Hepatotoxicity due to herbal medications and dietary supplements

INTRODUCTION — Complementary and alternative medical (CAM) therapies encompass a diverse group of practices that include the use of herbal and dietary supplements (HDS) (table 1) [1]. The use of CAM therapies in the United States and worldwide has been increasing steadily; however, the true prevalence is unknown [2]. In the United States, sales of CAMs in 1994 were $4 billion and increased to $14.8 billion (2008), to $30.2 billion (2012), and over $40 billion by 2019 [3-7]. The most commonly used CAM therapies are herbal and dietary supplements. Survey data have suggested that use of HDS ranges from 10 to over 75 percent depending on the country [8-13]. Up to 20 percent of adults report regular consumption of HDS [5,7,14]. In the United States in 2019, consumers spent over $9.6 billion on herbal products [2]. The most common reason reported for their use is for health improvement and health maintenance [15]. Women, persons younger than age 65 years, and individuals with a higher annual household income have been reported to be more likely to use CAM therapies [1,3,5,16,17].

Use of herbal preparations can be traced back thousands of years to ancient Egypt, China, India, and Sumeria, and formulations have been expanded upon over the centuries [18,19]. Many patients consider herbal remedies to be safe and completely free of unwanted side effects [20]. This is concerning since herbal products have biologic activity that can lead to severe hepatotoxicity or that interact with one another or with prescription medications. Fewer than 40 percent of patients disclose to their clinician that they are using these products, either alone or in conjunction with conventional drugs [1,3,21].

DATABASE OF DRUGS, HERBS, AND SUPPLEMENTS ASSOCIATED WITH HEPATOTOXICITY — It is estimated the number of herbal and dietary supplements (HDS) on the market is over 80,000 [6]. Over 1000 medications and herbal products have been implicated in the development of drug-induced liver injury, and the list continues to grow [13,22]. A searchable database of drugs, herbal medications, and dietary supplements associated with hepatotoxicity has been developed by the National Institutes of Health (NIH) [23]. Additionally, the NIH has developed a dietary supplement label database, containing over 165,000 product labels [24].

REGULATION — Regulation of herbal and dietary supplements (HDS) varies by country. In 1962, the US Kefauver-Harris Drug Amendment required that all over-the-counter and prescription products show proof of safety and efficacy [25]. Complementary and alternative medicines (CAMs) were assigned to the food supplement category and had a lower threshold of required evidence for safety [25]. An attempt by the US Food and Drug Administration (FDA) to develop stricter regulations was met with significant opposition by supplement manufacturers, lobbyists, and consumers.

The Dietary Supplement Health and Education Act (DSHEA) of 1994 was an amendment to the US Federal Food, Drug, and Cosmetic Act (FD&C) [26]. The DSHEA defines a dietary ingredient as a vitamin; mineral; herb or other botanical; amino acid; enzyme, organ or glandular tissue; dietary substance for use by humans to supplement the diet by increasing the total dietary intake; or a concentrate, metabolite, constituent, extract, or combination of the preceding substances [18,27]. These products may be in the form of tablets, capsules, soft gels, gel caps, liquids, teas, or powders. Medical foods are also covered under this regulation. The United States Pharmacopeia–National Formulary (USP–NF) is the official reference for herbal preparations [26]. The USP is an official public standards-setting authority for all prescription and over-the-counter medicines and other health care products manufactured or sold in the United States [28]. The USP also sets standards for food ingredients and dietary supplements.

Products in distribution prior to 1994 were allowed to remain on the market; their use was justified based upon experience from trial and error ("generally recognized as safe"), not rigorous scientific scrutiny. The FDA requires that manufacturers and distributors who wish to market products with "new dietary ingredients" notify the FDA about those ingredients. A new dietary ingredient is one that was not marketed prior to passage of the DSHEA in 1994. Manufacturers are responsible for reporting any side effects and ensuring that their product has not been adulterated. The FDA monitors product information, such as labeling and claims, and also monitors safety through a voluntary dietary supplement adverse event reporting system. Over the years, the FDA has issued warnings on products that have been reported to have hepatotoxicity [29]. The Federal Trade Commission (FTC) regulates advertising of dietary supplements [30].

The DSHEA requires that manufacturers establish product safety before marketing. Manufacturers must also ensure that the product label is truthful and not misleading. The label must contain a complete list of all ingredients contained in the product and the identity of the manufacturer. Manufacturers must also limit the claims that can be made (health claims, nutrient content claims, and structure/function claims) [31]. Structure/function claims are those which claim that the nutrient or dietary ingredient is intended to affect normal structure or function in humans (ie, "calcium builds strong bones"). They may also describe the means by which a nutrient or dietary ingredient acts to maintain such structure/function (ie, "fiber maintains bowel regularity"), or they may describe general wellbeing from consumption of the nutrient or dietary ingredient. Structure/function claims may also express a benefit related to a nutrient deficiency disease (ie, "vitamin C prevents scurvy"). In this last setting, they must include a statement telling how widespread the disease is in the United States. If a manufacturer makes a structure/function claim, the DSHEA requires a disclaimer: "This statement has not been evaluated by the FDA. This product is not intended to diagnose, treat, cure, or prevent any disease."

The Final Rule for Current Good Manufacturing Practices for Dietary Supplements was passed in 2007 as a further step in attempting to ensure the safety of HDS [32]. This rule requires that manufacturers follow certain standards in the production of HDS. Manufacturers are required to verify the contents of the product (ie, identify the ingredients, strength) and ensure that the product is contamination free. However, the Final Rule does not address the safety or efficacy of the product ingredient itself.

The DSHEA requires that a product's label accurately reflect its contents. However, there may be significant discrepancies between the ingredients listed on the label and the actual contents of the preparation. This is especially true of preparations containing multiple plants or herbs (ie, Chinese remedies), which may be adulterated either unintentionally or intentionally by cheaper, more toxic herbs, heavy metals, microbials, and pharmaceutical medicines (such as acetaminophen, aspirin, benzodiazepines, sildenafil, or steroids) [20,33-48]. As many as one-third or more of products can be adulterated [37,49-51]. There can also be variations between batches of the same herbal product from the same manufacturer [33,52-54]. It can be difficult to identify the agent responsible for the hepatotoxicity [55].

