Exclusion of trisialo-transferrin from carbohydrate-deficient transferrin measurement: advantage in patients with chronic liver disease?

BACKGROUND: Biological markers for chronic alcohol consumption like MCV or gammaGT or carbohydrate deficient transferrin (CDT) are useful, but far from being perfect. In patients with liver disease a reliable marker for chronic alcohol consumption as the underlying etiology is highly needed. Recently, a new ELISA based version of the carbohydrate-deficient-transferrin (CDT-TRISIALO (-)) assay has been developed, which measures asialo-, monosialo- and disialo transferrin, but excludes trisialo- transferrin; that modification suggests higher sensitivity and specificity in detecting recent alcohol consumption in patients. AIMS: The study goal was to evaluate the sensitivity, specificity, positive and negative predicitive value of this new carbohydrate-deficient-transferrin assay (CDT-TRISIALO (-)) in a group of patients with liver disease and to compare the results with that of the established CDT assay (CDT-TRISIALO (+)). PATIENTS AND METHODS: Our study population consisted of 110 consecutive patients (male: n = 80 [72.7 %], female: n = 30 [27.3 %]) with liver disease of the following etiologies: chronic alcohol consumption (n = 51 [46.4 %]; Out of them 30 alcohol abusing patients were assessed by cage = 1 and 21 alcohol dependent patients were assessed by cage = 2, chronic viral hepatitis (n = 33 [30.0 %]) including 25 [22.7 %] patients with chronic hepatitis C infection and 8 [7.3 %] patients with chronic hepatitis B infection), haemochromatosis (n = 4 [3.6 %]), mechanical cholestasis (n = 17 [15.5 %]) and other liver diseases (n = 5 [4.6 %] including autoimmune hepatitis (n = 2) and primary biliary cirrhosis (n = 3)). 27.3 % of our patients (n = 30) had no liver cirrhosis whereas the majority (72.7 %, n = 80) had liver cirrhosis.
RESULTS: In our population of liver disease patients the CDT-TRISIALO (-) assay had a sensitivity of 72.7 % and specificity of 58.1 % for recent alcohol consumption at the published cutoff level of 2.6 %. The positive predictive value was 34.0 % and the negative predictive value was 87.8 %. Sensitivity and specificity of the CDT-TRISIALO (+) assay at the recommended cutoff level of 4.7 % were similar, 77.3 % and 49.3 %, respectively. The positive and negative predictive values were 30.9 % and 88.1 %. CDTTRISIALO (+) and CDT-TRISIALO (-) levels increased significantly with higher Child-Pugh stages.
CONCLUSION: The newly developed carbohydrate deficient transferrin test (CDT-TRISIALO (-)) is of no advantage as compared to the established assay (CDT-TRISIALO (+)) when used in a patient population with liver disease. In that population, normal CDT-TRISIALO (-) helps to exclude recent alcohol consumption; this results from the high negative predictive value of a normal CDT-TRISIALO (-).