Association of glial fibrillary acid protein, Alzheimer's disease pathology and cognitive decline.

Increasing evidence shows that neuroinflammation is a possible modulator of tau spread effects on cognitive impairment in Alzheimer's disease. In this context, plasma levels of the glial fibrillary acidic protein (GFAP) have been suggested to have a robust association with Alzheimer's disease pathophysiology. This study aims to assess the correlation between plasma GFAP and Alzheimer's disease pathology, and their synergistic effect on cognitive performance and decline. A cohort of 122 memory clinic subjects with amyloid and tau positron emission tomography, MRI scans, plasma GFAP, and Mini-Mental State Examination (MMSE) was included in the study. A subsample of 94 subjects had a follow-up MMSE score at least one year after baseline. Regional and voxel-based correlations between Alzheimer's disease biomarkers and plasma GFAP were assessed. Mediation analyses were performed to evaluate the effects of plasma GFAP on the association between amyloid and tau PET, and tau PET and cognitive impairment and decline. GFAP was associated with increased tau PET ligand uptake in the lateral temporal and inferior temporal lobes in a strong left-sided pattern independently of age, gender, education, amyloid, and APOE status (ß=0.001, p < 0.01). The annual rate of MMSE change was significantly and independently correlated with both GFAP (ß=0.006, p < 0.01) and global tau SUVR (ß=4.33, p < 0.01), but not with amyloid burden. Partial mediation effects of GFAP were found on the association between amyloid and tau pathology (13.7%), and between tau pathology and cognitive decline (17.4%), but not on global cognition at baseline. Neuroinflammation measured by circulating GFAP is independently associated with tau Alzheimer's disease pathology and with cognitive decline, suggesting neuroinflammation as a potential target for future disease-modifying trials targeting tau pathology. Peretti et al. show that a circulatory marker of neuroinflammation-glial fibrillary acidic protein-is associated with tau pathology in lateral temporal and frontal regions in patients with Alzheimer's disease, independent of amyloid load. Neuroinflammation appears to modulate the association between amyloid and tau biomarkers.

A comparison of visual assessment and semi-quantification for the diagnostic and prognostic use of [18F]flortaucipir PET in a memory clinic cohort.

PURPOSE: [18F]Flortaucipir PET is a powerful diagnostic and prognostic tool for Alzheimer's disease (AD). Tau status definition is mainly based in the literature on semi-quantitative measures while in clinical settings visual assessment is usually preferred. We compared visual assessment with established semi-quantitative measures to classify subjects and predict the risk of cognitive decline in a memory clinic population. METHODS: We included 245 individuals from the Geneva Memory Clinic who underwent [18F]flortaucipir PET. Amyloid status was available for 207 individuals and clinical follow-up for 135. All scans were blindly evaluated by three independent raters who visually classified the scans according to Braak stages. Standardized uptake value ratio (SUVR) values were obtained from a global meta-ROI to define tau positivity, and the Simplified Temporo-Occipital Classification (STOC) was applied to obtain semi-quantitatively tau stages. The agreement between measures was tested using Cohen's kappa (k). ROC analysis and linear mixed-effects models were applied to test the diagnostic and prognostic values of tau status and stages obtained with the visual and semi-quantitative approaches. RESULTS: We found good inter-rater reliability in the visual interpretation of tau Braak stages, independently from the rater's expertise (k>0.68, p<0.01). A good agreement was equally found between visual and SUVR-based classifications for tau status (k=0.67, p<0.01). All tau-assessment modalities significantly discriminated amyloid-positive MCI and demented subjects from others (AUC>0.80) and amyloid-positive from negative subjects (AUC>0.85). Linear mixed-effect models showed that tau-positive individuals presented a significantly faster cognitive decline than the tau-negative group (p<0.01), independently from the classification method. CONCLUSION: Our results show that visual assessment is reliable for defining tau status and stages in a memory clinic population. The high inter-rater reliability, the substantial agreement, and the similar diagnostic and prognostic performance of visual rating and semi-quantitative methods demonstrate that [18F]flortaucipir PET can be robustly assessed visually in clinical practice.

The Taxonomy of Subjective Cognitive Decline: Proposal and First Clinical Evidence from the Geneva Memory Clinic Cohort.

