Advertisement

Factors associated with motor severity in vascular parkinsonism with normal dopamine transporter imaging

      Highlights

      • Periventricular hyperintensities correlated most strongly with the motor severity.
      • Diabetic patients had higher motor severity compared with non-diabetics.
      • They both had independent impacts on motor symptoms of vascular parkinsonism.

      Abstract

      Introduction

      To delineate the determinants of motor severity in vascular parkinsonism (VaP), we investigated the impact of regional white matter intensity (WMH) burden and co-morbidities on the motor score in the patients with VaP and normal dopamine transporter (DAT) imaging.

      Methods

      In this multicenter, retrospective study, we reviewed the records of 63 patients diagnosed with VaP and normal DAT imaging on 18F-FP-CIT PET. Signal hyperintensities in deep white matter (DWMH), periventricular (PVH), basal ganglia (BG) regions, and infratentorial foci (ITF) were rated according to Scheltens scale, a semi-quantitative visual rating system. Motor severity was assessed with Unified Parkinson's Disease Rating Scale (UPDRS) motor score. Regional hyperintensity scores, patients' demographics, and co-morbidities such as type 2 diabetes, hypertension, dyslipidemia, and previous stroke history were used as starting variables, and stepwise regression analysis was performed to select independent predictors of motor severity.

      Results

      PVH (R = 0.33, p = 0.008) and DWMH score (R = 0.31, p = 0.015) correlated with the motor severity, while BG and ITF scores did not. Diabetic patients had significantly higher motor scores compared with non-diabetics (34.7 (13.0) vs. 27.5 (12.4), p = 0.008). Other factors such as sex, BMI, hypertension, dyslipidemia, and previous history of stroke did not impact motor severity. In multivariate analysis, PVH scores and diabetes significantly correlated with motor severity.

      Conclusion

      PVH burden and diabetes were independent factors associated with motor severity in VaP with normal DAT imaging. Our results suggest that diabetes, along with white matter hyperintensities, may have a significant role in the development of motor symptoms in VaP.

      Keywords

      1. Introduction

      Vascular parkinsonism (VaP) is a clinical syndrome that generally refers to parkinsonian signs that are believed to be associated with ischemic cerebrovascular disease rather than primary neurodegenerative pathologies [
      • Rektor I.
      • Bohnen N.I.
      • Korczyn A.D.
      • Gryb V.
      • Kumar H.
      • Kramberger M.G.
      • de Leeuw F.E.
      • Pirtosek Z.
      • Rektorova I.
      • Schlesinger I.
      • Slawek J.
      • Valkovic P.
      • Vesely B.
      An updated diagnostic approach to subtype definition of vascular parkinsonism - Recommendations from an expert working group.
      ]. Clinically, VaP usually manifests with parkinsonism mainly involving gait disturbances ("lower body parkinsonism"), with variable extent of cognitive decline, pseudobulbar palsy, or pyramidal signs. However, the definition of this syndrome remains ambiguous, mainly due to lack of a clear understanding of its pathological background. Therefore, investigating the factors associated with the motor severity in VaP may be a starting point in understanding the pathophysiologic link between ischemic cerebral lesions and the parkinsonian signs. Many studies have already tried this approach, but the studies performed before the wide use of dopamine transporter (DAT) imaging possibly included the patients with concurrent Parkinson's disease (PD) [
      • Antonini A.
      • Vitale C.
      • Barone P.
      • Cilia R.
      • Righini A.
      • Bonuccelli U.
      • Abbruzzese G.
      • Ramat S.
      • Petrone A.
      • Quatrale R.
      • Marconi R.
      • Ceravolo R.
      • Stefani A.
      • Lopiano L.
      • Zappia M.
      • Capus L.
      • Morgante L.
      • Tamma F.
      • Tinazzi M.
      • Colosimo C.
      • Guerra U.P.
      The relationship between cerebral vascular disease and parkinsonism: the VADO study.
      ].
      To avoid confusion, an expert working group has proposed the classification of VaP into three subtypes: acute or subacute poststroke VaP, insidious onset VaP, and mixed or overlapping syndromes of VaP and neurodegenerative parkinsonism [
      • Rektor I.
      • Bohnen N.I.
      • Korczyn A.D.
      • Gryb V.
      • Kumar H.
      • Kramberger M.G.
      • de Leeuw F.E.
      • Pirtosek Z.
      • Rektorova I.
      • Schlesinger I.
      • Slawek J.
      • Valkovic P.
      • Vesely B.
      An updated diagnostic approach to subtype definition of vascular parkinsonism - Recommendations from an expert working group.
      ]. A separate study for each of these subtypes may provide a better understanding of underlying pathophysiology. Therefore, in this multicenter retrospective observational study, we focused on the insidious onset VaP with normal DAT imaging and investigated the factors associated with the severity of parkinsonian signs in these patients.

