Cochrane Alendronate, Risendronate Systematic Review

Alendronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women (Review)
Wells GA, Cranney A, Peterson J, Boucher M, Shea B, Welch V, Coyle D, Tugwell P
This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library 2011, Issue 9 http://www.thecochranelibrary.com
Alendronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women (Review) Copyright 2011 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
IN
TE
AL
SE
O N
LY
TABLE OF CONTENTS HEADER . . . . . . . . . . ABSTRACT . . . . . . . . . PLAIN LANGUAGE SUMMARY . BACKGROUND . . . . . . . OBJECTIVES . . . . . . . . METHODS . . . . . . . . . Figure 1. . . . . . . . . Figure 2. . . . . . . . . Figure 3. . . . . . . . . Figure 4. . . . . . . . . Figure 5. . . . . . . . . Figure 6. . . . . . . . . RESULTS . . . . . . . . . . Figure 7. . . . . . . . . Figure 8. . . . . . . . . Figure 9. . . . . . . . . Figure 10. . . . . . . . . Figure 11. . . . . . . . . Figure 12. . . . . . . . . Figure 13. . . . . . . . . Figure 14. . . . . . . . . Figure 15. . . . . . . . . Figure 16. . . . . . . . . Figure 17. . . . . . . . . Figure 18. . . . . . . . . Figure 19. . . . . . . . . Figure 20. . . . . . . . . Figure 21. . . . . . . . . Figure 22. . . . . . . . . DISCUSSION . . . . . . . . Figure 23. . . . . . . . . AUTHORS CONCLUSIONS . . ACKNOWLEDGEMENTS . . . REFERENCES . . . . . . . . CHARACTERISTICS OF STUDIES DATA AND ANALYSES . . . . . ADDITIONAL TABLES . . . . . FEEDBACK . . . . . . . . . WHATS NEW . . . . . . . . HISTORY . . . . . . . . . . CONTRIBUTIONS OF AUTHORS DECLARATIONS OF INTEREST . SOURCES OF SUPPORT . . . . INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 2 3 4 4 7 8 8 9 11 12 13 14 15 17 18 19 20 21 22 23 24 25 27 28 29 30 30 31 34 35 35 35 42 57 59 68 72 72 72 72 73 73
IN
TE
AL
SE
O N
LY
[Intervention Review]
Alendronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women
George A Wells1 , Ann Cranney2 , Joan Peterson3 , Michel Boucher4 , Beverley Shea5 , Vivian Welch6 , Doug Coyle7 , Peter Tugwell8
1 Department of Epidemiology and Community Medicine, University of Ottawa, Ottawa, Canada. 2 Division
Contact address: George A Wells, Department of Epidemiology and Community Medicine, University of Ottawa, Room H1-1, 40 Ruskin Street, Ottawa, Ontario, K1Y 4W7, Canada. gawells@ottawaheart.ca. Editorial group: Cochrane Musculoskeletal Group. Publication status and date: Edited (no change to conclusions), comment added to review, published in Issue 9, 2011. Review content assessed as up-to-date: 13 November 2007. Citation: Wells GA, Cranney A, Peterson J, Boucher M, Shea B, Welch V, Coyle D, Tugwell P. Alendronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women. Cochrane Database of Systematic Reviews 2008, Issue 1. Art. No.: CD001155. DOI: 10.1002/14651858.CD001155.pub2. Copyright 2011 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
Objectives
To assess the efcacy of alendronate in the primary and secondary prevention of osteoporotic fractures in postmenopausal women. Search strategy We searched CENTRAL, MEDLINE and EMBASE for relevant randomized controlled trials published between 1966 to 2007. Selection criteria Women receiving at least one year of alendronate, for postmenopausal osteoporosis, were compared to those receiving placebo and/or concurrent calcium/vitamin D. The outcome was fracture incidence. Data collection and analysis We undertook study selection and data abstraction in duplicate. We performed meta-analysis of fracture outcomes using relative risks and a > 15% relative change was considered clinically important. We assessed study quality through reporting of allocation concealment, blinding and withdrawals.
Alendronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women (Review) Copyright 2011 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd. 1
IN
Osteoporosis is an abnormal reduction in bone mass and bone deterioration leading to increased fracture risk. Alendronate belongs to the bisphosphonate class of drugs, which act to inhibit bone resorption by interfering with the activity of osteoclasts.
TE
Background
AL
ABSTRACT
SE
O N
LY
of Rheumatology, Ottawa Hospital, Ottawa, Canada. 3 Clinical Epidemiology Unit, Ottawa Civic Hospital / Loeb Research Institute, Ottawa, Canada. 4 HTA Development Canadian Agency for Drugs and Technologies in Health (CADTH), Ottawa, Canada. 5 CIET, Institute of Population Health, University of Ottawa, Ottawa, Canada. 6 Centre for Global Health, Institute of Population Health, University of Ottawa, Ottawa, Canada. 7 Epidemiology and Community Medicine, Ottawa Health Research Institute, Ottawa, Canada. 8 Department of Medicine, University of Ottawa, Ottawa, Canada
Main results Eleven trials representing 12,068 women were included in the review. Relative (RRR) and absolute (ARR) risk reductions for the 10 mg dose were as follows. For vertebral fractures, a signicant 45% RRR was found (RR 0.55, 95% CI 0.45 to 0.67). This was signicant for both primary prevention, with 45% RRR (RR 0.55, 95% CI 0.38 to 0.80) and 2% ARR, and secondary prevention with 45% RRR (RR 0.55, 95% CI 0.43 to 0.69) and 6% ARR. For non-vertebral fractures, a signicant 16% RRR was found (RR 0.84, 95% CI 0.74 to 0.94). This was signicant for secondary prevention, with 23% RRR (RR 0.77, 95% CI 0.64 to 0.92) and 2% ARR, but not for primary prevention (RR 0.89, 95% CI 0.76 to 1.04). There was a signicant 40% RRR in hip fractures (RR 0.60, 95% CI 0.40 to 0.92), but only secondary prevention was signicant with 53% RRR (RR 0.47, 95% CI 0.26 to 0.85) and 1% ARR. The only signicance found for wrist was in secondary prevention, with a 50% RRR (RR 0.50 95% CI 0.34 to 0.73) and 2% ARR.
Authors conclusions
PLAIN LANGUAGE SUMMARY
Alendronate for preventing fractures caused by osteoporosis in postmenopausal women This summary of a Cochrane review presents what we know from research about the effect of alendronate for preventing fractures (broken bones) caused by osteoporosis.
In women whose bone density is closer to normal, or who may not yet have had a fracture in the bones of their spine, alendronate: - probably prevents fractures in the spine - probably leads to no difference in fractures of the hip, wrist or bones other than the spine. We often do not have precise information about side effects and complications. This is particularly true for rare but serious side effects. Possible side effects may include digestive problems such as injury to the throat, esophagus and stomach and, less commonly, reduced blood supply to the jaw bone, which causes the bone tissue to break down.
What is osteoporosis and what is alendronate? Bone is a living, growing part of your body. Throughout your lifetime, new bone cells grow and old bone cells break down to make room for the new, stronger bone. When you have osteoporosis, the old bone breaks down faster than the new bone can replace it. As this happens, the bones lose minerals (such as calcium). This makes bones weaker and more likely to break even after a minor injury, like a little bump or fall. Women are more likely to get osteoporosis after menopause.
Alendronate belongs to the class of drugs called bisphosphonates. It is a type of medication that slows down the cells that break down the old bone.
IN
TE
- may prevent fractures in the spine, hip or wrist, or in bones other than the spine.
In women who have already been diagnosed with low bone density, putting them at risk for a fracture, or have already had a fracture in the bones of their spine, alendronate:
AL
SE
O N
At 10 mg per day, both clinically important and statistically signicant reductions in vertebral, non-vertebral, hip and wrist fractures were observed for secondary prevention (gold level evidence, www.cochranemsk.org). We found no statistically signicant results for primary prevention, with the exception of vertebral fractures, for which the reduction was clinically important (gold level evidence).
LY
For adverse events, we found no statistically signicant differences in any included study. However, observational data raise concerns regarding potential risk for upper gastrointestinal injury and, less commonly, osteonecrosis of the jaw.
The best estimate of what happens to women that have already been diagnosed with low bone density or have already had a fracture in the bones of their spine: Fracture of the spine - 12 out of 100 women had a fracture when taking a placebo - 6 out of 100 women had a fracture when taking alendronate Fracture in the hip or wrist - 2 out of 100 women had a fracture when taking a placebo - 1 out of 100 women had a fracture when taking alendronate
- 9 out of 100 women had a fracture when taking a placebo - 7 out of 100 women had a fracture when taking alendronate
The best estimate of what happens to women whose bone density is closer to normal or who may not yet have had a fracture in the bones of their spine:
- 3 out of 100 women had a fracture when taking a placebo
Fractures in bones other than the spine:
- 1 out of 100 women had a hip fracture when taking a placebo - 1 out of 100 women had a hip fracture when taking alendronate - 3 out of 100 women had a wrist fracture when taking a placebo - 4 out of 100 women had a wrist fracture when taking alendronate
- 13 out of 100 women had a fracture somewhere other than the spine when taking a placebo
BACKGROUND
Osteoporosis is in part a natural consequence of aging in postmenopausal women (Hodsman 2002). It is a skeletal disorder characterized by decreased bone mass, and deterioration in microarchitecture of bone resulting in an increased risk of fracture (NIH Consensus 2001). The most common consequences of osteoporosis are fractures of the hip, wrist and vertebrae (Hodsman 2002). Bone strength reects the integration of two main features: bone density and
IN
- 12 out of 100 women had a fracture somewhere other than the spine when taking alendronate
TE
AL
- 1 out of 100 women had a fracture when taking alendronate
SE
bone quality (Brown 2002). The clinical indicator of bone quality is a patients history of a fragility fracture. A fragility fracture is a fracture caused by an injury that would be insufcient to fracture normal bone (e.g. a fall from a standing height or less)(Brown 2002). The preferred method of evaluating bone density is the measurement of bone mineral density (BMD) of the spine and hip by Dual Energy X-ray Absorptiometry (DXA), which can be used to assess response to therapy (Hanley 2003). The interpretation of BMD results is based on comparison of a
3
Fracture of the spine
O N
LY
Fractures in bones other than the spine
Osteoporosis-related morbidity is associated with signicant medical and social consequences (Brown 2002). The major source of morbidity and mortality from osteoporosis is attributed to hip fractures. Hip fractures are not only associated with an increased mortality risk, but also inuence long-term function and independence. Fifty percent of women who sustain a hip fracture do not return to their previous functional state and become dependent on others for their daily activities (Brown 2002). The mortality associated with hip fractures in older women may be as high as 20% in the rst year (Cauley 2000). This excess mortality may not be directly attributable to the hip fracture, but to comorbid conditions (Browner 1996; Cooper 1993).
AL
Osteoporosis can be detected by BMD measurement or diagnosed by presence of osteoporosis-related fractures. The presence of preexisting osteoporotic fractures is an important risk factor for future fractures (Hodsman 2002). It is reported that 25% of women aged 80 have had at least one vertebral fracture (Melton 1989). Cauley 2000 demonstrated excess mortality in women who have experienced a clinical vertebral fracture. The cumulative lifetime fracture risk for a 50-year-old woman with osteoporosis is stated to be as high as 60% (Cummings 1989). As a result, effective fracture prevention would have a major impact on morbidity and a smaller but important impact on mortality in these women.
TE
IN
Prevention and treatment of osteoporosis can be complex, due to the multifactorial etiology of the disorder. Anabolic therapies directed at increasing bone formation, such as teriparatide (recombinant human parathyroid hormone (1-34))(Shukla 2003) are available and very expensive; however, most currently available osteoporosis drugs are anti-resorptive agents that act to decrease bone turnover. Anti-resorptive drugs include the bisphosphonates (e.g. etidronate, alendronate and risedronate). They are recommended as rst-line preventive agents in postmenopausal women with low BMD, and as rst-line agents for the treatment of postmenopausal women with osteoporosis (Brown 2002).
SE
OBJECTIVES
The aim of this systematic review was to assess the clinical efcacy of alendronate in the primary and secondary prevention of osteoporotic fractures in postmenopausal women receiving these agents, compared with untreated women over a follow-up period of at least one year.
METHODS
Criteria for considering studies for this review
Types of studies Randomized controlled trials (RCTs) with a duration of at least one year were included in this review. Types of participants We included only post-menopausal women, and accepted both primary and secondary prevention trials. A hierarchy was used to dene primary versus secondary prevention according to the information available. We selected a denition of primary and secondary prevention that gave more weight to study inclusion criteria, than baseline statistics. That is, if the inclusion criteria restricted the population to women whose bone density was at least 2 SD values below the peak bone mass, or the inclusion criteria
4
O N
patients BMD with the mean value for a young adult population. The T-score is the number of standard deviations (SDs) above or below the mean BMD for normal young adults (Brown 2002). The World Health Organization (WHO) Study Group on Osteoporosis dened osteoporosis as a hip BMD level of more than 2.5 SDs below the mean BMD for young, white, adult women (WHO 1994). Using the WHO denition, approximately 30% of postmenopausal women have osteoporosis (Kanis 1994; WHO 1994). It should be noted that there are limitations associated with the WHO denition. The predictive value of BMD measurement for fracture varies depending on the site selected, database used for comparison and the technology used. Furthermore, T-scores do not provide a good basis to establish comparable diagnostic thresholds between different regions of interest and different bone mass measurement techniques (Black 2001). As a result, betweensite and technique variability introduces potential for misclassication and inappropriate treatment of some individuals.
Bisphosphonates are stable analogues of naturally occurring pyrophosphates. The mechanism of action of these drugs is to inhibit bone resorption through their effects on osteoclast function (Brown 2002). Bisphosphonates are poorly absorbed and avidly taken up by bone on active sites of resorption. Alendronate is a second generation nitrogen containing bisphosphonate which is administered daily or once weekly (depending on formulation) and does not impair bone mineralization at doses that maximally inhibit bone resorption (Rodan 1993). The recommended doses for the prevention and treatment of osteoporosis in postmenopausal women are 5 mg/day (35 mg/week) or 10 mg/day (70 mg/week). Alendronate at 10 mg/day or greater, relative to control, has been shown to increase bone mineral density by 7.48% (95% CI, 6.12 to 8.85) after two to three years of treatment in the lumbar spine; 5.60% (95%CI, 4.80 to 6.39) after three to for years in the hip and 2.08% (95% CI, 1.53 to 2.63) after two to four years in the forearm (Cranney 2002). At 5 mg/day, bone mineral density was increased by 5.81% (95% CI, 4.32 to 6.29) after two to three years in the spine, 4.64% (95%CI, 4.27 to 5.01) after three to four years in the hip and 1.83% (95% CI, 1.47 to 2.20) after three to four years in the forearm. (Cranney 2002).
LY
restricted the population to women who had experienced previous vertebral compression fractures, then the trial was considered a secondary prevention study. If such inclusion criteria were not provided, then the baseline statistics were considered as follows: (a) we considered the trial as primary prevention if the average T-score (and SD) was such that it included women whose bone density was within 2 SD of the mean, or if the prevalence of vertebral fracture at baseline was less than 20%; and (b) when these data were not available, we considered a trial as secondary prevention if the average age was above 62 years. Types of interventions Treatment: Alendronate at any dose. Comparators: No treatment (including placebo or calcium and/or vitamin D). If the study used calcium and/or vitamin D controls, these same treatments would have to be given concurrently in the alendronate treatment groups. Types of outcome measures
Incidence of fractures, including vertebral, non-vertebral, hip and wrist fractures.
Search methods for identication of studies
The Cochrane Collaborative approach for identifying randomized controlled trials (RCTs) as described by Dickersin 1994, and modied for the Cochrane Musculoskeletal Group, guided our literature search. We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (1966 to November 2004), Current Contents and citations of relevant articles. No language restrictions were applied to the search strategy. The actual literature search was conducted in three stages. The rst stage was the basis for our systematic review published in 2002 (Cranney 2002) and the second and third stages involved updating the search. The rst search, for the period 1966-99, included CENTRAL, MEDLINE, EMBASE and Current Contents. This was followed by a MEDLINE search for the time period 1999-November 2004. This MEDLINE search was conrmed by a parallel literature search that was conducted for a companion bisphosphonate economic report (CADTH 2006). For the nal update (2004 - February 2007), we searched CENTRAL, MEDLINE and EMBASE. Literature search for MEDLINE using OVID interface
7. or/2-6 8. menopause/ 9. post-menopaus$.tw. 10. postmenopaus$.tw. 11. or/8-10 12. 7 and 11 13. 1 or 12 14. alendronate/ 15. alendronate.tw,rn. 16. fosamax.tw. 17. aminohydroxybutane bisphosphonate.tw. 18. or/14-17 19. 13 and 18 20. meta-analysis.pt,sh. 21. (meta-anal: or metaanal:).tw. 22. (quantitativ: review: or quantitativ: overview:).tw. 23. (methodologic: review: or methodologic: overview:).tw. 24. (systematic: review: or systematic: overview).tw. 25. review.pt. and medline.tw. 26. or/20-25 27. 19 and 26 28. clinical trial.pt. 29. randomized controlled trial.pt. 30. tu.fs. 31. dt.fs. 32. random$.tw. 33. (double adj blind$).tw. 34. placebo$.tw. 35. or/28-34 36. 19 and 35
AL
TE
1. osteoporosis, postmenopausal/ 2. osteoporosis/ 3. osteoporosis.tw. 4. exp bone density/ 5. bone loss$.tw. 6. (bone adj2 densit$).tw.
IN
SE
Data collection and analysis

Selection of studies Two reviewers examined each title generated from the search and identied potentially eligible articles, for which we obtained the abstracts. For abstracts consistent with study eligibility, the full article text was obtained. Overall, we only considered for inclusion studies for which ndings were published, either as full article or abstract. Data abstraction strategy Two independent reviewers abstracted all information and data using standardized data abstraction forms, with a third reviewer verifying the data. Abstraction included information on pertinent methodological aspects of the study design, characteristics of the participants, the specic dose of the study drug used and the outcomes assessed (e.g. number of vertebral, non-vertebral, hip and wrist fractures). For fracture data, we considered all reported fractures (whether clinical or radiographic). For the yearly data, our unit of analysis was number of patients sustaining a fracture. If an article reported yearly data, we used the time points available. For baseline denominators, we used the
5
O N
LY
For person year data, the unit of analysis, if available, was number of fractures. When these data were not available (i.e. in the majority of cases), we used the number of women sustaining a fracture. For denominators, we multiplied the number of women followed by the length of the study. For radiographic vertebral fractures, we used the number of women with available radiographs, if the number was reported in the article. For clinical fractures, we estimated the number of women followed over the duration of the study by taking the mean of the baseline and follow-up denominators. Strategy for quality assessment Two reviewers assessed each eligible RCT for quality based on allocation concealment. Research has shown that lack of adequate allocation concealment is associated with bias (Higgins 2005), and studies can be judged on the method of allocation concealment. The method for assigning participants to interventions should be robust against patient and clinician bias, and its description should be clear. The reviewers were required to indicate whether allocation concealment was adequate (A), unclear (B), or inadequate (C) as per The Cochrane Collaboration criteria as follows. Adequate: The following are some approaches that can be used to ensure adequate concealment schemes: centralized or pharmacycontrolled randomization, pre-numbered or coded identical containers which are administered serially to participants; on-site computer system combined with allocations kept in a locked unreadable computer le that can be accessed only after the characteristics of an enrolled participant have been entered or sequentially
IN
TE
AL
SE
O N
LY
same baseline denominator for each time point. For follow-up denominators, we used any yearly follow-up number of patients reported in the article, if available. If these were not available, we assumed a uniform drop-out rate for each year and calculated the denominators by determining the proportion of subjects that would have remained at the end of the year in question based on the number of withdrawals over the course of the study. If an article reported only end of study outcomes, these were used for our analysis with the exception of outcomes for which the numerator was zero for both treatment groups. In these instances, we included the outcome (with any necessary adjustments for follow-up denominators) for the earlier years in the duration of the study. For example, if a trial reported zero hip fractures for both treatment arms at the end of year three, we would also include in our analysis zero hip fractures for that trial at years one and two.
numbered; or sealed, opaque envelopes. Other approaches may include those similar to ones listed previously, along with reassurance that the person who generated the allocation scheme did not administer it. Inadequate: Approaches to allocation concealment that are considered inadequate include: alternation, use of case record numbers, dates of birth or day of the week and any procedure that is entirely transparent before allocation, such as an open list of random numbers. Unclear: When studies do not report any concealment approach, adequacy should be considered unclear. Examples include merely stating that a list or table was used, only specifying that sealed envelopes were used and reporting an apparently adequate concealment scheme in combination with other information that leads the reviewer to be suspicious. In addition, blinding and loss to follow-up were assessed. Data analysis For the analysis of fractures (i.e. vertebral, non-vertebral, hip and wrist), the relative risk (RR) of fracture was calculated. The methods we used for pooling the results are described elsewhere (Fleiss 1993). The pooled or weighted RRs, using the general inverse variance method for the weights were calculated. For the pooled results, site-specic 95% condence intervals (CIs) were calculated for vertebral, non-vertebral, hip and wrist fractures, and we tested for association using a chi-square test procedure, taking P value < 0.05 for presence of statistical association. Statistically signicant risk reductions were considered to be clinically important if a 15% or greater relative benet was shown. We also tested for homogeneity using a chi-square test procedure, taking the specic cut-off for presence of statistical heterogeneity as P value = 0.10 ( Fleiss 1993). In the event of signicant heterogeneity, a randomeffects model was used. If the relative risk reduction (RRR) was signicant (P value < 0.05), then the absolute risk reduction (ARR) and number needed to treat (NNT) were calculated. For these calculations, we based the ve-year risk of fracture in the untreated population on the FRACTURE Index (FI)(Black 2001), and the lifetime and veyear age-specic risks in the untreated population on the model by Doherty et al. (Doherty 2001) for predicting osteoporotic fractures in postmenopausal women (Figure 1; Figure 2; Figure 3; Figure 4).
Figure 1. Models for fracture risk in postmenopausal women: FRACTURE Index
IN
TE
AL
SE
O N
LY
7
Figure 2. Five year risk of fracture by Quintiles of the FRACTURE Index: Assessment with bone mineral density
Figure 3. Five year risk of fracture by Quintiles of the FRACTURE Index: Assessment without bone mineral density
IN
TE
AL
SE
O N
LY
8
Figure 4. Estimated ve year age-specic risks of rst and subsequent osteoporotic fractures (from Doherty et al 2001)
Trials varied as to the length of treatment (e.g. one to four years) and the number of patients available for study at the start of treatment (i.e. baseline denominator) compared to those available at different time points during the trial (i.e. follow-up denominators). The base case taken for the review of fractures considered the data available for the longest period of time for the treatment in the trial (i.e. all years) and used the baseline denominators for the number of patients in the trial. Data was initially pooled broadly across primary and secondary trials. The overall analysis was also considered using person-years of observation. In addition, we conducted subgroup analysis for: 1) primary versus secondary, 2) treatment duration and 3) treatment dose. Furthermore, we conducted sensitivity analysis for: 1) baseline denominators versus follow-up denominators, 2) xed versus random effects model and 3) baseline vertebral fracture rate. For the last sensitivity analysis, recall that the vertebral fracture criteria for a trial to be considered secondary was a prevalence of vertebral fracture at baseline of greater than 20%. A sensitivity analysis using different vertebral fracture rates (i.e. 100%, > 80%, > 60%, > 40%, > 20%) without the BMD and age criteria was conducted. Such sensitivity analysis allowed evaluating whether the effect of bisphosphonates on the secondary prevention of osteoporotic fractures varied, depending on how strictly secondary prevention was dened. Grading of evidence We graded results for the primary analyses using the system described in Evidence-based Rheumatology (Tugwell 2004), as recom-
IN
TE
AL
SE
mended by the Musculoskeletal Group. Platinum: To achieve the platinum level of evidence, a published systematic review that has at least two randomized controlled trials each satisfying the following is required: - sample sizes of at least 50 per group - if these do not nd a statistically signicant difference, they are adequately powered for a 20% relative difference in the relevant outcome; - blinding of patients and assessors for outcomes; - handling of withdrawals > 80% follow up (imputations based on methods such as Last Observation Carried Forward (LOCF) are acceptable); - concealment of treatment allocation. Gold: The gold level of evidence requires at least one randomized clinical trial meeting all of the following criteria for the major outcome(s) as reported: - sample sizes of at least 50 per group - if they do not nd a statistically signicant difference, they are adequately powered for a 20% relative difference in the relevant outcome; - blinding of patients and assessors for outcomes; - handling of withdrawals > 80% follow up (imputations based on methods such as LOCF are acceptable); - concealment of treatment allocation. Silver: The silver level of evidence requires a randomized trial that does
O N
LY
9
IN
TE
AL
SE
O N
not meet the above criteria for gold or platinum ranking, or evidence from at least one study of non-randomized cohorts that did and did not receive the therapy, or evidence from at least one high-quality case-control study. A randomized trial with a headto-head comparison of agents would be considered silver level ranking unless a reference were provided to a comparison of one of the agents to placebo showing at least a 20% relative difference. Bronze: The bronze level of evidence requires at least one high-quality case series without controls (including simple before/after studies in which patients act as their own control) or a conclusion derived from expert opinion based on clinical experience without reference to any of the foregoing (for example, argument from physiology, In addition, for the outcomes of vertebral and hip fractures we prepared Summary of Findings tables using the GRADE criteria from the GRADE working group (GRADE 2004), (Figure 5; Figure 6).
bench research or rst principles). Clinical relevance tables Clinical relevance tables were compiled under additional tables to improve the readability of the review. Results were presented within the context of both the study population and moderate-/high-risk women from the population at large. The number needed to treat (NNT) was calculated using the relative risk (RR) in combination with either the risk of fracture in the control group, or the ve-year FRACTURE Index (Black 2001). To do this, the Visual Rx NNT calculator (Cates 2004) was used. The weighted absolute risk difference was calculated using the risk difference (RD) statistic in RevMan and RR-1 was used to calculate the relative percent change (Table 1; Table 2).
LY
10
Figure 5. Summary of Findings for Primary Prevention
IN
TE
AL
SE
O N
LY
11
Figure 6. Summary of Findings for Secondary Prevention
IN
TE
AL
SE
O N
LY
12
RESULTS Description of studies

See: Characteristics of included studies; Characteristics of excluded studies. Quantity of research available The literature search revealed 1203 citations as depicted in Figure 7. Of these, 85 articles were retrieved for further review. A total of 74 articles were excluded for various reasons, including lack of appropriate control group (19)(Black 2003; Body 2002; Chailurkit 2003; Davas 2003; Evio 2004; Iwamoto 2004; Kushida 2004; Luckey 2004; Palomba 2002; Rittmaster 2000; Rizzoli 2002; Rozkydal 2003; Sahota 2000; Sambrook 2004a; Schnitzer 2000; Simon 2002; Sosa 2002; Tiras 2000; Vasikaran 1995); lack of fracture outcome (25)(Aki 2003; Boivin 2000; Bouxsein 1999; Chailurkit 2003; Chavassieux 1997; Cummings 2000; Dobnig 2006; Gonnelli 2002; Ho 2005; Johnell 2002; Kung 2000; Lau 2000; McClung 1998; Nenonen 2005; Ravn 1999a; Ravn 1999b;
Rhee 2006; Rossini 2000; Schneider 1999; Stepan 1999; Tutuncu 2005; Uusi-Rasi 2003; van der Poest 2000; Yen 2000; Yildirim 2005); lack of appropriate fracture data (i.e. reported as adverse events or unspecied)(8), (Adami 1995; Bell 2002; Bone 2000; Bonnick 1998; Downs 2000; Greenspan 2003; Hosking 2003; Murphy 2001); lack of randomization (2), (Heijckmann 2002; Sawka 2003); extension/discontinuation studies (6)(Bone 2004; Greenspan 2002a; McClung 2004; Ravn 1999c; Ravn 2000; Sambrook 2004b); duplicate report or earlier report of another study (7)(Adami 1993; Bettembuk 1999; Black 2000; Devogelaer 1996; Hochberg 1999; Seeman 1999; Tucci 1996); duration of therapy less than one year (7)(Cheng 2002; Chesnut 1993; Greenspan 2002b; Harris 1993; Malavolta 1999; Payer 2000; Rossini 1994). If we found duplicate reports of the same study in preliminary abstracts and articles, we analyzed the data from the most complete data set. In total, 11 trials met the selection criteria for inclusion in this report (Ascott Evans 2003, Black 1996; Bone 1997; Chestnut 1995; Cummings 1998; Durson 2001; Greenspan 2002; Greenspan 1998; Hosking 1998; Liberman 1995; Pols 1999).
IN
TE
AL
SE
O N
LY
13
Figure 7. Summary of literature search for alendronate
IN
TE
AL
SE
O N
LY
14
IN
TE
Figure 8. Weighted relative risk (RR) of fracture after alendronate (10 mg)
AL
Length of follow up ranged from one to four years, and mean age was 53 to 78 years. Eight trials excluded women with a history of gastrointestinal disease (Black 1996; Bone 1997; Cummings 1998; Durson 2001; Greenspan 2002; Hosking 1998; Liberman 1995; Pols 1999) and three trials evaluated fractures as the stated primary outcome (Black 1996; Cummings 1998; Durson 2001). All trials
SE
Trial characteristics The characteristics of the 11 selected trials are provided in the Characteristics of included studies table. A total of 12,068 women were enrolled; of these, 5525 received placebo. Three trials were in primary prevention (Ascott Evans 2003; Cummings 1998; Hosking 1998), and the other eight involved women with low BMD on densitometry and/or high prevalence of vertebral fracture (Black 1996; Bone 1997; Chestnut 1995; Durson 2001; Greenspan 2002; Greenspan 1998; Liberman 1995; Pols 1999). Three trials (including the largest secondary prevention trial--the Fracture Intervention Trial or FIT), used an initial dose of 5 mg and then switched to 10 mg for the nal years (Black 1996; Cummings 1998; Greenspan 1998). Two trials (Bone 1997; Hosking 1998) evaluated only doses of 5 mg or less, four trials (Ascott Evans 2003; Durson 2001; Greenspan 2002; Pols 1999) used only the 10 mg dose and two trials (Chestnut 1995; Liberman 1995) evaluated 5, 10 and 20 mg doses. For some endpoints, studies which administered both 5 and 10 mg did not report the results separately by dose (Black 1996; Cummings 1998; Greenspan 1998; Liberman 1995 ). To err on the conservative side, these endpoints were included in the 10 mg analysis.
administered at least 500 mg of calcium to all patients and vitamin D, at daily doses ranging from 125-400 IU, was administered in four trials (Black 1996; Cummings 1998; Greenspan 2002; Greenspan 1998). We did not include data from the HRT arm of the Hosking et al trial (Hosking 1998).
Risk of bias in included studies

Four trials concealed allocation (Black 1996; Bone 1997; Cummings 1998; Hosking 1998) and for the remainder it was unclear. Two trials achieved a loss to follow up of less than 5% (Black 1996; Cummings 1998); ve trials had losses to follow up from 5% to 20% (Ascott Evans 2003; Chestnut 1995; Hosking 1998; Liberman 1995; Pols 1999); three trials had losses to follow up over 20% (Bone 1997; Durson 2001; Greenspan 1998) and one trial did not report losses to follow up (Greenspan 2002). All studies but one were double blind (Durson 2001).
Effects of interventions
Effect on fractures A summary of the overall review of fractures for the base case (i.e. the longest treatment duration and use of the baseline denominators for the number of patients in the trial) for the standard 10 mg dose of alendronate is provided in Figure 8. In general, for vertebral, non-vertebral, hip and wrist fractures, the pooled estimate of the RR was signicant for secondary prevention but not for primary prevention (with the exception of vertebral fractures).
O N
LY
15
Vertebral fractures Vertebral fractures were reported in eight of 11 trials (Ascott Evans 2003; Black 1996; Bone 1997; Hosking 1998; Chestnut 1995; Cummings 1998; Durson 2001; Liberman 1995). In two trials, no fractures occurred in either treatment group (Ascott Evans 2003; Chestnut 1995) and two trials (Bone 1997; Hosking 1998) evaluated only the 5 mg dose The pooled estimate of RR of vertebral fractures from the four trials (Black 1996; Cummings 1998; Durson 2001; Liberman 1995) that could be analyzed for the 10 mg dose demonstrated a reduction (45%) in vertebral fractures (RR 0.55, 95%, CI 0.45 to 0.67). For details, please refer to Figure 8 and Comparison 01.01. This supports a fracture risk reduction with alendronate and results were consistent across the four trials (P = 0.61). There was a signicant reduction in vertebral fractures for both primary and secondary prevention trials. Estimates for the risk reduction were similar for the primary (RR 0.55, 95% CI, 0.38 to 0.80) and secondary (RR 0.55, 95% CI 0.43 to 0.69) prevention trials. Corresponding to the signicant RRR of 45% for the primary or secondary prevention of vertebral fractures, the absolute measures ARR and NNT of the 5-year risk of vertebral fracture after
IN
TE
AL
SE
treatment with alendronate were calculated for different levels of increasing risk as given by the FI. Results are provided in Figure 9, as well as for increasing age in Figure 10. For the illustrative case of the patient with a FI of 6-7, the ARR in vertebral fracture was 3.2% (i.e. a reduction in risk from 7.1% to 3.9%) and the NNT was 31 (i.e. 31 patients would need to be treated to avoid one vertebral fracture). Across the range of increasing FI risk, the ARR for vertebral fracture ranged from 0.5% to 5.0%, and the NNT to avoid one vertebral fracture ranged from 200 to 20. For the illustrative patient in the age group 60-64 years, the ARR for the rst vertebral fracture was 0.5% (i.e. a reduction in risk from 1.0% to 0.55%) and the NNT was 222 patients treated to avoid the rst fracture. The ARR for a subsequent fracture was 4.4% (i.e. a reduction in risk from 9.7% to 5.3%) and the NNT was 23 patients treated to avoid one subsequent fracture. For increasing age, the 5-year age-specic ARR for the rst vertebral fracture increased from 0.1% for the youngest age group (50-54 years) to 2.1% in the highest age group (90+ years). Accordingly, the NNT decreased from 1111 to 47. For the subsequent fracture, ARR increased from 0.2% to 12.6% and the NNT decreased from 444 to 8.
O N
LY
16
Figure 9. Five year FRACTURE Index specic risk of fracture after alendronate (10 mg)
IN
TE
AL
SE
O N
LY
17
Figure 10. Five year age-specic sk of rst and subsequent fracture after alendronate (10 mg)
Non-vertebral fractures Non-vertebral fractures were reported in nine trials (Ascott Evans 2003; Black 1996; Bone 1997; Chestnut 1995; Cummings 1998, Hosking 1998; Greenspan 1998; Liberman 1995; Pols 1999). One trial did not report fractures separately by treatment groups (Chestnut 1995); one trial reported that no fractures occurred in either treatment group (Ascott Evans 2003) and two trials evaluated only the 2.5 and 5 mg doses. (Bone 1997; Hosking 1998). The pooled estimate of the RR of non-vertebral fractures from the ve trials (Black 1996; Cummings 1998; Greenspan 1998; Liberman 1995; Pols 1999) that could be analyzed for the 10 mg dose demonstrated a signicant reduction (16%) in non-vertebral fractures (RR: 0.84, 95% CI 0.74 to 0.94). Details can be found in (Figure 8 and Comparison 02.01) and results were consistent across the ve trials (p = 0.29). Although the primary and secondary prevention trials differed in signicance of the reduction in risk of non-vertebral fractures, the non-signicant reduction (RR 0.89, 95% CI 0.76 to 1.04) in the one primary was not clearly
TE
AL
different from the signicant reduction (RR 0.77, 95% CI 0.64 to 0.92) in the four secondary prevention trials. Corresponding to the signicant RRR of 23% for the secondary prevention of non-vertebral fractures, the absolute measures ARR and NNT of the ve-year risk of non-vertebral fracture after treatment with alendronate were calculated for different levels of increasing risk as measured by the FI (Figure 9) and for increasing age (Figure 10). For the illustrative case of the patient with a FI of 67, the ARR in non-vertebral fracture was 4.6% (i.e. a reduction in risk from 19.8% to 15.2%) and the NNT was 22 (i.e. 22 patients need to be treated to avoid one non-vertebral fracture). Across the range of increasing FI risk, ARR for non-vertebral fracture ranged from 2.0% to 6.3% and NNT to avoid one non-vertebral fracture ranged from 50 to 16. For the illustrative patient in the age group 60-64 years, the ARR for the rst non-vertebral fracture was 0.7% (i.e. a reduction in risk from 3.1% to 2.4%) and the NNT was 140 patients treated to avoid the rst fracture. The ARR
18 Alendronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women (Review) Copyright 2011 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
IN
SE
O N
LY
IN
TE
Figure 11. Weighted relative risk of fracture after alendronate (10 mg) by years of treatment
Corresponding to the signicant RRR of 53% for the secondary prevention of hip fractures, the absolute measures: ARR and NNT of the ve-year risk of hip fracture after treatment with alendronate were calculated for different levels of increasing risk as measured by the FI. Results are provided in Figure 9 in addition to those for increasing age in Figure 10. For the illustrative case of the patient with a FI of 6-7, the ARR in hip fracture was 2.1% (i.e. a reduction in risk from 3.9% to 1.8%) and the NNT was 48 (i.e. 48 patients
AL
SE
for a subsequent fracture was 1.4% (i.e. a reduction in risk from 6.2% to 4.8%) and the NNT was 70 patients treated to avoid one subsequent fracture. For increasing age, the ve-year age-specic ARR for the rst non-vertebral fracture increased from 0.4% for the youngest age group (50-54 years) to 8.1% in the highest age group (90+ years) and the NNT decreased from 272 to 12. The ARR for subsequent fracture increased from 0.6% to 8.7% and the NNT decreased from 167 to 12. Hip fractures Hip fractures were reported in seven trials (Ascott Evans 2003; Black 1996; Cummings 1998; Greenspan 2002; Greenspan 1998; Liberman 1995; Pols 1999), with one trial reporting that no fractures occurred in either treatment group (Ascott Evans 2003). The pooled estimate of the RR of hip fractures from the six trials resulted in a signicant reduction (39%) in hip fractures (RR 0.61, 95% CI 0.40 to 0.92) as presented in (Figure 8 and Comparison 03.01). The results were consistent across the six trials (p = 0.84). The reduction in the risk of hip fractures for the one primary prevention trial (Cummings 1998) was not signicant (RR 0.79, 95% CI 0.44 to 1.44) compared to the signicant reduction demonstrated by the ve secondary prevention trials (RR 0.47, 95% CI 0.26 to 0.85).
need to be treated to avoid one hip fracture). Across the range of increasing FI risk, the ARR for hip fracture ranged from 0.2% to 4.6% and the NNT to avoid one hip fracture ranged from 500 to 22. For the illustrative patient in the age group 60-64 years, the ARR for the rst hip fracture was 0.1% (i.e. a reduction in risk from 0.2% to 0.1%) and the NNT was 943 patients treated to avoid the rst fracture. The ARR for a subsequent fracture was 0.1% (i.e. a reduction in risk from 0.2% to 0.1%) and the NNT was 943 patients treated to avoid one subsequent fracture. For increasing age, the ve-year age-specic ARR for the rst hip fracture increased from less than 0.05% for the youngest age group (50-54 years) to 11.1% in the highest age group (90+ years) and the NNT decreased from more than 272 to 9. For the subsequent fracture, the ARR increased from less than 0.05% to 12.1% and the NNT decreased from more than 472 to 8. Wrist fractures Wrist fractures were reported in ve trials (Black 1996; Cummings 1998; Greenspan 1998; Liberman 1995; Pols 1999) and one trial (Ascott Evans 2003) reported that no fractures occurred in either treatment group. Results were not consistent across the ve trials which reported wrist fractures (p = 0.0007)(Figure 11). The pooled estimate of the RR of wrist fracture from these ve trials resulted in a non-signicant reduction in fractures, either using the random effects (RR 0.68, 95% CI 0.34 to 1.37)(Figure 8) or xed effect (RR 0.83, 95%CI 0.65, 1.05) approach. We also could not identify a statistically signicant effect of alendronate when used for the primary prevention of wrist fractures (RR 1.19, 95%CI 0.87 to 1.62)(Figure 8 and Comparison 04.01). In the case of the four secondary prevention trials, a statistically signicant effect was observed (RR 0.50, 95%CI 0.34 to 0.73).
O N
LY
19
Additional analysis Person years Results were similar for vertebral, non-vertebral, hip and wrist fractures when person years were used, as illustrated in Figure 12. Notably, the pooled estimates of the RR for the secondary prevention trials all showed a signicant risk reduction of fracture for all fractures sites. Due to lack of consistency among the trials, a random-effects estimate was used for the pooled estimate for vertebral fractures; however, for wrist fractures, there was no inconsistency among the secondary prevention trials. The risk estimates obtained from the one primary prevention trial for vertebral fracture remained signicant although results were non-signicant for the other fracture sites.
Figure 12. Weighted relative risk (RR) of fracture after alendronate (10 mg): Person years
Subgroup analysis Treatment duration There were no trends over years of treatments that deviated from the overall RR estimates (Figure 11 and Comparisons 05.01; 06.01; 07.01; 08.01; 09.01), with the possible exception of hip fracture risk reduction being greater in later years. Treatment dose For the 5 mg dose of alendronate, fracture data were available for
20
IN
TE
AL
SE
O N
LY
vertebral and non-vertebral sites for secondary prevention trials, as well as non-vertebral for primary prevention. (Figure 13; Figure 14; Figure 15; Comparison 12.01) For vertebral fractures, the decrease in risk of fracture was statistically signicant and there was a further slight decrease in risk with the 5 mg dose (RR 0.40, 95%CI 0.29 to 0.55), compared to 10 mg (RR 0.55, 95%CI 0.43 to 0.69). For non-vertebral fractures, no signicant difference was found. Figure 13. Weighted relative risk of fractrue after alendronate by dose
IN
TE
AL
SE
O N
LY
21
Figure 14. Weighted relative risk of fracture after alendronate by dose: Person years
IN
TE
AL
SE
O N
LY
22
Figure 15. Weighted relative risk of fracture after alendronate by years of treatment and dose
Sensitivity analysis
Baseline versus follow-up denominators By using the data available for longest treatment duration, standard dose of alendronate (10 mg) and follow-up denominators for the number of patients in the trials, a summary of the overall review of fractures was prepared (Figure 16). These data are also provided by years of treatment. (Figure 17). The pooled estimates of the RR of fracture after alendronate were essentially the same as those obtained using the baseline denominators.
IN
TE
AL
SE
O N
LY
23
Figure 16. Weighted relative risk of fracture after alendronate by dose: Follow-up denominators
IN
TE
AL
SE
O N
LY
24
Figure 17. Weighted relative risk of fracture after alendronate by years of treatment and dose: Follow-up denominators
Random versus xed effects model There were a few instances where heterogeneity of the trial results was such that a random-effects model was required. In general, results obtained using the random- and xed-effects models were
IN
TE
AL
similar. Baseline vertebral fracture rate Using different baseline vertebral fracture rates (i.e. 100%, > 80%,
25 Alendronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women (Review) Copyright 2011 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
SE
O N
LY
IN
TE
AL
SE
O N
LY
26
> 60%, > 40%, > 20%) for dening secondary trials, a summary of the overall review of fractures was prepared (Figure 18). For vertebral fractures, the pooled estimates of the RR of fracture after alendronate were essentially the same as those obtained using the denition of secondary trials with the > 20% baseline fracture rate. Although the result for non-vertebral and hip fractures became non-signicant when the criteria increased from > 20% to > 40% and > 40% to > 60% respectively, the relative risk of fracture was exactly the same but the condence intervals were now slightly wider with the exclusion of the study by Liberman et al (Liberman 1995) for non-vertebral fractures and Greenspan et al (Greenspan 2002) for hip fractures. For wrist fractures, the non-signicant result was now signicant when a > 20% baseline fracture rate was used (without the BMD and age criteria) since a xed-effects model could now be used.
Figure 18. Weighted relative risk (RR) of fracture after alendronate (10 mg): sensitivity analysis by baseline prevalent vertebral fracture rate
IN
TE
AL
SE
O N
LY
27
Adverse events A summary of the adverse drug events reported in the 11 randomized placebo-controlled trials of alendronate is provided in Figure 19, Figure 20, Figure 21 and Figure 22. In general, the reported events were similar between alendronate and placebo. In particular, there were seven studies (Ascott Evans 2003; Black 1996; Cummings 1998; Greenspan 1998; Greenspan 2002; Hosking 1998; Pols 1999) reporting any upper gastrointestinal events resulting in an overall RR 1.03 (95% CI 0.98 to 1.08) and two studies (Black 1996; Cummings 1998) reporting esophageal ulcer, resulting in an overall RR 1.16 (95% CI 0.39 to 3.45). Figure 19. Summary of adverse drug events reported in randomized placebo-controlled trials of alendronate (part 1)
IN
TE
AL
SE
O N
LY
28
Figure 20. Summary of adverse drug events reported in randomized placebo-controlled trials of alendronate (part 2)
IN
TE
AL
SE
O N
LY
29
IN
TE
AL
SE
O N
LY
30
Study limitations
DISCUSSION
Summary of results
Secondary analyses showed that a dose of 5 mg, for secondary prevention resulted in a statistically signicant and clinically important reduction in vertebral fractures that is comparable to that of the 10 mg dose (RR 0.40, 95% CI 0.27 to 0.52 versus RR 0.55, 95% CI 0.43 to 0.69). It is worth noting that these results were driven primarily by the large Black trial (Black 1996). In Black, the 5 mg data represented the rst two years of follow up of the treatment group which was switched to 10 mg for the third year. No trials were available to assess the effectiveness of the 5 mg dose for the primary prevention of vertebral fractures. For non-vertebral fractures, no statistically signicant reductions were shown for primary or secondary prevention. There were no trials available to assess the efcacy of the 5 mg dose for hip and wrist fractures.
IN
TE
Based on the longest treatment duration and use of baseline denominators for the number of patients in the included trials, the main benet of alendronate was found to be in the secondary prevention of osteoporotic fractures. At a dose of 10 mg per day, alendronate results in a statistically signicant and clinically important reduction in vertebral, non-vertebral, hip and wrist fractures (gold level evidence). We also found that the use of alendronate 10 mg per day for the primary prevention of osteoporotic fractures is not associated with statistically signicant reductions in risk, with the exception of vertebral fractures for which the reduction was clinically important (gold level evidence).
The results of this systematic review are believed to be robust, as a comprehensive literature search was performed, inclusion and exclusion criteria were specied and a rigorous data analysis was conducted. A potential limitation of our approach may be that the update search (i.e. 2000 to 2004) did not include non-MEDLINE indexed journals. Recent empirical evidence indicates that this approach may have introduced a slight risk of bias into our meta-analysis. On average, such bias is estimated to result in a 6% variation in the pooled results (Egger 2003; Sampson 2003). Accordingly, given that the initial literature search (i.e. 1966 to 2000) was very extensive, the impact of only using MEDLINE for the search update is expected to be minimal, if any. This was conrmed by a parallel literature search update (i.e. 1999 to July 2004) that was conducted for an economic report published by The Canadian Agency for Drugs and Technologies in Health (CADTH 2006). The search update included etidronate, alendronate and risedronate (daily dose regimen only) in addition to teriparatide. A number of databases were searched (i.e. the Cochrane Library, MEDLINE, EMBASE, BIOSIS Previews, Toxle and PubMed) and no additional articles meeting the inclusion criteria were identied. While our methodology was robust, the results of our meta-analysis, however, are only as strong as the primary studies included. In keeping with this, the main limitations with regard to study quality were fracture assessment and classication, the lack of clarity of the concealment of allocation and large losses to follow up (primarily in the smaller studies). A potential source of heterogeneity is the lack of uniform denition of non-vertebral fracture. While some researchers may use a rather liberal denition (any fracture other than vertebral fracture), others may use a more conservative denition which includes only fractures of the hip, clavicle, humerus, wrist, pelvis or leg (Mayo Clin Proc 2005). Another consideration is the fact that fracture data was not the primary outcome for many of the trials. In par31
There were no substantive differences in the results whether baseline, end of study or person year denominators were used. Sensitivity analyses indicated that there were no major differences based on the percentage of baseline vertebral fractures. Further, no trends were found for years of treatment.
AL
SE
O N
LY
Toxicity and withdrawals Discontinuations due to adverse events or dropouts overall were available and analyzed for six (Ascott Evans 2003; Black 1996; Bone 1997; Cummings 1998; Durson 2001; Pols 1999) and ve (Ascott Evans 2003; Durson 2001; Greenspan 1998; Hosking 1998; Pols 1999) of the alendronate trials respectively. The pooled estimate demonstrated no statistical difference between alendronate and placebo for the risk of discontinuing medication due to adverse events (RR 0.95, 95% CI 0.83 to 1.09) or for dropouts overall (RR 1.10, 95% CI 0.94 to 1.29)(Comparison 13.01 and 13.02). Results were consistent across the trials.
Adverse drug events were similar between alendronate and placebo. There were also no statistically signicant difference in either the rate of treatment discontinuation due to adverse events (RR 0.95, 95% CI 0.83 to 1.09) or the overall withdrawal rate (RR 1.10, 95% CI 0.94 to 1.29), compared to placebo. Thus, it was concluded that study participants tolerated their alendronate treatment. Although no increased incidence of adverse effects was detected with alendronate, external to the randomized controlled trials, concerns exist regarding the potential risk of upper gastrointestinal events and osteonecrosis of the jaw.
ticular, three of the 11 alendronate trials (Black 1996; Cummings 1998; Durson 2001) had fractures as the primary outcome. There is another source of heterogeneity, and possible bias, related to some of the secondary prevention studies. It concerns the inclusion of participants with a low BMD but no proven fractures, and the difculty in discriminating between fracture types (traumatic versus pathological).
does not allow for the assessment of long-term toxicity associated with alendronate. Recently, there have been concerns regarding the potential risk of over suppressing bone turnover resulting in osteonecrosis of the jaw (ONJ)(Khosla 2007; Woo 2006). Although the majority of reported cases of ONJ have occurred in cancer patients receiving the intravenous bisphosphonates zolendrate or pamidronate (at higher cumulative doses than used in the treatment of postmenopausal osteoporosis), some osteoporosis patients receiving oral bisphosphonates have developed the condition as well. In a systematic review of cases reported in the medical literature, 13 of 368 bisphosphonate treated ONJ patients had received alendronate and one risedronate (Woo 2006). Since that publication, a review conducted by the American Society for Bone and Mineral Research (ASBMR) Task Force on Bisphosphonate-Associated ONJ has identied studies reporting a total of 67 cases (64 alendronate, two risedronate and one etidronate) among osteoporosis and Pagets disease patients (Khosla 2007). Most notably, an Australian study reported 30 of 114 ONJ cases related to alendronate (22 of whom were under treatment for osteoporosis) and two related to risedronate. The median time to onset, for alendronate, was 24 months. The most common triggering factor was dental extraction (Mavrokokki 2007). The ASBMR task force has pointed out that the incidence of ONJ in the general population not exposed to bisphosphonates is unknown, information on the incidence of ONJ is rapidly evolving and that ,often, the case ascertainment has been inadequate . They recommend that a hierarchy of evidence quality, based on the completeness of information across seven categories related to diagnosis and history, should be established for all future studies reporting cases of ONJ (Khosla 2007). No cases of ONJ were explicitly reported in any of the alendronate trials in our review. As well, a recent 10-year follow up of patients from the Black and Cummings trials (The Fracture Intervention Trial Long-term Extension (FLEX) reported that no cases were observed among the 662 women who were continued on alendronate at 5 or 10 mg for a total of 10 years or the 437 women who were switched to placebo following ve years of alendronate treatment (Black 2006). No difference between treatment groups were found for any other adverse events in FLEX. Finally, because RCTs are not designed to measure ADEs, particularly rare ones, it is common practice to include sources of information other than RCTs. While some reviewers include ADEs reported in observational studies, we elected to obtain information from the Canadian Adverse Drug Reaction Monitoring Program (CADRMP) (Health Canada 2005). Generalizability of ndings Generalizability of our ndings is limited by the controlled design of the trials included in our review. Study participants were carefully selected in these trials, and so utilization of the drugs in real
32
Lastly, a limitation of evaluating data on adverse effects from summary meta-analyses is that participants in RCTs tend to be healthier, with fewer co-morbid diseases, and therefore the results may not be generalizable to clinical practice. For alendronate, eight trials (Black 1996; Bone 1997; Cummings 1998; Durson 2001; Greenspan 2002; Hosking 1998; Liberman 1995; Pols 1999) excluded patients with pre-existing gastrointestinal (GI) disorders. Furthermore, RCTs are underpowered for rare effects and metaanalyses of these trials generally cannot provide conclusive information pertaining to drug toxicity. In addition, the heterogeneity of the adverse drug effects (ADEs) reported in the RCTs described in this review, including their nature, low occurrence and way they were assessed by investigators, made these improper for meta-analysis. Although a number of trials reported composite endpoints such as any GI event, as well as a number of individual GI events, only the two largest trials (Black 1996; Cummings 1998), for example, specically reported esophageal ulcers (an endpoint of particular interest, for which there were no statistically signicant differences between treatment and control). As well, the follow up for the included trials, which ranged between one and four years,
IN
TE
In regards to our own methodology, we acknowledge that the approach used to evaluate the effect of bisphosphonates over time may result in some estimates that are not robust. In particular, in order to determine the effect on the ve-year risk of fracture, we based our evaluation on the FRACTURE Index by Black et al (Black 2001), and for lifetime and ve-year age-specic risks, we used an existing model from Doherty et al (Doherty 2001). Although this latter approach allowed us to estimate the variation in risks between younger and older postmenopausal women, these estimates may be associated with a certain level of uncertainty. Nonetheless, we believe such information may be useful to decision-makers.
AL
SE
O N
An additional limitation is the length of follow up in the studies. It is difcult to extrapolate beyond the duration of the follow-up trials in the review with respect to the long-term impact on fractures. Ultimately, data from longer-term trials will help establish if the effect on fractures is maintained, increased or diminished.
LY
Another methodological limitation concerns the approach used for concealment of treatment allocation which was not reported for most trials. Four of the 11 trials (Black 1996; Bone 1997; Cummings 1998; Hosking 1998) concealed allocation, and for the remainder it was unclear.
life may vary substantially from study conditions. Furthermore, study participants were observed for periods of time varying from one to four years. Consequently, while our results provide support for efcacy (i.e. can the intervention have an effect on outcome?), they may possibly only provide partial information on the longterm effectiveness (i.e. does the intervention have an effect on outcome?) of alendronate in preventing osteoporotic fractures. From a safety perspective, we could not nd any statistically significant difference in either the rates of adverse drug events or withdrawal rates due to adverse drug events between patients receiving a bisphosphonate or patients receiving a placebo. However, outside controlled trials, concerns exist regarding the safe use of alendronate, for which esophageal ulcers and gastritis have been
IN
TE
AL
SE
O N
LY
reported (Kherani 2002). While such adverse events have mainly been identied through case reports and endoscopic studies, similar concerns are also reected in the proportions of gastrointestinal adverse drug reactions associated with the use of bisphosphonates reported to the CADRMP (Health Canada 2005). Indeed, GI adverse drug reactions represented 38% of all reactions reported for alendronate (Figure 23). These proportions should, however, be interpreted with caution, as adverse drug reactions are reported to CADRMP on a volunteer basis by health professionals, which means that several reactions may be unreported. Indeed, it is estimated that less than 10% of adverse reactions are reported to Health Canada (Health Canada 2005c). Also, a denite causeeffect relationship has not been established for these adverse drug reactions.
33
Figure 23. Adverse drug reactions reported to CADRMP for etidronate, alendronate and risedronate
IN
TE
AL
SE
O N
LY
34
Implications for research AUTHORS CONCLUSIONS Implications for practice

