Innovative biomarkers: tissue, liquid and AI

At IHU RespirERA, we are convinced that every patient is unique. This is why we conduct cutting-edge research to develop diagnostic and therapeutic approaches that are increasingly precise and adapted. The "Innovative Biomarkers: Tissue, Liquid, and AI" project is at the heart of this approach, aiming to transform the management of respiratory diseases.

Context

Respiratory diseases, and more specifically lung cancer, represent a major public health challenge. To better combat them, it is crucial to understand them in detail, diagnose them as early as possible, predict their evolution, and choose the most effective treatment for each patient. This is where biomarkers come in (measurable biological indicators, such as specific molecules or cellular characteristics, which provide information on health status or the presence of a disease).

Traditionally, the search for these biomarkers is often done through tissue biopsies (taking a small piece of tissue, for example from the lung). Although useful, these techniques can be invasive (requiring surgery or procedures that can be uncomfortable for the patient) and are sometimes limited for dynamic monitoring (monitoring the progression of the disease in real-time and repeatedly).

Furthermore, modern research has seen the massive emergence of "omics" data. This term includes the large-scale study of different components of our cells and our environment:

• Genomics (study of all our genes),
• Transcriptomics (study of RNA, the molecules that "read" genes to make proteins),
• Proteomics (study of all proteins),
• Metabolomics (study of small molecules resulting from cellular function, metabolites),
• Exposomics (study of all environmental exposures—pollution, diet, infections—a person has faced during their life).

When combined with medical imaging (CT scans, X-rays), we obtain massive and diverse volumes of information. For clinicians, manually analyzing such data is a challenge. This is where Artificial Intelligence (AI) acts as a powerful tool.

The challenge for IHU RespirERA: To exploit these new information sources and tools to revolutionize the care of patients with respiratory diseases.

Project Objectives

The "Innovative Biomarkers" project has clear ambitions to transform respiratory medicine:

Develop and integrate next-generation biomarkers:

• We seek to identify and validate new biomarkers from various sources. While classic tissue biopsies remain important, we place a major emphasis on liquid biopsy. This is a simple blood test that searches for traces of the disease (such as tumor DNA or cancer cells circulating in the blood). It is much less invasive and easier to repeat.
• IHU RespirERA aims to generalize the use of liquid biopsy in various clinical contexts: for monitoring metastatic cancer, for screening high-risk individuals, and for interception (detecting the very first signs of the disease before symptoms appear).

Better target treatments using biomarker panels:

• A key goal is to select panels (specific combinations) of biomarkers that are both prognostic (predicting disease evolution) and predictive (indicating which treatment is most likely to be effective).
• This will allow for better patient stratification into precise subgroups to offer truly personalized care.

Use Artificial Intelligence (AI) to boost analysis capabilities:

• AI is central to analyzing complex and massive biomarker data.
• It helps improve the diagnosis of both cancerous and non-cancerous respiratory pathologies (such as rare lung diseases).
• It enables the identification of molecular alterations and the prediction of genetic mutations.
• AI will also contribute to discovering new biomarkers and prioritizing drug candidates.

Added value for patients: Earlier and less painful diagnoses, better-targeted treatments, and a deeper understanding of their disease.

Methodology and Collaboration

To achieve these goals, IHU RespirERA relies on rigorous methodology and strategic collaborations:

• Liquid biopsy at the forefront: We use liquid biopsy to analyze circulating tumor DNA (ctDNA) and circulating tumor cells (CTC) using high-throughput sequencing techniques like NGS (Next-Generation Sequencing) and WGS (Whole Genome Sequencing).

• Integrated "multi-omics" approaches: We combine various analyses:
  • Spatial transcriptomics (which allows us to see which genes are active and where exactly they are active within a tissue).
  • Single-cell RNAseq (cell-by-cell analysis of gene activity, offering unrivalled resolution).
  • Epigenetics (the study of mechanisms that control gene activity without altering the DNA sequence itself).
  • Proteomics, metabolomics and exposomics (described above).

These approaches help us to characterise the molecular determinants of pathologies (understanding precisely which molecules and mechanisms are involved in the development and progression of respiratory diseases).

• Artificial Intelligence as a driver of analysis and discovery: AI is used at several levels:
  • For the standardisation of liquid biopsy data (ensuring that results obtained from different samples or using different techniques are comparable with each other).
  • For the combination and cross-analysis of different data sources (clinical patient information, multi-omic data, imaging results).
  • For the analysis of histological images (images of tissue observed under a microscope) and radiological images (scans, X-rays) in order to detect signs of disease that the human eye might miss or interpret more slowly.
  • To do this, we develop specific algorithms (computer programmes designed for specific tasks), for example for clustering (automatically grouping together patients or data that are similar) or variable selection (identifying, from among thousands of data points, those that are truly important for predicting a disease or response to treatment).

• Collaborations of excellence: This project would not be possible without a network of strong partners. Key collaborations include:
  • INRIA-3IA Côte d'Azur (National Institute for Research in Computer Science and Control, and Interdisciplinary Institute for Artificial Intelligence), a major player in the development of AI algorithms.
  • The clinical services of the founding partners of the RespirERA IHU (notably Nice University Hospital, Paris Public Hospital System, and the Georges-François Leclerc Centre in Dijon), which are essential for accessing patient data and clinically validating discoveries.
  • Industrial partnerships to accelerate the transfer of research innovations into concrete diagnostic tools and treatments for patients.

The commitment of IHU RespirERA: Pooling the best expertise and technologies to advance research and improve patients' lives.

Expected Progress and Impact

The advances resulting from the ‘Innovative Biomarkers: Tissue, Liquid and AI’ project will have major, tangible impacts for patients and the medical community:

• Less invasive diagnosis: The wider adoption of liquid biopsy (a simple blood test) will offer a much less invasive and easier-to-repeat alternative to traditional tissue biopsies. This will enable earlier diagnosis and better monitoring of disease progression and treatment effectiveness.
• More personalised and responsive respiratory medicine: Integrating innovative biomarker data with the analytical power of AI will enable better patient stratification. In practical terms, this means that we will be able to more accurately identify the subgroups of patients who will respond best to a given treatment, or those who are at higher risk of complications. Medical decisions will thus be optimised, leading to truly personalised medicine that can adapt quickly to the progression of each patient's disease.
• The discovery of new therapeutic targets: By better understanding the molecular mechanisms of respiratory diseases, this project will pave the way for the discovery of new targets (new weak points in diseased cells against which to develop drugs) and a much more detailed and comprehensive interpretation of biological data.
• More accurate and rapid diagnoses thanks to AI: The application of AI to the analysis of histology images (tissue slides viewed under a microscope) and radiology images (scans, MRIs) should significantly improve the accuracy, speed and reliability of diagnoses of lung cancer, as well as other complex conditions such as interstitial lung diseases (a group of rare diseases that affect the lung's supporting tissue).

The promise of IHU RespirERA: We are moving toward an era where respiratory diseases are detected earlier and treated more effectively, significantly improving patient quality of life and prognosis.