Plasma D-dimers: comparison of ELISA with a new, rapid, quantitative latex assay.

Abstract

A new, rapid quantitative latex assay (STA-Liatest D-Di) was compared with an enzyme-linked immunosorbent assay (ELISA) method, to exclude deep venous thrombosis (DVT) and pulmonary embolism (PE) in a retrospective case–control study including 300 consecutive hospitalized patients (233 clinically suspected of DVT and 67 clinically suspected of PE). Thromboembolic disease (TED) was defined by a well-established diagnostic strategy (Bosson et al. 1997; Ferretti et al. 1997) based on the following criteria: clinical probability, ultrasonography, ventilation/perfusion scintigraphy, helical tomodensitometry. TED was excluded in 146 patients who thus represented the controls in this study (49%) and ascertained in 154 patients (51%), divided into 107 DVT and 47 PE (with or without DVT). The value of plasma D-dimer, a degradation product of cross-linked fibrin, has been extensively evaluated over the last decade as a diagnostic tool to exclude either DVT or PE (Bounameaux et al. 1994). Non-quantitative latex assays were found to be unreliable (Elias et al. 1996) and ELISA methods have been validated as the reference method for plasma D-dimer measurement in suspected pulmonary embolism (Bounameaux et al. 1988; Bounameaux et al. 1991; Perrier et al. 1997) and deep venous thrombosis (Bounameaux et al. 1994; Janssen et al. 1997). However, ELISAs do not meet the needs of emergency diagnosis. D-dimers assays were performed with commercially available tests from Diagnostica Stago (Asnières, France). This is a new automated rapid quantitative latex assay, STA-Liatest D-Di, which allows measurement up to a concentration of 20 μg/ml (samples with a concentration > 4 μg/ml are automatically diluted 1/5) in less than 10 min. It was compared with ELISA Asserachrom D-Di test as the reference method. In the ELISA method, the plasma D-dimer to be measured is captured by a mouse monoclonal anti-human D-dimer antibody and sandwiched with a rabbit anti-fragment D antibody, coupled with peroxidase: bound peroxidase is revealed by its activity on the substrate o-phenylenediamine. In the STA-Liatest D-Di assay, latex microparticles are coated with two mouse monoclonal antibodies anti-human D-dimers: in the presence of plasma D-dimers, these latex particles aggregate and an increase in optical density is measured, proportional to the quantity of antigen fixed on latex particles. Plasma was stored at 37°C for 10 min before testing. The same aliquot was used simultaneously for both assays: one 96-well ELISA kit allowed 40 plasma samples to be analysed and the assays were batched accordingly. The results of vascular testing were not known when performing the laboratory assays. A positive threshold for D-dimers of 0.5 μg/ml was chosen as this had been previously validated for the ELISA (PIOPED Investigators 1990; Bounameaux et al. 1994). ELISA and STA-Liatest performances were determined by their sensitivity and specificity: the negative predictive value of these tests could not be evaluated as this was a case–control study. A contingency table and the Kappa concordance coefficient (Fermanian 1984) were used to compare STA-Liatest and ELISA performances. STA-Liatest D-dimer assay showed a sensitivity of 97.4% (95% confidence interval (CI) 94.9–99.9) and a specificity of 18.5% (95% CI 12.5–24.5); sensitivity was 95.8% for PE with a specificity of 40% while for DVT sensitivity was 98.1% with a specificity of 15%. True negatives with STA-Liatest were 18.5% while false negatives were 2.6% (95% CI 0.1–5.1) representing four cases: one distal DVT, one isolated proximal DVT and two DVT with pulmonary embolism. The same false negatives were found by ELISA testing but an additional patient with a distal DVT had 0.35 μg/ml plasma D-dimers raising the level of false negatives to 3.2% with ELISA. This patient had two distinct thrombi locations (tibial and fibular). In addition to this case, six patients without TED displayed discordant values between ELISA and STA-Liatest D-Di (Table 1). The number of patients with D-dimer STA-Liatest values above 5 μg/ml was 63, of whom 57 (90.5%) were cases, suggesting that D-dimer values above 5 μg/ml would allow these patients to be considered as having a very high probability of TED, whatever their initial clinical probability. The works shows the STA Liatest D-Di assay to be concordant with ELISA (Kappa concordance coefficient: 87%) and that the STA-Liatest D-Di assay is a reliable diagnostic investigation in DVT and PE. It is well-suited to emergency needs as it can be performed as an individual test in 10 minutes.