Author + information
- Received January 3, 2018
- Revision received March 8, 2018
- Accepted March 20, 2018
- Published online June 3, 2019.
- Andrea Guala, PhDa,∗∗ (, )
- Jose Rodriguez-Palomares, MD, PhDa,∗,
- Lydia Dux-Santoy, MSca,
- Gisela Teixido-Tura, MD, PhDa,
- Giuliana Maldonado, MDa,
- Laura Galian, MDa,
- Marina Huguet, MDa,
- Filipa Valente, MDa,
- Laura Gutiérrez, MDa,
- Teresa González-Alujas, MD, PhDa,
- Kevin M. Johnson, PhDb,
- Oliver Wieben, PhDb,
- Augusto Sao Avilés, MD, PhDa,
- David Garcia-Dorado, MD, PhDa and
- Arturo Evangelista, MD, PhDa
- aHospital Universitari Vall d’Hebron, Department of Cardiology, CIBER-CV, Vall d’Hebron Institut de Recerca, Universitat Autònoma de Barcelona, Barcelona, Spain
- bDepartments of Medical Physics and Radiology, University of Wisconsin–Madison, Madison, Wisconsin
- ↵∗Address for correspondence:
Dr. Andrea Guala, Department of Cardiology, Hospital Universitari Vall d’Hebron, Paseo Vall d’Hebron 119-129, 08035, Barcelona, Spain.
Objectives This study sought to ascertain whether patients with a bicuspid aortic valve (BAV) have an intrinsic alteration in regional aortic stiffness compared with patients with tricuspid aortic valve (TAV) and Marfan syndrome (MFS) patients with similar aortic sizes, as well as to assess the influence of ascending aorta (AAo) dilation on regional stiffness parameters in BAV patients.
Background Imaging biomarkers as predictors of BAV, MFS, and degenerative AAo aneurysms in TAV patients (DA-TAV) are lacking. Biomechanical characterization has been proposed as a possible tool for further aneurysm stratification.
Methods A total 234 subjects (136 BAV, 44 MFS, and 18 DA-TAV patients and 36 healthy control subjects) were included. The cardiac magnetic resonance protocol comprised 4-dimensional flow to assess AAo and descending aorta (DAo) pulse wave velocities (PWVs) and double-oblique, 2-dimensional, steady-state free-precession cine cardiac magnetic resonance to compute aortic distensibility (AD).
Results On adjusted analysis, nondilated BAV patients had similar PWV and AD as healthy control subjects in both AAo and DAo, whereas dilated BAV did not differ from DA-TAV. In contrast, AAo and DAo stiffness in MFS patients was markedly greater than in BAV patients, increasing slightly with dilation severity. AAo PWV showed a biphasic pattern in BAV patients: it first decreased and then increased throughout AAo dilation, with a clear turning point at 50 mm, whereas distensibility did not discern mildly dilated aorta. In multivariate analysis adjusted for clinical and demographic characteristics, only PWV was related to AAo dilation in BAV patients.
Conclusions The mechanical properties of AAo aneurysms are similar in BAV and TAV patients, whereas MFS patients have a stiffer aorta. Aortic stiffness strongly depends on dilation severity. AAo PWV resulted in a potentially clinically useful biphasic trend with respect to aneurysm diameter, whereas distensibility did not discern mildly dilated aorta. Beyond clinical risk factors, PWV but not AD was independently related to AAo dilation in BAV patients.
- ascending aorta aneurysm
- bicuspid aortic valve
- 4D flow CMR
- Marfan syndrome
- pulse wave velocity
Bicuspid aortic valve (BAV) is a congenital disorder frequently associated with ascending aorta (AAo) dilation, often called bicuspid aortopathy, and the concomitant increased risk of aortic aneurysm and dissection (1). Current clinical management of bicuspid aortopathy is based on the use of a threshold value for AAo diameter as a parameter for the indication of prophylactic aortic resection (2).
Wall shear stress abnormalities secondary to abnormal flow in the AAo have been related to aortic extracellular matrix dysregulation and elastic fiber degeneration (3) and to aortic dilation (4) in BAV patients; however, controversy remains about the eventual presence of intrinsic aorta wall alteration (1,5) in BAV patients. The fact that aortic dilation severity appears out of proportion to valve dysfunction (6), the presence of aortic aneurysm within families with BAV (5), and the histological findings of cystic medial degeneration and fibrillin-1 deficiency in the AAo wall similar to that observed in Marfan aortas (7) supported the existence of intrinsic abnormalities of the aorta wall.
