Antenatal renal pelvis dilatation: a predictor of vesicoureteral reflux?

Abstract

The purpose of our study was to evaluate the role of antenatally diagnosed renal pelvis dilatation (RPD) as a predictor of vesicoureteral reflux (VUR) in infants. All cases of RPD (i.e., those with a renal pelvis diameter equal to or greater than 5 mm) detected on routine antenatal sonography at our institution over a 3-year period were followed throughout infancy with serial sonographic evaluation. Infants with moderate to severe RPD on the initial postnatal study (i.e., those with a renal pelvis diameter greater than 15 mm) were further evaluated with voiding cystography or nuclear scintigraphy (99mTc-2,3-dimercaptosuccinic acid or 99mTc-benzoylmercaptoacetyltriglycerine). We have now reviewed all postnatal imaging on these infants. In addition, the records of all infants who presented to our institution with clinically symptomatic VUR during this same period have been retrieved. Antenatal sonograms were available for review in most of these cases. Seventy-six cases of RPD were detected on antenatal sonography at our institution over a 3-year period. Eight cases were lost to follow-up in early infancy. The remaining 68 infants had serial sonography performed throughout infancy. Twenty-five infants showed no evidence of urinary tract dilatation on postnatal sonography. Extrarenal pelves were present in eight cases, and 35 infants had moderate to severe RPD at 72 hr. On further evaluation of this last group with voiding cystography or nuclear scintigraphy, diagnoses were VUR (n = 6), pelvic-ureteric junction obstruction (n = 5), renal dysplasia (n = 1), and congenital megaureter (n = 1). VUR was detected in five male infants and one female infant. Twenty-two infants with moderate to severe RPD had neither an obstructive uropathy nor VUR. During this same period (July 1988-June 1991), 20 cases of VUR were detected at our institution. Antenatal sonograms were available for review in 16 of these cases. These showed evidence of RPD in five fetuses, whereas the remaining 11 antenatal studies were unremarkable. Our results show the positive predictive value of RPD for VUR to be 17%. Antenatally detected RPD, in isolation, is a weak predictor of VUR. Postnatal sonographic evaluation is, however, important in this group. Further investigation of infants showing moderate to severe RPD in the neonatal period is merited, as such investigation will lead to early detection of VUR in a significant number of cases (17% of RPD).