Hostname: page-component-cd9895bd7-hc48f Total loading time: 0 Render date: 2024-12-22T03:38:31.043Z Has data issue: false hasContentIssue false

Propofol as a Risk Factor for ICU-Acquired Weakness in Septic Patients with Acute Respiratory Failure

Published online by Cambridge University Press:  16 January 2017

Peter A. Abdelmalik*
Affiliation:
Department of Anesthesia and Critical Care Medicine, Neurocritical Care Division, The Johns Hopkins University School of Medicine, Baltimore, MD
Goran Rakocevic
Affiliation:
Department of Neurology, Thomas Jefferson University Hospitals, Philadelphia, PA.
*
Correspondence to: Peter Abdelmalik, Department of Anesthesia and Critical Care Medicine, Neurocritical Care Division, The Johns Hopkins University School of Medicine, 600 N Wolfe St., Phipps 455, Baltimore, MD, 21231. Email: pabdelm1@jhmi.edu
Rights & Permissions [Opens in a new window]

Abstract

Background: Critical illness polyneuropathy (CIN) and critical illness myopathy (CIM), together “ICU-Acquired weakness (ICUAW),” occur frequently in septic patients. One of the proposed mechanisms for ICUAW includes prolonged inactivation of sodium channels. Propofol, used commonly in patients with acute respiratory failure (ARF), primarily acts via enhancement of GABAergic transmission but may also increase sodium channel inactivation, suggesting a potential interaction. Methods: Electronic medical records and EMG reports of patients with ICUAW and a diagnosis of either sepsis, septicaemia, severe sepsis, or septic shock, concurrent with a diagnosis of acute respiratory failure (ARF), were retrospectively analyzed in a single center university hospital. Results: 74 cases were identified (50.0% men, age 58±14 years), and compared to age- and sex-matched controls. Of these, 51 (69%) had CIN, 19 (26%) had CIM, and 4 (5%) had both. Propofol exposure was significantly higher in patients with ICUAW compared to controls (63.5% vs. 33.8%, p<0.001). The odds ratio of developing ICUAW with propofol exposure was 3.4 (95% CI:1.7-6.7, p<0.001). Patients with ICUAW had significantly more days in hospital (59±44 vs. 30±23) and ICU (38±26 vs. 17±13), days dependent on mechanical ventilation (27±21 vs. 13±16), and rates of tracheostomy (79.7% vs. 36.5%) and gastrostomy (75.7% vs. 25.7%) (all p<0.001). They also received a significantly higher number of distinct intravenous antibiotics, cumulative days of antibiotic therapy, and exposure to vasopressors and paralytics. Conclusions: Propofol exposure may increase the risk of ICUAW in septic patients. An interaction through sodium channel inactivation is hypothesized.

Résumé

Facteur de risque du propofol dans le cas de patients septiques souffrant d’insuffisance respiratoire aiguë et de troubles de la faiblesse musculaire acquis aux soins intensifs. Contexte: La polyneuropathie et la myopathie sont deux troubles qui peuvent apparaître à la suite d’une maladie grave. Dans les deux cas, ils se manifestent par une faiblesse musculaire acquise aux soins intensifs ; ils surviennent aussi fréquemment chez des patients atteints de sepsie. Une des solutions apportées à ce syndrome clinique inclut notamment l’inactivation des canaux sodiques. Couramment utilisé pour soulager des patients souffrant d’insuffisance respiratoire aiguë, le propofol agit principalement en augmentant la transmission GABAergique mais pourrait aussi augmenter l’inactivation des canaux sodiques, ce qui suggère une potentielle interaction. Méthodes: Les dossiers médicaux électroniques et les rapports d’ÉMG de patients atteints de faiblesse musculaire acquise aux soins intensifs, chez qui on avait diagnostiqué une sepsie, une septicémie, une grave sepsie ou un choc septique de même qu’une insuffisance respiratoire aiguë, ont été analysés rétrospectivement au sein d’un centre hospitalier universitaire. Résultats: Au total, 74 cas ont été repérés (50,0 % d’hommes ; âge 58±14 ans) et comparés à des témoins du même sexe et du même âge. Sur ces 74 cas, 51 (69%) étaient atteints de polyneuropathie ; 19 (26%), de myopathie ; et 4 (5%), de ces deux troubles. En comparaison avec les témoins, l’exposition au propofol s’est révélée sensiblement plus élevée chez les patients atteints de faiblesse musculaire acquise aux soins intensifs (63,5% contre 33,8% ; p<0,001). Le risque relatif approché d’être aux prises avec un tel syndrome clinique après avoir été soulagé par du propofol était de 3,4 (95% IC : 1,7 – 6,7 ; p<0,001). Les patients aux prises avec ce syndrome ont passé beaucoup plus de jours à l’hôpital (59±44 contre 30±23) et aux soins intensifs (38±26 contre 17±13). Ils ont été aussi dépendants d’un système de ventilation mécanique pendant plus de jours (27±21 contre 13±16). Leurs taux de trachéostomie (79,7% contre 36,5%) et de gastrostomie (75,7% contre 25,7%), tous les deux p<0,001, se sont en outre avérés plus élevés. Finalement, on leur a administré par voie intraveineuse un nombre sensiblement plus élevé d’antibiotiques de divers types. Le cumul des jours comportant un traitement antibiotique et l’exposition à des agents vasopresseurs et paralytiques a également été plus élevé. Conclusions: Il se pourrait que l’exposition au propofol augmente le risque chez des patients atteints de sepsie et souffrant de faiblesse musculaire acquise aux soins intensifs. Selon nous, le tout pourrait s’expliquer par une interaction découlant de l’inactivation des canaux sodiques.

