n.162 – PROGNOSTIC VALUE OF 18F-FLORBETAPIR SCAN: A 36-MONTH FOLLOW UP ANALYSIS USING ADNI DATA

BACKGROUND-AIM
The Alzheimer’s Disease Neuroimaging Initiative(ADNI) provides a unique opportunity to investigate the relationship between -Amyloid neuropathology and patients’ long-term cognitive function change. We examined baseline 18F-florbetapir PET amyloid imaging status and 36-months change from baseline in cognitive performance in subjects with mild cognitive impairment(MCI).
METHODS
All ADNI subjects who underwent PET-imaging with 18F-florbetapir and had a clinical diagnosis of MCI at the visit closest to florbetapir imaging, were included. -Amyloid deposition was measured by florbetapir standard uptake value ratio(SUVR), and dichotomized as A+(SUVR>1.1) or A–(SUVR≤1.1). The change of cognitive scores including ADAS11, MMSE and CDR sum of boxes(CDR-SB) were evaluated every 6 months. Mixed-effect Model Repeated Measures(MMRM) was applied to detect the difference between A+ and A– subjects’ cognitive score change from baseline, adjusting for baseline age and cognitive scores. Clinically significant cognitive change(4 point decline on the ADAS 11) was also evaluated using a multivariate-logistic-regression-model with general estimating equation(GEE) to account for within-subjects correlation. Marginal Odds Ratio was used to evaluate the risk difference for a clinically significant cognitive change among A+ participants vs. A– participants.
RESULTS
Of 478 MCI-subjects who had at least one florbetapir scan, 153 had a cognitive evaluation at 36-month follow up. Of those, 79 were A– and 74 A+. At 36-month visit, the A+ vs. A– group score changed from baseline(LS means 4.03 vs. 0.26 for ADAS11;-2.61 vs.-0.40 for MMSE;1.53 vs. -0.11 for CDR-SB [p< 0.0001 all comparisons]). GEE analysis on clinically significant cognitive change showed a marginal Odds Ratio=2.18(95% CI:1.47–3.21) for A+ vs. A– groups.
CONCLUSION
MCI subjects with higher -Amyloid deposition, had greater deterioration in cognitive function over the 36-month follow up, while subjects with no  Amyloid accumulation tended to be stable on these measurements. This finding is consistent with previously published studies.

Scarica PDF