Validation of numerical outputs of IAEA software by the analysis of diuretic nephrogram in children with antenatally detected hydronephrosis.

UNLABELLED: Diuretic nephrogram is important diagnostic tool in the postnatal follow-up of asymptomatic antenatally detected hydronephrosis (HN). In the last decades, two quantitative indices of renal excretion, output efficiency and the residual kidney counts normalized to the 1-2min counts (normalized residual activity, NORA) have been proposed, that enhance the accuracy of technique to detect kidneys with obstruction. Unfortunately, in many nuclear medicine departments in developing countries the obsolete computer systems do not give the opportunity of sophisticated analysis of nephrogram. Almost a decade ago, the Nuclear Medicine Section of the International Atomic Energy Agency (IAEA) has developed non-commercial software for nephrogram processing on a simple p-computer, which allows access to the developments in this field. However, till now, the software has not been widely implemented in the nuclear medicine institutions in developing countries. Furthermore, the accuracy of numerical outputs of the software has not been assessed in comparison with commercial software. The aims of this study in children were: a) to calculate, by means of the International Atomic Energy Agency (IAEA) software, the values of the technetium-99m mercapto-acetyl-triglycine ((99m)Tc MAG3) parameters in three categories of kidneys: normal kidneys, obstructed kidneys and hypotonic unobstructed kidneys and b) to assess the accuracy of the obtained numerical parameters by comparing with the values published by other authors. Investigation was carried out on a sample of 62 children: 43 boys and 19 girls (median age: 16 months) with antenatally detected HN attributed to pelviureteric junction (PUJ) stenosis. Neither of kidneys had undergone pyeloplasty prior to our investigation. 130 nephrogram curves were analyzed. 22-minutes acquisition with 132 10sec images was applied. Furosemide was administered after 2min (F+2). Post-void static image was acquired at 60min. Two observers analyzed each study and classified kidneys into three categories. Group 1: 84 kidneys contralateral to hydronephrotic kidney, without structural abnormality on previous diagnostics; Group 2: 30 hypotonic non-obstructed kidneys; Group 3: 16 obstructed kidneys. Parameters analyzed were: output efficiency (OE), residual kidney counts at 20min normalized to the 1-2min counts (NORA20) and residual kidney counts on post-micturition acquisition normalized to the 1-2min counts (NORAPM). Results were presented as mean±SD. For group 1 they were: OE: 95±1.5%; NORA20: 0.25±0.06; NORAPM: 0.02±0.007. Results for group 2 were: OE: 87±7.8%; NORA20: 0.57±0.19; NORAPM: 0.03±0.02. For group 3: OE: 56±9.6%; NORA20: 2.16±0.33; NORAPM: 0.27±0.13. Linear regression analysis showed significant inverse linear correlation between NORA20 and ROE20 (R=-0.982; y=99.6-21.1x) at 0.01 level. ROC analysis revealed cutoff values of predicting obstruction at 71%, 1.62 and 0.11 for OE, NORA20 and NORAPM, respectively.

CONCLUSION: We have calculated in children by means of the IAEA software the values of three advanced parameters of the (99m)Tc MAG3 F+2 diuresis nephrogram for normal kidneys, hypotonic non-obstructed and obstructed kidneys. The overall results provided evidence of excellent agreement of obtained results with previously reported values of the quantitative parameters of renal washout. The parameters of IAEA software has been shown to be reliable in assessing kidney drainage. The nuclear medicine section of the IAEA should be encouraged to produce final version of the software and to release it through IAEA Web site.