Unraveling the Link: Seizure Characteristics and Ammonia Levels in Urea Cycle Disorder during Hyperammonemic Crises

Abstract

Background:

This retrospective clinical study performed at a single clinical center aimed to identify the prevalence of seizures in individuals with urea cycle disorders (UCD) both with and without hyperammonemic (HA) crises. In addition, we sought to correlate and study the utility of biochemical markers and EEG in detecting subclinical seizures during hyperammonemic crises.

Methods:

Medical records of individuals diagnosed with a UCD and who enrolled in Urea Cycle Disorders Consortium Longitudinal Study (UCDC-LS) at Children’s National Hospital between 2006 and 2022 were reviewed for evidence of clinical seizures and subclinical seizure during HA crises, and initial biochemical levels concurrently. These data were collected as part of an observational, longitudinal, natural history study conducted by the UCDC-LS (NCT00237315).

Results:

Eighty-five individuals with diagnosis of UCD were included in the analyses. Fifty-six of the 85 patients (66 %) experienced HA crises, with a total of 163 HA events. Seizures are observed in 13% of HA events. Among all HA events with concomitant EEG, subclinical seizures were identified in 27% of crises of encephalopathy without clinical seizures and 53% of crises with clinical seizures. The odd of seizures increases 2.65 (95% CI, 1.51–4.66) times for every 100 mcmol/l increase in ammonia and 1.14 (95%CI, 1.04–1.25) times for every 100 mcmol/l increase in glutamine.

Conclusion:

This study highlights the utility of EEG monitoring during crises for patients who present with clinical seizures or who have encephalopathy with high ammonia levels. During HA events, measurement of initial ammonia and glutamine can help determine risk for seizures and guide EEG monitoring decisions.

Introduction

Urea cycle disorders (UCD) are inherited inborn errors of metabolism caused by six enzymatic and two transporter protein deficiencies that result in impaired clearance of waste nitrogen. The incidence of UCD is 1 in 35,000 births in North America¹. The subtype of UCD is determined by the enzymatic or transport defects, which can be proximal in the mitochondria (carbamoyl phosphate synthetase 1 deficiency, CPS1D; N-acetylglutamate synthase deficiency, NAGSD; ornithine transcarbamylase deficiency, OTCD) or distal in the cytosol (argininosuccinic acid synthetase deficiency, ASSD; argininosuccinate lyase deficiency, ASLD; arginase 1 deficiency, ARG1D); transporter defects (Hyperornithinemia-hyperammonemia-homocitrullinuria syndrome, HHH and citrin deficiency) (figure 1)^{2, 3}. The most common UCD is OTC deficiency, an X-linked disorder, whereas all others UCDs are transmitted in an autosomal recessive mode of inheritance. UCD are characterized by accumulation of ammonia and/or other toxic substances, contributing to the manifestation of acute symptoms⁴; including vomiting, headache, irritability, poor feeding, lethargy, seizures, and encephalopathy. The severity and onset vary based on residual enzyme or transporter activity^{5, 6}.

Figure 1: The figure illustrates each individual enzymatic step involved in the urea cycle pathway³.

The proximal step of the urea cycle takes place in the mitochondria and involves the enzymes CPS1, NAGS, and OTC, which convert ammonia into the byproduct citrulline. Afterwards, citrulline exits the mitochondria and enters the cytosol, where it undergoes the distal step involving ASS, ASL, and ARG to produce ornithine and urea. Subsequently, ornithine returns into the urea cycle via ornithine transporter 1, and urea is excreted by the kidneys.

