Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Review - Intervention

Anticoagulación para el tratamiento a largo plazo de la tromboembolia venosa en personas con cáncer

Authors' declarations of interest

Version published: 19 June 2018 Version history

https://doi.org/10.1002/14651858.CD006650.pub5
Collapse all Expand all

Resumen

available in

Antecedentes

El cáncer aumenta el riesgo de eventos tromboembólicos, especialmente en personas bajo tratamiento anticoagulante.

Objetivos

Comparar la eficacia y la seguridad de la heparina de bajo peso molecular (HBPM) y los anticoagulantes orales directos (ACOD), los antagonistas de la vitamina K [AVK] y otros anticoagulantes para el tratamiento a largo plazo de la tromboembolia venosa (TEV) en personas con cáncer.

Métodos de búsqueda

Se realizó una búsqueda bibliográfica que incluyó una búsqueda electrónica exhaustiva en el Registro Cochrane central de ensayos controlados (Cochrane Central Register of Controlled Trials), MEDLINE (Ovid) y Embase (Ovid); una búsqueda manual de resúmenes de congresos; la verificación de las referencias de los estudios incluidos y una búsqueda de estudios en curso en registros de ensayos. Como parte del abordaje de la revisión sistemática continua, se realizan búsquedas en forma continua, y se incorpora la nueva evidencia después de que se identifica. Fecha de la última búsqueda: 14 de mayo 2021.

Criterios de selección

Ensayos controlados aleatorizados (ECA) que evalúan los efectos beneficiosos y perjudiciales del tratamiento a largo plazo con HBPM, ACOD, AVK u otros anticoagulantes en personas con cáncer y TEV sintomática.

Obtención y análisis de los datos

Se extrajeron datos por duplicado sobre el riesgo del sesgo y las características del estudio. Los desenlaces incluyeron: mortalidad por todas las causas, TEV recurrente, hemorragia grave, hemorragia leve, trombocitopenia y calidad de vida (CdV) relacionada con la salud. La certeza de la evidencia respecto del desenlace se evaluó con el método GRADE (GRADE handbook [manual GRADE]).

Resultados principales

De 3583 citas, 19 ECA cumplieron los criterios de elegibilidad.

Heparinas de bajo peso molecular versus antagonistas de la vitamina K
Ocho estudios que incorporaron a 2327 participantes compararon HBPM con los AVK. El metanálisis de cinco estudios probablemente no descartó un efecto beneficioso ni perjudicial de las HBPM en comparación con los AVK sobre la mortalidad hasta los 12 meses de seguimiento (razón de riesgos [RR] 1,00; intervalo de confianza [IC] del 95%: 0,88 a 1,13; diferencia de riesgos [DR] 0 menos por 1000; IC del 95%: 45 menos a 48 más; evidencia de certeza moderada). El metanálisis de cuatro estudios no descartó un efecto beneficioso ni perjudicial de las HBPM en comparación con los AVK sobre la hemorragia grave (RR 1,09; IC del 95%: 0,55 a 2,12; DR 4 más por 1000; IC del 95%: 19 menos a 48 más; evidencia de certeza moderada) ni sobre la hemorragia leve (RR 0,78; IC del 95%: 0,47 a 1,27; DR 38 menos por 1000; IC del 95%: 92 menos a 47 más; evidencia de certeza baja) o sobre la trombocitopenia (RR 0,94; IC del 95%: 0,52 a 1,69). El metanálisis de cinco estudios reveló que las HBPM probablemente redujeron la recurrencia de TEV en comparación con los AVK (RR 0,58; IC del 95%: 0,43 a 0,77; DR 53 menos por 1000; IC del 95%: 29 menos a 72 menos, evidencia de certeza moderada).

Anticoagulantes orales directos versus antagonistas de la vitamina K
Cinco estudios que incorporaron a 982 participantes compararon los ACOD con los AVK. El metanálisis de cuatro estudios posiblemente no descarte los efectos beneficiosos ni perjudiciales de los ADOC en comparación con los AVK sobre la mortalidad (RR 0,93; IC del 95%: 0,71 a 1,21; DR 12 menos por 1000; IC del 95%: 51 menos a 37 más; evidencia de certeza baja), la TEV recurrente (RR 0,66; IC del 95%: 0,33 a 1,31; DR 14 menos por 1000; IC del 95%: 27 menos a 12 más; evidencia de certeza baja), hemorragia grave (RR 0,77; IC del 95%: 0,38 a 1,57; DR 8 menos por 1000, IC del 95%: 22 menos a 20 más; evidencia de certeza baja), o sobre la hemorragia leve (RR 0,84; IC del 95%: 0,58 a 1,22; DR 21 menos por 1000; IC del 95%: 54 menos a 28 más; evidencia de certeza baja). Un estudio que informó sobre ACOD versus AVK se publicó como resumen, por lo tanto no se incluyó en el análisis principal.

Anticoagulantes orales directos versus heparinas de bajo peso molecular
Dos estudios que incorporaron a 1455 participantes compararon ACOD con HBPM. El estudio de Raskob no descartó un efecto beneficioso ni perjudicial de los ACOD en comparación con la HBPM sobre la mortalidad hasta los 12 meses de seguimiento (RR 1,07; IC del 95%: 0,92 a 1,25; DR 27 más por 1000; IC del 95%: 30 menos a 95 más; evidencia de certeza baja). Los datos también hallaron que posiblemente los ACOD revelaron una reducción en la recurrencia de TEV hasta 12 meses de seguimiento en comparación con la HBPM (RR 0,69; IC del 95%: 0,47 a 1,01; DR 36 menos por 1000; IC del 95%: 62 menos a 1 más; evidencia de certeza baja). Los ACOD podrían haber aumentado la hemorragia grave a los 12 meses de seguimiento en comparación con la HBPM (RR 1,71; IC del 95%: 1,01 a 2,88; DR 29 más por 1000; IC del 95%: 0 menos a 78 más; evidencia de certeza baja) y es probable que hayan aumentado la hemorragia leve hasta los 12 meses de seguimiento en comparación con la HBPM (RR 1,31; IC del 95%: 0,95 a 1,80; DR 35 más por 1000; IC del 95%: 6 menos a 92 más; evidencia de certeza baja). El segundo estudio sobre ACOD versus HBPM se publicó como resumen, por lo tanto no se incluyó en el análisis principal.

Idraparinux versus antagonistas de la vitamina K
Un ECA con 284 participantes comparó la inyección subcutánea semanal de idraparinux versus tratamiento estándar (anticoagulación parenteral seguida de warfarina o acenocumarol) durante tres o seis meses. Es posible que los datos no hayan descartado un efecto beneficioso ni perjudicial de idraparinux en comparación con los AVK sobre la mortalidad a los seis meses (RR 1,11; IC del 95%: 0,78 a 1,59; DR 31 más por 1000; IC del 95%: 62 menos a 167 más; evidencia de certeza moderada), la recurrencia de TEV a los seis meses (RR 0,46; IC del 95%: 0,16 a 1,32; DR 42 menos por 1000; IC del 95%: 65 menos a 25 más; evidencia de certeza baja) o sobre la hemorragia grave (RR 1,11; IC del 95%: 0,35 a 3,56; DR 4 más por 1000; IC del 95%: 25 menos a 98 más, evidencia de certeza baja).

Conclusiones de los autores

Para el tratamiento a largo plazo de TEV en personas con cáncer, la evidencia revela que las HBPM en comparación con los AVK probablemente produzcan una reducción importante en la TEV; además, los ACOD en comparación con las HBPM, probablemente reduzcan la TEV, aunque podrían aumentar el riesgo de hemorragia grave. La decisión de que una persona con cáncer y TEV comience con HBPM a largo plazo versus anticoagulación oral debe sopesar los efectos beneficiosos y perjudiciales, y debe integrar los valores y las preferencias del paciente respecto de los desenlaces importantes y las estrategias de tratamiento alternativas.

Nota editorial: esta es una revisión sistemática continua (living systematic review; LSR). Las LSR ofrecen nuevos métodos para la actualización de revisiones en que la revisión se actualiza continuamente, con la incorporación de nueva evidencia relevante a medida que esté disponible. Consultar el estado actual de esta revisión en la Base de Datos Cochrane de Revisiones Sistemáticas (Cochrane Database of Systematic Reviews).

PICOs

Population
Intervention
Comparison
Outcome

The PICO model is widely used and taught in evidence-based health care as a strategy for formulating questions and search strategies and for characterizing clinical studies or meta-analyses. PICO stands for four different potential components of a clinical question: Patient, Population or Problem; Intervention; Comparison; Outcome.

See more on using PICO in the Cochrane Handbook.

Resumen en términos sencillos

available in

Anticoagulantes para el tratamiento a largo plazo de los coágulos sanguíneos en las personas con cáncer

Antecedentes
Las personas con cáncer tienen mayor riesgo de presentar coágulos sanguíneos y podrían tener una respuesta diferente a los distintos tipos de anticoagulantes.

Características de los estudios
Se realizaron búsquedas en las bases de datos de ensayos clínicos sobre los efectos del tratamiento a largo plazo con diferentes anticoagulantes sobre la recurrencia de trombos en personas con cáncer y con un diagnóstico confirmado de trombosis venosa profunda (un coágulo sanguíneo en las extremidades) o de embolia pulmonar (un coágulo sanguíneo en los pulmones). Se incluyeron ensayos con cualquier tipo de cáncer, independientemente del tipo de tratamiento oncológico. Los ensayos analizaron la supervivencia, la recurrencia de coágulos sanguíneos, la hemorragia y los niveles sanguíneos de plaquetas (que participan en la coagulación de la sangre). La evidencia estaba actualizada hasta mayo de 2021.

Resultados clave
Se encontraron 18 ensayos que incorporaron a participantes con cáncer y coágulos sanguíneos. Los estudios hallaron que las heparinas de bajo peso molecular (HBPM; un tipo de anticoagulante que se inyecta en una vena) eran superiores a los antagonistas de la vitamina K (AVK, un tipo de anticoagulante administrado por vía oral) para reducir la recurrencia de coágulos sanguíneos. Los datos disponibles no proporcionaron una respuesta clara sobre los efectos de estos fármacos en la mortalidad ni los efectos secundarios de hemorragia. Los estudios también hallaron que los anticoagulantes orales de acción directa (ACOD, otro tipo de anticoagulante administrado por vía oral) podrían disminuir la recurrencia de coágulos sanguíneos en comparación con la HBPM, y al mismo tiempo podrían aumentar el riesgo de hemorragia. No hubo una respuesta clara cuando se compararon los ACOD (un nuevo tipo de anticoagulante oral) y los AVK (un tipo de anticoagulante oral más antiguo) respecto de la mortalidad, la recurrencia de coágulos sanguíneos y la hemorragia.

Fiabilidad de la evidencia
Cuando se compararon las HBPM con los AVK, se consideró que la certeza de la evidencia era moderada para la recurrencia de coágulos sanguíneos, la mortalidad al año y la hemorragia grave, y baja para la hemorragia leve.

Cuando se compararon los ACOD con los AVK, se consideró que la certeza de la evidencia era baja para la mortalidad, la recurrencia de coágulos sanguíneos y las complicaciones hemorrágicas.

Nota editorial: esta es una revisión sistemática continua. Las revisiones sistemáticas continuas ofrecen un nuevo método para la actualización de revisiones en el cual la revisión se actualiza continuamente, con la incorporación de nueva evidencia relevante según esté disponible. Consultar el estado actual de esta revisión en la Base de Datos Cochrane de Revisiones Sistemáticas (Cochrane Database of Systematic Reviews).

Authors' conclusions

Implications for practice

The decision for a person with cancer and venous thromboembolism (VTE) to start long‐term low molecular weight heparin (LMWHs) treatment or oral anticoagulation treatment should balance the benefits and harms and integrate people's values and preferences for outcomes and management options (Haynes 2002). While DOACs compared to LMWHs may show a likely reduction in VTE recurrence, it may show an increase in major bleeding.

Implications for research

There is a need for research assessing patients' values and preferences regarding long‐term anticoagulant agents for treating VTE. Researchers should consider making the raw data from randomized controlled trials (RCTs) available for individual participant data meta‐analysis. Further RCTs including subgroups of people with cancer should report separate results for these subgroups.

Summary of findings

Open in table viewer
Summary of findings 1. Low molecular weight heparin secondary prophylaxis compared to vitamin K antagonist secondary prophylaxis in people with cancer with venous thromboembolism

Low molecular weight heparin secondary prophylaxis compared to vitamin K antagonist secondary prophylaxis in patients with cancer with venous thromboembolism

Population: People with cancer with venous thromboembolism
Setting: Outpatient
Intervention: LMWH secondary prophylaxis
Comparison: VKA secondary prophylaxis

Outcomes

№ of participants
(studies)
Follow up

Certainty of the evidence
(GRADE)

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Risk with VKA secondary prophylaxis

Risk difference with LMWH secondary prophylaxis

All‐cause mortality (main analysis ‐ active cancer)
follow up: 6 months

1712
(4 RCTs)

⊕⊕⊝⊝
LOW 1 2

RR 0.99
(0.88 to 1.12)

Study population

374 per 1,000

4 fewer per 1,000
(45 fewer to 45 more)

All‐cause mortality (time‐to‐event)

1243
(2 RCTs)

⊕⊕⊝⊝
LOW 2 3

HR 0.94
(0.74 to 1.20)

Study population

374 per 1,000

18 fewer per 1,000
(81 fewer to 56 more)

Recurrent venous thromboembolism (main analysis ‐ active cancer)
follow up: 6 months

1712
(4 RCTs)

⊕⊕⊕⊝
MODERATE 1

RR 0.59
(0.44 to 0.80)

Study population

124 per 1,000

51 fewer per 1,000
(69 fewer to 25 fewer)

Recurrent venous thromboembolism (time‐to‐event)

1243
(2 RCTs)

⊕⊕⊕⊝
MODERATE 3

HR 0.49
(0.31 to 0.78)

Study population

124 per 1,000

61 fewer per 1,000
(84 fewer to 26 fewer)

Major bleeding (main analysis ‐ active cancer)
follow up: 6 months

1712
(4 RCTs)

⊕⊕⊝⊝
LOW 1 2 4

RR 1.09
(0.55 to 2.12)

Study population

43 per 1,000

4 more per 1,000
(19 fewer to 48 more)

Minor bleeding (main analysis ‐ active cancer)
follow up: 6 months

1712
(4 RCTs)

⊕⊝⊝⊝
VERY LOW 1 2 5

RR 0.78
(0.47 to 1.27)

Study population

174 per 1,000

38 fewer per 1,000
(92 fewer to 47 more)

Thrombocytopenia (main analysis‐ active cancer)
follow up: 6 months

138
(1 RCT)

⊕⊕⊝⊝
LOW 3 6

RR 0.94
(0.52 to 1.69)

Study population

254 per 1,000

15 fewer per 1,000
(122 fewer to 175 more)

*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).

CI: Confidence interval; RR: Risk ratio; OR: Odds ratio;

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1 Downgraded by one level due to serious risk of bias (allocation concealment unclear in one study, lack of blinding of participants and personnel in all the four studies, high risk of incomplete outcome data in one study, and high risk of selective reporting in one study).

