An Individual Participant Data Meta-Analysis of 13 Randomized Trials to Evaluate the Impact of Prophylactic Use of Heparin in Oncological Patients

Background: Parenteral anticoagulants may improve outcomes in patients with cancer by reducing the risk of venous thromboembolism (VTE) and through a direct anti-tumour effect. Study-level meta-analysis indicates a reduction in VTE and provide moderate certainty that a small survival benefit exists; it is unclear if patients with specific cancers benefit more or less. Utilizing data from randomized controlled trials (RCT), this individual participant data meta-analysis examines the impact of heparin on survival, VTE and major bleeding in oncological patients randomized to low-molecular weight heparin (LMWH) or no LMWH.

Methods: We performed a comprehensive systematic search for all RCTS (last search date March 2017) and contacted authors and sponsors to obtain individual participant data of patients with solid cancers and no other indication for prophylactic or therapeutic anticoagulation. We utilized the GRADE approach to evaluate the certainty of evidence and produce an evidence profile. All analyses followed the intention-to-treat principle. We calculated the impact on mortality through multivariable hierarchical models with patient-level variables as fixed effects and a categorical trial variable as a random effect. We adjusted the analysis for age, cancer type and metastasis status. To investigate whether intervention effects vary by predefined subgroups, including type of cancer, we tested interaction terms in the statistical model.

Results: A total of 18 RCTs (n=10,041 participants) were eligible for inclusion and we obtained data from 82.4% of the participants (13 RCTs, n=8,278; n=4,139 for LMWH and n=4,139 for no LMWH). The meta-analysis revealed an adjusted relative risk of mortality within one year of 0.99 (95% CI: 0.95, 1.03) and a hazard ratio of 0.97 (95% CI: 0.82, 1.14) after one year. The relative risk for VTE was 0.58 (95% CI: 0.48, 0.71), 0.57 (95% CI: 0.44, 0.74) for symptomatic deep vein thrombosis and 0.58 (95%CI: 0.44, 0.77) for symptomatic pulmonary embolism, separately. For every 1,000 patients treated, approximately 16 fewer would experience symptomatic DVT and 16 fewer would experience any PE. The adjusted relative risk for major bleeding throughout trial duration was 1.24 (95% CI: 0.91, 1.69; P=0.17). Subgroup analysis, by cancer type, of VTE occurrence throughout trial duration identified lung cancer OR=0.52 (95% CI: 0.39, 0.68; P<0.001) and pancreatic cancer OR=0.55 (95% CI: 0.35, 0.88; P=0.01) patients as experiencing the greatest benefit from LMWH treatment. The certainty of the evidence for the outcomes was moderate to high.

Conclusion: LMWH reduces risk of VTE without increasing risk of bleeding but does not improve survival across all patients.

Funding: Canadian Institutes of Health Research knowledge synthesis grant, KRS 126594

Registration: International Prospective Register for Systematic Reviews (PROSPERO), CRD42013003526.