Work

Integrated Analysis of Parenchymal and Vascular HRCT Patterns with Circulating Biomarkers in Severe COVID-19 Pneumonia

Year: 2026

Type: article

Abstract: Purpose: To explore the correlation between radiologic patterns on high-resolution computed tomography (HRCT) and circulating biomarkers of inflammation and endothelial activation in patients with COVID-19 pneumonia, with the aim of identifying imaging-biomarker phenotypes that may offer insights for clinical stratification. Materials and Methods: This prospective single-center study included 84 consecutive patients hospitalized with PCR-confirmed SARS-CoV-2 infection and respiratory failure. All underwent baseline HRCT, along with parallel biohumoral profiling, including inflammatory (IL-1Ra, IL-6, IL-10) and endothelial (Angiopoietin-2, sVCAM-1, sE-Selectin) biomarkers. HRCT scans were reviewed for parenchymal and vascular abnormalities (vascular tree-in-bud [TIB], vascular enlargement pattern [VEP]). Semi-quantitative scores were assigned for parenchymal (PS) and vascular (VS) involvement. Results: Patients with higher PS had significantly prolonged hospital stay (35 vs. 17 days; p = 0.014), increased ICU admission rates (68.8% vs. 21.4%; p = 0.003), and elevated serum levels of IL-1Ra, IL-6, and IL-10 (p < 0.05). At multivariable analysis, PS remained independently associated with ICU admission after adjustment for age, inflammatory burden, and comorbidities (p = 0.014). A high VS was associated with significantly increased Angiopoietin-2 levels (p = 0.036), although it did not directly correlate with ICU admission or mortality. A significant positive correlation was observed between PS and VS (r =0.392; p < 0.001). Conclusions: in this study, HRCT-based parenchymal and vascular patterns appear significantly correlated with biological processes occurring in severe COVID-19 pneumonia. These observations, although preliminary, may offer a conceptual basis for future studies exploring radiologic and biomarker-based stratification in severe respiratory infections.

Source: Diagnostics

DOI: 10.3390/diagnostics16040587