Safety and Effectiveness of Inebilizumab in Patients with Neuromyelitis Optica Spectrum Disorder: Interim Analysis of a Post-marketing Surveillance in Japan

Abstract

Introduction

Inebilizumab is used in Japan for relapse prevention in aquaporin-4 (AQP4) antibody-positive neuromyelitis optica spectrum disorder (NMOSD). Nationwide post-marketing surveillance (PMS) is underway to evaluate the real-world safety and effectiveness of inebilizumab in Japanese patients with NMOSD.

Methods

This PMS interim analysis assessed safety and effectiveness of inebilizumab in Japanese patients with NMOSD initiating treatment between June 2021 and November 2023, using data through June 2025.

Results

The safety analysis set included 228 patients, and the effectiveness analysis set included 224 patients after excluding three patients from N-MOmentum (NCT02200770) and one without confirmed anti-AQP4 antibody positivity. Twenty-one (safety analysis set) and 20 (effectiveness analysis set) patients discontinued treatment. In the safety analysis set, 194 patients (85.1%) were female, and the mean ± standard deviation (SD) age was 54.2 ± 13.1 years. Adverse drug reactions (ADRs) occurred in 74 patients (32.5%); 31 (13.6%) were serious. Common ADRs were infections (14.5%, n = 33); serious infections (5.7%, n = 13) included COVID-19 (1.3%, n = 3), COVID-19 pneumonia (0.9%, n = 2), and urinary tract infection (0.9%, n = 2). Two fatal ADRs (interstitial lung disease, sudden cardiac death) occurred, for which a causal relationship with inebilizumab could not be excluded. In the effectiveness analysis set, the annualized relapse rate (95% CI) decreased from 0.40 (0.33–0.47) per person-year during the 2 years before inebilizumab initiation to 0.10 (0.07–0.14) per person-year after inebilizumab initiation. From baseline to 1 year, the proportion of patients receiving neither concomitant oral glucocorticoids nor immunosuppressants increased from 5.4% to 33.0%, and the mean ± SD oral glucocorticoid dose decreased from 11.7 ± 7.9 mg/day to 4.3 ± 4.7 mg/day.

Conclusion

In this interim analysis, the safety profile of inebilizumab was consistent with prior evidence and no new safety concerns were identified. Inebilizumab reduced relapse rates and decreased concomitant use of oral glucocorticoids and immunosuppressants in patients with NMOSD.

UMIN Clinical Trials Registry

UMIN000044431.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s40120-026-01008-9.

Key Summary Points

Why carry out this study?

Inebilizumab has been used in Japan since 2021 for relapse prevention of neuromyelitis optica spectrum disorder (NMOSD), based largely on the results of the N-MOmentum study; however, patients in clinical practice often have diverse characteristics and comorbidities that exclude them from clinical trials.

We report the results of this interim analysis from an ongoing all-case post-marketing surveillance (PMS) evaluating the real-world safety and effectiveness of inebilizumab in Japanese patients with NMOSD.

What was learned from the study?

Inebilizumab treatment was associated with a reduction in the annualized relapse rate (from 0.40 at baseline to 0.10 after inebilizumab initiation), stability in disability (mean ± SD Expanded Disability Status Scale [EDSS] scores 4.26 ± 2.25 at baseline vs 3.74 ± 2.29 at 1 year), and marked reductions in glucocorticoid exposure.

Adverse drug reactions occurred in 32.5% of patients, most of which (14.5%) were infections; the safety profile was consistent with prior evidence and no new safety concerns were identified.

Our findings highlight the utility of inebilizumab as an effective and well-tolerated treatment option across a broad, heterogeneous NMOSD population in real-world clinical practice.

Introduction

Neuromyelitis optica spectrum disorder (NMOSD) is a chronic, inflammatory, autoimmune disease of the central nervous system (CNS) characterized by recurrent and debilitating episodes of optic neuritis and transverse myelitis [1]. These relapses contribute to irreversible disability, loss of independence, and a marked reduction in quality of life [2]. Occasionally, the diencephalon, brainstem, or cerebral hemispheres may also be involved [1]. Immunoglobulin G (IgG) autoantibodies targeting the water channel protein aquaporin-4 (AQP4) expressed on astrocytes initiate complement activation and lead to perivascular immune infiltration and secondary demyelination. The prevalence of NMOSD varies across studies, ranging from approximately 0.7 to 13.0 cases per 100,000 people worldwide [3–5], but is consistently higher in East Asian individuals (5.4 cases per 100,000 people in Japan) compared with white individuals (1.0–1.5 cases per 100,000 people) [6–8]. Therapeutic strategies aimed at reducing relapse rates are crucial for improving long-term outcomes in patients with NMOSD.

Inebilizumab (UPLIZNA®) is a humanized IgG1 monoclonal antibody targeting CD19⁺ B cells to induce B cell depletion via antibody-dependent cellular cytotoxicity [9]. By targeting a wider array of B cells, including disease-relevant CD19⁺ plasmablasts and plasma cells not targeted by anti-CD20 therapies such as rituximab, inebilizumab may more effectively address B cell-related immune mechanisms [10]. Inebilizumab received its first approval for the management and prevention of relapses in patients with NMOSD in the USA on June 11, 2020, followed by approval in Japan on March 23, 2021 [11, 12]. The approved regimen for patients with NMOSD consists of two initial 300 mg intravenous infusions administered 2 weeks apart, followed by single 300 mg intravenous infusions every 6 months beginning 6 months after the first dose [11]. Premedication with a corticosteroid, antihistamine, and antipyretic is required prior to each inebilizumab infusion to reduce the frequency and severity of infusion-related reactions (IRRs).

