Acute seizures are common and are defined as a transient occurrence of signs and/or symptoms due to abnormal excessive or synchronous neuronal activity in the brain.¹ Intrinsically, the brain has mechanisms in place to terminate excessive electrical activity. The mean duration of a secondarily generalized tonic-clonic (GTC) seizure is 53 to 62 seconds, and rarely lasts longer than 2 minutes.^2,3 However, some seizures do not stop and progress to status epilepticus (SE), which may be convulsive (CSE), with clinically apparent motor (clonic) rhythmic jerking and/or (tonic) stiffening, or nonconvulsive (NCSE), with seizure activity on electroencephalography (EEG), and subtle or no obvious clinical signs. Status epilepticus is a neurological emergency often requiring management in the intensive care unit (ICU) for causes or complications of SE, or both.

In 2015, the International League Against Epilepsy (ILAE) proposed a conceptual definition that applies to all types of SE: (1) SE starts as a condition resulting from failure of seizure-termination mechanisms or the initiation of pathological mechanisms that likely lead to continuous seizure activity, and (2) SE creates long-term consequences that begin to occur after the onset of SE, including neuronal death, neuronal injury, and alteration of neuronal networks. This definition hinges on the identification of the semiology of SE: the clinical manifestations of seizure activity (Table 16-1). Specifically for generalized CSE, criterion 1 is defined when seizures last longer than 5 minutes and criterion 2 occurs at the point that long-term consequences begin to appear, around 30 minutes.¹ Convulsive SE is also defined as recurrent seizures between which there is incomplete recovery of consciousness.⁴

The point at which focal seizures or nonconvulsive seizures become SE (criterion 1) and create long-term consequences (criterion 2) is less clear. Current proposed definitions suggest that focal CSE with impairment of consciousness is defined at 10 minutes with long-term injury developing at greater than 60 minutes.¹ In contrast, the diagnosis of NCSE relies on EEG. The Salzburg criteria is a unified, validated set of rules to define NCSE on EEG with a diagnostic accuracy of 92.5%.^5-7 An NCSE is defined on EEG as epileptiform discharges at a periodicity of greater than 2.5 Hz, or if discharges are slower, a clear evolution of the pattern over time or space, clinical or electrographic improvement with antiseizure drugs (ASDs), or subtle convulsive movements. The duration of time to fulfill criterion 1 is typically considered 10 minutes of continuous ictal activity or, for intermittent nonconvulsive seizures, more than 50% of a 1-hour EEG recording.⁶

Status epilepticus is associated with high mortality and morbidity and imposes a high financial burden on society. The annual incidence of SE is 10 to 41 per 100,000 people^8-13 and exhibits a U-shaped distribution across years of life, peaking both under 10 years of age and over 50 years of age.¹⁴ The annual costs of SE are estimated to be $4 billion.¹⁵ The overall mortality rate is approximately 20%.^14,16 Scoring systems have been devised to estimate outcome, including the Epidemiology-Based Mortality Score in Status Epilepticus and the Status Epilepticus Severity Score (STESS; Table 16-2).^17,18 The STESS is a validated scoring system,^19-22 with an overall sensitivity of 94% and a negative predictive value of 97%. It includes predictors of poor outcome, such as older age (> 65 years), impairment of consciousness, NCSE, and de novo onset of SE as underlying etiology. Other factors that have been reported to be associated with poor outcome include focal neurological signs, seizure duration, use of anesthetics for seizure control, injury severity scores such as the Acute Physiologic Assessment and Chronic Health Evaluation (APACHE), and other medical complications.^23-27

Early diagnosis and urgent high-quality treatment are essential to reduce the morbidity and mortality associated with prolonged status epilepticus, and to maximize the efficacy of medication treatment.²⁸

