Bone‐modifying agents for the prevention of bone loss in women with early or locally advanced breast cancer: a systematic review and network meta‐analysis

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To assess the effects of different bone‐modifying agents for women with breast cancer in early or locally advanced stages.

Using a network meta‐analysis, we will generate a clinically meaningful ranking of these agents according to safety and efficacy.

Background

Description of the condition

Breast cancer remains the most common cancer among women, accounting for 11.6% of all cancer occurrences (Bray 2018). In 2018, there were more than 2 million new cases of breast cancer worldwide. In the same year 626,679 women died from the disease (Bray 2018).

While long‐term outcomes are improving for women with breast cancer, rates of recurrence and death are still significantly high (Aft 2012; Dhesy‐Thind 2017). Anti‐cancer therapy can lead to a reduction in bone density and a higher risk of osteoporosis (Gnant 2009; Gnant 2011; Hadji 2014). Based on a guideline for supportive care in oncology, the incidence rate of breast cancer patients developing osteoporosis is 18.07 per 1000 person years (meaning on average 18.07 patients develop osteoporosis in 1000 breast cancer patients observed for one year; Khan 2011), which significantly enhances the risk of fractures. The incidence rate is largely influenced by factors such as therapy‐induced menopause as well as oestrogen suppression therapy. Supportive treatments (e.g. corticosteroids) may also harm the bone (Greep 2003; Hadji 2009). Osteoporosis may lead to bone fracture and this in turn may require surgery. Bone fracture is associated with an increased mortality rate; for example, a study by Kanis and colleagues found that men and women aged 50 years or older who had suffered a hip fracture had a much higher mortality rate soon after the event (Kanis 2003). In the same study, the five‐year mortality rates were increased in all age groups, compared to the general population in Sweden (Kanis 2003). That is why it is important to prevent the loss of bone density in breast cancer patients (Kanis 2008).

Premenopausal women with hormone receptor‐positive breast cancer receive treatment that involves tamoxifen alone, or suppression of ovarian function in combination with tamoxifen or aromatase inhibitors, all of which can lead to loss of bone density and an increased risk of osteoporosis (Gnant 2009; Gnant 2011; Hadji 2014; Hadji 2017). Postmenopausal women treated with aromatase inhibitors are also affected and show an increased fracture rate compared to women treated with tamoxifen (Gnant 2015; Hadji 2011; Kalder 2014; Rabaglio 2009). Chemotherapy also leads to loss of bone density (Greep 2003; Hadji 2009).

Description of the intervention

Bone‐targeted treatment is important for breast cancer patients without metastases (metastasis means cancer that has spread from a primary site, e.g. from the breast to the bone). Bone‐modifying agents like bisphosphates or receptor activator of nuclear factor‐kappa B ligand (RANKL)‐inhibitors may prevent cancer treatment‐induced bone loss (CTIBL). In patients with postmenopausal status (those who are either postmenopausal, or premenopausal with suppressed ovarian function), bone‐modifying agents may also prevent bone metastases and prolong overall survival (EBCTCG 2015; Gnant 2015). Bisphosphonates of interest are alendronate, clodronate, ibandronate, pamidronate, risedronate and zoledronate; the RANKL‐inhibitor of interest is denosumab.

How the intervention might work

To accomplish its functions, bone undergoes continuous destruction (resorption) carried out by osteoclasts, and formation by so‐called osteoblasts (Rodan 1998). Bisphosphonates are analogues of pyrophosphate that target osteoclastic cells, and are grouped into amino‐bisphosphonates or non‐amino‐bisphosphonates. Amino‐bisphosphonates affect the osteoclast metabolism by targeting the farnesyl diphosphate synthase, which is responsible for post‐translational modification of guanosine‐5'‐triphosphate‐binding proteins. Non‐amino‐bisphosphonates function by forming an analogue of adenosine triphosphate. The resulting metabolite has toxic properties and induces apoptosis (programmed cell death) of osteoclasts (Reyes 2016). Bisphosphonates therefore suppress bone resorption by promoting the apoptosis of osteoclasts. Bisphosphonates vary in terms of route of administration (oral or intravenous), dose, and frequency and duration of administration.

Denosumab, a fully humanised monoclonal antibody, functions by targeting and neutralising RANKL, which has been found to be a major contributor to the progression of bone metastases (Hanley 2012). RANKL is expressed by osteoblastic stromal cells and binds to RANK, thereby mediating osteoclast differentiation, activation and survival. RANKL is responsible for osteoclast‐mediated bone resorption (Hsu 1999; Yasuda 1998). Denosumab binds to RANKL with high affinity and blocks the interaction of RANKL and RANK. This action decreases bone resorption since the route mediating differentiation, activation and survival of bone resorptive osteoclasts is blocked (Bekker 2004). Denosumab is given subcutaneously at a fixed dose.

Since bisphosphonates and denosumab influence bone metabolism, there could be adverse events such as osteonecrosis of the jaw (a severe bone disease affecting the jaw bone, characterised by delayed wound healing after invasive procedures, as well as infection and death of the bone tissue) and hypocalcaemia (blood calcium levels under the normal range of 2.1 millimoles per litre (mmol/L) to 2.6 mmol/L, which may lead to further complications) (Tesfamariam 2019). Bisphosphonates are also known to have the ability to cause renal complications (Edwards 2013).

Why it is important to do this review

Bisphosphonates are recommended as an addition to usual treatment for postmenopausal women with breast cancer, since they may reduce the risk of fractures and bone recurrence (meaning the development of bone metastases), and prolong overall survival; they may also prevent therapy‐induced bone loss in pre‐ and postmenopausal patients (Dhesy‐Thind 2017; EBCTCG 2015; Gnant 2015; National Guideline Alliance 2018; O'Carrigan 2017). Denosumab has also been found to reduce fractures, but research on long‐term survival is still ongoing (Dhesy‐Thind 2017). An overall ranking of all different treatment options, focusing on different patient‐relevant outcomes, is still missing. Therefore a comparison is urgently needed to inform recommendations in national and international guidelines (Dhesy‐Thind 2017), and to help patients and healthcare providers with decision‐making.

Objectives

To assess the effects of different bone‐modifying agents for women with breast cancer in early or locally advanced stages.

Using a network meta‐analysis, we will generate a clinically meaningful ranking of these agents according to safety and efficacy.