In this context, biomechanical characterization has been proposed as a possible parameter for stratification of BAV aortopathy (8–11), especially if performed regionally (12,13). Aortic distensibility (AD) is the most widely used biomarker (4,8,14), whereas pulse wave velocity (PWV) is considered the gold standard for measuring arterial stiffness (10,15,16).
However, the choice of the most suitable mechanical biomarker is still under debate. First, both PWV and AD can depend on the aortic diameter (8). This condition, which has been most often neglected, can lead to misinterpretation of the results. Moreover, most studies investigating aortic mechanical properties in BAV patients were performed by echocardiography and not by cardiac magnetic resonance (CMR), which is considered the gold standard, and significant valvular dysfunction was not excluded. In addition, the evaluation of biomechanical differences among BAV, Marfan, and degenerative aneurysms is lacking.
The aims of the present study were 2-fold: 1) to ascertain whether BAV patients have intrinsic alterations in regional aortic stiffness compared with patients with tricuspid aortic valve (TAV) and Marfan syndrome (MFS) with similar aortic size and clinical conditions; and 2) to assess the influence of aortic dilation in regional aortic stiffness evaluation by PWV and AD in BAV patients.
We prospectively enrolled 198 consecutive patients (136 BAV, 44 MFS, and 18 TAV with degenerative aneurysm [DA-TAV]) and 36 healthy volunteers with TAV. Inclusion criteria were age >18 years, no significant aortic valvular disease by echocardiography (aortic regurgitation ≤II/IV, aortic velocity <3 m/s) or other cardiovascular disorders, and no history of surgical repair or aortic valve replacements. Subjects with DA-TAV were selected, excluding those who had a BAV or any genetic syndrome. All MFS patients had a TAV and were confirmed by genetic testing. All patients were Caucasian. The study was approved by the local ethics committee, and informed consent was obtained from all participants.
CMR studies were performed on a GE 1.5-T Signa scanner (GE Healthcare, Waukesha, Wisconsin). The protocol included several balanced steady-state free-precession cine images, which were used to assess BAV morphology and aortic areas, and a 4-dimensional (4D) phase-contrast CMR (4D flow CMR) acquisition with retrospective electrocardiographic gating during free breathing. A radially undersampled acquisition (PC VIPR [phase-contrast vastly undersampled isotropic projection]) with 5-point balanced velocity encoding (17) was used for 4D flow imaging of the entire thoracic aorta in approximately 10 min of total scan time. Data were acquired using the following parameters: velocity encoding 200 cm/s, field of view 400 × 400 × 400 mm, acquisition matrix 160 × 160 × 160, voxel size 2.5 × 2.5 × 2.5 mm, flip angle 8°, repetition time 4.2 to 6.4 ms, and echo time 1.9 to 3.7 ms. This dataset was reconstructed according to the nominal temporal resolution of each patient and was (5 × TR) 21 ms to 32 ms with no control for respiratory motion. Reconstructions were performed offline with corrections for background phase from concomitant gradients, eddy currents, and trajectory errors of the 3-dimensional radial acquired k-space (17). Intravenous contrast was not given to minimize patient risk.
Brachial systolic and diastolic pressures were taken immediately after the CMR study.
Biomechanical parameters calculation
AD was computed from double-oblique cine imaging in the AAo and descending aorta (DAo) at the pulmonary artery level. Time-resolved aortic contours were traced semiautomatically using ARTFUN software (INSERM U678, Paris, France) with minimal user interaction. Maximum (Amax) and minimum (Amin) areas were extracted, and AD was computed as: [(Amax − Amin) / Amin] / (systolic − diastolic blood pressure) (Figure 1). PWV was computed from 4D flow CMR data in the AAo and DAo, defined as the regions limited by the sinotubular junction, the third supra-aortic vessel, and the level of the diaphragm. Transit time was computed with a wavelet-based method, which is the most robust technique (18), using a total of 100 analysis planes perpendicular to the aortic centerline (Supplemental Appendix). Moreover, to obtain a diameter-independent biomechanical parameter, the Moens-Korteweg equation (12,15) was used to compute the product of arterial wall thickness and incremental Young modulus (Eh) as , where ρ is blood density and D is maximum vessel diameter. This derivation is valid under the normally accepted assumptions of inviscid and incompressible blood and a thin-walled, straight, uniform aorta.
