Alterations of the cerebral cortex in sporadic small vessel disease: A systematic review of in vivo MRI data

Abstract

Cerebral small vessel diseases of the brain are a major determinant of cognitive impairment in the elderly. In small vessel diseases, the most easily identifiable lesions, both at post-mortem evaluation and magnetic resonance imaging, lie in subcortical areas. However, recent results obtained post-mortem, particularly in severe cases, have highlighted the burden of cortex lesions such as microinfarcts and diffuse neuronal loss. The recent development of image post-processing methods allows now assessing in vivo multiple aspects of the cerebral cortex. This systematic review aimed to analyze in vivo magnetic resonance imaging studies evaluating cortex alterations at different stages of small vessel diseases. Studies assessing the relationships between small vessel disease magnetic resonance imaging markers obtained at the subcortical level and cortex estimates were reviewed both in community-dwelling elderly and in patients with symptomatic small vessel diseases. Thereafter, studies analyzing cortex estimates in small vessel disease patients compared with healthy subjects were evaluated. The results support that important cortex alterations develop along the course of small vessel diseases independently of concomitant neurodegenerative processes. Easy detection and quantification of cortex changes in small vessel diseases as well as understanding their underlying mechanisms are challenging tasks for better understanding cognitive decline in small vessel diseases.

Keywords

Cerebral small vessel disease magnetic resonance imaging cortex cortical thickness cortical folding voxel-based morphometry

Introduction

Age- and hypertension-related cerebral small vessel diseases (SVD) are a major determinant of cognitive decline and disability in the elderly.¹ The clinical and imaging spectra of SVD are wide. More than 80% of elderly subjects present magnetic resonance imaging (MRI) markers of SVD such as white matter hyperintensities (WMH), lacunes of presumed vascular origin, microbleeds, or dilated perivascular spaces.^1,2 In the general population, these markers were found associated with gait disturbances and cognitive performances and with a higher risk of stroke and dementia.^1,2 Besides, SVD are actually responsible for one fifth of ischemic strokes and represent a major cause of dementia after Alzheimer’s Disease (AD).² Thus, SVD are now considered a major public health issue.² Recently, an international collaborative effort refined the terminology used in MRI examination in the setting of SVD to facilitate and promote the generalization of research results in this field.³

In SVD, lesions most easily identified both at neuropathological evaluation and on conventional MRI lie in deep subcortical brain areas. This explains why the term ‘subcortical ischemic vascular dementia’ is often used to describe the severe cognitive decline presumably related to SVD,⁴ in line with a classification proposed in the 1970s to differentiate various subtypes of neurodegenerative diseases and separating cortical from subcortical dementias.⁵

There is now accumulating evidence in the literature supporting that the cortex is actually involved in SVD. Microinfarcts are commonly found in the cortex of elderly subjects at autopsy. They are associated with other markers of SVD and may represent a major determinant of cognitive impairment.^6–9 Cortical atrophy, of comparable severity to that of AD, has been also demonstrated in neuropathological studies of a small number of SVD cases with dementia.¹⁰ Although in vivo imaging is not yet capable of identifying the smallest lesions within the cortical mantle, recent results obtained using ultra-high field MRI showed that the largest intracortical infarcts become now identifiable.⁹ Although these data support that the cerebral cortex may play a key role in SVD, the natural history of cortical changes in SVD and their underlying mechanisms remain undetermined.

The purpose of this review was to analyze data estimating in vivo the extent of cortex alterations associated with the wide spectrum of SVD lesions. To do so, we searched for articles that evaluated the cortical alterations observed at different stages of SVD using validated quantitative MRI parameters. In contrast to severe cases for which the diagnosis of SVD can be made according to different sets of validated criteria, there is little agreement regarding the diagnosis of SVD in the absence of stroke, disability, or severe cognitive alterations. Thus, we organized our review into three main parts: (a) studies evaluating the relationships between cortex alterations and MRI markers of SVD at the subcortical level in community-dwelling subjects; (b) identical studies in patients with symptomatic SVD or high vascular risk profile; and (c) studies that compared patients with SVD to age- and sex-matched controls. Given that concomitant AD may promote cortex alterations before the apparition of clinical symptoms, a specific attention was paid to this aspect.

Material and methods

This systematic review follows the PRISMA statements (http://www.prisma-statement.org). Studies were identified by searching PubMed and through systematic scanning of reference lists. Articles from our personal databases were added to the search strategy. Only articles published in English were considered. The last search was run on June 23, 2015. We used the following search terms: white matter lesions, white matter hyperintensity, white matter disease, white matter damage, lacunes, lacunar infarcts, microbleeds, small vessel disease, cerebral microangiopathy, leukoaraiosis, leukopathy, vascular cognitive impairment, subcortical ischemic vascular disease, voxel-based morphometry, cortical thickness, gray matter, grey matter, cortical atrophy, sulcal morphology, cortical morphology, cortical gyri, cortical sulci, cortical gyrification, cortex morphology, and cortical folding. By contrast, multiple sclerosis, progressive multifocal leukoencephalopathy and AIDS were exclusion search terms. The exact search algorithm is provided as supplementary material.

Given the complex architecture of the cerebral cortex, with such a highly folded structure and its small thickness relatively to its surface area, specific MRI protocols are needed to accurately segment the cortex and capture its variability. Thus, only studies based on recognized post-processing methodology relying on high-resolution T1 MRI sequences (i.e. study of cortical thickness, of cortical gray matter using either voxel-based methods or direct quantification, or of sulcal morphology—see Figure 1) were considered. Studies based on high-resolution MRI but evaluating gray matter globally (with inclusion of basal ganglia and cortex without distinction in analyses) and those based on computed tomography were not included. Only studies comparing a SVD group to a control group or studies investigating relationships between a validated subcortical MRI marker of SVD and cortical alterations were considered. Studies including mostly AD patients were excluded since the pathophysiology of WMH in this subgroup remains controversial.¹¹ We also chose not to evaluate cortical alterations reported in the hippocampus given the very limited local gray-white matter contrast that prevents studying reliably the alterations of cortex in this structure. Studies performed in the CADASIL model were also not included in the present analyses but related results were confronted to those obtained in sporadic SVD when necessary.

Figure 1.

Main methods used to evaluate cortex alterations using in vivo MRI. Schematics of the different methods used to evaluate the cerebral cortex in vivo. Methods with registration involve positioning the MRI of different subjects in a common space to allow between subjects comparisons. This can be done through volume or surface-based registration. Methods without registration consist in extracting brain parameters from MRI in the native space of acquisition.

