Quantitative pH using chemical exchange saturation transfer and phosphorous spectroscopy
Zhuozhi Dai1,2, Phillip Zhe Sun3, Gang Xiao4, Gen Yan2, Yanlong Jia2, Zhiwei Shen5, Alan H. Wilman1, and Renhua Wu2,5

1Biomedical engineering, University of Alberta, Edmonton, AB, Canada, 2Medical Imaging, 2nd Affiliated Hospital of Shantou University Medical College, Shantou, China, People's Republic of, 3Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, MGH and Harvard Medical School, Charlestown, MA, United States, 4Math and Information Technology, Hanshan Normal University, Chaozhou, China, People's Republic of, 5Provincial key laboratory of medical molecular imaging, Shantou, China, People's Republic of

Synopsis

pH is a very important biochemical property that changes in many pathological states. Monitoring pH is of significance in early diagnosis and treatment therapy. However, there is lack of non-invasive methods to image pH in vivo effectively. Our study quantified the pH value using a chemical exchange saturation transfer (CEST) method in rat brain and established a strong correlation of pH between CEST and 31P-MRS in vivo, Pearson correlation factor is 0.819, P < 0.01. Because CEST imaging has superior spatial resolution to 31P-MRS, CEST may provide an alternative, straightforward, and effective way to obtain quantitative pH images in vivo.

Purpose

pH is a very important biochemical property that changes in many pathological states 1-4, such as tumor, stroke etc. Monitoring pH is of significance in early diagnosis and treatment therapy. However, there is lack of non-invasive methods to image pH in vivo effectively. The purpose of our study was to quantify the pH value using a chemical exchange saturation transfer (CEST) method in rat brain and to establish a correlation to pH measurement using phosphorous spectroscopy (31P-MRS).

Materials and Methods

MRI: Data were collected at 7.0 T on an animal MRI system using a standard 1H body coil and 31P surface coil for RF pulse transmitting and receiving. Image sequences included Z-spectra, T1 fitting, T2 fitting and 31P-MRS. For CEST, we used a continuous wave as the presaturated component, saturation time=3 s, saturation power=0.75 μT, 1 μT and 2 μT respectively, amino proton frequency offset=3.5ppm, reference offset=-3.5 ppm. For Z-spectra, we serially altered the frequency offset from -5.6 ppm to 5.6 ppm, in steps of 0.2 ppm. We used a Point Resolved Spectroscopy (PRESS) sequence as a readout component to obtain the Z-spectra. Seven water spectra signals were acquired for T1 measurement using a PRESS sequence with different TR, ranging from 1.1 s to 6 s. Six water spectra signals were obtained for T2 measurement with TEs from 6.5 to 70 ms. After 1H-imaging, the body coil was detuned and 31P-MRS were tuned to obtained the spectra in the same voxel.

Animal models: Brains of five normal adult male Sprague-Dawley (SD) rats were scanned in vivo. The animals were euthanized with overdose of Phenobarbital by intravenous injection after the first scanning to create whole brain stroke models, which were scanned immediately in the first 3 hours after stroke.

Analysis: All data were post-processed in Matlab and SPSS. Quantitative pH was calculated by fitting Z spectra and by 31P-MRS. Both the standard two pools quantifying exchange rates in chemical exchange saturation transfer agents using saturation power dependencies (QUESP) and a homemade Nuclear Overhauser Effect including constrained nonlinear multivariable (NICN) fitting methods were using to calculate the quantitative pH value. 31P-MRS used the chemical shift difference between pH-dependent inorganic phosphate and phosphocreatine to calculate the pH value. Assessment of the correlation between CEST pH and 31P-MRS pH was performed via Pearson correlation analysis. Means and standard variance in each method were calculated and analysed with one-way ANOVA to compare the mean differences, p < 0.01 was considered as statistically significant.

Results and discussion

The quantitative pH values from CEST effect and from 31P-MRS are shown in Figure 1. Statistical results indicated strong correlation, Pearson correlation factor is 0.819, P < 0.01 (Fig. 2). Moreover, AVONA results show F = 23.2, which was significant (P < 0.01), the effect size is 81.2%. The pH difference between normal brain tissues and stroke brain lesions was significant (P < 0.01). The difference between CEST and 31P-MRS methods was insignificant (P = 0.945 for stroke and P = 0.901 for normal tissues) (Fig. 3). Compare the two CEST fitting methods, the standard two pools QUESP method could not fit the in vivo data well because of the existing NOE effect, while our NICN method had a superior fitting both at CEST effect and NOE sides of the spectrum (Fig. 4).

Conclusion

Our study established a strong correlation of pH between CEST and 31P-MRS in vivo. Because CEST imaging has superior spatial resolution to 31P-MRS, CEST may provide an alternative, straightforward, and effective way to obtain quantitative pH images in vivo.

Acknowledgements

This study was funded in part by NSFC- 81471730 (RH. Wu),

MSF of Guangdong Province-B2013281 (ZZ. Dai).

References

1. Zhou, J., Payen, J.-F., Wilson, D. A., Traystman, R. J. & van Zijl, P. C. Using the amide proton signals of intracellular proteins and peptides to detect pH effects in MRI. Nature medicine 9, 1085-1090 (2003).

2. Zhuozhi Dai, Yanlong Jia, Gen Yan, Fei Duan, Gang Xiao, Zhiwei Shen, Hongfu Sun, Alan H. Wilman, and Renhua Wu. pH-weighted imaging in diabetes mellitus suffering acute cerebral ischemic stroke. ISMRM 2015.

3. Sun PZ, Lu J, Wu Y, Xiao G, Wu R. Evaluation of the dependence of CEST-EPI measurement on repetition time, RF irradiation duty cycle and imaging flip angle for enhanced pH sensitivity. Physics in Medicine and Biology. 2013;58(17):N229-N40. doi: 10.1088/0031-9155/58/17/n229.

4. Chen LQ, Pagel MD. Evaluating pH in the Extracellular Tumor Microenvironment Using CEST MRI and Other Imaging Methods. Advances in Radiology. 2015;2015.

Figures

Figure 1. Quantitative pH values from CEST effect and from 31P-MRS.


Figure 2. Correlation between CEST and from 31P-MRS, Pearson correlation factor is 0.819, P < 0.01.

Figure 3. The pH difference between normal brain tissues and stroke brain lesions was significant (P < 0.001). The difference between CEST and 31P-MRS methods was insignificant (P = 0.945 for stroke and P = 0.901 for normal tissues).

Figure 4. The standard two pools QUESP method (left) could not fit the in vivo data well because of the existing NOE effect, while our NICN method (right) had a superior fitting both at CEST effect and NOE sides.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4088