| dc.description.abstracteng | Heart failure with preserved ejection fraction (HFpEF) became an increasing problem during the last decades and accounts for more than 50% of all heart failure (HF) patients. Due to the lack of evidence based therapie and still poor survival rates, the prediction of prognosis in this special group is of major importance. Nevertheless, factors correlating with mortality in diastolic dysfunction (DD) and HFpEF remain largely unexplored and guidelines for prediction of prognosis do not exist. The present study evaluates the prognostic value of ten different neuropeptides from varying areas of pathophysiology in an existing collective of the prospective multicentre two-cohort study 'DIAST-CHF'. In addition to a basic cardiovascular examination, special laboratory of blood samples was carried out to determine the plasma concentrations of MRproADM, MRproANP, NTproANP, CTproAVP, NTproBNP, hsCRP, CTproET-1, PIIINP, aldosterone and renin in 1484 patients. The correlation between individual biomarkers and mortality and cardiovascular (CV) hospitalization was subsequently analysed in several Cox regression models, leading to hazard ratios (HR). In univariate regression logNTproBNP (HR 1,66), logCTproET-1 (HR 1,69), logMRproADM (HR 1,62) and logMRproANP (HR 1,53) show in each case highly significant (p<0,001) and for logNTproANP (HR 1,24; p=0,019), logPIIINP (HR 1,18; p=0,005) and logCTproAVP (HR 1,23; p=0,008) significant correlations to the endpoint. In this collective there is no significant correlation for aldosterone, hsCRP and renin (p=0,256-0,753). After mutual adjustments in multivariate Cox models there only remained a significant correlation for logNTproBNP (HR 1,28; p=0,014). This is founded in the correlation of the neuropeptides and the overlapping of prognostic informations. The examination of several neuropeptides with similar prognostic informations seemed to be more useful than examining singular neuropeptides. This led to the analysis of two factors (A/B) in final Cox regression model, which weighted either the cardial central or the peripheral prognostic informations. The present study shows in a collective of HFpEF- and risk patients, that all considered neuropeptides except hsCRP, aldosterone and renin correlated with the combined endpoint of mortality and CV-hospitalization after 5,2 years of follow-up and, that there are strong correlations between the neuropeptides amongst each other. An increase in plasma concentration of cardial central biomarkers raises the risk for mortality and CV-hospitalization more (factor A: HR 1,31; p=0,007) than those in peripheral biomarkers (factor B: HR 1,23; p=0,02). NTproBNP is best suited when determining a singular marker for predicting the prognosis in DD and HFpEF, especially due to its unrivalled valence and independence. | de |