Retrospective Report of Improvement Following Treatment With Intravenous Lipid Emulsion Therapy in Two Puppies With Hyperammonemia and Signs of Hepatic Encephalopathy Refractory to Conventional Therapy

ABSTRACT

Objective

To describe the adjunctive use of IV lipid emulsion (ILE) therapy in the medical management of two dogs with hepatic encephalopathy (HE). Clinical signs in both cases and plasma ammonia concentration in one case were documented to have improved immediately after ILE therapy.

Series Summary

Two dogs were presented with signs consistent with HE and were initially treated with conventional therapy including lactulose, antimicrobials, anticonvulsants, and IV fluids. Clinical signs and ammonia concentrations initially showed no or minimal improvement. The patients were then administered ILE therapy, after which clinical signs improved and, in one case, hyperammonemia improved.

New or Unique Information Provided

To the authors’ knowledge, this case series is the first report of improvement of clinical signs and ammonia concentration in dogs intentionally treated with ILE therapy for refractory signs of HE despite conventional therapy.

Abbreviations

HE

hepatic encephalopathy

ILE

intravenous lipid emulsion

PTT

partial thromboplastin time

RI

reference interval

1. Introduction

Hepatic encephalopathy (HE) is a complication of liver dysfunction that can be life‐threatening and encompasses a variety of neurologic signs [1, 2, 3]. Clinical signs of HE include a wide range of neurologic effects, from mild behavioral changes to seizures and coma. Decreased mobility, lethargy, ataxia, hypersalivation, head pressing, visual deficits, circling, vomiting, shaking, tremors, weakness, and varying levels of obtunded mentation are reported [1, 3, 4]. Portosystemic shunting is the most common cause of HE in dogs and cats. Other etiologies include cirrhosis, acute liver failure, and congenital urea‐cycle enzyme deficiency [1, 4]. The pathophysiology of HE is complex and incompletely understood. Hyperammonemia has long been thought to be a key component of HE and is often monitored. Ammonia has the advantage of being more readily measurable than many other molecules suspected to contribute to HE [1, 2, 4].

Intravenous lipid emulsion (ILE) is a formulation of triglycerides that form single‐layer spheres (liposomes). These formulations are used as a component of parenteral nutrition. Although prospective studies are limited, use of ILE as a component of therapy for a variety of toxicoses in human and animal medicine has been increasing [5, 6]. In 2021, a case report detailed the resolution of severe neurologic signs of HE in a dog after administration of ILE alone. Naproxen intoxication was initially suspected based on history and presentation. The dog (later diagnosed with an intrahepatic portosystemic shunt) showed complete resolution of neurologic signs within 30 min of ILE administration. The dog also showed improvement in hyperammonemia 24 h after presentation, after both ILE and conventional therapy for HE had been administered [7].

To the authors’ knowledge, no consensus guidelines for therapy of HE in dogs exist. Treatment of HE in human and veterinary patients traditionally focuses on addressing any precipitating factors, lactulose administration, antimicrobials (often metronidazole or ampicillin), anticonvulsants, and nutritional management. Ultimately, if possible, the underlying cause is addressed (e.g., shunt attenuation). Recently, investigations of l‐ornithine l‐aspartate and fecal transplantation have also shown promise [1, 2, 3]. In addition to ILE, treatment for the dog in the prior case report included lactulose, maropitant, pantoprazole, acetylcysteine, and Plasmalyte‐148 [7]. To the authors’ knowledge, this inadvertent treatment is the only report of using ILE for HE. Here, we report two puppies treated with ILE therapy for signs of suspected HE that were refractory to conventional initial interventions.

