#3745 Assay validation and quantification of new urinary kidney safety biomarker candidates for the detection of glomerular injury

Abstract Background and Aims Urine biomarkers for tubular kidney damage have been studied comprehensively and qualified for non-clinical safety assessments. Their use in cohort-level applications during Phase 1 trials with healthy volunteers is well-established, particularly when nephrotoxicity is a concern. Yet, there is a paucity of data on biomarkers for drug-induced glomerular damage, especially for identifying subclinical injury occuring before changes in GFR and/or albuminuria. The Translational Safety Biomarker Pipeline (TransBioLine) project is exploring a range of urinary protein biomarkers to detect drug-induced glomerular damage. Its goal is to address this gap by evaluating candidate glomerular injury biomarkers in at-risk patients. The project focuses on how a panel of biomarkers for both glomerular and tubular damage can provide insights into the mechanisms of drug-induced kidney injury. Here, we present findings from the learning phase studies, which were used to select protein biomarkers to be included in confirmatory phase studies aimed at supporting biomarker qualification. Method We present the results for glomerular protein biomarkers—nephrin (NPHS1), podocin (NPHS2), and podocalyxin (PODXL)—alongside tubular injury biomarkers KIM 1 and NGAL, as well as two vascular component markers, vascular cell adhesion molecule 1 (VCAM1) and matrix metalloproteinase 3 (MMP3). Using IA-LC-MS/MS assays, we quantified these urinary glomerular, tubular, and vascular protein biomarkers in urine samples from patients with confirmed glomerular injury (n = 48) and healthy volunteers (n = 50) as part of an ongoing biomarker qualification initiative. Results Geometric mean ratios (GMRs) calculated for each biomarker in diseased versus healthy cohorts were 3.83 for NGAL (2.17–6.76, P < 0.0001)), 1.43 for PODXL (1.08–1.89, P = 0.0128), 2.9 for MMP3 (1.93–4.38, P < 0.0001), and 6.04 for VCAM1 (4.16–8.78, P < 0.0001), while NPHS1 and NPHS2 showed no significance (p-values = 0.2034 and 0.7515, respectively). In cases with glomerular damage where eGFR was preserved (≥60 ml/min, n = 24), NPHS1, NPHS2, and PODXL displayed GMRs of 1.41, 1.74, and 2.16, respectively, compared to normal healthy volunteers. With respect to overall biomarker performance, VCAM1 and PODXL demonstrated the best discrimination between disease cases and controls with areas under the curve of 0.92 and 0.83, respectively. Conclusion Our findings highlight the potential of biomarkers like PODXL, VCAM1, and MMP3 for detecting glomerular damage in clinical urine samples. NPHS1, NPHS2, and PODXL may offer predictive value, particularly in cases with normal eGFR. To better understand the dynamics of these candidate biomarkers, longitudinal sample analysis is crucial, especially in the context of drug- or disease-related kidney damage. This will help clarify how these biomarkers compare to traditional metrics like eGFR and albuminuria and support their integration as valuable tools in drug development.