Definition of BAV phenotype and aortic dilation
BAV phenotypes were classified in 2 groups depending on whether there was fusion of the right and left cusps or fusion of the right and noncoronary cusps. Aortic dilation was defined when z-score was >2 considering AAo and root maximum diameters, age, sex, and body surface area (19,20).
The Kolmogorov-Smirnov test was used to evaluate distribution normality. Differences between groups for continuous parameters were assessed by Student’s t-test or ANCOVA if normally distributed and Mann-Whitney U test otherwise. Continuous demographic variables were expressed as mean ± SD if they had a normal distribution and as median [first–third] nonadjusted quartiles otherwise.
Multivariate linear regression analyses with a forward-selection procedure were used to identify independent determinants of aortic dilation. Variables were entered in the model if p <0.10 in univariate analyses. A 2-tailed p value <0.05 was considered statistically significant. SPSS version 19.0 software (IBM SPSS Statistics, Chicago, Illinois) was used for the analysis.
Demographic characteristics and aortic diameters among groups (BAV, MFS, DA-TAV, and healthy control subjects) are shown in Table 1. In the overall series, BAV patients had similar systolic blood pressure (136 ± 17 mm Hg vs. 132 ± 18 mm Hg; p = 0.345) but higher diastolic blood pressure (77 ± 9 mm Hg vs. 68 ± 12 mm Hg; p < 0.001) than TAV patients and no differences with respect to MFS patients (128 ± 17 mm Hg and 73 ± 12 mm Hg; p = 0.068 and p = 0.288, respectively). MFS subjects were younger (38 ± 13 years) than those with TAV (47 ± 16 years; p = 0.014) and BAV (51 ± 14 years; p < 0.001). AAo diameters were larger in BAV (45.1 ± 8.1 mm) than in either TAV (37.1 ± 11.0 mm; p < 0.001) or MFS (35.5 ± 5.9 mm; p < 0.001). Beta-blockers were used by 16% of BAV, 37% of DA-TAV, and 47% of MFS patients.
Biomechanical analyses of the 3 groups without aortic dilation are shown in Table 2. Nondilated BAV patients had similar AAo PWV and AD as the healthy control group after adjustment for age, diastolic blood pressure, and AAo diameter (Table 2). In contrast, AAo PWV and AD revealed stiffer aortas in MFS patients than in healthy control subjects and nondilated BAV patients after adjustment for age and AAo diameter. Thus, no intrinsic biomechanical impairment was found in the AAo of nondilated BAV patients, whereas strong differences were found when BAV and MFS patients were compared. Moreover, no differences with respect to BAV valve morphotype were found (Supplemental Table 1).
Dilated BAV and DA-TAV patients had similar AAo biomechanical properties after adjustment for age, diastolic blood pressure, and AAo diameter (Table 3). In contrast, the adjusted comparison showed greater AAo stiffness in dilated MFS patients than in dilated BAV patients. Thus, dilated BAV and TAV patients also had comparable AAo mechanical properties, but dilated MFS patients had a stiffer aorta.
AAo mechanical properties and aortic dilation in BAV
To assess the influence of AAo dilation on biomechanical parameters, we divided the BAV population into quintiles for maximum AAo diameter; cutoff values were 38.5, 43, 46.8, and 53.2 mm (Table 4). Statistical analysis was performed to compare each group with the immediate lower one (second vs. first group, third vs. second, and so on).
Although AAo PWV decreased from the first to the fourth quintile and increased from the fourth to the fifth (Figure 2), AAo AD was similar in the first 3 dilation quintiles and decreased from the third quintile onward. PWV trend highlighted a clear shift of biomechanical behavior at the fourth dilation degree, which corresponded to a mean diameter of 49.6 mm, with a marked increase in stiffness as the aorta was further enlarged. Figure 2 also includes the corresponding scatterplots.
