Management of Status Epilepticus in Adults
Currently in the US, phenytoin, valproic acid, phenobarbital, levetiracetam, lacosamide, diazepam, lorazepam are available in IV formulations. In February 2016, the FDA also approved brivaracetam (which also is available in an IV formulation) and in October of the same year IV carbamazepine. None of these AEDs has an FDA indication for SE, although they are widely used. Parenteral lacosamide has a success rate of 33% to 67.7% (200–400 mg over 3–5 min was the most common bolus dose) depending on its use as second or third AED [94–96]. In lacosamide-naive patients with RSE on continuous EEG monitoring, the success rate for cessation of SE was 15.7, 25.5, 58.8, and 82.4 % by 4, 12, 24, and 48 hours, respectively [97]. Alternatively, topiramate in doses 300–1600 mg/day per oro/nasogastric tube can be considered [98]. In a study of 35 patients with RSE treated with topiramate as an adjunct AED, the response rate was 86% (as the third AED), and remained stable at 67% after administration as the fourth to seventh AED. Overall, RSE was terminated in 71% of patients within 72 hours after first administration of topiramate [99]. Other studies, however, adjusting for co-variates, did not prove topiramate to be effective in RSE [100]. Clobazam, a unique oral 1,5-benzodiazepine with excellent absorption, has been also used in the treatment of RSE. Seventeen patients with RSE (11 with prior epilepsy) were successfully treated with clobazam, which was introduced after a median duration of 4 days and after a median of 3 failed AEDs. Termination of RSE within 24 hours of administration, without addition or modification of concurrent AED and with successful wean of anesthetic infusions, was seen in 13 patients, whereas indeterminate response was seen in another 3. Clobazam was deemed unsuccessful in 1 patient [101]. In another recent report of 70 episodes of RSE, clobazam was used in 24 (34.3%) of them. If clobazam was the last AED added to therapy before RSE termination, the success was attributed to this drug. Based on this definition, clobazam led to 6 episodes (25%) of successful RSE resolution [102]. If primary or metastatic brain tumor is the presumed cause of SE, a combination of IV phenytoin, IV levetiracetam (median dose 3 g/d) and enterically administered pregabalin (median dose 375 mg/day) led to 70% control of SE on average 24 hours after addition of the third AED [103]. However, the major treatment options, which should not be delayed in unresponsive RSE, are propofol or midazolam infusions at high rates and under continuous EEG monitoring. These infusions should be continued for at least 24 hours and then held to reassess the situation. By that time, cocurrent metabolic derangements and low AED levels from noncompliance should have been corrected. Prolonged and high-dose propofol should be avoided because of the risk for propofol infusion syndrome, especially if pressors/inotropes are co-infused [104].
Super-refractory SE
Should seizures continue or recur, stage 4 options for SRSE are considered [105]. Pentobarbital with shorter half-life is favored to phenobarbital. The main disadvantages of barbiturates are compromised neuro-exam (which has to be assessed frequently), cardiovascular depression and hypotension, respiratory depression with need for full ventilator support, cough suppression with increased risk for atelectasis and pneumonia, immunosuppression increasing the risk for infection or sepsis, immobility increasing the risk for thromboembolism and ileus mandating parenteral nutrition [106,107]. The depth and duration of the EEG suppression that must be achieved by barbiturates is unknown. Some experts recommend instead of burst-suppression pattern complete suppression or “flat record” because of better seizure control and fewer relapses [108]. Moreover, patients with more prolonged barbiturate treatment (> 96 hours) and those receiving phenobarbital at the time of pentobarbital taper are less likely to relapse [109]. European guidelines recommend titration of propofol and barbiturate to EEG burst-suppression, and midazolam to seizure suppression, maintained for at least 24 h [2]. In recent reviews, it was found that barbiturates control refractory and super-refractory SE in 64% of patients and are ineffective in only 5% [11,110].
If SE continues or recurs after emergence from barbiturate coma, ketamine may be an option [11,83]. Ketamine offers the advantage of NMDA receptor antagonism, which may be important in the late phase of SE and lacks cardiodepressant or hypotensive properties. Early [111] or late [112] use of ketamine has been reported in small case series with various success rates. In a recent multicenter retrospective study from North America and Europe, evaluating 58 patients with 60 RSE, ketamine was likely responsible for seizure control in 12% and possibly responsible in an additional 20%. No responses were observed when infusion rate was lower than 0.9 mg/kg/h or when ketamine was introduced 8 days or more after onset of SE or after failure of seven or more drugs [113].
If all these measures have failed, stage 4.2 treatment options are available (Table 2), but these are mostly based on small case series and expert opinions (except for the recent hypothermia study). Pyridoxine hydrochloride in an IV or enteral form at a dose of 100–300 mg/day for few days can be used in stage 4 or earlier stages, as it is a cofactor in the synthesis of the inhibitory neurotransmitter GABA [114]. There are no strong data for its effectiveness, but it can be used as a cheap and safe alternative [115]. Magnesium has been successfully used in 2 girls with juvenile Alper’s syndrome [116] and is the treatment of choice for eclamptic seizures. Ketogenic diet may also be an optionfor these patients [117]. Resection of the epileptic focus after mapping with intracranial EEG electrodes may be highly effective but cannot be used in many patients due to lack of focus or eloquence location [83,106,115]. Use of steroids, plasmapheresis or IVIG, followed by immunosuppression can be tried, but one should balance risks and benefits. These immunosuppressive or immunomodulating treatments should be especially considered in patients with NORSE or suspected autoimmune or paraneoplastic encephalitides, where AEDs usually have no effect [46]. These therapies though often precede the diagnosis, since it takes time for the autoantibody panel results to return and the treating physician has to make a decision to blindly start treatment for SRSE.
