Abstract
Aims
The trajectories of systolic function after admission for acute heart failure (HF) and their effect on clinical outcomes have not been fully elucidated. We aimed to assess changes in left ventricular ejection fraction (LVEF) between the index and 1 year after discharge and to examine their prognostic implications.
Methods and results
We extracted data from a prospective multicentre registry of patients hospitalized for acute HF and identified 1636 patients with LVEF data at admission and 1 year after discharge. We categorized them into five groups based on LVEF changes: HF with unchanged‐preserved EF [HFunc‐pEF (EF ≥ 50%); N = 527, 32.2%], unchanged‐mildly reduced EF [HFunc‐mrEF (EF 41–49%); N = 86, 5.3%], unchanged‐reduced EF [HFunc‐rEF (EF ≤ 40%); N = 377, 23.0%], worsened EF (HFworEF; N = 83, 5.1%), and improved EF (HFimpEF; N = 563, 34.4%). We then evaluated the subsequent composite outcome of cardiovascular death and HF readmission. During 1 year after discharge, 53% of patients with HF with reduced EF and 67% of those with HF with mildly reduced EF (HFmrEF) transitioned to other categories, whereas 92% of those with HF with preserved EF (HFpEF) remained within the same category. Patients with HFimpEF were more likely to be younger and had relatively preserved renal function, whereas those with HFworEF were the oldest and had more comorbidities among the five groups. After multivariable adjustment, patients with HFimpEF and HFunc‐pEF had a lower risk for composite outcomes when referenced to patients with HFunc‐rEF [hazard ratio (95% confidence interval), P‐value: 0.28 (0.16–0.49), P < 0.001, and 0.40 (0.25–0.63), P < 0.001, respectively]. Conversely, patients with HFunc‐mrEF and HFworEF had a comparable risk [0.44 (0.18–1.07), P = 0.07, and 0.63 (0.29–1.39), P = 0.26, respectively].
Conclusions
A substantial number of patients with HF experienced transitions to other categories after discharge. Notably, patients with decreased EF experienced a worse prognosis, even with slight decreases (e.g. HFpEF transitioning to HFmrEF). These findings emphasize the significance of longitudinal assessments of systolic function to better manage patients following acute decompensation.
Keywords: Acute heart failure, Left ventricular ejection fraction, Mortality, Trajectory
Introduction
The global prevalence of heart failure (HF) is estimated to be 64 million 1 and is expected to increase as the population ages. Moreover, despite advances in clinical management, HF remains critical, given that its prognosis is uniformly poor and that death occurs in 17–45% and 45–60% of patients within 1 and 5 years of admission, respectively. 2
Left ventricular ejection fraction (LVEF) has significant prognostic implications among patients with HF 3 and is critical in classifying HF and defining indications for clinical management. 4 , 5 , 6 However, as systolic function is not static over time with treatment responsiveness or disease progression, 7 LVEF assessment at a single time point may not provide precise long‐term prognostic information. 7 , 8 , 9 , 10 , 11 , 12 This may be more prominent in an episode of acute decompensation of HF, where elevated pressure in the left ventricle provokes its remodelling. 12 , 13 A previous study reported that many patients with HF with reduced EF (HFrEF) experience substantial changes in LVEF over time after acute HF. 12 Nonetheless, understanding the full spectrum of HF, including HF with mildly reduced EF (HFmrEF) and HF with preserved EF (HFpEF), requires knowing the LVEF trajectory, for which studies are currently scarce.
This study aimed to elucidate the clinical demographics, treatment patterns, and risk of clinical outcomes associated with dynamic changes in HF phenotypes, specifically LVEF, in patients hospitalized for acute decompensation. This study was conducted using data from a representative, prospective, contemporary, multicentre cohort study conducted in Japan. By investigating the associations between long‐term outcomes and LVEF trajectory across the full LVEF spectrum, we aimed to offer valuable insights for clinical decision‐making in managing patients with HF.
Methods
Data sources
This study was part of the West Tokyo Heart Failure (WET‐HF) registry, which is a prospective multicentre cohort registry designed to consecutively collect data of patients aged ≥20 years hospitalized for acute HF requiring urgent treatment. 14 , 15 Individual cardiologists clinically diagnosed acute HF at each institution based on the signs and symptoms of HF and the level of natriuretic peptides [i.e. brain natriuretic peptide (BNP) ≥ 100 pg/mL or N‐terminal pro‐brain natriuretic peptide (NT‐proBNP) ≥ 300 pg/mL] at admission. 4 , 16 Patients presenting with acute coronary syndrome or those who refused to participate in the registry were excluded. Participants were registered between January 2006 and December 2017 at six tertiary hospitals in the Tokyo area, and two additional institutions were added in April 2018 (WET‐HF2 registry), which included an update of the collected variables, such as data on medications and LVEF 1 year after discharge. 15
Patient data were entered into an electronic data‐capture system with a robust data query engine and system validation for data quality. The principal investigators (Y. S. and S. K.) conducted periodic queries to verify reporting quality at least once a year. Regarding the endpoints, all death‐related events were reviewed by the investigators and classified into the following groups: those requiring adjudication and those with a clearly defined mode of death. Subsequently, the Central Committee members (Y. S., S. K., and T. Y.) reviewed the abstract records and adjudicated the modes of death.
Before launching this registry, information on the objectives, social significance, and an abstract of this study was provided for clinical trial registration to the University Hospital Medical Information Network of Japan (UMIN000001171). The Institutional Review Board of each site approved the study protocol, and the study was conducted following the Declaration of Helsinki. According to the Ethical Guidelines for Medical and Health Research Involving Human Subjects and the Personal Information Protection Law in Japan, informed consent was obtained from each participant before the study.
