Effect of Gender on Treatment and Outcomes in Severe Aortic Stenosis




The aim of this study was to evaluate the effect of gender on operative rates and outcomes in men and women with severe aortic stenosis. An institutional echocardiographic database was used to identify all adult patients with severe aortic stenosis from 2004 through 2005. Only patients with class I indication for aortic valve replacement (AVR) during the period of follow-up were included in the study. Three hundred sixty-two patients were identified with severe aortic stenosis and class I indication for AVR (52% women). Overall operative rate for the cohort was 72%. In patients who underwent AVR, Kaplan–Meier survival rates were the same for men and women. Sixty-four percent of women versus 81% of men underwent AVR (p <0.001). After adjusting for multiple covariates, women had a 2.1-fold lower odds of undergoing AVR compared to men (p = 0.02). After matching for age and Society of Thoracic Surgery risk score, women underwent AVR at a 19% lower relative rate compared to men (p = 0.03); when stratified by gender, there was no difference in reasons for not undergoing AVR. In conclusion, despite similar outcomes after surgery, women with severe aortic stenosis are less likely than men to undergo AVR.


The effect of gender on treatment and outcomes in valvular heart disease has not been well studied. Bach et al queried a database of 5 million privately insured and a 5% sample of Medicare beneficiaries for patients with aortic valve disease and found that women were seen by specialists, underwent diagnostic tests, and underwent aortic valve replacement (AVR) at rates significantly lower than men. However, severity of valve disease in relation to diagnosis and treatment could not be assessed. Guidelines support AVR for patients with severe aortic stenosis (AS) who are symptomatic, asymptomatic and undergoing concomitant cardiac surgery, or who have an ejection fraction <50%. The purpose of this study was twofold: (1) to assess operative rates and outcomes stratified by gender and operative risk in patients with severe AS and class I indication for AVR and (2) to evaluate why medically managed patients in this cohort did not undergo AVR.


Methods


This study was approved by Massachusetts General Hospital’s internal review board. Because of the retrospective nature of the study, the need for patient consent was waived. The Massachusetts General Hospital echocardiographic database was used to identify all adult patients with severe AS from 2004 through 2005. Severe AS was defined as a mean gradient >40 mm Hg and an aortic valve area (AVA) <1 cm 2 . Four hundred sixty-seven patients were identified and comprehensive chart review was performed. Fifty-five patients (30 men, 25 women) were excluded because of incomplete medical records (47 patients), presence of prosthetic valve AS (4 patients), and an inability to determine whether the patient was alive or dead with the Social Security Death Index because they were traveling from a different country (4 patients). Of the remaining 412 patients, 368 patients were identified as having class I indication for AVR during the period of follow-up. Six patients (3 women, 3 men) were excluded because of moderate to severe mitral stenosis (mean transmitral gradient ≥5 mm Hg), leaving 362 patients for subsequent analysis. Baseline characteristics including time of initial diagnosis of severe AS, symptoms, co-morbidities, and echocardiographic data were collected. Follow-up period was defined as the date of development of a class I indication for AVR to the date that the Social Security Death Index was queried for each patient. If the patient was not operated on during the period of follow-up, the reason for no operation was assessed and assigned to 1 of 4 categories based on documentation in the medical record: advanced age (if age was documented as the major reason for the patient or the treating physician), co-morbidities other than advanced age, patient declined, or other.


Clinical variables were assessed at the time the patient developed a class I indication for AVR at our institution. A history of chest pain, dyspnea on exertion, and syncope were identified from medical records. Patients were considered asymptomatic if none of these symptoms were elicited in the history. Heart failure severity was assigned based on the New York Heart Association class index. Hypertension was defined as a blood pressure >140/90 mm Hg or need for antihypertensive medication. Chronic renal insufficiency was defined as a creatinine level ≥1.5 mg/dl. Diabetes mellitus was defined as a fasting blood glucose level >125 mg/dL or need for antidiabetic agents. Previous myocardial infarction was defined as a history of biomarker-positive infarction or evidence of scar on echocardiogram. Coronary artery disease was defined as a history of myocardial infarction, coronary artery bypass grafting, percutaneous coronary intervention, or >70% lesion (>50% for lesion in the left main coronary artery). Lung disease was considered present if a patient required daily inhalers. Peripheral vascular disease was defined as carotid stenosis >70% or requiring surgery, presence of an abdominal aortic aneurysm, history of peripheral vascular bypass surgery or stenting, or history of significant claudication. A patient was identified as having cancer if the cancer was actively being treated at the time of diagnosis of severe AS or at any time during the follow-up period. Data were collected from 2-dimensional transthoracic echocardiographic examinations or transesophageal echocardiogram. Peak and mean transaortic gradients were identified from continuous-wave Doppler measurements in ≥2 views. AVA was calculated using the continuity equation. AVA index was calculated as AVA/body surface area. Relative wall thickness was defined as (2 × posterior wall thickness)/left ventricular end-diastolic diameter.


