Background and Purpose
Studies on the impact of small vessel disease (SVD) on cognition generally focus on white matter hyperintensity (WMH) volume. The extent to which WMH location relates to cognitive performance has received less attention, but is likely to be functionally important. We examined the relation between WMH location and cognition in a memory clinic cohort of patients with sporadic SVD.
A total of 167 patients with SVD were recruited from memory clinics. Assumption-free region of interest-based analyses based on major white matter tracts and voxel-wise analyses were used to determine the association between WMH location and executive functioning, visuomotor speed and memory.
Region of interest-based analyses showed that WMHs located particularly within the anterior thalamic radiation and forceps minor were inversely associated with both executive functioning and visuomotor speed, independent of total WMH volume. Memory was significantly associated with WMH volume in the forceps minor, independent of total WMH volume. An independent assumption-free voxel-wise analysis identified strategic voxels in these same tracts. Region of interest-based analyses showed that WMH volume within the anterior thalamic radiation explained 6.8% of variance in executive functioning, compared to 3.9% for total WMH volume; WMH volume within the forceps minor explained 4.6% of variance in visuomotor speed and 4.2% of variance in memory, compared to 1.8% and 1.3% respectively for total WMH volume.
Our findings identify the anterior thalamic radiation and forceps minor as strategic white matter tracts in which WMHs are most strongly associated with cognitive impairment in memory clinic patients with SVD. WMH volumes in individual tracts explained more variance in cognition than total WMH burden, emphasizing the importance of lesion location when addressing the functional consequences of WMHs.
Citation: Biesbroek JM, Weaver NA, Hilal S, Kuijf HJ, Ikram MK, Xu X, et al. (2016) Impact of Strategically Located White Matter Hyperintensities on Cognition in Memory Clinic Patients with Small Vessel Disease. PLoS ONE 11(11): e0166261. https://doi.org/10.1371/journal.pone.0166261
Editor: Stephen D. Ginsberg, Nathan S Kline Institute, UNITED STATES
Received: July 20, 2016; Accepted: October 25, 2016; Published: November 8, 2016
Copyright: © 2016 Biesbroek et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: The Memory Aging and Cognition Centre is funded by a National Medical Council of Singapore Centre Grant (NMRC/CG/013/2013). Hugo J. Kuijf was financially supported by the project Brainbox (Quantitative analysis of MR brain images for cerebrovascular disease management), funded by the Netherlands Organisation for Health Research and Development (ZonMw) in the framework of the research programme IMDI (Innovative Medical Devices Initiative); project 104002002. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Cerebral small vessel disease (SVD) is a major cause of cognitive impairment, either by itself or in combination with other pathologies in particular Alzheimer’s disease (AD) . White matter hyperintensities (WMHs) are a major manifestation of SVD. Associations between WMH volume and cognitive functioning have been clearly established, but do show marked inter-individual variation [1,2]. Recent studies have highlighted the importance of WMH location in relation to cognitive functioning [3–8], which is consistent with the concept that disruption of strategic white matter tracts is a key mechanism through which subcortical vascular damage affects cognition . Identifying strategic white matter tracts in which WMHs have most impact on cognition would improve our understanding of the functional impact of SVD.
Current image analysis techniques allow integration of three-dimensional information of lesion location in individual patients and anatomical knowledge of major white matter tracts in relation to cognitive performance. Recent studies have applied these techniques to identify strategic WMH locations in patients with CADASIL (cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy), i.e. a rare form of hereditary SVD) [4,5], patients with manifest arterial disease , elderly stroke survivors (≥3 months post-stroke) , patients with AD  and community-based cohorts (i.e. patients without cognitive complaints) [6,7,11]. These studies suggest that WMHs located in frontal cortical-subcortical and frontal interhemispheric projections are associated with poor executive functioning and processing speed, while WMHs located in temporal and parieto-occipital white matter regions and the forceps major are associated with memory impairment. Furthermore, higher WMH volume within cholinergic projections has been associated with poor memory in AD patients . However, tract-specific techniques have not yet been applied to memory clinic patient with sporadic SVD, a population in which an evaluation of functional impact of SVD is clearly clinically relevant, also in the context of other co-occurring etiologies, in particular AD. In the current study in memory clinic patients with sporadic SVD, we therefore aimed to (1) determine whether the impact of WMHs on cognition depends on location, and to (2) identify strategic white matter tracts in which WMHs have high impact on cognitive functioning.