      2. Methods

      2.1 Subjects

      In three centers of South Korea (Ajou University Hospital, Dongtan Sacred Heart Hospital, and Yonsei University Wonju College of Medicine), we retrospectively reviewed medical records of the patients with the diagnosis of VaP between June 2014 and January 2016. The diagnosis was based on previously published clinical criteria [
      • Rektor I.
      • Bohnen N.I.
      • Korczyn A.D.
      • Gryb V.
      • Kumar H.
      • Kramberger M.G.
      • de Leeuw F.E.
      • Pirtosek Z.
      • Rektorova I.
      • Schlesinger I.
      • Slawek J.
      • Valkovic P.
      • Vesely B.
      An updated diagnostic approach to subtype definition of vascular parkinsonism - Recommendations from an expert working group.
      ]. Among the patients reviewed, those with abnormal DAT imaging (18F-FP-CIT PET) findings were excluded, except those with focal uptake defects due to lacunar infarction in the exact location. The patients with Evans ratio of more than 0.3 were also excluded to rule out the possible diagnosis of normal pressure hydrocephalus (NPH). Among the initial 75 patients of VaP with normal DAT imaging reviewed, 10 patients missing the initial Unified Parkinson's Disease Rating Scale (UPDRS) score were excluded in the analysis. Also, 2 patients who had more than 5-year interval between the acquisition of brain MRI and the diagnosis of VaP were excluded. Basic demographical data and clinical history of diabetes, hypertension, dyslipidemia, and previous ischemic or hemorrhagic stroke were obtained from chart review.

      2.2 Standard protocol approvals, registrations, and patient consent

      The study was approved by the Institutional Review Board of Ajou University Hospital (AJIRB-MED-MDB-21-253). The review board waived informed consent due to the retrospective design of the study.

      2.3 The assessment of the motor severity

      For assessment of the motor severity, we used UPDRS motor score (part 3) from the initial evaluation when the patient was first diagnosed. These patients were drug-naïve at the time, so the score represents the motor severity during the off-state.

      2.4 Semi-quantitative scoring of MRI hyperintensity lesion

      All MRI scans of VaP patients were performed on a 3.0 T MRI system at each center, and we used T2 and FLAIR sequence images for scoring. Signal hyperintensities in deep white matter (DWMH), periventricular (PVH), basal ganglia (BG) regions, and infratentorial foci (ITF) were rated according to the semi-quantitative visual rating system proposed by Scheltens and colleagues [
      • Scheltens P.
      • Barkhof F.
      • Leys D.
      • Pruvo J.P.
      • Nauta J.J.
      • Vermersch P.
      • Steinling M.
      • Valk J.
      A semiquantative rating scale for the assessment of signal hyperintensities on magnetic resonance imaging.
      ]. PVH score was rated separately for the frontal and occipital caps, and bands along lateral ventricles using a 0–2 scale. The total PVH score was the sum of the score in the three areas (0–6). DWMH score was rated separately for the frontal, temporal, parietal, and occipital lobes with the range of 0–6, in which 0 had no abnormality and 6 had a confluent lesion. The sum of the score from four lobes was used as total DWMH score (0–24). BG score was rated separately for the caudate, putamen, globus pallidus, thalamus, and internal capsule using a 0–6 scale. Total BG score was the sum of the score in these 5 regions (0–30). ITF score was rated separately for cerebellum, midbrain, pons, and medulla using a 0–6 scale. Total ITF score was the sum of the score in the four area (0–24). Thus, total hyperintensity score ranged from 0 to 84. MRI hyperintensities were scored blindly by DGP, JHY, and SYK. If the score was discordant between raters, the final score was determined by consensus.

      2.5 Quantitative analysis of 18F-FP-CIT PET

      All the patients had 18F-FP-CIT PET to assess DAT availability, but the images from only one center (Ajou University Hospital, n = 15) were eligible for quantification. The DAT images were quantified using PMOD Neuro tool (PNEURO, version 3.7, PMOD Technologies, Zürich, Switzerland), as described previously [
      • Park D.G.
      • Kim J.W.
      • An Y.S.
      • Chang J.
      • Yoon J.H.
      Plasma neurofilament light chain level and orthostatic hypotension in early Parkinson's disease.
      ]. Regional uptake values in the bilateral caudate nucleus and putamen were divided by those in the cerebellar cortex uptake to provide regional standardized uptake value ratios (SUVRs).