Status epilepticus itself is associated with multiple systemic complications (Table 16-3) and prognosis worsens with duration of time from seizure onset to treatment.^24,29-32 In humans, seizure activity lasting greater than 30 minutes is associated with significantly greater mortality than seizures lasting from 1 to 29 minutes (19% vs 2.6%).²⁴ Neuronal loss is observed after 40 minutes of seizure activity in animal models.³³ This early timeframe also affects treatment success as time-dependent pharmaco-resistance and self-perpetuation of SE occur after 15 to 30 minutes of seizure activity.³⁴ In fact, the duration of SE prior to treatment is one of the most important determinants of successful medical control of SE.³⁵ This is likely due to maladaptive changes that take place after sufficient stimulatory activity initiates SE. Within seconds to minutes, receptor trafficking, specifically internalization of synaptic γ-aminobutyric acid A (GABA_A) receptors and synaptic expression of N-methyl-D-aspartate (NMDA) and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors occurs.^36-38 Further changes in inhibitory and excitatory neuropeptides occur within minutes to hours and later gene expression alteration maintains this abnormal electrical circuitry.³⁹ As SE proceeds and become self-sustaining, GABAergic drugs like benzodiazepines and barbiturates lose effectiveness in time- and dose-dependent manners, while NMDA antagonists are usually more effective even late in the course of status epilepticus.^40,41 In an animal model, diazepam readily stopped seizure when given 10 minutes after SE onset, while its potency decreased by 20 times when given 30 minutes after SE. Phenytoin (PHT) also showed a similar time-dependent relationship.^42,43 The pathophysiology also illustrates the importance of appropriate drug choice and dosing. Inadequate drug dosing and/or route of administration significantly contribute to ineffective termination of SE and even mortality.^44,45 Overall quality of treatment is crucial for SE control.

Given the importance of timely and appropriate treatment and the high mortality associated with generalized CSE, guidelines propose algorithms for management. The most recent one, by the American Epilepsy Society proposes a 3-phase treatment: (1) a “stabilization phase” that should occur within 5 minutes of seizure onset and includes initial first aid and assessments; (2) an “initial therapy phase” that should occur in less than 20 minutes of onset and includes appropriate medical intervention; (3) a “second therapy phase” (20–40 minutes of seizure activity) when response to initial therapy should be apparent and a second-line agent should be administered, usually an intravenous (IV) formulation for rapid bioavailability; and (4) a “third therapy phase” (greater than 40 minutes of seizure activity), for which there is no clear guidance on treatment and includes either an anesthetic or another second-line therapy agent. The guideline also found strong evidence that the second therapy is often less effective than initial therapy.⁴⁶

Such timelines serve as guidance. Throughout management, the provider should be astutely monitoring the patient, anticipating the next step, and ready to quickly treat seizures (ie, having medication readily available). Effective seizure termination by an antiseizure drug is usually defined as cessation of electrical and/or clinical seizure activity within 20 minutes from time of administration without recurrence within 60 minutes.⁴⁷

Early Status Epilepticus: Primary Evaluation, Initial Management, and Therapy

The goals of initial SE management are to stop the seizures emergently, and to screen and treat for potentially life-threatening underlying causes of SE. These steps should take place quickly, whether in the prehospital, emergency department (ED), or ICU setting. As with any medical emergency, management begins by evaluating the patient’s airway, breathing, circulation, and IV access. A brief neurological assessment should focus on the patient’s mental status and any focal neurological deficits. Intubation for SE should be based on this clinical assessment. Laboratory studies should be sent concurrently, and a fingerstick glucose should be obtained (Table 16-4).

