Black Cohosh and Liver Injury
On February 9, 2006, the Australian Therapeutic Goods Administration (TGA) announced the following:

The Therapeutic Goods Administration (TGA) recently reviewed the safety of Black cohosh (Cimicifuga racemosa) following reports of possible liver problems internationally and in Australia. At the time of the review, there were 47 cases of liver reactions worldwide, including nine Australian cases. In Australia, four patients were hospitalized, including two who required liver transplantation. Although some reports are confounded by multiple ingredients, by more than one medication or by other medical conditions, there is sufficient evidence of a causal association between Black cohosh and serious hepatitis.

However, considering the widespread use of Black cohosh, the incidence of liver reaction appears to be very low. Following the safety review, the TGA has decided that medicines containing Black cohosh should include the following label statement:
Warning: Black cohosh may harm the liver in some individuals. Use under the supervision of a healthcare professional.

New products will need to comply with the requirement from the time of manufacture. For existing products, a phase-in period of twelve (12) months will be given to allow sponsors adequate time to comply with the new labelling requirements.

On July 18, 2006, the Medicines and Healthcare Products Regulatory Agency (MHRA) in the UK issued a press release stating that all black cohosh products sold there should carry the following label warning:

Black cohosh may rarely cause liver problems. If you become unwell (yellowing eyes/skin, nausea, vomiting, dark urine, abdominal pain, unusual tiredness), stop taking immediately and seek medical advice. Not suitable for patients with a previous history of liver disease.

Notwithstanding the Australian TGA's claim concerning the number of cases of liver damage linked to black cohosh (which include the adverse reaction reports filed with the UK MHRA and other health authorities), only five papers or letters have been published purporting to demonstrate a link between black cohosh (Actea racemosa) ingestion and subsequent liver injury. It is important to closely examine these published reports since, because of the process of peer review, these represent the best-documented evidence of any association with liver injury.

The first publication, from doctors at the Princess Alexandra Hospital in Brisbane, Australia, described six patients with evidence of severe hepatitis that was linked to taking a range of herbal products.¹ Two of these patients were taking black cohosh, although one was also taking other herbs including skullcap, a herb which can be substituted by Teucrium species, a known hepatotoxic genus.² The one case attributed to black cohosh alone (Case 1) was truly dramatic. Of the cases reported, the most serious illness occurred in this 47-year-old woman who was taking black cohosh for menopausal symptoms. She required liver transplantation even though, according to the publication, the patient had been taking the black cohosh for just one week. Histological examination of her liver confirmed severe hepatitis and early fibrosis. The patient did not exhibit eosinophilia and had no signs of any systemic disturbance. Serology for hepatitis A, B, and C was negative, but rechallenge with the herb was not performed "for ethical reasons." The dose of black cohosh taken was not specified, neither was the product.

The second publication, also from Australia, describes a 52-year-old woman with acute liver failure (Case 2).³ She had been taking an herbal formula containing 1:1 liquid extracts prescribed by a pharmacist. Black cohosh 1:1 was ten percent of the mixture, and the daily dose of the combination was 7.5 mL twice a day. The patient underwent successful liver transplantation. Although the authors stated that "Extensive investigation excluded other recognised causes of liver failure," they provided no details of what these investigations were. Analysis by the TGA was said to confirm the presence of goldenseal, black cohosh, and ginkgo in the herbal mixture. Other stated ingredients were ground ivy and oats seed.

The third, fourth, and fifth publications detail single case reports from the United States. One report describes the development of autoimmune hepatitis, which the authors claim was triggered by the use of black cohosh (Case 3).⁴ A 57-year-old diabetic woman presented with a two-week history of lethargy and fatigue. Her medications (all of which had been used for more than two years) included labetalol, fosinopril, verapamil, metformin, aspirin, and insulin. Three weeks before presentation, the patient began taking black cohosh tablets (unknown brand or dose) for hot flashes. Drug-induced autoimmune hepatitis, attributed to the black cohosh, was diagnosed. Tests for hepatitis A, B, and C were negative. The black cohosh was discontinued, and a tapering steroid course was instituted. Complete resolution of symptoms occurred within two weeks, and resolution of the abnormal liver function tests (LFTs) occurred within nine weeks. Follow-up liver chemistries remained normal two months after steroids were discontinued. However, at four months, the woman returned with a complaint of jaundice and fatigue. LFTs revealed AST 1260 U/L, ALT 1694 U/L, and bilirubin 9.2 mg/dL. The patient experienced rapid improvement on a second course of steroids, and long-term azathioprine was begun. The authors claimed that none of the patient's other medications have ever been implicated as triggers for autoimmune hepatitis, but provided no information as to whether these drugs were maintained or withdrawn from the patient. Rechallenge with black cohosh was not undertaken.

