Search

INTRODUCTION: Understanding longitudinal plasma biomarker trajectories relative to brain amyloid changes can help devise Alzheimer's progression assessment strategies. METHODS: We examined the temporal order of changes in plasma amyloid‐β ratio (Aβ42/Aβ40${{\rm A}\beta }_{42}/{{\rm A}\beta }_{40}$), glial fibrillary acidic protein (GFAP), neurofilament light chain (NfL), and phosphorylated tau ratios (p-tau181/Aβ42$\text{p-tau181}/\mathrm{A}{\beta}_{42}$, p-tau231/Aβ42$\text{p-tau231}/\mathrm{A}{\beta}_{42}$) relative to 11C‐Pittsburgh compound B (PiB) positron emission tomography (PET) cortical amyloid burden (PiB−/+). Participants (n = 199) were cognitively normal at index visit with a median 6.1‐year follow‐up. RESULTS: PiB groups exhibited different rates of longitudinal change in Aβ42/Aβ40(β=5.41×10−4,SE=1.95×10−4,p=0.0073)${{\rm A}\beta }_{42}/{{\rm A}\beta }_{40}\ ({\beta \ = \ 5.41 \times {{10}}^{ - 4},{\rm{\ SE\ }} = \ 1.95 \times {{10}}^{ - 4},\ p\ = \ 0.0073})$. Change in brain amyloid correlated with change in GFAP (r = 0.5, 95% CI = [0.26, 0.68]). The greatest relative decline in Aβ42/Aβ40${{\rm A}\beta }_{42}/{{\rm A}\beta }_{40}$ (−1%/year) preceded brain amyloid positivity by 41 years (95% CI = [32, 53]). DISCUSSION: Plasma Aβ42/Aβ40${{\rm A}\beta }_{42}/{{\rm A}\beta }_{40}$ may begin declining decades prior to brain amyloid accumulation, whereas p‐tau ratios, GFAP, and NfL increase closer in time. HIGHLIGHTS: Plasma Aβ42/Aβ40${{\rm A}\beta }_{42}/{{\rm A}\beta }_{40}$ declines over time among PiB− but does not change among PiB+.Phosphorylated‐tau to Aβ42 ratios increase over time among PiB+ but do not change among PiB−.Rate of change in brain amyloid is correlated with change in GFAP and neurofilament light chain.The greatest decline in Aβ42/Aβ40${{\rm A}\beta }_{42}/{{\rm A}\beta }_{40}$ may precede brain amyloid positivity by decades. [ABSTRACT FROM AUTHOR]

Background: New methods have recently been developed to model longitudinal amyloid biomarkers in AD. This work compares three methods in three different cohorts, and investigates associations of APOE‐e4 status with amyloid onset age and amyloid accumulation rates. Method: Participants from ADNI (N=1213), BLSA (N=205) and WRAP (N=272) with available amyloid PET imaging and APOE genotype were included in the study (Table 1). Amyloid (A) burden was quantified using established methods (ADNI: BAI processed SUVR, WM reference region; BLSA: SRTM‐LRSC DVR, PVC, cerebellum GM reference region; WRAP: graphical analysis DVR, cerebellum GM reference region). Group‐based trajectory modeling (GBTM), ordinary differential equation‐Gaussian process (ODE‐GP) and sampled iterative local linear approximation (SILLA) methods were used to model longitudinal amyloid accumulation, estimate A+ onset age, and align amyloid values by years A+ for each cohort. Forward and backward SUVR/DVR prediction, estimated age of A+ onset, and associations of DVR/SUVR residuals with age, time from reference scan, and SUVR/DVR, were compared for each model and cohort. The impact of APOE genotype on A+ onset and amyloid PET accumulation trajectories was investigated using discrete rate sample, Kaplan‐Meier curves, and accelerated failure time models. Result: All methods produced similar amyloid vs. years A+ curves for each cohort (Figure 1) with most notable discrepancies occurring in the A‐ range. DVR/SUVR prediction performance was similar across methods and was better for retrospective vs. prospective prediction. Amyloid accumulation rates and trajectories were similar for APOE‐e4+ and APOE‐e4‐ genotypes (Figure 2). Kaplan‐Meier curves were different between the APOE groups (log‐rank test, p<0.005 for each test; Figure 3), with the e4+ group exhibiting earlier A+ onset ages compared to e4‐ (Table 2) for all methods and cohorts. Conclusion: This work demonstrates similar performance of three novel methods for modeling longitudinal amyloid PET burden and estimating individualized A+ onset age. Results suggest that APOE‐e4 status affects the A+ onset age, but not the rate of amyloid PET accumulation. Ongoing work is investigating potential differences in amyloid accumulation trajectories and A+ onset age between APOE genotypes (e.g. e3e3, e3e4, e4e4,...) and is investigating relationships between A+ onset, APOE genotype and onset of cognitive impairment. [ABSTRACT FROM AUTHOR]

