Antiangiogenic therapy for high‐grade glioma
Abstract
Background
The most common primary brain tumours in adults are gliomas. Gliomas span a spectrum from low to high‐grade and are graded pathologically on a scale of one to four according to the World Health Organization (WHO) classification. High‐grade glioma (HGG) carries a poor prognosis. Grade IV glioma is known as glioblastoma (GBM) and carries a median survival in treated patients of about 15 months. GBMs are rich in blood vessels (i.e. highly vascular) and in a protein known as vascular endothelial growth factor (VEGF), which promotes new blood vessel formation (the process of angiogenesis). Antiangiogenic agents inhibit the process of new blood vessel formation and promote regression of existing vessels. Several antiangiogenic agents have been investigated in clinical trials in newly diagnosed and recurrent HGG, showing promising preliminary results. This review was undertaken to report on the benefits and harms associated with the use of antiangiogenic agents in the treatment of HGGs.
Objectives
To evaluate the efficacy and toxicity of antiangiogenic therapy in patients with high‐grade glioma. This intervention can be used in two broad groups of patients: those with first diagnosis as part of 'adjuvant' therapy, and those with recurrent or progressive disease. Comparisons will include the following.
• Treatment with antiangiogenic therapy versus placebo.
• Treatment (such as chemotherapy or chemoradiotherapy) with antiangiogenic therapy added versus the same treatment without the addition of antiangiogenic therapy.
Search methods
Searches were conducted to identify published and unpublished Randomised Controlled Trials (RCTs) starting in 2000; the following databases were searched: the Cochrane Central Register of Controlled Trials (CENTRAL), Issue 3, 2014; MEDLINE to April 2014 and EMBASE to April 2014. Proceedings of relevant oncology conferences since 2000 were handsearched.
Selection criteria
RCTs evaluating the use of antiangiogenic therapy versus control treatment without antiangiogenic therapy in the treatment of HGG.
Data collection and analysis
Review authors screened the search results and reviewed the abstracts of potentially relevant articles before retrieving the full text of eligible articles.
Main results
After a comprehensive literature search, seven eligible RCTs were identified (total of 2987 participants). Significant design heterogeneity was noted in the included studies, especially in the response assessment criteria used. All eligible studies were restricted to GBMs, and no eligible studies evaluated other HGGs. Four studies were available only in abstract form. We have reserved an overall assessment of the quality of the evidence until the final study publications are received. The three studies that have been published in full were judged to have low risk of bias. The seven trials of 2987 participants included in this systematic review did not show improvement in OS with the addition of antiangiogenic therapy (pooled hazard ratio (HR) 0.94, 95% confidence interval (CI) 0.86 to 1.02; P value 0.16). However, pooled analysis of PFS from six studies (2847 participants) showed improvement in PFS with the addition of antiangiogenic therapy (HR 0.74, 95% CI 0.68 to 0.81; P value < 0.00001). Bevacizumab was the antiangiogenic therapy more likely to yield favourable results. Pooled HR for PFS for bevacizumab studies (three studies with 1712 participants) was significant at 0.66 (95% CI 0.59 to 0.74; P value < 0.00001), and this was reflected in the lower hazard ratio reported in the pooled analysis of bevacizumab studies compared with the overall analysis. Nevertheless, this finding was not significant for OS (HR 0.92, 95% CI 0.83 to 1.02; P value 0.12). Similar to trials of antiangiogenic therapies in other solid tumours, adverse events related to this class of therapy included hypertension and proteinuria, poor wound healing and the potential for thromboembolic events, although generally, the occurrence of grade 3 events of this kind was low (< 14.1%), consistent with reported findings of studies of bevacizumab in other tumours.
Authors' conclusions
In patients with newly diagnosed GBM, the use of antiangiogenic therapy does not improve survival, despite evidence of improved progression‐free survival. Thus at this time, evidence is insufficient to support the use of antiangiogenic therapy in patients with newly diagnosed GBM on the basis of effects on survival.
Bevacizumab may confer a progression‐free survival benefit in GBM; however evidence in favour of using other antiangiogenic therapies in recurrent GBM is insufficient.
Although bevacizumab appears to prolong progression‐free survival in newly diagnosed and recurrent GBM, the impact of this on quality of life remains unclear.
Adequately powered, randomised, placebo‐controlled studies of bevacizumab in recurrent GBM (or HGG) are needed.
Not addressed here is whether subsets of patients with newly diagnosed GBM may benefit from antiangiogenic therapies and whether these therapies are useful in other high‐grade glioma histologies.
PICOs
Plain language summary
Drugs that target blood vessels in malignant brain tumours
Background
The most common primary brain tumours of adults are gliomas, which account for about two‐fifths of all primary brain tumours. Gliomas span a spectrum from low to high grade and are graded pathologically on a scale of one to four according to a classification by the World Health Organization (WHO). High‐grade glioma (HGG), including glioblastoma, or GBM, is difficult to treat and carries a poor prognosis.
These brain tumours form new blood vessels to help them grow. Drugs have been developed to reduce the formation of new blood vessels (angiogenesis) and to slow tumour growth. Bevacizumab, cediranib given directly and cilengitide given indirectly target blood vessel formation and have been studied in randomised clinical trials for the treatment of GBM.
Study characteristics
After a comprehensive search of the literature, seven eligible randomised clinical trials were identified (totaling 2987 participants). All eligible studies were restricted to GBMs, and no eligible studies included other brain tumour types. The largest trials were conducted in patients with newly diagnosed GBM who were treated with antiangiogenic therapy. Overall, the trials included in this systematic review did not show improvement in overall survival with the use of antiangiogenic therapy. However, clinical trials in bevacizumab‐treated GBM did prolong the time until tumour growth (progression‐free survival).
Key results
The adverse events observed were infrequent and were similar to those seen in trials of antiangiogenic therapies in other tumours. Adverse events included high blood pressure, loss of protein in the urine, poor wound healing and increased risk of blood clots.
In summary, we found insufficient evidence to show whether the antiangiogenic therapies evaluated so far prolong life in patients with high‐grade malignant brain tumours.
Authors' conclusions
Background
Description of the condition
High‐grade gliomas (HGGs), comprising World Health Organization (WHO) grade III tumours (e.g. anaplastic astrocytoma, anaplastic oligodendroglioma, anaplastic oligoastrocytoma) and grade IV tumours (e.g. glioblastoma, or GBM), represent 75% of primary brain tumours in adults (CBTRUS 2008; Louis 2007). Standard initial treatment for glioblastoma, a WHO grade IV tumour and the most common glioma variant, involves maximal surgical resection followed by radiotherapy with concurrent and then adjuvant chemotherapy with the DNA alkylator temozolomide (Khasraw 2010). The five‐year survival of patients with GBM is approximately 10% (Stupp 2009). WHO grade III tumours have a better prognosis than GBM but are likely to progress and follow a similar clinical course. Traditional prognostic factors include age, performance status, histology, symptom severity and extent of resection (Stupp 2009). The recursive partitioning analysis (RPA) classification uses a composite of these prognostic factors to define prognostic groupings (Curran 1993; Mirimanoff 2006). Molecular prognostic factors, including MGMT, 1p/19q LOH and isocitrate dehydrogenase (IDH) mutations, have been identified. MGMT is the promoter of hypermethylation of the methylguanine methyltransferase (MGMT) gene in glioblastoma. 1p/19q co‐deletion or loss of heterozygosity (LOH) involves loss of the short arm of chromosome 1 and the long arm of chromosome 19 in oligodendroglial tumours (1p/19q co‐deletion) and predicts a better prognosis. More recently identified mutations of prognostic importance include the genes encoding the enzyme IDH 1 and 2, which plays an important role in glucose metabolism (Yan 2009). These molecular markers are prognostic, but MGMT methylation has been associated with sensitivity to chemotherapy and radiotherapy (Stupp 2007).
Angiogenesis refers to the development of new blood vessels from pre‐existing vessels; abnormal angiogenesis has been implicated in disease processes such as malignant tumours (Fidler 1994; Folkman 1990). The dependence of tumour growth on the development of new blood vessels is now a well‐established aspect of cancer biology (Folkman 1971). Antiangiogenic therapy, the targeting of tumour blood vessels, interferes with tumour growth and spread in HGG, as described in the Description of the intervention section below, and is a novel anticancer strategy that has been the subject of various clinical trials in HGG and other cancers. Angiogenesis inhibitors have been developed that block new tumour blood vessel formation while inducing regression of existing tumour blood vessels.
