Periodic Reporting for period 6 - CARDIATEAM (CARdiomyopathy in type 2 DIAbetes mellitus)
Période du rapport: 2024-03-01 au 2025-02-28
Type 2 Diabetes (T2DM) epidemic is associated with serious comorbidities such as heart failure, the second most common cardiovascular disease after ischaemic heart disease in T2DM. CARDIATEAM is aiming to differentiate diabetic cardiomyopathy (DCM) against other forms of heart failure, identify its causal mechanism, and evaluate its impact on mortality. Thus, the identification of a novel phenotypic signature of DCM should allow to:
• predict cardiac function decline in T2DM patients
• allow for early preventive strategies
• facilitate tailored therapies to slow disease progression
• develop disease modeling for translatable preclinical models
Seven dedicated work packages (WP) have been set-up to establish a strong collaborative network of clinical and basic researchers. Enrolling 1,600 patients in a prospective clinical study, CARDIATEAM will provide a unique and highly standardized set of data and allow differentiation of related forms of heart failure. Deep phenotyping and unbiased machine learning will allow in a yet unrivalled precision to identify clusters of connected pathology, identification of affected biochemical pathways and specific biomarkers. A central database integrating historical data and data from clinical and experimental sources will permit novel bioinformatics-assisted visualization and modelling of interactions between phenotype, genetic, immune and metabolic pathways.
Building on these results, CARDIATEAM will enable DCM modelling, identify crucial pathophysiologic mechanisms and ultimately develop tailored treatments and robust preclinical models.
• predict cardiac function decline in T2DM patients
• allow for early preventive strategies
• facilitate tailored therapies to slow disease progression
• develop disease modeling for translatable preclinical models
Seven dedicated work packages (WP) have been set-up to establish a strong collaborative network of clinical and basic researchers. Enrolling 1,600 patients in a prospective clinical study, CARDIATEAM will provide a unique and highly standardized set of data and allow differentiation of related forms of heart failure. Deep phenotyping and unbiased machine learning will allow in a yet unrivalled precision to identify clusters of connected pathology, identification of affected biochemical pathways and specific biomarkers. A central database integrating historical data and data from clinical and experimental sources will permit novel bioinformatics-assisted visualization and modelling of interactions between phenotype, genetic, immune and metabolic pathways.
Building on these results, CARDIATEAM will enable DCM modelling, identify crucial pathophysiologic mechanisms and ultimately develop tailored treatments and robust preclinical models.
Building a Unique Patient Cohort and Data Infrastructure
Since its launch, CARDIATEAM has built a large, deeply characterized group of volunteers from across Europe. By the end of the most recent period (February 2025), 1,219 patients have been recruited, including people with and without type 2 diabetes and with different heart conditions. Recruitment was expanded and adapted to overcome challenges, such as those posed by the COVID-19 pandemic, with additional French centers and increased targets at some hospitals.. The infrastructure for collecting patient data-including advanced heart imaging (echocardiography, cardiac MRI), eye scans, and blood samples- is now fully operational, supported by a secure web-based database. This database integrates clinical, imaging, and molecular data, ensuring high-quality, standardized information for analysis.
Data Quality, Monitoring, and Analysis
All 17 participating centers are actively uploading data, with over 2,500 imaging studies and 290,000 data entries reviewed to date. Automated tools and dedicated data scientists have been deployed to monitor data quality, identify inconsistencies, and ensure the integrity of the information collected. Regular meetings and training sessions keep all partners aligned and maintain high standards across the network. Tests show that the quality of the imaging data is excellent and matches international standards.
Innovative Machine Learning and Pheno-Mapping Approaches
CARDIATEAM stands out for its use of advanced computational methods, including unsupervised machine learning and clustering, to analyze the large and complex dataset. These techniques help identify subgroups of patients with similar disease patterns and uncover hidden links between risk factors and heart dysfunction. Early analyses have highlighted the importance of metabolic and immune pathways in the development of DCM, pointing to potential new avenues for diagnosis and treatment.