2. Case Series

2.1. Case 1

A 14‐week‐old neutered male Chihuahua mix weighing 1.75 kg was presented to an emergency hospital for evaluation of progressive lethargy, “seeming wobbly,” and vomiting over a period of approximately 2 weeks. He was reportedly the smallest of his litter. On presentation, he was noted to be emaciated with generalized weakness and proprioceptive ataxia, and preferred to lie in lateral recumbency. Serum biochemistry analysis¹ showed ammonia of 167 µmol/L (284.4 µg/dL, reference interval [RI]: 0–98 µmol/L; 0–166.9 µg/dL), glucose of 14.7 mmol/L (268 mg/dL, RI: 4.1–7.9 mmol/L; 74–143 mg/dL), sodium of 172 mmol/L (172 mEq/L, RI: 144–160 mmol/L; 144–160 mEq/L), chloride of 131 mmol/L (131 mEq/L, RI: 109–122 mmol/L; 109–122 mEq/L), alanine aminotransferase (ALT) of 667 U/L (RI: 10–125 U/L), alkaline phosphatase (ALP) activity of 1104 U/L (RI: 23–212 U/L), gamma‐glutamyltransferase (GGT) activity of 25 U/L (RI: 0–11 U/L), and amylase of 211 U/L (114.1 Somogyi units, RI: 500–1500 U/L; 270.3–810.8 Somogyi units). A lactulose² enema was administered, along with a dose of midazolam³ (0.5 mg/kg, IV), before the puppy was transferred to a specialty and emergency hospital for continued care.

Upon transfer, the puppy remained mentally inappropriate with dull mentation and constant vocalization, and was unable to stand. Prothrombin time and partial thromboplastin time (PTT)⁴ were normal (17.0 s [RI: 11.0–17.0 s] and 88.0 s [RI: 72.0–102.0 s], respectively). A CBC⁵ showed WBC of 31.80 × 10⁹/L (31.80 K/µL, RI: 5.05–16.76 × 10⁹/L; 5.05–16.76 K/µL), neutrophils of 25.08 × 10⁹/L (25.08 K/µL, RI: 2.95–11.64 × 10⁹/L; 2.95–11.64 K/µL), monocytes of 2.77 × 10⁹/L (2.77 K/µL, RI: 0.16–1.12 × 10⁹/L; 0.16–1.12 K/µL), and platelet count of 26 × 10⁹/L (26 K/µL, RI: 148–484 × 10⁹/L; 148–484 K/µL). The thrombocytopenia was suspected to be spurious based on normal manual platelet estimate of approximately 236 × 10⁹/L (236 K/µL, RI: 160–500 × 10⁹/L; 160–500 K/µL). Recheck serum electrolytes⁶ showed serum chloride of 131 mmol/L (131 mEq/L, RI: 105–119 mmol/L; 105–119 mEq/L) and serum sodium of 167 mmol/L (167 mEq/L, RI: 145–157 mmol/L; 145–157 mEq/L). Blood glucose concentration⁷ was 5.2 mmol/L (95 mg/dL, RI: 3.3–6.3 mmol/L; 60–115 mg/dL), and serum ammonia⁶ was 106 µmol/L (180.5 µg/dL, RI: 0–99 µmol/L; 0–168.6 µg/dL). Levetiracetam⁸ (28.6 mg/kg, IV), maropitant⁹ (1 mg/kg, IV), and a cleansing enema (warm water), followed by a lactulose² retention enema (571.46 mg/kg diluted with warm water; retained for 30 min) were administered. The puppy was hospitalized with IV fluid therapy using a balanced electrolyte solution¹⁰ and D5W¹¹, with the rate adjusted with the goal of decreasing sodium by no more than 1 mmol/L/h (1 mEq/L/h) [8].

The next morning, 12 h after initiation of therapy, the patient remained dull and nonambulatory with a new rotary nystagmus and miotic pupils. Flumazenil¹² (0.02 mg/kg, IV), metronidazole¹³ (7 mg/kg, PO), pantoprazole¹⁴ (0.94 mg/kg, IV), and an additional cleansing enema followed by a lactulose retention enema were administered. At 16 h, blood glucose⁷ was 6.6 mmol/L (120 mg/dL, RI: 3.3–6.3 mmol/L; 60–115 mg/dL). An additional cleansing enema followed by a lactulose retention enema was administered at 19 h, along with lactulose (571 mg/kg, PO), immediately after which serum ammonia was 96 µmol/L (163.5 µg/dL, RI: 0–99 µmol/L; 0–168.6 µg/dL), serum sodium was 159 mmol/L (159 mEq/L, RI: 145–157 mmol/L; 145–157 mEq/L), and serum chloride was 121 mmol/L (121 mEq/L, RI: 105–119 mmol/L; 105–119 mEq/L). D5W was discontinued, and dextrose¹⁵ 2.5% was added to the balanced electrolyte solution at 21 h after initiation of conventional therapy. The reason for dextrose supplementation was not recorded. At 22 h, blood glucose concentration⁷ was 6.3 mmol/L (114 mg/dL, RI: 3.3–6.3 mmol/L; 60–115 mg/dL). At 23 h, levetiracetam (28.57 mg/kg, PO), pantoprazole (0.94 mg/kg, IV), and metronidazole (14 mg/kg, PO) were administered. The puppy's mentation remained dull.

