Biomarkers in diabetic kidney disease: good use and pitfalls.

Abstract

-ses, cystatin C-based formulas showed the best performance, followed closely by the new CKD-EPI creatinine-based equation. Their findings highlight the crave for more precise measures of GFR estimation and waiving of less efficient methods such as Cockcroft-Gault formula. When compared to MDRD equation the new CKD-EPI equation is more accurate and reduces the known effect of MDRD on sub-estimating GFR in normal and near-normal values (2). However, the CKD-EPI equation based on creatinine also requires a race definition, which is always problematic, particularly in highly admixture populations. This formula has not been largely validated outside US, Europe and Australia and this should be done in populations ethnically different from the population in whom the equation was derived and revalidated, such as Brazil. Small Brazilian studies yielded results suggesting that this equation may be applied in our population (3-5), but larger studies are still necessary. On the other hand, cystatin C seems to improve accuracy in comparison to creatinine, but is still not largely avail-able and its cost is still high. These problems could be minimized with a broader use of this marker, but to the present cystatin C use is still very much restricted to the research field.In the other manuscript, Polat and cols. (6) have analyzed several endothelial dys-function biomarkers in 73 patients with diabetes and either normoalbuminuria and microalbuminuria in comparison to healthy controls. The authors showed that al-though differences can be detected among the diabetic and non-diabetic groups, most biomarkers cannot discriminate those with and without albuminuria. Only sVCAM-1 showed a better performance in identifying those with microalbuminuria.Reflections raised by the two papers come in accordance with a very actual discus-sion on diabetic kidney disease. One of the most fearest complications of long-term poor glycemic control, CKD in diabetes still faces old problems regarding two very basic problems: diagnosis and classification. Diagnosis has always been supported by the recognition of long term hyperglycemia, increasing albuminuria occurring along to GFR loss (represented by the hyperfiltration, microalbuminuric and macroalbu-minuric phases), the presence of other microvascular injury, especially diabetic reti-nopathy, and absence of signs of other kidney diseases. However, it has been recently recognized that the classical description of the hyperfiltration, micro- and macroal-buminuric phases is limited, since a significant number of diabetic patients with CKD (and losing GFR over time) may in fact present normoalbuminuria and or micro-