Tratamientos sistémicos para la prevención de los eventos tromboembólicos venosos en pacientes con cáncer pediátrico y catéteres venosos centrales canalizados
Resumen
Antecedentes
Los eventos tromboembólicos venosos (ETV) ocurren en un 2,2% a un 14% de los pacientes con cáncer pediátrico y causan morbilidad y mortalidad significativas. La enfermedad maligna en sí, el tratamiento del cáncer y la presencia de catéteres venosos centrales (CVC) aumentan el riesgo de ETV.
Objetivos
El objetivo primario de esta revisión fue investigar los efectos de los tratamientos sistémicos preventivos en pacientes con cáncer pediátrico y CVC canalizados sobre los ETV sintomáticos o asintomáticos. Los objetivos secundarios de esta revisión fueron investigar los efectos adversos de los tratamientos sistémicos para la prevención de los ETV sintomáticos o asintomáticos en pacientes con cáncer pediátrico y CVC canalizados; e investigar los efectos de los tratamientos sistémicos en la prevención de los ETV sintomáticos o asintomáticos con infección relacionada al CVC en pacientes con cáncer pediátrico y CVC canalizados.
Métodos de búsqueda
Se hicieron búsquedas en el Registro Cochrane Central de Ensayos Controlados (Cochrane Central Register of Controlled Trials) (CENTRAL) (The Cochrane Library, número 8, 2012), MEDLINE (1966 hasta agosto 2012) y en EMBASE (1966 hasta agosto 2012). Además, se hicieron búsquedas en listas de referencias de artículos relevantes y en las actas de congresos de la International Society for Paediatric Oncology (SIOP) (desde 2006 hasta 2011), la American Society of Clinical Oncology (ASCO) (desde 2006 hasta 2011), la American Society of Hematology (ASH) (desde 2006 hasta 2011) y en la International Society of Thrombosis and Haematology (ISTH) (desde 2006 hasta 2011). Se examinó el International Standard Randomised Controlled Trial Number (ISRCTN) Register y el National Institute of Health (NIH) Register para obtener ensayos en curso (www.controlled‐trials.com) (agosto 2012) y se contactó con los autores de estudios elegibles, si se requería información adicional.
Criterios de selección
Ensayos controlados aleatorios (ECA) y ensayos clínicos controlados (ECC) que comparaban tratamientos sistémicos para prevenir los eventos tromboembólicos venosos (ETV) en pacientes con cáncer pediátrico y CVC canalizados versus una intervención de control o ningún tratamiento sistémico. Para la descripción de los eventos adversos, fueron elegibles para la inclusión los estudios de cohortes.
Obtención y análisis de los datos
Dos autores de la revisión seleccionaron los estudios de forma independiente, extrajeron los datos y realizaron la evaluación del riesgo de sesgo de los estudios incluidos. Los análisis se realizaron según las guías del Manual Cochrane de Revisiones Sistemáticas de Intervenciones (Cochrane Handbook for Systematic Reviews of Interventions).
Resultados principales
Tres ECA y tres ECC (con 1291 niños) investigaron la prevención de los ETV (heparina de bajo peso molecular [HBPM] n = 134; administración de suplementos de antitrombina [AT] n = 37; warfarina de dosis baja n = 31; administración de suplementos de plasma fresco congelado [PFC] o crioprecipitado n = 240; administración de suplementos de AT y HBPM n = 41). Sólo se administró AT y PFC y crioprecipitado en los casos de deficiencia de AT o fibrinógeno. De los seis ECA/ECC incluidos, cinco investigaron la prevención de los ETV en comparación con ninguna intervención (n = 737), y un ECC comparó la administración de suplementos de AT y HBPM con la administración de suplementos de AT (n = 71). Todos los estudios presentaban limitaciones metodológicas y hubo heterogeneidad clínica entre los estudios.
No se encontró ningún efecto significativo de los tratamientos sistémicos en comparación con ninguna intervención en la prevención de los ETV sintomáticos o asintomáticos ni diferencias en los eventos adversos (como hemorragia grave o leve; ninguno de los estudios informó trombocitopenia, trombocitopenia inducida por heparina [TIH], trombocitopenia inducida por heparina con trombosis [TIHT], la muerte como resultado de los ETV, la extracción del CVC debido a los ETV, infección relacionada con el CVC y síndrome postrombótico [SPT]) entre los grupos experimentales y de control. Se incluyeron dos estudios con grupos de participantes e intervenciones comparables para los metanálisis (n = 182). En el grupo experimental, 1/68 niños (1,5%) fueron diagnosticados con trombosis venosa sintomática, al igual que 4/114 (3,5%) en el grupo de control (mejor de los casos: cociente de riesgos [CR] 0,65; intervalo de confianza [IC] del 95%: 0,09 a 4,78). Estos estudios también evaluaron los ETV asintomáticos relacionados con el CVC: En el grupo experimental, 22/68 (32,4%) fueron diagnosticados con trombosis venosa asintomática, al igual que 35/114 (30,7%) en el grupo de control (mejor de los casos: CR 1,02; IC del 95%: 0,40 a 2,55). La heterogeneidad fue apreciable para este análisis: I2 = 73%.
La atribución de HBPM a la administración de suplementos de AT dio lugar a una reducción significativa en los ETV sintomáticos (prueba exacta de Fisher, p bilateral = 0,028) sin complicaciones hemorrágicas; no fueron evaluados los ETV asintomáticos, la trombocitopenia, la TIH, la TIHT, la muerte como resultado de los ETV, la extracción del CVC debido a los ETV, la infección relacionada con el CVC ni el SPT.
Se incluyeron cuatro estudios de cohortes para la evaluación de los eventos adversos. Tres estudios proporcionaron información sobre los episodios de hemorragia: Un participante presentó un accidente cerebrovascular isquémico‐hemorrágico. Un estudio proporcionó información sobre otros eventos adversos: No ocurrió ninguno.
Conclusiones de los autores
No se encontró ningún efecto significativo de los tratamientos sistémicos en comparación con ninguna intervención en la prevención de los ETV sintomáticos o asintomáticos en pacientes oncológicos pediátricos con CVC. Sin embargo, este resultado podría deberse al número bajo de participantes incluidos, que dio lugar a un poder estadístico reducido. En un ECC, que comparó un tratamiento sistémico con otro tratamiento sistémico, se identificó una reducción significativa de los ETV sintomáticos con el agregado de HBPM a la administración de suplementos de AT.
Todos los estudios investigaron la prevalencia de episodios de hemorragia grave o leve, y ninguno encontró una diferencia significativa entre los grupos de estudio. Ninguno de los estudios informó trombocitopenia, TIH, TIHT, la muerte como resultado de los ETV, la extracción del CVC debido a los ETV, infección relacionada con el CVC o SPT entre los participantes.
Según las pruebas actualmente disponibles, no se pueden brindar recomendaciones para la práctica clínica. Se necesitan ECA internacionales adicionales bien diseñados para explorar aún más los efectos de los tratamientos sistémicos en la prevención de los ETV. Los estudios futuros deben intentar lograr un poder estadístico adecuado con tamaños de la muestra alcanzables. La incidencia de ETV sintomáticos es relativamente baja; por lo tanto, quizá sea necesario seleccionar a participantes con factores de riesgo trombótico o investigar los ETV asintomáticos en su lugar.