Methods

Criteria for considering studies for this review

Types of studies

We will include studies if they are randomised controlled trials (RCTs). We will include both full‐text and abstract publications if sufficient information on study design, characteristics of participants (women with early or locally advanced breast cancer) and interventions (adjuvant bisphosphonates or RANKL‐inhibitors) are provided. We will include trials that include participants receiving these bone‐modifying agents in at least one treatment arm. In the case of cross‐over trials, only the first period of the trial will be analysed. There will be no limitations with respect to length of follow‐up.

We will exclude studies that are non‐randomised, case reports, and clinical observations.

Types of participants

We will include trials involving adult women (18 years of age and older) with a confirmed diagnosis of early or locally advanced breast cancer, meaning all stages without metastases, defined by the TNM Classification of Malignant Tumors (TNM) staging system showing any for T, any for N and 0 for M. In this staging system 'T' refers to the size and extent of the main (primary) tumour, 'N' refers to the number of nearby lymph nodes that have cancer and 'M' refers to whether the cancer has metastasised, which means that the cancer has spread from the primary tumour to other parts of the body (NCI 2015). By including M0, only studies with non‐metastasised participants will be included in this analysis. We will include both pre‐ and postmenopausal participants. We assume that participants who fulfil the inclusion criteria are equally eligible to be randomised to any of the interventions we plan to compare.

Types of interventions

We will include trials comparing different bone‐modifying agents with each other and with control regimens (placebo or no treatment) as adjuvant therapy for early or locally advanced breast cancer. We will consider any type of bisphosphonate or RANKL‐inhibitor, apart from radioactive bisphosphonates. We will not impose any restriction on the dose, route, frequency or duration of treatment with bone‐modifying agents. We plan to investigate the following comparisons. In order to establish fair comparisons, concomitant treatments should not differ between study arms.

Interventions

• Bisphosphonates (alendronate, clodronate, ibandronate, pamidronate, risedronate and zoledronate)

• RANKL‐inhibitors (denosumab)

• Placebo/no treatment

Comparisons

• Bisphosphonates versus placebo/no treatment

• RANKL‐inhibitors versus placebo/no treatment

• Bisphosphonates versus RANKL‐inhibitors

• One type of bisphosphonate versus another type of bisphosphonate

• One type of RANKL‐inhibitor versus another type of RANKL‐inhibitor

All these options are recommended in clinical guidelines for the prevention of therapy‐induced bone events in breast cancer patients in early or locally advanced stages. We plan to group interventions by evaluating different drug doses together as one drug of interest, according to the product characteristics.

For the patient population described above, all interventions and combinations of interventions will be analysed within a full network (for ideal network see Figure 1). We will include all RCTs comparing at least two study arms with one intervention of interest. Should no direct evidence from RCTs exist and the trials are considered sufficiently similar with respect to the patient population, indirect estimates of intervention effects will be obtained by means of the network calculations.

Figure 1.

Figure 1: ideal network with all possible comparisons of all treatment options

Types of outcome measures

We will include all trials fulfilling the inclusion criteria mentioned above, irrespective of reported outcomes. We will estimate the relative ranking of the competing interventions according to each of the following outcomes. To make sure outcomes are patient‐relevant, we held a meeting with patients and patient representatives during the development of the protocol, and discussed relevance and order of proposed outcomes.

Primary outcomes

• Bone density, defined as the amount of minerals (mostly calcium and phosphorous) contained in a certain volume of bone. This outcome is intended to analyse the potential prevention of cancer treatment‐induced bone loss (CTIBL).

• Quality of life.

Secondary outcomes

• Overall fracture rate, defined as the number of bone fractures of all kinds occurring during and after treatment with bone‐modifying agents. If possible, we will sub‐classify this outcome by site of fracture (vertebral and non‐vertebral).

• Pain response: we will consider all trials reporting on the proportion of participants with pain response.

• Overall survival (time‐to‐event outcome) or all‐cause mortality.

• Disease‐free survival (time‐to‐event outcome), defined as the length of time from diagnosis to the patient surviving without any signs or symptoms (distant, locoregional, or new primary symptoms in the contralateral breast, or as defined in the trial).

• Adverse events:

  • osteonecrosis of the jaw;

  • renal (we will consider all trials reporting renal adverse events. As bone‐modifying agents might be described with renal toxicity with variable expression, we consider renal adverse events to be clinically significant and requiring treatment or hospital admission);

  • bone pain (right after administration);

  • hypocalcaemia.

• Any bone recurrence, meaning the development of metastasis to the bone.

Method and timing of outcome measurement

• Bone density (or bone mineral density): assessed by using validated techniques, e.g. dual‐energy X‐ray absorptiometry, and given as mass of mineral per volume of bone. We will measure this outcome at baseline, six months, one year, two years, or at the longest reported follow‐up.

• Quality of life: assessed using validated generic and disease‐specific questionnaires. We will measure this outcome at baseline, six months, one year, two years, or at the longest reported follow‐up.

• Overall fracture rate: assessed by radiographic imaging and, if indicated, computed tomography or magnetic resonance imaging. We will measure fractures occurring at any time after participants were randomised to intervention or comparator groups.

• Proportion of participants with pain response: assessed using validated generic and disease‐specific questionnaires. We will not impose restrictions on pain assessment tools or the definition of pain response in the trials. Pain scores and analgesic consumption will be collected as part of this outcome. We will measure this outcome at baseline, six months, one year, two years, or at the longest reported follow‐up.

• Adverse events (osteonecrosis of the jaw, renal adverse events and further adverse events): grade 3 and 4 according to the common terminology criteria for adverse events (CTCAE) or as defined in the trial. We will measure this outcome at any time after participants were randomised to intervention or comparator groups.

• Overall survival or all‐cause mortality: defined as the time from randomisation to the date of death. If we are unable to retrieve the necessary information to analyse time‐to‐event outcomes, we will assess the number of events per treatment group for dichotomised outcomes. We will pay special attention regarding the transitivity assumption when it comes to the patient population and different treatments (see Assessment of heterogeneity), e.g. we will set date limits in order to compare more recent studies with each other rather than against older ones, since treatment of breast cancer has changed and older regimens might not be comparable with newer regimens. We will measure this outcome at six months, one year, two years, or at the longest reported follow‐up.

Outcomes to be included in the 'Summary of findings' table

• Bone density

• Quality of life

• Fracture rate

• Pain response

• Overall survival

• Adverse event: osteonecrosis of the jaw

• Adverse event: renal

Search methods for identification of studies

Electronic searches

We will search the following databases.

