Merck Manual

Please confirm that you are a health care professional

honeypot link
Professional Version

Hepatotoxins in Small Animals


Sharon A. Center

, DVM, DACVIM, Department of Clinical Sciences, College of Veterinary Medicine, Cornell University

Reviewed/Revised Aug 2023

Drug-associated Hepatotoxicity

Although many drugs have been associated with hepatic dysfunction, their influence on liver pathology varies depending on the pathomechanism of liver injury and the acinar zone of metabolic or circulatory disturbance.

Anticonvulsants and Hepatotoxicity in Small Animals

Many anticonvulsants can produce hepatotoxicity in animals. Primidone, phenytoin, phenobarbital, and zonisamide can each provoke idiosyncratic acute fulminant liver failure, or in the case of the first three drugs, chronic hepatitis or cholestatic liver disease.

Chronic phenobarbital can also lead to a diffuse progressive degenerative vacuolar hepatopathy (VH), characterized by proliferative foci with marginating bands of degenerative VH foci with a moth-eaten appearance. This can transform into a hepatocutaneous hepatopathy with or without cutaneous involvement (also known as necrolytic migratory erythema, superficial epidermal necrolysis, or hepatocutaneous hepatopathy) typified by severe plasma amino acid depletion and renal aminoaciduria with prodigious renal lysine and proline wasting. Being a cytochrome P450 enzyme inducer, phenobarbital probably also escalates hepatotoxin toxicity through increased reactive metabolite synthesis.

Glucocorticoids and Hepatotoxicity in Small Animals

Glucocorticoid treatment can produce hepatotoxicity in dogs and sometimes in cats in a diffuse glycogen-like VH (steroid hepatopathy). This is usually a benign and reversible process. Increased alkaline phosphatase (ALP) activity and to a lesser extent ALT activity are present as early as 2 days after glucocorticoid exposure in dogs and slowly resolve over weeks after discontinued treatment. Exposure to high-dose, longterm glucocorticoid administration can lead to a severe diffuse degenerative VH in dogs that may be associated with severe hepatomegaly, jaundice, nodular hepatopathy, and rarely ascites. Glucocorticoid treatment also may provoke hepatic lipidosis in inappetent cats, especially overconditioned hyporexic patients.

Chemotherapy Drugs and Hepatotoxicity in Small Animals

Hepatotoxicosis in humans is dose related to individual differences in pharmacogenetics affecting azathioprine (AZA) metabolism or is idiosyncratic. AZA, a purine analog, is metabolized to 6-mercaptopurine (6-MP) via reduction by glutathione (GSH) and other sulfhydryl-containing compounds; 6-MP is then enzymatically converted into 6-thiouric acid, 6-methyl-mercaptopurine, and 6-thioguanine. Intercalation of 6-MP into DNA blocks the de novo pathway of purine synthesis.

Liver injury is associated with oxidative stress secondary to GSH depletion (GSH utilization in conversion of AZA to 6-MP or in neutralizing reactive oxygen species [ROS] from other metabolic reactions) and mitochondrial damage leading to loss of ATP and cell necrosis. In rodent models, N-acetylcysteine (GSH precursor) and intravascular ROS scavenger protect against azathioprine hepatotoxicity.

Azathioprine is rarely hepatotoxic to dogs and is not advised for cats because of their low drug tolerance. In one retrospective review of dogs treated with azathioprine, 5 of 34 (15%) developed biochemical evidence of hepatotoxicosis (ALT activity > 2-fold baseline values) within 3–23 days of first treatment. The pattern of hepatotoxicosis in all dogs was mixed hepatocellular (median ALT activity: 9× baseline) and cholestatic (median ALP activity: 8× baseline). No dog developed jaundice, biochemical features consistent with liver failure, or ascites, and no dog died. Presumed hepatotoxicity stabilized and resolved with drug discontinuation or drug dose decrease. No liver biopsies were evaluated.

Based on extensive experimental work (rodents), dogs with AZA hepatotoxicity should have the drug discontinued and replaced by another immunomodulating medication or have a drug holiday followed by AZA dose decrease. Dogs should also receive a GSH donor, N-acetylcysteine IV if hospitalized, or bioavailable S-adenosylmethionine (SAMe) if oral medications are tolerated. It is unknown if preemptive treatment with a SAMe product would avert possibility of AZA hepatotoxicity.

Lomustine (CCNU) is a nitrosourea chemotherapeutic agent used mostly in dogs. Delayed progressive hepatotoxicity is the principal dose-limiting adverse effect in dogs. Hepatotoxicity is idiosyncratic, dose related, and irreversible with progressive portal tract and hepatocyte injury, eventually culminating in cirrhosis. ascites, and liver failure. Serious liver injury is estimated to impact up to 6% of dogs receiving lomustine.

Persistent toxic effects after a single lomustine dose were initially demonstrated by time-mortality curves in rodents and studies in healthy dogs in which continued increase of liver enzyme activity (ALT, ALP, gamma-glutamyl transferase [GGT]) persisted for weeks. Delayed hepatotoxicity may be first evident as early as 2 weeks but often is later (ie, 11 weeks or more) after drug administration.

CCNU undergoes rapid spontaneous decomposition and microsomal enzyme metabolism yielding chloroethyl moieties (that retain alkylating activity) and cyclohexyl isocyanates (can carbamylate amino groups in proteins). Pharmacologic studies detailed prolonged plasma residence of high concentrations of CCNU degradation products, eventually eliminated by renal and biliary elimination. This prolonged contact might contribute to delayed hepatotoxicity.

Prolonged concentrations of circulating degradation products may reflect metabolite-plasma protein binding or enterohepatic recirculation. The latter possibility is favored because of the tropism for hepatotoxicity. Experimental work demonstrated a protective influence of P450 enzyme induction (ie, phenobarbital) that might escalate metabolic detoxication and elimination of injurious metabolites. However, this has not been evaluated in dogs and has unknown impact on the chemotherapeutic effect of CCNU.

Metabolites of CCNU are electrophilic and attack nucleophilic sites on the DNA to form alkylated products, impacting gene transcriptional responses. It has been speculated that the isocyanate metabolite may carbamylate hepatocyte and cholangiocyte tubulin, interrupting cell division and bile section, respectively. Oxidative cell injury may also be involved. Pretreatment of dogs with bioavailable SAMe combined with low-dose silibinin (extract of milk thistle) was reported to be partially protective against CCNU hepatotoxicity (based on liver enzyme activity).

However, interpretation of this study is complicated by the absence of liver biopsies in dogs with suspected hepatotoxicity, no declaration of testing intervals for detection of delayed CCNU toxicity, and distribution of 6 of 7 Labrador Retrievers in the group (n = 25) not receiving nutraceuticals. These concerns regard the absence of a gold standard liver biopsy for definite characterization of liver injury, underdetection of latent liver injury that may develop over weeks to months after drug administration, and predisposition of Labrador Retrievers to copper-associated hepatopathy that could amplify any cause of liver injury.

Antiparasitic Drugs and Hepatotoxicity in Small Animals

Thiacetarsamide, an antiparasitic previously used as an adulticide to treat dirofilariasis, causes hepatotoxicity associated with its arsenical content. No hepatotoxicity has been noted secondary to melarsomine (also an arsenical), the current preferred heartworm adulticide. Thiacetarsamide led to hepatotoxicity in a subset of dogs in which it increased ALT activity and, in some dogs, jaundice. Increased liver enzyme activity after the first drug dose was used as an indication to suspend treatment; thereafter, hepatic injury resolved.

In mammals, the liver is the primary site for bioactivation and detoxication of arsenic, involving a series of oxidative methylation and reduction steps. This process generates injurious reactive oxygen species (ROS) that acutely decline hepatic antioxidant capacity. Discontinuing arsenical exposure until plasma liver enzymes declined theoretically allowed an induction response stepping up protective antioxidant enzymes and GSH rejuvenation. Theoretically, with thiacetarsamide, if an initial dose incited substantial ALT activity, treatment was discontinued and restarted when enzyme activity declined, affording greater antioxidant protection during the continued treatment protocol.

Mebendazole (methyl 5-benzoylbenzimidazole-2-carbamate) is an anthelmintic that was widely used in dogs for a short time 30 years ago. This drug provoked idiosyncratic hepatotoxicity in a few dogs, leading to fatal liver necrosis or chronic hepatitis. Chronic oxibendazole-diethylcarbamazine administration to dogs, also popular 30 years ago, also associated with hepatotoxicity in some dogs, leading to increased ALT and ALP activities, hyperbilirubinemia, periportal hepatitis, and evolving fibrosis. Predisposition for Doberman Pinschers was observed. Progressive injury and clinical signs resolved in many but not all dogs after drug discontinuation. In some dogs, idiosyncratic hepatotoxicity was lethal.

Nonsteroidal Anti-inflammatory Drugs and Hepatotoxicity in Small Animals

Many NSAIDs are mitochondrial toxins, and idiosyncratic hepatotoxicity has been observed with increased to least frequency with carprofen, deracoxib, etodolac, previcox, meloxicam, and ibuprofen. Carprofen was reported to induce idiosyncratic hepatic necrosis in some dogs, particularly Labrador Retrievers. Affected dogs may recover fully if NSAID-associated hepatotoxicity is recognized early and drug administration suspended.

