Leveraging medical records to identify patients at risk of neurodegenerative disease – LeMeReND
Leveraging medical records to identify patients at risk of neurodegenerative disease
Neurodegenerative diseases such as Alzheimer’s and Parkinson’s represent a major public health issue. Early identification of at-risk individuals is essential for implementing prevention strategies. LeMeReND aims to leverage longitudinal electronic health records (EHR) from multiple countries to identify biomedical risk factors, predict disease onset, and create a screening tool to support prevention efforts and personalized care.
LeMeReND seeks to develop EHR-based screening methods to identify individuals at risk of neurodegenerative diseases, addressing the challenges of early detection and personalized prevention strategies
Neurodegenerative diseases such as Alzheimer’s disease (AD), Parkinson’s disease (PD), dementia with Lewy bodies (LBD), and motor neuron diseases (MND) pose significant public health challenges due to their increasing prevalence and lack of curative treatments. Prevention strategies, particularly focused on early intervention, are crucial in mitigating the social and economic burden of these diseases.<br /><br />One of the central challenges is identifying individuals who are at risk of developing these diseases long before clinical symptoms appear. Early identification would allow for timely interventions, improving outcomes for at-risk individuals through both primary and secondary prevention measures. However, current methods for detecting risk factors are often inadequate, relying heavily on symptomatic presentation rather than pre-symptomatic or prodromal indicators.<br /><br />LeMeReND addresses this gap by utilizing vast electronic health records (EHR) databases from Australia, France, the UK, and Sweden to identify biomedical risk factors that can predict the future onset of neurodegenerative diseases. By analyzing data spanning more than 10 years, the project aims to uncover both common and disease-specific factors that contribute to disease onset, offering insights into how these diseases can be differentiated or grouped by their risk profiles.<br /><br />A central objective of LeMeReND is the development of a software tool that can be used by general practitioners (GPs) at the point of care. This tool will integrate patient history data and calculate a propensity score for the likelihood of developing AD, PD, MND, or LBD, allowing GPs to identify at-risk individuals early on. This approach could enhance recruitment for secondary prevention trials and optimize clinical resource allocation by prioritizing high-risk patients.<br /><br />Additionally, the project will explore the associations between these identified biomedical factors and brain MRI data, as well as genetic information, to better understand the underlying mechanisms driving disease progression. This will ultimately guide the development of tailored prevention strategies.<br /><br />LeMeReND also aims to assess the economic and social implications of such precision prevention approaches, exploring public acceptability and evaluating the feasibility of implementing such strategies in primary care settings in France, Australia, and Sweden.
LeMeReND utilizes a comprehensive and multidisciplinary approach, employing advanced methodologies and technologies to meet its objectives:
Electronic Health Records (EHR) Analysis: LeMeReND capitalizes on the availability of extensive longitudinal data from healthcare registries in four countries—Australia, France, the UK, and Sweden. These records include patient diagnoses, drug prescriptions, clinical care usage, and biological test results collected over more than a decade. By analyzing these large-scale datasets, the project will identify key biomedical risk factors associated with neurodegenerative diseases.
Longitudinal Data Analysis: The study leverages longitudinal modeling to analyze disease progression profiles. This allows researchers to examine how different biomedical factors evolve over time and contribute to the eventual onset of diseases such as Alzheimer’s, Parkinson’s, LBD, and MND.
Machine Learning: Advanced machine learning algorithms will be applied to the data to identify patterns that traditional statistical methods might overlook. This approach is particularly valuable when dealing with complex, high-dimensional data such as EHRs, where interactions between various risk factors may be nonlinear or obscured by noise.
Software Tool for Risk Stratification: LeMeReND aims to develop a software tool for use by general practitioners. This tool will ask targeted questions about a patient’s medical history and automatically calculate a propensity score for each neurodegenerative disease. This would help GPs identify at-risk patients during routine care and prioritize them for further investigation or inclusion in prevention trials.
Neuroimaging and Genetic Data Integration: The project will further integrate brain MRI data and genetic information from resources like the UK BioBank and GWAS summary statistics. By studying the relationship between identified risk factors and neuroimaging/genetic markers, LeMeReND aims to uncover disease mechanisms, providing insights into the biological processes that precede clinical symptoms.
Health Economics and Social Feasibility Studies: To ensure the proposed precision prevention strategies are viable, LeMeReND includes a component that evaluates the economic and social impact of implementing such measures. This involves assessing the cost-effectiveness of prevention strategies and studying the public acceptability of secondary prevention in France, Australia, and Sweden. The feasibility of incorporating these strategies into routine primary care will also be examined.
While LeMeReND is still in progress, anticipated key results include:
Identification of Biomedical Risk Factors: The project aims to identify a set of biomedical risk factors—based on diagnoses, drug prescriptions, and clinical data—that are predictive of the future onset of neurodegenerative diseases like AD, PD, LBD, and MND. These factors will be stratified according to their progression profiles, providing a clearer picture of how these diseases develop over time.
