Venous Thromboembolism: Diagnostic Advances and Unaddressed Challenges in Management

Abstract

Purpose of Review:

This review summarizes recent advances in developing targeted diagnostics for venous thromboembolism (VTE) and unaddressed knowledge gaps in patient management. Without addressing these critical data needs, the morbidity in VTE patients will persist.

Recent Findings:

Recent studies investigating plasma protein profiles in VTE patients have identified key diagnostic targets to address the currently unmet need for low-cost, confirmatory, point-of-care VTE diagnostics. These studies and a growing body of evidence from animal model studies have revealed the importance of inflammatory and vascular pathology in driving VTE, which are currently unaddressed targets for VTE therapy. To enhance the translation of preclinical animal studies, clinical quantification of thrombus burden and comparative component analyses between modeled VTE and clinical VTE are necessary.

Summary:

Lead candidates from protein profiling of VTE patients’ plasma offer a promising outlook in developing low cost, confirmatory, point-of-care testing for VTE. Additionally, addressing the critical knowledge gap of quantitatively measuring clinical thrombi will allow for an array of benefits in VTE management and informing the translatability of experimental therapeutics.

I. Introduction

Pathologic venous thromboembolism (VTE) is a lesion where thrombi form in the absence of bleeding, often in the deep veins of the lower extremities. These pathological thrombi may embolize to the pulmonary circuit, causing significant morbidity and mortality through obstruction of circulation or chronic thromboembolic pulmonary hypertension. Consequently, VTE encompasses deep vein thrombosis (DVT) and pulmonary embolism (PE) arising from DVT [1].

Venous Thrombus Structure

Venous thrombi are primarily composed of fibrin and red blood cells with relatively few platelets in contrast to arterial thrombi that have a significantly higher proportion of platelets[2]; however, all components have demonstrated important roles in regulating the thrombus size and resolution rates in animal models[3–7]. The physical measurements and weights of excised thrombi from animal models of DVT provide a measurement of thrombus burden; however, thrombus measurements are not estimated in clinical practice creating a knowledge gap in correlating animal model thrombus burden to clinical thrombus burden. Furthermore, few studies have directly compared animal model thrombi to clinical thrombi in terms of size and components [8, 9]. Conducting such comparative studies is crucial to enhancing translatability of experimental therapeutics.

Epidemiology of Venous Thromboembolism

Although estimates vary, the consensus on VTE strongly indicates a high global prevalence[10]. A retrospective analysis of a nationwide Swedish registry denoted a skewed distribution for VTE with respect to advancing age[11]. Analysis of the cohorts in the Cardiovascular Health Study and the Atherosclerosis Risk in Communities study revealed an 8% remaining lifetime risk of VTE in adults aged 45–64 years[12] with risk and incidence rates for VTE increasing significantly in subsequent decades[13–16]. While these distributions clearly demonstrate which age groups bear the major burden of VTE[11, 16, 17], etiology is not unifactorial. Multiple risk factors are often identified in patients presenting with VTE in the clinic. This makes VTE a difficult lesion to prevent and treat, given multiple formation and resolution mechanisms. With impaired venous resolution, DVTs may progress to morbid sequelae like postthrombotic syndrome (PTS), where chronic venous disease predisposes patients to chronic leg pain, dermatitis, ulcer formation, and recurrent VTE.

Virchow’s Triad: Provoking Factors for Venous Thromboembolism

VTE is a complex disorder with multiple etiological mechanisms presenting in the clinic[18] and correspondingly modeled in animal studies [19]. Although comprehension of the underlying mechanisms continues to be grounded in Virchow’s triad[20], significant changes to the original three components now offer a better understanding of the clinical contexts leading to VTE. The original triad stated that three key events cause VTE: blood flow stasis, vein wall injury, and a hypercoagulable state. Researchers and clinicians now appreciate and clinically incorporate the impact of significant risk factors not in the original triad ranging from malignancy to obesity and diabetes[21]. However, Virchow’s triad provides an important scaffold in characterizing VTE pathology and is used here to highlight the contribution of vascular cellular pathology to VTE development.

