Combined use of high-sensitivity st2 and nt-probnp for predicting major adverse cardiovascular events in coronary heart failure

Abstract

Background: Elevated serum soluble ST2 (sST2) level is marker of poor prognosis in chronic heart failure (HF). N-terminal pro-brain natriuretic peptide (NT-proBNP) is an important biomarker in cardiovascular diseases. This study aimed to address the incremental usefulness of the combined use of high-sensitivity sST2 and NT-proBNP for predicting the risk of major adverse cardiovascular events (MACEs) in patients with coronary heart disease (CHD). Methods: We used patient data along with data established by our research group to compare the performance of a combination of biomarkers reflecting ventricular fibrosis, remodeling, stretch and deterioration of cardiac function (sST2 and NT-proBNP) with that of established mortality risk factors [age, diabetes, systolic blood pressure, diastolic blood pressure, left ventricular ejection fraction (LVEF), body mass index (BMI), creatinine (Cr), uric acid (UA), glucose (Glu)] in stratifying the risk of MACEs in CHD patients. Results: The median follow-up period was 3.9 years, during which time there were 3,724 cases with CHD, 113 cases of cardiovascular death, 30 cases with myocardial infarction, 49 cases with stroke, 39 cases of non-cardiovascular death, 6 cases of peripheral arterial occlusion, 55 cases with HF, and 73 cases of revascularization. A total of 365 cases had MACEs. In the multivariate Cox proportional hazard model, both sST2 and NT-proBNP were significant predictors of MACEs. Both sST2 and NT-proBNP were separately incorporated into the model with established MACE risk factors, which significantly improved the C-statistics for predicting MACEs [0.758 (0.724–0.792)] and saw an estimated net improvement in reclassification of 1.03% (P<0.001) and an integrated discrimination improvement of 0.48% (P<0.001). The Hosmer-Lemeshow test showed that the models were well calibrated with and without the two biomarkers (P>0.344 for all comparisons). Moreover, the model incorporating the two biomarkers was shown to have a better global fit than the model with only the established MACE risk factors (P<0.001). Conclusions: A model incorporating sST2 and NT-proBNP was shown to outperform a model based on established MACE risk factors alone in stratifying risk of MACEs in a group of CHD patients.