Introduction
American and European classifications of heart failure (HF) are based on left ventricular (LV) ejection fraction (LVEF) measurements. The criteria for HF with preserved LVEF (HFpEF) are HF symptomology and LVEF ≥ 50%, which indicates LV diastolic dysfunction, despite normal LV systolic functionl [
1,
2]. In addition, HF with LVEF in the range from 40 to 50% is classified as an intermediate group, or HF with a mid-range LVEF (HFmrEF). The HFmrEF class is thought to represent primarily mild decrease in LV contractile performance, combined with features of diastolic dysfunction.
LVEF represents global LV function, and it is commonly used to indicate LV contractile performance in clinical practises. However, LVEF values are influenced by several factors extrinsic to the LV, such as the preload, afterload, and heart rate, in addition to the intrinsic contractile factor and LV dilatation [
3,
4]. Therefore, LVEF ≥ 50% does not accurately reflect the maintenance of normal LV contractile performance. Several previous studies performed detailed examinations of LV contractile performance in patients with LVEF ≥ 50%. Those studies demonstrated that mild impairment in LV contractile performance occurred even in patients with LVEF above 50%, or around 60% [
5‐
10].
We hypothesized that, among patients with LVEF ≥ 50%, some might have slightly impaired LV contractile performance that could not be detected based on LVEF values. We further hypothesized that these patients might be identified with a sophisticated cardiac function parameter that could detect marginally reduced LV contractile performance. Moreover, we hypothesized that these patients also had a potential risk of future HF development. We previously reported that the inertia force of late systolic aortic flow (IFLSAF) could be calculated from LV pressure recordings with a catheter-tipped micromanometer. We demonstrated that the IFLSAF displayed a masterful ability to detect slight impairment in LV contractile performance, and this could be used as a prognostic marker of poor outcome in patients with LVEF ≥ 50% as well as those with coronary artery disease [
9‐
11]. In the present study, with a focus on this LV contractile performance parameter of patients with normal or slightly decreased LVEF and no history of hospitalization for HF, we considered the threshold LVEF value to predict new-onset HF based on LV contractile performance in patients with preserved LVEF.
Discussion
The present study had four major findings. First, we found that mild decrease in LV contractile performance, which was reflected by a loss of the IFLSAF, was an important predictor of new-onset HF in patients with normal or slightly decreased LVEF (≥ 40%). Second, we found that, among patients with LVEF that ranged from 48 to 67%, a loss of the IFLSAF could strongly predict poor prognosis. Third, we demonstrated that the LVEF cut-off value of 58% could serve as a surrogate for determining whether the LV maintained the IFLSAF in patients with LVEF ≥ 40%. Finally, the IFLSAF was correlated with the LVMI, the LVEF, the peak -dP/dt, and the BNP level.
The LVEF has been widely used as a parameter of LV contractile performance for predicting the development of heart failure and cardiovascular death in the general population, in patients with asymptomatic reductions in LVEF, and in patients with symptomatic HF [
16,
17]. In addition, previous studies demonstrated that LVEF was related to differences in patient demographics, comorbid conditions, and response to therapies [
18]. Therefore, LVEF is considered important in classifications of patients with HF. However, it has been challenging to set a range of LVEF that represented no reduction in LV contractile performance. Several detailed examinations of LV contractile performance have observed slight impairment in the performance, even in patients with LVEF ≥ 50%, which is typically classified as a preserved LVEF [
5‐
10]. Similarly, in the current study, 15.3% of patients with LVEF ≥ 50% exhibited decreased LV contractile performance, which was reflected by a loss of the IFLSAF (18.8% in the whole study patients). Previous studies have used echocardiography to investigate the prevalence of decreased LV contractile performance in patients with LVEF ≥ 50%, patients with HFpEF risk factors, such as hypertension and diabetes mellitus, and patients with HFpEF [
7,
19‐
21]. According to the study by Shah et al. [
7], an abnormality in LV contractile performance could be detected by a reduction in the longitudinal strain of the LV. This LV strain reduction was shown to have prognostic utility in patients with HFpEF in the TOPCAT trial (Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist) [
22]. Their findings suggested that decreased LV contractile performance might underlie the pathophysiology in some patients with HFpEF. Consistently, in the present study, we also demonstrated that decreased LV contractile performance, reflected by a loss of the IFLSAF, was an independent predictor of new-onset HF and cardiovascular death in patients with LVEF ≥ 40%. Furthermore, we identified the range of LVEF over which the IFLSAF had prognostic power, based on a time-dependent ROC curve analysis. We found that, among patients with LVEF between 48 and 67%, mild impairment in LV contractile performance, indicated by a loss of the IFLSAF, could lead to future HF development.
HFmrEF was recently recognized as a new general category of HF, different from both HF with reduced LVEF (HFrEF) and HFpEF. However, the pathophysiological mechanisms underlying HFmrEF and the threshold LVEF values for differentiating among the three general categories of HF have not been elucidated. Our main findings suggested that mildly decreased LV contractile performance was associated with HF development in patients with LVEF ≥ 40%. Our findings also suggested that the extent of pathophysiological effects of the decrease in LV contractile performance on HF occurrence could be shown in patients with the LVEF from 48 to 67% who had not experienced hospitalization for HF previously. When patients have mild decrease in LV contractile performance that could lead to HF development, the condition might be classified as HFmrEF. In addition, when the classification of a normal LVEF is defined as no reduction in LV contractile performance, based on the maintenance of IFLSAF, the patients who have LVEF of around 67% or greater are considered with normal contractile performance.
