ST-elevation myocardial infarction (STEMI) remains an important cause of death, especially in developed countries, despite significant advances in its prevention and management.1 Primary percutaneous coronary intervention (PCI) has been shown to reduce death, reinfarction and stroke in STEMI patients.2 Accurate risk stratification plays an important role in appropriate therapeutic decision-making for these patients. The Global Registry of Acute Coronary Events (GRACE) and Thrombolysis in Myocardial Infarction (TIMI) risk scores are helpful in the identification of high-risk patients, however they require complex calculation methods that are not easily applicable before hospital admission or in the emergency department, and are therefore not widely used in clinical practice.

Brain natriuretic peptide (BNP) measurement is a simple tool that has been proposed to assist in risk stratification. BNP is secreted in response to cardiac hemodynamic stress mediated by volume and pressure overload.1 The inflammatory response that is triggered leads to myocyte necrosis and apoptosis, collagen deposition, fibrosis, hypertrophy and dilatation, resulting in left ventricular (LV) systolic dysfunction.1,3 Exogenous BNP infusion has proved to have a cardioprotective effect, preventing unfavorable LV remodeling through generation of 3′,5′-cyclic guanosine monophosphate and nitric oxide synthase, leading to reduced myocardial oxygen consumption, enhancement of myocardial relaxation, retardation of adrenergic activation, induction of vascular regeneration, and inhibition of cardiac fibroblast collagen synthesis and proliferation.3,4

BNP is a highly sensitive and specific indicator of the extent of myocardial infarction (MI) and remodeling that is strongly related to short-term (LV systolic dysfunction, ventricular arrhythmias, LV aneurysm) and long-term clinical outcomes, including heart failure (HF) and mortality.2,5–8 BNP's prognostic value is due to the fact that it is the sum of different risk markers. Detection of LV dysfunction during the early phase of MI through elevated plasma BNP levels should prompt close clinical follow-up and imaging studies to guide the therapeutic strategy needed in order to improve the long-term clinical outcome.5 After MI, BNP levels rise rapidly, reaching a peak after 16-18 hours, and usually fall over the next 2-3 days in a monophasic pattern. However, a biphasic pattern, with BNP levels rising once more within five days of MI, is observed in about half of patients and predicts worsening cardiac function and clinical HF.9 With the advent of primary PCI, the importance of remodeling has decreased significantly, however a significant proportion of STEMI patients still suffer impaired LV function, and rates of mortality and progression to HF remain high.3

The current guidelines recommend the use of BNP as a biomarker to provide additional prognostic information in patients with non-STEMI; however, no European or universal risk model has incorporated it for use in STEMI patients.1 To the best of our knowledge, there has been a single Dutch study evaluating a multimarker approach that includes BNP to improve mortality prognostication in STEMI patients.10 There is no evidence on this issue in the Portuguese population, despite its significant impact on clinical practice. The present study aims to assess the added prognostic value of BNP levels in a Portuguese cohort of STEMI patients regarding all-cause mortality and major adverse cardiac events (MACE).

The Portuguese Registry of Acute Coronary Syndromes (ProACS) is a national multicenter voluntary registry of patients admitted with acute coronary syndromes. Baseline demographic data, cardiovascular risk factors and other relevant personal background, clinical, laboratory, echocardiographic and angiographic data, and medications of all patients are recorded. The present analysis is a retrospective cohort study, but all data were collected prospectively during the index hospitalization. Fifteen Portuguese centers contributed to this study (four in the North region, four in the Center region, three in the Greater Lisbon area and four in the Lower Alentejo). Primary PCI was performed using standard techniques according to the institution's protocol and European guidelines for the management of patients with STEMI. The use of antiplatelet agents, beta-adrenergic blocking agents, angiotensin-converting enzyme inhibitors (ACEIs)/angiotensin receptor blockers (ARBs), statins or inotropic drug support was left at the clinician's discretion according to clinical protocols.

A total of 8036 consecutive STEMI patients were enrolled in this registry between October 2010 and January 2019. Of these, 2695 patients were excluded due to lack of information related to LV systolic function and coronary artery disease characterization, 22 due to lack of follow-up data and 3669 due to lack of BNP quantification. The remaining patients were stratified according to BNP levels: <100 pg/ml, 100-399 pg/ml and ≥400 pg/ml. These cut-off values were established in accordance with the upper limit of normal values in an acute setting set out in the European guidelines on heart failure.

The population was thus composed of 1650 subjects, of whom 1287 (78%) were followed for at least one year after the index hospitalization. Follow-up data were obtained by reviewing medical records or through telephone interviews with patients. The clinical outcomes analyzed were in-hospital adverse events and mortality and a composite of all-cause mortality and rehospitalization for cardiovascular causes at one year after discharge.

