Ventilator-Associated Lung Injury

Acute lung injury and ARDS are recognized as affecting the lungs heterogeneously. The distribution of edema fluid, ventilated versus flooded alveoli, and consequently the matching of ventilation and perfusion vary among gas exchange units. Moreover, it is recognized that the lung is capable of a brisk inflammatory response when injured or when ventilated mechanically, which may have local or systemic manifestations. The ARDSnet trial demonstrated improved outcomes from ALI/ARDS after ventilation with lower V_T and minute ventilation (V_E), resulting in lower airway pressures, less overdistension of recruitable alveoli, less shear stress on lung tissue, and lower mortality despite the paradox that most patients with ALI/ARDS do not die from an inability to oxygenate or ventilate. Rather, most such patients die in association with the multiple organ dysfunction syndrome—which has been linked closely with a rampant systemic inflammatory response. If less ventilation is better, it was hypothesized that more ventilation may be injurious or indeed provocative to the lung—leading to the concept of VILI.

Ventilator-induced lung injury occurs from excessive mechanical stress to the lung, either from excessive V_T or excessive airway pressure. Mechanical ventilation induces a pulmonary and systemic cytokine response, which can be minimized by limiting overdistension and phasic recruitment/derecruitment of lung. A substantial body of experimental and clinical data demonstrates that the mechanism of VILI is the proinflammatory response in the lung and the periphery, and that the response and injury are attenuated by lung-protective ventilation strategies. New modes of ventilation and protective ventilation are designed to minimize the deleterious effects of mechanical ventilation, which is a fundamental aspect of critical care management.

Proportional Assist Ventilation

Proportional assist ventilation (PAV) is a form of synchronized partial ventilatory assistance that augments the flow of gas to the patient in response to patient-generated effort. The ventilator augments the patient’s inspiratory effort without using preselected target volume or pressure. The purpose of PAV is to allow the patient to achieve a pattern of ventilation and breathing that is adequate and comfortable. The patient initiates and determines the depth and frequency of the breaths independently of the ventilator. Advantages to this type of ventilator support include greater comfort; reduction of peak airway pressure required to deliver the V_T; less likelihood of overventilation and overdistension of alveoli; preservation and enhancement of the patient’s own reflex, behavioral, and homeostatic control mechanisms; and improved efficiency of negative-pressure ventilation.

Effective use of PAV requires an understanding of the individual patient’s ventilatory mechanics. This entails measuring the patient’s airway resistance, compliance, and intrinsic positive end-expiratory pressure (auto-PEEP) to determine the ventilatory load and assistance the patient requires. Younes et al. proposed an innovative method for the noninvasive determination of passive elasticity during PAV. Once the patient’s elastance and resistance are determined, the PAV parameters are set followed by PEEP, adjusting the peak pressure limit to 30 cm H₂O, adjusting volume assist to 8% of elastance measured on PAV, and finally observing the patient’s ventilation, breathing pattern, and peak airway pressure. As a new ventilatory method, PAV can conceivably improve patient-ventilator interaction. Its true usefulness remains to be measured, and clinical usage is uncommon.

Pressure-Controlled Ventilation

Pressure-controlled ventilation (PCV) is pressure-limited timecycled breathing that is completely controlled by the ventilator, with no participation by the patient. Inspiratory airway pressure increases early in the respiratory cycle, and is maintained at that specified pressure throughout the remainder of the delivery phase. The major benefit of PCV relates to the inspiratory flow pattern and its benefit in gas delivery for patients with suppressed respiratory efforts. In pressure-cycled breathing, the inspiratory flow decreases exponentially during lung inflation in order to keep the airway pressure at the preselected value. Thus, this type of flow pattern can improve gas exchange. The primary disadvantage of PCV is the tendency for inflation volumes to vary with changes in the mechanical properties of the lungs. There is a proportional relationship between the lung inflation and the peak inflation pressure. When a constant peak inflation pressure is reached, the inflation volume decreases as the airway resistance increases or lung compliance decreases. The result is fluctuation in inflation volumes due to reliance on a specific pressure target.

