Semin Neurol. 2004;24(1) 

In This Article

Immunomodulating Therapies


Corticosteroids can suppress the immune system in several ways; the precise explanation for benefitin MG is unknown. Prednisone has been shown to reduce acetylcholine receptor antibody levels and this decrease often correlates with clinical improvement.[3] The ideal prednisone dose has not been established, but initial doses are usually 60 to 100 mg per day. The dose can be calculated at 1 to 1.5 mg/kg, but an initial dose of 100 mg per day is often recommended in adults. After 2 to 4 weeks, change to a dose of 100 mg every other day. This helps minimize side effects.[4] Most patients tolerate alternate-day dosing, but some feel poorly on the off day and feel better if they resume daily dosing. Patients with diabetes are better treated with the same prednisone dose each day to avoid fluctuations in blood sugar.

After beginning corticosteroid therapy at moderate to high doses, improvement is typically seen within 2 to 3 weeks.[5,6] A problem with starting treatment at high doses is that as many as 50% of patients experience a transient deterioration.[5,6,7] The exact cause of this is unknown but may involve prednisone-induced worsening of neuromuscular junction function.[6] This worsening can be serious; in one study, almost 10% of patients who experienced deterioration required intubation.[5] If an MG patient has significant weakness, high-dose prednisone therapy should only be initiated in the hospital where respiratory and bulbar function can be closely monitored. Unfortunately, deterioration can occur for as long as a week after initiation of corticosteroids. It is sometimes difficult to obtain approval for a hospitalization unless a patient already has significant weakness and requires plasmapheresis or IVIg (see below). For patients with milder baseline weakness, corticosteroids can be started at low doses and then built up slowly.[8] Therapy can be initiated with 10 to 20 mg per day and increased by 5 mg every 3 to 5 days. While this will minimize the risks of deterioration, the onset of improvement will be significantly prolonged. After reaching the target daily dose of 100 mg per day (within 6 to 8 weeks), the patient can be switched to alternate-day dosing.

A study from the University of Virginia has provided very good data regarding the efficacy of prednisone in MG.[5,6] In the treatment of 116 patients with prednisone, a marked improvement was seen in 52% (minor symptoms and return to activities of daily living), a moderate improvement in 15% (functional limitations), no improvement in 5%, and pharmacologic remission (asymptomatic on medication) in 28%. The mean time to maximum prednisone benefit was 5.5 months (range: 2 weeks to 6 years). This study also pointed out that only 14% of patients were able to be completely taken off of prednisone.

After improvement is achieved, the prednisone dose should be decreased. It is important not to taper prednisone too soon or too quickly. As with dose, there are no established guidelines for this. In general, patients can be maintained on 100 mg every other day for 2 to 4 months before tapering. It may take a week or two for deterioration to occur following a decrease in corticosteroid dose. Therefore, taper by only 5 mg per day every 2 weeks. When the prednisone dose is decreased to 20 mg every other day, taper even more slowly. Although we try to get patients completely off of prednisone, many patients will need to be maintained on low doses (5 to 10 mg every other day) for years or indefinitely.

Corticosteroids have several advantages in the treatment of MG. They are inexpensive and have a quick onset of action and an established track record. The main drawback is their adverse effects, ranging from cosmetic (weight gain, cushingoid facies) to serious medical problems (e.g., infection, diabetes mellitus, hypertension, osteoporosis, depression, psychosis). Although side effects cannot be avoided entirely, there are some measures that can help minimize these. One is using every other day dosing. To minimize weight gain, patients are placed on a calorie-restricted, low-carbohydrate, low-sodium diet. All patients are placed on calcium (1500 mg per day) and vitamin D (400 to 800 IU per day) supplementation to minimize bone mineral loss.[9] Bone density is measured at baseline and, if satisfactory, every 6 months. If osteopenia develops, patients are treated with a bisphosphonate, such as alendronate.[10] Histamine-2 blockers or proton pump inhibitors do not need to be prescribed routinely, rather only when symptoms of gastric irritation develop. Serum electrolytes, glucose, and blood pressure need to be monitored regularly.


