Introduction

Bromethalin is a commonly used rodenticide that causes neurologic dysfunction in both target and non-target species. The Environmental Protection Agency has requested a phase out of second-generation anticoagulant rodenticides, and as a result, the use of bromethalin rodenticides has increased since 2011.¹ In 2016, bromethalin was the most common cause of calls to a national poison control hotline in 17 states.²

Once ingested, bromethalin is converted into the toxic metabolite desmethylbromethalin by hepatic mixed function oxygenases.^3–5 The toxic mechanism of action is a decoupling of oxidative phosphorylation in mitochondria, resulting in a decrease in available adenosine triphosphate (ATP). Without ATP, the sodium-potassium-ATP pumps on the cell membrane are unable to function, resulting in a buildup of intracellular sodium. Increased intracellular sodium results in cytotoxic edema by causing an osmotic gradient and the subsequent movement of water into cells.^3–7 In addition, bromethalin is thought to induce cerebral lipid peroxidation, which results in damage to cell membranes and organelles. This cellular injury may disrupt the blood–brain barrier and thus cause vasogenic edema.^3,4,8 Cerebral edema formation results in increased intracranial pressure and neuronal damage. An evaluation of bromethalin toxicosis in a rodent model showed that the use of hyperosmolar agents decreased intracranial pressure.⁵

Clinical signs associated with bromethalin toxicosis can occur with exposures as low as 10% of the median lethal dose (LD50).⁴ Signs of toxicosis include tremors, ataxia, paraparesis, tetraparesis, seizures, patellar hyperreflexia, paddling, hyperexcitability, dysphonia, forelimb extensor rigidity, dysuria, absent menace, opisthotonos, and death.^3–7,9–11 These neurologic sequelae can be divided into two syndromes.³ A convulsant syndrome that can manifest as hyperexcitability, hyperesthesia, tremors, seizures, vocalization, central nervous system depression, and death can be seen with doses above the LD50, which is reported to be between 2.38–5.6 mg/kg in dogs.^3–5,7 At doses less than the LD50, a paralytic syndrome composed of ataxia, paresis or paralysis, and central nervous system depression can be seen.³ Once severe signs of toxicosis such as seizures, coma, or paralysis are present, the prognosis is considered to be poor to grave.^3,4 Onset of clinical signs with high-dose exposure typically occurs within 10 hr, with a range of 2–18 hr postingestion, whereas clinical signs may take between 1 and 7 days to manifest in animals who have ingested a dose below the LD50.^3–5,10 Death may occur within 15–63 hr after exposure.¹⁰ In patients who survive, clinical signs may be present for 1–2 wk.^3,4

The purpose of this report is to describe the successful management of severe symptomatic bromethalin toxicosis in a dog with a combination of intravenous lipid emulsion (ILE), hyperosmolar agents, ginkgo biloba, and supportive care.

Case Report

A 9 yr old, 5.1 kg, castrated male Norwich terrier presented to the Matthew J. Ryan Veterinary Hospital at the University of Pennsylvania as a referral for acute onset circling, seizures, obtunded mentation, and suspected bromethalin ingestion. The day prior to presentation, the patient had been outside unsupervised for ∼30 min. Several hours later, he appeared more lethargic, ate normally that evening, but vomited several hours after eating. Overnight, the patient began circling, pacing, and head pressing. By the following morning, the patient had become nonambulatory tetraparetic and developed generalized tremoring. That morning, the patient also defecated green feces. The clients had no known toxicants on their property. They spoke with their neighbor, who reported that he had recently scattered a bromethalin-based rodenticide around his property.

The patient was then presented to his primary care veterinarian, where a complete blood count, serum biochemistry panel, and a 4DX test^a for tick-borne diseases and heartworm disease were performed and were unremarkable. The patient had three generalized tonic-clonic seizures while at the clinic, which were each treated with 2 mg midazolam (0.4 mg/kg) IV. He was also given two 16 mg phenobarbital doses (3.1 mg/kg) IV to begin a phenobarbital load, and then referred for further care.