Materials and Methods
Patients with SVD on MRI were selected from the National University Health System Memory Ageing and Cognition Centre cohort. Patients in this cohort were recruited from two memory clinics in Singapore (National University Hospital and St. Luke’s Hospital) between December 2010 and September 2013. Patients with the following five referral diagnoses were eligible for inclusion in the National University Health System Memory Ageing and Cognition Centre cohort: 1) Cognitive impairment not dementia: impairment in at least one cognitive domain, but without meeting the DSM-IV (Diagnostic and Statistical Manual of Mental Disorders–Fourth Edition) criteria for dementia. A domain was considered to be impaired if subjects failed at least half of the tests (i.e. a score more than 1.5 SDs below established normal means on individual tests, after adjustment for age and education) in that domain. 2) Alzheimer’s disease (AD): diagnosed using the NINCDS-ADRDA criteria for probable or possible AD . 3) Vascular dementia (VaD): diagnosed using the NINDS-AIREN criteria . 4) Mixed dementia: defined as AD with significant cerebrovascular disease using the McKhann criteria . 5) No cognitive impairment: no objective cognitive impairment on formal neuropsychological tests, or functional loss. Patients with other diagnoses, significant neurological co-morbidities (e.g. Parkinson’s disease) or loss of functional independence (modified Rankin Scale >4) were not included in the cohort. All participants in the National University Health System Memory Ageing and Cognition Centre cohort underwent standardized clinical examination, neuropsychological testing (separate and independent from the primary testing at the memory clinics prior to study enrollment) and 3T MRI of the brain at the National University Hospital of Singapore.
A flow chart of patient enrollment for the present study is provided in Fig 1. Of all 250 consecutive patients with the abovementioned referral diagnoses, only patients with complete baseline MRI sequences (T1, T2 and T2 FLAIR) were eligible for the present study (n = 233). Next, we selected all patients who met the following operational definition of presence of cerebral SVD: 1) MRI showing either moderate to severe WMHs (Fazekas ≥2), lacunes, infarcts, or intracerebral microbleeds (n = 205) ; or 2) MRI showing mild white matter lesions (Fazekas = 1) and at least two of the following vascular risk factors: hypertension, hypercholesterolemia, diabetes mellitus, obesity (BMI >30), past or current smoking and clinically manifest arterial disease (n = 16). Of these 221 patients, we excluded 43 patients with a cortical infarct or a large subcortical infarct (>15 mm) or hemorrhage (>15 mm), because such lesions can result in the complete obliteration of white matter tracts, thereby interfering with our analyses in which WMH volume within specific tracts is related to cognition at a group level (see analysis section). Five additional patients were excluded due to missing neuropsychological data for one or more of the tests for executive functioning, visuomotor speed or memory. Finally, 6 patients were excluded due to failed lesion registration. This resulted in a final study sample of 167 patients.
Standard protocol approvals and patient consents
Ethical approval for this study was obtained from the National Healthcare Group Domain-Specific Review Board (DSRB). The study was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained by bilingual study coordinators in the preferred language of the patients. Informed consent was obtained face to face with the participant and an authorized family member present. No separate consent forms were used for cognitively impaired person; in such cases, demented participants indicated their willingness to join verbally. Legally acceptable representatives were approached to give consent on behalf of the individuals judged incapable of providing consent (e.g. severely demented patients). All queries were answered and clarified prior to the signing of the document to ensure that participants and family members had their concerns satisfactorily addressed.
MR investigations were performed on a 3 tesla Siemens Magnetom Trio Tim system, with a 32-channel head-coil, at the Clinical Imaging Research Centre of the National University of Singapore. The standardized protocol consisted of a 3D T1-weighted (1.0x1.0x1.0 mm3 voxels; TR 2300 ms; TE 1.9 ms; TI 900 ms; flip angle 9°; matrix 256x256), a 2D multislice T2-weighted (1.0x1.0x3.0 mm3 voxels; TR 3000 ms; TE 10.1 ms; matrix 247x256), a 2D multislice fluid-attenuated inversion recovery (FLAIR) (1.0x1.0x3.0 mm3; TR 9000 ms; TE 82 ms; TI 2500 ms; matrix 232x256), and a 2D multislice T2*-weighted image (1.0x1.0x1.5 mm3 voxels; TR 27 ms; TE 20 ms; flip angle 15°; matrix 192x256).
WMHs were manually segmented on T2 FLAIR sequences by a trained researcher (NAW) and reviewed by an experienced rater (JMB), who were blinded to the clinical data. The presence of infarcts and lacunes was initially scored independently by two raters (NAW and SH) and subsequently reviewed by another rater (JMB), followed by a consensus rating with another experienced rater (GJB) if there was uncertainty regarding lesion classification (which was the case in 23 patients (14%)). T1 and T2-weighted image sequences were used as reference to support proper lesion classification. Microbleeds were assessed by one rater (SH) on T2*-weighted images using the Brain Observer Micro-Bleeds Scale criteria . The interrater and intrarater reliability as expressed by k-statistic ranged from 0.59 to 0.80 for lacunes, and 0.75 to 0.89 for microbleeds. All raters were blinded to the clinical data during the segmentation and rating process. All lesions were defined in accordance with the internationally established STRIVE criteria, which include guidelines on how to differentiate lacunes from perivascular spaces . WMH segmentation was performed with in-house developed software based on MeVisLab (MeVis Medical Solutions AG, Bremen, Germany) [18,19]. Total brain volume and intracranial volume were quantified using the Statistical Parametric Mapping 12b unified segmentation approach  and used to compute the brain parenchymal fraction (BPF; (brain volume/intracranial volume)) as a measure of brain atrophy.
