IntroductionHuman epididymis protein 4 (HE4) is an emerging tumor marker in diagnosis of ovarian cancer. Together with the cancer antigen CA 125 it can improve sensitivity and specificity by calculating the risk of ovarian malignancy algorithm (ROMA) score. Our purpose is to validate performance characteristics of the electrochemiluminiscent Cobas e411 assay for HE4, including the endogenous interference study. Material and methodsTwo different levels of commercial control were analyzed for imprecision and trueness study during 20 days. Limit of blank and limit of detection was assessed with 20 replicates of a sample lacking HE4. Linearity was assessed by regression analysis mixing specimens, with high HE4 concentration and a devoid analyte one. In addition carry-over assay and endogenous interference study were conducted. ResultsThe imprecision obtained for each level were 1.7 and 2.8% for within run, 2.59 and 1.28%, for between run and 4.67 and 4.47% for total imprecision. The relative biases were 6.11 and 1.54% for each level. The limit of blank was 1.25 pmol/L and the limit of detection 2.21 pmol/L. The linearity study yielded the following regression equation y=-18.22+0.98x, demonstrating a linear behavior. There was not significant carry-over. The method was not affected by hemolysis, bilirrubin, neither lipemia. ConclusionsElectrochemiluminiscence method for HE4 showed an excellent imprecision and trueness. The limit of detection resulted lower than the declared by the manufacturer. Linearity was verified within the limits studied. In addition the assay is devoid of carry-over and free of significant endogenous interferences.
Read full abstract