PICO
Resumen en términos sencillos
Prevención de la trombosis en niños con cáncer y CVC canalizados
Los niños con cáncer se encuentran en mayor riesgo de trombosis que los niños que no presentan la enfermedad. La misma es resultado de la enfermedad en sí y también del tratamiento del cáncer y la presencia de un catéter venoso central. En esta revisión, se investigó si los tratamientos sistémicos pueden prevenir la trombosis. Se identificaron seis estudios; dos estudios investigaron las heparinas de bajo peso molecular, uno la administración de suplementos de antitrombina y uno la administración de suplementos de plasma fresco congelado o crioprecipitado; un estudio comparó la administración de suplementos de antitrombina con heparina de bajo peso molecular y suplementos de antitrombina, y otro investigó la warfarina. El agregado de heparinas de bajo peso molecular a los suplementos de antitrombina dio lugar a un número inferior de trombosis sintomáticas. Esto fue estadísticamente significativo. No fue posible detectar un efecto de los tratamientos sistémicos preventivos en comparación con ningún tratamiento, y no se observó ninguna diferencia en el número de participantes que sufrieron hemorragia grave o leve. Sin embargo, el número total de participantes fue muy pequeño; un estudio similar con una población mayor de participantes podría dar resultados diferentes.
Authors' conclusions
Background
Description of the condition
Venous thromboembolic disease (VTE) is an important complication in paediatric cancer patients. VTE comprises both deep vein thrombosis (DVT) and pulmonary embolism (PE), and the incidence varies between 2.2% and 14% (Knofler 1999; Molinari 2001; Wilimas 1998). This wide range can be explained by inconsistencies amongst study groups with regard to selection of study participants, definition of asymptomatic or symptomatic VTE and different diagnostic techniques.
The aetiology of VTE in paediatric cancer patients is multifactorial, consisting of both genetic and acquired factors (Blom 2005). Various genetic alterations lead to congenital defects in the clotting cascade, thereby contributing to the risk of VTE (Young 2008). More important is the role of acquired conditions, including the malignancy itself, cancer treatment and the presence of a central venous catheter (CVC) (Farinasso 2007; Lee 1999). It is known from adult studies that tumour cells can influence the haemostatic system directly. On the one hand, tumour cells produce procoagulant molecules, fibrinolytic molecules and inflammatory cytokines, thereby inducing a prothrombotic state. On the other hand, tumour cells directly activate platelets and monocytes or macrophages, or stimulate endothelial cells, which results in down‐regulation of anticoagulant properties and up‐regulation of procoagulant properties (Athale 2003). In children, haematological malignancies, such as acute lymphoblastic leukaemia (ALL), are frequently associated with thrombosis (Mitchell 2010). In the induction phase of treatment, patients receive induction chemotherapy with asparaginase and steroids, both inducing a prothrombotic state by suppression of anticoagulant proteins and by elevation of factor VIII/von Willebrand factor complex, respectively (Van Ommen 2009). Finally, a very important risk factor for VTE is the presence of a CVC. Paediatric oncology patients receive a CVC for the administration of chemotherapy, antibiotics, blood products and total parenteral nutrition and for withdrawal of blood (Andrew 1994; Farinasso 2007; Massicotte 1998; Revel‐Vilk 2009; Van Ommen 2001). Placement of the CVC and subsequent infused substances, such as total parenteral nutrition and chemotherapy, cause injury to the vessels and activate the clotting cascade (Cunningham 2006). The CVC itself also contributes to the development of thrombosis: The CVC compromises blood flow, and the material is thrombogenic (Athale 2007; Journeycake 2003).
Paediatric catheter‐related VTE causes significant morbidity and mortality (Heit 2000). Up to 4% of paediatric patients die as a direct result of VTE, predominantly because of pulmonary embolisms (Van Ommen 2009). These figures are not known for paediatric oncology patients. Morbidity includes mechanical obstruction of the CVC, CVC‐related infection, PE, recurrent VTE and the post‐thrombotic syndrome (PTS). Journeycake et al demonstrated an association between CVC‐related DVT and both CVC‐occlusion and CVC‐related infection (Journeycake 2006). In a retrospective study of 287 paediatric participants with malignancies and a CVC, 21 participants developed DVT. Among these participants, CVC‐occlusions occurred more frequently than in participants without DVT (6.2 vs 1.6 episodes per 1000 catheter days). Similarly, participants with a DVT were at higher risk for CVC‐related infection (2.7 vs 1.2 catheter‐related infections per 1000 catheter days). Both CVC‐related infections and mechanical obstructions often necessitate replacement of the CVC, resulting in additional anaesthesia, surgical interventions and subsequent treatment delay. Additionally, thrombi cause permanent damage to the vessel wall and make patients prone to recurrent VTE, especially as long as the risk factors, such as cancer treatment, persist. Besides, patients may develop PTS after symptomatic but also after asymptomatic thrombosis. The spectrum of PTS of the upper venous system is broad and varies from collateral veins in the skin to chronic superior caval vein syndrome (Van Ommen 2009).
Description of the intervention
As CVC‐related VTE leads to significant mortality and morbidity in paediatric oncology patients, efficacious and safe preventive strategies, including systemic treatments, are of great importance. In this review, we evaluated the efficacy and safety of systemic treatments for the prevention of VTE in paediatric cancer patients with CVCs. Systemic treatments currently available for the prevention of VTE include anticoagulants such as unfractionated heparin (UH), low molecular weight heparin (LMWH), vitamin K antagonists and antithrombin concentrates (Akl 2007).
Why it is important to do this review
In 2007 a Cochrane review was published by Akl et al describing the systemic prevention of VTE in adult cancer patients. These investigators found a trend for heparin (UH and LMWH) versus no intervention or placebo towards a reduction in symptomatic VTE. When they pooled data from all available anticoagulants, they did find a significant result (i.e. a reduction in risk ratio of 0.56, 95% confidence interval (CI) 0.34 to 0.92) (Akl 2007). In this review, however, they excluded two paediatric studies: one because patients with non‐malignant diseases were included and outcomes could not be retrieved for cancer patients separately (Massicotte 2003), and another by Ruud et al (Ruud 2006b) because of its focus on a paediatric population (Akl 2007).
Preventive treatments thus far have been studied most often in adult patients. The underlying malignancy and its treatment are important risk factors for VTE; therefore, these results cannot be extrapolated to paediatric patients. Children develop tumours that differ from adult tumours and are thus treated with different treatment regimens. Children with ALL, for example, are treated on protocols containing asparaginase. Asparaginase reduces antithrombin; supplementation of antithrombin would therefore be a suitable prophylaxis in these treatment regimens (Van Ommen 2009). In this way, haemostasis is different, especially in young children (Andrew 1994). Plasma values of the natural anticoagulant protein C are 50% of adult plasma values after birth and reach adult values during puberty. Finally, some preventive anticoagulants used in adults might not be useful in children. Many paediatricians prefer to give LMWH instead of vitamin K antagonists to children with cancer. In contrast to vitamin K antagonists, LMWH can easily be stopped and restarted in cases of thrombocytopenia and lumbar puncture. On the other hand, daily subcutaneous injections can be stressful for children, and the risk of anticoagulant treatments in children with cancer should not be underestimated. In the light of this, we conducted a systematic review to investigate the efficacy and safety of systemic treatments for the prevention of VTE in paediatric cancer patients with tunnelled CVCs.
Objectives
The primary objective of this review was to investigate the effects of preventive systemic treatments in paediatric cancer patients with tunnelled CVCs on (a)symptomatic VTE. Secondary objectives of this review were to investigate adverse effects of systemic treatments for the prevention of (a)symptomatic VTE in paediatric cancer patients with tunnelled CVCs; and to investigate the effects of systemic treatments in the prevention of (a)symptomatic VTE with CVC‐related infection in paediatric cancer patients with tunnelled CVCs.
