DIAGNOSTIC CHARACTERISTICS OF DIFFERENT CARBOHYDRATE-DEFICIENT TRANSFERRIN METHODS IN THE DETECTION OF PROBLEM DRINKING: EFFECTS OF LIVER DISEASE AND ALCOHOL CONSUMPTION

Alcohol & Alcoholism Vol. 38, No. 5, pp. 415–420, 2003 doi:10.1093/alcalc/agg102, available online at www.alcalc.oupjournals.org DIAGNOSTIC CHARACTERISTICS OF DIFFERENT CARBOHYDRATE-DEFICIENT TRANSFERRIN METHODS IN THE DETECTION OF PROBLEM DRINKING: EFFECTS OF LIVER DISEASE AND ALCOHOL CONSUMPTION PETRA ANTTILA1, KIMMO JÄRVI2, JAANA LATVALA1, JOAN E. BLAKE4 and ONNI NIEMELÄ1,3* Departments of 1Clinical Chemistry and 2Psychiatry, EP Central Hospital, Seinäjoki, 3University of Tampere, Finland and 4Addiction Research Foundation, Toronto, Ontario, Canada (Received 3 March 2003; first review notified 7 April 2003; in revised form 23 April 2003; accepted 1 May 2003) INTRODUCTION characteristics of the CDT assays may be markedly different from each other. Studies with the conventional CDTect method have found high diagnostic sensitivities, although also a high rate of false-positive results among women and in patients with high serum transferrin levels (Viitala et al., 1998). Concerns over the impact of liver status have also been raised (Tsutsumi et al., 1994; Niemelä et al., 1995; DiMartini et al., 2001). We designed the present study to compare the conventional CDTect method and %CDT assays for their diagnostic value to identify problem drinking in alcoholic patients with or without liver disease. Separate versions, which have been made available for the %CDT assay, allow the comparisons between the assays with (%CDT–TIA) or without (%CDT) reactivity towards the trisialofraction. The data suggests that the %CDT assay, which is devoid of such reactivity, offers several advantages over the previous CDT methods. Although serum carbohydrate-deficient transferrin (CDT) has been widely used as a marker of alcohol misuse, the interpretations of the assay results have suffered from a lack of uniform international standardization (Arndt, 2001; Helander et al., 2001a; Conigrave et al., 2002; Helander, 2002; Tagliaro et al., 2002; Whitfield, 2002). In healthy people, the most common transferrin isoform is tetrasialotransferrin, whereas, as a result of ethanol misuse, various degrees of deficiencies in the sialic acid content of transferrin may be generated (Stibler, 1991; Allen et al., 1994; Helander et al., 2001a). CDT refers to those isoforms that have a reduced sialic acid content in oligosaccharidic chains bound to amino acids 413 and 611. Usually this definition means isoforms with 0–2 or 0–3 sialic acid side-chains (Stibler, 1991; Allen et al., 1994; Arndt, 2001; Helander et al., 2001a; Helander, 2002). Recently, the relative importance of the various CDT isoforms in the clinical assays for detecting alcohol misuse has received increasing attention (Scouller et al., 2000; Whitfield, 2002). Although measurements with improved specificity towards such isoforms have been introduced, only a few studies are available on the comparisons between the different methods in clinical materials (Helander, 1999; Anton et al., 2001). The new assay approaches, which exclude the trisialo isoform (the isoform with three sialic acid side-chains) from the measurement, have recently shown high diagnostic accuracies in the detection of alcohol misuse (Legros et al., 2002). Previous studies have also indicated that the diagnostic SUBJECTS AND METHODS Study subjects We studied 62 patients with alcoholism who had a welldocumented history of continuous alcohol consumption or binge drinking. These included a sample of 33 with biopsyproven liver disease (seven females; 26 males), who had a history of continuous alcohol consumption for at least 5 years in amounts exceeding 80 g/day. The severity of alcoholic liver disease was assessed according to previously established clinical and laboratory (CCLI) and morphological (CMI) criteria (Blake and Orrego, 1991). The CCLI index combines laboratory and clinical data, which have been shown to be significantly associated with the prognosis of alcoholic patients. When the *Author to whom correspondence should be addressed at: EP Central Hospital, Laboratory, FIN-60220 Seinäjoki, Finland. Email: 415 Alcohol & Alcoholism 38 © Medical Council on Alcohol 2003; all rights reserved Abstract — Aims: Due to methodological heterogeneity, conflicting views have been expressed on the validity of CDT measurements in the detection of alcohol misuse. Methods: We compared the characteristics of the conventional CDTect method and the Axis turbidimetric CDT assays in the assessment of 62 alcoholics, who were either with (n = 33) or without (n = 29) liver disease, as analysed by combined clinical, laboratory, and morphological indices. Controls were 45 healthy volunteers who were either social drinkers or abstainers. Results: In the total sample of alcoholics, the sensitivity of the %CDT method, which excludes the trisialotransferrin isoform from the measurement, was 63% for men and 46% for women, as compared to 65% and 36% of CDTect, respectively. Both of these methods showed higher sensitivities than the %CDT–TIA method, which reacts with trisialotransferrin (32% and 25%, respectively). The assay specificities were 100% for men and 91% for women with %CDT, and 96% and 87% with the CDTect, respectively. The correlation between the CDTect and %CDT method was higher in men (r = 0.86) than in women (r = 0.57). The presence of liver disease was found to influence the results of the CDTect method, such that the highest CDT concentrations were observed in patients with mild to moderate liver disease, especially among women, whereas the %CDT method was less sensitive to the effect of liver pathology. The self-reported alcohol consumption from the 4 weeks prior to sampling showed a higher correlation between the %CDT results (r = 0.64, P < 0.0001) than with the CDTect results (r = 0.40; P < 0.01). Conclusions: The data indicate that the new %CDT method offers advantages over the previous versions of the CDT methods. The improved characteristics may be most useful in assays for excessive alcohol consumption in female alcoholics, patients with liver disease, and in patients with abnormal serum transferrin concentrations. 416 P. ANTTILA et al. Table 1. Main clinical and laboratory characteristics of the study population N Men/women Age GT Men Women MCV AST Men Women ALT ALP BIL ALB Alcoholics with liver disease Alcoholics without liver disease Controls 33 26/7 48 ± 9 309 ± 354*** 360 ± 371*** 84 ± 84 99 ± 4.2*** 82 ± 76*** 83 ± 81** 73 ± 50*** 99 ± 35*** 185 ± 59** 19 ± 12** 41 ± 5 29 25/4 46 ± 12 123 ± 139 110 ± 114 207 ± 267** 96 ± 4.8*** 41 ± 18 39 ± 16 53 ± 25 46 ± 20** 153 ± 48* 12 ± 6 43 ± 7 45 22/23 51 ± 16 25 ± 10 29 ± 11 20 ± 8 92 ± 3 30 ± 24 36 ± 32 23 ± 5 28 ± 15 126 ± 30 13 ± 5 42 ± 3 The values are mean ± SD. For GT and AST with sex-specific reference intervals the data are shown separately for men and women. ALB, albumin; ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BIL, bilirubin; GT, gamma glutamyltransfera (...truncated)