Pathophysiology of acute lung injury following esophagectomy

Esophagectomy is widely considered to be a high risk surgical approach. While improvements in surgical technique, perioperative monitoring, and postoperative care have led to a reduction in mortality following esophagectomy, early mortality remains between 5% and 10%.1,2 Moreover, postoperative complications, involving predominantly the cardiorespiratory system, occur in up to 50% of patients.3 Of the numerous respiratory complications that are associated with esophagectomy, acute lung injury (ALI) and its more severe form acute respiratory distress syndrome (ARDS) are of notable significance due to their prevalence and impact on early morbidity and mortality.4 Reported rates of ALI and ARDS following esophagectomy are 23.8% and 14.5% respectively, the latter associated with 50% mortality.4 While the existence of an accepted clinical definition for ALI/ARDS5 has greatly contributed to the wider study of ALI/ARDS, continued variability in reporting practices has meant that the pathophysiology of this syndrome in the context of esophagectomy remains incompletely understood. Notwithstanding, patient-specific factors, the extent and nature of surgical trauma, and frequent requirement for periods of one-lung ventilation (OLV) are all anticipated to contribute to the development of ALI/ARDS following esophagectomy through their influence on systemic and local inflammatory responses. Consequently, the aim of this review is to define the pathophysiology of ALI/ARDS in the setting of esophagectomy and to explore potential strategies for prediction and prevention of this complication. As first described by Ashbaugh et al. in 1967, ALI/ARDS is recognized as a constellation of clinical, radiological, and physical abnormalities that collectively define an underlying picture of diffuse damage to the alveolar capillary unit and associated pulmonary gas exchange disturbance.6 Central to this process is disruption of vascular endothelial and pulmonary epithelial linings and subsequent flooding of the alveolar and interstitial spaces with proteinaceous fluid, rich in inflammatory mediations and leukocytes.7 Causes of ALI/ARDS can be broadly divided between those that result in either direct or indirect lung injury. Common sources of direct lung injury include: pneumonia, aspiration, trauma, reperfusion, and inhalation injury. Alternatively, sepsis, shock, polytrauma, and blood transfusion are common sources of indirect lung injury.8 In patients undergoing esophagectomy precipitants of ALI/ARDS may reflect both local and systemic insults. Surgery is associated with the induction of an acute phase response that is proportional to the degree and nature of the trauma sustained. Accordingly, esophagectomy, often regarded as one of the most invasive surgical procedures, is prelude to a potent metabolic inflammatory response, signs of which are readily detectable both systemically and locally within the lungs. The normal stress response to injury was first described by Cuthbertson in the early 1930's.9 Although further details have since been added, Cuthbertson's original observation of a biphasic stress response is still upheld. The early ‘shock’ phase, occurring in response to anesthesia and tissue injury, is principally characterized by a hypometabolic state with an associated rise in catecholamine and glucocorticoid levels and hemodynamic instability. A subsequent hypermetabolic state is observed during which proteins and fats are consumed and body water and salts are conserved.10 While there is limited evidence as to the specific metabolic response to esophagectomy, Sato et al. have reported that resting energy expenditure is significantly increased following esophagectomy and that this response was attenuated in patients who underwent less invasive, transhiatal resection.11 While carbohydrate oxidation increased significantly after esophagectomy, the caloric contribution of fat decreased.11 Tashiro et al. also published evidence of enhanced skeletal muscle (protein) breakdown in patients following esophagectomy.12 While serious postoperative complications, including ALI/ARDS, are likely to augment these metabolic changes following esophagectomy, this has yet to be confirmed in clinical trials. A major feature of the acute phase response to surgery is the activation of cytokine networks and the release inflammatory mediators from immunocompetent cells. This in turn supports the propagation of both a local and systemic inflammatory response that is proportional to the degree of surgical trauma and is predicted to contribute to the development of ALI/ARDS following