Design and sample size considerations for Alzheimer's disease prevention trials using multistate models

Deep learning-based patient stratification for prognostic enrichment of clinical dementia trials

Dementia probably due to Alzheimer's disease is a progressive condition that manifests in cognitive decline and impairs patients' daily life. Affected patients show great heterogeneity in their symptomatic progression, which hampers the identification of efficacious treatments in clinical trials. Using artificial intelligence approaches to enable clinical enrichment trials serves a promising avenue to identify treatments. In this work, we used a deep learning method to cluster the multivariate disease trajectories of 283 early dementia patients along cognitive and functional scores. Two distinct subgroups were identified that separated patients into 'slow' and 'fast' progressing individuals. These subgroups were externally validated and independently replicated in a dementia cohort comprising 2779 patients. We trained a machine learning model to predict the progression subgroup of a patient from cross-sectional data at their time of dementia diagnosis. The classifier achieved a prediction performance of 0.70 ± 0.01 area under the receiver operating characteristic curve in external validation. By emulating a hypothetical clinical trial conducting patient enrichment using the proposed classifier, we estimate its potential to decrease the required sample size. Furthermore, we balance the achieved enrichment of the trial cohort against the accompanied demand for increased patient screening. Our results show that enrichment trials targeting cognitive outcomes offer improved chances of trial success and are more than 13% cheaper compared with conventional clinical trials. The resources saved could be redirected to accelerate drug development and expand the search for remedies for cognitive impairment.

Assessing Alzheimer's disease via plasma extracellular vesicle-derived mRNA

INTRODUCTION

Alzheimer's disease (AD), the most prevalent neurodegenerative disorder globally, has emerged as a significant health concern. Recently it has been revealed that extracellular vesicles (EVs) play a critical role in AD pathogenesis and progression. Their stability and presence in various biofluids, such as blood, offer a minimally invasive window for monitoring AD-related changes.

METHODS

We analyzed plasma EV-derived messenger RNA (mRNA) from 82 human subjects, including individuals with AD, mild cognitive impairment (MCI), and healthy controls. With next-generation sequencing, we profiled differentially expressed genes (DEGs), identifying those associated with AD.

RESULTS

Based on DEGs identified in both the MCI and AD groups, a diagnostic model was established based on machine learning, demonstrating an average diagnostic accuracy of over 98% and showed a strong correlation with different AD stages.

DISCUSSION

mRNA derived from plasma EVs shows significant promise as a non-invasive biomarker for the early detection and continuous monitoring of AD.

Highlights

The study conducted next-generation sequencing (NGS) of mRNA derived from human plasma extracellular vesicles (EVs) to assess Alzheimer's disease (AD).Profiling of plasma EV-derived mRNA shows a significantly enriched AD pathway, indicating its potential for AD-related studies.The AD-prediction model achieved a receiver-operating characteristic area under the curve (ROC-AUC) of more than 0.98, with strong correlation to the established Clinical Dementia Rating (CDR).

Comparing Symmetric Dimethylarginine and Amyloid-β42 as Predictors of Alzheimer's Disease Development

Background

Physicians may soon be able to diagnose Alzheimer's disease (AD) in its early stages using fluid biomarkers like amyloid. However, it is acknowledged that additional biomarkers need to be characterized which would facilitate earlier monitoring of AD pathogenesis.

Objective

To determine if a potential novel inflammation biomarker for AD, symmetric dimethylarginine, has utility as a baseline serum biomarker for discriminating prodromal AD from cognitively unimpaired controls in comparison to cerebrospinal fluid amyloid-β42 (Aβ42).

Methods

Data including demographics, magnetic resonance imaging and fluorodeoxyglucose-positron emission tomography scans, Mini-Mental State Examination and Functional Activities Questionnaire scores, and biomarker concentrations were obtained from the Alzheimer's Disease Neuroimaging Initiative for a total of 146 prodromal AD participants and 108 cognitively unimpaired controls.

Results

Aβ42 (p = 0.65) and symmetric dimethylarginine (p = 0.45) were unable to predict age-matched cognitively unimpaired controls and prodromal AD participants. Aβ42 was negatively associated with regional brain atrophy and hypometabolism as well as cognitive and functional decline in cognitively unimpaired control participants (p < 0.05) that generally decreased in time. There were no significant associations between Aβ42 and symmetric dimethylarginine with imaging or neurocognitive biomarkers in prodromal AD patients.