Description of the intervention
Vascular endothelial growth factor (VEGF) is a protein and a key regulator of new blood vessel formation. Activation of the VEGF receptor starts a number of processes that promote cell growth and survival of tumour blood vessels. In addition, VEGF mediates leakiness of blood vessels and is involved in promoting circulating progenitor cells from the bone marrow to distant sites of new vessel formation (Hicklin 2005; Zebrowski 1999). The anti‐VEGF antibody (bevacizumab) binds circulating angiogenic factor VEGF‐A, prevents the formation of new blood vessels and has activity in recurrent glioblastoma (Khasraw 2010; Friedman 2009; Kreisl 2009). Other antiangiogenic agents have also demonstrated favourable activity in glioblastoma. Antiangiogenic agents may be classified as direct, indirect or mixed, depending on their mechanism of action and target (Gasparini 2005), as blood vessel formation can be targeted by several mechanisms (Sivakumar 2005).
-
Binding of circulating angiogenic factors, such as VEGF‐A (e.g. anti‐VEGF antibodies (bevacizumab)) or decoys such as VEGF‐trap (aflibercept).
-
Blockade of angiogenic factor cell surface receptors (R) (e.g. with anti‐VEGF‐R antibodies or with small molecular intracellular tyrosine kinase inhibitors (TKIs) (cediranib)).
-
Imitators of endogenous angiogenesis inhibitors (e.g. angiostatin, endostatin, thrombospondin).
-
Inhibition of tumour and endothelial cell adhesion and migration by integrin inhibitors (cilengitide).
Adverse effects of antiangiogenesis have been observed consistently in clinical studies of these agents and thus are considered a "class" effect; these include bleeding, hypertension, delayed wound healing, gastrointestinal perforation and thromboses (Chinot 2014; Batchelor 2013; Gilbert 2014).
How the intervention might work
Antiangiogenic therapy has been shown to 'normalise' tumour‐associated blood vessel structure and function and thus may improve vessel leakiness and pressure within tumours (Khasraw 2010). The effectiveness of chemotherapy can be improved by this normalisation of tumour blood vessels, leading to a reduction in surrounding tumour interstitial oedema and pressure in a way that enhances delivery of the cytotoxic agent to the tumour (Khasraw 2010). Also, continued therapy may eventually lead to vessel regression, thus depriving cancers of their nutrient source. Antiangiogenic therapy may "paradoxically" improve tumour oxygenation, thus enhancing the effectiveness of radiation therapy (Scaringi 2013). Preclinical evidence indicates that antiangiogenic agents may enhance the effectiveness of chemotherapy for established tumours through various other mechanisms and may inhibit new tumour growth by inhibiting the vital process of angiogenesis (Kerbel 2008).
Why it is important to do this review
The purpose of this review is to find, organise and summarise high‐level evidence in terms of benefits and harms of antiangiogenic therapy in patients with HGGs to provide meaningful conclusions for clinical practice and further research.
As drugs have progressed to phase III studies, so has our understanding of the biology of HGGs. Antiangiogenic therapy has become a key focus of translational HGG research, but no systematic reviews or meta‐analyses of clinical trial results in this field have been published to date. Interpretation of emerging clinical data in the face of our improved understanding of the biology is important, so that evidence is applied to deliver the most effective treatment for each patient.
Objectives
To evaluate the efficacy and toxicity of antiangiogenic therapy in patients with high‐grade glioma. This intervention can be used in two broad groups of patients: those with first diagnosis as part of 'adjuvant' therapy, and those with recurrent or progressive disease. Comparisons will include the following.
-
Treatment with antiangiogenic therapy versus placebo.
-
Treatment (such as chemotherapy or chemoradiotherapy) with antiangiogenic therapy added versus the same treatment without the addition of antiangiogenic therapy.
Methods
Criteria for considering studies for this review
Types of studies
Only randomised controlled trials (RCTs) comparing antiangiogenic therapy versus a control treatment without antiangiogenic therapy in participants with high‐grade glioma were examined.
Types of participants
Patients with a histological diagnosis of WHO grade III glioma or grade IV glioma (glioblastoma).
Types of interventions
Studies were included when the agent evaluated was mechanistically described as an angiogenesis inhibitor. Therefore, agents targeting multiple molecular pathways were included only when at least one was clearly recognised as an important angiogenesis pathway. Studies that provided a secondary intervention in the intervention group (such as chemotherapy or radiotherapy) were included, as long as this secondary intervention was the same in the control group. The control group could receive placebo/best supportive care or an active intervention (such as chemotherapy), as long as antiangiogenic therapy was not provided. Studies from both the adjuvant setting (early diagnosis) and the relapse setting were included. When the protocol was designed, numerous other possibilities for comparison were considered; however, these have been discarded to simplify the analysis and avoid confusion; they can be found in Appendix 1.
Types of outcome measures
Studies including at least one of the following outcomes were considered for evaluation.
Primary outcomes
Overall survival (OS), defined as time from randomisation to death.
Secondary outcomes
-
Progression‐free‐survival (PFS), defined as time from randomisation to death from any cause or disease progression.
-
-
Disease progression as defined according to the widely used RECIST (Therasse 2000) criteria, which are used for solid tumours but not for brain tumours. Therefore, PFS was assessed according to the Macdonald criteria (Macdonald 1990) or the new International Working Party criteria (response assessment in neuro‐oncology (RANO)) (Wen 2010).
-
-
Debate in the literature regarding the best method of assessing progression radiologically is ongoing (Wen 2010). Although the Macdonald criteria have been accepted as the standard of care for assessment of progression, the RANO criteria have been developed by consensus to specifically deal with some of the issues associated with the Macdonald criteria. In particular, chemoradiotherapy is well recognised to cause 'pseudo‐progression' in 10% to 30% of all patients (Brandsma 2009; Wen 2010). Nevertheless, we used accepted radiological guidelines to determine response rates in this study. The Macdonald and RANO criteria are listed in Table 1 and Table 2.
| Response | Criteria |
| Complete response | Requires all of the following: complete disappearance of all enhancing measurable and non‐measurable disease sustained for at least 4 weeks; no new lesions; no corticosteroids; and stable or improved clinically |
| Partial response | Requires all of the following: ≥ 50% decrease compared with baseline in the sum of products of perpendicular diameters of all measurable enhancing lesions sustained for at least 4 weeks; no new lesions; stable or reduced corticosteroid dose; and stable or improved clinically |
| Stable disease | Requires all of the following: does not qualify for complete response, partial response or progression; and stable clinically |
| Progression | Defined by any of the following: ≥ 25% increase in sum of the products of perpendicular diameters of enhancing lesions; any new lesion; or clinical deterioration |
| Response | Criteria |
| Complete response | Requires all of the following: complete disappearance of all enhancing measurable and non‐measurable disease sustained for at least 4 weeks; no new lesions; stable or improved non‐enhancing (T2/FLAIR) lesions; and participant must be off corticosteroids or on physiological replacement doses only, and stable or improved clinically. In the absence of a confirming scan 4 weeks later, this response will be considered only stable disease |
| Partial response | Requires all of the following: ≥ 50% decrease, compared with baseline, in the sum of products of perpendicular diameters of all measurable enhancing lesions sustained for at least 4 weeks; no progression of non‐measurable disease; no new lesions; stable or improved non‐enhancing (T2/FLAIR) lesions on same or lower dose of corticosteroids compared with baseline scan; and participant must be taking a corticosteroid dose not greater than the dose at the time of baseline scan and is stable or improved clinically. In the absence of a confirming scan 4 weeks later, this response will be considered only stable disease |
| Stable disease | Stable disease occurs if participant does not qualify for complete response, partial response or progression (see next section) and requires the following: stable non‐enhancing (T2/FLAIR) lesions on same or lower dose of corticosteroids compared with baseline scan and clinically stable status. In the event that corticosteroid dose was increased for new symptoms and signs without confirmation of disease progression on neuroimaging, subsequent follow‐up imaging shows that this increase in corticosteroids was required because of disease progression; the last scan considered to show stable disease will be the scan obtained when the corticosteroid dose was equivalent to the baseline dose |
| Progression | Progression is defined by any of the following: ≥ 25% increase in sum of products of perpendicular diameters of enhancing lesions (compared with baseline if no decrease) on stable or increasing doses of corticosteroids; significant increase in T2/FLAIR non‐enhancing lesions on stable or increasing doses of corticosteroids compared with baseline scan or best response after initiation of therapy, not due to co‐morbid events; appearance of any new lesions; clear progression of non‐measurable lesions; or definite clinical deterioration not attributable to other causes apart from the tumour, or to decrease in corticosteroid dose. Failure to return for evaluation as a result of death or deteriorating condition should also be considered as progression |
-
Quality of life (QoL), when assessed using an objective grading measure such as those described by Mauer (Mauer 2008).
