Abstract
The o-toluidine-boric acid (o-TB) method for blood glucose estimation, which is being used widely in clinical laboratory, gave an unstable blue color following heating with a glucose. According to the recent reports by Ceriotti & Frank, the authors proposed a modified o-TB reagent to produce a stable chromogen in the case of direct estimation of plasma glucose without deproteinization.
This reagent which was prepared in the preliminary experiment consisted of 35% water, 50% glacial acetic acid, and 15% o-toluidine, while the contents of boric acid and thiourea were the same as the original reagent. Although this reagent enabled the authors to obtain a stable coloration at least for 24 hrs. when heated for 15 min., a slight turbidity was also produced, which was caused probably by protein in the plasma samples. The glucose values estimated were, therefore, slightly higher than those obtained with the other o-TB reagents or when compared with indicated values of commercial reference serums. In vitro addition of human γ-globulin or bovine serum albumin into a sample plasma did not cause any turbidity until 7 or 9g/dI of their final concentrations.
Addition of detergents such as Tween 20, 40, 60, 80, Tirton X-100, Brij 35, or glycerol, formamide, ethyleneglycol, xylene, toluene into the reagent did not eliminate the turbidity.
The authors finally recommended a reagent containing water, glacial acetic acid, o-toluidine, thiourea and boric acid in 10, 75, 15, 2.5 and 2.4% concentrations, respectively. The color intensity reached to a maximum by heating for 4-5 min., then decreased to a plateau beyond 15 min. Heating for 15 min., followed by cooling in running water produced a typical but slightly weakened blue color of glucose with absorption maximum at 635 nm. And no turbidity was seen in case of direct estimation of glucose without deproteinization. Higher accuracy was observed by using commercial reference serums and pooled serum, although slightly higher than those to the o-TB reagent. Excellent reproducibility was obtained and the coefficient of variation was 1.3-1.7%. By a comparative estimation of 100 patients' serums, good correlation was found between the glucose values assayed with the present and other commercial o-TB reagents. Their mean values were 151.4 and 152.4 mg/dl, respectively.
The present method was found to be suitable for a rapid and simple estimation of plasma glucose which does not require preceded deproteinization and give stable coloration by heating.
