ORIGINAL ARTICLE

Prospective, Randomized, Single-Blinded Comparative Trialof Intravenous Levetiracetam Versus Phenytoin for SeizureProphylaxis

Jerzy P. Szaflarski • Kiranpal S. Sangha •

Christopher J. Lindsell • Lori A. Shutter

Published online: 7 November 2009

� Humana Press Inc. 2009

Abstract

Background Anti-epileptic drugs are commonly used for

seizure prophylaxis after neurological injury. We per-

formed a study comparing intravenous (IV) levetiracetam

(LEV) to IV phenytoin (PHT) for seizure prophylaxis after

neurological injury.

Methods In this prospective, single-center, randomized,

single-blinded comparative trial of LEV versus PHT (2:1

ratio) in patients with severe traumatic brain injury (sTBI)

or subarachnoid hemorrhage (NCT00618436) patients

received IV load with either LEV or fosphenytoin followed

by standard IV doses of LEV or PHT. Doses were adjusted

to maintain therapeutic serum PHT concentrations or if

patients had seizures. Continuous EEG (cEEG) monitoring

was performed for the initial 72 h; outcome data were

collected.

Results A total of 52 patients were randomized

(LEV = 34; PHT = 18); 89% with sTBI. When control-

ling for baseline severity, LEV patients experienced better

long-term outcomes than those on PHT; the Disability

Rating Scale score was lower at 3 months (P = 0.042) and

the Glasgow Outcomes Scale score was higher at 6 months

(P = 0.039). There were no differences between groups in

seizure occurrence during cEEG (LEV 5/34 vs. PHT 3/18;

P = 1.0) or at 6 months (LEV 1/20 vs. PHT 0/14;

P = 1.0), mortality (LEV 14/34 vs. PHT 4/18; P = 0.227).

There were no differences in side effects between groups

(all P > 0.15) except for a lower frequency of worsened

neurological status (P = 0.024), and gastrointestinal

problems (P = 0.043) in LEV-treated patients.

Conclusions This study of LEV versus PHT for seizure

prevention in the NSICU showed improved long-term

outcomes of LEV-treated patients vis-a-vis PHT-treated

patients. LEV appears to be an alternative to PHT for

seizure prophylaxis in this setting.

Keywords Levetiracetam � Phenytoin � Fosphenytoin �Seizure prevention � ICU � SAH � TBI �Long-term outcomes � GCS � GOS � DRS

Introduction

Seizures in the setting of acute brain injury are common;

the chance of seizure occurrence depends, in part, on the

J. P. Szaflarski (&) � L. A. Shutter

Department of Neurology, University of Cincinnati Academic

Health Center, 260 Stetson Street, Rm. 2350, Cincinnati,

OH 45267-0525, USA

e-mail: jerzy.szaflarski@uc.edu

J. P. Szaflarski

Cincinnati Epilepsy Center at the University Hospital,

Cincinnati, OH, USA

J. P. Szaflarski � L. A. Shutter

The University of Cincinnati Neuroscience Institute, Cincinnati,

OH, USA

K. S. Sangha

Department of Pharmacy Services, The University Hospital,

Cincinnati, OH, USA

K. S. Sangha

James L. Winkle College of Pharmacy, University of Cincinnati,

Cincinnati, OH, USA

C. J. Lindsell

Department of Emergency Medicine, University of Cincinnati

Academic Health Center, Cincinnati, OH, USA

L. A. Shutter

Department of Neurosurgery, University of Cincinnati Academic

Health Center, Cincinnati, OH, USA

Neurocrit Care (2010) 12:165–172

DOI 10.1007/s12028-009-9304-y

severity of neurological injury [1]. Approximately 8.4%

of patients with subarachnoid hemorrhage have overt

seizures within the first 24 h of presentation and the

combined incidence of covert (as detected by EEG) and

overt seizures in patients with traumatic brain injury (TBI)

or subarachnoid hemorrhage (SAH) may reach 25–50%

[2–6]. As a consequence of seizures in the acute setting

there is an increase in secondary injuries including

aneurysmal rupture or re-rupture, intermittent, and sus-

tained increased intracranial pressure, hypoxia, physical

injury, and death. Any of these complications may

adversely affect the neurological status of patients with

brain injury and worsen their clinical outcome. Finally,

early seizures may be predictive of subsequent epilepsy

development [7, 8]. The prevalence of post-traumatic epi-

lepsy (approximately 6% of all epilepsies), the awareness of

the high incidence of seizures after neurological injury and

the contribution of seizures to secondary injury suggest the

use of prophylactic anti-epileptic drugs (AEDs) in this

setting [9].

Currently, the American Academy of Neurology sup-

ports the use of phenytoin (PHT) in the setting of acute

traumatic brain injury for seizure prevention [10]. But,

PHT carries high chance of potential side effects, medi-

cation interactions, and potential harmful reactions

including anticonvulsant hypersensitivity syndrome, rash

or Stevens–Johnson syndrome, tissue necrosis complicat-

ing medication extravasation, and purple glove syndrome.

