The Plasma Proteome and Risk of Future Venous Thromboembolism—Results from the HUNT Study in Thrombosis and Haemostasis

Starting with the discovery of the helical structure of DNA in 1953, and continuing through the completion of the Human Genome Project in 2003, medical science has focused on understanding the molecular structures and variations that lead to observed inheritance patterns and the pathogenesis of disease.[1] [2] Describing the structure and sequence of the human genome was a monumental advancement. However, the genome represents but one aspect of a complex biologic system in which genetically encoded information is differentially and dynamically expressed.[3] Efforts to more completely describe complex biological systems have led to the rapid development of other “-omics” including transcriptomics, proteomics, and metabolomics that link genotype to phenotype.[4] [5] Of these, proteomics, the large-scale systematic study of protein structure and function, holds perhaps the greatest potential for novel biomarker discovery. Proteins are the crucial signals between cells and tissues that drive cellular activity.[6]

Venous thromboembolism (VTE) is both a common and deadly acute illness, and there has been much interest in the clinical epidemiology, pathophysiology, diagnosis, and management of this condtion.[7] [8] These have included studies in various patient phenotypes and high-risk subgroups,[9] [10] [11] [12] focus on lifestyle and social factors,[13] and novel approaches for VTE prediction, such as machine learning.[14]

In this issue of the journal, the study by Brækkan et al represents the largest and most comprehensive proteomics study of incident VTE to date, including 294 VTE cases and 1,066 controls and 7,288 measured proteins. By comparison, other proteomic studies of both prevalent and incident VTE have had small sample sizes and analyzed a smaller number of proteins.[15] [16] [17] [18] [19] Brækkan et al identified five new candidate VTE biomarkers, namely, regulator of G protein signaling 3 (RGS3), collagen α-3 (VI):BPTI/Kunitz inhibitor, Histo-blood group ABO system transferase (BGAT), peroxidasin (PXDN), and human epididymis protein 4 (HE4), and reidentified two possible VTE biomarkers, namely, tumor necrosis factor soluble receptor II (TF sR-II) and coagulation factor VIII (FVIII).[20] [21] They also identified several enriched protein and biochemical pathways including the complement and coagulation cascades, MAPK signaling pathway, cytokine–cytokine receptor interaction, and PI3K-Akt signaling pathway.

In 2013, the National Heart, Lung, and Blood Institute (NHLBI), the National Cancer Institute (NCI), the American Association for Clinical Chemistry (AACC), and the US Food and Drug Administration (FDA) created a framework for the design of proteomic biomarker discovery and validation studies.[22] Since then, hundreds of analyses have been published using proteomics for biomarker discovery, mostly for chronic diseases like cancer.[23] [24] The COVID-19 pandemic did advance the use of proteomics in acute diagnostics and prognostics,[25] [26] [27] though proteomics remains relatively underutilized in the study of other acute illnesses, like VTE.[28] [29] [30] [31] [32] [33] Although the current study by Brækkan et al represents the largest proteomic study of VTE to date, it approaches VTE from an epidemiological perspective, as a disease with chronic risk factors that ultimately lead to a diagnosis. In this regard, VTE is analyzed in the same way one might analyze cancer. The current study must be distinguished from studies of acute, prevalent disease, which have heretofore been small.[15] [16] [19] The lack of proteomic research in acute illness likely results from the unique challenges of large-scale sample and data collection in acute care settings (like emergency departments where most VTEs are diagnosed), considering the necessary proteomic sample preparation and storage requirements.[34] These challenges will need to be overcome if we are to use modern -omic technology at the bedside of acutely symptomatic patients.

However, the need for more research into biomarkers that help clinicians diagnose and risk-stratify acute disease should not undermine the value of this important work. The current study not only identifies protein risk factors for VTE but also demonstrates how proteomics can improve our understanding of the pathogenesis of complex, acute illnesses like VTE. As such, this study provides a logical addition to large epidemiology studies that have identified genetic risk factors for VTE.[35] Importantly, the authors' unsupervised discovery approach is just a first, exploratory step, and like any good exploratory study, their results prompt us to ask many questions.

