See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/23243330

Treatment of Status Epilepticus

Article  in  Seminars in Neurology · August 2008

DOI: 10.1055/s-2008-1079339 · Source: PubMed

CITATIONS READS

48 1,149

2 authors:

Hiba Arif Haider Lawrence James Hirsch

Emory University Yale University
46 PUBLICATIONS   1,488 CITATIONS    315 PUBLICATIONS   10,270 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Saving the survivors of cardiac arrest step 2: accurate neuroprognostication View project

Yale Research View project

All content following this page was uploaded by Lawrence James Hirsch on 05 June 2014.

The user has requested enhancement of the downloaded file.

 Treatment of Status Epilepticus
Hiba Arif, M.D.,1 and Lawrence J. Hirsch, M.D.1

ABSTRACT

Status epilepticus (SE) is a neurological emergency that requires prompt diagnosis

and treatment, as delay is associated with a higher likelihood of poor response to treatment
and worse outcome. Lorazepam has been well established as a first-line therapy. Sub-
sequent steps are less established, and there are many reasonable options, including
intravenous fosphenytoin, valproate, midazolam, propofol, and phenobarbital. If intra-
venous access is not immediately available, rectal diazepam or nasal or buccal midazolam
should be given; this can also be used as out-of-hospital treatment to prevent or treat SE.

Downloaded by: University of Alberta. Copyrighted material.

For refractory SE, continuous intravenous midazolam and propofol, separately or in
combination, are rapidly effective, with pentobarbital remaining the gold standard for
prolonged cases. If a patient does not awaken after treatment, urgent electroencephalogram
(EEG) should be obtained to rule out nonconvulsive seizure activity. In refractory SE,
continuous EEG monitoring is required to recognize recurrence of seizure activity, as most
seizures will be nonconvulsive.

KEYWORDS: Status epilepticus, seizure, nonconvulsive status epilepticus, epilepsy,

continuous EEG monitoring

S tatus epilepticus (SE) is a medical and neuro- PRESENTATION AND DIAGNOSIS

logical emergency. Overall, mortality is
17 to Several types of SE exist. Clinically, the most im-
26%.1–3 An additional 10 to 23% of patients who portant distinction to make is between convulsive
survive SE are left with new or disabling neurological and nonconvulsive SE (NCSE), based on whether
deficits.2,4 or not rhythmic jerking of the extremities is observed.
Typically, patients who present with generalized
convulsive SE (GCSE) are expected to awaken grad-
DEFINITION ually after the motor features of seizures disappear. If
Traditionally, SE is defined as continuous or repeti- the level of consciousness does not improve by
tive seizure activity persisting for at least 30 minutes 20 minutes after cessation of movements, or the
without recovery of consciousness between attacks. mental status remains abnormal 30 to 60 minutes
More recently, authors have suggested that seizures after the convulsions cease, NCSE must be considered
exceeding 5 to 11 minutes should be considered and urgent electroencephalogram (EEG) is advised.
SE.5,6 For all practical purposes, a patient should be In a study by DeLorenzo and colleagues,7 14% of
considered to be in SE if a seizure persists for more patients treated successfully for convulsive SE were in
than 5 minutes, as very few single seizures will last NCSE when EEG was begun; of the patients who
this long. underwent continuous EEG monitoring (cEEG) after

1
Comprehensive Epilepsy Center, Columbia University, New York, [email protected]).
New York. Epilepsy; Guest Editor, Andres M. Kanner, M.D.
Address for correspondence and reprint requests: Lawrence J. Semin Neurol 2008;28:342–354. Copyright # 2008 by Thieme
Hirsch, M.D., Associate Clinical Professor of Neurology, Compre- Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001,
hensive Epilepsy Center, Columbia University, Neurological Institute, USA. Tel: +1(212) 584-4662.
Box NI-135, 710 West 168th Street, New York, NY 10032 (e-mail: DOI 10.1055/s-2008-1079339. ISSN 0271-8235.
342
 TREATMENT OF STATUS EPILEPTICUS/ARIF, HIRSCH 343

convulsive SE was controlled, 48% had nonconvulsive Table 2 Criteria for Nonconvulsive Seizure*
seizures. Any pattern lasting at least 10 seconds satisfying any one of
There is no clear consensus on the further classi- the following three primary criteria:
fication of NCSE. Many authors have classified NCSE Primary criteria
into generalized or partial onset, and into simple and  Repetitive generalized or focal spikes, sharp waves,
complex, based on whether or not there is alteration of spike-and-slow wave, or sharp-and-slow wave complexes
consciousness. In clinical practice, however, it is often at  3/sec
not possible to differentiate between these classifications  Repetitive generalized or focal spikes, sharp waves,
even with the help of EEG. In the intensive care unit spike-and-slow wave, or sharp-and-slow wave complexes
(ICU), the majority of seizures are nonconvulsive and at < 3/sec and the secondary criterion
will be missed without cEEG (Case 1).8 Nonconvulsive  Sequential rhythmic, periodic, or quasiperiodic waves at
seizures are more common than previously recognized;  1/sec and unequivocal evolution in frequency
nonconvulsive seizures have been reported in 34% of (gradually increasing or decreasing by at least 1/sec; e.g.,
neurological ICU patients,9 16% of severe head trauma 2 to 3/sec), morphology, or location (gradual spread into or
patients,10 and 8% of comatose patients that have no out of a region involving at least two electrodes). Evolution
clinical evidence of seizures.11 Nonconvulsive seizures in amplitude alone is not sufficient. Change in sharpness
have also been reported in 18 to 28% of patients with without other change in morphology is not enough to
intracerebral hemorrhage (ICH)12,13; such posthemor- satisfy evolution in morphology.
rhagic seizures are associated with neurologic worsening Secondary criterion
on the National Institutes of Health Stroke Scale (14.8

Downloaded by: University of Alberta. Copyrighted material.

