Chapter 3: Ventilator-Induced Lung Injury

Mechanical ventilation is lifesaving; it improves gas exchange, alters pulmonary mechanics, and decreases the work of the cardiopulmonary system. In spite of these beneficial effects, there are numerous potential side effects associated with mechanical ventilation, including

• Increased shunting and dead-space.

• Decreased cardiac output and renal blood flow.

• Increased risk of nosocomial pneumonia.

• Increased intracranial pressure.

But the concern that has received increasing attention over the past 20 years is ventilator-induced lung injury (VILI). It has become clear that the inappropriate application of mechanical ventilation can induce injury (Table 3-1) similar to acute respiratory distress syndrome (ARDS). In addition, inappropriate application of the mechanical ventilator has been implicated in the induction or extension of multiple system organ failure.

Historically, the lung injury most associated with mechanical ventilation was barotrauma. Disruption of the alveolar capillary membrane allows air to dissect along facial planes accumulating within the pleural space or other compartments, or the development of subcutaneous emphysema. It is reasonable to assume that the higher the ventilating pressure, the greater the likelihood of barotrauma. Early reports on ARDS and asthma where unlimited peak airway pressure was applied resulted in a higher incidence of barotrauma than more recent case series where high pressure and overdistention of the lungs was avoided. No clear, specific relationship between applied pressure and barotrauma is available. However, many clinicians agree that barotrauma occurs in the lungs ventilated with high alveolar pressures and large tidal volumes (V_T). The specific volume and pressure required to develop barotrauma is likely patient-specific.

High concentrations of inhaled oxygen result in the formation of oxygen-free radicals (eg, superoxide, hydrogen peroxide, hydroxyl ion). These free radicals can cause ultrastructural changes in the lung similar to acute lung injury. In animal models, inhalation of 100% oxygen causes death within 24 to 48 hours. Human volunteers breathing 100% oxygen develop inflammatory airway changes and bronchitis within 24 hours. There are also laboratory data to suggest that former exposure to bacterial endotoxin, inflammatory mediators, and sublethal levels of oxygen (≤ 85%) protect the lung from further injury when inspiring a high Fio₂.

Concern for oxygen toxicity should never prevent the use of a high Fio₂ in a patient who is hypoxemic. An Fio₂ of 1.0 should be administered whenever there is uncertainty about the Pao₂ and pulse oximetry is not available. However, Fio₂ should be lowered to the level resulting in a Pao₂ of 55 to 80 mm Hg (SpO₂ 88%-95%). The target Fio₂ is less than or equal to 0.60. Few clinicians feel the concern of oxygen toxicity is greater than the concern of tissue hypoxia. An exception is the patient treated with bleomycin. The combination of bleomycin and oxygen results in marked injury to the lungs. In this setting, the lowest Fio₂ should be used, tolerating a Pao₂ as low as 50 mm Hg (Spo₂ 85%-88%).

Primary determinants of lung injury are stress and strain. Stress is defined as the internal counterforce per unit area that balances an external load on a structure or the pressure gradient across a structure. Strain is the resulting deformation of the system as a result of the external load or the change in size or shape of the structure. From a pulmonary perspective, stress is the alveolar distending pressure (alveolar pressure minus pleural pressure) and strain is the ratio of volume change (V_T plus volume increase caused by positive end-expiratory pressure [PEEP]) to functional residual capacity during the application of the stress.

Lung strain of more than 2 is considered injurious to the lungs. Stress and strain are related by the specific lung elastance of 13.5 cm H₂O/mL. Thus, lung stress 13.5 cm H₂O/mL times lung strain. From a bedside perspective the surrogates for stress and strain are plateau pressure (Pplat) and V_T. Thus the highest alveolar distending pressure that theoretically prevents disruption of the alveolar capillary membrane is 27 cm H₂O; thus the Pplat limit of 30 cm H₂O is in close agreement with this value when the chest wall compliance is normal. In passively ventilated critically ill patients, the alveolar distending pressure never exceeds the (Pplat) and in most settings the (Pplat) is at least a few cm H₂O greater than the alveolar distending pressure.

