Парентеральные антикоагулянты у амбулаторных пациентов с раком
Abstract
Background
Anticoagulation may improve survival in patients with cancer through a speculated anti‐tumour effect, in addition to the antithrombotic effect, although may increase the risk of bleeding.
Objectives
To evaluate the efficacy and safety of parenteral anticoagulants in ambulatory patients with cancer who, typically, are undergoing chemotherapy, hormonal therapy, immunotherapy or radiotherapy, but otherwise have no standard therapeutic or prophylactic indication for anticoagulation.
Search methods
A comprehensive search included (1) a major electronic search (February 2016) of the following databases: Cochrane Central Register of Controlled Trials (CENTRAL) (2016, Issue 1), MEDLINE (1946 to February 2016; accessed via OVID) and Embase (1980 to February 2016; accessed via OVID); (2) handsearching of conference proceedings; (3) checking of references of included studies; (4) use of the 'related citation' feature in PubMed and (5) a search for ongoing studies in trial registries. As part of the living systematic review approach, we are running searches continually and we will incorporate new evidence rapidly after it is identified. This update of the systematic review is based on the findings of a literature search conducted on 14 August 2017.
Selection criteria
Randomized controlled trials (RCTs) assessing the benefits and harms of parenteral anticoagulation in ambulatory patients with cancer. Typically, these patients are undergoing chemotherapy, hormonal therapy, immunotherapy or radiotherapy, but otherwise have no standard therapeutic or prophylactic indication for anticoagulation.
Data collection and analysis
Using a standardized form we extracted data in duplicate on study design, participants, interventions outcomes of interest, and risk of bias. Outcomes of interested included all‐cause mortality, symptomatic venous thromboembolism (VTE), symptomatic deep vein thrombosis (DVT), pulmonary embolism (PE), major bleeding, minor bleeding, and quality of life. We assessed the certainty of evidence for each outcome using the GRADE approach (GRADE handbook [GRADE handbook]).
Main results
Of 6947 identified citations, 19 RCTs fulfilled the eligibility criteria. These trials enrolled 9650 participants. Trial registries' searches identified nine registered but unpublished trials, two of which were labeled as 'ongoing trials'. In all included RCTs, the intervention consisted of heparin (either unfractionated heparin or low molecular weight heparin). Overall, heparin appears to have no effect on mortality at 12 months (risk ratio (RR) 0.98; 95% confidence interval (CI) 0.93 to 1.03; risk difference (RD) 10 fewer per 1000; 95% CI 35 fewer to 15 more; moderate certainty of evidence) and mortality at 24 months (RR 0.99; 95% CI 0.96 to 1.01; RD 8 fewer per 1000; 95% CI 31 fewer to 8 more; moderate certainty of evidence). Heparin therapy reduces the risk of symptomatic VTE (RR 0.56; 95% CI 0.47 to 0.68; RD 30 fewer per 1000; 95% CI 36 fewer to 22 fewer; high certainty of evidence), while it increases in the risks of major bleeding (RR 1.30; 95% 0.94 to 1.79; RD 4 more per 1000; 95% CI 1 fewer to 11 more; moderate certainty of evidence) and minor bleeding (RR 1.70; 95% 1.13 to 2.55; RD 17 more per 1000; 95% CI 3 more to 37 more; high certainty of evidence). Results failed to confirm or to exclude a beneficial or detrimental effect of heparin on thrombocytopenia (RR 0.69; 95% CI 0.37 to 1.27; RD 33 fewer per 1000; 95% CI 66 fewer to 28 more; moderate certainty of evidence); quality of life (moderate certainty of evidence).
Authors' conclusions
Heparin appears to have no effect on mortality at 12 months and 24 months. It reduces symptomatic VTE and likely increases major and minor bleeding. Future research should further investigate the survival benefit of different types of anticoagulants in patients with different types and stages of cancer. The decision for a patient with cancer to start heparin therapy should balance the benefits and downsides, and should integrate the patient's values and preferences.
Editorial note:This is a living systematic review. Living systematic reviews offer a new approach to review updating in which the review is continually updated, incorporating relevant new evidence, as it becomes available. Please refer to the Cochrane Database of Systematic Reviews for the current status of this review.
PICO
Резюме на простом языке
Инъекционные средства для разжижения крови (антикоагулянты) у пациентов с раком
Актуальность
Свидетельства из исследований позволяют предполагать, что средства, разжижающие кровь, могут улучшить выживаемость больных раком. Эта польза может быть связана с их прямым противоопухолевым эффектом в дополнение к предотвращению образования тромбов (кровяных сгустков). Однако, антикоагулянты могут также увеличить риск кровотечения, которые могут быть серьезными и уменьшить выживаемость. Поэтому важно понимать плюсы и минусы лечения, чтобы пациенты и их врачи были осведомлены о балансе рисков и пользы.
Характеристика исследований
Мы провели поиск научной литературы на предмет исследований антикоагулянтов у людей с раком. Доказательства актуальны на 14 августа 2017 года. Мы включили 19 подходящих испытаний.
Основные результаты
Мы выбрали 19 клинических испытаний с участием 9650 людей с раком. Большинство испытаний включали участников с различными типами рака, в особенности мелкоклеточного рака легкого, немелкоклеточного рака легкого и рака поджелудочной железы. Все исследования были проведены в амбулаторных условиях. Результаты позволяют предположить, что влияние инъекционных антикоагулянтов на выживаемость остается неопределенным, и если эффект есть, то размер эффекта небольшой. Результаты также позволяют предположить, что инъекционные антикоагулянты снижают риск образования тромбов примерно наполовину и, возможно, увеличивают риск больших и малых кровотечений на 4 на 1000 и на 17 на 1000 соответственно. Влияние на качество жизни остается неопределенным.
Определенность доказательств
Мы оценили определенность доказательств как высокую в отношении симптоматической венозной тромбоэмболии и малых кровотечений, как среднюю ‐ в отношении смертности, больших кровотечений и качества жизни.
Редакторская пометка: Это живой систематический обзор. Живые систематические обзоры предлагают новый подход к обновлениям обзоров, при котором обзор постоянно обновляется, включая соответствующие новые доказательства по мере их появления. Пожалуйста, обратитесь к Кокрейновской базе данных систематических обзоров, чтобы узнать текущий статус этого обзора.
Authors' conclusions
Summary of findings
| Heparin prophylaxis compared with no prophylaxis in ambulatory patients with cancer without VTE receiving systemic therapy P: Ambulatory patients with cancer without VTE receiving systemic therapy I: Heparin prophylaxis C: No prophylaxis S: Outpatient | |||||
| Outcomes | № of participants | Quality of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
|---|---|---|---|---|---|
| Risk with No prophylaxis | Risk difference with Heparin prophylaxis | ||||
| Mortality | 9575 | ⊕⊕⊕⊝ | RR 0.98 | Study population | |
| 504 per 1000 | 10 fewer per 1000 | ||||
| Mortality | 5229 | ⊕⊕⊕⊝ | RR 0.99 | Study population | |
| 778 per 1000 | 8 fewer per 1000 | ||||
| Symptomatic VTE | 9036 | ⊕⊕⊕⊕ | RR 0.56 | Study population | |
| 68 per 1000 | 30 fewer per 1000 | ||||
| Major bleeding | 9592 | ⊕⊕⊕⊝ | RR 1.30 | Study population | |
| 14 per 1000 | 4 more per 1000 | ||||
| Minor bleeding | 9245 | ⊕⊕⊕⊕ | RR 1.70 | Study population | |
| 24 per 1000 | 17 more per 1000 | ||||
| Thrombocytopenia | 5832 | ⊕⊕⊕⊝ | RR 0.69 | Study population | |
| 105 per 1000 | 33 fewer per 1000 | ||||
| Quality of life impairment | 2241 | ⊕⊕⊕⊝ | ‐ | Macbeth 2016 (FRAGMATIC): " There was no difference between the two groups with respect to quality‐adjusted life years gained in the first year... No difference in overall quality of life at 6 months (P = .94) or at 12 months (P = .89)... Overall quality of life did not change significantly over the study period".Sideras 2006: "The QOL and SDS scores were similar, both at baseline and during the protocol period, in patients randomized to receive LMWH vs those not randomized to receive LMWH." | |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence | |||||
| 1 Downgraded one level due to serious imprecision: Confidence interval includes values suggesting clinically significant benefit and values suggesting no effect. 2 Downgraded one level due to serious imprecision: Confidence interval includes values suggesting clinically significant harm and values suggesting no effect. 3 Downgraded one level due to serious imprecision: Confidence interval includes values suggesting clinically significant benefit and values suggesting harm. 4 Downgraded one level due to serious risk of bias: Both studies were open‐label studies (lack of blinding may impact the patient‐reported subjective outcomes) | |||||
Background
Please refer to the glossary for the definitions of technical terms (Table 1).