 Significant improvement in clinical state or appearance of
vs. 18.6, p < 0.05), and with an increase in midline shift previously absent normal EEG patterns (such as posterior
( þ 2.7 mm vs. – 2.4 mm, p < 0.03).13 In a series of 102 dominant ‘‘alpha’’ rhythm) temporally coupled to acute
ICH patients who underwent cEEG at our center, administration of a rapidly acting AED. Resolution of the
electrographic seizures were associated with an increase epileptiform discharges leaving diffuse slowing without
in ICH volume of 30% or more between admission and clinical improvement and without appearance of previously
24-hour follow-up computed tomographic (CT) scan absent normal EEG patterns would not satisfy the secondary
(33% vs. 15%; odds ratio 9.5; 95% confidence interval, criterion.
1.7 to 53.8).12 Moreover, the many possible presenta-
*Satisfying these criteria is adequate for confirming nonconvulsive
tions of NCSE outside of the ICU setting (Table 1) may seizure activity. However, failing to meet these criteria does not rule
also be a cause of delay in diagnosis. Since an increasing out nonconvulsive seizure activity; clinical judgment and correlation
are required in this situation.
duration of seizure activity is associated with worse Reproduced with permission from Chong DJ, Hirsch LJ. Which EEG
outcomes,14 it is crucial to recognize and treat NCSE patterns warrant treatment in the critically ill? Reviewing the
evidence for treatment of periodic epileptiform discharges and
early. related patterns. J Clin Neurophysiol 2005;22:79–91, who modified
the criteria of Young GB, Jordan KG, Doig GS. An assessment of
nonconvulsive seizures in the intensive care unit using continuous
Table 1 Semiological Spectrum of Nonconvulsive EEG monitoring: an investigation of variables associated with
Seizures and Nonconvulsive Status Epilepticus mortality. Neurology 1996;47:83–89.
AED, antiepileptic drug.
Negative Symptoms Positive Symptoms
See Table 2 for the criteria for diagnosing def-
Anorexia Agitation/aggression
inite nonconvulsive seizures. A suggested method for
Aphasia/mutism Automatisms
performing a benzodiazepine trial to diagnose NCSE is
Amnesia Blinking
presented in Table 3. A common mistake is to give too
Catatonia Crying
high a dose of benzodiazepine; this results in marked
Coma Delirium
sedation, which has no diagnostic utility even if the
Confusion Delusions
EEG pattern rapidly resolves. Small incremental
Lethargy Echolalia
doses may allow observation of a clinical improvement
Staring Facial twitching
without marked sedation.
Laughter
Nausea/vomiting
Nystagmus/eye deviation
TREATMENT
Perseveration
Rapidity of treatment is key in the treatment of SE.
Psychosis
There are animal and human data to suggest that
Tremulousness
therapeutic interventions are most effective when
Reprinted from Jirsch J, Hirsch LJ. Nonconvulsive seizures:
developing a rational approach to the diagnosis and management
initiated early, and that efficacy decreases significantly
in the critically ill population. Clin Neurophysiol 2007;118:1660–1670, with increasing seizure duration.15,16 For all practical
with permission from Elsevier, who extracted (and expanded) from
Kaplan PW. Nonconvulsive status epilepticus in the emergency
purposes, treatment should be started after 5 minutes of
room. Epilepsia 1996;37:643–650. continuous seizure activity.
344 SEMINARS IN NEUROLOGY/VOLUME 28, NUMBER 3 2008

Table 3 Benzodiazepine Trial for the Diagnosis of the seizure threshold (Table 5) should be identified and
Nonconvulsive Status Epilepticus minimized.
Applies to: patients with rhythmic or periodic focal or
generalized epileptiform discharges on EEG with neurological
impairment Pharmacotherapy for SE
Monitoring: EEG, pulse oximetry, blood pressure, ECG, First-line medications control SE in 80% of patients
respiratory rate with dedicated nurse when initiated within 30 minutes, but in only 40% if
Antiepileptic drug trial: started after 2 hours of onset.15
 Sequential small doses of rapidly acting short-duration Benzodiazepines are the preferred initial therapy.
benzodiazepine such as midazolam at 1 mg/dose Most experts recommend IV lorazepam (0.1 mg/kg) as
 Between doses, repeated clinical and EEG assessment the first-line therapy. Patients who respond to first-line
 Trial is stopped after any of the following: agents will usually require maintenance therapy with a
 Persistent resolution of the EEG pattern (and exam repeated) second-line agent. Additional treatment must be pro-
 Definite clinical improvement vided quickly when patients continue to seize. The
 Respiratory depression, hypotension, or other adverse effect longer SE persists, the higher the risk for developing
 A maximum dose is reached (such as 0.2 mg/kg midazolam, refractory SE. There is a lack of prospective data for
though higher may be needed if on chronic second-line agents, but phenytoin (PHT) or fospheny-
benzodiazepines) toin is used most frequently.17
 Test is considered positive if there is resolution of the

Downloaded by: University of Alberta. Copyrighted material.

potentially ictal EEG pattern and either an improvement in
the clinical state or the appearance of previously absent Comparison of Initial Treatment Options of SE
normal EEG patterns (e.g., posterior dominant ‘‘alpha’’ Only a few prospective randomized trials have compared
rhythm) occurs; if EEG improves but patient does not, the initial treatment strategies for SE. The largest prospec-
result is equivocal tive study was the Veterans Affairs (VA) Status Epi-
Reprinted from Jirsch J, Hirsch LJ. Nonconvulsive seizures:
lepticus Cooperative study,18 which was a randomized,
developing a rational approach to the diagnosis and management double-blind, multicenter trial that compared four IV
in the critically ill population. Clin Neurophysiol 2007;118:1660–1670, treatments: lorazepam 0.1 mg/kg, diazepam 0.15 mg/kg
with permission from Elsevier.
EEG, electroencephalogram; ECG, electrocardiogram. plus PHT 18 mg/kg, phenobarbital 15 mg/kg, and PHT
18 mg/kg alone. In generalized convulsive SE, loraze-
pam was found to be most effective (65% for lorazepam
Initial Management alone vs. 58% for phenobarbital, 56% for diazepam plus
A sample protocol for the treatment of SE (as used in PHT, and 44% for PHT alone). The differences only
our center) is given in Table 4. Initial steps involve basic reached significance between lorazepam and PHT alone.
life support: administering oxygen, monitoring vital For subtle SE, no statistical difference was found be-
signs, and assessing and maintaining the airway. Intra- tween the groups and response was poor (only 8 to 24%
venous (IV) access should be established quickly, but responded to first treatment, vs. 44 to 65% for overt SE).
rectal, buccal, or nasal benzodiazepines should be given if All four treatment arms had similar complication rates.
there is any delay in obtaining IV access (see the No class I randomized controlled trials have been con-
following section on alternative modes of administra- ducted for second- or third-line therapy for SE.
tion). Laboratory studies should be sent (Table 4). A PHT or fosphenytoin is frequently recommended
fingerstick blood glucose level should be obtained and as a second-line agent. Seizure activity is often very
100 mg of thiamine and 50 mL of 50% glucose should be difficult to control once patients fail to respond to two
administered if glucose level is low or unknown. For AEDs. In the VA Cooperative Study,18 of the 38% of
seizures due to a metabolic derangement, correcting the patients with ‘‘overt’’ SE and the 82% of patients with
metabolic problem is often more effective than antiepi- ‘‘subtle’’ SE that continued to seize after receiving two
leptic drugs (AEDs). Fever, hypoxia, and hypotension AEDs, only 2% and 5%, respectively, stopped seizing
should also be treated concurrently, as these can exacer- after receiving a third agent. Once lorazepam failed, few
bate seizures and the associated neuronal injury. patients responded to PHT as a second-line agent
Other components of management include deter- (
5%); for this reason, some experts advocate progressing
mining whether there is a history of alcohol or drug use, directly to continuous IV (cIV) anesthetic drips once
previous epilepsy, possible acute neurological insult, lorazepam has failed (Table 4; also reviewed in Hirsch
obtaining a description of the seizure at onset if wit- and Arif19).
nessed, and brain imaging once the patient is stable and Although currently not approved by the Food and
seizures are controlled. Patients should not be pharma- Drug Administration (FDA) for use in SE, there is
cologically paralyzed unless cEEG is absolutely neces- growing evidence to support the efficacy of IV valproate
sary or being recorded. Use of medications that can lower (VPA) for SE. In a recent unblinded study from India,
 TREATMENT OF STATUS EPILEPTICUS/ARIF, HIRSCH 345

