Duration of Bisphosphonate Treatment for Patients with Osteoporosis: A Review of Clinical Effectiveness and Guidelines

Dave Marchand; Hannah Loshak

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Duration of Bisphosphonate Treatment for Patients with Osteoporosis: A Review of Clinical Effectiveness and Guidelines

CADTH Rapid Response Report: Summary with Critical Appraisal

Dave Marchand and Hannah Loshak.

Ottawa (ON): Canadian Agency for Drugs and Technologies in Health; 2019 Oct 4.

Abbreviations

ACP: American College of Physicians
AE: adverse event
AFF: atypical femoral fractures
AGREE II: Appraisal of Guidelines for Research and Evaluation Instrument II
AHRQ: Agency for Healthcare Research and Quality
ALN: alendronate
AMSTAR: A MeaSurement Tool to Assess systematic Reviews
AR: absolute risk
BIOSIS: BioSciences Information Service of Biological Abstracts
BIS: bisphosphonate
BMD: bone mineral density
CGC: clinical guidelines committee
CI: confidence interval
CINAHL: Cumulative Index to Nursing and Allied Health Literature
CRD: University of York Centre for Reviews and Dissemination
DXA: dual-energy x-ray absorptiometry
EMAS: European Menopause and Andropause Society
EMBASE: Excerpta Medica database
ETN: etidronate
FLEX: Fracture Interventional Trial Long Term Extension
FRAX: tool to evaluate fracture risk of patients
GP: general practitioner
GRADE: Grading of Recommendations, Assessment, Development and Evaluations
HORIZON-PFT: Health Outcomes and Reduced Incidence with Zoledronate One Yearly – Pivotal Fracture Trial
HR: hazard ratio
HRQoL: health related quality of life
HTA: health technology assessment
IBN: ibandronate
MA: meta-analysis
MEDLINE: Medical Literature Analysis and Retrieval System Online
MeSH: Medical Subject Heading
NHMRC: National Health and Medical Research Council
NHS: National Health Service
NIH: National Institutes of Health
NMA: network meta-analysis
NOGG: National Osteoporosis Guideline Group
NR: not reported
NRS: non-randomized study
OCEBM: Oxford Centre for Evidence-Based Medicine
ONJ: osteonecrosis of the jaw
OR: odds ratio
PM: postmenopausal
PROSPERO: International prospective register of systematic reviews
PubMed: public MEDLINE
RACGP: Royal Australian College of General Practitioners
RCT: randomized controlled trial
RR: relative risk
RSN: risedronate
SD: standard deviation
SEIOMM: Spanish Society for Research on Bone and Mineral Metabolism
SR: systematic review
WHO: World Health Organisation
WRHA: Winnipeg Regional Health Authority
ZLN: zoledronic acid

Context and Policy Issues

Throughout life bone tissue is constantly formed and resorbed; however, with age this process becomes less effective. Osteoporosis is a bone disease characterised by a deterioration of bone tissue in which bones become progressively less dense and more fragile.¹ When bones are severely weakened by osteoporosis they can fracture easily as a result of minimal trauma (such as falling from standing).² Despite overall skeletal fragility, fractures of the wrist, shoulder, vertebrae and hip are those most commonly associated with osteoporosis.¹^,²

Osteoporosis can affect Canadians of all ages; though, it is more common in those 50 years or older.³ In this population, it is estimated there were 57,413 hospitalizations for a total of 832,594 days from April 1, 2007 to March 31, 2008, costing $1.181 billion (2010 Canadian dollars) in emergency care, acute care admissions, and same day surgeries.⁴ Furthermore, females are disproportionately affected then males.¹^,⁴ There are many risk factors for osteoporosis, including cigarette smoking, long term use of corticosteroids or other high risk medications, and insufficient dietary intake of calcium and vitamin D.¹^,⁵ The presence of these, along with findings from a physical examination and laboratory investigations, may prompt further investigation with a bone mineral density (BMD) test using a dual X-ray absorptiometry (DXA) method.⁶ A DXA t-score result between −1 and – 2.5 standard deviations indicates low bone mass (osteopenia), while osteoporosis is a t-score at or below –2.5 standard deviations.¹^,³ This report will focus on patients with osteoporosis.

The goals of therapy for patients with osteoporosis include the prevention of fractures, prevention of associated disability and loss of independence, as well as the preservation or enhancement of bone mass.⁵ There are several pharmacological treatments for established osteoporosis, such as: bone resorption inhibitors, which slow down bone loss (e.g., bisphosphonates [BIS], denosumab, raloxifene, and estrogen), as well as bone formation agents, which stimulate the formation of bone tissue (e.g., teriparatide).⁵ This report will focus on the BIS class of pharmaceuticals, which are the mainstay of osteoporosis treatment and include: alendronate (ALN), etidronate (ETN), risedronate (RSN), and zoledronic acid (ZLN).

As with any drug therapy, there are adverse effects (AE) associated with BISs and their antifracture benefits ought to be weighed against any possible risks to the patient. Because of concerns about serious AEs, such as osteonecrosis of the jaw (ONJ), atrial fibrillation, and atypical femoral fractures (AFF), there have been suggestions of implementing drug holidays as part of the therapy, where the medication is stopped for a period of time, often several years.³^,⁵^,⁷^–¹² However, the optimal duration of therapy before a drug holiday could be considered and the length of the drug holiday (either indefinite or otherwise), is uncertain and remains a clinical controversy.

CADTH has previously reviewed the evidence for the use of bisphosphonates.¹³^–¹⁵ One report was a list of abstracts based on evidence available in August of 2019,¹⁵ one focused on a different patient population (patients with osteopenia and low risk fractures),¹⁴ and another focused on prevention of osteoporosis in patients treated with systemic corticosteroids.¹³ The objective of the current report is to evaluate the clinical evidence and evidence-based guidelines regarding various treatment durations of BISs for the treatment of osteoporosis.

Research Questions

What is the clinical effectiveness of various treatment durations and courses of bisphosphonates for osteoporosis?
What are the evidence based guidelines regarding length of treatment with bisphosphonates for osteoporosis?

Key Findings

One relevant health technology assessment and four systematic reviews (including one with meta-analysis) were identified as clinical evidence of various treatment duration and courses of bisphosphonates for osteoporosis.

The identified literature were heterogenous and revealed mixed conclusions regarding optimal treatment duration of bisphosphonates for osteoporosis. Incidence of atypical femoral fractures increased with duration of bisphosphonate treatment. Other types of adverse events were generally no different between the various durations of treatment reported, except for one study reporting four deaths (none considered associated with treatment) in their continuation group compared with none in the discontinuation group. No clear direction emerged regarding the effect of treatment duration on bone mineral density or on risks related to other types of fractures (e.g., vertebral, nonvertebral, morphometric).

Six evidence-based guidelines were identified regarding the length of treatment with bisphosphonates for osteoporosis. These offer no clear direction on optimal duration of bisphosphonate treatment or drug holidays; however, most guidelines are in concordance that regular patient monitoring and the individualisation of treatment in response to clinical and paraclinical manifestations.

The limitations of the included studies, such as the heterogeneity of primary studies included in the systematic reviews and the low-quality evidence upon which guideline recommendations were based, should be considered when interpreting the results.

Methods

Literature Search Methods

A limited literature search was conducted by an information specialist on key resources including Medline via OVID, the Cochrane Library, the University of York Centre for Reviews and Dissemination (CRD) databases, the websites of Canadian and major international health technology agencies, as well as a focused Internet search. The search strategy was comprised of both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concepts were bisphosphonates and osteoporosis. No filters were applied to limit the retrieval by study type. Where possible, retrieval was limited to the human population. The search was also limited to English language documents published between January 1, 2014 and August 16, 2019. Internet links were provided, where available.

Selection Criteria and Methods

One reviewer screened citations and selected studies. In the first level of screening, titles and abstracts were reviewed and potentially relevant articles were retrieved and assessed for inclusion. The final selection of full-text articles was based on the inclusion criteria presented in Table 1.

Table 1

Selection Criteria.

Exclusion Criteria

Articles were excluded if they did not meet the selection criteria outlined in Table 1 or they were duplicate publications. SRs that had broader inclusion criteria than the present review were examined in detail to ascertain whether data could be extracted from a relevant sub-set of included studies, rather than excluding the SR entirely. Primary studies that had a mixed population at baseline (i.e., some participants with osteopenia and others with osteoporosis) were only included if authors reported relevant results stratified by initial bone mineral density (BMD) t-score or baseline diagnosis. Primary studies retrieved by the search were excluded if they were captured in one or more included SR. Primary studies that were reported in multiple SRs are noted as such, and results were extracted from the publication reporting the most complete data.

Critical Appraisal of Individual Studies

The included Health Technology Assessment (HTA) and SRs were critically appraised by one reviewer using A Measurement Tool to Assess Systematic Reviews 2 (AMSTAR 2),¹⁶ and the guidelines were assessed with the Appraisal of Guidelines for Research Evaluation II (AGREE II) instrument.¹⁷ Summary scores were not calculated for the included studies; rather, a review of the strengths and limitations of each included study were described narratively.

Summary of Evidence

Quantity of Research Available

A total of 196 citations were identified in the literature search. Following screening of titles and abstracts, 183 citations were excluded and 13 potentially relevant reports from the electronic search were retrieved for full-text review. In addition, 10 potentially relevant publications were retrieved from the grey literature search for full-text review. Of these 23 potentially relevant articles, 12 publications were excluded for various reasons, while 11 publications met the inclusion criteria and were included in this report. These were comprised of one HTA,¹⁸ four systematic reviews (SR),¹⁹^–²² including one with meta-analysis (MA),²⁰ and six guidelines.⁶^,²³^–²⁷

Appendix 1 presents the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA)²⁸ flowchart of the study selection. Note that because the included HTA and SRs had broader inclusion criteria than the present review (i.e., were wider in scope), only subsets of primary studies from the included systematic reviews that met the selection criteria for the present review are described.

Appendix 6 includes seven additional references that did not meet the inclusion criteria of this report but may be of interest.

Summary of Study Characteristics

One HTA,¹⁸ four SRs,¹⁹^–²² including one MA,²⁰ and six guidelines⁶^,²³^–²⁷ were identified and included in this review. The HTA,¹⁸ met the inclusion criteria for this report; however, none of the primary studies included in the HTA met the eligibility criteria for this report; therefore, no summary of these primary studies could be provided. Detailed characteristics are available in Appendix 2, Table 2, and Table 3.

Study Design

The HTA,¹⁸ was published in 2016, three SRs were published in 2019,¹⁹^–²¹ and one in 2014.²² They sought out relevant RCTs,¹⁹^–²² clinical trials,²⁰^,²¹ cohort studies,²¹ and observational studies.¹⁹^,²⁰ There were nine,¹⁹ five,²⁰ seven,²¹ and four²² relevant primary studies from these SRs, with some study overlap as shown in Appendix 5 Table 8, resulting in 17 unique primary studies.