On presentation to our hospital, the patient was normothermic (100.0°F) and bradycardic with a heart rate of 35–50 bpm. The respiratory rate was 30 breaths/min, and there were normal bronchovesicular sounds in all lung fields. Abdominal palpation was unremarkable. The neurologic examination was consistent with diffuse intracranial disease. The patient was obtunded and responsive only to auditory stimuli. No menace response was present bilaterally, and the palpebral reflex was intact. The pupils were equal and midrange, with a slow-to-absent pupillary light response. The patient was nonambulatory tetraparetic, with a lack of proprioception in all four limbs as well as normal patellar and withdrawal reflexes. The bradycardia was confirmed to be a sinus bradycardia on an electrocardiogram. Systolic blood pressure, as measured by Doppler^b, was 100 mm Hg. An SpO₂ was mildly decreased at 93–94%. Venous blood gas, electrolytes, packed cell volume, and plasma total solids were normal.

A 20-gauge IV catheter was placed in the right cephalic vein. The patient was initially treated with one dose of hypertonic saline (HTS)^c combined with 6% hydroxyethyl starch^d (4.5 mL 23.4% HTS combined with 10.5 mL 6% hydroxyethyl starch^d to make a 7% solution, total dose 3 mL/kg IV). The head was elevated ∼15°, and oxygen was given via mask with an estimated FiO₂ of 0.30. After the administration of HTS, the patient remained dull but regained menace response. The patient was then given a dose of 20% sterile lipid solution^e (1.5 mL/kg IV followed by a constant-rate infusion of 0.25 mL/kg/min IV over 60 min). Because of its lipophilicity, phenobarbital was discontinued and levetiracetam^f (20 mg/kg IV q 8 hr) was started as an alternative anticonvulsant. Several hours after receiving HTS, the patient’s level of consciousness progressively decreased, and no menace response was present. The patient was given a dose of mannitol^g (0.5 g/kg IV), and after administration of mannitol, he became more alert but remained nonambulatory. The patient was also given ondansetron^h (0.2 mg/kg IV q 8 hr) because of the history of vomiting and concern that there was a risk of aspiration of gastric contents as a result of altered mental status. Intravenous fluids (Plasmalyte Aⁱ 4 mL/kg/hr) were administered. Three-view thoracic radiographs were unremarkable, and 6 hr after admission, the patient’s SpO₂ was measured within the normal range.

The patient was transferred for ongoing care to the intensive care unit 14 hr after admission. On presentation to the intensive care unit, the patient was obtunded, nonambulatory, had no menace response, minimal pupillary light response, and anisocoria, with the left pupil larger than the right. The patient’s serum was no longer lipemic, so he was administered a second dose of 20% sterile lipid solution^e (1.5 mL/kg IV followed by a constant-rate infusion of 0.25 mL/kg/min IV over 60 min). The patient was also maintained on mannitol (initial dose of 1 g/kg IV, then 0.5 g/kg IV q 6 hr), levetiracetam (20 mg/kg IV q 8 hr), ondansetron (0.2 mg/kg IV q 8hr), and Plasmalyte A (5 mL/kg/hr IV). Over the course of the day, the level of consciousness and cranial nerve deficits improved following a dose of mannitol and then gradually declined over the next 4–6 hr.

After a total of six doses of mannitol (3.5 g/kg), the patient remained dull and nonambulatory, but the anisocoria had resolved, and the menace response was consistently present. As the patient’s neurologic status consistently improved, no additional doses of mannitol were given. By 40 hr after admission, the patient began to be able to ambulate, although he remained tetraparetic and circled to the right. Forty-eight hours after admission, the patient was more alert, was offered food, and ate readily. Ginkgo biloba extract (60 mg [11.8 mg/kg] per os [PO] q 24 hr) was added to the treatment plan. Over the next 2 days, mentation and paresis improved considerably. The patient continued to circle to the right, and the previously reported cranial nerve deficits resolved. The patient was ultimately discharged on day 4 on levitaracetam (125 mg [24.5 mg/kg] PO q 8 hr), Cerenia (8 mg [1.6 mg/kg] PO q 24 hr), and ginkgo biloba (60 mg [11.8 mg/kg] PO q 24 hr). A serum desmethylbromethalin level collected prior to ILE therapy was not available for analysis; however, a post-ILE sample showed a level of 0.5 ppb, confirming exposure.¹⁶ At his 2 wk recheck examination, the patient had a normal neurologic examination.

