Amyloid PET is commonly interpreted using binary visual classification. However, quantitative assessment on the Centiloid scale identifies an intermediate range of amyloid burden not always captured by dichotomous interpretation. The biologic and clinical relevance of this intermediate category remain unclear, despite its implications for biomarker interpretation, prognostic stratification, and eligibility for amyloid-targeting therapies. Our aim was to characterize individuals within this intermediate range using multimodal biomarker profiles and longitudinal cognitive outcomes, as well as deriving Centiloid thresholds for early tau PET positivity, advanced tau PET positivity, and cognitive decline. Methods: We retrospectively analyzed participants who underwent amyloid PET between 2016 and 2024 at the Geneva Memory Center, including cognitively unimpaired individuals and patients with mild cognitive impairment or dementia. Global amyloid burden was quantified in Centiloids, and participants were categorized into low (<12), intermediate (12-37), or high (>37) groups. Group comparisons, receiver-operating-characteristic curve analyses, and linear mixed-effects models were used to assess fluid biomarker differences, define Centiloid thresholds for tau positivity and cognitive decline, and evaluate longitudinal change in Mini-Mental State Examination scores. Results: Among the 512 participants, 202 (39%) had low, 63 (12%) had intermediate, and 247 (48%) had high Centiloid values. The intermediate-Centiloid group showed biomarker profiles, including cerebrospinal fluid and plasma markers of amyloid, tau pathology, and neurodegeneration, between those of the low and high groups. Both the intermediate- and high-Centiloid groups exhibited faster cognitive decline than did individuals in the low-Centiloid group, with a moderate rate in the intermediate-Centiloid group. Thresholds of 13 and 14 Centiloids best identified individuals with cognitive decline and tau accumulation in the mesial temporal lobe, respectively, whereas 51 Centiloids best discriminated advanced neocortical tau involvement. Conclusion: Centiloid-based classification delineates biologically and clinically distinct stages along the amyloid continuum. Individuals with intermediate Centiloid values already show tau involvement and increased risk of cognitive decline, whereas higher amyloid levels are associated with advanced tau pathology. These findings support a 3-level interpretation of amyloid PET beyond binary classification and highlight the clinical relevance of the intermediate zone.

Protein deposition and neurodegeneration differently affect the brain spatially and temporally in Alzheimer's disease (AD). Here we used imaging transcriptomics to understand the biological and molecular properties underlying regional variability of neuroimaging phenotypes of amyloid, tau, and neurodegeneration assessed by PET and MRI. Brain patterns were estimated by contrasting imaging data between AD patients and healthy controls from two independent cohorts for replication (Geneva Memory Clinic and ADNI). Regional gene expression profiles were derived from brain-wide microarray measurements provided by the Allen human brain atlas (AHBA). Hypothesis-driven analyses assessed the spatial association between neuroimaging patterns and gene expression (gene-to-biomarker associations) for selected candidate genes for AD. Over-representation analysis (ORA) and gene set enrichment analysis (GSEA) were used to characterize molecular properties and biological pathways of genome-wide gene sets associated with regional AD pathologies in a data-driven manner. Regional patterns showed the highest amyloid load in frontal, parietal, and lateral temporal lobes, whereas tau deposition was most pronounced in medial temporal lobes and lateral temporoparietal areas. Neurodegeneration patterns were instead less widespread, involving mainly temporoparietal areas. Specific patterns of amyloid, tau and neurodegeneration were differently associated with AD-related genes. ORA and GSEA revealed that genes implicated in different aspects of protein synthesis (e.g. cytosolic ribosome, mitochondrion organization, and RNA metabolic processes) as well as immune regulation and neuroinflammation correlated exclusively with amyloid load, whereas genes involved in the synaptic organization, transmission, and function were associated to the severity of amyloid, tau, and neurodegeneration pathologies. GSEA confirmed that the gene-to-tau and gene-to-atrophy associations were related to similar biological pathways involving synaptic signaling and organization, while gene-to-hypometabolism associations were more related to cellular processes. Selective AD vulnerabilities were differently related to specific gene expression and molecular-biological properties, with a large set of genes associated with amyloid accumulation and a subset of genes conferring additional vulnerability to downstream tau. Our findings suggest that the spatial and temporal decoupling between amyloid deposition, tau deposition and neurodegeneration is explained by differential genetic expression but that shared mechanisms link upstream amyloid with subsequent tau pathology and loss of neuronal integrity.