Six guidelines were identified regarding osteoporosis that contained recommendations for the duration of BIS treatment.⁶^,²³^–²⁷ Four guidelines are published in 2017,²³^–²⁶ one in 2015,²⁷ and one in 2014.⁶ The first guideline, by the American College of Physicians (ACP), is an update to a 2011 SR by the Agency for Healthcare Research and Quality (AHRQ).²³ The second, by the European Menopause and Andropause Society (EMAS), is based on a SR of published literature up to 2017.²⁴ The third, by the National Osteoporosis Guideline Group is based on evidence from published SRs, MAs, and RCTs from a PubMed search.²⁵ The fourth guideline, by The Royal Australian College of General Practitioners (RACGP), is based on a SR to update their 2010 guidelines.²⁶ The fifth, published by the Winnipeg Regional Health Authority (WRHA) is based on a SR to update their 2010 guideline.⁶ Lastly, while the summary of the sixth guideline, published by the Spanish Society for Research on Bone and Mineral Metabolism (SEIOMM), was published in English, the full guideline was in Spanish; therefore, we were not able to extract data on study design.²⁷

As indicated in Table 3, guideline authors used a variety of evidence quality assessment methods, including the Grading of Recommendations Assessment, Development and Evaluation (GRADE) tool,²³ A Measurement Tool to Assess Systematic Reviews (AMSTAR),²⁵ the National Health and Medical Research Council (NHMRC) grades of recommendation,²⁶ and the Oxford Centre for Evidence-Based Medicine (OCEBM) tool.²⁷ One guideline,⁶ used a quality assessment tool developed for the 1998 Canadian clinical practice guidelines for diabetes, while the other did not report performing a quality assessment of the evidence.²⁴ Recommendations were developed by consensus,²⁶ a modified Delphi method,⁶ or by committee vote,²³^,²⁵ Recommendation development methods were unclear in one guideline²⁴ and it were unable to be assessed in another.²⁷

Country of Origin

The HTA¹⁸ and one SR¹⁹ were authored in the United Kingdom. Other SRs were authored in the United States of America²⁰^,²¹ and Norway.²²

The guidelines were developed in the United States of America,²³ Greece,²⁴ the United Kingdom,²⁵ Australia,²⁶ Spain,²⁷ and Canada.⁶

Patient Population

The HTA included studies with women 65 years of age or older or men 75 years of age or older.¹⁸ Two SRs included postmenopausal (PM) women,¹⁹^,²² with osteoporosis,²² or on osteoporosis medication for one year or more,¹⁹ Two SRs included men,²⁰^,²¹ the first considered men or PM women aged 50 years or older²⁰, while the second considered participants with osteoporosis or osteopenia who had received osteoporosis treatment for at least three years.²¹ The populations of the primary studies included within the SRs were heterogenous, including some studies with only women, only PM women, or mixed sexes; a variety of ages over 50 years; varying baseline characteristics (e.g., number of risk factors, number of fractures, BMD t-score); and varying number of participants up to 188,814.

The target population of the ACP guidelines are adults with low BMD or osteoporosis, with clinicians involved in their care as the intended user.²³ The European, United Kingdom, and Australian guidelines target men 50 years or older and PM women,²⁴^–²⁶ and the intended users are health care professionals involved in osteoporosis management.²⁵^,²⁶ Whereas we were unable to determine the intended user of the Spanish guidelines, their target population are men or PM women diagnosed with osteoporosis.²⁷ Lastly, the Canadian guidelines target women and men over the age of 50 and are intended for Manitoba health care providers as well as policy makers.⁶

Interventions and Comparators

The HTA evaluated BISs compared to each other, to placebo, or to other non-active treatment.¹⁸ In three SRs, various osteoporosis pharmacotherapies such as BISs (e.g., alendronate [ALN], risedronate [RSN], ibandronate, [IBN], etidronate [ETN], or zoledronic acid [ZLN]) or denosumab, were compared by varying treatment lengths.¹⁹^,²¹^,²² The fourth SR compared long-term (i.e. three or more years) osteoporosis drug treatments of any kind to controls (i.e., placebo or active control) or to a drug holiday of one year or more.²⁰ The interventions and comparators of the included studies were heterogenous and included various BISs (i.e., ALN, RSN, IBN, ETN, and ZLN), for various lengths of time (ranging from one to 10 years) and with various comparators such as placebo or no drug (i.e., a drug holiday).

The guidelines considered a broader scope of osteoporosis treatments, including BISs, and denosumab,⁶^,²³^–²⁷ and some also considered teriparatide, selective estrogen receptor modulators, estrogen, calcium, and vitamin D.⁶^,²³^,²⁵^–²⁷

Outcomes

The outcomes considered in the HTA included: fragility fractures, BMD at the femoral neck, mortality, adverse events (AE), compliance, health related quality of life (HRQoL), and health care resource use.¹⁸

The SRs considered outcomes relating to: bone turnover markers,¹⁹ BMD,¹⁹^–²² fractures,¹⁹^–²² AEs,²⁰^,²² osteonecrosis of the jaw (ONJ),¹⁹ and atypical femoral fractures (AFF).¹⁹^,²⁰

The outcomes of interest in three of the guidelines included: reduction in fractures,²³^,²⁴^,²⁶ BMD,²⁶ and AEs²³^,²⁴ Outcomes were not specified at the outset in two guidelines,⁶^,²⁵ and were not able to be assessed in another.²⁷

Summary of Critical Appraisal

Additional details regarding the strengths and limitations of the included publications are provided in Appendix 3, Table 4, and Table 5.

Health Technology Assessment

The HTA had several strengths, such as: clear objectives and inclusion criteria, reporting of key search terms and search strategies, provision of a list of included studies and summary of their characteristics, and a quality assessment of included studies using an appropriate tool. Assessors registered their protocol to an online database of systematic review protocols on health-related topics (i.e., PROSPERO) prior to the conduct of the review. Furthermore, they considered risk of bias when interpreting and discussing the results of individual studies. Attrition of 10% or greater was found in 63% of included studies, five trials were either open label or single blind and therefore at high risk of performance bias, and 28% of studies reported assessing outcomes in a blinded manner.¹⁸ These strengths of reporting increase confidence in the findings and the reproducibility of the HTA.

Systematic Reviews

Three SRs¹⁹^–²¹ (one with meta-analysis)²⁰ included clear objectives and inclusion criteria, while one SR did not.²² Two SRs did not describe included studies in adequate detail, with poor consistency in reporting the population, intervention, comparator, and research design.¹⁹^,²² All SRs reported on source of funding for the review.¹⁹^–²²

Reporting issues were apparent in one SR¹⁹ where discrepancies were identified between results stated in the SR and those in primary studies. These reporting issues decrease the confidence in the findings and the reproducibility of the Dennison, 2019¹⁹ SR.

One SR reported an a priori protocol registered to an online database of systematic review protocols on health-related topics (i.e., PROSPERO), performing study selection and data extraction in duplicate, ²⁰ while the others did not. A comprehensive literature search strategy, including hand searching of the bibliography of eligible studies, was used in two SRs,²⁰^,²¹ while the others did not.¹⁹^,²²

The quality of included studies was assessed by Fink, 2019²⁰ using Agency for Healthcare Research and Quality (AHRQ) criteria and by Nayak, 2019²¹ using the Cochrane risk of bias tool and the Newcastle-Ottawa quality assessment scale, while other SRs did not perform quality assessments, introducing uncertainty in their results.¹⁹^,²²

Furthermore, Fink, 2019²⁰ inferred fracture risk data from studies not designed to compare fracture incidence between groups. For instance, the primary study by Tonino, 2000²⁹ and the one by Bone, 2004³⁰ were designed to measure change in BMD and reported the number of incident fractures as AEs. In addition, their sample size ranged from 247 to 350 participants and it is unclear if they were adequately powered to report on fracture risk. Despite this, authors produced results on between-group relative risk differences of fractures which may lead to incorrect conclusions.

Authors of the SR with MA²¹ justified combining the data in a meta-analysis and they used the DerSimonian and Laird random-effect method, to calculate summary effect size estimates. However, unless all studies are of similar size, the DerSimonian and Laird method is known to be inefficient when estimating the between-study variance, but efficient when estimating treatment effect.³¹ The two cohort studies that were part of the MA had 39,502 participants³² and 183 participants.³³ As such, the meta-analysts were correct in restricting their inference to statements about the treatment effect only. In addition, the degree of inconsistency between the two studies (i.e., heterogeneity) was assessed as 94.3% (I² value); yet, authors did not discuss the likely impact on the results.²¹

Evidence-Based Guidelines

In the six guidelines⁶^,¹⁰^,²³^–²⁷ the scope and purpose were well described. While only one guideline is developed in Canada,⁶^,¹⁰ two others are developed in countries part of the Commonwealth of Nations,²⁵^,²⁶ with universal or near-universal health care coverage for core medical services, and their findings may be generalizable to the Canadian health care setting. The integral version of one guideline is only available in Spanish, restricting the assessment of strengths and limitations to information that is published in the English summary²⁷

None of the guidelines sought the views and preferences of the target population.⁶^,¹⁰^,²³^–²⁷ All of the guidelines provided an explicit link between the recommendations and the evidence, as well as the criteria for selecting the evidence and the strengths and limitations of the evidence.⁶^,¹⁰^,²³^–²⁷ The methods for formulating recommendations is described in four guidelines,⁶^,²³^,²⁵^,²⁶ but not in one,²⁴ and it is unable to be assessed in another.²⁷ Four guidelines indicate that their recommendations have been externally reviewed by experts prior to publication,⁶^,²³^,²⁵^,²⁶ while one guideline did not,²⁴ and it is unable to be assessed in another.²⁷ Three guidelines provided a procedures for future updates to their recommendations.²³^,²⁵^,²⁶

Summary of Findings

A detailed summary of findings and guideline recommendations is provided in Appendix 4, Table 6, and Table 7. The HTA,¹⁸ met the inclusion criteria for this report; however, none of the primary studies included in the HTA met the eligibility criteria for this report; therefore, no summary could be provided.

Clinical Evidence for Varying Treatment Durations and Courses of Bisphosphonates for Osteoporosis

Bone Mineral Density

Information regarding the effect of duration and course of BIS treatment for osteoporosis on BMD was available from four unique primary studies²⁹^,³⁰^,³³^,³⁴ included, with overlap, in three SRs.¹⁹^,²¹^,²²

A cohort study³³ in the Dennison, 2019¹⁹ SR reported that “BMD trends were similar in patients who sustained a fracture during the holiday versus those who did not sustain.”¹⁹ (p.1737) (no effect estimates or statistics provided).

Two unique primary studies²⁹^,³⁴ in the Nayak, 2019²¹ SR discussed mean percent change in BMD between various timepoints. The first primary study compared ALN 5 mg for seven years versus ALN 10 mg for seven years versus ALN 20 mg for two years, followed by 5 mg for three years, followed by two years of placebo.²⁹ BMD results measured at the lumbar spine between year six and seven favoured the continuation groups, while those measured at the femoral neck, trochanter, distal third of the forearm, and total body were no different between groups.²¹ In the second study, trialists compared ETN for seven years versus ETN for five years followed by two years of placebo.³⁴ BMD results measured at the spine favoured the continuation groups at the end of year seven.²¹ However, BMD results measured at the femoral neck, were non-statistically significant between the placebo group and the continuation group when measured at the end of year six.²¹ Similarly, BMD results measured at the distal radius, were non-statistically significant between the placebo group compared with the continuation group when measured at the end of year six,²¹ and again at the end of year seven.²¹

One additional primary study³⁰ in both the Nayak, 2019²¹ SR and Eriksen, 2014²² SR compared ALN 5 mg for 10 years versus ALN 10 mg for 10 years versus ALN 20 mg for two years followed by ALN 5 mg for three years followed by five years of placebo.³⁰ Nayak, 2019²¹ reported mean BMD percent change between years six to 10 and between years eight to 10. These results measured at the lumbar spine between years six to 10 as well as eight to 10 favoured the continuation groups,²¹ while those measured at the femoral neck were unfavourable in the placebo group but non significant in the ALN groups.²¹ However, Eriksen, 2014²² reported mean BMD percent change relative to pre-treatment levels for the lumbar spine, femoral neck, trochanter, and total hip, and results favoured the ALN 10 mg for 10 year group.²² The results of the placebo group were equal to, or better than, those of the ALN 5 mg for 10 year when looking at measurements of the lumbar spine, trochanter, and total hip.²²

Antifracture Efficacy

Information regarding the effect of duration and course of BIS treatment for osteoporosis on antifracture efficacy was available from two unique primary studies⁸^,³⁴ included in one SR.²¹

The first primary study compared ZLN for six years versus ZLN for three years plus three years of placebo.⁸ Odds ratio (OR) of fractures at the femoral neck (0.36; 95% CI, 0.15 to 0.77; P = 0.01) and hip (0.26; 95% CI, 0.08 to 0.69; P = 0.0113) favoured the six year treatment group compared with the placebo group.²¹ The second primary study compared ETN for seven years versus ETN for five years plus two years of placebo.³⁴ The number of morphometric vertebral fracture between the sixth and seventh year of the study favoured the continuation group (2.4%, n = 1/42) over the placebo group (10.2%, n = NR)).²¹

Bone turnover markers

Information regarding the effect of duration and course of BIS treatment for osteoporosis on bone turnover markers was available from one primary study³⁰ included in the Eriksen, 2014 SR.²² The primary study compared ALN 5 mg for 10 years versus ALN 10 mg for 10 years versus ALN 20 mg for two years followed by ALN 5 mg for three years followed by five years of placebo.³⁰ The authors reported that the ALN group (both dose groups combined) maintained a reduction in bone turnover markers, while the placebo group saw a “small increases in [bone turnover marker] levels (including [bone-specific alkaline phosphatase]) that were still below the pretreatment values at the end of the extension period”²² (p.131) (no effect estimates or statistics provided).

