Introduction
Despite modern reperfusion strategies and evidence-based pharmacotherapy, acute myocardial infarction (AMI) causes pathological left ventricular remodelling (LVR) and heart failure (HF), and therefore leads to adverse cardiovascular (CV) outcomes.1-4 AMI represents a major cause of HF.1,5-7 According to a Polish nationwide database of acute coronary syndromes (AMI-PL database), HF is one of the most frequent causes of recurrent hospitalization in patients after AMI in 1-year follow-up,8 indicating a need for an intensification of secondary prevention programs (including cardiac rehabilitation, smoking cessation, and improvement of risk stratification).3,4,9 Therefore, it is important to explore new sensitive and specific biomarkers that could help identify patients at risk of developing HF and adverse CV outcomes after AMI.
Galectin-3 (Gal-3) and soluble interleukin-1 receptor-like 1 (sST2) are promising biomarkers involved in LVR, resulting from inflammatory processes and fibrosis.10-14 There is evidence for a high prognostic value of both biomarkers in predicting outcomes in patients with chronic and acute HF.15 Biomarkers of myocardial fibrosis (including sST2 and Gal-3) were recommended as useful tools for additional risk stratification in the American guidelines for the management of HF.16
However, studies evaluating the role of Gal-3 and sST2 and their relationship with adverse outcomes in patients after AMI are insufficient. Therefore, we aimed to evaluate the association of Gal-3 and sST2 in patients with first-time ST-segment elevation myocardial infarction (STEMI) treated with primary percutaneous coronary intervention (PCI) with in-hospital and 1-year CV outcomes.
Patients and methods
Study population
The analysis was based on data collected in a prospective observational BIOSTRAT (Biomarkers for Risk Stratification After STEMI; ClinicalTrials.gov identifier, NCT03735719) study. The BIOSTRAT study included 117 consecutive white patients with first-time STEMI treated with primary PCI in the 1st Department of Cardiology, Medical University of Warsaw from October 2014 to April 2017. STEMI was diagnosed in accordance with the applicable guidelines.17 Main inclusion criteria were age 18 years or older and first-time STEMI treated with primary PCI. Main exclusion criteria were previous AMI, pre-existing HF (history of LV ejection fraction [LVEF] <50% or diagnosed HF with preserved LVEF), severe renal dysfunction (plasma creatinine level >220 mmol/l (approx. 2.5 mg/dl), and/or creatinine clearance <30 ml/min), severe liver disease, chronic inflammatory disease, current neoplastic disease, and life expectancy less than 1 year.
Informed written consent was obtained from each study participant. The trial protocol complied with the Declaration of Helsinki and was approved by the local ethics committee of the Medical University of Warsaw (decision no. KB/97/2014).
Baseline tests and measurements
Routine laboratory parameters, including complete blood count, glycemia, lipid profile, electrolytes, serum creatinine, and biomarkers such as cardiac troponin I, creatine kinase myocardial band, high-sensitive C-reactive protein (hs-CRP) and N-terminal pro-B-type natriuretic peptide (NT-proBNP) were measured using standard methods in the hospital laboratory. The concentration of NT-proBNP was measured using the Roche Elecsys 1010 analyzer (Roche Diagnostics, Mannheim, Germany). The concentration of hs-CRP was assessed using Cobas Integra 800 (Roche Diagnostics).
The highest concentrations of cardiac troponin I, NT-proBNP, and hs-CRP were subsequently included in the analyses.
Electrocardiogram, transthoracic echocardiogram and clinical examination were performed in each patient during index hospitalization. Demographic data, details on previous and current pharmacotherapy, clinical and angiographic characteristics, and medical history were also prospectively gathered.
Gal-3 and sST2 measurements
Additionally, blood samples were collected from all recruited patients for further measurements of serum Gal-3 and sST2 concentrations. To avoid the potential impact of PCI on biomarker concentration, we sampled blood after 72 to 96 hours after hospital admission, as per previous studies.18,19 Serum was prepared from blood samples by allowing the blood to clot for 60 min, followed by centrifugation at 3500 rpm for 15 min. Serum samples were then stored at –80°C. Measurements were performed after all patients had entered the study. Serum concentrations of Gal-3 were measured using Human Galectin-3 Quantikine ELISA Kit (BIOKOM, Janki, Poland), and plasma sST2, using Presage ST2 Assay (Genloxa, Puck, Poland).
Study endpoints
Patients were followed for 12 months. The primary endpoint was CV death or hospitalization for HF during 1-year follow-up. CV death was defined as deaths related to AMI, HF, sudden cardiac death, or stroke. Hospitalization for HF was considered as hospitalization with a primary diagnosis of HF supported by evidence of clinical signs of HF (rales, peripheral edema), pulmonary congestion on a chest radiograph, or a need for intravenous diuretics.
Secondary endpoints concerned in-hospital outcomes including 1) total length of index hospitalization, 2) length of stay in intensive cardiac care unit during index hospitalization, 3) in-hospital death; and events that occurred in 1-year follow-up including 4) CV death, 5) hospitalization for HF, and 6) MI.
Statistical analysis
For a between-group comparison, we used the Fisher exact test and the Mann–Whitney test for categorical and continuous variables, respectively. Categorical data were presented as numbers and percentages of patients. Normally distributed continuous data were expressed as mean (SD), while non-normally distributed continuous data were presented as median with an interquartile range (IQR). Pearson and Spearman correlation coefficients were used for parametric and nonparametric variables, respectively. The Cox proportional hazards regression model was performed to identify predictors of the primary endpoint. Receiver operating characteristic (ROC) curves were plotted for baseline Gal-3 and sST2 in relation to the primary endpoint. In addition, the Youden J statistic was performed to determine the optimal biomarker cutoff point for the prediction of the primary endpoint. A P value of less than 0.05 was considered significant. All tests were 2-tailed. SPSS software, version 22 (IBM SPSS Statistics 22, New York, New York, United States) was used for analysis.
Results
Baseline characteristics
A total of 117 consecutive first-time STEMI patients were included in the study. The median (IQR) age was 61.0 (50.5–67.0) years and 70% of patients were men. Median (IQR) baseline LVEF was 48% (41%–53%). Median (IQR) Gal-3 and sST2 concentrations were 7.1 (5.6–8.8) ng/ml and 23.4 (18.0–32.0) ng/ml, respectively.