| Term | Definition |
|---|---|
| Adjuvant therapy | Assisting in the amelioration or cure of disease |
| Anticoagulation | The process of hindering the clotting of blood especially by treatment with an anticoagulant |
| Antithrombotic | Used against or tending to prevent thrombosis (clotting) |
| Bacteremia | The presence of bacteria in the blood |
| Central venous line | Synthetic tube that is inserted into a central (large) vein of a patient to provide temporary intravenous access for the administration of fluid, medication or nutrients |
| Coagulation | Clotting |
| Deep vein thrombosis (DVT) | A condition marked by the formation of a thrombus within a deep vein (as of the leg or pelvis) that may be asymptomatic or be accompanied by symptoms (such as swelling and pain) and that is potentially life‐threatening if dislodgment of the thrombus results in pulmonary embolism |
| Fibrin | A white insoluble fibrous protein formed from fibrinogen by the action of thrombin, especially in the clotting of blood |
| Fondaparinux | An anticoagulant medication |
| Hemostatic system | The system that shortens the clotting time of blood and stops bleeding |
| Heparin | An enzyme occurring especially in the liver and lungs that prolongs the clotting time of blood by preventing the formation of fibrin. Two forms of heparin that are used as anticoagulant medications are: unfractionated heparin (UFH) and low molecular weight heparins (LMWH) |
| Impedance plethysmography | A technique that measures the change in blood volume (venous blood volume as well as the pulsation of the arteries) for a specific body segment |
| Kappa statistics | A measure of degree of non‐random agreement between observers and/or measurements of a specific categorical variable |
| Metastasis | The spread of cancer cells from the initial or primary site of disease to another part of the body |
| Oncogene | A gene having the potential to cause a normal cell to become cancerous |
| Osteoporosis | A condition that especially affects older women and is characterized by a decrease in bone mass with decreased density and enlargement of bone spaces producing porosity and brittleness |
| Parenteral nutrition | The practice of feeding a patient intravenously, circumventing the gut |
| Pulmonary embolism (PE) | Embolism of a pulmonary artery or one of its branches that is produced by foreign matter and most often a blood clot originating in a vein of the leg or pelvis and that is marked by labored breathing, chest pain, fainting, rapid heart rate, cyanosis, shock and sometimes death |
| Stroma | The supporting framework of an organ typically consisting of connective tissue |
| Thrombin | A proteolytic enzyme formed from prothrombin that facilitates the clotting of blood by catalyzing conversion of fibrinogen to fibrin |
| Thrombocytopenia | Persistent decrease in the number of blood platelets that is often associated with hemorrhagic conditions |
| Thrombosis | The formation or presence of a blood clot within a blood vessel |
| Vitamin K antagonists | Anticoagulant medications that are used for anticoagulation. Warfarin is a vitamin K antagonist |
| Warfarin | An anticoagulant medication that is a vitamin K antagonist, which is used for anticoagulation |
| Ximelagatran | An anticoagulant medication |
Description of the condition
Since the 1930s, scientists have been exploring the effects of anticoagulation on cancer (Smorenburg 2001). Studies have implicated the tumor‐mediated activation of the hemostatic system in both the formation of tumor stroma and in tumor metastasis (Dvorak 1986; Francis 1998; Levine 2003). There is also evidence that heparin inhibits expression of oncogenes and formation of thrombin and fibrin induced by cancer cells (Smorenburg 2001). In addition, heparin potentially inhibits intravascular arrest of cancer cells and thus metastasis (Smorenburg 2001).
Description of the intervention
Heparin and low molecular weight heparins (LMWHs), fondaparinux and danaparoid do not have intrinsic anticoagulant activity but potentiate the activity of antithrombin III in inhibiting activated coagulation factors. These agents constitute indirect anticoagulants as their activity is mediated by plasma cofactors. Recombinant hirudin, bivalirudin and argatroban directly inhibit thrombin and are classified as direct anticoagulants (Hirsh 2008). Heparin and its low molecular weight derivatives are not absorbed orally and must be administered parenterally by intravenous infusion or subcutaneous injections (Hirsh 1993).
How the intervention might work
Researchers have hypothesized that heparin may improve outcomes in patients with cancer through an anti‐tumor effect in addition to its antithrombotic effect (Thodiyil 2002). In a 1992 clinical trial comparing nadroparin, a LMWH, with unfractionated heparin in patients with proven deep vein thrombosis (DVT), nadroparin unexpectedly reduced mortality in the subgroup of patients with cancer (Prandoni 1992). At the same time, anticoagulants increase the risk for bleeding. In fact, in patients with venous thrombosis on anticoagulation, the risk of bleeding was higher if patients had cancer and correlated with the extent of cancer (Prandoni 2002). Heparins are also known to cause thrombocytopenia (reduced numbers of platelets) and heparin‐induced thrombocytopenia (HIT) syndrome (Girolami 2006).
Why it is important to do this review
We initially conducted this and other reviews on this topic and their updates to directly and better inform clinical practice guidelines. The last update of this systematic review, published in 2014 (Akl 2014 (parenteral)), identified 15 trials enrolling 7662 participants (Agnelli 2009 (PROTECHT); Agnelli 2012 (SAVE‐ONCO); Altinbas 2004; Haas 2012 (TOPIC 1); Haas 2012 (TOPIC 2); Kakkar 2004 (FAMOUS); Klerk 2005 (MALT); Lebeau 1994; Lecumberri 2013 (ABEL); Maraveyas 2012 (FRAGEM); Pelzer 2009 (CONKO‐2004); Perry 2010 (PRODIGE); Sideras 2006; van Doormaal 2011 (INPACT); Weber 2008). The included trials provided high‐certainty evidence for a reduction of venous thromboembolism (VTE) with heparin thromboprophylaxis compared to no heparin thromboprophylaxis. Since then, we have identified three eligible trials addressing this question (Khorana 2017 (PHACS); Macbeth 2016 (FRAGMATIC); Zwicker 2013 (MICRO TEC)) and the full‐text publication of a previously included abstract (Pelzer 2015 (CONKO‐004).
Living systematic review approach: Following the publication of this current 2017 update of the review, we will maintain it as a living systematic review. This means we will be continually running the searches and rapidly incorporating any newly identified evidence (for more information about the living systematic review approach being piloted by Cochrane, see Appendix 1. We believe a living systematic review approach is appropriate for this review for four reasons. First, the review addresses an important topic for clinical practice; patients with cancer have a relatively high rate of VTE, up to 17.7% (Ay 2010). In addition, the occurrence of VTE is associated with a 2.3 increased risk of death in patients with breast and non‐small cell lung cancer (NSCLC), 2.5 times lengthening of hospital stay among patients with lung cancer, and 50% higher total cost for patients with lung cancer (Chew 2008, Chew 2007; Connolly 2012). Second, there remains uncertainty in the existing evidence base; the 2014 update of this systematic review found a potential subgroup effect on all‐cause mortality at one year, with a possible higher reduction in mortality among patients with small cell lung cancer (SCLC) compared to other types of cancer. Third, we are aware of several recently published and ongoing trials in this area that will be important to incorporate in a timely manner. Fourth, we are planning to use this living systematic review as the basis of a living recommendation in a clinical practice guideline with the American Society of Hematology (Akl 2017). For more information about the living systematic review approach being piloted by Cochane, see Appendix 2.
Objectives
To evaluate the effectiveness and safety of parenteral anticoagulants in ambulatory patients with cancer. Typically, these patients are undergoing chemotherapy, hormonal therapy, immunotherapy or radiotherapy, but otherwise have no standard therapeutic or prophylactic indication for anticoagulation.
Methods
Criteria for considering studies for this review
Types of studies
Randomized controlled trials (RCTs).
Types of participants
Participants with cancer with no standard indication for prophylactic anticoagulation (e.g. for acute illness, for central venous line placement, perioperatively) or for therapeutic anticoagulation (e.g. for the treatment of deep vein thrombosis (DVT) or pulmonary embolism (PE)). Patients could have been of any age group (including children). Typically, these participants are undergoing chemotherapy, hormonal therapy, immunotherapy or radiotherapy.
Types of interventions
Intervention arm: parenteral anticoagulants, such as unfractionated heparin or low molecular weight heparin (LMWH).
Comparator intervention: placebo or no intervention.
The trial protocol should have planned to provide all other co‐interventions (e.g. chemotherapy) similarly.
Types of outcome measures
Primary outcomes
-
All‐cause mortality; pre‐specified at 12 months, 24 months and over the duration of the trial.
Secondary outcomes
-
Symptomatic DVT: DVT events had to be suspected clinically, and diagnosed using an objective diagnostic test such as: venography, 125I‐fibrinogen‐uptake test, impedance plethysmography or compression ultrasound.
-
PE: PE events had to be suspected clinically, and diagnosed using an objective diagnostic test such as: pulmonary perfusion/ventilation scans, computed tomography, pulmonary angiography or autopsy.
-
Major bleeding: we accepted the authors' definitions of major bleeding.
-
Minor bleeding: we accepted the authors' definitions of minor bleeding.
-
Health‐related quality of life: had to be measured using a validated tool.
-
Thrombocytopenia.
Search methods for identification of studies
Electronic searches
The search was part of a comprehensive search for studies of anticoagulation in patients with cancer. We did not use language restrictions. We conducted comprehensive searches on 14 August 2017, following the original electronic searches in January 2007, February 2010, February 2013, and February 2016 (last major search). We electronically searched the following databases: the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (starting 1946), and Embase (starting 1980; accessed via OVID) . The search strategies combined terms for anticoagulants, terms for cancer and a search filter for RCTs. We list the full search strategies for each of the electronic databases in Appendix 3, Appendix 4, and Appendix 5, respectively.
Living systematic review approach: Since the last major search in February 2016, we have been running searches monthly, using auto‐alerts to deliver the monthly yield by email. We will incorporate new evidence rapidly after it is identified. This update of the systematic review is based on the findings of a literature search conducted on 14 August 2017. We will review search methods and strategies approximately yearly, to ensure they reflect any terminology changes in the topic area, or in the databases.
Searching other resources
We handsearched the conference proceedings of the American Society of Clinical Oncology (ASCO, starting with its first volume, 1982 up to August 2017) and of the American Society of Hematology (ASH, starting with its 2003 issue up to August 2017).We also searched ClinicalTrials.gov and WHO International Clinical Trials Registry Platform for ongoing studies. We reviewed the reference lists of papers included in this review and of other relevant systematic reviews. We used the 'related citation' feature in PubMed to identify additional articles and 'citation tracking' of included studies in Web of Science Core Collection.In addition, experts in the field were contacted for information about unpublished and ongoing trials.
Living systematic review approach: We will search the conference proceedings of ASCO and ASH soon after their publications, ClinicalTrials.gov, and WHO International Clinical Trials Registry Platform on a monthly basis. As an additional step, we will contact corresponding authors of ongoing studies as they are identified and ask them to advise when results are available. We will continue to review the reference lists for any prospectively identified studies, with running the 'related citation' for all included studies on a monthly basis. Also, we will contact the corresponding authors of any newly included studies for advice as to other relevant studies. We will conduct citation tracking of included studies in Web of Science Core Collection on an ongoing basis, using citation alerts in Web of Science Core Collection.
