Introduction

Cerebrovascular alterations, one of the risk factors for Alzheimer’s disease (AD) pathogenesis, is classified as early change in AD progression. Aging, the main risk factor for Alzheimer's disease (AD), slows the velocity of cerebral blood flow. It progressively induces CBF reduction affected by cerebrovascular impairment and is related to the AD pathology cascade. The sluggish blood delivery to the tissue can facilitate β-amyloid (Aβ) accumulation and inhibit Aβ clearance. Since the brain attempts to maintain CBF homeostasis by regulating the caliber of cerebral arteries and arterioles¹, in the early stage of cerebrovascular alterations, cerebral arterial blood volume (CBVa) could be increased to compensate for slow blood. Detailed characterizations of early cerebrovascular changes – which to date has been limited to CBF measurement - could potentially provide a better understanding of AD pathophysiology.

Methods

A total of nine subjects, five healthy controls (age: 72.4 ± 12.1 years; 1/4 fe/males) and four MCI subjects (age: 79.2 ± 12.5 years; 4 males), were studied on a 3T system using a 32-channel head coil. The pulsed LL-ASL technique measures the evolution of ASL signal by acquiring data at multiple inversion times (TIs) after a single spin labeling pulse. Data (GE-EPI) were acquired with 15 readout steps after spin labeling with the time interval between TIs = 259 ms using multiband acquisition technique (acceleration factor = 5) with and without bipolar gradients alternating to separate the arterial component (b value = 3 s/mm²) ². A FOV/3 shift using CAIPIRINHA method was incorporated to improve the de-aliasing efficiency of multiband reconstruction ³. The imaging parameters were: voxel size = 3.69 x 3.69 x 4.0 mm³, TR/TE = 4 s/31 ms, FA = 30°, and 40 averages. An anatomical image (MPRAGE) was acquired and used for brain parcellation with Freesurfer (http://surfer.nmr.mgh.harvard.edu) to generate the ROIs. After motion correction, the ASL signals with and without the bipolar gradients were averaged for each ROI. The partial volume correction was performed for each voxel of the ASL images using GM, WM and CSF masks from structural images ⁴. ASL models were fit to these averaged ASL signals using a least-squares fitting algorithm ⁵. A modified Gaussian model was used to incorporate the dispersion effects for arterial blood and tissue ⁵. The perturbation effects of multiple LL-excitation pulses, which enhances arterial blood signal by suppression of tissue, was also incorporated into the ASL model.

Results

Fig. 1A and 1B shows visual differences in the ASL dynamics maps for arterial blood (above) and tissue (below) signal between MCI and controls with 259ms of temporal resolution for whole brain. The evolution of arterial blood delivery and the progression of tissue perfusion for whole GM and two ROIs (precuneus and anterior cingulate) are shown from 5 controls and 4 MCI patients (Fig. 1C). For the whole GM regions, the metrics for ASL signal dynamics in the controls versus patients were CBF = 60.3 ± 23.5 vs. 43.1 ± 17.8 ml/100g/min; transit time to artery = 1.24 ± 0.1 vs. 1.43 ± 0.19 s; transit time to capillary = 2.65 ± 0.94 vs. 3.25 ± 1.14 s (p < 0.05); and CBVa = 1.15 ± 0.30 vs. 0.49 ± 0.30 ml/100g. The precuneus region is more delayed than anterior cingulate area (Fig. 1C). Fig.2 shows the averaged metrics of ASL signal dynamics for multiple ROIs. Overall, the MCI subjects had slower delivery of arterial blood, smaller CBVa, prolonged capillary transit times, and smaller CBF in comparison to control subjects.

Discussion

Our measured longer blood transit time in the brains of MCI patients confirms and expands those findings from transcranial Doppler studies that found significantly lower blood velocity in the carotid arteries of AD patients, compared with healthy aged-matched controls 6. In the early stage of cerebrovascular alterations, CBVa increased to compensate for slow blood flow to maintain adequate supply of CBF (in units of ml/g/sec, i.e. blood volume x velocity) ². We observed this phenomenon in one of MCI subjects. However, the rest of our MCI patients already had decreased CBVa, and consequently decreased CBF. This may suggest that intervention to prevent or delay changes in transit time may have an impact on the development of MCI or clinical dementia.

Conclusion

Detailed quantifications of ASL signal demonstrated in this study give us better understanding of cerebrovascular mechanism of AD progression.

Acknowledgements

This work is supported by NIH 5P50 AG005133-31

References

1. Iadecola C, Davisson RL. Hypertension and cerebrovascular dysfunction. Cell Metab 2008;7(6):476-484. 2. Lee Y, Kim T. Assessment of hypertensive cerebrovascular alterations with multiband Look-Locker arterial spin labeling. J Magn Reson Imaging 2017. 3. Setsompop K, et al. Blipped-controlled aliasing in parallel imaging for simultaneous multislice echo planar imaging with reduced g-factor penalty. Magn Reson Med 2012;67(5):1210-1224. 4. Du AT, et al. Hypoperfusion in frontotemporal dementia and Alzheimer disease by arterial spin labeling MRI. Neurology 2006;67(7):1215-1220. 5. Chappell MA, et al. Modeling dispersion in arterial spin labeling: validation using dynamic angiographic measurements. Magn Reson Med 2013;69(2):563-570. 6. Sabayan B, et al. Cerebrovascular hemodynamics in Alzheimer's disease and vascular dementia: a meta-analysis of transcranial Doppler studies. Ageing Res Rev 2012;11(2):271-277.