Transcript Disclosures/ Conflicts of Interest

EE-41: New Multiparametric MRI
Quantification of the Trigeminal Spinal
Pathway after Gamma Knife Radiosurgery
in a Patient with Trigeminal Neuralgia
Authors
Michael Hoch, Kwan Chen, Amparo Wolf, Sohae Chung, Noam Ben-Eliezer,
Douglas Kondziolka, and Timothy Shepherd
Department of Neuroradiology and Neurosurgery
New York University Langone Medical Center
ASNR 54th Annual Meeting
May 23 - 26, 2016
Disclosures/ Conflicts of Interest:
No conflicts of interest from authors
Introduction – Trigeminal Nerve




Trigeminal nerve arises from the ventrolateral mid-pons and enters the
skull base through Meckel’s cave.
Peripheral components consist of motor and sensory branches. Motor
branches are found in the mandibular division and emerge from the motor
nucleus within the pons. Sensory information of touch and position from all
three divisions are conveyed by sensory afferents to the chief sensory
nucleus.
Pain and temperature descend the brainstem within the trigeminal spinal
tract to reach the trigeminal spinal nucleus within the medulla.
Second order axons decussate and ascend the brainstem within the
trigeminothalamic tract before terminating in the ventral posteromedial
nucleus of the thalamus.
Introduction – Trigeminal Neuralgia




Chronic neuropathic facial pain caused by trigeminal neuralgia (TN)
(aka tic douloureux) can be extremely debilitating. TN is most
commonly caused by neurovascular compression of the nerve root.
Medical management has numerous side effects, including
drowsiness, dizziness, headaches, aplastic anemia, and
agranulocytosis.
Surgical therapies include microvascular decompression and
ablative therapies such as radiosurgery.
Gamma Knife radiosurgery offers pain relief in a minimally-invasive
manner on an outpatient basis.
Problem Statement

There are no accepted imaging methods that correlate
GKRS treatment effectiveness with objective imaging
findings.

Gadolinium based MRI can identify the treated nerve, but
cannot predict outcomes.

CISS and other T2-weighted 3D gradient echo sequences are
not quantitative.
Objectives



Utilize novel MRI sequences and post-processing techniques, Track Density Imaging
and quantitative EMC T2 / Proton Density maps, to objectively measure imaging
findings after gamma knife radiosurgery
Track Density Imaging

Utilize simultaneous multislice (SMS) diffusion acquisition

Resolve / determine white matter fiber tractography

Generates super-resolution images where pixel intensity reflects probablistic
streamlines that traverses the voxel.

Color maps represent fiber orientation in conventional fashion.

Apparent Fiber Density provides direct measure of fiber density not subject to
fiber-tracking biases and errors.
Echo modulation curve (EMC) software package developed in-house enables
accurate T2 and proton density values (PD). More can be found at http://cai2r.net
Case Presentation





86yo, R dominant male presents with 5 year hx of classic R trigeminal neuralgia (V3
distribution). He had treatments of gamma knife radiosurgery 5 years ago (single
4mm isocenter, 80 Gy to the 100% isodose line) was located at the dorsal root entry
zone of the right trigeminal nerve. 1 year ago, he had another GKRS at the same
location (4mm isocenter, 50 Gy to the 100% isodose line). 9% isodose line
approached the surface of the brainstem. He describes the pain as “electrical
shocks” that has progressed to steadier aching. Pain lasts seconds to minutes,
located along R V3 distribution and R tongue. He also reports R V3 facial numbness.
No hx of multiple sclerosis, brain tumor, or facial trauma.
Medications include Neurontin, Trileptal, Baclofen, and periodic Oxycodone.
Physical exam unremarkable.
A third GKRS treatment for TN was planned. Patient underwent our 3T MRI
brainstem protocol for TDI and EMC T2 mapping.
Case Presentation
At the level of the mid pons there is
neurovascular compression of the
right trigeminal nerve (arrow) within
the cerebellar pontine angle cistern
by a dolichoectatic basilar artery on
CISS imaging. This is the cause of
the patient’s right trigeminal
neuralgia.
Case Presentation – TX plan 3rd GKRS
Methods – GKRS

