Ivermectin-induced blindness treated with intravenous lipid therapy in a dog

Abstract

Objective

To report a case of blindness due to the ingestion of ivermectin and subsequent successful treatment with intravenous lipid (IVL) therapy.

Case Summary

A female neutered Jack Russell Terrier was examined for acute onset of apparent blindness after being exposed to ivermectin the previous day. The dog appeared to be blind during initial examination. Pupillary light reflex, menace response, and dazzle reflex were not present in either eye. Fundic examination revealed small areas of linear retinal edema. Electroretinography (ERG) showed diminished activity in both eyes. Ivermectin was present in the serum on toxicological assay. Approximately 20 hours after exposure, IVL was infused. Within 30 minutes of initiating the infusion, the pupillary light reflexes returned in both eyes, and by the end of the infusion the patient behaved as if sighted. Fundic examination and ERG were unchanged at this time. The dog was tested for the multidrug resistance gene mutation and was unaffected.

New or Unique Information Provided

Ivermectin toxicity occurs in dogs with apparent blindness being a common clinical sign. This is the first case report of ivermectin-induced blindness evaluated with ERG before and after treatment with IVL in a dog unaffected by the multidrug resistance gene mutation. Treatment with an infusion of IVL therapy appeared to shorten the clinical course of disease in this patient without affecting ERG results.

Abbreviations

ERG

electroretinography

IVL

intravenous lipid

GABA

gamma-aminobutyric acid

MDR-1

multidrug resistance gene

PLR

pupillary light reflex

Introduction

An 11-year-old 10 kg female neutered Jack Russell Terrier presented to the Small Animal Emergency Service at the William R. Pritchard Veterinary Medical Teaching Hospital of the University of California, Davis, for acute onset blindness. The owners reported that the evening prior to presentation the farrier was de-worming horses with a product containing 1.87% ivermectin1 with the dog present in the pen. No abnormalities were noted with the dog on the previous day. General physical examination findings on presentation were unremarkable. The initial ophthalmic examination revealed a patient who behaved as if unsighted; no menace response was present in either eye and the patient ran into obstacles under both photopic and scotopic conditions. Both pupils were mydriatic in ambient light and the pupillary light reflexes (PLRs) were absent.

A board certified ophthalmologist performed a complete ophthalmic examination approximately 2 hours after presentation to the emergency service, including slit lamp biomicroscopy, indirect ophthalmoscopy, and electroretinography (ERG).2 Both eyes were open and appeared comfortable. No PLRs (either direct or consensual), menace responses, nor dazzle reflexes were present in either eye. No abnormalities were detected in the anterior segment or vitreous humor of either eye.

Fundic examination revealed small areas of linear retinal edema ventral and medial to optic disk in the nontapetal area of both eyes (Figure 1). Following a 25-minute period of dark adaption, an ERG was performed without sedation. Retinal responses from each eye were assessed separately by use of a series of 8 red-light flashes (findings were averaged). The b-wave amplitude was 75.56 μv in the right eye and 33.89 μv in the left (normal retinal function for this unit is >100 μv). While the specific origin of the ERG b-wave is controversial, it is currently thought to be the result of retinal bipolar cell activity.[1, 2] In common practice, b-wave amplitude is considered the key indicator of retinal function.[1-3]

Figure 1. Retinal examination of dog with ivermectin-induced blindness. The image depicts areas of retinal edema, right eye.

Results of a CBC were unremarkable. Serum biochemical analyses showed no clinically significant abnormalities. Serum was analyzed for macrocyclic lactones via liquid chromatography-mass spectrometry with results returning the following day positive for ivermectin at 1,500 parts per billion, but negative for abamectin, moxidectin, and selamectin. Testing for ABCB1-1∆ or multi-drug resistance (MDR)-1 allele mutation later revealed that the dog was unaffected.

Approximately 20 hours after exposure and 3.5 hours after clinical signs were noted, an IV catheter was placed, and a 20% intralipid emulsion3 was administered (1.5 mL/kg over a 10-minute period, then 0.25 mL/kg/min for a 90-minute period). Thirty minutes after initiation of the infusion, slight PLRs and a positive dazzle reflex were noted. At the end of the infusion the patient appeared to be sighted navigating around obstacles and avoiding walls, had a positive menace response, but still had bilateral mydriasis in ambient light. The linear areas of retinal edema were also still present. The ERG was repeated and the b-wave amplitudes were similar to the previous recordings with 55.83 μv for the right eye and 39.44 μv for the left eye. The dog was discharged to the owner the same day and has remained sighted ever since.

Approximately 9 months after initial evaluation the dog was reexamined. General physical examination was unremarkable except for a healing wound on the right antebrachium. Ophthalmic examination revealed normal pupil size in ambient light, present and brisk direct and consensual PLRs, and a positive menace response and dazzle reflex in both eyes. There was no evidence of any intraocular inflammation and the areas of retinal edema had resolved (Figure 2). The patient easily navigated a maze under both photopic and scotopic conditions. An ERG was not repeated at this time due to equipment malfunction. The patient subsequently returned for an ERG 2 months later (approximately 11 months after the initial evaluation). The b-wave amplitudes had improved in both eyes to 86.39 μv in the right eye and 84.17 μv in the left.

