Human serum transferrin glycosylation pattern : population differences, analytical methodology and application as a biomarker for testing of alcohol abuse and CDG

Abstract

Alcohol use and abuse is a major social and economic problem in many societies. Carbohydrate-deficient transferrin (CDT) is a widely used and highly specific biochemical alcohol marker of prolonged alcohol abuse. Increased knowledge about the clinical characteristics of CDT may lead to better possibilities of employing CDT as a biochemical alcohol marker of risky and heavy alcohol consumption and thus creating better opportunities for prevention and intervention of alcohol dependence. The aim of this thesis was to contribute to this knowledge by investigating the properties of CDT in different populations and evaluating the sensitivity and specificity of different analytical methodologies for analysis of CDT.

Serum samples (n=1387) from subjects originating from five different countries were analyzed with a HPLC candidate reference method for CDT. In non-drinkers there were minimal differences in the serum transferrin glycoform pattern with respect to different ethnicity, gender, age and BMI. When evaluating disialotransferrin, the primary glycoform in CDT, with respect to the same categories, no clinically significant differences were detected. Furthermore the overall test accuracy for identification of heavy drinkers (>210 g ethanol/week for men and >140 g ethanol/week for women) showed no gender difference.

Serum samples (n=178) were analyzed from subjects with clinical or pharmacological factors previously reported to cause false-positive CDT levels. Only ~5% showed a relative disialotransferrin level exceeding the upper limit for the reference interval, leading to a conclusion that earlier reports on reasons for false positive CDT values are linked with the methodology used rather than with true physiological influences.

When compared with a HPLC candidate reference method, the Bio-Rad %CDT HPLC test and the CEofix CDT assay proved to be appropriate for confirmatory and routine %CDT testing, showing an overall good correlation and agreement. The HPLC candidate reference method could readily be used for preliminary diagnosis of CDG and for assignment of cases to either CDG-I or CDG-II.