Temporal trajectory and left ventricular ejection fraction association of eight circulating biomarkers in first acute myocardial infarction patients: a 12-month prospective cohort study

Abstract

Aims

Our study aimed to explore the temporal trajectory of eight circulating biomarkers, measured serially over 12 months, in a prospective observational cohort of patients with acute myocardial infarction (AMI) and to investigate the association between these biomarkers and left ventricular ejection fraction (LVEF) during follow-up assessments.

Methods and results

We enrolled 155 patients admitted for a first AMI requiring percutaneous coronary intervention (PCI). Baseline characteristics, laboratory test results, and cardiac ultrasound examinations were collected at pre-PCI (H0), immediately post-PCI (H24), at discharge (D3), and at 6 months (M6) and 12 months (M12) post-PCI. Blood samples were analysed for established and emerging biomarkers described in left ventricular dysfunction: soluble suppression of tumorigenicity 2 (sST2), interleukin-6 (IL-6), osteopontin, angiopoietin-2, insulin-like growth factor-binding protein 2 (IGFBP-2), growth differentiation factor 15 (GDF-15), hepcidin, and galectin-3. Values at H24, D3, M6, and M12 were compared with value at H0. Three kinetic profiles were identified, with six biomarkers peaking during the acute MI phase. Crude relationships between clinical variables and the peak values (highest observed between H0 and D3) of each biomarker were studied. Peak levels of sST2, IL-6, osteopontin, and angiopoietin-2 demonstrated significant correlations with both baseline and follow-up LVEF values.

Conclusion

The assessment of the temporal trajectories of these biomarkers and their associations with LVEF suggests that sST2, IL-6, osteopontin, and angiopoietin-2 hold significant promise as companion biomarkers. These biomarkers may improve the identification of patients at risk for developing impaired LVEF following AMI, thereby enabling more targeted and effective management strategies.

Graphical abstract

Open in new tabDownload slide

Introduction

Acute myocardial infarction (AMI) related to coronary artery disease (CAD) remains a significant global health challenge, contributing substantially to mortality and disability worldwide. Despite advances in acute management that have reduced early mortality rates following AMI, the long-term mortality risk continues to be significant. Adverse cardiovascular events, including recurrent myocardial infarction (MI), heart failure (HF), and sudden cardiac death, contribute to an overall 10-year mortality rate of around 30% for AMI patients.¹ Additionally, while primary percutaneous coronary intervention (PCI) has successfully reduced AMI-related morbidity and mortality, the incidence of HF remains a concern. Approximately 5–10% of patients treated with PCI and optimal medication exhibit left ventricular (LV) remodelling within the first year after the acute event.² Consequently, AMI patients represent a high-risk group for whom HF screening and prevention are crucial. The identification of high-risk AMI patients involves evaluating a combination of clinical, anatomical, and functional assessments such as reduced left ventricular ejection fraction (LVEF).³ Indeed, studies have demonstrated a strong correlation between lower baseline LVEF values and an increased risk of HF incidence and progression.² Baseline LVEF provides a snapshot of cardiac function at a specific point in time, typically at the diagnosis of myocardial infarction, but may not capture the dynamic changes that occur during the remodelling process. Currently, natriuretic peptides, including NT-proBNP (N-terminal pro-B-type natriuretic peptide) and BNP (brain natriuretic peptide), are used for risk stratification of patients with acute coronary syndrome and to monitor the treatment of those with LV dysfunction. However, despite their clinical utility, NT-proBNP, and BNP have limitations due to confounding factors that can affect their interpretation and accuracy as diagnostic and prognostic tools.^4–6 Most studies on biomarker measurement for predicting LV remodelling or HF post-AMI have relied on a single baseline measurement. However, repeated measurements could better reflect biomarker dynamics and disease progression.

We hypothesized that combining serial measurements of both novel and established biomarkers could provide a powerful complementary approach to better identify patients at risk of developing impaired LVEF after AMI, independent of their baseline LVEF value. The primary objective of this study was to analyse the temporal trajectories of eight biomarkers over five serial blood measurements taken across 12 months in a prospective cohort of patients admitted for a first AMI treated with PCI. The secondary objective was to evaluate the association between these biomarkers and LVEF during follow-up assessments.

