Assessing recurrence risk following acute venous thromboembolism: use of algorithms

Machine Learning-Based Predictive Models for Patients with Venous Thromboembolism: A Systematic Review

BACKGROUND

 Venous thromboembolism (VTE) is a chronic disorder with a significant health and economic burden. Several VTE-specific clinical prediction models (CPMs) have been used to assist physicians in decision-making but have several limitations. This systematic review explores if machine learning (ML) can enhance CPMs by analyzing extensive patient data derived from electronic health records. We aimed to explore ML-CPMs' applications in VTE for risk stratification, outcome prediction, diagnosis, and treatment.

METHODS

 Three databases were searched: PubMed, Google Scholar, and IEEE electronic library. Inclusion criteria focused on studies using structured data, excluding non-English publications, studies on non-humans, and certain data types such as natural language processing and image processing. Studies involving pregnant women, cancer patients, and children were also excluded. After excluding irrelevant studies, a total of 77 studies were included.

RESULTS

 Most studies report that ML-CPMs outperformed traditional CPMs in terms of receiver operating area under the curve in the four clinical domains that were explored. However, the majority of the studies were retrospective, monocentric, and lacked detailed model architecture description and external validation, which are essential for quality audit. This review identified research gaps and highlighted challenges related to standardized reporting, reproducibility, and model comparison.

CONCLUSION

 ML-CPMs show promise in improving risk assessment and individualized treatment recommendations in VTE. Apparently, there is an urgent need for standardized reporting and methodology for ML models, external validation, prospective and real-world data studies, as well as interventional studies to evaluate the impact of artificial intelligence in VTE.

French vascular physicians' practices in indicating antiplatelet and anticoagulation therapy in venous thromboembolism

Background: As antiplatelet therapy becomes widespread, physicians face the issue patients being treated with antiplatelet agents (APA) and anticoagulants for venous thromboembolism (VTE). Bleeding risk of this combination is increased 1.5-2.5-fold. The aim of this survey is to assess French vascular physician's management of this combination at the beginning of the treatment and at 6 months of treatment for VTE. Patients and methods: French vascular physicians were surveyed between September and December 2017, using 4 fictional scenarios regarding a VTE event diagnosed in a patient under antiplatelet therapy plus isolated questions in an online questionnaire, sent by the French Society of Vascular Medicine to its members. Out of 1812 physicians, 179 returned valid questionnaires: the response rate was 9.9%. Results: Firstly 97.2% of respondents acknowledged extra risk with this combination; and 63% ceased antiplatelet therapy when initiating anticoagulants; while 36% did not. Secondly, four strategies emerged: 31.4% ceased APA and prescribed full-dose anticoagulants at initiation and at 6 months; 32% associated reduced-dose anticoagulation with APA at 6 months, regardless of what they decided at initiation; 16.5% prescribed isolated full-dose anticoagulants at initiation and reduced-dose at 6 months; lastly 11.2% associated full-dose anticoagulant with antiplatelet therapy at initiation and at 6 months. Conclusions: French vascular physicians adopted different strategies according to estimated risk/benefit ratio. Prospective randomized controlled trials should compare these strategies in order to make recommendations.

A Test in Context: D-Dimer

D-dimer is a soluble fibrin degradation product that results from ordered breakdown of thrombi by the fibrinolytic system. Numerous studies have shown that D-dimer serves as a valuable marker of activation of coagulation and fibrinolysis. Consequently, D-dimer has been extensively investigated for the diagnosis of venous thromboembolism (VTE) and is used routinely for this indication. In addition, D-dimer has been evaluated for determining the optimal duration of anticoagulation in VTE patients, for diagnosing and monitoring disseminated intravascular coagulation, and as an aid in the identification of medical patients at high risk for VTE. Thus, quantification of D-dimer levels serves an important role in guiding therapy. This review: 1) describes how D-dimer is generated; 2) reviews the assays used for its detection; and 3) discusses the role of D-dimer determination in these various conditions.

Occult cancer-related first venous thromboembolism is associated with an increased risk of recurrent venous thromboembolism

Essentials Recurrence risk after an occult cancer-related incident venous thromboembolism (VTE) is unknown. We compared the risk of VTE recurrence in occult-, overt- and non-cancer related first VTE. Patients with occult-cancer related first VTE had the highest risk of VTE recurrence. The high recurrence risk in occult cancer is likely due to the advanced cancers.