Benzodiazepines have Level A recommendation as first-line agent for SE.^4,46 An early in-hospital, randomized, double-blind control study compared lorazepam, diazepam plus phenytoin, phenytoin, and phenobarbital (PHB) as first-line treatment for status epilepticus and showed that lorazepam was superior.⁴⁷ Subsequently, 2 prehospital studies confirmed the role of benzodiazepine in the initial management of SE. In the first, lorazepam and diazepam aborted seizures in 59% and 42%, respectively, compared to 21% by placebo.⁹ In the second, intramuscular (IM) midazolam aborted seizures in 73.4% compared to 63.4% in the intravenous lorazepam group; midazolam was faster and statistically noninferior.⁴⁸ Lorazepam is recommended IV at a dose of 2 mg for children and adults less than 40 kg or 4 mg for adults more than 40 kg.^48,49 Intravenous lorazepam and IV diazepam have no significant difference (Level of Evidence [LOE] A).⁴⁶ Intravenous diazepam can be administered at a dose of 0.2 to 0.3 mg/kg for children or adults less than 40 kg or 10 mg for adults more than 40 kg. If IV access is not available, IM midazolam is recommended over IV lorazepam (LOE A) as a 5-mg dose for children and adults less than 40 kg or 10 mg for adults more than 40 kg. Rectal diazepam 15 to 20 mg is an alternative if IM midazolam is not immediately available; it similarly reduces the risk of progression to established SE compared to placebo (risk ratio [RR], 0.43). Benzodiazepines are associated with hypotension and respiratory depression, leading to the underdosing of these critical medications. However, in 1 randomized controlled trial, administration of lorazepam 4 mg IV led to a lower rate of complications (eg, hypotension, cardiac arrhythmia, respiratory depression requiring bag-valve mask or attempt at intubation) compared to patients treated with placebo (10% vs 22%; LOE A).^9,46

After seizures stop, further diagnostic studies may be indicated to identify the underlying etiology of the SE or its modifying factors. Up to two-thirds of SE identified in the ED occurs in patients with a history of epilepsy, and half have issues with their home medications.⁵⁰ Status epilepticus in patients who are hospitalized, on the other hand, usually results from an acute symptomatic cause, meaning that SE is provoked by either brain injury or a systemic illness occurring within 7 days of onset. Examples of acute symptomatic SE include stroke, traumatic brain injury, or hypoxic ischemic injury. Overall, acute symptomatic causes account for 48% to 63% of all hospitalized SE cases.^10,11,14 Although stroke is the most common cause of SE in the adult population, a growing proportion of SE is recognized as having an immune-mediated cause and manifestations may be subacute, or less clearly defined.⁵¹

Established Status Epilepticus: Starting a Second-Line Agent

Benzodiazepines fail to control SE in 35% to 45% of patients, which defines established SE.⁴⁷ If seizures persist or recur, a second-line agent should be administered immediately. Second-line ASDs are typically used to maintain seizure control if they are effective at terminating SE. Table 16-5 lists commonly used ASDs, including phenytoin/fosphenytoin (fPHT), valproic acid (VPA), levetiracetam (LEV), brivaracetam (BRV), lacosamide (LCS), and phenobarbital (PHB). Currently there is no high-quality data to support the use of 1 agent over another. Phenytoin/fPHT, VPA, and LEV are the most frequently used and recommended by current guidelines.^4,46 The most recent American Epilepsy Society guidelines include the following: “There is no difference in efficacy between IV lorazepam followed by IV phenytoin, IV diazepam plus phenytoin followed by IV lorazepam, and IV phenobarbital followed by IV phenytoin (Level A). Intravenous valproic acid has similar efficacy to IV phenytoin or continuous IV diazepam as second therapy after failure of a benzodiazepine (Level C). Insufficient data exist in adults about the efficacy of levetiracetam as either initial or second therapy (Level U).⁴⁶ A single-center prospective randomized control pilot study of 150 patients compared PHT, VPA, and LEV following lorazepam and showed that the 3 agents are safe and equally effective, controlling seizures in 71% overall.⁵² A multicenter randomized, controlled, blinded study comparing the effectiveness of fPHT, VPA, and LEV in established SE is currently being conducted.⁵³ Newer agents such as LCS and BRV will need to be compared in the future.

