Risk factors for EEG seizures in neonates treated with hypothermia: a multicenter cohort study

Treating Seizures and Improving Newborn Outcomes for Infants with Hypoxic-Ischemic Encephalopathy

Hypoxic-ischemic encephalopathy is the most common cause of neonatal seizures. Continuous electroencephalographic monitoring is recommended given high rates of subclinical seizures. Prompt diagnosis and treatment of seizures may improve neurodevelopmental outcomes. International League Against Epilepsy guidelines indicate that (1) phenobarbital remains the first-line treatment of neonatal seizures and (2) early discontinuation of antiseizure medications following resolution of acute provoked seizures, and prior to discharge home, is recommended. Long-term follow-up of these infants is necessary to screen for postneonatal epilepsy and support neurodevelopment.

Advances in Electroencephalographic Biomarkers of Neonatal Hypoxic Ischemic Encephalopathy

Electroencephalography (EEG) is a key objective biomarker of newborn brain function, delivering critical, cotside insights to aid the management of encephalopathy. Access to continuous EEG is limited, forcing reliance on subjective clinical assessments. In hypoxia ischaemia, the primary cause of encephalopathy, alterations in EEG patterns correlate with. injury severity and evolution. As HIE evolves, causing secondary neuronal death, EEG can track injury progression, informing neuroprotective strategies, seizure management and prognosis. Despite its value, challenges with interpretation and lack of on site expertise has limited its broader adoption. Technological advances, particularly in digital EEG and machine learning, are enhancing real-time analysis. This will allow EEG to expand its role in HIE diagnosis, management and outcome prediction.

A Comparison of Automatically Extracted Quantitative EEG Features for Seizure Risk Stratification in Neonatal Encephalopathy

PURPOSE

Seizures occur in up to 40% of neonates with neonatal encephalopathy. Earlier identification of seizures leads to more successful seizure treatment, but is often delayed because of limited availability of continuous EEG monitoring. Clinical variables poorly stratify seizure risk, and EEG use to stratify seizure risk has previously been limited by need for manual review and artifact exclusion. The goal of this study is to compare the utility of automatically extracted quantitative EEG (qEEG) features for seizure risk stratification.

METHODS

We conducted a retrospective analysis of neonates with moderate-to-severe neonatal encephalopathy who underwent therapeutic hypothermia at a single center. The first 24 hours of EEG underwent automated artifact removal and qEEG analysis, comparing qEEG features for seizure risk stratification.

RESULTS

The study included 150 neonates and compared the 36 (23%) with seizures with those without. Absolute spectral power best stratified seizure risk with area under the curve ranging from 63% to 71%, followed by range EEG lower and upper margin, median and SD of the range EEG lower margin. No features were significantly more predictive in the hour before seizure onset. Clinical examination was not associated with seizure risk.

CONCLUSIONS

Automatically extracted qEEG features were more predictive than clinical examination in stratifying neonatal seizure risk during therapeutic hypothermia. qEEG represents a potential practical bedside tool to individualize intensity and duration of EEG monitoring and decrease time to seizure recognition. Future work is needed to refine and combine qEEG features to improve risk stratification.

Hydrogen gas can ameliorate seizure burden during therapeutic hypothermia in asphyxiated newborn piglets

BACKGROUND

We previously reported that hydrogen (H2) gas combined with therapeutic hypothermia (TH) improved short-term neurological outcomes in asphyxiated piglets. However, the effect on seizure burden was unclear. Using amplitude-integrated electroencephalography (aEEG), we compared TH + H2 with TH alone in piglets 24 h after hypoxic-ischemic (HI) insult.

METHODS

After a 40-min insult and resuscitation, 36 piglets ≤24 h old were divided into three groups: normothermia (NT, n = 14), TH alone (33.5 ± 0.5 °C, 24 h, n = 13), and TH + H2 (2.1-2.7% H2 gas, 24 h, n = 9). aEEG was recorded for 24 h post-insult and its background pattern, status epilepticus (SE; recurrent seizures lasting >5 min), and seizure occurrence (Sz; occurring at least once but not fitting the definition of SE) were evaluated. Background findings with a continuous low voltage and burst suppression were considered abnormal.

RESULTS

The percentage of piglets with an abnormal aEEG background (aEEG-BG), abnormal aEEG-BG+Sz and SE was lower with TH + H2 than with TH at 24 h after HI insult. The duration of SE was shorter with TH + H2 and significantly shorter than with NT.

CONCLUSIONS

H2 gas combined with TH ameliorated seizure burden 24 h after HI insult.

IMPACT

In this asphyxiated piglet model, there was a high percentage of animals with an abnormal amplitude-integrated electroencephalography background (aEEG-BG) after hypoxic-ischemic (HI) insult, which may correspond to moderate and severe hypoxic-ischemic encephalopathy (HIE). Therapeutic hypothermia (TH) was associated with a low percentage of piglets with EEG abnormalities up to 6 h after HI insult but this percentage increased greatly after 12 h, and TH was not effective in attenuating seizure development. H2 gas combined with TH was associated with a low percentage of piglets with an abnormal aEEG-BG and with a shorter duration of status epilepticus at 24 h after HI insult.

The Impact of Phenobarbital on the Ability of Electroencephalogram to Predict Adverse Outcome in Asphyxiated Neonates during Therapeutic Hypothermia

OBJECTIVE

Classification of electroencephalogram (EEG) background has been established to predict outcome in neonates with hypoxic ischemic encephalopathy (HIE). However, the impact of phenobarbital therapy on the predictability of EEG background has not been studied. Our objective is to determine if EEG background after treatment with phenobarbital during therapeutic hypothermia (TH) remains a good predictor for brain injury in neonates with HIE.

STUDY DESIGN

This is a single-center, retrospective study of consecutive neonates with HIE who underwent TH and EEG monitoring from October 2017 to March 2021. Per institutional protocol, all infants received a dose of prophylactic phenobarbital and bumetanide therapy at the onset of TH for sedative and neuroprotective measures. The initial 3 hours of EEG background activity was classified based on national guidelines. Infants were separated into two groups based on EEG background scores: group 1 (normal-mild, n = 30) and group 2 (moderate-severe, n = 36). Brain magnetic resonance imaging (MRI) results were scored based on the National Institute of Child Health and Human Development (NICHD) criteria. Adverse outcomes were defined as death before MRI or NICHD brain injury score > 1A.

RESULTS

Infants in group 2 had lower Apgar scores at 5 minutes of age, severe acidemia, moderate to severe encephalopathy score, and earlier initiation of EEG monitoring than infants in group 1. Moderate to severe EEG background score was associated with presence of brain injury on MRI or death (p = 0.003), and this association remained significant even after adjustment for independent risk factors (odds ratio = 56.24 [95% confidence interval = 1.841-1718], p = 0.021).

CONCLUSION

Phenobarbital therapy does not affect the ability of EEG to predict adverse outcome in infants with perinatal asphyxia during TH.

KEY POINTS

· EEG has a clinical utility for predicting outcome in neonates with hypoxia-ischemia.. · Phenobarbital therapy is commonly used in neonates, and may impact EEG background findings.. · In spite phenobarbital therapy, moderate to severe EEG background abnormalities in infants with perinatal asphyxia during TH remain an excellent predictor for poor outcome..

Seizures after initiation of rewarming in cooled infants with hypoxic ischaemic encephalopathy

BACKGROUND

Seizures after initiation of rewarming from therapeutic hypothermia for neonatal encephalopathy are well recognised but not easy to predict.

METHODS

A secondary analysis was performed of NEOLEV2 trial data, a multicentre randomised trial of levetiracetam versus phenobarbital for neonatal seizures. Enrolled infants underwent continuous video EEG (cEEG) monitoring. The trial data were reviewed for 42 infants with seizures during therapeutic hypothermia and 118 infants who received therapeutic hypothermia but had no seizures on cEEG.

RESULTS

Overall, 112 of 160 (70%) had cEEG monitoring continued until rewarming was completed. Of the 42 infants with prior seizures, there were 30 infants with valid cEEG available and seizures occurred following the initiation of rewarming in 8 (26.6%). For the 118 seizure-naive infants, 82 (69.5%) continued cEEG until either rewarming was completed or 90 h of age and none had documented seizures.

CONCLUSION

Overall, just over a quarter of infants with prior seizures had cEEG evidence of at least one seizure in the 24 h after initiation of rewarming but no seizure-naive infant had cEEG evidence of seizure(s) on rewarming. Critically, by reporting the two groups separately, the data can provide guidance on the duration of EEG monitoring.

IMPACT

Infants with hypoxic ischaemic encephalopathy who have cEEG evidence of seizures during therapeutic hypothermia have a significant risk of further seizures on rewarming. For infants with hypoxic ischaemic encephalopathy but no cEEG evidence of seizures during therapeutic hypothermia, there is very little risk of de novo seizures. Ongoing work utilising large cohorts may generate EEG criteria that refine estimates of risk for rewarming seizures. Based on current experience, if seizures have occurred during therapeutic hypothermia for hypoxic ischaemic encephalopathy, the EEG monitoring should be continued during rewarming and for 12 h thereafter to minimise the risk of missing an event.

