Constructing longitudinal disease progression curves using sparse, short-term individual data with an application to Alzheimer's disease

Disease progression modeling with temporal realignment: An emerging approach to deepen knowledge on chronic diseases

The recent development of the first disease-modifying drug for Alzheimer's disease represents a major advancement in dementia treatment. Behind this breakthrough is a quarter century of research efforts to understand the disease not by a particular symptom at a given moment, but by long-term sequential changes in multiple biomarkers. Disease progression modeling with temporal realignment (DPM-TR) is an emerging computational approach proposed with this biomarker-based disease concept. By integrating short-term clinical observations of multiple disease biomarkers in a data-driven manner, DPM-TR provides a way to understand the progression of chronic diseases over decades and predict individual disease stages more accurately. DPM-TR has been developed primarily in the area of neurodegenerative diseases but has recently been extended to non-neurodegenerative diseases, including chronic obstructive pulmonary, autoimmune, and ophthalmologic diseases. This review focuses on opportunities for DPM-TR in clinical practice and drug development and discusses its current status and challenges.

Timing of Biomarker Changes in Sporadic Alzheimer's Disease in Estimated Years from Symptom Onset

OBJECTIVE

A clock relating amyloid positron emission tomography (PET) to time was used to estimate the timing of biomarker changes in sporadic Alzheimer disease (AD).

METHODS

Research participants were included who underwent cerebrospinal fluid (CSF) collection within 2 years of amyloid PET. The ages at amyloid onset and AD symptom onset were estimated for each individual. The timing of change for plasma, CSF, imaging, and cognitive measures was calculated by comparing restricted cubic splines of cross-sectional data from the amyloid PET positive and negative groups.

RESULTS

The amyloid PET positive sub-cohort (n = 118) had an average age of 70.4 ± 7.4 years (mean ± standard deviation) and 16% were cognitively impaired. The amyloid PET negative sub-cohort (n = 277) included individuals with low levels of amyloid plaque burden at all scans who were cognitively unimpaired at the time of the scans. Biomarker changes were detected 15-19 years before estimated symptom onset for CSF Aβ42/Aβ40, plasma Aβ42/Aβ40, CSF pT217/T217, and amyloid PET; 12-14 years before estimated symptom onset for plasma pT217/T217, CSF neurogranin, CSF SNAP-25, CSF sTREM2, plasma GFAP, and plasma NfL; and 7-9 years before estimated symptom onset for CSF pT205/T205, CSF YKL-40, hippocampal volumes, and cognitive measures.

INTERPRETATION

The use of an amyloid clock enabled visualization and analysis of biomarker changes as a function of estimated years from symptom onset in sporadic AD. This study demonstrates that estimated years from symptom onset based on an amyloid clock can be used as a continuous staging measure for sporadic AD and aligns with findings in autosomal dominant AD. ANN NEUROL 2024;95:951-965.

Novel avenues of tau research

INTRODUCTION

The pace of innovation has accelerated in virtually every area of tau research in just the past few years.

METHODS

In February 2022, leading international tau experts convened to share selected highlights of this work during Tau 2022, the second international tau conference co-organized and co-sponsored by the Alzheimer's Association, CurePSP, and the Rainwater Charitable Foundation.

RESULTS

Representing academia, industry, and the philanthropic sector, presenters joined more than 1700 registered attendees from 59 countries, spanning six continents, to share recent advances and exciting new directions in tau research.

DISCUSSION

The virtual meeting provided an opportunity to foster cross-sector collaboration and partnerships as well as a forum for updating colleagues on research-advancing tools and programs that are steadily moving the field forward.

Data-driven modelling of neurodegenerative disease progression: thinking outside the black box

Data-driven disease progression models are an emerging set of computational tools that reconstruct disease timelines for long-term chronic diseases, providing unique insights into disease processes and their underlying mechanisms. Such methods combine a priori human knowledge and assumptions with large-scale data processing and parameter estimation to infer long-term disease trajectories from short-term data. In contrast to 'black box' machine learning tools, data-driven disease progression models typically require fewer data and are inherently interpretable, thereby aiding disease understanding in addition to enabling classification, prediction and stratification. In this Review, we place the current landscape of data-driven disease progression models in a general framework and discuss their enhanced utility for constructing a disease timeline compared with wider machine learning tools that construct static disease profiles. We review the insights they have enabled across multiple neurodegenerative diseases, notably Alzheimer disease, for applications such as determining temporal trajectories of disease biomarkers, testing hypotheses about disease mechanisms and uncovering disease subtypes. We outline key areas for technological development and translation to a broader range of neuroscience and non-neuroscience applications. Finally, we discuss potential pathways and barriers to integrating disease progression models into clinical practice and trial settings.

