Tissue Characterization with Fast High Resolution Magic Angle Spinning (HRMAS) CEST Spectroscopy
Iris Yuwen Zhou1, Taylor Fuss1,2, Gang Xiao3, Takahiro Igarashi1, Lin Li1, Leo L. Cheng1,2, and Phillip Zhe Sun1

1Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Charlestown, MA, United States, 2Department of Pathology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, United States, 3Department of Mathematics and Statistics, Hanshan Normal University, Chaozhou, China, People's Republic of

Synopsis

Z-spectrum is conventionally acquired through multiple experiments with selective saturation at different frequency offsets of interest, leading to extreme long acquisition time. Here, we employ gradient-encoding to substantially accelerate the acquisition of Z-spectrum. This speedup in combination with higher spectral resolution provided by high resolution magic angle spinning (HRMAS) allows rapid quantification of chemical exchange rates of CEST agents, monitoring dynamic processes and fast tissue characterization. The approach was validated in phantom and used for characterization of brain tissues after ischemic stroke.

Purpose

CEST- or Z-spectrum is often obtained to characterize CEST agents and to quantify information related to molecule or microenvironment such as temperature and pH1,2. It is conventionally acquired through multiple experiments with selective saturation at different frequency offsets of interest, leading to extreme long acquisition time3. Here, we employ gradient-encoding to substantially accelerate the acquisition of Z-spectrum4. The utilization of high resolution magic angle spinning (HRMAS) further increases the spectral resolution for better identification and analysis of the spectrum.

Methods

Copper sulfate-doped (0.65 mM) phosphate-buffered saline with choline (10 mM) and varied concentrations of creatine (10, 20, 30 40 and 50 mM) in 1.5% agarose gel was used for phantom study. Middle cerebral artery occlusion (MCAO) was performed on six adult Wistar rats to induce ischemia stroke. 24 hours after MCAO, brain tissue samples from contralateral normal area or ipsilateral ischemia lesion were harvested, loaded into 4 mm Zirconia rotors with 1.0 μl of D2O and introduced into the HRMAS probe. Spectroscopic measurements of ex vivo tissue were carried out at 37 °C and at a HRMAS spinning rate of 4800 Hz on a 14.1T Bruker AVANCE spectrometer (Bruker BioSpin, Billerica, Massachusetts, USA). Z-spectra were acquired using a spin-echo sequence without (B1=0 µT) or with (B1=1, 1.5, 2, 3 µT) RF saturation (Figure 1). By applying a constant magnetic field gradient during the saturation period, the off-resonant data points in the Z-spectrum are generated by a gradient-induced change of the Larmor frequencies of the nuclei in the sample, such that they experience saturation with different off-resonance conditions depending on their position5.

Results

Representative HRMAS Z-spectra and corresponding CEST asymmetry spectra from the phantom study showed strong CEST signal from creatine at the offset of 1.8 ppm for different B1 power levels (Figure 2). The optimal B1 level can be found at 1.5 µT (Figure 3a). Figure 3b shows that the CESTR calculated from CEST asymmetry spectra at B1=1.5 µT increases linearly with creatine concentration. Figure 4 illustrates HRMAS Z-spectra of tissue samples from normal area or ischemic lesion of stroke brains. CEST effects can be observed at multiple offsets such as -3.5, -2.5, 2, 3.5 and 4.5 ppm, etc. and significant differences between the normal and lesion tissue samples can be found (Figure 4).

Discussion and Conclusion

Compared to conventional approach, this new method substantially reduce acquisition time by encoding the frequency offsets along one spatial dimension. It does not require the sample to have a homogeneous shape as it can be compensated for by normalizing to the 1D projection of the sample acquired with saturation off. This speedup in combination with higher spectral resolution provided by HRMAS allows rapid quantification of chemical exchange rates of CEST agents, monitoring dynamic processes and fast tissue characterization.

Acknowledgements

The study was supported in part by grants from NIH/1R01NS083654 and PHS NIH grants CA115746.

References

[1] van Zijl PC, et al. Magn. Reson. Med. 2011;65:927-48.

[2] Liu D, et al. Magn. Reson. Med. 2013;70:1070-81.

[3] Zaiss M, et al. Phys. Med. Biol. 2013;58:R221-69.

[4] Xu X, et al. Angew. Chem. Int. Ed. Engl. 2013;52:8281-4.

[5] Boutin C, et al. J. Phys. Chem. Lett. 2013;4:4172-76.

Figures

Figure 1 Conventional CEST spectroscopy is achieved through repeating multiple experiments with RF saturation at different frequency offsets. Fast HRMAS CEST spectroscopy utilizes gradient-encoding to substantially accelerate the acquisition with spin-echo for signal acquisition and HRMAS providing better spectral resolution. G1 is the gradient applied during saturation, G2 is the gradient applied during acquisition. Both gradients are applied along Z.

Figure 2 Gradient encoded HRMAS Z-spectra from a gel phantom containing 30 mM creatine and 10 mM choline at a spinning rate of 3600 Hz, acquired without (I0, top) or with (I, middle) saturation pulse at varied power levels. Z-spectra (bottom) were obtained by normalizing I to I0 and CEST asymmetry spectra were shown.

Figure 3 (a) CESTR (=(I--I+)/I0) at offset of 1.8ppm was found to be strongest at B1=1.5 µT. (b) CESTR as a function of creatine concentration.

Figure 4 HRMAS Z-spectra of tissue samples from contralateral normal area (red) or ipsilateral ischemic lesion (blue) at B1=1, 1.5, 2 µT and 37 °C and a spinning rate of 4800 Hz. Mean ± SEM presented. The difference between the normal area and ischemia lesion were compared with Student's t-test.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
2902