Patients with cancer have a high risk of deep vein thrombosis which often develops in the veins of the lower limbs. The initial evaluation of clinically suspected deep vein thrombosis includes the estimation of the clinical probability by calculating risk scores such as the Wells score and the measurement of the D-dimer. However, the usefulness and accuracy of the clinical scores and d-dimer test are lower in patients with cancer who often undergo unnecessary (serial) ultrasonography.

The aim of the current study is to analyze variables that predict the presence of thrombosis and may improve the accuracy of the Wells score. In addition, various cut-off of the d-dimer will be evaluated with the goal of improving the sensibility and specificity of the test.

Patients with cancer have a high risk of pulmonary embolism and deep vein thrombosis (DVT). The approach to clinically suspected deep vein thrombosis starts with the evaluation of the clinical probability of disease followed by the measurement of the circulating levels of the D-dimer. The presence of the thrombus is objectively demonstrated by compression ultrasonography. In patients without cancer and clinically suspected DVT, a low clinical probability combined with a normal d-dimer result exclude the presence of DVT and compression ultrasonography can be withheld. In patients with cancer, signs and symptoms of the cancer disease may mimic those of DVT shifting the result of the clinical probability scores towards high probabilities. In addition, the presence of cancer is part of the Wells score which further increases the chance to end up with a higher overall score and thus higher probability. Eventually, patients with cancer are more often classified as at high clinical probability than patients without cancer.

The d-dimer test is poorly specific in the presence of cancer disease and its results are often positive as the consequence of the cancer, the cancer treatments or the concomitant comorbidities.

As a result of the poor specificity of the clinical scores and d-dimer test, these initial diagnostic steps have limited usefulness in patients with cancer who often need to undergo ultrasonography.

The aim of the current study is to analyze variables that predict the presence of thrombosis and may improve the accuracy of the Wells score. In addition, various cut-offs of the D-dimer will be evaluated with the goal of improving the sensibility and specificity of the test.

Study design: prospective cohort of cancer patients with clinically suspected DVT.

Objectives: to evaluate the clinical and demographic variables that predict the presence of DVT and may improve the accuracy of the Wells score in cancer patients. The sensibility and specificity of different cut-offs of the D-dimer will be also evaluated.

Inclusion criteria All ambulatory cancer patients with clinically suspected DVT. All patients will be managed with the calculation of the Wells score, the measurement of the D-dimer. The presence of DVT will be ascertained by whole-leg ultrasonography performed independently of the results of the Wells and D-dimer. A phone contact/visit at 3 months will verify the development of venous thromboembolic complications in all patients in whom DVT was initially excluded by ultrasonography.

Inclusion Criteria:

• All patients with cancer and clinically suspected DVT of the lower limbs

Exclusion Criteria:

• Ongoing oral or parenteral anticoagulant therapy
• Unwillingness to participate

All patients with cancer and clinically suspected DVT of the lower limbs