Neonatal seizures: Etiologies, clinical characteristics, and radiological features: A cross-sectional study

Abstract

Seizures are a common clinical indication of central nervous system damage or abnormality in neonates. We aimed to identify the etiologies, clinical characteristics, and radiological features of neonatal seizures. This is a cross-sectional, retrospective, descriptive study using data obtained from the neonatal intensive care unit in King Abdulaziz Medical City (KAMC), a governmental, academic tertiary hospital in Riyadh, Saudi Arabia. The population of interest were neonates diagnosed with a neonatal seizure at KAMC between April 2015 and March 2019. A total of 61 patients with neonatal seizures were included in the study. The most common etiology was hypoxic-ischemic encephalopathy (43%). A total of 32 patients were full-term (52.5%). Around one-fifth of the study sample (21.3%) had a family history of neonatal seizures. Around 43.0% of the patients had epilepsy episodes. More than half of the patients (57.0%) were on one anti-seizure medication. Patients were followed up after 1 year, they had multiple comorbidities, including developmental delay, epilepsy, and cerebral palsy. Developmental delay was identified in 62.3% of the patients. A total of 19 patients have passed away (31%). Neonatal seizures are a common manifestation of neurologic disorders in neonates and are associated with high morbidity and mortality. Therefore, early identification of seizure etiology and proper management may help to improve the outcome.

1. Introduction

Seizures are a common clinical indication of central nervous system damage or abnormality in neonates.^[1] The incidence of clinical seizures in infants is about 1 to 3 per 1000.^[2] According to a previous study, the majority of newborn convulsions are attributable to known causes and are seldom associated with neonatal-onset epilepsy.^[3] Specific etiologies of seizure activity, such as metabolic diseases, are difficult to detect and, as a result, are often overlooked or misunderstood. Seizures associated with inherited metabolic disorders, for instance, provide a massive challenge to physicians due to their poor responsiveness to antiepileptic medicines.^[4] In addition, electroencephalographic findings are not specific, and other tests, such as magnetic resonance imaging, may yield normal results, making it difficult to establish a proper diagnosis.^[5] There are several neurodevelopmental prognostic variables among term newborns with seizures. There is evidence that late-onset seizures are linked with improved outcomes.^[6] Neonatal seizures were substantially related to an increased chance of developing epilepsy, cerebral palsy, headaches, and intellectual impairments.^[7,8]

Neonatal electroencephalography (EEG) is very important for the diagnosis of neonatal seizures since the majority of neonatal seizures are not noticeable.^[9] Previous literature reported that in epilepsy, EEG sensitivity ranges from 25% to 56%, which is relatively low.^[10] Some individuals experience symptoms resembling those of an epileptic seizure, but without any abnormal electrical activity in the brain.^[10,11] These are clinical manifestations that resemble seizures but are typically not epileptic in nature. These seizures, which are caused by mental duress or a physical condition, could be termed non-epileptic.^[10,11] For this reason, EEG background patterns have been used in the etiological categorization of newborns, particularly term neonates with seizures. Multiple studies have indicated that frequent electrographic seizures in newborns are substantially associated with severe brain damage and major developmental delay.^[12,13]

Acute provoked seizures and early-onset epilepsies are distinct seizure types with distinct underlying causes, clinical manifestations, and prognoses. Acute provoked seizures are triggered by specific factors (such as head trauma, fever, metabolic imbalances, and infections, or an isolated event such as brain injury).^[14,15] They occur in response to the imminent presence of the provocateur and are typically isolated occurrences of any age. Once the triggering factor is treated or eliminated, these seizures typically cease.^[15] EEG findings may be normal or may reveal abnormalities related to the underlying cause.^[15] Early-onset epilepsies, on the other hand, are a group of epileptic disorders that typically manifest in infancy or early childhood. Typically, developmental or genetic factors abnormalities. Early-onset epilepsies may manifest as recurrent seizures that occur over a prolonged period and commence in childhood (typically before the age of 3).^[14,16,17] In early-onset epilepsies, EEG recordings frequently reveal aberrant brain activity (including spikes, sharp waves, and other epileptiform discharges, even between seizures).^[16,17]

Inborn errors of metabolism (IEM) are an uncommon cause of newborn seizures compared to other causes.^[18] Due to an increased demand for glucose and oxygen, metabolic diseases have been related to brain changes such as cerebral hyperperfusion.^[19] In the newborn era, the diagnosis of IEM remains problematic. Not all IEMs are screened for in newborns. Neonatal convulsions resistant to medication should raise the possibility of IEMs.^[20] According to studies, inherited metabolic disorders are often linked with a bad prognosis.^[2]

Previous literature has demonstrated that the APGAR score is useful for defining the birth status of newborns and predicting their outcomes has been validated by research.^[21] In addition, it has been established that an APGAR score of less than 3 at 5 minutes of life is predictive of poor neurologic outcomes.^[22] This research aims to characterize the etiology, clinical features, and radiological characteristics of neonatal seizures.