ARG, arginase; ASL, argininosuccinate lyase; ASSD; Argininosuccinate synthase; CPS1, carbamoyl phosphate synthetase 1; NAGS, N-acetylglutamate synthase;OTC, Ornithine transcarbamylase

Whereas seizures associated with hyperammonemia have been reported previously in UCD^7–11, seizure characteristics and prevalence have not been systematically studied. Seizures can develop during metabolic crises caused by the accumulation of ammonia and/or other toxic metabolites. Ammonia passes through the blood-brain barrier and is converted into glutamine by glutamine synthetase. Subsequently, the cellular osmolarity increases due to the accumulation of glutamine, leading to cerebral edema¹². Other cellular mechanisms that occur during hyperammonemia including excitotoxicity due to increased levels of extracellular glutamine or ammonia, altered aquaporin 4 channels, disrupted energy/glucose metabolism, and impaired nitric oxide synthesis. These mechanisms may lead to seizures in UCD^{4, 13, 14}.

Prior studies have shown the occurrence of epilepsy in specific types of UCD, including ASLD and ARG1D^{15, 16}, but the overall prevalence of epilepsy in UCD or a comparison of the incidence of epilepsy in UCD is not known. There is also limited data in the literature regarding seizure type and the association between ammonia level and seizure occurrence. In this study, we sought to examine the prevalence of epilepsy, as opposed to symptomatic seizures, and the relationship between ammonia and seizure occurrence during hyperammonemic (HA) crises. We further sought to establish the utility of EEG monitoring in the clinical management of those with UCD.

Methods

We completed retrospective review of medical records of individuals with a molecularly confirmed UCD diagnosis who were enrolled in Urea Cycle Disorders Longitudinal Study (UCDC-LS) at Children’s National Hospital site between 2006 and 2022. Among a total of 161 individuals identified as having UCD, 12 asymptomatic carriers of the OTCD and 64 individuals with unavailable data were excluded. We focused on the remaining 85 individuals with UCD who had a confirmed diagnosis and whose medical records were available at Children’s National Hospital for review. Based on biochemical profiles or genetic testing, the individuals were diagnosed with NAGSD, CPS1D, OTCD, ASSD, ASLD, or ARG1D, or HHH. The following data were extracted from the review: presence or absence of seizures, age at seizure onset, seizure type(s), epilepsy diagnosis (yes/no), antiseizure medication treatment (ASM), the occurrence of seizure, and the first EEG report.

In individuals who had HA, information regarding neurological presentation, EEG monitoring reports, biochemical profiles including ammonia and glutamine and ASM treatments were gathered.

The onset of UCD was classified by age at clinical presentation as early onset (<28 days) and late onset (>28 days). Seizure type was classified by the ILAE task force on Neonatal Seizures (2020) for early onset seizures (<28 days)¹⁷ and Operational Classification of Seizure Types by the International League Against Epilepsy (ILAE 2017) for late onset seizures (>28 days)¹⁸. Seizure type was classified as specific to seizures during the HA crisis. We also established whether there was evidence for a diagnosis of epilepsy. Epilepsy according to ILAE 2014 is characterized by any of the following conditions: 1. at least two unprovoked (or reflex) seizures occurring greater than 24 hours apart, 2. one unprovoked (or reflex) seizure with a potential recurrence risk of more than 60% over the next 10 years, 3. diagnosis of an epilepsy syndrome¹⁹.

The study was conducted according to the Declaration of Helsinki and was approved by the Institutional Review Board (IRB) at Children’s National Medical Center

Statistical analysis

The statistical analysis was performed using SPSS program and SAS V9.4. Descriptive statistics were used to summarize demographic data, seizure characteristics, medical treatment, and biochemical profiles, which were reported by number, percentage, mean, and standard deviation for a normal distribution, and median and interquartile range for a non-normal distribution. We compared the onset of seizures with the type of UCD, its association with HA crisis, and the likelihood of the development of epilepsy using the chi-square test or Fisher’s Exact Test, with a 2-sided test and a p-value less than 0.05 considered statistically significant. We calculated odds ratios (ORs) and 95% confidence intervals (CIs) for the association between seizure occurrence and ammonia and glutamine levels using a Generalized Estimating Equation (GEE) with an exchangeable correlation structure.

