Surgery versus radiosurgery for people with single or solitary brain metastases

Abstract

This is the protocol for a review and there is no abstract. The objectives are as follows:

To assess the effectiveness and safety of surgery versus radiosurgery for people with single or solitary brain metastases.

Background

Description of the condition

The development of brain metastases is a frequent complication in people with cancer. In brain metastases cancer cells migrate from the place where they first formed (primary tumour) and travel, mainly through the blood, to the brain and form one or more tumours. Approximately 9.6% of all primary sites combined may spread to the brain during the course of the disease. Although many different malignant tumours have the ability to infiltrate the central nervous system, the most common primary tumours responsible for brain metastases are lung cancer (19.9%), breast cancer (5.1%), renal cell carcinoma (6.5%) and melanoma (6.9%). In contrast, other carcinomas, for instance prostate, oesophageal, oropharyngeal or non‐melanoma skin cancers, rarely infiltrate the brain (Barnholtz‐Sloan 2004; Bouffet 1997; Nayak 2012; Sundermeyer 2005). Brain metastases occur in more than 64% of people with lung cancer, and approximately 20% of those with breast cancer (Lassman 2003).

The haematogenous spread (when cancer cells are transported through the blood to distant sites of the body) is the most common mechanism of metastasis to the brain (Gavrilovic 2005) and as a consequence, the junction of the grey matter and white matter is the most frequent location, probably because blood vessels have a narrow diameter, acting as a trap for clumps of tumour cells (Delattre 1988).

Brain metastases, in the majority of cases, are multiple lesions that are diagnosed in later stages of the disease. However, in some cases brain metastases appear as the only deposit detected, either as a solitary brain metastasis, defined as "the only known metastasis of a tumour in the whole body which happens to be localised in the central nervous system" or as a single (also named singular) brain metastasis, defined as "a single cerebral metastasis with additional metastases in other organ systems” (Westphal 2003).

Description of the intervention

The most widely used therapeutic modalities for single or solitary brain metastases are surgery and also some forms of radiosurgery.

Surgery consists of either a biopsy or a resection of the metastasis by means of a neurosurgical technique. Resection can be either partial or complete, as confirmed by postoperative imaging. Radiosurgery is a type of external radiation therapy where ionising radiation beams are precisely focused on the brain metastasis (Pannullo 2011). Radiosurgery includes different technical options, for example robotic delivery of radiation, multiple convergent sources of cobalt and other technical devices adapted to linear accelerators (LINACs) (Flickinger 1994; Joseph 1996; Suh 2010).

How the intervention might work

In people with multiple brain metastases palliative radiotherapy or steroids, or both are the treatment of choice (Bradley 2004; Patchell 2003). However, in people with solitary or single brain metastasis a more radical approach using surgery or radiosurgery has shown positive results. A Cochrane review (Patil 2012) including three clinical trials, concluded from one of them (Andrews 2004) that people with only one brain metastasis may live longer when they receive radiosurgery in addition to whole brain radiation therapy (WBRT) versus WBRT alone. Another Cochrane review (Hart 2005) including three clinical trials (Mintz 1996; Patchell 1990; Vecht 1993) concluded that surgery and WBRT may reduce the proportion of deaths due to neurological cause and functionally independent survival. However, it did not show a clear improvement in overall survival.

Management of people newly diagnosed with single or solitary brain metastasis varies widely with location and extension of the primary tumour, histological subtype of the tumour, location of the metastasis and treatment facilities of the referred centre being the most relevant factors (Bradley 2004; Patchell 2003). People having two or more brain metastases (oligometastasis) are not generally considered candidates for surgery, therefore we will not consider this subgroup in this review.

Some studies have described a significant survival extension when single brain metastases are managed aggressively using surgery or radiosurgery of the lesion with or without whole brain irradiation (WBI) (Andrews 2004; Aoyama 2003; Patchell 1998).

Why it is important to do this review

Single and solitary brain metastases are infrequent. However, since more people are living longer with a primary diagnosis, the incidence of brain metastasis is increasing (Nayak 2012).

Choosing the most appropriate treatment for people with brain metastases is always a clinical challenge. It is imperative to balance the risks and benefits due to the incurable nature of the vast majority of metastatic cancer patients, even with a single brain deposit. The best‐described treatment strategies in the literature are surgery and radiosurgery, with or without WBI.

