P3825Long-term prognostic value of growth differentiation factor-15 in acute coronary syndrome

Introduction Biomarkers plays a critical role in diagnostic, prognostication, and decision-making in cardiovascular medicine. Growth differentiation factor-15 (GDF-15) has been reported as a potential biomarker in acute coronary syndrome (ACS). However, there is limited data on the long-term prognostic value after an ACS. Purpose To study the long-term prognostic value of GDF-15 in ACS. Methods We included patients with ACS who underwent coronary angiography. During angiography an arterial blood sample was collected. Plasma GDF-15 were measured and clinical data and long-term events were obtained. As previously reported, risk categories were defined as low risk (<1200ng/L), intermediate (1200–1800ng/L) and high risk (>1800ng/L). Incremental prognostic value of GDF-15 for all-cause death was assessed on top of a clinical model (GRACE score, LVEF<40% and age). Results A total of 358 patients were included; 157 as a low risk, 85 as an intermediate and 116 as a high risk. The median (IQR) age was 65 (56–74) years and 27.4% were female. Of all patients, 61.5% were admitted with non-ST-elevation myocardial infarction, 24.0% with ST-elevation myocardial infarction and 14.5% with unstable angina. Higher values of GDF-15 were consistently associated with an increased prevalence of cardiovascular risk factors. During 6 years of follow-up 54 patients died. Of those patients, 7 (4.5%) had values of GDF-15 below 1200ng/L, 6 (7.1%) between 1200–1800ng/L and 41 (35.3%) above 1800ng/L. After adjustment for a multivariate Cox regression model, GDF-15 >1800ng/L were independently associated with all-cause death (HR 4.5; 95% CI 1.8–11.6; p=0.002) and the composite of major adverse cardiovascular events (MACE) which were identified as all-cause death, nonfatal MI and heart failure (HR 2.5; 95% CI 1.4–4.4; p=0.001). For long-term all-cause death a significant increase of the c-statistic was seen after addition of GDF-15 to the clinical model 0.871 (95% CI 0.817–0.924; p=0.019) as well as net reclassification improvement (0.769; 95% CI 0.487–1.051; p<0.001) and integrated discrimination improvement (0.117; 95% CI 0.062–0.172; p<0.001). Of 18 events of heart failure, 17 occurred in patients with GDF>1800ng/L. A multivariate competing risk model showed a significant association between GDF-15>1800ng/L and incidence of heart failure (adjusted HR 30.8; 95% CI 4.1–231.5; p=0.001) but non-significant association were found for myocardial infarction. KM figures and all-cause death ROC curve Conclusions In the setting of ACS GDF-15 can predict long-term all-cause death, MACE and heart failure and provides incremental prognostic value beyond traditional risks factors in the long-term all-cause death.