Table 4 Status Epilepticus Adult Treatment Protocol: Table 4 (Continued )

Columbia University Comprehensive Epilepsy Center, Time
2008 (min) Action
Time
(min) Action Begin EEG monitoring as soon as possible if patient
does not rapidly awaken, or if any cIV treatment
0–5 Diagnose; give O2; ABCs; obtain IV access; begin is used.
ECG monitoring; draw blood for chem-7,
*The IV solution of diazepam can be given rectally if Diastat is not
magnesium, calcium, phosphate, CBC, LFTs, available; the IV solution of midazolam can be given by any of these
AED levels, ABG, troponin; toxicology screen routes.
y
Prolonged use of propofol at > 5 mg/kg/h increases the risk of the
(urine and blood) propofol infusion syndrome; see text.
6–10 Thiamine 100 mg IV; 50 mL of D50 IV unless cIV, continuous intravenous; ABCs, stabilize airway, breathing, and
circulation; ECG, electrocardiogram; CBC, complete blood count;
adequate glucose known LFTs, liver function tests; AED, antiepileptic drug; ABG, arterial blood
Lorazepam 4 mg IV over 2 minutes; if still seizing, gases; PR, per rectum; IM, intramuscularly.
repeat  1 in 5 minutes
If no rapid IV access give diazepam 20 mg PR or Misra and colleagues randomized 68 patients with
midazolam 10 mg intranasally, buccally, or IM* untreated convulsive SE to IV VPA (30 mg/kg) or
10–20 If seizures persist, begin fosphenytoin 20 mg/kg PHT (18 mg/kg); seizures were aborted in 66% in the
IV at 150 mg/min, with blood pressure and ECG VPA group versus 42% in the PHT group. If the first
monitoring. This step can be skipped initially, agent failed, patients were given the other agent. As the
second agent, VPA was effective in 79% and PHT

Downloaded by: University of Alberta. Copyrighted material.

especially if proceeding to midazolam or propofol,
or performed simultaneously with the next step; was effective in 25%. Tolerability between the two groups
if done simultaneously, administration rate did not differ.20 Several other case series suggest good
should be slowed; using IV valproate instead of efficacy for IV VPA in the treatment of different types of
fosphenytoin is a justifiable option as well, particularly SE, including partial-onset, nonconvulsive, absence, and
in those allergic to phenytoin or with hypotension. myoclonic status.21–23 Another recent study randomized
10–60 If seizures persist, give one of the following four
Table 5 Commonly Used Drugs That Can Lower the
options (intubation often necessary except
Seizure Threshold
for valproate):
1. cIV midazolam: load: 0.2 mg/kg; repeat Antibiotics, especially in the elderly or those with renal
0.2–0.4 mg/kg boluses every 5 minutes until impairment
seizures stop, up to a maximum total loading Imipenem
dose of 2 mg/kg. Initial cIV rate: 0.1 mg/kg/h. Penicillins
cIV dose range: 0.05–2.9 mg/kg/h. If still seizing, Cephalosporins
add or switch to propofol or pentobarbital. Isoniazid (treat overdose with pyridoxine)
or Metronidazole
2. cIV propofol: load: 1–2 mg/kg; repeat 1–2 mg/kg Antihistamines, including over-the-counter diphenhydramine
boluses every 3–5 minutes until seizures stop, (Benadryl)
up to maximum total loading dose of 10 mg/kg. Antipsychotics, especially clozapine and low potency
Initial cIV rate: 2 mg/kg/h. cIV dose range: phenothiazines (e.g., chlorpromazine)
1–15 mg/kg/hy . If still seizing, add or switch Antidepressants
to midazolam or pentobarbital. Maprotiline (Ludiomil)
or Bupropion (Wellbutrin)
3. IV valproate: 40 mg/kg over
10 minutes. If still Tricyclics, especially clomipramine; possibly least with
seizing, additional 20 mg/kg over
5 minutes. If still desipramine
seizing, add or switch to cIV midazolam or propofol. Baclofen
or Fentanyl
4. IV phenobarbital: 20 mg/kg IV at 50–100 mg/min. Flumazenil (benzodiazepine antagonist)
If still seizing, add or switch to cIV Ketamine
midazolam, propofol, or pentobarbital. Lidocaine
> 60 cIV pentobarbital: load: 5 mg/kg at up to 50 mg/min; Lithium (especially in overdosage)
repeat 5 mg/kg boluses until seizures stop. Meperidine (Demerol)
Initial cIV rate: 1 mg/kg/h. cIV dose range: Propoxyphene (Darvon)
0.5–10 mg/kg/h; traditionally titrated to Theophylline
suppression-burst on EEG but titrating to seizure Reprinted with permission from Hirsch LJ. Epilepsy and seizure
disorders. In: Marshall RS, Mayer SA, eds. On Call Neurology.
suppression is reasonable as well. Philadelphia, PA: WB Saunders; 2001:364–368. Copyright 2001,
Elsevier.
346 SEMINARS IN NEUROLOGY/VOLUME 28, NUMBER 3 2008

100 age- and sex-matched patients with benzodiazepine- midazolam is equal or superior to IV or rectal benzo-
refractory SE to IV VPA (n ¼ 50) or IV PHT (n ¼ 50).24 diazepines, primarily due to the more rapid administra-
Treatment was considered successful if all motor or tion.27–29 Thus, buccal or nasal midazolam—or another
EEG seizure activity ceased within 20 minutes after the rapid-acting benzodiazepine—should become the out-
beginning of the drug infusion and no seizure activity patient treatment of choice as soon as a practical form
recurred during the next 12 hours. Intravenous VPA was becomes widely available; a buccal midazolam prepara-
successful in 88% and IV PHT in 84% (p > 0.05) of tion is already available in several countries, but not the
patients with SE, with a significantly better response United States.
in patients with SE < 2 hours in duration (p < 0.05).
Of the patients who failed PHT and were crossed over
to VPA, seizure activity ceased in 57% (4/7); of the Refractory SE
5 patients who failed VPA and were crossed over to Most authors define refractory SE as generalized
PHT, 40% (2/5) stopped seizing (not significant). There convulsive or NCSE that continues clinically or
were no differences in secondary outcomes between the electrographically despite first- and second-line therapy
two groups (recurrence of seizures within the 12-hour (Case 2). The likelihood of developing refractory SE
study period or outcome at 7 days), although VPA was increases if SE is not treated early on. After failure of
slightly better tolerated (p > 0.05). Mild elevation of liver benzodiazepines and PHT/fosphenytoin, the traditional
enzymes was found in 8% (4/50) treated with IV VPA; treatment algorithm suggests loading with phenobarbi-
12% (6/50) treated with IV PHT developed hypotension, tal, followed by continuous IV pentobarbital if that fails.
and 4% (2/50) developed respiratory depression. We and others usually choose to proceed directly to

Downloaded by: University of Alberta. Copyrighted material.

