92 / 198
after sTBI. Subjects 18–75 yrs with sTBI were assessed for genetic
relationships with PTS. Genetic variants within neuronal glutamate
transporters, SCL1A1 and SCL1A6, were screened for 253 individ-
uals. Kaplan-Meier curves, compared with log-rank statistics, were
used to estimate seizure rates from admission to 3 yrs and 2 days post-
injury through 3 yrs post-injury for SNPs by genotype. Cox propor-
tional hazards regression was used to estimate hazard ratios for SNPs
significant in Kaplan-Meier, adjusting for known risk factors. 32 SNPs
were examined (SLC1A1: n
=
28, SLC1A6: n
=
4). 49 (19.37%) sub-
jects had PTS. Of these, 18 (36.73%) subjects seized within 7 days,
and 31 (63.27%) seized 8 days-3 yrs post-TBI. Correcting for multiple
comparisons, rs10974620 (SLC1A1) was significantly associated with
time to first seizure across the 3 yr follow-up (seizure rates: 77.1%
minor allele homozygous, 24.8% heterozygous, 16.6% major allele
homozygous; p
=
0.001). When beginning PTS follow-up on day 2,
rs7858819 (SLC1A1) was significantly associated with PTS risk
(seizure rates: 52.7% minor allele homozygous, 11.8% heterozygous,
21.1% major allele homozygous; p
=
0.002). After adjusting for cov-
ariates, rs10974620 but not rs7858819, remained significant
(rs10974620: p
=
0.018; HR: 5.24, 95%CI: 1.67–16.48, minor allele
versus major allele homozygous). Genetic variation within SCL1A1,
specifically rs1094620 and rs7858819, is associated with epilepto-
genesis following sTBI, as demonstrated by time-to-event analyses.
Future studies are needed to confirm findings, but variation within
neuronal glutamate transporter genes may represent a possible ther-
apeutic target for pharmacological PTS prevention.
Keywords: rehabilomics, epileptogenesis, genetic variation, neu-
ronal transmitters, glutamate, TBI
B4-03
COMPARISONOF FACTORS PREDICTINGPOST-TRAUMATIC
SEIZURE AT 1, 2, & 5 YEARS POST-INJURY: A TBIMS
ANALYSIS
Anne Ritter
1,2
, Amy Wagner
1,3,4
, TBI-MS PTS Writing Group
2
1
Univ Pittsburgh, Epidemiology, Pittsburgh, USA
2
Univ Pittsburgh, Physical Medicine/Rehab, Pittsburgh, USA
3
Univ Pittsburgh, Safar Center, Pittsburgh, USA
4
Univ Pittsburgh, Neuroscience, Pittsburgh, USA
Post-traumatic seizures (PTS) occur frequently following traumatic
brain injury (TBI). Published PTS prevalence rates depend on study
characteristics, and there is limited information regarding long-term
differences in PTS risk factors over recovery. Therefore, we devel-
oped logistic regression models, describing the most important fac-
tors related to PTS at 1, 2, & 5 years post-TBI. Data were collected
from the TBI Model Systems National Database, a multi-center
longitudinal study examining long-term recovery and outcomes after
TBI. PTS prevalence at each time-point was calculated. Baseline
history and injury characteristics were selected
a priori
from risk
factors identified in previous studies and biological plausibility.
Subjects with missing data were excluded at individual time-points.
For each model: Univariate logistic regression for each variable of
interest was performed; effect size and significance were calculated.