Alendronate demonstrated a clinically important benet in the secondary prevention of all osteoporotic fractures. At a dose of 10 mg per day, statistically signicant reductions in vertebral, nonvertebral, hip and wrist were observed. The secondary prevention population was dened as having a bone density of at least 2 SD values below peak bone mass or one or more vertebral compression fractures, or both. There were no statistically signicant reductions found for the primary prevention of osteoporotic fractures, with the exception of vertebral fractures, for which the reduction was clinically important. No increased incidence of adverse effects were detected with alendronate, but clinicians should be aware that outside of randomized controlled trials, concerns exist regarding the potential risk of upper gastrointestinal events and, less commonly, osteonecrosis of the jaw. The prevention of osteoporotic fractures is an important public health intervention. This is particularly true for hip and clinical vertebral fractures (i.e. fractures of the spine that present for medical attention). The RR of death following such fractures is six- to nine-fold greater in postmenopausal women aged 55 to 81 years with low BMD, which represents a typical postmenopausal population (Cauley 2000). In most cases, the mortality increase reects poor underlying health status and comorbidity, in addition to the fracture itself (Cauley 2000). Osteoporotic fractures are also associated with increase in morbidity, as it is reported that 50% of women who sustain a hip fracture do not return to their usual daily activities (Brown 2002), while 33% will require long-term care. Accordingly, reducing the incidence of such fractures can potentially increase the quality of life of patients with osteoporosis. Such interventions may also potentially decrease mortality. It has been suggested from clinical trials with the bisphosphonates (Black 2000b; Harris 1999; McClung 2001) that their effect in reducing non-vertebral fractures may be greater in patients with lower BMD who initiate treatment. The existing data have not fully resolved the question of whether important differences in risk reduction across groups of patients with varying degrees of osteoporosis exist. The impact of the bisphosphonates on the RR of non-vertebral fractures in populations without osteoporosis also merits further investigation. Additional research is needed to clarify the role of bisphosphonates in the primary prevention of osteoporotic fractures. There is also a need for further post-marketing safety. Finally, research into combination therapy with higher doses of vitamin D or anabolic agents would be merited as would research concerning adherence to bisphosphonate therapy. Given the morbidity consequences associated with osteoporotic fractures, preventing their recurrence can potentially lessen the need for community-based health services (e.g. home care). It may also reduce or delay the demand for long-term care beds. However, very little comparative information is currently available to support this (Hodsman 2002). There is also a lack of studies which evaluated the effect of bisphosphonates on hospital admissions (Hodsman 2002).
AL
IN
TE
REFERENCES
Bone 1997 {published data only} Bone HG, Downs RW Jr, Tucci JR, Harris ST, Weinstein RS, Licata AA, et al.Dose-response relationships for alendronate treatment in osteoporotic elderly women. Alendronate Elderly Osteoporosis Study Centers. Journal of Clinical Endocrinology & Metabolism 1997;82(1):26574. Chestnut 1995 {published data only} Chesnut CH III, McClung MR, Ensrud KE, Bell NH, Genant HK, Harris ST, et al.Alendronate treatment of the postmenopausal osteoporotic woman: effect of multiple dosages on bone mass and bone remodeling. American Journal of Medicine 1995;99(2):14452. Cummings 1998 {published data only} Cummings SR, Black DM, Thompson DE, Applegate
35
References to studies included in this review

Ascott Evans 2003 {published data only} Ascott-Evans BH, Guanabens N, Kivinen S, Stuckey BG, Magaril CH, Vandormael K, et al.Alendronate prevents loss of bone density associated with discontinuation of hormone replacement therapy: a randomized controlled trial. Archives of Internal Medicine 2003;163(7):78994. Black 1996 {published data only} Black DM, Cummings SR, Karpf DB, Cauley JA, Thompson DE, Nevitt MC, et al.Randomised trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. Fracture Intervention Trial Research Group.[see comment]. Lancet 1996;348(9041):153541.
SE
ACKNOWLEDGEMENTS
Thank you to Lara Maxwell and Marie Andree Nowlan from the Cochrane Musculoskeletal Group for their editorial assistance, and Tamara Rader for her assistance with the Consumer Summary.
O N
LY
WB, Barrett-Connor E, Musliner TA, et al.Effect of alendronate on risk of fracture in women with low bone density but without vertebral fractures: results from the Fracture Intervention Trial. [see comment]. JAMA 1998; 280(24):207782. Durson 2001 {published data only} Dursun N, Dursun E, Yalcin S. Comparison of alendronate, calcitonin and calcium treatments in postmenopausal osteoporosis. International Journal of Clinical Practice 2001; 55(8):5059. Greenspan 1998 {published data only} Greenspan SL, Parker RA, Ferguson L, Rosen HN, Maitland-Ramsey L, Karpf DB. Early changes in biochemical markers of bone turnover predict the long-term response to alendronate therapy in representative elderly women: a randomized clinical trial. Journal of Bone and Mineral Research 1998;13(9):14318. Greenspan 2002 {published data only} Greenspan S, Field-Munves E, Tonino R, Smith M, Petruschke R, Wang L, et al.Tolerability of once-weekly alendronate in patients with osteoporosis: a randomized, double-blind, placebo-controlled study. [see comment]. Mayo Clinic Proceedings 2002;77(10):104452. Hosking 1998 {published data only} Hosking D, Chilvers CE, Christiansen C, Ravn P, Wasnich R, Ross P, et al.Prevention of bone loss with alendronate in postmenopausal women under 60 years of age. Early Postmenopausal Intervention Cohort Study Group. New England Journal of Medicine 1998;338(8):48592. Liberman 1995 {published data only} Liberman UA, Weiss SR, Broll J, Minne HW, Quan H, Bell NH, et al.Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. The Alendronate Phase III Osteoporosis Treatment Study Group. [see comment]. New England Journal of Medicine 1995;333(22):143743.
salmon calcitonin on bone mass and biochemical markers of bone turnover in postmenopausal women with osteoporosis. Bone 1995;17(4):38390. Aki 2003 {published data only} Aki S, Gulbaba RG, Eskiyurt N. Effect of alendronate on bone density and bone markers in postmenopausal osteoporosis. Journal of Back & Musculoskeletal Rehabilitation 2003;17(1):2731. [MEDLINE: 669] Bell 2002 {published data only} Bell NH, Bilezikian JP, Bone HG III, Kaur A, Maragoto A, Santora AC, Study Group. Alendronate increases bone mass and reduces bone markers in postmenopausal AfricanAmerican women. Journal of Clinical Endocrinology & Metabolism 2002;87(6):27927. Bettembuk 1999 {published data only} Bettembuk P, Balogh A. [The effect of a one-year alendronate therapy on postmenopausal osteoporosis. (Results in Hungarian of an international multicenter clinical study)]. Orvosi Hetilap 1999;140(50):2799803. Black 2000 {published data only} Black DM, Thompson DE, Bauer DC, Ensrud K, Musliner T, Hochberg MC, et al.Fracture risk reduction with alendronate in women with osteoporosis: the Fracture Intervention Trial. FIT Research Group. [erratum appears in J Clin Endocrinol Metab 2001 Feb;86(2):938]. Journal of Clinical Endocrinology & Metabolism 2000;85(11):411824.
AL
Pols 1999 {published data only} Pols HA, Felsenberg D, Hanley DA, Stepan J, MunozTorres M, Wilkin TJ, et al.Multinational, placebocontrolled, randomized trial of the effects of alendronate on bone density and fracture risk in postmenopausal women with low bone mass: results of the FOSIT study. Fosamax International Trial Study Group. Osteoporosis International 1999;9(5):4618.
IN
TE
References to studies excluded from this review

Adami 1993 {published data only} Adami S, Baroni MC, Broggini M, Carratelli L, Caruso I, Gnessi L, et al.Treatment of postmenopausal osteoporosis with continuous daily oral alendronate in comparison with either placebo or intranasal salmon calcitonin. Osteoporosis International 1993;3 Suppl 3:S21S27. Adami 1995 {published data only} Adami S, Passeri M, Ortolani S, Broggini M, Carratelli L, Caruso I, et al.Effects of oral alendronate and intranasal
SE
Black 2003 {published data only} Black DM, Greenspan SL, Ensrud KE, Palermo L, McGowan JA, Lang TF, et al.The effects of parathyroid hormone and alendronate alone or in combination in postmenopausal osteoporosis. [see comment]. New England Journal of Medicine 2003;349(13):120715. Body 2002 {published data only} Body JJ, Gaich GA, Scheele WH, Kulkarni PM, Miller PD, Peretz A, et al.A randomized double-blind trial to compare the efcacy of teriparatide (recombinant human parathyroid hormone (1-34)) with alendronate in postmenopausal women with osteoporosis. [see comment]. Journal of Clinical Endocrinology & Metabolism 2002;87(10):452835.
Boivin 2000 {published data only} Boivin GY, Chavassieux PM, Santora AC, Yates J, Meunier PJ. Alendronate increases bone strength by increasing the mean degree of mineralization of bone tissue in osteoporotic women. Bone 2000;27(5):68794. Bone 2000 {published data only} Bone HG, Greenspan SL, McKeever C, Bell N, Davidson M, Downs RW, et al.Alendronate and estrogen effects in postmenopausal women with low bone mineral density. Alendronate/Estrogen Study Group. Journal of Clinical Endocrinology & Metabolism 2000;85(2):7206. Bone 2004 {published data only} Bone HG, Hosking D, Devogelaer JP, Tucci JR, Emkey RD, Tonino RP, et al.Ten years experience with alendronate for osteoporosis in postmenopausal women. [see comment]. New England Journal of Medicine 2004;350(12):118999.
36
O N
LY
Bonnick 1998 {published data only} Bonnick S, Rosen C, Mako B, DeLucca P, Byrned C, Melton M. Alendronate vs calcium for treatment of osteoporosis in postmenopausal women.. Bone 1998;350(5S):S476. Bouxsein 1999 {published data only} Bouxsein ML, Parker RA, Greenspan SL. Forearm bone mineral densitometry cannot be used to monitor response to alendronate therapy in postmenopausal women. Osteoporosis International 1999;10(6):5059. Chailurkit 2003 {published data only} Chailurkit LO, Jongjaroenprasert W, Rungbunnapun S, Ongphiphadhanakul B, Sae-tung S, Rajatanavin R. Effect of alendronate on bone mineral density and bone turnover in Thai postmenopausal osteoporosis. Journal of Bone & Mineral Metabolism 2003;21(6):4217. Chailurkit 2004 {published data only} Chailurkit LO, Aunphongpuwanart S, Ongphiphadhanakul B, Jongjaroenprasert W, Sae-tung S, Rajatanavin R. Efcacy of intermittent low dose alendronate in Thai postmenopausal osteoporosis. Endocrine Research 2004;30 (1):2936. Chavassieux 1997 {published data only} Chavassieux PM, Arlot ME, Reda C, Wei L, Yates AJ, Meunier PJ. Histomorphometric assessment of the long-term effects of alendronate on bone quality and remodeling in patients with osteoporosis. Journal of Clinical Investigation 1997;100(6):147580. Cheng 2002 {published data only} Cheng ZQ, Yin W, Fan JY, Ma TJ. [The efcacy of alendronate in the prevention and treatment of postmenopausal osteoporosis]. [Chinese]. Chung-Kuo i Hsueh Ko Hsueh Yuan Hsueh Pao Acta Academiae Medicinae Sinicae 2002;24(3):3069.
Dobnig 2006 {published data only} Dobnig H, Hofbauer LC, Viereck V, Obermayer-Pietsch B, Fahrleitner-Pammer A, Dobnig H, et al.Changes in the RANK ligand/osteoprotegerin system are correlated to changes in bone mineral density in bisphosphonate-treated osteoporotic patients. Osteoporosis International 2006;17(5): 693703. [MEDLINE: 13] Downs 2000 {published data only} Downs RW Jr, Bell NH, Ettinger MP, Walsh BW, Favus MJ, Mako B, et al.Comparison of alendronate and intranasal calcitonin for treatment of osteoporosis in postmenopausal women. Journal of Clinical Endocrinology & Metabolism 2000;85(5):17838. Evio 2004 {published data only} Evio S, Tiitinen A, Laitinen K, Ylikorkala O, Valimaki MJ. Effects of alendronate and hormone replacement therapy, alone and in combination, on bone mass and markers of bone turnover in elderly women with osteoporosis. Journal of Clinical Endocrinology & Metabolism 2004;89(2):62631. Gonnelli 2002 {published data only} Gonnelli S, Cepollaro C, Montagnani A, Martini S, Gennari L, Mangeri M, et al.Heel ultrasonography in monitoring alendronate therapy: a four-year longitudinal study. Osteoporosis International 2002;13(5):41521.
Cummings 2000 {published data only} Cummings SR, Palermo L, Browner W, Marcus R, Wallace R, Pearson J, et al.Monitoring osteoporosis therapy with bone densitometry: misleading changes and regression to the mean. Fracture Intervention Trial Research Group. JAMA 2000;283(10):131821.
IN
TE
Chesnut 1993 {published data only} Chesnut CH, Harris ST. Short term effect of alendronate on bone mass and bone remodeling in postmenopausal women. Osteoporosis International 1993;3(Suppl 3):S17S19.
AL
Davas 2003 {published data only} Davas I, Altintas A, Yoldemir T, Varolan A, Yazgan A, Baksu B. Effect of daily hormone therapy and alendronate use on bone mineral density in postmenopausal women. Fertility & Sterility 2003;80(3):53640. Devogelaer 1996 {published data only} Devogelaer JP, Broll H, Correa-Rotter R, Cumming DC, De Deuxchaisnes CN, Geusens P, et al.Oral alendronate induces progressive increases in bone mass of the spine, hip, and total body over 3 years in postmenopausal women with osteoporosis. [erratum appears in Bone 1996 Jul;19(1):78]. Bone 1996;18(2):14150.
SE
Greenspan 2002a {published data only} Greenspan SL, Emkey RD, Bone HG, Weiss SR, Bell NH, Downs RW, et al.Signicant differential effects of alendronate, estrogen or combination therapy on the rate of bone loss after discontinuation of treatment of postmenopausal osteoporosis. A randomized, double-blind, placebo-controlled trial. [summary for patients in Ann Intern Med. 2002 Dec 3;137(11):I31; PMID: 12459003]. Annals of Internal Medicine 2002;137(11):87583. Greenspan 2002b {published data only} Greenspan SL, Schneider DL, McClung MR, Miller PD, Schnitzer TJ, Bonin R, et al.Alendronate improves bone mineral density in elderly women with osteoporosis residing in long-term care facilities. A randomized, doubleblind, placebo-controlled trial. [summary for patients in Ann Intern Med. 2002 May 21;136(10):I54; PMID: 12020160]. Annals of Internal Medicine 2002;136(10): 7426. Greenspan 2003 {published data only} Greenspan SL, Resnick NM, Parker RA. Combination therapy with hormone replacement and alendronate for prevention of bone loss in elderly women: a randomized controlled trial. JAMA 2003;289(19):252533. Harris 1993 {published data only} Harris ST, Gertz BJ, Genant HK, Eyre DR, Survill TT, Ventura JN, et al.The effect of short term treatment with alendronate on vertebral density and biochemical markers of bone remodeling in early postmenopausal women. Journal of Clinical Endocrinology & Metabolism 1993;76(6): 13991406.
37
O N
LY
Heijckmann 2002 {published data only} Heijckmann AC, Juttmann JR, Wolffenbuttel BH. Intravenous pamidronate compared with oral alendronate for the treatment of postmenopausal osteoporosis. [see comment]. Netherlands Journal of Medicine 2002;60(8): 3159. Ho 2005 {published data only} Ho AY, Kung AW, Ho AYY, Kung AWC. Efcacy and tolerability of alendronate once weekly in Asian postmenopausal osteoporotic women. Annals of Pharmacotherapy 2005;39(9):142833. [MEDLINE: 109] Hochberg 1999 {published data only} Hochberg MC, Ross PD, Black D, Cummings SR, Genant HK, Nevitt MC, et al.Larger increases in bone mineral density during alendronate therapy are associated with a lower risk of new vertebral fractures in women with postmenopausal osteoporosis. Fracture Intervention Trial Research Group. Arthritis & Rheumatism 1999;42(6): 124654. Hosking 2003 {published data only} Hosking D, Adami S, Felsenberg D, Andia JC, Valimaki M, Benhamou L, et al.Comparison of change in bone resorption and bone mineral density with once-weekly alendronate and daily risedronate: a randomised, placebocontrolled study. Current Medical Research & Opinion 2003; 19(5):38394. Iwamoto 2004 {published data only} Iwamoto J, Takeda T, Sato Y, Uzawa M. Determinants of one-year response of lumbar bone mineral density to alendronate treatment in elderly Japanese women with osteoporosis. Yonsei Medical Journal 2004;45(4):67682.
mg daily in the treatment of postmenopausal osteoporosis. Menopause 2004;11(4):40515. Malavolta 1999 {published data only} Malavolta N, Zanardi M, Veronesi M, Ripamonti C, Gnudi S. Calcitriol and alendronate combination treatment in menopausal women with low bone mass. International Journal of Tissue Reactions 1999;21(2):519. McClung 1998 {published data only} McClung M, Clemmesen B, Daifotis A, Gilchrist NL, Eisman J, Weinstein RS, et al.Alendronate prevents postmenopausal bone loss in women without osteoporosis. A double-blind, randomized, controlled trial. Alendronate Osteoporosis Prevention Study Group. [see comment]. Annals of Internal Medicine 1998;128(4):25361. McClung 2004 {published data only} McClung MR, Wasnich RD, Hosking DJ, Christiansen C, Ravn P, Wu M, et al.Prevention of postmenopausal bone loss: six-year results from the Early Postmenopausal Intervention Cohort Study. Journal of Clinical Endocrinology & Metabolism 2004;89(10):487985. [MEDLINE: 190] Murphy 2001 {published data only} Murphy MG, Weiss S, McClung M, Schnitzer T, Cerchio K, Connor J, et al.Effect of alendronate and MK-677 (a growth hormone secretagogue), individually and in combination, on markers of bone turnover and bone mineral density in postmenopausal osteoporotic women. Journal of Clinical Endocrinology & Metabolism 2001;86(3):111625.
Kung 2000 {published data only} Kung AW, Yeung SS, Chu LW. The efcacy and tolerability of alendronate in postmenopausal osteoporotic Chinese women: a randomized placebo-controlled study. Calcied Tissue International 2000;67(4):28690.
IN
TE
Johnell 2002 {published data only} Johnell O, Scheele WH, Lu Y, Reginster JY, Need AG, Seeman E. Additive effects of raloxifene and alendronate on bone density and biochemical markers of bone remodeling in postmenopausal women with osteoporosis. [see comment]. Journal of Clinical Endocrinology & Metabolism 2002;87(3):98592.
AL
Kushida 2004 {published data only} Kushida K, Shiraki M, Nakamura T, Kishimoto H, Morii H, Yamamoto K, et al.Alendronate reduced vertebral fracture risk in postmenopausal Japanese women with osteoporosis: a 3-year follow-up study. Journal of Bone & Mineral Metabolism 2004;22(5):4628. Lau 2000 {published data only} Lau EM, Woo J, Chan YH, Grifth J. Alendronate prevents bone loss in Chinese women with osteoporosis. Bone 2000; 27(5):67780. Luckey 2004 {published data only} Luckey M, Kagan R, Greenspan S, Bone H, Kiel RD, Simon J, et al.Once-weekly alendronate 70 mg and raloxifene 60
SE
Nenonen 2005 {published data only} Nenonen A, Cheng S, Ivaska KK, Alatalo SL, Lehtimaki T, Schmidt-Gayk H, et al.Serum TRACP 5b is a useful marker for monitoring alendronate treatment: Comparison with other markers of bone turnover. Journal of Bone & Mineral Research 2005;20(8):180412. [MEDLINE: 576] Palomba 2002 {published data only} Palomba S, Orio F Jr, Colao A, di Carlo C, Sena T, Lombardi G, et al.Effect of estrogen replacement plus lowdose alendronate treatment on bone density in surgically postmenopausal women with osteoporosis. Journal of Clinical Endocrinology & Metabolism 2002;87(4):15028. Payer 2000 {published data only} Payer J Jr, Killinger Z, Masaryk P, Tomkova S, Kmecova Z, Ondrejkova J, et al.[Effect of alendronate therapy on bone turnover--results of a multicenter study]. [Slovak]. Vnitrni Lekarstvi 2000;46(10):68992. Ravn 1999a {published data only} Ravn P, Clemmesen B, Christiansen C. Biochemical markers can predict the response in bone mass during alendronate treatment in early postmenopausal women. Alendronate Osteoporosis Prevention Study Group. Bone 1999;24(3): 23744. Ravn 1999b {published data only} Ravn P, Hosking D, Thompson D, Cizza G, Wasnich RD, McClung M, et al.Monitoring of alendronate treatment and prediction of effect on bone mass by biochemical markers in the early postmenopausal intervention cohort study. Journal of Clinical Endocrinology & Metabolism 1999;84(7):23638.
38
O N
LY
Ravn 1999c {published data only} Ravn P, Bidstrup M, Wasnich RD, Davis JW, McClung MR, Balske A, et al.Alendronate and estrogen-progestin in the long-term prevention of bone loss: four-year results from the early postmenopausal intervention cohort study. A randomized, controlled trial. [see comment]. Annals of Internal Medicine 1999;131(12):93542. Ravn 2000 {published data only} Ravn P, Weiss SR, Rodriguez-Portales JA, McClung MR, Wasnich RD, Gilchrist NL, et al.Alendronate in early postmenopausal women: effects on bone mass during long-term treatment and after withdrawal. Alendronate Osteoporosis Prevention Study Group. Journal of Clinical Endocrinology & Metabolism 2000;85(4):14927. Rhee 2006 {published data only} Rhee Y, Kang M, Min Y, Byun D, Chung Y, Ahn C, et al.Effects of a combined alendronate and calcitriol agent (Maxmarvil) on bone metabolism in Korean postmenopausal women: a multicenter, double-blind, randomized, placebo-controlled study. Osteoporosis International 2006;17(12):18017. [MEDLINE: 7] Rittmaster 2000 {published data only} Rittmaster RS, Bolognese M, Ettinger MP, Hanley DA, Hodsman AB, Kendler DL, et al.Enhancement of bone mass in osteoporotic women with parathyroid hormone followed by alendronate. [see comment]. Journal of Clinical Endocrinology & Metabolism 2000;85(6):212934. Rizzoli 2002 {published data only} Rizzoli R, Greenspan SL, Bone G III, Schnitzer TJ, Watts NB, Adami S, et al.Two-year results of once-weekly administration of alendronate 70 mg for the treatment of postmenopausal osteoporosis. Journal of Bone & Mineral Research 2002;17(11):198896.
Sambrook 2004a {published data only} Sambrook PN, Geusens P, Ribot C, Solimano JA, FerrerBarriendos J, Gaines K, et al.Alendronate produces greater effects than raloxifene on bone density and bone turnover in postmenopausal women with low bone density: results of EFFECT (Efcacy of FOSAMAX versus EVISTA Comparison Trial) International. Journal of Internal Medicine 2004;255(4):50311. Sambrook 2004b {published data only} Sambrook PN, Rodriguez JP, Wasnich RD, Luckey MM, Kaur A, Meng L, et al.Alendronate in the prevention of osteoporosis: 7-year follow-up. Osteoporosis International 2004;15(6):4838. Sawka 2003 {published data only} Sawka AM, Adachi JD, Ioannidis G, Olszynski WP, Brown JP, Hanley DA, et al.What predicts early fracture or bone loss on bisphosphonate therapy?. Journal of Clinical Densitometry 2003;6(4):31522. Schneider 1999 {published data only} Schneider PF, Fischer M, Allolio B, Felsenberg D, Schroder U, Semler J, et al.Alendronate increases bone density and bone strength at the distal radius in postmenopausal women. Journal of Bone & Mineral Research 1999;14(8):138793.
AL
Rossini 2000 {published data only} Rossini M, Gatti D, Girardello S, Braga V, James G, Adami S. Effects of two intermittent alendronate regimens in the prevention or treatment of postmenopausal osteoporosis. Bone 2000;27(1):11922. Rozkydal 2003 {published data only} Rozkydal Z, Janicek P. The effect of alendronate in the treatment of postmenopausal osteoporosis. Bratislavske Lekarske Listy 2003;104(10):30913. Sahota 2000 {published data only} Sahota O, Fowler I, Blackwell PJ, Lawson N, Cawte SA, San P, et al.A comparison of continuous alendronate, cyclical alendronate and cyclical etidronate with calcitriol in the treatment of postmenopausal vertebral osteoporosis: a randomized controlled trial. Osteoporosis International 2000;11(11):95966.
IN
Rossini 1994 {published data only} Rossini M, Gatti D, Zamberlan N, Braga V, Dorizzi R, Adami S. Long-term effects of a treatment course with oral alendronate of postmenopausal osteoporosis. Journal of Bone & Mineral Research 1994;9(11):18337.
TE
SE
Schnitzer 2000 {published data only} Schnitzer T, Bone HG, Crepaldi G, Adami S, McClung M, Kiel D, et al.Therapeutic equivalence of alendronate 70 mg once-weekly and alendronate 10 mg daily in the treatment of osteoporosis. Alendronate Once-Weekly Study Group. Aging-Clinical & Experimental Research 2000;12(1):112.
Seeman 1999 {published data only} Seeman E. The antifracture efcacy of alendronate. International Journal of Clinical Practice 1999;Supplement. 101:405. Simon 2002 {published data only} Simon JA, Lewiecki EM, Smith ME, Petruschke RA, Wang L, Palmisano JJ. Patient preference for once-weekly alendronate 70 mg versus once-daily alendronate 10 mg: a multicenter, randomized, open-label, crossover study. Clinical Therapeutics 2002;24(11):187186. Sosa 2002 {published data only} Sosa M, Hernandez D, Segarra MC, Gomez A, de la PE, Betancor P. Effect of two forms of alendronate administration upon bone mass after two years of treatment. Journal of Clinical Densitometry 2002;5(1):2734. Stepan 1999 {published data only} Stepan JJ, Vokrouhlicka J. Comparison of biochemical markers of bone remodelling in the assessment of the effects of alendronate on bone in postmenopausal osteoporosis. Clinica Chimica Acta 1999;288(1-2):12135. Tiras 2000 {published data only} Tiras MB, Noyan V, Yildiz A, Yildirim M, Daya S. Effects of alendronate and hormone replacement therapy, alone or in combination, on bone mass in postmenopausal women with osteoporosis: a prospective, randomized study. Human Reproduction 2000;15(10):208792.
39
O N
LY
Tucci 1996 {published data only} Tucci JR, Tonino RP, Emkey RD, Peverly CA, Kher U, Santora AC. Effect of three years of oral alendronate treatment in postmenopausal women with osteoporosis. American Journal of Medicine 1996;101(5):488501. Tutuncu 2005 {published data only} Tutuncu L, Arslanhan N, Mungen E, Yilmaz S, Yergok YZ. Combined usage of raloxifene and alendronate: Effects on biochemical markers. [Turkish]. Nobel Medicus 2005; Vol. 1, issue 2. [MEDLINE: 519] Uusi-Rasi 2003 {published data only} Uusi-Rasi K, Kannus P, Cheng S, Sievanen H, Pasanen M, Heinonen A, et al.Effect of alendronate and exercise on bone and physical performance of postmenopausal women: a randomized controlled trial. Bone 2003;33(1):13243. van der Poest 2000 {published data only} van der Poest CE, Patka P, Vandormael K, Haarman H, Lips P. The effect of alendronate on bone mass after distal forearm fracture. Journal of Bone & Mineral Research 2000; 15(3):58693. Vasikaran 1995 {published data only} Vasikaran SD, Khan S, McCloskey EV, Kanis JA. Sustained response to intravenous alendronate in postmenopausal osteoporosis. Bone 1995;17(6):51720. Yen 2000 {published data only} Yen ML, Yen BL, Jang MH, Hsu SH, Cheng WC, Tsai KS. Effects of alendronate on osteopenic postmenopausal Chinese women. Bone 2000;27(5):6815.
randomized trial. 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):292738. Brown 2002 Brown JP, Josse RG. 2002 clinical practice guidelines for the diagnosis and management of osteoporosis in Canada. CMAJ 2002;167(10 Suppl):S134. Browner 1996 Browner WS, Pressman AR, Nevitt MC, Cummings SR. Mortality following fractures in older women. The study of osteoporotic fractures. Archives of Internal Medicine 1996; 156(14):15215. CADTH 2006 Coyle D, Hadj Tahar A, Murphy G, Perras C, Skidmore B, Boucher M, et al.Teriparatide and bisphosphonates of treatment of osteoporosis in women: a clinical and economic analysis. Canadian Agency for Drugs and Technologies in Health 2006. Cates 2004 Cates C. Visual Rx NNT Calculator 2.0. Dr. Chris Cates EBM website. Available from: URL: http:// www.nntonline.net/, 2004. Cauley 2000 Cauley JA, Thompson DE, Ensrud KC, Scott JC, Black D. Risk of mortality following clinical fractures. Osteoporosis International 2000;11(7):55661.
Black 2000b Black DM, Thompson DE, Bauer DC, Ensrud K, Musliner T, Hochberg MC. Fracture risk reduction with alendronate in women with osteoporosis: the Fracture Intervention Trial. FIT Research Group. J Clin Endocrinol Metab 2000; 85(11):411824.
IN
TE
Additional references
Yildirim 2005 {published data only} Yildirim K, Gureser G, Karatay S, Melikoglu MA, Ugur M, Erdal A, et al.Comparison of the effects of alendronate, risedronate and calcitonin treatment in postmenopausal osteoporosis. Journal of Back & Musculoskeletal Rehabilitation 2005;18(3-4):859. [MEDLINE: 540]
AL
Black 2001 Black DM, Steinbuch M, Palermo L, Dargent-Molina P, Lindsay R, Hoseyni MS, et al.An assessment tool for predicting fracture risk in postmenopausal women. Osteoporosis International 2001;12(7):51928. Black 2006 Black DM, Schwartz AV, Ensrud KE, Cauley JA, Levis S, Quandt SA, et al.Effects of continuing or stopping alendronate after 5 years of treatment: the Fracture Intervention Trial Long-term Extension (FLEX): a randomized trial. Effects of continuing or stopping alendronate after 5 years of treatment: the Fracture Intervention Trial Long-term Extension (FLEX): a
SE
Cooper 1993 Cooper C, Atkinson EJ, Jacobsen SJ, OFallon WM, Melton LJ III. Population-based study of survival after osteoporotic fractures. American Journal of Epidemiology 1993;137(9): 10015. Cranney 2002 Cranney A, Wells G, Willan A, Grifth L, Zytaruk N, Robinson V, et al.Meta-analyses of therapies for postmenopausal osteoporosis. II. Meta-analysis of alendronate for the treatment of postmenopausal women. Endocrine Reviews 2002;23(4):50816. Cummings 1989 Cummings SR, Black DM, Rubin SM. Lifetime risks of hip, Colles, or vertebral fracture and coronary heart disease among white postmenopausal women. Archives of Internal Medicine 1989;149(11):24458. Dickersin 1994 Dickersin K, Scherer R, Lefebvre C. Identifying relevant studies for systematic reviews. BMJ 1994;309(6964): 128691. Doherty 2001 Doherty DA, Sanders KM, Kotowicz MA, Prince RL. Lifetime and ve-year age-specic risks of rst and subsequent osteoporotic fractures in postmenopausal women. Osteoporosis International 2001;12(1):1623. Egger 2003 Egger M, Juni P, Bartlett C, Holenstein F, Sterne J. How important are comprehensive literature searches and the
40
O N
LY
assessment of trial quality in systematic reviews? Empirical study. Health Technology Assessment 2003;7(1):176. Fleiss 1993 Fleiss JL. The statistical basis of meta-analysis. Statistical Methods in Medical Research 1993;2(2):12145. GRADE 2004 Atkins D, Best D, Briss PA, Eccles M, Falck-Ytter Y, Flottorp S, et al.Grading quality of evidence and strength of recommendations.. BMJ 2004;328:1490. Hanley 2003 Hanley DA. Osteoporosis. In: Gray J editor(s). Therapeutic Choices. 4th Edition. Ottawa: Canadian Pharmaceutical Association, 2003:63746. Harris 1999 Harris ST, Watts NB, Genant HK, McKeever CD, Hangartner T, Keller M, et al.Effects of risedronate treatment on vertebral and nonvertebral fractures in women with postmenopausal osteoporosis: a randomized controlled trial. Vertebral Efcacy With Risedronate Therapy (VERT) Study Group. JAMA 1999;282(14):134452. Health Canada 2005 Health Canada, editors. Canadian Adverse Drug Reaction Monitoring Program. Alendronate. Adverse Reaction Database. Ottawa,. Ottawa, 2005. Health Canada 2005c Health Canada. Medeffect [website]. Health Canada. Ottawa, 2005. Higgins 2005 Higgins JPT, Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions 4.2.5 [updated May 2005]. In: The Cochrane Library, Issue 3, 2005. Chichester, UK: John Wiley & Sons, Ltd.
jaws in Australia. Journal of Oral and Maxillofacial Surgery 2007;65:41523. Mayo Clin Proc 2005 Alendronate and vertebral fracture risk [multiple letters].. Mayo Clin Proc YR:2005. 80 Vol. 80:123341. McClung 2001 McClung MR, Geusens P, Miller PD, Zippel H, Bensen WG, Roux C, et al.Effect of risedronate on the risk of hip fracture in elderly women. Hip Intervention Program Study Group. [see comment]. New England Journal of Medicine 2001;344(5):33340. Melton 1989 Melton LJ III, Kan SH, Frye MA, Wahner HW, OFallon WM, Riggs BL. Epidemiology of vertebral fractures in women. American Journal of Epidemiology 1989;129(5): 100011. NIH Consensus 2001 NIH Consensus Development Panel on Osteoporosis Prevention, Diagnosis, Therapy. Osteoporosis prevention, diagnosis, and therapy. JAMA 2001;285(6):78595.
AL
TE
Hodsman 2002 Hodsman AB, Hanley DA, Josse R. Do bisphosphonates reduce the risk of osteoporotic fractures? An evaluation of the evidence to date. CMAJ 2002;166(11):142630.
IN
Kanis 1994 Kanis JA, Melton LJ, Christiansen C, Johnston CC, Khaltaev N. The diagnosis of osteoporosis. Journal of Bone and Mineral Research 1994;9(8):113741. Kherani 2002 Kherani RB, Papaioannou A, Adachi JD. Long-term tolerability of the bisphosphonates in postmenopausal osteoporosis: a comparative review. Drug Safety 2002;25 (11):78190. Khosla 2007 Khosla S, Burr D, Cauley J, Dempster DW, Ebeling PR, Felsenberg D, et al.Bisphosphonate-associated osteonecrosis of the jaw: report of a task force of the American Society for Bone and Mineral Research. Journal of Bone and Mineral Research 2007;22(Khosla 2007):147991. Mavrokokki 2007 Mavrokokki T, Cheng A, Stein B, Gross A. Nature and frequency of bisphosphonate-associated osteonecrosis of the
SE
Rodan 1993 Rodan GA, Seedor JG, Balena R. Preclinical phamacology of alendronate. Osteoporosis International 1993;3(Suppl 3): S712. Sampson 2003 Sampson M, Barrowman NJ, Moher D, Klassen TP, Pham B, Platt R, et al.Should meta-analysts search Embase in addition to Medline?. Journal of Clinical Epidemiology 2003; 56(10):94355. Shukla 2003 Shukla V. Treating osteoporosis with teriparatide: many unknowns? [Issues in emerging health technologies issue 51]. Canadian Coordinating Ofce of Health Technology Assessment.. Ottawa: Canadian Coordinating Ofce of health Technology Assessment, 2003. Tugwell 2004 Tugwell P, Shea B, Boers M, Brooks P, Simon S, Strand V, Wells G, editors. In: Tugwell P, Shea B, Boers M, Brooks P, Simon S, Strand V, Wells G editor(s). Evidence-Based Rheumatology. London: BMJ Books, 2004. WHO 1994 Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: Report of a WHO study group. World Health Organ Tech Rep Ser 1994; Vol. 843: 1129. Woo 2006 Woo SB, Hellstein JW, Kalmar JR. Narrative [corrected] review: bisphosphonates and osteonecrosis of the jaws. Annals of Internal Medicine 2006;144(10):75361.
References to other published versions of this review

