Background Phosphatidylserine-dependent antiprothrombin antibodies (aPS/PT) are strongly correlated with lupus anticoagulant. A Task Force of scientists at the International Congress on Antiphospholipid Antibodies (APLA 2010) accepted that aPS/PT would potentially contribute for a better identification of antiphospholipid syndrome (APS). Accordingly,we carried out an Initial and a Validation retrospective, cross-sectional multi-centre study on aPS/PT.

Objectives To ascertain the value of aPS/PT for APS diagnosis.

Methods We conducted an initial study involving 8 centres from 7 countries.Demographics and clinical/laboratory data were retrospectively collected.A sample from the subjects was prepared at each institution.Specimens were blinded and determination of IgG aPS/PT performed at Inova Diagnostics, Inc.USA (Inova) using ELISA kits from two manufacturers:Medical and Biological Laboratories Co Ltd, Japan (MBL) and Inova.After completing the initial study, a Validation study with a new sample of subjects (5 different centres from 5 countries) was carried out using the same methodology.

Results The Initial study comprised 247 subjects.Statistically significant correlation in the IgG aPS/PT values was obtained with both ELISA kits (r=0.827, p<0.001). amples with concordant IgG aPS/PT results in both ELISAs (n=204) were subsequently analyzed (99 APS,58 non-APS and 47 healthy). IgG aPS/PT was more prevalent in APS patients (51%) than in those without (9%), OR: 10.8 [95%CI 4-29], p<0.0001. Sensitivity, specificity, positive (LR+) and negative (LR-) likelihood ratio of IgG aPS/PT for APS diagnosis were 51%, 94%, 8.9 and 0.5 respectively.The diagnostic accuracy of IgG aPS/PT for APS diagnosis was assessed by ROC curves showing AUC values of 0.782 and 0.777 for MBL and Inova assays, respectively. In the Validation study (n=214) a statistically significant correlation was found in the IgG aPS/PT titers with both ELISAs (r=0.827, p<0.001). IgG aPS/PT concordant samples were analyzed (n=182;76 APS,57 non-APS and 49 healthy). IgG aPS/PT was more frequent in APS patients (47%) than in those without (12%), OR: 6.4 [95%CI 2.6-16], p<0.0001. Sensitivity, specificity, LR+ and LR- for APS diagnosis were 47%, 93%, 7.2 and 0.6, respectively.AUC values were 0.757 and 0.759 for MBL and Inova assays, respectively.

Conclusions The performance of IgG aPS/PT using Inova and MBL ELISA kits is reliable. IgG aPS/PT detection is an easily performed laboratory parameter that may help in the diagnosis of APS.

Acknowledgements The authors want to acknowledge the contributions of the late Prof. Silvia Pierangeli, University of Texas Medical Branch, Galveston, TX, USA

Disclosure of Interest O. Amengual: None declared, R. Forastiero: None declared, M. Sugiura-Ogasawara: None declared, K. Otomo: None declared, O. Kenji: None declared, C. Favas: None declared, J. Delgado Alves: None declared, P. Žigon: None declared, A. Ambrožič: None declared, M. Tomšič: None declared, I. Ruiz Arruza: None declared, G. Ruiz Irastorza: None declared, M. Bertolaccini: None declared, G. Norman Employee of: Inova Diagnostics, Inc, San Diego, CA USA, Z. Shums: None declared, A. Jiro Employee of: Medical and Biological Laboratories, Co. Ltd, A. Murashima: None declared, A. Tebo: None declared, M. Gerosa: None declared, P. Meroni: None declared, I. Rodriguez-Pintό: None declared, R. Cervera: None declared, J. Swadzba: None declared, J. Musial: None declared, T. Atsumi Grant/research support from: Chugai Pharmaceutical Co., Ltd., Eisai Co., Ltd., Bristol-Myers Squibb Co., Astellas Pharma Inc., Daiichi Sankyo Co., Ltd. and Mitsubishi Tanabe Pharma Co., Speakers bureau: Astellas Pharma Inc. and Mitsubishi Tanabe Pharma Co.