Synopsis

In this work we acquired the first z-spectrum from the human liver in vivo at 7T. Glycogen and NOE peaks were observed, and their evolution over time was monitored after fasting and after a meal. Both the glycogen peak and the NOE peak at -1.7ppm were observed to increase 2-4 hours after a high carbohydrate meal.

Purpose

Glycogen is the primary short term energy store found predominantly in liver and muscle and plays a key role in glucose metabolism in health and disease (e.g. diabetes)¹. It can be studied with non-localised natural abundance ¹³C MRS² but experiments are time consuming and difficult particularly in patients with significant adipose tissue. Z-spectrum CEST imaging offers an alternative approach to map glycogen³ and has the potential to improve signal strength and reduce scan time compared to ¹³C MRS, without multi-nuclear equipment.

Previous studies have shown that glycoCEST is sensitive to in vivo changes in glycogen stores in humans at 3T when using MTR_asym analysis⁴, however a full z-spectrum will give additional information, and aid in identifying respiratory artefacts which greatly impact the signal in the z-spectrum in abdominal imaging. Furthermore imaging at 7T provides considerably enhanced sensitivity to CEST and NOE effects.

Aim

To observe changes to the z-spectrum in the liver at 7T in response to feeding.

Methods

4 subjects (age=24,26,36,55,2F) were fasted for +16 hours before being scanned on a 7T Philips Achieva system using a multi-transmit system using a 8TX/32RX fractionated dipole body array (MRCoils, Zaltbommel, Netherlands). Z-spectra were acquired at 64 off-resonance frequencies between ±100,000Hz using Semi-CW acquisition^5,6 (TFEPI readout, 3s saturation, B_1,max=1μT, single slice, acquisition time=10mins/spectrum). Each dynamic was acquired during a breath hold of 4.7s, with 4.3s spacing for the subject to breathe. Volume B₀ shimming was performed over the liver and RF shimming (phase nulling) was performed on a small target ROI using in-house software to maximise the B1+ in the ROI.

After the initial fasted scan, subjects were offered ad-libitum food, high in carbohydrates. They were then scanned at two hour intervals. One subject was then scanned 3 times over 30mins to assess repeatability.

The target ROI in the liver was masked, and spectra were averaged and B₀ corrected pixel-wise using in-house MATLAB scripts. After initial analysis, two peaks were quantified by measuring the area under a region in the spectrum: the expected glycogen peak between +0.8ppm and +1.8ppm, and an NOE peak between -2ppm and -1.17ppm.

A glycogen phantom was also scanned using a NOVA 32ch pTx head coil to confirm the expected peaks.

Results

Figure 1 shows the spectrum from the glycogen phantom at (a) 1.8μT, and (b) 0.3μT, which closely matched the saturation power achieved from the body coil. It appears that there is a second exchanging pool at +3.2ppm, which is not visible at lower powers. This is comparable to the signal reported from glucose by van Zijl³, suggesting there may have been some glucose present in the phantom.

Figure 2 shows that repeatability was good across the spectrum, although the peak assigned to glycogen increased over the course of 20mins, possibly due to glycogen storage.

Figure 3 shows the spectra acquired from 4 subjects. The general trend appears to be that the glycogen levels rise after two hours, and begin to fall after 4 hours, also illustrated in figure 4. Figure 5 shows that the NOE peak visible at -1.7ppm rises most after 4 hours.

Discussion

This is the first z-spectrum acquired in the liver at 7T. Figure 2 suggests multiple z-spectrum features that have been observed in other tissues which might be of interest, e.g. MT was evident on all spectra, which might be relevant in monitoring liver fibrosis.

We observed and quantified a glycogen peak which varied over subjects and with time after a meal. It was unexpected that the glycogen levels would drop back down to baseline after 4 hours, particularly since subjects were encouraged to keep eating. However there was variation in subject age, size, and weight, and future studies will use a more controlled feeding protocol and blood glucose monitoring.

The NOE peak observed at -1.7ppm is a source of great interest. It has previously seen to be visible in blood⁷, and so may be linked to increased blood flow into the liver. It also emphasizes that such data should not be quantified using asymmetry analysis.

Figure 3 shows the importance of acquiring the data breath-held. Noisy spectra were observed when subjects admitted to have fallen asleep, and indeed motion could be seen on the images from these time-points. This accounts for the apparent drop in glycogen after 4 hours in subject 4, where movement was observed on the first two spectra, altering the acquired signal. Future work will consider improved methods of respiratory control and monitoring.

Conclusion

Z-spectrum imaging performed in the liver at 7T yields quantifiable changes in peaks consistent with glycogen and also in the NOE signal located at -1.7ppm. Both of these signals rose two and four hours after a meal respectively.

Acknowledgements

References

1 Kalderon, B., Gopher, A., & Lapidot, A. (1986). Metabolic pathways leading to liver glycogen repletion in vivo, studied by GC‐MS and NMR. FEBS letters, 204(1), 29-32.

2 Sillerud, L. O., & Shulman, R. G. (1983). Structure and metabolism of mammalian liver glycogen monitored by carbon-13 nuclear magnetic resonance. Biochemistry, 22(5), 1087-1094.

3 Van Zijl, P. C., Jones, C. K., Ren, J., Malloy, C. R., & Sherry, A. D. (2007). MRI detection of glycogen in vivo by using chemical exchange saturation transfer imaging (glycoCEST). Proceedings of the National Academy of Sciences, 104(11), 4359-4364.

4 Deng, M., Chen, S. Z., Yuan, J., Chan, Q., Zhou, J., & Wáng, Y. X. J. (2016). Chemical exchange saturation transfer (CEST) MR technique for liver imaging at 3.0 Tesla: an evaluation of different offset number and an after-meal and over-night-fast comparison. Molecular Imaging and Biology, 18(2), 274-282.

5 Hoogduin H, Khlebnikov V, Keupp J, et al (2017) Semi continuous wave CEST with alternating sets of 4 transmit channels at 7T. MAGMA 30:S1–S152.

6 Keupp J, Baltes C, Harvey PR, Brink J van den (2011) Parallel RF Transmission based MRI Technique for Highly Sensitive Detection of Amide Proton Transfer in the Human Brain at 3T. Proc Intl Soc Mag Reson Med 19:710.

7 Shah, S. M., Mougin, O. E., Carradus, A. J., Geades, N., Dury, R., Morley, W., & Gowland, P. A. (2018). The z-spectrum from human blood at 7T. NeuroImage, 167, 31-40.