Treatment of Developmental/Epileptic Encephalopathy With Spike-Wave Activation in Sleep

The concept of electrical status epilepticus in sleep (ESES) was introduced in 1971 when Patry et al¹ described 6 children with severe neurodevelopmental delays and robust activation of epileptiform discharges in slow-wave sleep. The number of 1-second bins containing a spike-wave discharge divided by overall time assessed calculated a spike-wave index (SWI)—which if ≥85% in slow-wave sleep—became the defining threshold for this condition (Figure). 

Since 1971, a number of complex acronyms to define clinical and electrographic phenotypes have been proposed but disagreement has persisted. For many, ESES represented an electrographic phenotype whereas continuous spike-wave in sleep and Landau-Kleffner syndrome represented clinical syndromes. Other clinical phenotypes have included forme fruste epileptic encephalopathies with marked activation of epileptiform discharges in sleep, including atypical benign focal epilepsy (eg, pseudo-Lennox syndrome, atypical self-limited epilepsy with centrotemporal spikes [SeLECTS]) and epileptic opercular syndrome. Robust spike-wave activation in sleep (SWAS) characterizes each of these disorders, and the explicit expectation is that the specific neurologic phenotypes are caused by SWAS. 

Figure. Developmental/epileptic encephalopathy with spike-wave activation in sleep.
Abbreviations: ASM, antiseizure medication; DEE-SWAS, developmental/epileptic encephalopathy with spike-wave activation in sleep; ETX, ethosuximide; IVIg, intravenous immunoglobulin; LEV/BRV, levetiracetam/brivaracetam; SWI, spike-wave index; VPA, valproic acid; ZNS, zonisamide.

In 2022, the International League Against Epilepsy proposed new terminology for several epilepsy syndromes.² In this proposal, all conditions that may be defined as ESES or continuous spike-wave in sleep were placed under the broad umbrella of developmental/epileptic encephalopathy with spike-wave activation in sleep (DEE-SWAS). The acquired epileptic aphasia/auditory agnosia phenotype of Landau-Kleffner syndrome was retained as a clinically distinct EE-SWAS subtype. 

Due to the lack of an established electroclinical phenotype, there has been no standardization of therapy. Treatment protocols and even the threshold at which to treat can be variable. There is also a lack of consensus on how to measure a significant outcome other than by comparison of SWI before and after treatment, and the clinical effect may not be taken into consideration.

Treating the EEG to Treat the Patient
Treatments for DEE-SWAS include spike-suppressing antiseizure medications (ASMs; eg, valproic acid, levetiracetam, zonisamide, ethosuximide), high-dose benzodiazepines (eg, diazepam, clobazam), immunomodulatory therapies (eg, oral and intravenous [IV] steroids, IV immunoglobulins), uncommon ASMs (eg, sulthiame, acetazolamide, amantadine), ketogenic diet, and surgery (eg, multiple subpial transections, corpus callosotomy, focal resections). Some clarity was provided after a pooled analysis of 575 cases of treated ESES demonstrated that improvements in either cognition or EEG were most commonly observed in therapeutic trials of surgery (90%), steroids (81%), or benzodiazepines (68%); standard ASMs were least effective (49%).³

Benzodiazepines vs Steroids
Even among treatments such as benzodiazepines and steroid therapy, there is a paucity of data to support the most effective treatment regimen. With regard to benzodiazepines, some of the most commonly used treatment protocols have included high-dose diazepam.^4,5 Challenges to high-dose diazepam include the early recommendations to start this therapy with intensive care unit–level monitoring due to the risk of respiratory depression; most institutions now perform this on a standard inpatient floor due to the rarity of significant respiratory depression (Table 1). Clobazam is an attractive option that can be started in the outpatient setting for some children with concurrent epilepsy; this may provide a long-term option to help manage their DEE-SWAS phenotype and seizures.