Correlation analysis with baseline parameters
We also performed correlation analysis of baseline Gal-3 and sST2 concentrations with clinical parameters (Table 1). A correlation was found between Gal-3 and sST2 levels. Both Gal-3 and sST2 correlated positively with stay at an intensive cardiac care unit, NT-proBNP, and hs-CRP and inversely with glomerular filtration rate. Gal-3 correlated positively with age, Killip class, as well as the Thrombolysis in Myocardial Infarction (TIMI) and the Global Registry of Acute Coronary Events (GRACE) scores. sST2 correlated negatively with sodium level on admission and inversely with baseline and final TIMI Coronary Grade Flow. There was a nonsignificant correlation between sT2 and age.
Variable | Baseline Gal-3 | Baseline sST2 | ||
---|---|---|---|---|
Rho | P value | Rho | P value | |
Baseline characteristics | ||||
Age, y | 0.38 | <0.001 | 0.09 | 0.32 |
BMI, kg/m2 | 0.14 | 0.15 | 0.01 | 0.93 |
Clinical status, laboratory, and angiographic findings on admission | ||||
Killip class | 0.20 | 0.03 | 0.10 | 0.27 |
Hemoglobin, g/dl | –0.26 | 0.004 | –0.14 | 0.12 |
hs-CRP, mg/dl | 0.18 | 0.05 | 0.19 | 0.047 |
Troponin I, ng/ml | –0.45 | 0.64 | 0.03 | 0.74 |
NT-proBNP, pg/ml | 0.36 | 0.001 | 0.38 | <0.001 |
Gal-3, ng/ml | – | – | 0.48 | 0.04 |
sST2, ng/ml | 0.48 | 0.04 | – | – |
eGFR, ml/min/1.73 m2 | –0.30 | 0.001 | –0.11 | 0.23 |
Serum sodium, mmol/l | –0.09 | 0.36 | –0.20 | 0.03 |
Total cholesterol, mg/dl | –0.23 | 0.01 | –0.24 | 0.01 |
LDL, mg/dl | –0.12 | 0.07 | –0.19 | 0.059 |
TIMI score | 0.34 | <0.001 | 0.12 | 0.19 |
GRACE score | 0.38 | <0.001 | 0.10 | 0.26 |
Baseline TIMI grade flow | 0.10 | 0.31 | – 0.16 | 0.04 |
Final TIMI grade flow | –0.05 | 0.63 | –0.16 | 0.045 |
Echocardiographic and laboratory findings at discharge | ||||
Hemoglobin, g/dl | –0.19 | 0.05 | –0.03 | 0.78 |
Serum creatinine, mg/dl | 0.13 | 0.18 | 0.19 | 0.054 |
eGFR, ml/min/1.73 m2 | –0.36 | <0.001 | –0.24 | 0.01 |
Outcomes | ||||
Hospitalization length, d | 0.18 | 0.055 | 0.35 | <0.001 |
Time in ICCU, d | 0.35 | <0.001 | 0.45 | <0.001 |
Abbreviations: BMI, body mass index; eGFR, estimated glomerular filtration rate; Gal-3, galectin-3; GRACE, Global Registry of Acute Coronary Events; hs-CRP, high-sensitive C-reactive protein; ICCU, intensive cardiac care unit; LDL, low-density lipoprotein; NT-proBNP, N-terminal fragment of the prohormone brain natriuretic peptide; sST2, soluble interleukin‐1 receptor‐like 1; TIMI, Thrombolysis in Myocardial Infarction |
One-year follow-up
Data on survival was available for all patients (5 out of 117 patients [4.3%] died). Thirteen patients were lost to follow-up in terms of hospitalization for HF, leaving 104 patients (89%) for the composite primary endpoint analyses at 1 year. During 1-year follow-up, 9 patients (8.7%) reached the primary endpoint (3 died due to CV causes and 6 were hospitalized for HF). Among CV causes of death, HF-related death occurred in 2 patients and AMI-related death, in 1 patient (including 1 patient who died during index hospitalization after recruitment).
Patients who experienced the primary endpoint during follow-up had higher levels of baseline Gal-3 and sST2 (for sST2 nonsignificant). Baseline characteristics of patients who reached and who did not reach the primary endpoint at 1 year are presented in Supplementary material, Table S1.
Comparison of baseline characteristics, laboratory findings, clinical presentations, in-hospital and 1-year outcomes in patients with high and low Gal-3 and sST2 levels
ROC analysis revealed that the area under the curve (AUC) for Gal-3 and sST2 (for prediction of the primary endpoint) was 0.85 and 0.64, respectively (Figure 1). Gal-3 concentration of 9.57 ng/ml or higher had a sensitivity of 41%, a specificity of 91%, a negative predictive value of 92%, and a positive predictive value of 36% for prediction of the primary endpoint at follow-up (the Youden index). sST2 concentration of 45.99 ng/ml or higher had a sensitivity of 44%, a specificity of 97%, a negative predictive value of 95%, and a positive predictive value of 57% for prediction of the primary endpoint at follow-up (the Youden index).
Baseline characteristics, laboratory findings, clinical presentations, in-hospital and 1-year outcomes of patients with high and low Gal-3 and sST2 levels depending on established cutoffs are presented in Tables 2 and 3. Twenty-four patients (20.5%) had baseline Gal-3 levels equal to or above the upper limit of the established cutoff value of 9.57 ng/ml. Eight patients (6.8%) had baseline sST2 levels equal to or above the upper limit of the established cutoff value of 45.99 ng/ml. Higher concentrations of both biomarkers (above the cutoffs) were associated with a longer hospital stay (including a longer stay in intensive cardiac care unit) and more frequent occurrence of the primary endpoint. Patients with a Gal-3 level above the cutoff also experienced CV death more frequently.
