Ventilator-induced Lung Injury: Power to the Mechanical Power

We read with interest the study on “Mechanical Power and Development of Ventilator-induced Lung Injury” recently published by Cressoni et al.¹  in Anesthesiology. This study aimed to identify a power threshold for ventilator-induced lung injury (VILI) in noninjured lungs. The authors applied power (energy per breath times respiratory rate [RR]) above and below the threshold for VILI, as measured using computed tomography scans. Mechanical power is a function of transpulmonary driving pressure (ΔP,L), tidal volume (V_T), and RR. For this purpose, piglets were ventilated with higher power (V_T = 38 ml/kg and RR = 15 bpm) or at the same V_T with a different RR. The authors identified the mechanical power threshold for VILI as 12 J/min, conducted further experiments at RR = 35 bpm and at powers below and above the threshold, and noted that, if mechanical power is above 12 J/min, VILI may develop. However, the absolute value of power depends on animal size, lung volume, and RR, thus limiting direct translation of this threshold to the clinical setting. In addition to the concepts of ΔP,L, energy, and power, we should consider the concept of “intensity,” i.e., the distribution of power per unit of lung surface area. If mechanical power increases without changes in lung surface area, the intensity will be higher. On the other hand, if both power and lung surface area increase, e.g., due to lung recruitment, the “intensity” may reduce or remain constant. In the presence of regional inhomogeneities of aeration, measurement of lung surface area requires careful evaluation. Alveoli may open homogeneously, present areas of overdistension, or collapse with a similar overall surface area. In these situations, the regional “intensity” is higher in the presence of alveolar inhomogeneity.

Based on this interesting observation, we decided to conduct a further analysis of data from our group recently published in Anesthesiology.²  In this study, we evaluated the effects of different combinations of V_T, positive end-expiratory pressure (PEEP), and ΔP,L, at the same level of minute ventilation, on VILI in experimental acute respiratory distress syndrome (ARDS). These data enable us to investigate the impact of ΔP,L, energy, and power on VILI in two different conditions: (a) increasing V_T (from 6 to 22 ml/kg) and fixed PEEP level (3 cm H₂O) and (a) fixed V_T = 6 ml/kg with increasing PEEP levels (3 to 11 cm H₂O). We observed that, in the first condition, the correlations between power and gene expressions of interleukin-6 and amphiregulin were more robust than for energy and ΔP,L (fig. 1), whereas in the second condition, there was no correlation of ΔP,L, energy, or power with these biological markers.

In short, our data suggest that mechanical power, but not ΔP,L or energy, was the main determinant of activation of these markers of inflammation and alveolar stress in experimental ARDS. Additionally, in the presence of low V_T, the increase in PEEP levels did not significantly affect ΔP,L, energy, or mechanical power, thus suggesting that static stress plays a minor role in VILI (within the limits considered). These concepts apply to experimental ARDS during controlled mechanical ventilation and cannot be directly transferred to assisted mechanical ventilation.

Overall, the data of Cressoni et al.¹  and ours suggest that mechanical power applied should be taken into account to minimize VILI, whether in uninjured or injured lungs. This is a first step toward better understanding the role of mechanical power in ARDS. Further experimental and clinical studies are needed to assess the impact of the interactions between V_T, PEEP, ΔP,L, energy, mechanical power, and “intensity” on lung function and inflammation.

The authors express their gratitude to Mr. Filippe Vasconcellos (Porto Alegre, Rio Grande do Sul, Brazil) for his assistance in editing the manuscript and Ms. Marcella Rocco (Laboratory of Cellular and Molecular Physiology, Carlos Chagas Filho Biophysics Institute, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil) for her help in mathematical physics analysis.

The authors declare no competing interests.