Evaluation of an Automated CLIA System for the Determination of Carbohydrate Antigen 19-9

AIM: To compare analytical characteristics between two different methods, a automated chemiluminescent immunoassay (CLIA) versus a radioimmunometric assay (RIA), for the determination of carbohydrate antigen 19-9 (CA19-9). MATERIALS AND METHODS: One hundred and eighty-five blood samples from consecutively enrolled individuals (87 males and 98 females, aged 22-89 and 29-89 years, respectively) were evaluated. For both assays, a cut-off of 37 U/ml was used. RESULTS:Comparison between assays was analyzed using Passing-Bablok regression, which showed a high inter-assay correlation (r=0.926). The mean intra-assay coefficient of variation was 17.6% (range=6.1-50.7%) for the RIA (n=10 samples) and 5.2% (range=0.4-17.0%) for the CLIA assay (n=26 samples). CONCLUSION: This study showed that CLIA for CA19-9 has a good reliability on all samples analyzed and should be preferred to RIA with the aim of reducing costs, greater standardization and more harmonized results.