The next case report (Case 4) was that of a 50-year-old woman suffering from acute onset jaundice.⁵ The provisional diagnosis was autoimmune hepatitis, since tests for hepatitis A, B, and C, cytomegalovirus, and Epstein-Barr virus were all negative. In the five months prior to the onset of jaundice, the patient was taking black cohosh 500 mg daily for menopausal symptoms and was not on any other medications. The patient underwent liver transplantation after she failed to respond to initial treatment and the explanted liver showed features of acute hepatitis. No further details concerning the black cohosh usage by the patient (or its cessation) were provided, and a rechallenge was probably not undertaken.

The most recently published case report (Case 5) describes a 54-year-old woman who presented with an eight-week history of fatigue, forgetfulness, and a ten-pound unintentional weight loss.⁶ Her medications included levothyroxine 100 mcg daily and 1000 mg of black cohosh (product not specified). The patient had been taking the black cohosh product for eight months. She had been drinking about two glasses of wine nightly for several years. AST was 1014 U/L, and ALT came in at 1003 U/L. Bilirubin was mildly elevated at 2.4 mg/dL. Tests for standard infectious causes of liver disease were all negative, as were those for other likely causes. The patient deteriorated while in hospital and died during a liver transplant operation.

The authors attributed the cause of the liver damage to black cohosh, but also acknowledged that it was possible that the patient had pre-existing cryptogenic cirrhosis or marker-negative autoimmune liver disease. The biopsy was suggestive of autoimmune hepatitis or drug-induced chronic hepatitis with autoimmune features.

This case (Case 5) was first presented as a conferenc e report in 2004.⁷ Unfortunately, there are significant inconsistencies between the published and oral versions of this case report. The conference report lists the patient's concurrent medications as fluoxetine, propoxyphene, and acetaminophen, as well as the black cohosh and levothyroxine mentioned in the published report. Also, the time of black cohosh use is stated as about three months in the conference presentation, and the patient did apparently test positive for hepatitis B surface antibody and herpes simplex virus IgM. The patient had travelled to Mexico seven months earlier.

Background to Drug-Induced Idiosyncratic Liver Injury
The phenomenon of idiosyncratic hepatic reactions to drugs is well-documented. It also appears that this reaction does occur to certain herbs, e.g., chaparral (Larrea tridentata) and germander (Teucrium species). By definition, such reactions are rare and unpredictable and are not dose-related.⁸ There are two types of idiosyncratic hepatic injury: hypersensitivity and aberrant metabolism. The former develops one to five weeks after exposure to the drug and, since it is immune-mediated and acute, also involves a systemic reaction including rash, fever, and eosinophilia. The latter takes weeks to months to develop, and symptoms are confined to the liver.⁸ Diagnosis of drug-induced idiosyncratic liver injury (DILI) is very difficult and relies largely on circumstantial evidence. Factors taken into account include a temporal association, exclusion of other possible causes, a consistent latency period to those described above, presence or absence of hypersensitivity (systemic) features, positive response to drug removal (dechallenge), positive response to rechallenge, and a positive lymphocyte stimulation test (this last factor is quite controversial). Complicating this is the fact that DILI can mimic every known human liver disease.⁸

There are many confounding factors that could lead to incorrect associations between ingested medications or herbs and idiosyncratic liver injury. Many viruses that cause liver disease are still to be identified,⁹ and there are no tests for them. Tests are not always done for even known viruses. For example, a Dutch study published this year found that Hep E virus was a significant cause of unexplained hepatitis.¹⁰ Occult celiac disease has been suggested as a cause of unexplained raised alanine aminotransferase (ALT)and aspartate aminotrasferase (AST) levels.¹¹ Rare liver diseases may not be excluded.^12,13 Hair dye was a cause of DILI in a Japanese man.¹⁴ Other environmental factors could be implicated.