• The Cochrane Breast Cancer Group's (CBCG's) Specialised Register. The process of identifying studies and coding references is outlined on the CBCG's web site (breastcancer.cochrane.org/sites/breastcancer.cochrane.org/files/public/uploads/specialised_register_details.pdf). Trials with the key words 'early and locally advanced stages', 'bone‐modifying agents', 'bisphosphonates', 'RANKL‐inhibitors' and 'bone density' will be extracted and considered for inclusion in the review.

• CENTRAL (the Cochrane Library, latest issue); see Appendix 1.

• MEDLINE (via OvidSP) from 1946 to present; see Appendix 2.

• Embase (via Embase.com) from 1947 to present; see Appendix 3.

• The WHO International Clinical Trials Registry Platform (ICTRP) search portal (apps.who.int/trialsearch/Default.aspx) will be searched for all prospectively registered and ongoing trials; see Appendix 4.

• ClinicalTrials.gov (clinicaltrials.gov/); see Appendix 5.

Searching other resources

We will try to identify further studies from reference lists of identified relevant trials or reviews. A copy of the full article for each reference reporting a potentially eligible trial will be obtained. Where this is not possible, attempts will be made to contact trial authors to provide additional information.

Data collection and analysis

Selection of studies

We will apply Cochrane’s Screen4Me workflow to help assess the search results. Screen4Me comprises three components:

• known assessments, a service that matches records in the search results to records that have already been screened in Cochrane Crowd (Cochrane’s citizen science platform where the Crowd help to identify and describe health evidence) and labelled as 'RCT' or 'not an RCT';

• the RCT classifier, a machine‐learning model that distinguishes RCTs from non‐RCTs;

• Cochrane Crowd, if appropriate (crowd.cochrane.org).

More detailed information regarding evaluations of the Screen4Me components can be found in the following publications: Marshall 2018; McDonald 2017; Noel‐Storr 2018; Thomas 2017.

Two review authors (TJ, NS) will independently screen the results for eligibility for this review by reading the abstracts using Covidence software (Covidence). We will code the abstracts as either 'retrieve' or 'do not retrieve'. In the case of disagreement, or if it is unclear whether we should retrieve the abstract or not, we will obtain the full‐ text publication for further discussion. Independent review authors will eliminate studies that clearly do not satisfy the inclusion criteria, and obtain full‐text copies of the remaining studies. Two review authors (TJ, NS) will read these studies independently to select relevant studies; in the event of disagreement, a third author will adjudicate (AW). We will not anonymise the studies before assessment. We will include a PRISMA flow diagram in the full review that will show the status of identified studies (Moher 2009), as recommended in Section 11.2.1 of the Cochrane Handbook for Systematic Reviews of Interventions (Schunemann 2011). We will include studies in the 'Characteristics of included studies' table irrespective of whether measured outcome data are reported in a ‘usable’ way.

There will be no language restrictions and articles will be translated if they are not published in English. We will record the details of excluded studies in the 'Characteristics of excluded studies' table.

Data extraction and management

Two review authors (TJ, NS) will extract data using a standardised data extraction form. If the authors are unable to reach a consensus, we will consult a third review author (KK or AA) for a final decision. If required, we will contact the authors of specific studies for supplementary information (Higgins 2011). After agreement has been reached, we will enter data into Review Manager 5 (Review Manager 2014). We will extract the following information.

• General information: author, title, source, publication date, country, language, duplicate publications.

• Quality assessment: sequence generation, allocation concealment, blinding (participants, personnel, outcome assessors), incomplete outcome data, selective outcome reporting, other sources of bias.

• Study characteristics: trial design, aims, setting and dates, source of participants, inclusion/exclusion criteria, comparability of groups, statistical methods, power calculations, subgroup analysis, treatment cross‐overs, compliance with assigned treatment, time point of randomisation, length of follow‐up.

• Participant characteristics: participant details, baseline demographics, age, ethnicity, number of participants recruited/allocated/evaluated, participants lost to follow‐up, cancer type and stage, hormone receptor status, HER2 status, additional diagnoses, type and intensity of pain, menopausal status.

• Interventions: type, dose, route, frequency and cycles of treatment with bone‐modifying agents, producer information, prescription and administration of vitamin D and calcium, main cancer treatment (endocrine therapy, chemotherapy) and its details.

• Outcomes: bone density and measuring instrument, quality‐of‐life and measuring instrument, fracture rate (vertebral or non‐vertebral), pain response and how it is measured, analgesic consumption, overall survival or all‐cause mortality, disease‐free survival, adverse events (osteonecrosis of the jaw, renal, bone pain right after administration, hypocalcaemia), any bone recurrence.

• Notes: sponsorship/funding for trial and notable conflict of interest of authors' trial registry record information (e.g. national clinical trial numbers).

We will collate multiple reports of the same study, so that each study rather than each report is the unit of interest in the review. We will collect characteristics of the included studies in sufficient detail to populate a table of ‘Characteristics of included studies’ in the full review.

Assessment of risk of bias in included studies

We will complete a 'Risk of bias' table for each included study, using Review Manager 5 (Review Manager 2014). Two review authors (TJ, NS) will independently assess risk of bias for each study and if they are unable to reach a consensus, we will consult a third review author (AW) for a final decision. We will assess the following criteria, as outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a).

• Sequence generation

• Allocation concealment

• Blinding (participants and personnel)

• Blinding (outcome assessors)

• Incomplete outcome data

• Selective outcome reporting

• Other sources of bias

For each of the domains listed above, we will judge each study to be at either high, low or unclear risk of bias. We will present these judgements in the ‘Risk of bias’ tables, along with justification for our decision.

For performance bias (blinding of participants and personnel) and detection bias (blinding of outcome assessment), we will evaluate the risk of bias separately for each outcome. We will group these outcomes according to whether they were measured subjectively or objectively when reporting our findings in the 'Risk of bias' tables.

We define the outcome 'overall survival' as not being influenced by blinding of patients or outcome assessors (objective outcome). We define the following outcomes as not being influenced by blinding of patients (objective outcomes):

• bone density;

• overall fracture rate;

• disease‐free survival;

• adverse event: osteonecrosis of the jaw;

• adverse event: renal;

• adverse event: hypocalcaemia;

• any bone recurrence.

We define the following outcomes as subjective outcomes:

• quality of life;

• pain response;

• adverse event: bone pain (right after administration).

