Plasma proteins predict kidney function trajectories in type 2 diabetes.

Objective: The rate of decline in estimated glomerular filtration rate (eGFR) varies among patients with type 2 diabetes (T2D). We aimed to identify plasma proteins associated with diverse eGFR trajectories in T2D.

Methods: We performed latent class mixed models analysis among patients with T2D and relatively preserved kidney function (baseline eGFR ≥60 mL/min/1.73m2) from SMART2D (n=1285) and DN (n=798) cohorts to identify patterns of eGFR trajectories. Comprehensive proteomic association with eGFR trajectories was assessed using multivariable logistic regression in the SMART2D cohort.

Results: Three distinct eGFR trajectories group: slow decline (92.2%), progressive decline (4.0%), and accelerated decline (3.8%) were identified in SMART2D and validated in DN cohort. Participants in the accelerated decline group exhibited the highest risk of progression to end-stage kidney disease (ESKD) (log-rank test, p<0.0001). Among 1448 proteins analysed in the SMART2D cohort, 19 proteins, including KIM-1 (OR=2.95, 95% CI 2.01-4.32; p=2.95x10-8), MMP7 (OR=16.5, 95% CI 5.54-49.07; p=4.61x10-7), and VSIG4 (OR=7.38, 95%CI 3.22-16.89; p=2.24x10-6), were associated with accelerated decline and 1 protein (THBD) (OR=6.34, 95% CI 2.77-14.52; p=1.26x10-5) was associated with progressive decline, independent of traditional cardio-renal risk factors including baseline kidney function. Adding these proteins to clinical risk factors (age, sex, ethnicity, eGFR, uACR, HbA1c, diabetes duration, systolic blood pressure, triglyceride) improved AUC to 0.77 (delta 0.04,p=0.057) for progressive decline and 0.93 (delta 0.09,p<0.001) for accelerated decline.

Conclusion: Different plasma proteins are associated with progressive and accelerated eGFR decline, independent of traditional cardio-renal risk factors, some of which enhance eGFR trajectory prediction in patients with T2D.