Scalp high-frequency oscillations differentiate neonates with seizures from healthy neonates

OBJECTIVE

We aimed to investigate (1) whether an automated detector can capture scalp high-frequency oscillations (HFO) in neonates and (2) whether scalp HFO rates can differentiate neonates with seizures from healthy neonates.

METHODS

We considered 20 neonates with EEG-confirmed seizures and four healthy neonates. We applied a previously validated automated HFO detector to determine scalp HFO rates in quiet sleep.

RESULTS

Etiology in neonates with seizures included hypoxic-ischemic encephalopathy in 11 cases, structural vascular lesions in 6, and genetic causes in 3. The HFO rates were significantly higher in neonates with seizures (0.098 ± 0.091 HFO/min) than in healthy neonates (0.038 ± 0.025 HFO/min; P = 0.02) with a Hedge's g value of 0.68 indicating a medium effect size. The HFO rate of 0.1 HFO/min/ch yielded the highest Youden index in discriminating neonates with seizures from healthy neonates. In neonates with seizures, etiology, status epilepticus, EEG background activity, and seizure patterns did not significantly impact HFO rates.

SIGNIFICANCE

Neonatal scalp HFO can be detected automatically and differentiate neonates with seizures from healthy neonates. Our observations have significant implications for neuromonitoring in neonates. This is the first step in establishing neonatal HFO as a biomarker for neonatal seizures.

Characteristics of neonatal hypoxic-ischemic encephalopathy at high altitude and early results of therapeutic hypothermia

BACKGROUND

Altitude hypoxia and limited socioeconomic conditions may result in distinctive features of neonatal hypoxic-ischemic encephalopathy (HIE). Therapeutic hypothermia (TH) has not been used at altitude. We examined characteristics of HIE and early outcomes of TH in 3 centers at two high altitudes, 2 at 2,261 m and 1 at 3,650 m.

METHODS

The incidence of HIE at NICUs was noted. TH was conducted when personnel and devices were available in 2019~2020. Standard inclusion criteria were used, with the addition of admission age >6 hours and mild HIE. Demographic and clinical data included gestational age, gender, weight, Apgar score, ethnics, age on admission, age at TH and clinical degree of HIE. EEG was monitored for 96 hours during hypothermia and rewarming. MRI was performed before discharge.

RESULTS

There was significant difference in ethnics, HIE degree, age at TH across 3 centers. The overall NICU incidence of HIE was 4.0%. Among 566 HIE patients, 114 (20.1%) received TH. 63 (55.3%) patients had moderate/severe HIE. Age at TH >6 hours occurred in 34 (29.8%) patients. EEG discharges showed seizures in 7~11% of patients, whereas spikes/sharp waves in 94~100%, delta brushes in 50~100%. After TH, MRI showed moderate to severe brain injury in 77% of patients, and correlated with center, demographic and clinical variables (Ps≤0.0003). Mortality was 5% during hospitalization and 11% after discharge until 1 year.

CONCLUSIONS

At altitude, the incidence of HIE was high and brain injury was severe. TH was limited and often late >6 hours. EEG showed distinct patterns attributable to altitude hypoxia. TH was relatively safe.

TRIAL REGISTRATION

The study was registered on February 23, 2019 in Chinese Clinical Trial Register (ChiCTR1900021481).

Hypothermic Protection in Neocortex Is Topographic and Laminar, Seizure Unmitigating, and Partially Rescues Neurons Depleted of RNA Splicing Protein Rbfox3/NeuN in Neonatal Hypoxic-Ischemic Male Piglets

The effects of hypothermia on neonatal encephalopathy may vary topographically and cytopathologically in the neocortex with manifestations potentially influenced by seizures that alter the severity, distribution, and type of neuropathology. We developed a neonatal piglet survival model of hypoxic-ischemic (HI) encephalopathy and hypothermia (HT) with continuous electroencephalography (cEEG) for seizures. Neonatal male piglets received HI-normothermia (NT), HI-HT, sham-NT, or sham-HT treatments. Randomized unmedicated sham and HI piglets underwent cEEG during recovery. Survival was 2-7 days. Normal and pathological neurons were counted in different neocortical areas, identified by cytoarchitecture and connectomics, using hematoxylin and eosin staining and immunohistochemistry for RNA-binding FOX-1 homolog 3 (Rbfox3/NeuN). Seizure burden was determined. HI-NT piglets had a reduced normal/total neuron ratio and increased ischemic-necrotic/total neuron ratio relative to sham-NT and sham-HT piglets with differing severities in the anterior and posterior motor, somatosensory, and frontal cortices. Neocortical neuropathology was attenuated by HT. HT protection was prominent in layer III of the inferior parietal cortex. Rbfox3 immunoreactivity distinguished cortical neurons as: Rbfox3-positive/normal, Rbfox3-positive/ischemic-necrotic, and Rbfox3-depleted. HI piglets had an increased Rbfox3-depleted/total neuron ratio in layers II and III compared to sham-NT piglets. Neuronal Rbfox3 depletion was partly rescued by HT. Seizure burdens in HI-NT and HI-HT piglets were similar. We conclude that the neonatal HI piglet neocortex has: (1) suprasylvian vulnerability to HI and seizures; (2) a limited neuronal cytopathological repertoire in functionally different regions that engages protective mechanisms with HT; (3) higher seizure burden, insensitive to HT, that is correlated with more panlaminar ischemic-necrotic neurons in the somatosensory cortex; and (4) pathological RNA splicing protein nuclear depletion that is sensitive to HT. This work demonstrates that HT protection of the neocortex in neonatal HI is topographic and laminar, seizure unmitigating, and restores neuronal depletion of RNA splicing factor.

Evaluating Injury Severity in Neonatal Encephalopathy Using Automated Quantitative Electroencephalography Analysis: A Pilot Study

Quantitative analysis of electroencephalography (qEEG) is a potential source of biomarkers for neonatal encephalopathy (NE). However, prior studies using qEEG in NE were limited in their generalizability due to individualized techniques for calculating qEEG features or labor-intensive pre-selection of EEG data. We piloted a fully automated method using commercially available software to calculate the suppression ratio (SR), absolute delta power, and relative delta, theta, alpha, and beta power from EEG of neonates undergoing 72 h of therapeutic hypothermia (TH) for NE between April 20, 2018, and November 4, 2019. We investigated the association of qEEG with degree of encephalopathy (modified Sarnat score), severity of neuroimaging abnormalities following TH (National Institutes of Child Health and Development Neonatal Research Network [NICHD-NRN] score), and presence of seizures. Thirty out of 38 patients met inclusion criteria. A more severe modified Sarnat score was associated with higher SR during all phases of TH, lower absolute delta power during all phases except rewarming, and lower relative delta power during the last 24 h of TH. In 21 patients with neuroimaging data, a worse NICHD-NRN score was associated with higher SR, lower absolute delta power, and higher relative beta power during all phases. QEEG features were not significantly associated with the presence of seizures after correction for multiple comparisons. Our results are consistent with those of prior studies using qEEG in NE and support automated qEEG analysis as an accessible, generalizable method for generating biomarkers of NE and response to TH. Additionally, we found evidence of an immature relative frequency composition in neonates with more severe brain injury, suggesting that automated qEEG analysis may have a use in the assessment of brain maturity.

Risk of seizures in neonates with hypoxic-ischemic encephalopathy receiving hypothermia plus erythropoietin or placebo

BACKGROUND

An ancillary study of the High-Dose Erythropoietin for Asphyxia and Encephalopathy (HEAL) trial for neonates with hypoxic-ischemic encephalopathy (HIE) and treated with therapeutic hypothermia examined the hypothesis that neonates randomized to receive erythropoietin (Epo) would have a lower seizure risk and burden compared with neonates who received placebo.

METHODS

Electroencephalograms (EEGs) from 7/17 HEAL trial centers were reviewed. Seizure presence was compared across treatment groups using a logistic regression model adjusting for treatment, HIE severity, center, and seizure burden prior to the first dose. Among neonates with seizures, differences across treatment groups in median maximal hourly seizure burden were assessed using adjusted quantile regression models.

RESULTS

Forty-six of 150 (31%) neonates had EEG seizures (31% in Epo vs 30% in placebo, p = 0.96). Maximal hourly seizure burden after the study drug was not significantly different between groups (median 11.4 for Epo, IQR: 5.6, 18.1 vs median 9.7, IQR: 4.9, 21.0 min/h for placebo).

CONCLUSION

In neonates with HIE treated with hypothermia who were randomized to Epo or placebo, we found no meaningful between-group difference in seizure risk or burden. These findings are consistent with overall trial results, which do not support Epo use for neonates with HIE undergoing therapeutic hypothermia.

