7 Tesla MRI with Quantitative Susceptibility Mapping: qualitative grading and quantitative focal measurements of the substantia nigra as a potential biomarker for Parkinson’s disease
TL;DR
A combined QSM approach using visual grading followed by focal susceptibility measurement may provide added value in the imaging assessment of PD, and require validation in larger, preferably multicentre cohorts, with all readers blinded to disease status.
Abstract
Parkinson’s disease (PD) is characterized by dopaminergic neuronal loss in the substantia nigra (SN). Imaging biomarkers sensitive to nigral pathology are limited. This study evaluated susceptibility changes in the SN using quantitative susceptibility mapping (QSM) and susceptibility-weighted imaging (SWI) at 7 Tesla MRI, and explored associations with clinical variables. In this exploratory cross-sectional study, 76 healthy controls (HC) and 26 PD patients underwent 7 Tesla QSM; SWI was available in 53 HC and all PD patients. Grading of dorsal (SN) was performed on QSM and SWI using a four-point scale. Using Sectra IDS7, focal susceptibility values were measured in QSM images in anterior and dorsal SN ROIs. Correlation analyses with clinical variables were performed. Intra-rater reliability for repeated focal SN measurements ranged from moderate to excellent (ICC range 0.68–0.95). Focal dorsal SN susceptibility was significantly higher in PD patients compared with HC (p < 0.001). The area under the ROC curve for dorsal SN was 0.90 ( 95% CI 0.80–0.99). A two-step QSM approach combining visual grading and focal susceptibility measurement showed sensitivity of 88.5% and specificity of 89.5%. No clinical correlations remained significant after FDR correction. In this exploratory 7 Tesla MRI study, the focal dorsal SN susceptibility was significantly higher in PD patients. A combined QSM approach using visual grading followed by focal susceptibility measurement may provide added value in the imaging assessment of PD. These exploratory findings require validation in larger, preferably multicentre cohorts, with all readers blinded to disease status.