A multidimensional ODE-based model of Alzheimer's disease progression

Data-driven Alzheimer's disease (AD) progression models are useful for clinical prediction, disease mechanism understanding, and clinical trial design. Most dynamic models were inspired by the amyloid cascade hypothesis and described AD progression as a linear chain of pathological events. However, the heterogeneity observed in healthy and sporadic AD populations challenged the amyloid hypothesis, and there is a need for more flexible dynamical models that accompany this conceptual shift. We present a statistical model of the temporal evolution of biomarkers and cognitive tests that allows diverse biomarker paths throughout the disease. The model consists of two elements: a multivariate dynamic model of the joint evolution of biomarkers and cognitive tests; and a clinical prediction model. The dynamic model uses a system of ordinary differential equations to jointly model the rate of change of an individual's biomarkers and cognitive tests. The clinical prediction model is an ordinal logistic model of the diagnostic label. Prognosis and time-to-onset predictions are obtained by computing the clinical label probabilities throughout the forecasted biomarker trajectories. The proposed dynamical model is interpretable, free of one-dimensional progression hypotheses or disease staging paradigms, and can account for the heterogeneous dynamics observed in sporadic AD. We developed the model using longitudinal data from the Alzheimer's Disease Neuroimaging Initiative. We illustrate the patterns of biomarker rates of change and the model performance to predict the time to conversion from MCI to dementia.

Rates of regional tau accumulation in ageing and across the Alzheimer's disease continuum: an AIBL 18F-MK6240 PET study

BACKGROUND

Tau positron emission tomography (PET) imaging enables longitudinal observation of tau accumulation in Alzheimer's disease (AD). 18F-MK6240 is a high affinity tracer for the paired helical filaments of tau in AD, widely used in clinical trials, despite sparse longitudinal natural history data. We aimed to evaluate the natural history of tau accumulation, and the impact of disease stage and reference region on the magnitude and effect size of regional change.

METHODS

One hundred and eighty-four participants: 89 cognitively unimpaired (CU) beta-amyloid negative (Aβ-), 44 CU Aβ+, 51 cognitively impaired Aβ+ (26 with mild cognitive impairment [MCI] and 25 with dementia) had follow-up 18F-MK6240 PET for one to four years (median 1.48). Regional standardised uptake value ratios (SUVR) were generated. Two reference regions were examined: cerebellar cortex and eroded subcortical white matter.

FINDINGS

CU Aβ- participants had very low rates of tau accumulation in the mesial temporal lobe (MTL). In CU Aβ+, significantly higher rate of accumulation was seen in the MTL (particularly the amygdala), extending into the inferior temporal lobes. In MCI Aβ+, the rate of accumulation was greatest in the lateral temporal, parietal and lateral occipital cortex, and plateaued in the MTL. Accumulation was global in AD Aβ+, except for a plateau in the MTL. The eroded subcortical white matter reference region showed no significant advantage over the cerebellar cortex and appeared prone to spill-over in AD participants. Data fitting suggested approximately 15-20 years to accumulate tau to typical AD levels.

INTERPRETATION

Tau accumulation occurs slowly. Rates vary according to brain region, disease stage and tend to plateau at high levels. Rates of tau accumulation are best measured in the MTL and inferior temporal cortex in preclinical AD and in large neocortical areas, in MCI and AD.

FUNDING

NHMRC; Cerveau Technologies.

Estimating the longitudinal trajectory of cognitive function measurement using short-term data with different disease stages: Application in Alzheimer's disease

Alzheimer's disease (AD) is a chronic neurodegenerative disease characterized by a gradual decline in cognitive function over a few decades. The Mini-Mental State Examination (MMSE) is a widely used measure for evaluating global cognitive functioning. Characterizing the longitudinal trajectory of the MMSE in the population of interest is important to detect AD onset for preventive intervention. In this study, we formulate a new class of longitudinal trajectory modeling for MMSE from short-term individual data based on an ordinary differential equation. The proposed method models the relationship between individual decline speed of MMSE and the average MMSE using the fractional polynomial function model and subsequently estimates the longitudinal trajectory of MMSE by solving the ordinary differential equation for the estimated model. The appropriate model for trajectory estimation is selected based on the proposed criterion for quantifying the goodness of trajectory fit. The accuracy of the trajectory estimation of the proposed method was demonstrated via simulation studies. The proposed method was successfully applied to MMSE data from the Japanese Alzheimer's Disease Neuroimaging Initiative study.

