King's College London

Research portal

The comparison of real world and core laboratory antiphospholipid antibody ELISA results from antiphospholipid syndrome alliance for clinical trials & international networking (APS ACTION) clinical database and repository analysis

Research output: Contribution to journalArticle


Original languageEnglish
Pages (from-to)32-36
Number of pages5
JournalThrombosis Research
Publication statusPublished - 1 Mar 2019

King's Authors


Background: The APS ACTION International Clinical Database and Repository includes a secure web-based data capture system storing patient information including demographics, antiphospholipid antibodies (aPL)-related medical history, and aPL tests. Despite efforts at harmonization, inter-assay variability remains a problem in aPL testing. As a clinical repository open to researchers, ensuring comparability between assays and consistency in results between APS ACTION laboratories is essential to the validity of studies emerging from this network. Objective: To assess the level of agreement between an aPL-registry inclusion and core laboratory (core lab) anticardiolipin antibody (aCL) and anti-β 2 -glycoprotein-I antibody (aβ 2 GPI) ELISA testing results. Methods: Patients are recruited from 25 international centers based on positive aPL tests at inclusion. All samples are retested at the corresponding national APS ACTION core lab to confirm aPL positivity based on standard validated protocols. We analysed the categorical agreement, degree of linear association, and correlation between inclusion (local laboratory) and core lab aPL tests. Samples were included in this study only if results of aPL testing with ELISA at baseline were available. Results: 497 registry samples underwent confirmatory aPL tests. Categorical agreement between the inclusion and core lab values, as expressed by Cohen's kappa coefficients, ranged between 0.61 and 0.80 (as substantial agreement). The correlation between quantitative results in the aCL and aβ 2 GPI was better for IgM and IgA compared to IgG (Spearman rho 0.789 and 0.666 vs. 0.600 for aCL and rho 0.892 and 0.744 vs. 0.432 for aβ 2 GPI). Conclusions: The results of inclusion for aCL and aβ 2 GPI tests used for recruitment into the registry were in agreement to the results obtained by the APS ACTION core laboratories; aCL and aβ 2 GPI results showed very good categorical agreement. This agreement increased when considering high titer (>40 units) samples. APS ACTION is a reliable and useful research resource for APS.

View graph of relations

© 2018 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454