Page 64 - NNS 2014 Program & Abstract Book

Funding: NIDRR-H133A120087, DOD-W81XWH-071-0701, NIH-
R01HD048162, NIH-R01NS048178, NIH-R01NS049142.
Key words
depression, limbic systems, neuroimaging, traumatic brain injury
A2-21
CEREBROVASCULAR DYSFUNCTION AFTER TRAUMATIC
BRAIN INJURY: ASSESSMENT WITH MRI AND NIRS
Diaz-Arrastia, R.1–3, Moore, C.B.1–3, Amyot, F.1,2,4, Kenney, K.1–3,
Shenouda, C.1,5, Gandjbakhche, A.1,4, Wasserman, E.M.2,3
1Center for Neuroscience and Regenerative Medicine, Bethesda, USA
2USUHS, Bethesda, USA
3NINDS, Bethesda, USA
4NICHD, Bethesda, USA
5NIH Clinical Center, Bethesda, USA
Injury to small and medium-sized cerebral blood vessels is a recognized
consequence of traumatic brain injury (TBI). Clinical trials of therapies
directed at diffuse vascular injury (DVI) will require quantitative, non-
invasive methods to measure cerebrovascular reactivity (CVR), to se-
lect patients likely to benefit and to demonstrate physiologic efficacy in
early phase clinical trials. Measurement of the Blood Oxygen Level
Dependent (BOLD) response to hypercapnia, using magnetic resonance
imaging (MRI) or Near Infrared Spectroscopy (NIRS) shows promise
as a biomarker for DVI-directed theapies.
23 patients with moderate to severe TBI were studied during the
chronic stage ( > 6 months) after injury. 12 age and gender matched
uninjured controls were also studied. Region-specific CVR was as-
sessed using a hypercapnia challenge. Inhaled gas alternated between
room air and 5% carbon dioxide (CO2) mixed with 21% O2/74% N2
every minute in a block design. Hemodynamic response function was
acquired for 7 minutes using both MRI (whole brain) and NIRS
(frontal regions) in two separate sessions. End-tidal (EtCO 2) was
monitored continuously during the experiment using a capnograph
and used as a regressor in the analysis. Mean (SD) CVR measured
with BOLD-MRI was 0.31 (0.05) %BOLD/mmHg for the controls,
and 0.25 (0.12) %BOLD/mmHg for TBI subjects (p = 0.03). With
NIRS, CVR was 2.7 (0.3) lM/mmHg for controls and 1.7 (0.6) lM/
mmHg for TBI subjects (p = 0.01). Within subjects, CVR measured
using BOLD-MRI correlated highly with CVR measured with NIRS
(r = 0.83, p < 0.0001).
We conclude that measurement of CVR with either MRI or fNIRS
is practical and reliable. Patients in the chronic stage after TBI have
blunted and highly variable CVR. CVR measurements with either
method show promise as a biomarker for DVI-directed therapies.
Key words
BOLD-MRI, cerebrovascular reactivity, near-infrared
A2-22
CORRELATIVE MR AND PET-FDG IMAGING OF LESION
AND CONTRALATERAL CORTEX IN CONTROLLED COR-
TICAL IMPACT BRAIN INJURY
Shukla, D.K.1, Korotcov, A.1, Bosomtwi, A.1, Mathur, S.1, Wilson, C.1,
Jaiswal, S.1 , Broussard, E.1 , Jones, S.1,2, Byrnes, K.R.1,3, Selwyn, R.1,2
1Center for Neuroscience and Regenerative Medicine, Bethesda, USA
2Radiology, Uniformed Services University (USU), Bethesda, USA
3Anatomy, Physiology, and Genetics, USU, Bethesda, USA
Traumatic brain injury (TBI) diagnosis is primarily dependent on
computed tomography (CT) to indicate injury severity. However,
other modalities such as magnetic resonance imaging (MRI) and
positron emission tomography (PET) may provide improved sensi-
tivity and specificity for TBI. In the present study, imaging parameters
from MRI and PET were utilized to assess lesion and cortical prop-
erties after a controlled cortical impact (CCI) brain injury in a rat.
Male Sprague Dawley rats (N = 24) received a moderate CCI and
underwent T2-weighted, diffusion-weighted (DW), arterial spin la-
beling (ASL) MRI and 18F-fluorodeoxyglucose (FDG)-PET imaging
at baseline and at multiple time-points post-injury. Lesion volumes
were measured using a semi-automated algorithm and compared to
manually determined volumes on the images and on histology. T2-
values, apparent diffusion coefficient (ADC), cerebral blood flow
(CBF) and FDG uptake were obtained for lesion and contralateral
cortex. Lesion volume measurements were significantly correlated for
the algorithm, manual, and histological methods (ICC = 0.92). This
result demonstrates that T2 maps with semi-automated methods can
be used for an objective determination of TBI lesion volume. Un-
coupling between CBF (decrease) and FDG uptake (increase) was
detected in the lesion at days 7–8 along with vasogenic edema marked
by elevated T2 and ADC values. Uncoupling was also detected in the
contralateral cortex at day 11 but with cytotoxic edema indicated by
elevated T2 values and decreased ADC. In addition, cytotoxic edema
was also measured at day 2 in the contralateral cortex, possibly
serving as a marker for injury severity. Taken together, lesion volume,
edema (T2, ADC), and uncoupling (CBF, FDG) measurements may
provide researchers with a multiparametric model to stage or grade
injury severity and to possibly serve as a prognostic indicator of
outcomes.
Key words
arterial spin labeling, contralateral cortex, controlled cortical injury,
diffusion-weighted imaging, lesion volume, PET imaging, T2-weighted
imaging
A2-23
BASELINE ALTERATIONS OF MEDIAL TEMPORAL LOBE
SUBSTRUCTURES IN CONTACT-SPORT ATHLETES
Parekh, M., Mitchell, L., Han, N., Parivash, S., Douglas, D., Wilson,
E., Zeineh, M.
Stanford University, Stanford, USA
There is a high incidence rate of mild traumatic brain injury in contact-
sport athletes, such as football players. Understanding the relationship
between traumatic forces and brain injury may enable the prevention
of permanent cognitive deficits, and possibly even disorders such as
chronic traumatic encephalopathy (CTE). Neuropathologically, CTE
is characterized as a tauopathy with atrophy of the many portions of
the brains, including the medial temporal lobes (MTL). Our objective
was to determine if there were differences at baseline in MTL
structures in football players that likely have a long history of con-
cussive and sub-concussive forces compared to athletes who play a
non-contact sport such as volleyball. At the beginning of their season
and across two seasons, 47 football and 21 volleyball players were
scanned on a GE 3.0T whole-body scanner. Imaging sequences in-
cluded a 1.0x1.0x1.0 mm T1 volume as well as a 0.4x0.4x2 mm
oblique coronal T2-weighted FSE volume. We used the software
Automatic Segmentation of Hippocampal Subfields (ASHS) to seg-
ment the hippocampus and parahippocampal sub-regions. We blindly
edited the segmentations to (1) ensure exclusion of the amygdala, (2)
extend out hippocampal body/head subfields to encompass the entire
A-32