2 Downgraded by one level due to serious imprecision. Confidence interval includes suggests both potential harm and potential benefit.

3 Some concern with lack of blinding of patients and personnel.

4 Some concern with inconsistency. I2= 46%

5 Downgraded by one level due to serious inconsistency (I2= 78%)

6 Downgraded by two levels due to very serious imprecision. Confidence interval includes suggests both potential harm and potential benefit. Low number of events.

Open in table viewer
Summary of findings 2. Direct oral anticoagulant secondary prophylaxis compared to Vitamin K antagonist secondary prophylaxis in patients with active cancer with venous thromboembolism

Direct oral anticoagulant secondary prophylaxis compared to Vitamin K antagonist secondary prophylaxis in patients with active cancer with venous thromboembolism

Population: patients with active cancer with venous thromboembolism
Setting: Outpatient
Intervention: DOAC secondary prophylaxis
Comparison: VKA secondary prophylaxis

Outcomes

№ of participants
(studies)
Follow up

Certainty of the evidence
(GRADE)

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Risk with Vitamin K antagonist (VKA) secondary prophylaxis

Risk difference with Direct oral anticoagulant (DOAC) secondary prophylaxis

All‐cause mortality
follow up: range 6 months to 12 months

1060
(4 RCTs)

⊕⊕⊝⊝
LOW 1 2

RR 0.94
(0.72 to 1.23)

Study population

171 per 1,000

10 fewer per 1,000
(48 fewer to 39 more)

Recurrent venous thromboembolism
follow up: range 6 months to 12 months

1050
(4 RCTs)

⊕⊕⊝⊝
LOW 1 3

RR 0.63
(0.34 to 1.15)

Study population

49 per 1,000

18 fewer per 1,000
(32 fewer to 7 more)

Major bleeding
follow up: range 6 months to 12 months

1055
(4 RCTs)

⊕⊕⊝⊝
LOW 1 2

RR 0.77
(0.39 to 1.53)

Study population

37 per 1,000

8 fewer per 1,000
(23 fewer to 20 more)

Minor bleeding
follow up: range 6 months to 12 months

1055
(4 RCTs)

⊕⊕⊝⊝
LOW 1 2

RR 0.83
(0.57 to 1.23)

Study population

127 per 1,000

22 fewer per 1,000
(55 fewer to 29 more)

Thrombocytopenia ‐ not reported

‐

‐

‐

‐

‐

Health related quality of life
follow up: range 3 months to 12 months

8485
(1 RCT)

⊕⊕⊕⊝
MODERATE 4

‐

Prins 2014 (EINSTEIN DVT‐PE; n=8485 ): "in the general population of the EINSTEIN studies, patient‐reported satisfaction and quality of life was better in the rivaroxaban‐treated patients than in the group treated with enoxaparin and vitamin K antagonist, although we have not yet examined whether this is the same in patients with active cancer. Hence, it can be expected that quality of life will also be improved with rivaroxaban compared with long‐term injected low molecular‐weight heparin." The tool used was validated measure of treatment satisfaction – the Anti‐Clot Treatment Scale (ACTS))

*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).

CI: Confidence interval; RR: Risk ratio; OR: Odds ratio;

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1 Some concern with indirectness (study by Schulman et al (RECOVER I‐II) included patients with a diagnosis of cancer within five years before enrollment), however the weight of these studies was low and heterogeneity was very low.

2 Downgraded by two levels due to very serious imprecision. Confidence interval suggests both potential benefit and potential harm.

3 Downgraded by two levels due to very serious imprecision. Confidence interval suggests both potential benefit and potential no effect. Low number of events.

4 Downgraded by one level for serious indirectness. The study by Prins and colleagues (Prins 2014 ( EINSTEIN n=8485)) reports health related quality of life for the whole study population, without providing data for the cancer subgroup

Open in table viewer
Summary of findings 4. Direct oral anticoagulant secondary prophylaxis compared to Low molecular weight heparin secondary prophylaxis in patients with cancer with venous thromboembolism

Direct oral anticoagulant secondary prophylaxis compared to Low molecular weight heparin secondary prophylaxis in patients with cancer with venous thromboembolism

Patient or population: patients with cancer with venous thromboembolism
Setting: Outpatient
Intervention: DOAC secondary prophylaxis
Comparison: LMWH secondary prophylaxis

Outcomes

№ of participants
(studies)
Follow up

Certainty of the evidence
(GRADE)

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Risk with Low molecular weight heparin (LMWH) secondary prophylaxis

Risk difference with Direct oral anticoagulant (DOAC) secondary prophylaxis

All‐cause mortality
follow up: mean 6 months

2854
(5 RCTs)

⊕⊕⊝⊝
LOW 1 2

RR 0.97
(0.83 to 1.14)

Study population

248 per 1,000

7 fewer per 1,000
(42 fewer to 35 more)

Recurrent VTE
follow up: mean 6 months

2854
(5 RCTs)

⊕⊕⊝⊝
LOW 1 3

RR 0.63
(0.45 to 0.88)

Study population

87 per 1,000

32 fewer per 1,000
(48 fewer to 10 fewer)

Major bleeding
follow up: mean 6 months

2994
(5 RCTs)

⊕⊕⊝⊝
LOW 1 4

RR 1.20
(0.83 to 1.73)

Study population

35 per 1,000

7 more per 1,000
(6 fewer to 25 more)

Major GI bleeding
follow up: mean 6 months

1838
(4 RCTs)

⊕⊕⊝⊝
LOW 1 4

RR 1.16
(0.62 to 2.17)

Study population

20 per 1,000

3 more per 1,000
(8 fewer to 23 more)

Major upper GI bleeding
follow up: mean 6 months

1838
(4 RCTs)

⊕⊝⊝⊝
VERY LOW 1 5

RR 1.18
(0.51 to 2.76)

Study population

11 per 1,000

2 more per 1,000
(5 fewer to 19 more)

Major lower GI bleeding
follow up: mean 6 months

1838
(4 RCTs)

⊕⊝⊝⊝
VERY LOW 1 5

RR 1.10
(0.43 to 2.80)

Study population

9 per 1,000

1 more per 1,000
(5 fewer to 16 more)

Major non‐GI bleeding
follow up: mean 6 months

1838
(4 RCTs)

⊕⊝⊝⊝
VERY LOW 1 6

RR 0.84
(0.42 to 1.68)

Study population

19 per 1,000

3 fewer per 1,000
(11 fewer to 13 more)

Minor bleeding
follow up: mean 6 months

2854
(5 RCTs)

⊕⊕⊕⊝
MODERATE 1

RR 1.58
(1.15 to 2.16)

Study population

67 per 1,000

39 more per 1,000
(10 more to 78 more)

Minor GI bleeding
follow up: mean 6 months

1495
(2 RCTs)

⊕⊝⊝⊝
VERY LOW 6 7 8

RR 1.37
(0.41 to 4.64)

Study population

25 per 1,000

9 more per 1,000
(15 fewer to 92 more)

Minor upper GI bleeding
follow up: mean 6 months

1495
(2 RCTs)

⊕⊝⊝⊝
VERY LOW 6 7 9

RR 1.03
(0.04 to 25.97)

Study population

11 per 1,000

0 fewer per 1,000
(10 fewer to 267 more)

Minor lower GI bleeding
follow up: mean 6 months

1495
(2 RCTs)

⊕⊕⊝⊝
LOW 5 7

RR 1.54
(0.72 to 3.27)

Study population

15 per 1,000

8 more per 1,000
(4 fewer to 33 more)

Minor non‐GI bleeding
follow up: mean 6 months

1495
(2 RCTs)

⊕⊕⊕⊝
MODERATE 7 10

RR 2.37
(1.44 to 3.89)

Study population

31 per 1,000

42 more per 1,000
(14 more to 89 more)

*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).

CI: Confidence interval; RR: Risk ratio; OR: Odds ratio;

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1 Downgraded by one level due to serious risk of bias. Allocation concealment was not reported in one study, lack of blinding of patients and personnel in all studies, and high risk of bias related to incomplete outcome data.

2 Downgraded by one level due to serious imprecision. Confidence interval suggests both potential benefit and potential harm.

3 Downgraded by one level due to serious imprecision. Confidence interval suggests both potential benefit and potential no effect.

4 Downgraded by one level due to serious imprecision. Confidence interval suggests both potential harm and potential no effect.

5 Downgraded by two levels due to very serious imprecision. Confidence interval suggests both potential harm and potential no effect. Low number of events.

6 Downgraded by two levels due to very serious imprecision. Confidence interval suggests both potential harm and potential benefit. Low number of events.

7 Some concern with risk of bias. Lack of blinding of patients and personnel in both studies.

8 Downgraded by one level due to serious inconsistency (unexplained heterogeneity I2=64%) and due to some concern with risk of bias.

9 Downgraded by two levels due to very serious inconsistency (unexplained heterogeneity I2=74%.) and due to some concern with risk of bias.

10 Downgraded by one level due to serious imprecision. Low number of events.

Open in table viewer
Summary of findings 5. Idraparinux secondary prophylaxis compared to vitamin K antagonist secondary prophylaxis in people with cancer with venous thromboembolism

Idraparinux secondary prophylaxis compared to VKA secondary prophylaxis in people with cancer with VTE

Population: people with cancer with VTE receiving secondary prophylaxis

Setting: outpatient

Intervention: idraparinux prophylaxis

Control: VKA prophylaxis

Outcomes

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Risk with VKA secondary prophylaxis

Risk difference with idraparinux secondary prophylaxis

All‐cause mortality
follow‐up: mean 6 months

284
(1 RCT)

⊕⊕⊕⊝
Moderatea

RR 1.11
(0.78 to 1.59)

Study population

283 per 1000

31 more per 1000
(62 fewer to 167 more)

Recurrent VTE
follow‐up: mean 6 months

270
(1 RCT)

⊕⊕⊝⊝
Lowb

RR 0.46
(0.16 to 1.32)

Study population

77 per 1000

42 fewer per 1000
(65 fewer to 25 more)

Major bleeding
follow‐up: mean 6 months

270
(1 RCT)

⊕⊕⊝⊝
Lowc

RR 1.11
(0.35 to 3.56)

Study population

38 per 1000

4 more per 1000
(25 fewer to 98 more)

Minor bleeding – not reported

—

—

—

—

—

Health‐related quality of life – not reported

—

—

—

—

—

*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)

CI: confidence interval; RCT: randomized controlled trial; RR: risk ratio; VKA: vitamin K antagonist; VTE: venous thromboembolism.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate‐certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low‐certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low‐certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

aDowngraded one level due to serious imprecision, 95% CI was consistent with the possibility for important benefit (62 per 1000 absolute reduction) and possibility of important harm (167 per 1000 absolute increase), included 85 events.

bDowngraded two level due to very serious imprecision, 95% CI was consistent with the possibility of important benefit (65 fewer per 1000) and possibility of important harm (25 more per 1000); included 15 events.

cDowngraded two levels due to very serious imprecision, 95% CI was consistent with the possibility for important benefit (25 per 1000 absolute reduction) and possibility of important harm (98 per 1000 absolute increase), included 11 events.

Background

Please refer to the glossary for the definitions of technical terms (Table 1).

Open in table viewer
Table 1. Glossary

Term

Definition

Adjuvant therapy

A therapy given in addition to the primary treatment to decrease the risk of the cancer recurrence or to assist in the cure.

Anticoagulation

The process of hindering the clotting of blood especially by treatment with an anticoagulant.

Antithrombotic

Used against or tending to prevent thrombosis (clotting)

Coagulation

Clotting

Direct oral anticoagulants (DOAC)

Also known as NOACs are anticoagulant medications that require less monitoring compared to the traditional anticoagulants.

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 (as swelling and pain) and that is potentially life‐threatening if dislodgment of the thrombus results in pulmonary embolism.

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. 2 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 statistic

A measure of degree of nonrandom agreement between observers, measurements of a specific categorical variable, or both.

Metastasis

The spread of a cancer cells from the initial or primary site of disease to another part of the body.

Parenteral nutrition

The practice of feeding a person intravenously, circumventing the gastrointestinal tract.

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.

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. Warfarin is a vitamin K antagonist.

Warfarin

An anticoagulant medication that is a vitamin K antagonist that is used for anticoagulation.

Description of the condition

Cancer is associated with an increased risk of venous thromboembolism (VTE) of four‐ to six‐fold (Heit 2000). Cancer‐related interventions such as chemotherapy, hormonal therapy and indwelling central venous catheters also increase the risk of VTE (Heit 2000). Similarly, people undergoing surgery for cancer have a higher risk of VTE than people undergoing surgery for diseases other than cancer (Gallus 1997; Kakkar 1970). Furthermore, people with cancer and VTE have a higher risk of death than people with cancer alone or with VTE alone (Levitan 1999; Sorensen 2000).

People with cancer also have different benefits and risks from anticoagulant treatment than people without cancer. For instance, during oral anticoagulation therapy for VTE, people with cancer, compared with people without cancer, have a higher incidence of recurrent VTE (27.1 events per 100 participant‐years with cancer versus 9.0 events per 100 participant‐years without cancer; P = 0.003) and of major bleeding (13.3 events per 100 participant‐years with cancer versus 2.2 events per 100 participant‐years without cancer; P = 0.002) (Hutten 2000).

Description of the intervention

Low molecular weight heparins (LMWHs) 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. LMWHs are not absorbed orally and must be administered parenterally by subcutaneous injections (Hirsh 1993).

Direct oral anticoagulant (DOACs) are a new generation of medications with a rapid onset of action that allows a fixed‐dose treatment, and may simplify treatment of VTE by eliminating the need for an initial parenteral anticoagulation (Agnelli 2013).

Vitamin K antagonists (VKAs) have been the mainstay of oral anticoagulant therapy since the 1950s. Well‐designed clinical trials have shown the effectiveness of VKAs for the primary and secondary prevention of several venous and arterial thrombotic diseases (Ansell 2008).

How the intervention might work

Several systematic reviews have compared LMWHs, DOACs and VKAs in the long‐term treatment of VTE, but in populations not representative of people with cancer (Conti 2003; Iorio 2003; van der Heijden 2007). The review by van der Heijden and colleagues did not complete a preplanned subgroup analysis in people with cancer as the required data were not specifically reported (van der Heijden 2007). The review by Conti and colleagues did not conduct a meta‐analysis in the subgroup of people with cancer (Conti 2003). In the review by Iorio and colleagues, one meta‐analysis in the subgroup of people with cancer found no significant difference in mortality (odds ratio 1.13, 95% confidence interval (CI) 0.54 to 2.38).

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. This is fourth versions of this review (previous version published respectively in 2007, 2011, 2016 and cite them). The last major update of this Cochrane systematic review, published in 2018, identified 16 trials. It concluded that the LMWHs compared to VKAs probably produces an important reduction in VTE, and DOACs compared to LMWH may likely reduce VTE but may increase risk of major bleeding (Kahale 2018). Since 2018, we have identified three new eligible trials and one full‐text publication of a previously identified abstract) addressing this question.

