Nigral Iron Deposition Is Associated With Levodopa-Induced Dyskinesia in Parkinson's Disease

Abstract

Objective: To investigate iron deposition in the substantia nigra (SN) of Parkinson's disease (PD) patients associated with levodopa-induced dyskinesia (LID).

Methods: Seventeen PD patients with LID, 17 PD patients without LID, and 16 healthy controls were recruited for this study. The mean QSM values of the whole, left, and right SN were compared among the three groups. A multivariate logistic regression model was constructed to determine the factors associated with increased risk of LID. The receiver operating characteristic curve of the QSM value of SN in discriminating PD with and without LID was evaluated.

Results: The mean QSM values of the whole and right SN in the PD with LID were higher than those in the PD without LID (^∗ P = 0.03, ^∗ P = 0.03). Multivariate logistic regression analysis revealed that the QSM value of whole, left, or right SN was a predictor of the development of LID (^∗ P = 0.03, ^∗ P = 0.04, and ^∗ P = 0.04). The predictive accuracy of LID in adding the QSM value of the whole, left, and right SN to LID-related clinical risk factors was 70.6, 64.7, and 67.6%, respectively. The QSM cutoff values between PD with and without LID of the whole, left, and right SN were 148.3, 165.4, and 152.7 ppb, respectively.

Conclusion: This study provides the evidence of higher iron deposition in the SN of PD patients with LID than those without LID, suggesting that the QSM value of the SN may be a potential early diagnostic neuroimaging biomarker for LID.

Keywords: Parkinsion’s disease; dyskinesia; levodopa; quantitative susceptibility mapping; substantia nigra.

FIGURE 1

The definition of regions of interest (ROIs) included the left and right SN [(A) coronal image, (B) axial image, and (C) sagittal image].

FIGURE 2

Compare mean quantitative susceptibility mapping (QSM) value of the whole, left, and right SN among the three groups. (A) Analysis of variance (ANOVA) analysis of QSM value in the whole SN (SN_whole): Parkinson’s disease (PD) with levodopa-induced dyskinesia (LID) > PD without LID > healthy controls (HCs) (*P_{LID vs without LID} < 0.05, **P_{LID VS HCs} < 0.01, **P_{withoutLID VS HCs} < 0.01). (B) ANOVA analysis of QSM value in the left SN (SN_L): PD with LID>HCs (**P_{LID vs HCs} < 0.01) and PD without LID<HCs (**P_{without LID vs HCs} < 0.01). (C) ANOVA analysis of QSM value in the right SN (SN_R): PD with LID<PD without LID < HCs (*P_{LID VSwithout LID} < 0.05, **P_{LID vs HCs} < 0.01, *P_{withoutLIDvsHCs} < 0.05).

FIGURE 3

Receiver operating characteristic (ROC) curves showing the accuracy of QSM in differentiating between LID and HCs (A), PD without LID and HCs (B), and LID and PD without LID (C) in the SN (whole SN, left SN, and right SN, respectively). The diagnostic performance of the QSM value was defined by the area under the ROC curve (AUC). QSM, quantitative susceptibility mapping; PD, Parkinson’s disease (PD); LID, levodopa-induced dyskinesia; HCs, healthy controls; SN_whole, whole SN; SN_L, left SN; SN_R; right SN.