Impact of antibody specificity and calibration material on the measure of agreement between methods for cardiac troponin I

Available assays for cardiac troponin I (cTnI) yield numerically different results. The aim of this study was to compare patient values obtained from four cTnI immunoassays.We studied the Stratus(R) II assay, the Opus(R) II assay, the Access(R) assay, and a research-only cTnI heterogeneous immunoassay that uses the Dade Behring aca(R) plus immunoassay system equipped with two new noncommercial monoclonal antibodies. Because the aca plus cTnI assay is for research only, we first evaluated and analytically validated it for serum and citrated plasma. Initially, each method was calibrated using the method-specific calibrator supplied by each manufacturer; however, the aca plus cTnI assay was calibrated using patient serum pools containing cTnI and selected on the basis of increased creatine kinase MB isoenzyme and with values assigned by use of the Stratus cTnI assay. For method comparisons, individual patient sample cTnI values were determined and compared with the Stratus II assay.Passing and Bablock regression analysis yielded slopes of 1.44 (r = 0.96; n = 72) for the Opus II vs Stratus II assays; 0.07 (r = 0.91; n = 72) for the Access vs Stratus II assays; and 0.90 (r = 0.91, n = 72) for the aca plus vs Stratus II assays. The recalibration of each method with a Stratus II-assigned serum pool improved, but did not entirely eliminate, the slope differences between the different assays (range, 1.00-1.16). The observed scatter in the correlation curves remained.There is a need to further explore the specificities of these assays with respect to the different circulating forms of cTnI.