Acquired neuromuscular disorders have become increasingly recognized over the last two decades as a major cause of morbidity related to critical illness. Currently, acquired disorders are a much more common cause of severe generalized weakness in the ICU than are primary neuromuscular disorders.66 Although ICU-acquired muscle weakness is usually reversible, respiratory muscle involvement may lead to prolonged mechanical ventilation and delayed weaning. Once patients are successfully extubated, they often require prolonged physical rehabilitation and may be unable to perform basic activities of daily living for weeks to months.67 Of greater concern, recent studies have documented the persistence of significant weakness up to several years after hospital discharge, indicating that in some cases severe ICU-acquired weakness may result in permanent disability.68–71 Without question, the medical, economic, and psychosocial costs of ICU-acquired neuromuscular weakness represent a major problem in the current practice of critical care medicine.
Causes of Intensive Care Unit–Acquired Weakness
Three basic causes of ICU-acquired neuromuscular weakness have been identified: (1) persistent blockade of the neuromuscular junction after discontinuing a neuromuscular blocking agent (NMBA), (2) a sensorimotor axonal polyneuropathy, and (3) an acute myopathy.72–74 Although the clinical setting and physical examination may be of some help in elucidating the underlying cause of muscle weakness, in most cases diagnosis has been made by electrophysiologic studies, sometimes supplemented by biopsy of muscle or nerve. As will be discussed below, elucidation of the underlying cause of weakness is not always easy, and combined disorders of muscle and nerve may occur.75
Persistent Neuromuscular Junction Blockade
Residual blockade of the neuromuscular junction after discontinuation of an NMBA is rarely responsible for persistent neuromuscular weakness in the ICU.66 One possible exception is the use of vecuronium in patients with renal failure, in which case accumulation of the active metabolite 3-desacetylvecuronium may result in paralysis that is prolonged for several days.74 If residual neuromuscular paralysis is a consideration, a repetitive nerve stimulation protocol should be performed to see if there is a progressive decrement in the compound muscle action potential (CMAP) that would indicate residual blockade at the neuromuscular junction.
Critical Illness Polyneuropathy
In the 1980s Bolton and colleagues reported on a group of critically ill patients who developed generalized weakness after several weeks in the ICU.72,76 Affected patients, most of whom had underlying sepsis, often underwent neurologic evaluation because of difficulty in weaning from mechanical ventilation or because of diffuse limb weakness. Distal muscles were often affected most prominently, but cranial nerves did not appear to be involved. Sensory examination was difficult to perform, but deep tendon reflexes were either reduced or absent in most cases.72,76 Electrophysiologic evaluation with electroneurography (ENG) and EMG suggested that weakness was due to an axonal sensorimotor polyneuropathy. Key findings included a reduction in the CMAP in response to stimulation of motor nerves, often accompanied by a decrease in the sensory action potential (SAP) when sensory nerves were stimulated. The EMG revealed spontaneous fibrillations and positive sharp waves that were attributed to denervation. Survivors had improvement in their electrophysiologic studies over several months that paralleled clinical recovery. Autopsy findings included grouped fiber atrophy in muscle and axonal degeneration with loss of myelin in nerves, with distal segments most severely affected.76 Neither muscle nor nerve showed inflammation.76
Numerous additional studies dealing with critical illness polyneuropathy have been published in the last 20 years, and the reported findings have been similar to those initially described, including the diffuse distribution of weakness (other than the facial muscles), frequent difficulties with weaning, and occurrence in the setting of sepsis and systemic inflammatory response syndrome (SIRS), often with multiorgan dysfunction. The pathogenetic mechanism of acquired polyneuropathy in critical illness is not understood. It has been suggested that the peripheral nervous system may be one of the many target sites for tissue damage in the setting of sepsis, and that lack of effective vascular autoregulation and increased microvascular permeability could result in neural edema and capillary occlusion, thereby damaging peripheral nerves.72,77 As with other types of organ dysfunction in SIRS, neuronal injury may be cytokine-mediated.77
Intensive Care Unit–Acquired Myopathy (Critical Illness Myopathy)
Severe ICU-acquired myopathy was first reported 25 years ago,78 with numerous individual case reports and case series having been published subsequently.79–81 Myopathy most often occurred in the setting of mechanical ventilation for severe airflow obstruction, and affected individuals had typically been treated with concomitant corticosteroids and NMBAs. However, myopathy has also been documented in patients treated with corticosteroids without paralysis, and in critically ill septic patients who received neither corticosteroids nor NMBAs.73,82–85 Some authors have attempted to subdivide the acute myopathy that develops in the ICU into a “critical illness myopathy” associated with hypercatabolic (generally septic) states, a “thick filament myopathy” related to corticosteroids and (usually) NMBAs, and a “necrotizing myopathy” characterized by prominent myonecrosis.73 However, it is not clear that these are really distinct entities, and a recent editorial suggested the term critical illness myopathy be used as the sole descriptor for all instances of ICU-acquired myopathy.86