Operative mortality risk for AVR was assessed by the online Society of Thoracic Surgery (STS) score calculator because of its greater accuracy in assessing risk of AVR compared to the European System for Cardiac Operative Risk Evaluation score. Patients requiring double valve surgery or ascending aortic operation (n = 31, 22 men and 9 women) were excluded from analyses using the STS score because it does not support accurate mortality estimates in more complex surgeries. End points of the study were performance of AVR and all-cause mortality. All-cause mortality was determined by querying the Social Security Death Index.


All statistical analyses were performed using STATA 8.0 (STATA Corporation, College Station, Texas) and graphics were created with GraphPad Prism 5.0 (GraphPad Software, San Diego, California). Normality of data was assessed using the Shapiro–Wilk test. Continuous variables are expressed as mean ± SD or median (interquartile range). Group baseline characteristics were compared with Student’s t test or Mann–Whitney U statistic for continuous variables or Pearson chi-square or Fisher’s exact test for categorical variables, as appropriate. Multivariate logistic regression was performed to assess the probability of undergoing AVR. Survival analysis was performed using Kaplan–Meier product-limit analysis with log-rank comparison testing and multivariate Cox regression models. Multivariate analyses were performed by incorporating variables from univariate analysis that achieved a p value ≤0.05 and using a stepwise backward elimination protocol. Matching analysis was performed between women and men using the nearest-neighbor approach and requiring an age difference ≤1 year and an STS score difference ≤3%. A p value ≤0.05 was considered statistically significant.




Results


We identified 362 patients (52% women) with severe AS and class I indication for AVR. Baseline characteristics are presented in Tables 1 and 2 . Women were older than men and were less likely to have coronary disease or to have undergone coronary revascularization ( Table 1 ). Women had increased left ventricular ejection fraction (LVEF) and relative wall thickness at time of diagnosis indicating a greater relative degree of hypertrophy compared to men. There was no significant difference in the percentage of women versus men with LVEF <50%. Transaortic gradients, AVA index, degree of mitral regurgitation, and pulmonary artery systolic pressure were not significantly different between men and women.



Table 1

Baseline clinical characteristics































































































































Women Men p Value
(n = 190) (n = 172)
Age (years) 78 ± 10 72 ± 11 <0.001
Body surface area (m 2 ) 1.7 ± 0.3 2.0 ± 0.3 <0.001
Symptoms
Asymptomatic 7% (14) 6% (10) 0.55
Chest pain 23% (44) 35% (61) 0.01
Dyspnea on exertion 79% (151) 78% (135) 0.82
Syncope 13% (25) 9% (16) 0.25
New York Heart Association class
I 45% (85) 50% (86) 0.32
II 34% (65) 35% (60) 0.89
III 16% (30) 12% (20) 0.25
IV 5% (10) 3% (6) 0.41
Shock 2% (4) 2% (3) 0.80
Coronary artery disease 34% (64) 52% (90) <0.001
Previous myocardial infarction 13% (24) 19% (33) 0.09
Previous percutaneous intervention 8% (15) 16% (27) 0.02
Previous coronary artery bypass grafting 2% (4) 14% (24) <0.001
Diabetes mellitus 20% (38) 23% (39) 0.54
Hypertension 85% (161) 88% (151) 0.40
Chronic renal insufficiency 17% (32) 20% (35) 0.39
Lung disease 16% (31) 16% (27) 0.87
Peripheral vascular disease 8% (16) 14% (24) 0.10
Cancer 5% (9) 6% (11) 0.49

Measurements are presented as mean ± SD or percentage of patients (number).


