Urine Total Protein Measurement with the Vitros Dry Reagent Technology: Modification of Diluent to Re- solve Positive Bias of Diluted Samples, Patrick Lionel

The measurement of urinary total protein continues to have a place in the repertoire of clinical laboratory tests, particularly in the clinical assessment of patients with renal disease. Although the test volume is often modest, there exists the need to perform the assay on the major automated chemistry analyzers currently in use in most hospital laboratories. We have evaluated the performance of a new urine total protein procedure developed for dry reagent slide technology for use on the Vitros 950 analyzer (Johnson & Johnson). In this method, protein in the sample binds to a pyrocatechol violet–molybdate complex, resulting in an absorption shift. The method is calibrated with aqueous solutions of bovine serum albumin supplied by Johnson & Johnson. We compared the new method with that in current use on the Cobas Fara analyzer (Hoffmann–La Roche), which involves pyrogallol red dye binding of urine total protein [1] calibrated with human serum albumin (Biotrol). For this comparison we analyzed 53 patient samples (median 0.65 g/L, range 0.02–5.44 g/L, pyrogallol red results) and obtained the following Deming linear regression analysis: r 5 0.807, slope 5 3.01, intercept 5 21.19 g/L, Syux 5 0.72 g/L (x 5 Cobas Fara pyrogallol red, y 5 Vitros method). These results showed a strong positive bias with the Vitros method with respect to the pyrogallol red method. The differences between the procedures appeared greatest in samples that required dilution, i.e., those .2.00 g/L, the linear range of both methods. Dilutions, when required, were carried out with deionized water in accordance with both manufacturers’ instructions. Deming linear regression analysis of the samples that did not require dilution (n 5 36, median 0.23 g/L, range 0.02–1.51 g/L, pyrogallol red results) gave the following: r 5 0.980, slope 5 1.16, intercept 5 0.06 g/L, Syux 5 0.06 g/L. Thus the discrepancies between the two methods were markedly increased when dilutions were involved. At this point, three specimens were analyzed with Vitros methodology with various dilutions in deionized water; the results are given in Table 1. Clearly, urine total protein measurements with Vitros methodology are dilution dependent, showing an increasing positive bias the greater the dilution. Similar results were obtained with human serum albumin/deionized water solutions (data not shown). We hypothesize that the denaturing conditions of a deionized water dilution enhances the reaction of the proteins. Because urine contains a relatively high concentration of potassium, the three specimens were then diluted with 150 and 75 mmol/L KCl solutions and analyzed as before. These results are also shown in Table 1. The KCl concentrations were chosen to reflect the normal range of potassium excretion in urine (25–125 mmol per day [2]). They show that the use of a KCl solution effectively reduces the large positive bias observed when deionized water alone is used as the diluent. The same response to KCl was observed when human serum albumin solutions were analyzed (data not shown). To validate further the modified dilution protocol we compared the pyrogallol red and Vitros methods to a manual biuret reference procedure. This method provides essentially equal sensitivity for different proteins and is claimed to be the most accurate procedure for urine total