Understanding the mechanisms of ventilator-induced lung injury using animal models

Although mechanical ventilation provides benefits in many clinical situations, it can cause pulmonary structural damage [1], known as ventilator-induced lung injury (VILI), and hemodynamic instability [2]. This is in line with a series of potential harmful effects of mechanical ventilation, including increases in inflammatory infiltration and vascular permeability, hyaline membrane formation, and pulmonary edema. Death may occur during mechanical ventilation even with satisfactory blood gas exchange [3, 4].

The four main injury mechanisms associated with VILI are as follows: barotrauma/volutrauma caused by overstretching the lung tissues; atelectrauma, caused by repeated opening and closing of the alveoli resulting in shear stress; and biotrauma, the resulting biological response to tissue damage, which leads to lung and multi-organ failure [5].

Different static variables (peak, plateau, and driving pressures, positive end-expiratory pressure, and tidal volume) and dynamic variables (respiratory rate [RR], airflow amplitude, and inspiratory time fraction) can contribute to these mechanisms of VILI. Moreover, the potential for lung injury depends on tissue vulnerability, the energy applied per unit of time (mechanical power), and the duration of that exposure [6, 7]. This narrative review discusses the advantages and limitations of experimental VILI, elucidates the mechanisms underlying the pathogenesis, progression, and resolution of VILI, and analyzes the strategies that can mitigate VILI.

Advantages and limitations of experimental VILI

Experimental models allow researchers to investigate the mechanisms of VILI, which would be impossible and/or unethical in humans. Thus, different models of VILI have been developed and studied in diverse animal species in the last decades [8]. Some of the most common VILI models are summarized in Table 1. However, animal studies present some limitations that need to be considered in planning, conducting, and interpreting the results [9]. There are several physiologic and anatomic differences between humans and animals, which may influence the pulmonary response to an acute stimulus [10]. In this context, the RR is higher in mice (250–300 breaths per minute [bpm]) and rats (80–120 bpm) compared with humans (12–16 bpm). In addition, the lung structure of mice does not include bronchial arteries, and the size of the alveolus and the thickness of the alveolar-capillary membrane are smaller than those observed in rats and humans. Unlike the human lung, mice and rats have a monopodial airway branching pattern, whereas the human bronchial tree shows divisions with a dichotomic pattern (each bronchus is divided into two distal bronchi). In terms of inflammatory response, which is important during the development of VILI in animals, mice have lower rates of circulating neutrophils (10–25%) than humans (50–70%) and do not express defensins [11]. The baseline values and the names of neutrophil chemokines differ between rodents and humans, e.g., keratinocyte-derived chemokine in mice versus interleukin-8 in humans. Inter-species differences also exist between humans and pigs and/or piglets. Although the hemodynamics in humans and pigs are similar, the pulmonary vascular response to hypoxia (hypoxic vasoconstriction) is more pronounced in pigs than in humans [12]. To date, no available animal model perfectly mimics all key aspects of human VILI or acute respiratory distress syndrome (ARDS) [8, 13]; nevertheless, current models in use can help us better understand the mechanisms of VILI and develop new therapeutic approaches to mitigate lung damage. Selecting the animal model that most adequately fits the corresponding research question is of utmost importance.

Table 1 Common models of ventilator-induced lung injury

There are additional factors that should be explored further in preclinical studies, such as sex, age, and VILI resolution. Recently, sex was not associated with VILI susceptibility in mice [14]. These findings support the inclusion of both sexes in experimental studies rather than restricting the use of animals of a single sex [15]. Considering that most patients who undergo invasive mechanical ventilation are ~ 60 years old [16], the association between aging organs and mechanical ventilation should be explored further in future preclinical studies. There is insufficient evidence about pulmonary repair mechanisms in experimental VILI. The process after lung injury may involve resolution of alveolar/interstitial edema and inflammation, structural cell proliferation, and extracellular matrix organization [17]. Moreover, modulation of the redox capacity by the Nrf2-ARE pathway has been shown to increase resilience against oxidative stress during injurious mechanical ventilation [18]. In addition, therapy using a conditioned medium obtained from bone marrow and cryopreserved umbilical cord mesenchymal stem cells was able to reduce stretch-induced inflammation and cell death, thus enhancing VILI resolution [19].