Figure 2. Follow-up retinal examination of dog with ivermectin-induced blindness. Same area of right eye as depicted in Figure 1, 9 months after toxic event.

Discussion

Ivermectin toxicosis is well described in dogs.[4-10] A wide variety of clinical signs can occur including blindness, mydriasis, hypersalivation, ataxia, tremors, respiratory failure, obtundation, and death. Some intoxicated animals may show apparent blindness with or without the other generalized signs. The first reported retinal lesions with presumed ivermectin toxicosis were in 2 dogs in 1989[11] and have been subsequently reported in another dog with presumed ivermectin toxicosis[12] and in 2 dogs with confirmed ivermectin toxicosis.[13] Blindness due to suspected ivermectin ingestion has also been reported in a mule foal.[14]

Ivermectin, avermectin, milbemycin, and moxidectin are macrocyclic lactones and are commonly used in veterinary medicine. Gamma-aminobutyric acid (GABA) is an inhibitory amino acid neurotransmitter. GABA receptors, when activated by either endogenous GABA or ivermectin, cause cell membrane hyperpolarization via increased postysynaptic permeability to chloride ions. This effect is resolved with the application of picrotoxin, a GABA antagonist.[16] A second proposed mechanism for the action of ivermectin is by binding glutamate gated chloride channels which similarly causes inhibition of excitatory motor neurons.[16] In nematodes ivermectin results in rapid paralysis of movement and inability to feed due to pharyngeal muscle weakness resulting in death of the parasite.[15] Clinical signs of ivermectin toxicity in mammals are attributed to this enhancement of neuronal inhibition.[17, 18]

In mammals GABA mediated interneurons are largely present within the CNS, but may also be present on skeletal muscle and intestines.[19] P-glycoprotein (P-gp) is a large transmembrane protein encoded by the MDR-1 gene, and is responsible for limiting the penetration of ivermectin into the CNS. Toxicity in mammals requires a much larger dose as compared with that observed in nematodes.[15] The exceptions are dogs with a homozygous mutation in the ABCB1-1∆ allele that show increased sensitivity to ivermectin toxicosis.[20] In dogs without the ABCB1-1∆ polymorphism, clinical signs of toxicosis can occur at doses of 2.5 mg/kg.[21] In this Jack Russell Terrier who was negative for the ABCB1-1∆ polymorphism, the ingested dose was unknown.

The mechanism for ivermectin-induced blindness is not known at this time. Reports in dogs suggest that retinal pathology is involved in the process, but there may be a nerve and cortical component as well. Most cell types within the retina express GABA-ergic receptors[22] and GABA is considered the primary inhibitory neurotransmitter of the mammalian retina.[23] It is possible that if ivermectin crosses the blood-retinal barrier, neurons originating in the retina may be affected in a manner similar to the CNS.

This case report is of particular interest for a number of reasons. Because the dog had participated in a noninvasive clinical study about 6 months prior to the toxic event, the dog had received a complete ophthalmic examination by the same veterinary ophthalmologist who later examined this dog in connection with the toxic episode. The dog had been found to be free of any ophthalmic abnormalities at that time, although an ERG was not performed. Similar to 1 dog in a previous report,[13] and a mule foal[14] with suspected ivermectin-induced blindness, this patient did not exhibit an extinguished ERG reading even at presentation when she was blind and lacked PLRs. This would seem to eliminate the retina as the sole location for the changes in PLRs and vision loss and imply a multifactorial anatomic basis of this disease. Additionally, after the administration of intravenous lipid (IVL) therapy and subsequent return of vision, this patient did not demonstrate a significant change in fundic signs or ERG readings, and only a minimal improvement in PLRs. The fact that the dog had a return to vision without a significant change in PLRs would tend to rule out cranial nerve II as the location of the toxic effect as a lesion here would probably affect both PLRs and vision. Cranial nerve III dysfunction could cause PLR abnormalities without affecting vision. However, this would necessitate a lesion relatively distal in the course of the nerve, as more proximal lesions would also cause ptosis and strabismus, or ivermectin only affecting the superficial and medial layers of the nerve. Taken together, this would suggest that the blindness associated with ivermectin toxicity is not entirely retinal in origin and likely partially cortical in origin, with the PLR effects due to effects elsewhere.

Ivermectin-induced blindness occurred in 22% of suspected or diagnosed canine ivermectin toxicosis cases reported to the Animal Poison Control Center.[24] The dog in this report had exposure to ivermectin with simultaneous positive plasma levels for ivermectin confirming ingestion. Positive plasma levels may only be used to confirm ingestion and are less important than brain concentrations because the latter correlate more directly with the clinical consequences of intoxication.[25, 26] Current treatment recommendations for ivermectin toxicity are supportive in nature. Picrotoxin, a GABA antagonist, was reported to aid the recovery of 1 dog with ivermectin toxicity, but caused violent seizures and tachycardia and is not currently recommended.[27] Vision loss associated with ivermectin toxicity is temporary, and anecdotally, recovery is expected in 2–10 days[28] although the exact recovery time is unknown. As blindness can last many days, animals are typically discharged to owners to wait for vision to return which can create anxiety for owners and potentially the patient. In this case, IVL therapy was used, which was associated with a rapid return of vision in less than 90 minutes.

