Ivermectin Toxicosis in Dogs: A Retrospective Study

Introduction

Ivermectin is a macrolide antibiotic commonly used in the following ways: as a heartworm preventative in dogs and cats; as an anthelmenthic in ruminants, swine, and horses; and as an ear miticide in cats. Ivermectin is available in a variety of formulations including liquids, pastes, and tablets. At the dosage range of 6 to 24 μg/kg approved for heartworm prevention, ¹ ivermectin has a wide margin of safety. In dogs, toxicosis may occur when ivermectin is inadvertently overdosed by the owner or veterinarian; when dogs ingest large amounts of formulations intended for heartworm prevention; or when dogs are exposed to products intended for large animals. Even small amounts of these more concentrated ivermectin products may cause serious clinical effects in dogs.

Subpopulations of collies, border collies, Australian shepherds, Shetland sheepdogs, and mixes of these breeds have a unique sensitivity to ivermectin, milbemycin, moxidectin, loperamide, and other drugs. This is because these dogs have a defect in p-glycoprotein, an important component of the blood-brain barrier.² The defective p-glycoprotein results in accumulation of relatively high concentrations of these drug substrates in the central nervous system (CNS), even when relatively low doses of drug are administered. In situations of ivermectin overdosage, large amounts of the drug may persist in the CNS when the capacity for removal by p-glycoprotein is exceeded. Excessive accumulation of ivermectin in the CNS occurs at much lower doses in a p-glycoprotein-defective animal.

Signs of ivermectin toxicosis in dogs include lethargy, hypersalivation, ataxia, coma, mydriasis, clinical blindness, and seizures. These signs may progress to respiratory depression and death if no therapeutic intervention takes place.

Previous reports have indicated that many dogs (e.g., beagles and research animals) can tolerate ivermectin dosages of 2.50 mg/kg,³ while collies may develop serious ivermectin toxicosis at 0.20 mg/kg.⁴ In one reported case, a Doberman pinscher exhibited recumbency, tremors, seizures, and blindness at a dosage of approximately 3.50 mg/kg.⁵ The purposes of this study were to summarize the clinical signs that developed in dogs exposed to different dosages of ivermectin; to characterize the dosages at which clinical signs occurred according to breed; and to evaluate the breed’s possibility of carrying p-glycoprotein gene deletions using data reported to the American Society for the Prevention of Cruelty to Animals (ASPCA)Animal Poison Control Center (APCC).

Materials and Methods

Epidemiological and clinical information was obtained from the computerized record database of the APCC. The APCC is a 24-hour consultation service that receives calls from throughout the United States and Canada concerning incidents of animal poisoning from animal owners and practicing veterinarians. The APCC staff member collects the following information on each animal: species, breed, sex, age, and body weight; number of animals exposed and at risk; number of affected animals; source of exposure; amount of product ingested; and assurance of exposure (whether observed or evidenced). This information is entered into the AnTox database. Information is also obtained regarding time of onset, type and duration of clinical signs, serum biochemical alterations, pathological findings, and response to treatment. Follow-up calls are made as deemed necessary to update records as to progression of clinical signs, response to treatment, and final outcome. Information received is generally not confirmed by analytical methods.

On the basis of exposure history, clinical information available, expected clinical signs of toxicosis as described in the literature, amount ingested, and previous experience dealing with the agent, the APCC veterinarian categorizes each incident. Categories are toxicosis, suspected toxicosis, possible toxicosis, doubtful toxicosis, or exposure only (i.e., when no signs were present at the time of the call, and none were reported on follow-up). Incidents are categorized as toxicosis if all temporal, clinical, and historical data are consistent with the expected syndrome; as suspected toxicosis if clinical signs are characteristic of the expected syndrome, but some data are not available; as possible toxicosis if only a few signs are consistent with the expected syndrome; and as doubtful toxicosis if clinical signs and exposure history are not consistent with the expected syndrome. The APCC database was searched for records of dogs ingesting ivermectin between January 1998 and April 2005. Only cases involving reported single oral exposures were used in this report.

For this study, the authors divided cases into those involving dog breeds that are not generally considered to carry p-glycoprotein defects (NPD) and breeds that may genetically have p-glycoprotein defects (PD).⁶ Cases were examined separately. A PD breed was defined as a purebred or mix of the following breeds: collie, border collie, Australian shepherd, and Shetland sheepdog. The NPD breed distribution and frequency were evaluated using chi-square analysis compared to the average number of animals representing these breeds reported for all exposures in the AnTox database.

