Currently, cardiovascular disease ranks as the leading cause of mortality globally. Acute ST-segment elevation myocardial infarction (STEMI) remains one of the most serious acute manifestations of coronary artery disease [1]. Over the past few decades, significant advancements have been made in the management of STEMI, particularly with the widespread accessibility of primary percutaneous coronary intervention (PPCI). However, the in-hospital mortality rate for STEMI patients still ranges between 5% and 8%, and the 1-year mortality rate can be as high as 14.3% [2]. Following myocardial infarction, cardiomyocytes experience impaired energy metabolism. Concurrently, inflammatory responses, oxidative stress, ischemia-reperfusion injury, myocardial hypertrophy, and fibrosis occur. These factors collectively result in abnormal myocardial remodelling and facilitate the progression of heart failure (HF) [3]. The typical clinical manifestations include dyspnea, pulmonary rales, peripheral edema, and elevated B-type natriuretic peptide (BNP) levels. Therefore, the early identification of high-risk features of HF is crucial for improving the prognosis of STEMI patients. Growth stimulator gene 2 (ST2) belongs to the interleukin (IL)-1 receptor family and has two subtypes: transmembrane (ST2L) and soluble (sST2). ST2L binds to IL-33 and has cardioprotective effects. These effects mainly include anti-myocardial fibrosis, inhibition of cardiac hypertrophy, reduction of apoptosis, and improvement of cardiac function. During HF, the secretion of sST2 increases. It then competitively binds to IL-33, thereby reducing the cardioprotective effects of the ST2L-IL-33 complex. Both domestic and international HF guidelines have suggested [4, 5] that sST2, an indicator of myocardial fibrosis, is valuable for the risk stratification and prognostic assessment of HF patients. However, there is a paucity of information regarding the use of sST2 to predict the development of Killip class II-IV during hospitalisation in STEMI patients treated with PPCI. Therefore, the objective of this study was to investigate the predictive efficacy of sST2 for the development of Killip class II-IV during the hospitalisation of STEMI patients.

In this study, by analyzing the relationship between clinical indicators and the development of Killip class II–IV during hospitalisation in STEMI patients treated with PPCI, we demonstrated that sST2 was an independent predictor of in-hospital Killip class II–IV development in STEMI patients. Moreover, the predictive efficacy of the combination of sST2, first-visit HR, LVEF, and eGFR was superior to that of any single indicator.

Compared to other biomarkers such as NT-proBNP, the advantage of sST2 lies in its concentration being unaffected by age, renal function, intravascular volume, BMI, or atrial fibrillation [8]. This relative independence from common heart failure comorbidities implies its potential superiority in the prediction of heart failure [8]. NT-proBNP is a traditional and internationally recognized biomarker for heart failure but was not included in the final predictive model of this study. We considered two reasons for the lower-than-expected admission NT-proBNP levels in our study patients, which may have compromised its ability to objectively predict heart failure outcomes. First, the BMI of STEMI patients included in this study was generally high. The median BMI for all patients was 24.87 kg/m², with non-heart failure patients having a median BMI of 25.16 kg/m² and heart failure patients a median BMI of 24.14 kg/m² (no statistically significant difference between the groups). A BMI between 24.0 and 27.9 is classified as overweight. A study has confirmed that NT-proBNP levels are inversely correlated with BMI [9], possibly due to hemodilution, increased degradation of NT-proBNP by adipose tissue, and alterations in ventricular function. Second, acute myocardial infarction (AMI) patients often experience significant diaphoresis due to severe pain before hospitalisation or reduced oral intake, leading to intravascular volume depletion upon admission. Consequently, myocardial mechanical stress may not be significantly increased, resulting in normal or only mildly elevated NT-proBNP levels in the early hospital phase.

