Identifying Contributors to Aneurysmal Progression in the Marfan Aorta Using a Constrained Mixture Model

Introduction: Marfan syndrome (MFS) is caused by mutations to the FBN1 gene coding the glycoprotein fibrillin-1, primarily responsible for the long-term stability of elastic fibers in the aorta as well as promoting mechanosensing and mechanoregulation of extracellular matrix by mural cells. A major cause of premature death in individuals with MFS is aortic dissection following thoracic aortic aneurysm (TAA), localized dilatations in the aorta characterized by progressive loss of structural integrity and functionality. Understanding the mechanisms driving aneurysmal dilatation in MFS is critical for improved prediction of patient outcomes. Experimental studies on animal models of MFS have uncovered key histopathological features, and computational models of aortic growth and remodeling (G&R) have further demonstrated that many mechanical and mechanobiological factors coexisting in vivo can contribute to aneurysm progression. In this study, we extend an established finite element constrained mixture model of the thoracic aorta to capture TAA progression in MFS.

Materials and

Methods: We parameterized the model using histological and biomechanical data from the two most common mouse models of MFS (Fbn1C1041G/+ and Fbn1mgR/mgR), exhibiting a spectrum of aortic dilatation in response to aging and exogenous interventions. In situ two-photon microscopy and histological staining revealed alterations to aortic composition and microstructure, and biaxial mechanical testing quantified biomechanical changes in the thoracic aorta. In particular, strong correlations between elastin degradation and decreased elastic energy storage, increased circumferential stiffness, and increased luminal diameter were observed. We modeled these geometric and mechanical relationships using a 3D finite element implementation of a mechanobiologically equilibrated constrained mixture model by prescribing localized insults on a thick-walled cylindrical mesh to estimate long-term steady-state G&R evolutions, considering individual and combined contributors (compromised elastic fiber integrity, reduced collagen fiber cross-linking, and lost mechanosensing and mechanoregulation of extracellular matrix). Sequential simulations then allowed the construction of pseudo-time TAA progressions. Finally, we determined optimal insult parameters by fitting the model-predicted to experimentally-measured metrics with a surrogate management framework algorithm.

Results, Conclusions, and Discussions: Our modeling results indicated that individual contributors alone (e.g., reduced elastic fiber integrity) could not recapitulate the biomechanical data; rather, when multiple factors were considered in combination, the model was able to describe the data, confirming that dysfunctional mechanosensing and mechanoregulation of collagen fibers also play central roles in TAA progression. Fitting changes in stored energy and circumferential stiffness, critical measures of aortic function, we further found that the model successfully predicted other mechanical and geometric metrics not included in the optimization (Fig. 1). Optimal insult parameters further suggested a high model sensitivity to cellular mechanosensing and mechanoregulation relative to reduced integrity of elastic and collagen fibers, highlighting the importance of inter- and intra-lamellar elastin in facilitating cells' ability to assess and assemble the extracellular matrix. Aneurysmal dilatation in MFS thus results from multiple simultaneous contributors to progressive degeneration of the aortic wall, and computational mechanobiological models can help to disentangle these contributions.

Acknowledgements (Optional):