Variable | sST2 <45.99 ng/ml (n = 109) | sST2 ≥45.99 ng/ml (n = 8) | P value | Gal-3 <9.57 ng/ml (n = 93) | Gal-3 ≥9.57 ng/ml (n = 24) | P value | |
---|---|---|---|---|---|---|---|
Variable | sST2 <45.99 ng/ml (n = 109) | sST2 ≥45.99 ng/ml (n = 8) | P value | Gal-3 <9.57 ng/ml (n = 93) | Gal-3 ≥9.57 ng/ml (n = 24) | P value | |
Baseline characteristics | |||||||
Age, y | 60 (50.5–67) | 69.5 (51–78.8) | 0.19 | 59 (50–64.5) | 66.5 (55.3–78.8) | 0.01 | |
Male sex, n (%) | 77 (70.6) | 5 (62.5) | 0.69 | 70 (75.3) | 12 (50) | 0.02 | |
BMI, kg/m2 | 28.4 (24.6–30.5) n = 98 | 33.3 (29.3–34.7) n = 7 | 0.07 | 28.1 (24.4–30.4) n = 87 | 30.0 (28.6–33.7) n = 18 | 0.02 | |
Moderate valve disease, n (%) | 4 (3.7) | 1 (12.5) | 0.30 | 3 (3.2) | 2 (8.3) | 0.27 | |
Hypertension, n (%) | 65 (59.6) | 6 (75.0) | 0.48 | 55 (59.1) | 16 (66.7) | 0.64 | |
Atrial fibrillation, n (%) | 3 (2.8) | 2 (25) | 0.04 | 2 (2.2) | 3 (12.5) | 0.058 | |
Diabetes, n (%) | 21 (19.3) | 2 (25) | 0.66 | 17 (18.3) | 6 (25) | 0.57 | |
Chronic kidney disease, n (%) | 15 (13.8) | 5 (62.5) | 0.04 | 12 (12.9) | 8 (33.3) | 0.03 | |
COPD, n (%) | 4 (3.7) | 3 (37.5) | 0.01 | 4 (4.3) | 3 (12.5) | 0.15 | |
Prior stroke or TIA, n (%) | 4 (3.7) | 2 (25) | 0.054 | 3 (3.2) | 3 (12.5) | 0.10 | |
Peripheral artery disease, n (%) | 4 (3.7) | 3 (37.5) | 0.01 | 2 (2.2) | 5 (20.8) | 0.004 | |
Current or former smoking, n (%) | 80 (73.4) | 6 (75) | 1.00 | 70 (75.3) | 16 (66.7) | 0.44 | |
Clinical status and laboratory findings on admission | |||||||
Heart rate, bpm | 80 (70–90) | 90 (71–98.8) | 0.24 | 80 (70–90) | 80 (70–89) | 0.80 | |
SBP, mm Hg | 130 (120–141) | 142.5 (120–168.5) | 0.18 | 130 (120–140) | 132.5 (120–149.8) | 0.49 | |
DBP, mm Hg | 77 (70–90) | 83 (62.5–94.8) | 0.52 | 77 (70–90) | 80 (62.5–85) | 0.93 | |
Intravenous diuretics, n (%) | 33 (30.3) | 6 (75) | 0.02 | 26 (28) | 13 (54.2) | 0.03 | |
Killip class | 1 (1–2) | 2 (1–3) | 0.01 | 1 (1–1) | 2 (1–2) | <0.001 | |
TIMI score | 3 (2–5) | 5 (3–7) | 0.02 | 3 (2–4) | 5 (3–7) | <0.001 | |
GRACE score | 110 (95–127) | 135 (100.0–171) | 0.17 | 109 (94–123) | 132 (102–158) | 0.002 | |
Laboratory findings on admission | |||||||
Hemoglobin, g/dl | 14.3 (13.5–15.6) | 13.2 (12.2–13.8) | 0.02 | 14.3 (13.6–15.7) | 13.7 (12.6–14.3) | 0.01 | |
hs-CRP peak, mg/dl | 3.2 (1.7–7.4) n = 104 | 91.6 (3.2–208) | 0.01 | 3 (1.5–6.7) n = 88 | 7 (2.2–43.2) | 0.03 | |
Troponin I peak, ng/l | 31.91 (3.73–82.31) n = 105 | 40.04 (16.92–102.07) | 0.37 | 29.63 (3.58–82.31) n = 89 | 47.26 (16.92–95.67) | 0.12 | |
CK-MB peak, U/l | 74.55 (12.68–178.68) n = 108 | 116 (48.25–428.20) | 0.26 | 71.05 (7.13–173.20) n = 92 | 112.70 (55.95–241.18) | 0.07 | |
NT-proBNP peak, pg/ml | 886 (329–1945) n = 75 | 7194 (2343.3–14701.5) | <0.001 | 884 (283.8–1926.5) n = 64 | 3520 (762–5986) n = 19 | 0.003 | |
Serum creatinine, mg/dl | 0.93 (0.86–1.06) | 1.15 (0.87–1.31) | 0.12 | 0.94 (0.86–1.06) | 0.95 (0.81–1.22) | 0.90 | |
eGFR, ml/min/1.73 m2 | 92.8 (64.2–117.9) | 84.3 (58.1–107.9) | 0.30 | 96.3 (68.8–117.9) | 74.1 (58.4–102.2) | 0.06 | |
Serum sodium, mmol/l | 140.0 (138.4–141.9) | 138.5 (135.4–139.8) | 0.03 | 140 (137.9–142) | 139.5 (138.7–141.2) | 0.34 | |
Serum potassium, mmol/l | 3.9 (3.6–4.2) | 4 (3.7–4.3) | 0.53 | 3.97 (3.62–4.18) | 3.94 (3.55–4.38) | 0.76 | |
Total cholesterol, mg/dl | 187 (162–223) n = 105 | 145.5 (121.3–162.8) | 0.002 | 188 (163.8–226.5) n = 90 | 153 (140–188) n = 23 | 0.004 | |
LDL, mg/dl | 113 (84–145) n = 95 | 77 (61.8–95.3) | 0.01 | 115.5 (85–148) n = 80 | 84 (75–121) n = 23 | 0.03 | |
HDL, mg/dl | 44.5 (34.3–52) n = 104 | 43.5 (15.5–58.8) | 0.78 | 45 (35–53.5) n = 89 | 40 (33–50) n = 23 | 0.34 | |
Triglycerides, mg/dl | 136 (93–177) n = 103 | 103 (85.5–227.3) | 0.67 | 135.5 (91.3–181.3) n = 88 | 134 (91–157) n = 23 | 0.63 | |
Angiographic characteristics | |||||||
Infarct-related artery, n (%) | RCA | 46 (42.2) | 5 (62.5) | 0.29 | 39 (41.9) | 12 (50.0) | 0.50 |
LAD | 48 (44.0) | 3 (37.5) | 1.00 | 42 (45.2) | 9 (37.5) | 0.65 | |
