Brain amyloid PET interpretation approaches: from visual assessment in the clinic to quantitative pharmacokinetic modeling

Abstract

Cerebral β-amyloid burden can be measured by positron emission tomography (PET) tracers, including 11C-Pittsburgh Compound B (11C-PiB) and three fluorinated tracers that have been approved by regulatory agencies. This review focuses on the current clinical use of amyloid PET and the added utility of quantifying β-amyloid deposition. Using systematic review methods, a PubMed search identified studies on the added value of standardized uptake value ratio (SUVR) to visual reads. PubMed search was also performed to identify studies that illustrate important factors that cause variability in SUVR results, as well as key papers in the initial development of 11C-PiB. Among studies comparing visual interpretation and SUVR, there is high concordance between visual reads and SUVR for diagnosis classification, with evidence that SUVR decreases inter-reader variability in visual interpretation and improves accuracy. The variability among results of studies utilizing SUVR is likely multifactorial, including choice of reference regions, choice of target regions, and methods for determining SUVR thresholds. The initial 11C-PiB studies illustrate the importance of rigorous pharmacokinetic modeling and stepwise simpler quantitative methods when translating radiotracers. There is increasing evidence of the added value of SUVR in the clinic, in particular in borderline cases or for detection of subtle changes. SUVR provides a reasonably accurate measurement of amyloid plaque deposition compared to pharmacokinetic modeling methods.