Author + information
- Received March 7, 2017
- Revision received June 13, 2017
- Accepted July 27, 2017
- Published online September 25, 2017.
- Alberto Aimo, MDa,
- Giuseppe Vergaro, MDa,b,
- Vincenzo Castiglione, MSa,
- Andrea Barison, MD, PhDa,b,
- Emilio Pasanisi, MDb,
- Christina Petersen, MDb,
- Vladyslav Chubuchny, MDb,
- Alberto Giannoni, MD, PhDb,
- Roberta Poletti, MDb,
- Silvia Maffei, MDb,
- James L. Januzzi Jr., MDc,
- Claudio Passino, MDa,b and
- Michele Emdin, MD, PhDa,b,∗ ( )()
- aInstitute of Life Sciences, Scuola Superiore Sant’Anna, Pisa, Italy
- bCardiology Division, Fondazione Toscana Gabriele Monasterio, Pisa, Italy
- cCardiology Division, Massachusetts General Hospital and Harvard Clinical Research Institute, Boston, Massachusetts
- ↵∗Address for correspondence:
Prof. Michele Emdin, Scuola Superiore Sant’Anna and Fondazione Toscana Gabriele Monasterio, Via Giuseppe Moruzzi 1, Pisa 56124, Italy.
Objectives This study sought to investigate sex-related differences in reverse remodeling (RR).
Background RR, that is, the recovery from left ventricular (LV) dilation and dysfunction in response to treatment for heart failure (HF), is associated with improved prognosis.
Methods Data from patients with stable systolic HF (LV ejection fraction [LVEF] of <50%) undergoing 2 transthoracic echocardiograms within 12 ± 2 months were analyzed. Reverse remodeling was defined as a ≥15% reduction in LV end-systolic volume index.
Results A total of 927 patients were evaluated (68 ± 12 years; median LVEF = 35% [interquartile range: 30% to 43%]; 27% women). Ischemic HF was less often encountered in women (33% vs. 60%, respectively; p < 0.001), whereas most characteristics did not differ with regard to sex. Women showed a higher incidence of RR (41% vs. 27%, respectively; p < 0.001), despite similar baseline LV volume and function. RR was more frequent among women in the subgroups with either ischemic or nonischemic HF, as well as in all categories of systolic dysfunction (LVEF ≤35% or >35%, according to current indication for device implantation, and LVEF <40% or 40% to 50% according to the definition of HF with reduced or mid-range EF). In the whole population, female sex was an independent predictor of RR (hazard ratio: 1.54; 95% confidence interval: 1.11 to 2.14; p = 0.011), together with cause of HF, disease duration, and left bundle branch block. Female sex was again an independent predictor of RR in all LVEF categories.
Conclusions Reverse remodeling is more frequent among women, regardless of cause and severity of LV dysfunction. Female sex is an independent predictor of RR in all categories of LV systolic dysfunction.
Heart failure (HF) is the most common diagnosis in hospitalized patients ≥65 years of age (1). HF burden is increasing with the aging population, and its prognosis remains worse than that with most cancers (1). A major research effort is required to increase knowledge of factors influencing disease evolution and clinicians’ ability to improve outcome.
Adverse remodeling of the left ventricle (LV) is a progressive deterioration of cardiac structure and function, which can occur along the natural history of HF and is associated with worse outcome (2). Conversely, reverse remodeling (RR), i.e., the recovery from LV dilation and dysfunction in response to guideline-recommended HF therapy (3,4) has been associated with reduced all-cause (5) and cardiovascular (6) mortality, as well as hospitalizations for HF (6). Despite HF treatment, RR occurs in only approximately one-third of patients, with estimates ranging from 26% to 46% (5–9).
Among the most promising predictors of RR are nonischemic etiology, the absence or a lower extent of late gadolinium enhancement, lower disease duration, and plasma soluble suppression of tumorigenesis-2 (sST2) activity (5,7,9); furthermore, a multivariate score has been proposed, demonstrating fair accuracy in the prediction of RR (7). Nonetheless, prediction of RR remains challenging (3).
