[Analysis of the dilution deviation in CA19-9 measurement]

CA19-9 widely used as a tumor marker of the pancreas and a bile duct. There are a number of reports which describes the measured value discrepancies between RIA and non-RIA kits. RIA results also have shown lack of the linearity over 70 U/ml when the samples are diluted. The pH condition at assay reaction for RIA had been suggested as the major reason, it has been denied by the results from the same pH condition at assay reaction used by COBAS CORE CA19-9 EIA II. On the other hand, the lack of RIA antibody titer is indicated for the discordant results by changing the sample volume to reagent volume ratio in the reaction. Our further investigation also indicates that the specific Lewis blood type, i.e. Le (a-b+), shows the linearity issues by RIA. The discrepancies are not caused by the reaction pH, but the amount of the antibody used in the RIA kit is closely associated. Considering the CA19-9 antibody nature used in RIA kit, which covers broad molecular range, users need to pay more attention to setting up each laboratory's measuring range.