The detection of D-dimer in plasma by enzyme immunoassay: improved discrimination is obtained with a more specific signal antibody.

Most commercial D-dimer enzyme immunoassays employ two antibodies, a fibrin specific capture monoclonal and a less specific antibody cross-reacting with epitopes present on fibrinogen. Two different ELISA systems were compared to test whether this cross-reaction can lead to overestimation of the true levels of cross-linked fibrin derivatives. Both assays used the same D-dimer specific antibody for capture, DD-3B6/22, but utilized signalling antibodies which differed in fibrinogen reactivity. The assays gave low results for 210 normal samples (conventional ELISA, 39 +/- 45 ng/ml; non-fibrinogen reactive ELISA, 23 +/- 20 ng/ml) with a high degree of elevation for 53 patients with active thrombosis (conventional ELISA, 901 +/- 649 ng/ml; non-fibrinogen reactive ELISA, 1,906 +/- 1,725 ng/ml). However, a dramatic difference between results was seen when fibrinogenolysis was induced by treatment of 20 normal plasmas with high levels of t-PA and plasminogen. The D-dimer levels estimated with the conventional fibrinogen-reactive signal antibody rose 25-fold (normal, 36 +/- 27 ng/ml; treated, 833 +/- 272 ng/ml), whereas no increase was obtained with the non-fibrinogen reactive ELISA (normal, 24 +/- 21 ng/ml; treated, 27 +/- 22 ng/ml). These results suggest that a more accurate estimation of D-dimer levels in samples containing high levels of fibrinogen derivatives is achieved in assays incorporating non-fibrinogen reactive antibodies.