Poster #82 - Tanisha Johri

Clinically Characterizing Monogenic Variants in Pediatric Kidney Stone Disease

Tanisha Johri1, Rishi Suresh1, Jing Karchin PhD1, Thomas A. Reynolds3, Nikhil Lavu1, Laura Perez1, Gregory Tasian MD, MSc, MSCE 1,2 1Division of Pediatric Urology, The Children's Hospital of Philadelphia, Philadelphia PA 2Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA 3The Richard D. Wood Jr. Center for Fetal Diagnosis & Treatment, Children's Hospital of Philadelphia, Philadelphia, PA

Clinically Characterizing Monogenic Variants in Pediatric Kidney Stone Disease Pediatric kidney stone disease is a growing public health concern, with rising incidence leading to recurrent hospitalizations and long-term renal complications. While most cases are attributed to environmental or metabolic factors, a significant subset is driven by monogenic causes. Genetic testing in pediatric nephrolithiasis often reveals variants of uncertain significance (VUS), limiting their immediate clinical utility. This study integrates genomic and clinical data to improve understanding of these variants and support precision diagnostics. We conducted a cross-sectional analysis of 217 pediatric patients (ages 3 months to 24 years) with nephrolithiasis who underwent clinical genetic testing at the Children's Hospital of Philadelphia between 2020 and 2024. Over half (53%) had at least one genetic variant, with 224 unique variants across 122 genes identified. Notably, 72.6% of the variants were VUS. Pathogenic or likely pathogenic variants were most frequently found in SLC3A1, SLC7A9 (cystinuria), SLC34A4 (renal phosphate wasting), and HOGA1 (primary hyperoxaluria type 3). To support genotype-phenotype correlation, we utilized the Clinical Outcomes Data Archive (CODA), an internal tool that programmatically abstracts structured and unstructured data from electronic health records (EHR). CODA enables high-throughput, reproducible phenotyping at scale by reducing missingness and enhancing data fidelity. Preliminary abstraction includes medication use, demographics, comorbid conditions, and clinical outcomes-laying the groundwork for computational modeling of disease subtypes. This work highlights the feasibility of integrating clinical and genomic data through bioinformatics pipelines to improve variant interpretation. Ongoing efforts will apply CODA-derived phenotypes to refine VUS classification and identify patterns that distinguish monogenic from non-monogenic disease. Ultimately, our findings aim to inform diagnostic workflows, reduce time to diagnosis, and personalize care for children with nephrolithiasis.