Diagnostic Dilemmas and Therapeutic Challenges in Neonatal Seizures: A Neurology Perspective

1. Introduction

Seizures occur more frequently in the neonatal period than at any other time in life. Seizures and epileptic burden in the neonatal period- is correlated to neurodevelopmental outcomes and overall prognosis [1,2] Accurate diagnosis is essential for appropriate management; however, there are many diagnostic dilemmas. Seizures in the neonatal period are difficult to diagnose accurately. Many seizures are electrographic only with no clinical correlates, different kinds of paroxysmal events in neonates may mimic electroclinical seizures, and diagnosis based on clinical phenomena alone may result in inaccuracies [3,4,5]. There is no universally accepted definition of a neonatal seizure or neonatal status epilepticus [6,7,8,9] (Conventional EEG (CEEG) is the gold standard for diagnosis of a neonatal seizure. Continuous conventional video EEG monitoring (cVEEG) for 24 to 48 hours may be needed for detecting seizures in those at risk, to monitor therapy and to assess epileptic burden; it is not available for clinical care in most parts of the world [10,11,12,13,14,15].

There are many therapeutic challenges in the treatment of seizures in the neonate. Most neonatal seizures are provoked or due to symptomatic aetiology. Genetic, metabolic and structural epilepsies are less frequent causes for neonatal seizures. The treatment paradigm is often similar initially, till a specific aetiology is suspected or identified. When to treat, what to treat with, how much to treat, and how long to treat are still being debated, with no uniform strategy. Early distinction between provoked seizures and neonatal onset epilepsy is important to guide therapy. The emergence of targeted therapies (some quite expensive), while exciting, has added to the therapeutic challenges and inequities associated with management of neonates with seizures.

This review addresses many of the diagnostic dilemmas and therapeutic challenges encountered by a clinician in the management of neonates with seizures.

Figure 1. Challenges in diagnosis and management of neonatal seizures. EEG shows an electrographic seizure from the left hemisphere. EEG is double banana montage, calibration bars in figure. Picture shows baby having a CEEG. Abbreviations: ASM: Antiseizure medications, CEEG: conventional EEG.

Figure 1. Challenges in diagnosis and management of neonatal seizures. EEG shows an electrographic seizure from the left hemisphere. EEG is double banana montage, calibration bars in figure. Picture shows baby having a CEEG. Abbreviations: ASM: Antiseizure medications, CEEG: conventional EEG.

1.1. Classification of Seizures in Neonates

Seizures in neonates have unique EEG signatures and are predominantly electrographic only [10,16,17,18]. Clinical features, when present can be subtle and may be missed in a busy neonatal intensive care unit. Non-epileptic paroxysmal movements may mimic electroclinical seizures [5] and even experienced health professionals with expertise in neonatal seizures may be incorrect. The ILAE classification of Seizures and the Epilepsies: Modification for seizures in the neonate [6] addresses many of the issues identified in application of the previous classification schemes to seizures in the neonate [19,20,21]. It is the only ILAE classification scheme that recognises electrographic only seizures as a type of seizure. The classification emphasizes the need for a clinical event to have an EEG correlate to call it an epileptic seizure. This may protect babies with seizure mimics from receiving unnecessary antiseizure medications which may have acute and long-term adverse effects.

1.2. Definition of Neonatal Seizure

Substantial variability occurs in the definition of a neonatal seizure in the literature. Conventionally a neonatal seizure was defined as one in which a rhythmic EEG discharge lasted 10 seconds or longer, had an identifiable start and finish, and could exhibit variable morphology, frequency, and field [22,23,24]. Electrographic only seizures were those without clinical correlates, and electroclinical seizures had associated clinical phenomena [16,25]. However, the position paper by the ILAE Task Force on Neonatal Seizures [6] does not nominate a specific duration for an electrographic event to be called a seizure. The relevance of BERDS (brief EEG rhythmic discharges lasting less than 10s), which is considered as mini seizures by some groups, is considered as part of the ictal/interictal continuum [26,27]. The lack of a defined duration in the latest EEG definition of a seizure has the advantage of being able to include seizure types like myoclonic seizures and infantile spasms that often have an ictal discharge of less than 10s duration. However, lack of uniformity in identification of a seizure poses difficulties for clinical research studies and for comparison between future and past studies [11]