Our study focused on executive functioning and visuomotor speed, because these cognitive domains are commonly affected in patients with SVD , and on memory. The neuropsychological examination protocol and formation of compound scores for cognitive domains were hypothesis-driven and has been previously described . Neuropsychological tests were administered in the patient’s habitual language to avoid variance due to insufficient language skills. Executive functioning was assessed with the Colour Trails Test 1&2 and verbal category fluency tests for animals and food [23,24]. Visuomotor speed was assessed with the Digit Cancellation Task, Symbol Digit Modalities Task and Maze Test [25–27]. Memory was assessed with verbal word list  and visual picture  recall and recognition tests. Z-scores were calculated for each subtest based on the mean and SD values of the included patients; results for the Maze Test and Colour Trails Test 1&2 were multiplied by -1, so that lower scores indicated poorer performance. Compound z-scores for executive functioning, visuomotor speed and memory were calculated by taking the average of the z-scores of the subtests for each cognitive domain.
Generation of lesion maps
WMH segmentations were transformed to the T1 1-mm MNI-152 template using the following procedure . All registrations were performed with the elastix toolbox for registration . The results of the brain segmentation were used to create a brain tissue mask . The T2 FLAIR scans were transformed to the corresponding T1 scans using a linear registration. The T1 scans were transformed to the Alzheimer’s disease (AD) T1 brain template , with a linear registration followed by a non-linear registration. For the linear registration, the brain tissue masks of both the patients’ T1 scan and the AD template were aligned to create a rough initial estimate and the original scans were used in the subsequent non-linear registration. The AD template was transformed to the MNI-152 template with a linear and a non-linear registration. The intermediate registration step using the AD template proved more successful in transforming scans of patients with severe atrophy, because of a better match between patient and template. Finally, all computed transformations were composed into a single transformation step that was used to co-register the WMH maps directly to the MNI-152 template. Composing transformations avoids any intermediate interpolations and results in a more accurate registration. Visual checks of the results of this procedure were performed for all patients. Six patients (4%) were excluded because the registration was unsuccessful.
We performed our analyses in two subsequent steps. In the first step, we aimed to determine whether the impact of WMHs on cognitive functioning depends on location. In order to identify strategic white matter tracts in which WMHs are associated with cognitive functioning, independent of global WMH volume, we performed the following assumption-free (as opposed to hypothesis-driven) analyses: A) Region of interest (ROI)-based analyses based on the association between WMH volumes within specific white matter tracts and cognitive functioning, and B) voxel-based lesion-symptom mapping (VLSM) based on the association between the presence of WMHs and cognitive functioning for each affected voxel across the brain [33,34]. ROI-based analyses have the advantage of taking into account cumulative lesion burden within a white matter tract, whereas VLSM offers much higher spatial resolution (at the cost of lower statistical power due to more rigorous correction for multiple testing) without the need for predefining regions of interest. These two methods therefore have complementary strengths. For the ROI-based analyses, 11 regions of interest were created using a probabilistic white matter tract atlas with a probability threshold of 10% . Bilateral tracts were merged into a single region of interest (for example, the anterior thalamic radiation was formed by merging the left and right anterior thalamic radiation into a single region of interest). The regional WMH volumes within these 11 white matter tracts were entered as independent variables in a linear regression model, with the z-scores of cognitive performance as outcome, and age, sex, level of education, total normalized WMH volume (i.e. WMH volume after registration to the MNI-152 template) and the presence of lacunes as covariates. These correlations were adjusted for total WMH volume, because strategic white matter tracts were defined as tracts in which WMH volume is correlated with cognitive functioning independent of global lesion volume. Next, VLSM analyses were performed using the compound z-scores for executive functioning and visuomotor speed, after individualized correction for age, sex and level of education using linear regression. To determine if the observed voxel-wise associations were independent of total lesion burden, the analyses were repeated after additional correction for total normalized WMH volume (i.e. WMH volume after normalization to the MNI-152 template) and the presence of lacunes. VLSM was performed using Non-Parametric Mapping (settings: univariate analysis, Brunner-Munzel test) . Voxels that were damaged in less than 10 individuals were excluded from the analysis. Correction for multiple testing was done by applying a false discovery rate (FDR) with q<0.05. VLSM could not be performed for lacunes or microbleeds due to relatively low prevalence and small lesion size.
In the second step, we aimed to quantify and compare the impact of WMH lesion load in the strategic tracts as identified in the first step and other markers for SVD and brain injury, including total WMH burden on cognitive functioning and compare this to. For this purpose, ROI-based linear regression analyses were performed with cognitive functioning as the dependent variable, and the following measures separately entered as independent variables: global WMH volume, WMH volume within strategic white matter tracts (as identified in the first step), the presence of lacunes and microbleeds, and BPF. Finally, we performed a sensitivity analysis in which the ROI-based analyses were repeated after the exclusion of patients with AD.
Demographic and clinical characteristics of the study cohort are shown in Table 1. Forty-eight (28.7%) of the patients had one or more lacunes. WMHs were present in all patients. The cognitive profile of the study cohort is shown in Table 2.