These eight established and emerging biomarkers are described in LV dysfunction signalling and may reflect various stages of the pathological process: soluble suppression tumorigenicity 2 (sST2; myocardial fibrosis,⁷) interleukin-6 (IL-6; inflammation, myocardial fibrosis and hypertrophy,^8,9) osteopontin (inflammatory cell infiltration and matrix remodelling,^10,11) angiopoietin-2 (endothelial cell dysfunction and impaired angiogenesis,^12,13) hepcidin (oxidative stress and cardiomyocyte apoptosis,^14,15) insulin-like growth factor-binding protein 2 (IGFBP-2; LV dysfunction,^16,17) growth differentiation factor 15 (GDF-15; apoptosis, inflammation, and matrix remodelling^18–20) and galectin-3 (fibroblast activation and matrix remodelling^21–23).

Methods

Study population

This prospective, longitudinal cohort study was conducted at the Toulouse University Hospital Center in France, involving patients admitted to the intensive care unit for AMI between September 2018 and July 2021. The study protocol adhered to the ethical guidelines of the 1975 Declaration of Helsinki and Toulouse Hospital guidelines. It was reviewed and approved by the National Institution’s ethics committee on research on humans (Comité de Protection des Personnes, France), with approval number 2017-A01959-44.

Inclusion criteria included patients over 18 years of age admitted for a first AMI related to acute coronary syndrome, according to the ‘Third Universal Definition of MI’,²⁴ leading to a PCI.

Exclusion criteria were: patients resuscitated after sudden cardiac arrest by any method other than single, patients in cardiogenic shock, patients with severe extracardiac diseases affecting short-term life expectancy (risk of death within one year), patients scheduled for major non-cardiac surgery, those with known altered left ventricular function before inclusion, patients with a history of left, right, or global heart failure, patients with significant heart failure at hospital admission (KILLIP class 2 or higher), patients with acute coronary syndrome known for more than 24 h, or those with a previous MI.

The recruitment phase, initially planned for 24 months, was extended to 48 months due to the COVID-19 pandemic, followed by 12 months of monitoring for each patient. Baseline characteristics, laboratory test results, and cardiac ultrasound examinations were collected at pre-PCI (H0), immediately post-PCI time (H24), at discharge (D3), and at 6 months (M6) and 12 months (M12) post-PCI time, using electronic medical records (Figure 1). Cardiac ultrasound (transthoracic echocardiography, TTE) examinations were performed at D3, M6, and M12.

Consent

All patients included have signed an informed consent to participate in study.

Measurement of biomarkers

For each patient, plasma samples were collected at the specified five time points, centrifuged, aliquoted, and stored at −80°C. The concentrations of soluble suppression of tumorigenicity 2 (sST2) were determined using the Aspect-Plus ST2 lateral flow immunoassay (Critical Diagnostics, San Diego, USA), following the manufacturer's specifications. Galectin-3, hepcidin, osteopontin, IGFBP-2, angiopoietin-2, and IL-6 were measured using the Ella Automated Immunoassay System (Bio-Techne Corporation, Minneapolis, USA), adhering to the manufacturer's assay characteristics. Routine biomarkers were assessed in the Biochemistry Laboratory of the Toulouse University Hospital Center. High-sensitivity cardiac troponin T (hs-cTnT), NT-proBNP, and growth differentiation factor-15 (GDF-15) levels were measured in plasma aliquots using specific Elecsys® quantitative sandwich electro-chemiluminescence immunoassays on a Cobas 8000 analyzer (Roche Diagnostics, Mannheim, Germany), according to the manufacturer's guidelines. Total creatinine kinase activity was measured using a spectrophotometric assay based on the enzymatic reaction catalysed by creatinine kinase. Creatinine was measured using the CREP2 (creatinine plus ver.2 test), and C-reactive protein (CRP) levels were determined using the Tina-quant CRP IV test, both conducted on a Cobas 8000 analyzer (Roche Diagnostics, Mannheim, Germany), following the manufacturer's specified assay characteristics.