AE – Atypical Femoral Fractures

Information regarding the effect of duration and course of BIS treatment for osteoporosis on atypical femoral fractures (AFF) was available from five unique primary studies³⁵^–³⁹ included, with overlap, in two SRs.¹⁹^,²⁰

Three unique primary studies³⁶^,³⁷^,³⁹ in the Dennison, 2019¹⁹ SR discussed AFFs. The first was a combined cohort and case-control study of 59 cases of AFF and 263 controls (i.e., ordinary subtrochanteric or shaft fractures) who were also using BISs.³⁹ Dennison, 2019 reported that the risk of AFF after BIS discontinuation decreased by 70% per year since last use (OR = 0.28; 95% CI, 0.21 to 0.38).¹⁹ The second study was a cohort of 188,814 individuals, in an integrated health care provider (Kaiser Permanente) database, exposed to BIS.³⁶ Incidence of AFF after 0.1 to 0.9 years of exposure was 1.78/100,000 persons/year, while incidence after 8 to 9.9 years of exposure was 113.1/100,000 persons/year.³⁶ The third study was a SR of studies on AFF and other outcomes,³⁷ where Dennison, 2019¹⁹ reported an incidence of 3.0 to 9.8 AFFs per 100,000 person-years, and further described that the “relative risk increased with longer duration of [bisphosphonate] use, especially after more than 3 years […]”¹⁹ (p.1738)

One additional primary study³⁵ in the Dennison, 2019¹⁹ SR also had AFF results reported in the Fink, 2019²⁰ SR. The study was a combined cohort and case-control study of 172 cases of AFF and investigated their use of BISs.³⁵ Dennison, 2019¹⁹ reported an absolute risk (AR) of 11 (95% CI, 7 to 14) AFFs per 10,000 person-years of BIS use.¹⁹ After drug discontinuation, the risk decreased by 70% per year since the last use.¹⁹ Fink, 2019²⁰ further stratified these data and expresses them as OR after three to four years of use (40; 95% CI, 17 to 91), four to five years (116; 95% CI, 58 to 234), and after five years of use (93; 95% CI, 66 to 132).²⁰

One additional primary study³⁸ in the Fink, 2019²⁰ SR discussed AFFs. The primary study was a case-control comparing one or more year of current BIS use vs. past use.³⁸ The risk of AFFs with radiologic features was higher in current BIS users (HR = 3.36 [95% CI, 1.77 to 11.91] to 5.17 [95% CI, 2.0 to 13.36]).²⁰

AE – Other Fracture Risk

Information regarding the effect of duration and course of BIS treatment for osteoporosis on the risks relative to other types of fractures was available from nine unique primary studies²⁹^,³⁰^,³²^,³³^,⁴⁰^–⁴⁴ included, with overlap, in four SRs.¹⁹^–²²

Two unique primary studies⁴¹^,⁴² in the Dennison, 2019¹⁹ SR discussed risks relative to other types of fractures. The first was a cohort study of 160,369 women who had been established on BISs for three years with high refill compliance.⁴¹ The risk of hip fracture at a median follow-up of 2.7 years was significantly increased in the treatment interruption group compared with the continued user group (adjusted HR = 1.22; 95% CI, 1.11 to 1.34; P-value not reported), and was 1.8-fold increased after four years.¹⁹ The second study was a cohort study using the Swedish Prescribed Drug Register.⁴² Participants who persisted with therapy for greater than one year had a 60% lower risk (RR = 0.40; P = 0.001) of fracture during the first six months after discontinuation when compared with those who discontinue therapy within the first year.¹⁹

One additional primary study⁴⁰ in the Dennison, 2019¹⁹ SR also had results relative to other types of fractures reported in the Nayak, 2019²¹ and the Eriksen, 2014²² SRs. The study compared ALN 5mg for two years, then 10 mg for three years, then 5 mg for five years versus ALN 5 mg for two years, then 10 mg for eight years versus ALN 5 mg for two years, then 10 mg for three years, then placebo for five years.⁴⁰ Dennison, 2019¹⁹ reported a reduced nonvertebral fracture risk (values not reported), in those without a vertebral fracture at baseline, favouring the 10 year ALN continuation group versus the five year ALN continuation group.¹⁹ Nayak, 2019²¹ and Eriksen, 2014²² performed a different comparison, reporting nonvertebral fracture risks, in those without a vertebral fracture at baseline, favouring the 10 year ALN continuation group (both dose groups combined) compared with the placebo group (RR = 0.50; 95% CI, 0.26 to 0.96).²¹^,²² Nayak, 2019²¹ went on to report the morphometric vertebral fracture risks, in those without a vertebral fracture at baseline, as well as those with a vertebral fracture at baseline, also indicating in both cases that the results favoured the 10 year ALN continuation group (both dose groups combined) compared with the placebo group.²¹ However, in those with a vertebral fracture at baseline, the nonvertebral fracture risk, favoured the placebo group compared with the 10 year ALN continuation group (both dose groups combined) (RR = 1.11; 95% CI, 0.61 to 2.02).²¹ What is more, authors reported a clinical vertebral fracture risk favouring the 10 year ALN continuation group (both dose groups combined) compared with the placebo group (RR = 0.57; 95% CI, 0.23 to 1.40).²¹

Two primary study²⁹^,³⁰ in the Fink, 2019²⁰ SR also had results relative to other types of fractures reported in Nayak, 2019²¹ SR. The first primary study compared ALN 5 mg for seven years versus ALN 10 mg for seven years versus ALN 20 mg for two years, followed by 5 mg for three years, followed by two years of placebo.²⁹ Nayak, 2019²¹ reported the incidence of nonvertebral fracture between years six and seven as 6.6% (n = 8), 7.1% (n = 8), and 7.8% (n = 9) for the ALN 10 mg, 5mg, and placebo groups respectively,²¹ while Fink, 2019²⁰ reported no between-group difference (both dose groups combined vs. placebo group) with a RR = 0.87 (95% CI, 0.40 to 1.91).²⁰ Nayak, 2019 reported that incidence of clinical vertebral fractures between years six and seven favoured the ALN 5 mg group,²¹ while Fink, 2019²⁰ reported no between-group difference (both dose groups combined vs. placebo group) with a RR = 0.92 (95% CI, 0.40 to 2.10).²⁰ The second primary study compared ALN 5 mg for 10 years versus ALN 10 mg for 10 years versus ALN 20 mg for two years, followed by ALN 5 mg for three years, followed by five years of placebo.³⁰ Authors of Nayak, 2019²¹ reported rates of first nonvertebral fracture between years eight and 10 favouring the ALN 10 mg group,²¹ while Fink, 2019²⁰ reported no between-group differences (both dose groups combined vs. placebo group) with a RR = 0.81 (95% CI, 0.38 to 1.71).²⁰ As for morphometric vertebral fractures, Nayak, 2019²¹ reported a non significant between-group difference (no effect estimate or P-value reported) between years six and 10.²¹ Lastly, Fink, 2019²⁰ reported no between-group differences (both dose groups combined vs. placebo group) with a RR = 1.40 (95% CI, 0.52 to 3.74) for radiographic vertebral fractures.²⁰

One additional primary study⁴³ in Fink, 2019²⁰ discussed risks relative to other types of fractures. The primary study was a case-control of 1,855 women aged 68 years or older and compared one or more year of current BIS use vs. past use.⁴³ There was a higher risk of subtrochanteric or femoral shaft fracture in those with greater than five years of use (OR = 2.74; 95% CI, 1.25 to 6.02) compared past users.²⁰

Two unique primary studies³²^,³³ in the Nayak, 2019²¹ SR discussed risks relative to other types of fractures. The first primary study compared a BIS holiday versus persistent (i.e., therapy adherence of 50% or more) users.³² Adjusted hazard ratios (HR) for hip fractures, clinical vertebral fractures, and any osteoporotic fracture were no different between the BIS holiday group and the persistent use group.²¹ The second primary study compared three years of BIS continuation versus a three year holiday in participants who had three to five years of BIS treatment.³³ The HR for clinical fractures (1.40; 95% CI, 1.12 to 1.60; P = 0.0095) favoured the continuation group (11.9%; n = 16/135) compared with the drug holiday group (16.1%; n = 5/31).²¹ Nayak, 2019²¹ also conducted a MA of fracture risks as reported in the two aforementioned studies³²^,³³ and reported a non significant adjusted HR of any clinical osteoporotic fracture to be 1.13 (95% CI, 0.75 to 1.70) between individuals who discontinued BISs compared with those who continued therapy.²¹ However, these results were in the presence of high statistical heterogeneity (I² = 94.3%).²¹

One unique primary study⁴⁴ in the Eriksen, 2014²² SR reported a link between new morphometric vertebral fractures experienced by participants in the three years that followed a three year course of ZLN treatment and the participant’s baseline BMD t-score and occurrence of morphometric vertebral fracture during the three year course of ZLN treatment.²²

Other AEs

Information regarding the effect of treatment duration and courses of BISs for osteoporosis on other AEs was available from two unique primary studies²⁹^,³⁰ included, with overlap, in two SRs.²⁰^,²²

Two primary study²⁹^,³⁰ in the Fink, 2019²⁰ SR also had results relative to other types of AEs reported in Eriksen, 2014²² SR. The first primary study compared ALN 5 mg for seven years versus ALN 10 mg for seven years versus ALN 20 mg for two years, followed by 5 mg for three years, followed by two years of placebo.²⁹ Fink, 2019²⁰ reported no difference in risk for serious AEs (not defined), between the continuation groups (doses combined) and the placebo group, RR = 1.05 (95% CI, 0.57 to 1.96). Eriksen, 2014²² went on to specify that there were no significant difference (P-value not reported) between groups in the incidence of AEs, upper gastrointestinal AEs, and placebo due to AEs.²² The second primary study compared ALN 5 mg for 10 years versus ALN 10 mg for 10 years versus ALN 20 mg for two years followed by ALN 5 mg for three years followed by five years of placebo.³⁰ Fink, 2019²⁰ reported no difference in risk for serious AEs (not defined), between continuation (doses combined) and the placebo group, RR = 1.21 (95% CI, 0.75 to 1.96). Here too, authors of Eriksen, 2014²² went on to specify that there were no significant difference (P-value not reported) between groups in the incidence of AEs, upper gastrointestinal AEs, and discontinuation due to AEs.²² However, they reported four deaths (none considered associated with ALN treatment) in the ALN continuation groups (doses combined) compared with none in the placebo group.²²

Evidence-Based Guidelines Regarding Length of Treatment with Bisphosphonates for Osteoporosis

Six guidelines⁶^,²³^–²⁷ were identified regarding recommendations on the length of treatment with BISs for osteoporosis.