Results

We screened 934 articles, among which 78 full texts were analyzed in details (see Figure 2). Twenty-seven articles fulfilled the inclusion criteria. However, in some articles, data from the same cohort were repeatedly reported; therefore, the final number of distinct studies was 25. Nine were case–control studies, the relationships between the different subcortical MRI markers and cortex measures were evaluated in 16 others, 4 of which were longitudinal studies.

Figure 2.

Flow diagram.

Association between MRI markers of SVD at the subcortical level and estimates of cortex alterations in community-dwelling subjects (see Table 1)

We found eight studies, in which the relationships between MRI markers at the subcortical level and cortex alterations in community dwelling elderly subjects were analyzed.^11–18 Subjects with a medical history of neurological or psychiatric illnesses were excluded from half of studies.^15–18

Table 1.

Relationships between subcortical MRI markers and cortical measures in community-dwelling individuals (ordered by sample size N).

Study	Subject description							MRI			Statistical analysis			Results
Study	Study type	N	Age (mean ± SD)	Gender (female, %)	Cognitive status^a, MMSE (mean ± SD)	Cortical or territory infarcts	Associated prodromal AD	Field strength Protocol	Cortex measure	Subcortical marker assessed	Effect of age and gender	Adjusted for mult.comp. if needed	Association with cortical measure for	FL	PL		TL	OL
Raji et al.¹²	Cross-sectional	740	78.3 ± 3.9	432 (58%)	Normal,NA	Not excluded	Not controlled	1.5T 3DT1, T2	VBM	Visual grading of WMH	Adjusted	Yes	WMH: Yes Lacunes: NA Microbleeds: NA	✓			✓
Wen et al.¹¹	Cross-sectional	397	63 ± 1	179 (45%)	Normal, 29.3 ± 1.2	Not excluded	Not controlled	1.5T 3DT1, FLAIR	VBM	Volume of WMH	Adjusted	Yes	WMH: Yes Lacuens: NA Microbleeds: NA	✓	✓		✓
Liu et al.¹³	Longitudinal	185	77. ± 4.4	Not detailed	No dementia, 28.1 ± 1.3	Not excluded	Not controlled	3T 3DT1, FLAIR	Sulcal width Cortical thickness Cortical gray matter volume	Volume of WMH	Adjusted	Yes	WMH: No^b Lacunes: NA Microbleeds: NA	No regional analysis
Rossi et al.¹⁴	Cross-sectional	133	54.7 ± 9.8	94 (70%)	NA,28.9 ± 1.3	Excluded	Not controlled	1T 3DT1, FLAIR, 1Tesla	VBM	Presence/absence of WMH	Adjusted	No	WMH: Yes Lacunes: NA Microbleeds: NA	✓
Ota et al.¹⁵	Longitudinal	74	72.3 ± 3.8	35 (47%)	Normal, 27.8 ± 2.4	Excluded	Not controlled	1.5T 3DT1, T2	VBM	Visual rating of WMH	Adjusted	No	WMH: Yes^c Lacunes: NA Microbleeds: NA			✓			✓
Wirth et al.¹⁶	Cross-sectional	72	74.9 ± 5.7	48 (67%)	Normal, 28.8 ± 1.4	Not excluded	PET-FDG, PET-PIB	1.5T 3DT1	Cortical thickness	Volume of WML on T1	Adjusted	NA	WMH: Yes Lacunes: NA Microbleeds: NA	No regional analysis
Eckert et al.¹⁷	Cross-sectional	42	19-79	27 (64%)	Normal, ≥ 27	Not excluded	Not controlled	3T 3DT1, FLAIR	VBM	Visual rating of WMH	Adjusted	Yes	WMH: Yes Lacunes: NA Microbleeds: NA	✓
Kochunov et al.¹⁸	Cross-sectional	31	67.3 ± 6.7	20 (65%)	Normal,NA	Not measured	Not controlled	3T 3DT1, T2	Sulcal width Cortical thickness	Volume of WMH	Adjusted	NA	WMH: Yes (SW)/No (CT) Lacunes: NA Microbleeds: NA	No regional analysis

See supplementary material for detailed criteria.

Relationship between subcortical MRI markers at baseline and atrophy rate.

Relationship between subcortical MRI marker progression and atrophy rate.

SD: standard deviation; AD: Alzheimer’s disease; NA: not available; FL: frontal lobe; PL: parietal lobe; TL: temporal lobe; OL: occipital lobe; MMSE: mini mental state examination; mult.comp.: multiple comparisons; WMH: white matter hyperintensities of presumed vascular origin; VBM: voxel-based morphometry; SW: sulcal width; CT: cortical thickness.

Sample sizes were highly heterogeneous, ranging from 31 to 740 subjects.^11–18 Mean age was around 75 years in four studies,^12,13,15,16 lower in three others (55 in one and 65 in two),^11,14,18 while age ranged from 19 to 79 years in one.¹⁷

In all studies, subjects were scanned using 3DT1 1-mm isotropic sequences. Cortex measures were obtained using different techniques, with or without registration, as illustrated in Figure 1. Different methodologies were used to evaluate the burden of WMH: quantitative measures of WMH obtained by automated segmentation,^11,13,16,18 semi-quantitative scales,^12,15,17 or visual categorization according to WMH location.¹⁴ Lacunes or microbleeds were not evaluated in any of the studies.

The relationships between voxel-based morphometry (VBM) estimates of total cortical gray matter and the burden of WMH were evaluated in four cross-sectional studies that always showed significant inverse associations, independent of age.^11,12,14,17 This association was found independent of ApoE genotype in one study.¹² Results were not corrected for multiple comparisons in another one.¹⁴ Finally, in a longitudinal study of 74 subjects, a significant inverse association was also detected between VBM estimates of total cortical gray matter changes and the progression of WMH during a mean follow-up of 3.8 years.¹⁵

The relationships between cortical thickness and the volume of WMH were tested in three cross-sectional studies.^13,16,18 A significant inverse relationship was reported in one.¹⁶ In this study, subjects were divided between “abnormal” and “normal” cortical thickness.¹⁶ The threshold between “normal” and “abnormal” cortical thickness was pre-determined according to previous results obtained by contrasting a population of cognitively normal elderly subjects defined clinically and biologically (CDR 0 and normal memory function together with negative Florbetapir-PET imaging or biomarkers obtained through lumbar puncture) to a population of mild AD defined clinically (NINCDS ADRDA criteria, MMSE score between 20 and 26 and CDR of 0.5 or 1). A significantly higher volume of WMH was observed in subjects with “abnormal” cortical thickness compared with subjects with “normal” cortical thickness, independent of age,¹⁶ a finding that could not be explained by associated prodromal AD in the “abnormal” group as Aβ deposition on 11C-PIB-PET imaging did not differ between the two groups. By contrast, there was no significant association between cortical thickness and the volume of WMH in two other studies.^13,18

The relationships between sulcal width and the volume of WMH were evaluated in two different studies and revealed a significant association independent of age in one,¹⁸ but not in the other.¹³ However, the interpretation of these non-significant results is difficult as these analyses were not the primary aim of those studies and detailed results were not provided.^13,18

In summary, significant relationships between cortical gray matter volume and WMH are consistently observed with VBM approaches in community-dwelling elderly but links between WMH and other cortical parameters are inconsistently detected. The cortical impact of lacunes and/or microbleeds is not evaluated yet in these populations and only few data are available to determine whether the cortical alterations are actually secondary to effects of SVD rather than to concomitant prodromal AD.