2. Methods

2.1. Study design

This is a cross-sectional, retrospective study that was conducted at the neonatal intensive care unit in King Abdulaziz Medical City (KAMC). KAMC is a governmental, academic tertiary hospital in Riyadh, Saudi Arabia. The data were extracted for the period between April 2015 and March 2019.

2.2. Study population

Study subjects were recruited from the neonatal intensive care unit in KAMC. The population of interest in this particular study was the newborn children diagnosed with a neonatal seizure. The exclusion criterion was if the seizure occurred after 29 days of life.

2.3. Data collection

The etiologies, clinical characteristics, and radiological features of the patients were collected from the electronic medical records of the patients. The neonatal seizure in this study was diagnosed as per Volpe classification schema. It is a widely used classification tool for classifying seizures. Volpe classification schema is based on the paroxysmal clinical phenomena documented in medical records by consulting pediatric neurologists, attending neurologists, house staff, or neonatal nursing staff.^[23] In our study, the seizure types were categorized according to Volpe classification schema, meaning that the paroxysmal clinical phenomena documented in medical records (by consulting pediatric neurologists, attending neurologists, house staff, or neonatal nursing staff) were used to classify the seizure types based on the schema. Only the most prominent seizure type was considered for the study in cases where there were several seizure types. Electroencephalogram, electrolyte, and brain imaging examinations were performed on infants who had seizures if their clinical situation allowed it. All of the enrolled neonates had follow-up by phone call, during which developmental and seizure histories were gathered. The National Guard seizure protocol was used as the standard protocol for the management and investigation of neonatal seizures.

2.4. Ethical approval

This study was approved by the institutional review board at King Abdullah International Research Center in Riyadh, Saudi Arabia.

2.5. Data analysis

Summary statistics were presented as the mean (SD, standard deviation) for continuous variables and as frequencies and percentages for the categorical data. The data analysis was completed using SPSS Statistics Version 27.

3. Results

3.1. Patients’ baseline characteristics

Table 1 below presents the baseline characteristics of the patients. A total of 61 patients fit the inclusion criteria and were included in the study. Almost half of the patients (57.4%; n = 35) were males, and 32 were term births (52.5%). The median gestational age was 38 weeks (interquartile range (IQR): 23–39). The remaining 48.0% of the patients were preterm births (24.6% late preterm, 23.0% extreme preterm). The median birth weight was 2.75 kg (IQR: 0.5–3.1), and the median birth head circumference was 34.0 cm (IQR: 20.5–35.0). The average APGAR score at 5 minutes was 4 (SD: 2.6). More than half of the births (62.3%) were by cesarean section, and 36.1% were spontaneous vaginal births.

Table 1.

Patients’ baseline characteristics.

Variable	Category	Frequency	Percentage
Etiology (gestational age comorbidities)	Hypoxic-ischemic encephalopathy	26	42.6
	Intracranial hemorrhage	12	19.7
	Genetic syndromes	10	16.4
	Brain malformation	2	3.3
	Inborn errors of metabolism	1	1.6
	Others/unknown	10	16.4
Gestational age group
	Term	32	52.5
	Late preterm	15	24.6
	Extreme preterm	14	23.0
Gender
	Male	35	57.4
	Female	26	42.6
Family history
	Yes	13	21.3
	No	41	67.2
Pregnancy
	Eventful	40	65.6
	Uneventful	20	32.8
	Unknown	1	1.6
Birth history
	Spontaneous vaginal delivery	22	36.1
	Cesarean section	38	62.3
	Unknown	1	1.6
Resuscitation history at birth
	Oxygenation and bagging	7	11.5
	Positive pressure ventilation	14	23.0
	Intubation	33	54.1
	Unknown	7	11.5
Fetal deceleration		7	14.9
Congenital anomalies		14	23.0
Dysmorphism		13	21.3
Head size
	Normal	45	73.8
	Microcephaly	5	8.2
	Macrocephaly	4	6.6
EEG background
	Normal	24	39.3
	Abnormal	29	47.5
	Not Done	8	13.1
EEG epileptiform discharges		20	32.8
EEG burst suppression		9	14.8
Treated with cooling		20	32.8
Continuous variables		Mean/median	Standard Deviation/interquartile range
Gestational age (wk) (Median)		38.0	23–39
Apgar score at 5 min (Mean) (n = 60)		4.0	2.6
Cord pH (Mean) (n = 22)		6.9	0.1
Cord CO2 (Median) (n = 22)		76	0–88.8
Cord deficit (Mean) (n = 17)		−11.8	5.7
Blood gas pH (Mean) (n = 52)		7.1	0.2
Blood gas CO2 (Median) (n = 52)		49	23–55.25
Gas base deficit (Median) (n = 52)		−8.9	−27.0 to 4.38
Seizure onset (d) (Median) (n = 58)		1.0	1–5.25
Birth weight (kg) (Median) (n = 60)		2.75	0.5–3.1
Birth head circumference (cm) (Median) (n = 59)		34.0	20.5–35
Expired after (d) (Median) (n = 19)		5	1–16.5
Head circumference at follow-up (cm) (Mean) (n = 36)		43.75	23.5–47
Age at follow-up (Median) (n = 38)		13	1–23
Current age (mo) (Median) (n = 41)		31.0	14–37

