Risk factors of occult malignancy in patients with unprovoked venous thromboembolism.

Venous thromboembolism (VTE) can occur as the first manifestation of an underlying occult malignancy. It remains unclear whether or not a better selection of high risk patients might lead to more efficient occult cancer screening strategies. Our aim was to assess the predictors of occult malignancy diagnosis in patients with unprovoked VTE. Univariate analyses were performed to assess the effect of candidate predictors on occult cancer detection in patients enrolled in a prospective, multicenter, randomized, controlled study (MVTEP study) whose primary aim was to compare a limited screening strategy with a strategy combining limited screening and FDG PET/CT in patients with unprovoked VTE. This trial is completed and registered with ClinicalTrials.gov, number NCT00964275. Between March 3, 2009, and August 18, 2012, 399 patients were included. Five patients withdrew consent and refused the use of their data, and no VTE was confirmed in 2 patients who were excluded from this analysis. A total of 25 (6.4%) out of the 392 analysed patients received a new diagnosis of malignancyduring the 2-years follow-up. Age≥50years (p=0.01), male gender (p=0.04), leukocytes count (p=0.01), and platelets count (p=0.03) were associated with occult cancer detection. Patients with leukocytosis or thrombocytosis had a risk of cancer way above 10%. Previous VTE and smoker status (combining previous and current smokers) were not associated with occult cancer diagnosis (p>0.05). Demographic characteristics (age and sex), and laboratory tests (high platelets and leukocytes counts) may be associated with cancer detection in patients withunprovoked VTE.