Status epilepticus (SE) is a common pediatric neurological emergency. The definition of SE has evolved over time, to reconcile the likelihood of spontaneous seizure cessation based on pathophysiology, versus the operational urgency to achieve seizure termination and avoid adverse consequences. We have summarized some of the key definitions in Box I. The temporal evolution of SE is conventionally divided into early (5-30 min), established (30-60 min), and refractory (≥60 min) phases [1].

Box I Conceptual Evolution of the Definition of Status Epilepticus and Other Related Entities Conceptual basis for definition of status epilepticus • Mechanistic: failure of homeostatic mechanisms to control a seizure. • Physiological: ictal compromise of neuronal survival. • Operational: ictal duration to warrant emergent treatment. Definition of status epilepticus • 1904: Seizures occur so frequently that coma and exhaustion are continuous between seizures. (Clark and Prout) • 1940: Severest form of seizures in which the post-convulsive sleep of one attack is cut short by development of the next (seizure). (Wilson) • 1964: Seizure persists for a sufficient length of time or is repeated frequently enough to produce a fixed and enduring epileptic condition. (ILAE 1st International Classification of Epileptic Seizures) • 1981: Seizure persists for a sufficient length of time or is repeated frequently enough that recovery between attacks does not occur. (ILAE revision) • 1993: More than 30 minutes of (1) continuous seizure activity or (2) two or more sequential seizures without full recovery of consciousness between seizures. (AES working group on SE) • 1998: Two or more generalized convulsions, without full recovery of consciousness between seizures, or continuous convulsive activity for more than 10 minutes (overt generalized convulsive SE). (Veterans Affairs Study) • 1999: Operational definition: generalized convulsive SE in adults and children >5 years old refers to ³5 min of continuous seizures, or ³2 discrete seizures between which there is incomplete recovery of consciousness Mechanistic definition: failure of the “normal” factors that serve to terminate a typical generalized tonic-clonic seizure. (Lowenstein, Bleck, and Macdonald) • 2015: A condition resulting either from the failure of the mechanisms responsible for seizure termination or from the initiation of mechanisms which lead to abnormally prolonged seizures (after time point t1). It is a condition that can have long-term consequences (after time point t2), including neuronal death, neuronal injury, and alteration of neuronal networks, depending on the type and duration of seizures.Where: for tonic-clonic SE (t1=5 min, t2=30 min), focal SE with impaired consciousness (t1=10 min, t2>60 min), and absence SE (t1=10-15 min, t2 unknown) (ILAE Task Force on Classification of SE) Other entities • Refractory SE: SE that fails to respond to adequately used 1st (BDZ) and 2nd line (appropriately chosen non-BDZ) medications. • Super-refractory SE: SE that persists for 24 hours or more after initiation of 3rd line medications (anesthetics) or recurrence of seizure during withdrawal of the anesthetics. • NORSE: New onset of refractory SE without a clear acute or active structural, toxic, or metabolic cause in a patient without active epilepsy or other preexisting relevant neurological disorder. It is a clinical presentation and not a specific diagnosis. • FIRES: NORSE with a prior febrile infection between 2 weeks and 24 hours prior to onset of refractory SE, with or without fever at the actual onset of SE. It is a sub-category of NORSE. AES: American Epilepsy Society; ILAE: International League Against Epilepsy; FIRES: Febrile Infection Related Encephalopathy Syndrome; NORSE: New-Onset Refractory Status Epilepticus; SE: Status Epilepticus].

 
There are two common definitions of refractory SE (RSE), which may operationally converge. These include a convulsive seizure lasting longer than 60 minutes, which may be continuous or intermittent without return to baseline mental status; and an acute convulsive seizure that fails to respond to ³2 anti-seizure medications (ASMs) including at least one non-benzodiazepine ASM. In our experience, the latter criteria is more commonly used in practice. At present, there is a dearth of sufficient high-quality evidence to formulate a uniform management strategy for RSE, resulting in variability in treatment approaches and the choice of therapeutic endpoint(s). Partly because of this, about 15%-35% of RSE patients fail to achieve desired treatment response, and progress to super-refractory SE (SRSE), which is defined as an SE that continues for 24 hours or more despite anesthetic treatment, or recurs on an attempted wean of the anesthetic regimen [2,3].