Therefore, better treatment options than PHT are needed.

Oral and, more recently, intravenous (IV) levetiracetam

(LEV) have been studied in open-label trials or in a ret-

rospective fashion in the acute care setting [11–16]. For

example, we recently completed a review of 379 patients

who received AEDs for seizure prophylaxis in the NSICU

setting [16]. We have shown that when PHT was used prior

to the NSICU admission, it was frequently replaced during

the ICU stay with LEV monotherapy (P < 0.001) and that

patients treated with LEV monotherapy when compared to

other AEDs had lower complication rates and shorter

NSICU stays. Our results suggested that LEV may be a

desirable alternative to PHT. Based on our experiences

with IV LEV, we designed a standardized seizure pro-

phylaxis protocol for patients with severe TBI and high

grade SAH using IV LEV. This prospective, randomized

single-blinded study compares patients treated with IV

LEV to those treated with PHT as prophylactic AEDs in

the NSICU setting. The primary objective was to compare

the safety of LEV in critically ill NSICU patients to the

safety of PHT, which is currently the most commonly used

agent. The secondary objectives were to compare the rate

of clinically evident and sub-clinical seizures, and to

compare long-term outcomes between patients treated with

LEV and those treated with PHT.

Methods

Subject Recruitment, Screening and the Informed

Consent Process

This study was approved by the Institutional Review

Boards at the University of Cincinnati and The University

Hospital. The study was registered with www.Clinical

Trials.gov, Identifier: NCT00618436. Subjects were iden-

tified by the neuro-intensive care physicians from patients

admitted to the Neuroscience ICU (NSICU). Screening

procedures included a complete medical history, details of

the precipitating event, physical examinations, complete

baseline and ongoing vital sign assessments, neurological

evaluations, laboratory results, and diagnostic imaging

performed. Screening assessments were performed by the

clinician involved in the care of the patient to determine

subject eligibility criteria, especially GCS or Hunt–Hess

diagnosis.

Inclusion criteria for enrollment included: (1) traumatic

brain injury or subarachnoid hemorrhage admitted to the

hospital less than 24 h prior to randomization; (2) GCS score

3–8 (inclusive), or GCS motor score of 5 or less and abnor-

mal admission CT scan showing intracranial pathology; (3)

hemodynamically stable with a systolic BP C90 mmHg; (4)

at least one reactive pupil; (5) C17 years of age; and (6)

signed informed consent and HIPAA authorization for

research form. Exclusion criteria for enrollment included:

(1) no venous access; (2) spinal cord injury; (3) history of or

CT confirmation of previous brain injury such as brain

tumor, cerebral infarct, or spontaneous intracerebral hem-

orrhage; (4) hemodynamically unstable; (5) suspected

anoxic events; (6) other peripheral trauma likely to result in

liver failure; (7) age less than 17 years of age; (8) known

hypersensitivity to any anticonvulsant; (9) any treatment,

condition, or injury that contraindicated treatment with LEV

or PHT; and (10) inability to obtain signed informed consent

and HIPAA authorization for research.

General Design

This investigator initiated trial was originally designed to

enroll 52 patients with SAH and 52 patients with severe

traumatic brain injury (sTBI), but recruitment and funding

issues prompted a change in design to focus on sTBI and

stop enrollment at 52 patients. Randomization occurred as

soon as possible and up to 24 h after admission in the

NSICU and was done at a 2:1 ratio of LEV to PHT; sub-

jects were randomized and treatment group was assigned

by the pharmacy. Once enrolled, cEEG was initiated and

continued for up to 72 h. The study electrophysiologist

(J. P. Szaflarski) was blinded to the group assignment or

166 Neurocrit Care (2010) 12:165–172

diagnosis and reported results of the EEGs to the PI (L. A.

Shutter) on a daily basis. The managing physicians were

partially blinded in that they were not aware of which

group the patient was randomized into, but PHT levels

could be reviewed in the hospital laboratory computer. The

managing physicians were also unblinded to treatment

group if seizures occurred in order to optimize treatment.

All patients were treated with the standard of care for

TBI or SAH per NSICU protocols. TBI management pro-

tocols followed the Guidelines for Management of Severe

Traumatic Brain Injury [17]. SAH management protocols

were based on treatment algorithms developed with the

Neurosurgery Department’s Cerebrovascular Service at the

University Hospital. AED management used standardized

doses at the time of initiation, with adjustments in the

dosing performed by the study pharmacist (K. S. Sangha)

to maintain therapeutic levels of PHT. In the event of

seizures the study medication doses were escalated per

protocol until the maximum recommended dose was

reached. Maximum recommended doses were defined by a

measured therapeutic level of 20 lg/dl for PHT, and

1500 mg IV BID for LEV. Failure to suppress seizure

activity once at maximum dose resulted in the addition of

PHT to the current LEV dose or addition of LEV to the

current PHT dose. If this regimen did not provide benefit,

treatment with other AEDs was initiated. Any other con-

comitant medication and treatment required as the standard

of care for patient treatment were continued. All concurrent

drugs given and treatments provided were documented,

including dates of administration and reason for use.