Confusion with Idiopathic Hepatitis?
The experience of a liver transplant unit, which has been recently described, highlights some key issues behind the history and incidence of severe acute hepatitis – fulminant hepatic failure (FHF). All adult cases of FHF presenting to the Victorian Liver Transplant Unit (Australia) from 1988 to 2002 were analyzed. Eighty patients (mostly female) were referred, at a rate of approximately one case per million population per year. Mean age was about 38 years. Most cases were due to acetaminophen poisoning (36%) or idiopathic hepatitis (34%).¹⁵ Only five of the 80 cases were classified as drug-induced, making this causality a rare factor. Other main causes included hepatitis A (three cases), hepatitis B (eight cases), and Wilson's disease (six cases). The 27 cases (34%) of hepatitis due to unknown causes (idiopathic) is a staggering rate. These cases are also described as non-A non-B hepatitis, since patients are not positive for hepatitis A or B. In the USA, one study found that the most common cause of FHF was non-A non-B (idiopathic) hepatitis.¹⁶ (Note that this US study was published in 1995, well before the dramatic rise in herbal use in that country.) Presumably unidentified infections or environmental factors could cause these cases of idiopathic hepatitis. However, the authors of the Australian study state the following:

The strong female predominance of cases argues against a viral cause and raises the possibility that hormonal factors are involved, or that the condition is linked to autoimmune liver diseases. There is clearly a need for large, detailed, multicenter epidemiological studies to provide further clues to a possible etiology/ies of this syndrome.

The demographics of idiopathic hepatitis (female, late 30s to early 50s) and black cohosh use strongly overlap. Hence, there is a distinct possibility that some patients who develop idiopathic hepatitis might also be coincidentally taking black cohosh. The herb could then be mistakenly attributed as the cause.

Analysis of the Published Case Reports
The published case reports linking black cohosh to liver injury have some serious flaws. In particular, for all cases, the presence of black cohosh in the products being consumed was not definitely established. Moreover, in most cases, the name of the product and the dosage taken were not specified. Certainly, for Case 2, there is the assertion that the TGA identified black cohosh in the herbal mixture, but no details of the results or how this was done are provided. The fact that two herbs in the mixture could not be identified makes any argument for the involvement of black cohosh in Case 2 fundamentally flawed.

The issue of the botanical authenticity of black cohosh products in the US has been highlighted by a recent publication.¹⁷ Of 11 black cohosh products tested, three were found to be from the wrong species (Asian species of Actea), and one was a mixture of both black cohosh and an Asian Actea species. For the seven products containing only authentic black cohosh, there was significant product-to-product variability in phytochemical constituents.

The features of Case 1 are baffling. Since the problem reportedly developed after one week of taking black cohosh, this would have to be a hypersensitivity reaction if the herb was truly the cause. Yet the patient lacked any features of a hypersensitivity reaction (rash, fever, systemic reaction, eosinophilia). In fact, eosinophilia was specifically stated to be absent. Furthermore, the explant liver showed signs of early fibrosis, a phenomenon that could only develop after months of exposure to the causative agent. Clearly, on the evidence provided, one week of black cohosh is the least likely cause of the patient's liver damage.

For Case 3, the authors claimed that none of the drugs the patient was taking have been linked to autoimmune hepatitis. Yet a simple search revealed several cases for labetalol, where this drug may have indeed caused an idiosyncratic autoimmune hepatitis, including one overview report of 11 cases from the US FDA.^18-20 In Case 4, the authors justify their identification of black cohosh as the cause of the patient's liver injury on the basis of the two Australian publications. In their discussion, they attribute to black cohosh the presence of hepatotoxic alkaloids and salicylates. Such attributions are nonsensical, are unsupported by the literature,²¹ and above all cast doubt on the credibility and diligence of their overall analysis of the case.

Additional problems with the five cases include the lack of positive identification of black cohosh as the true cause by either a rechallenge or a lymphocyte stimulation test. While the latter can give false negatives, if positive, it would have provided more conclusive evidence of any link to black cohosh use. But the likelihood is that most of the authors involved never actually saw what the patients were taking (as evidenced by the appalling lack of product and dosage information) and therefore would have been unable to undertake such tests.

By Case 5, the literature evidence of an association between black cohosh and liver damage was considered to be so strong by the authors that they confidently identified black cohosh as the causative agent. This was despite their admission that other causes were possible. Furthermore, the inconsistencies already noted in the reporting of this case (conference report versus published article) create a great deal of uncertainty over the link to black cohosh. In particular, the cocktail of medications and alcohol that the patient was taking (according to the conference report) could well explain the hepatotoxic effect.²²

Analysis of All Current Case Reports
As mentioned earlier, in addition to the published reports, there are around 40 to 50 case reports of liver reactions to black cohosh that have been recorded by various government health authorities around the world. The European Medicines Agency Committee on Herbal Medicinal Products (HMPC) recently analyzed these cases, together with the published cases. Based on their evaluations, the HMPC concluded the following:²²

The HMPC evaluated 42 case reports of hepatotoxicity, collected from European National Competent Authorities (34 cases) as well as literature case reports (8 cases). Of these, only 16 cases were considered sufficiently documented to allow the Committee to assess if use of Cimicifugae racemosae rhizoma (Black Cohosh, root) could be linked to the liver injuries. As a result of the assessment, 5 cases were excluded, and 7 cases were considered unlikely to be related. In the remaining 4 cases (2 autoimmune hepatitis, 1 hepatocellular liver injury and 1 fulminant hepatic failure), there was a temporal association.