Study population
A total of 3641 consecutive patients with HF were registered in WET‐HF2 between April 2018 and December 2021. We excluded patients without LVEF data at admission [N = 143 (3.9%)], those who died during admission [N = 171 (4.7%)], those not followed up or who died during 1 year after discharge [N = 672 (18.5%)], and those without LVEF data 1 year after discharge [N = 1019 (28.0%)]. Subsequently, we categorized the remaining participants (N = 1636, 44.9%) into five groups based on changes in LVEF: (i) HF with unchanged‐preserved EF with LVEF ≥ 50% at admission and 1 year after discharge (HFunc‐pEF), (ii) HF with unchanged‐mildly reduced EF with LVEF 41–49% at admission and 1 year after discharge (HFunc‐mrEF), (iii) HF with unchanged‐reduced EF with LVEF ≤ 40% at admission and 1 year after discharge (HFunc‐rEF), (iv) HF with worsened EF with LVEF ≥ 50% at admission and LVEF ≤ 49% at 1 year after discharge or LVEF 41–49% at admission and LVEF ≤ 40% at 1 year after discharge (HFworEF), and (v) HF with improved EF with LVEF ≤ 40% at admission and LVEF ≥ 41% at 1 year after discharge or LVEF 41–49% at admission and LVEF ≥ 50% at 1 year after discharge (HFimpEF) (Figure 1 ).
Figure 1.
Study flowchart. EF, ejection fraction; HFimpEF, heart failure with improved ejection fraction; HFunc‐mrEF, heart failure with unchanged‐mildly reduced ejection fraction; HFunc‐pEF, heart failure with unchanged‐preserved ejection fraction; HFunc‐rEF, heart failure with unchanged‐reduced ejection fraction; HFworEF, heart failure with worsened ejection fraction; WET‐HF, West Tokyo Heart Failure.
Variables and clinical outcomes
The LVEF at admission, after the HF signs and symptoms stabilized, and 1 year after discharge was assessed via echocardiography using the modified Simpson's method by board‐certified physicians or echocardiographers. The New York Heart Association (NYHA) functional classification was assessed by the attending cardiologists. Estimated glomerular filtration rate (eGFR) was calculated using the Modification of Diet in Renal Disease Equation for Japanese Patients proposed by the Japanese Society of Nephrology. 17
The primary outcomes of this study were a composite of cardiovascular (CV) death and HF readmission 1–2 years after discharge. Secondary outcomes were all‐cause death and HF readmission 1–2 years after discharge. HF death, sudden cardiac death, and other aetiological deaths, including acute coronary syndrome, acute aortic syndrome, intracranial haemorrhage, stroke, and pulmonary thromboembolism, were considered CV deaths. Non‐CV deaths were caused by other factors (including unknown causes of death). Furthermore, we defined renal outcomes as follows: >40% eGFR decline or eGFR decline to <15 mL/min/1.73 m2 from the time of admission to 1 year after discharge.
Statistical analyses
We conducted a comparative analysis of the patient characteristics among the five groups. Parametric and non‐parametric variables were evaluated for differences using one‐way analysis of variance or the Kruskal–Wallis test, as appropriate. Additionally, significant differences between independent categorical variables were assessed using the χ 2 test. To examine changes in patient characteristics, including laboratory data and prescribed medications, between discharge and 1 year of follow‐up, we used McNemar's test, paired t‐test, or Wilcoxon's signed‐rank test, as appropriate, based on the distribution. Participants without relevant data were excluded from McNemar's test. The incidence of composite events, all‐cause death, or HF readmission from 1 to 2 years after discharge was estimated using the Kaplan–Meier survival function and compared among the five groups using the log‐rank test. Similarly, we compared the incidence of composite events among patients with HFrEF, HFmrEF, and HFpEF based on LVEF classifications at admission.
Multivariable Cox proportional hazards analysis was performed to assess the risk factors associated with clinical outcomes and the risk of composite outcomes across the five groups. The selection of covariates for the multivariable model was based on their clinical significance. The first analysis included age, sex, NYHA functional classification ≥III at 1 year, aetiology of HF (valvular heart disease vs. others), the presence of atrial fibrillation (AF), LVEF at 1 year, eGFR at 1 year, and haemoglobin at 1 year. The latter analysis included age, sex, NYHA functional classification ≥III at 1 year, eGFR at 1 year, and haemoglobin at 1 year. Similarly, we conducted a sensitivity analysis that excluded patients with a marked decline in LVEF during the first year [defined as ≥41% at admission and ≤40% at 1 year after discharge; N = 50 (1.4%)]. Finally, we performed a multivariable logistic analysis to identify predictors related to the HFimpEF phenotype.
Statistical significance was set at P < 0.05. All statistical analyses were performed using the R software (Version 4.2.2; R Foundation for Statistical Computing, Vienna, Austria).
Results
Heart failure classification and left ventricular ejection fraction trajectory
At admission, 806 (49.3%), 255 (15.6%), and 575 (35.1%) patients had HFrEF, HFmrEF, and HFpEF, respectively. A total of 162 (20.1%) and 267 (33.1%) patients with HFrEF transitioned to HFmrEF and HFpEF, respectively, within 1 year of discharge. Furthermore, 134 (52.5%) patients with HFmrEF improved, and 35 (13.7%) experienced a decline. In contrast, most patients with HFpEF (N = 527, 91.7%) remained in their initial classification, whereas a few (N = 48, 8.3%) experienced a decline in LVEF status (Figure 2 ). Based on LVEF trajectories, 527 (32.2%), 86 (5.3%), 377 (23.0%), 83 (5.1%), and 563 (34.4%) patients were classified as having HFunc‐pEF, HFunc‐mrEF, HFunc‐rEF, HFworEF, and HFimpEF, respectively (Figure 1 ).
Figure 2.
