Imaging-based biomechanical parameters for assessing risk of aortic dissection and rupture in thoracic aortic aneurysms

Abstract

OBJECTIVES

Imaging-based methods of measuring aortic biomechanics may provide superior and a more personalized in vivo risk assessment of patients with thoracic aortic aneurysms compared to traditional aortic size criteria such as maximal aortic diameter. We aim to summarize the data on in vivo imaging techniques for evaluation of aortic biomechanics.

METHODS

A thorough search of literature was conducted in MEDLINE, EMBASE and Google Scholar for evidence of various imaging-based biomechanics techniques. All imaging modalities were included. Data involving preclinical/animal models or exclusively focussed on abdominal aortic aneurysms were excluded.

RESULTS

The various imaging-based biomechanical parameters can be divided into categories of increasing complexity: strain-based, stiffness-based and computational modelling-derived. Strain-based and stiffness-based parameters are more simply calculated and can be derived using multiple imaging modalities. Initial studies are promising towards linking these parameters with clinically relevant end-points, including aortic dissection, though work is required for standardization. Computationally derived parameters provide detail of stress exerted on the aortic wall with great spatial resolution. However, they are highly dependent on the assumptions applied to the models, such as material properties of the aortic wall.

CONCLUSIONS

Imaging-based aortic biomechanics represent a major technical advancement for personalized in vivo risk stratification of patients with ascending thoracic aortic aneurysm. The next steps in clinical translation require large-scale validation of these markers towards predicting aortic dissections and comparison against the gold standard ex vivo aortic biomechanics as well as development of a user-friendly, low-cost algorithm that can be widely adopted.

Open in new tabDownload slide

INTRODUCTION

Patients with ascending thoracic aortic aneurysms (ATAA) are at risk of aortic dissection, rupture and sudden death. A basic tenet of aortic surgery is to identify patients at high risk of these events so that elective surgery can be performed with low perioperative risks. However, the rate of adverse aortic events for those with ATAAs is low (approximately 2% per patient-year), and patient selection for surgery remains a conundrum [1]. Current clinical guidelines for elective surgery are primarily based on size criteria such as the maximum aortic diameter [2, 3]. Aortic diameter and size thresholds provide a clinically feasible and widely used tool [4]. However, despite their extensive use, aortic size thresholds are limited as many small aneurysms dissect or rupture, while many large aneurysms remain stable [5, 6]. This is likely because diameter alone may not reflect the local microstructural changes and resultant biomechanical properties of the degenerating aortic tissue [7]. Complications of ATAAs are fundamentally a biomechanical problem where tissue fails when haemodynamic stresses exceed tissue strength. Several studies have established the link between tissue biomechanical properties ex vivo and the aortic wall microstructure [8–12]. More recently, associations have been between tissue biomechanics and failure properties [13, 14] as well as arterial haemodynamics [15, 16].

However, clinical translation of aortic biomechanics requires the ability to measure biomechanical properties in vivo. A variety of medical imaging modalities have been proposed, including strain-based, stiffness-based and computational modelling-based methods (Fig. 1). Used alongside conventional risk factors, imaging biomechanics may improve and personalize risk evaluation for aortic events. This review paper defines the imaging-based aortic biomechanics, describes the methods used to measure them, outlines their relative strengths and limitations and discusses their clinical relevance and utility in predicting dissection or rupture risk.

STRAIN-BASED METHODS

Strain magnitude

Strain is defined as a material’s relative deformation when placed under stress. The aorta undergoes strain throughout the cardiac cycle as haemodynamic pressures cause cyclic stretch and recoil. Cyclic deformation of the aorta can be imaged in vivo and used to calculate maximum strain magnitude as the ratio between maximum change in aortic diameter (from end diastole to peak systole) with respect to the diameter at end diastole (Fig. 2). It has been suggested that strain magnitude may reflect the histopathology of aortic aneurysms as medial degeneration results in elastin fragmentation, decreased elastin content and increased collagen content, and these changes contribute to stiffening of the aorta and a lower maximum strain.

Imaging methods to measure strain magnitude

Transthoracic echocardiography (TTE) and transesophageal echocardiography (TEE) have been used to measure aortic deformation and subsequently quantify circumferential aortic strain [17–19]. Recently, strain echocardiography imaging methods such as tissue Doppler imaging (TDI) [20, 21] and speckle tracking [22] echocardiography have been used to measure aortic deformation [23, 24]. Circumferential aortic wall strain can also be assessed by electrocardiographic (ECG) gated cardiac computed tomography (CT) via measurement of aortic cross-sectional area near peak systole and end-diastole [25, 26]. Moreover, cardiac MR imaging using 2D phase contrast [27, 28] or gradient echo (GRE) sequences [29, 30] provide excellent temporal resolution while also enabling calculation of 2- and 3-dimensional flow velocities.

