Immunosuppressive agents for myasthenia gravis
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
We will review the efficacy of immunosuppressive therapies in MG.
Background
The defining clinical feature of myasthenia gravis (MG) is fluctuating, painless muscle weakness that typically affects the extraocular, pharyngeal and proximal limb muscles. Severe respiratory or bulbar weakness constitutes a myasthenic crisis, which may require intubation and mechanical ventilation. It is an autoimmune disorder of neuromuscular transmission usually caused by antibodies to postsynaptic nicotinic acetylcholine receptors (AChRs) at the motor endplate (Drachman 1994). These antibodies reduce the number of functional AChRs, thus causing impaired neuromuscular transmission. About 15% of patients with generalised MG do not have detectable anti‐AChR antibodies (Vincent 1985). However, plasma from these patients contains various other immune factors and about 50% have an IgG antibody against the muscle‐specific kinase (MuSK) (Vincent 2003).
The annual incidence of MG is between 0.25 and 2.00 per 100,000 population (Vincent 2001). The natural history of the disease is characterised by exacerbations and remissions (Richman 2003). In untreated MG, the most disabling, sometimes life‐threatening phase is usually the first five to seven years (Grob 1981). Spontaneous remissions occur in about 10 to 15% of patients in the 10 years after onset (Oosterhuis 1981). With treatment, the mortality and morbidity from MG have markedly decreased in the last forty years (Oosterhuis 1988). Treatment options include symptomatic therapy with cholinesterase inhibitors and immunosuppression with glucocorticosteroids, azathioprine, ciclosporin, methotrexate, mycophenolate mofetil, tacrolimus, or cyclophosphamide. Thymectomy can be useful in many patients. Plasma exchange, immunoadsorption, and intravenous immunoglobulin are useful emergency therapies.
The first report of the use of cytotoxic immunosuppressive agents in a large series of MG patients was by Mertens and colleagues (Mertens 1969). They described 38 patients treated with 6‐mercaptopurine, azathioprine, methotrexate or actinomycin D. Some of these agents were combined or given alternately. From the data presented, it was impossible to determine which patients benefited from which drug. However, an overall improvement was observed in 84% of patients.
Matell et al. (Matell 1976) found improvement in 78% of 26 patients treated with azathioprine who had not responded to adrenocorticotropin hormone (ACTH) or glucocorticosteroids. In a further study, Mertens et al. reported 78 patients treated with azathioprine with or without steroids and/or thymectomy. There was improvement in 91% of patients (Mertens 1981). Witte et al. (Witte 1984) found that 15 of 18 patients (83%) improved when treated with azathioprine alone for longer than six months. Of 41 patients followed for more than three years, there was improvement in their MG when azathioprine was used both as monotherapy and in combination with prednisolone (Kuks 1991). In this study, of the 21 patients who were not on steroids, 19 (90%) either improved or remained stable on azathioprine. Thirteen (65%) of the 20 patients on steroids improved with the addition of azathioprine. In another study, 75% of 32 patients treated with azathioprine alone improved, compared with 70% of the 57 patients treated with azathioprine and steroids (Mantegazza 1988). The use of early 'high‐dose' immunosuppression with azathioprine and prednisolone resulted in 50% of patients achieving remission after two years, compared with only a 16% remission rate in those on a 'low‐dose' regimen (Heckmann 2001). Hohlfeld et al. (Hohlfeld 1985) evaluated the course of MG in patients treated with long‐term azathioprine until they entered stable clinical remission or near‐remission for longer than six months. After stopping azathioprine, eight of 15 patients had a clinical relapse within one year. These 15 patients, and another five who were treated identically, were re‐evaluated three years later. All these patients originally had marked MG, with QMG scores ranging from 0.9 to 3.0. Four of these 20 patients relapsed within 27 to 70 months. All of the relapsed patients achieved remission again on combined immunosuppressive treatment including azathioprine, apart from one who died from pulmonary embolism after plasma exchange (Michels 1988). Therefore this group of patients can be defined as a treatment‐responsive, originally severely affected group.