      2.6 Statistical analysis

      Continuous variables are presented as means (standard deviations), and categorical variables are presented as counts (percentages). For the comparison of continuous variables between two groups, two-tailed t-test was used in the case of gaussian distribution, and Mann-Whitney U test was used for nongaussian distribution of data. Chi-squared test was used to compare categorical variables. Pearson correlation and multiple linear regression analysis were used to assess correlations between continuous variables. Due to multicollinearity between the hyperintensity sub-scores, stepwise Akaike's Information Criterion (AIC) method was used in both directions to select the variables in a multiple regression model. P-values < 0.05 were considered to indicate statistical significance. All statistical analyses and plotting of graphs were performed using R software (version 3.5.3; (R Foundation for Statistical Computing, Vienna, Austria).

      3. Results

      3.1 Patient demographic and clinical characteristics

      A total of 63 patients (38 males (60.3%)) were included in the analysis (Supplementary Table). The mean age at diagnosis was 75.8 (5.6) years. The mean interval between the diagnosis and the brain MRI was 0.14 (0.64) years, with the majority of the patients (93.7%) having an interval of less than a year. Regarding vascular risk factors, 45 patients (71.4%) had a history of hypertension, 22 patients (34.9%) had diabetes (all of which had type 2 diabetes mellitus), and 22 patients (34.9%) had dyslipidemia. Body mass index (BMI) was 24.21 (2.88). 26 patients (41.3%) had a clinical history of previous strokes. The mean UPDRS motor score was 30.0 (13.0) with Hoehn and Yahr stage of 2.87 (0.96). The mean total white matter hyperintensities (WMH) score was 30.37 (10.18).

      3.2 Correlation between MRI hyperintensity score and motor severity

      Total MRI hyperintensity score and the motor severity showed a significant positive correlation. When tested with subscores of Scheltens scale, PVH score and DWMH score positively correlated with UPDRS motor score, whereas BG and ITF scores did not (Fig. 1). When DWMH scores were analyzed with anatomical distribution, only frontal (R = 0.29, p = 0.021) and parietal (R = 0.28, p = 0.025) WMH severity scores correlated with the motor severity (Supplementary Fig. 1).
      Fig. 1
      Fig. 1Correlations between motor severity and regional white matter hyperintensity burden. UPDRS, Unified Parkinson's Disease Rating Scale; WMH, white matter hyperintensities.

      3.3 Demographic and vascular risk factors (co-morbidities) correlated with motor severity

      Age significantly correlated with the motor severity (R = 0.30, p = 0.017). Diabetic patients had significantly higher motor severity compared with non-diabetic patients (34.7 (13.0) vs. 27.5 (12.4), p = 0.008; Supplementary Fig. 2). Other factors such as sex, BMI, hypertension, dyslipidemia, and previous history of stroke were not associated with the motor severity.

      3.4 Correlation between DAT imaging findings and motor severity

      16 patients showed "lesion defect" in their DAT imaging, a focal DAT uptake defect due to lacunar infarction in the striatum. When compared with those without lesion defect (n = 47), there was no significant difference in the motor score between the groups.
      There was no correlation between the motor severity score and striatal DAT uptake. Total WMH score (R = −0.54, p = 0.037) and BG hyperintensities score (R = −0.62, p = 0.014) negatively correlated with striatal DAT uptake, even with age-adjustments (p = 0.041 and 0.017, respectively).

      3.5 Multivariate regression analysis

      Table 1 shows the result of stepwise multiple regression analysis. In this model, PVH score and diabetes were independently associated with the motor severity score.
      Table 1Univariate and multivariate analysis in the correlation between UPDRS motor score and clinical parameters.
      Univariate analysis (n = 63)Multivariate analysis (n = 63)
      ΒpβP
      Age0.2990.017*
      Male sex−0.2550.043*
      BMI−0.1900.164−0.1570.224
      Diabetes0.2670.034*0.3130.016*
      Hypertension−0.0210.869
      Dyslipidemia0.0370.777
      History of previous strokes0.0280.8300.0520.694
      WMH scores
       PVH score0.3290.008*0.3520.007**
       DWMH score0.3050.015*
       BG score0.1190.353
       ITF score0.1950.127
      BMI, body mass index; UPDRS, Unified Parkinson's Disease Rating Scale; WMH, white matter hyperintensities; PVH, periventricular hyperintensities; DWMH, deep white matter hyperintensities; BG, basal ganglia; ITF, infratentorial; *, p < 0.05; **, p < 0.01.