We, herein, provide an overview of RSE and SRSE. However, we have not reviewed the epidemiology of RSE/SRSE, because of insufficient data for these entities separate from SE. We have also not discussed the pathophysiology of drug-resistance in SE, because the relative importance of different molecular mechanisms remains uncertain at present, and does not inform treatment approach. Finally, our discussion is focused on RSE/SRSE evolving from a convulsive seizure with impaired awareness, and does not address epilepsia partialis continua.

Causes and Risk Factors

The underlying causes of RSE/SRSE in a patient with no previous history of seizures, or possible triggers which may precipitate RSE/SRSE in a patient known to have epilepsy, are similar to those responsible for the more common convulsive SE. In general, any pathology which can trigger an acute symptomatic seizure can cause an SE which may progress to an RSE/SRSE. This includes neurological and systemic infections, acute vascular events, traumatic brain injury, and immune, metabolic, or toxic encephalopathies.

Febrile SE: Lessons from FEBSTAT

In the pediatric population, prolonged febrile seizures constitute the most common subgroup and account for up to 35% of all episodes of SE [4]. The importance of recognizing and promptly managing febrile SE has been underscored by the FEBSTAT study. In this study, 71% of patients with acute changes on brain magnetic resonance imaging (MRI) after febrile SE, were found to have obvious hippocampal sclerosis 1 year later [5,6]. Given the importance of hippocampal sclerosis as a substrate for drug-resistant temporal lobe epilepsy, this finding emphasizes the direct association between the two. Another important finding from FEBSTAT study was the association of seizure duration with the time to initial treatment with benzodiazepines (BDZs), where every 2.7 min delay in the initial treatment was associated with 1.3 min increase in seizure duration [6,7]. Another study which found shorter median duration of febrile SE (35 min) compared to FEBSTAT (68 min), potentially attributed it to shorter time to access emergency medical services and initial treatment [8].

Infectious Causes

The etiology of RSE/SRSE varies geographically, and studies from India have noted a predominance of infectious causes [9-11]. In a series of 148 adults with encephalitis, 18 were diagnosed with SE, predominantly in those with herpes simplex virus (HSV) infection or Japanese encephalitis [12]. Children were noted to be more susceptible to have encephalitis-related SE in this cohort. The diagnosis of HSV is important in SE/RSE patients, as treatment with acyclovir within 24 hours of onset has been shown to be associated with better prognosis [13]. A recent study reported convulsive SE in 41 patients with neurocysticercosis, although they did not progress to RSE/SRSE [14]. The duration of SE was found to be shorter in patients with single calcific lesion compared to those with degenerating cysts. In another study including 141 children presenting emergently with acute convulsive seizures, 49% were found to have neurocysticercosis, though again the progression to RSE was not reported [11]. There is a paucity of data on epidemiology of RSE/SRSE from India in other common infections including acute bacterial meningoencephalitis, cerebral malaria, and dengue. Differentiating acute symptomatic SE in a young child with a neurological infection, from a febrile SE, is a challenging and important consideration.