Methods
Criteria for considering studies for this review
Types of studies
-
Randomised controlled trials (RCTs) comparing participants given systemic treatments for the prevention of (a)symptomatic VTE with a control group.
-
Controlled clinical trials (CCTs) comparing participants given systemic treatments for the prevention of VTE with a control group.
-
Cohort studies investigating systemic treatments for the prevention of VTE without a control group. These studies were not used to evaluate the efficacy of prevention but to describe safety. The quality of cohort studies is considered by The Cochrane Collaboration to be too low for efficacy analyses (Higgins 2011).
Types of participants
Paediatric oncology patients (0 to 18 years) with a tunnelled central venous catheter. Oncology patients should account for at least 50% of the study population.
Types of interventions
Systemic treatments (in both prophylactic and therapeutic dosages) for the prevention of venous thrombo‐embolic events.
Types of outcome measures
Primary outcomes
-
(A)symptomatic VTE: deep venous thrombosis (diagnosis: clinical evaluation and confirmation by ultrasound or venography) and pulmonary embolism (diagnosis: clinical evaluation, angiography, ventilation/perfusion (V/Q) scan, spiral computed tomography (CT), magnetic resonance (MR) angiography, echocardiography) (Stein 2006; The PIOPED investigators 1990).
Secondary outcomes
-
Different adverse events: major bleeding, minor bleeding, thrombocytopenia, heparin‐induced thrombocytopenia (HIT), heparin‐induced thrombocytopenia with thrombosis (HITT) (Akl 2007).
-
Death due to (a)symptomatic VTE.
-
Removal of CVC because of (a)symptomatic VTE.
-
CVC‐related infection and CVC‐related bacteraemia as defined by the authors of the original studies (Mermel 2009).
-
Post‐thrombotic syndrome (Goldenberg 2010).
Search methods for identification of studies
See Cochrane Childhood Cancer Group methods used in reviews (Module CCG).
Electronic searches
We searched the following electronic databases: the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, issue 8, 2012); MEDLINE/PubMed (from 1945 to August 2012) and EMBASE/Ovid (from 1980 to August 2012).
The search strategies used for the different electronic databases (with a combination of controlled vocabulary and text words) are presented in the appendices (Appendix 1, Appendix 2, Appendix 3).
Searching other resources
We located information about trials not registered in CENTRAL, MEDLINE or EMBASE, published or unpublished, by searching the reference lists of relevant articles and review articles. We handsearched the conference proceedings of the International Society for Paediatric Oncology (SIOP) (from 2006 to 2011), the American Society of Clinical Oncology (ASCO) (from 2006 to 2011), the American Society of Hematology (ASH) (from 2006 to 2011) and the International Society on Thrombosis and Haemostasis (ISTH) (from 2006 to 2011). We scanned the International Standard Randomised Controlled Trial Number (ISRCTN) Register and the National Institute of Health (NIH) Register for ongoing trials (http://www.controlled‐trials.com; August 2012).
We did not impose language restrictions, and we will update the searches every two years.
Data collection and analysis
Selection of studies
After employing the search strategy described previously, two review authors independently identified studies meeting the inclusion criteria for this review. They were not blinded to the journal, author or institution. We resolved discrepancies between authors by consensus. If no agreement could be reached, we asked for the opinion of a third‐party arbitrator. We obtained in full for closer inspection any study seemingly meeting inclusion criteria on the grounds of the title or abstract, or both. We contacted authors for additional information when necessary. We clearly stated details of reasons for exclusion of any study considered for review and showed all excluded studies in a flow chart.
Data extraction and management
Two review authors independently performed data extraction using standardised forms. Data extraction included characteristics of the participant group involved, the intervention described, the outcome assessed and the follow‐up provided. We contacted authors for additional information when necessary. We resolved disagreements by consensus. If no agreement could be reached, we asked for the opinion of a third‐party arbitrator.
Assessment of risk of bias in included studies
Two review authors independently undertook the assessment of risk of bias of included RCTs and CCTs (i.e. selection bias, performance bias, detection bias, attrition bias and reporting bias). We used the risk of bias items as described in the module of the Childhood Cancer Group (Module CCG), which are based on the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011); to assess reporting bias, we compared the methods section of included studies with their results section. We contacted authors for additional information when necessary. We resolved discrepancies between authors by consensus. If no agreement could be reached, we asked for the opinion of a third‐party arbitrator.
We presented the results of the risk of bias assessment (i.e. how each trial scored on each risk of bias item) in the risk of bias table and in a methodological quality summary. We took the risk of bias in included studies into account in interpreting the results of the review.
For cohort studies, no risk of bias assessment was performed.
Measures of treatment effect
For dichotomous outcomes, we expressed the effect estimate as risk ratio (RR) and presented all results with corresponding 95% confidence intervals (CIs). When only one study was available and no events were reported in one of the treatment groups, it was not appropriate to calculate the RR, its 95% CI and the corresponding P value. For these outcomes, we calculated Fisher's exact P value instead, using PASW Statistics (SPSS) for Windows, version 19 (SPSS Inc, Chicago, Illinois, USA).
Unit of analysis issues
If trials other than those with a simple parallel design, such as cluster‐randomised trials or cross‐over trials, would have been included, we would have taken appropriate steps to avoid unit of analysis error. However, because only studies with a parallel design were included, this was not applicable.
Dealing with missing data
When relevant data were missing, we attempted to contact the authors to retrieve the missing data. We extracted data by allocation intervention, irrespective of compliance with the allocated intervention, to allow an intention‐to‐treat analysis. If this was not possible, we stated this and performed an 'as treated' analysis.
Assessment of heterogeneity
We assessed heterogeneity both by visual inspection of the forest plots and by completion of a formal statistical test for heterogeneity (i.e. the I2 statistic). If significant heterogeneity (I2 ≥ 50%) occurred, we explored possible reasons for it and took appropriate measures, such as using a random‐effects model to derive the estimation of treatment effects throughout the review.
Assessment of reporting biases
In addition to the evaluation of reporting bias as described in the 'Assessment of risk of bias' section, we planned to assess reporting bias by constructing a funnel plot when the number of included studies was sufficient (i.e. at least 10 studies included in a meta‐analysis) because otherwise, the power of the tests is too low to distinguish chance from real asymmetry (Higgins 2011). Because none of our meta‐analyses included at least 10 studies, this was not applicable.
We took the following measures to reduce reporting bias:
-
We searched multiple electronic databases, proceedings of scientific meetings and trial registries to deal with location and time lag bias;
-
We kept no language restriction in the search strategy; and
-
We excluded duplicate reports of the same study to avoid duplicate publication bias.
Data synthesis
We entered data into RevMan and undertook analyses according to the guidelines of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). The primary aim was to perform a pooled analysis; however, if groups were not comparable, we summarised the results descriptively. RCTs and CCTs were analysed separately. Ruud et al performed ultrasonography for asymptomatic VTE at one, three and six months (Ruud 2006a). We did not incorporate the difference in time points in our analyses; we included in our analyses the number of participants with an event, not the number of events.
Subgroup analysis and investigation of heterogeneity
We planned to look at patients with haematological and solid malignancies separately and to perform subgroup analyses for different age groups and for patients with inherited thrombophilia traits.
Sensitivity analysis
We performed a sensitivity analysis for the risk of bias criteria used (i.e. excluding studies with a high risk of bias for which the risk of bias was unclear and comparing the results of studies with a low risk of bias with the results of all available studies) for all analyses that included more than one study.