IMPACT

In the HEAL trial of erythropoietin (Epo) vs placebo for neonates with encephalopathy presumed due to hypoxic-ischemic encephalopathy (HIE) who were also treated with therapeutic hypothermia, electrographic seizures were detected in 31%, which is lower than most prior studies. Epo did not reduce the proportion of neonates with acute provoked seizures (31% in Epo vs 30% in placebo) or maximal hourly seizure burden after the study drug (median 11.4, IQR 5.6, 18.1 for Epo vs median 9.7, IQR 4.9, 21.0 min/h for placebo). There was no anti- or pro-convulsant effect of Epo when combined with therapeutic hypothermia for HIE.

Leveraging electronic medical record-embedded standardised electroencephalogram reporting to develop neonatal seizure prediction models: a retrospective cohort study

BACKGROUND

Accurate prediction of seizures can help to direct resource-intense continuous electroencephalogram (CEEG) monitoring to neonates at high risk of seizures. We aimed to use data from standardised EEG reports to generate seizure prediction models for vulnerable neonates.

METHODS

In this retrospective cohort study, we included neonates who underwent CEEG during the first 30 days of life at the Children's Hospital of Philadelphia (Philadelphia, PA, USA). The hypoxic ischaemic encephalopathy subgroup included only patients with CEEG data during the first 5 days of life, International Classification of Diseases, revision 10, codes for hypoxic ischaemic encephalopathy, and documented therapeutic hypothermia. In January, 2018, we implemented a novel CEEG reporting system within the electronic medical record (EMR) using common data elements that incorporated standardised terminology. All neonatal CEEG data from Jan 10, 2018, to Feb 15, 2022, were extracted from the EMR using age at the time of CEEG. We developed logistic regression, decision tree, and random forest models of neonatal seizure prediction using EEG features on day 1 to predict seizures on future days.

FINDINGS

We evaluated 1117 neonates, including 150 neonates with hypoxic ischaemic encephalopathy, with CEEG data reported using standardised templates between Jan 10, 2018, and Feb 15, 2022. Implementation of a consistent EEG reporting system that documents discrete and standardised EEG variables resulted in more than 95% reporting of key EEG features. Several EEG features were highly correlated, and patients could be clustered on the basis of specific features. However, no simple combination of features adequately predicted seizure risk. We therefore applied computational models to complement clinical identification of neonates at high risk of seizures. Random forest models incorporating background features performed with classification accuracies of up to 90% (95% CI 83-94) for all neonates and 97% (88-99) for neonates with hypoxic ischaemic encephalopathy; recall (sensitivity) of up to 97% (91-100) for all neonates and 100% (100-100) for neonates with hypoxic ischaemic encephalopathy; and precision (positive predictive value) of up to 92% (84-96) in the overall cohort and 97% (80-99) in neonates with hypoxic ischaemic encephalopathy.

INTERPRETATION

Using data extracted from the standardised EEG report on the first day of CEEG, we predict the presence or absence of neonatal seizures on subsequent days with classification performances of more than 90%. This information, incorporated into routine care, could guide decisions about the necessity of continuing EEG monitoring beyond the first day, thereby improving the allocation of limited CEEG resources. Additionally, this analysis shows the benefits of standardised clinical data collection, which can drive learning health system approaches to personalised CEEG use.

FUNDING

Children's Hospital of Philadelphia, the Hartwell Foundation, the National Institute of Neurological Disorders and Stroke, and the Wolfson Foundation.

Effect of neonatal seizure burden and etiology on the long-term outcome: data from a randomized, controlled trial

Background

Neonatal seizures are common, but the impact of neonatal seizures on long-term neurologic outcome remains unclear. We addressed this question by analyzing data from an early-phase controlled trial of bumetanide to treat neonatal seizures.

Methods

Neonatal seizure burden was calculated from continuous video-EEG data. Neurologic outcome was determined by standardized developmental tests and post-neonatal seizure recurrence.

Results

Of 111 enrolled neonates, 43 were randomized to treatment or control groups. There were no differences in neurologic outcome between treatment and control groups. A subgroup analysis was performed for 84 neonates with acute perinatal brain injury (57 HIE, 18 stroke, 9 ICH), most of whom (70%) had neonatal seizures. There was a significant negative correlation between seizure burden and developmental scores (p<0.01). Associations between seizure burden and developmental scores were stronger in HIE and stroke groups compared with ICH (p<0.05).

Conclusion

Bumetanide showed no long-term beneficial or adverse effects, as expected based on treatment duration versus duration of neonatal seizures. For neonates with perinatal brain injury, higher neonatal seizure burden correlated significantly with worse developmental outcome, particularly for ischemic versus hemorrhagic brain injury. These data highlight the need for further investigation of the long-term effects of both neonatal seizure severity and etiology.

Machine learning for the early prediction of infants with electrographic seizures in neonatal hypoxic-ischemic encephalopathy

OBJECTIVE

To assess if early clinical and electroencephalography (EEG) features predict later seizure development in infants with hypoxic-ischemic encephalopathy (HIE).

METHODS

Clinical and EEG parameters <12 h of birth from infants with HIE across eight European Neonatal Units were used to develop seizure-prediction models. Clinical parameters included intrapartum complications, fetal distress, gestational age, delivery mode, gender, birth weight, Apgar scores, assisted ventilation, cord pH, and blood gases. The earliest EEG hour provided a qualitative analysis (discontinuity, amplitude, asymmetry/asynchrony, sleep-wake cycle [SWC]) and a quantitative analysis (power, discontinuity, spectral distribution, inter-hemispheric connectivity) from full montage and two-channel amplitude-integrated EEG (aEEG). Subgroup analysis, only including infants without anti-seizure medication (ASM) prior to EEG was also performed. Machine-learning (ML) models (random forest and gradient boosting algorithms) were developed to predict infants who would later develop seizures and assessed using Matthews correlation coefficient (MCC) and area under the receiver-operating characteristic curve (AUC).

RESULTS

The study included 162 infants with HIE (53 had seizures). Low Apgar, need for ventilation, high lactate, low base excess, absent SWC, low EEG power, and increased EEG discontinuity were associated with seizures. The following predictive models were developed: clinical (MCC 0.368, AUC 0.681), qualitative EEG (MCC 0.467, AUC 0.729), quantitative EEG (MCC 0.473, AUC 0.730), clinical and qualitative EEG (MCC 0.470, AUC 0.721), and clinical and quantitative EEG (MCC 0.513, AUC 0.746). The clinical and qualitative-EEG model significantly outperformed the clinical model alone (MCC 0.470 vs 0.368, p-value .037). The clinical and quantitative-EEG model significantly outperformed the clinical model (MCC 0.513 vs 0.368, p-value .012). The clinical and quantitative-EEG model for infants without ASM (n = 131) had MCC 0.588, AUC 0.832. Performance for quantitative aEEG (n = 159) was MCC 0.381, AUC 0.696 and clinical and quantitative aEEG was MCC 0.384, AUC 0.720.

SIGNIFICANCE

Early EEG background analysis combined with readily available clinical data helped predict infants who were at highest risk of seizures, hours before they occur. Automated quantitative-EEG analysis was as good as expert analysis for predicting seizures, supporting the use of automated assessment tools for early evaluation of HIE.

Neuromonitoring in neonatal critical care part I: neonatal encephalopathy and neonates with possible seizures

The blooming of neonatal neurocritical care over the last decade reflects substantial advances in neuromonitoring and neuroprotection. The most commonly used brain monitoring tools in the neonatal intensive care unit (NICU) are amplitude integrated EEG (aEEG), full multichannel continuous EEG (cEEG), and near-infrared spectroscopy (NIRS). While some published guidelines address individual tools, there is no consensus on consistent, efficient, and beneficial use of these modalities in common NICU scenarios. This work reviews current evidence to assist decision making for best utilization of neuromonitoring modalities in neonates with encephalopathy or with possible seizures. Neuromonitoring approaches in extremely premature and critically ill neonates are discussed separately in the companion paper. IMPACT: Neuromonitoring techniques hold promise for improving neonatal care. For neonatal encephalopathy, aEEG can assist in screening for eligibility for therapeutic hypothermia, though should not be used to exclude otherwise eligible neonates. Continuous cEEG, aEEG and NIRS through rewarming can assist in prognostication. For neonates with possible seizures, cEEG is the gold standard for detection and diagnosis. If not available, aEEG as a screening tool is superior to clinical assessment alone. The use of seizure detection algorithms can help with timely seizures detection at the bedside.

Neonatal seizures: stepping outside the comfort zone

Seizures are the most common neurological disorders in newborns. Managing neonatal seizures is challenging, especially for neurologists who are not neonatal specialists. Acute brain injury during ischemic insult is a key component of seizure occurrence, while genetic and metabolic disorders play less prevalent but more severe roles. The diagnosis of neonatal seizure is ambiguous, as the subjective differentiation between seizure and nonepileptic events is difficult; therefore, electrographic recording is the gold standard for diagnosis. The detection of electrographic seizures by neonatologists is currently facilitated by amplitude-integrated electroencephalography availability in many neonatal intensive care units. Although it is less sensitive than conventional electroencephalography, it is better to record all risky neonates to filter the abnormal events as early as possible to enable the initiation of dedicated therapy at proper dose and time and facilitate the initial response to antiepileptic drugs. This, in turn, helps maintain the balance between unnecessary drug use and their neurotoxic effects. Moreover, the early treatment of electrographic seizures plays a vital role in the suppression of subsequent abnormal brain electricity (status epilepticus) and shortening the hospital stay. An explicit understanding of seizure etiology and pathophysiology should direct attention to the proper prescription of short- and long-term antiepileptic medications to solve the challenging issue of whether neonatal seizures progress to postneonatal epilepsy and long-term cognitive deficits. This review addresses recent updates in different aspects of neonatal seizures, particularly electrographic discharge, including their definition, etiology, classification, diagnosis, management, and neurodevelopmental outcomes.

