Purpose Little evidence of increased thrombotic risk is available in COVID-19 patients. Our purpose was to assess thrombotic risk in severe forms of SARS-CoV-2 infection. Methods All patients referred to 4 intensive care units (ICUs) from two centers of a French tertiary hospital for acute respiratory distress syndrome (ARDS) due to COVID-19 between March 3rd and 31st 2020 were included. Medical history, symptoms, biological data and imaging were prospectively collected. Propensity score matching was performed to analyze the occurrence of thromboembolic events between non-COVID-19 ARDS and COVID-19 ARDS patients. Results 150 COVID-19 patients were included (122 men, median age 63 [53; 71] years, SAPSII 49 [37; 64] points). Sixty-four clinically relevant thrombotic complications were diagnosed in 150 patients, mainly pulmonary embolisms (16.7%). 28/29 patients (96.6%) receiving continuous renal replacement therapy experienced circuit clotting. Three thrombotic occlusions (in 2 patients) of centrifugal pump occurred in 12 patients (8%) supported by ECMO. Most patients (> 95%) had elevated D-dimer and fibrinogen. No patient developed disseminated intravascular coagulation. Von Willebrand (vWF) activity, vWF antigen and FVIII were considerably increased, and 50/57 tested patients (87.7%) had positive lupus anticoagulant. Comparison with non-COVID-19 ARDS patients (n = 145) confirmed that COVID-19 ARDS patients (n = 77) developed significantly more thrombotic complications, mainly pulmonary embolisms (11.7 vs. 2.1%, p < 0.008). Coagulation parameters significantly differed between the two groups. Conclusion Despite anticoagulation, a high number of patients with ARDS secondary to COVID-19 developed life-threatening thrombotic complications. Higher anticoagulation targets than in usual critically ill patients should therefore probably be suggested.

Background: Neurotropism of SARS-CoV-2 and its neurological manifestations have now been confirmed. We aimed at describing delirium and neurological symptoms of COVID-19 in ICU patients. Methods: We conducted a bicentric cohort study in two French ICUs of Strasbourg University Hospital. All the 150 patients referred for acute respiratory distress syndrome due to SARS-CoV-2 between March 3 and May 5, 2020, were included at their admission. Ten patients (6.7%) were excluded because they remained under neuromuscular blockers during their entire ICU stay. Neurological examination, including CAM-ICU, and cerebrospinal fluid analysis, electroencephalography, and magnetic resonance imaging (MRI) were performed in some of the patients with delirium and/or abnormal neurological examination. The primary endpoint was to describe the incidence of delirium and/or abnormal neurological examination. The secondary endpoints were to describe the characteristics of delirium, to compare the duration of invasive mechanical ventilation and ICU length of stay in patients with and without delirium and/or abnormal neurological symptoms. Results: The 140 patients were aged in median of 62 [IQR 52; 70] years old, with a median SAPSII of 49 [IQR 37; 64] points. Neurological examination was normal in 22 patients (15.7%). One hundred eighteen patients (84.3%) developed a delirium with a combination of acute attention, awareness, and cognition disturbances. Eighty-eight patients (69.3%) presented an unexpected state of agitation despite high infusion rates of sedative treatments and neuroleptics, and 89 (63.6%) patients had corticospinal tract signs. Brain MRI performed in 28 patients demonstrated enhancement of subarachnoid spaces in 17/28 patients (60.7%), intraparenchymal, predominantly white matter abnormalities in 8 patients, and perfusion abnormalities in 17/26 patients (65.4%). The 42 electroencephalograms mostly revealed unspecific abnormalities or diffuse, especially bifrontal, slow activity. Cerebrospinal fluid examination revealed inflammatory disturbances in 18/28 patients, including oligoclonal bands with mirror pattern and elevated IL-6. The CSF RT-PCR SARS-CoV-2 was positive in one patient. The delirium/neurological symptoms in COVID-19 patients were responsible for longer mechanical ventilation compared to the patients without delirium/neurological symptoms. Delirium/neurological symptoms could be secondary to systemic inflammatory reaction to SARS-CoV-2. Conclusions and relevance: Delirium/neurological symptoms in COVID-19 patients are a major issue in ICUs, especially in the context of insufficient human and material resources. Trial registration: NA.

