Treatment of the Critically Ill Patient with Myasthenia Gravis

Myasthenia gravis (MG) is an uncommon disorder of immune pathogenesis that affects neuromuscular transmission and causes regional and generalized fatigable weakness. Its prevalence is 0.5 to five per 100,000 population (1). The effectiveness of anticholinesterase medications in ameliorating the symptoms was discovered in the 1930s by Walker, and the relationship of thymectomy to clinical improvement was first noted in 1939 by Blalock. Simpson proposed the role of the immune system in 1960, and antibodies to the acetylcholine receptor were discovered in 1973 by Patrick and Lindstrom.

The event that initiates the production of autoimmune acetylcholine receptor antibodies is unknown, but the increased incidence of autoimmune thyroid disease (10%), systemic lupus erythematosus, rheumatoid arthritis, pernicious anemia, and idiopathic thrombocytopenia in patients with myasthenia and their relatives suggests a more pervasive problem in the immune system. Furthermore, a genetic predisposition exists in young women with human leukocyte antigen types A1, B8, DRW3, and B12 (2). There are no important geographic factors, and curiously, few familial cases have been reported (3). The disease most commonly occurs in two age groups: young women (15 to 30 years old) and older men (60 to 75 years old) (4,5). About 10% of patients manifest a thymoma; most other patients have thymic enlargement owing to hyperplasia. Thymomas tend to occur in the older age group, and the long-term prognosis for these patients is far worse because of the severity of their myasthenia regardless of type of treatment or the extent of spread of the thymoma (6).

The relative importance of the roles of celland antibody-mediated damage to the acetylcholine receptor have not yet been determined, but it is clear that the main clinical feature of the disease, fluctuating skeletal muscle weakness, is related largely to binding of IgG antibodies to the postsynaptic acetylcholine receptor. There is destruction of the receptors and a reduction in the surface area of the postsynaptic membrane (7). In myasthenia, the normal process of receptor degradation is reduced from 7 days to less than 48 hours by a complement-mediated destruction and phagocytosis.

In the last two decades, treatment with corticosteroids, immunosuppressive medications, plasma exchange, and intravenous immunoglobulin has been added to the standard approach with anticholinesterase drugs, thymectomy, and advanced respiratory care to improve greatly the prognosis of patients with MG (8). At Columbia-Presbyterian Medical Center during the 1940s and 1950s, the mortality in patients with MG was greater than 30% (9). This declined to 12% during the 1960s, 3.3% during the 1970s, and even lower in the 1980s and 1990s (10).

CLINICAL FEATURES

The fluctuating nature of myasthenic weakness, truly a work-related fatigue, makes it a unique clinical disorder. Patients complain less of fatigue than of weakness
and this weakness is regional in the early stages of illness. In the majority, ocular muscles are affected first, resulting in ptosis and diplopia in about 40%; these muscles are ultimately affected in 90% of patients. The facial and oropharyngeal muscles are also commonly affected, resulting in dysarthria, dysphagia, and facial diplegia. On occasion, swallowing is impaired before the ocular signs appear. Virtually all patients with MG have some involvement of oropharyngeal and ocular muscles at some time during the course of illness. Isolated limb or diaphragm weakness is rare. An exacerbation with generalized weakness that includes respiratory muscles to the point of ventilatory failure is termed “myasthenic crisis.” Thirty percent of patients with MG develop some degree of respiratory muscle weakness, and 15% to 20% experience at least one episode of “crisis.” Numerous series of such patients with acute worsening have been reported, all emphasizing a high rate of morbidity and mortality (see the following for treatment of myasthenic crisis) (11, 12, 13 and 14).

TABLE 19.1. Modified Osserman classification of myasthenia gravis

Class 1	Patients with ocular involvement alone
Class 2	Mild weakness, not incapacitating, without oropharyngeal or respiratory muscle involvement
Class 3	Moderate weakness, not incapacitating, including oropharyngeal and respiratory muscle weakness
Class 4	Incapacitating weakness of any muscle system, including oropharyngeal and respiratory muscle weakness
Class 5	Life-threatening respiratory insufficiency requiring ventilatory assistance (crisis)

TABLE 19.2. Characteristics of myasthenic crisis

	1960-1980^a	1983-1994^b
No. patients	61	53
Average age	38 yr F; 62 y M	55 yr; 2:1 F:M
Duration before crisis (range)	21 mo, median; (2 mo-8 y)	8 mo, mean (20-82 wk)
Crisis within 1 year of onset of myasthenia gravis	36%	53%
Duration of mechanical ventilation (range)	2 wk median	13 d; (1 d-5 mo)
Mortality	42%	4% before extubation; 6% after extubation
^aData from Cohen MS, Younger D. Aspects of the natural history of myasthenia gravis: crisis and death. Ann NY Acad Sci 1981;377:670-677. ^bData from Thomas CS, Mayer SA, Gungor Y, et al. Myasthenic crisis: clinical features, mortality, complications and risk factors for prolonged mechanical ventilation. Neurology 1997;48:1253-1260.