Variables were ranked by level of significance, and those with
p
>
0.20 excluded. A main effects model was then fit. Forward and
backward regression were then performed, and variables excluded if
p
>
0.10 in forward, and if p
>
0.05 in backward regression. All
models controlled for injury year. Fit was assessed using Hosmer-
Lemeshow Goodness-of-Fit Test. At years 1, 2, & 5: 4,236, 2,991,
and 1,176 subjects had complete data. At each follow-up, acute
hospitalization seizure (all p
<
0.0001; OR: 3.37, 3.28, 5.13) and
craniectomy (all p
<
0.0001; OR: 2.86; 2.19; 2.51) were the most
significant predictors. All models include craniotomy, penetrating
TBI, and subdural hematoma. Additional significant variables in-
clude: (year-1) post-traumatic amnesia duration and pre-injury in-
carceration, (year-2) contusion load, race, and pre-injury cognitive
limitation, (year-5) contusion load and race. Some PTS risk variables
are dynamic, while the most significant remain consistent across the
follow-up period. Some variables (e.g. incarceration) may be proxies
for latent risk factors; others (e.g. acute-care seizure, craniectomy)
could inform risk associated with clinical practice, justify prophy-
lactic therapies development research, and support experimental
PTS model development.
Keywords: Rehabilomics, TBI, population risk factors, prediction
modeling, PTS
B4-04
EFFECTS OF LEVETIRACETAM AND GABAPENTIN COM-
BINATION THERAPY ON POST-TRAUMATIC NON-
CONVULSIVE SEIZURES (NCS) INDUCED BY A
PENETRATIN
Xi-Chun May Lu
1
, Ronald Tallarida
2
, Ying Cao
1
, Zhinlin Liao
1
,
Deborah Shear
1
, Frank Tortella
1
1
Walter Reed Army Institute of Research, Brain Trauma Neuropro-
tection & Neurorestoration/Psychiatry and Neuroscience, Silver
Spring, USA
2
Temple University, Dept. Pharmacology, Philadelphia, USA
When tested as monotherapies, levetiracetam (LEV) and gabapentin
(GBP) showed different anti-seizure dose-response profiles against
nonconvulsive seizures (NCS) induced by penetrating ballistic-like
brain injury (PBBI) in rats. LEV (12.5–100 mg/kg) reduced NCS
frequency and duration in a dose-dependent fashion whereas the
anti-seizure effects of GBP appeared to plateau across the dose-
range tested (12.5–25 mg/kg). The current study tested a series of
fixed-dose ratios of the LEV
+
GBP combination to determine if
combining these two drugs would produce additive or synergistic
effects. All rats received frontal PBBI, immediately followed by
continuous EEG monitoring for 72 h. LEV
+
GBP treatment (LEV/
GBP: 6.3/0.62, 12.6/1.25, 25/2.5, 50.7/5.0, 101.4/10.0 mg/kg) was
administered intravenously twice/day for three days, initiated 30 min
post-injury. Control animals received matching vehicle treatments.
Compared to vehicle-treated group, LEV
+
GPB combination ther-
apy reduced PBBI-induced NCS incidence from 69% (vehicle
group) to 27–65% and delayed NCS onset latency from 12.3h (ve-
hicle group) to 19.1–57.8h across all LEV
+
GBP treated groups.
Among the five dose ratios tested, the most significant anti-seizure
effects were afforded by the three highest dose ratios (25/2.5, 50.7/
5.0, and 101.4/10.0 mg/kg) as evidenced by the dose-dependent re-
duction in NCS frequency (34%, 45%, and 64%), and shortened
NCS duration (44%, 59% and 58%), respectively (p
<
0.05 vs. ve-
hicle treatment for each measure). However, the dose equivalence
analysis indicated that the observed anti-seizure effects of the
LEV
+
GBP combination failed to achieve additivity or synergism.
Consequently, these findings showed that a LEV
+
GBP combination
therapy replicated the dose-response profile of LEV monotherapy,
but did not benefit from the addition of GBP to improve their anti-
seizure activities. Hence, any advantage in simultaneous usage of
these two drugs appears limited.
This research was funded by the Army Combat Casualty Care
Research Program.
Keywords: Combination Therapy, Penetrating brain Injury, Leve-
tiracetam, Gabapentin, Isobolographic analysis
A-56