Data collection and analysis
Selection of studies
Two review authors independently screened the titles and abstracts of identified articles for eligibility. We retrieved the full text of articles judged as potentially eligible by at least one review author. Two review authors then independently screened the full‐text articles for eligibility using a standardized form with explicit inclusion and exclusion criteria. The two review authors resolved their disagreements by discussion or by consulting a third review author.
Living systematic review approach: For the monthly searches, we will immediately screen any new citations retrieved each month. As the first step of monthly screening, we will apply the machine learning classifier (RCT model) available in the Cochrane Register of Studies (CSR‐Web; Wallace 2017). The classifier assigns a probability (from 0 to 100) to each citation for being a true RCT. For citations that are assigned a probability score of less than 10, the machine learning classifier currently has a specificity/recall of 99.987% (James Thomas, personal communication). For citations assigned a score from 10 to 100, we will screen them in duplicate and independently. Citations that score 9 or less will be screened by Cochrane Crowd (Cochrane Crowd). Any citations that are deemed to be potential RCTs by Cochrane Crowd will be returned to the authors for screening.
Data extraction and management
Two review authors independently extracted data from each included study and resolved their disagreements by discussion. We aimed to collect data related to the following.
Participants
-
Number of participants randomized to each treatment arm.
-
Number of participants followed up in each treatment arm.
-
Number of withdrawals from treatment in each treatment arm.
-
Population characteristics (age, gender, co‐morbidity).
-
History of VTE.
-
Type of cancer (site and histology).
-
Stage of cancer.
-
Time since cancer diagnosis.
Interventions
-
Type of anticoagulant: unfractionated heparin, LMWH or fondaparinux.
-
Dose: prophylactic versus therapeutic (Table 2).
-
Duration of treatment.
-
Control: placebo or no intervention.
-
Co‐interventions including chemotherapy and hormonal therapy, immunotherapy and radiotherapy (type and duration).
| LMWH | Generic name | Prophylactic dose | Therapeutic dose |
|---|---|---|---|
| Lovenox | Enoxaparin | 40 mg once daily | 1 mg/kg twice daily |
| Fragmin | Dalteparin | 2500 to 5000 units once daily | 200 U/kg once daily or |
| Innohep, Logiparin | Tinzaparin | 4500 units once daily | 90 U/kg twice daily |
| Fraxiparine | Nadroparin | 35 to 75 anti‐Xa international units/kg once daily | 175 anti‐Xa int. units/kg once daily |
| Certoparin | Sandoparin | 3000 anti‐Xa international units once daily | — |
Outcomes
We extracted both time‐to‐event data (for the survival outcome) and dichotomous data (for all outcomes). For mortality, we collected data for the pre‐specified time point of 12 months, but also for 24 months and for over the duration of follow‐up.
For time‐to‐event survival data, we abstracted the log(hazard ratio (HR)) and its variance from trial reports. If these were not reported, we digitized the published Kaplan‐Meier survival curves and estimated the log(HR) and its variance using Parmar's methods (Parmar 1998). We also noted the minimum and maximum duration of follow‐up, which are required to make these estimates. We performed these calculations in Stata 9, using a specially written program, which yielded the reported log(HR) and variance when used on the data presented in Table V of Parmar 1998.
For dichotomous data, we extracted data necessary to conduct complete case analysis as the primary analysis. We collected all‐cause mortality at one year (time point defined a priori in the protocol) and at two years (time point defined post hoc based upon results reported in the individual RCTs). When we could not obtain the number of events at the time points of interest from the paper or from the authors, two review authors calculated these numbers independently and in duplicate from survival curves, if available (Altinbas 2004; Kakkar 2004 (FAMOUS)). We used the mean of the two estimates when they differed. We assessed agreement between the two authors for each estimated value by calculating the percentage difference, which is the difference between the two estimates divided by the denominator (number of people at risk for the event) and multiplied by 100. For some studies, where VTE is not reported as a separate outcome, we added the number of events of DVT and PE.
We attempted to contact study authors for incompletely reported data. We decided a priori to consider abstracts in the main analysis only if authors supplied us with full reports of their methods and results.
Other
We extracted from each included trial any information on the following points:
-
source of funding;
-
ethical approval;
-
conflict of interest.
Assessment of risk of bias in included studies
We assessed risk of bias at the study level using the Cochrane 'Risk of bias' tool (Cochrane Handbook). Two review authors independently assessed the risk of bias of each included study and resolved their disagreements by discussion. 'Risk of bias' criteria included:
-
adequate sequence generation;
-
allocation concealment;
-
blinding of participants and personnel;
-
blinding of outcome assessment;
-
percentage of follow‐up and whether incomplete outcome data were addressed;
-
whether the study was free of selective reporting; and
-
whether the study was stopped early for benefit.
See section on Dealing with missing data about assessing risk of bias associated with participants with missing data.
Measures of treatment effect
We collected and analyzed hazard ratios (HRs) for time‐to‐event data and risk ratios (RRs) for dichotomous data. None of the outcomes of interest was meta‐analyzed as a continuous variable.
Unit of analysis issues
The unit of analysis was the individual participant.
Dealing with missing data
Determining participants with missing data
It was not clear whether certain participant categories (e.g. those described as "withdrew consent" or "experienced adverse events") were actually followed up by the trial authors (versus had missing participant data) (Akl 2016). To deal with this issue, we made the following considerations:
-
"ineligible participants" and "did not receive the first dose" participant categories, which are defined prior to the initiation of the study intervention, most likely have missing participant data;
-
"withdrew consent", "lost to follow‐up" and "outcome not assessable" participant categories and other category explicitly reported as not being followed‐up, which are defined after the initiation of the study intervention, most likely have missing participant data;
-
"dead", "experienced adverse events", "non‐compliant", "discontinued prematurely" and similarly described participant categories are less likely to have missing participant data.
Dealing with participants with missing data in the primary meta‐analysis
In the primary meta‐analysis, we used a complete case analysis approach, i.e. we excluded participants considered to have missing data (Guyatt 2017).
For categorical data, we used the following calculations for each study arm.
-
Denominator: (number of participants randomized) ‐ (number of participants most likely with missing data, both pre‐ and post‐intervention initiation).
-
Numerator: number of participants with observed events (i.e. participants who suffered at least one event for the outcome of interest during their available follow‐up time).
For continuous data, we planned to use for each study arm the reported mean and standard deviation for participants actually followed up by the trial authors.
Assessing the risk of bias associated with participants with missing data
When the primary meta‐analysis of a specific outcome found a statistically significant effect, we conducted sensitivity meta‐analyses to assess the risk of bias associated with missing participant data. Those sensitivity meta‐analyses used a priori plausible assumptions about the outcomes of participants considered to have missing data. The assumptions we used in the sensitivity meta‐analyses were increasingly stringent in order to progressively challenge the statistical significance of the results of the primary analysis (Akl 2013; Ebrahim 2013).
For categorical data, and for RR showing a reduction in effect (RR < 1), we used the following increasingly stringent but plausible assumptions (Akl 2013):
-
for the control arm, relative incidence (RI) among those with missing data (lost to follow‐up (LTFU)) compared to those with available data (followed up, FU) in the same arm (RILTFU/FU) = 1; for the intervention arm, RILTFU/FU = 1.5;
-
for the control arm, RILTFU/FU = 1; for the intervention arm, RILTFU/FU = 2;
-
for the control arm, RILTFU/FU = 1; for the intervention arm, RILTFU/FU = 3;
-
for the control arm, RILTFU/FU = 1; for the intervention arm, RILTFU/FU = 5.
For RR showing an increase in effect (RR > 1), we switched the above assumptions between the control and interventions arms (i.e. used RILTFU/FU = 1 for the intervention arm).
Specifically, we used the following calculations for each study arm.
-
Denominator: (number of participants randomized) ‐ (number of participants most likely with missing data, pre‐intervention initiation).
-
Numerator: (number of participants with observed events) + (number of participants most likely with missing data post‐intervention initiation, with assumed events).
Assumed events are calculated by applying the a priori plausible assumptions to the participants considered most likely with missing data post‐intervention initiation.
For continuous data, we planned to use the four strategies suggested by Ebrahim and colleagues. The strategies imputed the means for participants with missing data based on the means of participants actually followed up in individual trials included in the systematic review. To impute standard deviation (SD), we used the median SD from the control arms of all included trials (Ebrahim 2013).
Assessment of heterogeneity
We assessed heterogeneity between trials by visual inspection of forest plots, by estimation of the percentage heterogeneity between trials which cannot be ascribed to sampling variation (Higgins 2003), and by a formal statistical test of the significance of the heterogeneity (Deeks 2001). If there was evidence of substantial heterogeneity, we attempted to investigate the possible reasons for this (see section on Subgroup analysis and investigation of heterogeneity).
Assessment of reporting biases
We assessed for selective outcome reporting by trying to identify whether the study was included in a trial registry, whether a protocol was available, and whether the methods section provided a list of outcomes. We compared the list of outcomes from those sources to the outcomes reported in the published paper. We also assessed for possible publication bias by creating an inverted funnel plot for the primary outcome of survival.
Data synthesis
For time‐to‐event data, we pooled the log(HRs) using a random‐effects model (DerSimonian 1986), and the generic inverse variance facility of RevMan 2014. For dichotomous data, we calculated the RR separately for each study. When analyzing data related to participants who were reported as not compliant, we attempted to adhere to the principles of intention‐to‐treat (ITT) analysis. We approached the issue of non‐compliance independently from that of missing data (Alshurafa 2012). We then pooled the results of the different studies using a random‐effects model. We assessed the certainty of evidence at the outcome level using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach (GRADE handbook).