Conclusions

Correlations were smaller between Aβ42 and neuropathological biomarkers over time and were absent in prodromal AD participants, suggesting a plateau effect dependent on age and disease stage. Evidence supporting symmetric dimethylarginine as a novel biomarker for AD as a single measurement was not found.

The theoretical problems of "prodrome" and "phenoconversion" in neurodegeneration

The recognition of and approach to prodromal symptoms, those which manifest before a diagnosis can be ascertained at the bedside, are of increasing interest in neurodegenerative research. A prodrome is conceived of as an early window into a disease, a critical time when putative disease-modifying interventions may be best suited for examination. Several challenges affect research in this area. Prodromal symptoms are highly prevalent in the population, can be nonprogressive for years or decades, and exhibit limited specificity in predicting conversion versus nonconversion into a neurodegenerative category within a time window feasible for most longitudinal clinical studies. In addition, there is a large range of biological alterations subsumed within each prodromal syndrome, forced to converge into the unifying nosology of each neurodegenerative disorder. Initial prodromal subtyping efforts have been developed but given the scarcity of prodrome-to-disease longitudinal studies, it is not yet clear whether any prodromal subtype can be predicted to evolve into the corresponding subtype of manifesting disease - a form of construct validity. As current subtypes generated from one clinical population are not faithfully replicated to others, it is likely that, lacking biological or molecular anchors, prodromal subtypes may only be applicable to the cohorts within which they were developed. Furthermore, as clinical subtypes have not aligned with a consistent pattern of pathology or biology, such might also be the fate of prodromal subtypes. Finally, the threshold defining the change from prodrome to disease for most neurodegenerative disorders remains clinical (e.g., a motor change in gait becoming noticeable to a clinician or measurable with portable technologies), not biological. As such, a prodrome can be viewed as a disease state not yet overt to a clinician. Efforts into identifying biological subtypes of disease, regardless of clinical phenotype or disease stage, may best serve future disease-modifying therapeutic strategies deployed not for a prodromal symptom but for a defined biological derangement as soon as it can be determined to lead to clinical changes, prodromal or not.

Change in the Quick Dementia Rating System Across Time in Older Adults with and without Cognitive Impairment

BACKGROUND

The Quick Dementia Rating System (QDRS) is a brief, informant-reported dementia staging tool that approximates scores on the Clinical Dementia Rating Scale in patients with Alzheimer's disease (AD).

OBJECTIVE

The current study sought to examine change in the QDRS across time, which is necessary for clinical and research efforts.

METHODS

One-hundred ten older adults (intact, mild cognitive impairment [MCI], mild AD, classified with Alzheimer's Disease Neuroimaging Initiative criteria) were rated on the QDRS by an informant and had an amyloid positron emission tomography scan at baseline. The informant re-rated each participant on the QDRS after one year. Dependent t-tests compared the entire sample and various subgroups (e.g., cognitive status, amyloid status) on baseline and follow-up QDRS scores.

RESULTS

In the entire sample, the Total score on the QDRS significantly increased (i.e., worsened) on follow-up (p < 0.001). When subgroups were analyzed, the MCI and mild AD subjects showed increasing (i.e., worsening) QDRS Total scores (both p < 0.001), but the intact subjects remained stable over time (p = 0.28). Additionally, those classified as being amyloid positive at baseline showed significantly increased QDRS Total scores at follow-up (p < 0.001) compared to those who were amyloid negative at baseline, whose QDRS Total scores remained stable over time (p = 0.63).

CONCLUSION

The QDRS can potentially demonstrate worsening functioning status across one year, especially in those who have MCI or mild AD and those who are amyloid positive. Therefore, the current results preliminarily suggest that the QDRS may provide an efficient tool for tracking progression in clinical trials in AD.