It is suspected that NSAID-associated hepatotoxicity is augmented by concurrent hepatic copper accumulation. Drug-associated injury is initiated by toxic oxidative metabolites (formed by centrilobular hepatocytes by cytochrome P450 metabolism) and worsened by the oxidative impact of colocalized copper. In the case of acute ingestion of toxic amounts of carprofen, activated charcoal is as effective as emesis in reducing realized total dose. Additionally, there is evidence that ibuprofen clearance is facilitated by administration of cholestyramine.

An extensive study of hepatotoxicity induced by the NSAID diclofenac in dogs characterized direct and indirect mechanisms leading to an immunoallergic form of chronic hepatitis that is not associated with marked liver enzyme activity, evidence of liver failure, or jaundice. Induced liver hypersensitivity involved microvesicular lipid vacuolation, glycogen depletion, and accelerated hepatocyte apoptosis, reflecting oxidative stress and mitochondrial dysfunction, macrophage activation, mastocytosis, complement activation, and erroneous programming of the innate and adaptive immune responses leading to a pyogranulomatous infiltrate. Toxicity associated with accelerated hepatocyte can evolve hepatocyte loss in the absence of increased enzyme activities or hyperbilirubinemia.

Anesthetic Agents and Hepatotoxicity in Small Animals

While uncommon, both halothane and methoxyflurane can provoke hepatotoxicity. The pathomechanism was categorized as an immunosensitization reaction in humans. Hepatotoxicity in dogs is rare and was anecdotally observed when these anesthetics were more commonly used. Prolonged or repeated exposure in susceptible patients was noted to progress to a necrotizing hepatopathy.

Propofol infusion can impair mitochondrial oxidation of long-chain fatty acids and mitochondrial oxidative phosphorylation, and it is noted to provoke lactic acidosis in some humans (infants and adults with propofol infusions of ≥ 48 hours). Propofol inhibits coenzyme Q at complex II and cytochrome C at complex IV of the electron transport chain, leading to failure of ATP production and oxidative injury. Affected patients develop hepatomegaly as well as muscle and cardiac effects; hepatic changes reflect passive congestion and diffuse hepatocyte microvesicular lipid vacuolation.

Adverse response to propofol may occur in cats with hepatic lipidosis (HL), a syndrome already complicated by imbalanced fatty acid oxidation and hepatic triglyceride exportation, GSH redox imbalance, and lactic acidosis. In some HL cats, propofol administration causes protracted sedation and Heinz body hemolysis, leading to anemia requiring blood transfusion. Onset of hemolysis and anemia is delayed for 6–12 hours after drug administration. This adverse reaction worsens jaundice and clinical status of affected cats.

Repeated administration of propofol to patients without antecedent liver disease (humans and dogs) is also known to provoke Heinz body hemolysis.

Xylitol and Hepatotoxicity in Small Animals

Xylitol, a commonly used artificial sweetener in human foods, may be an intrinsic hepatotoxin for dogs, with ingestion of small doses leading to intractable hypoglycemia and lethal hepatic failure. Toxicity may lead to death before liver enzyme activity increases. However, there is some evidence suggesting a breadth of individual responses to this toxin.

Xylitol is a pentahydroxy sugar alcohol with a sweetness index similar to sucrose. It is extraordinarily toxic to dogs with as little as 0.1 g/kg causing clinical hypoglycemia and hepatic necrosis in some dogs. While xylitol appears to be an intrinsic hepatotoxin for dogs, there is a breadth of individual tolerance. Some dogs consuming xylitol at toxic levels develop only a few days of self-resolving noncritical illness that on subsequent liver biopsies (collected during surgery for unrelated illnesses) reveal diffuse hepatocyte cytosolic lipofuscin accumulation, indicating diffuse oxidative injury.

A 2018 search of the American Society for the Prevention of Cruelty to Animals (ASPCA) Animal Poison Control Center’s (APCC) product database defined > 1,900 products containing xylitol. These include:

  • several vitamins (including chews, multivitamin tablets, and gummy vitamins)

  • nutritional supplements (eg, coenzyme Q10, 5-hydroxytryptophan, caffeine, omega-3 fish oils, melatonin)

  • numerous human foods (eg, chocolate, baked goods, puddings, syrup, fruit preserves, jellies, peanut butter, ice cream, nutritional/diet bars, drink powders, peanut butter)

  • various dental and other hygiene products to inhibit bacterial growth (eg, toothpaste, tooth wipes and towelettes for babies, oral lozenges, moisturizing mouth sprays and gels, mouthwash, exfoliating facial wipes, personal lubricants, deodorants, sunscreens, and night creams)

  • various medicinal products (eg, nicotine gums, oral drug suspensions, cold remedies, sublingual tablets, nasal sprays including aripiprazole disintegrating tablets, Bach flower remedies, and oral gabapentin solution)

Xylitol may be listed on product labels using a variety of synonyms, including wood sugar, birch sugar, and birch bark extract.

Pharmacokinetic studies in dogs demonstrate peak plasma concentrations within 30 minutes with metabolism primarily by the liver (70%–80%). In hepatocytes, xylitol is oxidized to D-xylulose and then subsequently phosphorylated to an intermediate in the pentose phosphate pathway. Further conversion to glyceraldehyde-6-phosphate or fructose-6-phosphate ultimately yields glucose, glycogen, or lactate.

In dogs, xylitol initiates a profound dose-related insulinotropic effect (2.5–7.0-fold greater insulin release compared to an equivalent amount of glucose). One study demonstrated it mediated a direct stimulation on pancreatic beta cells. This insulinotropic effect provokes severe clinical hypoglycemia.

Intoxicated dogs may develop vomiting, tremors, lethargy, weakness, collapse, and abdominal tenderness.

Initial clinicopathologic assessment of clinically affected dogs demonstrates acute extreme increases in serum ALT activity (> 10-fold upper reference limits), marked hypoglycemia, hypokalemia, and hypophosphatemia. Initially marked increases in ALT activity progressively decline over 3–4 days while ALP activity may increase modestly. Coagulopathy may manifest within 2 days of severe hepatotoxicosis, along with gradual onset of modest hyperbilirubinemia (reflecting heme release from necrotic hepatocytes and perhaps hemolyzed RBC secondary to systemic GSH depletion and hypophosphatemia).

Histopathologic changes associated with xylitol toxicosis in dogs includes centrilobular and midzonal hepatocyte necrosis, causing parenchymal collapse and a periportal hepatocyte ballooning degeneration. Hepatocytes display cytosolic lipofuscin accumulation reflecting oxidative injury.

Initial treatment for witnessed xylitol ingestion includes induction of emesis if the dog has not already commenced vomiting and if emesis induction is considered a safe procedure (dog not recumbent or collapsed). Emesis is induced by apomorphine given either by injection (0.03 mg/kg, IV, or 0.04 mg/kg, IM) or subconjunctival administration (crushed quarter of a 6 mg tablet, followed by copious flushing after emesis commences). Excessive sedation can be reversed with naloxone.

Alternatively, emesis may be initiated by oral administration of 3% active hydrogen peroxide (1–2.2 mL/kg, not to exceed 45 mL total dose); dose administration can be repeated one time (otherwise gastric irritation/ulceration may follow). Vomited material must be inspected for the xylitol-bearing product to assess status of decontamination. Ingestion of a 100% xylitol product presenting after 30 minutes may be difficult to decontaminate because of rapid xylitol absorption. Weakness associated with profound hypoglycemia puts a patient at risk for aspiration pneumonia. Activated charcoal is not recommended because of the rapidity of xylitol absorption and because xylitol binds poorly to this decontaminant.

Successful recovery can be achieved with supportive fluid therapy with judicious dextrose supplementation to achieve euglycemia and with correction of potassium and phosphate concentrations. Fluids should be supplemented with water-soluble vitamins. Intravenous N-acetylcysteine is given every 8 hours as a constant-rate infusion (CRI) over 20 minutes and oral bioavailable SAMe (20–40 mg/kg, PO on an empty stomach) initiated when oral medications can be tolerated. Parenteral vitamin K1 (0.5–1 mg/kg, SC or IM, every 12 hours, 3 doses) should be provided to all clinically affected dogs with transfusion of fresh frozen plasma reserved for those with documented coagulopathy.

Maropitant (1 mg/kg, PO or SC, every 24 hours) or ondansetron is used for control of nausea and vomiting. Intravenous antimicrobials are prescribed preemptively to protect against systemic dissemination of bacteria from enteric translocation during the acute phase of liver injury. During this time, Kupffer cell function may be compromised by altered sinusoidal perfusion (parenchymal collapse) and overwhelmed by phagocytosis of cellular debris of necrotic hepatocytes.

Dogs should be hospitalized for a minimum of 12–24 hours after known ingestion of xylitol due to risk of delayed-onset hypoglycemia, particularly with chewing gum exposure. Prognosis is generally good with early decontamination and effective management of hypoglycemia. Dogs with acute increases in ALT should be hospitalized, maintained on IV fluids, and supported, then sequentially monitored for developing coagulopathy until ALT activity demonstrates decline.

