WHAT THIS PAPER ADDS This meta-analysis provides complete evidence for the relative efficacy and safety of all major anticoagulant types in the treatment of cancer related venous thromboembolism, pointing out the often insufficient and moderate quality information in the existing literature, which highlights the need for further investigations. Data from previous studies showing that low molecular weight heparins (LMWHs) are more effective than, and equally safe as vitamin K antagonists (VKAs) are confirmed. However, anti-Xa agents are more effective and much safer than VKAs. In a direct comparison, direct oral anticoagulants are better but less safe than LMWHs, although the moderate quality of the evidence calls for further research results that are eagerly awaited.Objective/Background: The aim was to review the relative efficacy and safety of anticoagulation for managing venous thromboembolism (VTE) in patients with cancer. Methods: A systematic review and meta-analysis was carried out. On 17 May 2018 the MEDLINE and Scopus databases were searched for randomised controlled trials (RCTs). Eligible RCTs had to be performed in patients with cancer exclusively or to report results on a subset of patients with cancer. The main study outcomes (efficacy/recurrent VTE and safety/bleeding events) were expressed as risk ratios (RR) with a 95% confidence interval (CI). The quality of evidence was assessed following the GRADE method. Results: Twenty-three RCTs with 6980 patients were identified. Low molecular weight heparins (LMWHs) were more effective than vitamin K antagonists (VKAs) in preventing recurrent VTE (RR 0.58, 95% CI 0.45e0.75) and deep vein thrombosis (RR 0.44, 95% CI 0.29e0.69) but not pulmonary embolism (PE), bleeding, or overall mortality. Direct oral anticoagulants (DOACs) were more effective than VKAs in preventing recurrent VTE (RR 0.65, 95% CI 0.45e0.95) but not DVT, PE, overall mortality, or bleeding. However, anti-Xa DOACs were more effective (RR for VTE 0.64, 95% CI 0.42e0.97) and caused less bleeding than VKAs, although major bleeding was reduced only with DOACs not requiring initial parenteral anticoagulation (RR 0.45, 95% CI 0.21e0.97). In a direct comparison, DOACs were more effective than LMWHs in preventing VTE recurrence (RR 0.64, 95% CI 0.45e0.90) but caused more major bleeding (RR 1.75, 95% CI 1.10e2.77), with no difference in fatal bleeding and overall mortality. Quality of evidence, where sufficient, was mostly moderate or high. Conclusion: Compared with VKAs, LMWHs and DOACs are more effective in treating VTE, but the former caused less bleeding. DOACs are more effective than LMWHs in preventing VTE recurrence but may carry a higher risk of major bleeding, pending additional information by ongoing trials.