At the time of writing, we know of three randomised trials involving azathioprine in MG. None of these studies had a placebo arm because this was considered unethical. In two of these, prednisolone was tested against azathioprine (Bromberg 1997; Gajdos 1993). In the trial by Gajdos et al. (Gajdos 1993), where 41 patients were randomised and followed up over seven years, there were fewer treatment failures in patients on azathioprine, compared with those on prednisolone. However, no difference in the duration of improvement was demonstrated. The other trial (Bromberg 1997) showed no benefit in azathioprine over prednisolone, but only 10 patients were studied for one year. The third study was a randomised double‐blind trial of prednisolone with azathioprine versus prednisolone alone in 34 patients who were followed up for three years (Palace 1998). The addition of azathioprine was associated with a reduced maintenance dose of prednisolone, and with fewer treatment failures and longer remissions. All the trials of azathioprine in MG have recruited fewer patients than planned due to the exclusion criterion of previous use of immunomodulatory treatment.
Tindall et al. (Tindall 1987) carried out a placebo‐controlled double‐blind randomised trial of ciclosporin for six months in 20, mainly elderly patients with recent onset MG who had not previously had thymectomy or immunosuppressive therapy. Ciclosporin‐treated patients had significantly greater improvement of strength and a significant reduction in AChR antibody titres compared with untreated patients. Another study of ciclosporin versus placebo by the same group (Tindall 1993) in 39 patients with severe MG and a wide age range, all of whom were on long‐term steroid treatment and the majority having had thymectomy, also demonstrated significantly greater increase in strength and greater reduction in anti‐AChR antibody titres in the ciclosporin‐treated group at the end of six months. In another randomised double‐blind, controlled trial of ciclosporin versus azathioprine in 40 patients, which was published only as an abstract, the improvement of MG was not statistically different between the two groups (Schalke 1988). In an uncontrolled trial of ciclosporin in 10 patients with severe MG, eight patients improved at 12 months of treatment (Goulon 1988). Two further uncontrolled studies suggested improvement with ciclosporin in six (Nyberg‐Hansen 1988) and nine patients (Bonifati 1997) respectively.
Perez and colleagues (Perez 1981) reported stable remissions in 42 patients on cyclophosphamide. Thirty‐three of these patients received concomitant steroid therapy. In another study, 23 patients with poor disease control or steroid‐related side effects were treated for 12 months with pulse intravenous cyclophosphamide or placebo in a randomised double‐blind fashion (De Feo 2002). Cyclophosphamide produced significant improvement in muscle strength compared with placebo at 12 months.
The preliminary results of three open label trials of mycophenolate mofetil in MG have been published. In one (Cos 2000), there were 22 MG patients treated for between two and 18 months. In the other two studies, 12 (Ciafaloni 2001) and 32 patients (Chaudhry 2001) respectively received treatment for 12 months. In all three trials, at least two thirds of patients improved to some degree. A retrospective analysis of the use of mycophenolate mofetil in 85 MG patients revealed improvement in 73% of cases (Meriggioli 2003a). In another double‐blind placebo‐controlled pilot study of mycophenolate mofetil in the treatment of suboptimally controlled, stable MG (Meriggioli 2003b), mycophenolate mofetil resulted in greater improvement than placebo when assessed by quantitative myasthenia gravis scores, manual muscle testing scores, AChR antibody titres, and changes in magnitude of single fibre electromyographic (EMG) abnormalities. However, differences were statistically significant only on EMG findings. There were only 14 patients in the trial. In a further analysis of the same trial that compared pre‐ and post‐treatment single fibre EMG jitter measurements in the 11 patients that completed the trial, there was significant improvement in jitter in the treated group compared with controls (Meriggioli 2003c).
Konishi et al. (Konishi 2003) treated 19 patients with generalised MG, who had previously undergone thymectomy, in a 16‐week open trial of low dose tacrolimus (FK506). Seven patients (37%) showed improvement clinically at the end of the study. Significant reduction of AChR antibody titres were observed in 11 patients (58%) at the end of the study.
There are problems in comparing treatment trials in MG. In many studies there is no consideration of whether patients have had thymectomy and it is not possible to extract from the data how many patients in a treatment arm have had thymectomy and how many have not. In addition, in non‐operated patients treated with immunosuppressive agents, it is unknown how many of them had thymoma at the time of inclusion into the study. In studies conducted before 1980, the percentage of patients with and without AChR antibodies is not known. Therefore, some of these patients without AChR antibodies may have had antibodies to other distinct antigenic targets, such as MuSK. In most studies conducted after 1980, AChR antibody positivity is one of the inclusion criteria. Finally, there are no controlled or prospective trials of immunosuppressive treatment in children and adolescents.
