Contenu et objectifs du travail de recherche
Bien que différents types de complications aient été identifiées pour l’arthroplastie d’épaule, il n'existe à ce jour aucun indicateur clair permettant de les prédire. L’analyse détaillée des causes et risques de complications est difficile, notamment pour la contribution biomécanique, car celle-ci ainsi que les données cliniques sont rares et difficiles à obtenir. L'objectif principal de ce projet consiste à évaluer les risques statistiques de complications parmi un grand nombre de paramètres cliniques et biomécaniques. Des outils précédemment développés pour l’analyse de l'anatomie et biomécanique de l'épaule seront utilisés sur une série de patients ayant bénéficié d’une arthroplastie d’épaule, de façon à corréler la pathologie, le traitement et les complications avec des facteurs de risque potentiels cliniques et biomécaniques. Cette démarche sera appliquée à une cohorte multicentrique de patients ayant bénéficié d’une arthroplastie d’épaule, et complétée par un groupe contrôle de sujets sans pathologie de l'épaule. Cette base de données contiendra non seulement les données cliniques habituelles, mais aussi des données radiologiques et biomécaniques. Par conséquent, une partie importante du projet consistera à élaborer les méthodes nécessaires pour obtenir, recueillir, rassembler et analyser cette grande quantité de données.
Contexte scientifique et sociétal du projet de recherche
L'identification des paramètres responsables de l'échec de l’arthroplastie de l’épaule est nécessaire pour réduire le taux de complications. Ces paramètres, après validation, pourraient être intégrés aux scores cliniques préopératoires et contribuer à l’amélioration de la planification chirurgicale et du résultat de la chirurgie.
Background: Glenohumeral osteoarthritis is increasing worldwide, as in Switzerland, because of the aging population, leading to a steady increase in the number of total shoulder arthroplasties (TSA) observed in recent decades. This increase is also due to higher patient expectations regarding their quality of life, even after 60 years of age. However, although improvement in surgical outcome has been reported, especially for reverse TSA, significant concerns persist for both anatomical (aTSA) and reverse (rTSA) arthroplasty. Higher failure rate is reported for TSA compared to other main joint replacements such as hip or knee arthroplasties. While different complication types have been identified for both aTSA and rTSA, there is still a lack of clear indicators (markers/causes) able to predict the long-term success of these procedures. The complex biomechanical configuration of the glenohumeral joint could partly explain this and potential causes for the observed complications have been proposed, such as the pre-surgical state of rotator cuff muscles, the shape/orientation of the glenoid, or bone quality. However, evidence is scarce to support their distinct contributions to the observed complications, and their relative influence on the final clinical outcome has not yet been quantified.Objective: The main objective of this project is to perform a comprehensive analysis of all potential preoperative indicators/features, to characterize their relative contribution to the observed clinical outcomes, and determine which ones are critical to the success of TSA.Previous work: Over the past decade, the applicants have developed several methods and tools to analyze shoulder anatomy, and build patient-specific predictive models of the shoulder joint after TSA. This expertise will be integrated into a common framework enabling an in-depth evaluation of novel biomechanical markers on a large patient cohort.Methods: A multicenter cohort study will retrieve/recruit all patients who underwent TSA from five medical centers. This patient cohort will be supplemented with patients without shoulder disorders as a control group, to determine normative reference values. We target to include ~1000 pathological cases and ~1000 controls. This database will not only contain usual clinical data (patient characteristics, diagnosis, comorbidities, treatment, complications, and outcome), but also key radiological, anatomical, and biomechanical data. Therefore, an important part of the project will be to develop the methods needed to obtain, collect, assemble, and analyze this large amount of data. Specifically, we will develop and integrate automated methods to extract anatomical features and biomechanical properties from radiological images, using techniques such as statistical shape modeling and machine learning. This information will not only be used as anatomical markers, but will also constitute an input to personalize musculoskeletal and finite element models. These patient-specific numerical models will replicate the surgical procedure and therefore provide novel biomechanical markers, which were not previously accessible to clinicians, such as muscle elongations and forces during activities of daily living, humeral head translation and subacromial space, impingements between bones and implants, or bone deformations and stress states at bone-implant interfaces. The large amount of data collected for each patient will then be statistically analyzed and correlated with clinical outcome using multivariate proportional hazards regressions and machine learning methods.Significance: Identifying the markers responsible for aTSA and rTSA failure is required to improve surgical outcome and help decreasing the relatively high complication rate currently reported for TSA. These markers could eventually be integrated into preoperative clinical scores and improve surgical planning, helping surgeons to optimize their surgical procedure, implant choice and its positioning. The results of the project also offer the perspective of proposing a virtual test bench to further personalize the surgical treatment and develop novel implants.
Last updated:18.07.2023