The widespread use of mixed remedies and the lack of randomized trials make any review of the safety and efficacy of herbal remedies difficult. In addition, conformity to the USP–NF standards is voluntary and there are no regulatory standards for HDS. There has been a push in the United States to force major policy changes in oversight of the HDS industry as well as grant the FDA the ability to act against manufacturers providing misleading information. Both proposals have been met with resistance from the HDS industry. The European Food Safety Authority (EFSA) monitors claims made on herbal products [56]. Directive 2004/24/EC in the European Union requires that manufacturers register products and comply with specific manufacturing and quality standards [57]. Across the world, regulation and safety monitoring remain limited, and the World Health Organization (WHO) has developed a Traditional Medicine Strategy [58].

FREQUENCY OF HEPATOTOXICITY — Precise estimates of the frequency of hepatotoxicity attributable to herbal and dietary supplements (HDS) are unknown. Patients often do not report the use of herbal products to their clinicians [3,45]. As a result, hepatotoxicity associated with herbal use may be missed. The United States Drug-Induced Liver Injury Network (DILIN) found that 20 percent of cases of drug-induced liver injury (DILI) were attributed to HDS and HDS are the second most common class of agent to cause DILI [6,59-62]. Additionally, this proportion of DILI caused by HDS increased from 7 to 20 percent between 2004 and 2013 [59]. Other studies, depending on the geographic location and type of data collection, have implicated HDS in up to 73 percent of cases of DILI [49,57,59,63-73]. It has been estimated that fewer than 1 percent of adverse reactions to dietary supplements are reported [74].

DIAGNOSIS — Patients are often reluctant to discuss use of herbal and dietary supplements (HDS)/complementary and alternative medicines (CAMs) or do not recognize them as medications [45]. It is important for the clinician to discuss the use of these products with the patient in a nonjudgmental manner. Patients can then be educated on the potential adverse effects and herb-herb or drug-herb interactions that may occur. This may prevent the development of hepatotoxicity. However, the unreported use of herbal remedies must be considered as a possible etiology in any setting of clinical manifestations of liver injury.

The diagnosis of herb-induced liver injury (HILI) can be difficult. Clinical history may be helpful but is not always reliable. The relationship between the exposure and hepatic toxicity (latency) is not always clear and is quite variable. Patients may be taking multiple preparations or multi-ingredient products, making identification of a single offending agent impossible. Due to poor regulation of these products, HDS are prone to intentional or inadvertent adulteration with other ingredients, making identification of the responsible agent difficult [75]. Poor adherence to good manufacturing practices and to quality control contribute to risk of hepatotoxicity [76-78]. Many cases of hepatotoxicity resulted following ingestion of herbal products that contained untested combinations of products, adulterated products, or both [57]. Ingredients found in some HDS products have included unwanted chemicals, prescription medications, heavy metals, and pesticides [75]. HDS products may also vary in composition from batch to batch depending on manufacturing practices. These products are also frequently mislabeled [79]. Patients may also have confounding concomitant liver disease, such as alcohol-associated liver disease or metabolic dysfunction-associated steatohepatitis (MASH), which can produce similar clinical and laboratory features. There is no gold standard, and no specific tests or serum biomarkers exist to confirm a diagnosis of drug-induced liver injury (DILI) from either a prescription medication or an herbal remedy. Proving that an HDS causes HILI relies on both chronologic and clinical criteria [80,81].

A diagnosis of HILI should be suspected in patients with nonspecific symptoms developing after introduction of an HDS (such as nausea, anorexia, malaise, fatigue, right upper quadrant pain, or pruritus), and such symptoms prompt further evaluation. Features suggesting HILI include lack of illness prior to ingesting the HDS, clinical illness or biochemical abnormalities developing after beginning the offending agent, and improvement after the product is withdrawn. Patients may present with asymptomatic liver enzyme elevations, acute or chronic hepatitis (symptomatic malaise, nausea, vomiting, abdominal pain, etc), acute liver failure with coagulopathy and encephalopathy, or with signs and symptoms of cirrhosis. Jaundice is the most common presentation (46 percent) followed by abdominal pain and nausea [82]. Serum aminotransferases may be markedly elevated (up to >20 times the upper limit of normal). The clinical patterns of liver injury are similar to those seen with other forms of DILI. These patterns include hepatocellular, cholestatic, a mixture of the two, or vascular (ie, sinusoidal obstruction syndrome).

Hepatocellular injury is more frequently seen with HILI than with DILI, secondary to conventional medications [6,29,65,83]. If an immunologic reaction is suspected, the illness will generally recur upon reintroduction of the offending substance. For patients with suspected HILI, rechallenge is not advised. Unlike the more stereotypic presentations seen with hepatotoxicity due to most drugs, the clinical presentations of injury secondary to HDS are less typical and have been less well-defined [29]. Exceptions to this are hepatotoxicity related to anabolic steroids and pyrrolizidine alkaloid-containing products. The most common presentation is hepatocellular injury [84]. Female patients and patients presenting with a hepatocellular pattern of injury have a poorer prognosis [67,85,86]. Bile duct loss in the setting of cholestatic pattern of injury is also predictive of a poor outcome [84,87]. (See "Drug-induced liver injury".)

Liver biopsy may be useful if the diagnosis of HILI is uncertain. However, liver histology will not identify a specific product. Studies have reported that among patients who underwent liver biopsy, over 70 percent of patients have hepatocellular injury, followed by sinusoidal obstruction syndrome (approximately 10 percent) and cholestatic liver injury (9 percent) [82].

