European Journal of Cardio-Thoracic Surgery Advance Access published March 18, 2015
ORIGINAL ARTICLE
European Journal of Cardio-Thoracic Surgery (2015) 1–9 doi:10.1093/ejcts/ezv087
New insights in the association between bicuspid aortic valve phenotype, aortic configuration and valve haemodynamics Hans-Hinrich Sievers, Ulrich Stierle, Rebecca M.S. Hachmann and Efstratios I. Charitos* Department of Cardiac and Thoracic Vascular Surgery, University of Lübeck, Lübeck, Germany * Corresponding author. Department of Cardiac and Thoracic Vascular Surgery, University of Luebeck, Ratzeburger Allee 160, 23568 Luebeck, Germany. Tel: +49-451-5002108; fax: +49-451-5002051; e-mail: [email protected] (E.I. Charitos). Received 29 December 2014; received in revised form 4 February 2015; accepted 6 February 2015
Abstract OBJECTIVES: Patients with a bicuspid aortic valve (BAV) present with various phenotypes of the valve, the aortic root (AoR) and/or the ascending aorta (AAo) and various valve haemodynamics (vHs). The aim of the present study was to investigate the association between the above parameters. METHODS: Between February 1999 and April 2014, the preoperative aortographies of 828 surgical patients with BAV were evaluated. The exact BAV type was classified intraoperatively according to the number and spatial orientation of the raphes. RESULTS: On analysis of BAV phenotypes and aortic configurations, a weak pattern was revealed (P = 0.01) only for BAV type 0 and AoR dilatation. Including haemodynamics, certain significant patterns emerged: in insufficient BAVs, AoR dilatation was significantly more frequent in type 0, type 1 LR and type 1 RN, whereas AoR + AAo dilatation was more frequent in BAV type 1 LR. In stenotic BAVs, AAo dilatation alone was observed significantly more frequently in BAV type 1 LR and type 0. Combined vHs were associated with AAo and AoR + AAo dilatation in BAV type 2/unicuspid only. CONCLUSIONS: Associations between the two parameters, BAV type and aortic configuration, were only weak. With the inclusion of haemodynamics significant associations emerged but were not exclusive. In stenotic BAVs, aortic dilatation was more localized to AAo only, supporting the valve-related flow turbulence theory. Insufficient (even trace insufficient) BAVs and also combined lesions in BAV type 2/unicuspid were associated with a more extensive aortopathy (AoR + AAo) in younger patients, indicating a more aggressive gene-related aortopathy. Keywords: Bicuspid aortic valve • Ascending aorta • Aortic root • Haemodynamics • Aneurysm
INTRODUCTION The bicuspid aortic valve (BAV) is the most frequent congenital cardiac defect with an estimated prevalence of 1 to 2% in the general population, contributing considerably to the worldwide health burden. There is increasing evidence that BAV is not only a valvulopathy but a complex disease associated with different cardiovascular abnormalities [1] including the mitral valve [2], but more importantly an aortopathy of variable degree. This aortopathy bears an increased risk for life-threatening complications like rupture and dissection [3–5]. To reduce these possible complications, a better understanding of the relationship between valvulopathy and aortopathy is necessary. Patients with a BAV present with various phenotypes of the valve (BAV type), various configurations of the aorta ranging from normal configuration to dilatation of the aortic root (AoR) and/or the ascending aorta (AAo) as well as various degrees of valve haemodynamics
(vHs). A subject of intensive research is the association between BAV phenotype, aortic configuration and vH, and no definite consensus exists whether the dilated AoR and/or AAo are genetically determined or a result of the year-long, valve-related, haemodynamic stress. An answer to this question may have considerable therapeutic consequences, e.g. a more aggressive resection of the aortic dilatation during valve surgery or medical therapy if a genetic reason is suspected. The aim of the present study was 2-fold: First, we investigated whether the two variables (BAV phenotype and aortic configuration) were associated in BAV patients and, secondly, we aimed to investigate how these associations could be refined when vHs were included in the calculation in 828 patients referred for surgery on the aortic valve and/or AAo. This could probably provide some further information whether flow disturbance and/or genetics may be causative for aortic dilatation and have some implications for research on the aortopathy and its treatment during BAV surgery.
© The Author 2015. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ADULT CARDIAC
Cite this article as: Sievers H-H, Stierle U, Hachmann RMS, Charitos EI. New insights in the association between bicuspid aortic valve phenotype, aortic configuration and valve haemodynamics. Eur J Cardiothorac Surg 2015; doi:10.1093/ejcts/ezv087.