Results

Demographic data and seizure characteristic in general

We report on 85 UCD individuals who were evaluated clinically at our site and were enrolled in UCDC-LS at our site. The demographic data including UCD diagnosis are shown in table 1. In total, 29 individuals (34%) had a history of clinical seizures: 1 of 1 individual with NAGSD (100%), 3 of 4 with CPS1D (75%), 13 of 41 with OTCD (32%), 2 of 8 with ASSD (25%), 6 of 22 with ASLD (27%), and 4 of 6 with ARG1D (67%); (p = 0.12).

Table 1:

Demographic Data for individuals with Urea Cycle Disorders (UCD)

Demographic data for UCD (N= 85)
Onset
- Early onset of UCD n (%)	22 (26%)
- Late onset of UCD n (%)	63 (74%)
Sex (male, %)	38 (45%)
Type of UCD	n (%)
- NAGSD	1 (1%)
- CPS1D	4 (5%)
- OTCD	41 (48%)
- ASSD	8 (9%)
- ASLD	22 (26%)
- ARG1D	6 (7%)
- HHH	3 (4%)
A history of individuals having experienced a seizure (n=29, 34%)	n (%) of individuals had seizure
- NAGSD	1(100%)
- CPS1D	3 (75%)
- OTCD	13 (32%)
- ASSD	2 (25%)
- ASLD	6 (27%)
- ARG1D	4 (67%)
- HHH	0 (0%)

ARG1D, arginase 1 deficiency; ASLD, argininosuccinate lyase deficiency; ASSD; Argininosuccinate synthase deficiency; CPS1D, carbamoyl phosphate synthetase 1 deficiency; HHH, Hyperornithinemia, Hyperammonemia, and Homocitrullinuria Syndrome; NAGSD, N-acetylglutamate synthase deficiency; OTCD, Ornithine transcarbamylase deficiency; UCD, Urea cycle disorders

Seizure characteristics were classified as early onset or late onset (<28 Days vs >28 Days) (table 2). Early-onset seizures frequently occurred during HA crises, whereas late-onset seizures were noted to occur either during HA or without HA crises. Early-onset seizures manifested as clonic (50%), automatism (25%), or myoclonic (17%) or tonic (8%). Late-onset seizures manifested as focal (57%), generalized tonic-clonic seizures (43%) absence (21%) or generalized tonic (7%).

Table 2.

Comparison of Seizure Characteristics, Type of UCD, and Epilepsy Development Between Early Onset Seizure, Late Onset Seizure

Clinical seizure data in UCD (n=29)	Early onset seizure	Late onset seizure	P-value
Number of individuals with UCD	15	14	N/A
Age at onset of seizure (median, IQR)	4 (2–4.5) days	6.5 (4–8.5) years	N/A
Seizure rate by type of UCD n (%)	- NAGSD = 1 (7%) - CPSD = 2 (13%) - OTCD = 9 (60%) - ASSD = 2 (13%) - ASLD = 1 (7%) - ARG1D = 0 (0%) - HHH = 0 (0%)	- NAGSD = 0 (0%) - CPSD = 1 (7%) - OTCD = 4 (29%) - ASSD = 0 (0%) - ASLD = 5 (35%) - ARG1D = 4 (29%) - HHH = 0 (0%)	0.02
Seizure characteristic n, (%)	Automatism = 3/12 (25%)* Clonic seizure = 6/12 (50%) Tonic seizure = 1/12 (8%) Myoclonic seizure = 2/12 (17%)	Focal seizure = 8 (57%) GT = 1 (7%) GTC = 6 (43%) Absence seizure = 3 (21%)	N/A
Seizure association with crisis event	15 (100%)	7 (50 %)	0.002
Epilepsy diagnosis after first seizure Ten-year follow-up completion rate (%)	0 (0%) 53%	10 (71%) 93%	<0.001 N/A

ARG1D, arginase 1 deficiency; ASLD, argininosuccinate lyase deficiency; ASSD; Argininosuccinate synthase deficiency; CPS1D, carbamoyl phosphate synthetase 1 deficiency; HHH, Hyperornithinemia, Hyperammonemia, and Homocitrullinuria Syndrome; NAGSD, N-acetylglutamate synthase deficiency; N/A, not available; OTCD, Ornithine transcarbamylase deficiency; UCD, Urea cycle disorders

^*The seizure type of three individuals with UCD in early onset group was not well defined.

