Immunotherapy choice and maintenance for generalized myasthenia gravis in China

Abstract Aims To compare long‐term efficacy and safety of immunotherapeutic strategies as maintenance to prevent disease relapses of generalized myasthenia gravis (MG) in real‐world settings. Methods This is a retrospective cohort study on generalized MG conducted in seven major neurological centers across China. Eligible participants were patients with generalized MG who were under minimal manifestation status or better. Main outcome measures were probability of patients free of relapses and causes of drug discontinuation. Results Among 1064 patients enrolled, the median (interquartile range) age was 50.3 (37.0‐62.5) years and 641 (60.2%) were women. Disease relapse was significantly lower for rituximab (6.1%) compared with all the other monotherapies (hazard ratio [HR] = 0.18, 95% confidence interval [CI] 0.06 to 0.56, P = .0030). As combination therapies, tacrolimus in combination with corticosteroids reduced risk of disease relapses compared with azathioprine with corticosteroids (HR = 0.45, 95% CI 0.25 to 0.81, P = .0077) or mycophenolate mofetil with corticosteroids (HR = 0.32, 95% CI 0.15 to 0.67, P = .0020). Otherwise, lower‐dose corticosteroids or azathioprine as monotherapy significantly increased risk of disease relapses (HR = 2.78, 95% CI 1.94 to 3.99, P < .0001; HR = 2.14, 95% CI 1.42 to 3.23, P = .0003, respectively). The proportion of discontinuation was lowest in patients with rituximab (20.4%) as monotherapy and tacrolimus with corticosteroids (23.6%). Overall, combination treatment of immunosuppressants with corticosteroids had a lower rate of discontinuation compared with corresponding monotherapy (HR = 0.51, 95% CI 0.36 to 0.71, P < .0001). Conclusions Rituximab as monotherapy and tacrolimus with corticosteroids displayed better clinical efficacy as well as drug maintenance to prevent disease relapses in patients with generalized MG.


| INTRODUC TI ON
Myasthenia gravis (MG) is an autoantibody-mediated autoimmune disorder against postsynaptic membrane proteins at the neuromuscular junction, including acetylcholine receptor (AChR), muscle-specific kinase (MuSK), and lipoprotein receptor-related protein 4 (LRP4). 1 The bulk of generalized MG therapy include acetylcholinesterase inhibitors for symptomatic management as well as immunotherapeutic agents to inhibit autoimmunity. Most patients with generalized MG require induction therapy with glucocorticosteroids (steroids), or intravenous immunoglobulin (IVIG), plasma exchange (PE) for severe cases. And immunosuppressants are frequently used for maintenance. A repertoire of classical immunosuppressants, azathioprine, cyclophosphamide, mycophenolate mofetil, and tacrolimus have also been adopted. 2,3 Some efficacy of an anti-CD20 B-cell depleting agent, rituximab, is demonstrated by sustained clinical improvement with prolonged time to relapse and reduce the need of other immunosuppressants. [4][5][6][7] Treatments with these agents are commonly initiated in conjunction with steroids so that they may be gradually tapered to a lower maintenance dose, which varies among individual patients. 8 A previous network meta-analysis showed the efficacy and tolerability of the immunosuppressants were heterogeneous. 9 Presently, no consensus has been reached on an optimal therapeutic maintenance protocol in the prevention of relapses in generalized MG. 10 Using the MG registry of seven major centers throughout China, we analyzed the safety and efficacy of steroids and nonsteroidal immunotherapeutic agents in MG patients who had achieved Minimal Manifestation Status (MMS) or better during the long-term follow-up.