Trajectories of left ventricular ejection fraction. HFmrEF, heart failure with mildly reduced ejection fraction; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction.
Patient characteristics
We compared patient characteristics 1 year after discharge among the five groups (Table 1 ). Patients with HFimpEF were younger, were less likely to have a history of HF admission (18.3%), had relatively preserved renal function, and had the lowest BNP/NT‐proBNP levels among the five groups. Patients with HFworEF were the oldest, were less likely to receive HF‐specific medications, and had additional comorbidities, such as AF and previous stroke. Valvular heart disease was most prevalent in patients with HFunc‐pEF (48.6%), followed by those with HFworEF (27.3%). The prevalence of NYHA functional classification ≥III was similar between patients with HFunc‐rEF and those with HFworEF.
Table 1.
Patient characteristics
| HFunc‐pEF (N = 527) | HFunc‐mrEF (n = 86) | HFunc‐rEF (n = 377) | HFworEF (n = 83) | HFimpEF (N = 563) | P‐value | |
|---|---|---|---|---|---|---|
| Age (at discharge), years | 77.4 (12.1) | 73.8 (14.3) | 71.0 (13.0) | 79.4 (10.1) | 66.8 (15.3) | <0.001 |
| Female | 271 (51.4%) | 22 (25.6%) | 95 (25.2%) | 28 (33.7%) | 211 (37.5%) | <0.001 |
| Body mass index (at discharge), kg/m2 | 21.8 (4.1) | 22.0 (4.0) | 22.0 (4.6) | 21.9 (4.2) | 22.6 (4.7) | 0.09 |
| History of HF hospitalization | 131 (24.9%) | 23 (26.7%) | 155 (41.1%) | 23 (27.7%) | 103 (18.3%) | <0.001 |
| LVEF (at 1 year), % | 61.8 (6.7) | 45.4 (2.7) | 29.9 (7.1) | 39.2 (7.2) | 54.0 (7.7) | <0.001 |
| NYHA functional class ≥III (at 1 year) | 24 (4.6%) | 4 (4.7%) | 28 (7.4%) | 6 (7.2%) | 18 (3.2%) | 0.044 |
| Systolic blood pressure (at 1 year), mmHg | 124.5 (20.1) | 122.5 (19.7) | 113.4 (19.7) | 119.3 (19.4) | 124.8 (19.9) | <0.001 |
| Heart rate (at 1 year), b.p.m. | 72.1 (13.0) | 73.8 (16.5) | 73.9 (12.7) | 73.1 (12.1) | 72.9 (13.4) | 0.38 |
| Laboratory data (at 1 year) | ||||||
| Haemoglobin, g/dL | 11.9 (2.0) | 12.4 (2.3) | 12.8 (2.0) | 11.6 (1.9) | 13.0 (1.9) | <0.001 |
| eGFR, mL/min/1.73 m2 | 45.6 (19.6) | 45.6 (25.2) | 44.8 (20.7) | 38.5 (24.2) | 52.4 (21.3) | <0.001 |
| Serum sodium, mEq/L | 139.9 (2.8) | 140.1 (2.7) | 139.5 (3.5) | 138.9 (3.7) | 140.0 (3.1) | 0.006 |
| Serum potassium, mEq/L | 4.4 (0.5) | 4.5 (0.6) | 4.4 (0.5) | 4.3 (0.5) | 4.4 (0.5) | 0.005 |
| BNP, pg/mL | 134.6 (63.6–278.9) | 178.8 (105.5–318.9) | 231.6 (83.3–547.0) | 230.2 (107.0–485.0) | 60.1 (19.7–144.1) | <0.001 |
| NT‐proBNP, pg/mL | 1328 (511.8–3438) | 2215 (1044–7691) | 1645 (896.0–4724) | 2583 (1199–4163) | 583.0 (300.0–1396) | <0.001 |
| Aetiology | ||||||
| Ischaemic | 69 (13.1%) | 38 (44.2%) | 164 (43.5%) | 24 (28.9%) | 117 (20.8%) | <0.001 |
| Dilated | 6 (1.1%) | 5 (5.8%) | 106 (28.1%) | 6 (7.2%) | 111 (19.7%) | |
| Valvular | 256 (48.6%) | 21 (24.4%) | 40 (10.6%) | 23 (27.7%) | 95 (16.9%) | |
| Hypertensive | 13 (2.5%) | 1 (1.2%) | 0 | 1 (1.2%) | 7 (1.2%) | |
| Hypertrophic | 21 (4.0%) | 1 (1.2%) | 8 (2.1%) | 1 (1.2%) | 6 (1.1%) | |
| Amyloidosis | 4 (0.8%) | 1 (1.2%) | 3 (0.8%) | 0 | 2 (0.4%) | |
| Sarcoidosis | 1 (0.2%) | 0 | 7 (1.9%) | 0 | 9 (1.6%) | |
| Others/unknown | 157 (29.8%) | 19 (22.1%) | 49 (13.0%) | 28 (33.7%) | 216 (38.4%) | |
| Comorbidities | ||||||
| Atrial fibrillation | 239 (45.4%) | 34 (39.5%) | 138 (36.6%) | 37 (44.6%) | 180 (32.0%) | <0.001 |
| Hypertension | 361 (68.5%) | 59 (68.6%) | 227 (60.2%) | 57 (68.7%) | 327 (58.1%) | <0.001 |
| Diabetes | 146 (27.7%) | 33 (38.4%) | 141 (37.4%) | 32 (38.6%) | 161 (28.6%) | 0.004 |
| Dyslipidaemia | 187 (35.5%) | 42 (48.8%) | 168 (44.6%) | 28 (33.7%) | 170 (30.2%) | <0.001 |
| Previous stroke/TIA | 64 (12.1%) | 10 (11.6%) | 47 (12.5%) | 18 (21.7%) | 43 (7.6%) | 0.002 |
| COPD | 23 (4.4%) | 1 (1.2%) | 9 (2.4%) | 6 (7.2%) | 16 (2.8%) | 0.08 |
| Dialysis | 7 (1.3%) | 3 (3.5%) | 3 (0.8%) | 3 (3.6%) | 11 (2.0%) | 0.19 |
| Dementia | 33 (6.3%) | 6 (7.0%) | 11 (2.9%) | 4 (4.8%) | 14 (2.5%) | 0.011 |
| Malignant tumour | 4 (0.8%) | 0 | 1 (0.3%) | 0 | 6 (1.1%) | 0.49 |
