Deep Learning for Accurate Segmentation of Venous Thrombus from Black-Blood Magnetic Resonance Images: A Multicenter Study

Objective. Deep vein thrombosis (DVT) is the third-largest cardiovascular disease, and accurate segmentation of venous thrombus from the black-blood magnetic resonance (MR) images can provide additional information for personalized DVT treatment planning. Therefore, a deep learning network is proposed to automatically segment venous thrombus with high accuracy and reliability. Methods. In order to train, test, and external test the developed network, total images of 110 subjects are obtained from three different centers with two different black-blood MR techniques (i.e., DANTE-SPACE and DANTE-FLASH). Two experienced radiologists manually contoured each venous thrombus, followed by reediting, to create the ground truth. 5-fold cross-validation strategy is applied for training and testing. The segmentation performance is measured on pixel and vessel segment levels. For the pixel level, the dice similarity coefficient (DSC), average Hausdorff distance (AHD), and absolute volume difference (AVD) of segmented thrombus are calculated. For the vessel segment level, the sensitivity (SE), specificity (SP), accuracy (ACC), and positive and negative predictive values (PPV and NPV) are used. Results. The proposed network generates segmentation results in good agreement with the ground truth. Based on the pixel level, the proposed network achieves excellent results on testing and the other two external testing sets, DSC are 0.76, 0.76, and 0.73, AHD (mm) are 4.11, 6.45, and 6.49, and AVD are 0.16, 0.18, and 0.22. On the vessel segment level, SE are 0.95, 0.93, and 0.81, SP are 0.97, 0.92, and 0.97, ACC are 0.96, 0.94, and 0.95, PPV are 0.97, 0.82, and 0.96, and NPV are 0.97, 0.96, and 0.94. Conclusions. The proposed deep learning network is effective and stable for fully automatic segmentation of venous thrombus on black blood MR images.

Two Stage Intravital Imaging Mouse Model to Assess Venous Thromboembolism in Sickle Cell Disease

Sickle cell disease (SCD) patients have an increased risk of venous thromboembolism (VTE). Population-based studies demonstrated that VTE has a cumulative incidence rate of approximately 25% in adult SCD patients and is associated with higher risk of mortality. VTE in SCD is most commonly manifested as deep vein thrombosis (DVT) with associated pulmonary embolism (PE). Although autopsy studies have regularly discovered pulmonary thromboembolic lesions in SCD patients, the pathophysiology of VTE in SCD remains largely unknown mostly due to the lack of relevant animal VTE model. Understanding the mechanisms that promote VTE in SCD is imperative to identify its prevention and treatment measures. To elucidate the cellular, molecular and biophysical mechanisms of VTE in SCD we developed an innovative two stage intravital imaging analysis experimental model in SCD mice. DVT in SCD mice was induced by surgical ligation of femoral vein. Venous thrombus formation was visualized using intravital multi-photon-excitation (MPE) microscopy. Venous thrombus took the form of a large mass of elliptical shape which extended in the long-axis direction of the femoral vein. It was composed of fibrin, erythrocytes and sparse platelets. Over time, thrombus was infiltrated by migrating neutrophils. To trigger acute PE, femoral vein ligation was removed and the venous thrombus was observed to spontaneously detach and travel to the SCD mouse lung. This method allowed real time visualization of acute PE in vivo using MPE microscopy of intact lung in live breathing mice. Acute PE involved embolization of the pulmonary arterioles and the arteriolar bottle-necks located at the junction of pulmonary arterioles and capillaries. The embolization of arteriolar circulation led to loss of blood flow in the arterioles and the down-stream capillaries. Herein we introduce an intravital microscopy approach to probe VTE in SCD live mouse. Our model has potential application in investigating the molecular determinants of VTE associated with SCD as well as evaluating efficacy of new antithrombotic drugs. Disclosures Sundd: Bayer: Research Funding; CSL Behring Inc: Research Funding; Novartis: Membership on an entity's Board of Directors or advisory committees, Research Funding.

