Introduction

Cycad palms, also known as sago palms, are members of a plant family that predominated during the Mesozoic era. Their natural habitat is tropical and subtropical environments, including Florida, California, and the Gulf Coast region of the United States.¹ Their distribution is diverse, and they have been used for nutritional, decorative, and medicinal purposes. They are characterized by a large crown of compound leaves with a stout trunk and include Cycas circinalis, Cycas revolute, and Zamia floridana. The male cycads bear cones and the female plants produce flowers and seeds.

Toxicity has been documented in humans, cattle, sheep, horses, pigs, and dogs.^1,2–8 All parts of the plant are toxic, but the seeds contain the highest concentration of toxins, with one to two seeds being potentially lethal to an average-sized dog.⁹ Toxicity associated with cycad palm ingestion includes hepatotoxicity, neurotoxicity, and gastrointestinal disease.^1,2–8 Three toxins have been isolated: azoglycosides (cycasin, macrozamin, neocycasin), a neurotoxic amino acid (β-N-methylamino-L-alanine), and an unidentified high molecular weight compound.^1,3,5 Cycasin, the most significant azoglycoside toxin, is neurotoxic, hepatotoxic, carcinogenic, mutagenic, and teratogenic. β-N-methylamino-L-alanine, an excitatory neurotoxic amino acid, mimics the effect of glutamate by overstimulating N-methyl-D-aspartate receptors, resulting in a cascade of enzyme activation leading to neuronal death.¹⁰ The unidentified high molecular weight compound is suspected of causing hind limb paralysis and axonal degeneration in the central nervous system of cattle.^4,5,7,9

In canine patients, clinical signs of toxicity have been reported within 4 hr post-ingestion and include gastrointestinal disturbances, neurological abnormalities, and liver disease.^2–5,9 While clinical signs may be evident soon after ingestion, clinicopathologic abnormalities may not be apparent for up to 72 hr post-ingestion. These include elevations in bilirubin, alanine transaminase (ALT), aspartate aminotransferase (AST), and alkaline phosphatase (ALP), hypoproteinemia, hypoglycemia, thrombocytopenia, and prolonged clotting times.^2–5,9 Should patients survive the acute phase of toxicity, chronic liver disease may ensue secondary to widespread hepatic tissue damage and the lack of adequate functional mass.

A retrospective study by the National Animal Poison Control Center (NAPCC) in 1997 reviewed 60 cases of dogs that reportedly ingested cycad palms.⁴ Of those 60 dogs, 95% developed liver and gastrointestinal tract problems and 53.3% exhibited neurological signs.⁴ The mortality rate was 32.1%, but only 26.6% of the reported cases had confirmed ingestion.⁴ Another retrospective study in 2011 by Ferguson et al. evaluated 34 client-owned dogs with confirmed cycad palm toxicosis. This study revealed a higher case fatality rate of 50% and found initial ALT activity, total bilirubin concentration, coagulation times, and albumin (initial and nadir) as parameters useful in determining prognosis.²

The purpose of this retrospective study is to report physical examination findings, clincopathologic abnormalities, treatment modalities, and case outcome in 14 dogs with cycad palm poisoning and to identify any variables that may help predict survivors from nonsurvivors.

Materials and Methods

Medical records of Gulf Coast Veterinary Specialists (GCVS) were searched between 1998 and 2008 to identify dogs that were documented or suspected of cycad palm ingestion. Sixteen medical records were retrieved. Cases were included if the owner confirmed cycad palm ingestion or if plant material was present in the vomitus. Only dogs with documented ingestion that underwent treatment were included in the study.