Living review approach: since the publication of the 2018 update of the review, we are maintaining 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, 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; people with cancer being treated for VTE have a relatively high rate of VTE recurrence. For instance, during oral anticoagulation therapy for VTE, people with cancer, compared with people without cancer, have a higher incidence of recurrent VTE (27.1 events per 100 participant‐years with cancer versus 9.0 events per 100 participant‐years without cancer; P = 0.003) (Hutten 2000). Second, there remains uncertainty in the existing evidence in relation to the outcomes of mortality and bleeding. Third, we are aware of five ongoing eligible trials that will be important to incorporate in a timely manner. Fourth, this living systematic review may be used as part of a living guideline project  (Akl 2017). 

Objectives

  • To compare the efficacy and safety of anthithrombotics including low molecular weight heparins (LMWHs), direct oral anticoagulants (DOACs),  vitamin K antagonists (VKAs), and Idraparinuxfor the long‐term treatment of venous thromboembolism (VTE) in people with cancer.

  • To maintain this review as a living systematic review by continually running the searches and incorporating newly identified studies.

Methods

Criteria for considering studies for this review

Types of studies

Randomized controlled trials (RCTs).

Types of participants

People with active cancer with a confirmed diagnosis of VTE (deep venous thrombosis (DVT) or pulmonary embolism (PE)). Participants could have been of any age group (including children), with either solid or hematologic cancer, at any cancer stage and irrespective of the type of cancer therapy. VTE should have been diagnosed using an objective diagnostic test.

Cancer should have been diagnosed by the time of inclusion in the study.

We included studies with at least 75% of participants with a cancer status being active (i.e., we excluded studies with more than 25% of participants with non‐active cancer). If between 25% and 75% of the population had active cancer and outcome data for this subgroup of people were not reported, we excluded such studies from the main analysis and included them in the sensitivity analysis.

Types of interventions

Intervention arms consisted of long‐term treatment with antithrombtics including (beyond 3 months of initial treatment) :

  • LMWHs;

  • DOACs;

  • VKAs;

  • Idraparinux

We included any comparison of the three management options listed above (LMWHs versus VKAs, DOACs versus VKAs, DOACs versus LMWHs; idra). Cointerventions, if any, should have been balanced across the groups compared.

Types of outcome measures

Primary outcomes

  • All‐cause mortality.

Secondary outcomes

  • Symptomatic recurrent DVT: DVT events suspected clinically, and confirmed using an objective diagnostic test such as: venography, 125I‐fibrinogen‐uptake test, impedance plethysmography or compression ultrasound.

  • Symptomatic recurrent PE: PE events suspected clinically, and confirmed using an objective diagnostic test such as: pulmonary ventilation/perfusion scans, computed tomography, pulmonary angiography or autopsy.

  • Symptomatic VTE:

  • Major bleeding: we accepted the authors' definitions of major bleeding.

  • Minor bleeding: we accepted the authors' definitions of minor bleeding.

  • Thrombocytopenia: we accepted the authors' definitions of thrombocytopenia.

  • Health‐related quality of life measured using a validated tool.

  • Postphlebetic syndrome.

Search methods for identification of studies

Electronic searches

The search was part of a comprehensive search for studies of anticoagulation in people with cancer. We conducted a comprehensive search on 14 May 2021, following the a major search in February 2016. We electronically searched the following databases: the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (starting 1946, via Ovid), and Embase (starting 1980, via Ovid). We used no language restrictions.The search strategies combined terms for anticoagulants, terms for cancer and a search filter for RCTs. We used no language restrictions. The search strategy was revised by an information specialist (JP). We list the full search strategiy for each of the electronic databases in Appendix 2.

Living systematic review approach: We will be updating the searches using auto‐alerts on a monthly basis. We will publish an update every six months to incorporate any new identified evidence. This update of the systematic review is based on the findings of a literature search conducted on 14 May 2021. 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 May 2021) and of the American Society of Hematology (ASH, starting with its 2003 issue up to May 2021). We also searched ClinicalTrials.gov and World Health Organization (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.. In addition, we contacted experts in the field for information about unpublished work and ongoing trials.

Living review approach: we will search  the conference proceedings of ASCO and ASH soon after their publications and ClinicalTrials.gov and WHO International Clinical Trials Registry Platform. . We will continue to review the reference lists for any prospectively identified studies specify in differences between protocol

Data collection and analysis

Selection of studies

Four pairs of review authors (LAK, FA, MBK, CFM, VEDY, IT, FS, MB, IGT) independently screened the title and abstract of identified articles for potential eligibility. We retrieved the full text of articles judged potentially eligible by at least one review author. Then, the pairs of authors  independently screened the full‐text article for eligibility using a standardized form piloted on 500 RCTs with explicit inclusion and exclusion criteria (as detailed in the Criteria for considering studies for this review section) and resolved any disagreements by discussion or by consulting a third review author.

Living systematic review approach: for the monthly searches, we will immediately screen the 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% (Thomas 2017; Wallace 2017)). 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

Four pairs of review authors (LAK, FA, IT, IGT) independently extracted the data from each study and resolved any disagreements by discussion or by consulting a third review author. We aimed to collect data related to the following.

Participants

  • Number of participants randomized to each study arm.

  • Number of participants followed up in each study arm.

  • Number of participants who discontinued treatment in each arm.

  • Population characteristics (e.g., age, gender, co morbidities).

  • Type of cancer (site,histology).

  • Stage of cancer.

We defined active cancer as (1) non‐squamous cell or basal cell invasive cancer diagnosed within 6 months before enrollment, (2) cancer treated within the previous 6 months, (3) recurrent or metastatic cancer, or (4) reported as active cancer during the study.

We extracted outcome data for people with active cancer. If the study included both people with active cancer and people with non‐active cancer, we sought outcome data for the subgroup of people with active cancer.

Interventions

  • Type of anticoagulant.(LMWH, VKA, or DOAC)

  • Dosage or intensity of anticoagulation

  • Duration of long‐term and initial treatment

  • Cointerventions including chemotherapy, target therapy, immunotherapy, radiation therapy, or a combination of these (type and duration).

Outcomes

We extracted both time‐to‐event data (for the mortality and recurrence of VTE outcomes) and dichotomous data (for all outcomes).

For time‐to‐event 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 the method of Parmar (Parmar 1998). We also noted the minimum and maximum duration of follow‐up, which were 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 collected for each outcome and per arm number of events, number of participants randomized, and number of participants with incomplete data.

For the outcome major bleeding, we extracted events among people with gastrointestinal (GI) tract cancer and people without GI tract cancer. We further extracted major GI bleeding, major upper GI bleeding, major lower GI bleeding, and major non‐GI bleeding.

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, otherwise we included abstracts in the sensitivity analysis

None of the outcomes of interest were continuous variables.

Other

We extracted from each included trial any information on the following:

  • Ethical approval;

  • Source of funding;

  • Conflict of interest.

Assessment of risk of bias in included studies

We assessed risk of bias at the study level using Cochrane's 'Risk of bias' tool (Higgins 2011). Four pairs of review authors ((LAK, FA, IT, IGT) independently assessed the methodologic quality of each included study and resolved any disagreements by discussion. Methodologic criteria included:
 

  • Adequate  sequence generation;

  • Allocation concealment;

  • Blinding of participants and personnel;

  • Blinding of outcome assessment;

  • Incomplete outcome data ;

  • Selective outcome reporting;

  • Other bias (e.g., whether the study was stopped early for benefit).

See the Dealing with missing data section about assessing risk of bias associated with participants with missing data per outcome and across studies.

We attempted to contact the authors for any study domain that was unclear. We re‐evaluated our judgment when authors provided clarification.

Measures of treatment effect

We analyzed hazard ratio (HR) for time‐to‐event data and relative risk (RR) for dichotomous data, with 95% confidence intervals (CI). 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

Identifying participants with missing data

It was not clear whether certain categories of participants (e.g. those described as 'withdrew consent' or 'experienced adverse events') were actually followed up by the trial authors (versus had missing data) (Akl 2016). To identify participants with missing data, we followed the guidance suggested by Kahale et al (Kahale 2019) :

  • Definitely not missing data: (1) participants explicitly reported as followed‐up; (2) participants who died during the trial; (3) participants belonging to centres that were excluded.

  • Definitely missing data: (1) participants explicitly reported as not followed up; (2) participants with unclear follow‐up status and (a) excluded from the denominator of the analysis (i.e., complete case analysis); or (b) included in the denominator of the analysis and their outcomes were explicitly stated to be imputed. However, we did not treat them as missing data unless it is possible to obtain the number of observed/actual events (i.e., excluding imputed events) to avoid double counting.

  • Potentially missing data: Participants with unclear follow‐up status (e.g., included in the denominator of the analysis and their outcomes were not explicitly stated to be imputed).

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; Kahale 2020).

For categorical data, we used the following calculations for each study arm:

  • denominator: (number of participants randomized) – (number of participants definitely with missing data);

  • numerator: number of participants with observed events (i.e. participants who experienced at least one event for the outcome of interest during their available follow‐up time).

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 outcome 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 challenge the statistical significance of the results of the primary analysis progressively (Akl 2013; Kahale 2020).

For categorical data and for an RR showing a reduction in effect (RR < 1), we used the following increasingly stringent but plausible assumptions:

  • For the control arm, relative incidence (RI) among those with missing data (lost to follow‐up (LTFU)) compared with 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)

  • numerator: (number of participants with observed events) + (number of participants definitely with missing data with assumed events).

Assumed events are calculated by applying the a priori plausible assumptions to the participants definitely with missing data.

Assessment of heterogeneity

We assessed heterogeneity between trials by visual inspection of forest plots, estimation of the percentage heterogeneity between trials that could not be ascribed to sampling variation (I2 test; Higgins 2011), and by a formal statistical test of the significance of the heterogeneity (Deeks 2001). If there was evidence of substantial heterogeneity, we investigated and reported the possible reasons for this (see section on Subgroup analysis and investigation of heterogeneity).

Assessment of reporting biases

We planned to create funnel plots for outcomes including 10 or more trials .

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 Review Manager 5.4.1 (Review Manager 2020). For dichotomous data, we calculated the RR separately for each study. When analyzing data related to participants who were reported as non‐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 evidence at the outcome level using the GRADE approach for each of the following comparisons and outcomes (GRADE Handbook):

  • LMWH versus VKA

  • DOAC versus VKA

  • DOAC versus LMWH

  • Idraparinux versus VKA

We included in the primary analysis all studies that provided outcome data for particpants with active cancer, and in the sensitivity analysis all studies that included between 25% and 75% of their study sample participants with active cancer and did not provide outcome data for this subgroup of people.

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

Since anticoagulants might perfom differently in participants with GI tract cancer (CITATION), we planned to do a subgroup analysis for the outcome major bleeding among participants with GI tract cancer and those without GI tract cancer. We included studies that

  • Recruited only patients with GI tract cancer and studies that recruited only participants with non‐GI tract cancer;

  • Recruited both GI and non‐GI tract cancer if they provided outcome data for subgroups of participants with GI tract cancer and data for subgroups of participants with non‐GI tract cancer;

  • Recruited both GI tract and non‐GI tract cancer but did not provide subgroup data only if more than 75% of participants had GI tract cancer or more than 75% of participants had non‐GI tract cancer.

If the proportion of participants with GI tract cancer ranged between 25% and 75% and outcome data for this subgroup was not provided, we did not include such a study in the subgroup analysis.

For this subgroup analysis, we did not conduct complete case analysis for the primary analysis as noted under Dealing with missing data section. Instead we used the denominator reported in the analysis of each study report in order to balance numbers of participants with GI tract cancer and those without GI tract cancer.

 

Sensitivity analysis

 

We decided a priori to consider abstracts and completed studies published exclusively on ClinicalTrials.gov in the main analysis only if study authors supplied us with full reports of their methods and results; otherwise abstracts were included only in the sensitivity analysis.

As described earlier under Dealing with missing data section, we also planned for sensitivity meta‐analyses to assess the risk of bias associated with missing outcome data when the primary meta‐analysis of a specific outcome found a statistically significant effect.

In addition, we planned a sensitivity analysis including studies that included between 25% and 75% of their study sample participants with active cancer and did not provide outcome data for this subgroup of people.

Results

Description of studies

Results of the search

Figure 1 shows the study flow diagram. As of May 2021, the search strategy identified 3583 unique citations. The title and abstract screening identified 102 potentially eligible citations. The full‐text screening of the full texts of these 102  citations identified 19eligible RCTs published as full reports (Agnelli 2015 (AMPLIFY); Agnelli 2020 (Caravaggio); Deitcher 2006 (ONCENOX); El Mokadem 2020; Hull 2006 (LITE); Lee 2003 (CLOT); Lee 2015 (CATCH); Lopez‐Beret 2001; McBane 2019 (ADAM‐VTE); Meyer 2002 (CANTHANOX); Prins 2014 (EINSTEIN); Raskob 2016 (HOKUSAI); Raskob 2018 (HOKUSAI); Romera 2009; Schulman 2015 (RECOVER I‐II); van Doormaal 2010 (Van Gogh DVT trial); Young 2018 (SELECT‐D)), and two studies published as abstracts (Cesarone 2003; Mazilu 2014 (OVIDIUS)). Since the last major update in May 2018, we included the full text of two previously identified ongoing studies (Agnelli 2020 (Caravaggio); McBane 2019 (ADAM‐VTE), included the full text of a previously previously identified abstract (Young 2018 (SELECT‐D), included the full text of a new study (El Mokadem 2020). We identified five registered but unpublished trials: one terminated (Kamphuisen 2010 (Longheva)) and four ongoing (Kamphuisen 2010 (Longheva); Karatas 2015;  Meyer 2016 (CASTA‐DIVA); Ryun Park 2017 (PRIORITY); Schrag 2016 (CANVAS)). .

Figure 1
Study flow diagram.

Study flow diagram.

Included studies

We included 19 RCTs (55 reports) with  participants with cancer for which outcome data were available (see Characteristics of included studies table).

Eight RCTs compared LMWHs to VKAs for the long‐term treatment of VTE (Cesarone 2003; Deitcher 2006 (ONCENOX); Hull 2006 (LITE); Lee 2003 (CLOT); Lee 2015 (CATCH); Lopez‐Beret 2001; Meyer 2002 (CANTHANOX); Romera 2009); only one of these studies used a different initial anticoagulant in the two study arms (LMWH in the LMWH group and UFH in the VKA group) (Hull 2006 (LITE)). Four studies did not explictly specify whether patients had active cancer or did not provide outcome data for subgroups of patients with active cancer (Cesarone 2003; Hull 2006 (LITE); Lopez‐Beret 2001; Romera 2009).

Five RCTs compared DOACs to VKAs (Agnelli 2015 (AMPLIFY); Mazilu 2014 (OVIDIUS); Prins 2014 (EINSTEIN); Raskob 2016 (HOKUSAI); Schulman 2015 (RECOVER I‐II)). One study did not explictly specify whether patients had active cancer (Mazilu 2014 (OVIDIUS)). One RCT compared a once‐weekly subcutaneous injection of idraparinux for three or six months versus standard treatment (tinzaparin, enoxaparin or dose‐adjusted intravenous heparin followed by warfarin or acenocoumarol; van Doormaal 2010 (Van Gogh DVT trial)). 