rapid-acting, highly potent drips (midazolam or propo-
fol) rather than to phenobarbital once a patient has failed
Alternative Modes of Administering Treatment first- (and possibly second-) line drugs.
for SE and Seizure Clusters (Not IV or Oral) In a systematic review of the literature on treat-
Due to the clear importance of rapid treatment, recent ment of refractory SE, Claassen et al found no difference
studies have investigated out-of-hospital treatment of in mortality (48%) comparing 193 refractory SE patients
SE. Alldredge and associates showed that benzodiaze- treated with cIV propofol, cIV midazolam, or cIV pen-
pines were safe and effective when administered by tobarbital.30 This review also did not demonstrate any
paramedics for out-of-hospital SE in adults.25 In differences between propofol and midazolam for clinical
this study, 59% of patients with SE treated with IV endpoints such as acute treatment failure, breakthrough
lorazepam (2 mg, repeated as needed) in the field were seizures, or post-treatment seizures. By contrast, pento-
no longer seizing upon arrival at the emergency depart- barbital had a lower frequency of acute treatment failure
ment (vs. 43% with IV diazepam 5 mg [repeated as and breakthrough seizures, but this was confounded by
needed] and 21% with placebo). Importantly, the rate of the fact that pentobarbital was more often infused with a
respiratory depression or circulatory complications was titration goal of EEG background suppression and the
lower in the two benzodiazepine groups (10 to 11%) lack of cEEG in most pentobarbital-treated patients. In
compared with the placebo group (22.5%). This and our experience and others’,7,31 most breakthrough seiz-
other studies confirm that not giving benzodiazepines is ures in these patients are nonconvulsive (89% in our
more risky than giving them for prolonged convulsive midazolam series31) and will be unnoticed without
seizures. If widely practiced (as is occurring in many cEEG. Hypotension also occurred more often with
countries), this type of rapid prehospital treatment could pentobarbital (titrated to EEG background suppression)
have a major impact on the prevention of refractory SE. than with propofol or midazolam (usually titrated to
Investigated alternative modes of administration suppression of seizures).30
(primarily of benzodiazepines, especially midazolam) A recent series from Europe showed good out-
include buccal, intranasal, intramuscular, and rectal. comes when propofol was used in combination with IV
Good efficacy and rapid seizure control have been clonazepam32: of 31 episodes of SE, 67% were success-
demonstrated for all of these.26,27 Currently, the only fully treated with no major drug-related adverse events
FDA-approved version of these options is a rectal and a favorable mortality rate of only 22%. The mean
diazepam gel (Diastat, Xcel Pharmaceuticals, San Diego, propofol infusion rate in this study was 4.8 mg/kg/h
CA) available in prefilled syringes. However, buccal or (range, 2.1 to 13, with a goal to achieve suppression-
intranasal benzodiazepines (primarily midazolam) are burst), and the median duration of treatment was 3 days
easier to administer, more socially acceptable than the (range, 1 to 9). It may be that the concomitant IV
rectal route, and allow patients to treat themselves benzodiazepines allowed use of a lower and safer dose
during prolonged auras, simple partial seizures, or clus- of propofol; that is, the benzodiazepine infusion was
ters with recovery between seizures. Several prospective, used as a propofol ‘‘dose-sparing’’ technique. Other
randomized studies have shown that buccal or nasal treatments that have been used for refractory SE include
 TREATMENT OF STATUS EPILEPTICUS/ARIF, HIRSCH 347

very-high-dose phenobarbital (primarily in children, younger than 2 years (> 80%), whereas cryptogenic and
with serum levels 100 to 300 mg/mL), thiopental, remote symptomatic causes are more common in older
lorazepam or diazepam as a continuous infusion, children (p < 0.001).44
lidocaine, etomidate, isoflurane or other inhalational There is accumulating evidence to demonstrate
anesthetics, paraldehyde, electroconvulsive therapy, that nonconvulsive seizure (NCSz) and NCSE occur
transcranial magnetic stimulation, or neurosurgery. See frequently in critically ill children. A retrospective study
Robakis and Hirsch33 for a detailed discussion of treat- at our center examined a population of 117 critically ill
ment of prolonged highly refractory SE. children who underwent cEEG45; seizures were re-
corded in 51 (44%) of 117 patients. The vast majority
of these patients (75%) had purely NCSz, which would
Use of Other Standard Antiepileptic Drugs not have been diagnosed without cEEG. Those with
AEDs that are only available in an oral form can be given clinical seizures (not necessarily prolonged) prior to
via nasogastric tube or percutaneous endoscopic gastro- cEEG were more likely to have NCSz on cEEG
stomy in SE patients, though not for initial treatment. (83%) than those without prior seizures (17%).
Levetiracetam,34 topiramate,35 gabapentin, oxcarbaze- The general principles of management and the
pine, carbamazepine, and pregabalin have been used initial agents used in the pediatric population remain
for this purpose, particularly for prolonged SE refractory the same as for adults (Table 4). The initial drug of
to traditional first- and second-line agents, and to wean choice is IV lorazepam when an IV is available, at 0.05
off anesthetic-dose drips. These medications may be to 0.1 mg/kg, and may be repeated once after 5 minutes
helpful for preventing breakthrough and withdrawal if necessary.46 If IV access cannot be established or the

Downloaded by: University of Alberta. Copyrighted material.

seizures, particularly just prior to tapering cIV AEDs. child is in an out-of-hospital setting, rectal diazepam at
There is some preliminary evidence that topiramate36 0.2 to 0.5 mg/kg should be given, or nasal/buccal
and levetiracetam37 may have neuroprotective or anti- benzodiazepines as discussed above. Unlike adults,
epileptogenic properties as well. cIV propofol may be a poor option due to the seemingly
Levetiracetam has recently become available in an higher risk of multiorgan failure and the propofol
IV formulation; although not FDA-approved for use in infusion syndrome,47 though proper dosing may allow
SE, it is being tested in clinical trials already and results safe use in children.48 Intravenous VPA should be used
should be available soon. It is currently approved for up to cautiously if at all in children under age 2 or children
1500 mg IV given over 15 minutes as replacement for oral with possible metabolic disorders at any age due to the
dosing (1:1 ratio), but there are data showing safety and risk of fulminant hepatitis. A recent retrospective
tolerability in normal volunteers at higher and faster multicenter study from Japan studied the efficacy and
rates of up to 2500 mg in 5 minutes or 4000 mg over safety of midazolam in 358 children with SE.49 Mid-
15 minutes.38 Reports of the efficacy of IV levetiracetam azolam was administered as a bolus dose (0.25  0.21
in NCSE are beginning to emerge.39 Knake et al retro- mg/kg), followed if necessary by continuous infusion
spectively reviewed the use of IV levetiracetam in 18 (0.26  0.25 mg/kg/h). Of the 286 cases given bolus
episodes (16 patients) of SE (mean loading dose 944 mg, midazolam, the seizures stopped in 162/286 (56.6%). Of
maintenance dose 2166 mg/24 h).40 All patients received these 162 cases, 28 were seizure-free with no further need
at least a benzodiazepine (lorazepam in 94%) before IV for continuous midazolam infusion, 119 remained symp-
levetiracetam was given. SE was controlled in all patients tom-free with continuous midazolam infusion, and 15
by administration of a combination of different AEDs. cases showed seizure relapse. The incidence of adverse
In 16 episodes, IV levetiracetam was the last drug events was 19.2% in the group receiving the bolus only,
administered: only 2 patients required further treatment 28.2% in the group given continuous infusion not more
with other AEDs. Tracheal intubation was avoided in than 0.4 mg/kg/h, and 58.6% in those who received > 0.4
17/18 episodes (94%). All patients tolerated intravenous mg/kg/h. Respiratory distress was seen in 86 (24%) cases;
levetiracetam without any severe adverse events. only 71 (19.8%) required tracheal intubation.
As in adults, it is important in the pediatric
population to exclude NCSE by performing a prolonged
SE in Pediatric Patients EEG if a patient does not wake up after the clinical
Convulsive SE is the most common medical neuro- control of convulsive seizures or SE. In fact, prolonged
logical emergency in childhood.41 Etiology of SE in EEG monitoring may be warranted in pediatric patients
children differs from that in adults, with prolonged even after brief convulsive seizures. A recent study of 19
febrile seizures being the most common cause of SE; pediatric patients with NCSE reported that although
this is generally associated with low morbidity and most children (17/19) with NCSE had clinical (usually
mortality.42,43 The distribution of etiology is reported convulsive) seizures prior to onset of NCSE, in the
to be highly age-dependent. Acute symptomatic etiol- majority of cases (12/19) these seizures were brief,
ogies or febrile seizures are more common in children isolated convulsions.50
348 SEMINARS IN NEUROLOGY/VOLUME 28, NUMBER 3 2008

