Intensive care unit—acquired weakness (ICUAW) and muscle wasting in critically ill patients with severe sepsis and septic shock

Sepsis is a leading cause of death in intensive care units (ICU) [1‐3]. In order to establish consensus definitions, the terms systemic inflammatory response syndrome (SIRS), sepsis, severe sepsis and septic shock have been proposed [4‐7]. These terms define gradual stages of disease severity that correlate with mortality. Whereas SIRS defines a rather unspecific inflammatory host response due to both infectious and non-infectious origin, severe sepsis refers to a proven systemic infection associated with acute organ dysfunction. Criteria of septic shock are met once additional volume-refractory hemodynamic failure occurs [4, 5]. Recently, there has been great effort in studying the epidemiology of sepsis [2]. Data from the United States report 750,000 cases/year of severe sepsis [8] and an overall mortality of 29% for 2001. A 9% annual increase of severe sepsis incidence between 1977 and 2000 was reported [9]. Epidemiological trials from Germany indicate that 79,000 inhabitants (116/100,000) will suffer from sepsis each year and an additional 75,000 inhabitants (110/100,000) will be diagnosed with severe sepsis. Ninety-day mortality of severe sepsis amounts to 54%. With 60,000 sepsis-related deaths per year in Germany, sepsis remains a leading cause of death and a considerable burden for health care systems [10‐12].

Treatment of patients with sepsis should be implemented according to international guidelines and recommendations, e.g., as proposed by the Surviving Sepsis Campaign [4]. Early identification and therapy initiation seem of pivotal importance and were shown to significantly improve patients’ prognoses. In general, guidelines for the treatment of severe sepsis or septic shock aim to initiate early symptomatic organ support therapies; e.g. early optimization of cardiac preload, afterload and contractility in order to balance oxygen delivery with oxygen demand influences survival rates in sepsis [13]. Over the past years, several treatment approaches have been associated with beneficial outcomes in large-scale, mostly single-centre trials, including low-dose hydrocortisone [14], glycemic control [15, 16] and recombinant activated protein C [17]. Yet, these benefits could not be replicated by confirmative studies [18‐20].

Pathophysiologically, the first response to a severe infection consists in the activation of antigen-presenting immune cells. This occurs via pattern recognition receptors [1, 21‐23] and is accompanied by other immune mechanisms such as cytokine liberation, endothelium and complement activation, and release of oxygen radicals. Cytokines (tumour-necrosis factor alpha (TNF-α), interleukin (IL)-6 and IL-1) and complement factors (C3a and C5a) may then act as key mediators during this stage of inflammation [21]. Clinical signs of infection such as leukocytosis, increased respiratory rate or acute organ failure may then develop [1] and, as a protective measure, the immune system coordinates an immunological and endocrine counter-regulation which is mediated by up-regulation of key anti-inflammatory cytokines (e.g. IL-10, tumour-growth factor-ß) [7, 24‐27]. This may lead to prolonged hypo-inflammatory states (“immunoparalysis”) [27], which have been linked to secondary/nosocomial infection, prolonged ICU stay, chronic multiple organ failure and an increased mortality [21‐31]. Ongoing clinical trials investigate measures of immunostimulation [28‐30] or removal of inhibitory factors [31] to reverse this immunological condition.

Inflammation-mediated organ dysfunction represents a detrimental comorbidity of sepsis. As sepsis is frequently accompanied by short- and long-term affection of neuromuscular function, it has been postulated that the “motor unit” presents another system affected by inflammation-mediated multiple organ dysfunction: Sepsis is not only associated with involuntary loss of muscle mass, which is frequently referred to as muscle wasting, but also severe neuromuscular dysfunction resulting in “ICU-acquired weakness” (ICUAW). Patients with acute respiratory distress syndrome show a particularly high incidence for ICU-acquired weakness [32].

As mentioned above, muscle force capacity may remain stable in muscle wasting syndromes [36]. Muscle wasting can be triggered by other conditions than sepsis, including disuse, denervation, fasting, cancer, cardiac failure and renal dysfunction. As these conditions frequently coincide with sepsis, systematic research of sepsis-specific muscle wasting and ICUAW may be challenging in humans. This seems especially the case as sepsis is often accompanied by prolonged bedrest/immobilisation, application of sedatives, acute or chronic organ dysfunction, malignancy as an underlying disease, medication using glucocorticoids and others.

