Periodic Reporting for period 3 - REVERT (taRgeted thErapy for adVanced colorEctal canceR paTients)
Periodo di rendicontazione: 2023-01-01 al 2024-12-31
The REVERT project addresses the specific challenge of understanding at system level the pathophysiology of metastatic colorectal cancer (mCRC) in patients responding well or poorly to therapies, in order to design an optimal strategy for mCRC on a case-by-case basis, with therapeutic interventions modulated depending on the patient’s features. CRC is diagnosed in more than 400,000 people per year in Europe, almost 50% of whom are developing distant metastases. New molecular targeting drugs currently used to treat CRC might have severe side effects and lead to high treatment costs. Therefore, in consideration of the high prevalence of CRC, the frequent development of metastases, and the cost of commonly used drugs mCRC constitutes a high economic burden to the health and social systems within the EU. Accordingly, REVERT is building up an innovative artificial intelligence (AI)-based decision support system using the experience and the real-world data of several general Hospitals and Biobanks operating in the EU healthcare system, ultimately aiming at developing an improved and innovative model of combinatorial therapy (based on a personalised medicine approach) that identifies the most efficient and cost-effective therapeutic intervention for patients with unresectable mCRC
Over the whole period, the project results progressed toward the overall REVERT objective to create an AI-based decision support system (DSS) that can aid the oncologist in determining the best combinatorial treatment for patients with unresectable mCRC. The “FAIR - Data Management Plan and principle” was devised by the Coordinator, with the contribution of all partners and was subsequently updated in relation to the "AI Act" draft of June 2023. The virtual infrastructure of the REVERT DataBase (RDB) was created according to the principles of Privacy by Design and Privacy by Default and all partners contributed to the construction of the RDB by compiling clinical, histological and molecular information for both retrospectively and prospectively collected cases. As one of the major obstacles to the wide adoption of AI prediction systems in medicine is their opacity, which undermines clinicians’ trust, several approaches of explainable AI (XAI) have been pursued based on the retrospective analysis of RDB, as also envisaged in the EU AI Act. Moreover, to improve the performance of a model relying on a single algorithmic approach, the REVERT IT Partners defined an innovative approach to gain the trust of clinicians: the multi-party, responsibility-sharing AI DSS, in which trust is built by consensus of four different AI algorithms independently developed by the four centres, and independently applied. This AI DSS was validated in a clinical prospective pilot study (ClinicalTrials.gov PRS ID: NCT05396807) in which DSS assisted physicians in making more informed decisions about most suitable treatment options. Efficacy assessment was performed according to clinical outcome, showing that the Progression-Free Survival was improved compared to historical data. Other activities related to the identification of gene alterations important for CRC pathways showed that the identified genes have the potential to be used as targets and biomarkers for AI/ML programs for building prediction models. Meanwhile, organoids of CRC patient tissues were established and subjected to transcriptomic and miRNA epigenetic analyses showing a distinct expression signature for tumour organoids compared to tissue from normal CRC cells cultured as organoids and obtained from the same patients. Indeed, the established organoids displayed different rates of responses to already approved drugs which were useful to validate the REVERT AI DSS and a novel molecular algorithm – based on transcriptomic data – for response prediction to established chemotherapy regimens as well as novel combinational drugs. Next generation sequencing (NGS) analysis of prospective organoids of primary and mCRC further showed the involvement of genes regulating “Cell and Extracellular Matrix” interactions as well as pathways related to proteoglycans. Fluorescent MIPs targeting glycans were developed and used for staining CRC cell lines expressing sialic acid. Finally, epigenetic analysis suggested that observed effects are partly due to differential expression of specific miRNAs. Dissemination of the research results has been ensured through the REVERT project website (www.revert-project.eu) social media channels (i.e. LinkedIn and ) or specialised press and blogs, attendance/ presentation at National and International Meetings and OA publication in peer-reviewed journals (n=42) or dissemination and outreach magazines (n=3). The exploitation plan addressed key areas including market viability, the regulatory landscape for AI in healthcare, and intellectual property management. Commercial exploitation also aligned with those already present in the collaborating SMEs and has already shown the capability to enable the SMEs to highly improve the quality of their services and increase their visibility in the National and International markets
The main objective of the REVERT project was to build a sophisticated computational framework based on AI to predict patient responses to combinatorial therapies for mCRC care. Since the project start, explainability was identified as a key point in REVERT, no less important than typical accuracy and performance metrics, anticipating by a few years the guidelines and obligations embodied in the EU AI Act, which is intended to promote the uptake of human-centric and trustworthy AI. Accordingly, a series of ML-based predictive models (rational and explainable) relevant to the clinical outcome of a specific treatment were developed and an innovative approach of explainable artificial intelligence (AI) to gain the trust of clinicians was developed, namely the “multi-party, responsibility-sharing AI DSS”, in which trust is built by consensus of four different AI algorithms independently developed by the four IT centres and independently applied. This is well beyond the state of the art, as it represent the first approach combining multiple algorithms working together to generate shared probabilities and treatment recommendations in real-time. The AI-DSS was validated in a proof-of-concept clinical study demonstrating the significance of the AI DSS in assisting physicians in making more informed decisions about most suitable treatment options. To the best of our knowledge, this is the first time that an AI-based prediction model is validated in a specifically designed prospective clinical trial, posing the basis for the implementation of AI-based models in the clinical workflow in patient treatment. Another objective of REVERT was to identify biomarkers relevant for treatment selection of mCRC patients as already established drugs can be more efficiently used by accurately stratifying the patient groups that can mostly benefit from a given combinatorial treatment. The approach used in REVERT can be considered beyond the state of the art as it combined a multiomics approach of data analytics, in-silico models, patient-derived CRC organoids NGS and molecular testing of liquid biopsies with clinical outcome. This approach allowed the identification of several biomarkers relevant to the selection of personalized combinatorial treatment using either already established drugs or novel drugs in mCRC patients