Results
Description of studies
See Characteristics of included studies table; Characteristics of excluded studies table; and Figure 1. Searches of CENTRAL, MEDLINE and EMBASE identified 450 titles of reports of potentially relevant studies. We excluded 400 reports on the basis of title and abstract. The remaining 50 reports were screened by full text analysis. Of these reports, 45 were excluded. After searching the conference proceedings, we identified an additional 42 potentially relevant abstracts. One of the 42 potentially relevant abstracts was published in full text in 2010 (Harlev 2010), and another study was available as an abstract only (Brasseur 2007). The remaining 40 reports were excluded. We searched for ongoing studies in the ISRCTN Register and the NIH ClinicalTrials.gov register. None of the ongoing trials were relevant to this review. A complete list with reasons for exclusion is presented in the table Characteristics of excluded studies. Three studies were identified by searching reference lists of relevant studies and are included in this study (Abbott 2009, Elhasid 2001, Zaunschirm 1986). Thus we included ten studies. Four studies did not contain a control group and are presented in a separate overview Table (Table 1). The remaining six studies (three RCTs and three CCTs) were included for analyses.
Flow diagram of selection of studies.
| Study | Patients | Interventions | Outcomes |
| 18 paediatric oncology patients with ALL and a genetic predisposition for VTE | LMWH prophylaxis: 18 participants were found to harbour a genetic predisposition for VTE (heterozygosity for prothrombin gene mutation (n = 6), heterozygosity for factor V Leiden mutation (n = 12)) and received LMWH prophylaxis. Enoxaparin was administered 1 dd, 1 mg/kg, subcutaneously | 3/18 participants receiving prophylaxis developed a CVC‐related VTE: two symptomatic and one asymptomatic (discovered by cardiac ultrasound)
Adverse events or episodes of bleeding were not reported | |
| 8 paediatric oncology patients with ALL and increased risk of VTE | LMWH prophylaxis: All participants with increased risk of VTE were eligible for enoxaparin prophylaxis. 8/19 participants with increased risk of VTE received enoxaparin prophylaxis. The remaining 11 participants refused study participationa Enoxaparin was administered 1 dd, 1 mg/kg | 1/8 participants receiving prophylaxis developed a symptomatic VTE No episodes of bleeding occurred. Adverse events were not reported | |
| 13 paediatric oncology patients with ALL | FFP and/or AT III supplementation: Fibrinogen and AT III were determined every 2 to 3 days. FFP was administered if fibrinogen < 100 mg/dL, the amount calculated to raise fibrinogen levels to at least 120 mg/dL. FFP was given 3 dd. AT III was administered if AT III < 80% of normal but fibrinogen > 100 mg/dL, calculated to raise AT III levels to > 100% of normal (1 U/kg raises the level by 1%). AT III was given as a continuous infusion | 0/10 participants receiving FFP supplementation developed VTE 0/13 participants receiving AT III supplementation developed VTE One participant receiving FFP and AT III was diagnosed with disseminated intravascular coagulation preceding supplementation No episodes of bleeding or adverse events occurred | |
| 24 paediatric oncology patients with ALL | FFP supplementation: Fresh frozen plasma (FFP) was administered on alternate days (mean 11.5 ± 0.9 mL/kg). | 1/24 participants developed a VTE of the superior longitudinal sinus with ischaemo‐haemorrhagic stroke. No other episodes of bleeding were reported Adverse events were not reported |
ALL: acute lymphoblastic leukaemia; AT: antithrombin; CVC: central venous catheter; FFP: fresh frozen plasma; LMWH: low molecular weight heparin; VTE: venous thrombo‐embolic event.
Be aware: These studies cannot be used to evaluate the efficacy of prevention but can be used for the description of safety. The quality of cohort studies is considered by The Cochrane Collaboration to be too low for efficacy analyses (Higgins 2011).
aObtained from personal communication with the author.
Included studies
Characteristics of the included RCTs (n = 3: Massicotte 2003; Mitchell 2003; Ruud 2006) and CCTs (n = 3: Meister 2008; Elhasid 2001; Abbott 2009), including a total number of 1291 participants, are presented in the table Characteristics of included studies. Five studies compared systemic prophylactic treatment with no intervention. The following systemic treatments were investigated: LMWH (n = 2: Massicotte 2003; Elhasid 2001), antithrombin (AT; n = 1: Mitchell 2003), fresh frozen plasma (FFP) and/or cryoprecipitate (n = 1: Abbott 2009) and warfarin (n = 1: Ruud 2006). One study compared LMWH and AT supplementation with AT supplementation only (Meister 2008).
Four studies included only patients with ALL (Meister 2008; Mitchell 2003; Elhasid 2001; Abbott 2009), one study included all patients with malignancies (Ruud 2006) and one study included 92/186 patients with other, non‐malignant diseases (Massicotte 2003). All studies used different treatment schedules. Three studies investigated the effects of symptomatic VTE (Abbott 2009; Elhasid 2001; Meister 2008). Investigators monitored study participants closely and confirmed VTE with the appropriate radiodiagnostic test. One of these studies reported symptomatic VTE only in the central nervous system (Abbott 2009). Three studies investigated the effects of systemic prophylactic treatment on both symptomatic and asymptomatic VTE (Massicotte 2003; Mitchell 2003; Ruud 2006). One study defined that patients with symptomatic VTE were included in the intention‐to‐treat analysis but were excluded from further analyses (Ruud 2006). Asymptomatic VTE was assessed by different radiographic techniques. Massicotte et al investigated asymptomatic VTE with venography (Massicotte 2003), and Ruud et al performed ultrasound examination at one, three and six months after inclusion (Ruud 2006). Mitchell et al screened for asymptomatic VTE after completion of the induction phase of chemotherapy and during follow‐up (day 28 to a further three months) with the following radiographic tests: (1) bilateral venography or magnetic resonance imaging (MRI) of the upper body, (2) ultrasound of the upper body, (3) echocardiogram and (4) MRI of the head (Mitchell 2003). All studies investigated the occurrence of major and/or minor bleeding episodes.
For evaluation of adverse events, we included four publications on cohort studies (Table 1). All four studies investigated systemic treatments to prevent VTE: Two studies investigated LMWH (Harlev 2010; Mitchell 2010), one FFP supplementation (Brasseur 2007) and one FFP and/or AT supplementation (Zaunschirm 1986). Two studies selected patients who were at higher risk of VTE for systemic prophylaxis. Harlev et al selected ALL patients with a genetic predisposition for VTE (prothrombin gene mutation and heterozygosity for factor V Leiden mutation) (Harlev 2010). Mitchell et al developed a predictive model for identifying ALL patients at increased risk for VTE (Mitchell 2010). Patients with a score > 2.5 were eligible for systemic prophylaxis. All studies used different treatment schedules (Table 1).
Risk of bias in included studies
Data on the risk of bias assessment of the six included RCTs/CCTs are described in the risk of bias section of the Characteristics of included studies table and are presented in Figure 2 and Figure 3. All studies were found to have methodological limitations. For evaluation of internal validity, we assessed the risks of selection bias, performance bias, detection bias, attrition bias and reporting bias. We did not assess the risk of bias in the included cohort studies.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Selection bias
For evaluation of selection bias, we assessed random sequence generation and allocation concealment. Three of six studies were RCTs, and investigators performed random and concealed sequence allocation (Massicotte 2003; Mitchell 2003; Ruud 2006). The risk of selection bias in the three CCTs was high because no randomisation was performed (Abbott 2009; Elhasid 2001; Meister 2008).
