Differentiation of Parkinson's disease and Parkinsonism predominant multiple system atrophy in early stage by morphometrics in susceptibility weighted imaging

Abstract

Background and purpose: We previously established a radiological protocol to discriminate multiple system atrophy-parkinsonian subtype (MSA-P) from Parkinson's disease (PD). However, we do not know if it can differentiate early stage disease. This study aimed to investigate whether the morphological and intensity changes in susceptibility weighted imaging (SWI) of the lentiform nucleus (LN) could discriminate MSA-P from PD at early stages.

Methods: We retrospectively enrolled patients with MSA-P, PD and sex- and age-matched controls whose brain MRI included SWI, between January 2015 and July 2020 at the Movement Disorder Center. Two specialists at the center reviewed the medical records and made the final diagnosis, and two experienced neuroradiologists performed MRI analysis, based on a defined and revised protocol for conducting morphological measurements of the LN and signal intensity.

Results: Nineteen patients with MSA-P and 19 patients with PD, with less than 2 years of disease duration, and 19 control individuals were enrolled in this study. We found that patients with MSA- P presented significantly decreased size in the short line (SL) and corrected short line (cSL), ratio of the SL to the long line (SLLr) and corrected SLLr (cSLLr) of the LN, increased standard deviation of signal intensity (SIsd_LN, cSIsd_LN) compared to patients with PD and controls (P < 0.05). With receiver operating characteristic (ROC) analysis, this finding had a sensitivity of 89.5% and a specificity of 73.7% to distinguish MSA- P from PD.

Conclusion: Compared to PD and controls, patients with MSA-P are characterized by a narrowing morphology of the posterior region of the LN. Quantitative morphological changes provide a reference for clinical auxiliary diagnosis.

Keywords: Parkinson’s disease; diametral measurement; early duration; multiple system atrophy; susceptibility weighted imaging.

FIGURE 1

(A) Standard midsagittal plane. The thalamic syndesmosis (hollow red arrow), corpus callosum splenium (black dot) and 4th ventricle posterior edge (red dot). The parallel horizontal green and yellow line define the axis plane selection for lentiform nucleus and cerebrospinal fluid measurements: the green line is set at 3.5 mm (red vertical line) above the black dot, and the yellow line is set to cross the cusp of the red dot. The vertical blue line defines the anterior-posterior location of the lentiform nucleus short line and is set to cross the middle of thalamic syndesmosis. (B) Standard axis plane for lentiform nucleus measurement. The blue and red lines are used for the short and long lines of the lentiform nucleus measurement: the red line is the longest line near the middle area of the lentiform nucleus, and the blue line is defined by the vertical blue line in a. (C) F, 59, patient with probable multiple system atrophy-parkinsonian subtype with general weakness for 1 year, obviously in her lower limbs, walking drag, dizziness for 2 months, and turning over difficulty for 1 month. (D) F, 60, patient with Parkinson disease with an inflexible right upper limb for 2 years, right lower limb shake for 1 year, and the right upper limb shake for more than 2 months. The slice selection and measurement protocols were assigned according to our previous study. The smaller short line side (blue line in panel C and red line in panel D) was recorded as corrected short line (cSL), and the corrected ratio (cSLLr) was calculated using the long line with the same cSL side.

FIGURE 2

The box-scatter blot of the three groups with upper lower limit, upper lower quartile, and median line. The x-axis represents the three groups using different colors, and the y-axis represents the measured and calculated indexes. (A) corrected short line, cSL, (B) corrected short and long line ration, cSLLr, (C) mean signal intensity of lentiform nucleus, SIm_LN, (D) standard deviation of signal intensity of lentiform nucleus, SIsd_LN, (E) Short line, SL, (F) the ratio of short and long line, SLLr, (G) mean signal intensity of cerebrospinal fluid, SIm_CSF, (H) standard deviation of signal intensity of cerebrospinal fluid SIsd_CSF, (I) long line, LL, (J) Area, (K) normalized mean signal intensity, nSIm, and (L) corrected standard deviation of signal intensity, cSIsd.

FIGURE 3

The receiver operating characteristic (ROC) curves of six indexes for multiple system atrophy-parkinsonian subtype vs. Parkinson’s disease (A) and eight indexes multiple system atrophy- parkinsonian subtype vs. control group (B).