Duration of treatment

The first from the ACP, weakly recommends (based on low-quality evidence) a five year duration of treatment;²³ however, this recommendation is not specific regarding the type of pharmacological treatment.

The EMAS guidelines recommend (strength of evidence and recommendations not reported) the need to individualise treatment, taking into consideration the long-term efficacy of BISs, their safety, and the fracture risk of the patient.²⁴ Authors highlight that discontinuation of BISs should be considered in patients who have been treated for more than five years with ALN or more than three years with RSN, or ZLN (strength of evidence and recommendations not reported).²⁴

The NOGG guidelines differ in their recommendations on treatment duration, recommending (grade C) that a review should be performed at three years with ZLN and 5 years with oral BISs (e.g., ALN, IBN, RSN).²⁵ A decision on treatment continuation beyond three years with ZLN and 5 years with oral BISs, can generally be recommended (evidence level IIb, grade of recommendation B) in PM “individuals age ≥75 years, those with a history of hip or vertebral fracture, those who sustain a fracture while on treatment, and those taking oral glucocorticoids”.²⁵ (p.2) Authors further highlight that “There is no evidence to guide decisions beyond 10 years of treatment and management options in such patients should be considered on an individual basis”.²⁵ (p.2)

The RACGP guidelines recommend (grade D) that clinicians reconsider the need for BIS therapy after five to 10 years in PM women and men over 50 years of age with osteoporosis, when their BMD t-score is −2.5 or higher and there are no recent fractures.²⁶ If this BMD t-score threshold is not met or if there are new vertebral fractures, authors recommend (grade D) treatment continuation.²⁶

The SEIOMM guidelines do not indicate a particular treatment duration and recommend (grade D) that “treatment should last as long as necessary to decrease the risk of fractures to acceptable levels”.²⁷ (p.521) Authors suggest that a BMD t-score above −2.5 standard deviations and no recent fractures, or a BMD t-score above −2 standard deviations and one previous fracture could constitute an “acceptable level”.²⁷ They recommend (grade D) reassessing patients every three to five years and discontinue treatment once this “acceptable level” is achieved.²⁷

The WRHA guidelines recommend (grade D) that osteoporosis therapy should continue in individuals at high risk for fracture.⁶ While they recommend (strength of evidence and recommendations not reported) that individuals at moderate risk should have their osteoporosis therapy reassess every three to five years for the need to continue, discontinue, or initiate a drug holiday.⁶

Duration of drug holiday

The EMAS guidelines highlight the need to re-evaluate patients one to three years after BIS discontinuation and that the decision to resume treatment ought to depend on a reassessment of risk factors, new fractures, and possibly BMD (strength of evidence and recommendations not reported).²⁴

Similarly, the NOGG guidelines recommend (grade C) the need to re-evaluate patients where treatment has been discontinued, either after 18 months to three years, or after a new fracture.²⁵ Authors recommend (grade B) a reassessment of fracture risk using the FRAX tool or with femoral neck BMD. Also, if biochemical markers indicate a relapse from suppressed bone turnover and BMD has decreased following withdrawal, authors recommend (grade C) that resumption of treatment should be considered.²⁵

The RACGP guidelines recommend (grade D) that treatment should be restarted if there is evidence of bone loss, especially at the hip, or if a further minimal trauma fracture is sustained.²⁶ The re-assessment of these parameters are recommended (grade B) to be performed on an regular basis (frequency not specified).²⁶

Limitations

A number of limitations were identified in the critical appraisal as shown in Appendix 3, Table 4 and Table 5; however, additional limitations exist. The main limitations of this review are related to the heterogeneity of the study populations and the generalizability of the findings. Heterogeneity was apparent in the baseline patient characteristics, such as BMD, past fracture history, and past use of osteoporosis medication. Also, because SR data were often available only at the study level, it remains unclear whether any differences between outcomes were due to differences in population characteristics.

Eight of the 17 included primary studies were cohort or case-control studies,³²^,³³^,³⁵^–³⁷^,³⁹^,⁴²^,⁴³ with associated methodological limitations. Furthermore, some primary studies may have been too small to examine uncommon clinical events, as noted by the wide confidence intervals for some endpoints. In addition, there was an overall lack of comparative statistical analyses between treatment groups, and when performed, were usually done post-hoc.

Except for primary studies that queried databases for prescription refill statistics,⁴¹^,⁴² participant’s adherence with treatment was not reported which introduces uncertainty with regards to the magnitude of effects reported in the SRs.

Two SRs¹⁹^,²² did not perform a quality assessment of their included studies, which limits the overall reliability of their results. While the other two SRs²⁰^,²¹ reported that studies relevant to this report had a low or unclear risk of bias, introducing uncertainty in their results.

At least one SR²⁰ inferred fracture risk data from studies not designed to compared fracture incidence between groups. Their results on between-group relative risk differences of fractures may lead to incorrect conclusions.

All SRs¹⁹^–²² reported results without providing some, or all, associated P-values (where appropriate), which may have introduced an outcomes reporting bias, limiting the overall reliability of their results.

The applicability of the evidence to the Canadian setting is unclear since the country of origin of SR primary studies was inconsistently reported and only one evidence-based guideline was developed in Canada.

Guideline recommendations were generally based on low-quality evidence,²³^,²⁵^–²⁷ or did not consistently report the strength of evidence and recommendations.⁶^,²⁴

Although two SRs²⁰^,²¹ included males in their population eligibility criteria, one included primary studies³⁵ reported on this population; therefore, generalizability of results in males is unclear. This also suggests that additional research in this population is required.

Although data were identified regarding when to initiate a drug holiday, there was no clear evidence that emerged from the literature on the optimal duration of a drug holiday, suggesting that additional research in this area is required.

Conclusions and Implications for Decision or Policy Making

This report identified clinical evidence and evidence-based guidelines regarding treatment durations and courses of BISs for osteoporosis. One HTA,¹⁸ four SRs,¹⁹^–²² including one with MA,²⁰ and six guidelines⁶^,²³^–²⁷ were identified and included in this review. The HTA,¹⁸ met the inclusion criteria for this report; however, none of the primary studies included in the HTA met the eligibility criteria for this report; therefore, no summary could be provided.

The identified literature were heterogenous and revealed mixed conclusions regarding clinical evidence of treatment duration and courses of BISs for osteoporosis. No clear direction emerged regarding BMD results, most of which were non-statistically significant and between group comparisons were difficult because of heterogeneity in BIS used, the location of BMD measurements, the time point of the measurement, the duration of use, and participant’s history of fracture at baseline.¹⁹^,²¹^,²² Antifracture efficacy results favoured the drug continuation group for six years (ZLN) or seven (ETN).²¹ AFF results came mostly from cohort studies, which all concluded that an increased incidence occurred with increased duration of BIS use;¹⁹^,²⁰ however, the risk decreased by 70% per year after discontinuation.¹⁹ No clear direction emerged regarding risks related to other types of fractures, which favoured the drug discontinuation group or continuation group depending on the BIS used, the dose, the duration of use, the participant’s history of fracture at baseline, and the type of fracture sustained.¹⁹^–²² Other types of AEs were generally reported as no different between the drug continuation groups and discontinuation groups,²⁰^,²² except for four deaths (none considered associated with treatment) in a continuation group compared with none in the placebo group.²²

Evidence-based guidelines offer no clear direction on duration of BIS treatment. Two guidelines recommend a three to five year duration of treatment,²³^,²⁴ one recommends five to 10 years,²⁶ while others suggest an indeterminate period.⁶^,²⁵^,²⁷ However, most guidelines nuance their recommendations by advocating for regular patient monitoring and the individualisation of treatment in response to clinical and paraclinical manifestations.⁶^,¹⁰^,²³^–²⁷ Similarly, they offer no clear direction on duration of BIS drug holidays, but rather to base the decision on a periodic reassessment of the patient.²⁴^–²⁶

The limitations of the included studies should be considered when interpreting the results. The findings highlighted in this review come with a high degree of uncertainty. Further research investigating the clinical evidence of treatment duration and courses of BISs for osteoporosis, especially by way of large, methodologically-sound RCTs or well-designed meta-analyses, would help reduce this uncertainty.