Phenytoin/Fosphenytoin

Historically, PHT has been used for SE since the 1960s.⁷⁰ Fosphenytoin, its water-soluble prodrug, can be administered at a faster rate and with fewer side effects (notably subcutaneous tissue injury and pain, and cardiovascular effects) than standard PHT, which is administered with propylene glycol. Fosphenytoin requires approximately 15 minutes to undergo conversion to its active form; therefore, the overall timing of efficacy is similar. As a general rule, sodium channel blockers such as PHT, carbamazepine, and oxcarbazepine are effective for focal seizures; however, they can exacerbate primarily generalized seizures. Despite this, fPHT is currently recommended over PHT as second-line ASD for established SE. In patients with a history of primary generalized epilepsy, VPA is preferred.⁴

Valproic Acid

Valproic acid, or valproate, is a safe and well-tolerated drug with a lower risk of cardiovascular side effects than PHT, even in unstable or elderly patients.^71,72 A randomized controlled study of 100 patients showed both VPA and PHT to be effective after diazepam failure in controlling SE (84% and 88%, respectively).²⁴ A meta-analysis and systematic review confirmed similar efficacy of VPA and PHT.^49,73 Valproic acid is oxidized in hepatic mitochondria in addition to being glucuronidated and metabolized by the cytochrome P450 system. Valproic acid should be avoided in patients with mitochondrial or hepatobiliary disease.

Levetiracetam/Brivaracetam

Levetiracetam is commonly used due to minimal protein binding, nearly 100% bioavailability, no known drug-drug interactions, and renal rather than hepatic metabolism. No significant cardiorespiratory side effects have been noted with IV loading doses up to 4000 mg.⁵⁷ In a small multicenter retrospective study of 40 patients, early treatment appeared to be more effective than late (78% vs 46%).⁷⁴ However, others have observed that LEV may be less effective at terminating SE than VPA or PHT.⁵⁹

Brivaracetam, a recently approved medication similar to LEV, has a 20-fold higher affinity for the synaptic vesicle protein 2A (SV2A) ligand compared with LEV, and is now available in IV formulation.⁷⁵ Highly lipophilic, experimental data shows that it enters and acts faster in the brain than LEV.^69,76 Brivaracetam can be safely administered as 100 mg IV 2-min bolus or 15-minute infusion. Clinically insignificant electrocardiography (ECG) changes (sinus bradycardia, first degree AV block) have been observed with bolus therapy.⁶⁹ Brivaracetam is both hydrolyzed and metabolized by the cytochrome P450 system, and therefore, dosage adjustments are required for those with hepatic disease. Brivaracetam has not been studied in SE to date.

Lacosamide

Lacosamide is a novel agent that acts on sodium channels in a distinct way by enhancing their slow inactivation. Lacosamide is similar to LEV in its lack of drug-drug interactions and clinically relevant cardiopulmonary side effects using infusions of up to 400 mg under 5 minutes.⁷⁷ A dose-dependent prolongation in the PR interval on ECG and atrial arrhythmias has been reported.^64,78 Therefore, caution should be used in patients with preexisting arrhythmias, conduction block, or those on dromotropic agents. Lacosamide has not been systematically compared to fPHT, VPA, or LEV, although anecdotally its efficacy is similar.

Refractory Status Epilepticus

If a patient continues to have seizures despite a load of a second-line agent, SE is considered refractory (RSE). Most patients with generalized RSE require intubation,⁷⁹ and the use of sedatives and paralytics often used for induction masks ongoing motor activity. Importantly, almost half of those with clinical control of generalized SE exhibit seizures on EEG, suggesting electromechanical dissociation of SE that favors the development of NCSE in this context. Patients with generalized SE who do not stop seizing or those who remain comatose despite a second-line agent require continuous IV anesthetic (cIV) agents, such as midazolam, propofol, or pentobarbital (Table 16-6). In 2 systematic reviews, no treatment was found to be superior to another.^80,81 Continuous IV anesthetic dosing is titrated to cessation of electrographic seizures, or in some cases burst suppression, based largely on clinical preference. Generally, 24 to 48 hours of seizure control is recommended prior to cIV anesthetic weaning.⁴