Biomarker Discovery and Molecular Research
A key part of the project is the analysis of multiple layers of biological data-genes, proteins, and metabolites-from patient samples. This “omics” approach is revealing molecular signatures that may predict who is most at risk for DCM or who may benefit from specific treatments. Collaboration with our industry partners has enabled the measurement of additional biomarkers, enriching the dataset and increasing the potential for discovery.
Preclinical Models and Experimental Research
To complement patient studies, CARDIATEAM is developing and refining animal models that mimic DCM. These models help researchers explore how factors like diet, aging, and genetics contribute to heart disease in diabetes. Recent experiments have focused on mice exposed to metabolic stress, offering insights into the progression from metabolic disorders to heart failure. Tissue samples from these studies are being analyzed to identify early changes and potential therapeutic targets.
Knowledge Sharing and Scientific Impact
The consortium regularly meets to discuss progress, share results, and plan next steps. CARDIATEAM has published review articles on experimental models and the role of epigenetics in DCM, contributing to the global understanding of this condition. The project is active in scientific conferences and public events, promoting collaboration and raising awareness about the importance of personalized medicine in diabetes and heart disease.
Our next steps will focus on finding clear links between specific biological markers and clinical outcomes. By combining advanced data analysis with laboratory research, we hope to uncover the mechanisms behind diabetic cardiomyopathy.
Since its launch, CARDIATEAM has built a large, deeply characterized group of volunteers from across Europe. By the end of the most recent period (February 2025), 1,219 patients have been recruited, including people with and without type 2 diabetes and with different heart conditions. Recruitment was expanded and adapted to overcome challenges, such as those posed by the COVID-19 pandemic, with additional French centers and increased targets at some hospitals.. The infrastructure for collecting patient data-including advanced heart imaging (echocardiography, cardiac MRI), eye scans, and blood samples- is now fully operational, supported by a secure web-based database. This database integrates clinical, imaging, and molecular data, ensuring high-quality, standardized information for analysis.
Data Quality, Monitoring, and Analysis
All 17 participating centers are actively uploading data, with over 2,500 imaging studies and 290,000 data entries reviewed to date. Automated tools and dedicated data scientists have been deployed to monitor data quality, identify inconsistencies, and ensure the integrity of the information collected. Regular meetings and training sessions keep all partners aligned and maintain high standards across the network. Tests show that the quality of the imaging data is excellent and matches international standards.
Innovative Machine Learning and Pheno-Mapping Approaches
CARDIATEAM stands out for its use of advanced computational methods, including unsupervised machine learning and clustering, to analyze the large and complex dataset. These techniques help identify subgroups of patients with similar disease patterns and uncover hidden links between risk factors and heart dysfunction. Early analyses have highlighted the importance of metabolic and immune pathways in the development of DCM, pointing to potential new avenues for diagnosis and treatment.
Biomarker Discovery and Molecular Research
A key part of the project is the analysis of multiple layers of biological data-genes, proteins, and metabolites-from patient samples. This “omics” approach is revealing molecular signatures that may predict who is most at risk for DCM or who may benefit from specific treatments. Collaboration with our industry partners has enabled the measurement of additional biomarkers, enriching the dataset and increasing the potential for discovery.
Preclinical Models and Experimental Research
To complement patient studies, CARDIATEAM is developing and refining animal models that mimic DCM. These models help researchers explore how factors like diet, aging, and genetics contribute to heart disease in diabetes. Recent experiments have focused on mice exposed to metabolic stress, offering insights into the progression from metabolic disorders to heart failure. Tissue samples from these studies are being analyzed to identify early changes and potential therapeutic targets.
Knowledge Sharing and Scientific Impact
The consortium regularly meets to discuss progress, share results, and plan next steps. CARDIATEAM has published review articles on experimental models and the role of epigenetics in DCM, contributing to the global understanding of this condition. The project is active in scientific conferences and public events, promoting collaboration and raising awareness about the importance of personalized medicine in diabetes and heart disease.
