Age‐at‐Onset and APOE Related Heterogeneity in the Clinical, Plasma Biomarker, and Neuropathologic Presentation of Alzheimer’s Disease (AD)
Denis S. Smirnov,David P. Salmon,Douglas R. Galasko,Vanessa OF Goodwill,Lawrence A Hansen,Steven D. Edland,Annie OF Hiniker
DOI: https://doi.org/10.1002/alz.061763
2022-12-01
Abstract:Patients with sporadic early‐onset AD (EOAD) show less pronounced memory impairment, faster cognitive decline, and greater likelihood of focal cortical syndromes than those with later onset (LOAD), while often lacking the APOE ε4 allele classically associated with younger onset. This heterogeneity could arise from differences in the distribution of AD pathology or presence of concomitant pathologies. We examined clinical and cognitive heterogeneity in relation to age‐at‐onset, APOE genotype, and concomitant neuropathology (Lewy body, TDP‐43, and vascular) in patients with autopsy‐confirmed severe AD from the National Alzheimer’s Coordinating Center database (N=1750). In patients from the UCSD Alzheimer’s Disease Research Center (N=121) we examined how age‐related heterogeneity relates to differences in distribution of neurofibrillary tangle (NFT) pathology and plasma biomarkers of AD. Age‐at‐onset distribution in APOE ε4‐ patients was bimodal with best separation at age 63. Patients with EOAD (age ≤63) had faster cognitive and functional decline than LOAD regardless of APOE genotype; and were more likely to present with nonmemory cognitive complaints (OR=5.56, 99%CI: 3.32‐9.39), show greater executive dysfunction (β=2.59 SD, 99%CI: 2.10‐3.09), and receive a non‐AD clinical diagnosis (OR=1.54, 99%CI: 1.19‐1.97), particularly for ε4‐ EOAD (p’s<0.01, APOE × age‐at‐onset interactions). Despite this atypicality, groups did not differ in their high rates of AD biomarker positivity, and ε4‐ EOAD patients were most likely to have pure AD without concomitant pathology. TDP‐43 (OR=0.37; 99%CI: 0.18–0.72) and microvascular (OR=0.65; 99%CI: 0.47–0.88) pathology were less common in EOAD, while Lewy pathology was more common in ε4+ (OR=1.48; 99% CI, 1.02–2.17). Midfrontal NFT density was higher in EOAD than LOAD (β= 0.87 SD, 99%CI: 0.44‐1.29). Greater hippocampal NFT density was associated with APOE ε4+ (β=0.49 SD, 99%CI: 0.02‐0.96). Mediation analyses suggest that differences in midfrontal/hippocampal NFT ratio, rather than concomitant neuropathology, mediate the atypical cognitive profiles and decline of EOAD (p<0.01). A relatively more neocortical NFT distribution, rather than concomitant non‐AD pathology, accounts for the atypical, non‐memory dominant clinical presentation and rapid progression in EOAD. This suggests that selective cortical vulnerability may explain the earlier‐than‐expected onset and atypical profile in these individuals without the APOE ε4 risk allele.
clinical neurology
What problem does this paper attempt to address?