Surgery can be a reasonable option for some people with a solitary/single brain metastasis. However, morbidity and mortality associated with the procedure have to be taken into account. Complications include the increasing or onset of focal motor or sensory deficit, seizures and surgical wound and bone flap infection. In people with solitary and single brain metastasis, survival with surgery has been reported to be better than with radiosurgery (Bougie 2015; Muacevic 1999), although treatment‐related complications have been shown to be higher with surgery Bougie 2015.

Radiosurgery has shown comparable local control compared to surgery, and survival rates may be similar to surgery if patients receive equally aggressive treatment of the primary tumour. Since radiosurgery is a non‐invasive technique, complications are expected to be lower. Complications may include cerebral oedema, seizures and nausea (Muacevic 1999). Radiosurgery may also represent higher costs compared to surgery, due to set up expenses, and costs derived from follow‐up imaging and a possible salvage therapy.

This review aims to assess the effectiveness and safety of surgery versus radiosurgery for people with single or solitary brain metastasis. It is also possible that the meta‐analysis approach may overcome the limitations of small individual studies regarding rare conditions like solitary/single brain metastasis.

Objectives

To assess the effectiveness and safety of surgery versus radiosurgery for people with single or solitary brain metastases.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials (RCTs).

Types of participants

Population:

Adults (aged 18 years or more) with the following characteristics.

• Single or solitary brain metastasis of any size

• Biopsy‐proven malignancy (any tumour histology)

• No previous cranial radiation

• Any chemotherapy or target therapy must have been administered before the study intervention

Types of interventions

Intervention

Surgery (any neurosurgical technique). We will include trials where complete resection was intended. Post‐operative confirmation of metastasis resection with MRI would be desirable, but not necessary for inclusion. Partial resections, as a potential outcome of surgery, will be described and analysed with the intention‐to‐treat principle.

Comparison

Any type of radiosurgery: robotic delivery radiation; multiple convergent sources of cobalt or other technical devices adapted to LINACs, for example any form of LINAC stereotactic radiotherapy with or without re‐locatable frame using single or multiple fractions or Leksell gamma knife radiosurgery.

Co‐intervention

Studies with or without whole brain irradiation (WBRT) as co‐intervention are eligible as long as they compare surgery with radiosurgery.

Studies with or without chemotherapy or target therapy as co‐intervention for the treatment of single or solitary brain metastasis are eligible as long as they compare surgery with radiosurgery.

Studies where post‐intervention management varies between study groups, including non‐standardised post‐intervention management, will be eligible for inclusion. Variations in post‐intervention management between study groups will be carefully analysed.

Types of outcome measures

Primary outcomes

• Overall survival: length of time (in days, weeks or months) until death from any cause. We will assess survival from the time when patients were randomised.

• Adverse events: untoward medical events that may present during treatment with the study intervention, with or without a causal relationship with this treatment (Nebeker 2004). Examples of adverse events that we will consider in this review are: headache; nausea; vomiting; fatigue and seizures or other neurological toxicities; wound complications such as infection or dehiscence; other infections different from wound infection; haematoma or cerebrospinal fluid leak. Number of mild, moderate, and severe adverse events, and number of patients with adverse events will be described. Only moderate and severe adverse events will be taken into account when comparing the two study interventions.

Secondary outcomes

• Progression‐free survival (PFS): survival from time of randomisations to time of disease progression. We will consider disease progression to be an increase in the size of any lesion, development of new lesions, decline in performance status or worsening of symptoms.

• Quality of life: assessed through validated questionnaires of health‐related quality of life (HRQOL) (e.g. Karnofsky performance status (KFS), QLQ‐BN20, QLQ‐C15‐PAL, QLQ‐C30 or FACT‐G with Brain sub scale)

Search methods for identification of studies

Electronic searches

We will carry out electronic searches in the following databases:

The Cochrane Central Register of Controlled Trials (CENTRAL), (The Cochrane Library) Ovid MEDLINE: from 1950 to present EMBASE: from 1980 to present CINAHL: from 1982 to present

We will not apply any restrictions regarding language, publication status or date of publication.

The search strategy for MEDLINE is listed in Appendix 1.