Stasis of blood flow is crucial to replicating the pathology of clinical VTE in rodent models as the degree of obstruction-of-flow and availability of collateral flow impacts the incidence and size of the thrombus[22, 23]. However, stasis of blood flow is a broad characterization and may be more accurately recharacterized as pathologic disruption in the local hemodynamic profile, increasing the risk of VTE. These disruptions in the hemodynamic profile are often a function of the geometry of the vessel, further complicated by the presence of venous valves[24]. Valve failure leads to venous insufficiency causing retrograde blood flow[25], the prevalence of which increases with advancing age[26] and is a demonstrated risk factor for VTE[27]. Stasis of blood flow may cause local hypoxia in the vein wall potentiating VTE through increased inflammation [28]. Additionally, a growing body of work has demonstrated the effect of disturbed flow profiles on venous endothelial (EC) and smooth muscle cell (SMC) phenotypes potentially contributing to thrombus formation[29–31]. These disturbed flow profiles may condition the vasculature over extended periods of time potentiating chronic venous disease. Therefore there is a need for early identification, monitoring, and therapeutic targeting of chronic venous cellular pathology in addition to reducing hematologic thrombus burden or residual thrombus, to alleviate the chronic morbidity and mortality from VTE and its sequelae.

Vein wall injury directly implicates the importance of vascular and inflammatory pathology in VTE. However, this original characterization in Virchow’s triad is too broad. Vascular wall injury can be recharacterized as phenotypic dysfunction of the vascular cell components[32], which conditions and predisposes specific venous regions to thrombotic events. These specific venous regions can result from disruptions in blood flow, which conditions ECs and SMCs, as well as resident inflammatory cells[33]*, toward an activated, thrombotic, inflammatory, and fibrotic phenotype. Additionally, an ageing vasculature undergoes phenotypic changes in every component from thromboinflammatory ECs to stiffening connective tissues[34–36]. While in atherosclerosis the target of treatment is the vascular plaque until rupture culminates in a thrombotic event, VTE is not treated in an analogous manner, and physicians currently do not have the tools to inform them of venous vascular health in the ways which animal model studies indicate may be important to VTE pathology and prognosis.

Current treatment for VTE best addresses the third tenet of Virchow’s triad, hypercoagulability. Anticoagulation is the primary therapy for confirmed VTE and direct oral anticoagulants (DOACs) have demonstrated increased safety and comparable efficacy[37, 38] over the previous heparin and vitamin K antagonist regimen[39]. Providers follow treatment guidelines that rely on the presence or absence of provoking factors, which are inconsistently classified[1]. A current treatment debate revolves around whether guidelines should instead focus on the risk profile of the patient to develop VTE sequelae, such as PTS and recurrent VTE[40, 41]. Importantly, even aggressive interventions such as the mechanical removal of the thrombus did not alleviate the risk of PTS in the ATTRACT trial[42]. Opportunities remain for innovative targeting of the underlying vascular and inflammatory VTE mechanisms to improve therapeutic outcomes[43].

II. Diagnostic Development

VTE is difficult to diagnose for three key reasons: (i) Patients present with non-specific signs and symptoms[44] with a large portion of DVT and PE cases being silent[45]; (ii) The primary laboratory test available for VTE measures thrombus degradation products known as D-Dimers and has a high sensitivity but low specificity, requiring confirmatory imaging visualization for VTE diagnosis [46]; (iii) While imaging is required for confirmatory diagnosis, imaging is not advisable in all suspected VTE cases due to the burden of radiation, cost, and infrastructure of some modalities (e.g. equipment and skilled technologists). Additionally, ultrasound, an imaging modality without those burdens, has high rates of missed diagnoses due to low true positive and high false negative rates[47]. In settings where ultrasound is not available, has limited use,[48] or is inconclusive, secondary imaging modalities such as CT or MRI venography can provide 3D reconstructed images to estimate thrombus burden; however, standardized protocols for quantification are lacking. Pretest probability assessments such as the Wells and Caprini Scores help stratify risk and guide VTE diagnosis[49, 50]; however, they are inconsistently applied and can still lead to missed diagnoses. New diagnostic biomarkers with high sensitivity and specificity may prove helpful in supporting diagnosis, risk stratification, and guide treatment [51].

