Abstract
BACKGROUND: Worsening serum creatinine is common during treatment of acute decompensated heart failure (ADHF). A possible contributor to creatinine increase is diuresis-induced changes in volume of distribution (VD) of creatinine as total body water (TBW) contracts around a fixed mass of creatinine. Our objective was to better understand the filtration and nonfiltration factors driving change in creatinine during ADHF. METHODS: Participants in the ROSE-AHF trial with baseline to 72-hour serum creatinine; net fluid output; and urinary KIM-1, NGAL, and NAG were included (n=270). Changes in VD were calculated by accounting for measured input and outputs from weight-based calculated TBW. Changes in observed creatinine (Cr(observed)) were compared with predicted changes in creatinine after accounting for alterations in VD and non-steady state conditions using a kinetic GFR equation (Cr(72HR Kinetic)). RESULTS: When considering only change in VD, the median diuresis to elicit a ≥0.3 mg/dl rise in creatinine was -7526 ml (IQR, -5932 to -9149). After accounting for stable creatinine filtration during diuresis, a change in VD alone was insufficient to elicit a ≥0.3 mg/dl rise in creatinine. Larger estimated decreases in VD were paradoxically associated with improvement in Cr(observed) (r=-0.18, P=0.003). Overall, -3% of the change in eCr(72HR Kinetic) was attributable to the change in VD. A ≥0.3 mg/dl rise in eCr(72HR Kinetic) was not associated with worsening of KIM-1, NGAL, NAG, or postdischarge survival (P>0.05 for all). CONCLUSIONS: During ADHF therapy, increases in serum creatinine are driven predominantly by changes in filtration, with minimal contribution from change in VD.