A practical clinical classification for the severity of MG was devised by Osserman in 1958 and is still used with modifications by most centers that specialize in the treatment of MG (Table 19.1) (15,16).

The specialist in critical care can expect to treat severely ill patients in class 4 or 5. Only rarely does such a patient enter the hospital without a previously known diagnosis of MG. Most patients have been symptomatic for at least several months before developing respiratory compromise (Table 19.2). On occasion, a patient with unsuspected MG who has acute respiratory failure of unknown cause (17, 18 and 19) is admitted from the emergency room or transferred from another hospital. Rare instances of myasthenia presenting as stridor or isolated respiratory failure have been reported and may come to the attention of the intensivist (20). Another such idiosyncrasy occurs during obstetric delivery or general surgery, in which a previously undiagnosed myasthenic patient may develop respiratory failure as a result of the administration of neuromuscular blocking drugs,
agents known to greatly exaggerate the neuromuscular failure of MG (21).

Many other medications are known to worsen the symptoms of MG (22,23) but their propensity to uncover a previously unsuspected case is low in comparison to the neuromuscular blocking drugs. In this regard, the risk of the use of aminoglycoside and related antibiotics is often raised in the treatment in urinary or pulmonary infections. There is an undoubted effect on strength in marginally compensated individuals but most such patients are already intubated and the drugs can be used freely. In others less affected, surveillance for respiratory failure should be enhanced. Other medications, particularly β-blockers, present a similar but lesser risk.

Diagnosis

The diagnosis of MG depends on the presence of the appropriate clinical syndrome combined with electrodiagnostic features of impaired neuromuscular transmission (decremental response with repetitive nerve stimulation or increased “jitter” and blocking of impulses with single-fiber electromyography), a clinical response to cholinergic drugs [edrophonium (Tensilon)] or neostigmine test, and the presence in the serum of antibodies against the acetylcholine receptor. The sensitivity of these tests varies with the severity of the disease and the level of experience of the laboratory. The most frequently employed technique is repetitive stimulation of a muscle to elicit a decremental contraction. An increase in the variability of the timing of firing of linked muscle fibers within a motor unit, called “jitter,” can be demonstrated in 90% of weak muscles by the use of single-fiber electromyography. This reflects intermittent blocking of impulses at individual neuromuscular junctions. Acetylcholine receptor antibodies are present in high titer in 90% of patients with generalized or longstanding myasthenia, but in only 50% of patients with ocular signs alone (24). Restated, at least 10% of patients with typical MG are “seronegative” (25,26). The intravenous edrophonium test most clearly, but only briefly, ameliorates weakness (i.e., is positive) in patients with ocular signs (ptosis and strabismus) but the results are more difficult to assess in patients with limb weakness or respiratory failure alone. An alternative test using neostigmine, 2 mg intramuscularly is generally preferable in patients with respiratory failure or unusual patterns of limb weakness. The action of intramuscular neostigmine reaches a peak at 15 to 30 minutes, allowing for repeated clinical examination and measurements of vital capacity (VC) and maximal expiratory pressure in order to detect a definite increment.

In the critical care setting, the differential diagnosis of myasthenic crisis is limited mainly to other neuromuscular diseases that cause oropharyngeal and respiratory muscle weakness, namely Lambert-Eaton myasthenic syndrome, periodic paralysis, Guillain-Barré syndrome—particularly the oropharyngeal variant (Chapter 18), mitochondrial myopathies, muscular dystrophy, polymyositis, motor neuron diseases, botulism, organophosphate poisoning, and poison snake or spider bites (27,28). Central nervous system diseases with prominent brainstem features and processes that affect multiple cranial nerves may mimic the oculopharyngeal features of MG. Two such examples are a 35-year-old woman who underwent thymectomy for presumed myasthenic bilateral ptosis and ophthalmoparesis and was later found to have bilateral giant cavernous-carotid aneurysms; and a 68-year-old man who developed dysphagia and dysarthria followed by respiratory failure. He required intubation and mechanical ventilation, was treated with intensive plasma exchange, and improved over a 2-week period. After extubation, magnetic resonance imaging (MRI) of the brainstem revealed a cavernous angioma with recent hemorrhage in the medulla.

