Abdominal aortic aneurysm (AAA) disease involves a dilation of the aorta below the renal arteries. If the aneurysm becomes sufficiently dilated and tissue strength is less than vascular pressure, rupture of the aorta occurs entailing a high mortality rate. Despite improvements in surgical technique, the mortality rate for emergency repair remains high and so an accurate predictor of rupture risk is required. Inflammation and the associated recruitment of monocytes into the aortic wall are critical in the pathology of AAA disease, stimulating the degradation and remodeling of the vessel wall. Areas with high concentrations of macrophages may experience an increase in tissue degradation and therefore an increased risk of rupture. Determining the magnitude and distribution of monocyte recruitment can help us understand the pathology of AAA disease and add spatial accuracy to the existing rupture risk prediction models. In this study finite element computational fluid dynamics simulations of AAA haemodynamics are seeded with monocytes to elucidate patterns of cell deposition and probability of recruitment. Haemodynamics are first simulated in simplified AAA geometries of varying diameters with a patient averaged flow waveform inlet boundary condition. This allows a comparison with previous experimental investigations as well as determining trends in monocyte adhesion with aneurysm progression. Previous experimental investigations show a transition to turbulent flow occurring during the deceleration phase of the cardiac cycle. There has thus far been no investigation into the accuracy of turbulence models in simulating AAA haemodynamics and so simulations are compared using RNG κ − ε, κ − ω and LES turbulence models. The RNG κ − ε model is insufficient to model secondary flows in AAA and LES models are sensitive to inlet turbulence intensity. The probability of monocyte adhesion and recruitment depends on cell residence time and local wall shear stress. A near wall particle residence time (NWPRT)model is created incorporating a wall shear stress-limiter based on in vitro experimental data. Simulated haemodynamics show qualitative agreement with experimental results. Peaks of maximum NWPRT move downstream in successively larger geometries, correlating with vortex behaviour. Average NWPRT rises sharply in models above a critical maximum diameter. These techniques are then applied to patient-specific AAAs. Geometries are created from CT slices and velocity boundary conditions taken from Phase Contrast-MRI (PC-MRI) data for 3 patients. There is no gold standard for inlet boundary conditions and so simulations using 3 velocity components, 1 velocity component and parabolic flow profiles at the inlet are compared with each other and with PC-MRI data at the AAA midsection. The general trends in flow and wall shear stress are similar between simulations with 3 and 1 components of inlet velocity despite differences in the nature and complexity of secondary flow. Applying parabolic velocity profiles, however, can cause significant deviations in haemodynamics. Axial velocities show average to good correlation with PC-MRI data though the lower magnitude radial velocities produce high levels of noise in the raw data making comparisons difficult. Patient specific NWPRT models show monocyte infiltration is most likely at or around the iliac bifurcation.

Abdominal aortic aneurysm (AAA) is a chronic inflammatory disease with no available pharmacological treatment. AAA formation reduces the structural integrity of the vessel and increases the susceptibility to rupture. The inflammatory response within human aneurysmal tissue is characterized by increased expression of cyclooxygenase-2 (COX-2). Similarly, in a mouse model of the disease induced by chronic Angiotensin II (AngII) infusion, we have shown that COX-2 expression in the abdominal aortic smooth muscle layer increases early in the development of the disease. Furthermore, genetic or pharmacological inactivation of COX-2 prior to disease initiation reduces AAA incidence. The current study utilized nonhyperlipidemic mice to determine the effectiveness of COX-2 inhibition initiated after AAA formation. COX-2 inhibitor treatment was initiated 5 days after beginning the AngII infusion, a time-point where significant aneurysmal pathology is observed. COX-2 inhibition with celecoxib significantly reduced the incidence as well as severity of AAAs as compared to the control group. Celecoxib treatment also protected the mice from aortic rupture and death. AAA development is characterized by degradation of the aortic smooth muscle layer with loss of the contractile phenotype. We found that the effectiveness of celecoxib was associated with significantly increased mRNA expression of alpha-actin, SM22alpha and desmin, all of which are markers of a differentiated smooth muscle cell phenotype. Celecoxib treatment also decreased mRNA expression of a marker of dedifferentiated smooth muscle (hyaluronic acid synthase 2). We also examined the role of altered expression of COX-2 in the increased susceptibility of the abdominal segment to AAA formation. We found a prolonged and greater induction of COX-2 in the abdominal aortic smooth muscle layer in contrast to a transient induction of COX-2 in the other regions of the aorta throughout disease progression. Overall, these findings suggest that COX-2 plays an important role in AAA development in mice, and COX-2 inhibition with celecoxib attenuates progression of aneurysm development by maintaining a differentiated phenotype in abdominal aortic smooth muscle cells.

