CADTH Health Technology Review

Tocilizumab for Neuromyelitis Optica Spectrum Disorder

Rapid Review

Authors: Qiukui Hao, Aleksandra Grobelna

Key Messages

• Patients with neuromyelitis optica spectrum disorder treated with tocilizumab may experience decreased risks of relapse or disability progression compared to before treatment, but it is uncertain whether these changes can be entirely attributed to tocilizumab (TCZ) treatment.

• TCZ might be better than azathioprine for patients with neuromyelitis optica spectrum disorder in reducing relapse risk.

• TCZ might improve or have no impact on pain and fatigue measures for patients with neuromyelitis optica spectrum disorder.

• Adverse events were common during TCZ treatment, but patients treated with TCZ may experience adverse events at a similar or lower rate to those treated with azathioprine.

• One evidence-based guideline suggests that TCZ can be used as 1 of the second-line drugs to prevent long-term relapse for patients with neuromyelitis optica spectrum disorder who have no response to other immunosuppressants.

• We did not find any studies that compared TCZ to other immunosuppressants or any studies on the cost-effectiveness of TCZ for treating neuromyelitis optica spectrum disorder that met the inclusion criteria for this report.

Context and Policy Issues

Neuromyelitis optica spectrum disorder (NMOSD) (also named neuromyelitis optica or Devic disease) is a rare and relapsing chronic autoimmune disease of the central nervous system characterized by demyelination and axonal damage primarily targeting astrocytes in optic nerves and the spinal cord.¹ Neuroimaging characteristics (observed by MRI) and aquaporin-4(AQP4) autoantibodies are important in NMOSD diagnosis and the differential diagnosis from multiple sclerosis and myelin oligodendrocyte glycoprotein antibody-associated diseases.¹ The NMOSD incidence (0.037 to 0.73 per 100,000 person-years) and prevalence (0.7 to 10 per 100,000 persons) vary across studies, with the incidence and prevalence reported to be the highest in Afro-Caribbean regions and lowest in Australia and New Zealand.² In British Columbia, Canada, the mean incidence of NMOSD ranged from 0.14 to 0.60 per 100,000 person-years from 1986 to 2020 in Asian populations.³ Sex and ethnic groups were identified as influencing factors for these epidemiological data.

The median age of NMOSD onset is 32 to 41 years, with the peak incidence in middle-aged adults (30 to 59 years), but it also can affect children and older adults.^1,2 The prevalence of NMOSD among females was reported as 2.3 to 7.6 times greater than that among males.² People with NMOSD suffer from a wide range of symptoms that include severe visual loss, physical disability, pain, bladder dysfunction, intractable vomiting and hiccoughs, or other non-disease-specific symptoms (e.g., muscle, trunk, or leg pain).¹ The features of clinical manifestations for NMOSD typically reflect the acute episodes of optic neuritis (optic nerve inflammation), transverse myelitis (spinal cord inflammation), and area postrema syndrome.⁴ Unlike multiple sclerosis, patients with NMOSD are unlikely to experience a secondary progressive phase of the disease; however, relapse and attacks are closely related to disability and death.¹ Over half of the patients with NMOSD will be disabled and approximately one-third of patients will have died within 5 years after the first attack if the disease is left untreated.⁵

Acute attacks for NMOSD include the first attack and subsequent acute relapses defined as the worsening of original symptoms or the onset of new related symptoms that last for at least 24 hours and occur more than 30 days after the previous attack.⁶ Severe acute attacks can lead to paraplegia (paralysis of the legs), blindness, respiratory failure, or long-term disabilities.⁶ The most important goals of the management for NMOSD are to prevent disability and mortality by reducing the duration and severity of acute attacks or the risk of consequential relapses after the first attack.^4,6 For all patients with suspected NMOSD acute attack, several clinical practice guidelines recommend a high-dose glucocorticoid pulse as the first-line intervention and plasma exchange as the second-line treatment. Considering relapses could result in accumulated neurologic deficits closely related to poor prognosis, long-term immunotherapy (at least 5 years or lifetime treatment) is generally used once the diagnosis of NMOSD is made or after the management of the acute attacks.⁷

For long-term immunotherapies, available options include humanized monoclonal antibodies, azathioprine (AZA), mycophenolate mofetil (MMF), methotrexate, mitoxantrone, and oral glucocorticoids.⁷ Health Canada–approved the humanized monoclonal antibodies eculizumab and satralizumab as immunosuppressive drugs for preventing relapse in people with NMOSD.⁸ However, due to inconsistency in the limited available evidence, the current clinical practice guidelines have not established the optimal immunotherapies and treatment durations.⁷ Some clinicians prescribe the off-label use of tocilizumab (TCZ) or rituximab for patients with NMOSD as alternative immunosuppressive drugs.⁷

TCZ is a humanized monoclonal antibody against the interleukin-6 (IL-6) receptor that is approved in Canada for the treatment of moderately to severely active rheumatoid arthritis.^9,10 TCZ also can be used in patients hospitalized with Covid-19 pneumonia and other diseases.¹¹ In patients with NMOSD, IL-6 can stimulate abnormal B-cell response and increase AQP4 autoantibodies that play a crucial role in immunopathogenesis and astrocyte injury.¹² TCZ as 1 of the IL-6 inhibitors, had demonstrated immunological effects on immune cells and in reducing the AQP4 antibody levels,¹³ suggesting it may be a promising immunotherapy option for this condition. The purpose of this report is to review the clinical effectiveness and cost-effectiveness of TCZ for patients with NMOSD compared to alternative pharmacological therapies or no treatment. We also summarized relevant evidence-based guidelines regarding the use of TCZ in patients with NMOSD.

Research Questions

1. What is the clinical effectiveness of tocilizumab for patients with neuromyelitis optica spectrum disorder compared to alternative pharmacological therapies?

2. What is the clinical effectiveness of tocilizumab for patients with neuromyelitis optica spectrum disorder compared to no treatment?

3. What is the cost-effectiveness of tocilizumab for patients with neuromyelitis optica spectrum disorder?

4. What are the evidence-based guidelines regarding the use of tocilizumab for the treatment of patients with neuromyelitis optica spectrum disorder?