Platelet GPVI (Glycoprotein VI) and Thrombotic Complications in the Venous System

The immunoglobulin receptor GPVI (glycoprotein VI) is selectively expressed on megakaryocytes and platelets and is currently recognized as a receptor for not only collagen but also a variety of plasma and vascular proteins, including fibrin, fibrinogen, laminin, fibronectin, and galectin-3. Deficiency of GPVI is protective in mouse models of experimental thrombosis, pulmonary thromboembolism as well as in thromboinflammation, suggesting a role of GPVI in arterial and venous thrombus formation. In humans, platelet GPVI deficiency is associated with a mild bleeding phenotype, whereas a common variant rs1613662 in the GP6 gene is considered a risk factor for venous thromboembolism. However, preclinical studies on the inhibition of GPVI-ligand interactions are focused on arterial thrombotic complications. In this review we discuss the emerging evidence for GPVI in venous thrombus formation and leukocyte-dependent thromboinflammation, extending to venous thromboembolism, pulmonary thromboembolism, and cancer metastasis. We also recapitulate indications for circulating soluble GPVI as a biomarker of thrombosis-related complications. Collectively, we conclude that the current evidence suggests that platelet GPVI is also a suitable cotarget in the prevention of venous thrombosis due to its role in thrombus consolidation and platelet-leukocyte complex formation.

Matrine Impairs Platelet Function and Thrombosis and Inhibits ROS Production

Matrine is a naturally occurring alkaloid and possesses a wide range of pharmacological properties, such as anti-cancer, anti-oxidant, anti-inflammatory effects. However, whether it affects platelet function and thrombosis remains unclear. This study aims to evaluate the effect of matrine on platelet function and thrombus formation. Human platelets were treated with matrine (0–1 mg/ml) for 1 h at 37°C followed by measuring platelet aggregation, granule secretion, receptor expression by flow cytometry, spreading and clot retraction. In addition, matrine (10 mg/kg) was injected intraperitoneally into mice to measure tail bleeding time, arterial and venous thrombus formation. Matrine dose-dependently inhibited platelet aggregation and ATP release in response to either collagen-related peptide (Collagen-related peptide, 0.1 μg/ml) or thrombin (0.04 U/mL) stimulation without altering the expression of P-selectin, glycoprotein Ibα, GPVI, or αIIbβ3. In addition, matrine-treated platelets presented significantly decreased spreading on fibrinogen or collagen and clot retraction along with reduced phosphorylation of c-Src. Moreover, matrine administration significantly impaired the in vivo hemostatic function of platelets, arterial and venous thrombus formation. Furthermore, in platelets stimulated with CRP or thrombin, matrine significantly reduced Reactive oxygen species generation, inhibited the phosphorylation level of ERK1/2 (Thr202/Tyr204), p38 (Thr180/Tyr182) and AKT (Thr308/Ser473) as well as increased VASP phosphorylation (Ser239) and intracellular cGMP level. In conclusion, matrine inhibits platelet function, arterial and venous thrombosis, possibly involving inhibition of ROS generation, suggesting that matrine might be used as an antiplatelet agent for treating thrombotic or cardiovascular diseases.

Role of angiopoietin-2 in venous thrombus resolution and chronic thromboembolic disease

Defective angiogenesis, incomplete thrombus revascularisation and fibrosis are considered critical pathomechanisms of chronic thromboembolic pulmonary hypertension (CTEPH) after pulmonary embolism (PE). Angiopoietin-2 (ANGPT2) has been shown to regulate angiogenesis, but its importance for thrombus resolution and remodelling is unknown.ANGPT2 plasma concentrations were measured in patients with CTEPH (n=68) and acute PE (n=84). Tissue removed during pulmonary endarterectomy (PEA) for CTEPH was analysed (immuno)histologically. A mouse model of inferior vena cava ligation was used to study the kinetics of venous thrombus resolution in wild-type mice receiving recombinant ANGPT2 via osmotic pumps, and in transgenic mice overexpressing ANGPT2 in endothelial cells.Circulating ANGPT2 levels were higher in CTEPH patients compared to patients with idiopathic pulmonary arterial hypertension and healthy controls, and decreased after PEA. Plasma ANGPT2 levels were also elevated in patients with PE and diagnosis of CTEPH during follow-up. Histological analysis of PEA specimens confirmed increased ANGPT2 expression, and low levels of phosphorylated TIE2 were observed in regions with early-organised pulmonary thrombi, myofibroblasts and fibrosis. Microarray and high-resolution microscopy analysis could localise ANGPT2 overexpression to endothelial cells, and hypoxia and TGF-β1 were identified as potential stimuli. Gain-of-function experiments in mice demonstrated that exogenous ANGPT2 administration and transgenic endothelial ANGPT2 overexpression resulted in delayed venous thrombus resolution, and thrombi were characterised by lower TIE2 phosphorylation and fewer microvessels.Our findings suggest that ANGPT2 delays venous thrombus resolution and that overexpression of ANGPT2 contributes to thrombofibrosis and may thus support the transition from PE to CTEPH.