Type
Original Articles
Copyright
Copyright © The Canadian Journal of Neurological Sciences Inc. 2017 

Introduction

Critical illness polyneuropathy (CIN), and critical illness myopathy (CIM), are syndromes of generalized flaccid weakness with diminished reflexes that often develop simultaneously in critically ill patients, for which there are no other obvious neurological causes.Reference Fan, Cheek, Chlan, Gosselink, Hart and Herridge 1 Together known as ICU acquired weakness (ICUAW),Reference Schweickert and Hall 2 the diagnosis is often postulated when a patient with quadripareisis, or frank quadriplegia, fails to be weaned from mechanical ventilation. Many factors have been implicated in the pathogenesis of ICUAW. A meta-analysis identified the following as possible inciting and contributing factors: hyperglycemia; wide fluctuations of blood glucose in the context of sepsis and multi-organ failure; sepsis; systemic inflammatory response syndrome (SIRS); compromised renal excretion as well as renal replacement therapy; and catecholamine administration.Reference Stevens, Dowdy, Michaels, Mendez-Tellez, Pronovost and Needham 3 Others, such as patients’ age and gender, severity of illness, and use of certain medications (glucocorticoids, neuromuscular blockers, aminoglycosides, or midazolam) have been largely excluded as precipitating factors.Reference Stevens, Dowdy, Michaels, Mendez-Tellez, Pronovost and Needham 3

Several hypotheses exist regarding the pathophysiology of ICUAW, including sodium channel dysfunction.Reference Schweickert and Hall 2 , Reference Kress and Hall 4 - Reference Latronico and Bolton 8 Sepsis itself has also been implicated in causing sodium channel dysfunction, in both nerve Reference Novak, Nardelli, Cope, Filatov, Glass and Khan 9 and muscle,Reference Haeseler, Foadi, Wiegand, Ahrens, Krampfl and Dengler 10 by causing a depolarization of resting membrane potential and decreasing the population of available sodium channels in experimental models.

Propofol (2,6-diisopropylphenol) is a common sedative used in critically ill patients. First described for its use as a sedative in the late 1980s,Reference Grounds, Lalor, Lumley, Royston and Morgan 11 , Reference Newman, McDonald, Wallace and Ledingham 12 propofol has become a mainstay sedation drug in the ICU because of its rapid onset (1-2 minutes) and short duration of action (2-8 minutes).Reference Hughes, McGrane and Pandharipande 13 Propofol acts primarily as a direct gamma-aminobutyric acid (GABA) agonist at a non-benzodiazepine site. But it also inhibits the N-methyl-D-aspartate (NMDA) receptors and modulates calcium influx through slow calcium-ion channels.Reference Kotani, Shimazawa, Yoshimura, Iwama and Hara 14 Propofol has also been characterized as acting on sodium channels, both in the central and peripheral nervous systems. It has been shown to inhibit persistent sodium channels in in vitro brain preparations,Reference Martella, De Persis, Bonsi, Natoli, Cuomo and Bernardi 15 and block human skeletal muscle sodium channels in a voltage dependent manner.Reference Haeseler, Störmer, Bufler, Dengler, Hecker and Piepenbrock 16

Sodium channels are ubiquitous in excitable tissue and are crucial for axon salutatory conduction and muscle cell depolarization. Given the current hypothesis that sepsis causes sodium channel dysfunction and that propofol acts at sodium channels, we further hypothesized that propofol use would be an independent risk factor for the development of ICUAW. To date there have been a no electromyographical/nerve conduction studies (EMG/NCS) linking the use of propofol with ICUAW. However, diaphragmatic force was measured using bilateral anterior magnetic phrenic nerve stimulation (BAMPS) in a group of ten ICU patients, nine of whom were septic, and demonstrated an inverse relationship with diaphragmatic force and cumulative propofol dose,Reference Hermans, De Jonghe, Bruyninckx and Van Den Berghe 17 suggesting a potential interaction.

Methods

Subjects

The database inquires of patients’ health information were approved by the Thomas Jefferson University Hospital (TJUH) Institutional Review Board. For this type of study formal consent was not required. Patients with CIN, CIM, or both, were identified retrospectively in two ways. First, a keyword search of the MS Word documents (Microsoft, Redmond, WA) generated by two EMG machines at the Department of Neurology (Nicolet, Natus Neurology, Middleton, WI) using the keywords ‘intensive,’ ‘critical,’ ‘neurocritical,’ ‘ICU,’ and ‘intubated’ to identify all EMG studies performed on patients admitted to the medical (including cardiac and bone marrow), neurological, and surgical intensive care units. Patients with documented acute respiratory failure and either sepsis, septicemia, severe sepsis, or septic shock were included for further analysis. Only those patients with the finding of axonal sensorimotor neuropathy or/and myopathy in an ICU setting confirmed by electrophysiological studies were included.

Additionally, the inpatient Electronic Medical Records (EMR) system at TJUH was queried for all patients with a diagnosis of critical illness polyneuropathy (ICD 9: 357.82) or critical illness myopathy (ICD 9: 359.81), who also had concurrent diagnoses of acute respiratory failure (as a medical indication for the use of propofol, ICD 9: 518.81) and one of the following: sepsis (ICD 9: 995.91), septicemia (ICD 9: 038), severe sepsis (ICD 9: 995.92), or septic shock (ICD 9: 785.52).

Once a study cohort was identified, age- and sex-matched controls were identified by preforming another TJUH EMR database query screening all patients with a diagnosis of ARF and either sepsis, severe sepsis, or septic shock.