An imbalance between muscle protein synthesis and muscle protein degradation causing net loss of muscle mass is considered to present the main mechanism of muscle atrophy in muscle wasting and may result from decreased protein synthesis and/or increased protein degradation [36] (Fig. 1).

Even though studies on sepsis-induced muscle protein degradation received more attention, animal models of sepsis clearly indicate that sepsis also decreases protein synthesis in skeletal muscles [37] and preferentially inhibits myofibrillar and sarcoplasmatic protein synthesis within fast twitch muscles [38].

One central mechanism is decreased protein translation by lower amounts of active translation initiation factor complex eIF4E/eIF4G, possibly resulting from decreased phosphorylation of eIF4G and mammalian target of rapamycin (mTOR) [37]. Interestingly, mTOR activation presents an important downstream target of anabolic insulin/insulin like growth factor 1 (IGF-1) signalling, which may be impaired in sepsis since decreased insulin sensitivity presents a frequently observed complication of sepsis [39‐44]. Further indirect evidence for contribution of impaired insulin signalling to decreased protein synthesis comes from the observation that the proinflammatory cytokines IL-6 and TNF-α have been linked to both insulin resistance [45, 46] and muscle atrophy [47‐50] and that local IGF-1 application prevents sepsis-induced muscle atrophy [37], possibly by inhibition of sepsis-induced increases of muscle atrogin-1 and the proinflammatory cytokine IL-6 [51]. IL-1 may lead to decreased protein synthesis as well, as an animal model of abdominal sepsis showed cytokine-dependent decreases of phosphorylated eIF4G that was primarily attributed to this cytokine [52]. There is evidence that systemic cytokine response in sepsis results in local amplification of proinflammatory cytokines within muscle [36], possibly aggravating decreased protein synthesis in skeletal muscle during sepsis.

Nutritional aspects may contribute to decreased muscle protein synthesis as well. Administration of the essential branched chain amino acid leucine has been shown to increase protein synthesis in rat skeletal muscles during ageing, exercise or food-deprivation [53‐56]. However, under certain conditions including sepsis, the muscle may be resistant to leucine-stimulated protein synthesis [37]. Although the exact mechanisms of sepsis-induced leucine resistance remain to be elucidated, this phenomenon was accompanied by an 80% reduction in the amount of active eIF4E/eIF4G complex (besides other factors) and could be completely reversed by the combination of TNF binding protein and glucocorticoid receptor antagonist RU486 [57]. Exploring the underlying mechanism of sepsis-induced resistance to leucine-stimulated protein synthesis may contribute to a better understanding of the mechanisms involved in sepsis-mediated reductions of active translation initiation factor complexes, particularly the eIF4E/eIF4G complex.

Data from studies in cultured cells, animals and humans indicate an increase of sepsis-associated muscle protein degradation by several mechanisms, including the ubiquitin proteasome system (UPS) [58‐61] and lysosomal systems [50, 62‐64]. Calcium-dependent non-lysosomal calpains and pro-apoptotic pathways (caspases) have also been associated with sepsis-induced muscle atrophy [66‐68].It has been postulated that caspases and calpains are responsible for the cleavage of myofibrillar proteins preceding their proteasomal degradation [65‐67] and that cleavage is necessary as the proteasome cannot degrade intact myofibrillar proteins. Protein degradation by proteasomal and/or lysosomal systems may not be sepsis-specific as various other conditions associated with muscle wasting share similar or identical biochemical and transcriptional pathways.

During muscle wasting, defective insulin signalling may not only be involved in decreased muscle protein synthesis, but also in increased muscle proteolysis, since mTOR inhibits autophagy-induced lysosomal proteolysis [68]. mTOR activation is regulated by phosphorylated Akt, another key downstream effector of insulin signalling. Phosphorylated Akt leads to phosphorylation of FOXO3 transcription factors, which results in FOXO3 translocation from nucleus to cytoplasm and a thereby decreased transcription of atrophy gene atrogin-1 [69].

