Creatinine-based formulas are not ideal to estimate glomerular filtration rate in selected pediatric patients: data from a tertiary pediatric nephrology center.

Abstract

Evaluating glomerular filtration rate (GFR) remains challenging in pediatrics; new formulas were developed to increase performance of GFR estimation (eGFR). We aimed to evaluate the recently published formulas as applied to another pediatric population. A retrospective study was conducted in a cohort of 307 patients with a "kidney risk" (mean age 12.1 ± 4.5 years, sex ratio 1/1) assessed in a tertiary pediatric nephrology center and a mean measured GFR (mGFR) using plasma iohexol clearance of 85.5 ± 25.3 mL/min/1.73 m2; creatinine levels were measured by IDMS-standardized enzymatic method and cystatin C by immunonephelometry. The following eGFRs were calculated: Schwartz2009, Schwartz-Lyon, CKiDU25creat, and EKFC for eGFR using creatinine (eGFR-creat), CKiDU25cys and FAScys for eGFR using cystatin (eGFR-cys) as well as combined SchwartzCreat-Cys, average (CKiDU25creat-CKiDU25cys), and average (EKFC-FAScys) for eGFR using both biomarkers. The performance of the different formulas was evaluated compared to mGFR by absolute bias measurement and accuracy (p10%, p30%). Results are expressed as mean ± SD. Creatinine-based formulas and especially the new CKiDU25 and EKFC overestimate GFR, even in children with normal kidney function. However, the bias is constant with these two formulas whatever the age group or gender, contrary to the previously published formulas. In contrast, cystatin C-based equations and combined formulas showed good performance in all age groups and all medical conditions with an acceptable bias and p30%. In our pediatric population, the performance of all creatinine-based formulas is inadequate with significant GFR overestimation, mainly in subjects with mGFR > 75 mL/min/1.73 m2. Conversely, cystatin C-based or combined formulas have acceptable performance in patients followed in a tertiary pediatric nephrology unit.