Abstract
Background:
Cerebral microbleeds (CMBs) in patients with Parkinson’s disease (PD) or dementia with Lewy bodies (DLB) have not been adequately studied.
Objective:
This study aims to find a difference in the total number, prevalence, and common locations of CMBs between PD and DLB and evaluate 99 mTc-ECD SPECT subtraction images of these two diseases.
Methods:
We examined 112 patients with PD (53 males and 59 females; age: 77.4±3.6 years) and 28 age-matched patients with DLB (15 males and 13 females; age: 77.1±6.7 years) using brain magnetic resonance imaging (MRI) and 99 mTc-ECD SPECT subtraction imaging.
Results:
The total number of CMBs was higher in patients with DLB (41.2%) than in those with PD (11.5%), and the prevalence was significantly higher in the former (0.7±1.1) than the latter (0.2±0.5, p < 0.05). The odds ratio was 5.4 (95% confidence interval [CI]: 1.7–17.4). Furthermore, CMBs were commonly located in the basal ganglia of patients with PD (6 out of 87 patients) but in the occipital lobe of patients with DLB (8 out of 17 patients). 99 mTc-ECD SPECT subtraction imaging indicated lower cerebral blood flow in the posterior cingulate gyrus among the patients with CMB-positive DLB than among those with CMB-positive PD; additionally, the cerebral blood flow was lower in the bilateral basal ganglia and midbrain among patients with CMB-positive DLB compared to those with CMB-negative DLB.
Conclusion:
A reduction in occipital glucose metabolism may be related to CMBs in the occipital lobe of patients with DLB.
Keywords
INTRODUCTION
Parkinson’s disease (PD) and dementia with Lewy bodies (DLB) are neurodegenerative diseases characterized by an abnormal intraneuronal accumulation of misfolded α-synuclein within Lewy bodies and Lewy neurites [1]. Blood pressure abnormalities are generally associated with cerebrovascular diseases; however, previous studies have found no significant associations between blood pressure abnormalities and cerebral microbleeds (CMBs) in patients with PD and DLB [2, 3]. Therefore, this study examined the locations of CMBs using T2 star-weighted images (T2*WI) and assessed white matter changes using the Fazekas scale in fluid-attenuated inversion re-covery images from the brain magnetic resonance imaging (MRI) scans of patients with PD and DLB. Furthermore, it evaluated a new imaging method—subtracting the averaged 99 mTc-ECD SPECT images of patients with PD and DLB, focusing on both CMB-positive and CMB-negative groups of patients.
MATERIALS AND METHODS
Participants
Overall, 112 patients with PD (53 males and 59 females; age: 77.4±3.6 years) and 28 age-matched patients with DLB (15 males and 13 females; age: 77.1±6.7 years), who visited Okayama University Hospital between April 2010 and April 2020, were selected as participants. PD and DLB were diagnosed according to their clinical diagnostic criteria [4, 5]. Brain MRI scans were obtained from 90 patients with PD and 17 patients with DLB, including T1 weighted images, T2 weighted images, T2*WI, and fluid-attenuated inversion recovery images. The demographic characteristics of patients with PD and DLB were compared for systolic and diastolic blood pressure (BP), blood sugar (BS; mg/dL), low-density lipoprotein cholesterol (LDL-C; mg/dL), triglyceride (TG; mg/dL), and high-density lipoprotein cholesterol (HDL-C; mg/dL). The study was approved by the Ethical Committee of the Graduate School of Medicine, Dentistry and Pharmaceutical Science, Okayama University (approval No. K1603-031).
Cognitive and affective functions
Cognitive function was assessed using the Mini-Mental State Examination (MMSE) [6], Hasegawa Dementia Scale-Revised (HDS-R) [7], and the Frontal Assessment Battery (FAB) [8]. Affective function was assessed using the Geriatric Depression Scale (GDS) [9], Apathy Scale (AS) [10], and Abe’s behavioral and physiological symptoms of dementia (BPSD) score [11].
Neuroimaging
CMBs were defined as small (<10 mm), homogeneous, and having round foci with low signal intensity on T2*WI [12]. CMB locations in the fro-ntal, temporal, parietal, and occipital lobes, basal ganglia, and cerebellum were analyzed. The total number of CMBs was categorized into four subgr-oups: 0, 1, 2–4, and ≥5. White matter changes were assessed using the Fazekas rating grade [13], in which periventricular hyperintensity (PVH) and deep white matter hyperintensity (DWMH) lesions were graded as 0–III for PVH and 0–3 for DWMH. Furthermore, for comparisons of cerebral blood flow (CBF), 99 mTc-ethyl cysteinate dimer (ECD) SPECT scintigraphs [14] were obtained from 90 patients with PD and 14 patients with DLB; SPECT images were anatomically standardized using SPM8 [15]. An averaged SPECT image was created for each group of patients—1) PD with CMBs (PD CMB +), 2) PD without CMBs (PD CMB-), 3) DLB with CMBs (DLB CMB +), and 4) DLB without CMBs (DLB CMB-)—from these anatomically standardized images and by scaling voxel counts by a global mean. Subsequently, we compared the subtraction images of the PD CMB + , the PD CMB-, and the DLB CMB- groups with those of the DLB CMB + , the DLB CMB-, and the DLB CMB + groups, respectively.
Statistical analysis
Statistical analyses were performed using SPSS (Version 22.0). Comparisons between baseline characteristics (gender and age) and baseline scores of MMSE, HDS-R, FAB, GDS, and AS were conducted using the Mann-Whitney U test. In all statistical analyses, the significance level was assumed at p < 0.05.
