Introduction

Mushroom ingestion by dogs is common, with > 400 reports from the Pet Poison Helpline from 2018 to 2019 and 2,980 reports of potential mushroom ingestion in animals from the Animal Poison Control Center of the American Society for the Prevention of Cruelty to Animals from 2013 to 2017.¹ Although mushroom ingestion is often unnoticed and usually does not result in toxicosis, ingestion of some mushrooms can lead to serious liver damage and death. Hepatotoxicosis has been reported for dogs,^2,3 cats,⁴ and beef calves⁵ that ingested mushrooms.

The primary hepatotoxins found in mushrooms are amanitins. Amanitins are bicyclic octapeptides, and of the 8 known amanitins, α-amanitin is the most toxic.⁶ This toxin induces acute liver failure and is found in over 38 different species of mushroom from the genera Amanita (including Amanita phalloides and Amanita ocreata), Galerina, and Lepiota. However, not all Amanita spp have α-amanitin. Amanita phalloides, known as death cap, is the most well-known Amanita sp and is implicated in the majority of cases of α-amanitin toxicosis in people in North America.⁷ Amanita phalloides, originally native to Europe, is now found worldwide, having been introduced to non-native regions on the roots of imported trees. In many regions, A phalloides is commonly found alongside planted trees, but it is naturalized in California.⁸ On the west coast of the United States, it is found most commonly alongside oak trees; on the east coast, it is usually associated with pine trees.⁹ Amanita phalloides typically has a green to yellow cap, but its color can vary (eg, brown, tan, or white) because of environmental conditions.^9,10 Clusters of cases occur in certain years because of variations in weather conditions, including precipitation, that impact mushroom growth.^8,10 One gram of A phalloides can contain 1.5 to 2.5 mg of α-amanitin,¹¹ and 1 mushroom can contain up to 15 mg of α-amanitin.⁸ The estimated oral LD₅₀ of α-amanitin for people is 0.1 mg/kg (0.045 mg/lb),¹¹ and the oral LD₅₀ for methyl-α-amanitin for dogs is 0.5 mg/kg (0.23 mg/lb).¹²

Amanitins are rapidly absorbed from the small intestine through a sodium-dependent bile acid transporter. They travel to the liver via the portal vein and are actively transported into hepatocytes by another bile acid transporter identified in people as OAT-P1B3.¹³ Amanitins disrupt RNA transcription by inhibiting nuclear RNA polymerase, leading to deficient intracellular protein synthesis and then hepatocellular death.⁶ Other metabolically active cells, such as intestinal epithelial cells and proximal renal tubular cells, are similarly affected by amanitins.¹ Amanitins undergo enterohepatic recirculation and active reabsorption from the small intestine. Amanitins are not highly protein bound, and they are eliminated from the plasma through the kidneys within 48 hours of ingestion.^14,15

α-Amanitin toxicosis should be considered in dogs presenting with signs of acute liver failure, especially gastrointestinal signs, hypoglycemia, and marked coagulation abnormalities. Rarely, however, has mushroom exposure been observed, and other toxicoses, including acetaminophen, Sago palm (Cycad spp), xylitol, and blue-green algal microcystins may manifest similarly.¹⁶ Because of the difficulty in confirming a diagnosis of α-amanitin toxicosis without known ingestion of a hepatotoxic mushroom, a sensitive antemortem test for the detection of α-amanitin by liquid chromatography–mass spectrometry in urine was developed.¹⁷ The test can also be performed with mushroom fragments and other animal specimens, such as vomitus, serum, bile, and fresh (nonpreserved) liver and kidney tissues.

The purposes of the retrospective study reported here were to describe the clinical and clinicopathologic abnormalities, treatments, and outcomes for dogs confirmed with α-amanitin toxicosis and to identify and compare factors that differed between dogs that survived (survivors) and dogs that were euthanized or died (nonsurvivors).