In the pivotal international phase 2/3 N-MOmentum study, inebilizumab significantly reduced the risk of relapse compared with placebo in patients with NMOSD that are anti-AQP4 antibody positive (hazard ratio [HR] 0.27; 95% CI 0.15–0.50; P < 0.0001) [13]. Subgroup analyses in Asian patients demonstrated comparable efficacy and safety to the overall population [12]. However, N-MOmentum enrolled a pre-specified population, excluding individuals with severe disability (Expanded Disability Status Scale [EDSS] > 8.0), poorly controlled autoimmune conditions requiring immunosuppressive therapy, comorbidities requiring chronic high-dose corticosteroid therapy, or a history of malignancy [14]. Furthermore, oral corticosteroids (prednisone 20 mg/day or equivalent) were administered for only 2 weeks at the start of the study with a 1-week taper. In real-world clinical practice, however, inebilizumab is prescribed to a broad spectrum of patients with heterogeneous clinical profiles and may be used in combination with oral glucocorticoids and immunosuppressive agents for relapse prevention. Given the limited number of Japanese patients in the N-MOmentum study, post-marketing surveillance (PMS) encompassing all cases treated with inebilizumab has been conducted in accordance with the risk management plan (RMP). This PMS was designed to systematically collect data on patient characteristics and the safety and effectiveness of inebilizumab in real-world clinical practice, with the aim of supporting its appropriate use. Here, we present the interim findings from this PMS evaluating the real-world safety and effectiveness of inebilizumab in patients with NMOSD, including those using oral glucocorticoids.

Methods

Study Design

This was an ongoing, observational, all-case PMS investigating the safety and effectiveness of inebilizumab treatment in patients with NMOSD in Japan (UMIN-Clinical Trials Registry UMIN000044431).

Patient registration was conducted from June 1, 2021 to November 30, 2023, with the total surveillance period for this PMS from June 1, 2021 to May 31, 2027. All patients who received inebilizumab during the registration period were included, with each patient followed for up to 6 years after initiation of therapy. Patients discontinuing inebilizumab treatment are to be observed for 2 years after their final dose, or until the end of the surveillance period, whichever occurs first.

The results reported here are from the interim analysis of this ongoing PMS, using data collected between June 2021 and the data-lock date of June 10, 2025, at which time 1-year case report forms were available for most patients.

Participants

All patients who received at least one dose of inebilizumab during the registration period (June 1, 2021–November 30, 2023) were included. Patients who transferred to another institution after initiating inebilizumab remained eligible if the first dose of inebilizumab was administered before November 30, 2023.

Ethical Approval

This PMS has been mandated by the Ministry of Health, Labour and Welfare of Japan and is being conducted in compliance with the Good Post-marketing Study Practice (GPSP) ordinance and the principles of the Declaration of Helsinki (1964) and its later amendments [15]. The study protocol was approved through pre-consultation with the Pharmaceuticals and Medical Devices Agency (PMDA), in accordance with the GPSP.

Prior to enrolment, all patients were informed of the purpose and scope of the PMS by the physicians, and written consent for publication of de-identified data has been obtained.

Data Collection and Outcome Measures

Patient data were collected using standardized questionnaires completed by investigators at prespecified time points (i.e., 1, 12, 24, 36, 48, 60, and 72 months) after the start of inebilizumab treatment as per the observation period for each patient.

Adverse events (AEs) and adverse drug reactions (ADRs) were classified according to Japanese version of the Medical Dictionary for Regulatory Activities (MedDRA/J), Version 28.0. ADRs were defined as AEs for which a causal relationship to inebilizumab could not be excluded or was unknown in the opinion of the patient’s physician or study sponsor. Serious ADRs were defined as any ADRs resulting in death, life-threatening events, disability, potential disability, hospitalization or prolonged hospitalization for treatment, congenital diseases or abnormalities, and any other medically important events. Safety specifications defined in the RMP included infections, IRR, malignancy, progressive multifocal leukoencephalopathy (PML), decreased immune response, and hepatitis B virus reactivation.

Relapse was defined as a typical clinical sign of an acute inflammatory demyelinating event of the CNS, present or previously observed, lasting ≥ 24 h without fever or infection [16]. Diagnosis of relapse was based on the decision made by the attending physician and was not subject to adjudication by an independent committee, unlike in the N-MOmentum study. An attack was defined as either the initial attack or a relapse; relapse referred to events occurring after the initial attack.

Worsening of EDSS score was defined, in accordance with the criteria in the N-MOmentum study [13], as an increase of ≥ 2 points from a baseline score of 0, an increase of ≥ 1 point from a baseline score of 1–5, or an increase of ≥ 0.5 points from a baseline score of ≥ 5.5.