Table 2

Baseline echocardiographic characteristics




























































































Women Men p Value
(n = 190) (n = 172)
Ejection fraction (%) 69 (60–75) 62 (54–70) <0.001
Ejection fraction <50% 11% (21) 17% (30) 0.08
Ejection fraction ≤35% 4% (7) 6% (11) 0.24
Left ventricular outflow tract (cm) 1.9 ± 0.2 2.1 ± 0.2 <0.001
End-diastolic dimension (mm) 42 ± 5 49 ± 7 <0.001
Posterior wall thickness (mm) 11.9 ± 2.0 12.4 ± 2.0 0.02
Septal wall thickness (mm) 12.7 ± 2.0 13.1 ± 1.9 0.09
Relative wall thickness 0.58 ± 0.13 0.52 ± 0.11 <0.001
Aortic valve area (cm 2 ) 0.62 ± 0.15 0.70 ± 0.15 <0.001
Aortic valve area index 0.36 ± 0.09 0.35 ± 0.08 0.26
Aortic valve peak gradient (mm Hg) 91 ± 24 88 ± 22 0.28
Aortic valve mean gradient (mm Hg) 56 ± 15 54 ± 13 0.22
Mitral regurgitation grade 3 or 4 15% (29) 10% (18) 0.18
Aortic insufficiency grade 3 or 4 6% (11) 6% (11) 0.81
Bicuspid aortic valve 8% (15) 16% (28) 0.01
Pulmonary artery systolic pressure (mmHg) 47 ± 14 45 ± 14 0.31

Measurements are presented as mean ± SD, median (interquartile range), or percentage of patients (number).


Operative rate for the entire cohort was 72% (261 of 362). Mean period of follow-up for the cohort (n = 362) was 4.2 years. In women indications for AVR included (many patients had multiple indications) abnormal LVEF <50% in 11% (21 of 190), angina in 23% (44 of 190), dyspnea on exertion in 79% (151 of 190), syncope in 13% (25 of 190), and symptoms of congestive heart failure in 55% (105 of 190). In men indications for AVR included LVEF <50% in 17% (30 of 172), angina in 35% (61 of 172), dyspnea on exertion in 78% (135 of 172), syncope in 9% (16 of 172), symptoms of congestive heart failure in 50% (86 of 172), and asymptomatic but undergoing ascending aortic surgery for aortic aneurysm in 2% (4 of 172, all 4 had bicuspid aortic valves). Compared to women men were more likely to have angina (p = 0.01) and concomitant cardiac surgery for aortic aneurysm (p = 0.05, Fisher’s exact test) as an indication for AVR.


Overall, women had decreased survival compared to men (Cox hazard ratio 1.4, p = 0.04; Figure 1 ). As expected, patients who underwent AVR had improved survival (p <0.0001; Figure 1 ). In patients who underwent AVR mortality outcomes were similar for men and women ( Figure 1 ). Sixty-four percent of women (121 of 190) versus 81% of men (140 of 172) underwent AVR (p <0.001, Pearson chi-square test). Because women were older than men at baseline, regression analysis stratified by age was performed. Three age ranges were defined (<65, 65 to 80, >80 years) and operative rates within each group were compared for women and men. Multivariate regression analysis for the entire cohort identified gender as a significant predictor of undergoing AVR, with a 2.0-fold lower odds for women undergoing operation (p = 0.01) after adjusting for the 3 age groups ( Figure 2 ).




Figure 1


Kaplan–Meier survival curves in 362 patients with severe aortic stenosis and class I indication for aortic valve replacement stratified by male (black line) and female (gray line) gender (A) and by gender and operative status (B) . (A) Women exhibited increased mortality (Cox hazard ratio [HR] 1.4, p = 0.04). (B) Survival was significantly improved in men (n = 140) (solid black line) and women (n = 121) (solid gray line) undergoing aortic valve replacement versus men (n = 32) (dashed black line) and women (n = 69) (dashed gray line) not undergoing the operation (p <0.0001). Survival outcomes were similar for men and women who underwent aortic valve replacement (p = 0.89).



Figure 2


Comparison of operative rates stratified by gender and age in 362 patients with severe aortic stenosis and class I indication for aortic valve replacement showed. Comparison in women (gray bars) and men (black bars) for each age group (<65, 65 to 80, >80 years) was performed using Pearson chi-square analysis. Multivariate logistic regression for the entire cohort identified gender as a significant predictor of undergoing aortic valve replacement, with a 2.0-fold lower odds for women undergoing operation (odds ratio 0.51, p = 0.01) after adjusting for the 3 age groups (odds ratio 0.28, p <0.001; p = 0.87 for interaction).