Materials and Methods

Results

Clinicopathologic findings

Biochemical analyses were performed on the serum of 57 (97%) dogs at the time of initial evaluation at the hospital; biochemical analyses were not initially performed for 2 dogs that were suspected to have gastroenteritis but were performed when these dogs returned to the hospital within 24 hours of their initial presentation. All dogs with initial serum biochemical analyses had increased ALT activity (range, 144 to 32,395 U/L; reference interval, 18 to 121 U/L; Table 1), and ALT activity for many dogs could not be quantified with the in-hospital biochemistry analyzer without first diluting the serum. Therefore, subsequent samples were submitted to a veterinary diagnostic laboratory for definitive quantification. Alanine aminotransferase activity was recorded many times throughout hospitalization, and 27 of the 59 (46%) dogs ultimately had ALT activity > 10,000 U/L.

Table 1

Median (interquartile [25th to 75th percentile] range) values for specified variables for dogs that were euthanized or died (nonsurvivors) versus those that survived to hospital discharge (survivors) with confirmed α-amanitin toxicosis at an emergency and specialty hospital in northern California between 2006 and 2019.

Variable	No. of dogs	Reference interval	Nonsurvivors	Survivors	P value
Onset of clinical signs to presentation (h)	59	NA	12.0 (6.0 to 18.0)	12.0 (4.0 to 12.0)	0.53
Onset of clinical signs to resolution (death, euthanasia, or hospital discharge; h)	59	NA	36.0 (28.0 to 56.0)	100.0 (72.0 to 128.0)	< 0.01
Age (y)	59	NA	0.7 (0.33 to 5.0)	0.3 (0.2 to 0.8)	0.08
Weight (kg)	59	NA	10.4 (4.8 to 24.8)	11.5 (6.2 to 20.3)	0.84
Temperature at presentation (°C)*	59	37.2–39.2	38.9 (38.0 to 39.5)	39.1 (38.7 to 39.4)	0.33
ALT at presentation (U/L)	57	18–121	1,406 (605 to 5,965)	992 (420 to 3,558)	0.36
Highest ALT during hospitalization (U/L)	59	18–121	9,095 (4,525 to 13,263)	9,027 (4,922 to 12,839)	0.82
Glucose at presentation (mg/dL)	57	74–143	42.0 (27.0 to 67.0)	63.5 (44.5 to 102.5)	0.07
Lowest glucose during hospitalization (mg/dL)	59	74–143	32.0 (23.0 to 41.0)	58.0 (45.5 to 74.0)	< 0.01
ALP (U/L)	58	5–160	387.0 (129.0 to 572.0)	469.0 (263.0 to 1,226.0)	0.07
Cholesterol (mg/dL)	54	131–345	204.0 (157.0 to 259.0)	201.0 (182.5 to 285.5)	0.58
Bilirubin (mg/dL)	57	0.0–0.3	2.2 (1.3 to 2.8)	1.3 (0.7 to 3.2)	0.28
Albumin (g/dL)	58	2.7–3.9	2.7 (2.5 to 3.1)	3.0 (2.8 to 3.2)	0.34
Creatinine (mg/dL)	55	0.5–1.5	0.9 (0.5 to 1.4)	0.7 (0.4 to 1.0)	0.13
BUN (mg/dL)	58	9–31	15.0 (12.0 to 21.0)	11.0 (9.0 to 15.0)	0.04
PT at presentation (s)	37	11–14	> 100 (34.0 to > 100)	36.5 (26.0 to 96.0)	0.08
Highest PT during hospitalization (s)	53	11–14	> 100 (> 100 to > 100)	70.0 (42.0 to > 100)	0.02
aPTT at presentation (s)	31	60–93	301.0 (124.0 to > 350.0)	151.0 (97.0 to 208.0)	0.07
Highest aPTT during hospitalization (s)	48	60–93	> 350.0 (210.0 to > 350)	178.0 (139.0 to 301.0)	< 0.01

Values of P < 0.05 were considered significant.

^*To convert °C to °F, multiply by 9/5 and add 32.

NA = Not applicable.

Forty-six of 57 (81%) dogs with serum biochemical analyses at initial evaluation were hypoglycemic (glucose < 74 mg/dL), and 56 of 59 (95%) ultimately had hypoglycemia during hospitalization (Table 1). Hypoglycemia recorded during hospitalization was significantly (P < 0.01) worse for nonsurvivors, compared with survivors (Figure 1).