The use of oral glucocorticoids and concomitant immunosuppressive therapies was assessed at baseline and during follow-up. The daily mean glucocorticoid dose (prednisolone equivalent) was calculated at each time point, and changes in dose were analyzed descriptively. Differences in glucocorticoid use were explored according to key baseline clinical characteristics. Patients were stratified by time since the most recent NMOSD relapse (≤ 365 vs ≥ 366 days; higher relapse risk within 1 year) [17] and by baseline EDSS (< 7.0 vs ≥ 7.0; severe ambulatory disability at 7.0, essentially restricted to a wheelchair). Subgroup analyses were also conducted according to the presence or absence of coexisting autoimmune disease, diabetes, osteoporosis, hypertension, or dyslipidemia, as long-term glucocorticoid use is associated with complications such as hyperglycemia, hypertension, and osteoporosis [18].

Statistical Analysis

Continuous variables were summarized descriptively using mean (standard deviation [SD]), median (minimum, maximum), or median (interquartile range [IQR]) while categorical data were summarized using frequencies and percentages. No imputation of missing values was performed.

The number of mean relapses (times) and the annualized relapse rate (ARR) (times/patient-year) were calculated using the person-years method and its 95% confidence interval (CI) (times/patient-year) for the 2 years (i.e., 730 days) before the start of administration of inebilizumab, and from the start of inebilizumab administration to the last observation date, in the effectiveness analysis set.

Subgroup analyses were performed to explore the influence of patient characteristics on outcomes.

Results

Patient Disposition

Of the patients enrolled in this ongoing PMS, 233 provided consent for publication and were included in this analysis (Fig. 1). At data cutoff (10 June 2025), case report forms had been collected and locked for 228 patients, all of whom were included in the safety analysis set. Following exclusion of four patients (three patients who continued treatment in the N-MOmentum study and one patient whose anti-AQP4 antibody positivity could not be confirmed), the effectiveness analysis set comprised 224 patients.

Fig. 1.

Patient flow diagram. AQP4 aquaporin-4, CRF case report form

Baseline Demographics and Clinical Characteristics

Patient demographic and clinical characteristics for the safety and effectiveness sets are shown in Table 1. In the safety analysis set, patients were predominantly female (85.1%, n = 194; male 14.9 %, n = 34), the mean ± SD age was 54.2 ± 13.1 years, and the median duration of NMOSD was 6.4 (range 0.1–40.6) years. The most frequent manifestations at NMOSD diagnosis were acute myelitis (71.9%, n = 164) and optic neuritis (51.3%, n = 117). Comorbid conditions were present in 68.9% (n = 157) of patients, including hepatic impairment (n = 16), Sjögren’s syndrome (n = 15), and Hashimoto’s disease (n = 8). Other frequently reported comorbidities that may be influenced by glucocorticoid use included hypertension (n = 36), dyslipidemia (n = 35), diabetes mellitus (n = 34; diabetes, n = 27, steroid-induced diabetes, n = 4; type 2 diabetes, n = 3), osteoporosis (n = 24), and hyperlipidemia (n = 15).

Table 1.

Patient demographic and clinical characteristics (safety analysis set and effectiveness analysis set)

	Safety analysis set (N = 228)	Effectiveness analysis set (N = 224)
Age, years, mean ± SD	54.2 ± 13.1	54.4 ± 13.0
Sex, n (%)
Male	34 (14.9)	33 (14.7)
Female	194 (85.1)	191 (85.3)
Duration of NMOSD, years, median (min–max)	6.4 (0.1–40.6)	6.3 (0.1–40.6)
Symptoms at NMOSD diagnosis, n (%)
Optic neuritis	117 (51.3)	114 (50.9)
Acute myelitis	164 (71.9)	160 (71.4)
Area postrema symptoms (e.g., nausea, vomiting, hiccups)	40 (17.5)	40 (17.9)
Acute brainstem syndrome	31 (13.6)	31 (13.8)
Symptomatic narcolepsy or acute diencephalic syndrome with diencephalic MRI lesions	4 (1.8)	4 (1.8)
Symptomatic cerebral syndrome with brain MRI lesions	16 (7.0)	16 (7.1)
Time from most recent NMOSD relapse to inebilizumab initiation, n (%)
≤ 365 days	117 (51.3)	117 (52.2)
≥ 366 days	97 (42.5)	93 (41.5)
Unknown/missing	14 (6.1)	14 (6.3)
EDSS, mean ± SD	4.21 ± 2.35	4.23 ± 2.35
≥ 4.5, n (%)	97 (42.5)	97 (43.3)
≥ 7.0, n (%)	33 (14.5)	33 (14.7)
Unknown/missing, n (%)	31 (13.6)	30 (13.4)
Comorbidities, n (%)	157 (68.9)	154 (68.8)
Hepatitis B	3 (1.9)	3 (1.9)
HIV infection	0	0
Other infectious diseases	5 (3.2)	5 (3.2)
Rheumatoid arthritis	1 (0.6)	1 (0.6)
Hashimoto’s disease	8 (5.1)	8 (5.2)
Systemic lupus erythematosus	3 (1.9)	3 (1.9)
Sjögren’s syndrome	15 (9.6)	15 (9.7)
Myasthenia gravis	4 (2.5)	4 (2.6)
Pernicious anemia	1 (0.6)	1 (0.6)
Other autoimmune diseases	12 (7.6)	12 (7.8)
Renal impairment	5 (3.2)	5 (3.2)
Hepatic impairment	16 (10.2)	16 (10.4)
Malignancy	1 (0.6)	1 (0.6)
Other	134 (85.4)	131 (85.1)
Glucocorticoid/immunosuppressant use at inebilizumab initiation, n (%)
Glucocorticoid use only	142 (62.3)	141 (62.9)
Immunosuppressant use only	6 (2.6)	6 (2.7)
Both glucocorticoid and immunosuppressant use	65 (28.5)	65 (29.0)
Neither glucocorticoid nor immunosuppressant use	15 (6.6)	12 (5.4)
Immunosuppressant drugs used at inebilizumab initiation, n (%)
Cyclosporine	6 (2.6)	6 (2.7)
Azathioprine	23 (10.1)	23 (10.3)
Tacrolimus hydrate	40 (17.5)	40 (17.9)
Methotrexate	1 (0.4)	1 (0.4)
Mycophenolate mofetil	1 (0.4)	1 (0.4)

