Safety and efficacy of eculizumab in anti-acetylcholine receptor antibody-positive refractory generalised myasthenia gravis (REGAIN): a phase 3, randomised, double-blind, placebo-controlled, multicentre study

Articles

Safety and efficacy of eculizumab in anti-acetylcholine receptor antibody-positive refractory generalised myasthenia gravis (REGAIN): a phase 3, randomised, doubleblind, placebo-controlled, multicentre study James F Howard Jr, Kimiaki Utsugisawa, Michael Benatar, Hiroyuki Murai, Richard J Barohn, Isabel Illa, Saiju Jacob, John Vissing, Ted M Burns, John T Kissel, Srikanth Muppidi, Richard J Nowak, Fanny O’Brien, Jing-Jing Wang, Renato Mantegazza, in collaboration with the REGAIN Study Group*

Summary

Background Complement is likely to have a role in refractory generalised myasthenia gravis, but no approved therapies specifically target this system. Results from a phase 2 study suggested that eculizumab, a terminal complement inhibitor, produced clinically meaningful improvements in patients with anti-acetylcholine receptor antibody-positive refractory generalised myasthenia gravis. We further assessed the efficacy and safety of eculizumab in this patient population in a phase 3 trial. Methods We did a phase 3, randomised, double-blind, placebo-controlled, multicentre study (REGAIN) in 76 hospitals and specialised clinics in 17 countries across North America, Latin America, Europe, and Asia. Eligible patients were aged at least 18 years, with a Myasthenia Gravis-Activities of Daily Living (MG-ADL) score of 6 or more, Myasthenia Gravis Foundation of America (MGFA) class II–IV disease, vaccination against Neisseria meningitides, and previous treatment with at least two immunosuppressive therapies or one immunosuppressive therapy and chronic intravenous immunoglobulin or plasma exchange for 12 months without symptom control. Patients with a history of thymoma or thymic neoplasms, thymectomy within 12 months before screening, or use of intravenous immunoglobulin or plasma exchange within 4 weeks before randomisation, or rituximab within 6 months before screening, were excluded. We randomly assigned participants (1:1) to either intravenous eculizumab or intravenous matched placebo for 26 weeks. Dosing for eculizumab was 900 mg on day 1 and at weeks 1, 2, and 3; 1200 mg at week 4; and 1200 mg given every second week thereafter as maintenance dosing. Randomisation was done centrally with an interactive voice or webresponse system with patients stratified to one of four groups based on MGFA disease classification. Where possible, patients were maintained on existing myasthenia gravis therapies and rescue medication was allowed at the study physician’s discretion. Patients, investigators, staff, and outcome assessors were masked to treatment assignment. The primary efficacy endpoint was the change from baseline to week 26 in MG-ADL total score measured by worst-rank ANCOVA. The efficacy population set was defined as all patients randomly assigned to treatment groups who received at least one dose of study drug, had a valid baseline MG-ADL assessment, and at least one post-baseline MG-ADL assessment. The safety analyses included all randomly assigned patients who received eculizumab or placebo. This trial is registered with ClinicalTrials.gov, number NCT01997229. Findings Between April 30, 2014, and Feb 19, 2016, we randomly assigned and treated 125 patients, 62 with eculizumab and 63 with placebo. The primary analysis showed no significant difference between eculizumab and placebo (leastsquares mean rank 56·6 [SEM 4·5] vs 68·3 [4·5]; rank-based treatment difference −11·7, 95% CI −24·3 to 0·96; p=0·0698). No deaths or cases of meningococcal infection occurred during the study. The most common adverse events in both groups were headache and upper respiratory tract infection (ten [16%] for both events in the eculizumab group and 12 [19%] for both in the placebo group). Myasthenia gravis exacerbations were reported by six (10%) patients in the eculizumab group and 15 (24%) in the placebo group. Six (10%) patients in the eculizumab group and 12 (19%) in the placebo group required rescue therapy. Interpretation The change in the MG-ADL score was not statistically significant between eculizumab and placebo, as measured by the worst-rank analysis. Eculizumab was well tolerated. The use of a worst-rank analytical approach proved to be an important limitation of this study since the secondary and sensitivity analyses results were inconsistent with the primary endpoint result; further research into the role of complement is needed. Funding Alexion Pharmaceuticals.

Introduction Refractory generalised myasthenia gravis is a chronic, debilitating, rare disorder of severe muscle weakness

resulting from autoantibody-mediated destruction of the neuromuscular junction. Patients with refractory gen­ eralised myasthenia gravis, representing

www.thelancet.com/neurology Published online October 20, 2017 http://dx.doi.org/10.1016/S1474-4422(17)30369-1

Lancet Neurol 2017 Published Online October 20, 2017 http://dx.doi.org/10.1016/ S1474-4422(17)30369-1 See Online/Comment http://dx.doi.org/10.1016/ S1474-4422(17)30363-0 *All members of the Group are listed in the appendix Department of Neurology, University of North Carolina, Chapel Hill, NC, USA (Prof J F Howard Jr MD); Department of Neurology, Hanamaki General Hospital, Hanamaki, Japan (K Utsugisawa MD); Department of Neurology, University of Miami Miller School of Medicine, Miami, FL, USA (Prof M Benatar MD); Department of Neurology, International University of Health and Welfare, Narita, Japan (Prof H Murai MD); Department of Neurology, University of Kansas Medical Center, Kansas City, KA, USA (Prof R J Barohn MD); Neurology Department, Hospital Sant Pau, Universitat Autònoma Barcelona, Barcelona, Spain (Prof I Illa MD); Queen Elizabeth Neuroscience Centre, Wellcome Trust Clinical Research Facility, University Hospitals of Birmingham, Birmingham, UK (S Jacob MD); Rigshospitalet, Department of Neurology, University of Copenhagen, Copenhagen, Denmark (Prof J Vissing MD); Department of Neurology, University of Virginia Health System, Charlottesville, VA, USA (Prof T M Burns MD); Department of Neurology, Ohio State University, Columbus, OH, USA (Prof J T Kissel MD); Department of Neurology and Neurosciences, Stanford University School of Medicine, Stanford, CA, USA (S Muppidi MD); Department of

1

Articles

Neurology, Yale University School of Medicine, New Haven, CT, USA (R J Nowak MD); Alexion Pharmaceuticals, New Haven, CT, USA (F O’Brien PhD, J-J Wang MD); and Fondazione IRCCS Istituto Neurologico Carlo Besta, Milan, Italy (R Mantegazza MD) Correspondence to: Prof James F Howard Jr, Department of Neurology, University of North Carolina, Chapel Hill, NC, 27599−7025, USA [email protected]

Research in context Evidence before this study We searched MEDLINE, Embase, and PubMed databases up to July 23, 2017, for relevant clinical studies in refractory myasthenia gravis on the use of eculizumab, a terminal complement inhibitor, with no date or languge restrictions. The search terms used were “randomised controlled trial”, “eculizumab”, and “refractory myasthenia gravis”. We did not identify any reports of masked, placebo-controlled, randomised studies testing the use of eculizumab in refractory myasthenia gravis, except for a pilot phase 2 study of eculizumab in patients with anti-acetylcholine receptor antibody-positive refractory generalised myasthenia gravis published in 2013 that was funded by the same source as this study. Added value of this study Current immunosuppressant therapies have not been rigorously assessed in the setting of a prospective, randomised, placebo-controlled, clinical, phase 3 study in patients with refractory generalised myasthenia gravis.