| Study design and patients
This retrospective cohort study was based on the registry databases established in 7 independent neurological centers across China.
Patient data were collected from August 9, 2013, to September 30, 2019, and data were censored thereafter. Institutional review boards approved this study at each participating center, and informed consent was obtained from each patient. Inclusion criteria were as follows: (a) age ≥ 18 years; (b) patients who had received prior efficacious treatments and achieved treatment goals, that is, MMS or better, classified by MGFA Task Force postintervention status (PIS); patient presents no symptoms or functional limitations from MG but has some weakness on examination of some muscles; (c) steroids were used alone or in conjunction with a nonsteroidal immunotherapeutic agent, and these regimens were maintained at a constant dose ( Figure 1).
Stable doses of different immunosuppressants were as follows: azathioprine 2-3 mg/kg/d, tacrolimus 2-5 mg/d, and mycophenolate mofetil 1-3 g/d. Lower-dose steroids was defined as <0.25 mg/kg prednisone or equivalent prednisolone/methylprednisolone daily. 11 Combined regimens were defined by combination of lower-dose steroids and an immunosuppressant at the aforementioned dosages.
Higher-dose steroids were defined as ≥0.25 mg/kg prednisone or equivalent prednisolone/methylprednisolone daily. Rituximab was administered to reduce the frequency of CD19 + B cells to less than 1% in peripheral blood mononuclear cells, as quantified by flow cytometry. 12 If steroids were contraindicated or refused, a singular nonsteroidal immunotherapeutic agent was used instead. In such circumstances, early combination with steroids could be withdrawn only after the nonsteroidal immunotherapeutic agent had reached its maximal effects without disease progression or relapse.
Exclusion criteria were defined: (a) patients who participated in randomized clinical trials with unknown treatment allocation; (b) lack of follow-up data; (c) patients who did well on immunotherapeutic treatments and discontinued treatments on their own. Patients were censored at treatment discontinuation regarding drug survival.
proportion of discontinuation was lowest in patients with rituximab (20.4%) as monotherapy and tacrolimus with corticosteroids (23.6%). Overall, combination treatment of immunosuppressants with corticosteroids had a lower rate of discontinuation compared with corresponding monotherapy (HR = 0.51, 95% CI 0.36 to 0.71, P < .0001).
Conclusions: Rituximab as monotherapy and tacrolimus with corticosteroids displayed better clinical efficacy as well as drug maintenance to prevent disease relapses in patients with generalized MG.

K E Y W O R D S
efficacy, real-world, relapse, rituximab, tacrolimus

| Diagnosis of relapses and refractory MG
Disease relapses, MGC, discontinuation or switching and stated reasons, and adverse events were recorded. Relapse of MG was defined when a 3-point change or more was identified in MG-Composite (MGC) quantitative measure. 13 Refractory MG was defined as that postintervention status remained unchanged or worse after corticosteroids and at least 2 other immunosuppressant agents, used in adequate doses for an adequate duration, with persistent symptoms or side effects that limit functioning. 14

| Data collection and processing
The main clinical features of the disease were recorded: the subtypes of generalized MG onset according to the Myasthenia Gravis Demographic data (sex, date of birth, age at onset) as well as outcomes of assays for anti-AChR, MuSK, and LRP4 were also recorded.
Testing for low-affinity anti-AChR antibody was not available. Upon onset of generalized MG, all patients received induction treatments with high-dose steroids, and/or IVIG, PE to reach MMS status, which was maintained by steroids and/or immunosuppressants to prevent relapses. Steroids were tapered to stable doses for concomitant immunosuppressants or withdrawn after immunosuppressants started to exert an apparent maximal effect, based on clinicians' experience.
Adverse events were graded according to the National Cancer Institute's Common Terminology Criteria for Adverse Events (CTCAE 5.0).
Primary outcome was the probability of patients free of relapses. The baseline time was when the patients had MMS or better and concomitantly received stable dosage of various regimens.
Secondary outcomes included discontinuation of therapy, for any reason including disease relapses, pregnancy, and adverse events.
All outcomes were prespecified prior to data analysis. The assessment of efficacy of immunotherapeutic agents as monotherapy commenced upon treatment duration without concomitant steroids.

| Statistical analyses
Descriptive statistics of study cohort was performed. Continuous variables were expressed with median (IQR). Fisher's exact test was adopted to compare the categorical variables of age, gender, and previous MGFA subtype. Shapiro-Wilk method was used for normality test. Data that did not exhibit a normal/Gaussian distribution were analyzed via the nonparametric Kruskal-Wallis test. Kaplan-Meier curves and Cox proportional hazards models were used to visualize and compare the probability of disease relapse and drug F I G U R E 1 Treatment Strategies Before and After Study Inclusion. At disease onset, there were 2 protocols of corticosteroids usage to achieve remission or MMS. Patients received corticosteroids at high doses (1-1.5 mg/kg/d) and maintained 1-3 months to improve symptoms. The other patients resumed gradual corticosteroids escalation, starting with a dose of 10-20 mg/d and increased by 10 mg per week until the main symptoms improved. Then corticosteroids tapered. Immunosuppressants that aimed to reduce corticosteroids were added before MMS until stable doses were attained. Once the patients met criteria of inclusion, corticosteroids and immunosuppressants were used alone or in combination. The follow-up was ended when the patients discontinued stable maintenance treatments due to disease relapses, adverse events or pregnancy survival over timescale from first drug administration to outcome of interest endpoint date. Potential confounding variables: age, sex, historical MGFA subtypes, serostatus of auto-antibodies and followup time were examined and adjusted via sequential regression models. Cumulative incidences were estimated to compare the efficacy and rationale for therapy discontinuation over time between drug categories. To compare predictors between different treatments, we performed post hoc analyses for each treatment subgroup. Mixed effect Cox proportional hazards model was used to adjust the confounding factors, and centers were adjusted as a random effect factor. Statistical analyses and data processing were performed in R, Significance was set at a P value of less than .05.  (Table 1).

