Measurement of carbohydrate-deficient transferrin by isoelectric focusing/western

At present, the most reliable marker of recent and heavy alcohol intake is carbohydrate-deficient transferrin (CDT). While most CDT quantitation methods (including immunofixation and micro anion-exchange chromatography [MAEC] combined with radio-immunoassay [RIA]) either lack the precision required for diagnostic usage or are not commercially available, we recently described an isoelectric focusing/Western blotting (IEF/WB) procedure that provides sensitive and specific assessment of serum CDT content. However, a modified MAEC/RIA kit, supposedly more reliable than the original, is also being advanced as suitable for widespread clinical application. Therefore, we compared this modified MAEC/RIA procedure to the IEF/WB method of CDγ quantitation in the following 108 subjects; 53 alcoholics undergoing detoxification without clinical or histological evidence of liver disease, 24 recently drinking alcoholics with biopsy-proven liver disease, eight alcoholics abstinent for more than 30 days with biopsy-proven liver disease, seven non-drinking patients with non-alcoholic liver disease, and 16 healthy controls. Although CDT measurements by the two methods were correlated (r = 0.60, P <0.01), serum CDT values obtained with IEF/WB were nearly five-fold higher than those obtained with MAEC/RIA (e.g. 140.0 ± 58 versus 28.5 ± 16 mg/l among the active drinkers). Of the two methods, IEF/WB exhibited significantly greater sensitivity than MAEC/RIA for detecting recent, heavy drinking (75% versus 61%, P < 0.05) and generated no false positives whereas MAEC/RIA gave falsely elevated CDT levels in 37% of the abstinent alcoholics. Receiver operating characteristic (ROC) analysis, which was used to assess the accuracy of the two methods for discriminating alcoholics admitted for detoxification from healthy controls, suggested an increase in area under the ROC curve for IEF/WB compared to MAEC/RIA (0.92 versus 0.86, P = 0.09). For discriminating recent drinkers with liver disease from non-drinking patients with liver disease (including abstinent alcoholics), the area under the ROC curve was significantly greater for IEF/WB compared to MAEC/RIA (0.97 versus 0.68, P < 0.001). Serum CDT/total transferrin ratios, the latter of which was measured by ELISA, offered no advantage over serum CDT alone as determined by either method for distinguishing heavy drinking. Our results indicate that for detecting recent and heavy alcohol misuse, the IEF/ WB method of serum CDT measurement has a greater degree of diagnostic accuracy than the modified MAEC/RIA method, particularly in the presence of liver disease.