approximately 10–15% of all patients with myasthenia gravis, do not respond to long-term treatment with corticosteroids or multiple steroid-sparing immuno­ suppressive treatments, or they have intolerable sideeffects to these therapies or require ongoing treatment with either intravenous immuno­ globulin or plasma exchange.1,2 This hetero­ geneous patient population continues to have disease symptoms and persistent morbidities, despite substantial use of immuno­ suppressive treatments, including difficulties with speech, swallowing, and mobility, impairment of respiratory function, and extreme fatigue, with substantial negative effects on activities of daily living and quality of life.1,3 Patients with refractory generalised myasthenia gravis might also have frequent exacerbations, which can be life-threatening and require admission to hospital or intensive care, and cause episodes of respiratory failure that require mechanical ventilation. A recent US-based claims analysis reported that the frequency of myasthenic gravis crises and exacerbations, admissions to hospital, and emergency room visits was three times higher in patients with refractory generalised myasthenia gravis than in patients in the overall generalised myasthenia gravis population.4 Approximately 74–88% of patients with myasthenia gravis produce autoantibodies to the acetylcholine receptor.5–10 These autoantibodies are directly pathogenic at the postsynaptic membrane of the neuromuscular junction by initiating several processes, including accelerated endocytosis and degradation of acetylcholine receptors, and complement-mediated membrane damage and inflammation.11–16 The role of complement in damage to the neuro­ muscular junction is supported by animal models of 2

REGAIN is the only study in patients with anti-acetylcholine receptor antibody-positive refractory generalised myasthenia gravis. Eculizumab was well tolerated for more than 26 weeks, with few adverse events. The primary endpoint of the Myasthania Gravis-Activities of Daily Living score did not differ significantly between patients who received eculizumab and those who received placebo. Additional prespecified sensitivity and secondary analyses suggested efficacy of eculizumab. Implications of all the available evidence This study used several validated myasthenia gravis-specific, physician-reported and patient-reported outcome measures to capture a detailed picture of the patient population with refractory myasthenia gravis and the effects of eculizumab therapy. The data suggest that complement inhibition might be an approach to the treatment of anti-acetylcholine receptor antibody-positive refractory generalised myasthenia gravis, and should stimulate additional research.

myasthenia gravis in which inhibition of formation of the membrane attack complex or use of complementknockout animals has been shown to preserve the function of neuromuscular junctions in induced or passively transferred myasthenia gravis.17 Additionally, evidence showing that functional blockade of the complement protein C5 protects against severe disease in preclinical studies18,19 suggests that complement inhibition might be a potential therapeutic approach for myasthenia gravis. Further support for the role of complement is derived from clinical studies in which complement activation proteins were identified at the site of damage to neuromuscular junctions, and the presence of the membrane attack complex correlated with the site of the damage in patients with myasthenia gravis.20–22 There are currently no approved therapies for refractory generalised myasthenia gravis that specifically target the complement system; inhibition of terminal complement activation represents a novel, targeted approach that might prevent damage at the postsynaptic membrane of the neuromuscular junction in patients with anti-acetylcholine receptor antibody-positive refractory generalised myasthenia gravis. Eculizumab is a humanised monoclonal antibody that specifically binds with high affinity to human terminal complement protein C5, inhibiting enzymatic cleavage to the proteins C5a and C5b and preventing C5a-induced chemotaxis of proinflammatory cells and formation of C5b-induced membrane attack complex.23,24 Eculizumab is currently approved for the treatment of two lifethreatening, complement-mediated diseases: paroxysmal nocturnal haemoglobinuria (in nearly 50 countries) and atypical haemolytic uraemic syndrome (in nearly 40 countries, including the USA, European Union, and

www.thelancet.com/neurology Published online October 20, 2017 http://dx.doi.org/10.1016/S1474-4422(17)30369-1

Articles

Japan).25–27 It is also approved for refractory generalised myasthenia gravis in 28 countries with applications pending in the USA and Japan. Eculizumab’s mechanism of action suggests that it might also be effective in the management of patients with anti-acetylcholine receptor antibody-positive refractory generalised myasthenia gravis, a hypothesis supported by results from a pilot phase 2 study28 in which clinically meaningful improvements occurred in patients treated with eculizumab. We therefore did a phase 3 study (eculizumab for refractory generalised myasthenia gravis; REGAIN) to confirm the results of the pilot study and provide a comprehensive assessment of the safety and efficacy of eculizumab in this patient population.

Methods

Study design and participants We did a phase 3, randomised, double-blind, placebocontrolled study in 76 centres in 17 countries across North America, Latin America, Europe, and Asia (appendix). Independent ethics committees or institutional review boards provided written approval for the study protocol and all amendments. Patients with anti-acetylcholine receptor antibodypositive refractory generalised myasthenia gravis aged 18 years or older were eligible for the study if they had confirmed generalised myasthenia gravis, a positive serological test for anti-acetylcholine receptor antibodies, impaired activities of daily living defined as a Myasthenia Gravis-Activities of Daily Living (MG-ADL) score of 6 or higher (appendix), and class II–IV disease according to the Myasthenia Gravis Foundation of America (MGFA) classification system (appendix). Patients had to have received treatment with two or more immunosuppressive therapies, or at least one immunosuppressive therapy with intravenous immunoglobulin or plasma exchange given at least four times per year, for 12 months without symptom control. Patients with a history of thymic neoplasms, a thymectomy within 12 months prior to screening, exclusively ocular myasthenia gravis (MGFA class I), myasthenic crisis (MGFA class V), or use of intravenous immunoglobulin or plasma exchange within 4 weeks before randomisation or rituximab within 6 months before screening were excluded from the study. Full lists of inclusion and exclusion criteria are in the appendix. Potential patients were recruited from the investigators’ practice or through physicians’ referral. Written informed consent was obtained from all patients before they entered the study.

Randomisation and masking We randomly assigned participants (1:1) to 26 weeks of either intravenous eculizumab or intravenous placebo. Placebo was matched to eculizumab in appearance and supplied in identical containers to preserve masking. The randomisation was done centrally by an independent company (Almac, Souderton, PA, USA) using an

interactive voice or web response system with patients stratified to one of four groups based on MGFA clinical classification assigned at screening (IIa or IIIa, IVa, IIb or IIIb, and IVb). Randomisation was done across, rather than within, the centres because of the small numbers of patients anticipated at each centre. Investigators, patients, study personnel, and the funder remained masked to all treatment assignments for the study duration.

Procedures The schedule for eculizumab was induction dosing 900 mg on day 1 and weeks 1, 2, and 3; 1200 mg at week 4; and maintenance dosing 1200 mg every second week thereafter. Placebo was given on the same schedule. Patients receiving previous treatment with a cholinesterase inhibitor, oral corticosteroid, or other immuno­suppressive treatments were to maintain the dose and schedule of these medications throughout the study, unless there was a compelling medical need for adjustment. As a reflection of the severity of disease under study, rescue medication (such as high-dose corticosteroids, intravenous immuno­ globulin, or plasma exchange) was allowed under protocoldefined conditions, at the study physician’s discretion. All randomised patients were required to have been vaccinated against Neisseria meningitides. We assessed efficacy with the MG-ADL scale;29 the Quantitative Myasthenia Gravis (QMG) score;30 the Myasthenia Gravis Composite (MGC) scale;31 and the 15-item Myasthenia Gravis Quality of Life (MG-QOL15) questionnaire.32 A description and comparison of the key elements of each assessment tool is in the appendix. QMG and MGC assessments were done at least 10 h after the last dose of cholinesterase inhibitor. All assessments were scheduled to be done at screening, on day 1 predose (baseline), at weeks 1, 2, and 3 (weeks 1, 2, and 3 limited to the MG-ADL, QMG, and MGC), and at weeks 4, 8, 12, 16, 20, and 26, or at early termination. For patients who discontinued the study, a post-study followup assessment was done 8 weeks after their last dose of study drug. The overall study duration for an individual patient was up to 38 weeks, including screening and follow-up. The total treatment time was 26 weeks. After the 26-week treatment period, patients had the option to enter an extension study in which they would receive open-label eculizumab (NCT02301624), which is currently ongoing.