SUMMARY

Background Although venous thromboembolism (VTE) is associated with a high recurrence rate, the absolute recurrence rates for cancer-related VTE, particularly occult cancer, are not well known. Objectives To investigate the risk of VTE recurrence in patients with occult and overt cancer-related VTE. Methods Incident VTE events among participants of the first to sixth Tromsø surveys occurring in the period 1994-2012 were included. Occult cancer was defined as cancer diagnosed within a year following a VTE, and overt cancer was defined as cancer diagnosed within the 2 years before a VTE. Results Among 733 patients with incident VTE, 110 had overt cancer and 40 had occult cancer. There were 95 recurrent VTE events during a median of 3.2 years of follow-up. The 1-year cumulative incidence of VTE recurrence was 38.6% in subjects with occult cancer, 15.5% in subjects with overt cancer, and 3.8% in non-cancer subjects. The 1-year risk of recurrence was 12-fold (hazard ratio [HR] 12.4, 95% confidence interval [CI] 5.9-26.3) higher in subjects with occult cancer and four-fold (HR 4.3, 95% CI 2.0-9.2) higher in subjects with overt cancer than in non-cancer subjects. The occult cancers associated with VTE recurrence were typically located at prothrombotic sites (i.e. lung and gastrointestinal) and presented at advanced stages. The majority (69%) of recurrences in subjects with occult cancer occurred before or shortly after cancer diagnosis, and were therefore not treatment-related. Conclusion Our findings suggest that the increased risk of recurrence in patients with occult cancer is mainly attributable to the advanced cancers in these patients.

Extended non-vitamin K antagonist oral anticoagulation therapy for prevention of recurrent venous thromboembolism

Evidence from the use of traditional therapy (low-molecular-weight heparin/vitamin K antagonists) for venous thromboembolism (VTE) treatment and prevention suggests that extending treatment beyond the acute phase reduces recurrence. More recently, several non-vitamin K antagonist oral anticoagulants (NOACs) have been approved in the acute setting; accumulating evidence suggests continuing treatment with these agents beyond 12months offers additional benefits to patients with VTE. This review examines the evidence for NOAC use in longer-duration anticoagulation treatment, and discusses guidelines from major societies. Clinical data from the phase III extension studies for apixaban, dabigatran and rivaroxaban are presented, and the clinical and economic costs and benefits are examined. Evidence from other therapy areas utilising extended treatment regimens highlights the possible impact of factors relevant to extended anticoagulation therapy. Phase IV studies of NOACs are presented. US and European guidelines advise long-term therapy in certain instances, taking into account evidence on NOAC use in VTE accumulated recently. They support NOAC use where they have been selected as the initial therapy choice and therapy needs to be extended beyond 3months. The phase III extension studies demonstrate the benefits of extended NOAC use versus treatment cessation, with reduced recurrence rates versus placebo, although associated with a potential moderate increase in bleeding risk. Phase IV data are also emerging, with the recent XALIA study showing that a broad range of patients with VTE can benefit from continued rivaroxaban treatment; ongoing research will yield data on long-term use of the other NOACs in routine clinical practice.

Re-Initiation of Dabigatran and Direct Factor Xa Antagonists After a Major Bleed

Direct oral anticoagulants (DOACs) are a relatively recent addition to the oral anticoagulant armamentarium, and provide an alternative to the use of vitamin K antagonists such as warfarin. Regardless of the type of agent used, bleeding is the major complication of anticoagulant therapy. The decision to restart oral anticoagulation following a major hemorrhage in a previously anticoagulated patient is supported largely by retrospective studies rather than randomized clinical trials (mostly with vitamin K antagonists), and remains an issue of individualized clinical assessment: the patient's risk of thromboembolism must be balanced with the risk of recurrent major bleeding. This review provides guidance for clinicians regarding if and when a patient should be re-initiated on DOAC therapy following a major hemorrhage, based on the existing evidence.