Data obtained from the medical records included age, breed, sex, weight, time from ingestion to presentation at a primary care or emergency facility, time from ingestion to presentation at GCVS, clinical signs, physical examination findings, treatment modalities, and outcome. Results from the following diagnostic tests and procedures were reviewed when available: blood glucose, bile acids, complete blood count, packed cell volume and total solids, biochemistry profile, thyroid level, urinalysis, prothrombin time (PT), partial thromboplastin time (PTT), abdominal radiographs, abdominal ultrasound, and liver histopathology. Initial and serial laboratory abnormalities were reviewed. For the purpose of this study, initial presentation refers to the primary care or emergency facility the patients originally presented to post-ingestion, and initial laboratory values refers to blood work performed at these facilities prior to referral. All initial laboratory work was performed on in-house laboratory analyzers, with the exception of one patient. This patient's initial blood work was performed at a commercial reference laboratory^a and was unremarkable. All blood, fluid and tissue analysis performed at GCVS were submitted to a commercial reference laboratory^b. Blood lactate, PT, and PTT were performed on in-house analyzers^c,d. Case outcomes were determined by reviewing medical records or calling clients and referring veterinarians as necessary. The cause of death was recorded, differentiating patients who arrested, those who were euthanized, and those who died later of cycad-related disease.

Statistical analyses were performed with the aid of commercially available software^e. Patients were assigned to either the survivor or nonsurvivor group, with survivors defined as dogs that were discharged from the hospital and lived for a minimum of 3 mo. Analysis focused on variables recorded at the time of initial presentation as well as during treatment at GCVS and were compared via the Mann-Whitney U test for continuous data and the Fisher exact test for categorical data. Values of P < .05 were considered significant.

Results

Fourteen patients were included in this study: 12 ingested the seed; two ingested other portions of the plant. Median age was 3 yr, with a range of 8 mo to 5 yr. There was no significant difference in the distribution of male and female patients (six and eight dogs, respectively). Ten patients were neutered and four were intact. Breeds included three dachshunds, two Labrador retrievers, two Cavalier King Charles spaniels, two Yorkshire terriers, one golden retriever, one beagle, one poodle, one Jack Russell terrier and, one Pomeranian mix. Median weight was 6.25 kg (range 2.7–43.9 kg). For survivors, median weight was 4.5 kg (range 2.6–10.7 kg) and, for nonsurvivors, was 6.5 kg (range 4.4–43.9 kg), but this difference was not significant. Most patients presented during the fall months (36%), while equal numbers presented in the summer and winter months (28.5% each). The fewest patients presented during the spring months (7%).

Thirteen of 14 patients (93%) were evaluated by a primary care or emergency veterinarian prior to referral. The time from cycad palm ingestion to initial presentation at these facilities varied between these 13 patients. Six patients presented within hours, five patients within 1–2 days, and two patients presented 2–4 wk post-ingestion. All 13 had laboratory work performed at their respective referring facility. The fourteenth patient presented directly to GCVS within 12 hr post-ingestion. Initial laboratory abnormalities included elevated ALP, ALT, and AST, hypoglycemia, hyperglycemia, hyperbilirubinemia, decreased blood urea nitrogen (BUN), elevated BUN, hypocholesterolemia, prolonged PT and PTT, and thrombocytopenia (Table 1). Of these abnormalities, elevations in ALT, ALP, and total bilirubin and hypoglycemia and thrombocytopenia were most pronounced, but differences between survivors and nonsurvivors were not significant. Nine of 11 (82%) patients with elevated ALT presented within 2 days post ingestion. Six of these nine patients had ALT elevations noted within hours post exposure. Elevations in ALP were present in six (43%) patients. Hyperbilirubinemia was noted in five (36%) patients, four of which presented 2 or more days post exposure. Initial bilirubin levels were normal in all nine patients that presented within 24 hr of exposure. Five patients were hypoglycemic (36%), and all were nonsurvivors. Four out of the seven (57%) thrombocytopenic patients did not survive. Therapy prior to referral varied and included IV crystalloid and colloid fluid therapy, plasma transfusions, various antiemetics, gastroprotectants, antibiotics, and hepatoprotectants.

TABLE 1 Initial Laboratory Abnormalities Post Cycad Palm Ingestion

View Fullscreen

Time from ingestion to presentation at GCVS also varied between the 14 patients. Seven patients (50%) presented within 48 hr post ingestion, nine patients (64%) within 5 days, 12 patients (86%) within 14 days, and all 14 patients (100%) within 29 days. Median time to presentation was 4 days. Clinical signs included vomiting (92%), diarrhea (50%), decreased appetite (42%), lethargy (28%), ptyalism (21%), obtundation (21%), weight loss (14%), gastrointestinal bleeding (14%), seizures (14%), muscle tremors (7%), and head pressing (7%). Physical examination findings included dehydration (28%), icterus (28%), abdominal fluid wave (14%), abdominal discomfort (14%), and hyperemic mucus membranes (7%). Physical examination was unremarkable in three (21%) patients.