Searching other resources

We will search in the following clinical trials registries: ClinicalTrials.gov (www.clinicaltrials.gov), the UK Clinical Trials Gateway (www.ukctg.nihr.ac.uk), the EU Clinical Trials register (www.clinicaltrialsregister.eu) and the World Health Organization (WHO) International Clinical Trials Registry Platform Search Portal (apps.who.int/trialsearch). For identifying grey literature we will search in the following databases: Open Grey (www.opengrey.eu) () and we will also check the reference lists of all included trials and other systematic reviews identified in the electronic searches. As part of handsearching, we will search the following journals from the fields of oncology, neuro‐oncology, neuro‐surgery and radiotherapy to identify articles of trials published in the last three years:

• Radiotherapy and Oncology

• Journal of Clinical Oncology

• Seminars in Radiation Oncology

• Journal of Neurosurgery

• Neurosurgery

• Neuro‐Oncology

• Journal of Neuro‐Oncology

• World Journal of Surgical Oncology

• Journal of Radiotherapy in Practice

• Journal of Neurology, Neurosurgery and Psychiatry

• Cancer and Metastasis Reviews

In order to identify newly published studies, we will use the PubMed email alert service “My NCBI[A1] ” (National Center for Biotechnology Information), applying the search strategy described in Appendix 1.

If conference communications are identified or we have no access to a study report, we will contact at least one of the study authors by e‐mail. We will also contact principal researchers asking them about possible unpublished trials.

Data collection and analysis

Selection of studies

Two review authors will independently screen all references identified in the search for potential eligibility, by reading the title and abstract. At this point none of the outcomes listed will be required as part of the eligibility criteria for including studies. We will review the full texts of all potentially relevant articles. and will assess studies for eligibility irrespective of their publication status or language of publication. A third review author will resolve any disagreements about which articles are eligible. The review authors will be blinded to the study author’s name, study institution, journal or study results. We will include a flow‐chart of the study selection process, following the PRISMA recommendations (Moher 2009).

Data extraction and management

Using a pre‐tested data collection form (Appendix 2), two review authors will independently collect the following information from individual studies: study eligibility; study design (e.g. randomisation and blinding) and setting; participant characteristics; interventions and comparisons of interest; outcomes; loss to follow‐up; risk of bias (see section below); type of analysis (intention‐to‐treat, per protocol); study funding source; and any conflicts of interest stated by the investigators. Information will be collected in sufficient detail to complete a "Characteristics of included studies" table. We will classify interventions as:

• surgery alone;

• surgery and whole brain irradiation;

• radiosurgery alone, including any technique: radiosurgery using either gamma knife, cyberknife, LINAC, radiosurgery using linear accelerator and stereotactic radiotherapy using any kind of re‐locatable frame;

• any radiosurgery technique plus whole brain irradiation.

We will extract data in order to analyse data following the intention‐to‐treat principle. We will analyse participants in the groups to which they were randomised, regardless of whether or not they received the treatment they had been assigned, or whether or not they had been observed until the completion of the follow‐up period.

We will resolve any disagreements about the data extraction by discussion. If we exclude a potentially relevant study, we will provide the reasons in the 'Characteristics of excluded studies' table. We will also provide relevant information regarding on‐going trials that may be included in future versions of the review in the 'Characteristics of on‐going studies' table.

Assessment of risk of bias in included studies

Two review authors will independently assess the risk of bias using the Cochrane tool for assessing the risk of bias (Higgins 2011) and will judge the following domains in each included study: adequate sequence generation; allocation concealment; patient blinding; provider blinding; data collector blinding; outcome assessor blinding; analyst blinding; percentage of follow‐up and whether incomplete outcome data was addressed; whether the trial was free of selective reporting; and whether the trial was stopped early for benefit (i.e. stop recruiting patients before reaching the sample size or as soon as they find a positive result) (Guyatt 2012). Other bias reported by the study authors or identified by the review authors will also be considered.

If possible (10 or more studies), funnel plots will be used to assess for the potential existence of small study bias for the primary review outcomes (Sterne 2011).

Review authors will judge the risk of bias in each domain as 'low risk', 'high risk' or 'unclear risk'. We will resolve any disagreements by discussion or by consulting a third review author.

Review authors will explain their risk of bias judgements for individual studies in the 'Risk of bias' table. We will present judgements about each methodological quality item as percentages across all included studies in a ’Risk of bias’ graph, and summarise judgements about each methodological quality item for each included study in a ’Risk of bias’ summary chart.

