The impact of gender-related pathophysiologic features of severe aortic stenosis on transcatheter aortic valve implantation (TAVI) outcomes remains to be determined, as does the consistency of predictors of mortality between the genders. All consecutive patients who underwent TAVI at 6 institutions were enrolled in this study and stratified according to gender. Midterm all-cause mortality was the primary end point, with events at 30 days and at midterm as secondary end points. All events were adjudicated according to Valve Academic Research Consortium definitions. Eight hundred thirty-six patients were enrolled, 464 (55.5%) of whom were female. At midterm follow-up (median 365 days, interquartile range 100 to 516) women had similar rates of all-cause mortality compared with men (18.1% vs 22.6%, p = 0.11) and similar incidence of myocardial infarction and cerebrovascular accident. Gender did not affect mortality also on multivariate analysis. Among clinical and procedural features, glomerular filtration rate <30 ml/min/1.73 m 2 (hazard ratio [HR] 2.55, 95% confidence interval [CI] 1.36 to 4.79) and systolic pulmonary arterial pressure >50 mm Hg (HR 2.26, 95% CI 1.26 to 4.02) independently predicted mortality in women, while insulin-treated diabetes (HR 3.45, 95% CI 1.47 to 8.09), previous stroke (HR 3.42, 95% CI 1.43 to 8.18), and an ejection fraction <30% (HR 3.82, 95% CI 1.41 to 10.37) were related to mortality in men. Postprocedural aortic regurgitation was independently related to midterm mortality in the 2 groups (HR 11.19, 95% CI 3.3 to 37.9). In conclusion, women and men had the same life expectancy after TAVI, but different predictors of adverse events stratified by gender were demonstrated. These findings underline the importance of a gender-tailored clinical risk assessment in TAVI patients.
Highlights
- •
To analyze gender differences in outcomes after TAVI, and to identify gender-dependent predictors of midterm all-cause mortality, a retrospective study was conducted in a large international multicenter TAVI registry.
- •
Baseline and echocardiographic features in female and male patients were different.
- •
Women had similar rates of all-cause mortality compared with men and similar incidence of myocardial infarction and cerebrovascular accident.
- •
Among clinical and procedural features, glomerular filtration rate <30 ml/min/1.73 m 2 and systolic pulmonary arterial pressure >50 mm Hg independently predicted mortality in women, while insulin-treated diabetes, previous stroke, and an ejection fraction <30% were related to mortality in men.
- •
Postprocedural aortic regurgitation was independently related to midterm mortality in the 2 groups.
Transcatheter aortic valve implantation (TAVI) represents a new therapeutic option for patients with severe, symptomatic aortic stenosis (AS) ineligible for surgical aortic valve replacement, as TAVI was proved to reduce mortality compared with a conservative strategy and was not inferior to surgery at 2-year follow-up. Several baseline and procedural factors have been identified as predictors of early and midterm events after TAVI, but the impact of gender on TAVI outcomes remains a subject of debate. Gender-related pathophysiologic differences are well known in AS: female patients develop a greater degree of left ventricular hypertrophy and higher relative wall thickness, as their smaller heart size allows a larger increase in left ventricular mass. Studies on the effect of these baseline pathophysiologic features on outcomes after surgical aortic valve replacement showed an association between female gender and risk for cardiac morbidity but not mortality after adjusting for confounding baseline variables, as well as in patients >79 years of age. In contrast, female TAVI patients appear to have more postprocedural complications and obtain higher survival rates at midterm follow-up. To analyze gender differences in outcomes after TAVI and to identify gender-dependent predictors of midterm all-cause mortality, we conducted a retrospective study in a large international multicenter TAVI registry.
Methods
The present study is reported according to the Strengthening the Reporting of Observational Studies in Epidemiology statement.
This was a retrospective multicenter study, including all consecutive patients with severe symptomatic AS, considered at prohibitive or high surgical risk, referred for TAVI from June 2007 to December 2012 at 6 institutions (University of Turin, Turin, Italy; University of Padova, Padova, Italy; Istituto Humanitas, Milan, Italy; University Medical Center Utrecht, Utretcht, The Netherlands; University of Catania, Catania, Italy; and Istituto Clinico S. Ambrogio, Milan, Italy). Patients were divided into 2 groups according to gender. Baseline, procedural, and follow-up data were retrospectively documented in local dedicated registries. Data collection was approved by the ethics committee, and patients’ informed consent for procedures was always obtained. At each center, indications for TAVI were appraised after consensus by a team of cardiac surgeons, cardiologists, and anesthetists. Details on patient selection have been previously published. Workup included thorough multimodality imaging according to each center’s protocol. Preprocedural and discharge valvular and left ventricular function were evaluated by transthoracic echocardiography, performed by local experienced echocardiographers. Coronary anatomy and hemodynamic status were assessed by coronary angiography and left-sided cardiac and, when necessary, right-sided cardiac catheterization. Valvular anatomy and annular dimension were evaluated with transthoracic and transesophageal echocardiography, contrast angiography of the aortic root, and multislice computed tomography of the thoracic aorta. The vascular access site was assessed by color Doppler sonography and multislice computed tomography with contrast angiography of the aortoiliofemoral system. Transfemoral, trans-subclavian or transapical approach was performed according to single-center experience, with implantation of self-expanding CoreValve (Medtronic, Inc, Minneapolis, Minnesota) and balloon-expandable Edwards SAPIEN or SAPIEN XT devices (Edwards Lifesciences, Irvine, California).