esophagectomy. Cytokines elevated in the serum of patients following esophagectomy include interleukin (IL)-1, IL-2, IL-6, IL-8, IL-10, and tumor necrosis factor alpha (TNFα),13–24 all of which have been associated with the development of ALI/ARDS in other patient groups.7 Yamada et al. identified that the serum of patients who had undergone esophagectomy contained higher levels of IL-6 and IL-8 and that both these cytokines were highest in the serum of patients who developed pulmonary complications.15 Likewise, Morita and co-workers reported higher serum levels of IL-6 (postoperative days 0–2) and IL-8 (postoperative day 3) in patients who developed ALI after esophagectomy.24 Furthermore, administration of steroids and the adoption of lung protective ventilation strategies have both been independently shown to reduce postoperative serum cytokine levels and to improve pulmonary function.23,25 Several studies have reported finding increased levels of serum inflammatory mediators following esophagectomy including: hepatocyte growth factor (HGF), nitric oxide (NO), C-reactive protein, neutrophil elastase (NE), and high-mobility group box chromosomal protein-1 (HMGB-1).13,15,16,20,22,26,27 Of these mediators, high levels of serum HGF, NO, and HMGB-1 have been linked with the development of postoperative respiratory complications.15,27 Lung tissue is recognized to be a major source of cytokines during the early postoperative period following esophagectomy. Sakamoto et al. identified that within the first 24 hours after esophagectomy, the levels of IL-6 and IL-8 within drainage fluid from the thoracic cavity were 100-fold greater than in peripheral blood samples.28 Further investigations demonstrated elevated expression of IL-6 and IL-8 messenger RNA (mRNA) within leukocytes originating from pleural drainage fluid, but not peripheral blood, during the postoperative period. Abe et al. later gave support to these findings, identifying that the production of IL-6 was significantly greater in lung tissue samples following transthoracic esophagectomy and that IL-6 mRNA expression within the same samples was also elevated.17 Furthermore, the increased production of IL-6 within the lungs correlated with an observed peak in plasma IL-6 levels. Immunohistochemical staining revealed that the source of IL-6 production within the lung tissue was alveolar and bronchial epithelial cells and not alveolar macrophages as anticipated. Tsujimoto et al. recently reported finding higher pleural drainage fluid and serum IL-6 levels immediately after surgery and on postoperative day 1 in patients who developed pneumonia following esophagectomy.29 While Morita et al. could not confirm this finding in terms of pleural IL-6 levels, they did report higher levels of pleural IL-8 and IL-1β on postoperative day 3 in patients who developed ALI.24 Univariate analysis demonstrated that pleural IL-6 levels were associated with lowest postoperative PaO2/FiO2 ratio, a marker of ALI/ARDS.5 Changes in pulmonary inflammatory mediators are also recognized in patients who have undergone esophagectomy. Early experiments in both animals and humans identified that after transthoracic esophagectomy, the lungs release large quantities of the eicosanoid, thromboxane A2 (TXA2) into the systemic circulation; an event that coincided with an increase in both extravascular lung water and lung resistance and a decline in lung compliance.30 Pre-treatment with a thromboxane synthase inhibitor was observed to significantly inhibit postoperative lung injury.30 Schilling et al. later reported finding significantly increased TXB2 (a metabolite of TXA2) concentrations in blood taken from the postpulmonary, but not pre-pulmonary, circulation of patients who had developed ARDS after esophagectomy.31 An equivalent rise in postpulmonary TXB2 was not observed in control subjects and esophagectomy patients who did not develop ARDS. Schilling and co-workers performed a subsequent clinical trial in which patients undergoing esophagectomy were pre-treated with ketoconazole, a thromboxane synthase inhibitor.32 When compared with a group of historical controls, prophylactic treatment with ketoconazole significantly reduced the incidence of postoperative ALI.32 These findings may provide further evidence that TXA2, originating from the lungs, is associated with the development of ALI following esophagectomy. Interestingly, the same group found no difference in venous or arterial levels of leukotriene B4, a prominent marker of inflammation, in patients who developed ARDS following esophagectomy.31 Other inflammatory mediators which have been found to be elevated in the bronchoalveolar