Thrombin-Antithrombin Complexes and sICAM-1 Are Associated With In-Hospital Mortality in COVID-19 Patients With Inflammatory Storm

Abstract

Aim To evaluate whether endothelial activation (soluble intercellular adhesion molecule-1, sICAM-1) and thrombotic activation (thrombin-antithrombin complexes, TAT) markers provide prognostic information beyond standard inflammatory markers for in-hospital mortality in COVID-19 patients with cytokine storm. Materials and methods Prospective cohort study of 47 patients with severe SARS-CoV-2 pneumonia admitted November 2020–February 2021 to District Hospital, Debica, Poland. Patients were stratified by cytokine storm profile (CRP >50 mg/L or IL-6 >30 pg/mL) and survival status. Blood biomarkers included CRP, IL-6, sICAM-1, TAT, d-dimer, and antiphospholipid antibodies. Results Mortality rate was 40.4% (19/47 patients). Among patients with cytokine storm, TAT and sICAM-1 levels were significantly higher in non-survivors versus survivors (TAT: 15.4 (12.2–23.6) ng/mL vs. 11.6 (11.0–14.8) ng/mL, p=0.037; sICAM-1: 12.0 (5.6–35.3) ng/mL vs. 4.3 (2.5–23.9) ng/ml, p=0.033), whereas d-dimer and anticardiolipin IgM did not differentiate mortality outcomes. Elevated TAT was detected in 100% of patients, d-dimer in 93.6%, anticardiolipin IgM in 40.4%, and sICAM-1 in 21.3%. In the multivariable exploratory model based on Firth’s logistic regression, TAT ≥12.14 ng/mL and sICAM-1 ≥4.0 ng/mL emerged as potential independent predictors of mortality (OR: 6.27, 95% CI: 1.31–29.99 and OR: 8.68, 95% CI: 1.22–62.15, respectively). Traditional prognostic factors (diabetes on insulin, and blood desaturation) also discriminated poor outcomes. Conclusions sICAM-1 and TAT were associated with in-hospital mortality within the high-risk subgroup of COVID-19 patients with cytokine storm, suggesting prognostic information beyond conventional inflammatory markers and d-dimer. Given the limited sample size, these exploratory findings should be regarded as hypothesis-generating and require validation in larger, multicentre cohorts before informing risk stratification.