4D Flow MRI for assessing flow pulsatility along the carotid siphon in Alzheimer’s disease
Leonardo A Rivera-Rivera1, Tilman Schubert2, Kevin M Johnson1, Sterling C Johnson3, Oliver Wieben1,2, and Patrick Turski2

1Medical Physics, University of Wisconsin Madison, Madison, WI, United States, 2Radiology, University of Wisconsin Madison, Madison, WI, United States, 3Medicine, University of Wisconsin Madison, Madison, WI, United States

Synopsis

Cerebral arteries are often morphologically altered and dysfunctional in Alzheimer’s disease (AD). In this study, 4D flow MRI was used to assess flow pulsatility along the carotid siphon in patients with AD, mild cognitive impairment (MCI) and in healthy age matched controls. We found the physiologic dampening of pulsatility along the distal ICA is significantly diminished in patients with AD. With the large volume coverage and high temporal and spatial resolution, 4D flow MRI can provide additional biomarkers of vascular health that can contribute to the identifying patients who could benefit from interventions to improve circulatory system functions.

Purpose

There is recent evidence that suggests the tortuous cavernous segment of the internal carotid artery (ICA) may have attenuating effects on the pulsatile arterial flow on healthy young individuals [1]. In contrast, cerebral arteries are often morphologically altered and dysfunctional in Alzheimer’s disease (AD) [2]. Therefore, there is growing interest in the non-invasive assessment of cranial hemodynamics as potential systemic indicators of AD [3]. Recent advances in MR hardware, data acquisition, and reconstruction have facilitated 4D flow MRI in clinically feasible scan times, thereby providing dynamic velocity vector maps with volumetric coverage. With adequate spatial and temporal resolution, such 4D flow MRI approaches are well suited for comprehensive hemodynamic assessment of the larger vessels. Here we test the hypothesis that the physiologic dampening of pulsatility along the distal ICA is significantly diminished in patients with AD. We investigate local arterial blood flow patterns in the proximal middle cerebral artery (MCA) and ICA proximal to the carotid siphon in three age-matched groups: (1) patients with AD, (2) mild cognitive impairment (MCI) and in (3) older healthy controls with gated 4D flow MRI.

Methods

Subjects: The study population consisted of 20 AD patients (age range 61-89y, mean=73y, 6 F), 26 MCI patients (age range 52-87y, mean= 73y, 12 F) and 30 older control adults (age range 66-89y, mean= 74y, 17 F). MRI: Volumetric, time-resolved phase contrast (PC) MRI data with 3-directional velocity encoding were acquired on a 3T clinical MRI system (MR750, GE Healthcare) with an 8 channel head coil (Excite HD Brain Coil, GE Healthcare), and with a 3D radially undersampled sequence, PC VIPR [4]. Scan parameters Venc = 80 cm/s, imaging volume = 22x22x10 cm3, (0.7 mm)3 acquired isotropic spatial resolution, TR/TE=7.4/2.7ms, scan time ~ 7 min, retrospective cardiac gated into 20 cardiac phases with temporal interpolation[5]. Flow analysis: Vessel segmentation was performed in Matlab (The Mathworks, Natick, MA) on the PC angiograms, while interactive flow visualization and selection of planes for quantitative analysis were carried out in Ensight (CEI, Apex, NC), also using PC angiograms. For this purpose, flow analysis planes were manually placed orthogonal to the vessel orientation in 4 vessel segments (Fig. 1) (d, e): distal cervical Internal Carotid Artery (ICA) (left & right) and 5 mm from the Middle Cerebral Artery origin (left & right). 2D cine images series with through plane velocities were generated from the 4D flow MRI data and analyzed in a customized Matlab tool [6] that automatically detected the edge of the vessel wall. Pulsatility index ($$$PI=\frac{Q{max}-Q{min}}{Q_{mean}}$$$) and MCA/ICA PI ratios were calculated for the vessel segments and groups were compared with Student’s t-test (statistical significance for p<0.05).

Results

Results for the vessel analysis are summarized in Figures 2, 3 & 4. We found a significant increase in PI from ICA to MCA for all three groups especially pronounced in the AD cohort. The MCI/ICA ratio is significantly larger in the AD group compared to age matched controls, increasing by 13%. There is also significant increase in cervical ICA PI when AD and MCI are compared to controls. With the AD group reporting a 24 % increase in cervical ICA PI when compared to controls. The MCA PI was significantly higher in the AD cohort when compared to controls, reporting a 26% augmentation.