CADTH 2006b Wells GA, Cranney A, Bouchere M, Peterson J, Shea B, Robinson V, et al.Bisphosphonates for the primary
41
O N
LY
and secondary prevention of osteoporotic fractures in postmenopausal women: a meta-analysis [Technology report no 69]. Ottawa: Canadian Agency for Drugs and Technologies in Health 2006. Indicates the major publication for the study
IN
TE
AL
SE
O N
LY
42
CHARACTERISTICS OF STUDIES Characteristics of included studies [ordered by study ID]

Ascott Evans 2003 Methods Randomized controlled trial Primary prevention Duration: 1 year Blinding: double blind matching placebo, all study personnel blinded Withdrawals: Alendronate: 12/95 (12.6%) Placebo: 13/49 (26.5%) Total: 25/144 (17.4%) Source: 18 centres in 9 countries. Inclusion Criteria: Women under 80 yrs old who had been postmenopausal for at least 3 years. Previous HRT for at least 1 yr which had been discontinued at least 3 months prior to study. Low bone density between -3.5 and -1.5 of young normal Exclusion Criteria: History of osteoporotic fracture or metabolic bone disease. Recent bisphosphonate or other treatment known to affect bone metabolism Treatment N = 95 Control N = 49 Age: 67.3 (6.6); YSM: 11.5 (7.3) Calcium: not reported BMD: not reported T-score: -2.27 (0.65) Vertebral Fractures: 0% Alendronate 10 mg x 1 year vs placebo (Calcium 500mg/day) Vertebral, Non Vertebral, Hip and Wrist Fractures: Adverse experiences were recorded by blinded study personnel at each visit using non leading questions. No fractures were reported
Participants
Interventions
Outcomes
Notes
Risk of bias Bias Allocation concealment (selection bias) Authors judgement Unclear risk Support for judgement B - Unclear
IN
TE
AL
SE
O N
LY
43
Black 1996 Methods Randomized controlled trial Secondary prevention Blinding: double blind identical placebo, blinded radiologist Duration: 3 years Withdrawals: Alendronate (available radiographs): 44 (4.2%) Placebo (available radiographs): 37/1005 (3.7%) Total (available radiographs): 81/2027 (4.0%) Total (lost to follow up at close out): 78/2027 (3.8%) Source: Population-based listings in 11 metropolitan areas of the USA. Fracture Intervention Trial Inclusion Criteria: Age 55-81, postmenopausal for at least 2 years, femoral neck BMD 0.68 g/cm Hologic (2.1 SD below peak bone mass) or less. Exclusion Criteria: Peptic ulcer disease (single hospital admission for upper GI bleeding or 2 or more documented ulcers in previous 5 years), dyspepsia requiring daily treatment, abnormal renal function, major medical problems, severe malabsorption, uncontrolled hypertension, MI in previous 6 months, unstable angina, thyroid or parathyroid dysfunction, estrogen or calcitonin in previous 6 months, bisphosphonates or uorides at any time Treatment N = 1022 Control N = 1005 Age: 71.0 (5.6); YSM: not reported Calcium: 636 (407) mg/day BMD (hip): 0.57 g/cm 2 (0.07) T-score (hip): -3.3 Vertebral Fractures : 100% Alendronate 5 mg x 2 years then 10 mg x 1 year vs placebo (If intake < 1000 mg then received 500 mg Ca and 250 IU Vitamin D) Morphometric Vertebral Fractures: Lateral radiographs were taken at baseline 24 and 36 months intervals. Morphometry was performed with a translucent digitizer and cursor marking anterior, posterior and middle heights for each vertebra. Baseline fractures were dened a height of > 3 SD below the mean population level for that vertebrae. Incident fractures were dened as a decrease of 20% and at least 4 mm from baseline. Any questionable fractures were reviewed by the study radiologist. Technicians and radiologist were all blinded Clinical Vertebral Fractures: Fractures that came to medical attention and were reported by participants. Copy of radiograph was obtained and compared with baseline study radiograph. Incident clinical fracture was dened by a semiquantitative reading by the study radiologist Non Vertebral, Hip and Wrist Fractures: Clinical fractures were initially reported by participants and conrmed by a written radiological report. Excluded pathological fractures e.g. malignancies, excessive trauma, face and skull
Outcomes
Notes Risk of bias
IN
TE
Interventions
AL
SE
O N
LY
Participants
44
Black 1996
(Continued)
Bias Allocation concealment (selection bias)
Authors judgement Low risk
Support for judgement A - Adequate
Bone 1997 Methods Randomized controlled trial Secondary prevention Blinding: double blind Duration: 2 years Withdrawals: Total (for BMD analysis): 131/359 (36.5%)
Participants
Interventions
TE
Source: 15 clinical sites in USA. Subjects stratied so that 2/3 would be age 70-85 Inclusion Criteria: Women age 60-85 in good health apart from osteoporosis. Lumbar spine < -2.00 SD of peak bone mass (BMD 0.824g/cm or less by Hologic DXA or 9.44 g/cm or less by Lunar DXA) Exclusion Criteria: More than 1 spinal crush fracture or spinal anatomy otherwise unsuitable for DXA, history of recent major GI disease - peptic ulcer, esophageal ulcer, malabsorption, use of drug to inhibit gastric acid secretion for > 2 wks, chronic NSAID therapy, agents known to affect bone metabolism, unstable dose thyroid hormone replacement, uncorrected vitamin D deciency. Treatment N = 86, 89, 93 Control N = 91 Age: 70.4 (5.6); YSM: 24.2 (9.9) Calcium: 891(629) mg/day BMD: 0.71 g/cm 2 (0.08) T-score: -3.1 Vertebral Fractures: 37.4% Alendronate 1, 2.5, or 5 mg placebo. (500 mg calcium/day) Vertebral Fractures: Lateral thoracic and lumbar radiographs obtained at baseline and annual visits were sent to a central evaluation facility where they were evaluated by a single radiologist. Prevalent and incident fractures were scored using a semiquantitative scale as being intact (unfractured or questionably fractured), or fractured (mild - 20 to 25% height loss; moderate - 25-40% or severe - >40%) Non-vertebral Fractures: Reported at each centre based on clinical presentation and conrmatory radiographs
Outcomes
Notes Risk of bias Bias Authors judgement Support for judgement
IN
AL
SE
O N
LY
45
Bone 1997
(Continued)
Allocation concealment (selection bias)
Low risk
A - Adequate
Chestnut 1995 Methods Randomized controlled trial Secondary prevention Blinding: double blind unspecied Duration: 2 years Withdrawals: Total: 34/188 (18%) Loss to follow up: 34/188 (18.09%)
Participants
Interventions
Outcomes
TE
Notes Risk of bias Bias Allocation concealment (selection bias) Authors judgement Unclear risk Support for judgement B - Unclear
IN
AL
Source: Recruited by advertisements and medical announcements through seven centres in the USA Inclusion Criteria: Healthy women age 42 to 75 who were at least 5 years postmenopausal. BMD = 2 SD below young normal (0.88g/cm) Exclusion Criteria: Presence of spine or hip fractures attributable to osteoporosis. Any disease or drug therapy potentially affecting bone metabolism Treatment N = 32, 30, 32, 32 Control N = 31 Age: 63.04 (6.27); YSM: 15.6 (7.3) Calcium: 853 (516) mg/day BMD: 0.75 g/cm 2 (0.09) T-score: -2.7 Vertebral Fractures : 0% Alendronate 5mg/day, or 10 mg/day, for 2 years 20 mg/day or 40 mg/day for 1 year followed by 1 year of placebo or 40 mg for 3 months followed by 2.5 mg for 21 months vs placebo for 2 years (500 mg calcium/day) Vertebral Fractures: Lateral thoracic and lumbar radiographs were evaluated at each centre for presence of prevalent and incident fractures at baseline and completion of treatment Non-vertebral Fractures: Ascertainment not specied but patients were questioned about intercurrent health problems at each visit. This outcome was not included in the metaanalysis as non vertebral fractures werent broken down by study group
SE
O N
LY
46
Cummings 1998 Methods Randomized controlled trial Primary prevention Blinding: double blind, collection and review of data was blinded Study duration: 4 years Withdrawals: Total (lost to follow at close out) 160/4432 (3.6 %) Source: Recruited principally though mass mailings from 11 community-based clinical research centres in the USA. Fracture Intervention Trial Inclusion Criteria: Age 55-80, postmenopausal for at least 2 years, femoral neck BMD 0.68 g/cm or less Hologic. At the time of study this was thought to correspond to -2 SD below peak mass but subsequently found to correspond to a t-score of -1.6 based on the Third National Health and Nutritional Examination Survey. Consequently 1/3 of participants had higher BMD than expected Exclusion Criteria: Vertebral fractures, Peptic ulcer disease (single hospital admission for upper GI bleeding or 2 or more documented ulcers in previous 5 years), dyspepsia requiring daily treatment, abnormal renal function, major medical problems, severe malabsorption, uncontrolled hypertension, MI in previous 6 mos, unstable angina, thyroid or parathyroid dysfunction, estrogen or calcitonin in previous 6 mos, bisphosphonates or uorides at any time Treatment N = 2214 Control N = 2218 Age : 67.6 (6.1); YSM: not reported Calcium: 636 (400) mg/day BMD: 0.84 g/cm2 (0.13) T-score: -1.9 Vertebral Fractures : 0% Alendronate 5 mg for 2 yrs then increased to 10 mg for 2 years vs placebo (If intake < 1000 mg then received 500 mg Ca and 250 IU Vitamin D) Vertebral Morphometric Fractures: Lateral spine radiographs were obtained at baseline and 4 years. An incident fracture was dened as a decrease of 20% and 4 mm or more in any vertebral height which was conrmed by repeat measurement. All assessments were blinded Clinical Fractures: Dened as a fractured diagnosed by a physician. Self reports were conrmed by written reports of radiographs or other tests. Excluded pathologic fractures, trauma sufcient to fracture a young adult bone, facial and skull fractures
Participants
Interventions
Outcomes
Notes Risk of bias Bias Allocation concealment (selection bias) Authors judgement Low risk Support for judgement A - Adequate
IN
TE
AL
SE
O N
LY
47
Durson 2001 Methods Randomized controlled trial Secondary prevention Blinding: no blinding reported Duration: 1 year Withdrawals: Alendronate: 13/51 (25.5%) Control: 15/50 (30.0%) Total: 28/101 (28%) Source: Postmenopausal women applying to one centres department of physical medicine and rehabilitation in Turkey Inclusion Criteria: BMD of 2 SD or more below young adult mean at either lumbar spine or femoral neck Exclusion Criteria: Drug or alcohol abuse, bone metabolism disorder, active GI or liver disease, renal failure or calculi, treatment with specic therapy for osteoporosis, corticosteroids, malignancy, disorder of calcium metabolism, lumbar vertebrae abnormalities preventing evaluation of BMD. Treatment N = 51 Control N = 50 Age: 61 (7.8); YSM: 15.59 (8.04) Calcium: not reported BMD: 0.84 g/cm (0.08) T-score: -1.9 Vertebral Fractures: not reported
Participants
Interventions Outcomes
Notes Risk of bias Bias
IN
TE
AL
Authors judgement Unclear risk
Alendronate 10 mg/day plus calcium 1000 mg/day x 1 yr vs calcium 1000 mg Vertebral Fractures: Lateral and anteroposterior X-rays of thoracic and lumbar vertebrae were performed at baseline, 6 months and 12 months. A new vertebral fracture was dened as a decrease of 20% and at least 4 mm of height in any vertebrae
SE
O N
LY
B - Unclear
Support for judgement
48
Greenspan 1998 Methods Randomized controlled trial Secondary prevention Blinding: double blind, matching placebo Study duration: 2.5 years Loss to follow up: Alendronate:14/60 (23.3%) Placebo:15/60 (25.0%) Total: 29/120 (24.2%) Source: Unselected women from one city (Boston) in the USA were recruited by advertisement Inclusion Criteria: Healthy ambulatory community dwelling age 65 or older. Criteria not based on BMD Exclusion Criteria: History of any illness affecting bone and mineral metabolism - (renal, malignancy, hyperthyroidism, hyperparathyroidism, malabsorption), medications affecting bone metabolism, treatment for osteoporosis (bisphosphonates, HRT, calcitonin) within 1 year. Treatment N = 60 Control N = 60 Age: 70 (4.6); YSM: not reported Calcium: 719 (465) mg/day BMD: 0.57 g/cm2 (0.11) T-score: -4.3 Vertebral Fractures: not reported
Participants
Interventions
Outcomes Notes Risk of bias Bias
TE
AL
Alendronate 5 mg for year 1, 10 mg for year 2 vs placebo, (if Ca intake < 1000 mg - 250 mg Ca and/or 125 IU vitamin D/day) Non-vertebral Hip and Wrist: Ascertainment not described.
SE
IN
Greenspan 2002 Methods Randomized controlled trial Secondary prevention Blinding: double blind matching placebo Duration: 2 years Withdrawals: Not reported Source: Female residents in long term care facilities in 25 centre in USA Inclusion Criteria: Ambulatory, age 65 or older with BMD T-score of -2 or less at lumbar
49
Participants
O N
LY
B - Unclear
Greenspan 2002
(Continued)
Interventions
Outcomes Notes Risk of bias Bias Allocation concealment (selection bias)
Hip Fractures: Ascertainment not reported.
SE
Support for judgement B - Unclear
Hosking 1998 Methods
Participants
IN
TE
AL
Randomized controlled trial Primary prevention Blinding: double blind, blinded BMD measurement and analysis Duration: 2 years Withdrawals: Alendronate:2.5mg 92/499 (18.4%) Alendronate 5mg: 102/498 (20.5%) Placebo: 93/502 (18.5%) Total: 287/1499 (19.1%) Withdrawals Source: Recruited by direct mailing, advertisements or telephone. Multicentre - USA, UK and Denmark Inclusion Criteria: Postmenopausal at least 6 months (conrmed by FSH) and in good health. Only 10% of women at each centre were allowed to have a lumbar-spine BMD below 0.8 g/square metre (DEXA). Exclusion Criteria: Abnormal renal function, cancer, peptic ulcer or esophageal disease requiring prescription medication within the previous ve years, previous treatment with bisphosphonate or uoride, therapy with phosphate-binding antacid, HRT within previous 3 months, therapy with any drug which affects the skeleton.
50
O N
Alendronate 10 mg/day x 2yrs vs placebo (Vitamin D 400 IU/day and if dietary calcium was < 1500 mg/day they received calcium 500mg/day.)
LY
spine or hip. Exclusion Criteria: Disorders of bone mineralization, 25-hydroxycholecalciferol < 25 nmol/ L, untreated hyperthyroidism, recent major upper GI mucosal erosive disease, or use of bone active agents Treatment N = 163 Control N = 164 Age: 78.5 (range 65-91); YSM: not reported Calcium: not reported BMD: not reported T-score: (mean range hip and spine) -3.5 to-2.4 Fractures: (history of any) 55%
Hosking 1998
(Continued)
Treatment N = 499, 498 Control N = 502 Age: 53 (4); YSM: 6 (5) Calcium: 923 (505) mg/day BMD: 0.94 g/cm 2 (0.12) T-score: -1.0 Vertebral Fractures: NR Interventions Alendronate 2.5 or 5mg vs placebo (< 500mg calcium intake encouraged to increase) Vertebral and Non-Vertebral Fractures: Women were questioned about any symptoms at clinic visits every 3 months. All unfavourable clinical effects including fractures were evaluated with respect to severity, duration, seriousness, relation to study drug and outcome
Outcomes
Notes Risk of bias Bias Allocation concealment (selection bias)
Low risk
Authors judgement
SE
Liberman 1995 Methods
Participants
IN
TE
AL
Randomized controlled trial Secondary prevention Blinding: double blind, blinded radiologists Withdrawals: Alendronate: 97/597 (16.2%) Placebo: 65/397 (16.5%) Total: 162/994 (16.3 %) Duration: 3 years Source: Two multicentre studies, one in the US, and the other in Australia, Canada, Europe, Israel, Mexico, New Zealand, South America Inclusion Criteria: Postmenopausal (= 5 yrs) women age 45-80 with lumbar BMD at least 2.5 SD below premenopausal mean Exclusion Criteria: Other causes for osteoporosis (glucocorticoids, vitamin D deciency, Pagets, hyperparathyroidism), active peptic ulcer disease, abnormal renal or hepatic function, abnormalities of spine precluding assessment of BMD for 3 lumbar vertebrae, history or hip fracture, prior bisphosphonates, HRT, calcitonin, uoride, or anabolic steroid in previous 12 months. Treatment N = 597 Control N = 397 Age:64 (7); YSM: 16.5 Calcium: 739 (537) mg/day
51
O N
LY
A - Adequate
Liberman 1995
(Continued)
BMD: 0.71 T-score: -3.1 Vertebral Fractures: 21% Interventions Alendronate 5, 10 or 20/5 mg vs placebo (500 mg calcium/day) Vertebral fractures, vertebral deformities and height loss. (Deformities and height loss not included in this review.) Vertebral Fractures: Lateral thoracic and lumbar spine lms were obtained at baseline, one, two, and three years. Standard values for target-to-lm distance and centering were used at each centre. Vertebral heights were determined at a radiology centre by observers blinded to both treatment and sequence. All lms from each woman were digitized at the same time and anterior, middle and posterior vertebral heights were calculated with using computer software. Prevalent fractures were determined by comparing baseline vertebral height ratios to a reference group. A ratio of > -3 SD was considered a fracture. Incident fractures were dened as a reduction of at least 20% and 4 mm between baseline and follow up Non-Vertebral Fractures, Hip and Wrist: All reported symptomatic fractures were recorded with no attempt to exclude fractures on the basis of degree of trauma
Outcomes
AL
Authors judgement Unclear risk Randomized controlled trial Secondary prevention Blinding: double blind matching placebo Duration: 1 year Withdrawals: Alendronate: 118/950 (12.4%) Control: 93/958 (9.7%) Total: 211/1908 (11.1 %)
U
Pols 1999 Methods
Participants
IN
TE
Source: 153 centres in 34 countries in Europe, Latin America, Australia, Canada, South Africa and China. . Inclusion Criteria: Postmenopausal for at least 3 years, not older than age 85, lumbar BMD (L2-4) at least 2 SD below the premenopausal mean (=0.86 g/cm (DXA) or =0.98 g/cm (Hologic)). At lease 3 vertebrae from L1-L4 had to be evaluable by DXA to determine BMD. In good health and between 20% below and 50% above ideal body weight Exclusion Criteria: Other metabolic bone disease, disturbed parathyroid or thyroid, major GI disease (peptic ulcer or malabsorption, drug to inhibit gastric > 2 week within past 3
52
SE
O N
LY
Pols 1999
(Continued)
Interventions
Alendronate 10 mg, vs placebo (500mg calcium/day)
Outcomes
AL
U
Characteristics of excluded studies [ordered by study ID]
Study Adami 1993 Adami 1995 Aki 2003
IN
t-score calculated using the lumbar spine BMD [(LS BMD -1.047)/0.110]; YSM=Years Since Menopause; BMD=Bone Mineral Density; Txt=Treatment; HRT=Hormone Replacement Therapy
TE
Reason for exclusion Duplicate or earlier report of another study. Lack of appropriate fracture data (i.e. reported as adverse events or unspecied) Lack of fracture outcome.
53
SE
Non- Vertebral, Hip, Wrist: Clinical fractures were assessed through adverse event reporting. Supporting documentation for each fracture i.e. radiographs and/or radiology reports, hospital discharge reports with clinical diagnosis or conrmation by investigator/treating physician
O N
LY
mos), MI within 1 year, uncontrolled hypertension or angina, impaired renal function, bisphosphonates or uoride within 6 months, estrogen ipriavone or calcitonin within 4 months, anabolic steroids, glucocorticoid or progestin within 6 months, medication inuencing bone metabolism - vitamin A > 10,000 IU/day, vitamin D > 1000 IU/day, anticonvulsants, phosphate-binding antacids, Treatment N = 950 Control N = 958 Age: 62.8 (7.4); YSM: 15.9 (1.5) Calcium: Not available BMD: 0.72 g/cm2 (0.08) T-score: -2.97 Vertebral Fractures: not reported
(Continued)
Bell 2002 Bettembuk 1999 Black 2000 Black 2003 Body 2002 Boivin 2000 Bone 2000 Bone 2004 Bonnick 1998 Bouxsein 1999 Chailurkit 2003 Chailurkit 2004 Chavassieux 1997 Cheng 2002 Chesnut 1993 Cummings 2000 Davas 2003
Lack of appropriate fracture data (i.e. reported as adverse events or unspecied) Duplicate or earlier report of another study. Duplicate or earlier report of another study. Lack of an appropriate control group. Lack of an appropriate control group. Lack of fracture outcome.
Extension/discontinuation study.
Lack of fracture outcome. Lack of fracture outcome.
Lack of fracture outcome.
Duration of therapy < 1 year. Duration of therapy < 1 year.
Devogelaer 1996 Dobnig 2006 Downs 2000 Evio 2004 Gonnelli 2002
Greenspan 2002a Greenspan 2002b
IN
TE
Lack of fracture outcome. Lack of an appropriate control group. Duplicate or earlier report of another study. Lack of fracture outcome. Lack of appropriate fracture data (i.e. reported as adverse events or unspecied) Lack of an appropriate control group. Lack of fracture outcome. Extension/discontinuation study. Duration of therapy < 1 year.
54
AL
Lack of an appropriate control group.
SE
Lack of appropriate fracture data (i.e. reported as adverse events or unspecied)
O N
Lack of appropriate fracture data (i.e. reported as adverse events or unspecied)
LY
(Continued)
Greenspan 2003 Harris 1993 Heijckmann 2002 Ho 2005 Hochberg 1999 Hosking 2003 Iwamoto 2004 Johnell 2002 Kung 2000 Kushida 2004 Lau 2000 Luckey 2004 Malavolta 1999 McClung 1998 McClung 2004 Murphy 2001 Nenonen 2005 Palomba 2002 Payer 2000 Ravn 1999a Ravn 1999b Ravn 1999c Ravn 2000 Rhee 2006
Lack of appropriate fracture data (i.e. reported as adverse events or unspecied) Duration of therapy < 1 year. Non randomized. Lack of fracture outcome. Duplicate or earlier report of another study. Lack of appropriate fracture data (i.e. reported as adverse events or unspecied) Lack of an appropriate control group. Lack of fracture outcome. Lack of fracture outcome. Lack of an appropriate control group. Lack of fracture outcome.
Duration of therapy < 1 year. Lack of fracture outcome.
IN
TE
Extension study.
Lack of appropriate fracture data (i.e. reported as adverse events or unspecied) Lack of fracture outcome. Lack of an appropriate control group. Duration of therapy < 1 year. Lack of fracture outcome. Lack of fracture outcome. Extension/discontinuation study. Extension/discontinuation study. Lack of fracture outcome.
55
AL
SE
O N
LY
(Continued)
Rittmaster 2000 Rizzoli 2002 Rossini 1994 Rossini 2000 Rozkydal 2003 Sahota 2000 Sambrook 2004a Sambrook 2004b Sawka 2003 Schneider 1999 Schnitzer 2000 Seeman 1999 Simon 2002 Sosa 2002 Stepan 1999 Tiras 2000 Tucci 1996
Lack of an appropriate control group. Lack of an appropriate control group. Duration of therapy < 1 year. Lack of fracture outcome. Lack of an appropriate control group. Lack of an appropriate control group. Lack of an appropriate control group. Extension/discontinuation study. Non-randomized. Lack of fracture outcome. Lack of an appropriate control group.
Lack of an appropriate control group. Lack of an appropriate control group. Lack of fracture outcome. Lack of an appropriate control group.
Tutuncu 2005
Uusi-Rasi 2003 van der Poest 2000 Vasikaran 1995 Yen 2000 Yildirim 2005
IN
TE
Duplicate or earlier report of another study. Lack of fracture outcome. Lack of fracture outcome. Lack of fracture outcome. Lack of an appropriate control group. Lack of fracture outcome. Lack of fracture outcome.
AL
Duplicate or earlier report of another study.
SE
O N
LY
56
DATA AND ANALYSES
Comparison 1. Alendronate 10 mg vs Control - all years baseline denominators
Outcome or subgroup title 1 Vertebral Fractures 1.1 Vertebral primary 1.2 Vertebral secondary
No. of studies 6 2 4
No. of participants 7361 4576 2785
Statistical method Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI)
Effect size 0.55 [0.45, 0.67] 0.55 [0.38, 0.80] 0.55 [0.43, 0.69]
IN
1 Hip Fractures 1.1 Hip primary 1.2 Hip secondary
TE
Outcome or subgroup title
AL
1 Non Vertebral Fractures 1.1 Non-vertebral primary 1.2 Non vertebral secondary
6 2 4
9625 4576 5049
SE
No. of studies
No. of participants
O N
Statistical method Effect size 0.84 [0.74, 0.94] 0.89 [0.76, 1.04] 0.77 [0.64, 0.92] Statistical method Effect size 0.61 [0.40, 0.92] 0.79 [0.44, 1.44] 0.47 [0.26, 0.85] Statistical method Effect size 0.84 [0.66, 1.06] 1.19 [0.87, 1.62] 0.52 [0.36, 0.75]
Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI)
Outcome or subgroup title 1 Wrist Fractures 1.1 Wrist primary 1.2 Wrist secondary
LY
57
Comparison 5. Alendronate 10 mg vs Control - 1 year baseline denominators
Outcome or subgroup title 1 Fractures 1.1 Vertebral 1.2 Nonvertebral fractures 1.3 Hip fractures 1.4 Wrist fractures
No. of studies 4 3 2 2 2
No. of participants
Statistical method Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI)
Effect size Subtotals only 0.84 [0.43, 1.63] 0.52 [0.30, 0.89] 0.67 [0.11, 4.01] 0.40 [0.16, 1.04]
306 2052 2052 2052
Comparison 6. Alendronate 10 mg vs Control - 1 year baseline denominators
Outcome or subgroup title 1 Fractures 1.1 Vertebral primary 1.2 Vertebral secondary 1.3 Nonvertebral primary 1.4 Non vertebral secondary 1.5 Hip primary 1.6 Hip secondary 1.7 Wrist primary 1.8 Wrist secondary
No. of studies 4 1 2 1 1 1 1 1 1
No. of participants
SE
O N
Statistical method Effect size Subtotals only 0.0 [0.0, 0.0] 0.84 [0.43, 1.63] 0.0 [0.0, 0.0] 0.52 [0.30, 0.89] 0.0 [0.0, 0.0] 0.67 [0.11, 4.01] 0.0 [0.0, 0.0] 0.40 [0.16, 1.04] Statistical method Effect size Subtotals only 0.0 [0.0, 0.0] 0.50 [0.09, 2.71]
Comparison 7. Alendronate 10 mg vs Control - 2 years baseline denominators
IN
TE
144 162 144 1908 144 1908 144 1908
Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI)
Outcome or subgroup title 1 Fractures 1.1 Vertebral secondary 1.2 Hip secondary
AL
No. of participants 61 327
LY
58
Outcome or subgroup title 1 Fractures 1.1 Vertebral secondary 1.2 Nonvertebral secondary 1.3 Hip secondary 1.4 Wrist secondary
No. of participants
2623 3141 3141 3245
Outcome or subgroup title 1 Fractures 1.1 Vertebral primary 1.2 Nonvertebral primary 1.3 Hip primary 1.4 Wrist primary
No. of participants
O N
LY
Effect size Subtotals only 0.55 [0.38, 0.80] 0.89 [0.76, 1.04] 0.79 [0.44, 1.44] 1.19 [0.87, 1.62] Effect size Subtotals only 0.40 [0.29, 0.55] 1.58 [0.82, 3.05] 0.55 [0.26, 1.18] Effect size 0.95 [0.83, 1.09] 1.10 [0.94, 1.29]
59
R
No. of studies 4 3 1 1
N
AL
No. of participants 2807 999 184 No. of participants 8796 3273
4432 4432 4432 4432
IN
1 Fractures 1.1 Vertebral secondary 1.2 Non vertebral primary 1.3 Non vertebral secondary
TE
Outcome or subgroup title 1 Withdrawals due to side effects 2 Withdrawals overall
No. of studies 6 5
SE
Statistical method Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI)
Statistical method Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI)
ADDITIONAL TABLES
Table 1. Clinical Relevance Table for Fracture - Primary Prevention Trials
Outcome
# Patients # Control Trials Event Rate
Wt Absolute Wt Rel RD Change
% NNT B
Statistical Sig Quality Evidence Statistically signicant Gold
of
95% condence interval Verte4,576 (2) bral Fractures (Low Risk Woman) - Primary Prevention (alendronate 10 mg/day for 1-4 yrs) 95% condence interval Verte4,576 (2) bral Fractures (Moderate Risk Woman) - Primary Prevention (alendronate 10 mg/day for 1-4 yrs) 95% condence interval Non Vertebral 4,576 (2) Fractures (Trial Populations) - Primary Prevention (alen-
(-2, -1)
(-62, -20)
1.2 % (1 out of Not applicable -45% (I) 100)
SE
O N
(48, 148) 186 Statistically signicant Gold (135, 417) 42 Statistically signicant Gold (31, 95) Not applicable Not statis- Gold tically signicant
Verte4,576 (2) bral Fractures (Trial Populations) - Primary Prevention (alendronate 10 mg/day for 1-4 yrs)
3.4% (3 out of -2% 2 fewer -45% (I) 100) patients out of 100
66
AL R
(-62, -20) 5.3 % (5 out of Not applicable -45% (I) 100) (-62, -20) 13.0% -1% 1 fewer -11% (I) (13 out of 100 patient out of patients) 100
IN
TE
LY
60
(Continued)
dronate 10mg/day for 1-4 years) 95% condence interval Non Vertebral 4,576 (2) Fractures (Low Risk Woman) - Primary Prevention (alendronate 10mg/day for 1-4 years) 95% condence interval Non Vertebral 4,576 (2) Fractures (Moderate Risk Woman) - Primary Prevention (alendronate 10mg/day for 1-4 years) 95% condence interval (-3, 0) (-24, 4)
8.6% (9 out of Not applicable -11% (I) 100)
Not applicable Not statis- Gold tically signicant
AL
(-24, 4) 1.1% (1 out of 0% fewer pa- -21% (I) 100) tients out of 100 Not applicable Not statis- Gold tically signicant (-1, 0) (-56, 44) 0.4% (0 out of Not applicable -21% (I) 100) Not applicable Not statis- Gold tically signicant
Hip Fractures 4,576 (2) (Trial Populations) - Primary Prevention (alendronate 10mg/day for 1-4 yrs 95% condence interval Hip Fractures 4,576 (2) (Low Risk Woman) - Primary Prevention (alen-
IN
TE
16.5% (17 out Not applicable -11% (I) of 100)
SE
(-24, 4)
O N
LY
61
(Continued)
dronate 10mg/day for 1-4 yrs 95% condence interval Hip Fractures 4,576 (2) (Moderate Risk Woman) - Primary Prevention (alendronate 10mg/day for 1-4 yrs 95% condence interval Wrist Frac- 4,576 (2) tures (Trial Populations) - Primary Prevention (alendronate 10 mg/day for 1-4 yrs) 95% condence interval Wrist Frac- 4,576 (2) tures (Low Risk Woman) - Primary Prevention (alendronate 10 mg/day for 1-4 yrs) 95% condence interval Wrist Frac- 4,576 (2) tures (Moderate Risk Woman) - Primary Prevention (alendronate Not available (-56, 44)
(-56, 44)
AL R
(0, 2) (-13, 38) Not available Not applicable 19% (W) Not applicable Not statis- Gold tically signicant (-13, 38) Not applicable 19% (W) Not applicable Not statis- Gold tically signicant
IN
TE
3.1% (3 out of 1 1 more pa- 19% (W) 100) tient out of 100
SE
O N
LY
62
(Continued)
10 mg/day for 1-4 yrs) 95% condence interval Legend Primary prevention = bone density < 2 SD values below peak bone mass and/or no history of vertebral compression fractures For Trial Population rates are based on the event rate in the control group. Low and Moderate Risk, are 5 year community population risks derived from the following variables from the FRACTURE Index: age, fracture after 50 yrs., maternal hip fracture after 50 yrs., weight < 125 lbs, smoking, using arms to assist standing and BMD. Low = FRACTURE Index score 12, Moderate = FRACTURE Index score 5 (Black 2001) see Figure 1 Wt = weighted, RD = risk difference (-13, 38)
IN
TE
AL
SE
O N
LY
Wt Rel = weighted relative percent change, I = improvement
NNT B = number needed to benet
Gold level: At least one randomised clinical trial meets all of the following criteria for the major outcome (s) as reported: Sample sizes of at least 50 per group. If a statistically signicant difference is not found they must be powered for 20% relative difference in the relevant outcome. Blinding of patients and assessors for outcomes. Handling of withdrawals > 80% follow up (imputations based on methods such as Last Observation Carried Forward (LOCF) acceptable). Concealment of treatment
63
(Continued)
allocation. Silver level: Randomised trial does not meet the above criteria Table 2. Clinical Relevance Table for Fracture - Secondary Prevention Trials
Outcome
% NNT B
LY
Statistical Sig Quality Evidence Statistically signicant Gold
of
95% condence interval
IN
Verte2,785 (4) bral Fractures (Moderate Risk Woman) - Secondary Prevention (alendronate 10 mg for 1-3 yrs)
TE
5.3% (5 out of NA 100)
AL
(-8, -4)
Verte2,785 (4) bral Fractures (Trial populations) - Secondary Prevention (alendronate 10 mg for 1-3 yrs)
12.2% (12 out -6% 6 fewer -45% (I) of 100) patients out of 100
SE
(-57, -31) (15, 25) -45% (I) 42 Statistically signicant
O N
19 (34, 61) 20
Gold
(-57, -31)
Verte2,785 (4) bral Fractures (High Risk Woman) - Secondary Prevention (alendronate 10 mg for 1-3 yrs)
11.2% (11 out NA of 100)
-45% (I)
Statistically signicant
Gold
64
Table 2. Clinical Relevance Table for Fracture - Secondary Prevention Trials
(Continued)
95% condence interval Non Vertebral 5049 (4) Fractures (Trial Population)Secondary Prevention (alendronate 10mg/day for 1 - 3 yrs) 95% condence interval Non Vertebral 5049 (4) Fractures (Moderate Risk Woman) - Secondary Prevention (alendronate 10mg/day for 1 - 3 yrs) 95% condence interval Non Vertebral 5049 (4) Fractures (High Risk Woman) - Secondary Prevention (alendronate 10mg/day for 1 - 3 yrs) 95% condence interval Hip Fractures 5,376 (5) (Trial Population) - Secondary Prevention (alendronate 10mg/day for
(-57, -31)
(16, 29)
9.3% (9 out of -2% 2 fewer -23% (I) 100) patients out of 100
47
Gold
(-4, -1)
(-36, -8)
AL
SE
16.5% (17 out NA of 100)
-23% (I)
O N
(30, 135) 27 (17, 76) 16 (11, 46)
LY
Statistically signicant Gold
IN
TE
27.5% (28 out NA of 100)
(-36, -8)
-23%
Gold
(-36, -8) (I)
1.3% ( out of -1% 1 fewer -53% ( -74, - 146 100) patients out of 15) (I) 100
Gold
65
(Continued)
1-3 yrs) 95% condence interval Hip Fractures 5,376 (5) (Moderate Risk Woman) - Secondary Prevention (alendronate 10mg/day for 1-3 yrs) 95% condence interval Hip Fractures 5,376 (5) (High Risk Woman) - Secondary Prevention (alendronate 10mg/day for 1-3 yrs) 95% condence interval Wrist 5,153 (4) Fractures (Trial Population) - Secondary Prevention (alendronate 10mg/day for 1-3 yrs) 95% condence interval Wrist Frac- 5,153 (4) tures - Secondary Prevention (alendronate 10mg/day for 1-3 yrs) NA 8.9% (9 out of NA 100) (-1, 0) (104, 513)
1.9% (2 out of NA 100)
-53% ( -74, - 100 15) (I)
Gold
AL
SE
-53% ( -74, - 22 15) (I)
O N
(72, 351) Statistically signicant Gold
IN
TE
2.9% (3 out of -1% 1 fewer -50% (-66, - 69 100) patients out of 27) (I) 100
(16, 75)
LY
Statistically signicant Gold (-2, -1) (53, 128) NA -50% (-66, - NA 27) (I) Statistically signicant Gold
66
(Continued)
95% condence interval Wrist Frac- 5,153 (4) tures - Secondary Prevention (alendronate 10mg/day for 1-3 yrs) Legend Secondary prevention = bone density of at least 2 SD values below peak bone mass and/ or one or more vertebral compression fractures NA NA -50% (-66, - NA 27) (I) Statistically signicant Gold
For Trial Population rates are based on the event rate in the control group. Moderate and High Risk, are 5 year community population risks derived from the following variables in the FRACTURE Index: age, fracture after 50 yrs., maternal hip fracture after 50 yrs., weight < 125 lbs, smoking, using arms to assist standing and BMD. Moderate = FRACTURE Index score 5, High = FRACTURE Index score 8-13 (Black 2001) see Figure 1
IN
TE
AL
SE
O N
Wt = weighted, RD = risk difference
LY
Gold level: At least one randomised clinical trial meets all of the following criteria for the major outcome(s) as reported: Sample sizes of at least 50 per group. If a statistically signicant difference is not found they must be powered for 20% relative difference in the relevant outcome. Blinding of patients and assessors for outcomes. Handling of withdrawals > 80% follow up (imputations based on methods such
67
(Continued)
as Last Observation Carried Forward (LOCF) acceptable). Concealment of treatment allocation. Silver level: Randomised trial does not meet the above criteria.
FEEDBACK Feedback from Maryann Napoli, 2 June 2008