Exclusion criteria were hospital stay less than 48 hours; outside hospital transfer with outside hospital stay greater than seven days; outside hospital transfer with an existing diagnosis of acute respiratory failure and intubation prior to arrival; a history of new weakness preceding inpatient admission; or a documented history of myasthenia gravis, acute inflammatory demyelinating polyradiculoneuropathy (Guillan-Barre Syndrome, AIDP/GBS), chronic inflammatory demyelinating polyradiculoneuropathy (CIDP), polymyositis, amyotrophic lateral sclerosis (ALS) or other motor neuron disease. Patients with existing tracheostomy or gastrostomy were also excluded. Methods were consistent with guidelines from the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) for case-control studies.

Patient Data

Patients’ health information review included demographics; hospital and ICU length of stay; duration of mechanical ventilation; rates of tracheostomy and gastrostomy placement; concurrent medical diagnoses (using ICD 9 codes); pertinent medical histories; and the proportion requiring vasopressors, intravenous antibiotics, intravenous steroids, paralytics, titratable propofol or midazolam; and disposition at discharge. Bolus doses of sedation for bedside procedures were not included or reviewed.

The cumulative exposure to each aforementioned medications was tallied as days of exposure. Finally, pertinent laboratory values were reviewed, including blood glucose on admission; hemoglobin A1C, average serum glucose, protein, albumin, and pre-albumin noted on admission and repeated values; and initial and peak values of creatinine, creatine kinase, and lactate.

Statistics

Data were analyzed by SPSS version 21 (International Business Machines Corporation, Armonk, NY). Significance was noted at a p-value of<0.05.

Nominal data were analyzed by Chi square, or Fisher test where cases tallied less than five. Interval data were subjected to a test of normality using the Shapiro-Wilk test. Age, cumulative aminoglycoside exposure, and average serum glucose were analyzed via independent T test, and one-way ANOVA. All other values were analyzed using either the Mann Whitney U or the Kruskal Wallis test.

Bivariate logistic regression was also performed to identify independent risk factors for the diagnosis of ICUAW using a backward conditional approach, starting with known or hypothesized risk factors for ICUAW, including exposure to intravenous antibiotics, vasopressors, paralytics, propofol, midazolam, glucocorticoids; diagnoses of acute hepatitis; septic shock; and protein-calorie malnutrition. These variables were either previously hypothesized as potential inciting factors contributing to the occurence of ICUAW, or, the particular variable had a p value < 0.2 when compared between the ICUAW cohort and controls.

Results

In total, 3716 EMG reports were screened (237 Machine A [Thomas Jefferson Hospital, Gibbon Building]+472 Machine B [Jefferson Hospital for Neuroscience]+3007 [Jefferson Hospital for Neuroscience archived hard drive]). Of those, 37 cases were identified, one of which was performed in the outpatient setting. Of the 36 inpatient studies, 11 had a confirmed or suspected diagnosis of AIDP/GBS, three had a diagnosis of myasthenia gravis, one had CIDP, and one had a normal EMG. Of the 20 remaining cases, eight did not have documented sepsis or acute respiratory failure, leaving 12 cases of either CIN or CIM with documented sepsis and acute respiratory failure that underwent inpatient EMG testing.

Query of the TJUH EMR for CIN or CIM yielded 191 potential cases from the years 2000-2014. Of those, 57 did not have a diagnosis of sepsis; nine did not have a diagnosis of acute respiratory failure; 20 cases were repeat diagnoses; 25 were outside hospital transfers with outside hospital stay>7 days or intubated prior to arrival; 11 were weak on admission (one with polymyositis, one with motor neuron disease, two with GBS, one with CIDP and six with other diagnoses); and three had an incomplete chart. This left 66 cases, of which four were duplicates of those identified via EMG reports, thus leaving 62, plus the 12 identified via EMG reports, leaving 74 total cases.

Seventy-four age- and sex-matched controls were identified by querying the TJUH EMR for all patients with a diagnosis of ARF and either sepsis, severe sepsis, or septic shock from the years 2010-2014. This generated a list of 2080 patients. Of those, 1238 were screened in order to identify 74 matched controls.

Seventy-four cases of ICUAW were identified with a concurrent diagnosis of sepsis and ARF. Their demographics are depicted in Table 1, and their medical characteristics are depicted in Table 2. Of the 74 cases, 51 (68.9%) had a diagnosis of CIN, 19 (25.7%) had a diagnosis of CIM, and four (5.4%) had a diagnosis of both CIN and CIM. These three groups were statistically similar in the majority of the variables assessed, with the exception of the incidence of acute pancreatitis, the use of IV antibiotics, and serum protein measurements.

Table 1 Demographics and medical characteristics of patients with critical illness polyneuropathy (CIN), critical illness myopathy (CIM) or the combination of the two (Both)

Table 2 Medical treatments and laboratory results of patients with critical illness polyneuropathy (CIN), critical illness myopathy (CIM) or the combination of the two (Both)

The demographics of the 74 patients and 74 controls are compared in Table 3, and their medical characteristics are compared in Table 4. Patients with ICUAW had significantly longer hospital and ICU stays and days of mechanical ventilation, in addition to significantly higher rates of tracheostomy and gastrostomy placements. However, they had a significantly lower in-hospital mortality rate and a higher proportion of discharges to acute inpatient rehabilitation, as compared to controls, with no difference in the incidence of severe sepsis or septic shock, acute hepatitis, or pancreatitis. Additionally, there was no significant difference in the prevalence of cirrhosis or malignancy between the two groups.