Cx | 15 (13.8) | 0 | 0.59 | 12 (12.9) | 3 (12.5) | 1.00 | |
Extent of CAD, n (%) | 1-vessel | 64 (58.7) | 3 (37.5) | 0.28 | 55 (59.1) | 12 (50.0) | 0.49 |
2-vessel | 28 (25.7) | 5 (62.5) | 0.04 | 24 (25.8) | 9 (37.5) | 0.31 | |
3-vessel | 17 (15.6) | 0 | 0.60 | 14 (15.1) | 3 (12.5) | 1.00 | |
Angiographic characteristics | |||||||
TIMI grade flow | Baseline | 0 (0–1) | 0 (0–0) | 0.12 | 0 (0–1) | 0 (0–1) | 0.35 |
Final | 3 (3–3) | 3 (2–3) | 0.13 | 3 (3–3) | 3 (3–3) | 0.52 | |
Stent implantation, n (%) | 1 stent | 72 (66.1) | 6 (75.0) | 1.00 | 61 (65.6) | 17 (70.8) | 0.81 |
≥2 stents | 31 (28.4) | 2 (25.0) | 1.00 | 27 (29.0) | 6 (25.0) | 0.80 | |
Complete reva- scularization | 63 (57.8) | 3 (37.5) | 0.29 | 55 (59.1) | 11 (45.8) | 0.26 | |
Echocardiography | |||||||
Ejection fraction, % | 48 (42–53) | 38 (29–49) | 0.03 | 48 (41–54) | 46 (35–51) | 0.15 | |
LVEDD, mm | 4.8 (4.5–5.2) n = 108 | 5.3 (4.5–5.4) | 0.19 | 4.8 (4.5–5.2) n = 92 | 5.0 (4.3–5.4) | 0.62 | |
LVEDV, ml | 104 (80–127) | 130 (65–131) | 0.99 | 105 (79–124) | 112 (74–131) | 0.86 | |
LVESV, ml | 53 (40–70) | 86 (32–95) | 0.46 | 53 (41–70) | 59 (34–92) | 0.77 | |
LVHa, n (%) | 31 (32.6) | 3 (50.0) | 0.40 | 24 (28.6) | 10 (58.8) | 0.02 | |
LA dimension, mm | 3.8 (3.5–4.2) n = 108 | 4.1 (3.6–5.2) | 0.18 | 3.8 (3.5–4.2) n = 92 | 4.1 (3.6–4.2) | 0.08 | |
Clinical status and laboratory findings at discharge | |||||||
Heart rate, bpm | 70.0 (64.0–76.0) n = 108 | 80.0 (72.8–84.5) | 0.01 | 70.0 (64.0–80.0) | 71.0 (67.0–80.0) n=23 | 0.43 | |
SBP, mm Hg | 120.0 (110.0–130.8) n = 108 | 132.0 (103.8–138.8) | 0.55 | 120.0 (110.0–130.0) | 130.0 (105.0–141.0) n = 23 | 0.41 | |
DBP, mm Hg | 73.5 (60.0–80.00) n = 108 | 80.0 (65.5–80.00) | 0.64 | 75.0 (62.5–80.0) | 70.0 (60.0–80.00) n = 23 | 0.49 | |
Hemoglobin, g/dl | 13.7 (12.6–14.7) n = 99 | 12.9 (11.0–14.5) | 0.17 | 14.0 (12.7–14.8) n = 86 | 12.9 (11.5–14.2) n = 21 | 0.01 | |
Serum creatinine, mg/dl | 0.93 (0.81–1.10) n = 97 | 1.24 (0.83–1.45) | 0.054 | 0.95 (0.81–1.05) n = 84 | 0.93 (0.81–1.25) n = 21 | 0.42 | |
eGFR, ml/min/1.73 m2 | 91.36 (74.75–120.25) n = 97 | 57.14 (37.94–121.55) | 0.10 | 92.59 (78.17–112.06) n = 84 | 65.07 (54.60–106.38) n = 21 | 0.01 | |
Serum sodium, mmol/l | 141.3 (139.5–143.3) n = 97 | 140.8 (137.3–142.9) | 0.47 | 141.0 (139.5–143.3) n = 84 | 141.6 (139.4–143.0) n = 21 | 0.64 | |
Serum potassium, mmol/l | 4.4 (4.1–4.6) n = 97 | 4.6 (4.1–5.0) | 0.34 | 4.45 (4.19–4.69) n = 84 | 4.3 (4.1–4.7) n = 21 | 0.34 | |
Pharmacotherapy at hospital dischargeb | |||||||
ASA, n (%) | 108 (100) | 8 (100) | 1.00 | 93 (100) | 23 (100) | 1.00 | |
Clopidogrel, n (%) | 95 (88.0) | 7 (87.5) | 1.00 | 83 (89.2) | 19 (82.6) | 0.47 | |
Ticagrelor, n (%) | 13 (12.0) | 1 (12.5) | 1.00 | 10 (10.8) | 4 (17.4) | 0.47 | |
Anticoagulants, n (%) | 7 (6.5) | 2 (25.0) | 0.12 | 6 (6.5) | 3 (13.0) | 0.38 | |
Loop diuretics, n (%) | 24 (22.2) | 5 (62.5) | 0.02 | 18 (19.4) | 11 (47.8) | 0.01 | |
ACEI, n (%) | 104 (96.3) | 8 (100) | 1.00 | 89 (95.7) | 23 (100) | 0.58 | |
ARB, n (%) | 6 (5.6) | 1 (12.5) | 0.40 | 6 (6.5) | 1 (4.3) | 1.00 | |
β-Blocker, n (%) | 101 (93.5) | 8 (100) | 1.00 | 87 (93.5) | 22 (95.7) | 1.00 | |
Aldosterone antagonist, n (%) | 37 (34.3) | 1 (12.5) | 0.27 | 32 (34.4) | 6 (26.1) | 0.62 | |
Ivabradine, n (%) | 2 (1.9) | 0 | 1.00 | 2 (2.2) | 0 | 1.00 | |
Statin, n (%) | 104 (96.3) | 7 (87.5) | 0.31 | 89 (95.7) | 22 (95.7) | 1.00 | |