Performance bias and detection bias
For evaluation of performance bias, we assessed the blinding of participants and personnel. All studies were open label, and neither participants nor personnel were blinded; therefore the risk of performance bias was high in all studies (Abbott 2009; Elhasid 2001; Massicotte 2003; Meister 2008; Mitchell 2003; Ruud 2006).
For evaluation of detection bias, we assessed the blinding of outcome assessors for all outcomes separately. In three studies (all CCTs), the outcome assessors were not blinded for all evaluated outcomes (Abbott 2009; Elhasid 2001; Meister 2008). In two studies, radiographic tests were assessed for VTE by a central adjudication committee not involved in participant care and blinded to treatment group (Massicotte 2003; Mitchell 2003). In the third study, radiologists performing the ultrasound examination were blinded to the treatment assignment (Ruud 2006). Also, the results were not available to treating clinicians unless participants presented symptoms of VTE. All other outcomes (symptomatic VTE, major bleeding, minor bleeding, other adverse events) were not assessed by blinded clinicians. These outcomes were assessed by regular clinicians, and because these studies were open label, they were aware of the treatment group.
Attrition bias
For evaluation of attrition bias, we assessed the completeness of outcome data for all outcomes separately. The risk for attrition bias was low in two studies (two CCTs) (Abbott 2009; Elhasid 2001). In one study, the number of excluded patients and the reasons for exclusion were not specified; therefore the risk of attrition bias was unclear (Meister 2008). In the other three studies, outcome data were reported incompletely. In the study by Massicotte et al, 28/186 participants were not evaluable for the primary efficacy analysis (Massicotte 2003). Mitchell et al and Ruud et al excluded 24 and 11 participants, respectively, from analyses after randomisation (high risk of attrition bias) (Mitchell 2003; Ruud 2006).
Reporting bias
For evaluation of reporting bias, we assessed selective reporting. All outcomes described in the method sections were evaluated. Thus, the risk of reporting bias was low in all six included studies (Abbott 2009; Elhasid 2001; Massicotte 2003; Meister 2008; Mitchell 2003; Ruud 2006).
Effects of interventions
Not all studies allowed data extraction for all endpoints. See the Characteristics of included studies table for a more detailed description of the extractable endpoints of each study.
Symptomatic VTE
Data on the number of patients with symptomatic VTE could be extracted from all six included studies (Abbott 2009; Elhasid 2001; Massicotte 2003; Meister 2008; Mitchell 2003; Ruud 2006).
Three RCTs and two CCTs (Abbott 2009; Elhasid 2001; Massicotte 2003; Mitchell 2003; Ruud 2006) compared systemic intervention with no intervention. One CCT (Meister 2008) compared one systemic intervention with another systemic intervention.
Systemic intervention versus no intervention
Results from two RCTs were evaluated in a meta‐analysis with a total of 182 participants (Mitchell 2003; Ruud 2006) (Figure 4). The RCT by Massicotte et al (Massicotte 2003) was excluded from the meta‐analyses, as this study included 49% of participants with non‐malignant disease.
Forest plot of comparison: 1 Systemic preventive treatment versus no intervention, symptomatic VTE, outcome: 1.1 Meta‐analysis.
As outcome reporting was incomplete, we performed a best case scenario and an as treated analysis. A total of 68 participants were randomly assigned to systemic treatment (AT or warfarin) and 114 to no intervention. Outcomes were assessed in 54 participants in the experimental group and in 93 in the control group. One participant in the experimental group developed symptomatic VTE, along with four in the control group. We found no significant difference between systemic prophylactic treatment and no intervention (best case scenario: RR 0.65, 95% CI 0.09 to 4.78, P = 0.67, as treated; RR 0.64, 95% CI 0.09 to 4.74, P = 0.67). No heterogeneity was detected (I2 = 0%).
In the RCT by Massicotte et al that was not included in the meta‐analyses, 92 participants were randomly assigned to LMWH (78 analysed for all outcomes) and 94 to no intervention (80 analysed for all outcomes) (Massicotte 2003) (Figure 5). We found no significant difference between LMWH and no intervention (best case scenario: RR 1.02, 95% CI 0.21 to 4.93, P = 0.98 as treated; RR 1.03, 95% CI 0.21 to 4.93, P = 0.97).
Forest plot of comparison: 1 Systemic preventive treatment versus no intervention, symptomatic VTE, outcome: 1.2 Massicotte 2003.
The two CCTs were evaluated separately because the study populations were not comparable. In the first CCT (Elhasid 2001), 41 participants were allocated to LMWH and 50 to no intervention. No participants in the experimental group and one participant in the control group developed symptomatic VTE. We found no significant difference between LMWH and no intervention (Fisher's exact test, two‐sided P = 1.00) (Table 2). In the second study (Abbott 2009), 240 participants were randomly assigned to FFP and/or cryoprecipitate supplementation (if AT or fibrogen was too low) and 479 to no intervention. No participants in the experimental group and seven in the control group developed central nervous system VTE. This difference was not statistically significant (Fisher's exact test, two‐sided P = 0.10) (Table 2).
| Studya | Intervention | Experimental group (n) | Control group (n) | Outcome* | Events (experimental) | Events (control) | Two‐sided P value |
| Systemic preventive treatment versus no intervention | |||||||
| Elhasid (ITT) | LMWH versus no intervention | 41 | 50 | Symptomatic VTE | 0 | 1 | 1.00 |
| Abbott (ITT) | FFP and/or CRY supplementation versus no intervention | 240 | 479 | Symptomatic VTE | 0 | 7 | 0.10 |
| Massicotte (best case scenario) | LMWH versus no intervention | 92 | 94 | Major bleeding | 0 | 1 | 1.00 |
| Massicotte (as treated) | LMWH versus no intervention | 78 | 80 | Major bleeding | 0 | 1 | 1.00 |
| Mitchell (best case scenario) | AT versus no intervention | 37 | 72 | Minor bleeding | 2 | 0 | 0.12 |
| Mitchell (as treated) | AT versus no intervention | 25 | 60 | Minor bleeding | 2 | 0 | 0.094 |
| Systemic preventive treatment versus other systemic treatment | |||||||
| Meister (unclear whether ITT or best case scenario) | AT + LMWH versus AT | 41 | 71 | Symptomatic VTE | 0 | 9 | 0.028 |
ITT: intention to treat; LMWH: low molecular weight heparin; AT: antithrombin; CRY: cryoprecipitate; VTE: venous thrombo‐embolic event
aThe definitions used in the different studies can be found in the table of included studies (Characteristics of included studies).
Systemic intervention versus another systemic intervention
In one CCT (Meister 2008), 41 participants were randomly assigned to AT and LMWH and 71 to AT only. No participants in the experimental group and nine in the control group developed VTE. This difference was statistically significant (Fisher's exact test, two‐sided P = 0.028) (Table 2).
Symptomatic and asymptomatic VTE
Asymptomatic VTE was reported in three RCTs (Massicotte 2003; Mitchell 2003; Ruud 2006).
Systemic intervention versus no intervention
We analysed the results from two RCTs in a meta‐analysis, with a total number of 182 participants (Mitchell 2003; Ruud 2006) (Figure 6). The RCT by Massicotte et al (Massicotte 2003) was excluded from the meta‐analysis as this study included 49% of participants with non‐malignant disease. A total of 68 participants were randomly assigned to systemic treatment (AT or warfarin) and 114 to no intervention. Outcomes were assessed in 54 participants in the experimental group and in 93 participants in the control group. In the experimental group, 22 participants developed asymptomatic or symptomatic VTE versus 35 in the control group. We found no significant difference between systemic prophylactic treatment and no intervention (best case scenario: RR 1.02, 95% CI 0.40 to 2.55, as treated RR 1.05, 95% CI 0.62 to 1.78, P = 0.86). Heterogeneity was substantial: I2 = 73% in the best case analysis, and it was 30% in the as treated analysis.
