Extracellular Fluid Volume and Mortality after Kidney Transplantation

Abstract

Key Points:

1. Post-transplantation extracellular fluid volume (ECV) is not only influenced by recipient characteristics and allograft function, but also by transplantation-specific factors.

2. In multivariable cause-specific Cox analyses, increased ECV 3 months after kidney transplantation is independently associated with all-cause mortality but not graft loss.

3. Multivariable linear regression further identified increased 3-month ECV as an independent predictor of reduced GFR measurement at 12 months.

Background: High extracellular fluid volume (ECV) is associated with an increased risk of death in nontransplanted patients with CKD. By contrast, both the determinants and the prognosis value of ECV in kidney transplant recipients remain unclear.

Methods: We studied a bicentric prospective cohort of 2057 kidney transplant recipients who underwent GFR measurement (mGFR) 3 months after transplantation. We calculated ECV from iohexol plasma disappearance curve and analyzed its association with patient's characteristics and outcomes.

Results: The mean ECV and mGFR were 14.6±2.6 L/1.73 m² and 52±16 ml/min per 1.73 m², respectively. Multiple linear regression identified male sex, older donor and recipient ages, deceased donor, nonpreemptive transplantation, diabetes, cardiac arrhythmia, heart failure, proteinuria, and higher mGFR as independent factors associated with increased ECV. In multivariable cause-specific Cox analyzes, higher tertiles of ECV were associated with increased mortality (tertile 1 as reference; hazard ratio [95% confidence interval] for tertile 2: 1.65 [1.11 to 2.47]; tertile 3: 1.80 [1.18 to 2.74]) but not graft loss. Increased ECV, 3 months after transplantation, was a predictor of reduced mGFR at 12 months after adjusting for 3-month mGFR and other confounding factors (β coefficient: −0.06; 95% confidence interval [−0.09 to −0.02]).

Conclusions: An elevated ECV 3 months after kidney transplantation is independently associated with increased mortality and decreased mGFR at 12 months, but not with graft loss.

Graphical abstract

Figure 1

Correlation between ioECV and bisECV. Linear regression (A) and Bland–Altman analysis (B) of comparison of ECV measured with plasma ioECV or bisECV. Difference between ioECV and bisECV according to ioGFR (C). bisECV, bio-impedance spectroscopy; ECV, extracellular fluid volume; ioECV, iohexol disappearance curve.

Figure 2

Flow chart of the patients studied.

Figure 3

ioECV associates with a greater incidence of mortality but not allograft loss in KTRs. Cumulative incidence plots of ioECV/BSA tertiles and the composite outcome of all-cause mortality and allograft loss (A), all-cause mortality (B), and allograft loss (C). KTR, kidney transplant recipient.

Figure 4

Adjusted HRs and 95% CIs for the association of ioECV and mortality using penalized-splines estimator. CI, confidence interval; HR, hazard ratio.