The primary end point was all-cause mortality at midterm follow-up. Secondary end points were appraised at 30 days (periprocedural and spontaneous myocardial infarction, cerebrovascular accident, bleeding, major vascular complications, acute kidney injury, and cardiovascular and all-cause mortality) and at midterm follow-up (myocardial infarction, cerebrovascular accident, reintervention, and cardiovascular mortality). All events were adjudicated according to the Valve Academic Research Consortium definitions.
To assess all procedural and in-hospital outcomes, institutional electronic database and individual patient charts were consulted. Short and midterm outcomes with ≥6 months of follow-up were recorded by phone, formal queries to primary physicians, and ambulatory visits.
Categorical variables were compared using Fisher’s exact tests and are presented as counts and percentages. Continuous variables, presented as mean ± SD, were compared by using analysis of variance. A stepwise logistic regression analysis including all variables listed in Tables 1 and 2 with p values <0.20 in the univariate analysis was used to determine independent predictors of midterm mortality. Cox proportional-hazards regression was performed for the multivariate analyses. The multivariate models included all the covariates listed in Tables 1 to 3 with p values <0.20 in univariate analyses. Results are presented as hazard ratios (HR) for Cox proportional-hazards analysis with 95% confidence intervals (CIs). All statistical analyses were performed with SPSS version 20 (IBM Corporation, Armonk, New York).
| Variable | Female (n = 464) | Male (n = 372) | p Value |
|---|---|---|---|
| Age (years) | 82.5 ± 6.1 | 80.3 ± 6.9 | <0.0001 |
| Body Surface Area (m 2 ) | 1.64 ± 0.27 | 1.81 ± 0.13 | <0.0001 |
| Smoke (active) | 21 (4.5%) | 43 (11.6%) | <0.0001 |
| Hypertension ∗ | 384 (82.8%) | 304 (81.7%) | 0.69 |
| Hyperlipidemia † | 147 (31.6%) | 138 (37.1%) | 0.1 |
| Diabetes mellitus | 142 (30.6%) | 110 (29.5%) | 0.75 |
| Insulin treated diabetes | 63 (13.6%) | 32 (8.6%) | 0.02 |
| Prior myocardial infarction | 75 (16.2%) | 99 (26.6%) | <0.0001 |
| Prior coronary by-pass | 33 (7.1%) | 97 (26.1%) | <0.0001 |
| Prior percutaneous coronary revascularization | 146 (31.5%) | 162 (43.5%) | 0.001 |
| Prior stroke | 50 (10.8%) | 63 (16.9%) | 0.009 |
| Peripheral artery disease | 129 (27.8%) | 155 (41.7%) | <0.0001 |
| NYHA class | 0.48 | ||
| I | 10 (2.2%) | 14 (3.8%) | |
| II | 98 (21.1%) | 70 (18.8%) | |
| III | 296 (63.8%) | 237 (63.7%) | |
| IV | 60 (12.9%) | 51 (13.7%) | |
| Glomerular Filtration Rate (ml/min/1.73 m 2 ) | 50.6 ± 22.5 | 51.9 ± 24.8 | 0.44 |
| Glomerular Filtration Rate <30 ml/min | 84 (18.1%) | 65 (17.5%) | 0.81 |
| Dialysis | 47 (10.1%) | 31 (6.7%) | 0.48 |
| Chronic Obstructive Pulmonary Disease | 98 (40.3%) | 145 (61.7%) | 0.0001 |
| Logistic EuroSCORE | 20.9 ± 12.7 | 22.9 ± 15.4 | 0.06 |
∗ Blood pressure >140/90 mm Hg or previous pharmacologic treatment.