For any given material property, PWV is inversely related to total volume. To ascertain whether the decrease in AAo PWV was only due to an increase in aortic size or whether it was due to intrinsic differences in the wall properties, the trend of Eh with diameter quintiles was analyzed. The obtained trend was similar to the one encountered for PWV, with a decreasing part covering the first 4 dilation groups, followed by marked growth in the last diameter quintile (Figure 3, Supplemental Table 5).
Biomechanical parameters associated with AAO dilation
To investigate the main associations of AAo dilation in BAV patients, univariate and multivariate binary logistic analyses were performed separately to test PWV and AD (model 1 and model 2, respectively). Because diastolic blood pressure was related to AAo dilation in univariate analysis (Table 5), it was included in multivariate analyses. In multivariate analysis, PWV was the only biomechanical parameter independently related to dilation in BAV patients (p < 0.001).
Biomechanical parameters in DAo
Regional evaluation of PWV and AD was also conducted in the DAo. Subjects with BAV and TAV (healthy control subjects and those with DA-TAV) showed similar stiffness in the DAo, both with and without AAo dilation. In contrast, the DAo in MFS patients was stiffer than in the BAV and TAV groups (Supplemental Tables 2 to 4).
This study assessed regional PWV and distensibility by 4D flow and 2-dimensional (2D) cine CMR in a large cohort of BAV, DA-TAV, and MFS patients. The results showed that nondilated BAV patients and healthy control subjects had similar AAo and DAo stiffness (in terms of both PWV and AD), after adjustment for age, body surface area, diastolic blood pressure, and local diameter. Both groups strongly differed from MFS patients with normal aortic diameters. Similar differences were found among BAV, DA-TAV, and MFS groups with aortic dilation.
Analysis of the biomechanical properties of the AAo has been limited for technical reasons. To date, only 1 previous 2D phase-contrast CMR-based study measuring PWV exclusively in the AAo has been reported (13). Indeed, the studies that analyzed stiffness in this region more often also included the aortic arch and proximal DAo. A study with a large number of subjects found no significant changes in arch PWV (4), although another study with fewer cases (21) found arch PWV to be elevated in BAV. Considering our results, averaging the PWV over the whole thoracic aorta in patients with AAo dilation might cancel out these opposing regional differences. Our results are further in line with 2 studies showing MFS patients to have higher aortic PWV than TAV patients (without MFS) (13,22). Owing to its feasibility, different studies measured carotid-femoral PWV, which correlated with DAo PWV (23); however, this did not include information on the AAo (12). Confirming our results in the DAo, carotid-femoral PWV was similar in nondilated BAV and healthy control subjects (9) and higher in dilated BAV (9,24) and MFS patients (25,26). Regional analysis of the PWV could be particularly important in BAV patients given the localized nature of aortic dilation. To date, PWV computation from 4D flow CMR has been limited to feasibility and validation studies (27,28); however, the technology has several major advantages over 2D phase-contrast CMR. First, it is based on the same acquisition, thereby limiting errors due to heart rate variability. Moreover, the higher number of waveforms (compared with 2 for 2D phase-contrast CMR) permits spatially continuous evaluation (12). Nonetheless, 2D images can be acquired with a higher temporal resolution.
Compared with TAV subjects, the group of BAV patients had reduced AD in AAo and DAo, as reported previously (8,29–31). However, these discrepancies were due to differences in diameter and age between groups and disappeared when adjusted for these variables. Previous echocardiographic studies reported that nondilated BAV patients have higher distensibility than those with dilated BAV (8,29) and that biomechanical differences between BAV and TAV subjects were explained by the diameter (8). Furthermore, our study confirmed that AD is reduced in MFS patients (14,25).
We found no difference in aortic stiffness among valvular morphotypes (Supplemental Table 1), as found by histological (32), image-based (30), and tonometric-based (4,24) investigations. Moreover, in our series, AAo and DAo biomechanics were similar in BAV patients of both sexes (Supplemental Table 6), as reported elsewhere (33). We further found no difference related to beta-blocker administration in terms of PWV and AD in any of the patient groups (BAV, DA-TAV, and MFS).
Biomechanical parameters associated with AAo dilation
PWV showed the best association with AAo dilation in BAV patients on multivariate analysis, beyond the clinical risk factors. This might be a consequence of the several advantages of PWV over AD. The main advantages of PWV are that, unlike AD, it is a physical quantity, does not rely on any geometric or mechanical assumptions (15), and is not based on local pressure, the unavailability of which as a measurement is an unsurmountable methodological limitation (12,15). Of note, AD results were more sensitive to confounding factors such as age, aortic diameter, and diastolic blood pressure.