lavage (BALF) of patients following esophagectomy include: NE,33 granulocyte-colony stimulating factor (G-CSF),34 secretory leukocyte protease inhibitor (SLPI),35 and nitrogen-free radicals.36 Elevated levels of NE, G-CSF, and SLPI in the BALF of esophagectomy patients were associated with the development of postoperative respiratory complications.33–35 Despite their recognized role in the pathophysiology of ALI/ARDS, through the release of cytokines and inflammatory mediators, few studies have investigated the specific inflammatory cell response to esophagectomy. In one study, patients who developed septic complications (pneumonia) following esophagectomy were seen to have both phenotypic and functional alterations of monocytes characterized by both increased preoperative and postoperative hydrogen peroxide (H2O2) production and a prolonged postoperative decline in human leukocyte antigen-DR expression.37 Similarly, Katsuta et al. identified that patients with raised TNFα and IL-1β producing capacity of peripheral monocytes, both before and after esophagectomy, were at increased risk of developing postoperative respiratory complications.13 These patients exhibited increased plasma elastase and serum IL-6 postoperatively and went on to develop features of ALI/ARDS including reduce PaO2/FiO2 ratio and bilateral infiltrates on chest radiographs. Of the six patients who demonstrated this trait, five eventually developed pneumonia; this was in contrast to the remaining group of ‘normal’ esophagectomy patients, none of whom developed pneumonia. Kooguchi et al. observed that alveolar macrophages harvested from patients 24 hours after esophagectomy demonstrated significantly increase expression of iNOS, IL-6 and IL-8, and that the intensity of this expression was related to the development of post-operative respiratory failure.36 Finally, neutrophils harvested both preoperatively and postoperatively from patients undergoing esophagectomy were shown to possess reduced deformability secondary to changes in their cytoskeletal structure.38 It was suggested that such changes would promote neutrophil sequestration within the pulmonary microvasculature and contribute to lung injury. ARDS/ALI is defined a syndrome of inflammation and increased cellular permeability.5 Consequently, several groups have attempted to investigate the effect of esophagectomy on pulmonary vascular permeability.39,40 Rocker and co-workers presented evidence that transthoracic esophagectomy is associated with a significant increase in pulmonary vascular permeability.39 They observed that while vascular permeability was greatest on the thoracotomy (collapsed lung) side, an increase in contralateral vascular permeability was also seen. These findings demonstrate that esophagectomy necessitating single-lung collapse and OLV is associated with bilateral lung injury. A number of authors have sought to define risk factors associated with the development of ALI/ARDS following esophagectomy in an effort to predict and attenuate their effects. Poor study design and limited patient numbers however limits the strength of evidence that can be assimilated from many of these studies. Preoperative patient features reported to be associated with the development of ALI/ARDS following esophagectomy include: low body mass index,4 chronic respiratory comorbidity,41,42 cigarette smoking,4,42 and low forced expiratory volume in 1 second.42 Perioperative risk factors for ALI/ARDS include surgeon experience,4 traditional open (as opposed to minimally invasive) surgical techniques,41,43 duration of surgery,4,24 duration of OLV,4 colon interposition,24 anastomotic leak,4,24 high inspired oxygen concentration;42 and perioperative cardiorespiratory instability including the use of inotropes.4 Other studies have reported lower preoperative engagement in physical activity,44 higher perioperative fluid administration,45 lung tissue exposure to neoadjuvant radiotherapy,46 and earlier systemic inflammatory response syndrome47 as predictors of general postoperative pulmonary complications including ALI/ARDS. OLV is suggested to be a significant factor underlying the development of postoperative ALI/ARDS following transthoracic esophagectomy.48 Tandon et al. identified that the duration of OLV was an independent risk factor for the development of ALI/ARDS following esophagectomy.4 Baudouin presents four possible sources of lung injury relating to esophagectomy and the use of OLV (Fig. 1).48 Ischemia reperfusion injury: pulmonary ischemia reperfusion injury is known to occur as a result of lung collapse and re-expansion during OLV. Furthermore, animal models have confirmed that