A longitudinal study of Type-B aortic dissection and endovascular repair scenarios: Computational analyses

Abstract

Conservative medical treatment is commonly first recommended for patients with uncomplicated Type-B aortic dissection (AD). However, if dissection-related complications occur, endovascular repair or open surgery is performed. Here we establish computational models of AD based on radiological three-dimensional images of a patient at initial presentation and after 4-years of best medical treatment (BMT). Computational fluid dynamics analyses are performed to quantitatively investigate the hemodynamic features of AD. Entry and re-entries (functioning as entries and outlets) are identified in the initial and follow-up models, and obvious variations of the inter-luminal flow exchange are revealed. Computational studies indicate that the reduction of blood pressure in BMT patients lowers pressure and wall shear stress in the thoracic aorta in general, and flattens the pressure distribution on the outer wall of the dissection, potentially reducing the progressive enlargement of the false lumen. Finally, scenario studies of endovascular aortic repair are conducted. The results indicate that, for patients with multiple tears, stent-grafts occluding all re-entries would be required to effectively reduce inter-luminal blood communication and thus induce thrombosis in the false lumen. This implicates that computational flow analyses may identify entries and relevant re-entries between true and false lumen and potentially assist in stent-graft planning.

Introduction

Aortic dissection (AD) is one of the most dreadful cardiovascular diseases associated with high morbidity and mortality rates [1]. It is initiated by a tear or damage to the intima of the aortic wall, and followed by a surge of blood flowing into the aortic wall, splitting the original single lumen of the aorta into a true (TL) and false lumen (FL). Patients with uncomplicated Stanford Type-B AD (dissection begins distal to the supraaortic branches) are usually followed by best medical treatment (BMT) [2]; while, with the enlargement of aortic diameter or the presentation of disease related complications such as rupture or vital organ ischemia, endovascular repair or open surgery is required [3]. Although a variety of investigations have been conducted to improve management of AD, long-term prognosis often remains unsatisfactory; enlargement of aortic diameter, recurrence of dissection, and rupture are difficult to predict and to control [2].

Recently, computational hemodynamics has been increasingly used in analyzing cardiovascular diseases to assist clinical researches [4], [5]. A few of these studies have focused on AD, based on either artificially designed geometries [6] or patient-specific models [7], [8], [9]. These studies provided quantitative assessments of the hemodynamics in AD and improved our understandings about the pathogenesis of this disease. However, very few of them focused on the development of AD during treatment. Longitudinal studies targeting at the geometric and hemodynamic change of AD are still needed, which may provide insights on the progress of this disease and assist in decision-making of potential interventional treatments.

In this study, we establish computational models based on radiological scans of a patient at initial presentation (IP) and after 4-year follow-up (FU), when the blood pressure of the patient has been restricted to a normal level by BMT. The luminal geometric change was identified by comparing aortic diameter, volume, and tear sizes, between the IP and FU models. The inflow of the ascending aorta in the FU model was reduced in order to simulate the reduced left ventricular contractions by anti-hypertension treatment. Detailed flow and loading information of the AD system were provided by solving the 3D unsteady conservation equations for mass and momentum. Velocity, pressure and wall shear stress distributions were reported and salient hemodynamic features of the blood exchange between the TL and the FL were identified and compared between the two models. Finally, scenarios of potential endovascular repair were investigated by virtually occluding different tears in the FU model.