Azathioprine (Imuran) is an antimetabolite that blocks cell proliferation. Inhibition of T lymphocytes is the presumed mechanism for benefit. Acetylcholine receptor antibody levels are decreased with azathioprine treatment.[11]

Azathioprine is used most often in patients who have relapsed while on prednisone or as a steroid-sparing agent in patients who have been taking prednisone for long periods of time. Several clinicians start both corticosteroids and azathioprine concurrently. We do not use azathioprine as a first-line immunosuppressive agent instead of prednisone. This practice is more common in Europe.[11,12,13,14]

In retrospective studies of azathioprine therapy, 70 to 90% of MG patients improve whether the drug is used as a first- or second-line therapy.[11,12,13,14,15] However, the response is slow, ranging from 3 to 12 months. In an important double-blind study comparing the use of oral prednisolone versus prednisolone plus azathioprine, patients receiving both medications had fewer relapses, a higher incidence of remission, and could ultimately be managed on a lower corticosteroid dose.[16] The beneficial effects of azathioprine were not seen until after 18 months. On the basis of this data, corticosteroids and azathioprine are often combined in the initial treatment of patients with MG. We frequently still use prednisone monotherapy, especially in patients without risk factors for steroid-induced morbidity (e.g., obesity, diabetes, hypertension, osteopenia). This controlled study underscores the long latency before clear benefit from azathioprine may be seen. Therefore, for patients in whom a more rapid effect is needed, other treatments should be considered.

Azathioprine is started at 50 mg per day. If, after 1 week, there are no systemic side effects, the dose is increased to 2 to 3 mg/kg per day. Therefore, after the initial test dose, most MG patients are placed on 150 to 200 mg per day. Azathioprine is usually well tolerated, but there are a few limiting adverse effects. Within the first several weeks of treatment, ~10% of patients will have an idiosyncratic reaction consisting of fever, anorexia, nausea, vomiting, and abdominal pain.[17] Patients feel as if they have the flu. These symptoms resolve quickly after the drug is stopped. The same symptoms usually return if the patient is reexposed to azathioprine. Leukopenia and hepatotoxicity are important adverse effects. White blood cell counts and liver function tests should be monitored monthly. If the white blood cell count decreases below 4000 cells/mm3, the dose should be decreased. If the count falls below 3000 cells/mm3, azathioprine should be stopped. Medication should also be held if liver enzymes become significantly increased. In these circumstances, patients can be rechallenged with azathioprine, although in many patients, recurring toxicity will require discontinuation of this drug. Following long-term use of azathioprine, patients are at increased risk of developing malignancies.[18] Azathioprine is teratogenic and can impair fertility in women.


Cyclosporine (Sandimmune, Neoral), a drug designed to prevent rejection in organ transplantation patients, became popular for the treatment of myasthenia gravis in the 1990s. This drug inhibits helper T lymphocytes, facilitates suppressor T lymphocytes, and blocks the production and secretion of interleukin-2. Cyclosporine is unique among MG therapies in that it is the only treatment for which successful randomized, blinded, placebo-controlled trials have been performed.[19,20] Tindall and colleagues studied this drug in both relatively new-onset MG patients[19] and in corticosteroid-dependent patients.[20] In both studies, cyclosporine was more effective than placebo in improving clinical assessments and in lowering acetylcholine receptor antibody levels. The corticosteroid dose could be reduced with cyclosporine therapy.[19]

Clinical benefit can be seen 1 to 2 months after starting cyclosporine. The time to effect is slower than that seen with prednisone but faster that seen with azathioprine. The dose of cyclosporine is 4 to 6 mg/kg per day in two divided doses. Side effects include hirsutism, tremor, gum hyperplasia, paresthesias, anemia, and hepatotoxicity. However, hypertension and nephrotoxicity are the main treatment-limiting adverse effects of this therapy. Blood pressure, renal function, and trough plasma cyclosporine levels need to be followed monthly. The risk of malignancy with long-term use may also be increased.[21] Although effective, the risk of nephrotoxicity and hypertension has been significant in the replacement of cyclosporine with mycophenolate mofetil (MyM).

Mycophenolate Mofetil

MyM (CellCept) is another immunosuppressive agent used in organ transplantation. By selectively blocking purine synthesis it suppresses both T- and B-cell proliferation. A multicenter, double-blind, placebo-controlled study of MyM as a steroid-sparing agent is currently ongoing. However, based on retrospective case series[22,23,24] and a small, pilot, double-blind, placebo-controlled study,[25] MyM seems to be an effective adjunct therapy for MG. It appears to improve function and allow for decreased doses of prednisone and pyridostigmine.[24] Onset of benefit is similar to cyclosporine. In one large retrospective review, the mean time to objective improvement was 10 weeks (range: 4 to 40 weeks).[24] The mean time to maximal objective improvement was 27 weeks (range: 8 to 104 weeks).

Standard doses are 2000 to 3000 mg per day in two divided doses. Patients can be started on 500 mg twice a day and, after 1 to 2 weeks, increased to 1000 mg twice a day. MyM is well tolerated in most patients. The main adverse effects are gastrointestinal, primarily diarrhea. In transplant patients, myelosuppression is common, but in MG patients taking lower MyM doses, this does not seem to be a frequent problem.[22,23,24] Nevertheless, check blood counts monthly. In contrast to azathioprine and cyclosporine, MyM may not increase the risk of malignancy.[26] MyM appears to have efficacy similar to cyclosporine but with much less toxicity.


