Durability of the Rituximab Response in Acetylcholine Receptor Autoantibody–Positive Myasthenia Gravis

Key Points

Question  Is the rituximab response in treatment-refractory acetylcholine receptor autoantibody–positive myasthenia gravis (AChR+ MG) durable?

Findings  In this case series study of 16 patients with AChR+ MG who were treated with rituximab and followed up for 18 to 84 months, all patients were observed to have clinical improvement. Nine patients had a relapse within a mean of 36 months after the last treatment cycle; the remaining 7 continued to maintain clinical benefit during a mean follow-up of 47 months.

Meaning  Rituximab appears to be an effective option with sustained long-term benefit after treatment in patients with refractory AChR+ MG.

Importance  Myasthenia gravis (MG), an autoimmune disorder of neuromuscular transmission, is treated by an array of immunotherapeutics, many of which are nonspecific. Even with current therapies, a subset of patients has medically refractory MG. The benefits of B-cell–targeted therapy with rituximab have been observed in MG; however, the duration of these benefits after treatment is unclear.

Objective  To evaluate the durability of response to rituximab in the treatment of acetylcholine receptor autoantibody–positive (AChR+) generalized MG.

Design, Setting and Participants  This retrospective case series study included 16 patients with AChR+ MG referred to an MG clinic from January 1, 2007, to December 31, 2015. The patients were treated with rituximab and followed up for 18 to 84 months after treatment.

Main Outcomes and Measures  Assessment of long-term clinical response, durability of response and/or relapse rate, AChR autoantibody levels, adverse effects, and inflammatory markers.

Results  In the 16 patients (6 men and 10 women; median age, 42 [range, 18-69] years), clinical improvement was observed in parallel with complete withdrawal or reduction of other immunotherapies, with all patients achieving complete stable remission, pharmacologic remission, or minimal manifestations based on the Myasthenia Gravis Foundation of America postintervention status criteria. Nine patients (56%) had a relapse during a mean follow-up of 36 (range, 24-47) months. Seven patients (44%) remained relapse free with a mean follow-up of 47 (range, 18-81) months since the last rituximab treatment. All values were normalized to a pretreatment anti-AChR antibody level of 100% and the mean levels after each rituximab cycle were calculated. A 33% decrease was seen after cycle 1 of rituximab treatment (100% vs 67%; P = .004); 20% after cycle 2 (compared with cycle 1) (67% vs 47%; P = .008); and 17% after cycle 3 (compared with cycle 2) (47% vs 30%; P = .02). However, the serum cytokine levels measured were found to be unchanged.

Conclusions and Relevance  Rituximab therapy appears to be an effective option in patients with refractory AChR+ MG, who were observed to have a durable response after treatment. Identification of markers of disease relapse and sustained remission are critical next steps in the development of pathophysiology-relevant, evidence-based practice parameters for rituximab in the treatment of MG.

Myasthenia gravis (MG) is an autoimmune disorder affecting neuromuscular transmission with an estimated annual incidence of about 1.7 to 21.3 cases per 1 million person-years and prevalence as high as 15 to 179 per 1 million persons.¹Quiz Ref ID Control of symptoms can be initially achieved with acetylcholinesterase inhibitors; however, most patients require immunotherapy such as corticosteroids, azathioprine, cyclosporine, mycophenolate mofetil, plasma exchange, and intravenous immunoglobulin at some point in their disease course to achieve sustained clinical remission.^2-4 Thymectomy, regardless of the presence of a thymoma, is also considered a therapeutic option based on prior clinical experience; however, results and conclusions from the recently completed placebo-controlled thymectomy trial in nonthymomatous MG are pending at this time.⁵ Despite these therapeutic options, a subset of patients has medically refractory MG or intolerable adverse effects from medication.⁶