Statistical analysis

Categorical data were presented as numbers and percentages, while quantitative data were expressed as mean ± standard deviation or as median (interquartile range) for skewed distributions. To analyse the kinetics of each biomarker, the values observed at each time point (H24, D3, M6, and M12) were compared with those observed at H0 using repeated-measures one-way ANOVA, possibly with logarithmic transformation if necessary. The peak level of each biomarker was defined as the highest value observed during the acute phase between H0 and D3. Crude relationships between patient characteristics and the peak level of each biomarker were studied using non-parametric tests: Spearman’s correlation coefficient for continuous data and Mann–Whitney or Kruskal–Wallis tests for categorical data. All statistical analyses were conducted using Stata Statistical Software (StataCorp. 2023. Stata 18, Statistical software, StataCorp LLC).

Results

Population characteristics

We assessed a cohort of 155 patients (comprising both men and women) who experienced a first AMI event between September 2018 and July 2021, all treated PCI. Of the initial 155 participants, 119 subjects (77%) completed the study and were included in the present analysis. Reasons for exclusion from the final analysis included missing intermediate visits due to time constraints or the COVID-19 pandemic, as well as termination of the study due to refusal to continue. The baseline characteristics of the study population are presented in Table 1. In summary, the mean age was 61.4 ± 10.8 years, 80.7% were male, 69% were ST-elevation myocardial infarction (STEMI) patients, and the median LVEF at admission was 55% [interquartile range (IQR) 51–60].

Trajectories and kinetics of biomarkers over time

Figure 1 illustrates the trajectories and kinetics of biomarkers measured in AMI patients: sST2 (A), IL-6 (B), IGFBP-2 (C), osteopontin (D), angiopoietin-2 (E), hepcidin (F), GDF-15 (G), and galectin-3 (H) across the five measurement points. Comparing plasma levels of sST2 at H0 to H24, D3, M6, and M12 revealed a significant increase (P < 0.001) from baseline H0 (27.6 ng/mL, IQR [22.5–39.3]) to H24 (37.8 ng/mL, IQR [25–67.5]). This peak at H24 was followed by a significant decrease, with levels remaining stable from D3 to the 12-month measurement points. Similar patterns were observed for IL-6 and IGFBP-2, with significant increases (P < 0.001) from baseline H0 (IL-6: 6 pg/mL, IQR [4–12]; IGFBP-2: 227 ng/mL, IQR [181–305]) to H24 (IL-6: 17 pg/mL, IQR [10–32]; IGFBP-2: 301 ng/mL, IQR [277–378]). Osteopontin levels significantly increased (P < 0.001) from H0 (84 ng/mL, IQR [69–109]) to the D3 peak (131 ng/mL, IQR [90–172]), stabilizing thereafter at the 6 and 12-month measurements. Angiopoietin-2 displayed a slight decrease at H24 (1.4 ng/mL, IQR [1.1–1.7]) compared with H0 (1.5 ng/mL, IQR [1.2–1.9]), followed by a significant increase at D3 (1.6 ng/mL, IQR [1.2–2.1]), returning to baseline levels at M6 and M12. Hepcidin concentrations did not significantly change at H24, D3, or M12 compared with baseline H0, but a significant unexpected decrease was noted at M6 (28 ng/mL, IQR [18–49]) compared with baseline H0 (48 ng/mL, IQR [25–84]). GDF-15 concentrations exhibited a slight significant increase at H24 (1273 pg/mL, IQR [1005–1797]), followed by significant decreases at M6 (1019 pg/mL, IQR [780–1426]) and M12 (1066 pg/mL, IQR [800–1443]) compared with baseline H0. Galectin-3 levels showed a constant significant increase from baseline H0 (6.9 ng/mL, IQR [5.7–8.5]) to the M12 (10 ng/mL, IQR [8.4–11.5]) measurement points.