Myasthenic Crisis

This aspect of the disease is most likely to involve the neurointensivist. Two of the most extensive and useful clinical analyses of
myasthenic crisis are from the Columbia-Presbyterian Medical Center and in most ways reflect our general experience. Most points regarding the problem can be gleaned from a record review of 447 myasthenic patients between 1960 and 1980 (29), and a more recent survey comprising 53 patients with crisis (30). Of patients seen during the earlier 20-year period, 16% (61 patients) had at least one episode of crisis. Among 36 women, crisis occurred at a mean age of 38 years; the first crisis in men occurred later at a mean age of 62 years; the average age was 55 years in the later series. The interval between onset of myasthenia and the first crisis is shown in Table 19.2. The often-stated dictum, concordant with our own experience, that most crises occur within the first year of illness, was found in the more recent series of Thomas and colleagues (8 months average duration of illness before crisis) (30). One third of their patients had more than one episode of crisis, again usually within a year of the first. In up to one half of the episodes, there was no obvious cause for the exacerbation other than progressive myasthenic weakness. However, many patients had severe dysarthria and dysphagia at the time. In others, overt pneumonia or other respiratory infection precipitated respiratory failure but not always severe generalized weakness. Four patients in the older series had sudden cardiorespiratory arrest, two had respiratory depression secondary to sedative drugs, two had pulmonary embolism, and two patients, not known previously to have MG, were given curare during elective surgery and had prolonged ventilatory failure. Therefore, many of these cases did not conform to the usual definition of myasthenic crisis but were operationally equivalent because they resulted in admission to the intensive care unit (ICU).

In the later series, the mean VC just prior to intubation was high, 27 mL/kg, presumably reflecting the need to intubate some patients for protection of the airway rather than for ventilatory failure. Positive-pressure ventilation was required for varying periods, ranging from several hours to 1 year. In the more modern series, half of patients were extubated by 13 days, and 75% by 31 days. It is notable that three patients died from cardiac arrest during tracheostomy, presently a rare occurence.

Steroid therapy was being administered to 16 patients at the onset of crisis. These agents, used as treatment for crisis, were started for the first time in 11 patients who had been on a respirator for more than 2 weeks. In a retrospective analysis, it was impossible to ascertain if steroids had a beneficial effect in weaning the patient off the respirator.

The mortality rate of crisis has declined in recent years; presumably this reflects modern critical care techniques and possibly the introduction of plasma exchange and immunoglobulin therapy (see the following). In a review of patients by Chang and Fink from 1983 to 1991 (31), there had been no deaths from crisis, but in subsequent years there certainly have been several. In the later Columbia series the mortality rate was 10% overall, half of those dying after extubation. It is reemphasized that although plasma exchange and immune globulin are routinely used in the treatment of crisis, their efficacy remains uncertain, as discussed more extensively in the following.

Cholinergic Crisis

The frequency of cholinergic crisis in MG is somewhat uncertain. These crises indisputably occur with overdosage of cholinesterase-binding organophosphate toxins. Whereas much was made of the problem in the older literature, recent experience suggests their importance is overrated (32) in all likelihood because of the more cautious use of anticholinesterase drugs. The types and equivalent dosages of the cholinesterase inhibitors in clinical use are listed in Table 19.3.

Osserman and Genkins (15,16) described the syndrome as a worsening of myasthenic weakness caused by depolarization blockade of certain muscle groups from excessive anticholinesterase administration. To optimize the dosage of medications, they advocated an
edrophonium test. As originally described, 1 or 2 mg of edrophonium is injected intravenously. If weakness improved, true “myasthenic” crisis was assumed to be the problem, and the dosage of cholinesterase inhibitor was increased. If there was no improvement or if weakness worsened, “cholinergic” crisis was assumed, and the dosage of cholinesterase inhibitor was decreased. This technique is now little used in the management of patients with MG who have an exacerbation but it may have some utility in selected cases where the distinction is unclear. More often, patients are so weakened that they require intubation in any case, and the drugs are temporarily withdrawn.

TABLE 19.3. Cholinergic drug dosage equivalents and duration of action

	Equivalent dose	Onset	Maximum response
Pyridostigmine (Mestinon)	60 mg (p.o.)	40 min	1 h
Neostigmine (p.o.)	15 mg (p.o.)	1 h	1.5 h
Neostigmine (i.m.)	1.5 mg (i.m.)	30 min	1 h
Neostigmine	0.5 mg (i.v.)	Immediate	20 min
i.m., intramuscular; i.v., intravenous; p.o., oral.