Table 3 Demographics of patients with ICU-acquired weakness (ICUAW) versus controls

Table 4 Medical treatments and laboratory results of patients with ICU-acquired weakness (ICUAW) versus controls

ICUAW patients required significantly more vasopressor support, including number of vasopressors used and days of vasopressor exposure, as well as significantly higher intravenous antibiotic quantity and duration, but with no difference specifically in aminoglycoside exposure. ICUAW patients had significantly higher paralytic and propofol exposure, but no difference in the exposure to intravenous steroids (including glucocorticoids), and midazolam.

Of the laboratory values examined, ICUAW patients had significantly lower serum protein levels when initially measured, compared to controls. There were no statistical differences at either baseline or peak between ICUAW patients and controls in admission glucose, hemoglobin A1C, average serum glucose, albumin, prealbumin, creatinine, creatine kinase or lactate.

In order to identify independent predictors of ICUAW in patients with sepsis and ARF, a binary logistic regression was performed using a backwards conditional paradigm starting with the following factors: propofol exposure, midazolam exposure, vasopressor exposure, paralytic exposure, intravenous glucocorticoid exposure, intravenous antibiotic exposure, and diagnoses of acute hepatitis, septic shock, or protein calorie malnutrition. The results are displayed in Table 5. Propofol exposure was an independent predictor of ICUAW with an increased OR of 3.131 (95% CI: 1.532-6.398). Additionally, exposure to vasopressors was also found to be an independent predictor, with an increased OR of 3.655 (95%CI: 1.586-8.426). Diagnoses of both septic shock and protein calorie malnutrition neared significance with an OR of approximately 2, but ultimately had confidence intervals including 1.

Table 5 Logistic regression using a backwards conditional model to identify independent predictors of ICUAW. Initial starting variables included exposure to intravenous antibiotics, vasopressors, paralytics, propofol, midazolam, glucocorticoids, and diagnoses of acute hepatitis, septic shock and protein-calorie malnutrition

Discussion

Since the initial description of five patients with weakness associated with sepsis and respiratory failure,Reference Bolton, Gilbert, Hahn and Sibbald 18 sepsis has emerged as an important risk factor for development of ICUAW.Reference Schweickert and Hall 2 , Reference Kress and Hall 4 , Reference Lacomis 7 , Reference Latronico and Bolton 8 , Reference Latronico, Fenzi, Recupero, Guarneri, Tomelleri and Tonin 19 - Reference Garnacho-Montero, Madrazo-Osuna, García-Garmendia, Ortiz-Leyba, Jiménez-Jiménez and Barrero-Almodóvar 24 Although several pathophysiological mechanisms have been postulated as underlying causes of ICUAW, sodium channel dysfunction, in association with sepsis, is evolving as a unifying hypothesis for both CIN and CIM.Reference Schweickert and Hall 2 , Reference Kress and Hall 4 - Reference Latronico and Bolton 8 Interestingly, the dysfunction of sodium channels has also been implicated in the electrocardiogram-related changes associated in sepsis.Reference Rich, McGarvey, Teener and Frame 25

ICUAW can occur as quickly as three days after hospital admission for sepsis,Reference Khan, Harrison, Rich and Moss 26 in keeping with a possible acquired channelopathy. Several reports demonstrate changes in biophysical properties affecting the function of sodium channels in models of sepsis. Lipopolysaccharide has been shown to directly interact with voltage-gated sodium channels and reduce sodium channel availability in transfected cell lines.Reference Haeseler, Foadi, Wiegand, Ahrens, Krampfl and Dengler 10 In a rat model of chronic sepsis undergoing cecal ligation and perforation, patch clamp techniques demonstrated decreases in both sodium current and conductance, with a hyperpolarizing shift in the inactivation of sodium channels. A cause for this shift may be an upregulation of the Nav1.5 isoform and a downregulation of the Nav1.4 due to sepsis.Reference Teener and Rich 27 , Reference Friedrich, Hund, Weber, Hacke and Fink 28

Critically ill patients are routinely provided analgesia and sedation to prevent pain and anxiety, permit invasive procedures, reduce stress and oxygen consumption, and improve synchrony with mechanical ventilation.Reference Hughes, McGrane and Pandharipande 13 Propofol, a commonly used sedative for this indication, has also been studied under patch clamp conditions. In transfected cell lines, propofol antagonized voltage-gated skeletal muscle,Reference Haeseler, Störmer, Bufler, Dengler, Hecker and Piepenbrock 16 and CNSReference Rehberg and Duch 29 sodium channels in a concentration-dependent manner, again with a hyperpolarizing shift in the sodium channel inactivation curve.Reference Rehberg and Duch 29 In a similar fashion, propofol also caused a hyperpolarizing shift in the sodium channel inactivation curve in isolated rat ventricular myocytes in a concentration dependant manner,Reference Saint 30 which may explain the EKG attenuation noted in septic patients.Reference Rich, McGarvey, Teener and Frame 25

Aside from sodium channel dysfunction and electrical inexcitability, other possible mechanisms of ICUAW include muscle protein catabolism, disorganization of muscle ultrastructure with loss of thick myosin filaments, bioenergetics failure, and impaired microcirculation.Reference Schweickert and Hall 2 , Reference Kress and Hall 4 , Reference Bloch, Polkey, Griffiths and Kemp 6 , Reference Latronico and Bolton 8 Propofol is notorious for causing hemodynamic compromise, which may have a cumulative deleterious effect on the shock of sepsis. We observed that ICUAW patients with exposure to propofol had an increased need for vasopressor support, which is in keeping with this possibility and may extend to the compromise of the microcirculation. Additionally, shock and hypoperfusion are associated with multi organ failure, which has been clearly characterized as a risk factor for ICUAW.Reference Stevens, Dowdy, Michaels, Mendez-Tellez, Pronovost and Needham 3 , Reference Latronico and Bolton 8 , Reference de Letter, Schmitz, Visser, Verheul, Schellens and Op de Coul 31 , Reference De Jonghe, Sharshar, Lefaucheur, Authier, Durand-Zaleski and Boussarsar 32

