Suspicion-based screening for occult malignancy in lower-extremity venous thrombosis: a real-world cohort study

Malignancy screening after deep vein thrombosis (DVT) remains controversial. Extensive screening strategies have not demonstrated improved patient outcomes and may cause overdiagnosis. International guidelines recommend chest X-ray and sex-specific cancer screening, whereas Dutch and National Institute for Health and Care Excellence guidelines recommend a suspicion-based approach, with additional diagnostics only when clinical clues for malignancy are present. However, its effectiveness and safety in real-world practice remain uncertain.

The aim of this study was to evaluate the real-world performance of suspicion-based screening for malignancy in a cohort of patients with lower extremity venous thrombosis.

We retrospectively reviewed all patients with suspected DVT between 2016 and 2019. Based on ultrasound findings, patients were classified as having DVT, superficial vein thrombosis (SVT), or no thrombosis (controls). All patients underwent standardized history taking, physical examination, and limited laboratory testing. Additional diagnostics for malignancy were performed only when clinical suspicion existed. Long-term follow-up was the reference standard for malignancy diagnosis.

A total of 1285 patients (258 DVT, 179 SVT, and 848 controls) were followed for a median of 5 years. First-year malignancy incidence was higher in patients with DVT (6.2%) than in SVT (2.2%) or controls (2.3%). Additional diagnostics were performed in 57.8% of patients with DVT, 18.4% of patients with SVT, and 9.2% of controls (P < .001). Malignancies were detected in 4.6%, 1.1%, and 1.3% of these groups, respectively. Among patients without additional diagnostic testing, malignancy was diagnosed within 1 year in 0.8%, 1.1%, and 1.3% of these groups, respectively. The negative predictive value of suspicion-based screening was 98.2% in patients with DVT and 98.6% in both patients with SVT and controls.

In this real-world cohort, suspicion-based screening demonstrated a high negative predictive value with few missed malignancies, supporting a restrictive screening strategy in clinical practice.