Neuroprotection for hypoxic-ischemic encephalopathy: Contributions from the neonatal research network

Therapeutic hypothermia (TH) is now well established as the standard of care treatment for moderate to severe neonatal encephalopathy secondary to perinatal hypoxic ischemic encephalopathy (HIE) in infants ≥36 weeks gestation in high income countries. The Neonatal Research Network (NRN) contributed greatly to the study of TH as a neuroprotectant with three trials now completed in infants ≥36 weeks gestation and the only large randomized-controlled trial of TH in preterm infants now in the follow-up phase. Data from the first NRN TH trial combined with data from other large trials of TH affirm the safety and neuroprotective qualities of TH and highlight the importance of providing TH to all infants who qualify. In this review we will highlight the findings of the three NRN trials of TH in the term infant population and the secondary analyses that continue to inform the care of patients with HIE.

Association Between Early EEG Background and Outcomes in Infants With Mild HIE Undergoing Therapeutic Hypothermia

BACKGROUND

Hypoxic-ischemic encephalopathy (HIE) occurs in 1-4:1000 live births. Although neonates with moderate-severe HIE have been studied over several decades, newborns with mild HIE remain understudied, including seizure occurrence, electroencephalography (EEG) characteristics, and outcome.

METHODS

We conducted a retrospective cohort study of neonates ≥35 weeks of gestation with mild HIE who underwent therapeutic hypothermia to correlate the early EEG background pattern with clinical course and outcomes.

RESULTS

Of the included 29 neonates, 10 infants had a moderately to severely abnormal EEG background and 19 had either a normal or a mildly abnormal background. Those with moderately to severely abnormal background also had more multiorgan dysfunction (90% vs 42%, P = 0.02) and a higher incidence of subdural and intraventricular hemorrhages (80% vs 26%, P = 0.02). The overall seizure incidence was 20.7% and was significantly higher in newborns with more severely abnormal background compared to neonates with less abnormal background (50% vs 5%; P = 0.01; relative risk, 9.5; 95% confidence interval, 1.28-70.6). Seizure onset was between 11 and 63 hours of life. Regardless of the EEG background pattern, seizures were brief with an overall low seizure burden. None of the newborns with normal or mildly abnormal background had a new onset of seizures after 24 hours of recording or developed epilepsy during infancy.

CONCLUSIONS

In neonates with mild HIE, early moderately to severely abnormal EEG background is common and strongly associated with an increased risk for seizures.

Electroencephalogram background and head ultrasound together stratify seizure risk in neonates undergoing hypothermia

The benefits of continuous electroencephalography (cEEG) monitoring in the intensive care unit (ICU) are increasingly appreciated, though expanding indications for cEEG may strain resources. The current standard of care in babies with hypoxic-ischemic encephalopathy (HIE) undergoing therapeutic hypothermia (TH) includes cEEG monitoring throughout the entire TH and rewarming process (at least 72 h). Recent cEEG data demonstrate that most seizures occur within the first 24 h of monitoring. We hypothesized that abnormal head imaging and EEG background could stratify seizure risk in babies with HIE undergoing TH to identify candidates for early cEEG discontinuation. In this retrospective review of 126 neonates undergoing TH and cEEG, we identified seizures in 38 (30%) neonates, 33 (87%) of whom seized within the first 24 h of cEEG monitoring. EEG background was graded and demonstrated that 90% of neonates with seizures had a moderately/markedly abnormal background versus 33% of neonates who did not seize (p < 0.0001). Additionally, while head ultrasound (HUS) obtained before EEG did not stratify seizure risk alone, no neonates with both a normal/mildly abnormal EEG background and a normal HUS (0/25) experienced seizures in contrast to 60% (24/40) neonates with both an abnormal EEG background and an abnormal HUS (p < 0.0001). Our data suggest that neonates with abnormal EEG backgrounds and abnormal HUS should be monitored for seizures throughout TH and rewarming, while neonates with normal/mildly abnormal EEG backgrounds and normal HUS are at low risk of seizures after 24 h of monitoring, and thus would be candidates for early cEEG discontinuation.

Neonatal seizures: diagnostic updates based on new definition and classification

Neonatal seizures are the most common neurological symptoms caused by various etiologies in the neonatal period, but their diagnosis and treatment are challenging because their pathophysiology and electroclinical manifestations differ from those of patients in older age groups. Many seizures present as electrographic-only events without clinical signs or as obscure clinical manifestations that are difficult to distinguish from other neonatal behaviors. Accordingly, a new definition and classification of neonatal seizures was recently proposed by the International League Against Epilepsy Task Force on neonatal seizures, highlighting the role of electroencephalography in diagnosing and treating neonatal seizures. Neonatal seizures are defined as electrographic events with sudden, paroxysmal, and abnormal alteration of activity and divided into electroclinical seizures and electrographic-only seizures according to their clinical signs, thus excluding clinical events without an electrographic correlation. Seizure types are described by their predominant clinical features and divided into motor (automatisms, clonic, epileptic spasms, myoclonic, tonic, and sequential), nonmotor (autonomic and behavioral arrest), and unclassified. Although many neonatal seizures are acute reactive events caused by hypoxic-ischemic encephalopathy or vascular insults, structural, genetic, or metabolic etiologies of neonatal-onset epilepsy should also be thoroughly evaluated to determine their appropriate management.

Electrographic Seizures in Neonates with a High Risk of Encephalopathy

Background: Neonatal encephalopathy is caused by a wide variety of acute brain insults in newborns and presents with a spectrum of neurologic dysfunction, such as consciousness disturbance, seizures, and coma. The increased excitability in the neonatal brain appears to be highly susceptible to seizures after a variety of insults, and seizures may be the first clinical sign of a serious neurologic disorder. Subtle seizures are common in the neonatal period, and abnormal clinical paroxysmal events may raise the suspicion of neonatal seizures. Continuous video electroencephalographic (EEG) monitoring is the gold standard for the diagnosis of neonatal seizures. The aim of this study was to identify the prevalence of electrographic seizures and the impact of monitoring in neonates with a high risk of encephalopathy. Methods: We conducted this prospective cohort study in a tertiary neonatal intensive care unit over a 4-year period. Neonates with a high risk of encephalopathy who were receiving continuous video EEG monitoring were eligible. The patients were divided into 2 groups: (1) acute neonatal encephalopathy (ANE) and (2) other high-risk encephalopathy conditions (OHRs). The neonates’ demographic characteristics, etiologies, EEG background feature, presence of electrographic seizures and the impact of monitoring were analyzed. Results: A total of 71 neonates with a high risk of encephalopathy who received continuous video EEG monitoring were enrolled. In this consecutive cohort, 42 (59.2%) were monitored for ANE and 29 (40.8%) were monitored for OHRs. At the time of starting EEG monitoring, 54 (76.1%) of the neonates were term infants. The median gestational age at monitoring was 39 weeks (interquartile range, 37−41 weeks). The median total EEG monitoring duration was 64.7 h (interquartile range, 22.2−72.4 h). Electrographic seizures were captured in 25 of the 71 (35.2%) neonates, of whom 20 (80%) had electrographic-only seizures without clinical correlation. Furthermore, of these 20 neonates, 13 (65%) developed electrographic status epilepticus. Electrographic seizures were most commonly found in the ANE group (17, 40.5%) than in the OHRs group (8, 27.6%) (p = 0.013). Besides, normal/mild abnormality and inactive EEG background were less electrographic seizure than moderate and major abnormality EEG background (2 of 30, 6.7% vs. 23 of 41, 56.1%, p < 0.001). Finally, continuous video EEG monitoring excluded the diagnosis of electrographic seizures in two-thirds of the monitored neonates who had paroxysmal events mimicking seizures and led to a change in clinical management in 39.4% of the neonates. Conclusions: Our findings showed that monitoring could accurately detect seizures, and that it could be used to guide seizure medication management. Therefore, continuous video EEG monitoring has important clinical management implications in neonates with a high risk of encephalopathy.

Risk Factors for Clinical Seizures in Neonates with Hypoxic-ischemic Encephalopathy Treated with Therapeutic Hypothermia

BACKGROUND

This study aimed to assess the risk factors for clinical seizures in newborns treated with whole body cooling (WBC) for hypoxic ischemic encephalopathy (HIE).

METHODS

Infants with gestational age≥36 weeks and birth weight≥2.000 g who were treated with WBC due to HIE were retrospectively enrolled in this study. Patients were assigned to two groups: infants without clinical seizures (Group 1) and infants with clinical seizures (Group 2). The two groups were compared to determine the risk factors for the occurrence of clinical seizures.