Results

Fifty-five cases of ivermectin toxicosis were identified in PD dogs, and 117 cases were identified in NPD dogs. In all of these cases, the dogs were exposed to ivermectin orally, no other agents were known or suspected to be involved, and the cases were categorized as toxicosis or suspected toxicosis by the APCC veterinarian. In 66% of these cases, the agent was administered by the owner or a veterinarian either as a therapeutic dose, as an off-label dose, or as an overdose. In the remaining 34% of the cases, ivermectin ingestion was accidental.

Of the 55 PD dogs, 53 were reported as pure-bred, and two were reported to be mixed-breed dogs. Of the 53 purebred PD dogs, 24 (45%) were collies (smooth and rough); nine (16%) were border collies; three (6%) were Shetland sheepdogs; and 17 (32%) were Australian shepherds.

Of the 117 NPD dogs, the breeds most commonly involved were German shepherd dogs (n=23; 20%), Labrador retrievers (n=8; 7%), grey-hounds (n=7; 6%), Chihuahuas (n=6; 5%), Pit bull terriers (n=6; 5%), and mixed-breed dogs (n=11; 9%) [Table 1].

Overall in the APCC AnTox database, Labrador retrievers are involved in 15.7% of all canine cases of exposure to potentially toxic agents. However, in this study, Labrador retrievers were involved in only 4.5% of the total cases and represented only 7% of the dogs in the NPD group. German shepherd dogs, Pit bull terriers, and greyhounds accounted for more ivermectin cases when compared to the average of these breeds for all calls to the APCC. Of 172 ivermectin cases, German shepherd dogs were involved in 23 (13%) cases; Pit bull terriers and Chihuahuas were involved in six (3.5%) cases each; and greyhounds were involved in seven (4%) cases.

Overall at the APCC, German shepherd dogs are involved in 4.4% of canine cases, Chihuahuas are involved in 2.7%, Pit bull terriers are involved in 2.1%, and greyhounds are involved in <1% of canine cases. Chi-square analysis of frequency of ivermectin cases involving these breeds versus the expected frequency revealed that the larger number of greyhound and German shepherd dog cases, as well as the smaller number of Labrador retriever cases, are statistically significant with P<0.05 [Table 2].

Ages of the dogs ranged from 1.5 weeks to 15 years, with a mean age of 3.1 years and a median age of 1.5 years. In the PD dogs, ages ranged from 9 weeks to 14 years, with a mean age of 4.6 years and a median age of 3 years.

The amount of drug ingested was known exactly or could be estimated in 31 of the 55 PD dogs, and the dosages ranged from 0.08 mg/kg to 5 mg/kg. In NPD dogs, the amount ingested could be estimated in 72 of the 117 cases, and the dosages ranged from 0.20 mg/kg to 213 mg/kg. Lowest dosages that caused signs of ataxia and disorientation were 0.08 mg/kg in a PD dog and 0.20 mg/kg in 10 different NPD dogs.

The most frequent clinical signs reported to the APCC in both groups included ataxia, lethargy, mydriasis, hypersalivation, tremors, and recumbency [Table 3]. Ataxia was the most common sign reported in 105 (61%) of 172 symptomatic dogs, followed by lethargy (n=52; 30%), tremors (n=51; 30%), mydriasis (n=38; 22%), and blindness (n=38; 22%). Some signs were reported with similar frequency in both NPD and PD dogs, including ataxia (57% in NPD dogs; 62% in PD dogs), tremors (31% in NPD dogs; 20% in PD dogs), coma (15% in NPD dogs; 20% in PD dogs), and lethargy (31% in NPD dogs; 20% in PD dogs). Other signs were reported with very different frequencies between the two groups. These signs included recumbency (9% in NPD dogs; 27% in PD dogs) and seizures (4% in NPD dogs; 20% in PD dogs). Ataxia was reported in a broad range of dosages, from as low as 0.08 mg/kg in PD dogs up to dosages of 213 mg/kg in NPD dogs. Lethargy, tremors, and mydriasis were also seen over a similarly broad range of dosages. Death and euthanasia occurred most frequently at dosages >1 mg/kg in both NPD and PD dogs [Table 3]. Rarer signs that were reported in the 172 ivermectin cases were miosis (n=8; 5%), disorientation (n=14; 8%), and panting (n=7; 4%).

