BLI is characterised by the clinical triad of (1) apnea, (2) bradycardia, and (3) hypotension and may occur without obvious external injury to the chest.5 Additionally, the blast waves’ impact upon the lung results in tearing, haemorrhage, contusion, and oedema. The clinical sequel of BLI is a rapid respiratory deterioration and progressive hypoxia with resultant ventilation perfusion mismatch and subsequent acute respiratory distress syndrome (ARDS).3,6,7 The rapid ARDS picture that develops in BLI patients is a direct result of the high pressure wave front passing through the interfaces between air, alveolar, tissue and blood vessels. This pressure front causes chest wall displacement toward the spinal column, leading to transient high intrathoracic pressure. The elevated intrathoracic pressure leads to tearing of the alveolar septa, stripping of airway epithelium, and rupture of alveolar spaces with consequent alveolar hemorrhage, edema, and alveolovenous fistulae.7,8 Currently, there is no standardised assessment criteria for the diagnosis of BLI, however, is typically confirmed by clinicians from the following: chest radiographs showing a butterfly appearance (with or without pneumothorax) on admission and increased haziness in serial chest radiographs; the presence of burn injuries; and smoke inhalation of the upper airways as seen at bronchoscopy.
Interventional Lung Assist Devices
The notion of a lung assist device (LAD) was first raised in 1967 by Rashkind et al9 who proposed a pumpless oxygenator for temporary lung assist in cystic fibrosis, ARDS and congenital heart disease. This vision could not be realised with the technologies available at the time; however, over the past decades critical care medicine has made tremendous contributions to improve outcomes in patients suffering from acute lung injury (ALI).10 Key technologies for LAD include diffusion membranes to avoid plasma leakage in prolonged applications, long-term coating technologies and homogeneous distribution of blood flow. Currently, three concepts for LAD are being explored: (1) Interventional LAD for percutaneous attachment to the systemic circulation creating an arteriovenous shunt. This device is for single use and does not require a blood pump due to its insertion into the femoral artery and vein; (2) Intravascular gas exchange devices for single-needle venous access have been designed for implantation in the vena cava or the pulmonary artery. A pulsating balloon in the membrane bundle or an impeller blood pump can be employed to optimise blood flow around the gas exchange fibers or across the device; and (3) Total artificial lungs to completely replace pulmonary gas exchange function.11 LAD are viewed as an adjunct to mechanical ventilation that allow for optimised lung protective ventilation, thus giving the lungs time to heal and provide a bridge to recovery, or a bridge to transplantation after acute lung injury. Dembinski et al12 tested the safety and efficacy of a LAD (Delta Stream Rotary Blood Pump) in a controlled trial on animals with experimental ALI. The results from this study showed that in animals (N = six pigs) with ALI, hoemodynamics remained stable and gas transfer across the LAD was optimal with two animals showing a marked increase in PaO2 and carbon dioxide (CO2) removal was effective in all animals. Although the results from this study cannot be generalised to all ALI they give an insight into the potential benefits that LAD pose in treating ALI and BLI.Impetus for the literature review
During the period March till July 2005 at the 332nd Expeditionary Medical Group (EMDG), a United States Air Force Hospital (USAF) based in Iraq, several patients suffering severe BLI due to improvised explosive devices (IED) were treated. The management of these patients within the intensive care unit (ICU) became extremely difficult as they developed permissive hypercapnia, severe acidosis and ARDS. To prevent further deterioration and minimise ventilator induced lung injury these patients were trialed on an interventional LAD called the NovaLung.® This device showed promising results with significant improvements seen in patients’ acidosis, oxygenation and reduced ventilator support. Consequently, these improvements enabled the patients to be aeromedical evacuated to a level four military hospital in Germany. The trials conducted at the 332nd EMDG have not been formally reported in the literature; however it is suggested that timely diagnosis and correct treatment of BLI will result in improved outcomes.13
Aims
The aim of this paper is to identify through an extensive literature review: (1) the effectiveness of lung assist devices in blast lung injury; (2) the recommended treatment modalities for BLI; and (3) that this paper will provide current information ensuring medical and nursing staff within the Australian critical care setting become familiar with the management of BLI.