Step 1A: region of interest-based analyses
The aim of these assumption-free ROI-based analyses was to identify strategic white matter tracts in which WMHs are associated with cognitive functioning, independent of global lesion burden. The results are shown in Table 3. First, age, sex and level of education were entered as variables in the linear regression model. Model 1 shows that these variables—as expected—explained a large proportion of the variation for executive functioning (R2 = 37.0%, p<0.001), visuomotor speed (R2 = 49.8%, p<0.001) and memory (R2 = 31.2%, p<0.001). Second, presence of lacunes and total WMH volume were added to model 1 (model 2). Next, WMH volumes within the 11 white matter tracts were each separately added to model 2 (models 3a-k). WMH volume in the anterior thalamic radiation and forceps minor was significantly associated with both executive functioning and visuomotor speed, independent of total WMH volume and presence of lacunes. WMH volume within the forceps minor was significantly associated with memory, independent of total WMH volume and lacunes, but WMH volume within the anterior thalamic radiation was not. Correlations between WMH volume within the remaining nine tracts and cognitive functioning were not statistically significant.
Step 1B: voxel-based lesion-symptom mapping
The aim of this assumption-free voxel-wise analysis was to identify strategic voxels (as opposed to white matter tracts in step 1A) in which WMHs are associated with cognitive functioning, independent of global lesion burden. The spatial distribution of WMHs is displayed in the lesion prevalence map in Fig 2 (panel A). As expected, WMHs were most prevalent in periventricular and frontal white matter regions. The results of the VLSM analyses on executive functioning, visuomotor speed and memory are shown in Fig 2 (panels B-C). VLSM identified a large number of voxels with a significant association between the presence of WMHs and poor executive functioning, visuomotor speed and memory after correction for age, sex, level of education and multiple testing. Several of these voxels remained significant after additional correction for total WMH volume and the presence of lacunes (panels D-E of Fig 2). These voxels were almost exclusively located within the frontal subcortical white matter, mostly within the anterior thalamic radiation and forceps minor (Table 4).
(A) Lesion prevalence map. Voxels with white matter lesions in at least 10 patients are projected onto the MNI-152 template. (B-C): VLSM results for executive functioning (panel B), visuomotor speed (panel C) and memory (panel D) after correction for age, sex, level of education and multiple testing (settings: BM-test; FDR q<0.05). Significant voxels are shown in red. Z-coordinates: -20, -10, 0, 10, 20, 30, 40, 50. (E-F) VLSM results for executive functioning (panel E; coordinate Z-17) and visuomotor speed (panel F; coordinate Z-16) after additional correction for total WMH volume and presence of lacunes. One representative slice is shown per cognitive domain that depicts the location of several significant voxels (in red) in relation to the forceps minor (blue) and anterior thalamic radiation (green). The location of all significant voxels are provided in Table 4.
Step 2: quantification of the impact of WMHs in strategic tracts and other markers for SVD on executive functioning, visuomotor speed and memory
The explained variance in executive functioning, visuomotor speed and memory for each of the different imaging markers is provided in Table 5. WMH volume within the anterior thalamic radiation explained 6.8% of variance in executive functioning on top of age, sex and level of education (p<0.001), compared to 3.9% for total WMH volume (p = 0.001; on average only 11% of total WMH volume was contained in the anterior thalamic radiation). WMH volume within the forceps minor explained 4.6% of variance in visuomotor speed on top of age, sex and level of education (p<0.001), compared to 1.8% for total WMH volume (p = 0.013; on average only 4% of total WMHs were located within the forceps minor). WMH volume within the forceps minor explained 4.2% (p = 0.001) of variance in memory on top of age, sex and level of education (p<0.001), compared to 1.3% (p = 0.072) for total WMH volume. BPF explained 18.4% of variance in executive functioning (p<0.001), 11.7% of variance in visuomotor speed (p<0.001), and 15.6% of variance in memory (p<0.001) on top of age, sex and education. In a multivariable linear regression analysis, BPF was strongly correlated with WMH volume within the anterior thalamic radiation (ß -0.21, 95%CI -0.34 to -0.08; p = 0.002) and forceps minor (ß -0.23, 95%CI -0.36 to -0.11; p<0,001), independent of age, sex, education, lacunes and microbleeds, whereas total WMH volume was not (ß -0.04, 95%CI -0.17 to 0.09; p = 0.574). After adjustment for BPF, the correlations between total WMH volume, WMH volume within the anterior thalamic radiation and forceps minor, and executive functioning and visuomotor speed were attenuated, but remained statistically significant (Table 5), whereas the correlations between total WMH volume and WMH volume within the forceps minor and memory were no longer statistically significant. Lacunes (ß 0.03, 95%CI -0.10 to 0.16 for executive functioning; ß -0.02, 95%CI -0.14 to 0.09 for visuomotor speed; ß 0.10, 95%CI -0.04 to 0.23 for memory) and microbleeds (ß -0.04, 95%CI -0.16 to 0.09 for executive functioning; ß 0.00, 95%CI -0.11 to 0.12 for visuomotor speed; ß 0.00, 95%CI -0.13 to 0.13 for memory) were not significantly associated with cognitive functioning. In the sensitivity analysis, in which the ROI-based analyses were restricted to 125 patients without a diagnosis of AD, the main results were essentially the same (Table 6).