Forest plot of comparison: 2 Systemic preventive treatment versus no intervention, asymptomatic VTE and symptomatic VTE, outcome: 2.1 Meta‐analysis.
In the RCT by Massicotte et al, 92 participants were randomly assigned to LMWH and 94 to no intervention (Massicotte 2003) (Figure 7). Outcomes were measured in 78 and 80 participants, respectively. A total of 11 patients in the experimental group and 10 in the control group developed asymptomatic or symptomatic VTE. We found no significant difference between LMWH prophylaxis and no intervention (best case scenario: RR 1.12, 95% CI 0.50 to 2.52, P = 0.78 as treated; RR 1.13, 95% CI 0.51 to 2.50, P = 0.77).
Forest plot of comparison: 3 Systemic preventive treatment versus no intervention, asymptomatic VTE and symptomatic VTE, outcome: 3.2 Massicotte 2003.
Adverse events
Major bleeding
Systemic intervention versus no intervention
Major bleeding was mentioned as an outcome measure by all three RCTs and one CCT. We analysed the risk of major bleeding from two RCTs in a meta‐analysis (Mitchell 2003; Ruud 2006) (Figure 8). The RCT by Massicotte et al (Massicotte 2003) was excluded from the meta‐analyses because this study included 49% of participants with non‐malignant disease. A total of 68 participants were randomly assigned to systemic treatment (AT or warfarin) and 114 to no intervention. Bleeding complications were assessed in 54 and 93 participants, respectively. The risk of major bleeding was lower in participants receiving systemic prophylactic treatment compared with participants receiving no intervention, but this finding was not statistically significant (best case scenario: RR 0.39, 95% CI 0.05 to 3.31, P = 0.39 as treated: RR 0.38, 95% CI 0.05 to 3.08, P = 0.37). No heterogeneity was detected (I2 = 0%).
Forest plot of comparison: 3 Systemic preventive treatment versus no intervention, major bleeding, outcome: 3.1 Meta‐analysis.
No significant difference in major bleeding was found in the RCT by Massicotte et al between participants (n = 92) receiving LMWH and those given no intervention (n = 94) (Massicotte 2003). Bleeding complications were assessed in 78 and 80 participants, respectively (Fisher's exact test, two‐sided P = 1.00) (Massicotte 2003) (Table 2).
One CCT investigated only CNS haemorrhage, but no episodes of CNS haemorrhage occurred (Abbott 2009).
Systemic intervention versus another systemic intervention
In one CCT, no major bleeding episodes were reported (Meister 2008).
Minor bleeding
Systemic intervention versus no intervention
In two RCTs, episodes of minor bleeding were reported (Massicotte 2003; Mitchell 2003). One study reported 48 episodes of minor bleeding in 92 participants randomly assigned to LMWH prophylaxis and 41 episodes of minor bleeding in 94 participants randomly assigned to no intervention (Massicotte 2003) (Figure 9). Analyses of bleedings were performed in 78 and 80 participants, respectively. The difference was not statistically significant (best case scenario: RR 1.20, 95% CI 0.88 to 1.62, P = 0.24 as treated; RR 1.22, 95% CI 0.91 to 1.65, P = 0.19). In the other study, 2/37 participants randomly assigned to AT supplementation developed minor bleeding compared with 0 of 72 participants randomly assigned to no intervention (Mitchell 2003). Analyses of bleedings were performed in 25 and 60 participants, respectively. The difference was not statistically significant (best case scenario: Fisher's exact test, two‐sided P = 0.12, as treated; Fisher's exact, two‐sided P = 0.094) (Table 2).
Forest plot of comparison: 4 Systemic preventive treatment versus no intervention, minor bleeding, outcome: 4.1 Massicotte 2003.
Systemic intervention versus another systemic intervention
In one CCT, no minor bleeding episodes occurred (Meister 2008).
Any bleeding
One CCT did not state whether bleeding episodes were major or minor or were characterised as a combination (Elhasid 2001); no bleeding episodes were reported.
Thrombocytopenia
None of the studies reported thrombocytopenia or heparin‐induced thrombocytopenia and thrombosis (HITT) as an outcome measure.
Death due to (a)symptomatic VTE
None of the studies reported deaths due to VTE. Probably, the prevalence of death due to VTE is 0% in all studies.
Removal of CVC because of VTE
Systemic intervention versus no intervention
One study reported the number of CVCs prematurely removed because of CVC‐related infection (Ruud 2006). Four CVCs were removed because of CVC‐related infection. However among these four participants, two also had occlusive VTE, and one moderate jugular VTE. Whether these participants were treated in the experimental or the control group was not specified. No premature removal was mentioned in the other studies.
Catheter‐related infection and catheter‐related bacteremia
None of the studies provided adequate information on these outcomes.
Post‐thrombotic syndrome
None of the studies reported the number of participants who developed a post‐thrombotic syndrome.
Subgroup analyses
We planned to look at participants with haematological and solid malignancies separately or as specified by age group, but the included studies did not provide the data necessary, so no subgroup analysis was performed. None of the included studies provided outcome measures for participants with inherited thrombophilia traits, so no subgroup analyses were performed.
Sensitivity analyses for the used risk of bias criteria
No sensitivity analyses were possible because all studies included in the meta‐analyses had the same score for all risk of bias criteria.
Cohort studies (no control group included)
Cohort studies are described in detail in Table 1. In this section, we present only results for adverse events and episodes of bleeding, as this was the objective of including cohort studies. Treatment results are presented in Table 1. Three studies provided information on bleeding episodes: One participant developed an ischaemo‐haemorrhagic stroke. One study provided information on adverse events: No adverse events occurred.
Discussion
Paediatric oncology patients are at increased risk for the development of (a)symptomatic VTE. These thrombotic events cause significant morbidity and mortality in paediatric oncology patients during their cancer treatment.
In this review we evaluated the effects of systemic treatments in preventing VTE in paediatric cancer patients with tunnelled CVCs. We identified three RCTs and three CCTs. Five studies compared a systemic treatment with no intervention: two investigated LMWH, one warfarin, one AT and one FFP and/or cryoprecipitate supplementation. One CCT compared a systemic treatment with another systemic treatment: AT supplementation and LMWH prophylaxis with AT supplementation.
None of the studies comparing a systemic treatment with no intervention found a significant difference for (a)symptomatic VTE. Not all included studies were pooled. The meta‐analyses of two RCTs (Mitchell 2003; Ruud 2006) showed no significant difference between systemic prophylactic treatment (AT or warfarin) and no intervention for both symptomatic VTE and asymptomatic/symptomatic VTE. One CCT comparing systemic treatment with another systemic treatment (LMWH and AT supplementation vs AT supplementation) found a statistically significant reduction in symptomatic VTE in participants treated with LMWH and AT supplementation (Meister 2008). All included studies investigated major and/or minor bleeding episodes. None of the studies found a significant difference in major or minor bleeding episodes between groups. A meta‐analysis of two RCTs (Mitchell 2003; Ruud 2006) was performed, which showed no significant difference between systemic prophylactic treatment (AT or warfarin) and no intervention for major bleedings. None of the studies reported thrombocytopenia, HIT, HITT, death as a result of VTE, removal of CVC due to VTE, CVC‐related infection or PTS among participants. The prevalence of PTS in only 13 of 29 participants with catheter‐related asymptomatic thrombosis in the RCT of Mitchell et al was reported in a cross‐sectional study by Kuhle et al (Kuhle 2008). Seven of 13 participants (54%; 95% CI 25% to 81%) developed symptoms of PTS, which were graded as mild in six patients.
