ARTICLE IN BRIEF
Using a statistical analysis of images rendered through brain diffusion tensor tractography, investigators found that brain network connectivity in small-vessel disease (SVD) was disturbed compared with healthy controls. The association explains, in part, cognitive declines associated with SVD.
Declines in cognitive function and processing speed associated with small-vessel disease (SVD) can be partly explained by disruptions in the brain's structural network, a cross-sectional study has found. The report raises the possibility that advanced imaging tools that assess the strength and efficiency of the brain's structural network could provide reliable, clinically meaningful biomarkers for cognitive decline, the study authors and imaging experts not involved with the study told Neurology Today.
The investigators compared diffusion tensor images (DTIs) from patients with SVD and healthy controls, and used diffusion tensor tractography (DTT) to reconstruct and map the large-scale connectivity patterns from the DTI in the brain. They then used a statistical method (graph analysis) to characterize the organization of white matter connections by delineating a network as sets of interacting brain regions connected by white matter tracts.
Among their findings, they reported that cognitive function was impaired in SVD for executive function (p<0.0001) and processing speed (p<0.0001), but not for working memory (p=0.12) or long-term (episodic) memory (p=0.9).
They also found that the connectivity measures provided by the DTT offered anatomical and functional information that was not apparent on the MRI images alone.
“In multiple regression models controlling for confounding variables, associations with cognition were stronger for network measures than other MRI measures including conventional diffusion tensor imaging measures,” the British research team reported in the June 20 online edition of Neurology.
“Networks were less dense, connection weights were lower and measures of network efficiency were significantly disrupted” in SVD patients compared with controls, the study authors wrote.
“Dealing with the brain as a network is quite useful I think,” said the first author of the study, Andrew J. Lawrence, PhD, a research psychologist at the Centre for Clinical Neuroscience at St. George's University of London. “What we saw in small vessel disease is that the white-matter tracts are less coherent, and it was a diffuse effect. There weren't any connections that were completely lost, but there weren't any that were completely unaffected either.”
The cross-sectional study involved 115 patients with lacunar infarction and leukoaraiosis and 50 healthy individuals. The SVD patients were participating in a longitudinal study, not yet completed, investigating the relationship between MRI markers and cognition in SVD, the St. George's Cognition and Neuroimaging in Stroke Study. Structural connectivity was estimated between 90 cortical and subcortical brain regions, and the efficiency of those connections were measured with graph analysis.
A neuropsychologist obtained cognitive measures of premorbid IQ and four broad cognitive domains (executive function, processing speed, working memory and long-term memory).
A single, widely distributed subnetwork of 27 unique nodes [brain regions] and 29 edges [white matter tracts] jumped out as the most impaired. It involved a third of all inter-hemispheric connections, including all major subdivisions of the corpus callosum and intersected centrum semiovale white matter lesions. Other impaired association tracts, the study found, predominantly involved the prefrontal cortex, including fronto-frontal connections of the superior frontal gyrus and pathways between the inferior frontal cortex and parietal and temporal regions.
The study could not determine, however, whether the decreased efficiency is caused by the loss of edges in the network — the network “density in the language of graph analysis — or alterations in the arrangement of the network edges or both, said Dr. Lawrence. “At the moment, the techniques we have can't tell us which is driving what we observe,” he said.
With recent network-efficiency studies also finding the method to be a highly accurate marker of cognitive function in patients with type 2 diabetes and Alzheimer's disease, as well as in studies of healthy people, researchers and clinicians alike might soon adopt the method as a new standard for imaging, neurologists and neuroscientists told Neurology Today.
“The use of tractography is not yet found in most clinical settings, but that's tending to change, especially as the software matures,” said Arthur Toga, PhD, a researcher at the University of Southern California's Keck School of Medicine Laboratory of Neuro Imaging. “Until now, people have looked at SVD in terms of the burden of lacunes or white matter hyperintensities, but not in networks. You look at these images and they tell an extraordinary story of network changes in the brain distant from where the initial injury was.” In that regard, he said, “This study is an important milestone.”
The potential for using measures of structural network efficiency as a biomarker for disease progression, prognosis or treatment response is high, said Yael Reijmer, PhD, who coauthored a commentary accompanying the study, and who has been studying the technique as a postdoctoral researcher in the Stroke Research Center at Massachusetts General Hospital.
“Because it's a relatively new technique, we don't yet know enough about the reproducibility of these measures,” Dr. Reijmer said. “We need longitudinal studies. But so far the cross-sectional studies have been consistent, showing that these markers of connectivity are more sensitive than the measures of lacunes or microbleeds.”
If the technique holds up as a biomarker of cognitive impairment, “that would be huge,” said Jorge Sepulcre, MD, PhD, an assistant professor in the division of nuclear medicine and molecular imaging at Harvard Medical School.
“It could save us a lot of time even in clinical practice,” Dr. Sepulcre said. “In the past, clinical groups have been trying to use volume of lesions, location of lacunar infarcts, and such, to investigate cognitive impairment in SVD. This study is trying to jump to the level of network connectivity, aiming to increase detection capabilities of cognitive changes. Once these software tools are broadly available for getting this information from an MRI and putting it onto the doctor's desk, it may be easier to base clinical decisions on it.”
While commending the study as well designed and well conducted, Dr. Sepulcre said his one criticism was its use of the 90 brain regions obtained from the Automated Anatomical Labeling atlas, which has been widely used in network studies.
“It's a well known scheme,” Dr. Sepulcre acknowledged. “But the field has moved to a greater level of detail — the voxel level. That's why I'm not totally happy.” Even so, he added, “If they have found these impressive results with the 90 regions, which are at a much lower resolution than voxels, that probably means it's very robust. But I still want to see it applied it to a higher resolution.”
LINK UP FOR MORE INFORMATION:
•. Lawrence AJ, Chung AW, Morris RG, et al. Structural network efficiency is associated with cognitive impairment in small-vessel disease. Neurology 2014; Epub 2014 Jun 20.
•. Marshall RS, Reijmer YD. Connectivity at a crossroads: What white matter integrity can tell us about cognitive impairment. Neurology 2014; Epub 2014 Jun 20.
•. Reijmer YD, Leemans A, Caeyenberghs K, et al. Disruption of cerebral networks and cognitive impairment in Alzheimer disease. Neurology
2013; 80 (15): 1370–1377.
•. Lawrence AJ, Patel B, Morris RG, et al. Mechanisms of cognitive impairment in cerebral small vessel disease: multimodal MRI results from the St George's cognition and neuroimaging in stroke (SCANS) study.” PLoS One 8(4):e61014
•. Fischer FU, Wolf D, Scheurich A, et al. Association of structural global brain network properties with intelligence in normal aging. PLoS One
2014; 9(1): e86258.
•. Mueller S, Wang D, Fox MD, et al. Individual variability in functional connectivity architecture of the human brain. Neuron 2013;77(3): 586–595.