Purpose: Routine diagnostic abdominal computed tomography (CT) exams are conducted on millions of patients yearly for purposes other than evaluating kidney health. These scans contain rich, underutilized clinical data where extracted kidney image-derived phenotypes can serve as biomarkers, aiding in early disease detection. This study aims to examine these phenotypes in a large, diverse cohort and their corresponding links to demographic, genetic, and clinical data.

Materials and Methods: Clinical, demographic, and phenotypic traits were extracted from a large academic Penn Medicine biobank. A diverse cohort of 12,288 patient individuals was analyzed. Kidney volume was first extracted via the publicly available TotalSegmentator tool. Subsequent linear regression analyses of CT correlates were then performed among clinical laboratory measurements, such as hemoglobin, creatinine, and body-mass-index in R (version 4.3). A phenome-wide and genome-wide association study (PheWAS, GWAS) was performed using total kidney volume.

Results: Chronic renal failure, end stage renal disease, and hyperpotassemia were among dozens of significant phenotypes identified from the PheWAS using a Bonferroni threshold (p = ~0.00004, Fig. 1). Regressions for creatinine and hemoglobin were statistically significant at a p-value threshold of 0.001 with coefficients of -0.00000199 and 0.000000382 respectively. SNPs on Chromosome 11 linked to NDUFS3, a gene encoding an iron-sulfur protein in Complex I of the mitochondrial electron transport chain, were significant at the threshold of p = 1e-6.

Conclusions: This study provides clinical insights to how kidney volume is linked to laboratory, genetic, and demographic indicators in a diverse biobank. Kidney volume was significantly associated with several medical conditions and lab values.

Clinical Relevance/Application: This research elucidates the relationship between renal volume and its underlying clinical, laboratory, and genetic relationships. Understanding these broad and comprehensive clinical conclusions can provide insights that advance population health.