From April 2014 to May 2018, 484 patients presented to our tertiary center with acute chest pain suspected of being ACS, which developed within 6 hours. The initial hs-TnT (or 0-hour hs-TnT) was tested routinely. To estimate the interval change in cardiac enzymes, hs-TnT was re-checked 3 hours after presentation and defined as the 3-hour hs-TnT. The study flowchart is shown in Fig. 1. The following patients were excluded from the study: (1) those with duration of chest pain > 6 hours; (2) those with possible ‘rule-in’ MI; (3) those with medical conditions that could affect Tn levels (e.g., systemic infection, acute cerebrovascular accident, tachyarrhythmia, renal disease, heart failure, or other acute medical conditions rather than ACS); and (4) those who previously underwent an interventional cardiac procedure. Patients with a 0-hour hs-TnT ≥ 0.1 ng/mL or who presented with electrocardiogram findings suggesting ST-segment elevation were excluded because they were identified as a ‘rule-in’ MI. Patients with a history of coronary artery disease, previous percutaneous coronary intervention (PCI), or previous coronary artery bypass graft were also excluded. The patients who required cardiopulmonary resuscitation during admission, were in cardiogenic shock, or had chronic heart failure were excluded because these conditions may affect the initial Tn levels. Patients who did not undergo coronary angiography (CAG) were excluded. Ultimately, the study enrolled 163 patients. These were subdivided into hs-cTnT positive (n = 32) and negative (n = 131) interval change groups. This study was a prospective, non-randomized, observational study. The study protocol was approved by the Chonnam National University Hospital Institutional Review Board (IRB No. CNUH-2014-124). Informed consent was obtained from all study participants.

Blood samples were collected in serum-separating tubes. Samples were obtained from 484 patients immediately after emergency room visits. The hs-TnT level was measured by Roche e411 (Roche Diagnostics, Rotkreuz, Switzerland). An interval change in serial hs-TnT was defined as follows: (1) 3-hour hs-TnT ≥ 0.014 ng/mL and more than 50% greater than 0-hour hs-TnT in patients with a 0-hour hs-TnT < 0.014 ng/mL; or (2) 3-hour hs-TnT ≥ 0.014 ng/mL and more than 20% greater than 0-hour hs-TnT in patients with a 0-hour hs-TnT ≥ 0.014 ng/mL. The current cutoff value for the hs-TnT assay is 0.014 ng/mL [9]. Although the 2015 European Society of Cardiology (ESC) guidelines recommend invasive management for patients with acute chest pain if 3-hour hs-TnT is greater than 0.014 ng/mL [5], we defined positive or negative interval change groups according to an older guideline that subdivided patients with 3-hour hs-TnT ≥ 0.014 ng/mL into two groups [16]. The clinical manifestations of NSTE-ACS reference the 2015 ESC guidelines, as follows: (1) prolonged (over 20 minutes) angina at rest; (2) new-onset class II or III angina of the Canadian Cardiovascular Society (CCS) classification; (3) recent destabilization of previously stable angina with at least CCS class III angina (crescendo angina); and (4) post-MI angina. The term non-ST-segment elevation myocardial infarction (NSTEMI) was used when there was evidence of myocardial necrosis (defined as an elevated troponin value with at least one value above the 99th percentile upper reference limit) and the aforementioned clinical features [5]. In this study, hs-TnT was used as a cardiac biomarker to determine whether the subjects had unstable angina or MI. In this study, significant coronary artery stenosis was confirmed by CAG and defined as follows: (1) ≥ 50% diameter stenosis identified in the left main coronary artery at CAG; or (2) ≥ 70% diameter stenosis identified in coronary arteries other than the left main coronary artery at CAG. Multi-vessel disease was determined when there were significant coronary stenoses in two or more epicardial coronary arteries. A bifurcation lesion was defined as narrowing of a coronary artery adjacent to or involving the origin of a significant side branch [17]. Vasospasm was defined as a > 90% reduction in diameter on the ergonovine provocation test or a spontaneous, dynamic > 90% decrease in diameter observed during CAG.

