Why is Diuresis-related Weight Loss Not a Consistent Predictor of Clinical Outcomes in Heart Failure? -need to Account For Variability in Volume Status

Abstract

Introduction Change in body weight is a common metric used to assess adequacy of decongestive therapy in patients hospitalized for heart failure (HF). However, the findings of clinical studies linking this metric to clinical decongestion and outcomes are mixed. We propose this reflects the failure to take into account biologic variability in intravascular volume status which would impact response to therapy. Methods Data were prospectively collected in hospitalized patients requiring diuresis. Plasma volume (PV) was measured using I-131-labelled albumin indicator-dilution methodology. The cohort was stratified by median diuresis-related weight loss and analyzed for 1-year outcome of HF-related mortality or 1st re-hospitalization, and the relation of change in weight to change in PV admission to discharge. Results The cohort was composed of 138 patients with median weight loss of 6 (IQR -3.2, -8.1) kg during hospitalization (5 IQR 4, 6 days). Greater than median weight loss (Figure 1) did not result in better outcomes. At the time of hospital discharge the subgroup with greater weight loss (N=61) paradoxically demonstrated greater absolute (4.8±1.0 vs. 4.2±1.0 liters, p<0.001) and relative PV expansion(+44±29% vs. +33±23% above normal, p=0.023). Also demonstrated was marked heterogeneity in PV status in both subgroups (Figure 2). Volumes for the cohort ranged from 17% with normal PV, 23% mild-moderate PV expansion, to 60% with persisting severe (≥ +25%) PV expansion at hospital discharge. In a cohort subgroup of 62 patients where PV was measured at hospital admission and discharge there was no clinically significant association between change in weight and change in PV (r=0.327, p=0.01). Conclusions Diuresis-related changes in weight reflect clinical decongestion but do not consistently reflect better HF outcomes and, importantly, do not accurately reflect changes in intravascular volume. A basis for this discordance is the significant patient-to-patient variability in volume both at admission and discharge which confounds interpretation of surrogate metrics of decongestion. Failure to account for the marked heterogeneity in volume status contributes to the observed discordance between metrics of clinical decongestion and HF outcomes. Change in body weight is a common metric used to assess adequacy of decongestive therapy in patients hospitalized for heart failure (HF). However, the findings of clinical studies linking this metric to clinical decongestion and outcomes are mixed. We propose this reflects the failure to take into account biologic variability in intravascular volume status which would impact response to therapy. Data were prospectively collected in hospitalized patients requiring diuresis. Plasma volume (PV) was measured using I-131-labelled albumin indicator-dilution methodology. The cohort was stratified by median diuresis-related weight loss and analyzed for 1-year outcome of HF-related mortality or 1st re-hospitalization, and the relation of change in weight to change in PV admission to discharge. The cohort was composed of 138 patients with median weight loss of 6 (IQR -3.2, -8.1) kg during hospitalization (5 IQR 4, 6 days). Greater than median weight loss (Figure 1) did not result in better outcomes. At the time of hospital discharge the subgroup with greater weight loss (N=61) paradoxically demonstrated greater absolute (4.8±1.0 vs. 4.2±1.0 liters, p<0.001) and relative PV expansion(+44±29% vs. +33±23% above normal, p=0.023). Also demonstrated was marked heterogeneity in PV status in both subgroups (Figure 2). Volumes for the cohort ranged from 17% with normal PV, 23% mild-moderate PV expansion, to 60% with persisting severe (≥ +25%) PV expansion at hospital discharge. In a cohort subgroup of 62 patients where PV was measured at hospital admission and discharge there was no clinically significant association between change in weight and change in PV (r=0.327, p=0.01). Diuresis-related changes in weight reflect clinical decongestion but do not consistently reflect better HF outcomes and, importantly, do not accurately reflect changes in intravascular volume. A basis for this discordance is the significant patient-to-patient variability in volume both at admission and discharge which confounds interpretation of surrogate metrics of decongestion. Failure to account for the marked heterogeneity in volume status contributes to the observed discordance between metrics of clinical decongestion and HF outcomes.