toxicological impact of these differences in kava raw materials used for extraction needs to be clarified in future research.

Cultivars Used in German-Produced Kava Preparations

According to German Kava extraction companies, the raw material had to correspond to the definitions of the kava monograph in the German Drug Codex (DAC 86), which does not allow the use of aerial (above-ground) parts (F. Gaedcke [Finzelberg] and K-H Sensch [Gehrlicher Pharmazeutische Extrakte], personal communication, November 2006). The rules of drug registration in the EU require detailed batch analyses of the raw material; thus, deviations from the raw material quality as defined in the registration dossiers would immediately be recognized. However, as there is no definition of the cultivar quality to be used, the choice of the cultivars was based on established trading relationships rather than a consciously-made and deliberate decision.

The situation may be different in countries where kava is sold as a dietary or food supplement without strict controls. However, the original case reports of hepatotoxicity arose from the use of well-controlled Swiss/German preparations manufactured under drug regulations and related Good Manufacturing Practices, not with products produced as food or dietary supplements. In view of the high degree of batch documentation, the argument of inferior quality (e.g., through the use of aerial parts as proposed by some South Pacific scientists) is thus clearly wrong when it comes to the extracts ingested by the patients suffering from liver disorders. However, even though the use of aerial parts can be excluded, the choice of an adequate cultivar is a completely different and unregulated topic.

An analytical screening of commercial samples of kava products by high-performance liquid chromatography (HPLC) in the year 2000 (see Table 1) reflected a rather constant quality, with obviously only few kava cultivars being used for extraction. A mixture of various cultivars, of course, cannot be excluded; however, the extraction companies producing ethanolic kava extracts for German registered drugs did not usually buy from the open market, but rather from identifiable sources (K-H Sensch, personal communication, July 2001). The method used corresponded to the analytical conditions described by Grazca and Ruff,²² and it was applied by Lebot and Levesque²³ for a phytochemical screening of kava cultivars (normal phase, dioxan/hexane as mobile phase). Lebot and Levesque introduced the system of kavalactone signatures, attributing to each lactone a number in the sequence of its elution from the HPLC column: desmethoxyyangonin = 1; dihydrokavain = 2; yangonin = 3; kavain = 4; dihydromethysticin = 5; methysticin = 6. A signature is formed when the figures are sorted in the sequence of decreasing quantities of individual lactones in the sample. With the relative composition of kavalactones being genetically defined, this method of assigning a kavalactone signature not only allows one to draw conclusions regarding the migration of related kava cultivars throughout the South Pacific islands, but also to correlate the phytochemical composition to the local experience with preferences or dislikes for certain kava cultivars.^23-26 Of course, the preference for certain cultivars must not necessarily correlate with pharmacological or clinical effects; however, the dislikes for certain cultivars are mostly connected to the observation of adverse effects from kava drinking such as headache, nausea, vomiting, and hangover,²⁶ and are thus directly relevant to a form of direct, observational pharmacovigilance.

The HPLC method was adapted to the use of kava extract in drug or supplement dosage forms, such as tablets or extracts. The use of an accepted literature method was justified by the fact that the official monograph in DAC 86 does not have an HPLC method, but a thin-layer chromatography (TLC) determination with colorimetric quantification, based on the method published by Csupor.²⁷ This TLC method does not allow the detection of all 6 major lactones, but only of kavain, methysticin, and yangonin. The normal phase HPLC method—the one also used for the analytical screening of commercial kava products—was standard up to the year 2000, when a new reversed-phase method became known.^28,29 However, the 2 HPLC methods yield very similar results (changing to an RP column allows for a simpler analytical routine setup), in contrast to the comparison between the DAC TLC method and the HPLC methods.²⁹ There is a factor of 1.6 between the HPLC and the TLC methods,²⁹ which also explains the differences encountered in the declaration of kavalactone contents in various preparations, and even in clinical trials. Some of these preparations declared kavalactone contents according to the DAC method, whereas others referred to the HPLC method. This difference must be taken into consideration when the risk-benefit ratio of various preparations is compared.

Despite the large number of brands with different galenical formulations (i.e., methods and types of preparations), the majority of the extracts used in German-registered kava products were produced by only 4 extraction companies, among them a producer of an acetonic extract. This is reflected in the similarity of kavalactone signatures of the various brands. As some degree of shift in the composition of the kavalactones is to be expected through the extraction process with organic solvents, the exact cultivar cannot be determined retrospectively. However, our experience shows that the general proportions of the single kavalactones do not vary considerably before and after extraction. Thus, the approximate origin and overall quality of the raw material can still be detected in the extract preparation.

With the possibility of the use of a mixture of cultivars (which was, however, excluded by at least one extraction company), the question arises whether the analysis of kava extracts can yield meaningful hints on the raw material used for production. However, an extract signature pointing to a cultivar favored for daily kava drinking would not be a cause for concern from the toxicological point of view even if cultivars were mixed. Conversely, a signature pointing to a bad kava quality (one for which adverse effects are known) is a potential problem for the assessment of drug safety.