Recently, protein profiling of patients’ plasma has gained attention for VTE biomarker identification. This approach focuses on the minimally invasive blood draw, which is readily available to diagnosing physicians. A study by Iglesias et. al revealed complement factor H related 5 (CFHR5) protein was associated with VTE in multiple cohorts, demonstrating its usefulness as a potential diagnostic biomarker[52]**. Comparing a prospective cohort of suspected VTE patients and a confirmed case cohort of VTE patients who had already undergone primary anticoagulation therapy, they identified potential biomarker candidates by screening 756 antibodies targeting 408 proteins against patient plasma. This screen identified an antibody predicted to target sulfatase 1 (SULF1), which is responsible for the selective removal of 6-O sulfate groups on heparan sulfate, the most abundant vascular EC glycosaminoglycan[53, 54], as strongly correlated with VTE. In order to confirm that the antibody did indeed bind to its predicted target protein, they performed immunocapture mass spectrometry of patient plasma where CFHR5 produced a significantly larger signal, while not ruling out SULF1 as concurrently bound. Additionally, CFHR5 was associated with VTE independent of D-Dimer or C-reactive proteins; however, it was not significantly associated with recurrent VTE events. It is important to note that multiple studies profiling the proteins present in VTE patients’ plasma have identified biomarker candidates (Table 1), but consensus on any single biomarker is lacking[55–63]. In subsequent experiments performed by Iglesias et. al, incubation of platelet rich plasma with exogenous, recombinant CFHR5 was observed to enhance platelet activation and degranulation[52]**; however, it did not have a thrombotic effect on ECs when exogenous, recombinant CFHR5 was added to human umbilical vein ECs (HUVECs) and tested for thrombin generation and an array of inflammatory markers.

Table 1:

Leading Candidates for Diagnostic Biomarkers from Protein Profiling Studies of VTE Patient Plasma

Biomarker	Function	Reference
Complement Factor H Related Protein 5	Promotes complement activation by antagonizing complement factor H [77]	Iglesias et. al [52]**
Sulfatase 1	Selectively removes 6-O sulfates from heparan sulfate [53]
Parkinson’s disease protein 7 (DJ1)	Loss of function mutations in DJ-1 are associated with mitochondrial dysfunction and Parkinson’s disease among multiple other phenotypes [78]	Jensen et. al [55]
Protein Z	Vitamin K dependent cofactor for protein-z-dependent-protease to inactivate coagulation Factor Xa [79]
Transthyretin	Transports thyroxine and retinol-binding-proteins; transthyretin amyloidosis is associated with an increased risk of VTE [80]
Glutathione Peroxidase 3	Catalyzes the reduction of hydrogen peroxide through glutathione [81]	Bruzelius et. al [56] and Memon et. al[60]
Human Immunodeficiency Virus Type I Enhancer Binding Protein 1	Anti-inflammatory negative regulator of NF-κB [82]
Platelet Derived Growth Factor β	Demonstrates paracrine growth factor activity and chemoattractant for vascular smooth muscle cells and pericytes [83]
Von Willebrand Factor	Stabilization of coagulation Factor VIII and mediating platelet adhesion to subendothelial collagen [84]
Growth Arrest Specific 6	Vitamin K dependent TAM receptor ligand; Implicated in platelet and leukocyte recruitment, platelet aggregation, and endothelial activation [85]	Blostein et. al [57] and Schnegg-Kaufmann et. al [58]
Integrins subunits β1, β2, and β3	Extracellular matrix and inter-cellular adhesion receptors [86]	Song et. al [59]
Bleomycin Hydrolase	Regulates major histocompatibility complex class I ligand generation [87]	Memon et. al [60] and Khorana et. al [62]
Soluble Osteopontin	Macrophage and T cell chemoattractant mediating chronic inflammation [88]
P-Selectin	Mediates leukocyte and platelet adhesion to the endothelium [89]
Suppression of Tumorigenicity 2 Protein	Activates cardiac fibroblasts and induces collagen synthesis through neuropilin 1 [90]
Tissue Factor Pathway Inhibitor	Inhibits tissue factor and coagulation factor VIIa complexes as well as prothrombinase [91]
Transferrin Receptor Protein 1	Mediates transferrin binding and cellular iron uptake [92]
Glial Cell Line–Derived Neurotrophic Growth Factor	Neurotrophic growth factor for dopaminergic neurons [93]	Ten Cate et. al [61]
Interferon-γ	Cytokine with immunomodulatory and antitumor roles [94]
Interleukin-15 Receptor Subunit α	High affinity receptor for interleukin 15 [95]
Peptidyl Arginine Deiminase Type-2	Citrullinates fibrin, altering thrombus structure [96]
Polypeptide N-Acetylgalactosaminyltransferase 3	Catalyzes O-sulfation to initiate glycosaminoglycan chain formation [97]
Soluble C-Reactive Protein	Acute inflammatory phase reactant [98]	Khorana et. al [62]
Growth Hormone	Reduces von Willebrand factor – platelet interactions [99]
Interleukin-1 receptor type 1	Immunomodulatory cytokine receptor; inhibition of this receptor led to decreased venous thrombosis in tumor bearing mice [100]
Interleukin 8	Immunomodulatory cytokine demonstrated to accelerate rodent thrombus resolution [101]
Monocyte Chemotactic Protein 4	Induces monocyte tissue factor expression [102]
Stromal Cell–Derived Factor-1	Regulator of endothelial progenitor cell development into endothelial cells [103]
Thyroid-Stimulating Hormone	Primary stimulus for thyroid hormone production; abnormal levels of thyroid stimulating hormone, both low and high, have been associated with VTE [104]