Relevance to dissection or rupture risk

Aortic strain has been used to predict clinically relevant end-points in cohorts of patients with syndromic aneurysms. Lower maximum strain magnitude was independently associated with need for prophylactic aortic root replacement in young patients with connective tissue disorders [31]. Moreover, reduced longitudinal aortic strain magnitude predicted dissection in a Marfan cohort, though circumferential strain magnitude was not independently associated with dissection risk [32]. Additionally, several groups have reported an association between aortic strain and plasma markers of aneurysm. Maximal strain significantly correlated with increased matrix metalloproteinase (MMP)-2 levels; MMP-2 degrades aortic extracellular matrix (ECM) proteins including collagen and elastin and accelerates apoptosis of vascular smooth muscle cells (VSMC), causing weakening of the aortic wall [33]. Aortic strain is also found to be lower in patients with hypertension due to the adverse remodelling of the aortic wall and subsequent arterial stiffening, further reinforcing the importance of adequate clinical blood pressure control in ATAA patients [20]. In a recent study, a novel MRI-based imaging sequence termed displacement encoding with simulated echoes (DENSE) was applied to measure reduced aortic wall stretch percentage (analogous to aortic strain) in ATAA patients compared to healthy controls. It was also reported that aortic wall stretch was correlated with ex vivo tissue properties including peak elastic modulus, though this relationship was quantified in a small group of 5 samples [34]. Furthermore, there is an association between CT-derived maximal aortic strain and proposed plasma biomarkers of aortic aneurysms such as exosomal microRNAs (e.g. miR-26a and miR-320a) in ATAA patients [35]. These relationships highlight how adverse biomechanical changes co-occur alongside changes in the aortic tissue and local aneurysm microenvironment. However, a major limitation of aortic strain is that it does not consider the role of both the ambient blood pressure which affects initial deformation of the aortic wall and variations in blood pressure across the cardiac cycle. Both of these factors can vary significantly within the same patient and across patients and in turn independently affect the aortic strain measurements.

Three-dimensional strain mapping

The ascending aorta is unique in that the strain it experiences is the result of 2 unique forces: downwards traction due to left ventricular contraction and cyclic distension due to systolic pressurization. The net effect of these forces is a complex 3-dimensional motion profile of the ascending aorta that may not be well captured by 2-dimensional strain methods described above. ECG-gated cardiac CT quantifies this 3-dimensional motion and resolves it into longitudinal, in-plane and rotational components. Several CT approaches have been described for 3-dimensional aortic strain quantification, including methods that are based on finding corresponding points in systolic and diastolic aortic mesh models [36–38], and a recently described technique termed dynamic vascular deformation mapping (VDM-D) that uses a displacement field calculated by applying deformable image registration throughout the cardiac cycle [39]. As such, vascular deformation mapping allows measurement of pulsatile changes in aortic geometry in 3 dimensions.

Methods for vascular deformation mapping

The aorta is segmented in serial CT angiography (CTA) imaging using either manual [39–41] or semi-automatic [42] methods. Pairs of CTA volumes in systole and diastole are then aligned using rigid and/or deformable registration [41, 42]. Aligned volumes are then used to calculate and generate a mesh illustrating 3-dimensional voxel-wise deformations between peak systole and diastole. Additionally, the vascular deformation mapping approach can be extended to accurately track growth of the thoracic aorta over time using serial CTA scans (VDM-G), which are subsequently aligned using image registration to quantify the 3-dimensional aortic geometry, strain and aneurysm growth over time [41–43].

Relevance to dissection or rupture risk

Early results using VDM-D have shown clear differences in aortic root motion metrics among a mixed cohort of patients with normal aortas, ascending aneurysms and ascending aorta grafts [44]. Furthermore, VDM-G has been validated to produce accurate measurement of aortic growth both within and outside of the aneurysmal segment, and growth itself may indicate biomechanical compromise and elevated risk for dissection [43]. Additionally, 3-dimensional aortic strain measured using VDM-D can be combined with pulse pressure to derive stiffness-based metrics such as distensibility (see Section 3). Aortic distensibility quantified using VDM-D and growth rates measured using VDM-G were predictive of type B aortic dissection in a small cohort of 36 Marfan patients with thoracic aortic aneurysms [45]. Overall, vascular deformation mapping has shown early promise and validating these techniques in larger populations will be critical to facilitate potential clinical translation.

