Abstract
Background
CSF1R-related leukoencephalopathy is a type of autosomal dominant leukodystrophy caused by mutations in the colony stimulating factor 1 receptor (CSF1R) gene. Subcortical ischemic vascular dementia (SIVaD), which is caused by cerebral small vessel disease, is similar to CSF1R-related leukoencephalopathy in that it mainly affects subcortical white matter. In this study, we compared the patterns of white matter hyperintensity (WMH) and cortical thickness in CSF1R-related leukoencephalopathy with those in SIVaD.
Methods
Fourteen patients with CSF1R-related leukoencephalopathy and 129 with SIVaD were retrospectively recruited from three tertiary medical centers. We extracted and visualized WMH data using voxel-based morphometry to compare the WMH distributions between the two groups. Cortical thickness was measured using a surface-based method. Statistical maps of differences in cortical thickness between the two groups were generated using a surface model, with age, sex, education, and intracranial volume as covariates.
Results
Predominant distribution of WMH in the CSF1R-related leukoencephalopathy group was in the bilateral frontal and parietal areas, whereas the SIVaD group showed diffuse WMH involvement in the bilateral frontal, parietal, and temporal areas. Compared with the SIVaD group, the CSF1R-related leukoencephalopathy group showed more severe corpus callosum atrophy (CCA) and widespread cortical thinning.
Conclusions
To our knowledge, this is the first study using the automated MR measurement to capture WMH, cortical thinning, and CCA with signal changes in CSF1R-related leukoencephalopathy. It provides new evidence regarding differences in the patterns of WMH distribution and cortical thinning between CSF1R-related leukoencephalopathy and SIVaD.
Introduction
CSF1R-related leukoencephalopathy, previously known as adult-onset leukoencephalopathy with axonal spheroids and pigmented glia, is a rare autosomal dominant white matter disease caused by mutations in the colony stimulating factor 1 receptor (CSF1R) gene [1–4]. It has variable manifestations including cognitive decline, behavioral changes, motor dysfunction, and seizures [4–6]. As the disease progresses, patients eventually become bedridden with spasticity and rigidity [7]. Pathologically, CSF1R-related leukoencephalopathy is characterized by diffuse white matter degeneration and widespread axonal spheroids [2] observable as bilateral white matter lesions, cortical atrophy, a thinner corpus callosum (CC), and dilation of the lateral ventricles on magnetic resonance imaging (MRI) [1].
Given the heterogeneous clinical symptoms and characteristic but non-specific radiological findings of leukoencephalopathy, CSF1R-related leukoencephalopathy is often misdiagnosed as various other white matter diseases, including vascular dementia (VaD), primary progressive multiple sclerosis, other adult-onset leukodystrophies, toxic or metabolic disease, and inflammatory white matter diseases [8, 9]. Of these, VaD, especially subcortical ischemic vascular dementia (SIVaD), which is caused by cerebral small vessel disease, is the most common disease involving subcortical white matter and is also characterized by cognitive decline with motor deficits [10–12]. It is associated with vascular risk factors such as hypertension, diabetes mellitus or dyslipidemia [9]. Early detection of SIVaD is clinically important because managing these vascular risk factors can prevent its progression [13]. Alternatively, although microglia replacement by hematopoietic stem cell transplantation has been attempted [14], or a phase 2 study of VGL101 as a high-affinity human TREM2 agonistic monoclonal antibody (ClinicalTrials.gov ID NCT05677659) is ongoing, currently there are no approved disease-modifying therapy for CSF1R-related leukoencephalopathy. So an accurate and timely diagnosis is crucial for patients to save the extra money for extensive work-up and plan for long-term supportive therapy, and for physicians to optimize genetic counselling for family members.
A white matter imaging approach might be helpful in differentiating between CSF1R-related leukoencephalopathy and SIVaD, because both diseases are representative brain white matter diseases. Therefore, we compared the regional distributions of white matter hyperintensity (WMH) and changes in cortical thickness in CSF1R-related leukoencephalopathy with those in SIVaD to identify the patterns specific to the former using automated approaches. Although numerous neuroimaging studies on CSF1R-related leukoencephalopathy have been reported, this is the first study to automatically quantify WMH and measure cortical thinning. Even if this automated imaging approach may not be routinely applied in a clinical setting, we believe that the inferences drawn from the results of the present study will be helpful for differential diagnosis.
Methods
Participants
This study was approved by the institutional review boards of Samsung Medical Center (IRB approval no. 2019-12-034), Pusan National University Hospital (IRB approval no. H-1911-023-085), and Pusan National University Yangsan Hospital (IRB approval no. 05-2016-081). We obtained written consent from each patient. This study was conducted from December 27, 2019 and all data with fully anonymized were drawn from retrospective cohorts. Fourteen patients with CSF1R-related leukoencephalopathy and 129 with SIVaD were retrospectively recruited from three tertiary medical centers (Samsung Medical Center, Pusan National University Hospital, and Pusan National University Yangsan Hospital) between February 2009 and October 2022. All patients with CSF1R-related leukoencephalopathy were genetically confirmed. 3 out of 14 patients with CSF1R-related leukoencephalopathy were also pathologically confirmed. SIVaD was clinically diagnosed using Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition (DSM-IV) criteria for VaD and the following modified imaging requirements of the Fazekas ischemia and Erkinjuntti et al. criteria [14, 15]: (1) severe WMH on T2-weighted or fluid-attenuated inversion recovery (FLAIR) images (defined as periventricular WMH ≥ 10 mm and deep WMH ≥ 25 mm), and (2) deep grey matter lacunes. Patients with structural lesions including territorial cerebral infarction, brain tumors, WMH due to radiation injury, or multiple sclerosis, were excluded. They did not perform any genetic tests. For the cortical thickness and the mean volume of the CC analyses, 121 cognitively normal (CN) controls (mean age: 71.8 ± 7.0 years, male:female ratio, 47:74) with no history of neurological or psychiatric illness were selected from a cohort of volunteers at the Samsung Medical Center. All patients and CN controls underwent clinical evaluation, formal neuropsychological testing, and high-resolution 3.0-Tesla brain MRI (1 mm thickness T1-weighted and 2 or 5 mm thickness FLAIR imaging). Mean time intervals between disease onset and MRI study were 1.4 ± 1.7 years in CSF1R-related leukoencephalopathy and 3.6 ± 3.0 years in SIVaD groups, respectively. Mean durations between the neuropsychological tests and MRI of the patients with CSF1R-related leukoencephalopathy and SIVaD were 1.4 ± 2.7 months and 3.9 ± 10.5 months, respectively.