Aside from its effect at sodium channels, propofol itself may exacerbate ICUAW by other putative mechanisms. For example, anesthesia with profofol has been associated with rhabdomyolysis secondary to skeletal muscle breakdown, described in two patients receiving high rates of propofol infusion, with markedly elevated creatine kinase and histopathology demonstrating necrosis with swelling, loss of striation, and vacuole formation.Reference Stelow, Johari, Smith, Crosson and Apple 33 Propofol infusion syndrome is a rare but potentially lethal side effect of propofol, presenting with rhabdomyolysis in addition to metabolic acidosis, hyperkalaemia, hepatomegaly, renal failure, hyperlipidaemia, arrhythmia, and rapidly progressive cardiac failure.Reference Krajčová, Waldauf, Anděl, Duška and Krajcova 34 In vitro studies in cortical mixed neuronal/glial cultures have shown that relevant concentrations of propofol are involved in altered retrograde intracellular transport and neurite retraction via actin reorganization.Reference Turina, Bjornstrom, Sundqvist, Eintrei, Sciences and Medicine 35 Similar observations have not been reported in muscle tissue. Lastly, when exposed to supratherapeutic concentrations of propofol, mitochondria isolated from neurons of embryonic stem cells demonstrate increased fission and increased expression of mitochondrial permeability transition pore, inducing neurotoxicity.Reference Twaroski, Yan, Zaja, Clark, Bosnjak and Bai 36

The incidence of ICUAW is exceedingly high, occurring in approximately 50-70% of Intensive Care Unit (ICU) patients with diagnoses of sepsis or multi-organ failure.Reference Stevens, Dowdy, Michaels, Mendez-Tellez, Pronovost and Needham 3 , Reference Kress and Hall 4 In patients with a diagnosis of septic shock the incidence can rise to 70-80%.Reference Witt, Zochodne, Bolton, Grand’Maison, Wells and Young 20 , Reference Berek, Margreiter, Willeit, Berek, Schmutzhard and Mutz 23 , Reference Garnacho-Montero, Madrazo-Osuna, García-Garmendia, Ortiz-Leyba, Jiménez-Jiménez and Barrero-Almodóvar 24 However, our query of the EMR yielded a total of only 191 cases of appropriately coded CIN or CIM in the span of 14 years, suggesting ICUAW is grossly underdiagnosed. Additionally, a preponderance of ICUAW cases reported here were CIN, which contradicts previous reports suggesting CIMReference Latronico, Fenzi, Recupero, Guarneri, Tomelleri and Tonin 19 , Reference Koch, Spuler, Deja, Bierbrauer, Dimroth and Behse 37 or the combinationReference Khan, Harrison, Rich and Moss 26 occur more frequently. Both the underdiagnosis and possible misdiagnosis of patients with ICUAW may be in part due to the lack of proven effective interventions and the difficulty in making the diagnosis on the basis of the clinical exam alone. There is then the added complexity in obtaining supporting electrodiagnostics in the ICU in routine clinical practice, due to the ongoing nursing care of acutely ill patients, the presence of electrical artefacts and other limiting factors for quality studies such as patient’s sedation and edema, and the lack of clinically proven effective therapies.Reference Schweickert and Hall 2

In our cohort of ICUAW patients only 16% of diagnoses (12 patients) were made with EMG/NCS and only two patients underwent biopsy; these methods are recognized as the gold standard for diagnosis. Instead, the majority of CIN or CIM diagnoses were made on clinical grounds alone, without the use of electrodiagnostics or pathology. In a prospective study of mechanically ventilated ICU patients, using a clinical exam with low threshold criteria for the diagnosis of CIN (e.g. the presence of paresis or areflexia) was only 60% sensitive compared to EMG diagnosis,Reference Leijten, Poortvliet and de Weerd 38 suggesting many cases remain undiagnosed, as previously discussed. Direct muscle stimulation has been shown to differentiate CIN from CIM in comatose or encephalopathic critically ill patients.Reference Rich, Bird, Raps, McCluskey and Teener 39 The Medical Research Council (MRC) advocates using the MRC sum score as an initial diagnostic measure of muscle force in conscious patients who are suspected of having critical illness polyneuropathy or myopathy, where ICAW are arbitrarily diagnosed if the MRC sum score is less than 48.Reference De Jonghe, Sharshar, Lefaucheur, Authier, Durand-Zaleski and Boussarsar 32 , Reference Zhou, Wu, Ni, Ji, Wu and Zhang 40 However, our patients were intubated and sedated with several agents, including propofol, midazolam, and fentanyl, making routine use of this score in this patient population difficult. Future potential diagnostic modalities may include ultrasound,Reference Puthucheary, Rawal, McPhail, Connolly, Ratnayake and Chan 41 or peroneal nerve electrophysiological test (PENT), a simplified electrophysiological assessment of a single nerve which demonstrated a sensitivity of 100% and a specificity of 85.2% compared to comprehensive EMG/NCS in critically ill patients.Reference Latronico, Nattino, Guarneri, Fagoni, Amantini and Bertolini 42