Laboratory data was collected on presentation and serially while patients were treated at GCVS. Peaks in ALP, ALT, AST, and bilirubin levels, prolonged and shortened coagulation times and nadir albumin, BUN, cholesterol, and glucose levels were assessed (Table 2). Eleven out of 13 (85%) patients had an elevated ALP, all patients had an elevated ALT and nine of 10 patients (90%) had an elevated AST. Albumin levels were decreased in 11 of 13 (85%) patients. BUN was decreased in four of 13 (31%) patients and cholesterol levels were decreased in nine of 12 (75%) patients. Ten of 14 patients (71%) were hypoglycemic during treatment. Because blood glucose nadir levels were evaluated, some values reported in Table 2 were obtained with patients on IV dextrose supplementation. Bilirubin levels were elevated in 11 of 12 (92%) patients. All 14 patients became thrombocytopenic. PT was prolonged in 10 of 11 of patients (91%). Two were out of the reference range and were both nonsurvivors. PT was shortened in one (9%) patient. PTT was prolonged in three of six (50%) patients and decreased in two of six (33%) patients. Initial bile acids were elevated in all five patients when measured, including one nonsurvivor and four survivors.

TABLE 2 Peak and Nadir Laboratory Data Collected at GCVS

View Fullscreen

Peak and nadir hematological and biochemical values were evaluated with regard to survivors and nonsurvivors (Table 3). Two factors, nadir albumin and nadir platelets, were significantly different between the groups (P = .016 and P = .029, respectively). Albumin was decreased in all nonsurvivors (median 1.25 g/dL; range 0.4–2.1 g/dL), while three of five survivors demonstrated only mild decreases (median 2.6 g/dL; range 1.7–3.4 g/dL). Although all patients became thrombocytopenic, nonsurvivors reliably developed severe thrombocytopenia (<64,000) while survivors developed a range of thrombocytopenia from mild to severe (6000–144,000).

TABLE 3 Medians and Ranges for Continuous Data of Survivor and Nonsurvivor Groups

View Fullscreen

Abdominal ultrasound was performed in 12 patients. Findings included abdominal effusion (41%), thickened gallbladder wall (25%), hyperechoic liver (16%), hypoechoic liver (8%), microhepatica (8%), and ileus (8%). Three dogs (25%) had unremarkable ultrasound studies. Abdominal radiographs were performed in eight patients; seven were unremarkable, and one dog had decreased abdominal detail.

Twelve patients (86%) were admitted to the hospital and two (14%) were treated as outpatients. One outpatient presented 3 wk post ingestion after eating two seeds, while the second presented 10 days post ingestion, having only eaten a leafy portion of the plant. Both of these patients were counted among the five survivors.

All 12 inpatients were placed on IV crystalloid fluid therapy, and six (50%) patients were administered intravenous colloid. Two (14%) patients required dopamine to treat persistent hypotension. Dextrose was supplemented in nine (64%) patients. In patients with coagulation abnormalities and hemorrhage, five (36%) received plasma transfusion(s) and three (21%) received whole blood transfusion(s), while vitamin K1 therapy was initiated in nine (64%) patients. Thirteen of 14 (93%) patients received some combination of antibiotics, including ampicillin, amoxicillin, enrofloxacin, metronidazole, cefazolin, tylosin, neomycin, and/or erythromycin. Various gastroprotectants were administered to all patients and included sucralfate, omeprazole, pantoprazole, ranitidine, famotidine, cimetidine, and/or barium. Antiemetics were used in 11 (79%) patients and included metoclopramide, dolasetron, ondansetron, chlorpromazine, and maropitant. Cyproheptadine and mirtazapine were used in two (14%) patients to stimulate appetite. Pain management included tramadol, buprenorphine, or butorphanol. Eight (57%) patients required nutritional support in the form of partial or total parenteral nutrition, nasogastric tube enteral feedings, and/or oral syringe feedings. One (7%) patient received both phenobarbital and diazepam as anticonvulsants. Six (43%) patients were administered N-acetylcysteine. Abdominal effusion was treated in five (36%) patients with furosemide or spironolactone. Three (21%) patients were placed on lactulose, and four (28%) patients were administered activated charcoal. Steroid therapy was initiated in four (29%) patients. Liver supportive agents were administered to seven (50%)patients, which included ursodeoxycholic acid, S-adenosylmethionine, silybin, and vitamin E. There was no difference in treatment between survivors and nonsurvivors.