First, how should we interpret the results of this study given the diverse methodologies employed by existing proteomics platforms, and what can be done to harmonize proteomic results moving forward? The gold standard for identification and quantification of proteins is mass spectrometry (MS).[3] [36] Unfortunately, MS use remains limited due to both cost and time constraints.[6] [36] Accordingly, the use of high-throughput proteomic techniques has grown enormously. These techniques include antibody-based (e.g., Luminex and Simoa), aptamer (short oligonucleotide)-based (e.g., Somalogic), and proximity extension (oligonucleotide–antibody pair)-based (e.g., Olink) technologies.[37] Unfortunately, the results across platforms often differ, and do not necessarily correlate with the measurement of established clinical biomarkers. In the current study, for example, the Somalogic proteomics platform did not identify D-dimer and fibrinogen, two known protein biomarkers associated with VTE. Another study which directly compared 616 proteins measured by both Olink and Somalogic proteomics platforms found that only 236 (38%) were highly correlated.[36] This study found better specificity and higher phenotypic associations with the Olink platform and greater precision and analytic breadth with the Somalogic platform.[36] As Brækkan et al point out, high-throughput proteomic analyses are ideal for discovery but must be accompanied by robust validation prior to clinical implementation.[37]

Second, how do we envision these results guiding future research? One promising possibility for HUNT investigators is to leverage these data to perform multi-omic analyses that expand our understanding of VTE risk. VTE has an estimated 40 to 60% heritability,[35] [38] and large GWAS performed by the INVENT consortium (in which HUNT participates) has identified >135 novel loci associated with VTE.[35] These are in addition to well-known genetic risk factors (e.g., factor V Leiden, prothrombin G20210A) identified by earlier candidate-gene studies.[38] Proteomics has the potential to enhance genetic epidemiology by providing a functional layer of information on how genetic variations manifest at the protein level. It is estimated that after post-transcriptional changes there are approximately 100 protein isoforms of each gene.[6] Therefore, to truly understand the inheritable risk factors of a disease like VTE, one must consider not only genomics but also proteomics (and other -omics) and then use this knowledge to build a comprehensive understanding of the interplay of biological signals that lead to thromboembolic disease. It is only very recently with the significant advances in artificial intelligence that we are beginning to see the field of multi-omics spread beyond its infancy.[39] [40] If carefully considered, the integration of -omic science and machine learning technology holds great promise.

Finally, it is important to consider what these results do not tell us. The current study's approach was focused on identifying protein risk factors for incident VTE occurring up to 5 years after blood draw. It does not tell us what proteomic changes (i.e., diagnostic biomarkers) might be present at the time of acute, prevalent VTE, nor does it tell us about events or physiological changes that may have occurred between the study blood draw and VTE diagnosis. Considering the dynamic nature of protein expression, these considerations are important. One consequence is that we are unable to determine whether the proteins identified by Brækkan et al are, in fact, markers of incident VTE or, alternatively, markers of an intermediate phenotype like cancer (or another disease) that may have developed after cohort inception but prior to VTE diagnosis. The authors attempted to mitigate this possibility by excluding patients with prevalent cancer from their analysis, though 19% of initially cancer-free patients did develop cancer prior to VTE diagnosis. It is reassuring that the proteins identified by Brækkan et al fit well within our understanding of VTE as a disease of both coagulation and inflammation. Also, if a goal of this study is to aid in VTE risk prediction, a direct pathogenic connection between a protein and VTE may not be necessary. Nonetheless, future studies should involve detailed exploration of the pathophysiological role the identified proteins play in the development of VTE.

Overall, this work by Brækkan et al is an excellent step forward that demonstrates the feasibility of proteomic measurement in VTE and seems poised to advance our understanding of the epidemiology and pathophysiology of VTE. Their study raises key questions about clinical proteomic methodology, the possibility of new epidemiologic risk factors for VTE, and the promise of proteomics for acute VTE diagnosis. The greatest value in this study will likely be in the research that it inspires and the work that follows.