Data presented as median (IQR) unless otherwise indicated. a LVH was based on LVMI: LVMI >95 g/m2 for women, LVMI >115 g/m2 for men; b In patients who survived to hospital discharge (n = 108) SI conversion factors: to convert hemoglobin to g/l, multiply by 100; hs-CRP to nmol/l, by 95.24; troponin to μg/l, by 1; CK-MB to μ/l, by 0.0167; serum creatinine to μmol/l, by 76.25; total cholesterol and LDL and HDL cholesterol to mmol/l, by 0.0259; triglycerides to mmol/l, by 0.0113. Abbreviations: ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; ASA, acetylsalicylic acid; CAD, coronary artery disease; CK-MB, creatine kinase-muscle/brain; COPD, chronic obstructive pulmonary disease; Cx, circumflex artery; DBP, diastolic blood pressure; HDL, high-density lipoprotein; LA, left atrium; LAD, left anterior descending artery; LDL, low-density lipoprotein; LVEDD, left ventricular end-diastolic diameter; LVEDV, left ventricular end-diastolic volume; LVESV, left ventricular end-systolic volume; LVH, left ventricular hypertrophy; LVMI, left ventricular mass index; RCA, right coronary artery; SBP, systolic blood pressure; TIA, transient ischemic attack; others, see Table 1 |
Variable | sST2 <45.99 ng/ml (n = 109) | sST2 ≥45.99 ng/ml (n = 8) | P value | Gal-3 <9.57 ng/ml (n = 93) | Gal-3 ≥9.57 ng/ml (n = 24) | P value |
---|---|---|---|---|---|---|
In-hospital outcomes | ||||||
Hospitalization length, median (IQR), d | 8.0 (7.0–10.0) | 15.5 (12.3–24.8) | <0.001 | 8.0 (6.5–10.0) | 9.5 (7.3–13.8) | 0.03 |
Time in ICCU, median (IQR), d | 4.0 (3.0–5.0) | 11.0 (7.5–17.0) | <0.001 | 3.0 (3.0–5.0) | 5.5 (4.3–10.8) | <0.001 |
In-hospital death, n(%) | 1 (0.9) | 0 | 1.00 | 0 | 1 (4.2) | 0.21 |
1-year outcomes | ||||||
CV death or HF hospitalization, n (%) | 5 (5.2) n = 97 | 4 (57.1) n = 7 | 0.001 | 2 (2.4) n = 83 | 7 (33.3) n = 21 | <0.001 |
CV death, n (%) | 2 (1.8) | 1 (12.5) | 0.19 | 0 | 3 (12.5) | 0.01 |
HF hospitalization, n (%) | 3 (3.1) n = 97 | 4 (57.1) n = 7 | <0.001 | 2 (2.4) n = 83 | 5 (23.8) n = 21 | 0.003 |
MI, n (%) | 4 (4.1) n = 97 | 0 n = 7 | 1.00 | 2 (2.4) n = 83 | 2 (9.5) n = 21 | 0.18 |
Abbreviations: CV, cardiovascular; HF, heart failure; MI, myocardial infarction; others, see Table 2 |
Predictors of the primary endpoint
In the univariate Cox proportional hazards regression analysis, both baseline Gal-3 and sST2 (as continuous variables, as well as their newly-established cutoffs) were predictors of the primary endpoint (CV death or HF hospitalization), and of HF hospitalizations (Table 4). Furthermore, in contrast to sST2, Gal-3 predicted CV death.
Variable | HR (95% CI) | P value |
---|---|---|
Age, per 10 years | 1.97 (1.02–3.71) | 0.04 |
NT-proBNP, per 1000 pg/ml | 1.14 (1.04–1.25) | 0.01 |
Gal-3, per 1 ng/ml | 1.34 (1.17–1.54) | <0.001 |
Gal-3 ≥9.57 ng/ml | 15.94 (3.31–76.82) | 0.001 |
sST2, per 10 ng/ml | 1.63 (1.22–2.16) | 0.001 |
sST2 ≥45.99 ng/ml | 12.62 (3.37–47.20) | <0.001 |
Abbreviations: HR, hazard ratio; others, see Table 1 |
Gal-3 and sST2 remained significant predictors of the primary endpoint even after adjustment for age and NT-proBNP in multivariate analyses (Figure 2, Table 5).
Variable | Multivariate analyses | ||
---|---|---|---|
HR | 95% CI | P value | |
Age, per 10 years | 1.22 | 0.74–2.16 | 0.44 |
Gal-3 ≥9.57 ng/ml | 8.65 | 1.45–51.70 | 0.02 |
sST2 ≥45.99 ng/ml | 3.15 | 0.72–13.81 | 0.13 |
Age, per 10 years | 1.22 | 0.66–2.16 | 0.62 |
Gal-3 ≥9.57 ng/ml | 14.51 | 1.46–143.95 | 0.02 |
NT-proBNP, per 1000 pg/ml | 1.05 | 0.94–1.25 | 0.36 |
Age, per 10 years | 1.34 | 0.74–2.59 | 0.36 |
sST2 ≥45.99 ng/ml | 11.79 | 1.52–91.26 | 0.02 |
NT-proBNP, per 1000 pg/ml | 0.99 | 0.85–1.16 | 0.95 |
Abbreviations: see Tables 1 and 4 |
Discussion
In our study, concentrations of both Gal-3 and sST2 were higher in those with worse clinical presentation at baseline. Both Gal-3 and sST2 were associated with unfavorable in-hospital outcomes and were independent predictors of CV death or hospitalization for HF in 1-year follow-up (when regarded as continuous variables). However, if the newly-established cutoffs of both biomarkers were included in one multivariate model (together with age), only Gal-3 remained an independent predictor of the primary endpoint. Moreover, Gal-3 also predicted CV death alone. Furthermore, there was no difference in sST2 concentration between those who did and did not reach the primary endpoint during follow-up. Therefore, Gal-3 might be preferable to ST2 in risk stratification after STEMI.
