Periodic Reporting for period 2 - I3LUNG (Integrative science, Intelligent data platform for Individualized LUNG cancer care with Immunotherapy)
Okres sprawozdawczy: 2023-12-01 do 2025-05-31
Although immunotherapy is the new standard of care for advanced non-small cell lung cancer (NSCLC), less than 50 % of patients benefit from this treatment in the long term. Programmed death-ligand 1 biomarker is used to predict immunotherapy outcomes with limited efficacy, and other potential biomarkers have not yet been validated in randomised clinical trials. The EU-funded I3LUNG project aims to develop AI-based tools for improving patients’ survival and quality of life.
The project will set up a global platform comprising data from 2 000 patients for AI models validation; moreover, it will collect multi-omics data from 200 NSCLC patients for information integration and application in leading immunotherapy decisions.
The project will set up a global platform comprising data from 2 000 patients for AI models validation; moreover, it will collect multi-omics data from 200 NSCLC patients for information integration and application in leading immunotherapy decisions.
Since its launch, I3LUNG has established a robust foundation for clinical and translational research. Retrospective data collection has been completed across all clinical sites, surpassing the initial target cohort size with more than 2.200 patients enrolled. This cohort provides a uniquely rich basis for model development, as it includes clinical information coupled with radiological images, digital pathology slides, and genomic, transcriptomic, metabolomic, and immunological data. In parallel, a prospective study, launched in 2022 and still ongoing, had recruited nearly 200 patients by mid-2025, with more than 120 with the full panoply of omics data available. This ongoing recruitment serves as the clinical basis for validating the predictive models built between 2023 and 2024 on retrospective data, and for expanding them with longitudinal quality-of-life and behavioral information collected directly from the patients via a mobile application and psychological studies.
The technological backbone of the project is a secure data-sharing and elaboration system specifically designed for I3LUNG. The platform is already fully operational and currently hosts retrospective data, ongoing uploads of prospective datasets, and the integration of AI models. It also supports radiomics segmentation workflows, digital pathology processing, and standardized multi-omics pipelines that ensure harmonization of tissue and plasma analyses across centers.
On this foundation, the project is developing three AI-based devices, as decision-support systems. The Individual Patient Decision Aid System (IPDAS) is designed to provide patients with clear, personalized information about their disease and the available treatment options, while integrating their own quality-of-life priorities and preferences. After several cycles of refinement based on patient focus groups, clinician feedback, and pilot usability studies the system is now being piloted in the prospective trial. The Physician Decision Support System (PDSS) provides oncologists with user-friendly, transparent predictions of outcomes such as overall and progression-free survival to guide them in their interactions with patients and in selecting the treatment path most adapted to the individual patient. This tool has been finalized in prototype form, validated on retrospective cohorts, and is being refined for real-world application. In parallel, data will be made accessible near the end of the project to enable the wider scientific community to access and interrogate datasets and models within a secure environment.
I3LUNG has paid close attention to the rapidly evolving, cross-cutting topic of responsible and transparent innovation in healthcare. Ethical and legal project partners have completed an in-depth analysis of the regulatory and ethical challenges raised by AI in oncology, with particular focus on data protection, data sharing, initial assessment of medical device compliance, and patient rights. Psychological studies are assessing how patients and physicians perceive the use of AI in treatment decisions, the extent to which it is trusted, and its impact on the doctor–patient relationship, a core tenet of the Explainable AI methodologies that represent a crucial aspect the project promotes. A dedicated mobile app has been fully deployed to monitor patients’ self-reported and sensor-derived quality-of-life data. Health economic assessments have also been initiated, with preliminary budget impact models prepared and ready to be validated as soon as the prospective dataset reaches maturity at the end of 2025 and that could represent a useful framework to help healthcare policy experts and regulators steer towards effective integration of AI in healthcare systems including from the reimbursement and insurance coverage aspects of these novel technologies.