Our findings identify the anterior thalamic radiation and forceps minor as strategic white matter tracts in which WMHs are most strongly associated with cognitive functioning in memory clinic patients with SVD. Importantly, these tracts were identified through corroborative evidence from two statistically independent analysis methods: ROI-based regression analyses and non-parametric voxel-wise analyses. The relevance of WMH location is further emphasized by the results of the ROI-based regression analyses (Table 5), which demonstrate that WMH volume within these white matter tracts explains two to three times more variance in cognition than total WMH volume. This confirms that the impact of WMHs on cognitive functioning depends on location, and indicates that strategically located WMHs have far greater impact on executive functioning, visuomotor speed and memory than global WMH volume in memory clinic patients with sporadic SVD.
An inverse association between total WMH volume and cognitive performance has previously been clearly established at a group level [1,2,36]. However, it remains difficult to determine to what extent WMHs are the cause of cognitive impairment for an individual patient, because WMH burden is clearly not linearly linked to cognition at an individual level. Furthermore, the amount of variance in cognition in memory clinic patients and non-demented older individuals with SVD that is explained by total WMH volume is low, generally in the order of a few percent [3,37,38]. Therefore, the assessment of whether and to what extent clinical symptoms are caused by WMHs is often based on the clinician’s expert judgment . As total WMH volume can only explain a limited proportion of the inter-individual variability in cognitive functioning, it has been suggested that the location of WMHs might be more relevant than total WMH volume. This concept has been supported by previous lesion-symptom mapping studies in patients with CADASIL [4,5], cognitively preserved patients with manifest arterial disease , elderly stroke survivors , patients with AD  and a community-based cohort [6,7]. However, these studies have focused on patients without frank cognitive impairment, patients with a rare genetic cause of SVD, elderly stroke survivors or patients with AD. The domain of older patients with cognitive complaints and sporadic SVD in the memory clinic has—to our knowledge—thus far never been studied. As such, our study is the first to demonstrate a crucial role of strategically located WMHs in this clinically highly relevant population.
It should be noted that we chose to address the spectrum of patients with SVD in a memory clinic setting, rather than limiting our study to patients with a specific clinical diagnosis such as vascular dementia or AD. The rationale behind this is that in a memory clinic setting, SVD most commonly co-occurs with other pathologies . Hence, the impact of SVD on cognition in the setting of a memory clinic should, in our view, be addressed in the context of these other pathologies. We therefore included a representative cohort of patients with different levels of cognitive impairment with SVD in the context of different clinical diagnoses. This inclusion procedure served to ensure generalizability of our findings to patients who attend a memory clinic and have at least a minimal burden of SVD, regardless of the final diagnosis. Evidently, considering the impact of WMHs in the setting of other SVD and non-SVD etiologies required us to address other pathologies and imaging markers on cognitive performance. We performed a sensitivity analysis in which we excluded patients with AD. The results were essentially the same, although the impact of WMHs on cognition appeared to be even greater in this subset of patients (i.e. higher explained variance and standardized coefficients in the regression model), as might be expected when excluding patients who are diagnosed with dementia of non-vascular etiology. Brain atrophy, as measured by BPF, was the most relevant imaging marker in explaining variance in executive functioning (18.4%), visuomotor speed (11.7%) and memory (15.6%) independent of age, sex and education. This is not surprising considering that brain atrophy is a very robust marker for brain injury due to either SVD or Alzheimer pathology. The importance of brain atrophy as a marker for SVD is illustrated by a previous study in patients with CADASIL, a monogenic form of SVD, in which brain atrophy was more strongly associated with cognitive functioning than WMH and lacunar volume [4,39]. Of note, the mean age of these CADASIL patients was 48 years, suggesting that these patients had 'pure' SVD, without co-occurring AD . Our results indicate that in memory clinic patients with sporadic SVD, WMH volume (both global and within the anterior thalamic radiation and forceps minor) is of additional value in explaining variance in executive functioning and visuomotor speed independent of brain atrophy, though this was not the case for memory (Table 5). Of note, we also observed an interplay between WMH location and brain atrophy: WMH volume in the anterior thalamic radiation and forceps minor was strongly correlated with brain atrophy, independent of age, sex, education, microbleeds and lacunes, whereas global WMH volume was not. This explains our observation that the increase in explained variance in cognitive functioning was similar for global and regional WMH volumes when BPF was included in the model (Table 5). This interplay suggests that atrophy and WMHs in specific locations may be linked causally, which is in agreement with previous studies in AD patients in which an association between brain atrophy and WMHs located in the periventricular white matter and cholinergic pathways was found [10,12]. Whether this involves shared causal factors for WMHs and atrophy (i.e. certain processes causing both atrophy and WMHs in particular locations) or neurodegeneration secondary to WMHs remains to be determined. Of note, previous studies have shown that damage to subcortical fiber tracts causes secondary cortical neurodegeneration and focal thinning of connected cortices . In general, our findings underline the importance of taking lesion location into account when studying the role of SVD in cognitive decline and dementia, either alone or in combination with neurodegeneration. Further studies, particularly longitudinal, are needed to determine the relation between regional WMH volumes and other MRI markers for cognitive decline and dementia, their prognostic value, and whether regional WMH volumes have any value in predicting response to pharmacological treatment.