relatively short periods of lung ischemia, followed by reperfusion, are associated with the development of ALI.49 In humans, elevated markers of oxidative stress (a central feature of ischemia reperfusion injury), such as malondialdehyde, are observed in patients who have undergone OLV, supporting the involvement of ischemia reperfusion injury in the development of lung injury after esophagectomy.50 An alternative, indirect, method by which ischemia reperfusion may mediate lung injury after esophagectomy is related to the observation that intestinal perfusion, particularly at the proximal end of the reconstructed gastric tube, is significantly reduced during the operation.51 Although the importance of this pathway has so far not been investigated with respect to esophagectomy-related lung injury, it is well documented that ALI/ARDS may follow experimental intestinal ischemia reperfusion injury in animals.52 High fraction of inspired oxygen: often during OLV, the dependent lung is ventilated with a high fraction of inspired oxygen. It is proposed that high inspiratory oxygen may promote the formation of reactive oxygen species that can, in turn, induce lung damage.42,48 Ventilator-induced lung injury: it is accepted that ventilation of the lungs at an elevated tidal volume is a critical factor in the induction of ventilator-induced lung injury. This was verified by the ARDSNet trial which demonstrated a significant survival benefit associated with lower tidal volume ventilation.53 Consequently, the use of near normal tidal volumes during OLV may have a potentially adverse effect on the single ventilated lung. Several studies have reported benefits from using lung protective ventilatory strategies in patients undergoing esophagectomy.23,54 Pulmonary capillary stress failure: due to hypoxic pulmonary vasoconstriction in the collapsed lung during OLV, the ventilated lung is perfused with a greater proportion of the pulmonary blood flow. As a consequence of this vascular shunt, pulmonary arterial pressures are increased within the ventilated lung and can cause capillary stress failure.55 Sources of lung injury during esophagectomy and one-lung ventilation include: (1) ischemia (reperfusion injury) of the gastric tube, (2) ischemia of the collapsed lung, (3) production of reactive oxygen species within the ventilated lung secondary to high inspired oxygen concentration, (4) ventilator induced lung injury, (5) pulmonary capillary stress failure, (6) systemic release of inflammatory mediators affecting both lungs, and (7) reperfusion injury, secondary to re-expansion of the collapsed lung. (a) right lung, (b) aortic arch, (c) thoracic esophagus, (d) thoracic diaphragm, (e) stomach. All patients should undergo careful preoperative assessment in an attempt to identify and modify risk factors outlined above. Specific effort should be made to ensure smoking cessation and to optimize physical and nutrition status both in the preoperative and postoperative periods. Malnutrition and cachexia, frequently caused by upper gastrointestinal malignancy, are associated with immune dysfunction and skeletal muscle breakdown.12 Expiratory muscle weakness secondary to malnutrition is associated with higher rates of respiratory complications after upper abdominal surgery.56 Accordingly, exercise testing and training have been investigated as methods to improve respiratory outcomes. One method explored is inspiratory muscle training (IMT). While Dettling et al. reported that IMT led to preservation of postoperative respiratory muscle strength in patients undergoing esophagectomy, this was not shown to be associated with a reduction of postoperative lung infections.57 A recent randomized controlled pilot study did however demonstrate that postoperative respiratory complications were almost three times lower in patients receiving high intensity – compared with endurance – IMT prior to esophagectomy.58 Preoperative oral care and selective digestive decontamination have been suggested as methods to reduced infective pulmonary complications that are a major cause of ALI/ARDS, after esophagectomy.59 This hypothesis is based on the understanding that oropharyngeal flora are an important source of pathogens responsible for lung infection in critically ill patients. This is especially relevant following esophagectomy where recurrent laryngeal palsy, dysfunctional swallowing, and aspiration are frequently seen in the postoperative period. One study has reported that pre-emptive treatment of bacterial and viral airway colonization may prevent postoperative respiratory tract infection and ARDS.60 More research is however needed to further define the