EDSS Expanded Disability Status Scale, HIV human immunodeficiency virus, MRI magnetic resonance imaging, NMOSD neuromyelitis optica spectrum disorder, SD standard deviation

Over half (52.2%, n = 117) of patients in the effectiveness analysis set initiated inebilizumab treatment within a year of their last NMOSD attack, and 14.7% (n = 33) of patients had an EDSS score of ≥ 7.0 at baseline (Table 1). At inebilizumab treatment initiation, 92.0% (n = 206) were receiving glucocorticoids and 31.7% (n = 71) were receiving concomitant immunosuppressive therapy (Table 1). The types and percentages of immunosuppressant drugs used at inebilizumab initiation are summarized in Table 1.

Exposure to Inebilizumab Treatment

In the safety analysis set, the median duration of inebilizumab treatment was 379.0 (range 1–1102) days, and the median number of infusions was 4.0 (range 1–8). Twenty-one patients (9.2%) discontinued inebilizumab treatment during the observation period; reasons for discontinuation included AEs (n = 9), inadequate response (n = 5), transfer to another hospital (n = 2), and other reasons (n = 5), which included patient preference (n = 3) and switching to another drug (n = 2).

Safety

Adverse Drug Reactions

Of the 228 patients in the safety analysis set, ADRs occurred in 74 patients (32.5%), 31 (13.6%) of which were serious (Table 2). ADRs leading to discontinuation of inebilizumab occurred in five patients (2.2%)—pneumonia, hypersensitivity, neuralgia, interstitial lung disease, sudden cardiac death, anemia, and aplasia pure red cell. Anemia and aplasia pure red cell were observed in one patient, while hypersensitivity and neuralgia were observed in another patient. Two events in two patients, interstitial lung disease and sudden cardiac death, were fatal. Regarding the event of interstitial lung disease, the patient (a male patient in his 70s) with comorbid diabetes mellitus and atrial fibrillation had a history of pneumonia approximately 2 months before inebilizumab initiation. The patient had no smoking history and occupational exposure was unknown. Concomitant glucocorticoid use may have contributed to infection risk; however, a causal relationship to inebilizumab could not be excluded. It is possible that inebilizumab contributed to increased susceptibility to infection, which subsequently progressed to interstitial lung disease. The event of sudden cardiac death occurred in a separate patient (a female patient in her 60s). Although the cause was unknown, a causal relationship to inebilizumab could not be excluded.

Table 2.

Summary of adverse drug reactions and serious adverse drug reactions reported during the survey period (safety analysis set)