Image acquisition and processing
We acquired standardized 3D T1-weighted and FLAIR images from all patients using MRI scanners (Philips Achieva, the Netherlands; Siemens Skyra, Germany; Siemens Verio and Magnetom Avanto, Germany) at the three participating centers. The detailed MRI parameters for each center are described in S1 Table.
Assessment of WMH pattern
WMH data were extracted from FLAIR and T1-weighted images using the two Lesion Segmentation Tool (LST) algorithms. First, we used the Lesion Prediction Algorithm (LPA) [16] of the LST, which only requires FLAIR images and no additional parameters. The LPA results were compared with the actual FLAIR images, and if the extracted WMH values was different from the actual values, the other LST algorithm, the Lesion Growth Algorithm (LGA) [17], was used, which requires both FLAIR and T1-weighted images of a patient. We also had to input parameters and visually inspect the divisions produced by the different parameters and then use the optimal values determined in this manner. As the output of the LGA is in the T1 space, we linearly co-registered it to the FLAIR space using a Functional Magnetic Resonance Imaging of the Brain (FMRIB) software library (FMRIB’s Linear Image Registration Tool [18]) and nearest-neighbor interpolation. Additionally, to compare the WMH patterns between the CSF1R-related leukoencephalopathy and SIVaD groups, all WMH images were registered into Montreal Neurological Institute space, and significant WMHs were marked by color grade.
Measurement of WMH volume
Among the results obtained through the FreeSurfer process, automated segmentation and classification of WMH was used to create a mask for each patient. Then, the overlapping regions of each patient’s mask and WMH were identified and the corresponding areas were calculated. WMH volumes of 68 cortical regions of interests (ROIs) and five ROIs corresponding to the CC were then acquired.
Cortical thickness analyses
The publicly accessible, surface-based FreeSurfer package [19] was used for cortical thickness analysis. The analysis process included intensity normalization, registration to the Talairach space, exclusion of non-brain tissue, white matter segmentation, white matter boundary tessellation, smoothing of the tessellated surface, and automated topology correction. The deformable surface algorithm uses the tessellated surface as its starting point to delineate the white matter and then the pial boundary. The average distance from the white matter surface to the closest point on the pial surface and from that point back to the closest point on the white matter surface is used to determine the cortical thickness for each location on the tessellated white matter surface [20]. For all subjects, volumes were visually examined for reconstruction-related misclassifications after automated reconstruction. The recon-all script in FreeSurfer version 6.0 was used to obtain cortical thickness values for each T1-weighted scan. FreeSurfer output files were used to determine the thicknesses for all 68 ROIs of the Desikan–Killiany cortical parcellation. Additionally, the cortical thickness of 5 ROIs in the CC was also obtained resulting from FreeSurfer recon-all. Finally, the average CC volume was calculated using volume of genu, body and splenium from the result obtained through segmentation.
Neuropsychological tests
All patients underwent standardized neuropsychological testing using the Seoul Neuropsychological Screening Battery (SNSB) [21]. This battery consists of several tests that assess performance across five domains: attention, language and related functions, visuospatial functions, memory, and executive functions. Scores were considered abnormal when the result was lower than the 16th percentile for age- and education-matched norms. As the age of the norms ranged from 45 to 90 years, two patients with CSF1R-related leukoencephalopathy who were below 45 years of age were matched with 45-year-old norms.
Based on the neuropsychological test results, we used the following quantitatively scorable tests for each domain: the digit span backward task for attention, the Korean version of the Boston Naming Test (K-BNT) [22] for language and related functions, the Rey–Osterrieth Complex Figure Test (RCFT; copy) for visuospatial functions, the Seoul Verbal Learning Test (SVLT; a 20 min delayed recall trial of 12 items) for memory, and the Stroop Test (color reading of 112 items during a 2 min period) for executive functions. Z-scores for the subtest in each domain were used for comparing cognitive performance between the CSF1R-related leukoencephalopathy and SIVaD groups and for correlation analysis between WMH volume and cognitive impairment.
Statistical analyses
To compare the demographic and clinical variables of the CSF1R-related leukoencephalopathy and SIVaD groups, the Mann–Whitney U test (for continuous variables) and Fisher’s exact test (for categorical variables) were used. To evaluate the effects of regional WMH volume on cognitive outcomes in the CSF1R-related leukoencephalopathy group, multiple linear regression analysis was performed after controlling for age, sex, and education. SPSS version 24.0 software was used for all statistical analyses (SPSS Inc., Chicago, IL, USA).
WMH volume between the CSF1R-related leukoencephalopathy and SIVaD groups was compared using the Mann–Whitney U test. Frequency maps of WMH represented a considerably larger portion from the comparison between the two groups. Brain regions with a p-value of 0.05 or less were marked on the atlas in a color (ranging from yellow to red) corresponding to the p-value.
Localized differences in cortical thicknesses of the 68 cortical ROIs between the CSF1R-related leukoencephalopathy and SIVaD groups were examined using the Kruskal-Wallis test. Statistical maps of the differences in cortical thickness between the two groups were generated using a surface model, with age, sex, education, and intracranial volume as covariates. Only ROIs with p-values less than 0.05, as obtained using the ANCOVA, were considered as regions with significantly different cortical thickness.
Results
Demographic and clinical characteristics
The baseline demographic and clinical characteristics of the patients are presented in Tables 1 and 2. The mean age was significantly lower (p < 0.001) and educational level was significantly higher (p = 0.004) in the CSF1R-related leukoencephalopathy group compared with those in the SIVaD group. General cognitive indices, that is the Mini-Mental State Examination (MMSE) and Clinical Dementia Rating (CDR) scores, did not differ significantly between the two groups. With regard to the neuropsychological tests, only the visuospatial functions domain score was significantly lower in the CSF1R-related leukoencephalopathy group than in the SIVaD group (p = 0.013).
