- Email: [email protected]
Morphologic and Functional Markers of Aortopathy in Patients With Bicuspid Aortic Valve Insufficiency Versus Stenosis
Morphologic and Functional Markers of Aortopathy in Patients With Bicuspid Aortic Valve Insufficiency Versus Stenosis Evaldas Girdauskas, MD, PhD, Mina Rouman, MD, Kushtrim Disha, MD, Beatrix Fey, MD, Georg Dubslaff, MD, Yskert von Kodolitsch, MD, PhD, Hermann Reichenspurner, MD, PhD, Michael A. Borger, MD, PhD, and Thomas Kuntze, MD Department of Cardiovascular Surgery, University Heart Center, Hamburg; Departments of Cardiac Surgery and Radiology, Central Hospital Bad Berka, Bad Berka; Department of Cardiology, University Heart Center, Hamburg, Germany; and Columbia University Medical Center, New York, New York
Background. Bicuspid aortic valve (BAV)–associated aortopathy is heterogeneous and still insufficiently defined. We prospectively analyzed the morphologic and functional variables of aortopathy in patients undergoing operations for BAV insufficiency (BAV-AI) vs stenosis (BAV-AS). Methods. A total of 172 consecutive patients (71% male, 59 ± 10 years) underwent aortic valve replacement with or without proximal aortic operation for BAV-AS (n [ 137), and BAV-AI (n [ 35) from January 2012 through December 2014. All patients underwent preoperative cardiac magnetic resonance imaging to evaluate morphologic and functional variables of the aortic root. Magnetic resonance imaging data were used to guide sampling of aortic tissue intraoperatively (ie, from the area where flow jet impacts on the aortic wall [jet sample] and the opposite aortic wall [control sample]). Aortic wall lesions were graded based on the histologic sum score (range, 0 to 21). Expression and severity of aortopathy were quantified by means of
proximal aortic phenotype, indexed aortic diameters, and a sum score. Results. Cross-sectional aortic diameters were significantly larger in the BAV-AI group vs the BAV-AS group (47 ± 8 mm vs 41 ± 8 mm, p [ 0.001). Moreover, root dilatation phenotype was more frequent in the BAV-AI group (27% vs 6%, p [ 0.01) and was associated with a significantly larger aortic annulus diameter (32 ± 3 mm vs 27 ± 3 mm, p < 0.001). The histologic sum score was significantly different between the study groups (3.7 ± 2.6 BAV-AI vs 2.5 ± 1.4 BAV-AS, p [ 0.03). Logistic regression revealed a significant association between BAV-AI and indexed aortic diameter exceeding 22 mm/m2 (odds ratio, 4.7; p [ 0.007). Conclusions. Our study demonstrates that BAV functional phenotype correlates significantly with the expression and severity of bicuspid aortopathy.
B
etiologic entities [10, 11]. Distinct microstructural aortic wall lesions have been specifically demonstrated in patients with BAV-AI insufficiency vs BAV-AS [12, 13]. Moreover, recent meta-analysis demonstrated a 10-fold higher risk of aortic dissection in patients who undergo isolated aortic valve replacement (AVR) for BAV-AI compared with BAV-AS [14]. However, these data are limited by retrospective study design. Therefore, in our current study, we aimed to prospectively analyze the morphologic and functional parameters of aortopathy in patients undergoing operations for BAV-AI vs BAV-AS.
icuspid aortic valve (BAV)–associated aortopathy is heterogeneous and still insufficiently defined [1, 2]. Sophisticated imaging series [3], community-based echocardiographic follow-up data [4], and extensive surgical cohorts [5] have been pooled to evaluate a number of comprehensive classification systems based on BAV morphology [6] and the shape of the proximal aorta [7], or a combination of both factors [8, 9]. Several recent retrospective studies indicated that patients with BAV insufficiency (BAV-AI) vs BAV stenosis (BAV-AS) represent distinct clinical and pathologic phenotypes and therefore should be interpreted as distinct Accepted for publication May 20, 2016. Presented at the Fifty-second Annual Meeting of the Society of Thoracic Surgeons, Phoenix, AZ, Jan 23–27, 2016. Address correspondence to Dr Girdauskas, Department of Cardiovascular Surgery, University Heart Center Hamburg, Martinistraße 52, 20246 Hamburg, Germany; email: [email protected].
Ó 2016 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2016;-:-–-) Ó 2016 by The Society of Thoracic Surgeons
Patients and Methods The prospective institutional BAV database (Central Hospital, Bad Berka, Germany) included all consecutive patients with BAV disease who were referred for AVR operations, with or without simultaneous replacement of 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2016.05.085
2
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
the proximal aorta, from January 2012 through December 2014. Thuringian Chamber of Physicians Ethics Committee approved the study, and all patients gave written informed consent. Elective AVR operations were performed in 228 consecutive patients with BAV disease, with or without simultaneous proximal aortic replacement, at our institution. We excluded patients with Marfan syndrome and those undergoing combined procedures (ie, other than simultaneous aortic operations; n ¼ 31) or urgent or emergency operations (n ¼ 15). Nine BAV patients had contraindications for preoperative magnetic resonance imaging (MRI) examination and were excluded. The study cohort comprised 172 BAV patients (71% male), aged 59 10 years, who met these inclusion criteria. Predominant/isolated BAV-AS was identified in 137 patients (80%; BAV-AS group), and the remaining 35 BAV patients (20%) had pure aortic valve insufficiency (BAV-AI group). The primary end point of our study was the expression and severity of aortopathy in these two groups.
Definitions and Measurements Preoperative echocardiography and cardiac MRI were used to assess morphology and function of the aortic valve in all patients. Published guidelines were used to define aortic valve stenosis and insufficiency [15, 16]. Patients with mixed BAV disease were classified according to the predominant functional valve lesion; for example, BAV patients with severe stenosis and mild-tomoderate aortic valve insufficiency were classified as having BAV-AS. The final decision regarding the bicuspidality was made based on the intraoperative aortic valve description by the surgeon. Although transthoracic echocardiography failed to correctly identify aortic valve morphology in 54 patients (31%), cardiac MRI demonstrated 98% sensitivity and specificity in identifying BAV disease. The diameters of proximal aorta (ie, aortic annulus, sinuses of Valsalva, sinotubular junction, ascending aorta, aortic arch) were measured preoperatively by
Ann Thorac Surg 2016;-:-–-
cardiac MRI. All aortic diameters were measured as the largest observed cross-sectional diameter perpendicular to the aortic axis in a midvessel slice at end diastole using the inner edge–to–inner edge technique, in accordance with previously published recommendations for MRI measurements [17]. Similar to others [7–9], we determined aortic phenotype based on cardiac MRI data (Fig 1). Normal aorta phenotype was characterized by all aortic diameters less than 22 mm/m2 of body surface area and nonindexed aortic diameters of less than 40 mm (Fig 1A). Predominant aortic root dilatation was defined as maximal aortic dilatation exceeding 22 mm/m2 at the level of the sinuses of Valsalva or 40 mm in maximal diameter (Fig 1B). The middle ascending aorta phenotype was determined by maximal aortic diameters at the level of the middle ascending tubular aorta and exceeding 22 mm/m2 or 40 mm in maximal diameter (Fig 1C). Distal ascending/ aortic arch phenotype was diagnosed when maximal dimensions were measured at the level of distal ascending aorta or proximal aortic arch, or both, and exceeded 22 mm/m2 or 40 mm (Fig 1D).
Preoperative MRI Examination The following morphologic and functional variables were assessed: (1) cusp fusion pattern, (2) proximal aortic diameters, (3) proximal aortic phenotype, (4) angulation between the left ventricular (LV) outflow axis and the aortic root (angle LV/aorta). The protocol of cardiac MRI examination and the corresponding measurements have been described in detail in our previous publication [18]. Cusp fusion pattern, noted as right/left (R/L) vs right/ noncoronary (R/N) fusion of the BAV valve was determined from the preoperative MRI analysis and confirmed by intraoperative inspection. Cusp fusion pattern was determined by the orientation of the raphe and the shape of systolic BAV opening, by placing an steady-state free precession–function sequence parallel and just distal to the aortic valve annulus (Fig 2).
Fig 1. Four distinct proximal aortic phenotypes as identified by cardiac magnetic resonance imaging. (A) Normal aorta phenotype: all aortic diameters were less than 22 mm/m2 of body surface area and nonindexed aortic diameters were less than 40 mm. (B) Predominant aortic root dilatation: maximal aortic dilatation at the level of the sinuses of Valsalva exceeding 22 mm/m2 or 40 mm in maximal diameter. (C) Mid ascending aorta phenotype: maximal aortic diameters at the level of the mid ascending tubular aorta and exceeding 22 mm/m2 or 40 mm in maximal diameter. (D) Distal ascending/aortic arch phenotype: maximal dimensions at the level of distal ascending aorta or proximal aortic arch exceeded 22 mm/m2 or 40 mm.
Ann Thorac Surg 2016;-:-–-
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
3
Fig 2. Two different cusp fusion patterns in patients with bicuspid aortic valve disease: (A) right/noncoronary cusp fusion and (B) right/ left cusp fusion.
