Review: In patients with a first VTE, extended testing for undiagnosed cancer does not reduce mortality

Question
In patients with a first venous thromboembolism (VTE), does testing for undiagnosed cancer reduce cancer- and VTE-related mortality and morbidity compared with clinically indicated testing?
Review scope
Included studies compared the use of specific tests for cancer detection with those done at the physician's discretion or standard testing in patients who had a first unprovoked VTE (deep venous thrombosis [DVT] of the lower limb or pulmonary embolism [PE]) and no previous or clinically apparent cancer. Primary outcomes were all-cause, cancer-related, and VTE-related mortality. Other outcomes included VTE recurrence, frequency of underlying cancer diagnosis, and stage of diagnosed cancer.
Review methods
Cochrane Vascular Specialized Register to Feb 2017, Cochrane Register of Studies, CENTRAL (Issue 1, 2017), trial registries, relevant conference proceedings, and reference lists were searched for randomized controlled trials (RCTs) and quasi-RCTs. 4 RCTs (n = 1644, mean age 53 to 69 y, 50% to 67% men) met inclusion criteria: 2 compared extensive testing with testing at the discretion of the physician (n = 396), and 2 compared standard testing plus imaging (computed tomography [CT] of the abdomen and pelvis or 18-fluorodeoxyglucose positron emission tomography [PET]-CT of the chest, abdomen, and pelvis) with standard testing alone (n = 1248). All 4 RCTs had adequate allocation concealment and completeness of data, and 3 RCTs blinded outcome assessors. Blinding of patients and study staff was not considered feasible.
Main results
The main results are in the Table.
Conclusion
In patients with a first venous thromboembolism, testing for undiagnosed cancer beyond standard testing does not reduce all-cause mortality or cancer-related mortality.