EEG = electroencephalography.

The most common etiologies for the patients were hypoxic-ischemic encephalopathy, and intracranial hemorrhage, accounting for 42.6% and 19.7%, respectively (Fig. 1).

Figure 1.

Etiologies of neonatal seizure.

Almost one-fifth of the patients (21.3%) had a family history of seizure. At birth, more than half of the patients (54.1%) needed intubation, and 23.0% needed positive pressure ventilation (PPV). Fetal declaration was reported for 7 patients. Almost one-quarter of the patients (23.0%) were diagnosed with congenital anomalies. A similar proportion of the patients (21.3%) were diagnosed with dysmorphism. Figure 2 presents the etiology and associated gestational age comorbidities. The head size was normal for most of the patients (73.8%). The EEG background was normal for only 39.3% of the patients. EEG epileptiform discharges were identified in 32.8% of the patients, and 9 of them had EEG burst suppression.

Figure 2.

Etiology and associated gestational age comorbidities of the group (1: full term, 2: late preterm, 3: extreme preterm).

3.2. Seizure history

Table 2 below presents seizure profiles among the patients. The overall mean time for seizure onset was 1.0 days (IQR: 1–5.25). For the number of seizures before treatment, 45.9% of the patients had 1 seizure, 24.6% had 2 to 4 seizures, 16.4% had 5 to 10 seizures, and 3 patients had more than 10 seizures. The main type of seizure was focal tonic (36.1%), followed by clonic (23.0%), lip smacking/other (19.7%), myoclonic (seven patients), and subtle (two patients).

Table 2.

Patients seizure profile.

Variable	Category	Frequency	Percentage
Number of seizures before treatment
	1	28	45.9
	2–4	15	24.6
	5–10	10	16.4
	>10	3	4.9
	Unknown	1	1.6
Seizure type
	Tonic	22	36.1
	Clonic	14	23.0
	Myoclonic	7	11.5
	Subtle	2	3.3
	Other	12	19.7
Encephalopathy
	None	12	19.7
	Mild	18	29.5
	Moderate	16	26.2
	Severe	8	13.1
Tone
	Normal tone	23	37.7
	Hypotonia	26	42.6
	Hypertonia	1	1.6

Figure 3 presents seizure types in each etiology category. The seizure tone was identified as normal for 37.7% of the patients, 42.6% as hypotonic, and only 1 patient as hypertonic.

Figure 3.

Seizure types in each etiology category.

3.3. One year follow up and outcomes

Figure 4 presents the developmental outcome in each etiology category. The median follow-up age was 13.0 months (IQR: 1–23) for patients who survived the neonatal period. Developmental delay was identified in 62.3% of the patients. Around 43.0% of the patients had epilepsy episodes. Concerning cerebral palsy, at follow-up, 39.3% of the patients were diagnosed as spastic quadriplegic. Nineteen patients died (31.0%): 4 due to hypoxic ischemic encephalopathy (HIE), 4 due to genetic/metabolic causes, 2 with brain malformations, 1 with intracranial hemorrhage, and 7 from unknown causes. Ten deaths occurred during the first month of life, and the remaining 9 in the first year of life. Table 3 below presents the patients outcomes after 1 year of follow-up.

Figure 4.

