Surajith Nalantha Wanasundara PhD, Presenter: Nothing to Disclose

Michal John Wesolowski PhD, MSc, Abstract Co-Author: Nothing to Disclose

Mark C. Barnfield MSc, Abstract Co-Author: Nothing to Disclose

Michael Waller PhD, Abstract Co-Author: Nothing to Disclose

Anthony W. Murray BSC, MSc, Abstract Co-Author: Nothing to Disclose

Maria T. Burniston PhD, Abstract Co-Author: Nothing to Disclose

Paul S. Babyn MD, Abstract Co-Author: Nothing to Disclose

Carl Adam Wesolowski MD, FRCPC, Abstract Co-Author: Nothing to Disclose

The Tk-GV method using 4 plasma samples over 4-h can be used to accurately replace with very good precision those GFR values obtained from numerical integration with 9 samples over 8-h even when body fluid disturbances are suspected for a much less traumatic and labor intensive test especially for pediatric patients.

Glomerular filtration rate (GFR) values are calculated by dividing the administered dose by the total area under the plasma concentration curve (AUC) of a GFR marker. Measurement of GFR can assess renal function for a variety of clinical indications even when body fluid disturbances are suspected by using extensive time-sampling. However, it is clinically desirable to reduce the number of plasma samples and collection time required to calculate AUC without compromising its accuracy. This work presents calculation of GFR values of using 4 plasma samples obtained over 4-h that are as accurate as GFR values calculated using numerical integration (NAUC) from 9 plasma samples obtained over 8-h.

Following a single bolus injection of [99m]Tc-DTPA, 9 blood samples were obtained over 8-h, for 415 mostly pediatric studies in 150 subjects with suspected fluid disturbances. GFR values were calculated by two methods: 1) Numerical integration (NAUC) of the 9-sample, 8-h data including back-extrapolation from the first time samples and forward-extrapolation from the last time samples, and 2) Tikhonov regularized gamma variate fitting (Tk-GV) of 4 plasma sample data at approximately 5, 20, 60 and 240 min to adaptively minimize the relative error of the gamma variate’s exponential rate constant.

Passing-Bablok unbiased regression comparison showed GFR(Tk-GV, 4-samples, 4-h) to be equivalent to GFR(NAUC, 9-samples, 8-h) with a 4.0% interquartile range error. The GFR(NAUC) = GFR(Tk-GV) equivalence resulted from statistical indistinguishability with GFR(NAUC) = slope*GFR(Tk-GV)^power where the slope and power were statistically close to 1, i.e., the expected values and 95% confidence intervals were 0.990 (0.966 to 1.015) for the slope and 1.000 (0.994 to 1.006) for the power.

Wanasundara, S, Wesolowski, M, Barnfield, M, Waller, M, Murray, A, Burniston, M, Babyn, P, Wesolowski, C, Accurate Glomerular Filtration Rate Measurement Using Four Plasma Samples over Four Hours Despite Pathological Body Fluid Disturbances.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14005693.html