High correlation of quantitative susceptibility mapping and echo intensity measurements of nigral iron overload in Parkinson's disease

Abstract

Quantitative susceptibility mapping (QSM) and transcranial sonography (TCS) offer proximal evaluations of iron load in the substantia nigra. Our prospective study aimed to investigate the relationship between QSM and TCS measurements of nigral iron content in patients with Parkinson's disease (PD). In secondary analyses, we wanted to explore the correlation of substantia nigra imaging data with clinical and laboratory findings. Eighteen magnetic resonance imaging and TCS examinations were performed in 15 PD patients at various disease stages. Susceptibility measures of substantia nigra were calculated from referenced QSM maps. Echogenicity of substantia nigra on TCS was measured planimetrically (echogenic area) and by digitized analysis (echo-intensity). Iron-related blood serum parameters were measured. Clinical assessments included the Unified PD Rating Scale and non-motor symptom scales. Substantia nigra susceptibility correlated with echogenic area (Pearson correlation, r = 0.53, p = 0.001) and echo-intensity (r = 0.78, p < 0.001). Individual asymmetry indices correlated between susceptibility and echogenic area measurements (r = 0.50, p = 0.042) and, more clearly, between susceptibility and echo-intensity measurements (r = 0.85, p < 0.001). Substantia nigra susceptibility (individual mean of bilateral measurements) correlated with serum transferrin saturation (Spearman test, r = 0.78, p < 0.001) and, by trend, with serum iron (r = 0.69, p = 0.004). Nigral echogenicity was not clearly related to serum values associated with iron metabolism. Susceptibility and echogenicity measurements were unrelated to PD duration, motor subtype, and severity of motor and non-motor symptoms. The present results support the assumption that iron accumulation is involved in the increase of nigral echogenicity in PD. Nigral echo-intensity probably reflects ferritin-bound iron, e.g. stored in microglia.

Keywords: Magnetic resonance imaging; Parkinson disease; Substantia nigra; Transcranial sonography.

Fig. 1

Quantitative susceptibility mapping (QSM) of substantia nigra (SN) and topological correspondence with echo signals. (A) MRI susceptibility map of midbrain used for identification of bilateral SN and bilateral red nucleus. (B) MRI susceptibility map of midbrain with outlined ROIs of right-sided SN (blue edging) and nucleus ruber (NR; red edging). (C) MRI susceptibility map of brain corresponding to the scans shown in (A) and (B). The midbrain was surrounded (red line) for better recognition (arrow: left-sided SN). (D) Transcranial brain sonogram of axial transection through caudal midbrain in a patient with Parkinson’s disease. Enlarged echosignal of bilateral SN is visualized. The midbrain was surrounded (red line) for better recognition (arrow: left-sided SN). (E) Exact real-time fusion imaging with transcranial sonography and 3D referenced QSM shows the topological correspondence of SN demarcation by its echogenicity and its susceptibility. The midbrain was surrounded (red line) for better recognition (arrow: left-sided SN). (F) Referenced QSM image, with inversion of the image brightness values for better visualization. Note that the QSM plane shown in (E) and (F) has been reconstructed by the ultrasound system from the imported QSM DICOM 3D dataset according to the ultrasound plane shown in (D) and (E). The midbrain was surrounded (red line) for better recognition

Fig. 2

Visualization and measurement of substantia nigra echosignals in Parkinson’s patients. (A) Transcranial brain sonogram of axial transection at caudal midbrain level. The lower panel shows the zoomed midbrain in which the substantia nigra echo signals (arrow) on the side ipsilateral to the insonation were manually surrounded for planimetric measurement of echogenic area. (B) Screen shot of the MATLAB-based software tool “Cereb B-Mode Assist” used for digitized quantification of substantia nigra echo-intensity. The right panel shows the zoomed midbrain in which an elliptical ROI of 50 mm² was placed to cover the echo signals of the substantia nigra (arrow) on the side ipsilateral to the insonation. The automatic analysis results in the curve of the echo-intensity distribution in the ROI compared to a reference curve determined in healthy controls (lower left panel) and the calculated index value of the total substantia nigra echo-intensity (lower right panel)

Fig. 3

Diagrams showing the relationship between susceptibility and echogenicity measurements of the substantia nigra in Parkinson’s patients. (A) Relationship between susceptibility and digitally analyzed echo-intensity of the substantia nigra (Pearson test, p < 0.001). (B) Relationship between the susceptibility and the planimetrically measured echogenic area of the substantia nigra (p = 0.001)

Fig. 4

Diagrams showing the relationship between the measurement of circulating iron levels and the susceptibility of the substantia nigra in Parkinson’s patients. (A) Relationship between serum transferrin saturation and substantia nigra susceptibility (individual mean of bilateral measures; Spearman test, p < 0.001). (B) Relationship between serum iron and substantia nigra susceptibility (individual mean of bilateral measures; p = 0.004, not significant after Bonferroni correction with significance level set at p ≤ 0.003)