Autoimmune and Genetic Etiologies

Regarding RSE/SRSE, it is important to recognize two particular groups of diseases, which may prompt an etiology-specific treatment or prognostication early in the course of management. The first group includes autoimmune encephalitis. In children and adolescents, the leading autoimmune entity is anti-NMDA-receptor encephalitis, which can present with non-specific manifestations, such as, fever or headache in the prodromal stage. The next stage, involving the cerebral cortex, is the one which can present with RSE/SRSE in additional to behavioral symptoms. In children, these behavioral symptoms can take the form of new-onset temper tantrums in an otherwise well-adjusted child, unexplained episodes of aggression, and speech disturbances. Compared to the predominant temporal localization for seizures in adults with anti-NMDAR encephalitis, extra-temporal seizures, and sometimes even diffuse bilateral ictal onset can be seen in children, and should not exclude this consideration [15,16]. The other limbic encephalitis are also being increasingly recognized in children, and may often not occur as a para-neoplastic syndrome. In addition to seizures, which can progress to SE/RSE, other manifestations of limbic encephalitis can be relatively non-specific in children. The behavioral phenotype of limbic encephalitis in adults often consists of disturbances in recent memory, and affective symptoms including irritability, depression, or hallucinations. However, in children, non-specific disturbances of executive function (attention control, social inhibition, regulation of purposeful behavior) are often seen, and should be evaluated in a developmental context [17]. The second important etiologic subgroup of RSE/SRSE, which should be evaluated promptly, comprises of certain genetic epilepsies. These include, but are not limited to, Dravet syndrome and other sodium channelopathies, ring chromosome 20, pathogenic variants in polymerase-G or amino-acyl-tRNA synthetase genes affecting mito-chondrial function, and Angelman syndrome [18].

Seizure Triggers

About 16-38% SE episodes occur in children with a prior diagnosis of epilepsy, with low anti-seizure medication levels being the most common risk factor [4]. Low levels of regular anti-seizure medicines can, in turn, result from lack of adherence, scheduled withdrawal, insufficient dose, interaction with other concurrent medications, or growth spurt, etc. Other common precipitants of SE include inter-current illness, exposure to a known trigger (e.g. sleep deprivation in some idiopathic generalized epilepsy syndromes), or metabolic decompensation.

Diagnostic Evaluation

The goals of evaluation in RSE/SRSE include diagnostic verification, particularly in patients with subtle and non-convulsive SE (NCSE); monitoring for therapeutic efficacy; diagnosis of the underlying etiology and/or risk factors; and early recognition of multisystem complications. We have summarized these in Table I, and briefly discuss the role of electroencephalography (EEG) and neuroimaging below.

Table I Selected Investigations Useful in Management of Refractory and Super-refractory Status Epilepticus

EEG Monitoring

The foremost utility of EEG monitoring is to detect ongoing subclinical seizure or NCSE after the convulsive seizure has apparently terminated. Hence, professional societies have recommended continuous EEG monitoring to be initiated within 1 hour of SE, and continued for at least 24-48 hours, often longer in the presence of altered consciousness [19-21]. Although the EEG diagnosis of NCSE has been fraught with disagreement, recent studies have shown good diagnostic validity and inter-rater agreement for the Salzburg criteria [22,23]. In patients without known epileptic encephalopathy, occurrence of >2.5 Hz epileptiform dischargesmay be sufficient for the diagnosis of NCSE. In case of epileptiform discharges £2.5 Hz, or the ictal pattern limited to rhythmic slow waves, it is important to evaluate for electroencephalo-graphic improvement with a bolus of intravenous BDZ; associated subtle clinical manifestations; and typical spatiotemporal evolution in the morphology, frequency, or locus of the electroencephalographic pattern. The availability of synchronized video monitoring, allowing simultaneous review of video-EEG is important in such situations. The diagnosis of NCSE becomes more challenging in patients with a known epileptic encephalopathy. In such patients, diagnosis of NCSE requires evidence of a distinct increase in the locus, field, or frequency of epileptiform discharges, and an obvious change in clinical state, or response to an intravenous BDZ bolus [23]. Hence, in these patients, it is important to have a good understanding of the baseline EEG pattern prior to the onset of RSE, based on a personal review of previous EEGs.

Other periodic and rhythmic EEG patterns, which do not meet the criteria for NCSE, are often labelled as lying on an interictal-ictal continuum. Although, this terminology is an useful EEG descriptor, its clinical significance remains uncertain. However, recognition of periodic patterns including lateralized periodic discharges (LPDs) or generalized periodic discharges (GPDs) may be important because of some association with the risk of seizure recurrence and certain specific conditions, such as, LPDs with stroke and herpes encephalitis [24]. Rarely, other EEG patterns may suggest a specific etiology, for example, extreme delta brushes in anti-NMDAR encephalitis.

