Sex-Specific Differences at Presentation and Outcomes Among Patients Undergoing Transcatheter Aortic Valve Replacement

Context Whether female sex is associated with an increased risk for adverse events among patients having transcatheter aortic valve replacement (TAVR) is unclear. Contribution In this secondary analysis of data from the Placement of Aortic Transcatheter Valves trial, women who had TAVR had more postprocedural major bleeding and vascular complications and improved 1-year survival than men. Caution Other factors, such as paravalvular leak or patientprosthesis mismatch, may have confounded the results. Implication The sex-specific risk associated with TAVR is the reverse of that for surgical aortic valve replacement, for which female sex is associated with an increased risk for adverse events. During the past 50 years, the standard of care for patients with severe, symptomatic aortic stenosis (AS) has been surgical aortic valve replacement (SAVR), which has been shown to prolong survival and alleviate symptoms compared with medical therapy alone (1, 2). However, at least one third of patients with AS are deemed to be high-risk or inoperable for SAVR due to comorbidities or unfavorable anatomy (3). For these patients, transcatheter aortic valve replacement (TAVR) has emerged as an alternative (1, 2, 411). The PARTNER (Placement of Aortic Transcatheter Valves) trial was the first to show the safety and efficacy of TAVR compared with accepted standard therapies in high-risk and inoperable patients with AS (49). The results of this trial ultimately led to the approval of the SAPIEN transcatheter valve (Edwards Lifesciences) by the U.S. Food and Drug Administration. Since the PARTNER trial, many other trials and international registries have shown the benefits of TAVR in higher-risk patients (1012). Female sex has been shown to be associated with increased risk for adverse events after SAVR (1317) but has been shown to be a potentially favorable characteristic for patients having TAVR (1823). However, prior trials examining sex-based differences in patients having TAVR have been limited in size or have involved only single- or dual-center experiences. Furthermore, not all studies agree that TAVR may be more beneficial in women, with some finding no difference in outcomes or increased adverse events in women (2428). Given the conflicting data in the literature, we sought to perform a comprehensive analysis of sex-based differences in patients having TAVR by using the clinical trial structure of the PARTNER experience (encompassing all TAVR patients treated in not only the randomized PARTNER trials but also the randomized and nonrandomized continued access registries). We report the baseline demographic characteristics and core laboratoryassessed echocardiographic parameters of women and men treated with TAVR, as well as adjudicated 30-day and 1-year outcomes stratified by sex. Methods Study Design and Patients The PARTNER trial incorporated 2 parallel prospective, multicenter, randomized, active treatmentcontrolled clinical trials. Patients were divided into 2 cohorts. Cohort A comprised patients who were considered to be high-risk candidates for surgery, as defined by a Society of Thoracic Surgeons Predicted Risk of Mortality (STS-PROM) score of at least 8% (on a scale of 0% to 100%, with higher scores indicating greater risk for death within 30 days after surgery) (13) and/or by the presence of coexisting conditions associated with a 15% or greater risk for death at 30 days. Cohort B comprised those who were not considered to be suitable candidates for surgery because they had coexisting conditions associated with a 50% or greater probability of death or serious irreversible morbidity after surgery based on a multidisciplinary heart team assessment. In cohort A, patients were randomly assigned to TAVR or SAVR; in cohort B, patients were randomly assigned to TAVR or medical therapy. Patients in cohort A were enrolled from 11 May 2007 to 28 August 2009, and those in cohort B were enrolled from 11 May 2007 to 16 March 2009. After enrollment in the randomized PARTNER trial, sites were able to continue to consecutively enroll patients in a continued access registry using the same inclusion and exclusion criteria. The purpose of the continued access registry was to accumulate greater real-world experience with regard to the safety and efficacy of TAVR while approval by the U.S. Food and Drug Administration was pending. At the time of initiation of the registry, TAVR could not be performed in the United States unless the patient was formally enrolled in the registry or a comparable clinical trial. Because enrollment in the inoperable trial (cohort B) was completed 6 months before enrollment in the high-risk trial (cohort A) was completed, investigators decided to extend randomization of cohort B into the continued access registry to prevent enrollment bias in the ongoing high-risk trial. Once enrollment in the high-risk trial was completed, both cohorts were