Potential for xylitol-associated hepatotoxicity in cats was studied in six animals given escalated xylitol doses of 100, 500, and 1000 mg/kg in separate trials. There was no hypoglycemia or hepatotoxicity demonstrated.

Acetaminophen and Hepatotoxicity in Small Animals

Acetaminophen (N-acetyl-p-aminophenol, or APAP) can lead to acute fulminant hepatic failure (FHF) or hematotoxicity (chemically induced toxic effects in the blood) in dogs or cats. In clinical cases of APAP toxicosis, hematotoxicity is most common.

After APAP exposure, the parent drug is promptly detoxified by glucuronide or sulfate conjugation yielding water-soluble products that are renally excreted. However, depending on dose exposure, a portion of APAP is also rapidly oxidized by hepatic cytochrome P450s to an oxidative adduct, N-acetyl-p-benzoquinoneimine (NAPQI). With only small dose exposure, NAPQI is detoxified by GSH conjugation, forming water soluble cysteine or mercapturic acid conjugates.

With toxic APAP exposure, saturation of glucuronosyl and sulfate transferases (conjugating enzymes) permits accumulation of the oxidative NAPQI adduct, exceeding GSH reserves. Upon accumulation, NAPQ1 forms covalent bonds with protein macromolecules (sulfhydryl group bonds in cell membranes and enzyme groups), leading to hepatocyte cell, organelle (mitochondrial), and nuclear membrane oxidation with subsequent cell necrosis. Death from hematotoxicity is attributed to accumulation of para-aminophenol (PAP), normally a minor metabolite of APAP. This may transpire without evidence of hepatotoxicity after high-dose single APAP exposure.

In cats, a onetime oral dose of 95–150 mg/kg can provoke PAP toxicity, characterized by rapid onset (4–6 hours) of tachypnea (respiratory distress), cyanosis, and methemoglobinemia. This is followed by anorexia and lethargy over 24–72 hours, with potential lethal consequences. These patients can be rescued with IV N-acetylcysteine (NAC), administration of packed RBC, vitamin C administration, and administration of oral bioavailable SAMe.

In dogs, only sustained plasma APAP concentrations (eg, ≥ 140 mcg/mL for 20 hours) provoke hepatotoxicity. Experimental studies document canine hepatotoxicity with a single 500 mg/kg dose administered as a CRI over 90 minutes or with repeated SC dosing (ie, loading dose 750 mg/kg, SC, followed at 9 and 24 hours with 200 mg/kg, SC). Otherwise, single-dose exposure of up to 1,600 mg/kg causes lethal hematotoxicity, with microvascular, cardiac, and renal injury. Clinical occurrence of hepatotoxicosis in dogs is only encountered subsequent to repeated high-dose drug exposure or after ingestion of numerous tablets slowly absorbed because of delayed GI transit.

In dogs, hepatotoxicity is characterized by illness within 24 hours of drug exposure, including anorexia, vomiting, and diarrhea. Progressive illness (36–48 hours) follows with onset of hyperammonemia and neurologic signs progressing to coma and death within 72 hours. Extreme increases in ALT activity (> 500× baseline) develop within 48 hours of hepatotoxic drug exposure, along with modest increases in total bilirubin; substantial decreases in cholesterol, albumin, and BUN concentrations; marked hyperammonemia; and coagulopathy (failure to synthesize clotting factors). Clinical and clinicopathologic findings parallel onset of FLF, confirmed by histologic evaluation.

In dogs surviving hepatotoxicosis, lesser magnitudes of ALT activity (~15-fold above baseline) are observed. These animals also become hyperammonemic and display clinicopathologic findings consistent with synthetic failure of generally lesser magnitudes. Histologic features of hepatotoxicity are characterized by a centrilobular tropism, reflecting the dominant cytochrome P450 oxidase activity in this region (ie, NABQ1 accumulation). Injury causes centrilobular to panlobular hepatocyte necrosis with severe regional congestion in areas of parenchymal collapse.

Pre-exposure to P450 inducers such as phenobarbital or sodium thiopental, an analogue of phenobarbital, provokes severe APAP hematotoxicity due to accelerated formation of PAP (ie, deacetylation of APAP in hepatic microsomal enzymes). Hastened drug metabolism depletes GSH reserves, amplifying risk for oxidative injury. Indeed, the apparent high susceptibility of dogs to APAP toxicosis is thought to reflect their comparatively lower liver GSH concentrations relative to other species (rodents, humans, cats). Higher feline proclivity to lethal hematotoxicity reflects their limited glucuronidation capabilities, easily depleted sulfuration conjugation, and subsequent GSH depletion.

Dogs and cats are the only known species in which APAP toxicosis provokes hematotoxicity due to PAP accumulation, characterized by methemoglobinemia (apparent cyanosis, brown-colored blood and urine), Heinz body-associated hemolysis, and anemia. Methemoglobinemia and anemia critically impair oxygen transport, damage microvasculature, and provoke acute-onset facial swelling.

In species other than the dog and cat, PAP is removed by GSH conjugation and N-acetylation (orchestrated by two enzymes, N-acetyltrasnferase-1 [NAT-1] and N-acetyltransferase 2 [NAT-2]). Cats only have NAT-1 and dogs have no NAT enzymes. Thus, with toxic APAP exposure, PAP accumulates, causing hematotoxicity. There also is evidence for relatively poor methemoglobin reductase activity in feline and canine erythrocytes, predisposing these species to methemoglobinemia.

Antifungals and Hepatotoxicity in Small Animals

Ketoconazole hepatotoxicity in dogs may evolve a slow chronic subclinical injury, acute fulminant hepatic failure, persistent chronic hepatitis, or a chronic pyogranulomatous hepatitis. In many dogs, hepatocyte injury abates with drug discontinuation, leaving behind aggregates of dense lipofuscin-laden macrophages in central and portal regions representing scavenged remains of necrotic hepatocytes. Ketoconazole toxicosis is thought to reflect accumulation of a reactive drug adduct (N-deacetyl-ketoconazole [DAK]) that imparts oxidative injury and provokes hepatocyte ALT release in a dose- and concentration-related response. Cellular injury affiliates with covalent binding of ketoconazole metabolites with macromolecules, leading to hepatocyte GSH depletion.

Experimentally, DAK accumulation depletes hepatocyte and mitochondrial ATP, imposes oxidative membrane injury, provokes mitochondrial-initiated hepatocyte, and damages mitochondrial DNA. The latter injury is a serious pathomechanism of hepatotoxicosis, as mitochondria are largely devoid of nucleoprotein repair mechanisms. Itraconazole has also rarely been associated with hepatotoxicity with residual lipofuscin pigment observed in historically damaged regions. Hepatotoxicity resolves with drug discontinuation.

Dogs developing a pyogranulomatous injury pattern can have the injury progressively worsen even with drug withdrawal. A few dogs with ketoconazole-induced chronic hepatitis have responded to immunomodulation after drug withdrawal, along with administration of bioavailable SAMe, low-dose vitamin E, polyenylphosphatidylcholine, and ursodiol (ursodeoxycholic acid).

Because ketoconazole potently inhibits cytochrome P450s, coadministration with other drugs metabolized by these enzymes can lead to retention of toxic parent or metabolic adducts that may be injurious. One such injurious combination is ketoconazole with colchicine that lead to inappetence, vomiting, bloody diarrhea, muscle pain, and weakness associated with jaundice. Increased ALT, AST, ALP, and creatine kinase activities were documented. Liver biopsy demonstrated canalicular cholestasis, hepatocyte ring mitoses, and metaphase arrest, reflecting colchicines' impact on spindle fibers, rare necrotic or apoptotic hepatocytes, and diffuse lipofuscin pigment accumulation (likely reflecting ketoconazole-mediated oxidative stress).

Diagnostic findings were consistent with a transient necrotizing myopathy, acute hepatotoxicity, and gastrointestinal toxic effects with transient cortisol suppression by ketoconazole. Drug discontinuation with administration of ursodiol, bioavailable SAMe, and low-dose prednisolone resulted in gradual recovery of over 2 weeks. Cholestyramine may assist with removing ketoconazole undergoing enterohepatic circulation and should be considered in dogs with acute severe toxicity.

Terbinafine hepatotoxicity has rarely been characterized histologically in dogs. Marked acute increase in ALT activity (> 3,000 U/L) after 2 weeks of treatment was associated with moderate centrilobular hepatocyte necrosis, centrilobular parenchymal collapse, and diffuse cytosolic lipofuscin accumulation in hepatocytes (reflecting oxidative membrane injury). A mild lymphoplasmacytic central and lobular hepatitis were observed in biopsies collected shortly after onset of suspected hepatotoxicity. Clinical signs resolved with drug discontinuation.

In cats, griseofulvin can induce inappetence, lethargy, hyperbilirubinemia, and increased ALT activity. Apparent hepatotoxicity may reflect inhibition of ferrochelatase activity, as griseofulvin is a well-characterized suicidal substrate for this enzyme. Griseofulvin is also acknowledged to model acquired erythrocytic protoporphyria and to unveil previously unrecognized porphyric syndrome. Hepatotoxic effects typically resolve with drug discontinuation. Histologic lesions are characterized by prominent bile casts that may represent insoluble protoporphyrin pigment precipitation.