Methods

Literature Search Methods

A literature search was conducted by an information specialist on key resources, including MEDLINE, Embase, the Cochrane Database of Systematic Reviews, the International HTA Database, the websites of Canadian and major international health technology agencies, as well as a focused internet search. The search strategy comprised controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concepts were tocilizumab and neuromyelitis optica spectrum disorder (NMOSD). No filters were applied to limit the retrieval by study type. Where possible, retrieval was limited to the human population. The search was completed on February 2, 2023 and was limited to English-language documents published since January 1, 2018.

Selection Criteria and Methods

One reviewer screened citations and selected studies. In the first level of screening, titles and abstracts were reviewed and potentially relevant articles were retrieved and assessed for inclusion. The final selection of full-text articles was based on the inclusion criteria presented in Table 1.

Exclusion Criteria

Articles were excluded if they did not meet the selection criteria outlined in Table 1. We also excluded duplicate publications and citations that were published before 2018. Systematic reviews in which all relevant studies were captured in other more recent or more comprehensive systematic reviews were excluded. Primary studies retrieved by the search were excluded if they were captured in 1 or more included systematic reviews or reported the impacts of the intervention only on immune cells. Guidelines with unclear methodology were also excluded.

Critical Appraisal of Individual Studies

The included publications were critically appraised by 1 reviewer using the following tools as a guide: A Measurement Tool to Assess systematic Reviews 2 (AMSTAR 2)¹⁴ for systematic reviews, the Downs and Black checklist¹⁵ for randomized and non-randomized studies, and the Appraisal of Guidelines for Research and Evaluation (AGREE) II instrument¹⁶ for guidelines. Summary scores were not calculated for the included studies; rather, the strengths and limitations of each included publication were described narratively.

Summary of Evidence

Quantity of Research Available

A total of 54 citations were identified in the literature search. Following the screening of titles and abstracts, 24 citations were excluded, and 30 potentially relevant reports from the electronic search were retrieved for full-text review. Ten potentially relevant publications were retrieved from the grey literature search for a full-text review. Of these potentially relevant articles, 33 publications were excluded for various reasons, and 7 publications met the inclusion criteria and were included in this report. These comprised 4 systematic reviews, 2 non-randomized studies, and 1 evidence-based guideline. Appendix 1 presents the PRISMA¹⁷ flow chart of the study selection. Additional references of potential interest are provided in Appendix 6.

Summary of Study Characteristics

Four systematic reviews,^18-21 2 non-randomized studies,^22,23 and 1 evidence-based guideline²⁴ were included. Three systematic reviews^19-21 had broader inclusion criteria on interventions; only the subset of studies reporting on TCZ and its comparison were relevant to the current report. We only described the characteristics and results of this subset in this report. Appendix 5 includes a table describing the overlap in relevant primary studies in the included systematic reviews. Given this overlap, the summary of the systematic reviews in this report may repeat some primary study findings. Additional details regarding the characteristics of included publications are provided in Appendix 2.

Study Design

Three^18-20 of 4 systematic reviews included meta-analyses and 1 systematic review²¹ summarized the findings from primary studies narratively but did not conduct a meta-analysis. The number of primary studies involving TCZ that were included in the systematic reviews ranged from 7 to 27 primary studies, which comprised retrospective observational studies, case reports, and 1 randomized controlled trial (RCT). One review¹⁸ provided an indirect comparison of TCZ (7 primary studies) and AZA (19 primary studies) for 1 outcome using pooled effect size based on the test of interaction.²⁵ Three systematic reviews^18,20,21 included the same RCT (TANGO trial), which compared TCZ and AZA.²⁶ Two systematic reviews^20,21 were published in 2021, 1 systematic review¹⁹ was published in 2022 and the fourth systematic review¹⁸ was published in 2023 with the search up to July 21, 2022.

Two observational studies with a retrospective before-after design were published in 2023²³ and 2021.²² One study²³ retrospectively collected data for NMOSD patients treated with TCZ and compared the changes between baseline and after treatment. This study also employed adjusted logistic regression models to identify risk factors for relapse. The second study²² included NMOSD patients treated with TCZ, AZA, and rituximab and healthy controls. This study compared the differences between baseline and after treatment within each treatment group or across the 3 drug groups as well as the difference between NMOSD patients and the healthy controls on optical coherence tomography (OCT) measures.

One evidence-based clinical practice guideline was included from Latin America.²⁴ The guideline development group comprised experts in neurology who were involved in the diagnosis and care of NMOSD patients and conducted a systematic literature search (from 1990 to 2019) to identify relevant evidence. The ratings of the quality of evidence and strength of recommendations were not reported. The RAND-UCLA methodology for reaching formal consensus was used, and the recommendations were formulated based on formal consensus and voting processes.²⁴ The grade (i.e., appropriate, inappropriate, or uncertain) and proportion of expert agreement for recommendation statements were provided as outputs of the voting processes.

Country of Origin

Of the 4 systematic reviews,^18-21 2 systematic reviews were from Nepal,^20,21 the third systematic review was from China,¹⁸ and the fourth systematic review²¹ was from the US. One systematic review²⁰ presented the countries or regions of included primary studies. Of the 7 relevant studies, the RCT was conducted in 6 centres in China and the 6 observational studies were conducted in Germany, Spain, France, Japan, and the US.²⁰

The 2 observational studies^22,23 were single-centre studies that were conducted in the same hospital in China; and a group from Latin American developed the clinical practice guideline.²⁴

Patient Population

All 4 systematic reviews^18-21 involved patients with NMOSD, and the number of included patients ranged from 59 to 1,078. The age of patients ranged from 12 years to 68 years with mean age ranging from 29 years to 50 years in primary studies included in 1 systematic review;²⁰ mean ages were 42 years (TCZ studies) and 37 years (AZA studies) in the second systematic review;¹⁸ the mean age ranged from 29 to 50 in the primary studies included in the third systematic review;¹⁹ and age was not reported in the fourth systematic review.²¹ The study populations were predominantly female in 3 systematic reviews^18-20, with the proportion of females ranging from 86% to 92%. The fourth systematic review²¹ did not report the proportion of females. The proportion of patients with AQP4-positive serotype ranged from 14% to 100% across primary studies in 1 systematic review,¹⁹ 40% to 100% in the second review,²⁰ and was reported as 85.8% in patients with TCZ treatment in the third systematic review and 81.3% for the fourth systematic review.¹⁸ One review²⁰ noted that AZA and mycophenolate mofetil were the most common add-on drugs for TCZ, while the remaining systematic reviews^18,19,21 did not report the add-on drugs.