Predictors of all-cause mortality are listed in Table 3 . In addition to expected variables, relative wall thickness was an independent predictor of increased mortality. Women had a greater relative wall thickness compared to men (0.58 ± 0.13 vs 0.52 ± 0.11, p <0.001), indicating a greater relative degree of hypertrophy in women. For every 10% increment in relative wall thickness, the adjusted Cox regression mortality hazard ratio was 1.29 (95% confidence interval 1.12 to 1.50, p = 0.001).



Table 3

Predictors of all-cause mortality








































Variable Cox Hazard Ratio (95% confidence interval) p Value
Age (for every decade) 1.42 (1.13–1.78) 0.002
New York Heart Association class (for every class increase) 1.37 (1.10–1.70) 0.005
Previous myocardial infarction 2.13 (1.32–3.44) 0.002
Lung disease 2.41 (1.52–3.83) <0.001
Creatinine (every 1.0-mg/dl increase) 1.21 (1.05–1.39) 0.008
Left ventricular ejection fraction (every 5% increase) 0.89 (0.83–0.96) 0.002
Relative wall thickness (every 10% increase) 1.29 (1.12–1.50) 0.001
Undergoing aortic valve replacement 0.14 (0.09–0.22) <0.001


We found that 83% of men (143 of 172) with a class I indication for AVR were referred for evaluation by a cardiac surgeon versus 68% of women (130 of 190, p = 0.001, Pearson chi-square test). Of those referred for cardiac surgical evaluation, 98% of men (140 of 143) underwent AVR versus 93% of women (121 of 130, p = 0.07). Independent predictors of undergoing AVR are listed in Table 4 . Age, New York Heart Association class, and presence of chest pain were significant predictors of undergoing AVR. However, even after adjusting for these and other variables, gender remained an independent predictor of undergoing AVR, with men having a 2.1-fold increased odds of being operated on compared to women (p = 0.02; Table 4 ). When STS scores were added to the logistic regression model (odds ratio 0.90 for every 1% increase in score, p = 0.026), age was no longer a significant predictor of AVR. However, gender remained an independent predictor, with men having a 2.7-fold increased odds of operation (p = 0.001). Stratification of patients by STS score and gender revealed that most patients had an operative mortality estimated at <5% ( Figure 3 ). In this cohort of patients with the lowest risk, women were 12% less likely than men to undergo AVR (p = 0.04; Figure 3 ).



Table 4

Independent predictors of undergoing aortic valve replacement








































Variable Odds Ratio (95% confidence interval) p Value
Age 0.87 (0.83–0.91) <0.001
Chest pain 2.34 (1.11–4.94) 0.03
New York Heart Association class (every class decrease) 1.66 (1.21–2.28) 0.002
Absence of previous myocardial infarction 2.85 (1.31–6.19) 0.008
Absence of chronic renal insufficiency 2.48 (1.25–4.90) 0.010
Absence of cancer 7.88 (2.35–26.5) 0.001
Aortic valve mean gradient (every 10-mm Hg increase) 1.34 (1.05–1.71) 0.018
Male gender 2.08 (1.13–3.81) 0.018



Figure 3


Society of Thoracic Surgery operative risk scores were assessed for each patient. Operative risk was stratified into 3 groups (mortality <5%, 5% to 10%, >10%). Comparison of operative rates in women (gray bars) and men (black bars) for each risk group was performed using Pearson chi-square analysis, with a significant difference observed in those with Society of Thoracic Surgery scores <5%. Multivariate logistic regression model revealed a 1.9-fold lower odds for women undergoing operation (odds ratio 0.52, p = 0.01) after adjusting for the 3 risk groups (odds ratio 0.37, p <0.001; p = 0.83 for interaction).


A nearest-neighbor matching analysis was performed for men and women based on age (requiring ≤1-year difference) and STS score (≤3% difference) resulting in 89 matched pairs identified. Table 5 presents baseline characteristics of the matched men and women. In this cohort of matched patients, women were less likely to undergo AVR, with a 19% relative decrease in operative rates compared to men (p = 0.03; Figure 4 ).


Dec 22, 2016 | Posted by in CARDIOLOGY | Comments Off on Effect of Gender on Treatment and Outcomes in Severe Aortic Stenosis

Full access? Get Clinical Tree

Get Clinical Tree app for offline access