Background and objectives: Disease-related disability in aquaporin-4 antibody-positive neuromyelitis optica spectrum disorder (AQP4-NMOSD) is solely attributed to clinical attacks. However, few studies have assessed the relationship between attacks and residual disability in NMOSD. Thus, we aimed to quantify the contribution of clinical attacks to the residual disability in patients with AQP4-NMOSD.
Methods: This retrospective observational single-center study enrolled patients from the Oxford National NMO Service, with the inclusion criteria as (1) AQP4-NMOSD diagnosis and (2) availability of at least 1 disability score (Expanded Disability Status Scale [EDSS] or logarithm of the minimum angle of resolution [LogMAR] score) recorded ≥6 months after attack (defined as residual disability). The outcome measures were EDSS and LogMAR scores. Univariable and multivariable linear mixed-effect models were used to quantify the effect of clinical relapses on the outcomes.
Results: A total of 165 patients with AQP4-NMOSD (median onset age, 43 years, range 2-84; women, 140 [84.8%]; White European patients, 92 [55.8%]; African or African British patients, 40 [24.2%]; Asian or Asian British patients, 20 [12.1%]; multiracial or unknown racial patients, 13 [7.9%]) were included, with the median time of disability measurement since the last attack being 32 months (range 6-197). The mean increase in the EDSS score per relapse was 0.304 (95% CI 0.074-0.553, p < 0.001), with individual relapse phenotypes showing different effects: the transverse myelitis (TM) + optic neuritis (ON) phenotype contributed most, with an increase of 1.290 (95% CI 0.233-2.207, p = 0.017) per relapse, followed by brain plus other phenotypes (β = 0.782, 95% CI 0.029-1.03, p < 0.001) and isolated TM (β = 0.295, 95% CI 0.074-0.549, p < 0.001), while neither brain nor optic nerve relapse alone was associated with a residual change in the EDSS score. Older onset age was correlated with more severe motor disability where this mainly occurred early in the disease course while younger patients exhibited mild initial disability that worsened more significantly with relapses. Each ON attack led to a mean increase of 0.464 (95% CI 0.199-0.741, p < 0.001) in the LogMAR score. Race, sex, and timing of acute treatment did not significantly affect these disability outcomes (EDSS and LogMAR scores).
Discussion: The quantitative contribution of relapse to the residual disability in patients with AQP4-NMOSD varies across phenotypes, and this relapse-related disability progression may also vary by the onset age. Although this retrospective single-center study may need validation in other data sets, these findings may help predict disability and provide a modeling tool for longer term disability in the cost-effective analysis of newer interventions.