Review process
Literature on lung assist devices and treatment modalities for BLI in the critical care setting from January 1995 until March 2006 was reviewed using the CINAHL, MEDLINE, Cochrane Library, Blackwell Synergy, and ProQuest databases. Key words utilised were: blast lung injury, acute lung injury, blast lung, barotrauma, lung assist devices, trauma management, treatment modalities, acute respiratory distress syndrome and extracorporeal membrane oxygenation. Combinations of these words were also used (e.g. lung assist devices – blast lung, acute respiratory distress syndrome – lung assist device, extracorporeal membrane oxygenation – blast lung, treatment modalities – blast lung). The lack of research conducted in this area was demonstrated by the fact that the review produced no primary source articles reporting the use of lung assist devices in blast injured patients. Consequently, the review process was broadened to incorporated all research conducted on subjects with injuries that would reflect a blast lung type injury (e.g. ALI, ARDS, and barotrauma). This broadening of the review process identified only three relevant primary source articles reporting the use of LAD for ALI/ARDS in clinical trials and three retrospective case studies of BLI patients. This paper reviews these articles and Table 1 displays the authors, design, sample and main findings of the reviewed articles.
Literature review
What is the effectiveness of lung assist devices in blast injured patients?
The articles reviewed indicated that lung injury developing within 24 hours after an explosion is typically classified as ARDS or ALI, depending on the severity of injury.13 The three retrospective BLI case studies reviewed indicate that despite the severe hypoxemia caused by explosions, timely diagnosis and correct treatment will result in improved outcomes.8,13 Furthermore, Avidan et al8 was the only retrospective study to follow up the long term outcome of his cohort. Of the 28 survivors (one patient died 24 hrs after admission from sepsis and multiorgan failure), 75% responded to their telephone interview (median time of follow up was three years). This follow up reported that sixteen patients (76%) were free of respiratory symptoms and only five patients (24%) reported some degree of respiratory dysfunction leading them to conclude that BLI will have good outcomes if treated promptly and correctly.8
Clinical classification of BLI
Initiating the appropriate treatment for patients with BLI is dependent upon the severity of the blast injury and correct diagnosis. In assessing the presenting symptoms of BLI, Pizov et al14 attempted to develop a BLI severity score and suggests that stratification of the severity of lung injury produced by blasts may be useful in the treatment and prediction of patient outcomes. The proposed BLI severity score is based on three objective signs: hypoxemia (PaO2/FiO2 ratio), chest radiograph findings, and the presence of bronchopleural fistula. The score defined three levels of injury: (1) Mild: PaO2/FiO2 ratio >200 mmHg, localised lung infiltrates and no pneumothorax; (2) Moderate: PaO2/FiO2 ratio of 60 to 200 mmHg and diffuse (bilateral/unilateral) lung infiltrates with or without pneumothorax; and (3) Severe BLI: PaO2/ FiO2 ratio < 60 mmHg, bilateral lung infiltrates and bronchopleural fistula. Hypoxia exists in all patients with BLI and both Pizov et al14 and Sorkine et al13 have incorporated this objective measure into their lung injury scores. In Sorkine et al13 retrospective analysis patients were assigned lung injury scores (LIS) according to the criteria of Murray et al.15 The Murray score is a scale of 0 to 4 for four parameters: (1) evaluation of chest radiographs, (2) PaO2/FiO2 ratio, (3) level of positive end expiratory pressure (PEEP), and (4) lung compliance. The LIS was obtained through modification of the Murray score by adding individual criteria scores and dividing the sum by the number of variables used: no injury = 0, mild to moderate = 0.1 to 2.5, severe > 2.5, and maximal score = 4.0. On admission, the LIS for Sorkine et al13 patient group was 3.2 +/- 0.3 indicating major lung injury while Pizov et al14 group showed 33% with mild BLI, 40% had moderate, and 27% had severe BLI. Pizov et al14 criticises this modified Murray score or LIS because when applied to his patient group at 6 and 24 hours as it did not differentiate between patients with moderate BLI and those with severe BLI. Moreover, Pizov et al14 argues that the most critical and dynamic period in which primary blast lung injury develops is during the first 24 hours. However, 24 hours post blast injury Pizov et al14 study demonstrated good correlation between the proposed BLI score and the modified Murray score. The reliability and validity of both the BLI severity scoring system and LIS has not been established and its application in Pizov et al14 and Sorkine et al13 studies are limited due to the small sample size and retrospective analysis. Regrettably, Avidan et al8 was not able to retrospectively apply the Murray score or any other measure of injury severity due to the failure of the hospital to hold a trauma registry and the lack of relevant data in patient charts. Therefore, the comparability of this study is questionable despite that it represents the largest reported series of blast lung injuries in the literature to date.
