Exploring blood transcriptional signatures for thrombosis diagnosis in critically ill patients with COVID-19

Abstract

COVID-19 has revealed novel pathological mechanisms, particularly hypercoagulability, which leads to increased thrombotic risk in critically ill patients. This study investigates transcriptional signatures associated with thrombosis development in patients with COVID-19 in the intensive care unit and evaluates their predictive potential. We performed whole-blood transcriptional profiling of 57 mechanically ventilated patients with COVID-19, comparing those with thrombotic complications (TC; n = 36) with those without (non-TC, n = 21) using differential gene expression and machine learning approaches. Patients with TC showed greater transcriptome disruption and 283 differentially expressed genes compared with patients without TC. Key features included enhanced neutrophil activation, inflammatory responses, and monocyte activation alongside suppressed lymphocyte function. An orthogonal partial least squares discriminant analysis model achieved excellent classification performance (area under the curve [AUC] = 0.961; 95% confidence interval, 0.905-0.997). The maltase-glucoamylase (MGAM) gene was the top discriminatory biomarker outperforming traditional clinical markers such as D-dimer and C-reactive protein (AUC, 0.94). TC in critically ill patients with COVID-19 are characterized by distinct transcriptional signatures reflecting heightened neutrophil activation and inflammatory dysregulation. MGAM represents a novel potential biomarker that outperforms traditional clinical markers for identifying patients at high thrombotic risk, offering new opportunities for personalized risk stratification and management in severe COVID-19.