Women have traditionally been underrepresented in heart failure (HF) trials, and their baseline characteristics and outcomes after hospitalization for HF are unclear. We retrospectively analyzed the clinical characteristics and outcomes of patients according to gender in the Efficacy of Vasopressin Antagonism in Heart Failure Outcome Study with Tolvaptan (EVEREST) trial. EVEREST randomized 4,133 patients hospitalized for HF and ejection fraction of ≤40% to tolvaptan or placebo, in addition to standard therapy. The median follow-up was 9.9 months. Log-rank tests and multivariate Cox regression models were used to compare the hazards of all-cause mortality and cardiovascular mortality or HF hospitalization. Women constituted 1,058 (26%) of the study population. The baseline characteristics were similar except that the women were older with more hypertension and diabetes and less chronic renal insufficiency, previous myocardial infarction, previous coronary revascularization, and baseline defibrillator implantation (all p <0.001). The baseline use of evidence-based HF medical therapies was similar between genders (all p >0.30). Despite a high event rate, no difference was seen in all-cause mortality (men 27% vs women 24%, multivariate hazard ratio 1.04, p = 0.61) or cardiovascular mortality plus HF hospitalization (men 42% vs women 39%, multivariate hazard ratio 1.11, p = 0.10) on univariate analysis or after adjusting for baseline covariates. In conclusion, women hospitalized for worsening HF with an ejection fraction of ≤40% were older, had more hypertension, and had received fewer procedure-based interventions than men but had relatively similar HF medication usage and clinical findings. After hospitalization for HF, women have a similarly high risk of long-term HF morbidity and mortality compared with men.
Women constitute 60% of heart failure (HF) mortality, 52% of hospital discharges for HF, and 40% of patients with acute HF and a reduced left ventricular ejection fraction (LVEF). Large cohort studies and trial data have suggested the long-term prognosis with chronic HF and reduced LVEF is better for women than for men. In contrast, limited data suggest similar outcomes in men and women after hospitalization for heart failure (HHF), but the data have been conflicting. Registry data have shown similar lengths of stay and in-hospital mortality between men and women with HHF. The Euro-Heart Failure Survey II (EHFS-II) showed no difference in 1-year HF hospitalization and cardiovascular mortality stratified by gender, and the Organized Program To Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure (OPTIMIZE-HF) registry demonstrated no difference in 60- to 90-day mortality. However, a Dutch cohort study found that the after-HHF risk of death, adjusted for age and co-morbidity, was significantly less for women at 28 days, 1 year, and 5 years. No data from large, acute HF trials are available regarding long-term, gender-specific outcomes after HHF. Such information describing the effect of gender in the setting of HHF is important for understanding how to interpret the trial data for women.
Methods
The method and design of the EVEREST program ( Clinicaltrials.gov clinical trial no. NCT00071331 ) have been previously described. EVEREST was a prospective, international, multicenter, randomized, double-blind, placebo-controlled trial evaluating the short- and long-term safety and efficacy of adding tolvaptan, a vasopressin blocker, to optimal medical therapy in patients hospitalized for worsening HF with reduced LVEF. Background therapy was determined by the treating physician; however, specific guideline-based recommendations for treatment were included with the study protocol. From 2003 to 2006, 4,122 patients were enrolled at 359 centers in 20 countries and 4 geographic regions: North America, South America, Eastern Europe, and Western Europe.
The eligible patients were ≥18 years old with reduced LVEF (≤40%), had signs of volume expansion, had New York Heart Association class III/IV symptoms, and had been hospitalized for exacerbation of chronic HF within 48 hours. The exclusion criteria included surgery within 60 days of enrollment, cardiac mechanical support, biventricular pacemaker placement within 60 days before enrollment, co-morbid conditions with <6 months expected survival, acute myocardial infarction at hospitalization, hemodynamically significant uncorrected primary cardiac valve disease, refractory end-stage HF, hemofiltration or dialysis, supine systolic arterial blood pressure <90 mm Hg, serum creatinine >3.5 mg/dl, serum potassium >5.5 mEq/L, and hemoglobin <9 g/dl. A history of hypertension or hyperlipidemia was determined by the reported patient history.
The long-term primary end points were all-cause mortality (ACM) and the composite of cardiovascular mortality or hospitalization for HF (CVM + HHF). The mode of death and cardiovascular hospitalization were determined by an independent event committee for all patients during the follow-up period. The median follow-up was 9.9 months.