Background Current AI models that rely solely on chest radiographs (CXRs) or clinical and biological data have limitations. Our goal was to determine if incorporating clinical and biological data to CXR data improved accuracy in the diagnosis of pneumonia. Methods This retrospective study compared three AI models: an imaging model based on a convolutional neural network (CNN) trained on CXRs alone; a clinico-biological model based on a support vector machine (SVM) using clinical and biological data with no CXR; and a multimodal model integrating all three types of information. Data were extracted from two independent cohorts: a training set (PneumOld-CT, n = 200, median age 84 years (78.6–90.2)) and an independent test set (PACSCAN, n = 230, mean age 65 years +/- 20) for whom the reference diagnosis was determined a posteriori by a multidisciplinary expert panel using multimodal data. We assessed diagnostic performance by the area under the receiver operating characteristic curve (ROC-AUC) and we compared the models using DeLong’s test. Calibration curves and decision curve analysis (DCA) were also evaluated. Results In the independent test set, the multimodal AI model demonstrated significantly higher ROC-AUC than the imaging-based model (p < 0.05) or the clinico-biological model (p < 0.005). DCA confirmed a greater net clinical benefit for the multimodal model. Conclusion Integrating radiographic, clinical and biological data to develop a multimodal AI model significantly improved pneumonia diagnosis compared to a single-or a dual modality AI model. This multimodal approach has the potential to improve diagnostic support, especially in complex clinical scenarios.

This case highlights the rare coexistence of Wernicke's encephalopathy, probable progressive supranuclear palsy and probable Alzheimer's disease. The patient presented with acute confusion, repeated falls, postural instability, bradyphrenia, apathy and vertical gaze palsy. Clinical examination raised suspicion of overlapping neurological syndromes. Complementary investigations included brain MRI which revealed periventricular hyperintense regions on T2 and on fluid-attenuated inversion recovery (FLAIR) sequences, suggestive of Wernicke's encephalopathy. A subsequent ioflupane (123I) single-photon emission computed tomography (SPECT) scan showed left-sided dysfunction of presynaptic dopamine transporters, supporting a parkinsonian syndrome. Cerebrospinal fluid biomarkers finally demonstrated elevated total and phosphorylated-tau levels and a decreased Aβ42/Aβ40 ratio, consistent with Alzheimer's pathology. Timely intravenous thiamine was initiated. Once the presumed diagnoses were clinically confirmed, multidisciplinary management was implemented, including physiotherapy and caregiver support.This case illustrates how acute syndromes can unmask or exacerbate chronic neurodegenerative conditions, underscoring, in addition to the need for early intervention, the necessity of a thorough clinical approach. This is especially the case in older adults presenting with complex cognitive and motor symptoms.

Relative cerebral blood flow (rCBF), assessed using pulsed arterial spin labeling (pASL) MRI, and the standardized uptake value ratio (SUVr) in early-phase amyloid-PET (ePET) are used as proxies for brain perfusion. These methods have the potential to streamline clinical workflows and reduce the burden on patients by eliminating the need for additional procedures. While both techniques have shown good agreement with the gold standard for glucose metabolism assessment, F-fluorodeoxyglucose-PET, a direct comparison between them has yet to be fully clarified. This retrospective study aimed to compare perfusion-like data from pASL (rCBF) and ePET (SUVr) in a memory clinic cohort. We included 46 subjects (69 ± 8 years; 37 women) from the Geneva Memory Center (cognitively impaired-CI n = 29; cognitively unimpaired-CU n = 17), with available pASL and ePET. We evaluated the association between rCBF and SUVr values across 18 cortical and subcortical regions using linear regression and the within-subject coefficient of variation (wsCV). Regional differences between CU and CI groups were assessed using linear regression model corrected for age. We observed significant association between rCBF and SUVr in precuneus (β = 0.69, wsCV = 16.9), angular gyrus (β = 0.64, wsCV = 19.4), and hippocampus (β = 0.23, wsCV = 16.1). Additionally, significant differences in rCBF between CU and CI were also observed in the posterior cingulate, precuneus, calcarine, hippocampus, and composite (p < 0.05), while SUVr showed significant differences only in the hippocampus. Our findings indicate weak to moderate local correlations between the two techniques. However, both exhibited differing regional perfusion levels in CU and CI groups, with rCBF showing more regional differences between cognitive stages in comparison with SUVr.