View Full Size

Box-and-whisker plots of the lowest serum glucose concentrations recorded during hospitalization for 57 dogs that were euthanized or died (n = 45) because of α-amanitin toxicosis or survived (12) with treatment for α-amanitin toxicosis from 2006 to 2019 at a northern California veterinary emergency and specialty hospital. Survivors had lowest serum glucose concentrations that were significantly (P < 0.05) higher than nonsurvivors. The horizontal line within each box represents the median, boxes represent the interquartile (25th to 75th percentile) range, and whiskers indicate the maximum and minimum values. The black dot indicates an outlier.

Citation: Journal of the American Veterinary Medical Association 258, 5; 10.2460/javma.258.5.502

• Download Figure
• Download figure as PowerPoint slide

Prothrombin time and aPTT were determined for 55 (93%) dogs with an in-hospital analyzer because of suspected development of liver failure; 4 dogs did not undergo coagulation testing because they were euthanized at initial presentation. Results for 5 dogs were initially within reference intervals, but coagulation times became prolonged during hospitalization (Table 1). All evaluated dogs had prolongation of PT, aPTT, or both during hospitalization, with 35 (64%) dogs having had PT or aPTT results that were > 100 or 300 seconds, respectively.

Discussion

The retrospective study presented here provided a summary of the signalment; physical examination findings; results of serum biochemical analyses, coagulation tests, and a urine α-amanitin test; treatments; and outcomes for dogs that presented to an emergency and specialty veterinary hospital in northern California with amanitin toxicosis and provided a comparison of factors between dogs that survived to hospital discharge and dogs that were euthanized or died.

In the present study, the largest number of cases was in 2014 (n = 21), presumably attributable to variations in weather conditions, including precipitation. Multiple α-amanitin–containing mushroom species, including A ocreata and Galerina marginata, are present in northern California. The most common species is A phalloides, but some dogs may have ingested these other α-amanitin–containing mushrooms, because the test for detecting α-amanitin cannot distinguish between toxin from A phalloides and toxin from other mushroom species. For example, dogs with α-amanitin toxicosis in the winter and spring may have ingested A ocreata because A ocreata appears during these seasons. The California Poison Control Center did not identify an obvious increase in cases of human amanitin toxicosis in 2014,^b but clusters of cases during particular years have been previously reported,^8,10 including a cluster of 14 cases in the San Francisco Bay Area in 2016. The present study, however, did not reveal a correspondingly high number of dog cases in 2016 (n = 5). Awareness of case clusters of α-amanitin toxicosis could make this toxicosis the leading differential diagnosis and enable veterinarians to warn their clients of the risk of toxicosis in the area.

In veterinary medicine, diagnosis of α-amanitin toxicosis is often hindered because mushroom ingestion is rarely witnessed by a dog owner and testing for the presence of α-amanitin is uncommonly performed. In the present study, only 4 owners were aware that their dogs had eaten a mushroom. In areas where toxic mushrooms are prevalent, owners should be educated to monitor their pets closely for possible mushroom ingestion. If mushroom fragments are available, they can be submitted for α-amanitin detection or the mushroom species can be identified through PCR assay.¹⁹ Most dogs present relatively late in the course of toxicosis, such that it can only be definitively diagnosed by detection of α-amanitin. Confirmation of the toxin's presence is important because it can aid veterinarians in the treatment of the affected pet and inform owners of the risk of exposure for any other pets in the household. Additionally, confirmation could be a marker of how invasive mushroom species continue to expand throughout North America. A rapid lateral flow assay has been recently developed for the detection of α-amanitin in the urine of people and dogs,²⁰ and such a test could make α-amanitin detection possible in the hospital, improving the ability to identify affected dogs and begin appropriate treatment.

The present study revealed that dogs with trace (vs positive) urine α-amanitin had improved likelihood of survival. Whether this reflects a lower toxic dose, improved toxin clearance, or a delay in obtaining the sample for analysis is unknown.