The 2 non-randomized studies^22,23 included adults with NMOSD. The first study included 65 adult patients with a median age of 44.²³ The second study²² included 50 adults with NMOSD and 10 healthy controls with an age range of 18 years to 75 years (mean age from 41 years to 53 years for patients with NMOSD and 44 for healthy controls). The proportion of females was 90% or over for both studies.^22,23 The proportion of patients with AQP4-IgG seropositive was 83% in 1 study,²³ while all patients were AQP4-IgG seropositive in the second study.²²

The included clinical practice guideline was for the management of patients with NMOSD.²⁴ The intended users of the guideline were clinicians involved in the care of patients with NMOSD. There was no pediatrician in the guideline development panel; therefore, pediatric patients with NMOSD may not have been considered in this guideline.

Interventions and Comparators

Two systematic reviews^18,21 compared the effectiveness and safety of TCZ and AZA, and the remaining 2 systematic reviews^19,20 compared the before-and-after changes for TCZ without a comparator group. Two systematic reviews^18,19 did not report the dosage and route of administration of TCZ. The other 2 systematic reviews^20,21 reported TCZ can be administrated with IV with a dosage of 6 mg/kg to 8 mg/kg every 4 or 6 weeks, or subcutaneously with a dose of 162 mg every 1 to 2 weeks, and AZA can be administered orally with a dosage of 2 to 3 mg/kg per day.

One eligible retrospective observational study²³ included patients who were regularly administered IV TCZ (8 mg/kg) every 4, 6 or 8 weeks and compared outcomes before-and-after treatment. The second retrospective observational study²² compared the before-after change scores of OCT measures across 3 drug interventions: IV TCZ (8 mg/kg/month), oral AZA (2 mg/kg/day to 3 mg/kg/day) and IV rituximab (dosage depending on the proportion of CD19+ B-cells in peripheral blood mononuclear cells) and healthy controls.

The included guideline²⁴ considered most treatments for NMOSD that includes TCZ, AZA, eculizumab, inebilizumab, satralizumab, cyclophosphamide, mitoxantrone, IV methylprednisolone, plasmapheresis, mycophenolate mofetil, and rituximab.

Outcomes

All 4 systematic reviews^18-21 reported relapse-related outcomes and disability status. Two systematic reviews^20,21 reported adverse events and death. One systematic review²¹ also reported MRI-related disease activity, AQP-4 antibody titres, pain and fatigue (each measured by a numerical rating scale with an unreported range). The specific outcomes for relapse included: annualized relapse rate (ARR), relapse-free rate, hazard risk for relapse and time to first relapse. Disability status was assessed by the Expanded Disability Status Scale (EDSS)²⁷ and ambulatory status. EDSS is a formal scale with possible scores ranging from 0 to10 with higher scores indicating greater disability, while ambulatory status was not a formal scale with the following options: unrestricted, restricted, wheelchair, cane, or walker.²⁸ The mean or median follow-up duration ranged from 16 months to 78.9 months in 1 systematic review,¹⁸ 12 months to 31 months in the second systematic review,¹⁹ 12 months to 31.8 months in the third systematic review,²⁰ and 3 months to 80 months in the fourth systematic review.²¹

One retrospective observational before-and-after study²³ reported relapse-related outcomes (including ARR, relapse-free rate, time to the first relapse and risk factors for relapse), disability status, and adverse events with a median follow-up of 34 months. The other retrospective observational study²² reported the changes in visual acuity measures and imaging features (OCT) with a median follow-up ranging from 13 months to 14.6 months.

The included guideline²⁴ considered the relapse risk, disability, and safety outcomes. For some recommendation statements, the costs were considered.

Summary of Critical Appraisal

Systematic Reviews

In all 4 systematic reviews,^18-21 the objective was clearly described, multiple databases were searched, keywords of the search and study selection flow charts were provided, the review authors declared no conflicts of interest, lists of included articles were presented and the study characteristics were described. None of the systematic reviews^18-21 provided lists of excluded articles or assessed the funding sources in individual studies. Two systematic reviews^18,21 did not report performing a grey literature search. These limitations may result in the inappropriate exclusion of some studies, missing some unpublished studies or misidentification of potential publication bias. In 1 systematic review,²⁰ article selection and data extraction were done independently by 2 reviewers. In 2 systematic reviews,^18,21 it was unclear how article selection and data extraction were conducted. In the fourth systematic review,¹⁹ the article selection was done independently by 2 reviewers, but it was unclear how the data extraction was performed, The possibility of missing some relevant articles or errors in data extraction cannot be ruled out. Although 3 systematic reviews^18-20 assessed the risk of bias of included individual studies (and reported study quality to be “satisfactory” or “Newcastle-Ottawa Scale score ranged from 5 to 7” for observational studies [1 systematic review²⁰ presented the overall rating for 6 individual studies: 5 high risk of bias and 1 low risk of bias]; with “some concerns” in risk of bias for the included RCT), they did not assess the potential impact of study risk of bias on the interpretation of results. One systematic review²¹ did not assess the risk of bias of included individual studies, so the quality of these studies is unknown. The results may be driven by individual studies with high risk of bias.

One systematic review¹⁸ compared the effectiveness of TCZ and AZA, but only included the TCZ arm of the TANGO trial and left the direct comparison between the 2 drugs out of the pooled analysis. Instead, the systematic review authors pooled data from single-arm trials of TCZ and AZA, and then used the Douglas G test of interaction²⁵ to compare the difference in ARR between the 2 drug groups. This method did not adjust potential confounding factors and cannot take the advantage of randomization in the primary RCT, which only can provide indirect evidence and is likely an inappropriate analysis method for indirect comparisons. This systematic review¹⁸ also misinterpreted the P values for effect size and used a fixed-effect model for meta-analysis with moderate heterogeneity, which will result in a narrow confidence interval and in favour of interventions that compared a random-effect model. One systematic review²¹ summarized findings from individual studies narratively, and some outcome measures (fatigue and ambulatory status) were not validated.

Among the 3 systematic reviews^18,20,21 that included the TANGO trial (which provided the only direct comparison between TCZ and AZA among all included primary studies), 2 systematic reviews^18,20 assessed the risk of bias for the TANGO trial. One systematic review²⁰ reported the TANGO trial had a high risk of bias in the deviation from the intended intervention domain and an unclear risk of bias in the missing outcome data domain based on Cochrane’s risk of bias tool. This RCT was assessed as having “some concerns” by systematic review authors. The other systematic review¹⁸ reported that 2 stars for “selection,” 1 star for “comparability,” an 3 stars for “outcome” based on the Newcastle-Ottawa Scale (no further information was available about how the risk of bias assessments were conducted and the meaning of the risk of bias).