This study aimed to test the ability to visually detect the characteristic medial temporal and limbic hypometabolic pattern of limbic age-related TDP-43 encephalopathy (LATE) in 18F-FDG-PET of patients with amnestic mild cognitive impairment (aMCI) and evaluate its prognostic value. We included 70 patients with aMCI who underwent 18F-FDG-PET, amyloid-PET, tau-PET, and structural MRI, as well as baseline and follow-up cognitive evaluation. 18F-FDG-PET scans were analyzed visually with single-subject maps, categorized as normal, Alzheimer's disease (AD)-like, LATE-like, or other neurodegenerative diseases, while blinded from other data. Clinical and biomarker features as well as cognitive trajectories were compared between groups. 25 scans were classified as normal, 25 as AD-like, 12 as LATE-like, and 8 as others. Patients with AD-like patterns were younger, had lower MMSE scores, inferior-to-medial temporal metabolism ratio, and greater hippocampal atrophy and cortical tau load than subjects with normal scans. Patients with LATE-like patterns had lower MMSE scores and more hippocampal atrophy than subjects with normal scans. Patients with LATE-like patterns were significantly older, had greater inferior-to-medial temporal metabolism ratio, greater amygdalar atrophy, and lower cortical tau load than subjects with AD-like patterns. Only subjects classified as AD-like showed a faster cognitive decline than negative scans. LATE-like hypometabolic pattern in aMCI can identify a subgroup of subjects distinct from AD and controls in terms of clinical severity, medial temporal atrophy, cortical tau load, and cognitive decline, supporting the utility of 18F-FDG-PET as a biomarker that provides inferential support for the specific detection of LATE.

BACKGROUND:Tau-Positron Emission Tomography (PET) has become central in Alzheimer's disease (AD) research and clinical settings. Multiple preprocessing pipelines for tau-PET quantification have been described, with satisfactory performance but direct comparisons remain scarse. Our study evaluates the comparability of two commonly used PET preprocessing methods, respectively in native and standard spaces, in quantifying tau deposition and in their ability to discriminate AD patients. METHODS:209 subjects were included from the Geneva memory clinic including cognitively unimpaired (CU) individuals, mild cognitive impairment (MCI) and dementia patients. Images were processed in native and standard space using inferior cerebellar grey matter as reference region. Standardized uptake value ratios (SUVR) were extracted from AD-specific regions. Correlations between SUVR obtained by different methods and plasma biomarkers were assessed. ROC analyses compared the ability of the two methods to discriminate visually assessed tau status, amyloid-positive cognitively impaired from amyloid-negative CU, and subjects with declining cognition over time. RESULTS:SUVR from the two methods were strongly correlated across all regions. However, SUVR values obtained with standard space method showed higher values. SUVR in the medial temporal lobe from native space processing provided a greater accuracy in discriminating positive scans and identifying subjects with cognitive decline. For all other analyses methods performed equally well. The correlation with plasma biomarkers was comparably high with both methods. CONCLUSION:While preprocessing in native and standard space is adequate for quantifying 18F-Flortaucipir PET and for discriminating AD patients, higher accuracy can be obtained in the mesial temporal regions and to predict cognitive decline using processing in native space.

Diagnosing cerebral amyloid angiopathy (CAA) is essential when determining Alzheimer's disease (AD) patients’ eligibility for anti-amyloid immunotherapy since CAA is associated with amyloid-related imaging abnormalities. The Boston criteria are widely used in memory clinics to identify older adults with a high probability of CAA; however, the prevalence and clinical characteristics of CAA among cognitively impaired patients with AD remain unclear. This retrospective study aimed to determine the prevalence of CAA in cognitively impaired older adults with biomarker-confirmed AD (CI-AD) using the updated Boston criteria (v2.0), and to compare the cognitive, clinical, and biomarker profiles of CI-AD patients with and without CAA. 415 patients (mean age 73.8 ± 7.0 years) with probable AD, confirmed by lumbar puncture or positron emission tomography (PET) biomarkers, were retrospectively identified from the Geneva University Hospital Memory Center's database (June 2012 - July 2024). MRI scans were analyzed by a board-certified radiologist or neuroradiologist, independently reviewed by a trained image analyst, and imaging diagnoses confirmed by an experienced neurologist, the latter two blinded to clinical data. Participants were classified as having a high (AD-CAA) or low probability of CAA (AD-nCAA) using the Boston criteria. Patient characteristics and biomarker distributions were compared between groups using the Chi-squared test or Fisher's exact test for categorical variables, and the Mann-Whitney U test for continuous variables. 29% ( n  = 119) of AD patients were classified as AD-CAA, whereas 71% ( n  = 296) were classified as AD-nCAA. Clinical severity, global cognition, verbal episodic memory, and executive functions were comparable between groups. AD-CAA patients were older, more likely to use antiplatelet therapy, and exhibited a higher prevalence of cardiovascular disease, despite having similar cardiovascular risk factors. No significant differences between groups were observed in fluid AD biomarkers, PET amyloid burden, or medial temporal atrophy assessed through imaging. The prevalence of CAA among AD patients was lower than pathology-based estimates. AD patients from memory clinics may exhibit comparable cognitive profiles and similar patterns of AD biomarkers regardless of their CAA probability. These findings underscore the importance of further research in the application of the Boston criteria in memory clinic patients with AD.

