Abstract
Background and Objectives Cr-51 EDTA clearance (CL) from plasma is used to estimate glomerular filtration rate (GFR). We propose that current methods for analysing the raw Cr-51 EDTA measurements over-simplifies the disposition of Cr-51 EDTA and therefore could produce biased GFR estimates. The aim of this study was to develop a population pharmacokinetic model for Cr-51 EDTA disposition and to compare model-predicted GFR to other methods of estimating renal function. Patients and Methods Data from 40 individuals who received similar to 7.4 MBq of Cr-51 EDTA, as an intravenous bolus, were available for analysis. Plasma radioactivity (counts/min) was measured from timed collection points at 2, 4, 6 and 24 h after the dose. A population analysis was conducted using NONMEM (R) version 7.2. Model-predicted GFR was compared with other methods for estimating renal function using mean prediction error (MPE). Results A two-compartment pharmacokinetic model with first-order elimination best fit the data. Compared with the model predictions, creatinine CL from 24 h urine data was unbiased. The commonly used 'slope-intercept' method for estimating isotopic GFR was positively biased compared with the model (MPE 15.5 mL/min/1.73 m(2) [95% confidence interval {CI} 8.9-22.2]. The Cockcroft Gault, Modification of Diet in Renal Disease (MDRD) and Chronic Kidney Disease Epidemiology Collaboration (CKD-Epi) equations led to negatively biased GFR estimates (MPE - 19.0 [95% CI - 25.4 to - 12.7], -20.1 [95% CI -27.2 to -13.1] and -16.5 [95% CI -22.2 to -10.1] mL/min/1.73 m(2), respectively). Conclusions The biased GFR estimates were most obvious in patients with relatively normal renal function. This may lead to inaccurate dosing in patients who are receiving drugs with a narrow therapeutic range where dosing is adjusted according to GFR estimates (e.g.carboplatin).