Living systematic review approach: Whenever new evidence (studies, data or information) that meets the review inclusion criteria is identified, we will immediately assess risk of bias and extract the data and incorporate it in the synthesis, as appropriate. We will not adjust the meta‐analyses to account for multiple testing given the methods related to frequent updating of meta‐analyses are under development (Simmonds 2017).
Subgroup analysis and investigation of heterogeneity
We planned to explore heterogeneity by conducting subgroup analyses based on the characteristics of participants (type and stage of cancer, and whether participants were on cancer treatment or not). In particular, we conducted subgroup analyses for patients with (1) lung cancer (either SCLC or NSCLC) versus those with non‐lung cancer; (2) patients with advanced cancer versus those with non‐advanced cancer. We included in the lung versus non lung subgroup analysis data from:
-
studies that recruited only patients with lung cancer (either SCLC or NSCLC) and studies that recruited only patients with non‐lung cancer;
-
studies that recruited both lung and non‐lung cancer if they provided data for subgroups of patients with lung cancer AND data for subgroups of patients with non‐lung cancer;
-
studies that recruited both lung and non‐lung cancer but did not provide subgroup data, if more than 75% of participants had lung cancer or more than 75% of participants had non‐lung cancer.
Similarly for the subgroup analysis for non‐advanced cancer. We planned to assess the credibility of subgroup effect, when statistically significant, using the criteria suggested by Sun 2010.
Sensitivity analysis
We planned for sensitivity analyses excluding trials at high risk of bias. As described above, we also planned for sensitivity meta‐analyses to assess the risk of bias associated with missing participant data when the primary meta‐analysis of a specific outcome found a statistically significant effect.
Results
Description of studies
Results of the search
Figure 1 shows the study flow diagram. As of August 2017, the search strategy identified a total of 6947 unique citations. The title and abstract screening identified 192 potentially eligible citations. The full‐text screening of the full texts of these 192 citations identified 18 eligible RCTs published as full reports (Agnelli 2009 (PROTECHT); Agnelli 2012 (SAVE‐ONCO); Altinbas 2004; Haas 2012 (TOPIC 1); Haas 2012 (TOPIC 2); Kakkar 2004 (FAMOUS); Khorana 2017 (PHACS); Klerk 2005 (MALT); Lebeau 1994; Lecumberri 2013 (ABEL); Macbeth 2016 (FRAGMATIC); Maraveyas 2012 (FRAGEM); Pelzer 2015 (CONKO‐004); Perry 2010 (PRODIGE); Sideras 2006; van Doormaal 2011 (INPACT); Weber 2008; Zwicker 2013 (MICRO TEC)), and one RCT published as abstract (Vadhan‐Raj 2013) We had also identified two eligible studies published as abstracts but for which we were not able to obtain the necessary data from the authors: Salat 1990, Chazouilleres 1994,. We identified nine registered but unpublished trials: four completed (Borad 2011 (PGPC1); Germonpre 2008 (SYRINGES); Kakkar 2010 (GASTRANOX); Okuno 1999); two terminated (Chibauldel 2008 (PAM07); Pandya 2002); two ongoing (Lars 2008 (RASTEN); Meyer 2017 (PROVE)); and one withdrawn prior to enrolment (Rosovsky 2009).
Study flow diagram.
Included studies
The 19 included studies had 9650 participants. One study used unfractionated heparin as the intervention (Lebeau 1994), another used ultra‐LMWH (Agnelli 2012 (SAVE‐ONCO)), while the other 17 used LMWH as the intervention (Agnelli 2009 (PROTECHT); Altinbas 2004; Haas 2012 (TOPIC 1); Haas 2012 (TOPIC 2); Kakkar 2004 (FAMOUS); Khorana 2017 (PHACS); Klerk 2005 (MALT); Lecumberri 2013 (ABEL); Macbeth 2016 (FRAGMATIC); Maraveyas 2012 (FRAGEM); Pelzer 2015 (CONKO‐004); Perry 2010 (PRODIGE); Sideras 2006; Vadhan‐Raj 2013; van Doormaal 2011 (INPACT); Weber 2008; Zwicker 2013 (MICRO TEC)). We did not identify any study using fondaparinux as the intervention.
Agnelli and colleagues (PROTECHT trial) recruited 1150 ambulatory participants with metastatic or locally advanced cancer (Agnelli 2009 (PROTECHT)). Participants were randomized to receive a prophylactic dose of nadroparin or placebo, each with concomitant chemotherapy. The primary efficacy outcomes were symptomatic DVT, and PE. The secondary efficacy outcomes were asymptomatic thromboembolic events incidentally diagnosed, and survival at the end of study treatment and at 12 months. Study outcomes included survival, asymptomatic VTE, and minor and major bleeding. Follow‐up was about 90% in each group.
Agnelli and colleagues (SAVE‐ONCO trial) recruited 3212 participants with advanced metastatic or locally advanced cancer. Of the participants, 91% had an ECOG performance status of zero or one and 42% had at least one risk factor for VTE (Agnelli 2012 (SAVE‐ONCO)). Participants were randomized to receive either subcutaneous injection of semuloparin or placebo for a minimum of three months. The study outcomes included mortality, PE, symptomatic DVT, major bleeding and minor bleeding. Follow‐up data were available for 99% of participants for mortality and VTE outcomes. The minimum duration of follow‐up was up to three days after last injection, with a median of 3.5 months. The maximum duration of follow‐up was 12 months.
Altinbas and colleagues recruited 84 participants with both limited and extensive SCLC and an Eastern Cooperative Oncology Group (ECOG) status < 3 (Altinbas 2004). The ECOG performance Status scale ranges from zero (fully active) to five (dead) (Oken 1982). Participants were randomized to receive either a prophylactic dose of a LMWH (dalteparin) or placebo for 18 weeks or less, in combination with chemotherapy in case of disease progression. Study outcomes included mortality (at 12 and 24 months), symptomatic DVT and bleeding. Follow‐up was complete (100%). The minimum and maximum duration of follow‐up were two and 33 months, respectively. Hazard ratios (HRs) were estimated from published survival curves.
Haas and colleagues conducted two multi‐centre double‐blind studies and recruited 900 ambulatory participants receiving chemotherapy for disseminated metastatic breast carcinoma (Haas 2012 (TOPIC 1); n = 353) or stage III/IV NSCLC carcinoma (Haas 2012 (TOPIC 2); n = 547). Participants were randomized to receive either subcutaneous certoparin or placebo for six months. The study outcomes included mortality, confirmed VTE, PE, DVT, thrombocytopenia, major bleeding and minor bleeding. A number of participants were not included in the intention‐to‐treat (ITT) analysis but it is not reported whether they were followed up. The minimum duration of follow‐up was six months.
Kakkar and colleagues recruited 385 participants with advanced stage III or IV malignant disease of the breast, lung, gastrointestinal tract, pancreas, liver, genitourinary tract, ovary or uterus, and a minimum life expectancy of three months (Kakkar 2004 (FAMOUS)). Participants were randomized to receive either a prophylactic dose of a LMWH (dalteparin) or placebo for 12 months, with no restriction on concomitant chemotherapy or radiotherapy. The study outcomes included mortality (at 12, 24 and 36 months), symptomatic VTE (PE, DVT), major bleeding and minor bleeding. Follow‐up data were available for 374 participants (97%). The minimum duration of follow‐up was not reported. The maximum duration of follow‐up was 81 months. HRs were estimated from published survival curves, assuming all participants were followed up for 77 months.
Khorana and colleagues conducted a multi‐center study and recruited 98 participants with cancer and a Khorana risk score of ≥ 3 (Khorana 2017 (PHACS)). Participants were randomized to subcutaneous dalteparin or observation for a period of 12 weeks. The study outcomes included symptomatic DVT and PE, and clinically significant major and non‐major bleeding. Follow‐up was complete (100%). The study was terminated early due to low accrual.
Klerk and colleagues (MALT trial) recruited 302 participants with different types of advanced solid malignant tumors and a minimum life expectancy of one month (Klerk 2005 (MALT)). Participants were randomized to receive either a LMWH or a placebo for six weeks, each with concomitant chemotherapy or radiotherapy. Study outcomes included mortality (at six, 12 and 24 months), major bleeding, non‐major bleeding and combined major and non‐major bleeding. Follow‐up was complete (100%). The minimum duration of follow‐up was not reported, whereas the maximum duration was 84 months. The HR and its standard error were reported.
Lebeau and colleagues recruited 277 participants with both limited and extensive small cell lung cancer (SCLC), 78% of which had a Karnofsky Performance Scale Index > 80 (Lebeau 1994). The Karnofsky Performance Scale Index ranges from zero (dead) to 100 (normal) (Karnofsky 1948). Participants were randomized to receive either a prophylactic dose of UFH for five weeks or no intervention, in combination with chemotherapy. The study outcomes were mortality (at 12, 24 and 36 months) and bleeding. Follow‐up was complete (100%). The minimum duration of follow‐up was not reported. The maximum duration of follow‐up was 59 months. HRs were estimated from published survival curves, assuming all participants were followed up for 59 months.
Lecumberri and colleagues recruited 38 participants diagnosed with limited SCLC in a multicenter, open‐label study (Lecumberri 2013 (ABEL)). Participants were randomized to receive standard chemoradiotherapy or the same therapy plus bemiparin for a maximum of 26 weeks. The study outcomes included all‐cause mortality, incidence of VTE, major and minor bleeding, and thrombocytopenia. All outcomes were assessed at 18 months. Follow‐up was complete (100%).
Macbeth and colleagues recruited 2202 participants diagnosed with lung cancer (Macbeth 2016 (FRAGMATIC)). Participants were on standard anticancer treatment and randomized to subcutaneous dalteparin or no anticoagulation. The study outcomes included overall survival and bleeding. The median duration of follow‐up was 23.1 months.
Maraveyas and colleagues recruited 123 participants with non‐resectable, recurrent or metastatic pancreatic adenocarcinoma, Karnofsky performance status (KPS) of 60 to 100, and estimated life expectancy of more than 12 weeks (Maraveyas 2012 (FRAGEM)). Participants were randomized to receive either subcutaneous dalteparin or placebo. The study outcomes included mortality, all‐type VTE, DVT, and PE. Data from a range of 55 to 62 participants were used for different outcome assessments. All outcomes were assessed at 12 weeks and one year follow‐up.