There were some promising data regarding hypothermia use in these desperate situations [118,119] until the HYBERNATUS study, conducted in France, was recently published. In this study, 270 patients with convulsive SE were randomized in to hypothermia (32° to 34°C for 24 hours) in addition to standard care or to standard care alone. A Glasgow Outcome Scale score of 5 (primary outcome) occurred in 49% of patients in the hypothermia group and in 43% in the control group (a nonstatistical difference). Secondary outcomes, including mortality at 90 days, RSE on day 1, SRSE and functional sequelae on day 90 were not different except for the rate of progression to EEG-confirmed SE on the first day, which was lower in the hypothermia group (11% vs. 22% in the controls). Adverse events were more frequent in the hypothermia group than in the control group [120].
Additional anecdotal treatments are presented in Table 2, but their efficacy is questionable.
This staged management approach may change in the future to a more physiologic and rational treatment with polytherapy based on synaptic receptor trafficking during SE [63]. For example, in an animal model of severe SE, combinations of a benzodiazepine with ketamine and valproate, or with ketamine and brivaracetam, were more effective and less toxic than benzodiazepine monotherapy [121]. Allopregnalonone, a metabolite of progesterone, is an endogenous, naturally occurring neuroactive steroid produced in the ovary, the adrenal cortex and the central nervous system. It is a potent positive allosteric modulator of synaptic and extrasynaptic GABAA receptors with antiepileptic activity [122]. Neuroactive steroids, such as allopregnanolone, are currently evaluated in SE.
Outcomes
SE still carries significant mortality and morbidity. Distinct variants of SE carry different mortalities, and the range is quite broad: from zero mortality for absence or complex partial SE in ambulatory patients [12], to 19% to 27% 30-day mortality for generalized tonic-clonic SE [20,123] and to 64.7% 30-day mortality for subtle SE [20]. Variables playing an important role in the outcome are the underlying cause (regarded by most authorities the most important variable), the duration of SE (mortality 32% if persistent for > 1 hour vs 2.7% if < 1 hour), the type of SE, the treatment administered, and the age of the patient (children have better outcomes than adults) [123–125]. The etiology of SE still remains the most important prognostic factor, with alcohol and AED-withdrawal/noncompliance having the best outcomes; structural brain injuries, such as anoxia-ischemia, vascular lesions, or brain tumors, have the worst prognosis.
The most resistant cases pose significant dilemmas regarding the length of treatment using barbiturate coma and the potential for acceptable prognosis or the need to withdraw life support. For RSE, for example, in-hospital mortality is 31.7% and 76.2% of patients reach poor functional outcome. Long-term outcomes are also dismal: at 1 year post-discharge, 74% are dead or in a state of unresponsive wakefulness, 16% severely disabled, and only 10% have no or minimal disability [126]. Duration of drug-induced coma, arrhythmias requiring intervention, and pneumonia are associated with poor functional outcome, whereas prolonged mechanical ventilation with mortality and seizure control without burst-suppression or isoelectric EEG are associated with good functional outcome [127,128].
Treatment with barbiturates may contribute to these outcomes, although it is very challenging to prove causality in such a complex and prolonged ICU environment. Some data have shed light towards that direction: in a recent retrospective study of 171 patients with SE, of which 37% were treated with IV anesthetic drugs, there was a higher risk for infections and a 2.9-fold relative risk for death after adjustment for confounders in the group treated with IV anesthetics compared to the group without these agents [129].
The SE Severity Score (STESS, range 0–6) is a prognostic score for survival [130] and can be used as a scaffold for discussions with families and covariate adjustment tool for research. A favorable score of 0–2 has a negative predictive value of 0.97 for survival and likelihood to return to baseline clinical condition in survivors, although an unfavorable score (3–6) had a positive predictive value for death of only 0.39 [131].
The risk for recurrence of afebrile SE in a population-based study in Minnesota has been estimated at 31.7% over a 10-year follow-up period. The risk for recurrence was about 25% regardless of the underlying etiology, except in those patients with SE occurring in the setting of a progressive brain disorder (who had a 100% risk). Female gender, generalized (vs partial) SE and lack of response to the first AED after the initial episode of SE were independent factors for recurrence [132].
Conclusion
Despite the use of better diagnostic tools (continuous video EEG), advances in technology in the ICU, and availability of new AEDs, SE still carries significant mortality and morbidity, which depends mainly on age and etiology. The current treatment is still staged, with supportive measures and benzodiazepine administration remaining the mainstay initially and followed by older and newer AEDs and anesthetics for resistant cases. With the advance of pathophysiologic mechanisms elucidation at a molecular/receptor level, combinations of AEDs may become the foundation of future SE control.
Corresponding author: Panayiotis N. Varelas, MD, PhD, FNCS, Division Head, Neuro-Critical Care Service, Henry Ford Hospital, K-11, 2799 W. Grand Blvd., Detroit, MI 48202, pvarela1@hfhs.org.
Financial disclosures: Dr. Varelas was local principal investigator for a super-refractory status epilepticus study sponsored by Sage Therapeutics.
Author contributions: conception and design, ARR, PNV; analysis and interpretation of data, PNV; drafting of article, PNV; critical revision of the article, ARR, PNV; administrative or technical support, PNV; collection and assembly of data, ARR, PNV.