Background Although abdominal aortic aneurysm (AAA) growth is non-linear, serial measurements of aneurysm diameter are the mainstay of aneurysm surveillance and contribute to decisions on timing of intervention. Aneurysm biology plays a key part in disease evolution but is not currently routinely assessed in clinical practice. Magnetic Resonance Imaging (MRI) and Positron Emission Tomography-Computed Tomography (PET-CT) provide insight into disease processes on a cellular or molecular level, and represent exciting new imaging biomarkers of disease activity. Macrophage-mediated inflammation may be assessed using ultrasmall superparamagnetic particles of iron oxide (USPIO) MRI and the PET radiotracer 18FSodium Fluoride (18F-NaF) identifies microcalcification which is a response to underlying necrotic inflammation. The central aim of this thesis was to investigate these imaging modalities in patients with AAA. Methods and Results USPIO MRI: MULTI-CENTRE STUDY In a prospective multi-centre observational cohort study, 342 patients (85.4% male, mean age 73.1±7.2 years, mean AAA diameter 49.6±7.7mm) with asymptomatic AAA ≥4 cm anteroposterior diameter underwent MRI before and 24-36 hours after intravenous administration of USPIO. Colour maps (depicting the change in T2* caused by USPIO) were used to classify aneurysms on the basis of the presence of USPIO uptake in the aneurysm wall, representing mural inflammation. Intra- and inter-observer agreement were found to be very good, with proportional agreement of 0.91 (kappa 0.82) and 0.83 (kappa 0.66), respectively. At 1 year, there was 29.3% discordant classification of aneurysms on repeated USPIO MRI and at 2 years, discordance was 65%, suggesting that inflammation evolves over time. In the observational study, after a mean of 1005±280 days of follow up, there were 126 (36.8%) aneurysm repairs and 17 (5.0%) ruptures. Participants with USPIO enhancement (42.7%) had increased aneurysm expansion rates (3·1±2·5 versus 2·5±2·4 mm/year; difference 0·6 [95% confidence intervals (CI), 0·02 to 1·2] mm/year, p=0·0424) and had higher rates of aneurysm rupture or repair (69/146=47·3% versus 68/191=35·6%; difference 11·7%, 95% CI 1·1 to 22·2%, p=0·0308). USPIO MRI was therefore shown to predict AAA expansion and the composite of rupture or repair, however this was not independent of aneurysm diameter (c-statistic, 0·7924 to 0·7926; unconditional net reclassification -13·5%, 95% confidence intervals -36·4% to 9·3%). 18F-NaF PET-CT: SINGLE-CENTRE STUDY A sub-group of 76 patients also underwent 18F-NaF PET-CT, which was evaluated using the maximum tissue-to-background ratio (TBRmax) in the most diseased segment (MDS), a technique that showed very good intra- (ICC 0.70-0.89) and inter-observer (ICC 0.637-0.856) agreement. Aneurysm tracer uptake was compared firstly in a case-control study, with 20 patients matched to 20 control patients for age, sex and smoking status. 18F-NaF uptake was higher in aneurysm when compared to control aorta (log2TBRmax 1.712±0.560 vs. 1.314±0.489; difference 0.398 (95% CI 0.057, 0.739), p=0.023), or to non-aneurysmal aorta in patients with AAA (log2TBRmax 1.647±0.537 vs. 1.332±0.497; difference 0.314 (95% CI 0.0685, 0.560), p=0.004). An ex vivo study was performed on aneurysm and control tissue, which demonstrated that 18F-NaF uptake on microPET-CT was higher in the aneurysm hotspots and higher in aneurysm tissue compared to control tissue. Histological analysis suggested that 18F-NaF was highest in areas of focal calcification and necrosis. In an observational cohort study, aneurysms were stratified by tertiles of TBRmax in the MDS and followed up for 510±196 days, with 6 monthly serial ultrasound measurements of diameter. Those in the highest tertile of tracer uptake expanded more than 2.5 times more rapidly than those in the lowest tertile (3.10 [3.58] mm/year vs. 1.24 [2.41] mm/year, p=0.008) and were also more likely to experience repair or