Pelzer and colleagues recruited 312 chemotherapy‐naive participants with advanced pancreatic cancer (Pelzer 2015 (CONKO‐004)). Participants were randomized to receive or not to receive additional LMWH (enoxaparin) starting simultaneously with palliative systemic chemotherapy. Study outcomes included overall survival, symptomatic VTE, asymptomatic subclinical DVT and major bleeding. Follow‐up for overall survival was about 95.7% in the intervention group and 93.4% in the control group. The median duration of follow‐up was 30.4 weeks.
Perry and colleagues recruited 186 participants with newly diagnosed malignant glioma (Perry 2010 (PRODIGE)). Participants were randomized to receive a prophylactic dose of LMWH (dalteparin) or placebo. Study outcomes included objectively documented symptomatic DVT or PE (primary outcome), bleeding (major and all bleeding), quality of life and death. The duration of follow‐up was 12 months.
Sideras and colleagues recruited 141 participants with different types of advanced cancer and a minimum life expectancy of 12 weeks and ECOG state zero to two (Sideras 2006). Participants were randomized either to a prophylactic dose of a LMWH (dalteparin) or to placebo or no intervention. Study outcomes included overall survival (at 12, 24 and 36 months), VTE and major bleeding. Follow‐up data were available for 138 participants (98%). The minimum duration of follow‐up was not reported, whereas the maximum duration of follow‐up was 24 months. The authors supplied us with unpublished data, giving the HR and its standard error.
Vadhan‐Raj and colleagues recruited 75 participants with metastatic or locally advanced pancreatic cancer (Vadhan‐Raj 2013). Participants were randomized to receive dalteparin 5000 U SQ daily for 16 weeks during chemotherapy or chemotherapy alone. Assessed outcomes were VTE, DVT and PE. Participants were followed‐up for 16 weeks. The study reported complete follow up.
van Doormaal and colleagues recruited 503 participants with prostate carcinoma, NSCLC, or with a locally advanced pancreatic cancer (van Doormaal 2011 (INPACT)). Participants were randomized to receive either subcutaneous nadroparin or no nadroparin. The median duration of follow‐up was 10.5 months in the nadroparin group and 10.4 months in the control group. The study outcomes included mortality (at one, two and three years versus at five, 10, 15, 20, 25, 30, 35, 40 months), PE, DVT, major bleeding and minor bleeding. The percentage of participants lost to follow‐up was 0.8% and 3.5% from the nadroparin group and the control group respectively.
Weber and colleagues recruited 20 participants with advanced cancer and an estimated life expectancy of less than six months (Weber 2008). Participants were randomized to receive either a prophylactic dose of LMWH (nadroparin) or no treatment, each with concomitant anticancer treatment. Study outcomes included mortality, VTE (including PE and DVT), minor and major bleeding, and thrombocytopenia. Follow‐up was complete (100%). The minimum duration of follow‐up was reported as three months for mortality, whereas the maximum was 18 months for all outcomes.
Zwicker and colleagues recruited 34 participants with advanced cancer and high tissue factor‐bearing microparticles (Zwicker 2013 (MICRO TEC)). Participants were randomized to subcutaneous enoxaparin or observation. The study outcomes included incidence of symptomatic VTE for a follow‐up duration of two months. The trial was originally designed as a phase III, then re‐adapted to a phase II randomized clinical trial.
Excluded studies
We excluded 99 studies (130 reports) from this review for the following reasons: not population of interest (hospitalized): n = 11; not population of interest (surgical): n = 29; not population of interest (patients with central venous catheter (CVC)): n = 7; not population of interest (patients with VTE): n = 18; not population of interest (no participants with cancer): n = 2; not intervention of interest (oral): n = 9; not intervention of interest (aspirin): n = 7; not intervention of interest (different route of administration): n=4; not comparison of interest: n = 7; not design of interest: n = 33; not outcomes of interest: n = 3.
Risk of bias in included studies
The judgments for the risk of bias are summarized in Figure 2 and Figure 3, respectively.
'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.
Allocation
We judged allocation to be adequately concealed in 13 of the 19 studies (Agnelli 2009 (PROTECHT); Agnelli 2012 (SAVE‐ONCO); Haas 2012 (TOPIC 1); Haas 2012 (TOPIC 2); Kakkar 2004 (FAMOUS); Klerk 2005 (MALT); Lebeau 1994; Macbeth 2016 (FRAGMATIC); Pelzer 2015 (CONKO‐004); Perry 2010 (PRODIGE); Sideras 2006; van Doormaal 2011 (INPACT); Weber 2008), and not concealed in two studies (Altinbas 2004; Lecumberri 2013 (ABEL)). Four studies did not report on allocation concealment (Khorana 2017 (PHACS); Maraveyas 2012 (FRAGEM); Vadhan‐Raj 2013Zwicker 2013 (MICRO TEC)).
Blinding
Blinding of participants and personnel (performance bias)
We judged participants and personnel to be definitely blinded in three studies (Agnelli 2009 (PROTECHT); Klerk 2005 (MALT); Perry 2010 (PRODIGE) and probably blinded in four studies (Agnelli 2012 (SAVE‐ONCO); Haas 2012 (TOPIC 1); Haas 2012 (TOPIC 2); Kakkar 2004 (FAMOUS). We judged nine studies as definitely not blinded (Altinbas 2004; Khorana 2017 (PHACS); Lebeau 1994; Lecumberri 2013 (ABEL); Macbeth 2016 (FRAGMATIC); Pelzer 2015 (CONKO‐004); Sideras 2006; van Doormaal 2011 (INPACT); Weber 2008) and three as probably not blinded (Maraveyas 2012 (FRAGEM); Vadhan‐Raj 2013Zwicker 2013 (MICRO TEC).
Blinding of outcome assessment (detection bias)
We judged outcome assessors to be definitely blinded in two studies (Klerk 2005 (MALT); Perry 2010 (PRODIGE) and probably blinded in nine studies (Agnelli 2009 (PROTECHT); Agnelli 2012 (SAVE‐ONCO); Haas 2012 (TOPIC 1); Haas 2012 (TOPIC 2); Kakkar 2004 (FAMOUS); Khorana 2017 (PHACS); Lecumberri 2013 (ABEL); Pelzer 2015 (CONKO‐004); van Doormaal 2011 (INPACT). We judged four studies as definitely not blinded due to their open‐label design (Altinbas 2004; Macbeth 2016 (FRAGMATIC); Sideras 2006; Weber 2008) and four as probably not blinded. (Lebeau 1994; Maraveyas 2012 (FRAGEM); Vadhan‐Raj 2013Zwicker 2013 (MICRO TEC). However, we judged risk of bias in relation to detection bias as low when reporting on objective outcomes (for all 19 studies) and high when reporting on patient‐reported subjective outcomes (for two studies Macbeth 2016 (FRAGMATIC); Sideras 2006).
Incomplete outcome data
Eight studies reported a complete follow‐up rate (Altinbas 2004; Khorana 2017 (PHACS); Klerk 2005 (MALT); Lebeau 1994; Lecumberri 2013 (ABEL); Weber 2008;Vadhan‐Raj 2013; Zwicker 2013 (MICRO TEC).
Agnelli and colleagues reported an approximate 90% follow‐up rate in the PROTECHT trial (Agnelli 2009 (PROTECHT)). In the SAVE‐ONCO trial, follow‐up data were reported per outcome as follows: for mortality and VTE outcomes, approximately 99% in both the intervention and control groups; for bleeding outcome, 88% in the intervention group and 95% in the control group (Agnelli 2012 (SAVE‐ONCO)).
Kakkar and colleagues reported an approximate 97% follow‐up rate in both the intervention and contro groups (Kakkar 2004 (FAMOUS)). Pelzer and colleagues reported a 95% follow‐up rate in the intervention group and 93% in the control group for the outcome overall survival (Pelzer 2015 (CONKO‐004)). Sideras and colleagues reported a 98% follow‐up rate (Sideras 2006). van Doormaal and colleagues reported a 97.85% follow‐up rate (van Doormaal 2011 (INPACT)). Macbeth and colleagues reported a 94% follow‐up rate in the intervention group and 97% in the control group (Macbeth 2016 (FRAGMATIC)).
Only one study reported follow‐up data per outcome and not per participant (Maraveyas 2012 (FRAGEM)). The follow‐up rates for the outcomes overall survival, VTE incidence, and toxicity ranged between 93% and 98%.
In both studies by Haas and colleagues, it is not reported whether participants not included in the intention‐to‐treat (ITT) analyses were followed up (Haas 2012 (TOPIC 1); Haas 2012 (TOPIC 2)). All participants in the intervention group and 99% of participants in the control group were included in the analysis for TOPIC 1, whereas 98% in the intervention group and 97% in the control group were included for TOPIC 2.
In the study by Perry and colleagues, it is not reported whether participants were followed up among those that did not receive first dose, withdrew consent, or discontinued treatment (Perry 2010 (PRODIGE)). We judged the risk of attrition bias as high since those participants represent 37% of the intervention group and 53% of the control group.
Selective reporting
The outcomes listed in the methods section were reported in the results section for 13 studies (Agnelli 2009 (PROTECHT); Agnelli 2012 (SAVE‐ONCO); Haas 2012 (TOPIC 1); Haas 2012 (TOPIC 2); Kakkar 2004 (FAMOUS); Khorana 2017 (PHACS); Klerk 2005 (MALT); Lebeau 1994; Maraveyas 2012 (FRAGEM); Pelzer 2015 (CONKO‐004); Sideras 2006; van Doormaal 2011 (INPACT); Weber 2008). Seven studies are registered in ClinicalTrials.gov (Agnelli 2009 (PROTECHT); Agnelli 2012 (SAVE‐ONCO); Khorana 2017 (PHACS); Lecumberri 2013 (ABEL); Maraveyas 2012 (FRAGEM); Perry 2010 (PRODIGE); van Doormaal 2011 (INPACT)). One study is registered in the ISRCTN registry (Pelzer 2015 (CONKO‐004)).
