Measurement of glomerular filtration rate using nonradioactive Iohexol: comparison of two one-compartment models

Radioisotopic methods for the determination of the glomerular filtration rate (GFR) are highly accurate but require the collection of multiple blood and urine samples and are costly to perform due to personnel, material, and analysis costs. Nonradioactive methods of GFR determination have the potential of minimizing procedure costs while preserving accuracy. We determined the GFR simultaneously by 125I-iothalamate and nonradioactive iohexol clearance methods in 41 adults. The study group consisted of 54% males, with a mean age of 50.7 (range 28-79) years and a mean GFR by 125I-iothalamate clearance of 66.5 +/- 28.3 (range 10-118) ml/min. The iohexol concentrations were measured by a simplified high-performance liquid chromatography method that did not require sample preparation. The iohexol plasma clearance was calculated by both a new one-compartment model as well as by Jacobsson's one-compartment model. Using Jacobsson's single-sample model and data from the 240-min point, there was an excellent correlation between 125 I-iothalamate and nonradioactive iohexol clearance values: r2 = 0.95, standard error of the estimate = 11.4 ml/min, and intrapatient coefficient of variation = 16.9%. However, this formula tended to overestimate GFRs < 30 ml/min and to underestimate GFRs > 80 ml/min. The new one-compartment model is a modification of Bubeck's model, originally used for the determination of renal plasma blood flow. Using this modified model, there was an excellent correlation between 125I-iothalamate and nonradioactive iohexol clearance values at all levels of GFR tested: r2 = 0.95, standard error of the estimate = 9.2 ml/min, and intrapatient coefficient of variation = 13.7%. In conclusion, the determination of the plasma clearance of iohexol by a nonradioactive technique and a monoexponential model is a simple and accurate method of determining the GFR in patients with varying degrees of renal impairment.