Acute respiratory distress syndrome (ARDS) accounts for approximately 10% of all ICU admissions and 23% of patients requiring mechanical ventilation (MV). Despite advances in care, hospital mortality remains high, ranging from 34% in mild cases to 46% in severe ARDS. Positive-pressure MV remains the cornerstone of ARDS management. However, when excessive stress and strain are applied to the lung parenchyma, it can exacerbate lung injury, leading to ventilator-induced lung injury (VILI). VILI substantially contributes to morbidity and mortality in ARDS. Strategies that reduce tidal volume (Vt), driving pressure (ΔP, defined as plateau pressure minus PEEP), and respiratory rate (RR) can lower the mechanical power (PowerRS), i.e., the energy delivered to the lungs by the ventilator. This reduction in pulmonary stress and strain may lessen VILI and potentially improve survival. Nonetheless, reducing Vt to <6 ml/kg in order to achieve plateau pressures <23-25 cm H₂O, driving pressures <9-11 cm H₂O, and RR <15-20/min can result in severe hypercapnia. This, in turn, may increase intracranial pressure, promote pulmonary hypertension, impair myocardial contractility, reduce renal perfusion, and trigger endogenous catecholamine release. Thus, such "ultraprotective" MV strategies are not feasible for most ARDS patients managed with conventional ventilation. The neutral findings of the REST trial further suggested that low-flow extracorporeal CO₂ removal (ECCO₂R) devices may provide insufficient CO₂ clearance to enable ultraprotective ventilation while adequately controlling respiratory acidosis. Moreover, since partial lung derecruitment may occur with substantial Vt reduction, extracorporeal membrane oxygenation (ECMO) may be necessary, particularly in patients with PaO₂/FiO₂ <120-130 at the time of Vt reduction. Therefore, respiratory extracorporeal life support (ECLS)-ranging from high-flow ECCO₂R to mid-flow venovenous ECMO (VV-ECMO)-can be employed in this setting. These modalities facilitate further reductions in ventilatory intensity while ensuring adequate oxygenation and CO₂ removal.

Inclusion Criteria:

1. Intubation and Invasive mechanical ventilation ≤ 7 days

2. Presence of all of the following conditions for ≤48 hours:

   • 80 ≤ PaO2/FiO2 ≤300 with PEEP >5 cmH2O
   • Bilateral opacities not fully explained by effusions, lobar/lung collapse, or nodules
   • Respiratory failure not fully explained by cardiac failure or fluid overload

3. One of the following criteria (with Vt set at 6 mL/kg PBW):

   • DeltaP ≥15 cm H2O OR
   • Ventilatory ratio ≥2.2

4. Signed Informed consent from a close relative or surrogate or a family member. According to the specifications of emergency inclusion, randomization without the close relative/surrogate consent could be performed if the patient is unable to give his/ger consent and when the close relative/surrogate/family member are absent. Close relative/surrogate/family member consent will be asked as soon as possible after randomization. The patient will be asked as soon as possible to give his/her consent for the continuation of the trial when his/her condition will allow.

5. Social security registration (AME excluded)

Exclusion Criteria:

1. Age <18 years
2. Pregnancy or breastfeeding
3. Catheter access to femoral vein or jugular vein impossible
4. Expected duration of mechanical ventilation < 48 hours
5. Chronic restrictive or obstructive (COPD) respiratory insufficiency with home ventilation or oxygen therapy
6. Currently receiving ECLS therapy
7. Severe cardiac failure or ongoing acute coronary syndrome
8. Heparin-induced thrombocytopenia
9. Severe underlying pre-existing condition with expected six-month mortality >50%
10. Contraindication for systemic anticoagulation (including platelet count <50G/L)
11. Patient moribund, decision to limit therapeutic interventions
12. Acute brain injury or irreversible neurological pathology
13. Bone marrow transplantation within the last 1 year
14. Actual body weight exceeding 1 kg per centimeter of height
15. Prior enrolment in the trial

The procedures carried out with the French data privacy authority (CNIL, Commission nationale de l'informatique et des libertés) do not provide for the transmission of the database, nor do the information and consent documents signed by the patients.

Consultation by the editorial board or interested researchers of individual participant data that underlie the results reported in the article after deidentification may nevertheless be considered, subject to prior determination of the terms and conditions of such consultation and in respect for compliance with the applicable regulations.

Beginning 3 months and ending 3 years following article publication. Requests out of these time frame can also be submitted to the sponsor

Researchers who provide a methodologically sound proposal