Like critical illness polyneuropathy, acute myopathy is usually manifested by generalized muscle weakness that becomes apparent once withdrawal of sedation allows an assessment of neuromuscular function. Although reflexes may be depressed or absent if myopathy is severe, sensation is intact. In severe cases, patients are unable to move their limbs, resulting in the commonly used descriptor “acute quadriplegic myopathy.” On occasion, patients may not open their eyes in response to verbal commands, giving the impression of coma. When less severe, patients may have predominantly distal weakness, and both foot drop and wrist drop are common. In contrast to the frequent diaphragmatic involvement in critical illness polyneuropathy, asthmatic patients who develop myopathy often have relatively well-preserved inspiratory muscle function despite near-quadriplegic limb weakness.80
Serum creatine kinase (CK) levels may be elevated or normal, and can neither establish nor exclude an underlying myopathy as the cause of muscle weakness. The EMG reveals either normal or decreased CMAP amplitudes, depending on the degree of muscle involvement. Like critical illness polyneuropathy, needle EMG often reveals spontaneous fibrillations and positive sharp waves in resting muscle. However, with voluntary contractions there is a characteristic myopathic pattern of abundant low-amplitude, short-duration polyphasic units with early recruitment.87 Pathologic findings in acute myopathy have been variable, with fiber atrophy and vacuolization in the absence of inflammation being the most common finding.73 Although necrosis may be prominent, more often it is minimal or absent.73 A key finding on electron microscopy is a selective loss of myosin, with relative sparing of actin and Z bands.73 This pattern is very characteristic of ICU-acquired myopathy and is rarely seen in other disorders of muscle. A reduction in the myosin:actin ratio has also been documented by electrophoresis of needle aspirates of muscle, potentially offering a minimally invasive technique of diagnosing acute ICU-acquired thick filament myopathy.88
The underlying reason for the often near-quadriplegic muscle weakness in ICU-acquired myopathy is uncertain. Although myosin loss may interfere with normal contractility, another potential mechanism of weakness may be temporary muscle inexcitability. In a study of patients with severe ICU-acquired myopathy related to corticosteroids and NMBAs, Rich and colleagues found that they were unable to induce an action potential even when the muscle was stimulated directly.89 Muscles later regained their ability to generate an action potential during clinical recovery.89 In a subsequent study, inability to induce an action potential by direct muscle stimulation was also documented in a more heterogenous group of patients with ICU weakness, including a number of patients who were septic and had not received corticosteroids or NMBAs.90 Studies in experimental animals suggest that muscle inexcitability may be due to impairment of sodium fast channels.91
Polyneuropathy, Myopathy, or “Polyneuromyopathy”
Differentiation of an axonal motor polyneuropathy and severe myopathy by EMG often rests upon the pattern of recruitment seen with voluntary muscle contraction, and inability to cooperate with the voluntary portion of the needle EMG may preclude reliable diagnosis. Marked reduction in sensory nerve amplitudes would indicate neuropathy, but tissue edema may limit the accuracy of sensory ENG, and it has been claimed that 40% to 60% of patients with critical illness polyneuropathy have normal sensory exams. Furthermore, the routine histopathologic findings in muscle biopsy specimens may be similar in myopathy and critical illness polyneuropathy. Therefore, a diagnostic dilemma is presented by patients who have a normal sensory ENG (or in whom sensory evaluation cannot be performed) and are unable to perform voluntary muscle contractions.
A recent study evaluated eight critically ill patients in whom a standard ENG/EMG and muscle histopathology could not differentiate a pure motor axonal neuropathy and a primary myopathy.92 In all cases electron microscopy revealed a profound selective loss of myosin and normal sural nerve histology, indicating the presence of a primary myopathy rather than motor neuropathy as the cause of weakness.92 A similar conclusion was reached when direct muscle stimulation was used to differentiate motor neuropathy and myopathy.93 As noted earlier, muscle inexcitability has been documented in acute myopathy, but a motor neuropathy should not preclude generation of a normal CMAP in response to direct muscle stimulation. In a study of 22 consecutive patients in whom standard ENG/EMG was nondiagnostic, all but 1 patient had evidence of myopathy by direct muscle stimulation and quantitative EMG.93 Another group has noted that with use of direct muscle stimulation they have never been able to document a single case of pure axonal motor neuropathy, and suggest that previously reported patients given the latter diagnosis almost certainly had a myopathy.94 Direct muscle stimulation could eventually prove to be useful in the investigation of severe weakness in the ICU, because coma or deep sedation does not preclude its use.84 It should be appreciated, however, that this technique is not as well-standardized as the routine EMG, and individual laboratories may need to establish their own normative data.
