EuroPOND will develop a data-driven statistical and computational modeling framework for neurological disease progression. This will enable major advances in differential and personalized diagnosis, prognosis, monitoring, and treatment and care decisions, positioning Europe as world leaders in one of the biggest societal challenges of 21st century healthcare. The inherent complexity of neurological disease, the overlap of symptoms and pathologies, and the high comorbidity rate suggests a systems medicine approach, which matches the specific challenge of this call. We take a uniquely holistic approach that, in the spirit of systems medicine, integrates a variety of clinical and biomedical research data including risk factors, biomarkers, and interactions. Our consortium has a multidisciplinary balance of essential expertise in mathematical/statistical/computational modelling; clinical, biomedical and epidemiological expertise; and access to a diverse range of datasets for sporadic and well-phenotyped disease types. The project will devise and implement, as open-source software tools, advanced statistical and computational techniques for reconstructing long-term temporal evolution of disease markers from cross-sectional or short-term longitudinal data. We will apply the techniques to generate new and uniquely detailed pictures of a range of important diseases. This will support the development of new evidence-based treatments in Europe through deeper disease understanding, better patient stratification for clinical trials, and improved accuracy of diagnosis and prognosis. For example, Alzheimer’s disease alone costs European citizens around €200B every year in care and loss of productivity. No disease modifying treatments are yet available. Clinical trials repeatedly fail because disease heterogeneity prevents bulk response. Our models enable fine stratification into phenotypes enabling more focussed analysis to identify subgroups that respond to putative treatments.

Myasthenia Gravis (MG) is a prototypic autoimmune disease causing muscle weakness and fatigability, mostly treated by chronic immunosuppressive (IS) therapy. MG clinical heterogeneity, fluctuating symptoms with unpredictable disease course, and inter-individual variation in response to treatments, including conventional IS and the emerging biological drugs, highlight the need to adopt safe, predictive and preventive personalised medicine (PM) strategies, still lacking in MG. The MG-PerMed project will employ an interdisciplinary approach, combining pre-clinical, clinical, artificial intelligence (AI) and bioethic research, to achieve PM for MG. Pharmacogenetic, pharmaco-miR and serological biomarkers associated with clinical features and response to therapies will be validated in three different MG populations (Italian, French and Israeli). Integration of biological and real-world clinical data by AI will lead to the development of a clinical decision support tool (MG-CDST), to guide clinicians in the choice of the best therapeutic program for individual patients/patient subgroups, enabling both early prediction of the optimal therapy and on-treatment disease monitoring to prevent MG symptom worsening and crisis. For the first time, a MG-CDST-based PM approach will be prospectively validated. The project outcomes promise to significantly change the MG treatment flow-chart, shifting from the “one-fits-all” approach to personalised care. MG-CDST adoption into the clinical practice should prospectively lead to: treatment failure prevention, prevention of IS drug-related adverse events, prevention of disease exacerbations, and in turn an improved MG patient compliance to treatment and a better quality of life. MG-dedicated Apps will allow patient/caregiver involvement in therapeutic decisions. Data dissemination will promote PM adoption in consensus guidelines for MG, and potentially other autoimmune diseases in which current treatment is chronic immunosuppression.