Uroepithelial thickening improves detection of vesicoureteral reflux in infants with prenatal hydronephrosis.

INTRODUCTION

Postnatal evaluation of prenatal hydronephrosis (PNH) often includes a voiding cystourethrogram (VCUG) for VUR assessment. Despite limited supporting data, VCUG is currently recommended if postnatal renal and bladder ultrasound (RBUS) reveals moderate/severe hydronephrosis (HN) or hydroureter (HU). Recent studies have shown VUR is more accurately diagnosed by using certain sonographic findings as criteria for obtaining VCUG. Uroepithelial thickening (UET) of the renal pelvis is a finding associated with high-grade vesicoureteral reflux (HGVUR); however, the clinical significance of UET with PNH has not been studied.

OBJECTIVE

We sought to determine if the presence of UET implies increased risk for VUR, and to investigate whether UET can improve the test characteristics of RBUS for VUR.

STUDY

DESIGN

We retrospectively analyzed postnatal RBUS and VCUG findings in infants ≤30 days undergoing evaluation for "prenatal hydronephrosis" over an 11-year period. We used logistic regression to identify factors associated with VUR. Test characteristics of RBUS for HGVUR were compared based on the presence of UET and two criteria sets to define abnormal RBUS. Criteria set 1 consisted of HN SFU grade 3-4 and/or HU; criteria set 2 was defined by the presence of two of following: UET, HU, duplication, and/or renal dysmorphia.

RESULTS

Of 135 patients, 39 (29%) had VUR, of whom 16 (41%) had HGVUR. UET was significantly associated with VUR (p < 0.001), and the sensitivity for HGVUR based on UET alone was 94%. On multivariable analysis, UET, HU, duplication, and renal dysmorphia remained significant independent predictors of HGVUR. Compared to criteria 1, using criteria 2 resulted in 43 fewer VCUGs, and significant improvement in sensitivity and specificity for HGVUR (Table).

DISCUSSION: Consistent with previous studies, HN alone on postnatal RBUS has little value in predicting the presence or severity of VUR. This study is the largest known series to evaluate UET in the setting of PNH, and our results demonstrate that UET, as well as HU, duplication and renal dysmorphia, are independent sonographic findings predicting HGVUR. Using our proposed criteria, the probability of HGVUR is fourfold more than the prevalence described in the literature, and importantly, when compared to the criteria recommended by the SFU and AUA, would have resulted in 53% fewer VCUGs while missing zero cases of HGVUR.

CONCLUSION

In infants with PNH, the sonographic findings of UET, HU, duplication and renal dysmorphia independently indicate greater risk of HGVUR, and the sensitivity and specificity of RBUS for HGVUR is markedly improved when at least two of the four are present.