See Online for appendix

Outcomes The primary efficacy endpoint was the change in MG-ADL total score from baseline to week 26 for eculizumab compared with placebo, measured by worst-rank ANCOVA. Prespecified secondary efficacy end­points were ordered as follows: (1) change from baseline in QMG total score, (2) responder analysis of the MG-ADL score (≥3-point improvement), (3) responder analysis of the QMG score (≥5-point improvement), (4) change from baseline in MGC total score, and (5) change from baseline in MG-QOL15

www.thelancet.com/neurology Published online October 20, 2017 http://dx.doi.org/10.1016/S1474-4422(17)30369-1

3

Articles

medication use (definitions in the appendix), changes in vital signs (appendix), electrocardiography, and clinical laboratory variables were reported.

170 patients assessed for eligibility

44 excluded because they did not meet inclusion criteria

126 randomised

63 assigned and received placebo

2 discontinued treatment 2 withdrawal by patient

63 assigned eculizumab 62 received eculizumab 1 randomised in error*

5 discontinued treatment 4 adverse events 1 withdrawal by patient

61 completed the study

57 completed the study

63 included in modified intention-to-treat analysis and safety analysis

62 included in modified intention-to-treat analysis and safety analysis

Figure 1: Trial profile For the prespecified outcome change from baseline in worst-rank ANCOVA, 103 patients were ranked by the change from baseline (rank 1–103), 103 completed 26 weeks without rescue therapy, 22 were in the lowest-rank group, ranked by time to event (rank 104–125), one had a myasthenia gravis crisis and rescue therapy (rank 125), 17 other patients required rescue therapy by study end, and four discontinued for any reason. For the post-hoc change from baseline in worst-rank ANCOVA, 106 patients were ranked by the change from baseline (rank 1–106), 103 completed 26 weeks without rescue therapy, three discontinued and did not have an adverse outcome related to myasthenia gravis, 19 were in the lowest-rank group, ranked by time to event (rank 107–125), one had a myasthenia gravis crisis and rescue therapy (rank 125), 17 other patients required rescue therapy by study end, and one had an adverse outcome related to myasthenia gravis. *Patient had a serious adverse event (myocardial infarction) during screening and was inadvertently randomised.

total score. We used hierarchical testing of primary and secondary endpoints to address multiplicity in the study. All change-from-baseline secondary endpoint analyses were assessed at week 26 using a worst-rank ANCOVA, as undertaken for the MG-ADL primary endpoint analysis. The prespecified responder analyses assessed the proportion of patients who received no rescue therapy and met the threshold for improvement in total score from baseline to week 26. For this study, clinical deterioration was defined as one of the following: myasthenia gravis crisis, substantial symptomatic worsening (to a score of 3 or a 2-point worsening on any single MG-ADL item, excluding ocular items), or health in jeopardy if rescue therapy was not given, as determined by the treating physician. The terms clinical deterioration, clinical worsening, and exacerbation are used interchangeably in this report. Clinical assessors of the efficacy outcome measures were trained and certified. Assessments of myasthenia gravis were scheduled for approximately the same time of day and, preferably, by the same assessors. To assess safety, the incidence of adverse events, serious adverse events, admissions to hospital, protocoldefined clinical deterioration with or without rescue 4

Statistical analysis We calculated the sample size to detect a mean-ranked difference of 3 points between the eculizumab and placebo groups based on a two-sided type I error of 5% and 90% power, assuming a common SD of 4. A sample size of approximately 92 patients (46 per treatment group) was required after adjusting for 15% of patients dropping out. The mean-ranked difference of 3 (SD 4) was based on mean changes from baseline to week 26 of 4 points on the MG-ADL scale for eculizumab and 1·50 points for placebo (SD 3·25); and 7 points on the QMG for eculizumab and 3 points for placebo (SD 6), based on the results of the myasthenia gravis pilot phase 2 study.28 Efficacy analyses were done with the full analysis set, defined as all randomly assigned patients who received at least one dose of study drug, had a valid baseline MG-ADL assessment, and at least one post-baseline MG-ADL assessment. Efficacy analyses were also done on the per-protocol set (not presented in this report). Safety analyses were done with the safety analysis set, defined as all randomly assigned patients who received at least one dose of study drug or placebo. To account for the potential effect of rescue medication on subsequent efficacy assessments, as well as other poor outcomes (in the order of death, myasthenia gravis crisis, rescue usage or discontinuation), worst-rank analysis was used. Patients were ranked from 1 (best outcome) to 125 (worst outcome). For each category of poor outcome type, patients were ranked on the basis of the time to that event. For the remaining patients, rank was assigned on the basis of endpoint change from baseline to week 26 (with rank 1 equivalent to the largest improvement). A similar approach was followed to assess QMG, MGC, and MG-QOL15. Prespecified sensitivity analyses were done to assess the effect of key assumptions on the overall conclusions of the study. Sensitivity analyses of the four validated assessments (MG-ADL, QMG, MGC, and MG-QOL15; appendix) included repeated-measures analyses with observed changes from baseline at each visit, with and suppressive use as a without adjusting for immuno­ covariate. Modelling assumptions for both the primary and sensitivity analyses were assessed. All worst-rank ANCOVA and repeated-measures models included treatment indicator, baseline covariate, and the random­ isation stratification variable (pooled to the categories of II, III, and IVa or II, III, and IVb because of the small numbers of patients in MGFA classes IVa and IVb). ANCOVA analyses used last observation carried forward (LOCF) for missing week 26 assessments. No imputation of missing data was done for the repeated-measures analyses. To understand the results of the worst-rank ANCOVA analytical approach, we did a post-hoc worst-

www.thelancet.com/neurology Published online October 20, 2017 http://dx.doi.org/10.1016/S1474-4422(17)30369-1

Articles

Eculizumab Placebo (n=62) (n=63)

Total (n=125)

Eculizumab Placebo (n=62) (n=63)

Total (n=125)

Age at diagnosis (years)

38·0 (17·8)

38·1 (19·6)

38·1 (18·6)

(Continued from previous column)

Age at first study dose (years)

47·5 (15·7)

46·9 (18·0)

47·2 (16·8)

Immunosuppressive treatments at baseline (day 1) Corticosteroids

47 (76%)

51 (81%)

Male

21 (34%)

22 (35%)

43 (34%)

Azathioprine

20 (32%)

21 (33%)

41 (33%)

Female

41 (66%)

41 (65%)

82 (66%)

Mycophenolate mofetil

18 (29%)

16 (25%)

34 (27%)

Asian

3 (5%)

16 (25%)

19 (15%)

Black or African American

0

3 (5%)

3 (2%)

Sex

Race

98 (78%)

Cyclosporine

8 (13%)

9 (14%)

17 (14%)

Tacrolimus

5 (8%)

6 (10%)

11 (9%)

Methotrexate

5 (8%)

4 (6%)

9 (7%)

Rituximab

0

0

0

0

White

53 (85%)

42 (67%)

95 (76%)

Other

6 (10%)

2 (3%)

8 (6%)

Cyclophosphamide

2 (3%)

31·4 (9·0)

30·5 (8·4)

30·9 (8·7)

Previous thymectomy

37 (60%)

31 (49%)

68 (54%)

9·9 (8·1)

9·2 (8·4)

9·6 (8·2)

Mean time from thymectomy to first dose in REGAIN (years)

11 (8·51)

11·3 (9·67)

11·1 (8·99)

Previous long-term intravenous immunoglobulin therapy

18 (29%)

17 (27%)

35 (28%)

BMI (kg/m²) Myasthenia gravis duration (years)

2 (2%)

MG-ADL score

10·5 (3·1)

9·9 (2·6)

10·2 (2·8)

QMG score

17·3 (5·1)

16·9 (5·6)

17·1 (5·3)

MGC score

20·4 (6·1)

18·9 (6·0)

19·6 (6·1)

MG-QOL15 score

33·6 (12·2)

