ABSTRACT
BACKGROUND: Monitoring of renal function is crucial in pediatric oncology. The use of creatinine to estimate glomerular filtration rate (GFR) is hampered by its dependency on muscle mass. Muscle wasting is common in children with cancer, leading to overestimation of GFR. Data on cystatin C are sparse in pediatric oncology, although this marker could be particularly useful in this population. PROCEDURE: Inulin clearance, estimated GFR using serum cystatin C according to Filler (eGFRcys) and serum creatinine according to Schwartz (eGFRcrea) were measured in 68 children with malignancy and 121 controls. We analyzed the difference between measured and estimated GFR and performance, bias and accuracy. RESULTS: Multiple linear regression analysis showed overestimation of GFR by eGFRcrea in females (B = -21.18; P = 0.001), and in patients with malignancy (B = -21.77; P = 0.014). eGFRcys overestimated GFR in females (B = -10.47; P = 0.001), but was independent of treatment for malignancy. Agreement with gold standard in detecting GFR below 90 ml/min/1.73 m(2) is better for eGFRcys (AUC 0.854) than for eGFRcrea (AUC 0.675) in the group with cancer. They performed comparably in the control group. Bland-Altman analysis showed considerable bias for eGFRcrea compared to eGFRcys (-14.3 ml/min/1.73 m(2) vs. -7.3 ml/min/1.73 m(2)). The proportion of estimates within 30% of true GFR for eGFRcrea (72.1%) was lower than for eGFRcys (82.4%) in the group with cancer. In the control group eGFRcrea (84.3%) outperformed eGFRcys (76.0%). When using the 50% limits of agreement, eGFRcys outperformed eGFRcrea in both groups. CONCLUSION: Cystatin C more accurately detects mildly impaired renal function than creatinine in children receiving treatment for malignancy.
