Affiliation: | aRheumatology and Clinical Immunology, Ospedale Civile di Brescia, Piazza Spedali Civili 1, 25125 Brescia, Italy bGeneva University Hospital, Switzerland cHopital de la Conception, Marseille, France dIstituto Auxologico, Università di Milano, Italy eDepartment Int. Med., Hopital de la Pitiè, Paris, France |
Abstract: | Antiphospholipid ELISAs are part of the Antiphospholipid Antibodies Syndrome classification criteria, having the same diagnostic value as lupus anticoagulant. However, sometimes their results appear scarcely meaningful especially when wide metanalyses studies are performed, probably because of their well-known inter-laboratory variability. The application of a common protocol was shown to improve the test reproducibility, but this observation did not have any influence on the routine performances.After discussion among experts at the European level, we identified four conditions named “minimal requirements” considered useful to decrease the inter-laboratory variability: - (1) to run the samples in duplicate;
(2) to determine the cut off level in each laboratory analysing at least 50 samples from normal subjects, possibly age- and sex-matched with the patient population usually attending the Centre; (3) to calculate the cut-off level in percentiles; (4) to use stable external controls in the test.
A collaborative study involving 36 European centres proved that the use of monoclonal anti-beta2 glycoprotein I antibodies, HCAL (IgG) and EY2C9 (IgM) as standards, can help to reduce the inter-laboratory coefficient of variation both in anticardiolipin (aCL) and anti-β2GPI (anti-beta2 glycoprotein I) ELISA. Therefore, we propose HCAL and EY2C9 as external controls, but other monoclonal or polyclonal preparations may be considered. During an interactive workshop held last May in Italy, 16 companies producing these tests agreed to consider the introduction of the “requirements” in their products. We suggest to adopt these “requirements” particularly in clinical studies, in order to compare more easily the literature data. |