STIFFNESS-BASED METHODS

Stiffness is defined as the extent to which material resists deformation in response to loading. Thus, higher stresses are required to deform a stiff vessel to the same extent as a more compliant vessel. Previous ex vivo testing has shown that aneurysmal thoracic aortic tissue is significantly stiffer than healthy aortic tissue [46]. Pathological changes in aortic tissue stiffness result from aortic remodelling, mediated in part by mechanobiological responses of vascular wall cells [47]. For example, circumferential and longitudinal stresses contribute to increased apoptosis and transition of medial VSMCs from a contractile to a proliferative, synthetic phenotype, which secrete enzymes (e.g. MMPs) that participate in ECM remodelling and preferential elastin degradation [48]. Meanwhile, extended mechanical stresses can transform fibroblasts into myofibroblasts which also contribute to ECM remodelling, excessive collagen deposition and tissue stiffening [49]. As such, in vivo imaging-based measures of stiffness are of significant interest.

Distensibility, compliance and arterial stiffness index

These related parameters quantify aortic deformation due to changes in blood pressure during the cardiac cycle (Table 1). Distensibility can be defined as maximal aortic strain per unit pulse pressure. Compliance is defined as the aortic deformation per unit pulse pressure. Therefore, stiffer materials will have lower distensibility and compliance. Arterial stiffness index (ASI—also denoted $β1$⁠) is defined as the natural log ratio of systolic and diastolic blood pressures per unit aortic strain and is therefore inversely related to compliance and distensibility.

Methods to measure distensibility, compliance and ASI

Estimating aortic deformation and blood pressure changes across the cardiac cycle are necessary for calculation of these 3 parameters (Fig. 2). The previously discussed imaging techniques have all been used to compute aortic deformation for the purpose of calculating distensibility, compliance and ASI parameters. Systolic and diastolic pressure can then be measured using brachial cuff pressure.

Relevance to dissection or rupture risk

Aortic distensibility has been used to independently predict ascending aneurysm growth in a BAV cohort, though ASI failed to predict aneurysm growth in the same study [50]. Moreover, pre-dissection distensibility was reduced and ASI was elevated in 7 patients who went on to experienced type A aortic dissections compared to non-aneurysmal, age- and size-matched controls [51]. Multiple studies also report that aortic distensibility is reduced in patients with BAV aortopathy and syndromic aneurysms compared to idiopathic degenerative aneurysms [52–54], potentially reflecting the higher propensity for disease progression, aneurysm growth and adverse aortic events in these states. Notably, a recent study demonstrated that pre-dissection ASI was elevated in patients who went on to experience type A aortic dissection compared to undissected ascending aneurysms and ASI was predictive of future dissection events with an area under the curve of 93.37% [55]. Overall, these results show promise for the predictive utility of imaging-based in vivo stiffness parameters in thoracic aortic aneurysms. Reproducing these very promising results with larger, multicentre cohorts will be the next step towards clinical translation.

Cardiac cycle pressure modulus and cardiac cycle stress modulus

Cardiac cycle pressure modulus (CCPM) and cardiac cycle stress modulus (CCSM) are in vivo stiffness parameters that are derived by measuring circumferential aortic deformation using TEE imaging and the corresponding arterial line blood pressure tracing throughout the cardiac cycle [56]. These data are used to recreate the pressure-stretch loop recreated and CCPM is then the average slope of this loop. The CCSM was then defined by applying Laplace’s law which relates pressure to circumferential wall stress [57]. Both CCPM and CCSM were strongly correlated with the ex vivo biomechanical parameter energy loss, which in turn is a measure of the viscoelastic properties of aortic tissue and correlates well histopathological changes in aneurysmal aortas and susceptibility to dissection [57]. This is important because in vivo biomechanics are rarely validated against ex vivo aortic tissue properties. While this validation is a strength, derivation of these parameters in clinical settings could be limited by the requirement of invasive blood pressure measurement.

Pulse wave velocity

Pulse wave velocity (PWV) is another in vivo parameter that characterizes vessel stiffness (Fig. 3). PWV measures the speed of fluid wave propagation through a vessel and is related to the vessel stiffness by the Moens–Korteweg equation [58]. A stiffer vessel will produce a higher PWV. Practically, PWV is measured as the speed of propagation of the systolic wave between proximal and distal sampling sites [59].