Of these four cases, only two were rated as "probable" using the Roussel UCLAF Causality Assessment Method (RUCAM). Not surprisingly, these were two of the published cases (already described in this column article as Cases 3 and 4). As previously outlined, there are several flaws in these two case reports, not the least of which was the failure to identify black cohosh in the products used.

Conclusions
The demographics of patients with non-A non-B (idiopathic) hepatitis closely match those of the black cohosh user. Hence, the most likely and rational explanation of some of the cases described is that they are idiopathic hepatitis, mistakenly attributed to black cohosh because of the common use of this herb. Once one mistaken case is described in the literature, however poor its quality, it is likely that others will follow in a process akin to a self-fulfilling prophecy. Separate or confounding issues are the common adulteration of black cohosh products with Asian species of Actea and the coadministration of many drugs known to cause liver damage. Hopefully, the regulators will not be motivated to act on such poor-quality case reports. The association of black cohosh and DILI remains unproven on the current evidence.

Notes
1. Whiting PW, Clouston A, Kerlin P. MJA. 2002; 177: 432-435.
2. Mills S, Bone K. The Essential Guide to Herbal Safety. New York: Churchill Livingstone, USA, 2005;581-584.
3. Lontos S, Jones RM, Angus PW, et al. MJA. 2003; 179: 390-391.
4. Cohen SM, O'Connor AM, Hart J, et al. Menopause. 2004; 11(5): 575-577.
5. Levitsky J, Alli TA, Wisecarver J, et al. Dig Dis Sci. 2005; 50(3): 538-539.
6. Lynch CR, Folkers ME, Hutson WR. Liver Transpl. 2006; 12: 989-992.
7. Cohen SM. In Workshop on the Safety of Black Cohosh in Clinical Studies. National Institutes of Health, Bethesda, Maryland, November 22, 2004. Available at: http://nccam.nih.gov/news/pastmeetings/blackcohosh_mtngsumm.pdf
8. Lewis JH. Best practice of medicine. 2000.Available at: http://merck.micromedex.com/index.asp?page=bpm_brief&article_id+CPM02hP373.
(January 3, 2007: Complete link does not work.)
9. Bowden DS, Moaven LD, Locarnini SA. MJA. 1996; 164: 87-89.
10. Waar K, Herremans MM, Vennema H, et al. J Clin Virol. 2005; 33(2): 145-149.
11. Bardella MT, Vecchi M, Conte D, et al. Hepatology. 1999; 29(3): 654-657.
12. Tordjmann T, Grimbert S, Genestie C, et al. Gastroenterol Clin Biol. 1998; 22(3): 305-310.
13. Gaya DR, Thorburn D, Oien KA, et al. J Clin Pathol. 2003; 56(11): 850-853.
14. Tokumoto Y, Horiike N, Onji M, et al. Intern Med. 2003; 42(11): 1104-1106.
15. Gow PJ, Jones RM, Dobson JL, et al. J Gastroenterol Hepatol. 2004; 19(2): 154-159.
16. Hoofnagle JH, Carithers RL Jr, Shapiro C, et al. Hepatology. 1995; 21: 240-252.
17. Jiang B, Kronenberg F, Nuntanakorn P, et al. J Agric Food Chem. 2006; 54: 3242-3253.
18. Marinella MA. J Clin Hypertens. 2002; 4(2): 120-121.
19. Stronkhorst A, Bosma A, van Leeuwen DJ. Neth J Med. 1992; 40(3-4): 200-202.
20. Clark JA, Zimmerman HJ, Tanner LA. Ann Intern Med. 1990; 113(3): 210-213.
21. Mills S, Bone K. The Essential Guide to Herbal Safety. New York: Churchill Livingstone, USA, 2005; 269-272.
22. The European Medicines Agency Committee on Herbal Medicinal Prod. (reports). Available at: http://www.emea.eu.int/pdfs/human/hmpc/26925906en.pdf and http://www.emea.eu.int/pdfs/human/hmpc/26925806en.pdf.