Predicting Alzheimer disease with β-amyloid imaging: results from the Australian imaging, biomarkers, and lifestyle study of ageing.

ObjectiveBiomarkers for Alzheimer disease (AD) can detect the disease pathology in asymptomatic subjects and individuals with mild cognitive impairment (MCI), but their cognitive prognosis remains uncertain. We aimed to determine the prognostic value of -amyloid imaging, alone and in combination with memory performance, hippocampal atrophy, and apolipoprotein E epsilon 4 status in nondemented, older individuals. MethodsA total of 183 healthy individuals (age = 72.0 7.26 years) and 87 participants with MCI (age = 73.7 8.27) in the Australian Imaging, Biomarkers, and Lifestyle study of ageing were studied. Clinical reclassification was performed after 3 years, blind to biomarker findings. -Amyloid imaging was considered positive if the C-11-Pittsburgh compound B cortical to reference ratio was 1.5. ResultsThirteen percent of healthy persons progressed (15 to MCI, 8 to dementia), and 59% of the MCI cohort progressed to probable AD. Multivariate analysis showed -amyloid imaging as the single variable most strongly associated with progression. Of combinations, subtle memory impairment (Z score = -0.5 to -1.5) with a positive amyloid scan was most strongly associated with progression in healthy individuals (odds ratio [OR] = 16, 95% confidence interval [CI] = 3.7-68; positive predictive value [PPV] = 50%, 95% CI = 19-81; negative predictive value [NPV] = 94%, 95% CI = 88-98). Almost all amnestic MCI subjects (Z score -1.5) with a positive amyloid scan developed AD (OR = ; PPV = 86%, 95% CI = 72-95; NPV = 100%, 95% CI = 80-100). Hippocampal atrophy and epsilon 4 status did not add further predictive value. InterpretationSubtle memory impairment with a positive -amyloid scan identifies healthy individuals at high risk for MCI or AD. Clearly amnestic patients with a positive amyloid scan have prodromal AD and a poor prognosis for dementia within 3 years. Ann Neurol 2013;74:905-913