Table 1. Demographic and clinical characteristics of the included patients.
| CSF1R-related leukoencephalopathy (N = 14) | SIVaD (N = 129) | p-value | |
|---|---|---|---|
| Age (mean ± SD), years | 49.79 ± 6.13 | 76.05 ± 7.25 | <0.001 |
| Female, N (%) | 7 (50%) | 96 (74.4%) | 0.065* |
| Education (mean ± SD), years | 12.43 ± 5.16 | 7.97 ± 5.39 | 0.004 |
| Interval between onset and MRI, years | 1.38 ± 1.66 | 3.59 ± 2.96 | 0.004 |
| Interval between MRI and neuropsychological tests, months | 1.38 ± 2.67 | 3.93 ± 10.53 | 0.453 |
| Neuropsychological test results (z-score) | |||
| Attention | -1.56 ± 0.96 | -1.19 ± 1.28 | 0.461 |
| Language and related functions | -2.41 ± 1.77 | -1.99 ± 1.87 | 0.568 |
| Visuospatial functions | -15.32 ± 9.25 | -3.25 ± 3.13 | 0.013 |
| Memory | -2.43 ± 1.11 | -1.82 ± 1.18 | 0.186 |
| Executive functions | -4.66 ± 3.01 | -2.35 ± 1.59 | 0.162 |
| MMSE | 17.29 ± 7.96 | 18.09 ± 5.61 | 0.625 |
| CDR | 1.09 ± 0.63 | 1.18 ± 0.65 | 0.670 |
| ICV, mL | 1762.83 ± 256.85 | 1583.37 ± 132.04 | 0.020 |
| Mean WMH volume, mL (%) | |||
| Frontal lobe | 12.85 ± 29.15 (69.33%) | 28.28 ± 15.0 (55.95%) | <0.001 |
| Parietal lobe | 4.45 ± 8.06 (24%) | 16.97 ± 9.74 (33.57%) | <0.001 |
| Temporal lobe | 0.74 ± 1.78 (4%) | 2.68 ± 1.92 (5.30%) | <0.001 |
| Occipital lobe | 0.49 ± 0.89 (2.67%) | 2.62 ± 1.88 (5.18%) | <0.001 |
Abbreviations: SIVaD, subcortical ischemic vascular dementia; SD, standard deviation; MMSE, Mini-Mental State Examination; CDR, Clinical Dementia Rating; ICV, intracranial volume; WMH, white matter hyperintensity
*Fisher’s exact test
Table 2. Detailed clinical characteristics of included patients with CSF1R-related leukoencephalopathy.
| Case no. | Sex | Age at onset | Age at MRI | Duration of disease at initial MRI (year) | Family history | Initial symptoms | Dementia | Psychiatric symptoms | Pyramidal signs | Parkinsonism | Seizure | CSF1R gene mutation |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | M | 44 | 47 | 3 | + | Bradykinesia | - | - | + | + | - | p.R782H |
| 2 | F | 44 | 44 | 0 | - | Optic ataxia | + | - | - | - | - | p.F849del |
| 3 | F | 45 | 50 | 5 | - | Bradykinesia, postural instability | + | + | - | + | - | p.P878S |
| 4 | M | 47 | 49 | 2 | + | Dysarthria, gait disturbance | - | - | - | + | - | p.A823V |
| 5 | F | 51 | 52 | 1 | - | Cognitive impairment | + | + | - | - | - | p.I794T |
| 6 | M | 44 | 45 | 1 | - | Cognitive impairment | + | + | - | + | - | p.G747* p.F971Serfs*7 |
| 7 | M | 51 | 55 | 4 | - | Personality change | + | + | - | + | + | p.I794T |
| 8 | M | 52 | 56 | 4 | + | Personality change | + | + | - | + | + | p.C892_A894del |
| 9 | F | 60 | 61 | 1 | - | Cognitive impairment, gait disturbance | + | - | - | + | + | p.I794T |
| 10 | F | 56 | 58 | 2 | + | Bradykinesia | + | - | + | + | - | c.2442+5G>A |
| 11 | F | 41 | 41 | 0 | + | Gait disturbance, stuttering, hesitant speech | + | + | + | - | - | p.A781V |
| 12 | M | 39 | 42 | 3 | + | Tremor | + | + | - | + | - | pM766V |
| 13 | M | 44 | 45 | 1 | + | Cognitive impairment | + | + | + | - | - | p.A781V |
| 14 | F | 39 | 40 | 1 | + | Cognitive impairment | + | + | - | + | - | p.R782H |
Regional distributions of WMH in the CSF1R-related leukoencephalopathy and SIVaD groups
Mean WMH volumes in the frontal, parietal, temporal, and occipital lobes were significantly higher in the SIVaD group than in the CSF1R-related leukoencephalopathy group (p < 0.001). Mean WMH volumes in the frontal and parietal lobes in both groups were higher than those in the temporal and occipital lobes (Table 1). In the probability map showing overlaid WMH from all patients in each group, the CSF1R-related leukoencephalopathy group tended to involve the bilateral frontal and parietal areas more than other areas, whereas the SIVaD group showed diffuse involvement in the bilateral frontal, parietal, and temporal areas (Fig 1). Spatial location and frequency maps based on group comparisons also demonstrated profound WMH in the bilateral frontal and parietal areas, with relatively spared temporal and occipital areas, in the CSF1R-related leukoencephalopathy group. In contrast, the SIVaD group showed more diffuse WMH involvement in the bilateral frontal, parietal, and temporal areas (Fig 2).
Fig 1. Comparison of the probability maps.
It obtained after being overlaid WMH derived from FLAIR image of each patient with (A) CSF1R-related leukoencephalopathy and (B) SIVaD. The maps show WMH as color overlays (the color bar ranges from 0 to 1 reflecting the percentage of patients with WMH in each image voxel). Abbreviations: SIVaD, subcortical ischemic vascular dementia; WMH, white matter hyperintensity.
Fig 2. Topographical frequency maps of WMH in a single format.
The color-overlaid areas represent more significant WMH in (A) the CSF1R-related leukoencephalopathy group than in the SIVaD group and (B) the SIVaD group than in the CSF1R-related leukoencephalopathy group. Abbreviations: WMH, white matter hyperintensity; SIVaD, subcortical ischemic vascular dementia.