A better understanding of RR predictors would be useful for clinical planning, prognostic discussions, and decision making about potential need for care escalation. For example, device implantation could be deferred when a patient profile points to a higher probability of recovery.
Notably, although sex-related differences have been described for clinical characteristics and prognosis of HF (10,11), the influence of sex on RR remains uncertain. The present study investigated the relationship between sex and RR in a large cohort of patients with systolic HF.
Patient population and study protocol
Stable outpatients with previously diagnosed chronic systolic HF (left ventricular ejection fraction [LVEF] <50%, determined using transthoracic echocardiography [TTE]), evaluated at the Fondazione Toscana Gabriele Monasterio, Pisa, Italy, from 1999 to 2015 and undergoing 2 TTE examinations within 12 ± 2 months, were retrospectively studied.
The following baseline investigations had been performed: complete clinical evaluation, 12-lead electrocardiogram, and blood sampling, including N-terminal pro–B-type natriuretic peptide (12). All investigations were performed within 1 week. Patients had received guideline-recommended therapy for HF and had undergone clinical follow-up examinations at our outpatient clinic every 3 to 6 months, as clinically indicated.
Standard, 2-dimensional TTE images were obtained (Philips IE33 ultrasound machine, with X5-1 transducer; Philips Medical Systems, Palo Alto, California). Wall thickness, chamber volumes, and indices of systolic and diastolic functions had been assessed according to American Society of Echocardiography and European Association of Cardiovascular Imaging guidelines, with volumes measured using the biplane method of disks (modified Simpson's rule) (13–17). All parameters associated with cardiac chamber size had been normalized for body surface area. Sex-specific cutoffs for normalized LV volumes and mass had also been considered according to current guidelines (15). RR was defined as ≥15% left ventricular end-systolic volume index (LVESVi) reduction across the 2 TTE examinations, as previously reported (18,19).
Patients were also stratified according to baseline LV systolic functions: LVEF ≤35 versus >35% (threshold cutoff for insertable cardioverter-defibrillator therapy for primary prevention); LVEF <40% (threshold that defines HF with reduced EF); and 40% to 50%, the definition of HF with mid-range ejection fraction, according to the latest European Society of Cardiology guidelines (13).
Statistical analysis was performed by using SPSS Statistics version 22, 2013 edition (IBM, Armonk, New York]). Normal distribution was assessed by using the Kolmogorov-Smirnov test; variables with normal distribution are mean ± standard deviation, whereas those with non-normal distribution are median and interquartile range (IQR). Mean differences among groups were evaluated using the unpaired Student t test or Mann-Whitney U test, when appropriate. Discrete variables were compared by using the chi-square test with Yates correction or the Fisher exact test. The Spearman rank correlation test was used to determine the correlation between different numerical variables. Predictors of RR were assessed by using logistic regression analysis. Hazard ratios (HRs) with corresponding 95% confidence intervals (CIs) were calculated by using binomial logistic regression analysis. A p value of <0.05 was considered statistically significant.
Sex and baseline patient characteristics
Out of 927 patients studied, 250 were women (27%; mean age 68 years in both sexes). Most women (n = 225; 90%) in the analysis were “aged ≥49 years” (i.e., the mean age of natural menopause in the Framingham Heart Study) (20). HF caused by ischemia was observed in 486 patients (52%), and less often in women than in men (33% vs. 60%, respectively; p < 0.001) (Table 1). Disease duration and most baseline parameters were not significantly different, including LVEF, LVESVi, and left ventricular end-diastolic volume index (LVEDVi), although LVEDVi exceeded a sex-specific cutoff (15) more often among women (Table 1). A total of 530 patients (57%) had LVEF ≤35% and 610 (66%) with LVEF <40%; both sexes were similarly represented across all LVEF categories (Table 1). After therapy optimization during baseline assessment, no sex-related differences were present with regard to drug and device therapy, with the exception of defibrillator implantation either alone or with cardiac resynchronization therapy (CRT) (Table 1).