Background: The analysis of the longitudinal progression of plasma biomarkers and amyloid imaging might lead to new insights in the early pattern of biomarker changes. Method: Plasma Aβ42/Aβ40, pTau231/Aβ42, pTau181/Aβ42, glial fibrillary acidic protein (GFAP), and neurofilament light chain (NFL) were measured for n = 223 subjects and 1673 total visits in the Wisconsin Registry for Alzheimer's Prevention (WRAP) using Quanterix Simoa assays at the University of Gothenburg. Global distribution volume ratio (DVR) was computed from 11C‐Pittsburgh compound B PET scans. Inclusion criteria included Amyloid or CSF Amyloid positive status, and at least one plasma serial value. Trajectories spanning age 46.9 to 80.3 years old for the DVR and plasma biomarkers were estimated by fitting a vector field to the longitudinal data (https://arxiv.org/abs/2206.15215). A Louvain community clustering analysis was performed. The time from Amyloid positivity (DVR = 1.19; centiloid of 22) was also computed for all individual trajectories that crossed this threshold. Result: The Louvain community clustering algorithm resulted in three distinct communities (Fig 1): no increase in DVR or plasma biomarkers (Green, n = 88(39.5%)); increase in DVR and all plasma biomarkers except NFL (Red, 41(18.4%)); and mixed DVR plus GFAP and NFL increase (Blue, 94(42.2%)). Time since Amyloid positivity (TSAP) was computed for the subset of subjects (n = 84) expected to cross the Amyloid positivity threshold. The Louvain communities as a function of TSAP are shown in Fig 2. Conclusion: We identified three communities regarding the longitudinal progression of amyloid PET and plasma biomarker levels. Plasma biomarkers were dynamic up to 30 years before Amyloid positivity. [ABSTRACT FROM AUTHOR]

Background: High‐frequency biomarkers (HFB) are measurements collected weekly, on a daily basis or even more frequently. They are designed to detect early progression from one cognitive state to a later stage. Examples include home‐based measurements, e.g., computer usage and sleep patterns. In randomized controlled trials (RCT), we aim to detect the difference in trajectories of HFB between placebo and experimental group. Current statistical analysis e.g. linear mixed effect with random intercept models detect changes in the mean trajectory but fail to detect changes in the covariance structure, thus requiring longer trials. We examined whether more careful modeling, using historical data, would reduce the sample size needed for a RCT? Objective: (1) to present a novel statistical methodology for analyzing RCTs whose outcomes are HFB based on historical data of cognitively normal subjects, and (2) to evaluate this methodology using simulated data first, then using data from the Oregon Center for Aging and Technology (ORCATECH). Methods: The statistical methodology proposed proceeds in three phases (Figure 1). Highlights include the use of historical data for cognitively normal subjects and the use of a Gaussian process with parameterized mean function and covariance structure. Examples of results from simulated data are shown in Figure 2, demonstrating the superior power of the methodology compared to the linear mixed‐effects model. ORCATECH's weekly computer usage data was used (Figure 3). We computed an RCT effect size with 86 cognitively normal subjects and 11 subjects who were cognitively normal at baseline, but would eventually develop MCI within 139 weeks on average (sd=72). Results: Using this new approach, we found an effect size of W=0.616 (Area Under the curve), outperforming a baseline method which yielded an effect size of W=0.515. Respective total sample sizes for an RCT with alpha=0.05 and power=0.8 are 74 and 1028, for the new versus the conventional approach, respectively. Conclusions: We have shown that a careful statistical procedure allows for maximally benefiting from HFB by reducing the sample size needed for a 100 week RCT. More experiments will be necessary to validate these findings. [ABSTRACT FROM AUTHOR]