-
Adverse events, classified according to WHO or National Cancer Institute Common Terminology Criteria (NCI‐CTC), including percentage of treatment‐related deaths.
Search methods for identification of studies
No language restriction was applied in all searches.
Electronic searches
Searches were conducted to identify published and unpublished RCTs. Because of the relatively recent availability of targeted antiangiogenic drugs, a literature search starting in 2000 was considered sufficient for the purpose of this review. The following databases were searched.
The Cochrane Central Register of Controlled Trials (CENTRAL),Issue 3, 2014; MEDLINE and EMBASE to April 2014; and databases of ongoing trials:
http://www.controlled‐trials.com, http://www.clinicaltrials.nci.nih.gov, http://www.eortc.be, http://www.update‐software.com/National/nrr‐frame.html, http://www.CenterWatch.com/
MEDLINE, EMBASE and CENTRAL search strategies are listed in Appendix 2, Appendix 3, and Appendix 4.
Searching other resources
Reference lists from trials selected by electronic searching were handsearched to identify further relevant trials. Published abstracts between the years 2000 and 2009 from conference proceedings of the European Society for Medical Oncology (published in Annals of Oncology), the European Council for Clinical Oncology (published in European Journal of Cancer), the Society of Neuro‐oncology (SNO), the European Association of Neuro‐oncology (EANO), the World Federation of Neuro‐oncology (WFNO) and the American Society for Clinical Oncology (ASCO) were handsearched. In addition, members of relevant Cancer Groups (EANO, SNO), experts in the field and manufacturers of relevant drugs were asked to provide details of outstanding clinical trials and of any relevant unpublished material.
Data collection and analysis
Selection of studies
Two review authors (MK and MA) independently assessed the titles and abstracts retrieved by the search strategy for potential eligibility. Full‐text articles of potentially eligible studies (see criteria in section Types of interventions) were obtained when possible. These were assessed for eligibility independently and in a blinded fashion (to study authors, journal, drug company, institutions and results) by two of the review authors (MK and MA). It was agreed that disagreement would be resolved by consensus with a third review author (any of the other review authors).
Abstracts and unpublished data were agreed to be included only if sufficient information on study design and characteristics of participants, interventions and outcomes was available. Further information or final results were to be sought from the primary study author.
Data extraction and management
Data extraction was performed independently by two review authors (MK and MA). Data were entered into RevMan5 for analysis. The following were recorded for each eligible trial: study design, participants, setting, interventions and quality components, duration of follow‐up, efficacy outcomes, biomarker analyses and adverse effects. For studies with more than one publication, data on all outcomes were extracted from the most recent publication. Short‐term adverse events were to be highlighted if considered significant.
Two review authors (MK and MA) independently extracted details of study population, interventions and outcomes by using a standardised data extraction form. Blinding of study participants and investigators was also assessed. Differences in data extraction were resolved by consensus with a third review author (any two review authors), referring back to the original article. No disagreement among review authors regarding selection of studies was reported.
The data extraction form included the following items.
-
General information: title, study authors, source, contact address, country, published/unpublished, language and year of publication, sponsor of trial.
-
Trial characteristics, including study design, duration/follow‐up and quality assessment as specified above.
-
Participants: inclusion and exclusion criteria, sample size, baseline characteristics, similarity of groups at baseline, withdrawals and losses to follow‐up.
-
Interventions: dose, route and timing of chemotherapy, antiangiogenic therapy and comparison intervention.
-
Outcomes.
-
For time to event (survival and disease progression), we extracted hazard ratios (HRs) and 95% confidence intervals (CIs), log rank Chi2, log rank P values, numbers of events, numbers of participants per group and median and one‐, two‐, three‐ and five‐year survival rates.
-
For dichotomous outcomes (radiological response and adverse events), we extracted the number of participants in each group who experienced the outcome of interest and the number of participants assessed at endpoint to estimate the risk ratio (RR).
-
For continuous outcomes (quality of life (QoL) measures), we extracted the final or change value and standard deviation of the outcome of interest and the number of participants analysed at endpoint for each treatment group to estimate differences in means and 95% CIs between treatment arms.
-
When possible, data for Intention‐to‐treat analysis were extracted for all outcomes.
HRs and their 95% CIs were estimated directly or indirectly from the published data (Altman 2001). HRs can be estimated (under some assumptions) from log‐rank Chi2, log‐rank P values, observed to expected event ratios and ratios of median survival times or time point survival rates (Machin 1997; Parmar 1998).
The time points at which outcomes were collected and reported were recorded.
Assessment of risk of bias in included studies
All studies that met the inclusion criteria were assessed independently for quality by two review authors (MK and MA), and disagreements were resolved by a third review author if required. The risk of bias for each included study was assessed using the Cochrane risk of bias tool (Higgins 2008).
Each study was assessed independently by two review authors for the use of random allocation to comparison groups. Trials that permitted cross‐over for participants after disease progression were also included.
Measures of treatment effect
Data analysis
Summary statistics for the primary endpoints (time to event data) are presented as HRs (Cox 1972).
-
For dichotomous outcomes, we estimated RRs.
-
For continuous outcomes, we estimated differences in means between treatment arms, and if the outcome had not been evaluated by using the same measurement scale across studies, standardised mean differences (SMDs) were used.
Unit of analysis issues
Measurement of PFS was based on different response criteria. Nevertheless, use of the hazard ratio for comparisons minimises the significance of this issue, as the control group is subject to the same response criteria. Nevertheless, the different response criteria must be considered when the pooled analysis of PFS data is interpreted.
Dealing with missing data
The first author of the most recent publication was contacted in cases of missing data. Specifically, the first and senior authors of the BELOB study (randomised phase II study of bevacizumab versus bevacizumab plus lomustine versus lomustine in patients with recurrent GBM) were contacted to try to obtain a hazard ratio for PFS, but this information was not available.
Assessment of heterogeneity
Heterogeneity between studies was assessed by the Cochrane Q‐test, with a significance threshold of alpha = 0.1. and by estimation of the percentage of heterogeneity between trials that cannot be ascribed to sampling variation (Higgins 2003).
In cases of substantial heterogeneity, the extra variation was incorporated into the analysis by using a random‐effects model.
The following factors were considered as possible sources of heterogeneity.
-
Differing clinical settings (adjuvant versus recurrent disease).
-
Different applications of antiangiogenic treatment (scheduling, etc.).
-
Different types of angiogenesis inhibitors (as classified above).
-
Differences in prognostic factors between studies.
-
Study quality.
These factors were considered in the sensitivity and subgroup analyses, except in cases of differing prognostic factors, as limited data were available for this analysis (apart from analysis by setting, that is, adjuvant versus recurrent).
Assessment of reporting biases
.
Data synthesis
Results were pooled in a meta‐analysis.
-
For time‐to‐event data, HRs were pooled using the generic inverse variance facility of RevMan 5.
In trials with multiple treatment groups, the antiangiogenic intervention was compared with the specific control comparator. Specifically, when multiple interventions were provided, the shared control group was split across the two interventions. The review authors believed that this approach was more advantageous than combining all experimental groups, as it allowed investigations of heterogeneity across intervention arms (Cochrane Handbook for Systematic Reviews of Interventions, Section 16.5.4).
Fixed‐effect models were used for all meta‐analyses, and random‐effects models with inverse variance weighting were used for all meta‐analyses in cases of substantial heterogeneity (DerSimonian 1968).