Female sex and RR
The mean interval between the 2 TTE examinations was 12 months (11–13) among both men and women (p = 0.391). In the whole population, RR incidence was 31% (Online Table 1) and was higher among women than among men (41% vs. 27%, respectively; p < 0.001). Indeed, compared with men, women showed greater percentages of reduction in both LVEDVi (p = 0.037) and LVESVi (p < 0.001) and a greater increase in LVEF (p = 0.003) (Figure 1). RR occurred more frequently in the female sex regardless of HF etiology. Indeed, among patients with ischemic HF, RR was observed in 28 women (35%) versus 91 men (35%); p = 0.002; in the nonischemic subgroup, 71 women (33%) and 90 men (33%) experienced RR (p = 0.003).
Among all characteristics listed in Table 1, several univariate predictors of RR were found and are reported in Table 2. The 5-year intervals of baseline examination (1999 to 2004, 2005 to 2009, and 2010 to 2015) were also evaluated but did not emerge as univariate predictors. Female sex resulted as an independent predictor of RR, together with shorter disease duration, nonischemic cause, and left bundle branch block (Table 2).
Female sex and RR in patients with severe systolic dysfunction
There were 530 patients (57%) with LVEF ≤35%; their baseline characteristics are reported in Online Table 2. RR occurred more often in women (n = 101 [40%]) than in men (n = 184 [28%]; p < 0.001) (Figure 2) with similar baseline LVEF and LV volume indexes (data not shown). Female sex emerged as independent predictor of RR, together with nonischemic etiology lower disease duration, and left bundle branch block (Table 3).
Among patients with LVEF ≤35% and not undergoing CRT (n = 292 [31%]), RR was more often observed in women than in men (n = 33 [41%] vs. n = 60 [28%]; p < 0.001) (Figure 2), despite similar baseline LVEF (30% [IQR: 26% to 35%] vs. 29% [IQR: 25% to 33%], respectively; p = 0.057). Of note, women more often exceeded the 35% LVEF threshold for defibrillator implantation than men (n = 39 [48%] vs. n = 72 [34%], respectively; p = 0.027). When higher LVEF values were considered, 60 of 272 men (22%) crossed the 40% LVEF cutpoint versus 39 of 93 women (42%; p < 0.001). For the 45% LVEF value, these percentages were 14% and 32% for men and women, respectively (p < 0.001).
Among patients with LVEF ≤35% and undergoing CRT (n = 238), RR occurred more often in the female sex (n = 24 [46%] vs. n = 50 [27%], respectively; p = 0.008). Baseline LVEF did not differ significantly (baseline LVEF: 25%; IQR: 21% to 31% vs. baseline LVEF: 28%; IQR: 22% to 30%; p = 0.793).
The subgroup of HF with reduced EF was composed of 610 patients (66%) (Online Table 3). Again, LV measurements between the 2 sexes did not differ (data not shown). Women developed RR more often (n = 65 [42%] vs. n = 128 [28%], respectively; p = 0.001) (Figure 2), and female sex was an independent predictor of RR (Table 3).
Female sex and RR in patients with mild systolic dysfunction
There were 397 patients (43%) with LVEF >35% (Online Table 2). Baseline LVEF and LV volume indexes were similar (data not shown). RR occurred more often among women (n = 101 [40%] vs. n = 184 [28%], respectively; p < 0.001) (Figure 2).
Also in the subgroup with HF and mid-range EF (n = 317, 34%), women developed RR more often than men (n = 36, 39% vs. n = 56, 25%; p = 0.002) (Figure 2), despite a similar severity of LV dilation and dysfunction (data not shown).
Only female sex, lower age, and increased LVEDVi were univariate predictors of RR in the subgroup with LVEF >35%, female sex being the only independent predictor (HR: 1.61; 95% CI: 1.01 to 2.59; p = 0.047). Among patients with HF with mid-range EF, female sex and lower age were univariate predictors, and female sex was again an independent predictor of RR (HR: 1.82; 95% CI: 1.08 to 3.06; p = 0.025).