Safety ad efficacy of direct oral anticoagulants for extended treatment of venous thromboembolism

Currently available anticoagulants have limitations for long-term treatment of venous thromboembolism (VTE). We have evaluated the efficacy and safety of direct oral anticoagulants (DOACs) for extended treatment of VTE. Four randomized controlled trials (RCTs) comparing DOACs (apixaban, rivaroxaban, and dabigatran) with placebo or warfarin for extended treatment of VTE were published. Primary efficacy outcome was recurrent VTE or VTE-related death, and primary safety outcome was major bleeding. DOACs significantly lower the risk of recurrent VTE or VTE-related death compared to placebo/warfarin, as well as all-cause mortality. Risk of major bleeding is not different with DOACs compared to placebo/warfarin. However, DOACs are associated with a significantly higher rate of the composite of major and clinically relevant bleeding compared to placebo. In conclusion, DOACs are effective and safe for the extended treatment of VTE, and may reduce the risk of all-cause mortality.

Re-initiation of dabigatran and direct factor Xa antagonists after a major bleed

Direct oral anticoagulants (DOACs) are a relatively recent addition to the oral anticoagulant armamentarium, and provide an alternative to the use of vitamin K antagonists such as warfarin. Regardless of the type of agent used, bleeding is the major complication of anticoagulant therapy. The decision to restart oral anticoagulation following a major hemorrhage in a previously anticoagulated patient is supported largely by retrospective studies rather than randomized clinical trials (mostly with vitamin K antagonists), and remains an issue of individualized clinical assessment: the patient’s risk of thromboembolism must be balanced with the risk of recurrent major bleeding. This review provides guidance for clinicians regarding if and when a patient should be re-initiated on DOAC therapy following a major hemorrhage, based on the existing evidence.

The epidemiology of venous thromboembolism

Venous thromboembolism (VTE) is categorized by the U.S. Surgeon General as a major public health problem. VTE is relatively common and associated with reduced survival and substantial health-care costs, and recurs frequently. VTE is a complex (multifactorial) disease, involving interactions between acquired or inherited predispositions to thrombosis and VTE risk factors, including increasing patient age and obesity, hospitalization for surgery or acute illness, nursing-home confinement, active cancer, trauma or fracture, immobility or leg paresis, superficial vein thrombosis, and, in women, pregnancy and puerperium, oral contraception, and hormone therapy. Although independent VTE risk factors and predictors of VTE recurrence have been identified, and effective primary and secondary prophylaxis is available, the occurrence of VTE seems to be relatively constant, or even increasing.

Gene microarray analyses for potential biomarkers of single and recurrent venous thromboembolism

Venous thromboembolism is a major cause of morbidity and mortality with a high recurrence rate. The present study aimed to explore the molecular mechanisms and potential biomarkers of single venous thromboembolism (SVTE) and recurrent venous thromboembolism (RVTE). The microarray dataset GSE19151 was downloaded from Gene Expression Omnibus, which contained data from whole blood samples from 63 healthy controls, 32 SVTE and 38 RVTE patients. Differentially expressed genes (DEGs) in the SVTE and RVTE groups compared with those in the controls were identified using the t-test, followed by clustering analysis of DEGs and samples. Functional and pathway enrichment analyses were performed for DEGs in patients with RVTE and SVTE, as well as specific DEGs in patients with RVTE. The identified 42 DEGs in RVTE were mainly enriched in biological processes of cellular protein metabolism, gene expression and translational elongation as well as in pathways associated with ribosomes, Parkinson's disease and oxidative phosphorylation. In SVTE, 20 DEGs were identified, which were mainly involved in biological processes of biopolymer biosynthesis, translational elongation and cellular protein metabolism as well as pathways associated with ribosomes and cardiac muscle contraction. In RVTE, 22 specific DEGs were mainly involved in translational elongation, negative regulation of the force of heart contraction by chemical signals, cell proliferation, ribosomal pathways and protein export. The identified DEGs of SVTE, including COX7C and UQCRQ, may be potential biomarkers for SVTE, and the specific DEGs of RVTE, including ADRBK1, NDUFA5 and ATP5O, may be potential biomarkers for RVTE.