| Pacemaker | 31 (5.9%) | 7 (8.1%) | 20 (5.3%) | 14 (16.9%) | 21 (3.7%) | <0.001 |
| ICD | 3 (0.6%) | 2 (2.3%) | 39 (10.3%) | 1 (1.2%) | 5 (0.9%) | <0.001 |
| CRT | 0 | 1 (1.2%) | 18 (4.8%) | 0 | 1 (0.2%) | <0.001 |
| Medications at discharge | ||||||
| Beta‐blocker | 279 (61.7%) | 64 (81.0%) | 314 (90.0%) | 51 (68.0%) | 465 (82.6%) | <0.001 |
| ACE‐I/ARB | 202 (44.9%) | 48 (60.8%) | 234 (67.2%) | 31 (41.3%) | 380 (67.5%) | <0.001 |
| ARNI | 9 (2.0%) | 4 (5.1%) | 22 (6.3%) | 1 (1.2%) | 40 (7.7%) | 0.001 |
| MRA | 132 (29.4%) | 23 (29.1%) | 205 (59.6%) | 27 (36.0%) | 252 (44.8%) | <0.001 |
| SGLT2i | 34 (7.6%) | 17 (21.5%) | 99 (28.5%) | 17 (22.7%) | 90 (16.0%) | <0.001 |
| Ivabradine | 1 (0.2%) | 1 (1.3%) | 11 (3.2%) | 1 (1.3%) | 18 (3.2%) | 0.006 |
| Loop diuretic | 323 (71.9%) | 59 (74.7%) | 293 (84.4%) | 48 (64.0%) | 306 (54.4%) | <0.001 |
| Thiazide diuretic | 13 (2.9%) | 5 (6.3%) | 12 (3.5%) | 5 (6.7%) | 4 (0.7%) | 0.002 |
| Tolvaptan | 95 (21.1%) | 28 (35.4%) | 97 (28.0%) | 21 (28.0%) | 68 (12.1%) | <0.001 |
| Digoxin | 13 (2.9%) | 1 (1.3%) | 7 (2.0%) | 1 (1.3%) | 14 (2.5%) | 0.81 |
| Antiplatelet | 157 (34.9%) | 42 (52.5%) | 153 (44.2%) | 32 (42.7%) | 153 (27.2%) | <0.001 |
| Anticoagulant | 257 (57.2%) | 44 (55.7%) | 184 (53.2%) | 38 (50.7%) | 251 (44.6%) | 0.072 |
| Statin | 182 (40.4%) | 46 (58.2%) | 199 (57.2%) | 38 (50.7%) | 247 (43.9%) | <0.001 |
ACE‐I, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; ARNI, angiotensin receptor neprilysin inhibitor; BNP, brain natriuretic peptide; COPD, chronic obstructive pulmonary disease; CRT, cardiac resynchronization therapy; eGFR, estimated glomerular filtration rate; HF, heart failure; HFimpEF, heart failure with improved ejection fraction; HFunc‐mrEF, heart failure with unchanged‐mildly reduced ejection fraction; HFunc‐pEF, heart failure with unchanged‐preserved ejection fraction; HFunc‐rEF, heart failure with unchanged‐reduced ejection fraction; HFworEF, heart failure with worsened ejection fraction; ICD, implantable cardioverter defibrillator; LVEF, left ventricular ejection fraction; MRA, mineralocorticoid receptor antagonist; NT‐proBNP, N‐terminal pro‐brain natriuretic peptide; NYHA, New York Heart Association; SGLT2i, sodium–glucose cotransporter‐2 inhibitor; TIA, transient ischaemic attack.
Values are mean (standard deviation), median (interquartile range), or n (%).
Clinical outcomes
The crude incidence of the composite outcomes was highest in patients with HFunc‐rEF, followed by those with HFworEF, HFunc‐pEF, HFunc‐mrEF, and HFimpEF [21.9 vs. 18.5 vs. 11.9 vs. 10.8 vs. 5.4 (per 100 person‐years); log‐rank P < 0.001] (Figure 3 A and Supporting Information, Table S1 ). This trend persisted in the incidence of all‐cause death and HF readmission (log‐rank P < 0.001) (Supporting Information, Figures S1 and S2 , respectively). In contrast, the incidence of the composite outcome was similar among patients with HFrEF, HFmrEF, or HFpEF based on their classification at admission (log‐rank P = 0.81; Figure 3 B ). Regarding the renal outcome, the incidence of >40% decline in eGFR from the time of admission and 1 year after discharge and eGFR of <15 mL/min/1.73 m2 at 1 year after discharge was highest in patients with HFunc‐pEF (7.9%) and HFworEF (18.4%), respectively (Supporting Information, Table S2 ).
Figure 3.
Kaplan–Meier curves for composite outcomes (A) among the five groups and (B) among patients with heart failure with reduced ejection fraction (HFrEF), heart failure with mildly reduced ejection fraction (HFmrEF), and heart failure with preserved ejection fraction (HFpEF) based on left ventricular ejection fraction classifications at admission. HFimpEF, heart failure with improved ejection fraction; HFunc‐mrEF, heart failure with unchanged‐mildly reduced ejection fraction; HFunc‐pEF, heart failure with unchanged‐preserved ejection fraction; HFunc‐rEF, heart failure with unchanged‐reduced ejection fraction; HFworEF, heart failure with worsened ejection fraction.