Nine dogs (64%) died or were euthanized as a direct result of cycad palm intoxication: four arrested, three were euthanized during hospitalization, and two were euthanized shortly after discharge. Of the survivors, two dogs (14%) had complete resolution of clinical signs and laboratory abnormalities 2 and 3.5 mo later, respectively. Two of the five survivors were clinically normal but continued to have mildly elevated liver enzymes and were placed on liver supplements and supportive diets. One of these survivors developed ammonium acid urate cystic calculi (Patient 12). The last survivor (Patient 13) developed chronic liver failure and ammonium biurate crystalluria. A laparoscopic liver biopsy was consistent with acute hepatic insult characterized by periportal and random neutrophils, lymphoplasmacytic and histiocytic inflammation with rare individual hepatocellular necrosis, and segmental congestion and hemorrhage. After 5 mo of supportive care, the patient improved clinically and only had a mildly elevated ALP (132 U/L).

A survivor who continued to have elevated liver enzymes 8 yr post ingestion was humanely euthanized after developing acute liver failure following an anesthetic procedure. Liver histopathology revealed marked, widespread, chronic cholangiohepatitis with biliary hyperplasia, random multifocal acute hepatic necrosis, and suppurative hepatitis.

Discussion

The purpose of this retrospective study was to evaluate the manifestation and outcome of cycad palm toxicosis in dogs and to identify any significant findings that may differentiate survivors from nonsurvivors. When comparing the outcome of the patients in the current study to two previously published retrospective studies, the case fatality rate is higher (67%) than that reported by Ferguson et al. (50%), and both are substantially higher than that reported by the NAPCC toxicology consultation center by Albretsen et al. (32.1%).^2,4 In the NAPCC study, data were included from patients who were presumed to have, but had not definitively, consumed cycad, potentially deflating the actual case fatality rate. Time to presentation at GCVS may explain the higher mortality rate in the current study when compared to Ferguson et al. Ninety-one percent of patients treated in the prior study had presented within 36 hr of clinical signs or ingestion, whereas only seven dogs (50%) had presented to GCVS within 48 hr. The resultant delay in treatment may have increased the likelihood of a poor outcome.

Although all parts of cycad palms are considered toxic, the seeds are reported to have the highest concentration of toxins.⁹ With regards to the dogs in this study, 12 patients ingested the seed and two patients ingested other portions of the plant. Of the survivors, two ate the leafy part of the plant, one ate half of a seed, one ate a whole seed, and the last survivor ate two seeds. It is possible ingesting the leafy part of the plant or a small amount of seeds exposed survivors to a relatively small amount of toxin, which naturally improved their chances of survival.

In all cases, the offending plants were located on the property where the patients resided, as most clients were not aware of its toxicity. A majority of the exposures occurred during the fall months, which correlates with the most comfortable time of year in southeastern Texas for dogs to be outside. More importantly, this coincides with the time of year when cycad seeds mature on the female plant during its natural reproductive cycle.¹¹

There was no difference between survivors and nonsurvivors regarding physical examination findings, treatments, clinical signs, and weight. Clinical signs were similar to those previously reported and were primarily gastrointestinal and neurologic.^2–5 Decontamination via emesis occurred in four patients, but these patients continued to display evidence of toxicity, demonstrating the rapid onset of toxicosis post-exposure. After vomiting the ingested cycad, one patient underwent endoscopy and colonoscopy in an attempt to retrieve any remnants of the plant. Although no additional material was found, the patient developed acute liver failure and was humanely euthanized. It is important to note that decontamination may help diminish the severity of toxicity, but it will not prevent liver insult or clinical disease.

