Non-amnestic phenotypes of alzheimer’s disease, early age of onset and apoe genotype are associated with tau, not aβ-pet

not available. FEATURED RESEARCH SESSIONS F1-01 AMYLOID, TAU, AND STRUCTURAL IMAGING IN ATYPICAL AD AND NON-AD DEMENTIAS F1-01-01 NEUROPATHOLOGIC HETEROGENEITY OFALZHEIMER’SDISEASEANDRELATED DISORDERS Melissa E. Murray,Mayo Clinic Florida, Jacksonville, FL, USA. Contact e-mail: murray.melissa@mayo.edu Figure 1. Mean PET SUVR maps within each clinical group and statistical group comparison. Left panel shows marginal (covariate-adjusted) means derived from the voxelwise ANCOVA model, illustrating the estimated average maps in each group, adjusting for age (fixed at the overall sample mean 1⁄4 64yo) and CDR-SB (fixed at the overall sample mean 1⁄4 4). Background: Alzheimer’s disease (AD) is neuropathologicallydefined as a multi-proteinopathy involving abnormally aggregated intracellular tau as neurofibrillary tangles and extracellular accumulation of amyloid-b plaques. As we move toward an era where biomarkers of amyloid, neurodegeneration, and tau are being integrated with syndromal cognitive stage, we move away from dependence on clinical presentation to define AD as recommended by the NIA-AA Research Framework. Methods: Autopsied individuals with antemortem MRI and PET imaging will be reviewed. To complement the overall session we will focus the discussion on autopsied individuals who were clinically diagnosed as AD dementia, dementia with Lewy bodies, or frontemporal dementia spectrum. Results: Neuropathologic and neuroimaging studies have demonstrated the importance of recognizing individuals with AD whose pattern of tau pathology may diverge from what is expected based upon Braak tangle staging. These atypical AD subtypes may inform our understanding of disease progression and impact on clinical syndrome, which appears to be specific to tau and not amyloid. Given that age is the strongest risk factor for AD, we must also consider the contribution of age-related tau in astroglial (ARTAG) or argyrophilic grains disease. Moreover, the impact of the TDP-43 proteinopathy, hippocampal sclerosis, will be discussed for potential contribution to variability in neurodegeneration biomarkers. Lewy body pathology is found in more than 50% of autopsied AD cases, whereas amyloid-b plaques are found in nearly all autopsied dementia with Lewy body patients. Oppositely, frontemporal dementia cases with underlying TDP-43 for the most part lack significant amyloid-b pathology. An emphasis on this overlap of co-existing pathologies (or lack thereof) will be discussed. Conclusions:The cross-talk between the aerial view provided of the brain from the neuroimaging perspective and the ground-level knowledge of the neurobiology from neuropathology studies are critical toward advancing our understanding of these complex diseases. F1-01-02 NON-AMNESTIC PHENOTYPES OF Figure 2. Older age is associated with lower cortical Flortaucipir, but not PIB SUVR. All models included CDR-SB as a covariate. The lower panel ALZHEIMER’S DISEASE, EARLYAGE OF ONSETAND APOE GENOTYPE ARE ASSOCIATEDWITH TAU-, NOTAb-PET illustrates the histogram of each correlation map presented above, showing that the vast majority of gray matter (GM) voxels had negative associations between Flortaucipir SUVR and age in all groups. Renaud La Joie, Adrienne Visani, Orit H. Lesman-Segev, Viktoriya Bourakova, Mustafa Janabi, Zachary A. Miller, James P. O’Neil, David Perry, Julie Pham, Howard J. Rosen, Bruce L. Miller, William J. Jagust, Gil D. Rabinovici, University of California, San Francisco, San Francisco, CA, USA; University of California San Francisco, San Francisco, CA, USA; Lawrence Berkeley National Laboratory, Berkeley, CA, USA; Memory and Aging Center, Weill Institute for