Metabolites related to eGFR: Evaluation of candidate molecules for GFR estimation using untargeted metabolomics

Abstract Background Metabolomics can be used to identify novel metabolites related to renal function and that could therefore be used for estimating GFR. We evaluated metabolites replicated and related to eGFR in 3 studies (CKD and general population). Methods Metabolomics was performed by GC–MS. The Progredir Cohort ( n  = 454, class 3 and 4 CKD) was used as the derivation study and adjusted linear regression models on eGFR-CKDEPI were built. Bonferroni correction was applied for selecting metabolites to be independently validated in the Diabetic Nephropathy Study ( n  = 56, macroalbuminuric DN) and in the Baependi Heart Study (BHS, n  = 1145, general population). Results In the Progredir Cohort, 72 metabolites where associated with eGFR. Of those, 11 were also significantly associated to eGFR in the DN Study and 8 in the BHS. Four metabolites were replicated and significantly associated to eGFR in all 3 studies: d -threitol, myo-inositol, 4-deoxierythronic acid and galacturonic acid. In addition, pseudouridine was strongly correlated to eGFR only in the 2 CKD populations. Conclusions Our results demonstrate metabolites that are potential biomarkers of renal function: d -threitol, myo-inositol, 4-deoxierythronic acid, galacturonic acid and pseudouridine. Further investigation is needed to determine their performance against otherwise gold-standard methods, most notably among those with normal eGFR.