Re: Blood flow analysis of the aortic arch using computational fluid dynamics Free

Common diseases of the thoracic aorta such as dilatation, aneurysm or valve abnormalities can result in severe consequences such as aortic rupture or dissection. Patient management therefore often requires surgical intervention including the repair or replacement of the aortic valve and segments of the thoracic aorta itself. Surgical intervention and aortic repair have been highly successful with excellent patient outcome. However, the specific patterns of aortopathy are highly variable between patients, resulting in heterogeneous clinical phenotypes affecting distinct segments of the aorta. Despite the complex and heterogeneous nature of aortic disease, contemporary guidelines for surgical intervention are based solely on simple geometry parameters (maximal aortic diameter and growth rate) to guide patient management and surgical repair of the aorta. As a result, the timing of the intervention and aortic resection strategies are based on empirically obtained parameters and patient-specific risk stratification is lacking. A number of recent studies have shown that altered aortic haemodynamics in patients with common aortic diseases have the potential to provide a more individualized understanding of aortic pathology. Specifically, altered flow patterns such as helix and vortex flow and their impact on the aortic wall as quantified by wall shear stress (WSS) may represent novel diagnostic tools for improved patient-specific evaluation of complex aortic pathologies.

In this context, the recent paper ‘Blood Flow Analysis of the Aortic Arch Using Computational Fluid Dynamics’ by Numata et al. presents the use of computational fluid dynamics (CFD) for the detailed numerical simulation of changes in aortic haemodynamics in patients with aneurysmal segments of the thoracic aorta [ 1 ]. Computational studies allow for flexible, parametric investigations to test the effects of isolated factors (such as valve morphology) on metrics of interest or prototype surgical intervention [ 2 , 3 ]. In their study, they used patient-specific computer models of the thoracic aorta in combination with CFD to generate 3D visualization of changes in aortic blood flow patterns and to quantify regional changes in WSS, which are known to be implicated in vascular remodelling. Their findings in a feasibility study with 6 patients clearly demonstrate the potential of advanced imaging and computational techniques such as patient-specific CFD for the improved evaluation of aortic pathologies. In addition, the findings illustrate the advantage of volumetric 3D flow analysis by CFD—which cannot be achieved by standard diagnostic tools such as Doppler echocardiography or 2D phase contrast magnetic resonance imaging (MRI)—allowing for the evaluation of the impact of a focal pathology (segmental dilatation) on changes in 3D haemodynamics affecting the entire aorta.

In addition to modelling aortic blood flow, CFD permits simulation and outcome prediction for different types of vascular interventions under well-controlled conditions that imitate blood flow within the human circulation. The study by Numata et al. includes an interesting application of this approach in 1 patient by simulating various approaches to intraoperative management with different blood flows during right sub-clavian arterial perfusion. Although data in larger patient cohorts coupled with patient outcomes for different interoperative approaches are still needed, the approach and findings presented in their study are highly innovative and have the potential to help vascular surgeons better understand the blood flow characteristics associated with particular treatment approaches.

While such CFD studies are capable of reproducing patient-specific haemodynamic conditions with high temporal and spatial resolution far surpassing any imaging modality, there are a number of drawbacks as acknowledged by the authors when discussing the limitations of their approach. CFD approaches alone do not guarantee the fidelity to reproduce in vivo haemodynamics, due to inherent model assumptions such as in-flow velocity profiles, blood rheology, choice of turbulence model and parameters, and the need for high-quality data for geometry and flow boundary conditions [ 2 , 3 ]. Grid resolution errors are a further possible source of uncertainty. Some of these modelling uncertainties can be quantified, but typically only with significant additional computational expense and analytical complexity [ 4 ].

As an alternative, echo cardiography and respiration-controlled 4D flow MRI have been shown to enable comprehensive assessment of cardiovascular haemodynamics in the heart and great vessels in vivo without the need for any model assumptions [ 5–7 ]. 4D flow MRI is uniquely suited for the study of complex aortic 3D blood flow as a single, free-breathing 8–12 min acquisition and can provide blood velocity in three orthogonal directions with full volumetric coverage of the aorta. A number of groups have demonstrated the high potential of in vivo 4D flow MRI techniques and efficient data analysis workflows to identify altered 3D flow characteristics related to cardiovascular pathologies, including common aortic abnormalities. These reports include the development of novel tools to quantify metrics of altered aortic haemodynamics such as WSS and changes in flow patterns [ 8 ], similar to the metrics used in the study by Numata et al.

It should be noted that both in vivo 4D flow MRI and patient-specific CFD have successfully characterized aortic haemodynamics. However, both techniques have unique complimentary advantages and limitations, such as high resolution (CFD), insensitivity to phase offsets (CFD) versus insensitivity to boundary conditions (4D flow MRI) and direct measurement of in vivo 3D blood flow velocities (4D flow MRI). Of particular interest is the ability to quantify turbulence, which has been demonstrated with MRI [ 5 ]. Given the contribution of turbulence to large pressure gradients and their suspected relationship to vascular remodelling, the combination of CFD and 4D flow MRI to measure turbulence is a particularly interesting application of these powerful tools. Future studies should aim to combine the various advantages of CFD and 4D flow to enhance the assessment of cardiovascular haemodynamics and patient-specific planning of surgical intervention.

Funding

National Institutes of Health (grant numbers R01HL115828 and K25HL119608).

REFERENCES