98 Poster Display II. Prognosis 359 Myostatin serum levels in heart failure E. Zamora 1 , J. Lupon 1 , R. Simo 2 , A. Galan 1 , A. Urrutia 1 , D. Mas 1 , B. Gonzalez 1 , V. Valle 1 1 Hospital Universitari Germans Trias i Pujol, Badalona, Spain; 2 Hospital Universitari Vall d’Hebron, Barcelona, Spain Background: muscle wasting is quite frequent in heart failure (HF). However, the mechanisms involved in this process and its clinical and prognostic implications are fare to be elucidated. In recent years, myostatin (MYOS) has became as an important modulator of muscle catabolism. Objective: to assess the MYOS and MYOS pro-peptide precursor (pro- MYOS) serum levels determined by commercial ELISA in patients with HF, and to analyse the possible relationship of such peptides with age, sex, several currently used parameters of HF severity, metabolic param- eters, and also with the prognosis. In addition the possible correlation of these peptides with TNF-alfa soluble receptor 2 (TNFR2) and with NT-proBNP serum levels was also evaluated. Patients: 70 patients (70% men, median age 72.3 years [34.9-82.7]), who where in NYHA functional class I (15), II (15), III (35) and IV (5) were included in the study. Median LVEF was 32% (range 8-72%). Etiology of HF was mainly ischemic heart disease (66%), followed by dilated cardiomyopathy (10%). Mean follow-up was 538±40 days. Results: 13 patients (18.6%) died during follow-up. Mean value of MYOS was 12.3±7.2 ng/ml. Median value of pro-MYOS was 0.1 ng/ml (range 0.1-15). Median value of TNFR2 was 2.42 ng/ml (range 0.08- 25.75). Median value of NT-proBNP was 899.5 pg/ml (range 57-25398). No significant correlation was found between MYOS and pro-MYOS. We did not found any relationship between MYOS or pro-MYOS and NYHA functional class groups. In contrast, both NT-proBNP (p<0.001) and TNFR2 (p=0.001) were significantly related to NYHA functional class. No correlation was found between MYOS or pro-MYOS with TNFR2 and NT-proBNP serum levels. A lack of significant correlation was observed between MYOS or pro-MYOS with age, sex, LVEF,etiol- ogy of HF, duration of HF symptoms, and glucose, total protein, albumin, creatinine, Tn I, urates or cholesterol serum levels. Pro-MYOS (r=0.24, p=0.04) but not MYOS correlated weakly with CK serum levels. Finally there was no relationship between MYOS (alive: 11.9±7.5; deceased: 14±5.4; p=0.3) or pro-MYOS (alive: median 0.1, range 0.1-11.4; de- ceased: median 0.1, range 0.1-15; p=0.39) and mortality. In contrast, both TNFR2 (p=0.005) and NT-proBNP (p<0.001) levels were signifi- cantly related with mortality. Conclusions: MYOS and pro-MYOS serum levels determined by com- mercial ELISA were unrelated with the currently used parameters of HF severity. In addition, MYOS and pro-MYOS seem not to be useful indicators of prognosis in the HF patients included in our study. 360 Chromogranin A and C-terminal endothelin-1 precursor fragment add independent prognostic information to amino-terminal proBNP in patients with acute destabilized heart failure B. Dieplinger 1 , A. Gegenhuber 2 , J. Struck 3 , W. Poelz 4 , W. Langsteger 5 , M. Haltmayer 1 , T. Mueller 1 1 Konventhospital Barmherzige Brueder Linz, Department of Laboratory Medicine, Linz, Austria; 2 Krankenhaus Bad Ischl, Department of Internal Medicine, Bad Ischl, Austria; 3 B.R.A.H.M.S AG, Research Department, Hennigsdorf/Berlin, Germany; 4 University of Linz, Institute for Applied System Sciences and Sta, Linz, Austria; 5 St. Vincent s Hospital Linz, Department of Nuclear Medicine and Endocrinol, Linz, Austria Objective: The aim of this study was to evaluate the prognostic value of chromogranin A (CgA) and C-terminal endothelin-1 precursor fragment (CT-proET-1) in patients with acute destabilized heart failure. Design: Outcome study. Setting: Tertiary care hospital. Patients: 137 consecutive patients with acute destabilized heart failure attending the emergency department were prospectively