INTRODUCTION
	1. Total Listed Cases
	2. Duplicate Entries
	3. Side Effects With No Relationship to the Ingestion of Kava
	4. Causal relationship with concomitant medication probable
	5. Doubtful Causality
	6. Causal relationship cannot be evaluated due to information status
	7. Causal relationship probable at monograph-conforming dosage
	8. Probable cause relationship related to overdosage of kava extract
	9. Literature

Analysis of Kava Side Effects Reports Concerning the Liver

Translation to English by Lindenmaier M and Brinckmann J
31. December 2001

Analysis of hepatotoxic reactions listed by the BfArM (German Federal Institute for Drugs and Medical Devices)

The BfArM has recently informed industry and media about reported side effects that may be associated with the ingestion of kava preparations (2;3). In this preliminary information, it is stated that the Institute proposes to revoke marketing authorization for kava containing drugs including homeopathic preparations with a final concentration of D6. Henceforth, the manufacturers have been given the opportunity, within 4 weeks time of receipt of the letter, to respond with its position concerning the proposed measures. Even though the announcements are rather serious, the need for urgency comes as a surprise: with regard to the drug safety protocol developed in Switzerland in 2000, of which the German health authorities also had knowledge of, the information status has not fundamentally changed. Yet on October 19, 2000, the German health authorities stated in a 10 part report that the BfArM had no intentions of conducting a new risk evaluation for kava products in Germany (1). The announcement of a drug safety protocol through the BfArM was based on a listing of 24 cases of side effects in connection with serious hepatotoxic effects ranging up to liver failure, (cholestatic) hepatitis or cirrhosis of the liver. In 18 of these cases, the BfArM classified an association with kava as probable or possible. In one case, the adverse effect on the liver was fatal. Five cases were without any concomitant medication. Two reports could not be evaluated due a lack of clinical data. Also, in the cases involving co-medication, the BfArM considered kava to be responsible for the side effect. Serological investigations, as far as they were carried out, were negative in all cases. Closer inspection of the presented cases provides, however, another outcome and raises considerable questions with regard to the BfArM's accuracy or carefulness in association with sensible procedures. For example, the report regarding virus serology is misleading: such investigations were conducted in only the fewest of cases, and these were primarily the ones reported in Switzerland. The adverse event case reports from Switzerland are collectively characterized as being representative, while the evaluation of the listings by the BfArM is far from compliance with the current standards required to fulfill relevant European guidelines. When one examines the reactions in detail, it appears that the BfArM's classification of causality linked to kava, is, to a large extent, incomprehensible, and arbitrary. Moreover, in its evaluation of cases, the BfArM had not taken into consideration various existing pieces of information, for example those with regard to other possible causes. One extreme example may be concerning the aforementioned lethal case: in this instance, it was known to the Institute that the cause of liver failure was several years of alcohol abuse, and that kava was not involved in the genesis of the liver symptoms. The autopsy had shown that the cirrhotic process had already started long before the adminstration of kava began! The second, internal listing documented that this circumstance was known to the BfArM , but this listing has not been made accessible to the manufacturers for use in rebuttal statements regarding the notice of possible marketing revocation. This second listing contained a compilation of all known suspected cases (32 in all), including those reported in Switzerland and those published in the literature. This listing is indeed carefully conveyed as being an „official“ paper, however it still contains a range of obvious errors. Because the „non-official“ second listing of the BfArM is complete in regard to the sources referred to in this evaluation, the 24 reports of side effects initially reported by BfArM are consequently not being used as the basis of discussion, but rather the 32 cases that were entered into the second listing. In addition to the BfArM's two listings, other sources of information for the present case evaluation include the Interkantonalen Kontrollstelle (IKS) der Schweiz (Swiss Intercantonal Agency for Control of Medicines), the pharmacovigilance databank of the WHO, as well as concerned product manufacturers. The listing of cases that are suspected to be kava -related by the Arzneimittelkomission der deutschen Ärzteschaft (AKDÄ) (Drug Commission of the German Physician's Association) is not included. The AKDÄ's listing of the most recent adverse event reports contains no indications of liver toxicity from kava products. Moreover, the AKDÄ does not release the product names publicly, which prevents a meaningful use of the data a priori . Both lists of the BfArM contain repeated reports. Moreover, if one removes the cases in which a relationship with kava is either improbable or doubtful, and one also takes into consideration the cases that are not possible to evaluate due to insufficient documentation, the following picture results: Total cases 32 Double listings 4 No relationship with kava 3 Causal relationship with concomitant medication 11 Causal relationship with kava doubtful, however not definitively ruled out 4 Causal relationship with kava is not able to be evaluated 6 Causal relationship with kava probable with overdose and/or misuse of kava 3 Causal relationship with kava probable with dosage conforming to that recommended in the monograph 1 Consequently, there remains, in total, one single case of liver side effects with a confirmed relationship to kava . In three more cases, a relationship is probable, however in these cases kava was not taken within the recommended dosage of maximum 120 mg kava lactones daily. One case can even be identified as drug abuse. Only this last relevant case became known after finalization of the Swiss drug safety protocol in the autumn of 2000. The other relevant observations were already the subject of the same use / risk evaluation, which led to the report of the Federal Institute on October 19, 2000, wherein it was decided that there was no necessity for a new evaluation of kava preparations. In the final analysis, the safety protocol is based on a single, recently discovered case involving the use of an undetermined quantity of one kava product. A conservative estimate of about 250 million daily doses of ethanolic kava extract products have been ingested over the last 10 years. During the same time period there were only two known cases, wherein a causal relationship with the ingestion of an ethanolic extract of kava appeared probable. Therefore, the incidence ratio is calculated to be 0.008 cases per 1,000,000 daily doses. By comparison, the incidence rate for benzodiazepine is 0.90 (Bromazepam), 1.23 (Oxazepam) or 2.12 (Diazepam) cases of hepatic side effects per million daily doses. Therefore, a restriction of marketing kava products will mean that the patients involved will have to revert to an alternative medication with a 112–265 fold increase in hepatic risk. On the basis of the portrayed connections, there has neither been a change in the state of drug safety data compared with the preceding year, nor will there be a minimization of risk for the population by restricting the trade of kava preparations. Bielefeld, November 2001 redinomedica AG Dr. Mathias Schmidt

www.mauikava.com
cliff@mauikava.com
©2009