Aortic biomechanical properties at different aortic dilation severities
In BAV patients, AAo AD presented steady values at mild degrees of dilation, while showing a progressive stiffening at greater diameters (Figure 2). By contrast, AAo PWV resulted in a biphasic pattern, with a gradual reduction in stiffness from nondilated to mildly dilated aortas, followed by sharp stiffening at severe dilation (from 50 mm). To analyze aortic stiffness independent of local diameter, the elastance–wall thickness product (Eh) was calculated. Showing a biphasic trend, this parameter confirmed the results obtained from PWV (Figure 3).
A biphasic trend similar to our findings was shown in a previous invasive study (29), which found the elastic modulus at mild aortic dilation to be lower than in healthy control subjects while strongly rising at higher diameters. Although this analysis was performed in TAV patients, in view of our results it should also be valid for BAV patients.
We speculate that several factors participate in the creation of a change in AAo stiffness at increasing diameter in BAV patients. At lower diameters, the cumulative damage of the matrix results in impairment of the mechanical resistance of the wall. Despite histological analysis recognizing cystic medial necrosis, smooth muscle cell apoptosis, and elastin fragmentation as possible sources of aorta wall weakening (32), this is the first study to report it. However, at greater diameters, the considerable increase in tension due to dilation (Laplace law) implies an increase in the role of collagen filaments (34). Because collagen is much stiffer than elastin, it results in an exponential increase in aortic stiffness.
This trend could have strong implications for the clinical use of noninvasive biomarkers of aortic stiffness. First, the concomitant dependence of AD on diameter (8) and the reduction in wall stiffness at mild dilation (Eh) (Figure 3) might preclude the use of this parameter for monitoring of mildly dilated patients. In contrast, the biphasic trend of AAo PWV could be favorable for clinical monitoring, although longitudinal studies are needed to test whether this trend might be exploited as an early marker of aneurysm progression.
The importance of our findings lies in differentiating the aortic mechanical properties of BAV and MFS patients and in demonstrating that no intrinsic mechanical impairment is found in the aorta of BAV patients. Thus, it seems that BAV patients without additional specific risk factors do not benefit from a different clinical management than the DA-TAV population, as suggested by current guidelines on management of aortic diseases.
Reflecting the prospective, consecutive nature of our study, DA-TAV patients were older than subjects with nondilated TAV and BAV, whereas MFS patients were younger than BAV patients. These characteristics reflect the heterogeneous etiology of dilation in these groups. To overcome this age dispersion, we performed adjusted analysis, stratifying by age. Brachial pressure was used in the computation of AD. Although commonly used, this information can be a leading source of error. Although the hallmark of aortopathy in MFS patients is aortic root dilation, distensibility and PWV were measured in the tubular AAo. This allowed direct comparison with both BAV and TAV dilated subjects and was supported by prior studies that, like ours, found impaired mechanical properties in the tubular aorta in MFS patients even in the absence of aorta dilation (13,14). Moreover, distensibility was not corrected for through-plane motion. The computation of PWV can be impacted by the relatively low time resolution of 4D flow acquisition, especially for high PWV values. However, the algorithm for transit time acquisition implemented has been shown to provide stable results at time resolutions substantially lower than that used in the present study (18). Because it is very difficult to reliably measure aorta wall thickness, we could not further differentiate the role of average material properties from its thickness. However, our conclusions cannot be reversed by the very limited difference in thickness between BAV and TAV (35) or with different grades of dilation (29). We further acknowledge the absence of correlation with events and aneurysm progression over time. Moreover, our results might only be valid for Caucasian patients.
BAV and TAV patients have similar regional aorta mechanical properties, both with and without aortic dilation. By contrast, dilated and nondilated MFS patients have much greater aortic stiffness than corresponding BAV patients. In BAV, AAo PWV was first decreased and then increased throughout AAo dilation, with a turning point at a diameter of around 50 mm.