RESULTS
Table 1 shows the demographic characteristics of patients with PD and DLB; it indicates that no statistically significant differences were found in systolic and diastolic BP (mmHg), BS (mg/dL), LDL-C (mg/dL), or TG (mg/dL) between the two gr-oups. However, a significant difference was found in relation to HDL-C (mg/dL). The average cognitive functions, as measured by MMSE, HDS-R, and FAB scores, were significantly higher in patients with PD (27.5±2.0, 27.3±2.5, and 14.9±2.7, respectively, p < 0.001) than in those with DLB (20.1±4.2, 19.3±7.0, and 10.0±4.8, respectively). Furthermore, the affective functions of those with DLB were more impaired in terms of their GDS (7.1±4.2, p < 0.05), apathy (21.0±11.3, p < 0.05), and Abe’s BPSD scores (9.6±8.1, p < 0.001) compared to patients with PD (5.3±4.2, 14.1±9.0, and 2.4±4.5, respectively) (see Table 1).
Demographics and microbleed results in PD and DLB
ABS, Abe’s behavioral and physiological symptoms of dementia score; AS, Apathy Scale; BP, blood pressure; BS, blood sugar; DLB, dementia with Lewy bodies; DWMH, deep white matter hyperintensity; FAB, Frontal Assessment Battery; GDS, Geriatric Depression Scale; HDL-C, high-density lipoprotein cholesterol; HDS-R, Hasegawa Dementia Scale-Revised; LDL-C, low-density lipoprotein cholesterol; MB, microbleed; MMSE, Mini-Mental State Exam; MRI, magnetic resonance imaging; PD, Parkinson’s disease; PVH, periventricular hyperintensity; TG, triglyceride **Odds ratio 5.4 (95% confidence interval: 1.7–17.4).
With respect to the MRI images, the number of total CMBs was higher in patients with DLB (41.2%) than in those with PD (11.5%), and their prevalence was significantly higher in the DLB group (0.7±1.1, p < 0.05) than in the PD group (0.2±0.5); the odds ratio was 5.4 (95% confidence interval [CI]: 1.7–17.4). CMBs were commonly located in the basal ganglia of patients with PD (6 out of 87 patients), and in the occipital lobe of patients with DLB (8 out of 17 patients). The MRI-based Fazekas PVH and DWMH scales and MIBG scintigraphy did not show any significant differences between the two groups (see Table 1).
A comparison of the subtracted average 99 mTc-ECD SPECT images between the PD CMB + and the DLB CMB + groups (see Fig. 1, left) reveal fewer CBF in the posterior cingulate gyrus of the latter group (marked with arrows) than in the former one; whereas, comparing the averaged images of the PD CMB- and the DLB CMB- groups exhibited no significant differences in terms of CBF. The subtracted SPECT images of the DLB CMB- and the DLB CMB + groups (see Fig. 1, right) reveal fewer CBF in the bilateral basal ganglia (marked with arrows) and midbrain (marked with arrowhead) in the latter than in the former group.
The subtracted 99 mTc-ECD SPECT averaged image of the PD CMB + and the DLB CMB + groups (left panels) shows fewer CBF in the posterior cingulate gyrus in the DLB CMB + group than in the PD CMB + group (arrows). The subtracted SPECT image of the DLB CMB- and the DLB CMB + groups (right panels) shows fewer CBF in the bilateral basal ganglia (arrows) and midbrain (arrowhead) in the DLB CMB + group than in the DLB CMB- group.
DISCUSSION
In this study, the prevalence ratios of CMBs in patients with PD and DLB were 11.5% and 41.2%, respectively, similar to previous reports of 17.4% in PD [2] and 17% –45% in DLB [16–18]. Our findings indicate that CMBs are commonly located in the basal ganglia of patients with PD, consistent with a previous study [2]. Furthermore, studies suggest that CMBs are commonly located in the lobar region [16], frontal or deep areas [17], and occipital or frontal lobes [18] of patients with DLB. However, in this study, CMBs were found to be commonly located in the occipital lobe of the patients with DLB; notably, the prevalence ratio of CMBs in patients with DLB was 5.4 times higher than in those with PD (see Table 1, odds ratio).
Patients with DLB exhibit a significant reduction in occipital glucose metabolism [19, 20], which is linked to occipital hypoperfusion. Additionally, white matter spongiform change and gliosis in the occipital lobe are prominently observed in autopsied patients with DLB [21]. The reduction in occipital glucose metabolism as well as the pathological changes could be related to the prevalence of CMBs in the occipital lobe of DLB patients.
This is the first study evaluating subtraction of the average 99 mTc-ECD SPECT images of patients with PD and DLB, focusing on whether patients are CMB-positive or CMB-negative. The study revealed a reduction in occipital glucose metabolism related to the presence of CMBs in the occipital lobe of patients with DLB.
Footnotes
ACKNOWLEDGMENTS
This study was partly supported by a Grant-in-Aid for Scientific Research (B) 17H0419611, (C) 17H0975609, 17K1082709, 20K12044, and Grants-in-Aid from the Research Committees (Kaji R, Toba K, and Tsuji S) from the Japan Agency for Medical Research and Development 7211700121, 7211800049, and 7211800130. In addition, we are grateful to Mr. Masahiro Nakashima (Department of Radiological Technology, Graduate School of Health Sciences, Okayama University) and Mr. Yoshitake Takahashi (FUJIFILM Toyama Chemical Co., Ltd.) for efficient technical support in creating the subtraction images of averaged 99 mTc-ECD SPECT images of patients with PD and DLB.