Static ventilator variables associated with VILI

Peak airway pressure (Ppeak,RS), plateau airway pressure (Pplat,RS), positive end-expiratory pressure (PEEP), driving pressure (∆P,RS), and tidal volume (VT) are static ventilator variables associated with VILI (Fig. 1).

Fig. 1figure 1

Static and dynamic ventilatory variables that contribute to ventilator-induced lung injury (VILI). DP driving pressure, Ppeak peak airway pressure, PEEP positive end-expiratory pressure, Pplat plateau airway pressure, VT tidal volume

Peak airway pressure

In pressure-controlled ventilation (PCV), Ppeak,RS is the maximum pressure during inspiration and depends on the elastic and resistive components (airway, lung tissue) and equipment (endotracheal tube diameter and length) [20, 21]. PCV is usually associated with lower Ppeak,RS compared with volume-controlled ventilation (VCV) due to the different flow profiles, but this difference is less important when the option of ramp flow is used in VCV. In 1974, Webb and Tierney [22] showed that healthy rats ventilated with high Ppeak,RS (45 cmH2O) and zero PEEP presented perivascular and alveolar edema, lung overdistension, and barotrauma. On the other hand, a Ppeak,RS of 45 cmH2O and PEEP of 10 cmH2O did not result in edema. In 2017, Katira et al. [23] reproduced the classic study of Webb and Tierney to clarify these different responses, focusing on heart–lung interaction in healthy rats. They showed that high Ppeak,RS impairs right ventricular filling and pulmonary perfusion, resulting in right ventricular failure and dilation. This scenario is in line with endothelial cell injury and capillary stress failure, which may facilitate microvascular leakage of protein and water into the alveoli, yielding high permeability pulmonary edema. Thus, this preclinical study showed that increased Ppeak,RS values should be avoided due to adverse heart–lung interactions.

Plateau airway pressure

Pplat,RS is calculated during a period when airflow is stopped at end inspiration and reflects end-inspiratory alveolar pressure. Pplat,RS can be affected by changes in VT and respiratory system compliance (C,RS) but not by changes in airflow and airway resistance [24]. The effects of four levels of Pplat,RS (15, 20, 25, and 30 cmH2O) on alveolar-capillary barrier permeability to proteins were studied in a model of lung damage induced by hypertonic solution. Pplat,RS between 20 and 25 cmH2O was associated with epithelial and endothelial cell damage as well as increased permeability [25].

Because Pplat,RS can be affected by the properties of the chest wall, the chest wall component needs to be subtracted from the respiratory system, thus yielding the transpulmonary plateau pressure (Pplat,L) that is associated with the development of VILI. Limiting Pplat,RS to ≤ 28 cmH2O was found to be effective in reducing the risk of overdistension and is widely accepted.

Positive end-expiratory pressure

PEEP reflects the end-expiratory pressure remaining in the airways and, thus, the static preload of the respiratory system. The use of low PEEP levels may not be sufficient to reduce alveolar collapse and lung edema [26]. However, higher PEEP may cause lung overdistention in the more compliant areas of the lungs and hemodynamic impairment. How to best set the PEEP in experimental models of ARDS is still challenging and the following strategies have been described to date: PEEP titrated according to oxygenation, respiratory system compliance or driving pressure, transpulmonary pressure (esophageal pressure), and imaging (computed tomography scan, electrical impedance tomography) [27]. Nevertheless, there are controversies regarding the best PEEP to use in clinical ARDS; it should be set according to each patient considering lung function (arterial blood gases and mechanics), imaging findings (degree of recruitability), and phenotype (hypo- versus hyperinflammatory).

Respiratory system driving pressure

∆P,RS is defined as Pplat,RS-PEEP or VT normalized to C,RS [28], and ΔP,L is defined as the difference between ∆P,L at end inspiration and ∆P,L at end expiration. ∆P,L can be calculated as:

$$\Delta },_}} \, = \,\left( },_}}} \, - \,}_},} - }}} } \right)\, - \,(}_}}} \, - \,}_},} - }}} ).$$

Both ∆P,RS and ΔP,L have been shown to correlate positively with stress and strain [29, 30]. In experimental endotoxin-induced ARDS, different combinations of VT and PEEP were used to create a range of ∆P,L. The combination of a VT of 6 ml/kg and the lowest PEEP and ∆P,L to maintain oxygenation within a normal range minimized VILI even in the presence of alveolar collapse [31]. In agreement with these results, Güldner et al. [32] observed that atelectrauma led to less inflammation than volutrauma strategies (Fig. 2). This strategy of keeping the collapsed lung closed is known as “permissive atelectasis”.

