Comparison of GFR estimation equations using creatinine, cystatin C, and their combination in pediatric hematology-oncology: no single equation is superior across subgroups.

Abstract

Background: Accurate assessment of renal function is essential in treating pediatric patients dosed with nephrotoxic chemotherapy. The validity of the bedside Schwartz, 5-covariate St. Jude (5SJ), CKiD-CysC-U25, combined Cr-CysC-based CysPed, and the serum creatinine-BUN-cystatin C-based CKiD (CKiD Cr-CysC) equations were evaluated in pediatric hematology and oncology patients.

Methods: A retrospective analysis was conducted comparing estimated glomerular filtration rate (eGFR) to measured GFR (mGFR) obtained from technetium- 99 m diethylenetriaminepentaacetic acid (^{99 m}Tc-DTPA) clearance between January 2016 and May 2022. The influence of corticosteroid use and inflammation in our patient population was evaluated for effect on serum cystatin C (CysC) concentrations and mGFR.

Results: All equations agreed within 2 SD of the mean difference with mGFR, but the 5SJ equation had the smallest bias followed closely by the CysPed equation. Overall accuracy (P30) was assessed, and the 5SJ, CKiD Cr-CysC, CysPed, and CKiD-Cys-U25 exhibited comparable performance. In our patient population, we did not observe an effect of corticosteroids (cumulative dosage of > 0.5 mg/kg within the past 14 days) or the presence of inflammation (CRP > 1.2 mg/L) on cystatin C concentrations or mGFR.

Conclusions: In our pediatric hematology and oncology patient population, no one estimating equation demonstrated superior accuracy and bias overall and in all subgroups. Neither corticosteroid use nor elevated CRP influenced serum CysC concentrations or eGFR.