Further studies are needed to investigate the role of CFHR5 on VTE pathology and EC phenotype under physiologic and pathologic flow conditions. These studies would help further elucidate the role of SULF1 and its target, EC heparan sulfate, the most abundantly expressed vascular glycosaminoglycan[64], to VTE pathology. Additionally, there is also a paucity of investigation of the role of heparan sulfate in murine models of DVT and PE, which may be warranted given clinical indications of their effect in treating the pathology. Notably, Sulodexide, which contains heparan sulfate[65], has been demonstrated in randomized controlled trials to have beneficial adjunctive effects to primary VTE therapy by improving odds against all-cause mortality, cardiovascular mortality, and DVT without increasing bleeding risk[66, 67]. However, Sulodexide is not FDA approved.

Inflammatory and vascular biomarkers have also been investigated and reviewed to increase the accurate diagnosis of VTE. A recent study using mendelian randomization analysis of VTE patients, further stratified by cardiovascular comorbidities and modifiable risk factors, identified both VTE associated proteins and pharmaceutical targets. One of their most promising candidates was coagulation factor XI, which was found to mediate independent associations between obesity indicators, smoking, and insomnia with VTE[63]. Audu et. al summarize P-selectin, fibrin monomer complex, factor VIII, ICAM-1, and circulating DNA in conjunction to D-dimer as potential novel biomarkers in aiding accurate diagnosis of VTE[51]. However, to date no clinical grade biomarker is available to provide robust increases in accuracy of VTE diagnosis, highlighting further investigative need in this area.

III. Monitoring and Therapeutics Development

Monitoring VTE progression effectively to predict prognosis is a pressing clinical challenge. The DACUS study supported the importance of close monitoring of residual thrombus and vascular patency to determine the duration of anticoagulation after first VTE[68]. The multinational REVERSE trial did not find value in monitoring of residual venous obstruction to predict recurrent VTE[69]. However, other studies have indicated utility in residual thrombus monitoring as an important measure to predict VTE recurrence[70] and to determine morbidity from PTS[71]. Thus, there is still debate around the utility of residual venous thrombus monitoring in predicting incidence of sequelae such as PTS[72].