Publication History

Article published online:
25 April 2025

• References

• 1 Mersha TB. From Mendel to multi-omics: shifting paradigms. Eur J Hum Genet 2024; 32 (02) 139-142
• 2 Lander ES, Linton LM, Birren B. et al; International Human Genome Sequencing Consortium. Initial sequencing and analysis of the human genome. Nature 2001; 409 (6822) 860-921
• 3 Roberts DS, Loo JA, Tsybin YO. et al. Top-down proteomics. Nat Rev Methods Primers 2024; 4 (01) 38
• 4 Dai X, Shen L. Advances and trends in omics technology development. Front Med (Lausanne) 2022; 9: 911861
• 5 Dayon L, Cominetti O, Affolter M. Proteomics of human biological fluids for biomarker discoveries: technical advances and recent applications. Expert Rev Proteomics 2022; 19 (02) 131-151
• 6 Timp W, Timp G. Beyond mass spectrometry, the next step in proteomics. Sci Adv 2020; 6 (02) eaax8978
• 7 Silverstein MD, Heit JA, Mohr DN, Petterson TM, O'Fallon WM, Melton III LJ. Trends in the incidence of deep vein thrombosis and pulmonary embolism: a 25-year population-based study. Arch Intern Med 1998; 158 (06) 585-593
• 8 Roy DC, Wang TF, Lun R. et al. Circulating blood biomarkers and risk of venous thromboembolism in cancer patients: a systematic review and meta-analysis. Thromb Haemost 2024; 124 (12) 1117-1133
• 9 De Corte P, Milhoranca I, Mechsner S, Oberg AS, Kurth T, Heinemann K. Unravelling the causal relationship between endometriosis and the risk for developing venous thromboembolism: a pooled analysis. Thromb Haemost 2025; 125 (04) 385-394
• 10 Huang Z, Zheng Z, Pang L. et al. The association between obstructive sleep apnea and venous thromboembolism: a bidirectional two-sample Mendelian randomization study. Thromb Haemost 2024; 124 (11) 1061-1074
• 11 Lanting VR, Takada T, Bosch FTM. et al. Risk of recurrent venous thromboembolism in patients with cancer: an individual patient data meta-analysis and development of a prediction model. Thromb Haemost 2024; . Epub ahead of print
• 12 Granat LM, Li H, Ondeck M. et al. Venous thromboembolism post-allogeneic hematopoietic cell transplant: risk factors, incidence, and outcomes. Thromb Haemost 2025; 125 (02) 155-162
• 13 Liang B, Zha L, Peng B. et al. Association of cardiovascular health with the incidence of venous thromboembolism: a prospective study of 275,149 participants from the UK Biobank. Thromb Haemost 2024; 124 (10) 948-957
• 14 Danilatou V, Dimopoulos D, Kostoulas T, Douketis J. Machine learning-based predictive models for patients with venous thromboembolism: a systematic review. Thromb Haemost 2024; 124 (11) 1040-1052
• 15 Gade IL, Riddersholm SJ, Stilling-Vinther T, Brøndum RF, Bennike TB, Honoré B. A clinical proteomics study of exhaled breath condensate and biomarkers for pulmonary embolism. J Breath Res 2023; 18 (01) 016007
• 16 Granholm F, Bylund D, Shevchenko G, Lind SB, Henriksson AE. A feasibility study on the identification of potential biomarkers in pulmonary embolism using proteomic analysis. Clin Appl Thromb Hemost 2022 ;28:10760296221074347
• 17 Jensen SB, Latysheva N, Hindberg K, Ueland T. Plasma lipopolysaccharide-binding protein is a biomarker for future venous thromboembolism: results from discovery and validation studies. J Intern Med 2022; 292 (03) 523-535
• 18 Ten Cate V, Prochaska JH, Schulz A. et al. Protein expression profiling suggests relevance of noncanonical pathways in isolated pulmonary embolism. Blood 2021; 137 (19) 2681-2693
• 19 Ganesh SK, Sharma Y, Dayhoff J. et al. Detection of venous thromboembolism by proteomic serum biomarkers. PLoS One 2007; 2 (06) e544