Relationships between biomarkers, left ventricular ejection fraction, and clinical variables

Spearman rank correlation coefficients between variables are presented in Table 2. The peak level of each biomarker was defined as the highest value observed during the acute phase between H0 and D3. Crude relationships between patient characteristics and the peak level of each biomarker were analysed. In the absence of a significant peak during the acute phase, which was the case for hepcidin and galectin-3, no correlation study was conducted.

Soluble suppression of tumorigenicity 2

The peak sST2 concentration at H24 correlated positively with NT-proBNP peaks at H24 (ρ = 0.47, P < 0.001), D3 (ρ = 0.45, P < 0.001), M6 (ρ = 0.33, P < 0.001), and M12 (ρ = 0.35, P < 0.001). Hs-cTnT levels during the acute phase H0 (ρ = 0.27, P = 0.004), H24 (ρ = 0.41, P < 0.001), and D3 (ρ = 0.50, P < 0.001) also correlated positively with sST2 peak. Additionally, total creatine kinase (ρ = 0.49, P < 0.001) and C-reactive protein (ρ = 0.35, P = 0.035) levels measured at H24 were positively correlated with sST2 peak. LVEF values at baseline (ρ = −0.31, P < 0.001), M6 (ρ = −0.29, P = 0.002), and M12 (ρ = −0.39, P < 0.001) inversely correlated with sST2 peak (scatterplots provided in Supplementary material online, Data), and a significant correlation was found with the type of AMI (P = 0.044), a higher sST2 value is found in STEMI patients compared with NSTEMI patients. No significant association was shown with traditional cardiovascular risk factors.

Interleukin-6

The IL-6 peak concentration at H24 correlated positively with NT-proBNP levels at all time points: H0 (ρ = 0.30, P = 0.002), H24 (ρ = 0.47, P < 0.001), D3 (ρ = 0.52, P < 0.001), M6 (ρ = 0.38, P < 0.001) and M12 (ρ = 0.34, P < 0.001), with hs-cTnT levels during the acute phase H0 (ρ = 0.36, P < 0.001), H24 (ρ = 0.47, P < 0.001) and D3 (ρ = 0.50, P < 0.001), with total creatine kinase levels at H0 (ρ = 0.28, P = 0.003), H24 (ρ = 0.44, P < 0.001) and D3 (ρ = 0.40, P < 0.001), and C-reactive protein levels (ρ = 0.38, P < 0.001). LVEF values at baseline (ρ = −0.30, P < 0.001) and M12 (ρ = −0.19, P = 0.047) inversely correlated with IL-6 peak (scatterplots provided in Supplementary material online, Data), and a significant association was found with age (ρ = 0.26, P = 0.005). No significant associations were found with the type of AMI.

Osteopontin

The osteopontin peak concentration at H24 correlated positively with NT-proBNP levels at all time points: H0 (ρ = 0.19, P = 0.05), H24 (ρ = 0.37, P < 0.001), D3 (ρ = 0.29, P = 0.002), M6 (ρ = 0.34, P < 0.001) and M12 (ρ = 0.29, P = 0.003), with hs-cTnT levels during the acute phase: H0 (ρ = 0.21, P = 0.026), H24 (ρ = 0.32, P < 0.001) and D3 (ρ = 0.25, P = 0.008), with total creatine kinase levels at H24 (ρ = 0.24, P = 0.013) and D3 (ρ = 0.24, P = 0.012), and C-reactive protein (ρ = 0.21, P = 0.03). Osteopontin peak inversely correlated with LVEF values at baseline (ρ = −0.20, P = 0.033), M6 (ρ = −0.21, P = 0.028) and M12 (ρ = 0.24, P = 0.01) (scatterplots provided in Supplementary material online, Data). A notable correlation was observed with the AMI category (P = 0.012), a higher osteopontin value is found in STEMI patients compared with NSTEMI patients. However, no significant associations were evident with traditional cardiovascular risk factors, except for hypertension (P = 0.022).