Assessment of a polygenic hazard score for the onset of pre-clinical Alzheimer's disease

Abstract

Background

With a growing number of loci associated with late-onset (sporadic) Alzheimer’s disease (AD), the polygenic contribution to AD is now well established. The development of polygenic risk score approaches have shown promising results for identifying individuals at higher risk of developing AD, thereby facilitating the development of preventative and therapeutic strategies. A polygenic hazard score (PHS) has been proposed to quantify age-specific genetic risk for AD. In this study, we assessed the predictive power and transferability of this PHS in an independent cohort, to support its clinical utility.

Results

Using genotype and imaging data from 780 individuals enrolled in the Australian Imaging, Biomarkers and Lifestyle (AIBL) study, we investigated associations between the PHS and several AD-related traits, including 1) cross-sectional Aβ-amyloid (Aβ) deposition, 2) longitudinal brain atrophy, 3) longitudinal cognitive decline, 4) age of onset. Except in the cognitive domain, we obtained results that were consistent with previously published findings. The PHS was associated with increased Aβ burden, faster regional brain atrophy and an earlier age of onset.

Conclusion

Overall, the results support the predictive power of a PHS, however, with only marginal improvement compared to apolipoprotein E alone.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-022-08617-2.

Predicting Symptom Onset in Sporadic Alzheimer's Disease: "How Long Do I Have?"

Predicting not just if but when cognitively normal individuals will develop the onset of Alzheimer's disease (AD) dementia seems increasingly feasible, as evidenced by converging findings from several approaches and cohorts. These estimates may improve the efficiency of clinical trials by better identifying cognitively normal individuals at high risk of developing AD symptoms. As models are refined, the implications of disclosing estimates of the age of AD symptom onset must be examined, since telling a cognitively normal individual the age they are expected to develop AD symptoms may have different implications than disclosing increased risk for AD dementia.

Predicting Symptom Onset in Sporadic Alzheimer Disease With Amyloid PET

BACKGROUND AND OBJECTIVES

To predict when cognitively normal individuals with brain amyloidosis will develop symptoms of Alzheimer disease (AD).

METHODS

Brain amyloid burden was measured by amyloid PET with Pittsburgh compound B. The mean cortical standardized uptake value ratio (SUVR) was transformed into a timescale with the use of longitudinal data.

RESULTS

Amyloid accumulation was evaluated in 236 individuals who underwent >1 amyloid PET scan. The average age was 66.5 ± 9.2 years, and 12 individuals (5%) had cognitive impairment at their baseline amyloid PET scan. A tipping point in amyloid accumulation was identified at a low level of amyloid burden (SUVR 1.2), after which nearly all individuals accumulated amyloid at a relatively consistent rate until reaching a high level of amyloid burden (SUVR 3.0). The average time between levels of amyloid burden was used to estimate the age at which an individual reached SUVR 1.2. Longitudinal clinical diagnoses for 180 individuals were aligned by the estimated age at SUVR 1.2. In the 22 individuals who progressed from cognitively normal to a typical AD dementia syndrome, the estimated age at which an individual reached SUVR 1.2 predicted the age at symptom onset (R ² = 0.54, p < 0.0001, root mean square error [RMSE] 4.5 years); the model was more accurate after exclusion of 3 likely misdiagnoses (R ² = 0.84, p < 0.0001, RMSE 2.8 years).

CONCLUSION

The age at symptom onset in sporadic AD is strongly correlated with the age at which an individual reaches a tipping point in amyloid accumulation.