Due to the persistent, severe mentation changes at approximately 24 h after initiation of conventional therapy, informed consent was obtained from the owners for initiation of ILE therapy. ILE¹⁶ was administered as a 1.5‐mL/kg bolus over 15 min, followed by a constant rate infusion (CRI) at 0.25 mL/kg/min for 45 min. Shortly after ILE, the puppy's mentation notably improved. The mentation became quiet to bright, and the puppy was weakly ambulatory and interactive. The previous head tilt, nystagmus, and hypersalivation resolved. An additional cleansing enema and then lactulose retention enema were administered at 25 h. Lactulose was also administered orally, and the patient ate one‐quarter of a can of a puppy food diet¹⁷. At 29 h, the puppy was fed one‐quarter of a can of commercial, prescription dog food for dogs with liver disease¹⁸. At this time, the puppy's mentation was quiet, alert, and responsive.

Overnight, therapy was continued with maropitant (1 mg/kg, IV, q 24 h), levetiracetam¹⁹ (28.6 mg/kg, PO, q 8 h), pantoprazole (0.94 mg/kg, IV, q 12 h), metronidazole (13.6 mg/kg, PO, q 12 h), and balanced electrolyte solution with 2.5% dextrose supplementation. Two additional cleansing enemas followed by lactulose retention enemas were administered (for a total of six in hospital). The next day (40 h after initiation of conventional therapy), recheck serum ammonia was 50 µmol/L (85.1 µg/dL, RI: 0–99 µmol/L; 0–168.6 µg/dL), and serum sodium and chloride were within normal limits: 154 mmol/L (154 mEq/L, RI: 145–157 mmol/L; 145–157 mEq/L) and 115 mmol/L (115 mEq/L, RI: 105–119 mmol/L; 105–119 mEq/L), respectively. Blood glucose concentration⁷ was 7.8 mmol/L (142 mg/dL, RI: 3.3–6.3 mmol/L; 60–115 mg/dL). The puppy's mentation remained markedly improved. The puppy was ambulatory, apparently visual, and excited and interactive with the family.

Due to financial constraints, the owners declined further diagnostic tests, and the patient was discharged with prescriptions for metronidazole, lactulose, maropitant²⁰, levetiracetam, omeprazole²¹, and the commercial, prescription liver diet. Table 1 presents a timeline of ammonia concentrations and key interventions for this case.

TABLE 1.

Key aspects of clinical course in a 14‐week‐old neutered male Chihuahua mix with PSS treated for HE with ILE in addition to conventional therapy.

Time (h:min)	Ammonia in µmol/L (µg/dL) [RI: 0–99 µmol/L; 0–168.6 µg/dL] a
0:00	167 (284.4)
0:17	Initiate conventional therapy
9:04	106 (180.5)
19:34	96 (163.5)
23:35	Initiate ILE
39:38	50 (85.1)

Abbreviations: HE, hepatic encephalopathy; ILE, intravenous lipid emulsion; PSS, portosystemic shunt; RI, reference interval.

^a The reference interval is 0–99 µmol/L (0–168.6 µg/dL) for all except the first ammonia concentration. The ammonia concentration on Day 1 at 4:08 p.m. was obtained from a different blood chemistry machine and is 0–98 µmol/L (0–166.9 µg/dL).