Subgroup analysis and investigation of heterogeneity
At the time of this review, data were insufficient to permit analysis of preplanned subgroups specifically, with the exception of antiangiogenic therapy alone versus antiangiogenic treatment in combination with chemotherapy.
Sensitivity analysis
A sensitivity analysis was performed after the main meta‐analysis was completed for both primary and secondary outcomes. Several subgroups were analysed to deal with the sources of heterogeneity listed above. Specifically, a separate analysis for bevacizumab‐containing antiangiogenic therapy was conducted apart from the main analysis to account for the different types of antiangiogenic therapy used across studies, with bevacizumab appearing most frequently. The GLARIUS study (bevacizumab, irinotecan, and radiotherapy versus standard temozolomide and radiotherapy in newly diagnosed, MGMT non‐methylated glioblastoma patients) was excluded as part of the sensitivity analysis because of its high risk of bias, the outlying nature of study results and the fact that a differing chemotherapeutic backbone was noted across the two arms; however this did not change results of the analysis significantly.
Review updates
Four of the seven studies analysed in this review were available only in abstract form and details were restricted to what was presented at scientific conferences and additional details from clinical trials.gov. Updates of this review will be undertaken as soon as all eligible studies are published in full in the peer‐reviewed literature. In future reviews, we will consider the mechanism of action and whether the drug is an antibody or a small molecule and a direct, indirect or mixed inhibitor for classification, as well as different types of chemotherapy, schedules of antiangiogenic agents and combinations of different antiangiogenic strategies. Studies using antiangiogenic therapies as second‐line treatment will be classified according to their prior use in first‐line treatment as well. This classification will be adopted depending on the availability of relevant studies in the future. As soon as results of studies using combinations of antiangiogenic and other targeted drugs become available, further comparisons will be performed.
Results
Description of studies
Results of the search
A broad search yielded 133 references through MEDLINE, 624 references from EMBASE and 252 references from CENTRAL. Conference searches yielded an additional 10 references. Duplicates were deleted. Abstracts were reviewed independently by two review authors (MK and MA), and articles that obviously did not meet the inclusion criteria were excluded at this stage.
Included studies
Seven eligible RCTs were identified, four of which were reported in conference abstracts and three in published journal articles. The seven eligible RCTs identified included 2987 participants in total for this analysis.
Randomised trials in newly diagnosed GBM
The Avastin in Glioblastoma (AVAGLIO) study (Chinot 2014)
This phase III, double‐blind, placebo‐controlled trial evaluated bevacizumab in participants with newly diagnosed glioblastoma.
In this study, four to seven weeks after surgical resection of glioblastoma, 921 participants were randomly assigned to standard temozolomide (TMZ: 75 mg/m2/d) and radiotherapy (2 Gy five days a week; maximum 60 Gy)
with bevacizumab (10 mg/kg i.v. every two weeks) or placebo. Four weeks after completion of radiotherapy, participants commenced six cycles of standard maintenance temozolomide (150 to 200 mg/m2 p.o. days one through five, every 28 days). Bevacizumab or placebo was continued every two weeks during the maintenance phase and then every three weeks at a dose of 15 mg/kg until disease progression or toxicity. OS and PFS (investigator assessed) were co‐primary endpoints. Secondary endpoints included PFS assessed by independent central review, one‐year survival rates, health‐related quality of life (European Organisation Quality of Life Questionnaire (EORTC QLQ‐C30) and Brain Tumour Module (BN20) scales) and safety. Exploratory endpoints included performance status and use of corticosteroids.
One of the secondary endpoints of AVAGLIO required comparison of QoL between treatment arms, as measured by the EORTC QLQ‐C30 scale and its companion brain tumour module BN20; 78% to 91% of evaluable participants without disease progression completed each assessment during the first year. Baseline scores were comparable between arms. Although no difference between treatment arms was noted in health‐related quality of life (HRQoL) score changes over time, the addition of bevacizumab did not worsen or improve participant response over time compared with placebo. However, bevacizumab‐treated participants experienced longer time to HRQoL deterioration compared with those in the control arm (Taphoorn 2013).
In AVAGLIO, radiological progression was defined as a 25% or greater increase in the size of enhancing lesions, with unequivocal progression of existing non‐enhancing lesions or any new lesions. Clinical progression was defined as worsening neurological symptoms. Stratification included age, performance status, MGMT status, extent of surgical resection, Mini Mental State Examination (MMSE) and corticosteroid and antiepileptic use.
The Radiation Therapy Oncology Group (RTOG) 0825 study (Gilbert 2014)
This phase III, double‐blind, placebo‐controlled trial evaluated bevacizumab in participants with newly diagnosed glioblastoma. Researchers enrolled 621 patients with newly diagnosed glioblastoma stratified on the basis of MGMT methylation status and molecular profile. Participants then received three weeks of radiation therapy and daily TMZ plus bevacizumab, or a continuation of their standard therapy plus placebo. Primary objectives were OS and PFS. Secondary endpoints included toxicity, symptom burden, HRQoL, neurocognitive function and identification of participant subsets more likely to benefit from bevacizumab.
As part of this study, net clinical benefit was measured using longitudinal measures of patient‐reported outcomes (PROs), including the MD Anderson Symptom Inventory‐Brain Tumor Module (MDASI‐BT) and the EORTC QLQ‐C30/BN20. These were completed at baseline and longitudinally (weeks 6, 10, 22, 34 and 46) .
Participants also completed neurocognitive testing with the Hopkins Verbal Learning Test‐Revised (HVLT‐R), the Trail Making Test (TMT) and Controlled Oral Word Association (COWA) at baseline and longitudinally (week 6, 10, 22, 34 and 46). Six neurocognitive test scores, as well as the Clinical Trial Composite (CT COMP) score (i.e. average performance across all neurocognitive tests), were examined over time.
The Cilengitide, Temozolomide, and Radiation Therapy in Treating Patients With Newly Diagnosed Glioblastoma and Methylated Gene Promoter Status (CENTRIC) study (Stupp ASCO 2013)
The CENTRIC study combined cilengitide with standard treatment for participants with newly diagnosed glioblastoma and methylated MGMT gene promoter (CENTRIC). It was an international phase III trial that enrolled 545 participants with GBM harbouring MGMT promoter methylation, while adding cilengitide (2000 mg twice weekly i.v.) to standard RT/TMZ compared with RT/TMZ without cilengitide. The primary endpoint was OS, and secondary endpoints included PFS, safety and tolerability, pharmacokinetics, QoL and general health and work status.
Bevacizumab, irinotecan, and radiotherapy versus standard temozolomide and radiotherapy in newly diagnosed, MGMT‐nonmethylated glioblastoma patients: First results from the randomized multicenter GLARIUS trial. (Glarius 2013) Herrlinger ASCO 2013 )
This was a randomised phase II study in participants with MGMT non‐methylated newly diagnosed glioblastoma. A total of 170 participants were randomly assigned in a 2:1 ratio to bevacizumab and irinotecan (Bev/Iri) or temozolamide (TMZ). All participants received radiation therapy at a standard dose for six weeks. Participants in the Bev/Iri arm received four cycles of bevacizumab over the course of radiation and then received Bev/Iri every two weeks until disease progression. The primary endpoint of the study was six‐month PFS. Secondary endpoints included overall survival, steroid use and toxicity. Cross‐over to bevacizumab and irinotecan following progression on temozolomide was permitted. Of 54 participants in the temozolomide arm, 29 crossed over to bevacizumab and irinotecan.
Cilengitide, Temozolomide, and Radiation Therapy in Treating Patients With Newly Diagnosed Glioblastoma and Unmethylated Gene Promoter Status (The CORE study) (CORE 2013)
This was a randomised, three‐arm, open‐label, phase II study of cilengitide in MGMT non‐methylated newly diagnosed GBM. The first intervention arm of the study included cilengitide at standard dosing (2000 mg i.v. twice weekly) through chemoradiation (six weeks) and as maintenance until week 34. The second intervention arm of the study provided cilengitide at intensive dosing (2000 mg i.v. five times weekly) throughout chemoradiation, then 2000 mg i.v. twice weekly as maintenance until week 34. The control arm and both intervention arms included standard chemoradiotherapy with temozolomide 75 mg/m2 followed by six cycles of maintenance temozolomide at 150 to 200 mg/m2 on days one through five, every 28 days.