Benzodiazepines and Hepatotoxicity in Small Animals

Hepatotoxicity has been documented with both oral and parenterally administered benzodiazepines. Idiosyncratic diazepam hepatotoxicity in cats causes fulminant hepatic failure, characterized by panlobular necrosis with retention only of progenitors differentiating as cholangiocytes (ductular reaction), Similar toxic effects have been observed with oxazepam and alprazolam.

Signs of benzodiazepine-associated hepatotoxicity are peracute in onset after only a few treatments. Cats with diazepam hepatotoxicity were most often medicated for behavior modification or as treatment of feline lower urinary tract disease.

Unfortunately, idiosyncratic benzodiazepine hepatotoxicity is usually fatal. Proactive monitoring of liver enzymes can identify adverse reactions early in their course, allowing prompt drug discontinuation and aggressive supportive care (including blood component therapy, fluid and electrolyte stabilization, and antibiotic protection against translocated bacteria). Kupffer cell dysfunction is suspected because of the severity of parenchymal collapse, deviated sinusoidal perfusion, and extreme necrosis with accumulation of sinusoidal cellular debris.

Other Drug-Associated Hepatotoxicity in Small Animals

In dogs, trimethoprim-sulfadiazine can cause idiosyncratic hepatotoxicity that may involve an immune-mediated pathogenesis. Furthermore, dogs as a species are unable to rapidly detoxify sulfa drug metabolites, which likely makes more dogs susceptible to adverse reactions. Reversible cholestatic hepatopathy and acute/subacute massive fatal hepatic necrosis have been documented. In some cases, only a few treatments with appropriate conventional dosing has initiated toxicity.

Danazol, an impeded androgen, can be hepatotoxic and cause idiosyncratic reversible jaundice in dogs. Androgenic anabolics can induce hepatic lipidosis in anorectic cats or in cats fed a protein-restricted diet. Androgenic anabolics also increase risk of hepatocellular carcinoma.

Tetracyclines including doxycycline rarely lead to idiosyncratic hepatic necrosis in dogs and cats. However, any tetracycline may augment risk for hepatocellular lipid accumulation.

Methimazole treatment in cats for control of hyperthyroidism can cause idiosyncratic or immune-mediated hepatotoxicity associated with inappetence, jaundice, and increased ALT and AST activities. Hepatic lesions are characterized by hepatocyte degeneration and necrosis. Injury resolves after drug discontinuation.

Mitotane hepatotoxicity is uncommon and usually encountered within the first 1–4 weeks of treatment.

Clinical signs include prodigious increases in ALP activity with marked increases in ALT and GGT activity and normal or variably increased total bilirubin.

Recovery typically follows drug withdrawal.

Environmental or Foodborne Hepatotoxins

Other specific hepatotoxins documented in dogs include foodborne aflatoxins, toxins derived from Amanita mushrooms (amanitin); blue-green algae (microcystin); cycad-associated (sago palm) cycasin and beta-methylamino L-alanine, a neurotoxic amino acid; jerky treats or rawhides manufactured in countries with low manufacturing standards; essential oil treatment; and even air freshener pulsing aerosols containing essential oils.

Aflatoxicosis Hepatotoxicity in Small Animals

Aflatoxins are potent hepatotoxic and hepatocarcinogenic agents produced primarily by strains of the fungus Aspergillus. Involved fungi are widely distributed in nature, making aflatoxicosis an ever-present danger. Food contamination with aflatoxins can occur in the field or after harvest during storage or processing. Although many different aflatoxins exist, aflatoxin B1 (AFB1) is the most common and potent hepatotoxin and remains inert until bioactivated by host hepatocytes.

Hepatotoxicosis is dose related and caused predominantly by cytochrome P450-catalyzed formation of electrophilic and arylating metabolites. The most noxious product of AFB1 (8,9-epoxide) is formed within minutes in hepatocytes, overwhelming GSH detoxification. The ensuing GSH depletion contributes to liver injury, increasing vulnerability to oxidative injury.

Depending on the amount of AFB1 ingested, aflatoxicosis can present as a peracute syndrome or as chronic liver injury. Accumulated AFB1 8,9-epoxide causes hepatic injury via irreversible covalent binding to cell and organelle proteins (structural, enzymatic, nucleoproteins).

Binding to intracellular and mitochondrial macromolecules, DNA and RNA, impairs gene transcription (ie, compromises RNA polymerase), DNA replication and repair, hepatic protein synthesis, and cell energy production. Mitochondrial toxicity impairs fatty acid oxidation and the electron transport chain and initiates direct and indirect apoptotic pathways. Hepatocarcinogenesis, confirmed in several species, is not recognized in dogs to date. In humans, hepatocellular carcinoma is linked to a specific mutation in the p53 tumor-suppressor gene.

Different species have varying susceptibility to aflatoxin hepatotoxicity, with dogs being highly susceptible. Canine vulnerability is thought to reflect the inherently lower hepatocyte GSH concentrations in dogs compared to other species, lower activity of GSH S-transferase, and interindividual pharmacogenetic differences in cytochrome P450 AFB1 bioactivation to the AFB1 8,9 epoxide. Sporadic reporting of individual and regionally widespread aflatoxicosis in dogs has occurred since the 1970s. Since 2005, four foodborne outbreaks have been documented in dogs (US, Israel [likely the same source as in US], Brazil, and South Africa). Outbreaks were due to corn contamination.

Aflatoxicosis is usually an insidious process because the toxin has no taste or smell distinguishing its presence. That said, some dogs literally refused to consume contaminated food during the 2005 US outbreak, presumably because of a taste or smell related to growth of mold, rather than the toxin. Some dogs die acutely from aflatoxicosis with no recognized antecedent illness. These have aflatoxicosis suspected on the basis of necropsy findings (hepatic injury pattern). In dog kennels, these sentinel deaths have not averted aflatoxicosis of other dogs because of the practice of feeding a common food source.

Clinical Signs

Wide interindividual differences in clinical signs are observed among affected dogs. This diversity may reflect differences in pharmacogenetics (cytochrome P450 activity, GSH transferase activity) affecting toxin biotransformation or detoxification, the amount and duration of toxin exposure, nutritional status of the patient (poor nutritional support can decreased hepatic GSH concentrations), interactions with other toxic insults, antecedent disease processes or drugs that might induce P450 activity, patient age (younger animals have greater risk), and nutritional and hormonal status.

With these considerations in mind, clinical signs of canine aflatoxicosis typically include anorexia, lethargy, vomiting and diarrhea, occasional polyuria and polydipsia, and gradual onset jaundice. Initial illness with slow chronic hepatotoxicosis is eventually followed by hematochezia or melena as well as development of an abdominal effusion (modified transudate, rarely hemorrhagic).

Peripheral edema and abdominal effusion may emerge after IV fluid administration. Severe hepatic synthetic failure and hepatic parenchymal injury (parenchymal collapse) lead to loss of clotting factors, consumptive coagulopathy, and terminal signs of hepatic encephalopathy associated with severe enteric hemorrhage.

In addition to hepatic synthetic failure, gastrointestinal bleeding may be a coumarin-related effect of aflatoxin as well as portal hypertension, provoking diapedesis of RBCs from capillaries into the enteric lumen. Acquired portosystemic shunts develop in dogs surviving > 4 weeks after onset of clinical signs. In some dogs with severe illness, enteral nutrition is thwarted by severe gastric and enteric atony that fails to respond to prokinetic agents.

Not all dogs die of aflatoxicosis despite demonstration of clinical illness. Resolution of abdominal effusion manifesting 60 days after illness onset was followed by resolution and apparent full recovery in some dogs given supportive care. Preexistent necroinflammatory liver disease may increase vulnerability of some dogs. Conversely, inappetence due to unrelated causes (ie, hypoadrenocorticism associated with anorexia, severe diffuse vacuolar hepatopathy that may have slowed toxin delivery to hepatocytes or altered hepatocyte metabolism, and intestinal malabsorption that deterred toxin absorption) has protected some dogs from fatal toxicity.

Experimentally, pretreatment with phenobarbital increases metabolism of AFB1 to less toxic metabolites with biliary elimination compared to formation of the AFB1 8,9 epoxide adduct. There is no benefit from phenobarbital after toxin exposure.

Clinicopathologic Features

There are no distinguishing hematologic features; thrombocytopenia often coordinates with development of coagulopathy. Electrolyte status reflects dehydration, losses incurred from vomiting and diarrhea, and third-space fluid sequestration associated with fluid therapy, hypoalbuminemia, and loss of microvascular integrity (suspected to be toxin related).

While hepatotoxicosis is often heralded by marked increases in ALT and AST and lesser concentrations of ALP and GGT, aflatoxicosis is associated with wide variability in liver enzyme activities. This likely reflects the mode of hepatocyte death associated with aflatoxicosis orchestrated by mitochondrial-initiated apoptotic pathways (programed cell death) rather than necroinflammatory, cytotoxic, or coagulative necrosis. Further relevant to aflatoxicosis is that toxic effects impairing hepatocyte synthetic capabilities also blunt enzyme activity.