Retrospective Observational Studies

The 2 retrospective observational studies^22,23 clearly reported study objectives, inclusion and exclusion criteria, intervention details, participant characteristics, outcome measures, and the main findings. The 2 studies used appropriate statistical analysis methods for comparing changes before-and-after the intervention. Both studies^22,23 were funded by academic institutions in China, which may have little influence on potential publication bias. Neither study^22,23 conducted sample size calculations, which may result in underpowering for detecting the changes in visual acuity or pain measures. One study²³ used a before-after design without a control group. Due to the nature of the design, it is difficult to attribute the before-and-after changes to the effect of TCZ. In this study,²³ the authors did not describe the secondary outcome measures well (disability and pain) and the covariates in the logistic regression models. In the other study,²² the authors did not provide detailed information on changes in visual acuity measures, the number of eyes in each drug group, and the covariates in the generalized estimating equations.

Evidence-Based Guideline

The critical appraisal of the included clinical practice guideline²⁴ was performed in a previous CADTH report on rituximab for NMOSD.²⁹ More details about the quality of the guideline can be found in the reports. Briefly, the guideline had clear descriptions of scope, population, target users, guideline developers, and was externally peer-reviewed. The guideline had a transparent literature search and recommendation development methods based on consensus and voting, but the evidence assessment and strength of the recommendations were unclear. Although several authors declared the conflicts of interest, how these were addressed is unclear.

Additional details regarding the strengths and limitations of included publications are provided in Appendix 3.

Summary of Findings

Three included systematic reviews^18,20,21 included the TANGO trial, which was an RCT that compared TCZ and AZA.²⁶ Two reviews^18,20 only included data from the TCZ arm of the trial, and the third systematic review²¹ included data from the 2 arms; however, they included the interim analysis for the TANGO trial as the final results of this trial were not available at that time. The directions of all reported outcomes in the interim analysis were the same as the final trial result report. In this report, we summarized the main findings of the trial based on the final report²⁶ rather than the interim analysis. Appendix 4 presents the main study findings.

Due to some overlap in the studies included in the systematic reviews (Appendix 5), the pooled estimates from these systematic reviews may contain some of the same data. For the systematic reviews without a meta-analysis that narratively summarized the findings on the same outcome,^20,21 to avoid duplication of results, outcome data from an individual primary study are only reported once.

Clinical Effectiveness of TCZ

Relapse-Related Outcomes

Four systematic reviews^18-21 reported on ARR, mostly based on findings from single-arm trials. Three systematic reviews^18-20 showed that pooled estimates of ARR were statistically significantly reduced after TCZ treatment compared with before TCZ treatment. One systematic review²¹ reported mean or median changes in ARR for the individual included studies and showed that ARR values were numerically lower after TCZ treatment than before TCZ treatment. The included before-after study²³ also showed that the ARR values were statistically significantly reduced after TCZ treatment compared with before TCZ treatment. One systematic review¹⁸ reported that ARR was statistically significantly lower with TCZ than AZA, based on indirect comparisons of pooled TCZ and AZA findings from separate, single-arm studies. Moderate to substantial statistical heterogeneities were found in this systematic review¹⁸ for TCZ and AZA arms.

One systematic review²¹ (presenting findings from the TANGO trial²⁶) and 1 before-after study²³ reported relapse-free rates. The proportion of patients who were relapse-free was 76.9% from 1 observational study with a median follow-up of 34.1 months after TCZ treatment.²³ The systematic review²¹ narratively summarized the findings of 2 individual included case series and reported 70.3% of patients were relapse-free after TCZ treatment with an unclear follow-up period. The proportion of patients who were relapse-free at the end of the study (up to 92 weeks of follow-up) was statistically significantly higher in the TCZ arm (89%) compared with the AZA arm (56%) in the TANGO trial.²⁶ The before-after study²³ showed that the median time to first relapse was 62 weeks for patients treated with TCZ, and identified an infusion interval of more than 4 weeks and pGFAP greater than 220 pg/mL as statistically significant risk factors for relapse with very wide confidence intervals (i.e., low precision). There was a trend for TCZ monotherapy to increase the risk of relapse, but this did not reach statistical significance.

Disability

Three systematic reviews^18-20 reported that pooled estimates of EDSS were statistically significantly reduced after TCZ treatment compared with before TCZ treatment. Moderate statistical heterogeneities were found in 2 systematic reviews.^18,19 One systematic review²¹ narratively summarized EDSS findings from 10 observational studies (case reports included) and reported 6 individual studies with “significant” EDSS improvement, 2 studies with “mild” improvement, and 2 studies with “unchanged” in EDSS after TCZ treatment. Levels of EDSS improvement were not otherwise defined by systematic review authors. This systematic review also narratively summarized the difference in EDSS between patients treated with TCZ and AZA from the TANGO trial. Presenting findings from the TANGO trial publication,²⁶ there was a statistically significantly lower disability progression rate measured by EDSS in the TCZ group than in the AZA group after 12 weeks of treatment (5 of59 versus 15 of 59, HR: 0.288; 95% CI, 0.105 to 0.759; P = 0.0087). The study authors concluded the same at 24 weeks of treatment; however, the statistical test results for comparing the 2 groups on disability progression were inconsistent within the trial publication at this time point (2 of 59 versus 5 of 59, HR: 0.221; 95% CI, 0.047 to 1.042; P = 0.0309 in the main text; P = 0.0004 in Table 2). The TANGO trial considered the minimally important difference for EDSS as validated for patients with multiple sclerosis (i.e., “disability progression” was defined as an increase in EDSS of at least 1.0 if baseline score was 5.5 or less; at least of 0.5 if baseline score was great than 5.5). The included before-after study²³ also showed that the median EDSS scores were statistically significantly reduced after TCZ treatment compared with before treatment both for AQP4-antibody positive and negative patients.

One systematic review²¹ identified a case series (with 12 patients) that reported on ambulatory status and found that most patients (n = 9) had stable ambulatory status after TCZ treatment.