SE that is not controlled by first-line treatment increased the risk for recurrent SE. Both partial SE and
(refractory SE) may require prolonged treatment with good therapeutic response to the initial AED therapy
other agents in many cases, and knowledge of the were associated with a decreased risk of recurrent SE.
prognosis is important in making therapeutic decisions
in these patients. Although the prognosis of SE in
general is better in children than adults, mortality of How Aggressively Should SE Be Treated?
refractory SE in children seems more similar to that in There are no studies demonstrating a convincing differ-
adults (16 to 43.5%).51–53 Etiology is a very important ence in outcome based on the goal of treatment (seizure
determinant of outcome, and acute symptomatic refrac- suppression vs. burst suppression vs. flat line) independ-
tory SE carries a higher mortality and morbidity. Sahin ent of etiology. Krishnamurthy and Drislane reported
and colleagues54 studied the long-term outcome in seven persistent seizure control after pentobarbital taper in 3/3
previously normal children with refractory SE and pre- patients who never reached suppression-burst, 6/12 who
sumed encephalitis. High-dose pentobarbital or mida- reached suppression-burst but never a flat record, and
zolam or both was administered to achieve a 17/20 who reached a flat record (complete suppression).
suppression-burst pattern on EEG, aiming for an inter- Recurrence of electrographic status predicted clinical
burst interval of > 5 seconds. Duration of treatment relapse as well. Isolated epileptiform discharges during
ranged from 11 to 146 days. All seven patients developed pentobarbital treatment did not correlate with out-
intractable epilepsy on long-term follow-up, as well as come.57 The same authors also found that clinical relapse
severe learning problems and attention disorders. It is after pentobarbital withdrawal predicted higher mortal-
unclear whether the poor outcome results from severe ity (mortality of 8/9 vs. 9/26, p < 0.005).34 Thus, these

Downloaded by: University of Alberta. Copyrighted material.

brain injury caused by the acute insult alone or if retrospective, uncontrolled data (the only kind of data
prolonged refractory SE contributes to neurologic mor- available for this question) suggest possible benefit from
bidity. However, these results do suggest that refractory a period of intense EEG suppression. The most recent
SE may need to be treated more aggressively. After attempt to answer this by Rossetti and colleagues, who
discharge, a latent period was seen before seizure recur- reviewed 49 episodes of refractory SE, concluded that
rence in five of the patients, with a mean time to seizure outcome was independent of the agent of choice and
recurrence of 5.3 months; this raises the question of the extent of EEG suppression.58 Similarly, it is still
whether different agents, such as those providing neuro- debatable how aggressively to treat NCSE.59,60 While
protection during refractory SE or antiepileptogenesis nonconvulsive seizures have definite potential for neu-
after refractory SE, would help prevent this evolution ronal injury, aggressive treatment also carries signifi-
into refractory chronic epilepsy. cant risk. Only a randomized trial can truly address this
issue. One retrospective study in the critically ill elderly
found that aggressive treatment with IV benzodiaze-
Tapering Off Continuous Infusions pines was associated with worse outcome despite sim-
In all patients treated with continuous infusions of an ilar severity of illness.61 When comparing outcome via
AED, these should be continued for 12 to 24 hours after conservative non-ICU treatment due to advance direc-
seizures are stopped before a gradual taper is started. tives to outcome with ICU care, aggressive ICU man-
Our practice is to taper off continuous infusions over agement prolonged stay by 17 days with no difference

24 hours. If seizures recurred with a prior taper, it may in outcome.
be necessary to treat longer and taper more slowly the
next time while maintaining high therapeutic levels of
other AEDs. Maintaining patients on phenobarbital FUTURE DIRECTIONS
may increase the chances of weaning off of pentobarbital IN THE MANAGEMENT OF SE
without seizure recurrence.55 A randomized trial is needed to determine the best
treatment for SE that does not stop with lorazepam.
Although IV lorazepam is an excellent and proven first-
Recurrence of SE line treatment, the steps after this are less clear. The
Hesdorffer and colleagues recently studied the risk of exact roles of IV VPA and levetiracetam are still unclear
recurrence of SE in a population-based sample.56 Among (particularly the latter), but they are useful additions to
the 183 episodes of first afebrile SE, the risk of recurrent the armamentarium of drugs used to treat SE. Buccal/
SE was 31.7% over a 10-year follow-up period. The risk nasal midazolam has great promise for out-of-hospital
of recurrence was
25% for those with acute sympto- treatment that may greatly decrease delays in diagnosis
matic SE, remote symptomatic SE, and idiopathic cryp- and treatment, thus preventing refractory SE and possi-
togenic SE. Recurrence was 100% for those with SE bly improving outcomes. Randomized trials are also
associated with a progressive symptomatic disease. needed to help decide the goal or depth of treatment:
Female gender and progressive symptomatic etiology cessation of seizures versus suppression of background.
 TREATMENT OF STATUS EPILEPTICUS/ARIF, HIRSCH 349

Animal studies of prolonged status will also be helpful in CONCLUSION

this regard. SE is a neurological emergency. Rapid diagnosis and
The next important focus for research in treating treatment are much more important than the choice of
SE is likely to be neuroprotection, and animal research agent used. Out-of-hospital treatment of prolonged
will probably be most important in the coming decade. seizures or clusters is now available and should be
There is preliminary evidence that some of the newer utilized. Widespread use of benzodiazepines by care-
AEDs have neuroprotective and antiepileptogenic prop- givers or paramedics, including buccal, nasal, and rectal
erties (i.e., they may prevent neuronal injury and future administration, could significantly improve prevention
epilepsy). Other potential neuroprotective agents or of refractory SE.
methods that may be worthy of study in SE include NCSE remains an underdiagnosed entity; it
hypothermia, glutamate receptor antagonists, calcium should be promptly considered in anyone with unex-
channel blockers, antioxidants, and erythropoietin. plained impairment or fluctuation of mental status/
Development of a reliable neuronal injury marker behavior, especially if there is a history of seizures or
will also be helpful in determining which patients and epilepsy risk factors, in children, in coma, and after
EEG patterns require treatment, and may have a role convulsive SE if the patient does not rapidly awaken.
as a surrogate marker for outcome. The work of Further research is needed to determine how
DeGiorgio62 and others suggests that neuron-specific aggressively to treat critically ill patients with NCSE,
enolase, or NSE, may be an appropriate marker for this. which agent to use, when to stop treatment, and how to
After isolated seizures, NSE is usually normal or ele- interpret periodic discharges and other equivocal EEG
vated for a few hours only. However, after SE it is often patterns.

