The Effect of Lacunar Infarcts on White Matter Tract Integrity
Background and Purpose—Lacunar infarcts may cause disturbances of the white matter (WM) structure remote from the primary lesion. Here, we used diffusion MRI and tractography to (1) spatially characterize microstructural abnormalities along WM tracts containing a lacunar infarct and (2) relate abnormalities in remote parts of the affected WM tract to cognitive outcome.
Methods—In 17 participants with a lacunar infarct, we reconstructed the affected WM tract using fiber tractography. The corresponding nonlesioned tract in the contralateral hemisphere served as a control tract. Diffusion parameters (fractional anisotropy and mean diffusivity) were plotted along the tract and related to measures of memory, executive functioning and information processing speed.
Results—Diffusion abnormalities remote from the lacune were present in the affected tract compared with the control tract up to 2 cm from the lacune (9% to 17% decrease in fractional anisotropy, 11% to 27% increase in mean diffusivity; P<0.05). The severity of these abnormalities attenuated with increasing distance to the primary lesion. Furthermore, the degree of remote WM disturbances was related to worse cognitive functioning on all 3 domains, independent of the size of the lacune (r=0.6–0.8; P<0.05).
Conclusions—Lacunar infarcts are associated with abnormalities in the affected WM tract that extend centimeters beyond the lesion visible on conventional MRI. These secondary WM abnormalities may contribute to the cognitive deficits observed in patients with subcortical infarcts.
Lacunar infarcts are generally considered as focal lesions, visible on conventional MRI images. However, it is suggested that ischemic infarcts have more widespread effect on the white matter (WM) microstructure attributable to, for example, secondary degeneration of the affected WM tract or inflammatory responses.1,2 The microstructural alterations of the WM can be studied in vivo using diffusion tensor imaging.3 Recently, it is has become possible to spatially characterize WM diffusion abnormalities along the pathway of a specific tract. This method can reveal localized diffusion abnormalities that may not be apparent using an ROI-based or a tract-averaged approach.4 Until now, however, no study has used this method to examine whether lacunar infarcts are associated with local or more widespread disruptions of the WM structure, and whether these disruptions contribute to worse cognitive outcome. In the present study, we examined the WM microstructure of tracts containing a lacunar infarct and evaluated whether potential WM abnormalities within the affected tract, but remote from the primary lesion, are related to worse cognitive functioning.
Participants (n=17; age range, 58–87; 57% men) were selected from an ongoing research program on brain MRI markers of vascular cognitive impairment at the University Medical Center Utrecht5,6 (for details see Methods in the online-only Data Supplement). To be eligible for the present study, participants should have had a dedicated 3T MRI scan and no diagnosis of dementia or a Mini Mental State Examination <26, or other known neurological disease apart from small vessel disease/noninvalidating stroke. Participants with an infarct in or near the corresponding tract in the contralateral hemisphere, and with a large cortical infarct (>1.5 cm) were excluded. Of the 17 patients, 12 had multiple lacunar infarcts on MRI and 5 of them reported a history of clinical manifest stroke or transient ischemic attack.
MRI data were acquired on a Philips 3.0 Tesla scanner using a standardized protocol,5 including a diffusion-weighted scan (single-shot spin echo echo planar imaging sequence, 48 contiguous slices, acquired isotropic voxel size 2.50 mm, 45 directions, b-value: 1200 s/mm2, 1 b=0 s/mm2), a 3-dimensional T1, and a fluid-attenuated inversion recovery scan. Details on the data processing can be found in the Methods in the online-only Data Supplement. Total WM hyperintensity load was assessed on fluid-attenuated inversion recovery scans using a visual rating scale. Cognition was assessed by the Rey Auditory Verbal Learning Test (memory), the Stroop Color-Word Test (information processing speed [card I+II] and executive functioning [card III]), and the Category Verbal Fluency Test (executive functioning).
For each individual, the fractional anisotropy (FA) and mean diffusivity (MD) of the lesioned tract were expressed as percentage of the control tract: (FAlesioned/FAcontrol)×100%. Thus, we obtained a relative measure of FA and MD for each segment, which is independent of interindividual differences in diffusion tensor imaging metrics related to confounders, such as age and small vessel disease.
