BACKGROUND: The N-MOmentum trial, a double-blind, randomized, placebo-controlled, phase 2/3 study of inebilizumab in neuromyelitis optica spectrum disorder (NMOSD), enrolled participants who were aquaporin-4-immunoglobulin G (AQP4-IgG)-seropositive (AQP4+) or -seronegative (AQP4-). This article reports AQP4- participant outcomes. METHODS: AQP4-IgG serostatus was determined for all screened participants by a central laboratory, using a validated, fluorescence-observation cell-binding assay. Medical histories and screening data for AQP4- participants were assessed independently by an eligibility committee of three clinical experts during screening. Diagnosis of NMOSD was confirmed by majority decision using the 2006 neuromyelitis optica criteria. Myelin oligodendrocyte glycoprotein-immunoglobulin G (MOG-IgG) serology (using a clinically validated, flow cytometry assay) and annualized attack rates (AARs) were evaluated post hoc. Efficacy outcomes were assessed by comparing pre-study and on-study AARs in treated participants. RESULTS: Only 18/50 AQP4- screened participants (36%) were initially considered eligible for randomization; 16 were randomized and received full treatment, 4 to placebo (1 MOG-IgG-seropositive [MOG+]) and 12 to inebilizumab (6 MOG+). The most common reason for failure to pass screening among prospective AQP4- participants was failure to fulfill the 2006 NMO MRI criteria. In inebilizumab-treated AQP4- participants, on-study AARs (95% confidence interval [CI]) calculated from treatment initiation (whether from randomization or when received at the start of the open-label period) to the end of study were lower than pre-study rates: for all AQP4- participants (n = 16), mean (95% CI) AAR was 0.048 (0.02-0.15) versus 1.70 (0.74-2.66), respectively. For the subset of AQP4-/MOG+ participants (n = 7), AAR was 0.043 (0.006-0.302) after treatment versus 1.93 (1.10-3.14) before the study. For the subset of AQP4-/MOG- participants (n = 9), post-treatment AAR was 0.051 (0.013-0.204) versus 1.60 (1.02-2.38). Three attacks occurred during the randomized controlled period in the AQP4- inebilizumab group and were of mild severity; no attacks occurred in the AQP4- placebo group. The low number of participants receiving placebo (n = 4) confounds direct comparison with the inebilizumab group. No attacks were seen in any AQP4- participant after the second infusion of inebilizumab. Inebilizumab was generally well tolerated by AQP4- participants and the adverse event profile observed was similar to that of AQP4+ participants. CONCLUSION: The high rate of rejection of AQP4- participants from enrollment into the study highlights the challenges of implementing the diagnostic criteria of AQP4- NMOSD. An apparent reduction of AAR in participants with AQP4- NMOSD who received inebilizumab warrants further investigation.