2.2. Case 2

A 14‐week‐old intact male Newfoundland weighing 14.2 kg was presented to an emergency and specialty hospital for evaluation of dull mentation, restlessness, nausea, ptyalism, and decreased appetite. On presentation, the puppy had a dull to stuporous mentation, delayed to absent menace response, decreased gag reflex, a thin body condition, and severe ptyalism. Blood work^5,6 abnormalities at the time of admit included a serum ammonia of 380 µmol/L (647.1 µg/dL, RI: 0–99 µmol/L; 0–168.6 µg/dL), PTT of 103.0 s (RI: 72.0–102.0 s), serum ALP of 367 U/L (RI: 46–337 U/L), serum amylase of 154 U/L (83.2 Somogyi units, RI: 300–1300 U/L; 162.2–702.7 Somogyi units), BUN of 2.1 mmol/L (6 mg/dL, RI: 2.5–10.4 mmol/L; 7–29 mg/dL), serum sodium of 161 mmol/L (161 mEq/L, RI: 145–157 mmol/L; 145–157 mEq/L), HCT of 0.36 L/L (36.3%, RI: 0.37–0.62 L/L; 37.3%–61.7%), hemoglobin of 122 g/L (12.2 g/dL, RI: 131–205 g/L; 13.1–20.5 g/dL), mean corpuscular volume of 54.2 fL (RI: 61.6–73.5 fL), mean corpuscular hemoglobin of 18.2 pg (RI: 21.2–25.9 pg), monocytes of 1.46 × 10⁹/L (1.46 K/µL, RI: 0.16–1.12 × 10⁹/L; 0.16–1.12 K/µL), basophils of 0.48 × 10⁹/L (0.48 K/µL, RI: 0.00–0.10 × 10⁹/L; 0.00–0.10 K/µL), platelet count of 136 × 10⁹/L (136 K/µL, RI: 148–484 × 10⁹/L; 148–484 K/µL), and mean platelet volume of 14.4 fL (RI: 8.7–13.2 fL). Urinalysis^{22 , 23} obtained via cystocentesis showed WBC (2/hpf [RI: 0–5/hpf]), RBC (2/hpf [RI: 0–5/hpf]), bacterial rods, nonsquamous epithelial cells (1–2/hpf), ammonium biurate crystals (6–20/hpf), and unclassified crystals (1–5/hpf). A urine culture²³ later showed no growth.

The puppy was hospitalized with maropitant⁹ (1 mg/kg, IV, q 24 h), balanced electrolyte solution¹⁰ (60 mL/kg/day), pantoprazole¹⁴ (1 mg/kg, IV, q 12 h), ondansetron²⁴ (0.5 mg/kg, IV, q 8 h), and ampicillin–sulbactam²⁵ (31 mg/kg, IV, q 8 h). A cleansing enema (warm water) followed by a lactulose² retention enema (469.5 mg/kg; diluted in warm water; retained for 30 min) was administered and continued every 6 h while hospitalized. An abdominal ultrasound²⁶ showed an intrahepatic portosystemic shunt, as well as mild microhepatica, moderate renomegaly, a small amount of mineral debris in the urinary bladder, and fluid throughout the gastrointestinal tract with decreased peristalsis consistent with functional ileus. A recheck serum ammonia 4.5 h after the initiation of conventional therapy was 193 µmol/L (328.7 µg/dL, RI: 0–99 µmol/L; 0–168.6 µg/dL), and the puppy's mentation remained markedly obtunded. Due to the persistent severe mentation changes, informed consent was obtained from the owners for initiation of ILE therapy. Approximately 5 h after initiation of conventional therapy, ILE¹⁶ was initiated as a 4.6‐mL/kg bolus over 15 min, followed by a CRI at 0.22 mL/kg/min for 1 h¹⁰. A recheck ammonia an hour later (6 h after initiation of conventional therapy) was 79 µmol/L (134.5 µg/dL, RI: 0–99 µmol/L; 0–168.6 µg/dL), and the mentation had markedly improved to quiet, alert, and responsive. At 7 h, lactulose (146 mg/kg, PO, q 8 h) was initiated as well as levetiracetam⁸ (60 mg/kg, IV loading dose, followed by 18.2 mg/kg, PO, q 8 h).