role of such strategies in preventing postesophagectomy lung injury. The surgical approach to esophagectomy has been specifically identified as an important factor in the development of postoperative pulmonary complications.2 Recent studies have shown lower rates of ALI/ARDS following minimally invasive esophagectomy.41,43 Minimally invasive esophagectomy is associated with a shorter systemic inflammatory response61 and in the case of transhiatal procedure can avoid thoracotomy and OLV. While the adoption of less invasive surgical techniques is shown to reduced pulmonary complications, it may however jeopardize the opportunity for complete oncological clearance, in turn affecting long-term survival.62 Several ongoing randomized controlled trials are tasked with finding further evidence for the clinical efficacy of minimally invasive esophagectomy. Protective ventilator strategies may be in to attenuate lung injury associated with esophagectomy and OLV. One randomized controlled trial which the of and the adoption of lower tidal volumes during OLV reported that these led to a reduction in the systemic inflammatory response IL-6 and lung and earlier This study did not however identify a reduction in postoperative A randomized controlled trial did however report a reduction in pulmonary including ALI/ARDS, in patients undergoing minimally invasive esophagectomy with lung protective ventilation tidal volume and study that with of ventilation was both and in the context of esophagectomy in the following esophagectomy is common and may contribute to lung injury and respiratory While several recent studies have reported the of early following esophagectomy, often as of an there are as yet there are no studies at the early and its to the development of ALI/ARDS. This an where further research is Several studies have investigated the use of ventilation for the of ALI/ARDS following esophagectomy. While the use of in this context remains due to its potential on anastomotic it has been shown to reduce the risk of and to from thoracotomy and abdominal influence respiratory of and in turn the development of respiratory may be to systemic due to the with respiratory and of Furthermore, has been found to attenuate the systemic inflammatory response following et al. have also demonstrated the potential benefit of anesthesia in the prevention of respiratory In to preoperative physical training with postoperative has been as an important factor in the prevention of pulmonary Despite there are limited studies which for a direct between the use of and the prevention of respiratory complications following esophagectomy. One study did identified lower rates of respiratory complications in those patients receiving postoperative The methods and duration of following esophagectomy may also an opportunity to respiratory While a are frequently to the and gastric prolonged for up to days was not shown to major respiratory complications compared with early by Furthermore, one study reported higher rates of respiratory tract in patients with as opposed to after esophagectomy, a finding that was to Similarly, other authors have shown that after esophagectomy was to drainage in the prevention of postoperative respiratory complications, including respiratory A number of studies have investigated the of a of on the development of esophagectomy respiratory complications, including of these studies are limited by and use of historical control but a role for 1 the that have been shown to have either direct or indirect benefit in terms of ALI/ARDS after esophagectomy. with possible for the prevention of ALI/ARDS following esophagectomy of six randomized controlled trials. PaO2/FiO2 is of the for ALI/ARDS.5 acute lung lung NE, neutrophil systemic inflammatory response with possible for the prevention of ALI/ARDS following esophagectomy of six randomized controlled trials. PaO2/FiO2 is of the for ALI/ARDS.5 acute lung lung NE, neutrophil systemic inflammatory response Despite a major source of morbidity and mortality following esophagectomy, the pathophysiology of ALI/ARDS in this context is and remains incompletely understood. The review a and of the the local and systemic inflammatory changes that are observed in the perioperative period. of these changes can be linked to the development of lung injury and for Further randomized controlled trials are needed to confirm the benefit of both clinical and for the prevention of ALI/ARDS following esophagectomy. should also be made at the of preoperative assessment to identify and known risk factors of postoperative patients should be to undergo less invasive procedures, but not to the of oncological