BACKGROUND: The use of automated immunometric methods for the detection of anti-thyroid peroxidase antibodies (TPOAb), the main serological marker of autoimmune thyroid diseases (AITD), has expanded in recent years. However, it is not known whether these new automated platforms have improved the diagnostic performance of TPOAb assays. The aim of this study was to evaluate the potential improvement of the inter-method agreement of current automated third generation systems, 12 years after a previous study, which had assessed the analytical variability between semi-automated second generation methods of TPOAb detection. METHODS: Eight pools of sera from patients with chronic lymphocytic thyroiditis, exhibiting different TPOAb concentrations, were collected from routine laboratory diagnostics and distributed to seven companies throughout Italy. All automated third generation methods were calibrated against the Medical Research Council (MRC) reference preparation 66/387. RESULTS: The overall mean variability (CV) was 93.6% when results were expressed in part as arbitrary Units (U/mL) and in part as International Units (IU/mL). The conversion of all values in IU/mL resulted in a significant decrease of CV (49.8%). The CV expressed as COM (cut-off concentration multiples) was 64.0%. Agreement of qualitative results was 95.3% with a pronounced difference in the threshold values proposed by manufacturers (range 3.2-35.0 IU/mL). CONCLUSIONS: These findings confirm the improvement of harmonisation between different methods of automated third generation TPOAb assays. Nevertheless, further efforts should be made in the definition of the positive cut-off concentration to avoid misclassification of AITD patients as well as in a new international reference preparation and in the autoantigen purification modality.
BACKGROUND: The use of automated immunometric methods for the detection of anti-thyroid peroxidase antibodies (TPOAb), the main serological marker of autoimmune thyroid diseases (AITD), has expanded in recent years. However, it is not known whether these new automated platforms have improved the diagnostic performance of TPOAb assays. The aim of this study was to evaluate the potential improvement of the inter-method agreement of current automated third generation systems, 12 years after a previous study, which had assessed the analytical variability between semi-automated second generation methods of TPOAb detection. METHODS: Eight pools of sera from patients with chronic lymphocytic thyroiditis, exhibiting different TPOAb concentrations, were collected from routine laboratory diagnostics and distributed to seven companies throughout Italy. All automated third generation methods were calibrated against the Medical Research Council (MRC) reference preparation 66/387. RESULTS: The overall mean variability (CV) was 93.6% when results were expressed in part as arbitrary Units (U/mL) and in part as International Units (IU/mL). The conversion of all values in IU/mL resulted in a significant decrease of CV (49.8%). The CV expressed as COM (cut-off concentration multiples) was 64.0%. Agreement of qualitative results was 95.3% with a pronounced difference in the threshold values proposed by manufacturers (range 3.2-35.0 IU/mL). CONCLUSIONS: These findings confirm the improvement of harmonisation between different methods of automated third generation TPOAb assays. Nevertheless, further efforts should be made in the definition of the positive cut-off concentration to avoid misclassification of AITD patients as well as in a new international reference preparation and in the autoantigen purification modality.
Authors: F D'Aurizio; P Metus; A Ferrari; B Caruso; R Castello; D Villalta; A Steffan; K Gaspardo; F Pesente; N Bizzaro; E Tonutti; S Valverde; C Cosma; M Plebani; R Tozzoli Journal: Auto Immun Highlights Date: 2017-06-19