Discussion

As with many toxins, induction of emesis is considered to be a first-line treatment if possible. In this patient, exposure to bromethalin had occurred too many hours prior to presentation to allow for successful decontamination. Additionally, as bromethalin undergoes enterohepatic recirculation, activated charcoal is often recommended. In this patient, his altered mental status did not allow for safe administration of oral medications because of the concern that he would aspirate, so activated charcoal was not given. In patients with recent exposure who are neurologically appropriate, induction of emesis and administration of activated charcoal should be considered.

Although a decrease in serum desmethylbromethalin levels secondary to ILE therapy has previously been reported in an asymptomatic dog, this is the first report of successful treatment of a clinically affected dog.¹² The serum desmethylbromethalin level in this patient was 0.5 ppb and was obtained after two doses of ILE. Heggem-Perry et a1. demonstrated a serum desmethylbromethalin decrease from 4 ppb to 1 ppb 1 hr after the administration of ILE.¹² The serum desmethylbromethalin level at which clinical signs occur is unknown, and given the length of time between the initial administration of ILE and sample collection, it is unclear how this compares with the previous report. Clinical signs in bromethalin toxicosis can be divided into the following two syndromes: a convulsant syndrome at doses above the LD50 and a paralytic syndrome at doses below the LD50.³ Although the total amount of bromethalin that this dog ingested is unknown, the observation of severe clinical signs affecting the brain indicates that the dose ingested was likely greater than the LD50. It should be noted that although the patient’s signs were severe and associated with intracranial disease, our initial neurologic examination was affected by the prior administration of phenobarbital and midazolam; however, the presence of circling and seizures prior to presentation indicates that enough bromethalin had been ingested to cause intracranial signs.

Although ILE has been shown to alter pharmacokinetics from as early as the 1970s, it has only come into widespread clinical use in the veterinary field in the last decade and has been documented to be useful in the management of other toxicoses associated with lipophilic compounds such as local anesthetics, macrocyclic lactones, permethrin, and ibuprofen.^13–15 The hypothesized mechanism of action of ILE is the creation of a “lipid sink,” which sequesters a lipophilic toxin in the plasma compartment.¹³ Adverse effects are uncommon, but anaphylaxis, contamination of the emulsion, lipemia, lipid emboli, hepatomegaly, splenomegaly, icterus, thrombocytopenia, increased clotting times, hemolysis, and alterations in pulmonary function in human patients with acute respiratory distress syndrome have been observed.¹³

The usefulness of ILE in treating a toxicity can be predicted based on the log P of a substance, with P representing the partition coefficient of that substance between octanol and water.¹³ A log P > 1 indicates that a substance is highly lipophilic and likely to be sequestered in the intravascular space if ILE is given, whereas those with a log P < 1 are more hydrophilic. The log P of bupivacaine, for which ILE therapy has been well established, is 3.4.¹⁷ In contrast, the log P for bromethalin is 6.2, and the predicted log P for desmethylbromethalin is 6.6, indicating that it may be ideally suited for ILE therapy.^18,19 As this is a single case report, it is not possible to say whether the ILE given to this patient was the cause of his recovery. Although his recovery was delayed from the time of ILE administration, a delayed clinical response to ILE has previously been reported in a case of ibuprofen toxicosis despite nondetectable serum ibuprofen levels after ILE administration.¹⁵ Given that the severity of the clinical presentation of symptomatic bromethalin toxicosis and the minimal adverse effects observed in small animal patients who receive ILE, it is a promising therapy for this toxicosis, although further study is indicated to evaluate its effect on outcome.

