The Promise of Metabolomics in Decelerating CKD Progression in Children

Metabolomics is defined as the quantification of small molecules (usually <1500 Daltons) within a biologic specimen (e.g., plasma) (1). Detectable metabolites belong to various chemical classes like amino acids or lipids. Their concentrations are the result of the interaction between genes, the microbiome, and the environment (1). Metabolomic studies provide information on physiologic variation and pathophysiologic alterations (2).

The kidneys are an excretory organ with a major role in the handling of small molecules. The kidney-metabolome relationship is complex and highly dependent on the kidneys’ health status (1). Abnormal metabolomic profiles can be indicative of the pathogenesis and prognosis of CKD. Metabolomics has therefore been recognized as a valuable tool in the field of nephrology, particularly in the search of new biomarkers for the (early) diagnosis and prognosis of CKD (2).

First applications of metabolomics in nephrology examined plasma or dialysate from individuals with kidney failure because of a long-standing interest in the potential of metabolites as uremic toxins (3). Moreover, population-based studies such as the Atherosclerosis Risk in Communities (ARIC) Study were assessed to identify metabolites associated with estimated glomerular filtration rate (eGFR) (3). Although such results may help to detect new kidney function markers that may, for example, allow early detection of CKD, research interest has now extended to the understanding of the underlying pathogenesis of CKD progression.

The importance of this research field is underscored by the variable progression patterns of CKD patients, ranging from stable trajectories to rapid progression (2). A better understanding of the pathophysiologic processes promoting CKD progression may help to identify patients at higher risk for CKD progression and enable individualized therapeutic approaches. Progress in this field depends on the availability of large prospective cohorts of CKD patients (4).

There are several commonalities between pediatric and adult CKD populations, e.g., variability in progression patterns as well as higher comorbidity and mortality rates compared with a healthy population. But there are also distinct differences, such as comorbidity patterns and underlying causes of CKD, with children developing CKD mostly due to congenital anomalies of the kidney and urinary tract (5). Large pediatric CKD studies are therefore warranted, but scarce.

The prospective Chronic Kidney Disease in Children (CKiD) Study constitutes such a remarkable enterprise (6,7). By including children aged 1–16 years with mildly to moderately reduced kidney function (eGFR, 30–90 mL/min per 1.73 m²), the study group aims to tackle several research questions including the evaluation of risk factors for CKD progression. While eventually >1000 children were enrolled and followed over time, the power of this study is still more limited than corresponding studies among adults, such as the Germany Chronic Kidney Disease (GCKD) Study, the Chronic Renal Insufficiency Cohort (CRIC) Study, or the French Chronic Kidney Disease-Renal Epidemiology and Information Network (CKD-REIN), which all had ≥3000 participants (4,7). Nevertheless, CKiD is still one of the first large studies of a pediatric CKD population with available metabolite measurements.

In this issue of CJASN, Denburg et al. report on their recent findings based on 645 CKiD participants for whom metabolite measurements and follow-up information were available (7). The aim of this investigation was to identify novel biomarkers of CKD progression in children. CKD progression as the primary end point of interest was defined as initiation of kidney replacement therapy or a 50% reduction in eGFR. After a median follow-up time of approximately 5 years, a remarkable proportion of the study population (209 of 645; 32%) experienced this end point.

Denburg et al. measured metabolite levels in plasma of 825 metabolites using the platform by Metabolon, Inc. (Durham, NC), a company with much experience in profiling large cohorts (e.g., the GCKD Study) (7,8). Their technique is based on an untargeted protocol using ultra-high performance liquid chromatography tandem mass spectrometry. The advantage of this protocol is that—without prior restriction to a priori selected metabolites—many, even yet unknown metabolites can be measured in a reproducible manner, thereby providing a more comprehensive snapshot of the metabolome (1). The disadvantage of this technique is that no absolute concentration of these metabolites can be provided (i.e., without unit) and measurement comparability across studies and tissues is limited.

In light of two limitations, namely limited sample size and thus power as well as lack of external validation cohorts, Denburg et al. conducted a sophisticated statistical analysis using a combination of standard and advanced approaches (7). Because of the limited study size, a false discovery rate method in addition to the usual, more conservative Bonferroni correction was applied to define statistical significance in this study, thereby enabling the discovery of more associations (9). Furthermore, several sensitivity analyses including the assessment of robustness and variability of significant findings using resampling were conducted to ensure internal validity of study results (10).

As a result, Denburg et al. report on seven metabolites associated with CKD progression in children with an eGFR≥60 ml/min per 1.73 m², and one metabolite (tetrahydrocortisol sulfate) in patients with an eGFR<60 ml/min per 1.73 m² (7). Higher levels of all metabolites other than tetrahydrocortisol sulfate were associated with higher risk for CKD progression. Among them, the two amino acids C-glycosyltryptophan (also known as C-mannosyltryptophan; tryptophan metabolism) and pseudouridine (pyrimidine metabolism) have already been repeatedly reported by different cross-sectional and prospective studies in adults (1,2). In a recent study, C-glycosyltryptophan was also found significantly associated with kidney failure and mortality when levels were measured from urine (8).

Other metabolites reported by Denburg et al. might be progression markers specific for children (7). However, the absence of reports from adult studies might also be due to the fact that those metabolites were not assessed previously as a consequence of, for example, the ongoing curation and improvement of metabolomics platforms such as the one by Metabolon. In any case, follow-up studies of these metabolites are warranted, especially concerning tetrahydrocortisol sulfate with its observed protective effect on CKD progression.

Finally, the magnitude of the reported effect sizes of significantly associated metabolites is noteworthy. Denburg et al. reported hazard ratios of, for example, pseudouridine from various analyses ranging from 10 to 62 (main analysis: 39) (7). In contrast, a hazard ratio of 2.32 for pseudouridine (similarly measured and transformed) was reported from an analysis of adult CKD patients and the outcome of kidney failure, when adjusted for baseline eGFR (11). Although the effect may indeed be larger in children than in adults, the interpretation of these reported effect sizes generally requires caution. Because of the nature of measurements being semi-quantitative and of limited comparability, the effect estimates are difficult to interpret. To further assess their associations with CKD progression, absolute quantification of metabolite levels would be required.

Overall, the findings of this study are promising but not of immediate relevance for daily patient care. Several further steps would be necessary for translation: first, absolute concentrations of metabolites are required to ensure reproducible measurements that enable comparability with other studies, also over time. Second, the external validation of findings is important. In the context of prognostic research, the evaluation of the predictive ability of these measurements is required. Association analyses as presented here do not provide data on prediction error or predictive value, especially in comparison with other known prognostic factors (e.g., eGFR) (7). Another line of research might necessitate the verification of the causal relationship between certain metabolite concentrations and CKD progression in order to, for example, identify novel therapeutic targets. Finally, in order to implement findings into routine patient care, an assay that allows quantification of candidate metabolites in an easy way, e.g., through standardized bench-top nuclear magnetic resonance spectroscopy set within routine clinical labs, and at low costs is another point to be tackled in the future.

In conclusion, Denburg et al. identified plasma metabolites associated with CKD progression in children (7). This is an important first step on the long road to translate research findings toward patient benefit.

Disclosures

All authors have nothing to disclose.

Funding

The work of UTS was supported by German Federal Ministry of Education and Research (BMBF) grant 01ZX1912B within the framework of the e:Med research and funding concept.