RESULTS

A total of 25 patients (Group 1=10 and Group 2=15) were included in the study. Prothrombin time (PT) was determined as independent risk factor for clinical seizures (p=0.046) and the odds ratio for the effect of PT was found as 1.475 (%95 CI:1.006-2.299). PT (area under the curve [AUC]=0.764; p=0.041), and increased cardiac troponin-I (cTnI) (AUC=0.935; p=0.002) were found to be significant risk factors for predicting the occurrence of clinical seizures. The optimal PT cut-off value was 22.7 sec, with a sensitivity and specificity of 45.4% and 90%, respectively; as well as positive and negative predictive value of 83.3% and 60.0%, respectively. The chest compression in the delivery room, severely abnormal amplitude integrated electroencephalography and high encephalopathy score were also found risk factors for occurrence of clinical seizures.

CONCLUSION

Chest compression in the delivery room, high encephalopathy score, prolonged PT, and increased cTnI are significant factors for clinical seizures in newborns treated with WBC for HIE.

Ensemble Learning Using Individual Neonatal Data for Seizure Detection

Objective: Sharing medical data between institutions is difficult in practice due to data protection laws and official procedures within institutions. Therefore, most existing algorithms are trained on relatively small electroencephalogram (EEG) data sets which is likely to be detrimental to prediction accuracy. In this work, we simulate a case when the data can not be shared by splitting the publicly available data set into disjoint sets representing data in individual institutions. Methods and procedures: We propose to train a (local) detector in each institution and aggregate their individual predictions into one final prediction. Four aggregation schemes are compared, namely, the majority vote, the mean, the weighted mean and the Dawid-Skene method. The method was validated on an independent data set using only a subset of EEG channels. Results: The ensemble reaches accuracy comparable to a single detector trained on all the data when sufficient amount of data is available in each institution. Conclusion: The weighted mean aggregation scheme showed best performance, it was only marginally outperformed by the Dawid–Skene method when local detectors approach performance of a single detector trained on all available data. Clinical impact: Ensemble learning allows training of reliable algorithms for neonatal EEG analysis without a need to share the potentially sensitive EEG data between institutions.

Association Between Increased Seizures During Rewarming After Hypothermia for Neonatal Hypoxic Ischemic Encephalopathy and Abnormal Neurodevelopmental Outcomes at 2-Year Follow-up: A Nested Multisite Cohort Study

Importance

Compared with normothermia, hypothermia has been shown to reduce death or disability in neonatal hypoxic ischemic encephalopathy but data on seizures during rewarming and associated outcomes are scarce.

Objective

To determine whether electrographic seizures are more likely to occur during rewarming compared with the preceding period and whether they are associated with abnormal outcomes in asphyxiated neonates receiving hypothermia therapy.

Design, Setting, and Participants

This prespecified nested cohort study of infants enrolled in the Optimizing Cooling (OC) multicenter Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) Neonatal Research Network trial from December 2011 to December 2013 with 2 years' follow-up randomized infants to either 72 hours of cooling (group A) or 120 hours (group B). The main trial included 364 infants. Of these, 194 were screened, 10 declined consent, and 120 met all predefined inclusion criteria. A total of 112 (90%) had complete data for death or disability. Data were analyzed from January 2018 to January 2020.

Interventions

Serial amplitude electroencephalography recordings were compared in the 12 hours prior and 12 hours during rewarming for evidence of electrographic seizure activity by 2 central amplitude-integrated electroencephalography readers blinded to treatment arm and rewarming epoch. Odds ratios and 95% CIs were evaluated following adjustment for center, prior seizures, depth of cooling, and encephalopathy severity.

Main Outcomes and Measures

The primary outcome was the occurrence of electrographic seizures during rewarming initiated at 72 or 120 hours compared with the preceding 12-hour epoch. Secondary outcomes included death or moderate or severe disability at age 18 to 22 months. The hypothesis was that seizures during rewarming were associated with higher odds of abnormal neurodevelopmental outcomes.

Results

A total of 120 newborns (70 male [58%]) were enrolled (66 in group A and 54 in group B). The mean (SD) gestational age was 39 (1) weeks. There was excellent interrater agreement (κ, 0.99) in detection of seizures. More infants had electrographic seizures during the rewarming epoch compared with the preceding epoch (group A, 27% vs 14%; P = .001; group B, 21% vs 10%; P = .03). Adjusted odd ratios (95% CIs) for seizure frequency during rewarming were 2.7 (1.0-7.5) for group A and 3.2 (0.9-11.6) for group B. The composite death or moderate to severe disability outcome at 2 years was significantly higher in infants with electrographic seizures during rewarming (relative risk [95% CI], 1.7 [1.25-2.37]) after adjusting for baseline clinical encephalopathy and seizures as well as center.

Conclusions and Relevance

Findings that higher odds of electrographic seizures during rewarming are associated with death or disability at 2 years highlight the necessity of electroencephalography monitoring during rewarming in infants at risk.

Trial Registration

ClinicalTrials.gov Identifier: NCT01192776.

Current and Future Uses of Continuous EEG in the NICU

Continuous EEG (cEEG) is a fundamental neurodiagnostic tool in the care of critically ill neonates and is increasingly recommended. cEEG enhances prognostication via assessment of the background brain activity, plays a role in predicting which neonates are at risk for seizures when combined with clinical factors, and allows for accurate diagnosis and management of neonatal seizures. Continuous EEG is the gold standard method for diagnosis of neonatal seizures and should be used for detection of seizures in high-risk clinical conditions, differential diagnosis of paroxysmal events, and assessment of response to treatment. High costs associated with cEEG are a limiting factor in its widespread implementation. Centralized remote cEEG interpretation, automated seizure detection, and pre-natal EEG are potential future applications of this neurodiagnostic tool.

Expanding Access to Continuous EEG Monitoring in Neonatal Intensive Care Units

PURPOSE

Neonatal seizures are common and difficult to identify clinically because the majority are subclinical and correct identification of electroclinical seizures based on semiology is unreliable. Therefore, continuous EEG monitoring (CEEG) is critical for seizure identification in neonates and is recommended as the gold standard method in American Clinical Neurophysiology Society guidelines. Despite these recommendations, barriers to implementing widespread CEEG exist.

METHODS

To expand access to CEEG for at-risk neonates, a framework for providing remote CEEG was established at two network hospital neonatal intensive care units. Utilization and clinical impact were tracked as a quality improvement study.

RESULTS

In a 27-month period from June 2017 through September 2019, 76 neonates underwent CEEG between the two network neonatal intensive care units. Electrographic seizures occurred in about one quarter of records (18/76; 24%), though their incidence varied by CEEG indication. Care notes indicated that CEEG impacted clinical care in three quarters of cases (57/76; 75%). Continuous EEG impacted decisions to treat with anti-seizure medications in approximately one half of patients (impact: 28/57 [49%]; no impact 29/57 [51%]), and CEEG impacted prognostic discussions in approximately two thirds of patients (impact: 39/57 [68%]; no impact 18/57 [32%]).

CONCLUSIONS

Establishment of a remote CEEG program for neonates is feasible, effective at identifying seizures, and improves the quality of care provided to neonates hospitalized at these network hospitals.

Neurological and developmental outcomes following neonatal encephalopathy treated with therapeutic hypothermia

In randomized trials, therapeutic hypothermia (TH) is associated with reduced prevalence of the composite outcome mortality or neurodevelopmental morbidity in infants with neonatal encephalopathy (NE). Following systematic review, the reduction in prevalence of both mortality and infant neuromorbidity is clear. Among three trials reporting school age outcomes, the effects of NE and TH suggest that such benefit persists into middle childhood, but none of the major trials were powered to detect differences in these outcomes. Cognitive, educational and behavioural outcomes are all adversely affected by NE in children without moderate or severe neuromorbidity. High-quality longitudinal studies of neurocognitive and educational outcomes following NE in the era of TH, including studies incorporating multimodal neuroimaging assessments, are required to characterise deficits more precisely so that robust interventional targets may be developed, and resource planning can occur. Understanding the impact of NE on families and important educational, social, and behavioural outcomes in childhood is critical to attempts to optimise outcomes through interventions.

Bedside and laboratory neuromonitoring in neonatal encephalopathy

Several bedside and laboratory neuromonitoring tools are currently used in neonatal encephalopathy (NE) to assess 1) brain function [amplitude-integrated electroencephalogram (aEEG) and EEG], 2) cerebral oxygenation delivery and consumption [near-infrared spectroscopy (NIRS)] and 3) blood and cerebrospinal fluid biomarkers. The aim of the review is to provide the role of neuromonitoring in understanding the development of brain injury in these newborns and better predict their long-term outcome. Simultaneous use of these monitoring modalities may improve our ability to provide meaningful prognostic information regarding ongoing treatments. Evidence will be summarized in this review for each of these modalities, by describing (1) the methods, (2) the clinical evidence in context of NE both before and with hypothermia, and (3) the research and future directions.