Development of a Risk Stratification Tool: A significant outcome will be the creation of a software tool that general practitioners can use to assess a patient’s risk for developing neurodegenerative diseases. This tool will provide a propensity score based on their medical history, enabling early identification and prioritization of high-risk individuals for preventive measures or further clinical studies.
Integration of Neuroimaging and Genetic Data: LeMeReND will also link identified biomedical factors with neuroimaging and genetic data. This integration aims to elucidate the biological mechanisms driving disease progression and identify potential new therapeutic targets.
Economic and Social Insights: The project will provide an in-depth evaluation of the economic benefits of early detection and prevention strategies, highlighting potential cost savings for healthcare systems. Additionally, LeMeReND will shed light on public attitudes towards secondary prevention and the practicalities of incorporating such strategies into primary care in various countries.
The outstanding feature of LeMeReND lies in its interdisciplinary approach, combining expertise from epidemiology, statistics, machine learning, health economics, and public health. This approach allows the project to address multiple dimensions of neurodegenerative disease prevention, from risk prediction and patient stratification to the social and economic implications of large-scale preventive measures.
One of the most promising future prospects is the application of the risk stratification tool at the point of care. This tool has the potential to transform how general practitioners identify at-risk individuals, making precision prevention a reality in everyday clinical practice. By helping healthcare providers prioritize patients for preventive interventions or inclusion in secondary prevention trials, this tool could greatly accelerate research into new treatments and interventions.
Another exciting prospect is the possibility of uncovering new therapeutic targets through the integration of neuroimaging and genetic data with EHR-derived risk factors. This could lead to breakthroughs in understanding the biological mechanisms behind these diseases and open the door to novel treatments that target the early stages of neurodegeneration.
Finally, LeMeReND’s work on evaluating the feasibility and acceptability of precision prevention measures will inform future health policies, ensuring that the strategies developed are both practical and widely accepted by the public. The project’s findings could serve as a blueprint for the global adoption of preventive measures for neurodegenerative diseases, ultimately reducing the burden on healthcare systems and improving patient outcomes worldwide.
Neurodegenerative diseases represent one of the main public health issues in our western societies and one of the greatest challenges in drug
development. Prevention policies have become essential to address these issues: primary prevention to prevent disease onset by acting on actionable
risk factors, or secondary prevention to slow disease progression with very early therapeutic interventions, ideally at pre-symptomatic stages. Key to
the implementation of such prevention measures is the identification of at-risk patients, at the point of care, and preferably long before disease onset.
Our project, LeMeReND, proposes to use electronic health records (EHR) to identify biomedical risk factors through studying previous diagnoses (preclinical
comorbidities), drug prescription, clinical care usage, and biological test results. This analysis will use longitudinal data in EHR registries
including millions of patients who have been followed for at least 10 years before diagnosis in 4 different healthcare systems: Australia, France, the UK
and Sweden and across 4 therapeutic areas: Alzheimer’s disease, Parkinson’s disease, dementia with Lewy bodies and motor neuron diseases. We
will identify the biomedical risk factors that are common to these diseases and the ones differentiating them.
We will stratify patients based on the progression profile of their exposure to the set of risk factors, in order to design tailored primary prevention
measures. We will also design a screening tool which will give each patient a propensity score to develop one of these neurodegenerative diseases.
Such a tool could be deployed at the point of care to prioritise at-risk individuals for further inclusion in secondary prevention trials. We will evaluate the
economic and social benefits of this new generation of precision prevention measures. We will study the public acceptability of a secondary-prevention
effort, among the French population, and the feasibility of its implementation in primary care practices in France, Australia, and Sweden. Eventually,
we will progress our understanding of the genetic and imaging markers of the disorders by studying the identified prodromal biomedical factors, using
the UK BioBank and GWAS summary statistics. This will progress our understanding of the pathological processes which result in an increased risk to
develop a specific neurodegenerative disease.
LeMeReND gathers a multidisciplinary research group with a leading expertise in epidemiology, statistics and machine learning, in particular for the
analysis of longitudinal EHR data. Partners have demonstrated a strong track record on neurodegenerative diseases (Sweden, France, Australia),
analyses of large-scale data including neuroimaging (France), genetics (Australia), longitudinal modelling (Sweden, France), and machine learning
(Australia, France). An expert team in health economics and health policy complements the consortium.
LeMeReND will therefore provide invaluable insights to inform health policies and highlight possible new therapeutic targets. It will provide unique
screening tools to facilitate the large-scale recruitment of patients in secondary prevention trials.
Project coordination
Stanley DURRLEMAN (Institut du cerveau)
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
Partnership
ICM Institut du cerveau
AMSE Aix Marseille School of economics
Help of the ANR 432,980 euros
Beginning and duration of the scientific project:
March 2022
- 36 Months