The IL-33-ST2L pathway can immunologically inhibit the development of atherosclerosis via helper T-cell 2 and macrophage 2 phenotype responses. Conversely, high levels of sST2 may promote plaque progression [10]. Therefore, sST2 can be regarded as a marker for both early and late post-infarction remodelling. Jenkins et al. [11] classified a cohort of 1401 AMI patients into three cardiovascular risk classes according to early sST2 values. Class 2 (37 sST2 $\le$ 72.3 ng/mL) and class 3 (sST2 $>$72.3 ng/mL) were associated with a higher risk of death within the first 30 days and during the first 5-year follow-up. AMI patients with sST2 $>$72.3 ng/mL are more prone to the activation of neurohormonal and fibrotic signalling pathways, which increases the risk of adverse myocardial remodelling and heart failure [11]. For AMI patients during hospitalisation, sST2 values can guide discharge decisions, and a 30% reduction in sST2 values at discharge is recommended [12]. In this study, sST2 was measured early (within 24 hours) in STEMI patients treated with PPCI. Based on the ROC curve, the optimal cut-off value for sST2 was 77.3 ng/mL. This was confirmed by one-way logistic regression analysis, LASSO regression analysis, and multifactorial logistic regression analysis, indicating that sST2 $>$77.3 ng/mL is an independent risk factor for the development of HF in STEMI patients. The 77.3 ng/mL value obtained in this study is close to the 72.3 ng/mL value derived by Jenkins et al. [11], further validating the accuracy and reliability of this study.

Acute kidney injury (AKI) and chronic kidney disease (CKD) are indicators of poor prognosis in AMI. As renal function deteriorates, numerous metabolic pathways are disrupted. These include alterations in cardiac volume and pressure status, accelerated atherosclerosis, electrolyte metabolism disorders, the presence of uremic toxins, and oxidative stress, all of which can reduce cardiac function and lead to heart failure [13]. The present study revealed that eGFR levels were significantly lower in the heart failure group compared to the non-heart failure group, and renal insufficiency was an independent risk factor for in-hospital heart failure in STEMI patients. This is consistent with previous studies. Yandrapalli et al. [14] found that among 237,549 AMI survivors, 13.8% had concurrent AKI, 16.5% had concurrent CKD, and 3.4% had concurrent end-stage renal disease (ESRD). Moreover, in-hospital heart failure was more prevalent among AMI patients with renal insufficiency. In patients with renal insufficiency, elevated fibroblast growth factor levels are associated with an increased risk of left ventricular hypertrophy, which is related to diastolic dysfunction, HF, and death [15].

It is widely recognized that the admission heart rate is an important predictor of in-hospital mortality in patients with acute coronary syndrome (ACS). Jensen et al. [16] proposed that a heart rate $>$80 bpm in ACS patients should be considered a marker of adverse events during hospitalisation, regardless of the ACS type. In this study, patients in the HF group had a significantly higher initial heart rate than those in the non-HF group. Based on the optimal cut-off value calculated from the ROC curve, a heart rate $>$94 beats/min was an independent risk factor for in-hospital heart failure in STEMI patients treated with PPCI.

Our study has several limitations. We acknowledge that Killip class II–IV serves as a clinical severity indicator rather than a definitive HF diagnosis, which ideally requires comprehensive assessment incorporating biomarkers (BNP/NT-proBNP) and imaging studies. This may introduce diagnostic misclassification bias. As a single-center study with limited sample size, external validation through multicenter cohorts is needed. Unmeasured confounders (e.g., socioeconomic factors, medication adherence) were not systematically assessed. In future studies, we plan to implement this risk stratification tool in clinical practice to guide personalized treatment decisions and ultimately improve patient outcomes.

This study verifies that sST2 is an independent predictor for the development of in-hospital Killip class II-IV in STEMI patients. The combination of sST2, first-visit HR, LVEF, and eGFR can enhance the predictive value of Killip class II–IV development in STEMI patients, facilitating early risk stratification and has the potential to improve outcomes in these patients.

The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

YZ, GYD, YS and RJ designed the research study. YZ, FD and YFZ performed data collection. YZ, YQZ and MG analyzed the data. YZ, GYD and RJ drafted the manuscript. All authors contributed to critical revision of the manuscript for important intellectual content. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

The study was carried out in accordance with the guidelines of the Declaration of Helsinki and approved by the ethics committee of the TEDA International Cardiovascular Hospital (ethical approval number: [2023]-0310-1). Due to the retrospective design of the study, the requirement for written informed consent from patients was waived.

National Key Clinical Specialty Construction Project — “Breaking Barriers” Action Construction Project for the Improvement of Medical Quality in Tianjin; Construction Project of Key Medical Disciplines (Specialties) in Tianjin - Science (Cardiology) - TJYXZDXK - 020A; Demonstration Project for the Reform and High - quality Development of Public Hospitals in Tianjin Binhai New Area; Science and Technology Project of Tianjin Binhai New Area Health Commission (2022BWKQ006).

The authors declare no conflict of interest.