In our study there was a small number of events, but the results are in line with a recent study assessing the same composite endpoint in patients with a first anterior STEMI treated with pPCI. In this study, 20 out of 103 patients (19.4%) died or were admitted for HF within a 6-month follow-up. Gal-3, measured within 48 hours after STEMI, significantly predicted the composite endpoint after adjustment for age, gender, renal and ventricular function as well as troponin and NT-proBNP values.20
Gal-3 plays an important role in various biological processes, but the most acknowledged role of Gal-3 is participation in fibrosis.10,21,22 Gal-3 is produced by activated macrophages that stimulate inflammation and proliferation of myofibroblasts and collagen deposition.10 Sanchez-Mas et al23 observed based on experimental data that Gal-3 increases in myocardium after AMI with the maximum concentration achieved in the infarcted area during the first week, with a gradual decrease over the following weeks. It appears that the increase in concentration of Gal-3 in the early phase after myocardial infarction contributes to the activation of repair functions in the damaged zone in order to maintain the geometry and function of the heart. However, in a longer perspective, chronic activation leads to tissue fibrosis and accelerates adverse LVR.23
The ST2 molecule is a soluble glycoprotein belonging to the interleukin-1 receptor family, and is secreted by inflammatory cells, cardiomyocytes, and endothelium.11 ST2 has 2 clinically relevant isoforms—transmembrane (ST2 ligand) and soluble (sST2) circulating in the bloodstream.11,24 The balance between these 2 forms of ST2 guarantees an appropriate biological effect. Elevated sST2 triggers myocardial fibrosis.11,24 In an experimental study, sST2 concentrations increased steadily after AMI with maximum expression on the first day.11
It is also known that Gal-3 and interleukin 33/ST2 pathways are involved in the pathogenesis of atherosclerosis, in which the inflammatory substrate is one of the main causes of instability of atherosclerotic plaques.24,25 Tsai et al26 showed that Gal-3 levels were significantly higher in patients with AMI than in healthy controls. In our study, we observed lower median Gal-3 concentrations (7.1 ng/ml) in patients following first-time AMI than was reported previously by Szadkowska et al27 and van der Velde et al28 (13.0 and 13.4 ng/ml, respectively). However, these differences can be explained by more restrictive exclusion criteria related to potential fibrosis processes (ie, exclusion of patients with neoplasms, advanced chronic kidney disease, previous HF) in our study. We also only enrolled patients with STEMI, excluding patients with non–ST-segment elevation myocardial infarction. In addition, previously, it has also been shown that elevated Gal-3 and sST2 concentrations were observed in patients with hypertension, diabetes, prior AMI, and prior HF—factors which may bias the biomarkers’ measurements between studies.19,29
The first studies on Gal-3 and sST2 were in the field of HF and showed that higher levels of circulating Gal-3 and sST2 were associated with worse prognosis in those patients.15,29,30 The American Heart Association recommendations have even considered Gal-3 and sST2 to be valuable prognostic markers in acute and chronic HF (class IIb recommendation, level of evidence B).16 The Food and Drug Administration approved threshold values of 17.8 ng/ml for Gal-3 and 35.0 ng/ml for sST2 for additional risk stratification in patients with chronic and acute HF.28 However, there is still a need to assess the clinical utility and specific cutoffs of both biomarkers to help clinicians conduct better risk stratification in patients with AMI but without previous HF. Only few studies have described the impact of Gal-3 and sST2 on outcomes after AMI.
Subanalysis of the PROVE IT-TIMI 22 (Pravastatin or Atorvastatin Evaluation and Infection Therapy-Thrombolysis in Myocardial Infarction 22) trial showed that Gal-3 concentrations above median value of 16.7 μg/l, measured within 7 days after acute coronary syndrome (100 patients after AMI or unstable angina), were associated with a higher risk of HF development in a 2-year follow-up.19 In our study, the cutoff Gal-3 level for the primary endpoint occurrence was equal to or above 9.57 ng/ml. This Gal-3 threshold was associated with a worse clinical condition during index hospitalization, as well as a higher rate of unfavorable outcomes during index hospitalization and follow-up. Similarly, the study performed by Tsai et al26 (196 patients with first-time STEMI) revealed that a Gal-3 level equal to or above 7.67 ng/ml was the most powerful predictor of death and development of HF in a 30-day postinfarction period. Moreover, this association was observed regardless of the severity of coronary artery lesions, LVEF, and serum creatinine.26 Of note, in our study patients with higher Gal-3 concentrations had higher risk in GRACE and TIMI scores, which identify high-risk patients after MI.
In our study, median sST2 concentration in the study group was 23.4 ng/ml, while in a study by Jenkins et al29 (1401 patients after AMI) the median sST2 value was 48.7 ng/ml. However, their study relied on heterogeneous diagnoses—79% of patients had non–ST-segment elevation myocardial infarction. Increased concentrations of sST2 were associated with an increased risk of death and HF development during 5 years of follow-up, independently of other prognostic factors.29 Liu et al31 showed that sST2 concentrations above 58.7 ng/ml have highest specificity in predicting either MACEs (defined as the composite of all-cause death, HF, and nonfatal AMI) or mortality at 1-year after STEMI. In our study, the cutoff sST2 value of 45.99 ng/ml identified patients at high risk of the primary endpoint.
Another analysis based on the BIOSTRAT study assessed the association of Gal-3 and sST2 and changes in their concentrations after 1 year with the development of HF which showed that baseline Gal-3 and sST2 concentrations have higher clinical value than measurements obtained after 1 year.32
In the CORONA study, Gal-3 was not correlated with worse prognosis after adjusting for NT-proBNP in older patients with ischemic chronic HF, hence Gal-3 may have limited the application of risk stratification in older patients.33 However, our study showed that this does not apply to patients after STEMI. Gal-3 (but not sST2) correlated with age but both Gal-3 and sST2 were independent predictors of the primary endpoint even after adjusting for age.
In our study, there was no association between sST2 and Gal-3 concentrations and maximum troponin I. An explanation for this is that these biomarkers are involved in distinct pathophysiological pathways than that which are already known. AMI provokes an inflammatory response with the migration of a multitude of cells and regulators into the infarcted and noninfarcted areas. This process initiates reparative changes in the early phase after AMI.2 This could be the reason for higher values of hs-CRP in both groups of sST2 and Gal-3 upper cutoffs values. However, chronic activation of these processes leads to tissue fibrosis, adverse LVR, and development of HF. Szadkowska et al27 found that elevated Gal-3 (>16 ng/ml) concentrations during hospitalization in patients after AMI were associated with a higher risk of HF and atrial fibrillation. In our study, patients with higher levels of Gal-3 and sST2 were more likely to have a higher Killip class on admission, and more frequently required diuretics intravenously during hospitalization and orally at discharge. Consequently, elevated Gal-3 and sST2 concentrations were associated with increased risk of combined endpoint—CV death and HF hospitalization—as well as the risk of HF hospitalization itself.
