Abstract: FR-PO0332
Comparison of KidneyintelX.dkd and Kidney Failure Risk Equation for Predicting Progression of Diabetic Kidney Disease
Session Information
- Diabetic Kidney Disease: Progression, Predictive Tools, Therapeutics, and Outcomes
November 07, 2025 | Location: Exhibit Hall, Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Diabetic Kidney Disease
- 702 Diabetic Kidney Disease: Clinical
Authors
- Coca, Steven G., Icahn School of Medicine at Mount Sinai, New York, New York, United States
- Nadkarni, Girish N., Icahn School of Medicine at Mount Sinai, New York, New York, United States
- Fleming, Fergus, Renalytix plc, London, England, United Kingdom
Background
Accurate and reliable prediction of progressive decline in kidney function among patients with diabetic kidney disease (DKD) is critical to enable timely interventions and referrals. The Kidney Failure Risk Equation (KFRE) is a validated risk equation for predicting end stage kidney disease (ESKD). The kidneyIntelX.dkd biomarker-based test has received FDA approval for predicting risk of progression as defined by a ≥40% decline in eGFR or kidney failure in patients with stages G1-G3b DKD. We sought to compare the predictive performance of KFRE vs. kidneyIntelX.dkd for the two kidney endpoints.
Methods
This analysis included 657 patients with early-stage DKD (G1-G3b) from the clinical validation study for kidneyintelX.dkd and were scored using the FDA-approved algorithm. KFRE was calculated using the standard input variables (age, sex, eGFR, and uACR). Predictive performance was assessed using C statistics for both endpoints: sustained ≥40% eGFR decline and kidney failure (eGFR <15). Coefficients of variation (CV%) were also assessed to determine robustness against variability in input features.
Results
Mean eGFR was 64 ml/min/1.73 m2 and median uACR was 52 mg/g. There were 77 (12%) events for 40% eGFR decline and 15 (2.2%) ESKD events over a median follow-up of 3.8 years. For prediction of kidney failure, KidneyIntelX achieved an C statistic of 0.95 compared to 0.90 for KFRE. For ≥40% decline in eGFR, KidneyIntelX achieved an C statistic of 0.78 versus 0.69 for KFRE. In contrast, KFRE performance showed greater sensitivity to known within preson variability in uACR and eGFR inputs, with a CV% of 30%, while KidneyIntelX demonstrated much greater stability in response to expected variation of the input features (overall CV% of 3%).
Conclusion
KidneyIntelX outperformed KFRE in predicting both kidney outcomes in patients with DKD, with higher discrimination and greater robustness to input variability. These findings support the use of KidneyIntelX as a more accurate and reliable tool for risk stratification in clinical practice, particularly in early stage disease where a 40% decline event is more actionable, enabling better-informed decision-making for patients at risk of progressive DKD.
Funding
- Commercial Support – Renalytix