BACKGROUND:The Three-Objects-Three-Places (3O3P) test is a 5-min screen for episodic memory impairment due to Alzheimer's disease, known for its briefness and easy administration, culture- and language-free nature, and the absence of specific equipment. However, no studies have validated its potential in memory clinic cohorts. The aim of this study was to test its convergent, discriminant, and known-group validities and to define thresholds for its clinical use. METHODS:We included 2062 cognitively unimpaired (CU), mild cognitive impairment (MCI) and dementia patients from the Geneva Memory Center cohort who underwent the 3O3P test in the context of clinical practice. Convergent and discriminant validities were assessed using an exploratory factor analysis. The known-group validity was assessed in CU vs. MCI and dementia using the area under the curve (AUC). 3O3P test scores vs. amyloid and tau positivity, neurodegeneration, and cognition (ATNC) were assessed using the Kruskal-Wallis test. The 3O3P test cut-offs were calculated using sensitivity, specificity, PPV, NPV, and accuracy. RESULTS:Mean age was 72 years (SD = 11), 60% were female, mean education was 13 years (SD = 4), and mean MMSE was 25 (SD = 5). The 3O3P and Delayed Total Recall tests loaded strongly on the "memory" factor and weakly on "non-memory" factors. The 3O3P test can discriminate CU vs. MCI (AUC = 0.71) and dementia (AUC = 0.92). Higher 3O3P scores were associated with lower prevalence of ATNC (p < 0.001). A 3O3P value of 7 can detect MCI and dementia patients. CONCLUSIONS:The 3O3P test has demonstrated good convergent, discriminant, and known-group validity in a large memory clinic population.

Purpose As dual-phase amyloid-PET can evaluate amyloid (A) and neurodegeneration (N) with a single tracer injection, dual-phase tau-PET might be able to provide both tau (T) and N. Our study aims to assess the association of early-phase tau-PET scans and 18 F-fluorodeoxyglucose (FDG) PET and their comparability in discriminating Alzheimer’s disease (AD) patients and differentiating neurodegenerative patterns. Methods 58 subjects evaluated at the Geneva Memory Center underwent dual-phase 18 F-Flortaucipir-PET with early-phase acquisition (eTAU) and 18 F-FDG-PET within 1 year. A subsample of 36 participants also underwent dual-phase amyloid-PET (eAMY). Standardized uptake value ratios (SUVRs) were calculated to assess the correlation of eTAU and their respective 18 F-FDG-PET and eAMY scans. Hypometabolism and hypoperfusion maps and their spatial overlap were also evaluated at the individual level visually and semiquantitatively. Receiver operating characteristic analyses were performed to compare the discriminative power of eTAU, FDG, and eAMY SUVR between A-/T- and A+/T + participants. Results Strong positive correlations were found between eTAU and FDG SUVRs ( r = 0.84, p < 0.001) and eTAU and eAMY SUVRs ( r > 0.87, p < 0.001). Clusters of significant hypoperfusion with good correspondence to hypometabolism topographies were found at the individual level, independently of the underlying neurodegenerative patterns. Both eTAU and FDG SUVRs significantly distinguished A+/T + from A-/T- individuals (AUC eTAU =0.604, AUC FDG =0.748) with FDG performing better than eTAU ( p = 0.04). eAMY and eTAU SUVR showed comparable discriminative power. Conclusion Early-phase 18 F-Flortaucipir-PET can provide perfusion information closely related to brain regional glucose metabolism and perfusion measured by early-phase amyloid-PET, even if less accurate than FDG-PET as a biomarker for neurodegeneration.

BACKGROUND:Diagnosing cerebral amyloid angiopathy (CAA) is essential when determining Alzheimer's disease (AD) patients' eligibility for anti-amyloid immunotherapy since CAA is associated with amyloid-related imaging abnormalities. The Boston criteria are widely used in memory clinics to identify older adults with a high probability of CAA; however, the prevalence and clinical characteristics of CAA among cognitively impaired patients with AD remain unclear. This retrospective study aimed to determine the prevalence of CAA in cognitively impaired older adults with biomarker-confirmed AD (CI-AD) using the updated Boston criteria (v2.0), and to compare the cognitive, clinical, and biomarker profiles of CI-AD patients with and without CAA. METHOD:415 patients (mean age 73.8 ± 7.0 years) with probable AD, confirmed by lumbar puncture or positron emission tomography (PET) biomarkers, were retrospectively identified from the Geneva University Hospital Memory Center's database (June 2012 - July 2024). MRI scans were analyzed by a board-certified radiologist or neuroradiologist, independently reviewed by a trained image analyst, and imaging diagnoses confirmed by an experienced neurologist, the latter two blinded to clinical data. Participants were classified as having a high (AD-CAA) or low probability of CAA (AD-nCAA) using the Boston criteria. Patient characteristics and biomarker distributions were compared between groups using the Chi-squared test or Fisher's exact test for categorical variables, and the Mann-Whitney U test for continuous variables. RESULT:29% (n = 119) of AD patients were classified as AD-CAA, whereas 71% (n = 296) were classified as AD-nCAA. Clinical severity, global cognition, verbal episodic memory, and executive functions were comparable between groups. AD-CAA patients were older, more likely to use antiplatelet therapy, and exhibited a higher prevalence of cardiovascular disease, despite having similar cardiovascular risk factors. No significant differences between groups were observed in fluid AD biomarkers, PET amyloid burden, or medial temporal atrophy assessed through imaging. CONCLUSION:The prevalence of CAA among AD patients was lower than pathology-based estimates. AD patients from memory clinics may exhibit comparable cognitive profiles and similar patterns of AD biomarkers regardless of their CAA probability. These findings underscore the importance of further research in the application of the Boston criteria in memory clinic patients with AD.

