Acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) are the most frequent complications of the COVID-19 pandemic. In these conditions, hypoxemia may result from : i) a pulmonary vascular dilatation resulting from an impaired hypoxic pulmonary vasoconstriction and leading to ventilation-perfusion mismatching within the lungs and ii) thrombosis-mediated perfusion defects. Pulmonary vascular dilation might be due to a relative failure of the physiological acute hypoxic pulmonary vasoconstriction, in the context of an over-activation of a regional vasodilatation cascade, as part of a dysfunctional inflammatory process. Perfusion abnormalities associated with pulmonary vascular dilation are suggestive of intrapulmonary shunting toward areas where gas exchange is impaired, ultimately leading to a worsening ventilation-perfusion mismatch, a regional hypoxia and a profound hypoxemia.

Increased plasma levels of VEGF have been reported in moderate to severe COVID-19 pneumonia, highlighting the role of VEGF in the pathophysiology of the disease. A better prognosis has been reported in critically ill patients with lower levels of growth factors, HGF and VEGF-A at the time of ICU admission. Recent data of the study NCT 04275414 by Pang J et al have suggested that patients receiving a single-dose of bevacizumab have improved their oxygen support status in 92% of cases during a 28-day follow-up period, as compared with 62% of cases in an external cohort receiving standard care.

Correcting endothelial permeability and vasodilatation with VEGF-targeted therapy could allow repair damaged vascular endothelium, have an indirect anti-inflammatory effect (limiting alveolar exudation of circulating inflammatory and procoagulant mediators) and improve oxygenation and therefore reduce the proportion of patients with severe forms requiring ICU referral and finally patient death. This clinical trial will therefore focus on the specific efficacy of bevacizumab in COVID-19 patients with severe hypoxemia.

Inclusion Criteria:

• Patients included in the CORIMUNO-19 cohort
• Patients hospitalized in conventional ward or in the ICU belonging to the following groups: OMS Progression scale 6, 7, 8 AND no acute pulmonary embolism on CT-scan performed in the preceding 72 hours no pulmonary evident bacterial coinfection or superinfection evaluated by non-invasive procedures (serology, antigens, nasopharynx PCR, sputum examination, blood cultures…)

Exclusion Criteria:

• Patients in OMS progression class 9
• Patients with exclusion criteria to the CORIMUNO-19 cohort
• Pregnancy
• Active cancer with ongoing treatment
• acute use of NIV for COPD exacerbation or cardiac decompensation associated to COVID-19
• Oxygen patient requiring long-term oxygen before hospitalization
• Patient already included in an interventional research
• Risk of bleeding especially hemoptysis, active venous or arterial thromboembolic disease and recent surgery during the last 3 weeks
• Hypersensitivity to the active substance (bevacizumab) or to any of the excipients (sucrose, succinic acid, disodium edetate, polysorbate 80, sodium hydroxide, water for injection
• Hypersensitivity to Chinese Hamster Ovary (CHO) cell products or other recombinant human or humanised antibodies
• Persistant uncontrolled arterial hypertension after using to anti-hypertensive drugs
• Current documented bacterial infection not controlled by antibiotics.
• Active viral diseases (especially active herpes, chickenpox, shingles),
• Active tuberculosis or disseminated strongyloidiasis
• patient with known active hepatitis or with increased level of SGOT or SGPT ≥5N
• Patient with anormal laboratory results: Absolute neutrophil count (ANC) ≤ 1.0 x 109/L, Platelets (PLT) < 50 G /L