Measures of treatment effect

Relative treatment effect

We will use intention‐to‐treat data. For binary outcomes, we will extract the number of patients and number of events per arm and calculate risk ratios (RRs) with 95% confidence intervals (CIs) for each trial. This will apply for the outcomes overall fracture rate (fractures might be defined as evident fractures with associated symptoms or as defined in the trials themselves, while vertebral fractures might be defined as 20% to 25% or greater height reduction, measured by radiographs), pain response, bone density (assessed with dual‐energy X‐ray absorptiometry scans, or as reported in trials if different), and adverse events like osteonecrosis of the jaw and renal adverse events.

We will calculate continuous outcomes as mean differences (MDs) when assessed with the same instruments; otherwise we will calculate standardised mean differences (SMDs) with 95% CIs. This will apply for the outcome quality of life.

For time‐to‐event outcomes, we will extract hazard ratios (HRs) from published data, according to Parmar 1998 and Tierney 2007. This will apply for the outcomes overall and disease‐free survival.

Since we plan to conduct a network meta‐analysis, we will define the direction for every RR or HR we are reporting and add 'RR or HR smaller than 1.0 favours...' when reporting results. We will not report pairwise meta‐analysis results since these have been shown elsewhere (O'Carrigan 2017).

Relative treatment ranking

We will obtain a treatment hierarchy using P‐scores (Rücker 2015). P‐scores allow ranking of treatments on a continuous zero‐to‐one scale in a frequentist network meta‐analysis.

Unit of analysis issues

Studies with multiple treatment groups

As recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), for studies with multiple treatment groups we will combine arms as long as they can be regarded as subtypes of the same intervention.

When arms cannot be pooled this way, we will include multi‐armed trials using a network meta‐analysis approach that accounts for the within‐study correlation between the effect sizes by re‐weighting all comparisons of each multi‐armed study (Rücker 2012; Rücker 2014). For pairwise meta‐analysis, we will treat multi‐armed studies as multiple independent comparisons and will not combine these data in any analysis.

Dealing with missing data

As suggested in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), we will take the following steps to deal with missing data.

Whenever possible, we will contact the original investigators to request relevant missing data. If the number of participants evaluated for a given outcome is not reported, we will use the number of participants randomised per treatment arm as the denominator. If only percentages but no absolute number of events are reported for binary outcomes, we will calculate numerators using percentages. If estimates for mean and standard deviations are missing, we will calculate these statistics from reported data whenever possible, using approaches described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). If standard deviations are missing and we are not able to calculate them from reported data, we will calculate values according to a validated imputation method (Furukawa 2006). If data are not reported numerically but graphically, we will estimate missing data from figures. We will perform sensitivity analyses to assess how sensitive results are to imputing data in some way. We will address the potential impact of missing data on the findings of the review in the 'Discussion' section.

Assessment of heterogeneity

Assessment of clinical and methodological heterogeneity within treatment comparisons

To evaluate the presence of clinical heterogeneity, we will generate summary statistics for the important clinical and methodological characteristics across all included studies. Within each pairwise comparison, we will assess the presence of clinical heterogeneity by visually inspecting the similarity of these characteristics.

Assessment of transitivity across treatment comparisons

To infer about the assumption of transitivity, we will assess whether the included interventions are similar when they are evaluated in RCTs with different designs; for example, whether combinations of two drugs are administered the same way in studies comparing them to other combinations of two drugs and in those comparing combinations of two drugs to combinations of three drugs. Furthermore, we will compare the distribution of the potential effect modifiers across the different pairwise comparisons.

Assessment of statistical heterogeneity and inconsistency

To evaluate the presence of heterogeneity and inconsistency in the entire network, we will give the generalised heterogeneity statistic Q_total and the generalised I² statistic, as described in Schwarzer 2015. We will use the decomp.design command in the R package netmeta 1.0‐1 (R Core Team 2018, Rücker 2019) for decomposition of the heterogeneity statistic into a Q statistic for assessing the heterogeneity between studies with the same design and a Q statistic for assessing the designs inconsistency to identify the amount of heterogeneity/inconsistency within, as well as between, designs.

To evaluate the presence of inconsistency locally, we will compare direct and indirect treatment estimates of each treatment comparison. This can serve as a check for consistency of a network meta‐analysis (Dias 2010). For this purpose, we will use the netsplit command in the R package netmeta 1.0‐1, which enables the splitting of the network evidence into direct and indirect contributions (R Core Team 2018; Rücker 2019). For each treatment comparison, we will present direct and indirect treatment estimates plus the network estimate using forest plots. In addition, for each comparison we will give the z‐value and P value of test for disagreement (direct versus indirect). It should be noted that in a network of evidence there may be many loops, and with multiple testing there is an increased likelihood that we might find an inconsistent loop by chance. Therefore, we will be cautious when deriving conclusions from this approach.

If we find substantive heterogeneity or inconsistency (or both), we will explore possible sources by performing pre‐specified sensitivity and subgroup analyses (see below). In addition, we will review the evidence base, reconsider inclusion criteria and discuss the potential role of unmeasured effect modifiers to identify further sources.

We will interpret I² values according to Chapter 9.5.2 of the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2011), as follows.

• 0% to 40%, might not be important.

• 30% to 60% may represent moderate heterogeneity.

• 50% to 90% may represent substantial heterogeneity.

• 75% to 100% represents considerable heterogeneity.

We will use the P value of the Chi² test only for describing the extent of heterogeneity and not for determining statistical significance. In addition, we will report Tau², the between‐study variance in random‐effects meta‐analysis. In the event of excessive heterogeneity that is unexplained by subgroup analyses, we will not report outcome results as the pooled effect estimate of the network meta‐analysis, but provide a narrative description of the results of each study.

Assessment of reporting biases

In pairwise comparisons with at least 10 trials, we will examine the presence of small‐study effects graphically by generating funnel plots. We will use linear regression tests (Egger 1997) to test for funnel plot asymmetry. A P value less than 0.1 will be considered significant for this test (Sterne 2011). We will additionally consider comparison‐adjusted funnel plots and the accompanying regression test to assess selection bias. We will examine the presence of small‐study effects for the primary outcome only. Moreover, we will search study registries to identify trials that are completed but not published.

Data synthesis

Methods for direct treatment comparisons

Pairwise comparisons are part of the network meta‐analysis, thus we do not plan to perform additional pairwise meta‐analyses. In order to outline available direct evidence, we will provide forest plots for pairwise comparisons, without giving an overall estimate. Only in the case where data are not sufficient to be combined in a network meta‐analysis, e.g. in the case of inconsistency, we will perform pairwise meta‐analyses according to recommendations provided in Chapter 9 of the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2011). We will use the random‐effects model. We will use the R package meta for statistical analyses (Schwarzer 2007; R Core Team 2018). When trials are too clinically heterogenous to be combined, we will perform only subgroup analyses without calculating an overall estimate.