Biochemical features reflecting impaired hepatic synthesis in dogs with aflatoxicosis include:

  • development of hypoalbuminemia, hyperbilirubinemia, and hypocholesterolemia

  • prolonged prothrombin time (PT) and activated partial thromboplastin time (aPTT) clotting times (factor insufficiency)

  • low antithrombin and protein C activity

  • hypofibrinogenemia

Coagulopathy likely reflects impaired protein synthesis as well as development of a consumptive coagulopathy with a possibility of released thromboplastic substances from damaged hepatocytes, decreased clearance of activated proteases, and a potential contribution from a suspected coumarin-like effect of aflatoxin. In some dogs the kidneys manifest injury by the presence of granular casts (only observed in animals that die).

While there are clear differences in median values of clinicopathologic parameters between dogs that survive and dogs that do not in large retrospective studies, these observations lack utility for single-patient prognosis. Rather, it is important to sequentially evaluate relative changes in parameters over time, usually over multiple weeks (4–16 weeks) of survival. Positive indicators included increase in cholesterol and albumin concentrations, decline of hyperbilirubinemia, and increased activity of protein C and antithrombin.

Variability in clinicopathologic parameters displayed among dogs with confirmed aflatoxicosis (2005 outbreak in North America), irrespective of survival outcome, included increased ALT (83%), AST (79%), ALP (44%), and GGT (44%) activities; hypocholesterolemia (72%); and subnormal protein C (96%) and antithrombin (96%) on initial assessments. Abnormal protein C, antithrombin, and cholesterol preceded development of clinical signs or were detected in dogs presenting only with hyporexia. On sequential testing, abnormalities in protein C, antithrombin, and cholesterol persisted beyond several weeks whereas serum liver enzyme activities remained nonspecific mildly increased or were within reference range in some dogs.

Markedly increased serum liver enzyme activities and overt hyperbilirubinemia (total bilirubin > 2.5 mg/dL, consistent with jaundice) were inconsistent laboratory features of early intoxication. This corroborates with experimental findings in dogs with chronic aflatoxicosis in which hyperbilirubinemia usually developed after ≥ 1 month of toxin exposure.

Abdominal Ultrasound Imaging

Ultrasound imaging discloses abnormal hepatic architecture in chronically affected dogs and dogs developing portal hypertension. Hyperechoic parenchyma reconciles with diffuse microvesicular lipid vacuolation of aflatoxicosis with hypoechoic hepatic nodules observed in dogs with regenerative nodules. Anechoic abdominal effusion, a thick edematous gallbladder wall, and large lymph nodes were observed in dogs with portal hypertension. Occasionally, APSSs were observed (color-flow Doppler ultrasonography) after 4–6 weeks of illness.

Cytologic Features

Cytologic evaluation of liver aspirates from dogs with aflatoxicosis discloses diffuse microvesicular hepatocyte lipid vacuolation, rare degenerative hepatocytes, and sparse mononuclear infiltrates (lymphocytes, macrophages).


Definitive confirmation of aflatoxicosis is based on histologic changes in liver (see below) and confirmation of toxic levels of aflatoxin in consumed food, circumstantially linked with patient illness. Several samples of food from different areas of a bag of dry food should be analyzed because the toxin can be unevenly distributed.

Attempting to measure aflatoxin B1 in liver is usually unsuccessful due to its exceptionally short half-life (minutes), AFB1 being instantaneously converted to the AFB1-8,9, epoxide and other metabolites. As lapses of several days between tissue sampling and ingestion of contaminated foods are common in clinical patients, the determination of AFB1 or its metabolites is often unrewarding. That said, collection of urine or blood samples in animals still consuming the suspected aflatoxin-contaminated food may reveal presence of the M1 metabolite, M1 having a half-life < 12 hours.

One study analyzing the M1 metabolite in liver from dogs dying of aflatoxicosis confirmed its presence in 7 of 8 samples. However, this complex analytic procedure is not routinely used as a diagnostic test.

Different species vary greatly in their susceptibility to AFB1, with dogs being highly susceptible. While a zero tolerance for aflatoxin is suggested by the Food and Drug Administration for humans, a legal limit of 20 mg/kg (ppb) food is imposed for human and dog foods. For dog food, toxic concentrations > 60 mg/kg (ppb) feed and lethal concentrations of 500–1,000 mg/kg (ppb) food are recognized.

In outbreaks of canine foodborne aflatoxicosis since 2005, the causal source was corn distributed in manufactured foods and in one instance was home-cooked polenta added to a homemade ration. While the thermal stability of aflatoxin B1 is disrupted at 160°C, polenta is made in boiling water (~100°C).

Among these outbreaks, aflatoxin food contamination ranged from 48 to 800 ppb food (2005, US, commercial food) up to 4,946 ppb food (2011, South Africa, commercial food) and 1,640 to 1,770 ppb (2011, southern Brazil, polenta scraps). Based on experimental work, dogs exposed chronically to 0.05–0.3 mg AFB1/kg of feed over 6–8 weeks demonstrated chronic illness, eventually dying of liver failure and hemorrhagic diathesis with a cirrhotic liver.

The median lethal dose (LD50) of AFB1 for dogs is 0.5–1.5 mg AFB1/kg with death occurring within 3 days; the LD50 of AFB1 for cats is reported as 0.3–0.6 mg/kg. It is rare to encounter aflatoxicosis in pet cats, likely because of their carnivore dietary formulations.

Microscopic Hepatic Features

Histopathologic features consistent with aflatoxicosis are not pathognomonic for this cause of hepatic injury; rather, they reflect a pattern observed in hepatotoxicosis targeting nucleoprotein integrity, gene transcriptional responses, mitochondrial function, and oxidative damage. The magnitude of liver enzyme activity does not reconcile with the severity of hepatocellular injury observed in clinical patients with aflatoxicosis. This phenomenon was also described in dogs sequentially monitored with chronic experimental aflatoxicosis.

A consistent histologic feature is microvesicular hepatocyte lipid vacuolation, initially with a centrilobular tropism that becomes panlobular. Additional features include canalicular bile plugs reflecting hepatocyte canalicular membrane injury, areas of parenchymal collapse (hepatocyte degeneration, initially centrilobular), with moderate to marked Kupffer cell and macrophages accumulation of lipofuscin (pigment reflecting oxidized membrane debris) and iron (released from the hemoprotein cytochrome P450 with hepatocyte death).

Depending on chronicity, variable ductular reaction (cholangiocyte proliferation) and bile duct hyperplasia reflect regeneration of viable progenitor cells and interference with bile flow imposed by tissue injury/remodeling. Variable mixed lymphocyte and macrophage infiltrates in central and portal regions reflect nonspecific reactions to cell injury and response to cytokine elaboration by activated macrophages. Aflatoxin B1 has been shown experimentally to provoke macrophage activation, thereby increasing inflammatory cytokine expression. In the later stage of injury, nodular hyperplasia and dissecting fibrosis are evident.

Necrotic hepatocytes are rarely observed. Rather, apoptotic hepatocytes with condensed cytoplasm with nonfragmented pyknotic nuclei without adjacent inflammatory infiltrates are observed. Hepatocyte death in aflatoxicosis coordinates with downregulation of antioxidant enzymes (catalase and GSH peroxidase) that are typically upregulated in a necroinflammatory process or oxidative cause of cell death, and upregulation of genes involved with death receptor-initiated apoptosis (associated with mitochondrial injury).


Despite intensive supportive care including judicious administration of fluids, potassium, water-soluble vitamins vitamin K, blood component therapy, N-acetylcysteine (NAC) constant-rate infusion every 8 hours, oral SAMe in dogs that could tolerate oral medications, oral vitamin E supplements, control of gastric acidity and esophageal reflux, and antimicrobial coverage, mortality rate among dogs with inappetence, vomiting, and evidence of synthetic failure was approximately 65%.

Lower mortality rate was observed in dogs without or with only minor clinical illness, despite evidence of liver injury and modest hemostatic abnormalities. When confronted with an outbreak of suspected aflatoxicosis, determination of cholesterol concentrations and measurement of protein C and antithrombin activities, looking for subnormal values, can be used to incriminate exposure. These parameters appear to function as biomarkers of aflatoxin-suppressed synthetic function.

Liver enzymes and total bilirubin concentrations cannot be relied upon to accurately identify these dogs. Dogs with suspected exposure should be treated with removal of the suspected toxin-laden food, administration of hepatoprotectants (IV NAC recommended early) combined with oral bioavailable SAMe and vitamin E, and parenteral vitamin K administration every few days. Blood component therapy is reserved for animals with spontaneous bleeding.

Sequential surveillance of cholesterol, albumin, protein C, antithrombin, and if jaundiced, total bilirubin allows prediction of recovery over several weeks. These animals are given hospitalized care for only a few days until it is clear they are not destabilizing. No treatment can accelerate recovery. Recovery occurs if there are enough hepatocytes with replicatory and synthetic capabilities.

