Sex Differences in the Outcomes of Elderly Patients With Heart Failure With Preserved Ejection Fraction 


 Background: It has been shown the impacts of sex on patients' outcomes with preserved ejection fraction (HFpEF), but little is known about the impacts of sex on elderly patients with HFpEF.Methods: A secondary analysis was conducted to evaluate the impacts of sex on outcomes of patients who were ≥70 years of age with HFpEF from the Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist Trial (TOPCAT). The primary outcome was composed of cardiovascular (CV) mortality, HF hospitalization. Secondary outcomes included all-cause mortality and all-cause hospitalization. Cox regression models were used to determine sex differences in outcomes.Results: A total of 1619 patients were included: 898 (55.5%) women and 721 (44.5%) men. Their age ranged from 70 to 94 years, similar between women and men. Women had fewer comorbidities than men. The rate of primary outcome was lower in women than in men (18.9% vs. 28.1%, p=0.002), including CV mortality (10.6% vs. 15.4%, p=0.039) and HF hospitalization (13.5% vs. 19.0%, p=0.033). After adjustment for baseline characteristic, the Cox regression analysis showed that woman was a protective factor for CV mortality (hazard ratio [HR] 0.53, 95% confidence interval [CI] 0.40-0.73), HF hospitalization (HR 0.71, 95% CI 0.55-0.93) and all-cause mortality (HR 0.59, 95% CI 0.47-0.75). Although a significant reduction in all-cause mortality associated with spironolactone in women was observed even after adjustment (HR: 0.68; 95% CI: 0.48-0.96; p=0.028), there is not a significant multivariate sex-treatment interaction (p interaction=0.190).Conclusion: Among elderly patients with HFpEF, women had fewer comorbidities and better outcomes. Clinical trial registration: NCT00094302 (TOPCAT). Registered 15 October 2004, https://www.clinicaltrials.gov/ct2/show/NCT00094302


Background
The risk of heart failure with preserved ejection fraction (HFpEF) increases with advancing age. Patients with HFpEF are older than HF patients with reduced ejection fraction (HFrEF); With aging, the incidence and prevalence of HFpEF increase more rapidly [1]. These elderly patients are under-represented in randomized clinical trials, and the clinical features, event rates, response for treatments could be different from young patients.
Indeed, sex differences existed in almost every facet of HF (both HFrEF and HFpEF), including baseline characteristics, risk factors, pathophysiology, response to drugs, and different outcomes [2,3]. Community-based studies indicated that women are substantially different from men concerning clinical features and event rates [4,5]. The Irbesartan in heart failure with Preserved Ejection Fraction (I-PRESERVE) trial, including about 2500 women patients, showed that women had a lower risk of mortality or hospitalization for both cardiovascular and non-cardiovascular events, suggesting women had better overall prognosis [6]. The large randomized placebo-controlled trial TOPCAT showed no signi cant interactions between spironolactone arm and sex concerning the primary outcome in a pre-speci ed subgroup analysis of the full TOPCAT cohort [7]. In contrast, the secondary analysis restricted to the Americas observed a signi cant reduction in all-cause mortality associated with spironolactone in women but not in men [8]. These ndings suggested a positive effect of spironolactone in women with HFpEF but not in men.
However, information regarding the impacts of sex on outcomes and response to spironolactone in elderly patients with HFpEF is limited. In the present study, we did a post-hoc, exploratory subgroup analysis in elderly patients from the TOPCAT study to discuss differences in baseline characteristics, outcomes, and response to spironolactone between elderly women and men with HFpEF.