Summary
Reply Thank you for your feedback about the Plain language summary. The CSMG has been and continues to be actively involved in research to summarise our reviews in a format that is useful and comprehensible to consumers. Your feedback addresses the issues around providing absolute event rates and absolute differences to consumers, in particular for dichotomous outcomes. In summaries that we have published in earlier reviews, review authors presented primarily absolute event rates without differences. In this review, differences were provided, but event rates were not. Unfortunately we do not have high quality evidence to determine the best presentation, but we continue to investigate and explore the options. User testing of a variety of Plain language summary formats, which is funded by the Cochrane Collaboration Opportunity Fund, is pending and we look forward to the results. As you have suggested, we have edited this Plain language summary to present absolute event rates.
IN
Feedback: I like the new format of the plain language summary, particularly the way you have expressed best estimates. But there is a glaring omission: You should also provide an estimate of the number of hip fractures, vertebral fractures, etc out of 100 women who do NOT take alendronate. If this and future PLS do not include numerical information about the no-treatment options, then it will not be clear to readers what their chances of having a hip fracture, vertebral fracture, etc. are to begin with. People will want to know one fewer than what? Submitter agrees with default conict of interest statement: I certify that I have no afliations with or involvement in any organization or entity with a nancial interest in the subject matter of my feedback.
TE
Date of Submission: 02-Jun-2008 Name: Maryann Napoli Email Address: mnapoli2@ix.netcom.com Personal Description: Occupation a consumer advocate
AL
SE
O N
LY
68
Contributors George A. Wells, Nancy Santesso, Tamara Rader.
Feedback from Mark Porcheret, 22 December 2010
Summary Date of Submission: 22-Dec-2010 Name: Mark Porcheret Email Address: m.porcheret@keele.ac.uk Personal Description: Occupation GP Research fellow Feedback: In gure 6 on page 20 the corresponding risk for the outcome hip fractures for the moderate risk population is stated to be 19 per 1000 (CI 5 to 16). Could you conmr this is a typo and should be 9 per 1000. Also, I am trying to interpret this data for an EBP group I facilitate and have a question: the risks in gure 6 are 5yr risks comparing people on alendronate to those not taking it. But the heading states it is for alendronate for 1-3yrs. So, I am not sure how to phrase the evidence for use in the consultation. Does that data mean, for example for vertebral factures in the moderate risk populations, that: If you take a 1000 people like you, those at moderate risk, in 5yrs time 63 will have had a vertebral fracture, but if all 1000 took alendronate for the 5 years only 29 will have had a vertebral fracture. This is how such a comparison would normally be comminicated with the period of risk / benet being the same as the period of treatment, but the gure presents 5yr risks for 1-3yrs of treatment, which complicates the message. Your views on this would be welcome before the meeting we have to discuss this on 12th January 2011. Many thanks Mark Porcheret Submitter agrees with default conict of interest statement: I certify that I have no afliations with or involvement in any organization or entity with a nancial interest in the subject matter of my feedback. Reply
Contributors George A Wells, Lara Maxwell.
IN
Many thanks for your feedback on this review. Please nd below our response to your comment after consultation with the lead author, George Wells. Regarding your rst point, yes, we can conrm that there is a typo here and it should state 9 per 1000. We will x this for the next issue of the Library. For your second point, yes, the risks in the table are for 5 year risks and your interpretation is correct (though the 63 in your email should be 53 as from the table). The 1-3 years in the heading simply indicates that the relative risk came from studies of 1 to 3 years duration (since this is the best evidence that we have) but were modeled on a 5-year time horizon. We agree this heading is confusing so we will take the 1-3 out of the heading and explain this in a footnote in the table for the next issue. We also noticed that the highrisk population should state 62/1000 instead of 62/100. Thanks again for bringing this to our attention.
TE
AL
SE
O N
LY
69
Feedback from Aaron M Tejani, 17 December 2010
Summary Date of Submission: 17-Dec-2010 Name: Aaron M Tejani Email Address: aaron.tejani@fraserhealth.ca Personal Description: Occupation a pharmacist Feedback: In the review, alendronate was found to have a statistically signicant benet for primary vertebral fracture prevention (45% RRR, 95% CI of 0.38 to 0.80). The results were compiled from two studies that were conducted by Ascott-Evans et al. (2) (2003) and Cummings et al. (3) (1998) (as shown by Analysis 1.1 gure). Based on closer inspection of these two studies we have some specic concerns. The review (1) cited that vertebral fractures did not occur in neither the treatment nor control group in the study by Ascott-Evans et al. (2) as shown in the Analysis 1.1 gure. However, the study did not report the incidence of vertebral fractures. It is incorrect to assume that no vertebral fractures occurred simply because they were not reported in the trial publication. The authors of the review need to clarify if unpublished vertebral fracture information was received from Ascott-Evans et al. The other study that contributed to the vertebral fracture meta-analysis for alendronate was Cummings et al. 2003 (3). In this trial vertebral fractures were solely radiographically determined and not clinical fractures. The review authors should emphasize that the data used in the meta-analysis of alendronate in primary prevention from Cummings et al. 2003 was only for non-clinical vertebral fractures. If this review is revised, we urge the authors to also clarify whether alendronate reduces the risk of clinical vertebral fractures in secondary prevention. It would be very useful if these authors also clarify the non-clinical versus clinical fracture issue for all the oral bisphosphonate reviews.(4,5) 1. Wells GA, Crannery A, Peterson J, Boucher M, Shea B, Welch V, Coyle D, Tugwell P. Alendronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women. Cochrane Database of Systematic Reviews 2008, Issue 1. Art. No.: CD001155. DOI: 10.1002/14651858.CD001155.pub2. 2. Ascott-Evans BH, Guanabens N, Kivinen S, Stuckey BG, Magaril CH, Vandormael K, et al. Alendronate prevents loss of bone density associated with discontinuation of homrone replacement therapy: a randomized controlled trial. Archives of Internal Medicine 2003; 163(7): 789-94.
4. Wells GA, Cranney A, Peterson J, Boucher M, Shea B, Welch V, Coyle D, Tugwell P. Etidronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women. Cochrane Database of Systematic Reviews 2008, Issue 1. Art. No.: CD003376. DOI: 10.1002/14651858.CD003376.pub3. 5. Wells GA, Cranney A, Peterson J, Boucher M, Shea B, Welch V, Coyle D, Tugwell P. Risedronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women. Cochrane Database of Systematic Reviews 2008, Issue 1. Art. No.: CD004523. DOI: 10.1002/14651858.CD004523.pub3. Submitter agrees with default conict of interest statement: I certify that I have no afliations with or involvement in any organization or entity with a nancial interest in the subject matter of my feedback.
IN
3. Cummings SR, Black DM, Thompson DE, Appleggate WB, Barrett-Connor E, Musliner TA, et al. Effect of alendronate on risk of fracture in women with low bone density but without vertebral fractures: results from the Fracture Intervention Trial. JAMA 1998; 280(24):2077-82.
TE
AL
SE
O N
LY
70
Reply Dear Dr Tejani, Thank you for your feedback regarding our review. With respect to the Ascott-Evans study, we determined the absence of vertebral fractures from the information provided in the publication. (We did not request additional data from any study authors as the selective response often elicited to such a request may introduce a source of bias.) Fractures, in this paper, were assessed as adverse events. Adverse events were recorded by study personnel at each visit using non-leading questions. In reporting these events, in the results section of the paper, the authors stated that no fractures were reported during the study. We agree that it may have been presumptuous of us to assume the absence of vertebral fractures from this report and we would not be adverse to removing this study from the analysis. Doing so, however, would not substantially alter our results or conclusions. Your concern regarding the Cummings paper is a point well taken. This was the only primary prevention study which reported both radiographic and clinical fractures as separate outcomes. Since the other papers (with the exception of Ascott-Evans having 0 events) reported only radiographic vertebral fractures we decided, for the sake of consistency, to use the radiographic outcome for Cummings in the meta-analysis. (As well, all of the alendronate studies in the secondary prevention analysis reported only radiographic fractures.) Your suggestion that we clarify the issue surrounding clinical and non clinical fractures for future bisphosphonate updates is a good one and will be included in the next update of the review. We hope that these responses are helpful. Please do not hesitate to contact us should you have any further questions or concerns.
Geroge A. Wells, Elizabeth Ghogomu.
Summary Name: George hannah Email Address: george.hannah@nhs.net Personal Description: Occupation GP
Reply Dear Dr Hannah, The NNTs reported in our review are derived from the pooling of trials varying in length between 1 and 4 years. (The majority of subjects stemmed from trials which had follow ups of 3 or 4 years). These studies were designed to evaluate the current use of alendronate and did not provide for an off drug follow up period. The estimates reported in our review cannot be extrapolated to a future date at which the patient is no longer taking the medication. You had also asked if 5 years was a standard treatment length. We did not make any recommendations in the review regarding length of treatment. We used 5 year risk estimates in our Summary of Findings tables and in Figure 9 as the FRACTURE Index (Black 2001) used to classify women according to baseline risk factors was based on 5 year community population risks. With respect to the effect of length of treatment on the NNTs, we did provide a breakdown of treatment effects by year of follow up in Figure 11,which could, theoretically, be translated into NNTs. These estimates, however, should be viewed with caution as in some cases they are comprised of interim time point data.
IN
Feedback: I am uncertain what length of treatment of alendronate is being referred to in the review, to produce the NNT. When you may be treating women in their 60s to prevent fractures in their 80s? Does treatment have to be ongoing to produce the benet? Is 5 years a standard length. How does this affect NNT? ie is the NNT in year one, less than year 2, year 3 etc of treatment. Submitter has modied conict of interest statement: I certify that I have no afliations with or involvement in any organization or entity with a nancial interest in the subject matter of my feedback.
TE
AL
Feedback from George Hannah, 5 November 2010
SE
Contributors
O N
LY
71
Contributors George A. Wells, Elizabeth Ghogomu.
WHATS NEW
Last assessed as up-to-date: 13 November 2007.
Date 8 July 2011
Event Feedback has been incorporated
Description Responses to queries about: 1. clincal and non-clinical fractures 2. NNT calculation.
HISTORY
Protocol rst published: Issue 3, 2005 Review rst published: Issue 1, 2008
Date 13 January 2011 12 August 2008 28 May 2008
Event Feedback has been incorporated Feedback has been incorporated Amended
CONTRIBUTIONS OF AUTHORS
Ann Cranney was involved with the conception of the review, data abstraction, analysis, interpretation and revision of the nal report. Joan Peterson screened the literature, was involved in the data abstraction, quality assessment and analysis of the primary trials and contributed signicantly to the writing of the report. Michel Boucher assisted in the design of the analysis, reporting and interpretation of the ndings and was involved in the writing of the report. Beverley Shea was involved in developing the protocol and conducting the systematic review. Vivian Robinson was involved in developing the protocol and conducting the systematic review. Douglas Coyle assisted with the design of the review and reviewed the analysis. Peter Tugwell provided clinical rheumatology expertise and methodological guidance.
IN
George Wells was involved in the conception, design and implementation of the project and contributed signicantly to the writing of the report.
TE
AL
U
Description Amendments in Figure 6: Summary of Findings for Secondary Prevention Absolute event rates included in the Plain language summary. Converted to new review format.CMSG ID C004-R
SE
O N
LY
72
DECLARATIONS OF INTEREST
None.
SOURCES OF SUPPORT Internal sources

Ottawa Hospital Research Institute, Canada.
External sources
Canadian Agency for Drugs and Technologies in Health, Canada.
INDEX TERMS Medical Subject Headings (MeSH)
Female; Humans
IN
TE
AL
MeSH check words
SE
Alendronate [ therapeutic use]; Bone Density Conservation Agents [ therapeutic use]; Fractures, Bone [ prevention & control]; Fractures, Spontaneous [prevention & control]; Hip Fractures [prevention & control]; Osteoporosis, Postmenopausal [ drug therapy]; Randomized Controlled Trials as Topic; Spinal Fractures [prevention & control]
O N
LY
73
Risedronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women (Review)
Wells GA, Cranney A, Peterson J, Boucher M, Shea B, Welch V, Coyle D, Tugwell P
Risedronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women (Review) Copyright 2010 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
IN
TE
AL
SE
O N
LY
TABLE OF CONTENTS HEADER . . . . . . . . . . ABSTRACT . . . . . . . . . PLAIN LANGUAGE SUMMARY . BACKGROUND . . . . . . . OBJECTIVES . . . . . . . . METHODS . . . . . . . . . Figure 1. . . . . . . . . Figure 2. . . . . . . . . Figure 3. . . . . . . . . Figure 4. . . . . . . . . Figure 5. . . . . . . . . Figure 6. . . . . . . . . RESULTS . . . . . . . . . . Figure 7. . . . . . . . . Figure 8. . . . . . . . . Figure 9. . . . . . . . . Figure 10. . . . . . . . . Figure 11. . . . . . . . . Figure 12. . . . . . . . . Figure 13. . . . . . . . . Figure 14. . . . . . . . . Figure 15. . . . . . . . . Figure 16. . . . . . . . . Figure 17. . . . . . . . . Figure 18. . . . . . . . . Figure 19. . . . . . . . . Figure 20. . . . . . . . . Figure 21. . . . . . . . . Figure 22. . . . . . . . . DISCUSSION . . . . . . . . Figure 23. . . . . . . . . AUTHORS CONCLUSIONS . . ACKNOWLEDGEMENTS . . . REFERENCES . . . . . . . . CHARACTERISTICS OF STUDIES DATA AND ANALYSES . . . . . ADDITIONAL TABLES . . . . . WHATS NEW . . . . . . . . HISTORY . . . . . . . . . . CONTRIBUTIONS OF AUTHORS DECLARATIONS OF INTEREST . SOURCES OF SUPPORT . . . . NOTES . . . . . . . . . . . INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 2 3 4 4 7 8 8 9 11 12 13 14 15 17 18 19 20 21 22 23 25 26 27 28 29 30 30 31 34 34 35 35 39 49 52 60 60 60 61 61 61 62
IN
TE
AL
SE
O N
LY
[Intervention Review]
Risedronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women
George A Wells1 , Ann Cranney2 , Joan Peterson3 , Michel Boucher4 , Beverley Shea5 , Vivian Welch6 , Doug Coyle7 , Peter Tugwell8
1 Cardiovascular Research Reference Centre, University of Ottawa Heart Institute, Ottawa, Canada. 2 Division of Rheumatology, Ottawa Hospital, Ottawa, Canada. 3 Clinical Epidemiology Unit, Ottawa Civic Hospital / Loeb Research Institute, Ottawa, Canada. 4 HTA Development Canadian Agency for Drugs and Technologies in Health (CADTH), Ottawa, Canada. 5 Institute of Population Health, University of Ottawa, Ottawa, Canada. 6 Centre for Global Health, Institute of Population Health, University of Ottawa, Ottawa, Canada. 7 Epidemiology and Community Medicine, Ottawa Health Research Institute, Ottawa, Canada. 8 Centre for Global Health, Institute of Population Health, Department of Medicine, Ottawa Hospital, Ottawa, Canada
Contact address: George A Wells, Cardiovascular Research Reference Centre, University of Ottawa Heart Institute, Room H1-1, 40 Ruskin Street, Ottawa, Ontario, K1Y 4W7, Canada. gawells@ottawaheart.ca. Editorial group: Cochrane Musculoskeletal Group. Publication status and date: Edited (no change to conclusions), published in Issue 7, 2010. Review content assessed as up-to-date: 13 November 2007.
Citation: Wells GA, Cranney A, Peterson J, Boucher M, Shea B, Welch V, Coyle D, Tugwell P. Risedronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women. Cochrane Database of Systematic Reviews 2008, Issue 1. Art. No.: CD004523. DOI: 10.1002/14651858.CD004523.pub3. Copyright 2010 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
Osteoporosis is an abnormal reduction in bone mass and bone deterioration leading to increased fracture risk. Risedronate belongs to the bisphosphonate class of drugs which act to inhibit bone resorption by interfering with the activity of osteoclasts. Objectives
To assess the efcacy of residronate in the primary and secondary prevention of osteoporotic fractures in postmenopausal women. Search strategy We searched CENTRAL, MEDLINE and EMBASE. Relevant randomized controlled trials published between 1966 to 2007 were identied. Selection criteria Women receiving at least one year of risedronate for postmenopausal osteoporosis were compared to those receiving placebo or concurrent calcium/vitamin D or both. The outcome was fracture incidence. Data collection and analysis We carried out study selection and data abstraction in duplicate. Study quality was assessed through the reporting of allocation concealment, blinding and withdrawals. Meta-analysis was preformed using relative risks and a >15% relative change was considered clinically important.
Risedronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women (Review) Copyright 2010 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd. 1
IN
TE
Background
AL
ABSTRACT
SE
O N
LY
Main results Seven trials were included in the review representing 14,049 women. Relative (RRR) and absolute (ARR) risk reductions for the 5 mg dose were as follows. Risk estimates for primary prevention were available only for vertebral and non vertebral fractures and showed no statistically signicant effect of risedronate on fractures. For secondary prevention, a signicant 39% RRR in vertebral fractures (RR 0.61, 95% CI 0.50 to 0.76) with 5% ARR was found. For nonvertebral fractures, a signicant 20% RRR (RR 0.80, 95% CI 0.72 to 0.90) with 2% ARR and for hip fractures there was a signicant 26% RRR (RR: 0.74, 95% CI 0.59 to 0.94) with a 1% ARR. When primary and secondary prevention studies were combined, the reduction in fractures remained statistically signicant for both vertebral (RR 0.63, 0.51 to 0.77) and non vertebral fractures (RR 0.80, 0.72 to 0.90) For adverse events, no statistically signicant differences were found in any of the included studies. However, observational data has led to concerns regarding the potential risk for upper gastrointestinal injury and, less commonly, osteonecrosis of the jaw. Authors conclusions
Risedronate for preventing fractures caused by osteoporosis in postmenopausal women This summary of a Cochrane review, presents what we know from research about the effect of Risedronate for preventing fractures (broken bones) caused by osteoporosis.
- probably prevents fractures in the bones of the spine and in bones other than in the spine; - may prevent hip fractures;
- may not lead to any difference in wrist fractures.
In women whose bone density is closer to normal or who may not yet have had a fracture in the bones of their spine, risedronate: may not lead to any difference in fractures in the bones of the spine, hip fractures or wrist fractures; there is not enough information to tell if Risedronate prevents fractures in bones other than in the spine. We do not have precise information about side effects and complications. This is particularly true for rare but serious side effects. Possible side effects may include digestive problems such as damage to the throat, esophagus and stomach and, less commonly, reduced blood supply to the jaw bone, which causes the bone tissue to breakdown .
What is osteoporosis and what is risedronate? Bone is a living, growing part of your body. Throughout your lifetime, new bone cells grow and old bone cells break down to make room for the new, stronger bone. When you have osteoporosis, the old bone breaks down faster than the new bone can replace it. As this happens, the bones lose minerals (such as calcium). This makes bones weaker and more likely to break even after a minor injury, like a little bump or fall. Women who have gone through menopause are more likely to get osteoporosis than other people.
IN
TE
In women who have already been diagnosed with low bone density putting them at risk for fracture or have already had a fracture in the bones of their spine, risedronate:
AL
SE
O N
At 5 mg/day a statistically signicant and clinically important benet in the secondary prevention of vertebral, non-vertebral and hip fractures was observed, but not for wrist. The level of evidence for secondary prevention is Gold (www.cochranemsk.org) for vertebral and non-vertebral and Silver for hip and wrist. There were no statistically signicant reductions in the primary prevention of vertebral and non-vertebral fractures. The level of evidence is Silver.
LY
Risedronate belongs to the class of drugs called bisphosphonates. It is a type of medication that slows down the cells that break down the old bone.
Best estimate of what happens to women that have already been diagnosed with low bone density putting them at risk for fracture or have already had a fracture in the bones of their spine, who take Risedronate:
Fracture of the spine - 14 out of 100 women had a fracture when taking a placebo - 9 out of 100 women had a fracture when taking Risedronate
Fracture in the hip - 3 out of 100 women had a fracture when taking a placebo - 2 out of 100 women had a fracture when taking Risedronate
Fracture in the wrist
- 3 out of 100 women had a fracture when taking Risedronate.
Fractures in bones other than the spine
Best estimate of what happens to women whose bone density is closer to normal or who may not yet have had a fracture in the bones of their spine who take risedronate: - there is no difference in the number of women out of 100 who will have a spine fracture. This may be the result of chance. - for hip and wrist fractures, it is not possible to calculate the effect because no one had fractures of the hip or wrist in the studies. - there is not enough information to tell if Risedronate prevents fractures in bones other than in the spine.
BACKGROUND
Osteoporosis is in part a natural consequence of aging in postmenopausal women (Hodsman 2002). It is a skeletal disorder characterized by decreased bone mass and deterioration of microar-
IN
TE
- 8 out of 100 women had a fracture when taking Risedronate
AL
SE
chitecture of bone resulting in an increased risk of fracture (NIH Consensus 2001). The most common consequences of osteoporosis are fractures of
3
O N
LY
the hip, wrist and vertebrae (Hodsman 2002). Bone strength reects the integration of two main features: bone density and bone quality (Brown 2002b). The clinical indicator of bone quality is a patients history of a fragility fracture. A fragility fracture is a fracture caused by an injury that would be insufcient to fracture normal bone (for example, a fall from a standing height or less)(Brown 2002b). The preferred method of evaluating bone density is the measurement of bone mineral density (BMD) of the lumbar spine and hip by Dual Energy X-ray Absorptiometry (DXA), which can be used to assess response to therapy (Hanley 2003). The interpretation of BMD results is based on comparison of a patients BMD with the mean value for a young adult population. The T-score is the number of standard deviations (SDs) above or below the mean BMD for normal young adults (Brown 2002b). The World Health Organization (WHO) Study Group on Osteoporosis dened osteoporosis as a hip BMD level of more than 2.5 SDs below the mean BMD for young, white, adult women (WHO 1994). Using the WHO denition, approximately 30% of postmenopausal women have osteoporosis (Kanis 1994; WHO 1994). It should be noted that there are limitations associated with the WHO denition. The predictive value of BMD measurement for fracture varies depending on the site selected, database used for comparison and the technology used. Furthermore, T-scores do not provide a good basis to establish comparable diagnostic thresholds between different regions of interest and different bone mass measurement techniques (Black 2001). As a result, betweensite and technique variability introduces potential for misclassication and inappropriate treatment of some individuals. Osteoporosis can be detected by BMD measurement or diagnosed by presence of osteoporosis-related fractures. The presence of preexisting osteoporotic fractures is an important risk factor for future fractures (Hodsman 2002). It is reported that 25% of women aged 80 have had at least one vertebral fracture (Melton 1989) and Cauley et al., (Cauley 2000) demonstrated excess mortality in women who have experienced a clinical vertebral fracture. The cumulative lifetime fracture risk for a 50-year-old woman with osteoporosis is stated to be as high as 60% (Cummings 1989). As a result, effective fracture prevention would have a major impact on morbidity and a smaller but important impact on mortality in these women.
conditions (Browner 1996; Cooper 1993). Prevention and treatment of osteoporosis can be complex, due to the multifactorial etiology of the disorder. New anabolic therapies directed at increasing bone formation, such as teriparatide (recombinant human parathyroid hormone (1-34)), (Shukla 2003) are available, however most currently available osteoporosis drugs are anti-resorptive agents that act to decrease bone turnover. One class of anti resorptive drugs includes the bisphosphonates: etidronate, alendronate and risedronate. They are recommended as rst-line preventive agents in postmenopausal women with low BMD and as rst-line agents for the treatment of postmenopausal women with osteoporosis (Brown 2002b). Bisphosphonates are stable analogues of naturally occurring pyrophosphates. The mechanism of action of these drugs is to inhibit bone resorption through their effects on osteoclasts (Brown 2002b). Bisphosphonates are poorly absorbed and avidly taken up by bone on active sites of resorption. Risedronate is a nitrogen containing pyridinyl third generation bisphosphonate which is administered daily or once weekly (depending on formulation). The recommended dose for the prevention and treatment of osteoporosis in postmenopausal women is 5 mg/day (35 mg / week) . Risedronate at 5 mg, relative to control, has been shown to increase bone mineral density after 1.5 to three years of treatment by 4.54% (95% CI 4.12 to 4.97) in the lumbar spine, and 2.75% (95% CI 2.32 to 3.17) in the femoral neck (Cranney 2002). At a dose of 2.5 mg increases at the lumbar spine and femoral neck were 2.94% (95% CI 1.55 to 4.34) and 1.71%, (95% CI 1.17 to 2.25) (Cranney 2002).
AL
TE
IN
Osteoporosis-related morbidity is associated with signicant medical and social consequences (Brown 2002b). The major source of morbidity and mortality from osteoporosis is attributed to hip fractures. Hip fractures are not only associated with an increase mortality risk but also inuence long-term function and independence. Fifty per cent of women who sustain a hip fracture do not return to their previous functional state and become dependent on others for their daily activities (Brown 2002b). The mortality associated with hip fractures in older women may be as high as 20% in the rst year (Cauley 2000). This excess mortality may not be directly attributable to the hip fracture, but to comorbid
SE
OBJECTIVES
The aim of this systematic review was to assess the clinical efcacy of risedronate in the primary and secondary prevention of osteoporotic fractures in postmenopausal women receiving these agents compared with untreated women over a follow-up period of at least one year.
METHODS
Criteria for considering studies for this review
Types of studies Randomized controlled trials (RCTs) with a duration of at least one year were included in this review.
4
O N
LY
Types of participants The population group of interest was post menopausal women. Both primary and secondary prevention trials were accepted. A hierarchy was used to dene primary versus secondary prevention according to the information available. We selected a denition of primary and secondary prevention that gave more weight to study inclusion criteria than baseline statistics. That is, if the inclusion criteria restricted the population to women whose bone density was at least 2 SD values below the peak bone mass or the inclusion criteria restricted the population to women that had experienced previous vertebral compression fractures, then the trial was considered a secondary prevention study. If such inclusion criteria were not provided then the baseline statistics were considered as follows: (a) we considered the trial as primary prevention if the average T-score (and SD) was such that it included women whose bone density was within 2 SD of the mean or if the prevalence of vertebral fracture at baseline was less than 20%; and (b) when these data were not available, we considered a trial as secondary prevention if the average age was above 62 years. Types of interventions
Search methods for identication of studies

The Cochrane Collaborative approach for identifying randomized controlled trials (RCTs) as described by Dickersin et al., (Dickersin 1994) and modied for the Cochrane Musculoskeletal Group, guided our literature search. We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE from 1966 to November 2004, Current Contents, and citations of relevant articles. No language restrictions were applied to the search strategy. The actual literature search was conducted in three stages. The rst stage was the basis for our systematic review published in 2002 (Cranney 2002) and the second and third stages involved updating the search. The rst search, for the time period 1966 to December 2000, included Cochrane Controlled Trials Register, MEDLINE, EMBASE, Current Contents, and handsearching of conference abstracts and FDA proceedings. This was followed by a MEDLINE search for the time period 2000 to November 2004. This MEDLINE search was conrmed by a parallel literature search that was conducted for a companion bisphosphonate
IN
TE
Incidence of fractures, including vertebral, non-vertebral, hip and wrist fractures.
Types of outcome measures
AL
Treatment: Risedronate at any dose. Comparators: No treatment (including placebo or calcium or vitamin D or both). If the study used calcium or vitamin D controls or both, these same treatments would have to be given concurrently in the risedronate treatment group.
economic report CADTH 2006. For the nal update (2004 to February 2007), we searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE and EMBASE. Search Strategy MEDLINE Using OVID Interface 1. osteoporosis, postmenopausal/ 2. osteoporosis/ 3. osteoporosis.tw. 4. exp bone density/ 5. bone loss$.tw. 6. (bone adj2 densit$).tw. 7. or/2-6 8. menopause/ 9. post-menopaus$.tw. 10. postmenopaus$.tw. 11. or/8-10 12. 7 and 11 13. 1 or 12 14. risedronate.tw. 15. 13 and 14 16. meta-analysis.pt,sh. 17. (meta-anal: or metaanal:).tw. 18. (quantitativ: review: or quantitativ: overview:).tw. 19. (methodologic: review: or methodologic: overview:).tw. 20. (systematic: review: or systematic: overview).tw. 21. review.pt. and medline.tw. 22. or/16-21 23. 15 and 22 24. clinical trial.pt. 25. randomized controlled trial.pt. 26. tu.fs. 27. dt.fs. 28. random$.tw. 29. (double adj blind$).tw. 30. placebo$.tw. 31. or/24-30 32. 15 and 31
SE
Data collection and analysis

Selection of studies Two review authors examined each title generated from the search and identied potentially eligible articles, for which we obtained the abstracts. We obtained the full article text for abstracts consistent with study eligibility. Overall, we only considered published studies for inclusion, either as the full article or abstract. Data abstraction strategy Two independent review authors abstracted all information and data using standardized data abstraction forms with a third review author verifying the data. Abstraction included information on pertinent methodological aspects of the study design, characteristics of the participants, the specic dose of the study drug used and the outcomes assessed (e.g. number of vertebral, non-verte5
O N
LY
bral, hip and wrist fractures). For fracture data, we considered all reported fractures (whether clinical or radiographic). For the yearly data, our unit of analysis was number of patients sustaining a fracture. If an article reported yearly data, we used the time points available. For baseline denominators we used the same baseline denominator for each time point. For follow-up denominators we used any yearly follow-up number of patients reported in the article, if available. If these were not available, we assumed a uniform drop-out rate for each year and calculated the denominators by determining the proportion of participants that would have remained at the end of the year in question based on the number of withdrawals over the course of the study. If an article reported only end of study outcomes, these were used for our analysis with the exception of outcomes for which the numerator was zero for both treatment groups. In these instances, we included the outcome (with any necessary adjustments for follow-up denominators) for the earlier years in the duration of the study. For example, if a trial reported zero hip fractures for both treatment arms at the end of year three, we would also include in our analysis zero hip fractures for that trial at years one and two. For person year data, the unit of analysis, if available, was number of fractures. When these data were not available (that is, in the majority of cases), we used the number of women sustaining a fracture. For denominators, we multiplied the number of women followed by the length of the study. For radiographic vertebral fractures we used the number of women with available radiographs, if the number was reported in the article. For clinical fractures, we estimated the number of women followed over the duration of the study by taking the mean of the baseline and follow-up denominators. Strategy for quality assessment Two review authors assessed each eligible RCT for quality based on allocation concealment. Research has shown that lack of adequate allocation concealment is associated with bias, (Higgins 2005) and studies can be judged on the method of allocation concealment. The method for assigning participants to interventions should be robust against patient and clinician bias and its description should be clear. The review authors were required to indicate whether allocation concealment was adequate (A), unclear (B), or inadequate (C) as per the Cochrane Collaboration criteria as follows: Adequate: The following are some approaches that can be used to ensure adequate concealment schemes: centralized or pharmacycontrolled randomization, pre-numbered or coded identical containers which are administered serially to participants, on-site com-
puter system combined with allocations kept in a locked unreadable computer le that can be accessed only after the characteristics of an enrolled participant have been entered or sequentially numbered or sealed, opaque envelopes. Other approaches may include those similar to ones listed previously, along with reassurance that the person who generated the allocation scheme did not administer it. Inadequate: Approaches to allocation concealment that are considered inadequate include: alternation, use of case record numbers, dates of birth or day of the week and any procedure that is entirely transparent before allocation, such as an open list of random numbers. Unclear: When studies do not report any concealment approach, adequacy should be considered unclear. Examples include merely stating that a list or table were used, only specifying that sealed envelopes were used and reporting an apparently adequate concealment scheme in combination with other information that leads the review author to be suspicious. In addition, blinding and loss to follow up were assessed. Data analysis For the analysis of fractures (i.e. vertebral, non-vertebral, hip and wrist), we calculated the relative risk (RR) of fracture using a xedeffect model. The methods we used for pooling the results are described elsewhere by Fleiss, (Fleiss 1993). We calculated the pooled or weighted RRs using the general inverse variance method for the weights. For the pooled results, we calculated site-specic 95% condence intervals (CIs) for vertebral, non-vertebral, hip and wrist fractures and we tested for association using a chi-squared test procedure taking P value < 0.05 for presence of statistical association. Statistically signicant risk reductions were considered to be clinically important if a 15% or greater relative benet was shown. We also tested for homogeneity using a chi-squared test procedure taking the specic cut-off for presence of statistical heterogeneity as P = 0.10 ( Fleiss 1993). In the event of signicant heterogeneity, a random-effects model was used. If the relative risk reduction (RRR) was signicant (P < 0.05), then the absolute risk reduction (ARR) and number needed to treat to benet (NNTB) were calculated. For these calculations, the ve year risk of fracture in the untreated population was based on the FRACTURE Index (FI) of Black et al., (Black 2001) and the lifetime and ve year age-specic risks in the untreated population were based on the model by Doherty et al., (Doherty 2001) for predicting osteoporotic fractures in postmenopausal women (Figure 1; Figure 2; Figure 3; Figure 4.)
IN
TE
AL
SE
O N
LY
Figure 1. Models for fracture risk in postmenopausal women: FRACTURE Index
IN
TE
AL
SE
O N
LY
7
Figure 2. Five year risk of fracture by Quintiles of the FRACTURE Index: Assessment with bone mineral density
Figure 3. Five year risk of fracture by Quintiles of the FRACTURE Index: Assessment without bone mineral density
IN
TE
AL
SE
O N
LY
8
Figure 4. Estimated ve year age-specic risks of rst and subsequent osteoporotic fractures (from Doherty et al 2001)
Trials varied as to the length of treatment (for example, one to four years) and the number of patients available for study at the start of treatment (that is, baseline denominator) compared to those available at different time points during the trial (that is, followup denominators). The base case taken for the review of fractures considered the data available for the longest period of time for the treatment in the trial (that is, all years) and used the baseline denominators for the number of patients in the trial. Data was initially pooled broadly across primary and secondary trials. The overall analysis was also considered using person years of observation. In addition, we conducted subgroup analysis for: 1) primary versus secondary, 2) treatment duration and 3) treatment dose. Furthermore, we conducted sensitivity analysis for: 1) baseline denominators versus follow-up denominators, 2) xed versus random-effects model, and 3) baseline vertebral fracture rate. For the last sensitivity analysis, recall that the vertebral fracture criteria for a trial to be considered secondary was a prevalence of vertebral fracture at baseline of greater than 20%. A sensitivity analysis using different vertebral fracture rates (i.e. 100%, > 80%, > 60%, > 40%, > 20%) without the BMD and age criteria was conducted. Such sensitivity analysis allowed evaluating whether the effect of bisphosphonates on the secondary prevention of osteoporotic fractures varied, depending on how strictly secondary prevention was dened. Grading of evidence Results from the primary analyses were graded according to the system described in the 2004 book Evidence-based Rheumatol-
IN
TE
AL
SE
ogy (Tugwell 2004) and recommended by the Cochrane Musculoskeletal Group: Platinum To achieve the platinum level of evidence, a published systematic review that has at least two randomized controlled trials each satisfying the following is required. Sample sizes of at least 50 per group - if these do not nd a statistically signicant difference, they are adequately powered for a 20% relative difference in the relevant outcome. Blinding of patients and assessors for outcomes. Handling of withdrawals > 80% follow up (imputations based on methods such as Last Observation Carried Forward (LOCF) are acceptable). Concealment of treatment allocation. Gold The gold level of evidence requires at least one randomized clinical trial meeting all of the following criteria for the major outcome(s) as reported: Sample sizes of at least 50 per group - if these do not nd a statistically signicant difference, they are adequately powered for a 20% relative difference in the relevant outcome. Blinding of patients and assessors for outcomes. Handling of withdrawals > 80% follow up (imputations based on methods such as LOCF are acceptable). Concealment of treatment allocation.
O N
LY
9
IN
TE
AL
SE
O N
Silver The silver level of evidence requires a randomized trial that does not meet the above criteria for gold or platinum ranking or evidence from at least one study of non-randomized cohorts that did and did not receive the therapy, or evidence from at least one high quality case-control study. A randomized trial with a headto-head comparison of agents would be considered silver level ranking unless a reference were provided to a comparison of one of the agents to placebo showing at least a 20% relative difference. Bronze The bronze level of evidence requires at least one high quality case series without controls (including simple before/after studies in which patients act as their own control) or a conclusion derived In addition, for the outcomes of vertebral and hip fractures we prepared Summary of Findings tables using the GRADE criteria from the GRADE working group (GRADE 2004), (Figure 5; Figure 6).
from expert opinion based on clinical experience without reference Clinical relevance tables Clinical relevance tables were compiled under Additional tables to improve the readability of the review. Results were presented within the context of both the study population and moderate/ high risk women from the population at large. The number needed to treat to benet (NNTB) was calculated using the relative risk (RR) in combination with either the risk of fracture in the control group, or the ve year FRACTURE Index (Black 2001). To do this, the Visual Rx NNT calculator (Cates 2004) was used. The weighted absolute risk difference was calculated using the risk difference (RD) statistic in RevMan and RR-1 was used to calculate the relative per cent change (Table 1; Table 2).
LY
10
Figure 5. Summary of Findings for Primary Prevention
IN
TE
AL
SE
O N
LY
11
Figure 6. Summary of Findings for Secondary Prevention
IN
TE
AL
SE
O N
LY
12
RESULTS Description of studies

See: Characteristics of included studies; Characteristics of excluded studies. Quantity of research available The literature search revealed 419 citations as depicted in Figure 7. Of these, 36 articles were retrieved for further scrutiny. A total of 29 articles were excluded for various reasons including lack of appropriate control group (4), (Brown 2002b; Harris 2001b; Harris 1999b; Kushida 2004) lack of fracture outcome (8), (Delmas 1997; Dobnig 2006; Eriksen 2002; Hooper 1999; Hu 2005;
IN
TE
AL
SE
O N
LY
Leung 2005; Li 2005; Yildirim 2005) lack of appropriate fracture data (that is, reported as adverse events and unspecied) (1), (Hosking 2003) lack of randomization (5), (Goa 1998; Licata 1997; Reszka 1999; Singer 1995; Watts 1998) extension/discontinuation study (2), (Sorensen 2003; Ste-Marie 2004) duplicate report or earlier report of another study (7), (Eastell 2003; Miller 1999; Reginster 2001; Ribot 1999; Roux 2004; Watts 1999; Watts 2003) duration of therapy less than one year (1) (Zegels 2001) and no extractable data (1). (McClung 1998 1) If duplicate reports of the same study were found in preliminary abstracts and articles, the data from the most complete data set was analyzed. In the end, seven trials met the selection criteria (Clemmesen 1997; Fogelman 2000; Harris 1999; Hooper 2005; McClung 2001; Mortensen 1998; Reginster 2000). (See Figure 7.)
13
Figure 7. Summary of literature search for risedronate
IN
TE
AL
SE
O N
LY
14
Figure 8. Weighted relative risk of fracture after risedronate (5mg)
IN
TE
AL
Trial characteristics The characteristics of the seven selected trials are provided in the Characteristics of included studies table. In total, there were 14,049 women enrolled in these trials, and of these, 4746 received placebo. Two trials were primary prevention (Hooper 2005; Mortensen 1998) and the other ve studied women with low BMD on densitometry or who had experienced previous fractures or both. (Clemmesen 1997; Fogelman 2000; Harris 1999; McClung 2001; Reginster 2000) All trials had three treatment arms that included two arms involving risedronate at different doses (i.e. 2.5 mg or 5 mg) or different schedules or both and a placebo arm. In our analysis, we did not include data from the two risedronate arms in any single meta-analysis in order to avoid duplication of data in the pooled estimates. In the case of the McClung study, results were not reported separately in the form of raw data for the 2.5 mg and 5 mg doses. We therefore used the combined dose data in our primary 5 mg analysis. Length of follow up ranged from two to three years and mean age was 51 to 78. No trial excluded women with a history of gastrointestinal disease and in three trials fractures were evaluated as the stated primary outcome (Harris 1999; McClung 2001; Reginster 2000). All trials with the exception of one (Mortensen 1998) administered 1000 mg of calcium to all patients. Three trials administered up to 500
IU of vitamin D (Harris 1999; McClung 2001; Reginster 2000).
Risk of bias in included studies

