Highlights
- Large-scale proteomics identified 23 proteins associated with incident noncancer venous thromboembolism (VTE), with 15 being novel markers not previously linked to the disease.
- Mendelian randomization (MR) provides evidence for a causal role of TIMD4, TIMP4, and Cystatin-C in the pathogenesis of VTE, highlighting non-clotting pathways.
- Identified pathways include extracellular matrix (ECM) regulation, immunity, vascular endothelium interactions, and vascular senescence, suggesting a much broader etiology for VTE than Virchow’s Triad.
- Clinical decision-making for intensive cardiovascular risk management in older adults is heavily influenced by frailty and age, though digital monitoring can mitigate clinical inertia.
Background
Venous thromboembolism (VTE), encompassing deep vein thrombosis and pulmonary embolism, remains a leading cause of cardiovascular morbidity and mortality globally. Despite the established roles of hypercoagulability, stasis, and endothelial injury, many incident VTE cases occur in individuals without clear traditional risk factors. Traditional risk scores often focus on clinical parameters (e.g., age, surgery, immobilization) but fail to capture the underlying molecular heterogeneity of the disease. The emergence of high-throughput proteomic platforms—such as aptamer-based SomaScan and antibody-based Olink—offers a transformative opportunity to map the systemic molecular signatures that precede clinical VTE, potentially revealing novel therapeutic targets and predictive biomarkers.
Key Content
Large-Scale Proteomic Discovery: The ARIC, CHS, MESA, and HUNT Meta-Analysis
Recent high-quality longitudinal evidence has shifted the focus toward a broader biological understanding of VTE. In a seminal study involving 20,737 participants from four major cohorts (ARIC, CHS, MESA, and HUNT), researchers utilized the SomaScan platform to measure approximately 5,000 to 7,000 baseline plasma proteins. Following a maximum follow-up of 10 to 29 years, 1,371 incident noncancer VTE events were recorded.
The discovery meta-analysis identified 23 proteins that met the false discovery rate-adjusted significance threshold (P < 0.05). Most importantly, 15 of these proteins were previously unrelated to VTE pathophysiology. Among the top-tier proteins validated across cohorts were transgelin, sushi, von Willebrand factor type A, and TIMP4 (metalloproteinase inhibitor 4). These findings suggest that vascular structural integrity and extracellular matrix remodeling are as critical to VTE risk as the coagulation cascade itself.
Causal Inference through Mendelian Randomization
To distinguish between mere association and true causality, Mendelian randomization (MR) was applied to the 15 newly identified proteins. This technique uses genetic variants as instrumental variables to mimic randomized controlled trials. Significant evidence for causality was found for T-cell immunoglobulin and mucin domain-containing protein 4 (TIMD4). Furthermore, suggestive causal roles were identified for TIMP4 and Cystatin-C (CST3).
Notably, the direction of association in the MR analyses was opposite to that seen in the observational proteomics for TIMP4 and TIMD4. This divergence often indicates a biological feedback mechanism or a compensatory response in the pre-clinical state, where the body might upregulate certain inhibitors to counteract pro-thrombotic shifts. Such nuances are critical for drug development, where timing and dosage of potential inhibitors must be meticulously calculated.
Replication and Cross-Platform Validation
Scientific rigor was maintained through external replication using the UK Biobank (UKB) study, which utilized the Olink proteomics platform. Of the 16 top VTE proteins available on the Olink panel, 11 were successfully replicated in the UKB cohort. This cross-platform validation (Aptamer vs. Antibody-based detection) provides robust evidence that these markers are not technological artifacts but biological realities across diverse populations.
Clinical Context: Managing Risks in Aging Populations
The identification of these markers occurs within a broader clinical landscape where cardiovascular risk management is becoming increasingly complex. Discrete choice experiments have shown that while clinicians generally favor intensive systolic blood pressure targets (≤130 mm Hg) to reduce overall cardiovascular risk, their preference wanes significantly when dealing with older adults (age >80), patients with a history of falls, or those with moderate frailty. Integrating novel proteomic risk scores could help clinicians better identify which ‘high-risk’ older patients would benefit from aggressive intervention versus those where the risks of falls and adverse events outweigh the benefits.
Intersection with Metabolic and Endocrine Health
Secondary analyses of NCD (non-communicable disease) burdens, particularly the bidirectional relationship between thyroid dysfunction and diabetes, further complicate the VTE landscape. Hypothyroidism has been linked to accelerated atherosclerosis and potential shifts in the coagulative state, yet it remains a neglected priority in many global NCD agendas. The interplay between thyroid hormones and kidney function also modulates cardiovascular risk, suggesting that a multi-organ, proteomic-led approach is necessary for future personalized medicine.
Expert Commentary
The identification of markers like TIMP4 and TIMD4 signifies a shift from ‘hemostasis-centric’ research to ‘systems-biology-centric’ research in VTE. The role of extracellular matrix regulation and immune-vascular interaction suggests that inflammation and vascular senescence are potent drivers of thrombus formation in the venous system.
One significant controversy remains the clinical applicability of these 5,000-protein panels. While they are invaluable for discovery, the transition to clinical practice will likely require the development of smaller, targeted multiplex assays focusing on the 10-15 most predictive and causal markers. Furthermore, the MR data for CST3 showed consistent directionality, making it a particularly attractive candidate for future risk stratification tools. Clinicians should remain aware that as we move toward digital health monitoring and personalized proteomic profiles, the management of ‘clinical inertia’ will depend on our ability to translate these complex data points into actionable insights at the bedside.
Conclusion
We have entered a new era of VTE research where high-throughput proteomics has unveiled pathways involving immunity, vascular senescence, and ECM regulation. The validation of 23 proteins—15 of which were novel—across multiple longitudinal cohorts and different proteomic platforms provides a high degree of confidence in these findings. Future research must now focus on longitudinal changes in these protein levels and their integration with existing clinical risk scores to improve prevention and personalize treatment for this silent cardiovascular killer.
References
- Tang W, Li A, Austin TR, et al. Novel Plasma Proteomic Markers and Risk of Venous Thromboembolism. Circulation. 2026;153(11):810-825. PMID: 41693575.
- O’Hagan E, Livingstone A, Gadsden T, et al. Preferences for Antihypertensive Prescribing in Older Adults: A Discrete Choice Experiment. J Am Coll Cardiol. 2026. PMID: 41778956.
- The public health burden of diabetes mellitus and thyroid disease: twin epidemics. Nat Rev Endocrinol. 2026;22(4):242-254. PMID: 41507499.
- Nutritional adequacy of the EAT-Lancet diet: a Swedish population-based cohort study. Lancet Planet Health. 2027;101416. PMID: 41692025.