Correlation of electroencephalogram background evolution with the degree of brain injury in neonates with hypoxic-ischemic encephalopathy

OBJECTIVES

To study the correlation of electroencephalogram (EEG) background evolution with the degree of brain injury in neonates with hypoxic-ischemic encephalopathy (HIE).

METHODS

A retrospective analysis was performed for 56 neonates with HIE who underwent continuous video electroencephalogram (cVEEG) and brain magnetic resonance imaging (MRI) examinations. According to clinical symptoms, they were divided into a mild group with 3 neonates, a moderate group with 36 neonates, and a severe group with 17 neonates. EEG background grading and MRI score were determined for each group to analyze the correlation of EEG background evolution with the degree of brain injury.

RESULTS

Compared with the moderate group, the severe group had significantly lower gestational age and Apgar score at 5 minutes after birth, a significantly higher resuscitation score, significantly lower base excess in umbilical cord blood or blood gas within 1 hour, a significantly higher proportion of neonates on mechanical ventilation, and a significantly higher incidence rate of short-term adverse outcomes (P<0.05). For the neonates in the mild and moderate groups, MRI mainly showed no brain injury (67%, 2/3) and watershed injury (67%, 16/24) respectively, and EEG showed mild abnormality in 62% (13/21) of the neonates on the 3rd day after birth. For the neonates in the severe group, MRI mainly showed basal ganglia/thalamus + brainstem injury (24%, 4/17) and whole brain injury (71%, 12/17), and EEG showed moderate or severe abnormalities on the 3rd day after birth. EEG background grading was correlated with clinical grading, MRI score, and short-term outcome on days 1, 2, 3 and 7-14 after birth (P<0.01). The highest correlation coefficient between EEG grading and MRI score was observed on the 3rd day after birth (rs=0.751, P<0.001), and the highest correlation coefficients between EEG grading and clinical grading (rs=0.592, P=0.002) and between EEG grading and short-term outcome (rs=0.737, P<0.001) were observed 7-14 days after birth. Among the neonates with severe abnormal EEG, the neonates without brain electrical activity had the highest MRI score, followed by those with status epileptics and persistent low voltage (P<0.05).

CONCLUSIONS

There is a good correlation between EEG background grading and degree of brain injury in neonates with HIE, which can help to evaluate the degree and prognosis of brain injury in the early stage.

Seizure Control in Neonates Undergoing Screening vs Confirmatory EEG Monitoring

OBJECTIVE

To determine whether screening continuous EEG monitoring (cEEG) is associated with greater odds of treatment success for neonatal seizures.

METHODS

We included term neonates with acute symptomatic seizures enrolled in the Neonatal Seizure Registry (NSR), a prospective, multicenter cohort of neonates with seizures. We compared 2 cEEG approaches: (1) screening cEEG, initiated for indications of encephalopathy or paralysis without suspected clinical seizures; and (2) confirmatory cEEG, initiated for the indication of clinical events suspicious for seizures, either alone or in addition to other indications. The primary outcome was successful response to initial seizure treatment, defined as seizures resolved without recurrence within 30 minutes after initial loading dose of antiseizure medicine. Multivariable logistic regression analyses assessed the association between cEEG approach and successful seizure treatment.

RESULTS

Among 514 neonates included, 161 (31%) had screening cEEG and 353 (69%) had confirmatory cEEG. Neonates with screening cEEG had a higher proportion of successful initial seizure treatment than neonates with confirmatory cEEG (39% vs 18%; p < 0.0001). After adjusting for covariates, there remained a greater odds ratio (OR) for successful initial seizure treatment in the screening vs confirmatory cEEG groups (adjusted OR 2.44, 95% confidence interval 1.45-4.11, p = 0.0008).

CONCLUSIONS

These findings provide evidence from a large, contemporary cohort of neonates that a screening cEEG approach may improve odds of successful treatment of acute seizures.

CLASSIFICATION OF EVIDENCE

This study provides Class III evidence that for neonates a screening cEEG approach, compared to a confirmatory EEG approach, increases the probability of successful treatment of acute seizures.

Management of seizures in neonates with neonatal encephalopathy treated with hypothermia

Neonatal encephalopathy (NE) is the most common etiology of acute neonatal seizures - about half of neonates treated with therapeutic hypothermia for NE have EEG-confirmed seizures. These seizures are best identified with continuous EEG monitoring, as clinical diagnosis leads to under-diagnosis of subclinical seizures and over-treatment of events that are not seizures. High seizure burden, especially status epilepticus, is thought to augment brain injury. Treatment, therefore, is aimed at minimizing seizure burden. Phenobarbital remains the mainstay of treatment, as it is more effective than levetiracetam and easier to administer than fosphenytoin. Emerging evidence suggests that, for many neonates, it is safe to discontinue the phenobarbital after acute seizures resolve and prior to hospital discharge.

Convolutional neural networks ensemble model for neonatal seizure detection

BACKGROUND

Neonatal seizures are a common occurrence in clinical settings, requiring immediate attention and detection. Previous studies have proposed using manual feature extraction coupled with machine learning, or deep learning to classify between seizure and non-seizure states.

NEW METHOD

In this paper a deep learning based approach is used for neonatal seizure classification using electroencephalogram (EEG) signals. The architecture detects seizure activity in raw EEG signals as opposed to common state-of-art, where manual feature extraction with machine learning algorithms is used. The architecture is a two-dimensional (2D) convolutional neural network (CNN) to classify between seizure/non-seizure states.

RESULTS

The dataset used for this study is annotated by three experts and as such three separate models are trained on individual annotations, resulting in average accuracies (ACC) of 95.6 %, 94.8 % and 90.1 % respectively, and average area under the receiver operating characteristic curve (AUC) of 99.2 %, 98.4 % and 96.7 % respectively. The testing was done using 10-cross fold validation, so that the performance can be an accurate representation of the architectures classification capability in a clinical setting. After training/testing of the three individual models, a final ensemble model is made consisting of the three models. The ensemble model gives an average ACC and AUC of 96.3 % and 99.3 % respectively.

COMPARISON WITH EXISTING METHODS

This study outperforms previous studies, with increased ACC and AUC results coupled with use of small time windows (1 s) used for evaluation.

CONCLUSION

The proposed approach is promising for detecting seizure activity in unseen neonate data in a clinical setting.

Does the first hour of continuous electroencephalography predict neonatal seizures?

OBJECTIVE

Prolonged continuous video-electroencephalography (cEEG) is recommended for neonates at risk of seizures. The cost and expertise required to provide a real-time response to detected seizures often limits its utility. We hypothesised that the first hour of cEEG could predict subsequent seizures.

DESIGN AND SETTING

Retrospective multicentre diagnostic accuracy study.

PATIENTS

266 term neonates at risk of seizure or with suspected seizures.

INTERVENTION

The first hour of cEEG was graded by expert and novice interpreters as normal, mildly, moderately or severely abnormal; seizures were identified.

MAIN OUTCOME MEASURES

Association between abnormalities in the first hour of cEEG and the presence of seizures during total cEEG monitoring.

RESULTS

50/98 (51%) of neonates who developed seizures had their first seizure in the first hour of cEEG monitoring. The 'time-to-event' risk of seizure from 0 to 96 hours was 0.38 (95% CI 0.32 to 0.44) while the risk in the first hour was 0.19 (95% CI 0.15 to 0.24). cEEG background was normal in 48% of neonates, mildly abnormal in 30%, moderately abnormal in 13% and severely abnormal in 9%. Inter-rater agreement for determination of background was very good (weighted kappa=0.81, 95% CI 0.72 to 0.91). When neonates with seizures during the first hour were excluded, an abnormal background resulted in 2.4 times increased risk of seizures during the subsequent monitoring period (95% CI 1.3 to 4.4, p<0.003) while a severely abnormal background resulted in a sevenfold increased risk (95% CI 3.4 to 14.3, p<0.0001).

CONCLUSIONS

The first hour of cEEG in at-risk neonates is useful in identifying and predicting whether seizures occur during cEEG monitoring up to 96 hours. This finding enables identification of high-risk neonates who require closer observation.

Aminophylline for renal protection in neonatal hypoxic-ischemic encephalopathy in the era of therapeutic hypothermia

BACKGROUND

Neonates with hypoxic-ischemic encephalopathy (HIE) frequently develop acute kidney injury (AKI). Aminophylline has been shown to reduce severe renal dysfunction in neonates after perinatal asphyxia. However, the effect of aminophylline on renal function in neonates undergoing hypothermia has not been studied.

METHODS

A single-center, retrospective chart review of neonates cooled for moderate/severe HIE who received aminophylline for AKI was conducted to assess changes in urine output (UOP) and serum creatinine (SCr). Comparisons were also made to control neonates matched for hours of life who were cooled but unexposed to aminophylline.

RESULTS

Sixteen neonates cooled for HIE received aminophylline starting at 25 ± 14 h of life. Within 12 h of starting aminophylline, UOP increased by 2.6 ± 1.9 mL/kg/h. SCr declined by 0.4 ± 0.2 mg/dL in survivors over the first 4 days. When compared to control neonates, UOP increase was greater in the aminophylline group (p < 0.001). SCr declined in survivors in both groups, although baseline SCr was higher in the aminophylline group.