Treatment allocation was concealed in two trials (Harris 1999; Hooper 2005) and was unclear for the other ve . All seven trials had losses to follow up that exceeded 20%. Two trials (Clemmesen 1997; Hooper 2005) had losses to follow up between 20 and 30%, three trials between 30 and 40% (Fogelman 2000; Harris 1999; McClung 2001) and two trials exceeded 40% (Mortensen 1998; Reginster 2000). All trials were double blind.
Effects of interventions
Effect on fractures A summary of the overall review of fractures with the standard dose of risedronate (5 mg) for the base case (i.e. the longest treatment duration in the trial and using the baseline denominators for the number of patients) is provided in (Figure 8). Primary prevention estimates were available only for vertebral and non- vertebral fractures, neither of which were statistically signicant. For each of vertebral, non-vertebral, and hip fracture, the pooled estimate of the RR of fracture was signicant for secondary prevention. The exception was wrist fracture, which was not signicant.
SE
O N
LY
15
IN
TE
AL
Vertebral fractures For primary prevention, Mortensen (Mortensen 1998) reported no fractures leaving Hooper (Hooper 2005) as the only primary prevention study reporting vertebral fractures. The estimate of the RR was not statistically signicant (RR 0.97, 95% CI 0.42 to 2.25). Vertebral fractures were reported in four secondary prevention trials (Clemmesen 1997; Fogelman 2000; Harris 1999; Reginster 2000). One study was not included in analysis due to the apparent inclusion of an off-drug treatment period in the data (Clemmesen 1997). The pooled estimate of the RR of vertebral fractures from the three trials (Fogelman 2000; Harris 1999; Reginster 2000) that could be analyzed demonstrated a signicant reduction (39%) in vertebral fractures (RR 0.61, 95% CI 0.50 to 0.76) as detailed in Comparison 01, 01. This demonstrates a fracture risk reduction with risedronate 5 mg daily and results are consistent across studies (P = 0.75). For both primary and secondary prevention trials combined the pooled estimate (RR 0.63, 95% CI 0.51 to 0.77) was similar to that for that for secondary prevention alone Corresponding to the signicant RRR of 39% for the secondary prevention of vertebral fractures, the absolute measures ARR and
SE
NNT of the ve year risk of vertebral fracture after treatment with risedronate were calculated for different levels of increasing risk as given by the FI. Results are provided in (Figure 9) as well as for increasing age in (Figure 10). For the illustrative case of the patient with a FI of 6 to 7, the ARR in vertebral fracture was 2.8% (that is, a reduction in risk from 7.1% to 4.3%) and the NNTB was 36 (that is, 36 patients need to be treated to avoid one vertebral fracture). Across the range of increasing FI risk, the ARR for vertebral fracture ranged from 0.5% to 4.4% and the NNT to avoid one vertebral fracture ranged from 214 to 23. For the illustrative patient in the age group 60 to 64 years, the ARR for the rst vertebral fracture was 0.4% (that is, a reduction in risk from 1.0% to 0.6%) and the NNT was 256 patients treated to avoid the rst fracture. The ARR for a subsequent fracture was 3.8% (that is, a reduction in risk from 9.7% to 5.9%) and the NNT was 26 patients treated to avoid one subsequent fracture. For increasing age, the ve-year age-specic ARR for the rst vertebral fracture increased from 0.1% for the youngest age group (50 to 54 years) to 1.8% in the highest age group (90+ years) and the NNT decreased from 1282 to 55. For the subsequent fracture, ARR increased from 0.2% to 10.9% and the NNT decreased from 513 to 9.
O N
LY
16
Figure 9. Five year FRACTURE Index specic risk of fracture after risedronate (5mg)
IN
TE
AL
SE
O N
LY
17
Figure 10. Five year age-specic risk of rst and subsequent fracture after risedronate (5 mg)
Non-vertebral fractures One primary prevention trial (Hooper 2005) reported non-vertebral fractures. The risk estimate (RR 0.81, 95% CI 0.25 to 2.58) was not statistically signicant. Non-vertebral fractures were reported in ve secondary prevention trials (Clemmesen 1997; Fogelman 2000; Harris 1999; McClung 2001; Reginster 2000). After excluding the study with a possible off drug treatment period, the pooled estimate of the RR of nonvertebral fractures from four trials (Fogelman 2000; Harris 1999; McClung 2001; Reginster 2000) that could be analyzed demonstrated a signicant 20% reduction in non-vertebral fractures (RR 0.80, 95% CI 0.72 to 0.90). Results are provided in Comparison 02, 01 and they were consistent across studies (P = 0.43). For both primary and secondary prevention trials combined the pooled estimate (RR 0.80, 95% CI 0.72 to 0.90) was the same that for that for secondary prevention alone as the Hooper trial comprised only 1.04% of the total weight. Corresponding to the signicant RRR of 20% for the secondary prevention of non-vertebral fractures, the absolute measure: ARR and NNT of the ve year risk of non-vertebral fracture after treatment with risedronate were calculated for different levels of increasing risk as given by the FI (Figure 9) and for increasing age (Figure 10). For the illustrative case of the patient with a FI of 6 to 7, the ARR in non-vertebral fracture was 4.0% (that is, a reduction in risk from 19.8% to 15.8%) and the NNT was 25 (that is, 25 patients need to be treated to avoid one non-vertebral fracture). Across the range of increasing FI risk, the ARR for non-vertebral fracture ranged from 1.7% to 5.5% and the NNT to avoid one
IN
TE
AL
SE
non-vertebral fracture ranged from 58 to 18. For the illustrative patient in the age group 60 to 64 years, the ARR for the rst nonvertebral fracture was 0.6% (i.e. a reduction in risk from 3.1% to 2.5%) and the NNT was 161 patients treated to avoid the rst fracture. The ARR for a subsequent fracture was 1.2% (that is, a reduction in risk from 6.2% to 5.0%) and the NNT was 81 patients treated to avoid one subsequent fracture. For increasing age, the ve year age-specic ARR for the rst non-vertebral fracture increased from 0.3% for the youngest age group (50 to 54 years) to 7.0% in the highest age group (90+ years). Accordingly, the NNT decreased from 313 to 14. For the subsequent fracture, ARR increased from 0.5% to 7.5% and NNT decreased from 192 to 13. Hip fractures Hip fractures were reported in three of the secondary prevention trials (Harris 1999; McClung 2001; Reginster 2000). The pooled estimate of the RR of hip fractures from the three trials showed a signicant 26% reduction in hip fractures (RR 0.74, 95% CI 0.59 to 0.94). Results are provided in Comparison 03, 01 and they were consistent across trials (P = 0.95). Corresponding to the signicant RRR of 26% for the secondary prevention of hip fractures, the absolute measures ARR and NNT of the ve year risk of hip fracture after treatment with risedronate were calculated for different levels of increasing risk as given by the FI (Figure 9) and for increasing age (Figure 10). For the illustrative case of the patient with a FI of 6 to 7, the ARR for hip fracture was 1.0% (that is, a reduction in risk from 3.9% to 2.9%) and the
O N
LY
18
Figure 11. Weighted relative risk (RR) of fracture after risedronate (5 mg): Person years
IN
TE
AL
SE
O N
NNT was 99 (that is, 99 patients need to be treated to avoid one hip fracture). Across the range of increasing FI risk, the ARR for hip fracture ranged from 0.1% to 2.3% and the NNT to avoid one hip fracture ranged from 962 to 44. For the illustrative patient in the age group 60 to 64 years, the ARR for the rst hip fracture was 0.05% (i.e. a reduction in risk from 0.2% to 0.15%) and the NNT was 1923 patients treated to avoid the rst fracture. The ARR and NNT for a subsequent fracture were the same. For increasing age, the ve year age-specic ARR for the rst hip fracture increased from less than 0.05% for the youngest age group (50 to 54 years) to 5.4% in the highest age group (90+ years) and the NNT decreased from more than 1923 to 18. For the subsequent fracture, ARR increased from less than 0.05% to 5.9% and the NNT decreased from more than 962 to 17. Wrist fractures Wrist fractures were reported in two of the secondary trials (Harris 1999; Reginster 2000). The pooled estimate of the RR of wrist fractures from the two trials showed a 33% reduction which was
not signicant (RR 0.67, 95% CI 0.42 to 1.07). Results are given in Comparison 04, 01 and they were consistent across trials (P = 0.81). Additional analysis Person years Results were similar for vertebral, non-vertebral, hip and wrist fractures when using person years as depicted in (Figure 11). Of note, the pooled estimates of the RR for the secondary prevention trials all showed a signicant risk reduction of fracture, with the exception of wrist fractures. Subgroup analysis Treatment duration No trends over years of treatments were found that deviated from the overall RR estimates. For details, please refer to Figure 12 and Comparisons 05, 01; 06, 01; 07, 01; 08, 01; 09, 01; 10, 01; 11, 01.
LY
19
Figure 12. Weighted relative risk of fracture after risedronate (5 mg) by years of treatment
Treatment dose For risedronate 2.5 mg, fracture data were only available for vertebral and non-vertebral sites. There was no protective effect observed for the primary prevention of vertebral fractures. For the secondary prevention trials, the decrease in risk of vertebral fracture was signicant and there was a further, albeit slight, decrease in risk with the 2.5 mg dose, compared with 5 mg (Figure 13; Figure 14; Figure 15; Comparison 12, 01; Comparison 12, 02). For non-vertebral fractures, the decrease in risk was not statistically signicant for either primary or secondary prevention but was more pronounced with the 2.5 mg dose compared with 5 mg.
IN
TE
AL
SE
O N
LY
20
Figure 13. Weighted relative risk of fracture after risedronate by dose
IN
TE
AL
SE
O N
LY
21
Figure 14. Weighted relative risk of fracture after risedronate by dose: Person years
IN
TE
AL
SE
O N
LY
22
Figure 15. Weighted relative risk of fracture after risedronate by years of treatment and dose
IN
TE
AL
SE
O N
LY
23
IN
TE
AL
SE
O N
LY
24
Sensitivity analysis Baseline versus follow-up denominators By using the data available for the longest treatment duration, standard dose of risedronate (5 mg) and follow-up denominators for the number of patients in the trials, a summary of the overall review of fractures was prepared (Figure 16). These data are also provided by years of treatment (Figure 17). The pooled estimates of the RR of fracture after risedronate were essentially the same as those obtained using the baseline denominators, with the exception that the pooled estimate of the RR of wrist fractures from two secondary prevention trials (Harris 1999; Reginster 2000) demonstrated a signicant reduction (39%) in wrist fractures (RR 0.61, 95% CI 0.38 to 0.96) for risedronate 5 mg.
Figure 16. Weighted relative risk of fracture after risedronate by dose: Follow-up denominators
IN
TE
AL
SE
O N
LY
25
Figure 17. Weighted relative risk of fracture after risedronate by years of treatment and dose: Follow-up denominators
IN
TE
AL
SE
O N
LY
26
Random versus xed-effect model There were a few instances where heterogeneity of the trial results was such that a random-effects model was needed (that is, only involving the person years analysis). In general, results obtained using the random and xed-effect models were similar. Baseline vertebral fracture rate Using different baseline vertebral fracture rates (i.e. 100%, > 80%, > 60%, > 40%, > 20%) for dening secondary trials, a summary of the overall review of fractures was prepared (Figure 18). For
vertebral, non-vertebral and wrist fractures, the pooled estimates of the RR of fracture after risedronate were essentially the same as those obtained using the denition of secondary trials with the > 20% baseline fracture rate. Although the result for hip fractures became non-signicant when the criteria increased from > 40% to > 60%, the relative risk of fracture was about the same (0.74 compared to 0.81) but the condence interval was now wider since the large 2001 trial by McClung et al (McClung 2001) with over 6000 participants was excluded from the analysis.
Figure 18. Weighted relative risk (RR) of fracture after risedronate (5 mg): sensitivity analysis by baseline prevalent vertebral fracture rate
IN
TE
AL
SE
O N
LY
27
Adverse events A summary of the adverse drug reactions reported in the six randomized placebo-controlled trials of risedronate is provided in Figure 19; Figure 20; Figure 21 and Figure 22. In general, the reported events were similar between risedronate and placebo. In particular, there were ve studies reporting any upper gastrointestinal events resulting in an overall RR 1.01 (95% CI 0.94 to 1.09) and four studies reporting esophageal ulcer resulting in an overall RR 0.75 (95% CI 0.39 to 1.47). Figure 19. Summary of adverse drug events reported in randomized placebo-controlled trials of risedronate (part 1)
IN
TE
AL
SE
O N
LY
28
Figure 20. Summary of adverse drug events reported in randomized placebo-controlled trials of risedronate (part 2)
IN
TE
AL
SE
O N
LY
29
Figure 21. Summary of adverse drug events reported in randomized placebo-controlled trials of risedronate (part 3
Figure 22. Summary of adverse drug events reported in randomized placebo-controlled trials of risedronate (part 4)
IN
TE
AL
SE
O N
LY
30
Toxicity and withdrawals Discontinuations due to adverse events or dropouts overall were available and analyzed for ve (Fogelman 2000; Harris 1999; Hooper 2005; McClung 2001; Reginster 2000) and six (Clemmesen 1997; Fogelman 2000; Harris 1999; Hooper 2005; McClung 2001; Reginster 2000) of the risedronate trials respectively. The pooled estimate demonstrated no statistical difference between risedronate and placebo for the risk of discontinuing medication due to adverse events (RR 0.96, 95% CI 0.88 to 1.05) or for dropouts overall (RR 0.96, 95% CI 0.91 to 1.00) (Comparison 13.01, 13.02). Results were consistent across the trials.
placebo. In keeping with this, it was concluded that study participants tolerated their risedronate treatment. Although no increased incidence of adverse effects were detected with risedronate, external to the randomized controlled trials, concerns exist regarding the potential risk of upper gastrointestinal events and osteonecrosis of the jaw. Study limitations The results of this systematic review are believed to be robust, as we performed a comprehensive literature search, inclusion and exclusion criteria were specied and we conducted a rigorous data analysis. A potential limitation of our approach may be that the update search (that is, 2000 to 2004) did not include non-MEDLINE indexed journals. Recent empirical evidence indicates that this approach may have introduced a slight risk of bias into our meta-analysis. On average, such bias is estimated to result in a 6% variation in the pooled results (Egger 2003; Sampson 2003). Accordingly, given that the initial literature search (that is, 1966 to 2000) was very extensive, the impact of only using MEDLINE for the search update is expected to be minimal, if any. This was conrmed by a parallel literature search update (that is, 1999 to July 2004) that was conducted for an economic report for the Canadian Agency for Drugs and Technologies in Health. The search update included etidronate, alendronate and risedronate (daily dose regimen only) in addition to teriparatide. We searched a number of databases (i.e. The Cochrane Library, MEDLINE, EMBASE, BIOSIS Previews, Toxle, PubMed) and no additional articles meeting the inclusion criteria were identied (CADTH 2006). While our methodology was robust, the results of our meta-analysis, however, are only as strong as the primary studies included. In keeping with this, the main limitations with regard to study quality were fracture assessment and classication, the lack of clarity of the concealment of allocation and large numbers of withdrawals. A potential source of heterogeneity is the lack of uniform denition of non-vertebral fracture. While some researchers may use a rather liberal denition (any fracture other than vertebral fracture), others may use a more conservative denition which includes only fractures of the hip, clavicle, humerus, wrist, pelvis or leg (Mayo Clin Proc 2005). Also, the statistical power to detect heterogeneity was extremely limited for some tests due to the low number of fractures in some categories. Another consideration is the fact that fracture data was not the primary outcome for some of the trials. In particular, four of the seven risedronate trials had fractures as the primary outcome. There is another source of heterogeneity, and possible bias, related to some of the secondary prevention studies. It concerns the inclusion of participants with a low BMD but no proven fractures and the difculty in discriminating between fracture types (traumatic versus pathological). Also, at times, our criteria for categorizing a study as a secondary prevention trial differed from those used by study investigators. An example is the study by McClung et al (McClung 2001) which included two
31
DISCUSSION
Based on longest treatment duration and use of baseline denominators for the number of patients in each of the included trials, risedronate at 5 mg per day was associated with a statistically signicant and clinically important reduction in the secondary prevention of vertebral, non-vertebral and hip fractures but not wrist. The level of evidence was Gold for vertebral and non-vertebral fractures and Silver for hip and wrist. For primary prevention, one trial reported vertebral and non-vertebral fractures but no statistically signicant reductions were found (Silver level evidence).
Secondary analyses showed that a dose of 2.5 mg, for secondary prevention resulted in a statistically signicant and clinically important reduction in vertebral fractures that is comparable to that of the 5 mg dose (Gold level evidence). Only one primary prevention trial reported vertebral fractures for the 2.5 mg dose and no reduction was found (Silver level evidence). For non-vertebral fractures, no statistically signicant reductions were shown for either primary or secondary prevention (Silver level evidence). In the case of hip fractures, McClung et al (2001), randomized women to receive placebo, 2.5 mg risedronate or 5.0 mg risedronate. This was done separately for younger (70 to 79 years) and older (80 + years) women. Although the results were not generally reported separately by dose and age groups, the authors did note that the effects of dose (2.5 mg versus 5.0 mg) for the younger age group were similar (for 2.5 mg RR 0.5 (95% CI 0.3 to 0.9); for 5 mg RR 0.7 (95% CI 0.4 to 1.1). There were no trials available to assess the efcacy of the 2.5 mg dose for wrist fractures. There were no substantive differences in the results whether baseline, end of study, or person year denominators were used. Sensitivity analyses indicated that there were no major differences based on the percentage of baseline vertebral fractures. Further, no trends were found for years of treatment.
Adverse drug events observed with risedronate were similar to the ones observed with placebo. The use of risedronate was not associated with any statistically signicant difference in withdrawals due to adverse events (RR 0.96; 95% CI 0.88 to 1.05) or overall withdrawals (RR 0.96; 95% CI 0.91 to 1.00), when compared to
IN
TE
AL
SE
O N
LY
age subgroups. In our review, we considered that both subgroups evaluated the effect of risedronate in secondary prevention of osteoporotic fractures, although we acknowledge study investigators may not have intended the same. However both subgroups satised two of our secondary prevention criteria, that is, baseline fracture rate and age. Indeed, both age subgroups had a baseline fracture rate > 20% (primary prevention studies were required to have a baseline fracture rate of < 20%). In addition, study participants in both subgroups were over 62 years of age (we considered a trial as secondary prevention if the average age was above 62 years). Finally, the exploration of differences in effect between primary and secondary prevention trials was a particular concern in our review.
which ranged between one and three years, does not allow for the assessment of long term toxicity associated with risedronate. Recently, there have been concerns regarding the potential risk of over suppressing bone turnover resulting in osteonecrosis of the jaw (ONJ), (Khosla 2007; Woo 2006). Although the majority of reported cases of ONJ have occurred in cancer patients receiving the intravenous bisphosphonates zolendrate or pamidronate (at higher cumulative doses than used in the treatment of postmenopausal osteoporosis), some osteoporosis patients receiving oral alendronate or risedronate have developed the condition as well. In a systematic review of cases reported in the medical literature, 13 of 368 bisphosphonate treated ONJ patients had received alendronate and one risedronate (Woo 2006). Since that publication, a review conducted by the American Society for Bone and Mineral Research (ASBMR) Task Force on Bisphosphonate-Associated ONJ has identied studies reporting a total of 67 cases (64 alendronate, 2 risedronate and 1 etidronate) among osteoporosis and Pagets disease patients (Khosla 2007). Most notably, an Australian study reported 30 of 114 ONJ cases related to alendronate (22 of whom were under treatment for osteoporosis) and 2 related to risedronate. The median time to onset, for alendronate, was 24 months. The most common triggering factor was dental extraction (Mavrokokki 2007). The ASBMR task force has pointed out that the incidence of ONJ in the general population not exposed to bisphosphonates is unknown, information on the incidence of ONJ is rapidly evolving and that, often, the case ascertainment has bee inadequate. They recommend that a hierarchy of evidence quality, based on the completeness of information across seven categories related to diagnosis and history, should be established for all future studies reporting cases of ONJ (Khosla 2007). No cases of ONJ were reported in any of the risedronate trials in our review. Finally, because RCTs are not designed to measure ADEs, particularly rare ones, it is common practice to include sources of information other than RCTs. While some reviewers include ADEs reported in observational studies, we elected to obtain information from the Canadian Adverse Drug Reaction Monitoring Program (CADRMP, see below) (Health Canada 2005). Generalizability of ndings Generalizability of our ndings is limited by the controlled design of the trials included in our review. Study participants were carefully selected in these trials and so utilization of the drugs in real life may vary substantially from study conditions. Furthermore, study participants were observed for periods of time varying from one to three years. Consequently, while our results provide support for efcacy (that is, can the intervention have an effect on outcome?) they may possibly only provide partial information on the long-term effectiveness (that is, does the intervention have an effect on outcome?) of bisphosphonates in preventing osteoporotic fractures.
32
Another methodological limitation concerns the approach used for concealment of treatment allocation which was not reported for most trials (that is, classied as unclear). Treatment allocation was concealed in two trials (Harris 1999; Hooper 2005) and was unclear for the other ve trials. The length of follow up in the studies (two to three years) is an additional limitation. It is difcult to extrapolate beyond the duration of the follow-up trials in the review with respect to the long-term impact on fractures. Ultimately, data from longer-term trials will help establish if the effect on fractures is maintained, increased or diminished.
In regards to our own methodology, we acknowledge that the approach used to evaluate the effect of bisphosphonates over time may result in some estimates that are not robust. In particular, in order to determine the effect on the ve year risk of fracture we based our evaluation on the FRACTURE Index by Black et al. (Black 2001) and for lifetime and ve year age-specic risks we used an existing model from Doherty et al. (Doherty 2001). Although this latter approach allowed us to estimate the variation in risks between younger and older postmenopausal women, these estimates may be associated with a certain level of uncertainty. Nonetheless, we believe such information may be useful to decision-makers. Lastly, a limitation of evaluating data on adverse effects from summary meta-analyses is that participants in RCTs tend to be healthier with fewer co-morbid diseases and therefore the results may not be generalizable to clinical practice. None of the six risedronate trials, however, explicitly excluded patients with pre existing GI disorders. Furthermore, RCTs are underpowered for rare effects and meta-analyses of these trials generally cannot provide conclusive information pertaining to drug toxicity. In addition, the heterogeneity of the adverse drug events (ADEs) reported in the RCTs described in this review, including their nature, low occurrence, and way they were assessed by investigators, were obstacles for meta-analysis. As well, the follow up for the included trials,
IN
TE
AL
SE
O N
LY
From a safety perspective, we could not nd any statistically signicant difference in either the rates of adverse drug events or withdrawal rates due to adverse drug events between patients receiving a bisphosphonate or patients receiving a placebo. However, outside controlled trials, concerns exist regarding the safe use of these drugs, especially alendronate and, to a lesser extent, risedronate, for which esophageal ulcers and gastritis have been reported (Kherani 2002). While such adverse events have mainly been identied through case reports and endoscopic studies, similar concerns are also reected in the proportions of gastrointesti-
nal adverse drug reactions associated with the use of risedronate reported to the CADRMP (Health Canada 2005). Indeed, GI adverse drug reactions represented 35% of reports for risedronate (Figure 23). These proportions should however be interpreted with caution as adverse drug reactions are reported to CADRMP on a volunteer basis by health professionals, which means that several reactions may be unreported. Indeed, it is estimated that less than 10% of adverse reactions are reported to Health Canada (Health Canada 2005c). Also, a denite cause-effect relationship has not been established for these adverse drug reactions.
IN
TE
AL
SE
O N
LY
33
Figure 23. Adverse drug reactions reported to CADRMP for etidronate, alendronate and risedronate
IN
TE
AL
AUTHORS CONCLUSIONS Implications for practice
Risedronate demonstrated a clinically important benet in the secondary prevention of the majority of osteoporotic fractures. At a dose of 5 mg per day, statistically signicant reductions in vertebral, non-vertebral and hip fractures were observed (but not wrist). No statistically signicant reduction in the primary prevention of vertebral and non-vertebral fractures and no evidence was available for the primary prevention of hip or wrist fractures. No increased incidence of adverse effects were detected with risedronate, but
34 Risedronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women (Review) Copyright 2010 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
SE
O N
LY
clinicians should be aware that outside of randomized controlled trials, concerns exist regarding the potential risk of upper gastrointestinal events and, less commonly, osteonecrosis of the jaw. The prevention of osteoporotic fractures is an important public health intervention. This is particularly true for hip and clinical vertebral fractures (i.e. fractures of the spine that present for medical attention). The RR of death following such fractures is six- to nine-fold greater in postmenopausal women aged 55 to 81 years with low BMD, which represents a typical postmenopausal population (Cauley 2000). In most cases, the mortality increase reects poor underlying health status and comorbidity in addition to the fracture itself (Cauley 2000). Osteoporotic fractures are also associated with increase in morbidity, as it is reported that 50% of women who sustain a hip fracture do not return to their usual daily activities (Brown 2002b) while 33% will require long-term care. Accordingly, reducing the incidence of such fractures can potentially increase the quality of life of patients with osteoporosis. Such interventions may also potentially decrease mortality.
fully resolved the question of whether important differences in risk reduction across groups of patients with varying degrees of osteoporosis exist. The impact of risedronate on the RR of nonvertebral fractures in populations without osteoporosis also merits further investigation. Additional research is also needed to clarify the role of risedronate in the primary prevention of osteoporotic fractures. There is also a need for additional post marketing safety data. Finally, research into combination therapy with higher doses of vitamin D or anabolic agents would be merited as would research concerning adherence to bisphosphonate therapy. Given the morbidity consequences associated with osteoporotic fractures, preventing their recurrence can potentially lessen the need for community-based health services (for example, home care). It may also reduce or delay the demand for long-term care beds. However, very little comparative information is currently available to support this (Hodsman 2002). There is also a lack of studies which evaluated the effect of bisphosphonates on hospital admissions (Hodsman 2002).
Implications for research
It has been suggested from clinical trials with the bisphosphonates (Black 2000; Harris 1999; McClung 2001) that their effect in reducing non-vertebral fractures may be greater in patients with lower BMD who initiate treatment. The existing data have not
Fogelman 2000 {published data only} Fogelman I, Ribot C, Smith R, Ethgen D, Sod E, Reginster JY. Risedronate reverses bone loss in postmenopausal women with low bone mass: results from a multinational, double-blind, placebo-controlled trial. BMD-MN Study Group. Journal of Clinical Endocrinology & Metabolism 2000;85(5):1895900. Harris 1999 {published data only} Harris ST, Watts NB, Genant HK, McKeever CD, Hangartner T, Keller M, et al.Effects of risedronate treatment on vertebral and nonvertebral fractures in women with postmenopausal osteoporosis: a randomized controlled trial. Vertebral Efcacy With Risedronate Therapy (VERT) Study Group. JAMA 1999;282(14):134452. Hooper 2005 {published data only} Hooper MJ, Ebeling PR, Roberts AP, Graham JJ, Nicholson GC, DEmden M, et al.Risedronate prevents bone loss in
IN
Clemmesen 1997 {published data only} Clemmesen B, Ravn P, Zegels B, Taquet AN, Christiansen C, Reginster JY. A 2-year phase II study with 1-year of follow-up of risedronate (NE-58095) in postmenopausal osteoporosis. Osteoporosis International 1997;7(5):48895.
TE
AL
REFERENCES
early postmenopausal women: a prospective randomized, placebo-controlled trial. Climacteric 2005;8:25162. McClung 2001 {published data only} McClung MR, Geusens P, Miller PD, Zippel H, Bensen WG, Roux C, et al.Effect of risedronate on the risk of hip fracture in elderly women. Hip Intervention Program Study Group.[see comment]. New England Journal of Medicine 2001;344(5):33340. Mortensen 1998 {published data only} Mortensen L, Charles P, Bekker PJ, Digennaro J, Johnston CC Jr. Risedronate increases bone mass in an early postmenopausal population: two years of treatment plus one year of follow-up. Journal of Clinical Endocrinology & Metabolism 1998;83(2):396402. Reginster 2000 {published data only} Reginster J, Minne HW, Sorensen OH, Hooper M, Roux C, Brandi ML, Lund B, Ethgen D, Pack S, Roumagnac I, Eastell R. Randomized trial of the effects of risedronate on vertebral fractures in women with established postmenopausal osteoporosis. Vertebral Efcacy with Risedronate Therapy (VERT) Study Group. Osteoporosis International 2000;11(1):8391.
35
SE
ACKNOWLEDGEMENTS
Thank you to Lara Maxwell and Marie Andree Nowlan from the Musculoskeletal Group for their editorial assistance and Tamara Rader for her assistance with the Consumer Summaries.
O N
LY

Brown 2002 {published data only} Brown JP, Kendler DL, McClung MR, Emkey RD, Adachi JD, Bolognese MA, et al.The efcacy and tolerability of risedronate once a week for the treatment of postmenopausal osteoporosis. Calcied Tissue International 2002;71(2): 10311. Delmas 1997 {published data only} Delmas PD, Balena R, Confravreux E, Hardouin C, Hardy P, Bremond A. Bisphosphonate risedronate prevents bone loss in women with articial menopause due to chemotherapy of breast cancer: a double-blind, placebocontrolled study. Journal of Clinical Oncology 1997;15(3): 95562. Dobnig 2006 {published data only} Dobnig H, Hofbauer LC, Viereck V, Obermayer-Pietsch B, Fahrleitner-Pammer A, Dobnig H, et al.Changes in the RANK ligand/osteoprotegerin system are correlated to changes in bone mineral density in bisphosphonate-treated osteoporotic patients. Osteoporosis International 2006;17(5): 693703. [MEDLINE: 13] Eastell 2003 {published data only} Eastell R, Barton I, Hannon RA, Chines A, Garnero P, Delmas PD. Relationship of early changes in bone resorption to the reduction in fracture risk with risedronate. Journal of Bone & Mineral Research 2003;18(6):10516.
controlled study. Current Medical Research & Opinion 2003; 19(5):38394. Hu 2005 {published data only} Hu YF, Sun ZQ. Quality of life in the treatment assessment of postmenopausal osteoporosis. [Chinese]. Zhong Nan da Xue Xue Bao Yi Xue Ban = Journal of Central South University Medical Sciences 2005;30(3):299303. [MEDLINE: 708] Kushida 2004 {published data only} Kushida K, Fukunaga M, Kishimoto H, Shiraki M, Itabashi A, Inoue T, et al.A comparison of incidences of vertebral fracture in Japanese patients with involutional osteoporosis treated with risedronate and etidronate: a randomized, double-masked trial. Journal of Bone & Mineral Metabolism 2004;22(5):46978. Leung 2005 {published data only} Leung JY, Ho AY, Ip TP, Lee G, Kung AW, Leung JYY, et al.The efcacy and tolerability of risedronate on bone mineral density and bone turnover markers in osteoporotic Chinese women: a randomized placebo-controlled study. Bone 2005;36(2):35864. [MEDLINE: 274] Li 2005 {published data only} Li Y, Zhang Z, Deng X, Chen L, Li Y, Zhang Z, et al.Efcacy and safety of risedronate sodium in treatment of postmenopausal osteoporosis. Journal of Huazhong University of Science and Technology 2005;Medical Sciences. 25(5):5279. [MEDLINE: 292] Licata 1997 {published data only} Licata AA. Bisphosphonate therapy.[see comment]. [Review]. American Journal of the Medical Sciences 1997; 313(1):1722. McClung 1998 1 {published data only} McClung M, Bensen WG, Bolognese M, Bonnick S, Ettinger M, Harris ST. Risedronate increases bone mineral density at the hip, spine and radius in postmenopausal women with low bone mass. Osteoporosis International 1998;8(Suppl 3):111. Miller 1999 {published data only} Miller P, Roux C, McClung M, Adami S, Eastell R, Ethgen D. Risedronate reduces hip fractures in patients with low femoral neck bone mineral density [abstract]. Arthritis and Rheumatism 1999;42:S287. Reginster 2001 {published data only} Reginster JY. Risedronate increases bone mineral density and reduces the vertebral fracture incidence in postmenopausal women.[see comment]. Clinical & Experimental Rheumatology 2001;19(2):1212. Reszka 1999 {published data only} Reszka AA, Halasy-Nagy JM, Masarachia PJ, Rodan GA. Bisphosphonates act directly on the osteoclast to induce caspase cleavage of mst1 kinase during apoptosis. A link between inhibition of the mevalonate pathway and regulation of an apoptosis-promoting kinase. Journal of Bone and Joint Surgery 1999;274(49):3496773. Ribot 1999 {published data only} Ribot C, Smith R, Fogelman I, Pack S, Ethgen D. Risedronate increases bone mineral density and is well
36
Harris 2001 {published data only} Harris ST, Eriksen EF, Davidson M, Ettinger MP, Moffett JA Jr, Baylink DJ, et al.Effect of combined risedronate and hormone replacement therapies on bone mineral density in postmenopausal women.[see comment]. Journal of Clinical Endocrinology & Metabolism 2001;86(5):18907. Hooper 1999 {published data only} Hooper M, Ebeling P, Roberts A. Risedronate prevents bone loss in early postmenopausal women [abstract]. Calcied Tissue International 1999, (Suppl 1):80. Hosking 2003 {published data only} Hosking D, Adami S, Felsenberg D, Andia JC, Valimaki M, Benhamou L, et al.Comparison of change in bone resorption and bone mineral density with once-weekly alendronate and daily risedronate: a randomised, placebo-
IN
Harris 1999b {published data only} Harris ST, Wasnich R, Ettinger M, Davidson M, Bosch L, Chines A. The effects of risedronate plus estrogen compared with estrogen alone in postmenopausal women. Journal of Bone & Mineral Research 1999;14(Suppl 1):S404.
TE
Goa 1998 {published data only} Goa KL, Balfour JA. Risedronate. [Review] [36 refs]. Drugs & Aging 1998;13(1):8391.
Eriksen 2002 {published data only} Eriksen EF, Melsen F, Sod E, Barton I, Chines A. Effects of long-term risedronate on bone quality and bone turnover in women with postmenopausal osteoporosis. Bone 2002;31 (5):6205.
AL
SE
O N
LY
tolerated in postmenopausal women with low bone mineral density. Journal of Bone & Mineral Research 1999;14:S163. Roux 2004 {published data only} Roux C, Seeman E, Eastell R, Adachi J, Jackson RD, Felsenberg D, et al.Efcacy of risedronate on clinical vertebral fractures within six months. Current Medical Research & Opinion 2004;20(4):4339. Singer 1995 {published data only} Singer FR, Minoofar PN. Bisphosphonates in the treatment of disorders of mineral metabolism. Advances in Endocrinology and Metabolism 1995;6:25988. Sorensen 2003 {published data only} Sorensen OH, Crawford GM, Mulder H, Hosking DJ, Gennari C, Mellstrom D, et al.Long-term efcacy of risedronate: a 5-year placebo-controlled clinical experience. Bone 2003;32(2):1206. Ste-Marie 2004 {published data only} Ste-Marie LG, Sod E, Johnson T, Chines A, Ste-Marie LG, Sod E, et al.Five years of treatment with risedronate and its effects on bone safety in women with postmenopausal osteoporosis. Calcied Tissue International 2004;75(6): 46976. [MEDLINE: 275] Watts 1998 {published data only} Watts NB. Treatment of osteoporosis with bisphosphonates. [Review]. Rheumatic Diseases Clinics of North America 1998; 27(2):41939.
alendronate in women with osteoporosis: the Fracture Intervention Trial. FIT Research Group. The Journal of Clinical Endocrinology and Metabolism 2000;85(11): 411824. Black 2001 Black DM, Steinbuch M, Palermo L, Dargent-Molina P, Lindsay R, Hoseyni MS, et al.An assessment tool for predicting fracture risk in postmenopausal women. Osteoporos International 2001;12(7):51928. Brown 2002b Brown JP, Josse RG. 2002 clinical practice guidelines for the diagnosis and management of osteoporosis in Canada. CMAJ 2002;167(10 Suppl):S134. Browner 1996 Browner WS, Pressman AR, Nevitt MC, Cummings SR. Mortality following fractures in older women. The study of osteoporotic fractures. Archives of Internal Medicine 1996; 156(14):15215. CADTH 2006 Coyle D, Hadj Tahar A, Murphy G, Perras C, Skidmore B, Boucher M, et al.Teriparatide and Bisphosphonates of Treatment of Osteoporosis In Women: A Clinical and Economic analysis. Canadian Agency for Drugs and Technologies in Health 2006. Cates 2004 Cates C. Visual Rx NNT Calculator 2.0. http:// www.nntonline.net/ 2004. Cauley 2000 Cauley JA, Thompson DE, Ensrud KC, Scott JC, Black D. Risk of mortality following clinical fractures. Osteoporos Int 2000;11(7):55661. Cooper 1993 Cooper C, Atkinson EJ, Jacobsen SJ, OFallon WM, Melton LJ, III. Population-based study of survival after osteoporotic fractures. Am J Epidemiol 1993;137(9):10011005. Cranney 2002 Cranney A, Tugwell P, Adachi J, Weaver B, Zytaruk N, Papaioannou A, Robinson V, Shea B, Wells G, Guyatt G. Meta-analyses of therapies for postmenopausal osteoporosis. III. Meta-analysis of risedronate for the treatment of postmenopausal osteoporosis. Endocr Rev 2002;23(4): 517523. Cranney 2003 Cranney A, Waldegger L, Zytaruk N, Shea B, Weaver B, Papaioannou A, Robinson V, Wells G, Tugwell P, Adachi JD, Guyatt G. Risedronate for the prevention and treatment of postmenopausal osteoporosis. Cochrane Database of Systematic Reviews 2003, Issue 4. Cummings 1989 Cummings SR, Black DM, Rubin SM. Lifetime risks of hip, Colles, or vertebral fracture and coronary heart disease among white postmenopausal women. Arch Intern Med 1989;149(11):24452448.
37
Yildirim 2005 {published data only} Yildirim K, Gureser G, Karatay S, Melikoglu MA, Ugur M, Erdal A, et al.Comparison of the effects of alendronate, risedronate and calcitonin treatment in postmenopausal osteoporosis. Journal of Back & Musculoskeletal Rehabilitation 2005;18(3-4):8595. [MEDLINE: 540] Zegels 2001 {published data only} Zegels B, Eastell R, Russell RG, Ethgen D, Roumagnac I, Collette J, et al.Effect of high doses of oral risedronate (20 mg/day) on serum parathyroid hormone levels and urinary collagen cross-link excretion in postmenopausal women with spinal osteoporosis. Bone 2001;28(1):10812.
Black 2000 Black DM, Thompson DE, Bauer DC, Ensrud K, Musliner T, Hochberg MC, et al.Fracture risk reduction with
IN
TE
Watts 2003 {published data only} Watts NB, Josse RG, Hamdy RC, Hughes RA, Manhart MD, Barton I, et al.Risedronate prevents new vertebral fractures in postmenopausal women at high risk.[see comment]. Journal of Clinical Endocrinology & Metabolism 2003;88(2):5429.
Watts 1999 {published data only} Watts NB, Roux C, Genant H, Adami S, Hangartner T, Miller P. Risedronate reduces vertebral fracture risk after the rst year of treatment in postmenopausal women with established osteoporosis. Journal of Bone & Mineral Research 1999;14(Suppl 1):S136.
AL
SE
O N
LY
Dickersin 1994 Dickersin K, Scherer R, Lefebvre C. Identifying relevant studies for systematic reviews. BMJ 1994;309(6964): 12861291. Doherty 2001 Doherty DA, Sanders KM, Kotowicz MA, Prince RL. Lifetime and ve-year age-specic risks of rst and subsequent osteoporotic fractures in postmenopausal women. Osteoporos Int 2001;12(1):1623. Egger 2003 Egger M, Juni P, Bartlett C, Holenstein F, Sterne J. How important are comprehensive literature searches and the assessment of trial quality in systematic reviews? Empirical study. Health Technol Assess 2003;7(1):176. Fleiss 1993 Fleiss JL. The statistical basis of meta-analysis. Stat Methods Med Res 1993;2(2):121145. GRADE 2004 Atkins D, Best D, Briss PA, Eccles M, Falck-Ytter Y, Flottorp S, Guyatt GH, Harbour RT, Haugh MC, Henry D, Hill S, Jaeschke R, Leng G, Liberati A, Magrini N, Mason J, Middleton P, Mrukowicz J, OConnell D, Oxman AD, Phillips B, Schnemann HJ, Edejer TT, Varonen H, Vist GE, Williams JW Jr, Zaza S, GRADE Working Group. Grading quality of evidence and strength of recommendations.. BMJ Jun 19;(7454):1490 2004;328: 1490.
Kanis 1994 Kanis JA, Melton LJ, Christiansen C, Johnston CC, Khaltaev N. The diagnosis of osteoporosis. Journal of Bone and Mineral Research 1994;9(8):113741. Kherani 2002 Kherani RB, Papaioannou A, Adachi JD. Long-term tolerability of the bisphosphonates in postmenopausal osteoporosis: a comparative review. Drug Safety 2002;25 (11):78190. Khosla 2007 Khosla S, Burr D, Cauley J, Dempster DW, Ebeling PR, Felsenberg D, et al.Bisphosphonate-associated osteonecrosis of the jaw: report of a task force of the American Society for Bone and Mineral Research. Journal of Bone and Minereral Research 2007;22:147991. Mavrokokki 2007 Mavrokokki T, Cheng A, Stein B, Gross A. Nature and Frequency of Bisphosphonate-Associated Osteonecrosis of the Jaws in Australia. Journal of Oral Maxillofacial Surgery 2007;65:41523.
TE
Harris 2001b Harris, S.T, Eriksen, E.F, Davidson, M, Ettinger, M.P, Moffett, A. H, Baylink, D.J. Effect of combined risedronate and hormone replacement therapies on bone mineral density in postmenopausal women. Journal of Clinical Endocrinology and Metabolism 2001;86(5):18907.
Hanley 2003 Hanley DA. Osteoporosis. In: Gray J editor(s). Therapeutic Choices. 4th Edition. Ottawa: Canadian Pharmaceutical Association, 2003:63746.
AL
Health Canada 2005 Canadian Adverse drug Reaction Monitoring Program. Risedronate. In: Health Canada, editor(s). Adverse reaction database. Ottawa, 2005. Health Canada 2005c Health Canada. Medeffect [website]. Health Canada YR: 2005. Ottawa. Higgins 2005 Higgins JPT, Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions 4.2.5 [updated May 2005]. In: The Cochrane Library, Issue 3, 2005. Chichester, UK: John Wiley & Sons, Ltd. Hodsman 2002 Hodsman AB, Hanley DA, Josse R. Do bisphosphonates reduce the risk of osteoporotic fractures? An evaluation of the evidence to date. CMAJ 2002;166(11):142630.
IN
SE
Mayo Clin Proc 2005 Alendronate and vertebral fracture risk [multiple letters]. Mayo Clinic Proceedings 2005;80:123341.
Melton 1989 Melton LJ, III, Kan SH, Frye MA, Wahner HW, OFallon WM, Riggs BL. Epidemiology of vertebral fractures in women. American Journal of Epidemiology 1989;129(5): 100011. NIH Consensus 2001 NIH Consensus Development Panel on Osteoporosis Prevention, Diagnosis, Therapy. Osteoporosis prevention, diagnosis, and therapy. JAMA 2001;285(6):78595. Sampson 2003 Sampson M, Barrowman NJ, Moher D, Klassen TP, Pham B, Platt R, et al.Should meta-analysts search Embase in addition to Medline?. Journal of Clinical Epidemiology 2003; 56(10):94355. Shukla 2003 Shukla V. Treating osteoporosis with teriparatide: many unknowns? [Issues in emerging health technologies issue 51]. Ottawa: Canadian Coordinating Ofce of health Technology Assessment, 2003. Tugwell 2004 Tugwell P, Shea B, Boers M, Simons L, Strand V, Wells G. Evidence-based Rheumatology. BMJ Books, 2004. WHO 1994 WHO. Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: Report of a WHO study group. World Health Organ Technical Report Series 1994;843:1129. Woo 2006 Woo SB, Hellstein JW, Kalmar JR. Narrative [corrected] review: bisphosphonates and osteonecrosis of the jaws. Annals of Internal Medicine 2006;144(10):75361.
38
O N
LY
References to other published versions of this review

CADTH 2006b Wells GA, Cranney A, Bouchere M, Peterson J, Shea B, Robinson V, et al.Bisphosphonates for the primary and secondary prevention of osteoporotic fractures in postmenopausal women: a meta-analysis [Technology report no 69]. Ottawa: Canadian Agency for Drugs and Technologies in Health 2006. Indicates the major publication for the study
IN
TE
AL
SE
O N
LY
39
CHARACTERISTICS OF STUDIES Characteristics of included studies [ordered by study ID]

Clemmesen 1997 Methods Randomized controlled trial Secondary prevention Blinding: double blind unspecied Study length: 2 years on drug plus 3rd year drug free follow up Withdrawals: Continuous: 15/44 (34.1%) Cyclic: 11/44 (25.0%) Placebo: 13/44 (29.5%) Total: 39/132 (29.5%)
Participants
Interventions
Outcomes
Notes Risk of bias
IN
TE
Risedronate 2.5 mg/day or cyclical risedronate 2.5 mg/day for 2 wk followed by placebo for 10 wk vs. placebo (Calcium 1000 mg/day) This study was not included in the analysis as the assessment period for fractures appeared to include the off drug treatment period Vertebral Fractures: Lateral thoracic and lumbar radiographs performed under standardized conditions. At the end of study radiographs were displayed simultaneously in chronological order for masked assessment of vertebrae T4 to L5. Vertebral anterior and posterior heights were measured to the nearest millimetre with a transparent ruler. Belgian site incident fractures were dened as = 15% height reduction in anterior to posterior ratio or in the anterior or posterior wall as compared to adjacent vertebrae. For the Danish site the reduction was 25% Non-vertebral Fractures: Subjects reported all adverse events and were questioned about intercurrent symptoms and illnesses at every visit
Source: Outpatients attending clinics which specialized in osteoporosis. Study was carried out in two centres one in Denmark and one in Belgium Inclusion Criteria: Healthy age 53-81, postmenopausal at least 1 year. Had at least one but no more than 4 vertebral fractures and at least 3 intact lumbar vertebrae Exclusion Criteria: Estrogen or calcitonin within 12 mos, any bisphosphonate or uoride, medications affecting bone metabolism, secondary causes of osteoporosis. Treatment N: 44, 44 Control N: 44 Age: 68.3 (5.7); YSM: 20.3 (7.3) Calcium: not reported BMD: 0.78 (0.14) g/cm2 T-score: -2.4 Fractures: 100%
AL
SE
O N
LY
40
Clemmesen 1997
(Continued)
Item Allocation concealment? Fogelman 2000 Methods
Authors judgement Unclear
Description B - Unclear
Interventions
Outcomes
Notes Risk of bias

