Effect of cerebral blood flow changes on F-18-florbetaben SUVR

1180 Introduction: Quantitation of amyloid-PET data is critical to monitor amyloid-beta plaque changes both in longitudinal observational studies and in interventional trials investigating amyloid-beta-modifying treatments for Alzheimer’s disease (AD). Standardized uptake value ratios (SUVRs) are widely used to assess the amyloid-beta plaque load with PET, however, may be biased by cerebral blood flow (CBF) changes. These effects may be relevant in interventional trials where an amyloid-beta-modifying drug is administered and drug effects on CBF are unknown. Objectives: The objective of this study was to assess the impact of CBF changes on 18F-florbetaben SUVR. Methods: Dynamic 18F-florbetaben PET scans were acquired in 10 patients with mild to moderate AD dementia based on clinical diagnosis (69±7 y) and 10 age-matched HCs (67±8 y) (Becker et al. JNM 2013). Arterial samples were collected after tracer injection and corrected for metabolites. Arterial input functions and rate constants (K1, k2, k3 and k4) from each subject were used to generate simulated time-activity curves (TACs). Three type of simulations considering CBF changes up to 20% were conducted: 1) regional CBF change affecting only cortical regions; 2) regional CBF change affecting only the reference region (RR) and 3) global CBF change affecting both cortical and RR. The simulated TACs were used to derive SUVR at different time points using three RR (CGM, pons, SWM). Results: At pseudo-equilibrium (90-110 min), a 20% regional CBF reduction affecting the RR caused a decrease of the composite SUVR of -0.7±0.9% when using the CGM as RR, and an increase in the composite SUVR of 1.3±1.2% and 9.4±1.7% when using the pons and SWM as RR, respectively. When the 20% CBF reduction affected only the target region, the composite SUVR showed a decrease of -1.8±1.7%, irrespective of the RR used. A 20% reduction of the CBF affecting both target and RR caused a decrease of the composite SUVR using CGM (-2.5±1.3%) and pons (-0.5±1.8%) an increase when using SWM (7.4±2.4%). Conclusions: Regional and global CBF changes can bias the amyloid load estimates using a florbetaben PET SUVR approach. The SUVR using CGM as RR had lowest influence of CBF changes making this approach sufficient for most clinical research applications. However, for longitudinal studies in which a large CBF change might occur and maximum quantification accuracy is desired, dynamic acquisition and non-invasive kinetic analysis is recommended. Dynamic data acquisition time can be shortened using a dual-time window approach as shown recently (Bullich et al. JNM 2018).