The angle LV/aorta was determined at peak systole in the LV inflow-outflow view (Fig 3). LV outflow axis was modeled as a vector between the LV apex and the midpoint of the LV outflow tract (15 mm below aortic valve plane). Two study radiologists (B.F. and G.D.), blinded to the intraoperative findings, performed the MRI analyses.
p ¼ 0.002). Moreover, significantly larger aortic valve prostheses were used in the BAV-AI group (25.7 2.0 mm vs 23.8 1.9, p ¼ 0.01). One patient in the BAV-AS group (0.6%) died inhospital of multiorgan failure induced by low cardiac output syndrome. A perioperative stroke occurred in 2 patients (1.2%). The in-hospital course in all remaining patients was uneventful.
Study Population Demographics and intraoperative variables are summarized in Table 1. Both study subgroups were quite different in baseline characteristics. In particular, the patients in BAV-AI group were significantly younger and more frequently male. The BAV-AS group had a tendency toward higher prevalence of comorbidities, including arterial hypertension and diabetes. The 137 BAV-AS patients underwent an elective AVR operation that was accompanied by a simultaneous supracoronary ascending aortic replacement in 22 patients (16%). Aortic root replacement by means of composite graft implantation was performed in only 3 BAV-AS patients (2%). Aortic valve repair, with or without simultaneous aortic root replacement, was successful in 30 BAV-AI patients (86%), and the remaining 5 patients (14%) underwent an AVR operation. Proximal aortic operations were significantly more prevalent in the BAV-AI than in the BAV-AS group (71% vs 18%,
Histologic Examination MRI data were used to obtain 2 aortic specimens, a “jet sample” and a “control sample” intraoperatively. The jet sample was obtained from the area of contact between the systolic transvalvular flow jet and the aortic wall, as determined by preoperative MRI analysis (ie, segment of aortic circumference that was in direct contact with the flow jet). The control sample was collected from the opposite aortic wall. Intraoperative collection of aortic tissue has been described in detail previously [18]. Histologic examination of the aortic tissue and assessment of aortic wall lesions by means of histologic grading scale [19] has been described previously [18]. Briefly, seven histologic criteria were used to grade aortic wall lesions: fibrosis, atherosclerosis, medionecrosis, cystic medial necrosis, changes in smooth muscle cell orientation, elastic fragmentation, and periaortic inflammation. Each variable was graded from 0 (no change) to 3 (most Fig 3. Measurement of angle left ventricle (LV)/aorta on cardiac magnetic resonance imaging. (A) Large angle between the LV outflow axis and the aortic root (i.e. angle LV/Aorta 68 ). (B) Small angle between the LV outflow axis and the aortic root (i.e. angle LV/Aorta 35 ).
4
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
Ann Thorac Surg 2016;-:-–-
Table 1. Demographics and Intraoperative Variables in Both Subgroups Variablea Age, y Male sex Body surface area, m2 NYHA Functional Class Diabetes History of smoking Arterial hypertension COLD Valve size, mm Aortic root replacement Ascending aortic replacement Hemiarch replacement
BAV-AS (n ¼ 137)
BAV-AI (n ¼ 35)
p Value
61 9 93 (68) 1.98 0.2 2.4 0.6 20 (15) 32 (23) 89 (65) 8 (6) 23.8 1.9 3 (2) 22 (16)
49 11 30 (86) 1.92 0.3 2.6 0.5 3 (9) 9 (26) 16 (46) 1 (3) 25.7 2.0 11 (31) 14 (40)
0.001 0.04 0.2 0.4 0.1 0.7 0.04 0.6 0.01 0.003 0.01
6 (4)
2 (6)
expressed as mean standard deviation with range. Continuous variables were tested for normal distribution using the Shapiro-Wilk test. Statistical analyses were performed with IBM SPSS 23.0 software (IBM Corp, Armonk, NY). Two-tailed Student t test for continuous variables and the c2 test for categoric variables were used for univariate comparisons. Multivariate logistic regression was used to identify independent predictors of an indexed aortic diameter exceeding 22 mm/m2. All variables with a p of less than 0.1 in the univariate analysis were included in the logistic regression model. All p values of 0.05 or less were considered statistically significant. Intrarater/interrater reliability in measurements of angle LV/aorta was published in our previous report [18].
Results
0.8
a
Categoric data are presented as numbers (%) and continuous data as mean standard deviation.
BAV-AI ¼ bicuspid aortic valve insufficiency; BAV-AS ¼ bicuspid aortic valve stenosis; COLD ¼ chronic obstructive lung disease; NYHA ¼ New York Heart Association.
severe change), which was based on the worst area observed at light microscopy. The values of all seven histologic qualities were summarized in a histologic sum score (0 to 21) for the jet and control aortic samples. All histologic analyzes were performed by 2 experienced pathologists who were blinded to the group affiliation (BAV-AS vs BAV-AI) and the collection site of aortic specimens (jet sample vs control sample).
Statistical Analysis Standard definitions were used for patient variables and outcomes. Categoric variables are expressed as percentages, and normally distributed continuous variables are
Proximal Aortic Diameters and Aortic Phenotype Cross-sectional aortic diameters and proximal aortic phenotypes in the BAV-AS and BAV-AI groups are summarized in Table 2. The study cohorts differed significantly in maximal cross-sectional aortic diameters, with markedly larger aortic annulus and sinuses of Valsalva diameters in the BAV-AI group (43.2 6.7 mm vs 36.4 5.7 mm, p ¼ 0.001). This difference persisted after indexing proximal aortic dimensions for body surface area (23.7 4.0 mm/m2 vs 21.1 3.9 mm/m2, p ¼ 0.001). Furthermore, the percentage of patients with an indexed maximal aortic diameter exceeding 22 mm/m2 was significantly higher in the BAV-AI vs BAV-AS group (60% vs 34%, p ¼ 0.01). We found significant differences in the prevalence of various aortic phenotypes between the study subgroups (Table 2). In particular, the prevalence of root dilatation phenotype was significantly higher in the BAV-AI vs BAV-AS group (29% vs 7%, p ¼ 0.01). Corresponding to this finding, replacement of the aortic root was more frequent in the BAV-AI
Table 2. Proximal Aortic Diametersa in Both Subgroups Variableb Aortic annulus, mm Sinuses of Valsalva, mm Sinotubular junction, mm Middle ascending aorta, mm Proximal aortic arch, mm Indexed aortic diameter, mm/m2 Indexed aortic diameter >22 mm/m2 Proximal aortic phenotype Normal aorta Root dilatation Middle ascending Distal ascending
BAV-AI (n ¼ 35)
p Value
27.1 3.2 36.4 5.7 31.3 5.1 40.2 7.2 33.1 4.6 21.1 3.9 47 (34)
32.3 3.5 43.2 6.7 35.1 3.9 42.7 8.5 32.1 3.6 23.7 4.0 21 (60)
0.001 0.001 0.01 0.2 0.3 0.001 0.01
61 (44) 10 (7) 63 (47) 3 (2)
a Based on preoperative magnetic resonance imaging data. deviation.
BAV-AI ¼ bicuspid aortic valve insufficiency;
BAV-AS (n ¼ 137)
b
10 10 15 0
(29) (29) (42) (0)
0.1 0.001 0.7 0.8
Categoric data are presented as numbers (%) and continuous data as mean standard
BAV-AS ¼ bicuspid aortic valve stenosis.
Ann Thorac Surg 2016;-:-–-
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
group (31% vs 2%, p ¼ 0.003). The middle ascending aorta phenotype was the most common type of proximal aortopathy in both study subgroups (47% vs 42%, p ¼ 0.7). Distal ascending/proximal aortic arch phenotype was very uncommon in our study population (2% vs 0%, p ¼ 0.8).
Cusp Fusion Pattern The fusion of R/L coronary cusps (BAV type I) was the most common BAV morphotype in both study subgroups (83% BAV-AI vs 72% BAV-AS, p ¼ 0.2; Table 3). However, we found a significantly higher prevalence of R/M cusp fusion (BAV type I) in the BAV-AS subgroup (28% vs 11%, p ¼ 0.03). Moreover, there were 2 patients with BAV type 0 morphotype in the BAV-AI group (6% vs 0%, p ¼ 0.6).
Functional Aortic Root Variables All MRI-analyzed functional/rheologic variables of transvalvular flow are presented in Table 3. Eccentric systolic transvalvular flow jet hitting the right lateral/right posterior segment of the tubular ascending aorta (ie, greater curvature) was found in 136 of 137 patients (99%) in the BAV-AS group. However, only 2 BAV-AI patients (6%) demonstrated similar eccentric systolic flow jet, and the remaining patients showed homogeneous systolic transvalvular flow propagation along the central axis of aortic root. There was no significant difference in angle LV/aorta in the BAV-AS vs BAV-AI group (49.80 10.50 vs 45.70 9.20, p ¼ 0.08) nor was the prevalence of patients with angle LV/aorta exceeding 500 significantly different between study groups (ie, 43% vs 31%, p ¼ 0.3). Table 3. Morphologic and Functional Variables of the Aortic Roota Variableb BAV cusp fusion pattern Type I R/L Type I right/noncoronary Type 0 Eccentric transvalvular flow jet Angle LV/aorta (degrees) Angle LV/aorta >500 Histologic sum score Jet sample Control sample Difference [jet sample – control sample]
BAV-AS (n ¼ 137)
BAV-AI (n ¼ 35)
p Value
99 (72) 38 (28) 0 (0) 136 (99)
29 (83) 4 (11) 2 (6) 2 (6)
0.2 0.03 0.6 0.001
49.8 10.5 59 (43)
45.7 9.2 11 (31)
0.1 0.3
2.5 1.4 1.7 1.6 1.0 0.9
3.7 2.6 3.5 1.9 0.2 0.7
0.03 0.01 0.04
a
b Based on preoperative magnetic resonance imaging data. Categoric data are presented as numbers (%) and continuous data as mean standard deviation.
BAV-AI ¼ bicuspid aortic valve insufficiency; aortic valve stenosis; LV ¼ left ventricle.
BAV-AS ¼ bicuspid
5
Histologic Changes in the Aortic Wall The histologic sum score in the jet sample and the control sample in both study subgroups is reported in Table 3. The histologic sum scores were significantly higher in the BAV-AI group than in the BAV-AS group (3.7 2.6 vs 2.5 1.4, p ¼ 0.03; Fig 4). Histologic changes in the individual subcategories of sum score in both groups consisted most commonly of medial necrosis and elastic fragmentation (data not shown). Moreover, we were able to demonstrate significant differences in the pattern of histologic lesions between jet and control samples in both study subgroups (Fig 4). The histologic sum score was significantly higher in the jet sample than in the control sample in the BAV-AS group (2.5 1.4 vs 1.7 1.6, p ¼ 0.04); however, the difference between the histologic sum score in the jet sample and sum score in the control sample in the BAV-AI group was not significant (Table 3). Logistic regression analysis was performed to identify predictors of indexed cross-sectional aortic diameter exceeding 22 mm/m2 (Table 4). BAV-AI (odds ratio, 4.7; p ¼ 0.007) and the angle LV/aorta (odds ratio, 1.3; p ¼ 0.01) were significantly associated with the indexed crosssectional aortic diameter exceeding 22 mm/m2.
Comment BAV morphotype and patterns of proximal aortic disease have been used to propose distinct clinical BAV phenotypes to explain the differences in clinical presentation, natural history, and prognosis of BAV patients [6–9]. Although the evidence of BAV aortopathy heterogeneity has gained increasing recognition in the last decade, data on individual aortic phenotypes are still scarce. Most of the published natural history and follow-up studies have consisted of mixed BAV cohorts and included different stages of BAV disease. Previous retrospective and autopsy studies demonstrated some important baseline differences between BAV-AS vs BAV-AI patients in that pure/predominant BAV insufficiency occurs in younger patients, more frequently male, and accounts for only 10% to 15% of BAV lesions [20, 21]. These differences in demographics between those two BAV cohorts have been shown to correlate with distinct patterns of aortopathy in BAV-AS vs BAV-AI patients [21, 22]. Furthermore, differences in the pathobiologic mechanisms of aortopathy have been suggested by investigations of the histologic and extracellular matrix protein changes [12, 13], also consistent with the clinical evidence from post-AVR follow-up studies [11]. A recent meta-analysis demonstrated a 10-fold higher risk of aortic dissection in patients who undergo isolated AVR for BAV insufficiency compared with BAV stenosis [14]. In our current study, we aimed to prospectively address the issue of BAV phenotypic heterogeneity and severity and expression of bicuspid aortopathy in a consecutive cohort of patients referred for AVR operations due to BAV disease. We prospectively performed cardiovascular
6
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
Ann Thorac Surg 2016;-:-–-
Fig 4. Histologic sum score in bicuspid aortic valve subgroups with aortic stenosis (BAV-AS) vs aortic insufficiency (BAV-AI). The bold horizontal line in the box indicates the mean; the top and bottom borders of the box mark standard deviation, and the whiskers mark the range.
MRI analysis, MRI-guided sampling of aortic tissue during the AVR operation, and quantitative histologic examination of this aortic tissue. A modified MRI-based aortic phenotype classification was implemented, which may coincide with the classification by Kang and colleagues [9]. The main finding of our study is that the expression and severity of bicuspid aortopathy was significantly different between BAV-AI and BAV-AS subgroups. We were able to demonstrate significant differences in proximal aortic diameters and phenotypes (Table 2) and in histologic lesions of aortic tissue (Table 3) between the BAV-AI and BAV-AS groups. This prospective study confirms previous retrospective findings that dilatation of the aortic annulus and aortic root (ie, sinuses of Valsalva) is significantly worse and more frequent in the BAV-AI subgroup, as demonstrated by larger aortic diameters, higher prevalence of root dilatation phenotype, and resultant aortic root operation, as well as higher values of the histologic sum score. Our Table 4. Predictors of Indexed Proximal Aortic Diameter Exceeding 22 mm/m2 by Logistic Regression) Variable a
Right/left cusp fusion Arterial hypertension Age Angle LV/aorta BAV insufficiencyb a
Compared variable).
b
Odds Ratio
95% CI
p Value
1.5 1.4 1.1 1.3 4.7
0.5–4.6 0.2–1.7 0.9–1.2 1.1–1.7 1.7–14.5
0.4 0.3 0.8 0.01 0.007
with right/noncoronary cusp fusion (categoric Compared with BAV stenosis (categoric variable).
BAV ¼ bicuspid aortic valve; ventricle; OR ¼ odds ratio.
CI ¼ confidence interval;
LV ¼ left
more conservative surgical strategy with regard to aortopathy in BAV-AS patients (ie, 34% BAV-AS patients had indexed aortic diameters exceeding 22 mm/m2, whereas only 18% subsequently underwent simultaneous aortic operations) was supported retrospectively by our previous finding of very low risk of adverse aortic events in BAV-AS patients with a moderately dilated aorta after an isolated AVR operation [11]. In line with the previous findings, we found significant differences in demographics between the study subgroups. Interestingly, the prevalence of specific cusp fusion patterns was also different between the study groups: type I R/N fusion was significantly more frequent in the BAV-AS than in the BAV-AI patients (Table 3). This finding is in accordance with the data published previously by Aicher and coauthors [23], demonstrating that R/N cusp fusion was found only rarely (9%) in BAVAI patients referred for BAV repair operations. Interestingly, the proximal aortic phenotype was still inhomogeneous in the BAV-AI group. Although root dilatation was significantly more common in the BAV-AI group (29% vs 7%, p ¼ 0.001), the middle ascending phenotype was still the most common type of aortopathy in both study cohorts. Furthermore, there was a relevant subgroup of BAV-AI patients (29%) with a normally sized aorta who presented with aortic valve insufficiency due to prolapse of the fused cusp and marked dilatation of aortoventricular junction. Thus, the above valve-function/ aortic-dilatation associations, although statistically significant, are not absolute, and there is heterogeneity even within subgroups of valve function, as reported previously from large observational series [24]. Eccentric systolic transvalvular flow jet was demonstrated in 136 of 137 BAV-AS patients (99%) but in only 2 patients (6%) in the BAV-AI group (Fig 5). The
Ann Thorac Surg 2016;-:-–-
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
7
Fig 5. Two-dimensional phasecontrast magnetic resonance imaging shows propagation of transvalvular flow in patients with bicuspid aortic valve (A) stenosis vs (B) insufficiency shown.
angulation between the LV outflow axis and aortic root (angle LV/aorta), which has been recently demonstrated to significantly influence aortopathy in BAV stenosis patients [18], was not significantly different in the BAV-AI vs BAV-AS patients (Table 3). Given the lack of eccentricity in transvalvular flow patterns of BAV-AI patients, the calculation of angle jet/aorta, which correlated significantly with the BAV-AS associated aortopathy in our previous study [18], was not possible in the BAV-AI cohort. We therefore excluded this variable from the current analysis. The histologic sum score in the jet and control sample was significantly higher in the BAV-AI subgroup than in the BAV-AS subgroup. Moreover, there was a relevant difference in the pattern of histologic lesions between jet and control samples in both study subgroups. On the one hand, the sum score was significantly higher in the jet sample than in the control sample in the BAV-AS group, which may indicate the effect of the transvalvular flow jet in the progression of BAV-AS–associated aortopathy. On the other hand, histologic lesions between the jet and control sample were very comparable in BAV-AI patients, which may be a marker of more diffuse aortopathy, potentially less influenced by the rheologic factors. Logistic regression analysis revealed that BAV insufficiency and the angle LV/aorta were significantly associated with the indexed aortic diameter exceeding 22 mm/m2 (Table 4). The significant association between the angle LV/aorta and the severity of bicuspid aortopathy may be further explained by the 80% prevalence of BAV-AS patients in our study population. In our previous study, we were able to demonstrate a significant effect of this “subvalvular component” on the expression and severity of BAV-AS–associated aortopathy [18]. The lack of comparison with a tricuspid aortic valve group is a major limitation of our study. Only patients
with BAV disease referred for an AVR operation were enrolled. Although this clinical entity represents a consecutive surgical BAV cohort, our findings may not be generalizable to the whole BAV population. The use of z-score approach for definition of proximal aortic dilatation might have been more appropriate in the era of individualized medical therapy. Moreover, we have no longitudinal MRI data on the development of BAV aortopathy in patients with BAV-AI vs BAV-AS and, therefore, may not answer the question regarding aortopathy progression over time in the study subgroups. Data to support a causative relationship vs a correlative relationship between BAV phenotype and aortopathy are lacking. We conclude that BAV functional phenotype correlates significantly with the expression and severity of bicuspid aortopathy. Our prospective analysis supports further the hypothesis that patients with BAV-AI vs BAV-AV may represent distinct clinical and pathologic entities and should be analyzed separately when reporting any relevant events in the future BAV studies.
References 1. Della Corte A. Phenotypic heterogeneity of bicuspid aortopathy: a potential key to decode the prognosis? Heart 2014;100:96–7. 2. Girdauskas E, Borger MA. Bicuspid aortic valve and associated aortopathy: an update. Semin Thorac Cardiovasc Surg 2013;25:310–6. 3. Mahadevia R, Barker AJ, Schnell S, et al. Bicuspid aortic cusp fusion morphology alters aortic three-dimensional outflow patterns, wall shear stress, and expression of aortopathy. Circulation 2014;129:673–82. 4. Detaint D, Michelena HI, Nkomo VT, et al. Aortic dilatation patterns and rates in adults with bicuspid aortic valves: a comparative study with Marfan syndrome and degenerative aortopathy. Heart 2014;100:126–34.
8
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
5. Della Corte A, Bancone C, Dialetto G, et al. The ascending aorta with bicuspid aortic valve: a phenotypic classification with potential prognostic significance. Eur J Cardiothorac Surg 2014;46:240–7. 6. Pag e M, Mongeon FP, Stevens LM, et al. Aortic dilation rates in patients with bicuspid aortic valve: correlations with cusp fusion phenotype. J Heart Valve Dis 2014;23:450–7. 7. Fazel SS, Mallidi HR, Lee RS, et al. The aortopathy of bicuspid aortic valve disease has distinctive patterns and usually involves the transverse aortic arch. J Thorac Cardiovasc Surg 2008;135:901–7. 8. Schaefer BM, Lewin MB, Stout KK, et al. The bicuspid aortic valve: an integrated phenotypic classification of leaflet morphology and aortic root shape. Heart 2008;94: 1634–8. 9. Kang JW, Song HG, Yang DH, et al. Association between bicuspid aortic valve phenotype and patterns of valvular dysfunction and bicuspid aortopathy: comprehensive evaluation using MDCT and echocardiography. JACC Cardiovasc Imaging 2013;6:150–61. 10. Padang R, Bannon PG, Jeremy R, et al. The genetic and molecular basis of bicuspid aortic valve associated thoracic aortopathy: a link to phenotype heterogeneity. Ann Cardiothorac Surg 2013;2:83–91. 11. Girdauskas E, Disha K, Secknus M, Borger M, Kuntze T. Increased risk of late aortic events after isolated aortic valve replacement in patients with bicuspid aortic valve insufficiency versus stenosis. J Cardiovasc Surg (Torino) 2013;54: 653–9. 12. Roberts WC, Vowels TJ, Ko JM, et al. Comparison of the structure of the aortic valve and ascending aorta in adults having aortic valve replacement for aortic stenosis versus for pure aortic regurgitation and resection of the ascending aorta for aneurysm. Circulation 2011;123:896–903. 13. Girdauskas E, Rouman M, Borger MA, Kuntze T. Comparison of aortic media changes in patients with bicuspid aortic valve stenosis versus bicuspid valve insufficiency and proximal aortic aneurysm. Interact Cardiovasc Thorac Surg 2013;17:931–6. 14. Girdauskas E, Rouman M, Disha K, et al. Aortic dissection after previous aortic valve replacement for bicuspid aortic valve disease. J Am Coll Cardiol 2015;66:1409–11.
Ann Thorac Surg 2016;-:-–-
15. Baumgartner H, Hung J, Bermejo J, et al. Echocardiographic assessment of valve stenosis: EAE/ASE recommendations for clinical practice. Eur J Echocardiogr 2009;10:1–25. 16. Lancellotti P, Tribouilloy C, Hagendorff A, et al; Scientific Document Committee of the European Association of Cardiovascular Imaging. Recommendations for the echocardiographic assessment of native valvular regurgitation: an executive summary from the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging 2013;14:611–44. 17. Burman ED, Keegan J, Kilner PJ. Aortic root measurement by cardiovascular magnetic resonance: specification of planes and lines of measurement and corresponding normal values. Circ Cardiovasc Imaging 2008;1:104–13. 18. Girdauskas E, Rouman M, Disha K, et al. Aortopathy in patients with bicuspid aortic valve stenosis: role of aortic root functional parameters. Eur J Cardiothorac Surg 2016;49: 635–43; discussion 643–4. 19. Bechtel M, Noack F, Sayk F, et al. Histopathological grading of ascending aortic aneurysm: comparison of patients with bicuspid versus tricuspid aortic valve. J Heart Valve Dis 2003;12:54–61. 20. Sabet HY, Edwards WD, Tazelaar HD, et al. Congenitally bicuspid aortic valves: a surgical pathology study of 542 cases (1991 through 1996) and a literature review of 2,715 additional cases. Mayo Clin Proc 1999;74:14–26. 21. Della Corte A, Bancone C, Quarto C, et al. Predictors of ascending aortic dilatation with bicuspid aortic valve: a wide spectrum of disease expression. Eur J Cardiothorac Surg 2007;31:397–404. 22. Hahn RT, Roman MJ, Mogtader AH, Devereux RB. Association of aortic dilation with regurgitant, stenotic and functionally normal bicuspid aortic valves. J Am Coll Cardiol 1992;19:283–8. 23. Aicher D, Kunihara T, Abou Issa O, Brittner B, Gr€aber S, Sch€ afers HJ. Valve configuration determines long-term results after repair of the bicuspid aortic valve. Circulation 2011;123:178–85. 24. Keane MG, Wiegers SE, Plappert T, Pochettino A, Bavaria JE, Sutton MG. Bicuspid aortic valves are associated with aortic dilatation out of proportion to coexistent valvular lesions. Circulation 2000;102:III35–9.
DISCUSSION DR DEREK R. BRINSTER (New York, NY): That was an excellent talk. I saw a brief slide on this. We know that the most common fusion pattern is right-left and that the most common aortopathy is the type I that you describe with ascending aortic root. The importance of your study is whether that was related to the aortic stenosis or not, is that correct? Is that the summary here? DR GIRDAUSKAS: That is the summary, yes. DR BRINSTER: So you found basically that the commissure pattern related to the aortopathy? DR GIRDAUSKAS: Yes. It was published already by Schaefer’s group. He analyzed specifically patients who were repaired for bicuspid insufficiency and reported, similar to our study, a very high prevalence of patients with right/left fusion and only about 10% of patients with right/noncoronary fusion. So it is different from the stenosis subgroup. DR TOHRU ASAI (Shiga, Japan): It was an excellent presentation. I have a question. The aortopathy in ascending aorta in the proximal arch is quite different from the ordinary. In my series
we have 3 cases of aortic replacement even after aortic valve replacement (AVR). In this phenotype difference do you closely follow-up that post-AVR patient by some aortic area, difference of aortic area to be replaced? DR GIRDAUSKAS: We published these studies on post-AVR follow-up separately for patients with stenosis and insufficiency phenotypes, and actually the main message from these is that over 10 to 15 years after isolated AVR surgery, we see only very few patients who had adverse aortic events in the stenosis subgroup, it was about 2% to 3% only, but we saw a lot of patients coming back for aortic surgery who underwent previous isolated AVR for pure insufficiency. So it is quite a remarkable difference between those two functional phenotypes. DR JEHANGIR APPOO (Calgary, AB, Canada): Thank you. A nice study. It is an interesting question right now in terms of aortopathy with bicuspid valves: Is it due to a genetic association with aortic wall weakness or is it due to a flow pattern? I was a little bit confused by what you said. I thought you said that patients with insufficiency had a higher incidence of aortopathy, but when you did the histological studies, it
Ann Thorac Surg 2016;-:-–-
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
9
was the patients with aortic stenosis that had the difference, and there was no difference between the control vs the jet lesion?
DR GIRDAUSKAS: I think that in patients with stenosis it is mostly a flow-related problem and in insufficiency patients there is a major impact of genetic component.
DR GIRDAUSKAS: I would come back to a slide. Actually what we saw was the histological sum score in the insufficiency subgroup, which was significantly higher.
DR APPOO: Interesting. Thank you.
DR APPOO: But no difference between the control versus the jet lesion, right? DR GIRDAUSKAS: In patients with bicuspid aortic valve (BAV) insufficiency we saw no difference between those two samples taken from different parts of the aorta, which indicates that the flow is not the issue in these patients, and in the stenosis subgroup, there was a significant difference between the jet sample and the control sample. In the jet area the histological score was higher, which indicates an impact of a transvalvular flow jet. DR APPOO: But those patients did not have aortic dilatation, right? DR GIRDAUSKAS: Some of them. DR APPOO: So the aneurysm that we see with bicuspid valve, do you think it is flow related or do you think it is actually a weakness in the wall?
DR KAZIMIERZ JAN WIDENKA (Rzesz ow, Poland): I really enjoyed your presentation, and I have a question. How many patients in the group with aortic insufficiency could have Marfan syndrome? That could have an impact on your results. DR GIRDAUSKAS: Thank you for this question. Actually, Marfan patients were excluded completely from this study. We had no Marfans. DR WIDENKA: Was this histologically or genetically confirmed or just from the history of the patient? DR GIRDAUSKAS: Actually none of them had typical phenotypes of Marfan syndrome. I saw very rarely patients with Marfan syndrome having bicuspid aortic valve. Actually it is a different phenotype, in our experience. DR WIDENKA: A few percent of patients with Marfan syndrome will have bicuspid aortic valve. It is not a large number, but there is, and that could have influence on the results. DR GIRDAUSKAS: Okay. Thank you for this comment.
Background. Bicuspid aortic valve (BAV)–associated aortopathy is heterogeneous and still insufficiently defined. We prospectively analyzed the morphologic and functional variables of aortopathy in patients undergoing operations for BAV insufficiency (BAV-AI) vs stenosis (BAV-AS). Methods. A total of 172 consecutive patients (71% male, 59 ± 10 years) underwent aortic valve replacement with or without proximal aortic operation for BAV-AS (n [ 137), and BAV-AI (n [ 35) from January 2012 through December 2014. All patients underwent preoperative cardiac magnetic resonance imaging to evaluate morphologic and functional variables of the aortic root. Magnetic resonance imaging data were used to guide sampling of aortic tissue intraoperatively (ie, from the area where flow jet impacts on the aortic wall [jet sample] and the opposite aortic wall [control sample]). Aortic wall lesions were graded based on the histologic sum score (range, 0 to 21). Expression and severity of aortopathy were quantified by means of
proximal aortic phenotype, indexed aortic diameters, and a sum score. Results. Cross-sectional aortic diameters were significantly larger in the BAV-AI group vs the BAV-AS group (47 ± 8 mm vs 41 ± 8 mm, p [ 0.001). Moreover, root dilatation phenotype was more frequent in the BAV-AI group (27% vs 6%, p [ 0.01) and was associated with a significantly larger aortic annulus diameter (32 ± 3 mm vs 27 ± 3 mm, p < 0.001). The histologic sum score was significantly different between the study groups (3.7 ± 2.6 BAV-AI vs 2.5 ± 1.4 BAV-AS, p [ 0.03). Logistic regression revealed a significant association between BAV-AI and indexed aortic diameter exceeding 22 mm/m2 (odds ratio, 4.7; p [ 0.007). Conclusions. Our study demonstrates that BAV functional phenotype correlates significantly with the expression and severity of bicuspid aortopathy.
B
etiologic entities [10, 11]. Distinct microstructural aortic wall lesions have been specifically demonstrated in patients with BAV-AI insufficiency vs BAV-AS [12, 13]. Moreover, recent meta-analysis demonstrated a 10-fold higher risk of aortic dissection in patients who undergo isolated aortic valve replacement (AVR) for BAV-AI compared with BAV-AS [14]. However, these data are limited by retrospective study design. Therefore, in our current study, we aimed to prospectively analyze the morphologic and functional parameters of aortopathy in patients undergoing operations for BAV-AI vs BAV-AS.
icuspid aortic valve (BAV)–associated aortopathy is heterogeneous and still insufficiently defined [1, 2]. Sophisticated imaging series [3], community-based echocardiographic follow-up data [4], and extensive surgical cohorts [5] have been pooled to evaluate a number of comprehensive classification systems based on BAV morphology [6] and the shape of the proximal aorta [7], or a combination of both factors [8, 9]. Several recent retrospective studies indicated that patients with BAV insufficiency (BAV-AI) vs BAV stenosis (BAV-AS) represent distinct clinical and pathologic phenotypes and therefore should be interpreted as distinct Accepted for publication May 20, 2016. Presented at the Fifty-second Annual Meeting of the Society of Thoracic Surgeons, Phoenix, AZ, Jan 23–27, 2016. Address correspondence to Dr Girdauskas, Department of Cardiovascular Surgery, University Heart Center Hamburg, Martinistraße 52, 20246 Hamburg, Germany; email: [email protected].
Ó 2016 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2016;-:-–-) Ó 2016 by The Society of Thoracic Surgeons
Patients and Methods The prospective institutional BAV database (Central Hospital, Bad Berka, Germany) included all consecutive patients with BAV disease who were referred for AVR operations, with or without simultaneous replacement of 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2016.05.085
2
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
the proximal aorta, from January 2012 through December 2014. Thuringian Chamber of Physicians Ethics Committee approved the study, and all patients gave written informed consent. Elective AVR operations were performed in 228 consecutive patients with BAV disease, with or without simultaneous proximal aortic replacement, at our institution. We excluded patients with Marfan syndrome and those undergoing combined procedures (ie, other than simultaneous aortic operations; n ¼ 31) or urgent or emergency operations (n ¼ 15). Nine BAV patients had contraindications for preoperative magnetic resonance imaging (MRI) examination and were excluded. The study cohort comprised 172 BAV patients (71% male), aged 59 10 years, who met these inclusion criteria. Predominant/isolated BAV-AS was identified in 137 patients (80%; BAV-AS group), and the remaining 35 BAV patients (20%) had pure aortic valve insufficiency (BAV-AI group). The primary end point of our study was the expression and severity of aortopathy in these two groups.
Definitions and Measurements Preoperative echocardiography and cardiac MRI were used to assess morphology and function of the aortic valve in all patients. Published guidelines were used to define aortic valve stenosis and insufficiency [15, 16]. Patients with mixed BAV disease were classified according to the predominant functional valve lesion; for example, BAV patients with severe stenosis and mild-tomoderate aortic valve insufficiency were classified as having BAV-AS. The final decision regarding the bicuspidality was made based on the intraoperative aortic valve description by the surgeon. Although transthoracic echocardiography failed to correctly identify aortic valve morphology in 54 patients (31%), cardiac MRI demonstrated 98% sensitivity and specificity in identifying BAV disease. The diameters of proximal aorta (ie, aortic annulus, sinuses of Valsalva, sinotubular junction, ascending aorta, aortic arch) were measured preoperatively by
Ann Thorac Surg 2016;-:-–-
cardiac MRI. All aortic diameters were measured as the largest observed cross-sectional diameter perpendicular to the aortic axis in a midvessel slice at end diastole using the inner edge–to–inner edge technique, in accordance with previously published recommendations for MRI measurements [17]. Similar to others [7–9], we determined aortic phenotype based on cardiac MRI data (Fig 1). Normal aorta phenotype was characterized by all aortic diameters less than 22 mm/m2 of body surface area and nonindexed aortic diameters of less than 40 mm (Fig 1A). Predominant aortic root dilatation was defined as maximal aortic dilatation exceeding 22 mm/m2 at the level of the sinuses of Valsalva or 40 mm in maximal diameter (Fig 1B). The middle ascending aorta phenotype was determined by maximal aortic diameters at the level of the middle ascending tubular aorta and exceeding 22 mm/m2 or 40 mm in maximal diameter (Fig 1C). Distal ascending/ aortic arch phenotype was diagnosed when maximal dimensions were measured at the level of distal ascending aorta or proximal aortic arch, or both, and exceeded 22 mm/m2 or 40 mm (Fig 1D).
Preoperative MRI Examination The following morphologic and functional variables were assessed: (1) cusp fusion pattern, (2) proximal aortic diameters, (3) proximal aortic phenotype, (4) angulation between the left ventricular (LV) outflow axis and the aortic root (angle LV/aorta). The protocol of cardiac MRI examination and the corresponding measurements have been described in detail in our previous publication [18]. Cusp fusion pattern, noted as right/left (R/L) vs right/ noncoronary (R/N) fusion of the BAV valve was determined from the preoperative MRI analysis and confirmed by intraoperative inspection. Cusp fusion pattern was determined by the orientation of the raphe and the shape of systolic BAV opening, by placing an steady-state free precession–function sequence parallel and just distal to the aortic valve annulus (Fig 2).
Fig 1. Four distinct proximal aortic phenotypes as identified by cardiac magnetic resonance imaging. (A) Normal aorta phenotype: all aortic diameters were less than 22 mm/m2 of body surface area and nonindexed aortic diameters were less than 40 mm. (B) Predominant aortic root dilatation: maximal aortic dilatation at the level of the sinuses of Valsalva exceeding 22 mm/m2 or 40 mm in maximal diameter. (C) Mid ascending aorta phenotype: maximal aortic diameters at the level of the mid ascending tubular aorta and exceeding 22 mm/m2 or 40 mm in maximal diameter. (D) Distal ascending/aortic arch phenotype: maximal dimensions at the level of distal ascending aorta or proximal aortic arch exceeded 22 mm/m2 or 40 mm.
Ann Thorac Surg 2016;-:-–-
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
3
Fig 2. Two different cusp fusion patterns in patients with bicuspid aortic valve disease: (A) right/noncoronary cusp fusion and (B) right/ left cusp fusion.
The angle LV/aorta was determined at peak systole in the LV inflow-outflow view (Fig 3). LV outflow axis was modeled as a vector between the LV apex and the midpoint of the LV outflow tract (15 mm below aortic valve plane). Two study radiologists (B.F. and G.D.), blinded to the intraoperative findings, performed the MRI analyses.
p ¼ 0.002). Moreover, significantly larger aortic valve prostheses were used in the BAV-AI group (25.7 2.0 mm vs 23.8 1.9, p ¼ 0.01). One patient in the BAV-AS group (0.6%) died inhospital of multiorgan failure induced by low cardiac output syndrome. A perioperative stroke occurred in 2 patients (1.2%). The in-hospital course in all remaining patients was uneventful.
Study Population Demographics and intraoperative variables are summarized in Table 1. Both study subgroups were quite different in baseline characteristics. In particular, the patients in BAV-AI group were significantly younger and more frequently male. The BAV-AS group had a tendency toward higher prevalence of comorbidities, including arterial hypertension and diabetes. The 137 BAV-AS patients underwent an elective AVR operation that was accompanied by a simultaneous supracoronary ascending aortic replacement in 22 patients (16%). Aortic root replacement by means of composite graft implantation was performed in only 3 BAV-AS patients (2%). Aortic valve repair, with or without simultaneous aortic root replacement, was successful in 30 BAV-AI patients (86%), and the remaining 5 patients (14%) underwent an AVR operation. Proximal aortic operations were significantly more prevalent in the BAV-AI than in the BAV-AS group (71% vs 18%,
Histologic Examination MRI data were used to obtain 2 aortic specimens, a “jet sample” and a “control sample” intraoperatively. The jet sample was obtained from the area of contact between the systolic transvalvular flow jet and the aortic wall, as determined by preoperative MRI analysis (ie, segment of aortic circumference that was in direct contact with the flow jet). The control sample was collected from the opposite aortic wall. Intraoperative collection of aortic tissue has been described in detail previously [18]. Histologic examination of the aortic tissue and assessment of aortic wall lesions by means of histologic grading scale [19] has been described previously [18]. Briefly, seven histologic criteria were used to grade aortic wall lesions: fibrosis, atherosclerosis, medionecrosis, cystic medial necrosis, changes in smooth muscle cell orientation, elastic fragmentation, and periaortic inflammation. Each variable was graded from 0 (no change) to 3 (most Fig 3. Measurement of angle left ventricle (LV)/aorta on cardiac magnetic resonance imaging. (A) Large angle between the LV outflow axis and the aortic root (i.e. angle LV/Aorta 68 ). (B) Small angle between the LV outflow axis and the aortic root (i.e. angle LV/Aorta 35 ).
4
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
Ann Thorac Surg 2016;-:-–-
Table 1. Demographics and Intraoperative Variables in Both Subgroups Variablea Age, y Male sex Body surface area, m2 NYHA Functional Class Diabetes History of smoking Arterial hypertension COLD Valve size, mm Aortic root replacement Ascending aortic replacement Hemiarch replacement
BAV-AS (n ¼ 137)
BAV-AI (n ¼ 35)
p Value
61 9 93 (68) 1.98 0.2 2.4 0.6 20 (15) 32 (23) 89 (65) 8 (6) 23.8 1.9 3 (2) 22 (16)
49 11 30 (86) 1.92 0.3 2.6 0.5 3 (9) 9 (26) 16 (46) 1 (3) 25.7 2.0 11 (31) 14 (40)
0.001 0.04 0.2 0.4 0.1 0.7 0.04 0.6 0.01 0.003 0.01
6 (4)
2 (6)
expressed as mean standard deviation with range. Continuous variables were tested for normal distribution using the Shapiro-Wilk test. Statistical analyses were performed with IBM SPSS 23.0 software (IBM Corp, Armonk, NY). Two-tailed Student t test for continuous variables and the c2 test for categoric variables were used for univariate comparisons. Multivariate logistic regression was used to identify independent predictors of an indexed aortic diameter exceeding 22 mm/m2. All variables with a p of less than 0.1 in the univariate analysis were included in the logistic regression model. All p values of 0.05 or less were considered statistically significant. Intrarater/interrater reliability in measurements of angle LV/aorta was published in our previous report [18].
Results
0.8
a
Categoric data are presented as numbers (%) and continuous data as mean standard deviation.
BAV-AI ¼ bicuspid aortic valve insufficiency; BAV-AS ¼ bicuspid aortic valve stenosis; COLD ¼ chronic obstructive lung disease; NYHA ¼ New York Heart Association.
severe change), which was based on the worst area observed at light microscopy. The values of all seven histologic qualities were summarized in a histologic sum score (0 to 21) for the jet and control aortic samples. All histologic analyzes were performed by 2 experienced pathologists who were blinded to the group affiliation (BAV-AS vs BAV-AI) and the collection site of aortic specimens (jet sample vs control sample).
Statistical Analysis Standard definitions were used for patient variables and outcomes. Categoric variables are expressed as percentages, and normally distributed continuous variables are
Proximal Aortic Diameters and Aortic Phenotype Cross-sectional aortic diameters and proximal aortic phenotypes in the BAV-AS and BAV-AI groups are summarized in Table 2. The study cohorts differed significantly in maximal cross-sectional aortic diameters, with markedly larger aortic annulus and sinuses of Valsalva diameters in the BAV-AI group (43.2 6.7 mm vs 36.4 5.7 mm, p ¼ 0.001). This difference persisted after indexing proximal aortic dimensions for body surface area (23.7 4.0 mm/m2 vs 21.1 3.9 mm/m2, p ¼ 0.001). Furthermore, the percentage of patients with an indexed maximal aortic diameter exceeding 22 mm/m2 was significantly higher in the BAV-AI vs BAV-AS group (60% vs 34%, p ¼ 0.01). We found significant differences in the prevalence of various aortic phenotypes between the study subgroups (Table 2). In particular, the prevalence of root dilatation phenotype was significantly higher in the BAV-AI vs BAV-AS group (29% vs 7%, p ¼ 0.01). Corresponding to this finding, replacement of the aortic root was more frequent in the BAV-AI
Table 2. Proximal Aortic Diametersa in Both Subgroups Variableb Aortic annulus, mm Sinuses of Valsalva, mm Sinotubular junction, mm Middle ascending aorta, mm Proximal aortic arch, mm Indexed aortic diameter, mm/m2 Indexed aortic diameter >22 mm/m2 Proximal aortic phenotype Normal aorta Root dilatation Middle ascending Distal ascending
BAV-AI (n ¼ 35)
p Value
27.1 3.2 36.4 5.7 31.3 5.1 40.2 7.2 33.1 4.6 21.1 3.9 47 (34)
32.3 3.5 43.2 6.7 35.1 3.9 42.7 8.5 32.1 3.6 23.7 4.0 21 (60)
0.001 0.001 0.01 0.2 0.3 0.001 0.01
61 (44) 10 (7) 63 (47) 3 (2)
a Based on preoperative magnetic resonance imaging data. deviation.
BAV-AI ¼ bicuspid aortic valve insufficiency;
BAV-AS (n ¼ 137)
b
10 10 15 0
(29) (29) (42) (0)
0.1 0.001 0.7 0.8
Categoric data are presented as numbers (%) and continuous data as mean standard
BAV-AS ¼ bicuspid aortic valve stenosis.
Ann Thorac Surg 2016;-:-–-
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
group (31% vs 2%, p ¼ 0.003). The middle ascending aorta phenotype was the most common type of proximal aortopathy in both study subgroups (47% vs 42%, p ¼ 0.7). Distal ascending/proximal aortic arch phenotype was very uncommon in our study population (2% vs 0%, p ¼ 0.8).
Cusp Fusion Pattern The fusion of R/L coronary cusps (BAV type I) was the most common BAV morphotype in both study subgroups (83% BAV-AI vs 72% BAV-AS, p ¼ 0.2; Table 3). However, we found a significantly higher prevalence of R/M cusp fusion (BAV type I) in the BAV-AS subgroup (28% vs 11%, p ¼ 0.03). Moreover, there were 2 patients with BAV type 0 morphotype in the BAV-AI group (6% vs 0%, p ¼ 0.6).
Functional Aortic Root Variables All MRI-analyzed functional/rheologic variables of transvalvular flow are presented in Table 3. Eccentric systolic transvalvular flow jet hitting the right lateral/right posterior segment of the tubular ascending aorta (ie, greater curvature) was found in 136 of 137 patients (99%) in the BAV-AS group. However, only 2 BAV-AI patients (6%) demonstrated similar eccentric systolic flow jet, and the remaining patients showed homogeneous systolic transvalvular flow propagation along the central axis of aortic root. There was no significant difference in angle LV/aorta in the BAV-AS vs BAV-AI group (49.80 10.50 vs 45.70 9.20, p ¼ 0.08) nor was the prevalence of patients with angle LV/aorta exceeding 500 significantly different between study groups (ie, 43% vs 31%, p ¼ 0.3). Table 3. Morphologic and Functional Variables of the Aortic Roota Variableb BAV cusp fusion pattern Type I R/L Type I right/noncoronary Type 0 Eccentric transvalvular flow jet Angle LV/aorta (degrees) Angle LV/aorta >500 Histologic sum score Jet sample Control sample Difference [jet sample – control sample]
BAV-AS (n ¼ 137)
BAV-AI (n ¼ 35)
p Value
99 (72) 38 (28) 0 (0) 136 (99)
29 (83) 4 (11) 2 (6) 2 (6)
0.2 0.03 0.6 0.001
49.8 10.5 59 (43)
45.7 9.2 11 (31)
0.1 0.3
2.5 1.4 1.7 1.6 1.0 0.9
3.7 2.6 3.5 1.9 0.2 0.7
0.03 0.01 0.04
a
b Based on preoperative magnetic resonance imaging data. Categoric data are presented as numbers (%) and continuous data as mean standard deviation.
BAV-AI ¼ bicuspid aortic valve insufficiency; aortic valve stenosis; LV ¼ left ventricle.
BAV-AS ¼ bicuspid
5
Histologic Changes in the Aortic Wall The histologic sum score in the jet sample and the control sample in both study subgroups is reported in Table 3. The histologic sum scores were significantly higher in the BAV-AI group than in the BAV-AS group (3.7 2.6 vs 2.5 1.4, p ¼ 0.03; Fig 4). Histologic changes in the individual subcategories of sum score in both groups consisted most commonly of medial necrosis and elastic fragmentation (data not shown). Moreover, we were able to demonstrate significant differences in the pattern of histologic lesions between jet and control samples in both study subgroups (Fig 4). The histologic sum score was significantly higher in the jet sample than in the control sample in the BAV-AS group (2.5 1.4 vs 1.7 1.6, p ¼ 0.04); however, the difference between the histologic sum score in the jet sample and sum score in the control sample in the BAV-AI group was not significant (Table 3). Logistic regression analysis was performed to identify predictors of indexed cross-sectional aortic diameter exceeding 22 mm/m2 (Table 4). BAV-AI (odds ratio, 4.7; p ¼ 0.007) and the angle LV/aorta (odds ratio, 1.3; p ¼ 0.01) were significantly associated with the indexed crosssectional aortic diameter exceeding 22 mm/m2.
Comment BAV morphotype and patterns of proximal aortic disease have been used to propose distinct clinical BAV phenotypes to explain the differences in clinical presentation, natural history, and prognosis of BAV patients [6–9]. Although the evidence of BAV aortopathy heterogeneity has gained increasing recognition in the last decade, data on individual aortic phenotypes are still scarce. Most of the published natural history and follow-up studies have consisted of mixed BAV cohorts and included different stages of BAV disease. Previous retrospective and autopsy studies demonstrated some important baseline differences between BAV-AS vs BAV-AI patients in that pure/predominant BAV insufficiency occurs in younger patients, more frequently male, and accounts for only 10% to 15% of BAV lesions [20, 21]. These differences in demographics between those two BAV cohorts have been shown to correlate with distinct patterns of aortopathy in BAV-AS vs BAV-AI patients [21, 22]. Furthermore, differences in the pathobiologic mechanisms of aortopathy have been suggested by investigations of the histologic and extracellular matrix protein changes [12, 13], also consistent with the clinical evidence from post-AVR follow-up studies [11]. A recent meta-analysis demonstrated a 10-fold higher risk of aortic dissection in patients who undergo isolated AVR for BAV insufficiency compared with BAV stenosis [14]. In our current study, we aimed to prospectively address the issue of BAV phenotypic heterogeneity and severity and expression of bicuspid aortopathy in a consecutive cohort of patients referred for AVR operations due to BAV disease. We prospectively performed cardiovascular
6
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
Ann Thorac Surg 2016;-:-–-
Fig 4. Histologic sum score in bicuspid aortic valve subgroups with aortic stenosis (BAV-AS) vs aortic insufficiency (BAV-AI). The bold horizontal line in the box indicates the mean; the top and bottom borders of the box mark standard deviation, and the whiskers mark the range.
MRI analysis, MRI-guided sampling of aortic tissue during the AVR operation, and quantitative histologic examination of this aortic tissue. A modified MRI-based aortic phenotype classification was implemented, which may coincide with the classification by Kang and colleagues [9]. The main finding of our study is that the expression and severity of bicuspid aortopathy was significantly different between BAV-AI and BAV-AS subgroups. We were able to demonstrate significant differences in proximal aortic diameters and phenotypes (Table 2) and in histologic lesions of aortic tissue (Table 3) between the BAV-AI and BAV-AS groups. This prospective study confirms previous retrospective findings that dilatation of the aortic annulus and aortic root (ie, sinuses of Valsalva) is significantly worse and more frequent in the BAV-AI subgroup, as demonstrated by larger aortic diameters, higher prevalence of root dilatation phenotype, and resultant aortic root operation, as well as higher values of the histologic sum score. Our Table 4. Predictors of Indexed Proximal Aortic Diameter Exceeding 22 mm/m2 by Logistic Regression) Variable a
Right/left cusp fusion Arterial hypertension Age Angle LV/aorta BAV insufficiencyb a
Compared variable).
b
Odds Ratio
95% CI
p Value
1.5 1.4 1.1 1.3 4.7
0.5–4.6 0.2–1.7 0.9–1.2 1.1–1.7 1.7–14.5
0.4 0.3 0.8 0.01 0.007
with right/noncoronary cusp fusion (categoric Compared with BAV stenosis (categoric variable).
BAV ¼ bicuspid aortic valve; ventricle; OR ¼ odds ratio.
CI ¼ confidence interval;
LV ¼ left
more conservative surgical strategy with regard to aortopathy in BAV-AS patients (ie, 34% BAV-AS patients had indexed aortic diameters exceeding 22 mm/m2, whereas only 18% subsequently underwent simultaneous aortic operations) was supported retrospectively by our previous finding of very low risk of adverse aortic events in BAV-AS patients with a moderately dilated aorta after an isolated AVR operation [11]. In line with the previous findings, we found significant differences in demographics between the study subgroups. Interestingly, the prevalence of specific cusp fusion patterns was also different between the study groups: type I R/N fusion was significantly more frequent in the BAV-AS than in the BAV-AI patients (Table 3). This finding is in accordance with the data published previously by Aicher and coauthors [23], demonstrating that R/N cusp fusion was found only rarely (9%) in BAVAI patients referred for BAV repair operations. Interestingly, the proximal aortic phenotype was still inhomogeneous in the BAV-AI group. Although root dilatation was significantly more common in the BAV-AI group (29% vs 7%, p ¼ 0.001), the middle ascending phenotype was still the most common type of aortopathy in both study cohorts. Furthermore, there was a relevant subgroup of BAV-AI patients (29%) with a normally sized aorta who presented with aortic valve insufficiency due to prolapse of the fused cusp and marked dilatation of aortoventricular junction. Thus, the above valve-function/ aortic-dilatation associations, although statistically significant, are not absolute, and there is heterogeneity even within subgroups of valve function, as reported previously from large observational series [24]. Eccentric systolic transvalvular flow jet was demonstrated in 136 of 137 BAV-AS patients (99%) but in only 2 patients (6%) in the BAV-AI group (Fig 5). The
Ann Thorac Surg 2016;-:-–-
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
7
Fig 5. Two-dimensional phasecontrast magnetic resonance imaging shows propagation of transvalvular flow in patients with bicuspid aortic valve (A) stenosis vs (B) insufficiency shown.
angulation between the LV outflow axis and aortic root (angle LV/aorta), which has been recently demonstrated to significantly influence aortopathy in BAV stenosis patients [18], was not significantly different in the BAV-AI vs BAV-AS patients (Table 3). Given the lack of eccentricity in transvalvular flow patterns of BAV-AI patients, the calculation of angle jet/aorta, which correlated significantly with the BAV-AS associated aortopathy in our previous study [18], was not possible in the BAV-AI cohort. We therefore excluded this variable from the current analysis. The histologic sum score in the jet and control sample was significantly higher in the BAV-AI subgroup than in the BAV-AS subgroup. Moreover, there was a relevant difference in the pattern of histologic lesions between jet and control samples in both study subgroups. On the one hand, the sum score was significantly higher in the jet sample than in the control sample in the BAV-AS group, which may indicate the effect of the transvalvular flow jet in the progression of BAV-AS–associated aortopathy. On the other hand, histologic lesions between the jet and control sample were very comparable in BAV-AI patients, which may be a marker of more diffuse aortopathy, potentially less influenced by the rheologic factors. Logistic regression analysis revealed that BAV insufficiency and the angle LV/aorta were significantly associated with the indexed aortic diameter exceeding 22 mm/m2 (Table 4). The significant association between the angle LV/aorta and the severity of bicuspid aortopathy may be further explained by the 80% prevalence of BAV-AS patients in our study population. In our previous study, we were able to demonstrate a significant effect of this “subvalvular component” on the expression and severity of BAV-AS–associated aortopathy [18]. The lack of comparison with a tricuspid aortic valve group is a major limitation of our study. Only patients
with BAV disease referred for an AVR operation were enrolled. Although this clinical entity represents a consecutive surgical BAV cohort, our findings may not be generalizable to the whole BAV population. The use of z-score approach for definition of proximal aortic dilatation might have been more appropriate in the era of individualized medical therapy. Moreover, we have no longitudinal MRI data on the development of BAV aortopathy in patients with BAV-AI vs BAV-AS and, therefore, may not answer the question regarding aortopathy progression over time in the study subgroups. Data to support a causative relationship vs a correlative relationship between BAV phenotype and aortopathy are lacking. We conclude that BAV functional phenotype correlates significantly with the expression and severity of bicuspid aortopathy. Our prospective analysis supports further the hypothesis that patients with BAV-AI vs BAV-AV may represent distinct clinical and pathologic entities and should be analyzed separately when reporting any relevant events in the future BAV studies.
References 1. Della Corte A. Phenotypic heterogeneity of bicuspid aortopathy: a potential key to decode the prognosis? Heart 2014;100:96–7. 2. Girdauskas E, Borger MA. Bicuspid aortic valve and associated aortopathy: an update. Semin Thorac Cardiovasc Surg 2013;25:310–6. 3. Mahadevia R, Barker AJ, Schnell S, et al. Bicuspid aortic cusp fusion morphology alters aortic three-dimensional outflow patterns, wall shear stress, and expression of aortopathy. Circulation 2014;129:673–82. 4. Detaint D, Michelena HI, Nkomo VT, et al. Aortic dilatation patterns and rates in adults with bicuspid aortic valves: a comparative study with Marfan syndrome and degenerative aortopathy. Heart 2014;100:126–34.
8
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
5. Della Corte A, Bancone C, Dialetto G, et al. The ascending aorta with bicuspid aortic valve: a phenotypic classification with potential prognostic significance. Eur J Cardiothorac Surg 2014;46:240–7. 6. Pag e M, Mongeon FP, Stevens LM, et al. Aortic dilation rates in patients with bicuspid aortic valve: correlations with cusp fusion phenotype. J Heart Valve Dis 2014;23:450–7. 7. Fazel SS, Mallidi HR, Lee RS, et al. The aortopathy of bicuspid aortic valve disease has distinctive patterns and usually involves the transverse aortic arch. J Thorac Cardiovasc Surg 2008;135:901–7. 8. Schaefer BM, Lewin MB, Stout KK, et al. The bicuspid aortic valve: an integrated phenotypic classification of leaflet morphology and aortic root shape. Heart 2008;94: 1634–8. 9. Kang JW, Song HG, Yang DH, et al. Association between bicuspid aortic valve phenotype and patterns of valvular dysfunction and bicuspid aortopathy: comprehensive evaluation using MDCT and echocardiography. JACC Cardiovasc Imaging 2013;6:150–61. 10. Padang R, Bannon PG, Jeremy R, et al. The genetic and molecular basis of bicuspid aortic valve associated thoracic aortopathy: a link to phenotype heterogeneity. Ann Cardiothorac Surg 2013;2:83–91. 11. Girdauskas E, Disha K, Secknus M, Borger M, Kuntze T. Increased risk of late aortic events after isolated aortic valve replacement in patients with bicuspid aortic valve insufficiency versus stenosis. J Cardiovasc Surg (Torino) 2013;54: 653–9. 12. Roberts WC, Vowels TJ, Ko JM, et al. Comparison of the structure of the aortic valve and ascending aorta in adults having aortic valve replacement for aortic stenosis versus for pure aortic regurgitation and resection of the ascending aorta for aneurysm. Circulation 2011;123:896–903. 13. Girdauskas E, Rouman M, Borger MA, Kuntze T. Comparison of aortic media changes in patients with bicuspid aortic valve stenosis versus bicuspid valve insufficiency and proximal aortic aneurysm. Interact Cardiovasc Thorac Surg 2013;17:931–6. 14. Girdauskas E, Rouman M, Disha K, et al. Aortic dissection after previous aortic valve replacement for bicuspid aortic valve disease. J Am Coll Cardiol 2015;66:1409–11.
Ann Thorac Surg 2016;-:-–-
15. Baumgartner H, Hung J, Bermejo J, et al. Echocardiographic assessment of valve stenosis: EAE/ASE recommendations for clinical practice. Eur J Echocardiogr 2009;10:1–25. 16. Lancellotti P, Tribouilloy C, Hagendorff A, et al; Scientific Document Committee of the European Association of Cardiovascular Imaging. Recommendations for the echocardiographic assessment of native valvular regurgitation: an executive summary from the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging 2013;14:611–44. 17. Burman ED, Keegan J, Kilner PJ. Aortic root measurement by cardiovascular magnetic resonance: specification of planes and lines of measurement and corresponding normal values. Circ Cardiovasc Imaging 2008;1:104–13. 18. Girdauskas E, Rouman M, Disha K, et al. Aortopathy in patients with bicuspid aortic valve stenosis: role of aortic root functional parameters. Eur J Cardiothorac Surg 2016;49: 635–43; discussion 643–4. 19. Bechtel M, Noack F, Sayk F, et al. Histopathological grading of ascending aortic aneurysm: comparison of patients with bicuspid versus tricuspid aortic valve. J Heart Valve Dis 2003;12:54–61. 20. Sabet HY, Edwards WD, Tazelaar HD, et al. Congenitally bicuspid aortic valves: a surgical pathology study of 542 cases (1991 through 1996) and a literature review of 2,715 additional cases. Mayo Clin Proc 1999;74:14–26. 21. Della Corte A, Bancone C, Quarto C, et al. Predictors of ascending aortic dilatation with bicuspid aortic valve: a wide spectrum of disease expression. Eur J Cardiothorac Surg 2007;31:397–404. 22. Hahn RT, Roman MJ, Mogtader AH, Devereux RB. Association of aortic dilation with regurgitant, stenotic and functionally normal bicuspid aortic valves. J Am Coll Cardiol 1992;19:283–8. 23. Aicher D, Kunihara T, Abou Issa O, Brittner B, Gr€aber S, Sch€ afers HJ. Valve configuration determines long-term results after repair of the bicuspid aortic valve. Circulation 2011;123:178–85. 24. Keane MG, Wiegers SE, Plappert T, Pochettino A, Bavaria JE, Sutton MG. Bicuspid aortic valves are associated with aortic dilatation out of proportion to coexistent valvular lesions. Circulation 2000;102:III35–9.
DISCUSSION DR DEREK R. BRINSTER (New York, NY): That was an excellent talk. I saw a brief slide on this. We know that the most common fusion pattern is right-left and that the most common aortopathy is the type I that you describe with ascending aortic root. The importance of your study is whether that was related to the aortic stenosis or not, is that correct? Is that the summary here? DR GIRDAUSKAS: That is the summary, yes. DR BRINSTER: So you found basically that the commissure pattern related to the aortopathy? DR GIRDAUSKAS: Yes. It was published already by Schaefer’s group. He analyzed specifically patients who were repaired for bicuspid insufficiency and reported, similar to our study, a very high prevalence of patients with right/left fusion and only about 10% of patients with right/noncoronary fusion. So it is different from the stenosis subgroup. DR TOHRU ASAI (Shiga, Japan): It was an excellent presentation. I have a question. The aortopathy in ascending aorta in the proximal arch is quite different from the ordinary. In my series
we have 3 cases of aortic replacement even after aortic valve replacement (AVR). In this phenotype difference do you closely follow-up that post-AVR patient by some aortic area, difference of aortic area to be replaced? DR GIRDAUSKAS: We published these studies on post-AVR follow-up separately for patients with stenosis and insufficiency phenotypes, and actually the main message from these is that over 10 to 15 years after isolated AVR surgery, we see only very few patients who had adverse aortic events in the stenosis subgroup, it was about 2% to 3% only, but we saw a lot of patients coming back for aortic surgery who underwent previous isolated AVR for pure insufficiency. So it is quite a remarkable difference between those two functional phenotypes. DR JEHANGIR APPOO (Calgary, AB, Canada): Thank you. A nice study. It is an interesting question right now in terms of aortopathy with bicuspid valves: Is it due to a genetic association with aortic wall weakness or is it due to a flow pattern? I was a little bit confused by what you said. I thought you said that patients with insufficiency had a higher incidence of aortopathy, but when you did the histological studies, it
Ann Thorac Surg 2016;-:-–-
GIRDAUSKAS ET AL BAV-ASSOCIATED AORTOPATHY
9
was the patients with aortic stenosis that had the difference, and there was no difference between the control vs the jet lesion?
DR GIRDAUSKAS: I think that in patients with stenosis it is mostly a flow-related problem and in insufficiency patients there is a major impact of genetic component.
DR GIRDAUSKAS: I would come back to a slide. Actually what we saw was the histological sum score in the insufficiency subgroup, which was significantly higher.
DR APPOO: Interesting. Thank you.
DR APPOO: But no difference between the control versus the jet lesion, right? DR GIRDAUSKAS: In patients with bicuspid aortic valve (BAV) insufficiency we saw no difference between those two samples taken from different parts of the aorta, which indicates that the flow is not the issue in these patients, and in the stenosis subgroup, there was a significant difference between the jet sample and the control sample. In the jet area the histological score was higher, which indicates an impact of a transvalvular flow jet. DR APPOO: But those patients did not have aortic dilatation, right? DR GIRDAUSKAS: Some of them. DR APPOO: So the aneurysm that we see with bicuspid valve, do you think it is flow related or do you think it is actually a weakness in the wall?
DR KAZIMIERZ JAN WIDENKA (Rzesz ow, Poland): I really enjoyed your presentation, and I have a question. How many patients in the group with aortic insufficiency could have Marfan syndrome? That could have an impact on your results. DR GIRDAUSKAS: Thank you for this question. Actually, Marfan patients were excluded completely from this study. We had no Marfans. DR WIDENKA: Was this histologically or genetically confirmed or just from the history of the patient? DR GIRDAUSKAS: Actually none of them had typical phenotypes of Marfan syndrome. I saw very rarely patients with Marfan syndrome having bicuspid aortic valve. Actually it is a different phenotype, in our experience. DR WIDENKA: A few percent of patients with Marfan syndrome will have bicuspid aortic valve. It is not a large number, but there is, and that could have influence on the results. DR GIRDAUSKAS: Okay. Thank you for this comment.