Autoreactive B cells have a clear pathogenic role in the development of MG, and B-cell–directed therapy has emerged as a highly effective tool in managing other autoimmune diseases such as rheumatoid arthritis and neuromyelitis optica.^7,8Quiz Ref ID Several groups^9-18 have also observed the benefits of rituximab, a chimeric anti-CD20 monoclonal antibody, in patients with MG. In addition to significant clinical improvement, rituximab also allowed for tapering and subsequent discontinuation of other immunotherapies in patients with acetylcholine receptor autoantibody–positive (AChR+) and muscle-specific kinase autoantibody–positive MG.⁹ However, the duration of clinical remission among patients treated with rituximab while not receiving other immunotherapy, especially considering the prior refractory nature of their disease, is unclear. The number of rituximab treatment cycles necessary to achieve long-term remission also remains unknown. Herein, we report our experience with the long-term effects of rituximab in 16 patients with refractory AChR+ MG who were followed up for 18 to 84 months, representing to our knowledge the longest follow-up of a single MG cohort to date.

This retrospective study included patients with generalized MG referred to the Yale Myasthenia Gravis Clinic, New Haven, Connecticut, from January 1, 2007, to December 31, 2015. Sixteen patients with AChR+ refractory generalized disease and a minimum of 12 months of follow-up after completion of the initial set of rituximab treatment cycles were identified (Table). This study was approved by the institutional review board of Yale University as part of an observational study examining the treatment and disease course of MG. All patients provided written informed consent.

Disease was defined as refractory when the immunotherapy dosage could not be lowered without clinical relapse, inadequate clinical control of the disease was achieved during the immunotherapy regimen, or severe adverse effects due to current immunosuppressive therapy were present. Pretreatment and posttreatment immunotherapy regimens, clinical symptoms, and examination findings were evaluated. The Myasthenia Gravis Foundation of America (MGFA) clinical classification criteria¹⁹ and postintervention status were used to assign clinical state a minimum of 12 months after completion of the initial set of rituximab cycles. The number of administered rituximab treatment cycles, time since the last treatment cycle, and times to relapse and postrelapse treatments were reviewed. In addition, anti-AChR antibody levels, measured by conventional radioimmunoassays (reference value, ≤0.02 nmol/L; Mayo Medical Laboratories), were assessed in patients before initiation of rituximab therapy, at the end of each cycle, at the time of clinical relapse, and at the last follow-up. Statistical analysis was performed using GraphPad Prism (GraphPad Software), and P values were calculated using nonparametric Friedman and Wilcoxon matched-pairs signed rank tests.

Because no established infusion protocol for rituximab use in MG currently exists, we used a standard protocol adopted from the non–Hodgkin lymphoma regimen of 4 weekly infusions of 375 mg/m². One cycle is defined as 1 infusion per week for 4 consecutive weeks. The interval between cycles was 6 months. Infusions were completed per protocol in the outpatient infusion center at our institution. Our patients were treated with an initial 2- to 4-cycle regimen. The number of cycles was mainly based on reaching a symptom-free state and patient toleration of tapering or withdrawal of other immunotherapies (ie, corticosteroids). The number of rituximab treatment cycles or interval between cycles was not dictated by B-cell counts.

All prior immunotherapies were reviewed. These included prednisone, which is the standard first-line agent; plasma exchange; azathioprine; mycophenolate mofetil; and intravenous immunoglobulin.

In addition to clinical follow-up, we reviewed the infusion center notes and monitored complete blood cell counts and liver function test profiles to evaluate the safety profile. Periodic measurement of B-cell counts was performed per best medical practice after completion of the initial set of cycles from a safety perspective (ie, safety of vaccinations, etc). However, owing to the retrospective nature of this work, we did not have complete longitudinal data on B-cell counts for analysis.

Preinfusion and postinfusion serial blood samples were collected from 4 patients and 10 healthy control individuals after obtaining informed consent. Serum was obtained by centrifugation of whole-blood samples and was cryopreserved at −80°C. A fluorescent multiplexed magnetic bead–based screening assay (R&D Systems, Inc) was performed in accordance with the manufacturer’s protocol for the following 10 cytokines: interleukin 4 (IL-4), IL-5, IL-6, IL-10, IL-17A, IL-17F, tumor necrosis factor, interferon γ, vascular endothelial growth factor, and resistin. The samples were analyzed in duplicate and diluted 2-fold. Results were expressed as a mean value in pictograms per milliliter. We used the Mann-Whitney test to determine statistical significance (P < .05).