Study design and patients
In the present study, we used TOPCAT clinical data obtained from the National Heart, Lung, and Blood Institute's Biological Specimen and Data Repository Information Coordinating Center (BioLINCC, Calverton, Maryland). Patients were eligible if they were diagnosed with symptomatic HF and LVEF ≥ 45% combined with either a hospitalization for HF within 12 months before inclusion or an elevated natriuretic peptide level (brain natriuretic peptide [BNP] ≥100 pg/mL or N-terminal pro-BNP [NT-proBNP] ≥ 360 pg/mL) within 60 days before inclusion. Patients had to be aged ≥ 50 years, have controlled systolic blood pressure <140 mmHg (or ≤ 160 mmHg if the patient was taking three or more medications to control blood pressure), and a serum potassium level < 5.0 mmol/L. The main exclusion criteria were life expectancy < 3 years, estimated glomerular ltration rate (eGFR) < 30 ml/min/1.73 m 2 body surface area or serum creatinine ≥ 2.5 mg/dL. The detailed inclusion and exclusion criteria were described in the main study publication [7,9]. Patients were randomized in a double-blind fashion to receive either spironolactone or placebo therapy. For this study, 1619 elderly patients (age ≥ 70 years) [10,11] were selected from the TOPCAT trial, and a secondary analysis was conducted.

De nitions and endpoints
The mean follow-up time was 3.3 years. The primary outcomes were composed of cardiovascular (CV) mortality, HF hospitalization. Secondary outcomes for variables. The presence of sex differences in outcomes was compared between men and women within the placebo arm, spironolactone arm, and all patients.
All categorical variables were compared using the Fisher's exact test or χ 2 test, and continuous variables were compared using the t-test or the Mann-Whitney U test, as appropriate. Time-to-event curves were obtained using the Kaplan-Meier method. Univariate and multivariate Cox proportional hazards model was used to explore the associations between sex and the outcomes. Effects of intervention versus placebo on the outcomes were analyzed by sex. Adjustment variables included race, NYHA class, myocardial infarction (MI), percutaneous coronary intervention (PCI), coronary artery bypass grafting (CABG), atrial brillation (AF), hypertension, dyslipidemia, chronic obstructive pulmonary disease (COPD), systolic blood pressure (SBP), diastolic blood pressure (DBP), heart rate, body mass index (BMI), baseline estimated glomerular ltration rate (eGFR) and baseline potassium level. Statistical analyses were performed using Stata/S.E. version 15.1 software (Stata Corp., College Station, Texas). Sex-treatment interaction was analyzed using Empower Stats. A p-value <0.05 was considered signi cant throughout.

Results
Characteristics of the patients according to sex The overall cohort (n = 1619) strati ed by sex were described in Table 1. Of the 1619 elderly patients, 898(55.5%) were women, and 721(44.5%) were men.
Their age ranged from 70 to 94 years, and the mean age was similar between women and men. Women from the Americas (including the United States, Canada, Brazil, Argentina) or Russia/Georgia were 533(59.4%), 365(40.6%), respectively.
The baseline characteristics of each group were listed in Table 1 were lower in women than in men. Moreover, we observed the elderly women with HFpEF had higher LVEF, a higher rate of NYHA functional classes III-IV, and lower Kansas City Cardiomyopathy Questionnaire (KCCQ) scores than men in the present study.
Regarding the use of medications, there were no signi cant differences between men and women in the use of angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor blockers (ARB), beta-blockers, calcium channel blockers (CCB), or diuretics. Men were signi cantly more likely to take statin, warfarin, any hypoglycemic, other anti-hypertensive, and CV medications. Moreover, no matter in the placebo arm or the spironolactone arm, differences between women and men are the same as the whole cohort.