With some caution, the available analytical data partially presented in Table 1 point to the use of kava cultivars preferred for daily drinking in Samoa and Vanuatu—with one exception. Obviously, two very different cultivars were used for the preparation of the acetonic extract in 1999/2000—the time when the first relatively well-documented kava liver case reports came up in Switzerland. The finding of signatures typical for a preferred Samoan cultivar in one batch, and a typical Tudey signature in another batch is not an analytical artifact. In fact, it correlates with the reports of the suppliers in Samoa and Vanuatu stating that they delivered the raw material (Eddie Wilson, personal communication, October 2000). With the sources of the cultivars used in the production of the acetonic kava extract known, one cultivar could be identified as the Samoan “noble” kava cultivar Ava Laau; the other was the “no-drink” type Palisi from Vanuatu. Material from both cultivars was obtained and submitted to phytochemical analyses, for which the results were presented on the International kava conference in Fiji.³⁰ From the toxicological point of view, the fact that there were 2 essentially different qualities of the acetonic extract on the market may be a mere coincidence, yet still a point to consider for the interpretation of the liver case reports.

The signatures 426531 and 462531 correspond to a preferred quality typically found in Samoa, 254631 and 246531 are cultivars typically used in Vanuatu, whereas the signature 526431 corresponds to a typical “Tudey” quality, a “no-drink” kava type from Vanuatu.

Inadequate Kava Cultivars

The use of a “no-drink” kava in the preparation of the acetonic kava extract is not circumstantial. The cultivar Palisi is a very proliferative and fast-growing cultivar with a high yield of kavalactones already after one year of cultivation. Palisi was avoided for traditional kava drinking prior to the selection as a raw material for cultivation. The reason for the avoidance of Palisi in traditional kava drinking is that Palisi is known for various adverse effects, such as hangover, scaly skin, or watery eyes. There is no hint of adverse effects on the liver; however, this is probably the case because the material was never used regularly enough to gain corresponding experience.

Unfortunately, the traditional experience is vanishing in the South Pacific, and with the cultivation of Palisi still existing, the local kava market is now flooded with material of doubtful quality. Already, hitherto unknown adverse effects such as scaly skin or watery eyes are becoming more common in Vanuatu. If kava is ever to return on the world market, the quality must be assured in a reproducible way. Clearly attributable serious adverse events from kava drinking in any state of the South Pacific would most likely be the ultimate end of kava.

Proposals for Kava Quality Control

With the new regulations for the registration of drugs or the production of food supplements in Europe, there is now an increasing awareness of quality matters and traceability in the supply of herbal raw material. Should kava come back to the European market as a registered drug or even under food status, the new rules asking for traceable raw material (EU directive 178/2002/EC) already apply. Appropriate quality specifications of kava must therefore be defined to prevent foreseeable future problems.

The big question is how to define kava quality. The question might seem more complicated than it is. The South Pacific states can look back on a wealth of traditional experience. The most logical approach would be to orient the definition at the local experience and support this experience with a scientifically-based screening. Much work was already done by the CIRAD (Centre de Cooperation Internationale en Recherche Agronomique pour le Développement; Center for International Cooperation in Agricultural Research for Development) in the past,^23-26 and the consequences from this work can immediately be used. More than 200 cultivars of kava were identified throughout the South Pacific islands, and the genetic relationships partly examined.

In conclusion, there are quality parameters that can already be defined and applied. The European pharmaceutical industry will set up standards for future kava quality. Kava raw material not meeting the standards will simply be rejected. These quality specifications will probably encompass the following:

• No use of stem peelings or chips from the aerial parts of kava stumps. A certain percentage of root chips from underground parts may be acceptable.

• Traceable origin of the roots as defined by the new regulations for drugs and food supplements in the EU. A method to ensure the traceability of the kava roots will have to be installed in cooperation with the farmers and traders, if possible analogous to the system of controlled origin as successfully done with wine, as proposed by V. Lebot.²¹

• Use of “noble” cultivars, preferentially with high relative kavain content and low contents of methysticin and desmethoxyyangonin. The control of this parameter is easily made by standard HPLC methods. Limits would have to be defined according to the identification of suitable cultivars. For practical reasons the extract producers should be able to select from a variety of suitable cultivars that might be compiled into a positive list, without being restricted to one single acceptable type. Natural variability in phytochemical composition must also be taken into account in the process of defining a suitable kava quality for use in herbal medicinal products.

• Use of healthy and fully matured plants of minimum 3 years of age. This parameter cannot be controlled by European producers and must be guaranteed by the farmers.

Quality control procedures are not a single-sided matter. They must start from the farmer’s field by choice of plant variety and an appropriate cropping system, and must be implemented throughout the marketing channel down to the final recipient. It is not only in the best interest of all parties involved to produce the best possible kava quality, but it is also an investment in the future of kava.

Mathias Schmidt, PhD, is a pharmacist with long-term experience in kava research. He specializes in medicinal plant research with issues related to herbal safety and quality: pharmacovigilance, clinical studies, toxicology, and traceability projects for herbal raw material. He is affiliated with HerbResearch Germany, a consultancy company in the field of herbal medicines and food supplements. E-mail: Schmidt@herbresearch.de.

Table 1. Analytical profiles of some German kava products commercially available in 1999/2000 (Source: Gehrlicher Pharmaceutical Extracts, data from an HPLC screening of commercial samples of German registered kava products).