Premium
P4‐352: Imaging brain amyloid in non‐demented young adults with Down syndrome using PiB‐PET
Author(s) -
Cohen Ann,
Christian Bradley,
Handen Ben,
Johnson Sterling,
Hartley Sigan,
Seltzer Marsha,
Murali Dhanabalan,
Bulova Peter,
Tumuluru Rameshwari,
Bergnic Michael,
Matan Cristy,
Price Julie
Publication year - 2011
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2011.09.098
Subject(s) - pittsburgh compound b , psychology , down syndrome , trisomy , asymptomatic , medicine , amyloid (mycology) , dementia , pathology , disease , psychiatry , biology , genetics
heimer’s disease (AD) (MCI-c) from those that remain stable (MCI-s). Longitudinal percentage volume changes (PVC) in hippocampal, parahippocampal and entorhinal cortex (ERC) and left/right asymmetry were compared to standard cross-sectional measures. Methods: Data on 58 MCI-c and 239 MCI-s subjects was acquired from the public ADNI consortium database. Multiple 1.5T scanners were used to obtain MP-RAGE scans at baseline, six-months and one-year. Hippocampal volume and parahippocampal/ERC thickness were measured using FreeSurfer and used to calculate PVC and asymmetry measuresbetweenbaseline&six-months, six-months&one-year and baseline & one-year. A mixed-model logistic regression analyses using a generalised linear approachwith randomeffects for centre and controlling for age and genderwas used to examine the extent towhich normalised volumes could predict group membership. Akaike information criterion (AIC) and Bayesian information criterion (BIC) were used to identify the best models which were then reported in terms of sensitivity, specificity and balanced accuracy. Results: Cross-sectional volumes at one-year were found to be the best model (Sensitivity 1⁄4 71%, Specificity 1⁄4 69%, Accuracy 1⁄4 70%) with both the left hippocampus (p<0.001) and right ERC (p1⁄4 0.05) significantly predicting conversion. The best longitudinal models from each group were PVCs between baseline and one-year (Sensitivity 1⁄4 64%, Specificity 1⁄4 67%, Accuracy 1⁄4 66%) with the left ERC as the most accurate predictor (p 1⁄4 0.001), and asymmetry changes between six-months and one-year (Sensitivity1⁄4 61%, Specificity1⁄4 64%, Accuracy1⁄4 62%). Combining cross-sectional volumes at baseline with PVC measures from baseline to six-months improved the ability to predict conversion (Sensitivity 1⁄4 70%, Specificity 1⁄4 72%, Accuracy 1⁄4 71%). Conclusions: Longitudinal PVC and asymmetry measures were not as accurate at predicting group membership as cross-sectional volumes. However, the ability to predict conversion was improved by combining longitudinal measures with cross-sectional volumes. Consistent with previous findings, we confirm that the left hippocampus and right ERC are the most accurate cross-sectional predictors of conversion, but PVCs of the left ERC are amore important longitudinalmarker of conversion.