Methods for indirect and mixed comparisons

Should the data be considered sufficiently similar to be combined, we will perform a network meta‐analysis using the frequentist weighted least squared approach described by Rücker 2012. We will use a random‐effects model, taking into account the correlated treatment effects in multi‐arm studies. We will assume a common estimate for the heterogeneity variance across the different comparisons. To evaluate the extent to which treatments are connected, we will give a network plot for our primary and secondary outcomes. For each comparison, we will give the estimated treatment effect along with its 95% CI. We will present the results graphically using forest plots, with placebo/no treatment as the reference treatment. We will use the R package netmeta for statistical analyses (R Core Team 2018; Rücker 2019).

Certainty in the evidence

Two review authors (TJ, NS) will independently rate the certainty of the evidence for each outcome. We will use the GRADE system to rank the certainty in the evidence, using GRADEpro GDT software (GRADEpro GDT) and the guidelines provided in Chapter 12.2 of the Cochrane Handbook for Systematic Reviews of Interventions (Schunemann 2011) and specifically for network meta‐analyses (Puhan 2014).

The GRADE approach to assess the certainty in the body of evidence for each outcome by performing network meta‐analysis uses five domains: study limitations (risk of bias of included studies); indirectness (relevance to the review question); inconsistency (looking at heterogeneity and incoherence); imprecision (e.g. confidence intervals); and publication bias.

The GRADE assessment of the evidence for each outcome results in one of the four following categories.

• High certainty: we are very confident that the true effect lies close to that of the effect estimate.

• Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the effect estimate, but there is a possibility that it is substantially different.

• Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the effect estimate.

• Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the effect estimate.

The GRADE system uses the following criteria for assigning a GRADE level to a body of evidence (Schunemann 2011).

• High quality: randomised trials; or double‐upgraded observational studies.

• Moderate quality: downgraded randomised trials; or upgraded observational studies.

• Low quality: double‐downgraded randomised trials; or observational studies.

• Very low quality: triple‐downgraded randomised trials; or downgraded observational studies; or case series/case reports.

We will downgrade our assessment of the evidence if there is:

• serious (‐1) or very serious (‐2) limitation to study quality;

• important inconsistency (‐1);

• some (‐1) or major (‐2) uncertainty about directness;

• imprecise or sparse data (‐1);

• high probability of reporting bias (‐1).

'Summary of findings' table

We will include a 'Summary of findings' table to present the main findings in a transparent and simple tabular format. In particular, we will include key information concerning the certainty in the evidence, the magnitude of effect of the interventions examined, and the sum of available data on the outcomes defined under 'Outcomes to be included in the 'Summary of findings' table'. We will need to adapt the table which is developed using the GRADEpro GDT online program (GRADEpro GDT) to comply with the results of the network meta‐analysis.

Subgroup analysis and investigation of heterogeneity

We will perform subgroup analyses related to factors which might have an effect on the outcomes, as follows.

• Premenopausal participants versus postmenopausal participants (premenopausal patients with ovarian suppression therapy (gonadotropin‐releasing hormone antagonist (GnHR) analogues, ovariectomy, radiomenolysis) will still be considered as premenopausal).

• Participants receiving endocrine therapy versus those not receiving endocrine therapy, or hormone receptor (HR)‐positive versus HR‐negative, also compared to human epidermal growth factor receptor 2 (HER2)‐positive.

• Type of endocrine therapy (e.g. tamoxifen alone versus aromatase inhibitor alone versus ovarian function suppression (OFS) in combination with tamoxifen versus OFS in combination with aromatase inhibitor).

• Type of bone‐modifying agent (bisphosphonate versus RANKL inhibitor).

• Bisphosphonates of the first (non‐amino bisphosphonates: etidronate, clodronate) and second generation (amino‐bisphosphonates: alendronate, risedronate, pamidronate, ibandronate, zoledronate), independently.

• Duration of bone modifying agent: one year versus two‐to‐five years.

• Participants with high risk of relapse (defined as receiving chemotherapy additionally to endocrine therapy) versus participants only receiving endocrine therapy.

• Participants with status N1, N2, N3 versus status N0.

Sensitivity analysis

We will perform a sensitivity analysis to test the robustness of our results by analysing studies with low risk of bias only. We will judge studies as being at high risk of bias overall if they are at high risk for two or more of the 'Risk of bias' domains.

Appendices

Appendix 1. CENTRAL via CRSO

#1 MESH DESCRIPTOR Breast Neoplasms EXPLODE ALL TREES

#2 breast near cancer*

#3 breast near neoplasm*

#4 breast near carcinom*

#5 breast near tumour*

#6 breast near tumor*

#7 breast near malignan*

#8 #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7

#9 MESH DESCRIPTOR Diphosphonates EXPLODE ALL TREES

#10 (diphosphonate* or diphosph*nate*)

#11 (bisphosph*nate* or biphosph*nate*)

#12 (diphosphonic* or bisphosphonic*)

#13 #9 OR #10 OR #11 OR #12

#14 MESH DESCRIPTOR Alendronate EXPLODE ALL TREES

#15 (alendronat* or aledronic*)

#16 (fosamax* or binosto* or adronat* or alendros* or onclast*)

#17 #14 OR #15 OR #16

#18 MESH DESCRIPTOR Clodronic Acid EXPLODE ALL TREES

#19 (clodronic* or clodronat*)

#20 (bonefos* or clasteon* or difosfonal* or ossiten* or mebonat* or loron*)

#21 Cl2MDP

#22 #18 OR #19 OR #20 OR #21

#23 MESH DESCRIPTOR Etidronic Acid EXPLODE ALL TREES

#24 (etidronic* or etidronat*)

#25 (didronel* or xidifon* or dicalcium or xidiphon*)

#26 (HEDP or EHDP)

#27 #23 OR #24 OR #25 OR #26

#28 MESH DESCRIPTOR Technetium Tc 99m Medronate EXPLODE ALL TREES

#29 (medronat* or medronic*)

#30 (Technetium near2 Tc 99m near2 Medronat*)

#31 (Tc‐99m‐MDP or Tc‐MDP)

#32 #28 OR #29 OR #30 OR #31

#33 (ibandronic* or ibandrovic* or ibandronat*)

#34 (bon*iva* or bondronat* or adronil*)

#35 (RPR102289A or RPR‐102289A)

#36 (BM210955 or BM‐210955)

