IC‐P‐036: SPATIAL CORRELATION BETWEEN GRAY MATTER CHANGES IN CHILDHOOD, FDG METABOLISM IN YOUNG ADULTS, AND BETA‐AMYLOID DEPOSITION IN THE ELDERLY

been explored by previous studies. Regarding quantification of brain amyloid burden on 11C-PIB brain PET images, we tried to find out newmethods which can represent better than conventional method. In this study, we made correlation between CSF biomarkers (Ab1-42, total tau (t-tau), tau phosphorylated at tyrosine 181 (p-tau), t-tau/ Ab1-42, p-tau/ Ab1-42) and brain PIB burden using various quantitative methods in samples from patients with clinically well characterized AD and healthy normal controls. Methods: We prospectively enrolled 18 normal controls (NCs) and 16 patients with ADwhowere clinically diagnosed according to NIA-AA revised diagnostic criteria. All subjects ( n1⁄434) underwent history taking, neuropsychological test, lumbar puncture for CSF study, brain MRI and PIB PET. CSF sample collection and processing. The CSF was collected via standard sterile techniqueWithin 1 hour after collection, samples were centrifuged at low speed and aliquoted (500mL) into polypropylene tubes prior to freezing at -80 C. The samples were analyzed for Ab1-42, t-tau and p-tau by Luminex xMAP bead-based method (INNO-BIA AlzBio3) INNO-BIA reagents included monoclonal capture/detection antibodies 4D7A3/3D6 for Ab142, AT120/HT7 for t-tau, and AT270/HT7 for p-tau. PIB PET image processing Static PET acquisition for 20min was performed 40min after intravenous bolus injection of 11C-PIB (17.863.3 mCi) in each subject. Details of image processing is described in poster No . Ratios of SUVmean, region to cerebellum (SUVRs) or product of SUVR and volume of voxels (PSV) with uptake above SUVmean of cerebellum (1⁄4 cerebellar threshold) in CCVs were calculated in five different methods (Method 1) Mean SUVR without threshold (conventional method) (Method 2) Mean SUVR above cerebellar threshold (SUVRcbll) (Method 3) Total volume (TV) of voxels above cerebellar threshold (TVcbll) (Method 4) Sum of PSVs above cerebellar threshold (PSVcbll) (Method 5) Volume-weighted SUVR (vwSUVR). Correlations between CSF biomarkers such as Ab142, t-tau, p-tau, t-tau/ Ab1-42, p-tau/ Ab1-42 and PET-originated parameters such as SUVR, TV , PSV were examined. Mann-Whitney test and ROC analysis were used to compare the diagnostic performances of five different methods. P<0.05 was considered significant. Results: Comparing AD vs NC, it showed significant difference in MMSE, CDR(SOB), level of CSF biomarkers and PIB burden using 5methods. (Table1) In ROC analysis, Ab1-42 among CSF biomarkers and SUVRcbll (or vwSUVR) among PIB quantitative methods showed best diagnostic performance for discriminating AD from NC (Table2) All of the PIB quantitative method revealed significant correlation with each of the CSF anayltes. (Figure1) SUVRcbll showed highest correlation with Ab1-42, p-tau, p-tau/ Ab1-42 and vwSUVR with t-tau, t-tau/ Ab1-42. Conclusions: Both CSF biomarkers and 11C-PIB brain PET images are useful for diagnosis of AD Among CSF biomarkers, Ab1-42 outperformed any other CSF analytes. Among quantitative methods for PIB burden, mean SUVR above cerebellar threshold and volume-weighted SUVR showed better diagnostic performance for AD and better correlation with each of the CSF analytes than the conventional method. Table 1 Baseline characteristics of two groups.