Regional WMH and cognitive functions in the CSF1R-related leukoencephalopathy group
The WMH volume of any cerebral region in the CSF1R-related leukoencephalopathy group was not found to be associated with attention, language, visuospatial, memory, or executive function (Table 3).
Table 3. Effects of regional WMH on cognitive function in the CSF1R-related leukoencephalopathy group.
| Regional WMH | Attention | Language and related function | Visuospatial function | Memory | Executive function | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| Average difference (SE) | P | Average difference (SE) | P | Average difference (SE) | P | Average difference (SE) | P | Average difference (SE) | P | |
| Frontal lobe | 0.09x10-4 (0.05x10-3) | 0.848 | 0.03x10-2 (0.07x10-2) | 0.674 | -0.02x10-2 (0.08x10-2) | 0.747 | -0.01x10-3 (0.02x10-2) | 0.963 | 0.03x10-1 (0.06x10-1) | 0.636 |
| Parietal lobe | 0.02x10-3 (0.09x10-3) | 0.815 | 0.04x10-2 (0.01x10-1) | 0.766 | -0.03x10-2 (0.02x10-1) | 0.852 | 0.01x10-2 (0.04x10-2) | 0.806 | 0.04x10-1 (0.04) | 0.929 |
| Temporal lobe | -0.05x10-2 (0.09x10-2) | 0.588 | 0.04x10-1 (0.01) | 0.769 | -0.05x10-1 (0.02) | 0.759 | -0.02x10-2 (0.04x10-1) | 0.973 | 0.18 (0.12) | 0.240 |
| Occipital lobe | 0.04x10-2 (0.01x10-1) | 0.727 | 0.04x10-1 (0.02) | 0.821 | -0.09x10-1 (0.02) | 0.664 | 0.01x10-1 (0.05x10-1) | 0.853 | 0.05 (0.14) | 0.742 |
Cortical thickness in the CSF1R-related leukoencephalopathy & SIVaD groups
Compared with the CN group participants, patients in both groups showed widespread cortical thinning across the bilateral frontotemporoparietal areas, with relatively spared occipital regions (Fig 3A and 3B, Table 4), although the CSF1R-related leukoencephalopathy group showed more widespread cortical thinning than the SIVaD group (Fig 3C and 3D, Table 4).
Fig 3. Three-dimensional statistical maps of cortical thickness.
Colored areas represent regions with significant cortical atrophy in (A) the CSF1R-related leukoencephalopathy group compared with that in the CN group, (B) the SIVaD group compared with that in the CN group (C) the CSF1R-related leukoencephalopathy group compared with that in the SIVaD group, and (D) the SIVaD group compared with that in the CSF1R-related leukoencephalopathy group. The covariates were age, sex, education, and intracranial volume. Abbreviations: CN, cognitively normal; SIVaD, subcortical ischemic vascular dementia.
Table 4. Comparison of cortical thickness between the CSF1R-related leukoencephalopathy and SIVaD groups.
| CSF1R-related leukoencephalopathy (N = 14) | SIVaD (N = 129) | CN (N = 121) | p-value | CSF1R-related leukoencephalopathy vs. SIVaD | CSF1R-related leukoencephalopathy vs. CN | SIVaD vs. CN | |
|---|---|---|---|---|---|---|---|
| Frontal lobe, mm | 2.19 ± 0.19 | 2.41 ± 0.14 | 2.47 ± 0.10 | <0.001 | <0.001 | <0.001 | <0.001 |
| Parietal lobe, mm | 2.00 ± 0.16 | 2.22 ± 0.13 | 2.25 ± 0.10 | <0.001 | <0.001 | 0.004 | <0.001 |
| Temporal lobe, mm | 2.41 ± 0.26 | 2.57 ± 0.17 | 2.62 ± 0.12 | <0.001 | <0.001 | <0.001 | <0.001 |
| Occipital lobe, mm | 1.95 ± 0.15 | 1.96 ± 0.11 | 1.96 ± 0.11 | 0.052 | 0.012 | 0.104 | 0.519 |
Abbreviations: CC, corpus callosum; SIVaD, subcortical ischemic vascular dementia; CN, cognitively normal
Moreover, the CSF1R-related leukoencephalopathy group demonstrated greater cortical thinning across the entire cerebral cortex, except in the bilateral ventromedial prefrontal cortices, left parahippocampal gyrus, right entorhinal cortex, bilateral lingual gyri, and posterior cingulate cortex, compared with the SIVaD group (Fig 3C and Table 4). Cortical thickness in the right entorhinal cortex was significantly lesser in the SIVaD group than in the CSF1R-related leukoencephalopathy group (Fig 3D and Table 4).
Corpus callosum in the CSF1R-related leukoencephalopathy and SIVaD groups
The mean volume of the CC in the CSF1R-related leukoencephalopathy group was significantly smaller than that in the SIVaD group (p = 0.007) and was the lowest in the genu. In contrast, the mean WMH volume in the CC in the CSF1R-related leukoencephalopathy group (2.86 ± 6.54 mL) tended to be larger than that in the SIVaD group (1.29 ± 0.61 mL), although the difference was not significant (p = 0.068). In the CSF1R-related leukoencephalopathy group, the mean WMH volume in the genu of the CC was the largest, followed by that in the body and splenium. In the SIVaD group, the mean WMH volume of body of the CC was the largest, followed by that of the genu and splenium. The regional proportion of WMH volume in the CC differed significantly between the CSF1R-related leukoencephalopathy and SIVaD groups (p < 0.001) (Table 5). The localized atrophy pattern of the CC on the CSF1R-related leukoencephalopathy compared with normal control differed from its regional WMH pattern; however, it did not differ between the CSF1R-related leukoencephalopathy and SIVaD groups.
