Transient hyperammonaemia following epileptic seizures in cats

Abstract

Objectives

The aim of this study was to determine whether transient postictal hyperammonaemia exists in cats.

Methods

The medical records of all feline patients that presented at a Swedish veterinary hospital between 2008 and 2018 were retrospectively reviewed to find those that had a recent or ongoing epileptic seizure. To qualify for inclusion, the medical record had to include information on at least one ammonia value taken in close proximity to, or during, an active seizure, the cat must have exceeded the normal upper limit of blood ammonia concentration on initial testing (reference interval 0–95 μmol/l), and there needed to be a follow-up ammonia value available within a maximum of 3 days.

Results

Five cats were included in the study, and they had blood ammonia concentrations on initial testing ranging from 146 to 195 µmol/l. They were all retested within a period of 2 h to 3 days of the original reading. All five cats had a spontaneous decrease in ammonia levels without any specific treatment for hyperammonaemia.

Conclusions and relevance

Pursuant to the findings of this retrospective study, transient hyperammonaemia may be noted after epileptic seizure in cats. Consequently, a differential diagnostic list in feline patients with hyperammonaemia could, depending on the context, include non-hepatic-related pathologies, such as epileptic seizures.

Introduction

Epileptic seizure is one of the most common neurological presentations in cats, with a reported incidence of approximately 2.1% ¹ in a referral hospital population. According to the International Veterinary Epilepsy Task Force, research that attempts to uncover the underlying cause of epileptic seizures should, at a minimum, include data on metabolic biochemistry tests, such as fasting bile acids and/or ammonia, so that hepatic encephalopathy can be ruled out. ²

Elevated ammonia levels can occur either as a result of increased ammonia production or as a result of decreased detoxification. Ammonia is produced primarily in the gastrointestinal tract by bacterial metabolism of amino acids, urea and glutamine. The liver is responsible for detoxification of ammonia to urea before its excretion in urine.^3,4 Hyperammonaemia can develop in patients with hepatic failure or shunting of portal blood, and it may also occur secondary to urea cycle enzyme deficiencies. ⁵ The ammonia load can also be excessive as a result of muscle activity. ⁶ The kidneys are another source of ammonia and its release changes with acute disturbances of the acid–base balance. ⁶ Ammonia can be metabolised to glutamine by astrocytes and other extrahepatic cells. ³ In the case of liver failure, astrocytes detoxify a larger portion of ammonia, which can ultimately cause astrocyte swelling and encephalopathy. ⁷

In cats, hyperammonaemia is mainly associated with acute and chronic hepatic encephalopathy with portovascular abnormalities and cirrhosis as differentials. ⁸ However, in human medicine, recent studies have shown a link between transient hyperammonaemia and epileptic seizures in cases without concurrent signs of liver failure.^9–13 Nakamura et al ⁹ claimed that idiopathic epilepsy alone can cause transient elevated ammonia levels. This was supported by Tomita et al, ¹⁴ who showed that hyperammonaemia exists during generalised seizures. Moreover, they found that the measurement of ammonia alone is important as an individual marker in the diagnosis of generalised seizures. ¹⁴ Furthermore, Hung et al ¹¹ concluded that the incidence of transient postictal hyperammonaemia was 67.8% in human patients, with a subsequent reduction in ammonia levels within 8 h. To date, in veterinary medicine, transient postictal hyperammonaemia has only been described in one canine case report. ¹⁵

The relationship between seizures and transient increased ammonia levels is not fully understood, but human medical literature has suggested a few physiological mechanisms. Under normal conditions, muscle activity alone produces ammonia. ¹⁶ During extensive muscle contractions, as in seizures, the muscle production of ammonia overwhelms liver metabolism, possibly owing to reduced blood flow. The net result is hyperammonaemia.^17,18

The study by Hung et al ¹¹ investigated the relationship between general tonic–clonic seizures and non-general tonic–clonic seizures and ammonia elevation. The group with general tonic–clonic seizures had significantly higher ammonia levels, with patients with status epilepticus having even higher levels. Another suggested mechanism is an increase in ammonia levels with acidosis. ¹⁹ Acidosis interferes with glycolytic pathways in the red blood cells and produces ammonia in the same way muscle cells do.^19,20 Nakamura et al ⁹ found a strong relationship between ammonia elevation and acidosis in patients with idiopathic epilepsy.