Inverse-Ratio Ventilation

Inverse-ratio ventilation (IRV) is a combination of PCV (hence, PC-IRV) with a prolonged inspiratory time (I). One way to increase I is to decrease the inspiratory flow rate, such as increasing the I:expiratory time (E) ratio from the usual 1:4 to 2:1 (up to 4:1). Benefits of IRV include improved oxygenation and the prevention of alveolar collapse as confirmed in cases of ARDS and neonatal respiratory distress syndrome. On occasion, the early use of PC-IRV can facilitate tapering of high fractions of inspired oxygen (F_IO₂) and decrease high PEEP and peak inspiratory pressures (PIPs). The downside to PC-IRV is that is can lead to stacking of breaths (auto-PEEP), with high airway pressures, hyperinflation and barotrauma, CO₂ retention, and metabolic acidosis. Another adverse effect observed with IRV is the potential to cause decreased cardiac output due to auto-PEEP because increased transthoracic pressure decreases venous return.

Mandatory Minute Ventilation

Mandatory minute ventilation (MMV) is a mode of mechanical ventilation in which the minimum level of V_E needed by the patient is provided. If the patient’s spontaneous ventilation is insufficient to meet the predetermined V_E, the ventilator provides the difference. Conversely, if the patient’s spontaneous breathing exceeds the target V_E no ventilator support is provided. This mode is one of the so-called “closed-loop” ventilation modes (Table 1) because the ventilator varies its parameters in response to the patient’s own intrinsic ventilatory requirements. The major advantage of MMV is the capability to vary ventilatory support according to the response of the patient. This mode of mechanical ventilation is best suited for patients with severe neuromuscular disease or drug overdose, or patients heavily sedated. One of the main disadvantages with MMV is that alveolar ventilation may not be matched equally with exhaled V_E, thus diminishing closing volumes and leading to atelectasis. None of the closed-loop modes, MMV included, has been tested sufficiently on critically ill patients to recommend widespread incorporation into practice.

Airway Pressure Release Ventilation

Airway pressure release ventilation (APRV) (Table 2) has been used as an alternative mode of mechanical ventilation in patients with acute respiratory failure. APRV, which has been available in some ventilator models since the mid-1990s, allows for the unloading during exhalation of any positive pressure provided during inhalation in order to facilitate the egress of the tidal breath. Release of airway pressure from an elevated baseline simulates exhalation. Technically, APRV is time-triggered, pressure-limited, time-cycled mechanical ventilation. Conceptualizing APRV as continuous positive airway pressure (CPAP) with regular, brief, intermittent releases of airway pressure may facilitate understanding. It can augment alveolar ventilation in the patient breathing spontaneously, or provide full support to the apneic patient.

Advantages of APRV include lower peak airway pressure, lower intrathoracic pressure, lower V_E, minimal effect on cardiac output, and improved matching of ventilation and perfusion. The mode may facilitate spontaneous breathing by the patient. Sedation requirements may be decreased, and neuromuscular blockade should be avoided altogether. Patient-ventilator dyssynchrony is believed not to develop. Disadvantages of APRV include pressure control of ventilation, increased effects of airway and circuit resistance on ventilation, decreased transpulmonary pressure, and potential interference with spontaneous ventilation. Facilitated exhalation may make APRV beneficial in patients with bronchospasm or small-airways disease. This mode of ventilation can be used as a weaning mode. Although increasingly popular, the advantage of APRV over other modes of ventilation is unproved.

The terminology of APRV differs somewhat from other modes of mechanical ventilation, and has yet to be standardized. Four important terms include pressure high (P_high), pressure low (P_low), time high (T_high), and time low (T_low). The P_high term describes the baseline airway pressure (the higher of the two pressures), alternatively called CPAP, inflating pressure, or the P1 pressure. The P_low term describes the airway pressure resulting from the release of pressure (alternatively called PEEP, release pressure, or the P2 pressure). The T_high time refers to the time during which P_high is maintained (T1), whereas T_low refers to the duration of time when airway pressure is released (T2). Mean airway pressure can be calculated from the following equation:

Application of APRV to the patient must be individualized, as standard approaches have yet to emerge. Initial settings are deduced partly from the result of conventional mechanical ventilation, which should be attempted initially for most patients. The plateau airway pressure (P_plat) from conventional ventilation (if not higher than 35 cm H₂O) is converted to P_high, aiming for a V_E of 2–3 l/min (lower than with conventional ventilation). The P_low pressure is set initially at 0 cm H₂O. The setting for T_high is a minimum of 4 seconds, and T_low is set at approximately 0.8 seconds (0.5–1.0 second). Spontaneous breating is permitted. At these settings, mean airway pressure is 29 cm H₂O. Rarely, a higher P_high (40–45 cm H₂O) is needed for patients with low compliance (e.g., morbid obesity, abdominal distention). For all patients, T_high is lengthened progressively to 12–15 seconds (usually in 1- to 2-second increments as lung mechanics improve). Longer T_high prevents the cyclical opening and closing of small airways that is believed to be a cause of VILI. The T_low parameter is optimized when expiratory flow decreases to 25%–50% of peak expiratory flow.

Clinical improvement may not be immediate after transition to APRV (as is the case with IRV). Clinical studies have shown that maximum clinical improvement may not occur until 8–16 hours after the transition. After improvement, weaning from APRV is guided by general principles of weaning. Weaning from APRV is accomplished primarily by manipulation of P_high and T_high. High pressure is decreased in increments of 2–3 cm H₂O down to about 15 cm H₂O, and T_high is lengthened progressively to 12–15 seconds (usually in 1- to 2-second increments). Minute ventilation must be monitored carefully for signs of hypoventilation during the transition. The goal is to switch the patient to pure CPAP of 6–12 cm H₂O, at which point the patient may be extubated—all conditions permitting.

Some confusion arises with similar modes of ventilation. BiPAP differs from APRV only in the timing of T_high and T_low. The latter is longer in BiPAP. Intermittent mandatory pressure release ventilation (IMPRV)—similar to APRV and rarely used—synchronizes the release of pressure with the patient’s spontaneous effort. In IMPRV, all spontaneous breaths are pressure-supported ventilation (PSV) to reduce the work of breathing. However, the rationale for IMPRV is considered dubious by some because dysynchrony appears not to occur with APRV.

High-Frequency Ventilation

High-frequency ventilation (HFV) is a ventilatory strategy that has been used with success for respiratory failure in neonates and children. This modality utilizes limited high mean airway pressures at low V_T (often smaller than anatomic dead space) and a highfrequency respiratory rate (2.5–30 Hz) to achieve adequate ventilation while at the same time preventing alveolar overdistension. This mode is conceptually attractive because it achieves many of the goals of lung-protective ventilation. Used most often to support complex thoracic surgical procedures (e.g., one-lung or split-lung ventilation) or to facilitate fiberoptic bronchoscopy or healing of bronchopleural fistula, HFV has been shown to be safe and effective for ventilation of small numbers of adult patients with severe ARDS who have failed conventional ventilation. However, large-scale trials are needed.

Permissive Hypercapnia

Permissive hypercapnia is an adjunctive protective ventilatory strategy. Permissive hypercapnia defines a ventilatory strategy for acute respiratory failure in which the lungs are ventilated with a low V_T, permitting PaCO₂ levels to increase. Permissive hypercapnia aims to avoid hyperinflation-induced lung trauma, as described initially by limiting the plateau airway pressure (as a surrogate of static alveolar pressure) to approximately 30–35 cm H₂O while allowing PaCO₂ to increase absent any contraindications (such as increased intracranial pressure).

Hickling et al. introduced the concept of permissive hypercapnia, reporting that reducing the peak inspiratory airway pressure to a maximum of 20–30 cm H₂O while allowing PaCO₂ to increase resulted in a decreased mortality rate of 16% for 50 consecutive patients with ARDS. Amato et al. reported similar results in the first controlled study on the use of permissive hypercapnia in patients with ARDS. If the ARDSnet low V_T protocol is adhered to as a ventilatory strategy, permissive hypercapnia may provide further improvement in outcomes in patients with ALI. In addition, experimental evidence suggests that VILI may cause release of inflammatory mediators—increasing the likelihood of multiple-organ dysfunction syndrome.

