Abstract
1253
Objectives Cerebral amyloid load and patterns of hypometabolism/-perfusion are complementary information in differential diagnosis of dementia. Due to a high extraction fraction, early distribution of the amyloid ligand C-11-PIB reflects tissue perfusion with good signal-to-noise ratio. Our aim was to exploratively evaluate the diagnostic potential of early PIB distribution by intraindividual comparison with F-18-FDG uptake.
Methods 4 pts. with Alzheimer's disease (AD), 8 pts. with dementia with lewy bodies (DLB) and 12 controls underwent PIB- and FDG-PET (max. 800MBq PIB and 370MBq FDG). Tracer concentration 0 to 4min p.i. (PIB) and 40 to 52min p.i. (FDG) were coregistered, spatially normalized (SPM2), smoothed (12mm) and scaled using a wholebrain-ROI. For each pat., two z-images (FDG and PIB) were calculated from voxelwise comparison with the control group. Each pair of z-images was rated visually. Furthermore, we calculated for each pat. the percentage volume with which clusters of reduced tracer concentration overlapped (100% = cluster size at FDG-z>2.5).
Results The hypometabolic volume (FDG:z>2.5) ranged from 60 to 225ml per pat.. 56% (median across pts.) of the respective voxels also showed PIB-reductions (z>2). The extent of PIB reductions (z>2.5) outside the FDG-cluster (z>2) was only 10% (median). The largest mismatch was found in a DLB pat. with moderate temporooccipital hypometabolism not evident in PIB (overlap: 32%). In patients with clear-cut hypometabolism, z-images from PIB- and FDG-PET were rated equivalent.
Conclusions In AD and DLB with clear-cut hypometabolism, early PIB distribution seems to be of equal diagnostic value as FDG-uptake and may support clinical diagnosis if an FDG scan is not available. Generally, findings in PIB-perfusion tended to be of lower significance than in FDG-PET, however, there might be room for further optimizing PIB-quantification for diagnostic purposes (reference tissue? modelling?)