The medical and social histories were obtained using a questionnaire. The conventional cardiac risk factors included were hypertension, diabetes, dyslipidemia, stroke, and smoking status. Body data such as height (m) and weight (kg) were measured by experienced nurses and body mass index (BMI, kg/m²) was calculated. Blood was tested in all subjects on admission, including white blood cell count, hemoglobin level, platelet count, serum glucose, serum creatinine, and hs-TnT. The hs-TnT was measured at presentation and 3 hours later. All study subjects underwent echocardiography and CAG. The left ventricular ejection fraction was considered an indicator of ventricular systolic function on transthoracic echocardiography. Patients who had an interval change underwent CAG between admission and hospital discharge, while patients who were negative for an interval change underwent CAG within 1 to 2 months of admission via a hospital visit. The coronary arteries were assessed quantitatively and qualitatively using CAG. If significant coronary artery stenosis was found at CAG, PCI was performed simultaneously.

All statistical analyzes were performed using SPSS for Windows version 25.0 (IBM Co., Armonk, NY, USA). Continuous variables are expressed as the mean ± standard deviation. Discrete (categorical) variables are expressed as percentages and counts. Student’s t test was used to analyze the difference between two independent groups of continuous variables. In the analysis of categorical variables, the difference between the observed and expected frequencies was confirmed by Pearson’s chi-square test. To demonstrate the impact of the interval change in hs-TnT on the presence of significant coronary artery stenosis, the odds ratio (OR) and 95% confidence interval (CI) were calculated using univariable and multivariable logistic regression analysis. The ORs were adjusted according to the variables that had statistical significance (p < 0.05) in the univariable analysis in the multivariable model.

Table 1 compares the baseline clinical characteristics. There was no significant difference between the two groups in age (60.09 ± 10.40 years vs. 57.79 ± 11.74 years, p = 0.312), proportion ≥ 65 years old (34.4% vs. 28.2%, p = 0.495), or proportion of males (71.9% vs. 65.6%, p = 0.502). The positive interval change group had higher 0-hour hs-TnT level (0.020 ± 0.014 ng/mL vs. 0.011 ± 0.008 ng/mL, p = 0.001), proportion of 0-hour hs-TnT ≥ 0.014 ng/mL (56.3% vs. 16.0%, p = 0.001), and 3-hour hs-TnT level (0.150 ± 0.206 ng/mL vs. 0.010 ± 0.007 ng/mL, p = 0.050) than the negative interval change group. The laboratory tests did not differ significantly, except for the serum glucose level (151.90 ± 57.92 ng/mL vs. 115.07 ± 29.68 mg/dL, p = 0.031). There was no difference in conventional cardiac risk factors such as hypertension, diabetes mellitus, and dyslipidemia in both groups.

Table 2 summarizes the angiographic findings of both groups. There was a marked difference in the presence of significant coronary artery stenosis (68.8% vs. 23.7%, p = 0.001). There was a higher proportion of reduced initial thrombolysis in myocardial infarction (TIMI) flow (i.e., initial TIMI flow 0, I, and II) in the positive interval change group (53.1% vs. 13.7%, p < 0.001). The positive interval change group had a higher rate of bifurcation lesions (12.5% vs. 2.3%, p = 0.028), but a lower proportion of vasospastic components (6.3% vs. 31.3%, p = 0.003). A higher proportion of the positive interval change group underwent PCI (65.6% vs. 22.9%, p < 0.001).

The univariable and multivariable analyses of predictors of significant coronary stenosis are described in Tables 3 and 4, respectively. In the univariable analyses, age, 0-hour hs-TnT, 3-hour hs-TnT, interval change of serial hs-TnT, serum glucose level and history of diabetes mellitus were associated with significant coronary stenosis (Table 3). Multivariable analysis was performed on these significant variables, and age (adjusted OR, 1.054; 95% CI, 1.013 to 1.097; p = 0.024), interval change of serial hs-TnT (adjusted OR, 10.196; 95% CI, 3.571 to 29.111; p = 0.001), and history of diabetes mellitus (adjusted OR, 3.390; 95% CI, 1.216 to 9.451; p = 0.013) were identified as independent predictors of the presence of significant coronary artery stenosis (Table 4).