Five studies compared DOACs to LMWHs (Agnelli 2020 (Caravaggio); El Mokadem 2020; McBane 2019 (ADAM‐VTE); Raskob 2018 (HOKUSAI); Young 2018 (SELECT‐D)). We also identified five ongoing studies comparing DOACs to LMWHs (Kamphuisen 2010 (Longheva); Karatas 2015; Meyer 2016 (CASTA‐DIVA); Ryun Park 2017 (PRIORITY); Schrag 2016 (CANVAS))).

Agnelli and colleagues recruited 169 participants with active cancer and VTE, a subgroup in the AMPLIFY trial, and followed them up for six months (Agnelli 2015 (AMPLIFY)). Participants were randomized to receive apixaban (10 mg twice daily for seven days followed by 5 mg twice daily) or enoxaparin (1 mg kg twice daily for at least five days) followed by dose‐adjusted warfarin (target international normalized ratio (INR) of 2 to 3). Assessed outcomes were mortality, recurrent VTE and major bleeding. It is of note that 25 participants (13 in the apixaban group and 12 in the enoxaparin/warfarin group) without cancer or a history of cancer at baseline were diagnosed with cancer after treatment assignment. 

Agnelli and colleagues recruited and analized 1155 participants with active cancer from 119 centers in 9 European countries, Israel and The U.S.A.(Agnelli 2020 (Caravaggio)). Participants were randomized to receive apixaban 10 mg twice daily for seven days then 5 mg twice daily for a total of six months or dalteparin 200 IU/kg once daily for the first month and than 150 IU/kg daily for a total of six months. Assessed outcomes were all cause mortality, recurrent VTE, recurrent DVT, PE, major and minor bleeding. Gastrointestinal bleeding was also reported. Participants were followed up for six months.

Cesarone and colleagues recruited 199 participants with cancer and DVT (Cesarone 2003). The authors did not report whether the participants had active cancer or not. Participants were randomized to receive enoxaparin 100 Ul/kg twice daily or coumadin (dose adjusted to keep INR close to 3) for three months. Assessed outcomes were mortality and major outcome event in the three‐month period. 

Deitcher and colleagues recruited 102 participants with active cancer with acute symptomatic VTE (Deitcher 2006 (ONCENOX)). Participants were randomized to receive enoxaparin subcutaneous twice daily (1.0 mg/kg) for five days followed by once daily enoxaparin for 175 days or enoxaparin subcutaneous twice daily (1.0 mg/kg) for five days then warfarin starting day two of enoxaparin for 180 days. Assessed outcomes were mortality, recurrent VTE, and major and minor bleeding. .

El Mokadem and colleagues recruited and analyzed 138 participants with active cancer and acute DVT (El Mokadem 2020). Participants were randomized to receive apixaban 10 mg twice daily for seven days then 5 mg twice daily for a total of six months or enoxaparin 1 mg/kg/sc every 12 h for a total of six months. Assessed outcomes were mortality,recurrence of DVT and PE, major bleeding, and minor bleeding. Participants were followed up for six months.

Hull and colleagues recruited 200 participants with cancer with acute symptomatic proximal vein thrombosis (Hull 2006 (LITE)). Participants were randomized to receive tinzaparin 175 anti‐Xa/kg subcutaneously daily for 12 weeks or UFH either 5000 U or 80 U/kg for five days followed by VKAs (target INR 2 to 3) for 12 weeks. Assessed outcomes were mortality, recurrent VTE, major and minor bleeding, and thrombocytopenia. Participants were followed up for one year. 

Lee and colleagues recruited 676 participants with cancer and proximal DVT, PE or both in the CLOT study (Lee 2003 (CLOT)). Participants were randomized to receive dalteparin 200 IU per kilogram once daily for five to seven days and a coumarin derivative for six months (target INR 2.5) or dalteparin alone for six months (200 IU per kilogram once daily for one month, followed by a daily dose of approximately 150 IU per kilogram for five months). Assessed outcomes were mortality, recurrent VTE, and major and minor bleeding. Participants were followed up for six months. 

Lee and colleagues recruited 900 participants with active cancer and objectively documented proximal DVT or PE in the CATCH study (Lee 2015 (CATCH)). Participants were randomized to receive tinzaparin 175 IU/kg once daily for six months or conventional therapy with tinzaparin 175 IU/kg once daily for five to 10 days followed by warfarin at a dose adjusted to maintain the INR within the therapeutic range (2 to 3) for six months. Assessed outcomes were mortality, recurrent VTE, and major and non‐major bleeding. Participants were followed up for six months. 

Lopez‐Beret and colleagues recruited 35 participants with cancer and symptomatic DVT of the lower limb, a subgroup of 158 participants recruited (Lopez‐Beret 2001a). Participants were randomized to receive nadroparin 1.025 anti‐Xa IU/10 kg twice daily for three days then 1.025 anti‐Xa IU/10 kg twice daily, after the third month, nadroparin was switched to once daily, or nadroparin 1.025 anti‐Xa IU/10 kg twice daily for three days then acenocoumarol (target INR 2 to 3) for three to six months. Assessed outcome available for the cancer subgroup was mortality. Participants were followed up for 12 months. 

Mazilu and colleagues recruited 46 participants with paraneoplastic DVT (Mazilu 2014 (OVIDIUS)). Participants were randomized to receive either fixed‐dose dabigatran or adjusted‐dose acenocoumarol. Assessed outcomes were mortality, recurrent VTE and bleeding. 

McBane and colleagues recruited 300 participants with active cancer from 28 centers the U.S.A.(Mc Bane 2019 (ADAM‐VTE)). Participants were randomized to receive apixaban 10 mg twice daily for seven days then 5 mg twice daily for a total of six months or dalteparin 200 IU/kg for the first month and than 150 IU/kg once daily for a total of six months. Assessed outcomes were major bleeding, clinically‐relevant non‐major bleeding, any recurrence of DVT, PE, fatal PE, arterial thromboembolism and mortality.Gastrointestinal bleeding was also reported. Participants were followed up for six months.

Meyer and colleagues recruited 146 participants with cancer and VTE (Meyer 2002 (CANTHANOX)). Participants were randomized to receive enoxaparin 1.5 mg/kg daily for three months or enoxaparin 1.5 mg/kg daily for four days followed by warfarin (target INR 2 to 3) for three months. Outcomes assessed were mortality, recurrent VTE and major bleeding. Participants were followed up for three months. The study noted that 52% of participants had ongoing cancer treatment in the warfarin group versus 76% in the enoxaparin group. 

Prins and colleagues recruited 459 participants with active cancer at baseline and DVT or PE, a subgroup of the EINSTEIN‐DVT and EINSTEIN‐PE studies (Prins 2014 (EINSTEIN)). Participants were randomized to receive rivaroxaban 15 mg twice daily for 21 days, followed by 20 mg once daily. Participants assigned to the enoxaparin and VKA group received enoxaparin subcutaneously 1.0 mg/kg bodyweight twice daily and either oral warfarin or acenocoumarol (target INR 2 to 3), started within 48 hours of randomization. Enoxaparin was discontinued when the INR was 2 or more for two days consecutively and the participant had received at least five days of enoxaparin treatment. The dose of the VKA was adjusted to maintain an INR of 2 to 3. Assessed outcomes were mortality, recurrent VTE, major bleeding and clinically relevant bleeding. Participants were followed up for 12 months. 

Raskob and colleagues recruited 208 participants with active cancer and DVT or PE (Raskob 2016 (HOKUSAI)). Participants were randomized to receive LMWH for at least five days followed by oral edoxaban 60 mg once daily (edoxaban group) or warfarin (or placebo) started concurrently with the study regimen of heparin. Assessed outcomes were recurrent VTE, major and non‐major bleeding, and mortality. Participants were followed up for one year. 

Raskob and colleagues recruited 1050 participants with active cancer and VTE (Raskob 2018 (HOKUSAI)). Participants were randomized to receive LMWH for at least five days followed by oral edoxaban 60 mg once daily (edoxaban group) or subcutaneous dalteparin 200 IU per kilogram bodyweight once daily for one month followed by dalteparin 150 IU per kilogram once daily (dalteparin group). Assessed outcomes were recurrent VTE, major and non‐major bleeding, and mortality. Participants were followed up for one year. 

Romera and colleagues recruited 69 participants with cancer with symptomatic proximal DVT, a subgroup of 241 recruited participants (Romera 2009). All participants were given tinzaparin fixed dose 175 IU anti‐Xa per kg bodyweight once daily. The participants randomized to tinzaparin received this regimen for six months without dosage adjustments. The participants randomized to oral anticoagulants were given acenocoumarol 3 mg orally, which was subsequently adjusted to achieve a regular INR between 2 and 3 for six months. This group received tinzaparin until the INR reached at least 2 on two consecutive measurements. The assessed outcome for the cancer subgroup was recurrent VTE. Participants were followed up for one year. 

Schulman and colleagues recruited 221 participants with active cancer and VTE, a subgroup of the RECOVER and RECOVER‐II trials (Schulman 2015 (RECOVER I‐II)). Participants were randomized to receive warfarin adjusted to achieve an INR of 2 to 3 or dabigatran fixed‐dose 150 mg twice daily. In both randomization arms, initial treatment was with a parenteral anticoagulant (UFH, LMWH or fondaparinux) until the INR or sham INR became at least 2 for two consecutive days. Assessed outcomes were symptomatic recurrent VTE and VTE‐related death, major bleeding and clinically relevant non‐major bleeding. Participants were followed up for six months. The study authors reported complete follow‐up.

Van Doormaal and colleagues recruited 284 participants with active cancer and DVT, a subgroup of the Van Gogh DVT trial (van Doormaal 2010 (Van Gogh DVT trial)). Participants were randomized to receive idraparinux for three or six months or VKA. The study noted that 66% of the idraparinux group and 69% of the VKAs group had active cancer. Assessed outcomes were mortality, recurrent VTE and bleeding. Participants were followed up for six months. 

Young and colleagues recruited 406 participants with active cancer and VTE (Young 2017 (SELECT‐D)). Participants were randomized to receive rivaroxaban 15 mg twice daily for three weeks then 20 mg once daily for a total of six months or dalteparin 200 IU/kg daily for one month and 150 IU/kg daily for a total of six months. Assessed outcomes were recurrent VTE, mortality, and major and clinically non‐major bleeding. Gastrointestinal bleeding was also reported. Participants were followed up for six months.

Excluded studies

We excluded 46reports) from this review for the following reasons: not population of interest (26);not intervention of interest (3):; not design of interest (10) and outcome data for cancer subgroup not reported (7). See Characteristics of excluded studies table.

Risk of bias in included studies

The judgments for the risk of bias are summarized in Figure 2 and Figure 3.

Figure 2
Risk of bias graph: review authors' judgments about each risk of bias item presented as percentages across all included studies.

Risk of bias graph: review authors' judgments about each risk of bias item presented as percentages across all included studies.

Figure 3
Risk of bias summary: review authors' judgments about each risk of bias item for each included study.

Risk of bias summary: review authors' judgments about each risk of bias item for each included study.

Allocation

Random sequence was definitely generated in 14 studies (Agnelli 2015 (AMPLIFY); Agnelli 2020 (Caravaggio); El Mokadem 2020; Hull 2006 (LITE); Lee 2003 (CLOT); Lee 2015 (CATCH); McBane 2019 (ADAM‐VTE); Meyer 2002 (CANTHANOX); Prins 2014 (EINSTEIN); Raskob 2016 (HOKUSAI); Raskob 2018 (HOKUSAI);  Schulman 2015 (RECOVER I‐II); van Doormaal 2010 (Van Gogh DVT trial); Young 2018 (SELECT‐D)).

Five studies were judged to be at low risk of selection bias because minimal information about random sequence generation was provided (Cesarone 2003; Deitcher 2006 (ONCENOX); Lopez‐Beret 2001; Mazilu 2014 (OVIDIUS); Romera 2009).

We judged allocation to be adequately concealed in 12 studies (Agnelli 2015 (AMPLIFY); Agnelli 2020 (Caravaggio) Lee 2003 (CLOT); Lee 2015 (CATCH); McBane 2019 (ADAM‐VTE); Meyer 2002 (CANTHANOX); Prins 2014 (EINSTEIN); Raskob 2016 (HOKUSAI); Raskob 2018 (HOKUSAI); Schulman 2015 (RECOVER I‐II); van Doormaal 2010 (Van Gogh DVT trial)). Eight studies did not report allocation concealment (Cesarone 2003; Deitcher 2006 (ONCENOX); El Mokadem 2020; Hull 2006 (LITE); Lopez‐Beret 2001; Mazilu 2014 (OVIDIUS); Raskob 2018 (HOKUSAI); Romera 2009).

Blinding

Blinding of participants and personnel (performance bias)

We judged participants and personnel to be definitely blinded in four studies (Agnelli 2015 (AMPLIFY); Deitcher 2006 (ONCENOX); Raskob 2016 (HOKUSAI); Schulman 2015 (RECOVER I‐II)),  definitely not blinded in 11 studies (Agnelli 2020 (Caravaggio); Hull 2006 (LITE);  Lee 2003 (CLOT); Lee 2015 (CATCH); Lopez‐Beret 2001; Meyer 2002 (CANTHANOX); Prins 2014 (EINSTEIN); Raskob 2018 (HOKUSAI); Romera 2009; van Doormaal 2010 (Van Gogh DVT trial); Young 2018 (SELECT‐D) , and probably not blinded in three studies (Cesarone 2003; El Mokadem 2020; McBane 2019 (ADAM‐VTE)). One study did not report on blinding of participants and personnel (unclear risk of bias; Mazilu 2014 (OVIDIUS)).

Blinding of outcome assessment (detection bias)

 

We judged outcome assessors to be definitely blinded in 10 studies (Agnelli 2015 (AMPLIFY); Agnelli 2020 (Caravaggio); Hull 2006 (LITE); Lee 2003 (CLOT); Lee 2015 (CATCH); Lopez‐Beret 2001; Meyer 2002 (CANTHANOX); Prins 2014 (EINSTEIN); Raskob 2016 (HOKUSAI); Raskob 2018 (HOKUSAI); Romera 2009; Schulman 2015 (RECOVER I‐II); van Doormaal 2010 (Van Gogh DVT trial)), definitely not blinded in one study (Young 2018 (SELECT‐D)), and probably not blinded in four studies (Cesarone 2003; Deitcher 2006 (ONCENOX); El Mokadem 2020; McBane 2019 (ADAM‐VTE))  One study was not clear about the blinding of outcome assessors (Mazilu 2014 (OVIDIUS)).

Incomplete outcome data

We assessed the risk of bias associated with missing data for each outcome with a significant effect (please see Effects of interventions section). Seven studies did not report follow‐up data for the cancer subgroup, but we assumed complete follow‐up taking into consideration the small sample size (Agnelli 2015 (AMPLIFY); Lopez‐Beret 2001; Prins 2014 (EINSTEIN); Raskob 2016 (HOKUSAI); Romera 2009; van Doormaal 2010 (Van Gogh DVT trial)) ).Six studies were judged to be at low risk of incomplete outcome data because the rates of missing outcome data were lower that the event rate in the studies (Agnelli 2020 (Caravaggio); Deitcher 2006 (ONCENOX); Hull 2006 (LITE); Lee 2003 (CLOT); Meyer 2002 (CANTHANOX); Young 2018 (SELECT‐D)), whereas  five studies were judged to be at high risk of incomplete outcome data because the rate of missing outcome data was higher that the event rate in the study (Cesarone 2003; El Mokadem 2020; Lee 2015 (CATCH); McBane 2019 (ADAM‐VTE); Raskob 2018 (HOKUSAI)). One study reported complete follow‐up based on communication with author( Schulman 2015 (RECOVER I‐II)). The risk of incomplete outcome data was not clear in one study (Mazilu 2014 (OVIDIUS)).