The next phase of the project (2025-2027) will focus on finalizing the prospective recruitment to achieve the target of fully profiled patients, finalizing multimodal data pipelines for the definitive validation of the models that in turn will allow the scaling up of the deployment of decision-support tools in clinical practice and delivering validated predictive models for immunotherapy outcomes.
For more information on the project, please visit our website www.i3lung.eu (and its dedicated 'for patients' content - https://i3lung.eu/for-patients/(odnośnik otworzy się w nowym oknie)) and social media accounts (https://www.linkedin.com/company/i3lung(odnośnik otworzy się w nowym oknie); https://www.instagram.com/i3lung/(odnośnik otworzy się w nowym oknie))
The technological backbone of the project is a secure data-sharing and elaboration system specifically designed for I3LUNG. The platform is already fully operational and currently hosts retrospective data, ongoing uploads of prospective datasets, and the integration of AI models. It also supports radiomics segmentation workflows, digital pathology processing, and standardized multi-omics pipelines that ensure harmonization of tissue and plasma analyses across centers.
On this foundation, the project is developing three AI-based devices, as decision-support systems. The Individual Patient Decision Aid System (IPDAS) is designed to provide patients with clear, personalized information about their disease and the available treatment options, while integrating their own quality-of-life priorities and preferences. After several cycles of refinement based on patient focus groups, clinician feedback, and pilot usability studies the system is now being piloted in the prospective trial. The Physician Decision Support System (PDSS) provides oncologists with user-friendly, transparent predictions of outcomes such as overall and progression-free survival to guide them in their interactions with patients and in selecting the treatment path most adapted to the individual patient. This tool has been finalized in prototype form, validated on retrospective cohorts, and is being refined for real-world application. In parallel, data will be made accessible near the end of the project to enable the wider scientific community to access and interrogate datasets and models within a secure environment.
I3LUNG has paid close attention to the rapidly evolving, cross-cutting topic of responsible and transparent innovation in healthcare. Ethical and legal project partners have completed an in-depth analysis of the regulatory and ethical challenges raised by AI in oncology, with particular focus on data protection, data sharing, initial assessment of medical device compliance, and patient rights. Psychological studies are assessing how patients and physicians perceive the use of AI in treatment decisions, the extent to which it is trusted, and its impact on the doctor–patient relationship, a core tenet of the Explainable AI methodologies that represent a crucial aspect the project promotes. A dedicated mobile app has been fully deployed to monitor patients’ self-reported and sensor-derived quality-of-life data. Health economic assessments have also been initiated, with preliminary budget impact models prepared and ready to be validated as soon as the prospective dataset reaches maturity at the end of 2025 and that could represent a useful framework to help healthcare policy experts and regulators steer towards effective integration of AI in healthcare systems including from the reimbursement and insurance coverage aspects of these novel technologies.
The next phase of the project (2025-2027) will focus on finalizing the prospective recruitment to achieve the target of fully profiled patients, finalizing multimodal data pipelines for the definitive validation of the models that in turn will allow the scaling up of the deployment of decision-support tools in clinical practice and delivering validated predictive models for immunotherapy outcomes.
For more information on the project, please visit our website www.i3lung.eu (and its dedicated 'for patients' content - https://i3lung.eu/for-patients/(odnośnik otworzy się w nowym oknie)) and social media accounts (https://www.linkedin.com/company/i3lung(odnośnik otworzy się w nowym oknie); https://www.instagram.com/i3lung/(odnośnik otworzy się w nowym oknie))
The I3LUNG project has generated a set of preliminary results towards the integration of multimodal data and AI in the context of individualized lung and broader cancer care. These results include the development of predictive models based on retrospective and prospective cohorts, and the establishment of a decision support system for clinicians, as well as an individual patient decision aid system to enhance shared decision-making. The potential impacts are substantial in terms of improving treatment selection in non-small cell lung cancer immunotherapy, supporting evidence-based clinical decisions, and enhancing patient involvement in care choices - all items being scalable to other tumor types and pathologies.