Table 5. Comparison of corpus callosum volume with WMH volume.
| CSF1R-related leukoencephalopathy (N = 14) | SIVaD (N = 129) | CN (N = 121) | p-valuea* | |
| Mean CC volume, mL | 0.38 ± 0.45 | 0.43 ± 0.03 | 0.46 ± 0.01 | <0.001 |
| Genu, mL | 0.08 ± 0.01 | 0.90 ± 0.01 | 0.10 ± 0.00 | <0.001 |
| Body, mL | 0.23 ± 0.03 | 0.25 ± 0.02 | 0.27 ± 0.01 | <0.001 |
| Splenium, mL | 0.08 ± 0.01 | 0.09 ± 0.01 | 0.10 ± 0.00 | <0.001 |
| CSF1R-related leukoencephalopathy (N = 14) | SIVaD (N = 129) | p-valueb | ||
| Mean WMH volume of CC, mL | 2.86 ± 6.54 | 1.29 ± 0.61 | 0.068 | |
| Genu (%) | 67.78% | 26.50% | <0.001 | |
| Body (%) | 16.42% | 49.44% | ||
| Splenium (%) | 15.80% | 24.06% | ||
Abbreviations: CC, corpus callosum; SIVaD, subcortical ischemic vascular dementia; CN, cognitively normal; WMH, white matter hyperintensity
a: Kruskal-Wallis test
b: Mann-Whitney U test
*Post hoc analyses showed significant differences between all three groups.
Discussion
In this study, we characterized the differences in regional patterns of WMH and changes in cortical thickness between the CSF1R-related leukoencephalopathy and SIVaD groups using automated imaging techniques. The main findings of our study were as follows: (1) the CSF1R-related leukoencephalopathy group showed predominant subcortical white matter involvement of the bilateral frontoparietal areas, whereas the SIVaD group had more diffuse involvement of the bilateral frontoparietal and temporal regions; (2) thinning across almost the entire cortical region was greater in the CSF1R-related leukoencephalopathy group than in the SIVaD group; and (3) the CC was more atrophied and affected by WMH in the CSF1R-related leukoencephalopathy group than in the SIVaD group. Specifically, WMH in the CSF1R-related leukoencephalopathy group was most prominent in the genu of the CC. Overall, our findings provide new evidence regarding the different WMH distribution and cortical thinning patterns in CSF1R-related leukoencephalopathy compared with those in SIVaD.
Our first finding of predominant WMH in frontoparietal areas with a relatively spared temporal area in the CSF1R-related leukoencephalopathy group compared with that in the SIVaD group is in line with previous neuroimaging reports demonstrating prominent patchy or confluent and symmetric or asymmetric WMH in the frontal or frontoparietal areas in CSF1R-related leukoencephalopathy on visual inspection [6, 8, 23–25]. WMH on T2-weighted or FLAIR images in patients with CSF1R-related leukoencephalopathy has been pathologically confirmed as widespread white matter degeneration with loss of myelin and axons, abundant axonal spheroids, and pigmented macrophages [25–27]. According to the radiologic or pathologic staging of white matter lesions in CSF1R-related leukoencephalopathy, these lesions develop preferentially in the frontal or frontoparietal areas and extend to the occipitotemporal regions over time [28, 29]. As mean duration from onset to date of MRI of the patients in our study was 1.4 ± 1.7 years, which corresponds to early-stage disease considering the average disease duration of 6.8 ± 5.4 years [30], the finding of predominant WMH in the frontoparietal areas in our patients seems to be consistent with the temporal and spatial schemes of white matter lesions in CSF1R-related leukoencephalopathy. However, how and why CSF1R mutation triggers white matter degeneration in the frontal or frontoparietal areas remains unclear. Konno et al. recently suggested that CSF1R-related leukoencephalopathy is a primary microgliopathy, and that microglial dysfunction induced by CSF1R mutation is the primary cause of white matter degeneration [1]. Nevertheless, the pathophysiological links between aberrant microglia-mediated white matter degeneration and temporal/spatial neuropathological processes needs to be further elucidated. In comparison, the SIVaD group showed more widespread WMH without topographically specific patterns. WMH in SIVaD reflects partial demyelination, axonal loss, gliosis, and enlarged perivascular spaces, mainly induced by hypoxic stress or chronic ischemia caused by fibrohyalinotic or arteriolosclerotic vessel changes [31, 32]. A previous study using voxel-based approach showed the spatial distribution of WMH in elderly individuals depended on various vascular risk factors [33]. Therefore, WMH in our SIVaD group might be diffusely present in regions related to different pathogeneses or vulnerabilities based on patient-specific vascular risk factors [33–36]. Another possibility is that the SIVaD group was older and had a longer disease duration, which might have resulted in more diffuse WMH involvement [36, 37].
Second, although both groups showed significant cortical thinning throughout almost the entire cortex compared with that in CN controls, the thinning was more widespread in the CSF1R-related leukoencephalopathy group than in the SIVaD group. Cortical atrophy, one of the MRI characteristics of CSF1R-related leukoencephalopathy [29, 30], is usually associated with white matter lesions, but tends to appear as more generalized atrophy regardless of the localized pattern of WMH [38], which is in agreement with our results. Considering that neuropathological studies have shown that cortical neurons themselves are observed to be relatively spared, cortical atrophy or thinning in CSF1R-related leukoencephalopathy may be secondary to massive damage to both axons and myelin [38]. To date, no studies have reported cortical thinning patterns in CSF1R-related leukoencephalopathy. However, there has been one automated volumetric study comparing the brain volume fractions in CSF1R-related leukoencephalopathy with those in other diseases [39]. In that study, the authors revealed no significant differences of grey and white matter fractions between patients with CSF1R-related leukoencephalopathy and multiple sclerosis and speculated that their results could be attributable to a type II error, because the number of patients in each group was too small (n = 5). Another study exploring the imaging features of CSF1R-related leukoencephalopathy and cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) reported that predominant bifrontoparietal white matter lesions with cortical atrophy were more distinctive features of CSF1R-leukoencephalopathy than of CADASIL, whereas predominant temporal white matter lesions without cortical atrophy were more distinctive features of CADASIL than of CSF1R-leukoencephalopathy. However, the study also included only a small number of patients and used visual assessment of MRI scans [29]. In this regard, our study implemented more advanced methods and included relatively more patients, and thus may be more accurate and reliable than most previous studies that used visual ratings or scales and had smaller sample sizes.