30·7 (12·7)

32·1 (12·5)

Previous long-term plasma exchange therapy

4 (6%)

10 (16%)

14 (11%)

62 (50%)

History of myasthenia gravis exacerbations

46 (74%)

52 (83%)

98 (78%)

History of myasthenia gravis crisis

13 (21%)

10 (16%)

23 (18%)

Any hospital admissions for myasthenia gravis since diagnosis

47 (76%)

48 (76%)

95 (76%)

Any hospital admissions for myasthenia gravis in the last 2 years

30 (48%)

29 (46%)

59 (47%)

Number of patients with ICU visits in the past 2 years

9 (15%)

7 (11%)

16 (13%)

Number of days in ICU per hospital admission

5·7 (5·8)

5·3 (3·3)

5·5 (4·7)

15 (24%)

14 (22%)

29 (23%)

MGFA classification by randomisation stratification* Class IIa or IIIa Class IVa Class IIb or IIIb Class IVb

30 (48%)

32 (51%)

4 (6%)

2 (3%)

6 (5%)

25 (40%)

26 (41%)

51 (41%)

3 (5%)

3 (5%)

6 (5%)

Previous use of immunosuppressive treatments ≥3

31 (50%)

34 (54%)

65 (52%)

≥2

61 (98%)

62 (98%)

123 (98%)

Types of immunosuppressive treatments used before study enrolment Corticosteroids

58 (94%)

62 (98%)

120 (96%)

Azathioprine

47 (76%)

47 (75%)

94 (75%)

Mycophenolate mofetil

27 (44%)

29 (46%)

56 (45%)

Cyclosporine

18 (29%)

18 (29%)

36 (29%)

9 (15%)

11 (17%)

20 (16%)

Tacrolimus Rituximab

7 (11%)

7 (11%)

14 (11%)

Methotrexate

6 (10%)

8 (13%)

14 (11%)

Cyclophosphamide

3 (5%)

3 (5%)

6 (5%)

Time from start of medication to first dose in REGAIN (years) Corticosteroid

7·4 (7·6)

6·2 (6·2)

6·8 (6·9)

Cyclosporine

5·4 (5·5)

6·3 (6·9)

5·9 (6·1)

Methotrexate

4·8 (4·2)

7·4 (8·1)

5·9 (6·0)

Azathioprine

4·1 (3·1)

7·2 (9·6)

5·7 (7·3)

Tacrolimus

3·9 (3·3)

1·5 (1·1)

2·6 (2·5)

Mycophenolate mofetil

3·6 (4·5)

1·9 (1·2)

2·8 (3·5)

(Table 1 continues in next column)

rank analysis. In the post-hoc analysis, criteria for assignment to the lowest ranks were restricted to patients who died or had adverse myasthenia gravis outcomes (myasthenia gravis crisis, use of myasthenia gravis rescue medication, and worsening of myasthenia gravis). Prespecified responder analyses measured the proportion of patients with a reduction of at least 3 points in MG-ADL total score from baseline to week 26 without rescue therapy, and the proportion of patients with a

Any previous ventilator support

Data are mean (SD) or n (%). BMI=body-mass index. MG-ADL=Myasthenia Gravis Activities of Daily Living. QMG=Quantitative Myasthenia Gravis. MGC=Myasthenia Gravis Composite. MG-QOL15=15-item Myasthenia Gravis Quality of Life questionnaire. MGFA=Myasthenia Gravis Foundation of America. ICU=intensive-care unit. *Additional breakdown of MGFA classifications are in the appendix.

Table 1: Baseline characteristics and medical history of modified intention-to-treat population

reduction of at least 5 points in QMG total score from baseline to week 26 without rescue therapy, using the Cochran-Mantel-Haenszel test stratified by MGFA classification. We also did additional analyses using increasingly higher improvement point thresholds (≥4, 5, 6, 7, or 8 on MG-ADL and ≥6, 7, 8, 9, or 10 on QMG). An additional post-hoc worst-rank ANCOVA was done, which mirrored the primary analysis except that discon­tinuation unrelated to myasthenia gravis worsening was not considered a poor outcome (appendix). Safety data were summarised using descriptive statistics. All statistical analyses were done with SAS version 9.4. There was no data monitoring committee for this study.

www.thelancet.com/neurology Published online October 20, 2017 http://dx.doi.org/10.1016/S1474-4422(17)30369-1

5

Articles

The funder of the study had a role in study design, study conduct, and data collection. The funding source was responsible for the statistical analysis plan and protocol as well as the final clinical study report. In addition, the medical writer was an employee of the funding source and additional employees provided a review of the manuscript. All authors had full access to all the data in the study, provided final approval of the manuscript content, and had responsibility for the decision to submit for publication.

Results Between April 30, 2014, and Feb 19, 2016, 170 patients were screened, of whom 126 were randomly assigned to treatment and 125 received study drug (62 received eculizumab and 63 received placebo; figure 1). Of the 125 treated patients, 118 completed the study and seven Eculizumab (n=62)

Placebo (n=63) Difference (95% CI)

p value*

Prespecified worst-rank ANCOVA score† MG-ADL‡

56·6 (4·5)

68·3 (4·5)

−11·7 (−24·3 to 0·96)

QMG

54·7 (4·5)

70·7 (4·5)

−16·0 (−28·5 to −3·4)

0·0698 0·0129

MGC

57·3 (4·5)

67·7 (4·5)

−10·5 (−23·1 to 2·1)

0·1026

MG-QOL15

55·5 (4·6)

69·7 (4·5)

−14·3 (−27·0 to −1·6)

0·0281

Prespecified sensitivity repeated-measures model analysis with immunosuppressive treatments as covariate§ MG-ADL

−4·1 (0·5)

−2·3 (0·5)

−1·8 (−3·2 to −0·5)

0·0077

QMG

−4·6 (0·6)

−1·7 (0·6)

−2·9 (−4·6 to −1·2)

0·0007

−7·9 (1·0)

−4·6 (1·0)

−3·3 (−5·9 to −0·6)

0·0168

−13·8 (1·6)

−6·7 (1·6)

−7·1 (−11·3 to −3·0)

0·0009

−15·4 (−27·8 to −2·9)

0·0160

MGC MG-QOL15

Post-hoc sensitivity worst-rank ANCOVA score¶ MG-ADL

54·8 (4·5)

70·2 (4·4)

QMG

53·9 (4·5)

71·6 (4·4)

−17·7 (−30·1 to −5·3)

0·0055

MGC

56·1 (4·5)

69·0 (4·4)

−12·9 (−25·4 to −0·5)

0·0414

MG-QOL15

54·6 (4·5)

70·6 (4·5)

−16·0 (−28·6 to −3·4)

0·0134

Data are changes from baseline at week 26, given as least-squares mean rank (SEM) for prespecified and post-hoc sensitivity worst-rank ANCOVA scores and least-squares mean (SEM) for prespecified sensitivity repeated-measures model analysis with immunosuppressive treatments as covariates. Responder analyses and repeated-measures model sensitivity analyses are in figures 2 and 3. For all analyses, the eculizumab group was compared with the placebo group and all hypothesis testing was two-sided and done at the 0·05 level of significance, unless otherwise specified. Estimates of treatment effect on efficacy parameters were accompanied by two-sided 95% CIs for the effect size. MGADL=Myasthenia Gravis Activities of Daily Living. QMG=Quantitative Myasthenia Gravis. MGC=Myasthenia Gravis Composite. MG-QOL15=15-item Myasthenia Gravis Quality of Life questionnaire. *Nominal p values, except for prespecified worst-rank ANCOVA change from baseline in MG-ADL. †Prespecified worst-rank ANCOVA of each patient: ranked from 1 (best) to 125 (worst), whereby the patient who had a myasthenia gravis crisis was ranked lowest and patients who received rescue therapy or dropped out of the study were ranked lowest according to time to event; all other patients were ranked higher according to change from baseline to week 26 or last observation carried forward (LOCF). ‡Primary analysis. §Data from all patients (including a covariate for immunosuppressive treatment) at each assessment timepoint were used to compare eculizumab and placebo groups. ¶Post-hoc sensitivity worst-rank ANCOVA of each patient: ranked from 1 (best) to 125 (worst), whereby the patient who had a myasthenia gravis crisis was ranked lowest and patients who received rescue therapy or dropped out of the study with myasthenia gravis exacerbations without rescue therapy were ranked according to time to event; all other patients (including those who dropped out for reasons unrelated to myasthenia gravis exacerbations) were ranked higher according to change from baseline to week 26 or LOCF.