Differences in outcomes between women and men
Rates of primary and secondary outcomes strati ed by sex for the placebo arm and spironolactone arm were summarized in Table 2. In all patients, the rates of all outcomes were signi cantly lower in women than in men. In the placebo arm, women had lower rate of primary outcome (18.9% vs. 28.1%, p = 0.002), CV mortality (10.6% vs. 15.4%, p = 0.039), HF hospitalization (13.5% vs. 19.0%, p = 0.033), and all-cause mortality (31.5% vs. 50.4%, p < 0.001) than men. The rate of all-cause hospitalization was numerically lower in women than in men, but there were no statistically signi cant differences. In the spironolactone arm, rates of primary outcome (17.9% vs. 26.8%, p = 0.002), CV mortality (7.3% vs. 14.2%, p = 0.001), all-cause mortality (13.0% vs. 25.7%, p < 0.001), all-cause hospitalization (45.5% vs 53.4%, p = 0.026) in women were signi cantly lower than in men. HF hospitalization rates were numerically lower in women than men, but there was no statistically signi cant difference. Kaplan-Meier curves for primary and secondary outcomes strati ed by sex summarized in Figure 1 and Figure 2. Sex-speci c univariate analysis showed that women had lower rates of all outcomes in all patients and the placebo arm. In the spironolactone arm, there were no signi cant differences in HF hospitalization (hazard ratios [HR] 0.73, 95% CI: 0.51-1.04, p = 0.083) and all-cause hospitalization (HR 0.71, 95% CI: 0.50-1.01, p = 0.058) between women and men. sex-speci c multivariate HR in the placebo arm and spironolactone arms for all outcomes adjusted for region, NYHA class, MI, PCI, CABG, AF, hypertension, dyslipidemia, COPD, SBP, DBP, BR, BMI, baseline eGFR, and baseline potassium levels were summarized detailly in Supplemental Table 1 and Figure 3. In the whole cohort, women had a signi cantly lower risk of the primary outcome, CV mortality, HF hospitalization, and all-cause mortality after adjusting the covariates. Importantly, both in the placebo arm and the spironolactone arm, there was a reduced likelihood of the primary outcome, CV mortality, and all-cause mortality in women but not in men.