Post hoc analysis was performed of the data from the EVEREST trial. The baseline characteristics were compared using Student’s t test or Wilcoxon rank-sum test for continuous variables and chi-square tests for categorical variables. The univariate time-to-event comparisons between men and women were made using log-rank tests. Cox proportional hazards models were used to assess multivariate models. Proportional hazards assumptions were verified. Covariates shown to have prognostic value for acute HF were chosen from the demographic (region, age), clinical (admission systolic blood pressure, presence of atrial fibrillations, or atrial flutter on baseline electrocardiogram), and laboratory values (B-type natriuretic peptide/N-terminal pro-B-type natriuretic peptide, ejection fraction, serum sodium, blood urea nitrogen, and QRS duration). Although no variable had >9% of the data missing (maximum missing 8.13% for B-type natriuretic peptide), these variables combined created a 13.8% rate of missing data, with a greater percentage of missing data among men (14.2% vs 12.6%). PROC MI and PROC MIANALYZE in SAS were used to impute values for these covariates. Complete case Cox regression models were also fit and similar results were obtained. All analysis was run in SAS, version 9.2 (Cary, North Carolina).
Results
Of the study population, 1,058 were women (26%). Variations were present in baseline characteristics between men and women ( Table 1 ). Compared with the men, the women were older, more likely to have a history of hypertension, and less likely to have a history of a myocardial infarction. Women also had lower rates of procedure-based interventions, including coronary artery bypass grafting, percutaneous coronary intervention, and implantable cardioverter defibrillator at baseline (all p <0.001). Both women and men had similar signs and symptoms at presentation and a comparable EF (mean 28.8% vs 27%, p <0.005). The laboratory values were similar between both groups. Despite a greater median B-type natriuretic peptide level in men, the range of the midquartiles was similar. The baseline use of β blockers, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and mineralocorticoid receptor antagonists was similar between men and women (all p >0.30). Antithrombotic agents and nitrites were the only medications with statistically significant differences between men and women; however, the absolute difference in the percentage of use was 5% and 7%, respectively.
| Characteristic | Men | Women | p Value |
|---|---|---|---|
| Subjects (total n = 4,133) | 3,075 (74%) | 1,058 (26%) | — |
| Age (years) | 64.9 ± 11.9 | 68.2 ± 11.4 | <0.001 |
| Previous hospitalization for heart failure | 2,418 (79%) | 832 (79%) | 0.94 |
| Coronary artery disease | 2,206 (72%) | 705 (67%) | 0.002 |
| Previous myocardial infarction | 1,642 (53%) | 445 (42%) | <0.001 |
| Hypertension | 2,084 (68%) | 848 (80%) | <0.001 |
| Hypercholesterolemia | 1,528 (50%) | 475 (45%) | 0.01 |
| Atrial arrhythmia on admission | 894 (29%) | 301 (28%) | 0.70 |
| Diabetes mellitus | 1,155 (38%) | 443 (42%) | 0.01 |
| Previous coronary artery bypass surgery | 740 (24%) | 122 (12%) | <0.001 |
| Chronic renal insufficiency | 887 (29%) | 220 (21%) | <0.001 |
| Peripheral vascular disease | 649 (21%) | 217 (21%) | 0.68 |
| Percutaneous coronary intervention | 590 (19%) | 148 (14%) | <0.001 |
| Implantable cardioverter defibrillator | 502 (16%) | 98 (9%) | <0.001 |
| Previous stroke | 352 (12%) | 119 (11%) | 0.95 |
| Severe chronic obstructive pulmonary disease | 330 (11%) | 86 (8%) | 0.02 |