The development of RR has been repeatedly shown to predict survival benefit from multiple HF treatments, including guideline-recommended medical therapy and CRT. In this large cohort of unselected patients under treatment for systolic HF, women showed a greater improvement in LV geometry and systolic function, and a higher incidence of RR. Although women were more likely to have nonischemic HF (well known to have more frequent RR than ischemic HF), in adjusted analyses that included HF cause and duration, female sex still predicted RR, despite less use of CRT in women and similar LV volumes and function at baseline. Additionally, female sex predicted incident RR across LVEF categories, including those patients whose baseline LVEF was either ≤35% or <40% or milder degrees of systolic dysfunction. Besides improving prognosis significantly, the increase in LVEF in women led to less indication for need of defibrillator therapy. Although RR has been widely studied as a function of therapeutic intervention, to our knowledge, a sex-specific impact on RR is a novel finding.
The reasons why underlying higher incidence of RR among women and female sex predicts RR are not clear. In the present study, baseline clinical, biohumoral, and echocardiographic characteristics were homogeneous in both sexes; the only differences worth noting were a higher frequency of nonischemic cause and smaller LV chamber size among women. Although this suggests the mechanism of RR in women may relate to smaller LV size or higher frequency of nonischemic status, in adjusted analyses, RR was still predicted by female sex. Thus, biological differences may help to explain our finding. Important differences in myocardial remodeling between female and male subgroups in response to different types of cardiac stress or injury (including aging and pressure and volume overload and myocardial infarction) have been found in both experimental and clinical settings, as reviewed by Piro et al. (21). In particular, the remodeling process has been consistently reported to be more favorable in women than in men. The observed differences have led to considerable speculation regarding the underlying cause, and the effects of estrogen on vascular and cardiac cells are deemed to play a crucial role (21). However, most women present with HF after menopause, when circulating estrogen levels are low, thus questioning their protective effects. In this regard, it is important to distinguish between ovarian synthesis of estrogens, which is subject to dramatic changes during the course of life, particularly after menopause, and intramyocardial synthesis, which seems stable over the lifetime span (22). Cardiomyocytes and cardiac fibroblasts all express estrogen receptors, the activation of which promotes positive myocardial remodeling through both genomic and nongenomic effects (23–26). Of note, estrogen receptors can be recruited to activate estrogen-responsive genes even in the absence of estrogens (27), and an up-regulation of nongenomic estrogen receptor signaling has been speculated to occur in post-menopausal women (21). Whereas estrogens (either endogenous or derived from hormone replacement therapy) foster cardiac recovery, fragmentary evidence points to a detrimental role of androgens, which have been reported to promote myocardial hypertrophy, fibrosis, and apoptosis (28–30). Therefore, sex-related differences in RR and prognosis could stem from both a positive effect of estrogens in women and a deleterious effect of androgens in men.
Beyond a sex hormone difference, it is also reasonable to speculate that women display a more prominent response to HF drug and/or device treatment. Indeed, in prior studies, female sex was an independent predictor of super-response to CRT plus defibrillator (defined as an absolute LVEF increase of 18 ± 3%) (31); in the present study, women undergoing CRT developed RR more often than men. With regard to HF medications, variability in drug responses has been observed in a few post hoc analyses of clinical trials (10,11,32,33). Sex-related differences may be explained by estrogen effects on drug metabolism and/or actions on target cells, although clear data are lacking (34,35). Most drugs do not appear to confer greater prognostic benefit in women (10), but data regarding RR are limited, with the exception of data for eplerenone, which has been shown to reduce myocardial infarction-induced cardiac remodeling preferentially in female sex (36). Last, it is possible women are simply more adherent to drug treatment, as previously proposed in one study (37). Future studies are required to better define the mechanism of enhanced RR in women.
Whatever its explanation, the greater propensity of women to RR has a significant clinical value. Indeed, RR means improved clinical status, need for less intensive follow-up and support strategies, less likelihood for device therapy, and as demonstrated by previous studies, better prognosis (5,6). Patient sex should therefore be considered during the process of risk stratification, which in turn informs therapy decision making.