CONCLUSIONS

Aminophylline use in neonates with HIE undergoing hypothermia was associated with an increase in UOP and a decline in SCr. A randomized trial will be needed to establish a potential renal protective role of aminophylline.

IMPACT

The renal protective effect of aminophylline in neonates with HIE has not yet been studied in the context of therapeutic hypothermia. Aminophylline exposure in neonates cooled for HIE was associated with increased UOP and a similar decline in SCr when compared to control infants unexposed to aminophylline. Improved renal function after receiving aminophylline in this observational cohort study suggests the need for future randomized trials to establish the potential benefit of aminophylline in the HIE population undergoing hypothermia.

Neonatal Seizures Revisited

Seizures are the most common neurological disorder in newborns and are most prevalent in the neonatal period. They are mostly caused by severe disorders of the central nervous system (CNS). However, they can also be a sign of the immaturity of the infant’s brain, which is characterized by the presence of specific factors that increase excitation and reduce inhibition. The most common disorders which result in acute brain damage and can manifest as seizures in neonates include hypoxic-ischemic encephalopathy (HIE), ischemic stroke, intracranial hemorrhage, infections of the CNS as well as electrolyte and biochemical disturbances. The therapeutic management of neonates and the prognosis are different depending on the etiology of the disorders that cause seizures which can lead to death or disability. Therefore, establishing a prompt diagnosis and implementing appropriate treatment are significant, as they can limit adverse long-term effects and improve outcomes. In this review paper, we present the latest reports on the etiology, pathomechanism, clinical symptoms and guidelines for the management of neonates with acute symptomatic seizures.

Sex specific correlation between GABAergic disruption in the dorsal hippocampus and flurothyl seizure susceptibility after neonatal hypoxic-ischemic brain injury

Since neonatal hypoxia-ischemia (HI) disrupts the hippocampal (Hp) GABAergic network in the mouse and Hp injury in this model correlates with flurothyl seizure susceptibility only in male mice, we hypothesized that GABAergic disruption correlates with flurothyl seizure susceptibility in a sex-specific manner. C57BL6 mice were exposed to HI (Vannucci model) versus sham procedures at P10, randomized to normothermia (NT) or therapeutic hypothermia (TH), and subsequently underwent flurothyl seizure testing at P18. Only in male mice, Hp atrophy correlated with seizure susceptibility. The number of Hp parvalbumin positive interneurons (PV⁺INs) decreased after HI in both sexes, but TH attenuated this deficit only in females. In males only, seizure susceptibility directly correlated with the number of PV⁺INs, but not somatostatin or calretinin expressing INs. Hp GABA_B receptor subunit levels were decreased after HI, but unrelated to later seizure susceptibility. In contrast, Hp GABA_A receptor α1 subunit (GABA_ARα1) levels were increased after HI. Adjusting the number of PV⁺ INs for their GABA_ARα1 expression strengthened the correlation with seizure susceptibility in male mice. Thus, we identified a novel Hp sex-specific GABA-mediated mechanism of compensation after HI that correlates with flurothyl seizure susceptibility warranting further study to better understand potential clinical translation.

Continuous Electroencephalography Monitoring in Critically Ill Infants and Children

Continuous video electroencephalography (CEEG) monitoring of critically ill infants and children has expanded rapidly in recent years. Indications for CEEG include evaluation of patients with altered mental status, characterization of paroxysmal events, and detection of electrographic seizures, including monitoring of patients with limited neurological examination or conditions that put them at high risk for electrographic seizures (e.g., cardiac arrest or extracorporeal membrane oxygenation cannulation). Depending on the inclusion criteria and clinical characteristics of the population studied, the percentage of pediatric patients with electrographic seizures varies from 7% to 46% and with electrographic status epilepticus from 1% to 23%. There is also evidence that epileptiform and background CEEG patterns may provide important information about prognosis in certain clinical populations. Quantitative EEG techniques are emerging as a tool to enhance the value of CEEG to provide real-time bedside data for management and prognosis. Continued research is needed to understand the clinical value of seizure detection and identification of other CEEG patterns on the outcomes of critically ill infants and children.

Early Electroencephalogram Background Could Guide Tailored Duration of Monitoring for Neonatal Encephalopathy Treated with Therapeutic Hypothermia

OBJECTIVE

To evaluate whether features of the early electroencephalographic (EEG) background could guide the optimal duration of continuous video EEG monitoring for seizure detection in newborn infants treated with therapeutic hypothermia for hypoxic ischemic encephalopathy (HIE).

STUDY DESIGN

Retrospective cohort study of 114 consecutive infants treated with therapeutic hypothermia for moderate to severe HIE at a level IV neonatal intensive care unit (NICU) between 2012 and 2018. All infants were monitored with continuous video EEG through cooling and rewarming. Archived samples from the first 24 hours of these EEG traces were reviewed systematically and classified by background characteristics.

RESULTS

Electrographic seizures occurred in 56 of the 114 infants (49%). Seizure onset was within the first 24 hours after initiation of continuous video EEG in 49 if these 56 infants (88%), between 24 and 48 hours in 4 infants (7%), and >72 hours in 3 infants (5%). Infants with a normal or mildly abnormal EEG background either had seizure onset within the first 24 hours or never developed seizures. Four patients with seizure onset between 24 and 48 hours had markedly abnormal EEG backgrounds. The 3 patients with seizure onset beyond 72 hours had moderate or severely abnormal early continuous video EEG backgrounds.

CONCLUSIONS

The use of early continuous video EEG background categorization may be appropriate to guide the duration of continuous video EEG for infants with HIE treated with therapeutic hypothermia. Some infants may reasonably be monitored for 24 hours rather than throughout cooling and rewarming without a significant risk of missed seizures. This could have significant implications for continuous video EEG resource utilization.

Continuous EEG Monitoring in Neonates: One Size Does Not Fit All

[Box: see text]

EEG Monitoring of the Epileptic Newborn

PURPOSE OF REVIEW

Although differentiating neonatal-onset epilepsies from acute symptomatic neonatal seizures has been increasingly recognized as crucial, existing guidelines, and recommendations on EEG monitoring are mainly based on acute symptomatic seizures, especially secondary to hypoxic-ischemic encephalopathy. We aimed to narratively review current knowledge on neonatal-onset epilepsies of genetic, metabolic, and structural non-acquired origin, with special emphasis on EEG features and monitoring.

RECENT FINDINGS

A wide range of rare conditions are increasingly described, reducing undiagnosed cases. Although distinguishing features are identifiable in some, how to best monitor and detect less described etiologies is still an issue. A comprehensive approach considering onset, seizure evolution, ictal semiology, clinical, laboratory, EEG, and neuroimaging data is key to diagnosis. Phenotypic variability prevents precise recommendations, but a solid, consistent method moving from existing published guidelines helps in correctly assessing these newborns in order to provide better care, especially in view of expanding precision therapies.

Acute Symptomatic Seizures in Neonates

Seizures are an important sign of neurologic dysfunction in neonates, and they most often represent acute brain injury such as hypoxic-ischemic encephalopathy, stroke, or intracranial hemorrhage (acute symptomatic seizures). Clinical identification of seizures is not reliable since seizures in neonates often do not have an apparent clinical correlate; therefore, electroencephalography should be used to accurately diagnose and manage neonatal seizures. Seizures are refractory to initial loading doses of standard medications in >50% of cases. Since seizures are commonly associated with adverse acute and long-term outcomes, and the seizures themselves may result in additional brain injury, it is important to quickly recognize, diagnose, and treat seizures in neonates. Local practice pathways may optimize efficiency in assessment and treatment for affected newborns. Herein, we review the etiology, methods of diagnosis, treatment, and current knowledge gaps for neonatal seizures.

Seizure Prediction Models in the Neonatal Intensive Care Unit

PURPOSE

Conventional video-EEG monitoring is required to diagnose seizures accurately in neonates. This tool is resource-intense and has limited availability in many centers. Seizure prediction models could help allocate resources by improving efficiency in which conventional video-EEG monitoring is used to detect subclinical seizures. The aim of this retrospective study was to create a neonate-specific seizure prediction model using clinical characteristics and EEG background findings.

METHODS

We conducted a 3-year retrospective study of all consecutive neonates who underwent conventional video-EEG monitoring at a tertiary care pediatric hospital. Variables including age, EEG indication, high-risk clinical characteristics, and EEG background informed seizure prediction models based on a multivariable logistic regression model. A Cox proportional hazard regression model was used to construct time to first EEG seizure.

RESULTS

Prediction models with clinical variables or background EEG features alone versus combined clinical and background EEG features were created from 210 neonates who met inclusion criteria. The combined clinical and EEG model had a higher area under the curve for combined sensitivity and specificity to 83.0% when compared to the clinical model (76.4%) or EEG model (66.2%). The same trend of higher sensitivity of the combined model was found for time to seizure outcome.

CONCLUSIONS

While both clinical and EEG background features were predictive of neonatal seizures, the combination improved overall prediction of seizure occurrence and prediction of time to first seizure as compared with prediction models based solely on clinical or EEG features alone. With prospective validation, this model may improve efficiency of patient-oriented EEG monitoring.