Another limitation of this study is the significantly higher mortality rate in the control group, at 44.6% compared to the ICUAW group at 16.2%. Recently published mortality rates for sepsis are approximately 30%,Reference Mouncey, Osborn, Power, Harrison, Sadique and Grieve 43 whereas severe sepsis and septic shock may reach as high as 50%.Reference Carson, Kress, Rodgers, Vinayak, Campbell-Bright and Levitt 44 - Reference Annane, Aegerter, Jars-Guincestre and Guidet 48 The diagnosis of CIN or CIM in the setting of sepsis or septic shock may have introduced a selection bias for survivors of the acute medical illness (e.g. sepsis and multi-organ failure), which translated into decreased mortality, but also significantly higher hospital and ICU length of stay, days of mechanical ventilation, rates of tracheostomy and gastrostomy placements, and patients who were discharged to acute rehabilitation or long term care facilities (Table 2). This is in contrast to previous reports suggesting that ICUAW was a risk factor for increased in-hospital mortality.Reference Latronico, Fenzi, Recupero, Guarneri, Tomelleri and Tonin 19 , Reference Garnacho-Montero, Madrazo-Osuna, García-Garmendia, Ortiz-Leyba, Jiménez-Jiménez and Barrero-Almodóvar 24 , Reference Khan, Harrison, Rich and Moss 26 Furthermore, it is possible that the control group of septic patients with ARF may have also had undiagnosed ICUAW.

There was no statistical difference between ICUAW patients and controls with respect to the proportion of patients with the diagnosis of septic shock; age, initial, or peak serum lactate; or creatinine or glucose control, suggesting a similar level of medical acuity between the two groups. However, the use ICU scoring systems such as the APACHE II or III, SAPS II, and SOFA were not documented during the patients’ hospitalization. These classifications would have allowed for the expected mortality of the respective cohorts to be calculated, and may have helped clarify the large differences in observed mortality between the two groups.

We did, however, note that patients with ICUAW required significantly more intravenous antibiotics and vasopressors which, in light of the difference in mortality, may have been an artefact of exposure bias and longer hospitalizations, and thus must be interpreted cautiously.

Lastly, propofol exposure was assumed to be a consequence of ARF and mechanical ventilation, which were inclusion criteria for the study. In some cases, however, propofol exposure may have been the result of isolated bolus doses given for bedside procedures and not as a continuous infusion for sedation with mechanical ventilation. Additionally, the diagnosis of ICUAW may have preceded the propofol exposure in some cases. An examination of the duration of propofol exposure, measured by days of exposure, attempted to address these potential confounders. However, we estimated propofol exposure based on days of exposure, which is a crude estimate of exposure; actual individual doses may have varied considerably from patient to patient. Accurate hourly infusion rates of propofol were not charted, thus restricting the ability to explore the presence of a potential dose response relationship between propofol and ICUAW in patients with sepsis and ARF.

Both sepsis and propofol act to increase the inactivation of sodium channels. The data presented here, although significantly limited by the difficulty in making the diagnosis of ICUAW on clinical grounds alone, suggest that propofol exposure may be an independent risk factor in the development of ICUAW in patients with sepsis. On the other hand, the onset of ICUAW may prolong the need for mechanical ventilation in patients with ARF, which would in turn require longer exposure to propofol and other sedatives. Further studies, both basic science and clinical, with supporting EMG/NCS and biopsy, are warranted to confirm the supposition that propofol may act synergistically at sodium channels, along with sepsis, causing sodium channel dysfunction, leading to the impairment of excitable membranes.

Ackowledgements

We would like to thank Ms. Micky Triocci for her help querying the EMR, and Ms. Karlene Copeland for her help in querying the EMG reports.

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Disclosures

Peter Abdelmalik and Goran Rakocevic have no financial or other interests to disclose.

Statement of Authorship

Peter Abdelmalik and Goran Rakocevic both oversaw the experimental design, data retrieval, data analysis and manuscript preparation.