The clinical course of α-amanitin toxicosis is multiphasic.^1,21 During a latency period of 6 to 24 hours after ingestion, clinical signs are usually absent. Physicians use a patient's report of a latency period to help them differentiate A phalloides poisoning from other mushroom poisonings, which typically cause vomiting within 1 to 2 hours of ingestion. During the gastrointestinal phase of toxicosis, vomiting, intestinal cramping, and diarrhea last 12 to 24 hours, secondary to enterocyte membrane damage by biologically active phallotoxins. After the gastrointestinal phase, about 36 to 48 hours after mushroom ingestion, is a convalescent phase during which clinical signs abate. The final phase is α-amanitin–induced liver and kidney failure, culminating in CNS dysfunction secondary to hepatic encephalopathy and cerebral edema.

In the present study, the time of mushroom ingestion was rarely known, and the distinct phases of illness were not always identified. In the 4 cases in which mushroom fragments were pulled from dogs' mouths, clinical signs were not noted until 12 to 24 hours after ingestion, consistent with the latency period reported in people. Only 13 (22%) dogs went through a recognizable convalescent phase before progression to liver failure. Presumably, some owners ignored or simply did not identify mild gastrointestinal signs in their dogs. Possibly, however, dogs have increased susceptibility to α-amanitin toxicosis or they ingest comparatively higher amounts of toxin (vs people), resulting in a more rapid progression to fulminant liver failure. An immediate progression from severe gastroenteritis to fulminant organ failure has been described in people with severe toxicosis, and clinical signs are likely dose dependent. By the time most dogs in the present study were brought to the hospital, clinical and biochemical evidence of liver failure was already present. Because of the inconsistency to recognize the distinct phases of α-amanitin toxicosis, whether those phases are even present for dogs, and possible differences in the ingested amounts of toxin between people and dogs, identification of distinct phases of toxicosis should not be attempted to screen dogs for toxicosis.

The dogs of the present study had increased serum activities of ALT and ALP, hypoglycemia, coagulopathy, and encephalopathy attributable to acute, extensive hepatocellular necrosis and subsequent compromised hepatic blood flow and bile production. These signs of liver failure are not specific to α-amanitin toxicosis; however, early-onset hypoglycemia was a distinguishing biochemical abnormality.

Forty-six of 57 (81%) dogs were found to be hypoglycemic at presentation, and 7 of the remaining 11 dogs became hypoglycemic during hospitalization. α-Amanitin depletes hepatic glycogen in people and animals,²² and this mechanism is the most likely explanation for the early-onset hypoglycemia noted in the dogs of the present study. A second possible mechanism of hypoglycemia is α-amanitin–induced insulin release from pancreatic beta cells.²³ In 1 hypoglycemic dog reported here, a serum insulin-to-glucose ratio was determined at admission to the hospital, and insulin concentration was undetectable (< 2 μU/mL; reference interval, 5.2 to 41.5 μU/mL), which was an expected result with marked hypoglycemia (< 60 mg/dL) and, therefore, a result inconsistent with α-amanitin–induced insulin release. However, study is needed on the role of α-amanitin on insulin release from the pancreatic beta cells and hypoglycemia in affected dogs.

Identification of hypoglycemia helped to distinguish affected dogs from those with other causes of gastroenteritis. Six dogs in the present study were initially treated as outpatients for suspected gastroenteritis, yet returned to the hospital within 24 hours with progressive clinical signs. At initial presentation, these dogs had serum glucose concentrations within reference intervals or did not have serum biochemical analyses performed. When a vomiting dog was hypoglycemic, α-amanitin toxicosis was considered as a differential diagnosis; other differential diagnoses included xylitol toxicosis, severe anaphylaxis, sepsis, and porto-systemic shunt. Some clinicians were concerned that some of the 22 (37%) dogs with rectal temperatures > 39.2°C had sepsis because they did not expect dogs with toxicosis to present with high rectal temperatures. Although hypoglycemia was significantly worse during hospitalization for nonsurvivors, the decision on whether to treat should not be solely influenced by the presence of marked hypoglycemia because 5 of the 13 surviving dogs presented with serum glucose concentrations of 20, 30, 44, 45, and 72 mg/dL.

