The accuracy of flow cytometric cell-based assay to detect anti-myelin oligodendrocyte glycoprotein (MOG) antibodies determining the optimal method for positivity judgement.

To illustrate the accuracy of the fluorescence-activated cell sorting cell-based assay (FACS-CBA) and to detect anti-myelin oligodendrocyte glycoprotein (MOG) antibodies and ascertain the optimal method for positivity judgement, referencing the findings of microscopic CBA. We tested serum anti-MOG antibodies in 57 patients with central nervous system inflammatory disorders (CIDs), 30 healthy controls (HCs), and 63 disease controls (DCs) by FACS-CBA. To assess the diagnostic performance of 2 positive judgement methods for FACS-CBA, we evaluated the ratio of positive cells (RPC) and median fluorescence intensity (MFIratio); samples from 57 CIDs and 3 antiaquaporin-4 antibody-positive patients whose anti-MOG antibody levels were relatively high but negative by FACS-CBA were tested by microscopic CBA. Blinded to the RPC and MFIratio results, we classified the acquired dot plot into 3 patterns-"upright," "broadband," and "oblique"-as pattern analysis. The sample with the highest RPC in CIDs was subjected to serial dilution analysis. Finally, we analyzed the clinical and laboratory data of anti-MOG antibody-positive patients in the acute phase. Referencing results by microscopic CBA and receiver-operating characteristic curve analysis, the area under the curve, sensitivity, specificity, and cutoff value were 0.952, 92%, 94%, and 1.52 for RPC and 0.931, 79%, 94%, and 6.39 for MFIratio, respectively, suggesting the optimality of RPC for positive judgement. Titers by microscopic CBA analysis significantly correlated with RPC (P = .031). In the validation study, the positive rate of RPC for anti-MOG antibodies was 42.1% in CIDs, but 0% in HCs and DCs (both P < .001). In the pattern analysis, all anti-MOG antibody-positive patients but none of the HCs and DCs exhibited the "oblique" pattern. Serial dilution curve analysis fit a quaternary polymodal. FACS-CBA using RPC analysis for anti-MOG antibodies displayed relatively higher specificity, sensitivity, and semiquantitative property, indicating it could become another acceptable test to detect anti-MOG antibodies.