Measures of treatment effect

Time‐to‐event outcomes (i.e. overall survival and survival free of brain relapses) will be analysed as hazard ratios (HRs), and adverse events will be analysed as dichotomous data using risk ratios (RRs). For quality of life, we will use either mean differences (MD) or standardised mean differences (SMD), depending on whether the outcomes are reported using the same or different scales. We will report all effect measures with 95% confidence intervals (CI).

Unit of analysis issues

The unit of analysis for all pre‐defined outcomes will be individual participants, except for the outcome, 'adverse events', which we will also analyse by number of adverse events. When analysing outcomes, we will take into account the level at which randomisation occurs.

Dealing with missing data

Where data are missing, we will contact study authors to request the necessary information. If data are still missing, after attempts to retrieve information have been exhausted, we will report the available results. Missing outcome data will not be imputed for any of the outcomes in the review.

Assessment of heterogeneity

We will assess heterogeneity of effect sizes by visual inspection of forest plots and we will measure heterogeneity using the I² statistic (Higgins 2003). If substantial heterogeneity is detected, we will attempt to explain the differences between studies based on the clinical characteristics (i.e. setting, characteristics of participants, co‐morbidity and treatments) and the methodological characteristics (i.e. randomisation, study quality and analytical method). If included studies are different, we will not combine their results, we will present the results using a narrative approach instead.

Assessment of reporting biases

We will assess potential reporting bias of the review using a funnel plot when 10 or more studies are available (Sterne 2011). The funnel plot will illustrate variability among trials. If asymmetry is detected, we will check other causes such as the methodological quality of studies, the influence of small studies or the presence of heterogeneity.

Data synthesis

If possible, we will undertake a meta‐analysis of the results for all primary and secondary outcomes. For time‐to‐event data (overall survival and survival free of brain relapses), we will use hazard ratios (HR); for adverse events (dichotomous outcomes) we will use risk ratios; and for quality of life we will use mean differences (MD) or standardised mean differences (SMD) as appropriate. We will express all effect measures together with 95% confidence intervals (CI).

We will combine the results using a random‐effects model (Higgins 2003). Differences among studies will be analysed based on clinical characteristics and methodological characteristics of included studies. If the included studies are too diverse for a meta‐analysis, we will present the results using a narrative approach. Data synthesis will be carried out using Review Manager(RevMan 2014).

To interpret the findings and to rate the quality of evidence we will use the GRADE approach (Guyatt 2011). First, we will analyse the overall quality of evidence for each outcome individually, downgrading the evidence from 'high quality' to 'moderate', 'low' or 'very low' depending on the risk of bias, indirectness of evidence, inconsistency, imprecision of effect estimates and potential publication bias. We will take this analysis into account in our conclusions. We will use the GRADEpro Guideline Development Tool to produce a 'Summary of findings' table with the results of this analysis (GRADEpro GDT 2015).

Subgroup analysis and investigation of heterogeneity

When possible, we will carry out subgroup analyses by the location of the primary tumour (lung cancer, breast cancer, renal cell carcinoma, melanoma, and others), by the clinical presentations (solitary or single) and by the use of whole brain radiation or chemotherapy, or both as co‐interventions.

Sensitivity analysis

If meta‐analysis is performed, we will conduct sensitivity analyses to assess the effect on the primary outcome, based on the methodological quality of included studies: studies classified as 'overall low' risk of bias versus those classified as 'overall high' or 'unclear' risk of bias".

Any special sensitivity analysis will take into account possible variations in post‐intervention management between study groups and studies.

Appendices

Appendix 1. Medline search strategy (Ovid)

1 Radiosurgery/ 2 (radiosurg* or stereotactic or linear accelerator or cyberknife or gamma knife or linac).mp. 3 1 or 2 4 surgery.fs. 5 exp Neurosurgical Procedures/ 6 (surg* or neurosurg* or excis*).mp. 7 4 or 5 or 6 8 exp Brain Neoplasms/ 9 ((brain or cerebral or cerebellum) adj5 (tumor* or tumour* or neoplas* or cancer* or carcinoma* or malignan* or metast*)).mp. 10 8 or 9 11 3 and 7 and 10 12 randomized controlled trial.pt. 13 controlled clinical trial.pt. 14 randomized.ab. 15 placebo.ab. 16 clinical trials as topic.sh. 17 randomly.ab. 18 trial.ti. 19 12 or 13 or 14 or 15 or 16 or 17 or 18 20 11 and 19 21 exp animals/ not humans.sh. 22 20 not 21

key:

mp=title, abstract, original title, name of substance word, subject heading word, keyword heading word, protocol supplementary concept word, rare disease supplementary concept word, unique identifier, fs=floating subheading, ab=abstract, sh=subject heading, pt=publication type