Treatment with Study Medications

The PHT group received a loading dose of fos-PHT 20 mg/

kg PE IV, maximum of 2000 mg, given over 60 min and

was then started on a PHT maintenance dose (5 mg/kg/day,

rounded to nearest 100 mg dose, IV every 12 h given over

15 min). PHT serum levels were checked on days 2 and 6

after randomization and dosing was adjusted by the phar-

macist as needed to maintain therapeutic serum levels of

10–20 lg/dl. The LEV group received a loading dose of

20 mg/kg IV, rounded to the nearest 250 mg over 60 min

then started on maintenance dose (1000 mg, IV every 12 h

given over 15 min) as prophylaxis. LEV dose was adjusted

as needed for therapeutic effect up to 1500 mg every 12 h

(3000 mg/day) as maximum dose if seizures occurred.

Patients were maintained on study medications for 7 days.

If there were no seizures at that time (clinical or sub-

clinical), study medication was discontinued. Intravenous

medications were used for the entire 7 days.

Study medication was supplied by the study sponsor

(LEV) or by the investigators (fos-PHT and PHT) and

stored and dispensed by the investigational pharmacy

service at University Hospital. The project coordinator was

notified about the randomization and authorized dispensing

the appropriate medication based on the physician’s order.

The investigational pharmacy service maintained records

of the receipt and distribution of medications used in this

clinical trial to provide drug accountability.

Safety and Efficacy Monitoring

Safety

The primary outcome measure was the incidence of clinical

adverse events. Patients were evaluated daily during the

hospital stay for seizures, fever, neurological changes, car-

diovascular, hematologic and dermatologic abnormalities,

liver failure, renal failure, and death. Each adverse event was

classified by the PI as attributable or possibly attributable to

the study drug versus other adverse events (unlikely related to

the study drug, unrelated to the study drug, or unknown).

Serious adverse events for this study were defined as those that

resulted in death, prolonged hospitalization, life threatening

events, persistent or significant disability, or is an important

medical event that may not be immediately life threatening or

result in death or hospitalization but based upon appropriate

medical judgment may have jeopardized the subject, or may

require medical or surgical intervention to prevent one of the

other outcomes listed in the definitions above.

Efficacy

The secondary endpoints were seizure frequency and long-

term outcomes (seizures, Glasgow Outcomes Scale-

Extended (GOSE), Disability Rating Scale (DRS)). All

patients were monitored on continuous EEG (cEEG) for

72 h or until awake and following commands. Since over

50% of initial seizure activity in these patients consists of

subclinical non-convulsive seizures, as observed in a

number of studies [6, 18, 19] and about 93% of these

seizures occur within the first 2 days of admission to the

ICU, we stopped cEEG as patients awakened, or by 72 h

after admission if there were no seizures.

Outcome Measures

The clinical research coordinator remained blinded to patient

study medication and conducted all outcome assessments.

Data dictionary with explicit, pre-specified data definitions

was used. Neurological outcomes were assessed using the

GOSE and DRS at time of hospital discharge and again at 3

and 6 months after admission. Seizure frequency, any

adverse events, prescribed medications, and a Resource

Utilization Questionnaire were also documented at the 3 and

6 month follow-up.

Neurocrit Care (2010) 12:165–172 167

Statistical Analyses

Initially, the two groups were characterized using descriptive

statistics. Medians and ranges are used for continuous vari-

ables, frequencies, and percentages are used for categorical

variables. Comparisons between groups were based on

Fisher’s Exact tests for categorical variables or a Mann–

Whitney U-test for continuous variables. Generalized linear

models were used to test for differences between groups

adjusted for confounding factors. Statistical analyses were

conducted using SPSS version 17.0 (SPSS Inc., Chicago, IL).

Results

Demographic data of the 52 enrolled patients are presented

in Table 1. A total of 18 patients were enrolled in the PHT

arm and 34 in the LEV arm; and 88.5% were diagnosed

with sTBI. There were no differences in PHT verus LEV

groups in baseline characteristics, including GCS at

admission (4 vs. 5; P = 0.42), GCS at 24 h (3 vs. 3;

P = 0.99), and interventions performed (all P > 0.5).

There were no differences in early seizure occurrence

between the PHT versus LEV groups (3/18 vs. 5/34;

P = 1.0, respectively) or death (4/18 vs. 14/34; P = 0.227).

