PERFORMANCE OF CREATININE AND CYSTATIN-BASED ESTIMATES OF GFR IN PAEDIATRIC KIDNEY TRANSPLANT RECIPIENTS. COMPARISON OF THE CKD-EPI AGE- ADJUSTED CREATININE EGFR WITH PREVIOUS FORMULAS

Abstract BACKGROUND AND AIMS Estimation of glomerular filtration rate (GFR) is essential for renal function evaluation and the surveillance of kidney transplant recipients. We aimed to compare the GFR estimation by different formulas using creatinine and/or cystatin C, in paediatric kidney transplant recipients and to compare the chronic kidney disease-epidemiology collaboration (CKD-EPI) equation with age-adjusted creatinine values for height-independent estimated GFR (eGFR) calculation. We also intend to evaluate the impact of body mass index (BMI) influence on eGFR. METHOD Data on paediatric renal transplant patients (n = 35) with stable graft function for more than 6 months post-transplantation were enrolled. eGFR was calculated using the revised Schwartz (bedside Cr), CKD-EPI (Cr), CKiD (CysC), Zappitelli (CysC), combined Schwartz (Cr/CysC) and combined Zappitelli (Cr/CysC) formulas. The estimation by CKD-EPI formula used age and sex-adjusted creatinine values to adult levels with 40 as assigned age (CKD-EPI40). RESULTS A total of 35 patients were included, 26 (76%) were male, with a median age of 16 (4–18) years and a median BMI of 21.27 (15.20–32.29) with a Z-score median of 3.26 (2–5). The median eGFR obtained by the Schwartz [64.39 (23.96–115.16) mL/min/1.73 m2] and by the CKD-EPI40 equation [64.99 (13.98-141.80) mL/min/1.73 m2] was similar (P = .176). When compared with the Schwartz estimation, Cys-based eGFR were significantly lower (by CKiD-CysC, 49.48 (25.47–84.68) (P < .001), by Zappitelli formula 51.31 (24.74–92.61) (P < .001), by combined Schwartz 51.37 (24.07–85.31), (P < .001), by combined Zappitelli 54.83 (20.64–109.68).The correlations between eGFR by all formulas were strong and statistically significant, especially between the combined Schwartz and the combined Zappitelli (r = 0.97). The same pattern of differences between the eGFR estimations was observed when BMI groups (normal versus overweight) were analysed separately, with all formulas being significantly different for Schwartz except for CKD-EPI. CONCLUSION As expected, formulas with creatinine were similar to each other but significantly different when compared with formulas that included cystatin. As demonstrated in previous studies, the combined formulas appear to be the closest estimates of the GFR calculated by exogenous methods. Thus, estimates that consider only creatinine may overestimate the GFR. The eGFR obtained by the adjusted CDK-EPI formula is similar to that obtained by the Schwartz bedside, unlike the other formulas. In this study, we did not find differences regarding the BMI, probably due to the number of this sample.