Our next steps will focus on finding clear links between specific biological markers and clinical outcomes. By combining advanced data analysis with laboratory research, we hope to uncover the mechanisms behind diabetic cardiomyopathy.
CARDIATEAM will go beyond the state-of-the-art since it will investigate mechanisms underlying DCM and prognosis.
Expected results of CARDIATEAM will be
• establishment of a prospective cohort of 1,600 patients within 2 years and with a follow-up of 3 years, phenotyping the patients with echocardiography, CMR, retinography and -omics
• Application of unsupervised machine learning algorithms to improve cardiac phenotyping & identification of DCM (WP4)
• Provide a sex- and age- based stratification approach of T2DM patients at risk of DCM (WP 2)
• Identification of causal mechanisms and pathways responsible for DCM (WP 2, WP3, WP5 & WP7)
• Identification of new potential therapeutic targets for preventing or alleviating DCM (WP6)
• Application of disease modelling to develop DCM preclinical models (WP7)
• New taxonomy of DCM to be communicated to health agencies, practitioners and patients (WP1)
The results of CARDIATEAM will impact clinical care with the stratification of patients into risk groups of developing DCM, earlier diagnosis of DCM and an improvement of therapy thanks to better assessment of underlying pathophysiology and identification of new biomarkers.
The outcome and results of CARDIATEAM will have significant impact on the efficiency of R&D in the field of Diabetes & Cardiovascular Disease.
The deep molecular and phenotypic characterization of DCM with the discovery and validation of respective biomarkers will allow an improvement in developing therapeutic options by:
- discovery and validation of biomarkers being predictive for risk and progression to DCM as well as to monitor efficacy of treatment options.
- development appropriate (animal) models of relevance for human DCM to profile and develop drug candidates for this medical indication.
- stratification of patients with DCM for clinical drug trials; this would enable smaller and focused clinical studies to evaluate efficacy of drug candidates in this indication earlier, faster and cheaper to progress and introduce them into clinical practice. Finally, DCM cluster analysis will potentially allow the rational repositioning of existing treatments.
Expected results of CARDIATEAM will be
• establishment of a prospective cohort of 1,600 patients within 2 years and with a follow-up of 3 years, phenotyping the patients with echocardiography, CMR, retinography and -omics
• Application of unsupervised machine learning algorithms to improve cardiac phenotyping & identification of DCM (WP4)
• Provide a sex- and age- based stratification approach of T2DM patients at risk of DCM (WP 2)
• Identification of causal mechanisms and pathways responsible for DCM (WP 2, WP3, WP5 & WP7)
• Identification of new potential therapeutic targets for preventing or alleviating DCM (WP6)
• Application of disease modelling to develop DCM preclinical models (WP7)
• New taxonomy of DCM to be communicated to health agencies, practitioners and patients (WP1)
The results of CARDIATEAM will impact clinical care with the stratification of patients into risk groups of developing DCM, earlier diagnosis of DCM and an improvement of therapy thanks to better assessment of underlying pathophysiology and identification of new biomarkers.
The outcome and results of CARDIATEAM will have significant impact on the efficiency of R&D in the field of Diabetes & Cardiovascular Disease.
The deep molecular and phenotypic characterization of DCM with the discovery and validation of respective biomarkers will allow an improvement in developing therapeutic options by:
- discovery and validation of biomarkers being predictive for risk and progression to DCM as well as to monitor efficacy of treatment options.
- development appropriate (animal) models of relevance for human DCM to profile and develop drug candidates for this medical indication.
- stratification of patients with DCM for clinical drug trials; this would enable smaller and focused clinical studies to evaluate efficacy of drug candidates in this indication earlier, faster and cheaper to progress and introduce them into clinical practice. Finally, DCM cluster analysis will potentially allow the rational repositioning of existing treatments.