The patients death were evaluated in detail. An early death

attributed to the injury itself occurred in six patients (PHT 2/

18 vs. LEV 4/34; P = 0.150); in the other cases families

decided to withdraw care early (within 30 days after injury)

in five patients (PHT 0/18 vs. LEV 5/34; P = 1.00) and late

(1–6 months after injury) in seven patients (PHT 2/18 vs.

LEV 5/34; P = 1.00) based on quality of life issues. The

seizures that occurred were all non-convulsive in nature.

The overall duration of PHT treatment was 7 (3–7) days

vs. 7 (1–7) days with LEV (P = 0.969). There were no

differences in PHT versus LEV groups in other short- and

long-term outcomes including GCS at 7 days (6 vs. 7;

P = 0.58) and GOS at discharge (2 vs. 2; P = 0.33),

3 months (3 vs. 3; P = 0.61), and 6 months (3 vs. 3;

P = 0.89; Table 2). There were no differences between

PHT and LEV groups in the occurrence of fever, increased

intracranial pressure (ICP), stroke, hypotension, arrhyth-

mia, thrombocytopenia/coagulation abnormalities, liver

abnormalities, renal abnormalities, or early death (all

P > 0.15). LEV-treated patients experienced worsening

neurological status less frequently (P = 0.024) and had

less gastrointestinal problems (P = 0.043); there was

tendency toward lower incidence of anemia in patients

treated with PHT (P = 0.076). In the PHT group, the mean

PHT serum concentrations were 17.7 mcg/ml on day 2 and

15.2 mcg/ml on day 6.

Tables 3 and 4 show the characteristics of patients who

survived in each study arm, and their outcomes. Overall,

surviving patients treated with LEV experienced better

outcomes than surviving patients treated with PHT

including lower DRS at 3 and 6 months (P = 0.006 and

P = 0.037, respectively) and higher GOSE at 6 months

(P = 0.016). Finally, after adjusting for GCS at admission,

there were no differences in DRS at discharge (P = 0.472),

but at 3 months, the DRS was 5.2 points lower (95%CI

0.2–10.3) among those treated with LEV compared with

Table 1 Characteristics of subjects in the study grouped by study

arm

PHT LEV P value

N = 18 N = 34

Demographics

Age 35 18–80 44 17–75 0.802

Male 13 72.2 26 76.5 0.747

Female 5 27.8 8 23.5

Diagnosis

SAH 2 11.1 4 11.8 1.000

TBI 16 88.9 30 88.2

GCS

On scene

Eyes 1 1–4 1 1–4 0.917

Verbal 1 1–4 1 1–5 0.643

Motor 2 1–6 1 1–6 0.777

Total 4 3–14 5 3–15 0.718

In emergency department

Eyes 1 1–4 1 1–4 0.801

Verbal 1 1–5 1 1–5 0.645

Motor 2 1–6 2 1–6 0.376

Total 4 3–15 5 3–14 0.419

Best in first 24 h

Eyes 1 1–4 2 1–4 0.090

Verbal 1 1–5 1 1–5 0.527

Motor 5 3–6 5 1–6 0.277

Total 8 5–15 9 3–15 0.301

Worst in first 24 h

Eyes 1 1–3 1 1–3 0.221

Verbal 1 1–4 1 1–5 0.449

Motor 1 1–6 1 1–6 0.938

Total 3 3–12 3 3–14 0.991

Interventions

ICP monitor 15 83.3 29 85.3 1.000

Licox 14 77.8 22 64.7 0.529

Craniotomy 6 33.3 14 41.2 0.766

Hematoma evacuation 4 22.2 9 26.5 1.000

Decompression 3 16.7 9 26.5 0.507

Data are given as median and range or frequency and percent.

P values were from Fisher’s Exact tests or Mann–Whitney U-tests as

appropriate; GCS Glasgow Coma Scale, SAH subarachnoid hemor-

rhage, TBI traumatic brain injury, ICP intracranial pressure

168 Neurocrit Care (2010) 12:165–172

those treated with PHT (P = 0.042). At 6 months, the

difference was 3.7 points (95%CI -1.0 to 8.5), but this was

not statistically significant (P = 0.118). The GOSE was

not different at discharge or 3 months, but at 6 months it

was 1.5 points higher for those treated with LEV compared

with those treated with PHT (95%CI 0.1–3.0; P = 0.039).

There was no difference in overall seizure control between

study arms (3/14 vs. 3/20; P = 1.000).

Discussion

This first randomized, single-blinded trial of treatment with

LEV versus PHT in patients with sTBI and/or SAH shows

that LEV is at least as safe as PHT when used for seizure

prevention in the NSICU setting, and that the short- and

long-term outcome measures favor the use of LEV. This

includes significantly better side-effects profile of LEV

(less worsening of neurological status and less gastroin-

testinal problems) when compared to PHT and significantly

improved outcomes at 3 and 6 months including higher

GOSE and lower DRS in the LEV-treated patients.

