To (1) identify clinical and treatment factors associated with AQP4-IgG seroreversion and (2) evaluate whether achieving seronegativity influences subsequent disease activity.

We conducted a retrospective cohort study of adults with AQP4-IgG–positive NMOSD, as defined by the 2015 International Consensus Diagnostic Criteria, treated at Siriraj Hospital between May 2013 and July 2025, who had ≥2 serial AQP4-IgG measurements. Serostatus was determined using a fixed cell-based assay (positivity defined as titer ≥1:10). Primary outcomes were predictors of seroreversion and the association between seronegativity and relapse risk. Predictors were assessed with a multivariable Cox model; significance set at P<0.05.

Eighty-three patients were included (89.2% female); 35 (42.2%) achieved seroreversion. Median baseline age was similar between those who seroreverted and those who remained seropositive. In multivariable analysis, treatment with a high-efficacy therapy for ≥6 months was the strongest predictor of seroreversion (hazard ratio [HR] 4.51, 95% CI 2.07–9.84; P<0.001). A higher disability score within six months before serostatus testing was associated with a lower likelihood of seroreversion (HR 0.76, 95% CI 0.62–0.93; P=0.005), as was a greater number of attacks before treatment initiation (HR 0.76, 95% CI 0.67–0.87; P<0.001). Notably, no patient experienced a clinical relapse after achieving seronegativity.

High-efficacy therapy, lower disability, and fewer prior attacks independently predicted AQP4-IgG seroreversion. These findings should be interpreted cautiously given two limitations: the retrospective design limits causal inference, and reliance on a fixed cell-based assay may affect serological validity and comparability. Larger, prospective studies are needed to confirm these associations and to evaluate AQP4-IgG seroreversion as a useful prognostic biomarker in NMOSD.