From February 2019 to June 2019, 103 adult patients diagnosed with ICM admitted to Hunan Provincial People's Hospital were prospectively enrolled in this study. ICM was defined according to the following criteria: (1) clear evidence of myocardial ischemia or myocardial infarction. At least one of the following conditions need to be fulfilled: 1. coronary arteriography demonstrates more than 75% stenosis in the left main coronary artery or left anterior descending artery or two vessels or more with greater than 75% stenosis, meanwhile, resting electrocardiogram/dynamic electrocardiogram /electrocardiogram treadmill exercise test or echocardiography/dobutamine stress echocardiography indicates myocardial ischemia, 2. patients underwent coronary revascularization or had a history of myocardial infarction[12]; (2) enlargement of the left ventricle due to myocardial ischemia or myocardial infarction (echocardiography revealed that left ventricular end-diastole diameter ≥ 50 mm for women or ≥ 55 mm for men); and (3) clinical manifestations of heart failure. Patients with the following conditions were excluded: (1) ventricular septal perforation, ventricular aneurysm or papillary muscle dysfunction leading to mitral regurgitation; (2) women who were pregnant; and (3) patients with a history of cancer. This research was conducted in compliance with the Declaration of Helsinki and was approved by the Ethics Committee of Hunan Provincial People's Hospital (2018-50). Informed consent was obtained from all enrolled patients.

Basic demographic data, such as age, sex, biochemical tests and New York Heart Association Functional Classification (NYHA-FC), were collected at baseline. The Modification of Diet in Renal Disease (MDRD) formula was used to calculate the estimated glomerular filtration rate (eGFR). eGFR(ml/min/1.73m²) = 30,849 × [Scr(μmol/L)]^−1.154 × (age)^−0.203 × (0.742 if patient is female). NT-proBNP was determined at baseline using the chemiluminescence immunoassay method (Wantaikairui, XiaMen, China) in the Department of Laboratory Medicine of Hunan Provincial People's Hospital.

All patients underwent echocardiographic examination within 3 days after admission using GE Vivid E9 machines with a 3.5 MHz probe (General Electric Company, U.S.). Echocardiographic examination was performed according to the American Society of Echocardiography recommendations [13]. A parasternal long-axis M-mode image was acquired to measure the left ventricular end diastolic diameter (LVEDD). The left ventricular ejection fraction (LVEF) was measured by a modified version of Simpson's biplane method. left ventricular diastolic function was determined using a combination of mitral spectral pulse-wave Doppler and tissue Doppler of the mitral annulus according to the American Society of Echocardiography recommendations [14]. A dobutamine stress echocardiogram was performed to evaluate whether myocardial ischemia existed in some patients.

Patients were followed up in the first, third, sixth, and twelfth months and every 6 months following discharge from the hospital. The end point events were defined as the time from the date of blood sampling to cardiovascular death and/or heart failure-related rehospitalization. The follow-up time was calculated from the time of blood sampling to February 15, 2020.

Peripheral venous blood samples were collected the following morning after admission. Blood samples were centrifuged at 1000 rpm for 15 min, and the upper serum was frozen and stored at -80℃. Serum levels of HE4 were measured using a chemiluminescence microparticle immunoassay (Abbott, Germany) in the Department of Nuclear Medicine of Hunan Provincial People's Hospital. The detectable dose of human HE4 ranged from 15 to 1500 pmol/L. The investigator was blinded with respect to clinical data.

The data are presented as the mean ± standard deviation (\(\overline{x} \pm s\)) for continuous variables with normal distribution, as interquartile ranges (IQRs) for continuous variables with nonnormal distribution, and as percentages for categorical variables. Statistical differences between two groups were analyzed using an independent Student’s t-test for normally distributed variables and the Mann–Whitney U test for nonnormally distributed variables. Chi-square test was used for categorical variables. Pearson or Spearman correlation coefficients were used as appropriate to explore correlations between HE4 and other variables. The optimal cutoff value of HE4 was determined by receiver operating characteristic (ROC) curve analysis. Univariate Cox regression analysis was used to explore the predictive value of each variable for events, and then a forward stepwise multivariate Cox regression model was further utilized to explore independent predictors. Kaplan–Meier plots with log-rank tests were used to illustrate outcomes of patients grouped by the optimal cutoff value of HE4. ROC analysis was performed using Med Calc 13.0 (Med Calc Software), and other statistical analyses were performed using SPSS version 19.0 (SPSS, Inc.). A 2-sided P < 0.05 was considered statistically significant.

A total of 103 patients with ICM were enrolled, 7 of whom were lost to follow-up and were excluded from the analysis. Among the remaining 96 patients, 63 (65.6%) were males. The median age was 71 (65–77) years. The median left ventricular ejection fraction (LVEF) was 33 (29–42)%. Twenty-eight patients (29.2%) presented with NYHA-FC II, 40 patients (41.7%) presented with NYHA-FC III symptoms, and 28 patients (29.2%) presented with NYHA-FC IV symptoms. The median follow-up was 263 (153–313) days. Forty-five patients (47%) experienced events, 12 patients died, and 33 patients were hospitalized due to heart failure (Table 1). Three patients received cardiac resynchronization therapy, and 2 patients received implantable cardioverter defibrillators.

As NYHA-FC increased, HE4 levels became significantly higher (NYHA-FC II: 75.30 pmol/L, IQR 57.18–97.75; NYHA-FC III: 121.90 pmol/L, IQR 98.25–174.85; NYHA-FC IV: 203.60 pmol/L, IQR 121.93–317.18) (Fig. 1). HE4 levels correlated positively with NT-proBNP (r = 0.602, P < 0.001) and negatively with eGFR (r = − 0.697, P < 0.001). No significant correlation was observed between HE4 and left ventricular ejection fraction (r = − 0.082, P = 0.426).

Patients with events had a similar sex distribution and body mass index (BMI) compared to patients without events. There was no difference in complications or treatment regimen using angiotensin-converting enzyme inhibitor (ACEI), angiotensin receptor blocker (ARB), angiotensin receptor-neprilysin inhibitor (ARNI), β-blockers or percutaneous coronary intervention (PCI) between patients with and without events. However, patients with events exhibited less favorable NYHA-FC and renal function, older age, and higher NT-proBNP and HE4.

In univariate analysis, HE4, NT-proBNP, eGFR, NYHA-FC and age were all significant predictors of events (Table 2). Multivariable forward stepwise Cox analysis was performed including HE4, NT-proBNP, eGFR, NYHA-FC and age. The results showed that HE4 [χ²: 9.602, hazard ratio (HR): 1.003, 95% confidence interval (CI): 1.001–1.005, P = 0.002] and age [χ², 4.557; HR: 1.044, 95% CI: 1.003–1.085, P = 0.033] were independent predictors of events (Table 3). The forward stepwise and backward stepwise analyses yielded the same results.

The ROC analysis showed that HE4 [area under the curve (AUC) = 0.800] was the best predictor for events, followed by NT-proBNP (AUC = 0.766), age (AUC = 0.753), and NYHA-FC (AUC = 0.678) (Table 4 and Fig. 2).

Patients with HE4 > 100.2 pmol/L exhibited a worse survival rate than those with HE4 ≤ 100.2 pmol/L (log-rank rest: 17.662, p < 0.001). The unadjusted HR was 4.826 (95% CI: 2.148–10.846), and the HR after adjustment for age and sex was 3.372 (95% CI: 1.409–8.065). The Kaplan–Meier curves are shown in Fig. 3.