In this study, serum ALT activity was increased but highly variable. Surprisingly, ALT activity did not significantly differ between survivors and nonsurvivors. Although 27 (46%) dogs had ALT activity > 10,000 U/L, some dogs that had ALT activity < 10,000 U/L progressed to fulminant liver failure. Thirteen (22%) dogs had ALT activity < 500 U/L at initial evaluation. Because the clinical course of α-amanitin toxicosis is rapid, mild to moderate (vs severe) hepatocellular damage, as denoted by ALT activity, does not exclude α-amanitin toxicosis. For this reason, ALT activity may not have significantly differed between survivors and nonsurvivors.

Coagulation testing was performed for 55 dogs, and results for all dogs were abnormal at least once during hospitalization. The time of testing varied in relation to α-amanitin exposure because of the variability in time between the onset of clinical signs and presentation to the hospital. Also, coagulation tests were not part of the minimum database, and clinicians ordered coagulation testing only once they suspected liver failure. Prolongation of PT and aPTT during hospitalization was greater for nonsurvivors, compared with survivors, but was not significantly different at presentation. However, deciding whether to euthanize solely on the basis of prolonged PT and aPTT may be erroneous because many of the survivors had evidence of coagulopathies.

Despite years of research and experience, treatment of people with α-amanitin toxicosis is not standardized⁷; treatments frequently include decontamination, supportive care, and mitigation of enterohepatic recirculation and transport of α-amanitin into hepatocytes.^7,24 For people and dogs, decontamination and mitigation of enterohepatic recirculation are frequently limited by the delay between mushroom ingestion and presentation at the hospital. Although decontamination is often initiated at presentation, studies⁷ of α-amanitin kinetics reveal that plasma concentrations of α-amanitin are low, urinary excretion of toxin is rapid, and decontamination is most valuable in the first 12 to 24 hours after mushroom ingestion. However, α-amanitin undergoes enterohepatic recirculation, so some physicians advocate for aggressive and prolonged decontamination, such as administration of multiple doses of activated charcoal.¹¹ Only 18 (31%) dogs in the present study were administered activated charcoal.

Other methods of disrupting enterohepatic recirculation of toxins that have been investigated include administration of cholestyramine to adsorb bile acids and octreotide to reduce gallbladder motility and performance of biliary drainage procedures.^8,11,24,25 Placement of biliary drainage catheters is challenging in veterinary patients because of small patient size, gallbladder anatomy, and limited access to needed instrumentation. Ultrasound-guided cholecystocentesis is commonly performed in veterinary medicine.²⁶ However, it may be risky for patients with coagulopathies, and unlike a biliary drainage catheter, the procedure only allows for 1-time rather than continuous biliary drainage. Ultrasound-guided cholecystocentesis was performed in only 3 dogs in the present study because the risk of bleeding was considered high, given that most dogs had prolonged PT and aPTT by the time α-amanitin toxicosis was suspected or diagnosed. Also, the benefit of 1-time removal of only a small amount of bile was likely low, considering that studies^15,21 show biliary excretion of α-amanitin accounts for < 10% to 20% of its excretion. Although dogs with indwelling biliary drainage catheters placed prior to experimental ingestion of Amanita mushrooms had less severe toxicosis (vs those without biliary drainage catheters), it was likely because of poor toxin absorption from the small intestine without the aid of bile, not because biliary drainage prevented or lessened enterohepatic recirculation of toxin.²⁶ Clinical assessment of the impact of cholecystocentesis on patient outcomes is lacking and merits additional research. Cholecystocentesis may be helpful for patients identified within the first 24 hours of toxicosis. Although cholecystocentesis was performed in 3 dogs, none of these dogs survived (all 3 were euthanized).

