Abstract
The aim of this study was to validate an automated immunoturbidimetric assay used to quantify human albumin in urine and to accurately measure canine albumin concentrations in both urine and cerebrospinal fluid. The partial homology existing between human and canine albumin limited the accuracy of the human assays in measuring canine albumin without method modifications. Thus, the assay was modified by calibrating the analyzer with calibrators made in the laboratory containing known concentrations of canine albumin. To prepare the set of calibrators, the albumin concentration of pooled sera of healthy dogs was assessed in 5 replicates using the BromocresolGreen assay. Pooled samples were aliquoted and serially diluted to obtain the expected concentrations of albumin (0.5, 1, 5, 13, and 30 mg/dl) for establishing the canine calibration curve. Thereafter, the performance was assessed by analyzing canine urine and CSF. The modified assay accurately quantified canine albumin in both specimens, as indicated by the following. Intra- and interassay variability was 0.92% and 2.74%, respectively; recovery was 99.66% and 99.07% in urine and 105.02% in CSF. No interference was detected when hemolysate and glucose were added to urine. The test was linear within the verified range (0–225 mg/dl). These results demonstrate that the modified human albumin immunoturbidimetric assay can be a useful tool in the veterinary diagnostic laboratory. It is accurate and tends itself to automatization on chemistry analyzers.