NNS 2015 Program & Abstract Book

following injury. Controls groups received an IV administration of saline.

DCE-MRI analysis indicated that there is a significant increase in GI

vascular permeability in injured animals compared to uninjured animals

at 48 hours after injury. We observed a significant decrease in GI vascular

permeability following a single IV injection of Ang-1 (300

g) compared

to saline treated animals. In addition, Ang-1 treatment produced a

qualitative improvement in GI morphological outcome. SCI produced

disruption in GI villi compared to naive, uninjured control mice in H&E

stained GI tract sections. Ang-1 (300

g) treated animals exhibited re-

duced GI villi damage compared to vehicle-treated subjects. Taken to-

gether the data suggests that promoting vascular stability following SCI

by an IV administration of Ang-1 not only reduces GI vascular perme-

ability but also appears to preserve intestinal villi.

Keywords: gastrointestinal, dynamic contrast enhanced imaging,

vascular permeability

S12 Open Communication: Clinical

S12-01

A PROGNOSTIC MODEL FOR DETERMINING ONE-

MONTH OUTCOMES IN MILD TRAUMATIC BRAIN INJURY

Hayley Falk

, HeadSMART Investigators

Johns Hopkins University School of Medicine, Emergency Medicine,

Baltimore, USA

There are currently no tools for aiding emergency physicians in edu-

cating mild traumatic brain injury (mTBI) patients regarding the

prognosis of their injury. We sought to derive a model for identifying

mTBI patients at risk for incomplete recovery from their symptoms at

1-month after injury. We analyzed data from a prospective cohort of

TBI patients presenting to an urban emergency department (ED) (The

Head

injury

erum

arkers for

ssessing

esponse to

rauma

[HeadSMART] cohort). Subjects presenting within 24 hours of injury

were interviewed on the day of injury. Telephone interviews were

performed at 1-month after injury to determine TBI outcomes. In-

complete recovery was defined as Glasgow Outcome Scale Extended

(GOSE)

8. Prognostic models were built using univariable and mul-

tivariable logistic regression methods and stepwise selection proce-

dures. A total of 194 subjects were enrolled between April 2014 and

February 2015. Of this number, 108 were mTBI patients with a pre-

senting Glasgow Coma Scale (GCS) of 14 or 15; a negative head CT

scan; and 1-month follow-up data were included in this analysis. Within

this subpopulation, 52.8% (57/108) had GOSE

8 at 1 month. Predictor

variables included in the final prognostic model were altered mental

status at presentation (AMS), gender, race (African-American or non-

African American), work-related injury, dangerous injury mechanism

(ejection from motor vehicle, pedestrian struck, fall from height

3ft or

5 stairs), and other injury (solid organ injury or bony fracture). This

model discriminated between subjects with and without incomplete

recovery with an area under the receiver operator curve (AUC) of 0.82

(95% CI: 0.73 – 0.88). A HeadSMART30 score was computed by

assigning a score of 2 for AMS, 1 for female gender, 2 for African-

American, 2 for work-related injury, 1 for dangerous mechanism and 1

for other injury. Subjects with a HeadSMART30

7; 7 and 8; 9 and 10;

and greater than 10 had a 1-month risk of incomplete recovery of 0%,

27%, 71% and 86% respectively. This study provides preliminary ev-

idence that prognostication of mTBI outcome using readily available

clinical and demographic data is feasible.

Keywords: prognostic models, GOSE, TBI

S12-02

AN INITIAL EVALUATION OF THE NINDS PHENOTYPING

COMMON DATA ELEMENTS FOR TRAUMATIC BRAIN

INJURY

John Dsurney

, Shannon McNally

, Andre van der Merwe

, Leighton

Chan

1,2

National Institutes of Health, Clinical Center, Bethesda, USA

Center for Neuroscience and Regenerative Medicine, Phenotyping

Core, Rockville, USA

Introduction:

In 2010, the NIH and other federal agencies identified a

list of Common Data Elements (CDE) that might be used in traumatic

brain injury (TBI) research. The selection of these instruments was not

empirically based, but was guided by their availability in the public

domain, the availability of alternate forms, and, most importantly,

expert opinion regarding the utility of the tests. The present study

undertakes an empirical examination of these instruments, comparing

their performance to other tests.

Methods:

A total of 111 (62% male) subjects who had sustained a

closed head injury were seen at 30, 90, 180 and/or 365 days post

injury. Subjects were administered eight of the ten original ‘‘core’’

neuropsychological CDE’s, two ‘‘core’’ CDE’s were replaced by

equivalent ‘‘supplemental’’ measures. Subjects were also adminis-

tered seven additional ‘‘supplemental’’ CDE’s, as well as additional

well validated, commonly used neuropsychological tests. The per-

centage of individuals classified as ‘‘impaired’’ (scoring less than one

standard deviation below the mean) was calculated for each time point

and by severity.

Results:

Our cohort included 60 mild, 33 moderate, and 18 severe

patients with TBI. Of the original CDE’s, the Trail Making Tests

(TMT) A and B and California Verbal Learning Test (CVLT-2) were

the most sensitive, identifying impairment regardless of patient se-

verity or time since injury. Other original CDE tests, such the Wis-

consin Card Sort did not perform as well. In addition, some other tests,

such as the Booklet Category Test, Sea Shore Rhythm Test, Con-

trolled Oral Word Association Test, Grooved Pegboard, and Finger

Tapping Tests consistently identified impairment, outperformed the

original CDE’s.

Conclusions:

Only some of the current CDE’s were useful in our

cohort. These included: TMT A and B, BSI and CVLT-2. In addition,

a number of tests not included as original CDE’s demonstrated sen-

sitivity in the evaluation of TBI subjects and are recommended for use

in this population.

Keywords: Common Data Elements, Neuropsychological, Out-

comes, Phenotyping

S12-03

ADOLESCENT TRAUMATIC BRAIN INJURY INCREASES

ALCOHOL CONSUMPTION AND REWARD IN FEMALE

MICE

Zachary Weil

, Kate Karelina

, Kristopher Gaier

, Timothy

Corrigan

, John Corrigan

Ohio State University Wexner Medical Center, Department of Neu-

roscience, Columbus, USA

Ohio State University Wexner Medical Center, Department of Phy-

sical Medicine and Rehabilitation, Columbus, USA

Traumatic brain injury (TBI) is inextricably and bidirectionally linked

with alcohol use. Some estimates implicate alcohol intoxication in one-

third to one-half of all TBI cases. Alcohol use following injury can

A-147