Death occurred in three PD dogs, all at unknown dosages after exhibiting signs of ataxia, mydriasis, and tremors. Two dogs in the PD group were reported to have been euthanized. One of these dogs became comatose after ingesting 0.15 mg/kg of ivermectin. The dog subsequently aspirated activated charcoal and was euthanized. The other PD dog that was euthanized ingested 2.5 mg/kg of ivermectin and developed ataxia, tremors, and bradycardia. When the dog became comatose, the owners elected euthanasia. Of the NPD dogs, death occurred at dosages between 1 and 2.5 mg/kg in two dogs and at unknown dosages in two dogs; all of these dogs showed signs of CNS depression, ataxia, mydriasis, and tremors or disorientation. Seven of the NPD dogs were reported to have been euthanized. Three of these dogs were puppies of about 4 weeks of age that had been exposed to dosages of about 6.25 mg/kg. The remaining four dogs were adults exposed to dosages ranging from 10 to 213 mg/kg; one dosage was unknown.

Discussion

Ivermectin binds to receptors for gamma-aminobutyric acid (GABA), an inhibitory neurotransmitter, and acts as a GABA agonist, enhancing inhibition.⁷ Binding of ivermectin to the neuronal membrane also potentiates the presynaptic release of GABAand enhances the binding of GABA to postsynaptic receptors.³ Gamma-aminobutyric acid opens chloride channels and allows an influx of chloride ions that cause an inhibitory effect through membrane hyperpolarization. Receptors for GABA are widespread in invertebrate parasites, and treatment with ivermectin leads to paralysis and death of these target species. In mammals, GABA receptors are confined to the CNS.⁸ At therapeutic doses in normal dogs, ivermectin is excluded from the CNS by p-glycoprotein, a component of the blood-brain barrier that removes various xenobiotics from the CNS.⁸

Some dogs carry a genetically determined defect in p-glycoprotein that allows accumulation of ivermectin (and a variety of other drugs) within the CNS; this defect has been identified with increased frequency in several breeds of dogs. This p-glycoprotein defect is an autosomal recessive trait in dogs. A study of collies in the northwestern United States demonstrated that 33% of client-owned dogs tested were homozygous for the defect, and 14% were carriers.⁹ In gene samples from 4000 dogs, the homozygous deletion was determined to exist in 1.7% of 178 Australian shepherds tested, in 31% of 263 collies, in 1.1% of 190 Shetland sheepdogs, and in 0% of 151 Old English sheepdogs.⁶ Thirty percent of Australian shepherds, 46.8% of collies, 14.7% of Shetland sheepdogs, and 7.3% of Old English sheepdogs tested were heterozygous for the trait.⁶

Ivermectin dosages as low as 0.15 to 0.34 mg/kg have been reported to cause serious clinical signs in breeds considered to be p-glycoprotein deficient.^4,10–11 In the authors’ study, one PD dog showed clinical signs at a dosage of <0.10 mg/kg, and 17 dogs exhibited signs at dosages ranging from 0.10 to 0.40 mg/kg; these findings indicated that PD dogs could show clinical signs at dosages lower than previously reported. However, since only one dog showed signs at <0.1 mg/kg, this may be a rare occurrence.

In NPD dogs, 3.50 mg/kg was the lowest dosage reported in the literature to cause severe clinical signs.⁵ Dosages of 2.50 mg/kg have been reported to cause mydriasis and tremors, while 10 mg/kg caused tremors and ataxia in experimentally dosed dogs.³ In the current study, 11 of the NPD dogs showed signs of serious toxicosis at dosages of 0.10 to 0.40 mg/kg. In addition, 20 other dogs showed signs of toxicosis at <2.50 mg/kg. This information supports the possibility that some dogs considered as NPD breeds may be at risk for clinical signs at dosages much lower than previously reported.

Explanations for this apparent sensitivity of some members of NPD breeds to ivermectin might include conditions that could cause a disruption in the blood-brain barrier, such as infectious or inflammatory disease, prior cranial trauma, or neoplastic disease. However, in this study, none of these conditions were reported or suspected by the pet owners or veterinarians. Very young animals also may have an immature blood-brain barrier.¹² Of the 31 NPD dogs showing signs at relatively low doses, 11 were puppies between 12 days and 12 weeks of age. These puppies may have had an immature blood-brain barrier, predisposing them to toxicosis. Nineteen other dogs were adults (aged 6 months to 15 years) and would not be expected to have any defect in the blood-brain barrier due to immaturity (in one case, the age was not reported). Humans are reported to have a decrease in p-glycoprotein function¹³ as they age; this may be the case in dogs as well. Three of these animals were older (aged 11, 13, and 15 years) and may have had an increased susceptibility due to age-related decreased p-glycoprotein function.