Perfexion Gamma Knife

Leksell Model G stereotactic frame

Single 4mm isocenter placed along nerve, anterior to basilar artery.
Maximum dose 65 Gy at the 100% isodose line. 8% isodose line
approached the brainstem surface.
Methods – Image Acquisition
•
Patient underwent contrast enhanced MRI with 20-channel head and neck
coil (3T, Siemens Skyra).
•
Quantitative T2 mapping multiecho spin echo with 15 echos at 10ms
intervals (TE=10-150ms, TR 5100ms, 384x234 matrix, iPAT 2, 2mm slice
thickness, 2 averages, 6 minutes).
•
High angular resolution diffusion-weighed imaging (HARDI) for TDI
parameters: 3mm isotropic, 80x80 matrix, SMS 2-slice acceleration, 256
directions, b=2500 s/mm2, 8 b=0 images, TR/TE=3816/98ms, 50 slices,
phase encoding in AP direction (14 mins).
• For susceptibility-induced distortion correction, one image b=0 with
phase encoding in PA direction.
•
Other sequences include 2mm dual-echo T2, 5mm axial T2, FLAIR, SWI,
axial 0.5mm isotropic T2 SPACE.
Methods - Postprocessing
•
Quantitative T2 maps: Maps are generated from an echo modulation
curve T2 fitting software developed in-house. The matlab and C++
scripts can be found on http://cai2r.net.
•
Track Density Imaging: DWI images first corrected for susceptibility
with FSL (FMRIB software library). Constrained spherical
deconvolution was used to model multiple fiber-orientations. Wholebrain probablistic fiber-tracking was performed with 10 million seeds.
Super-resolution maps were generated with 500-um isotropic
resolution. Spherical Deconvolution Informed Filtering of
Tractograms (SIFT2) was applied to show biologically accurate
measures of fiber connectivity while using complete streamline
reconstruction.
Methods - Quantification
•
MRI Data Quantification: Regions of interest (ROI) were manually
drawn over trigeminal nerve root, spinal tract at the level of medulla,
spinal nucleus at the level of the pons, and spinothalamic tract at the
level of midbrain.
•
ROIs were placed on quantitative T2, PD and TDI maps. Apparent
Fiber Density (AFD) values were obtained from the diffusion data.
•
Diseased right side was compared to the normal left.
Results – EMC T2 and PD
Mid Pons Level: EMC Mapping demonstrates increased (A) PD and (B)
T2 values within the right trigeminal spinal tract compared to the left.
No appreciable change is seen in the T2 MSE image (C).
Results – EMC T2 and PD
EMC Mapping demonstrates increased T2 and PD values within the
right trigeminal nerve, spinal tract, and nucleus compared to the left. No
change is seen in the Trigeminothalamic tract.
Results – EMC T2 and PD
Track Density Imaging &
Direction-Encoded Color TDI
demonstrate decreased
streamlines within the right
trigeminal nerve (arrows) and
spinal tract (ovals) at the level
of the mid pons (A&B).
At the rostral medulla there
are decreased streamlines in
the right trigeminal spinal
nucleus (circles in C&D).
Results – EMC T2 and PD
AFD values were decreased
in the R trigeminal nerve and
spinal tract.
The values were similar in
both the spinal nucleus and
trigeminothalamic tracts.
Greatest difference was seen
in the trigeminal nerve.
Discussion




Simultaneous multislice (SMS) acquisition has enabled us to
shorten data acquisition times of high angular DWI mapping
sequences to clinically feasible scan times.
Track Density Imaging post-processing tool is based on high
angular resolution diffusion acquisitions. While other
algorithms, namely DTI, are available, TDI generates superresolution images derived from whole-brain probablistic
tractography. These contrasts have been histologically
validated in animal models. We can quantify fiber density from
TDI by using the apparent fiber density (AFD) measurement.
EMC T2 mapping models allows accurate quantification of T2
and PD values in clinically feasible time scales.
These measures enable us to further understand anatomical
changes that occur after GKRS.
References
Andersson JLR, Skare S, Ashburner J. How to correct susceptibility distortions in spin-echo echo-planar images: application to
diffusion tensor imaging. Neuroimage [Internet]. 2003 Oct;20(2):870–88. Available from:
http://www.ncbi.nlm.nih.gov/pubmed/14568458
Ben-Eliezer N, Sodickson DK, and Block KT. Rapid and accurate T2 mapping from multi-spin-echo data using Bloch-simulationbased reconstruction. Magn Reson Med 2015; 73(2): 809-817.
Ben-Eliezer N, Sodickson DK, Shepherd T, Wiggins G, Block KT. Accelerated and motion-robust in vivo T2 mapping from radially
undersampled data using Bloch-simulation-based iterative reconstruction. Magn Reson Med 2015 Apr 17. doi: 10.1002mrm.25558.
Burchiel KJ. A new classification for facial pain. Neurosurgery 2003; 53: 1164-1166.
Calamante F, Tournier JD, Jackson GD, Connelly A. Track-density imaging (TDI): Super-resolution white matter imaging using
whole-brain track-density mapping. Neuroimage. 2010;53(4):1233–43.
Calamante F, Tournier JD, Heidemann RM, Anwander A, Jackson GD, Connelly A. Track density imaging (TDI): Validation of super
resolution property. Neuroimage. 2011;56(3):1259–66.
Kondziolka, Douglas, et al. "Stereotactic radiosurgery for trigeminal neuralgia: a multi institutional study using the gamma unit."
Journal of neurosurgery 84.6 (1996): 940-945.
Setsompop, Kawin, et al. "Blipped‐controlled aliasing in parallel imaging for simultaneous multislice echo planar imaging with reduced
g‐factor penalty." Magnetic Resonance in Medicine 67.5 (2012): 1210-1224.
Tournier JD, Calamante F, Connelly A. Robust determination of the fibre orientation distribution in diffusion MRI: Non-negativity
constrained super-resolved spherical deconvolution. Neuroimage. 2007;35(4):1459–72.
Thank You!