Barotrauma (pneumothorax, pneumomediastinum) is a well-described complication of Acute Respiratory Distress Syndrome (ARDS), especially in patients with coronavirus disease 2019 (COVID-19) (16.1% in COVID-19, and about 6% in non-COVID-19 ARDS). Macklin effect was recently discovered by our group as an accurate radiological predictor of barotrauma in COVID-19 ARDS; the Investigators also found that density histograms automatically extracted from chest CT images provide a reliable insight into lung composition

. Since lung frailty is a major issue also in non-COVID-19 ARDS, the Investigators want to confirm the predictive role of Macklin effect also in this setting. In addition, the Investigators aim to explore inflammatory profiling to decipher different biological aspects of the same clinical issue. Finally, the Investigators want to develop a specific management algorithm for patients diagnosed, according to our findings, with a specific ARDS sub phenotype characterized by increased lung frailty

Barotrauma occurs frequently in acute respiratory distress syndrome (ARDS), and has a difficult, non-standardized management. Unfortunately, mortality rates remain high (> 60% in COVID-19 ARDS, around 46% in non-COVID-19 ARDS). Interestingly, data from COVID-19 patients suggested that barotrauma may occur also in spontaneously breathing patients with ARDS. Accordingly, frailty of lung parenchyma represents a major issue in ARDS. Protective mechanical ventilation (i.e. ventilation with low tidal volume and low airway pressures) remains a cornerstone of supportive management of ARDS. Unfortunately, mechanical ventilation may worsen pulmonary damage (ventilator-induced lung injury) and, in high-risk patients, may induce barotrauma even when ventilator settings are maintained within the "safe" limit of protective ARDS. Early identification of high-risk features could therefore allow clinicians to individualize management of high-risk patients, by tailoring respiratory support and potentially select candidates for advanced support (i.e. extracorporeal membrane oxygenation) before development of overt barotrauma.

Macklin effect is a well-described radiological sign originally intended to differentiate between "peripheral" (distal airway rupture, "respiratory" barotrauma) and "central" (lesion to large airways/esophaegal injury) causes of air leakage in the mediastinum. However, the Investigators recently identified Macklin effect as a strong radiological predictor of barotrauma development in mechanically ventilated COVID-19 ARDS patients (sensitivity: 89.2%; specificity: 95.6%). In our cohort, radiologically-detected Macklin effect was identified 8-12 days before development of pneumomediastinum/pneumothorax. These preliminary results have been confirmed in a subsequent multicenter study (sample size 697 patients; sensitivity: 100%; specificity: 99.8%).

Furthermore, preliminary data suggest that early application of awake veno/venous extracorporeal membrane oxygenation (ECMO) before invasive mechanical ventilation in COVID-19 patients with severe ARDS and at high-risk for barotrauma (defined as presence of Macklin effect on chest CT imaging) might result in no barotrauma events with a low intubation rate.

Concurrently, a hyper inflammatory sub phenotype has been associated with overall worse outcome both in terms of mortality and ventilator-free days in ARDS. Moreover, the occurrence of lung injury during mechanical ventilation has been proven to be significantly related to the recruitment of mast cells via CXCL10/CXCR3 signaling .

In this view, confirmation of Macklin effect predictive role and identification of further, novel laboratory biomarkers could provide instruments for early risk stratification in ARDS patients.

Taken together, i) quantitative imaging analysis and ii) systemic inflammatory profiling could decipher different biological aspects of the same clinical issue, possibly laying foundation for the definition of a multimodality signature of lung frailty in ARDS patients..

Accordingly, the driving hypotheses of this retrospective/prospective study is that identification of a novel ARDS sub phenotype characterized, irrespective of the underlying etiology, by increased lung frailty could substantially improve the poor prognosis routinely associated with this condition, possibly being a landmark for personalized management strategies.

To further validate the role of Macklin effect, the Investigators will:

• evaluate the accuracy of Macklin effect in a retrospective cohort of 350 ARDS patients (COVID-19 and non-COVID-19)
• identify, throughout densitometry, machine learning and artificial intelligence-based approaches, novel imaging biomarkers characteristics of higher lung frailty in the same cohort.

In the main prospective study, the Investigators will:

• analyse the following biomarkers in the serum and bronchoalveolar lavage fluid of 100 ARDS patients prospectively enrolled: Interleukin-8 (IL-8), Interleukin (IL)-6, IL-1Ra, IL-18, interferon (IFN ), Angiopoietin-2 (Ang-2), Tumour Necrosis Factor receptor-1 (TNFr1), Plasminogen Activator Inhibitor-1(PAI-1), Receptor for Advanced Glycation Endproducts (RAGE), Intercellular adhesion molecule-1 (ICAM-1), Surfactant Protein D (SPD), protein C, Von Willebrand Factor (VWF), CXCL10/CXCR3, and metalloproteases (MMP9, MMP10).
• Develop a specific management algorithm for ARDS patients at high risk for barotrauma by collecting clinical and outcome data from 10 ARDS patients receiving unconventional management (e.g. awake ECMO, ultraprotective ventilation, etc)

Inclusion Criteria:

• Clinical and radiological signs of ARDS, according to Berlin criteria, requiring ICU admission;
• Obtain duly signed informed consent
• Availability of at least one chest CT scan during hospital stay

Exclusion Criteria:

• Poor quality imaging (because of motion/respiratory artefacts).

Patients with ARDS diagnosis will be enrolled. With the exception of blood and BAL samples collection, all patients enrolled in the prospective cohort will follow their normal course of diagnosis/treatment without any further requirements for the scope of this study other than those provided in the normal management of patients with this condition. Furthermore, any decision regarding drug or procedure will be made by the physician based on her/his clinical judgment, independently from the decision to include the patient in this study. No follow-up medical and/or imaging examinations/laboratory analysis after hospital discharge are needed by the present study design.