Multiparametric approach to congestion for predicting long-term survival in heart failure

Abstract

BackgroundCongestion is a marker of adverse prognosis in patients with heart failure (HF). In addition to brain natriuretic peptide (BNP), estimated plasma volume status (ePVS), bioimpedance vector analysis (BIVA), and blood urea nitrogen/creatinine ratio (BUN/Cr) are emerging as new markers for congestion. The aim of this study was to evaluate the prognostic value of BNP, ePVS, BIVA, and BUN/Cr in HF. MethodsWe analyzed the data from 436 patients with acute or chronic heart failure (AHF, n=184, and CHF, n=252, respectively). BNP, ePVS, hydration index (HI%), and BUN/Cr were collected from all patients at admission. The endpoint was all-cause mortality. ResultsNinety-two patients died after a median follow-up of 463 days (IQR: 287–669). The cumulative mortality of all of the patients was 21% (31% and 13% in AHF and CHF, respectively, p<0.0001). The optimal cut-offs for death occurrence were BNP: >441pg/mL, ePVS: >5.3dL/gr, HI: >73.8%, BUN/Cr: >25. Multivariate Cox regression analysis maintained an independent predictive value for mortality (HR 2. 1, HR 2.2, HR 2.1, and HR 1.7; C-index 0.756). AHF status was no longer associated with death. Together, these variables explained 40% of the risk of death (R2 adjusted=0.40). Patients with all four parameters below or above their optimal cut-off had mortality rates of 4% and 59%, respectively. ConclusionsBNP, ePVS, BIVA, and BUN/Cr at admission provide independent and complementary prognostic information in patients with HF and, when combined, explain the 40% risk of death in these patients independent from the acute or chronic HF condition.