Downloaded by: University of Alberta. Copyrighted material.

elevated for days. NSE correlates with the duration of As the population ages, it is likely that we will be
SE and the outcome. It is most elevated in patients with seeing more patients with SE, especially the nonconvul-
subtle or nonconvulsive seizures. It is also elevated in sive and refractory forms. Technology is now available to
patients with SE but without any acute brain injury, study in detail cerebral blood flow, brain metabolism,
further evidence that seizures themselves, including intracranial pressure, EEG, neuronal injury markers, and
nonconvulsive ones, can cause neuronal injury. other parameters in these patients. Active research in
Education about the prevalence of subclinical these areas, as well as in neuroprotection and antiepi-
seizures and NCSE is immensely important. Diagnosis leptogenesis, will continue to keep SE an evolving,
is often delayed, and this delay has been found to highly active topic in the coming years. We hope the
correlate with a worse prognosis.14 Education and train- lessons learned will translate into more effective treat-
ing regarding the pitfalls of EEG interpretation will also ment and improved outcomes.
be important; the review of simultaneous video is quite
helpful when attempting to distinguish between ictal
patterns and artifact or encephalopathic patterns. CASE STUDIES
Whether or not further brain injury is occurring
during NCSE or periodic discharges can be studied in Case 1: Poststroke NCSE
many ways, perhaps most effectively with animal mod- A 67-year-old man with hypertension and atrial fibrilla-
els. In humans, there are now many potential markers tion on chronic warfarin presented with dizziness, left
of neuronal injury and noninvasive methods of studying hemiparesis, and lethargy. Initial head CT was negative
the underlying pathophysiology during and after SE: and international normalized ratio was subtherapeutic.
magnetic resonance imaging (MRI) with diffusion In the emergency room, the neurology consultant noted
weighted imaging (often shows focal bright signal rhythmic facial twitching and the patient was loaded
representing restricted diffusion during and after SE), with fosphenytoin 18 mg/kg. No further twitching was
functional MRI, serum and cerebrospinal fluid (CSF) seen, and initial overnight EEG showed right hemi-
markers such as NSE discussed above, positron emis- sphere slowing and occasional isolated epileptiform dis-
sion tomography, single photon emission computed charges on the right, but no seizures. Fosphenytoin was
tomography, MR spectroscopy (e.g., for lactate), and discontinued after several days and levetiracetam was
other more invasive monitoring techniques such as started instead to minimize drug interactions and po-
brain-tissue oxygen and cerebral microdialysis (for tential long-term adverse effects. Repeat head CT
parenchymal interstitial fluid lactate, pyruvate, gluta- (Fig. 1) showed hemorrhagic conversion of a moderate
mate, glycerol, neurotransmitters, and more). It is to large right middle cerebral artery infarct and a smaller
hoped that all these techniques, combined with clinical left medial parietal hemorrhagic infarct. The patient was
trials and experience, will enable more definitive evi- stable, but developed medical complications including
dence-based and physiology-based recommendations pneumonia, renal insufficiency, and a deep vein throm-
and improved outcomes in management of patients bosis. On day 14 of his hospital stay he had a witnessed
with SE. convulsion, followed by continued impaired mental
350 SEMINARS IN NEUROLOGY/VOLUME 28, NUMBER 3 2008

Levetiracetam was increased to 1500 mg twice a

day and PHT (100 mg) was given three times a day per
feeding tube. However, levels were low and erratic with
enteric administration so fosphenytoin was resumed IV
at 200 mg twice a day instead, maintaining total PHT
serum levels at
8 to 12 mg/mL, which corresponded to
a measured free PHT of 2.0 to 3.0 mg/mL. Seizures
rapidly decreased to a few isolated nonconvulsive seiz-
ures per day that resolved over a few days as the patient’s
medical condition improved and PHT levels were maxi-
mized. Due to occasional agitation, levetiracetam was
changed to gabapentin, rapidly increased to 900 mg
three times a day. His mental status improved gradually,
he resumed speaking and following simple commands,
but he was left with significant neurological impairment.
He was discharged to a nursing home.

COMMENT
This case shows NCSE after an isolated but pro-
longed convulsive seizure in an acutely ill inpatient in

Downloaded by: University of Alberta. Copyrighted material.

Figure 1 Case 1: Head computed tomography image
showing bilateral hemorrhagic infarcts, much larger on the the poststroke setting. Medication management was
right (left side of figure). complicated by the desire to avoid drug interactions in
a complex medical patient (by avoiding the strong
status and leg twitching. He was given 4 mg of loraze- P450 enzyme-inducer PHT), probable poor gastro-
pam and was reloaded with fosphenytoin 18 mg/kg. All intestional absorption (preventing confidence in en-
motor activity ceased after
15 minutes, as fosphenytoin teral medications, especially ones without readily
was being administered. Urgent EEG was arranged and available serum levels), and agitation (which can occur
showed nearly continuous seizures from the left posterior from some antiepileptic medications, particularly lev-
quadrant (see EEG, Figs. 2, 3, and 4) lasting up to 4 etiracetam). EEG monitoring played an important
minutes each with only a minute or so between seizures. role in following this patient’s nonconvulsive seizures
There was no detectable clinical correlate to any of these as they could not otherwise be identified. Seizures
electrographic seizures. increase metabolic demand and cerebral blood flow

Figure 2 Case 1: Start of typical seizure from left posterior quadrant.

 TREATMENT OF STATUS EPILEPTICUS/ARIF, HIRSCH 351

Downloaded by: University of Alberta. Copyrighted material.