One study reported on all outcomes except for two listed in the methods section (quality of life and cognition assessment) (Perry 2010 (PRODIGE)). The outcomes of interest were all reported but were not listed in the methods section for one study (Altinbas 2004).
One study had a published protocol and reported on all outcomes listed in the protocol (Pelzer 2015 (CONKO‐004)). One study that also had a published protocol reported on all outcomes listed in the protocol except for four that will be reported elsewhere (health economics, health‐related quality of life, dyspnea and biomarker studies) (Macbeth 2016 (FRAGMATIC)).
Selective reporting bias was unclear in the study published as an abstract (Vadhan‐Raj 2013).
Other potential sources of bias
We questioned whether in the study by Agnelli and colleagues the follow‐up time "occurring between randomization and 3 days after the last injection of the study drug" could have potentially led to differential follow‐up time between the two groups (Agnelli 2012 (SAVE‐ONCO)). However, the authors report that "the duration of treatment was similar in the two study groups, with a median of approximately 3.5 months".
Klerk and colleagues reported that "chemotherapy was more frequently administered during the period of study treatment in participants receiving placebo, whereas radiotherapy was more frequently given to participants receiving nadroparin"; thus 25% of the nadroparin group and 34% of the placebo group received chemotherapy; 32% of the nadroparin group and 18% of the placebo group received radiotherapy. Having different co‐interventions between the two groups might lead to performance bias (Klerk 2005 (MALT)).
Three studies were stopped early due to insufficient accrual (Khorana 2017 (PHACS); Perry 2010 (PRODIGE); Sideras 2006).
We judged that in the study by Lebeau and colleagues participants received similar co‐interventions although brain and thoracic irradiation depended on response to treatment. In that study, 11% and 7%, respectively of participants randomized to heparin and control groups received radiotherapy (Lebeau 1994).
In the study by Pelzer and colleagues, the related abstracts published in 2005 and 2007 reported a target recruitment of 540 patients whereas 312 patients were recruited into the trial (Pelzer 2015 (CONKO‐004)).
The study by Zwicker and colleagues was originally designed as a phase III randomized clinical trial then re‐adapted to a phase II trial. Also, the trial is described as underpowered (Zwicker 2013 (MICRO TEC)).
Effects of interventions
All‐cause mortality
All‐cause mortality at 12 months
Meta‐analysis of the 18 randomized controlled trials (RCTs), including 9575 participants, found that the use of heparin compared to no heparin has no effect on mortality rates at 12 months: risk ratio (RR) 0.98; 95% confidence interval (CI) 0.93 to 1.03; risk difference (RD) 10 fewer per 1000; 95% CI 35 fewer to 15 more (see Analysis 1.1). The I2 value indicates that the percentage of the variability in effect estimates that is due to heterogeneity rather than sampling error (chance) is moderate (I2 = 31%). The inverted funnel plot for the primary outcome of mortality at one year did not suggest publication bias, but there were relatively few trials to permit an accurate assessment (Figure 4). The certainty of evidence was moderate due to imprecision (summary of findings Table 1). Appendix 6 includes the GRADE Evidence Profile (a more detailed version of the summary of findings Table 1).
Funnel plot of comparison: 1 Heparin versus placebo, outcome: 1.1 Mortality at 12 months‐ Main analysis.
In a subgroup analysis of participants with lung cancer (either SCLC or NSCLC) (Altinbas 2004; Haas 2012 (TOPIC 2); Lebeau 1994; Lecumberri 2013 (ABEL); Macbeth 2016 (FRAGMATIC); van Doormaal 2011 (INPACT)), versus other types of cancer (that is neither SCLC or NSCLC) (Haas 2012 (TOPIC 1); Klerk 2005 (MALT); Maraveyas 2012 (FRAGEM); Pelzer 2015 (CONKO‐004); van Doormaal 2011 (INPACT); Weber 2008), the test for subgroup difference was not statistically significant (P value = 0.47).
In a subgroup analysis of participants with advanced cancer (including participants with extensive SCLC) (Agnelli 2009 (PROTECHT); Agnelli 2012 (SAVE‐ONCO); Altinbas 2004; Kakkar 2004 (FAMOUS); Klerk 2005 (MALT); Lebeau 1994; Maraveyas 2012 (FRAGEM); Pelzer 2015 (CONKO‐004); Sideras 2006; van Doormaal 2011 (INPACT); Weber 2008; Zwicker 2013 (MICRO TEC)), versus participants with non‐advanced cancer (including participants with limited SCLC) (Altinbas 2004; Haas 2012 (TOPIC 1); Haas 2012 (TOPIC 2); Khorana 2017 (PHACS); Lebeau 1994; Lecumberri 2013 (ABEL); Macbeth 2016 (FRAGMATIC); Perry 2010 (PRODIGE)), the test for subgroup effect was not statistically significant (P value = 0.56).
All‐cause mortality at 24 months
In a meta‐analysis of 14 RCTs, including 5229 participants, we found that heparin compared to no heparin has no effect on mortality rates at 24 months: RR 0.99; 95% CI 0.96 to 1.01; RD 8 fewer per 1000; 95% CI 31 fewer to 8 more (see Analysis 1.4). The I2 value indicates that the percentage of the variability in effect estimates that is due to heterogeneity rather than sampling error (chance) is moderate (I2 = 27%). The certainty of evidence was moderate due to imprecision (summary of findings Table 1).
In a subgroup analysis of participants with advanced cancer (including participants with extensive SCLC) (Kakkar 2004 (FAMOUS); Klerk 2005 (MALT); Pelzer 2015 (CONKO‐004); Sideras 2006; van Doormaal 2011 (INPACT); Weber 2008), versus participants with non‐advanced cancer (including participants with limited SCLC) (Altinbas 2004; Haas 2012 (TOPIC 1); Haas 2012 (TOPIC 2); Lebeau 1994; Lecumberri 2013 (ABEL); Macbeth 2016 (FRAGMATIC); Maraveyas 2012 (FRAGEM); Perry 2010 (PRODIGE)), the test for subgroup effect was not statistically significant (P value = 0.97)
All‐cause mortality ‐ time‐to‐event analysis
Fifteen studies, including 8388 participants, reported data allowing their inclusion in the time‐to‐event meta‐analysis. Meta‐analysis indicated that heparin compared to no heparin has no effect on reduction in the risk of death (hazard ratio (HR) 0.93; 95% CI 0.84 to 1.03) (see Analysis 1.6). The I2 value indicates that the percentage of the variability in effect estimates that is due to heterogeneity rather than sampling error (chance) may represent moderate heterogeneity (I2 = 64%).
Symptomatic venous thromboembolism (VTE)
Meta‐analysis of 16 RCTs, including 9036 participants, found that heparin reduces the risk of symptomatic VTE compared to no heparin: RR 0.56; 95% CI 0.47 to 0.68; RD 30 fewer per 1000; 36 fewer to 22 fewer (see Analysis 1.7). The I2 value indicates that the percentage of the variability in effect estimates that is due to heterogeneity rather than sampling error (chance) is not important (I2 = 0%). Results did not change in a sensitivity analysis including the study published as abstract (Vadhan‐Raj 2013): RR 0.56, 95% CI 0.46 to 0.67. Since the primary meta‐analysis found a statistically significant effect, and in order to assess the risk of bias associated with missing participant data, we conducted sensitivity meta‐analyses using the a priori plausible assumptions detailed in the Methods section. The effect estimate remained significant across all four stringent assumptions (Appendix 7). Analysis 1.9 and Analysis 1.10 respectively show the separate analyses for PE and symptomatic DVT. The inverted funnel plot for symptomatic VTE did not suggest publication bias, but there were relatively few trials to permit an accurate assessment (Figure 5). The certainty of evidence was high (summary of findings Table 1).
Funnel plot of comparison: 1 Heparin versus placebo, outcome: 1.7 Symptomatic VTE‐ Main analysis.
In a subgroup analysis of participants with lung cancer (either SCLC or NSCLC), (Agnelli 2009 (PROTECHT); Agnelli 2012 (SAVE‐ONCO); Altinbas 2004; Haas 2012 (TOPIC 2)Lecumberri 2013 (ABEL); Macbeth 2016 (FRAGMATIC)) versus participants with any type of cancer (that is neither SCLC or NSCLC), (Kakkar 2004 (FAMOUS); Khorana 2017 (PHACS); Sideras 2006; van Doormaal 2011 (INPACT); Zwicker 2013 (MICRO TEC)) the test for subgroup effect was not statistically significant (P value 0.21).
Major bleeding
Meta‐analysis of 18 RCTs, including 9592 participants, showed that heparin likely increases the risk of major bleeding compared to no heparin: RR 1.30; 95% CI 0.94 to 1.79; RD 4 more per 1000; 95% CI 1 fewer to 11 more) (see Analysis 1.11). The I2 value indicates that the percentage of the variability in effect estimates that is due to heterogeneity rather than sampling error (chance) may represent no heterogeneity (I2 = 0%). The certainty of evidence was moderate due to imprecision (summary of findings Table 1).
In a subgroup analysis of participants with lung cancer (either SCLC or NSCL) ( Altinbas 2004; Haas 2012 (TOPIC 2); Lebeau 1994; Lecumberri 2013 (ABEL); ; Macbeth 2016 (FRAGMATIC)), versus participants with any type of cancer (that is neither SCLC or NSCLC) (Haas 2012 (TOPIC 1); Klerk 2005 (MALT); Pelzer 2015 (CONKO‐004); Perry 2010 (PRODIGE); Weber 2008), the test for subgroup effect was not statistically significant (P value = 0.61).
