Protein Biomarkers May Indicate Early Kidney Impairment, Study Finds

A study published in PLOS One found significant associations between several protein biomarkers and estimated glomerular filtration rate (eGFR), and this overall finding could serve as an indication of early kidney impairment when treatment is most effective.

The study evaluated 71 cardiovascular disease (CVD)-related plasma proteins with chronic kidney disease (CKD) traits for this study, because CVD and CKD have similar risk factors. The discovery sample for the analysis included participants from the FHS Offspring cohort (1998-2001). Participants were excluded if they had a medical record-confirmed diagnosis of heart failure or myocardial infarction at the baseline visit. A total of 2873 participants were included in this study.

The FHS Third Generation cohort was used for longitudinal discovery analyses. All protein biomarkers associated with eGFR in the FHS Offspring cohort were validated with 3951 participants in the FHS Third Generation cohort by using data from the first (2002-2005) and second (2009-2011) exams of this cohort.

The mean (SD) age of the Offspring and Third Generation participants were 60 (9) and 40 (9) years, respectively, and the mean (SD) eGFRs were 84 (16) mL/min/1.73m² and 102 (14) mL/min/1.73m². There were 7% and less than 1% of Offspring and Third Generation participants who had CKD, respectively.

There were 37 protein biomarkers that were significantly associated with eGFR in cross-sectional analysis, with 36 having an inverse association with eGFR. Alpha-1-microglobulin (A1M), adipsin, beta-2-microglobulin (β2M), cystatin C, myoglobin, and resistin were significantly associated with a change in R2 of at least 5%. There were 20 protein biomarkers that were associated with eGFR in both cohorts.

There were 27 protein biomarkers that were significantly associated with the annual change in eGFR (ΔeGFR) in the Offspring cohort, with 25 protein biomarkers that were inversely associated with ΔeGFR. β2M and cystatin C were associated with increases in R2 (annual changes in R2, 3.3% and 2.5%, respectively). There were 12 biomarkers that were significantly associated with ΔeGFR in the Third Generation cohort.

Eight protein biomarkers were consistently associated with continuous kidney function in both cohorts in the cross-sectional and longitudinal analyses. Adipsin, A1M, β2M, cystatin C, epidermal growth factor–containing fibulin-like extracellular matrix protein 1, myoglobin, soluble receptor for advanced glycation endproducts, and tissue inhibitor of metalloproteinases 1 were all associated with eGFR and ΔeGFR.

There were 35 biomarkers that were associated with prevalent CKD, 34 of which were associated with greater odds of CKD. There were also 5 protein biomarkers that were associated with higher odds of new-onset CKD and 17 protein biomarkers that were associated with higher odds of rapid eGFR decline.

The sutdy authors noted some limitations to their findings. They used a single creatinine measurement at baseline and follow-up rather than multiple eGFR values, which means their definition of CKD does not mirror the clinical diagnosis of CKD. In addition, the younger and healthier cohort limited the ability to identify biomarkers of changing kidney function and CKD.

The researchers concluded that the results of the study validate previous research on proteins that may be used to detect impaired kidney function early when treatment is most beneficial. Significant associations were found between 20 protein biomarkers with eGFR and 12 proteins with ΔeGFR, with causal associations of reduced kidney function for 2 proteins.

“Further studies are necessary to determine if any of the proteins identified by Mendelian randomization can serve as useful biomarkers for CKD either individually or in combination with known and validated biomarkers,” the authors concluded.

Reference

Keshawarz A, Hwang SJ, Lee GY, et al. Cardiovascular disease protein biomarkers are associated with kidney function: the Framingham heart study. PLoS One. Published online May 11, 2022. doi:10.1371/journal.pone.0268293