H.-H. Sievers et al. / European Journal of Cardio-Thoracic Surgery
2
MATERIALS AND METHODS Study population Between February 1999 and April 2014 a total of 1580 patients with intraoperatively classified BAVs and without known connective tissue disorders like transforming growth factor beta (TGF-β) receptor 1 or 2 mutation, Marfan, Turner, Loeys-Dietz or EhlersDanlos syndrome underwent cardiac surgery in our institution. In this investigation, we excluded patients with previous cardiac surgery (n = 74), aortic valve endocarditis (n = 32) and acute type A dissection (n = 17). From the remaining patients a total of 904 aortographies obtained during cardiac catheterization were available. In 33 of these patients, imaging of the AAo was not performed and in 43 patients the shape of the AAo could not be clearly determined. The final sample consisted of 828 patients. In all 828 patients, the BAV type was intraoperatively classified under direct vision. The primary indication for surgery was the valve lesion in 790 patients and AoR/AAo dilatation in 38 patients. These 38 patients had trace or mild insufficient BAVs. To determine the aortic configuration, all images were evaluated in a blinded fashion. Patient demographics and preoperative data are given in Table 1. All patients provided informed consent and the local ethics committee of the University of Lübeck approved the present study.
Classifications and haemodynamic definitions The terminology of the aortic valve was used as recently reported [6]. The classification of BAV has been published previously [7] and is presented in brief in Fig. 1. The term BAV denotes an aortic valve with the appearance of a ‘fishmouth’-like opening. To classify the complete spectrum of BAV comprehensively, the so-called ‘unicuspid’ valves were included in the BAV type 2/unicuspid group. The classification of the aortic dilatation configurations was slightly modified from the Fazel classification [8] (Fig. 2). The Fazel
classification is based on four clusters, cluster I: AoR dilatation, cluster II: AAo dilatation, cluster III: AAo dilatation tapering into the arch and cluster IV: AoR and AAo dilatation combined. The modification in this study entails the unification of Fazel cluster II and III. Thus, in the present study three aortic dilatation configurations were used: AoR dilatation, AAo dilatation and combined AoR + AAo dilatation. An aortic segment was determined as dilated when the diameter was obviously larger than normal (Fig. 2). The maximal and mean diameter of the ascending aorta and sinus for Fazel cluster I was 43 mm (36.3 ± 6.7 mm) and 54 mm (40.4 ± 6.5 mm), respectively.; for Fazel cluster II and III 70 mm (48.7 ± 6.1 mm) and 53 mm (38.4 ± 5.2 mm), respectively; and for Fazel cluster IV 80 mm (53.3 ± 7.7 mm) and 71 mm (44.8 ± 7.0 mm), respectively. The vHs were determined by echocardiography. For the purposes of this work vHs were categorized in stenosis due to leaflet calcification (no insufficiency grade>I, orifice area of 1 cm² or less or mean pressure gradient of >50 mmHg), insufficiency (no stenotic component, maximal valvular pressure gradient of 20 mmHg, no calcification of the valve intraoperatively) and combined lesion.
Statistical analyses Simple statistical tests (such as the t-test, χ 2 test, Mann–Whitney U-test, analysis of variance and Kruskal–Wallis tests) were employed where appropriate to identify differences in the demographics. Three-way contingency tables of the BAV type, aortic configuration and vHs were constructed and analysed using log-linear models [9]. To evaluate whether there is an association between the three variables, we hypothesized that the three variables are independent (null hypothesis) and investigated whether our data provide significant evidence to the contrary, evaluating whether there is mutual independence between the three variables (BAV type ? aortic configuration ? vH). To visualize the observed and expected frequencies in the threeway contingency tables as well as their deviations, mosaic plots were constructed [10]. The mosaic plots provide a quantitative visualization
Table 1: Demographic and preoperative clinical data of 828 patients
Age (years) Gender (male) BSA (m²) Hypertension LVEF (%) 50 Unknown NYHA I II III IV Unknown Diabetes COPD
Total (n = 828)
BAV type 0 (n = 62)
BAV type 1 (n = 649)
BAV type 2/unicuspid (n = 117)
P-values
56.8 ± 15.0 631 (76.2) 1.977 ± 0.200 (n = 797) 558 (67.4)
55.0 ± 14.5 47 (75.8) 1.941 ± 0.197 (n = 62) 40 (64.5)
59.1 ± 14.3 499 (76.9) 1.980 ± 0.197 (n = 620) 461 (71.0)
45.0 ± 13.6 85 (72.6) 1.979 ± 0.197 (n = 115) 57 (48.7)
ORIGINAL ARTICLE
European Journal of Cardio-Thoracic Surgery (2015) 1–9 doi:10.1093/ejcts/ezv087