Minor bleeding
Meta‐analysis of 16 RCTs, including 9245 participants, found that heparin causes an increase in the risk of minor bleeding compared to no heparin: RR 1.70; 95% CI 1.13 to 2.55; RD 17 more per 1000; 3 more to 37 more) (see Analysis 1.13). The I2 value indicates that the percentage of the variability in effect estimates that is due to heterogeneity rather than sampling error (chance) may represent moderate heterogeneity (I2 = 53%). Since the primary meta‐analysis found a statistically significant effect, and in order to assess the risk of bias associated with missing participant data, we conducted sensitivity meta‐analyses using the a priori plausible assumptions detailed in the Methods section. The effect estimate did not lose significance across all four stringent assumptions (Appendix 7). The certainty of evidence was high (summary of findings Table 1).
Thrombocytopenia
Meta‐analysis of 12 RCTs, including 5832 participants, failed to show or to exclude a beneficial or detrimental effect of heparin on the risk of thrombocytopenia compared to no heparin (RR 0.69; 95% CI 0.37 to 1.27; RD 33 fewer per 1000; 95% CI 66 fewer to 28 more) (see Analysis 1.14). The I2 value indicates that the percentage of the variability in effect estimates that is due to heterogeneity rather than chance may represent high heterogeneity (I2 = 83%). The certainty of evidence was moderate due to imprecision (summary of findings Table 1).
Health‐related quality of life
Two studies assessed quality of life, one using the Uniscale and the Symptom Distress Scale (SDS) (Sideras 2006), the other using the Hospital Anxiety and Depression Score and EQ‐5D (Macbeth 2016 (FRAGMATIC)). Both studies concluded that the scores for the two scales were similar for the two study groups, both at baseline and at follow‐up. The certainty of evidence was moderate due to risk of bias (summary of findings Table 1).
Sensitivity analyses
The sensitivity analysis excluding the one study at high risk of bias, Altinbas 2004, from the analyses did not change the results significantly. We have presented above the sensitivity meta‐analyses to assess the risk of bias associated with missing participant data.
Discussion
Summary of main results
Parenteral anticoagulation (with either unfractionated heparin or low molecular weight heparin (LMWH)) appears to have no effect on mortality in patients with cancer, who have no therapeutic or prophylactic indication for anticoagulation. While parenteral anticoagulation reduces venous thromboembolism (VTE), it likely increases major bleeding and minor bleeding. We did not identify any study using fondaparinux as an anticoagulant.
Overall completeness and applicability of evidence
The included studies recruited patients with a variety of cancer types and stages, which should increase the applicability of the results. The results apply best to LMWH, given that only one study evaluated unfractionated heparin. Unfortunately, not enough data were available to evaluate the impact of the intervention on bleeding outcomes or on quality of life. The latter outcome is important given the potential burden of daily subcutaneous injections.
As mentioned above, we identified three eligible studies for which we were not able to obtain the necessary data from the authors. Chazouilleres 1994 recruited 51 participants with unresectable hepatocellular carcinoma and reported a lower short‐term mortality rate with LMWH. Salat 1990 did not report on mortality outcome. Vadhan‐Raj 2013a randomized 75 participants with metastatic or locally advanced pancreatic cancer and reported a trend towards a reduction in VTE.
Quality of the evidence
Our systematic approach to searching, study selection and data extraction should have minimized the likelihood of missing relevant studies. The certainty of evidence was high for symptomatic VTE and minor bleeding, moderate for mortality, major bleeding and quality of life.
Potential biases in the review process
The inclusion of different types of cancer in the same study precluded us from conducting the subgroup analyses to explore effect modifiers such as type and stage of cancer. The interpretation of findings is also limited by not including data from the trials published as abstracts only. Also, for two studies we had to calculate the number of mortality events at 12 and 24 months from the survival curves (Altinbas 2004; Kakkar 2004 (FAMOUS)). Also, there might be potential bias associated with multiple testing in the planned meta‐analyses and currently there are no plans to adjust meta‐analyses for multiple testing.
Agreements and disagreements with other studies or reviews
A recent review by Che and colleagues assessed the effect of LMWH compared with no heparin in patients with cancer with no history of VTE (Che 2013). Similar to our findings, the review found that LMWH significantly reduced the risk of VTE and increased the risk of bleeding. Moreover, this study did not focus on the type of intervention or type of participants, for example the pooled participants included patients being started on thromboprophylaxis due the placement of a central venous catheter (CVC), or in the perioperative setting. Our review eligibility criteria focused on parenteral anticoagulation in ambulatory patients with cancer, i.e. reducing clinical heterogeneity.
Another Cochrane systematic review conducted by Di Nisio and colleagues assessed the efficacy and safety of primary thromboprophylaxis in ambulatory patients with cancer receiving chemotherapy (Di Nisio 2016). The review found that LMWH, when compared with inactive control, significantly reduced the incidence of symptomatic VTE, whereas there was no statistically significant effects on major bleeding, asymptomatic VTE, minor bleeding, one‐year mortality, symptomatic arterial thromboembolism, superficial thrombophlebitis or serious adverse events. The authors included various interventions for both prophylactic and therapeutic purposes in different populations. The interventions included parenteral anticoagulants (LMWH, unfractionated heparin), oral agents (Vitamin K antagonists (VKA), direct oral anticoagulants (DOAC), aspirin, antithrombin), and placebo. The populations included patients without VTE, with VTE, with multiple myeloma, and pediatrics. We tackled most of these comparisons in separate Cochrane reviews (Akl 2014 (initial); Akl 2014 (long‐term); Akl 2014 (oral))
Another recent publication by Phan and colleagues, studying the efficacy of heparin‐based medications for prevention of VTE, found a significant reduction in VTE with an odds ratio (OR) of 0.56 (95% confidence interval (CI) 0.45 to 0.71) (Phan 2014). However, that review had limitations in comparison to ours. That review did not include four studies we deemed to be eligible (Altinbas 2004; Pelzer 2015 (CONKO‐004); Sideras 2006; Weber 2008). The reported reason for not including two of these studies was that VTE was not assessed (Altinbas 2004; Sideras 2006). There was no reference to the two other studies (Sideras 2006; Weber 2008). Secondly, Phan 2014 included in the review the Young 2009 trial, assessing anticoagulation in patients with a CVC. This introduced increased clinical heterogeneity. We have included that trial in a separate Cochrane review evaluating prophylaxis for catheter‐related thrombosis (Akl 2014 (CVC)). Unlike the review conducted by Phan 2014, we did not include in the VTE meta‐analysis the trial conducted by Klerk and colleagues (Klerk 2005 (MALT)) because the number of VTE events reported pertains to participants who discontinued the study drug prematurely because they developed VTE; the paper does not report the total number of VTE observed in the trial. Moreover, Phan 2014 focused solely on VTE and did not assess other patient‐important outcomes, such as mortality.
Similary, another systematic review conducted by Ben Aharon and colleagues assessing the efficacy and safety of primary thromboprophylaxis with LMWH in ambulatory participants with solid malignancies (Ben‐Aharon 2014) found that primary prophylaxis with LMWH reduced symptomatic VTE (RR and the rate of PE especially in the subgroup of participants with lung and pancreatic cancers. They found no significant effect for anticoagulation on one‐year mortality or major bleeding.
Another systematic review conducted by Zhang and colleagues assessed whether anticoagulation improves survival and VTE outcomes in participants with lung cancer exclusively with no indication for anticoagulation (Zhang 2013). Anticoagulation showed a survival benefit, prolonged life expectancy, and reduced the risk of VTE in participants with lung cancer with no indication for anticoagulants, especially for those with SCLC, whereas our review included a wider range of patients with various types of cancer.
Study flow diagram.
'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.
Funnel plot of comparison: 1 Heparin versus placebo, outcome: 1.1 Mortality at 12 months‐ Main analysis.
Funnel plot of comparison: 1 Heparin versus placebo, outcome: 1.7 Symptomatic VTE‐ Main analysis.