The present study has several strengths. First, we applied an extensive, standardized neuropsychological test battery and MRI scan protocol. Second, we included a substantial number of memory clinic patients with a high level of cognitive dysfunction and high WMH burden and, consequently, could include a large number of white matter tracts in our analyses. Finally, we have used two independent hypothesis-free methods (ROI-based linear regression models and non-parametric VLSM) that both identified the anterior thalamic radiation and forceps minor as strategic white matter tracts, thereby providing converging evidence. An additional benefit of combining these methods is that they have complementary strengths as VLSM provides very high spatial resolution (association per voxel), while tract-based analyses take into account the cumulative burden of lesions in a specific tract. There are also several potential limitations. First, we were unable to assess the potential influence of lacune and microbleed location on cognitive performance in this study. Despite the fact that these lesions were quite common, their small volume did not permit lesion-symptom mapping. Second, we have not analyzed potential lateralized functions of white matters tracts, since we merged bilateral tracts for the tract-based analyses. Previous studies, including some lesion-symptom mapping studies on the cognitive impact of vascular lesions, suggest that the left hemisphere plays a prominent role in processing speed tasks . In our VLSM analyses (after correction for age, sex, education, lacunes and total WMH volume), we observed no clear lateralization of significant voxels for executive functioning, visuomotor speed or memory. Third, the scan of each individual had to be registered to the MNI-152 template in order to perform VLSM. This brain template is based on the brains of young healthy subjects. The degree of deformation of an individual’s brain depends on innate brain size and acquired atrophy. In subjects with an innate smaller brain or severe atrophy (in which case the white matter tracts will be more densely packed), the brain and WMHs are proportionally enlarged to match the template during registration. In order to optimize the quality of the registration procedure for patients with severe atrophy, we have used an Alzheimer atlas as intermediate step in the registration procedure, followed by registration to the MNI atlas. Fourth, WMH due to sporadic SVD only rarely affect the temporal regions: only four patients had WMH in the cingulum of the hippocampus, whereas all 167 patients had a certain amount of WMH in the forceps minor. This explains why no association was found between WMH and memory in the cingulum of the hippocampus even though this tract is known to be crucially involved in memory processes. Fifth, we did not distinguish between specific etiologies underlying the WMHs, such as cerebral amyloid angiopathy (CAA) or hypertensive vasculopathy . Of note, the prevalence of microbleeds in our cohort is higher than generally reported in patients with AD and the general elderly population [42–44], but it should be noted that this is also due to the inclusion criteria for the current study, which focused on memory clinic patients manifesting small vessel disease. Finally, it would be interesting to apply a multi-level analysis in which markers of Alzheimer pathology and measures of microstructural integrity are integrated with data on the presence and location of vascular lesions. However, these data were not available for all patients and such an approach might require an even larger study sample.
The current study identifies the anterior thalamic radiation and forceps minor as strategic white matter tracts in which WMHs are most strongly associated with executive functioning, visuomotor speed and memory performance in memory clinic patients with SVD. Our findings underline the concept that WMHs located in crucial white matter tracts most strongly affect cognition. Therefore, future studies on the pathogenesis of vascular cognitive impairment and the role of sporadic SVD in dementia should not only focus on total WMH volume, but should also take WMH location into account.
Nick A. Weaver is a medical student participating in the Honours program of the Faculty of Medicine, UMC Utrecht. Hugo J. Kuijf was financially supported by the project Brainbox (Quantitative analysis of MR brain images for cerebrovascular disease management), funded by the Netherlands Organisation for Health Research and Development (ZonMw) in the framework of the research programme IMDI (Innovative Medical Devices Initiative); project 104002002.
- Conceptualization: JMB NAW GJB CPLHC.
- Data curation: SH CPLHC.
- Formal analysis: JMB NAW.
- Funding acquisition: CPLHC GJB.
- Investigation: JMB NAW GJB CPLHC.
- Methodology: JMB NAW GJB CPLHC.
- Project administration: CPLHC.
- Resources: SH CPLHC.
- Software: HJK.
- Supervision: GJB CPLHC.
- Validation: JMB NAW.
- Visualization: JMB NAW.
- Writing – original draft: JMB NAW.
- Writing – review & editing: GJB CPLHC SH HJK MKI XX BYT NV AP.
