Ethical approval
The animals were handled according to the European and national regulations for the protection of experimental animals (2010/63/UE and RD 53/2013). Ethical approval (Ethical Committee CEEA/OH 002-001/2022 Ref. PROEX 034.2/22) was provided by the Institutional Animal Care and Use Committee of the Puerta de Hierro-Segovia de Arana Health Research Institute, Madrid, Spain. Our experiments were performed in accordance with the ARRIVE guidelines for experimentation in animal research
Experimental protocol
20 male Wistar Han rats aged 4 – 5 months were used. Their mean weight was 357 ± 21 g. An online data supplement (S1) provides a detailed experimental protocol. After the induction of general anaesthesia and adequate monitoring (SpO2, temperature, invasive arterial blood pressure), we inserted a 16-gauge polyethylene catheter into the trachea. It was connected to an anaesthesia Flow-i C20 (Getinge, Solna, Sweden) machine through a “Y connector” with a minimum dead space (< 0.2 mL) and a three-way stopcock (figure E1). Then we started MV as follows: pressure – controlled continuous mechanical ventilation (PC-CMV) with PEEP = 5 cm H2O, DP = 14 cm H2O, FiO2 = 0.5, sevoflurane 3% and RR = 40 bpm (adjusted to obtain an end-tidal CO2 between 40-45 mmHg). We established ARDS following the validated model proposed by Germann and Häffner13 through repeated bronchoalveolar lavages (BAL). In our case, we performed 5 to 7 BAL with 10 mL/kg of warmed saline solution 0.9% to generate a severe ARDS according to the Berlin definition of ARDS2. After the BAL, we waited at least 15 minutes before obtaining blood gas samples with PC – CMV but with FiO2 = 0.8 to ensure post- ARDS oxygenation (SpO2 > 92%). The material obtained in the first two BALs was stored in a test tube and transported to the laboratory.
Individualized DP titration
After setting an individualized PEEP for each animal in the individualized DP group (a detailed protocol of experimentation is provided in an online data supplement), we calculated for each animal their individualized DP (adjusted to pulmonary elastance). We changed the ventilatory mode to continuous positive airway pressure (CPAP) for the DP titration, with the individualized PEEP. We interrupted the connection between the anaesthesia machine and the rat by closing the Flow - I port from the 3-way stopcock. Then we introduced a known volume of 14 mL of oxygen (in steps of 1mL each step) through the stopcock, and we measured Pmean for each mL that we introduced. We introduced the data in the software Excel 16.59 and obtained a pressure-volume graph.
An example is presented in Figure 1 – B. According to the Young’s curve, the maximum slope of the graph would represents the elastic limit, followed by the flat zone, which indicates the transition zone. We considered this point as the individualized driving pressure. It is important to emphasize that the first part of the graph (‘elastic behavior’) corresponds to a flat zone in the animal’s graph because the individualized PEEP already opens the lung.
Experimental groups
A computer-generated random list selected one of the two groups. In the standard group 10 animals received 120 minutes of MV with the standard ventilation parameters described previously (PEEP = individualized PEEP, FiO2 = 0,8, RR = 55 bpm, DP = 14 cm H2O always checking that TV was < 6mL/kg). In the experimental group, 10 animals received MV for 120 minutes with the following parameters: PEEP = individualized PEEP, DP = individualized DP, FiO2 = 0.8, RR = 55 bpm). Hence, the total MV time was 150 minutes: 30 minutes during the induction and 120 minutes with the parameters we described. After this time, we finished the experiment, obtained arterial blood for blood gas analysis, and performed two BALs with the same solution and volume as before. The obtained material was kept in a test tube and transported to the laboratory. The animals were sacrificed, and their lungs were collected (an online data supplement (S1) provides a detailed experimental protocol).
BAL analysis
Additional details on the BAL processing can be found in S1.Samples were analyzed using a FACSort flow cytometer. We also measured the concentrations of tumor necrosis factor (TNF)-α, interleukin (IL)-6, and IL-1β using enzyme-linked immunosorbent assays (ELISAs) (Elabscience, China) according to the manufacturer's protocol.
Histopathological grading of VILI
After processing the lungs (S1), two pathologists’ experts in interstitial pneumopathies blindly analyzed lung injury through the most used score to analyze VILI: the ALI score14,15. This score has four items: a) alveolar capillary congestion, b) hemorrhage, c) infiltration of neutrophils into the airspace or the vessel wall, and thickness of the alveolar wall, and d) alveolar wall thickness/hyaline membrane formation. Every item received punctuation ranged from zero to four, corresponding zero to normal findings, one to mild (< 25% of the surface), two to moderate (25 – 50%), three severe (50 – 75%), and four to very intense (> 75%) lung involvement. The score used for the results was the mean score for each sample from both pathologists.
Statistical analysis.
Sample size and power calculations were based on previous data from similar studies and determined using the Granmo 7.12 software program (Institut Municipal d’Investigació Mèdica, Spain). Ten animals per group were deemed adequate to accept an alpha risk of 0.05 and a beta risk of 0.05 in a two-sided test. A normality test (Kolmogorov-Smirnov test with Dallar – Wilkinson – Llilliefor p-value) was performed to analyze quantitative variables. If the sample followed the normal distribution, quantitative data were shown as mean [95% interval confidence for the mean], and a t-test was performed to check for differences between groups. If they did not, quantitative data are shown as median [p25, p75]. If the data distribution was parametric, a t-test was performed to check for differences between groups, and a Mann-Whitney test was performed if it did not. Qualitative variables are presented as %, and after checking if the collected variable met the minimum characteristics required, a χ2 test with Fisher correction was performed to check for differences between groups. Statistical analysis was performed using Prism Graphpad 9.0 software.