IN
TE
Risedronate 2.5 mg/day or risedronate 5 mg/day vs. placebo. (Calcium 1000 mg/day) Vertebral Fractures: Lateral and anterior-posterior thoracic and lumbar (T4-L4) radiographs were taken at baseline and lateral radiographs taken at end of study. Fractures were identied by quantitative morphometry guidelines of US National Osteoporosis Foundation Working Group on Vertebral Fractures and veried by a radiologist. A fracture was dened by any vertebral height ratio falling 3 SD of the study population mean Non-Vertebral: Adverse events (including vertebral and non-vertebral fractures) were assessed throughout the study via spontaneous reports and direct questioning
AL
Participants
Source: 13 centres in France, UK, the Netherlands, Belgium and Germany Inclusion Criteria: Women up to 80 years who had been postmenopausal 1 year. Mean lumbar T-score of -2 or less Exclusion Criteria: Hyperthyroidism, hyperparathyroidism, or osteomalacia within 1 year, cancer, abnormalities affecting lumbar BMD measurement, drugs within 6 - 12 months affecting bone metabolism including parenteral Vitamin D = 10,000 IU. Note: patients with previous or ongoing upper GI disease were not excluded Treatment N: 184, 177 Control N: 180 Age: 64.7 (7.2); YSM: 17.7 (9.4) Calcium: not reported BMD: 0.74 (0.08) t-score: -2.9 Fractures: 30%
SE
Randomized controlled trial Secondary prevention Blinding: double blind unspecied Study length: 2 (2.5 mg patients from 9 of 13 centres (N = 76) discontinued prior to the end of the study due to protocol amendment.) Withdrawals: 2.5 mg: 111/184 (60.0%) 5.0 mg: 38/177 (21.5%) Placebo: 37/180 (20.6%) Total: 186/543 (34.3%) [not including protocol amendment: 110/467 (23.6%)]
O N
LY
Fogelman 2000
(Continued)
Item Allocation concealment? Harris 1999 Methods
Authors judgement Unclear
Description B - Unclear
Interventions
Outcomes
Notes
IN
TE
Risedronate 2.5 mg/day or risedronate 5 mg/day vs. placebo. (Calcium 1000 mg/day and if 25-hydroxyvitamin D level was < 40 nmol/L they received up to 500 IU/ day cholecalciferol Vertebral Fractures: Lateral thoracolumbar T4-L4 spine radiographs were obtained at baseline and annually. Prevalent and incident fractures were diagnosed quantitatively and semi-quantitatively. Quantitative assessment dened an incident fracture as a 15% decrease of anterior, posterior or middle vertebral height in a vertebra that was normal at baseline. In the semiquantitative assessment a new fracture was diagnosed if the grade changed from 0 (normal) to 1(mild) 2 (moderate), or 3 (severe). An independent radiologist adjudicated discrepancies between methods. Radiologists were blinded Non-vertebral Fractures: Radiographically conrmed fractures of the clavicle, humerus, wrist, pelvis, hip or leg, regardless of relationship to trauma
AL
SE
Participants
Source: 110 centres in North America Inclusion Criteria: Ambulatory women up to 85 years old who had been post menopausal at least 5 years. Minimum two radiographically conrmed vertebral T4-L4 fractures (ratio of anterior or middle vertebral height to the posterior height was = 0.8) or 1 vertebral fracture and lumbar BMD T-score = -2SD of young adult (0.83g/cm Hologic or 0.94g/cm Lunar) Exclusion Criteria: Conditions that could interfere with evaluation of spinal osteoporosis, calcitonin, calcitriol or Vitamin D within one month, anabolic steroids, or HRT within 3 months, bisphosphonates, uoride within 6 months. Note: Patients with previous or ongoing upper GI disease or use of medications associated with GI intolerance (NSAIDS or asprin) were not excluded. Treatment N: 817 Control N: 821, 820 Age: 69 (7.3); YSM: 24 (9.9) Calcium: not reported BMD: 0.83 g/cm2 (0.16) t-score: -2.4 Fractures: 81%
O N
Randomized controlled trial Secondary prevention Blinding: double blind, identical placebo, investigators and other study personnel remained blinded Study length: 3 (2.5 mg/day dose discontinued at 1 year due to protocol amendment.) Withdrawals:2.5 mg: 238/817 (29.1%) 5.0 mg: 332/821 (40.4%) Placebo: 370/820 (45.1%) Total: 940/2458 (38.2%)
LY
Harris 1999
(Continued)
Risk of bias Item Allocation concealment? Authors judgement Yes Description A - Adequate
Hooper 2005
Participants
Interventions
Outcomes
Notes Risk of bias Item Authors judgement Description
IN
TE
Source: 11 centres in Australia. Inclusion Criteria: Healthy women who were post menopausal 6- 36 months age with FSH > 50 MIU/ ml, and estradiol < 20 pg/ml. BMD T score > -2.5 (0.76g/cm2 Hologic or 0.8 g/cm2 Lunar). Exclusion Criteria: Hyperparathyroidism, hyperthyroidism, osteomalacia or treatments affecting bone metabolism. Patients not excluded due to GI disease or medications with potential to irritate GI tract Treatment N: 2.5 mg 127, 5 mg 129 Control N: 125 Age: 52.6 (3.3); YSM 3.9 (5.6) Calcium: not reported BMD: 1.079 (12) t-score: -0.410 (-0.1073) Fractures: 18.3% Risedronate 2.5 mg/day or 5.0 mg/day vs. placebo (Calcium 1000 mg/day) Vertebral Fractures: Lumbar and thoracic lateral and anterior-posterior radiographs were taken at baseline and 2 years. Deformity was conrmed by visual inspection and dened as vertebral height ratio below 3 SD of study population mean. Incident fracture was dened as a loss of 14% or more in anterior, posterior or middle vertebral height in a vertebra that was normal at baseline Non-vertebral Fractures: Were monitered as adverse events. At each visit patients were questioned regarding medically related changes in well being with particular attention to fractures and GI events
AL
SE
O N
LY
43
Methods
Randomized controlled trial Primary prevention Blinding: double blind, matching placebo Study length: 2 years Withdrawals: Risedronate 2.5 mg: 28/128 (21.9%) Risedronate 5 mg: 26/129 (20.2%) Placebo: 33/126 (26.2%) Total: 87/383 (22.8%)
Hooper 2005
(Continued)
Allocation concealment?
Yes
A - Adequate
McClung 2001 Methods Randomized controlled trial Secondary prevention Blinding: double blind, identical placebo Study length: 3 years Withdrawals:2.5/ 5.0 mg combined: 2197/6197 (35.5%) placebo: 1127/3134 (35.9%) Total: 3324/9331 (35.6%)
Participants
Interventions
Outcomes
Notes Risk of bias Item Authors judgement Description
IN
Source: 2 cohorts from 183 centres in North America, Europe, New Zealand and Australia Inclusion Criteria: Group 1 - Women age 70-79 with osteoporosis dened as femoral neck BMD T-score > 4 SD below peak bone mass (originally calculated according to densitometer reference data base and later according to Third National Health and Nutrition Examination Survey) or > -3 SD and at least one clinical risk factor i.e. difculty standing from a sitting position, poor tandem gait, fall-related injury in previous year, psychomotor score = 5 on Clifton Modied Gibson Spiral Maze test, smoking during previous 5 years, maternal history of hip fracture, previous hip fracture, hip axis length = 11.1cm. Group 2 - Women age 80 or older with at least one non-skeletal risk factor for hip fracture, a femoralneck t score < -4 or a femoral-neck t score < -3 plus a hip axis length = 11.1cm Exclusion Criteria: Major medical illness, recent history of cancer, another metabolic bone disease within 1 year, important lab test abnormalities, recent drugs affecting bone, allergy to bisphosphonates, bilateral hip fractures. Note: Patients with previous or ongoing upper GI or use of medications associated with GI intolerance (NSAIDS, asprin, proton pump inhibitors or antacids) were not excluded Treatment N: 3093, 3104 Control N: 3134 Age: 78.0 (9.7); YSM: 31.8 (19.3) Calcium: not reported BMD: not reported t-score (femoral): -3.7 Fractures: 42% Risedronate 2.5 mg/day or risedronate 5 mg/day vs. placebo (Calcium 1000 mg/day and if 25-hydroxyvitamin level was < 40 nmol/L, they received vitamin D 500 IU/day) Hip Fractures: Radiographically conrmed. Non-vertebral Fractures: Radiographically conrmed osteoporotic fractures of wrist leg, humerus, hip, pelvis or clavicle
TE
AL
SE
O N
LY
44
McClung 2001
(Continued)
Allocation concealment? Mortensen 1998 Methods
Unclear
B - Unclear
Interventions
Outcomes
Notes
IN
TE
Risedronate 5 mg/day or cyclical risedronate 5 mg/day for 1st 2 wk of every calendar month followed by placebo for remainder vs. placebo (Not required to take supplemental calcium.) Vertebral fractures - Thoracic and lumbar radiographs were taken at baseline, 7, 13 and 25 months and 12 months after treatment cessation. Vertebral deformities were dened as a 25% or more decrease in anterior, mid or posterior, height of a vertebra as compared to baseline and were evaluated for safety purposes Non-vertebral fractures- Ascertainment not specied but adverse events were recorded at all patient visits. This endpoint was not included in the meta-analysis as the evaluation period appeared to include the follow up time in which patients were off drug
AL
SE
Participants
Source: Two centres - USA and Denmark Inclusion Criteria: Ambulatory, active women 6-60 month postmenopausal (as measured by FSH and estradiol) weighing 45-90 kg (within 25% of normal height and weight). Lumbar spine within 2 SD of age matched bone mass Exclusion Criteria: Bisphosphonates, thyroid hormone therapy, glucocorticoids = 5 mg/day, anabolic agents, calcitonin, vitamin D > 400 IU/day, calcium > 1500/day, diuretics, anticonvulsants > 1 month in previous 6 months, HRT > 1 month in previous 6 months, uoride > 1 month ever, any bone disease including hyperparathyroidism, alcohol or drug abuse, psychiatric disease, any evidence of osteoporosis - vertebral deformity or osteoporosis related fracture of hip or wrist, bilateral oophorectomy, or articial menopause. Note: Nothing in criteria to indicate exclusion of those with upper GI disease. Treatment N: 37, 38 Control N: 36 Age: 51.2 (3.8); YSM: 2.7 (1.7) Calcium: 977 (535)mg/d BMD: 0.94 (0.11) g/cm2 T-score: -1.0 Fractures: 0% (excluded)
O N
Randomized controlled trial Primary prevention Blinding: double blind, identical placebo Study length: 2 on drug plus 3rd year drug free follow up. (Patients given option to continue in study after completing 1st year.) Withdrawals: Continuous: 20/37 (54.1%) Cyclic: 14/38 (36.8%) Placebo: 15/36 (41.7%) Total:49/111 (44%)
LY
Mortensen 1998
(Continued)
Risk of bias Item Allocation concealment? Reginster 2000 Authors judgement Unclear Description B - Unclear
Interventions
Outcomes
IN
TE
Risedronate 2.5 mg/day or risedronate 5 mg/day vs. placebo (Calcium 1000 mg/day and if 25-hydroxyvitamin level was < 40 nmol/l, they received vitamin D 500 IU/day) Primary Outcome: Vertebral fractures Vertebral Fractures: Lateral thoracolumbar T4-L4 spine radiographs were obtained at baseline, 1,2 and 3 years and were read in the order they were taken at a single radiology department. Prevalent and incident fractures were diagnosed quantitatively and semi-quantitatively. Quantitative assessment dened incident fracture as a 15% decrease of anterior, posterior or middle vertebral height in a vertebra that was normal at baseline. In the semiquantitative assessment a new fracture was diagnosed if the grade changed from 0 (normal) to 1(mild) 2 (moderate), or 3 (severe). An independent radiologist adjudicated discrepancies between methods Non-vertebral Fractures: Radiographically conrmed fractures of the clavicle, humerus, wrist, pelvis, hip
46
AL
Participants
Source: 80 European and Australian centres. VERT study. Inclusion Criteria: Ambulatory women up to 85 years old who had been post menopausal at least 5 years. Minimum two radiographically conrmed vertebral T4-L4 fractures Exclusion Criteria: Conditions that could interfere with evaluation of spinal osteoporosis, calcitonin, calcitriol or Vitamin D within one month, anabolic steroids, or HRT within 3 months, bisphosphonates, uoride within 6 months. Note: Patients with previous or ongoing upper GI disease or use of medications associated with GI intolerance (NSAIDS or asprin) were not excluded Treatment N: 408, 407 Control N: 407 Age: 71.0 (7.0); YSM: 24.4 (8.5) Calcium: not reported BMD: 0.79 g/cm2 (0.15) T-score: -2.7 Fractures: 100%
SE
O N
LY
Methods
Randomized controlled trial Secondary prevention Blinding: double blind unspecied Study length: 3 (2.5 mg/dose discontinued at 2 years due to protocol amendment) Withdrawals: 2.5 mg (at 1 year): 76/408 (18.6%) 5.0 mg (at 3 years): 156/407 (38.3%) Placebo (at 3 years): 186/407 (45.7%) Total for 5.0 mg and placebo: 342/814 (40%)
Reginster 2000
(Continued)
or leg, regardless of relationship to trauma Notes Risk of bias Item Allocation concealment? Authors judgement Unclear Description B - Unclear
Study Brown 2002 Delmas 1997 Dobnig 2006 Eastell 2003
Reason for exclusion
Eriksen 2002 Goa 1998 Harris 1999b Harris 2001 Hooper 1999 Hosking 2003
IN
TE
Lack of fracture outcome. Lack of fracture outcome. Duplicate report or earlier report of another study. Lack of fracture outcome. Non randomized. Lack of an appropriate control group. Lack of an appropriate control group. Lack of fracture outcome. Lack of appropriate fracture data (i.e. reported as adverse event and unspecied.)
47
AL
Characteristics of excluded studies [ordered by study ID]
SE
BMD: Lumbar Bone Mineral Density Calcium: Baseline calcium intake GI: gastointestinal HRT: hormone replacement therapy NSAID: Non steroidal anti inammatory drugs Participant Characteristics: Unless othewise specied, participant characteristics are presented as mean (SD) SD: Standard deviation t-score calculated using the lumbar spine BMD [(LS BMD -1.047)/0.110] vs: versus YSM: Years Since Menopause
O N
LY
(Continued)
Hu 2005 Kushida 2004 Leung 2005 Li 2005 Licata 1997 McClung 1998 1 Miller 1999 Reginster 2001 Reszka 1999 Ribot 1999 Roux 2004 Singer 1995 Sorensen 2003 Ste-Marie 2004 Watts 1998 Watts 1999 Watts 2003
Lack of fracture outcome. Lack of an appropriate control group. Lack of fracture outcome. Lack of fracture outcome. Non randomized. No extractable data. Duplicate report or earlier report of another study. Duplicate report or earlier report of another study. Non randomized.
Duplicate report or earlier report of another study. Duplicate report or earlier report of another study.
Extension / discontinuation study. Extension study.
Non randomized.
Yildirim 2005 Zegels 2001
IN
TE
Duplicate report or earlier report of another study. Duplicate report or earlier report of another study. Lack of fracture outcome. Duration of therapy < 1 year.
AL
Non randomized.
SE
O N
LY
48
DATA AND ANALYSES
Comparison 1. Risedronate 5 mg vs Control - all years baseline denominators
Outcome or subgroup title 1 Vertebral Fractures 1.1 Vertebral Primary 1.2 Vertebral secondary
Effect size 0.63 [0.51, 0.77] 0.97 [0.42, 2.25] 0.61 [0.50, 0.76]
IN
1 Hip Fractures 1.1 Hip primary 1.2 Hip secondary
TE
AL
1 Non Vertebral Fractures 1.1 Non Vertebral Primary 1.2 Non vertebral secondary
5 1 4
12397 254 12143
SE
No. of studies
No. of participants
O N
Statistical method Effect size 0.80 [0.72, 0.90] 0.81 [0.25, 2.58] 0.80 [0.72, 0.90] Statistical method Effect size 0.74 [0.59, 0.94] Not estimable 0.74 [0.59, 0.94] Statistical method Effect size 0.67 [0.42, 1.07] Not estimable 0.67 [0.42, 1.07]
Outcome or subgroup title 1 Wrist Fractures 1.1 Wrist primary 1.2 Wrist secondary
LY
49
Comparison 5. Risedronate 5 mg vs Control - 1 year baseline denominators
Outcome or subgroup title 1 Fractures 1.1 Vertebral primary 1.2 Vertebral secondary 1.3 Hip primary 1.4 Wrist primary
No. of participants
Effect size Subtotals only Not estimable 0.40 [0.27, 0.59] Not estimable Not estimable
73 2455 73 73
Comparison 6. Risedronate 5 mg vs Control - 2 years baseline denominators
Outcome or subgroup title 1 Fractures 1.1 Vertebral primary 1.2 Vertebral secondary 1.3 Non vertebral primary 1.4 Non vertebral secondary 1.5 Hip primary 1.6 Wrist secondary
No. of studies 3 2 1 1 1 1 1
No. of participants
SE
O N
Statistical method Effect size Subtotals only 0.97 [0.42, 2.25] 0.48 [0.21, 1.08] 0.81 [0.25, 2.58] 0.55 [0.22, 1.34] Not estimable Not estimable Statistical method Effect size Subtotals only 0.66 [0.37, 1.18] 0.63 [0.31, 1.28] Not estimable Not estimable
IN
TE
Outcome or subgroup title 1 Fractures 1.1 Vertebral 1.2 Non vertebral 1.3 Hip 1.4 Wrist
327 357 254 357 73 73
Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI)
AL
No. of participants 684 611 73 73
Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI)
LY
50
Outcome or subgroup title 1 Fractures 1.1 Vertebral secondary 1.2 Non vertebral secondary 1.3 Hip secondary 1.4 Wrist secondary
No. of participants
2455 11786 11786 2455
Comparison 9. Risedronate 2.5 mg vs Control - 1 year baseline denominators
Outcome or subgroup title 1 Fractures 1.1 Vertebral
No. of studies 2 2
No. of participants
O N
Statistical method
LY
Effect size Subtotals only 0.54 [0.38, 0.76] Effect size Subtotals only 0.92 [0.52, 1.62] 0.50 [0.21, 1.19] Effect size Subtotals only 1.08 [0.48, 2.46] 0.78 [0.35, 1.76] 0.49 [0.13, 1.92] 0.51 [0.17, 1.53]
51
2452
Comparison 10. Risedronate 2.5 mg vs Control - 2 years baseline denominators
TE
1 Fractures 1.1 Vertebral 1.2 Non vertebral
N
No. of studies 2 2 2 No. of studies 2 1 1 1 1
AL
No. of participants 540 540 No. of participants 252 288 252 288
U
Comparison 11. Risedronate 2.5 mg vs Control - 2 years baseline denominators
Outcome or subgroup title 1 Fractures 1.1 Vertebral primary 1.2 Vertebral secondary 1.3 Non vertebral primary 1.4 Non vertebral secondary
IN
SE
Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI)
Comparison 12. Risedronate 2.5 mg vs Control - all years baseline denominators
Outcome or subgroup title 1 Vertebral fractures 1.1 Primary prevention 1.2 Secondary prevention 2 Non-vertebral fractures 2.1 Primary prevention 2.2 Secondary prevention
No. of studies 4 1 3 2 1 1
No. of participants 2992 252 2740 540 252 288
Statistical method Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI)
Effect size 0.62 [0.46, 0.83] 1.08 [0.48, 2.46] 0.57 [0.42, 0.78] 0.50 [0.21, 1.19] 0.49 [0.13, 1.92] 0.51 [0.17, 1.53]
Outcome or subgroup title 1 Withdrawals due to side effects 2 Withdrawals overall
No. of studies 5 6
No. of participants 9204 12486
O N
Statistical method % NNTB
LY
Effect size 0.96 [0.88, 1.05] 0.96 [0.91, 1.00] Statistical Sig Quality Evidence of
52
Table 1. Clinical Relevance Table for Fracture Primary Prevention Trials
Verte327 (2) bral Fractures - (Trial Population) Primary Prevention 5mg/day for 2 yrs 95% condence interval Verte327 (2) bral Fractures - (Low Risk Woman) Primary Prevention 5mg/day for 2 years
TE
Outcome
IN
6.2% (6 out of 0% 0 fewer -3% (I) 100) patients out of 100
AL
(-5, 5)
ADDITIONAL TABLES
U
Not applicable Not Statis- Silver tically signicant (-58, 125) 1.2% (1 out of Not applicable -3% (I) 100) Not applicable Not Statis- Silver tically signicant
SE
Risk Ratio (M-H, Fixed, 95% CI) Risk Ratio (M-H, Fixed, 95% CI)
(Continued)
95% condence interval Verte327 (2) bral Fractures - (Moderate Risk Woman) Primary Prevention 5mg/ day for 2 yrs 95% condence interval Non Vertebral 254 (1) Fractures (Trial Population) Primary Prevention (5 mg/day for 2 yrs) 95% condence interval Non Vertebral 254 (1) Fractures - (Low Risk Woman) Primary Prevention(5 mg/day for 2 yrs) 95% condence interval
(-58, 125)
Not applicable Not Statis- Silver tically signicant
(-58, 125)
4.8% (5 out of -1% 1 fewer -19% (I) 100) patient out of 100
(-3, 4)
IN
TE
AL
SE
(-75, 158) Not applicable Not Statis- Silver tically signicant (-75, 158) 16.5% (17 out Not applicable -19% (I) of 100) Not applicable Not Statis- Silver tically signicant (-75, 158)
Non Vertebral 254 (1) Fractures - (Moderate Risk Woman) Primary Prevention(5 mg/day for 2 yrs) 95% condence interval
O N
LY
53
(Continued)
Hip Fractures 73 (1) - (Trial Population) Primary Prevention (5mg/day for 2 yrs) 95% condence interval Hip Fractures 73 (1) - (Low Risk Woman) Primary Prevention (5mg/day for 2 yrs) 95% condence interval Hip Fractures 73 (1) - (Moderate Risk Woman) Primary Prevention (5mg/ day for 2 yrs) 95% condence interval Wrist Frac- 73 (1) tures - (Trial Population) Primary Prevention (5mg/ day for 2 yrs) 95% condence interval Wrist Frac- 73 (1) tures - (Low Risk Woman) Primary Prevention (5mg/ day for 2 yrs) 95% condence interval
0.0 (0 out of 0% 0 fewer Not estimable 100) patients out of 100
(-5, 5)
TE
0% (0 out of -0% 0 fewer Not estimable 100) patients out of 100
AL
Not applicable Not statis- Silver tically signicant (-5, 5) Not available Not applicable Not estimable Not applicable Not statis- Silver tically signicant
IN
1.9% (2 out of Not applicable Not estimable 100)
SE
O N
0.4% (0 out of Not applicable Not estimable Not applicable Not Statis- Silver 100) (I) tically signicant
LY
54
(Continued)
Wrist Frac- 73 (1) tures - (Moderate Risk Woman) Primary Prevention (5mg/day for 2 yrs) 95% condence interval Legend Secondary prevention = bone density of at least 2 SD values below peak bone mass and/ or one or more vertebral compression fractures
Not available
Not applicable Not estimable
Not applicable Not statis- Silver tically signicant
For Trial Population rates are based on the event rate in the control group. Low and Moderate Risk, are 5 year community population risks derived from the following variables in the FRACTURE Index: age, fracture after 50 yrs, maternal hip fracture after 50 yrs., weight < 125 lbs, smoking, using arms to assist standing and BMD. Low = FRACTURE Index score 12, Moderate = FRACTURE Index score 5 (Black 2001) see Figure 1
IN
TE
AL
SE
O N
LY
Gold level: At least one randomised clinical trial meets all of the following criteria for the major outcome (s) as reported: Sample sizes of at least 50 per group. If a statistically signicant difference is not found they must be powered for 20% relative difference in the relevant outcome. Blinding of patients and assessors for outcomes. Handling of withdrawals > 80% follow up (imputations based on methods such as Last Observation Carried
55
(Continued)
Outcome
O N
% NNTB
Table 2. Clinical Relevance Table for Fracture Secondary Prevention Trials
LY
19 Statistically signicant (15, 30) 49 Statistically signicant (38, 79) 23 Statistically signicant
Forward (LOCF) acceptable). Concealment of treatment allocation. Silver level: Randomised trial does not meet the above criteria
Statistical Sig Quality Evidence Gold
of
Verte2,812 (3) bral Fractures - (Moderate Risk Woman) Secondary Prevention 5 mg/day for 3 years 95% condence interval Verte2,812 (3) bral Fractures - (High Risk Woman) Secondary Prevention 5 mg/ day for 3 yrs
AL
(-8, -3) (-50, -24) 5.3% (5 out of Not applicable -39% (I) 100)
Verte2,812 (3) bral Fractures - (Trial Population) Secondary Prevention 5 mg/ day for 3 yrs
SE
IN
TE
Gold
(-50, -24)
Gold
56
(Continued)
95% condence interval Non Vertebral 12,143 (4) Fractures - (Trial Population) Secondary Prevention(5 mg/ day for 2-3 yrs) 95% condence interval Non Vertebral 12,143 (4) Fractures - (Moderate Risk Woman) Secondary Prevention(5 mg/ day for 2-3 yrs) 95% condence interval Non Vertebral 12,143 (4) Fractures - (High Risk Woman) Secondary Prevention(5 mg/ day for 2-3 yrs) 95% condence interval Hip Fractures 11,786 (3) - (Trial Population) Secondary Prevention (5 mg/day for 3 yrs)
(-50, -24)
(18, 38)
49
Gold
O N
31 (22, 61) 19 (13, 37) 138
(-3, -1)
(-28, -10)
(35, 98)
LY
Statistically signicant Gold Statistically signicant Gold Statistically signicant Silver
57
AL R N
IN
TE
2.8 (3 out of -1% 1 fewer -26% (I) 100) patient out of 100
U
(-28, -10) (-28, -10)
SE
(Continued)
95% condence interval Hip Fractures 11,786 (3) - (Moderate Risk Woman) Secondary Prevention (5 mg/day for 3 yrs) 95% condence interval Hip Fractures 11,786 (3) - (High Risk Woman) Secondary Prevention (5 mg/day for 3 yrs) 95% condence interval Wrist Frac- 2,455 (2) tures - (Trial Population) Secondary Prevention (5 mg/day for 3 yrs) 95% condence interval
(-1, 0)
(-41, -6)
(88, 596)
203
Silver
(-41, -6)
(129, 878)
O N
45 (29, 192)
LY
Statistically signicant Silver
58
IN
TE
3.5% (4 out of -1% 1 fewer -33% (I) 100) patient out of 100
AL
(-2, 0)
U
Not available
SE
(-41, -6) Not applicable Not statis- Silver tically signicant (-58, 7) Not applicable -33% (I) Not applicable Not statis- Silver tically signicant (-58, 7) (I)
Wrist Frac- 2,455 (2) tures - (Moderate Risk Woman) Secondary Prevention (5 mg/day for 3 yrs) 95% condence interval
(Continued)
Wrist Frac- 2,455 (2) tures - (High Risk Woman) Secondary Prevention (5 mg/day for 3 yrs) 95% condence interval Legend Secondary prevention = bone density of at least 2 SD values below peak bone mass and/ or one or more vertebral compression fractures
Not available
Not applicable -33% (I)
Not applicable Not statis- Silver tically signicant
(-58, 7)
For Trial Population rates are based on the event rate in the control group. Moderate and High Risk, are 5 year community population risks derived from the following variables in the FRACTURE Index: age, fracture after 50 yrs, maternal hip fracture after 50 yrs, weight < 125 lbs, smoking, using arms to assist standing and BMD. Moderate = FRACTURE Index score 3-4, High = FRACTURE Index score 813 (Black 2001) see Figure 1
IN
TE
AL
SE
O N
LY
Gold level: At least one randomised clinical trial meets all of the following criteria for the major outcome (s) as reported: Sample sizes of at least 50 per group. If a statistically signicant difference is not found they must be powered for 20% relative difference in the relevant outcome. Blinding of patients and assessors for outcomes. Handling of withdrawals > 80% follow up (imputations based on methods such as Last Observation Carried
59
(Continued)
WHATS NEW
Last assessed as up-to-date: 13 November 2007.
Date 13 August 2008 28 May 2008
Event Amended Amended
Description
Absolute event rates included in the Plain language summary Converted to new review format. CMSG ID C072-R
HISTORY
Review rst published: Issue 4, 2003
Date
IN
TE
Event
AL
14 November 2007
New citation required and conclusions have changed
SE
O N
Description See published notes for details on update.
LY
Forward (LOCF) acceptable). Concealment of treatment allocation. Silver level: Randomised trial does not meet the above criteria
60
CONTRIBUTIONS OF AUTHORS
George Wells was involved in the conception, design and implementation of the project and contributed signicantly to the writing of the report. Ann Cranney was involved with the conception of the review, data abstraction, analysis, interpretation and revision of the nal report. Joan Peterson screened the literature, was involved in the data abstraction, quality assessment and analysis of the primary trials and contributed signicantly to the writing of the report. Michel Boucher assisted in the design of the analysis, reporting and interpretation of the ndings and was involved in the writing of the report. Beverley Shea was involved in developing the protocol and conducting the systematic review. Vivian Robinson was involved in developing the protocol and conducting the systematic review. Douglas Coyle assisted with the design of the review and reviewed the analysis. Peter Tugwell provided clinical rheumatology expertise and methodological guidance.
DECLARATIONS OF INTEREST
None at present.
The Cochrane Funding Arbitration Panel recommended withdrawal of the original review from The Cochrane Library. The original review was externally supported by Merck and Proctor & Gamble to support research staff.
SOURCES OF SUPPORT Internal sources
Ottawa Health Research Institute, Canada.
Canadian Agency for Drugs and Technologies in Health, Canada.
NOTES
This review updated a previously published review (Cranney 2003) of risedronate conceived, conducted and completed, in part, by the authors of this report. There were four general differences in the manner in which the two reviews were conducted. First, the current review included articles published after the previous review was completed (updated to Feb 2007). Second, although both published and unpublished data were used in the previous reviews, only published data were used in this review. Third, in the previous review the random-effects model was always used whereas in the present review the base analysis used the xed-effect model unless the results were heterogeneous. Fourth, although the previous review considered both BMD and fracture data, this review was only concerned with fractures and therefore the denition of primary and secondary prevention put more of an emphasis on the fracture study inclusion criteria.
Our updated literature search retrieved an additional trial (Hooper 2005) which was published after the previous review. There were two changes in the choice of the included trials. First, the trials by Mortensen (Mortensen 1998) at 5 mg per day (for non-vertebral fracture data) and Clemmesen (Clemmesen 1997) at 2.5 mg per day (for vertebral and non-vertebral fracture data) were included in the previous review but were excluded for the current review since they appeared to include an off drug treatment period. That
IN
TE
External sources
AL
This current review was updated without the support of any industry sponsor.
SE
O N
LY
is, the fractures may have occurred during the follow-up period on no treatment and not during the active treatment phase of the study. Second, in the previous review, additional data were obtained from the authors for the McClung trial (McClung 2001) but this additional information was not used in the current review.
INDEX TERMS Medical Subject Headings (MeSH)

Bone Density Conservation Agents [adverse effects; therapeutic use]; Etidronic Acid [adverse effects; analogs & derivatives; therapeutic use]; Fractures, Bone [ prevention & control]; Hip Fractures [prevention & control]; Osteoporosis, Postmenopausal [ drug therapy; prevention & control]; Randomized Controlled Trials as Topic; Spinal Fractures [prevention & control]
Female; Humans
IN
TE
AL
SE
O N
LY
62
MeSH check words
Strontium ranelate for preventing and treating postmenopausal osteoporosis (Review)

ODonnell S, Cranney A, Wells GA, Adachi JD, Reginster JY
Strontium ranelate for preventing and treating postmenopausal osteoporosis (Review) Copyright 2006 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd
IN
TE
AL
SE
O N
LY
1
TABLE OF CONTENTS ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CRITERIA FOR CONSIDERING STUDIES FOR THIS REVIEW . . . . . . . . . . . . . . . SEARCH METHODS FOR IDENTIFICATION OF STUDIES . . . . . . . . . . . . . . . . METHODS OF THE REVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . DESCRIPTION OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . METHODOLOGICAL QUALITY . . . . . . . . . . . . . . . . . . . . . . . . . . . RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . AUTHORS CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . POTENTIAL CONFLICT OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SOURCES OF SUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Characteristics of included studies . . . . . . . . . . . . . . . . . . . . . . . . . . Characteristics of excluded studies . . . . . . . . . . . . . . . . . . . . . . . . . . ADDITIONAL TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Table 01. Clinical relevance table strontium ranelate 2 g per day: Fractures & safety data . . . . . . . Table 02. Clinical relevance table strontium ranelate 2 g per day: BMD data . . . . . . . . . . . Table 03. Other adverse events from additional sources (EMEA 2004* and Servier**) . . . . . . . . ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparison 01. Fractures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparison 02. BMD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparison 03. Adverse Events . . . . . . . . . . . . . . . . . . . . . . . . . . . COVER SHEET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . GRAPHS AND OTHER TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 01.01. Comparison 01 Fractures, Outcome 01 Verterbral fractures . . . . . . . . . . . . Analysis 01.02. Comparison 01 Fractures, Outcome 02 Non-vertebral fractures . . . . . . . . . . Analysis 02.01. Comparison 02 BMD, Outcome 01 Lumbar spine BMD not adjusted for strontium content Analysis 02.02. Comparison 02 BMD, Outcome 02 Lumbar spine adjusted for strontium content . . . . Analysis 02.03. Comparison 02 BMD, Outcome 03 Femoral neck . . . . . . . . . . . . . . . Analysis 02.04. Comparison 02 BMD, Outcome 04 Total hip . . . . . . . . . . . . . . . . Analysis 03.01. Comparison 03 Adverse Events, Outcome 01 Total withdrawls . . . . . . . . . . Analysis 03.02. Comparison 03 Adverse Events, Outcome 02 Withdrawals due to adverse events . . . . Analysis 03.03. Comparison 03 Adverse Events, Outcome 03 Number of emergent adverse events . . . . Analysis 03.04. Comparison 03 Adverse Events, Outcome 04 Serious adverse events . . . . . . . . . Analysis 03.05. Comparison 03 Adverse Events, Outcome 05 Diarrhea . . . . . . . . . . . . . Analysis 03.06. Comparison 03 Adverse Events, Outcome 06 Gastritis . . . . . . . . . . . . . Analysis 03.07. Comparison 03 Adverse Events, Outcome 07 Deaths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 3 4 4 5 5 6 7 7 10 11 12 12 12 12 15 15 16 17 17 18 18 19 19 19 19 19 21 21 21 22 23 24 25 25 26 27 28 28 29 29
IN
TE
AL
SE
O N
LY
Strontium ranelate for preventing and treating postmenopausal osteoporosis (Review)

