THE METASTRA PROJECT: DEVELOPMENT AND VALIDATION OF COMPUTATIONAL TOOLS TO STRATIFY PATIENTS WITH VERTEBRAL METASTASES AT RISK OF FRACTURE

Le résumé fourni par la source

Patients surviving with cancer frequently develop metastases in their remaining lifetime. Metastases affect the spine in 30–70% cases, compromising the affected vertebrae. Vertebral fractures occur in 30% patients. Currently, clinicians can rely on their sole experience when choosing between operating to stabilise the spine (often exposing patients to unnecessary surgeries), and leaving the patient exposed to a high fracture risk (with dramatic consequences). The standard-of-care are scoring systems (e.g. Spine Instability Neoplastic Score, SINS), which give little consideration to the spine biomechanics, and the structure of the vertebrae affected. Such scoring systems fail to provide clear indications in 60% patients. Aim of the METASTRA project: to improve the stratification of patients with vertebral metastases by evaluating their risk of fracture through a combination of dedicated reliable computational models. Methods: The HEU-funded METASTRA project relies on the expertise of biomechanicians, modellers, clinicians, experts in verification, validation, uncertainty quantification and certification from 15 partners across Europe. The model based on Explainable Artificial Intelligence (AI) is expected to stratify most patients with limited effort, based on parameters semi-automatically extracted from their medical records and images. The cases which are not reliably stratified through the AI, are examined through more detailed personalised physiology-based biomechanical (VPH) models. Results so far: the METASTRA project is now in its second year (of five). An unprecedented retrospective multicentric retrospective study is being completed, with medical files and clinical images of over 2000 metastatic patients. This is allowing to train the AI models. In parallel, an ex-vivo study on 120 cadaveric metastatic spine segments is running, collecting input multiscale data (morphology, strains, displacements, failure loads and mode of failure) to build detailed personalised VPH models. The current standard-of-care often results in under-treatment (≈30% cases, where 90% of such fractures need surgical treatment), or over-treatment (≈30% patients). The METASTRA decision support system is expected to cut down the indeterminate diagnoses from the current 60% down to 20%of cases, with significant benefits for the patient population. Also, significant cost saving is expected. The current cost related to metastatic vertebral fracture and unneeded surgery is around 11100€/year in Europe. METASTRA will allow better use of resources by: (i) reducing unneeded surgeries (a saving of ≈1400M€/year is estimated in the EU), and (ii) preventing fractures (saving ≈1000M€/year thanks to timely surgery instead of late surgery). Acknowledgements: METASTRA project funded by the European Union, HEU topic HLTH-2022-12-01, grant 101080135.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.