For evaluation of adverse events, we included four publications on cohort studies (Brasseur 2007; Harlev 2010; Mitchell 2010; Zaunschirm 1986). These studies did not have a control arm. Two studies investigated LMWH, one study FFP and/or AT III supplementation and one study FFP supplementation only. Three studies provided information on bleeding episodes: One participant developed an ischaemo‐haemorrhagic stroke (Brasseur 2007; Mitchell 2010; Zaunschirm 1986). One study investigated other adverse events: None occurred (Zaunschirm 1986).
Although we did not detect a significant effect of systemic preventive treatments, this does not imply that we proved there was no effect. We included six studies with a total number of 1291 participants. Because of different outcome definitions and heterogeneity in both study design and included participants, our meta‐analysis comprised only 182 participants (two studies). Furthermore, all studies were underpowered individually. Three studies were CCTs and included the maximum number of participants treated within a certain period of time without a pre‐defined power analysis (Abbott 2009; Elhasid 2001; Meister 2008). Despite the relatively large number of participants included in the study by Abbott et al (n = 719), the number of events was still low, and the difference between cohorts was not statistically significant (Abbott 2009). Not all participants received the intervention (FFP supplementation n = 86, cryoprecipitate supplementation n = 163) because supplementation is administered only when AT or fibrinogen levels drop below pre‐defined levels. Therefore in such studies, larger numbers are required to attain adequate power. The three RCTs did perform power calculations beforehand but were still underpowered. The PARKAA (Prophylactic Antithrombin Replacement in Kids with ALL Treated with Asparaginase) study was never developed to detect a significant difference (Mitchell 2003). The primary objective of this study was to determine the prevalence of (a)symptomatic VTE in the non‐treated arm. The aim of the antithrombin supplementation was to achieve safety data, not to investigate efficacy. Investigators in the PROTEKT (Prophylaxis of Thromboembolism in Kids) study sought to include 600 participants (Massicotte 2003). The authors assumed that 25% of participants would develop CVC‐related VTE and calculated a sample size (with a two‐sided type I error of 5%; 80% power) of 600 participants to demonstrate a relative risk reduction of 40% for participants in the experimental group. However, the study was prematurely closed at the request of the sponsor because of low accrual. Ruud et al aimed to include 140 participants (Ruud 2006). They assumed a 25% reduction in CVC‐related VTE with low‐dose warfarin (power 90%; significance level 5%) and calculated a sample size of 140 participants. After including 73 participants, they performed an interim analysis, and the study had to be stopped prematurely because of futility; the incidence of VTE was higher in the experimental group than in the control group. The sample size required for adequate power is determined by pretest probability and expected risk reduction. In cases of low pretest probability (i.e. low incidence), a larger sample size is required to detect a significant effect. We calculated the incidence of (a)symptomatic VTE in participants who received no systemic preventive treatments included in the RCTs of this review (as treated analysis). The incidence of symptomatic VTE was 4.0% (7/173), and the incidence of asymptomatic VTE was 22% (38/173). To detect a hypothetical 50% risk reduction in symptomatic or asymptomatic VTE (power 80%; two‐sided type I error 5%), a sample size of 1239 or 196 participants per group is required. As can be learned from the studies included in this review, such samples and thus adequate power within paediatric oncology are very difficult to achieve.
In the light of the higher incidence rate of asymptomatic VTE, it might be reasonable to base power calculations on these incidence numbers. However, the clinical relevance of asymptomatic VTE is not well established thus far. A known complication of asymptomatic VTE is the post‐thrombotic syndrome (PTS), which occurs in about 50% of ALL patients, as shown by Kuhle et al (Kuhle 2008). The precise prevalence of pulmonary embolism, death and recurrent VTE after asymptomatic VTE is still unknown. Thus, until the clinical relevance of asymptomatic VTE is completely elucidated, prophylactic studies should be based on established clinical relevant outcomes (i.e. symptomatic VTE).
To get at least a partial answer on the question of whether systemic treatment prevents symptomatic VTE, it would be worthwhile to select a group of patients at higher risk for (symptomatic) VTE. Mitchell et al developed and validated a predictive model to identify children treated for ALL who were at increased risk for VTE (Mitchell 2010). In this model, participants were scored and were subsequently allocated to a low‐risk group (score ≤ 2.5) or a high‐risk group (score > 2.5). Scores were based on potential risk factors for VTE (steroid treatment, treatment with Escherichia coli asparaginase, presence of CVC and genetic thrombophilic abnormalities) and were validated in a cohort of 339 participants with ALL. The rate of symptomatic VTE at 3.5 months was 2.5% in the low‐risk group and 64.7% in the high‐risk group with reduced thrombosis‐free survival for children in the high‐risk group (P < 0.001).
Adult studies, on the other hand, generally have less trouble including sufficient participants for adequate power. In a recently published Cochrane review by Di Nisio et al, the authors compared 1436 participants treated with LMWH with 1028 participants receiving placebo and found a significant reduction in symptomatic VTE in the group of participants treated with LMWH (RR 0.62, 95% CI 0.41 to 0.93) (Di Nisio 2012). In 2007 another systematic Cochrane review was published (and was updated in 2009) that investigated systemic treatments for the prevention of VTE (Akl 2007). Investigators also compared participants receiving LMWH (n = 465) with those given placebo (n = 313) and found a trend toward reduction in symptomatic DVT in the experimental group (RR 0.49, 95% CI 0.17 to 1.39). Both studies included one (other) paediatric study, included in this report (Ruud 2006, Mitchell 2003). However, neither of these reviews incorporated a full overview of paediatric evidence with the goal of making recommendations for paediatric oncology practise.
Similar to our findings, the reviews mentioned above did not detect a significant increase in major or minor bleeding among study participants receiving systemic prophylaxis (Akl 2007; Di Nisio 2012). Although we did see a trend toward lower risk of major bleeding in the experimental group in one of the included studies, this is likely to be caused by chance rather than indicating a true difference (Massicotte 2003). Massicotte et al reported increased episodes of minor bleeding, compared with the other included studies, in both experimental and control groups (Massicotte 2003). The authors applied a broad definition of minor bleeding as described in the study protocol, and most of these episodes would not be registered if this was not explicitly required by a prospective study. Bleeding is an important known side effect of VTE prophylaxis; therefore, future studies should thoroughly investigate major and minor bleeding episodes, so the clinical relevance of VTE prophylaxis can be properly weighed against the risks of such treatments.
Finally, the risk of bias in the included studies varied. However, at this time, this is the best available evidence based on RCTs and CCTs that compared systemic treatment with no intervention and systemic treatment with another systemic intervention in paediatric cancer patients with tunnelled CVCs. Although an RCT is the best study design for adequate determination of efficacy, provided that the design and the execution are correct, CCTs can also provide reliable information. Because of the high risk of bias associated with other study designs, we did not include cohort studies without control groups in our efficacy analyses. These studies were included for evaluation of adverse events, but no risk of bias assessment was performed.
Flow diagram of selection of studies.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Forest plot of comparison: 1 Systemic preventive treatment versus no intervention, symptomatic VTE, outcome: 1.1 Meta‐analysis.