INTRODUCTION:Whether Alzheimer's disease pathology involves white matter pathways connecting the locus coeruleus (LC) to the entorhinal cortex (EC) is unclear. In this cross-sectional observational study, we investigated the microstructural integrity of the LC-EC pathway in relation to amyloid, tau, and neurodegeneration (ATN) biomarkers along the cognitive spectrum from normal cognition to dementia. METHODS:One hundred twenty-four participants underwent clinical assessment, diffusion-weighted imaging, structural magnetic resonance imaging (N), amyloid (A), and tau (T) positron emission tomography. Diffusivity indices were assessed in the LC-EC tract using a probabilistic atlas, and linear models were used to assess associations with ATN markers and cognition. RESULTS:Differences in LC-EC microstructural parameters were observed in participants with Braak stage > I versus Braak 0 (p < 0.020), N+ versus N- (p < 0.001), and cognitively impaired versus unimpaired (p < 0.019). LC-EC mean diffusivity was associated with Mini-Mental State Examination score even after accounting for ATN markers (p = 0.015). DISCUSSION:Our results suggest that LC-EC diffusivity provides complementary information over ATN biomarkers in explaining cognitive impairment. HIGHLIGHTS:Locus coeruleus-entorhinal cortex (LC-EC) tract microstructure is associated with tau and especially neurodegeneration markers. LC-EC tract microstructure is more sensitive to tau pathology and neurodegeneration than tracts commonly affected in Alzheimer's disease. LC-EC diffusivity measures provide complementary information over amyloid, tau, and neurodegeneration (ATN) biomarkers.

The identification of Alzheimer's disease (AD) pathology at an early stage utilizing plasma biomarkers has attracted significant interest due to its potential for improving global screening programs. Different forms of p-tau measured in plasma, mainly p-tau217, have demonstrated similar diagnostic accuracy when compared to traditional biomarkers. Moreover, plasma biomarkers associated with neuroinflammation and neurodegeneration, namely GFAP and NfL, respectively, have been found to be elevated in patients with amyloidosis and tau accumulation, even though these conditions are not exclusive to AD. Our goal was to see if there were any differences in plasma biomarkers between amyloid (A+) and tau (T+) positive groups based on PET scans and compare their diagnostic accuracy in a Geneva Memory Center non-demented group. Plasma, amyloid-PET, and tau-PET were assessed in a total of 100 subjects (CU = 33; MCI = 67). Plasma biomarkers included were p-tau217, p-tau231, p-tau181, GFAP, and NfL. The differences in each plasma biomarker among A+/T+ groups were tested using Kruskal-Wallis tests. Additionally, we calculated the receiver operating characteristic (ROC) and underlying area under the curve (AUC) in A+ and T+. All plasma biomarkers revealed significant differences between A-/T- and A+/T+ (p < 0.05), where p-tau217 showed a higher effect size (δ = 0.98) in comparison with other biomarkers (δ range = 0.31 - 0.67). Moreover, p-tau217 was the only biomarker displaying significant differences between A-/T- and A+/T- (p < 0.001; δ = 0.89). Lastly, p-tau217 showed a higher AUC for detecting amyloid and tau (AUC average = 0.96) than other p-tau forms (AUC average = 0.77) and GFAP/NfL (AUC average = 0.69). Plasma p-tau217 revealed superior performance in AD pathology detection when compared to other p-tau forms and plasma biomarkers of neuroinflammation/neurodegeneration. Our results suggest the potential of p-tau217 to identify AD pathology in a non-demented population.

