Plasma exchange (PE) and immunoadsorption (IA) are recognized as effective ways to treat attacks in AQP4 antibody-positive NMOSD, but high-quality evidence was lacking. To evaluate the efficacy and safety of PE/IA plus intravenous methylprednisolone (IVMP) in NMOSD attacks using propensity scores to match IVMP as control. Patients were from a prospective observational cohort study. Stratification and interval propensity score matching (PSM) were used to reduce selection bias by matching baseline characteristics (gender, age, time to IVMP, EDSS at attack) between PE/IA + IVMP and IVMP group (in a ratio of 1:2). The primary endpoint of efficacy was EDSS change at 6 months. Adverse events and changes in laboratory tests were recorded. Four hundred and eleven attacks of 336 patients were screened for PSM, and 90 attacks (30 PE/IA + IVMP and 60 IVMP) were included in the analysis. There were significant differences in EDSS [6.25 vs. 6.75; IQR (4.50-8.38 vs. 5.00-8.00), p = 0.671] and visual acuity [median logMAR = 0.35 vs. 1.00; IQR (0.30-0.84 vs. 0.95-1.96), p = 0.002] change between two groups at 6 months. PE/IA + IVMP treatment demonstrated predictive capacity for good recovery as indicated by an area under the curve (AUC) of 0.726. Fibrinogen reduction was found during PE/IA + IVMP treatment [n = 15 (50.00%)], but no severe adverse events led to apheresis treatment discontinuation. After PSM analysis, IVMP+PE/IA in acute attack of NMOSD achieved better and continuous improvement in neurological function within 6 months compared with IVMP alone. PE/IA treatment showed a good safety profile.
Read full abstract