Selective reporting

We did not suspect selective reporting of outcomes for any of the studies except for Cesarone 2003 where results for outcomes of interest were not reported individually, and all results were reported under the term "major outcome," in addition we suspected selective reporting in Lee 2015 (CATCH) where authors failed to report on some of the outcomes mentioned in the study protocol. The cancer subgroup data were missing for a large number of studies. Reporting bias was not clear in one study (Mazilu 2014 (OVIDIUS).

Other potential sources of bias

Another potential source of bias was the screening for asymptomatic VTE in three studies (Lopez‐Beret 2001; Meyer 2002 (CANTHANOX); Romera 2009). The full‐text of the abstract Cesarone 2003 was never published.

Effects of interventions

See: Summary of findings 1 Low molecular weight heparin secondary prophylaxis compared to vitamin K antagonist secondary prophylaxis in people with cancer with venous thromboembolism; Summary of findings 2 Direct oral anticoagulant secondary prophylaxis compared to Vitamin K antagonist secondary prophylaxis in patients with active cancer with venous thromboembolism; Summary of findings 4 Direct oral anticoagulant secondary prophylaxis compared to Low molecular weight heparin secondary prophylaxis in patients with cancer with venous thromboembolism; Summary of findings 5 Idraparinux secondary prophylaxis compared to vitamin K antagonist secondary prophylaxis in people with cancer with venous thromboembolism

Low molecular weight heparin versus vitamin K antagonist

Seven RCTs compared LMWH to VKA for the long‐term treatment of VTE in patients with cancer (Cesarone 2003; Deitcher 2006 (ONCENOX); Hull 2006 (LITE); Lee 2003 (CLOT); Lee 2015 (CATCH); Meyer 2002 (CANTHANOX); Romera 2009). The initial treatment was heparin in both arms for all the seven studies.

We identified one ongoing study comparing LMWH to VKA for the long‐term treatment of cancer participants with VTE that was terminated due to slow inclusion of patients (Kamphuisen 2010 (Longheva)).

All‐cause mortality up to six months

Meta‐analysis of four RCTs including 1712 participants, found that LMWH may result in little to no difference on mortality up to six months when compared to VKA (Relative Risk (RR) 0.99, 95% CI 0.88 to 1.13; Risk Difference (RD) 4 fewer per 1000, 95% CI 45 fewer to 45 more; I2 0%; low certainty evidence; Analysis 1.1) (Deitcher 2006 (ONCENOX); Lee 2003 (CLOT); Lee 2015 (CATCH); Meyer 2002 (CANTHANOX)). The certainty of evidence was rated down to low due to serious risk of bias and serious imprecision (summary of findings Table 1).

The results were consistent in a sensitivity analysis including the study published as an abstract (RR 0.99, 95% CI 0.88 to 1.12) (Cesarone 2003),in a sensitvity analysis including the studies that did not specify whether the cancer was active in the included patients (RR 1.00, 95% CI 0.89 1.13) (Hull 2006 (LITE); Lopez‐Beret 2001), and in a sensitivity analysis including the study that used a different initial anticoagulant in the two study arms (RR 1.00, 95% CI 0.89 to 1.12) (Hull 2006 (LITE)).

We did not create a funnel plot for the outcome of mortality due to the low number of included trials.

All‐cause mortality: time‐to‐event analysis

Two studies including 810 participants reported data allowing their inclusion in the time‐to‐event analysis (Lee 2003 (CLOT); Meyer 2002 (CANTHANOX)). Meta‐analysis indicated that LMWHmay reduce mortality slightly compared to VKA (HR 0.94, 95% CI 0.74 to 1.20; RD 18 fewer per 1000, 95% CI 81 fewer to 56 more; I2 16%; low certainty evidence Analysis 1.2). The certainty of evidence was rated down to low due to serious risk of bias and serious imprecision (summary of findings Table 1). The results were consistent in a sensitivity analysis including data provided by the author for the study that used a different initial anticoagulant in the two study arms (HR 0.96, 95% CI 0.81 to 1.14) (Hull 2006 (LITE)).

Recurrent venous thromboembolism up to six months

None of the studies reported DVT and PE as separate outcomes. Meta‐analysis of four studies including 956 participants found that LMWH probably reduces recurrent VTE up to six months compared to VKA (RR 0.59, 95% CI 0.44 to 0.80; RD 51 fewer per 1000, 95% CI 69 fewer to 25 fewer; I2 0%; moderate certainty evidence; Analysis 1.3) (Deitcher 2006 (ONCENOX); Lee 2003 (CLOT); Lee 2015 (CATCH); Meyer 2002 (CANTHANOX)).The certainty of evidence was rated down to moderate due to serious risk of bias (summary of findings Table 1).

The results were consistent in a sensitvity analysis including the studies that did not specify whether the cancer was active in the included patients (RR 0.59, 95% CI 0.45 to 0.79) (Hull 2006 (LITE); Romera 2009) and in a sensitivity analysis including the study that used a different initial anticoagulant in the two study arms (RR 0.56, 95% CI 0.42 to 0.74) (Hull 2006 (LITE)).

Since the primary meta‐analysis found a statistically significant effect, and in order to assess the risk of bias associated with missing 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 3).

We did not create a funnel plot for the outcome of recurrent VTE due to the low number of included trials.

Recurrent venous thromboembolism: time‐to‐event analysis

Two studies including 810 participants reported data allowing their inclusion in the time‐to‐event meta‐analyses. We used time‐to‐event data reported by two studies (Lee 2003 (CLOT); Meyer 2002 (CANTHANOX)). Meta‐analysis showed that LMWH probably reduces recurrent VTE (HR 0.49, 95% CI 0.31 to 0.78; RD 61 fewer per 1000, 95% CI 84 fewer to 26 fewer; I2 0%; moderate certainty evidence; Analysis 1.4).The certainty of evidence was rated down to moderate due to serious risk of bias (summary of findings Table 1).

The results were consistent in a sensitivity analysis (Hull 2006 (LITE); Romera 2009) including data provided by the author for the study that used a different initial anticoagulant in the two study arms (HR 0.47, 95% CI 0.32 to 0.71) (Hull 2006 (LITE)).

We did not create a funnel plot for the outcome of recurrent VTE due to the low number of included trials.

Major bleeding up to six months

Meta‐analysis of four studies including 1712 participants found than LMWH may result in little to no difference in major bleeding up tp six months compared to VKA (RR 1.09, 95% CI 0.55 to 2.12; RD 4 more per 1000, 95% CI 19 fewer to 48 more; I2 = 46%; low certainty evidence; Analysis 1.5) (Deitcher 2006 (ONCENOX); Lee 2003 (CLOT); Lee 2015 (CATCH); Meyer 2002 (CANTHANOX)). The certainty of evidence was rated down to low due to serious risk of bias and serious imprecision (summary of findings Table 1).

The results were different in a sensitivity analysis including the study published as an abstract (RR 0.85, 95% CI 0.41 to 1.74) (Cesarone 2003) and in a senstivity analysis including the study that did not specify whether the cancer was active in the included patient and used a different initial anticoagulant in the two study arms for the outcome of major bleeding (RR 1.07, 95% CI 0.64 to 1.78) (Hull 2006 (LITE)).

We did not create a funnel plot for the outcome of major bleeding due to the low number of included trials.

Minor bleeding up to six months

Meta‐analysis of four studies including 1712 participants found than LMWH may reduce minor bleeding up tp six months compared to VKA but the evidence is very uncertain (RR 0.78, 95% CI 0.47 to 1.27; RD 38 fewer per 1000, 95% CI 92 fewer to 47 more; I2 = 78%; very low certainty evidence; Analysis 1.6) (Deitcher 2006 (ONCENOX); Lee 2003 (CLOT); Lee 2015 (CATCH); Meyer 2002 (CANTHANOX)). The certainty of evidence was rated down to very low due to serious risk of bias, serious inconsistency, and serious imprecision (summary of findings Table 1).

The results were consistent in a sensitivity analysis including the study that did not specify whether the cancer was active in the included patient and used a different initial anticoagulant in the two study arms (RR 0.84, 95% CI 0.56 to 1.27) (Hull 2006 (LITE)).

We did not create a funnel plot for the outcome of minor bleeding due to the low number of included trials.

Thrombocytopenia up to six months

One study including 138 participants assessed thrombocytopenia (Meyer 2002 (CANTHANOX)). The study found that LMWH may reduce thrombocytopenia up to six months slightly compared to VKA (RR 0.94, 95% CI 0.52 to 1.69; RD 15 fewer per 1000, 95% CI 122 fewer to 175 more; low certainty evidence; Analysis 1.7). The certainty of evidence was rated down to low due to very serious imprecision (summary of findings Table 1).

The results were consistent in a sensitivity analysis including the study that did not specify whether the cancer was active in the included patient and used a different initial anticoagulant in the two study arms (RR 1.10, 95% CI 0.67 to 1.79) (Hull 2006 (LITE)).

We did not create a funnel plot for the outcome of thrombocytopenia due to the low number of included trials.

Health‐related quality of life up to six months

None of the studies reported health‐related quality of life.

Postphlebitic syndrome up to six months

None of the studies reported postphlebitic syndrome.

Direct oral anticoagulants versus vitamin K antagonists

Five RCTs compared DOAC to VKA for the long‐term treatment of VTE in patients with cancer (Agnelli 2015 (AMPLIFY); Mazilu 2014 (OVIDIUS); Prins 2014 (EINSTEIN); Raskob 2016 (HOKUSAI); Schulman 2015 (RECOVER I‐II)). The initial treatment was heparin in both arms for two studies (Raskob 2016 (HOKUSAI); Schulman 2015 (RECOVER I‐II)), DOAC vs LMWH for two other studies (Agnelli 2015 (AMPLIFY); Prins 2014 (EINSTEIN), and DOAC vs VKA in one study (Mazilu 2014 (OVIDIUS).

All‐cause mortality six to 12 months

Meta‐analysis of five RCTs, including 1060 participants found that DOAC may reduce mortality from six to 12 months slightly compared to VKA (RR 0.94, 95% CI 0.72 to 1.23; RD 10 fewer per 1000, 95% CI 48 fewer to 39 more; I2 = 0%; low certainty evidence; Analysis 2.1) (Agnelli 2015 (AMPLIFY); Prins 2014 (EINSTEIN); Raskob 2016 (HOKUSAI); Schulman 2015 (RECOVER I‐II)). The certainty of evidence was rated down to low due to very serious imprecision (summary of findings Table 2).

The results were consistent in a sensitivity analysis including the study published as an abstract and that did not specify whether the cancer was active in the included patients (RR 0.92, 95% CI 0.71 to 1.19) (Mazilu 2014 (OVIDIUS)).

We did not create a funnel plot for the outcome of all‐cause mortality due to the low number of included trials.

Recurrent venous thromboembolism six to 12 months

None of the studies reported DVT and PE as separate outcomes. Meta‐analysis of four studies including 1050 participants found that DOAC may reduce recurrent VTE from six to to 12 months slightly compared to VKA (RR 0.63, 95% CI 0.34 to 1.15; RD 18 fewer per 1000, 95% CI 32 fewer to 7 more; I2 = 0%, low certainty evidence; Analysis 2.2) (Agnelli 2015 (AMPLIFY); Prins 2014 (EINSTEIN); Raskob 2016 (HOKUSAI); Schulman 2015 (RECOVER I‐II)). The certainty of evidence was rated down to low due to very serious imprecision (summary of findings Table 2).

We did not create a funnel plot for the outcome of recurrent VTE due to the low number of included trials.

Major bleeding six to 12 months

Meta‐analysis of four studies including 1055 participants found that DOACs may result in little to no difference in major bleeding compared to VKA (RR 0.77, 95% CI 0.39 to 1.53; RD 8 fewer per 1000, 95% CI 23 fewer to 20 more; low certainty evidence; I2 = 0%; Analysis 2.3) (Agnelli 2015 (AMPLIFY); Prins 2014 (EINSTEIN); Raskob 2016 (HOKUSAI); Schulman 2015 (RECOVER I‐II)).The certainty of evidence was rated down to low due to very serious imprecision (summary of findings Table 2).

We did not create a funnel plot for the outcome of major bleeding due to the low number of included trials.

Minor bleeding six to 12 months

Meta‐analysis of four studies including 1055 participants found that DOACs may reduce minor bleeding from six to to 12 months slightly compared to VKA (RR 0.83, 95% CI 0.57 to 1.23; RD 22 fewer per 1000, 95% CI 55 fewer to 29 more; low‐certainty evidence; Analysis 2.4) (Agnelli 2015 (AMPLIFY); Prins 2014 (EINSTEIN); Raskob 2016 (HOKUSAI); Schulman 2015 (RECOVER I‐II)). The certainty of evidence was rated down to low due to very serious imprecision (summary of findings Table 2).

We did not create a funnel plot for the outcome of minor bleeding due to the low number of included trials.

Thrombocytopenia six to 12 months

None of the studies reported thrombocytopenia.

Health‐related quality of life six to 12 months

Two studies assessed health‐related quality of life; the first used the Anti‐Clot Treatment Scale (ACTS) (Prins 2014 (EINSTEIN)), while the other did not report the tool used for assessment (Mazilu 2014 (OVIDIUS)). Prins and colleagues assessed the outcome for the study population (8485 participants) without reporting on the cancer subgroup (655 participants). They reported that HRQoL was better in the rivaroxaban‐treated participants than in the group treated with enoxaparin and VKAs (no further statistical data reported). The certainty of evidence was rated down to low due to very serious imprecision (summary of findings Table 2).

The study by Mazilu and colleagues, published as an abstract, reported that HRQoL was better in the dabigatran group due to the fact that there was no need for monthly blood tests as in the acenocoumarol group (Mazilu 2014 (OVIDIUS)).

Postphlebitic syndrome

None of the studies reported postphlebitic syndrome.

Direct oral anticoagulants versus low molecular weight heparins

Five RCTs compared DOAC to LMWH for the long‐term treatment of VTE in patients with cancer (Agnelli 2020 (Caravaggio); El Mokadem 2020; McBane 2019 (ADAM‐VTE); Raskob 2018 (HOKUSAI); Young 2018 (SELECT‐D)). The initial treatment was DOAC vs LMWH for four studies (Agnelli 2020 (Caravaggio); El Mokadem 2020; McBane 2019 (ADAM‐VTE); Young 2018 (SELECT‐D)) and heparin in both arms for one study (Raskob 2018 (HOKUSAI)).

We identified four ongoing studies comparing DOACs to LMWHs for the long‐term treatment of cancer participants with VTE;two are completed and not published yet (Meyer 2016 (CASTA‐DIVA); Schrag 2016 (CANVAS)); one is still recruiting (Ryun Park 2017 (PRIORITY)); and one ongoing study that was terminated since recruitment was not as expected (Karatas 2015).