The purpose of this study was to investigate whether a link between transient postictal hyperammonaemia and epileptic seizures can be found in feline patients.

Materials and methods

The case records of cats that presented between 1 January 2008 and 1 January 2018 to the Anicura Läckeby Animal Hospital in Kalmar (Sweden) were reviewed retrospectively. Information regarding signalment, such as age, sex and breed, was included.

Cats with status epilepticus, cluster seizures or a single seizure within hours of admission or on arrival that underwent a full neurological examination were included in the study, with each record including information about the timing and description of the seizure activity.

Additional inclusion criteria were that the biochemistry information collected included at least one ammonia value obtained directly at the time of admission, that the cat exceeded the normal upper limit of blood ammonia concentration on initial testing, and that one follow-up ammonia value repeated within a maximum of 3 days was available. The blood samples had been immediately centrifuged following collection and analysed within 1 h on a Catalyst Dx (IDEXX) with the use of dry-slide technology. This technology involves a chemical reaction between ammonia and bromophenol blue, where the ensuing colour change is read optically. ²¹ The reference interval (RI) for ammonia is 0–95 µmol/l. As part of the blood database inclusion criteria, each record also had to contain a more extensive set of blood values (haematology, as well as the additional biochemical values of sodium, potassium, calcium, total protein, urea nitrogen, creatinine, glucose, alanine aminotransferase and alkaline phosphatase). Acidosis was not evaluated in this study. If present, findings on abdominal ultrasound were also noted.

The data obtained regarding the outcome of the patients were recorded as survival time in years. Zero years indicated immediate euthanasia or death. In cases where the cause of death was known, such information was noted.

Results

Of the 147 cats that presented with a recent or ongoing epileptic seizure, 101 were subsequently excluded from the study as no data concerning ammonia levels had been noted.

Of the 46 cats that remained, 36 had normal ammonia values. On initial blood sampling, 10 cases exceeded the normal upper limit of blood ammonia concentration. Five of those cases were rechecked within 2 h to 3 days of initial testing, and these were the final cases enrolled in this study. Their initial ammonia levels ranged from 146 to 195 µmol/l (RI 0–95 µmol/l). All five cats displayed a spontaneous decrease to normal ammonia levels on subsequent testing without any specific treatment for hyperammonaemia. Details of these five cats with transient hyperammonaemia are shown in Table 1, including breed, age, type of seizure at presentation, history of previous seizures, initial ammonia level, follow-up ammonia level, whether abdominal ultrasound was performed, survival time in years and cause of death.

Table 1

Five cats with elevated ammonia levels

Cat	Breed	Age at presentation (years)	Type of seizure at presentation	History of previous seizures	Initial ammonia value (RI 0–95 µmol/l)	Follow-up ammonia value (RI 0–95 µmol/l)	Ultrasoundperformed	Consciousnesslevel at first presentation	Survival time (years)	Causeof death
1	Domestic shorthair	4	Cluster	No	153	31	Yes	Impaired	2	Unknown
2	Birman	14	One seizure	Yes	186	47	Yes	Good	5	Euthanased due to anorexia and renal failure
3	Domestic shorthair	4	One seizure	Yes	146	46	Yes	Impaired	4	Euthanased due to renal failure
4	Domestic shorthair	3	Cluster	No	187	0	Yes	Impaired	0	Euthanased due to aggressive behaviour
5	Domestic shorthair	1	Status epilepticus	No	195	33	No	Impaired	0	Complications of status epilepticus