Permissive hypercapnia has not been widely implemented to near its physiologic limits (e.g., PaCO₂ up to 80 mm Hg, arterial pH down to 7.20) because of a relative paucity of controlled studies showing clear benefit from the application of this strategy in ARDS, and because of concerns over physiologic consequences of the associated hypercapnia on the central nervous, cardiovascular, and renal systems. The absolute level of PaCO₂ and the permissible degree of acidosis is debated, as is the concern of alveolar derecruitment and possible worsening of ventilation-perfusion mismatching. The PaCO₂ is directly proportional to the rate of CO₂ production by oxidative metabolism (VCO₂) and inversely proportional to the rate of CO₂ elimination by alveolar ventilation (V_A). An equation that illustrates the relationship between each source is PaCO₂ = k (VCO₂/V_A). The three major sources of hypercapnia include increased CO₂ production, hypoventilation, and increased dead space ventilation. Metabolic CO₂ production is an essential factor in promoting hypercapnia only in patients with underlying lung disease.

Contraindications and adverse effects of permissive hypercapnia include cerebral edema or high intracranial pressure, convulsions, depressed cardiac function, arrhythmias, increased pulmonary vascular resistance, tachypnea, increased work of breathing, dyspnea, respiratory distress, headache, sweating, and biochemical disturbances related to acidosis.

Liquid Ventilation

Use of fluids to facilitate gas exchange has been under scrutiny for many years. Due to the fact that mechanical ventilation with gas may cause barotrauma, exacerbate ALI (causing structural damage to the lungs), and induce the release of inflammatory mediators, alternative means of supporting pulmonary gas exchange while preserving lung structure and function are desirable. Much research has focused on the use of perfluorocarbon (PFC) liquids to deliver biologic agents to diseased lungs, generally by one of two modalities. The first is total liquid ventilation, in which the lungs are filled with PFC to a volume equivalent to functional residual capacity (FRC), then ventilating the PFC-filled lung with oxygen. Total liquid ventilation has been largely abandoned owing to its logistical complexity.

The second technique of liquid ventilation (partial liquid ventilation, PLV) involves intratracheal administration of PFC in a volume equivalent to FRC, followed by standard gas mechanical ventilation of the PFC-filled lung. In infants with biochemically immature lungs, liquid ventilation may minimize the effect of barotrauma. There is evidence that liquid ventilation may eliminate surface-active forces, providing effective gas exchange with minimal risk for barotrauma. Airway toilet may be improved as debris floats upward to the meniscus, where it can be removed. Notably, the debris can be so voluminous as to cause airway obstruction. Thus, pulmonary toilet must be diligent. However, PFC is volatile and requires frequent “topping off” to maintain sufficient volume. Moreover, PFC is radioopaque and creates a bilateral “white-out” on chest x-ray that makes radiographic interpretation impossible.

Perfluorocarbon liquids may have anti-inflammatory properties in the alveolar space. The anti-inflammatory effects of liquid ventilation in ALI are from inhibition of neutrophil and macrophage function, and the dilution of inflammatory debris in the airways. PFC liquids are currently used clinically in a number of ways, such as intravascular PFC emulsions for volume expansion, improving oxygen-carrying capacity, angiography, and intracavitary PFC liquid for image contrast enhancement and vitreous fluid replacement. However, no agent for liquid ventilation in the United States has been approved for clinical use. Several factors complicated the phase 2 and 3 clinical studies of PLV. In a prospective, randomized, controlled pilot study of 90 adults with ALI/ARDS, with PaO₂:F_IO₂ greater than 60 but less than 300, PLV did not affect ventilator-free days (the primary endpoint), mortality, or any other clinical factors. Criticisms of this study included slow recruitment (entry criteria were relaxed after 45 patients), lack of a weaning protocol, and a disproportionate number of patients over age 55 in the PLV group. However, a post hoc analysis found significantly more rapid discontinuation of mechanical ventilation and a trend toward more ventilator-free days in the PLV group among younger patients. The authors suggested further evaluation, particularly in certain well-defined (especially younger) patients.

