WP 3: Prediction of VTE in arterial cardiovascular diseases
In the initial four-year period, we extensively investigated the bidirectional relationship between arterial and venous thrombosis. The next step will be to identify risk factors and triggers of VTE in these patient groups, and build prediction models that can stratify MI and stroke patients into high and low risk of VTE in order to guide clinical decisions on targeted prevention.
Leader: Professor Sigrid Brækkan
In the initial four-year period, we extensively investigated the bidirectional relationship between arterial and venous thrombosis. We found that the risk of VTE was substantially increased in the first 6 months after an event of MI or ischemic stroke, and declined rapidly thereafter (Rinde et al JTH 2016). This risk was not explained by common atherosclerotic risk factors, and carotid atherosclerosis was not associated with future risk of VTE (Hald et al ATVB 2014;34:226-30). Family history of MI was associated with VTE (Lind et al Circ Cardiovasc Genet 2014;7:684-91), but this association was not explained by common genetic prothrombotic risk factors. We estimated that 15% of all the VTE cases in the population could be attributed to MI or stroke. The next step will be to identify risk factors and triggers of VTE in these patient groups, and build prediction models that can stratify MI and stroke patients into high and low risk of VTE in order to guide clinical decisions on targeted prevention. Current knowledge on risk factors and triggers of VTE in MI and stroke patients, particularly in the first months after the event, is scarce, and no risk prediction model exists for these patient groups. With an annual VTE incidence of 1.1 mill, MI and stroke accounts for 165 000 VTE events in Europe each year. Current anticoagulant regimens efficiently prevent first VTE (50-70% efficacy), but at the expense of a substantial risk of major and life-threatening bleedings (2-3%/year). Targeted prevention with anticoagulants should therefore only be offered to subjects with high VTE risk. A prediction model able to identify 80% of patients at risk, with 80% compliance to the model, and 60% efficacy of the anticoagulant preventive treatment, would imply that 63,000 VTE events could be avoided in Europe each year, if such a model existed.
In WP3, we will create cohorts of subjects with MI and stroke derived from the Tromsø and HUNT studies, respectively (follow-up 1994-2016). Altogether these cohorts will comprise around 600 (Tromsø) and 1200 (HUNT) patients with MI and stroke. Information on MI- and stroke-related risk factors (e.g. type, severity etc.) and potential triggers of VTE (e.g. immobilization, paralysis, surgery, infections etc.) will be collected by systematic review of medical records, and genetic and protein biomarkers of VTE (identified in WP1) will be analyzed from stored samples. We will use the Tromsø MI and stroke cohort to build prediction models for VTE based on clinical risk factors, biomarkers and genetic risk factors (derivation cohort). Then, we will externally validate the predictive capability of these models in MI and stroke patients derived from the HUNT study (validation cohort).