Florbetapir PET Measurements of Amyloid Plaque Deposition Using a White Matter Reference Region are More Closely Correlated with Cross-Sectional and Longitudinal Measures of Alzheimer-Disease Related Glucose Hypometabolism

BackgroundA white matter reference region for in vivo PET-based amyloid imaging may offer advantages due to its larger anatomical size, reduced susceptibility to scatter correction error, and location at the center-of-field view (Lowe et al. 2018, Lopez-Gonzales et al. 2019). The purpose of this study was to compare the association between Florbetapir-measured mean cortical amyloid deposition and an Alzheimers disease-related glucose metabolism marker, when two different reference regions are used for Standardized Uptake Value Ratio (SUVR) calculations. We found that mean cortical Florbetapir Standardized Uptake Value Ratios using a white matter reference region (FBP mcSUVRwm) were more strongly correlated with the AD-related glucose measure, Hypometabolic Convergence Index (HCI, Chen, et al., 2011) than SUVRs calculated using a cerebellar reference region of interest (FBP mcSUVRcb). MethodsBaseline and 2.16{+/-}0.37 year follow-up FBP and fluorodeoxyglucose PET data from 1,238 mild AD dementia, mild cognitive impairment (MCI), and cognitively unimpaired (CU) participants from AD Neuroimaging Initiative (ADNI) were used to compare associations between cross-sectional and longitudinal FBP SUVRs and HCI measurements using a whole cerebellar and white matter reference region. ResultsCross-sectionally, partial correlations between mean cortical SUVR and HCI measurements were significantly stronger (Steigers Test p<1.0E-16) using a white matter reference region (r=0.59 [p=4.51E-107]) than using a cerebellar reference region (r=0.40 [p=9.30E-46]). Longitudinally, partial correlations between mean cortical SUVR and HCI changes were also significantly stronger (Steigers Test p=2.6E-11) using a white matter reference region (HCI/mcSUVRwm r=0.29 [p=1.26E-11]; HCI/mcSUVRcb (r=-0.04 [p=0.36]). Overall, post-hoc within group cross-sectional and longitudinal analysis were significantly stronger when a white matter reference region was used (Cross Sectional Analysis - CU: (Steigers Test p<0.01); MCI: (Steigers Test p=4.63E-11); AD: (Steigers Test p=7.95E-06); Longitudinal Analysis - MCI: (Steigers Test p=3.94E-04); AD: (Steigers Test p=0.02)). ConclusionsThis study further supports the use of FBP mcSUVRwm measurements in detecting and tracking AD-related amyloid deposition.