INTRODUCTION: Subjective cognitive decline (SCD) is characterized by subjective cognitive concerns without objective cognitive impairment and is considered a risk factor for cognitive decline and dementia. However, most SCD patients will not develop neurodegenerative disorders, yet they may suffer from minor psychiatric, neurological, or somatic comorbidities. The aim of the present study was to provide a taxonomy of the heterogeneous SCD entity and to conduct a preliminary validation using data from a memory clinic sample. METHODS: Participants were fifty-five SCD individuals consecutively recruited at the Geneva Memory Center. Based on clinical reports, they were classified into three clinically pre-defined subgroups: (i) those with psychological or psychiatric comorbidities (Psy), (ii) those with somatic comorbidities (SomCom), (iii) and those with no apparent cause (NAC). Baseline demographics, clinical, cognitive, and biomarker differences among the SCD subgroups were assessed. Longitudinal cognitive changes (average 3 years follow-up) were modeled using a linear mixed model. RESULTS: Out of the 55 SCD cases, 16 were SomCom, 18 Psy, and 21 NAC. 47% were female, mean age was 71 years. We observed higher frequency of APOE ε4 carriers in NAC (53%) compared to SomCom (14%) and Psy (0%, p = 0.023) and lower level of plasma Aß42 in NAC (6.8 ± 1.0) compared to SomCom (8.4 ± 1.1; p = 0.031). SomCom subjects were older (74 years) than Psy (67 years, p = 0.011), and had greater medial temporal lobe atrophy (1.0 ± 1.0) than Psy (0.2 ± 0.6) and NAC (0.4 ± 0.5, p = 0.005). SomCom has worse episodic memory performances (14.5 ± 3.5) than Psy (15.8 ± 0.4) and NAC (15.8 ± 0.7, p = 0.032). We observed a slightly steeper, yet not statistically significant, cognitive decline in NAC (ß = -0.48) compared to Psy (ß = -0.28) and SomCom (ß = -0.24). CONCLUSIONS: NAC features a higher proportion of APOE ε4 carriers, lower plasma Aß42 and a trend towards steeper cognitive decline than SomCom and Psy. Taken together, these findings suggest that NACs are at higher risk of cognitive decline due to AD. The proposed clinical taxonomy might be implemented in clinical practice to identify SCD at higher risk. However, such taxonomy should be tested on an independent cohort with a larger sample size.

Insights into single-timepoint ASL hemodynamics: what visual assessment and spatial coefficient of variation can tell.

PURPOSE: Arterial spin labeling (ASL) represents a noninvasive perfusion biomarker, and, in the study of nonvascular disease, the use of the single-timepoint ASL technique is recommended. However, the obtained cerebral blood flow (CBF) maps may be highly influenced by delayed arterial transit time (ATT). Our aim was to assess the complexity of hemodynamic information of single-timepoint CBF maps using a new visual scale and comparing it with an ATT proxy, the "coefficient of spatial variation" (sCoV). MATERIAL AND METHODS: Individual CBF maps were estimated in a memory clinic population (mild cognitive impairment, dementia and cognitively unimpaired controls) and classified into four levels of delayed perfusion based on a visual rating scale. Calculated measures included global/regional sCoVs and common CBF statistics, as mean, median and standard deviation. One-way ANOVA was performed to compare these measures across the four groups of delayed perfusion. Spearman correlation was used to study the association of global sCoV with clinical data and CBF statistics. RESULTS: One hundred and forty-four participants (72 ± 7 years, 53% women) were included in the study. The proportion of maps with none, mild, moderate, and severe delayed perfusion was 15, 20, 37, and 28%, respectively. SCoV demonstrated a significant increase (p < 0.05) across the four groups, except when comparing none vs mild delayed perfusion groups (pBonf > 0.05). Global sCoV values, as an ATT proxy, ranged from 67 ± 4% (none) to 121 ± 24% (severe delayed) and were significantly associated with age and CBF statistics (p < 0.05). CONCLUSION: The impact of ATT delay in single-time CBF maps requires the use of a visual scale or sCoV in clinical or research settings.

Neuroimaging-guided diagnosis of possible FTLD-FUS pathology: a case report.