Treatment focused on supportive care of gastroenteritis, acute liver failure, and disseminated intravascular coagulation (DIC). Activated charcoal, which was previously documented to have a protective effect, was administered to only four patients, of which only two survived.² Due to the small sample size we were not able to assess the effects of activated charcoal on patient outcome. Six patients were administered N-acetylcysteine for its hepatoprotective properties. Of these dogs, only two survived, despite early administration. For the same reason, it is equally difficult to evaluate the usefulness of this drug in cycad toxicosis.

Because the majority of dogs (13 of 14) were initially evaluated by a primary care or emergency veterinarian, and the time from exposure to referral varied, peak and nadir laboratory values during hospitalization at GCVS were evaluated, as well as the laboratory data from initial presentation at the referring facility. In addition, toxins, by nature, can have a “hit and run” effect on the body, causing delayed changes in laboratory values. Despite the fact that values may be affected within hours post-ingestion, evaluating data collected relatively soon after exposure may not provide a complete clinical picture of the extensive effects of cycad palm toxicosis.

Two variables were found to be significant when evaluating differences between survivors and nonsurvivors: nadir albumin levels and nadir platelet counts. These differences may allow for some insight into survival and prognosis. The nonsurvivor group was found to be moderately to severely hypoalbuminemic, while the survivors were found to have normal or only mild decreases in albumin measurements at their nadirs. Ferguson et al. presented similar findings, as the nadir albumin concentration was found to be lower in nonsurvivors compared to survivors.² Hypoalbuminemia may be secondary to decreased production due to acute liver failure, loss via the gastrointestinal tract or urinary system, or vasculitis.¹²

Nadir platelet counts were found to be different between survivors and nonsurvivors. Although all patients had some degree of thrombocytopenia, the nonsurvivors had significantly lower nadir platelet counts, seven of nine having nadir platelet counts of 40,000 or less. Only one survivor experienced such severe thrombocytopenia. Thrombocytopenia may be a consequence of DIC and consumption due to gastrointestinal hemorrhage.

Although hypoalbuminemia and thrombocytopenia are common clinicopathologic manifestations of cycad palm toxicosis, there does not seem to be a predictable time frame within which nadir values occur. Nadir albumin was measured at a median of 14 days post-ingestion, but over a range of 2 to 61 days. Likewise, nadir platelet counts were measured at a median of 8 days post-ingestion, but over a range of 1 to 48 days.

Of the initial laboratory abnormalities, elevations in ALT, ALP, and total bilirubin, hypoglycemia, and thrombocytopenia were most pronounced. These findings are similar to what had previously been reported.^2–5 Although significant differences between survivors and nonsurvivors with respect to these laboratory values were not noted in the present study, Ferguson et al. did find significant differences in ALT on presentation between the two groups. The relevance of this finding was questioned by the authors, however, as ALT values were considered mild to moderately increased for both groups.² In the present study, ALT was elevated in 79% of the patients on initial presentation, and 66% of these had elevations noted within hours post-ingestion, suggesting rapid hepatocellular damage. With serial measurements, 100% of patients had elevations in ALT within 72 hr of presentation to GCVS. Likewise, 50% of patients were thrombocytopenic at initial presentation, but all patients eventually became thrombocytopenic, and 13 of 14 patients had significant thrombocytopenia within 4 days. In two previous case reports in which serial blood work was evaluated, all patients involved had similar changes in ALT and platelet counts.^3,5 This knowledge may prove useful when a patient has gastroenteritis and cycad palm ingestion is suspected. Should serial measurements of ALT and platelet counts remain normal, the likelihood of ingestion diminishes.

