Analytical validation of an HPLC assay for urinary albumin

Background: Microalbuminuria is the earliest clinical finding for renal disease. Diabetic individuals often produce modified forms of albumin, perhaps due to impaired lysosomal processing, that are undetectable by common immunoassays but accurately measured by HPLC. Methods: We evaluated the performance of a commercially available, FDA-approved HPLC assay (AusAm Biotechnologies, NY) and compare results to our immunoturbidimetric assay (ITA, Beckman-Coulter, CA) using random urine specimens from 32 nondiabetic and 60 type 1 and 2 diabetic subjects. Results: The HPLC assay was linear to 963 mg/l with a limit of detection of 6.1 mg/l. Within-run and between-run precision was < 2% and 7-10%, respectively. Unpreserved urine was stable for at least 3 days at room temperature and 10 days at 4 degrees C. In both diabetic and nondiabetic subjects urinary albumin concentrations were higher by HPLC than by ITA, and many more diabetic and nondiabetic individuals were classified as microalbuminuric by HPLC than by ITA. The HPLC assay showed acceptable performance; however, because urinary albumin concentrations are higher in apparently healthy nondiabetic as well as diabetic subjects, different cutpoints will be necessary to accurately differentiate microalbuminuria. Conclusions: Prospective studies are necessary to determine whether the HPLC assay can effectively detect microalbuminuria earlier than current assays without a concomitant increase in the false positive rate. (c) 2005 Elsevier B.V. All rights reserved.