Neurosciences, University of California San Francisco, San Francisco, CA, USA; University of California San Francisco, San Francisco, CA, USA; University of California, Berkeley, Berkeley, CA, USA. Contact e-mail: renaud.lajoie@gmail.com Background: The factors underlying AD ‘atypical variants’ (e.g. early symptom onset or non-amnestic phenotypes) are poorly understood. We aimed to relate a patient’s clinical presentation of AD, age, and APOE genotype to the amount and topography of b-amyloid and tau pathologies using positron emission tomography (PET) with PIB and Flortaucipir. Methods:We studied 86 patients with a clinical diagnosis of dementia or MCI due to AD and a positive PIB-PET (Table 1), including 17 patients who also met criteria for logopenic variant primary progressive aphasia (lvPPA, i.e. language-predominant AD) and 18 who met criteria for Posterior Cortical Atrophy (PCA, visuospatial-predominant AD). Standardized Uptake Value Ratio (SUVR) images of PIB (50-70 min) and Flortaucipir (80-100 min) were calculated using reference regions from MRI (cerebellar cortex and inferior cerebellar cortex, respectively), and warped to template space. Statistical models assessed Figure 3. APOE-ε4 is specifically associated with increased Flortaucipir SUVR in the medial temporal lobe. Left. Voxelwise ANCOVA analyses of Flortaucipir, comparing 41 ε4 carriers (32 ε3/ε4 + 9 ε4/ε4) to 35 ε4 non-carriers (34 ε3/ε3 + 1 ε2/ε3); two patients carrying ε2/ε4 were not included in this analysis. No significant cluster was found in the ε4 carriers < ε4 non-carriers contrast, even at the uncorrected p<0.001 threshold. Right. Confirmation of the voxelwise finding based on the average Flortaucipir SUVR extracted from the medial temporal lobe in native space using FreeSurfer (combining hippocampus, amygdala, and entorhinal cortex). Bars indicate median and quartiles within each group and the effect size is given as Cohen’s d (with 95% confidence interval) Table 2 Associations between global cortical PET SUVR and clinical/demographic characteristics DV: PIBcortex F P h 2 P Post-hoc Clinical group 1.2 0.31 0.029 Age 1.0 0.32 0.012 CDR-SB 1.0 0.31 0.013 DV: Flortaucipircortex F P h 2 P Clinical group 5.5 0.006 0.120 PCA>AD; lvPPA>AD Age 18.7 <0.001 0.188 Negative association CDR-SB 8.3 0.005 0.093 Positive association The table reports two separate ANCOVA models (one for PIB, one for Flortaucipir); all three predictors were included in each model. Podium Presentations: Sunday, July 22, 2018 P200 relationships between PET-SUVR, clinical presentation, and age, while controlling for disease severity using the clinical dementia rating scale sum of boxes (CDR-SB). Results:Global cortical PIB SUVR was not correlated with clinical group, age, or CDR-SB (p’s >0.3, Table 2). Voxelwise analyses showed high betweengroup regional similarity in PIB patterns, with slightly increased occipital SUVR in PCA at an uncorrected threshold (Figure 1, top). In contrast, global cortical Flortaucipir was associated with all clinical and demographic variables (Table 2). Patterns of Flortaucipir strongly varied across groups, with increased occipito-parietal and dorsal frontal SUVR in PCA and higher left temporal and left inferior frontal SUVR in lvPPA (Figure 1, bottom). Elevated cortical Flortaucipir was associated with younger age in all clinical groups (Figure 2, top), and this effect was mainly observed in parietal and dorsal frontal cortices, independent of phenotype (Figure 2, bottom). APOE genotype was not significantly correlated with PIB SUVR, but the presence of ε4 was associated with a focal Flortaucipir increase in the medial temporal lobe (Figure 3). Conclusions: AD phenotypes were strongly related to patterns of tau, but not amTable 1 Demographics n IvPPA PCA Other AD