After multivariable adjustment, patients with HFimpEF and HFunc‐pEF, and HFunc‐mrEF had a lower risk for the composite outcome [hazard ratio (95% confidence interval), P‐value: 0.28 (0.16–0.49), P < 0.001, and 0.40 (0.25–0.63), P < 0.001, respectively] than those with HFunc‐rEF. Conversely, patients with HFunc‐mrEF and HFworEF had a comparable risk [0.44 (0.18–1.07), P = 0.07, and 0.63 (0.29–1.39), P = 0.26, respectively] (Figure 4 ). We conducted a sensitivity analysis to compare the incidence of clinical outcomes by excluding patients who exhibited a significant reduction in LVEF, which we defined as a decrease from ≥41% at admission to ≤40% at 1 year after discharge. This resulted in the exclusion of 50 (1.4%) patients from the analysis; however, we observed results directionally similar to those obtained in the full cohort (Supporting Information, Figures S3 – S6 ). Finally, age, LVEF, and haemoglobin level were independently associated with composite outcomes, all‐cause death, and HF readmission (Supporting Information, Table S3A–C ).
Figure 4.
Multivariable Cox regression analysis for the composite outcomes. Forest plots showing the hazard ratios for the composite outcomes of cardiovascular death and heart failure readmission. CI, confidence interval; HFimpEF, heart failure with improved ejection fraction; HFunc‐mrEF, heart failure with unchanged‐mildly reduced ejection fraction; HFunc‐pEF, heart failure with unchanged‐preserved ejection fraction; HFunc‐rEF, heart failure with unchanged‐reduced ejection fraction; HFworEF, heart failure with worsened ejection fraction.
Changes in patient characteristics over 1 year
Changes in patient characteristics between admission and 1 year after discharge for each group are described in Supporting Information, Table S4A–E . The NT‐proBNP levels remained unchanged for >1 year after discharge, except in patients with HFimpEF. The eGFR levels decreased in all five groups. The proportion of use of beta‐blocker and angiotensin‐converting enzyme inhibitor or angiotensin receptor blocker (ACE‐I/ARB) remained unchanged, whereas those of angiotensin receptor neprilysin inhibitor (ARNI) and sodium–glucose cotransporter‐2 inhibitor (SGLT2i) increased in patients with HFimpEF and HFunc‐rEF. Conversely, for patients with HFworEF, the proportion of individual HF‐specific medications differed insignificantly over 1 year.
Predictors for the heart failure with improved ejection fraction phenotype
Following multivariable adjustment, younger age, female sex, LVEF, no history of HF admission, NYHA functional class (≤II at discharge), haemoglobin level, and use of ACE‐I/ARB at discharge were independently associated with the HFimpEF phenotype (Table 2 ).
Table 2.
Predictors of heart failure with improved ejection fraction
| Variables | OR (95% CI) | P‐value |
|---|---|---|
| Age (per 1 year increase) | 0.98 (0.97–0.99) | 0.001 |
| Female sex | 1.90 (1.46–2.48) | <0.001 |
| BMI at discharge (per 1 kg/m2 increase) | 0.99 (0.96–1.02) | 0.64 |
| No prior HF admission | 2.28 (1.71–3.08) | <0.001 |
| Valvular heart disease (vs. others) | 1.10 (0.80–1.51) | 0.55 |
| NYHA functional class ≤II (at discharge) | 2.06 (1.16–3.85) | 0.017 |
| LVEF at admission (per 1% increase) | 0.94 (0.93–0.95) | <0.001 |
| eGFR at discharge (per 1 mL/min/1.73 m2 increase) | 1.00 (0.99–1.01) | 0.47 |
| Haemoglobin at discharge (per 1 g/dL increase) | 1.11 (1.05–1.18) | <0.001 |
| Atrial fibrillation | 0.96 (0.74–1.23) | 0.73 |
| Hypertension | 1.06 (0.82–1.37) | 0.68 |
| Diabetes | 0.91 (0.70–1.19) | 0.49 |
| Beta‐blocker (at discharge) | 1.28 (0.89–1.19) | 0.19 |
| ACE‐I/ARB (at discharge) | 1.37 (1.05–1.79) | 0.021 |
| MRA (at discharge) | 0.96 (0.75–1.24) | 0.76 |
| Loop diuretic (at discharge) | 0.86 (0.62–1.20) | 0.37 |
ACE‐I, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; BMI, body mass index; CI, confidence interval; eGFR, estimated glomerular filtration rate; HF, heart failure; LVEF, left ventricular ejection fraction; MRA, mineralocorticoid receptor antagonist; NYHA, New York Heart Association; OR, odds ratio.
Discussion
The present study investigated LVEF trajectories between admission for acute HF and 1 year after discharge and their influence on subsequent clinical outcomes. We found that a substantial number of patients with HFrEF or HFmrEF experienced improved systolic function after acute decompensation. Furthermore, given that HF phenotypes based on LVEF trajectories rather than LVEF status at baseline were significantly associated with subsequent clinical outcomes, longitudinal evaluations of systolic function provide valuable prognostic implications.