New insights in the association between bicuspid aortic valve phenotype, aortic configuration and valve haemodynamics Hans-Hinrich Sievers, Ulrich Stierle, Rebecca M.S. Hachmann and Efstratios I. Charitos* Department of Cardiac and Thoracic Vascular Surgery, University of Lübeck, Lübeck, Germany * Corresponding author. Department of Cardiac and Thoracic Vascular Surgery, University of Luebeck, Ratzeburger Allee 160, 23568 Luebeck, Germany. Tel: +49-451-5002108; fax: +49-451-5002051; e-mail: [email protected] (E.I. Charitos). Received 29 December 2014; received in revised form 4 February 2015; accepted 6 February 2015
Abstract OBJECTIVES: Patients with a bicuspid aortic valve (BAV) present with various phenotypes of the valve, the aortic root (AoR) and/or the ascending aorta (AAo) and various valve haemodynamics (vHs). The aim of the present study was to investigate the association between the above parameters. METHODS: Between February 1999 and April 2014, the preoperative aortographies of 828 surgical patients with BAV were evaluated. The exact BAV type was classified intraoperatively according to the number and spatial orientation of the raphes. RESULTS: On analysis of BAV phenotypes and aortic configurations, a weak pattern was revealed (P = 0.01) only for BAV type 0 and AoR dilatation. Including haemodynamics, certain significant patterns emerged: in insufficient BAVs, AoR dilatation was significantly more frequent in type 0, type 1 LR and type 1 RN, whereas AoR + AAo dilatation was more frequent in BAV type 1 LR. In stenotic BAVs, AAo dilatation alone was observed significantly more frequently in BAV type 1 LR and type 0. Combined vHs were associated with AAo and AoR + AAo dilatation in BAV type 2/unicuspid only. CONCLUSIONS: Associations between the two parameters, BAV type and aortic configuration, were only weak. With the inclusion of haemodynamics significant associations emerged but were not exclusive. In stenotic BAVs, aortic dilatation was more localized to AAo only, supporting the valve-related flow turbulence theory. Insufficient (even trace insufficient) BAVs and also combined lesions in BAV type 2/unicuspid were associated with a more extensive aortopathy (AoR + AAo) in younger patients, indicating a more aggressive gene-related aortopathy. Keywords: Bicuspid aortic valve • Ascending aorta • Aortic root • Haemodynamics • Aneurysm
INTRODUCTION The bicuspid aortic valve (BAV) is the most frequent congenital cardiac defect with an estimated prevalence of 1 to 2% in the general population, contributing considerably to the worldwide health burden. There is increasing evidence that BAV is not only a valvulopathy but a complex disease associated with different cardiovascular abnormalities [1] including the mitral valve [2], but more importantly an aortopathy of variable degree. This aortopathy bears an increased risk for life-threatening complications like rupture and dissection [3–5]. To reduce these possible complications, a better understanding of the relationship between valvulopathy and aortopathy is necessary. Patients with a BAV present with various phenotypes of the valve (BAV type), various configurations of the aorta ranging from normal configuration to dilatation of the aortic root (AoR) and/or the ascending aorta (AAo) as well as various degrees of valve haemodynamics
(vHs). A subject of intensive research is the association between BAV phenotype, aortic configuration and vH, and no definite consensus exists whether the dilated AoR and/or AAo are genetically determined or a result of the year-long, valve-related, haemodynamic stress. An answer to this question may have considerable therapeutic consequences, e.g. a more aggressive resection of the aortic dilatation during valve surgery or medical therapy if a genetic reason is suspected. The aim of the present study was 2-fold: First, we investigated whether the two variables (BAV phenotype and aortic configuration) were associated in BAV patients and, secondly, we aimed to investigate how these associations could be refined when vHs were included in the calculation in 828 patients referred for surgery on the aortic valve and/or AAo. This could probably provide some further information whether flow disturbance and/or genetics may be causative for aortic dilatation and have some implications for research on the aortopathy and its treatment during BAV surgery.
© The Author 2015. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ADULT CARDIAC
Cite this article as: Sievers H-H, Stierle U, Hachmann RMS, Charitos EI. New insights in the association between bicuspid aortic valve phenotype, aortic configuration and valve haemodynamics. Eur J Cardiothorac Surg 2015; doi:10.1093/ejcts/ezv087.