Of the 16 patients in the study (6 men and 10 women; median age, 42 [range, 18-69] years), 15 were receiving prednisone before initiating rituximab therapy (Table). Eight patients were also receiving a corticosteroid-sparing agent (azathioprine in 6 and mycophenolate mofetil in 2). Quiz Ref IDThirteen had undergone thymectomy (thymoma [5 patients], thymic hyperplasia [2 patients], and normal thymus [6 patients]). Eight patients were treated with 2 cycles; 7 patients, with 3 cycles; and 1 patient, with 4 cycles. A change in clinical status to improved was observed in parallel with complete withdrawal or reduction of other immunotherapies with all patients achieving complete stable remission, pharmacologic remission, or minimal manifestations (MM) based on the MGFA postintervention status criteria. After completing the initial set of rituximab treatment cycles, 10 patients (63%) achieved complete stable remission; 3 patients (19%), pharmacologic remission (with azathioprine in patient 10 and with prednisone in patients 11 and 13). The remaining 3 patients (19%) achieved MM-0 (minimal manifestations but no therapy for MG). The 13 patients (with complete stable remission and MM-0) who were able to taper and discontinue all other immunotherapies were able to do so in a mean of 8.3 (range, 1-15) months since the last infusion.

No infusion reactions were seen. One patient developed leukopenia (white blood cell count, 2700/µL [to convert to ×10⁹ per liter, multiply by 0.001]) after the second cycle, but this resolved without intervention. Treatment had to be stopped in 1 patient owing to an unplanned pregnancy during the second cycle. She went on to have an uncomplicated pregnancy and delivery.

Nine of 16 patients (56%) experienced a relapse in a mean of 36 (range, 24-47) months after the last rituximab treatment cycle (Figure 1). The 4 patients who had received 2 cycles had a relapse within a mean follow-up of 33 months. The 4 patients who had received 3 cycles had a relapse within a mean follow-up of 36 (range, 29-44) months. One patient received 4 cycles and had a relapse at 47 months. All of these patients improved again after further immunosuppression therapy (intravenous immunoglobulin or plasma exchange in 7; high-dose prednisone in 1; and an additional cycle of rituximab in 4; some received more than 1 treatment for relapse). The MGFA postintervention status at their most recent follow-up was pharmacologic remission (n = 7) or MM-1 (MM with some form of immunotherapy for MG) (n = 2).

Seven of 16 patients (44%) remained clinically stable with follow-up ranging from 18 to 81 months (mean follow-up, 47 months) (Figure 1). The MGFA postintervention status at their most recent follow-up was complete stable remission (n = 5), pharmacologic remission (n = 1), or MM-0 (n = 1).

Because we observed a mean time to relapse of 36 months, we compared durability of benefit in patients followed up for more than 48 months (12 patients) and 48 months or less (4 patients) after completion of the initial treatment regimen. Of those patients with follow-up of more than 48 months, 8 of 12 patients had a relapse in a mean follow-up of 37 (range, 29-47) months. The remaining 4 patients did not have a relapse, with a mean follow-up period of 66 (range, 51-81) months. Of those patients with a follow-up of 48 months or less, 1 of 4 had a relapse at 24 months. The remaining 3 patients did not have a relapse with a mean follow-up period of 22 (range, 18-24) months.

We also reviewed the time from diagnosis to initiation of first treatment with rituximab in our cohort, which was a mean of 36 (range, 9-90) months. The mean duration of disease before treatment in the relapse group was 31 (range, 9-61) months; in the nonrelapse group, 41 (range, 10-90) months (Figure 1). We observed no difference in the time from diagnosis to initiation of rituximab therapy and response durability based on these data.