Time course of phosphorylated-tau181 in blood across the Alzheimer's disease spectrum

Tau phosphorylated at threonine 181 (p-tau181) measured in blood plasma has recently been proposed as an accessible, scalable, and highly specific biomarker for Alzheimer’s disease. Longitudinal studies, however, investigating the temporal dynamics of this novel biomarker are lacking. It is therefore unclear when in the disease process plasma p-tau181 increases above physiological levels and how it relates to the spatiotemporal progression of Alzheimer’s disease characteristic pathologies. We aimed to establish the natural time course of plasma p-tau181 across the sporadic Alzheimer’s disease spectrum in comparison to those of established imaging and fluid-derived biomarkers of Alzheimer’s disease. We examined longitudinal data from a large prospective cohort of elderly individuals enrolled in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (n = 1067) covering a wide clinical spectrum from normal cognition to dementia, and with measures of plasma p-tau181 and an ¹⁸F-florbetapir amyloid-β PET scan at baseline. A subset of participants (n = 864) also had measures of amyloid-β_1–42 and p-tau181 levels in CSF, and another subset (n = 298) had undergone an ¹⁸F-flortaucipir tau PET scan 6 years later. We performed brain-wide analyses to investigate the associations of plasma p-tau181 baseline levels and longitudinal change with progression of regional amyloid-β pathology and tau burden 6 years later, and estimated the time course of changes in plasma p-tau181 and other Alzheimer’s disease biomarkers using a previously developed method for the construction of long-term biomarker temporal trajectories using shorter-term longitudinal data. Smoothing splines demonstrated that earliest plasma p-tau181 changes occurred even before amyloid-β markers reached abnormal levels, with greater rates of change correlating with increased amyloid-β pathology. Voxel-wise PET analyses yielded relatively weak, yet significant, associations of plasma p-tau181 with amyloid-β pathology in early accumulating brain regions in cognitively healthy individuals, while the strongest associations with amyloid-β were observed in late accumulating regions in patients with mild cognitive impairment. Cross-sectional and particularly longitudinal measures of plasma p-tau181 were associated with widespread cortical tau aggregation 6 years later, covering temporoparietal regions typical for neurofibrillary tangle distribution in Alzheimer’s disease. Finally, we estimated that plasma p-tau181 reaches abnormal levels ∼6.5 and 5.7 years after CSF and PET measures of amyloid-β, respectively, following similar dynamics as CSF p-tau181. Our findings suggest that plasma p-tau181 increases are associated with the presence of widespread cortical amyloid-β pathology and with prospective Alzheimer’s disease typical tau aggregation, providing clear implications for the use of this novel blood biomarker as a diagnostic and screening tool for Alzheimer’s disease.

Association of β-Amyloid Level, Clinical Progression, and Longitudinal Cognitive Change in Normal Older Individuals

OBJECTIVE

To determine the effect of β-amyloid (Aβ) level on progression risk to mild cognitive impairment (MCI) or dementia and longitudinal cognitive change in cognitively normal (CN) older individuals.

METHODS

All CN from the Australian Imaging Biomarkers and Lifestyle study with Aβ PET and ≥3 years follow-up were included (n = 534; age 72 ± 6 years; 27% Aβ positive; follow-up 5.3 ± 1.7 years). Aβ level was divided using the standardized 0-100 Centiloid scale: <15 CL negative, 15-25 CL uncertain, 26-50 CL moderate, 51-100 CL high, >100 CL very high, noting >25 CL approximates a positive scan. Cox proportional hazards analysis and linear mixed effect models were used to assess risk of progression and cognitive decline.

RESULTS

Aβ levels in 63% were negative, 10% uncertain, 10% moderate, 14% high, and 3% very high. Fifty-seven (11%) progressed to MCI or dementia. Compared to negative Aβ, the hazard ratio for progression for moderate Aβ was 3.2 (95% confidence interval [CI] 1.3-7.6; p < 0.05), for high was 7.0 (95% CI 3.7-13.3; p < 0.001), and for very high was 11.4 (95% CI 5.1-25.8; p < 0.001). Decline in cognitive composite score was minimal in the moderate group (-0.02 SD/year, p = 0.05), while the high and very high declined substantially (high -0.08 SD/year, p < 0.001; very high -0.35 SD/year, p < 0.001).

CONCLUSION

The risk of MCI or dementia over 5 years in older CN is related to Aβ level on PET, 5% if negative vs 25% if positive but ranging from 12% if 26-50 CL to 28% if 51-100 CL and 50% if >100 CL. This information may be useful for dementia risk counseling and aid design of preclinical AD trials.

