Background: PIB PET imaging was added to the Alzheimer's Disease Neuroimaging Initiative (ADNI) study in 2007 to provide a publically accessible, imaging biomarker data base for use in clinical trials of anti-amyloid therapies as well as for other academic studies. Data on the ADNI site allow correlation of PIB status with other biomarkers such as CSF Abeta42, t-tau, and p-tau measured in the same subjects. Methods: As of February 2009, 103 baseline PIB PET studies from 12 ADNI sites had been collected and posted on the UCLA Laboratory of NeuroImaging (LONI) website after completion of requisite quality control and pre-processing steps by the University of Michigan, which included summing over 50-70 min post-injection, co-registration to Talaraich space, re-orientation along AC-PC axis, standardization to a uniform voxel size and resolution, and normalization to cerebellar grey matter. The ADNI LONI database included PIB PET studies in 19 controls (MMSE range 24 - 30, mean 29 ± 1; Age: 78 ± 5 yrs), 65 mild cognitive impairment (MCI) (MMSE range 20 - 30, 27 ± 2; Age: 75 ± 8 yrs), and 19 Alzheimer's disease (AD) (MMSE range 15 - 26, 22 ± 3; Age: 73 ± 9 yrs) subjects. One year longitudinal PIB PET studies have been completed in 67 subjects (15 controls, 42 MCI, and 10 AD). CSF Abeta42 and tau measures were collected in 11 controls, 34 MCI, and 10 AD subjects. Results: Positive amyloid deposition (PIB+) was defined when the average SUVR (standardized uptake value ratio in region relative to the cerebellum) of four regions (frontal, anterior cingulate, precuneus, parietal cortex) exceeded 1.50. With this cut-off value, a large percentage of the elderly control (47%) and MCI (71%) subjects were PIB+, while 2 clinically diagnosed AD subjects were clearly PIB-. Dichotomous categorization of all subject groups showed substantial agreement (90%) between PIB PET and CSFAbeta42 measures and modest agreement (75%) between PIB PET and p-tau. Conclusions: Multi-site PIB PET data have been successfully collected and analyzed. Comparison of PIB status with CSF measures in ADNI subjects provided substantial agreement with Abeta42 and lesser agreement with p-tau and t-tau.
FULL TEXT