References

1. Fan, E, Cheek, F, Chlan, L, Gosselink, R, Hart, N, Herridge, MS, et al. An Official American Thoracic Society Clinical Practice Guideline: The Diagnosis of Intensive Care Unit–acquired Weakness in Adults. Am J Respir Crit Care Med [Internet]. 2014;190(12):1437-1446.CrossRefGoogle ScholarPubMed
2. Schweickert, WD, Hall, J. ICU-acquired weakness. Chest. 2007;131(5):1541-1549.Google Scholar
3. Stevens, RD, Dowdy, DW, Michaels, RK, Mendez-Tellez, P a., Pronovost, PJ, Needham, DM. Neuromuscular dysfunction acquired in critical illness: A systematic review. Intensive Care Med. 2007;33(11):1876-1891.Google Scholar
4. Kress, JP, Hall, JB. ICU-acquired weakness and recovery from critical illness. N Engl J Med [Internet]. 2014;370:1626-1635.Google Scholar
5. Z’Graggen, WJ, Lin, CSY, Howard, RS, Beale, RJ, Bostock, H. Nerve excitability changes in critical illness polyneuropathy. Brain. 2006;129(9):2461-2470.CrossRefGoogle ScholarPubMed
6. Bloch, S, Polkey, MI, Griffiths, M, Kemp, P. Molecular mechanisms of intensive care unit-acquired weakness. Eur Respir J. 2012;39(4):1000-1011.Google Scholar
7. Lacomis, D. Electrophysiology of neuromuscular disorders in critical illness. Muscle Nerve [Internet]. 2013, Mar [cited 2014 May 1]47(3):452-463.Google Scholar
8. Latronico, N, Bolton, CF. Critical illness polyneuropathy and myopathy: A major cause of muscle weakness and paralysis. Lancet Neurol [Internet] Elsevier Ltd 2011;10(10):931-941.Google Scholar
9. Novak, KR, Nardelli, P, Cope, TC, Filatov, G, Glass, JD, Khan, J, et al. Inactivation of sodium channels underlies reversible neuropathy during critical illness in rats. J Clin Invest. 2009;119(5):1150-1158.CrossRefGoogle ScholarPubMed
10. Haeseler, G, Foadi, N, Wiegand, E, Ahrens, J, Krampfl, K, Dengler, R, et al. Endotoxin reduces availability of voltage-gated human skeletal muscle sodium channels at depolarized membrane potentials. Crit Care Med. 2008;36(4):1239-1247.CrossRefGoogle ScholarPubMed
11. Grounds, RM, Lalor, JM, Lumley, J, Royston, D, Morgan, M. Propofol infusion for sedation in the intensive care unit: preliminary report. Br Med J (Clin Res Ed). 1987;294(6569):397-400.Google Scholar
12. Newman, LH, McDonald, JC, Wallace, PG, Ledingham, IM. Propofol infusion for sedation in intensive care. Anaesthesia. 1987;42(9):929-937.Google Scholar
13. Hughes, CG, McGrane, S, Pandharipande, PP. Sedation in the intensive care setting. Clin Pharmacol Adv Appl. 2012;4(1):53-63.Google Scholar
14. Kotani, Y, Shimazawa, M, Yoshimura, S, Iwama, T, Hara, H. The experimental and clinical pharmacology of propofol, an anesthetic agent with neuroprotective properties. CNS Neurosci Ther [Internet]. 2008, Jan [cited 2014 May 14] 14(2):95-106.CrossRefGoogle ScholarPubMed
15. Martella, G, De Persis, C, Bonsi, P, Natoli, S, Cuomo, D, Bernardi, G, et al. Inhibition of persistent sodium current fraction and voltage-gated L-type calcium current by propofol in cortical neurons: implications for its antiepileptic activity. Epilepsia [Internet]. 2005, May 46(5):624-635.CrossRefGoogle ScholarPubMed
16. Haeseler, G, Störmer, M, Bufler, J, Dengler, R, Hecker, H, Piepenbrock, S, et al. Propofol blocks human skeletal muscle sodium channels in a voltage-dependent manner. Anesth Analg [Internet]. 2001, May 92(5):1192-1198.Google Scholar
17. Hermans, G, De Jonghe, B, Bruyninckx, F, Van Den Berghe, G. Interventions for preventing critical illness polyneuropathy and critical illness myopathy. Cochrane Database Syst Rev. 2014:(1):1-68.Google Scholar
18. Bolton, CF, Gilbert, JJ, Hahn, a F, Sibbald, WJ. Polyneuropathy in critically ill patients. J Neurol Neurosurg Psychiatry. 1984;47(11):1223-1231.Google Scholar
19. Latronico, N, Fenzi, F, Recupero, D, Guarneri, B, Tomelleri, G, Tonin, P, et al. Critical illness myopathy and polyneuropathy.pdf. Lancet. 1996;347:1579-1582.Google Scholar
20. Witt, NJ, Zochodne, DW, Bolton, CF, Grand’Maison, F, Wells, G, Young, GB, et al. Peripheral nerve function in sepsis and multiple organ failure. Chest J. 1991;99(1):176-184.Google Scholar
21. Zochodne, DW, Bolton, CF, Wells, G a, Gilbert, JJ, Hahn, a F, Brown, JD, et al. Critical illness polyneuropathy. A complication of sepsis and multiple organ failure. Brain. 1987;110 ( Pt 4): 819-841.Google Scholar
22. Bolton, CF, Laverty, D a, Brown, JD, Witt, NJ, Hahn, a F, Sibbald, WJ. Critically ill polyneuropathy: electrophysiological studies and differentiation from Guillain-Barré syndrome. J Neurol Neurosurg Psychiatry. 1986;49(5):563-573.Google Scholar
23. Berek, K, Margreiter, J, Willeit, J, Berek, a, Schmutzhard, E, Mutz, NJ. Polyneuropathies in critically ill patients: a prospective evaluation. Intensive Care Med. 1996;22(9):849-855.Google Scholar
24. Garnacho-Montero, J, Madrazo-Osuna, J, García-Garmendia, J, Ortiz-Leyba, C, Jiménez-Jiménez, F, Barrero-Almodóvar, a., et al. Critical illness polyneuropathy: Risk factors and clinical consequences. A cohort study in septic patients. Intensive Care Med. 2001;27(8):1288-1296.Google Scholar
25. Rich, MM, McGarvey, ML, Teener, JW, Frame, LH. ECG changes during septic shock. Cardiology. 2002;97(4):187-196.Google Scholar
26. Khan, J, Harrison, TB, Rich, MM, Moss, M. Early development of critical illness myopathy and neuropathy in patients with severe sepsis. Neurology. 2006;67(8):1421-1425.Google Scholar
27. Teener, JW, Rich, MM. Dysregulation of sodium channel gating in critical illness myopathy. J Muscle Res Cell Motil. 2006;27(5-7):291-296.Google Scholar
