Neuroanatomical Signature of the Transition from Normal Cognition to MCI in Parkinson's Disease

Iman Beheshti,Jarrad Perron,Ji Hyun Ko

DOI: https://doi.org/10.14336/AD.2024.0323

2024-06-07

Abstract:The progression of Parkinson's disease (PD) is often accompanied by cognitive decline. We had previously developed a brain age estimation program utilizing structural MRI data of 949 healthy individuals from publicly available sources. Structural MRI data of 244 PD patients who were cognitively normal at baseline was acquired from the Parkinson Progression Markers Initiative (PPMI). 192 of these showed stable normal cognitive function from baseline out to 5 years (PD-SNC), and the remaining 52 had unstable normal cognition and developed mild cognitive impairment within 5 years (PD-UNC). 105 healthy controls were also included in the analysis as a reference. First, we examined if there were any baseline differences in regional brain structure between PD-UNC and PD-SNC cohorts utilizing the three most widely used atrophy estimation pipelines, i.e., voxel-based morphometry (VBM), deformation-based morphometry and cortical thickness analyses. We then investigated if accelerated brain age estimation with our multivariate regressive machine learning algorithm was different across these groups (HC, PD-SNC, and PD-UNC). As per the VBM analysis, PD-UNC patients demonstrated a noticeable increase in GM volume in the posterior and anterior lobes of the cerebellum, sub-lobar, extra-nuclear, thalamus, and pulvinar regions when compared to PD-SNC at baseline. PD-UNC patients were observed to have significantly older brain age compared to both PD-SNC patients (p=0.009) and healthy controls (p<0.009). The increase in GM volume in the PD-UNC group could potentially indicate an inflammatory or neuronal hypertrophy response, which could serve as a biomarker for future cognitive decline among this population.