RI = reference interval

Four cats were domestic shorthair and one was a Birman. The age of the cats varied from 1 to 14 years. The type of seizure activity prior to blood sampling included status epilepticus in one case, cluster seizures in two cases and a single seizure in two cases. The initial ammonia level varied from 146–195 µmol/l with an RI of 0–95 µmol/l. Follow-up ammonia levels were all below the RIs within 2 h to 3 days after initial blood sampling without any specific treatment. Four cats had a normal abdominal ultrasound examination report and one cat died before an ultrasound could be performed. Four cats had impaired consciousness on admission, and all but one regained consciousness in time for the follow-up ammonia sample. One cat died after 2 years of unknown causes, one was euthanased after 5 years due to anorexia and kidney failure, one cat was euthanased after 4 years owing to a diagnosis of acute kidney failure, one cat was immediately euthanased due to aggressive behaviour and the remaining cat died owing to what was regarded as a complication of status epilepticus with respiratory failure, bradycardia and ultimately cardiac arrest.

All cats investigated had extensive blood analysis to exclude metabolic causes of the seizures. Blood analyses were determined to be within normal limits, including liver enzyme levels. One cat had a normal bile acid stimulation test despite elevated ammonia levels, whereas the remaining four cats did not undergo a bile acid stimulation test.

Discussion

A proper understanding of transient postictal hyperammonaemia is important in daily veterinary practice. According to the findings of this retrospective study, transient postictal hyperammonaemia can occur in cats. In summary, all five patients with initial elevated ammonia levels had subsequent ammonia levels within normal limits when rechecked within 2 h to 3 days. Even more importantly, the results of this study illustrate that transient postictal hyperammonaemia can be found in cats with epileptic seizures without concurrent signs of liver failure. Of the whole population of feline seizure patients, 22% had an abnormally high ammonia value in the post-seizure period. This possible distinction is an important one as the current main differentials for the increased measurements of plasma ammonia are acute and chronic hepatic encephalopathy, with portovascular abnormalities and cirrhosis as the main causes of the seizures. Such a narrow focus on hepatic differentials could discourage owners from continuing treatment based on incomplete information. Therefore, epileptic seizures should be an important addition to the differential list of feline patients with hyperammonaemia. Importantly, in addition to seizures and liver disease, hyperammonaemia can also develop secondary to acquired and congenital amino acid deficiencies, urinary tract infections with urease-producing bacteria and cobalamin deficiency.^5,22–25 It was beyond the scope of this project to test for the potential impact of possible metabolic defects as possible alternative agents of change with regard to hyperammonaemia. Such an investigation could be the subject of future research.

The postictal state shows a rich variation in neurological deficits and behavioural abnormalities. Altered states of consciousness with disorientation, tiredness and exhaustion are typically seen, ² and as neurologists more frequently encounter patients in their postictal state than during their seizure, it is important to recognise this syndrome. Several hypotheses have been proposed to explain the postictal state, but it is a complex aspect of a seizure disorder. Four of five cats with hyperammonaemia at initial blood sampling in this study had decreased mentation in their postictal state. Such impaired consciousness in the postictal state together with hyperammonaemia has also been noted in the human medical literature. A study by Liu et al ¹² showed elevated ammonia levels in 11/17 seizure patients. Of the seven patients with impaired consciousness at the time of blood sampling, all seven had elevated ammonia levels. Of the 10 with full recovery of consciousness at the time of blood sampling, four had elevated ammonia levels. Liu et al ¹² could therefore show an association between transient hyperammonaemia and postictal confusion. Wilkinson et al ¹⁷ also suggested that increased systemic ammonia concentration is involved in deficits in cognition, decision-making and skilled performance. Knowledge of the link between ammonia and mentation can be of use in patients with decreased mental status in a clinical setting. For example, in an emergency setting, it is often difficult to determine the cause of a sudden lapse in consciousness level and the owner may not have witnessed the epileptic seizures. Blood ammonia analysis could therefore be a possible diagnostic tool to indicate epileptic seizures and postictal syndrome. Recognising the postictal period is also important to avoid inappropriate and unnecessary treatments, as spontaneous improvement is the rule in most cases.