Fig. 2figure 2

Lung morphology at expiration and inspiration in experimental ARDS, mechanically ventilated with low tidal volume (VT = 6 ml/kg) and progressively increased positive end-expiratory pressure (PEEP). With low VT and low PEEP, aerated lungs (baby lung) are ventilated and collapsed lungs are at rest. With progressive increase in PEEP, at low VT, areas of lung collapse reduce, areas of overdistension increase, and areas of alveolar lung heterogeneity and pendelluft arise; these areas are concentrated around the collapsed units, which present the highest lung stress. At the highest PEEP, the area of lung collapse reduces but even though lung overdistension remains increased, the degree of lung stress and the biological impact on lung tissue reduce because the area associated with pendelluft is no longer observed

Static and dynamic ΔP,L were compared in experimental ARDS. Using the same protective VT, pressure-support ventilation (PSV) resulted in similar static ΔP,L but higher dynamic ΔP,L compared with PCV, leading to higher expression of biomarkers associated with inflammation in PCV [33]. This preclinical study suggested that the main determinant of lung injury is, therefore, the static rather than dynamic ΔP,L.

Tidal volume

Experimental models were also helpful in determining that overdistention rather than inspiratory pressure per se caused lung damage yielding volutrauma. In this context, Dreyfuss et al. [34] reported lung edema in animals ventilated with high VT (40 ml/kg), but such edema did not develop when rats underwent ventilation with increased airway pressures with the use of straps around their abdomens and chests, which reduced the VT (19 ml/kg).

Mechanical ventilation with low VT (4–6 ml/kg) induces repetitive opening and closing of airways and lung units, promoting epithelial cell damage, hyaline membrane formation, and lung edema, which has been named atelectrauma [1]. Interestingly, considering the “baby lung” in ARDS, the shear stress in atelectatic areas induces less lung damage (4–5 times lower) than the force at the edges between aerated and atelectatic lung regions [31, 35].

Recently, Felix et al. [36] showed that in experimental ARDS, lung damage caused by high VT (22 ml/kg) could be attenuated if VT increased slowly enough to progressively (0.5 ml/kg/min) reduce mechanical heterogeneity and allow the epithelial and endothelial cells, as well as the extracellular matrix of the lung, to adapt. In contrast, extending the adaptation period (0.25 ml/kg/min) increased cumulative power and did not prevent lung damage.

Dynamic ventilator variables associated with VILI

The dynamic ventilator variables associated with VILI are the RR, inspiratory airflow, and the inspiratory to expiratory time ratio (Fig. 1).

Respiratory rate

Whereas VT is set to match lung size, RR is usually set to maintain appropriate minute ventilation and meet the patient’s metabolic demand. In contrast to other ventilator variables, RR has been largely neglected compared with other potential variables that cause lung damage. However, when lungs are heterogeneously aerated, as shown in normal lungs [37] and a double-hit VILI model [38], high RR can amplify microstresses and regional strains, thus causing VILI. This phenomenon was shown to be modulated by the degree of pulmonary aeration [39]. The mechanisms of extracellular matrix, epithelial, and endothelial cell adaptation associated with different velocities of increases in RR were recently investigated in rats with experimental ARDS [40]. The animals received abrupt or different gradual increases of RR during protective ventilation. Longer RR adaptation resulted in less lung damage compared with abrupt RR increases. By promoting a gradual increase in RR, alveolar units remain open and better accommodate the stress (reduced airway pressures) for the same strain (VT). On the other hand, by promoting an abrupt increase in RR and shortening inspiratory time, only fast alveolar units remain open, which may favor alveolar overdistension, more heterogeneity, and lung damage. Thus, fast alveolar units, which better accommodate strain, tend to overdistend [41, 42]. After application of the recruitment maneuver, the fraction of slow alveolar units tends to decrease [43], as does the propensity of alveolar units to become atelectatic, which may decrease regional tidal strains and heterogeneity.