A total of 13 patients in our cohort had a thymectomy. Among the 6 patients who underwent thymectomy less than 12 months before starting treatment, 4 had a relapse in a mean time of 36 (range, 29-44) months. Among the 7 patients who underwent thymectomy more than 12 months before starting treatment, 4 had a relapse in a mean time of 35 (range, 24-47) months. The relapse rate was 67% in the group with thymectomy less than 12 months before rituximab treatment (mean, 6.2 months) and 57% in the group with thymectomy more than 12 months before rituximab treatment (mean, 34 months). The relapse rates appear similar between these 2 groups. Only 1 of 3 patients who did not undergo thymectomy had a relapse at 39 months. However, no firm conclusions can be drawn from such a small number of patients.

The following statistically significant decreases in anti-AChR antibody levels was observed after treatment with the initial set of rituximab cycles (Figure 2): 33% after cycle 1 (100% vs 67%; P = .004); 20% after cycle 2 (compared with cycle 1) (67% vs 47%; P = .008); and 17% after cycle 3 (compared with cycle 2) (47% vs 30%; P = .02). In the patients who did not have a relapse (eFigure 1A in the Supplement), a sustained decrease in anti-AChR antibody level was noted until their last follow-up (mean antibody level before rituximab treatment, 24.05 nmol/L; after last cycle, 10.8 nmol/L; and at last follow-up, 15.65 nmol/L; P = .01). In the patients who had a relapse after initial cycles of rituximab treatment (eFigure 1B in the Supplement), no significant difference was noted in the anti-AChR antibody levels at relapse (mean antibody level after last cycle, 2.72 nmol/L; at the time of relapse, 2.87 nmol/L; P = .22). However, 1 patient had a dramatic rise in anti-AChR antibody level at the time of a second relapse that occurred in the setting of a new diagnosis of stage IV adenocarcinoma with an unknown primary site of origin.

Levels of 8 of the 10 cytokines measured (IL-4, IL-5, IL-6, IL-10, IL-17A, IL-17F, tumor necrosis factor, and interferon γ) were below the level of detection for the assay, implying that their levels in serum were not elevated. No statistically significant differences in serum levels of resistin and vascular endothelial growth factor were observed between healthy donor and patient serum samples before the initiation of rituximab treatment (eFigure 2 in the Supplement). We also found no appreciable change in the concentrations of resistin and vascular endothelial growth factor with rituximab treatment or after achieving a state of clinical remission, during longitudinal follow-up. One of the 4 patients studied experienced a relapse, but there was no significant difference observed in the levels of these 2 cytokines before and after relapse.

The need for additional treatments, particularly for patients who do not respond to or have intolerable adverse effects from existing immunosuppressive therapy, has led to an interest in targeted immunotherapies. Rituximab is an appealing choice owing to its B-cell–targeting mechanism of action and the precedence for its use in the treatment of other autoimmune diseases, such as rheumatoid arthritis.^7,20,21

Quiz Ref IDIn this retrospective analysis of 16 patients with refractory AChR+ MG, rituximab appears to have a durable response. Our results support the hypothesis that rituximab can be helpful in managing refractory MG. These findings are in agreement with previous reports of its benefit.^9,10,22,23

The number of rituximab treatment cycles necessary to achieve disease remission has been unclear. A minimum of 2 cycles appear to be needed, because most of the patients in our cohort required approximately 1 year to taper other therapies. As observed in previous studies,¹⁰ some patients may need additional cycles. Although patients treated with more cycles tended to have a longer-lasting response, our sample size in each group is too small to draw any firm conclusions. To date, guidelines on when to discontinue or repeat rituximab treatment are yet to be established. As a practical matter, patients with evidence of clinical disease relapse and a minimum of 6 months since the last cycle should be considered for retreatment in the absence of medical contraindications.

The relapse rate in our cohort was 56%, typically occurring about 3 years after the last rituximab treatment cycle. This relapse rate is similar to those of previously reported cohorts,¹⁰ but after a longer duration of disease stability. After an induction regimen, a mean time to relapse of 17 (ranging, 6-34) months was observed in an independent study.¹⁰ The patients with relapse in our cohort were able to achieve clinical improvement again after treatment with further immunosuppression. Thymectomy is certainly a possible confounder. Acknowledging the limitation of our sample size, thymectomy status, timing of thymectomy or pathologic findings in the thymus did not seem to influence disease relapse or durability of response in our cohort.

