Cerebral Amyloid Angiopathy Pathology and Its Association With Amyloid-β PET Signal

Abstract

Background and objectives: To determine the contribution of cerebral amyloid angiopathy (CAA) to Pittsburgh compound B (PiB)-PET tracer retention.

Methods: Participants from the Mayo Clinic Study of Aging and Mayo Clinic Alzheimer's Disease Research Center with antemortem PiB-PET imaging for β-amyloid (Aβ) who later underwent autopsy were included in this study. Pathologic regional leptomeningeal, parenchymal, capillary CAA, and Aβ plaque burden were calculated from one hemisphere. Regional lobar amyloid standardized uptake value ratio (SUVR) on PET was calculated from the same hemisphere sampled at autopsy. Single- and multiple-predictor linear regression models were used to evaluate the relative contributions of pathologically determined regional CAA and Aβ plaques to antemortem PiB-PET SUVR.

Results: Forty-one participants (30 male, 11 female) with a mean (SD) age at death of 75.7 (10.6) years were included. Twenty-seven (66%) had high PiB signal with a mean (SD) of 2.3 (1.2) years from time of PET scan to death; 24 (59%) had a pathologic diagnosis of Alzheimer disease. On multivariate analysis, CAA was not associated with PiB-PET SUVR, while plaques remained associated with PiB-PET SUVR in all regions (all p < 0.05). In patients without frequent amyloid plaques, CAA was not associated with PiB-PET in any region.

Discussion: We did not find evidence that pathologically confirmed regional CAA burden contributes significantly to proximal antemortem regional PiB-PET signal, suggesting that amyloid PET imaging for measurement of cortical amyloid burden is unconfounded by CAA on a lobar level. Whether CAA burden contributes to PiB-PET signal in patients with severe CAA phenotypes, such as lobar hemorrhage, requires further investigation.

Figure 1. Distribution of Severity of Pathologic Cerebral Amyloid Angiopathy (CAA), Neuritic Plaques, and β-Amyloid (Aβ) Pittsburgh Compound B (PiB)–PET Standardized Uptake Value Ratio (SUVR) by Brain Region

(A) Distribution of parenchymal + leptomeningeal CAA severity by brain region. A score of 0 indicates no CAA and 6 indicates severe combined parenchymal and leptomeningeal CAA. (B) Distribution of neuritic plaques (Consortium to Establish a Registry for Alzheimer's Disease [CERAD] score) with 0 indicating no neuritic plaques and 3 indicating frequent neuritic plaque. (C) Distribution of Aβ PiB-PET SUVR by brain region.

Figure 2. β-Amyloid (Aβ)–PET Imaging and Gradient Recalled Echo (GRE) Sequences in a 76-Year-Old Man With Moderate Global Cerebral Amyloid Angiopathy (CAA) and No Significant Aβ Plaque Burden

(A) Aβ-PET and (B) MRI GRE sequences of a 76-year-old man at time of death without significant amyloid β plaque burden (Thal phase 1, Braak stage 1, Consortium to Establish a Registry for Alzheimer's Disease 0) and at least moderate global CAA (frontal: severe leptomeningeal, moderate parenchymal, present capillary; temporal: severe leptomeningeal, mild parenchymal, absent capillary; parietal: severe leptomeningeal, moderate parenchymal, present capillary; occipital: mild leptomeningeal, moderate parenchymal, present capillary; hippocampal: mild leptomeningeal, mild parenchymal, present capillary). Pathologic diagnosis was frontotemporal dementia secondary to TDP-43 associated with progranulin mutation. This patient had minimal microbleed burden and was amyloid negative with a global cortical standardized uptake ratio (SUVR) of 1.384.