BLI patient management and treatment Modalities
In the retrospective study conducted by Pizov et al14 following primary resuscitation or surgery all patients with primary BLI (N=15) were admitted to the ICU. Similarly, in the Avidan et al8 study of BLI patients (N=29) all required ICU admission. Although all the patients included in the Avidan et al8 study were admitted to ICU, the authors acknowledge that a BLI patient may have been admitted to a ward bed and they were unable to retrospectively identify patients in this category. Additionally, even though the Sorkine et al13 study group of blast lung injured patients (N=17) were all admitted to ICU, they represented only 5.6% of the total number of casualties injured during three bomb blasts in Tel Aviv. All three of the retrospective studies had similar distributions in the mechanism of lung injury with N=24 (83%) patients reviewed in Avidan et al8 study injured in a closed space (defined as a bus or café); N=10 (59%) of patients in Sorkine et al13 study were victims of an explosion in a bus; and all of the patients within the Pizov et al14 study were in two civilian bus explosions.
Analysis of BLI patient records across all three retrospective studies revealed that to correct hypoxaemia and respiratory distress endotracheal intubation and mechanical ventilation was initiated either at the scene, during initial resuscitation in the emergency department, in the operating theatre, or in the ICU. In the Avidan et al8 study, 22 patients (76%) required intubation and mechanical ventilation, 14 patients (93%) in the Pizov et al14 study and 17 patients (100%) in the Sorkine et al13 retrospective study. Sorkine et al13 notes that in ARDS cases caused by other means, the large area of ruptured lung in BLI patients makes them prone to develop unique complications from mechanical ventilation. Furthermore, positive pressure ventilation and PEEP should be avoided whenever possible because of the risk of pulmonary alveolar rupture and subsequent arterial air embolism.14 Air embolisms were clinically suspected in two patients (7%) in the Avidan et al8 study, however, the mode of ventilation, level of PEEP, and outcomes of these patients is not documented in the results. In the Pizov et al14 study of the five patients with mild BLI, one received oxygen through a face mask whereas the others received volume-controlled (VCV) or pressure support ventilation (PSV) with PEEP that did not exceed 5 cm H2O. Of the six patients with moderate BLI two were ventilated with VCV and four with pressure controlled inverse-ratio ventilation (PCIRV) and received PEEP levels up to 15 cm H2O.14 Similarly, the levels of PEEP in the Avidan et al8 cohort ranged from 0 to 15 cm H2O (median 7.5 cm H2O). It is impossible to compare this with the Sorkine et al13 study as the results makes no reference to the PEEP levels used in the mechanical ventilation of its patients.
A possible explanation for this lack of data in the Sorkine et al13 study is that he used a respiratory management strategy based on volume controlled synchronised intermittent mandatory ventilation (VC-SIMV) with small tidal volumes (VT) and low peak inspiratory pressures (PIP) together with permissive hypercapnia. In limiting PIP through reduced VT, the authors argue that ventilator induced lung damage is reduce. Furthermore, permissive hypercapnia which results in alveolar hypoventilation, respiratory acidosis, and lower ventilatory pressures may limit pulmonary over distension in severe lung injury.13 The initial mechanical ventilation variables for the 17 patients in this study were a VT of 6.4 +/- 0.6 ml/ kg and a mandatory ventilator respiratory rate of 17 +/- 5 bpm, resulting in PIP of 35 +/- 0.3 cm H2O.13 Shortly after instituting mechanical ventilation with limited VT, four patients developed elevated PaCO2 (93 +/- 12.4 mmHg) which resulted in an associated reduction in arterial pH (7.13 +/- 0.08). Until this time Sorkine et al13 made no attempt to control PaCO2 levels until the arterial pH fell below 7.20, at which time the mandatory respiratory rate was increased in increments of two breaths per minute until the pH rose to greater than 7.50. These low pH values responded to the increased respiratory rate and Sorkine et al13 state that the patients experienced no adverse metabolic or haemodynamic effects as measured by changes in mean arterial pressure (MAP), cardiac index (CI), central venous pressure (CVP), or peripheral vascular resistance (PVR). Of the 17 patients, two patients (12%) died while in ICU from severe penetrating head injuries. Despite the authors’ assurances that there was no organ system dysfunction related to the respiratory acidosis they have not outlined which statistical tool was used to measure the significance level in these baseline physiological measures. Therefore, due to the lack of controlled studies utilising this ventilatory strategy the conclusions made by Sorkine et al13 should be viewed cautiously.