Protein N-linked homocysteine is associated with recurrence of venous thromboembolism

BACKGROUND

Recently, protein N-linked homocysteine (Hcy) has been measured in healthy subjects and patients with marked hyperhomocysteinemia. Since elevated total Hcy (tHcy) levels are associated with increased risk of venous thromboembolism (VTE), we aimed to investigate protein N-linked Hcy levels in patients with VTE.

METHODS

We studied 200 consecutive patients with VTE (89 men, 111 women, aged from 17 to 83 years), including 57 subjects with a subsequent episode of VTE (recurrent VTE) during 24 months of follow-up. Protein N-linked Hcy was assayed using high-performance liquid chromatography with an on-column derivatization with o-phthaldialdehyde and fluorescence detection.

RESULTS

The median protein N-linked Hcy was 1.404 μM (interquartile range [IQR] 0.859-2.066), while the median tHcy (IQR) was 9.1 μM (6.8-11.2). In the whole group protein N-linked Hcy correlated only with C-reactive protein (CRP; r = 0.44, p < 0.0001). In patients with recurrent VTE protein N-linked Hcy correlated with C-reactive protein (r = 0.43, p < 0.0001), tHcy (r = 0.42, p = 0.001) and age (r = 0.32, p = 0.014), but not with thrombophilia, unprovoked VTE or the current anticoagulation. Hyperhomocysteinemia, defined as tHcy ≥ 15 μM (n = 14.7%), was not associated with higher protein N-linked Hcy. Patients with recurrent VTE had higher levels of protein N-linked Hcy compared to those who experienced a single episode of VTE (1.553 μM, 1.157-2.445 vs. 1.27 μM, 0.826-1.884; p = 0.002). Multiple regression adjusted for potential confounders showed that the only independent predictor of protein N-linked Hcy in the upper quartile was CRP > 3mg/L (odds ratio 3.04, 95% confidence interval 2.12-4.36, p < 0.0001).

CONCLUSION

Elevated protein N-linked Hcy concentrations, indicating enhanced protein homocysteinylation in vivo, characterize patients with recurrent VTE and this phenomenon is associated with enhanced inflammatory state.

Immediate Discharge and Home Treatment With Rivaroxaban of Low-risk Venous Thromboembolism Diagnosed in Two U.S. Emergency Departments: A One-year Preplanned Analysis

Objectives

The study hypothesis was that a target‐specific anticoagulant would allow successful home treatment of selected patients with deep vein thrombosis (DVT) and pulmonary embolism (PE) diagnosed in two urban emergency departments (EDs).

Methods

A protocol was established for treating low‐risk DVT or PE patients with rivaroxaban and clinic, follow‐up at both 2 to 5 weeks, and 3 to 6 months. Patients were determined to be low‐risk by using a modified version of the Hestia criteria, supplemented by additional criteria for patients with active cancer. Acceptable outcome rates were defined as venous thromboembolism (VTE) recurrence ≤ 2.1% or bleeding ≤ 9.4% during treatment. VTE recurrence required positive imaging of any VTE. The International Society of Thrombosis and Hemostasis definition of major or clinically relevant nonmajor bleeding was used.

Results

From March 2013 through April 2014, a total of 106 patients were treated. Seventy‐one (68%) had DVT, 30 (28%) had PE, and five (3%) had both, representing 51% of all DVTs and 27% of all PEs diagnosed in both EDs during the period of study. The 106 patients have been followed for a mean (±SD) of 389 (±111) days (range = 213 to 594 days). No patient had VTE recurrence, and no patient had a major or clinically relevant bleeding event while on therapy (none of the 106, 0%, 95% confidence interval [CI] = 0% to 3.4%). However, three patients 2.8% (95% CI = 1% to 8%) had recurrent DVT after cessation of therapy.

Conclusions

Patients diagnosed with VTE and immediately discharged from the ED while treated with rivaroxaban had a low rate of VTE recurrence and bleeding.

Rationale and design of three observational, prospective cohort studies including biobanking to evaluate and improve diagnostics, management strategies and risk stratification in venous thromboembolism: the VTEval Project

INTRODUCTION

Venous thromboembolism (VTE) with its two manifestations deep vein thrombosis (DVT) and pulmonary embolism (PE) is a major public health problem. The VTEval Project aims to investigate numerous research questions on diagnosis, clinical management, treatment and prognosis of VTE, which have remained uncertain to date.