Table 2: Efficacy and sensitivity analyses at week 26

6

A

Minimum point improvement

Role of the funding source

dropped out prematurely. Discontinuation because of adverse events was the most common reason. Treatment groups were generally well matched regarding demographic characteristics, disease status, and medical history (table 1, appendix). 78% of patients received concomitant corticosteroids and more than 80% received other immunosuppressive therapies during the study. Mean baseline scores for all assessment tools reflected a refractory population that had many symptoms and persistent morbidities despite concomitant use of immunosuppressive therapies. Half of all patients reported their worst MGFA classification since diagnosis as class IV or V (table 1). For the primary endpoint, the difference between the groups in mean-ranked difference in change in MG-ADL total score from baseline to week 26 for the eculizumab and placebo groups (least-squares mean rank 56·6 [SEM 4·5]) vs 68·3 [4·5]; rank-based treatment difference −11·7 [95% CI −24·3 to 0·96]) did not achieve significance (p=0·0698; table 2). The change in QMG total score from baseline to week 26, as measured by the worst-rank ANCOVA, showed a benefit with eculizumab compared with placebo (p=0·0129; table 2). In the responder analysis for the MGADL and the QMG, a higher proportion of patients Eculizumab (n=62) Placebo (n=63)

MG-ADL

8

p=0·0176

7

p=0000·7

6

p=0·0072

5

p=0·0182

4

p=0·0358

3

p=0·0229

B

Minimum point improvement

This study is registered with ClinicalTrials.gov, number NCT01997229.

21%

6%

34%

10%

39%

18%

45%

25%

55%

37%

60%

40%

QMG

10

p=0·0043

9

p=0·0036

8

p=0·0005

7

p=0·0024

6

p=0·0021

5

p=0·0018 60

40

16%

2%

23%

5%

27% 34%

5% 11%

39%

14%

45%

19%

20

0

20

40

Patients (n)

Figure 2: Prespecified responder analyses done at 26 weeks Data are minimum point improvements on the outcome measures of (A) the Myasthenia Gravis Activities of Daily Living (MG-ADL) scale and (B) the Quantitative Myasthenia Gravis (QMG) scale. Nominal p values for the more stringent improvements in MG-ADL (≥4, 5, 6, 7, and 8) and QMG (≥6, 7, 8, 9, and 10) are provided to aid interpretation, but they were not prespecified.

www.thelancet.com/neurology Published online October 20, 2017 http://dx.doi.org/10.1016/S1474-4422(17)30369-1

Articles

B 0

–1

–1

0· p=

56

0· p=

0·

p=

0· 00

06

p=

0·

00

56 0· 00 p=

p=

0· 00

53

21 00 0·

22

p=

–7

D

MGC

0

MG-QOL 15

0 –2

–2

–4

–4

–4·8

–5·4

–6 –8

p=0·0166

95

–10

0

2

4

6

10

12

p=

Time (weeks)

18

20

22

24

26

00

10

0· 00

0· p=

–18

p=

p=

0· 00

76

0· 01

–16

28

93

0· 00

–14

p=

34 p=

0·

16

–12·6

02

–12

p=

14

p=

p=

8

0· 01

00

01 0·

p=

–12

63

0· 03

08

00 0·

p=

p=

0·

–10

24

00

07

–8·1 03

–8

p< 0· 00 0 0· 00 1 76

0· 03

–6

p=

Change from baseline total score (mean, 95% CI)

–6

p=

00 p=

p=

p=

0·

0·

01

96 00

–7

07

0· p=

Eculizumab (n=62) Placebo (n=63)

–4·6

58

01

00

83

46

–5 0·

–6

C

00 p=0·0644 71

p=0·0125

p=

–5

–4 –4·2 0·

–4

–3

p= 0 p= ·00 0· 02 05 05 0· 00 08

–3

–1·6

–2

72

–2·3

–2

QMG

02

MG-ADL

0

p=

Change from baseline total score (mean, 95% CI)

A

the treatment effect occurring by week 12 and sustained to week 26 (figure 3). The mean change from baseline at week 26 in MG-ADL and QMG score was greater with eculizumab than with placebo based on repeatedmeasures analyses with and without immunosuppressive therapies as a covariate (figure 3, table 2). The prespecified worst-rank criteria assigned all discontinued patients to the lowest ranks regardless of reason for discontinuation (figure 1). Of the seven patients who discontinued (figure 1), four met protocol criteria for clinical worsening (two in the placebo group and two in the eculizumab group). The other three, all of whom received eculizumab, discontinued the study because of adverse events that did not reflect worsening of myasthenia gravis and did not receive rescue therapy, and all three had a clinically meaningful benefit in response to treatment (appendix). The adverse effects that led to discontinuation were Moraxella lacunata bacteraemia, bowel perforation, and adenocarcinoma of prostate gland. At the time of discontinuation, these three patients had 3, 7, and 7-point improvements in MG-ADL, respectively. In the post-hoc analysis, where those patients who discontinued but were

04

achieved a clinically meaningful response with eculizumab than with placebo (figure 2). For both endpoints, increasing the stringency of the responder definition with higher thresholds revealed a more substantial difference between eculizumab and placebo at each threshold (figure 2). The worst-rank ANCOVA for MGC did not differ between groups. The score for the MG-QOL15 showed a benefit with eculizumab compared with placebo (p=0·0281). With the repeated-measures analysis, patients receiving eculizumab showed an initial improved MGC total score by week 1, and initial improved MG-QoL15 total score by week 4, with most of the treatment effect occurring by week 12 and sustained to week 26 (figure 3). Mean change from baseline at week 26 in both scores was greater with eculizumab compared with placebo based on repeated-measures analyses with and without immuno­ suppressive therapies as a covariate (figure 3, table 2). Further sensitivity analyses are in the appendix. In sensitivity analyses, patients receiving eculizumab showed an initial improvement in MG-ADL total score by week 1, and QMG total score by week 2, with most of

0

2

4

6

8

10

12

14

16

18

20

22

24

26

Time (weeks)

Figure 3: Prespecified sensitivity analyses for four outcome measures The analyses used repeated measures without immunosuppressive therapy as a covariate and show the change in scores over time for (A) the Myasthenia Gravis Activities of Daily Living (MG-ADL) scale, (B) the Quantitative Myasthenia Gravis (QMG) scale, (C) the Myasthenia Gravis Composite (MGC) scale, and (D) the 15-item Myasthenia Gravis Quality of Life (MG-QOL15) questionnaire. p values are nominal.

www.thelancet.com/neurology Published online October 20, 2017 http://dx.doi.org/10.1016/S1474-4422(17)30369-1

7

Articles

Eculizumab (n=62) Placebo (n=63) Total (n=125) Admissions to hospital

9 (15%)

Discontinuations because of adverse events

4 (6%)

18 (29%)

Patient reports of myasthenia gravis exacerbations

6 (10%)

15 (24%)

21 (17%)

Rescue therapy used during the 26-week treatment period

6 (10%)

12 (19%)

18 (14%) 5 (4%)

0

27 (22%) 4 (3%)

High-dose corticosteroids

0

5 (8%)

Plasmapheresis or plasma exchange

3 (5%)

4 (6%)

7 (6%)

Intravenous immunoglobulin

4 (6%)

6 (10%)

10 (8%)

Other

1 (2%)

2 (3%)

3 (2%) 22 (18%)

Most common adverse events* (≥10% in either group) Headache

10 (16%)

12 (19%)

Upper respiratory tract infection

10 (16%)

12 (19%)

22 (18%)

Nasopharyngitis

9 (15%)

10 (16%)

19 (15%)

Nausea

8 (13%)

9 (14%)

17 (14%)

Diarrhoea

8 (13%)

8 (13%)

16 (13%)

Myasthenia gravis

6 (10%)

11 (17%)

17 (14%)

Data are n (%). Reports of myasthenia gravis exacerbations and crises were also reported separately from adverse events and serious adverse events. Common adverse events and serious adverse events were reported by the physician in accordance with their clinical discretion. Investigators were not required to report each myasthenia gravis exacerbation as an adverse event under the term myasthenia gravis unless it was a serious adverse event. *Preferred term in the Medical Dictionary for Regulatory Activities.