BACKGROUND AND OBJECTIVES:Tau accumulation pattern shows substantial variability in Alzheimer disease (AD), and 4 distinct spatiotemporal trajectories were distinguished using a data-driven approach called the Subtype and Stage Inference (SuStaIn). A visual method to validate and identify these subtypes is a requirement for their clinical translation. Our study aimed to provide a standardized topographic method for identifying tau patterns visually using tau-PET in a clinical setting. METHODS:Participants in this prospective study were included from the memory clinic of Geneva University Hospital. Inclusion criteria required participants to have undergone at least 1 18F-Flortaucipir tau-PET scan and a Mini-Mental State Examination (MMSE) within a 1-year time frame. All scans were classified into different tau subtypes (limbic [S1], medial temporal lobe-sparing [S2], posterior [S3], and lateral temporal [S4]) using both visual rating and SuStain algorithm. A subgroup underwent amyloid-PET and clinical follow-up. Cohen's κ tested the agreement between raters and between visual and automated subtypes. Chi-squared and Kruskal-Wallis tests assessed differences in clinical and biomarker features between subtypes, whereas differences in cognitive trajectories were tested using linear mixed-effects models, controlling for age, sex, and clinical and tau stages. RESULTS:A total of 245 tau-PET scans of individuals ranging from cognitively unimpaired to mild dementia (mean age: 68.25 years, 52% women) were included and classified into different tau pattern subtypes. A substantial agreement between raters was found in visually interpreting tau subtypes (κ > 0.65, p < 0.001) and a fair agreement between visual and automated subtypes (κ = 0.39, p < 0.001), with the automated approach more likely to classify a scan as tau negative and lower agreement between methods in more severe cases and AD clinical variants. Regarding the visual classification, individuals with S2 subtype were younger than S1 and S3, had lower MMSE and verbal fluency scores than S4 and S1, showed higher global tau burden than other subtypes, and a steeper cognitive decline. DISCUSSION:Visual classification reliably identified 4 tau patterns that differ in global tau load, clinical features, and long-term outcomes, suggesting its clinical usefulness for the detection of higher-risk AD variants. A clinically implementable classification of subtypes with faster decline is paramount for personalized diagnosis, accurate prognosis, and treatment.

This study investigated sex differences in the associations between Alzheimer’s disease (AD) biomarkers, cognitive performance, and decline in memory clinic settings. 249 participants (females/males:123/126), who underwent tau-PET, amyloid-PET, structural MRI, and plasma glial fibrillary acidic protein (GFAP) measurement were included from Geneva and Lausanne Memory Clinics. Mann-Whitney U tests investigated sex differences in clinical and biomarker data. Linear regression models estimated the moderating effect of sex on the relationship between biomarkers and cognitive performance and decline. Sex differences in cognitive decline were further evaluated using longitudinal linear mixed-effect models with three-way interaction effects. Women and men present similar clinical features, amyloid, and neurodegeneration. Women had higher tau load and plasma levels of GFAP than men (p < 0.05). Tau associations with amyloid (standardized β = 0.54,p < 0.001), neurodegeneration (standardized β=-0.44,p < 0.001), and cognition (standardized β=-0.48,p < 0.001) were moderated by a significant interaction with sex. Specifically, the association between amyloid and tau was stronger among women than men (standardized β=-0.19,p = 0.047), whereas the associations between tau and cognition and between tau and neurodegeneration were stronger among men than in women (standardized β=-0.76,p = 0.001 and standardized β=-0.56,p = 0.044). Women exhibited faster cognitive decline than men in the presence of severe cortical thinning (p < 0.001). Women showed higher tau load and stronger association between amyloid and tau than men. In individuals with high tau burden, men exhibited greater neurodegeneration and cognitive impairment than women. These findings support that sex differences may impact tau deposition through an upstream interplay with amyloid, leading to downstream effects on neurodegeneration and cognitive outcomes.

Plasma biomarkers have been increasingly studied in Alzheimer’s disease due to their potentially high accessibility, affordability, and low invasiveness. Recent studies have shown that baseline plasma levels are capable of predicting cognitive decline in cognitively unimpaired subjects (CU) and with mild cognitive impairment (MCI). Despite the fact that neuroimaging biomarkers are also strong predictors, it is still unclear how well they perform when compared to plasma biomarkers in predicting cognitive deterioration. Our goal was to evaluate how plasma (i.e., p-tau181, p-tau231, p-tau217, GFAP, and NfL) and traditional neuroimaging biomarkers (i.e., amyloid-PET centiloid, tau-PET global-SUVr, and hippocampal volume) are able to predict cognitive decline in non-demented individuals. A total of 100 subjects were included in this study, namely 67 MCI and 33 CU from the Geneva Memory Center cohort. The study applied linear mixed-effect regression models to evaluate how baseline biomarker levels predict cognitive decline. Moreover, sample sizes for future AD clinical trials were calculated for a subset of amyloid-positive subjects based on how well each biomarker could predict the disease. Cognitive decline was only significantly predicted in the MCI subsample by baseline plasma NfL (β = -0.53, p = 0.02) p-tau217 (β = -0.59, p = 0.009), and tau-SUVR (β = -0.91, p = 0.002). Lastly, we have proven that the sample sizes for future treatment trials targeting AD can be reduced by up to 70% in individuals who exhibit both amyloid-PET positivity by incorporating NfL as a requirement for inclusion. On the other hand, the inclusion of p-tau217 might decrease amyloid-positive sample sizes only by 10%. Tau-SUVR was the best predictor of cognitive decline in the MCI stage, followed by plasma p-tau217, and NfL. However, when the goal is to identify at-risk subjects for future AD clinical trials with amyloid-PET positivity in the inclusion criteria, the NfL seems to be the most appropriate biomarker to evaluate. This implies a strong association between amyloid-PET and p-tau217, whereas NfL may be more appropriate to identify individuals at-risk for developing cognitive decline.