Continued use of the offending product increases morbidity once hepatotoxicity has developed [34,88]. The patient may even increase the use of the herbal preparation to help manage the new symptoms, further worsening the liver injury. For patients with elevated liver enzymes, other causes of liver disease must be ruled out. Patients often use herbal remedies to treat their chronic disease and may, in fact, exacerbate it [4,5,52]. Continued ingestion of the herbal preparation in the face of ongoing hepatotoxicity may lead to acute liver failure, sinusoidal obstruction syndrome, or cirrhosis [65,89].

The Council for International Organizations of Medical Sciences (CIOMS) developed a series of standard designations for drug-induced liver disorders and classification of injury [90]. The US Food and Drug Administration Drug Hepatotoxicity Steering Committee proposed modifications to the CIOMS classification scheme for classifying hepatotoxicity in clinical trials (table 2) [91]. These guidelines are used as markers of hepatotoxicity in clinical trials.

A number of scales have been developed that attempt to codify causality of drug toxicity into objective criteria [92]. The best-known scale is the Roussel Uclaf Causality Assessment Method (RUCAM) scale. Other scales, including the Maria & Victorino scale and the Naranjo scale, are simpler clinical diagnostic scales [81,93-95]. Studies that have compared these models suggest that the RUCAM scale may have better discriminative ability [80,92,96]. While these scales have not been validated in the setting of herbal medications, they may be useful (table 3). Another scale in development, the Revised Electronic Causality Assessment Method (RECAM), updates and expands upon the RUCAM [97].

None of these scales address all risk factors, and none are used routinely in clinical practice [81]. In addition, none were developed specifically for HDS-induced hepatotoxicity. One of the most important diagnostic criteria remains the exclusion of other causes of liver injury. RUCAM has been validated independently in the setting of drug rechallenge [94]. The Drug-Induced Liver Injury Network (DILIN) developed the DILIN Causality Scoring System to adjudicate the causality of drug-induced injury for patients enrolled into its prospective clinical trial [98]. This model relies on structured expert opinion, and when compared with the RUCAM, it produces higher agreement rates and likelihood scores. However, intraobserver variability remains significant. Unfortunately, the DILIN scale is not a clinically viable option for determining causality since it relies on expert opinion.

Key elements for attributing liver injury to an HDS are similar to those for other drug products and include:

●Exposure must precede the onset of liver injury (although the latent period is highly variable).

●Underlying liver disease should be excluded.

●Injury may improve when the HDS is stopped (although, in some cases, injury may initially worsen for days or weeks while, in cases of acute liver failure, declining liver biochemical tests may indicate deterioration rather than improvement).

●Liver injury may have recurred more rapidly and severely after repeated exposure.

●None of the causality assessments models have been validated in the setting of HDS hepatotoxicity. They are less reliable given the possibility of product variability and contamination [84].

TREATMENT — Early recognition of toxicity is important to permit assessment of severity and monitoring for acute liver failure. The mainstay of therapy for herb-induced liver injury is withdrawal of the offending toxin. (See "Acute liver failure in adults: Etiology, clinical manifestations, and diagnosis".)

Glucocorticoids are of uncertain benefit for most forms of herbal and dietary supplement (HDS) hepatotoxicity, although they may have a role for treating patients with hypersensitivity reactions [99]. Our practice is to give glucocorticoids to patients with hypersensitivity reactions who have progressive cholestasis despite stopping HDS and who have biopsy features that resemble those seen in autoimmune hepatitis. In addition, we give glucocorticoids to patients with extrahepatic manifestations of a hypersensitivity reaction that warrant glucocorticoid treatment (eg, severe pulmonary involvement in patients with drug reaction with eosinophilia and systemic symptoms). (See "Overview of autoimmune hepatitis", section on 'Histology' and "Drug reaction with eosinophilia and systemic symptoms (DRESS)", section on 'Management'.)

Patients should be followed by serial biochemical measurements. Hepatology consultation may be warranted.

Recovery should be expected in the majority of patients after discontinuing the HDS. More advanced disease requires supportive therapy, with acute liver failure often leading to death or need for liver transplantation. In this setting, mortality can be over 80 percent without transplantation. The development of jaundice (bilirubin greater than two times the upper limit of normal, in the setting of an alanine aminotransferase greater than three times the upper limit of normal) following introduction of an HDS potentially portends a poor outcome and should also prompt immediate referral to a center with expertise in hepatology [67,100,101]. Patients who recover from HDS-induced hepatotoxicity generally have a favorable prognosis.

In the United States, the US Food and Drug Administration (FDA) records HDS toxicity. Cases of HDS hepatotoxicity should be reported to 1-800-332-1088 or https://www.fda.gov/safety/reporting-serious-problems-fda/reporting-health-professionals.

SPECIFIC HEPATOTOXIC HERBS — The many reports in the literature of the toxic effects of herbal remedies underscore the view that not all natural products are harmless. Hepatotoxicity may be the most frequent adverse reaction to herbal remedies [102,103]. It is often difficult to pinpoint the product that has caused the injury since most products do not contain a single ingredient. Plant extracts are regarded as multi-ingredient products since the product is a concentration of phytochemicals from the plant [7]. Some of the more frequently reported offending herbs include Chinese remedies and teas (Jin Bu Huan [104,105], Ma-Huang [106-108]), germander [89,109-114], valerian [115], mistletoe [116], skullcap [115-117], chaparral [118-122], comfrey [123-127], herbal teas containing toxic alkaloids [125,128-130], pennyroyal oil [131], and kava (table 4) [103,132-136]. Although liver abnormalities are frequently transient, cases of chronic liver disease and acute liver failure have been described [20,136,137].

The list that follows includes some of the herbs that have been associated with hepatotoxicity. However, the list of herbal medications, dietary supplements, and medications associated with hepatic injury is long. A more complete listing can be found in the searchable database of drugs, herbal medications, and dietary supplements developed by the National Institutes of Health [138].

