Ventilator-Induced Lung Injury (VILI)

Etiology

The predominant mechanisms by which the ventilator-induced lung injury occurs include alveolar overdistention (volutrauma), barotrauma, atelectotrauma, and inflammation (biotrauma). Other mechanisms that are attributed include adverse heart-lung interactions, deflation related, and effort induced injuries.[4] Related factors being studied in this context also include heterogeneous local lung mechanics, alveolar stress frequency, and stress failure of pulmonary capillaries. Variation in the expression of genetically determined inflammatory mediators has been known to affect VILI susceptibility.[5] 

In a study on  332 mechanically ventilated patients who were not having ARDS at the initiation of ventilation, the risk factors found for ventilator-induced lung injury were larger tidal volume, blood product transfusion, acidemia, and history of restrictive lung disease.[6] Though factors such as respiratory acidosis, respiratory rate, pulmonary vascular pressures, and body temperature are found to be associated with ventilator-induced lung injury, many experts consider them only as second-order effects at this stage.

The excessive stretch from high tidal volumes results in volutrauma. Faridy et al., in their study on dogs, found that increasing the tidal volume and decreasing the PEEP resulted in lower lung volumes for similar transpulmonary pressures.[7] They concluded that high tidal volume and lower PEEP resulted in high surface -active forces, which could cause collapse and inflammation of the lungs. Dreyfuss et al., in a 1988 paper, described the development of pulmonary edema in animals undergoing ventilation at high tidal volumes. It was also noticed that such edema did not develop in animals with similar airway pressures when the lower tidal volume is ensured with straps around the chest and abdomen.[8] A randomized controlled trial by Amato et al., and subsequently, the landmark ARDS Nett study, have indisputably proved that low tidal volume ventilation improves the morality in ARDS patients.[9][3] Even a higher tidal volume due to high patient efforts on non-invasive ventilation could result in self-inflicted lung injury.[4]

Barotrauma is a pressure-related lung injury. Limiting the inflation pressure to prevent overdistension has conventionally been used as a part of the lung-protective strategy(i.e., plateau pressure < 28 to 30 cms H2O) for ARDS patients. Air leaks, pneumothoraces, and pneumomediastinum could result from overdistention. One also needs to understand that regional lung overdistention is a key factor for such ventilator-induced lung injuries. However, the evaluation of local lung mechanics is experimental at this stage. Transpulmonary pressure, which is the difference between alveolar pressure and pleural pressure, is the pressure that keeps the lung inflated when the airflow is zero at end inspiration. Hence there is a strong relation between transpulmonary pressure and tidal volume. Plateau pressure has been used as a surrogate of transpulmonary pressure at the bedside despite certain inherent limitations.

Animal experiments have shown that cyclical opening and closing of the atelectatic alveoli during the respiratory cycle could damage the adjacent non- atelectatic alveoli and airways by shear stress forces.[10][11][12] This mechanism is called atelectotrauma. The application of optimal PEEP is important in the prevention of atelectrauma. Higher PEEP can cause alveolar overdistension, and lower PEEP may be inadequate to stabilize the alveoli and keep them open. The first in vivo study on VILI was published in 1974 by Webb and Tierney who found that rats ventilated at high airway pressures without PEEP died shortly with florid hemorrhagic pulmonary edema, and this could be mitigated by the application of PEEP  while maintaining the same airway pressure.[13]

Biotrauma is the release of inflammatory mediators from the cells in the injured lungs in response to volutrauma and atelectotrauma. In ventilator-induced lung injury, the neutrophils, macrophages, and probably alveolar epithelial cells secrete various inflammatory mediators, including TNF-alpha, interleukins 6 & 8, transcription factor nuclear factor(NF)-kB, and matrix metalloproteinase-9. These cytokines could trigger detrimental effects locally and systemically, resulting in multiorgan failure.

Adverse heart-lung interaction could result in ventilator-associated lung injury, especially in the setting of high tidal volume and low PEEP, as observed in animal studies.[13][14] During inspiration, the pulmonary blood flow is significantly decreased due to compression of the right ventricular cavity by the expanding lung, and the blood flow is exaggerated during expiration. This results in a  cyclic occurrence of high flow- no flow-high flow state, which damages pulmonary capillaries' endothelium, termed capillary stress failure. An endothelial injury could result in increased permeability of capillaries, promoting leakage of protein and water, resulting in pulmonary edema. Within a period of about 20 minutes, left ventricular failure with pulmonary edema ensues as a result of right ventricular failure and RV dilatation, pushing the interventricular septum towards the left ventricle, thereby increasing the left ventricular end-diastolic pressure.