	N = 228
	ADR	Serious ADR
Patients with an event, n (%)	74 (32.5)	31 (13.6)
Infections and infestations	33 (14.5)	13 (5.7)
Bronchitis	1 (0.4)	–
Cellulitis	1 (0.4)	1 (0.4)
Cystitis	2 (0.9)	–
Gastroenteritis	1 (0.4)	–
Herpes zoster	2 (0.9)	1 (0.4)
Infection	1 (0.4)	1 (0.4)
Parotitis	1 (0.4)	1 (0.4)
Pharyngitis	1 (0.4)	–
Pneumonia	2 (0.9)	1 (0.4)
Pneumonia pneumococcal	1 (0.4)	1 (0.4)
Sinusitis	1 (0.4)	–
Upper respiratory tract infection	2 (0.9)	–
Urinary tract infection	4 (1.8)	2 (0.9)
Coronavirus infection	4 (1.8)	1 (0.4)
Arthritis bacterial	1 (0.4)	1 (0.4)
COVID-19	11 (4.8)	3 (1.3)
COVID-19 pneumonia	2 (0.9)	2 (0.9)
Neoplasms benign, malignant, and unspecified (including cysts and polyps)	1 (0.4)	1 (0.4)
Acute myelomonocytic leukemia	1 (0.4)	1 (0.4)
Blood and lymphatic system disorders	2 (0.9)	1 (0.4)
Anemia	1 (0.4)	1 (0.4)
Aplasia pure red cell	1 (0.4)	1 (0.4)
Iron deficiency anemia	1 (0.4)	–
Immune system disorders	2 (0.9)	1 (0.4)
Hypersensitivity	1 (0.4)	1 (0.4)
Hypogammaglobulinemia	1 (0.4)	–
Endocrine disorders	1 (0.4)	–
Autoimmune thyroiditis	1 (0.4)	–
Metabolism and nutrition disorders	1 (0.4)	–
Hyperuricemia	1 (0.4)	–
Psychiatric disorders	2 (0.9)	2 (0.9)
Anxiety	1 (0.4)	1 (0.4)
Depressive symptom	1 (0.4)	1 (0.4)
Nervous system disorders	5 (2.2)	2 (0.9)
Hypoesthesia	2 (0.9)	–
Myasthenia gravis	1 (0.4)	1 (0.4)
Neuralgia	1 (0.4)	1 (0.4)
Somnolence	1 (0.4)	–
Eye disorders	1 (0.4)	1 (0.4)
Visual impairment	1 (0.4)	1 (0.4)
Respiratory, thoracic, and mediastinal disorders	2 (0.9)	2 (0.9)
Acute respiratory distress syndrome	1 (0.4)	1 (0.4)
Interstitial lung disease	1 (0.4)	1 (0.4)
Gastrointestinal disorders	7 (3.1)	1 (0.4)
Colitis ischemic	1 (0.4)	1 (0.4)
Diarrhea	2 (0.9)	–
Gastritis erosive	1 (0.4)	–
Glossitis	1 (0.4)	–
Nausea	1 (0.4)	–
Pancreatic cyst	1 (0.4)	–
Stomatitis	1 (0.4)	–
Skin and subcutaneous tissue disorders	4 (1.8)	–
Blister	1 (0.4)	–
Pruritus	1 (0.4)	–
Skin swelling	1 (0.4)	–
Nail bed inflammation	1 (0.4)	–
Chronic pigmented purpura	1 (0.4)	–
Musculoskeletal and connective tissue disorders	6 (2.6)	2 (0.9)
Arthralgia	1 (0.4)	–
Muscle spasms	2 (0.9)	2 (0.9)
Myalgia	1 (0.4)	–
Pain in extremity	1 (0.4)	–
Rheumatoid arthritis	1 (0.4)	–
Renal and urinary disorders	2 (0.9)	1 (0.4)
Renal impairment	2 (0.9)	1 (0.4)
General disorders and administration site conditions	10 (4.4)	3 (1.3)
Feeling abnormal	1 (0.4)	–
Injection site erythema	1 (0.4)	–
Malaise	1 (0.4)	–
Edema peripheral	3 (1.3)	–
Pyrexia	3 (1.3)	1 (0.4)
Sudden cardiac death	1 (0.4)	1 (0.4)
Lithiasis	1 (0.4)	1 (0.4)
Investigations	18 (7.9)	2 (0.9)
Blood immunoglobulin G decreased	7 (3.1)	–
Blood immunoglobulin M decreased	3 (1.3)	–
Immunoglobulins decreased	4 (1.8)	–
Lymphocyte count decreased	4 (1.8)	2 (0.9)
Hepatic enzyme increased	1 (0.4)	–
Anti-aquaporin-4 antibody	1 (0.4)	–
Total complement activity increased	1 (0.4)	–
Injury, poisoning, and procedural complications	21 (9.2)	3 (1.3)
Infusion-related reaction	21 (9.2)	3 (1.3)
Surgical and medical procedures	1 (0.4)	1 (0.4)
Oxygen therapy	1 (0.4)	1 (0.4)

ADR adverse drug reaction

Safety Specifications

Infections, IRRs, hepatitis B virus reactivation, PML, decreased immune response, and malignancies were monitored as safety concerns.

Infections were reported in 33 patients (14.5%), 13 (5.7%) of which were serious. Serious infections included COVID-19 (n = 3), COVID-19 pneumonia (n = 2), urinary tract infection (UTI) (n = 2), and cellulitis, herpes zoster, infection (unspecified), parotitis, pneumonia, pneumonia pneumococcal, coronavirus infection, and arthritis bacterial (n = 1 each).

IRRs occurred in 21 patients (9.2%), three (1.3%) of which were classified as serious. Three serious IRRs occurred in patients during the first dose; of these, one patient discontinued the infusion before completion and was subsequently recovering. The remaining two patients received the full dose. The three patients were reported to have received their subsequent dose of inebilizumab.

Malignancy (acute myelomonocytic leukemia) was reported in one patient (0.4%). No cases of hepatitis B virus reactivation or PML were observed. With respect to decreased immune response, no patients were reported to have received live vaccines or live attenuated vaccines since 4 weeks before inebilizumab treatment initiation.

Effectiveness

Of the 224 patients included in the effectiveness analysis set, 20 experienced at least one relapse during inebilizumab treatment period. The ARR following inebilizumab treatment was 0.10 (95% CI 0.07–0.14) relapses per patient-year, which compared favorably with 0.40 (95% CI 0.33–0.47) relapses per patient-year during the 2 years prior to inebilizumab initiation (Fig. 2). Among the 20 patients who experienced relapse, five discontinued inebilizumab treatment: four due to an inadequate response to treatment and one due to AEs. Thirteen patients experienced one relapse, while seven patients experienced two or more relapses. All seven patients who experienced multiple relapses had a duration of NMOSD of at least 5 years (Supplementary Table S1).

Fig. 2.