Epilepsy in urea cycle disorder

Epilepsy was diagnosed in 10 UCD individuals (12%) (table 3 and figure 2) at median onset of 7.3 years (IQR 4.5–8.5 years). A significantly higher proportion of individuals with late onset seizures (71%) have been diagnosed with epilepsy compared to those with early onset seizures (0%) (p-value=<0.001). The rate of epilepsy was higher in those with ASLD and ARG1D, compared to OTCD (5 ASLD (22%), 3 ARG1D (50%), and 2 OTCD (5%), p-value=0.03). Interictal EEGs were abnormal in 8 of 10 individuals who developed epilepsy, including focal epileptiform discharges (2 frontal, 2 central, 1 temporal), generalized spike wave (4), slow background (7), and focal slow wave (2). Furthermore, 70% of UCD individuals with epilepsy have been seizure-free more than 1 year since the last appointment, while the rest (30%) continue to have ongoing seizures at their last follow up appointment.

Table 3.

Generalized Estimating Equation estimates for seizure

Initial biochemical profiles	Odds ratio (95%CI)	Standard Error	P-value
Ammonia level per every 100 mcmol/l	2.65 (1.51–4.66)	0.76	<0.001
Glutamine level per every 100 mcmol/l	1.14 (1.04–1.25)	0.05	0.004

Figure 2: The Sankey diagram illustrates epilepsy development in individuals with UCD.

ARG1D, arginase 1 deficiency; ASLD, argininosuccinate lyase deficiency; ASSD; Argininosuccinate synthase deficiency; CPS1D, carbamoyl phosphate synthetase 1 deficiency; HHH, Hyperornithinemia, Hyperammonemia, and Homocitrullinuria Syndrome; NAGSD, N-acetylglutamate synthase deficiency; OTCD, Ornithine transcarbamylase deficiency; UCD, Urea cycle disorders

Seizure events during metabolic crisis

In our cohort, 56 of 85 UCD individuals had a total of 163 HA events (range 1–13 episodes per individuals). Clinical presentation during events included encephalopathy (69%), seizure (13 %), behavioral change (6%), and ataxia (5%). Overall, seizures were identified in 22 events including 19 with clinical seizures only (although 4 of these did not undergo EEG monitoring), 3 with subclinical seizures only, and 8 with both (figure 3). Seizures occurred more often in those with early onset (15) compared to those with late onset (7) UCD. In instances where EEG monitoring was employed, EEG detected subclinical seizures in 8 of 15 (53%) HA crises in association with clinical seizures, while subclinical seizures were detected in 3 of 11 (27%) episodes of encephalopathy-only symptoms. The ammonia level in encephalopathy-only events with EEG monitoring (median 328, IQR 144–553 mcmol/l) was higher than in those individuals without EEG monitoring (median 157.5, IQR 110–210 mcmol/l). Individuals with higher ammonia had a 2.65 (95% CI, 1.51–4.66) greater odds of seizure per every 100 mcmol/l increase in ammonia and those with higher glutamine had a 1.14 (95%CI, 1.04–1.25) greater odds of seizure per every 100 mcmol/l increase in glutamine. The initial biochemical profiles of individuals with and without epilepsy are compared in figure 4.

Figure 3:

Diagrams illustrate the detection of subclinical clinical seizure in seizures and encephalopathy events.