Randomised trials in recurrent GBM
Cediranib in Combination With Lomustine Chemotherapy in Recurrent Glioblastoma (REGAL) (Batchelor 2013)
This was a phase III, randomised, multi‐centre study comparing cediranib as monotherapy and in combination with lomustine versus lomustine alone. A total of 325 participants with disease progression after receiving only one prior systemic temozolomide‐containing chemotherapy regimen were randomly assigned in a 2:2:1 ratio to one of three arms: cediranib monotherapy 30 mg orally daily; cediranib 20 mg orally daily in combination with oral lomustine 110 mg/m2 once every six weeks; or lomustine 110 mg/m2 once every six weeks in combination with placebo. The primary endpoint was assessment of the relative efficacy of cediranib—given alone or in combination with lomustine—versus lomustine alone by independent central radiographic assessment of PFS. The REGAL study is the only randomised controlled phase III study that has investigated recurrent GBM.
A randomized phase II study of bevacizumab versus bevacizumab plus lomustine versus lomustine single agent in recurrent glioblastoma: (The Dutch BELOB study). (Belob 2013)
This was a randomised three‐arm phase II study of 140 participants with GBM at first recurrence treated with (1) bevacizumab 10 mg/kg every two weeks versus (2) bevacizumab every two weeks plus lomustine 110/m2 every six weeks for six cycles capped at 200 mg or (3) single‐agent lomustine. The primary endpoint was OS at nine months. Secondary endpoints included median PFS, six‐month PFS, QoL, deterioration‐free survival and safety in the form of safety monitoring in the first 10 participants. For the purposes of this analysis, the single‐agent lomustine arm was considered the control group, and the two other groups were the intervention groups.
In total, seven studies with 2987 participants reported data for overall survival. All studies except BELOB reported overall survival in the form of hazard ratios and were eligible to be pooled. In one study (BELOB), the hazard ratio was estimated using the Parmar method (Parmar 1998; Cochrane Handbook for Systematic Reviews of Interventions, Section 7.7.6) from a survival curve provided in the conference proceedings. Although BELOB was a three‐arm study of lomustine (control), lomustine and bevacizumab and bevacizumab monotherapy, hazard ratios were calculated for datasets of bevacizumab‐containing regimens compared with lomustine to allow inclusion in this analysis. Although the GLARIUS study used different chemotherapy arms (irinotecan vs temozolomide), the study authors believed that pooling of data from this study was appropriate, as both groups had a chemotherapy backbone, albeit different. Limitations of the data from the GLARIUS study were accounted for in the sensitivity analysis.
Five of the seven studies compared the addition of antiangiogenic therapy in the first‐line or adjuvant setting, whereas the other two studies were conducted in the relapse setting. Additional variability was noted in the interventions provided across studies. Four of the seven studies utilised bevacizumab as antiangiogenic treatment, with two studies of cilengitide and one of cediranib. Among the adjuvant studies, GLARIUS used a different cytotoxic chemotherapy (irinotecan) of uncertain efficacy in gliomas compared with all of the other studies, which used standard chemoradiation with temozolomide. Both of the relapse studies used lomustine as the comparator in three‐arm studies, although two arms did not provide any chemotherapy at all.
Excluded studies
All non‐randomised, single‐arm studies and studies that provided an antiangiogenic agent but no control arm or a historical control arm were excluded. Of note, Bevacizumab Alone and in Combination With Irinotecan in Recurrent Glioblastoma (the BRAIN study) (Friedman 2009) and the National Cancer Institute study (Kreisl 2009), which was used to support Food and Drug Administration approval of bevacziumab in the USA for recurrent GBM, were not eligible for inclusion in this analysis because neither study had a control arm that did not provide antiangiogenic therapy.
Risk of bias in included studies
As four of the studies were published only in the form of conference proceedings, information was inadequate for review authors to fully evaluate risk of bias; these studies are thus awaiting classification. Information from the REGAL, AVAGLIO and RTOG 0825 studies was available in full and was of high quality with low risk of bias (all risk of bias items were adequate). Available information for the included studies awaiting further classification is listed below. It is planned for this review to be updated once details of these studies become available in full publication format.
All of the included studies were RCTs. Information regarding the method of randomisation was available for three studies—REGAL, AVAGLIO and RTOG 0825—all of which used computer‐generated randomisation sequences.
Allocation
For three of the studies, information regarding allocation concealment was available. For the REGAL and AVAGLIO studies, unique tracking IDs were assigned and managed through an interactive web response system (IWRS). RTOG 0825 allocation and randomisation were managed through the centralised cancer trials support unit (CTSU).
Blinding
For three of the studies, complete information regarding blinding was available. For the REGAL study, participants and investigators were blinded. Placebo and cediranib were identical as packaged, and participants could be unblinded only in the event of a medical emergency. For the RTOG 0825 study, both participants and investigators were blinded. Investigators were blinded to the MGMT status of the participant. Salvage bevacizumab was available on progression, resulting in unblinding once a code‐breaking form was submitted. For AVAGLIO, both participants and investigators were blinded. Unblinding was permitted on progression only when subsequent therapy was determined. In the intervention group, 31% of participants continued bevacizumab beyond progression, and 48% crossed‐over to bevacizumab in the control group. The CENTRIC and CORE studies were open‐label. All of the studies underwent centralised radiological review, except the BELOB study, for which the method of assessment of response is unknown from the current data.
Incomplete outcome data
For most of the studies, outcome data capture was near complete, thereby minimising the risk of attrition bias. In the AVAGLIO study, among 921 participants, 26 dropped out before the study commenced, with a further 23 in the second phase of the study, and a further 28 in the third phase. The intervention and control groups were well matched in terms of dropout rates. In the CENTRIC study, only one participant was lost to follow‐up, and the rate at which participants did not receive the assigned intervention was low in both groups (nine and 15, respectively). In the RTOG 0825 study, 16 participants dropped out, eight in each arm. In the REGAL study, 10 participants were lost to follow‐up and 12 participants did not receive the allocated intervention (five, six and one participant, respectively). The BELOB study had a similarly low dropout rate, with only three participants dropping out across the entire study. In the CORE study, only two participants were lost to follow‐up, and most participants received the allocated intervention. The dropout rate was not available for the GLARIUS study. Moreover, all studies used intention‐to‐treat analysis, further limiting the risk of attrition bias.
Selective reporting
All trials were registered at clinicaltrials.gov or at Eudra‐CT, and investigators reported preplanned outcomes, thereby limiting the risk of selective reporting. The AVAGLIO study did include an unplanned analysis of "deterioration‐free survival" as a surrogate for the QoL data; this was not prespecified in the trial protocol, although QoL was a planned endpoint .
Other potential sources of bias
All studies used intention‐to‐treat analysis, and when information was available, all had similar schedules for follow‐up between control and experimental arms. The RTOG 0825 study described additional scans (such as dynamic magnetic resonance imaging (MRI)) that were performed periodically and did not relate to the study's outcome assessments. The REGAL (cediranib) study was published in Journal of Clinical Oncology (2013). The AVAGLIO study and RTOG 0825 were published recently in New England Journal of Medicine (2014). The GLARIUS, CENTRIC and BELOB studies were reported at the American Society for Clinical Oncology Meeting 2013. The CORE study was presented recently (CORE 2013) in abstract form at meetings of the 2013 European Society of Medical Oncology and later the Society of Neuro‐oncology.
Effects of interventions
Overall survival
Meta‐analysis of all seven trials of antiangiogenic therapy (n = 2987 participants) found no observed differences in OS, with a fixed‐effect pooled HR of 0.94 (95% CI 0.86 to 1.02; P value 0.16) (Figure 1). Significant heterogeneity was observed (I2 = 0.61) and a random‐effects meta‐analysis confirms this (HR 0.93, 95% CI 0.80 to 1.08; P value 0.32; I2 = 0.61). This heterogeneity most likely is due to differences between the studies in terms of participant population and clinical setting as well as interventions applied.