All‐cause mortality up to six months

Meta‐analysis of five RCTs, including 2854 participants found that DOAC may result in little to no difference in mortality up to six months compared to LMWH (RR 0.97, 95% CI 0.84 to 1.14; RD 7 fewer per 1000, 95% CI 42 fewer to 35 more; I2 = 25%; low certainty evidence; Analysis 3.1) (Agnelli 2020 (Caravaggio); El Mokadem 2020; McBane 2019 (ADAM‐VTE); Raskob 2018 (HOKUSAI); Young 2018 (SELECT‐D)). The certainty of evidence was rated down to low due to serious risk of bias and serious imprecision (summary of findings Table 4).

We did not create a funnel plot for the outcome of all cause mortality due to the low number of included trials.

Recurrent venous thromboembolism up to six months

Meta‐analysis of five RCTs, including 2854 participants found that DOAC may reduce recurrent VTE up to six months compared to LMWH (RR 0.63, 95% CI 0.45 to 0.88; RD 32 fewer per 1000, 95% CI 48 fewer to 10 more; I2 = 18%; low certainty evidence; Analysis 3.2) ) (Agnelli 2020 (Caravaggio); El Mokadem 2020; McBane 2019 (ADAM‐VTE); Raskob 2018 (HOKUSAI); Young 2018 (SELECT‐D)). The certainty of evidence was rated down to low due to serious risk of bias and serious imprecision (summary of findings Table 4).

We did not create a funnel plot for the outcome of recurrent VTE due to the low number of included trials.

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 3).

Major bleeding up to six months

Meta‐analysis of five RCTs, including 2854 participants found that DOAC may result in little to no difference in major bleeding up to six months compared to LMWH (RR 1.20, 95% CI 0.83 to 1.73; RD 7 more per 1000, 95% CI 6 fewer to 25 more; I2 = 18%; low certainty evidence; Analysis 3.3) (Agnelli 2020 (Caravaggio); El Mokadem 2020; McBane 2019 (ADAM‐VTE); Raskob 2018 (HOKUSAI); Young 2018 (SELECT‐D)). The certainty of evidence was rated down to low due to serious risk of bias and serious imprecision (summary of findings Table 4). A subgroup analysis of participants with GI tract cancer versus participants without GI tract cancer found no subgroup difference (among people with GI tract cancer RR 1.47, 95% CI 0.76 to 2.84; among people without GI tract cancer RR 1.05, 95% CI 0.63 to 1.73) (Analysis 3.3) ((Agnelli 2020 (Caravaggio); El Mokadem 2020;Raskob 2018 (HOKUSAI); Young 2018 (SELECT‐D)).

Meta‐analysis of four RCTs, including 1838 participants found that DOAC may result in little to no difference in major GI tract bleeding up to six months compared to LMWH (RR 1.16, 95% CI 0.62 to 2.17; RD 3 more per 1000, 95% CI 8 fewer to 23 more; I2 = 0%; low certainty evidence; Analysis 3.4) (Agnelli 2020 (Caravaggio); El Mokadem 2020; McBane 2019 (ADAM‐VTE); Raskob 2018 (HOKUSAI); Young 2018 (SELECT‐D)). The certainty of evidence was rated down to low due to serious risk of bias and serious imprecision (summary of findings Table 4).

Meta‐analysis of four RCTs, including 1838 participants found that DOAC may have little to no difference in major upper GI tract bleeding up to six months compared to LMWH but the evidence is very uncertain (RR 1.18, 95% CI 0.51 to 2.76; RD 2 more per 1000, 95% CI 5 fewer to 19 more; I2 = 0%; very low certainty evidence; Analysis 3.5) (Agnelli 2020 (Caravaggio); El Mokadem 2020; McBane 2019 (ADAM‐VTE); Young 2018 (SELECT‐D)). The certainty of evidence was rated down to very low due to serious risk of bias and very serious imprecision (summary of findings Table 4).

Meta‐analysis of four RCTs, including 1838 participants found that DOAC may have little to no difference in major lower GI tract bleeding up to six months compared to LMWH but the evidence is very uncertain (RR 1.10, 95% CI 0.43 to 2.80; RD 1 more per 1000, 95% CI 5 fewer to 16 more; I2 = 0%; very low certainty evidence;Analysis 3.6) (Agnelli 2020 (Caravaggio); El Mokadem 2020; McBane 2019 (ADAM‐VTE); Young 2018 (SELECT‐D)). The certainty of evidence was rated down to very low due to serious risk of bias and very serious imprecision (summary of findings Table 4).

Meta‐analysis of four RCTs, including 1838 participants found that DOAC may have little to no difference in major non‐GI tract bleeding up to six months compared to LMWH but the evidence is very uncertain (RR 0.84, 95% CI 0.62 to 1.68; RD 3 fewer per 1000, 95% CI 11 fewer to 13 more; I2 = 0%; very low certainty evidence;Analysis 3.7) (Agnelli 2020 (Caravaggio); El Mokadem 2020; McBane 2019 (ADAM‐VTE); Young 2018 (SELECT‐D)). The certainty of evidence was rated down to very low due to serious risk of bias and very serious imprecision (summary of findings Table 4).

We did not create a funnel plot for the outcome of recurrent VTE due to the low number of included trials.

Minor bleeding up to six months

Meta‐analysis of five RCTs, including 2854 participants found that DOAC probably reduces minor bleeding up to six months compared to LMWH (RR 1.58, 95% CI 1.15 to 2.16; RD 39 more per 1000, 95% CI 10 more to 78 more; I2 = 23%; moderate certainty evidence; Analysis 3.8) (Agnelli 2020 (Caravaggio); El Mokadem 2020; McBane 2019 (ADAM‐VTE); Raskob 2018 (HOKUSAI); Young 2018 (SELECT‐D)).The certainty of evidence was rated down to moderate due to serious risk of bias (summary of findings Table 4).

Since the primary meta‐analysis found a statistically significant effect, and in order to assess the risk of bias associated with missing data, we conducted sensitivity meta‐analyses using the a priori plausible assumptions detailed in the Methods section. The effect estimate lost significance with RI 3 and RI 5, indicating high risk of bias associated with missing data (Appendix 3).

Meta‐analysis of two RCTs, including 1495 participants found that DOAC may have little to no difference in minor GI tract bleeding up to six months compared to LMWH but the evidence is very uncertain (RR 1.37, 95% CI 0.41 to 4.64; RD 9 more per 1000, 95% CI 15 fewer to 92 more; I2 = 64%; very low certainty evidence; Analysis 3.9) (Agnelli 2020 (Caravaggio); Young 2018 (SELECT‐D)). The certainty of evidence was rated down to very low due to serious inconsistency and very serious imprecision (summary of findings Table 4).

Meta‐analysis of two RCTs, including 1495 participants found that DOAC may have little to no difference in minor upper GI tract bleeding up to six months compared to LMWH but the evidence is very uncertain (RR 1.03, 95% CI 0.04 to 25.97; RD 0 more per 1000, 95% CI 10 fewer to 267 more; I2 = 74%; very low certainty evidence; Analysis 3.10) (Agnelli 2020 (Caravaggio); Young 2018 (SELECT‐D)). The certainty of evidence was rated down to very low due to very serious inconsistency and very serious imprecision (summary of findings Table 4).

Meta‐analysis of two RCTs, including 1495 participants found that DOAC may have little to no difference in minor lower GI tract bleeding up to six months compared to LMWH (RR 1.54, 95% CI 0.72 to 3.27; RD 8 more per 1000, 95% CI 4 fewer to 33 more; I2 = 0%; low certainty evidence; Analysis 3.11) (Agnelli 2020 (Caravaggio); Young 2018 (SELECT‐D)). The certainty of evidence was rated down to low due to very serious imprecision (summary of findings Table 4).

Meta‐analysis of two RCTs, including 1495 participants found that DOAC probably increases minor non‐ GI tract bleeding up to six months compared to LMWH (RR 2.37, 95% CI 1.44 to 3.89; RD 42 more per 1000, 95% CI 18 fewer to 89 more; I2 = 3%; moderate certainty evidence; Analysis 3.12) (Agnelli 2020 (Caravaggio); Young 2018 (SELECT‐D)). The certainty of evidence was rated down to moderate due to serious imprecision (summary of findings Table 4).

Thrombocytopenia up to six months

None of the studies reported thrombocytopenia.

Health‐related quality of life up to six months

None of the studies reported HRQoL.

Postphlebitic syndrome up to six months

None of the studies reported postphlebitic syndrome.

Once‐weekly idraparinux versus vitamin K antagonists

One RCT with 284 participants compared once‐weekly subcutaneous injection of idraparinux versus standard treatment (parenteral anticoagulation followed by warfarin or acenocoumarol) for three or six months (van Doormaal 2010 (Van Gogh DVT trial)).

All‐cause mortality up to six months

The trial found thatidraparinux may reduce all‐cause mortality up to six months slightly compared to VKAs on mortality at six months (RR 1.11, 95% CI 0.78 to 1.59; RD 31 more per 1000, 95% CI 62 fewer to 167 more; low‐certainty evidence).

Recurrent venous thromboembolism up to six months

The trial found that idraparinux may reduce recurrent VTE up to six months slightly compared to VKA (RR 0.46, 95% CI 0.16 to 1.32; RD 42 fewer per 1000, 95% CI 65 fewer to 25 more; low‐certainty evidence).

Major bleeding up to six months

The trial found that idraparinux may result in little to no difference in major bleeding up to six months slightly compared to VKA (RR 1.11, 95% CI 0.35 to 3.56; RD 4 more per 1000, 95% CI 25 fewer to 98 more; low‐certainty evidence).

Minor bleeding up to six months

The study did not report minor bleeding.

Thrombocytopenia up to six months

The study did not report thrombocytopenia.

Health‐related quality of life up to six months

The study did not report HRQoL.

Postphlebitic syndrome

The study did not report postphlebitic syndrome.

We did not create funnel plots due to the low number of included trials for each outcome.

Discussion

Summary of main results

For the long‐term treatment of VTE in people with cancer, LMWH compared to VKAmay result in little to no difference on mortality and major bleeding up to six months, probably reduces recurrent VTE up to six months, may reduce minor bleeding up tp six months but the evidence is very uncertain, and may reduce thrombocytopenia up to six months slightly.

Direct oral anticoagulants compared to VKA may reduce mortality, recurrent VTE from six to 12 months slightly and may result in little to no difference in major bleeding compared to VKA.

Direct oral anticoagulants compared to LMWH may result in little to no difference in mortality, major bleeding, major GI tract bleeding, up to six months, may reduce recurrent VTE up to six months, may have little to no difference in major upper GI tract bleeding, major lower GI tract bleeding, major non‐GI tract bleeding, minor GI tract bleeding, minor upper GI tract bleeding, but the evidence is very uncertain, probably reduces minor bleeding up to six months, may have little to no difference in minor lower GI tract bleeding, and probably increases minor non‐ GI tract bleeding up to six months

Idraparinux compared to VKA may reduce all‐cause mortality, recurrent VTE, up to six months slightly, and may result in little to no difference in major bleeding up to six months slightly.

 

Overall completeness and applicability of evidence

While the reduction in VTE with LMWHs was expected to reduce thrombosis‐related mortality, this did not translate into an observed reduction in all‐cause mortality. This finding was not apparently explained by an increase in any specific‐cause mortality (e.g. fatal bleeding), but might have been due to the lack of power to detect a reduction in all‐cause mortality. Similarly, the size of the available evidence was not large enough to rule out beneficial or harmful effects for many comparisons (e.g. effects of LMWHs versus VKAs on bleeding, effects of DOACs vs VKA on bleeding, effects of DOACs vs LMWH on mortality, bleeding).

Stratifying bleeding outcomes based on type of cancer (GI tract vs non‐GI tract) did not find any subgroup effect when comparing DOAC to LWMH. When adding patients with non‐active cancer (or unclear whether active cancer) to patients with active cancer, the results of the meta‐analyses for all studied outcomes remained unchanged. were unable to conduct subgroup analyses based on histologic type (GI tract) or stage of cancer (active) because of the lack of data DELETE ONCE CONFIRMED BY EA. In the absence of evidence for the contrary, we assumed that the results of this review applied to people with any type or stage of cancer.

For comparisons including DOACs, four RCTs used Apixaban (Agnelli 2015 (AMPLIFY); Agnelli 2020 (Caravaggio); El Mokadem 2020; McBane 2019 (ADAM‐VTE)); three RCTs used Rivaroxaban (Prins 2014 (EINSTEIN); Raskob 2016 (HOKUSAI); Raskob 2018 (HOKUSAI); two RCTs used Edoxaban (Raskob 2016 (HOKUSAI); Raskob 2018 (HOKUSAI)) raskob 2016 2018 ; dabigatran mazilu schulman we were unable to conduct subgroup analyses based on different agents and hence we cannot make any judgment in related to class effect. The majority of studies used Apixaban CHARBEL, SOME LITERATURE ON CLASS EFFECT

majority of stufies used apixaban ‐ unclear evidence regarding edoaxaban or other classess of DPAC

Quality of the evidence

Our systematic approach to searching, study selection and data extraction should have minimized the likelihood of missing relevant studies.

When comparing LMWHs to VKAs, we judged the certainty of evidence to be moderate for recurrent VTE due to serious risk of bias, and moderate for mortality at one year, and major bleeding due to both imprecision and risk of bias and low for minor bleeding due to imprecision, risk of bias and inconsistency..

We downgraded recurrent VTE by one level due to serious risk of bias, allocation concealment unclear in two studies, high risk of selective reporting and high risk of incomplete outcome data in one study, and lack of blinding of participants and personnel in all included studies. We downgraded the outcomes of mortality and major bleeding by one level due to both risk of bias and imprecision, in addition we downgraded the outcome of minor bleed by two levels, one for inconsistency and one for risk of bias and imprecision combined. The lack of allocation concealment in two of the studies did not affect the results when conducting a sensitivity analysis after removing those studies that had a combined weight of 6.5%, but we were concerned about the lack of blinding of participants and personnel in all included studies in addition to high risk of bias in the CATCH trial that represented 43.1% of the weight, so we decided to downgrade by one level due to both concerns about imprecision and risk of bias.

When comparing DOACs to VKAs, we judged the certainty of evidence to be moderate for HRQoL due to serious indirectness, low for mortality, recurrent VTE, and major and minor bleeding due to serious imprecision and serious indirectness.

When comparing DOACs to LMWHs, we judged the certainty of evidence to be low for mortality, VTE recurrence, and major and minor bleeding due to serious risk of bias and serious imprecision.

When comparing idraparinux to VKAs, we judged the certainty of evidence to be moderate for mortality due to serious imprecision and low for recurrent VTE and major bleeding due to very serious imprecision, Taking into consideration the wide CIs, the low number of events and the fact that only one study is providing data for this comparison.pote

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 was also limited by not including data from the trials published as abstracts only. A potential bias of our review might be the limitation of the electronic search strategy to participants with cancer, while the data needed for this review came from studies not restricted to this subgroup. 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. A major limitation of this review was that we were unable to include in the meta‐analyses 11 eligible RCTs with subgroups of participants with cancer because relevant data were not reported and not obtainable from the authors.