Use of quantitative EEG has been increasing recently in neuro-critical care practice, due to integration of analysis software into EEG reading suites [25]. During management of RSE/SRSE, quantitative analysis of EEG trends may be helpful in evaluation of the suppression ratios over the course of treatment, recognition of periodic patterns or seizure recurrence, and comparison of power ratios in canonical EEG frequency bands. However, there is a paucity of studies validating use of quantitative EEG for prediction of outcomes in RSE/SRSE. Hence, it is important to verify any observations on quantitative trends, with a careful review of raw EEG. In our practice, the two most common uses of quantitative EEG in the management of RSE/SRSE include temporal analysis of suppression ratios, and quantitation of seizure burden once the specific imprint of a given seizure type is recognized on the density spectral array or amplitude integration. A detailed discussion of the methods, utility, and pitfalls of quantitative EEG is outside the scope of this paper.

The second most important role of EEG monitoring in the management of RSE/SRSE is to establish the therapeutic target. In these children, clinical evaluation for continued seizure activity is often non-informative, and becomes more so as the anesthetic infusions are increased, or neuromuscular blockers are used to facilitate airway control. At this point, the EEG usually shows electrographic bursts interrupted by periods of relative diffuse amplitude suppression. The optimal EEG target for treatment of RSE/SRSE is not strictly defined, and have included, suppression of seizures, varying degrees of suppression-burst pattern, and nearly complete EEG suppression [26]. Our practice is to aim for complete suppression of seizures, and achieve 50%-70% suppression ratio. Once this therapeutic goal is achieved, EEG monitoring is continued for 24-48 hours, when attempts are first made to wean off the infusions. EEG monitoring is important at this point to recognize the occurrence of emergence seizures or bursts with epileptiform elements, though these are not always treated individually.

Neuroimaging

The primary goal of neuroimaging in RSE/SRSE is to help identify the etiology, which in some cases, may lead to a specific urgent intervention. Secondly, identification of a potentially epileptogenic lesion may prompt conside-ration of emergency epilepsy surgery for RSE/SRSE in centers where such expertise is available. However, the interpretation of MRI studies early in the course of an RSE/SRSE can be challenging due to seizure-related or treatment-related changes. Specifically, it may be worthwhile to distinguish vasogenic edema reflected by T2/FLAIR hyper-intensities, from cytotoxic edema predominantly manifested as diffusion restriction. This is because vasogenic edema can sometimes point to an underlying etiology, whereas cytotoxic edema often represents compromised neuronal integrity from RSE/SRSE. In many patients there is an overlap between the two, and it is helpful to follow their neuroimaging over time. It is also important to look for confounding factors, for example, vigabatrin can be associated with both T2 hyper-intensities and restricted diffusion, predominantly in corpus callosum and deep gray structures [27,28]. In the FEBSTAT study, about 10% of children with a single febrile SE episode were found to have altered signal in the hippocampal region. After one year, 86% of them showed hippocampal volume loss and 71% had evidence of hippocampal sclerosis [5], emphasizing the need for longitudinal neuroimaging to ascertain neuronal injury in patients with RSE/SRSE.

Treatment

The primary goal of treatment is the termination of clinical and electrographic SE. However, the objectives of treatment in RSE/SRSE are somewhat different from those in early/established SE. In patients with early/established SE, the treatment efforts are focused on rapid seizure control to avoid neuronal injury. However, by definition, RSE represents a situation where pathophysiological mechanisms that support drug-response in established SE have been overwhelmed or altered, and hence, conventional treatment approaches are unlikely to be successful. Further, in SRSE, it is very likely that excitotoxic and other processes which compromise neuronal survival, have already been activated. Hence, the goal of treatment in RSE/SRSE is to limit or reverse these processes, prevent their downstream consequences, and salvage end-organ function to the maximum possible extent [29]. Clinically, this means that the treatment plan should encompass not only seizure control, but to also avoid, anticipate, and manage multisystem dysfunction resulting from ongoing seizure activity, from medications, and from prolonged unconsciousness and immobility [30]. While the intensive care related to RSE/SRSE is outside the scope of this paper, we have briefly summarized the first and second line treatment for SE, and then discussed specific therapeutic approaches to RSE/SRSE in more detail.
 