In terms of participant population, although some minor differences in molecular profile were noted (for instance, use of the better prognosis MGMT‐methylated participants in the cilengitide study compared with the worse prognosis MGMT‐unmethylated participants in GLARIUS and CORE), the major participant‐related difference was the treatment setting (see Included studies above).
Pooled data overall survival—antiangiogenic therapy versus no antiangiogenic therapy
In a meta‐analysis of two studies (n = 292 participants) of antiangiogenic therapy without chemotherapy, the HR for OS was 1.26 (95% CI 0.96 to 1.65; P value 0.10). No heterogeneity was observed (I2 = 0).
Pooled data overall survival—antiangiogenic therapy with chemotherapy versus chemotherapy
Meta‐analysis of seven studies of antiangiogenic therapy with chemotherapy (n = 2806 participants) showed significant improvement in OS (Figure 2), with an HR of 0.91 (95% CI 0.83 to 1.00; P value 0.04). However, given that significant heterogeneity was observed (I2 = 60%), when a random‐effects model was used, the result was not significant (HR 0.87, 95% CI 0.75 to 1.02; P value 0.10), largely as a result of the large treatment effect in the BELOB study seen in the treatment arm of bevacizumab with lomustine, and in the GLARIUS study of bevacizumab with irinotecan. If these two studies are excluded (n = 2635 participants), the corresponding HR increases to 0.95 (95% CI 0.86 to 1.04), but the heterogeneity becomes much less (I2 = 45%).
Forest plot of meta‐analysis of overall survival, antiangiogenic therapy plus chemotherapy (Analysis 1.3).
Pooled data overall survival—bevacizumab versus no bevacizumab
Although not a prespecified subgroup analysis, given the larger number of studies with bevacizumab, a separate analysis of all bevacizumab‐containing regimens was performed (see Differences between protocol and review). Meta‐analysis of four studies of bevacizumab (n = 1852 participants) revealed no observed difference in OS, with a fixed‐effect HR of 0.92 (95% CI 0.83 to 1.02; P value 0.12). Heterogeneity was significant (I2 = 71%), and a random‐effects meta‐analysis confirmed this result (HR 0.87, 95% CI 0.69 to 1.09; P value 0.23). In this instance, the heterogeneity is more difficult to explain, particularly as the intervention was the same across groups. One study (BELOB) was conducted in the relapse setting, but exclusion of these results from the analysis does not significantly alter the heterogeneity.
Although we initially intended to break down the data by treatment setting and by treatment, given the small number of studies, this was not appropriate at this stage (Figure 3).
Forest plot of meta‐analysis of overall survival, bevacizumab (Analysis 1.4)
Pooled data overall survival—adjuvant setting
Meta‐analysis of five studies of antiangiogenic therapy in the adjuvant setting (n = 2522 participants) showed no observed difference in OS (Figure 1), with a fixed‐effect HR of 0.92 (95% CI 0.84 to 1.01; P value 0.10). Heterogeneity was significant (I2 = 56%); therefore a random‐effects meta‐analysis was performed, which demonstrated similar results (HR 0.90, 95% CI 0.76 to 1.05; P value 0.17). This was likely to be driven by the large treatment effect seen in the intervention arm of the GLARIUS study, part of which may be due to the different cytotoxic chemotherapy backbone (irinotecan). If this study were excluded (n = 2352), heterogeneity would be reduced (I2 = 51%) and the HR would be largely unchanged at 0.94 (95% CI 0.85 to 1.03; P value 0.20).
Pooled data overall survival—recurrent (relapse) setting
In a meta‐analysis of two studies (n = 465) of antiangiogenic therapy in the recurrent setting, the fixed‐effect HR for OS was 1.02 (95% CI 0.84 to 1.24; P value 0.86). However, statistical heterogeneity was significant (I2= 0.73), and a random‐effects meta‐analysis confirms this (OS HR 1.02, 95% CI 0.70 to 1.48; P value 0.93; I2 = 0.73). Heterogeneity appears to be due to contrasting results in OS observed in the BELOB study (using the lomustine control group (n = 46) when bevacizumab was used alone (n = 50); fixed‐effect OS HR 1.12, 95% CI 0.77 to 1.63), compared with bevacizumab and lomustine (n = 44; fixed‐effect OS HR 0.58, 95% CI 0.39 to 0.86). When the combination arm of the BELOB study is omitted, the fixed‐effect pooled OS HR is 1.22 (favouring no antiangiogenic therapy) (95% CI 0.97 to 1.53; P value 0.08; heterogeneity I2 = 0).
Progression‐free survival
Meta‐analysis of six studies (n = 2847 participants) yielded data for PFS. The BELOB study did not report PFS, and we were unable to obtain these data from the study authors. Meta‐analysis of all trials of antiangiogenic therapy using a fixed‐effect model found an observed difference in PFS (HR 0.74, 95% CI 0.68 to 0.81; P value < 0.00001) (Figure 4). Heterogeneity was significant(I2 = 75%), and a random‐effects meta‐analysis confirms this (HR 0.73, 95% CI 0.61 to 0.88; P value 0.0008). This heterogeneity is most likely due to differences between the studies in terms of participant population and clinical setting, as well as to the intervention applied (as described in the overall survival section above). No meta‐analysis was done for studies of antiangiogenic therapies without chemotherapy or of antiangiogenic therapies in the relapse setting, as PFS data were available for only one study in both of these instances.
Forest plot of meta‐analysis of progression‐free survival, categorized by setting (Analysis 2.1)
Pooled data PFS: antiangiogenic therapy with chemotherapy versus chemotherapy
Meta‐analysis of six studies of antiangiogenic therapy with chemotherapy (n = 2716 participants) revealed an observed difference in PFS (HR 0.72, 95% CI 0.66 to 0.79; P value < 0.00001). Heterogeneity was significant (I2 = 75%). Random‐effects model meta‐analysis confirmed similar results (HR 0.70, 95% CI 0.58 to 0.84; P value 0.0002). Heterogeneity was reduced significantly if the GLARIUS study was excluded because of the magnitude of its effect (I2 = 43%). In a sensitivity analysis without the GLARIUS study (n = 2546), the HR for PFS remained significant at 0.75 (95% CI 0.69 to 0.82; P value < 0.0001) (Figure 5).
Forest plot of meta‐analysis of progression‐free survival, antiangiogenic therapy plus chemotherapy (Analysis 2.2).
Pooled data PFS: bevacizumab versus no bevacizumab
Meta‐analysis of three adjuvant studies of bevacizumab (n = 1712 participants) revealed significant observed differences in PFS (HR 0.66, 95% CI 0.59 to 0.74; P value < 0.00001). Significant heterogeneity was reported (I2 = 87%) in these studies, as for the greater pooled analysis of PFS. Therefore, a random‐effects meta‐analysis was performed with similar results (HR 0.57, 95% CI 0.54 to 0.82; P value 0.002). This may be attributable to the results of the GLARIUS study, which, unlike the Gilbert and CENTRIC studies, was undertaken in an MGMT unmethylated GBM population—a prognostically worse group that paradoxically did better here with the addition of irinotecan and bevacizumab to standard radiotherapy as opposed to TMZ (Figure 6).
Pooled data PFS: adjuvant setting
Meta‐analysis of five studies for PFS in the adjuvant setting (n = 2522 participants) showed a difference in observed PFS (HR 0.72, 95% CI 0.66 to 0.79; P value < 0.00001). Even when the outlier GLARIUS results were excluded, the pooled analysis for PFS remained significant at 0.75 (95% CI 0.68 to 0.82; P value < 0.00001). Heterogeneity was significant, particularly when the GLARIUS results were included (I2 = 79%). This was reduced when the GLARIUS results were omitted (I2 = 55%). Findings of a random‐effects meta‐analysis excluding the GLARIUS results were consistent with results of the fixed‐effect analysis (HR 0.76, 95% CI 0.66 to 0.88; P value 0.0002).