Figure 3 Case 1 continued: Middle of seizure, almost 2 minutes after EEG in Fig. 2, now involving entire left hemisphere.

and release glutamate, all of which are likely to Case 2: Refractory SE Requiring 1 Month
exacerbate neuronal injury in the acute stroke setting of Iatrogenic Coma but with Good Outcome
such as this. As is typical in critically ill patients, the A 63-year-old healthy woman presented with viral symp-
occasional clinical seizure was just the tip of the toms for a few days including fever, followed by a con-
iceberg. It is interesting that his seizures seemed to vulsion. She was loaded with fosphenytoin in the
arise from the smaller left parietal hemorrhagic infarct emergency room but her mental status remained impaired
rather than the large one on the right. and intermittent focal seizures continued, consisting of

Figure 4 Case 1 continued: End of seizure, 1 minute after EEG in Fig. 3. After seizure offset, seizure activity is replaced by
periodic lateralized epileptiform discharges recurring at just under 1 per second, also maximal in the left posterior quadrant.
352 SEMINARS IN NEUROLOGY/VOLUME 28, NUMBER 3 2008

Figure 5 Case 2: Onset of typical nonconvulsive seizure from the right temporal region with spread to the right parasagittal
region.

eye deviation and facial twitching. Lorazepam was given myopathy improved and resolved completely after
several times. Lumbar punctures showed 13 to 22 white
6 months. Subsequent MRIs showed mild bilateral
blood cells per mm3 (mostly lymphocytes), mildly elevated hippocampal atrophy and increased signal suggesting
protein concentration, and normal glucose concentration. hippocampal sclerosis. She returned to work part time
Brain MRI was normal as were all blood tests. CSF, and functioned well. One year later, she underwent

Downloaded by: University of Alberta. Copyrighted material.

herpes simplex virus, polymerase chain reaction, and other epilepsy monitoring for refractory seizures, averaging a
encephalitis tests were negative. EEG monitoring showed couple per week. Bilateral independent temporal
NCSE (see EEG, Figs. 5 and 6). The patient was treated lobe seizures were identified as well as occipital simple
with many agents, including ceftriaxone, acyclovir, mid- partial seizures. She received a vagus nerve stimulator,
azolam infusion, propofol infusion (while following crea- which helped modestly. She has since retired and now
tine phosphokinase, pH, and triglycerides), IV VPA, travels the world with her husband. She continues to have
levetiracetam, topiramate (after off propofol), oxcarbaze- about 1 complex partial seizure per week.
pine (while following sodium closely), and prolonged
pentobarbital coma for 4 weeks. Every 1 to 4 days, an COMMENT
attempt to wean her off of pentobarbital was made, but This case reminds us that the prognosis in refractory SE,
seizures recurred each time for several weeks. Phenobar- although poor overall, is not that bad for those with no
bital was maintained to attempt to minimize withdrawal severe brain injury. It is not unusual to require a few
seizures as pentobarbital was tapered, and this was even- weeks or more of iatrogenic coma for encephalitis-
tually successful after just over 1 month in iatrogenic related SE and still have good functional outcome,
coma. No organism was ever identified for her presumed though most are left with epilepsy that can be severe.
encephalitis. It is possible that earlier identification and treatment of
The patient gradually awoke. Intermittent isolated NCSE could improve outcome even more, and perhaps
seizures were tolerated. Over the subsequent 2 months, prevent the permanent hippocampal injury and related
she gradually returned to baseline mental status except for seizures that many of these patients, including this one,
mild-moderate memory dysfunction. Her critical illness develop.

Figure 6 Case 2 continued: Continuation of seizure, remaining maximal on the right but spreading to the left, then ending at
the end of this EEG segment. Left sided periodic discharges continued.
 TREATMENT OF STATUS EPILEPTICUS/ARIF, HIRSCH 353

REFERENCES 21. Sinha S, Naritoku DK. Intravenous valproate is well tolerated

1. Logroscino G, Hesdorffer DC, Cascino G, et al. Time trends in unstable patients with status epilepticus. Neurology 2000;
in incidence, mortality, and case-fatality after first episode of 55:722–724
status epilepticus. Epilepsia 2001;42:1031–1035 22. Limdi NA, Shimpi AV, Faught E, et al. Efficacy of rapid
2. Claassen J, Lokin JK, Fitzsimmons BF, et al. Predictors of IV administration of valproic acid for status epilepticus.
functional disability and mortality after status epilepticus. Neurology 2005;64:353–355
Neurology 2002;58:139–142 23. Peters CN, Pohlmann-Eden B. Intravenous valproate as an
3. DeLorenzo RJ, Hauser WA, Towne AR, et al. A prospective, innovative therapy in seizure emergency situations including
population-based epidemiologic study of status epilepticus in status epilepticus: experience in 102 adult patients. Seizure
Richmond, Virginia. Neurology 1996;46:1029–1035 2005;14:164–169
4. Lowenstein DH, Alldredge BK. Status epilepticus. N Engl J 24. Agarwal P, Kumar N, Chandra R, et al. Randomized study of
Med 1998;338:970–976 intravenous valproate and phenytoin in status epilepticus.
5. Lowenstein DH, Bleck T, Macdonald RL. It’s time to revise Seizure 2007;16:527–532
the definition of status epilepticus. Epilepsia 1999;40:120–122 25. Alldredge BK, Gelb AM, Isaacs SM, et al. A comparison of
6. Jenssen S, Gracely EJ, Sperling MR. How long do most lorazepam, diazepam, and placebo for the treatment of out-
seizures last? A systematic comparison of seizures recorded in of-hospital status epilepticus. N Engl J Med 2001;345:631–
the epilepsy monitoring unit. Epilepsia 2006;47:1499–1503 637
7. DeLorenzo RJ, Waterhouse EJ, Towne AR, et al. Persistent 26. Lahat E, Goldman M, Barr J, et al. Comparison of intranasal
nonconvulsive status epilepticus after the control of con- midazolam with intravenous diazepam for treating febrile
vulsive status epilepticus. Epilepsia 1998;39:833–840 seizures in children: prospective randomised study. BMJ
8. Claassen J, Mayer SA, Kowalski RG, et al. Detection of 2000;321:83–86
electrographic seizures with continuous EEG monitoring in 27. Scott RC, Besag FM, Neville BG. Buccal midazolam and

Downloaded by: University of Alberta. Copyrighted material.