Agreements and disagreements with other studies or reviews

We identified seven published systematic reviews comparing LMWHs or DOACs to VKAs in the long‐term treatment of VTE (Conti 2003; Iorio 2003; Laporte 2011; Noble 2008; Posch 2015; Romera‐Villegas 2010; Vedovati 2015). We review below the findings of the two most recent reviews.

Posch and colleagues compared LMWHs or DOACs to VKAs for the long‐term treatment of VTE in participants with cancer including six RCTs comparing LMWHs to VKAs and four RCTs comparing DOACs to VKAs (Posch 2015). The meta‐analysis found a significant reduction of recurrent VTE in favor of LMWHs (RR 0.6, 95% CI 0.45 to 0.79) and a non‐significant difference in major bleeding episodes (RR 1.07, 95% CI 0.66 to 1.73; p = 0.80). There was no significant difference in recurrent VTE and major bleeding when comparing DOACs to VKAs (recurrent VTE: RR 0.65, 95% CI 0.38 to 1.09; major bleeding: RR 0.72, 95% CI 0.39 to 1.35). These results were in agreement with our study.

Vedovati and colleagues compared DOACs to VKAs in the long‐term treatment of VTE in participants with cancer (Vedovati 2015). Meta‐analysis of five RCTs showed no significant difference in VTE recurrence when comparing DOACs to VKAs (RR 0.63, 95% CI 0.37 to 1.1). These results were in agreement with our study.

Study flow diagram.

Figures and Tables -
Figure 1

Study flow diagram.

Navigate to figure in ReviewOpen in new tab

Risk of bias graph: review authors' judgments about each risk of bias item presented as percentages across all included studies.

Figures and Tables -
Figure 2

Risk of bias graph: review authors' judgments about each risk of bias item presented as percentages across all included studies.

Navigate to figure in ReviewOpen in new tab

Risk of bias summary: review authors' judgments about each risk of bias item for each included study.

Figures and Tables -
Figure 3

Risk of bias summary: review authors' judgments about each risk of bias item for each included study.

Navigate to figure in ReviewOpen in new tab

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 1: All‐cause mortality (up to 6 months) (main analysis ‐ active cancer)

Figures and Tables -
Analysis 1.1

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 1: All‐cause mortality (up to 6 months) (main analysis ‐ active cancer)

Navigate to figure in ReviewOpen in new tab

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 2: All‐cause mortality (time‐to‐event)

Figures and Tables -
Analysis 1.2

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 2: All‐cause mortality (time‐to‐event)

Navigate to figure in ReviewOpen in new tab

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 3: Recurrent venous thromboembolism (up to 6 months) (main analysis ‐ active cancer)

Figures and Tables -
Analysis 1.3

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 3: Recurrent venous thromboembolism (up to 6 months) (main analysis ‐ active cancer)

Navigate to figure in ReviewOpen in new tab

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 4: Recurrent venous thromboembolism (time‐to‐event)

Figures and Tables -
Analysis 1.4

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 4: Recurrent venous thromboembolism (time‐to‐event)

Navigate to figure in ReviewOpen in new tab

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 5: Major bleeding (up to 6 months) (main analysis ‐ active cancer)

Figures and Tables -
Analysis 1.5

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 5: Major bleeding (up to 6 months) (main analysis ‐ active cancer)

Navigate to figure in ReviewOpen in new tab

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 6: Minor bleeding (up to 6 months) (main analysis ‐ active cancer)

Figures and Tables -
Analysis 1.6

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 6: Minor bleeding (up to 6 months) (main analysis ‐ active cancer)

Navigate to figure in ReviewOpen in new tab

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 7: Thrombocytopenia (up to 6 months) (main analysis‐ active cancer)

Figures and Tables -
Analysis 1.7

Comparison 1: Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA), Outcome 7: Thrombocytopenia (up to 6 months) (main analysis‐ active cancer)

Navigate to figure in ReviewOpen in new tab

Comparison 2: Direct oral anticoagulants (DOAC) versus vitamin K antagonists (VKA), Outcome 1: All‐cause mortality (6‐12 months)

Figures and Tables -
Analysis 2.1

Comparison 2: Direct oral anticoagulants (DOAC) versus vitamin K antagonists (VKA), Outcome 1: All‐cause mortality (6‐12 months)

Navigate to figure in ReviewOpen in new tab

Comparison 2: Direct oral anticoagulants (DOAC) versus vitamin K antagonists (VKA), Outcome 2: Recurrent venous thromboembolism (6‐12 months)

Figures and Tables -
Analysis 2.2

Comparison 2: Direct oral anticoagulants (DOAC) versus vitamin K antagonists (VKA), Outcome 2: Recurrent venous thromboembolism (6‐12 months)

Navigate to figure in ReviewOpen in new tab

Comparison 2: Direct oral anticoagulants (DOAC) versus vitamin K antagonists (VKA), Outcome 3: Major bleeding (6‐12 months)

Figures and Tables -
Analysis 2.3

Comparison 2: Direct oral anticoagulants (DOAC) versus vitamin K antagonists (VKA), Outcome 3: Major bleeding (6‐12 months)

Navigate to figure in ReviewOpen in new tab

Comparison 2: Direct oral anticoagulants (DOAC) versus vitamin K antagonists (VKA), Outcome 4: Minor bleeding (6‐12 months)

Figures and Tables -
Analysis 2.4

Comparison 2: Direct oral anticoagulants (DOAC) versus vitamin K antagonists (VKA), Outcome 4: Minor bleeding (6‐12 months)

Navigate to figure in ReviewOpen in new tab

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 1: All‐cause mortality (6 months)

Figures and Tables -
Analysis 3.1

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 1: All‐cause mortality (6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 2: Recurrent VTE (6 months)

Figures and Tables -
Analysis 3.2

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 2: Recurrent VTE (6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 3: Major bleeding (6 months)

Figures and Tables -
Analysis 3.3

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 3: Major bleeding (6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 4: Major GI bleeding (6 months)

Figures and Tables -
Analysis 3.4

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 4: Major GI bleeding (6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 5: Major upper GI bleeding (6 months)

Figures and Tables -
Analysis 3.5

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 5: Major upper GI bleeding (6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 6: Major lower GI bleeding (6 months)

Figures and Tables -
Analysis 3.6

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 6: Major lower GI bleeding (6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 7: Major non‐GI bleeding (6 months)

Figures and Tables -
Analysis 3.7

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 7: Major non‐GI bleeding (6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 8: Minor bleeding (6 months)

Figures and Tables -
Analysis 3.8

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 8: Minor bleeding (6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 9: Minor GI bleeding (6 months)

Figures and Tables -
Analysis 3.9

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 9: Minor GI bleeding (6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 10: Minor upper GI bleeding (6 months)

Figures and Tables -
Analysis 3.10

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 10: Minor upper GI bleeding (6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 11: Minor lower GI bleeding (6 months)

Figures and Tables -
Analysis 3.11

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 11: Minor lower GI bleeding (6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 12: Minor non‐GI bleeding (6 months)

Figures and Tables -
Analysis 3.12

Comparison 3: Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH), Outcome 12: Minor non‐GI bleeding (6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 4: Idraparinux versus vitamin K antagonists (VKA), Outcome 1: All‐cause mortality (up to 6 months)

Figures and Tables -
Analysis 4.1

Comparison 4: Idraparinux versus vitamin K antagonists (VKA), Outcome 1: All‐cause mortality (up to 6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 4: Idraparinux versus vitamin K antagonists (VKA), Outcome 2: Recurrent VTE (up to 6 months)

Figures and Tables -
Analysis 4.2

Comparison 4: Idraparinux versus vitamin K antagonists (VKA), Outcome 2: Recurrent VTE (up to 6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 4: Idraparinux versus vitamin K antagonists (VKA), Outcome 3: Major bleeding (up to 6 months)

Figures and Tables -
Analysis 4.3

Comparison 4: Idraparinux versus vitamin K antagonists (VKA), Outcome 3: Major bleeding (up to 6 months)

Navigate to figure in ReviewOpen in new tab

Comparison 4: Idraparinux versus vitamin K antagonists (VKA), Outcome 4: Minor bleeding (up to 6 months)

Figures and Tables -
Analysis 4.4

Comparison 4: Idraparinux versus vitamin K antagonists (VKA), Outcome 4: Minor bleeding (up to 6 months)

Navigate to figure in ReviewOpen in new tab
Summary of findings 1. Low molecular weight heparin secondary prophylaxis compared to vitamin K antagonist secondary prophylaxis in people with cancer with venous thromboembolism

Low molecular weight heparin secondary prophylaxis compared to vitamin K antagonist secondary prophylaxis in patients with cancer with venous thromboembolism

Population: People with cancer with venous thromboembolism
Setting: Outpatient
Intervention: LMWH secondary prophylaxis
Comparison: VKA secondary prophylaxis

Outcomes

№ of participants
(studies)
Follow up

Certainty of the evidence
(GRADE)

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Risk with VKA secondary prophylaxis

Risk difference with LMWH secondary prophylaxis

All‐cause mortality (main analysis ‐ active cancer)
follow up: 6 months

1712
(4 RCTs)

⊕⊕⊝⊝
LOW 1 2

RR 0.99
(0.88 to 1.12)

Study population

374 per 1,000

4 fewer per 1,000
(45 fewer to 45 more)

All‐cause mortality (time‐to‐event)

1243
(2 RCTs)

⊕⊕⊝⊝
LOW 2 3

HR 0.94
(0.74 to 1.20)

Study population

374 per 1,000

18 fewer per 1,000
(81 fewer to 56 more)

Recurrent venous thromboembolism (main analysis ‐ active cancer)
follow up: 6 months

1712
(4 RCTs)

⊕⊕⊕⊝
MODERATE 1

RR 0.59
(0.44 to 0.80)

Study population

124 per 1,000

51 fewer per 1,000
(69 fewer to 25 fewer)

Recurrent venous thromboembolism (time‐to‐event)

1243
(2 RCTs)

⊕⊕⊕⊝
MODERATE 3

HR 0.49
(0.31 to 0.78)

Study population

124 per 1,000

61 fewer per 1,000
(84 fewer to 26 fewer)

Major bleeding (main analysis ‐ active cancer)
follow up: 6 months

1712
(4 RCTs)

⊕⊕⊝⊝
LOW 1 2 4

RR 1.09
(0.55 to 2.12)

Study population

43 per 1,000

4 more per 1,000
(19 fewer to 48 more)

Minor bleeding (main analysis ‐ active cancer)
follow up: 6 months

1712
(4 RCTs)

⊕⊝⊝⊝
VERY LOW 1 2 5

RR 0.78
(0.47 to 1.27)

Study population

174 per 1,000

38 fewer per 1,000
(92 fewer to 47 more)

Thrombocytopenia (main analysis‐ active cancer)
follow up: 6 months

138
(1 RCT)

⊕⊕⊝⊝
LOW 3 6

RR 0.94
(0.52 to 1.69)

Study population

254 per 1,000

15 fewer per 1,000
(122 fewer to 175 more)

*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).

CI: Confidence interval; RR: Risk ratio; OR: Odds ratio;

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1 Downgraded by one level due to serious risk of bias (allocation concealment unclear in one study, lack of blinding of participants and personnel in all the four studies, high risk of incomplete outcome data in one study, and high risk of selective reporting in one study).

2 Downgraded by one level due to serious imprecision. Confidence interval includes suggests both potential harm and potential benefit.

3 Some concern with lack of blinding of patients and personnel.

4 Some concern with inconsistency. I2= 46%

5 Downgraded by one level due to serious inconsistency (I2= 78%)

6 Downgraded by two levels due to very serious imprecision. Confidence interval includes suggests both potential harm and potential benefit. Low number of events.

Figures and Tables -
Summary of findings 1. Low molecular weight heparin secondary prophylaxis compared to vitamin K antagonist secondary prophylaxis in people with cancer with venous thromboembolism
Navigate to table in Review
Summary of findings 2. Direct oral anticoagulant secondary prophylaxis compared to Vitamin K antagonist secondary prophylaxis in patients with active cancer with venous thromboembolism

Direct oral anticoagulant secondary prophylaxis compared to Vitamin K antagonist secondary prophylaxis in patients with active cancer with venous thromboembolism

Population: patients with active cancer with venous thromboembolism
Setting: Outpatient
Intervention: DOAC secondary prophylaxis
Comparison: VKA secondary prophylaxis

Outcomes

№ of participants
(studies)
Follow up

Certainty of the evidence
(GRADE)

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Risk with Vitamin K antagonist (VKA) secondary prophylaxis

Risk difference with Direct oral anticoagulant (DOAC) secondary prophylaxis

All‐cause mortality
follow up: range 6 months to 12 months

1060
(4 RCTs)

⊕⊕⊝⊝
LOW 1 2

RR 0.94
(0.72 to 1.23)

Study population

171 per 1,000

10 fewer per 1,000
(48 fewer to 39 more)

Recurrent venous thromboembolism
follow up: range 6 months to 12 months

1050
(4 RCTs)

⊕⊕⊝⊝
LOW 1 3

RR 0.63
(0.34 to 1.15)

Study population

49 per 1,000

18 fewer per 1,000
(32 fewer to 7 more)

Major bleeding
follow up: range 6 months to 12 months

1055
(4 RCTs)

⊕⊕⊝⊝
LOW 1 2

RR 0.77
(0.39 to 1.53)

Study population

37 per 1,000

8 fewer per 1,000
(23 fewer to 20 more)

Minor bleeding
follow up: range 6 months to 12 months

1055
(4 RCTs)

⊕⊕⊝⊝
LOW 1 2

RR 0.83
(0.57 to 1.23)

Study population

127 per 1,000

22 fewer per 1,000
(55 fewer to 29 more)

Thrombocytopenia ‐ not reported

‐

‐

‐

‐

‐

Health related quality of life
follow up: range 3 months to 12 months

8485
(1 RCT)

⊕⊕⊕⊝
MODERATE 4

‐

Prins 2014 (EINSTEIN DVT‐PE; n=8485 ): "in the general population of the EINSTEIN studies, patient‐reported satisfaction and quality of life was better in the rivaroxaban‐treated patients than in the group treated with enoxaparin and vitamin K antagonist, although we have not yet examined whether this is the same in patients with active cancer. Hence, it can be expected that quality of life will also be improved with rivaroxaban compared with long‐term injected low molecular‐weight heparin." The tool used was validated measure of treatment satisfaction – the Anti‐Clot Treatment Scale (ACTS))

*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).

CI: Confidence interval; RR: Risk ratio; OR: Odds ratio;

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1 Some concern with indirectness (study by Schulman et al (RECOVER I‐II) included patients with a diagnosis of cancer within five years before enrollment), however the weight of these studies was low and heterogeneity was very low.

2 Downgraded by two levels due to very serious imprecision. Confidence interval suggests both potential benefit and potential harm.

3 Downgraded by two levels due to very serious imprecision. Confidence interval suggests both potential benefit and potential no effect. Low number of events.