- 1. Gorelick PB, Scuteri A, Black SE, Decarli C, Greenberg SM, Iadecola C, et al. Vascular Contributions to Cognitive Impairment and Dementia. Stroke 2011;42:2672–2713. pmid:21778438
- 2. Prins ND, Scheltens P. White matter hyperintensities, cognitive impairment and dementia: an update. Nat Rev Neurol 2015;11:157–165. pmid:25686760
- 3. Biesbroek JM, Kuijf HJ, van der Graaf Y, Vincken KL, Postma A, Mali WP, et al. Association between subcortical vascular lesion location and cognition: a voxel-based and tract-based lesion-symptom mapping study. PLoS One 2013;8:e60541. pmid:23593238
- 4. Duering M, Zieren N, Hervé D, Jouvent E, Reyes S, Peters N, et al. Strategic role of frontal white matter tracts in vascular cognitive impairment: a voxel-based lesion-symptom mapping study in CADASIL. Brain 2011;134:2366–2375. pmid:21764819
- 5. Duering M, Gonik M, Malik R, Zieren N, Reyes S, Jouvent E, et al. Identification of a strategic brain network underlying processing speed deficits in vascular cognitive impairment. Neuroimage 2013;66:177–183. pmid:23153965
- 6. Duering M, Gesierich B, Seiler S, Pirpamer L, Gonik M, Hofer E, et al. Strategic white matter tracts for processing speed deficits in age-related small vessel disease. Neurology 2014;82:1946–1950. pmid:24793184
- 7. Smith EE, Salat DH, Jeng J, McCreary CR, Fischl B, Schmahmann JD, et al. Correlations between MRI white matter lesion location and executive function and episodic memory. Neurology 2011;76:1492–1499. pmid:21518999
- 8. Bolandzadeh N, Davis JC, Tam R, Handy TC, Liu-Ambrose T. The association between cognitive function and white matter lesion location in older adults: a systematic review. BMC Neurol. 2012 Oct 30;12:126. pmid:23110387
- 9. Burton EJ, Kenny RA, O'Brien J, Stephens S, Bradbury M, Rowan E, et al. White matter hyperintensities are associated with impairment of memory, attention, and global cognitive performance in older stroke patients. Stroke. 2004 Jun;35(6):1270–5. pmid:15118186
- 10. Ramirez J, McNeely AA, Scott CJ, Stuss DT, Black SE. Subcortical hyperintensity volumetrics in Alzheimer's disease and normal elderly in the Sunnybrook Dementia Study: correlations with atrophy, executive function, mental processing speed, and verbal memory. Alzheimers Res Ther. 2014 Aug 11;6(4):49. pmid:25478020
- 11. O'Brien JT, Wiseman R, Burton EJ, Barber B, Wesnes K, Saxby B, et al. Cognitive associations of subcortical white matter lesions in older people. Ann N Y Acad Sci. 2002 Nov;977:436–44. pmid:12480784
- 12. McNeely AA, Ramirez J, Nestor SM, Zhao J, Gao F, Kiss A, et al. Cholinergic subcortical hyperintensities in Alzheimer's disease patients from the Sunnybrook Dementia Study: relationships with cognitive dysfunction and hippocampal atrophy. J Alzheimers Dis. 2015;43(3):785–96. pmid:25114078
- 13. McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA work group under the auspices of department of health and human services task force on Alzheimer’s disease. Neurology 1984;34:939–944. pmid:6610841
- 14. Wiederkehr S, Simard M, Fortin C, van Reekum R. Validity of the clinical diagnostic criteria for vascular dementia: a critical review. J Neuropsychiatry Clin Neurosci 2008;20:150–177. pmid:18451186
- 15. Fazekas F, Chawluk JB, Alavi A, Hurtig HI, Zimmerman RA. MR signal abnormalities at 1.5 T in Alzheimer's dementia and normal aging. AJR Am J Roentgenol. 1987 Aug;149:351–356. pmid:3496763
- 16. Cordonnier C, Potter GM, Jackson CA, Doubal F, Keir S, Sudlow CL, et al. Improving interrater agreement about brain microbleeds: development of the Brain Observer Microbleed Scale (BOMBS). Stroke 2009; 40:94–99. pmid:19008468
- 17. Wardlaw JM, Smith EE, Biessels GJ, Cordonnier C, Fazekas F, Frayne R, et al. Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. Lancet Neurol 2013;12:822–838. pmid:23867200
- 18. Kuijf HJ. Image processing techniques for quantification and assessment of brain MRI. Dissertation, Utrecht University. 2013. Available: http://www.isi.uu.nl/Research/Publications/publicationview.php?id=2545
- 19. Ritter F, Boskamp T, Homeyer A, Laue H, Schwier M, Link F, et al. Medical Image Analysis: A visual approach. IEEE Pulse 2011;2:60–70. pmid:22147070
- 20. Ashburner J, Friston KJ. Unified segmentation. Neuroimage 2005;26:839–851. pmid:15955494
- 21. Charlton RA, Morris RG, Nitkunan A, Markus HS. The cognitive profiles of CADASIL and sporadic small vessel disease. Neurology 2006;66:1523–1526. pmid:16717212
- 22. Dong Y, Gan DZ, Tay SZ, Koay WI, Collinson SL, Hilal S, et al. Patterns of neuropsychological impairment in Alzheimer's disease and mixed dementia. J Neurol Sci 2013;333:5–8. pmid:23978422
- 23. Elia LFD, Satz P, Uchiyama CL, Uchiyama CL, White T. Color Trails Test Professional Manual. Florida: Psychological Assessment Resources, Inc; 1996.