Adverse events
Some differences were observed in the toxicity profiles associated with different regimens, particularly dependent upon whether antiangiogenic therapy was combined with cytotoxic therapy. Toxicity data from the three published studies are presented in Table 3. Studies of bevacizumab reported similar rates of common and serious toxicities, including wound complications (1.6% to 3.3%), hypertension (4.2% to 11.3%), thromboembolic complications (4.6% to 7.7%) and gastrointestinal perforation (0.8% to 1.1%). Overall, the other antiangiogenic therapies were well tolerated, although in the REGAL study, an excess of haematological toxicity was seen in the cediranib group combined with the lomustine groups (38.3% thrombocytopaenia). Toxicity data were not formally meta‐analysed because of the paucity of data on toxicity reported across studies.
| Grade ≥ 3 adverse event | REGAL (%, c = cediranib, l = cediranib + lomustine) | AVAGLIO (%) | RTOG 0825 (%, crt = chemoradiotherapy, a = adjuvant chemotherapy) |
| Haemorrhage | c: 0 l: 0.8 | 3.3 | crt: 0 a: 1.6 |
| Wound‐healing complications | NR | 3.3 | crt: 1 a: 1.6 |
| Thromboembolic events (arterial ATE, venous VTE) | c: 3.1 l: 4.9 | ATE: 5 VTE: 7.6 | crt: 4.6 a: 7.7 |
| Hypertension | c: 14.1 l: 6.5 | 11.3 | crt: 1.3 a: 4.2 |
| Proteinuria | NR | 5.4 | NR |
| GI perforation | NR | 1.1 | crt: 0 a: 0.8 |
| Thrombocytopaenia | c: 1.6 l: 38.3 | 15 | crt: 10.2 a: 11.1 |
| Fatigue | c: 16,4 l: 15.4 | 7.4 | crt: 0 a: 13.1 |
| Diarrhoea | c: 6.3 l: 5.7 | NR | NR |
Quality of life
Published data are inadequate to allow formal assessment and pooling of QoL endpoints. However, the two adjuvant studies (Chinot 2014; Gilbert 2014) that have been published in full differ in the reported impact of antiangiogenic therapy on patient QoL. The secondary endpoint of the AVAGLIO study protocol was EORTC QLQ‐C30, and in the publication, significant improvement in global health status and in Karnofsky performance score was noted. All participants were required to fill in the QoL questionnaires. The paper reports a measure of deterioration‐free survival that includes survival. Deterioration was classified as a 10‐point deterioration from baseline seen as sustained damaged or death. This was not included in the trial protocol.
The RTOG 0825 study showed that over time, an increased symptom burden, a worse QoL and a decline in neurocognitive function were more frequent in the bevacizumab group. This has been reported only in abstract form (Armstrong 2013; Wefel 2013). The trial had an optional net clinical benefits study, in which 80% of trial participants initially took part. It is important to note that participants were not required to fill in forms upon progression, and as a result of consistent improvement in progression‐free survival in the bevacizumab arm, more participants completed the various questionnaires throughout the study in the intervention arm. For example, during week 34, 107 participants in the bevacizumab arm completed the forms compared with 72 in the control group. Consequently, significant deterioration in quality of life across many domains was observed for participants given the bevacizumab‐containing regimen.
In summary, the impact of antiangiogenic therapy, including bevacizumab, on QoL remains unclear.
Sensitivity analysis
As overall quality was difficult to thoroughly assess in six of the seven studies, which were obtained in the form of conference proceedings only, a sensitivity analysis was not performed to assess study quality. We therefore confined our sensitivity analysis to differing classes of agents and treatment regimens, as previously described (Analysis 1.2: Analysis 1.3; Analysis 1.4; Analysis 2.2; Analysis 2.3). Data were insufficient for a pooled analysis on the basis of MGMT status or molecular profile at this stage. In general, the subgroup analysis showed that OS and PFS largely remain the same when studies with differing interventions and settings are considered. Moreover, the findings were unchanged when the very large treatment effect of the GLARIUS study was omitted; this study incorporated a different cytotoxic chemotherapy (irinotecan) for the control group.
Discussion
Summary of main results
The trials included in this systematic review did not show improvement in overall survival with antiangiogenic therapy. However, they did demonstrate significant overall improvement in PFS. These divergent results reflect uncertainty regarding the efficacy of antiangiogenic therapy in high‐grade glioma and further question the correlation between radiological response assessment to determine PFS versus OS in HGG.
Assessment of PFS was based largely on MRI assessments. Radiological progression may or may not represent true biological progression of the malignancy. Some patients with treated GBM who have an MRI abnormality may indeed have the so called pseudoprogression that may reflect tumour necrosis rather than disease progression. Alternatively, imaging may show pseudoresponse that might indicate improvement in appearance of the tumour area on MRI as a result of the antiangiogenic effect, making tumour blood vessels less leaky and reducing contrast enhancement on MRI despite ongoing tumour progression. Using the newly developed RANO criteria (Wen 2010) will hopefully minimise the impact of this issue in future clinical trials.
It should be noted that some of the agents evaluated in this review are direct angiogenic pathway inhibitors that are more potent than antiangiogenic therapy compared with agents indirectly inhibiting blood vessel formation. This may explain the greater observed effects of bevacizumab and cediranib, which are direct inhibitors of VEGF and VEGFR signalling, respectively, compared with cilengitide, which is thought to indirectly mediate antiangiogenic effects through inhibition of tumour and endothelial cell adhesion and migration. These differences may account for the PFS prolongation observed with bevacizumab that was not seen with cilengitide. It is noteworthy that the hazard ratio of the largest cilengitide study (CENTRIC) was 1.0.
Most of the randomised studies focused on newly diagnosed GBM. In recurrent HGG, the two randomised studies (evaluating cediranib and bevacizumab, respectively) demonstrated no benefit. With the exception of the relatively small BELOB study, with 50 participants per study arm, no adequately powered, randomised, placebo‐controlled studies for recurrent HGG (or GBM) have been conducted with bevacizumab, and the question of survival benefit of bevacizumab in recurrent HGG thus remains unanswered.
Hazard ratios for the pooled meta‐analyses of combination therapy studies containing bevacizumab as the antiangiogenic therapy were more likely to report favourable results than those for other antiangiogenic agents alone; this was also reflected in the lower hazard ratio in the pooled analysis of bevacizumab studies compared with the overall pooled analyses.
Subanalyses of the pooled studies raised questions regarding the direction of future research. Although not significant, inferior outcomes have been suggested in the group of participants who received antiangiogenic therapy alone (shown in Figure 1 for cediranib alone and for bevacizumab alone compared with the combination of these agents with lomustine). The BELOB study was associated with inferior efficacy results, observed when bevacizumab was administered alone compared with when it was given with chemotherapy (lomustine). A trend toward greater benefit was seen when antiangiogenic therapy was given with chemotherapy as compared with the efficacy seen with single‐agent bevacizumab. This was observed in the bevacizumab studies BELOB and GLARIUS (Figure 2). This finding is consistent with data on use of bevacizumab in other malignancies such as colorectal cancer or lung cancer, for which bevacizumab is active only when combined with chemotherapy. This has to be interpreted with caution because of the relatively small size of the BELOB and GLARIUS studies.
These results suggest that combination therapy should be used instead of single‐agent antiangiogenic therapies in future trials to achieve synergy between antiangiogenic and chemotherapeutic agents .
The quality of life data from the two largest studies with bevacizumab in HGG that have been presented are not consistent. The AVAGLIO results suggest that quality of life is improved in patients receiving bevacizumab compared with placebo, but investigators used a measure of deterioration‐free survival that was not prespecified in the trial protocol. However, the RTOG 0825 study suggests the opposite—that quality of life of patients with HGG taking bevacizumab may be worse than that of patients who have not received bevacizumab. The RTOG 0825 study provided optional participation in the net clinical benefits component of the study and did not require completion of the form on progression. As such, a larger number of participants receiving the bevacizumab‐containing regimen completed the questionnaires, and this may account for the divergence in results.
Gliomas and HGGs are an especially important group of tumours because of their disproportionate impact on patient well‐being and QoL, in addition to longevity. However, the impact of antiangiogenic therapy, including that of bevacizumab, on QoL remains unclear.