The treatment of ivermectin toxicosis with IVL has recently been described[29, 30] although there are anecdotal reports of its use for years. In one case of an ABCB1-1∆ unaffected Border Collie[31] that presented with both ophthalmologic and neurologic symptoms, the dog's dazzle response returned within 6 hours. The PLR was incomplete and visual ability improved over the next 24 hours. The return to vision appeared to take longer than the time course for the case reported here, but in both cases the dazzle response returned within hours of initiating IVL therapy.

This seemingly positive response to lipid therapy was not appreciated in 3 dogs that were homozygous for the ABCB1-1∆ gene mutation.[30] This patient and the Border Collie were negative for the mutation, yet both appeared to have a clinical response to IVL therapy, while 3 homozygous dogs did not. This raises the possibility that IVL therapy may only be helpful if the dog is not a homozygous mutant. Possible support for this is that P-gp is expressed in renal tubular epithelial and biliary canalicular cells,[31] which may reduce the excretion rates during ivermectin toxicosis in ABCB1-1∆ homozygous mutant dogs. P-gp also transports ivermectin from the brain into the blood. Without this ability, homozygous mutant dogs would rely solely on a concentration gradient for diffusion of ivermectin from the brain into the infused lipid. Without P-gp, the ivermectin partitioned in the circulating lipid could back diffuse into the brain, limiting the potential helpful effects of IVL therapy. In wild type dogs, ivermectin transport into circulating lipid may be an active process and back diffusion into the brain would be limited. Further investigations are needed to investigate the role of IVL therapy in ABCB1-1∆ homozygous mutant dogs.

Intravenous lipid has also been used in the treatment of moxidectin toxicosis in a puppy, which may have shortened the clinical course of disease,[32] and recently IVL therapy was used to treat successfully lidocaine toxicity in a cat.[33] In human medicine, IVL infusions have been used to treat toxicities for lipid soluble drugs such as local anesthetics,[34, 35] a severe case of organophosphate poisoning,[36] antidepressants,[37-40] and cardiac drugs.[41-43]

The mechanism by which lipids improve clinical signs of toxicity associated with the macrocyclic lactones is unclear. A proposed mechanism is through a “lipid sink” where ivermectin is removed from the CNS or retina by sequestering it into the plasma lipid phase. If a drug is lipophilic, theoretically this allows a higher drug concentration in the plasma with less available to be toxic in the tissues. A drug is deemed lipophilic if its log P is >1.0. This “lipid sink” theory is supported with increasing plasma levels of clomipramine,[37] mepivacaine,[44] bupropion,[39] and ivermectin[29] when toxicoses are treated with lipid emulsions. The log P for these drugs are 3.30, 1.89, 3.47, and 3.50, respectively.[45]

Although 20% intralipid solutions have been used in parenteral nutrition safely, there is no clinical evidence on safety of short-term large boluses of lipid solutions. Short-term infusions of soybean oil-based lipid emulsions have been shown to decrease neutrophil function in dogs.[46] Other potential adverse effects are pancreatitis, fat emboli, phlebitis, and hypersensitivity reactions. However, no adverse outcome related to IVL therapy for the treatment of intoxication has been reported except for laboratory difficulty in analyzing lipemic blood samples.[47]

The optimal dose of intralipid for IVL therapy is unknown. The dose of 1.5 mL/kg bolus then 0.25 mL/kg/min was chosen based on this being the most commonly cited dose in human case reports. A review of IVL therapy to treat local anesthetic toxicity in humans suggests a dose of 1.5 mL/kg bolus that can be repeated up to 3 times in cardiac arrests then 0.25–0.5 mL/kg/min.[48] On this basis it is reasonable to use this dose in veterinary medicine as well until further studies show an optimal dose.

The present case report shows a definitive diagnosis of ivermectin ingestion with acute onset of blindness that resolved temporally after a lipid infusion. The patient's fundic changes were not altered with lipid therapy, nor were the ERG results different, despite an apparent clinical improvement. The improvement observed in this patient however cannot definitively be attributed to the IVL therapy. Further assessment of plasma levels would have been useful to support a “lipid sink” theory, but as they do not necessarily correlation to brain levels or clinical signs, they were not performed. At this time IVL therapy appears to be a reasonable treatment option for ivermectin-induced blindness although its safety and efficacy in canine patients has not been fully determined at this time. Further studies on the benefit of lipid emulsions to treat lipid soluble toxicities such as ivermectin are indicated.

Footnotes

1. 1

   Duramectin Paste, Durvet, Blue Springs, MO.

2. 2

   RetinoGraphics, Inc., Norwalk, CT.

3. 3

   Intralipid, for Baxter Healthcare Corporation by Fresenius Kabi, Uppsala, Sweden.