Treatment effect between women and men
Univariate HRs for all outcomes were summarized in Table 3. In women patients, the primary outcome occurred in 84 patients (10.4%) taking placebo and 81 patients (10.0 %) taking spironolactone, with a corresponding HR of 0.95 (95% CI: 0.70 to 1.29). Rates of CV mortality, HF hospitalization, all-cause mortality, and all-cause hospitalization were also numerically lower in patients treated with spironolactone but did not reach statistical signi cance (p 0.05 for all outcomes). The spironolactone treatment effect was similar in men patients.
Multivariate HRs and interaction terms between sex and treatment response for outcomes were summarized in Supplemental Figure 1 and Table 3. There was no signi cant reduction in primary outcome associated with spironolactone in women (HR 0.91, 95% CI: 0.67-1.25, p=0.580) and men (HR 0.88, 95% CI: 0.66-1.17, p=0.377). Rates of CV mortality, HF hospitalization, and all-cause hospitalization all had no signi cant difference between placebo and spironolactone arm both in women and men (p >0.05). Although there was a decreased rate of all-cause mortality r in women treated with spironolactone (10.0% vs. 7.3%, HR 0.68, 95% CI: 0.48-0.96) but not in men, the sex-treatment interaction was not signi cant (p interaction=0.190).
Patients older than 70 years of age from the TOPCAT study were included in the present study, and sex differences in baseline characteristics and sex-speci c advantages in outcomes were found: (1) elderly women with HFpEF had fewer comorbidities and better health conditions than men; (2) Elderly women had a lower risk of CV mortality than men either receiving placebo or spironolactone; (3) There was no signi cant reduction in primary outcome associated with spironolactone in elderly women and men, but elderly women taking spironolactone had lower all-cause mortality than those taking placebo.
Most patients with HF are above 65 years of age, and in many contemporary studies, a signi cant proportion of patients are 70 years of age or older. A prospective study enrolled 1203 patients with HFpEF from 11 Asian regions grouped patients into very young (<55 years), young (55-64 years), older (65-74 years), and elderly (≥75 years) [12], which showed that patients from different age groups have different clinical characteristics and outcomes. However, elderly patients are under-represented in randomized clinical trials. Speci cally, prior reports from HF trials that examined sex differences included age ≥60 years old in the I-PRESERVE (Irbesartan in heart failure with Preserved ejection fraction study [6], age ≥21 years old in the DIG-PEF [13], and age ≥ 50 years old in the TOPCAT-Americas study [14]. Moreover, there were prominent sex differences in baseline characteristics in patients with HFpEF, but few studies focused on the sex differences of elderly patients with HFpEF. Therefore, in the present study, we included patients from the TOPCAT trial restricted to the elderly. We observed that among the elderly patients, women with HFpEF were more likely to have higher BMI, chronic kidney disease, and hypertension than men but less likely to have coronary artery diseases, tobacco use, AF, and COPD. These ndings were in agreement with data derived from the I-PRESERVE study and the TOPCAT study. Another meta-analysis[8] including 4458 women and 4010 men with HFpEF enrolled in CHARM-Preserved (Candesartan in Heart failure: Assessment of Reduction in Mortality and morbidity) (EF≥45%), I-Preserve, and TOPCAT-Americas indicated that women were older and more often have obesity and hypertensive but less likely to have coronary artery disease or atrial brillation. These ndings suggested elderly women with HFpEF had fewer comorbidities than men.
Moreover, we observed elderly women with HFpEF had higher LVEF, a higher rate of NYHA functional classes III-IV, and lower KCCQ scores than men. The results indicated elderly women were prone to have better heart function but had more symptoms and worse quality of life. This phenomenon was not unexpected. An observational study suggested that other factors may explain low life quality in women with HFpEF instead of HF itself, and the reason was multifactorial [15].
Although women with HFpEF were more likely to comorbid with hypertension, there were no signi cant differences between men and women in using antihypertensive drugs, including ACEI, ARB, beta-blocker, CCB, or diuretics. According to previous studies[16 -19], we speculated that these ndings might be attributed to different pharmacokinetics and pharmacodynamics. Women taking ACEI, ARB, and beta-blocker had higher plasma drug concentration than men. Men were more likely to be taking statin, warfarin, or hypoglycemic medications than women; these differences may be explained by the fact that men were more likely to comorbid with AF, diabetes, and dyslipidemia in the present study.
Remarkably, elderly women had lower CV mortality rates and HF hospitalization than men among the patients taking placebo in the present study. This nding was in contrast with Merrill's report [14], which showed that women and men present with similar clinical outcomes. Speci cally, CV mortality and HF hospitalization rates were numerically lower in women than in men (2.15% vs. 2.68%; 3.79% vs. 4.43%, respectively), but there were no statistically signi cant differences. Because our cohort only included patients of ≥70 years of age, the above difference might be caused by different age groups. The secondary analysis using data from the I-PRESERVE study indicated that women had signi cantly lower mortality rates and HF hospitalization than men[6].
Moreover, the meta-analysis[8] indicated the risk of the primary outcome was lower in women (HR 0.80, 95% CI: 0.73-0.88), as was the risk of cardiovascular death (HR 0.70, 95% CI: 0.62-0.80), but there was no difference in the rate for the rst hospitalization for heart failure (HR 0.92, 95% CI: 0.82-1.02). Among the patients taking spironolactone, elderly women had a lower CV mortality rate than men but not HF hospitalization. After adjusting for baseline characteristics, elderly women, who either took a placebo or spironolactone, had a lower risk of CV mortality. However, there was no difference in HF hospitalization risk between elderly women and men from the placebo arm and spironolactone arm. These ndings suggested that women were at a lower risk of the primary composite endpoint than men due to a substantially lower risk of cardiovascular death but not HF hospitalization.
Previous individual patient-data meta-analysis [20] has shown that mineralocorticoid receptor antagonists (MRAs) reduced morbidity and mortality in elderly patients with HF, a bene cial effect that was more marked in patients with HFrEF but homogenous across HFrEF and HFpEF. Speci cally, in elderly patients with HFpEF from the TOPCAT-Americas trial, the primary outcome of cardiovascular death or HF hospitalization occurred in 116 patients (32.7%) taking placebo and 111 patients (28.1%) taking MRA, with a corresponding HR of 0.83 (95% CI: 0.63 to 1.08), suggesting spironolactone therapy failed to be associated with reduced CV mortality and HF hospitalization in elderly patients. After strati ed by sex in the present analysis, there was no signi cant reduction in CV mortality and HF hospitalization associated with spironolactone in elderly women and men. The result kept in line with a previous study that showed the interaction between spironolactone and sex was nonsigni cant for the CV mortality and HF hospitalization in TOPCAT overall [7], and the TOPCAT study restricted to the Americas [14]. Although there was no signi cant sex-treatment interaction, spironolactone treatment had a signi cantly lower multivariate risk of the all-cause mortality in elderly women but not in men, suggesting a possible sex difference in spironolactone treatment concerning allcause mortality.