The puppy continued to show markedly improved mentation and ate commercial, prescription liver diet¹⁸ well in the hospital. Approximately 13 h after initiation of conventional therapy, a dose of trazodone²⁷ (5.5 mg/kg, PO) was administered for anxiety and vocalization. This resulted in profound sedation that lasted several hours. A recheck ammonia at 14 h was 221 µmol/L (376.4 µg/dL, RI: 0–99 µmol/L; 0–168.6 µg/dL). Recheck ammonia and electrolytes the next day (approximately 26 h after initiation of conventional therapy) showed a serum ammonia of 63 µmol/L (107.3 µg/dL, RI: 0–99 µmol/L; 0–168.6 µg/dL) and a serum chloride of 120 mmol/L (120 mEq/L, RI: 105–119 mmol/L; 105–119 mEq/L).

The puppy was discharged on Day 2 with prescriptions for amoxicillin–clavulanate²⁸, levetiracetam²⁹, lactulose, maropitant³⁰, and commercial, prescription liver diet¹⁰. At recheck approximately 48 h after initiation of conventional therapy, serum ammonia was 57 µmol/L (97.1 µg/dL, RI: 0–99 µmol/L; 0–168.6 µg/dL), and the puppy continued to show clinical improvement. Initially, the puppy was managed medically and showed improvements in both ammonia level stabilization on recheck blood work and behavioral signs observed at home. Eight months later, the dog underwent transjugular coil embolization. Since intervention, the dog has been managed for intermittently decreased appetite at home, entropion with corneal ulcers, and asymptomatic thrombocytopenia, but is otherwise doing well. Table 2 presents a timeline of ammonia concentrations and key interventions for Case 2.

TABLE 2.

Key aspects of clinical course in a 14‐week‐old intact male Newfoundland with PSS treated for HE with ILE in addition to conventional therapy.

Time (h:min)	Ammonia in µmol/L (µg/dL) [RI: 0–99 µmol/L; 0–168.6 µg/dL]
0:00	Initiate conventional therapy
0:31	380 (647.1)
4:34	193 (328.7)
4:52	Initiate ILE
6:20	79 (134.5)
13:15	Trazodone administration
13:58	221 (376.4)
25:48	63 (107.3)
47:40	57 (97.1)

Abbreviations: HE, hepatic encephalopathy; ILE, intravenous lipid emulsion; PSS, portosystemic shunt; RI, reference interval.

3. Discussion

The puppies in this series showed improvement of clinical signs and resolution of hyperammonemia associated with suspected HE after the administration of ILE. Previously, these puppies showed persistent marked mentation changes despite conventional therapy with antimicrobials, lactulose, IV fluids, and supportive care. The timeline of improvement is such that ILE is suspected to have played a role in this clinical improvement, although this interpretation is limited by the retrospective review and the concurrent administration of conventional therapy. Ammonia concentration was improving before ILE administration in both cases but remained markedly elevated in Case 2. In both cases, the ammonia concentration decreased further after ILE administration.

Diagnosing HE in veterinary patients is typically a diagnosis of exclusion. Signs of cerebral and hepatic dysfunction, along with presence of hyperammonemia, are suggestive of HE [1]. The pathophysiology of HE remains an area of active study. Ammonia is a normal product of the metabolic breakdown of proteins and other nitrogen sources but is toxic at increased concentrations [1, 2, 4]. In patients with significant hepatic dysfunction, ammonia concentration increases due to impaired hepatic urea synthesis. An alternate pathway for ammonia detoxification is through glutamine production in the brain. Increasing glutamine buildup within the astrocytes leads to cerebral edema and, ultimately, neurologic impairment (HE). However, there are many other mechanisms thought to contribute to HE, including other neurotoxins such as manganese, oxidative stress, and disruption of the gastrointestinal microbiome [1, 2, 3].