Association between MRI markers of SVD at the subcortical level and estimates of cortex alterations in patients with symptomatic SVD or with high cardio- or neuro-vascular risk profile (see Table 2)

We found 10 studies based on five different cohorts of individuals with symptomatic SVD or high cardio- or neuro-vascular risk profile in which the links between conventional MRI markers measured at the subcortical level and cortex alterations were evaluated.^19–28 The profile of patients included in these studies was largely heterogeneous.

Table 2.

Relationships between subcortical MRI markers and cortical measures in populations at high cardio-neurovascular risk (ordered by sample sizes N).

Study	Subject description							MRI			Statistical analysis			Results
Study	Recruitment criteria	N	Age (mean ± SD)	Gender (female, %)	Cognitive status^a, MMSE (mean ± SD)	Cortical or territory infarcts	Associated prodromal AD	Field strength Protocol	Cortex measure	Subcortical markers assessed	Effect of age and gender	Adjusted for mult.comp.	Association with cortical measure for	FL	PL	TL	OL
Appelman et al.¹⁹	Symptomatic atheroma (SMART-MR cohort)	840	58.0 ± 10.1	168 (20%)	Independently living, Not provided	Excluded	Not controlled	1.5T T1, T2, FLAIR, IR, 4 mm	Volume of cortical gray matter	Volume of WMH, Presence of lacunes	Adjusted	NA	WMH: Yes Lacunes: No Microbleeds: NA	No regional analysis
Geerlings et al.²⁰	Symptomatic atheroma (SMART-MR cohort)	808	58 ± 10	About 20%	Independently living, Not provided	Only patients with infarcts	Not controlled	1.5T T1, T2, FLAIR, IR, 4 mm	Volume of cortical gray matter	Volume of WMH, > 1 or < 2 lacunes	Adjusted	NA	WMH: NA Lacunes: No Cortical infarcts: Yes Microbleeds: NA	No regional analysis
Kloppenborg et al.²¹	Symptomatic atheroma (SMART-MR cohort)	565	57 ± 9.5	106 (18.8%)	Independently living, Not provided	Excluded	Not controlled	1.5T T1, T2, FLAIR, IR, 4 mm	Volume of cortical gray matter	Volume of WMH, Presence of lacunes	Adjusted	NA	WMH: Yes^b/Yes^c, Lacunes: No^b/Yes^c, Microbleeds: NA	No regional analysis
Reid et al.²²	Incidental WMH or lacunes (RUN-DMC cohort)	485	64.2 ± 3.3	215 (44%)	No dementia, 28.1 ± 1.5	Included	Not controlled	1.5T 3DT1, FLAIR	Cortical thickness	Volume of WMH	Adjusted	Yes	WMH: No Lacunes: NA Microbleeds: NA	Significant links in secondary analyses in different age groups
Tuladhar et al.²³	Incidental WMH or lacunes (RUN-DMC cohort)	426	65.1 ± 7.6	198 (46%)	No dementia, 29 (IQR 27–30)	Excluded	Not controlled	3DT1, FLAIR	Cortical thickness	Volume of WMH, Number of lacunes	Adjusted	Yes	WMH: Yes Lacunes: NA Microbleeds: NA	✓		✓
Mok et al.²⁴ and Xiong et al.²⁵	Ischemic stroke with confluent WMH	100	75.2 ± 7.6	48 (48%)	Normal to demented, 21.6 ± 5.4 (10–30)	Included	PIB-PET imaging in seven patients	1.5T T1, FLAIR, T2, T2^b, 5 mm	Volume of cortical gray matter	Volume of WMH, Total volume of infarcts, Number of microbleeds	Adjusted	NA	WMH: Yes Lacunes: No Microbleeds: No	✓
Fein et al.²⁶	Lacunes on MRI	87	73 ± 7	31 (36%)	Normal to demented, 26 ± 3	Excluded	Eight autopsy cases, none of them meet NIA-Reagan criteria for AD	1.5T High-res 2DT1,T2, PD	Volume of cortical gray matter	Volume of WMH, Volume and number of lacunes	Not adjusted	NA	WMH: Yes Lacunes: No Microbleeds: NA	Some negative regional associations between lacunes and cortical gray matter volume
Grau-Olivares et al.^27,28	Stroke related to recent SSI, cross-sectional	40	70.8 ± 12.2	21 (53%)	Normal vs. MCI	Excluded	Not controlled	1.5T 3DT1 3 mm, FLAIR	Voxel-based cortical gray matter analyses	Visual rating of WMH, Number of lacunes	Not adjusted	Yes	WMH: Yes Lacunes: NA Microbleeds: NA	✓		✓
Grau-Olivares et al.^27,28	Stroke related to recent SSI, longitudinal	30	68.4 ± 11.7	12 (30%)	Normal vs. MCI	Excluded	Not controlled	1.5T 3DT1 3 mm, FLAIR	Voxel-based cortical gray matter analyses	Visual rating of WMH, Number of lacunes	Adjusted	Yes	WMH: No Lacunes: NA Microbleeds: NA

See supplementary material for detailed criteria.

Relationship between subcortical MRI markers at baseline and atrophy rate.

Relationship between subcortical MRI marker progression and atrophy rate.

SD: standard deviation; AD: Alzheimer’s disease; NA: not available; FL: frontal lobe; PL: parietal lobe; TL: temporal lobe; OL: occipital lobe; MMSE: mini mental state examination; mult.comp.: multiple comparisons; WMH: white matter hyperintensities of presumed vascular origin; IQR: interquartile range; SSI: small subcortical infarct; MCI: mild cognitive impairment.

