ABSTRACT
Background and Aim: We conducted a multidisciplinary study to investigate the correlations between clinical-laboratory-imaging data and histopathologic pulmonary patterns in patients died from severe CoViD-19. Materials and Methods: We analyzed lung autoptic tissue from consecutive CoViD-19 patients between February 29 and June 30, 2020. We considered three samples for each pulmonary lobe per patient. The pre-specified histopathological patterns were: exudative diffuse alveolar damage (DAD), proliferative DAD, organizing pneumonia, acute fibrinous organizing pneumonia, interstitial pneumonia, bronchopneumonia, arteriolar thrombi, intracapillary megakaryocytes, and areas of normal lung. Hierarchical cluster analysis was performed. Results: Among 92 autopsies, 48 patients had complete clinical data. Four clusters were identified. Length-of-stay in ICU and in hospital (p<0.0001), days on mechanical ventilation (p<0.0001), days on positive pressure airway (p<0.0001), mean positive endexpiratory pressure PEEP (p=0.007), PEEP x days on mechanical ventilation (p=0.003), PEEP x days on positive pressure airway (p=0.003), worst serum albumin (p=0.017), interleukin 6 (p=0.047), and kidney SOFA (p=0.001) differed between clusters. The cluster characterized by prevalence of exudative-proliferative DAD and lung megakaryocytosis had the greater difference from the others. Conclusions: Our research sheds light on the correlations between clinical-laboratory-imaging pictures and histopathologic findings, with clues on the impact of therapeutic strategies on lung tissues.
ABSTRACT
BACKGROUND: Since the first observations of patients with COVID-19, significant hypoalbuminaemia was detected. Its causes have not been investigated yet. OBJECTIVE: We hypothesized that pulmonary capillary leakage affects the severity of respiratory failure, causing a shift of fluids and proteins through the epithelial-endothelial barrier. METHODS: One hundred seventy-four COVID-19 patients with respiratory symptoms, 92 admitted to the intermediate medicine ward (IMW) and 82 to the intensive care unit (ICU) at Luigi Sacco Hospital in Milan, were studied. RESULTS: Baseline characteristics at admission were considered. Proteins, interleukin 8 (IL-8) and interleukin 10 (IL-10) in bronchoalveolar lavage fluid (BALF) were analysed in 26 ICU patients. In addition, ten autopsy ultrastructural lung studies were performed in patients with COVID-19 and compared with postmortem findings in a control group (bacterial pneumonia-ARDS and H1N1-ARDS). ICU patients had lower serum albumin than IMW patients [20 (18-23) vs 28 (24-33) g L-1 , P < 0.001]. Serum albumin was lower in more compromised groups (lower PaO2 -to-FiO2 ratio and worst chest X-ray findings) and was associated with 30 days of probability of survival. Protein concentration was correlated with IL-8 and IL-10 levels in BALF. Electron microscopy examinations of eight out of ten COVID-19 lung tissues showed loosening of junctional complexes, quantitatively more pronounced than in controls, and direct viral infection of type 2 pneumocytes and endothelial cells. CONCLUSION: Hypoalbuminaemia may serve as severity marker of epithelial-endothelial damage in patients with COVID-19. There are clues that pulmonary capillary leak syndrome plays a key role in the pathogenesis of COVID-19 and might be a potential therapeutic target.