Adverse Events

Two systematic reviews^20,21 and 1 before-and-after study²³ reported adverse events in patients with NMOSD who received TCZ treatment. One systematic review²⁰ showed that the pooled estimate of the proportion of patients with any adverse events was 56%, and the proportion of patients with serious adverse events was 11%. One systematic review²¹ that included findings from the TANGO trial reported that the proportions of patients with at least 1 adverse event were 97% (61% treatment-related adverse events as determined by the trial authors) in the TCZ arm and 95% (83% treatment-related adverse events as determined by the trial authors) in the AZA arm and the proportions of patients with at least 1 serious adverse event was 8% and 15%, respectively (without statistical test). The serious adverse events included pneumonia, herpes zoster infection, deep vein thrombosis, basal ganglia hemorrhage, and myelitis.

Mortality

One systematic review²¹ identified 3 deaths. One occurred in 1 of the included observational studies, and 2 occurred in the TANGO trial (1 for each arm). The trial and the systematic review authors noted these deaths are unlikely related to the treatment.

Pain and Fatigue

One systematic review²¹ and 1 before-after study reported on pain outcomes measured by the numerical rating scale. The systematic review showed that pain scores were numerically lower after 12 months of TCZ treatment compared with before TCZ treatment (without statistical test). However, the before-after study²³ reported the pain score remained unchanged after TCZ treatment (median follow-up of 34.1 months) compared with before TCZ treatment (without statistical test). Based on findings from individual included studies, 1 systematic review²¹ reported that the general fatigue scores (no details on the measures) were numerically lower after TCZ treatment compared with before TCZ treatment (without statistical test).

Other Outcomes

One observational study²² reported eye- and vision-related outcomes in patients with NMOSD who were treated with TCZ, rituximab, or AZA and compared the changes in eye measures between patients and heathy controls. This study concluded that there were no changes in LogMAR visual acuity or low-contrast letter acuity either in NMOSD^ON+ eyes (eyes with a history of acute optic neuritis attack) or NMOSD^ON- eyes (eyes without a history of acute optic neuritis attack) after TCZ, AZA and rituximab treatment compared with baseline. No data were reported to support conclusions on visual acuity. The study also reported OCT measures of eyes and noted that patients treated with TCZ and rituximab did not display macular ganglion cell complex or peripapillary retinal nerve fibre layer thinning in NMOSD^ON- eyes, while patients treated with AZA showed statistically significant macular ganglion cell complex or peripapillary retinal nerve fibre layer thinning in NMOSD^ON- eyes. NMOSD patients treated with AZA showed statistically significant greater thinning in macular ganglion cell complex or peripapillary retinal nerve fibre layer measures compared with patients treated with TCZ and rituximab in NMOSD^ON- eyes. These measures may serve as indicators of retinal ganglion cell loss.

One systematic review²¹ also reported MRI-related disease activity measures and AQP4-and myelin oligodendrocyte glycoprotein antibody titres in patients with NMOSD after TCZ treatment but noted that the 2-antibody status was not correlated with a higher risk of relapse.

Cost-Effectiveness of TCZ

No cost-effectiveness evidence regarding TCZ for patients with NMOSD was identified; therefore, no summary can be provided.

Evidence-Based Guidelines Regarding the Use of TCZ

The clinical practice guidelines²⁴ from Latin America presents statements on various treatments for patients with NMOSD, described in detail in a previous CADTH report.²⁹ For long-term relapse prevention, the guideline recommends the early start of immunosuppressant treatments to reduce disease activity and prevent NMOSD attacks. Immunosuppressant treatments addressed by this guideline include AZA, MMF, rituximab, TCZ, eculizumab, satralizumab, inebilizumab, and oral steroids.

TCZ is suggested as 1 of the second-line treatment options for patients who have no response to other immunosuppressants to prevent long-term relapse prevention (Appropriate, 90% agreement). Eculizumab, inebilizumab, and satralizumab also can be used under these conditions; the guideline did not favour a particular humanized monoclonal antibody for second-line therapy. When starting immunosuppressant treatment (AZA, MMF, or rituximab), oral steroid tapering or maintained therapy were suggested in this guideline; however, this guideline did not mention whether adding TCZ should be combined with the oral steroid therapy. This guideline did not report the strength of the recommendations or the quality of evidence informing the recommendations, though the authors cited the TANGO trial²⁶ and 1 observational study to support the statement regarding TCZ. The 2 studies were captured by systematic reviews included in this report.

Limitations

Although we identified 4 systematic reviews^18-21 and 2 retrospective observational studies^22,23 to answer the research question on the clinical effectiveness of TCZ, most included primary studies in these systematic reviews were uncontrolled before-after studies or case reports, except for 1 RCT. This RCT (with “some concerns” according to systematic review authors²⁰) overlapped among 3 systematic reviews, and an overlap matrix of relevant primary studies (Appendix 5) reveals that more recent systematic reviews do not substantially build on the evidence base. Due to the nature of the observational studies, potential selection bias, recall bias or performance bias in the body of evidence cannot be ruled out. Potentially inappropriate indirect comparisons of pooled estimates may amplify the bias. The quantity and quality of evidence regarding pain and fatigue were limited. The death events (n = 3) were low among included studies, and the comparison of TCZ and other immunosuppressants on mortality was not determined.

TCZ and other antibodies are usually expensive compared with traditional oral drugs (e.g., AZA), and patients underwent prolonged infusion intervals mainly due to economic burden.²³ Cost-effectiveness will be essential for future decision-making. However, we did not find any cost-effectiveness studies of TCZ among patients with NMOSD.

Most eligible patients for the included studies were adults and most (over 80%) of them are female, which reflects the prevalence of NMOSD worldwide.² However, the data did not allow us to comment on the subgroup effects of the pediatric population and the male population. When applying the evidence to these minor populations, one might need to consider the applicability of the evidence to the target population. The management of patients with NMOSD involved many drugs and the usual care may differ across different studies. Observational studies, especially for uncontrolled before-after studies, may not account for these differences, and it is unclear to what extent the difference in outcome measures between studies can be attributed to different usual care therapies. The variabilities in add-on drugs, treatment intervals, or comorbidities in the primary studies were probably some reasons for the observed heterogeneity in the pooled estimates. The current report did not identify comparisons of TCZ versus other immunosuppressant treatments such as satralizumab, MMF, methotrexate, cyclophosphamide, cyclosporine, prednisone, eculizumab, inebilizumab, and rituximab. Furthermore, the identified guideline²⁴ did not address the place in therapy for TCZ relative to other second-line treatment options. Together, it remains unclear which patients may benefit from TCZ compared to alternative immunosuppressant therapies, particularly other humanized monoclonal antibodies.

