Clinicopathologic and11C-Pittsburgh compound B implications of Thal amyloid phase across the Alzheimer’s disease spectrum

Abstract

Thal amyloid phase, which describes the pattern of progressive amyloid-β plaque deposition in Alzheimer’s disease, was incorporated into the latest National Institute of Ageing – Alzheimer’s Association neuropathologic assessment guidelines. Amyloid biomarkers (positron emission tomography and cerebrospinal fluid) were included in clinical diagnostic guidelines for Alzheimer’s disease dementia published by the National Institute of Ageing – Alzheimer’s Association and the International Work group. Our first goal was to evaluate the correspondence of Thal amyloid phase to Braak tangle stage and ante-mortem clinical characteristics in a large autopsy cohort. Second, we examined the relevance of Thal amyloid phase in a prospectively-followed autopsied cohort who underwent ante-mortem ¹¹C-Pittsburgh compound B imaging; using the large autopsy cohort to broaden our perspective of ¹¹C-Pittsburgh compound B results. The Mayo Clinic Jacksonville Brain Bank case series (n = 3618) was selected regardless of ante-mortem clinical diagnosis and neuropathologic co-morbidities, and all assigned Thal amyloid phase and Braak tangle stage using thioflavin-S fluorescent microscopy. ¹¹C-Pittsburgh compound B studies from Mayo Clinic Rochester were available for 35 participants scanned within 2 years of death. Cortical ¹¹C-Pittsburgh compound B values were calculated as a standard uptake value ratio normalized to cerebellum grey/white matter. In the high likelihood Alzheimer’s disease brain bank cohort (n = 1375), cases with lower Thal amyloid phases were older at death, had a lower Braak tangle stage, and were less frequently APOE-ε4 positive. Regression modelling in these Alzheimer’s disease cases, showed that Braak tangle stage, but not Thal amyloid phase predicted age at onset, disease duration, and final Mini-Mental State Examination score. In contrast, Thal amyloid phase, but not Braak tangle stage or cerebral amyloid angiopathy predicted ¹¹C-Pittsburgh compound B standard uptake value ratio. In the 35 cases with ante-mortem amyloid imaging, a transition between Thal amyloid phases 1 to 2 seemed to correspond to ¹¹C-Pittsburgh compound B standard uptake value ratio of 1.4, which when using our pipeline is the cut-off point for detection of clear amyloid-positivity regardless of clinical diagnosis. Alzheimer’s disease cases who were older and were APOE-ε4 negative tended to have lower amyloid phases. Although Thal amyloid phase predicted clinical characteristics of Alzheimer’s disease patients, the pre-mortem clinical status was driven by Braak tangle stage. Thal amyloid phase correlated best with ¹¹C-Pittsburgh compound B values, but not Braak tangle stage or cerebral amyloid angiopathy. The ¹¹C-Pittsburgh compound B cut-off point value of 1.4 was approximately equivalent to a Thal amyloid phase of 1–2.