In a second trial, the hypothesis was tested that PLV would increase the number of ventilator-free days compared to conventional mechanical ventilation—and would decrease 28-day all-cause mortality compared to conventional ventilation. Adult patients with ALI who had been on mechanical ventilation for less than 120 hours (with a PaO₂:F_IO₂ below 200, F_IO₂ above 0.5, and PEEP above 5 cm H₂O) were enrolled. There was no improvement in 28-day all-cause mortality. The mean number of ventilator-free days was reduced significantly. The aggregate results have caused a substantial loss of enthusiasm for the clinical use of PLV.

Surfactant Administration

Acute lung injury is characterized by pulmonary and endothelial inflammation, which causes pulmonary edema, destruction and impaired synthesis of surfactant with atelectasis and reduced pulmonary compliance, hypoxemia from ventilation/perfusion mismatching, and subsequent pulmonary hypertension and fibrosis. Studies have examined the intratracheal administration of surfactant as a means of improving both oxygenation and ventilation. Although surfactant is indisputably effective for neonatal respiratory distress syndrome, clinical trials of adult patients have failed to demonstrate a benefit for any of several variations of natural or synthetic surfactant or components thereof.

Inhaled Nitric Oxide

Inhaled nitric oxide (NO) is a selective pulmonary vasodilator that acts on the alveolar endothelium to produce regional vasodilation in well-ventilated lung units where it is distributed. NO at a dose of 40 parts per million (PPM) has been demonstrated to improve ventilation-perfusion mismatching, hypoxemia, and pulmonary hypertension in patients who have ALI/ARDS. In contrast, systemic vasodilators may actually cause pulmonary vasodilation in nonventilated lung, thereby abrogating hypoxic vasoconstriction and leading to hypoxemia and exacerbated ventilation-perfusion inequality. NO is also a bronchodilator and has anti-inflammatory properties that have been helpful in lung transplant patients. When inhaled, it diffuses into the blood stream—where it is metabolized rapidly and excreted via the urine, minimizing systemic effects. Three randomized controlled trials suggest that NO may improve oxygenation for up to 72 hours, but neither survival nor a shorter duration of mechanical ventilation was observed. In addition, rebound pulmonary hypertension has been reported with cessation of therapy. Therefore, inhaled NO is not recommended for therapy of ALI/ARDS even as rescue therapy.

Independent Lung Ventilation

Acute lung injury is recognized to be heterogeneous with the lung, but also may be heterogeneous between lungs (e.g., massive aspiration of gastric content confined to one lung). The left lung is smaller than the right in human beings. Moreover, when lung injury is asymmetric, differences in compliance exist between the lungs. Consequently, conventional mechanical ventilation delivers a larger V_T to the more compliant lung—which may cause overdistension and VILI. In addition, overdistension disrupts blood flow through alveolar vessels—diverting flow to the underventilated lung and worsening ventilation-perfusion mismatch. Independent lung ventilation (ILV) has been described to ventilate the more diseased lung while avoiding overdistension of the more normal lung.

Using a dual-lumen endotracheal tube with a bifurcated tip such that each main-stem bronchus can be intubated separately, gas can be delivered using two ventilators dedicated one to each lung. Typically, the V_T is set equal for both lungs. However, this may be an irrational approach both anatomically and physiologically because higher airway pressure may be anticipated in the more injured lung. Alternatively, ventilator settings may be adjusted to produce equal P_plat in both lungs. The ventilators may be managed independently, and the lungs may be monitored independently by pressure measurements, compliance calculations, and capnography.

Independent lung ventilation is seldom used anymore. Modern mechanical ventilators and lung-protective ventilation strategies have obviated many of the difficulties that made ILV attractive when described initially in the 1970s and early 1980s. The dual-lumen endotracheal tube is challenging to position and keep positioned. Patients require heavy sedation and often neuromuscular blockade. Combinations of ILV with newer modes of ventilation have not been described.

Extracorporeal Membrane Oxygenation

Extracorporeal membrane oxygenation (ECMO) provides oxygenation of blood and removal of CO₂ via an extracorporeal circuit. ECMO consists of the application of intermediate-term cardiopulmonary bypass for the treatment of potentially reversible cardiac or pulmonary failure for patients of any age. If successful, sufficient gas exchange permits reduced support of positive-pressure ventilation—thereby lowering the incidence of barotrauma-induced lung injury associated with mechanical ventilation. ECMO thus provides relative “lung rest” to the acutely injured lungs and facilitates recovery.

