Moderator's view: Estimating glomerular filtration rate—the past, present and future

Nephrol Dial Transplant (2013) 28: 1404–1406 doi: 10.1093/ndt/gfs607 Moderator’s view: Estimating glomerular filtration rate— the past, present and future 1 Stein I. Hallan1,2 Department of Medicine, Division of Nephrology, St. Olav University Hospital, Trondheim, Norway, and Ron T. Gansevoort3 2 Department of Cancer Research and Molecular Medicine, Faculty of Medicine, Norwegian University of Science and Technology, Trondheim, Norway and 3 Department of Nephrology, University of Groningen, University P O L A R V I E W S I N N E P H R O LO G Y Modern renal physiology arose during the scientific revolution of the 1840s in Europe with mechanistic experiments playing a central role. Carl Ludwig, a German physiologist, was the first to describe the principle of glomerular ultrafiltration driven by physical forces alone [1]. However, an understanding of the fundamental mechanisms of kidney function awaited the advent of micropuncture techniques by Wearn and Richards in 1924 [2]. Since then, numerous research groups have provided us with knowledge on glomerular haemodynamics, filtration barrier function and the processes of tubular reabsorption and secretion. We also know that tubular and interstitial pathological processes can affect glomerular function. Therefore, the glomerular filtration rate (GFR) is now regarded as the best overall measure of kidney function, reflecting reduced clearance of waste products and also to some extent disturbed regulation of electrolytes, water balance and hormone production. Inulin clearance was established early as the gold standard method for measuring the GFR. This substance completely fulfils the requirements of no extra-renal clearance, free filtration and no tubular reabsorption or secretion [3]. However, the method is complicated from both an analytical and a practical point of view with the need for continuous intravenous infusion. It has, therefore, seldom been used in clinical practice and nor is it widely used in research institutions today. Plasma clearance of exogenous markers like iothalamate, iohexol and EDTA is available instead for measuring the GFR in most hospitals and are considered equally accurate. Although considered as reference methods, their accuracy is not very high with coefficients of variation reported in the range of 5–20% [4]. Although less complicated than the classical inulin clearance technique, these methods are also very expensive and take hours per patient. © The Author 2013. Published by Oxford University Press on behalf of ERA-EDTA. Serum creatinine and various formulas based on this cheap, fast and universally available analyte, therefore, remains the cornerstone of kidney function assessment. The relationship between serum creatinine and GFR is non-linear and influenced by several non-kidney variables, and numerous formulas for estimating the GFR have been published over the past 40 years. The Cockcroft Gault formula gained worldwide popularity despite being based on only 249 hospitalized males with kidney function measured as 48-h urine creatinine clearance, which includes tubular creatinine secretion [5]. Later studies have shown that the formula gives a suboptimal performance as a proxy for true GFR in several settings including in the elderly and in advanced kidney disease, where tubular creatinine secretion can lead to 35% overestimation of GFR [6–8]. In 1999, Levey et al. [9] published a new formula based on 1628 chronic kidney disease (CKD) patients from the Modification of Diet in Renal Disease (MDRD) study with the GFR measured as urinary iothalamate. These patients all had a GFR of <60 mL/min/1.73 m2. The formula was found to be superior to the Cockcroft Gault formula and has since gained worldwide popularity and has undergone several modifications, among which to account for differences in the measurement technique for creatinine. However, accuracy was still only moderate, especially in the GFR range >60 mL/min/ 1.73 m2. Ten years later, Levey et al., therefore, published a new serum creatinine-based formula [the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula] using the highest available standards: serum creatinine traceable to isotope-dilution mass spectrometry standards from an impressive 8254 subjects, iothalamate clearance for measuring true GFR, separate study groups for developing and validating the formula, a sufficiently high number of black participants and participants with a wide range of kidney functions [10]. 1404 Medical Center Groningen, Groningen, the Netherlands Correspondence and offprint requests to: Stein Hallan; E-mail: 1405 Estimating glomerular filtration rate P O L A R V I E W S I N N E P H R O LO G Y impede correct medication dosing and all kinds of preventive medicines for kidney disease. Most doctors and health authorities now agree that there should be an increased focus on early diagnosis of CKD and interventions trying to slow down the progression towards end-stage renal disease (ESRD). Late referral could lead us back to those times where nephrologists were only focusing on lifesaving renal replacement therapy without time for institution of renoprotective therapies in an early disease phase. Hence, the nephrology community will heavily depend on creatinine-based formulas in the foreseeable future, especially since the estimated glomerular filtration rate (eGFR) has also become a strong predictor of renal endpoints and cardiovascular and all-cause mortality as well. Low kidney function contributes directly to harmful pathophysiological mechanisms and is also a marker of prevalent subclinical general vascular disease. Clearly, the CKD-EPI formula does not represent any revolution in GFR estimation. As stated by both sides in the current debate, it is more of an evolution. The current knowledgebase strongly indicates that the CKD-EPI equation is superior to the MDRD formula [13, 14]. The difference is not very big, but the nephrology community should take advantage of all possible steps going forward in this field: bias is reduced by 50% and accuracy is improved by 18%. The eGFR distribution in the general population will shift somewhat upwards and become more plausible, e.g. one-third of all CKD Stage 3a patients will be shifted upwards to the 60–89 mL/ min/1.73 m2 range [13]. These reclassified patients have a substantially lower mortality and ESRD risk compared with those who remain in the eGFR stage. We also know that the costs of implementing the CKD-EPI formula in clinical practice are minimal, and other more accurate new creatinine-based formulas are unlikely to be published in the coming years. There are, therefore, many strong arguments supporting the view that European and North American laboratories should change to the CKD-EPI formula now. A more difficult question is what we should do in the future. Creatinine-based equations do not seem to have much potential for further improvement, so some kind of non-creatinine kidney (...truncated)