Albuminuria is widely used to monitor chronic kidney disease (CKD) and to assess treatment response, but it has recognized limitations: it is variable and predominantly reflects glomerular injury. The authors evaluated two urinary protein markers that target tubular biology: urinary clusterin (uCLU, reported as uCLU-to-creatinine ratio, uCLU/Cr), associated with tubular injury, and urinary epidermal growth factor (uEGF, uEGF-to-creatinine ratio, uEGF/Cr), associated with tubular repair. These markers were examined as mechanism-informed pharmacodynamic biomarkers that might complement albuminuria by capturing distinct intrarenal pathways engaged by specific therapies.
The analysis used two randomized CKD trial populations. Effects of an endothelin receptor antagonist (ERA, atrasentan) on uCLU/Cr and endothelin-1 (ET-1) were assessed in the atrasentan enrichment phase of the SONAR trial (Study of Diabetic Nephropathy with Atrasentan). Effects of an SGLT2 inhibitor (SGLT2i, dapagliflozin) on uEGF/Cr were examined in the DAPA-CKD trial across different CKD etiologies, diabetes status, and levels of glycemic control. Associations between urinary uEGF/Cr and intrarenal EGF messenger RNA (mRNA) expression were investigated using kidney biopsy single-cell RNA sequencing data.
After 6 weeks of treatment with atrasentan in SONAR, uCLU/Cr decreased by 46.3% (95% confidence interval [CI]: -57.8 to -37.1). Concurrently, serum endothelin-1 (ET-1) increased by 23.4% (95% CI 19.2–27.4). The report notes no observed effect of atrasentan on serum clusterin and only a numerical (non-specified) decrease in urinary ET-1. Baseline uCLU/Cr correlated significantly with urinary ET-1 (Pearson r = 0.65, P < .0001), supporting a link between tubular injury marker levels and the endothelin pathway.
In the DAPA-CKD population, dapagliflozin attenuated the decline in uEGF/Cr over 1 year, indicating preservation of a marker linked to tubular repair. The effect displayed heterogeneity across CKD etiologies: relative increases in uEGF/Cr were greatest in diabetic kidney disease and absent in glomerulonephritis and hypertensive nephropathy. Treatment effects were also larger in participants with type 2 diabetes and in those with higher baseline HbA1c. These findings suggest that the effect of SGLT2 inhibition on tubular repair biomarkers varies by underlying disease mechanism and metabolic context.
Using kidney biopsy single-cell RNA sequencing data, tubular EGF mRNA expression correlated positively with uEGF/Cr. This supports a mechanistic connection between urinary uEGF measurements and intrarenal EGF gene expression, strengthening the biological plausibility of uEGF as a marker of tubular repair processes.
The results indicate that urinary clusterin and uEGF act as mechanism-informed pharmacodynamic biomarkers reflecting distinct intrarenal pathways engaged by specific therapies: uCLU/Cr for tubular injury linked to endothelin receptor antagonism and uEGF/Cr for tubular repair linked to SGLT2 inhibition. These biomarkers complement albuminuria by assessing tubular pathobiology in addition to glomerular injury, potentially providing a more comprehensive biological assessment of treatment effects and explaining heterogeneity in responses across CKD etiologies.
By capturing tubular injury and repair, uCLU/Cr and uEGF/Cr could inform risk stratification and guide personalized therapeutic strategies in CKD. Their differential responses to ERA and SGLT2i suggest potential utility as pharmacodynamic markers in clinical trials and eventual clinical monitoring to detect mechanism-specific effects not apparent from albuminuria alone. Heterogeneous effects across diseases and metabolic states highlight the need to consider CKD etiology and diabetes control when interpreting biomarker changes.
The abstract reports primary biomarker changes, correlations, and heterogeneity by CKD etiology and diabetes status. It does not provide complete methodological details in this summary, such as sample sizes, assay methods, adjustment covariates, statistical models beyond reported percentages and confidence intervals, or long-term clinical outcome associations for these biomarker changes. Further details and full data would be found in the full article and supporting materials referenced by the trial publications.