Venous thrombosis and D-dimers : a new approach in diagnostic management

Abstract

In the diagnosis of deep vein thrombosis (DVT), serial compression ultrasonography is a safe but inefficient strategy as the overall prevalence of DVT in outpatients suspected of thrombosis is low (17-24%) and only 0.9-2.3% of the patients have DVT detected by the second ultrasound. This inefficiency has lead to the development of new non-invasive strategies by which the need for ultrasound can be reduced by using the pre-test clinical probability (PCP) score and D-dimer measurements. As degradation products of cross-linked fibrin, D-dimer levels will rise in the event of fibrinolysis and are an indirect measurement of thrombus formation. In this thesis, the role of the D-dimer assay in the diagnostic management of DVT is refined. We demonstrate that it is valid to measure D-dimer concentrations using heparin plasma on routine chemical analyzers in stead of the standard citrate plasma. This finding implicates a significant reduction of the turn-around time. Furthermore, we found D-dimer concentrations to remain stable under various (pre)analytic conditions, making comparison of different studies more feasible. In a second study, we studied the course of D-dimer concentrations in patients undergoing embolisation of pulmonary arteries because of arteriovenous malformations. D-dimer concentrations increased almost immediately after formation of a clot and they remained elevated for at least 8 hours after onset. We describe the performance of a rapid latex D-dimer assay in the exclusion of DVT and tried to define variables that might restrict the use of the D-dimer in terms of reduced sensitivity, negative predictive value or specificity. We show that the overall sensitivity and negative predictive value of this D-dimer assay are high. The D-dimer test has a good performance in patients with previous venous thromboembolism, but we do not recommend its use in patients using oral anticoagulants because of reduced sensitivity. The D-dimer tests in patients with cancer and in patients over 70 years old are usually positive and may not be worthwhile in terms of low specificity. The key-question is whether the D-dimer test can replace or reduce the need for CUS in outpatients suspected for having DVT. In a multicenter management study in 812 patients, we show the safety of withholding anticoagulation in patients suspected of having deep venous thrombosis with the combination of a normal D-dimer concentration and a non-high clinical score. This strategy reduces the need for ultrasonography by approximately 30%. In 537 of these patients, extra samples were taken to compare five different D-dimer assays. We show that these D-dimer assays all have a high sensitivity and negative predictive value, but none of the assays reached an NPV of >98% at standard cut-off values. For exclusion of DVT, we therefore recommend combining the D-dimer assay with other non-invasive tests in order to reach failure rates less than 1.0%. The use of D-dimer assays with a very low specificity might not be worthwhile from an economical point of view, as a positive test will necessitate additional testing in the majority of the patients. Venous thromboembolism can be related to malignancy, but routine screening for cancer in every patient with deep venous thrombosis is a matter of debate. We show that high D-dimer concentrations at presentation and during the first days of treatment are indicators of an increased change for overt or occult forms of cancer, especially in patients younger than 60 years.