      4. Discussion

      In the present study, we found that the PVH score and the presence of diabetes had independent positive correlations with the motor severity score.
      The positive correlation between PVH and motor severity in the VaP has been addressed in previous studies. In the VADO study, among the patients with normal FP-CIT SPECT findings (n = 48), PVH score positively correlated with UPDRS motor score (p = 0.05) [
      • Antonini A.
      • Vitale C.
      • Barone P.
      • Cilia R.
      • Righini A.
      • Bonuccelli U.
      • Abbruzzese G.
      • Ramat S.
      • Petrone A.
      • Quatrale R.
      • Marconi R.
      • Ceravolo R.
      • Stefani A.
      • Lopiano L.
      • Zappia M.
      • Capus L.
      • Morgante L.
      • Tamma F.
      • Tinazzi M.
      • Colosimo C.
      • Guerra U.P.
      The relationship between cerebral vascular disease and parkinsonism: the VADO study.
      ]. However, a small sample size of those with the clinical diagnosis of VaP and normal FP-CIT SPECT (n = 28) limited the interpretation of data. Also, in a study investigating the association of MRI hyperintensities and falls in the elderly population, PVH and frontal DWMH scores were independently associated with the history of falls [
      • Rektor I.
      • Bohnen N.I.
      • Korczyn A.D.
      • Gryb V.
      • Kumar H.
      • Kramberger M.G.
      • de Leeuw F.E.
      • Pirtosek Z.
      • Rektorova I.
      • Schlesinger I.
      • Slawek J.
      • Valkovic P.
      • Vesely B.
      An updated diagnostic approach to subtype definition of vascular parkinsonism - Recommendations from an expert working group.
      ]. Likewise, the study comparing clinical features between VaP and PD showed more frequent postural instability in the VaP patients with PVH and DWMH [
      • Benitez-Rivero S.
      • Marin-Oyaga V.A.
      • Garcia-Solis D.
      • Huertas-Fernandez I.
      • Garcia-Gomez F.J.
      • Jesus S.
      • Caceres M.T.
      • Carrillo F.
      • Ortiz A.M.
      • Carballo M.
      • Mir P.
      Clinical features and 123I-FP-CIT SPECT imaging in vascular parkinsonism and Parkinson's disease.
      ]. Similar results are also found in the patients with PD. A recent study reported the association between axial parkinsonian symptoms and PVH, and between bradykinesia and DWMH [
      • Lee Y.
      • Ko J.
      • Choi Y.E.
      • Oh J.S.
      • Kim J.S.
      • Sunwoo M.K.
      • Yoon J.H.
      • Kang S.Y.
      • Hong J.Y.
      Areas of white matter hyperintensities and motor symptoms of Parkinson disease.
      ]. Considering that VaP mainly manifests with axial symptoms commonly referred to as "lower-body parkinsonism," this may be in accordance with our results.
      Contrary to common belief, BG hyperintensities score did not correlate with motor severity; but it did correlate with quantitatively measured DAT availability. A recent study with 23 VaP patients with visually normal DAT imaging showed a negative correlation between PVH and quantitative DAT availability [
      • Lee Y.H.
      • Lee S.
      • Chung S.J.
      • Yoo H.S.
      • Jung J.H.
      • Baik K.
      • Ye B.S.
      • Sohn Y.H.
      • Yun M.
      • Lee P.H.
      The pattern of FP-CIT PET in pure white matter hyperintensities-related vascular parkinsonism.
      ]. That study targeted the patients without basal ganglia lesions, but according to our results, DAT availability more readily correlates with BG hyperintensity score than PVH score. We suspect that DAT availability may be influenced by BG ischemia (even when basal ganglia is visually normal in MRI) because BG and PVH scores correlate with each other (R = 0.42, P < 0.001). Moreover, considering that DAT availability in PD patients does not necessarily reflect their motor severity, we can expect the same for the patients with VaP.
      Only a few studies report the association between diabetes and motor severity. A cohort study including 882 subjects without PD or dementia, diabetes was associated with worsening rigidity and gait (which are typical parkinsonian signs associated with VaP), but not with bradykinesia or tremor [