1.3. Definition of Neonatal Status Epilepticus

The lack of a universally accepted definition of neonatal status epilepticus (NSE) makes diagnosis and management challenging [7]

The definition of the ILAE task force on Classification [28]) of Status epilepticus is conceptual, with two operational dimensions. The first is the length of the seizure and the time point (t1) beyond which the seizure should be regarded as “continuous seizure activity.” The second time point (t2) is the time of ongoing seizure activity after which there is a risk of long-term consequences. In the case of convulsive (tonic-clonic) status epilepticus. both time points (t1 at 5 minutes and t2 at 30 minutes) are based on animal experiments and clinical research. The relevance and applicability of these time points to neonatal status epilepticus, based on current knowledge, is unclear [7] (nagarajan and ghosh, 2025). The ILAE task force on classification of seizures, modified for neonates [6], does not address this issue. Based on a review of outcomes, the likelihood of spontaneous seizure termination, and probability of response to antiseizure medications, it has been proposed that any single neonatal seizure lasting more than 5 minutes be considered the equivalent of t1-early neonatal status epilepticus [8]).

A scoping review by the ILAE Neonatal Task force [9] of the available literature demonstrated substantial variations in definition of neonatal status epilepticus. The review concluded that the common definitions were based around a 30-minute seizure duration criterion, but evidence was insufficient to support a 30-minute cut-off to define prolonged seizures or indicate that seizures exceeding this burden were more likely to be pharmaco-resistant or associated with worse outcomes. Some studies suggest a seizure burden of 10-13 minutes may be a more realistic definition of t2. The neonatal task force intends to develop a standardised approach to assessing and describing neonatal seizure burden and defining NSE.

2. Diagnosis of Neonatal Seizures

Neonatal seizures are reliant on EEG for diagnosis. Conventional EEG (CEEG) is the gold standard for diagnosis and monitoring, with amplitude integrated EEG (aEEG) being a less accurate but more widely available and used tool.

CEEG vs aEEG

CEEG is the gold standard for diagnosis and monitoring of neonatal seizures – both electrographic only and electroclinical seizures [6,7]. (Most CEEG machines and studies are based on synchronised Video-EEG (VEEG). Continuous monitoring with VEEG (cVEEG) over 24 to 48 hours is used for detecting seizures (when a brief recording may miss them), and to monitor therapy and the overall epileptic burden. cVEEG may be also useful to identify infrequent non-epileptic events and to estimate the seizure burden. However, cVEEG is not available in most neonatal intensive care units across the world. Expertise in acquiring and interpreting neonatal EEG is important as neonatal seizures may have unique EEG signatures. Normal background patterns vary, are dependent on gestational and chronological age, and influenced by therapeutic interventions [10,22,29,30,31,32] .Amplitude integrated EEG (aEEG) is a simplified tool that provides a visual trend over time [10,33] and is used far more commonly in clinical practice. A recent Cochrane review [34] comparing aEEG with CEEG for detection of neonatal seizures concluded aEEG has only moderate sensitivity and specificity for detecting ‘neonates with seizures’, and its ability to detect ‘individual seizures’ varies widely. aEEG may not be sufficiently accurate for diagnosing neonatal seizures as it can under-diagnose or over-diagnose seizures and lead to inadequate or unnecessary treatment with antiseizure medications[35,36,37,38]. Use of CEEG monitoring will potentially optimise antiseizure medication use in neonates and help prognosticate outcomes [35,39] In clinical practice often combinations of standard recordings (<I hour) and more prolonged recordings are undertaken. More recently quantitative analysis of EEG (QEEG) and trend analysis have become available to clinicians; they condense hours of data into functional maps and graphs based on frequency, amplitude and rhythmicity changes [40]. Combined displays of aEEG[41] and CEEG, ECG, along with QEEG and trend analysis will promote greater interaction between neonatologists and child neurologists and is likely to contribute to better management and outcomes for neonates with seizures.