Lastly, the CSF1R-related leukoencephalopathy group had a relatively smaller CC volume with more profound WMH, especially in the CC genu, than the SIVaD group. Although thinning of the CC with signal abnormalities is often observed in demyelinating diseases and other leukoencephalopathies [40–42], this is one of the most common MRI findings in CSF1R-related leukoencephalopathy [30] and has been included as one of the core features of the recently proposed diagnostic criteria for CSF1R-related leukoencephalopathy [43]. Similar to our study, Kinoshita et al. found that the CC index (CCI), determined based on the width and length of the CC, was significantly lower in CSF1R-related leukoencephalopathy than in VaD [44]. However, the CCI was manually calculated, and the WMH in the CC was not quantitatively measured; in addition, patients with VaD in that study were heterogeneous. Given that we applied automated quantification methods and our SIVaD group was rather homogeneous, our study and the study by Kinoshita et al. have limited comparability. Interestingly, we observed that the regional WMH pattern of the CC was not consistent with its localized CC atrophy pattern, but was consistent with the subcortical WMH pattern. This corresponds with the results of Sundal et al.’s study, which showed that WMH in the CC and CC atrophy (CCA) were not correlated with each other [38]. Regarding the underlying mechanism of CCA in SIVaD, the direct or indirect impact of vascular injury on subcortical white matter is believed to contribute to CCA, while “peculiar morphological alteration of microglia” has been suggested as the pathologic basis of CCA in CSF1R-related leukoencephalopathy [45].
The present study had a few limitations. First, the number of patients with CSF1R-related leukoencephalopathy was relatively small because of the rarity of this disease [8]. In addition, small number of patients made it difficult to identify the most discriminative MRI feature between the two groups. Second, most enrolled patients happened to have early-stage CSF1R-related leukoencephalopathy; it is possible that patients with a later stage of the disease would show more confluent WMH with severe cortical thinning on brain MRI, which may make it more challenging to distinguish CSF1R-related leukoencephalopathy from SIVaD. Finally, because genetic testing was not performed in patients with SIVaD, there might be a possibility that the patients with genetic leukoencephalopathy or genetic vascular leukoencephalopathy were included in the SIVaD group, even though they met the clinical diagnostic criteria for SIVaD. Despite these limitations, the strength of our study is that it is the first to use automated quantitative evaluations for WMH and cortical thinning in CSF1R-related leukoencephalopathy compared with those of SIVaD. Therefore, our findings can be expected to be more reliable than those based on the visual assessment methods used in most previous studies.
Conclusion
In conclusion, we observed predominantly bilateral frontoparietal WMH patterns, profound cortical thinning, and CCA with WMH as unique MRI features of CSF1R-related leukoencephalopathy rather than SIVaD. To the best of our knowledge, this is the first study to use automated MR measurements to capture WMH, cortical thinning, and CCA with signal changes in CSF1R-related leukoencephalopathy. Our study highlights the capacity and utility of automated MRI technique for differentiating CSF1R-related leukoencephalopathy from other white matter diseases, even with a relatively small number of patients.
Supporting information
(DOCX)
Acknowledgments
The authors acknowledge the patients and their families for their cooperation.
Data Availability
All relevant data are within the manuscript and its Supporting Information files.
Funding Statement
this research was supported by the National Institute of Health research project [No. 2021-ER1004-01], and the Korea National Institute of Health (KNIH) research project [No. 2024-ER1001-00]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
References
- 1.Konno T, Kasanuki K, Ikeuchi T, Dickson DW, Wszolek ZK. CSF1R-related leukoencephalopathy: A major player in primary microgliopathies. Neurology. 2018;91(24):1092–104. doi: 10.1212/WNL.0000000000006642 ; PubMed Central PMCID: PMC6329328. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Jin C, Washimi Y, Yoshida K, Hashizume Y, Yazawa I. Characterization of spheroids in hereditary diffuse leukoencephalopathy with axonal spheroids. J Neurol Sci. 2015;352(1–2):74–8. doi: 10.1016/j.jns.2015.03.033 . [DOI] [PubMed] [Google Scholar]
- 3.Di Donato I, Stabile C, Bianchi S, Taglia I, Mignarri A, Salvatore S, et al. A Novel CSF1R Mutation in a Patient with Clinical and Neuroradiological Features of Hereditary Diffuse Leukoencephalopathy with Axonal Spheroids. J Alzheimers Dis. 2015;47(2):319–22. doi: 10.3233/JAD-150097 . [DOI] [PubMed] [Google Scholar]
- 4.Tian WT, Zhan FX, Liu Q, Luan XH, Zhang C, Shang L, et al. Clinicopathologic characterization and abnormal autophagy of CSF1R-related leukoencephalopathy. Transl Neurodegener. 2019;8:32. doi: 10.1186/s40035-019-0171-y ; PubMed Central PMCID: PMC6886209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Axelsson R, Roytta M, Sourander P, Akesson HO, Andersen O. Hereditary diffuse leucoencephalopathy with spheroids. Acta Psychiatr Scand Suppl. 1984;314:1–65. . [PubMed] [Google Scholar]
- 6.Makary MS, Awan U, Kisanuki YY, Slone HW. Adult-onset leukoencephalopathy with axonal spheroids and pigmented glia: Clinical and imaging characteristics. Neuroradiol J. 2019;32(2):139–42. doi: 10.1177/1971400918822136 ; PubMed Central PMCID: PMC6410452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sundal C, Wszolek ZK. CSF1R-Related Adult-Onset Leukoencephalopathy with Axonal Spheroids and Pigmented Glia. In: Adam MP, Everman DB, Mirzaa GM, Pagon RA, Wallace SE, Bean LJH, et al., editors. GeneReviews((R)). Seattle (WA)1993. [Google Scholar]