Volutrauma refers to lung parenchymal damage caused by mechanical ventilation that is similar to ARDS (see Figure 3-1). Volutrauma is VILI manifested by an increase in the permeability of the alveolar capillary membrane, the development of pulmonary edema, the accumulation of neutrophils and proteins, the disruption of surfactant production, the development of hyaline membranes, and a decrease in compliance of the respiratory system (Table 3-2). The term volutrauma is used because the induced injury is the result of alveolar overdistention. Clinically, volume is used as a surrogate for pressure as it is impossible to measure local overdistention at the bedside. The pressure that has been used as a surrogate for local overdistention is Pplat. A Pplat more than 30 cm H₂O increases the likelihood of VILI; Pplat should be kept as low as possible.

Chest Wall Compliance

Because alveolar distention is determined by the difference between alveolar and pleural pressure, the chest wall has a role in determining the extent of overdistention. When the chest wall is stiff (low compliance), a high Pplat may be associated with less risk of overdistention. That is, a stiff chest wall (eg, abdominal distention, massive fluid resuscitation, chest wall deformity, chest wall burns, morbid obesity) protects the lungs from VILI.

Active Breathing Efforts

The alveolar distending pressure can change markedly on a breath-by-breath basis in a spontaneous breathing patient. This most commonly occurs in pressure-targeted ventilation with high inspiratory efforts by the patient. When the airway pressure is constant and the patient forcefully inhales, the alveolar distending pressure may exceed what is expected by the airway pressure setting. For example, if the pressure control is 25 cm H₂O and the patient's effort decreases the pleural pressure to –10 cm H₂O, alveolar distending pressure is 35 cm H₂O, 10 cm H₂O greater than expected with an airway pressure setting of 25 cm H₂O. During pressure-targeted ventilation, the contribution of patient's effort to alveolar distending pressure must be appreciated.

Preexisting Injury

Preexisting injury increases the likelihood of VILI. This is called the two-hit process of lung injury. Previous injury predisposes the lungs to a greater likelihood of ventilator-induced injury. The use of lung-protective ventilatory strategies is thus necessary for all patients. The risk of developing ARDS is reduced if lung-protective ventilation strategies are implemented from the onset of mechanical ventilation (eg, volume and pressure limitation).

Another mechanism for the development of VILI is the repetitive recruitment and derecruitment (opening and closing) of unstable lung units (atelectrauma) during each ventilatory cycle. The junction between an open and a closed alveolus serves as a stress riser (Figure 3-2). It has been estimated that more than 100 cm H₂O of stress is created when 30 cm H₂O alveolar distending pressure is applied adjacent to a collapsed alveolus. As illustrated in Figure 3-3, when PEEP is applied, the effect of a given distending pressure is attenuated and the extent of VILI reduced. The method used to determine the amount of PEEP in ARDS that prevents de-recruitment in a given alveolar unit is controversial. However, the best PEEP level prevents de-recruitment at end exhalation.

Overdistending V_T and repetitive opening and closing of unstable alveoli result in the activation of inflammatory mediators within the lung. Numerous proinflammatory (cytokines, chemokines) and anti-inflammatory mediators are activated by injurious ventilatory patterns. These mediators increase edema formation, neutrophil migration, and relaxation of vascular smooth muscle. Injurious ventilatory patterns increase systemic inflammatory-mediator response compared to a lung protective ventilation strategy.

Bacteria instilled into the lungs of otherwise healthy animals produce bacteremia when inappropriate ventilatory patterns are employed. Translocation of bacteria is minimized if lung-protective ventilatory strategies are used.

Preliminary and controversial animal data suggest a role for vascular volume, ventilator rate, and body temperature on VILI. Higher vascular infusion volumes, rapid respiratory rates, and high body temperature potentially cause greater injury. These potentially injurious factors are probably most important in the presence of nonprotective ventilatory strategies.

Injurious ventilatory patterns not only cause VILI but may also cause or extend multiple organ dysfunction syndrome (MODS). Disruption of the alveolar-capillary membrane allows leakage of pulmonary inflammatory mediators into the bloodstream, allowing downstream organ failures (Figure 3-4).