4 Downgraded by one level for serious indirectness. The study by Prins and colleagues (Prins 2014 ( EINSTEIN n=8485)) reports health related quality of life for the whole study population, without providing data for the cancer subgroup

Figures and Tables -
Summary of findings 2. Direct oral anticoagulant secondary prophylaxis compared to Vitamin K antagonist secondary prophylaxis in patients with active cancer with venous thromboembolism
Navigate to table in Review
Summary of findings 4. Direct oral anticoagulant secondary prophylaxis compared to Low molecular weight heparin secondary prophylaxis in patients with cancer with venous thromboembolism

Direct oral anticoagulant secondary prophylaxis compared to Low molecular weight heparin secondary prophylaxis in patients with cancer with venous thromboembolism

Patient or population: patients with cancer with venous thromboembolism
Setting: Outpatient
Intervention: DOAC secondary prophylaxis
Comparison: LMWH secondary prophylaxis

Outcomes

№ of participants
(studies)
Follow up

Certainty of the evidence
(GRADE)

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Risk with Low molecular weight heparin (LMWH) secondary prophylaxis

Risk difference with Direct oral anticoagulant (DOAC) secondary prophylaxis

All‐cause mortality
follow up: mean 6 months

2854
(5 RCTs)

⊕⊕⊝⊝
LOW 1 2

RR 0.97
(0.83 to 1.14)

Study population

248 per 1,000

7 fewer per 1,000
(42 fewer to 35 more)

Recurrent VTE
follow up: mean 6 months

2854
(5 RCTs)

⊕⊕⊝⊝
LOW 1 3

RR 0.63
(0.45 to 0.88)

Study population

87 per 1,000

32 fewer per 1,000
(48 fewer to 10 fewer)

Major bleeding
follow up: mean 6 months

2994
(5 RCTs)

⊕⊕⊝⊝
LOW 1 4

RR 1.20
(0.83 to 1.73)

Study population

35 per 1,000

7 more per 1,000
(6 fewer to 25 more)

Major GI bleeding
follow up: mean 6 months

1838
(4 RCTs)

⊕⊕⊝⊝
LOW 1 4

RR 1.16
(0.62 to 2.17)

Study population

20 per 1,000

3 more per 1,000
(8 fewer to 23 more)

Major upper GI bleeding
follow up: mean 6 months

1838
(4 RCTs)

⊕⊝⊝⊝
VERY LOW 1 5

RR 1.18
(0.51 to 2.76)

Study population

11 per 1,000

2 more per 1,000
(5 fewer to 19 more)

Major lower GI bleeding
follow up: mean 6 months

1838
(4 RCTs)

⊕⊝⊝⊝
VERY LOW 1 5

RR 1.10
(0.43 to 2.80)

Study population

9 per 1,000

1 more per 1,000
(5 fewer to 16 more)

Major non‐GI bleeding
follow up: mean 6 months

1838
(4 RCTs)

⊕⊝⊝⊝
VERY LOW 1 6

RR 0.84
(0.42 to 1.68)

Study population

19 per 1,000

3 fewer per 1,000
(11 fewer to 13 more)

Minor bleeding
follow up: mean 6 months

2854
(5 RCTs)

⊕⊕⊕⊝
MODERATE 1

RR 1.58
(1.15 to 2.16)

Study population

67 per 1,000

39 more per 1,000
(10 more to 78 more)

Minor GI bleeding
follow up: mean 6 months

1495
(2 RCTs)

⊕⊝⊝⊝
VERY LOW 6 7 8

RR 1.37
(0.41 to 4.64)

Study population

25 per 1,000

9 more per 1,000
(15 fewer to 92 more)

Minor upper GI bleeding
follow up: mean 6 months

1495
(2 RCTs)

⊕⊝⊝⊝
VERY LOW 6 7 9

RR 1.03
(0.04 to 25.97)

Study population

11 per 1,000

0 fewer per 1,000
(10 fewer to 267 more)

Minor lower GI bleeding
follow up: mean 6 months

1495
(2 RCTs)

⊕⊕⊝⊝
LOW 5 7

RR 1.54
(0.72 to 3.27)

Study population

15 per 1,000

8 more per 1,000
(4 fewer to 33 more)

Minor non‐GI bleeding
follow up: mean 6 months

1495
(2 RCTs)

⊕⊕⊕⊝
MODERATE 7 10

RR 2.37
(1.44 to 3.89)

Study population

31 per 1,000

42 more per 1,000
(14 more to 89 more)

*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).

CI: Confidence interval; RR: Risk ratio; OR: Odds ratio;

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1 Downgraded by one level due to serious risk of bias. Allocation concealment was not reported in one study, lack of blinding of patients and personnel in all studies, and high risk of bias related to incomplete outcome data.

2 Downgraded by one level due to serious imprecision. Confidence interval suggests both potential benefit and potential harm.

3 Downgraded by one level due to serious imprecision. Confidence interval suggests both potential benefit and potential no effect.

4 Downgraded by one level due to serious imprecision. Confidence interval suggests both potential harm and potential no effect.

5 Downgraded by two levels due to very serious imprecision. Confidence interval suggests both potential harm and potential no effect. Low number of events.

6 Downgraded by two levels due to very serious imprecision. Confidence interval suggests both potential harm and potential benefit. Low number of events.

7 Some concern with risk of bias. Lack of blinding of patients and personnel in both studies.

8 Downgraded by one level due to serious inconsistency (unexplained heterogeneity I2=64%) and due to some concern with risk of bias.

9 Downgraded by two levels due to very serious inconsistency (unexplained heterogeneity I2=74%.) and due to some concern with risk of bias.

10 Downgraded by one level due to serious imprecision. Low number of events.

Figures and Tables -
Summary of findings 4. Direct oral anticoagulant secondary prophylaxis compared to Low molecular weight heparin secondary prophylaxis in patients with cancer with venous thromboembolism
Navigate to table in Review
Summary of findings 5. Idraparinux secondary prophylaxis compared to vitamin K antagonist secondary prophylaxis in people with cancer with venous thromboembolism

Idraparinux secondary prophylaxis compared to VKA secondary prophylaxis in people with cancer with VTE

Population: people with cancer with VTE receiving secondary prophylaxis

Setting: outpatient

Intervention: idraparinux prophylaxis

Control: VKA prophylaxis

Outcomes

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Relative effect
(95% CI)

Anticipated absolute effects* (95% CI)

Risk with VKA secondary prophylaxis

Risk difference with idraparinux secondary prophylaxis

All‐cause mortality
follow‐up: mean 6 months

284
(1 RCT)

⊕⊕⊕⊝
Moderatea

RR 1.11
(0.78 to 1.59)

Study population

283 per 1000

31 more per 1000
(62 fewer to 167 more)

Recurrent VTE
follow‐up: mean 6 months

270
(1 RCT)

⊕⊕⊝⊝
Lowb

RR 0.46
(0.16 to 1.32)

Study population

77 per 1000

42 fewer per 1000
(65 fewer to 25 more)

Major bleeding
follow‐up: mean 6 months

270
(1 RCT)

⊕⊕⊝⊝
Lowc

RR 1.11
(0.35 to 3.56)

Study population

38 per 1000

4 more per 1000
(25 fewer to 98 more)

Minor bleeding – not reported

—

—

—

—

—

Health‐related quality of life – not reported

—

—

—

—

—

*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)

CI: confidence interval; RCT: randomized controlled trial; RR: risk ratio; VKA: vitamin K antagonist; VTE: venous thromboembolism.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate‐certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low‐certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low‐certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

aDowngraded one level due to serious imprecision, 95% CI was consistent with the possibility for important benefit (62 per 1000 absolute reduction) and possibility of important harm (167 per 1000 absolute increase), included 85 events.

bDowngraded two level due to very serious imprecision, 95% CI was consistent with the possibility of important benefit (65 fewer per 1000) and possibility of important harm (25 more per 1000); included 15 events.

cDowngraded two levels due to very serious imprecision, 95% CI was consistent with the possibility for important benefit (25 per 1000 absolute reduction) and possibility of important harm (98 per 1000 absolute increase), included 11 events.

Figures and Tables -
Summary of findings 5. Idraparinux secondary prophylaxis compared to vitamin K antagonist secondary prophylaxis in people with cancer with venous thromboembolism
Navigate to table in Review
Table 1. Glossary

Term

Definition

Adjuvant therapy

A therapy given in addition to the primary treatment to decrease the risk of the cancer recurrence or to assist in the cure.

Anticoagulation

The process of hindering the clotting of blood especially by treatment with an anticoagulant.

Antithrombotic

Used against or tending to prevent thrombosis (clotting)

Coagulation

Clotting

Direct oral anticoagulants (DOAC)

Also known as NOACs are anticoagulant medications that require less monitoring compared to the traditional anticoagulants.

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 (as swelling and pain) and that is potentially life‐threatening if dislodgment of the thrombus results in pulmonary embolism.

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. 2 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 statistic

A measure of degree of nonrandom agreement between observers, measurements of a specific categorical variable, or both.

Metastasis

The spread of a cancer cells from the initial or primary site of disease to another part of the body.

Parenteral nutrition

The practice of feeding a person intravenously, circumventing the gastrointestinal tract.

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.

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. Warfarin is a vitamin K antagonist.

Warfarin

An anticoagulant medication that is a vitamin K antagonist that is used for anticoagulation.
Figures and Tables -
Table 1. Glossary
Navigate to table in Review
Comparison 1. Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA)

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1.1 All‐cause mortality (up to 6 months) (main analysis ‐ active cancer) Show forest plot

4

1712

Risk Ratio (M‐H, Random, 95% CI)

0.99 [0.88, 1.12]

1.2 All‐cause mortality (time‐to‐event) Show forest plot

2

810

HR (IV, Random, 95% CI)

0.94 [0.74, 1.20]

1.3 Recurrent venous thromboembolism (up to 6 months) (main analysis ‐ active cancer) Show forest plot

4

1712

Risk Ratio (M‐H, Random, 95% CI)

0.59 [0.44, 0.80]

1.4 Recurrent venous thromboembolism (time‐to‐event) Show forest plot

2

810

HR (IV, Random, 95% CI)

0.49 [0.31, 0.78]

1.5 Major bleeding (up to 6 months) (main analysis ‐ active cancer) Show forest plot

4

1712

Risk Ratio (M‐H, Random, 95% CI)

1.09 [0.55, 2.12]

1.6 Minor bleeding (up to 6 months) (main analysis ‐ active cancer) Show forest plot

4

1712

Risk Ratio (M‐H, Random, 95% CI)

0.78 [0.47, 1.27]

1.7 Thrombocytopenia (up to 6 months) (main analysis‐ active cancer) Show forest plot

1

138

Risk Ratio (M‐H, Random, 95% CI)

0.94 [0.52, 1.69]
Figures and Tables -
Comparison 1. Low molecular weight heparins (LMWH) versus vitamin K antagonists (VKA)
Navigate to table in Review
Comparison 2. Direct oral anticoagulants (DOAC) versus vitamin K antagonists (VKA)

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

2.1 All‐cause mortality (6‐12 months) Show forest plot

4

1060

Risk Ratio (M‐H, Random, 95% CI)

0.94 [0.72, 1.23]

2.2 Recurrent venous thromboembolism (6‐12 months) Show forest plot

4

1050

Risk Ratio (M‐H, Random, 95% CI)

0.63 [0.34, 1.15]

2.3 Major bleeding (6‐12 months) Show forest plot

4

1055

Risk Ratio (M‐H, Random, 95% CI)

0.77 [0.39, 1.53]

2.4 Minor bleeding (6‐12 months) Show forest plot

4

1055

Risk Ratio (M‐H, Random, 95% CI)

0.83 [0.57, 1.23]
Figures and Tables -
Comparison 2. Direct oral anticoagulants (DOAC) versus vitamin K antagonists (VKA)
Navigate to table in Review
Comparison 3. Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH)

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

3.1 All‐cause mortality (6 months) Show forest plot

5

2854

Risk Ratio (M‐H, Random, 95% CI)

0.97 [0.83, 1.14]

3.2 Recurrent VTE (6 months) Show forest plot

5

2854

Risk Ratio (M‐H, Random, 95% CI)

0.63 [0.45, 0.88]

3.3 Major bleeding (6 months) Show forest plot

5

2994

Risk Ratio (M‐H, Random, 95% CI)

1.20 [0.83, 1.73]

3.3.1 GI tract cancer

4

854

Risk Ratio (M‐H, Random, 95% CI)

1.47 [0.76, 2.84]

3.3.2 Non‐GI tract cancer

4

1853

Risk Ratio (M‐H, Random, 95% CI)

1.05 [0.63, 1.73]

3.3.3 GI tract cancer not specified

1

287

Risk Ratio (M‐H, Random, 95% CI)

0.20 [0.01, 4.04]

3.4 Major GI bleeding (6 months) Show forest plot

4

1838

Risk Ratio (M‐H, Random, 95% CI)

1.16 [0.62, 2.17]

3.5 Major upper GI bleeding (6 months) Show forest plot

4

1838

Risk Ratio (M‐H, Random, 95% CI)

1.18 [0.51, 2.76]

3.6 Major lower GI bleeding (6 months) Show forest plot

4

1838

Risk Ratio (M‐H, Random, 95% CI)

1.10 [0.43, 2.80]

3.7 Major non‐GI bleeding (6 months) Show forest plot

4

1838

Risk Ratio (M‐H, Random, 95% CI)

0.84 [0.42, 1.68]

3.8 Minor bleeding (6 months) Show forest plot

5

2854

Risk Ratio (M‐H, Random, 95% CI)

1.58 [1.15, 2.16]

3.9 Minor GI bleeding (6 months) Show forest plot

2

1495

Risk Ratio (M‐H, Random, 95% CI)

1.37 [0.41, 4.64]

3.10 Minor upper GI bleeding (6 months) Show forest plot

2

1495

Risk Ratio (M‐H, Random, 95% CI)

1.03 [0.04, 25.97]

3.11 Minor lower GI bleeding (6 months) Show forest plot

2

1495

Risk Ratio (M‐H, Random, 95% CI)

1.54 [0.72, 3.27]

3.12 Minor non‐GI bleeding (6 months) Show forest plot

2

1495

Risk Ratio (M‐H, Random, 95% CI)

2.37 [1.44, 3.89]
Figures and Tables -
Comparison 3. Direct oral anticoagulants (DOAC) versus low molecular weight heparins (LMWH)
Navigate to table in Review
Comparison 4. Idraparinux versus vitamin K antagonists (VKA)

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

4.1 All‐cause mortality (up to 6 months) Show forest plot

1

284

Risk Ratio (M‐H, Random, 95% CI)

1.11 [0.78, 1.59]

4.2 Recurrent VTE (up to 6 months) Show forest plot

1

270

Risk Ratio (M‐H, Random, 95% CI)

0.46 [0.16, 1.32]

4.3 Major bleeding (up to 6 months) Show forest plot

1

270

Risk Ratio (M‐H, Random, 95% CI)

1.11 [0.35, 3.56]

4.4 Minor bleeding (up to 6 months) Show forest plot

1

270

Risk Ratio (M‐H, Random, 95% CI)

0.70 [0.30, 1.60]
Figures and Tables -
Comparison 4. Idraparinux versus vitamin K antagonists (VKA)
Navigate to table in Review