White matter pathology in Parkinson's disease: The effect of imaging protocol differences and relevance to executive function

Abstract

Diffusion magnetic resonance imaging is increasingly used as a non-invasive method to investigate white matter structure in neurological and neuropsychiatric disease. However, many options are available for the acquisition sequence and analysis method. Here we used Parkinson's disease as a model neurodegenerative disorder to compare imaging protocols and analysis options. We investigated fractional anisotropy and mean diffusivity of white matter in patients and age-matched controls, comparing two datasets acquired with different imaging protocols. One protocol prioritised the number of b value acquisitions, whilst the other prioritised the number of gradient directions. The dataset with more gradient directions was more sensitive to reductions in fractional anisotropy in Parkinson's disease, whilst the dataset with more b values was more sensitive to increases in mean diffusivity. Moreover, the areas of reduced fractional anisotropy were highly similar to areas of increased mean diffusivity in PD patients. Next, we compared two widely used analysis methods: tract-based spatial statistics identified reduced fractional anisotropy and increased mean diffusivity in Parkinson's disease in many of the major white matter tracts in the frontal and parietal lobes. Voxel-based analyses were less sensitive, with similar patterns of white matter pathology observed only at liberal statistical thresholds. We also used tract-based spatial statistics to identify correlations between a test of executive function (phonemic fluency), fractional anisotropy and mean diffusivity in prefrontal white matter in both Parkinson's disease patients and controls. These findings suggest that in Parkinson's disease there is widespread pathology of cerebral white matter, and furthermore, pathological white matter in the frontal lobe may be associated with executive dysfunction. Diffusion imaging protocols that prioritised the number of directions versus the number of b values were differentially sensitive to alternative markers of white matter pathology, such as fractional anisotropy and mean diffusivity.

Imaging changes associated with cognitive abnormalities in Parkinson’s disease

“…White matter changes in these regions were consistently reported in previous TBSS findings (Agosta et al 2013a, b;Deng et al 2013;Hattori et al 2012;Matsui et al 2007;Melzer et al 2013;Rae et al 2012). We detected group differences with an increase in MD (but not in FA) in our patients.…”

“…Contrary to gray matter findings, in which regional changes are more common in non-demented PD, white matter changes are reported in more diffuse brain areas even in non-demented PD (Hattori et al 2012;Kim et al 2013;Melzer et al 2013;Theilmann et al 2013;Zheng et al 2014), and the frontal region was one of the most consistently reported regions for white matter changes (Agosta et al 2013b;Deng et al 2013;Gattellaro et al 2009;Rae et al 2012;Zhan et al 2012;Zhang et al 2011). Moreover, variability in FA and/ or MD was primarily associated with cognitive measures (Agosta et al 2013b;Gallagher et al 2013;Rae et al 2012;Theilmann et al 2013;Zheng et al 2014. ) Despite the evidence of structural abnormalities and variability associated with clinical and cognitive manifestations in non-demented PD, whether structural abnormalities account for specific clinical sequelae in PD patients is still unclear.…”

Increased connectivity between sensorimotor and attentional areas in Parkinson’s disease

“…Previous studies evaluating cerebral changes induced by Parkinson's disease found white-matter diffusivity alterations [71,72]. However, the methodology in these studies is somewhat different than ours, with one study specifically evaluating white matter regions of interest (ROIs) [71] while the other evaluating different protocols and analyses [72]. The latter emphasizes that choice of protocol can be an important factor affecting diffusion findings, and it is possible that our study was affected by the choice of protocol due to technical limitations of our MRI system (e.g., number of b values).…”

Abnormal functional connectivity density in Parkinson's disease

“…Given that FC is closely related to white matter connections as well as monosynaptic or polysynaptic structural connectivity [42][43][44], alterations in FCD may be related to the suboptimal integrity of the underlying structural connectivity. In fact, many previous studies have identified widespread white matter pathology in PD patients according to altered fractional anisotropy, mean diffusivity and track density [45][46][47][48]. Moreover, the presence of cortical Lewy bodies and Lewy neurites in PD patients was indicative of altered synaptic structural connectivity [49,50].…”