BACKGROUND AND OBJECTIVES:Limbic-predominant age-related TDP-43 encephalopathy (LATE) is a neurodegenerative condition often overlapping clinically with Alzheimer disease (AD). No established in vivo biomarkers exist for LATE. The amygdalar atrophy scale (AAS) is a visual MRI rating tool scoring the amygdala as AAS0 (no atrophy), AAS1 (mild-to-moderate atrophy), or AAS2 (severe atrophy) and is associated with TDP-43 neuropathology in limbic regions. This study describes the clinical, neuroimaging, and longitudinal cognitive profiles of individuals classified using a proposed framework that combines the AAS with established AD biomarkers. METHODS:This retrospective study included individuals attending the Geneva Memory Center (2014-2022) who underwent T1-weighted MRI, as well as clinical and neuropsychological assessments. Participants were categorized with an etiologic grouping defined as No-AD/No-LATE (AAS0, negative AD biomarkers), AD (AAS0, positive AD biomarkers), LATE (AAS1-2, negative AD biomarkers), or AD/LATE (AAS1-2, positive AD biomarkers). Group differences were compared using Kruskal-Wallis or χ2 tests. Differences in MRI brain volumes and cortical thickness between groups were examined with analysis of covariance. Cognitive trajectories using Mini-Mental State Examination (MMSE) scores over 30 months were estimated with linear mixed-effects models accounting for baseline age and education. RESULTS:The analytic sample included 469 participants (mean age 71.2 ± 8.5 years, 53% female). The No-AD/No-LATE group exhibited the highest MMSE scores (N = 181, 27.9 ± 2.2), outperforming AD (N = 146, 24.2 ± 4.9), LATE (N = 36, 26.2 ± 3.2), and AD/LATE (N = 106, 22.5 ± 5.3) groups (p < 0.001). On the Free and Cued Selective Reminding Test, the No-AD/No-LATE group showed the highest scores (15.1 ± 1.8); the LATE and AD groups performed similarly (13.3 ± 3.5 and 12.5 ± 4.0), both exceeding the AD/LATE group (10.6 ± 4.1) (p < 0.001). LATE and AD/LATE groups showed lower volumes/thicknesses in TDP-43-related regions compared with AD and No-AD/No-LATE groups. Longitudinally, the LATE group maintained cognitive stability as the No-AD/No-LATE group (β = -0.010, 95% CI -0.096 to 0.075, p = 0.813), whereas both AD (β = -0.136, 95% CI -0.184 to -0.088, p < 0.001) and AD/LATE (β = -0.139, 95% CI -0.196 to -0.082, p < 0.001) groups showed faster decline over time. DISCUSSION:Patients identified as LATE by the integration of AAS with AD biomarkers had generally mild amnestic cognitive impairment, significant limbic atrophy, and slow decline over time, all features consistent with previous autopsy series. Moreover, AD/LATE cases were more impaired than those with LATE or AD alone. Prospective and neuropathologic validation is warranted.

