Développement d'une base de données cartographiques et de modèles statistiques pour la caractérisation du cartilage articulaire

Osteoarthritis (OA) is the most common form of disease causing pain and disability in millions of people. Indeed, according to the United Nations, 130 people in the world will suffer from osteoarthritis, of which 40 million will be severely disabled as a result of the disease. Known as the most common type of arthritis, OA is characterized by the degeneration of the articular cartilage. The knee is one of the joints most commonly affected by OA. Unfortunately, OA in the knee is often diagnosed when severe cartilage lesions (up to the bone) are present. In late-stage OA, treatment options are limited. Nevertheless, since OA is a disease that develops over decades, it offers a wide range of time to alter its course. Thus, the detection of cartilage quality at a pre-osteoarthritic stage will allow to initiate effective treatments at an early stage of the disease. Therefore, there is still an urgent need to design a tool providing quantitative and reliable evaluation of cartilage integrity. For more than a decade, our research group has been interested in this issue. Over the years, our researchers have studied the electromechanical properties of articular cartilage, called streaming potential. This phenomenon, which consists of electrical potentials generated by cartilage loading, led to the design of a medical device, the Arthro-BST. This device is an arthroscopic probe, with 37 microelectrodes lying on its semi-spherical indenter tip, making it possible to measure these streaming potentials. Numerous studies have shown that streaming potentials reflect the structure, function and composition of cartilage as well as being sensitive to cartilage degeneration. This technology makes it possible to accurately diagnose cartilage abnormalities and therefore, innovate the evaluation of this tissue. However, the Arthro-BST remains a tool for measuring compression-induced streaming potentials of cartilage. Thus, the general objective of this thesis is to transform the current device into a precise grading tool for cartilage quality. The first experimental study aimed at examining the correlation of the quantitative parameter (QP) of the Arthro-BST with conventional techniques of human cartilage assessment. Although several previous studies have demonstrated the correlation of streaming potentials with histological, biochemical or mechanical properties of cartilage in animal models, it was essential to study human tissues for clinical application. Non-destructive electromechanical measurements using the Arthro-BST were first performed on entire articular surfaces from human cadaveric knees. Osteochondral cores were harvested from these samples for histological, biochemical and mechanical assessments. The results of this study demonstrated that the electromechanical QP correlates strongly with the histological and mechanical properties of human cartilage while a low correlation with the content of proteoglycans and water was observed. The strong correlation between the electromechanical QP and the histological scores suggested that the streaming potentials accurately reflects the quality of the tissue and the integrity of the collagen network. Moreover, the strong correlation of the electromechanical properties with the mechanical properties demonstrated that the streaming potentials represent the functional properties of the cartilage. Subsequently, the second experimental study aimed to examine the capacity of the electromechanical device, as well as an automated indentation technique (developed by Biomomentum), to characterize entire human articular surfaces rapidly and non-destructively to distinguish between early degenerated and healthy articular cartilage. Effectively, we studied the ability of the electromechanical probe to accurately evaluate cartilage quality and the rapidity in which electromechanical data were obtained. To do this, the samples were mapped macroscopically, electromechanically, mechanically and by measuring cartilage thickness. Then, osteochondral cores were harvested from macroscopically healthy and degenerated areas for histological, biochemical and unconfined compression assessments. The macroscopic characterization allowed defining 3 distinct regions on each articular surface: region I was macroscopically degenerated, region II was macroscopically normal but adjacent to regions I and region III was the remaining normal articular surface. Each osteochondral core was assigned to one of the 3 regions and statistical analyses made it possible to reveal the sensitivity of each of these characterizations. The results demonstrated that only the electromechanical parameters obtained by the arthroscopic probe and the mechanical parameters obtained by the automated indentation technique allowed to distinguish early cartilage degeneration that appears visually normal but is adjacent to a defect. These techniques have not only shown superior sensitivity in regard to conventional techniques (histological or biochemical) but were also much faster. Thus, with these promising results, these two mapping techniques could be used in research such as cartilage repair studies where entire articular surfaces could be rapidly and non-destructively assessed or for clinical cartilage assessment using the electromechanical probe. Taken together, the encouraging results of these two experimental studies revealed the diagnostic potential of the Arthro-BST and motivated its further development into a grading tool for articular cartilage quality. First, it is important to mention that to be clinically useful, the current output of the device must be translated into a language easily implemented by surgeons. Among the existing macroscopic classification systems, the International Cartilage Repair Society (ICRS) grade was chosen. This grading system is based on cartilage lesion severity and is represented by a nominal scale of 0 to 4. Despite the subjectivity of this score, it is widely used by arthroscopists. Thus, it was agreed that the electromechanical QP would be converted into an electromechanical score analogous to the ICRS grade, which means being represented on a continuous scale ranging from 0 to 4. To do this, an electromechanical database of healthy cartilage was built to establish a baseline for the electromechanical QP. Following this, an examination of the electromechanical properties in function of cartilage degeneration (based on the ICRS score) was performed on 100 articular surfaces issued from patients undergoing total knee replacement surgeries. Statistical analysis on the effect of patient-specific and location-specific characteristics on the electromechanical QP was conducted. This analysis revealed that electromechanical properties were not affected by patient-specific characteristics within each ICRS grade, but were significantly different across the articular surface. Thus, it is crucial to consider the location-specific information during cartilage evaluation to provide an accurate diagnostic. The relationship between the electromechanical QP and the ICRS grade was established to construct a continuous and quantitative electromechanical score analogous to ICRS grade. Finally, the reliability of this novel score was assessed by comparing it with ICRS grade on a subset of human articular surfaces. The electromechanical score not only made it possible to distinguish cartilage lesions not detected macroscopically but also offered more accurate grading of the lesions. To conclude, this new grading functionality has been implemented in the Arthro-BST. Despite limitations such as the extended period of preservation of cadaveric articular surfaces and the inability to independently diagnose normal or much degenerated cartilage (cartilage thinning), this project revolutionizes the assessment of this tissue by being objective and quantitative. This tool is essential to prevent, improve treatment and evaluate accurately articular cartilage. In addition, an ongoing project and complementary to this one has the overall objective of making the grading algorithm completely autonomous by analyzing the raw signals of compression-induced streaming potentials of cartilage to differentiate visually normal and severely degraded cartilage due to cartilage thinning. Keywords: Articular cartilage, knee, streaming potential, electromechanics, diagnostic, human