critically ill patients. Neurology 2004;62:1743–1748 rectal diazepam for treatment of prolonged seizures in
9. Jordan KG. Continuous EEG monitoring in the neuro- childhood and adolescence: a randomised trial. Lancet 1999;
science intensive care unit and emergency department. J Clin 353:623–626
Neurophysiol 1999;16:14–39 28. Baysun S, Aydin OF, Atmaca E, et al. A comparison of
10. Vespa PM, Nuwer MR, Nenov V, et al. Increased incidence buccal midazolam and rectal diazepam for the acute
and impact of nonconvulsive and convulsive seizures after treatment of seizures. Clin Pediatr (Phila) 2005;44:771–776
traumatic brain injury as detected by continuous electro- 29. McIntyre J, Robertson S, Norris E, et al. Safety and efficacy
encephalographic monitoring. J Neurosurg 1999;91:750– of buccal midazolam versus rectal diazepam for emergency
760 treatment of seizures in children: a randomised controlled
11. Towne AR, Waterhouse EJ, Boggs JG, et al. Prevalence of trial. Lancet 2005;366:205–210
nonconvulsive status epilepticus in comatose patients. Neu- 30. Claassen J, Hirsch LJ, Emerson RG, et al. Treatment of
rology 2000;54:340–345 refractory status epilepticus with pentobarbital, propofol, or
12. Claassen J, Jette N, Chum F, et al. Electrographic seizures midazolam: a systematic review. Epilepsia 2002;43:146–
and periodic discharges after intracerebral hemorrhage. 153
Neurology 2007;69:1356–1365 31. Claassen J, Hirsch LJ, Emerson RG, et al. Continuous EEG
13. Vespa PM, O’Phelan K, Shah M, et al. Acute seizures after monitoring and midazolam infusion for refractory non-
intracerebral hemorrhage: a factor in progressive midline shift convulsive status epilepticus. Neurology 2001;57:1036–1042
and outcome. Neurology 2003;60:1441–1446 32. Rossetti AO, Reichhart MD, Schaller MD, et al. Propofol
14. Young GB, Jordan KG, Doig GS. An assessment of treatment of refractory status epilepticus: a study of 31
nonconvulsive seizures in the intensive care unit using episodes. Epilepsia 2004;45:757–763
continuous EEG monitoring: an investigation of variables 33. Robakis TK, Hirsch LJ. Literature review, case report, and
associated with mortality. Neurology 1996;47:83–89 expert discussion of prolonged refractory status epilepticus.
15. Lowenstein DH, Alldredge BK. Status epilepticus at an Neurocrit Care 2006;4:35–46
urban public hospital in the 1980s. Neurology 1993;43(3 Pt 1): 34. Rossetti AO, Bromfield EB. Levetiracetam in the treatment
483–488 of status epilepticus in adults: a study of 13 episodes. Eur
16. Mazarati AM, Baldwin RA, Sankar R, et al. Time-depend- Neurol 2005;54:34–38
ent decrease in the effectiveness of antiepileptic drugs during 35. Towne AR, Garnett LK, Waterhouse EJ, et al. The use of
the course of self-sustaining status epilepticus. Brain Res topiramate in refractory status epilepticus. Neurology 2003;
1998;814(1–2):179–185 60:332–334
17. Claassen J, Hirsch LJ, Mayer SA. Treatment of status 36. Niebauer M, Gruenthal M. Topiramate reduces neuronal
epilepticus: a survey of neurologists. J Neurol Sci 2003;211 injury after experimental status epilepticus. Brain Res 1999;
(1–2):37–41 837(1–2):263–269
18. Treiman DM, Meyers PD, Walton NY, et al. A comparison 37. Klitgaard H. Levetiracetam: the preclinical profile of a new
of four treatments for generalized convulsive status epilepti- class of antiepileptic drugs? Epilepsia 2001;42(S4):13–18
cus. Veterans Affairs Status Epilepticus Cooperative Study 38. Ramael S, Daoust A, Otoul C, et al. Levetiracetam
Group. N Engl J Med 1998;339:792–798 intravenous infusion: a randomized, placebo-controlled
19. Hirsch LJ, Arif H. Status epilepticus. Continuum: Lifelong safety and pharmacokinetic study. Epilepsia 2006;47:1128–
Learning Neurol 2007;13:121–151 1135
20. Misra UK, Kalita J, Patel R. Sodium valproate vs phenytoin 39. Farooq MU, Naravetla B, Majid A, et al. IV levetiracetam in
in status epilepticus: a pilot study. Neurology 2006;67:340– the management of non-convulsive status epilepticus. Neuro-
342 crit Care 2007;7:36–39
354 SEMINARS IN NEUROLOGY/VOLUME 28, NUMBER 3 2008

40. Knake S, Gruener J, Hattemer K, et al. Intravenous epilepticus in children: a meta-analysis. J Child Neurol 1999;
levetiracetam in the treatment of benzodiazepine refractory 14:602–609
status epilepticus. J Neurol Neurosurg Psychiatry 2008;79: 52. Sahin M, Menache CC, Holmes GL, et al. Outcome of
588–589 severe refractory status epilepticus in children. Epilepsia
41. DeLorenzo RJ, Pellock JM, Towne AR, et al. Epidemiology 2001;42:1461–1467
of status epilepticus. J Clin Neurophysiol 1995;12:316–325 53. Kim SJ, Lee DY, Kim JS. Neurologic outcomes of pediatric
42. Chin RF, Neville BG, Peckham C, et al. Incidence, cause, epileptic patients with pentobarbital coma. Pediatr Neurol
and short-term outcome of convulsive status epilepticus in 2001;25:217–220
childhood: prospective population-based study. Lancet 2006; 54. Sahin M, Menache CC, Holmes GL, et al. Prolonged
368:222–229 treatment for acute symptomatic refractory status epilepticus:
43. Hesdorffer DC, Logroscino G, Cascino G, et al. Incidence of outcome in children. Neurology 2003;61:398–401
status epilepticus in Rochester, Minnesota, 1965–1984. 55. Krishnamurthy KB, Drislane FW. Relapse and survival after
Neurology 1998;50:735–741 barbiturate anesthetic treatment of refractory status epilepti-
44. Shinnar S, Pellock JM, Moshe SL, et al. In whom does status cus. Epilepsia 1996;37:863–867
epilepticus occur: age-related differences in children. Epilepsia 56. Hesdorffer DC, Logroscino G, Cascino GD, et al. Recurrence
1997;38:907–914 of afebrile status epilepticus in a population-based study in
45. Jette N, Claassen J, Emerson RG, et al. Frequency and Rochester, Minnesota. Neurology 2007;69: 73–78
predictors of nonconvulsive seizures during continuous 57. Krishnamurthy KB, Drislane FW. Depth of EEG suppres-
electroencephalographic monitoring in critically ill children. sion and outcome in barbiturate anesthetic treatment for
Arch Neurol 2006;63:1750–1755 refractory status epilepticus. Epilepsia 1999;40:759–762
46. Eriksson K, Kälviäinen R. Pharmacologic management of 58. Rossetti AO, Logroscino G, Bromfield EB. Refractory status
convulsive status epilepticus in childhood. Expert Rev epilepticus: effect of treatment aggressiveness on prognosis.

Downloaded by: University of Alberta. Copyrighted material.

Neurother 2005;5:777–783 Arch Neurol 2005;62:1698–1702
47. Hanna JP, Ramundo ML. Rhabdomyolysis and hypoxia 59. Jordan KG, Hirsch LJ. Nonconvulsive status epilepticus
associated with prolonged propofol infusion in children. (NCSE), treat to burst-suppression: pro and con. Epilepsia
Neurology 1998;50:301–303 2006;47(suppl l):41–45
48. van Gestel JP, Blusse van Oud-Alblas HJ, Malingre M, et al. 60. Jirsch J, Hirsch LJ. Nonconvulsive seizures: developing a
Propofol and thiopental for refractory status epilepticus in rational approach to the diagnosis and management in the
children. Neurology 2005;65:591–592 critically ill population. Clin Neurophysiol 2007;118:1660–
49. Hayashi K, Osawa M, Aihara M, et al. Efficacy of 1670
intravenous midazolam for status epilepticus in childhood. 61. Litt B, Wityk RJ, Hertz SH, et al. Nonconvulsive status
Pediatr Neurol 2007;36:366–372 epilepticus in the critically ill elderly. Epilepsia 1998;39:
50. Tay SK, Hirsch LJ, Leary L, et al. Nonconvulsive status 1194–1202
epilepticus in children: clinical and EEG characteristics. 62. DeGiorgio CM, Heck CN, Robinowicz AL, Gott PS,
Epilepsia 2006;47:1504–1509 Smith T, Correale J. Serum neuron-specific enolase in the
51. Gilbert DL, Gartside PS, Glauser TA. Efficacy and mortality major subtypes of status epilepticus. Neurolgy 1999;52:746–
in treatment of refractory generalized convulsive status 749

View publication stats