Using data from the Swedish Heart Failure Registry, Savarese et al. reported that 26% and 39% of the patients with HFrEF and HFpEF, respectively, transitioned to other LVEF categories over time. 11 In addition, Tsuji et al. reported that in the Chronic Heart Failure Analysis and Registry in the Tohoku District‐2 Study, 40% of patients with HFrEF transitioned to HFmrEF or HFpEF within a year; however, most patients with HFpEF remained in the same category (only 10% transitioned to other LVEF categories). 10 However, these investigations were limited to patients with stable HF in an outpatient setting. Comprehending the LVEF trajectory in vulnerable patients hospitalized for acute HF is essential, as most of those with HF experience acute decompensation in their lifetime. Our study demonstrated that more than half of the patients with HFrEF and HFmrEF had improved systolic function after discharge, underscoring the significance of conducting longitudinal assessments of systolic function, particularly following acute HF. Furthermore, Albert et al. reported a similar trend towards LVEF improvement during a 6 month follow‐up in patients with HFrEF hospitalized for acute decompensation. 12 In contrast, the lower prevalence of LVEF reduction among patients with HFpEF in our study may be attributed to the underrepresentation of patients with HFpEF who underwent post‐discharge echocardiographic assessment. Moreover, the relatively small proportion of individuals with an ischaemic aetiology in our cohort compared with those in Western cohorts may have influenced this finding.
The risk of clinical adverse events in patients with acute HF is assessed traditionally by measuring LVEF during initial admission, 5 , 6 and outcomes are determined after a follow‐up period using an intention‐to‐treat approach. However, our analysis, with no differences in clinical outcomes among the LVEF classifications at admission, highlighted the significant limitation of using only LVEF at a specific time point in the conventional assessment of acute HF. While our study demonstrated the significance of dynamic changes in LVEF, we did not propose a new risk stratification scheme to replace conventional risk assessment. Our primary objective was to highlight that the risk of patients with HF may change continuously, emphasizing the significance of regular longitudinal evaluations with LVEF reassessment and management strategies. Furthermore, we extended the previous literature by providing crucial information on the LVEF trajectory across the full LVEF spectrum of patients with HF following acute decompensation.
Previous studies have reported that HFimpEF has a better prognosis than other LVEF phenotypes 7 , 8 , 9 , 10 , 11 , 12 ; however, HFworEF, which shows decreased LVEF over time, has a poor prognosis. 10 , 11 Our results are comparable with those of previous studies. Notably, even though excluding patients with a marked decline in LVEF during the first year (defined as ≥41% at admission and ≤40% at 1 year after discharge), patients with HFworEF (i.e. ≥50% at admission and 41–49% at 1 year after discharge) had worst clinical outcomes. We found that younger age, female sex, and lower LVEF at admission were independently associated with the HFimpEF phenotype in this study, which is consistent with previous reports. 7 , 10 , 11 , 12 Additionally, a previous study indicated that incorporating guideline‐directed medical therapy (GDMT) caused a higher improvement rate in LVEF, 18 , 19 implying that clinicians should be cautious in initiating or titrating GDMT and monitoring changes in LVEF in patients with HF. 5 , 6 GDMT should be offered promptly unless contraindicated. Moreover, this study revealed a significant association between ACE‐I/ARB use and the HFimpEF phenotype, further underscoring the significance of GDMT. Given that poor medical adherence is observed especially in older patients, there is an increasing focus on digital health‐based management systems that facilitate medication adherence. 20 Adopting these tools may benefit optimizing medication therapies.
Advanced age, low haemoglobin levels, and AF were independently associated with adverse clinical outcomes. Considering the close association of aging with comorbidities and frailty, 21 these findings suggest the clinical significance of frailty in the prognosis of patients with HF. Individualized approaches based on multidomain assessments (i.e. clinical, psychocognitive, functional, and social), such as cardiac rehabilitation, nutritional therapy, and comorbidity management, are warranted. 22 Despite the critical significance of frailty as a therapeutic target, there is insufficient evidence regarding interventions towards preventing or reversing frailty. 23 Further studies are required to address this challenge.
Our study had some limitations. First, as this was an observational study, managing patients with HF varied according to the discretion of the attending physicians. Second, this study included only patients who could be followed up for 1 year after discharge. Therefore, selection bias could not be excluded. Third, the participants were mostly Japanese. Considering that clinical management (e.g. implementation of GDMT) and race could affect LVEF improvement, 24 our findings may not apply to patients in Western countries. Fourth, as described above, the use of HF‐specific medications, particularly ARNI and SGLT2i, remained suboptimal for over 1 year after discharge. Therefore, we did not assess the effects of these medications on the improvement in LVEF.
Conclusions
Many patients with HF transitioned to different categories after acute decompensation. Furthermore, the clinical outcomes exhibited significant variations depending on the LVEF trajectories, irrespective of the HF phenotype. Notably, patients with a decreased EF experienced a worse prognosis, even with slight decreases (e.g. HFpEF transitioning to HFmrEF). Therefore, longitudinal assessments of systolic function are essential following hospitalization for HF.
Funding
This study was supported by a Grant‐in‐Aid for Young Scientists [Japan Society for the Promotion of Science (KAKENHI), 18K15860 (Y.S.), 23K15168 (Y.S.)], a Grant‐in‐Aid for Scientific Research (C) [23591062 (T.Y.), 26461088 (T.Y.), 16K09469 (Y.N.), 17K09526 (T.K.), 18K08056 (T.Y.), 20K08408 (T.K.), 20K08482 (Y.N.), 21K08142 (T.Y.), 23K07584 (T.K.)], a Grant‐in‐Aid for Scientific Research (B) [16H05215 (S.K.), 20H03915 (S.K.)], a Health Labour Sciences Research Grant [14528506 (S.K.)], the Sakakibara Clinical Research Grant for Promotion of Sciences (2012–20), a grant from the Japan Agency for Medical Research and Development [201439013C (S.K.)], and the Grant‐in‐Aid for Clinical Research from the Japanese Circulation Society [2019 (Y.S.)].