Amanita phalloides (Death Cap Mushroom) Hepatotoxicity in Small Animals

Amanita phalloides (death cap mushroom) is responsible for the majority of fatal hepatotoxicity induced by mushrooms in humans and animals. However, there are numerous amatoxin-containing mushroom species within the genera of Amanita, Galerina, and Lepiota.

The principal toxins are alpha- and beta-amanitin, with alpha-amanitin proposed as more potent.

These toxins have great hepatotoxic threat because they have high stability against heat, freezing, dehydration, light exposure, and longterm storage, and they are water soluble, acid resistant, and not biotransformed by endogenous enteric or bacterial enzymes. Amatoxins are readily absorbed from the gastrointestinal tract, are not bound to albumin, and thus are rapidly distributed to liver and kidneys within 48 hours of ingestion.

The liver is the primary target of amatoxin toxicity and the first organ in contact with ingested toxin. Amatoxin accumulation in hepatocytes is facilitated by organic anion transporting proteins (OATP). Although amatoxins do not undergo biometabolism, up to 60% are eliminated in bile and may undergo enterohepatic circulation, augmenting hepatotoxicity. This effect has been documented in dogs when extracorporeal biliary drainage survives lethal amatoxin ingestion. Most toxin is eliminated in urine during the first 4 days after ingestion; high renal exposure can result in nephrotoxicity.

The primary mechanism of amanitin hepatotoxicity is covalent binding of the toxin with and inhibition of RNA polymerase II in the cell nucleus. Inhibition of messenger RNA (mRNA) synthesis arrests protein synthesis and eventually leads to cell death. Cells with high rates of protein synthesis are most vulnerable; this coordinates with the initial gastrointestinal signs and later development of hepatic and renal injuries. The process of hepatocyte injury involves apoptosis pathways but is synergistic with tumor necrosis factor (TNF)-alpha–provoked oxidative injury and cell necrosis.

Clinical Signs

Clinical evidence of A phalloides syndrome is not obvious until extensive liver or renal injury has occurred; this ranges from hours to several days after exposure. Four phases of amanitin toxicosis are characterized:

  • The initial latent phase lasts for up to 12 hours.

  • The gastrointestinal phase develops ~6–24 hours after toxin ingestion, persisting up to 36 hours. Clinical signs include abrupt onset of nausea, vomiting and diarrhea (often bloody), abdominal pain, and hematuria. Dehydration, electrolyte disequilibrium, and hypoglycemia may develop. Immediate gastrointestinal signs may reflect phallotoxins also present in toxigenic mushrooms.

  • The hypoglycemic phase reflects glycogen utilization in the face of hepatic synthetic failure (impaired gluconeogenesis) and can lead to death if euglycemia is not achieved with IV fluid supplementation.

  • The final hepatorenal phase reflects severe progressive hepatic necrosis and is characterized by development of jaundice and clinicopathologic markers of escalating liver failure, worsening coagulopathy, and emergence of hepatic encephalopathy and renal failure.

Notably, dogs ingesting large amounts of amanitin may die without developing classic phased progression.

Presumptive diagnosis of amanitin hepatotoxicity is usually based on circumstantial association of observed mushroom ingestion, clinical signs, biochemical features, and histologic liver lesions. However, definitive detection of alpha-amanitin in serum, urine, gastric contents, suspect mushrooms, liver, or kidneys can be done by select veterinary toxicology laboratories. The Meixner test for rapid amatoxin detection does not provide definitive identification and can be misleading.

Serum, urine, and tissue sections from animals dying from suspected amanitin toxicosis should be frozen if subsequent definitive testing is desired. Serum and urine samples should be saved as early in the illness prodrome as possible because with time the circulating toxin diminishes. In dogs an LD50 of 0.11 mg/kg alpha-amanitin had been determined.

Clinicopathologic Features

Experimental alpha-amanitin hepatotoxicity in dogs given 0.09 mg/kg, PO, once (study done to evaluate utility of IV silibinin over 30 years ago), provides details regarding canine amanitin toxicity. No clinical signs were manifest for 16 hours. Thereafter, vomiting and bloody diarrhea commenced, peaking at ~48 hours.

Survivors improved rapidly after ~60 hours. Extreme increases in ALT (~180× baseline), AST (50× baseline), and lesser increases in ALP (6× baseline), GGT (2× baseline), and bilirubin (6× baseline) developed by 48 hours. Within this time frame, marked prolongation of the prothrombin clotting time was also documented. Dogs that died of amanitin toxicosis became azotemic (increased BUN and creatinine concentrations) and developed enteric bleeding and terminal hepatic encephalopathy between 35 and 54 hours.

Hyperammonemia reflects decreased hepatic ammonia detoxification due to FHF and may be provoked by renal azotemia and enteric hemorrhage. While the magnitude of documented transaminase activity is extreme, it is possible that these are somewhat limited by severe hepatic synthetic failure. With onset of synthetic failure, hypoglycemia, hypoalbuminemia, and hypocholesterolemia simultaneously emerge.

Dogs surviving experimental amatoxin toxicosis had normalization of clinicopathologic abnormalities by 192 hours. Despite attempts to normalize alpha-amanitin dosing among studied dogs, 4 of 12 died. This may reflect variable distribution of toxin within the mushroom extract or individual variation in toxin tolerance.

Bleeding diathesis is an important complication of amanitin hepatotoxicity that requires administration of fresh frozen plasma. Simple vitamin K administration cannot arrest this process. Hepatic synthetic failure provokes coagulopathy due to loss of pro- and anticoagulant proteins, altered activity of fibrinolytic proteases, and release of tissue thromboplastin from hepatic necrosis. Consumptive coagulopathy escalates bleeding complications and causes the appearance of schistocytes in the early phase of tissue injury. Bleeding complications and development of disseminated intravascular coagulation (DIC) complicate tolerance of liver injury and contribute to multiorgan failure in lethal amanitin hepatotoxicity.

While nephrotoxicity is common, it is far less critical than hepatotoxicity. Renal azotemia secondary to acute tubular necrosis is associated with active urine sediment characterized by the appearance of numerous hyaline and granular casts. Rarely, an acquired Fanconi renal tubular acidosis also develops.

Gross Liver Features

In acute lethal toxicity, the liver is swollen, may have a friable texture and mottled red and yellow pallor, and may display subcapsular hemorrhage. The yellow pallor reflects diffuse lipid vacuolation of residual hepatocytes.

Microscopic Hepatic Features

Histologic features are dominated by massive centrilobular to panlobular coagulative and hemorrhagic necrosis with loss of sinusoidal architecture. Rare residual viable hepatocytes abut normal-appearing portal elements; these hepatocytes demonstrate microvesicular lipid vacuolation and vacuolar degeneration. Centrilobular and subcapsular congestion and hemorrhage are usually obvious. There are limited inflammatory infiltrates.


There is no internationally accepted protocol for treatment of amanitin hepatotoxicity in humans; an IV form of silibinin is the preferred intervention. With acute presentation, decontamination is critical. Unfortunately, many poisoned dogs are not presented until after onset of clinical signs, such that gastric lavage is likely too late. After evacuating the stomach (self-induced vomiting, therapeutically initiated vomiting, or possible gastric lavage), activated charcoal (with sorbitol if there is no diarrhea) is used absorb the remaining toxin. This may also interrupt the enterohepatic circulation of amanitin. Oral and rectal administration routes may be used.

There are no data supporting utility of cholestyramine for capture of amatoxin. Concurrently, critical supportive care includes the following:

  • IV fluid therapy

  • electrolyte adjustments

  • maintenance of euglycemia with supplemental IV glucose

  • blood component therapy (usually fresh frozen plasma) to address complicating coagulopathy

  • antioxidants (IV N-acetylcysteine, oral bioavailable SAMe)

  • combinations of drugs thought to compete with amanitin OATP transport into hepatocytes and that block its enterohepatic circulation (high-dose penicillin G and IV silibinin; see below)

Adequate fluid therapy may be renal protective because amanitin is easily filtered via glomeruli; this toxin does not undergo renal tubular reabsorption.

Patient survival depends on the extent of hepatic injury, ability of surviving hepatocytes to regenerate, and management of systemic complications secondary to acute hepatic failure. In humans, liver transplantation is the salvage therapy for amanitin-induced FHF. Extracorporeal elimination (hemodialysis, hemoperfusion, or plasmapheresis) for removal of circulating amatoxin is thought to have limited value because amatoxins are only detected in plasma at the early phase of toxicosis and only for a short period of time. Newer extracorporeal techniques involving fractionated plasma separation and adsorption systems scavenge both protein-bound and water-soluble moieties and have shown benefit in humans with preclinical toxicity.

Specific measures aimed at limiting amanitin cellular uptake are predicated on the basis of experimental data from canine studies and observational deductions derived from human patients with Amanita hepatotoxicity. These include use of high-dose penicillin G or benzylpenicillin and high-dose IV silibinin to competitively block amanitin OATP transport. In humans, penicillin is dosed at 1,000,000 U/kg on day 1, 500,000 U/kg on days 2 and 3, and then discontinued. While there are multiple reports demonstrating circumstantial benefit, there is no consensus regarding efficacy of this treatment. In dogs, penicillin G given at a dose of 1,000 mg/kg, IV, every 5 hours after toxin ingestion appreciably decreased hepatic amanitin accumulation.