References

1.: Raisz L. Chapter 37: Osteoporosis. In: Porter RS, Kaplan JL, eds. The Merck Manual of Diagnosis and Therapy. 19th ed. Whitehouse Station (NJ): Merck Sharp & Dohme Corp.; 2011.
2.: Chapter 44: Agents affecting mineral ion homeostatis and bone turnover. In: Hilal-Dandan R, Brunton LL, eds. Goodman and Gilman’s Manual of Pharmacology and Therapeutics. 2nd ed. New York (NY): McGraw-Hill Medical; 2014.
3.: O’Connell MB, Borchert JS. Chapter 73: Osteoporosis and other metabolic bone diseases. In: DiPiro JT, Talbert RL, Yee GC, Matzke GR, Wells BG, Posey LM, eds. Pharmacotherapy: A Pathophysiologic Approach. 9th ed. New York (NY): McGraw-Hill Education; 2014.
4.: Tarride J-E, Hopkins RB, Leslie WD, et al. The burden of illness of osteoporosis in Canada. Osteoporos Int. 2012;23(11):2591–2600. [PMC free article: PMC3483095] [PubMed: 22398854]
5.: Hanley DA. Chapter 83: Osteoporosis. Compendium of Therapeutic Choices. Ottawa (ON): Canadian Pharmacists Association; 2014.
6.: Anderson K, Meade M, Anthony L, et al. Clinical practice guidelines for the diagnosis and management of osteoporosis. Winnipeg (MB): Winnipeg Regional Health Authority; 2014: https://www.wrha.mb.ca/extranet/eipt/files/EIPT-031-001.pdf. Accessed 2019 Sep 23.
7.: Black DM, Schwartz AV, Ensrud KE, et al. Effects of continuing or stopping alendronate after 5 years of treatment the fracture intervention trial long-term extension (FLEX): a randomized trial. JAMA. 2006;296(24):2927–2938. [PubMed: 17190893]
8.: Black DM, Reid IR, Boonen S, et al. The effect of 3 versus 6 years of zoledronic acid treatment of osteoporosis: A randomized extension to the HORIZON-Pivotal Fracture Trial (PFT). J Bone Miner Res. 2012;27(2):243–254. [PMC free article: PMC3427916] [PubMed: 22161728]
9.: Watts N, Chines A, Olszynski W, et al. Fracture risk remains reduced one year after discontinuation of risedronate. Osteoporos Int. 2008;19(3):365–372. [PubMed: 17938986]
10.: Papaioannou A, Morin S, Cheung A, et al. Clinical practice guidelines for the diagnosis and management of osteoporosis in Canada: Background and technical report. Toronto (ON): Osteoporosis Canada; 2010: https://osteoporosis.ca/wp-content/uploads/Osteoporosis_Guidelines_2010_Background_And_Technical_Report.pdf. Accessed 2019 Sep 23.
11.: Watts N, Bilezikian J, Camacho P, et al. American Association of Clinical Endocrinologists medical guidelines for clinical practice for the diagnosis and treatment of postmenopausal osteoporosis. Endocr Pract. 2010;16(Supplement 3):1–37. [PMC free article: PMC4876714] [PubMed: 21224201]
12.: Watts NB, Diab DL. Long-term use of bisphosphonates in osteoporosis. J Clin Endocrinol Metab. 2010;95(4):1555–1565. [PubMed: 20173017]
13.: Bisphosphonates for the prevention of osteoporosis in patients treated with systematic corticosteroids: a review of the clinical evidence and guidelines (CADTH Rapid response report: summary with critical appraisal). Ottawa (ON): CADTH; 2011: https://cadth.ca/sites/default/files/pdf/htis/aug-2011/RC0293_bispho_final.pdf. Accessed 2019 Sep 25.
14.: Bisphosphonate use in patients with osteopenia and low risk of fractures: clinical effectiveness and guidelines (CADTH Rapid response report: summary of abstracts). Ottawa (ON): CADTH; 2015: https://cadth.ca/bisphosphonate-use-patients-osteopenia-and-low-risk-fractures-clinical-effectiveness-and-guidelines. Accessed 2019 Sep 25.
15.: Duration of bisphosphonate treatment for patients with osteoporosis: clinical evidence and guidelines (CADTH Rapid response report: reference list). Ottawa (ON): CADTH; 2019: https://cadth.ca/duration-bisphosphonate-treatment-patients-osteoporosis-clinical-evidence-and-guidelines. Accessed 2019 Sep 25.
16.: Shea BJ, Reeves BC, Wells G, et al. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ. 2017;358:j4008. [PMC free article: PMC5833365] [PubMed: 28935701]
17.: AGREE Next Steps Consortium. The Appraisal of Guidelines for Research Evaluation II (AGREE II) Instrument. Hamilton, ON: AGREE Enterprise; 2017: https://www.agreetrust.org/wpcontent/uploads/2017/12/AGREE-II-Users-Manual-and-23-item-Instrument-2009-Update-2017.pdf. Accessed 2019 Sep 25.
18.: Davis S, Martyn-St James M, Sanderson J, et al. A systematic review and economic evaluation of bisphosphonates for the prevention of fragility fractures. Health Technol Assess. 2016;20(78):1–406. [PMC free article: PMC5107884] [PubMed: 27801641]
19.: Dennison EM, Cooper C, Kanis JA, et al. Fracture risk following intermission of osteoporosis therapy. Osteoporos Int. 2019;30(9):1733–1743. [PubMed: 31175404]
20.: Fink HA, MacDonald R, Forte ML, et al. Long-term drug therapy and drug discontinuations and holidays for osteoporosis fracture prevention: a systematic review. Ann Intern Med. 2019;171(1):37–50. [PubMed: 31009947]
21.: Nayak S, Greenspan SL. A systematic review and meta-analysis of the effect of bisphosphonate drug holidays on bone mineral density and osteoporotic fracture risk. Osteoporos Int. 2019;30(4):705–720. [PMC free article: PMC6499675] [PubMed: 30623214]
22.: Eriksen EF, Diez-Perez A, Boonen S. Update on long-term treatment with bisphosphonates for postmenopausal osteoporosis: a systematic review. Bone. 2014;58:126–135. [PubMed: 24120384]
23.: Qaseem A, Forciea MA, McLean RM, Denberg TD, for the Clinical Guidelines Committee of the American College of Physicians. Treatment of low bone density or osteoporosis to prevent fractures in men and women: a clinical practice guideline update from the American College of Physicians. Ann Intern Med. 2017;166(11):818–839. [PubMed: 28492856]
24.: Anagnostis P, Paschou SA, Mintziori G, et al. Drug holidays from bisphosphonates and denosumab in postmenopausal osteoporosis: EMAS position statement. Maturitas. 2017;101:23–30. [PubMed: 28539165]
25.: Compston J, Cooper A, Cooper C, et al. UK clinical guideline for the prevention and treatment of osteoporosis. Arch Osteoporos. 2017;12(1):43. [PMC free article: PMC5397452] [PubMed: 28425085]
26.: The Royal Australian College of General Practitioners, Osteoporosis Australia. Osteoporosis prevention, diagnosis and management in postmenopausal women and men over 50 years of age. 2nd ed. East Melbourne, Victotria (AU): Royal Australian College of General Practitioners; 2017: https://www.racgp.org.au/FSDEDEV/media/documents/Clinical%20Resources/Guidelines/Joint%20replacement/Osteoporosis-guidelines.pdf. Accessed 2019 Sep 23.
27.: González-Macias J, Del Pino-Montes J, Olmos JM, Nogués X, on behalf of the Comisión de Redacción de las Guías de Osteoporosis de la SEIOMM. Clinical practice guidelines for posmenopausal, glucocorticoid-induced and male osteoporosis. Spanish Society for Research on Bone and Mineral Metabolism (3rd updated version 2014). Rev Clin Esp. 2015;215(9):515–526. [PubMed: 26434811]
28.: Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol. 2009;62(10):e1–e34. [PubMed: 19631507]
29.: Tonino RP, Meunier PJ, Emkey R, et al. Skeletal benefits of alendronate: 7-year treatment of postmenopausal osteoporotic women. Phase III osteoporosiss treatment study group. J Clin Endocrinol Metab. 2000;85(9):3109–3115. [PubMed: 10999794]
30.: Bone HG, Hosking D, Devogelaer J-P, et al. Ten years’ experience with alendronate for osteoporosis in postmenopausal women. N Engl J Med. 2004;350(12):1189–1199. [PubMed: 15028823]
31.: Jackson D, Bowden J, Baker R. How does the DerSimonian and Laird procedure for random effects meta-analysis compare with its more efficient but harder to compute counterparts? J Stat Plan Inference. 2010;140(4):961–970.
32.: Adams AL, Adams JL, Raebel MA, et al. Bisphosphonate drug holiday and fracture risk: a population-based cohort study. J Bone Miner Res. 2018;33(7):1252–1259. [PubMed: 29529334]
33.: Mignot MA, Taisne N, Legroux I, Cortet B, Paccou J. Bisphosphonate drug holidays in postmenopausal osteoporosis: effect on clinical fracture risk. Osteoporos Int. 2017;28(12):3431–3438. [PubMed: 28875236]
34.: Miller PD, Watts NB, Licata AA, et al. Cyclical etidronate in the treatment of postmenopausal osteoporosis: efficacy and safety after seven years of treatment. The American Journal of Medicine. 1997;103(6):468–476. [PubMed: 9428829]
35.: Schilcher J, Koeppen V, Aspenberg P, Michaëlsson K. Risk of atypical femoral fracture during and after bisphosphonate use. Acta Orthop. 2015;86(1):100–107. [PMC free article: PMC4366670] [PubMed: 25582459]
36.: Dell RM, Adams AL, Greene DF, et al. Incidence of atypical nontraumatic diaphyseal fractures of the femur. J Bone Miner Res. 2012;27(12):2544–2550. [PubMed: 22836783]
37.: Khow KSF, Shibu P, Yu SCY, Chehade MJ, Visvanathan R. Epidemiology and postoperative outcomes of atypical femoral fractures in older adults: A systematic review. J Nutr Health Aging. 2017;21(1):83–91. [PubMed: 27999854]
38.: Koh JH, Myong JP, Yoo J, et al. Predisposing factors associated with atypical femur fracture among postmenopausal Korean women receiving bisphosphonate therapy: 8 years’ experience in a single center. Osteoporos Int. 2017;28(11):3251–3259. [PubMed: 28748389]
39.: Schilcher J, Michaëlsson K, Aspenberg P. Bisphosphonate use and atypical fractures of the femoral shaft. N Engl J Med. 2011;364(18):1728–1737. [PubMed: 21542743]
40.: Schwartz AV, Bauer DC, Cummings SR, et al. Efficacy of continued alendronate for fractures in women with and without prevalent vertebral fracture: The FLEX Trial. J Bone Miner Res. 2010;25(5):976–982. [PubMed: 20200926]
41.: Curtis J, Chen R, Li Z, et al. The impact of bisphosphonate drug holidays on fracture rates. Paper presented at: American Society for Bone and Mineral Research (ASBMR) Annual Meeting. 2018; Orlando (FL).
42.: Ström O, Landfeldt E, Garellick G. Residual effect after oral bisphosphonate treatment and healthy adherer effects—the Swedish adherence register analysis (SARA). Osteoporos Int. 2015;26(1):315–325. [PubMed: 25297890]
43.: Park-Wyllie LY, Mamdani MM, Juurlink DN, et al. Bisphosphonate use and the risk of subtrochanteric or femoral shaft fractures in older women. JAMA. 2011;305(8):783–789. [PubMed: 21343577]
44.: Cosman F, Cauley J, Eastell R, et al. Who is at highest risk for new vertebral fractures after 3 years of annual zoledronic acid and who should remain on treatment [conference abstract 1248]. J Bone Miner Res. 2011;26:(Suppl 1):S79.
45.: MacLean C, Newberry S, Maglione M, et al. Systematic review: comparative effectiveness of treatments to prevent fractures in men and women with low bone density or osteoporosis. Ann Intern Med. 2008;148(3):197–213. [PubMed: 18087050]
46.: Brouwers MC, Kho ME, Browman GP, et al. AGREE II: advancing guideline development, reporting and evaluation in health care. Can Med Assoc J. 2010;182(18):E839–E842. [PMC free article: PMC3001530] [PubMed: 20603348]

Appendix 1. Selection of Included Studies

Appendix 2. Characteristics of Included Publications

Table 2Characteristics of Included Health Technology Assessment, Systematic Reviews, and Meta-Analysis

First Author, Publication Year, Country

Study Designs and Numbers of Primary Studies Included

Population Characteristics

Intervention and Comparator(s)

Clinical Outcomes, Length of Follow-Up

Health Technology Assessment

Davis, 2016¹⁸

United Kingdom

Study design: SR including NMA of relevant RCTs and NRSs

Literature search strategy: Authors performed literature searches in several databases including: MEDLINE, EMBASE, the Cochrane Library, CINAHL, Web of Science, BIOSIS preview, Clinicaltrials.gov, and WHO International Clinical Trials Registry Platform from 2008 until September 2014. No limiters were applied to study retrieval.

Number of studies included: In total, 59 articles were included; however, none were relevant to this report.

Quality assessment tool: The Cochrane Risk of Bias Tool.

Objective: To evaluate the clinical effectiveness, safety, and cost-effectiveness of BISs for the prevention of fragility fractures.

Women 65 years of age or older or men 75 years of age or older. Alternatively, study participants could be younger if they also presented with certain risk factors.

Interventions:

BISs

Comparators:

BISs compared to each other
placebo
Other non-active treatment

Relevant Outcomes:

Fragility fractures
BMD at the femoral neck
Mortality
AEs
Compliance
HRQoL
Health care resource use

Follow-up: not applicable

Systematic Reviews and Meta-Analysis

Dennison, 2019¹⁹

United Kingdom

Study design: SR of relevant RCTs and observational studies.

Literature search strategy: Authors performed literature searches in PubMed, EMBASE, Cochrane library, NHS Evidence, Epistemonikos, and NIH trial registry records for the period of 2011 to 2016. This was supplemented by a search of relevant conference abstracts for 2017 to 2018.

Number of studies included: In total, 38 articles were included, with nine relevant for this review: a secondary analysis⁴⁰ of the FLEX RCT study, a SR,³⁷ cohort studies,³³^,³⁵^,³⁶^,³⁹^,⁴¹^,⁴² and case-control studies.³⁵^,3⁹

Note that Mignot, 2017³³ and Schwartz, 2010⁴⁰ are also included in the Nayak, 2019²¹ SR herein. Similarly, Schilcher, 2015³⁵ is also included in the Fink, 2019²⁰ SR herein, and Schwartz, 2010⁴⁰ is also included in the Eriksen, 2014²² SR herein.

Quality assessment tool: NR

Objective: To review the clinical evidence on osteoporosis therapy duration and decision making around continuation or discontinuation of therapy.

PM women on osteoporosis medication for 1 y or more.

Included studies:

A secondary analysis⁴⁰ of the FLEX RCT study: PM women 61 to 86 y old, previously treated with 4 or 5 y of ALN.
SR: 23 studies on AFF,14 on epidemiology, and 11 on treatment outcomes.³⁷
Cohort study: 160,369 women who had been established on BISs for 3 y with high refill compliance.⁴¹
Cohort study: using the Swedish Prescribed Drug Register.⁴²
Cohort study: 188,814 individuals in Kaiser Permanente database exposed to BIS: 142 cases of AFF³⁶
A combined cohort and case-control study: 59 cases of AFF and 263 controls, while on BIS therapy.³⁹
A combined cohort and case-control study: 172 cases.³⁵
Cohort study: 183 PM women with osteoporosis who continued or discontinued BISs for 3 to 5 y.³³

Interventions:

BISs (i.e., ALN, RSN, IBN, ZLN) or denosumab

Comparators: Medication continuation versus discontinuation

Relevant Outcomes:

Bone turnover markers (types not specified)
BMD
Fracture
Osteonecrosis of the jaw
AFF

Follow-up: NR

Fink, 2019²⁰

United States of America

Study design: SR with MA of relevant trials and observational studies.

