Outcome predictive values of the Society of Fetal Urology (SFU) Grading System and Urinary Tract Dilation (UTD) Classification in patients with high-grade ureteropelvic junction obstruction-like prenatal hydronephrosis

ABSTRACT

Objectives@#This study aimed to determine the surgical predictive value of both SFU and UTD classifications in a specific subset of patients presenting with high-grade (SFU 3 and 4; UTD P2 and P3) UPJO-like hydronephrosis on prenatal ultrasound. Furthermore, this study also aimed to determine the likelihood of spontaneous resolution of high grade UPJO-like hydronphrosis based on both grading systems. @*Methods@#Patients who presented with high-grade hydronephrosis on prenatal ultrasound based on the SFU grading system (Grades 3 and 4) were included in this study. The prenatal renal ultrasounds of these patients were reclassified by a single interpreter according to the UTD classification. Logistic regression was used to test the predictive value of SFU and UTD; ROC curves were plotted accordingly. Kaplan-Meier curves were used to model time to operation and mean time to operation was computed with a 95% confidence interval. Breslow Test was used to determine significant differences in survival curves across the different SFU grades and UTD classifications.@*Results@#Of the 163 patients in the database who presented with prenatal hydronephrosis, 25 patients presented with high-grade UPJO-like hydronephrosis (50 renal units). Logistic regression revealed that the SFU grading system was able to explain only 18.7% of the variance of the occurrence of pyeloplasty, thus, was a poor predictor of the occurrence of surgery. In contrast, logistic regression of the UTD classification was able to explain 47.3% of the occurrence of pyeloplasy with an accuracy of 86% making it a good predictor of surgical intervention. Both SFU and UTD classifications were poor predictors of spontaneous resolution. Mean time to pyeloplasty from the time of diagnosis was 2.98 years (95% CI 2.45-3.53) Kaplan-Meier curve analysis for the time of pyeloplasty for the SFU grading system revealed no significant difference in the time to operation among the different SFU grades (p=0.110) while for the UTD classification, there was a significant difference in time to pyeloplasty across the different UTD classes with the higher classes correlating to a shorter time to pyeloplasty. (p<0.05) @*Conclusion@#The UTD classification system is a good predictor of surgery in patients presenting with high-grade UPJO-like hydronephrosis on prenatal ultrasound with a predictive accuracy of 86%. High-grade hydronephrosis based on the UTD classification equates to a shorter time to surgical intervention from the time of diagnosis necessitating closer follow-up of these patients.