Growth differentiation factor 15

The peak concentration of GDF-15 at H24 exhibited a positive correlation with heart rate (ρ = 0.31, P = 0.001) and C-reactive protein (ρ = 0.41, P < 0.001), while displaying an inverse correlation with eGFR at H24 (ρ = −0.30, P = 0.003), D3 (ρ = −0.26, P = 0.009), M6 (ρ = −0.44, P < 0.001), and M12 (ρ = −0.41, P < 0.001) measurements. No significant associations were found between LVEF values and traditional cardiovascular risk factors, except for age (ρ = 0.31, P = 0.002), nor with the AMI category.

Insulin-like growth factor-binding protein 2

The peak concentration of IGFBP-2 at H24 exhibited positive correlations with NT-proBNP at H24 (ρ = 0.18, P = 0.05), D3 (ρ = 0.27, P = 0.004), and M12 (ρ = 0.33, P < 0.001) levels. IGFBP-2 peak levels inversely correlated with eGFR at H24 (ρ = −0.20, P = 0.033). No significant associations were found with LVEF values, nor with the AMI category. Positive correlations were observed with age (ρ = 0.29, P = 0.002) and inverse correlations with BMI (ρ = −0.30, P = 0.001).

Angiopoietin-2

The peak concentration of angiopoietin-2 at D3 displayed positive correlations with NT-proBNP at H24 (ρ = 0.31, P = 0.001), D3 (ρ = 0.35, P < 0.001), M6 (ρ = 0.29, P = 0.004), and M12 (ρ = 0.28, P = 0.005). Hs-cTnT levels at H24 (ρ = 0.34, P < 0.001) and D3 (ρ = 0.37, P < 0.001) showed positive correlations with angiopoietin-2. Total creatine kinase levels at H24 (ρ = 0.39, P < 0.001) and D3 (ρ = 0.26, P = 0.008), as well as C-reactive protein (ρ = 0.19, P = 0.052), were positively correlated with the angiopoietin-2 peak. Negative correlations were observed with LVEF at baseline (ρ = −0.25, P = 0.009), M6 (ρ = −0.26, P = 0.006), and M12 (ρ = −0.33, P < 0.001) measurements (scatterplots provided in Supplementary material online, Data). No significant associations were found with traditional cardiovascular risk factors, nor with the AMI category.

Discussion

The temporal trajectory of biomarkers and their association with adverse cardiovascular outcomes following AMI remains an area with limited evidence. While previous research in cardiovascular disease risk assessment has predominantly relied on single biomarker measurements, serial assessments allow for the dynamic monitoring of disease progression and response to therapy over time.

Our data, for the first time, assessed the kinetic trajectory of multiple biomarkers in an AMI patient cohort across five time points, spanning from hospitalization to a 12-month follow-up period. Among the measured biomarkers, we identified three distinct biomarker profiles that offer valuable insights that can be leveraged in clinical practice to tailor monitoring and therapeutic strategies accordingly. For instance, patients with a rapidly rising biomarker profile may benefit from more aggressive interventions, while those with a stable or declining profile might be candidates for conservative management. In addition, these profiles could serve as dynamic markers for monitoring disease progression and evaluating the effectiveness of treatments thereby improving patient outcomes. The first profile is characterized by sST2, IL-6, osteopontin, GDF-15, angiopoietin-2, and IGFBP-2, demonstrating significant changes and a peak concentration at H24 post-PCI, followed by a decline at 6 and 12 months. Our findings align with prior research describing an acute-phase concentration peak, primarily attributed to the proinflammatory state in patients with AMI.^9,25–28 Given their association with key pathological processes, such as cardiomyocyte death, endothelial dysfunction, immune cell infiltration, and myofibroblast activation, these biomarkers could serve as valuable tools in clinical practice. Monitoring these biomarkers may aid in the early identification of patients at risk for adverse cardiac remodelling, and assist clinicians to optimize antifibrotic including ACE inhibitors and angiotensin receptor blockers and, select patients for anti-inflammatory therapy.

