Introduction

Patients with Parkinson’s disease (PD) are known to have dopaminergic denervation in the substantia nigra (SN). Prior studies have shown that nigrosome-1, a subregion of SN, and neuromelanin contrast are useful imaging surrogates to detect PD pathology.^1,2 The nigrosome-1 is radiologically delineated as a region of signal hyperintensity on T2* susceptibility-weighted imaging (SWI) in healthy subjects, but lost in PD.³ Dark pigmentation of the SN is attributed to neuromelanin, which has T1-shortening properties manifest on neuromelanin-sensitive imaging (NMS), and reduced in PD.

High resolution susceptibility map-weighted images (SMWI) has recently been developed to improve visualization of the nigrosome-1 region of the substantia nigra (SN) on 3T imaging.⁴ SMWI utilizes susceptibility weighting mask derived from quantitative susceptibility mapping (QSM) to enhance the contrast for the SWI magnitude images.

In this study, we evaluated the sensitivity of quantitative SMWI & NMS in differentiating PD from HC.

Methods

This study was approved by the local ethics board. Twenty-eight PD patients clinically diagnosed by a movement disorder specialist and 25 healthy controls (HC) were included in the study (Figure 1). All participants underwent motor assessments (Figure 1), viz the Unified Parkinson’s Disease Rating Scale (UPDRS), Hoehn and Yahr Scale (H&Y), and brain MRI scan on a 3T scanner.

3D high resolution T2* SWI images were acquired using a multi-echo gradient echo sequence with TR=48ms, TE=13.77/26.39/39ms, FA=20°, in-plane resolution= 0.5x0.5x1mm³, number of slices = 32, scan duration 4:09mins. An oblique-coronal imaging plane oriented perpendicular to the midbrain structure was chosen (Figure 2).⁴

A high resolution NMS imaging was performed using a T1 TSE sequence with TR/TE=938/15 ms, voxel size=0.5x0.5x3mm³, number of slices=13, scan duration 10:25mins. The imaging tilting was copied from the SWI acquisition and the first slice placement follows SWI acquisition.

QSM and SMWI images were reconstructed from the multi-echo GRE images using SMWI software (Seoul National University, Seoul, South Korea).⁴

Region-of-interest (ROI) of the SN was manually drawn on SMWI and NMS images, whilst blinded to subject status (Figure 3). SMWI images were inspected in the superoinferior-anteroposterior direction to identify the last slice where the inferior tail of the red nucleus was still visible. The SN mask was drawn along its margins on the next consecutive 3 slices. A background ROI mask was placed over the decussation of the superior cerebellar peduncles, which lies inferior to the SN.⁵

On the NMS images, the SN mask was drawn on the slice corresponding to the first SMWI slice containing the SN mask. The SN mask was drawn on this and the next inferior slice.

Histogram analysis was performed on the mask metrics from both QSM/SMWI and NMS. The difference in the mean susceptibility values from the SN and background masks were computed from the QSM images. Chi-square test was used to compare the sex distribution between PD patients and healthy controls. Mann-Whitney U test or independent Student’s t-test was performed according to the normality of each continuous variable distribution.

Results

Clinical demographics were reported in Figure 1. Representative images of NMS, QSM and SMWI in a HC and PD patient were shown in Figure 4. SN masks on SWMI and NMS were similar in volume between PD and HC (p=0.675).

Mean quantitative values from the SN masks on QSM/SMWI and NMS images are tabulated in Figure 5. Susceptibility values in the SN was significantly higher in PD than HC (p<0.0001). Given that SMWI was derived from the QSM, similar results were also found in the SN masks on SMWI, where the mean signal in the SN was significantly lower in PD than HC.

There was no significant difference in mean signal of the SN masks on NMS between PD and HC (p=0.104). After filtering for voxels which are higher than background signal, the SN masks in HC were significantly bigger than PD.

Discussion

To our knowledge, this is the first study evaluating the sensitivity of QSM/SMWI and NMS imaging in PD. Our quantitative SMWI results confirmed and added to the existing literature showing the utility of SMWI as a tool to visualize nigrosome-1.^1,4,6 SMWI heightens the contrast between regions with high iron content such as SN and its surrounding structures, as shown by the strong differentiation between PD and HC in our results. As such, it would be an ideal tool for radiologists in clinical diagnosis.

Although NMS imaging was also able to distinguish PD and HC well, it was not as robust as SMWI in sensitivity. NMS images needed additional post-processing to accentuate the difference between PD and HC by filtering out signal contribution from the background which are visually distracting. Besides, the SN boundaries on NMS images were not as crisp as those on SMWI images, as the viable, neuromelanin-containing dopaminergic neurons appeared more diffuse distributed on NMS images, especially in PD (Figure 4).

Conclusion

Both QSM/SMWI and NMS offered excellent differentiation of PD and HC through targeted imaging of nigrosome-1 and neuromelanin-containing dopaminergic neurons respectively. Nevertheless, QSM/SMWI offered more robust results compared to NMS and may be preferred for use in the clinical setting.

Acknowledgements

We would like to thank National Medical Research Council, Singapore for their funding support.