E3 ubiqitin ligases, such as atrogin-1 and muscle ring finger protein 1 (MuRF-1), represent substrate specific enzymes involved in the UPS that prevent unselective degradation by the proteasome and have been shown to be markedly induced during experimental atrophic conditions [70, 71]. Whereas myosin heavy chain and other myofibrillar proteins including myosin binding proteins were identified as MuRF-1 substrates [72, 73], the only atrogin-1 substrates identified so far include MyoD and translation initiation factor eIF3-f, which are both known to regulate protein synthesis [74, 75]. It is therefore possible, that atrogin-1 leads to muscle atrophy by selective breakdown of key regulators of protein synthesis [76]. We emphasise once more, that sepsis-mediated muscle wasting and ICU-acquired weakness should not be used synonymously, as ICUAW weakness represents its own entity, characterised by additional factors.

ICUAW presents a severe and frequent complication of critical illness, confronting intensivists around the globe with various difficulties. It is believed that ICUAW can affect more than half of all ICU patients [77]. This substantial neuromuscular complication of critical illness is associated with increased rates of morbidity and mortality, substantially affecting both short- and long-term clinical outcomes in septic patients [78‐85]. By causing prolongation of ICU stay and rehabilitation, ICUAW leads to an increased risk of secondary complications and a higher demand on already limited resources of health care systems. ICUAW is further associated with decreases in health-related quality of life (HRQOL) which may be impaired for years after ICU discharge [85, 86]. Modern critical care must therefore no longer solely focus on survival, but particularly consider HRQOL after ICU discharge.

ICUAW refers to the bedside diagnosis of pronounced weakness in ICU patients without plausible aetiology other than critical illness [35]. Whereas the patients’ history usually reveal exposure to ICUAW-associated risk factors (see below), physical examination of awake and cooperative patients with typical ICUAW shows symmetric weakness and decreased tone. This primarily affects the lower limbs but may extend to tetraplegia in more severe cases, explaining its prior terminology of acute quadriplegic myopathy. Muscles innervated by cranial nerves are normally—yet not necessarily—spared, while respiratory muscle function is frequently abnormal. Deep tendon reflexes may be normal, decreased or absent [35]. Whenever suspecting ICUAW, it is fundamental to rule out prolonged neuromuscular blockade (involvement of cranial nerve-innervated muscles), pre-existing neuromuscular dysfunction and other conditions as alternative causes. Currently, ICUAW is frequently evidenced by difficulties in weaning from mechanical ventilation or unexpected problems with mobilisation. However, as weaning from respiratory support may not be initiated for several days or even weeks, it seems important to consider possible ICUAW before development of its full extent in order to prevent or at least attenuate underlying pathology by minimising risk factor exposure [34].

After all, earlier detection of ICUAW is possible and can be obtained by assessment of voluntary maximum strength in ICU patients, either by hand dynamometry or according to the Medical Research Council (MRC)-score [35]. The MRC-score grades manually tested strength from 0 (no movement observed) to 5 (muscle contracts normally against full resistance) in three functional muscle groups of each extremity, with mean MRC-scores of <4 (antigravity strength) indicating ICUAW. It has shown good interobserver reliability in patients with Guillan–Barré syndrome [87]. Values of <11-kg force for men and <7-kg force for women at dominant-hand dynamometry have also been described to identify ICU-acquired weakness in previously healthy individuals [88]. Nevertheless—and in spite of daily wake up calls, both approaches share one major limitation—the requirement for patient cooperation and consciousness, which may be inadequate due to delirium resulting from sedation and/or septic encephalopathy. Electrophysiologic testing offers an additional approach to estimate neuromuscular dysfunction in unconscious patients incapable of voluntary contraction. Moreover, electrophysiologic testing after direct muscle stimulation can be conducted in unconscious/sedated patients and has been shown to predict ICUAW with high sensitivity and specificity in mechanically ventilated, sedated patients and differentiates between primary nerve or muscle dysfunction [89]. However, as it requires some expertise and certain devices to perform these measurements, use of this technique is currently restricted to larger facilities or experts.