enrolled. Main outcome measures: Plasma concentrations of CgA, CT-proET-1, and amino-terminal proBNP (NT-proBNP) were measured at baseline. The endpoint was defined as all-cause mortality; the study participants were followed up for 365 days. Results: Decedents (n=41) had higher median plasma concentrations of CgA (9.7 vs. 6.0 nmol/L; p=0.002), CT-proET-1 (120 vs. 72 pmol/L; p=0.006), and NT-proBNP (5112 vs. 2610; p <0.001) at baseline than survivors (n=96). Applying univariate Cox proportional-hazards regres- sion analyses increased CgA (≥4.9 nmol/L), CT-proET-1(≥111 pmol/L), and NT-proBNP (≥5350 ng/L) revealed similar risk ratios of 2.84 (95% CI, 1.26-6.39) for CgA, 2.84 (95% CI, 1.54-5.24) for CT-proET-1, and 2.36 (95% CI, 1.28-4.33) for NT-proBNP. When our cohort was stratified according to increased CgA and NT-proBNP, and to CT-proET-1 and NT- proBNP, respectively, multivariable Cox proportional-hazards regression analyses showed the highest risk for death in patients with both increased CgA and NT-proBNP (risk ratio, 3.79; 95% CI, 1.36-10.57), and in patients with either increased CT-proET-1 or NT-proBNP or both (risk ratios, ≥3.2). Conclusions: Our study demonstrates that determination of CgA and CT-proET-1 plasma concentrations at the initial presentation of patients with acute destabilized heart failure in the emergency department adds independent prognostic information to NT-proBNP measurement. 361 Prognostic value of HSCRP in recently decompensated HF patients P. Lourenco 1 , C. Paulo 2 , J. Mascarenhas 2 , F. Frioes 3 , J.P. Araujo 3 , A. Azevedo 3 , P. Bettencourt 3 1 Oporto, Portugal; 2 Hospital S. Joao, Internal Medicine, Porto, Portugal; 3 Hospital S. Joao/Faculdade de Medicina Porto, Internal Medicine, Porto, Portugal Background: The prognostic value of high sensitivity C-reactive pro- tein (hsCRP) in heart failure (HF) patients is not well established. It is generally accepted that a higher hsCRP has a negative prognostic impact. Objectves: To study the prognostic value of hsCPR after an episode of acute HF. Methods: We prospectively recruited 278 patients admitted to our In- ternal Medicine Department for acute HF (de novo or decompensated chronic HF). All patients were given treatment to the discretion of the attending physician. The level of hsCRP was measured at discharge in 225 patients. The mean (SD) age was 72 (13) years, 64% were men, 64.7% had arterial hypertension, 42.2% were diabetic and 54.2% had ischemic heart disease. Median discharge hsCRP was 10.8 (range: 0.8 to 194.3mg/L). Patients were followed for 3 months. Endpoint was defined as all cause death or readmission. Infection during admission was defined by the cause of decompensation being infectious and/or the occurrence of an infectious complication during hospital stay. Within these groups patients were classified in three categories according to tertiles of hsCRP. One-hundred and sixteen patients had no infection detected (1st and 2nd tertiles of hsCRP: 5.0 and 12.2, respectively) and in the remaining 109 an infectious condition was present (1st and 2nd tertiles of CRP 8.8 and 27.1mg/L, respectively). The prognostic value of hsCRP was assessed by use of a Cox regression model. Results: In the infected group hsCRP was not a good predictor of adverse outcome. In the non-infected group the hazard ratio of those with hsCRP >12.2mg/L in comparison with those below 5.0mg/L was 2.4 (95% confidence interval: 1.1-5.2; p=0.034). After adjustment for BNP, the hazard ratio was 1.8 (95%CI: 0.8-5.2; p=0.13). Conclusions: hsCRP had no prognostic interest when HF patients had an infectious condition. In the non-infected group there was a 2.4 fold increase in risk of adverse outcome between extreme tertiles. However, the predictive power of hsCRP in the non-infected group was not inde- pendent from that of BNP at discharge. Our results do not support the use of hsCRP as a prognostic determinant in HF decompensation.