Tsang-Wei Tu1, Georgios Z Papadakis2, Zsofia Kovacs2, William Reid2, Dima Hammoud 2, and Joseph Frank2
1Radiology and Imaging Sciences, National Institutes of Health, Bethesda, MD, United States, 2Radiology and Imaging Sciences, National Institutes of Health
Synopsis
The current study investigated the
CEST sensitivity and specificity for simultaneous glucose (glucoCEST) and
glutamate (gluCEST) detection on a phantom and a rodent brain injury. Using a
median saturation power (2μT) and short saturation duration (2s),
the MTRasym
at 1.2, 2.1 and 2.9ppm could detect the glucose content, while the MTRasym at 3ppm could measure the
glutamate in vivo. The glucoCEST
results were compared in parallel to the 18F-FDG PET detecting the
tissue uptake of glucose, while the gluCEST data were shown comparable to 18F-DPA
PET radio-ligands binding to translocator protein (TSPO) for imaging microglial
activation in the injured brain.
PURPOSE
The
molecular CEST imaging previously has been shown sensitive to the glucose and
glutamate by measuring the exchangeable protons from the hydroxyl and amine
group respectively for targeting glucose metabolism and inflammation
1.
The current study investigated the
feasibility of CEST imaging for a simultaneous detection on the proton transfer
for glucose (glucoCEST) and glutamate
(gluCEST). The
CEST imaging parameters were optimized on a test phantom consisting of glucose
(Glc) and glutamate (Glu), together with other concentrated molecules in the
brain: myo-inositol (MI), lactate (Lac), N-Acetylaspartic acid (NAA), and a
mixture of them to test the sensitivity and specificity of glucose and
glutamate detection by CEST. A preliminary in vivo experiment was performed on
a rat brain with injury and compared to the
18F-FDG positron
emission tomography (PET) detecting the glucose uptake and
18F-DPA
for imaging microglial activation in vivo.
METHODS
The
phantom experiments were done using a Doty radiofrequency quadrature coil (Doty
Scientific, Inc., SC) on a Bruker 9.4T scanner (Bruker Corp., MA). The CEST
data were acquired by 2D Rapid Acquisition with Relaxation Enhancement (RARE)
sequence with (SMT) and without (S0) magnetization
transfer (MT) preparation pulses: TR 3.10s, TE 10.39ms; RARE factor 4; in plane
resolution 200μm2, slice thickness 0.8mm. The effect of varying MT
preparation pulses on the sensitivity and specificity for glucose detection
were investigated by arraying the saturation power from 0.5μT to 7μT and saturation
duration from 0.5s to 3s. The CEST Z-spectrum were acquired from the MT
frequency offsets (Δω) from −2.0kHz to +2.0kHz with 50Hz frequency stepping, as
a result, 81 points were sampled to delineate the entire Z-Spectrum from -5
part per million (ppm) to +5ppm. CEST data were corrected for B0/B1 field inhomogeneity
using WASSR water referencing2: 0.2 μT for saturation power and 0.1s
for saturation duration, Δω −0.4kHz to +0.4kHz with 10Hz frequency stepping. The
asymmetry of magnetization transfer ratio (MTRasym) maps were derived for
glucoCEST (1.2, 2.1, 2.9ppm) and gluCEST (3ppm) maps by calculating the area
under the curve at the related chemical shifts. A focal injured rat brain was
scanned using the same imaging parameters with the optimized MT saturation
pulses (2μT, 2s). 18F-FDG and 18F-DPA PET were performed
24 and 48 hours after MRI scans, respectively, using Siemens micro-PET/CT scanner.
Standardized uptake values were calculated for the PET data. RESULTS
The MTRasym of the Glc phantom by saturation
power 2μT, and duration 2s showed peaks at 1.2, 2.1, 2.9ppm, and that of Glu
phantom exhibited a broad coverage of peaks from 1 to 3.5 ppm (Fig. 1). The
glucoCEST map showed the most contrast to Glc, while the gluCEST displayed
predominant contrast to the Glu. The CEST Z-spectrum and MTRasym of the rat
brain showed different patterns between the normal and injured tissue (Fig. 2).
The injured tissue comparing to the normal tissue exhibited decreased MTRasym at
the hydroxyl chemical shifts, where the increased MTRasym was seen for the
amine chemical shift range at 3ppm. Comparable results are seen in the
assessment of glucose uptake between the glucoCEST and 18F-FDG PET
(Fig. 3A), and for the inflammation markers between gluCEST and 18F-DPA
PET (Fig. 3B).DISCUSSION
CEST
imaging has been previously shown sensitive to detect glucose uptake in tumor3
and neuroinflammation in Alzheimer's disease4 by selectively mapping
the chemical shifts of the hydroxyl and amine group. In the current study, we
further demonstrated the possibility to simultaneously detecting the glucose
and glutamate content by optimized CEST acquisition parameters to show the
decreased glucose metabolism and increased inflammation at the same time. The
current finding suggests that the CEST technique has strong potentials for the
simultaneous detection of the glucose and glutamate, without using radioactive
isotopes, for repetitively monitoring the metabolism and inflammation in the
TBI. Acknowledgements
No acknowledgement found.References
[1]
van Zijl, MRM, 2011;
[2] Kim, MRM, 2009;
[3] Walker-Samuel, Nat.Med, 2013;
[4] Haris,
NMR Biomed, 2013