COMPARISON OF DIFFERENT METHODS FOR DETECTING CARBOHYDRATE-DEFICIENT TRANSFERRIN

Different methods for detecting carbohydrate-deficient transferrin (CD T) were compared. In addition, their efficiency for detecting alcohol abuse among men not having clinical evidence of liver disease was stud ied in controls (n = 26), weekend (n = 16) end daily (n = 12) heavy dr inkers, and alcoholics (n = 28). Comparisons were made between anion e xchange separation of iron-saturated transferrin (Tf) by microcolumns (CDTect) and by the Fast Protein Liquid Chromatography (FPLC% and FPLC -MG), followed by double-antibody radioimmunoassay of collected fracti ons. Tf fractions with pl greater than or equal to 5.7 were also measu red by two different isoelectric focusing (IEF) methods, followed by i mmunofixation (SA-IEF-CDT and IEF-CDT-TOT), the latter method being us ed also for detection of asialo-transferrin (IEF-CDT-AS). The cut-off was 20 units/liter for CDTect, 4.4% of total Tf for SA-IEF CDT, and th e mean + 2 sd of the control group for FPLC-MG (as mg/liter of Tf), FP LC-%, IEF-CDT-TOT, and IEF-CDT-AS (all as percentage of Tf). The overa ll accuracies (combining sensitivity and specificity) for detecting he avy drinkers of CDTect, FPLO (mg/liter), FPLC (%), SA IEF-CDT, IEF-CDT -TOT, and IEF-CDT-AS were 63%, 59%, 61%, 74%, 57%, and 63%, respective ly; for detecting alcoholics, 87%, 83%, 81%, 89%, 67%, and 76%, respec tively. In conclusion, the methods were in rather good agreement with each other. Diagnostic characteristics among heavy drinkers and correl ations between methods differed slightly, probably depending on the ab ility of different methods to separate and detect asialo-, monosialo-, and disialotransferrin. The overall accuracy among alcoholics was sim ilar using each of the methods tested. The same was true among heavy d rinkers, although SA-IEF-CDT had the highest sensitivity and overall a ccuracy among this group, making the method slightly better than the o ther ones.