Growth Differentiation Factor-15 Levels at Admission Provide Incremental Prognostic Information on All-Cause Long-term Mortality in ST-Segment Elevation Myocardial Infarction Patients Treated with Primary Percutaneous Coronary Intervention

Abstract

Introduction: To investigate the additive prognostic value of growth differentiation factor (GDF-15) levels in ST-segment elevation myocardial infarction (STEMI) patients treated with primary percutaneously coronary intervention (pPCI) with 10-year mortality on top of clinical characteristics and known cardiac biomarkers. Methods: Baseline serum GDF-15 levels were measured in 290 STEMI patients treated with pPCI in the MISSION! intervention trial conducted from February 1, 2004 through October 31, 2006. The incremental prognostic value of GDF-15 and NTproBNP levels was evaluated on top of clinical characteristics using Cox proportional hazards analysis, Chi-square models and C-index. Outcome was 10-year all-cause mortality. Results: Mean age was 59.0 ± 11.5 years and 65 (22.4) patients were female. A total of 37 patients died during a follow-up of 9.4 (IQR 8.8–10.0) years. Multivariable Cox regression revealed GDF-15 and NTproBNP levels above median to be independently associated with 10-year all-cause mortality [HR GDF-15, 2.453 (95% CI 1.064–5.658), P = 0.04; HR NTproBNP, 2.413 (95% CI 1.043–5.564), P = 0.04] after correction for other clinical variables. Stratified by median GDF-15 (37.78 pmol/L) and NTproBNP (11.74 pmol/L) levels, Kaplan–Meier curves showed significant better survival for patients with GDF-15 and NTproBNP levels below the median versus above the median. The likelihood ratio test showed a significant incremental value of GDF-15 (P = 0.03) as compared with a model with clinically important variables and NTproBNP. The C-statistics for this model improved from 0.82 to 0.84 when adding GDF-15. Conclusion: GDF-15 levels at admission in STEMI patients are independently associated with 10-year all-cause mortality rates and could improve risk stratification on top of clinical variables and other cardiac biomarkers.