H.-H. Sievers et al. / European Journal of Cardio-Thoracic Surgery
2
MATERIALS AND METHODS Study population Between February 1999 and April 2014 a total of 1580 patients with intraoperatively classified BAVs and without known connective tissue disorders like transforming growth factor beta (TGF-β) receptor 1 or 2 mutation, Marfan, Turner, Loeys-Dietz or EhlersDanlos syndrome underwent cardiac surgery in our institution. In this investigation, we excluded patients with previous cardiac surgery (n = 74), aortic valve endocarditis (n = 32) and acute type A dissection (n = 17). From the remaining patients a total of 904 aortographies obtained during cardiac catheterization were available. In 33 of these patients, imaging of the AAo was not performed and in 43 patients the shape of the AAo could not be clearly determined. The final sample consisted of 828 patients. In all 828 patients, the BAV type was intraoperatively classified under direct vision. The primary indication for surgery was the valve lesion in 790 patients and AoR/AAo dilatation in 38 patients. These 38 patients had trace or mild insufficient BAVs. To determine the aortic configuration, all images were evaluated in a blinded fashion. Patient demographics and preoperative data are given in Table 1. All patients provided informed consent and the local ethics committee of the University of Lübeck approved the present study.
Classifications and haemodynamic definitions The terminology of the aortic valve was used as recently reported [6]. The classification of BAV has been published previously [7] and is presented in brief in Fig. 1. The term BAV denotes an aortic valve with the appearance of a ‘fishmouth’-like opening. To classify the complete spectrum of BAV comprehensively, the so-called ‘unicuspid’ valves were included in the BAV type 2/unicuspid group. The classification of the aortic dilatation configurations was slightly modified from the Fazel classification [8] (Fig. 2). The Fazel
classification is based on four clusters, cluster I: AoR dilatation, cluster II: AAo dilatation, cluster III: AAo dilatation tapering into the arch and cluster IV: AoR and AAo dilatation combined. The modification in this study entails the unification of Fazel cluster II and III. Thus, in the present study three aortic dilatation configurations were used: AoR dilatation, AAo dilatation and combined AoR + AAo dilatation. An aortic segment was determined as dilated when the diameter was obviously larger than normal (Fig. 2). The maximal and mean diameter of the ascending aorta and sinus for Fazel cluster I was 43 mm (36.3 ± 6.7 mm) and 54 mm (40.4 ± 6.5 mm), respectively.; for Fazel cluster II and III 70 mm (48.7 ± 6.1 mm) and 53 mm (38.4 ± 5.2 mm), respectively; and for Fazel cluster IV 80 mm (53.3 ± 7.7 mm) and 71 mm (44.8 ± 7.0 mm), respectively. The vHs were determined by echocardiography. For the purposes of this work vHs were categorized in stenosis due to leaflet calcification (no insufficiency grade>I, orifice area of 1 cm² or less or mean pressure gradient of >50 mmHg), insufficiency (no stenotic component, maximal valvular pressure gradient of 20 mmHg, no calcification of the valve intraoperatively) and combined lesion.
Statistical analyses Simple statistical tests (such as the t-test, χ 2 test, Mann–Whitney U-test, analysis of variance and Kruskal–Wallis tests) were employed where appropriate to identify differences in the demographics. Three-way contingency tables of the BAV type, aortic configuration and vHs were constructed and analysed using log-linear models [9]. To evaluate whether there is an association between the three variables, we hypothesized that the three variables are independent (null hypothesis) and investigated whether our data provide significant evidence to the contrary, evaluating whether there is mutual independence between the three variables (BAV type ? aortic configuration ? vH). To visualize the observed and expected frequencies in the threeway contingency tables as well as their deviations, mosaic plots were constructed [10]. The mosaic plots provide a quantitative visualization
Table 1: Demographic and preoperative clinical data of 828 patients
Age (years) Gender (male) BSA (m²) Hypertension LVEF (%) 50 Unknown NYHA I II III IV Unknown Diabetes COPD
Total (n = 828)
BAV type 0 (n = 62)
BAV type 1 (n = 649)
BAV type 2/unicuspid (n = 117)
P-values
56.8 ± 15.0 631 (76.2) 1.977 ± 0.200 (n = 797) 558 (67.4)
55.0 ± 14.5 47 (75.8) 1.941 ± 0.197 (n = 62) 40 (64.5)
59.1 ± 14.3 499 (76.9) 1.980 ± 0.197 (n = 620) 461 (71.0)
45.0 ± 13.6 85 (72.6) 1.979 ± 0.197 (n = 115) 57 (48.7)