rupture (15.3% vs. 5.6%, log-rank p=0.043). In multivariable analyses, 18F-NaF uptake on PET-CT emerged as an independent predictor of AAA expansion (p=0.042) and rupture or repair (HR 2.49, 95% CI1.07, 5.78; p=0.034), even when adjusted for age, sex, body mass index, systolic blood pressure, current smoking and, crucially, aneurysm diameter. Conclusion These are the largest USPIO MRI and PET-CT studies in AAA disease to date and the first to investigate 18F-NaF. Both USPIO MRI and 18F-NaF PET-CT are able to predict AAA expansion and the composite of rupture and repair, with 18F-NaF PETCT emerging as the first imaging biomarker that independently predicts expansion and AAA events, even after adjustment for aneurysm diameter. This represents an exciting new predictor of disease progression that adds incremental value to standard clinical assessments. Feasibility and randomised clinical trials are now required to assess the potential of this technique to change the management and outcome of patients with AAA.

The aim of this thesis is the problematic of abdominal aortic aneurysm wall stress in relation to a different material behavior of intraluminal thrombus, which is in most cases present in the aneurysmal volume. In this thesis, the influence of neglecting the patient-specific material properties of the intraluminal thrombus, on the aneurysmal wall stress, obtained from finite element stress-strain analysis, is investigated. In terms of solution method selection, a system approach was applied so that the solution method was selected in order to respect a system of essential variables as much as possible. The first part of this thesis is focused on a description of the problematic and the human cardiovascular system with important aspects contributing to development and growth of the abdominal aortic aneurysm. Next, this part of the thesis includes chapters devoted to the intraluminal thrombus in terms of its basic characteristics (anatomy, physiology, pathology), structure and its influence on processes within the abdominal aneurysm. The second part of this work is devoted to the accomplishment of the first and second goal of this thesis, which is analyzing the available literature to obtain mean population stiffness values of the intraluminal thrombus and conducting biaxial experimental tests of provided samples of intraluminal thrombus. The experimental testing was conducted in order to obtain the patient-specific mechanical properties, which are used as the inputs in the finite element analysis. The experimental testing confirmed the stiffness negligibility of the intraluminal thrombus’s outer layer, which is mentioned is several studies, however, the influence of this layer on resulting aneurysmal wall stress has been to this date not tested. The dominant part of this thesis is focused on the third goal of this work, which is a comparison of aneurysmal wall stress obtained from the finite element computation that included mechanical properties of intraluminal thrombus obtained either from the literature analysis of experimental testing. This part includes discerption of idealized geometry model development, which was used to analyze the sensibility of computed stresses on a number of ILT layers representing different material properties. In order to obtain this analysis, a macro was created prescribing each element of the intraluminal thrombus finite element mesh with material properties derived from its distance from the lumen. Next, this chapter contains description of patient-specific geometry models development, material models, and boundary conditions selection. In the end of this part, results of the finite element computations are presented together with their statistical analysis. Within the last part of this thesis, discussion of results and conclusions of this thesis is included. Also, an overview of important aspects entering computational modeling of abdominal aortic aneurysms is presented.