Hyperbilirubinemia was noted in only 36% of patients on initial presentation, and bilirubin levels were normal in nine of 10 (90%) patients that presented within 24 hr post-exposure. However, almost all patients (11 of 12; 92%) developed hyperbilirubinemia during hospitalization at GCVS. The increases in bilirubin are due to ongoing hepatocellular damage resulting in cholestasis and failure to adequately uptake, conjugate, and excrete bilirubin.¹²

No significant differences were noted between survivors and nonsurvivors with respect to blood glucose levels, either at initial presentation or during hospitalization at GCVS. Of the patients that were hypoglycemic on initial presentation, none survived. These patients had blood glucose levels lower than 65 mg/dL, and all presented within 48 hr post-exposure, indicating rapid, early progression of disease. While patients were hospitalized at GCVS, maintaining normal blood glucose was a major therapeutic goal and was achieved using dextrose supplementation in seven of 14 patients (50%). Therefore, nadir blood glucose levels during hospitalization were artificially elevated and statistically biased.

It is important to recognize that the cause of hypoglycemia and acute coagulopathy is likely multifactorial, and that they are manifestations of DIC in combination with liver failure from hepatic necrosis. The liver plays a major role in glucose metabolism and coagulation homeostasis. It stores glycogen and is the site of gluconeogenesis. Naturally, hepatic failure would impair normal glucose metabolism, resulting in hypoglycemia. The liver synthesizes coagulation factors and inhibitors, activates vitamin K-dependent factors, and clears activated coagulation products. In cases of widespread hepatic necrosis, the liver becomes the source of consumption of coagulation factors and, as a result of synthetic failure, cannot compensate for this consumption. This leads to a range of coagulation factor deficiencies and potentially DIC.

Histopathologic findings in the patients who had liver biopsies agree with changes noted in previous publications.^2,5,13,14 Patient 13 had a laparoscopic biopsy performed 3 wk post-ingestion, and the results were consistent with acute hepatic insult characterized by inflammation, hepatocellular necrosis, congestion, and hemorrhage. The other patient had liver histopathology performed years post-ingestion, and results were consistent with a chronic process, with widespread inflammation, biliary hyperplasia, necrosis, and fibrosis being the major findings.

There appears to be two phases of cycad palm toxicity: an acute phase directly related to cycad toxins, characterized by hepatobiliary insult, neurologic dysfunction, and gastrointestinal disturbances, followed by a chronic phase secondary to widespread liver damage and the lack of functioning liver mass, evidenced by liver enzyme elevation/dysfunction and/or chronic liver failure seen in one of the survivors (Patient 13). When evaluating the survivors, two dogs fully recovered after the acute phase of the disease, both having normal liver enzymes on reevaluation. Yet the other three survivors continued to have residual liver damage, evidenced by chronic liver enzyme elevation. In addition, two of these survivors developed urinary complications associated with liver dysfunction, with one patient developing ammonium acid urate cystic calculi (Patient 12) and another developing ammonium biurate crystalluria (Patient 13). This pattern of acute hepatic insult followed by chronic sequelae is not an uncommon finding with hepatotoxins and has been reported with lomustine, phenobarbital, and aflatoxins.^15–17 It is interesting to note that a survivor (Patient 13) developed and subsequently recovered from chronic liver failure. Even in the event of severe hepatotoxicity from cycad palm ingestion, recovery is possible.

Limitations of the present study are those inherent to its retrospective nature as well as the small sample size. In addition, the laboratory values obtained at the initial veterinary facilities were performed on a variety of in-house analyzers. Although reference ranges of the different analyzers were nearly identical, quality control and user experience cannot be assessed. The range of time from ingestion to initial presentation makes evaluation of initial laboratory values challenging. Patients 2 and 13, who presented 2 and 4 wk post-exposure, respectively, could have been manifesting the chronic sequelae of toxicosis. Additionally, cases evaluated were almost entirely referred, and it is possible that the patients represent dogs that were more severely affected by the toxins. The quantity of cycad palm each patient ingested was also unknown. This information could have been useful in estimating a toxic dose and its association with mortality.

Conclusion

In conclusion, cycad palm toxicosis is associated with severe gastroenteritis, neurologic dysfunction, and hepatopathy that may result in acute and/or chronic liver failure or residual liver enzyme elevation. At this time, it is unknown why the responses of individual dogs vary from case to case, but it may be associated to the quantity of toxin ingested. Changes in clinicopathologic variables, such as nadir albumin and nadir platelets, may help differentiate survivors from nonsurvivors.