The real-world impact of biologics for NMOSD: A retrospective single-center study compared with natural course and conventional treatments in Japanese

Abstract

BackgroundNeuromyelitis optica spectrum disorder (NMOSD), a central nervous system inflammatory disease associated with aquaporin-4 immunoglobulin G (AQP4-IgG), is conventionally treated with oral steroids and immunosuppressants (IS) in Japan. Several biologics which show great efficacy in the clinical trials have been developed recently. However, studies on their efficacy, especially those comparing them with conventional treatments in real-world situations are lacking. ObjectiveHere, we conducted a single-center retrospective cohort study in Japan comparing the efficacy of biologics, over conventional drugs, in treating AQP4-IgG-positive NMOSD. MethodsWe extracted the medical history of patients with AQP4-IgG-positive NMOSD who visited the Tohoku University Hospital between 2000 and 2023, from the hospital patient database. All patients were diagnosed according to the international consensus diagnostic criteria for NMOSD 2015. We then classified the disease duration of each patient into four periods based on their prescription history as: no-treatment, prednisolone monotherapy (PSL-mono), immunosuppressants (IS) treatment, and biologics (Bio) treatment. Subsequently, the efficacy of Bio treatment, over the conventional treatment, in alleviating AQP4-IgG-positive NMOSD was estimated. We used univariate Poisson regression analysis to compare the annualized relapse rate (ARR), log-rank test for the first attack, and the hazard ratios (HR)—calculated using multivariate Andersen-Gill model for recurrent attacks—of the Bio and conventional treatment period groups. The safety of each treatment period group was assessed by comparing infection and mortality rates. ResultsA total of 109 patients (92 % females) met the eligibility criteria of the study. We could extract a total of 1,283 patient years with 289 NMOSD attacks from their medical history data. The mean ARR of no-treatment group was 0.60. Most of the Bio group initially received combined treatments with PSL or IS. The mean ARR of the Bio group was 0.01 [95 % confidence interval (CI): 0.002 to 0.08], which was significantly lower than the PSL-mono group (0.16, 95 % CI: 0.13 to 0.19, p = 0.03) and the IS group (0.17, 95 % CI: 0.13 to 0.22, p = 0.02). In the survival analysis, the Bio group showed a significantly prolonged attack-free period than the other groups, suggesting its potential in reducing 79 % of relapses in the no-treatment group, 33 % in the PSL-mono group, and 31 % in the IS group during the two years. The multivariate analysis using Andersen-Gill model showed that the Bio group had significantly lower HR (log HR −2.75, 95 % CI: −4.71 to −0.8, p = 0.006), relative to the PSL-mono group. Importantly, the patients needed significantly lower PSL (median 5 mg/day) during the Bio group treatment period than in the PSL-mono group treatment period (median 10 mg/day). Further, the concomitant use of IS could be stopped safely in all patients who were on Bio treatment. All treatment period groups showed similar safety profiles. ConclusionIn patients with AQP4-IgG-positive NMOSD, biologics demonstrated more efficacy than conventional PSL and/or IS treatment, without increasing infection and mortality rates.