First- and Second- line Drug Treatments

The American Epilepsy Society guidelines published in 2016, have synthesized available evidence for drug-treatment of early and established SE [31]. There is consensus that BDZs represent the initial treatment of choice for SE, with equipoise among intramuscular midazolam (MDZ), intravenous lorazepam (LZP), and intravenous diazepam (DZP). This equipoise is based on three class I randomized controlled trials, which are summarized below. The Veterans Affairs study compared four interventions including DZP followed by phenytoin (PHT), LZP, phenobarbital (PHB), and PHT alone [32]. There were no significant differences among these 4 interventions on intention-to-treat (ITT) analysis. However, on pairwise comparisons in patients with generalized convulsive SE, LZP was superior to PHT, for cessation of all motor and EEG seizure activity within 20 min of starting drug infusion and no recurrence of seizure activity during the next 40 min. The Rapid Anticonvulsant Medication Prior to Arrival Trial (RAMPART), showed equivalence between intramuscular MDZ and intravenous LZP for the proportion of patients who achieved seizure termination without need for additional rescue [33]. However, the auto-injectors used in RAMPART are not available commercially, and whether conventional intramuscular MDZ injection will have similar efficacy is open to question. The third class I trial, showed equivalence between intravenous LZP and DZP for the proportions of patients with seizure termination, need for assisted ventilation, recurrence within 1 hour, and time to SE termination [34].

Only when these three options are not available, or not feasible due to challenges in achieving intravenous access in a convulsing child, other treatments may be considered, including rectal DZP, intranasal MDZ, and buccal MDZ. A network meta-analysis of 16 studies, specifically looking at non-venous treatments for acute convulsive seizures, found intramuscular and intranasal MDZ to be superior to other comparators for clinically relevant outcomes [35]. Specifically, this meta-analysis excluded data from RAMPART in their quantitative synthesis, thus independently supporting the efficacy of conventional intramuscular MDZ and encouraging the use of intranasal MDZ. From a practical standpoint, MDZ has the advantage of not needing refrigeration for storage. In US, MDZ nasal spray has been recently approved to treat seizure clusters and acute repetitive seizures [36]. Intranasal formulations of DZP and rapid systemic delivery of alprazolam (staccato alprazolam) by inhalation of thermally generated aerosol from a single-use hand-held device, are currently undergoing evaluation [37,38].

Appropriate first-line drug-treatment results in seizure termination in about 65% of patients. However, the subsequent choice of treatment is not driven by rigorous evidence. Traditionally, intravenous PHT has been the most commonly used 2nd line treatment for SE. Fosphenytoin (FOS) has been preferred over PHT, due to the ability for faster infusion, less risk of cardiac arrhythmias, and decreased incidence of local tissue reaction. However, a meta-analysis of 22 studies has found the efficacy of PHT to be below, though not significantly, to that of valproate (VPA), PHB, and levetiracetam (LEV) [39]. This meta-analysis also concluded that there is insufficient data for the use of lacosamide (LAC) in the treatment of SE. Recently, two different open-label trials: ConSEPT study from Australia and New Zealand, and EcLiPSE study from United Kingdom, have compared LEV and PHT in children with BDZ-refractory SE [40,41]. Clinical cessation of seizure activity 5 min after infusions were not statistically different between the two groups. Overall, 50%-70% of participants achieved primary efficacy endpoint. An ongoing trial in US, Established Status Epilepticus Treatment Trial (ESETT), is also comparing LEV, FOS, and VPA [42]. Given the lack of class I trials, decisions about 2nd line treatment are often based on local availability, cost, and most importantly patient-specific factors.