In five studies based on three different cohorts, patients were recruited if they had at least one cerebral infarct, most often attributable to SVD. Two reports were based on the same small cohort, in which patients were recruited after a first small subcortical ischemic stroke with separate reports of cross-sectional,²⁷ and longitudinal analyses.²⁸ Another cohort comprised 100 patients with any type of ischemic stroke and confluent WMH (60 from these 100 patients had a small subcortical ischemic stroke) and was the basis of two different reports.^24,25 In the fifth study, patients were included if they had a history of small subcortical ischemic stroke or a radiologic report of lacune.²⁶ The sample size of these cohorts ranged from 40 to 100 subjects, mean age was 72 years and mean MMSE score around 25.

Two larger studies (426 and 503 subjects) based on the RUN-DMC cohort, included patients if they had WMH or lacunes on MRI performed for any acute or subacute symptom attributable to SVD.^22,23 Those patients were clinically healthier and younger than those recruited in the five above-mentioned studies, with a mean age of 65 years and a mean MMSE score of 28. They also had a relatively low burden of MRI markers of SVD (mean volume of WMH = 12.96 ml; presence of lacunes in 21% of cases) compared with the three first cohorts (mean volume of WMH = 25.06 ml; presence of lacunes in 60 to 100% of patients).

Finally, three studies were based on the SMART-MR cohort that included independently living individuals with symptomatic atherosclerosis (coronary artery disease, stroke or TIA, peripheral arterial disease or abdominal aortic aneurysm).^19–21 Their sample size was large (565 to 840 subjects) and the mean age of patients was 58 years.

The presence of cortical infarcts was an exclusion criterion in 6 of 10 studies.^{19,21,23,26–28} The burden of WMH was evaluated quantitatively in eight studies^19–26 and using a semi-quantitative visual score in two.^27,28 In seven studies, lacunes were defined as lesions appearing hyperintense on T2 sequence and hypointense on T1 sequence and Flair sequences, and rated either by splitting the population according to the presence and absence of lacunes^19,21 or by assessing the number and/or volume of lacunes.^20,26–28 Lacunes were not assessed in one study²² or assessed globally with other infarct types in two others.^24,25 The number of microbleeds was only assessed in two studies based on the same cohort.^24,25 Attempt to rule out AD diagnosis was made in three studies.^24–26 In one, eight brain autopsies were performed, blind to clinical and MRI data and none of them met the NIA-Reagan Institute criteria supporting a high likelihood of AD.²⁶ In the two others based on the same cohort, 11C-PIB-PET imaging was performed in 7 out of the 100 recruited patients and showed no significant PIB retention.^24,25

The relationships between the total volume of cortical gray matter and WMH was tested in four cross-sectional studies based on three different cohorts.^19,24–26 All showed significant negative associations independent of age. The results obtained in one longitudinal study showed that both the baseline volume and the increase of periventricular WMH were significantly associated with a larger decrease of the total cortical gray matter volume after a mean follow-up of 3.9 years.²¹ A possible bias in some of these studies originating from the same SMART-MR cohort, including the longitudinal study,^19,21 was the use of 4-mm thick MR slices that may impact estimates for a highly folded structure such as the cerebral cortex.²⁹

The relationships between the total volume of cortical gray matter and lacunes were tested in five cross-sectional studies and in one longitudinal study. By contrast to the consistent associations observed between cortex alterations and the burden of WMH, no association was observed between the total volume of cortical gray matter and the presence,¹⁹ number,²⁰ or volume²⁶ of lacunes in cross-sectional studies. There was no association either between the volume of cortical gray matter and the total volume of infarcts, including lacunes, in 100 patients with confluent WMH included after any type of ischemic stroke.^24,25 By contrast, a significant association between a reduction of the total volume of cortical gray matter and the progression of lacunes, independent of age, gender, baseline brain volume, and follow-up time, was reported in a single study.²¹ Finally, no significant association was found between cortical gray matter volume and microbleeds in two studies investigating this aspect.^24,25

The relationships between cortical thickness and WMH were tested in two studies originating from the RUN-DMC cohort. The results were inconsistent showing a lower cortical thickness related to the volume of WMH in one study²³ and no significant association in another one, except in secondary subgroup analyses.²² A possible explanation might be that about 50 patients with cortical infarcts were not excluded from analyses in this latter study although such lesions are another prominent source of cortical atrophy. We found no study that evaluated the relationships between cortical thickness and lacunes or microbleeds.

Finally, VBM analyses were performed in two studies based on a single small cohort. Cross-sectional analyses showed a significant association between cortical gray matter estimates and the extent of WMH that did not persist in longitudinal analyses.^27,28 This discrepancy may be explained by the 25% of loss of follow-up in the longitudinal study and by the lack of adjustment for multiple comparisons in cross-sectional analyses. No association between cortical estimates obtained through VBM analyses and lacunes was reported.^27,28

In summary, despite the strong heterogeneity of the samples of patients with SVD or high risk profile, significant inverse relationships between most MRI estimates of cortex structure and the extent of WMH were consistently reported. By contrast, the relationships between cortex alterations and lacunes or microbleeds have been more rarely assessed and showed mostly non-significant associations. Some efforts were made to exclude concomitant AD in these studies, reinforcing the hypothesis of cortex alterations in the context of SVD patients, even in the absence of dementia.

Cortex alterations in patients with SVD by comparison with age- and sex-matched controls

Patients with SVD diagnosed using clinical criteria (see Table 3)

We found five studies including patients with cognitive impairment or dementia attributed to SVD (SVD_CI and SVD_D, respectively) that were compared with control subjects.^{10,30,32–34} Validated criteria were used to define cognitive impairment and dementia and to attribute them to SVD. Nonetheless, these criteria were variable among studies (see supplementary material). For instance, the presence of focal neurological signs was mandatory in three studies,^30,32,34 but not in the two others,^10,33 and the presence of cortical infarcts was an explicit exclusion criterion in three studies^10,30,34 but not in two.^32,33 In one study, post-mortem data were available in 4 of 18 cases and dementia was definitively attributed to SVD in all cases.¹⁰ By contrast, in the other studies, no attempt was made to exclude concomitant AD.^30,32–34 Inclusion criteria for controls were also highly variables. However, a normal cognitive status was requested in all studies and a normal MRI (definition of which varied among studies) in all but one study.³³

Table 3.

Case–control studies of patients with SVD according to clinical and neuroimaging data (ordered by sample sizes N).

