CARBOHYDRATE-DEFICIENT TRANSFERRIN IS NOT AFFECTED BY SERUM SEPARATORS

TORSTEN ARNDT 2 DIETRICH CZYLWIK 2 ROLF HACKLER 2 ANGELIKA HELWIG-ROLIG 1 2 THOMAS GILG 0 2 0 Ludwig-Maximilians-Universit8t, Institut ftlr Rechtsmedizin , Frauenlobstrasse 7a, D-80337 MOnchen, Germany 1 Abt. Klinische Chemie und Zentrallabor, Baldingerstrasse, D-35033 Marburg 2 Woscientia, Institut ftlr Laboruntersuchungen Ingelheim GmbH , Hamburger Strasse 1, D-55218 Ingelheim; Klinikum der Phihpps-Universitat , 'Med. Zentrum filr Innere Medizin, Abt. Kardiologie We studied me possible effects on serum carbohydrate-deficient transferrin (CDT) determination by a CDTect (Pharmacia) method of serum isolation in four different types of bloodcollection tubes, namely: (1) glass tubes (glass Vacutainer tubes with no additive); (2) S-Monovette Neutral tubes (plastic tubes with no additive); (3) S-Monovette Serum tubes (plastic tubes with kaolincoated plastic granulate coagulation accelerator); and (4) S-Monovette Serum/Gel tubes (plastic tubes with kaolin-coated plastic granulate and a polymerized acrylamide resin). Using Passing and Bablok regression analysis, we did not observe significant differences in CDT concentrations determined in 58 serum samples using any of these four blood-collection systems. - INTRODUCTION Carbohydrate-deficient transferrin (CDT), the most specific laboratory marker for chronic alcohol abuse so far, is measured in serum. Serum is usually obtained after clotting of the blood sample at room temperature followed by centrifugation. A coagulation accelerator (e.g. kaolin) and/or a separating gel (e.g. a polymerized acrylamide resin) are well-established tools to speed up and improve the separation of cells and blood clot from serum. During centrifugation, the serum separator moves, because of its welldefined density, between the packed cell layer and serum, inhibiting redistribution between blood clot and serum. However, the effects of serum separators on analyte concentrations are known. Thus, in serum samples stored in contact with the blood clot, Dasgupta et al. (1994, 1996) found significantly decreased concentrations of com*Author to whom correspondence should be addressed. monly monitored therapeutic drugs, e.g. phenytoin, phenobarbital, lidocaine, chinidine, and carbamazepine. As they showed by chemical extraction of the barrier gels, these decreases were due to adsorption of the analytes to the gel. Losses of tricyclic antidepressants (e.g. amitriptyline, imipramine, and clomipramine and their monodemethylated metabolites) in serum samples obtained from blood tubes containing a serum separator have also been demonstrated, e.g. by Nyberg and Martensson (1986) and Levy et al. (1987). In some cases, these losses were greater than 40% (Nyberg and Martensson, 1986). Variations in serum potassium concentrations after repeated centrifugation of stored tubes (Hue et al., 1991), after mail transport (Sandberg et al., 1988) or owing to an incomplete separating gel (Eichhorn et al., 1997) have been described. Altered concentrations of some organic solvents, e.g. ethylbenzene and xylene (Streete and Flanagan, 1993) or organochlorines (Longnecker et al., 1996) have also been reported. We have often by colleagues Tube types Glass vs plastic* Glass vs plasticykaolin Glass vs plastic/kaolin/gel Plastic vs plastic/kaolin Plastic vs plastic/kaolin/gel Plastic/kaolin vs plastic/kaolin/gel Slope Regression function 95% CI 1.000-1.231 0.933-1.200 0.833-1.030 0.9OO-1.050 0.75O-1.OOO 0.750-1.000 Intercept 95% CI -4.115-0.000 -4.100-0.433 -1.500-1.833 -1.500-1.200 -1.000-4.000 0.000-4.500 * Glass: Vacutainer No Additives; plastic: S-Monovette Neutral; plastic/kaolin: S-Monovette Serum; plastic/kaolin/gel: S-Monovette Serum/Gel. Since all 95% confidence intervals (CI) included the values 1 (slope) and 0 (intercept), the equality of CDT values in the serum obtained from the blood-collection tubes studied is demonstrated. whether similar effects have been studied for measurement of CDT. To our knowledge, there is no appropriate information available so far. We have therefore studied in the present paper the effects on CDT of four widely distributed bloodcollection tubes. MATERIALS AND METHODS The four widely used blood-collection tubes used in the present work were: (1) S-Monovette Neutral (a plastic tube without additives); (2) S-Monovette Serum (a plastic tube with kaolincoated plastic granulate as a coagulation accelerator); (3) S-Monovette Serum/Gel (a plastic tube with kaolin-coated plastic granulate and a polymerized acrylamide resin); and (4) Vacutainer No Additive tube (a glass blood tube). The first three of these tube types were supplied by Sarstedt (Niimbrecht, Germany), the Vacutainer tube was provided by Becton-Dickinson (Heidelberg, Germany). Blood was taken from 58 healthy persons by venipuncture using Multifly Sets (Sarstedt, Ntimbrecht, Germany). The tubes were filled in a randomized order. Serum was obtained after clotting for 30-45 min at room temperature followed by centrifugation at 2000 g for 10 min. CDT was determined by the CDTect-ElA assay in accordance with the instructions of the manufacturer. For statistical analysis, we used the method of Passing and Bablok (1983), which is independent of the assignment of the tubes to the X- and y-axes. Statistical calculations could therefore be reduced from 12 to six possible blood-tube combinations. RESULTS AND DISCUSSION The means (and ranges) of the CDT concentrations measured in 58 serum samples obtained from each type of blood-collection tube were (in U/l) as follows: 19.6 (7-51) for Vacutainer No Additive (a glass blood tube), 19.5 (8-48) for S-Monovette Neutral (plastic tube with no additive), 19.1 (7-62) for S-Monovette Serum (plastic tube with kaolincoated plastic granulate as a coagulation accelerator), and 18.6 (8-48) for S-Monovette Serum/ Gel (plastic tube with kaolin-coated plastic granulate and a polymerized acrylamide resin). The differences in means are clearly below the upper limits for the within-day and pure betweenday coefficients of variation of the CDTect-EIA assay of <9.2% and <14% respectively (Arndt et al, 1998*). Table 1 shows the correlation functions and the corresponding 95% confidence intervals (CI) of the slopes and intercepts of these six tube combinations. Figure 1 illustrates the correlation of CDT values measured in serum samples which were obtained from the Vacutainer No Additive (glass) tube and the S-Monovette Serum/Gel (plastic, kaolin and gel) tube as the bloodcollection tubes which differed most in our CDT AND SERUM SEPARATION DEVICES Vacutainer No Additives Fig. 1. Comparison of carbohydrate-deficient transferrin (CDT) values obtained in serum from a glass bloodcollection tube without coagulation accelerator or serum separator (Vacutainer No Additives) and a plastic tube with kaolin-coated plastic granulate and polymerized acrylamide resin as a serum separator (S-Monovette Serum/Gel). The Passing and Bablok (1983) regression function is Y (...truncated)