A critical review of continuous conventional EEG (cVEEG) for seizure detection in the intensive care units highlighted the gaps in evidence, practical problems, funding shortfalls and areas of future [11]research [13]. It concluded that while cVEEG detects electrographic seizures in a significant proportion of at-risk neonates, children, and adults in the ICU, conferring poorer neurological outcomes, and guiding treatment in many settings, the health economic benefits of treating such seizures remain to be proven. The American Clinical Neurophysiology Society review [14] identified that while there was evidence for the accuracy of cVEEG for seizure diagnosis and prognosis formulation in a variety of conditions, there were gaps in evidence directly demonstrating that cVEEG use improves long-term outcomes for neonates.

Optimising cVEEG resource allocation is important [11,42] as cost, personnel and equipment requirements are high. A review on EEG monitoring in critically ill children [43,44] (concluded that cVEEG monitoring has the potential to be cost effective if electrographic status epilepticus identification and management improves patient outcomes by as little as 3%. Their study supported monitoring of critically ill children for 24 hours because the cost to identify a child with electrographic seizures was modest. It is important for child neurologists, and all health professionals involved in neonatal care to advocate for availability of CEEG and cVEEG, and expertise to interpret it, across the world.

Neonatal automated background analysis and seizure detection models along with advances in machine and deep learning may help reduce cost and make cVEEG monitoring more feasible [15,18,45,46,47,48,49].. Artificial intelligence may play a significant role in future in integrating biological models, neurophysiology, neuroimaging, genetics, neurobiochemical markers, with clinical profile for better outcomes for babies with seizures. Implementation of digital health strategies incorporating real-time and continuous video-aEEG/EEG monitoring has the potential to improve diagnostic accuracy for electrographic seizures, optimize antiseizure medication stewardship, and inform early neuroprotective interventions [12,50,51,52].

Neonatal neurocritical care (NNCC) has emerged as a specialised discipline to address the neurological needs of babies that are critically ill with impairment of brain function. Remote NNCC, supported by telemedicine and digital health tools, offers a promising approach to extend specialized neurological care beyond tertiary centres, particularly to underserved or resource-limited settings [51,53,54].

2.1. Role of EEG in Identifying Aetiology and in Prognostication in Neonatal Seizures

Distinctive EEG patterns and electroclinical corelates have been described for gene related, metabolic and structural epilepsies [3,30,55]. The CEEG background may help prognostication and monitoring of responses to interventions [22,56,57].

Multimodal neuromonitoring in NICU, that includes EEG, ECG, blood pressure, cerebral tissue oxygenation and respiratory parameters, has been shown to be useful in predicting outcomes, and developing bundles of care for neonates with encephalopathy and seizures [7,30,58,59].

2.2. Aetiology

The therapeutic interventions for neonates with seizures is initially often similar irrespective of aetiology [7,60,61,62] However, it is essential to distinguish symptomatic seizures from epilepsies starting in the neonate [63] to individualise treatment bundles. Hypoxic ischaemic encephalopathy, neonatal strokes, intracranial bleeds, infections and metabolic disturbances remain the most frequent causes of seizures in the neonate [23,60] Genetic, metabolic and structural aetiologies (or combinations) are important to recognise early [64,65]. Children with post-neonatal epilepsy had a higher burden of pathogenic variants in epilepsy-associated genes compared to those without post-neonatal epilepsy [66]. Although targeted interventions such as disease specific and disease modifying treatments remain limited to a few conditions at present (Ziobro 2021), recent advances with unique approaches such as small molecules, cell therapy and other forms of genetic therapy are exciting[67,68]. Newborn screening programmes for metabolic disorders has been developed for many years to identify neonates with severe but treatable conditions [69] The feasibility of genome sequencing using DNA from dried blood spots for early actionable conditions is being explored [70] The challenges in this area are the ethical and social issues, limited availability across the world of appropriate genetic testing and newborn screening, the time to an aetiological diagnosis, absence of precision therapy for most conditions, and impact of a genetic diagnosis on the family. This sometimes leads to frustration, families seeking alternate therapies and anger at health systems and professionals [71,72,73,74].

3. Therapeutic Challenges

3.1. Treatment with Anti-Seizure Medication

3.1.1. What and When Should You Treat

There is consensus that both electrographic only and electroclinical seizures need treatment [1,7,74,75,76].

When should one start an antiseizure medication (ASM): should it be at the first seizure, or only after recurrent or prolonged seizures? There is no known optimal time to start antiseizure medication. Treatment may be time critical, so most neonatal seizures are treated when diagnosed [17,33,62,76,77,78].

