Plasma proteome dynamics in venous thromboembolism

Venous thromboembolism (VTE), including deep vein thrombosis (DVT) and pulmonary embolism (PE), is the third leading cause of morbidity and mortality among cardiovascular diseases. The current diagnostic approach assesses clinical probability through decision rules (e.g., Wells score) and measuring plasma D-dimer levels. While D-dimer helps rule out VTE in low-probability cases, patients with medium to high probability require imaging to confirm diagnosis. However, fewer than 20% of computed tomography pulmonary angiograms (CTPAs) performed for suspected PE actually confirm the diagnosis. Additionally, 25% of patients with unprovoked VTE experience a recurrence within 5 years after stopping anticoagulation. There is a clear need for better plasma marker-based tools that could enable early decision-making without relying on imaging, increasing patient safety and reducing costs. Such tools could also help identify patients at the highest risk of recurrence after unprovoked VTE, supporting decisions on extended anticoagulation where the benefits outweigh bleeding risks. As VTE is an intravascular disease, blood properties, endothelial function, and their interactions are critical in thrombosis.Proteomic analysis can help identify new biomarkers for diagnosis and risk assessment. Recent advancements in proteomics, particularly mass spectrometry and affinity-based proximity extension assays, have enabled deeper characterization of the plasma proteome and the identification of changes in protein profiles in VTE that could improve diagnosis. In our study, we have profiled plasma samples from patients with acute VTE and after a first VTE event. We used the high-throughput proximity extension assay, as part of the generation of a comprehensive blood atlas covering 59 diseases, designed to explore the circulating human proteome and identify plasma proteins as potential signatures for disease diagnosis, prognosis, and treatment management. We semi-quantified 1161 unique proteins in 48 acute VTE and 98 post-VTE plasma samples. Using differential expression analysis and machine learning, we identified VTE-related protein signatures by comparing them with circulating protein profiles from healthy, cardiovascular, and pan-disease cohorts. Our results highlight significant up- and downregulation of proteins involved in the regulation of coagulation, complement activation, endothelial function, and fibrinolysis. The most relevant proteins were selected and applied to predictive models. Our findings demonstrate the potential of proteomics technology and the open-access blood atlas for identifying proteomic signatures associated with VTE and other diseases. In this context, our study provides novel insights into the dynamics of the plasma proteome in venous thrombosis and could contribute to improved clinical decision-making in patient risk stratification and personalized treatments for VTE.