P4669Quantification of fluid status using bioelectrical impedance spectroscopy (BIS): a predictor of outcome in patients with valvular heart disease

• Published in: European heart journal 2019-10, Vol.40 (Supplement_1)
• Main Authors: Dona, C, Kammerlander, A K, Karner, B K, Keil, A M, Seitlhuber, M S, Duca, F D, Nitsche, C N, Koschutnik, M K, Binder, C B, Aschauer, A S, Bonderman, D B, Hengstenberg, C H, Mascherbauer, J M
• Format: Article
• Language: English
• ISSN: 0195-668X; 1522-9645
• DOI: 10.1093/eurheartj/ehz745.1051

Abstract Background Volume overload, which may finally lead to cardiac decompensation, is a major threat in valvular heart disease (VHD) patients. In clinical practice, fluid overload is estimated by evaluation of leg edema, pulmonary congestion, weight gain or distension of jugular veins; however, these parameters lack both specificity and sensitivity. Bioelectrical impedance spectroscopy (BIS) is an easy, non-invasive and reliable way to determine the extent of fluid overload. BIS it broadly used in patients on chronic haemodialysis to guide therapy. Whether fluid status as measured by BIS is associated with outcome in VHD patients without obvious volume overload is unknown. Methods Stable patients with moderate or severe VHD as diagnosed by transthoracic echocardiography (TTE) underwent fluid status assessment by BIS at baseline and were prospectively followed. The primary endpoint was a composition of heart failure hospitalisation and cardiovascular death. Kaplan-Meier estimates and multivariable Cox-regression analysis were used to identify factors associated with outcome. This study was registered at clinicaltrials.gov (NCT03372512). Results 232 patients (46.6% female, 72±13 years) were included in the study. 23.7% suffered from aortic stenosis (2.5% moderate, 21.2% severe), 49.6% from mitral regurgitation (21.5% moderate, 28.1% severe). In 61.6% of the patients additional tricuspid regurgitation was present (23.2% moderate, 38.4% severe). Median overhydration (OH) was +0.6L, and patients were stratified according to this cut-off into two groups. Fluid status by BIS was not associated with diabetes (p=0.776), coronary artery disease (p=0.504), renal function (p=0.824), left ventricular ejection fraction (p=0.785), NYHA functional class (p=0.809), or leg edema (p=0.492). During a follow-up of 8.8±7.5 months a total of 85 events (36.6%) occurred. 71 patients (30.6%) underwent invasive treatment for VHD (either surgical or transcatheter) and were censored at the time of intervention. These were not treated as an event unless the primary endpoint occurred prior to valve intervention. Patients with fluid overload (OH≥0.6L) were more likely to experience an event (log-rank, p=0–013; Figure 1). By univariable Cox-regression fluid overload was significantly associated with outcome (per 1L: HR 1.152 [1.073–1.236]; p