Kidney Week

Abstract: TH-PO0354

Baseline and One-Year Percent Changes in Concentration of Plasma Proteins as Predictors of ESKD in the Chronic Renal Insufficiency Cohort (CRIC) with Type 2 Diabetes

Session Information

• Diabetic Kidney Disease: From Early Biomarkers to Novel Therapeutic Targets
  November 06, 2025 | Location: Exhibit Hall, Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Diabetic Kidney Disease

• 702 Diabetic Kidney Disease: Clinical

Authors

• Kobayashi, Hiroki, Joslin Diabetes Center, Boston, United States

Hiroki Kobayashi, MD, PhD, FASN

• Md Dom, Zaipul, Joslin Diabetes Center, Boston, Massachusetts, United States

Zaipul Md Dom, PhD

• Satake, Eiichiro, Joslin Diabetes Center, Boston, Massachusetts, United States

Eiichiro Satake, MD, PhD

• Tye, Sok Cin, Joslin Diabetes Center, Boston, Massachusetts, United States

Sok Cin Tye, PhD

• Lash, James P., Division of Nephrology, Department of Medicine, College of Medicine, University of Illinois at Chicago, Chicago, Illinois, United States

James P. Lash, MD, FASN

• Fleming, Fergus, Renalytix AI Inc, New York, New York, United States

Fergus Fleming, MSc

• Galecki, Andrzej, Department of Biostatistics, University of Michigan, Ann Arbor, Michigan, United States

Andrzej Galecki

• Krolewski, Andrzej S., Joslin Diabetes Center, Boston, Massachusetts, United States

Andrzej S. Krolewski, MD, PhD

Background

The Joslin Kidney Panel (JKP), comprising 21 circulating proteins linked to inflammation, fibrosis, and tubular injury, predicts end-stage kidney disease (ESKD) risk in diabetes at baseline. The added value of measuring 1-year changes in these proteins for ESKD risk prediction remains unexplored.

Methods

In a case-cohort study within the Chronic Renal Insufficiency Cohort (CRIC), we analyzed 194 patients with type 2 diabetes and impaired renal function. Of these, 65 progressed to ESKD within 5 years from baseline, while 129 remained event-free. Baseline eGFR was lower in progressors (35 vs. 49 mL/min/1.73m²) and declined further after 1 year (27 vs. 47 mL/min/1.73m²). JKP proteins were measured at baseline and 1-year follow-up using OLINK proximity extension assays. Percent change was calculated as [(1-year value – baseline)/baseline] × 100. We computed c-statistics for plasma concentration of each of the JKP protein at baseline, 1-year percent change, and the index that combined both. Logistic regression utilizing LASSO analysis selected the most informative subset of indices.

Results

C-statistics for baseline concentrations of the 21 JKP proteins ranged from 0.71 to 0.83 (the highest for KIM) and ranged from 0.70 to 0.80 (the highest for WFDC2) for percent changes. Because the correlations between baseline and percent change were minimal to modest (Spearman’s ρ = <0.30), we constructed combined indices that were weighted according to logistic regression coefficients. These indices improved c-statistics from 0.70 to 0.90 (the highest for eGFR and WFDC2). LASSO analysis selected five key indices—eGFR, KIM1, WFDC2, CD27, and SYND1—yielding almost perfect multimarker c-statistic of 0.931 (95% CI 0.892–0.970).

Conclusion

Both baseline levels and 1-year percent changes in the JKP proteins independently enhance ESKD risk prediction in type 2 diabetes. A prognostic model combining baseline values and 1-year percent changes for four JKP proteins and eGFR achieved highly accurate discrimination for ESKD risk in type 2 diabetes patients with impaired kidney function. Validation in broader populations is needed (Data provided by NIDDK-CR, a program of the National Institute of Diabetes and Digestive, and Kidney Diseases).

Funding

• NIDDK Support

Digital Object Identifier (DOI)

• doi: 10.1681/ASN.2025fzkxzks8