👤 Dan Mungas

Machine learning enables scalable quantification of neuropathology, offering deeper phenotyping of Alzheimer's disease (AD). In this validation study, we quantified amyloid-beta (Aβ) deposits, evaluating multiple brain regions across institutions, and evaluated associations with clinical, demographic, and genetic factors in persons pathologically diagnosed with AD. All linear models were adjusted for sex, age of death, ethnicity, and center. We analyzed densities (#/mm2) of cored plaques, diffuse plaques, and cerebral amyloid angiopathy (CAA) in 273 individuals from 3 Alzheimer's Disease Research Centers. Formalin-fixed paraffin-embedded sections of frontal, temporal, and parietal cortices were immunostained and digitized, generating 799 whole-slide images (WSIs). Following log transformation, mixed-effects modeling revealed the parietal cortex had the highest cored plaque densities (P < .001); the temporal cortex had the highest diffuse plaque (P < .001); CAA showed no regional differences. Wilcoxon rank-sum test, and covariates adjusted linear models showed ApoE ε4- status was associated with higher cored plaque densities in the temporal lobe (P = .04). ApoE ε4+ status was associated with diffuse plaques in the temporal lobe (P = .001), and CAA in the frontal lobe (P = .004). These findings provide further validation and provide exploratory associations advancing deeper phenotyping of AD. Show less

People with mild cognitive impairment (MCI) are candidates for early intervention, but not all progress to Alzheimer's disease (AD) dementia. Identifying a subgroup at highest risk may improve treatment targeting. We analyzed data from participants with MCI enrolled in the Alzheimer's Disease Neuroimaging Initiative (ADNI). Cognitive domains included memory, executive functioning, language, and visuospatial abilities. We evaluated baseline performance and 6-month change scores, using proportional hazards models to estimate associations with time to conversion to AD dementia. The strength of association varied by domain, but in general both baseline performance and 6-month change were associated with conversion. The strongest effects observed for memory and language. Observed associations were largely independent of established risk biomarkers, including APOE genotype, structural MRI measures, and CSF biomarkers. 6-month change scores on cognitive tests may help identify a high-risk subgroup of persons with MCI likely to progress to AD dementia. Systematic review. The authors reviewed the literature using traditional (e.g. PubMed) sources. There is a modest literature on change scores in the context of the AD clinical spectrum, but few investigations have evaluated whether short-term changes may be able to identify a high-risk subgroup of people with MCI. The authors have published a systematic review of this literature (Jutten et al. 2020) and appropriately refer to relevant citations here.Interpretation: Our findings suggest that short-term changes in cognition may be useful as part of a strategy to identify subsets of people with MCI who are at highest risk of conversion. Findings were clearest for memory and language. Domain-specific changes appeared to be independent from other biomarkers used to identify people at highest risk. Domain-specific changes did not appear to be better than changes in global cognition as measured by the MMSE or the CDR-sum of boxes.Future directions: Short-term changes in cognition may be useful to help identify a subgroup of people with MCI at highest risk of conversion to AD dementia. Future work could consider time frames shorter than the 6-month data we had available, better characterizing changes with more than 2 time points, or developing strategies that combine changes in cognition with other biomarkers to identify a subgroup of people with MCI to target for treatment. Show less