Serum Cystatin C-Based Formulas for Prediction of Glomerular Filtration Rate in Patients with Chronic Kidney Disease

Background: The present study, involving a large group of patients with chronic kidney disease (CKD), compares different serum cystatin C-based equations for prediction of the glomerular filtration rate (GFR). Methods: A total of 592 adult patients with CKD were enrolled in the study. Serum cystatin C was determined in each patient by an immunonephelometric method. Their GFR was estimated using 5 equations based on serum cystatin C: (1) the Larsson formula, (2) the Hoek formula, (3) the Grubb formula, (4) the simple cystatin C formula (GFR = 100/cystatin C) and (5) our own cystatin C formula (GFR = 90.63 x cystatin C-1.192). The actual GFR was measured using (51)CrEDTA clearance. Results: The mean (51)CrEDTA clearance was 47 ml/min/1.73 m(2); the mean serum cystatin C concentration was 2.68 mg/l. Receiver operating characteristic curve analysis (cutoff for GFR: 60 ml/min/1.73 m(2)) showed no difference between the cystatin C formulas with regard to diagnostic accuracy. All equations underestimated the measured GFR except the simple cystatin C formula, which slightly overestimated the measured GFR. All equations lacked precision. The ability to correctly estimate the patient's GFR was high for all equations (87.3-91.9%), except for the Larsson formula, with which 29.2% of subjects were misclassified. Conclusions: Our results indicate that all serum cystatin C-based equations, excluding the Larsson formula, are reliable markers of the GFR in patients with CKD, and for daily clinical practice the simplest formula (100/cystatin C) could be accurate enough for GFR estimation. Copyright (C) 2009 S. Karger AG, Basel