Intracerebral hemodynamic abnormalities in patients with Parkinson's disease: Comparison between multi-delay arterial spin labelling and conventional single-delay arterial spin labelling

Abstract

Purpose

The purpose of this study was to analyze the intracerebral abnormalities of hemodynamics in patients with Parkinson's disease (PD) through arterial spin labelling (ASL) technique with multi-delay ASL (MDASL) and conventional single-delay ASL (SDASL) protocols and to verify the potential clinical application of these features for the diagnosis of PD.

Materials and methods

Perfusion data of the brain obtained using MDASL and SDASL in patients with PD were compared to those obtained in healthy control (HC) subjects. Intergroup comparisons of z-scored cerebral blood flow (zCBF), arterial transit time (zATT) and cerebral blood volume (zCBV) were performed via voxel-based analysis. Performance of these perfusion metrics were estimated using area under the receiver operating characteristic curve (AUC) and compared using Delong test.

Results

A total of 47 patients with PD (29 men; 18 women; mean age, 69.0 ± 7.6 (standard deviation, [SD]) years; range: 50.0−84.0 years) and 50 HC subjects (28 men; 22 women; mean age, 70.1 ± 6.2 [SD] years; range: 50.0−93.0 years) were included. Relative to the uncorrected-zCBF map, the corrected-zCBF map further refined the distributed brain regions in the PD group versus the HC group, manifested as the extension of motor-related regions (P_FWE < 0.001). Compared to the HC subjects, patients with PD had elevated zATT and zCBV in the right putamen, a shortened zATT in the superior frontal gyrus, and specific zCBV variations in the left precuneus and the right supplementary motor area (P_FWE < 0.001). The corrected-zCBF (AUC, 0.90; 95% confidence interval [CI]: 0.84−0.96) showed better classification performance than uncorrected-zCBF (AUC, 0.84; 95% CI: 0.75−0.92) (P = 0.035). zCBV achieved an AUC of 0.89 (95% CI: 0.82−0.96) and zATT achieved an AUC of 0.66 (95% CI: 0.55−0.77). The integration model of hemodynamic features from MDASL provided improved performance (AUC, 0.97; 95% CI: 0.95−0.98) for the diagnosis of PD by comparison with each perfusion model (P < 0.001).

Conclusion

ASL identifies impaired hemodynamics in patients with PD including regional abnormalities of CBF, CBV and ATT, which can better be mapped with MDASL compared to SDASL. These findings provide complementary depictions of perfusion abnormalities in patients with PD and highlight the clinical feasibility of MDASL.