| Systolic blood pressure (mm Hg) | 119.0 ± 19.0 | 124.9 ± 20.8 | <0.001 |
| Diastolic blood pressure (mm Hg) | 72.5 ± 12.5 | 73.2 ± 13.4 | 0.14 |
| Heart rate (beats/min) | 79.6 ± 15.6 | 80.5 ± 15.7 | 0.15 |
| Dyspnea | 2,740 (91%) | 939 (91%) | 0.74 |
| Jugular venous distention ≥10 cm | 843 (28%) | 239 (23%) | 0.003 |
| Rales | 2,449 (81%) | 846 (82%) | 0.53 |
| Peripheral edema | 2,699 (89%) | 925 (90%) | 0.82 |
| Heart murmur | 1,721 (57%) | 625 (61%) | |
| Ejection fraction (%) | 27.0 ± 8.0 | 28.8 ± 8.2 | <0.001 |
| Blood urea nitrogen (mg/dl) | <0.001 | ||
| Median | 27 | 25 | |
| 25th, 75th | 20, 36 | 18, 34 | |
| Creatinine (mg/dl) | <0.001 | ||
| Median | 1.3 | 1.1 | |
| 25th, 75th | 1.1, 1.6 | 0.9, 1.4 | |
| Sodium (mEq/L) | <0.001 | ||
| Median | 140 (137, 142) | 141 | |
| 25th, 75th | 138, 143 | ||
| QRS duration (ms) | <0.001 | ||
| Median | 124 | 121 | |
| 25th, 75th | 98, 152 | 93, 146 | |
| B-type natriuretic peptide (pg/ml) | <0.001 | ||
| Median | 731 | 616.2 | |
| 25th, 75th | 304, 1,513.6 | 259, 1,451.2 | |
| N-terminal pro-B-type natriuretic peptide (pg/ml) | 0.08 | ||
| Median | 4,468.9 | 5,470 | |
| 25th, 75th | 2,122, 9,218 | 2094, 11,349 | |
| New York Heart Association class | 0.05 | ||
| III | 1,820 (61%) | 584 (57%) | |
| IV | 1,184 (39%) | 438 (43%) | |
| Medication use | |||
| Diuretics | 2,977 (97%) | 1,025 (97%) | 0.91 |
| Furosemide | 2,698 (88%) | 933 (88%) | 0.70 |
| Angiotensin-converting enzyme inhibitors/angiotensin receptor blockers | 2,578 (84%) | 901 (85%) | 0.31 |
| β-Blockers | 2,162 (70%) | 741 (70%) | 0.87 |
| Aldosterone-blocking agents | 754 (55%) | 283 (53%) | 0.37 |
| Digoxin | 1,349 (44%) | 466 (44%) | 0.92 |
| Nitrates | 520 (38%) | 243 (45%) | 0.003 |
| Calcium channel blockers | 311 (10%) | 129 (12%) | 0.06 |
| Intravenous inotropes | 34 (2%) | 7 (1%) | 0.11 |
| Hydralazine | 98 (3%) | 23 (2%) | 0.09 |
| Aspirin | 873 (64%) | 339 (63%) | 0.84 |
| Lipid-lowering agents | 537 (39%) | 204 (38%) | 0.64 |
| Antithrombotic agents | 1,174 (86%) | 432 (81%) | 0.01 |
Despite a high event rate, no difference was seen in ACM or CVM + HHF on univariate analysis ( Table 2 ). Figure 1 presents the age-adjusted Kaplan-Meier curves for the primary end points in women compared with men. After adjusting for baseline covariates, no significant difference was seen in ACM and CVM + HHF for women compared to men ( Table 2 ). Additional models adjusting for discharge medications and additional co-morbidities (diabetes, hypertension, and renal insufficiency) continued to show no difference in outcomes between men and women. A statistically significant interaction was noted between gender and atrial fibrillation or flutter on admission, such that among those with atrial arrhythmia, women had a greater hazard ratio for ACM than those without atrial arrhythmia. These results, showing no difference in outcomes, were consistent in the initial complete set analysis and the model with inputted data for missing covariates.
| Cox Regression Models | All-Cause Mortality | CVM + and HFH | ||
|---|---|---|---|---|
| HR (95% CI) | p Value | HR (95% CI) | p Value | |
| Female gender | 0.89 (0.77–1.02) | 0.11 | 0.93 (0.83–1.04) | 0.20 |
| Female gender + base ⁎ | 1.02 (0.88–1.18) | 0.78 | 1.09 (0.97–1.22) | 0.12 |
| Female gender + base ⁎ + medications † | 0.96 (0.82–1.12) | 0.59 | 1.06 (0.94–1.19) | 0.36 |
| Female gender + base ⁎ + co-morbidities ‡ | 1.04 (0.90–1.20) | 0.63 | 1.11 (0.99–1.25) | 0.07 |
| Female gender + base ⁎ + medications † + diabetes + chronic renal insufficiency | 0.97 (0.83–1.13) | 0.72 | 1.07 (0.95–1.21) | 0.24 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