Figure 4:

Comparison of Biochemical Profiles in individual with and without seizures

All seizure events were controlled with ASM, either as monotherapy (61%) or polytherapy (39%), along with ammonia reduction therapies such as nitrogen scavenger treatment (100%) and hemodialysis (67%). Antiseizure medications included phenobarbital in 67% of the seizure events, phenytoin in 19%, levetiracetam in 19%, midazolam in 10%, and topiramate in 5%.

Discussion

Our retrospective study reveals the seizure prevalence and seizure characteristics in association with UCD at Children’s National Hospital over the past two decades. In comparison to previous reports ^7–10, the prevalence of seizures in our study was 34% manifesting either early (15 individuals) or later (14 individuals) in childhood. All early-onset seizures observed in the study were symptomatic seizures associated with HA crises, with a higher rate in proximal UCD and ASSD. Individuals with early-onset seizures often had high levels of ammonia during crisis periods²⁰. In distal UCD such as ASLD and ARG1D, seizures usually present during later childhood and may not be directly related to HA episodes, suggesting the influence of other epileptogenic factors. An increase in arginine can lead to an increase in the epileptogenic guanidino compound in association with ARG²¹ and MRI spectroscopy in ASLD reveals an increase in guanidinoacetate, also epileptogenic²². The ASL conditional knockout mouse demonstrates increased seizure sensitivity by enhanced firing rate in the ASLD locus coeruleus resulting from a decrease in nitric oxide level and tyrosine hydroxylase activity.

In this study, we classified seizure semiology based on ILAE classification. The most common seizure types in neonatal onset were clonic, whereas focal seizures predominated in late-onset seizures. Three UCD individuals (OTCD, ARG1D, and ASLD individuals) had absence seizures with the EEG showing typical 3–3.5 Hz generalized spike discharges, albeit atypical in presentation given the presence of concomitant developmental delays. Previous publications have also reported absence seizures in 59% of ASLD and 66% of ARG1D^{15, 16}, but there is no report in individuals with OTCD. The pathogenesis of absence seizures in UCD is not well understood, but the ictogenesis of absence seizures in experimental models is largely due to increased inhibition of thalamocortical neurons²³. Magnetic resonance spectroscopy (MRS) in UCD shows a decrease in myoinositol in the thalamic nucleus⁴, with levels being inversely correlated with brain glutamine levels. Therefore, we hypothesize that involvement of the thalamus in UCD may be responsible for absence seizures. Recently, absence seizures have been increasingly identified in several genetic disorders, especially in diseases that require precision therapy including SCN8A related epilepsy (loss-of-function)²⁴, glucose transporter type 1 deficiency (GLUT1 deficiency)²⁵ and UCD. This article expands the inherited metabolic differential diagnosis that may present as absence seizure with developmental delay.

The overall prevalence of epilepsy in UCD has not been reported previously. We have found that 12% of people with UCD have epilepsy. Most epilepsy individuals in our study had distal urea cycle disorders (80%). The highest rate of epilepsy was observed in individuals with ASLD and ARG1D, consistent with a prior study finding that epilepsy affected 46–60% of ASLD and 66% of ARG1D^{15, 16, 26, 27}. The pathophysiology of development of epilepsy may result from brain injury due to prior HA crisis or the presence of other epileptogenic metabolites known to accumulate in those specific enzyme deficiencies.

Additionally, our findings suggest that neonatal seizures do not significantly increase the risk of developing future epilepsy. All seizures during the neonatal period were provoked by hyperammonemia and subsequently resolved. This information may help physicians decide whether or not to continue antiseizure medication for patients who present during the neonatal period. Similar to a prior study¹⁶, a few individuals with ASL and epilepsy had neonatal-onset seizures, but they did not correlate with an increased risk of developing epilepsy. Notably, the follow-up period in those with early onset seizures has not reached ten years in half of individuals; therefore, future research must investigate the relationship between epilepsy and other risk factors, a larger population, and an extended follow-up period. The relationship between epilepsy and neurocognitive disorders in those with UCD is another important area for exploration.