Appendix 2. Data collection form

Review: Surgery versus radiosurgery for patients with single or solitary brain metastases
Study ID (surname of first author and year first full report of study was published e.g. Smith 2001)
Report ID
General Information
Date form completed	(dd/mm/yyyy)
Name/ID of person extracting data
Reference citation
Study author contact details
Publication type	(e.g. full report, abstract, letter)
Study eligibility
1. Type of study (Randomised controlled trials)	(yes or no)
2. Participants (Adults aged 18 years or more, with single or solitary brain metastasis of any size, a biopsy‐proven malignancy of any tumour histology, no previous cranial radiation, and any chemotherapy or target therapy administered before the study intervention)	(yes or no)
3. Types of intervention (Any neurosurgical technique where complete resection was intended)	(yes or no)
4. Types of comparison (Any type of radiosurgery: robotic delivering of radiation, multiple convergent sources of cobalt or other technical devices adapted to linear accelerators e.g. any form of LINAC stereotactic radiotherapy with or without re‐locatable frame using single or multiple fractions or Leksell gamma knife radiosurgery.	(yes or no)
Inclusion (Do not proceed if the study does not meet the four eligibility criteria)	(Included or excluded)
Reason for exclusion
Notes(any other information you consider important)
Methods (Descriptions as stated in report/paper)
Country	(where the study was conducted)
Design	(e.g. parallel, cluster)
Was the study multicentre?	(if yes, state No. of centres)
Funders of the trial
Duration of trial	(state start date and end date of trial)
Duration of participation	(from start of recruitment to last follow‐up)
Ethical approval needed/ obtained for study	(yes, no, unclear)
Notes (any other information you consider important)
Participants (Include comparative information for each intervention or comparison group if available)
Population description	(describe any risk factors, and criteria for diagnosing the metastasis)
Setting	(from where were participants enrolled?)
Inclusion criteria
Exclusion criteria
Method of recruitment of participants	(e.g. phone, mail, clinic patients)
Total no. randomised
No. of participants assigned to each group
No. of participants receiving the intended treatment
No. of participants analysed
Withdrawals and exclusions	(if not provided below by outcome)
Age
Sex
Race/Ethnicity
Characteristics of the primary tumour and metastasis	(time to diagnosis, time to metastasis, characteristics of brain metastasis, solitary vs single?)
Notes (any other information you consider important)
Intervention group
Intervention name
No. randomised to group	(specify whether no. people or clusters)
Details of the intervention	(type of surgery, details of the technique)
	(Was a post‐operative confirmation of metastasis resection with MRI done?)
Co‐interventions	(any additional interventions given e.g. whole brain irradiation, chemotherapy or other target therapy. Has the co‐intervention been used equally in both study groups)
Notes (any other information you consider important)
Comparator group
Intervention name	E.g. robotic delivering of radiation, LINAC stereotactic radiotherapy, Leksell gamma knife radiosurgery or other multiple convergent sources of cobalt or other technical devices adapted to linear accelerators)
No. randomised to group
Details of the intervention	(brand, doses, frequency, duration)
Co‐interventions	(any additional interventions given e.g. whole brain irradiation, chemotherapy or other target therapy. Has the co‐intervention been used equally in both study groups)
Notes (any other information you consider important)
Outcome characteristics
Outcome 1 name	Overall survival: survival from the intervention (surgery or radiosurgery) until death from any cause.
Time points measured	(specify whether from start or end of intervention) (How long was the follow‐up for this outcome?)
Time points reported
Person measuring/ reporting
Imputation of missing data	(e.g. assumptions made for ITT analysis)
Notes (any other information you consider important)
Outcome 2 name	Any adverse event
Time points measured	(specify whether from start or end of intervention) (How long was the follow‐up for this outcome?)
Person measuring/ reporting
Notes (any other information you consider important)
Outcome 3 name	Survival free of brain relapses: survival from the intervention until the diagnosis of a new brain metastasis by imaging, either by computed tomography (CT) or by magnetic resonance (MRI).
Time points measured	(specify whether from start or end of intervention) (How long was the follow‐up for this outcome?)
Person measuring/ reporting
Imputation of missing data	(e.g. assumptions made for ITT analysis)
Notes (any other information you consider important)
Outcome 4 name	Quality of life: assessed through validated questionnaires
Time points measured
Time points reported
Person measuring/ reporting
How was quality of life assessed? (measurement scale)
Scales: upper and lower limits (indicate whether high or low score is good)