Methods to measure PWV

PWV can be measured non-invasively using applanation tonometry, measuring pulse transit time between the carotid and femoral arteries. The transit distance of the fluid wave between carotid and femoral sampling sites is estimated using the subtraction method which relies on body surface measurements [60]. The resulting tonometry-based carotid-femoral PWV (cfPWV) measurements correlate well with gold-standard invasive intra-aortic PWV measurements. However, cfPWV provides only a global measure of aortic stiffness and there may be significant segmental variations in aortic biomechanics (i.e. between the ascending, descending thoracic, and abdominal aorta) [61, 62]. In contrast, imaging methods such as echocardiography [24] and cardiac MRI [30] (Fig. 2) have been proposed to estimate transit times between any 2 sampling sites and thus enable segmental quantification of PWV. ECG gated 2D phase contrast (PC) MRI images can be acquired at 2 specified proximal and distal sampling sites to calculate the velocity profiles and arrival times of the systolic wave at both sites. Additionally, transit distance can be estimated directly in MR studies by measuring the centreline distance between sampling sites on a 2D oblique-sagittal view of the aorta. PWV measurements using 2D PC MR are reproducible with low variability across multiple centres and operators and correlate well with both tonometry-derived cfPWV and invasive intra-aortic PWV, but with lower temporal resolution [30]. More recently, studies have also used 4D flow MRI acquisitions to determine transit time [63].

Relevance to dissection or rupture risk

Elevated PWV has been associated with thoracic aneurysm development and progression. Several studies have reported elevated cfPWV is associated with faster aneurysmal growth during short-term follow-up in a BAV cohort [64] and a broader cohort of thoracic aortic aneurysms [65, 66]. Additionally, elevated PWV measured using 4D flow MRI (sampling sites selected at the sinotubular junction and the left subclavian artery) was an independent predictor of progressive ascending aorta dilatation in BAV patients [67]. Multiple studies have also reported elevated PWV in higher risk patient groups. PWV was elevated in the ascending aorta of Marfan patients with ATAA compared to non-Marfan patients with ATAA after adjusting for age, blood pressure and ascending aortic maximal diameter [67]. This again may reflect higher risk for aortic dissection in syndromic aneurysms. Moreover, PWV measured using 2D PC MRI in the proximal aorta (sampling sites in ascending and proximal descending thoracic aorta at level of pulmonary trunk) and distal regions (proximal descending thoracic aorta to aortic bifurcation) was higher in BAV patients compared to age-matched healthy controls [54]. However, a study measuring ‘global’ PWV across the entire aorta using 4D flow MRI found no difference between patients with thoracic aortopathies and healthy controls [68]. This discrepancy between results could be due to differences in biomechanics across various aortic segments. There is also significant heterogeneity in imaging techniques and sampling sites which adds to difficulty in comparing results across studies and highlights the need for further standardization for imaging-based aortic PWV. Nonetheless, MRI-based aortic PWV correlates well with cfPWV which has been standardized and validated in large cohorts [63].

PARAMETERS BASED ON COMPUTATIONAL MODELLING

Computational methods, such as finite element analysis (FEA) and computational fluid dynamics (CFD), use well-defined structural and fluid mechanics models to predict tissue failure by simulating the interaction of complex wall geometry, aortic tissue material properties, and haemodynamic loads. Computational modelling approaches have been proposed to estimate both aortic wall stresses and near wall stresses (Fig. 4). Patient-specific aortic geometry obtained from imaging data and material properties of aortic tissue are prerequisites for computational modelling. Computational methods, such as FEA, most often involve discretizing, or dividing, complex geometries into many small finite elements such as small cubes and applying mathematical equations to predict the behaviour of each small element. Then, by summing the individual behaviours of each element, the overall behaviour of the complex geometry can be predicted. Here we outline 2 parameters based on computational modelling.

Peak wall stress

The distribution of stresses acting on the aortic wall can be calculated using FEA based on the aortic geometry derived from imaging data and assumptions about the material properties. This analysis yields the magnitude and the location of circumferential and longitudinal peak wall stress (PWS), which can indicate the risk and area of probable failure, respectively.