A 2021 case report detailed the resolution of severe neurologic signs of HE in a dog after administration of ILE alone. On presentation, possible ingestion of a toxic dose of naproxen was reported. The dog was stuporous, recumbent, cortically blind, and showed torticollis. ILE was administered (1.5 mL/kg slow IV bolus, followed by CRI at 0.25 mL/kg/min IV for 45 min) to aid in decontamination for naproxen. The dog (later diagnosed with an intrahepatic portosystemic shunt) showed complete resolution of neurologic signs within 30 min of ILE administration. The dog also showed significant improvement in hyperammonemia 24 h after presentation (after both ILE and conventional therapy for HE) [7]. To the authors’ knowledge, this inadvertent treatment is the only report of using ILE for HE in dogs. Although not a routine therapy for people with HE, ILE has been used as a component of parenteral nutrition in human patients with signs of HE and is considered safe when used as such [9].

The use of ILE for treatment of various toxicoses is prevalent in veterinary medicine. However, the mechanism(s) of action of ILE remains an active area of debate. Historically, the lipid “sink” and later lipid “shuttle” theories have been most emphasized. The lipid sink theory proposes that the lipids scavenge lipophilic compounds from the bloodstream and prevent them reaching target organs. However, the dose of lipid administered would likely be inadequate to explain the effects on toxin concentrations and clinical signs by this means alone [10]. The lipid shuttle theory proposes that lipids scavenge lipophilic compounds from the target organs and expedite transport for storage, metabolism, and excretion. This change in distribution of toxin is proposed to decrease concentration in toxin‐susceptible organs, increase the volume of distribution, and increase clearance. These proposed mechanisms emphasize the importance of the lipophilicity of the target molecule [5, 10, 11]. Lipophilicity is often approximated by the log P (the octanol–water partition coefficient). Positive log P values indicate that the compound is more soluble in lipid than in water, whereas negative values indicate that the solution is more soluble in water [10]. ILE is traditionally considered an effective therapy for toxins with log P >1.0, but the full scope of ILE's mechanism of action is more complex [5, 7]. Studies indicate that positive effects of ILE depend not just on the lipophilicity but also on a mixture of physiochemical factors. These include both the pH of the blood and the thermodynamic environment, as well as various characteristics of the target compound. Important characteristics include the charge of the target molecule and its dissociation constant, which determines how much of the target molecule will be in the charged state at physiologic pH. The liposomes in ILE have a strong negative surface charge that attracts positively charged molecules. Therefore, the surface charge of the target molecule may be equally as important as its lipophilicity in determining its relative binding affinity to ILE [10].

In addition to these “scavenging” effects, various non‐scavenging mechanisms are proposed to contribute to ILE's effectiveness in decontamination [5, 10]. These mechanisms include direct cardiac effects, such as changes in inotropy and lusitropy, and vascular effects such as increased vasoconstriction (and thereby increased blood pressure and preload). ILE has also been shown to reduce ischemia–reperfusion injury, although the mechanism is incompletely understood [10]. To further confound interpretation, log P values often differ between resources, because log P may be either calculated or measured. Different algorithms can be used for calculated values, and the accuracy of experimental methods varies [12]. Log P is also pH independent and therefore does not take into account the effect of the charged molecules of the target substance that are present at physiologic pH. The log D, or distribution coefficient, is pH dependent [10] and may be a more predictive metric for future study.

The potential mechanism of ILE in patients with HE is unknown and is confounded by the fact that the pathophysiology of HE is incompletely understood. ILE scavenging and non‐scavenging effects could feasibly impact one or several of the compounds that play a role in HE, including ammonia. Ammonia is considered a lipophilic compound [1, 7, 13] and is at first glance a reasonable target to consider for ILE therapy. However, the reported log P of ammonia is estimated to be 0.23 [14], which is below the >1.0 log P threshold previously described for predicting expected interaction with ILE [15]. The unionized and lipophilic ammonia (NH₃) is what crosses the blood–brain barrier and affects the central nervous system. However, ammonia in circulation is almost entirely in the ionized (positively charged) form of the ammonium ion (NH₄ ⁺) [16]. As discussed, the charge of a molecule may be as important as its lipophilicity in the activity of ILE [10]. Small fluctuations in factors such as pH may alter the proportion of ionized to nonionized ammonia in circulation and thereby shift the distribution of ammonia to the central nervous system [16]. ILE may impact this balance by scavenging either the lipophilic ammonia (NH₃), the positively charged ammonium (NH₄ ⁺), or both. Additionally, because the pathophysiology of HE is not fully understood, it is possible that the observed improvement in mentation after ILE therapy in these puppies and the previously published case report [7] is related to ILE's effects on other neurotoxic compounds involved in HE's pathogenesis. Alternatively, the non‐scavenging effects of ILE, including improved cardiovascular function and decreased ischemia–reperfusion injury [10], may have contributed to improved overall status without directly scavenging toxic substances.