enrolled in a nonrandomized continued access registry. Enrollment in this registry concluded on 9 January 2012. The same data collection procedures and follow-up used for the randomized PARTNER trial were used for the continued access registry. The same clinical events committee, adjudication process, and core laboratory interpretation of echocardiographic data were also used for both the randomized and continued access portions of the PARTNER trial. The institutional review board at each site approved the study, and all patients provided written informed consent. End Points The 30-day and 1-year frequencies of all-cause mortality, cardiovascular mortality, rehospitalization, stroke, major vascular complications, major bleeding, myocardial infarction, acute kidney injury, and need for a permanent pacemaker were recorded per the PARTNER trial protocol (4, 5). Major vascular complications were defined as any thoracic aortic dissection; access site or access-related vascular injury leading to death, need for significant blood transfusions, or unplanned percutaneous or surgical intervention; or distal embolization from a vascular source. Major bleeding was defined as a bleeding event that caused death or permanent disability, caused or prolonged a hospitalization, or required an open or endovascular procedure or a transfusion. All adverse events were fully adjudicated by an independent clinical events committee. Echocardiograms were obtained at baseline; at discharge or 7 days, whichever came first; and at 30 days, 6 months, and 1 year after the procedure. All echocardiograms were analyzed at an independent core laboratory that followed the American Society of Echocardiography standards for echocardiography core laboratories (29). Statistical Analysis The current analysis was an as-treated analysis, which included all TAVR patients from cohort A, cohort B, and the continued access registry. Patients were stratified on the basis of sex. Continuous variables are presented as means and SDs and were compared using the t test; categorical variables were compared by using the chi-square test or the Fisher exact test, as appropriate. KaplanMeier techniques were used for outcomes that contained all-cause death. All nonmortality outcomes were modeled by using a proportional subdistribution hazards regression to account for the competing risk for death; the corresponding hazard ratios (HRs) and P values are reported from this model. For each sex, the cumulative incidence for 30-day and 1-year outcomes is summarized in the presence of all-cause death. A 2-sided level of 0.05 was used to indicate statistical significance. To assess the association between sex and all-cause mortality at 1 year, Cox multivariable regression analyses were performed. The multivariable model was created by forcing sex into a stepwise model selection procedure with clinically relevant candidate variables using an entry/stay criterion of a P value less than 0.10. Variables of clinical interest were included if the variable was significant (P< 0.10) in the univariate Cox model. A supplemental model included moderate or severe paravalvular regurgitation at discharge and any major vascular complication as a time-dependent covariate along with variables selected in the final model. Using the same variables as those selected for the multivariable model for mortality, we tested for an association between sex and rehospitalization while accounting for the competing risk for death. We also tested for an interaction between sex and treatment approach (transfemoral vs. transapical TAVR) on 1-year all-cause mortality. Analyses were done using the FREQ, TTEST, UNIVARIATE, PHREG, and LIFETEST procedures in SAS, version 9.2 (SAS Institute). Clinical Trial Registration The PARTNER trial was registered at ClinicalTrials.gov on 14 September 2007, slightly more than 4 months after the first patient was enrolled on 11 May 2007. Only 41 (1.3%) of the 3216 total patients were enrolled before registration of the trial, and no interval analysis was conducted between enrollment of the first patient and registration of the trial. Role of the Funding Source The PARTNER trial was funded by Edwards Lifesciences, and the protocol was developed collaboratively by the sponsor and the steering committee. The funding source had no involvement in the design, analysis, or interpretation of this substudy or the decision to publish the manuscript. Results Baseline Characteristics All 2559 patients who had TAVR in the PARTNER trial were analyzed on the basis of their sex (Table 1). Of these, 2000 were in the nonrandomized continued access registry, 40 were in the randomized continued access registry, and 519 were in the randomized trial. Five patients in cohort A, 1 in cohort B, and none in the continued access registries were lost to follow-up. A total of 47.7% of all patients were women, and 52.3% were men. Table 1. Study Population The baseline presenting characteristics varied between sexes in terms of demographic variables and comorbidities