Methods to estimate PWS

Imaging data acquired using either MR or CT angiograms can be segmented to obtain 3D patient-specific vessel geometry, which is then discretized into a mesh of finite elements [69–72]. Aortic wall thickness is difficult to measure from imaging and thus a uniform thickness is usually assumed (typically 1.7–1.8 mm) [70–73]. Previous studies assume uniform, incompressible and isotropic material properties in addition to using a hyperelastic model of aortic tissue [70–73]. Most studies then use ‘population-average’ parameters to define aortic tissue material properties because non-invasively estimating tissue properties for individual unoperated patients is challenging. Haemodynamic loads in the model can be generated from patient-specific clinically measured blood pressures. Based on the model geometry, tissue material properties and loading conditions, the stresses on the aortic wall can then be estimated using FEA software (Fig. 4).

Relevance to dissection or rupture risk

Several studies have found weak correlations between ascending aortic diameter and PWS, especially for small aneurysms [70–73]. This reflects that aortic peak wall stress is a complex marker and simpler measurements such as aortic size and diameter are inadequate and over-simplified. Preliminary data have linked elevated peak longitudinal wall stress directly to clinical end-points of most interest; it independently predicted all-cause mortality within 3 years in ATAA patients, though there were only 2 cases of aortic dissection in this cohort of 270 patients [72]. In terms of limitations, modelling assumptions used to calculate PWS can have a significant impact on results [69]. There is generally a trade-off between number of model assumptions and computational time; while more realistic models may be more desirable, these models may be too time-consuming and computationally expensive for widespread clinical use.

Peak wall shear stress

Wall shear stress (WSS) represents the near-wall tangential forces from blood flow on the vessel wall. WSS can be estimated using CFD simulations created based on imaging-derived aortic geometry or directly from 4D flow MRI (Fig. 5). WSS influences function of endothelial cells [47] and is hypothesized to influence risk of aortic dissection.

Methods to estimate peak WSS

MRI or CT angiography images of the aorta are first segmented to define a patient-specific geometry which is subsequently discretized into finite elements, as described above [74]. Computational fluid dynamics simulation requires parameterization of the fluid and vessel wall. Typically, blood is assumed to be a non-Newtonian fluid [74]. While some studies have modelled the aortic wall as rigid [15, 16], recent studies have used 3D fluid–structure interaction (FSI) modelling to simulate deformations of the aortic wall which produce significantly different WSS calculations [75]. Alternatively, 4D flow MR imaging can be used directly to obtain 3D visualization of aortic haemodynamics and estimate WSS in vivo [76–79]. However, as the spatial resolution of this technique is low, 4D flow MR imaging is commonly combined with computational modelling to provide boundary conditions for CFD simulations (Fig. 4).

Relevance to dissection or rupture risk

The magnitude of WSS (1–2 Pa) is very small compared to transmural wall stresses (300–400 kPa). Nonetheless, higher WSS measured using 4D flow MRI may predict ascending aneurysm dilatation and growth rates in BAV patients [80, 81]. This association may be underpinned by the effect of mechanotransduction of higher WSS on tissue remodelling and biomechanics mediated by reduced wall thickness, reduced elastin and reduced VSMC counts, particularly at earlier stages of disease [15, 78, 79]. However, other studies implicate abnormally low WSS in adverse degenerative changes. Lower WSS was measured using 4D flow MRI in patients with dilated ascending aorta compared to age-matched healthy volunteers [77] and low WSS was correlated with lower elastin levels and higher medial degeneration in the ascending aorta [82]. Moreover, areas of elevated in vivo WSS measured using 4D flow MRI have also been associated with adverse ex vivo tissue biomechanics including higher stiffness quantified using the low stretch tangent modulus [83]. Thus, the relationship between abnormal WSS and aneurysm risk may be ‘U-shaped’ and differ based on aortic segment, stage of disease (early vs late) and pathophysiology (e.g. BAV aortopathy). Ongoing investigation alongside standardization is required to fully characterize the relationship between WSS and adverse aortic events.

SUMMARY AND CONCLUSIONS

There have been exciting multi-pronged advances in the development of in vivo imaging parameters for aortic biomechanics, each with relative advantages and disadvantages (Table 2). Strain-based methods provide simplicity and can be measured using versatile imaging modalities; however, further work is required to verify that strain can indeed independently predict aortic events in large cohorts of both syndromic and non-syndromic ATAA. Meanwhile, stiffness-based methods such as distensibility and ASI have significantly more evidence suggesting their utility in predicting both future ATAA growth and adverse aortic events. However, for certain stiffness-based techniques such as aortic PWV, there is still some heterogeneity between definitions and measurement techniques in literature which will require further standardization across future studies. On the other hand, computational modelling potentially offers precise and high-resolution prediction of aortic wall dynamics, though significant modelling assumptions, high computational times and technical expertise required to conduct computer simulations could pose barriers to clinical translation.