ODonnell S, Cranney A, Wells GA, Adachi JD, Reginster JY
Status: New This record should be cited as: ODonnell S, Cranney A, Wells GA, Adachi JD, Reginster JY. Strontium ranelate for preventing and treating postmenopausal osteoporosis. Cochrane Database of Systematic Reviews 2006, Issue 3. Art. No.: CD005326. DOI: 10.1002/14651858.CD005326.pub2.
ABSTRACT
Background Strontium ranelate is a new anti-osteoporosis therapy therefore, its benets and harms need to be known. Objectives To determine the efcacy and safety of strontium ranelate for the treatment and prevention of postmenopausal osteoporosis. Search strategy We searched MEDLINE (1996 to March 2005), EMBASE (1996 to week 9 2005), the Cochrane Library (1996 to Issue 1 2005), reference lists of relevant articles and conference proceedings from the last two years. Additional data was sought from authors and industry sponsors. Selection criteria We included randomized controlled trials (RCTs) of at least one year duration comparing strontium ranelate versus placebo reporting fracture incidence, bone mineral density (BMD), health related quality of life and/or safety outcomes in postmenopausal women. Treatment (versus prevention) population was dened as women with prevalent vertebral fractures and/or lumbar spine BMD T score < -2.5 SD. Data collection and analysis Two reviewers independently determined study eligibility, assessed trial quality and extracted the relevant data. Disagreements were resolved by consensus. RCTs were grouped by dose of strontium ranelate and treatment duration. Where possible, meta-analysis was conducted using the random effects model. Main results A total of four trials met our inclusion criteria, three of which investigated the effects of strontium ranelate compared to placebo in a treatment population (doses ranged from 0.5 to 2 g daily) and one, in a prevention population (doses 0.125, 0.5 and 1 g daily). In osteoporotic, postmenopausal women a 37% reduction in vertebral fractures (two trials, n = 5082, RR 0.63, 95% CI 0.56 to 0.71) and a 14% reduction in non-vertebral fractures (two trials, n = 6572, RR 0.86, 95% CI 0.75 to 0.98) was demonstrated over a three year period with 2 g of strontium ranelate daily. An increase in BMD at all sites was shown with the same dose: lumbar spine BMD (two trials, n = 1614, WMD adjusted for strontium content 5.44, 95% CI 3.41 to 7.46 and WMD not adjusted 11.29, 95% CI 10.22 to 12.37 over two years), femoral neck and total hip (two trials, n = 4230, WMD 8.25, 95% CI 7.84 to 8.66 and WMD 9.83, 95% CI 9.39 to 10.26 respectively over three years). One gram of strontium ranelate daily in postmenopausal women without osteoporosis increased BMD at all sites over a two year period: lumbar spine (one trial, n = 59, WMD adjusted for strontium content 2.39, 95% CI 0.15 to 4.63 and WMD not adjusted 6.68, 95% CI 5.16 to 8.20), femoral neck (one trial, n= 60, WMD 2.52, 95%CI 0.96 to 4.09) and total hip (one trial, n = 60, WMD 1.02, 95% CI 0.48 to 1.56). In both the treatment and prevention populations, lower doses of strontium ranelate were superior to placebo with the highest dose of strontium ranelate demonstrating the greatest reduction
Strontium ranelate for preventing and treating postmenopausal osteoporosis (Review) Copyright 2006 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd 1
IN
TE
AL
SE
O N
LY
This version rst published online: 19 July 2006 in Issue 3, 2006. Date of most recent substantive amendment: 24 May 2006
in vertebral fractures and increase in BMD. There is some evidence to suggest that 2 g of strontium ranelate daily compared to placebo may have a benecial effect on health related quality of life in postmenopausal women after three years of treatment. Two grams of strontium ranelate daily increased the risk of diarrhea (RR 1.38%, 95% CI 1.02 to 1.87); however, adverse events did not affect the risk of discontinuing strontium ranelate nor did it increase the risk of serious side effects, gastritis or death. Additional data obtained suggests that the risk of vascular system disorders including venous thromboembolism (two trials, n = 6669, 2.2% versus 1.5%, OR 1.5, 95% CI 1.1 to 2.1) and pulmonary embolism (two trials, n = 6669, 0.8% versus 0.4%, OR 1.7, 95% CI 1.0 to 3.1) as well as nervous system disorders such as headaches (3.9% versus 2.9%), seizures (0.3% versus 0.1%), memory loss (2.4% versus 1.9%) and disturbance in consciousness (2.5% versus 2.0%) is slightly increased with taking 2 g of strontium ranelate daily over a 3 to 4 year period. Authors conclusions There is silver level evidence to support the efcacy of strontium ranelate for the reduction of vertebral fractures (and to a lesser extent non-vertebral fractures) in postmenopausal osteoporotic women and an increase in BMD (all sites) in postmenopausal women with and without osteoporosis. Diarrhea may occur however, adverse events leading to study withdrawal were not signicantly increased in the strontium ranelate group. Potential risks to the vascular and neurological system associated with taking 2 g of strontium ranelate daily need to be further explored and quantied.
Strontium ranelate for osteoporosis in women after menopause This summary of a Cochrane review presents what we know from research about the effect of strontium ranelate for osteoporosis in women after menopause. The review shows that: There is silver level evidence that for treatment of osteoporosis in women after menopause, 2 g of strontium ranelate daily over 3 years decreases fractures in the spine and slightly decreases fractures not in the spine. Most women do not have side effects that would cause them to stop taking strontium ranelate. However, other research shows that harms could include a chance of blood clots and seizures, memory loss and consciousness. What is osteoporosis and how can strontium ranelate help? Osteoporosis is a condition in which bone loss occurs. Bone loss leads to weak brittle bones that can break easily, even during everyday activities. Breaks (fractures) of the spine or non-spine (e.g. wrist and hip) are the most common type. There are many drugs and minerals that work to treat osteoporosis. Strontium ranelate is a drug that decreases the chance of fractures by slowing the loss of bone and possibly by building new bone. It is a new drug and therefore its benets and harms need to be known. What are the results of this review? Women in the studies took 2 g of strontium ranelate or a placebo (fake tablets or powder). After 2 to 3 years, the number of fractures that occurred and bone mineral density was measured. Bone mineral density is a lab test to measure how dense or strong bones are in the hip, spine or neck. The higher the bone density the better.
strontium ranelate decreases spine fractures -13 out of 100 women had spine fractures taking strontium ranelate -21 out of 100 women had spine fractures taking a placebo strontium ranelate may decrease fractures that are not in the spine -11 out of 100 women had non-spine fractures taking strontium ranelate -13 out of 100 women had non-spine fractures taking a placebo strontium ranelate increases bone mineral density Harms of strontium ranelate In women after menopause who have osteoporosis: strontium ranelate did not cause side effects that would make them stop taking it
Strontium ranelate for preventing and treating postmenopausal osteoporosis (Review) Copyright 2006 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd 2
IN
Benets of strontium ranelate In women after menopause who have osteoporosis:
TE
AL
SE
O N
LY
strontium ranelate did not lead to serious side effects, stomach infections, back pain or death strontium ranelate increased diarrhea -7 out of 100 women had diarrhea taking strontium ranelate -5 out of 100 women had diarrhea taking a placebo Other research shows that harms could include a chance of blood clots, and seizures, memory loss and consciousness. The cause of these neurological side effects are not known. This review has several limitations which include difculty interpreting the change in bone mineral density due to the unique aspects of strontium in bone and incomplete follow-up of some patients within the individual trials.
BACKGROUND Osteoporosis is a skeletal disorder characterized by low bone mass and micro-architectural deterioration of bone resulting in an increase in bone fragility and risk of fracture (1993 Consensus; NIH 2001). It most often affects postmenopausal women as reductions in circulating levels of estrogen lead to accelerated bone turnover and resorption. The most common sites of osteoporotic fracture are the wrist, hip and spine however, osteoporotic fractures do occur at other sites (Seeley 1991). Osteoporotic fractures are a major burden for the individual, their families and society (Johnell 2004; Kanis 2003; Cauley 2000). Individuals who suffer osteoporotic fractures, particularly spine or hip, must deal with the complications which include reductions in health-related quality-adjusted life years, increased morbidity and mortality (Cauley 2000; Johnell 2004; Tosteson 2001). Furthermore, the direct and indirect expenditures associated with the care of these individuals, particularly hip fracture patients, is costly (US Dept Health 2004; Ray 1997). In Europe, the number of osteoporotic fractures was estimated to be 3.79 million in 2000 and the associated total direct costs were 31.7 billion Euros (Kanis 2004). In the United States, approximately 1.5 million osteoporotic fractures occur each year (US Dept Health 2004) and the cost of fractures was estimated to be 20 billion US dollars in 1995 (Ray 1997). With the aging of the population, and the age-specic increases in osteoporotic fracture rates, it has been suggested that these costs will more than double in the coming decades (Burge 2003).
AL
The bone fragility which characterizes this disease is a result of an imbalance in bone remodeling (bone resorption exceeds bone formation) and an increase in the rate of remodeling at the tissue level (Seeman 2002). Risk factors associated with fragility fracture include advancing age, prior fragility fracture, family history of osteoporosis/fracture and low bone mineral density (BMD) (Brown 2002). A working group of the World Health Organization in 1994 proposed that an individual with a BMD more than 2.5 standard deviations (SD) below the young adult mean has osteoporosis (WHO 1994). Furthermore, it has been estimated that for every one SD reduction of BMD, there is an increase in relative risk of fracture of approximately 1.5 to 2.6 (Marshall 1996).
IN
TE
SE
Effective therapies are available and have been demonstrated to reduce the relative risk of fracture by 40 to 60% (Cranney 2002). Pharmacotherapy for prevention and treatment of osteoporosis includes two primary types of drugs, anti resorptive and anabolic agents. Anti resorptive agents increase bone strength by decreasing the number of bone multcellular units. This reduces resporption and prevents further structural damage of trabecular bone and by reducing cortical porosity. In contrast, anabolic agents increase bone strength by increasing bone mass due to an increase in the number of bone multicellular units. As result the magnitude of the formation phase is greater than the resorption phase (Riggs 2005). The majority of the agents currently available for the treatment of osteoporosis are anti resorptive (e.g. bisphosphonates, estrogen, selective estrogen modulators and calcitonin) and there are a few anabolic agents (e.g. intermittent recombinant human parathyroid hormone and uoride) (Sorbera 2003). A novel oral agent, strontium ranelate, has been suggested to simultaneously decrease bone resorption and stimulate bone formation although there is some controversy surrounding its mechanism of action. Strontium ranelate consists of two divalent cation atoms of stable strontium (natural element) and an organic moiety (ranelic acid) which dissociates at the gastro-intestinal level. Strontium is a cation (i.e. positively charged ion) and physically closely related to calcium, an active component of the skeleton. Ranelic acid is an organic, highly polar molecule without pharmacological activity (EMEA 2004). In vitro, strontium ranelate has been suggested to have a dual effect on bone however, in vivo long term dosing of strontium ranelate in OVX rats and monkeys resulted in increased bone formation but non-signicant trends of bone resorption. In human studies (phase III trials), there is some evidence of increases in bone formation markers (serum bone-specic alkaline phosphatase and C-terminal propeptide of type I procollagen) and decreases in markers of bone resorption (serum C-telopeptide and urinary N-telopeptide cross links) from the third month of treatment (2 g of strontium ranelate daily) up to three years. Potential mechanisms of action include activation of calcium-sensing receptor or induction of cellular differentiation. The proposed in3
O N
LY
dication is for treatment of postmenopausal osteoporosis in order to reduce the risk of fracture (EMEA 2004). Different doses of strontium ranelate have been tested. In a two year randomized controlled trial doses from 0.5 g to 2 g per day were tested in 353 women with postmenopausal osteoporosis (Meunier 2002) and in another two year randomized controlled trial, doses from 0.125 g to 1 g per day were evaluated in 160 early postmenopausal women (Reginster 2002-1). In both trials, the primary efcacy endpoint was BMD and results showed a clear dose response. All tested doses were superior to placebo with the highest dose of strontium ranelate (2 g per day) demonstrating the greatest increase in BMD after adjusting for bone strontium content over two years (Reginster 2003-1). As a result, 2 g of strontium ranelate per day is considered the recommended daily dose and was the only dose evaluated in the two phase III trials (Meunier 2004-1; Reginster 2005). Given the potential advantages of strontium ranelate in the prevention and treatment of osteoporosis, and that it is a new therapeutic agent, it is important that the benets and harms of this therapy are fully explored through a systematic review of the literature. OBJECTIVES
investigated multiple interventions where the effect of strontium ranelate could not be separated out were not included. Types of outcome measures Efcacy measures: 1. The primary efcacy outcome was the number of women with incident vertebral and non-vertebral fractures (a feasible outcome for a treatment population). Asymptomatic vertebral fractures were included if they were either quantitatively or semiquantatively ascertained via a radiographic examination as well, symptomatic (or clinical) vertebral fractures as dened by acute back pain and radiographical ndings were also included. Non-vertebral fractures included all appendicular type fractures except fractures of the coccyx, skull, jaw, face, ankle, ngers or toes as they are not considered to be osteoporotic related (Meunier 2004-1; Reginster 2005). 2. The secondary efcacy outcome to fractures (or surrogate outcome) was the mean percent change in BMD of the lumbar spine, femoral neck and total hip measured by Dual Energy X-Ray Absorptiometry (DXA) at baseline and yearly intervals (a relevant outcome for both prevention and treatment populations). Strontium ranelate has a higher atomic number than calcium. When present in bone as it attenuates x-rays to a greater extent than calcium resulting in an overestimation of BMD as measured by DXA (Blake 2005). As a result, BMD measurements should be adjusted for strontium content in order to avoid such an artifactual increase in BMD. The correction used to adjust for the strontium content in bone in the lumbar spine BMD measurements has been described as an indirect method and based on: 1) the correlation between the strontium content measured in the iliac crest on bone biopsy and the area under the curve of the integrated strontium plasma curve; and 2) the correlation between the bone strontium content measured in lumbar vertebrae and the iliac crest in monkeys. However, given that no correlation has been established between femoral neck and iliac crest bone strontium content is not adjusted for at the other BMD sites (Meunier 2002). 3. Health Related Quality of Life (a relevant outcome for a treatment population). 4. Safety measures include the following (relevant for both prevention and treatment populations): i) Total withdrawals (the total number of withdrawals after enrolment in the study). ii) Withdrawals due to adverse events (withdrawals as a result of an adverse event) iii) Number of emergent adverse events (adverse events that developed during the study). iv) Serious adverse events (adverse events that were immediately life-threatening, or resulted in hospitalization, disability, malignant disease or death) (Reginster 2002-1). v) Number of adverse events affecting the gastrointestinal system (e.g. diarrhea or gastritis)
4
CRITERIA FOR CONSIDERING STUDIES FOR THIS REVIEW Types of studies
Randomized placebo or active comparator-controlled trials of at least one year duration were included in this review. Studies were excluded if they were not truly randomized (e.g. patients randomized using date of birth) but not on the basis of language of publication. Types of participants Postmenopausal women, in which menopause was either surgically or naturally induced, were included. Types of intervention Trials that investigated the effect of strontium ranelate versus placebo or an active comparator were included however, trials that
IN
TE
To assess the clinical efcacy and safety of strontium ranelate in the prevention and treatment of osteoporosis compared to placebo or active comparator in postmenopausal women through a systematic review of the literature. The following major endpoints were used for this purpose: 1) Fractures (vertebral and non-vertebral); 2) BMD; 3) Health related quality of life and; 4) Safety. A treatment (versus prevention) population was dened as postmenopausal women with prevalent vertebral fractures and/or lumbar spine BMD T score < -2.5 SD.
AL
SE
O N
LY
vi) Deaths
SEARCH METHODS FOR IDENTIFICATION OF STUDIES See: methods used in reviews. Strontium ranelate is a relatively new drug and it was anticipated that most of the trials would have been published in the past ve years therefore, our search focused on this time period only. Our search aimed to identify all trials of strontium ranelate for either the prevention or treatment of postmenopausal osteoporosis and we employed the following approaches (based on Cochrane search strategy outlined by Robinson and Dickersin) (Robinson 2002): An electronic search of MEDLINE (1996 to March 2005), EMBASE (1996 to week 9 2005) and the Cochrane Library (1996 to Issue 1 2005). Our search strategy included MeSH terms such as osteoporosis, postmenopausal and strontium ranelate in addition, complementary free text words. We limited the search to randomized controlled trials and supplemented it to include previously completed systematic reviews. A review of reference lists of relevant articles for additional published trials.
20 (double adj blind$).tw. 21 placebo$.tw. 22 or/15-21 23 22 and 14 24 Strontium/ 25 strontium.tw. 26 ranelate.tw. 27 prevos.tw. 28 or/24-27 29 28 and 23 30 limit 29 to yr=1996 - 2005 Embase Search Strategy 1 exp osteoporosis/ 2 bone loss$.tw. 3 osteoporos#s.tw. 4 bone density/ 5 (bone adj2 (density or fragil$)).tw. 6 bone mass.tw. 7 bmd.tw. 8 exp Fracture/ 9 fracture$.tw. 10 or/1-9 11 postmenopause/ 12 (post menopaus$ or postmenopaus$ or post-menopaus$).tw. 13 11 or 12 14 10 and 13 15 random$.tw. 16 factorial$.tw. 17 crossover$.tw. 18 placebo$.tw. 19 (singl$ adj blind$).tw. 20 (doubl$ adj blind$).tw. 21 assign$.tw. 22 allocat$.tw. 23 volunteer$.tw. 24 randomized controlled trials/ 25 double-blind method/ 26 single-blind method/ 27 or/15-26 28 27 and 14 29 STRONTIUM/ 30 strontium.tw. 31 ranelate.tw. 32 prevos.tw. 33 or/29-32 34 27 and 33 35 limit 34 to yr=1996 - 2005
MEDLINE Search Strategy 1 *Osteoporosis/ 2 osteoporos#s.tw. 3 bone loss$.tw. 4 Bone Density/ 5 (bone adj2 (density or fragil$)).tw. 6 bone mass.tw. 7 bmd.tw. 8 exp Fractures, Bone/ 9 fracture$.tw. 10 or/1-9 11 Postmenopause/ 12 (post menopaus$ or postmenopaus$ or post-menopaus$).tw. 13 11 or 12 14 10 and 13 15 clinical trial.pt. 16 randomized controlled trial.pt. 17 tu.fs. 18 dt.fs. 19 random$.tw.
IN
TE
Lastly, additional information was sought from authors and industry sponsors.
AL
A hand search of abstracts from Osteoporosis International, Journal of Bone and Mineral Research, Calcied Tissue International and FDA proceedings from the last two years.
SE
METHODS OF THE REVIEW Study selection, quality assessment and data extraction:
5
O N
LY
Grading the strength of the evidence per outcome: We used the ribbon grading system as described in the 2004 Evidence-based Rheumatology BMJ book (Tugwell 2004) to grade the strength of the evidence per outcome. The ribbon grading system uses four categories to rank the evidence from research studies: Platinum, Gold, Silver and Bronze. The ranking is given according to different criteria, including sample size, blinding, handling of withdrawals and concealment allocation (Tugwell 2004). The ranking of the efcacy outcomes (i.e. fractures and BMD) is included in the synopsis, methodological quality of included studies and the clinical relevance tables (see Additional Tables - 01 and 02) of this review.
IN
TE
AL
The bibliographic record (i.e. title, authors, keywords and abstract) retrieved from the search were assessed by two independent reviewers (SOD, AC) for potential eligibility based on the reviews a priori eligibility criteria. Those records deemed potentially eligible, or those in which there was not enough information, underwent a full text review to conrm their inclusion. Methodological quality was assessed by two independent reviewers (SOD, AC) using a validated instrument by Jadad (Jadad 1996). This checklist includes three items pertaining to descriptions of randomization, blinding, and the inclusion of data for dropouts and withdrawals, with a total score of ve. Studies with a score less than or equal to two were considered low quality studies. Allocation concealment was also evaluated by two independent reviewers (SOD, AC) using the allocation component of a validated instrument (Schulz 1995). As outlined in the Cochrane Reviewers Handbook, the allocation concealment was determined to be: A) adequate i.e. central randomization; numbered or coded bottles or containers; drugs prepared by the pharmacy; serially numbered, opaque, sealed envelopes; or other description that contained elements convincing of concealment, B) unclear i.e. authors either did not report an allocation concealment approach at all or reported an approach that was neither adequate nor inadequate, C) inadequate i.e. alternation or reference to case record numbers or to dates of birth, and D) not used. Next, data were independently extracted by both reviewers (SOD, AC) using a data extraction form designed specically for this review. Details of the study population, duration of intervention, baseline demographic data, and the outcomes were collected. Differences with respect to article eligibility, quality assessment and data extraction were resolved by referring to the original publication and establishing consensus (Alderson 2003).
DESCRIPTION OF STUDIES A total of 80 potentially relevant studies were identied from the electronic and hand search strategy outlined above and screened for retrieval. Of these, 62 were excluded as they did not meet the reviews eligibility criteria and 18 underwent a full text review (Meunier 2003; Meunier 2004-3; Meunier 2002; Meunier 2004-1; Reginster 2002-1; Reginster 2005; Boivin 2003; Meunier 20042; Naveau 2004; Pors 2004; Reginster 2002-2; Reginster 20031; Reginster 2003-2; Reginster 2004; Reginster 2004-1; Sorbera 2003; Uebelhart 2003; Marquis 2005). Of these 18 records, a total of 13 were excluded as a result of being either a review publication (Meunier 2004-2; Reginster 2003-1; Meunier 2004-3; Boivin 2003; Naveau 2004; Pors 2004; Reginster 2002-2; Reginster 2003-2; Reginster 2004; Reginster 2004-1; Sorbera 2003; Uebelhart 2003) or a description of the study protocol (Meunier 2003). The remaining ve studies met our eligibility criteria, four of which were primary studies (Meunier 2002; Meunier 2004-1; Reginster 2002-1; Reginster 2005) and one, an abstract, that was a companion paper to the included study by Meuneir et al., 2004 (Marquis 2005). There was no previous systematic review on this topic. All four included primary studies were randomized placebo controlled trials (Meunier 2002; Meunier 2004-1; Reginster 20021; Reginster 2005). Three investigated the efcacy of strontium ranelate in a treatment population (Meunier 2002; Meunier 20041; Reginster 2005) and one included a prevention population (Reginster 2002-1). The mean age of the postmenopausal women studied ranged from 54.2 (Reginster 2002-1) to 76.7 years (Reginster 2005). None of the women had a previous vertebral fracture in one study (Reginster 2002-1), approximately half had a prior vertebral fracture in one (Reginster 2005) and all of the women had a prior vertebral fracture in two (Meunier 2002; Meunier 20046
Data analysis: Prior to the pooling, we developed hypotheses that might account for heterogeneity of study results and compared groups according to: 1) treatment duration, 2) dose and, 3) prevention versus treatment populations. Where possible, the analyses were based on intention-to-treat data from the individual clinical trials. For fractures and safety outcomes, a weighted relative risk was determined for the number of women with either incident fractures or adverse events using Review Manager 4.2.7 (Fleiss
SE
O N
LY
1993). For BMD, a weighted mean difference (WMD) of the percent change between treatment and control groups for different BMD sites including lumbar spine, femoral neck and total hip was calculated. Analyses of the four trials were conducted using an available data set as it was not possible to carry out an intention to treat analysis with the published data. Meta-analysis was conducted according to random effects model. Heterogeneity of the treatment effect was calculated using a chi-square test with n -1 degrees of freedom; where n is the number of studies and the I2 statistic (Fleiss 1993; Higgins 2003). In addition to relative measures, the absolute risk reduction (ARR) was calculated and for those outcomes that were statistically signicant, the number needed to treat (NNT) was determined. The NNT was calculated by taking the inverse of the ARR (NNT = 1/ARR) where ARR is the control event rate minus the treatment event rate. These results are summarized in the clinical relevance tables of this review (see Additional Tables - 01 and 02).
The quality of the included studies was assessed using the Jadad instrument (Jadad 1996). Quality scores, percent lost to follow-up and allocation concealment grades are summarized in the Characteristics of Included Studies Table of this review. All four studies were adequately reported as randomized and described adequate methods regarding the sequence of randomization. All reported that the trial was double blind and of these, two indicated that the recipients of care were unaware of their assigned intervention (Reginster 2002-1; Reginster 2005). One reported that the persons responsible for assessing outcomes were unaware of the assigned intervention (Meunier 2004-1) and one reported that the recipients, those providing the care and persons responsible for assessing outcomes were all unaware of the assigned intervention (Meunier 2002). A description of withdrawals was adequately provided in all trials. All trials had a methodological quality score of greater than three out ve on the Jadad checklist (mean 4.25, range 4-5).
IN
TE
METHODOLOGICAL QUALITY
1). The womens mean BMD T score was < - 2.5 SD in three (Meunier 2002; Meunier 2004-1; Reginster 2005) and > - 2.5 SD in one (Reginster 2002-1). With respect to dosages of strontium ranelate received, the three treatment studies included the recommended daily dose of strontium ranelate (2g) (Meunier 2002; Meunier 2004-1; Reginster 2005) whereas in the prevention trial; the highest daily dose was 1 g (Reginster 2002-1). The compliance rate ranged from 80% (Reginster 2005) to 93% (Meunier 2002). All four studies included a calcium supplement in both treatment and control groups which ranged in dose from 500 mg (Meunier 2002; Reginster 2002-1) to 1000 mg (Meunier 2004-1; Reginster 2005) daily. In three trials, women in the treatment and control groups also received vitamin D supplements which ranged from 400 to 800 IU daily based on serum concentrations of 25-hydroxyvitamin D (Reginster 2005; Meunier 2004-1) or 800 IU daily (Meunier 2002). No other osteoporotic treatments were administered. Three studies assessed vertebral fractures (Meunier 2002; Meunier 2004-1; Reginster 2005) and two included non-vertebral fractures (Meunier 2004-2; Reginster 2005). All four studies measured BMD, three of which assessed BMD of the lumbar spine (Meunier 2002; Reginster 2002-1; Meunier 2004-1), three at the total hip (Reginster 2002-1; Meunier 2004-1; Reginster 2005) and four at the femoral neck (Meunier 2002; Reginster 2002-1; Meunier 2004-1; Reginster 2005). Quality of life was assessed in two trials (Meunier 2004-1; Reginster 2005). Total withdrawals, withdrawals due to adverse events, emergent adverse events, serious adverse events and deaths were reported in all four studies (Meunier 2004-1; Reginster 2002-1; Reginster 2005; Meunier 2002) whereas the number of individuals that developed diarrhea or gastritis was reported in three (Meunier 2002; Meunier 20041; Reginster 2005). Additional details regarding the characteristics of the included studies are presented in the Characteristics of Included Studies Table of this review.
Despite the high quality of the included studies, three had losses to follow-up greater than 20% (Meunier 2002; Meunier 2004-1; Reginster 2005) and only one provided an adequate description of allocation concealment (Meunier 2002). Lastly, the analyses of the four trials were conducted using an available analysis set, which is not preferred but close to the intention to treat principle (Alderson 2003). For the efcacy outcomes of this review (fractures and BMD), a silver level of evidence has been assigned as none of the randomized trials met all of the criteria required for a gold level ranking i.e. sample size of at least 50 in each group, blinding of patients and assessors for outcomes, loss to follow-up < 20% and adequate allocation concealment (Tugwell 2004).
RESULTS FRACTURES:
AL
SE
Vertebral Fractures: Vertebral fractures were determined by the quantitative morphometric assessment method by Genant (Meunier 2002; Meunier 2004-1) and semi-quantative visual assessments (Meunier 20041; Reginster 2005). Patients were not obligated to undergo a vertebral x-ray in one of the phase III trials however, x-rays were obtained for the largest number of patients as possible (total of 3640 patients or 71%) (Reginster 2005). In osteoporotic women, 2 g of strontium ranelate per day demonstrated a 41% relative reduction in radiographic vertebral fractures over a one year period (three trials, n=5254, RR 0.59, 95% CI 0.46 to 0.74) (Meunier 2002; Meunier 2004-1; Reginster 2005) and a 37% relative reduction over a three year period (two trials, n = 5082, RR 0.63, 95% CI 0.56 to 0.71) compared to placebo (Meunier 2004-1; Reginster 2005). The chi-square test for heterogeneity of treatment effect was not signicant in either of these analyses (i.e. p > 0.1). There was only one trial (n=1442) that investigated the effects of 2 g of strontium ranelate versus placebo per day on symptomatic or clinical vertebral fractures in osteoporotic women (Meunier 2004-1). The results from this trial demonstrated a 52% relative reduction in risk of a symptomatic fracture (RR 0.48, 95% CI 0.29 to 0.80) over a one year period and a 38% relative reduction (RR 0.62, 95% CI 0.47 to 0.83) over three years. With respect to the lower doses of strontium ranelate, one trial demonstrated a 31% relative reduction in radiographic vertebral fractures in osteoporotic women using 0.5 g of strontium ranelate versus placebo per day over a two year period however, this was not statistically signicant (RR 0.69, 95% CI 0.47 to 1.01) (Meunier 2004-1). The same study showed a 6% relative reduction in radiographic vertebral fractures using 1.0 g of strontium ranelate daily over the same time frame again, this was not statistically signicant (RR 0.94, 95% CI 0.68 to 1.30) (Meunier 2004-1). Non-vertebral Fractures:
7
O N
LY
Non-vertebral fractures were reported by study investigators based on radiographic evaluation or written documentation provided e.g. radiological report, copy of the hospitalization/emergency department report (Meunier 2004-1, Reginster 2005). In osteoporotic women, 2 g of strontium ranelate per day demonstrated a 14% relative reduction in all non-vertebral fractures (including hip but excluding fractures of the coccyx, skull, jaw, face, ankle, ngers and toes) over a three year period compared to placebo (two trials, n = 6572, RR 0.86, 95% CI 0.75 to 0.98) (Meunier 2004-1; Reginster 2005). However, the upper boundary of the condence interval approximates one. The chi-square test for heterogeneity of treatment effect was not signicant. One study (n = 4932) reported on major osteoporotic non-vertebral fractures dened as fractures of the wrist, pelvis and sacrum, ribs-sternum, clavicle, humerus or hip only in women with osteoporosis and found a 19% relative reduction taking 2 g strontium ranelate daily compared to placebo although the upper boundary of the condence interval approaches one (RR 0.81, 95% CI 0.66 to 0.98) (Reginster 2005). There were no studies that evaluated the effects of lower doses of strontium ranelate on the incidence of non-vertebral fractures. Hip: There was only one trial that assessed the efcacy of 2 g of strontium ranelate per day versus placebo on the relative reduction of hip fractures in women with osteoporosis (Reginster 2005) therefore, we were unable to estimate a pooled relative risk. Hip fractures, similar to other non-vertebral fractures, were determined by a radiological evaluation or by a report from a hospitalization. After three years, the relative risk of a hip fracture was 15% (RR 0.85, 95% CI 0.61 to 1.19) in the total group of women (n=4932) versus 36% for a subgroup at high risk of hip fracture (n=1977) dened by age > 74 years and a femoral neck BMD T-score < - 3 (Looker 1991) (RR 0.64, 95% CI 0.41 to 0.99). In both instances the treatment effect was not statistically signicant. However, this study was not powered for this efcacy outcome.
of 2 g of strontium ranelate daily versus placebo in osteoporotic women on lumbar spine BMD over a three year period and the WMD demonstrated an increase in lumbar spine BMD relative to placebo (WMD adjusted for strontium content 8.09, 95% CI 7.22 to 8.96 and WMD not adjusted for strontium content 14.39, 95% CI 13.40 to 15.38) (Meunier 2004-1). In terms of the lower doses of strontium ranelate, one study (n=63) investigated the effects of 0.125 g of strontium ranelate daily versus placebo on women without osteoporosis and found a nonsignicant increase in lumbar spine BMD over a two year period (WMD not adjusted for strontium content 0.37, 95% CI -1.57 to 2.31) (Reginster 2002-1). Women, with and without osteoporosis, taking 0.5 g of strontium ranelate daily compared to placebo showed a non-signicant increase in lumbar spine BMD over a two year period when BMD was adjusted for strontium content (two trials, n = 232, WMD 1.01, 95% CI -0.63 to 2.66). However, when BMD was not adjusted for strontium content, the increase was signicant (WMD 3.59, 95% CI 1.66 to 5.51) (Meunier 2002; Reginster 2002-1). The chi square test for heterogeneity of treatment effect for the analyses where BMD was not adjusted for strontium content was approaching signicance (p=0.16) and I2 = 49.5%. The inclusion of a treatment (Meunier 2002) versus prevention (Reginster 2002-1) population may explain this nding. Lastly, women with and without osteoporosis, taking 1 g of strontium ranelate daily compared to placebo demonstrated an increase in lumbar spine BMD over a two year period (Meunier 2002; Reginster 2002-1) (two trials, n = 232, WMD adjusted for strontium content 2.14, 95% CI 0.70 to 3.58 and WMD, not adjusted for strontium content 6.68, 5.16 to 8.20). The chi square test for heterogeneity of treatment effect was not signicant. Femoral Neck BMD: The effects of taking 2 g of strontium ranelate daily versus placebo in osteoporotic women demonstrated an increase in femoral neck BMD over two year (two trials, n=1614, WMD 5.73, 95% CI 5.15 to 6.32) (Meunier 2002; Meunier 2004-1) and three year period (two trials, n=4230, WMD 8.25%, 95% CI 7.84 to 8.66) (Meunier 2004-1; Reginster 2005). The chi-square test for heterogeneity of the treatment effect was not signicant for both of these analyses. With respect to the lower doses of strontium ranelate, one trial (n = 63) explored the effects of 0.125 g/day in women without osteoporosis and found a non signicant increase in femoral neck BMD in favour of those receiving the placebo over a two year period (WMD -1.47, 95% CI -3.68 to 0.74) (Reginster 20021). The effects of 0.5 g of strontium ranelate daily versus placebo demonstrated a non-signicant increase in femoral neck BMD (two trials, n = 232, WMD 1.00, 95% CI -0.52 to 2.52) whereas 1 g of strontium ranelate daily versus placebo showed a signicant increase (two trials, n = 233, WMD 2.52, 95% CI 0.96 to 4.09) over a two year period (Meunier 2002; Reginster 2002-1). The chi
8
BMD:
Lumbar Spine BMD: Using 2 g of strontium ranelate daily versus placebo, an increase in lumbar spine BMD was demonstrated in osteoporotic women over a two year period (two trials, n = 1614, WMD adjusted for strontium content 5.44, 95% CI 3.41 to 7.46 and WMD, not adjusted for strontium content 11.29, 95% CI 10.22 to 12.37) (Meunier 2002; Meunier 2004-1). However, the chi-square test for heterogeneity of treatment effect was signicant for the analysis involving the results adjusted for strontium content (p=0.04) and I2 = 76.6%. We investigated sources of clinical heterogeneity and a possible explanation relates to the difference in timing and methods of the strontium content calculation from bone-biopsy samples between the two trials (Meunier 2004-1; Meunier 2002). There was only one trial (n=1442) that investigated the effects
IN
TE
AL
SE
O N
LY
square test for heterogeneity of treatment effect was not signicant for either of these analyses. Total Hip BMD: One study (n=1442) demonstrated an increase in total hip BMD in osteoporotic women on 2 g of strontium ranelate daily compared to placebo over a two year period (WMD 1.16, 95% CI 1.051.27) (Meunier 2004-1) and two trials (n=4230) demonstrated an increase in total hip BMD over a three year period (WMD 9.83, 95% CI 9.39 to 10.26). The chi-square test for heterogeneity of treatment effect for the latter analyses was not signicant. Only one trial investigated the effects of lower doses of strontium ranelate on total hip BMD and found a non-signicant increase with 0.125 g of strontium ranelate daily (n= 63, WMD 0.67, 95% CI -1.18 to 2.52) and signicant increases with 0.5 g/day (n=65, WMD 2.02, 95% CI 0.47 to 3.57) and 1 g/day (n=60, WMD 4.09, 95% CI 2.09 to 6.09) over a two year period (Reginster 2002-1). Health Related Quality of Life: Quality of life was assessed using the SF-36 questionnaire in two trials (Meunier 2004-1; Reginster 2005) as well as the Quality of Life questionnaire in Osteoporosis (QUALIOST) in one (Meunier 2004-1). The QUALIOST is a 23 item, disease specic (vertebral osteoporosis) questionnaire, with a global score and two subscores: physical and emotional. In both trials, women completed the quality of life assessments at baseline and every six months throughout the duration of the trial (Meunier 2003). The results are not within the published literature however, unpublished data demonstrated that 2 g of strontium ranelate daily compared to placebo has a benecial effect on quality of life as dened by the QUALIOST in a subset of osteoporotic women (n=1240) after three years of treatment (global score p = 0.016, emotional and physical scores p = 0.019 and 0.032 respectively) (Marquis 2005). Furthermore, results from a back pain assessment included in the QUALIOST questionnaire conducted every six months, revealed that the occurrence of back pain was signicantly reduced by 29% in the strontium ranelate group as compared to the placebo group over three years with a signicant effect in the rst year (p = 0.006) (Marquis 2005).
ference in the risk of withdrawal (RR 0.87, 95% CI 0.36 to 2.11) (Meunier 2002; Reginster 2002-1). The chi square test was approaching signicance (p=0.11) and I2 = 60.7% which may be explained by the inclusion of a treatment (Meunier 2002) versus prevention (Reginster 2002-1) population. Withdrawals due to adverse events: A total of three trials (n = 6847) reported the safety of using the recommended daily dose of 2 g strontium ranelate versus placebo through withdrawals due to adverse events (Meunier 2002; Meunier 2004-1; Reginster 2005). The pooled estimate of the relative risk was 1.20 (95% CI 0.96 to 1.50) with 22% of the strontium ranelate treated patients versus 19.1% of the controls having withdrawn due to an adverse event however, this nding was not significant (p=0.12). The chi-square test for heterogeneity of the treatment effect was borderline signicant (p=0.10) and I2 =57.0%. This may be attributed to differences in the study populations baseline characteristics including age and frailty (i.e. fracture prevalence). Similarly, for those women taking 0.5 g/day, there was no signicant difference in the risk of withdrawals due to adverse events (two trials, n=256, RR 1.10, 95% CI 0.56 to 1.80) (Meunier 2002; Reginster 2002-1). The chi square test for heterogeneity was not signicant. Number of emergent adverse events: A total of three trials (n = 6847) using the recommended daily dose of 2 g of strontium ranelate versus placebo daily did not nd a signicant difference in the number of emergent adverse events (RR 0.99, 95% CI 0.98 to 1.01). (Meunier 2004-1; Meunier 2004-1; Reginster 2005). Similarly, two trials (n=256) using 0.5 g of strontium ranelate versus placebo daily did not nd a signicant difference in the risk of developing an adverse event (RR 0.93, 95% CI 0.86 to 1.00) (Meunier 2002; Reginster 2002-1). The chi-square test for heterogeneity of the treatment effect was not signicant for either of these analyses. Serious adverse events: A total of three trials (n = 6847) reported the number of participants that developed a serious adverse event using 2 g of strontium ranelate daily versus placebo (Meunier 2002; Meunier 20041; Reginster 2005). Serious adverse events occurred in 24.09% of the strontium ranelate treated patients versus 23.97% of the controls. The pooled estimate of the relative risk was 1.01 (95% CI 0.92 to 1.09) demonstrating a non signicant difference between the two groups. Similarly, two trials (n=256) reported the number of serious adverse events using 0.5 g of strontium ranelate versus placebo daily and found no signicant difference in the relative risk of developing a serious adverse event (RR 0.81, 95% CI 0.44 to 1.48) (Meunier 2002; Reginster 2002-1). The chi-square test for heterogeneity of the treatment effect was not signicant for either of these analyses. Diarrhea: A total of three trials (n = 6847) reported the number of participants that developed diarrhea using the recommended dose of
9
ADVERSE EVENTS: All adverse event data was reported by dose regardless of study duration i.e. two years (Meunier 2002; Reginster 2002-1) and three years (Meunier 2004-1; Reginster 2005). Total withdrawals: A total of three trials (n = 6847) using the recommended dose of 2 g strontium ranelate versus placebo daily did not nd a signicant difference in the risk of withdrawals (RR 0.98, 95% CI 0.91 to 1.05) (Meunier 2002; Meunier 2004-1; Reginster 2005). The chi-square test for heterogeneity of the treatment effect was not signicant. Similarly, two trials (n=256) using 0.5 g of strontium ranelate versus placebo daily did not demonstrate a signicant dif-
IN
TE
AL
SE
O N
LY
strontium ranelate versus placebo daily (Meunier 2004-1; Meunier 2002; Reginster 2005). Diarrhea occurred in 6.5% of the strontium ranelate treated patients versus 4.7% in the controls. The pooled estimate of the relative risk was 1.38 (95% CI 1.02 to 1.87) with an overall signicant effect (p = 0.04). The chi-square test for heterogeneity of the treatment effect was not signicant. Gastritis: A total of three trials (n = 6847) reported the number of participants that developed gastritis using 2 g of strontium ranelate versus placebo daily (Meunier 2002; Meunier 2004-1; Reginster 2005). Gastritis occurred in 2.7% of the strontium ranelate treated patients and 3.4% of the controls. The pooled estimate of the relative risk was 0.81 (95% CI 0.56 to 1.17). The chi-square test for heterogeneity of the treatment effect was not signicant. Deaths: A total of three trials (n = 6847) reported the total number of deaths using 2 g of strontium ranelate versus placebo daily (Meunier 2002; Meunier 2004-1; Reginster 2005). A total of 4.97% of the strontium ranelate treated patients versus 5.37% of the controls died during the follow-up period. The pooled estimate of the relative risk was 0.99 (95% CI 0.64 to 1.53). The chi-square test for heterogeneity of the treatment effect was borderline signicant with p= 0.17 and I2 = 42.8%. This may be attributed to differences in the mean age of the study population in addition to their frailty. Additional data obtained suggests that the risk of vascular system disorders including venous thromboembolism (two trials, n = 6669, 2.2% versus 1.5%, OR 1.5, 95% CI 1.1 to 2.1) and pulmonary embolism (two trials, n = 6669, 0.8% versus 0.4%, OR 1.7, 95% CI 1.0 to 3.1) as well as nervous system disorders such as headaches (3.9% versus 2.9%), seizures (0.3% versus 0.1%), memory loss (2.4% versus 1.9%) and disturbance in consciousness (2.5% versus 2.0%) is slightly increased with taking 2 g of strontium ranelate daily over a 3 to 4 year period.
there was little heterogeneity of treatment effect however, the upper boundary of the condence interval of the non-vertebral fractures approaches one indicating that the data may be consistent with a null effect. Furthermore, the impact that 2 g of strontium ranelate daily has on reducing the risk of a hip fracture remains unclear as the only included study with data was not powered for this outcome. Although data from direct comparisons with other antiosteoporotic treatments are lacking, the reduction in the relative risk of vertebral fracture seems similar to the other available therapies which have been shown to reduce the relative risk of recurrent fracture by 40 to 60% (Cranney 2002). The greater reduction in risk of vertebral fractures compared to non-vertebral fractures may be explained by the greater effect that strontium ranelate has on the vertebral versus non-vertebral bone mineral density. The results of this review support the use of 2 g of strontium ranelate daily in osteoporotic postmenopausal women for increasing BMD at all sites over a two to three year period. In a prevention population, 1 g of strontium ranelate daily was demonstrated to increase BMD at all sites compared to placebo over a two year period. In both the treatment and prevention populations, lower doses of strontium ranelate were superior to placebo with the highest dose of strontium ranelate demonstrating the greatest increase in BMD over a two year period. While the increase in BMD in patients taking 2 g of strontium ranelate daily is impressive, caution is necessary when in interpreting these results. As previously mentioned, the combined effect of strontium distribution in bone and increased x-ray absorption of strontium compared to calcium leads to an amplication of BMD measurement by DXA (Ortolani 2006). While the included trials described and performed a correction to adjust for the strontium content in bone for the lumbar spine BMD measurements, there is considerable uncertainty about the accuracy of the results which arises from the small number of participants in whom iliac crest bone biopsy was performed and the reliance on animal data for the correction factor for inferring bone strontium content in the spine (Blake 2005). Although limited, there is evidence from one phase III trial to suggest that 2 g of strontium ranelate daily compared to placebo has a benecial effect on health related quality of life in a subset of postmenopausal women (n=1240) after three years of treatment. In keeping with this, results from a back pain assessment included in the quality of life questionnaire revealed that the occurrence of back pain was signicantly reduced by 29% in the strontium ranelate group as compared to the placebo group over three years with a signicant effect in the rst year (p = 0.006) (Marquis 2005). These ndings are presumably due to a reduction in the consequences related to osteoporosis such as vertebral fractures. Overall incidence rates for adverse events with strontium ranelate did not differ from placebo regardless of dose. There was a statistically signicant increase in the risk for diarrhea in patients treated with 2 g of strontium ranelate daily relative to placebo. However; there was no signicant difference in the number of withdrawals
10
DISCUSSION
A total of four trials met our inclusion criteria, three of which investigated the effects of strontium ranelate compared to placebo in a treatment population (doses ranged from 0.5 to 2 g daily) and one, in a prevention population (doses 0.125, 0.5 and 1 g daily). The included studies were presumably conducted in calcium and vitamin D replete postmenopausal women. There is silver level evidence to support the use of 2 g of strontium ranelate daily in osteoporotic postmenopausal women to reduce the risk of vertebral and to a lesser extent, non-vertebral fractures. The pooled estimate of the relative risk for vertebral and non-vertebral fractures over a three year follow-up period were consistent with a reduction of 37% for vertebral fractures and 14% for nonvertebral fractures. Both estimates were statistically signicant and
IN
TE
AL
SE
O N
LY
due to side effects, number of emergent events, serious adverse events, gastritis or deaths regardless of the dose analyzed. Additional data obtained from the scientic report by the European Agency for the Evaluation of Medicinal Products (EMEA 2004), in addition to the industry sponsor (Servier), has illustrated some concern of a slight increased risk in vascular and neurological disorders as well as abnormal laboratory ndings. This information is based on results from the two phase III trials which focused on the recommended dose of 2 g of strontium ranelate daily (Meunier 2004-1; Reginster 2005) and has been summarized below and within the appended table entitled Other adverse events from additional sources (see Additional Tables -02). Disorders of the vascular system were present in 26.3% of the patients in the strontium ranelate group versus 24.4% in the placebo; Estimated difference = 1.9% (95% CI -0.2 to 4.0) with an increased reporting rate of adverse events of venous thromboembolism (2.2% versus 1.5%, OR 1.5, 95% CI 1.1 to 2.1) and pulmonary embolism (0.8% versus 4.5%, OR 1.7, 95% CI 1.0 to 3.1) over a three year period. Furthermore, the absolute number of patients that suffered a pulmonary embolism resulting in discontinuation of treatment or death was slightly increased in the strontium ranelate group compared to controls (i.e. 0.2% versus 0.1%). The cause of this increased risk of vascular disorders is not understood. Nervous system disorders were found in 20.9% of the patients in the strontium ranelate group versus 18.9% in the placebo; Estimated difference = 2.0 % (95% CI 0.1 to 3.9) with an increased reporting of seizures (0.3% versus 0.1%), memory loss (2.4% versus 1.9%) and disturbances in consciousness (2.5% versus 2.0%) over a four year period. Again, the etiology of the increased risk in neurological disorders is not clear. Lastly, mean baseline serum creatine kinase levels increased in both groups however, this increase was signicantly greater in the strontium ranelate group (31.3 + 80.8 IU/L) compared to the controls (13.1 + 46.6 IU/L); Estimated difference = 18.2 IU/L (95% CI 14.8 to 21.6). The serum creatine kinase levels was greater than the upper limit of normal on at least one occasion in 29.4% (789/2680) of the women in the strontium ranelate group versus 17.6% (475/2705) of the controls (RR 1.68, 95% CI 1.52 to 1.85) providing evidence of strontium ranelate impacting skeletal muscle cell integrity. In light of these ndings targetted surveillance will be needed (EMEA 2004).
lished studies resulting in pooling of proportions and limiting our ability to adjust for differences in patient populations. This review will be updated every two years (or earlier) depending on the emergence of new evidence. Review updates will entail repeating, at periodic intervals, the steps involved in the original review. If new evidence addresses important variables that were not included in the original review we will consider including them. In such instances, we will recheck whether any of their earlier identied studies had such information that was overlooked. Furthermore, should we decide to include a new analysis strategy in our updated review we understand that any new analysis strategies represents a substantive change to the review requiring editorial critique through the Cochrane Collaborations established editorial process.
AUTHORS CONCLUSIONS Implications for practice
IN
There is silver level evidence to support the usefulness of strontium ranelate in reducing fractures in postmenopausal osteoporotic women and increasing BMD in women with/without osteoporosis. Pooled estimates using 2 g of strontium ranelate daily compared to placebo in osteoporotic women over a three year period are consistent with a reduction in vertebral fractures (37%); however, there is less of a reduction in non-vertebral fractures (14%) and the effect on hip fractures remains unclear. Strontium ranelate increased BMD at all sites in both treatment and prevention populations and while lower doses of strontium ranelate were superior to placebo, the highest dose demonstrated the greatest increase. There is some evidence to suggest that 2 g of strontium ranelate daily compared to placebo may have a benecial effect on health related quality of life in postmenopausal women after three years of treatment. Diarrhea may occur however, adverse events leading to study withdrawal were not signicantly increased in the strontium ranelate group. Potential risks to the vascular and neurological system associated with taking 2 g of strontium ranelate daily need to be further explored and quantied. Implications for research Further monitoring of the quality, effectiveness and safety of strontium ranelate is essential especially in the prevention of osteoporosis. Additional research is required to conrm its mechanism of action. Long term fracture data are needed to conrm the effect that strontium ranelate has on bone health in both prevention and treatment populations. Long term safety data is required with particular attention to be paid to bone mineralization, continued fracture efcacy, neurological and vascular system disorders, specically venous and pulmonary thromboembolism and skeletal muscle integrity. Future trials to evaluate the impact of strontium ranelate treatment on BMD, including the effect on the elimination of bone strontium in patients switching to other
11
Our systematic review has several limitations. Firstly, while the methodological quality of the included studies was high based on the Jadad instrument, three of the four studies had losses to followup greater than 20% (Meunier 2002; Meunier 2004-1; Reginster 2005) and three had unclear descriptions of allocation concealment (Meunier 2004-1; Meunier 2004-1; Reginster 2005). Losses to follow-up can threaten the validity of the trial since the event rate may be very different in those lost to follow-up versus those who completed the trial and failure to conceal the participants treatment allocation could also bias the treatment effect in either direction. Secondly, access to aggregate data only within the pub-
TE
AL
SE
O N
LY
anti-resorptive treatments, are needed. In addition, comparative trials evaluating the efcacy of strontium ranelate relative to other osteoporosis therapies such as bisphosphonates and intermittent recombinant human parathyroid hormone are required.
POTENTIAL CONFLICT OF INTEREST None known.
ACKNOWLEDGEMENTS The authors would like to thank Louise Falzon of the Cochrane Musculoskeletal Group for her assistance in developing the search strategy and Patricia Chatelain from Servier who provided additional data not found within the published literature.
SOURCES OF SUPPORT External sources of support Canadian Institutes for Health Research CANADA Internal sources of support Ottawa Health Research Institute CANADA
Meunier 2004-1 {published data only} Meunier PJ, Roux C, Seeman E, Ortolani S, Badurski JE, Spector TD, Cannata J, Balogh A, Lemmel E, Pors-Nielson S, Rizzoli R, Genant HK, Reginster J. The effects of strontium ranelate on the risk of vertebral fracture in women with postmenopausal osteoporosis. New england journal of medicine 2004;350(5):459468. Reginster 2002-1 {published data only} Reginster JY, Deroisy R, Dougados M, Jupsin I, Colette J, Roux C. Prevention of early postmenopausal bone loss by strontium ranelate: The randomized, two-year, double-masked, dose-ranging, placebocontrolled PREVOS study. Osteoporosis international 2002;13:925 931. Reginster 2005 {published data only} Reginster JY, Seeman E, de Vernejoul MC, Adami S, Compston J,
IN
Meunier 2002 {published data only} Meunier PJ, Slosman DO, Delmas PD, Sebert JL, Brandi ML, Albanese C, Lorenc R, Pors-Nielsen S, de Vernejoul MC, Roces A, Reginster JY. Strontium ranelate: Dose-dependent effects in established postmenopausal vertebral osteoporosis - a 2-year randomized placebo controlled trial. Journal of clinical endocrinology and metabolism 2002; 87(5):20602066.
TE
AL
REFERENCES
Phenekos C, Devogelaer JP, Diaz Curiel M, Sawicki A, Goemaere S, Sorensen OH, Felsenberg D, Meunier PJ. Strontium ranelate reduces the risk of nonvertebral fractures in postmenopausal women with osteoporosis: TROPOS study. Journal of clinical endocrinology and metabolism 2005;90(5):28162822.
U
Boivin 2003 Boivin G. Bone mineralization and mineral status. Therapie 2003; 58(5):409413. Meunier 2003 Meunier PJ, Reginster JY. Design and methodology of the phase 3 trials for the clinical development of strontium ranelate in the treatment of women with postmenopausal osteoporosis. Osteoporosis international 2003;14 Suppl(3):S66S76. Meunier 2004-2 Meunier PJ, Roux C, Seeman E, Pagalilauan G, Laya M. Strontium ranelate prevented vertebral fractures in postmenopausal women with osteoporosis. Evidence-based obstetrics and gynecology 2004;6(4):216 217. Meunier 2004-3 Meunier PJ, Roux C, Seeman E, Cheung A. Strontium ranelate reduced the risk of vertebral fractures in postmenopausal women os12
SE
O N
LY
teoporosis. Evidence-based medicine 2004;9(5):149. Naveau 2004 Naveau B. Strontium: a new treatment for osteoporosis. Joint bone spine 2004;71(4):261263. Pors 2004 Pors NS. The biological role of strontium. Bone 2004;35(3):583 588. Reginster 2002-2 Reginster JY. Strontium ranelate in osteoporosis. Current pharmaceutical design 2002;8(21):19071916. Reginster 2003-1 Reginster JY, Meunier PJ. Strontium ranelate phase 2 dose-ranging studies: PREVOS and STRATOS studies. Osteoporosis international 2003;14(Suppl 3):S56S65. Reginster 2003-2 Reginster JY, Deroisy R, Jupsin I. Strontium ranelate: a new paradigm in the treatment of osteoporosis. Drugs today (Barc ) 2003;39(2):89 101. Reginster 2004 Reginster JY, Devogelaer JP. Treatment of postmenopausal osteoporosis in 2004. Revue medicale de liege 2004;59(11):633647. Reginster 2004-1 Reginster JY, Lecart MP, Deroisy R, Lousberg C. Strontium ranelate: a new paradigm in the treatment of osteoporosis. Expert opinion on investigational drugs 2004;13(7):857864. Sorbera 2003 Sorbera LA, Castaner J, Leeson PA, Bayes M. Strontium ranelate: Treatment and prevention of osteoporosis bone resorption inhibitor bone formation stimulant. Drugs of the future 2003;28(4):328335.
in alveolar bone in monkeys (Macaca fascicularis). Bone 2001;29(2): 176179. Burge 2003 Burge RT, King AB, Balda E, Worley D. Methodology for estimating current and future burden of osteoporosis in state populations: application to Florida in 2000 through 2025. Value health 2003;6: 574583. Canalis 1996 Canalis E, Hott M, Deloffre P, Tsouderos Y, Marie PJ. The divalent strontium salt S12911 enhances bone cell replication and bone formation in vitro. Bone 1996;18(6):517523. Cauley 2000 Cauley JA, Thompson DE, Ensrud KC, Scott JC, Black D. Risk of mortality following clinical fractures. Osteoporosis international 2000; 11(7):556561. Cranney 2002 Cranney A, Guyatt G, Grifth L, Wells G, Tugwell P, Rosen C. Metaanalyses of therapies for postmenopausal osteoporosis. IX: Summary of meta-analyses of therapies for postmenopausal osteoporosis. Endocrine reviews 2002;23(4):570578. Cummings 2002 Cummings SR. Epidemiology and outcomes of osteoporotic fractures. Lancet 2002;359:17611767. EMEA 2004 EMEA. PROTELOS. The committee for medicinal products for human use 2004. Fleiss 1993 Fleiss, JL. Statistical basis of meta-analysis. Statistical methods in medical research 1993;2(2):121145. Goeree 1996 Goeree R, OBrien B, Petitt D, Cuddy L, Ferraz M, Adachi JD. An Assessment of the burden of illness due to osteoporosis in Canada. Journal of the society of obstetricians and gynecologists of Canada 1996; (Supplement):1524. GRADE Working Group GRADE Working Group. Grading quality of evidence and strength of recommendations. British medical journal 2005;328. Higgins 2003 Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. British medical journal 2003;327:557 560. Jackson 2000 Jackson SA, Tenenhouse A, Robertson L, the CaMos Study Group. Vertebral fracture denition from population-based data: Preliminary results from the Canadian Multicenter Osteoporosis Study (CaMos). Osteoporosis international 2000;11:680687. Jadad 1996 Jadad A, Moore RA, Carroll D, Jenkinson D, Reynolds DJ, Gavaghan DJ, McQuay HJ. Assessing the quality of reports of randomized clinical trials: is blinding necessary?. Controlled clinical trials 1996; 17:112. Jagal 1996 Jagal S, Sherry P, Schatzker J. The impact and consequences of hip fracture in Ontario. Canadian journal of surgery 1996;39:105111.
13
Alderson 2003 Alderson P, Green S, Higgins JPT. Cochrane Reviewers Handbook 4.2.2. http://www.cochrane.org/resources/handbook/hbook.htm 2003. Blake 2005 Blake GM, Fogelman I. Long-term effect of strontium ranelate treatment on BMD. Journal of bone and mineral research 2005;20(11): 19011904. Brown 2002 Brown JP, Josse RG. 2002 Clinical practice guidelines for the diagnosis and management of osteoporosis in Canada. Canadian medical association journal 2002;167(10 Suppl):S134. Buehler 2001 Buehler J, Chappuis P, Saffar JL, Tsouderos Y, Vignery A. Strontium ranelate inhibits bone resorption while maintaining bone formation
IN
TE
1993 Consensus Consensus Development Conference: Diagnosis, Prophylaxis, and Treatment of Osteoporosis. American journal of medicine. 1993; Vol. 94, issue 3:561.
Uebelhart 2003 Uebelhart D, Frey D, Frey-Rindova P, Goerres G, Michel BA. Therapy of osteoporosis: bisphosphonates, SERMs, teriparatide and strontium. Zeitschrift fr rheumatologie 2003;62(6):512517.
AL
SE
O N
LY
Jaglal 1998 Jaglal S. Incidence and impact. In: BadleyE, WilliamsJI editor(s). Patterns of health care in Ontario: Arthritis and related conditions. North York: Institute for Clinical Evaluative Sciences, 1998:Chapter 8. Johnell 2004 Johnell O, Kanis JA. An estimate of the worldwide prevalence, mortality and disability associated with hip fracture. Osteoporosis international 2004;15(11):897902. Kanis 2003 Kanis JA, Oden A, Johnell O, De Laet C, Jonsson B, Oglesby AK. The components of excess mortality after hip fracture. Bone 2003; 32:468473. Kanis 2004 Kanis JA, Johnell O. Requirements for DXA for the management of osteoporosis in Europe. Osteoporosis international 2004. Looker 1991 Looker AC, Wahner HW, Dunn WL, Calvo MS, Harris TB, Heyse SP, Johnston CC Jr, Lindsay RL. Proximal femur bone mineral levels of U.S. adults. Osteoporosis international 1995;5:389409. Marie 1985 Marie PJ, Garba MT, Hott M, Miravet L. Effect of low doses of stable strontium on bone metabolism in rats. Mineral electrolyte and metabolism 1985;11(1):513.
Ortolani 2006 Ortolani S, Vai S. Strontium ranelate: An increased bone quality leading to vertebral antifracture efcacy at all stages. Bone 2006;38: S19S22. Ray 1997 Ray NF, Chan JK, Thamer M, Melton LJ III. Medical expenditures for the treatment of osteoporosis fractures in the United States in 1995: report from the National Osteoporosis Foundation. Journal of bone and mineral research 1997;12:2435. Riggs 2005 Riggs BL, Partt AM. Drugs used to treat osteoporosis: The critical need for a uniform nomenclature based on their action on bone remodeling. Journal of bone and mineral research 2005;20(2):177 184. Robinson 2002 Robinson KA, Dickersin K. Development of a highly sensitive search strategy for the retrieval of reports of controlled trials using PubMed. International journal of epidemiology 2002;31(1):150153.
Neer 2001 Neer RM, Arnaud CD, Zanchetta JR, Prince R, Gaich GA, Reginster JY, et al.Effect of parathyroid hormone (1-34) on fractures and bone mineral density in postmenopausal women with osteoporosis. New england journal of medicine 2001;344(19):14341441. NIH 2001 NIH consensus development panel on osteoporosis prevention and diagnosis. Osteoporosis prevention, diagnosis, prophylaxis and treatment. Journal of american medical association 2001;285:785795.
IN
Melton 1992 Melton LJ, III, Chrischilles EA, Cooper C, Lane AW, Riggs BL. Perspective. How many women have osteoporosis?. Journal of bone and mineral research 1992;7(9):10051010.
TE
Marshall 1996 Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurence of osteoporotic fractures. British medical journal 1996;312:12541259.
Marquis 2005 Marquis P, De La Loge C, Diaz-Curiel M, Spector T, Meunier PJ. Benecial effects of strontium ranelate on the quality of life in patients with vertebral osteoporosis (SOTI Study). Osteoporosis international 2005;16(Suppl 3):S54 (P223).
AL
Marie 2001 Marie PJ, Ammann P, Boivin G, Rey C. Mechanisms of action and therapeutic potential of strontium in bone. Calced tissue international 2001;69(3):121129.
Seeley 1991 Seeley DG, Browner WS, Nevitt MC, Genant HK, Scott JC, Cummings SR. Which fractures are associated with low appendicular bone mass in elderly women? The Study of Osteoporotic Fractures Research Group. Annals of internal medicine 1991;115(11):837842.
SE
Seeman 2002 Seeman E. Pathogenesis of bone fragility in women and men. Lancet 2002;359(9320):18411850. Slosman 1999 Slosman D, Provvedini DM, Meunier PJ, Delmas PD, Sebert JL, De Vernejoul MC, et al.The use of different dual X-ray absorptiometry brands in a multicenter clinical trial. Journal of clinical densitometry 1999;2:3744. Tosteson 2001 Tosteson AN, Gabriel SE, Grove MR, Moncur MM, Kneeland TS, Melton LJ, III. Impact of hip and vertebral fractures on qualityadjusted life years. Osteoporosis international 2001;12:10421049. Tugwell 2004 Tugwell P, Shea B, Boers M, Brooks P, Simon LS, Strand V, et al.Evidence-based rheumatology. BMJ Publishing Group, 2004. US Dept Health 2004 U.S. Department of Health and Human Service. Bone health and osteoporosis: a report of the surgeon general. Ofce of the Surgeon General, Rockville, and MD 2004:1436. WHO 1994 WHO Study Group. Assessment of fracture risk and its application to screening for postmenopausal osteoporosis. Report of a WHO study group. World Health Organization technical report 1994; Vol. 843:1129.
O N
Schulz 1995 Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. Journal of american medical association 1995;273(5):408412.
LY
14
TABLES
Characteristics of included studies

