Biomarkers of CKD in Children

The global and United States burden of CKD among children is increasing.¹ The global prevalence of CKD in children <20 years old increased from 14,301,902.99 (uncertainty interval [UI], 12,311,407.92–16,839,916.37) in 1990 to 17,330,656.23 (UI, 14,545,431.31–20,740,504.91) in 2017.¹ This increase outpaced that expected by population growth because the CKD prevalence rate also increased from 623.33/100,000 persons (UI, 536.57–733.94) in 1990 to 667.96/100,000 persons (UI, 560.61–799.38) in 2017.¹ Trends are similar in the United States because prevalence has also increased from 427,925.73 (UI, 360,242.73–513,222.29) to 508,105.50 (UI, 420,435.90–613,844.54), and prevalence rate has increased from 580.62/100,000 persons (UI, 488.78–696.35) to 612.70/100,000 persons (UI, 506.99–740.21).¹ The drivers of this increase in burden of CKD among children are not entirely clear.¹

CKD in children is associated with serious consequences, including delayed growth and development, increased risk of cardiovascular disease, and death. Early recognition of CKD and tools to predict its progression and outcomes are important not only for risk stratification—and potentially, as a strategy to enrich enrollment in clinical trials—but also, they may contribute to a greater understanding of the underlying biologic mechanisms and may help identify potential therapeutic targets to halt CKD progression and reduce the burden of health loss associated with this disease.

In this issue of JASN, Greenberg et al.² conducted a prospective analysis of 651 children with CKD (from the Chronic Kidney Disease in Children multicenter cohort study) to examine the relationship between kidney injury molecule 1 (KIM-1), monocyte chemoattractant protein-1 (MCP-1), TNF receptor-1 (TNFR-1), TNFR-2, soluble urokinase-type plasminogen activator receptor (suPAR), and YKL-40 and the risk of CKD progression (defined as a composite end point of eGFR decline >50% or incident ESKD). The median age was 11 years old (interquartile range, 8–15), and median baseline eGFR was 53 ml/min per 1.73 m² (interquartile range, 40–67). After a median follow-up of 5.7 years, 223 children (34%) progressed to the composite end point (21% with incident ESKD and 13% with >50% decline in eGFR), reflecting the unfortunate sizable toll of CKD progression among children with CKD. Children with plasma KIM-1, TNFR-1, or TNFR-2 concentrations in the highest quartile were at a significantly higher risk of CKD progression compared with those with biomarker concentrations in the lowest quartile. In analyses stratified by type of kidney disease, KIM-1 was associated with the composite outcome in children with nonglomerular causes of CKD, whereas TNFR-1 and TNFR-2 were associated with the composite outcome in children with glomerular causes of CKD. However, interaction analyses suggested that the association between each of these biomarkers and the composite outcome was not modified by type of kidney disease—suggesting that these biomarkers may not reliably serve to differentiate type of CKD. Interestingly, plasma MCP-1, suPAR, and YKL-40 were not independently associated with CKD progression in this study. Analyses of area under the curve, integrated discrimination improvement, and net reclassification improvement suggested that addition of these biomarkers (to routinely measured clinical variables, including eGFR) did not result in substantial improvement in predictive performance.

The study is notable in that it is a large longitudinal study of a well characterized cohort of children with a diverse range of CKD including over 30 primary diagnoses encompassing both glomerular and nonglomerular causes. The authors set out to evaluate potential utility of a battery of biomarkers representing several mechanistic pathways, including tubular injury (KIM-1), inflammation (MCP-1, TNFR-1, TNFR-2, and suPAR), and repair (YKL-40). The results show for the first time that elevated levels of KIM-1, TNFR-1, or TNFR-2 are independently associated with risk of CKD progression in children. The addition of these biomarkers to clinical models that also included eGFR did not improve predictive performance substantially—a reflection of both the strength of eGFR as a functional biomarker in prediction of outcomes and its limitations; eGFR decline occurs after substantial chronic injury, and it does not provide insight into the mechanism and type of underlying kidney disease. The strength of eGFR as a functional biomarker and its many limitations add to the challenges of identifying and validating novel biomarkers of CKD that can significantly improve predictive performance beyond what can be achieved with inclusion of eGFR.³ In this specific study and as the authors indicate, the relatively small number of cohort participants may have underpowered the predictive performance of the examined biomarkers.

The findings in this pediatric cohort are generally consistent with studies in adults, including a report by Coca et al.⁴ in which analyses of the Action to Control Cardiovascular Risk in Diabetes trial and the Veterans Administration Nephropathy in Diabetes study of people with type 2 diabetes mellitus showed that plasma TNFR-1, TNFR-2, and KIM-1 were each independently associated with risk of eGFR decline. In the study by Coca et al.⁴—which is comparatively much larger and performed in adults—addition of the three biomarkers to a prediction model that also included clinical variables improved risk prediction for eGFR decline.

Also notable in this study is the finding that suPAR, plasma MCP-1, and YKL-40 were not independently associated with CKD progression. In particular, the findings on suPAR are not congruent with prior reported results, and they likely emanate from differences in assays and underlying differences in the characteristics of the study populations.

Overall, this study illustrates both the promise and the challenge of biomarkers in children with CKD. Identification and validation of CKD biomarkers have been a challenge in adults, and it is even more so in children, where cohorts are generally smaller and etiologies are vastly more diverse. The meritorious work by Greenberg et al.² illuminates our understanding of relationship of several biomarkers and CKD in children and establishes a solid foundation for future investigations. Continued progress in biomarker discovery and validation holds great promise to improve our ability to define the disease earlier and with greater precision, risk stratify patients, and predict outcomes, and it may also aid in the identification of potential therapeutic targets.

Disclosures

None.