Silibinin, a flavonolignan derived from seeds of the milk thistle fruit, represents a mixture of two diastereomers, silibinin A and silibinin B. Mechanistically, silibinin competitively inhibits OATP amanitin transport into hepatocytes and during its subsequent enterohepatic circulation. Silibinin also affords benefit via suppression of TNF-alpha released from damaged tissues (thought to contribute to live injury) and via its antioxidant properties. IV silibinin is a bioavailable microcrystalline hydrosoluble ester of silibinin (with succinic anhydride); intravenous administration allows high-dose delivery not attainable by oral silibinin containing products. This product is not licensed for sale in the US and is acquired under a compassionate use assignment or from a European distributor.

The human protocol for amanitin toxicosis is to provide IV silibinin at 5 mg/kg during the first hour of presentation, followed by 20 mg/kg/d as a constant-rate infusion for 3 days. Alternatively, 20–50 mg/kg/d, IV, is continued for 3 to 4 days. There is no recommended limit on duration of silibinin administration because there are no recognized adverse effects. Studies in dogs with amanitin hepatotoxicosis suggest similar safety.

In early canine studies, IV silymarin administered at 50 mg/kg at 5 hours after amatoxin ingestion, repeated at 30 mg/kg at 24 hours, afforded hepatoprotection. In humans, IV silibinin administered up to 48 hours after amanitin ingestion also appears to prevent severe liver damage.

Microcystin: Blue-Green Algae (Cyanobacteria) Hepatotoxicity in Small Animals

Cyanotoxins derived from cyanobacteria (blue-green algae) are classified as hepatotoxic, neurotoxic, and dermatotoxic. Among these are microcystins, prominent hepatotoxins produced by freshwater blue-green algae (primarily Microcystis aeruginosa). Microcystins also are produced by other Microcystis species and well as other algae genera (eg, Anabaena, Oscillatoria, Nostoc, Planktothrix). Microcystins are monocyclic heptapeptides, the most common being microcystin-LR (MC-LR).

Microcystin toxins derive from blue-green algal blooms encountered during hot, dry weather, in shallow stagnant water, particularly water contaminated by runoff drainage containing a moderate to high nutrient content (ie, from agribusinesses [food-producing animals, agriculture] or suburban neighborhoods with heavy fertilizer use). The toxin is released from dying algal cells and is typically encountered in water with floating blue-green algae surface scum. Cyanobacterial blooms represent a worldwide environmental problem because they may develop in fresh or brackish water and even marine ecosystems.

The most extensively studied pathomechanism of MC-LR hepatotoxicity is its inhibition of protein phosphatases (protein phosphatase-1 [PPA1] and protein phosphatase 2A [PPA2A]), leading to increased protein phosphorylation. The dynamic balance between protein phosphorylation and dephosphorylation (catalyzed by phosphatases and kinases, respectively) regulates protein metabolism. Inhibition of phosphatase activity greatly impacts cellular homeostasis. Specifically, as PP2A is a major cell protein phosphatase, and its inhibition can adversely influence cell proliferation, cell death pathways (apoptosis, necrosis), cell mobility, cytoskeletal dynamics, cell cycle control, and numerous signaling pathways.

The liver is the prime target of MC-LR toxicosis because this toxin is preferentially and predominantly transported and accumulated in the liver. Severe hepatotoxicity induced by MC-LR has been documented in numerous mammals, including mice, rats, guinea pigs, sheep, swine, cattle, dogs, and humans.

MC-LR cannot penetrate cell membranes by simple diffusion; enteric uptake, access to the portal circulation, and hepatic localization are facilitated by organic anion transporting polypeptides (OATP). This toxin undergoes some degree of biotransformation, including GSH conjugation, biliary excretion, and at least some degree of enterohepatic circulation. Its enterohepatic circulation of this toxin has been exploited for therapeutic rescue (ie, cholestyramine toxin trapping).

In addition to its extreme hepatotoxic effects, chronic MC-LR exposure also has an oncogenic impact in humans, in whom it increases risk for hepatocellular carcinoma (recognized in China). Because MC-LR may cause acute or chronic hepatic injury leading to death and increases risk for primary hepatic neoplasia, the WHO has established a guideline limit of MC-LR in drinking water as 1 mcg/L (< 0.001 ppm). However, doses of MC-LR only 5-fold higher than this standard can provoke toxic effects in human hepatocytes.

After ingestion and enteric uptake, MC-LR is rapidly cleared from plasma within one hour as it is transported into hepatocytes via OATP. Toxicity of MC-LR is mediated by formation of a nonreversible covalent bond with the catalytic subunit of PPA1 and PPA2A, the toxin binding to cysteine residues in these enzymes. The toxin also binds and disables sulfhydryl (SH) groups in other proteins and antioxidants, including GSH and other sulfhydryl-bearing antioxidant proteins (eg, thioredoxin).

Flagrant cytoskeletal disruption occurs in acute severe MC-LR hepatotoxicity. Ordinarily, the cytoskeleton is a well-organized network of intracellular filaments (ie, microfilaments, microtubules, and intermediate filaments) that provide a scaffoldlike substructure essential for maintaining parenchymal architecture. This cytoskeletal structure maintains cell shape and influences cell division, migration, adhesion, contraction/relaxation, bile flow, and signal transduction.

In severe acute MC-LR, disruption of hepatocyte shape (cell rounding with loss of polyhedral silhouette), loss of intercellular adhesions, and disruption of hepatic cord structure and sinusoidal (endothelial) integrity lead to severe hemorrhagic necrosis, congestion, and parenchymal collapse. When severe, injury escalates to fulminant hepatic failure and circulatory shock.

Light microscopic and ultrastructural changes associated with MC-LR hepatotoxicity support both apoptotic and necrotic cell death pathways. Hepatocyte death in low-dose MC-LR toxicosis is apoptotic; however, with acute severe hepatotoxicity, organ injury is dominated by sinusoidal vascular leakage and severe centrilobular hemorrhagic necrosis. With high-dose exposure, plasma activities of liver enzymes and total bilirubin rapidly escalate within a few hours of toxin exposure. However, with apoptotic cell injury associated with lower-dose MC-LR hepatotoxicity, loss of sinusoidal architecture, widening of the space of Disse, dilation of bile canaliculi, and decrease in intercellular hepatocyte contact precede changes in membrane permeability and release of cytosolic enzymes (eg, ALT).

Widespread oxidative injury in MC-LR hepatotoxicity derives from GSH exhaustion (utilization in toxin conjugation, or scavenging reactive oxygen species [ROS]); decline in protectant antioxidant enzymes (GSH peroxidase, GSH-Px, GSH-reductase, superoxide dismutase [SOD], catalase); and escalated lipid peroxidation impacting cell and organelle membranes. Kidneys also may be similarly affected by high-level toxin ingestion. Oxidative injury can be initiated by mitochondrial injury disrupting the chemico-electrical gradient essential for oxidative phosphorylation and ATP synthesis. Microcystin is also known to bind and inactivate mitochondrial ATP synthase. Mitochondrial dysfunction also promotes accumulation of ROS and the mitochondrial permeability transition (MPT) that triggers apoptosis.

Clinical Signs

Rapid-onset lethargy, anorexia, and vomiting follow within 1–2 hours of toxin ingestion, with abrupt development of jaundice in dogs challenged by severe MC-LR exposure. With severe toxicity, continued inappetence and vomiting, escalating jaundice, and development of coagulopathy (surface hemorrhages, epistaxis, hematemesis, melena) ensue. However, exposure to sublethal levels of MC-LR may initially pass unnoticed without clinical signs or biochemical abnormalities.

Clinicopathologic Features

Dogs with single high-dose MC-LR exposure develop extreme increases in hepatic transaminase activity (ie, ALT and AST, 80–200× reference limits), with modest increases in ALP activity (~1.5–2.0× reference limits) and marked hyperbilirubinemia (7.0–12× reference limits) within 24 hours. Liver enzymes escalate over several days with fold increases in transaminases far exceeding those of ALP. Hyperbilirubinemia also worsens. In severe acute toxicity, lethal fulminant liver failure is associated with coagulopathy, hypocholesterolemia, hyperammonemia, and hepatic encephalopathy.

Recovery is possible in severely impacted dogs given supportive care (IV fluids, vitamin K, blood component therapy, antioxidants, and cholestyramine). Complete resolution of enzyme activity and hyperbilirubinemia may take up to several months in survivors. Exposure to sublethal levels of MC-LR may pass unnoticed without alarming increases in liver enzymes or hyperbilirubinemia.

Gross Hepatic Features

In acute severe lethal MC-LH hepatotoxicity, the liver appears swollen with rounded margins and has a dark red color due to the severe centrilobular hemorrhagic necrosis. Jaundiced tissues and surface hemorrhages are common in dogs with FHF. Animals with chronic sublethal MC-LR may have no overt changes.