In addition to medical stabilization, the treatment of early/established SE proceeds concurrently with attempts to ascertain the etiology of SE. The specific treatments are driven by presumptive or definite etiology, often determined by local epidemiology, and investigative resources. A few suggestions are offered in Table II.

Table II Suggested Etiology-Specific Treatments in Status Epilepticus

Third-line Treatment: Drug Therapy for RSE

There is a further paucity of clinical trials for drug treatment of RSE. Designing comparative effectiveness trials in this population has been challenging due to several reasons, including heterogeneity of 1st and 2nd line treatment choices, concurrent use of multiple other medications affecting the nervous system, variability in the time course of administration of different medications, and the level of supportive intensive care. This is exemplified by a multicenter trial which was terminated inconclusively, because only 14 patients could be recruited over 3 years, against an estimated sample size of 150 [43]. Another trial of brexanolone (an aqueous formulation of the neuroactive steroid allopregnanolone) in 25 patients with SRSE was also inconclusive regarding efficacy end point, though it showed reasonable tolerability [44].

Conventionally, the three most commonly used anesthetic agents for the treatment of RSE include MDZ, short-acting barbiturates (pentobarbital/thiopentone), and propofol (Table III). Currently, MDZ is perhaps the most commonly used drug for RSE due to faster onset of action and short duration of effect [45]. However, use of MDZ in RSE/SRSE is fraught with several issues including development of tolerance, prolonged half-life with continued use, and potential for interactions with other drugs, nephrotoxicity, hepatotoxicity, and cardio-respiratory depression [45,46]. Several studies have reported on the use of MDZ for RSE/SRSE, using different doses and treatment targets, as reviewed elsewhere [45]. A meta-analysis including 111 children showed that MDZ was as effective as other coma inducing medications, but had a lower mortality [47]. However, another study that compared MDZ and DZP in 40 children and found a similar efficacy (86% and 89%, respectively), reported MDZ to be associated with a higher recurrence (57% vs 16%) and higher mortality (38% vs 10.5%) [48]. Seizure control has been reported as occurring within 0.3-1.1 hours [45]. Breakthrough seizures have been reported in 47-57% of patients [49].

Table III Continuous Infusions for Treatment of Refractory Status Epilepticus

To compare, barbiturates have been used for the longest period, and believed to have higher efficacy. A meta-analysis including 193 adults with RSE compared pentobarbital, MDZ, and propofol. Although pentobarbital was associated with a significantly lower incidence of short-term treatment failure, breakthrough seizures, and the need to change to a different medication, it was also associated with a significantly higher frequency of hypotension [50]. One of the important confounding factors in this study was that the target of treatment in patients receiving pentobarbital tended to be burst-suppression, while it was limited to seizure suppression in those treated with MDZ and propofol. In pediatric RSE/SRSE the efficacy of pentobarbital has been reported to be 74-100% in heterogeneous studies [47,51]. Some experts believe that barbiturates have a neuroprotective effect and have an additional anti-seizure efficacy from their ability to lower core body temperature [52]. However, barbiturates have more potent cardiorespiratory suppression, immunosuppression, longer half-life with storage in the lipid compartment resulting in slow recovery, and problems of auto-induction and zero-order kinetics. Intravenous formulations of pentobarbital also contain propylene glycol, which at the high cumulative doses necessary to achieve burst suppression, may cause hyperosmolality and lactic acidosis that can progress to cardiac arrhythmias, refractory hypotension, renal dysfunction, and multi-organ failure [53]. It is possible that advances in intensive care, and the increasing practice for pre-emptive control of ventilation and perfusion, may help mitigate some of these challenges.