Discussion

These results differ from those reported in a study based on PC MRI on a healthy young population, where flow pulsatility was attenuated along the carotid syphon, possibly due to the contorted shaped of the cervical ICA [1]. Our results indicate an age-associated increased arterial rigidity and decreased arterial compliance, significantly aggravated in the AD cohort [3]. This pathology reduces the Windkessel effect, progressing to increase pulse wave velocity and possibly contributing to reduced glymphatic transport of amyloid from the extracellular space and cognitive decline.

Conclusions

This study demonstrates the feasibility of hemodynamic analysis over a large vascular territory in the context of Alzheimer’s disease with 4D flow MRI within a 7 minute acquisition. All three groups of elderly patients and controls showed significant increase in flow pulsatility from ICA to MCA, indicating age associated arterial wall stiffening. However, the AD group reported significantly higher MCA/ICA PI ratio than the controls, suggesting a further diminished potentially protective effect from arterial pulsation for downstream cerebral vasculature. With the large volume coverage and high temporal and spatial resolution demonstrated here, 4D flow MRI can provide additional biomarkers of vascular health that can contribute to identifying patients who could benefit from interventions to improve circulatory system functions.

Acknowledgements

We gratefully acknowledge funding by the NIH (NIA grant P50-AG033514 and NIGMS R25GM083252) as well as GE Healthcare for their assistance and support. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

References

[1] Schubert T, Santini F, Stalder AF, Bock J, Meckel S, Bonati L, et a. Dampening of Blood-Flow Pulsatility along the Carotid Siphon: Does Form Follow Function? AJNR 2011;32:1107-1112.

[2] Roher AE. Editorial Comment: Cardiovascular system participation in Alzheimer's disease pathogenesis. Journal of Internal Medicine. 2015; 277(4): 426-28.

[3] Roher AE, Garami Z, Tyas SL, Maarouf CL, Kokjohn TA, Belohlavek M, et al. Transcranial Doppler ultrasound blood flow velocity and pulsatility index as systemic indicators for Alzheimer’s disease. Alzheimers Dement. 2011;7:445–455.

[4] Gu T, Korosec FR, Block WF, Fain SB, Turk Q, Lum D, et al. PC VIPR: a high-speed 3D phase-contrast method for flow quantification and high resolution angiography. AJNR Am J Neuroradiology 2005;26(4):743–749.

[5] Liu J, Redmond MJ, Brodsky EK, Alexander AL, Lu A, Thornton FJ, et al. Generation and visualization of four dimensional MR angiography data using an undersampled 3-D projection trajectory. IEEE Trans Med Imaging. 2006; 25(2):148–157.

[6] Stalder AF, Russe MF, Frydrychowicz A, Bock J, Henning J, Markl M. Quantitative 2D and 3D phase contrast MRI: optimized analysis of blood flow and vessel wall parameters. Magn Reson Med 2008;60:1218–31.

Figures

Figure1: PC VIPR data shown as (a)sagittal, (b)coronal, (c)axial MIP images of the PC angiogram, demonstrating large volumetric coverage and isotropic spatial resolution. Segmented views of the ICA(d) and MCA(e) with blood flow distribution and velocity vectors. (f)Pulsatile flow waveform through the cardiac cycle generated by the PC VIPR data.

Figure 2: Pulsatility Index (PI) in the cervical ICA (blue) and MCA (red) for 76 subjects including patients with AD, MCI and the age matched controls. Significant differences between ICA and MCA are indicated with a “*”.

Figure 3: PI numerical values in the ICA, MCA for 76 subjects including patients with AD, MCI and the age matched controls. The MCI/ICA PI ratios are also shown.

Figure 4: (a) P-values obtain with unpaired Student’s t-test comparing PI among cohorts; (b) P-values obtain with paired Student’s t-test comparing PI in the ICA and MCA. (Statistical significance for p<0.05)

Figure 5: MCA/ICA PI ratios for 76 subjects including patients with AD, MCI and the age matched controls. Statistically significant differences were found between AD and age matched controls are indicated with a “*”.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
0151