Limitations to our study include the fact that women constituted only 27% of our study cohort. Nevertheless, this was a nonselected group of patients, and women were more represented than in most clinical trials of systolic HF, in which percentages ranged from 0% (38,39) to 32% (11). No information was available regarding hormone replacement therapy or compliance to treatment. Furthermore, although we carefully adjusted our models for demographic and morphologic differences between women and men, it is possible our findings are influenced by the higher prevalence of nonischemic HF and generally smaller LV size in women; in this respect, cutoff values established in healthy subjects for LV size may not be completely adequate in a disease setting, yet they currently represent the only reference point for such type of adjustment (15). Finally, no hard outcomes (i.e., all-cause mortality, cardiovascular mortality, or HF hospitalizations) were included in the present study. Rather, this analysis focused on RR, which was used as a surrogate endpoint given prior research demonstrating RR predicts survival benefit; nevertheless, the link between RR and better prognosis could be further explored in dedicated studies.
In a large population of patients with systolic HF, RR incidence was higher among women despite the homogeneity of clinical, biohumoral, and echocardiographic characteristics. Female sex was an independent predictor of RR, both in the whole population and in patients with severe systolic dysfunction.
COMPETENCY IN MEDICAL KNOWLEDGE: The greater propensity for RR in females has a significant clinical value. Indeed, RR means improved clinical status, less need for intensive follow-up and support strategies, less likelihood for device therapy, and ultimately, better prognosis. Patient sex should therefore be considered during the process of risk stratification, which in turn, informs therapy decision making.
TRANSLATIONAL OUTLOOK: Clarifying the mechanisms of sex-related differences in RR would lead to a better understanding of the remodeling process and possibly to the detection of novel therapeutic approaches.
For supplemental tables, please see the online version of this paper.
Dr. Januzzi has received grant support from Siemens, Singulex, and Prevencio; consulting income from Roche Diagnostics, Critical Diagnostics, Sphingotec, Phillips, and Novartis; and participates in clinical endpoint committees for Novartis, Amgen, Janssen, and Boehringer Ingelheim. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
John R. Teerlink, MD, served as Guest Editor for this paper.
- Abbreviations and Acronyms
- confidence interval
- cardiac resynchronization therapy
- heart failure
- hazard ratio
- left ventricular
- left ventricular end-diastolic volume index
- left ventricular ejection fraction
- left ventricular end-systolic volume index
- reverse remodeling
- transthoracic echocardiogram
- Received March 7, 2017.
- Revision received June 13, 2017.
- Accepted July 27, 2017.
- 2017 American College of Cardiology Foundation
- Konstam M.A.,
- Kramer D.G.,
- Patel A.R.,
- Maron M.S.,
- Udelson J.E.
- Wilcox J.E.,
- Fonarow G.C.,
- Ardehali H.,
- et al.
- Merlo M.,
- Pyxaras S.A.,
- Pinamonti B.,
- Barbati G.,
- Di Lenarda A.,
- Sinagra G.
- Choi J.O.,
- Kim E.Y.,
- Lee G.Y.,
- et al.
- Lupon J.,
- Gaggin H.K.,
- de Antonio M.,
- et al.
- Masci P.G.,
- Schuurman R.,
- Andrea B.,
- et al.
- Simon T.,
- Mary-Krause M.,
- Funck-Brentano C.,
- Jaillon P.
- O'Meara E.,
- Clayton T.,
- McEntegart M.B.,
- et al.
- Ponikowski P.,
- Voors A.A.,
- Anker S.D.,
- et al.
- Nagueh S.F.,
- Smiseth O.A.,
- Appleton C.P.,
- et al.
- Rudski L.G.,
- Lai W.W.,
- Afilalo J.,
- et al.
- Zhang Q.,
- Chan Y.S.,
- Liang Y.J.,
- et al.
- Piro M.,
- Della Bona R.,
- Abbate A.,
- Biasucci L.M.,
- Crea F.
- Grohe C.,
- Kahlert S.,
- Lobbert K.,
- Vetter H.
- Mahmoodzadeh S.,
- Eder S.,
- Nordmeyer J.,
- et al.
- Levin E.R.
- Baron S.,
- Escande A.,
- Alberola G.,
- Bystricky K.,
- Balaguer P.,
- Richard-Foy H.
- Zwadlo C.,
- Schmidtmann E.,
- Szaroszyk M.,
- et al.
- Hsu J.C.,
- Solomon S.D.,
- Bourgoun M.,
- et al.