Literature search strategy: Authors performed literature searches in MEDLINE, EMBASE, and the Cochrane library for the period of Jan 1995 to Oct 2018. This was supplemented by a hand search of bibliographies and a search of trial registry records. Results were limited to English language studies.

Number of studies included: In total, 48 articles were included, with 5 relevant for this review: 2 extensions²⁹^,³⁰ of a phase III RCT, a combination cohort case-control study,³⁵ and two case-control studies.³⁸^,⁴³

Note that Bone, 2004³⁰ and Tonino, 2000²⁹ are also included in the Nayak, 2019²¹ SR herein. Similarly, Schilcher, 2015³⁵ is included in the Dennison, 2019¹⁹ SR herein.

Quality assessment tool: Risk of bias and strength of the evidence were assessed using criteria from the AHRQ.

Objective: To review the effects of long-term osteoporosis drug treatment, discontinuation, and holidays.

Men or PM women aged 50 y or older who were being investigated or treated for fracture prevention.

Included studies:

Two extensions of a phase III RCT, the first in 350 PM women with osteoporosis,²⁹ and the second in 247 PM women with osteoporosis.³⁰
A combined cohort and case-control study: 172 cases.³⁵
Case-control: 172 PM women.³⁸
Case-control: 1,855 women aged 68 y or older.⁴³

Interventions:

Long-term (i.e., 3 or more y) osteoporosis drug treatments

Comparators:

Control,
Continuation versus discontinuation (cessation for 1 or more y after 1 or more y of use)

Included studies:

Two extensions of a phase III RCT comparing ALN 5 mg or 10 mg for 7 y²⁹ or 10 y³⁰ vs. ALN 20 mg for 2 y, then 5 mg for 3 y, then placebo for 2 y²⁹ or 5 y.³⁰
Cohort and case-control comparing BIS vs. no BIS.³⁵
Case-control comparing one or more year of current BIS use vs. past use.³⁸
Case-control comparing one or more year of current BIS use vs. past use.⁴³

Relevant Outcomes:

Incidence of clinical fractures
Incidence of radiographic fractures
DXA BMD
AEs

Follow-up: 3 or more years of treatment, 1 or more year of discontinuation after one or more year of use.

Nayak, 2019²¹

United States of America

Study design: SR of relevant clinical studies, controlled trials, and cohort studies.

Literature search strategy: Authors performed literature searches in PubMed, EMBASE, and the Cochrane Library up to January 2018. This was supplemented by a manual search of the reference lists of included studies. No limiters were applied to study retrieval.

Number of studies included: In total, 13 articles were included, with 7 relevant for this review: An extension⁸ of the HORIZON-PFT study, a secondary analysis⁴⁰ of the FLEX RCT study, 2 extensions²⁹^,³⁰ of a phase III RCT, an extension study³⁴ of an RCT, and cohort studies.³²^,³³

Note that Schwartz, 2010,⁴⁰ Bone, 2004,³⁰ and Tonino, 2000²⁹ are also included in the Eriksen, 2014²² SR herein. Similarly, Bone, 2004,³⁰ and Tonino, 2000²⁹ are also included in the Fink, 2019²⁰ SR herein, as well as Mignot, 2017³³ and Schwartz, 2010⁴⁰ are also included in the Dennison, 2019¹⁹ SR.

Quality assessment tool: for clinical trials: the Cochrane Collaboration’s risk of bias tool. For cohort studies: the Newcastle-Ottawa quality assessment scale.

Objective: To summarize the evidence on the effect of osteoporosis treatment drug holidays or discontinuation on BMD and fracture risk.

Participants with osteoporosis or osteopenia who had received osteoporosis treatment for at least 3 y.

Included studies:

RCT: 1,233 PM women who previously received ZLN for 3 y for osteoporosis.⁸
Secondary analysis⁴⁰ of the FLEX RCT study: 1,099 PM women 61 to 86 y old, previously treated with 4 or 5 y of ALN.
Two extensions of a phase III RCT, the first in 350 PM women with osteoporosis,²⁹ and the second in 247 PM women with osteoporosis.³⁰
An extension study of an RCT of 193 PM women³⁴
Cohort: 39,502 women aged 45 y or older with 3 or more y of prior BIS use with 50% or greater adherence.³²
Cohort: 183 PM women with osteoporosis who continued or discontinued BISs for 3 to 5 y.³³

Interventions:

Continue treatment

Comparators:

Discontinue treatment (drug holiday)

Included studies:

RCT comparing ZLN for 6 y vs. ZLN for 3 y plus 3 y of placebo,⁸
A secondary analysis⁴⁰ of the FLEX RCT study, which compared continuing ALN (various doses) after 5 y vs. discontinuing (placebo).
Two extensions of a phase III RCT comparing ALN 5 mg or 10 mg for 7 y²⁹ or 10 y³⁰ vs. ALN 20 mg for 2 y, then 5 mg for 3 y, then placebo for 2 y²⁹ or 5 y.³⁰
An extension study of an RCT of ETN or placebo for 3 y, then all participants received ETN for years 4 and 5, then a re-randomization to ETN or placebo for years 6 and 7.³⁴
Cohort comparing BIS holiday (average of 3.1 y long after an average of 5.2 y of prior use) vs. persistent users (i.e., 50% or greater adherence) and non-persistent users (i.e., less than 50% adherence).³²
Cohort comparing continuation or discontinuation of BISs (44% ALN, 40% RSN, 11% ZLN, 5% IBN).³³

Relevant Outcomes:

BMD
Fracture risk

Follow-up: Not applicable

Eriksen, 2014²²

Norway

Study design: SR of relevant controlled clinical trials.

Literature search strategy: Authors performed a literature search in PubMed up to January 2013.

Number of studies included: In total, 9 articles were included, with 4 relevant for this review: two extensions¹³^,¹⁴ of a phase III RCT,²⁹^,³⁰ a secondary analysis⁴⁰ of the FLEX RCT study, and a secondary analysis⁴⁴ of the HORIZON-PFT study.

Note that Schwartz, 2010,⁴⁰ Bone, 2004,³⁰ and Tonino, 2000²⁹ are also included in the Nayak, 2019²¹ SR herein. Similarly, Bone, 2004,³⁰ and Tonino, 2000²⁹ are also included in the Fink, 2019²⁰ SR herein, and Schwartz, 2010⁴⁰ is also included in the Dennison, 2019¹⁹ SR.

Quality assessment tool: NR

Objective: To summarize the long-term data relating to BISs in PM women with osteoporosis.

Participants were PM women with osteoporosis

Included studies:

Two extensions of a phase III RCT, the first in 350 PM women with osteoporosis,²⁹ and the second in 247 PM women with osteoporosis.³⁰
A secondary analysis⁴⁰ of the FLEX RCT study: 1,099 PM women 61 to 86 y old, previously treated with 4 or 5 y of ALN.
A secondary analysis of the HORIZON-PFT study: 1,223 women who had already received 3 y of ZLN.⁴⁴

Interventions: Continued treatment with either:

Comparators:

Discontinuation of treatment

Included studies:

Two extensions of a phase III RCT comparing ALN 5 mg or 10 mg for 7 y²⁹ or 10 y³⁰ vs. ALN 20 mg for 2 y, then 5 mg for 3 y, then placebo for 2 y²⁹ or 5 y.³⁰
A secondary analysis⁴⁰ of the FLEX RCT study, which compared continuing ALN (various doses) after 5 y vs. discontinuing (placebo).
A secondary analysis of the HORIZON-PFT study which compared 6 y of ZLN vs. 3 y of ZLN plus 3 y of placebo.⁴⁴

Relevant Outcomes:

BMD
Fracture
AEs

Follow-up: Not applicable

: AE = adverse event; AFF = atypical femoral fractures; AHRQ = Agency for Healthcare Research and Quality; ALN = alendronate; BIOSIS = BioSciences Information Service of Biological Abstracts; BIS = bisphosphonate; BMD = bone mineral density; CINAHL = Cumulative Index to Nursing and Allied Health Literature; DXA = dualenergy x-ray absorptiometry; EMBASE = Excerpta Medica database; ETN = etidronate; FLEX = Fracture Interventional Trial Long Term Extension; HORIZON-PFT = Health Outcomes and Reduced Incidence with Zoledronate One Yearly – Pivotal Fracture Trial; HRQoL = health related quality of life; IBN = ibandronate; MEDLINE = Medical Literature Analysis and Retrieval System Online; NHS = National Health Service; NIH = National Institutes of Health; NMA = network meta-analysis; NR = not reported; PM = postmenopausal; PubMed = public MEDLINE; RCT = randomized controlled trial; RSN = risedronate; SR = systematic review; WHO = World Health Organisation; ZLN = zoledronic acid.