In our study we found that late-onset seizures pose a greater risk of progressing to epilepsy if they occur in the absence of a HA crisis. Our findings indicate that a single late onset seizure in distal UCD may warrant ASM since the risk of additional seizures exceeds 60%¹⁹. This would be congruent with the ILAE diagnostic criteria for epilepsy. The majority of patients in our study showed favorable responses to ASM monotherapy or polytherapy. Three out of ten individuals with epilepsy continue to experience seizures since their last documented follow up appointment. The response to antiseizure treatment differs among studies^{16, 26}.

Hyperammonemic crisis event

UCD individuals often present with HA crises. In the population studied, two-thirds of UCD individuals experienced crises. Encephalopathy was the most common neurological symptom at presentation, followed by clinical seizures, as observed in the previous study ²⁸.

Nonconvulsive seizures (NCS) have been observed in previous UCD case series during HA crises ^{9, 29}. During crises, individuals with UCD often experience lethargy, which can be caused by either hyperammonemia or a concurrent presence of NCS. As a result, NCS might be underrecognized in cases of hyperammonemic encephalopathy. Physicians might attribute the encephalopathy solely to hyperammonemia without considering that NCS could be the cause. Similarly, in patients with clinical seizures, even if symptoms subside after treatment, it is not always indicative of the absence of ongoing seizures. Therefore, Early detection and treatment of NCS are crucial because NCS can cause further neuronal damage. Additionally, EEG monitoring for NCS allows physicians to evaluate the effectiveness of treatments and adjust ASM, thereby improving prognosis and preventing neurological deterioration. However, there is no consensus surveillance protocol recommending the use of EEG to detect nonconvulsive status epilepticus in UCD individuals³⁰. In our retrospective study, EEG was not performed in all events. The decision to use EEG monitoring is based upon the suspicion of NCS or to monitor encephalopathy. Most individuals with overt seizures had EEG monitoring for the detection of NCS, but only a few presenting with encephalopathy without overt clinical seizures were monitored via continuous EEG. EEG was employed more frequently in events with high ammonia; half undergoing continuous EEG had ammonia concentrations more than 328 mmol/l. This factor could potentially impact the rate of NCS during crisis as observed in our research. Nevertheless, our findings indicate a high rate of NCS detection in these two groups (53% of crises with clinical seizures and 27% of crises of encephalopathy without clinical seizures). Moreover, the probability of seizure increases significantly as ammonia and glutamine concentrations rise. Therefore, we recommend continuous EEG monitoring for NCS detection in individuals with UCD who have overt seizures or encephalopathy with high ammonia levels. In our own center we have developed an algorithm for monitoring individuals with UCD and HA encephalopathy or seizures (unpublished, manuscript in submission).

The purpose of treatment during a HA crisis is to eliminate ammonia and other neurotoxic substances from the brain. Standard treatment of UCD involves the reduction of natural protein from the diet, addition of a nitrogen scavenger (sodium benzoate and sodium phenylbutyrate or phenylacetate), and supplementation with arginine or citrulline ^{2, 30}. During crises, neurological complications such as seizures and brain edema require immediate management to prevent significant morbidity or even mortality. In addition, adequate seizure control with ASM treatment is crucial for preventing further brain injury. The individuals in our study were controlled with ASM monotherapy or polytherapy for seizures, along with standard treatment for ammonia reduction.

Our research has some limitations. Based on our results, CPS1D and ARG1D have a higher rate of seizures, although this was not statistically significant. Due to the small number of certain UCD types, our ability to detect a statistical difference when comparing the occurrence of seizures and UCD types may be limited. As a result, further multicenter trials should be conducted. In addition, our retrospective study gathers the initial ammonia level at the time point closer to the presenting symptom during crises, but there are still some gaps in the time points between symptoms and the biochemical profiles. However, we can use the initial ammonia level as a predictor of seizure occurrence during crises. Further investigation is required to determine the optimal cut-point for EEG monitoring, and not all crisis events were monitored by EEG, suggesting that some seizures may be underdiagnosed. For these reasons, it requires the implementation of prospective research designs and the recruitment of larger sample sizes. It should be noted that certain cases of severe hyperammonemia may have convulsions after ammonia reduction, potentially attributed to brain edema ⁹.