Study	Patients description								Controls			MRI		Statistical analysis		Results
	Patients	Cognitive criteria	N	Age (mean ± SD)	Gender (female, %)	MMSE (mean ± SD)	Effect of associated cortical or territory infarcts	Effect of associated prodromal AD	N	Age (mean ± SD)	Gender (female, %)	Field strength Protocol	Cortex measure	Effect of age and gender	Adjusted for mult. comp. if needed	Significant global group differences	Significant regional group differences
	Patients	Cognitive criteria	N	Age (mean ± SD)	Gender (female, %)	MMSE (mean ± SD)	Effect of associated cortical or territory infarcts	Effect of associated prodromal AD	N	Age (mean ± SD)	Gender (female, %)	Field strength Protocol	Cortex measure	Effect of age and gender	Adjusted for mult. comp. if needed	Significant global group differences	FL	PL	TL	OL
Seo et al.³⁰	(1) SVD related cognitive impairment	Adapted from MCI criteria ³¹,Score < 16th percentile of norm (battery of test)	34	70.6 ± 6.4	15 (44.1%)	26.2 ± 2.2	Territory infarcts were an exclusion criterion	Not controlled	96	67.7 ± 6.3	54 (56.3%)	1.5T 3DT1	(1) Mean cortical thickness in native space (2) Surface-based regional analyses of cortical thickness	Adjusted	Yes	Mean cortical thickness	✓		✓	✓
	(2) SVD related dementia	DSM IV, Erkinjuntti	20	74.2 ± 6.1	11 (55.0%)	19.0 ± 5.1	Territory infarcts were an exclusion criterion	Not controlled	96	67.7 ± 6.3	54 (56.3%)	1.5T 3DT1		Adjusted	Yes	Mean cortical thickness	✓		✓	✓
Liu et al.³²	SVD related dementia	Modified NINDS AIREN,( + Hachinski ischemic scale ≥ 7 on MRI)	34	69.0 ± 7.9	13 (38%)	16 ± 5	Not detailed	Not controlled	35	68.0 ± 5.8	13 (37%)	3T 3DT1, FLAIR, T2	(1) Cortical gray matter volume (2) Mean cortical thickness in native space (3) Cortical gray matter density (4) Surface-based regional analyses of cortical thickness	Matched	Yes	Total cortical gray matter volume,Mean cortical thickness	Cortical gray matter density
																	✓	✓	✓
																	Cortical thickness
																	✓	✓	✓
Mueller et al.³³	(1) SVD related dementia	ADDTC	11	71.6 ± 10.3	2 (18%)	22.5 ± 4.2	Not detailed	Not controlled	42	75.9 ± 6.4	18 (43%)	1.5T 3DT1,T2	Cortical gray matter volume	Adjusted	No	NA	✓	✓	✓	✓
Mueller et al.³³	(2) SVD related cognitive impairment	Not detailed	21	74.0 ± 5.6	4 (19%)	27.6 ± 0.8	Not detailed	Not controlled	42	75.9 ± 6.4	18 (43%)	1.5T 3DT1,T2	Cortical gray matter volume	Adjusted	No	NA				✓
Du et al.¹⁰	SVD related dementia	ADDTC	18	74.2 ± 8.3	6 (33%)	19.9 ± 4.7	Territory infarcts were an exclusion criterion	Autopsy confirmed diagnosis in 4/18 patients	38	75.0 ± 4.3	18 (47%)	1.5T 3DT1	Cortical gray matter volume	Matched	NA	Total cortical gray matter volume	No regional analysis
Li et al.³⁴	SVD related dementia	DSM IV, Erkinjuntti	9	62.3 ± 6.1	4 (44%)	15.4 ± 3.6	Territory infarcts were an exclusion criterion	Not controlled	9	60.4 ± 4.2	4 (44%)	1.5T 3DT1, T2	Cortical gray matter density	Matched	Yes	NA	✓	✓	✓

SD: standard deviation; AD: Alzheimer’s disease; FL: frontal lobe; PL: parietal lobe; TL: temporal lobe; OL: occipital lobe; MMSE: mini mental state examination; mult.comp.: multiple comparisons; WMH: white matter hyperintensities of presumed vascular origin; NA: not applicable; MCI: mild cognitive impairment.

Sample sizes in these five studies were small (mean number of patients: 29, mean number of controls: 52). Mean age was about 70 years in four studies^10,30,32,33 and close to 60 years in one.³⁴ Controls were age- and sex-matched in three studies^10,32,34 and analyses were adjusted for age and gender in the two others.^30,33 Patients and controls were in all cases evaluated using the same scanner, MRI sequence and post-processing method based on 3DT1 millimetric isotropic resolution.

The mean cortical thickness measured over the whole cortex was compared between SVD_D patients and controls in two studies and between SVD_CI patients and controls in one, with significantly lower measures observed in patients in all cases.^30,32 In additional regional analyses, group differences were consistently observed in frontal and temporal areas.^30,32

The total cortical gray matter volume was compared between SVD_D patients and controls in two studies, showing significant lower volumes in patients in both of them.^10,32 The volume of cortical gray matter measured in each cerebral lobe (without reporting global measures) was compared between SVD_D or SVD_CI patients and controls in a single study.³³ Significantly lower cortical gray matter volume were observed in SVD_D patients in all cerebral lobes and in SVD_CI patients only in the occipital lobe.³³

Finally, local cortical gray matter estimates determined using VBM were compared between SVD_D patients and controls in two studies, with significantly lower estimates observed in patients in both studies, in all cortical regions but the occipital lobe.^32,34

Additionally, the relationships between the burden of subcortical MRI markers of SVD and cortex alterations in SVD_D patients were tested in only one study, showing, using uncorrected additional analyses, that a higher number of lacunes was associated with a lower cortical thickness in frontal, parietal, temporal, and insular lobes.³² By contrast, the volume of WMH was not related to cortical thickness.³² These relationships were also tested in one study after pooling SVD_D and SVD_CI patients, showing a significant linear negative association between the volume of WMH and that of cortical gray matter in all lobes, but this was also observed in controls.³³ The association between lacunes and cortex alterations was not evaluated in this study.³³

In summary, significant and consistent alterations of cortex estimates were reported using various methodologies in patients with cognitive impairment or dementia attributed to SVD by comparison with age- and sex-matched controls. However, with the exception of post-mortem data confirming the diagnosis of SVD in four cases in one study, few data are available to ensure that cortex alterations are actually related to SVD and are not secondary to a degenerative disorder.