Special modes of ventilation or unconventional therapies were applied in both Pizov et al14 and Avidan et al8 more severely lung injured patients. In the Pizov et al14 study four severe BLI patients developed extreme hypoxia (PaO2/FiO2 < 60 mmHg) together with bronchopleural fistulae resulting in the following management: independent lung ventilation (one patient), extracorporeal membrane oxygenation (ECMO) (one patient), and a combination of nitric oxide (NO) inhalation and high frequency jet ventilation (HFJV) (two patients). Similarly, three patients in the Avidan et al8 study were trialed on HFJV (two patients), NO inhalation (one patient) and excluding the patient trialed on ECMO, both studies patients’ oxygenation improved, flow reduced through the bronchopleural fistula and lower ventilation pressures were required. Although NO has become increasing popular for the treatment of severe ARDS, it is not the first choice for treatment of hypoxaemia in lung injury.16 Moreover, only three BLI patients were treated with NO making it impossible to draw any conclusions from the results of these studies. Pizov et al14 states that HFJV is recommended for ventilation of patients with bronchopleural fistula, however, most clinicians would argue that they can be ventilated adequately with conventional mechanical ventilation.17 Despite this argument, the results from these studies demonstrated that four patients with severe BLI (two with a bronchopleural fistula) were successfully ventilated with HFJV. As with patients treated with NO, no direct correlation can be made in regards to the effectiveness of HFJV and improved outcomes of severely lung injured patients due to the small sample size of the studies and the significance of its combination with NO in the Pizov et al14 study. The patient in the Avidan et al8 study trialed on ECMO two hours after the explosion experienced severe refractory hypoxaemia, shock, massive haemoptysis, and intrapulmonary bleeding increased upon heparin administration. Ultimately this patient died and the use of ECMO as a last resort in a severely BLI patient was stated by Avidan et al.8 Regrettably, the authors did not detail the type of ECMO device used or how it was managed in combination with mechanical ventilation.
Effectiveness of LAD in ARDS and ALI
The prospective study conducted by Liebold et al18 was the first clinical report on the use of LAD applied to patients suffering from severe ARDS. Patients were selected for this study based on the consensus that if pulmonary injury could not be reversed the patient would die. Consequently, 20 patients (aged 41 +/- 16 years) with ARDS and failing conventional respiratory therapy (HFJV, surfactant replacement, prone positioning) were recruited through referral to either a cardiothoracic, surgical or medical ICU. The minimum heamodynamic requirements for inclusion in this study were a cardiac output (CO) > 6 L/min, and MAP > 70 mmHg which resulted in septic and cardiac failure patients being excluded from the study group. The authors aim was to test the feasibility and effectiveness of a pumpless extracorporeal LAD in patients with ARDS.18 Liebold et al18 justified the use of the pumpless membrane oxygenator (MO) (Quadrox Spezial™, Jostra Inc., Hirrlingen, Germany) because it is based on heparin coated hollow fibre technology thereby reducing the requirements for systemic anticoagulation, the risk of thrombus formation, and haemolysis of blood through a mechanical pump. Similarly, the studies conducted by David and Heinrichs19 and Reuttimann et al20 are both a case report of a patient with severe ARDS who are trialed on the same pumpless interventional LAD (NovaLung™, GmbH, Hechingen, Germany). The similarities between these studies ceases there as David et al19 utilises the LAD in combination with high frequency oscillatory ventilation (or HFJV), Reuttimann et al20 aimed for apneic ventilation, while Liebold et al18 reduced mechanical ventilation to achieve a more normal inspiratory: expiratory (I:E) ratio, reduced PEEP, and a reduced maximum airway pressure (Pmax). The mechanism of injury for each patient which resulted in the development of ARDS in these studies also varies dramatically. All Liebold et al18 patients’ developed ARDS from differing causes (e.g. pneumonia, lung contusion) while Reuttimann et al20 case report details the treatment a 15 year old girl who fell 15 metres down a rock face and developed severe ARDS two weeks into her admission following a bacterial pneumonia. In contrast, David et al19 reports on a 30 year old male who aspirated paraffin oil whilst fire eating, and rapidly developed hypoxaemia and ARDS (within eight hours). Initially, David et al19 trialed pressure controlled ventilation (PCV), with PEEP levels increased to 20 cm H2O, mean airway pressures of 27 cm H2O, and despite repeated recruitment maneuvres, oxygenation did not improve responded to the increased respiratory rate and Sorkine et al13 state that the patients experienced no adverse metabolic or haemodynamic effects as measured by changes in mean arterial pressure (MAP), cardiac index (CI), central venous pressure (CVP), or peripheral vascular resistance (PVR). Of the 17 patients, two patients (12%) died while in ICU from severe penetrating head injuries. Despite the authors’ assurances that there was no organ system dysfunction related to the respiratory acidosis they have not outlined which statistical tool was used to measure the significance level in these baseline physiological measures. Therefore, due to the lack of controlled studies utilising this ventilatory strategy the conclusions made by Sorkine et al13 should be viewed cautiously.