Annualized relapse rate (times per person-year) 2 years before inebilizumab treatment and following inebilizumab treatment in effectiveness analysis set. CI confidence interval

Mean ± SD EDSS scores were 4.26 ± 2.25 at baseline (n = 168), 3.80 ± 2.27 at 6 months (n = 155), and 3.74 ± 2.29 at 1 year (n = 142) after inebilizumab treatment. Worsening of EDSS was observed in five patients (3.2%) at 6 months and seven patients (4.9%) at 1 year, indicating stability of overall disability status in the vast majority of patients (Table 3).

Table 3.

Change in EDSS during the survey period (effectiveness analysis set)

Time point	Change	n (%)
6 months	Decreased	26 (16.8)
	Unchanged	121 (78.1)
	Increased	8 (5.2)
	Totala	155 (100.0)
	Worseningb	5 (3.2)
1 year	Decreased	30 (21.1)
	Unchanged	103 (72.5)
	Increased	9 (6.3)
	Totala	142 (100.0)
	Worseningb	7 (4.9)

^aTotal refers to the sum of patients classified as having decreased, unchanged, or increased EDSS scores. The analysis included only those patients with EDSS data available both before the initiation of inebilizumab and at least once after treatment began. “Decreased” was defined as any reduction in EDSS score compared with baseline; “unchanged” as no change from baseline; and “increased” as any higher EDSS score after treatment initiation

^bWorsening was defined according to established EDSS thresholds: for patients with a baseline EDSS of 0, worsening was defined as an increase of ≥2 points; for baseline scores of 1.0–5.0, as an increase of ≥1 point; and for baseline scores ≥5.5, as an increase of ≥0.5 points

EDSS Expanded Disability Status Scale

Change in Use of Oral Glucocorticoids and Immunosuppressants

In the effectiveness analysis set, 29.0% of patients (n = 65) were receiving both oral glucocorticoids and immunosuppressants at inebilizumab initiation, which decreased to 14.3% (n = 29) at week 26, and 10.2% (n = 18) at week 52 (Fig. 3a). The mean daily oral glucocorticoid dose also decreased from 11.7 mg at inebilizumab initiation to 4.3 mg at week 52 (Fig. 3b). The proportion of patients not using oral glucocorticoids increased from 8.0% (n = 18) at baseline to 37.5% (n = 66) at week 52, and the proportion of patients receiving ≥ 20 mg/day decreased from 21.0% (n = 47) to 1.1% (n = 2) (Fig. 3c). Steroid withdrawal syndrome was reported in one patient as an AE not causally related to inebilizumab.

Fig. 3.

a Proportion of patients according to oral glucocorticoid and immunosuppressant use, b Changes in daily mean ± SD dose of oral glucocorticoid dose (prednisolone equivalent) from initiation of inebilizumab to week 52, and c Proportion of patients according to oral glucocorticoid use during the course of inebilizumab treatment, in the effectiveness analysis set

Subgroup analyses showed that patients who had experienced an NMOSD attack within 365 days before starting inebilizumab required higher mean daily doses of oral glucocorticoids and were less likely to have minimal (< 5 mg/day) or no glucocorticoid use through to week 52 compared with those who had been relapse-free for ≥ 366 days (Supplementary Fig. S1a). Similarly, patients with an EDSS score of ≥ 7.0 at baseline also had higher glucocorticoid doses through to week 52 than those with an EDSS score of < 7.0 at baseline (Supplementary Fig. S1b). Importantly, in each of these subgroups, the mean daily oral glucocorticoid dose decreased from initiation of inebilizumab to week 52. No major differences were observed according to the presence or absence of coexisting autoimmune disease, diabetes, osteoporosis, hypertension, or dyslipidemia, although some small numerical trends were noted (Supplementary Fig. S1c–g).

Laboratory Findings

Data on the percentage of peripheral blood B cells, and IgG and IgM concentrations were available for only a subset of patients in the safety analysis set (228 patients); therefore, the following results should be interpreted with caution. In the safety analysis set, the mean ± SD percentage of B cells among total peripheral blood lymphocytes decreased from 6.473 ± 5.661% (n = 32) at baseline, to 1.424 ± 1.536% (n = 24) at week 2, 1.171 ± 2.524% (n = 21) at 6 months, and 1.271 ± 1.750% (n = 17) at 1 year (Supplementary Fig. S2a). Although the number of evaluable patients was small, this downward trend was consistent with the pharmacological mechanism of B cell depletion.

Mean ± SD IgG concentrations decreased slightly over time, from 869.71 ± 364.93 mg/dL (n = 121) at baseline, to 804.66 ± 272.80 mg/dL (n = 80) at week 2, 740.19 ± 244.39 mg/dL (n = 96) at 6 months, and 716.17 ± 243.24 mg/dL (n = 77) at 1 year (Supplementary Fig. S2b).

A similar trend was observed for IgM concentrations (Supplementary Fig. S2c).

Discussion

The results of this interim analysis from an ongoing, nationwide PMS provide important real-world evidence regarding the safety and effectiveness of inebilizumab in 228 patients with NMOSD in Japan. There were no new safety concerns identified regarding inebilizumab use, and the ARR was lower after treatment initiation compared with the 2 years before treatment. Concomitant use of oral glucocorticoids and immunosuppressants also tended to decrease after inebilizumab initiation compared with before initiation.

