Biomechanical and Clinical Studies in EndoVascular Aortic Repair

Abstract

Objectives This thesis investigates biomechanical and clinical performances of endovascular repair for thoracic aortic dissection (AD) and aneurysm. Insights from both medical and bio-engineering perspectives are pursued with the aim of providing scientific data that will help guide endovascular aortic treatment. Materials and Methods All patients enrolled in the International Registry of Acute Aortic Dissection from 1996 through 2015 were reviewed to better stratify patients with retrograde dissection extension. Four-dimensional imaging was used to quantify thoracic aortic pulsatility, before and after thoracic endovascular aortic repair (TEVAR). A new formal definition is proposed, defining changes in length as the result of a force pulling up and down the proximal end of the aorta, referred to as extensibility. In addition, we quantified changes in longitudinal and radial aortic strains induced by TEVAR, using an ex-vivo porcine model as well as uniaxial tensile testing. Modern computational modelling techniques were used to study adverse cardiovascular modelling and aortic thrombosis following TEVAR. Results Acute retrograde dissection into the arch or ascending aorta was not uncommon, occurring in about 7% [101/1433] of acute AD cases. Early mortality was comparable for those managed with medical, surgical, or endovascular treatment. A favourable early mortality was observed in patients with retrograde extension limited to the arch (8.6%) versus into the ascending aorta (18.6%). Four-dimensional imaging revealed considerable aortic deformations throughout the thoracic aorta in both the longitudinal and circumferential direction, with notable differences between the ascending and descending aorta. Moreover, we found that TEVAR was followed by an increase of pulsatile strains in the non-stented aortic segments. This was most evident in the longitudinal direction of the arch, with an average strain increase of 77% in thoracic aneurysm patients managed with TEVAR. Our experimental data showed that TEVAR stiffened the aorta by two-fold. This resulted in a proximal strain mismatch between the non-stented and stented aorta. Tensile testing revealed that aortic tissue was most vulnerable in the longitudinal direction, in particular proximally close to the arch. Assessing adverse cardiovascular remodelling following TEVAR, we found a significant33% increase of left ventricular mass after one-year follow-up. Computational modelling showed that regions of high virtual thrombus potential were associated with aortic thrombosis, while virtual surgical repair and branched stent-grafting both reduced such thrombus potential. Conclusion A subset of patients with acute retrograde AD originating from a tear in the descending aorta might be managed non-operatively with acceptable in-hospital results, particularly among those with proximal extension limited to the arch. Significant aortic pulsatility highlights the importance of dynamic imaging for stent-graft sizing. Additionally, the acute strain mismatch between stiff stent-grafts and elastic aortic tissue might play a role in TEVAR-related complications. Moreover, TEVAR was associated with significant aortic stiffening. This diminishes the Windkessel function, increasing impedance and cardiac workload, as observed in this thesis. These data highlight the need for more compliant devices, in particular for TEVAR of the highly pulsatile proximal thoracic aorta. Finally, computational modelling enriches our understanding of hemodynamic complex cases and may help guide repair strategies.