The recent development of tau-specific positron emission tomography (PET) tracers enables in vivo quantification of regional tau pathology, one of the key lesions in Alzheimer's disease (AD). Tau PET imaging may become a useful biomarker for clinical diagnosis and tracking of disease progression but there is no consensus yet on how tau PET signal is best quantified. The goal of the current study was to evaluate multiple whole-brain and region-specific approaches to detect clinically relevant tau PET signal. Two independent cohorts of cognitively normal adults and amyloid-positive (Aβ+) patients with mild cognitive impairment (MCI) or AD-dementia underwent [18F]AV-1451 PET. Methods for tau tracer quantification included: (i) in vivo Braak staging, (ii) regional uptake in Braak composite regions, (iii) several whole-brain measures of tracer uptake, (iv) regional uptake in AD-vulnerable voxels, and (v) uptake in a priori defined regions. Receiver operating curves characterized accuracy in distinguishing Aβ- controls from AD/MCI patients and yielded tau positivity cutoffs. Clinical relevance of tau PET measures was assessed by regressions against cognition and MR imaging measures. Key tracer uptake patterns were identified by a factor analysis and voxel-wise contrasts. Braak staging, global and region-specific tau measures yielded similar diagnostic accuracies, which differed between cohorts. While all tau measures were related to amyloid and global cognition, memory and hippocampal/ entorhinal volume/thickness were associated with regional tracer retention in the medial temporal lobe. Key regions of tau accumulation included medial temporal and inferior/middle temporal regions, retrosplenial cortex, and banks of the superior temporal sulcus. Our data indicate that whole-brain tau PET measures might be adequate biomarkers to detect AD-related tau pathology. However, regional measures covering AD-vulnerable regions may increase sensitivity to early tau PET signal, atrophy and memory decline.

Memory decline accompanies Aβ accumulation in otherwise healthy, Aβ-negative older adults. Amyloid increases within the negative range may represent the earliest detectable indication of pathology with domain-specific cognitive consequences.

Objective: To examine the clinical and biomarker characteristics of patients with amyloid-negative Alzheimer disease (AD) and mild cognitive impairment (MCI) from the Alzheimer's Disease Neuroimaging Initiative (ADNI), a prospective cohort study. Methods:We first investigated the reliability of florbetapir2 PET in patients with AD and patients with MCI using CSF-Ab 1-42 as a comparison amyloid measurement. We then compared florbetapir2 vs florbetapir1 patients with respect to several AD-specific biomarkers, baseline and longitudinal cognitive measurements, and demographic and clinician report data.Results: Florbetapir and CSF-Ab 1-42 1/2 status agreed for 98% of ADs (89% of MCIs), indicating that most florbetapir2 scans were a reliable representation of amyloid status. Florbetapir2 AD (n 5 27/177; 15%) and MCI (n 5 74/217, 34%) were more likely to be APOE4-negative (MCI 83%, AD 96%) than their florbetapir1 counterparts (MCI 30%, AD 24%). Florbetapir2 patients also had less AD-specific hypometabolism, lower CSF p-tau and t-tau, and better longitudinal cognitive performance, and were more likely to be taking medication for depression. In MCI only, florbetapir2 participants had less hippocampal atrophy and hypometabolism and lower functional activity questionnaire scores compared to florbetapir1 participants. Conclusions:Overall, image analysis problems do not appear to be a primary explanation of amyloid negativity. Florbetapir2 ADNI patients have a variety of clinical and biomarker features that differ from their florbetapir1 counterparts, suggesting that one or more non-AD etiologies (which may include vascular disease and depression) account for their AD-like phenotype. Alzheimer's Disease Neuroimaging Initiative; AGD 5 argyrophilic grain disease; MCI 5 mild cognitive impairment; metaROI 5 previously validated region of interest; MMSE 5 Mini-Mental State Examination; MPRAGE 5 magnetization-prepared rapid gradient echo; RAVLT 5 Rey Auditory Verbal Learning Test; SUVR 5 standardized uptake value ratio; TBM-SyN 5 tensorbased morphometry-symmetric diffeomorphic image normalization method; VBM 5 voxel-based morphometry; WM 5 white matter.The rate of b-amyloid (Ab) negativity in clinically diagnosed Alzheimer disease (AD) varies across a variety of study populations and as a function of APOE genotype status.1-6 Previous studies of patients with clinically diagnosed AD have shown that 12% were negative on amyloid PET in a recent meta-analysis, 7 and 10%-25% of APOE4-negative patients with AD did not meet the neuropathologic criteria for AD at autopsy. 8,9 Older adults with an amnestic profile that is suggestive of AD comprise a diverse group with heterogeneous pathology. Hippocampal sclerosis, argyrophilic grain disease, vascular dementia, Lewy body disease, and frontotemporal dementia have been observed at autopsy in addition to AD pathology 10 and in Ab2 cases with an antemortem AD diagnosis.

