Should we sweat the small (micro) things?

Thirty years ago, a physician whose patient had diabetes and was urine protein dipstick–positive could only say, “You will most likely progress to end-stage renal disease in 10 years and need dialysis or a kidney transplant, and there is not much we can do about it.” Following the landmark microalbuminuria study of Mogensen and Christensen in 1984 (1 ), clinicians could say to patients with diabetes and microalbuminuria, “The good news is, we can now tell you 20 years ahead of time that you will likely progress to end-stage renal disease, but the bad news is that we still cannot do anything about it.” However, by the early 1990s, it was shown that treatment of microalbuminuric diabetic patients with ACE inhibitors to control blood pressure, diet modification, and intensive glucose control could slow or prevent progression to end-stage renal disease (ESRD) (2 ). Microalbumin quickly became a textbook example of a new laboratory test with profound impact on treatment and long-term outcomes of patients with a chronic disease. Since then, the term microalbumin has been abandoned in favor of urine albumin because there is nothing “small” or “low molecular weight” about the albumin measured in urine. Urine albumin assays are simply immunoassays capable of measuring albumin concentrations that are more than 3 orders of magnitude lower than those encountered in serum, which are measured by dye-binding methods. Since its first use in patients with diabetes, urine albumin concentration has shown to predict progression to all-cause chronic kidney disease (CKD) and be an independent risk factor for cardiovascular disease (3 ). Today, numerous clinical practice guidelines exist for the use of urine albumin to predict progression to CKD and cardiovascular risk (4 ). Two of the most cited are the Kidney Disease Improving Global Outcomes (KDIGO) and the American Diabetes Association (ADA) Standards of Medical Care in Diabetes (5, 6 ). Both of these define increased albumin excretion as an albumin:creatinine ratio (ACR) of 30 g/mg creatinine (mg/g creatinine) and severely increased albumin excretion as 300 g/mg creatinine (formerly called macroalbuminuria). The KDIGO guidelines define normal albumin excretion in “young” adults as 10 mg/g creatinine, thus leaving a gray-zone albumin excretion of 10 –30 mg/g creatinine. It is interesting that these values, which correspond to approximately 30 mg/24 h, are remarkably close to the original cut point of 15 g/min (20 mg/24 h) determined by Mogensen and Christensen in a study of only 43 patients! In this issue of Clinical Chemistry, Bachmann et al. point out the rather dismal current state of the art for urine albumin measurement (7 ). In a well-controlled and -performed study, 332 patient urine samples were analyzed by 17 different commercial methods and a reference isotope dilution mass spectrometry (IDMS) method. The careful control of sample handling, strict statistical treatment of the causes of bias, and the broad concentration range examined make the conclusions from this study indisputable. Namely, differences in median bias between methods can be as much as 37%– 45% across the concentration interval of 15–1000 mg/L. It is hoped that such a definitive study as this one will convince manufacturers to harmonize their methods to remove this variable from urine albumin measurement. Working together, the laboratory and clinical communities have done a remarkable job of standardizing and/or harmonizing tests used in clinical guidelines such as cholesterol (http://www.cdc.gov/ labstandards/lsp.html), glycated hemoglobin (Hb A1c) (http://www.ngsp.org), and serum/plasma creatinine (8 ). It is likely that because of studies such as that of Bachmann et al. the community can do the same for urine albumin. However, fixing the biases among urine albumin methods will not address the numerous other issues that confound the measurement of urine albumin excretion rates. These issues are addressed in a great detail in a recent review written by several of the same authors as the current study, but the points warrant repeating (4 ). First, there is a very large intraindividual variability, with estimates ranging as high as 103% but centering around 28%– 47%. There may be less intraindi1 Division of Laboratory and Genomic Medicine, Department of Pathology and Immunology, and 2 Renal Division, Department of Medicine, Washington University School of Medicine, St. Louis, MO. * Address correspondence to this author at: Division of Laboratory and Genomic Medicine, Washington University School of Medicine, 660 S. Euclid Ave., Campus Box 8118, St. Louis, MO 63110. Fax 314-362-1461; e-mail [email protected]. Received November 22, 2013; accepted December 2, 2013. Previously published online at DOI: 10.1373/clinchem.2013.218800 3 Nonstandard abbreviations: ESRD, end-stage renal disease; CKD, chronic kidney disease; KDIGO, Kidney Disease Improving Global Outcomes; ADA, American Diabetes Association; ACR, albumin creatinine ratio; IDMS, isotope dilution mass spectrometry; Hb A1c, glycated hemoglobin; CVi, coefficient of intraindividual variability; GFR, glomerular filtration rate; PTH, parathyroid hormone. Clinical Chemistry 60:3 000 – 000 (2014) Editorials