Conflict of interest
Dr Y.S. has received research grants from the SECOM Science and Technology Foundation and Uehara Memorial Foundation and honoraria from Otsuka Pharmaceutical Co., Ltd. and Ono Pharmaceutical Co., Ltd. Dr S.K. has received an unrestricted research grant from Novartis Pharmaceutical Co. and honoraria from Pfizer. The other authors declare no conflicts of interest.
Supporting information
Table S1. Incidence of clinical outcomes.
Table S2. Crude incidence of renal outcomes.
Table S3. Associations between patient characteristics and clinical outcomes.
Table S4. Changes in patient characteristics between discharge and 1 year follow up.
Figure S1. Kaplan–Meier curves for all‐cause death among the five groups.
Figure S2. Kaplan–Meier curves for heart failure rehospitalization among the five groups.
Figure S3. Kaplan–Meier curves for composite outcomes in a sensitivity analysis with excluding patients with a marked decline in LVEF during the first year.
Figure S4. Kaplan–Meier curves for all‐cause death in a sensitivity analysis with excluding patients with a marked decline in LVEF during the first year.
Figure S5. Kaplan–Meier curves for heart failure rehospitalization in a sensitivity analysis with excluding patients with a marked decline in LVEF during the first year.
Figure S6. Multivariable Cox regression analysis for the composite outcomes in a sensitivity analysis with excluding patients with a marked decline in LVEF during the first year.
Acknowledgements
The authors are grateful to the members of the WET‐HF investigators. Additionally, we would like to thank Editage (www.editage.jp) for English language editing.
Nakamaru, R. , Shiraishi, Y. , Kohno, T. , Nagatomo, Y. , Akiyama, H. , Motoya, Y. , Fukui, M. , Yajima, T. , Yoshikawa, T. , and Kohsaka, S. (2024) Treatment patterns and trajectories in patients after acute heart failure hospitalization. ESC Heart Failure, 11: 692–701. 10.1002/ehf2.14635.
References
- 1. James SL, Abate D, Abate KH, Abay SM, Abbafati C, Abbasi N, et al. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: A systematic analysis for the Global Burden of Disease Study 2017. Lancet 2018;392:1789‐1858. doi: 10.1016/S0140-6736(18)32279-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Ferreira JP, Kraus S, Mitchell S, Perel P, Piñeiro D, Chioncel O, et al. World Heart Federation roadmap for heart failure. Glob Heart 2019;14:197‐214. doi: 10.1016/j.gheart.2019.07.004 [DOI] [PubMed] [Google Scholar]
- 3. Wong M, Staszewsky L, Latini R, Barlera S, Glazer R, Aknay N, et al. Severity of left ventricular remodeling defines outcomes and response to therapy in heart failure. J Am Coll Cardiol 2004;43:2022‐2027. doi: 10.1016/j.jacc.2003.12.053 [DOI] [PubMed] [Google Scholar]
- 4. Bozkurt B, Coats AJ, Tsutsui H, Abdelhamid M, Adamopoulos S, Albert N, et al. Universal definition and classification of heart failure. J Card Fail 2021;27:387‐413. doi: 10.1016/j.cardfail.2021.03.003 [DOI] [PubMed] [Google Scholar]
- 5. McDonagh TA, Metra M, Adamo M, Gardner RS, Baumbach A, Böhm M, et al. 2021 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J 2021;42:3599‐3726. doi: 10.1093/eurheartj/ehab368 [DOI] [PubMed] [Google Scholar]
- 6. Heidenreich PA, Bozkurt B, Aguilar D, Allen LA, Byun JJ, Colvin MM, et al. 2022 AHA/ACC/HFSA guideline for the management of heart failure: A report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation 2022;145:e895‐e1032. doi: 10.1161/CIR.0000000000001062 [DOI] [PubMed] [Google Scholar]
- 7. Lupón J, Gavidia‐Bovadilla G, Ferrer E, de Antonio M, Perera‐Lluna A, López‐Ayerbe J, et al. Dynamic trajectories of left ventricular ejection fraction in heart failure. J Am Coll Cardiol 2018;72:591‐601. doi: 10.1016/j.jacc.2018.05.042 [DOI] [PubMed] [Google Scholar]
- 8. Basuray A, French B, Ky B, Vorovich E, Olt C, Sweitzer NK, et al. Heart failure with recovered ejection fraction: Clinical description, biomarkers, and outcomes. Circulation 2014;129:2380‐2387. doi: 10.1161/CIRCULATIONAHA.113.006855 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Kalogeropoulos AP, Fonarow GC, Georgiopoulou V, Burkman G, Siwamogsatham S, Patel A, et al. Characteristics and outcomes of adult outpatients with heart failure and improved or recovered ejection fraction. JAMA Cardiol 2016;1:510‐518. doi: 10.1001/jamacardio.2016.1325 [DOI] [PubMed] [Google Scholar]
- 10. Tsuji K, Sakata Y, Nochioka K, Miura M, Yamauchi T, Onose T, et al. Characterization of heart failure patients with mid‐range left ventricular ejection fraction—A report from the CHART‐2 Study: Characterization of HFmrEF. Eur J Heart Fail 2017;19:1258‐1269. doi: 10.1002/ejhf.807 [DOI] [PubMed] [Google Scholar]
- 11. Savarese G, Vedin O, D'Amario D, Uijl A, Dahlström U, Rosano G, et al. Prevalence and prognostic implications of longitudinal ejection fraction change in heart failure. JACC Heart Fail 2019;7:306‐317. doi: 10.1016/j.jchf.2018.11.019 [DOI] [PubMed] [Google Scholar]