Compared with patients that were anti-AQP4 antibody positive in the pivotal N-MOmentum study, patients in this PMS were older (mean 54.2 vs 43.0 years), had a lower ARR (0.40 vs 1.63 times per person-year) and had longer disease duration (mean 9.1 vs 2.6 years) [14]. Of note, this PMS enrolled patients exclusively in Japan, whereas the N-MOmentum study enrolled patients with NMOSD from 25 countries worldwide; Asian patients accounted for 21% of the anti-AQP4 antibody-positive population, with only a limited number of patients enrolled from Japan [13]. Furthermore, the N-MOmentum study excluded patients with autoimmune diseases requiring immunosuppressive therapy or a history of malignancy whereas this PMS included such patients.

Adverse Drug Reactions

Infections and IRRs were the most frequently reported ADRs in this PMS. Infections occurred in 33 patients (14.5%) and were serious in 13 patients (5.7%). Although no increase in the incidence of infections was observed over 4 years in the N-MOmentum study [14], continued monitoring of infections in real-world clinical practice will be conducted in this PMS. When interpreting these real-world infection data, it should be noted that many patients received concomitant oral glucocorticoids and immunosuppressants at inebilizumab initiation, whereas concomitant immunosuppressant use was not permitted in the N-MOmentum study, which may limit direct comparability between this PMS and the N-MOmentum study.

Although data were collected from a limited number of patients, the percentage of peripheral blood B cells decreased rapidly following initiation of inebilizumab, while serum IgG and IgM concentrations gradually declined during treatment. In the N-MOmentum study, rapid depletion of B cells and gradual reductions in serum IgG and IgM concentrations were reported, with no clear correlation observed between Ig levels and the risk of infection [14]. A previous study in Korea found that, in patients receiving long-term rituximab treatment, sustained memory B cell depletion, low baseline IgG, prior mitoxantrone treatment, and obesity may predispose patients to rituximab-induced hypogammaglobulinemia, but reduced immunoglobulin levels were not associated with an increased incidence of infections [19].

Respiratory tract infections were the most frequently reported infections, including COVID-19 (n = 11). UTI (n = 4) was also frequently reported. Among these, two cases were classified as serious; one occurred in a patient with an EDSS score of 8.5, and the other occurred in a patient whose EDSS score was unknown. Neurogenic bladder and severe motor impairment have been reported as risk factors for UTI in NMOSD [20], and in the N-MOmentum study, all three patients with a baseline EDSS score of 8.0 developed a UTI [12]. In patients with a high degree of disability, careful monitoring for possible signs and symptoms of UTI and dysuria is therefore recommended.

IRRs occurred in 21 patients (9.2%), most commonly either during or after the first infusion. This trend is consistent with the N-MOmentum study [14], which reported an increased incidence in IRRs during the first dose and a decrease with subsequent doses. Of the 14 patients who experienced IRRs during or after the first dose, 13 received premedication with a corticosteroid, antihistamine, and antipyretic while one received corticosteroids alone.

No cases of PML or hepatitis B virus reactivation were observed.

Relapse and Clinical Considerations

In this interim analysis, the ARR decreased from 0.40 relapses per patient-year in the 2 years before inebilizumab initiation to 0.10 thereafter, suggesting that inebilizumab may have contributed to relapse suppression. However, the follow-up period after treatment initiation remains short, and continued observation is necessary to evaluate long-term effectiveness and cumulative relapse-free rates. In the N-MOmentum trial, relapses in patients that are anti-AQP4 antibody positive predominantly occurred within the first year and subsequently declined [14]. Although reference to the N-MOmentum trial is useful for contextualizing the findings of this PMS, differences in study design and concomitant medication restrictions should be considered. In particular, oral glucocorticoids were permitted during the first 3 weeks of treatment, whereas other immunosuppressants were not allowed, which may limit direct comparability between studies [14]. In addition, other factors that may influence relapse risk should be considered, such as the degree of B cell depletion [21] and the potential persistence of antibody production by CD19-negative long-lived plasma cells [22].

Of the 20 patients who experienced relapse in this PMS, five discontinued treatment: four due to insufficient response (two after two relapses and two after a single relapse) and one due to a relapse classified as an adverse event. In contrast, three other patients continued inebilizumab treatment despite experiencing multiple relapses. According to Holian and Weinshenker, even a single relapse may prompt a treatment change; however, patients are more likely to continue treatment if their relapse is relatively mild, recovery is satisfactory, sufficient time has elapsed since the last relapse, or multiple prior treatments have failed [23]. These considerations reflect general clinical decision-making and should not be interpreted as a recommendation to continue inebilizumab following relapse.

Concomitant Oral Glucocorticoid Use

In the effectiveness analysis set, the mean daily oral glucocorticoid dose decreased by approximately 63% from inebilizumab initiation to week 52. Over the same period, the proportion of patients receiving ≥ 20 mg/day decreased from 21.0% to 1.1%, whereas the proportion not receiving oral glucocorticoid increased from 8.0% to 37.5%.

