We measured serum levels of carbohydrate deficient transferrin (CDT) in 420 subjects: 100 healthy blood donors, 82 healthy employees, 70 abstaining patients with different chronic nonalcoholic liver disease, 16 abstaining patients with alcoholic fatty liver, 50 abstaining patients with alcoholic liver cirrhosis, 25 abusing patients with alcoholic fatty liver, 41 abusing patients with alcoholic liver cirrhosis, and 36 patients with alcohol dependence syndrome with a daily ethanol consumption of 173 +/- 120 g the last 4 weeks before blood was drawn. In controls the serum level of CDT was significantly higher in females compared with males (17.7 +/- 5.1 and 13.7 +/- 3.8 units/liter, respectively), and the upper normal limit was defined as 27 and 20 units/liter. Sixty-two of 102 (60.8%) abusing patients with alcoholic liver disease had increased levels of CDT compared with 1 of 66 abstaining (1.5%) patients with alcoholic liver disease, and 10 of 70 (14.3%) abstaining patients with nonalcoholic liver disease among them 3 with primary biliary cirrhosis and 2 with chronic autoimmune hepatitis. No correlation was found between serum CDT and gamma-glutamyltranspeptidase (GGT), AST, ALT, and mean red cell volume (MCV). The sensitivity and specificity for serum CDT was 61 and 92%, respectively, compared with 85 and 18% for GGT and 70 and 66% for MCV. No advantage was gained by using the CDT/transferrin ratio. Our study confirms that CDT is a specific marker for chronic alcohol abuse, except in few patients with other chronic liver diseases. Serum CDT seems to be a better indicator of abstention than GGT; AST and MCV in patients with alcoholic liver disease. However, in our hands CDT is not so sensitive for alcohol abuse in patients with liver disease as reported earlier in unselected alcoholics.

The performance of CDTect (Kabi-Pharmacia, Uppsala, Sweden) and two versions of AXIS % CDT (AXIS Biochemicals, Oslo, Norway) for determining carbohydrate deficient transferrin (CDT) was examined in 502 consecutive patients admitted to the Department of Medicine, Aker University Hospital. The sensitivity for detecting an alcohol consumption > or = 50 g/day for the last 4 weeks was 69% for CDTect, 65% for AXIS % CDT, version 1 (AX CDT 1) and 50% for AXIS % CDT, version 2 (AX CDT 2). The specificity at the same level of alcohol consumption, markedly differed between the two methods: 92%, 76% and 90% for CDTect, AX CDT 1 and AX CDT 2, respectively. The variation coefficient (day-to-day) was 10%, 22% and 10% for CDTect, AX CDT 1 and AX CDT 2, respectively.

No abstract

No abstract