Table 3: Treatment-emergent safety outcomes in all treated patients

improving on the MG-ADL were not assigned to the lowest ranks, patients receiving eculizumab had a greater treatment effect than those receiving placebo across all four outcome measures (table 2). No deaths or cases of meningococcal infection occurred during the study. One event of myasthenia gravis crisis occurred in a patient receiving eculizumab. This patient did not respond to eculizumab (MG-ADL score increased from 13 at baseline to 18 at last assessment), discontinued from the study because of the crisis, and died 90 days after the last eculizumab dose from crisis-related complications. The most common adverse events were headache, upper respiratory tract infection, and naso­ pharyngitis (table 3). Most adverse events were mild to moderate in severity and unrelated to the study drug. Four patients, all treated with eculizumab, discontinued because of an adverse event (figure 1), including the patient who had a myasthenia gravis crisis. Serious adverse events were reported in nine (15%) patients receiving eculizumab and 18 (29%) patients receiving placebo (table 4). The most frequently reported serious adverse events were infections (four events in two patients receiving eculizumab [3%] vs seven events in six patients receiving placebo [10%]; table 4). Six (10%) patients receiving eculizumab and 12 (19%) receiving placebo required rescue therapy (table 3). Exacerbations were reported by six (10%) patients in the eculizumab group and 15 (24%) in the placebo group (table 3), three of whom (all in the placebo group) did not require rescue therapy. There were no clinically meaningful changes from baseline in mean vital sign 8

measurements, bodyweight, or electrocardiography, haematology, or chemistry parameters in either treatment group (appendix).

Discussion In this phase 3 study, the mean-ranked difference in change in MG-ADL between eculizumab and placebo was not statistically significant. Inhibition of terminal complement activation is a biologically rational approach to prevent damage at the neuro­ muscular junction in patients with anti-acetylcholine receptor antibody-positive refractory generalised myasthenia gravis. REGAIN was designed to compare eculizumab, a terminal complement inhibitor, with placebo in these patients. However, the worst-rank approach we used was a statistical method that enabled patients who needed temporary rescue medication to be included in the efficacy analysis by treating rescue medication use or discontinuation for any reason as a negative outcome. The derived use of ranks rather than actual change scores in the worst-rank analysis makes it difficult to translate the implications of the study results into clinical practice. To provide a comprehensive clinical picture of treatment effect that reflects both the physicians’ and patients’ perspectives, we included prespecified sensitivity and secondary analytical approaches in the analysis plan. The sensitivity analyses of changes in the MG-ADL, QMG, and MG-QOL15 scores showed improvements with eculizumab compared with placebo, and results for the MGC were consistent with those for MG-ADL (the change from baseline to week 26, as measured by the worst-rank analysis, were not statistically significant for both the MG-ADL and the MGC, but in all other analyses a benefit of eculizumab compared with placebo was observed in both measures). An emphasis on the proportion of patients who exceed established clinically meaningful differences is especially important in a rare disease. Using thresholds above the established clinically important differences, defined as the threshold value that patients and clinicians perceived as clinically meaningful (appendix) for the responder analyses of MG-ADL and QMG,33,34 as well as progressively increasing stricter thresholds, two to three times as many patients improved with eculizumab treatment compared with placebo. A portion of patients in the placebo group also achieved a clinically meaningful change in both MG-ADL and QMG scores. This might reflect the well known effect of placebo responsiveness in patients with neurological disorders participating in clinical trials, and could account for some of the response seen in the placebo group during this study in which patients were followed up more frequently than in clinical practice.35 Alternatively, this observation might partly reflect the known variability of myasthenic symptoms over short periods of time or greater potential for variability when baseline scores are higher (as expected in patients with refractory disease).

www.thelancet.com/neurology Published online October 20, 2017 http://dx.doi.org/10.1016/S1474-4422(17)30369-1

Articles

Patients in REGAIN were maintained on stable immunosuppressive therapies throughout the study and, even with concomitant immunosuppressive therapy, those in the eculizumab group had larger improvements than those in the placebo group, as shown by responder and repeated-measures analyses. The prespecified repeated-measures sensitivity analyses were specifically done to determine if background immunosuppressive therapy use confounded interpretation of study results; it was shown that with or without controlling for their use, there was a separation between treatment arms favouring eculizumab over placebo. Using the repeated-measures analyses, the benefit of eculizumab compared with placebo occurred within the first 4 weeks of treatment, with most of the effect achieved by 12 weeks. Commonly used steroid-sparing agents might require 6 months or longer36–38 to achieve clinical benefit, although the extent and rate of steroid reduction that constitutes steroid sparing has not been defined. Overall, eculizumab was well tolerated, and the safety profile was consistent with its known safety profile in paroxysmal nocturnal haemoglobinuria and atypical haemolytic uraemic syndrome25–27 and post-marketing experience in these indications over 10 years. No increase occurred in serious infections or reports of meningococcal infections, and no new safety concerns were identified in this patient population. Numerically fewer myasthenia gravis exacerbations, reduced use of rescue medication, and fewer admissions to hospital occurred in the eculizumab group compared with the placebo group, all of which contribute to overall disease burden. The strengths of this study included a rigorous randomised, placebo-controlled study design and use of several validated outcome measures, including physician-reported and patient-reported outcomes, as well as multiple analytical approaches. However, this study highlights an important limitation in the use of a worst-rank analytic approach. Because discontinuation from a clinical study could be viewed as a negative outcome, the prespecified worst-rank analyses assigned all patients who discontinued the study to the pooroutcome group, regardless of the reason for discontinuation or myasthenia gravis status at the time of discontinuation. The results of the post-hoc analysis illustrate the importance of differentiating patients with poor myasthenia gravis outcomes (myasthenia gravis crisis or clinical worsening with or without rescue therapy) from patients with poor outcomes unrelated to myasthenia gravis when applying the worst-rank analytical approach. The sample size used in this study might not have been adequate to allow for the informative use of the worst-rank analytical approach. This study offered the opportunity to characterise patients with anti-acetylcholine receptor antibodypositive refractory generalised myasthenia gravis, adding to the scarce characterisation of such patients in the

Eculizumab (n=62) Placebo (n=63) Total (n=125) Patients with any treatment-emergent serious adverse event*

9 (15%)

18 (29%)

27 (22%)

Myasthenia gravis

5 (8%)

8 (13%)

13 (10%)

Pyrexia

2 (3%)

0

2 (2%)

Upper respiratory tract infection

0

2 (3%)

2 (2%)

Apnoea

0

1 (2%)

1 (1%)

Bacteraemia

1 (2%)

0

1 (1%)

Acute cholecystitis

0

1 (2%)

1 (1%)

Deep-vein thrombosis

0

1 (2%)

1 (1%)