Some of the NPD dogs may have had a p-glycoprotein defect similar to the PD dogs. Although many breeds represented in this paper were not found to have the same gene deletion as collies,^6,14 the defect could be much rarer in these breeds and may not have been detected in the population studied by Neff. Other genes might code for a different p-glycoprotein defect that has not yet been identified. Other mechanisms not involving p-glycoprotein also may cause some dogs to be more sensitive to ivermectin.

In this report, German shepherd dogs and greyhounds were found to be involved in ivermectin cases significantly more frequently than breed averages for all exposures reported to the APCC, and Labrador retrievers were involved significantly less frequently than expected. Greyhounds and German shepherd dogs may be overrepresented in this study because of the inclusion of a single case involving a litter of six greyhound puppies administered ivermectin by the owner and a single case of a breeding kennel that exposed eight German shepherd dogs (it was unknown if these animals were related). However, even when removing those eight dogs from the calculations, German shepherd dogs are still statistically significantly overrepresented.

One explanation for the difference in breed distribution may be the fact that in a majority (66%) of these cases, the agent was administered at incorrect dosages by an owner or veterinarian, and the exposure did not occur through the animals’ curiosity. This may be the reason Labrador retrievers are underrepresented. Another possibility is that p-glycoprotein defects are present in some dogs in the population. White German shepherd dogs have been documented to have the p-glycoprotein defect.¹⁴ The ASPCA APCC does not routinely record animal coloring, so it is unknown if the German shepherd dogs in these cases were white. The dosages in the German shepherd dogs ranged from 0.2 to 8.27 mg/kg; three cases were exposed to an unknown amount. Of these cases, 10 dogs were dosed <2.5 mg/kg, a dosage that has not previously been reported to cause serious clinical signs in NPD dogs. These findings suggest that German shepherd dogs and greyhounds may be more susceptible than most breeds to ivermectin and other drugs that are substrates of p-glycoprotein. Also, the related greyhounds and perhaps the related German shepherd dogs may have had the same p-glycoprotein deficit, which may not be present in the general population as frequently.

Conclusion

It is prudent to use caution when treating a PD dog with ivermectin or other drugs that are p-glycoprotein substrates, such as milbemycin, moxidectin, loperamide, and some chemotherapeutics. Ivermectin is used off-label at dosages of 0.4 to 0.6 mg/kg per day to treat demodectic mange,¹ and it is important for the clinician to be aware that toxicosis could take place following single or multiple doses, even in breeds not traditionally considered to be sensitive to ivermectin. Owners whose dogs are receiving off-label dosages of ivermectin should be educated on initial signs of toxicosis (e.g., mydriasis, lethargy, ataxia) and be advised to discontinue medications if these effects develop. Testing dogs prior to administering these higher doses of ivermectin is also prudent. Testing for the MDR1-1Δ. genotype can now be done at Washington State University.¹⁴ At dosages of ivermectin used for heartworm prevention (0.006 mg/kg), clinical signs would not be expected, even in a PD dog.

In cases reported to the APCC, 16 (9%) of 172 symptomatic animals died or were euthanized. Thirteen deaths occurred at dosages >5 mg/kg or unknown amounts. This indicates that with proper treatment, the prognosis for NPD dogs exposed to <5 mg/kg can be considered good. At dosages >5 mg/kg, the prognosis is guarded for cases of exposure in any breed of dog. General supportive care is indicated for these dogs, including multiple doses of activated charcoal (1 to 2 g/kg q 8 hours for six doses) to interrupt the enterohepatic recirculation that the drug undergoes and thereby reduce the half-life.

Overall, ivermectin is a safe drug when used at dosages appropriate for heartworm prevention. Caution should be used in all breeds when utilizing off-label dosages for endoparasitism or ectoparasitism, as some animals may become symptomatic at dosages previously not considered to cause clinical problems. This study also suggests that caution should be used when treating German shepherd dogs and greyhounds at dosages exceeding heartworm preventative dosages, as some dogs of these breeds may have an increased sensitivity to ivermectin and possibly other p-glycoprotein substrates.