Developmental outcome in each etiology category.

Table 3.

Patients outcomes at 1 yr of follow-up.

Variable	Category	Frequency	Percentage
Developmental delay in 1 yr
	None	23	37.7
	Mild	11	18.0
	Moderate	14	23.0
	Severe	13	21.3
Expired		19	31.1
Epilepsy		26	42.6
Cerebral palsy	No	20	32.8
	Spastic quadriplegic	24	39.3
	Hemiplegic	9	14.8
	Diplegic	8	13.1

3.4. Acute anti-seizure medication

In total, 57% of the patients received 1 drug, 28% 2 drugs, 9 patients 3 and more. Phenobarbital was administered to 60 patients (98.4%), phenytoin to 19 (31.1%), levetiracetam to 13 (21.3%), midazolam to 3, and clonazepam to one. In all, 31 patients were discharged on maintenance therapy, while 15 patients did not need maintenance therapy.

4. Discussion

Seizures are the most conspicuous manifestation of neurologic dysfunction in the newborn era. It is crucial to determine the underlying cause of newborn seizures. Etiology dictates prognosis and outcome and directs therapy techniques.^[23] Accurate diagnosis may lead to etiology-specific treatment and may prevent central nervous system impairment that might otherwise arise if the underlying illness were left untreated. It may be important to treat the underlying cause of seizures in order to manage the seizures themselves.^[24] In our study, HIE was the most common seizure etiology among late preterm and term neonates (42.6%), whereas intracranial hemorrhage was more common among extremely preterm neonates (19.7%), which is similar to other internationally reported studies.^[3] HIE is a form of brain dysfunction caused by a reduction in oxygen or blood supply to the brain. HIE may develop before, during, or after labor and delivery.^[25,26] The length of time the brain is deprived of oxygen or blood flow may affect the severity of a brain injury. Children with HIE may have no long-term repercussions. Some may have mild to moderate impairments as a result of HIE, while others may be severely disabled. A HIE-related brain injury may result in developmental delay, cognitive disability, cerebral palsy, or epilepsy. As a baby matures, the symptoms of HIE may become increasingly apparent.^[25,26]

In our study, genetic and metabolic causes accounted for 16.4% of all causes of neonatal seizures, and these were primarily among preterm neonates. It is usually challenging to discern a significant etiological difference between genetic, metabolic, and structural forms. Metabolic diseases and deformities (structural abnormalities) do, in fact, have a genetic basis, despite the fact that it is not always simple to identify the underlying pathogenic mutation. Decoupling the etiologies may consequently be difficult or perhaps inappropriate.^[27] Regarding this, the most recent International League Against Epilepsy Task Force on Neonatal Seizures advocated a reorganized, less restrictive, and nonhierarchical diagnostic categorization.^[28] Multiple categories (genetic, metabolic, structural, immunological, and unknown) that are no longer regarded as mutually exclusive may be used to classify epilepsy. Since genetic, metabolic, and structural factors account for 10 to 15% of all newborn seizures, this technique is important for giving an accurate diagnostic evaluation, enabling a sufficient and early treatment start, and reducing adverse outcomes.^[1,16,29,30]

In our study, individuals with HIE, cerebral bleeding, and metabolic/genetic factors were more likely to have poor outcomes, comparable to the results published by Baudou in 2019.^[31] A total of 36.8% of the deceased had HIE, 5.3% had cerebral hemorrhage, 36.8% had genetic syndromes, 10.5% had brain malformations, and 5.3% had inherited metabolic disorders. Despite the fact that 73.1% of those with HIE and 91.7% of those with cerebral hemorrhage survived, we observed the opposite tendency, with 70% of those with genetic disorders and 100% of those with IEM (n = 1) and brain malformations (n = 2) passing away.

In terms of gestational age, 69.2% of patients with HIE were full-term, 75% of patients with cerebral hemorrhage were very preterm, and 50% of patients with genetic disorders were full-term; and 30% were late preterm neonates. The 2 individuals with brain malformations were a term baby and a late preterm infant. A late preterm infant had an inherited metabolic disorder. A previous study in Russia examined the psychomotor development of newborns with neonatal seizures (NS) born at various gestational ages.^[32] This study included 52 infants with neonatal seizures who were born at various gestational ages. Only 32.7% of the infants evaluated had no developmental delay on any of the 5 Bayley-III scores. On at least 1 measure, 62.3% of patients had a significant developmental delay (score composite >= 70).^[32] Another study compared neurodevelopment and social-emotional development between moderate and late preterm (MLPT) infants and term-born control infants at age 2 years.^[33] Compared with term-born infants, MLPT children showed lower cognitive, linguistic, and motor development at age 2 years. Compared to term-born infants, MLPT infants had increased risks of developmental delay, with adjusted odds ratios of 1.8 (95% CI, 1.1–3.0) for cognitive delay, 3.1 (95% CI, 1.8–5.2) for linguistic delay, and 2.4 (95% CI, 1.3–4.5) for motor delay.^[33]