Quantitative systems pharmacology model of the amyloid pathway in Alzheimer's disease: Insights into the therapeutic mechanisms of clinical candidates

Despite considerable investment into potential therapeutic approaches for Alzheimer's disease (AD), currently approved treatment options are limited. Predictive modeling using quantitative systems pharmacology (QSP) can be used to guide the design of clinical trials in AD. This study developed a QSP model representing amyloid beta (Aβ) pathophysiology in AD. The model included mechanisms of Aβ monomer production and aggregation to form insoluble fibrils and plaques; the transport of soluble species between the compartments of brain, cerebrospinal fluid (CSF), and plasma; and the pharmacokinetics, transport, and binding of monoclonal antibodies to targets in the three compartments. Ordinary differential equations were used to describe these processes quantitatively. The model components were calibrated to data from the literature and internal studies, including quantitative data supporting the underlying AD biology and clinical data from clinical trials for anti-Aβ monoclonal antibodies (mAbs) aducanumab, crenezumab, gantenerumab, and solanezumab. The model was developed for an apolipoprotein E (APOE) ɛ4 allele carrier and tested for an APOE ɛ4 noncarrier. Results indicate that the model is consistent with data on clinical Aβ accumulation in untreated individuals and those treated with monoclonal antibodies, capturing increases in Aβ load accurately. This model may be used to investigate additional AD mechanisms and their impact on biomarkers, as well as predict Aβ load at different dose levels for mAbs with known targets and binding affinities. This model may facilitate the design of scientifically enriched and efficient clinical trials by enabling a priori prediction of biomarker dynamics in the brain and CSF.

ALZT-OP1: An experimental combination regimen for the treatment of Alzheimer's Disease

INTRODUCTION

For Alzheimer's disease (AD) treatment, US FDA granted accelerated approval for aducanumab due to its amyloid-β (Aβ)-lowering effects, notwithstanding the reported poor correlation between amyloid plaque reduction and clinical change for this drug. The diversification of drug targets appears to be the future of the AD field and from this perspective, drugs modulating microglia dysfunction and combination treatment regimens offer some promise.

AREAS COVERED

The aim of the present article was to provide a comprehensive review of ALZT-OP1 (cromolyn sodium plus ibuprofen), an experimental combination treatment regimen for AD, discussing their mechanisms of action targeting Aβ and neuroinflammation, examining the role of microglia in AD and offering our own insights on the role of present and alternative approaches directed toward neuroinflammation.

EXPERT OPINION

Enrolling high-risk participants with elevated brain amyloid could help to slow cognitive decline in secondary prevention trials during AD preclinical stages. Long-term follow-up indicated that non-steroidal anti-inflammatory drugs use begun when the brain was still normal may benefit these patients, suggesting that the timing of therapy could be crucial. However, previous clinical failures and the present incomplete understanding of the Aβ pathophysiological role in AD put this novel experimental combination regimen at substantial risk of failure.

Limitations of clinical trial sample size estimate by subtraction of two measurements

In planning randomized clinical trials (RCTs) for diseases such as Alzheimer's disease (AD), researchers frequently rely on the use of existing data obtained from only two time points to estimate sample size via the subtraction of baseline from follow-up measurements in each subject. However, the inadequacy of this method has not been reported. The aim of this study is to discuss the limitation of sample size estimation based on the subtraction of available data from only two time points for RCTs. Mathematical equations are derived to demonstrate the condition under which the obtained data pairs with variable time intervals could be used to adequately estimate sample size. The MRI-based hippocampal volume measurements from the Alzheimer's Disease Neuroimaging Initiative (ADNI) and Monte Carlo simulations (MCS) were used to illustrate the existing bias and variability of estimates. MCS results support the theoretically derived condition under which the subtraction approach may work. MCS also show the systematically under- or over-estimated sample sizes by up to 32.27 % bias. Not used properly, such subtraction approach outputs the same sample size regardless of trial durations partly due to the way measurement errors are handled. Estimating sample size by subtracting two measurements should be treated with caution. Such estimates can be biased, the magnitude of which depends on the planned RCT duration. To estimate sample sizes, we recommend using more than two measurements and more comprehensive approaches such as linear mixed effect models.

What have we learned from past failures of investigational drugs for Alzheimer's disease?

INTRODUCTION

In the last 15 years, huge efforts against Alzheimer's disease (AD) with drugs targeting β-amyloid (Aβ) and tau have produced poor clinical results. Aducanumab, a recently FDA-approved anti-Aβ monoclonal antibody has been greeted with distrust by most experts, hospitals and insurance companies for its level of efficacy and poor tolerability.

AREA COVERED

We reviewed literature on Alzheimer trials using PubMed, meeting abstracts and ClnicalTrials.gov and discuss what we can learn from past failures of investigational drugs for Alzheimer's disease, especially anti-Aβ and anti-tau drugs.

