Clearance of iohexol, chromium-51-ethylenediaminetetraacetic acid, and creatinine for determining the glomerular filtration rate in pigs with normal renal function: comparison of different clearance techniques

RATIONALE AND OBJECTIVES: We wanted to improve determination of the glomerular filtration rate (GFR) with plasma clearance techniques because the alternative-renal clearance techniques-may involve inaccurate urine sampling or risk of urinary tract infection when bladder catheterization becomes necessary. Therefore, we compared the renal and plasma clearances of iohexol and chromium-51-ethylenediaminetetraacetic acid (51Cr-EDTA), as well as endogenous creatinine clearance, in 19 normal pigs using different techniques. METHODS: After an intravenous bolus injection of the GFR markers, 16 plasma samples were used to plot the marker concentrations versus time for 4.5 hr. Urine was collected during nine 30-min periods. Plasma clearance was calculated by dividing the dose of marker with the area under the plasma concentration curve (AUC) from the time of injection to infinity using one-compartment (ClAUC-slope) and three-compartment (ClAUC-3comp) models. The renal clearance was calculated by dividing the amount of marker excreted in the urine in a period with the AUC in the same period. This AUC was determined by integrating the total area in the period (Clren adv)--our reference method representing the "true" GFR--or by using the arithmetic mean of the plasma concentrations of the marker at the beginning and end of the urine collection period (Clren simple). Creatinine clearance was determined according to Clren simple. RESULTS: Renal clearances of iohexol and 51Cr-EDTA were significantly higher than creatinine clearance (P = .0002). There was no significant difference between the renal clearances of iohexol and 51Cr-EDTA or between their plasma clearances. The two mathematical methods of calculating the renal clearance of iohexol were highly correlated (rs = .99), as were the two methods of calculating its plasma clearance (rs = .95). Because of the extrarenal clearance of the markers, the plasma clearance methods for iohexol and 51Cr-EDTA always overestimated the true GFR. ClAUC-3comp was the method closest to the true GFR. For iohexol, the median overestimation of the GFR was higher with ClAUC-slope when early plasma samples (30-120 min) after injection of the marker were used (5.5 ml.min-1.10 kg-1) than when late samples (180-270 min) were used (4.0 ml.min-1.10 kg-1). After subtracting the median extrarenal clearances of iohexol and 51Cr-EDTA (previously determined in nephrectomized pigs) from their plasma clearances (ClAUC-3comp), the median overestimation of the true GFR was reduced from 2.0 to 1.1 ml.min-1.10 kg-1 with iohexol and from 2.1 to 1.3 ml.min-1.10 kg-1 with 51Cr-EDTA. CONCLUSION: GFR determination with plasma clearance techniques can be improved in three- and one-compartment models by taking late plasma samples and by subtracting the extrarenal plasma clearance of the species. One-compartment models can be improved by determining a correction formula in the species for the early parts of the decay curve of the plasma concentration of the marke