Neurohumorale Aktivierung in einem kardiovaskulären Risikokollektiv - Einfluss von diastolischer oder systolischer Dysfunktion

Abstract

Diastolic dysfunction (DD) and especially diastolic heart failure (DHF) are characterized by high prevalence and mortality. According to newer studies, up to 50% of patients suffering from symptoms of heart failure (HF) show a preserved left ventricular ejection fraction (LVEF). Unfortunately, the diagnostic process for detecting DD using echocardiography is not well standardized and still under investigation. A serological marker would be helpful to reliably detect and monitor these entities. Therefore, we studied stable fragments derived from the synthesis of the neurohumoral peptides BNP (NT-proBNP), ANP (MR-proANP), Adrenomedullin (MR-proADM), Endothelin-1 (CT-proET-1) and Vasopressin (CT-proAVP). We conducted the multi centre, prospective and community-based cohort study Diast-CHF consisting of 1687 patients between 50 and 85 years with at least one risk factor for DD or a history of heart failure. For each patient with preserved left ventricular function (LVEF ≥50%) the grade of DD was determined using a scheme established within the German Competence Network on Heart Failure. The analysis of the serological parameters was performed using previously published methods (B.R.A.H.M.S. AG Hennigsdorf, Germany). The mean age of all included patients was 65.98 years, mean BMI was 29.12 kg/m². The cohort consisted of 49.9% men and 50.1% women. Arterial hypertension was the most frequent risk factor (78.7%). Echocardiographically determined E/A ratio and E/e ratio was 0.94 and 13.42, respectively. The grades of DD (LVEF≥ 50%) were distributed as follows: no DD (grade 0): 16.5%, DD grade I: 57.1%, DD grade II-III: 14.7%. In 3.7% of all patients DD could not be classified according to the given scheme. A LVEF <50% was observed in 8.0% of all patients. Higher grades of DD were characterized by significantly higher serum levels of the peptides: NT-proBNP: 120.10 pg/ml (grade 0), 141.76 pg/ml (grade I), 228.51 pg/ml (grade II-III), MR-proANP: 89.90 pmol/l, 96.22 pmol/l, 120.38 pmol/l. MR-proADM: 0.54 nmol/l, 0.61 nmol/l, 0.62 nmol/l. CT-proET-1: 53.89 pmol/l, 56.98 pmol/l, 58.58 pmol/l. CT-proAVP: 4.76 pmol/l, 5.81 pmol/l, 5.86 pmol/l. Patients with LVEF <50% had the significantly highest serum levels: NT-proBNP: 727.58 pg/ml, MR-proANP: 169.19 pmol/l, MR-proADM: 0.74 nmol/l, CT-proET-1: 68.88 pmol/l, CT-proAVP: 9.22 pmol/l. Furthermore, patients suffering from symptoms of HF showed augmented serum peptide levels compared to asymptomatic patients. This elevation was significant in all cases except for CT-proAVP. Regarding important echocardiographic parameters of DD there was a significant positive correlation between E/e and the peptides with CT-proAVP being the only exception. The E/A ratio showed a positive significant correlation with NT-proBNP and MR-proANP while the correlation with MR-proADM, CT-proET-1 and CT-proAVP was also significant but negative. In addition, we performed a ROC-analysis in order to describe the diagnostic accuracy of the peptides for detecting and discriminating certain grades of cardiac dysfunction. While MR-proADM yielded the highest area under the curve (AUC) for detecting any grade of DD (0.62) and DD grade I (0.62), NT-proBNP was superior for detecting patients with reduced LVEF (AUC 0.75). Both peptides showed an AUC of 0.62 when discriminating between grade I and grade II-III of DD. This study is the first to describe the changes of serum levels of different neurohumoral markers in presence and progression of DD and their possible role as surrogate parameters in a large and not preselected sample. The major findings yielded by this study were: In presence of DD the serum levels of the mentioned peptides were significantly elevated while higher grades of DD were characterized by significantly higher peptide levels. Furthermore, the serum levels correlated significantly with both the presence of HF symptoms and echocardiographically determined and prognostically relevant parameters of DD. However, the peptide fragments yielded relatively low AUC for detecting cardiac dysfunction and were unable to reliably detect cardiac dysfunction when used exclusively.