Diverticulitis

1 (2%)

0

1 (1%)

Endocarditis

1 (2%)

0

1 (1%)

Gastritis

0

1 (2%)

1 (1%)

Gastroenteritis

0

1 (2%)

1 (1%)

General physical health deterioration

0

1 (2%)

1 (1%)

Hyperglycaemia

0

1 (2%)

1 (1%)

Intentional overdose†

0

1 (2%)

1 (1%)

Intestinal perforation

1 (2%)

0

1 (1%)

Lymphocyte count decreased

0

1 (2%)

1 (1%)

Lymphopenia

1 (2%)

0

1 (1%)

Metastases to bone

1 (2%)

0

1 (1%)

Myasthenia gravis crisis

1 (2%)

0

1 (1%)

Prostate cancer

1 (2%)

0

1 (1%)

Pulmonary embolism

0

1 (2%)

1 (1%)

Tonsillitis

0

1 (2%)

1 (1%)

Bacterial urinary tract infection

0

1 (2%)

1 (1%)

Varicella

0

1 (2%)

1 (1%)

Data are n (%). Serious adverse events were reported by the physician in accordance with their clinical discretion. Investigators were not required to report each myasthenia gravis exacerbation as an adverse event under the term myasthenia gravis unless it was a serious adverse event. *Preferred term in the Medical Dictionary for Regulatory Activities. †Overdose on concomitant medications (pyridostigmine and paracetamol); no unintentional overdoses were reported as serious adverse events.

Table 4: Serious treatment-emergent adverse events in all treated patients

literature.1 Maximum disease worsening in myasthenia gravis has been reported in the first 2–3 years after diagnosis,39 but has never been described for refractory disease. In this study, patients were on background immunosuppressive therapy; however, patients continued to be at risk of exacerbations and hospital admissions before the start of REGAIN, despite a mean disease duration of 10 years. After 26 weeks of treatment, the numbers of exacerbations and hospital visits for patients receiving eculizumab were half those for patients receiving placebo, suggesting that over-activation of the complement system might contribute to the risk of disease exacerbation. Effective therapies to treat patients with antiacetylcholine receptor antibody-positive refractory generalised myasthenia gravis are crucially needed. This was the first phase 3 clinical study specifically targeting this patient population that is difficult to treat and has a substantial disease burden. The primary endpoint did not achieve statistical significance. Interpretation of any trial, however, should consider all evidence from the

www.thelancet.com/neurology Published online October 20, 2017 http://dx.doi.org/10.1016/S1474-4422(17)30369-1

9

Articles

study.40 The preplanned analyses and outcomes from REGAIN suggest a potential benefit of eculizumab treatment in  patients with anti-acetylcholine receptor antibody-positive refractory generalised myasthenia gravis. These results should stimulate further research into the role of complement in this disease. Contributors JFH Jr, KU, MB, HM, RJB, TMB, JTK, FO’B, J-JW, and RM contributed to the concept or design of the study. JFH Jr, KU, MB, HM, II, SJ, JV, JTK, SM, RJN, FO’B, J-JW, and RM contributed to acquisition, analysis, or interpretation of the data. All authors contributed to drafting the report or revised it critically for important intellectual content. Declaration of interests JFH Jr received research support and grants from Alexion Pharmaceuticals, research support from the Centers for Disease Control and Prevention (CDC), grants from the National Institutes of Health (NIH; National Institute of Neurological Disorders and Stroke [NINDS], National Institute of Arthritis and Musculoskeletal and Skin Diseases, Environmental Medicine, Asthma and Lung Biology), grants from UCB Biosciences, research support from the Muscular Dystrophy Association, and non-financial support from Alexion Pharmaceuticals, Alnylam Pharmaceuticals, Argenx BVBA, RA Pharmaceuticals, and Toleranzia AB. KU has received honoraria for consultancies from Alexion Pharmaceuticals. MB received research support from NIH (NINDS, National Center for Advancing Translational Sciences), CDC, the Food and Drug Administration Orphan Products Development Program, US Department of Defense, ALS Association, Muscular Dystrophy Association, and Eli Lilly and Company. KU is also the site investigator for Alexion Pharmaceuticals, Cytokinetics, and Neuraltus, and has consultantships with Alnylam Pharmaceuticals, RA Pharmaceuticals, Voyager Therapeutics, UCB Pharma, Denali, and Mitsubishi-Tanabe Pharmaceuticals. HM received fees from Alexion Pharmaceuticals, Biogen, Novartis, Bayer, Mitsubishi-Tanabe, and the Japan Blood Products Organization. RJB received personal fees from NuFactor, Plan 365, and Option Care, and research support from Novartis, Sanofi/Genzyme, Biomarin, IONIS, Teva, Cytogenetics, and Eli Lilly and Company. SJ is an advisory board member for Alnylam Pharmaceuticals and a member of International Advisory Board for Alexion Pharmaceuticals. JV received research and travel support, speaker honoraria from Sanofi/Genzyme, Ultragenyx Pharmaceuticals, Santhera Pharmaceuticals and aTyr Pharma, and has served as consultant on advisory boards for Sanofi/Genzyme, aTyr Pharma, Ultragenyx Pharmaceuticals, Santhera Pharmaceuticals, Sarepta Therapeutics, Novo Nordisk, Alexion Pharmaceuticals, and Stealth Biotherapeutics within the past 3 years. JTK received research and travel support from Alexion Pharmaceuticals, NIH (NINDS), aTyr Pharma, Cytokinetics, and AveXis. SM has been an advisory board member for Alnylam Pharmaceuticals, and a consultant for Alexion Pharmaceuticals, Alnylam Pharmaceutical, and Lundbeck Pharmaceuticals. RJN received research support from the NIH (NINDS and NIADS), Alexion Pharmaceuticals, Genentech, Grifols, and Myasthenia Gravis Foundation of America, and has consultantships with Alexion Pharmaceuticals, RA Pharma, Shire, and Grifols. RM received funding for research and congress participation from Sanofi-Genzyme, Teva, Bayer, and BioMarin, and has participated in scientific advisory boards for BioMarin, Alexion Pharmaceuticals, and Argenx BVBA. FO’B and J-JW are employed by and own stocks in Alexion Pharmaceuticals. II and TMB declare no competing interests. Acknowledgments This study was funded by Alexion Pharmaceuticals (New Haven, CT, USA). We thank the patients who took part and their families as well as the REGAIN principal investigators, subinvestigators, and study coordinators (appendix). We also thank Gary Cutter (UAB School of Public Health, New Haven, Connecticut, USA) for his expert advice early in the study, Laura Herbelin (University of Kansas Medical Center, KA, USA) for providing standardised training and certification of the MGADL, QMG, and MGC scales, Angela Kaya (Alexion Pharmaceuticals) for medical writing support, Róisín Armstrong, Kenji Fujita, and Gus Khursigara (Alexion Pharmaceuticals) for critical review of the