In our study, congenital abnormalities were found among 16% of HIE patients, 27.3% among intracranial hemorrhage patients, 75% among genetic syndromes, and 50% among brain malformations, but were not evident in the 1 patient with IEM. Similarly, dysmorphism was evident in only 8.7% HIE cases, none with intracranial hemorrhage, 70% among genetic syndromes, 1 (50%) with brain malformations, and none among IEM. Preterm newborns have a distinct etiological profile than full-term infants.^[34] Hypoxic-ischemic encephalopathy is the most common cause of newborn convulsions in full-term infants, followed by localized ischemia (stroke), brain abnormalities, and metabolic disorders. The majority of seizures recorded in premature infants are caused by intraventricular hemorrhage and infections.^[34]

In our study, the mean APGAR score was 2.19 (SD: 1.98) for HIE, 5.17 (SD: 2.04) for intracranial hemorrhage, 5.20 (SD: 2.53) for genetic syndromes, and 6.0 for both cases of brain malformation, and the single case of IEM. The mean gestational ages were between 36 and 38 weeks for all etiologies except intracranial hemorrhage, which was 28.1 weeks (6.5 SD). The relationship between Apgar score and gestational age has been shown in prior research. Apgar scores are correlated with the maturity level of newborn infants.^[35,36] The 1- and 5-minute Apgar scores correlated closely with gestational age. Respiratory efforts, muscular tone, and reflex were the primary predictors of a diminishing Apgar score as gestational age decreased.^[35]

Multiple studies recommend a scoring system for prognostic evaluation following newborn convulsions. One of the scores is dependent on birth weight, status epilepticus, the Apgar score, the neurologic examination at seizure start, cerebral ultrasonography, and the success of anticonvulsant medication.^[16] Another study found a correlation between gestational age, seizures, abnormal EEG, underlying etiology, and the quantity of medicines used to manage the seizure.^[37]

This research has a number of limitations that need attention. First, the sample size of the research was quite small. Second, some patients were born outside of the hospital, and incomplete information was acquired. Lastly, several individuals passed away before the cause of their newborn convulsions was identified. Therefore, it is advised to conduct prospective research with more regular genetic testing and longer follow-up periods.

5. Conclusion

This research explored the causes, clinical manifestations, and radiological manifestations of newborn seizures. Seizures in infants may be indicative of serious underlying medical conditions requiring immediate treatment. It indicated that seizures are a frequent symptom of neurologic abnormalities in newborns and are linked with significant morbidity and death. Investigating the causes and symptoms can aid in the identification of these conditions, guiding the healthcare team in the development of a comprehensive treatment and management strategy. Therefore, early diagnosis of the seizure cause and appropriate care may contribute to an improved prognosis.

Acknowledgments

King Abdullah International Medical Research Center (KAIMRC), Ministry of National Guard, Riyadh, Saudi Arabia.

Author contributions

Conceptualization: Mohammed Almuqbil.

Data curation: Mohammed Almuqbil.

Formal analysis: Mohammed Almuqbil.

Funding acquisition: Mohammed Almuqbil.

Investigation: Mohammed Almuqbil, Yousof Alrumayyan, Shahad Alattas, Duaa Baarmah, Waleed AlTuwaijri, Ahmed AlRumayyan, Mohammed Tala AlRifai, Asma Al Madhi, Hassan Al-shehri, Saif Alsaif.

Methodology: Mohammed Almuqbil.

Project administration: Mohammed Almuqbil.

Resources: Mohammed Almuqbil.

Software: Mohammed Almuqbil.

Supervision: Mohammed Almuqbil.

Validation: Mohammed Almuqbil.

Visualization: Mohammed Almuqbil.

Writing – original draft: Mohammed Almuqbil.

Writing – review & editing: Mohammed Almuqbil, Yousof Alrumayyan, Shahad Alattas, Duaa Baarmah, Waleed AlTuwaijri, Ahmed AlRumayyan, Mohammed Tala AlRifai, Asma Al Madhi, Hassan Al-shehri, Saif Alsaif.