Echo-enhanced voiding urosonography for detection of vesicoureteric reflux in children.

Abstract

Received: 4 January 2003 Accepted: 18 January 2003 Published online: 14 February 2003 Springer-Verlag 2003 Sir, We read with great interest the paper by McEwing et al. [1] on the use of echo-enhanced sonography for detection of vesicoureteric reflux (VUR) in a selected group of children, and we would like to comment briefly on the methodology and results that we have obtained with this new and attractive technique. In a survey presented at the 8th Workshop on Sonographic Diagnosis of Vesicoureteric Reflux (Heidelberg, 2000), echo-enhanced voiding urosonography (EEVUS) had been used in over 20 centres worldwide. This non-ionising technique had been applied to various age groups and for various indications, mainly for investigation of children with urinary tract infection (UTI) or as a control study in patients with previously diagnosed VUR. EEVUS was compared with voiding cystourethrography (VCUG) in the majority of studies and the sensitivity of EEVUS varied between 67 and 100%. In a study from Slovenia and in our study, EEVUS was compared with direct radionuclide cystography (DRCG) [2, 3]. In our study, we investigated 40 children with 80 nephroureteral units (NUU) using EEVUS as the initial investigation; using the same catheter we subsequently performed DRCG. Both methods detected VUR in 18/ 80 NUU, both were negative in 48/ 80 NUU. DRCG alone was positive in 9/80 NUU and EEVUS alone was positive in 5/80 NUU (v=0.643, P=0.423). If DRCG was considered as the gold standard, then calculated sensitivity and specificity of EEVUS were 67 and 90%, respectively. If both techniques were considered referent, then DRCG had a sensitivity of 84% with a specificity of 85%, and EEVUS had a sensitivity of 72% with a specificity of 92%. In this series, EEVUS missed one grade-IV reflux in a girl. We concluded that DRCG should be performed through the same catheter only in those children with no VUR by EEVUS. In total, EEVUS detected VUR in 17 patients (23 NUU). If EEVUS was considered a screening method, then 23 patients should have the second study (DRCG) and this would not justify the high costs of EEVUS. The unsatisfactory results in our study were most likely due to dependence on the operator’s skill (‘‘learning curve’’) and on the equipment used during the study. Harmonic tissue imaging was not available on our US machines at the time of the study. The group from Heidelberg found excellent sensitivity (100%) of EEVUS compared with VCUG for detection of VUR when harmonic options were used [4, 5]. Similar to us [3], McEwing et al. [1] and Kenda et al. [2] used only conventional US imaging and this was probably the reason for low sensiPediatr Radiol (2003) 33: 286–287 DOI 10.1007/s00247-003-0881-1 LETTER TO THE EDITOR