Host infection by a micro-organism triggers systemic inflammation, innate immunity and complement pathways, but also haemostasis activation. The role of thrombin and fibrin generation in host defence is now recognised, and thrombin has become a partner for survival, while it was seen only as one of the “principal suspects” of multiple organ failure and death during septic shock. This review is first focused on pathophysiology. The role of contact activation system, polyphosphates and neutrophil extracellular traps has emerged, offering new potential therapeutic targets. Interestingly, newly recognised host defence peptides (HDPs), derived from thrombin and other “coagulation” factors, are potent inhibitors of bacterial growth. Inhibition of thrombin generation could promote bacterial growth, while HDPs could become novel therapeutic agents against pathogens when resistance to conventional therapies grows. In a second part, we focused on sepsis-induced coagulopathy diagnostic challenge and stratification from “adaptive” haemostasis to “noxious” disseminated intravascular coagulation (DIC) either thrombotic or haemorrhagic. Besides usual coagulation tests, we discussed cellular haemostasis assessment including neutrophil, platelet and endothelial cell activation. Then, we examined therapeutic opportunities to prevent or to reduce “excess” thrombin generation, while preserving “adaptive” haemostasis. The fail of international randomised trials involving anticoagulants during septic shock may modify the hypothesis considering the end of haemostasis as a target to improve survival. On the one hand, patients at low risk of mortality may not be treated to preserve “immunothrombosis” as a defence when, on the other hand, patients at high risk with patent excess thrombin and fibrin generation could benefit from available (antithrombin, soluble thrombomodulin) or ongoing (FXI and FXII inhibitors) therapies. We propose to better assess coagulation response during infection by an improved knowledge of pathophysiology and systematic testing including determination of DIC scores. This is one of the clues to allocate the right treatment for the right patient at the right moment.

In sepsis, inflammation and thrombosis are both the cause and the result of interactions between circulating (for example, leukocytes and platelets), endothelial and smooth muscle cells. Microparticles are proinflammatory and procoagulant fragments originating from plasma membrane generated after cellular activation and released in body fluids. In the vessel, they constitute a pool of bioactive effectors pulled from diverse cellular origins and may act as intercellular messengers. Microparticles expose phosphatidylserine, a procoagulant phospholipid made accessible after membrane remodelling, and tissue factor, the initiator of blood coagulation at the endothelial and leukocyte surface. They constitute a secretion pathway for IL-1β and up-regulate the proinflammatory response of target cells. Microparticles circulate at low levels in healthy individuals, but undergo phenotypic and quantitative changes that could play a pathophysiological role in inflammatory diseases. Microparticles may participate in the pathogenesis of sepsis through multiple ways. They are able to regulate vascular tone and are potent vascular proinflammatory and procoagulant mediators. Microparticles' abilities are of increasing interest in deciphering the mechanisms underlying the multiple organ dysfunction of septic shock.

Cardiogenic shock is usually defined as primary cardiac dysfunction with low cardiac output leading to critical organ hypoperfusion, and tissue hypoxia, resulting in high mortality rate between 40% and 50% despite recent advances. Many studies have now evidenced that cardiogenic shock not only involves systemic macrocirculation, such as blood pressure, left ventricular ejection fraction, or cardiac output, but also involves significant systemic microcirculatory abnormalities which seem strongly associated with the outcome. Although microcirculation has been widely studied in the context of septic shock showing heterogeneous alterations with clear evidence of macro and microcirculation uncoupling, there is now a growing body of literature focusing on cardiogenic shock states. Even if there is currently no consensus regarding the treatment of microcirculatory disturbances in cardiogenic shock, some treatments seem to show a benefit. Furthermore, a better understanding of the underlying pathophysiology may provide hypotheses for future studies aiming to improve cardiogenic shock prognosis.