We observed higher BNP levels in elderly patients and in those with classic cardiovascular risk factors and previous history of cardiovascular disease, CKD, COPD and cancer. Patients with higher BNP levels had lower systolic BP at admission, higher heart rate and Killip class, and higher creatinine and lower hemoglobin levels, which are known predictors of worse outcome. Their coronary anatomy was more complex, with multivessel disease, left main and left anterior descending lesions, also features associated with a worse prognosis. A higher proportion of these patients were proposed for medical therapy, probably reflecting significant surgical risk and frailty. In patients referred for revascularization, there was a trend, albeit not statistically significant, toward coronary artery bypass grafting or a hybrid approach, probably reflecting more severe coronary disease. These features were also in accordance with the more frequent need for advanced therapeutic devices and ventilatory support in this group.

BNP is a polypeptide secreted by the ventricles in response to excessive stretching of cardiomyocytes, and increased BNP may represent a greater extent of myocardial ischemic territory, which would explain the positive correlation between elevated BNP and MACE in MI patients.5 In our study, patients with high BNP levels had lower LVEF. Previous studies have shown an association between elevated BNP level and multivessel disease, independently of LV systolic dysfunction and congestive HF.4

Higher BNP is associated with short-term complications, including HF, ventricular arrhythmias and LV aneurysm formation, which is in accordance with our results, although we did not specifically assess the presence of LV aneurysms.11 HF was the most frequent in-hospital complication, followed by AF and cardiogenic shock. In-hospital mortality was more than seven times higher in the highest than in the lowest BNP group. At discharge, patients with higher BNP were more likely to be prescribed drugs to control cardiovascular risk factors (aspirin and statins) and for AF control and anticoagulation. The rate of neurohormonal therapy was lower in the high BNP group, except for aldosterone antagonists, probably as a result of their in-hospital clinical course requiring greater use of inotropes and levosimendan, which could have limited the use of neurohormonal HF drugs. The 30-day mortality of patients with STEMI treated with PCI ranges between 5.3 and 7.3%, and is higher in patients with HF (16.5%).12 Although overall mortality in our study was low, it proved that high BNP was associated not only with worse in-hospital outcomes but also with one-year outcomes, which is in line with previous studies.13 BNP levels probably represent the sum of well-known prognostic markers in coronary artery disease, the presence and extent of LV ischemia and dysfunction, suggesting that BNP therapy-guided interventions might improve mortality after STEMI.14 Furthermore, in patients with type 2 diabetes and renal dysfunction, independently of glomerular filtration rate, BNP is an important predictor of mortality, cardiovascular death, MI, HF, and stroke.15 In survivors, patients’ quality of life after hospital discharge is affected primarily by the development of HF, recurrence of ischemic events, and death. Close monitoring of STEMI patients enables the identification of high-risk patients who might benefit from more regular attention and long-term follow-up.12 One-year mortality in our study was 4.6% and 9.0% were hospitalized for cardiovascular causes.

The results of our study are supported by the findings of previous studies aiming to identify STEMI patients at high risk of poor clinical outcomes. Accurate, comprehensive and simple risk assessment plays an important role in appropriate therapeutic decision-making for these patients. Higher-intensity treatments may be appropriate in patients with higher risk scores; however, their inappropriate use in low-risk patients may expose them to adverse effects. Although existing risk scores present good predictive value, their use is limited due to their complexity, the need to carry out various calculations, and the effect of procedure-related variables. BNP can be rapidly measured during admission and is increasingly shown to be predictive of short- and long-term outcomes following STEMI, independently of the presence of clinical HF during hospitalization, providing significant prognostic information in addition to that of Killip class and TIMI risk score in STEMI patients.16–18

Our study presents some limitations. First, it was a retrospective observational study and thus susceptible to inherent limitations, including selection bias. Since this is a voluntary registry, it does not represent all the cardiology centers in Portugal and, even in the centers that did participate, we cannot guarantee that all STEMI patients are included. Also, STEMI patients suffering out-of-hospital death were probably not included. Second, BNP was collected for clinical reasons and this decision was made at the discretion of the attending physician, thus there was only one measurement, which does not allow an analysis of BNP kinetics, and the timing of the measurements was variable. BNP values may have been determined at admission, reflecting the baseline severity of the population, during hospitalization, or at discharge, reflecting the evolution of STEMI itself and implemented therapies. Thirdly, there were no data regarding adherence to medical treatment after discharge.

Patients with higher BNP levels were significantly older, had more cardiovascular risk factors and comorbidities, more severe clinical presentation and complications, and needed more aggressive treatment. These differences might explain their higher in-hospital and one-year mortality and adverse outcomes. Importantly, BNP levels during the index hospitalization were a powerful prognostic biomarker for one-year all-cause mortality and MACE in patients with STEMI, which may indicate a need for more regular monitoring of these patients at follow-up.

The authors have no conflicts of interest to declare.