The second profile encompasses hepcidin, which did not exhibit notable variations along the trajectory profile. Hepcidin, a peptide hormone involved in iron metabolism and inflammation,²⁹ lacks prior kinetic studies during AMI. Based on its role and our data, we can only hypothesize that hepcidin, by reducing iron availability, could impact erythropoiesis and tissue repair processes post-AMI.¹⁴ However, there is evidence that iron deficiency (ID), which can be influenced by hepcidin levels, can exacerbate HF. Recently, it was reported that mice spontaneously develop cardiac ID long term after MI and identified cardiac hepcidin suppression as a potential mechanism.¹⁵ Pre-emptive iron supplementation prevents cardiac ID and attenuates post-MI remodelling in these mice, suggesting that iron modulation strategies could be explored in post-AMI care. Although we did not include hepcidin in the second part of our study, further research involving a larger sample size and the measurement of additional iron biomarkers could provide deeper insights into the potential role of hepcidin.

The third kinetic profile is defined by the persistent elevation of galectin-3 concentration, a β-galactoside-binding lectin involved in inflammation, fibrosis, and tissue repair.²² The sustained increase in galectin-3 concentration from baseline to 12 months is consistent with previous studies demonstrating an early rise after the acute event. This rise reflects the inflammatory response and the initiation of reparative processes in the injured myocardium.³⁰ Given its role in promoting collagen deposition and fibrotic changes in the myocardium,^31–33 the persistent elevation of galectin-3 underscores its importance as a marker for ongoing fibrosis. This underscores the importance for clinicians to prioritize antifibrotic therapies and supports ongoing research into novel therapeutic targets for chronic fibrosis, as well as regenerative strategies.

In the second part, our study extends the data on kinetic trajectories with correlation analyses between biomarker peak values during the acute phase and clinical variables at baseline and follow-ups. The significant correlations between sST2, IL-6, osteopontin, IGFBP-2, and angiopoietin-2 biomarkers and NT-proBNP levels at hospitalization and follow-up points showed that these markers reflect key pathophysiological processes in HF genesis, such as cardiac stress, inflammation, fibrosis, metabolic disturbances, and vascular remodelling. The lack of correlation between GDF-15 and NT-proBNP levels suggests that while GDF-15 is involved in processes relevant to HF, it is influenced by a range of factors such as renal dysfunction and systemic inflammatory diseases. Consequently, GDF-15 may reflect broader systemic responses rather than directly correlating with cardiac-specific stress and severity, as indicated by NT-proBNP. Circulating sST2 and osteopontin levels are significantly higher in STEMI patients, who generally experience more severe cardiac dysfunction and have a poorer prognosis. Additionally, sST2 and osteopontin are significantly inversely correlated with LVEF values at baseline and follow-ups, confirming their pathophysiological role in inflammation, stress response, and fibrosis, all of which are associated with adverse remodelling and deteriorated myocardial function.^25,26,31,32 Therefore, our data underscore the utility of these biomarkers as indicators for assessing LV dysfunction adding value to traditional biomarkers.

The significant correlation between the peak level of IL-6 with NT-proBNP, hsTnT, and baseline LVEF suggested that higher IL-6 levels is associated with increased myocardial necrosis leading to reduced LVEF. In STEMI patients, IL-6 has been shown to be upregulated at the site of coronary occlusion and to promote cardiomyocyte death, contributing to reduced LVEF and, high IL-6 values at admission were associated with long-term cardiovascular mortality.^33,34 Acute and sustained elevation of IL-6 measured 4 months after STEMI were also significantly associated with reduced LVEF⁹ and peak IL-6 measured at 24 h post-STEMI is associated with cardiac function measured at 4 months.³⁵ Therefore, these findings support our observations. The significant weak correlation observed with LVEF during follow-ups may suggest that as the acute phase of inflammation resolves, other factors begin to influence cardiac function. Clinical trials using tocilizumab, a monoclonal antibody that neutralizes IL-6, demonstrate reduced inflammation in individuals with ACS, suggesting a beneficial effect of IL-6 inhibition in patients with CAD.^36,37 As describe in the recent review from Matter et al.;³⁸ the challenge remains in optimizing anti-inflammatory interventions, and ongoing studies on novel anti-inflammatory therapies in the post-MI setting are anticipated to provide crucial insights.

