Résumé
Background: Autoimmune nodopathy (AN) is a disabling peripheral nerve disorder mediated by four pathogenic autoantibodies targeting the node of Ranvier: anti-neurofascin-155 (Nfasc155), anti-Nfasc155 and Nfasc186 (PanNfasc), anti-contactin-1, and anti-contactin-associated protein 1 autoantibodies. Several autoantibody detection assays exist; however, which assay yields the best performance remains unclear. We evaluated the performance of five diagnostic assays individually and in combinations to identify a gold standard assay for diagnosing AN and improve diagnostic accuracy.Methods: Sera from 290 individuals from a European cohort including healthy controls and patients with Guillain–Barré syndrome, chronic inflammatory demyelinating polyradiculoneuropathy, monoclonal gammopathy, Charcot–Marie–Tooth disease, and AN were tested using five diagnostic assays for AN: ImmunoDOT (D-tek, Mons, Belgium), ELISA, an in-house cell-based assay (CBA), a commercial CBA, and peripheral nerve immunohistochemistry (IHC). The sensitivity (Se) and specificity (Sp) of the individual assays and their combinations were estimated using Bayesian latent class analysis. Results: The overall Se was 84% (95% confidence interval [CI], 79–88) for ImmunoDOT, 86% (81–90) for ELISA, 96% (93–98) for the in-house CBA, 92% (89–96) for the commercial CBA, and 91% (87–94) for IHC. For all assays, the overall Sp was ≥ 97%. Among the tested assays, in-house CBA showed the best performance, with the highest Se and Sp. The assay combinations demonstrated excellent analytical performance (Se ≥ 91% and Sp ≥ 98%). Inter-laboratory validation revealed excellent repeatability, with a Gwet’s agreement coefficient of > 0.94 for most assays.Conclusion: While no assay emerged as a gold standard for AN diagnosis, using a combination of CBA and ImmunoDOT showed enhanced diagnostic accuracy. We recommend to make a positive diagnostic when at least 1 positive result is obtained with 2 different assays.