An important consideration in all patients with hyperammonaemia is false-positive results due to inappropriate blood sample handling as ammonia ions are released by red blood cells over time. It is recommended that plasma and red blood cells should be separated as soon as possible and plasma should be analysed within 30 mins. ⁸ For the present study, the blood samples were analysed within 1 h, as was the case in one human article. ¹⁰ The labile nature of ammonium ions makes many clinicians hesitant to use them as a marker in routine clinical practice. However, contrary to such concerns, in an unpublished study undertaken at Läckeby Animal Hospital, ammonia in plasma in healthy cats was analysed after 10 mins, 1 h and 6 h. Despite the great variability in time, the mean value over time was 7±6 µmol/l. ²⁶ This result is similar to an article on humans where the mean rate of increase in ammonia for normal whole blood per hour was 6 µmol/l at 22°C.27 The rate of ammonia increase is affected by erythrocyte and platelet counts, as well as plasma activity of gamma-glutamyl transferase and alanine aminotransferase. ²⁷ Ammonia tests can be a good tool to use in emergency situations because an exact number is not important postictally, and a maximum time of 1 h before analysing the sample is therefore tolerable.

Another concern is the accuracy of in-house laboratory ammonia measurements compared with measurements performed by a reference laboratory. In-house dry-slide technology was used in this study and it is reliable when used by trained personnel using free-flowing blood without massage. Importantly, this method allows exclusion of hyperammonaemia and is suitable for screening, but confirmation of abnormal results by a reference laboratory using quantitative methods is recommended. As a normal follow-up ammonia test is expected after an epileptic seizure, in-house laboratory testing can be of great value.^28–30

Another consideration is the postprandial effect on ammonia levels. Post-feeding hyperammonaemia in the mesenteric blood was found in patients with liver cirrhosis. ³¹ Similarly, Kogika et al ³² found a significant difference in pre- and postprandial ammonia levels in the fresh plasma from healthy cats. However, when Walker et al ³³ performed ammonia tolerance tests on venous blood of healthy dogs and dogs with suspected liver disease, they concluded that ammonia measurement is a useful test to diagnose congenital portosystemic vascular abnormalities, but that it has poor sensitivity in detecting hepatocellular disease. Subsequently, they claimed that severe generalised liver disease must be present to cause abnormally high ammonia levels. To support the latter claim, Wright et al ³⁴ suggested that hyperammonaemia could never occur in the presence of adequate blood flow. In conclusion, on the issue of the possible influence of postprandial effects, the findings of the current study can most probably not be attributed to such a state. Nevertheless, a clinical assessment of the fasting blood ammonia concentration might be advantageous to more definitively rule out postprandial effects.

Finally, as this was a retrospective study, there were several limitations. Many of the cats that presented with epileptic seizures did not meet the inclusion criteria owing to a lack of analysed ammonia values. Therefore, the total number of cats included in the study was low. In addition, the actual diagnosis of epileptic seizures was based on the medical history and background provided by the owner. Another limitation is that the time between the first and second blood samples was not standardised, and it is not known how fast the ammonia level decreased in patients. Finally, one cat did not have a further diagnostic evaluation of the liver with ultrasound, and 4/5 cats did not have other blood tests, such as bile acid testing. Other liver pathologies could therefore have been missed and not picked up by the liver biochemical analysis that was performed in this study (alanine aminotransferase and alkaline phosphatase).

Despite the limitations cited above, the results of this study provide inspiration for future veterinary medicine studies on this important subject matter. A future prospective study with a larger cohort of feline patients of different breeds, ages and comorbidities could further investigate not only repeated ammonia levels with a standardised protocol, but also use repeated measures to examine concurrent transient levels of mentation in veterinary medicine, as well as complementary ultrasound imaging of the liver. As epileptic seizures are one of the most common neurological presentations in cats, all additional future findings to help diagnose and help estimate the prognosis of feline patients with such neurological clinical signs will be of great help to clinicians in daily clinical practice.

Conclusions

The results of this study show that transient postictal hyperammonaemia can be found in cats with epileptic seizures. Consequently, epileptic seizures could, depending on the context, be a non-hepatic pathology of interest to be included in the differential list in feline patients with hyperammonaemia.