Abstract: TH-PO0350
Predictive Value of Established and Novel Biomarkers for Kidney Outcomes in CKD: A Comparative Validation Study
Session Information
- Top Trainee Posters - 1
November 06, 2025 | Location: Exhibit Hall, Convention Center
Abstract Time: 01:30 PM - 01:36 PM
Category: Diabetic Kidney Disease
- 702 Diabetic Kidney Disease: Clinical
Authors
- Moedt, Erik, Universitair Medisch Centrum Groningen Afdeling Klinische Farmacie en Farmacologie, Groningen, GR, Netherlands
- Jongs, Niels, Universitair Medisch Centrum Groningen Afdeling Klinische Farmacie en Farmacologie, Groningen, GR, Netherlands
- Januzzi, James, Massachusetts General Hospital Corrigan Minehan Heart Center, Boston, Massachusetts, United States
- Jardine, Meg, The George Institute for Global Health, Sydney, New South Wales, Australia
- Neuen, Brendon Lange, The George Institute for Global Health, Sydney, New South Wales, Australia
- Pollock, Carol A., Kolling Institute of Medical Research, St Leonards, New South Wales, Australia
- Hansen, Michael K., Janssen Research & Development LLC, Spring House, Pennsylvania, United States
- Bakker, Stephan J.L., Universitair Medisch Centrum Groningen Afdeling Interne Geneeskunde, Groningen, GR, Netherlands
- Heerspink, Hiddo Jan L., Universitair Medisch Centrum Groningen Afdeling Klinische Farmacie en Farmacologie, Groningen, GR, Netherlands
Background
CKD is a heterogeneous disease with involvement of multiple pathophysiological mechanisms and processes. Combining biomarkers reflecting distinct pathophysiological processes may enhance prognostic accuracy and enable a more comprehensive risk assessment than single biomarkers.
Methods
We measured kidney injury molecule-1 (KIM-1), tumor necrosis factor receptor (TNFR)-1, TNFR-2, interleukin-6 (IL-6), and neuroblastoma suppressor of tumorigenicity 1 (NBL1) in plasma, and albumin and epidermal growth factor (EGF) in urine, on biobanked samples from CREDENCE participants. Longitudinal kidney outcome was defined as a composite of ≥40% decline in eGFR, kidney failure, and renal death. Associations were evaluated using multivariate Cox regression, with HRs expressed per SD increase. A risk prediction model was developed using bootstrapping and elastic net regression, and internally validated through 5-fold cross-validation. The performance was assessed using C-statistics.
Results
We included 2055 participants (mean 57.0 eGFR mL/min/1.73m2; median UACR 883.0 mg/g) with available samples. Baseline UACR, KIM-1, TNFR1, TNFR2, EGF and NBL1 were each independently associated with kidney outcome. A combined biomarker-based risk score demonstrated a stronger association with kidney outcome than any of the individual markers, achieving a C-statistic of 0.81, whereas individual risk markers had C-statistics of 0.77 or lower (Fig. 1). The risk score demonstrated consistent performance upon cross-validation, with C-statistics ranging from 0.76 to 0.84. Compared to placebo, canagliflozin significantly increased EGF and reduced UACR, KIM-1, TNFR1, TNFR2, and the risk score (Fig. 1). Changes in UACR, KIM-1, TNFR2, EGF, and NBL1 from baseline to year 1 were also independently associated with kidney outcome, with the change in risk score showing the strongest association (Fig. 1).
Conclusion
A combined biomarker approach improved risk stratification over individual markers, underscoring the value of integrating markers from distinct pathophysiological pathways to better capture risk progression of CKD.