Pyrrolizidine alkaloids — Pyrrolizidine alkaloids (PAs) are a group of naturally occurring alkaloids based on the structure of pyrrolizidine. More than 660 PAs and PA N-oxides have been identified in over 6000 plant species throughout the world (3 to 5 percent of the world's flowering plants) (table 5) [139,140]. They are used for a variety of conditions, including arthritis, gout, and infections. Over 50 percent of PAs exhibit hepatotoxicity [129,140]. One of the historical examples of hepatotoxicity was reported following the use of "bush tea." Bush tea was brewed predominantly from the Senecio plant, which is high in PAs.

Other plant families that have shown hepatotoxicity after exposure include: Boraginaceae (Heliotropium, Trichodesma, Symphytum officinale [Comfrey]), Compositae (Senecio [Bush Teas], Eupatorium), Crotalaria (Leguminosae), Echinacea (Echinacea purpurea) [141], Germander (Teucrium chamaedrys), Greater Celandine (Chelidonium majus), and Scrophul-ariaceae (Castilleja) [123-125,128-130,142-144]. PA poisoning is common in areas where traditional remedies are frequently used (such as Africa and India) [145,146]. In Western countries, it is seen more sporadically. Comfrey has been removed from the market in France following numerous reports of liver damage. It remains widely available in the United States despite US Food and Drug Administration (FDA) requests to remove comfrey products from the market [125,147].

Alkaloid toxicity usually results in moderate to severe liver damage, particularly to the hepatic central vein and sinusoidal endothelium. Hemorrhagic necrosis, hepatomegaly, ascites, and endothelial cell proliferation are also seen. This leads to occlusion of the hepatic veins, resulting in hepatic sinusoidal obstruction syndrome (formerly veno-occlusive disease), a progressive form of portal hypertension, which often progresses to hepatic failure [148,149]. Sinusoidal obstruction syndrome is the characteristic histologic feature of PA hepatotoxicity [144]. (See "Hepatic sinusoidal obstruction syndrome (veno-occlusive disease) in adults".)

The precise mechanism of hepatic injury is unknown but appears to result from accumulation of highly reactive electrophilic metabolites produced via the cytochrome P450 (CYP450) enzyme system [20,123,131]. The increased concentration of P450 enzymes within the hepatic centrilobular region correlates with the changes seen histologically. The pyrrolic and pyrrole-like metabolites are highly reactive and lead to the formation of deoxyribonucleic acid (DNA) and protein adducts as well as glutathione conjugates [144,150]. The damage is compounded by depletion of glutathione or induction of CYP450 3A4 [18,151]. PAs have also been shown to induce apoptosis [151,152].

Nonthrombotic luminal occlusion of the small centrilobular veins leads to hepatic congestion and subsequent hemorrhagic parenchymal necrosis [20,153]:

●With acute injury, patients typically present with sudden onset abdominal pain, jaundice, conjugated hyperbilirubinemia, and elevated aminotransferases. Hepatomegaly and ascites are common.

●The subacute and chronic forms are more insidious and may mimic cirrhosis and portal hypertension from other causes. However, histology shows a distinctive pattern of centrilobular injury [148].

Hepatotoxicity from PAs is dose and time dependent and reproducible [154]. Approximately one-half of patients with hepatotoxicity develop limited lesions and generally proceed to complete recovery once the offending herb has been discontinued [148,154]. More extensive disease can lead to cirrhosis, hepatic failure, and death [20,154]. The acute form is rapidly fatal in 20 to 40 percent of patients (with worse prognosis in adults compared with children). Approximately 15 percent with acute disease will progress to subacute or chronic injury, succumbing within several years to end-stage liver disease. Of the remainder, most develop cirrhosis and portal hypertension [34,148,154,155]. Death may occur two weeks to more than two years after poisoning, but patients may recover almost completely if the herb intake is discontinued and liver damage is not severe.

Ayurvedic herbs — Ayurvedic medicine originated in ancient India. These herbal and dietary supplement (HDS) products have been used most commonly on the Indian subcontinent in the practice of traditional medicine (Ayurveda), but their use is expanding worldwide. Ayurveda uses many medicinal plants that contain toxic pyrrolizidine alkaloids [156,157]. They include Psoralea corylifolia, Centella asiatica, Acacia catechu, Eclipta alba, and Vetivexia [66,157-159]. Both acute and chronic hepatitis have been reported. Additionally, many of these products have been found to be adulterated with heavy metals, including lead, mercury, or arsenic [46,160]. Ashwagandha (Withania somnifera) is popular in Ayurvedic medicine. Although generally considered safe, there have been reports of cholestatic liver injury following its ingestion [161,162].

Germander — The blossoms of germander (Teucrium chamaedrys) have been used for thousands of years for a variety of symptoms [163]. It is used in tea form or alcohol-based extracts. The herb is said to have choleretic properties, although this has never been scientifically proven. It is also used for dyspepsia, hypertension, diabetes, gout, and obesity [139]. Liver injury can present as acute hepatitis, chronic hepatitis, or acute liver failure [89,109,110,163,164]. The herb has been removed from the market in France because of its hepatotoxicity [112,113]. The majority of cases of hepatotoxicity have occurred at the manufacturers' recommended doses (600 to 1600 mg/day). Clinical symptoms generally develop after approximately two to three months of ingestion and include a nonspecific hepatitis, which generally runs a benign course [20]. Typical features include anorexia, nausea, abdominal pain, and jaundice associated with a marked elevation in serum aminotransferases. Complete recovery is usually seen within two to six months following discontinuation of the herb. Fatal cases have been described, as have more insidious presentations with progression to cirrhosis [18,20,89,111]. Cirrhosis generally develops in patients who have ingested the offending herb for longer periods of time or in larger quantities.

Germander contains toxic alkaloids, the furanic neoclerodane diterpenes (NCDs), which are oxidized by the CYP450 3A4 enzyme system into hepatotoxic reactive metabolites [109,165]. The toxic metabolites accumulate and covalently bind to cellular proteins, which in association with the depletion of the free-radical scavenger glutathione, leads to acute centrilobular hepatocyte necrosis in vivo. In vitro studies have revealed progressive apoptosis and cytoskeletal disorganization [165].