• 20 Rietveld IM, Lijfering WM, le Cessie S. et al. High levels of coagulation factors and venous thrombosis risk: strongest association for factor VIII and von Willebrand factor. J Thromb Haemost 2019; 17 (01) 99-109
• 21 Yuan S, Titova OE, Zhang K. et al. Plasma protein and venous thromboembolism: prospective cohort and mendelian randomisation analyses. Br J Haematol 2023; 201 (04) 783-792
• 22 Skates SJ, Gillette MA, LaBaer J. et al. Statistical design for biospecimen cohort size in proteomics-based biomarker discovery and verification studies. J Proteome Res 2013; 12 (12) 5383-5394
• 23 Kwon YW, Jo HS, Bae S. et al. Application of proteomics in cancer: recent trends and approaches for biomarkers discovery. Front Med (Lausanne) 2021; 8: 747333
• 24 Xiao GG, Recker RR, Deng HW. Recent advances in proteomics and cancer biomarker discovery. Clin Med Oncol 2008; 2: 63-72
• 25 Suran M. After the genome—a brief history of proteomics. JAMA 2022; 328 (12) 1168-1169
• 26 Demichev V, Tober-Lau P, Nazarenko T. et al; PA-COVID-19 Study group. A proteomic survival predictor for COVID-19 patients in intensive care. PLOS Digit Health 2022; 1 (01) e0000007
• 27 Zhong W, Altay O, Arif M. et al. Next generation plasma proteome profiling of COVID-19 patients with mild to moderate symptoms. EBioMedicine 2021; 74: 103723
• 28 Forrer A, Schoenrath F, Torzewski M. et al. Novel blood biomarkers for a diagnostic workup of acute aortic dissection. Diagnostics (Basel) 2021; 11 (04) 615
• 29 Marto JP, Carvalho ASG. G Mollet I, et al. Proteomics to identify new blood biomarkers for diagnosing patients with acute stroke. J Am Heart Assoc 2023; 12 (22) e030021
• 30 Gustafsson S, Lampa E, Jensevik Eriksson K. et al. Markers of imminent myocardial infarction. Nat Cardiovasc Res 2024; 3 (02) 130-139
• 31 Oved K, Cohen A, Boico O. et al. A novel host-proteome signature for distinguishing between acute bacterial and viral infections. PLoS One 2015; 10 (03) e0120012
• 32 D'Onofrio V, Heylen D, Pusparum M. et al. A prospective observational cohort study to identify inflammatory biomarkers for the diagnosis and prognosis of patients with sepsis. J Intensive Care 2022; 10 (01) 13
• 33 Strehlow M, Gisondi MA, Caretta-Weyer H. et al. 2023 Society for Academic Emergency Medicine Consensus Conference on Precision Emergency Medicine: development of a policy-relevant, patient-centered research agenda. Acad Emerg Med 2024; 31 (08) 805-816
• 34 Kavallaris M, Marshall GM. Proteomics and disease: opportunities and challenges. Med J Aust 2005; 182 (11) 575-579
• 35 Thibord F, Klarin D, Brody JA. et al; Global Biobank Meta-Analysis Initiative; Estonian Biobank Research Team; 23andMe Research Team; Biobank Japan; CHARGE Hemostasis Working Group. Cross-ancestry investigation of venous thromboembolism genomic predictors. Circulation 2022; 146 (16) 1225-1242
• 36 Katz DH, Robbins JM, Deng S. et al. Proteomic profiling platforms head to head: Leveraging genetics and clinical traits to compare aptamer- and antibody-based methods. Sci Adv 2022; 8 (33) eabm5164
• 37 Cui M, Cheng C, Zhang L. High-throughput proteomics: a methodological mini-review. Lab Invest 2022; 102 (11) 1170-1181
• 38 Zöller B, Svensson PJ, Dahlbäck B, Lind-Hallden C, Hallden C, Elf J. Genetic risk factors for venous thromboembolism. Expert Rev Hematol 2020; 13 (09) 971-981
• 39 Flores JE, Claborne DM, Weller ZD, Webb-Robertson BM, Waters KM, Bramer LM. Missing data in multi-omics integration: recent advances through artificial intelligence. Front Artif Intell 2023; 6: 1098308
• 40 Kang M, Ko E, Mersha TB. A roadmap for multi-omics data integration using deep learning. Brief Bioinform 2022; 23 (01) bbab454