Comparison 1: Heparin versus placebo, Outcome 1: Mortality at 12 months‐ Main analysis
Comparison 1: Heparin versus placebo, Outcome 2: Mortality at 12 months‐ Subgroups Lung vs non‐Lung Cancer
Comparison 1: Heparin versus placebo, Outcome 3: Mortality at 12 months‐ Subgroups Advanced vs non‐Advanced
Comparison 1: Heparin versus placebo, Outcome 4: Mortality at 24 months‐ Main Analysis
Comparison 1: Heparin versus placebo, Outcome 5: Mortality at 24 months‐ Subgroups Advanced vs non‐Advanced
Comparison 1: Heparin versus placebo, Outcome 6: Mortality over duration of study
Comparison 1: Heparin versus placebo, Outcome 7: Symptomatic VTE‐ Main analysis
Comparison 1: Heparin versus placebo, Outcome 8: Symptomatic VTE‐ Subgroups Lung vs non‐Lung Cancer
Comparison 1: Heparin versus placebo, Outcome 9: PE
Comparison 1: Heparin versus placebo, Outcome 10: Symptomatic DVT
Comparison 1: Heparin versus placebo, Outcome 11: Major bleeding‐ Main analysis
Comparison 1: Heparin versus placebo, Outcome 12: Major bleeding‐ Subgroups Lung vs non‐Lung Cancer
Comparison 1: Heparin versus placebo, Outcome 13: Minor bleeding
Comparison 1: Heparin versus placebo, Outcome 14: Thrombocytopenia
| Heparin prophylaxis compared with no prophylaxis in ambulatory patients with cancer without VTE receiving systemic therapy P: Ambulatory patients with cancer without VTE receiving systemic therapy I: Heparin prophylaxis C: No prophylaxis S: Outpatient | |||||
| Outcomes | № of participants | Quality of the evidence | Relative effect | Anticipated absolute effects* (95% CI) | |
|---|---|---|---|---|---|
| Risk with No prophylaxis | Risk difference with Heparin prophylaxis | ||||
| Mortality | 9575 | ⊕⊕⊕⊝ | RR 0.98 | Study population | |
| 504 per 1000 | 10 fewer per 1000 | ||||
| Mortality | 5229 | ⊕⊕⊕⊝ | RR 0.99 | Study population | |
| 778 per 1000 | 8 fewer per 1000 | ||||
| Symptomatic VTE | 9036 | ⊕⊕⊕⊕ | RR 0.56 | Study population | |
| 68 per 1000 | 30 fewer per 1000 | ||||
| Major bleeding | 9592 | ⊕⊕⊕⊝ | RR 1.30 | Study population | |
| 14 per 1000 | 4 more per 1000 | ||||
| Minor bleeding | 9245 | ⊕⊕⊕⊕ | RR 1.70 | Study population | |
| 24 per 1000 | 17 more per 1000 | ||||
| Thrombocytopenia | 5832 | ⊕⊕⊕⊝ | RR 0.69 | Study population | |
| 105 per 1000 | 33 fewer per 1000 | ||||
| Quality of life impairment | 2241 | ⊕⊕⊕⊝ | ‐ | Macbeth 2016 (FRAGMATIC): " There was no difference between the two groups with respect to quality‐adjusted life years gained in the first year... No difference in overall quality of life at 6 months (P = .94) or at 12 months (P = .89)... Overall quality of life did not change significantly over the study period".Sideras 2006: "The QOL and SDS scores were similar, both at baseline and during the protocol period, in patients randomized to receive LMWH vs those not randomized to receive LMWH." | |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence | |||||
| 1 Downgraded one level due to serious imprecision: Confidence interval includes values suggesting clinically significant benefit and values suggesting no effect. 2 Downgraded one level due to serious imprecision: Confidence interval includes values suggesting clinically significant harm and values suggesting no effect. 3 Downgraded one level due to serious imprecision: Confidence interval includes values suggesting clinically significant benefit and values suggesting harm. 4 Downgraded one level due to serious risk of bias: Both studies were open‐label studies (lack of blinding may impact the patient‐reported subjective outcomes) | |||||
| Term | Definition |
|---|---|
| Adjuvant therapy | Assisting in the amelioration or cure of disease |
| Anticoagulation | The process of hindering the clotting of blood especially by treatment with an anticoagulant |
| Antithrombotic | Used against or tending to prevent thrombosis (clotting) |
| Bacteremia | The presence of bacteria in the blood |
| Central venous line | Synthetic tube that is inserted into a central (large) vein of a patient to provide temporary intravenous access for the administration of fluid, medication or nutrients |
| Coagulation | Clotting |
| Deep vein thrombosis (DVT) | A condition marked by the formation of a thrombus within a deep vein (as of the leg or pelvis) that may be asymptomatic or be accompanied by symptoms (such as swelling and pain) and that is potentially life‐threatening if dislodgment of the thrombus results in pulmonary embolism |
| Fibrin | A white insoluble fibrous protein formed from fibrinogen by the action of thrombin, especially in the clotting of blood |
| Fondaparinux | An anticoagulant medication |
| Hemostatic system | The system that shortens the clotting time of blood and stops bleeding |
| Heparin | An enzyme occurring especially in the liver and lungs that prolongs the clotting time of blood by preventing the formation of fibrin. Two forms of heparin that are used as anticoagulant medications are: unfractionated heparin (UFH) and low molecular weight heparins (LMWH) |
| Impedance plethysmography | A technique that measures the change in blood volume (venous blood volume as well as the pulsation of the arteries) for a specific body segment |
| Kappa statistics | A measure of degree of non‐random agreement between observers and/or measurements of a specific categorical variable |
| Metastasis | The spread of cancer cells from the initial or primary site of disease to another part of the body |
| Oncogene | A gene having the potential to cause a normal cell to become cancerous |
| Osteoporosis | A condition that especially affects older women and is characterized by a decrease in bone mass with decreased density and enlargement of bone spaces producing porosity and brittleness |
| Parenteral nutrition | The practice of feeding a patient intravenously, circumventing the gut |
| Pulmonary embolism (PE) | Embolism of a pulmonary artery or one of its branches that is produced by foreign matter and most often a blood clot originating in a vein of the leg or pelvis and that is marked by labored breathing, chest pain, fainting, rapid heart rate, cyanosis, shock and sometimes death |
| Stroma | The supporting framework of an organ typically consisting of connective tissue |
| Thrombin | A proteolytic enzyme formed from prothrombin that facilitates the clotting of blood by catalyzing conversion of fibrinogen to fibrin |
| Thrombocytopenia | Persistent decrease in the number of blood platelets that is often associated with hemorrhagic conditions |
| Thrombosis | The formation or presence of a blood clot within a blood vessel |
| Vitamin K antagonists | Anticoagulant medications that are used for anticoagulation. Warfarin is a vitamin K antagonist |
| Warfarin | An anticoagulant medication that is a vitamin K antagonist, which is used for anticoagulation |
| Ximelagatran | An anticoagulant medication |
| LMWH | Generic name | Prophylactic dose | Therapeutic dose |
|---|---|---|---|
| Lovenox | Enoxaparin | 40 mg once daily | 1 mg/kg twice daily |
| Fragmin | Dalteparin | 2500 to 5000 units once daily | 200 U/kg once daily or |
| Innohep, Logiparin | Tinzaparin | 4500 units once daily | 90 U/kg twice daily |
| Fraxiparine | Nadroparin | 35 to 75 anti‐Xa international units/kg once daily | 175 anti‐Xa int. units/kg once daily |
| Certoparin | Sandoparin | 3000 anti‐Xa international units once daily | — |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Mortality at 12 months‐ Main analysis Show forest plot | 18 | 9575 | Risk Ratio (M‐H, Random, 95% CI) | 0.98 [0.93, 1.03] |
| 1.2 Mortality at 12 months‐ Subgroups Lung vs non‐Lung Cancer Show forest plot | 12 | 4768 | Risk Ratio (M‐H, Random, 95% CI) | 0.94 [0.86, 1.03] |
| 1.2.1 Lung Cancer (SCLC or NSCLC) | 6 | 3204 | Risk Ratio (M‐H, Random, 95% CI) | 0.89 [0.73, 1.08] |
| 1.2.2 non‐Lung Cancer | 7 | 1564 | Risk Ratio (M‐H, Random, 95% CI) | 0.95 [0.88, 1.03] |
| 1.3 Mortality at 12 months‐ Subgroups Advanced vs non‐Advanced Show forest plot | 18 | 9575 | Risk Ratio (M‐H, Random, 95% CI) | 0.98 [0.94, 1.03] |
| 1.3.1 Advanced cancer | 12 | 6115 | Risk Ratio (M‐H, Random, 95% CI) | 0.98 [0.93, 1.02] |
| 1.3.2 Non‐advanced cancer | 8 | 3460 | Risk Ratio (M‐H, Random, 95% CI) | 0.92 [0.75, 1.12] |
| 1.4 Mortality at 24 months‐ Main Analysis Show forest plot | 14 | 5229 | Risk Ratio (M‐H, Random, 95% CI) | 0.99 [0.96, 1.01] |
| 1.5 Mortality at 24 months‐ Subgroups Advanced vs non‐Advanced Show forest plot | 14 | 5229 | Risk Ratio (M‐H, Random, 95% CI) | 0.99 [0.96, 1.01] |
| 1.5.1 Advanced cancer | 6 | 1554 | Risk Ratio (M‐H, Random, 95% CI) | 0.98 [0.93, 1.03] |
| 1.5.2 Non‐advanced cancer | 8 | 3675 | Risk Ratio (M‐H, Random, 95% CI) | 0.98 [0.93, 1.04] |
| 1.6 Mortality over duration of study Show forest plot | 15 | Hazard Ratio (IV, Random, 95% CI) | 0.93 [0.84, 1.03] | |
| 1.7 Symptomatic VTE‐ Main analysis Show forest plot | 16 | 9036 | Risk Ratio (M‐H, Random, 95% CI) | 0.56 [0.47, 0.68] |
| 1.8 Symptomatic VTE‐ Subgroups Lung vs non‐Lung Cancer Show forest plot | 11 | 8090 | Risk Ratio (M‐H, Random, 95% CI) | 0.52 [0.43, 0.63] |
| 1.8.1 Lung Cancer (SCLC or NSCLC) | 6 | 4217 | Risk Ratio (M‐H, Random, 95% CI) | 0.53 [0.41, 0.68] |
| 1.8.2 Non‐lung cancer | 7 | 3873 | Risk Ratio (M‐H, Random, 95% CI) | 0.51 [0.37, 0.70] |
| 1.9 PE Show forest plot | 14 | 8867 | Risk Ratio (M‐H, Random, 95% CI) | 0.61 [0.47, 0.80] |
| 1.10 Symptomatic DVT Show forest plot | 14 | 8867 | Risk Ratio (M‐H, Random, 95% CI) | 0.46 [0.33, 0.63] |
| 1.11 Major bleeding‐ Main analysis Show forest plot | 18 | 9592 | Risk Ratio (M‐H, Random, 95% CI) | 1.30 [0.94, 1.79] |
| 1.12 Major bleeding‐ Subgroups Lung vs non‐Lung Cancer Show forest plot | 10 | 4163 | Risk Ratio (M‐H, Random, 95% CI) | 1.62 [1.02, 2.56] |
| 1.12.1 Lung Cancer (SCLC or NSCLC) | 5 | 3035 | Risk Ratio (M‐H, Random, 95% CI) | 1.45 [0.77, 2.73] |
| 1.12.2 Non‐lung cancer | 5 | 1128 | Risk Ratio (M‐H, Random, 95% CI) | 1.84 [0.94, 3.58] |
| 1.13 Minor bleeding Show forest plot | 16 | 9245 | Risk Ratio (M‐H, Random, 95% CI) | 1.70 [1.13, 2.55] |
| 1.14 Thrombocytopenia Show forest plot | 12 | 5832 | Risk Ratio (M‐H, Random, 95% CI) | 0.69 [0.37, 1.27] |