Objectives
We will review the efficacy of immunosuppressive therapies in MG.
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled and quasi‐randomised controlled trials.
Types of participants
Patients of any age with any type of autoimmune MG of all degrees of disease severity, with or without concomitant use of glucocorticosteroids or immunomodulatory therapies, including intravenous immunoglobulin and plasma exchange. Patients with or without thymectomy will be included.
Types of interventions
Any immunosuppressive agent used in the treatment of MG will be considered. Any randomised or quasi‐randomised controlled trial or branch of a trial evaluating the efficacy of one of these immunosuppressive drugs with or without immunomodulation versus placebo and/or glucocorticosteroids. This does not include treatment with glucocorticosteroids alone (Schneider 2004), plasma exchange (Gajdos 2002) or intravenous immunoglobulin (Gajdos 2003), which are the subjects of other Cochrane systematic reviews.
Types of outcome measures
In each of the randomised trials, different outcome measures were used for the evaluation of the efficacy of immunosuppressive agents in MG. Comparable outcome measures were used in only a few studies. We propose to transform data on weakness/fatigue from the various scores and scales into the simple categories of improved and not improved defined in 'Methods of the review' below.
Primary outcome measure
(1) Improvement or lack of improvement at six months.
Secondary outcome measures (where data are available)
(1) Improvement or lack of improvement at 12 months.
(2) Need for other treatment (e.g. dose of steroid or pyridostigmine) at six months.
(3) Number of episodes of worsening during the first year (deterioration of more than 20 per cent in the Quantitative MG Score for Disease Severity (QMG)).
(4) Serum acetylcholine receptor antibody titre after at least six months of therapy.
(5) The occurrence of one or more adverse events at any time after the introduction of treatment. If the time patients are at risk of such events varies with the study this will be taken into account in a weighted meta‐analysis.
Search methods for identification of studies
We will search the Cochrane Neuromuscular Disease Group Register, MEDLINE (from January 1966 to the present) and EMBASE (from January 1980 to the present) for randomised controlled trials using the search terms 'myasthenia gravis', 'azathioprine', 'cyclosporin A', 'cyclosporine A', 'ciclosporin', 'cyclophosphamide', 'methotrexate', 'mycophenolate mofetil', 'tacrolimus', 'FK‐506', 'immunosuppressive drug', 'immunosuppressive agent', and 'immunosuppression'. We will check the bibliographies of published randomised trials and contact the authors to identify additional published or unpublished data.
Data collection and analysis
Titles and abstracts identified from the register will be checked by two authors. The full text of all potentially relevant studies will be obtained for independent assessment by the authors. The authors will decide which trials fit the inclusion criteria and grade their methodological quality. Disagreements about inclusion criteria will be resolved by discussion between the authors. Data extraction will be performed onto a specially designed form by a single author and checked by the other two. Missing data will be obtained from the authors whenever possible.
The assessment of methodological quality of the trials will include concealment of allocation which we will grade using the Cochrane approach:‐ grade A: adequate, grade B: unclear, and grade C: inadequate. The methodological quality assessment will take into account patient blinding, observer blinding, exactness of the description of the clinical status, explicit outcome criteria, how studies deal with baseline differences between the experimental groups, and completeness of follow‐up. We will grade these items A: adequate, B: moderate risk of bias, and C: inadequate. If the information is unavailable, the item will be graded inadequate. If agreement is poor, we will reassess the studies and reach agreement by consensus.
We will try to perform a plausible transformation of the scoring system of each study into a simple categorised system: improved in the case of any improvement, and not improved in the case of no change in or worsening of the condition. These categories will be extracted from the individual papers by approximation of the relevant information whenever possible. The Myasthenia Gravis Foundation of America (MGFA) have recommended that to measure response to therapy, the Quantitative MG Score for Disease Severity (QMG) and MGFA Postintervention Status scales should be used (Jaretzki 2000). We will define improvement categories from these two scales.