There are many clinical examples in aortic clinics where in vivo imaging-based biomechanical parameters may provide power tools to assist with challenges in clinical decision-making. For example, consider a young patient with no known genetic condition and a strong family history of aortopathy and aortic dissection presenting with a 42 mm thoracic aneurysm. This patient would benefit from a personalized, quantitative and non-invasive measure of aortic biomechanics that would indicate their specific susceptibility to adverse aortic events and therefore decide on timing of surgery. In another example, consider an older adult with clinical frailty who presents with a large ascending aneurysm. In such a case, the risk of future adverse aortic events needs to be balanced with high surgical risk. In vivo biomechanics could help counsel patients on these competing risks and enable more informed, collaborative decision-making. Finally, the use of in vivo biomechanical markers would provide a quantitative end-point of aortic risk that could be used to guide frequency of clinic and imaging follow-up as well as to assess effectiveness of antihypertensive and other medical prevention strategies.

To this end, we propose a call to action to (1) emphasize greater collaboration between research and clinical groups across multiple centres with biomechanics expertise to enable greater coordination and design larger and diverse multicentre prospective cohort studies. Such collaborative efforts would quickly drive this field forward as opposed to individual small studies in silos. (2) Furthermore, increased collaboration would facilitate the standardization of methods required for validation and translation in vivo biomechanics to clinical practice. Standard imaging protocols and biomarker definitions (i.e. sampling sites used for parameters such as strain and stiffness parameters) will be crucial for comparison and synthesis of evidence across studies. (3) We also call for engagement of a diverse team of multidisciplinary clinicians, scientists and engineers. Future studies need to validate in vivo biomechanical parameters against ex vivo tissue biomechanics, which will require surgical contribution to obtain tissue samples and engineering expertise test tissue mechanical properties. (4) Finally, we call for greater engagement of industry leaders and experts in future studies. Measurement of in vivo parameters must be cost-effective and easily compatible with existing hardware and software used in clinical settings, to deploy such solutions will require industry support.

While aortic diameter is still used clinically for characterizing risk of adverse aortic events in patients with thoracic aortic aneurysms, imaging-based biomechanical parameters offer a powerful set of tools to augment patient risk stratification. Ongoing research efforts are making substantial progress to address steps in the above call to action. We expect that this field will soon yield a set of non-invasive, patient-specific markers of rupture and dissection risk to guide timely and appropriate intervention in patients with thoracic aortic aneurysms.

FUNDING

This research was supported by a University of Toronto EMHSeed Award, the Canada First Research Excellence Fund (CFREF) Medicine by Design Initiative, American Association for Thoracic Surgery (AATS) Foundation Surgical Investigator Program and Melanie Munk Professorship in Aortic Biomechanical Research.

Conflict of interest: K.H. receives honoraria from Sanofi. Other authors declare no conflicts.

DATA AVAILABILITY

No new data were generated or analysed in support of this research.

Author contributions

Nitish Bhatt: Investigation; Methodology; Visualization; Writing—original draft; Writing—review & editing. Hijun Seo: Conceptualization; Investigation; Methodology; Writing—original draft; Writing—review & editing. Kate Hanneman: Investigation; Resources; Writing—review & editing. Nicholas Burris: Investigation; Writing—review & editing. Craig A. Simmons: Conceptualization; Investigation; Supervision; Writing—review & editing. Jennifer C.-Y. Chung: Conceptualization; Investigation; Supervision; Writing—review & editing

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Luca Di Marco, John Elefteriades, Santi Trimarchi and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• ASI

  Aortic stiffness index

 
• CCPM

  Cardiac cycle pressure modulus

 
• CCSM

  Cardiac cycle stress modulus

 
• CT

  Computed tomography

 
• MRDR

  Maximum rate of diastolic recoil

 
• MRSD

  Maximum rate of systolic distension

 
• PWS

  Peak wall stress

 
• PWV

  Pulse wave velocity

 
• TDI

  Tissue Doppler imaging

 
• TEE

  Transesophageal echocardiography

 
• TTE

  Transthoracic echocardiography

 
• VDM

  Vascular deformation mapping