Conclusions

Individuals with UCD are at risk for developing epilepsy, particularly if they present with late-onset seizures unrelated to a HA crisis. During HA events, there was a significant association between the levels of ammonia and the risk for occurrence of seizures. The measurement of initial ammonia and glutamine can help physicians consider the risk of seizures and implement appropriate neuromonitoring. This study also highlights the utility of EEG monitoring during HA crises for individuals with clinical seizures and/or encephalopathy with high ammonia levels to identify subclinical seizures and provide appropriate treatment. Understanding the prevalence of epilepsy in the UCD population is essential in understanding cognitive morbidity in these disorders.

Table 4:

Clinical Characteristics of the UCD Cases Diagnosed with Epilepsy

Individual number	Type	Sex	Age onset	Seizure onset	Crisis Events	Seizure type	Comorbid	EEG	Treatment	Current seizure
1	OTCD	M	5 D	7.5 Y	3	Absence seizure, Focal seizure, GTC	ID, ADHD	3–4 Hz generalized spike, normal background	LTG Failed-LEV	No seizure
2	ARGD	F	4.5 Y	4.5 Y	2	Focal seizure	ID, spastic diplegia	Focal spike over central, temporal region, Slow BG	LEV	No seizure
3	ASLD	M	2 Y	2.5 Y	>10	GT, GTC	ID	Slow BG	PHT, TPM	No seizure
4	ASLD	F	5 Y	5 Y	N/A	Focal seizure	ID, migraine, anxiety	Focal spike and focal slow over frontal region, Slow BG	LTG	No seizure
5	OTCD	F	5 Y	8.5 Y	3	Focal seizure	No	Normal	LEV	No seizure
6	ASLD	M	N/A	8.5 Y	0	GTC	ADHD	Focal spike over frontal, central region Slow BG	LEV	No seizure
7	ARGD	F	N/A	4 Y	2	GTC	ID, spastic diplegia	Normal	CBZ	No seizure
8	ASLD	F	6 Y	13 Y	N/A	Focal seizure, GTC	ID, headache	3–4 Hz generalized spike, Slow BG	PHB	Ongoing seizure
9	ASLD	M	N/A	7 Y	N/A	Absence seizure, GTC	ID, ASD	3–4 Hz generalized spike, Slow BG	TPM	Ongoing seizure
10	ARGD	M	N/A	8.5 Y	6	Absence seizure, Focal seizure	ID, Febrile seizure, spastic diplegia	2.5–3 Hz generalized spike, Slow BG, focal slow over frontal region	LEV Failed-TPM Past medication: ETX (response)	Ongoing seizure

ADHD, Attention-deficit/hyperactivity disorder; ARG1D, arginase 1 deficiency; ASD, Autistic spectrum disorder; ASLD, argininosuccinate lyase deficiency; BG, Background; CBZ, Carbamazepine; D, day; ETX, ethosuximide; F, female; GTC, generalized tonic-clonic seizure; ID, intellectual disability; LEV, Levetiracetam; LTG, Lamotrigine; M, male; N/A, Not available; Ornithine transcarbamylase deficiency, OTCD; PHB, phenobarbital; TPM, topiramate; UCD, Urea cycle disorders; Y, year

Highlight.

A late-onset seizure poses an epilepsy risk.

Distal UCD has a higher rate of epilepsy than proximal UCD.

Initial ammonia and glutamine levels help in seizure prediction.

EEG monitoring found subclinical seizures in half of clinical seizures crises.

Data Availability:

For any inquiries regarding further information, please contact the corresponding author at agropman@childrensnational.org.