- 12. Albert J, Lezius S, Störk S, Morbach C, Güder G, Frantz S, et al. Trajectories of left ventricular ejection fraction after acute decompensation for systolic heart failure: Concomitant echocardiographic and systemic changes, predictors, and impact on clinical outcomes. J Am Heart Assoc 2021;10:e017822. doi: 10.1161/JAHA.120.017822 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Perrino C. Intermittent pressure overload triggers hypertrophy‐independent cardiac dysfunction and vascular rarefaction. J Clin Invest 2006;116:1547‐1560. doi: 10.1172/JCI25397 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Kitakata H, Kohno T, Kohsaka S, Shiraishi Y, Parizo JT, Niimi N, et al. Prognostic implications of early and midrange readmissions after acute heart failure hospitalizations: A report from a Japanese multicenter registry. J Am Heart Assoc 2020;9:e014949. doi: 10.1161/JAHA.119.014949 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Nakamaru R, Shiraishi Y, Niimi N, Kohno T, Nagatomo Y, Takei M, et al. Phenotyping of elderly patients with heart failure focused on noncardiac conditions: A latent class analysis from a multicenter registry of patients hospitalized with heart failure. J Am Heart Assoc 2023;12:e027689. doi: 10.1161/JAHA.122.027689 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Maddox TM, Januzzi JL, Allen LA, Breathett K, Butler J, Davis LL, et al. 2021 update to the 2017 ACC expert consensus decision pathway for optimization of heart failure treatment: Answers to 10 pivotal issues about heart failure with reduced ejection fraction. J Am Coll Cardiol 2021;77:772‐810. doi: 10.1016/j.jacc.2020.11.022 [DOI] [PubMed] [Google Scholar]
- 17. Matsuo S, Imai E, Horio M, Yasuda Y, Tomita K, Nitta K, et al. Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis 2009;53:982‐992. doi: 10.1053/j.ajkd.2008.12.034 [DOI] [PubMed] [Google Scholar]
- 18. Udelson JE, Feldman AM, Greenberg B, Pitt B, Mukherjee R, Solomon HA, et al. Randomized, double‐blind, multicenter, placebo‐controlled study evaluating the effect of aldosterone antagonism with eplerenone on ventricular remodeling in patients with mild‐to‐moderate heart failure and left ventricular systolic dysfunction. Circ Heart Fail 2010;3:347‐353. doi: 10.1161/CIRCHEARTFAILURE.109.906909 [DOI] [PubMed] [Google Scholar]
- 19. Wilcox JE, Fang JC, Margulies KB, Mann DL. Heart failure with recovered left ventricular ejection fraction. J Am Coll Cardiol 2020;76:719‐734. doi: 10.1016/j.jacc.2020.05.075 [DOI] [PubMed] [Google Scholar]
- 20. Guasti L, Dilaveris P, Mamas MA, Richter D, Christodorescu R, Lumens J, et al. Digital health in older adults for the prevention and management of cardiovascular diseases and frailty. A clinical consensus statement from the ESC Council for Cardiology Practice/Taskforce on Geriatric Cardiology, the ESC Digital Health Committee and the ESC Working Group on e‐Cardiology. ESC Heart Fail 2022;9:2808‐2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Richter D, Guasti L, Walker D, Lambrinou E, Lionis C, Abreu A, et al. Frailty in cardiology: Definition, assessment and clinical implications for general cardiology. A consensus document of the Council for Cardiology Practice (CCP), Association for Acute Cardio Vascular Care (ACVC), Association of Cardiovascular Nursing and Allied Professions (ACNAP), European Association of Preventive Cardiology (EAPC), European Heart Rhythm Association (EHRA), Council on Valvular Heart Diseases (VHD), Council on Hypertension (CHT), Council of Cardio‐Oncology (CCO), Working Group (WG) Aorta and Peripheral Vascular Diseases, WG e‐Cardiology, WG Thrombosis, of the European Society of Cardiology, European Primary Care Cardiology Society (EPCCS). Eur J Prev Cardiol 2022;29:216‐227. doi: 10.1093/eurjpc/zwaa167 [DOI] [PubMed] [Google Scholar]
- 22. Gorodeski EZ, Goyal P, Hummel SL, Krishnaswami A, Goodlin SJ, Hart LL, et al. Domain management approach to heart failure in the geriatric patient. J Am Coll Cardiol 2018;71:1921‐1936. doi: 10.1016/j.jacc.2018.02.059 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Ijaz N, Buta B, Xue QL, Mohess DT, Bushan A, Tran H, et al. Interventions for frailty among older adults with cardiovascular disease. J Am Coll Cardiol 2022;79:482‐503. doi: 10.1016/j.jacc.2021.11.029 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. DeVore AD, Hellkamp AS, Thomas L, Albert NM, Butler J, Patterson JH, et al. Improvement in left ventricular ejection fraction in outpatients with heart failure with reduced ejection fraction: Data from CHAMP‐HF. Circ Heart Fail 2020;13:e006833. doi: 10.1161/CIRCHEARTFAILURE.119.006833 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. Incidence of clinical outcomes.
Table S2. Crude incidence of renal outcomes.
Table S3. Associations between patient characteristics and clinical outcomes.
Table S4. Changes in patient characteristics between discharge and 1 year follow up.
Figure S1. Kaplan–Meier curves for all‐cause death among the five groups.
Figure S2. Kaplan–Meier curves for heart failure rehospitalization among the five groups.
Figure S3. Kaplan–Meier curves for composite outcomes in a sensitivity analysis with excluding patients with a marked decline in LVEF during the first year.
Figure S4. Kaplan–Meier curves for all‐cause death in a sensitivity analysis with excluding patients with a marked decline in LVEF during the first year.
Figure S5. Kaplan–Meier curves for heart failure rehospitalization in a sensitivity analysis with excluding patients with a marked decline in LVEF during the first year.
Figure S6. Multivariable Cox regression analysis for the composite outcomes in a sensitivity analysis with excluding patients with a marked decline in LVEF during the first year.