The Quantitative MG Score for Disease Severity (QMG)
This scoring system is based on quantitative testing of sentinel muscle groups. The specific scoring system recommended (Barohn 1998) is a modification of earlier systems (Besinger 1983; Tindall 1987). Its interexaminer reliability has been confirmed (Barohn 1998). The QMG consists of 13 independent assessments (Table 1). Each item is graded 0 (normal), 1 (mild), 2 (moderate) and 3 (severe), with a total score ranging from 0 to 39.
| Items tested |
| Time to develop diplopia with prolonged lateral gaze |
| Time to develop ptosis with prolonged upgaze |
| Facial strength |
| Ability to swallow 4 oz (1/2 cup) of water |
| Speech after counting aloud from 1 to 50 |
| Time patient can outstretch either arm |
| Vital capacity |
| Right and left grip strength |
| Time patient can lift head 45° in a supine position |
| Time patient can outstretch either leg |
For calculation, grades of each item are added and divided by the number of items tested. On serial testing, slight alterations of QMG (within 0.3 score grades) will be considered as 'no change', a decrease of more than 0.3 as 'improvement' and an increase of more than 0.3 as 'worsening'. Using our transformation, a decrease of more than 0.3 will be categorised as improved, while not improved will include the subsets 'no change' and 'worsening'. When QMG is used, we will subdivide improved to 'mild improvement' (a reduction of 0.3 to 1.0) and 'marked improvement' (a reduction of more than 1.0).
MGFA Postintervention Status
This scale (Table 2) is designed to assess the clinical state of MG patients at any time after institution of treatment (Jaretzki 2000). The MGFA have recommended that criteria for change in a patient's status should be defined in each study protocol based on quantitative assessment of strength in pertinent or sentinel muscles, and what constitutes a sustained substantial change in medication should also be defined specifically in the protocol.
| Category | Clinical state |
| Complete Stable Remission (CSR) | No symptoms or signs of MG. Not received therapy for MG for at least one year. |
| Pharmacologic Remission (PR) | Same as above, but continues to take some form of therapy for MG. Patients on cholinesterase inhibitors are excluded as their use suggests the presence of weakness. |
| Minimal Manifestations (MM) | No symptoms of functional limitations, but has some weakness on examination. |
| Minimal Manifestations MM‐0 | Not received MG treatment for at least one year. |
| Minimal Manifestations MM‐1 | Continues to receive some form of immunosuppression, but no cholinesterase inhibitors or other symptomatic therapy. |
| Minimal Manifestations MM‐2 | Received only low‐dose cholinesterase inhibitors (<120 mg pyridostigmine/day) for at least one year. |
| Minimal Manifestations MM‐3 | Received cholinesterase inhibitors or other symptomatic therapy and some form of immunosuppression during the past year. |
| Change in status | |
| Improved (I) | A substantial decrease in pretreatment clinical manifestations, or a sustained substantial reduction in medications, or a specific decrease in QMG score, as defined in the protocol. |
| Unchanged (U) | No substantial change in pretreatment clinical manifestations or reduction in medications, or a change in QMG score, as defined in the protocol. |
| Worse (W) | A substantial increase in pretreatment clinical manifestations, or a substantial increase in medications, or a specific increase in QMG score, as defined in the protocol. |
| Exacerbation (E) | Fulfilled criteria of CSR, PR or MM, but subsequently developed clinical findings greater than permitted by these criteria. |
| Died of MG (D of MG) | Died of MG, of complications of therapy, or within |
Using our transformation, the category improved will comprise 'I' in Table 2, while not improved will encompass categories 'U', 'W', 'E' and 'D of MG' in Table 2.
Apart from the stipulated primary and secondary outcome measures we will compare the data of those studies in which identical, or similar primary or secondary outcome measures, have been used.
We will calculate a weighted treatment effect across trials using the Cochrane statistical package, Review Manager (RevMan). The results will be expressed as relative risks (RRs) with 95% confidence intervals (CIs) and risk differences (RDs) with 95% CIs for dichotomous outcomes, and weighted mean differences (WMDs) and 95% CIs for continuous outcomes. We will analyse all the primary and secondary outcomes under consideration. When the trials of the same immunosuppressive agent do not use the QMG or MGFA Postintervention Status scales, or when different scales are used in each study, a description of the trial results will be given.
