Objective: Identifying novel biomarkers to predict disease progression is of paramount importance in patients with Parkinson’s disease (PD). In this study, we utilized Structural Magnetic Resonance Imaging (sMRI)-based multivariate summary score of gray matter volume (GMV) using Mahalanobis distance (MD) to capture the GMV heterogeneity in individual patients and (1) examined whether MD of GMV (M_GMV) could predict the rate of progression of motor symptoms, and (2) investigated the changes in M_GMV over time.

Background: Longitudinal sMRI studies have reported GMV abnormalities in widespread brain regions and were associated with disease progression in PD[1]. A metric that summarizes the GMV heterogeneity from multiple brain regions into a single score, to predict motor severity and its longitudinal changes has not yet been explored in PD.

Method: sMRI (T1-weighted MRI) and clinical motor scores (MDS-UPDRS-III; medication OFF) data were obtained from the Parkinson’s Progression Markers Initiative database (www.ppmi-info.org/data); for 40 PD patients with sMRI at both baseline (BL) and 48-month follow-up (48M), and 55 healthy controls (HC) at BL only [table 1]. GMV was calculated from 86 regions of interest (ROIs) using Hammers atlas[2] in Computational Anatomy Toolbox (CAT12, http://dbm.neuro.uni-jena.de/cat/). Later, GMV was divided by total intracranial volume of each participant to obtain normalized GMV (GMV_N) and derived a subject-specific summary score of GMV_N using MD defined as M_GMV = √(s-µ)τ .C-1.(s-µ), where, s represents a vector of subject GMV_N in each ROI, µ the average GMV_N calculated from HC, and C a covariance matrix between ROIs across the control population. We tested whether M_GMV could predict changes in motor severity, measured by ∆ MDS-UPDRS-III using multiple linear regression, ∆ denotes the changes in the MDS-UPDRS-III scores from BL to 48M. To assess the changes in M_GMV over time, we conducted linear mixed-effect model.

Results: M_GMV at BL significantly predicted changes in motor severity (β = 4.99; adjusted R²= 0.28; p = 0.005) [figure 1]. Longitudinal analyses from BL to 48M showed a tendency of increased M_GMV over time, however, it was not significant (β = 0.11; p = 0.09) [figure 2].

Conclusion: In conclusion, we identified an imaging biomarker based on a patient-specific summary score of gray matter volume to predict long-term motor outcomes in patients with PD.

References: [1] E. Sarasso, F. Agosta, N. Piramide, M. Filippi, Progression of grey and white matter brain damage in Parkinson’s disease: a critical review of structural MRI literature, J. Neurol. 268 (2021) 3144–3179.
[2] A. Hammers, R. Allom, M.J. Koepp, S.L. Free, R. Myers, L. Lemieux, T.N. Mitchell, D.J. Brooks, J.S. Duncan, Three-dimensional maximum probability atlas of the human brain, with particular reference to the temporal lobe, Hum. Brain Mapp. 19 (2003) 224–247.

« Back to 2022 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/patient-specific-multivariate-gray-matter-volumetric-distance-to-predict-disease-severity-in-parkinsons-disease/