Patients with SVD diagnosed according to radiological criteria (see Table 4)

In four studies, patients diagnosed with SVD according to neuroimaging criteria were compared with age- and sex-matched control subjects.^35–38 Inclusion criteria were variable. In one study, 56 patients with type 2 diabetes and various load of WMH in presence or absence of lacunes were included,³⁵ but only 46% of them were considered to have SVD (defined as a volume of WMH superior to the highest tertile and/or presence of lacune(s)). Furthermore, no significant difference was detected globally between patients and controls regarding their volume of WMH. In another study, 16 stroke- and dementia-free subjects aged 50 to 75, regularly followed in a health examination center were included if they only had periventricular WMH.³⁷ In a third study, 11 stroke- and dementia-free patients who presented initially with cognitive complaints were considered to have SVD if they had confluent WMH.³⁸ In the last study, the SVD group comprised 21 patients with a history of lacunar stroke and 6 with incidental WMH and/or lacunes on brain MRI obtained after any neurological symptom.³⁶ A PET-FDG was performed to exclude patients with suspected associated prodromal AD in this last study, but no detail was provided according to their interpretation. No attempt to exclude concomitant AD was made in the three other studies. Results of MRI were used to exclude controls with confluent³⁸ or “any” WMH^36,37 in three studies, but the presence of cortical infarcts was not an explicit exclusion criterion in any study and was even tolerated in one.³⁵ A history of any neurological or psychiatric disorder including stroke and cognitive impairment was an exclusion criterion for controls in all studies.

Table 4.

Case–control studies of patients with SVD according to neuroimaging data (ordered by sample sizes N).

Study	Patients description								Controls			MRI		Statistical analysis		Results
	Patients	Cognitive criteria	N	Age (mean ± SD)	Gender (female, %)	MMSE (mean ± SD)	Effect of associated cortical or territory infarcts	Effect of associated prodromal AD	N	Age (mean ± SD)	Gender (female, %)	Field strength Protocol	Cortex measure	Effect of age and gender	Adjusted for mult.comp. if needed	Significant global group differences	Significant regional group differences
	Patients	Cognitive criteria	N	Age (mean ± SD)	Gender (female, %)	MMSE (mean ± SD)	Effect of associated cortical or territory infarcts	Effect of associated prodromal AD	N	Age (mean ± SD)	Gender (female, %)	Field strength Protocol	Cortex measure	Effect of age and gender	Adjusted for mult.comp. if needed	Significant global group differences	FL	PL	TL	OL
Brundel et al.³⁵	Patients with type 2 diabetes	Independently living individuals, absence of psychiatric or neurological disorder, but history of non-invalidating stroke was tolerated	56	70.0 ± 5.2	29 (52%)	–	Territory infarcts tolerated	Not controlled	30	68.1 ± 4.3	17 (57%)	1.5T 3DT1, FLAIR	(1) Total cortical gray matter volume (2) Mean cortical thickness (3) Surface-based regional analyses of cortical thickness	Adjusted	Yes	Total cortical gray matter volume	Cortical gray matter volume
Brundel et al.³⁵	Patients with type 2 diabetes		56	70.0 ± 5.2	29 (52%)	–	Territory infarcts tolerated	Not controlled	30	68.1 ± 4.3	17 (57%)	1.5T 3DT1, FLAIR		Adjusted	Yes	Total cortical gray matter volume			✓
Preul et al.³⁶	Lacunar stroke or incidental WMH	Independently living patients	27^a	58.3 ± 7.8	7 (26%)	–	Territory infarcts were an exclusion criterion	PET (FDG?) in cases suspicious of AD	27	58.3 ± 7.1	7 (26%)	3T 3DT1, T2	(1) Mean cortical thickness (2) Total cortical gray matter volume	Matched	No	Mean cortical thickness	✓	✓	✓
Zi et al.³⁷	Periventricular WMH (no detailed threshold)	No history of stroke, no dementia	16	62.0 ± 4.9	9 (56%)	–	Not detailed	Not controlled	16	61.5 ± 3.7	9 (56%)	3T T1, T2, FLAIR, DTI	Surface-based regional analyses of cortical thickness (no 3DT1)	Matched	Yes	NA	✓		✓	✓
Quinque et al.³⁸	WMH and cognitive complaints	No history of psychiatric or neurologic disorder including stroke	11	61.4 ± 6.3	4 (36%)	27.6 ± 1.5	Not detailed	Not controlled	21	66.0 ± 6.7	8 (38%)	3T 3DT1, T2, FLAIR, DTI	Voxel-based gray matter analyses	Matched	Yes	Yes	✓

21 with lacunar stroke, 6 with incidental WMH.

Sample sizes were small (mean number of patients: 16, mean number of controls: 24). Mean age was about 60 years in three studies^36–38 and 70 years in the other.³⁵ Controls were age- and sex-matched in three studies^36–38 and analyses were adjusted for age and gender in the other.³⁵ Patients and controls were assessed on the same scanner using the same sequence in all cases and post-processing relied on MRI with 3DT1 1 mm isotropic resolution in all studies but one that used T1 weighted sequence with a slice thickness of 5 mm.³⁷

The mean cortical thickness measured over the whole cortex was compared between SVD patients and controls in two studies,^35,36 reporting in both cases lower cortical thickness in patients, that reached significance in only one study.³⁶ This was possibly related to the different profiles of patients selected in these studies: the difference was significant in the sample including cases with severe SVD (21 out of 26 patients had lacunar stroke)³⁶ but not in diabetic patients among which some had a diagnosis of SVD.³⁵

The total cortical gray matter volume was compared between SVD patients and controls in the same two studies. The total volume of cortical gray matter was lower in patients in these two studies,^35,36 but statistical significance was reached only in the diabetic sample.³⁵ Whether this difference is directly related to the role of diabetes³⁵ or due to technical issues³⁶ remains undetermined.

Surface-based regional analyses of cortical thickness was performed in two studies^35,37 and showed a significantly lower cortical thickness in patients compared with controls in both. This reduction was detected diffusely except in the parietal lobe in one study³⁷ but analyses were not corrected for multiple comparisons. It was found only in the temporal lobe in the other study.³⁵ Whether technical issues explain these differences could not be totally excluded.

Finally, VBM estimates of cortical gray matter were compared in only one study between a very small sample of 11 patients having WMH on MRI performed for cognitive complaints and controls. In this study, significantly lower estimates of cortical gray matter in two clusters located in the frontal lobes bilaterally were detected in SVD patients compared with controls.³⁸ These results remain difficult to interpret given the small number of patients and the very focal asymmetric clusters detected.

In summary, in patients with SVD defined according to a wide range of MRI criteria, significantly lower cortex estimates were consistently observed by comparison with age- and sex-matched controls. However, these studies included only small sample sizes, were in all cases cross-sectional, and efforts were rarely made to exclude concomitant prodromal AD as a source of the observed cortical alterations.

Discussion

The results of the present systematic review support the hypothesis that the cerebral cortex is diffusely altered in SVD, from the earliest stages in subjects sampled from the general population and having only WMH on their MRI to latest stages, in patients with stroke and/or dementia attributed to SVD, thus challenging the concept of “subcortical ischemic vascular dementia.” Although the results of most articles included in the present review are consistent, important questions remain unanswered.

