REGION-SPECIFIC ATROPHY AS MEASURED BY CORTICAL GRAY MATTER VOLUME IS ASSOCIATED WITH BOTH REGIONAL AND TOTAL CORTICAL AMYLOID-BETA BURDEN IN COGNITIVELY NORMAL INDIVIDUALS AT RISK FOR ALZHEIMER'S DISEASE

Abstract

Introduction Models of preclinical Alzheimer's disease (AD) have proposed that cerebral amyloidosis and subsequent neurodegeneration may occur years, if not decades prior to observable cognitive decline. We previously demonstrated that total cortical amyloid-beta burden was inversely associated with cortical gray matter volume but not episodic memory in a sample of cognitively normal, middle-aged participants, with a first-degree family history and varying genetic risks of AD (Mecca et al., 2018). The current study utilizes an exploratory approach to determine the regional contribution of cortical amyloid deposition to gray matter volume and neuropsychological test performance in pre-symptomatic individuals at varying genetic risk for AD. Methods As described previously (Mecca et al., 2018), cognitively normal participants aged 50-66 with a first-degree family history for AD were genetically screened to select three groups: APOE genotype e4e4 (n=15), e3e4 (n=15), and e3e3 (n=15), matched for age and sex. Participants were then studied with [11C]PiB PET, MRI, and neuropsychological testing. PET and MR images were co-registered for application of a ROI template (AAL for SPM2) to generate regional time-activity curves with cerebellum as the reference region. MRI images were segmented into gray matter, white matter, and CSF. Mask images of these segments were smoothed to the system resolution (∼6mm), and gray matter values of BPND were partial volume corrected (PVC). PVC-BPND was then computed using SRTM2 for frontal, posterior cingulate, precuneus, lateral parietal, lateral temporal, medial temporal, occipital, and basal ganglia ROIs. A cortical BPND (cortical amyloid) was also calculated by taking a weighted average of only the frontal, posterior cingulate, precuneus, lateral parietal, and lateral temporal regions. To quantify GM within each ROI, we applied inverse transformations of the AAL from MNI to subject space and calculated GM fraction as the number of voxels segmenting a GM divided by the total number of voxels in a region. Results The study sample consisted of 24 females and 21 males with an average age of 59.12 (SD 4.72) and 16.33 (SD 1.86) years of education. Participants demonstrated an average MMSE score of 29.66 (SD 0.91), GDS score of 0.71 (SD 1.25), WAIS-III FSIQ score of 115.04 (SD 13.32), and WRAT-3 Reading score of 109.47 (SD 9.64). As previously reported, mean cortical BPND was 0.145 ± 0.131 for APOE-e3e3, 0.297 ± 0.299 for APOE-e3e4, and 0.389 ± 0.222 for APOE-e4e4 participants. An exploratory analysis was performed to examine the regional contributions to the association between cortical amyloid burden and gray matter volume. There were significant inverse associations between gray matter fraction and cortical amyloid in the frontal (spearman r=-0.459, p=0.002), lateral temporal (spearman r=-0.546, p Conclusions Based on our regional exploratory analyses, the inverse association between cortical amyloid burden and gray matter volume is strongest in the frontal, lateral, and medial temporal lobes. This is consistent with models of preclinical AD in which neurodegeneration occurs early in the disease process and before manifest cognitive decline. Interestingly, although the occipital lobeis traditionally thought to be a region of low amyloid deposition, we find the regional gray matter volume is significantly correlated with both regional and cortical amyloid. This perhaps implicates the occipital lobe as part of the cortical signature of AD, at least in cognitively normal middle-aged individuals with an increased risk of developing AD. It is also noted that medial temporal lobe atrophy isn't associated with regional amyloid, but instead significantly correlated with total cortical amyloid deposition. This is expected given the medial temporal lobe's reputation as a known area of early atrophy but not amyloid deposition in AD. This research was funded by This research was supported by the Alzheimer's Association IIRG-07-60026 (CHvD) and National Institute on Aging (P50-AG047270). APM was supported in part by the NIH/NIA K23-AG057794. RSO was supported in part by the NIMH R25 IMPORT grant (#R25MH071584).