Phenytoin is an established standard AED in the setting of

acute traumatic brain injury. In fact, the American Academy

of Neurology suggests using PHT for seizure prevention in

the first 7 days after traumatic brain injury [10]. But, the

Table 2 Outcomes and complication data for 52 patients with trau-

matic brain injury or subarachnoid hemorrhage enrolled in the study

PHT LEV P value

N = 18 N = 34

Injury Severity Scale 27 16–50 28 9–50 0.953

GCS, day 7 6 3–15 7 3–15 0.581

GCS, discharge 10 3–15 10 5–5 0.617

GOSE, discharge 2 1–3 2 1–4 0.334

DRS discharge 23 7–30 24 7–30 0.547

GOSE 3 months 3 1–5 3 1–7 0.612

DRS 3 months 13 5–30 15 0–30 0.959

GOSE 6 months 3 1–7 3 1–8 0.892

DRS 6 months 9 0–30 17 0–30 0.787

Fever 10 55.6 18 52.9 1.000

Increased intracranial pressure 8 44.4 13 38.2 0.769

Stroke 3 16.7 7 20.6 1.000

Worsen neurologic status 9 50.0 6 17.6 0.024

Hypotension 2 11.1 7 20.6 0.470

Cardiac arrhythmia 6 33.3 14 41.2 0.766

Anemia 4 22.2 17 50.0 0.076

Platelets low 3 16.7 5 14.7 1.000

Coagulation deficits 1 5.6 2 5.9 1.000

Dermatological 0 0.0 0 0.0 –

Liver function tests 0 0.0 2 5.9 0.538

Renal 1 5.6 2 5.9 1.000

Gastrointestinal 4 22.2 1 2.9 0.043

Early death 2 11.1 4 11.8 0.150

Care withdrawn early 0 0.0 5 14.7 1.000

Care withdrawn late 2 11.1 5 14.7 1.000

Length of stay 15 4–31 14 3–49 0.616

AED duration 7 3–7 7 1–7 0.969

Seizures at follow–up 0 0.0 1 5.3 1.000

AED, 3 months 3 21.4 4 21.1 1.000

AED, 6 months 2 18.2 2 13.3 1.000

GCS Glasgow Coma Scale, GOSE Glasgow Outcomes Scale-Exten-

ded, DRS Disability Rating Scale

Table 3 Characteristics of surviving patients in the two study groups

PHT LEV P value

N = 14 N = 20

Demographics

Age 30 18–80 39 18–66 0.904

Male 10 71.4 16 80.0 0.689

Female 4 28.6 4 20.0

Diagnosis

SAH 2 14.3 2 10.0 1.000

TBI 12 85.7 18 90.0

GCS

On scene

Eyes 1 1–4 1 1–4 0.647

Verbal 1 1–4 1 1–5 0.434

Motor 1 1–6 5 1–6 0.183

Total 2 3–14 7 3–14 0.121

In emergency department

Eyes 1 1–4 1 1–4 0.622

Verbal 1 1–5 1 1–5 0.489

Motor 2 1–6 5 1–6 0.170

Total 4 3–15 7 3–14 0.183

Best in first 24 h

Eyes 1 1–4 3 1–4 0.022

Verbal 1 1–5 1 1–5 0.413

Motor 5 3–6 6 3–6 0.027

Total 7 5–15 10 5–15 0.036

Worst in first 24 h

Eyes 1 1–3 1 1–3 0.872

Verbal 1 1–4 1 1–5 0.928

Motor 2 1–6 2 1–6 0.439

Total 4 3–12 5 3–14 0.341

Interventions

ICP monitor 12 85.7 15 75.0 0.672

Licox 11 78.6 12 60.0 0.295

Craniotomy 5 35.7 7 365.0 1.000

Hematoma evacuation 3 21.4 4 20.0 1.000

Decompression 2 14.3 5 25.0 0.672

Neurocrit Care (2010) 12:165–172 169

availability of newer AEDs questions the use of PHT as the

first line AED in this setting; some even suggest that it may

be reasonable to use LEV instead of PHT for seizure pro-

phylaxis in the setting of intracranial surgery and NSICU

management [13, 20]. LEV has been used in the NSICU

setting for several years now and numerous studies have

reported on oral as well as IV use of this medication for the

treatment or prevention of seizures in this setting. However,

existing studies of LEV safety and efficacy in the NSICU

setting are limited by their methodology, relying either on

retrospective chart review or an open-label design [11–16].

Therefore, previous studies do not provide strong evidence

that LEV affords better short- and long-term outcomes when

compared to other treatments.

In the present study, we compared in a blinded and

randomized fashion these two AEDs to show that treating

patients with LEV in the setting of TBI and/or SAH pro-

vides long-term benefits over PHT. We note that while the

baseline characteristics of both groups including severity of

injury are similar (as demonstrated by similar GCS scores

in the first 24 h and at discharge), two different measures of

long-term outcome, the GOSE and DRS, at 3 and 6 months

favor LEV as the AED of choice in this setting. We

speculate that the better neurological outcomes seen here

may afford patients treated with LEV higher chance of

return to the society as productive members.