First, given the fact that AD can promote cortex atrophy before the appearance of clinical symptoms,^39,40 it is important to keep in mind that the reported cortex alterations are not necessarily the results of SVD. Only in 5 among the 25 studies identified in the present review, the authors attempted to exclude a possible role of concomitant prodromal AD. 11C-PIB-PET imaging was performed in 72/72 subjects in one study,¹⁶ in 7/100 individuals in the other²⁴ and ruled out PIB binding consistent with AD diagnosis. FDG-PET imaging was obtained in an additional study for patients with suspected AD who were excluded when cortical hypoperfusion consistent with AD was detected.³⁶ Finally, AD was excluded according to autopsy data in only two studies (in 4/18 subjects in one study,¹⁰ in 8/87 subjects in the other²⁶). Nonetheless, it is important to note that, given the frequency of mixed cases, even those studies could not formally exclude an effect of associated neurodegenerative illnesses.^40–42 In this respect, it is important to emphasize that cortical alterations were also demonstrated in hereditary SVD affecting younger patients.^43–46 Anyway, further studies including a larger number of patients with quantification of both SVD and neurodegenerative processes are needed to determine the exact extent of cortex alterations in SVD without associated AD, and the relative impact of AD and of SVD on cortex metrics in patients having both SVD and AD.

Second, the nature of the observed cortex alterations also remains incompletely understood. No study trying to disentangle the effects on cortical gray matter volume or cortical thickness from that on cortical surface area or cortex morphology was done yet. Additionally, as effect sizes were rarely provided in the different studies, it was not possible to compare the results of the different studies by other means than by their significance. Also, technical issues or global brain alterations could lead to the detection of group differences, particularly with the use of VBM estimates, that may not be related to cortex alterations.⁴⁷ Additionally, cortical thickness, but not cortex morphology, is strongly dependent on gray to white matter contrast, which is clearly altered in SVD,⁴⁸ while cortex morphology is strongly dependent on the underlying white matter structure.⁴⁹

The mechanisms linking subcortical markers of SVD with cortical alterations remain to be determined. The inverse links repeatedly observed between the extent of WMH and cortex alterations, independent of age, might be secondary to remote degeneration.⁴⁵ Alternatively, WMH in subcortical areas might also be in part secondary to cortical atrophy through retrograde degeneration.^11,33,50 Another explanation could be that WMH and cortical alterations share common underlying mechanisms such as chronic hypoperfusion.^11,24 By contrast to that of WMH, the association between lacunes or microbleeds with cortex alterations were only rarely and inconsistently observed precluding any conclusion regarding their own impact. The observed impact of lacunes on remotely connected cortex in a genetic model of SVD suggests that lacunes can actually lead to cortex alterations through secondary degeneration,⁴⁵ which may even promote amyloid deposition.⁴²

The relationships between cortical alterations and cognitive status were analyzed in 10 out of the 25 selected studies. While memory and global cognition were associated with alterations of cortex measures in few studies,^26,32 the most consistent finding is that executive functions and processing speed^12,16,23,37 are significantly related to the cortical alterations detected in SVD. This association was found independent of age, gender, and of level of education. Some results even suggest that cortical alterations in SVD may impact cognition independently of WMH volume and sometimes, even with a stronger effect than that found only with WMH.^23,24 Further studies are needed to determine the impact of cortex alterations independently of other MRI markers of SVD, particularly global brain atrophy.

Altogether, despite an important heterogeneity in SVD definition, methodologies, post-processing methods and strategies in data analysis, the results of these studies provide strong evidence that diffuse alterations of the cerebral cortex may occur early during the course of SVD. Available data also support that SVD should no longer be considered as a subcortical disorder. Further studies are needed to better depict the mechanisms and consequences of cortical alterations in SVD.

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by grants from the Network of European Funding for Neuroscience Research grant (01EW1207), from Fondation Leducq (Transatlantic Network of Excellence on the Pathogenesis of Small Vessel Disease of the Brain, ), and from the Ile-de-France regional health agency.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Authors’ contributions

EJ and HC designed the review. RP and FDG collected and analyzed the literature. RP, FDG, and EJ wrote the initial draft of the review. All authors reviewed the manuscript.

Supplementary material

Supplementary material for this paper can be found at

References

Pantoni

. Cerebral small vessel disease: from pathogenesis and clinical characteristics to therapeutic challenges. Lancet Neurol 2010; 9: 689–701.

Wardlaw

Smith

Dichgans

. Mechanisms of sporadic cerebral small vessel disease: insights from neuroimaging. Lancet Neurol 2013; 12: 483–497.

Wardlaw

Smith

Biessels

. Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. Lancet Neurol 2013; 12: 822–838.

Roman

Erkinjuntti

Wallin

. Subcortical ischaemic vascular dementia. Lancet Neurol 2002; 1: 426–436.

Albert

Feldman

Willis

. The 'subcortical dementia' of progressive supranuclear palsy. J Neurol Neurosurg Psychiatry 1974; 37: 121–130.

Kovari

Gold

Herrmann

. Cortical microinfarcts and demyelination significantly affect cognition in brain aging. Stroke 2004; 35: 410–414.

Sonnen

Larson

Crane

. Pathological correlates of dementia in a longitudinal, population-based sample of aging. Ann Neurol 2007; 62: 406–413.

Smith

Schneider

Wardlaw

. Cerebral microinfarcts: the invisible lesions. Lancet Neurol 2012; 11: 272–282.

van Veluw

Hilal

Kuijf

. Cortical microinfarcts on 3T MRI: Clinical correlates in memory-clinic patients. Alzheimers Dement 2015; 11: 1500–1509.

10.

Schuff

Laakso

. Effects of subcortical ischemic vascular dementia and AD on entorhinal cortex and hippocampus. Neurology 2002; 58: 1635–1641.

11.

Wen

Sachdev

Chen

. Gray matter reduction is correlated with white matter hyperintensity volume: a voxel-based morphometric study in a large epidemiological sample. NeuroImage 2006; 29: 1031–1039.

12.

Raji

Lopez

Kuller

. White matter lesions and brain gray matter volume in cognitively normal elders. Neurobiol Aging 2012; 33: 834.e7–834.e16.

13.

Liu

Sachdev

Lipnicki

. Limited relationships between two-year changes in sulcal morphology and other common neuroimaging indices in the elderly. NeuroImage 2013; 83: 12–17.

14.

Rossi

Boccardi

Sabattoli

. Topographic correspondence between white matter hyperintensities and brain atrophy. J Neurol 2006; 253: 919–927.

15.