Conventional mechanical ventilation is the mainstay of treatment for severe respiratory failure associated with trauma. However, when extensive lung injury is present, conventional ventilation may not be sufficient to prevent hypoxia and may exacerbate pulmonary damage by barotrauma. The logic of ECMO for severe pulmonary failure is that borderline patients may be saved if their lungs are allowed to rest and heal rather than endure the morbidity of the high-level ventilator support necessary to achieve adequate gas exchange. ECMO has been used successfully to manage critically ill adult trauma patients and offers an additional treatment modality. However, ECMO is not available in all centers.

The first adult managed successfully with prolonged ECMO was a trauma patient cared for by Hill et al., reported in 1972. Subsequently, a National Institutes of Health (NIH)-sponsored randomized multi-institutional trial (reported in 1979) failed to demonstrate improved survival in adults managed with ECMO. In contrast, Cordell-Smith et al. found that a high proportion of trauma patients treated with ECMO for severe lung injury survived. This outcome appears to compare favorably with conventional ventilation techniques and may have a role in patients who develop acute severe respiratory failure associated with trauma. ECMO may be considered for support of severe ARDS affecting adult patients when all other treatment options have failed. Injury of the thoracic aorta, even if contained, is considered a contraindication to ECMO use. The paradox is that ECMO appears to achieve the best outcomes when utilized relatively early. The mortality from respiratory failure increases the longer a patient is mechanically ventilated before initiation of ECMO.

Over the years, ECMO therapy has undergone substantial changes in indications, technique, and materials. Technical progress has been made in the pumps, oxygenators, and coating of artificial surfaces, leading to greater biocompatibility and a lower rate of procedure-related complications. The potential of new inline pumps in combination with a decreasing incidence of procedure-related complications may lead to a reevaluation of the role of ECMO in the therapy of ARDS. New techniques for insertion of intravascular oxygenators (IVOX) and extracorporeal CO₂ removal (ECCO₂R) devices highlight some of the technical advances being made. Unfortunately, the technical advancements have yet to translate to improved survival in clinical trials.

In a prospective controlled trial using ECMO in patients with ARDS and severe ALI, Zapol et al. showed that refractory hypoxia tempered enthusiasm for ECMO use by demonstrating no survival benefit. However, the control group and treatment group both had very high mortality rates (control 91% vs. ECMO 90%, vs. ∼30% in the ARDSnet trial), and the study did not use a lung protection ventilation strategy. The only randomized, prospective, controlled study utilizing ECCO₂R similarly did not demonstrate any survival benefit from ECCO₂R use. The IVOX device attempts to accomplish the same objectives as ECMO through placement of a membrane within a major vein such as the vena cava. Its intended patient population is intensive care unit (ICU) patients with severe potentially reversible acute respiratory failure. Initially examined in its current form in 1982, the main function of IVOX is to provide transport for oxygen and CO₂ across its microporous hollow capillaries. Phase I and phase II IVOX clinical trial observations note that IVOX managed limited but statistically significant amounts of oxygen and CO₂. Further trials are pending.

Prone Positioning

Prone positioning (PP)—in which the patient is positioned prone, most commonly using a specialty bed—improves oxygenation by decreasing ventilation-perfusion mismatching. Ventilated lung segments tend to be in nondependent portions of lung, whereas perfused lung segments (and higher pulmonary vascular pressures and hence lung edema) tend to be in dependent portions of lung. Positive-pressure ventilation exacerbates the mismatching. PP reverses the dependency of the lung, and newly dependent wellventilated lung segments are well perfused for several hours until the effects of gravity and positive airway pressures restore the previous conditions over a period of several hours. Drainage of secretions may also be facilitated in the prone position. The large multicenter randomized controlled trial by Gattinoni et al. showed significant improvement in PaO₂:F_IO₂ and in 10-day mortality, but this modest advantage did not persist beyond ICU discharge. Complications with PP include pressure sores, need for increased sedation, facial edema, and difficulty maintaining airway patency or restoring it when the patient is inverted. Recently, facilitated PP using a specialized bed has shown promise. Thus, this intervention may yet prove useful for management of severely hypoxic patients with ARDS.