#37 #33 OR #34 OR #35 OR #36

#38 (pamidronat* or pamidronic* or amidronat*)

#39 MESH DESCRIPTOR Risedronate Sodium EXPLODE ALL TREES

#40 (risedronic* or risedronat*)

#41 (actonel* or atelvia* or benet* or optinate*)

#42 (NE58095 or NE‐58095)

#43 #40 OR #41 OR #42

#44 (zoledronic* or zoledronat*)

#45 (zometa* or zomera* or aclasta* or reclast* or aredia* or orazol*)

#46 (m05BA08 or CGP‐42446$ or CGP42446$ or zol‐446 or zol446)

#47 #44 OR #45 OR #46

#48 (neridronat* or neridronic*)

#49 ("AHHexBP" or "6AHHDP" or "6‐AHHDP" or nerixia)

#50 #48 OR #49

#51 (tiludronat* or tiludronic*)

#52 (skelid* or tildren* or sr 42329 or sr42329 or sr 41319b or sr41319b)

#53 #51 OR #52

#54 (incadronat* or incadronic*)

#55 (cimadronat* or cimadronic*)

#56 (bisphonal* or YM175 or YM 175)

#57 #54 OR #55 OR #56

#58 MESH DESCRIPTOR RANK Ligand EXPLODE ALL TREES

#59 (rank* near3 ligand*)

#60 RANK ligand inhibitor*

#61 (protein* near2 RANKL) or (protein* near2 TRANCE)

#62 (osteoclast* near2 differentiation factor*)

#63 (osteoclast* near2 ligand*)

#64 Tumor Necrosis Factor‐Related Activation‐Induced Cytokin*

#65 #58 OR #59 OR #60 OR #61 OR #62 OR #63 OR #64

#66 MESH DESCRIPTOR Receptor Activator of Nuclear Factor‐kappa B EXPLODE ALL TREES

#67 ((receptor activator* near3 nf‐kappab) or (receptor activator* near3 nuclear factor kappab))

#68 ((receptor activator* near3 nf‐kappa) or (receptor activator* near3 nuclear factor kappa))

#69 tnfrsf11a

#70 (trance r or trance receptor*)

#71 #66 OR #67 OR #68 OR #69 OR #70

#72 MESH DESCRIPTOR Denosumab EXPLODE ALL TREES

#73 denosumab*

#74 (xgeva* or prolia*)

#75 (AMG162 or AMG‐162)

#76 #73 OR #74 OR #75

#77 romosozumab*

#78 (cdp 7851 or cdp7851)

#79 (AMG 785 or AMG785)

#80 evenity*

#81 #77 OR #78 OR #79 OR #80

#82 blosozumab*

#83 (Ly2541546 or Ly 2541546)

#84 #82 or #83

#85 #13 OR #17 OR #22 OR #27 OR #32 OR #37 OR #38 OR #43 OR #47 OR #50 OR #53 OR #57 OR #71 OR #76 OR #81 or #84

#86 #8 AND #85

key: *: truncation, near#: adjacent within # number of words

Appendix 2. MEDLINE (via Ovid)

#	Searches
1	exp BREAST NEOPLASMS/
2	(breast adj6 cancer*).tw.
3	(breast adj6 neoplasm*).tw.
4	(breast adj6 carcinoma*).tw.
5	(breast adj6 tumo?r*).tw.
6	or/1‐5
7	exp DIPHOSPHONATES/
8	(disphosphonate* or diphosph#nate*).tw,kf,ot,nm.
9	(bisphosph#nate* or biphosph#nate*).tw,kf,ot,nm.
10	(diphosphonic* or bisphosphonic*).tw,kw,ot,nm.
11	or/7‐10
12	ALENDRONATE/
13	(alendronat* or aledronic*).tw,kf,ot,nm.
14	(fosamax* or binosto* or adronat* or alendros* or onclast*).tw,kf,ot,nm.
15	or/12‐14
16	CLODRONIC ACID/
17	(clodronic* or clodronat*).tw,kf,ot,nm.
18	(bonefos* or clasteon* or difosfonal* or ossiten* or mebonat* or loron*).tw,kf,ot,nm.
19	Cl2MDP.tw,kf,ot,nm.
20	or/16‐19
21	ETIDRONIC ACID/
22	(etidronic* or etidronat*).tw,kf,ot,nm.
23	(didronel* or xidifon* or dicalcium or xidiphon*).tw,kf,ot.
24	(HEDP or EHDP).tw,kf,ot.
25	or/21‐24
26	TECHNETIUM TC 99M MEDRONATE/
27	(medronat* or medronic*).tw,kf,ot,nm.
28	(Technetium adj2 Tc 99m adj2 Medronat*).tw,kf,ot,nm.
29	or/26‐28
30	IBANDRONIC ACID/
31	(ibandronic* or ibandrovic* or ibandronat*).tw,kf,ot,nm.
32	(bon?iva* or bondronat* or adronil*).tw,kf,ot,nm.
33	(RPR102289A or RPR‐102289A).tw,kf,ot,nm.
34	(BM210955 or BM‐210955).tw,kf,ot,nm.
35	or/30‐34
36	PAMIDRONATE/
37	(pamidronat* or pamidronic* or amidronat*).tw,kf,ot,nm.
38	or/36‐37
39	RISEDRONIC ACID/
40	(risedronic* or risedronat*).tw,kf,ot,nm.
41	(actonel* or atelvia* or benet* or optinate*).tw,kf,ot,nm.
42	(NE58095 or NE‐58095).tw,kf,ot,nm.
43	or/39‐42
44	ZOLEDRONIC ACID/
45	(zoledronic* or zoledronat*).tw,kf,ot,nm.
46	(zometa* or zomera* or aclasta* or reclast* or aredia* or orazol*).tw,kf,ot,nm.
47	(m05BA08 or CGP‐42446$ or CGP42446$ or zol‐446 or zol446).tw,kf,ot,nm.
48	or/44‐47
49	(neridronat* or neridronic*).tw,kf,ot,nm.
50	(AHHexBP or 6AHHDP or 6‐AHHDP or nerixia).tw,kf,ot,nm.
51	or/49‐50
52	(tiludronat* or tiludronic*).tw,kf,ot,nm.
53	(skelid* or tildren* or sr 42329 or sr42329 or sr 41319b or sr41319b).tw,kf,ot,nm.
54	or/52‐53
55	(incadronat* or incadronic*).tw,kf,ot,nm.
56	(cimadronat* or cimadronic*).tw,kf,ot,nm.
57	(bisphonal* or YM175 or YM 175).tw,kf,ot,nm.
58	or/55‐57
59	(olpadronat* or olpadronic*).tw,kf,ot,nm.
60	(ig 8801 or ig8801).tw,kf,ot,nm.
61	or/59‐60
62	RANK Ligand/
63	(rank* adj3 ligand*).tw,kf,ot,nm.
64	((protein* adj2 RANKL) or (protein* adj2 TRANCE)).tw,kf,ot,nm.
65	(osteoclast* adj2 differentiation factor*).tw,kf,ot,nm.
66	(osteoclast* adj2 ligand*).tw,kf,ot,nm.
67	tumor necrosis factor related activation induced cytokine.tw,kf,ot,nm.
68	or/62‐67
69	RECEPTOR ACTIVATOR OF NUCLEAR FACTOR‐KAPPA B/
70	((receptor activator* adj3 nf‐kappab) or (receptor activator* adj3 nuclear factor kappab)).tw,kf,ot,nm.
71	((receptor activator* adj3 nf‐kappa) or (receptor activator* adj3 nuclear factor kappa)).tw,kf,ot,nm.
72	tnfrsf11a.tw.
73	(trance r or trance receptor*).tw,kf,ot,nm.
74	or/69‐73
75	DENOSUMAB/
76	denosumab*.tw,kf,ot,nm.
77	(xgeva* or prolia*).tw,kf,ot,nm.
78	(AMG162 or AMG‐162).tw,kf,ot,nm.
79	or/75‐78
80	romosozumab*.tw,kf,ot,nm.
81	(cdp 7851 or cdp7851).tw,kf,ot,nm.
82	(AMG 785 or AMG785).tw,kf,ot,nm.
83	evenity*.tw,kf,ot,nm.
84	or/80‐83
85	blosozumab*.tw,kf,ot,nm.
86	(Ly2541546 or Ly 2541546).tw,kf,ot,nm.
87	or/85‐86
88	15 or 20 or 25 or 29 or 35 or 38 or 43 or 48 or 51 or 54 or 58 or 61 or 68 or 74 or 79 or 84 or 87
89	randomized controlled trial.pt.
90	controlled clinical trial.pt.
91	randomi?ed.ab.
92	placebo.ab.
93	drug therapy.fs.
94	randomly.ab.
95	trial.ab.
96	groups.ab.
97	or/89‐96
98	exp ANIMALS/ not HUMANS/
99	97 not 98
100	6 and 88 and 99