Neonatal seizures usually last 1-5 minutes in duration[16,79]. Higher seizure burden and ictal fractions (percentage of recording with EEG seizure activity) have been associated with increased mortality and adverse neurodevelopmental outcomes [80,81,82,83] . However, a precise seizure duration beyond which a neonatal seizure is unlikely to self-terminate has not been established. In estimating seizure burden various metrics are included in different studies –longest seizure, ictal fraction, ictal EEG correlates of the seizure, number of electrodes involved, characteristics of the ictal discharge, the amount of brief rhythmic discharges, total duration of seizures and epileptic burden [6,7,8,26].

Balancing the risk of a high seizure burden on neurodevelopmental outcome versus possible neurotoxic effects of ASM continues to remain a challenge [17,52,84,85,86,87] . Exposure to antiseizure medication may induce an increase in neuronal death, upset the balance between survival and death of neurons in the developing brain, and result in maladaptive plasticity. Chronic exposure to ASM may result in comorbidities such as depression, anxiety, And impairments in memory and learning. [88]

3.1.2. With What to Treat

Phenobarbital is considered the first line therapy for neonatal seizures [7,60,62,75,78] irrespective of aetiology, based on current evidence. In specific situations some studies suggest that an alternative ASM may have more efficacy: for example phenytoin for seizures with arterial ischaemic stroke, levetiracetam in very low birth weight infants [89,90] and phenytoin, carbamazepine or other sodium channel drug if a channelopathy is suspected.

At the onset of seizures, often along with or after first ASM, neonates need initiation of bundles of care: cardiorespiratory stabilisation, correction of metabolic perturbances such as glucose or electrolyte abnormalities, starting antibiotics if appropriate and undertaking tests to establish aetiology (EEG monitoring, structural, metabolic, genetic).

If seizures have not stopped with the first ASM, phenytoin, levetiracetam, midazolam or lignocaine may be used as second line drugs (levetiracetam may the preferred second line in neonates with cardiac dysfunction), in accordance with ILAE guidelines [60] A trial of pyridoxine may be warranted if Vitamin B6 dependant epilepsy is a possibility. If seizures continue, then a wide range of options such as vigabatrin, topiramate, lacosamide, perampanel, and ketogenic diet are often considered [7,76].

When second line ASMs are not effective aetiology specific interventions often become more relevant – removal of toxins, supplementation of missing cofactors, dietary interventions (ketogenic diets, elimination or addition of specific nutrients). Targeted therapy considerations such as Vigabatrin in some babies with Tuberous sclerosis (TSC) for prevention and treatment of neonatal seizures, quinidine in KCNT1 related DEE (Developmental and Epileptic Encephalopathy), ganaxolone for CDKL5 related DEE, and gene therapies may be available for some aetiologies in some parts of the world [1,64,67,75,76,91,92].

Seizure cessation is the primary aim of treatment. Efficacy of antiseizure medications are evaluated for seizure free outcomes or a predetermined (often 50%) reduction in seizures The treatment goals need to be better defined and individualized, based on short and long-term outcomes. Long term neurodevelopmental outcomes, though important, are difficult to predict in the acute situation and are also dependant on aetiology. Future trials [93] ( soul 26), with innovative designs, may explore other features such as postmenstrual age, aetiology, timing of seizures and other seizure metrics for analysing seizure burden and meaningful reduction with antiseizure medication.

3.1.3. How Long to Treat

When to stop ASM in symptomatic seizures is important and often a balancing act. ASMs maybe harmful to the neonatal brain while seizures are also harmful [94]).

It is safe for most neonates to stop ASM after resolution of acute provoked seizures, prior to discharge from the hospital, often in the first week [60,94,95]. A small number of neonates, for example with severe HIE, hypoglycaemia or infections may never stop seizing and progress to epilepsy in the neonatal period, needing continuation of ASM [23,24,94].

Neonates with genetic developmental and epileptic encephalopathies (DEE), cortical malformation disorders with early onset of seizures, metabolic disorders with seizures as a manifestation, also need continuation of ASM or other interventions beyond the neonatal period [15,30,60,61,75]. Duration of therapy should be individualised and prolonged ASM treatment should be avoided whenever possible.

