Kidney Week

Abstract: FR-PO702

Predictive Value of Initial Diagnostic Imaging for Congenital Ureteropelvic Junction Obstruction Requiring Pyeloplasty

Session Information

• Pediatric Nephrology - 1
  October 25, 2024 | Location: Exhibit Hall, Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Pediatric Nephrology

• 1900 Pediatric Nephrology

Authors

• Solarin, Adaobi, Kidney and Urinary Tract Center, Abigail Wexner Research Institute at Nationwide Children's Hospital, Columbus, Ohio, United States

Adaobi Solarin, MBBS

• Joshi, Shivam, The Ohio State University Wexner Medical Center, Columbus, Ohio, United States

Shivam Joshi, MS, MBA

• Gupta, Sudipti, Kidney and Urinary Tract Center, Abigail Wexner Research Institute at Nationwide Children's Hospital, Columbus, Ohio, United States

Sudipti Gupta, PhD

• Ching, Christina B., Kidney and Urinary Tract Center, Abigail Wexner Research Institute at Nationwide Children's Hospital, Columbus, Ohio, United States

Christina B. Ching, MD

• Becknell, Brian, Kidney and Urinary Tract Center, Abigail Wexner Research Institute at Nationwide Children's Hospital, Columbus, Ohio, United States

Brian Becknell, MD, PhD

Background

Hydronephrosis is one of the most common findings on prenatal ultrasound. While prenatal hydronephrosis resolves in most instances, 6% of cases represent ureteropelvic junction obstruction (UPJO) that will require pyeloplasty, and differentiating between obstructive and non-obstructive hydronephrosis remains an unmet challenge. Usually, a combination of factors such as the Society of Fetal Urology (SFU) grade, drainage pattern, and differential renal function (DRF) on diuretic renogram are used in assessing the severity of hydronephrosis. We sought to identify predictive features on initial postnatal imaging for pyeloplasty among children with suspected UPJO.

Methods

We conducted a prospective, single-center study of children with prenatal hydronephrosis, postnatal imaging suggestive of isolated UPJO without other congenital anomalies, and two years of follow-up. Demographic, clinical and radiologic data from initial postnatal imaging were retrieved. Chi-squared test or Fisher's exact test was used to examine associations between categorical variables and surgical intervention, and Wilcoxon rank sum test identified the differences in the distribution of continuous variables across surgical interventions. A logistic regression model was utilized to identify potential predictors of surgical intervention.

Results

There were 98 cases of UPJO, of whom 49 (50%) underwent pyeloplasty. The median age at initial imaging was 4.4 months, 77 (79%) were male and 81 (84%) Caucasian. The initial SFU grade and gravity assisted drainage (GAD) on radionuclide imaging were associated with pyeloplasty. Logistic regression analysis indicated that age (p=0.0090), GAD (p=0.0009), and DRF (p=0.0422) are significant predictors of pyeloplasty. The model explained 45% of the variance in the outcome and correctly classified 83% (AUC=0.83) of cases undergoing pyeloplasty.

Conclusion

The initial diagnostic imaging has shown an excellent performance in predicting surgical intervention with age, DRF and GAD as significant predictors. Ongoing studies seek to extend these observations in an independent patient cohort and superimpose candidate urinary biomarkers to further discriminate UPJO from non-obstructive hydronephrosis at the time of initial postnatal imaging.

Funding

• NIDDK Support