Cystatin C Serum Concentrations Underestimate Glomerular Filtration Rate in Renal Transplant Recipients

Abstract

Cystatin C, a cationic 13.3-kDa protein (1) has recently been described as a promising endogenous marker of glomerular filtration rate (GFR) (2) in both adults (2)(3) and children (4)(5). Correlation of serum cystatin C concentrations with the results of inulin and 51Cr-ethylenediaminetetraacetic acid clearance examinations was superior to serum creatinine (2)(3)(4). Although patients with a wide range of renal function were examined, the potential effect of underlying disease on this correlation has not been studied yet. In a recent longitudinal study on 24 pediatric renal transplant recipients, cystatin C serum concentrations were found to underestimate GFR compared with the results of a 24-h creatinine clearance (6). To validate these findings, we extended our previous study on the correlation between inulin clearance (CIn) and serum cystatin C concentration in children (4). One hundred and eighty-four children (87 girls, 97 boys) with a mean age of 11.2 ± 4.5 years (mean ± SD; range, 0.24–17.96 years) had undergone an inulin clearance examination between 1985 and 1995 at the Department of Pediatric Nephrology at Hannover Medical School. Forty-four of these (19 boys, 25 girls; mean age, 13.9 ± 3.2 years) had received a renal transplantation (RTx). Serum samples from CIn examinations had been stored at −20 °C for 6.9 ± 2.8 years (mean ± SD) until measurement of cystatin C, which has been shown to be stable for such extended periods of storage (4). Cystatin C was measured by particle-enhanced immunoturbidimetry (2)(3) using the Cystatin C PET-Kit (Dako) on a Hitachi 717 automated analyzer (Roche Diagnostics). Data are presented as means ± SE. Statistical analysis was performed with StatView Ver. 4.01 statistical software (Abacus Concepts) for Macintosh computers (Apple), using standard parametric tests as indicated. P values below 0.05 were considered statistically significant. Correlation was assessed by linear regression analysis and comparison of the slopes using parametric tests (7). Cystatin C serum concentrations were significantly higher in the RTx group compared with the control group (2.5 ± 0.1 vs 2.1 ± 0.1 mg/L; P = 0.002, unpaired t-test), although CIn was not different between the two groups (50.5 ± 2.5 vs 53.4 ± 2.4 mL · min−1 · 1.73 m−2; P = 0.423). Linear regression between the reciprocal of serum cystatin C and CIn (Fig. 1) showed a significantly lower slope of the regression line for the RTx patients (0.0046; 95% confidence interval, 0.0033–0.0059) than for the control group (0.0073; 95% confidence interval, 0.0060–0.0086; P = 0.002, two-sided t-test). Bivariate plot of CIn vs reciprocal of cystatin C in renal transplant (•) and nontransplant (○) patients. Linear regression analysis gave 1/cystatin C = 0.186 + 0.0046 · CIn (r = 0.76; P <0.0001) for renal transplant patients and 1/cystatin C = 0.149 + 0.0073 · CIn (r = 0.82; P <0.0001) for nontransplant patients. In the transplanted children, relDevCys differed significantly from zero (25.1% ± 2.7%; 95% confidence interval, 19.4–30.7%; P <0.0001, two-tailed t-test). Linear regression analysis failed to show a significant correlation between relDevCys and cyclosporin A trough concentrations (r = 0.12; P = 0.34), cyclosporin A dose (mg · m−2 · day−1; r = 0.13; P = 0.30) or prednisolone dose (mg · kg−1 · day−1; r = 0.188; P = 0.13) in those 66 children receiving immunosuppressive drugs. The present study confirms our previous finding that cystatin C serum concentrations are increased in RTx compared with non-RTx children having the same GFR. Estimation of GFR from serum cystatin C concentrations by use of a regression formula derived in non-RTx patients with similar renal function led to an underestimation of ∼25%. This may account for some of the scatter observed in the Bland-Altman analysis in our previous report (4), which demonstrated the limitations in estimating GFR from the reciprocal of cystatin C serum concentrations by linear regression with CIn. If such regression formulae are to be used in clinical practice, care must be taken for a specific calibration in RTx patients. Newman et al. (8) have also noted increased cystatin C concentrations in spite of creatinine values within the reference interval in adult renal transplant recipients on discharge from hospital. In contrast, Plebani et al. (9) found cystatin C concentrations within the reference interval in renal transplant recipients with CIn above 70 mL · min−1 · 1.73 m−2. Even at a GFR of 10–15 mL · min−1 · 1.73 m−2, they reported cystatin C concentrations of only 2.5 mg/L, which is one-half of the expected value for this degree of renal functional impairment (2)(3)(4)(10). In the present study, the slope of the regression line of the RTx group was significantly lower than the non-RTx group. Bearing in mind that the y-axis represents the reciprocal of cystatin C, this finding is suggestive of a constant increase in cystatin C concentrations independent of GFR. Several mechanisms