In a 2018 publication, Katira et al. showed that abrupt deflation after sustained inflation could cause ventilator-induced lung injury in rat models.[15] Extrapolating these findings into human scenarios, any abrupt disconnection from the mechanical ventilator could potentially cause lung injury by loss of PEEP with resultant alveolar collapse. The authors have termed this phenomenon as lung deflation injury. The authors attribute this phenomenon to decreased cardiac output during sustained inflation, which is usually compensated by increased systemic vascular resistance to maintain blood pressure. When there is sudden deflation, the cardiac output becomes normalized but faces significant afterload due to the increased systemic vascular resistance, which results in back pressure causing increased left ventricular end-diastolic pressure and pulmonary edema. Another contributory factor is the high pulmonary forward blood flow after sudden deflation causing sudden high pressure in capillaries, causing capillary injury. The authors suggest avoiding open suctioning and slow lowering of PEEP in ventilated ARDS patients. Abrupt disconnection from NIV could also cause potential harm, as per the authors.

The following points need to be noted to understand the concept of effort-induced lung injury(self-inflicted lung injury). Early paralysis has been known to improve lung function and mortality.[16][17][18]. A post hoc analysis of the LUNG SAFE study showed that patients with severe ARDS fared worse on NIV than mechanical ventilators.[19][20] Airway pressure release ventilation(APRV) in a single centered randomized trial showed high mortality in the interventional group, promoting spontaneous breathing.[21] 

Patients with already injured lung are much more susceptible to effort induced lung injury. The proposed mechanisms of lung injury during spontaneous efforts at breathing include increased pleural negative pressure during spontaneous efforts causing increased transpulmonary pressure resulting in higher tidal volumes causing volutrauma, pendelluft phenomenon resulting in tidal recruitment of injured alveoli, increased transvascular pressure predisposing to pulmonary edema in volume cycled mode, and patient-ventilator asynchrony.[4] 

Driving pressure is the difference between the plateau pressure and PEEP, and is also derived by dividing Vt by static compliance of the respiratory system (Crs). In 2002, Estenssoro et al. first described the consistent ability of driving pressure values in the first week to identify survivors versus non-survivors in ARDS patients (along with other variables, including P/F ratio and SOFA).[22] Amato et al., in a 2015 meta-analysis on more than 3500 patients, showed that driving pressure is the physical variable that has correlated best with mortality.[23] A driving pressure-based ventilatory strategy to prevent lung injury in ARDS patients on a ventilator has been proposed and debated actively.[24][25]

Epidemiology

Data regarding the incidence of ventilator-associated lung injury is unavailable at this stage. However, the injury is expected to be more frequent in ARDS patients. When the lungs are pathologically injured, the mechanical ventilation (especially in the absence of lung-protective strategies) will further damage the injured lungs via the previously mentioned mechanisms. A study on 332 mechanically ventilated patients without acute lung injury at the initiation of mechanical ventilation observed ventilator-induced lung injury in 24 % of the patients within the first five days.[6]

Pathophysiology

Ventilator-induced lung injury mostly occurs in patients with underlying physiological insults such as sepsis, trauma, and major surgery, where the immune system is already primed for a cascading response to mechanical lung injury. Volutrauma, atelectrauma, and biotrauma are the key mechanisms of ventilator-induced lung injury, although each component's relative contribution is unclear at this stage. Alveolar distention and injury cause increased alveolar permeability, alveolar and interstitial edema, alveolar hemorrhage, and formation of hyaline membranes, resulting in diminished functional surfactant with resultant alveolar collapse.[12][26][27]

History and Physical

Clinical diagnosis of ventilator-induced lung injury is made at the bedside with a high degree of suspicion and ruling out of other causes that could closely mimic the picture. The patient on a mechanical ventilator typically develops worsening hypoxemia with low PaO2 and a fall in saturation. X-ray chest will show bilateral diffuse alveolar/interstitial infiltrates without cardiac enlargement. A CT scan thorax may show heterogeneous consolidation and atelectasis with focal areas of hyperlucencies suggestive of alveolar overdistension.