Study Methods Meunier 2002 Randomized placebo controlled trial Duration = 2 years N = 353 Postmenopausal women (mean age 66.2, mean (SD) Lumbar spine (LS) T-score by group: -3.80 (0.94) to -.3.97 (0.95), previous vertebral fracture 100%) Treatment Primary outcome: LS bone mineral density (BMD) Strontium ranelate 0.5 g OR 1 g OR 2 g VERSUS placebo daily (calcium supplement 500 mg daily and vitamin D 800 IU daily) BMD: Lumbar spine and femoral neck Fractures: vertebral (deformities) Lost to follow-up: 81/353 (22.9%) Quality Score: 5/5 A Adequate Meunier 2004-1
Participants
Interventions Outcomes Notes Allocation concealment Study Methods
Participants
Interventions
Outcomes Notes
Allocation concealment Study Methods
Participants
IN
TE
Strontium ranelate 2g VERSUS placebo daily (calcium supplement up to 1000 mg daily based upon dietary intake and vitamin D 400-800 IU daily based on serum concentration of 25-hydroxyvitamin D) BMD: Lumbar spine, femoral neck and total hip Fracture: Vertebral and non-vertebral Lost to follow-up: 389/1649 (23.6%) Quality score: 4/5
B Unclear Reginster 2002-1 Randomized placebo controlled trial Duration = 2 years N = 160 Postmenopausal women (mean age 54.2, mean LS T-score by group: -1.3 to -1.5, previous vertebral fracture 0%) Prevention Primary outcome: LS BMD
15
Randomized placebo controlled trial Duration = 3 years N = 1649 Postmenopausal women (mean age 69.3, mean (SD) LS T-score by group: -3.5 (1.3) to -3.6 (1.2), previous vertebral fracture 100%) Treatment Primary outcome: Vertebral fractures
AL
SE
O N
LY
Interventions Outcomes Notes Allocation concealment Study Methods
Strontium ranelate 125 mg OR 500 mg OR 1 g VERSUS placebo daily (calcium supplement 500 mg daily as calcium carbonate) BMD: Lumbar spine, femoral neck and total hip Lost to follow-up: 17/160 (10.6%) Quality score: 4/5 B Unclear Reginster 2005 Randomized placebo controlled trial Duration = 5 years (main statistical analysis over 3 years) N = 5091 Postmenopausal women (mean age 76.7, mean (SD) LS T-score -2.83 (1.63) to -3.24 (1.53), previous vertebral and non-vertebral fracture 55.4 VERSUS 54.2%) Treatment Primary outcome: Non-vertebral fractures Strontium ranelate 2 g VERSUS placebo daily (calcium supplement up to 1000 mg daily based upon dietary intake and vitamin D 400-800 IU daily based on serum concentration of 25-hydroxyvitamin D)
Participants
Interventions
Outcomes
Notes
Lost to follow-up: 1771/5091 (34.8%) Quality Score: 4/5 B Unclear
Study Boivin 2003 Meunier 2003 Meunier 2004-2 Meunier 2004-3 Naveau 2004 Pors 2004 Reginster 2002-2 Reginster 2003-1 Reginster 2003-2 Reginster 2004 Reginster 2004-1 Sorbera 2003 Uebelhart 2003
Reason for exclusion
IN
TE
Review publication
Review of study design Review publication Review publication
Review publication Review publication Review publication Review publication Review publication Review publication
Review publication
Review publication Review publication

16
Characteristics of excluded studies
AL
Allocation concealment
SE
BMD: Femoral neck and total hip Fracture: Vertebral, non-vertebral, major non-vertebral (hip, wrist, pelvis and sacrum, ribs and sternum, clavicle, humerus) and hip
O N
LY
Characteristics of excluded studies (Continued )

ADDITIONAL TABLES
Table 01. Clinical relevance table strontium ranelate 2 g per day: Fractures & safety data
Ctrl grp: # of women 7/100 Rel % Improvement 41%
Outcome Vertebral fracture (1 year) Vertebral fracture (3 years) Nonvertebral fracture (3 years) Withdrawals due to adverse events Serious adverse events Emergent adverse events Gastritis
ER in Sr grp (%) 117/2621 (4.5%)
ER in Ctrl RR (95% grp (%) CI) 206/2633 (7.8%) 0.59 (0.46, 0.74) 0.63 (0.56, 0.71) 0.86 (0.75, 0.98)
ARD (95% CI) -4% (-7 , 1)
NNT/NNH Sr grp: # (95% CI) of women 25 (14, 1) 4/100
Quality of evidence Silver
341/2536 (13.4%)
543/2546 (21.3%)
-9% (-13, -4)
11 (8, 25)
13/100
21/100
37%
Silver
AL
828/3437 (24.1%)
816/3404 (24.0%)
1.01 (0.92, 1.09)
0% (-2, 2)
SE
762/3439 (22.1%)
650/3408 (19.1%)
1.20 (0.96, 1.50)
3% (1, 6)
O N
6/100 4/100 #= Number
345/3305 (10.4%)
398/3267 (12.5%)
-2% (-3, 0)
50 (33, 0)
11/100
LY
13/100
14%
Silver
Silver
Silver
3028/3439 3019/3408 0.99 (88.0%) (88.6%) (0.98, 1.01)
-1% (-2, 1)
Silver
IN
TE
93/3439 (2.7%)
115/3408 (3.4%)
0.81 (0.56, 1.17) 1.38 (1.02, 1.87) RR = Relative risk
-1% (-2, 1)
Silver
Diarrhea
222/3439 (6.5%)
160/3408 (4.7%)
2% (0, 3)
50 (0, 33)
-38%
Silver
Legend:
ER = Event rate
Ctrl = Controls
ARD = Absolute risk difference
NNT = Number needed to treat NNH = Number needed to harm
Rel = Relative
Sr = Strontium ranelate
CI = Condence interval
17
Table 02. Clinical relevance table strontium ranelate 2 g per day: BMD data
# in Sr: Ctrl group 804:810 Quality of Evidence Silver
Outcome (scale) Lumbar spine BMD - not adjusted (2 years) Lumbar spine BMD - adjusted (2 years) Femoral neck BMD (2 years) Femoral neck BMD (3 years) Total hip BMD (3 years) most representative study: Meunier, 20041*; Reginster, 2005**
Mean (SD) Ctrl -1.02 (6.10)*
WMD (95% CI) Abs change 11.29 (10.22, 12.37) 5.44 (3.41, 7.46) 11.29
Rel change -11.1%
804:810
-1.18 (6.26)*
5.44
-4.6%
Silver
804:810 2112:2118 2112:2118 # = Number
-2.18 (5.06)* -2.57 (5.80)** -2.74 (5.80)** SD = Standard deviation Ctrl = Controls
5.73 (5.15, 6.32) 8.25 (7.84, 8.66) 9.83 (9.39, 10.26) WMD = Weighted mean difference
5.73 8.25 9.83
-2.6% -3.2%
Silver Silver Silver
O N
Abs = Absolute ED (95% CI) 1.1% (0.4, 1.9)* 2.0% (0.1, 3.0)* 1.0% (0.1, 1.9)* -
Table 03. Other adverse events from additional sources (EMEA 2004* and Servier**)
Adverse event (AE) VASCULAR SYSTEM DISORDERS Thrombosis Venous thromboembolism at 3 years Venous thromboembolism at 4 years Pulmonary embolism (PE) Fatal PE Sr (n = 3352) 880 (26.3%)* 111 (3.3%)*
AL
Control (n= 3317)
SE
Sr = Strontium ranelate
LY
-3.6% Rel = Relative
OR (95% CI) 1.5 (1.1, 2.1)* 1.7 (1.0, 3.1)* 18
809 (24.4%)* 72 (2.2%)* 50 (1.5%)** 61 (1.8%)** 15 (0.4%)** 3 (0.1%)* 3 (0.1%)* 627 (18.9%)* 97 (2.9%)* 3 (0.1%)** 63 (1.9%)** 66 (2.0%)**
1.9% (-0.2, 4.0)*
IN
PE leading to treatment discontinuation NERVOUS SYSTEM DISORDERS Headaches Seizures at 4 years Memory loss at 4 years Disturbance in consciousness
TE
R
75 (2.2%)** 87 (2.6%)** 25 (0.8%)** 6 (0.2%)* 7 (0.2%)* 699 (20.9%)* 131 (3.9%)* 9 (0.3%)** 79 (2.4%)** 83 (2.5%)**
Table 03. Other adverse events from additional sources (EMEA 2004* and Servier**) (Continued )
Adverse event (AE) at 4 years LABORATORY RESULTS Serum creatine kinase Data obtained from the EMEA* and Servier** 31.3 (80.8) IU/L* Sr = Strontium ranelate 13.1 (46.6) IU/L* 18.2 (14.8; 21.6) IU/L* ED= Estimated difference CI = Condence interval ANALYSES OR = Odds ratio Sr (n = 3352) Control (n= 3317) ED (95% CI) OR (95% CI)
01 Verterbral fractures 02 Non-vertebral fractures
Relative Risk (Random) 95% CI Relative Risk (Random) 95% CI
O N
Statistical method Statistical method
Outcome title
No. of studies
No. of participants
Statistical method
LY
Comparison 01. Fractures
Effect size Subtotals only Subtotals only
Comparison 02. BMD

Outcome title 01 Lumbar spine BMD not adjusted for strontium content 02 Lumbar spine adjusted for strontium content 03 Femoral neck 04 Total hip No. of studies No. of participants
SE
Effect size
Comparison 03. Adverse Events

Outcome title
R
No. of studies
N
No. of participants
AL
Weighted Mean Difference (Random) 95% CI Subtotals only Weighted Mean Difference (Random) 95% CI Subtotals only Weighted Mean Difference (Random) 95% CI Subtotals only Weighted Mean Difference (Random) 95% CI Subtotals only
01 Total withdrawls 02 Withdrawals due to adverse events 03 Number of emergent adverse events 04 Serious adverse events 05 Diarrhea 06 Gastritis 07 Deaths
TE
Effect size Subtotals only Subtotals only Subtotals only Subtotals only Subtotals only Subtotals only Subtotals only
Relative Risk (Random) 95% CI Relative Risk (Random) 95% CI Relative Risk (Random) 95% CI Relative Risk (Random) 95% CI Relative Risk (Random) 95% CI Relative Risk (Random) 95% CI Relative Risk (Random) 95% CI
IN
COVER SHEET Title Authors Contribution of author(s) Strontium ranelate for preventing and treating postmenopausal osteoporosis ODonnell S, Cranney A, Wells GA, Adachi JD, Reginster JY Siobhan ODonnell prepared the protocol and review for submission.
19
Ann Cranney conceptualized the idea and supervised all stages of the review. George Wells provided statistical support. Jonathan Adachi provided feedback on drafts. Jean-Yves Reginster facilitated the attainment of unpublished data and provided feedback on drafts. All co-reviewers reviewed and approved the nal review. Issue protocol rst published Review rst published Date of most recent amendment Date of most recent SUBSTANTIVE amendment Whats New Date new studies sought but none found Date new studies found but not yet included/excluded Date new studies found and included/excluded Date authors conclusions section amended Contact address 2005/2 2006/3 24 May 2006 24 May 2006
Information not supplied by author Information not supplied by author
DOI Cochrane Library number Editorial group Editorial group code
IN
TE
AL
CD005326 HM-MUSKEL
Information not supplied by author
Ms Siobhan ODonnell Research Coordinator Clinical Epidemiology Program Ottawa Health Research Institute 1053 Carling Avenue, C-414 Ottawa ON K1Y 4E9 CANADA E-mail: sodonnell@ohri.ca Tel: 613-798-5555 Fax: 613-761-5492 10.1002/14651858.CD005326.pub2
Cochrane Musculoskeletal Group
SE
O N
LY
20
GRAPHS AND OTHER TABLES Analysis 01.01.

Review: Comparison: 01 Fractures Outcome: 01 Verterbral fractures Study Strontium n/N 01 2 g/day - 1 Year Meunier 2002 Meunier 2004-1 Reginster 2005 Subtotal (95% CI) 22/85 44/719 51/1817 2621 28/87 85/723 93/1823 2633 22.2 37.6 40.2 0.80 [ 0.50, 1.29 ] 0.52 [ 0.37, 0.74 ] 0.55 [ 0.39, 0.77 ] 0.59 [ 0.46, 0.74 ] Placebo n/N Relative Risk (Random) 95% CI Weight (%) Relative Risk (Random) 95% CI
Comparison 01 Fractures, Outcome 01 Verterbral fractures
Strontium ranelate for preventing and treating postmenopausal osteoporosis
Total events: 117 (Strontium), 206 (Placebo) Test for heterogeneity chi-square=2.34 df=2 p=0.31 I =14.4% Test for overall effect z=4.47 02 2 g/day - 3 Years Meunier 2004-1 Reginster 2005 Subtotal (95% CI) 139/719 202/1817 2536 222/723 321/1823 2546 p<0.00001
O N SE U
0.1 0.2 0.5 1 2 5 10 Favours strontium Favours placebo
LY
43.8 56.2 100.0 Placebo n/N Relative Risk (Random) 95% CI Weight (%) 32.6 67.4 100.0
100.0
0.63 [ 0.52, 0.76 ] 0.63 [ 0.54, 0.74 ] 0.63 [ 0.56, 0.71 ]
Total events: 341 (Strontium), 543 (Placebo) Test for heterogeneity chi-square=0.00 df=1 p=0.98 I =0.0% Test for overall effect z=7.37 p<0.00001
Review:
Outcome: 02 Non-vertebral fractures Study
TE
Comparison: 01 Fractures
IN
Strontium ranelate n/N
Analysis 01.02.
Comparison 01 Fractures, Outcome 02 Non-vertebral fractures
AL
Relative Risk (Random) 95% CI
01 2 g/day - 3 Years Meunier 2004-1 Reginster 2005 Subtotal (95% CI)
112/826 233/2479 3305
122/814 276/2453 3267
0.90 [ 0.71, 1.15 ] 0.84 [ 0.71, 0.99 ] 0.86 [ 0.75, 0.98 ]
Total events: 345 (Strontium ranelate), 398 (Placebo) Test for heterogeneity chi-square=0.29 df=1 p=0.59 I =0.0% Test for overall effect z=2.22 p=0.03
21
Analysis 02.01.
Review: Comparison: 02 BMD
Comparison 02 BMD, Outcome 01 Lumbar spine BMD not adjusted for strontium content
Outcome: 01 Lumbar spine BMD not adjusted for strontium content Study Strontium Ranelate N 01 0.5 g/day - 2 Years Meunier 2002 Reginster 2002-1 Subtotal (95% CI) 80 35 115 p=0.0003 5.87 (6.59) 1.90 (3.59) 87 30 117 1.25 (5.30) -0.75 (3.02) 47.5 52.5 100.0 4.62 [ 2.80, 6.44 ] 2.65 [ 1.04, 4.26 ] 3.59 [ 1.66, 5.51 ] Mean(SD) N Placebo Mean(SD) Weighted Mean Difference (Random) 95% CI Weight (%) Weighted Mean Difference (Random) 95% CI
Test for heterogeneity chi-square=2.52 df=1 p=0.11 I =60.4% Test for overall effect z=3.65 02 1 g/d - 2 Years Meunier 2002 Reginster 2002-1 Subtotal (95% CI) 86 29 115 p<0.00001 8.33 (8.65) 5.53 (5.12) 87 30 117 1.25 (5.30) -0.75 (3.02)
O N SE
-100.0 -50.0 0 50.0 100.0 Favours placebo Favours strontium
Test for heterogeneity chi-square=0.27 df=1 p=0.61 I =0.0% Test for overall effect z=8.63 03 2 g/day - 2 Years Meunier 2002 Meunier 2004-1 Subtotal (95% CI) 85 719 804 p<0.00001 13.61 (8.87) 9.99 (9.83) 87 723 810 1.25 (5.30) -1.02 (6.10)
Test for heterogeneity chi-square=1.27 df=1 p=0.26 I =21.3% Test for overall effect z=20.59
IN
TE
AL
LY
50.3 49.7 100.0 20.8 79.2 100.0
7.08 [ 4.94, 9.22 ]
6.28 [ 4.13, 8.43 ]
6.68 [ 5.16, 8.20 ]
12.36 [ 10.17, 14.55 ] 11.01 [ 10.17, 11.85 ] 11.29 [ 10.22, 12.37 ]
22
Analysis 02.02.
Review: Comparison: 02 BMD
Comparison 02 BMD, Outcome 02 Lumbar spine adjusted for strontium content
Outcome: 02 Lumbar spine adjusted for strontium content Study Strontium Ranelate N 01 0.5 g/day - 2 Years Meunier 2002 Reginster 2002-1 Subtotal (95% CI) 80 35 115 p=0.2 3.11 (5.92) -0.78 (3.42) 87 30 117 1.23 (5.31) -0.98 (3.14) 48.2 51.8 100.0 1.88 [ 0.17, 3.59 ] 0.20 [ -1.40, 1.80 ] 1.01 [ -0.63, 2.66 ] Mean(SD) N Placebo Mean(SD) Weighted Mean Difference (Random) 95% CI Weight (%) Weighted Mean Difference (Random) 95% CI
Test for heterogeneity chi-square=1.98 df=1 p=0.16 I =49.5% Test for overall effect z=1.20 02 1 g/day - 2 Years Meunier 2002 Reginster 2002-1 Subtotal (95% CI) 86 29 115 p=0.004 3.20 (7.16) 1.41 (5.33) 87 30 117 1.23 (5.31) -0.98 (3.14)
O N SE
Test for heterogeneity chi-square=0.08 df=1 p=0.78 I =0.0% Test for overall effect z=2.92 03 2 g/day - 2 Years Meunier 2002 Meunier 2004-1 Subtotal (95% CI) 85 719 804 p<0.00001 5.45 (6.80) 5.13 (8.56) 87 723 810 1.23 (5.31) -1.18 (6.26)
IN
TE
AL
LY
58.7 41.3 100.0 41.9 58.1 100.0
1.97 [ 0.09, 3.85 ] 2.39 [ 0.15, 4.63 ] 2.14 [ 0.70, 3.58 ]
4.22 [ 2.39, 6.05 ] 6.31 [ 5.54, 7.08 ] 5.44 [ 3.41, 7.46 ]
23
Analysis 02.03.
Review: Comparison: 02 BMD Outcome: 03 Femoral neck Study Strontium Ranelate N 01 0.5 g/day - 2 Years Meunier 2002 Reginster 2002-1 Subtotal (95% CI) 80 35 115 p=0.2 0.24 (3.01) 0.46 (4.70) 87 30 117 Mean(SD) N
Comparison 02 BMD, Outcome 03 Femoral neck
Placebo Mean(SD)
Weighted Mean Difference (Random) 95% CI
Weight (%)
-0.47 (9.39) -0.87 (4.46)
53.5 46.5 100.0
0.71 [ -1.37, 2.79 ] 1.33 [ -0.90, 3.56 ] 1.00 [ -0.52, 2.52 ]
Test for heterogeneity chi-square=0.16 df=1 p=0.69 I =0.0% Test for overall effect z=1.29 02 1 g/day - 2 Years Meunier 2002 Reginster 2002-1 Subtotal (95% CI) 86 30 116 p=0.002 1.41 (3.45) 2.45 (4.78) 87 30 117 -0.47 (9.39) -0.87 (4.46)
O N SE
Test for heterogeneity chi-square=0.80 df=1 p=0.37 I =0.0% Test for overall effect z=3.16 03 2 g/day - 2 Years Meunier 2002 Meunier 2004-1 Subtotal (95% CI) 85 719 804 p<0.00001 5.36 (8.22) 3.55 (6.47) 87 723 810 -0.47 (9.39) -2.18 (5.06)
Test for heterogeneity chi-square=0.01 df=1 p=0.94 I =0.0% Test for overall effect z=19.22 04 2 g/day - 3 Years Meunier 2004-1 Reginster 2005 Subtotal (95% CI) 719 1393 2112 p<0.00001 5.51 (7.58) 5.65 (7.90)
723
AL
-2.79 (5.67)
1395
-2.57 (5.80)
2118
Test for heterogeneity chi-square=0.03 df=1 p=0.86 I =0.0%
IN
TE
Test for overall effect z=39.17
LY
55.3 44.7 100.0 4.9 95.1 100.0 35.7 64.3 100.0
1.88 [ -0.22, 3.98 ] 3.32 [ 0.98, 5.66 ] 2.52 [ 0.96, 4.09 ]
5.83 [ 3.19, 8.47 ] 5.73 [ 5.13, 6.33 ] 5.73 [ 5.15, 6.32 ]
8.30 [ 7.61, 8.99 ] 8.22 [ 7.71, 8.73 ] 8.25 [ 7.84, 8.66 ]
24
Analysis 02.04.
Review: Comparison: 02 BMD Outcome: 04 Total hip Study Strontium ranelate N 01 2 g/day - 3 Years Meunier 2004-1 Reginster 2005 Subtotal (95% CI) 719 1393 2112 p<0.00001 7.33 (7.67) 7.09 (8.71) 723 1395 2118 Mean(SD) N
Comparison 02 BMD, Outcome 04 Total hip
Placebo Mean(SD)
Weight (%)
-2.49 (5.84) -2.74 (6.19)
38.8 61.2 100.0
9.82 [ 9.12, 10.52 ] 9.83 [ 9.27, 10.39 ] 9.83 [ 9.39, 10.26 ]
-100.0
-50.0
50.0
100.0
Favours placebo
Favours strontium
Analysis 03.01.
Review: Comparison: 03 Adverse Events Outcome: 01 Total withdrawls Study
Comparison 03 Adverse Events, Outcome 01 Total withdrawls
Placebo n/N
SE
Relative Risk (Random) 95% CI Weight (%)
O N
1 2 5 10 Favours placebo
LY
Relative Risk (Random) 95% CI 59.6 40.4 100.0 1.26 [ 0.71, 2.24 ] 0.50 [ 0.19, 1.33 ] 0.87 [ 0.36, 2.11 ] 1.5 16.0 82.5 100.0 1.23 [ 0.69, 2.19 ] 1.07 [ 0.90, 1.28 ] 0.96 [ 0.88, 1.03 ] 0.98 [ 0.91, 1.05 ] 25
01 0.5 g/d Meunier 2002 Reginster 2002-1 Subtotal (95% CI) 20/85 5/40 125
Test for heterogeneity chi-square=2.55 df=1 p=0.11 I =60.7% Test for overall effect z=0.31 02 2 g/d Meunier 2002 Meunier 2004-1 Reginster 2005 Subtotal (95% CI)
TE
p=0.8 20/87 198/826 839/2526 3439 p=0.5
Total events: 25 (Strontium ranelate), 27 (Placebo)
IN
Total events: 1057 (Strontium ranelate), 1069 (Placebo) Test for heterogeneity chi-square=2.01 df=2 p=0.37 I =0.3% Test for overall effect z=0.65
AL
17/91 10/40 131 17/91 182/814 870/2503 3408
0.1 0.2 0.5 Favours strontium
Analysis 03.02.
Review:
Comparison 03 Adverse Events, Outcome 02 Withdrawals due to adverse events
Comparison: 03 Adverse Events Outcome: 02 Withdrawals due to adverse events Study Strontium ranelate n/N 01 0.5 g/d Meunier 2002 Reginster 2002-1 Subtotal (95% CI) 15/85 4/40 125 14/91 6/40 131 76.1 23.9 100.0 1.15 [ 0.59, 2.23 ] 0.67 [ 0.20, 2.18 ] 1.01 [ 0.56, 1.80 ] Placebo n/N Relative Risk (Random) 95% CI Weight (%) Relative Risk (Random) 95% CI
Total events: 19 (Strontium ranelate), 20 (Placebo) Test for heterogeneity chi-square=0.61 df=1 p=0.43 I =0.0% Test for overall effect z=0.02 02 2 g/d Meunier 2002 Meunier 2004-1 Reginster 2005 Subtotal (95% CI) 11/87 140/826 611/2526 3439 14/91 95/814 541/2503 3408 p=1
LY
8.3
0.82 [ 0.39, 1.71 ] 1.45 [ 1.14, 1.85 ] 1.12 [ 1.01, 1.24 ] 1.20 [ 0.96, 1.50 ]
O N SE
36.4 55.4 100.0
IN
TE
AL
26
Analysis 03.03.
Review:
Comparison 03 Adverse Events, Outcome 03 Number of emergent adverse events
Comparison: 03 Adverse Events Outcome: 03 Number of emergent adverse events Study Strontium ranelate n/N 01 0.5 g/d Meunier 2002 Reginster 2002-1 Subtotal (95% CI) 70/85 37/40 125 83/91 39/40 131 42.7 57.3 100.0 0.90 [ 0.80, 1.02 ] 0.95 [ 0.86, 1.05 ] 0.93 [ 0.86, 1.00 ] Placebo n/N Relative Risk (Random) 95% CI Weight (%) Relative Risk (Random) 95% CI
Total events: 107 (Strontium ranelate), 122 (Placebo) Test for heterogeneity chi-square=0.48 df=1 p=0.49 I =0.0% Test for overall effect z=1.89 02 2 g/d Meunier 2002 Meunier 2004-1 Reginster 2005 Subtotal (95% CI) 78/87 730/826 2220/2526 3439 83/91 711/814 2225/2503 3408 p=0.06
LY
3.2
0.98 [ 0.89, 1.08 ] 1.01 [ 0.98, 1.05 ] 0.99 [ 0.97, 1.01 ] 0.99 [ 0.98, 1.01 ]
O N SE
22.9 73.9
100.0
IN
TE
AL
27
Analysis 03.04.
Review: Comparison: 03 Adverse Events Outcome: 04 Serious adverse events Study
Comparison 03 Adverse Events, Outcome 04 Serious adverse events
Placebo n/N
Weight (%)
01 0.5 g/d Meunier 2002 Reginster 2002-1 Subtotal (95% CI) 11/80 5/40 120 17/87 4/40 127 76.1 23.9 100.0 0.70 [ 0.35, 1.41 ] 1.25 [ 0.36, 4.32 ] 0.81 [ 0.44, 1.48 ]
Total events: 16 (Strontium ranelate), 21 (Placebo) Test for heterogeneity chi-square=0.63 df=1 p=0.43 I =0.0% Test for overall effect z=0.69 02 2 g/d Meunier 2002 Meunier 2004-1 Reginster 2005 Subtotal (95% CI) 16/85 188/826 624/2526 3437 17/87 188/814 611/2503 3404 p=0.5
LY
1.9
0.96 [ 0.52, 1.78 ] 0.99 [ 0.83, 1.18 ] 1.01 [ 0.92, 1.11 ] 1.01 [ 0.92, 1.09 ]
O N SE
22.5 75.6
100.0
Analysis 03.05.
Review: Comparison: 03 Adverse Events Outcome: 05 Diarrhea Study
TE
R
Placebo n/N Relative Risk (Random) 95% CI
AL
Comparison 03 Adverse Events, Outcome 05 Diarrhea
Weight (%)
Meunier 2002
IN
01 2 g/d
3/87 50/826
6/91 29/814 125/2503 3408
4.9 31.6 63.5 100.0
0.52 [ 0.13, 2.03 ] 1.70 [ 1.09, 2.66 ] 1.34 [ 1.07, 1.68 ] 1.38 [ 1.02, 1.87 ]
Meunier 2004-1 Reginster 2005 Subtotal (95% CI)
169/2526 3439
0.1 0.2
0.5
10
Favours strontium
Favours placebo
28
Analysis 03.06.
Review: Comparison: 03 Adverse Events Outcome: 06 Gastritis Study Strontium ranelate n/N 01 2 g/d Meunier 2002 Meunier 2004-1 Reginster 2005 Subtotal (95% CI) 5/87 30/826 58/2526 3439
Comparison 03 Adverse Events, Outcome 06 Gastritis
Placebo n/N
Weight (%)
2/91 45/814 68/2503 3408
5.0 40.7 54.3 100.0
2.61 [ 0.52, 13.12 ] 0.66 [ 0.42, 1.03 ] 0.85 [ 0.60, 1.19 ] 0.81 [ 0.56, 1.17 ]
0.01
0.1
Favours strontium
Analysis 03.07.
Review: Comparison: 03 Adverse Events Outcome: 07 Deaths Study Strontium ranelate n/N 01 0.5 g/d Meunier 2002 x Reginster 2002-1 Subtotal (95% CI) 4/85 0/40 125
Comparison 03 Adverse Events, Outcome 07 Deaths
AL
Placebo n/N
U
Relative Risk (Random) 95% CI Weight (%)
SE
O N
1 10 100 Favours placebo
LY
Relative Risk (Random) 95% CI 3/91 0/40 100.0 0.0 100.0 1.43 [ 0.33, 6.19 ] Not estimable 1.43 [ 0.33, 6.19 ] 131 3/91 21/814 159/2503 3408 2.2 34.3 63.5 100.0 0.15 [ 0.01, 2.85 ] 1.36 [ 0.78, 2.37 ] 0.88 [ 0.71, 1.10 ] 0.99 [ 0.64, 1.53 ]
0.001 0.01 0.1 Favours strontium 1 10 100 1000 Favours placebo
Total events: 4 (Strontium ranelate), 3 (Placebo) Test for heterogeneity: not applicable Test for overall effect z=0.48 02 2 g/d p=0.6
Meunier 2002
Meunier 2004-1 Reginster 2005 Subtotal (95% CI)
Total events: 171 (Strontium ranelate), 183 (Placebo) Test for heterogeneity chi-square=3.50 df=2 p=0.17 I =42.8% Test for overall effect z=0.06 p=1
IN
TE
0/87 29/826 142/2526 3439
29

Cochrane Alendronate, Risendronate Systematic Review

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cochrane Alendronate, Risendronate Systematic Review

Uploaded by

Copyright:

Available Formats

Alendronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women (Review)

Wells GA, Cranney A, Peterson J, Boucher M, Shea B, Welch V, Coyle D, Tugwell P

PLAIN LANGUAGE SUMMARY

- 3 out of 100 women had a fracture when taking a placebo

Fractures in bones other than the spine:

- 1 out of 100 women had a fracture when taking alendronate

Fracture of the spine

Fractures in bones other than the spine

Criteria for considering studies for this review

Incidence of fractures, including vertebral, non-vertebral, hip and wrist fractures.

Search methods for identication of studies

Data collection and analysis

Figure 1. Models for fracture risk in postmenopausal women: FRACTURE Index

Figure 5. Summary of Findings for Primary Prevention

Figure 6. Summary of Findings for Secondary Prevention

RESULTS Description of studies

Figure 7. Summary of literature search for alendronate

Risk of bias in included studies

Implications for research AUTHORS CONCLUSIONS Implications for practice

References to studies included in this review

References to studies excluded from this review

References to other published versions of this review

CHARACTERISTICS OF STUDIES Characteristics of included studies [ordered by study ID]

Notes Risk of bias

Bias Allocation concealment (selection bias)

Authors judgement Low risk

Support for judgement A - Adequate

Notes Risk of bias Bias Authors judgement Support for judgement

Allocation concealment (selection bias)

Notes Risk of bias Bias

Allocation concealment (selection bias)

Support for judgement

Outcomes Notes Risk of bias Bias

Allocation concealment (selection bias)

Support for judgement

Outcomes Notes Risk of bias Bias Allocation concealment (selection bias)

Hip Fractures: Ascertainment not reported.

Authors judgement Unclear risk

Hosking 1998 Methods

Notes Risk of bias Bias Allocation concealment (selection bias)

Liberman 1995 Methods

Support for judgement

Notes Risk of bias Bias Allocation concealment (selection bias)

Pols 1999 Methods

Alendronate 10 mg, vs placebo (500mg calcium/day)

Notes Risk of bias Bias Allocation concealment (selection bias)

Characteristics of excluded studies [ordered by study ID]

Study Adami 1993 Adami 1995 Aki 2003

Lack of fracture outcome. Lack of fracture outcome.

Lack of fracture outcome.

Duration of therapy < 1 year. Duration of therapy < 1 year.

Greenspan 2002a Greenspan 2002b

Lack of an appropriate control group.

Lack of appropriate fracture data (i.e. reported as adverse events or unspecied)

Lack of appropriate fracture data (i.e. reported as adverse events or unspecied)

Duration of therapy < 1 year. Lack of fracture outcome.

Lack of an appropriate control group.

Duplicate or earlier report of another study.

DATA AND ANALYSES

Comparison 1. Alendronate 10 mg vs Control - all years baseline denominators

No. of participants 7361 4576 2785

Comparison 2. Alendronate 10 mg vs Control - all years baseline denominators

Comparison 4. Alendronate 10 mg vs Control - all years baseline denominators

1 Hip Fractures 1.1 Hip primary 1.2 Hip secondary