METHODS AND ANALYSIS

The VTEval Project consists of three observational, prospective cohort studies on VTE comprising cohorts of individuals with a clinical suspicion of acute PE (with or without DVT), with a clinical suspicion of acute DVT (without symptomatic PE) and with an incidental diagnosis of VTE (PE or DVT). The VTEval Project expects to enrol a total of approximately 2000 individuals with subsequent active and passive follow-up investigations over a time period of 5 years per participant. Time points for active follow-up investigations are at months 3, 6, 12, 24 and 36 after diagnosis (depending on the disease cohort); passive follow-up investigations via registry offices and the cancer registry are performed 48 and 60 months after diagnosis for all participants. Primary short-term outcome is defined by overall mortality (PE-related death and all other causes of death), primary long-term outcome by symptomatic VTE (PE-related death, recurrence of non-fatal PE or DVT). The VTEval Project includes three 'all-comer' studies and involves the standardised acquisition of high-quality data, covering the systematic assessment of VTE including symptoms, risk profile, psychosocial, environmental and lifestyle factors as well as clinical and subclinical disease, and it builds up a large state-of-the-art biorepository containing various materials from serial blood samplings.

ETHICS AND DISSEMINATION

The VTEval Project has been approved by the local data safety commissioner and the responsible ethics committee (reference no. 837.320.12 (8421-F)). Trial results will be published in peer-reviewed journals and presented at national and international scientific meetings.

TRIAL REGISTRATION NUMBER

NCT02156401.

Epidemiology of venous thromboembolism

Thrombosis can affect any venous circulation. Venous thromboembolism (VTE) includes deep-vein thrombosis of the leg or pelvis, and its complication, pulmonary embolism. VTE is a fairly common disease, particularly in older age, and is associated with reduced survival, substantial health-care costs, and a high rate of recurrence. VTE is a complex (multifactorial) disease, involving interactions between acquired or inherited predispositions to thrombosis and various risk factors. Major risk factors for incident VTE include hospitalization for surgery or acute illness, active cancer, neurological disease with leg paresis, nursing-home confinement, trauma or fracture, superficial vein thrombosis, and-in women-pregnancy and puerperium, oral contraception, and hormone therapy. Although independent risk factors for incident VTE and predictors of VTE recurrence have been identified, and effective primary and secondary prophylaxis is available, the occurrence of VTE seems to be fairly constant, or even increasing.

B blood group: A strong risk factor for venous thromboembolism recurrence

BACKGROUND

Non-O blood group patients are at higher risk of first episode of venous thromboembolism (VTE). However, only little is known about the risk of recurrence according to the blood group. In this study, we aimed to determine the impact of ABO blood group on VTE recurrence.

METHODS

We prospectively recruited 106 consecutive patients with a first documented episode of pulmonary embolism (PE). Patients were followed at least 12months after anticoagulation discontinuation. The main endpoint was recurrence of symptomatic VTE.

RESULTS

Data from 100 patients were analyzed. Median follow-up was 28months [24-34.8]. PE was unprovoked in 48 patients. Mean anticoagulation duration was 5.3±2.2months. The rate of VTE recurrence was 12.7 per 100 patient-years (30 recurrences). B blood group patients had a 2.7-fold increased risk of VTE recurrence (95%CI 1.1-6.2, p=0.03). On multivariate analysis, B blood group was the strongest independent predictor of VTE recurrence (Hazard Ratio (HR) 2.6, 95%CI 1.1-6.1, p=0.04). In contrast, A and AB blood groups were not associated with VTE recurrence. VTE recurrences were less frequent in O blood group compared to non-O patients (HR 0.5, 95%CI 0.2-1.1, p=0.09). O blood group women had a 5-fold decreased risk of VTE recurrence (HR 0.2, 95%CI 0.1-0.8, p=0.01).

CONCLUSIONS

Non-O blood groups, beyond being involved in the occurrence of a first VTE event, also contribute to VTE recurrence. B blood group is strongly associated with VTE recurrence, thus high-risk B blood group patients could benefit from long-term anticoagulation therapy after a first VTE event.