In subgroup analyses stratified by time since the most recent NMOSD attack, the mean daily oral glucocorticoid dose at baseline was twofold higher in patients who had experienced an attack within 365 days before inebilizumab initiation than in those without an attack for ≥ 366 days, which may reflect higher disease activity and greater clinical need at treatment start. Nevertheless, oral glucocorticoid doses decreased over time even in this subgroup, with a lower proportion of patients receiving ≥ 20 mg/day and a higher proportion receiving < 5 mg/day or no oral glucocorticoids.

When stratified by baseline disability, the mean daily oral glucocorticoid dose was lower in patients with an EDSS < 7.0 than in those with an EDSS ≥ 7.0, and both groups showed reductions during the treatment period. However, by week 52, the proportion of patients not receiving oral glucocorticoids increased to approximately 45% in the EDSS < 7.0 group, whereas it remained < 10% in the EDSS ≥ 7.0 group, suggesting that higher baseline disability may be associated with continued glucocorticoid use.

Overall, oral glucocorticoid reduction was observed during inebilizumab treatment across subgroups, including the subgroup with a recent attack (within 365 days). Given that the extent and pace of tapering may vary by patient characteristics (e.g., disease activity and physical condition), it will be important to establish safer and more effective strategies for glucocorticoid dose reduction tailored to the characteristics of each patient.

Limitations

The PMS was observational in the real-world clinical practice and is therefore subject to reporting bias. The absence of a control group prevents direct comparisons and limits causal inferences regarding the safety and effectiveness of inebilizumab. The non-interventional design resulted in incomplete data for laboratory tests. Relapse assessment was based on the physician judgment and may not be as rigorous as in clinical trials, potentially affecting relapse assessment. This report represents an interim analysis, and the study is ongoing. Longer-term follow-up is therefore needed to confirm the sustained safety and effectiveness of inebilizumab, including infection risk, and other long-term outcomes.

Conclusions

This interim analysis of a PMS study in Japan shows the real-world safety and effectiveness of inebilizumab for prevention of relapse in patients with NMOSD. Treatment was generally well tolerated, with no new safety concerns identified. ADRs occurred in approximately one-third of patients, most of which were non-serious and manageable. The ARR was lower after inebilizumab initiation compared with the 2 years prior to treatment, and the majority of patients remained relapse-free during treatment period. Additionally, concomitant use of oral glucocorticoids and immunosuppressive therapies decreased over the course of inebilizumab treatment. As the 6-year surveillance period continues, further data will clarify the long-term safety and sustained effectiveness of inebilizumab in Japanese patients with NMOSD.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contribution

Study design and methodology: Kazuo Fujihara, Kyoko Kato; Project administration: Kyoko Kato, Shinya Hirota; Data acquisition: Hideaki Hida; Analyses (including statistical): Hideaki Hida; Visualization: Shinya Hirota; Investigation, manuscript writing, feedback, and revisions: Kazuo Fujihara, Shinya Hirota, Muneyoshi Kudo, Hideaki Hida, Satoshi Yuki, Takashi Narisawa and Kyoko Kato.

Funding

This study and payment of the Rapid Service fee for publication was funded by Tanabe Pharma Corporation. The study sponsor participated in the design, execution and analysis of the study, and the conception, planning, writing and decision to publish the manuscript.

Data Availability

The datasets generated during this PMS are not publicly available because consent for public data sharing was not obtained.

Declarations

Conflict of Interest

Kazuo Fujihara has received funding and medical writing support from Tanabe Pharma Corporation; consulting fees from Tanabe Pharma Corporation, Chugai/Roche, Argenyx, and UCB; Payment or honoraria for lectures, presentations, speakers’ bureaus, manuscript writing and educational events from Tanabe Pharma Corporation, Eisai, Biogen, Novartis, Chugai/Roche, Alexion, Sanofi, Teijin, Asahi Kasei Medical, and UCB; support for attending meetings and/or travel from Roche/Chugai, Novartis, and Asahi Kasei Medical; participation in a Data Safety Monitoring Board or Advisory Board for Tanabe Pharma Corporation, Biogen, Novartis, Chugai/Roche, Alexion, Sanofi, UCB, and Amgen; Paid or unpaid leadership or fiduciary role in board, society, committee or advocacy group including President Emeritus (unpaid) of the Pan-Asian Committee for Treatment and Research In Multiple Sclerosis (PACTRIMS), board member (unpaid) for Japan Multiple Sclerosis Society and European Charcot Foundation, and committee and board member (unpaid) on the executive committee of the International Medical and Scientific Board. Kazuo Fujihara is an Editorial Board member of Neurology and Therapy. Kazuo Fujihara was not involved in the selection of peer reviewers for the manuscript nor any of the subsequent editorial decisions. Shinya Hirota, Hideaki Hida, Kyoko Kato, Muneyoshi Kudo, Satoshi Yuki, and Takashi Narisawa are employees of Tanabe Pharma Corporation.

Ethical Approval

This PMS has been mandated by the Ministry of Health, Labour and Welfare of Japan and is being conducted in compliance with the Good Post-marketing Study Practice (GPSP) ordinance and the principles of the Declaration of Helsinki (1964) and its later amendments. The study protocol was approved through pre-consultation with the Pharmaceuticals and Medical Devices Agency (PMDA), in accordance with the GPSP. Prior to enrolment, all patients were informed of the purpose and scope of the PMS by the physicians, and written consent for publication of de-identified data has been obtained.