One area of concern regarding the use of ILE to treat HE is its potential to interact with drugs used in conventional treatment of HE. The log P values of ammonia and some commonly administered therapeutic drugs are reported in Table 3, though as above, the effect of ILE on the pharmacokinetics of these medications may not be as simple as their log Ps suggest. Lactulose is administered orally or per rectum because it is nonabsorbable and promotes ion trapping of ammonium in the gastrointestinal tract, decreases bacterial production of ammonia, and acts as an osmotic laxative [3]. Lactulose therapy is thus considered less likely to be affected by ILE. However, some antimicrobials commonly used to treat HE are lipophilic. For example, metronidazole is weakly lipophilic [17] and ampicillin is lipophilic [18]. Their lipophilicity raises a concern for decreased bioavailability of these antimicrobials in patients with HE treated with ILE. Levetiracetam is a hydrophilic molecule with a negative log P¹² and may be less likely to be affected by administration of ILE. Propofol is sometimes also used for control of seizures in HE [19]. It is insoluble in water and is formulated in a lipid emulsion [20]; therefore, propofol is likely to be affected by the scavenging effects of ILE. However, no adverse events were noted with ILE administration in a case report of a dog with moxidectin and praziquantel intoxication in which the patient was managed with propofol for seizure activity [11]. Both exogenous and endogenous benzodiazepines are generally strongly lipophilic, including diazepam [21, 22]. Midazolam is hydrophilic in solutions of low pH but, at physiologic pH, is also strongly lipophilic [23]. Benzodiazepines are therefore likely to be affected by the scavenging effects of ILE if used for anticonvulsant effects. Indeed, apparent improvement of clinical signs with ILE after suspected benzodiazepine overdose has been reported in human medicine [24]. It is worth noting that benzodiazepines are a controversial treatment for seizures in cases of HE, and endogenous benzodiazepines may play a role in the pathophysiology of HE. The reversal agent, flumazenil, has been proposed as a potential adjunctive HE treatment to reverse the effects of endogenous benzodiazepines [1]. The lipophilicity of exogenous benzodiazepines and their endogenous analogs may in fact contribute to ILE's beneficial effects in patients with HE.

TABLE 3.

Log P of ammonia and of some therapeutics commonly used to treat hepatic encephalopathy.

Compound	Log P
Ammonia	0.23 [14]
Ampicillin	1.06 [18]
Diazepam	2.90 [22]
Levetiracetam	−0.67, −0.49 [12]
Metronidazole	0.75 [17]
Propofol	4.16 [20]

Note: Octanol–water partition coefficients (Log P values) of ammonia and of a selection of drugs sometimes used for conventional therapy of hepatic encephalopathy. Positive log p values indicate that the compound is more soluble in lipid than in water, whereas negative values indicate that the compound is more soluble in water [10].

Abbreviation: Log P, octanol–water partition coefficient.