Table 3Characteristics of Included Guidelines

Intended Users, Target Population	Intervention and Practice Considered	Major Outcomes Considered	Evidence Collection, Selection, and Synthesis	Evidence Quality Assessment	Recommendations Development and Evaluation	Guideline Validation
American College of Physicians, 2017, United States of America²³
Intended users: All clinicians Target population: Adults with low BMD or osteoporosis	BISs (ALN, RSN, IBN, ZLN), denosumab, teriparatide, selective estrogen receptor modulators, estrogen, calcium, vitamin D	Reduction in fractures (total, vertebral, nonvertebral, spine, hip, wrist, other); AEs	AHRQ conducted a SR of literature published between January 2, 2005 and June 3, 2011. The review was updated to October 2016 using PubMed.	GRADE system, AGREE II	CGC used an informal process to evaluate and formulate the recommendations based on evidence. The final recommendations are voted by CGC members and ACP’s Board of Regent.	Guideline undergoes Independent peer-review at Annals of Internal Medicine. Occasionally, external reviewers and experts are invited.
European Menopause and Andropause Society, 2017, Greece²⁴
Intended users: NR Target population: PM women or older men (50 y or older) diagnosed with osteoporosis	BISs (ALN, RSN, ZLN), denosumab	Fracture risk, AEs	SR of literature published up to January 31, 2017 with searches in MEDLINE, Scopus, EMBASE, and Cochrane databases.	NR	Three independent researchers reviewed eligible studies during systematic review. Recommendations were reviewed by EMAS board members.	Not externally peer reviewed
National Osteoporosis Guideline Group, 2017, United Kingdom²⁵
Intended users: Health care professionals involved in osteoporosis management Target population: PM women or older men (50 y or more)	BISs (ALN, RSN, IBN, ZLN), denosumab, teriparatide, selective estrogen receptor modulators, estrogen, calcium, vitamin D	Not specified, fracture risk, AEs	Where available, SRs, MAs, and RCTs were used to form the evidence base. PubMed searches were conducted.	SRs and MAs were graded using AMSTAR. Recommendations were systematically graded based on the level of evidence. Grade A = evidence levels Ia (from MAs of RCTs) and Ib (from at least one RCT) Grade B = evidence levels IIa (from at least one NRS), IIb (from at least one other type of quasi-experimental study) and III (from non-experimental studies) Grade C = evidence level IV (from expert opinion)	The National Osteoporosis Guideline Development Writing Group, consisting of the Guideline Development Group and Expert Advisory Group, provided the guideline content. Only the Guideline Development Group was involved in the voting of recommendations.	Recommendations were reviewed by stakeholders and external reviewers
Royal Australian College of General Practitioners, 2017, Australia²⁶
Intended users: Health care professionals involved in osteoporosis management in older patients Target population: PM women or older men (50 y or older) diagnosed with osteoporosis	BISs (ALN, RSN, IBN, ZLN), denosumab, teriparatide, selective estrogen receptor modulators, estrogen, Strontium ranelate, calcium, vitamin D	Fracture, BMD	SR of literature published between 2006 and February 2016. Some studies used for the 2010 guideline are included.	NHMRC grades of recommendation: Grade A = Body of evidence can be trusted to guide practice Grade B = Body of evidence can be trusted to guide practice in most situations Grade C = Body of evidence provides some support for recommendation, but care should be taken in its application Grade D = Body of evidence is weak, and recommendation must be applied with caution	Twelve-member expert Working Group developed recommendations through review of literature, and Working Group consensus when insufficient evidence was available	Review by GP subject matter experts and the RACGP’s Expert Committee for Quality Care. The guideline was also endorsed by the RACGP Council.
Spanish Society for Research on Bone and Mineral Metabolism, 2015, Spain²⁷
Intended users: Unable to assess (full guideline not available in English) Target population: PM women or men diagnosed with osteoporosis	BISs (ALN, ETN, RSN, IBN, ZLN), denosumab, teriparatide, selective estrogen receptor modulators, Strontium ranelate, calcium, vitamin D	Not specified, fracture risk, AEs	Unable to assess. Multiple databases used (MEDLINE, Cochrane, CINAHL).	OCEBM criteria used for levels of evidence. OCEBM grades of recommendation: Grade A = consistent level 1 studies (RCTs) Grade B = consistent level 2 (cohort) or level 3 (case-control) studies or extrapolations of level 1 studies Grade C = level 4 studies (case series & low-quality cohort or case-control studies) or extrapolations of level 2/3 studies Grade D = level 5 tests (expert opinions, inconclusive studies or inconsistency problems)	Unable to assess	Unable to assess
Winnipeg Regional Health Authority, 2014, Canada⁶^,¹⁰
Intended users: Health care providers in Manitoba, including direct care staff, policy makers, educators, administrators and interprofessional care team members Target population: Women and men over age 50	BISs (ALN, RSN, ZLN), denosumab, teriparatide, selective estrogen receptor modulators, calcium, vitamin D	Not specified, fracture risk	SR of literature on fracture risk assessment published between Jan 1990 to September 19, 2010. To identify further studies, they used an existing SR on osteoporosis therapies by MacLean et al., 2008⁴⁵	Recommendations were graded based on the level of evidence. Grade A = Need supportive level 1+ (systematic overview of MA of RCTs) or 1 (adequately powered RCTs) evidence plus consensus Grade B = Need supportive level 2+ (systematic overview or MA of level 2 RCTs) or 2 (RCTs that do not meet level 1 criteria) evidence plus consensus Grade C = Need supportive level 3 (NRS or cohort studies) evidence plus consensus Grade D = Any lower level of evidence (before-after, case-control, case series, case report studies) supported by consensus	The Best Practice Guidelines Committee, consisting of participants from across Canada with methodological and content expertise, drafted the guidelines. An expert panel, consisting of members of the Osteoporosis Canada Scientific Advisory Council, members of stakeholder organizations, family physicians and experts from across Canada, used a modified Delphi method to discuss the recommendations.	The Guidelines Committee and the Executive Committee of the Osteoporosis Canada Scientific Advisory Council reviewed the recommendations.

: ACP = American College of Physicians; AE = adverse event; AGREE II = Appraisal of Guidelines for Research and Evaluation Instrument II; AHRQ = Agency for Healthcare Research and Quality; ALN = alendronate; AMSTAR = A MeaSurement Tool to Assess systematic Reviews; BIS = bisphosphonate; BMD = bone mineral density; CGC = clinical guidelines committee; CINAHL = Cumulative Index to Nursing and Allied Health Literature; EMAS = European Menopause and Andropause Society; ETN = Etidronate; GP = general practitioner; GRADE = Grading of Recommendations, Assessment, Development and Evaluations; IBN = ibandronate; MA = meta-analysis; MEDLINE = Medical Literature Analysis and Retrieval System Online; NHMRC = National Health and Medical Research Council; NOGG = National Osteoporosis Guideline Group; NR = not reported; NRS = non-randomized study; OCEBM = Oxford Centre for Evidence-Based Medicine; PM = postmenopausal; PubMed = public MEDLINE; RACGP = Royal Australian College of General Practitioners; RCT = randomized controlled trial; RSN = risedronate; SEIOMM = Spanish Society for Research on Bone and Mineral Metabolism; SR = systematic review; WRHA = Winnipeg Regional Health Authority; ZLN = zoledronic acid.

Appendix 3. Critical Appraisal of Included Publications

Table 4Strengths and Limitations of Included Health Technology Assessment, Systematic Reviews, and Meta-Analysis using AMSTAR 2¹⁶

Strengths	Limitations
Health Technology Assessment
Davis, 2016, United Kingdom¹⁸
The review has a clear population, intervention, comparator, and outcomes as inclusion criteria An a priori protocol was registered with PROSPERO (CRD42013006883) The choice of included study designs was justified Review authors used a comprehensive literature search strategy, included additional references identified via the bibliography of eligible studies, contacted experts in the field, and sought out grey literature Study selection and data extraction were performed in duplicate A list of excluded studies was provided Authors assessed the risk of bias in the RCTs using the Cochrane risk of bias tool. Attrition of 10% or greater was found in 63% of included studies, 5 trials were either open label or single blind and therefore at high risk of performance bias, and 28% of studies reported assessing outcomes in a blinded manner.¹⁸ Risk of bias in individual studies was accounted for in the interpretation and discussion of the results of the review Authors reported their competing interests and they are unlikely to have influenced the results of the review Review authors reported on source of funding for the included studies Source of funding for the review was disclosed (National Institute for Health Research) and editorial independence was reported	Although the search strategy was thorough, a formal assessment of publication bias was not undertaken
Systematic Reviews and Meta-Analysis
Dennison, 2019, United Kingdom¹⁹
Inclusion criteria for the review has a clear population, intervention, comparator, and outcomes Source of funding for the review were disclosed and editorial independence was reported	Reporting discrepancies were identified between the SR and some primary studies Referencing discrepancies were also identified, where the bibliographic references didn’t match the citation An a priori protocol was not reported for the review The choice of included study designs was not justified Although the search strategy was provided, publication restrictions are not justified. Also, it was not clear if the reference lists of included studies were searched, and if trial/study registries were searched The quality of included studies was not assessed Study selection and data extraction were not reported as completed in duplicate A list of excluded studies was not provided Included studies were not described in detail A satisfactory technique for assessing the risk of bias in included studies was not reported Risk of bias in individual studies was not accounted for in the interpretation and discussion of the results of the review Review authors did not report on source of funding for the included studies Although authors reported their competing interested, a statement on how these were managed was not included
Fink, 2019, United States of America²⁰
Inclusion criteria for the review has a clear population, intervention, comparator, and outcomes An a priori protocol was registered with PROSPERO (CRD42018087006) The choice of included study designs was justified Review authors used a comprehensive literature search strategy and they included additional references identified via the bibliography of eligible studies, as well as grey literature if it included sufficient information to assess eligibility and risk of bias Study selection and data extraction were performed in duplicate Authors assessed the risk of bias for outcomes of interest as low, medium, or high on the basis of criteria from the AHRQ Risk of bias in individual studies was accounted for in the interpretation and discussion of the results of the review Authors reported their competing interests and they are unlikely to have influenced the results of the review Source of funding for the review were disclosed and editorial independence was reported	A list of excluded studies was not provided Review authors did not report on source of funding for the included studies
Nayak, 2019, United States of America²¹
Inclusion criteria for the review has a clear population, intervention, comparator, and outcomes Review authors used a comprehensive literature search strategy which included a hand search of the bibliography of eligible studies. However, it is not clear if they searched trial registries or attempted to retrieve grey literature Authors assessed the risk of bias in individual studies using the Cochrane Collaborations’ risk of bias tool for clinical trials and the Newcastle-Ottawa quality assessment scale for cohort studies Review authors extracted information on source of funding for the included studies, but it was not reported The authors justified combining the data in a meta-analysis and they used a random-effects (DerSimonian and Laird) method, to calculate summary effect size estimates Risk of bias in individual studies was accounted for in the interpretation and discussion of the results of the review Authors reported their competing interests (none) Source of funding for the review were disclosed and it was unlikely to have influenced the findings of the review	An a priori protocol was not reported for the review The choice of included study designs was not justified Study selection and data extraction were not reported as being performed in duplicate A list of excluded studies was not provided Although between-study heterogeneity was assessed, they did not discuss the likely impact on the results Potential publication bias was not reported, therefore its likely impact on the results of the review could not be assessed.
Eriksen, 2014, Norway²²
The choice of included study designs was justified Source of funding for the review were disclosed (no funding received) and editorial independence declared	Although authors described the population and intervention inclusion criteria for the review, it was not clear what comparator and outcomes were of interest, if any An a priori protocol was not reported for the review Review authors did not use a comprehensive literature search strategy. Only one database was searched and there was no mention of hand searching the bibliography of eligible studies, searching trial registries, or attempting to retrieve grey literature Study selection and data extraction were not reported as being performed in duplicate A list of excluded studies was not provided Included studies were not described in adequate detail and there was poor consistency in reporting the population, intervention, comparator, and research design Quality of included studies as well as risk of bias were not assessed Authors did not report on the source of funding of the included studies Although authors noted that there was heterogeneity of the populations, interventions, and lengths of follow up, they did not discuss its likely impact on the results Although authors reported their competing interests (such as those with drug manufacturers), they did not discuss the manner by which these were managed, therefore we could not assess how these may have influenced the review findings

: AHRQ = Agency for Healthcare Research and Quality; PROSPERO = International prospective register of systematic reviews; SR = systematic review.

Table 5Strengths and Limitations of Guidelines using AGREE II⁴⁶

Item	Guideline
Item	American College of Physicians, 2017²³	European Menopause and Andropause Society, 2017²⁴	National Osteoporosis Guideline Group, 2017²⁵	Royal Australian College of General Practitioners, 2017²⁶	Spanish Society for Research on Bone and Mineral Metabolism, 2015²⁷	Winnipeg Regional Health Authority, 2014⁶^,¹⁰
Domain 1: Scope and Purpose
1. The overall objective(s) of the guideline is (are) specifically described.	Yes	Yes	Yes	Yes	Yes	Yes
2. The health question(s) covered by the guideline is (are) specifically described.	Yes	Yes	Yes	Yes	Yes	Yes
3. The population (patients, public, etc.) to whom the guideline is meant to apply is specifically described.	Yes	Yes	Yes	Yes	Yes	Yes
Domain 2: Stakeholder Involvement
4. The guideline development group includes individuals from all relevant professional groups.	Yes	Yes	Yes	Yes	Yes	Yes
5. The views and preferences of the target population (patients, public, etc.) have been sought.	No	No	No	No	Unable to assess^a	No
6. The target users of the guideline are clearly defined.	Yes	No	Yes	Yes	Unable to assess^a	Yes
Domain 3: Rigour of Development
7. Systematic methods were used to search for evidence.	Yes	Yes	Yes	Yes	Unable to assess^a	Yes
8. The criteria for selecting the evidence are clearly described.	Yes	Yes	Yes	Yes	Yes	Yes
9. The strengths and limitations of the body of evidence are clearly described.	Yes	Yes	Yes	Yes	Yes	Yes
10. The methods for formulating the recommendations are clearly described.	Yes	No	Yes	Yes	Unable to assess^a	Yes
11. The health benefits, side effects, and risks have been considered in formulating the recommendations.	Yes	Yes	Yes	Yes	Yes	Yes
12. There is an explicit link between the recommendations and the supporting evidence.	Yes	Yes	Yes	Yes	Yes	Yes
13. The guideline has been externally reviewed by experts prior to its publication.	Yes	No	Yes	Yes	Unable to assess^a	Yes
14. A procedure for updating the guideline is provided.	Yes	No	Yes	Yes	Unable to assess^a	No
Domain 4: Clarity of Presentation
15. The recommendations are specific and unambiguous.	Yes	Yes	Yes	Yes	Yes	Yes
16. The different options for management of the condition or health issue are clearly presented.	Yes	Yes	Yes	Yes	Yes	Yes
17. Key recommendations are easily identifiable.	Yes	Yes	Yes	Yes	Yes	Yes
Domain 5: Applicability
18. The guideline describes facilitators and barriers to its application.	Not within scope	Not within scope	Yes	Yes	Unable to assess^a	Not within scope
19. The guideline provides advice and/or tools on how the recommendations can be put into practice.	Yes	Yes	Yes	Yes	Unable to assess^a	Yes
20. The potential resource implications of applying the recommendations have been considered.	Not within scope	Not within scope	Yes	Yes	Unable to assess^a	Not within scope
21. The guideline presents monitoring and/or auditing criteria.	Yes	Yes	Yes	Yes	Yes	Yes
Domain 6: Editorial Independence
22. The views of the funding body have not influenced the content of the guideline.	Yes	Yes	Yes	Yes	Unable to assess^a	Yes
23. Competing interests of guideline development group members have been recorded and addressed.	Yes	Yes	Yes	Yes	Unable to assess^a	Yes