| Study population
There were no differences in the protocol and execution of followups between centers and treatments.     Table 2 and Figure 3A).

| MG relapses
As combined therapy, tacrolimus with lower-dose steroids had lower risks of relapses than azathioprine with lower-dose steroids or mycophenolate mofetil with lower-dose steroids (HR = 0.45, 95% CI 0.25 to 0.81, P = .0077; HR = 0.32, 95% CI 0.15 to 0.69, P = .0020, respectively). There was no significant difference in risk reduction between azathioprine with lower-dose steroids and mycophenolate mofetil with lower-dose steroids (HR = 1.19, 95% CI 0.69 to 2.05, P = .5297; Figure 3B). We identified no significant differences in disease relapse or drug discontinuation between anti-AChR positive and seronegative cohorts.

| Predictors of relapse risk
The analysis for independent treatment response predictors re- Neither diabetes mellitus nor gender was associated with higher relapse risk ( Figure S1).

| Association of relapses with thymoma types under different treatments
The efficacy of study regimens in generalized MG patients with thymoma was also analyzed. B2/B3/C thymoma types are more aggressive and an independent risk factor for postoperative myasthenic progression or crisis in MG patients. Therefore, they were separately  Table 2). However, similar efficacy was seen in the treatment groups of combined therapies as well as rituximab as monotherapy, irrespective of thymoma subtype.

| Drug discontinuation
The   Figure 5C, D, and E). Overall, lower drug discontinuation rates were found in the patients with tacrolimus or rituximab ( Figure 5F and G).
The adverse events varied in each treatment group, but most were CTCAE grade 1. Hyperglycemia, liver dysfunction, and infection were the most adverse events in our cohort (Table S1).
A higher percentage of patients receiving mycophenolate mofetil with lower-dose steroids (12.7%) and mycophenolate mofetil as monotherapy (8.1%) discontinued treatment due to pregnancy. The lowest discontinuation rates by pregnancy were recorded in patients on azathioprine with lower-dose steroids (1.5%) and tacrolimus with lower-dose steroids (1.8%).

| Treatments for refractory cases
In this study population, 75 patients (7.0%) met the criteria for refrac-

| D ISCUSS I ON
Patients with generalized MG may be at a high risk of relapses if they discontinued immunotherapeutic treatments. 16  Similarly, patients who received cyclosporine or methotrexate with daily lower-dose steroids were also not enrolled due to small number. In addition, though initial therapy choice and the threshold for switching therapies were at the treating physicians' discretion, patient-oriented features such as price considerations were not deliberated in this study. In regard to change of therapy for efficacy reasons, our data support that such switches were prompted by objective evidence of disease relapses, manifested by significant changes in QMC scale. However, switching owing to safety or tolerability issues cannot be demonstrated in objective terms. Systemic recording of adverse events, as in prospective clinical trials, is lacking in the present study due to practical considerations of reporting.
Lastly, newer immunotherapeutic agents, such as eculizumab, were not included in the study.

| CON CLUS IONS
Patients who had attained MMS or better and then received stable treatments as maintenance demonstrate that regimens of rituximab as monotherapy and tacrolimus with lower-dose steroids displayed improved efficacy and tolerability compared with other common immunotherapeutic regimens. Further studies are needed to clarify long-term head-to-head comparative safety and effective outcomes of these different immunotherapies in large scale populations.

ACK N OWLED G M ENTS
The study was supported by grants from National Science Foundation Center for Neuroinflammation for support; Samuel Shi for editorial assistance.

CO N FLI C T O F I NTE R E S T
The authors have no conflicts of interest to declare.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets generated and analyzed during the current study are available from the corresponding author on reasonable requests.