Poley (Teucrium polium) is in the same genus as germander and is used for diabetes (in Israel), gastric complaints (in North Africa), fever (in Italy), and as a poultice (in Spain). It has also been linked to hepatotoxicity and acute liver failure [114,166,167]. Teucrium capitatum and T. viscidum have also been reported to cause hepatotoxicity [168,169].

Greater celandine — Greater celandine (Chelidonium majus) is a member of the poppy family that grows mainly in Europe and Asia but has been introduced into North America. It is considered a cleansing herb and is said to stimulate pancreatic enzymes and bile. In fact, it has been shown to increase biliary flow in experimental models [170]. Therefore, it is often used for hepatobiliary (jaundice, hepatitis, and gallbladder) and digestive tract (dyspepsia, gastroenteritis, irritable bowel syndrome, constipation, anorexia, stomach cancer, intestinal polyps) complaints [171]. The plant contains over 20 alkaloids [171]. When used orally, it has been implicated in numerous cases of liver injury [143,171-178]. Injury generally develops after one to six months of use. Jaundice is generally the predominant symptom, but anorexia, nausea, vomiting, and abdominal pain and itching have been reported. Liver enzyme elevations range from 100 to over 4000 units/L, and injury is hepatocellular, resembling acute viral hepatitis in the majority of cases [178]. However, cholestatic hepatitis has also been reported [143,172]. The mechanism of hepatotoxicity is unclear. Liver injury appears to show a female predominance [178,179]. Histologic examination of the involved liver has suggested metabolic hepatotoxicity [143,172,178]. There have been no reports of fatalities and discontinuing generally leads to resolution of liver injury over two to six months. Greater celandine has been banned from use in many European countries.

Chaparral — Chaparral (Larrea tridentata) is indigenous to the southwestern United States and is derived from the leaves of the creosote bush (greasewood). It is said to have antioxidant, antiamebic, antibacterial, antifungal, and antiviral properties. It is used for diarrhea, cramps, chest pain, upper respiratory tract infections, rheumatism, cutaneous venereal lesions, chronic skin disorders, and weight loss [154]. It has been popular among patients with human immunodeficiency virus (HIV) infection [180]. It is available in the form of tea, capsules, tablets, and salves [118]. Scientific evidence regarding chaparral's effectiveness is lacking.

Patients with chaparral-induced hepatotoxicity typically present with acute hepatitis [118]. Marked elevation in the serum aminotransferase and bilirubin values may be seen. Alkaline phosphatase may be mildly elevated. Acute hepatocellular injury of varying degrees of severity has been reported, including mild cholestatic hepatitis, severe collapse of parenchymal architecture, hepatic fibrosis, and massive liver failure requiring transplantation [118-122,181]. The onset of disease occurs 3 to 52 weeks after the ingestion and resolves in 1 to 17 weeks in most cases [118]. Progression to cirrhosis or acute liver failure requiring liver transplantation has been reported [118]. There have been no reported cases of chaparral hepatotoxicity published in over 10 years [138].

Atractylis gummifera and callilepis laureola — Atractylis gummifera grows in the Mediterranean and is used as an antipyretic, emetic, abortifacient, and diuretic, and hepatotoxicity continues to be reported [20,182,183]. Toxicity is secondary to two diterpenoid glucosides: atractyloside and carboxyatractyloside [184]. These toxins inhibit mitochondrial oxidative phosphorylation by interacting with a mitochondrial adenine nucleotide translocator, inducing oxidative stress [184,185]. The onset of toxicity is rapid, occurring within hours after ingestion. It often leads to acute liver and kidney failure and subsequent death.

Callilepis laureola (Impila) is native to South Africa [186,187]. Like A. gummifera, it contains atractyloside. Its toxicity has been linked to depletion of cellular glutathione. It has a similar rapid onset of action, and mortality rates are over 90 percent by five days after ingestion. There have been reports that supplementation with N-acetylcysteine may reduce its toxicity [186].

Pennyroyal — Pennyroyal tea and leaf extracts are generally considered to be safe. However, pennyroyal oil (Mentha pulegium or Hedeoma pulegoides) is a highly toxic agent that can cause both hepatic and neurologic injury even in small doses (ie,15 mL) [138]. It is frequently used to flavor herbal remedies and teas (particularly mint teas) and is often used for digestive disorders, liver disorders, amenorrhea, gout, colds, increased micturition, and for skin diseases. It also continues to be used as an abortifacient. Acute toxicity leads to gastrointestinal distress and central nervous system effects within one to two hours of ingestion, and fatal ingestions can occur with ingestion of >15 mL of the oil [188].

Pennyroyal oil contains 85 percent pulegone. CYP450 1A2 and 2E1 produce oxidative metabolites (menthofuran) of pulegone that are associated with severe hepatic toxicity with centrilobular degeneration and massive hepatocellular necrosis [154]. These metabolites bind to cellular proteins and deplete hepatic glutathione [131,189,190]. The use of N-acetylcysteine may help minimize hepatic injury [188].

Mistletoe — Mistletoe is used in traditional medication and refers to two unrelated plants: American mistletoe (Phoradendron leucarpum) and European mistletoe (Viscum album). They are purported to be effective in treating asthma, epilepsy, arteriosclerosis, cardiac arrhythmia, cramps, blood loss, nervous conditions, high blood pressure, joint pain, dizziness, and infertility [116]. There have been case reports of hepatitis [116]. Lectins in mistletoe products induce apoptosis and can also stimulate the immune system [29]. However, controversy remains as to whether mistletoe actually causes liver disease. Both appear to be safe at conventional doses and are an unlikely cause of symptomatic liver injury [138]. In reported cases, the ingested capsules were also laced with Skullcap (Scutellaria galericulata). Skullcap and Valerian (Valeriana officinalis) have been implicated in liver injury, although there are no experimental data supporting their toxicity [20,115].