The primary studies of the included systematic reviews^18-21 were conducted in Asia, Europe, and the US. The clinical practice guideline²⁴ was from Latin America and considered the local cost in Latin America. Thus, the generalizability of these findings to settings in Canada is uncertain.

Conclusions and Implications for Decision- or Policy-Making

Four systematic reviews^18-21 and 2 observational studies^22,23 were identified to address the clinical effectiveness of TCZ for patients with NMOSD. Evidence from before-and-after studies without a control group suggests that patients treated with TCZ had statistically significantly reduced ARR and EDSS compared with before TCZ treatment. Evidence from 2 systematic reviews^18,21 indicated that TCZ may be better than AZA to manage patients with NMOSD in reducing the relapse risk and increasing the relapse-free rate. However, these findings were based on a single RCT²⁶ and an indirect comparison of pooled findings from single-arm trials, which did not account for confounding variables. One before-and-after study²³ without a control group, identified infusion interval of more than 4 weeks and pGFAP greater than 220 pg/mL as risk factors for relapse, but the confidence intervals were very wide (i.e., low precision). Using TCZ may slow disability progression compared with AZA; however, some observational studies showed the EDSS scores were unchanged after TCZ treatment. The adverse events were common during the treatment. However, the proportion of adverse events was similar between TCZ and AZA, and the proportions of treatment-related adverse events and serious adverse events were numerically higher in the AZA group than in the TCZ group. The serious adverse events included pneumonia, herpes zoster infection, deep vein thrombosis, basal ganglia hemorrhage, and myelitis. Based on a retrospective study,²² TCZ treatments may have little or no impact on visual acuity function but may have some benefits in OCT measures. The impacts of TCZ on pain and fatigue were inconsistent: some studies in 1 included systematic review²¹ showed numerically lower scores (pain and fatigue), while another study²³ with a longer follow-up did not show any changes in pain measure scores.

One evidence-based clinical practice guideline²⁴ from Latin America suggests that TCZ can be used for patients with NMOSD who have no response to other immunosuppressants to prevent long-term relapse. However, this guideline did not report the strength of the recommendation or the quality of evidence informing the recommendation. The available evidence identified for this report suggests TCZ can be potentially used before AZA when considering the relapse outcomes and disability status, but the quality of evidence was low. Future guidelines may need to consider all related evidence regarding benefits and harms to make transparent and trustworthy recommendations. Considering the limitations of the body of evidence, such as the limitation of observational studies with before-and-after study designs without a control group, and the potentially inappropriate indirect comparison of pooled estimates, these findings should be interpreted with caution. In addition, the generalizability of the evidence to settings in Canada or minor NMOSD populations (e.g., pediatric, male) was unclear.

Cost-effectiveness will be helpful for decision- or policy-making. However, no evidence regarding the cost-effectiveness of TCZ for patients with NMOSD was identified that met the inclusion criteria for this report. TCZ requires IV or subcutaneous administration; patients’ values or preferences regarding injection and oral medicine also may affect the decision-making.³⁰

To compare the clinical effectiveness of TCZ to other pharmacological therapies or no treatment, we should consider both direct and indirect evidence. The evidence synthesis by 1 included systematic review¹⁸ omitted the direct comparison from the TANGO trial that directly compared TCZ versus AZA. To help decision-making regarding the use of TCZ for NMOSD, we need to consider it relative to other immunosuppressant treatments, including other humanized monoclonal antibodies recommended for second-line therapy. However, an available systematic review with network meta-analysis³¹ did not include TCZ. When possible, given the availability of trials that address the relative effects of AZA versus other treatments, a systematic review with network meta-analysis may be helpful to answer the question. Future primary studies with a robust methodology, such as those with prospective designs or RCTs, comparing TCZ with other immunosuppressants, are also required.

References

1.Glisson CC. Neuromyelitis optica spectrum disorders (NMOSD): clinical features and diagnosis. In: Post TW, ed. UpToDate. Waltham (MA): UpToDate; 2023: https://www.uptodate.com/contents/neuromyelitis-optica-spectrum-disorders-nmosd-clinical-features-and-diagnosis. Accessed 2023 Mar 3.

2.Papp V, Magyari M, Aktas O, et al. Worldwide incidence and prevalence of neuromyelitis optica: a systematic review. Neurology. 2021;96(2):59-77. PubMed

3.Lee JD, Guimond C, Yee IM, et al. Incidence of multiple sclerosis and related disorders in Asian populations of British Columbia. Can J Neurol Sci. 2015;42(4):235-241. PubMed

4.Huda S, Whittam D, Bhojak M, Chamberlain J, Noonan C, Jacob A. Neuromyelitis optica spectrum disorders. Clin Med (Lond). 2019;19(2):169-176. PubMed

5.Wingerchuk DM, Hogancamp WF, O'Brien PC, Weinshenker BG. The clinical course of neuromyelitis optica (Devic's syndrome). Neurology. 1999;53(5):1107-1114. PubMed

6.Ma X, Kermode AG, Hu X, Qiu W. NMOSD acute attack: understanding, treatment and innovative treatment prospect. J Neuroimmunol. 2020;348:577387. PubMed

7.Glisson CC. Neuromyelitis optica spectrum disorders (NMOSD): treatment and prognosis. In: Post TW, ed. UpToDate. Waltham (MA): UpToDate; 2023: https://www.uptodate.com/contents/neuromyelitis-optica-spectrum-disorders-nmosd-treatment-and-prognosis. Accessed 2023 Mar 3.

8.Pharmacoeconomic report: satralizumab (enspryng): (Hoffmann-La Roche Limited): indication: neuromyelitis optica spectrum disorder. (CADTH Common Drug Review). Ottawa (ON): CADTH; 2021: https://www.ncbi.nlm.nih.gov/books/NBK583281/. Accessed 2023 Mar 3.

9.Scott LJ. Tocilizumab: a review in rheumatoid arthritis. Drugs. 2017;77(17):1865-1879. PubMed

10.CADTH Canadian Drug Expert Committee recommendation: tocilizumab (actemra - Hoffmann-La Roche Limited): indication: giant cell arteritis. Ottawa (ON): CADTH; 2018: https://www.ncbi.nlm.nih.gov/pubmed/30427624. Accessed 2023 Mar 3.

