A multicenter comparison of MOG-IgG cell-based assays

Patrick J. Waters, Lars Komorowski, Mark Woodhall, Sabine Lederer, Masoud Majed, Jim Fryer, John Mills, Eoin P. Flanagan, Sarosh R. Irani, Amy C. Kunchok, Andrew McKeon, Sean J. Pittock

Research output: Contribution to journalArticlepeer-review

58 Scopus citations


sTo compares 3 different myelin oligodendrocyte glycoprotein-immunoglobulin G (IgG) cell-based assays (CBAs) from 3 international centers.MethodsSerum samples from 394 patients were as follows: acute disseminated encephalomyelitis (28), seronegative neuromyelitis optica (27), optic neuritis (21 single, 2 relapsing), and longitudinally extensive (10 single, 3 recurrent). The control samples were from patients with multiple sclerosis (244), hypergammaglobulinemia (42), and other (17). Seropositivity was determined by visual observation on a fluorescence microscope (Euroimmun fixed CBA, Oxford live cell CBA) or flow cytometry (Mayo live cell fluorescence-activated cell sorting assay).ResultsOf 25 samples positive by any methodology, 21 were concordant on all 3 assays, 2 were positive at Oxford and Euroimmun, and 2 were positive only at Oxford. Euroimmun, Mayo, and Oxford results were as follows: clinical specificity 98.1%, 99.6%, and 100%; positive predictive values (PPVs) 82.1%, 95.5%, and 100%; and negative predictive values 79.0%, 78.8%, and 79.8%. Of 5 false-positives, 1 was positive at both Euroimmun and Mayo and 4 were positive at Euroimmun alone.ConclusionsOverall, a high degree of agreement was observed across 3 different MOG-IgG CBAs. Both live cell-based methodologies had superior PPVs to the fixed cell assays, indicating that positive results in these assays are more reliable indicators of MOG autoimmune spectrum disorders.

Original languageEnglish (US)
Pages (from-to)E1250-E1255
Issue number11
StatePublished - Mar 12 2019

ASJC Scopus subject areas

  • Clinical Neurology


Dive into the research topics of 'A multicenter comparison of MOG-IgG cell-based assays'. Together they form a unique fingerprint.

Cite this