Correlation of various published radionuclide glomerular filtration rate estimation techniques and proposed paediatric normative data

Abstract

The aim of this study is to assess the comparability and interchangeability of the radionuclide glomerular filtration rate (GFR) using different published techniques, and propose normative data for paediatrics. A total of 476 paediatric oncology patients aged 2-17 years, referred between January 2001 and December 2008 for GFR estimation, were reviewed for any potential cause of renal impairment. Sixty-nine patients met the stringent inclusion criteria, and were included in the study. GFR estimation was carried out using either technetium-99m diethylene triamine penta-acetic acid (⁹⁹mTc-DTPA) or chromium-51 EDTA (⁵¹Cr-EDTA). Multiple GFR results were calculated from the same blood sample data (counts/min/ml), according to previously published GFR estimation techniques using one to three blood samples. These techniques were slope-intercept, slope-only and half life. For slope-intercept techniques, GFR was normalized to body surface area or extracellular fluid volume. The GFR values obtained using different techniques were highly variant. The intraclass correlation (ICC) for different methods was moderate (ICC=0.56-0.66). A reliable empiric formula to allow conversion of GFR values from one technique to another could not be derived because of this variability, with some exceptions. ⁵¹Cr-EDTA yielded the same or lower variability than ⁹⁹mTc-DTPA. The British Nuclear Medicine Society-recommended method had the lowest coefficient of variation, with a mean value of 116 (SD 22) normalized to 1.73 m² for ⁵¹Cr-EDTA using two samples. The GFR values obtained from different calculation techniques are not readily interchangeable or comparable, with some exceptions. For both ⁹⁹mTc-DTPA and ⁵¹Cr-EDTA, the British Nuclear Medicine Society-recommended technique appears to be the most robust, with the least coefficient of variation.