Therefore, it may be concluded that increased levels of Gal-3 and sST2 after AMI reflect myocardial damage and may help in the early identification of patients at higher risk of HF development. It may have important implications for postdischarge follow-up and highlights the need for therapies which impact adverse cardiac remodeling, such as angiotensin-converting enzyme inhibitors, aldosterone receptor blockers or mineralocorticoid receptor antagonists. There are several studies showing the potential advantageous influence of renin-angiotensin-aldosterone antagonists, as well as genetic therapies on the reduction of biomarker levels.34-38 Other mediators of inflammatory processes (ie, leukotrienes) have recently been intensively studied for their role in coronary artery disease. Further studies will show if novel anti-inflammatory strategies added to conventional therapy will reduce cardiovascular risk.39
Limitations
The main limitations of our study relate to the small sample size and relatively short follow-up which translated into a small number of events. Additionally, 13 patients (11.1% of the 117 patients) were lost to follow-up in terms of hospitalization for HF. Therefore, in order to maintain an adequate events per predictor variable value, we were not able to include more potentially significant variables in the Cox proportional hazards regression model.28 Moreover, a certain proportion of data (including NT-proBNP concentrations) for some of the patients was missing (as indicated in tables). Still, given the correlation of both Gal-3 and sST2 with NT-proBNP, as well as the fact that NT-proBNP concentration predicted the primary endpoint, we decided it was important to include NT-proBNP in our multivariate models. However, we also performed a separate multivariate analysis including age and both of the studied biomarkers (Gal-3 and sST2), but not NT-proBNP. Gal-3 and sST2 remained significant predictors of the primary endpoint in all those models.
Further studies with longer follow-up are still needed to determine the predictive value and clinical utility of serum levels of sST2 and Gal-3 in patients with AMI.
Conclusion
In patients with first-time STEMI treated with primary PCI, Gal-3 and sST2 predicted CV death or hospitalization for HF at 1 year. Concentrations of both biomarkers above the established cutoffs (≥9.57 ng/ml for Gal-3 and ≥45.99 ng/ml for ST2) were associated with worse clinical presentation at baseline, as well as adverse in-hospital and 1-year outcomes. Assessment of these 2 biomarkers of inflammation and fibrosis may play an important role in CV risk stratification after AMI; however, Gal-3 may be considered a more preferable option.
Agnieszka Kapłon-Cieślicka, MD, PhD, 1st Department of Cardiology, Medical University of Warsaw, ul. Banacha 1a, 02-097 Warszawa, Poland, phone: +48225992958, email: agnieszka.kaplon@gmail.com
September 7, 2019.
October 23, 2019.
October 23, 2019.
The BIOSTRAT study is supported by 2 grants from Medical University of Warsaw (1WR/NM2/14, to AT; 1WR/NM4/16, to AW).
All authors made substantial contributions to the concept and design of the study. AT, AKC, KO, MB, AW, PS, and JM researched data, conducted data interpretation. AT, KO, and MP performed statistical analysis. AT, KO, and AKC wrote manuscript. All authors reviewed the manuscript and approved its final version.
None declared.
Tymińska A, Kapłon-Cieślicka A, Ozierański K, et al. Association of galectin-3 and soluble ST2 with in-hospital and 1-year outcomes in patients with ST-segment elevation myocardial infarction treated with primary percutaneous coronary intervention. Pol Arch Intern Med. 2019; 129: 770-780. doi:10.20452/pamw.15030
- 1.
- Ponikowski P, Voors AA, Anker SD, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur J Heart Fail. 2016; 18: 891-975.Crossref
- 2.
- Sutton MG, Sharpe N. Left ventricular remodeling after myocardial infarction: pathophysiology and therapy. Circulation. 2000; 101: 2981-2988.Crossref
- 3.
- Trzeciak P, Gierlotka M, Poloński L, Gąsior M. Treatment and outcomes of patients under 40 years of age with acute myocardial infarction in Poland in 2009-2013: an analysis from the PL-ACS registry. Pol Arch Intern Med. 2017; 127: 666-673.Crossref
- 4.
- Karwowski J, Gierlotka M, Gąsior M, et al. Relationship between infarct artery location, acute total coronary occlusion, and mortality in STEMI and NSTEMI patients. Pol Arch Intern Med. 2017; 127: 401-411.Crossref
- 5.
- Kapłon-Cieślicka A, Tymińska A, Peller M, et al. Diagnosis, clinical course, and 1-year outcome in patients hospitalized for heart failure with preserved ejection fraction (from the Polish cohort of the European Society of Cardiology Heart Failure Long-Term Registry). Am J Cardiol. 2016; 118: 535-542.Crossref
- 6.
- Tymińska A, Kapłon-Cieślicka A, Ozierański K, et al. Anemia at hospital admission and its relation to outcomes in patients with heart failure (from the Polish cohort of 2 European Society of Cardiology Heart Failure Registries). Am J Cardiol. 2017; 119: 2021-2029.Crossref
- 7.
- Balsam P, Ozierański K, Kapłon-Cieślicka A, et al. Differences in clinical characteristics and 1-year outcomes of hospitalized patients with heart failure in ESC-HF Pilot and ESC-HF-LT registries. Pol Arch Intern Med. 2019; 129: 106-116.
- 8.
- Gierlotka M, Zdrojewski T, Wojtyniak B, et al. Incidence, treatment, in-hospital mortality and one-year outcomes of acute myocardial infarction in Poland in 2009-2012-nationwide AMI-PL database. Kardiol Pol. 2015; 73: 142-158.Crossref
- 9.
- Ozierański K, Witkowska A, Wojtyniak B, et al. Smoking ban in public places and myocardial infarction hospitalizations in a European country with high cardiovascular risk: insights from the Polish nationwide AMI-PL database. Pol Arch Intern Med. 2019; 129: 386-391.
- 10.
- Meijers WC, van der Velde AR, Pascual-Figal DA, Boer RA. Galectin-3 and post-myocardial infarction cardiac remodeling. Eur J Pharmacol. 2015; 763: 115-121.Crossref
- 11.
- Weinberg EO, Shimpo M, De Keulenaer GW, et al. Expression and regulation of ST2, an interleukin-1 receptor family member, in cardiomyocytes and myocardial infarction. Circulation. 2002; 106: 2961-2966.Crossref
- 12.
- Tomaniak M, Sygitowicz G, Filipiak KJ, et al. Dysregulations of miRNAs and galectin-3 may underlie left ventricular dilatation in patients with systolic heart failure. Kardiol Pol. 2018; 76: 1012-1014.Crossref
- 13.
- Sygitowicz G, Tomaniak M, Filipiak KJ, et al. Galectin-3 in patients with acute heart failure: preliminary report on first Polish experience. Adv Clin Exp Med. 2016; 25: 617-623.Crossref
- 14.