- 24. Carew TG, Lamar M, Cloud BS, Grossman M, Libon DJ. Impairment in category fluency in ischemic vascular dementia. Neuropsychology 1997;11:400–412. pmid:9223144
- 25. Diller L, Ben-Yishay Y, Gerstman LJ. Studies in cognition and rehabilitation in hemiplegia. New York: New York University Medical Center Institute of Rehabilitation Medicine; 1974.
- 26. Porteus SD. The Maze Test and Clinical Psychology. Palo Alto, CA: Pacific Books; 1959.
- 27. Smith A. Symbol Digit Modalities Test (SDMT) manual, revised ed. Los Angeles: Western Psychological Services; 1982.
- 28. Sahadevan S, Tan NJ, Tan TC, Tan S. Cognitive testing of elderly Chinese from selected community clubs in Singapore. Ann Acad Med Singapore 1997;26:271–7. pmid:9285015
- 29. Wechsler D. Wechsler Memory Scale-Revised. San Antonio, TX: Harcourt Brace Jovanovich; 1997.
- 30. Fonov VS, Evans AC, McKinstry RC, Almli CR, Collins DL. Unbiased nonlinear average age-appropriate brain templates from birth to adulthood. Neuroimage 2009;47: S102.
- 31. Klein S, Staring M, Murphy K, Viergever MA, Pluim JP. Elastix: A toolbox for intensity-based medical image registration. IEEE Transactions on Medical Imaging 2010;29:196–205. pmid:19923044
- 32. Mega MS, Dinov ID, Mazziotta JC, Manese M, Thompson PM, Lindshield C, et al. Automated brain tissue assessment in the elderly and demented population: construction and validation of a sub-volume probabilistic brain atlas. Neuroimage 2005;26:1009–1018. pmid:15908234
- 33. Rorden C, Karnath H. Using human brain lesions to infer function: a relic from a past era in the fMRI age? Nature 2004;5:813–815.
- 34. Rorden C, Karnath H-O, Bonilha L. Improving lesion-symptom mapping. J Cogn Neurosci 2007;19:1081–1088. pmid:17583985
- 35. Hua K, Zhang J, Wakana S, Jiang H, Li X, Reich DS, et al. Tract probability maps in stereotaxic spaces: Analyses of white matter anatomy and tract-specific quantification. NeuroImage 2008;39:336–347. pmid:17931890
- 36. Kloppenborg RP, Nederkoorn PJ, Geerlings MI, van den Berg E. Presence and progression of white matter hyperintensities and cognition: a meta-analysis. Neurology 2014;82:2127–2138. pmid:24814849
- 37. Reijmer YD, Leemans A, Caeyenberghs K, Heringa SM, Koek HL, Biessels GJ. Disruption of cerebral networks and cognitive impairment in Alzheimer disease. Neurology 2013;80:1370–1377. pmid:23486876
- 38. van Norden AG, van Uden IW, de Laat KF, van Dijk EJ, de Leeuw FE. Cognitive function in small vessel disease: the additional value of diffusion tensor imaging to conventional magnetic resonance imaging: the RUN DMC study. J Alzheimers Dis 2012;32:667–676. pmid:22850313
- 39. Chabriat H, Hervé D, Duering M, Godin O, Jouvent E, Opherk C, et al. Predictors of Clinical Worsening in Cerebral Autosomal Dominant Arteriopathy With Subcortical Infarcts and Leukoencephalopathy: Prospective Cohort Study. Stroke. 2016 Jan;47(1):4–11. pmid:26578659
- 40. Duering M, Righart R, Wollenweber FA, Zietemann V, Gesierich B, Dichgans M. Acute infarcts cause focal thinning in remote cortex via degeneration of connecting fiber tracts. Neurology. 2015 Apr 21;84(16):1685–1692. pmid:25809303
- 41. Biffi A, Greenberg SM. Cerebral amyloid angiopathy: a systematic review. J Clin Neurol. 2011 Mar;7(1):1–9. pmid:21519520
- 42. Sepehry AA, Lang D, Hsiung GY, Rauscher A. Prevalence of Brain Microbleeds in Alzheimer Disease: A Systematic Review and Meta-Analysis on the Influence of Neuroimaging Techniques. AJNR Am J Neuroradiol. 2016 Feb;37(2):215–22. pmid:26635284
- 43. Vernooij MW, van der Lugt A, Ikram MA, Wielopolski PA, Niessen WJ, Hofman A, et al. Prevalence and risk factors of cerebral microbleeds: the Rotterdam Scan Study. Neurology. 2008 Apr 1;70(14):1208–14. pmid:18378884
- 44. Poels MM, Vernooij MW, Ikram MA, Hofman A, Krestin GP, van der Lugt A, et al. Prevalence and risk factors of cerebral microbleeds: an update of the Rotterdam scan study. Stroke. 2010 Oct;41(10 Suppl):S103–6. pmid:20876479