Abstract
Cell-based assays (CBA) have increased the sensitivity of the neuromyelitis optica (NMO)-IgG/aquaporin-4-antibody detection compared to classical tissue-based indirect assays. We describe the sensitivity of an optimized immunohistochemistry (IHC-o) to detect NMO-IgG/aquaporin-4-antibody in comparison with that of two CBA: an in-house (CBA-ih) and a commercial (CBA-c) assay (Euroimmun, Germany). Coded serum from 103 patients with definite NMO and 122 inflammatory controls were studied by IHC-o, CBA-ih, and CBA-c. IHC-o used the same protocol described to detect antibodies against cell surface antigens. CBA-ih used live cells transfected with the aquaporin-4-M23-isoform. The sensitivity of the IHC-o was 74.8% (95% confidence interval [CI] 65-83) and was similar to that of the CBA-ih 75.7% (95% CI 66-84) and the CBA-c 73.8% (95% CI 64-82). The specificity of the three assays was 100% (95% CI 97-100). Interassay concordance was high, 100 of 103 samples were coincident in all techniques. The optimized immunohistochemistry proves to be as sensitive and specific as the cell-based assays. This assay extends the available tools for NMO-IgG/aquaporin-4-antibody detection.