Special modes of ventilation or unconventional therapies were applied in both Pizov et al14 and Avidan et al8 more severely lung injured patients. In the Pizov et al14 study four severe BLI patients developed extreme hypoxia (PaO2/FiO2 < 60 mmHg) together with bronchopleural fistulae resulting in the following management: independent lung ventilation (one patient), extracorporeal membrane oxygenation (ECMO) (one patient), and a combination of nitric oxide (NO) inhalation and high frequency jet ventilation (HFJV) (two patients). Similarly, three patients in the Avidan et al8 study were trialed on HFJV (two patients), NO inhalation (one patient) and excluding the patient trialed on ECMO, both studies patients’ oxygenation improved, flow reduced through the bronchopleural fistula and lower ventilation pressures were required. Although NO has become increasing popular for the treatment of severe ARDS, it is not the first choice for treatment of hypoxaemia in lung injury.16 Moreover, only three BLI patients were treated with NO making it impossible to draw any conclusions from the results of these studies. Pizov et al14 states that HFJV is recommended for ventilation of patients with bronchopleural fistula, however, most clinicians would argue that they can be ventilated adequately with conventional mechanical ventilation.17 Despite this argument, the results from these studies demonstrated that four patients with severe BLI (two with a bronchopleural fistula) were successfully ventilated with HFJV. As with patients treated with NO, no direct correlation can be made in regards to the effectiveness of HFJV and improved outcomes of severely lung injured patients due to the small sample size of the studies and the significance of its combination with NO in the Pizov et al14 study. The patient in the Avidan et al8 study trialed on ECMO two hours after the explosion experienced severe refractory hypoxaemia, shock, massive haemoptysis, and intrapulmonary bleeding increased upon heparin administration. Ultimately this patient died and the use of ECMO as a last resort in a severely BLI patient was stated by Avidan et al.8 Regrettably, the authors did not detail the type of ECMO device used or how it was managed in combination with mechanical ventilation.
Effectiveness of LAD in ARDS and ALI
The prospective study conducted by Liebold et al18 was the first clinical report on the use of LAD applied to patients suffering from severe ARDS. Patients were selected for this study based on the consensus that if pulmonary injury could not be reversed the patient would die. Consequently, 20 patients (aged 41 +/- 16 years) with ARDS and failing conventional respiratory therapy (HFJV, surfactant replacement, prone positioning) were recruited through referral to either a cardiothoracic, surgical or medical ICU. The minimum heamodynamic requirements for inclusion in this study were a cardiac output (CO) > 6 L/min, and MAP > 70 mmHg which resulted in septic and cardiac failure patients being excluded from the study group. The authors aim was to test the feasibility and effectiveness of a pumpless extracorporeal LAD in patients with ARDS.18 Liebold et al18 justified the use of the pumpless membrane oxygenator (MO) (Quadrox Spezial™, Jostra Inc., Hirrlingen, Germany) because it is based on heparin coated hollow fibre technology thereby reducing the requirements for systemic anticoagulation, the risk of thrombus formation, and haemolysis of blood through a mechanical pump. Similarly, the studies conducted by David and Heinrichs19 and Reuttimann et al20 are both a case report of a patient with severe ARDS who are trialed on the same pumpless interventional LAD (NovaLung™, GmbH, Hechingen, Germany). The similarities between these studies ceases there as David et al19 utilises the LAD in combination with high frequency oscillatory ventilation (or HFJV), Reuttimann et al20 aimed for apneic ventilation, while Liebold et al18 reduced mechanical ventilation to achieve a more normal inspiratory: expiratory (I:E) ratio, reduced PEEP, and a reduced maximum airway pressure (Pmax). The mechanism of injury for each patient which resulted in the development of ARDS in these studies also varies dramatically. All Liebold et al18 patients’ developed ARDS from differing causes (e.g. pneumonia, lung contusion) while Reuttimann et al20 case report details the treatment a 15 year old girl who fell 15 metres down a rock face and developed severe ARDS two weeks into her admission following a bacterial pneumonia. In contrast, David et al19 reports on a 30 year old male who aspirated paraffin oil whilst fire eating, and rapidly developed hypoxaemia and ARDS (within eight hours). Initially, David et al19 trialed pressure controlled ventilation (PCV), with PEEP levels increased to 20 cm H2O, mean airway pressures of 27 cm H2O, and despite repeated recruitment maneuvres, oxygenation did not improve.
