(1) Background: We aimed to analyze the characteristics associated with the in-hospital mortality, describe the early CT changes expressed quantitatively after tocilizumab (TOC), and assess TOC timing according to the oxygen demands. (2) Methods: We retrospectively studied 101 adult patients with severe COVID-19, who received TOC and dexamethasone. The lung involvement was assessed quantitatively using native CT examination before and 7−10 days after TOC administration. (3) Results: The in-hospital mortality was 17.8%. Logistic regression analysis found that interstitial lesions above 50% were associated with death (p = 0.01). The other variables assessed were age (p = 0.1), the presence of comorbidities (p = 0.9), the oxygen flow rate at TOC administration (p = 0.2), FiO2 (p = 0.4), lymphocyte count (p = 0.3), and D-dimers level (p = 0.2). Survivors had a statistically significant improvement at 7−10 days after TOC of interstitial (39.5 vs. 31.6%, p < 0.001), mixt (4.3 vs. 2.3%, p = 0.001) and consolidating (1.7 vs. 1.1%, p = 0.001) lesions. When TOC was administered at a FiO2 ≤ 57.5% (oxygen flow rate ≤ 13 L/min), the associated mortality was significantly lower (4.3% vs. 29.1%, p < 0.05). (4) Conclusions: Quantitative imaging provides valuable information regarding the extent of lung damage which can be used to anticipate the in-hospital mortality. The timing of TOC administration is important and FiO2 could be used as a clinical predictor.
Keywords: quantitative chest CT; severe COVID-19; tocilizumab timing.