Performance Characteristics of Commercial Immunoassays for the Detection of IgG and IgM Antibodies to β2 Glycoprotein I and an Initial Assessment of Newly Developed Reference Materials for Assay Calibration. Issue 6 (5th June 2016)

• Record Type: Journal Article
• Authors: Willis, Rohan; Pierangeli, Silvia S.; Jaskowski, Troy D.; Malmberg, Elisabeth; Guerra, Marta; Salmon, Jane E.; Petri, Michelle; Branch, D. Ware; Tebo, Anne E.
• Is Part Of: American journal of clinical pathology. Volume 145:Issue 6(2016)
• Page Start: 796
• Page End: 805
• Publication Date: 2016-06-05
• Subjects: Antiphospholipid syndrome -- Immunoassays -- β2 glycoprotein I -- Antibodies -- Standardization -- Harmonization; Diagnosis, Laboratory -- Periodicals; Pathology -- Periodicals; 616.07
• DOI: 10.1093/ajcp/aqw065
• Languages: English
• ISSNs: 0002-9173

Abstract

Abstract : Objectives: To investigate the performance characteristics and impact of newly developed reference calibrators on the commutability between anti–β2 glycoprotein I (anti–β2 GPI) immunoassays in antiphospholipid syndrome (APS) and/or systemic lupus erythematosus (SLE). Methods: Immunoglobulin G (IgG) and immunoglobulin M (IgM) anti–β2 GPI immunoassays from four manufacturers were evaluated. Serum samples from 269 patients (APS only, n = 31; SLE and APS, n = 83; SLE only, n = 129; pregnancy-related clinical manifestations without APS, n = 26) and 162 women with histories of successful pregnancies were tested. Results were expressed in kit-specific arbitrary units and in the calibrator reference units (RUs) based on 99th percentile cutoff values. Diagnostic accuracies, correlation between kits, and specific clinical manifestations in APS were investigated. Results: The sensitivities of the assays ranged from 15.8% to 27.2% (IgG) and 12.3% to 15.8% (IgM) while specificities ranged from 79.4% to 86.5% (IgG) and 80.6% to 84.5% (IgM). There was moderate to almost perfect interassay reliability (Cohen κ, 0.69-0.98), and Spearman correlation coefficients were generally improved when results of the IgG determinations were expressed in RUs. Conclusions: Although qualitative agreements between immunoassays for both antibody isotypes are acceptable, correlations with APS clinical manifestations were kit dependent. Only the use of IgG reference material improved quantitative correlations between assays.