Currently, the only promising medical treatments for patients with HF are for patients with decreased LV contractile performance (LVEF < 35 ~ 40%). When we reconsider the HF classification from the viewpoint of responses to medical treatments for HF, the classifications may not be based on the already established LVEF value, but on LV contractile performance, irrespective of whether or not the LVEF is ≥ 40%. A sub-analysis, reported by Solomon et al. in the TOPCAT trial [
23], demonstrated that the estimated benefits of mineralocorticoid receptor antagonist were stronger in patients with LVEF at the lowest end of the spectrum than in patients with LVEF at the highest end of the spectrum in patients with LVEF ≥ 45%. This finding demonstrated the importance of differentiating patients with HF based on LV contractile performance, due to different responses to medical treatments, even among patients with LVEF ≥ 40%. Although, to date, no convincing medical treatment has been shown to reduce morbidity or mortality in patients with LVEF ≥ 40%, we speculate that by identifying patients with HFmrEF based on LV contractile performance, we could treat a larger proportion of HF patients with drug therapy regimens that were designed for patients with HFrEF.
In this study, we performed a ROC curve analysis to assess the best threshold LVEF value for determining whether the LV was with the IFLSAF. We demonstrated that LVEF values < 58% could predict a loss of the IFLSAF with high sensitivity (85.4%) in patients with LVEF ≥ 40%. This finding indicated that patients with LVEF < 58% had a potential risk of decrease in LV contractile performance that could lead to the development of HF in the future, despite a LVEF > 40%. Therefore, taking decrease in LV contractile performance as a risk factor for new-onset HF, we propose that the upper LVEF cut-off value between HFmrEF and HFpEF should be around 58%. Several previous reports have supported the notion that the LVEF threshold for defining HFpEF should be raised above the 50% threshold commonly used. Some community-based cohort studies demonstrated that the persons with LVEF 55 to 60% had greater risk for morbidity and mortality compared to those with LVEF 60% [
24‐
26]. In addition, patients with HFpEF that had LVEF < 55% were reported to be significantly associated with a risk of the LVEF declining to below 50%, which means that the patients would shift to more severe decrease in LV contractile performance [
27]. Furthermore, when study patients were classified based on another threshold LVEF value a bit higher than 55%, the clinical features of HFpEF were heterogeneous. According to a sub-analysis in the I-PRESERVE Study (Irbesartan in Heart Failure with Preserved Ejection Fraction) [
28], the prognostic impact of LVEF on HFpEF was significantly different when LVEF was below 60% compared to when LVEF was 60% or greater. Additionally, in a sub-analysis of the J-MELODIC trial (Japanese Multicenter Evaluation of Long- vs. short-acting Diuretics In Congestive heart failure), we previously demonstrated that HFpEF was heterogeneous based on the prognostic utility of BNP levels in the LVEF ranges (40–60% and ≥ 60%) [
29]. Furthermore, Solomon et al. also demonstrated the effect of angiotensin receptor-neprilysin inhibition (ARNI) in patients with symptomatic heart failure and LVEF ≥ 45% in the PARAGON-HF trial (The Prospective Comparison of ARNI with Angiotensin-receptor blockers Global Outcomes in Heart Failure with Preserved Ejection Fraction) [
30]. Although ARNI did not reduce the rate of total hospitalizations for HF and death from cardiovascular causes among whole study patients, for the patients with relatively lower LVEF (LVEF ≤ 57%), ARNI significantly reduced morbidity and mortality. The PARAGON-HF trial did demonstrate the efficacy of ARNI in patients with the very similar LVEF range where a decrease in LV contractile performance was observed as the loss of the IFLSAF in the current study. These findings support that a LVEF value < 58% indicates mild decrease in LV contractile performance in patients with LVEF ≥ 40% from the viewpoint of drug effects.
This study had several limitations. First, the study design was retrospective, and we analysed data from a single institution. We recruited study patients from those who underwent diagnostic cardiac catheterization for the evaluation of coronary artery disease concomitantly with a sophisticated measurement of LV pressure with a catheter-tipped micromanometer. It may be associated with lower proportion of female in the study patients. Second, all patients had some HFpEF risk factors, such as hypertension and diabetes mellitus, but no history of hospitalization for HF. Because this study focused on the potential risk of the impairment of LV contractile performance in patients with LVEF ≥ 40% for new-onset HF, the patients with a history of hospitalization for HF were excluded. When we discuss a threshold LVEF value for distinguishing between HFpEF and HFmrEF, we should give consideration to the distinction in the mechanism underlaying HF between new-onset and recurrence HF in patients with LVEF ≥ 40%. In addition, all our patients were in sinus rhythm. Thus, a future prospective study is needed to strengthen our conclusions, with a larger study cohort that includes more patients with a history of hospitalization for HF and atrial fibrillation. Third, we focused on the IFLSAF as a marker of LV contractile performance. Although, compared to LVEF, the IFLSAF could detect small reductions in LV contractile performance, the IFLSAF measurement was somewhat complicated and invasive, because it required LV pressure recordings with a catheter-tipped micromanometer. This procedure is not practical for the bed-side management of patients with HF. Consequently, a more practical parameter for representing LV contractile performance, which could be measured by non-invasive approach such as echocardiography or cardiac magnetic resonance imaging, might be necessary. Finally, we did not address changes in LV contractile performance over the course of HF, nor did we investigate the association between changes in LV contractile performance and prognosis.
In conclusion, we demonstrated that mild reduction in LV contractile performance, indicated by a loss of the IFLSAF, was one of the prognostic indicators for new-onset HF in patients with LVEF ≥ 40%. We showed that an LVEF ≥ 58% could be taken as a surrogate for the IFLSAF maintenance. Moreover, medical treatments that are efficacious for HFrEF might be applicable for the patients with their LVEF < 58%.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.