The lipophilicity of medications that the patient is receiving must be evaluated prior to commencing ILE therapy, as a secondary consequence of the use of ILE is the sequestration of prescribed lipophilic medications in the plasma compartment along with the toxin. The log P of phenobarbital is 1.5, which is high enough that its distribution is likely to be affected be ILE therapy.²⁰ In comparison, the log P of levetiracetam is –0.3, indicating that it is quite hydrophilic, and its distribution is unlikely to be altered by ILE therapy.²¹

The patient in this report also received a single dose of HTS with 6% hydroxyethyl starch^d as well as repeated doses of mannitol. The combination of HTS and 6% hydroxyethyl starch^d was administered initially to treat presumptive intracranial hypertension, to maintain cerebral perfusion, and to increase his blood pressure. The beneficial hemodynamic effects of HTS may be longer lasting when combined with 6% hydroxyethyl starch^d compared with administering HTS alone.²² This combination was chosen initially rather than mannitol given the patient’s initial blood pressure and concern that mannitol might induce hypotension. During the infusion, the electrocardiogram was monitored continuously, and his blood pressure was monitored before and after infusion.

The use of mannitol as an osmotic diuretic was reported to decrease intracranial pressure in a rat model of bromethalin toxicity, although the total dose and duration of efficacy was not reported.⁵ For osmotherapy to be effective, plasma mannitol concentrations need to be >5 mg/dL, which can be achieved for 4–6 hr after a dose of 0.5–1 g/kg in humans.²³ The finding of the aforementioned study is consistent with this case report, in which the patient would show improvement in mentation and cranial nerve deficits following the administration of mannitol, with a slow decline over 4–6 hr. In addition to being an osmotic diuretic, mannitol improves blood rheology and improves oxygen delivery to the brain.²³ It is also able to bind free radicals in vitro, although the clinical significance of this effect in vivo is unknown.²³

One concern with giving repeated doses of mannitol is the development of the rebound phenomenon, which describes a rise in intracranial pressure to baseline or above baseline following the administration of hyperosmolar agents.²³ It is thought to be at least in part caused by penetration of the osmotic agent into the tissues, reducing and possibly reversing the osmotic gradient. Serum osmolality can be used to direct therapy, as it is not recommended to exceed a serum osmolality of ≥320 mosm/kg; however, this technology was not available to our service at the cageside.²³ Serum electrolytes were measured on a venous blood gas every 12 hr and remained normal. In this patient’s case, we consistently observed an improvement in his neurologic status postadministration and then a slow decline over several hours. When the patient had improved and his neurologic status remained stable postadministration of mannitol, this treatment was discontinued. The optimal duration of hyperosmolar therapy cannot be determined based on this case report, and it is likely based on individual patient characteristics, response to therapy, and total bromethalin dose ingested if known.

Once the patient had regained enough neurologic function to be safely offered food, he was also given ginkgo biloba extract. In a rat model, subjects were given either 100 mg/kg of ginkgo biloba extract or placebo at the same time as a dose 1 mg/kg of bromethalin, and the subjects given ginkgo biloba had decreased clinical signs compared with rats who received the saline placebo.⁸ Ginkgo biloba is postulated to have hypoxia-protective effects and has previously been shown to speed recovery in rats given triethylin, which is another compound that uncouples oxidative phosphorylation.⁸ It is unclear if the coadministration of ginkgo biloba decreased gastrointestinal absorption of bromethalin in the rodent model or protected the subjects from hypoxia-induced injury. Given that it is inexpensive, readily accessible, and has been shown to be relatively safe in dogs, it was elected to administer it to this patient once it was determined that he was neurologically appropriate enough to take oral medications; however, the efficacy of this treatment is unknown.²⁴

Conclusion

In summary, this report describes the successful treatment of severe bromethalin toxicosis in a dog. A return to normal neurologic function cannot be determined to be directly related to a single agent or a combination of therapies, as used in this particular case. Nonetheless, the patient’s return to normal neurologic function is encouraging. Further research is indicated to investigate the efficacy of ILE therapy and optimal dosing of hyperosmolar solutions in bromethalin toxicosis.