Kava-kava — Kava (Piper methysticum) is used as a dietary supplement and has been used in the South Pacific for centuries for complaints of anxiety, sleeplessness, and menopausal symptoms. Numerous reports of severe hepatotoxicity and liver failure have been described in Europe and the United States [103,132-136,191-193]. Hepatotoxicity has been reported within a few weeks to up to two years (average 2 to 24 weeks) following ingestion [191].

Clinically, reports of toxicity range from transient elevation in aminotransferases to acute liver failure and death [103,138]. Liver injury has been characterized histologically with hepatic necrosis, cholestasis, and lobular inflammation. The mechanism of hepatotoxicity is unclear, but multiple hepatotoxic components have been identified, including pipermethystine and flavokavain B [194,195]. The aerial portions of the plant contain the toxic alkaloid pipermethysticin and are generally not consumed [194,195].

Based upon case reports of serious toxicity, the FDA issued a consumer advisory regarding this toxic herbal product. Details of the advisory are available online. The product has been banned from many European markets (including Germany, France, Great Britain, and Switzerland). The link of toxicity to kava remains debated [194]. Some propose that the aqueous preparation is safe [196] while others have shown that even this preparation leads to hepatotoxicity [197].

Green tea extracts — Green tea is widely consumed, is safe in moderation, and may have beneficial effects. Concentrated extracts of green tea (Camellia sinensis L.) have been associated with hepatocellular injury, particularly in high doses (>800 mg) or prolonged exposure [7,198-201]. Numerous supplement extracts are available and lead to heightened exposure to the catechin polyphenols. These extracts have been used in multiple weight-loss products, many of which have been removed from the market due to toxicity.

The precise mechanism of hepatotoxicity remains unclear. The extraction process increases the catechin content as well as potential contaminants [7]. Catechins such as epigallocatechin-3-gallate (EGCG) are likely to be involved. EGCG constitutes over 50 percent of the catechin content. EGCG induces reactive oxygen species similar to other herbal products [202]. This may overwhelm hepatic drug-metabolizing capability [201]. A genetic predisposition may also predispose to injury (HLA-B*35:01, C*04:01), and females are more susceptible than males [198,203,204]. Hepatotoxicity is increased in those who are fasting [141,205]. The pattern of injury is predominantly hepatocellular, but mixed and cholestatic presentations have been reported [6,202,206]. Liver injury generally develops within three months of use and usually resolves with discontinuation of the product; however, fatalities have been reported. The United States Pharmacopeia–National Formulary has proposed a cautionary labeling requirement on products with green tea extract, which has not been implemented [201].

Anabolic (muscle-building) supplements — Anabolic steroid use has been associated with many types of liver injury, including hepatitis, intrahepatic cholestasis, peliosis hepatis, hepatocellular adenoma, and hepatocellular carcinoma [207,208]. Selective androgen receptor modulators (SARMs) are found in some supplements marketed for muscle-building, and they mimic the effects of testosterone and anabolic steroids [209]. Despite regulation and known adverse effects, including hepatotoxicity, use of these agents remains widespread, and patients continue to present with liver injury [210,211]. The classic presentation is one of hepatocellular injury with severe jaundice at a median of approximately 70 days following use [212]. Patients who present with markedly elevated bilirubin levels are also prone to acute kidney injury [29]. The FDA has issued public health advisories warning consumers to avoid body-building products containing anabolic steroids or steroid-like compounds [213,214]. (See "High-risk dietary supplements: Patient evaluation and counseling", section on 'Sports supplements'.)

Black cohosh — Black cohosh (Cimicifuga racemose syn. Actaea racemosa) is indigenous to eastern North America. It is widely available, and the rhizomes and roots are commonly used for the treatment of peri- and postmenopausal symptoms. The product contains many compounds, including phenolic acids, flavonoids, volatile oils, triterpene glycosides, and tannins [141]. There have been multiple reports of hepatotoxicity and adulteration (particularly with Asian Cimicifuga species) [215,216]. Reported toxicity has been an autoimmune-like liver injury, and liver failure necessitating transplantation has been reported [141,215,217-220]. However, the causality between black cohosh and hepatotoxicity remains uncertain. Subsequent meta-analyses have suggested that this product was effective and safe [221,222].

Flavocoxid — Flavocoxid is a prescription medical food that is a blend of two flavonoids, baicalin and catechins (derived from the plants Scutellaria baicalensis and Acacia catechu, respectively). It has been used to treat osteoarthritis. The FDA recommended discontinuation of flavocoxid because of reports of serious adverse events, including drug-induced liver injury (DILI) [223]. In a prospective study of 877 patients, four patients (0.4 percent) developed liver injury [224]. The association between flavocoxid use and liver injury was deemed highly likely in three cases and possible in one. In all four cases, liver tests returned to normal within 3 to 12 weeks of discontinuing the flavocoxid.

Kratom — Kratom (Mitragyna speciosa) is native to Southeast Asia and has been traditionally used to treat diarrhea or for its stimulant effect. In the United States, most users take the product for pain, depression, or as an anxiolytic [225]. Its effects are secondary to mitragynine and 7-hydroxymitragynine effects on the mu-opioid (agonist) receptors. (See "Opioid withdrawal: Medically supervised withdrawal during treatment for opioid use disorder", section on 'Kratom'.)

Kratom has been associated with liver injury [226,227]. Liver injury may be severe; both hepatocellular and mixed patterns have been reported. Despite discontinuation of the product, bilirubin may continue to rise before normalizing. Kratom products have been banned in some countries [7]. The FDA attempted to classify Kratom as a schedule I drug, but lobbyists prevented this [227,228]. This has resulted in increasing reports of injury.