- Wiśniowska-Śmiałek S, Dziewięcka E, Holcman K, et al. Kinetics of selected serum markers of fibrosis in patients with dilated cardiomyopathy and different grades of diastolic dysfunction of the left ventricle. Cardiol J. 2018 Nov 28.
- 15.
- Bayes-Genis A, de Antonio M, Vila J, et al. Head-to-head comparison of 2 myocardial fibrosis biomarkers for long-term heart failure risk stratification: ST2 versus galectin-3. J Am Coll Cardiol. 2014; 63: 158-166.Crossref
- 16.
- Yancy CW, Jessup M, Bozkurt B, et al. 2017 ACC/AHA/HFSA focused update of the 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Failure Society of America. Circulation. 2017; 136: e137-e161.Crossref
- 17.
- Steg PG, James SK, Atar D, et al. ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation. Eur Heart J. 2012; 33: 2569-2619.Crossref
- 18.
- Di Tano G, Caretta G, De Maria R, et al. Galectin-3 predicts left ventricular remodelling after anterior-wall myocardial infarction treated by primary percutaneous coronary intervention. Heart. 2017; 103: 71-77.Crossref
- 19.
- Grandin EW, Jarolim P, Murphy SA, et al. Galectin-3 and the development of heart failure after acute coronary syndrome: pilot experience from PROVE IT-TIMI 22. Clin Chem. 2012; 58: 267-273.Crossref
- 20.
- Di Tano G, Caretta G, De Maria R, et al. Galectin-3 and outcomes after anterior-wall myocardial infarction treated by primary percutaneous coronary intervention. Biomark Med. 2018; 12: 21-26.Crossref
- 21.
- Yang RY, Rabinovich GA, Liu FT. Galectins: structure, function and therapeutic potential. Expert Rev Mol Med. 2008; 10: e17.Crossref
- 22.
- Ho JE, Liu C, Lyass A, et al. Galectin-3, a marker of cardiac fibrosis, predicts incident heart failure in the community. J Am Coll Cardiol. 2012; 60: 1249-1256.Crossref
- 23.
- Sanchez-Mas J, Lax A, Asensio-Lopez MC, et al. Galectin-3 expression in cardiac remodeling after myocardial infarction. Int J Cardiol. 2014; 172: e98-e101.Crossref
- 24.
- Kakkar R, Lee RT. The IL-33/ST2 pathway: therapeutic target and novel biomarker. Nat. Rev. Drug Discov. 2008; 7: 827-840.Crossref
- 25.
- Papaspyridonos M, McNeill E, de Bono JP, et al. Galectin-3 is an amplifier of inflammation in atherosclerotic plaque progression through macrophage activation and monocyte chemoattraction. Arterioscler Thromb Vasc Biol. 2008; 28: 433-440.Crossref
- 26.
- Tsai TH, Sung PH, Chang LT, et al. Value and level of galectin-3 in acute myocardial infarction patients undergoing primary percutaneous coronary intervention. J Atheroscler Thromb. 2012; 19: 1073-1082.Crossref
- 27.
- Szadkowska I, Wlazeł RN, Migała M, et al. The association between galectin-3 and clinical parameters in patients with first acute myocardial infarction treated with primary percutaneous coronary angioplasty. Cardiol J. 2013; 20: 577-582.Crossref
- 28.
- van der Velde AR, Lexis CP, Meijers WC, et al. Galectin-3 and sST2 in prediction of left ventricular ejection fraction after myocardial infarction. Clin Chim Acta. 2016; 452: 50-57.Crossref
- 29.
- Jenkins WS, Roger VL, Jaffe AS, et al. Prognostic value of soluble ST2 after myocardial infarction: a community perspective. Am J Med 2017; 130: 1112.e9-1112.e15.Crossref
- 30.
- van der Velde AR, Gullestad L, Ueland T, et al. Prognostic value of changes in galectin-3 levels over time in patients with heart failure: data from CORONA and COACH. Circ Heart Fail. 2013; 6: 219-226.Crossref
- 31.
- Liu X, Hu Y, Huang W, et al. Soluble ST2 for prediction of clinical outcomes in patients with ST-segment elevation myocardial infarction receiving primary PCI. Int Heart J. 2019; 60: 19-26.Crossref
- 32.
- Tymińska A, Kapłon-Cieślicka A, Ozierański K, et al. Association of galectin-3 and soluble ST2, and their changes, with echocardiographic parameters and development of Heart failure after ST-segment elevation myocardial infarction. Disease Markers. 2019 Oct 10. .Crossref
- 33.
- Gullestad L, Ueland T, Kjekshus J, et al. The predictive value of galectin-3 for mortality and cardiovascular events in the Controlled Rosuvastatin Multinational Trial in Heart Failure (CORONA). Am Heart J. 2012; 164: 878-883.Crossref
- 34.
- Yu L, Ruifrok WP, Meissner M, et al. Genetic and pharmacological inhibition of galectin-3 prevents cardiac remodeling by interfering with myocardial fibrogenesis. Circ Heart Fail. 2013; 6: 107-117.Crossref
- 35.
- Anand IS, Rector TS, Kuskowski M, et al. Baseline and serial measurements of galectin-3 in patients with heart failure: relationship to prognosis and effect of treatment with valsartan in the Val-HeFT. Eur J Heart Fail. 2013; 15: 511-518.
- 36.
- Lax A, Sanchez-Mas J, Asensio-Lopez MC, et al. Mineralocorticoid receptor antagonists modulate galectin-3 and interleukin-33/ST2 signaling in left ventricular systolic dysfunction after acute myocardial infarction. JACC Heart Fail. 2015; 3: 50-58.
- 37.
- Vittinghoff E, McCulloch CE. Relaxing the rule of ten events per variable in logistic and Cox regression. Am J Epidemiol 2007; 165: 710-718.
- 38.
- Xia J, Qu Y, Yin C, Xu D. Preliminary study of beta-blocker therapy on modulation of interleukin-33/ST2 signaling during ventricular remodeling after acute myocardial infarction. Cardiol J. 2017; 24: 188-194.
- 39.
- Stodółkiewicz E, Rewerska B, Rzeszutko M, et al. Leukotriene biosynthesis in coronary artery disease. Results of the Leukotrienes and Thromboxane In Myocardial Infarction (LTIMI) study. Pol Arch Intern Med. 2018; 128: 43-51.