While the exact molecular mechanisms contributing to ICUAW remain to be elucidated, five central risk factors of ICUAW have been repeatedly reported [34]. As there are currently no specific therapies, minimising exposure to these risk factors is crucial in order to prevent this devastating neuromuscular complication. Possibly the most important risk factor complex comprises conditions leading to multiple organ failure, particularly severe sepsis and septic shock. Some authors actually consider ICUAW an additional organ failure following severe sepsis and septic shock. The other four risk factors of ICUAW involve muscle inactivity, disturbances of glucose metabolism resulting in hyperglycaemia, administration of corticosteroids, and use of neuromuscular blocking agents (Fig. 1).

Depending on electrophysiologic or histological documentation of neuropathy and/or myopathy, ICUAW can be further subclassified as critical illness myopathy (CIM), critical illness polyneuropathy (CIP), or critical illness myoneuropathy (CIMN), which is more frequent than previously thought and applies to ICUAW patients with coinciding neuropathy and myopathy [35]. Myopathy is likely to present the predominant feature of ICUAW and has been shown to precede neuropathy in patients with CIMN [89]. Yet, as both histology and electrophysiology are currently not obtained during clinical routine, subclassification of ICUAW may seem of less importance to clinicians dealing with this complication. We just recently demonstrated that myopathy occurs earlier and more frequent, whereas additional polyneuropathy develops later and less frequent. By showing that additional polyneuropathy was associated with longer ICU length of stay, we suggest that differential diagnosis of ICUAW becomes more important to clinicians [105].

Three characteristics are commonly described in patients with CIM or CIMN [106‐108]:

Unspecific but predominant type II (fast twitch) muscle fibre atrophy has been repeatedly described within muscle tissue from CIM/CIMN biopsies [35]. Selective but patchy loss of myosin filaments can be visualised by electron microscopy and is considered a hallmark of CIM, explaining its additional terminology of thick filament myopathy [35]. In contrast to these morphologic alterations, non-excitable muscle membrane is detected by electrophysiologic testing [43, 108, 109] and has been shown to predict ICUAW in mechanically ventilated, sedated patients with high sensitivity and specificity [89]. All three characteristics may result in different mechanisms leading to ICUAW, as loss of muscle mass (atrophy) correlates with decreased maximum force, thick filament loss represents an additional reduction of force generating capacity by additional myofilament dysfunction, and non-excitable muscle membrane may be considered as an incapability of the muscle to generate contraction-preceding action potentials. Less frequently described histological signs of CIM/CIMN include acute necrosis [110, 111], regeneration [112] as well as loss of myofibrillar ATPase staining [113], the latter affecting both type I and II muscle fibres.

A number of subcellular sites involved in excitation contraction coupling may be affected in sepsis-induced myopathy [36]. These include the sarcolemma, the sarcoplasmatic reticulum, the contractile apparatus and the mitochondria. As described above, one of the key features of CIM/CIMN is that skeletal muscle becomes electrically inexcitable, which has led to the concept that CIM/CIMN could represent an acquired channelopathy involving dysregulation of sodium channels located at the sarcolemma. Experimental studies describe sarcolemmal injuries in diaphragm and limb muscle that were in part associated with excessive nitric oxide generation [114‐116]. Altered calcium homeostasis has been observed in a number of studies on skeletal muscle during sepsis with calcium level increases in some subcellular compartments and decreases in others [36]. Contractile protein dysfunction resulting in reduced force pCa relationship has been reported in diaphragm and limb muscle in a number of sepsis models [117‐120]. It is likely that free radical generation is involved in this mechanism, as either a superoxide scavenger or NO synthesis inhibitor significantly attenuated reduced endotoxin-induced force pCa relationships. TNF-α has been linked to decreases in titanic force generation due to changes in myofilament function, as well [121, 122]. A considerable amount of data indicates that sepsis also causes profound alterations in respiratory muscle mitochondrial function [36]. Limb muscle biopsies from septic patients show decreased mitochondrial content and lower concentrations of energy rich phosphates [123].