The significance of GDF-15 for patients with diastolic dysfunction and heart failure with preserved ejection fraction

Heart failure (HF) is a disease of high prevalence and mortality. Depending on the respective study, more than 50% of patients with HF suffer from heart failure with preserved ejection fraction (HFPEF). In patients with heart failure and reduced ejection fraction (HFREF), the use of biomarkers plays a central part in medical diagnosis and monitoring of the disease course. According to the guidelines of the European Society of Cardiology (ESC), the natriuretic peptides NTproBNP and BNP should be implemented in the diagnosis of HFPEF. An increasing number of biomarkers are currently being tested with respect to their utility for the diagnosis of HEPEF. The importance of GDF-15 for patients with HFPEF has not been sufficiently examined yet. Therefore, this study investigated the clinical relevance and diagnostic utility of GDF-15 plasma concentrations in patients with diastolic dysfunction (DD) and HFPEF and, furthermore, compared these results with other established biomarkers. For this purpose, a total of 330 patients with ascertained HF or positive Framingham criteria were selected from the prospective multicenter study “Prävalenz und Verlauf der diastolischen Dysfunction und der diastolischen Herzinsuffizienz” (Diast-CHF). Patients with HFPEF (n=142) and DD (n=103) were grouped predominantly according to echocardiographic criteria. In order to compare the clinical results, a group of healthy persons (n=85) was additionally included. The median age of study participants with HFPEF was 73 years and for patients with DD 61 years, while the median age of healthy controls was 53 years. Among patients grouped as HEPEF, the percentage of female patients was 64%, which was higher than for patients with DD (59%) but less than for healthy individuals (74%). Serum levels of GDF-15 were measured using a kit developed by Roche Diagnostics which is not commercially available yet. In patients with HFPEF, the serum concentration of GDF-15 reached a median of 1.64 ng/ml and was significantly higher than in the group of healthy controls (0.84 ng/ml). In the group of DD patients, the median concentration of GDF-15 was 0.96 ng/ml, which was significantly lower as compared to patients with HFPEF. I found a significant positive association between serum-concentration of GDF-15 and the severity of diastolic dysfunction. In addition, GDF-15 correlated with several echocardiographic parameters that were used for the diagnosis of HFPEF. Using ROC analysis, the biomarkers BNP (NTproBNP), ANP (MRproANP), adrenomedullin (MRproADM), endothelin-1(CTproET1) and vasopressin (CTproAVP) were compared for their utility to detect HFPEF. For GDF-15 the area under the curve (AUC) was 0.907, and for MRproADM this value was 0.932, while the AUCs for MRproANP, NTproBNP, CTproET1 and CTproAVP were lower than that for GDF-15. However, GDF-15 appeared less suitable for the diagnosis of DD. Furthermore, in controls I found a significant association between serum concentrations of GDF-15 and reduced physical capacity. Thus, GDF-15 seems to be well suited as a biomarker to classify the clinical severity grade of patients with HFPEF and DD. This clinical relevance should be investigated by research still to come in a larger group of patients