11.Ghosn L, Chaimani A, Evrenoglou T, et al. Interleukin-6 blocking agents for treating COVID-19: a living systematic review. Cochrane Database Syst Rev. 2021;3(3):CD013881. PubMed

12.Chihara N, Aranami T, Sato W, et al. Interleukin 6 signaling promotes anti-aquaporin 4 autoantibody production from plasmablasts in neuromyelitis optica. Proc Natl Acad Sci U S A. 2011;108(9):3701-3706. PubMed

13.Liu Y, Zhang H, Zhang TX, et al. Effects of tocilizumab therapy on circulating B Cells and T helper cells in patients with neuromyelitis optica spectrum disorder. Front Immunol. 2021;12:703931. PubMed

14.Shea BJ, Reeves BC, Wells G, et al. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ. 2017;358:j4008. PubMed

15.Downs SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health. 1998;52(6):377-384. PubMed

16.Agree Next Steps Consortium. The AGREE II Instrument. [Hamilton, ON]: AGREE Enterprise; 2017: https://www.agreetrust.org/wp-content/uploads/2017/12/AGREE-II-Users-Manual-and-23-item-Instrument-2009-Update-2017.pdf. Accessed 2023 Mar 3.

17.Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol. 2009;62(10):e1-e34. PubMed

18.Tang Q, Yao M, Huang Y, Bian J, Wang Y, Ji W. A comparison of the efficacy of tocilizumab versus azathioprine for neuromyelitis optica spectrum disorder: a study protocol for systematic review and meta-analysis. Medicine (Baltimore). 2023;102(4):e32748. PubMed

19.Luitel P, Ghimire A, Upadhyay D, Ojha R. Efficacy of monoclonal antibodies in neuromyelitis optica: an updated systematic review with meta-analysis. Clin Exp Neuroimmunol. 2022;13(4):194-207.

20.Kharel S, Shrestha S, Ojha R, Guragain N, Ghimire R. Safety and efficacy of interleukin-6-receptor inhibitors in the treatment of neuromyelitis optica spectrum disorders: a meta-analysis. BMC Neurol. 2021;21(1):458. PubMed

21.Lotan I, McGowan R, Levy M. Anti-IL-6 therapies for neuromyelitis optica spectrum disorders: a systematic review of safety and efficacy. Curr Neuropharmacol. 2021;19(2):220-232. PubMed

22.Zeng P, Du C, Zhang R, et al. Optical coherence tomography reveals longitudinal changes in retinal damage under different treatments for neuromyelitis optica spectrum disorder. Front Neurol. 2021;12:669567. PubMed

23.Yang S, Zhang C, Zhang TX, et al. A real-world study of interleukin-6 receptor blockade in patients with neuromyelitis optica spectrum disorder. J Neurol. 2023;270(1):348-356. PubMed

24.Carnero Contentti E, Rojas JI, Cristiano E, et al. Latin American consensus recommendations for management and treatment of neuromyelitis optica spectrum disorders in clinical practice. Mult Scler Relat Disord. 2020;45:102428. PubMed

25.Altman DG, Bland JM. Interaction revisited: the difference between two estimates. BMJ. 2003;326(7382):219. PubMed

26.Zhang C, Zhang M, Qiu W, et al. Safety and efficacy of tocilizumab versus azathioprine in highly relapsing neuromyelitis optica spectrum disorder (TANGO): an open-label, multicentre, randomised, phase 2 trial. Lancet Neurol. 2020;19(5):391-401. PubMed

27.Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology. 1983;33(11):1444-1452. PubMed

28.Lotan I, Charlson RW, Ryerson LZ, Levy M, Kister I. Effectiveness of subcutaneous tocilizumab in neuromyelitis optica spectrum disorders. Mult Scler Relat Disord. 2019;39:101920. PubMed

29.Banerjee S, Butcher R. Rituximab for the treatment of neuromyelitis optica spectrum disorder. (CADTH Health Technology Review). Can J Health Technol. 2021;1(2). https://www.cadth.ca/sites/default/files/pdf/htis/2021/RD0058%20Rituximab%20for%20NMOSD%20Final.pdf. Accessed 2023 Mar 3.

30.González-González JG, Diaz Gonzalez-Colmenero A, Millan-Alanis JM, et al. Values, preferences and burden of treatment for the initiation of GLP-1 receptor agonists and SGLT-2 inhibitors in adult patients with type 2 diabetes: a systematic review. BMJ Open. 2021;11(7):e049130. PubMed

31.Huang W, Wang L, Zhang B, Zhou L, Zhang T, Quan C. Effectiveness and tolerability of immunosuppressants and monoclonal antibodies in preventive treatment of neuromyelitis optica spectrum disorders: a systematic review and network meta-analysis. Mult Scler Relat Disord. 2019;35:246-252. PubMed

Appendix 1: Selection of Included Studies

Figure 1: Selection of Included Studies

Appendix 2: Characteristics of Included Publications

Note that these tables have not been copy-edited.

Table 2: Characteristics of Included Systematic Reviews

AQP4 = aquaporin 4; ARR = annualized relapse rate; AZA = azathioprine; EDSS = Expanded Disability Status Scale; MOG = myelin oligodendrocyte glycoprotein; NA = not applicable; NMO = neuromyelitis optica; NMOSD = neuromyelitis optica spectrum disorder; NR = not reported; TCZ = tocilizumab.

^aThe systematic review included the interim analysis for the RCT due to the final results of this study have not been published, in this table, the RCT data from the final publication.²⁶

Table 3: Characteristics of Included Primary Clinical Studies

AZA = azathioprine; IQR = Interquartile range; logMAR = logarithm of the minimum angle of resolution; OCT = optical coherence tomography; ON = Optic neuritis; TCZ = tocilizumab.

Appendix 3: Critical Appraisal of Included Publications

Note that these tables have not been copy-edited.

Table 5: Strengths and Limitations of Systematic Reviews Using AMSTAR 2¹⁴

AMSTAR 2 = A MeaSurement Tool to Assess systematic Reviews 2; ISPOR = International Society for Pharmacoeconomics and Outcomes Research; NA = not applicable; NR = not reported. RoB = risk of bias; RCT = randomized controlled trial; NOS = Newcastle-Ottawa Scale.

Table 6: Strengths and Limitations of Clinical Studies Using the Downs and Black Checklist¹⁵

IQR = interquartile range; OCT = Optical Coherence Tomography; ORs = odds ratios; CI = confidence interval.