There are descriptions of various steroid therapy regimens in the literature including dexamethasone, hydrocortisone, adrenocorticotropic hormone, prednisone, and IV methylprednisolone.^6-8 The most efficacious steroid formulation and method of delivery (ie, oral daily vs oral pulse vs IV pulse) have not been established in the management of DEE-SWAS, nor has the exact duration of treatment. Again, the choice of steroid and duration tends to fall to expert opinion and preferences or experiences. There are reports describing dual benzodiazepine and steroid treatments in refractory cases, akin to combination therapy of steroids and vigabatrin in infantile spasms (Table 2).⁹

A survey of treatment preferences in North America in 2014 suggested that high-dose benzodiazepines were the most common first choice with diazepam being the more common medication over clobazam.¹⁰ These findings were recapitulated in a 2021 multicenter collaborative research project that sought to evaluate current treatment practices in the United States.¹¹

The Randomized European Trial of Steroids vs Clobazam Usage for Encephalopathy With Electrical Status Epilepticus in Sleep (RESCUE ESES) attempted to answer the “steroids vs benzodiazepines” debate but was unsuccessful due to early termination of the study secondary to poor enrollment. The trial assessed clobazam doses up to 1.2 mg/kg/d and steroid regimens that could include oral prednisolone 2 mg/kg/d up to 60 mg/d for 1 month followed by 1 to 2 mg/kg/d up to 60 mg/d for the next 5 months or IV methylprednisolone 20 mg/kg/d for 3 days, repeated every 4 weeks for 6 months. The frequency of side effects was comparable between groups including 45% of children receiving corticosteroids (predominantly weight gain) and 52% of children receiving clobazam (predominantly fatigue and behavioral disturbance). There was an improvement noted in intelligence quotient in 25% of children in the corticosteroid arm which was not observed in the clobazam group; however, this change was not reflected in the cognitive sum score. There was no difference between groups in EEG responder rate (a reduction in SWI ≥25%) or number of individuals with SWI <50%. Despite the underpowered nature of the study, the authors conclude that the intelligence quotient change is enough to support early corticosteroid treatment over clobazam treatment.¹² Recommended dosing for ASMs including corticosteroids is provided in Table 3.

Epilepsy Surgery and Neuromodulation Therapies
Whereas resective surgery is not often considered in the treatment of interictal activity, it should be considered when the interictal burden is contributing to epileptic encephalopathy because resective surgery can potentially improve cognition.¹³ Successful surgical treatment of epileptic encephalopathy improves development in people with infantile spasms with unilateral structural pathology.¹⁴ In the case of DEE-SWAS, surgical workup is often not considered unless there is concurrent refractory focal epilepsy; however, studies have shown that hemispherectomy leads to resolution of SWAS and cognitive improvement in cases with a unilateral structural etiology. Given that cortical malformations and thalamic injury are the most common structural etiologies, an approach with disconnection of the thalamus is imperative for successful surgical treatment.¹⁵

Multiple subpial transections (MSTs) were previously considered for the theoretical benefit of interrupting spike-wave generation through disrupting cortical layers while preserving eloquent cortex, but results of the initial case series were not able to be duplicated, and the practice largely fell out of fashion.^16,17 In the attempt to disrupt secondary bilateral synchrony, callosotomy has also been considered including in nonlesional cases. The number of individuals treated with this modality has been small, with only 1 individual demonstrating resolution of ESES.¹⁸ The utility of callosotomy would also be questionable in the setting of DEE-SWAS, given that it could impair language acquisition in a developing child. 

A meta-analysis by van den Munckhof et al³ found that surgical treatments (MST, corpus callosotomy, hemispherotomy, focal or lobar resection, or other disconnection), when pooled together, were successful in 90% of individuals, with 80% of participants reporting cognitive improvement and 80% of participants experiencing improvement in EEG background. The overall odds ratio of treatment response was largest for the surgical group at 9.8. Even after subgroup analysis of individuals treated consecutively with different modalities, surgery had the best response rate at 93%. This argues that for individuals with focal or hemispheric structural etiologies, surgery should be considered for the treatment of DEE-SWAS, even in the absence of refractory seizures. However, half of these cases (31 of 62) were MSTs. Given that MSTs are no longer common clinical practice, updated data on more modern surgical approaches are needed. 

