Alterations and test-retest reliability of functional connectivity network measures in cerebral small vessel disease

While structural network analysis consolidated the hypothesis of cerebral small vessel disease (SVD) being a disconnection syndrome, little is known about functional changes on the level of brain networks. In patients with genetically defined SVD (CADASIL, n = 41) and sporadic SVD (n = 46), we independently tested the hypothesis that functional networks change with SVD burden and mediate the effect of disease burden on cognitive performance, in particular slowing of processing speed. We further determined test-retest reliability of functional network measures in sporadic SVD patients participating in a high-frequency (monthly) serial imaging study (RUN DMC-InTENse, median: 8 MRIs per participant). Functional networks for the whole brain and major subsystems (i.e., default mode network, DMN; fronto-parietal task control network, FPCN; visual network, VN; hand somatosensory-motor network, HSMN) were constructed based on resting-state multi-band functional MRI. In CADASIL, global efficiency (a graph metric capturing network integration) of the DMN was lower in patients with high disease burden (standardized beta = -.44; p [corrected] = .035) and mediated the negative effect of disease burden on processing speed (indirect path: std. beta = -.20, p = .047; direct path: std. beta = -.19, p = .25; total effect: std. beta = -.39, p = .02). The corresponding analyses in sporadic SVD showed no effect. Intraclass correlations in the high-frequency serial MRI dataset of the sporadic SVD patients revealed poor test-retest reliability and analysis of individual variability suggested an influence of age, but not disease burden, on global efficiency. In conclusion, our results suggest that changes in functional connectivity networks mediate the effect of SVD-related brain damage on cognitive deficits. However, limited reliability of functional network measures, possibly due to age-related comorbidities, impedes the analysis in elderly SVD patients.

Disruption of rich club organisation in cerebral small vessel disease

Cerebral small vessel disease (SVD) is an important cause of vascular cognitive impairment. Recent studies have demonstrated that structural connectivity of brain networks in SVD is disrupted. However, little is known about the extent and location of the reduced connectivity in SVD. Here they investigate the rich club organisation-a set of highly connected and interconnected regions-and investigate whether there is preferential rich club disruption in SVD. Diffusion tensor imaging (DTI) and cognitive assessment were performed in a discovery sample of SVD patients (n = 115) and healthy control subjects (n = 50). Results were replicated in an independent dataset (49 SVD with confluent WMH cases and 108 SVD controls) with SVD patients having a similar SVD phenotype to that of the discovery cases. Rich club organisation was examined in structural networks derived from DTI followed by deterministic tractography. Structural networks in SVD patients were less dense with lower network strength and efficiency. Reduced connectivity was found in SVD, which was preferentially located in the connectivity between the rich club nodes rather than in the feeder and peripheral connections, a finding confirmed in both datasets. In discovery dataset, lower rich club connectivity was associated with lower scores on psychomotor speed (β = 0.29, P < 0.001) and executive functions (β = 0.20, P = 0.009). These results suggest that SVD is characterized by abnormal connectivity between rich club hubs in SVD and provide evidence that abnormal rich club organisation might contribute to the development of cognitive impairment in SVD. Hum Brain Mapp 38:1751-1766, 2017. © 2017 Wiley Periodicals, Inc.

Diffusion tensor imaging of the hippocampus and verbal memory performance: the RUN DMC study

BACKGROUND

Cerebral small vessel disease (SVD) and hippocampal atrophy are related to verbal memory failures and may ultimately result in Alzheimer's disease. However, verbal memory failures are often present before structural changes on conventional MRI appear. Changes in microstructural integrity of the hippocampus, which cannot be detected with conventional MRI, may be the underlying pathological substrate. With diffusion tensor imaging (DTI), we investigated the relation between the microstructural integrity of the hippocampus and verbal memory performance in 503 nondemented elderly with SVD.

METHODS

The Radboud University Nijmegen Diffusion tensor and Magnetic resonance imaging Cohort study is a prospective cohort study among 503 nondemented elderly with cerebral SVD aged between 50 and 85 years. All participants underwent T1 MPRAGE, fluid-attenuated inversion recovery, DTI scanning and the Rey Auditory Verbal Learning Test. After manual segmentation of the hippocampi, we calculated the mean diffusivity (MD) and fractional anisotropy in both hippocampi. The relation between memory performance and hippocampal DTI parameters was adjusted for age, sex, education, depressive symptoms, hippocampal, and white-matter lesions volume and lacunar infarcts.

RESULTS

We found inverse relations between hippocampal MD and verbal memory performance (β = -0.22; P < 0.001), immediate recall (β = -0.22; P < 0.001), delayed recall (β = -0.20; P < 0.001), and forgetting rate (β = -0.13; P = 0.025), most pronounced in participants with a normal hippocampal volume.

CONCLUSION

Microstructural integrity of the hippocampus assessed by DTI is related to verbal memory performance in elderly with SVD, also in participants with an intact appearing hippocampus. Changes in hippocampal microstructure may be an early marker of underlying neurodegenerative disease, before macrostructural (i.e., volumetric) changes occur.