Predictive value of overt and non-overt volume overload in patients with high- or low-gradient aortic stenosis undergoing transcatheter aortic valve implantation.

Abstract

The plasma volume status (PVS) is considered a marker of non-overt cardiac congestion and is of prognostic value. Patients with low-flow, low-gradient (LFLG) aortic stenosis (AS) suffer from impaired left ventricular function and show signs of heart failure (HF). We hypothesized that PVS might predict post-interventional rehospitalization and cardiovascular mortality in high-risk patients undergoing transcatheter aortic valve implantation (TAVI). In this retrospective, observational analysis, PVS before transfemoral TAVI was calculated by a formula taking into account hematocrit and weight. The predictive performance of PVS was compared with that of prior cardiac decompensation (PCD). In the entire cohort of n=2,458 patients, PVS >-4% (high plasma volume) identified patients (n=1,013) with a higher post-interventional 1-year mortality rate than patients (n=1,445) with a PVS ≤-4% (low plasma volume). However, PVS lost prognostic independence when adjusted for anemia, whereas PCD did not. A high PVS and PCD were not correlated, and both parameters similarly revealed a low sensitivity and specificity but a high negative predictive value (NPV) for future HF events. PVS was not different between control patients (n=1,512) and those with intermediate (paradoxical LFLG-AS, n=327) or high risk scores (LFLG-AS, n=239). The accuracy of high PVS in predicting adverse events in these subpopulations was the same as in the study population overall. Kaplan-Maier analyses demonstrated similar prognostic impacts for PVS and PCD. PVS and PCD represent two independent parameters of volume overload with unfavorable prognostic significance. Pre-interventional PVS does not appear to be suitable for predicting clinical outcomes in high-risk patients undergoing TAVI.