† Total cholesterol >190 mg/dl or previous pharmacologic treatment.
| Variable | Female (n = 464) | Male (n = 372) | p Value |
|---|---|---|---|
| Ejection Fraction (%) | 54 ± 12.4 | 49.9 ± 12.7 | <0.0001 |
| End diastolic volume (cc) | 84 ± 31 | 115 ± 46 | <0.0001 |
| End systolic volume (cc) | 35 ± 21 | 65 ± 30 | <0.0001 |
| Hypertrophy index | 0.6 ± 0.1 | 0.5 ± 0.1 | 0.03 |
| Ejection Fraction <30% | 22 (4.7%) | 35 (9.4%) | 0.005 |
| Second degree diastolic dysfunction | 82 (20%) | 67 (14%) | 0.001 |
| Systolic Pulmonary Arterial Pressure (mmHg) | 42.7 ± 13.2 | 44 ± 12.6 | 0.24 |
| Systolic Pulmonary Arterial Pressure >50 (mmHg) | 79 (17%) | 69 (18.5%) | 0.56 |
| Severe mitral valve insufficiency | 15 (3.2%) | 10 (2.7%) | 0.65 |
| Aortic Valve Area (cm 2 ) | 0.68 ± 0.19 | 0.79 ± 0.2 | <0.0001 |
| Aortic Valve Area Index (cm 2 /m 2 ) | 0.34 ± 0.12 | 0.34 ± 0.17 | 0.9 |
| Mean Aortic Pressure Gradient (mmHg) | 54.3 ± 20.1 | 47.5 ± 16.1 | <0.0001 |
| Variable | Female (n = 464) | Male (n = 372) | p Value |
|---|---|---|---|
| Valve | <0.0001 | ||
| Core Valve | 263 (56.7%) | 157 (42.2%) | |
| Edwards Sapien | 201 (43.3%) | 215 (57.8%) | |
| Device size (mm) | <0.0001 | ||
| 23 | 157 (33.8%) | 14 (3.8%) | |
| 26 | 247 (53.2%) | 164 (44.1%) | |
| 29 | 58 (12.5%) | 175 (47%) | |
| 31 | 1 (0.2%) | 19 (5.1%) | |
| Approach | 0.31 | ||
| Transapical | 143 (31.2%) | 107 (28.8%) | |
| Transfemoral | 266 (57.3%) | 211 (56.7%) | |
| Trans-subclavian | 49 (9.2%) | 52 (14%) | |
| Other | 6 (1.3%) | 2 (0.5%) | |
| Fluoroscopy Time | 22.7 ± 11.6 | 22.8 ± 10.8 | 0.9 |
| Conversion to open-heart surgery | 3 (0.6%) | 3 (0.8%) | >0.9 |
| Post procedural grade III aortic regurgitation | 7 (1.5%) | 8 (2.1%) | 0.48 |
| Post procedural Ejection fraction | 54.6 ± 10.8 | 51.2 ± 12 | 0.016 |
| Post procedural Mean aortic gradient (mmHg) | 10.6 ± 6.7 | 12.7 ± 6.4 | 0.001 |
Results
Of the 836 enrolled patients, 464 (55.5%) were female and 372 (44.5%) were male. Baseline clinical, echocardiographic, and procedural characteristics are listed in Tables 1 and 2 . Women were older and had smaller body surface areas. Furthermore, smaller aortic valve areas, greater mean transvalvular aortic gradients, higher ejection fractions, smaller volumes, and higher hypertrophy indexes were observed in women. In addition, women more frequently had insulin-treated diabetes. Men were more likely to be active smokers and had more co-morbidities, as reflected by the incidence of peripheral vascular disease, stroke, previous myocardial infarction, coronary percutaneous intervention, surgical coronary revascularization, and chronic obstructive pulmonary disease. Despite these dissimilarities in co-morbidities, there was no significant difference in logistic European System for Cardiac Operative Risk Evaluation score between the 2 groups, although a trend toward higher risk was discerned in men. New York Heart Association class distribution was similar in the 2 groups. Women more frequently received CoreValve devices, were given smaller devices compared with men, and showed similar rates of postprocedural aortic regurgitation. Procedural features are summarized in Table 3 .
Thirty-day all-cause mortality ( Table 4 ) was similar in women and men (6.5 vs 5.6%, p = 0.62), as were the rates of stroke, myocardial infarction, bleeding, major vascular complications, and acute kidney injury. Women more frequently reported minor vascular complications (18.5 vs 13.7%, p = 0.04), showed a trend toward a higher bleeding rate (35.1 vs 29.8%, p = 0.06) driven by the higher rate of life-threatening bleeding (14.9 vs 10.8%, p = 0.078), and more frequently received blood transfusions (44.6% vs 31.7%, p <0.001).