      • Arvanitakis Z.
      • Wilson R.S.
      • Schneider J.A.
      • Bienias J.L.
      • Evans D.A.
      • Bennett D.A.
      Diabetes mellitus and progression of rigidity and gait disturbance in older persons.
      ]. On the other hand, for PD, diabetes is increasingly recognized as a risk factor for both disease onset and motor severity [
      • Chohan H.
      • Senkevich K.
      • Patel R.K.
      • Bestwick J.P.
      • Jacobs B.M.
      • Bandres Ciga S.
      • Gan-Or Z.
      • Noyce A.J.
      Type 2 diabetes as a determinant of Parkinson's disease risk and progression.
      ]. Diabetes was independently associated with postural instability and gait difficulty, which is similar to what we found in VaP.
      The possible mechanisms underlying the link between diabetes and parkinsonism are still under investigation. Previous animal studies showed hypoinsulinemia decreased basal dopamine concentrations and DAT mRNA level in the substantia nigra [
      • Craft S.
      • Watson G.S.
      Insulin and neurodegenerative disease: shared and specific mechanisms, the Lancet.
      ]. Higher insulin resistance correlated with dopaminergic dysfunction, and the synaptic dysfunction was found in the mouse models of both type 1 and type 2 diabetes [
      • Craft S.
      • Watson G.S.
      Insulin and neurodegenerative disease: shared and specific mechanisms, the Lancet.
      ]. Imaging studies also provide some evidence. Diffusion tensor imaging (DTI) studies showed that diabetes was associated with abnormal white matter integrity, possibly related to the potential contribution of diabetes to small vessel disease. Even in the patients with normal appearing white matter, these microstructural alterations were associated with gait disorders and parkinsonism [
      • Wang H.C.
      • Hsu J.L.
      • Leemans A.
      Diffusion tensor imaging of vascular parkinsonism: structural changes in cerebral white matter and the association with clinical severity.
      ].
      In addition, diabetic patients are not only at risk for hyper-glycemia but also for hypo-glycemic attacks. For example, we have previously reported a case of reversible parkinsonism with normal DAT imaging, which was associated with basal ganglia edema resulted from a hypoglycemic attack [
      • Gil Y.E.
      • Yoon J.H.
      Hypoglycemia-induced parkinsonism with vasogenic basal ganglia lesion.
      ]. More research is needed on whether recurrent hypoglycemic attacks can cause persistent parkinsonism.
      This study is subject to several limitations. First, due to the retrospective design, only patients who underwent both MRI and DAT scans were enrolled in the study, which may have led to selection bias. Patients with very mild symptoms or a definitive clinical diagnosis may have been more hesitant to undergo additional imaging studies. Second, we did not quantify WMH volume, but used semi-quantitative scale with visual assessment. However, Scheltens scale correlated well with automated WMH volume measurements, according to a previous study [
      • Shim Y.
      • Yoon B.
      • Hong Y.J.
      • Cho A.H.
      • Yang D.-W.
      A semi-automated method for measuring white matter hyperintensity volume.
      ]. Third, although we excluded the patients with Evans ratio greater than 0.3, we cannot completely rule out the possibility of NPH enrolled in our study. Lastly, we acknowledge the ongoing controversy of VaP as a distinct clinical entity [
      • Vizcarra J.A.
      • Lang A.E.
      • Sethi K.D.
      • Espay A.J.
      Vascular Parkinsonism: deconstructing a syndrome.
      ]. Our results present the possibility of clinical parkinsonism arising from PVH and diabetes in the patients with normal DAT imaging, supporting the distinction of VaP as a separate entity.
      In conclusion, PVH burden and diabetes independently correlated with motor severity in VaP with normal DAT imaging. Thus, our results suggest that diabetes, along with white matter hyperintensities, may have a significant role in the development of motor symptoms in VaP.