Cannabidiol (CBD) — CBD is a non-psychotropic, biologically active component of marijuana (Cannabis sativa). There has been little oversight into the manufacture of CBD products, and sale of both substandard and adulterated products have been reported [7]. Clinical trials with CBD noted that up to 20 percent of those taking it developed elevated liver enzymes, generally after several months of chronic use. The mechanism of hepatic injury remains unclear.

CHINESE HERBAL REMEDIES — Over 11,000 species of medicinal plants are used in China [141]. Use of traditional Chinese medicines (TCMs) has grown globally. Many TCMs have beneficial effects [36]. However, published reports on hepatotoxicity with TCM have been increasing [36,103,104,108,229-232]. Evidence regarding the incidence and prevalence rates of hepatotoxicity with these products is uncertain [233]. Studies of drug-induced liver injury (DILI) that include TCMs report widely varying data, with TCMs accounting for 4 to 55 percent of cases, depending on the patient cohort [232]. Specific products include Ba Jiao Lian (Dysosma pleianthum), Bol Gol Zhee (Fructus psoraleae), Chi R Yun (Breynia officinalis), Jin Bu Huan (Lycopodium serratum), Jing Tian San Qi (Sedum aizoon syn. Stonecrop), Ma-Huang (Ephedra), Shan Chi (Gynura segetum), Sho Saiko To, and Shou Wu Pian (Polygonum multiflorum).

Much of herbal-induced liver injury (HILI) due to TCMs are believed to be due to herbs containing pyrrolizidine alkaloids [234]. TCMs that have been shown to cause sinusoidal obstruction syndrome include Jing Tian San Qi (Sedum aizoon syn. Stonecrop), Shan Chi (Gynura segetum), and Senecio vulgaris [235-237]. The TCM Shi Can contains Teucrium chamaedrys (Germander) or other Teucrium species, which are a known cause of HILI [89].

The clinical course for patients with hepatotoxicity ranges from elevated liver enzymes [229,230] to acute liver failure and death or need for liver transplantation [137]. Up to 70 percent of these products produce reactive metabolites that cause hepatotoxicity [238]. It is often difficult to identify the agent responsible for the hepatotoxicity since most Chinese preparations are either mixtures of multiple herbal preparations (typically four to six different herbs) or have been adulterated with substituted herbs, heavy metals, or pharmaceutical medicines [20,33,35,37-40,50,51,66]. Generally, a single herb is considered primary (the "king" or "sovereign"), while the additional herbs are thought to act synergistically with the main product or to modify its toxicity [13]. Toxicity is generally observed with preparations that contain aconitine, podophyllin, or anticholinergics [230].

Jin Bu Huan — Jin Bu Huan (Lycopodium serratum) has been used for over 1000 years as a sedative and analgesic [104,163,164]. It has morphine-like properties due to the alkaloid levo-tetrahydropalmatine. It was banned for use in the United States in the 1990s following multiple reports of toxicity [104,105,239].

Both acute and chronic hepatotoxicity have been reported [104,105,239]. Patients with hepatotoxicity present at a mean of 20 weeks (7 to 52 weeks) after ingestion with acute hepatitis associated with fever, nausea, vomiting, fatigue, pruritus, abdominal pain, hepatomegaly, and jaundice. Plasma aminotransferase levels are significantly elevated. The liver biopsies that have been performed appear consistent with a DILI, demonstrating eosinophils and focal hepatocellular necrosis [104,105]. Fibrosis and microvesicular steatosis have also been reported [104].

Symptoms resolve over a mean of eight weeks (2 to 30 weeks) with discontinuation of the herb and have promptly recurred when patients were rechallenged [104,105]. The hepatotoxic mechanism is not known but may be immunologic given the histologic presentation [163].

Ma-Huang (ephedra) — Ma-Huang (Ephedra sinica), an ephedra alkaloid, has been used for centuries for cough, bronchitis, fever, joint symptoms, edema, bone pain, to induce weight loss, and as a central nervous system stimulant [106,108]. The active ingredient, ephedrine, has been associated with psychiatric and autonomic system adverse events. Overall, the cardiovascular and neurologic side effects are more common than liver-related effects. Use of the drug is associated with reports of hepatotoxicity, including hepatitis, acute liver failure, and exacerbation of autoimmune hepatitis [106-108,240,241]. The mechanism of toxicity is not known. Affected patients present in similar fashion to those taking Jin Bu Huan [106,108]. Biopsies revealed diffuse severe hepatocellular necrosis and polymorphonuclear cell infiltrates with occasional eosinophils. Abnormal liver tests usually normalize within four months of discontinuation of drug [241]. Ephedra was withdrawn from the United States market in 2004.

Syo-saiko-to — Syo-saiko-to (Dai-saiko-to, xiao-chai-hu-tang, da-chai-hu-tang, TJ-19) is an herbal product used in China to treat fever and gastrointestinal disorders. It has been used in Japan for treating chronic hepatitis, particularly hepatitis C virus infection. This remedy contains numerous herbs (Bupleurium root, Pinelliae tuber, Scutellaria baicalensis root, ginseng root, ginger rhizome, glycyrrhiza root, and jujube fruit). The components in this product differ in their proportion of herbal products. There is some evidence that these products have anti-inflammatory, antioxidant, antifibrotic, and antitumorigenic activity [242]. There have been multiple reports of hepatotoxicity associated with this product; however, some occurring up to three months after ingestion and recurring with rechallenge [243-245]. The offending component is suspected to be the Chinese skullcap (Scutellaria baicalensis). The histologic picture is one of centrilobular confluent or spotty necrosis, microvesicular steatosis, acidophil bodies, and possibly granulomas [243]. Patients should be cautioned against using this herbal remedy.

Chaso or Onshido — Chaso and Onishido are used for weight loss. The toxic ingredient in these formulations is thought to be N-nitroso-fenfluramine, which depletes adenosine triphosphate (ATP) and impairs mitochondrial oxidative phosphorylation [29]. This results in diffuse hepatic necrosis and nonspecific inflammation. Cases of acute liver failure, death, and liver transplantation have been reported [246].