Ota

Nemoto

Sato

. Relationship between white matter T2 hyperintensity and cortical volume changes on magnetic resonance imaging in healthy elders. Int J Geriatr Psychiatry 2011; 26: 886–892.

16.

Wirth

Villeneuve

Haase

. Associations between Alzheimer disease biomarkers, neurodegeneration, and cognition in cognitively normal older people. JAMA Neurol 2013; 70: 1512–1519.

17.

Eckert

Keren

Roberts

. Age-related changes in processing speed: unique contributions of cerebellar and prefrontal cortex. Front Hum Neurosci 2010; 4: 10.

18.

Kochunov

Thompson

Coyle

. Relationship among neuroimaging indices of cerebral health during normal aging. Hum Brain Mapp 2008; 29: 36–45.

19.

Appelman

Vincken

van der Graaf

. White matter lesions and lacunar infarcts are independently and differently associated with brain atrophy: the SMART-MR study. Cerebrovasc Dis 2010; 29: 28–35.

20.

Geerlings

Appelman

Vincken

. Brain volumes and cerebrovascular lesions on MRI in patients with atherosclerotic disease. The SMART-MR study. Atherosclerosis 2010; 210: 130–136.

21.

Kloppenborg

Nederkoorn

Grool

. Cerebral small-vessel disease and progression of brain atrophy: the SMART-MR study. Neurology 2012; 79: 2029–2036.

22.

Reid

van Norden

de Laat

. Patterns of cortical degeneration in an elderly cohort with cerebral small vessel disease. Hum Brain Mapp 2010; 31: 1983–1992.

23.

Tuladhar

Reid

Shumskaya

. Relationship between white matter hyperintensities, cortical thickness, and cognition. Stroke 2015; 46: 425–432.

24.

Mok

Wong

Xiong

. Cortical and frontal atrophy are associated with cognitive impairment in age-related confluent white-matter lesion. J Neurol Neurosurg Psychiatry 2011; 82: 52–57.

25.

Xiong

Wong

. Predictors for cortical gray matter volume in stroke patients with confluent white matter changes. J Neurol Sci 2014; 338: 169–173.

26.

Fein

Di Sclafani

Tanabe

. Hippocampal and cortical atrophy predict dementia in subcortical ischemic vascular disease. Neurology 2000; 55: 1626–1635.

27.

Grau-Olivares

Bartres-Faz

Arboix

. Mild cognitive impairment after lacunar infarction: voxel-based morphometry and neuropsychological assessment. Cerebrovasc Dis 2007; 23: 353–361.

28.

Grau-Olivares

Arboix

Junque

. Progressive gray matter atrophy in lacunar patients with vascular mild cognitive impairment. Cerebrovasc Dis 2010; 30: 157–166.

29.

Wonderlick

Ziegler

Hosseini-Varnamkhasti

. Reliability of MRI-derived cortical and subcortical morphometric measures: effects of pulse sequence, voxel geometry, and parallel imaging. NeuroImage 2009; 44: 1324–1333.

30.

Seo

Ahn

Yoon

. Cortical thinning in vascular mild cognitive impairment and vascular dementia of subcortical type. J Neuroimaging 2010; 20: 37–45.

31.

Petersen

Smith

Waring

. Mild cognitive impairment: clinical characterization and outcome. Archiv Neurol 1999; 56: 303–308.

32.

Liu

Gui

. The pattern of brain gray matter impairments in patients with subcortical vascular dementia. J Neurol Sci 2014; 341: 110–118.

33.

Mueller

Mack

Mungas

. Influences of lobar gray matter and white matter lesion load on cognition and mood. Psychiatry Res 2010; 181: 90–96.

34.

Zheng

. A voxel-based morphometric analysis of cerebral gray matter in subcortical ischemic vascular dementia patients and normal aged controls. Int J Med Sci 2011; 8: 482–486.

35.

Brundel

van den Heuvel

de Bresser

. Cerebral cortical thickness in patients with type 2 diabetes. J Neurol Sci 2010; 299: 126–130.

36.

Preul

Lohmann

Hund-Georgiadis

. Morphometry demonstrates loss of cortical thickness in cerebral microangiopathy. J Neurol 2005; 252: 441–447.

37.

Duan

Zheng

. Cognitive impairments associated with periventricular white matter hyperintensities are mediated by cortical atrophy. Acta Neurol Scand 2014; 130: 178–187.

38.

Quinque

Arelin

Dukart

. Identifying the neural correlates of executive functions in early cerebral microangiopathy: a combined VBM and DTI study. J Cereb Blood Flow Metab 2012; 32: 1869–1878.

39.

Jagust

Zheng

Harvey

. Neuropathological basis of magnetic resonance images in aging and dementia. Ann Neurol 2008; 63: 72–80.

40.

Viswanathan

Rocca

Tzourio

. Vascular risk factors and dementia: how to move forward? Neurology 2009; 72: 368–374.

41.

Jellinger

. Alzheimer disease and cerebrovascular pathology: an update. J Neural Transm (Vienna) 2002; 109: 813–836.

42.

Garcia-Alloza

Gregory

Kuchibhotla

. Cerebrovascular lesions induce transient beta-amyloid deposition. Brain 2011; 134(Pt 12): 3697–3707.

43.

Jouvent

Poupon

Gray

. Intracortical infarcts in small vessel disease: a combined 7-T postmortem MRI and neuropathological case study in cerebral autosomal-dominant arteriopathy with subcortical infarcts and leukoencephalopathy. Stroke 2011; 42: e27–e30.

44.

Viswanathan

Gray

Bousser

. Cortical neuronal apoptosis in CADASIL. Stroke 2006; 37: 2690–2695.

45.

Duering

Righart

Csanadi

. Incident subcortical infarcts induce focal thinning in connected cortical regions. Neurology 2012; 79: 2025–2028.

46.

De Guio

Reyes

Vignaud

. In vivo high-resolution 7 Tesla MRI shows early and diffuse cortical alterations in CADASIL. PloS One 2014; 9: e106311.

47.

Ashburner

. Computational anatomy with the SPM software. Magn Reson Imag 2009; 27: 1163–1174.

48.

De Guio

Reyes

Duering

. Decreased T1 contrast between gray matter and normal-appearing white matter in CADASIL. AJNR Am J Neuroradiol 2014; 35: 72–76.

49.

Mangin

Jouvent

Cachia

. In-vivo measurement of cortical morphology: means and meanings. Curr Opin Neurol 2010; 23: 359–367.

50.

Capizzano

Acion

Bekinschtein

. White matter hyperintensities are significantly associated with cortical atrophy in Alzheimer's disease. J Neurol Neurosurg Psychiatry 2004; 75: 822–827.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.09 MB