Alzheimer’s disease (AD) is a form of dementia characterized by a long preclinical phase during which amyloid (Aβ) and tau proteins aggregate into plaques and tangles, respectively. The prevailing hypothesis holds that Aβ aggregation sets off a series of events, including tau aggregation and neurodegeneration, which ultimately leads to cognitive decline. Study of Aβ and tau aggregation has been facilitated in recent years by advances in positron emission tomography (PET), which allow serial imaging to understand what early factors affect clinical outcomes and progression to AD. Through this work, we know that older adults with high Aβ have greater levels of neurodegeneration and are at higher risk for developing AD. However, Aβ may be at its most toxic before it aggregates into insoluble plaques. This may mean that by the time global thresholds for Aβ positivity are reached, soluble forms of Aβ may have already begun the AD pathological cascade. As such, it is important to identify early levels of abnormal Aβ before cognitive decline and, to the extent that it is possible, before Aβ pathology is high everywhere.
The first project focuses on identifying early regions of Aβ deposition in cognitively normal older adults through PET imaging and creating a time course of regional Aβ burden. We found that parietal and frontal regions exhibited slightly earlier Aβ pathology than other studied regions, but that they all reached their peak accumulation rates within an eight-year span. Aβ accumulation in all brain regions studied was also similarly associated with apolipoprotein ε4 allele, a genetic risk factor for sporadic AD, and tau pathology. These results indicate that spread from one region to another is unlikely. Neurodegeneration is widely thought to occur later in AD than Aβ and tau aggregation and can be studied with both PET and MRI measures. These measures do not always concur, so it can be useful to have both measures available. However, the main PET measure of neurodegeneration, [18F] Fluorodeoxyglucose (FDG), which measures glucose metabolism, has been discontinued in many research studies because it adds an additional scan and radiation exposure. Thus, the second project investigates the use of relative delivery (R1), a proxy of cerebral blood flow derived from dynamic PET scans (which are often collected in research protocols), as an alternative measure to FDG. We found that R1 and FDG were highly correlated with one another and with MRI-based neurodegeneration measures. They were also similarly associated with cognition in AD patients. This confirms that R1 can be used instead of FDG in situations where the latter is unavailable and may be particularly useful when dynamic PET scans are already being acquired.
Although cerebral blood flow is a proxy for neurodegeneration, it also reflects vascular function. A major question is whether vascular dysfunction leads to Aβ and tau accumulation, or whether changes in cerebral blood flow simply reflect neurodegeneration. Longitudinal investigations of Aβ, tau, and cerebral blood flow can help answer this question. The third project investigates change in R1 related to baseline Aβ/tau, and vice versa. We found that baseline Aβ predicted change in R1, but the opposite was not true, indicating that Aβ accumulation is upstream of cerebral blood flow and therefore it is less likely that blood flow changes drive amyloid deposition. Tau and R1 were correlated cross-sectionally and showed weak bidirectional longitudinal relationships that will require further study to untangle.
Taken together, these results further our understanding of the temporal dynamics of AD pathology and lend support to the Aβ cascade hypothesis. Our results indicate that the identification of a stable “early Aβ” region across studies with Aβ-PET is unlikely, and that cerebral blood flow measured with R1 is a proxy for neurodegeneration rather than reflecting early vascular pathology that incites AD.