Adverse effects of ILE administration (both for treatment of toxicosis and as a component of parenteral nutrition) have been reported in both human and veterinary medicine. ILE is considered to have a wide safety margin, and many of the reports of adverse events are from nutritional administration, where higher total doses over longer treatment durations are common [5, 6]. Adverse effects reported in people associated with use of ILE for treatment of intoxications include kidney injury, pulmonary effects (including acute respiratory distress syndrome), cardiac embolism, and “fat overload” syndrome [5]. There is also a risk for promotion of bacterial growth and susceptibility to infection, especially associated with nutritional administration (human and veterinary medicine) but also reported with its use for toxicoses in people [5, 6]. ILE administration can cause lipemia and, subsequently, spurious biochemical abnormalities on routine blood tests [5, 6]. Pancreatitis is reported in both human and veterinary medicine [5, 11]. A cat treated with ILE for permethrin toxicosis developed corneal lipidosis, although this resolved within 1 week [5]. Hemolysis, facial pruritus, and pain associated with extravasation have been reported in veterinary medicine [5, 11]. Patients should be monitored for hypersensitivity reactions, which are common in pigs but do not appear to be as common in people, dogs, or cats [5]. In a retrospective study from Germany of canine and feline patients (313 dogs, 100 cats) treated for various toxicoses with ILE at a referral hospital, 6% of treated patients experienced potential adverse effects (20/313 dogs, 4/100 cats). These adverse events included neurologic and behavioral signs (altered consciousness, ataxia, abnormal behavior), bradycardia, hyperthermia, gastrointestinal signs, respiratory distress, facial swelling, and thrombophlebitis. In people, recurrence of signs from intoxicants with long half‐lives is also reported after ILE therapy [5]. It is worth noting that the most common clinical signs of intoxication on presentation were neurologic (83.3%), followed by altered general behavior (35.4%) [15]. This means that adverse events associated with ILE administration can be difficult to differentiate from clinical signs of toxicosis in many cases.

An obvious limitation of this small case series is the fact that at the time of ILE administration, both puppies were receiving conventional therapy for HE. ILE therapy was initiated because their severe clinical signs had been apparently refractory to those conventional therapies. It is impossible to know whether the clinical improvements were due to ILE therapy or continued administration of conventional therapy, or a combination of these. However, the rapid improvement of clinical signs in both puppies after ILE administration does highlight this emerging therapy as a potential adjunctive treatment and warrants prospective study.

The ammonia concentration in these patients also showed improvement after ILE administration. However, ammonia concentration was decreasing before the administration of ILE in both cases. Additionally, in Case 1, there was a long delay between ILE and the next ammonia concentration evaluated, during which time conventional therapies were continued. The inconsistent improvement in ammonia concentration compared with the improvement of clinical signs may be because the severity of hyperammonemia is generally not well correlated with the severity of signs of HE [3, 19]. Additionally, ammonia testing is notoriously fraught with error due to its labile nature, and lipemia can affect results [3, 7, 19]. Although no overt lipemia was noted in these patients, the potential for interference cannot be ruled out. Additionally, in Case 2, trazodone administration was suspected to have contributed to the setback in mentation, as administration of sedatives can precipitate signs of HE [1, 2]. Also, trazodone‐induced hepatotoxicity is a potential side effect of administration in people and has been reported in one dog [25]. Liver values were not monitored around the time of administration, and trazodone‐induced hepatotoxicity cannot be ruled out in that case.

In Case 1, dextrose supplementation was provided despite normoglycemia. This was suspected to be due to clinician concern for hepatic glucose regulation in the anorexic, underconditioned patient. It is not known what role, if any, dextrose supplementation played in the progression of clinical signs. As both puppies were receiving targeted therapies to reduce ammonia levels, their combined effects make it difficult to isolate and quantify the specific role of ILE in ammonia reduction.

Although the exact mechanism of action for ILE therapy remains unclear, its use in veterinary medicine for treating toxicoses is increasing. Additionally, ILE may have potential as an adjunctive treatment for patients with hyperammonemia or signs of HE that respond poorly to conventional therapies.

Author Contributions

Rachel K. Lucia: conceptualization, investigation, writing – original draft, writing – review and editing, visualization. John C. Gicking: writing – review and editing, supervision, conceptualization. Virginia Sinnott‐Stutzman: supervision, writing – review and editing, conceptualization.

Ethics Statement

The authors confirm that the ethical policies of the journal, as noted on the journal's author guidelines page, have been adhered to. No ethical approval was required as this is a case series article with no original research data.

Conflicts of Interest

At the time of manuscript submission, Virginia Sinnott‐Stutzman was an assistant editor for the Journal but only participated in the review process as an author. The authors declare no other conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.