key: exp # /: explode # MeSH subject heading, tw: text word, kf: keyword heading word, ot: original title, ti: title, nm: substance name, pt: publication type, ab: abstract, fs: floating subheading; *: truncation, #: mandated wild card character stands for one character; ?: mandated wild card character stands for one or no character, adj#: adjacent within # number of words searchline #89‐#99: Cochrane RCT‐Filter sensitivity‐maximizing version

Appendix 3. EMBASE via Ovid SP

#	Searches
1	exp BREAST/
2	exp BREAST DISEASE/
3	(1 or 2) and exp NEOPLASM/
4	exp BREAST TUMOR/
5	exp BREAST CANCER/
6	exp BREAST CARCINOMA/
7	(breast* adj5 (neoplas* or cancer* or carcin* or tumo* or metasta* or malig*)).ti,ab.
8	or/3‐7
9	exp BISPHOSPHONIC ACID DERIVATIVE/
10	(diphosphonate* or diphosph#nate*).tw,kw,ot.
11	(bisphosph#nate* or biphosph#nate*).tw,kw,ot.
12	(diphosphonic* or bisphosphonic*).tw,kw,ot.
13	or/9‐12
14	ALENDRONIC ACID/
15	(alendronat* or aledronic*).tw,kw,ot.
16	(fosamax* or binosto* or adronat* or alendros* or onclast* or alend*).tw,kw.
17	or/14‐16
18	CLODRONIC ACID/
19	(clodronic* or clodronat*).tw,kw.
20	(bonefos* or clasteon* or difosfonal* or ossiten* or mebonat* or loron*).tw,kw.
21	Cl2MDP.tw,kw.
22	or/18‐21
23	ETIDRONIC ACID/
24	(etidronic* or etidronat*).tw.
25	(didronel* or xidifon* or dicalcium or xidiphon*).tw,kw.
26	(HEDP or EHDP).tw,kw.
27	or/23‐26
28	MEDRONATE TECHNETIUM TC 99M/
29	(medronat* or medronic*).tw,kw.
30	(Technetium adj2 Tc 99m adj2 Medronat*).tw,kw.
31	(Tc 99m MDP or Tc MDP).tw,kw.
32	or/28‐31
33	IBANDRONIC ACID/
34	(ibandronic* or ibandrovic* or ibandronat*).tw,kw.
35	(bon?iva* or bondronat* or adronil*).tw,kw.
36	(RPR102289A or RPR‐102289A).tw,kw.
37	(BM210955 or BM‐210955).tw,kw.
38	or/33‐37
39	PAMIDRONIC ACID/
40	(pamidro* or pamidronic* or amidronat*).tw,kw.
41	or/39‐40
42	RISEDRONIC ACID/
43	(risedronic* or risedronat*).tw,kw.
44	(actonel* or atelvia* or benet* or optinate*).tw,kw.
45	(NE58095 or NE‐58095).tw,kw.
46	(cgp 23339 or cgp23339 or cgp 23339a or cgp 23339a).tw,kw.
47	or/42‐46
48	ZOLEDRONIC ACID/
49	(zoledronic* or zoledronat*).tw,kw.
50	(zometa* or zomera* or aclasta* or reclast* or aredia* or orazol*).tw,kw.
51	(m05BA08 or CGP‐42446 or CGP42446 or CGP‐42446a or CGP42446a or zol‐446 or zol446).tw,kw.
52	or/48‐51
53	NERIDRONIC ACID/
54	(neridronat* or neridronic*).tw,kw.
55	(AHHexBP or 6AHHDP or 6‐AHHDP or nerixia).tw,kw.
56	or/53‐55
57	TILUDRONIC ACID/
58	(tiludronat* or tiludronic*).tw,kw.
59	(skelid* or tildren* or sr 42329 or sr42329 or sr 41319b or sr41319b).tw,kw.
60	or/57‐59
61	INCADRONIC ACID/
62	(incadronat* or incadronic*).tw,kw.
63	(cimadronat* or cimadronic*).tw,kw.
64	(bisphonal* or YM 175 or YM175).tw,kw.
65	or/61‐64
66	OLPADRONIC ACID/
67	(olpadronat* or olpadronic*).tw,kw.
68	(ig 8801 or ig8801).tw,kw.
69	or/66‐68
70	OSTEOCLAST DIFFERENTIATION FACTOR/
71	(osteoclast* adj2 differentiation factor*).tw,kw.
72	(osteoclast* adj2 ligand*).tw,kw.
73	(rank* adj3 ligand*).tw,kw.
74	((protein* adj2 RANKL) or (protein* adj2 TRANCE)).tw,kw.
75	tumor necrosis factor related activation induced cytokine.tw,kw.
76	or/70‐75
77	"RECEPTOR ACTIVATOR OF NUCLEAR FACTOR KAPPA B"/
78	((receptor activator* adj3 nf‐kappab) or (receptor activator* adj3 nuclear factor kappab)).tw,kw.
79	((receptor activator* adj3 nf‐kappa) or (receptor activator* adj3 nuclear factor kappa)).tw,kw.
80	tnfrsf11a.tw.
81	(trance r or trance receptor*).tw,kw.
82	or/77‐81
83	DENOSUMAB/
84	denosumab*.tw,kw.
85	(xgeva* or prolia*).tw,kw.
86	(AMG162 or AMG‐162).tw,kw.
87	or/83‐86
88	ROMOSOZUMAB/
89	(cdp 7851 or cdp7851).tw,kw.
90	(AMG 785 or AMG785).tw,kw.
91	evenity*.tw,kw.
92	or/88‐91
93	BLOSOZUMAB/
94	Blosozumab*.tw,kw.
95	(Ly2541546 or Ly 2541546).tw,kw.
96	or/93‐95
97	13 or 17 or 22 or 27 or 32 or 38 or 41 or 47 or 52 or 56 or 60 or 65 or 69 or 76 or 82 or 87 or 92 or 96
98	8 and 97
99	RANDOMIZED CONTROLLED TRIAL/
100	CONTROLLED CLINICAL STUDY/
101	random*.ti,ab.
102	RANDOMIZATION/
103	INTERMETHOD COMPARISON/
104	placebo.ti,ab.
105	(compare or compared or comparison).ti.
106	(open adj label).ti,ab.
107	((double or single or doubly or singly) adj (blind or blinded or blindly)).ti,ab.
108	DOUBLE BLIND PROCEDURE/
109	parallel group$1.ti,ab.
110	(crossover or cross over).ti,ab.
111	((assign$ or match or matched or allocation) adj5 (alternate or group$1 or intervention$1 or patient$1 or subject$1 or participant$1)).ti,ab.
112	(controlled adj7 (study or design or trial)).ti,ab.
113	(volunteer or volunteers).ti,ab.
114	trial.ti.
115	or/99‐114
116	8 and 97 and 115