Is outcome/tool validated?
Imputation of missing data	(e.g. assumptions made for ITT analysis)
Notes (any other information you consider important)
Data and analysis (Descriptions as stated in report/paper)
Outcome 1. Overall survival: survival from the intervention until death from any cause.
Comparison
Outcome
Subgroups
Time point	(specify from start or end of intervention)
Results	Intervention # Event Total in group Control # Event Total in group
Any other results reported	(e.g. odds ratio, risk difference, CI or P value)
No. participants moved from other group	# Reason
Unit of analysis	(by individuals, cluster/groups or body parts)
Statistical methods used and appropriateness of these	Was any adjustment done?
Notes (any other information you consider important)
Outcome 2. Any adverse event
Comparison
Outcome
Subgroups
Time point	(specify from start or end of intervention)
Results	Intervention # Event Total in group Control # Event Total in group
Any other results reported	(e.g. odds ratio, risk difference, CI or P value)
No. missing participants	# Reason
No. participants moved from other group	# Reason
Unit of analysis	(by individuals, cluster/groups or body parts)
Statistical methods used and appropriateness of these	Was any adjustment done?
Notes (any other information you consider important)
Outcome 3. Survival free of brain relapses
Comparison
Outcome
Subgroups
Time point	(specify from start or end of intervention)
Results	Intervention # Event Total in group Control # Event Total in group
Any other results reported	(e.g. odds ratio, risk difference, CI or P value)
No. participants moved from other group	# Reason
Unit of analysis	(by individuals, cluster/groups or body parts)
Statistical methods used and appropriateness of these	Was any adjustment done?
Notes (any other information you consider important)
Outcome 4. Quality of life: assessed through validated questionnaires
Comparison
Outcome
Subgroups
Time point	(specify from start or end of intervention)
Results	Intervention Mean, Median Standard deviation Total in group Control Mean, Median Standard deviation Total in group
Any other results reported	(e.g. odds ratio, risk difference, CI or P value)
No. missing participants	# Reason
No. participants moved from other group	# Reason
Unit of analysis	(by individuals, cluster/groups or body parts)
Statistical methods used and appropriateness of these	Was any adjustment done?
Notes (any other information you consider important)
Risk of Bias assessment
Random sequence generation (selection bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Allocation concealment (selection bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Blinding of participants and personnel for outcome 1 (Overall survival) (performance bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Blinding of participants and personnel for outcome 2 (Any adverse event) (performance bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Blinding of participants and personnel for outcome 3 (Survival free of brain relapses) (performance bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Blinding of participants and personnel for outcome 4 (Quality of life) (performance bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Blinding of outcome 1 assessment (Overall survival) (detection bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Blinding of outcome 2 assessment (Any adverse event) (detection bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Blinding of outcome 3 assessment (Survival free of brain relapses) (detection bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Blinding of outcome 4 assessment (Quality of life) (detection bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Incomplete outcome 1 data (Overall survival) (attrition bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Incomplete outcome 2 data (Any adverse event) (attrition bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Incomplete outcome 3 data (Survival free of brain relapses) (attrition bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Incomplete outcome 4 data (Quality of life) (attrition bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Selective outcome reporting? (reporting bias)	Low risk High risk Unclear (include direct quotes where available with explanatory comments)
Notes (any other information you consider important)

What's new

Date	Event	Description
21 September 2016	Amended	Contact details updated.

Contributions of authors

All review authors drafted and approved the final protocol.

Sources of support

Internal sources

• Nil, Other.

External sources

• Nil, Other.

Declarations of interest

Rafel Fuentes: no conflict of interest related to this protocol Dimelza Osorio: no conflict of interest related to this protocol José Expósito Hernandez: no conflict of interest related to this protocol Daniel Simancas‐Racines: no conflict of interest related to this protocol Xavier Bonfill Cosp: no conflict of interest related to this protocol

Edited (no change to conclusions)