LEV is known to potently suppress seizures in animal

models of both, focal and secondary generalized epilepsies

[21–23]. Further, pretreatment with LEV can prevent or

delay the development of kindled seizures [23, 24]. In

addition, LEV has been shown to be neuroprotective in

animal models of brain injury [25, 26]. Because of that

suppressive effect on seizures the results of our trial are not

surprising. Surprising, though, is the fact that we were able

to show results favoring LEV in this setting despite rela-

tively small number of patients enrolled in the trial. Pre-

clinical studies support our findings and raise the possi-

bility that patients at high risk for seizure development may

benefit from prophylactic use of LEV instead of PHT

during periods of acute brain insult, i.e., invasive neuro-

surgical procedures, trauma, stroke, subarachnoid, or

intracerebral hemorrhage. Further, the results of our study

suggest that a randomized, double-blind trial of LEV in this

setting evaluating short- and long-term outcomes is war-

ranted in order to evaluate the neuroprotective and anti-

epileptogenic effects of this AED in humans.

Our single-blinded trial used cEEG for the monitoring

of possible subclinical (covert) seizures in the enrolled

patients. Continuous EEG monitoring has been used to

determine the incidence of early seizures and for prog-

nostication in patients admitted to the NSICU, and is

considered to be the new standard of care in this setting

[2, 27]. Knowledge of the EEG characteristics is known to

affect treatment and predict outcome [28, 29]. In one of

the first studies of EEG utility for the detection of non-

convulsive seizures and status epilepticus in the ICU

setting, Privitera et al. [30] found that out of 198 cases

with altered consciousness but no clinical convulsions, 74

(37%) showed EEG and clinical evidence of definite or

probable nontonic–clonic seizures or status epilepticus. In

a study of critically ill patients admitted to a neurological

ICU setting, 19% of the patients monitored with cEEG

had seizures, of which 92% had no overt clinical signs of

seizure activity; 88% seizures detected in the first 24 and

93% in the first 48 h [18]. Although the incidence of

seizures in our study is somewhat lower (8/52; 15%), this

is likely related to the fact that patients in our study were

selected based on the presence of severe TBI and not

based on the suspicion that they may or may not be

Table 4 Outcomes and complication data for surviving patients only

PHT LEV P value

N = 14 N = 20

Injury Severity Scale 26 16–50 28 9–38 0.439

GCS, day 7 6 3–15 10 4–15 0.138

GCS, discharge 10 3–15 11 6–15 0.396

GOSE, discharge 3 2–3 3 2–4 0.545

DRS discharge 22 7–29 22 7–26 0.436

GOSE 3 months 3 2–5 4 2–7 0.107

DRS 3 months 11 5–23 5 0–23 0.006

GOSE 6 months 3 3–7 5 3–8 0.016

DRS 6 months 6 0–20 3 0–17 0.037

Fever 7 50.0 9 45.0 1.000

Increased intracranial pressure 6 42.9 4 20.0 0.252

Stroke 2 14.3 0 0.0 0.162

Worsen neurologic status 6 42.9 1 5.0 0.012

Hypotension 0 0.0 3 15.0 0.251

Cardiac arrhythmia 3 21.4 8 40.0 0.295

Anemia 3 21.4 8 40.0 0.295

Platelets low 1 7.1 2 10.0 1.000

Coagulation deficits 1 7.1 2 10.0 1.000

Dermatological 0 0.0 0 0.0 –

Liver function tests 0 0.0 0 0.0 –

Renal 0 0.0 0 0.0 –

Gastrointestinal 3 21.4 1 5.0 0.283

Early death 0 0.0 0 0.0 –

Care withdrawn early 0 0.0 0 0.0 –

Care withdrawn late 0 0.0 0 0.0 –

Length of stay 15 4–31 13 7–28 0.545

AED duration 7 3–7 7 6–7 0.602

Seizures at follow-up 0 0.0 1 5.6 1.000

AED, 3 months 3 21.4 3 16.7 1.000

AED, 6 months 2 18.2 2 13.3 1.000

170 Neurocrit Care (2010) 12:165–172

having seizures. Our study is therefore likely more

reflective of clinical practice, where prophylactic treat-

ments are not given on suspicion of seizure occurrence,

but to prevent seizure occurrence among all patients; the

minor discrepancy in the detected seizure incidence is

expected in this instance.

Studies using intermittent or cEEG recordings have

clearly documented high incidence of subclinical seizures

in sTBI/SAH patient population [6, 18, 19, 30]. In

patients with sTBI, studies have confirmed that early

PHT use (up to 7 days post-TBI) reduces the risk of

early seizures (relative risk reduction: 0.37, 95%CI

0.18–0.74); similar short-term effect was observed in a

single trial of carbamazepine [10]. Availability and

administration of IV AEDs is especially relevant for the

critically ill patients with altered levels of consciousness.