New-onset pulmonary infections and pulmonary edema are the commonest differential diagnosis to be ruled out initially. A thorough bedside clinical evaluation needs to be performed to exclude new-onset bronchospasm or crackles. Evidence for pneumothorax, pleural effusion, limb edema, ascites, and intra-abdominal hypertension has to be evaluated. History of drug allergy or blood product transfusion needs to be clarified. Ventilatory settings need to be cross-checked to look for any contributing factors for acute lung injury.

Evaluation

Evaluation should be targeted at ruling out other associated etiologies causing hypoxemia in ventilated patients. Aspiration, infective pneumonia, auto PEEPing, acute coronary syndromes, venous, fat, air or amniotic fluid embolism, pneumothoraces, pleural effusion, and intra-abdominal distension are to be ruled out. A thorough clinical evaluation bedside chest X-ray, ultrasound of the chest and abdomen along with a 12 lead ECG with a bedside ECHO can throw light to rule out most of the above etiologies. Auto PEEPing could be detected by analyzing the flow-time curve.

Chest X-ray in ventilator-induced lung injury could be almost similar to ARDS patients. It will show bilateral diffuse alveolar interstitial shadows without cardiomegaly. A CT scan thorax can show bilateral heterogeneous consolidation and atelectasis with focal areas of hyperlucency consistent with alveolar distension.

Laboratory investigations include lipase levels, cardiac enzymes & blood culture, and other body secretions. A lower limb Doppler and CT pulmonary angiogram may be necessary in certain cases. Fiber-optic bronchoscopy and biopsy are rarely indicated.

Gattinoni et al., in a 2016 paper, hypothesized that ventilator-related etiology of lung injury could be converted into a single variable called mechanical power. Mechanical power can be computed from tidal volume/driving pressure, flow, PEEP, and respiratory rate. They have proposed a simple ventilator software where the mechanical power equation (hence identifying the contribution of each component causing ventilator-induced lung injury) could be easily computed and analyzed at the bedside.[28]

A 2019 review mentions that quantitative bedside measurement of dynamic elastance (E) and the interpretation of the way it varies as a function of time and PEEP could be utilized to evaluate not only the degree and nature of lung injury but also the degree of contributions each from volutrauma, and atelectrauma.[29]

Treatment / Management

The most important measure to prevent ventilator-induced lung injury is to select appropriate ventilatory settings that prevent overdistension of alveoli, causing volutrauma and biotrauma, and atelectrauma.

The concept of “baby lung’ in ARDS represents the relatively small areas of aerated normal lung (which is just the size of a baby’s lung), which needs to be protected from injury during mechanical ventilation. Since most of the remaining alveoli are non-aerated and collapsed, delivery of a large tidal volume could overinflate the baby lung areas inciting lung injury. The baby lung is not a fixed anatomic structure since redistribution of dependent atelectasis occurs in prone positioning. The ideal tidal volume might have been the tidal volume required to ventilate the baby's lung, which has been evaluated only in physiologic studies at this stage.

In ARDS patients, a low tidal volume strategy of 6 ml /Kg of predicted body weight(PBW) has been shown to prevent overdistension of the alveoli and improved mortality when compared with a higher tidal volume (i.e., 12 ml/Kg of predicted body weight) as shown in the ARDS Nett trial.[3] In non-ARDS patients, a meta-analysis of 15 small randomized control trials and 5 large observational studies concluded that a tidal volume of 6-8 ml/Kg of predicted body weight is associated with improved survival.[30] Patients with mild to moderate ARDS who are on a trial of non-invasive ventilation could generate high efforts with large tidal volumes, which have the potential to harm the lung. It is better to abort NIV in such settings and consider early intubation.[4](A1)

PEEP is an important aspect of the ARDS ventilatory strategy, offering protection from atelectrauma apart from improving alveolar recruitment & oxygenation. PEEP needs to be carefully titrated since an inappropriately high PEEP can cause overdistension injury, and a lower PEEP could be insufficient to stabilize and keep the alveoli open. The PEEP is most commonly selected at the bedside for a given FiO2 based on the PEEP selection criteria adopted at the landmark ARDS Nett trial. Optimal PEEP titration based on pressure-volume curve analysis, transpulmonary pressure measurements, CT, and ultrasound pictures have been tried in various studies though proved clinical benefits with such strategies are lacking. Closed suction catheters should be preferred in mechanically ventilated ARDS patients to avoid abrupt disconnection and PEEP derecruitment, which could cause hypoxemia as well as lung deflation injury.