There are conflicting results claiming that Alzheimer disease signature neurodegeneration may be more, less, or similarly advanced in individuals with β-amyloid peptide (Aβ)-negative (Aβ−) suspected non-Alzheimer disease pathophysiology (SNAP) than in Aβ-positive (Aβ+) counterparts. OBJECTIVE To examine patterns of neurodegeneration in individuals with SNAP compared with their Aβ+ counterparts. DESIGN, SETTING, AND PARTICIPANTS A longitudinal cohort study was conducted among individuals with mild cognitive impairment (MCI) and cognitively normal individuals receiving care at Alzheimer's Disease Neuroimaging Initiative sites in the United States and Canada for a mean follow-up period of 30.5 months from August 1, 2005, to June 30, 2015. Several neurodegeneration biomarkers and longitudinal cognitive function were compared between patients with distinct SNAP (Aβ− and neurodegeneration-positive [Aβ−N+]) subtypes and their Aβ+N+ counterparts. MAIN OUTCOMES AND MEASURES Participants were classified according to the results of their florbetapir F-18 (Aβ) positron emission tomography and their Alzheimer disease-associated neurodegeneration status (temporoparietal glucose metabolism determined by fluorodeoxyglucose F 18 [FDG]-labeled positron emission tomography and/or hippocampal volume [HV] determined by magnetic resonance imaging: participants with subthreshold HV values were regarded as exhibiting hippocampal volume atrophy [HV+], while subthreshold mean FDG values were considered as FDG hypometabolism [FDG+]). RESULTS The study comprised 265 cognitively normal individuals (135 women and 130 men; mean [SD] age, 75.5 [6.7] years) and 522 patients with MCI (225 women and 297 men; mean [SD] age, 72.6 [7.8] years). A total of 469 individuals with MCI had data on neurodegeneration biomarkers; of these patients, 107 were Aβ−N+ (22.8%; 63 FDG+, 82 HV+, and 38 FDG+HV+) and 187 were Aβ+N+ (39.9%; 135 FDG+, 147 HV+, and 95 FDG+HV+ cases). A total of 209 cognitively normal participants had data on neurodegeneration biomarkers; of these, 52 were Aβ−N+ (24.9%; 30 FDG+, 33 HV+, and 11 FDG+HV+) and 37 were Aβ+N+ (17.7%; 22 FDG+, 26 HV+, and 11 FDG+HV+). Compared with their Aβ+ counterparts, all patients with MCI SNAP subtypes displayed better preservation of temporoparietal FDG metabolism (mean [SD] FDG: Aβ-N+, 1.25 [0.11] vs Aβ+N+, 1.19 [0.11]), less severe atrophy of the lateral temporal lobe, and lower mean (SD) cerebrospinal fluid levels of tau (59.2 [32.8] vs 111.3 [56.4]). In MCI with SNAP, sustained glucose metabolism and gray matter volume were associated with disproportionately low APOE ε4 (Aβ-N+, 18.7% vs Aβ+N+, 70.6%) and disproportionately high APOE ε2 (18.7% vs 4.8%) carrier prevalence. Slower cognitive decline and lower rates of progression to Alzheimer disease (Aβ-N+, 6.5% vs Aβ+N+, 32.6%) were also seen in patients with MCI with SNAP subtypes compared with their Aβ+ counterparts. In cognitively normal individuals, neurodegeneration biomarkers did not differ between Aβ−N+ and Aβ+N+ cases. CONCLUSIONS AND RELEVANCE In...

Identifying patterns in the world requires noticing not only unusual occurrences, but also unusual absences. We examined how people learn from absences, manipulating the extent to which an absence is expected. People can make two types of inferences from the absence of an event: either the event is possible but has not yet occurred, or the event never occurs. A rational analysis using Bayesian inference predicts that inferences from absent data should depend on how much the absence is expected to occur, with less probable absences being more salient. We tested this prediction in two experiments in which we elicited people's judgments about patterns in the data as a function of absence salience. We found that people were able to decide that absences either were mere coincidences or were indicative of a significant pattern in the data in a manner that was consistent with predictions of a simple Bayesian model.