Clinicopathological correlates in atypical Alzheimer's disease: evaluating anatomical distributions of neurofibrillary tangles and neuropsychological profiles
MetadataShow full item record
This study aims to discover whether there is a correlation between atypical clinical presentations of Alzheimer’s disease (AD) and atypical distribution patterns of AD pathology. To provide a measure of the atypical clinical presentations, we obtained standardized neuropsychological test scores for a group of 345 subjects of the Boston University Alzheimer’s Disease Center cohort that had received a clinical or pathological diagnosis of AD. Each of the neuropsychological test scores included in our analyses was classified into one of five cognitive domains, according to the primary domain each test assesses: memory, executive function, attention, visuospatial function, and language. From these test scores, global cognitive performance scores and individual domain performance scores were calculated for a subset of 53 subjects that had brain tissue slides available for pathological analysis. Difference scores were computed for each domain, providing a within-subject comparison of performance between each individual cognitive domain and overall cognitive performance. For these same 53 subjects, tissue slides from six brain regions were obtained and digitally scanned. Neurofibrillary tangle (NFT) quantification was performed for all tissue slides using a computer algorithm modified to recognize AT8 staining patterns. NFT densities were then calculated for five general brain regions (frontal, parietal, temporal, limbic and occipital). In addition, a global NFT density score was computed for each subject, averaging NFT densities across all regions. From these densities, difference scores were calculated for each brain region individually, providing a measure of how each region’s NFT density compares to the overall brain NFT density. Multiple linear regressions analyses were performed with five pairs of cognitive domain difference scores and region NFT density difference scores: memory difference scores and limbic difference scores, executive function difference scores and frontal difference scores, attention difference scores and parietal difference scores, visuospatial difference scores and occipital difference scores, and language difference scores and occipital difference scores. Though we expected to observe significant negative correlations between each of the five difference score pairs, the only statistically significant correlation observed was between memory difference scores and limbic difference scores (β= -0.361, p<0.05). These results suggest that poorer performance in memory-related neuropsychological tests, when compared to global cognitive performance, can predict higher NFT densities in limbic regions when compared to the overall brain pathology. Although no other difference score pairs showed any statistically significant correlations, many study limitations, including small sample size and simplifications in analysis, should be addressed in the future to provide better understanding of these atypical presentations of AD and their underlying pathologies.