Because α-amanitin is filtered through the glomeruli, maintaining urine output is essential for its elimination. However, diuresis through IV administration of crystalloid fluids or administration of loop diuretics (eg, furosemide) does not increase α-amanitin elimination or alter outcome. Also, hemodialysis and hemoperfusion are ineffective in eliminating toxin.²⁷ All treated dogs in the present study were administered lactated Ringer solution (IV) to correct hypovolemia secondary to fluid loss through the gastrointestinal tract and hypovolemic shock and, later, to maintain normovolemia. Additionally, dogs received treatments that addressed hypoglycemia and signs of nausea and received antimicrobials (IV), a standard practice for people with acute liver failure or hepatic encephalopathy because of increased risk of sepsis and multiorgan failure.²⁷ Many dogs were also administered lactulose orally, although lactulose may have little benefit in people for resolving hepatic encephalopathy secondary to acute liver failure.²⁸ Patients with acute liver failure can develop cerebral edema and intracranial hypertension.²⁸ These changes inevitably contributed to the progression of hepatic encephalopathy in several dogs of the present study, and when dogs became obtunded (excluding hypoglycemia-induced obtundation that resolved with restoration of euglycemia), they were euthanized or died.

A review⁷ of > 2,000 cases of human α-amanitin toxicosis in North America and Europe revealed that treatments associated with better outcomes included silybin and N-acetylcysteine.²⁸ Silybin has a direct protective effect in α-amanitin toxicosis.^13,29,30 Silybin is the most potent of the flavonoligans in silimarin, which is derived from milk thistle. Silybin has poor bioavailability, but its absorption after oral administration improves when it is complexed with phosphatidylcholine, thus creating a lipophilic-phytosome complex.^31,32 In people, silybin inhibits the transporter system OATP1B3 that transports α-amanitin into hepatocytes, such that hepatocyte uptake of α-amanitin is mitigated.¹³ Silybin may also be beneficial by inhibiting proinflammatory tumor necrosis factor-α, disrupting enterohepatic recirculation, and stimulating intrahepatocellular protein synthesis.³³

Silybin is protective in dogs when it is administered IV within 5 hours of α-amanitin ingestion.^29,30 The efficacy of oral or delayed (> 5 hours) administration of silybin is unknown. Silybin is typically administered IV to people with α-amanitin toxicosis, with an initial dose of 5 mg/kg, then 20 mg/kg/d (9.1 mg/lb/d) as a constant rate infusion. A silybin product^c has been available for IV administration in Europe since 1984 and is thought to be of benefit.^33,34 People who are not administered silybin IV should be administered 20 to 50 mg/kg/d, PO.¹¹

Through the years of the present study, attending clinicians increasingly believed in the importance of silybin as a treatment for dogs with α-amanitin toxicosis. Therefore, dogs with α-amanitin toxicosis after 2013 were managed with high and frequent doses of silybin. Dogs were administered high doses of silybin 2 to 4 hours prior to starting activated charcoal in an attempt to prevent silybin adsorption by the charcoal. Because the bioavailability of milk thistle products varies, administration of high doses of milk thistle complexed with phosphatidylcholine is recommended.^31,32 Treatments varied in the present study, so their effect on patient outcome (survival vs nonsurvival) could not be compared.

Death rates of people with A phalloides toxicosis are between 10% and 30%.¹⁰ A review³⁵ of 27 cases in San Francisco between 1997 and 2014 indicates fatal toxicosis in 3 (11%). Of the 50 dogs that were hospitalized for treatment in the present study, 37 (74%) were euthanized or died.

The 13 surviving dogs recovered fully and were assessed by their owners as healthy within 1 to 2 days after hospital discharge. Each returned for reexamination a few weeks later and had serum ALT activity and biochemical markers of liver function (eg, albumin, cholesterol, and bilirubin) within reference intervals. One dog developed mild azotemia 2 weeks after hospital discharge, but the azotemia resolved at a reexamination 1 month later. Of those dogs not lost to follow-up, many were doing well up to 5 years later.

The findings of the present study indicated that veterinarians should consider α-amanitin toxicosis when dogs present with severe gastrointestinal signs accompanied by increased serum ALT activity and hypoglycemia. If toxicosis is suspected, urine should be collected for toxin detection and supportive care should be initiated. Alanine aminotransferase activity was not significantly different between survivors and nonsurvivors; therefore, the magnitude of increased ALT activity should not be used for treatment (vs euthanasia) decision-making. Although death was frequent, the survival rate among dogs that were treated was 26%. Ongoing research in human and veterinary medicine might provide new treatment options to mitigate α-amanitin's effects and improve patient outcomes.