BACKGROUND: This case report presents a patient with progressive memory loss and choreiform movements. CASE PRESENTATION: Neuropsychological tests indicated multi-domain amnestic mild cognitive impairment (aMCI), and neurological examination revealed asymmetrical involuntary hyperkinetic movements. Imaging studies showed severe left-sided atrophy and hypometabolism in the left frontal and temporoparietal cortex. [18F]Flortaucipir PET exhibited moderately increased tracer uptake in hypometabolic areas. The diagnosis initially considered Alzheimer's disease (AD), frontotemporal degeneration (FTD), and corticobasal degeneration (CBD), cerebral hemiatrophy syndrome, but imaging and cerebrospinal fluid analysis excluded AD and suggested fused-in-sarcoma-associated FTD (FTLD-FUS), a subtype of the behavioural variant of FTD. CONCLUSIONS: Our case highlights that despite the lack of specific FUS biomarkers the combination of clinical features and neuroimaging biomarkers can guide choosing the most likely differential diagnosis in a complex neurological case. Imaging in particular allowed an accurate measure of the topography and severity of neurodegeneration and the exclusion of AD-related pathology.

Comparison of plasma and neuroimaging biomarkers to predict cognitive decline in non-demented memory clinic patients.

BACKGROUND: Plasma biomarkers of Alzheimer's disease (AD) pathology, neurodegeneration, and neuroinflammation are ideally suited for secondary prevention programs in self-sufficient persons at-risk of dementia. Plasma biomarkers have been shown to be highly correlated with traditional imaging biomarkers. However, their comparative predictive value versus traditional AD biomarkers is still unclear in cognitively unimpaired (CU) subjects and with mild cognitive impairment (MCI). METHODS: Plasma (Aß42/40, p-tau181, p-tau231, NfL, and GFAP) and neuroimaging (hippocampal volume, centiloid of amyloid-PET, and tau-SUVR of tau-PET) biomarkers were assessed at baseline in 218 non-demented subjects (CU = 140; MCI = 78) from the Geneva Memory Center. Global cognition (MMSE) was evaluated at baseline and at follow-ups up to 5.7 years. We used linear mixed-effects models and Cox proportional-hazards regression to assess the association between biomarkers and cognitive decline. Lastly, sample size calculations using the linear mixed-effects models were performed on subjects positive for amyloid-PET combined with tau-PET and plasma biomarker positivity. RESULTS: Cognitive decline was significantly predicted in MCI by baseline plasma NfL (ß=-0.55), GFAP (ß=-0.36), hippocampal volume (ß = 0.44), centiloid (ß=-0.38), and tau-SUVR (ß=-0.66) (all p < 0.05). Subgroup analysis with amyloid-positive MCI participants also showed that only NfL and GFAP were the only significant predictors of cognitive decline among plasma biomarkers. Overall, NfL and tau-SUVR showed the highest prognostic values (hazard ratios of 7.3 and 5.9). Lastly, we demonstrated that adding NfL to the inclusion criteria could reduce the sample sizes of future AD clinical trials by up to one-fourth in subjects with amyloid-PET positivity or by half in subjects with amyloid-PET and tau-PET positivity. CONCLUSIONS: Plasma NfL and GFAP predict cognitive decline in a similar manner to traditional imaging techniques in amyloid-positive MCI patients. Hence, even though they are non-specific biomarkers of AD, both can be implemented in memory clinic workups as important prognostic biomarkers. Likewise, future clinical trials might employ plasma biomarkers as additional inclusion criteria to stratify patients at higher risk of cognitive decline to reduce sample sizes and enhance effectiveness.

Validating the accuracy of deep learning for the diagnosis of pneumonia on chest x-ray against a robust multimodal reference diagnosis: a post hoc analysis of two prospective studies.

