Tamara van Donge

Kidney injury biomarkers in pediatrics 171 9 are exposed to nephrotoxic drugs during the time their glomerulogenesis is still in full swing (until 34-35 weeks of gestation), and therefore it is much more challenging to obtain and interpret reference values in this specific patient population. The aims of this literature review are to (i) give an overview about the current research stage concerning KI biomarkers in pediatrics, (ii) provide a synopsis of reference values for KI biomarkers in the pediatric population, (iii) identify possible knowledge gaps, and (iv) examine which biomarkers are subjected to increased interest and might be considered as an appropriate candidate for optimal kidney monitoring in neonates, infants and children. Methods A literature search on reference values of KI biomarkers was performed, followed by an exploratory analysis of the retrieved data. Literature search This review is based on a systematic PubMed search ( “Reference Values” [Mesh] AND “Kidney” [Mesh] AND “Biomarkers” [Mesh]) conducted on the 1 st of April 2019. Using specific Mesh terms for distinct biomarkers (e.g. NGAL and kidney injury molecule-1, etc.) resulted in limited results and lack of relevant papers. Therefore, references of significant papers were examined, and an additional supplementary search was performed to ascertain that no duplicated studies were included and no relevant studies were missed. The following exclusion criteria were applied; papers containing adult or animal data, review articles, studies without specific biomarker values, studies without pro- or retrospective study design, ill patient population, therapy with nephrotoxic drugs and biomarker values expressed over creatinine values. In order to perform an extensive analysis, studies which compare ill pediatric patients with healthy pediatric controls were included as well, where only the healthy control pediatric population was considered. Exploratory data analysis Summary data (median, mean, confidence interval, interquartile range, etc.), study design, sample size, study population and the specific biomarker values were retrieved. In case solely figures were available, PlotDigitalizer was used to retrieve median or min-max values. Additionally, median values are converted to means in order to perform a weighted mean comparison. 35 An exploratory data analysis has been performed to combine the results of multiple studies in order to integrate the retrieved data while accounting for sample size, based on different age groups.