Recruitment maneuvers should reduce ventilator-associated injury in theory.[31] However, due to concerns regarding complications (e.g., hemodynamic compromise, pneumothorax) and uncertainty regarding clinical benefits, they are not applied widely. High-frequency oscillatory ventilation(HFOV) is theoretically promising in providing a low tidal volume(even lower than dead space and high frequency). However, the landmark OSCILLATE and OSCAR trials failed to provide any clinical superiority in ARDS patients.[32][33](A1)

The use of neuromuscular agent has been known to reduce the cytokine levels in previous studies.[18] The ACURASYS study on 340 patients published in 2010 showed approximately 10% morality benefit at 28 days and 90 days in ARDS patients who received a neuromuscular agent for 48 hours.[16] The morality benefit attributed in the cisatracurium arm is likely due to decreased multiorgan dysfunction due to decreased biotrauma resulting from decreased effort induced lung injury.(A1)

Prone position ventilation has been known to increase the homogeneity of ventilation in animal studies, thus protecting from lung injury.[34][35] A 2010 metanalysis of seven trials involving 1724 patients showed a reduction in absolute mortality by 10 % in severely hypoxemic patients with ARDS with a PaO2: FiO2 ratio < 100. The landmark PROSEVA trial on 466 patients with severe ARDS with a PaO2: FiO2 ratio < 150 also showed a significant 28-day mortality difference of 16.8 % compared to patients who were not prone.[36](A1)

Partial or total extracorporeal support (ECMO/ECCO2-R) have been conceptually promising in the prevention of ventilator related lung injury. However, data supporting clinical benefits prompting the routine initiation of extracorporeal support are scanty at this stage.

Anti-inflammatory strategies and the use of mesenchymal stem cells have been employed in animal studies to prevent the consequences of ventilator-induced lung injury.[37][38] However, their clinical utility in humans is yet to be proven. A 2017 Chinese study found that both ketamine and propofol could increase the pulmonary function index in patients with a ventilatory induced injury. Ketamine was found to be superior to propofol as an anti-inflammatory agent in reducing IL-1β, Caspase-1, and NF-κB.[39](B3)

Differential Diagnosis

Ventilator-induced lung injury is a clinical diagnosis made by ruling out common causes of respiratory deterioration on a ventilator and common etiologies of ARDS. New-onset pneumonia, cardiogenic pulmonary edema, endobronchial intubation, lung collapse, pneumothorax, pulmonary embolism (thrombotic a well as non-thrombotic), auto PEEPing, pleural effusion, and abdominal distension could give rise to respiratory deterioration. Sepsis, aspiration, infectious pneumonia, severe trauma and/or fractures, pulmonary contusion, burns, inhalational injuries(e.g., smoke, gases, near drowning), and transfusion-associated lung injury need to be ruled out. ARDS could also occur after post cardiopulmonary bye pass and major surgeries, pancreatitis, and in patients following haematopoetic stem cell transplant.

Prognosis

There is potential for increased morbidity and mortality in patients with ventilator-induced lung injury. The multiorgan failure secondary to biotrauma might be a major contributor to mortality, though higher tidal volumes per se are also proved to increase the mortality. There is a theoretical potential for long term respiratory disability, recurrent pulmonary infections, and cor-pulmonale in the survivors with extensively injured lungs.

Complications

Ventilator-induced lung injury is associated with increased morbidity and mortality. Complications of ventilatory induced lung injury include pulmonary edema, barotrauma, worsening hypoxemia resulting in prolongation of mechanical ventilation. Though higher tidal volumes are proved to increase mortality, multiorgan failure resulting from biotrauma is a major contributor. If extensively inflamed & injured areas get fibrosed, there is a potential for long-term respiratory disability and cor-pulmonale in survivors. Recurrent infections are also common in fibrosed pulmonary tissue.