The probability of seizures during continuous EEG monitoring in high-risk neonates

OBJECTIVE

We evaluated the impact of monitoring indication, early electroencephalography (EEG), and clinical features on seizure risk in all neonates undergoing continuous EEG (cEEG) monitoring following a standardized monitoring protocol.

METHODS

All cEEGs from unique neonates 34-48 weeks postmenstrual age monitored from 1/2011-10/2017 (n = 291) were included. We evaluated the impact of cEEG monitoring indication (acute neonatal encephalopathy [ANE], suspicious clinical events [SCEs], or other high-risk conditions [OHRs]), age, medication status, and early EEG abnormalities (including the presence of epileptiform discharges and abnormal background continuity, amplitude, asymmetry, asynchrony, excessive sharp transients, and burst suppression) on time to first seizure and overall seizure risk using Kaplan-Meier survival curves and multivariable Cox proportional hazards models.

RESULTS

Seizures occurred in 28% of high-risk neonates. Discontinuation of monitoring after 24 hours of seizure-freedom would have missed 8.5% of neonates with seizures. Overall seizure risk was lower in neonates monitored for ANE compared to OHR (P = .004) and trended lower compared to SCE (P = .097). The time course of seizure presentation varied by group, where the probability of future seizure was less than 1% after 17 hours of seizure-free monitoring in the SCE group, but required 42 hours in the OHR group, and 73 hours in the ANE group. The presence of early epileptiform discharges increased seizure risk in each group (ANE: adjusted hazard ratio [aHR] 4.32, 95% confidence interval [CI] 1.23-15.13, P = .022; SCE: aHR 10.95, 95% CI 4.77-25.14, P < 1e-07; OHR: aHR 56.90, 95% CI 10.32-313.72, P < 1e-05).

SIGNIFICANCE

Neonates who undergo cEEG are at high risk for seizures, and risk varies by monitoring indication and early EEG findings. Seizures are captured in nearly all neonates undergoing monitoring for SCE within 24 hours of cEEG monitoring. Neonates monitored for OHR and ANE can present with delayed seizures and require longer durations of monitoring. Early epileptiform discharges are the best early EEG feature to predict seizure risk.

Practice variation in anti-epileptic drug use for neonatal hypoxic-ischemic encephalopathy among regional NICUs

Background

While intercenter variation (ICV) in anti-epileptic drug (AED) use in neonates with seizures has been previously reported, variation in AED practices across regional NICUs has not been specifically and systematically evaluated. This is important as these centers typically have multidisciplinary neonatal neurocritical care teams and protocolized approaches to treating conditions such as hypoxic ischemic encephalopathy (HIE), a population at high risk for neonatal seizures. To identify opportunities for quality improvement (QI), we evaluated ICV in AED utilization for neonates with HIE treated with therapeutic hypothermia (TH) across regional NICUs in the US.

Methods

Children’s Hospital Neonatal Database and Pediatric Health Information Systems data were linked for 1658 neonates ≥36 weeks’ gestation, > 1800 g birthweight, with HIE treated with TH, from 20 NICUs, between 2010 and 2016. ICV in AED use was evaluated using a mixed-effect regression model. Rates of AED exposure, duration, prescription at discharge and standardized AED costs per patient were calculated as different measures of utilization.

Results

Ninety-five percent (range: 83–100%) of patients with electrographic seizures, and 26% (0–81%) without electrographic seizures, received AEDs. Phenobarbital was most frequently used (97.6%), followed by levetiracetam (16.9%), phenytoin/fosphenytoin (15.6%) and others (2.4%; oxcarbazepine, topiramate and valproate). There was significant ICV in all measures of AED utilization. Median cost of AEDs per patient was $89.90 (IQR $24.52,$258.58).

Conclusions

Amongst Children’s Hospitals, there is marked ICV in AED utilization for neonatal HIE. Variation was particularly notable for HIE patients without electrographic seizures, indicating that this population may be an appropriate target for QI processes to harmonize neuromonitoring and AED practices across centers.

Levetiracetam optimal dose-finding as first-line treatment for neonatal seizures occurring in the context of hypoxic-ischaemic encephalopathy (LEVNEONAT-1): study protocol of a phase II trial

INTRODUCTION

Therapeutic schedules for treating neonatal seizures remain elusive. First-line treatment with phenobarbital is widely supported but without strong scientific evidence. Levetiracetam (LEV) is an emerging and promising antiepileptic drug (AED). The aim of this phase II trial is to determine the benefits of LEV by applying a strict methodology and to estimate the optimal dose of LEV as a first-line AED to treat seizures in newborns suffering from hypoxic-ischaemic encephalopathy.

METHODS AND ANALYSIS

LEVNEONAT-1 is an open and sequential LEV dose-finding study. The optimal dose is that which is estimated to be associated with a toxicity not exceeding 10% and an efficacy higher than 60%. Efficacy is defined by a seizure burden reduction of 80% after the loading dose. Four increasing dose regimens will be assessed including one loading dose of 30, 40, 50 or 60 mg/kg followed by eight maintenance doses (ie, a quarter of the loading dose) injected every 8 hours. A two-patient cohort will be necessary at each dose level to consider an upper dose level assignment. The maximal sample size expected is 50 participants with a minimum of 24 patients or fewer in the case of a high rate of toxicity. Patients will be recruited in five neonatal intensive care units beginning in October 2017 and continuing for 2 years. In parallel, the LEV pharmacokinetics will be measured five times (ie, 30 min; 4 and 7 hours after the loading dose; 1-3 hours and 12-18 hours after the last maintenance dose).

ETHICS AND DISSEMINATION

Ethics approval has been obtained from the regional ethical committee (2016-R25) and the French Drug Safety Agency (160652A-31). The results will be published in a peer-reviewed journal. The results will also be presented at medical meetings.

TRIAL REGISTRATION NUMBER

NCT02229123; Pre-results.

Seizure classification, etiology, and management

The first weeks of life are a time of heightened risk for seizures due to age-dependent physiologic features of the developing brain that lead to increased neuronal excitation and decreased inhibition. Usually, seizures in neonates are a symptom of an acute brain injury; seizures are only rarely due to neonatal-onset epilepsy syndromes. Neonatal seizures are harmful to the developing brain; early and accurate diagnosis is critical. For suspected seizures, EEG monitoring should be initiated as soon as is feasible, in order to evaluate for events of concern, screen for subclinical seizures, and assess the EEG background. Amplitude-integrated EEG can provide excellent complementary data, particularly with regard to evolution of background patterns, but has limited sensitivity to detect individual neonatal seizures. An urgent and systematic approach to precise etiologic diagnosis is key for optimal management and estimates of prognosis. Evaluation of the seizure etiology must occur in parallel with initiation of appropriate treatment. It is critical that neonatologists and neurologists develop hospital-specific, consensus-based practice pathways for neonatal seizure evaluation and treatment. Such practice pathways can streamline medical decision making, facilitate rapid medication administration, and potentially decrease seizure burden and optimize outcomes. Herein, the pathophysiology, epidemiology, treatment, and long-term management considerations for neonatal seizures are presented.

Sedation and Analgesia Influence Electroencephalography Monitoring in Pediatric Neurocritical Care

OBJECTIVES

We assessed neuroactive medication use in critically ill children who require neurological consultation and evaluated the associations between administration of these medications and continuous electroencephalography (cEEG) utilization and seizure frequency.

METHODS

We evaluated exposure to sedatives, analgesics, anesthetics, and paralytics in consecutive patients (0 days to 18 years) for whom neurological consultation was requested in three intensive care units (ICUs) [neonatal (NICU), pediatric (PICU), and cardiothoracic (PCTU)]) at one children's hospital. We assessed cEEG usage and seizure incidence in relation to drug exposure.

RESULTS

From November 2015 to November 2016, 300 consecutive patients were evaluated (93 NICU, 139 PICU, and 68 PCTU). Ninety-seven (32%) were receiving ≥1 sedative infusion at the time of consultation [NICU 7 (8%), PICU 50(36%), PCTU 40 (58%%]; 91 (30%) received ≥1 paralytic agent within the preceding 24 hours. Continuous electroencephalography was performed more often for patients treated with sedative infusions (81 of 97 versus 133 of 203, P = 0.001) and paralytic medications (80 of 91 versus 134 of 209, P < 0.001) within 24 hours preceding consultation than those who were not. Sixty-eight of 214 (32%) had electrographic seizures (65 of 68 within initial 24 hours of monitoring); seizures were less common among patients who had received sedative infusions (18 of 81 versus 51 of 133, P = 0.014). In multivariable analysis of seizure likelihood, only younger age was associated with increased risk (P = 0.037).

CONCLUSIONS

Critically ill infants and children are frequently treated with sedatives, anesthetics, analgesics, and paralytics. Neuroactive medications limit bedside neurological assessments and, in this cohort, were associated with increased cEEG usage. Our data underscore the need to study the effect of these medications on clinical care and long-term outcomes.