Cyclophosphamide (Cytoxan) is a nitrogen mustard alkylating agent that blocks cell proliferation, affecting both T and B cells. There are limited data regarding the benefits of cyclophosphamide in MG.[27,28] Cyclophosphamide is a potent immunosuppressive medication, but due to significant potential adverse effects, such as myelosuppression, hemorrhagic cystitis, and an increased risk of malignancy, it is not typically used in MG. As outlined above, there are several effective, less toxic immunosuppressive agents available. Cyclophosphamide may be of benefit in MG patients refractory to other treatments.[28] Recently, a novel approach to treatment of refractory patients has been proposed. Drachman and colleagues used a one-time course of high-dose cyclophosphamide to "reboot" the immune system.[29] Stem cells are resistant to cyclophosphamide, allowing the immune system to be reconstituted after high doses of cyclophosphamide that ablate cellular components in the circulation. Based on prior success in animal MG models and in the treatment of other autoimmune diseases in humans, Drachman and colleagues treated three refractory MG patients.[29] All three tolerated the procedure well, showed significant clinical improvement, and were able to decrease the dose of other immunosuppressive medications. Benefit has persisted for several years. The best approach to treating refractory MG patients is uncertain. Further studies are needed to compare chronic conventional immunosuppressive medications, repeated pulse-dose cyclophosphamide, and one-time high-dose cyclophosphamide. The long-term benefits and risks of these therapies need to be determined.


Plasmapheresis, a procedure that removes acetylcholine receptor antibodies from the circulation, is an effective treatment for MG.[30] However, it is surprising that a controlled trial of plasmapheresis in MG has never been performed. Improvement following plasmapheresis occurs within a few days, much faster than for other immunomodulating therapies. Plasmapheresis is an established therapy for patients in myasthenic crisis. It is also often used to improve strength (if necessary) in patients prior to undergoing thymectomy. Plasmapheresis is also used for the treatment of weak patients admitted to the hospital for initiation of corticosteroids. Infrequently, chronic, intermittent plasmapheresis is used in patients refractory to other therapies.

The effects of a course of plasmapheresis last only several weeks. Plasmapheresis can be performed through peripheral veins, but usually a large-bore, double-lumen subclavian catheter is necessary. The risks of a chronic indwelling catheter (pneumothorax in the short term and infection and thrombosis in the long term) make chronic plasmapheresis a relatively unattractive long-term treatment option. A course of plasmapheresis usually consists of four to six exchanges, removing 3 to 5 L of plasma each treatment, performed daily or every other day. There is no precise prescription governing the number, volume, and frequency of exchanges. The course of plasmapheresis is dependent to a large extent on how well a patient tolerates the procedure and how quickly and thoroughly the patient responds. Plasmapheresis produces large fluid shifts and patients are susceptible to hypotension and, in individuals at risk, myocardial infarction.

Intravenous Immunoglobulin

Although clear efficacy demonstrated by a placebo-controlled trial is lacking, there is experimental and anecdotal evidence for benefit of IVIg in MG.[31,32,33,34,35,36,37] The response rate is ~60 to 70%, and improvement occurs within days to weeks. The specific mechanism of action of IVIg with respect to MG is not known.[38]

A randomized study comparing IVIg to plasmapheresis in 87 MG patients with exacerbations found the two therapies to be equally effective; IVIg had fewer side effects.[39] However, a retrospective series found plasmapheresis to be superior to IVIg in patients in myasthenic crisis.[40] There is a report of four patients in myasthenic crisis who had no response to IVIg but who improved with plasmapheresis.[41] A double-blind, placebo-controlled trial of IVIg in mild to moderately weak outpatients was attempted, but enrollment was stopped prematurely due to a lack of available IVIg.[42] This underpowered study did not find any significant differences; trends toward improvement with IVIg were seen in an open-label portion. Our anecdotal experience is that patients respond more quickly to plasmapheresis. Therefore, we generally prefer this treatment to IVIg for myasthenic crisis.

The initial dose of IVIg for MG and other neuromuscular diseases is usually 2 g/kg. This is generally administered over 2 to 5 days. Slower rates of infusion are preferable in older patients and those with renal insufficiency or congestive heart failure. Intravenous immunoglobulin is often used for treatment of MG outpatients refractory to other immunomodulating therapies. In this situation, following a 2 gm/kg dose, two or three additional infusions of 0.4 g/kg are given at monthly intervals. After these, reevaluate to determine if further treatments are needed. Because of expense, insurance companies are often reluctant to approve IVIg for this indication.