Traditionally, medications such as phenytoin, phenobar-

bital, and valproic acid have been used because of well-

defined and easily monitored therapeutic drug levels, and

the availability of the IV formulations. Unfortunately

liver toxicity, hypotension, hematologic abnormalities,

drug interactions, and sedation are only a few of the

many adverse effects of these agents that can become

problematic in critically ill patients. Availability of

newer prophylactic medications that could be easily

initiated with fewer side effects could be beneficial in

this patient population. Because of the high frequency of

clinical and subclinical seizures in this setting, sTBI

presents an ideal human target for further investigations

of AEDs such as LEV in prevention of seizures and

epilepsy and the use of cEEG appears to be indicated in

this setting as many seizures may go unnoticed in the

severely compromised NSICU patients.

Limitations of this study should be noted. While the

cEEG interpretation was provided by a physician blinded

to group assignment, the clinicians managing the patients

were not always blinded to the treatment. Therefore,

although unlikely, we cannot exclude the possibility that

some of the group differences are due to the biases of the

clinicians managing the critically ill patients. Second, the

statistical analyses have not been adjusted for multiple

comparisons. Given the number of statistical tests, it is

possible some of them may have been statistically signifi-

cant purely by chance. While the physiologic plausibility

and magnitude of differences suggest the effects are not

spurious, this possibility cannot be excluded. Finally, only

6-months follow-up data are available. Longer term data

and additional clinical assessments including long-term

seizure outcomes may be needed before replacing PHT

with LEV in this clinical setting can be firmly advocated.

But, until more definitive studies are conducted, we believe

LEV is an acceptable PHT replacement for seizure pro-

phylaxis in patients with TBI and/or SAH.

Conclusions

This single-blinded, randomized study of LEV versus PHT

in the NSICU setting showed that patients treated with

PHT vis-a-vis LEV have the same outcomes in respect to

death or seizures, but LEV results in less undesirable side

effects and better long-term outcomes as measured with

GOSE and DRS. Therefore, LEV may be a suitable alter-

native to PHT in seizure prevention in patients with sTBI

or SAH in the NSICU setting.

Acknowledgments This study was supported by a grant from UCB

Inc., Principal Investigator: Lori A. Shutter, MD. This study was

presented in part at the Neurocritical Care Society Meetings in 2008

and 2009. Data Safety Monitoring Board included Drs. Andrew

Ringer, MD (Department of Neurosurgery) and Michael D. Privitera,

MD (Department of Neurology).

Disclosure of Conflicts of Interest Jerzy P. Szaflarski, MD, PhD

has received grant support from the American Academy of Neurol-

ogy, Davis Phinney Foundation/Sunflower Revolution, National

Institutes of Health, UCB Pharma Inc., and The University Research

Council at the University of Cincinnati. He has served as a paid

consultant and/or speaker for Abbott Laboratories, American Acad-

emy of Neurology, Pfizer and UCB, Inc. Kiranpal S. Sangha,

Pharm.D—has nothing to disclose. Christopher J. Lindsell, PhD—has

received grant support from Abbott POC. Lori A. Shutter, MD has

received grant support for the Department of Defense, National

Institute of Health, and UCB Pharma, Inc. She has served as a paid

consultant and/or speaker for Integra Lifesciences and the Brain

Trauma Foundation.

References

1. Hesdorffer D. Risk factors. In: Engel J, Pedley T, editors. Epi-

lepsy: a comprehensive textbook. Philadelphia: Wolters Kluwer,

Lippincott–Raven Publishers; 2007. p. 57–63.

2. Claassen J, Hirsch LJ, Frontera JA, et al. Prognostic significance

of continuous EEG monitoring in patients with poor-grade sub-

arachnoid hemorrhage. Neurocrit Care. 2006;4:103–12.

3. Claassen J, Jette N, Chum F, et al. Electrographic seizures and

periodic discharges after intracerebral hemorrhage. Neurology.

2007;69:1356–65.

4. Lowenstein DH. Epilepsy after head injury: an overview. Epi-

lepsia. 2009;50(Suppl 2):4–9.

5. Szaflarski JP, Rackley AY, Kleindorfer DO, et al. Incidence of

seizures in the acute phase of stroke: A population-based study.

Epilepsia. 2008;49(6):974–81.

6. Vespa PM, Nuwer MR, Nenov V, et al. Increased incidence and

impact of nonconvulsive and convulsive seizures after traumatic

brain injury as detected by continuous electroencephalographic

monitoring. J Neurosurg. 1999;91:750–60.

7. Kilpatrick CJ, Davis SM, Hopper JL, Rossiter SC. Early seizures

after acute stroke. Risk of late seizures. Arch Neurol. 1992;49:

509–11.

