Utility of additional abdominopelvic CT in detecting occult cancer in patients with unprovoked venous thromboembolism

Aim To analyse the positive rate for cancer on additional abdominopelvic computed tomography (CT) in patients with unprovoked venous thromboembolism (VTE), evaluate the subsequent emergence of a cancer diagnosis in the clinical follow-up period, and identify any possible predictive factors of cancer in this cohort, which may allow better selection of patients for additional imaging. Materials and methods Consecutive adult patients with VTE were retrospectively identified in two large teaching hospitals between January 2013 and June 2014, including a subset of those with unprovoked VTE. Relevant demographic data were extracted and analysed. All patients had a minimum of 12 months clinicoradiological follow-up. Results One thousand four hundred and forty-six patients with VTE were deemed eligible, of which 699 (48%) were male; the median age (range) was 66 (16–102) years. The prevalence of pre-existing cancer in these patients was 343/1446 (24%), and 388/1446 (27%) were classified as unprovoked VTE. In 12/1446 (0.8%), cancer was diagnosed synchronously with VTE on the initial imaging investigation. Additional screening imaging was performed in 232/388 (60%) including abdominopelvic CT in 205 (53%) patients with unprovoked VTE. Only five additional cancers were identified, all of these occurring in patients with clinical symptoms suspicious for cancer. None of the additional CT examinations identified any clinically occult cancer in asymptomatic patients, and subsequent mean follow-up of 22 (SD=6) months also failed to reveal any further cancer diagnosis. Conclusion Contrary to the National Institute of Health and Care Excellence (NICE) guidance, the yield of performing additional abdominopelvic CT as a screening tool for occult cancer in asymptomatic patients with unprovoked VTE is negligible. A more selective and clinically-driven assessment of these patients is recommended.