Microscopic Features

Acute severe MC-LH hepatotoxicity leads to massive centrilobular and midzonal hepatocyte hemorrhagic necrosis with congestion in areas of parenchymal collapse. Few remaining viable hepatocytes abut portal tracts. Retained viable hepatocytes have poorly defined cell-to-cell attachments and display mild microvesicular lipid vacuolation. Based on studies in rodents, chronic low-dose MC-LH exposure leads to centrilobular microvesicular lipid vacuolation, loss of centrilobular sinusoidal organization, and abnormal nuclear chromatin clumping without evidence of apoptotic or necrotic cell injury. These changes are completely reversible within several months after removal of chronic low-dose MC-LH exposure.


Supportive care as described for FHF is indicated for dogs with severe acute MC-LH hepatotoxicity. Because oxidant injury is an early pathomechanism, IV N-acetylcysteine is advised. Oral bioavailable SAMe (20 mg/kg, PO, every 24 hours) and water-soluble vitamin E (10 U/kg, PO, every 24 hours) are given when the animal can tolerate enteral medications, and phosphatidylcholine (25–50 mg/kg, PO, every 24 hours) is advised to support membrane recovery and stabilization.

Because MC-LR undergoes enterohepatic circulation and has been shown to bind with cholestyramine, this binding resin should be used to remove recycling toxin. As there appears to be site-specific binding in the ileum, if oral cholestyramine is not possible because of vomiting, administration by high rectal installation is advised. High-dose IV silymarin has also been shown to be beneficial; this form of silymarin is commonly available in Europe but not in the US.

Sago Palm, Cycad Hepatotoxicity in Small Animals

There are many species in the hepatotoxic sago palm family; these plants are commonly used as yard ornamentals in temperate climates and sold as bonsais in large retail stores. Sago palm ingestion can lead to severe unremitting hepatic injury in dogs. Cycasin, in seeds, roots and leaves of cycads, is carcinogenic when orally ingested. Enteric bacterial hydroxylation by beta-ᴅ-glucosidase releases methylazoxymethanol (MAMA), the toxic metabolite.

Toxicologic impact of MAMA is similar to dimethylnitrosamine, another well-characterized and potent hepatotoxin. These toxins methylate cellular RNA and DNA, thereby inhibiting protein synthesis. MAMA specifically methylates guanine moieties in DNA and RNA, interfering with template function; it also potently inhibits nuclear and nucleolar RNA synthesis by inhibiting RNA polymerase II activity.

High-dose acute MAMA toxicosis causes centrilobular hepatocyte necrosis, regional parenchymal collapse and congestion, reactive inflammation (neutrophilic and macrophages), minor microvascular hepatic lipid vacuolation in periportal regions, and Kupffer cell necrosis. Decline in Kupffer cell function increases risk for complicating bacterial infections. MAMA induces distinct cytologic changes in hepatocytes that usually persist and are identifiable on cytologic evaluation of hepatic aspirates. Abnormal cytologic features persist for months beyond apparent recovery from acute toxicosis.

Acute salvage of severely intoxicated dogs may not arrest MAMA toxicity. The MAMA toxin is thought to slowly and spontaneously decompose with no evidence supporting enterohepatic circulation, hepatic metabolism, or conjugation to a water-soluble product.

After acute survival, some dogs develop proliferative foci with variable sinusoidal and bridging fibrosis. Retrospective evaluations of dogs 4–6 months after recovery from acute toxicosis are not published. Nevertheless, it is acknowledged that some dogs present months after the toxic acute incident with severe hepatic remodeling, splanchnic hypertension, ascites, enteric hemorrhage, hepatic encephalopathy, or problems related to ammonium biurate urolithiasis.

Hypertrophy of endoplasmic reticulum noted in experimental studies of MAMA speculated some degree of cytochrome P450 induction. Experimental pretreatment (rodents) with phenobarbital before MAMA administration decreases the severity of acute toxicity. There are no investigations regarding the impact of phenobarbital after acute MAMA toxicosis in dogs.

Clinical Signs

Clinical signs vary in respect to the lag time between toxin ingestion and initial presentation. In acutely intoxicated dogs (< 24 hours), signs commonly include vomiting and in declining frequency, diarrhea, hyporexia, lethargy, ptyalism, polydipsia, neurologic signs (ataxia, confusion, somnolence, obtundation, head pressing, rare seizures), GI bleeding (melena or hematochezia), and muscle tremors. During acute presentation, physical examination may be within normal limits. Within 24–48 hours, findings may include dehydration, jaundice (< 30% initially, progresses over time), abdominal discomfort, and abdominal effusion (< 15% initially, may develop over time).

Clinicopathologic Features

Although clinical signs may manifest within 4 hours of cycad ingestion, biochemical defects are often delayed for up to 72 hours. Initial ALT and AST activities may be within reference limits (within 24 hours) but then become mild to modestly increased. Increases in ALP activity lag behind trends in transaminases and are usually more modest unless there is progressive injury with tissue remodeling and biliary hyperplasia (ductular reaction). Variably, hypoalbuminemia, hyperbilirubinemia, hypoglycemia, and hypocholesterolemia may manifest.

Prognosis is not possible based on simple first inspection of test results. Rather, case outcome is more related to sequential trends in test results over the first several weeks, then monitored biweekly. Coagulopathies often manifest several days after acute presentation but also may appear after several weeks.


Abdominal radiographs are often within normal limits, with the exception of decreased detail attributed to abdominal effusion. Abdominal ultrasonography may be within normal limits or disclose normal to small liver size (small liver with chronic toxicity), altered parenchymal echogenicity (hyperechoic to hypoechoic), abdominal effusion, a thick gallbladder wall consistent with lymphedema, and intestinal atony.

Hepatic Gross and Histologic Features

In dogs dying from acute MAMA hepatotoxicosis, gross appearance of the liver may show scattered red foci (reflecting centrilobular necrosis and congestion) or show pallor (reflecting severe anemia rather than lipid vacuolation). At necropsy, all tissues are typically jaundiced, with numerous surface hemorrhages, evidence of enteric bleeding (melena or hematochezia), and abdominal effusion. In dogs dying from chronic MAMA toxicity, the liver is small and nodular with APSSs and ascites. Jaundice, visceral hemorrhages, and enteric bleeding are also evident.

Histologic features of acute toxicosis (0–2 days) include prominent centrilobular to midzonal hepatocyte and sinusoidal degeneration/necrosis, with severe regional congestion, parenchymal collapse, and reactive neutrophilic infiltrates. Conspicuous expansion of the space of Disse and damaged sinusoidal endothelium (fenestra widening) distort sinusoidal organization and are associated with loss of intercellular adhesions. Modest microvesicular lipid vacuolation develops in periportal hepatocytes.

Numerous hepatocellular cytologic aberrations are also observed, including cell individuation (detachment from reticulin scaffolding and intercellular tight junctions); loss of polyhedral hepatocyte shape; marked hepatocyte anisocytosis and anisokaryosis; appearance of megalohepatocytes and cells with marked karyomegaly; clumped nuclear chromatin; prominent large, dense, and sometimes vacuolated nucleoli; and mitotic derangements generating polynucleated cells. Variable canalicular cholestasis also be observed.


Retrospective reports of dogs experiencing cycad toxicosis describe recoveries of 34% (n = 60, with suspected cycad toxicity, poison control data) and in dogs with confirmed cycad toxicosis 50% (n = 34) and 64% (n = 14). Two clinical phases of cycad toxicosis are recognized in dogs:

  • an acute phase characterized by gastrointestinal and neurologic signs, with evidence of a modest hepatic insult (based on finding normal or mildly increased liver enzyme activity and lack of hyperbilirubinemia)

  • a latent phase characterized by increasing liver enzyme activity, jaundice, and progressive synthetic failure, with development of sinusoidal hypertension causing portal hypertension and abdominal effusion

In some dogs acute toxicosis is seemingly managed successfully only to discover serious latent hepatotoxicity weeks to months later (ammonium biurate crystalluria, acquired portosystemic shunting, and severe hepatic parenchymal remodeling). Thus, it is important to explain to an owner this possibility.

Initial decontamination is usually achieved via patient vomiting. If not, vomiting should be initiated if safe (see treatment for fulminant hepatic failure Fulminant Hepatic Failure in Small Animals Fulminant hepatic failure is characterized by the abrupt critical decline in liver function in a patient without preexistent liver disease, associated with a 2-3–fold increase in liver enzymes... read more ). Cleansing enemas followed by an activated charcoal enema should also be considered if plant material remains within the GI tract. Activated charcoal should be administered after vomiting ceases to be productive. If substantial gastric debris is noted on abdominal ultrasonography, endoscopy can be used to assist in decontamination.

Administration of activated charcoal was strongly associated with survival in one study. There is anecdotal observation that cholestyramine may accelerate recovery from MAMA toxicity. However, there is no information regarding enterohepatic circulation of this toxin. Otherwise, supportive care is used to judiciously adjust hydration and electrolyte status, provide water-soluble vitamins and antioxidant support (ie, NAC given IV, bioavailable SAMe if oral medications can be used), water-soluble vitamin E (10 U/kg, PO), antibiotics to protect against translocating opportunistic bacteria, and blood component therapy as needed. Treatment of hepatic encephalopathy may also be needed.

quiz link

Test your knowledge

Take a Quiz!