The utility of neuromodulation including transcranial magnetic stimulation, transcranial direct current stimulation, responsive neurostimulation, vagus nerve stimulation, and deep brain stimulation for DEE-SWAS has yet to be formally investigated, with forthcoming clinical trials anticipated.¹⁹ Experience is limited to case reports, such as with vagus nerve stimulation.²⁰ Noninvasive approaches including transcranial magnetic stimulation and transcranial direct current stimulation are particularly attractive for potential use in DEE-SWAS cases that may resolve with age. More broadly, individuals with nonlesional, medically refractory DEE-SWAS without resective targets may be candidates for neuromodulation.¹³ Bilateral thalamic responsive neurostimulation or deep brain stimulation might be of particular interest, given the role of the thalamus in the corticothalamic generation of sleep-activated spike-wave discharges, especially given that perinatal thalamic injury is a common structural risk factor for DEE-SWAS.²¹ Nuances regarding targeting particular thalamic nuclei, especially in injured thalami with perturbed internal architecture, remain prospective challenges. However, beyond their theoretical impact, as is the case with many of the commonly used therapies in pediatric epilepsy, neuromodulation devices do not have Food and Drug Administration clearance for the treatment of pediatric patients with DEE-SWAS; therefore, further studies are needed to determine their safety and efficacy in this population, and any current use would be considered off-label. 

Defining Treatment Success
Quantifying meaningful treatment responses in SWAS—or in the degree of encephalopathy itself—is a challenge. Stipulating SWI as a reliable measure of SWAS, many assess SWI change against a defining threshold, such as 50% or 85%. However, universal thresholds may prove arbitrary among a heterogeneous population and limited by variability in visual EEG quantification methods. Furthermore, small cross-threshold changes (eg, 55% to 45%) may fail to demonstrate clinical significance. Quantifying change in absolute terms (eg, a 50-percentage point reduction from 80% to 30%) or relative terms (eg, 50% reduction from 60% to 30%) may be more significant; however, it remains unclear which is more clinically meaningful. Spatial or morphologic EEG features beyond SWI may prove to be clinically important as well. Non–SWI-related measures, such as sleep spindle metrics, should also be further evaluated as potential treatment biomarkers. 

The impact of encephalopathy is quantified using cognitive and behavioral measures, but objective assessment is stymied in part by limited feasibility and availability of testing methods. Recurrent comprehensive neuropsychologic assessment remains impractical. No tool exists that can quickly and reliably track ongoing responses across multiple treatments weeks to months apart, especially considering the breadth of cognitive abilities in this heterogeneous population. Psychoeducational testing as part of an individualized education plan through school districts is not always widely available or equitable. Thus, treatment response is often subjective and reported by caregivers. 

Measures of SWAS and encephalopathy are not always synchronous or proportional. SWI may change although clinical response lags or fails to improve, potentially changing the causative relationship between SWAS and encephalopathy. Conversely, caregivers may report considerable subjective improvement without appreciable changes in SWI or objective neuropsychologic measures, challenging the sensitivity of these measures. Ongoing work reconfiguring traditional developmental milestones into developmental “inch stones” may better capture potentially clinically meaningful changes in individuals with DEE.²² 

Conclusions and Future Directions
The updated nomenclature of DEE-SWAS from the International League Against Epilepsy represents an important unifying first step in connecting clinicians and researchers. Thoughtful monitoring and assessment for at-risk individuals with serial EEGs may help predict those at risk of conversion to a DEE-SWAS phenotype and initiate preventive or prophylactic therapy. Azeem et al²³ found that individuals with perinatal stroke showed higher spike frequency and lower delta frequency preceding onset of ESES. This could lead to the consideration of a trial similar to the Long-Term, Prospective Study Evaluating Clinical and Molecular Biomarkers of Epileptogenesis in a Genetic Model of Epilepsy–Tuberous Sclerosis Complex (EPISTOP) and the Preventing Epilepsy Using Vigabatrin In Infants With Tuberous Sclerosis Complex (PREVeNT; NCT02849457) trial.^24,25 Such prospective trials could inherently evaluate the efficacy and tolerability of different treatment paradigms. The failures of attempted clinical trials such as RESCUE ESES should not dissuade future involvement from clinicians and industry. The aforementioned efforts to improve and standardize treatments and measurements of efficacy should lead to more clinical trials and measurable successful outcomes.