We will define and analyse the following subgroups: patients with or without thymectomy, with or without steroid co‐medication, and with different types of disease severity as defined by the MGFA Clinical Classification (see Table 3) (Jaretzki 2000).
| Class I | Any ocular muscle weakness |
| Class II | Mild weakness affecting other than ocular muscles |
| Class IIa | Predominantly affecting limb, axial muscles, or both |
| Class IIb | Predominantly affecting oropharyngeal, respiratory muscles, or both |
| Class III | Moderate weakness affecting other than ocular muscles |
| Class IIIa | Predominantly affecting limb, axial muscles, or both |
| Class IIIb | Predominantly affecting oropharyngeal, respiratory muscles, or both |
| Class IV | Severe weakness affecting other than ocular muscles |
| Class IVa | Predominantly affecting limb and/or axial muscles |
| Class IVb | Predominantly affecting oropharyngeal, respiratory muscles, or both |
| Class V | Defined by intubation, with or without mechanical ventilation, except when employed during routine postoperative management. The use of a feeding tube without intubation places the patient in class IVb |
When the QMG is used, as a supplementary test we will subdivide the four groups (improved, not improved, immunosuppressive agent‐treated and placebo‐treated patients) into eight groups differentiating mild from marked improvement and no change from worsening.
We will search for relevant complications and adverse events that can be related to immunosuppressive treatment in non‐randomised literature (Dukes 2000). Since only a few studies address adverse events we do not specify an exact time frame in which an adverse event has had to have been recorded to be taken into consideration but appropriate adjustments will be made if the time at risk varies between studies.
We will discuss the cost and cost‐effectiveness of immunosuppressive agents in MG where there is information available to do this.
| Items tested |
| Time to develop diplopia with prolonged lateral gaze |
| Time to develop ptosis with prolonged upgaze |
| Facial strength |
| Ability to swallow 4 oz (1/2 cup) of water |
| Speech after counting aloud from 1 to 50 |
| Time patient can outstretch either arm |
| Vital capacity |
| Right and left grip strength |
| Time patient can lift head 45° in a supine position |
| Time patient can outstretch either leg |
| Category | Clinical state |
| Complete Stable Remission (CSR) | No symptoms or signs of MG. Not received therapy for MG for at least one year. |
| Pharmacologic Remission (PR) | Same as above, but continues to take some form of therapy for MG. Patients on cholinesterase inhibitors are excluded as their use suggests the presence of weakness. |
| Minimal Manifestations (MM) | No symptoms of functional limitations, but has some weakness on examination. |
| Minimal Manifestations MM‐0 | Not received MG treatment for at least one year. |
| Minimal Manifestations MM‐1 | Continues to receive some form of immunosuppression, but no cholinesterase inhibitors or other symptomatic therapy. |
| Minimal Manifestations MM‐2 | Received only low‐dose cholinesterase inhibitors (<120 mg pyridostigmine/day) for at least one year. |
| Minimal Manifestations MM‐3 | Received cholinesterase inhibitors or other symptomatic therapy and some form of immunosuppression during the past year. |
| Change in status | |
| Improved (I) | A substantial decrease in pretreatment clinical manifestations, or a sustained substantial reduction in medications, or a specific decrease in QMG score, as defined in the protocol. |
| Unchanged (U) | No substantial change in pretreatment clinical manifestations or reduction in medications, or a change in QMG score, as defined in the protocol. |
| Worse (W) | A substantial increase in pretreatment clinical manifestations, or a substantial increase in medications, or a specific increase in QMG score, as defined in the protocol. |
| Exacerbation (E) | Fulfilled criteria of CSR, PR or MM, but subsequently developed clinical findings greater than permitted by these criteria. |
| Died of MG (D of MG) | Died of MG, of complications of therapy, or within |
| Class I | Any ocular muscle weakness |
| Class II | Mild weakness affecting other than ocular muscles |
| Class IIa | Predominantly affecting limb, axial muscles, or both |
| Class IIb | Predominantly affecting oropharyngeal, respiratory muscles, or both |
| Class III | Moderate weakness affecting other than ocular muscles |
| Class IIIa | Predominantly affecting limb, axial muscles, or both |
| Class IIIb | Predominantly affecting oropharyngeal, respiratory muscles, or both |
| Class IV | Severe weakness affecting other than ocular muscles |
| Class IVa | Predominantly affecting limb and/or axial muscles |
| Class IVb | Predominantly affecting oropharyngeal, respiratory muscles, or both |
| Class V | Defined by intubation, with or without mechanical ventilation, except when employed during routine postoperative management. The use of a feeding tube without intubation places the patient in class IVb |