EXPERT OPINION

It is our opinion that previous failures of anti-AD drugs suggest that soluble Aβ and tau are not appropriate drug targets. In addition, pivotal clinical trials of future clinical candidates should avoid major protocol amendments and futility analyses. Study protocols should adopt better measures to protect study blinding and minimize the potential introduction of major biases in the evaluation of clinical results. Finally, alternative biological targets should be pursued as well as more multimodal approaches to addressing neurodegeneration in AD.

Pursuit of precision medicine: Systems biology approaches in Alzheimer's disease mouse models

Alzheimer's disease (AD) is a complex disease that is mediated by numerous factors and manifests in various forms. A systems biology approach to studying AD involves analyses of various body systems, biological scales, environmental elements, and clinical outcomes to understand the genotype to phenotype relationship that potentially drives AD development. Currently, there are many research investigations probing how modifiable and nonmodifiable factors impact AD symptom presentation. This review specifically focuses on how imaging modalities can be integrated into systems biology approaches using model mouse populations to link brain level functional and structural changes to disease onset and progression. Combining imaging and omics data promotes the classification of AD into subtypes and paves the way for precision medicine solutions to prevent and treat AD.

Diagnostic Value of the CSF α-Synuclein Real-Time Quaking-Induced Conversion Assay at the Prodromal MCI Stage of Dementia With Lewy Bodies

OBJECTIVE

To investigate whether the cerebrospinal fluid (CSF) α-synuclein (α-syn) real-time quaking-induced conversion (RT-QuIC) assay accurately identifies patients with mild cognitive impairment due to probable Lewy body disease (MCI-LB).

METHODS

We applied α-syn RT-QuIC to 289 CSF samples obtained from two independent cohorts, including 81 patients with probable MCI-LB (70.7±6.6 y, 13.6% F, MMSE 26.1±2.4), 120 with probable MCI-AD (68.6±7.4 y, 45.8% F, MMSE 25.5±2.8), and 30 with unspecified MCI (65.4±9.3 y, 30.0% F, MMSE 27.0±3.0). Fifty-eight individuals with no cognitive decline or evidence of neurodegenerative disease and 121 individuals lacking brain α-syn deposits at the neuropathological examination were used as controls.

RESULTS

RT-QuIC identified MCI-LB patients against cognitively unimpaired controls with 95% sensitivity, 97% specificity, and 96% accuracy, and showed 98% specificity in neuropathological controls. The accuracy of the test for MCI-LB was consistent between the two cohorts (97.3% vs. 93.7%). Thirteen percent of MCI-AD patients also had a positive test; of note, 44% of them developed one core or supportive clinical feature of dementia with Lewy bodies (DLB) at follow-up, suggesting an underlying LB co-pathology.

CONCLUSIONS

These findings indicate that CSF α-syn RT-QuIC is a robust biomarker for prodromal DLB. Further studies are needed to fully explore the added value of the assay to the current research criteria for MCI-LB.

CLASSIFICATION OF EVIDENCE

This study provides Class III evidence that CSF α-syn RT-QuIC accurately identifies patients with MCI due to LB disease.

Theoretical impact of the AT(N) framework on dementia using a community autopsy sample

The AT(N) research framework categorizes eight biomarker profiles using amyloid (A), tauopathy (T), and neurodegeneration (N), regardless of dementia status. We evaluated associations with dementia risk in a community-based cohort by approximating AT(N) profiles using autopsy-based neuropathology correlates, and considered cost implications for clinical trials for secondary prevention of dementia based on AT(N) profiles. We used Consortium to Establish a Registry for Alzheimer's Disease (moderate/frequent) to approximate A+, Braak stage (IV-VI) for T+, and temporal pole lateral ventricular dilation for (N)+. Outcomes included dementia prevalence at death and incidence in the last 5 years of life. A+T+(N)+ was the most common profile (31%). Dementia prevalence ranged from 14% (A-T-[N]-) to 79% (A+T+[N]+). Between 8% (A+T-[N]-) and 68% (A+T+[N]-) of decedents developed incident dementia in the last 5 years of life. Clinical trials would incur substantial expense to characterize AT(N). Many people with biomarker-defined preclinical Alzheimer's disease will never develop clinical dementia during life, highlighting resilience to clinical expression of AD neuropathologic changes and the need for improved tools for prediction beyond current AT(N) biomarkers.