Background: Accumulation of beta‐amyloid plaques, an early hallmark neuropathology of Alzheimer's disease (AD), begin more than a decade prior to the onset of clinical symptoms. Understanding the temporal progression of early amyloid accumulation is important for predicting who will become amyloid‐positive and when. Based on observations from several longitudinal amyloid positron emission tomography (PET) studies that the rate of overall amyloid burden is closely associated with baseline level of amyloid, we develop a formal mathematical model capturing this phenomenon and use it to (i) align temporal trajectories of amyloid accumulation across individuals, (ii) create a fine‐grained timeline of amyloid accumulation, and (iii) estimate the age at onset of amyloid positivity per individual. Method: We used (PiB) PET scans from the Wisconsin Registry for Alzheimer's Prevention (WRAP) and the Baltimore Longitudinal Study of Aging (BLSA) to assess global amyloid burden cDVR, see Figure 1. A first‐order autonomous model for the slope of cDVR as a function of cDVR was fit to the data. We used Bayesian modeling with a Gaussian process (GP) prior and independent homoscedastic noise. Result: The best fit was obtained with a quadratic GP kernel (Figure 2). Since the system is autonomous, the trajectories are parallel and thus can be registered by translation. The progression of the Amyloid burden is thus visualized as a function of the time (in years) from PIB positivity in Figure 3. The estimated age when the PIB cDVR becomes positive is estimated. This age depends on APOE status, with a significant difference between the APOE groups [3,3] and [3,4]. p‐value=.008 in WRAP, p‐value=.054 In BLSA. Conclusion: Large studies on aging allow studying the time‐progression of the amyloid burden. We found that an autonomous ordinary differential equation allows for a good explanatory model of the PIB‐DVR data in the WRAP as well as in the BLSA studies. Still, a few participants do not fit the data closely. It remains to understand if this is due to noisy measurements or to specific conditions. Also, alternative dynamical systems, including Hamiltonian, still need to be investigated and compared. [ABSTRACT FROM AUTHOR]

This paper describes a feasibility study of a method for delineating the tentorium cerebelli in magnetic esonance imaging (MRI) brain scans. The tentorium cerebelli is a thin sheet of dura matter covering the cerebellum and separating it from the posterior part of the temporal lobe and the occipital lobe of the cerebral hemispheres. Cortical structures such as the parahippocampal gyrus can be indistinguishable from tentorium in magnetized prepared rapid gradient echo and T1-weighted MRI scans. Similar intensities in these neighboring regions make it difficult to perform accurate cortical analysis in neuroimaging studies of schizophrenia and Alzheimer's disease. A semi-automated, geometric, intensitybased procedure for delineating the tentorium from a whole-brain scan is described. Initial and final curves are traced within the tentorium. A cost function, based on intensity and Euclidean distance, is computed between the two curves using the Fast Marching method. The initial curve is then evolved to the final curve based on the gradient of the computed costs, generating a series of intermediate curves. These curves are then used to generate a triangulated surface of the tentorium. For 3 scans, surfaces were found to be within 2 voxels from hand segmentations. [ABSTRACT FROM AUTHOR]

Background The delineation of the relative temporal trajectories of specific cognitive measures associated with Alzheimer's disease (AD) is important for evaluating preclinical markers and monitoring disease progression. Methods We characterized the temporal trajectories of measures of verbal episodic memory, short-term visual memory, and mental status using data from 895 participants in the Baltimore Longitudinal Study of Aging. Results The California Verbal Learning Test (CVLT) immediate recall was the first measure to decline, followed by CVLT delayed recall. However, further along the disease progression scale, CVLT delayed recall and visual memory changed more rapidly than CVLT immediate recall. Conclusions Our findings reconcile reports of early changes in immediate recall with greater reliance on delayed recall performance in clinical settings. Moreover, the utility of cognitive markers in evaluating AD progression depends on the stage of cognitive decline, suggesting that optimal endpoints in therapeutic trials may vary across different stages of the disease process. [ABSTRACT FROM AUTHOR]