key:

exp # /: explode # MeSH subject heading, tw: text word, kw: keyword, ti: title, ab: abstract, $, *: truncation, #: mandated wild card character stands for one character; ?: mandated wild card character stands for one or no character, *,$: wildcard, adj#: adjacent within # number of words

Appendix 4. WHO ICTRP

WHO ICTRP search strategy

Basic search 1. breast cancer AND bisphosphonate 2. breast cancer AND diphosphonates 3. breast cancer AND denusomab 4. breast cancer AND RANK

Advanced search Condition: Breast cancer Intervention: diphosphonate OR bisphosphonate OR bisphosphonate Recruitment status: ALL

Condition: Breast cancer Intervention: zoledronate OR zoledronic OR clodronate OR clodronic OR etidronic OR etidronate Recruitment status: ALL

Condition: Breast cancer Intervention: medronate OR medronic OR alendronate OR aledronic OR ibandronate OR ibandronic Recruitment status: ALL

Condition: Breast cancer Intervention: pamidronate OR pamidronic OR risedronate OR risedronic OR tiludronate OR tiludronic Recruitment status: ALL

Condition: Breast cancer Intervention: incadronate OR incadronic OR olpadronate OR olpadronic OR neridronate OR neridronic Recruitment status: ALL

Condition: Breast cancer Intervention: rank ligand OR rankl OR denosumab Recruitment status: ALL

Appendix 5. Clinicaltrials.gov

ClinicalTrials.gov search strategy

Basic Search Breast cancer AND bisphosphonates Breast cancer AND Diphosphonates Breast cancer AND Denosumab Breast cancer AND “RANK ligand”

Advanced search Conditions: Breast cancer* OR breast neoplasm* OR breast carcinoma* Interventions:

bisphosphonates OR diphosphonates OR zoledronate OR "zoledronic acid" OR clodronate OR "clodronic acid" OR "etidronic acid" OR

etidronate OR

medronate OR medronic OR alendronate OR aledronic OR ibandronate OR ibandronic OR pamidronate OR pamidronic OR risedronate OR risedronic OR tiludronate OR tiludronic OR incadronate OR incadronic OR olpadronate OR olpadronic OR neridronate OR neridronic OR rank OR osteoclast OR denosumab

Recruitment: All studies Study type: Interventional studies Gender: Studies with Female Participants

Contributions of authors

Drafting the protocol: Tina Jakob Study selection: Ina Monsef, Tina Jakob, Nicole Skoetz Extracting data from studies: Tina Jakob, Nicole Skoetz Entering data into Review Manager 5: Tina Jakob Carrying out the analysis: Anne Adams, Kathrin Kuhr Interpreting the analysis: Tina Jakob, Achim Wöckel, Christian Maurer Drafting the final review: Tina Jakob Disagreement resolution: Achim Wöckel, Christian Maurer Updating the review: Tina Jakob

Sources of support

Internal sources

• No sources of support supplied

External sources

• Federal Ministry of Education and Research, Germany.

  Grant number: 01KG1806

Declarations of interest

Tina Jakob: awarded a grant from the Federal Ministry of Education and Research to perform this systematic review and this does not lead to a conflict of interest Ina Monsef: none known Kathrin Kuhr: awarded a grant from the Federal Ministry of Education and Research to perform this systematic review and this does not lead to a conflict of interest Anne Adams: awarded a grant from the Federal Ministry of Education and Research to perform this systematic review and this does not lead to a conflict of interest Achim Wöckel: none known Christian Maurer: awarded a travel grant to attend ASCO 2017 from Amgen. Nicole Skoetz: awarded a grant from the Federal Ministry of Education and Research to perform this systematic review and this does not lead to a conflict of interest

New