28. Friedrich, O, Hund, E, Weber, C, Hacke, W, Fink, RH a. Critical illness myopathy serum fractions affect membrane excitability and intracellular calcium release in mammalian skeletal muscle. J Neurol. 2004;251(1):53-65.CrossRefGoogle ScholarPubMed
29. Rehberg, B, Duch, DS. Suppression of central nervous system sodium channels by propofol. Anesthesiology. 1999;91(2):512-520.Google Scholar
30. Saint, D. The effects of propofol on macroscopic and single channel sodium currents in rat ventricular myocytes. Br J Pharmacol. 1998;124(4):655-662.Google Scholar
31. de Letter, M, Schmitz, PI, Visser, LH, Verheul, FA, Schellens, RL, Op de Coul, DA, et al. Risk factors for the development of polyneuropathy and myopathy in critically ill patients. Crit Care Med. 2001;29(12):2281-2286.Google Scholar
32. De Jonghe, B, Sharshar, T, Lefaucheur, JP, Authier, FJ, Durand-Zaleski, I, Boussarsar, M, et al. Paresis Acquired in the Intensive Care Unit. JAMA. 2002;288(22):2859-2867.Google Scholar
33. Stelow, EB, Johari, VP, Smith, SA, Crosson, JT, Apple, FS. Propofol-associated rhabdomyolysis with cardiac involvement in adults: Chemical and anatomic findings. Clin Chem. 2000;46(4):577-581.Google Scholar
34. Krajčová, A, Waldauf, P, Anděl, M, Duška, F, Krajcova, A. Propofol infusion syndrome : a structured review of experimental studies and 153 published case reports. Crit Care [Internet]. 2015;19(1):1-9.Google Scholar
35. Turina, D, Bjornstrom, K, Sundqvist, T, Eintrei, C, Sciences, H, Medicine, E. Propofol Alters Vesicular Transport in Rat Cortical Neuronal Cultures. J Physiol Pharmacol. 2011;61(1):119-124.Google Scholar
36. Twaroski, DM, Yan, Y, Zaja, I, Clark, E, Bosnjak, ZJ, Bai, X. Altered Mitochondrial Dynamics Contributes to Propofol-induced Cell Death in Human Stem Cell–derived Neurons. Anesthesiology [Internet]. 2015;123(5):1067-1083.CrossRefGoogle ScholarPubMed
37. Koch, S, Spuler, S, Deja, M, Bierbrauer, J, Dimroth, A, Behse, F, et al. Critical illness myopathy is frequent: accompanying neuropathy protracts ICU discharge. J Neurol Neurosurg Psychiatry. 2011;82(3):287-293.Google Scholar
38. Leijten, FS, Poortvliet, DC, de Weerd, a W. The neurological examination in the assessment of polyneuropathy in mechanically ventilated patients. Eur J Neurol [Internet]. 1997;4(2):124-129.CrossRefGoogle ScholarPubMed
39. Rich, MM, Bird, SJ, Raps, EC, McCluskey, LF, Teener, JW. Direct muscle stimulation in acute quadriplegic myopathy. Muscle and Nerve. 1997;20(6):665-673.Google Scholar
40. Zhou, C, Wu, L, Ni, F, Ji, W, Wu, J, Zhang, H. Critical illness polyneuropathy and myopathy: A systematic review. Neural Regen Res. 2014;9(1):101-110.Google Scholar
41. Puthucheary, Z a, Rawal, J, McPhail, M, Connolly, B, Ratnayake, G, Chan, P, et al. Acute skeletal muscle wasting in critical illness. JAMA [Internet]. 2013;310(15):1591-1600.Google Scholar
42. Latronico, N, Nattino, G, Guarneri, B, Fagoni, N, Amantini, A, Bertolini, G. Validation of the peroneal nerve test to diagnose critical illness polyneuropathy and myopathy in the intensive care unit: the multicentre Italian CRIMYNE-2 diagnostic accuracy study. F1000Research [Internet]. 2014:1-11; Available from: http://f1000research.com/articles/3-127/v1.Google ScholarPubMed
43. Mouncey, PR, Osborn, TM, Power, GS, Harrison, D a, Sadique, MZ, Grieve, RD, et al. Trial of Early, Goal-Directed Resuscitation for Septic Shock for the ProMISe Trial Investigators. N Engl J Med. 2015;372(14):1301-1311.Google Scholar
44. Carson, SS, Kress, JP, Rodgers, JE, Vinayak, A, Campbell-Bright, S, Levitt, J, et al. A randomized trial of intermittent lorazepam versus propofol with daily interruption in mechanically ventilated patients. Crit Care Med. 2006;34(5):1326-1332.Google Scholar
45. Angus, DC, Linde-Zwirble, WT, Lidicker, J, Clermont, G, Carcillo, J, Pinsky, MR. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med. 2001;29(7):1303-1310.Google Scholar
46. Rusconi, AM, Bossi, I, Lampard, JG, Szava-Kovats, M, Bellone, A, Lang, E. Early goal-directed therapy vs usual care in the treatment of severe sepsis and septic shock: a systematic review and meta-analysis. Intern Emerg Med [Internet] Springer Milan 2015, Available from: http://link.springer.com/10.1007/s11739-015-1248-y.Google Scholar
47. Annane, D, Bellissant, E, Cavaillon, J-M. Septic shock. Lancet. 2005;365(9453):63-78.Google Scholar
48. Annane, D, Aegerter, P, Jars-Guincestre, MC, Guidet, B. Current epidemiology of septic shock: The CUB-Rea network. Am J Respir Crit Care Med. 2003;168(2):165-172.Google Scholar
Figure 0

Table 1 Demographics and medical characteristics of patients with critical illness polyneuropathy (CIN), critical illness myopathy (CIM) or the combination of the two (Both)

Figure 1

Table 2 Medical treatments and laboratory results of patients with critical illness polyneuropathy (CIN), critical illness myopathy (CIM) or the combination of the two (Both)

Figure 2

Table 3 Demographics of patients with ICU-acquired weakness (ICUAW) versus controls

Figure 3

Table 4 Medical treatments and laboratory results of patients with ICU-acquired weakness (ICUAW) versus controls

Figure 4

Table 5 Logistic regression using a backwards conditional model to identify independent predictors of ICUAW. Initial starting variables included exposure to intravenous antibiotics, vasopressors, paralytics, propofol, midazolam, glucocorticoids, and diagnoses of acute hepatitis, septic shock and protein-calorie malnutrition