11C-PiB PET can underestimate brain amyloid-β burden when cotton wool plaques are numerous

Abstract

Abstract Subjects with familial Alzheimer’s disease due to PSEN1 mutations develop high cortical fibrillar amyloid-β load but often have lower cortical 11C-Pittsburgh compound-B retention than subjects with sporadic Alzheimer’s disease. We hypothesized this is influenced by limited interactions of Pittsburgh compound-B with cotton wool plaques, an amyloid-β plaque type common in familial Alzheimer’s disease but rare in sporadic Alzheimer’s disease. Histological sections of frontal and temporal cortex, caudate nucleus, and cerebellum were obtained from 14 cases with sporadic Alzheimer’s disease, 12 cases with familial Alzheimer’s disease due to PSEN1 mutations, two relatives of a PSEN1 mutation carrier but without genotype information, and three non-Alzheimer’s disease cases. Sections were processed immunohistochemically using amyloid-β-targeting antibodies and the fluorescent amyloid stains cyano-Pittsburgh compound-B and X-34. Plaque load was quantified by percent area analysis. Frozen homogenates from the same brain regions from five sporadic Alzheimer’s disease and three familial Alzheimer’s disease cases were analyzed for 3H-Pittsburgh compound-B in-vitro binding and concentrations of amyloid-β1-40 and amyloid-β1-42. Nine sporadic Alzheimer’s disease, three familial Alzheimer’s disease, and three non-Alzheimer’s disease subjects had 11C-Pittsburgh compound-B PET with standardized uptake value ratios calculated using the cerebellum as the reference region. Cotton wool plaques were present in the neocortex of all familial Alzheimer’s disease cases and one sporadic Alzheimer’s disease case, in the caudate nucleus from four familial Alzheimer’s disease cases, but not in the cerebellum. Cotton wool plaques immunolabeled robustly with 4G8 and amyloid-β42 antibodies but weakly with amyloid-β40 and amyloid-βN3pE antibodies and had only background cyano-Pittsburgh compound-B fluorescence despite labeling with X-34. Relative to amyloid-β plaque load, cyano-Pittsburgh compound-B plaque load was similar in sporadic Alzheimer’s disease while in familial Alzheimer’s disease it was lower in the neocortex and the caudate nucleus. In both regions, insoluble amyloid-β1-42 and amyloid-β1-40 concentrations were similar in familial Alzheimer’s disease and sporadic Alzheimer’s disease groups, while 3H-Pittsburgh compound-B binding was lower in the familial Alzheimer’s disease than the sporadic Alzheimer’s disease group. Higher amyloid-β1-42 concentration associated with higher 3H-Pittsburgh compound-B binding in sporadic Alzheimer’s disease but not familial Alzheimer’s disease. 11C-Pittsburgh compound-B retention correlated with region-matched postmortem amyloid-β plaque load, however, familial Alzheimer’s disease cases with abundant cotton wool plaques had lower 11C-Pittsburgh compound-B retention than sporadic Alzheimer’s disease cases with similar amyloid-β plaque loads. Pittsburgh compound-B has limited ability to detect amyloid-β aggregates in cotton wool plaques and may underestimate total amyloid-β plaque burden in brain regions with abundant cotton wool plaques.