- 8.Sundal C, Lash J, Aasly J, Oygarden S, Roeber S, Kretzschman H, et al. Hereditary diffuse leukoencephalopathy with axonal spheroids (HDLS): a misdiagnosed disease entity. J Neurol Sci. 2012;314(1–2):130–7. doi: 10.1016/j.jns.2011.10.006 ; PubMed Central PMCID: PMC3275663. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Saitoh BY, Yamasaki R, Hiwatashi A, Matsushita T, Hayashi S, Mitsunaga Y, et al. Discriminative clinical and neuroimaging features of motor-predominant hereditary diffuse leukoencephalopathy with axonal spheroids and primary progressive multiple sclerosis: A preliminary cross-sectional study. Mult Scler Relat Disord. 2019;31:22–31. doi: 10.1016/j.msard.2019.03.008 . [DOI] [PubMed] [Google Scholar]
- 10.Wardlaw JM, Valdes Hernandez MC, Munoz-Maniega S. What are white matter hyperintensities made of? Relevance to vascular cognitive impairment. J Am Heart Assoc. 2015;4(6):001140. doi: 10.1161/JAHA.114.001140 ; PubMed Central PMCID: PMC4599520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.van Dijk EJ, Prins ND, Vrooman HA, Hofman A, Koudstaal PJ, Breteler MM. Progression of cerebral small vessel disease in relation to risk factors and cognitive consequences: Rotterdam Scan study. Stroke. 2008;39(10):2712–9. doi: 10.1161/STROKEAHA.107.513176 . [DOI] [PubMed] [Google Scholar]
- 12.Shinkawa A, Ueda K, Kiyohara Y, Kato I, Sueishi K, Tsuneyoshi M, et al. Silent cerebral infarction in a community-based autopsy series in Japan. The Hisayama Study. Stroke. 1995;26(3):380–5. doi: 10.1161/01.str.26.3.380 . [DOI] [PubMed] [Google Scholar]
- 13.Gow AJ, Bastin ME, Munoz Maniega S, Valdes Hernandez MC, Morris Z, Murray C, et al. Neuroprotective lifestyles and the aging brain: activity, atrophy, and white matter integrity. Neurology. 2012;79(17):1802–8. doi: 10.1212/WNL.0b013e3182703fd2 . [DOI] [PubMed] [Google Scholar]
- 14.Fazekas F, Kleinert R, Offenbacher H, Schmidt R, Kleinert G, Payer F, et al. Pathologic correlates of incidental MRI white matter signal hyperintensities. Neurology. 1993;43(9):1683–9. doi: 10.1212/wnl.43.9.1683 . [DOI] [PubMed] [Google Scholar]
- 15.Erkinjuntti T, Inzitari D, Pantoni L, Wallin A, Scheltens P, Rockwood K, et al. Research criteria for subcortical vascular dementia in clinical trials. J Neural Transm Suppl. 2000;59:23–30. doi: 10.1007/978-3-7091-6781-6_4 . [DOI] [PubMed] [Google Scholar]
- 16.Schmidt P. Bayesian inference for structured additive regression models for large-scale problems with applications to medical imaging: lmu; 2017. [Google Scholar]
- 17.Schmidt P, Gaser C, Arsic M, Buck D, Forschler A, Berthele A, et al. An automated tool for detection of FLAIR-hyperintense white-matter lesions in Multiple Sclerosis. Neuroimage. 2012;59(4):3774–83. doi: 10.1016/j.neuroimage.2011.11.032 . [DOI] [PubMed] [Google Scholar]
- 18.Jenkinson M, Bannister P, Brady M, Smith S. Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage. 2002;17(2):825–41. doi: 10.1016/s1053-8119(02)91132-8 . [DOI] [PubMed] [Google Scholar]
- 19.Fischl B. FreeSurfer. Neuroimage. 2012;62(2):774–81. doi: 10.1016/j.neuroimage.2012.01.021 ; PubMed Central PMCID: PMC3685476. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Fischl B, Dale AM. Measuring the thickness of the human cerebral cortex from magnetic resonance images. Proc Natl Acad Sci U S A. 2000;97(20):11050–5. doi: 10.1073/pnas.200033797 ; PubMed Central PMCID: PMC27146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ahn HJ, Chin J, Park A, Lee BH, Suh MK, Seo SW, et al. Seoul Neuropsychological Screening Battery-dementia version (SNSB-D): a useful tool for assessing and monitoring cognitive impairments in dementia patients. J Korean Med Sci. 2010;25(7):1071–6. doi: 10.3346/jkms.2010.25.7.1071 ; PubMed Central PMCID: PMC2890886. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kim H, Na DL. Normative data on the Korean version of the Boston Naming Test. J Clin Exp Neuropsychol. 1999;21(1):127–33. doi: 10.1076/jcen.21.1.127.942 . [DOI] [PubMed] [Google Scholar]
- 23.Freeman SH, Hyman BT, Sims KB, Hedley-Whyte ET, Vossough A, Frosch MP, et al. Adult onset leukodystrophy with neuroaxonal spheroids: clinical, neuroimaging and neuropathologic observations. Brain Pathol. 2009;19(1):39–47. doi: 10.1111/j.1750-3639.2008.00163.x ; PubMed Central PMCID: PMC2757058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Van Gerpen JA, Wider C, Broderick DF, Dickson DW, Brown LA, Wszolek ZK. Insights into the dynamics of hereditary diffuse leukoencephalopathy with axonal spheroids. Neurology. 2008;71(12):925–9. doi: 10.1212/01.wnl.0000325916.30701.21 ; PubMed Central PMCID: PMC2843529. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Mao C, Zhou L, Zhou L, Yang Y, Niu J, Li J, et al. Biopsy histopathology in the diagnosis of adult-onset leukoencephalopathy with axonal spheroids and pigmented glia (ALSP). Neurol Sci. 2020;41(2):403–9. doi: 10.1007/s10072-019-04116-7 . [DOI] [PubMed] [Google Scholar]
- 26.Kim EJ, Shin JH, Lee JH, Kim JH, Na DL, Suh YL, et al. Adult-onset leukoencephalopathy with axonal spheroids and pigmented glia linked CSF1R mutation: Report of four Korean cases. J Neurol Sci. 2015;349(1–2):232–8. doi: 10.1016/j.jns.2014.12.021 . [DOI] [PubMed] [Google Scholar]
- 27.Terada S, Ishizu H, Yokota O, Ishihara T, Nakashima H, Kugo A, et al. An autopsy case of hereditary diffuse leukoencephalopathy with spheroids, clinically suspected of Alzheimer’s disease. Acta Neuropathol. 2004;108(6):538–45. doi: 10.1007/s00401-004-0920-5 . [DOI] [PubMed] [Google Scholar]