Biomarkers would be very helpful in guiding clinicians as to whom to offer additional cycles. Anti-AChR antibody levels can be helpful in assessing the response to treatment because these levels were noted to decrease after the administration of rituximab. However, their role in predicting relapse is less clear because we did not note any significant increase in the levels at the time of relapse, although our small sample size limits definitive conclusion.

Our analysis of 10 cytokines that have been associated with the immunopathogenesis of MG^24,25 did not reveal any appreciable changes with B-cell depletion, clinical remission, or relapse. Acknowledging the limitation of sample size, further studies are needed to attribute value to these cytokines as biomarkers and to identify other indicators of disease activity and response to therapy.

The effects of rituximab on B cells as well as putative T-cell–mediated immune dysregulation in MG need further investigation. We plan to apply recently developed assays^26,27 to the prospective clinical trial of rituximab in MG currently under way.²⁸

All patients followed up in this study tolerated rituximab with no severe hematologic derangements or infusion reactions. Although the most common adverse effect is an infusion reaction, progressive multifocal leukoencephalopathy is also of concern after rituximab therapy^29,30; however, the relative risk is thought to be low.³¹ A recent case report³² has described the occurrence of progressive multifocal leukoencephalopathy in a patient with seronegative MG, having been treated with rituximab in addition to prednisone, azathioprine, mycophenolate mofetil, intravenous immunoglobulin, and plasma exchange at different times during the course of disease. What role aggressive, long-term immunosuppression therapy had in this case is unclear. Nevertheless, clinical monitoring per best medical practice standards and minimizing combination immunosuppressive regimens is required and strongly advised when considering the initiation of rituximab therapy.

We found B-cell depletion therapy to be an effective option with sustained long-term benefit after treatment in patients with refractory AChR+ MG. This study represents, to our knowledge, one of the largest single-center studies with extended long-term follow-up. A prospective, placebo-controlled clinical trial is currently under way that will further help to evaluate the efficacy, safety, and pharmacodynamics of rituximab in MG.²⁸Quiz Ref ID Identification of markers of disease activity, responsiveness to therapy, clinical relapse, and remission are critical next steps in the development of evidence-based practice parameters for rituximab in the treatment of MG as well as other potential target therapeutics.

Corresponding Author: Richard J. Nowak, MD, MS, Program in Clinical and Translational Neuromuscular Research, Division of Neuromuscular Medicine, Department of Neurology, Yale University School of Medicine, PO Box 208018, New Haven, CT 06520 (richard.nowak@yale.edu).

Accepted for Publication: August 26, 2016.

Published Online: November 21, 2016. doi:10.1001/jamaneurol.2016.4190

Author Contributions: Drs Robeson and Kumar contributed equally to this study and are co–first authors. Dr Nowak had full access to all the data in the study and takes full responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Robeson, Kumar, Keung, Goldstein, O’Connor, Nowak.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Robeson, Kumar, Nowak.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Robeson, Kumar, Nowak.

Study supervision: Nowak.

Conflict of Interest Disclosures: Dr Nowak reports receiving grant U01NS084495-01A1 from the National Institute of Neurological Disorders and Stroke (NINDS), National Institutes of Health (NIH), to conduct the currently underway prospective phase 2 trial of rituximab in myasthenia gravis along with drug and placebo provided by Genentech through an investigator-initiated trial agreement, which is separate from the research in this article. Dr O’Connor reports receiving honoraria (speaking fees) from Genentech. No other disclosures were reported.

Funding/Support: Dr Nowak was supported in part, by award U01NS084495-01A1 from the NINDS of the NIH. This study was supported, in part, by grant R01AI114780 from the National Institute of Allergy and Infectious Diseases of the NIH (Dr O’Connor).

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.