Perioperative Pharmacological Thromboprophylaxis in Patients with Cancer: A Systematic Review and Meta-analysis.

Qiang Guo,Bin Huang,Jichun Zhao,Yukui Ma,Ding Yuan,Yi Yang,Xiaojiong Du
DOI: https://doi.org/10.1097/sla.0000000000002074
IF: 13.787
2017-01-01
Annals of Surgery
Abstract:Objective: To compare the relative effects between pharmacological thromboprophylaxis and no anticoagulation.Background: The efficacy and safety of pharmacological thromboprophylaxis in cancer patients undergoing surgery need to be quantified to guide management.Methods: We searched multiple electronic databases (up to March 31, 2016) for trials of cancer patients undergoing surgery that assessed the relative benefits and harms of perioperative pharmacological thromboprophylaxis. Relative risks (RRs) with 95% confidence intervals (CI) were estimated.Results: A total of 39 studies were enrolled in this review. Patients with pharmacological thromboprophylaxis had a relatively reduced incidence of deep venous thrombosis (DVT) compared with those without (0.5% vs 1.2%, RR 0.51, 95% CI 0.27-0.94; P = 0.03) but a significantly increased incidence of bleeding events (RR 2.51, 95% CI 1.79-3.51; P < 0.0001). The incidence of pulmonary embolism (PE) (RR 1.77, 95% CI 0.76-4.14; P = 0.19) and mortality related to venous thromboembolism (VTE) (1/2,811 vs 2/3,380) were similar between the pharmacological thromboprophylaxis group and the no pharmacological thromboprophylaxis group. Low-molecular-weight heparin (LMWH) reduced the incidence of DVT compared with unfractionated heparin (UFH) (RR 0.81, 95% CI 0.66-1.00; P = 0.05), and standard extended thromboprophylaxis after cancer surgery significant decreased the incidence of DVT as compared with conventional thromboprophylaxis (RR 0.57, 95% CI 0.39-0.83; P = 0.003).Conclusions: Routine pharmacological thromboprophylaxis for cancer patients undergoing surgery needs to be carefully considered, because although thromboprophylaxis is associated with lower VTE events, there is a higher incidence of clinically significant bleeding events. If pharmacological thromboprophylaxis is to be used, extended thromboprophylaxis started preoperatively with LWMH might be the most effective strategy.
What problem does this paper attempt to address?