Clinical DVT is replicated in animal models using the stasis or stenosis methods, in which inferior vena cava (IVC) ligation or constriction, respectively, are used in combination with mechanical EC injury and collateral cauterization[19, 23]. These models provide insight into the mechanisms causing the incidence and burden of VTE and its sequelae. Other models, such as ferric chloride or electrolytic models[73], are more useful in studying initiation, embolization, and recurrence of VTE. The primary output of the venous stasis or stenosis models is the incidence and weight of the thrombi followed by further characterization of the thrombotic and vein wall tissue in response to genetic or pharmacologic treatments. Current clinical guidelines qualitatively characterize DVT as proximal or distal, describe extent of occlusion, identify collaterals, and describe the acute or chronic nature of the lesion. Clinical outcomes are measured by all-cause mortality, recurrent VTE, and PTS incidence. Unfortunately, there is no animal model of long term sequelae like PTS, disconnecting clinical VTE outcomes and experimental model outcomes. Quantitative measurements of clinical venous thrombi can reduce subjectivity in evaluation of the thrombi and support data driven decision making with quantification of initial thrombus size and subsequent thrombus regression in response to anticoagulation therapy. These measurements would enable comparison between patients receiving different therapeutic regimens or varying clinical contexts. While limitations in such an approach exist due to variations in estimates and lack of standardization and complexity of thrombus anatomy, estimates of thrombus dimensions (e.g. length and volume) would aid in clinically correlating laboratory model results by providing an analogous measure of clinical thrombus burden (Figure 1).

Figure 1:

Original Image: Clinical estimates of thrombus burden from measurement of thrombi through ultrasound or manually and externally are a critical knowledge gap needed to correlate animal model thrombus burden to clinical thrombus burden. Image created with Biorender.com.

Animal models have shed light on the role of the vasculature and inflammatory system in mediating both resolution of the thrombus and recanalization of the thrombus obstructed vessel wall, as well as the impact of these processes on long term vascular health, including predisposition to PTS. DeRoo et. al uncovered diverse cell populations present in the vein wall 24 hours after IVC ligation in the stasis DVT model compared to sham surgical mice by performing single cell RNA sequencing on IVC tissue[33]*. In the vein wall of DVT induced mice, they identified early neutrophil presence and a lack of fibrotic phenotype, which they attributed to acute inflammatory and thrombotic events.

Inflammatory mediators of VTE formation and resolution have been highlighted in recent animal model studies as well. Tang et. al performed mass-spectrometry analysis of the vein wall in IVC-stenosis induced DVT in rats, implicating EC SIRT1 in ameliorating thrombus burden through NF-κB deacetylation[74]. Lapointe et. al demonstrate the importance of mast cell protease 4 protein in 2 murine models [75], following on from Ponomaryov et. al demonstrating the importance of mast cell and EC communication in DVT formation[76]. These studies strongly implicate the cross-talk of vascular, hematologic, and inflammatory systems in predisposing, forming, and resolving the thrombus; and are recent characterizations in a growing body of work investigating the vascular and inflammatory mediators of VTE. Unfortunately, clinical tools targeting these underlying mechanisms have yet to be developed, due to critical knowledge gaps in comparative thrombus analyses between model thrombi and clinical thrombi as well as analogous measures of thrombus burden.

VI. Future Directions

Emerging evidence underscores the role of EC and inflammatory cell components in the formation, progression, and resolution of VTE. Translational studies investigating the role of adjunctive therapy targeting vascular and inflammatory mechanisms in addition to anticoagulation for VTE and its sequelae are warranted to address whether adjunctive therapy will improve long term clinical outcomes for VTE patients.

VII. Conclusion

Diagnostic development and adjunctive therapy to first-line DOACs remain key areas of translational investigation to improve long term outcomes for VTE patients. Measurements of thrombi to estimate clinical thrombus burden is a critical knowledge gap that can be addressed in the near term to allow for correlation of animal model results to clinical thrombus burden.

Key Points:

• Protein profiling of patient plasma has generated candidates for biomarker-based diagnosis of venous thromboembolism, providing an alternative or adjunctive tool to preclinical test scores, the D-dimer assay, and imaging.

• Quantification of clinical thrombus burden offers various benefits ranging from reduced subjectivity in patient evaluation to allowing for an analogous measure and output to correlate animal model studies to clinical thrombus progression.

• Specifically targeting the vascular and immuno-inflammatory systems remain opportunistic areas in treating the underlying pathological mechanisms and long-term sequelae of venous thromboembolism.