5016
Comparison between glucoCEST at 3T and Blood Glucose Sampling in HumansAnina Seidemo1, Anna Rydhög2, Ronnie Wirestam1, Xiang Xu3,4, Akansha Ashvani Sehgal3,4, Yi Zhang5, Frederik Testud6, Pia C Sundgren2,7,8, Peter C van Zijl3,4, and Linda Knutsson1,3
1Department of Medical Radiation Physics, Lund University, Lund, Sweden, 2Centre for Medical Imaging and Physiology, Skåne University Hospital, Lund and Malmö, Sweden, 3Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, MD, United States, 4F.M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, United States, 5College of Biomedical Engineering and Instrument Science, Zhejiang University, Hangzhou, China, 6Siemens Healthcare AB, Malmö, Sweden, 7Institution of Clinical Sciences/Diagnostic Radiology, Lund University, Lund, Sweden, 8Lund University Bioimaging Center, Lund University, Lund, Sweden
Synopsis
Dynamic glucose-enhanced (DGE) MRI uses chemical exchange saturation transfer (CEST) to retrieve information about microcirculation using D-glucose as a contrast agent.We performed glucoCEST imaging in four healthy volunteers at 3T and compared arterial input functions (AIFs) and DGE signal in white matter to measured venous blood glucose levels after glucose infusion. An increase in DGE signal following the glucose infusion was observed in cerebral arteries, but not in white matter. We observed a similarity in shape of the AIF and the blood glucose curve, and that a higher blood glucose level corresponds to a higher DGE signal.
Introduction
Dynamic glucose-enhanced (DGE) MRI can provide tumor enhancement and information about microvasculature using chemical exchange saturation transfer (CEST) with D-glucose as a biodegradable contrast agent 1, referred to as glucoCEST. The hydroxyl protons in glucose are saturated using a frequency-selective RF pulse and exchange subsequently with non-saturated water protons, leading to a lowering of the water signal where glucose is present. The CEST-effect benefits from higher field strengths 2, but DGE imaging has been performed also at 3T 3. Since reports are accumulating of adverse effects using gadolinium 4,5, safe clinical alternatives are desirable. We performed in-vivo glucoCEST imaging at 3T to investigate the clinical potential of DGE MRI by comparing dynamic DGE MRI curves in cerebral arteries and in white matter to venous blood glucose levels (BGL) measured by blood gas analysis. Since arterial concentration curves (i.e. arterial input functions, AIFs) are often required to extract different perfusion-related parameters it would be of interest to examine how the measured AIFs correspond to the sampled venous BGLs.Methods
Four consenting fasting volunteers were scanned on a 3T MAGNETOM Prisma (Siemens Healthcare, Erlangen, Germany) using a prototype CEST sequence with turbo spin echo readout. One axial slice with in-plane resolution 1.7x1.7 mm2 and slice thickness 5 mm was imaged with TR/TE = 6000/7.1 ms. Saturation was achieved using 10 Gaussian-shaped pulses (B1=1.6 $$$\mu$$$T) of 100 ms duration and 10 ms interpulse delay at a single saturation offset of 2 ppm from the water resonance frequency. Acquisition of 126 glucoCEST images was accomplished at a temporal resolution of 6 seconds and with a total scan time of less than 13 minutes, with intravenous administration of 50 ml of D-glucose in one arm after 3 minutes of imaging. The infusion duration was 90 seconds for three subjects and 80 seconds for one subject. Venous blood samples were collected in the contralateral arm before and after the glucose infusion and during the glucoCEST data acquisition, and the BGL was measured using a blood gas analyzer. The baseline BGL for each participant was subtracted from the measured BGL to obtain relative BGLs. Retrospective motion correction was performed using Elastix 6. DGE images were calculated from the signal difference between each glucoCEST image and the averaged pre-infusion images normalized to a non-saturated image (S0). A positive DGE-signal can therefore be interpreted as an increased glucose concentration in the voxel. Dynamic response curves were calculated in arteries and in white matter as the average of 4-8 voxels. DGE area-under-curve (AUC) images were calculated as the sum of DGE image signal values over two-minute intervals.Results and Discussion
An increase in DGE signal following the glucose infusion was observed in all subjects in several cerebral arteries. The signal change in white matter was small compared to the arterial signal and sometimes negative. The arterial DGE signal change over the glucoCEST image acquisition time is shown together with the signal in white matter and the venous BGL in figure 1. Figure 2 shows arterial DGE signal as a function of BGL for all participants. The relationship between DGE signal and BGL varies between individuals, but the general trend is a higher DGE-signal for a higher BGL. A linear regression analysis resulted in R2=0.55. One pre-infusion and one post-infusion DGE AUC map are presented in figure 3. Both hypo- and hyperintense areas can be observed in the post-infusion AUC map. Vessels show mainly hyperintensity while cerebrospinal fluid is often characterized by negative DGE signal. The DGE signal is sensitive to partial-volume effects and to motions such as head movements and to physiological response to the administered glucose in terms of ventricular volumetric changes and vessel dilation or restriction.The relationship between the BGL in a peripheral vein and the AIF is complicated and depends on insulin response and glucose transport over the blood-brain barrier, but also on the characteristics of the CEST-sequence/effect (origin of signal, acquisition, T2-relaxation). However, the similarity in shape of the AIFs and the venous blood glucose curve suggests that the blood glucose curves canserve as a guideline to how a given individual responds to the glucose infusion, and therefore support the evaluation of the DGE-signal and help differentiate DGE-signal from movements. The difference in DGE signal between blood and brain tissue may be useful to guide the interpretation in tumor imaging.Conclusion
AIFs can be measured in DGE images obtained at 3T, and a higher venous BGL did generally correspond to a higher DGE signal. Insignificant or negative response to glucose infusion was observed in white matter.Acknowledgements
No acknowledgement found.References
1. Xu X, et al. Dynamic Glucose-Enhanced (DGE) MRI: Translation to Human Scanning and First Results in Glioma Patients. Tomography. 2015;1(2):105-114.
2. van Zijl PCM, Yadav NN. Chemical exchange saturation transfer (CEST): What is in a name and what isn’t? Magnetic Resonance in Medicine. 2011;65:927-48.
3. Xu X, et al. Dynamic Glucose Enhanced Imaging at 3T: First Human Data. Proceedings of the 25th Annual Meeting of the ISMRM, 2017, Honolulu, p.0193.
4. Thomsen HS, et al. Nephrogenic systemic fibrosis and gadolinium-based contrast media: updated ESUR Contrast Medium Safety Committee guidelines. European radiology. 2013;23(2):307-318.
5. Gulani V, et al. Gadolinium Deposition in the Brain: Summary of Evidence and Recommendations. The Lancet Neurology. 2017;16(7):564-70.
6. Klein S, et al. Elastix: a toolbox for intensity based medical image registration. IEEE Trans Med Imaging. 2010;29(1):196–205.
Figures
Figure 1. DGE dynamic response curves in arteries, white matter and measured venous blood glucose level for four healthy volunteers (a-d). The blue area represents the glucose infusion duration. Each subject shows an individual response to the glucose infusion, which can be observed both in the AIFs and in the venous-blood glucose curves. The reason for the large signal fluctuations in the AIF for subject (a) is probably motion.
Figure 2. Arterial DGE signal and corresponding relative blood glucose level for four healthy volunteers.
Figure 3. DGE AUC maps in one subject for a two-minute interval before and after glucose infusion. A change in the DGE signal relative to the baseline is visible after glucose infusion.