In adults, propofol is used for management of RSE/SRSE due to its quick onset of action and prompt recovery on withdrawal. Studies in adults have shown that propofol infusion terminates RSE/SRSE in 67% of patients [54]. Although propofol induces burst-suppression within 35 minutes of initiation, frequent titration may be needed to maintain adequate suppression [55]. A retrospective study of 33 children aged 4 months to 15 years with RSE reported that propofol was more effective than thiopental in terminating seizures (64% vs 55%). However, adverse effects including rhabdomyolysis and hypertrigly-ceridemia, prompted discontinuation in 18% of patients, with recovery after discontinuation of propofol [56]. Use of propofol has been restricted in children due to the risk of propofol infusion syndrome (cardiac failure, rhabdo-myolysis, metabolic acidosis, renal failure, and sometimes death) and propensity to induce dyskinesias which can mimic breakthrough seizures [57]. Risk factors reported to be associated with a higher likelihood of the infusion syndrome include high doses, prolonged use, concurrent use of catecholamines and corticosteroids, and possibly malnutrition. However, there have been some concerns that this syndrome may not be limited to propofol, and could in fact, result from drug-induced cerebral suppression [58]. Also relevant to pediatric practice are the reported fatalities on concurrent use of propofol and ketogenic diet (KD) [59].

Ketamine

Acceptance for the use of ketamine in RSE/SRSE has been steadily increasing due to favorable hemodynamics and a different mechanism of action than conventional anesthetics. Given the GABA receptor endocytosis in RSE/SRSE, and the NMDA glutamate receptor antagonism of ketamine, there is a theoretical advantage for its use instead of, or more commonly in addition to, other anesthetic agents. A systematic review of 25 class IV studies analyzed 244 SE episodes (37 in children) treated with ketamine [60]. Although authors reported 73% of children to ‘respond’, the heterogeneity in the definition of response was not addressed. Our practice is to start with MDZ infusion, and optimize it to achieve seizure suppression and >50% burst suppression. In non-responders, we then consider ketamine concurrent to MDZ. In patients who do not respond, we finally start pentobarbital and wean off MDZ and ketamine (Fig. 1). This is different from published multicenter experience from US, where ketamine was typically used after pentobarbital [61]. Salient pharmacological aspects of these drugs are summarized in Table III.
 
Therapeutic Target for RSE

As we have already alluded, there is a lack of consensus regarding optimal target for treatment of RSE/SRSE. Different therapeutic endpoints have included seizure suppression, burst suppression, or electro-cerebral silence. Moreover, the extent of burst-suppression that is associated with best outcomes, remains undefined [62]. This results in variability in clinical practice as shown by a study including 35 adult RSE patients, which reported that patients remained within the target suppression range (defined as 65-95%) only 0-29% (median 8%) of the total time under treatment [63]. Our practice is to increase infusions rapidly to achieve a suppression ratio of 50-70% and complete seizure control, as soon as possible after failure of 2nd line treatment, and to frequently review EEG to ensure that this degree of suppression is maintained. We try to wean off anesthetic infusions once this treatment target is maintained for 24-48 hours. However, this duration is not based on rigorous evidence, and a shorter duration of burst suppression may be beneficial in some patients, whereas several cycles may be needed in patients with SRSE. Longer pharmacological coma of up to 5-7 days may have to be tolerated in particularly challenging cases. Simultaneously, additional ASMs preferably with complimentary mechanisms of action, short half-life, and low incidence of drug inter-actions, should be added to facilitate ongoing seizure suppression and smooth weaning of infusions. Topira-mate, LEV, LAC, and more recently, brivaracetam and perampanel have been used for this purpose [26].

The effectiveness of this empirical practice was recently assessed in a cohort of 111 RSE patients recruited prospectively over two years in an observational study. MDZ was the most frequently used initial anesthetic agent (78%), and pentobarbital was most frequently used agent after MDZ failure (82%) [64]. Treating physicians used up to four cycles of serial anesthetic therapy in these patients, and seizure termination was achieved in 94% patients by the second cycle. However, other studies have shown regional differences in the use of therapeutic coma for RSE treatment, lack of effect on overall mortality, and increased length of hospital stay, related costs, and adverse effects [65,66].