Table 7: Strengths and Limitations of Guideline Using AGREE II¹⁶

AGREE II = Appraisal of Guidelines for Research and Evaluation II; NA = not applicable; NR = not reported.

Appendix 4: Main Study Findings

Note that these tables have not been copy-edited.

Table 8: Summary of Findings by Outcome — Relapse Risk

AQP4-ab = aquaporin antibody; ARR = annualized relapse rate; AZA = azathioprine; CI = confidence interval; HR = hazard ratio; IQR = interquartile range; MD = mean difference; NA = not applicable; NR = not reported; OR = odds ratio; pGFAP = plasma glial fibrillary acidic protein; SMD = standardized mean difference, TCZ = tocilizumab.

^aThe systematic review included the interim analysis for the RCT as the final results of this study were not yet published. In this table, the RCT data are reported from the final publication.²⁶

Table 9: Summary of Findings by Outcome — Disability

CI = confidence interval; EDSS = Expanded Disability Status Scale; MD = mean difference; NA = not applicable; NR = not reported. SMD = standardized mean difference TCZ = tocilizumab.

Table 10: Summary of Findings by Outcome — Adverse Events

AZA = azathioprine; CI = confidence interval; DVT = deep vein thrombosis; NA = not applicable; NR = not reported; RCT = randomized controlled trial; SMD = standardized mean difference TCZ = tocilizumab; UTI = urinary tract infection.

^aThe systematic review included the interim analysis for the RCT as the final results of this study were not yet published. In this table, the RCT data are reported from the final publication.²⁶

Table 11: Summary of Findings by Outcome — Pain and Fatigue

IQR = interquartile range; NA = not applicable; NR = not reported; SD = standard deviation.

^aP values for comparisons were not available.

Table 12: Summary of Findings by Outcome — Optical Coherence Tomography Measures

CI = confidence interval; FLV = focal loss volume; mGCC = macular ganglion cell complex; NA = not applicable; NMOSD = neuromyelitis optica spectrum disorder; NR = not reported; NS = not statistically significant; pRNFL = peripapillary retinal nerve fibre layer thickness; SD = standard deviation; TCZ = tocilizumab; μm = micrometres.

Table 13: Summary of Findings by Outcome — Other outcomes

AQP4 = aquaporin 4; MRI = NA = not applicable; NR = not reported. SMD = standardized mean difference; TCZ = tocilizumab.

Table 14: Summary of Recommendations in Included Guidelines

IST = immunosuppressant treatment; NMOSD = neuromyelitis optica spectrum disorder.

Appendix 5: Overlap Between Included Systematic Reviews

Note that this table has not been copy-edited.

Table 15: Overlap in Relevant Primary Studies Between Included Systematic Reviews

^aLotan et al. 2021 also included 12 relevant articles that were not included in the other 3 systematic reviews and are therefore not presented in this table.

^bIncluded interim analysis presented in conference abstract.

^cConference abstract.

^dUpdated case series.

Appendix 6: References of Potential Interest

Previous CADTH Reports

Rituximab for the treatment of neuromyelitis optica spectrum disorder. (Health Technology Review). Ottawa (ON): CADTH: 2021 Feb 9: https://www.cadth.ca/rituximab-treatment-neuromyelitis-optica-spectrum-disorder. Accessed 2023 Mar 3.

Clinical review report: eculizumab (soliris). Alexion Pharma Canada Corp. Indication: neuromyelitis optica spectrum disorder. (CADTH Common Drug Review). Ottawa (ON): CADTH; 2020: https://www.cadth.ca/sites/default/files/cdr/clinical/sr0640-soliris-nmosd-clinical-review-report.pdf. Accessed 2023 Mar 3.

Clinical review report: satralizumab (enspryng): (Hoffmann-La Roche Limited). Indication: neuromyelitis optica spectrum disorder. Ottawa (ON): CADTH; 2021: https://www.cadth.ca/sites/default/files/cdr/clinical/sr0663-enspryng-clinical-review-report.pdf. Accessed 2023 Mar 3.

Systematic Reviews

Xie Q, Zheng T, Sun M, Sun J, Wang M. A meta-analysis to determine the efficacy and safety of tocilizumab in neuromyelitis optica spectrum disorders. Mult Scler Relat Disord. 2020;45:102421. PubMed

Kastrati K, Aletaha D, Burmester GR, et al. A systematic literature review informing the consensus statement on efficacy and safety of pharmacological treatment with interleukin-6 pathway inhibition with biological DMARDs in immune-mediated inflammatory diseases. RMD Open. 2022;8(2):09. PubMed

Randomized Controlled Trials

Zhang C, Zhang M, Qiu W, et al. Safety and efficacy of tocilizumab versus azathioprine in highly relapsing neuromyelitis optica spectrum disorder (TANGO): an open-label, multicentre, randomised, phase 2 trial. Lancet Neurol. 2020;19(5):391-401. PubMed

Additional References

Ringelstein M, Ayzenberg I, Lindenblatt G, et al. Interleukin-6 receptor blockade in treatment-refractory MOG-IgG-associated disease and neuromyelitis optica spectrum disorders. Neurol Neuroimmunol Neuroinflamm. 2022;9(1):01. PubMed

Rigal J, Pugnet G, Ciron J, Lepine Z, Biotti D. Off-label use of tocilizumab in neuromyelitis optica spectrum disorders and MOG-antibody-associated diseases: a case-series. Mult Scler Relat Disord. 2020;46:102483. PubMed

Lotan I, Charlson RW, Ryerson LZ, Levy M, Kister I. Effectiveness of subcutaneous tocilizumab in neuromyelitis optica spectrum disorders. Mult Scler Relat Disord. 2019;39:101920. PubMed

Du C, Zeng P, Han JR, et al. Early initiation of tocilizumab treatment against moderate-to-severe myelitis in neuromyelitis optica spectrum disorder. Front Immunol. 2021;12:660230. PubMed

Carreon Guarnizo E, Hernandez Clares R, Castillo Trivino T, et al. Experience with tocilizumab in patients with neuromyelitis optica spectrum disorders. Neurologia (Engl Ed). 2022;37(3):178-183. PubMed

Xie Q, Sun M, Sun J, Zheng T, Wang M. New progress in the treatment of neuromyelitis optica spectrum disorder with monoclonal antibodies (review). Exp Therapeutic Medicine. 2021;21(2):148. PubMed