Shui-zhi-zi or sansisi — These products contain Gardenia jasminoides and are used to treat fever, liver disease, or cancer. This product contains geniposide, which has been shown to have dose dependent hepatotoxicity [247].

Polygonum multiflorum — This product has been used to improve fertility and to enhance hair color. Reports of hepatoxicity have been reported and are frequently associated with the HLA-B*35:01 allele [231,248].

WEIGHT-LOSS PRODUCTS — Numerous proprietary preparations have been associated with hepatotoxicity, including Hydroxycut [249-254], Herbalife [44,255-260], Enzyte [261], N.O.-XPLODE [212], OxyELITE [262], SlimQuick [203], and Exolise [203,206].

Herbalife products — These products are sold worldwide in the form of drinks, tablets, capsules, and energy bars. There are multiple reports of hepatotoxicity with these products, ranging from mild to severe enough to require liver transplantation [44,255,257-259,263-265]. The pattern of liver injury is hepatocellular in the majority of cases; however, there are also reports of both mixed and cholestatic injury. The mechanism of toxicity remains unknown.

Hydroxycut — Hydroxycut is a weight-loss product that contains multiple herbal products. The original formulation contained ephedra and was withdrawn from the market in 2004.The product was reformulated, and 23 reported cases of liver injury were reported. In 2009, the US Food and Drug Administration (FDA) issued a warning regarding the product’s potential hepatotoxicity and recommended that consumers discontinue use of the products carrying the brand name Hydroxycut [253,263,266,267]. The manufacturer voluntarily recalled all of its products and again reformulated the herbal mix [263]. Reformulations have included different alkaloid products, including Camellia sinensis (117 mg [epigallocatchein-3-gallate [EGCG]), yohimbe extract, garcinia cambogia, and green tea extract [7]. Hepatotoxicity generally presents within the first three months of taking the product. Patients generally have hepatocellular injury with elevated liver enzymes; however, both cholestatic injury and mixed acute hepatitis with jaundice have been reported. Although infrequent, there are still reports of liver injury in the setting of Hydroxycut.

Supplements containing Garcinia cambogia — Extracts of garcinia contain xanthones, benzophenones, amino and organic acids, including hydroxycitric acid [138]. Dietary supplements containing Garcinia cambogia are widely marketed for weight loss. As their use has gained popularity in the United States, cases of acute hepatitis and acute liver failure requiring liver transplantation have been reported [268-270]. It has been postulated that the active ingredient, hydroxycitric acid, is the cause of liver injury. This ingredient was banned by the FDA because of hepatotoxicity. Additionally, the FDA has recommended that specific products containing Garcinia cambogia be avoided as they are often contaminated with sibutramine [271]. In a case report, sibutramine was associated with hepatotoxicity, and it was removed from the US market for cardiovascular toxicity [272]. (See "Obesity in adults: Drug therapy", section on 'Therapies not recommended'.)

OxyELITE Pro — The original product contained 1,3-dimethylamylamine (DMAA), which had been noted to have caused acute liver injury [273]. Following a report of seven patients who developed severe hepatitis or acute liver failure while taking the product OxyELITE Pro [274], the initial formulation was removed from the market in 2013. The formulation was changed by the manufacturer to remove DMAA and add aegeline [275]. Aegeline is extracted from the fruit Aegle marmelos (bael) and has been used in Ayurvedic medicine. Its main alkaloid is N-[2-hydroxy-2(4-methoxyphenyl) ethyl]-3-phenyl-2-propenamide [276]. Following its release, further cases of severe liver injury were reported [234,262,274]. A total of 55 cases of liver injury were reported with the new formulation [6,62,275]. The majority of those affected were young (mean age 33.9 years) and male (60 percent). Many patients required hospitalization and liver transplantation, and deaths were reported [262]. The FDA banned the product in October 2013 and the manufacturer issued a recall of all products in November 2013.

OTHER HEPATOTOXIC HERBS — There are many case reports of hepatotoxicity with other herbal preparations. These include Aloe (A. vera; A. barbadensis, A. arborescens), Bajiaolian, Borage (Borago officinalis), Broom Corn (Sorghum vulgare), Camphor oil, Centella asiatica (gotu kola), Margosa oil (Azadirachza indica), Cocaine, Coltsfoot, Copaltra (Coutarea latiflora and Centaurium erythraea), Kombucha tea, LipoKinetix (a capsule used for weight loss that contains multiple ingredients including usnic acid from the lichen Usnea barbata), Noni juice (Morinda citrifolia), Cascara sagrada, Sassafras (Sassafras albidum), and Senna (Cassia angustifolia) [20,34,103,154,158,163,164,172,206,246,256,277-287].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Drug-induced liver injury".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Drug-induced hepatitis (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Background – Herbal remedies may be associated with hepatotoxicity, making it important to discuss the use of herbal and dietary supplements (HDS) and complementary and alternative medical (CAM) therapies with patients to prevent potential complications, such as herb-drug interactions. (See "Overview of herbal medicine and dietary supplements".)

The use of herbal remedies should be considered as a possible etiology in any setting of clinical manifestations of liver injury.

●Clinical features and diagnosis – Proving that an herbal or dietary supplement causes liver injury relies on both chronologic and clinical criteria. Features suggesting drug toxicity include lack of illness prior to ingesting the drug, clinical illness or biochemical abnormalities developing after beginning the drug, and improvement after the product is withdrawn. (See 'Diagnosis' above.)

Obtaining a sample of the implicated HDS or CAM therapies for testing is possible, but expertise is not widely available. Costs are generally prohibitive, and most research laboratories are capable of testing only for certain compounds.

●Management – The mainstay of therapy for herbal hepatotoxicity is withdrawal of the offending toxin and supportive care. (See 'Treatment' above.)

●Database of hepatoxic drugs and supplements – A searchable database of hepatotoxic drugs, herbal medications, and dietary supplements has been developed by the National Institutes of Health.