3.1.4. Neuroprotective Strategies

Neonatal seizures may contribute to adverse long-term outcomes by independently adding further brain injury to the initial insult [23,85,96]. Many promising interventions (such as magnesium sulphate, erythropoietin, topiramate, citicoline, melatonin, bumetanide, correction of metabolic abnormalities, Dexmetomidine, possible immune modulation, stem cell therapy), have been and are being evaluated [1,67,75,97,98,99,100,101] for efficacy.

Therapeutic hypothermia (TH) remains the only intervention with robust evidence for neuroprotective efficacy. TH is well established as a standard treatment for neonates with moderate and severe HIE and has been shown to reduce mortality and severe neurodevelopmental disability [102,103]. Evidence based recommendation from the ILAE task force on treatment of neonatal seizures [60] states TH may reduce seizure burden in neonates with HIE. It’s role in mild HIE is being explored.

Research in neonatal encephalopathy and neonatal seizures is primarily based in high income countries (HIC), while neonates in low and middle-income countries are disproportionately affected [102,104,105,106]). The challenges raised by this situation in clinical care of ill neonates is illustrated by the HELIX trial[107] that concluded that TH should not be offered as treatment for neonatal encephalopathy in low-income and middle-income countries (LMICs). They reported that TH did not reduce the combined outcome of death or disability at 18 months after neonatal encephalopathy in LMICs, but significantly increased death alone. A systematic review[105] discusses the differences in some outcomes from LMICs and HICs and concludes further research is needed to explain the differences.

3.2. Prognosis and Neurodevelopmental Outcome

Seizures are an ominous sign in the neonatal period, often occurring in the first hours to days after birth. Seizures occur frequently, with wide variance in estimates of incidence: 1-15/1000 live births [3,108,109,110] The neonatal period is a unique time of brain development where there is rapid growth and maturation of cortical neurons and formation of exuberant synapses and connections. It is a period when genetic programs are influenced by early experiences. As a result, the brain reacts differently to injury and the consequences of injury may include the disruption of the proper sequence of development [111] Animal studies indicates that seizures adversely affect the young brain, and result in longstanding changes in signalling properties of neurons [85,86,87]

Neonatal seizures are the most common manifestation of neurological disorders in the newborn period and an important determinant of long-term outcome. The estimation of risk of adverse outcomes due to neonatal seizures is complex, as many factors, including aetiology, are involved. Several studies indicate that seizures themselves, irrespective of underlying aetiology contribute to adverse outcomes [82,96,108,110]. In the acute situation it is important to utilise history, clinical examination and profile, all investigations available – EEG, MRI, blood tests, metabolic and genetic tests, and response to treatment to prognosticate outcome.

Neurodevelopmental outcomes after neonatal seizures have improved with time, though disparities persist across the globe. Neonatal seizures are associated with increased risk of adverse outcomes (25-70%) that often co-occur[2,15,22,81,83,108,109,112,113,114,115]. Global developmental delay (upto 45%), epilepsy (20 -35%) and cerebral palsy (20-35%) are the most frequent adverse outcomes. Intellectual Disability is also common (10-15%), and autism, ADHD and impaired social behaviour (4 - 30%) also occur. Despite advances in NNNC, mortality (7-58%) still occurs in preterm and term neonates with seizures and is higher with NSE[116]. The large variance is due to different study cohorts, in different parts of the world, studied at different timeframes and different modalities of assessment. The high incidence of neurodevelopmental disorders, the variance in prevalence and the multidomain involvement seen, highlight the need for standardised follow up programs.

4. Conclusion

Many challenges remain in the clinical care of neonates with seizures. Diagnostic dilemmas, inequities in resources, patchy clinical expertise, availability of CEEG, ability to undertake appropriate investigations in a timely fashion, choice of ASMs, neuroprotective strategies, good long term follow-up studies, and the paucity of randomised clinical trials to establish best clinical practise across the gestational ages, aetiologies and socioeconomic scenarios need to be addressed. Advocacy for the neonate with seizures is required to address the inequities in resources and expertise across the world. Breakthroughs in mechanistic understanding, in neuromonitoring, innovative trial design, and novel therapeutic interventions promote enthusiasm and optimism for clinicians caring for neonates with seizures.: