Gender disparities in short- and long-term outcomes have been documented in cardiac and valvular heart surgery. However, there is a paucity of data regarding these differences in the bicuspid aortic valve (BAV) population. The aim of this study was to examine gender-specific differences in short- and long-term outcomes after surgical aortic valve (AV) replacement in patients with BAV. A retrospective analysis was performed in 628 consecutive patients with BAV who underwent AV surgery from April 2004 to December 2013. To reduce bias when comparing outcomes by gender, propensity score matching obtained on the basis of potential confounders was used. Women with BAV who underwent AV surgery presented with more advanced age (mean 60.7 ± 13.8 vs 56.3 ± 13.6 years, p <0.001) and less aortic regurgitation (29% vs 44%, p <0.001) and had a higher risk for in-hospital mortality (mean Ambler score 3.4 ± 4.4 vs 2.5 ± 4.0, p = 0.015). After propensity score matching, women received more blood products postoperatively (48% vs 34%, p = 0.028) and had more prolonged postoperative lengths of stay (median 5 days [interquartile range 5 to 7] vs 5 days [interquartile range 4 to 6], p = 0.027). Operative, discharge, and 30-day mortality and overall survival were not significantly different. In conclusion, women with BAV who underwent AV surgery were older, presented with less aortic regurgitation, and had increased co-morbidities, lending higher operative risk. Although women received more blood products and had significantly longer lengths of stay, short- and long-term outcomes were similar.
Bicuspid aortic valve (BAV) disease affects approximately 1.3% of the population worldwide, with a male predominance of approximately 3:1. A common complication of BAV disease is aortic valve dysfunction, either stenosis or regurgitation or both. Aortic valve replacement (AVR) is usually performed at an earlier age in patients with BAV than in those with trileaflet aortic valves. In the Olmsted County series, the average age for BAV surgery was approximately 27 years younger than in patients with tricuspid aortic valves. In addition to valvular abnormalities, dilatation of the aortic root and/or the ascending aorta (aortopathy) is also present in approximately 20% to 84% of patients with BAV, depending on the definition. Although gender differences have been widely reported in patients with BAV disease, there is a paucity of data regarding gender-related presentation and impact on surgical outcomes. We sought to compare short- and long-term outcomes between the genders in patients with BAV who underwent AVR.
Methods
Data for this project were obtained from the Cardiovascular Research Database in the Clinical Trial Unit of the Bluhm Cardiovascular Institute at Northwestern Memorial Hospital. This database was approved by the Institutional Review Board at Northwestern University (project STU00012288). Any subjects refusing participation in the project were not included in the analysis. The database was queried for patients with BAV who underwent AVR from April 2004 to December 2013. The accuracy of 2-dimensional echocardiography in establishing aortic valve structure and morphology has been evaluated in patients with aortic stenosis who underwent AVR. The determination of a bicuspid valve and morphology was made at the time of surgery and, if unclear from the operative report, was based on preoperative magnetic resonance imaging (first choice) or echocardiography. The distinction between BAV and degenerative fusion of tricuspid aortic valve leaflets was based on the classification system previously described by Sievers and Schmidtke.
After excluding patients with ventricular assist devices or cardiac trauma (n = 7), missing aortic measurements (n = 59), and preoperative aortic dissection (n = 6), we arrived at the final cohort of 628 patients. Operations were performed by 6 surgeons, with 3 of them accounting for >90% of these operations.
To reduce potential confounding due to baseline imbalances, propensity score (PS)–matching methods were used. The genders were compared using 1-to-1 PS matching on the basis of a greedy algorithm with a caliper of size 0.1 logit PS SD units, adjusting for the following factors: age, body mass index (BMI), the left ventricular ejection fraction, maximal aortic diameter, surgeon, aneurysm resection (yes or no), severe aortic stenosis, aortic regurgitation grade ≥2+, mitral valve surgery, elective status, repeat sternotomy, previous cardiovascular intervention, family history of coronary artery disease, diabetes, hypercholesterolemia, hypertension, chronic obstructive pulmonary disease, previous heart failure, New York Heart Association functional class III or IV symptoms, and medications.
Baseline demographics, pre-, intra-, and postoperative characteristics, and outcomes are listed in Table 1 . Balance in characteristics in the PS-matched groups was assessed on the basis of formal statistical testing and confirmed visually using bean plots, a particular type of kernel-smoothed histogram ( Figure 1 ). Postsurgery medium-term overall survival was summarized using Kaplan-Meier curves, and gender comparisons were based on the log-rank test ( Figure 2 ). One-way analysis of variance or Kruskal-Wallis tests were used to compare continuously distributed variables. Chi-square or Fisher’s exact tests were used to compare discrete variables. Statistical significance was declared at a 2-sided 5% α level, and there were no multiplicity adjustments. Statistical analyses were performed using SAS version 9.3 (SAS Institute Inc., Cary, North Carolina) and R version 2.12.2 (R Foundation for Statistical Computing, Vienna, Austria).
| Men (N=478) | Women (N=150) | P-value | |
|---|---|---|---|
| Preoperative | |||
| Age (years) | 56.3 ± 13.6 | 60.7 ± 13.8 | <.001 |
| Body Mass Index (kg/m 2 ) | 28.2 ± 4.4 | 27.6 ± 7.2 | 0.22 |
| Repeat Sternotomy | 57 (12%) | 16 (11%) | 0.67 |
| Prior Valve Surgery | 40 (8%) | 18 (12%) | 0.18 |
| NYHA Functional Class III or IV Symptoms | 93 (20%) | 37 (25%) | 0.18 |
| Left Ventricular Ejection Fraction (Q1, Q3) | 60.0 (53.0, 65.0) | 61.0 (60.0, 65.0) | <.001 |
| Severe Aortic Stenosis † (N=620) | 307 (65%) | 117 (79%) | 0.001 |
| Aortic Insufficiency ≥ 2+ | 228 (48%) | 44 (29%) | <.001 |
| Aortic Valve Area (cm 2 ) | 1.4 ± 1.2 | 0.9 ± 0.7 | <.001 |
| Aortic Valve Mean Gradient (mmHg) (N=551) | 34.2 ± 21.3 | 43.4 ± 22.6 | <.001 |
| Aortic Valve Peak Gradient (mmHg) (N=419) | 53.6 ± 33.6 | 68.9 ± 35.3 | <.001 |
| Intraoperative | |||
| Perfusion Time (minutes) | 114.0 (81.0, 169.0) | 83.5 (64.0, 126.0) | <.001 |
| Cross Clamp Time (minutes) | 95.0 (71.0, 134.0) | 72.5 (55.0, 104.0) | <.001 |
| Ascending Aortic Replacement (N=602) | 82 (18%) | 15 (11%) | 0.040 |
| Bioprosthetic Aortic Valve Implant | 423 (88%) | 140 (94%) | 0.390 |
| Aortic Valve Implant Size ≤ 21mm (%) | 10 (2%) | 63 (44%) | <.001 |
| Coronary Artery Bypass Grafting | 92 (19%) | 18 (12%) | 0.042 |
| Postoperative | |||
| Postoperative Blood Products | 165 (35%) | 72 (48%) | 0.003 |
| Any Complication ∗ | 171 (36%) | 52 (35%) | 0.80 |
| Post-Operative Length of Stay (Days) | 5.0 (4.0, 6.0) | 5.0 (5.0, 7.0) | 0.011 |
| Total ICU Hours | 27.4 (23.2, 46.7) | 31.2 (24.0, 69.5) | 0.014 |
| Readmission within 30 Days | 59 (12%) | 13 (9%) | 0.24 |
| 30-Day Mortality | 2 (0%) | 4 (3%) | 0.014 |
∗ Complications include re-operation for bleeding, peri-operative myocardial infarction, deep sternal infection, septicemia, post-operative stroke, prolonged ventilation > 24 hours and renal failure.
† Severe aortic stenosis was defined as maximum aortic velocity of >4 m/s or mean gradient > 40 mmHg.
The calculation of hemodynamic parameters was done according to the American Society of Echocardiography guidelines for the evaluation of prosthetic valves. All Doppler measurements were obtained with standard Doppler beam alignment and averaged over 3 cardiac cycles in patients with sinus rhythm and over 5 cardiac cycles in those with atrial fibrillation. Aortic regurgitant severity was graded as none to severe (0 to 4+) on the basis of color flow imaging. Severe aortic stenosis was defined as a maximum aortic velocity >4 m/s or a mean gradient >40 mm Hg. Transaortic pressure gradient (ΔP) was calculated from velocity (V) on the basis of the simplified Bernoulli equation: ΔPmax = 4V 2 .
Results
In the original groups, women with BAV who underwent AVR were significantly older than men (60.7 ± 13.8 vs 56.3 ± 13.6 years, p <0.001). Women presented with more aortic stenosis (78% vs 65%, p = 0.001), less aortic insufficiency (29% vs 48%, p <0.001), and smaller maximal aortic diameter (38.8 ± 9.3 vs 43.6 ± 7.9 mm, p <0.001) and therefore underwent aortic aneurysm repair less frequently (33% vs 49%, p <0.001). As expected, fusion of the left and right coronary cusps was the most common aortic valve phenotype and did not differ significantly by gender (67% in men vs 59% in women, p = 0.27). Women had higher risk for in-hospital mortality (mean Ambler score 3.4 ± 4.4 vs 2.5 ± 4.0, p = 0.015). Intraoperatively, women had shorter median perfusion (83.5 vs 114 minutes, p <0.001) and cross-clamp times (72.5 vs 95 minutes, p <0.001) and received smaller valve prostheses (p <0.001). Postoperatively, women had longer intensive care unit (ICU) median hours of stay (31.2 vs 27.4 hours, p = 0.014) and required more postoperative blood products (48% vs 35%, p = 0.003) ( Table 1 ). Kaplan-Meier curves showed similar survival outcomes by gender (p = 0.74) ( Figure 2 ). The mean follow-up time was 4.4 ± 2.6 years for men and 4.2 ± 2.8 years for women.
The bean plots indicate excellent-quality PS matching in the resulting 124 pairs ( Figure 1 ). In PS-matched analyses, women had more prolonged postoperative lengths of stay (LOS) (median 5 [interquartile range 5 to 7] vs 5 days [interquartile range 4 to 6], p = 0.027) and more postoperative blood products (48% vs 34%, p = 0.028). Short-term outcomes, including 30-day mortality (1% in men vs 2% in women, p = 0.56), were not significantly different ( Table 2 ). Mean Ambler scores were not significantly different: 3.3 ± 4.1 for women and 2.5 ± 4.0 for men (p = 0.14). Long-term overall survival was similar between genders (p = 0.60) ( Figure 2 ). The mean follow-up time was 4.3 ± 2.6 years for men and 4.2 ± 2.7 years for women. There were no aortic valve reinterventions during follow-up.