The advantages of IVIg are the low risk of side effects compared with immunosuppressive medications and plasmapheresis. Most adverse effects are related to the rate of infusion and include headache, light-headedness, and chills. However, nephrotoxicity can occur in patients with underlying renal insufficiency and hypertension.[43] This appears to be related to the osmolality and depends largely on sugar content, which varies among brands of IVIg. Although uncommon, significant thrombotic events, including myocardial infarction and stroke, may occur.[44,45]


In 1939, Blalock and colleagues[46] reported the remission of generalized MG in a 21-year-old woman following the removal of a thymoma. Since then, thymectomy, with or without the presence of thymoma, has gained widespread acceptance as a form of treatment for MG. Thymectomy was therefore the first attempt at "immunotherapy" for MG and continues to be one of the most frequently utilized forms of treatment for this disease. Key considerations regarding thymectomy include which patients are most likely to respond, when thymectomy should be performed, whether all myasthenics should undergo this procedure, and whether the type of surgical procedure is important.

Gronseth and Barohn[47] recently reviewed 21 thymectomy series from 1953 to 1998, involving 8490 MG patients. Patients who underwent thymectomy were more likely to achieve medication-free remission, become asymptomatic, or at least show clinical improvement compared with patients who did not have thymectomies (median relative rates: 2.1, 1.6, and 1.7, respectively). However, among the patients undergoing thymectomy, the median rate of remission, asymptomatic state, or improvement were only 25, 39, and 70%, respectively (the mean rates are ~50% lower for each). Therefore, although a patient undergoing thymectomy may be two times more likely to experience improvement, the majority of patients having this surgery will not experience remission or become completely asymptomatic. A recent multivariate analysis involving a total of 756 MG patients found thymectomy to be significantly associated with remission; however, the odds ratio was only 1.6.[48]

Even in the most favorable reports, the response to thymectomy was not immediate; patients and physicians need to be aware of this. In one early study, among patients who obtained remission, only 25% achieved remission in the first year, 40% the second year, and 55% in the third year.[49] Other series have shown that the "benefit" of a remission from thymectomy may take as long as 7 to 10 years.[50,51] These prolonged latencies to improvement raise the question of whether other confounding factors might account for the perceived benefit.

Thymectomies can be done in children but probably should be avoided during the first few years of life.[50] Whether or not ocular MG patients should have thymectomies is controversial.[52,53] We generally do not advise thymectomy for these patients. In addition, thymectomy is probably less effective in the elderly due to atrophy of the thymus. An age limit for thymectomy has not been established, and some authors do not regard age as exclusionary. We typically do not recommend thymectomy to patients over 60.

For the reasons outlined above, there is much uncertainty about the benefit of thymectomy. In a 1990 survey of 56 experts, only three expressed no reservations in recommending thymectomy for patients with generalized MG.[54] For several years, there has been a recognized need for randomized studies of thymectomy.[47,52,55,56] A multicenter, multinational study is currently being planned.[57] Our current practice is to start treating an MG patient with pyridostigmine and, if necessary, a course of immunomodulating therapy. If patients do not show a good response or cannot be decreased to a small dose of medication, then thymectomy is considered. Some authors suggest that thymectomy is most effective when performed soon after diagnosis,[58,59] but significant methodological issues limit interpretation of those data.[47] We tend to favor thymectomy for newly diagnosed young women, although, admittedly, this is not based on firm empirical data. We do present the option of thymectomy to all patients. We explain there may be an increased likelihood of improvement, describe the limitations of available data and point out that most patients, despite surgery, will not go into remission or become entirely asymptomatic.

Another controversy is whether a sternal splitting surgical approach is superior to techniques offering less exposure, such as transcervical approach or fiberoptic thorascopy. A similar lack of controlled data hampers comparison of different surgical procedures.[60,61,62,63,64,65] With current techniques, surgical mortality and morbidity from thymectomy are each less than 1%.[47]

The presence of a thymoma is, of course, the one absolute indication for thymectomy. All newly diagnosed MG patients should have a computed tomography scan or magnetic resonance imaging of the chest to look for a thymoma. Chest X-rays are not sensitive enough to exclude thymoma.[66] Interestingly, MG may sometimes develop weeks or months after removal of a thymoma.[67]

Other Immunosuppressant Therapies

Tacrolimus, a transplant medication similar in mode of action and toxicity to cyclosporine, has been found effective in MG in few small series.[68,69] It may have less nephrotoxicity than cyclosporine.[70] Rituximab, a monoclonal antibody directed against antigens on B cells, devised to treat lymphomas, has theoretical applicability to the treatment of MG and is relatively safe. There are two case reports describing benefit in MG.[71,72]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.