10

manuscript, Cindy Lane (Alexion Pharmaceuticals) for clinical study oversight, and Charlotte Cookson and Ruth Gandolfo (Oxford PharmaGenesis, Oxford, UK) who provided editorial assistance in the production of the manuscript (funded by Alexion Pharmaceuticals). References 1 Suh J, Goldstein JM, Nowak RJ. Clinical characteristics of refractory myasthenia gravis patients. Yale J Biol Med 2013; 86: 255–60. 2 Buzzard KA, Meyer NJ, Hardy TA, Riminton DS, Reddel SW. Induction intravenous cyclophosphamide followed by maintenance oral immunosuppression in refractory myasthenia gravis. Muscle Nerve 2015; 52: 204–10. 3 Sanders DB, Wolfe GI, Benatar M, et al. International consensus guidance for management of myasthenia gravis: executive summary. Neurology 2016; 87: 419–25. 4 Engel-Nitz NM, Boscoe AN, Wolbeck R, Johnson J, Silvestri N. Clinical and economic burden of refractory generalized myasthenia gravis in the United States. J Neuromuscul Dis 2016; 3 (suppl 1): S198–99. 5 Lindstrom JM, Seybold ME, Lennon VA, Whittingham S, Duane DD. Antibody to acetylcholine receptor in myasthenia gravis. Prevalence, clinical correlates, and diagnostic value. Neurology 1976; 26: 1054–59. 6 Mantegazza R, Pareyson D, Baggi F, et al. Anti AChR antibody: relevance to diagnosis and clinical aspects of myasthenia gravis. Ital J Neurol Sci 1988; 9: 141–45. 7 Vincent A, McConville J, Farrugia ME, et al. Antibodies in myasthenia gravis and related disorders. Ann NY Acad Sci 2003; 998: 324–35. 8 Vincent A, Newsom-Davis J. Acetylcholine receptor antibody as a diagnostic test for myasthenia gravis: results in 153 validated cases and 2967 diagnostic assays. J Neurol Neurosurg Psychiatry 1985; 48: 1246–52. 9 Oh SJ, Kim DE, Kuruoglu R, Bradley RJ, Dwyer D. Diagnostic sensitivity of the laboratory tests in myasthenia gravis. Muscle Nerve 1992; 15: 720–24. 10 Somnier FE. Clinical implementation of anti-acetylcholine receptor antibodies. J Neurol Neurosurg Psychiatry 1993; 56: 496–504. 11 Biesecker G, Gomez CM. Inhibition of acute passive transfer experimental autoimmune myasthenia gravis with Fab antibody to complement C6. J Immunol 1989; 142: 2654–59. 12 Christadoss P. C5 gene influences the development of murine myasthenia gravis. J Immunol 1988; 140: 2589–92. 13 Karachunski PI, Ostlie NS, Monfardini C, Conti-Fine BM. Absence of IFN-gamma or IL-12 has different effects on experimental myasthenia gravis in C57BL/6 mice. J Immunol 2000; 164: 5236–44. 14 Piddlesden SJ, Jiang S, Levin JL, Vincent A, Morgan BP. Soluble complement receptor 1 (sCR1) protects against experimental autoimmune myasthenia gravis. J Neuroimmunol 1996; 71: 173–77. 15 Fumagalli G, Engel AG, Lindstrom J. Ultrastructural aspects of acetylcholine receptor turnover at the normal end-plate and in autoimmune myasthenia gravis. J Neuropathol Exp Neurol 1982; 41: 567–79. 16 Conti-Tronconi B, Tzartos S, Lindstrom J. Monoclonal antibodies as probes of acetylcholine receptor structure. 2. Binding to native receptor. Biochemistry 1981; 20: 2181–91. 17 Tuzun E, Huda R, Christadoss P. Complement and cytokine based therapeutic strategies in myasthenia gravis. J Autoimmun 2011; 37: 136–43. 18 Zhou Y, Gong B, Lin F, Rother RP, Medof ME, Kaminski HJ. Anti-C5 antibody treatment ameliorates weakness in experimentally acquired myasthenia gravis. J Immunol 2007; 179: 8562–67. 19 Soltys J, Kusner LL, Young A, et al. Novel complement inhibitor limits severity of experimentally myasthenia gravis. Ann Neurol 2009; 65: 67–75. 20 Engel AG, Lambert EH, Howard FM. Immune complexes (IgG and C3) at the motor end-plate in myasthenia gravis: ultrastructural and light microscopic localization and electrophysiologic correlations. Mayo Clin Proc 1977; 52: 267–80. 21 Sahashi K, Engel AG, Lambert EH, Howard FM Jr. Ultrastructural localization of the terminal and lytic ninth complement component (C9) at the motor end-plate in myasthenia gravis. J Neuropathol Exp Neurol 1980; 39: 160–72.

www.thelancet.com/neurology Published online October 20, 2017 http://dx.doi.org/10.1016/S1474-4422(17)30369-1

Articles

22 Nakano S, Engel AG. Myasthenia gravis: quantitative immunocytochemical analysis of inflammatory cells and detection of complement membrane attack complex at the end-plate in 30 patients. Neurology 1993; 43: 1167–72. 23 Rollins SA, Sims PJ. The complement-inhibitory activity of CD59 resides in its capacity to block incorporation of C9 into membrane C5b-9. J Immunol 1990; 144: 3478–83. 24 Rother RP, Rollins SA, Mojcik CF, Brodsky RA, Bell L. Discovery and development of the complement inhibitor eculizumab for the treatment of paroxysmal nocturnal hemoglobinuria. Nat Biotechnol 2007; 25: 1256–64. 25 Hillmen P, Young NS, Schubert J, et al. The complement inhibitor eculizumab in paroxysmal nocturnal hemoglobinuria. N Engl J Med 2006; 355: 1233–43. 26 Zuber J, Fakhouri F, Roumenina LT, Loirat C, Fremeaux-Bacchi V, French Study Group for a HCG. Use of eculizumab for atypical haemolytic uraemic syndrome and C3 glomerulopathies. Nat Rev Nephrol 2012; 8: 643–57. 27 Zuber J, Le Quintrec M, Krid S, et al. Eculizumab for atypical hemolytic uremic syndrome recurrence in renal transplantation. Am J Transplant 2012; 12: 3337–54. 28 Howard JF Jr, Barohn RJ, Cutter GR, et al. A randomized, double-blind, placebo-controlled phase II study of eculizumab in patients with refractory generalized myasthenia gravis. Muscle Nerve 2013; 48: 76–84. 29 Wolfe GI, Herbelin L, Nations SP, Foster B, Bryan WW, Barohn RJ. Myasthenia gravis activities of daily living profile. Neurology 1999; 52: 1487–89. 30 Barohn RJ, McIntire D, Herbelin L, Wolfe GI, Nations S, Bryan WW. Reliability testing of the quantitative myasthenia gravis score. Ann N Y Acad Sci 1998; 841: 769–72.

31 Burns TM, Conaway M, Sanders DB, MG-Qol Study Group. The MG Composite: a valid and reliable outcome measure for myasthenia gravis. Neurology 2010; 74: 1434–40. 32 Burns TM, Grouse CK, Conaway MR, Sanders DB, MG-QOL15 Study Group. Construct and concurrent validation of the MG-QOL15 in the practice setting. Muscle Nerve 2010; 41: 219–26. 33 Muppidi S. The myasthenia gravis-specific activities of daily living profile. Ann N Y Acad Sci 2012; 1274: 114–19. 34 Katzberg HD, Barnett C, Merkies IS, Bril V. Minimal clinically important difference in myasthenia gravis: outcomes from a randomized trial. Muscle Nerve 2014; 49: 661–65. 35 Murray D, Stoessl AJ. Mechanisms and therapeutic implications of the placebo effect in neurological and psychiatric conditions. Pharmacol Ther 2013; 140: 306–18. 36 Muscle Study Group. A trial of mycophenolate mofetil with prednisone as initial immunotherapy in myasthenia gravis. Neurology 2008; 71: 394–99. 37 Pasnoor M, He J, Herbelin L, et al. A randomized controlled trial of methotrexate for patients with generalized myasthenia gravis. Neurology 2016; 87: 57–64. 38 Sanders DB, Hart IK, Mantegazza R, et al. An international, phase III, randomized trial of mycophenolate mofetil in myasthenia gravis. Neurology 2008; 71: 400–06. 39 Grob D, Brunner N, Namba T, Pagala M. Lifetime course of myasthenia gravis. Muscle Nerve 2008; 37: 141–49. 40 Pocock SJ, Stone GW. The primary outcome fails—what next? N Engl J Med 2016; 375: 861–70.

www.thelancet.com/neurology Published online October 20, 2017 http://dx.doi.org/10.1016/S1474-4422(17)30369-1

11