Overall completeness and applicability of evidence
The analysed studies were heterogeneous in terms of interventions applied, clinical settings and study designs. This review included different pharmacological interventions that directly or indirectly target angiogenesis. Bevacizumab, cilengitide and cediranib are different agents in terms of their mechanisms of action, their antiangiogenic effects and their pharmacodynamic and pharmacokinetic effects. The studies also differed in terms of clinical settings (some were first‐line and others were recurrent), but all were restricted to glioblastoma, and participants had similar baseline demographic characteristics across studies. Nevertheless, the sensitivity analysis accounted for this, and the overall survival results were largely unchanged. Furthermore, as the review was performed by relying on abstracts submitted at five conferences and only one journal article, information is missing, particularly with regard to details of study design and conduct and toxicity data.
Quality of the evidence
We have reserved an overall assessment of the quality of the evidence until the final study publications in journal form are received. The three publications that have been published in full are of good quality. Furthermore, the fact that all studies were randomised, had relatively balanced baseline groups and low rates of dropout and used intention‐to‐treat analysis, and that all but two were double‐blinded, suggests overall high quality of the studies included in this review.
Potential biases in the review process
No prespecified subset analysis examined trials of antiangiogenic therapy alone or antiangiogenic therapy with chemotherapy and bevacizumab alone. However, at the time of writing of the review, it is noted that the differing classes of agents would form the basis of the sensitivity analysis.
Agreements and disagreements with other studies or reviews
Not applicable.
Forest plot of meta‐analysis of overall survival, categorised by setting (Analysis 1.1).
Forest plot of meta‐analysis of overall survival, antiangiogenic therapy plus chemotherapy (Analysis 1.3).
Forest plot of meta‐analysis of overall survival, bevacizumab (Analysis 1.4)
Forest plot of meta‐analysis of progression‐free survival, categorized by setting (Analysis 2.1)
Forest plot of meta‐analysis of progression‐free survival, antiangiogenic therapy plus chemotherapy (Analysis 2.2).
Forest plot of meta‐analysis of progression‐free survival, bevacizumab (Analysis 2.3).
Comparison 1 Hazard for OS, Outcome 1 Hazard for OS.
Comparison 1 Hazard for OS, Outcome 2 Hazard for OS antiangiogenic therapy alone.
Comparison 1 Hazard for OS, Outcome 3 Hazard for OS antiangiogenic plus chemotherapy.
Comparison 1 Hazard for OS, Outcome 4 Hazard for OS bevacizumab.
Comparison 2 Hazard for PFS, Outcome 1 Hazard for PFS.
Comparison 2 Hazard for PFS, Outcome 2 Hazard for PFS antiangiogenic therapy plus chemotherapy.
Comparison 2 Hazard for PFS, Outcome 3 Hazard for PFS bevacizumab.
| Response | Criteria |
| Complete response | Requires all of the following: complete disappearance of all enhancing measurable and non‐measurable disease sustained for at least 4 weeks; no new lesions; no corticosteroids; and stable or improved clinically |
| Partial response | Requires all of the following: ≥ 50% decrease compared with baseline in the sum of products of perpendicular diameters of all measurable enhancing lesions sustained for at least 4 weeks; no new lesions; stable or reduced corticosteroid dose; and stable or improved clinically |
| Stable disease | Requires all of the following: does not qualify for complete response, partial response or progression; and stable clinically |
| Progression | Defined by any of the following: ≥ 25% increase in sum of the products of perpendicular diameters of enhancing lesions; any new lesion; or clinical deterioration |
| Response | Criteria |
| Complete response | Requires all of the following: complete disappearance of all enhancing measurable and non‐measurable disease sustained for at least 4 weeks; no new lesions; stable or improved non‐enhancing (T2/FLAIR) lesions; and participant must be off corticosteroids or on physiological replacement doses only, and stable or improved clinically. In the absence of a confirming scan 4 weeks later, this response will be considered only stable disease |
| Partial response | Requires all of the following: ≥ 50% decrease, compared with baseline, in the sum of products of perpendicular diameters of all measurable enhancing lesions sustained for at least 4 weeks; no progression of non‐measurable disease; no new lesions; stable or improved non‐enhancing (T2/FLAIR) lesions on same or lower dose of corticosteroids compared with baseline scan; and participant must be taking a corticosteroid dose not greater than the dose at the time of baseline scan and is stable or improved clinically. In the absence of a confirming scan 4 weeks later, this response will be considered only stable disease |
| Stable disease | Stable disease occurs if participant does not qualify for complete response, partial response or progression (see next section) and requires the following: stable non‐enhancing (T2/FLAIR) lesions on same or lower dose of corticosteroids compared with baseline scan and clinically stable status. In the event that corticosteroid dose was increased for new symptoms and signs without confirmation of disease progression on neuroimaging, subsequent follow‐up imaging shows that this increase in corticosteroids was required because of disease progression; the last scan considered to show stable disease will be the scan obtained when the corticosteroid dose was equivalent to the baseline dose |
| Progression | Progression is defined by any of the following: ≥ 25% increase in sum of products of perpendicular diameters of enhancing lesions (compared with baseline if no decrease) on stable or increasing doses of corticosteroids; significant increase in T2/FLAIR non‐enhancing lesions on stable or increasing doses of corticosteroids compared with baseline scan or best response after initiation of therapy, not due to co‐morbid events; appearance of any new lesions; clear progression of non‐measurable lesions; or definite clinical deterioration not attributable to other causes apart from the tumour, or to decrease in corticosteroid dose. Failure to return for evaluation as a result of death or deteriorating condition should also be considered as progression |
| Grade ≥ 3 adverse event | REGAL (%, c = cediranib, l = cediranib + lomustine) | AVAGLIO (%) | RTOG 0825 (%, crt = chemoradiotherapy, a = adjuvant chemotherapy) |
| Haemorrhage | c: 0 l: 0.8 | 3.3 | crt: 0 a: 1.6 |
| Wound‐healing complications | NR | 3.3 | crt: 1 a: 1.6 |
| Thromboembolic events (arterial ATE, venous VTE) | c: 3.1 l: 4.9 | ATE: 5 VTE: 7.6 | crt: 4.6 a: 7.7 |
| Hypertension | c: 14.1 l: 6.5 | 11.3 | crt: 1.3 a: 4.2 |
| Proteinuria | NR | 5.4 | NR |
| GI perforation | NR | 1.1 | crt: 0 a: 0.8 |
| Thrombocytopaenia | c: 1.6 l: 38.3 | 15 | crt: 10.2 a: 11.1 |
| Fatigue | c: 16,4 l: 15.4 | 7.4 | crt: 0 a: 13.1 |
| Diarrhoea | c: 6.3 l: 5.7 | NR | NR |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Hazard for OS Show forest plot | 7 | 2987 | Hazard Ratio (Fixed, 95% CI) | 0.94 [0.86, 1.02] |
| 1.1 Adjuvant | 5 | 2522 | Hazard Ratio (Fixed, 95% CI) | 0.92 [0.84, 1.01] |
| 1.2 Recurrent | 2 | 465 | Hazard Ratio (Fixed, 95% CI) | 1.02 [0.84, 1.24] |
| 2 Hazard for OS antiangiogenic therapy alone Show forest plot | 2 | 292 | Hazard Ratio (Fixed, 95% CI) | 1.26 [0.96, 1.65] |
| 3 Hazard for OS antiangiogenic plus chemotherapy Show forest plot | 7 | 2806 | Hazard Ratio (Fixed, 95% CI) | 0.91 [0.83, 1.00] |
| 4 Hazard for OS bevacizumab Show forest plot | 4 | 1852 | Hazard Ratio (Fixed, 95% CI) | 0.92 [0.83, 1.02] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Hazard for PFS Show forest plot | 6 | 2847 | Hazard Ratio (Fixed, 95% CI) | 0.74 [0.68, 0.81] |
| 1.1 Adjuvant | 5 | 2522 | Hazard Ratio (Fixed, 95% CI) | 0.72 [0.66, 0.79] |
| 1.2 Recurrent | 1 | 325 | Hazard Ratio (Fixed, 95% CI) | 0.90 [0.70, 1.15] |
| 2 Hazard for PFS antiangiogenic therapy plus chemotherapy Show forest plot | 6 | 2716 | Hazard Ratio (Fixed, 95% CI) | 0.72 [0.66, 0.79] |
| 3 Hazard for PFS bevacizumab Show forest plot | 3 | 1712 | Hazard Ratio (Fixed, 95% CI) | 0.66 [0.59, 0.74] |