Forest plot of comparison: 1 Systemic preventive treatment versus no intervention, symptomatic VTE, outcome: 1.2 Massicotte 2003.
Forest plot of comparison: 2 Systemic preventive treatment versus no intervention, asymptomatic VTE and symptomatic VTE, outcome: 2.1 Meta‐analysis.
Forest plot of comparison: 3 Systemic preventive treatment versus no intervention, asymptomatic VTE and symptomatic VTE, outcome: 3.2 Massicotte 2003.
Forest plot of comparison: 3 Systemic preventive treatment versus no intervention, major bleeding, outcome: 3.1 Meta‐analysis.
Forest plot of comparison: 4 Systemic preventive treatment versus no intervention, minor bleeding, outcome: 4.1 Massicotte 2003.
Comparison 1 Systemic preventive treatment versus no intervention, symptomatic VTE, Outcome 1 Meta‐analysis.
Comparison 1 Systemic preventive treatment versus no intervention, symptomatic VTE, Outcome 2 Massicotte 2003.
Comparison 2 Systemic preventive treatment versus no intervention, asymptomatic VTE and symptomatic VTE, Outcome 1 Meta‐analysis.
Comparison 2 Systemic preventive treatment versus no intervention, asymptomatic VTE and symptomatic VTE, Outcome 2 Massicotte 2003.
Comparison 3 Systemic preventive treatment versus no intervention, major bleeding, Outcome 1 Meta‐analysis.
Comparison 4 Systemic preventive treatment versus no intervention, minor bleeding, Outcome 1 Massicotte 2003.
| Study | Patients | Interventions | Outcomes |
| 18 paediatric oncology patients with ALL and a genetic predisposition for VTE | LMWH prophylaxis: 18 participants were found to harbour a genetic predisposition for VTE (heterozygosity for prothrombin gene mutation (n = 6), heterozygosity for factor V Leiden mutation (n = 12)) and received LMWH prophylaxis. Enoxaparin was administered 1 dd, 1 mg/kg, subcutaneously | 3/18 participants receiving prophylaxis developed a CVC‐related VTE: two symptomatic and one asymptomatic (discovered by cardiac ultrasound)
Adverse events or episodes of bleeding were not reported | |
| 8 paediatric oncology patients with ALL and increased risk of VTE | LMWH prophylaxis: All participants with increased risk of VTE were eligible for enoxaparin prophylaxis. 8/19 participants with increased risk of VTE received enoxaparin prophylaxis. The remaining 11 participants refused study participationa Enoxaparin was administered 1 dd, 1 mg/kg | 1/8 participants receiving prophylaxis developed a symptomatic VTE No episodes of bleeding occurred. Adverse events were not reported | |
| 13 paediatric oncology patients with ALL | FFP and/or AT III supplementation: Fibrinogen and AT III were determined every 2 to 3 days. FFP was administered if fibrinogen < 100 mg/dL, the amount calculated to raise fibrinogen levels to at least 120 mg/dL. FFP was given 3 dd. AT III was administered if AT III < 80% of normal but fibrinogen > 100 mg/dL, calculated to raise AT III levels to > 100% of normal (1 U/kg raises the level by 1%). AT III was given as a continuous infusion | 0/10 participants receiving FFP supplementation developed VTE 0/13 participants receiving AT III supplementation developed VTE One participant receiving FFP and AT III was diagnosed with disseminated intravascular coagulation preceding supplementation No episodes of bleeding or adverse events occurred | |
| 24 paediatric oncology patients with ALL | FFP supplementation: Fresh frozen plasma (FFP) was administered on alternate days (mean 11.5 ± 0.9 mL/kg). | 1/24 participants developed a VTE of the superior longitudinal sinus with ischaemo‐haemorrhagic stroke. No other episodes of bleeding were reported Adverse events were not reported | |
| ALL: acute lymphoblastic leukaemia; AT: antithrombin; CVC: central venous catheter; FFP: fresh frozen plasma; LMWH: low molecular weight heparin; VTE: venous thrombo‐embolic event. Be aware: These studies cannot be used to evaluate the efficacy of prevention but can be used for the description of safety. The quality of cohort studies is considered by The Cochrane Collaboration to be too low for efficacy analyses (Higgins 2011). aObtained from personal communication with the author. | |||
| Studya | Intervention | Experimental group (n) | Control group (n) | Outcome* | Events (experimental) | Events (control) | Two‐sided P value |
| Systemic preventive treatment versus no intervention | |||||||
| Elhasid (ITT) | LMWH versus no intervention | 41 | 50 | Symptomatic VTE | 0 | 1 | 1.00 |
| Abbott (ITT) | FFP and/or CRY supplementation versus no intervention | 240 | 479 | Symptomatic VTE | 0 | 7 | 0.10 |
| Massicotte (best case scenario) | LMWH versus no intervention | 92 | 94 | Major bleeding | 0 | 1 | 1.00 |
| Massicotte (as treated) | LMWH versus no intervention | 78 | 80 | Major bleeding | 0 | 1 | 1.00 |
| Mitchell (best case scenario) | AT versus no intervention | 37 | 72 | Minor bleeding | 2 | 0 | 0.12 |
| Mitchell (as treated) | AT versus no intervention | 25 | 60 | Minor bleeding | 2 | 0 | 0.094 |
| Systemic preventive treatment versus other systemic treatment | |||||||
| Meister (unclear whether ITT or best case scenario) | AT + LMWH versus AT | 41 | 71 | Symptomatic VTE | 0 | 9 | 0.028 |
| ITT: intention to treat; LMWH: low molecular weight heparin; AT: antithrombin; CRY: cryoprecipitate; VTE: venous thrombo‐embolic event aThe definitions used in the different studies can be found in the table of included studies (Characteristics of included studies). | |||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Meta‐analysis Show forest plot | 2 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 1.1 Best case scenario | 2 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 0.65 [0.09, 4.78] |
| 1.2 As treated analysis | 2 | 147 | Risk Ratio (M‐H, Random, 95% CI) | 0.64 [0.09, 4.74] |
| 2 Massicotte 2003 Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 Best case scenario | 1 | 186 | Risk Ratio (M‐H, Random, 95% CI) | 1.02 [0.21, 4.93] |
| 2.2 As treated analysis | 1 | 158 | Risk Ratio (M‐H, Random, 95% CI) | 1.03 [0.21, 4.93] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Meta‐analysis Show forest plot | 2 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 1.1 Best case scenario | 2 | 182 | Risk Ratio (M‐H, Random, 95% CI) | 1.02 [0.40, 2.55] |
| 1.2 As treated analysis | 2 | 147 | Risk Ratio (M‐H, Random, 95% CI) | 1.05 [0.62, 1.78] |
| 2 Massicotte 2003 Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 Best case scenario | 1 | 186 | Risk Ratio (M‐H, Random, 95% CI) | 1.12 [0.50, 2.52] |
| 2.2 As treated analysis | 1 | 158 | Risk Ratio (M‐H, Random, 95% CI) | 1.13 [0.51, 2.50] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Meta‐analysis Show forest plot | 2 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 1.1 Best case scenario | 2 | 182 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.39 [0.05, 3.31] |
| 1.2 As treated analysis | 2 | 147 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.38 [0.05, 3.08] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Massicotte 2003 Show forest plot | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 1.1 Best case scenario | 1 | 186 | Risk Ratio (M‐H, Random, 95% CI) | 1.20 [0.88, 1.62] |
| 1.2 As treated analysis | 1 | 184 | Risk Ratio (M‐H, Random, 95% CI) | 1.22 [0.91, 1.65] |