Limitations
Firstly, this is a post-hoc, exploratory subgroup analysis that strati ed the TOPCAT subjects older than 70 years according to sex and treatment arm, and all ndings are hypothesis-generating only. Secondly, the dose and treatments could vary between the Americas and other regions (Russia and Georgia) [21,22], which might in uence researchers to analyze treatment response. Finally, the present study was underpowered to assess differences in outcomes and response to treatment between sex above the age of 75 years. However, there was likely to be better outcomes in this very elderly woman.
Our analysis showed that elderly women with HFpEF had fewer comorbidities and more likely to have better heart function but had more symptoms and worse quality of life than men. Importantly, elderly women were at a lower risk of the primary composite endpoint than men due to a substantially lower risk of cardiovascular death but not HF hospitalization. It is worth noting that there was a possibility of reducing all-cause mortality associated with spironolactone therapy in elderly women.

PERSPECTIVES AND SIGNIFICANCE
Although our research showed that spironolactone has no signi cant effect on improving the prognosis of men and women with HFpEF, the current analysis showed that women with HFpEF have worse clinical symptoms but better outcomes than men with HFpEF. This study suggests that exploring the in-depth mechanism of HFpEF prognostic differences caused by sex differences may help discover new targets for HFpEF treatment in the future.
Abbreviations ABBREVIATIONS/ ACRONYMS FULL NAME ACE angiotensin-converting enzyme ARB angiotensin receptor blockers AF. atrial brillation BMI body mass index BUN blood urea nitrogen CABG coronary artery bypass grafting CCB calcium channel blockers CV cardiovascular DBP diastolic blood pressure eGFR estimated glomerular ltration rate HFpEF heart failure with a preserved ejection fraction HFrEF heart failure with a reduced ejection fraction HGB hemoglobin HR. heart rate KCCQ Kansas City Cardiomyopathy Questionnaire MI myocardial infarction PCI percutaneous coronary intervention SBP systolic blood pressure Declarations Ethics approval and consent to participate No administrative permissions were required to access the raw data, and the data used in this study were anonymized.

Consent for publication
Written informed consent for publication was obtained from all participants.

Availability of data and material
The datasets used or analyzed during the current study are available from the National Heart, Lung, and Blood Institute's Biological Specimen and Data Repository Information Coordinating Center (BioLINCC, Calverton, Maryland).

Competing interests
The authors declare that they have no con ict of interest.

Funding
This research was partly supported by the Natural Science Foundation of China (NSFC) Projects 81670269 (to S Zhou) and 81801394 (to S Tai).

Authors' contributions
The work presented here was carried out in collaboration with all authors. SZ de ned the study theme and methods. JS and ST collected clinical data, analyzed the data, interpreted the results, and wrote the paper. YG, LT, HY, XL, ZH X, LF are the attending doctor responsible for reviewing and giving suggestions to the manuscript. JS and ST contributed equally to this article. All authors have contributed to, seen, and approved the manuscript.    Table 3. Values are n (%). Cox proportional hazards model to explore the associations between sex and the outcomes. Abbreviations as in Tables 1. Figure 1 Kaplan-Meier survival curves for primary outcomes and components strati ed by sex according to treatment. There was a signi cant association between sex and the primary outcome, CV mortality in either placebo arm or spironolactone arm. Women had a signi cantly lower rate of all the primary outcome in all patients; CV=cardiovascular diseases, HF= heart failure; HR= hazard ratio