Accelerating Alzheimer's disease drug discovery and development: what's the way forward?

Introduction: As the global burden of Alzheimer's disease (AD) grows, an effective disease-modifying therapy remains a distant prospect following the repeated failure of multiple therapeutics targeting β-amyloid and (it seems) tau over many years of costly effort. The repeated failure of single-target therapies to meaningfully modify disease progression raises major questions about the validity of many aspects of drug development in this area, especially target selection.Area covered: The authors explore the critical questions raised by a review of the collective experience to date, relating to why findings with non-clinical models and clinical biomarkers so frequently fail to translate to positive outcomes in clinical trials, which alternatives should be considered, and how we can design and conduct clinical trials that can successfully identify and quantify meaningful benefits in the future.Expert opinion: It is our opinion that we must recognize and accept the need to consider less specific, more multimodal approaches to addressing neurodegeneration in AD if we are to make progress - and we must avoid repeating the well intentioned, but ultimately erroneous, assumptions of the past.

Suppressing aberrant phospholipase D1 signaling in 3xTg Alzheimer's disease mouse model promotes synaptic resilience

Current approaches in treatment of Alzheimer's disease (AD) is focused on early stages of cognitive decline. Identifying therapeutic targets that promote synaptic resilience during early stages may prevent progressive memory deficits by preserving memory mechanisms. We recently reported that the inducible isoform of phospholipase D (PLD1) was significantly increased in synaptosomes from post-mortem AD brains compared to age-matched controls. Using mouse models, we reported that the aberrantly elevated neuronal PLD1 is key for oligomeric amyloid driven synaptic dysfunction and underlying memory deficits. Here, we demonstrate that chronic inhibition using a well-tolerated PLD1 specific small molecule inhibitor is sufficient to prevent the progression of synaptic dysfunction during early stages in the 3xTg-AD mouse model. Firstly, we report prevention of cognitive decline in the inhibitor-treated group using novel object recognition (NOR) and fear conditioning (FC). Secondly, we provide electrophysiological assessment of better synaptic function in the inhibitor-treated group. Lastly, using Golgi staining, we report that preservation of dendritic spine integrity as one of the mechanisms underlying the action of the small molecule inhibitor. Collectively, these studies provide evidence for inhibition of PLD1 as a potential therapeutic strategy in preventing progression of cognitive decline associated with AD and related dementia.

Design and sample size considerations for Alzheimer's disease prevention trials using multistate models

BACKGROUND/AIMS

Clinical trials for Alzheimer's disease have been aimed primarily at persons who have cognitive symptoms at enrollment. However, researchers are now recognizing that the pathophysiological process of Alzheimer's disease begins years, if not decades, prior to the onset of clinical symptoms. Successful intervention may require intervening early in the disease process. Critical issues arise in designing clinical trials for primary and secondary prevention of Alzheimer's disease including determination of sample sizes and follow-up duration. We address a number of these issues through application of a unifying multistate model for the preclinical course of Alzheimer's disease. A multistate model allows us to specify at which points during the long disease process the intervention exerts its effects.

METHODS

We used a nonhomogeneous Markov multistate model for the progression of Alzheimer's disease through preclinical disease states defined by biomarkers, mild cognitive impairment and Alzheimer's disease dementia. We used transition probabilities based on several published cohort studies. Sample size methods were developed that account for factors including the initial preclinical disease state of trial participants, the primary endpoint, age-dependent transition and mortality rates and specifications of which transition rates are the targets of the intervention.

RESULTS

We find that Alzheimer's disease prevention trials with a clinical primary endpoint of mild cognitive impairment or Alzheimer's disease dementia will require sample sizes of the order many thousands of individuals with at least 5 years of follow-up, which is larger than most Alzheimer's disease therapeutic trials conducted to date. The reasons for the large trial sizes include the long and variable preclinical period that spans decades, high rates of attrition among elderly populations due to mortality and losses to follow-up and potential selection effects, whereby healthier subjects enroll in prevention trials. A web application is available to perform sample size calculations using the methods reported here.

CONCLUSION

Sample sizes based on multistate models can account for the points in the disease process when interventions exert their effects and may lead to more accurate sample size determinations. We will need innovative strategies to help design Alzheimer's disease prevention trials with feasible sample size requirements and durations of follow-up.