High prevalence of increased trisialotransferrin concentrations in patients with anorexia nervosa: implications for determination of carbohydrate-deficient transferrin.

Background: Carbohydrate-deficient transferrin (asialo-+monosialo-+disialotransferrin, CDT) is currently the most specific laboratory marker of chronic alcohol abuse. We tested whether previous findings of false-positive CDT results for anorexia nervosa patients have been due to invalid CDT analysis methods or anorexia nervosa by itself.

Methods: Serum CDT from 49 anorexia nervosa patients, 14 bulimia nervosa patients and 22 healthy controls (all adolescent, female and age-matched) was determined in a retrospective study by HPLC (Clin-Rep-CDT-in-serum-online, cut-off > or =1.8%, Recipe), by capillary electrophoresis (Capillarys-CDT, cut-off > or =1.3%, Sebia) and (due to limited surplus serum volume for a subset of 18 anorexia nervosa patients with increased trisialotransferrin detected by HPLC) by immunoassay based on anion-exchange CDT and non-CDT fractionation (%CDT-TIA, cut-off > or =2.6% CDT, Bio-Rad).

Results: HPLC and capillary electrophoresis: No false-positive CDT results were obtained. Asialo- and monosialotransferrin were not detected and disialotransferrin (CDT) was in each case clearly below the test-specific cut-offs. Trisialotransferrin (a non-CDT isoform) was increased (cut-off > or =5.0% for HPLC) in 33 anorexia patients, 2 bulimia patients and 2 controls. %CDT-TIA: 8 false-positive CDT results of > or =2.6% out of the 18 samples tested (CDT-range/mean/median 2.6-4.6/3.2/2.8%).

Conclusions: Anorexia nervosa does not cause by itself increased CDT results. False-positive CDT values from the past are most likely due to an incomplete separation of trisialotransferrin from CDT and thus overdetermination of CDT. Immunological CDT testing without confirmatory analysis by HPLC or CE is no longer acceptable.