Recently, it has been suggested that some patients with ICU-acquired weakness have both an axonal polyneuropathy and myopathy. In one study muscle and nerve biopsies were performed in 24 patients with ICU-acquired weakness, all of whom had ENG/EMG evidence suggestive of polyneuropathy.95 All but one had pathologic evidence of a myopathy and one third also had an axonal lesion. The authors concluded that polyneuropathy and myopathy often coexist. A recent prospective study of ICU-acquired paresis found that 22 patients in whom ENG/EMG could be performed uniformly had evidence of a sensorimotor axonal neuropathy, but each of the 10 patients who also underwent muscle biopsy had evidence of myopathy (type II fiber atrophy with myosinolysis).96 This again suggested that both polyneuropathy and myopathy may coexist and together contribute to ICU-acquired weakness.
Although clarification of the underlying mechanisms of ICU-acquired weakness may be important from a research perspective, it is uncertain how it will affect individual patients in the ICU since there is no specific therapy. Indeed, it has been suggested that electrophysiologic and pathologic studies are of little practical value in the evaluation of weakness that is clearly acquired after the onset of critical illness, and that a purely clinical approach is adequate to define the presence of a critical illness weakness syndrome.97 From a practical standpoint, the primary concern with omission of electrophysiologic studies would be missing a rare case of the Guillain-Barré syndrome, a disorder with specific therapy that may be triggered by acute nonneurologic critical illness (see above). In addition, the prognosis for recovery may be different with a pure myopathy and a severe axonal polyneuropathy (see below).
Intensive Care Unit–Acquired Weakness: Incidence and Risk Factors
A number of studies have attempted to define the incidence of ICU-acquired weakness and the risk factors for its development. In general, two types of patient cohorts have been examined: those with status asthmaticus, and a more general ICU population in which sepsis, SIRS, and multiorgan failure were dominant.
Asthmatic patients who developed acute myopathy have uniformly received systemic corticosteroids while undergoing mechanical ventilation, and in the great majority of cases neuromuscular paralysis had been used to facilitate mechanical ventilation.79–81 Two large retrospective studies found an identical 30% incidence of myopathy among mechanically ventilated asthmatic patients who had received an NMBA.80,98 The risk of myopathy increases with the duration of paralysis. One prospective study found a striking relationship between the duration of paralysis and incidence of myopathy.79 In a second retrospective study the mean duration of paralysis for patients who developed myopathy was 3.2 days, and 90% of patients had been paralyzed for more than 24 hours.80 A third study also reported a mean duration of paralysis of 3 days among asthmatic patients who developed myopathy.98 Although it was initially believed that myopathy may be related to use of steroidal NMBAs (vecuronium and pancuronium), the incidence of myopathy is similar with atracurium (a nonsteroidal NMBA) and vecuronium.80
These reports, as well as numerous additional publications, have suggested that mechanically ventilated patients with status asthmaticus who are treated with corticosteroids should be at minimal risk of developing acute myopathy as long as they have minimal or no exposure to NMBAs. Unfortunately, the data linking use of NMBAs with acute myopathy has been based almost entirely on retrospective analyses. In the single prospective study of myopathy in status asthmaticus, patients were uniformly paralyzed.79 More recently, myopathy has been reported in mechanically ventilated patients with airflow obstruction who received corticosteroids and deep sedation, without paralysis.83 Furthermore, our recent unpublished clinical experience has included a number of patients who developed severe myopathy despite minimal or no exposure to NMBAs. Therefore, while prolonged paralysis appears to increase the risk of myopathy in status asthmaticus, avoidance of NMBAs does not eliminate the risk of neuromuscular weakness.
In experimental animals a thick filament myopathy analogous to that seen in humans with status asthmaticus can be produced by combined denervation and corticosteroids, but not by either intervention alone.99 Denervation results in a marked increase in the number of cytosolic corticosteroid receptors, but a similar effect is seen after limb immobilization by pinning and casting.100,101 It is possible that the apparent enhancement of corticosteroid myotoxicity by prolonged paralysis in individuals with status asthmaticus is related to the induction of total muscle inactivity rather than denervation per se. This could explain the occurrence of severe myopathy in corticosteroid-treated patients who underwent deep sedation for many days.83