Tamara van Donge

Chapter 9 182 during the day. There are multiple arguments against this approach. 41 First, as mentioned before, serum creatinine is an unreliable marker for KI in the pediatric population, and therefore by applying this correction possible biomarker trends will be masked. Second, the creatinine excretion is unstable in patients with acute kidney injury; therefore, the changes in biomarker/creatinine ratio cannot be correctly interpreted unless the direct change of each biomarker is known. 20,30 Based on the current knowledge and available data, it seems that serum creatinine and estimated glomerular filtration rate are insensitive markers to detect early kidney injury, particularly in pediatric patients. Even in the field of pharmacometrics, serum creatinine (or glomerular filtration rate) is often identified as an important covariate in pediatric population pharmacokinetic or physiologically-based pharmacokinetic models. 46 Without doubt, kidney function and renal clearance are important factors and can influence the pharmacokinetics of a drug. Moreover, due to high renal perfusion and the tubular reabsorption and secretion, the kidney is at a high risk for drug-related KI. For example, repetitive dosing of aminoglycosides can result in renal cortical accumulation and toxicity. 47 Furthermore, product labels for most nephrotoxic drugs recommend a dose reduction in case of decreased glomerular filtration rate. This early dose reduction in nephrotoxic drugs should avoid irreversible drug-related KI and highlights, once again, the problem of late detection of kidney damage using serum creatinine as a KI biomarker. However, glomerular filtration rate reflects mainly the glomerular filtration, and therefore is not sensitive to detect early drug-related tubular damage. When identifying serum creatinine or estimated glomerular filtration rate as a significant covariate in a pediatric population pharmacokinetic model of nephrotoxic drugs, this indicates that the clearance of the drug of interest is influenced by glomerular function. One can debate the value of implementing an inaccurate surrogate marker for kidney function in such a population pharmacokinetic model. Wilbaux et al. performed an extensive literature research on antibiotic population pharmacokinetic models in neonates. Out of 68 assessed population pharmacokinetic models concerning antibiotics in neonates, 22 identified creatinine (serum or plasma creatinine, creatinine clearance or renal function) as a significant covariate for antibiotic clearance. 46 In addition, critically ill patients and patients with chronic diseases are often suffering from muscle mass loss, nutrition changes and cachexia, resulting in the risk of glomerular filtration rate overestimation and consequently nephrotoxic drug overdosing. This shows that even in case of normal serum creatinine values significant KI, particularly in the tubule, is possible. What actions are needed? Over the last few decades, numerous candidates of KI biomarkers for measuring kidney function have been considered and investigated. 1 A great advantage of these KI biomarkers is the possibility to more precisely localize the specific site of KI (Table 1). Nevertheless,