8. So EL, Annegers JF, Hauser WA, O’Brien PC, Whisnant JP.

Population-based study of seizure disorders after cerebral

infarction. Neurology. 1996;46:350–5.

9. Temkin NR. Preventing and treating posttraumatic seizures: the

human experience. Epilepsia. 2009;50(Suppl 2):10–3.

Neurocrit Care (2010) 12:165–172 171

10. Chang BS, Lowenstein DH. Practice parameter: antiepileptic

drug prophylaxis in severe traumatic brain injury: report of the

Quality Standards Subcommittee of the American Academy of

Neurology. Neurology. 2003;60:10–6.

11. Jones KE, Puccio AM, Harshman KJ, et al. Levetiracetam versus

phenytoin for seizure prophylaxis in severe traumatic brain

injury. Neurosurg Focus. 2008;25:E3.

12. Michaelides C, Thibert R, Shapiro M, Kinirons P, John T,

Manchharam D, et al. Tolerability and dosing experience of

intravenous levetiracetam in children and infants. Epilepsy Res.

2008;81:143–7.

13. Milligan T, Hurwitz S, Bromfield E. Efficacy and tolerability of

levetiracetam versus phenytoin after supratentorial neurosurgery.

Neurology. 2008;71:665–9.

14. Patel NC, Landan IR, Levin J, Szaflarski J, Wilner AN. The use

of levetiracetam in refractory status epilepticus. Seizure. 2006;15:

137–41.

15. Ruegg S, Naegelin Y, Hardmeier M, Winkler DT, Marsch S, Fuhr

P. Intravenous levetiracetam: treatment experience with the first

50 critically ill patients. Epilepsy Behav. 2008;12:477–80.

16. Szaflarski JP, Meckler JM, Szaflarski M, Shutter LA, Privitera

MD, Yates SL. Levetiracetam use in critically ill patients.

Neurocrit Care. 2007;7:140–7.

17. Bratton SL, Chestnut RM, Ghajar J, et al. Guidelines for the

management of severe traumatic brain injury. J Neurotrauma.

2007;24(Suppl 1):S7–95.

18. Claassen J, Mayer SA, Kowalski RG, Emerson RG, Hirsch LJ.

Detection of electrographic seizures with continuous EEG mon-

itoring in critically ill patients. Neurology. 2004;62:1743–8.

19. Rhoney DH, Tipps LB, Murry KR, Basham MC, Michael DB,

Coplin WM. Anticonvulsant prophylaxis and timing of seizures

after aneurysmal subarachnoid hemorrhage. Neurology. 2000;55:

258–65.

20. Fountain N. Should levetiracetam replace phenytoin for seizure

prophylaxis after neurosurgery? Epilepsy Curr. 2009;9:71–2.

21. Klitgaard H, Matagne A, Gobert J, Wulfert E. Evidence for a

unique profile of levetiracetam in rodent models of seizures and

epilepsy. Eur J Pharmacol. 1998;353:191–206.

22. Loscher W, Honack D, Rundfeldt C. Antiepileptogenic effects of

the novel anticonvulsant levetiracetam (ucb L059) in the kindling

model of temporal lobe epilepsy. J Pharmacol Exp Ther. 1998;

284:474–9.

23. Loscher W, Reissmuller E, Ebert U. Anticonvulsant efficacy of

gabapentin and levetiracetam in phenytoin-resistant kindled rats.

Epilepsy Res. 2000;40:63–77.

24. Stratton SC, Large CH, Cox B, Davies G, Hagan RM. Effects of

lamotrigine and levetiracetam on seizure development in a rat

amygdala kindling model. Epilepsy Res. 2003;53:95–106.

25. Hanon E, Klitgaard H. Neuroprotective properties of the novel

antiepileptic drug levetiracetam in the rat middle cerebral artery

occlusion model of focal cerebral ischemia. Seizure. 2001;10:

287–93.

26. Marini H, Costa C, Passaniti M, et al. Levetiracetam protects

against kainic acid-induced toxicity. Life Sci. 2004;74:1253–

64.

27. Vespa P. Continuous electroencephalography for subarachnoid

hemorrhage has come of age. Neurocrit Care. 2006;4:99–100.

28. Kilbride RD, Costello DJ, Chiappa KH. How seizure detection by

continuous electroencephalographic monitoring affects the pre-

scribing of antiepileptic medications. Arch Neurol. 2009;66:723–8.

29. Orta DS, Chiappa KH, Quiroz AZ, Costello DJ, Cole AJ. Prog-

nostic implications of periodic epileptiform discharges. Arch

Neurol. 2009;66:985–91.

30. Privitera M, Hoffman M, Moore JL, Jester D. EEG detection of

nontonic–clonic status epilepticus in patients with altered con-

sciousness. Epilepsy Res. 1994;18:155–66.

172 Neurocrit Care (2010) 12:165–172