COMPETENCY IN MEDICAL KNOWLEDGE: BAV and TAV patients have similar stiffness in the ascending and descending aorta, whereas MFS patients have a stiffer aorta. Aortic dilation exerts similar effects on the local mechanical properties of BAV and TAV patients, with MFS patients still being different. BAV cusp fusion pattern does not imply different aortic stiffness. An increase in aorta size correlates with a decrease in local PWV at low to mild dilation and with an increase at greater dilation. Distensibility and PWV should always be considered in context with aorta size.
TRANSLATIONAL OUTLOOK: Given that the advantages of 4D flow CMR permit regional PWV quantification, longitudinal studies are needed to define the predictive value of this parameter beyond maximum diameter as a marker of aorta complications.
The authors thank Christopher François (University of Wisconsin-Madison) and Rob van der Geest (Leiden University Medical Center) for advice on 4D flow sequence parameters and Roberto García Álvarez (GE Healthcare) for technical support. The authors are also grateful to Christine O’Hara for English revisions.
↵∗ Drs. Guala and Rodriguez-Palomares contributed equally to this work and are joint first authors.
This study was funded by Instituto de Salud Carlos III through the projects PI11/01081 and PI14/0106 (co-funded by ERDF/ESF), La Marató de TV3 (project number 20151330), and Ministerio de Economía y Competitividad through Retos-Colaboración 2016 (RTC-2016-5152-1) and Beca Philips de la Societat Catalana de Cardiologia 2017. Dr. Guala has received funding from the European Union Seventh Framework Programme FP7/People under grant agreement number 267128. Dr. Johnson has received a research grant from GE Healthcare through The University of Wisconsin-Madison (no personal compensation); has received research grant support from Myocardial Solutions; and holds a consulting agreement with Vertex Pharmaceuticals. Dr. Wieben has received a research grant from GE Healthcare through The University of Wisconsin-Madison (no personal compensation). All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- 4-dimensional (time-resolved 3-dimensional imaging)
- ascending aorta
- aortic distensibility
- bicuspid aortic valve
- cardiac magnetic resonance
- descending aorta
- degenerative aneurysms with tricuspid aortic valve
- elastance–wall thickness product (product of wall thickness and incremental Young modulus)
- Marfan syndrome
- pulse wave velocity
- tricuspid aortic valve
- Received January 3, 2018.
- Revision received March 8, 2018.
- Accepted March 20, 2018.
- 2019 American College of Cardiology Foundation
- Hiratzka L.F.,
- Creager M.A.,
- Isselbacher E.M.,
- et al.
- Guzzardi D.G.,
- Barker A.J.,
- Van Ooij P.,
- et al.
- Bissell M.M.,
- Hess A.T.,
- Biasiolli L.,
- et al.
- Biner S.,
- Rafique A.M.,
- Ray I.,
- Cuk O.,
- Siegel R.J.,
- Tolstrup K.
- Hahn R.T.,
- Roman M.J.,
- Mogtader A.H.,
- Devereux R.B.
- Tzemos N.,
- Lyseggen E.,
- Silversides C.,
- et al.
- Erbel R.,
- Aboyans V.,
- Boileau C.,
- et al.
- Coutinho T.,
- Turner S.T.,
- Kullo I.J.
- Whitlock M.C.,
- Hundley W.G.
- Cavalcante J.L.,
- Lima J.A.C.,
- Redheuil A.,
- Al-Mallah M.H.
- Johnson K.M.,
- Lum D.P.,
- Turski P.A.,
- Block W.F.,
- Mistretta C.A.,
- Wieben O.
- Bargiotas I.,
- Mousseaux E.,
- Yu W.,
- et al.
- Grotenhuis H.B.,
- Ottenkamp J.,
- Westenberg J.J.M.,
- Bax J.J.,
- Kroft L.J.M.,
- de Roos A.
- Rogers W.J.,
- Hu Y.L.,
- Coast D.,
- et al.
- Hirata K.,
- Triposkiadis F.,
- Sparks E.,
- Bowen J.,
- Wooley C.F.,
- Boudoulas H.
- Salvi P.,
- Grillo A.,
- Marelli S.,
- et al.
- Wentland A.L.,
- Wieben O.,
- François C.J.,
- et al.
- Longobardo L.,
- Carerj M.L.,
- Pizzino G.,
- et al.
- Hickson S.S.,
- Butlin M.,
- Graves M.J.,
- et al.