Inspiratory airflow

The inspiratory airflow can also be adjusted in some modes of ventilation, which is also a potential cause of VILI [44]. The shear stress at the top of the cells within the respiratory bronchi increased injury. In this context, in situ experiments have shown that healthy lungs support magnitudes of shear stress (15 dyn/cm2) at all alveolar opening velocities in the physiologic range. However, for a lung with increased viscosity of intra-alveoli fluid, shear stress may increase by several orders of magnitude, enough to induce epithelial cell injury [45]. Some reports have associated high inspiratory airflow profiles with gas exchange, the work of breathing, cardiovascular function, and lung damage [46, 47]. Not only the inspiratory airflow amplitude can be harmful but also the airflow waveform (e.g., constant versus decelerating) may be a relatively neglected and modifiable determinant of VILI risk in ARDS [6, 48].

Expiratory airflow: addressing expiration

Traditionally, less attention is given to the expiration phase than to inspiration during controlled mechanical ventilation. Nevertheless, the passive de-pressurization of the respiratory system in conventional ventilation modes predisposes closure of the distal airway and atelectasis formation. However, during so-called flow-controlled ventilation (FCV), airflows during inspiration and expiration are actively controlled and constant, whereas the airway pressure alternates between a peak and end-expiratory pressure, creating a triangular airway pressure profile [49, 50]. Thereby, FCV avoids zero-flow conditions. Along with physiologic improvements, FCV was shown to reduce VILI compared with conventional ventilation [49, 51]. Furthermore, Wittenstein et al. [50] showed that, regardless of fluid status, FCV reduced the mechanical power mainly due to the resistive component compared with VCV during one-lung ventilation. By actively controlling the expiratory phase, the appearance of intrinsic PEEP may be prevented, which in turn promotes better air exhalation among alveoli with different time constants. In a recent preclinical study, Busana et al. [52] studied healthy pigs randomized to a control group and a valve group, where the expiratory flow was controlled through a variable resistor, but all the pigs were ventilated under similar VT, PEEP, and inspiratory airflow. No differences were observed in respiratory mechanics, gas exchange, hemodynamics, wet-to-dry ratios, and histology, whereas the decrease in end-expiratory lung impedance was significantly greater in the control group compared with those that used the variable resistor. The authors concluded that the reduction in expiratory flow occurred mostly across the endotracheal tube and partly in the respiratory system. The beneficial effect of the variable resistor at the expiratory phase may also be dependent on heterogeneous and injured lungs at baseline [53].

Effects of inspiratory to expiratory time ratios

In a model of mild ARDS, mechanical ventilation with increased inspiratory to expiratory ratios (2:1) led to increased gene expression of biological markers associated with inflammation and alveolar epithelial cell damage, whereas a reduced inspiratory to expiratory ratio (1:2) increased markers of endothelial cell damage, and an inspiratory to expiratory of 1:1 minimized lung damage [54]. Similar results were observed in another preclinical study using high VT and prolonged inspiratory time [55].

Mechanical power as a hub for the development of VILI

Mechanical power (MP) is the mechanical energy delivered from the ventilator to the respiratory system and has been considered to be a unifying driver of VILI [49,50,51].

The following formula for MP was described in 2016 [56]:

$$}\, = \,\left( }^} \, \times \,\left[ \, \times \,},_}}} \, + \,}\, \times \,\left( \, + \,}:}} \right)/0\, \times \,}:}\, \times \,}} \right)\, + \,\Delta }\, \times \,}} \right]} \right)\, \times \,}.$$

Not all combinations of the three pressure components of energy (elastic, resistive, and PEEP components) and RR are equally hazardous. Doubling RR increases MP by 1.4-fold, doubling PEEP increases MP by twofold, whereas doubling VT increases MP by fourfold [49]. The increase in transpulmonary MP has been associated with the development of VILI [50]. Moreover, even at low VT, high MP promoted VILI [51]. In short, all combined variables of MP must be considered together [51, 52]. In a study of experimental ARDS in pigs, MP was positively correlated with pulmonary neutrophilic inflammation, which is a mainstay of ARDS pathogenesis [57]. Different formulas for MP have been described [58]. The most simplified version is based on the classic equation of motion:

$$}\, = \,0.0\, \times \,}_}} \, \times \,}\, \times \,(},_}}} \, - \,\Delta },_}}} /).$$

This formula computes three components, i.e., static PEEP × volume, elastic, and resistive; other formulas compute only the elastic and/or resistive component [59]. Whether the static PEEP × volume component should be included in the MP formula or not is a topic of intense debate [

Comments (0)

No login
gif