P3‐227: MILD COGNITIVE IMPAIRMENT: ASSOCIATION OF NEURODESTRUCTION MARKERS WITH REGIONAL ATROPHY OF THE BASAL FOREBRAIN CHOLINERGIC SYSTEM AND ITS PROJECTING CORTICAL AREAS

Previous studies showed highly significant changes of cortical thickness and volumetric measurements of the basal forebrain cholinergic system (BFCS) in patients with Alzheimer's disease (AD) compared to cognitive healthy controls (HC). In this study we analysed structural parameters in participants with mild cognitive impairment (MCI) - a risk state of AD. The aim of the study was to identify possible associations of cerebro-spinal-fluid (CSF) markers on structural magnetic resonance imaging (MRI) in these patients group. We used 169 high-resolution 3-D structural MRI datasets from four centers of the European DTI Study on Dementia (EDSD) including 103 participants with MCI following the Petersen criteria and 66 HC. 67 participants had additional CSF analysis, 51 of this group had positive tau, phospho-tau or amyloid 1-42 levels. We used a subregional specific map of the BFCS (following Mesulam's nomenclature) which is based on post mortem MRI in cranio data for the voxel-based morphometry. Cortical thickness was analysed using FreeSurfer software. Comparing the CSF positive MCI group (MCIp) with the CSF negative MCI (MCIn), three subregions of the BFCS showed significant differences: Ch4p (p=0.018), Nucleus subputaminalis (p=0.024) and the Ch3 (p=0.024) had significant less volume in the MCIp group than in the MCIn group. Cortical thickness of areas projecting from Ch4p (e.g. temporal pole (p=0.009) and medial temporal lobe (p=0.015)) was significantly reduced in MCIp compared to MCIn. The parahippocampal cortex (p=0.03) - which cholinergic source Ch2 did not show any significant differences between the MCIn and MCIp groups - showed a significant thinning in the MCIp group compared to the MCIn. Structural changes in the BFCS and the cortical thickness are significantly pronounced in patients with MCI and positive CSF marker compared to MCI patients without positive CSF marker. Our data supports the notion of a corticotopy of BFCS subnuclei (Ch4p) as proposed by post mortem findings in non human primates, but the influence of CSF marker of neurodestruction on the BFCS still needs further investigation. In a further step, we will compare BFCS and cortical thickness changes in MCI with cognitively healthy controls.