      Funding

      This research was supported by the Basic Science Research Program through the National Research Foundation of Korea funded by the Ministry of Education (no. NRF-2018M3A9E8023859 (JHY)).

      Declarations of interest

      None.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Rektor I.
        • Bohnen N.I.
        • Korczyn A.D.
        • Gryb V.
        • Kumar H.
        • Kramberger M.G.
        • de Leeuw F.E.
        • Pirtosek Z.
        • Rektorova I.
        • Schlesinger I.
        • Slawek J.
        • Valkovic P.
        • Vesely B.
        An updated diagnostic approach to subtype definition of vascular parkinsonism - Recommendations from an expert working group.
        Park. Relat. Disord. 2018; 49: 9-16
        • Antonini A.
        • Vitale C.
        • Barone P.
        • Cilia R.
        • Righini A.
        • Bonuccelli U.
        • Abbruzzese G.
        • Ramat S.
        • Petrone A.
        • Quatrale R.
        • Marconi R.
        • Ceravolo R.
        • Stefani A.
        • Lopiano L.
        • Zappia M.
        • Capus L.
        • Morgante L.
        • Tamma F.
        • Tinazzi M.
        • Colosimo C.
        • Guerra U.P.
        The relationship between cerebral vascular disease and parkinsonism: the VADO study.
        Park. Relat. Disord. 2012; 18: 775-780
        • Scheltens P.
        • Barkhof F.
        • Leys D.
        • Pruvo J.P.
        • Nauta J.J.
        • Vermersch P.
        • Steinling M.
        • Valk J.
        A semiquantative rating scale for the assessment of signal hyperintensities on magnetic resonance imaging.
        J. Neurol. Sci. 1993; 114: 7-12
        • Park D.G.
        • Kim J.W.
        • An Y.S.
        • Chang J.
        • Yoon J.H.
        Plasma neurofilament light chain level and orthostatic hypotension in early Parkinson's disease.
        J. Neural. Transm. 2021;
        • Benitez-Rivero S.
        • Marin-Oyaga V.A.
        • Garcia-Solis D.
        • Huertas-Fernandez I.
        • Garcia-Gomez F.J.
        • Jesus S.
        • Caceres M.T.
        • Carrillo F.
        • Ortiz A.M.
        • Carballo M.
        • Mir P.
        Clinical features and 123I-FP-CIT SPECT imaging in vascular parkinsonism and Parkinson's disease.
        J. Neurol. Neurosurg. Psychiatry. 2013; 84: 122-129
        • Lee Y.
        • Ko J.
        • Choi Y.E.
        • Oh J.S.
        • Kim J.S.
        • Sunwoo M.K.
        • Yoon J.H.
        • Kang S.Y.
        • Hong J.Y.
        Areas of white matter hyperintensities and motor symptoms of Parkinson disease.
        Neurology. 2020; 95: e291-e298
        • Lee Y.H.
        • Lee S.
        • Chung S.J.
        • Yoo H.S.
        • Jung J.H.
        • Baik K.
        • Ye B.S.
        • Sohn Y.H.
        • Yun M.
        • Lee P.H.
        The pattern of FP-CIT PET in pure white matter hyperintensities-related vascular parkinsonism.
        Park. Relat. Disord. 2021; 82: 1-6
        • Arvanitakis Z.
        • Wilson R.S.
        • Schneider J.A.
        • Bienias J.L.
        • Evans D.A.
        • Bennett D.A.
        Diabetes mellitus and progression of rigidity and gait disturbance in older persons.
        Neurology. 2004; 63: 996-1001
        • Chohan H.
        • Senkevich K.
        • Patel R.K.
        • Bestwick J.P.
        • Jacobs B.M.
        • Bandres Ciga S.
        • Gan-Or Z.
        • Noyce A.J.
        Type 2 diabetes as a determinant of Parkinson's disease risk and progression.
        Mov. Disord. 2021;
        • Craft S.
        • Watson G.S.
        Insulin and neurodegenerative disease: shared and specific mechanisms, the Lancet.
        Neurology. 2004; 3: 169-178
        • Wang H.C.
        • Hsu J.L.
        • Leemans A.
        Diffusion tensor imaging of vascular parkinsonism: structural changes in cerebral white matter and the association with clinical severity.
        Arch. Neurol. 2012; 69: 1340-1348
        • Gil Y.E.
        • Yoon J.H.
        Hypoglycemia-induced parkinsonism with vasogenic basal ganglia lesion.
        Park. Relat. Disord. 2018; 49: 112-113
        • Shim Y.
        • Yoon B.
        • Hong Y.J.
        • Cho A.H.
        • Yang D.-W.
        A semi-automated method for measuring white matter hyperintensity volume.
        Dement. Neurocogn. Disord. 2013; 12: 21-28
        • Vizcarra J.A.
        • Lang A.E.
        • Sethi K.D.
        • Espay A.J.
        Vascular Parkinsonism: deconstructing a syndrome.
        Mov. Disord. 2015; 30: 886-894