BACKGROUND: Artificial intelligence (AI) seems promising in diagnosing pneumonia on chest x-rays (CXR), but deep learning (DL) algorithms have primarily been compared with radiologists, whose diagnosis can be not completely accurate. Therefore, we evaluated the accuracy of DL in diagnosing pneumonia on CXR using a more robust reference diagnosis. METHODS: We trained a DL convolutional neural network model to diagnose pneumonia and evaluated its accuracy in two prospective pneumonia cohorts including 430 patients, for whom the reference diagnosis was determined a posteriori by a multidisciplinary expert panel using multimodal data. The performance of the DL model was compared with that of senior radiologists and emergency physicians reviewing CXRs and that of radiologists reviewing computed tomography (CT) performed concomitantly. RESULTS: Radiologists and DL showed a similar accuracy on CXR for both cohorts (p ≥ 0.269): cohort 1, radiologist 1 75.5% (95% confidence interval 69.1-80.9), radiologist 2 71.0% (64.4-76.8), DL 71.0% (64.4-76.8); cohort 2, radiologist 70.9% (64.7-76.4), DL 72.6% (66.5-78.0). The accuracy of radiologists and DL was significantly higher (p ≤ 0.022) than that of emergency physicians (cohort 1 64.0% [57.1-70.3], cohort 2 63.0% [55.6-69.0]). Accuracy was significantly higher for CT (cohort 1 79.0% [72.8-84.1], cohort 2 89.6% [84.9-92.9]) than for CXR readers including radiologists, clinicians, and DL (all p-values < 0.001). CONCLUSIONS: When compared with a robust reference diagnosis, the performance of AI models to identify pneumonia on CXRs was inferior than previously reported but similar to that of radiologists and better than that of emergency physicians. RELEVANCE STATEMENT: The clinical relevance of AI models for pneumonia diagnosis may have been overestimated. AI models should be benchmarked against robust reference multimodal diagnosis to avoid overestimating its performance. TRIAL REGISTRATION: NCT02467192 , and NCT01574066 . KEY POINT: • We evaluated an openly-access convolutional neural network (CNN) model to diagnose pneumonia on CXRs. • CNN was validated against a strong multimodal reference diagnosis. • In our study, the CNN performance (area under the receiver operating characteristics curve 0.74) was lower than that previously reported when validated against radiologists' diagnosis (0.99 in a recent meta-analysis). • The CNN performance was significantly higher than emergency physicians' (p ≤ 0.022) and comparable to that of board-certified radiologists (p ≥ 0.269).

Head-to-head study of diagnostic accuracy of plasma and cerebrospinal fluid p-tau217 versus p-tau181 and p-tau231 in a memory clinic cohort.

BACKGROUND AND OBJECTIVE: Phosphorylated tau (p-tau) 217 has recently received attention because it seems more reliable than other p-tau variants for identifying Alzheimer's disease (AD) pathology. Thus, we aimed to compare the diagnostic accuracy of plasma and CSF p-tau217 with p-tau181 and p-tau231 in a memory clinic cohort. METHODS: The study included 114 participants (CU = 33; MCI = 67; Dementia = 14). The p-tau variants were correlated versus continuous measures of amyloid (A) and tau (T)-PET. The p-tau phospho-epitopes were assessed through: (i) effect sizes (δ) between diagnostic and A ± and T ± groups; (ii) receiver operating characteristic (ROC) analyses in A-PET and T-PET. RESULTS: The correlations between both plasma and CSF p-tau217 with A-PET and T-PET (r range 0.64-0.83) were stronger than those of p-tau181 (r range 0.44-0.79) and p-tau231 (r range 0.46-0.76). Plasma p-tau217 showed significantly higher diagnostic accuracy than p-tau181 and p-tau231 in (i) differences between diagnostic and biomarker groups (δrange: p-tau217 = 0.55-0.96; p-tau181 = 0.51-0.67; p-tau231 = 0.53-0.71); (ii) ROC curves to identify A-PET and T-PET positivity (AUCaverage: p-tau217 = 0.96; p-tau181 = 0.76; p-tau231 = 0.79). On the other hand, CSF p-tau217 (AUCaverage = 0.95) did not reveal significant differences in A-PET and T-PET AUC than p-tau181 (AUCaverage = 0.88) and p-tau231 (AUCaverage = 0.89). DISCUSSION: Plasma p-tau217 demonstrated better performance in the identification of AD pathology and clinical phenotypes in comparison with other variants of p-tau in a memory clinic cohort. Furthermore, p-tau217 had comparable performance in plasma and CSF. Our findings suggest the potential of plasma p-tau217 in the diagnosis and screening for AD, which could allow for a decreased use of invasive biomarkers in the future.