Impact of APOE-ε4 carriage on the onset and rates of neocortical Aβ-amyloid deposition

Neocortical Aβ-amyloid deposition, one of the hallmark pathologic features of Alzheimer's disease (AD), begins decades prior to the presence of clinical symptoms. As clinical trials move to secondary and even primary prevention, understanding the rates of neocortical Aβ-amyloid deposition and the age at which Aβ-amyloid deposition becomes abnormal is crucial for optimizing the timing of these trials. As APOE-ε4 carriage is thought to modulate the age of clinical onset, it is also important to understand the impact of APOE-ε4 carriage on the age at which the neocortical Aβ-amyloid deposition becomes abnormal. Here, we show that, for 455 participants with over 3 years of follow-up, abnormal levels of neocortical Aβ-amyloid were reached on average at age 72 (66.5-77.1). The APOE-ε4 carriers reached abnormal levels earlier at age 63 (59.6-70.3); however, noncarriers reached the threshold later at age 78 (76.1-84.4). No differences in the rates of deposition were observed between APOE-ε4 carriers and noncarriers after abnormal Aβ-amyloid levels had been reached. These results suggest that primary and secondary prevention trials, looking to recruit at the earliest stages of disease, should target APOE-ε4 carriers between the ages of 60 and 66 and noncarriers between the ages of 76 and 84.

Longitudinal evaluation of the natural history of amyloid-β in plasma and brain

Plasma amyloid-β peptide concentration has recently been shown to have high accuracy to predict amyloid-β plaque burden in the brain. These amyloid-β plasma markers will allow wider screening of the population and simplify and reduce screening costs for therapeutic trials in Alzheimer's disease. The aim of this study was to determine how longitudinal changes in blood amyloid-β track with changes in brain amyloid-β. Australian Imaging, Biomarker and Lifestyle study participants with a minimum of two assessments were evaluated (111 cognitively normal, 7 mild cognitively impaired, 15 participants with Alzheimer's disease). Amyloid-β burden in the brain was evaluated through PET and was expressed in Centiloids. Total protein amyloid-β 42/40 plasma ratios were determined using ABtest® assays. We applied our method for obtaining natural history trajectories from short term data to measures of total protein amyloid-β 42/40 plasma ratios and PET amyloid-β. The natural history trajectory of total protein amyloid-β 42/40 plasma ratios appears to approximately mirror that of PET amyloid-β, with both spanning decades. Rates of change of 7.9% and 8.8%, were observed for total protein amyloid-β 42/40 plasma ratios and PET amyloid-β, respectively. The trajectory of plasma amyloid-β preceded that of brain amyloid-β by a median value of 6 years (significant at 88% confidence interval). These findings, showing the tight association between changes in plasma and brain amyloid-β, support the use of plasma total protein amyloid-β 42/40 plasma ratios as a surrogate marker of brain amyloid-β. Also, that plasma total protein amyloid-β 42/40 plasma ratios has potential utility in monitoring trial participants, and as an outcome measure.

Predict Disease Progression With Reaction Rate Equation Modeling of Multimodal MRI and PET

Neurodegenerative dementia often has multiple types of underlying pathology, for example, beta-amyloid, misfolded tau, chronic neuroinflammation and neurodegeneration may coexist in Alzheimer’s disease. However, the relationship between them is often unclear, in other words, whether one pathology is upstream or downstream of others can be very difficult to investigate directly. This is partly because the underlying pathology in dementia may precede detectable symptoms by several years if not decades. The time scale associated with disease progression in dementia generally exceeds that in conventional longitudinal imaging studies in humans, so it is difficult to directly observe the temporal ordering of different pathologies. Also, animal studies are not always transferable to patients due to obvious differences between the two systems. To investigate the disease progression and relationships among underlying pathological changes, we propose a novel computational modeling approach for multimodal MRI and PET inspired by reaction rate equation in chemical kinetics. We also discuss the possibility and prerequisites to use cross-sectional data to generate preliminary hypothesis for future longitudinal studies. It has been shown that the rate of change in some biomarkers can be approximated by the average trajectory across patients at different stages of disease severity in cross-sectional studies. The relationship modeled in our approach is akin to that in the control theory, and can be assessed by demonstrating that the presence of one disease related biomarker predicts dynamics in another. We argue that the proposed framework has important implications for trials targeting different pathologies in dementia.