While IGFBP-2 has been associated to mortality risk in heart failure¹⁷ and poor prognosis in patients with acute coronary syndrome,¹⁶ we did not find any association between IGFBP-2 and LVEF at baseline or follow-ups. IGFBP-2’s primary role is in regulating insulin-like growth factors, which are involved in long-term growth and metabolic processes. Whereas these factors can influence cardiac function, their impact is likely more indirect and long-term, rather than directly correlating with immediate changes in LVEF. Peak level of GDF-15 was not correlated with LVEF at baseline or follow-ups, although a former study has evidenced that higher GDF-15 levels have been associated with LV remodelling.¹⁹ However, GDF-15 is a stress-responsive cytokine produced by both cardiovascular and non-cardiovascular cell types under pathological conditions; instead, it reflects the cellular stressors prevalent in many chronic disease states, including HF, which are related to co-morbidities, ageing, and even lifestyle. In the PARADIGM-HF trial, the lack of response of GDF-15 to valsartan and sacubitril/valsartan in HF patients with reduced ejection fraction led the authors to conclude that GDF-15 represents ‘an integrated biomarker of multiple co-morbidities rather than a specific reflection of cardiovascular health’.³⁹ Thus, this broad response profile may diminish the direct relationship between GDF-15 levels and specific measures of cardiac function such as LVEF.

Study limitations

Our observational study has some limitations to note. Firstly, being a single-centre study may limit the applicability of the findings to more diverse populations. Additionally, the relatively small sample size, while sufficient for detecting significant differences, may reduce the statistical power to detect associations and may limit the generalizability of the findings. Conducting multi-centre studies or subsequent validation with larger sample sizes independent cohorts would be beneficial to confirm and extend our findings. Secondly, selecting biomarkers based on pre-existing hypotheses can introduce selection bias. Since complex diseases often involve networks of biomarkers rather than isolated ones, this complexity may not have been fully explored. Thirdly, the lack of standardized methods for measuring biomarkers in clinical routine may affect replicability in other laboratories or with other techniques.

Conclusion

In conclusion, our observational study successfully achieved its objectives by analysing the temporal trajectories of eight established and emerging biomarkers involved in the pathophysiological processes of LV remodelling and HF. Among these, sST2, IL-6, osteopontin, and angiopoietin-2 were significantly associated with LVEF impairment at 12 months. These findings underscore the additive value of these biomarkers in identifying high-risk post-MI patients, both as companion biomarkers and as potential novel therapeutic targets. Further studies will be needed to validate their use as tools for risk stratification. Finally, incorporating multi-biomarker-guided strategies could enhance individualized patient care by tailoring treatment interventions to specific risk profiles, ultimately optimizing patient outcomes.

Lead author biography

Dr Cecile Vindis, PhD, is Director of Research at the French National Institute of Health and Medical Research. She started her research works as a group leader at the Institute of Metabolic and Cardiovascular Diseases in Toulouse where she headed a team working on mechanisms involved in cell survival (ER stress, autophagy/mitophagy) and death during atherosclerotic plaque progression; as well as translational research projects on new cardiovascular risk biomarkers with the cardiologists of the University Hospital of Toulouse. She is now group leader at the Clinical Investigation Center where she conducts translational research to discover new mechanisms and biomarkers involved in coronary artery disease.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Supplementary material

Supplementary material is available at European Heart Journal Open online.

Authors’ contribution

All authors contributed to the article and approved the submitted version.

Acknowledgements

We would like to express our sincere gratitude to the nurses, the patients, and all investigators involved in this study, without whom the study would not have been possible. We thank Roche Diagnostics for generously providing the reagents for the GDF-15 assay. We are grateful to D. Tayac for assistance in calibrating the GDF-15 assay, as well as the technical staff of the Clinical Biochemistry Laboratory at CHU Rangueil, Toulouse.

Funding

This work was supported by a European Union project (funding by Eurostars program 2017) to M.E. and H.F. and by ‘La Fédération Française de Cardiologie’ to C.V. and M.E.

Conflict of interest: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

References