L'Estimation de la filtration glomérulaire en 2014: Intérêts et limites des tests et formules

The accurate estimation of the glomerular filtration rate (GFR) is a goal of multiple interests regarding clinical, research and public health aspects. The strong relationship between progressive loss of renal function and mortality underlines the need for early diagnosis and close follow-up of renal diseases. Creatinine is the commonest biomarker of GFR in use. By reason of non-renal determinants of GFR, it is required to integrate creatinine values within equations that take in account its most important determinants (i.e. age, sex). The CKD-EPI 2009 equation is now recommended as the first line equation to estimate GFR within the general population. In this indication, it should replace MDRD that tends to overestimate the prevalence of stage 3 chronic kidney disease with GFR around 60 ml/min. However, many questions remain about the accuracy of GFR equations in specific situations such as extremes of age or body weight. The identification of new biomarkers, less determined by non-renal determinants, is of importance. Among these biomarkers, cystatin-C is more accurate to estimate GFR when it is combined to creatinine (i.e. equation CKD-EPI 2012). However the indications for using cystatin-C instead of creatinine alone are still unclear and its use remains limited in routine practice. In conclusion, neither biomarker nor equation gives an accurate estimation for the whole range of GFR and for all patient populations. Limits of prediction are relying on both biomarker's properties and the range of GFR that is concerned, but also rely on the measurement methods. Therefore, it is crucial to interpret the estimated GFR according to the strengths and weaknesses of the equation in use.SCOPUS: ar.j