BACKGROUND:Protein deposition and neurodegeneration differently affect the brain spatially and temporally in Alzheimer's disease (AD). Here we used imaging transcriptomics to understand the biological and molecular properties underlying regional variability of neuroimaging phenotypes of amyloid, tau, and neurodegeneration assessed by PET and MRI. METHOD:Brain patterns were estimated by contrasting imaging data between AD patients and healthy controls from two independent cohorts for replication (Geneva Memory Clinic and ADNI). Regional gene expression profiles were derived from brain-wide microarray measurements provided by the Allen human brain atlas (AHBA). Hypothesis-driven analyses assessed the spatial association between neuroimaging patterns and gene expression (gene-to-biomarker associations) for selected candidate genes for AD. Over-representation analysis (ORA) and gene set enrichment analysis (GSEA) were used to characterize molecular properties and biological pathways of genome-wide gene sets associated with regional AD pathologies in a data-driven manner. RESULT:Regional patterns showed the highest amyloid load in frontal, parietal, and lateral temporal lobes, whereas tau deposition was most pronounced in medial temporal lobes and lateral temporoparietal areas. Neurodegeneration patterns were instead less widespread, involving mainly temporoparietal areas. Specific patterns of amyloid, tau and neurodegeneration were differently associated with AD-related genes. ORA and GSEA revealed that genes implicated in different aspects of protein synthesis (e.g. cytosolic ribosome, mitochondrion organization, and RNA metabolic processes) as well as immune regulation and neuroinflammation correlated exclusively with amyloid load, whereas genes involved in the synaptic organization, transmission, and function were associated to the severity of amyloid, tau, and neurodegeneration pathologies. GSEA confirmed that the gene-to-tau and gene-to-atrophy associations were related to similar biological pathways involving synaptic signaling and organization, while gene-to-hypometabolism associations were more related to cellular processes. CONCLUSION:Selective AD vulnerabilities were differently related to specific gene expression and molecular-biological properties, with a large set of genes associated with amyloid accumulation and a subset of genes conferring additional vulnerability to downstream tau. Our findings suggest that the spatial and temporal decoupling between amyloid deposition, tau deposition and neurodegeneration is explained by differential genetic expression but that shared mechanisms link upstream amyloid with subsequent tau pathology and loss of neuronal integrity.

INTRODUCTION:Resilience, the ability to maintain cognition or brain integrity despite Alzheimer's disease (AD) pathology, is often quantified using the residual approach. However, the variability in methodology and correction methods for this approach raises concerns about the interpretability of findings across studies. METHODS:We assessed brain resilience (BR) and cognitive resilience (CR) in a memory clinic population using the residual approach. We compared non-corrected and corrected residuals' associations with risk factors using linear regression models, and their impact on longitudinal cognition using linear mixed-effects models. RESULTS:Corrected versus non-corrected BR yielded distinct, often opposing, associations. For example, glial fibrillary acidic protein (GFAP) was negatively associated with non-corrected BR (β = -0.33; p < 0.01) but positively with corrected BR (β = 0.5, p < 0.001). Only corrected CR measures yielded significant associations. Only corrected residuals predicted cognitive decline. DISCUSSION:The observed discrepancies raise questions about the reliability of the residual approach in accurately capturing resilience. HIGHLIGHTS:Corrected and non-corrected residuals show distinct associations with risk factors. Corrected and non-corrected residuals show different predictions of cognitive decline. These approaches may reflect general brain health rather than true resilience mechanisms.

Resilience, the ability to maintain normal cognition (cognitive resilience, CR) or brain integrity (brain resilience, BR) despite neuropathological burden, is often quantified using the residual approach. This method calculates residuals from a linear regression where the dependent variable is brain volume or cognition, and the independent variable is neuropathology. Residuals can be corrected to render them independent from the dependent variable, but comparative analyses of residual correction methods within memory clinic populations are scarce. Our study aims to bridge this gap by comparing non-corrected and corrected residual approaches. 112 MCI patients from the Geneva Memory Center who underwent amyloid-PET (A), tau-PET (T), MRI (N), and clinical and neuropsychological assessments were included. Standardized residuals were extracted from linear regression models between hippocampal volume and AT (BR) or MMSE scores and ATN (CR). We compared: (i) non-corrected residuals (nBR and nCR); (ii) residuals corrected by regressing the dependent variable out of the residuals (cBR and cCR); (iii) non-corrected residuals with the dependent variable included as covariate (covBR and covCR). Associations between CR or BR and demographic, clinical, cognitive, and biomarker variables were investigated. Linear mixed models explored the impact of CR and BR on cognitive trajectories over time. Results were compared with non-residual based statistical models. Corrected and non-corrected approaches yielded significantly different results. Several factors were associated with nBR, but not with cBR or covBR, such as age (nBR: β = -0.399, p < 0.001; cBR: β = -0.048, p = 0.641; covBR: β = -0.012, p = 0.614). Conversely, several factors were associated with cBR and covBR, bur not nBR, such as female sex (nBR: β = -0.102, p = 0.619; cBR: β = -0.506, p = 0.012; covBR: β = -0.108, p = 0.013). Similar results were observed with CR. Longitudinal assessments also indicated differences in the relationship between corrected/non-corrected resilience and cognitive trajectories. For example, there was a significant interaction of time x cBR on MMSE scores (p < 0.001), but not of time x nBR (p = 0.226) Researchers should be aware that the choice of residual correction method significantly influences study conclusions. Our findings emphasize the necessity of adopting sophisticated modeling approaches to accurately capture resilience.