The relative FA and MD values of the lesioned tract were compared with the control tract using a Wilcoxon signed-rank test. This was performed for the whole tract, after excluding the segment(s) containing the infarct, and for each individual segment along the tract (Figure 1).
In addition, we related the relative FA/MD of the lesioned tract to cognitive functioning, controlling for age, sex, and level of education. Again, the infarct-containing segment(s) was excluded from the analysis. In secondary models, we adjusted these correlations for the side (left/right) and volume of the lacune and for total WM hyperintensity load.
Details on the reconstructed tracts are given in the online-only Data Supplement (Table I in the online-only Data Supplement). Total tract volume did not differ between tracts indicating that tractography was reliably performed in both hemispheres and was not affected by the lacune (Table I in the online-only Data Supplement). Also, total WM hyperintensity load, and the mean FA and MD were not different between the affected and nonaffected hemispheres (P>0.30).
After exclusion of the infarct-containing tract segment(s), diffusion abnormalities were present in the affected tract compared with the control tract, indicated by an overall decrease in FA (93±8%; P=0.002) and an increase in MD (110±13%; P=0.004). Figure 2 shows the relative FA and MD for each segment along the tract. The difference in FA and MD relative to the control tract gradually attenuated with increasing distance to the primary lesion.
A greater decrease in FA in the affected tract relative to the control tract was related to worse executive functioning (Change is Z score per 10% decrease in FA: verbal fluency: −0.85 SD, P=0.024) and information processing speed (Stroop task card I+II: −1.11 SD, P<0.001). A greater increase in MD was also related to slowing of information processing (verbal fluency: −0.49 SD, P=0.033) and to worse memory performance (word-recall task: −0.45 SD, P=0.024). Correlation plots are given in Figure II in the online-only Data Supplement. The correlations remained significant after controlling for the side or the size of the lacune (all P<0.05). Controlling for total WM hyperintensity load only attenuated the relation between MD and memory (word-recall task: −0.38, P=0.107); all other associations remained significant. Repeating the same analyses in WM tracts in the affected hemisphere not containing a lacune revealed no significant relation between relative FA/MD and cognition (see Results in the online-only Data Supplement).
This is the first study to assess the impact of a lacunar infarct on the WM structure along a tract using diffusion tensor imaging and tractography. Our results show structural abnormalities remote from the lacune that are not visible on conventional MRI. The severity of the WM abnormalities attenuates with increasing distance to the primary lesion. This pattern suggests that the mechanisms responsible for the degradation in WM structure are directly related to the ischemic lesion. Furthermore, the degree of secondary WM abnormalities were related to worse cognitive functioning independent of the size of the lacune itself.
Potential mechanisms underlying the alterations in diffusion measures include axonal degeneration, demyelination, and chronic inflammatory processes/reactions in the vicinity of small infarcts.1,7 Previous studies in patients with stroke have demonstrated WM diffusion abnormalities in the so-called normal appearing WM.8 Our findings indicate that subcortical infarcts may directly contribute to these diffusion abnormalities by remote effects on the WM structure along affected WM tracts.
Strengths of this study include the comprehensive scan protocol, including high-resolution diffusion MRI data, and the along-tract-based analysis approach. A limitation of this study is the limited sample size, the fact that we were not blinded for the lacune during tract reconstruction, and lack of information on the time since infarction. However, the fact that we focused on cavitating lacunes indicates that all infarcts were examined at a chronic stage. Future longitudinal studies should further examine how secondary WM abnormalities contribute to the cognitive deficits observed in patients with subcortical infarcts.
The help of S. Heringa, M. Brundel, and W. Bouvy with the recruitment of the data is gratefully acknowledged.
The research of Dr. Biessels is supported by VIDI grant 91711384 from the Netherlands Organisation for Health Research and Development (ZonMw) and by grant 2010T073 from the Netherlands Heart Foundation.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.113.001321/-/DC1.
- Received February 28, 2013.
- Accepted March 26, 2013.
- © 2013 American Heart Association, Inc.