| dc.description.abstract | Degenerative mitral valve disease impacts around 2% of the population and leads to mitral regurgitation because of valvular incompetence. Two distinct phenotypes of degenerative mitral valve disease are fibroelastic deficiency and diffuse myxomatous degeneration, commonly referred to as Barlow’s disease. These phenotypes both show significant microstructural changes, albeit with different pathophysiological mechanisms. The disrupted microstructure is associated with mechanically incompetent leaflets, which affect the function of the mitral valve.
Numerical simulations can complement the clinical evaluation of mitral regurgitation caused by degenerative mitral valve disease, which conventionally relies on descriptive assessments. However, accurate numerical simulations require accurate material modeling. This is essential to represent the altered state of the degenerative mitral valve leaflets in comparison to a healthy and functional mitral valve. Achieving this demands an understanding of mechanical behavior, insights into microstructural changes, and the application of suitable constitutive models.
In this dissertation, we adopt a multidisciplinary protocol to examine the collagen structure of the mitral valve leaflets and its mechanical behavior. We utilized second harmonic generation microscopy in combination with tissue clearing to facilitate in-depth imaging of collagen fibers. Additionally, a custom-designed planar biaxial machine was fabricated to characterize the mechanical behavior of the mitral valve leaflets. This protocol was applied to porcine mitral valve leaflets, providing evidence of how collagen fiber distribution dictates the mechanics of the mitral valve leaflets. An appropriate constitutive model was also employed for the associated tissue modeling, showing good agreement (average r2 = 0.94). Moreover, it was found that the collagen structure could be represented by a mean direction and a dispersion with a single family of fibers, despite variations in collagen fiber direction and dispersion across the entire thickness of the mitral valve leaflets.
Next, we improved the methodology by first combining forward and backward scattering from second harmonic generation microscopy to further facilitate collagen imaging up to 2500μm, and second, by using a morphology-preserving tissue clearing. We applied the protocol to tissue samples harvested peroperatively from one patient diagnosed with fibroelastic deficiency and another patient diagnosed with Barlow’s disease. Both Barlow’s disease and fibroelastic deficiency samples showed a high dispersion value, indicating collagen remodelling. We also developed the first disease-specific finite element model from echocardiography of patients diagnosed with fibroelastic deficiency and Barlow’s disease. This model incorporated the diseased geometry, boundary condition, and material model. With the detailed information on the structure of collagen fibers in the diseased samples, we employed a constitutive model that could exclude collagen fibers under compression. As collagen fibers are wavy, it is intuitive to assume they bear no load under compression. This is very important in the mitral valve, which is folded during systole to ensure complete closure. It was shown that excluding compressed collagen fibers improved the point-to-mesh distance error. The average error value for the simulation of Barlow’s disease was 2.02±1.8mm and 2.37±2.28 mm, with and without fiber exclusion, respectively. For fibroelastic deficiency, this value was 1.05 ± 0.79mm and 1.14 ± 0.86 mm, again with and without fiber exclusion, respectively.
Finally, we examined the collagen structure of twenty diseased mitral valve leaflets from eighteen patients, nine diagnosed with fibroelastic deficiency and nine with Barlow’s disease. Additionally, six control samples were obtained postmortem from three individuals and their collagen structure were examined. Due to the more conservative tissue resections in current practice, obtaining sufficiently large samples for mechanical testing has become challenging. However, we have shown that the collagen fibers govern the mechanical competency of mitral valve leaflets. Therefore, this study provides valuable insights into the mechanics of diseased mitral valve leaflets and also the remodelling in degenerative mitral valve disease. The collagen fiber dispersion in the control and diseased samples was significantly different. In the control group, the collagen fibers were highly aligned, whereas they were completely dispersed in the diseased samples, particularly in cases of Barlow’s disease. Moreover, for the diseased samples, two distinct preferred orientations were observed. The angle between the preferred orientations in the diseased samples differed significantly from the control (p < 0.00001), whereas there was no significant difference between the Barlow’s disease and fibroelastic deficiency samples (p > 0.38). In fact, it was observed that the two preferred orientations in the control tissue are close and, in some cases, merge into a single family of collagen fibers. However, in the diseased samples, there are two distinct preferred orientations with an average angle of almost 40˚.
In an era where a multitude of treatment solutions are being presented, a more comprehensive understanding of the remodeling processes and the biomechanics of degenerative mitral valves could facilitate customized repairs, aiming not only to achieve optimal blood flow dynamics but also to restore the normal biomechanics of the mitral valve. It may also provide pathophysiological explanations for the differences observed in fibroelastic deficiency and Barlow’s disease, as the main phenotypes of degenerative mitral valve disease. | en_US |
