Over the past several years there have been numerous reports of possible liver toxicity associated with the use of various black cohosh (Actaea racemosa, Ranunculaceae, syn. Cimicifuga racemosa) preparations, popular for treating symptoms associated with menopause. Although some regulatory agencies and related bodies have reviewed these cases and have announced some preliminary cautions (for example, the European Medicines Agency [EMEA]), critical analyses have questioned the causality of such cases. In this paper, researchers from the Teaching Hospital of the Johann Wolfgang Goethe University of Frankfurt/Main, Hanau, Germany, analyzed case reports that regulators have previously considered as demonstrating “possible” or “probable” causality from black cohosh.

The EMEA has received 42 case reports of hepatotoxicity in patients taking black cohosh root for menopausal symptoms. In 26 of the cases, there was insufficient documentation to support a causal relationship between black cohosh use and hepatotoxicity. In 5 cases the association was deemed unrelated, and in 7 cases the association was determined to be unlikely. However, the EMEA has proposed a causal relationship for 4 cases. The purpose of this article was to analyze those 4 cases.

Details regarding the clinical course, co-medicated drugs, and diagnostic work-up were obtained from published case-study and EMEA reports. Causality was assessed with a structured algorithm based on the method of the Council for International Organizations of Medical Sciences. The algorithm has 3 steps: a pre-test, main-test, and post-test. The pre-test assesses each herb or drug consumed by the patient and qualitatively assesses if causality is unrelated or not assessable. In the main-test, each herb or drug is quantitatively evaluated for causality of histological changes. The post-test is a qualitative assessment (yes/no answers) of other diseases that are rare and may not have been considered in the previous steps.

Three of the 4 cases did not report the brand name of the black cohosh product, the ingredients, the manufacturer, the plant part used, or the solvent used in the extraction process. Information on duration of use is missing in one case and conflicting in another.

In Case 1, the patient was taking 4 concomitant medications that are known to be potentially hepatotoxic. After one week of taking black cohosh (dose and brand not known), liver-associated symptoms were noticed. Black cohosh treatment was discontinued, and a tapering steroid course was initiated, since drug-induced autoimmune hepatitis was suspected. Resolution of the abnormal liver function tests occurred within 9 weeks, and follow-up liver chemistries remained normal 2 months after steroid discontinuation. At 4 months, the liver disease recurred with rapid improvement by a second course of steroids and long-term azathioprine. A discussion of this case revealed that there is no certainty that the product taken by this patient included only black cohosh or any black cohosh. The patient recalled seeing the words “black cohosh” on the bottle but could not remember any details about the label or the bottle, which she had discarded. Another salient point is that drugs that cause autoimmune hepatitis are usually taken for a long time; in this case, black cohosh was taken only briefly. Since the concomitantly taken conventional drugs were not discontinued (including labetolol, fosinopril, verapamil, metformin, aminosalicylic acid, and aspirin), it is possible that they may have caused or triggered the autoimmune hepatitis in this patient with prior autoimmune polymyositis (inflammation of muscle tissue). Based on these findings and the changes in alanine transaminase (ALT) liver enzymes, the authors conclude that the liver disease was unrelated to black cohosh. The EMEA classified the case as having probable causality; however, the EMEA evaluated only one of the concomitant pharmaceutical drugs.

In Case 2, the patient took 500 mg/day of black cohosh root for 5 months prior to presentation with clinical signs of liver disease. The patient regularly consumed alcohol and was taking numerous conventional drugs that are known as potential hepatotoxins (valaciclovir and ibuprofen). The patient was also being treated for herpes virus, so she could have had herpetic liver disease. She recovered after liver transplantation. The EMEA classified the case as having probable causality with black cohosh; however, the EMEA did not evaluate the concomitant drugs because in the published case report the patient had denied taking other drugs or alcohol initially. This information was documented at a later date by other physicians/researchers. The authors conclude that causality is unrelated to black cohosh since a non-drug cause (herpetic hepatitis) is highly probable and qualitative and quantitative causality could not be assessed because conventional drug treatments were not stopped and changes in liver enzymes could not be assessed.

In Case 3, the patient took 1000 mg/day of black cohosh for either 3 months or 8 months (according to the authors, one report documents 3 months of use and another report documents 8 months of use). She stopped taking the black cohosh at first presentation of liver disease; however, she did not stop taking concomitant conventional medications with known potential hepatotoxicity (fluoxetine, propoxyphene, acetaminophen). When black cohosh was discontinued, there was a decrease in ALT, followed by an unexpected increase in ALT above the initial levels. EMEA suggested a possible causality, and the authors of the case report reported a probable causality. However, neither took into account the concomitant medications. Also, the patient had antibodies for a herpes infection and died during liver transplantation. The authors conclude that the patient most probably had herpetic liver disease, so they conclude that the causality to black cohosh must be excluded.

In Case 4, the patient used 80 mg/day of black cohosh for an unknown period of time. The plant part, brand, and manufacturer are unknown. There is not much data available, so the case cannot be examined with the pre-test or main-test. The EMEA reported the case as having only possible causality, citing that the data were not presented in detail. Due to the lack of adequate details, the authors of this critical analysis conclude that the overall causality is unable to be determined.

The authors conclude that the clinical analysis and structured causality assessment reveal that no valid evaluation was possible for one patient due to lack of basic information, and the remaining 3 cases had no convincing evidence that the liver diseases were caused by black cohosh. These 3 patients were all treated with steroids for acute drug-induced hepatocellular jaundice and fulminant liver failure. The authors note that there is no evidence of steroids benefiting this condition, and that since early antiviral therapy is necessary for herpetic liver disease, steroid therapy should not be considered unless all viral causes have been safely excluded. It is interesting that the reanalysis of the data indicate that the EMEA drew inaccurate conclusions. Vigorous causality assessments using a diagnostic algorithm are essential to determine causality for any severe adverse event.

—Heather S. Oliff, PhD