Evaluating prophylactic heparin in ambulatory patients with solid tumours: a systematic review and individual participant data meta-analysis
Holger J Schünemann,Matthew Ventresca,Mark Crowther,Matthias Briel,Qi Zhou,Simon Noble,Fergus Macbeth,Gareth Griffiths,David Garcia,Gary H Lyman,Marcello Di Nisio,Alfonso Iorio,Lawrence Mbuagbaw,Ignacio Neumann,Nick van Es,Melissa Brouwers,Gordon Guyatt,Michael B Streiff,Maura Marcucci,Tejan Baldeh,Ivan D Florez,Ozlem Gurunlu Alma,Ziad Solh,Patrick M Bossuyt,Lara A Kahale,Walter Ageno,George Bozas,Harry R Büller,Bernard Lebeau,Ramon Lecumberri,Charles Loprinzi,Robert McBane,Kostandinos Sideras,Anthony Maraveyas,Uwe Pelzer,James Perry,Clara Klerk,Giancarlo Agnelli,Elie A Akl
DOI: https://doi.org/10.1016/S2352-3026(20)30293-3
Abstract:Background: Study-level meta-analyses provide high-certainty evidence that heparin reduces the risk of symptomatic venous thromboembolism for patients with cancer; however, whether the benefits and harms associated with heparin differ by cancer type is unclear. This individual participant data meta-analysis of randomised controlled trials examines the effect of heparin on survival, venous thromboembolism, and bleeding in patients with cancer in general and by type. Methods: In this systematic review and meta-analysis we searched MEDLINE, Embase, and The Cochrane Library for randomised controlled trials comparing parenteral anticoagulants with placebo or standard care in ambulatory patients with solid tumours and no indication for anticoagulation published from the inception of each database to January 14, 2017, and updated it on May 14, 2020, without language restrictions. We calculated the effect of parenteral anticoagulant administration on all-cause mortality, venous thromboembolism occurrence, and bleeding related outcomes through multivariable hierarchical models with patient-level variables as fixed effects and a categorical trial variable as a random effect, adjusting for age, cancer type, and metastatic status. Interaction terms were tested to investigate effects in predefined subgroups. This study is registered with PROSPERO, CRD42013003526. Findings: We obtained individual participant data from 14 of 20 eligible randomised controlled trials (8278 [79%] of 10 431 participants; 4139 included in the low-molecular-weight heparin group and 4139 in the control group). Meta-analysis showed an adjusted relative risk (RR) of mortality at 1 year of 0·99 (95% CI 0·93-1·06) and a hazard ratio of 1·01 (95% CI 0·96-1·07). The number of patients with venous thromboembolic events was 158 (4·0%) of 3958 with available data in the low-molecular-weight heparin group compared with 279 (7·1%) of 3957 in the control group. Major bleeding events occurred in 71 (1·7%) of 4139 patients in the control population and 88 (2·1%) in the low-molecular-weight heparin group, and minor bleeding events in 478 (12·1%) of 3945 patients with available data in the control group and 652 (16·6%) of 3937 patients in the low-molecular-weight heparin group. The adjusted RR was 0·58 (95% CI 0·47-0·71) for venous thromboembolism, 1·27 (0·92-1·74) for major bleeding, and 1·34 (1·19-1·51) for minor bleeding. Prespecified subgroup analysis of venous thromboembolism occurrence by cancer type identified the most certain benefit from heparin treatment in patients with lung cancer (RR 0·59 [95% CI 0·42-0·81]), which dominated the overall reduction in venous thromboembolism. Certainty of the evidence for the outcomes ranged from moderate to high. Interpretation: Low-molecular-weight heparin reduces risk of venous thromboembolism without increasing risk of major bleeding compared with placebo or standard care in patients with solid tumours, but it does not improve survival. Funding: Canadian Institutes of Health Research.