Fig 1.
Analyzed regions of interest (ROIs) in the 18F-florbetepir PET scans.
Six individual cortical ROIs and a cerebellum ROI were analyzed for regional 18F-florbetapir binding and to calculate a composite ROI (see Methods). Representative trans-axial images from a Aβ negative (left) and positive (right) study demonstrate each ROI: (A) parietal cortex and precuneus; (B) frontal cortex; (C) anterior and posterior cingulate gyrus; (D) temporal cortex; and (E) cerebellum. Darker regions indicate higher 18F-florbetapir retention.
Fig 2.
Cognitive and motor characteristics of the study population.
A. Total DRS-2 score was significantly lower in PD-MCI participants (134.3, 131.7 to 136.9) as compared to PD-NC (139.3, 138.1 to 140.6) (t(57) = 4.12, p < 0.001). B. Total DRS-2 score was not significantly different among H&Y stages (F(5,53) = 0.54, p = 0.75). C. UPDRS part 3 score was significantly higher in PD-MCI (24.5, 19.9 to 29.2) as compared to PD-NC (17.1, 14.5 to 19.7) (t(59) = 3.07, p = 0.003). D. UPDRS part 3 score was significantly higher with advanced H&Y stage (1.0: 13.5, 0.2 to 26.8; 1.5: 14.8, 12.4 to 17.1; 2.0: 18.2, 15.5 to 20.9; 2.5: 17.3, 11.8 to 22.9; 3.0: 28.1, 22.3 to 33.9; 4.0: 48) (F(5,55) = 6.57, p < 0.001). There was no association between motor phenotype (TD vs. PIGD) and H&Y stage (data not shown). Data depicted are mean ± 95% C.I. Significance identified by one-way ANOVA and Tukey post-hoc test. *** p < 0.005, **** p < 0.001.
Table 1.
Demographic characteristics in 18F-florbetapir positive versus negative scans.
Table 2.
Cognitive diagnosis and test results in 18F-florbetapir positive versus negative scans.
Table 3.
Association between mild cognitive impairment and 18F-florbetapir PET imaging.
Table 4.
Composite and regional Aβ amyloid retention in participants with and without cognitive impairment.
Fig 3.
No correlation between composite 18F-florbetapir ROI SUVr and global cognition or memory performance.
(A) Total DRS-2 score did not significantly correlate with composite ROI SUVr values, either analyzed in aggregate (F(1,57) = 2.39, p = 0.13) or by cognitive diagnosis (for PD-NC, F(1,39) = 0.02, p = 0.90; for PD-MCI, F(1,16) = 2.15, p = 0.16). (B) DRS-2 memory subscore did not significantly correlate with composite ROI SUVr values, either analyzed in aggregate (F(1,57) = 3.80, p = 0.06) or by cognitive diagnosis (for PD-NC, F(1,39) = 0.00, p = 0.97; for PD-MCI, F(1,16) = 3.03, p = 0.10).
Fig 4.
Association between posterior cingulate gyrus 18F-florbetapir SUVr and memory performance.
Regional 18F-florbetapir SUVr in the posterior cingulate gyrus significantly correlated with memory performance on the DRS-2 (A, rho = -0.446, p < 0.0005), which remained significant after adjusting for age, sex, disease duration, education, and presence of an APOE ε4 allele (rho = -0.316, p = 0.02). SUVr in this region also significantly correlated with immediate free total recall on the Hopkins Verbal Learning Test (HVLT-R) on unadjusted analysis (B, rho = -0.263, p = 0.04) but not after adjustment (rho = -0.263, p = 0.054). Outlying participant indicated by closed diamond.