Baseline and longitudinal grey matter changes in newly diagnosed Parkinson's disease: ICICLE-PD study

Abstract

Mild cognitive impairment in Parkinson's disease is associated with progression to dementia (Parkinson's disease dementia) in a majority of patients. Determining structural imaging biomarkers associated with prodromal Parkinson's disease dementia may allow for the earlier identification of those at risk, and allow for targeted disease modifying therapies. One hundred and five non-demented subjects with newly diagnosed idiopathic Parkinson's disease and 37 healthy matched controls had serial 3 T structural magnetic resonance imaging scans with clinical and neuropsychological assessments at baseline, which were repeated after 18 months. The Movement Disorder Society Task Force criteria were used to classify the Parkinson's disease subjects into Parkinson's disease with mild cognitive impairment (n = 39) and Parkinson's disease with no cognitive impairment (n = 66). Freesurfer image processing software was used to measure cortical thickness and subcortical volumes at baseline and follow-up. We compared regional percentage change of cortical thinning and subcortical atrophy over 18 months. At baseline, cases with Parkinson's disease with mild cognitive impairment demonstrated widespread cortical thinning relative to controls and atrophy of the nucleus accumbens compared to both controls and subjects with Parkinson's disease with no cognitive impairment. Regional cortical thickness at baseline was correlated with global cognition in the combined Parkinson's disease cohort. Over 18 months, patients with Parkinson's disease with mild cognitive impairment demonstrated more severe cortical thinning in frontal and temporo-parietal cortices, including hippocampal atrophy, relative to those with Parkinson's disease and no cognitive impairment and healthy controls, whereas subjects with Parkinson's disease and no cognitive impairment showed more severe frontal cortical thinning compared to healthy controls. At baseline, Parkinson's disease with no cognitive impairment converters showed bilateral temporal cortex thinning relative to the Parkinson's disease with no cognitive impairment stable subjects. Although loss of both cortical and subcortical volume occurs in non-demented Parkinson's disease, our longitudinal analyses revealed that Parkinson's disease with mild cognitive impairment shows more extensive atrophy and greater percentage of cortical thinning compared to Parkinson's disease with no cognitive impairment. In particular, an extension of cortical thinning in the temporo-parietal regions in addition to frontal atrophy could be a biomarker in therapeutic studies of mild cognitive impairment in Parkinson's disease for progression towards dementia.

Keywords: Parkinson’s disease; dementia; neurodegeneration; neuroimaging.

Mild cognitive impairment in Parkinson’s disease (PDMCI) is associated with progression to dementia in a majority of patients. Mak et al. reveal accelerated cortical thinning in patients with PDMCI compared to non-cognitively impaired patients and healthy controls. Patterns of cortical thinning may constitute biomarkers for increased dementia risk.

Figure 1

Reduced regional cortical thickness in PD-MCI compared to healthy controls at baseline. No significant differences in baseline cortical thickness were found between PD-MCI and PD-NC, and between PD-NC and healthy controls. The colour bar shows the logarithmic scale of P-values (−log₁₀).

Figure 2

Vertex-wise comparisons of percentage change in cortical thinning over 18 months. (A) PD-MCI < HC; (B) PD-NC < HC; (C) PD-MCI < PD-NC. The colour bar shows the logarithmic scale of P-values (−log₁₀). Lh = left hemisphere; Rh = right hemisphere.

Figure 3

Association between cognitive measures and regional cortical thickness at baseline. The colour bar shows the logarithmic scale of P-values (−log₁₀). PD = Parkinson’s disease; Lh = left hemisphere; Rh = right hemisphere.