: NOTE:
a: The integral version of this guideline is only available in Spanish, restricting the assessment of strengths and limitations to information that is published in the English summary²⁷

Appendix 4. Main Study Findings and Authors’ Conclusions

Table 6. Summary of Findings in Included Health Technology Assessment, Systematic Reviews and Meta-Analysis (PDF, 455K)

Table 7. Summary of Recommendations in Included Guidelines (PDF, 214K)

Appendix 5. Overlap between Included Systematic Reviews

Table 8Primary Study Overlap between Included Systematic Reviews

Primary Study Citation	Systematic Review Citation
Primary Study Citation	Dennison, 2019¹⁹	Fink, 2019²⁰	Nayak, 2019²¹	Davis, 2016¹⁸	Eriksen, 2014²²
Adams, 2018³²			✕
Black, 2006⁷			✕
Black, 2012⁸			✕		✕
Bone, 2004³⁰		✕	✕		✕
Cosman, 2011⁴⁴					✕
Curtis, 2018⁴¹	✕
Dell, 2012³⁶	✕
Khow, 2017³⁷	✕
Koh, 2017³⁸		✕
Mignot, 2017³³	✕		✕
Miller, 1997³⁴			✕
Park-Wyllie, 2011⁴³		✕
Schilcher, 2011³⁹	✕
Schilcher, 2015³⁵	✕	✕
Schwartz, 2010⁴⁰	✕		✕		✕
Ström, 2015⁴²	✕
Tonino, 2000²⁹		✕	✕		✕

Appendix 6. Additional References of Potential Interest

Clinical Practice Guidelines

Unclear Methodology

Best Practice Advocacy Centre New Zealand. Bisphosphonates: addressing the duration conundrum. Dunedin (NZ): bpac^nz; 2019: https://bpac.org.nz/2019/docs/bisphosphonates.pdf. Accessed 2019 Oct 4. See: Determining an appropriate duration of bisphosphonate use
Kaiser Permanente Guideline Oversight Group. Osteoporosis screening, diagnosis, and treatment guideline. Seattle (WA): Kaiser Foundation Health Plan of Washington; 2019: https://wa.kaiserpermanente.org/static/pdf/public/guidelines/osteoporosis.pdf. Accessed 2019 Oct 4. See: Recommended pharmacologic options
Adler RA, El-Hajj Fuleihan G, Bauer DC, et al. Managing osteoporosis in patients on long-term bisphosphonate treatment: report of a Task Force of the American Society for Bone and Mineral Research. J Bone Miner Res. 2016 Jan;31(1):16–35. PubMed: PM26350171 [PMC free article: PMC4906542] [PubMed: 26350171]
Toward Optimized Practice (TOP) Osteoporosis CPG Committee. Diagnosis and management of osteoporosis: clinical practice guideline. Edmonton (AB): TOP; 2016 Feb: http://www.topalbertadoctors.org/download/1907/Osteoporosis%20CPG.pdf. Accessed 2019 Oct 4. See: DISCONTINUING BISPHOSPHONATE THERAPY

Position Statements

Simm PJ, Biggin A, Zacharin MR, et al. Consensus guidelines on the use of bisphosphonate therapy in children and adolescents. J Paediatr Child Health. 2018 Mar;54(3):223–233: PubMed: PM29504223 [PubMed: 29504223]
Meier C, Uebelhart B, Aubry-Rozier B, et al. Osteoporosis drug treatment: duration and management after discontinuation. A position statement from the SVGO/ASCO. Swiss Med Wkly. 2017;147:w14484. PubMed: PM28871570 [PubMed: 28871570]
Lee SH, Gong HS, Kim TH, et al. Position statement: drug holiday in osteoporosis treatment with bisphosphonates in South Korea. J Bone Metab. 2015 Nov;22(4):167–174. PubMed: PM26713307 [PMC free article: PMC4691590] [PubMed: 26713307]

About the Series

CADTH Rapid Response Report: Summary with Critical Appraisal

ISSN: 1922-8147

Version: 1.0

Acknowledgement: Yan Li

Funding: CADTH receives funding from Canada’s federal, provincial, and territorial governments, with the exception of Quebec.

Suggested citation:

Duration of BIS Treatment for Patients with Osteoporosis: A Review of Clinical Effectiveness and Guidelines. Ottawa: CADTH; 2019 October. (CADTH rapid response report: summary with critical appraisal).

Disclaimer: The information in this document is intended to help Canadian health care decision-makers, health care professionals, health systems leaders, and policy-makers make well-informed decisions and thereby improve the quality of health care services. While patients and others may access this document, the document is made available for informational purposes only and no representations or warranties are made with respect to its fitness for any particular purpose. The information in this document should not be used as a substitute for professional medical advice or as a substitute for the application of clinical judgment in respect of the care of a particular patient or other professional judgment in any decision-making process. The Canadian Agency for Drugs and Technologies in Health (CADTH) does not endorse any information, drugs, therapies, treatments, products, processes, or services.

While care has been taken to ensure that the information prepared by CADTH in this document is accurate, complete, and up-to-date as at the applicable date the material was first published by CADTH, CADTH does not make any guarantees to that effect. CADTH does not guarantee and is not responsible for the quality, currency, propriety, accuracy, or reasonableness of any statements, information, or conclusions contained in any third-party materials used in preparing this document. The views and opinions of third parties published in this document do not necessarily state or reflect those of CADTH.

CADTH is not responsible for any errors, omissions, injury, loss, or damage arising from or relating to the use (or misuse) of any information, statements, or conclusions contained in or implied by the contents of this document or any of the source materials.

This document may contain links to third-party websites. CADTH does not have control over the content of such sites. Use of third-party sites is governed by the third-party website owners’ own terms and conditions set out for such sites. CADTH does not make any guarantee with respect to any information contained on such third-party sites and CADTH is not responsible for any injury, loss, or damage suffered as a result of using such third-party sites. CADTH has no responsibility for the collection, use, and disclosure of personal information by third-party sites.

Subject to the aforementioned limitations, the views expressed herein are those of CADTH and do not necessarily represent the views of Canada’s federal, provincial, or territorial governments or any third party supplier of information.

This document is prepared and intended for use in the context of the Canadian health care system. The use of this document outside of Canada is done so at the user’s own risk.

This disclaimer and any questions or matters of any nature arising from or relating to the content or use (or misuse) of this document will be governed by and interpreted in accordance with the laws of the Province of Ontario and the laws of Canada applicable therein, and all proceedings shall be subject to the exclusive jurisdiction of the courts of the Province of Ontario, Canada.

The copyright and other intellectual property rights in this document are owned by CADTH and its licensors. These rights are protected by the Canadian Copyright Act and other national and international laws and agreements. Users are permitted to make copies of this document for non-commercial purposes only, provided it is not modified when reproduced and appropriate credit is given to CADTH and its licensors.

Except where otherwise noted, this work is distributed under the terms of a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International licence (CC BY-NC-ND), a copy of which is available at http://creativecommons.org/licenses/by-nc-nd/4.0/

Bookshelf ID: NBK551872PMID: 31893493

PubReader
Print View
Cite this Page
Marchand D, Loshak H. Duration of Bisphosphonate Treatment for Patients with Osteoporosis: A Review of Clinical Effectiveness and Guidelines [Internet]. Ottawa (ON): Canadian Agency for Drugs and Technologies in Health; 2019 Oct 4.
PDF version of this title (1.5M)

In this Page

Context and Policy Issues
Research Questions
Key Findings
Methods
Summary of Evidence
Conclusions and Implications for Decision or Policy Making
References
Selection of Included Studies
Characteristics of Included Publications
Critical Appraisal of Included Publications
Main Study Findings and Authors’ Conclusions
Overlap between Included Systematic Reviews
Additional References of Potential Interest

Other titles in this collection

CADTH Rapid Response Reports

Related information

NLM Catalog
Related NLM Catalog Entries
PMC
PubMed Central citations
PubMed
Links to PubMed

Recent Activity

Clear Turn Off Turn On

Duration of Bisphosphonate Treatment for Patients with Osteoporosis: A Review of...
Duration of Bisphosphonate Treatment for Patients with Osteoporosis: A Review of Clinical Effectiveness and Guidelines

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

Bookshelf

Duration of Bisphosphonate Treatment for Patients with Osteoporosis: A Review of Clinical Effectiveness and Guidelines

Abbreviations

Context and Policy Issues

Research Questions

Key Findings

Methods

Literature Search Methods

Selection Criteria and Methods

Table 1

Exclusion Criteria

Critical Appraisal of Individual Studies

Summary of Evidence

Quantity of Research Available

Summary of Study Characteristics

Study Design

Country of Origin

Patient Population

Interventions and Comparators

Outcomes

Summary of Critical Appraisal

Health Technology Assessment

Systematic Reviews

Evidence-Based Guidelines

Summary of Findings

Clinical Evidence for Varying Treatment Durations and Courses of Bisphosphonates for Osteoporosis

Bone Mineral Density

Antifracture Efficacy

Bone turnover markers

AE – Atypical Femoral Fractures

AE – Other Fracture Risk

Other AEs

Evidence-Based Guidelines Regarding Length of Treatment with Bisphosphonates for Osteoporosis

Duration of treatment

Duration of drug holiday

Limitations

Conclusions and Implications for Decision or Policy Making

References

Appendix 1. Selection of Included Studies

Appendix 2. Characteristics of Included Publications

Table 2Characteristics of Included Health Technology Assessment, Systematic Reviews, and Meta-Analysis

Table 3Characteristics of Included Guidelines

Appendix 3. Critical Appraisal of Included Publications

Table 4Strengths and Limitations of Included Health Technology Assessment, Systematic Reviews, and Meta-Analysis using AMSTAR 216

Table 5Strengths and Limitations of Guidelines using AGREE II46

Appendix 4. Main Study Findings and Authors’ Conclusions

Appendix 5. Overlap between Included Systematic Reviews

Table 8Primary Study Overlap between Included Systematic Reviews

Appendix 6. Additional References of Potential Interest

Clinical Practice Guidelines

Unclear Methodology

Position Statements

About the Series

Suggested citation:

Views

In this Page

Other titles in this collection

Related information

Similar articles in PubMed

Recent Activity

Table 4Strengths and Limitations of Included Health Technology Assessment, Systematic Reviews, and Meta-Analysis using AMSTAR 2¹⁶

Table 5Strengths and Limitations of Guidelines using AGREE II⁴⁶