- 28.Oyanagi K, Kinoshita M, Suzuki-Kouyama E, Inoue T, Nakahara A, Tokiwai M, et al. Adult onset leukoencephalopathy with axonal spheroids and pigmented glia (ALSP) and Nasu-Hakola disease: lesion staging and dynamic changes of axons and microglial subsets. Brain Pathol. 2017;27(6):748–69. doi: 10.1111/bpa.12443 ; PubMed Central PMCID: PMC8029200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Bender B, Klose U, Lindig T, Biskup S, Nagele T, Schols L, et al. Imaging features in conventional MRI, spectroscopy and diffusion weighted images of hereditary diffuse leukoencephalopathy with axonal spheroids (HDLS). J Neurol. 2014;261(12):2351–9. doi: 10.1007/s00415-014-7509-2 . [DOI] [PubMed] [Google Scholar]
- 30.Konno T, Yoshida K, Mizuno T, Kawarai T, Tada M, Nozaki H, et al. Clinical and genetic characterization of adult-onset leukoencephalopathy with axonal spheroids and pigmented glia associated with CSF1R mutation. Eur J Neurol. 2017;24(1):37–45. doi: 10.1111/ene.13125 ; PubMed Central PMCID: PMC5215554. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Gouw AA, Seewann A, van der Flier WM, Barkhof F, Rozemuller AM, Scheltens P, et al. Heterogeneity of small vessel disease: a systematic review of MRI and histopathology correlations. J Neurol Neurosurg Psychiatry. 2011;82(2):126–35. doi: 10.1136/jnnp.2009.204685 . [DOI] [PubMed] [Google Scholar]
- 32.Black S, Gao F, Bilbao J. Understanding white matter disease: imaging-pathological correlations in vascular cognitive impairment. Stroke. 2009;40(3 Suppl):S48–52. doi: 10.1161/STROKEAHA.108.537704 . [DOI] [PubMed] [Google Scholar]
- 33.Rostrup E, Gouw AA, Vrenken H, van Straaten EC, Ropele S, Pantoni L, et al. The spatial distribution of age-related white matter changes as a function of vascular risk factors—results from the LADIS study. Neuroimage. 2012;60(3):1597–607. doi: 10.1016/j.neuroimage.2012.01.106 . [DOI] [PubMed] [Google Scholar]
- 34.Griffanti L, Jenkinson M, Suri S, Zsoldos E, Mahmood A, Filippini N, et al. Classification and characterization of periventricular and deep white matter hyperintensities on MRI: A study in older adults. Neuroimage. 2018;170:174–81. doi: 10.1016/j.neuroimage.2017.03.024 . [DOI] [PubMed] [Google Scholar]
- 35.Ryu WS, Woo SH, Schellingerhout D, Chung MK, Kim CK, Jang MU, et al. Grading and interpretation of white matter hyperintensities using statistical maps. Stroke. 2014;45(12):3567–75. doi: 10.1161/STROKEAHA.114.006662 . [DOI] [PubMed] [Google Scholar]
- 36.Pantoni L. Cerebral small vessel disease: from pathogenesis and clinical characteristics to therapeutic challenges. Lancet Neurol. 2010;9(7):689–701. doi: 10.1016/S1474-4422(10)70104-6 . [DOI] [PubMed] [Google Scholar]
- 37.Schmidt R, Ropele S, Enzinger C, Petrovic K, Smith S, Schmidt H, et al. White matter lesion progression, brain atrophy, and cognitive decline: the Austrian stroke prevention study. Ann Neurol. 2005;58(4):610–6. doi: 10.1002/ana.20630 . [DOI] [PubMed] [Google Scholar]
- 38.Sundal C, Van Gerpen JA, Nicholson AM, Wider C, Shuster EA, Aasly J, et al. MRI characteristics and scoring in HDLS due to CSF1R gene mutations. Neurology. 2012;79(6):566–74. doi: 10.1212/WNL.0b013e318263575a ; PubMed Central PMCID: PMC3413763. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Granberg T, Hashim F, Andersen O, Sundal C, Karrenbauer VD. Hereditary diffuse leukoencephalopathy with spheroids—a volumetric and radiological comparison with multiple sclerosis patients and healthy controls. Eur J Neurol. 2016;23(4):817–22. doi: 10.1111/ene.12948 . [DOI] [PubMed] [Google Scholar]
- 40.Yamauchi H, Pagani M, Fukuyama H, Ouchi Y, Nagahama Y, Matsuzaki S, et al. Progression of atrophy of the corpus callosum with deterioration of cerebral cortical oxygen metabolism after carotid artery occlusion: a follow up study with MRI and PET. J Neurol Neurosurg Psychiatry. 1995;59(4):420–6. doi: 10.1136/jnnp.59.4.420 ; PubMed Central PMCID: PMC486080. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Evangelou N, Konz D, Esiri MM, Smith S, Palace J, Matthews PM. Regional axonal loss in the corpus callosum correlates with cerebral white matter lesion volume and distribution in multiple sclerosis. Brain. 2000;123 (Pt 9):1845–9. doi: 10.1093/brain/123.9.1845 . [DOI] [PubMed] [Google Scholar]
- 42.Lee DY, Fletcher E, Martinez O, Zozulya N, Kim J, Tran J, et al. Vascular and degenerative processes differentially affect regional interhemispheric connections in normal aging, mild cognitive impairment, and Alzheimer disease. Stroke. 2010;41(8):1791–7. doi: 10.1161/STROKEAHA.110.582163 ; PubMed Central PMCID: PMC2922914. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Konno T, Yoshida K, Mizuta I, Mizuno T, Kawarai T, Tada M, et al. Diagnostic criteria for adult-onset leukoencephalopathy with axonal spheroids and pigmented glia due to CSF1R mutation. Eur J Neurol. 2018;25(1):142–7. doi: 10.1111/ene.13464 ; PubMed Central PMCID: PMC5741468. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Kinoshita M, Kondo Y, Yoshida K, Fukushima K, Hoshi K, Ishizawa K, et al. Corpus callosum atrophy in patients with hereditary diffuse leukoencephalopathy with neuroaxonal spheroids: an MRI-based study. Intern Med. 2014;53(1):21–7. doi: 10.2169/internalmedicine.53.0863 . [DOI] [PubMed] [Google Scholar]
- 45.Kinoshita M, Oyanagi K, Kondo Y, Ishizawa K, Ishihara K, Yoshida M, et al. Pathologic basis of the preferential thinning of thecorpus callosum in adult-onset leukoencephalopathy with axonal spheroids and pigmented glia (ALSP). eNeurologicalSci. 2021;22:100310. doi: 10.1016/j.ensci.2021.100310 ; PubMed Central PMCID: PMC7844436. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(DOCX)
Data Availability Statement
All relevant data are within the manuscript and its Supporting Information files.