Skip to main content
PMC home page
Genetic Testing and Molecular Biomarkers logoLink to Genetic Testing and Molecular Biomarkers
. 2014 Feb 1;18(2):117–122. doi: 10.1089/gtmb.2013.0331

Bisphosphonates and Risk of Subtrochanteric, Femoral Shaft, and Atypical Femur Fracture: Sensitivity and Trim and Fill Studies

Jie Liu 1, Hong-xin Zhang 2, Xiong-xiong Lu 3, Jia-jia Hu 4, Lian-fu Deng 1,✉
PMCID: PMC3926151  PMID: 24205872

Abstract

Objective: This study carried out sensitivity analysis and trim-fill analysis between bisphosphonates and subtrochanteric, femoral shaft, and atypical femur fracture. Methods: A random-effects model was used finally. Sensitivity, trim and fill, and publication bias analyses were done. Results: Under a random-effects model (I2=87.535), the Z-value=5.672, p-value of test of null<0.001. Bisphosphonate exposure was associated with an increased risk of atypical femur fracture (3.243 [95% CI 2.160–4.870]). When any study is removed, the remaining sensitivity analysis results are still significant. Trim and fill results show that two studies were missed. After filling them, a funnel plot of precision by log risk ratio was more symmetrical. Conclusion: This study suggests that (1) there is an increased risk of subtrochanteric, femoral shaft, and atypical femur fracture in bisphosphonate users; (2) any single study does not influence the total sensitivity; (3) two studies have been lost, theoretically.

Introduction

Osteoporosis is a progressive systemic skeletal disease characterized by low bone mass and deterioration of bone microarchitecture. It has a relationship with bone fractures that significantly increases the morbidity and mortality of affected patients. It mainly affects postmenopausal women; about 50% of women older than 50 years will sustain an osteoporosis-related fracture during their lifetime (Ioannidis et al., 2009; Papaioannou et al., 2009). According to the radiographic appearance, femoral fractures were classified as stress fractures or nonstress fractures. Medications, such as bisphosphonates, are therefore increasingly used to prevent and treat osteoporosis. A number of studies showed that bisphosphonates could reduce the overall risk of fracture in patients with osteoporosis (Black et al., 2006). However, concerns have been raised about the potential risk of these drugs during long-term use. Some other studies showed an increased risk of subtrochanteric and femoral shaft fractures with the use of bisphosphonates (Park-Wyllie et al., 2011; Schilcher et al., 2011; Vestergaard et al., 2011). Among them, stress fractures (commonly called atypical femoral fractures) were more common in patients who received bisphosphonates than in those who did not (Lenart et al., 2009; Girgis et al., 2010). Some studies also showed that bisphosphonates could negatively affect bone remodeling and lead to increased microdamage (Mashiba et al., 2000; Odvina et al., 2005). However, these reports are controversial and the detailed mechanism of this phenomenon is unknown.

More recently, Gedmintas et al. (2013) published a meta-analysis that indicated that there is an increased risk of subtrochanteric, femoral shaft, and atypical femoral fractures among bisphosphonate users. However, they did not perform two very important statistics, sensitivity analysis and trim and fill analysis. To analyze this issue more thoroughly and deeply, we performed this comprehensive study.

Materials and Methods

Identification of eligible studies

We searched the MEDLINE and EMBASE databases, from January 1990 to October 2012 for studies examining the association of bisphosphonate use and atypical femur fracture, as well as subtrochanteric and femoral shaft fractures. Only studies published in English were included. The keywords for searching were the same as Gedmintas et al's (2013).

Data extraction and quality evaluation

The following information was independently extracted from the identified studies by two participants: first author, journal, year of publication, study design, ethnicity of the study population, the number of cases and controls or risk ratios (RR) and 95% confidence interval (CI), country in which the study was conducted, and what type of bisphosphonate was used. The results were compared and any disagreement was discussed and resolved by consensus. According to Gedmintas et al. (2013), the Newcastle-Ottawa Quality Assessment Scale was used independently by the two authors (J.L. and H.X.Z.) to help determine the quality of studies included in the study.

Statistical analysis

This study examined the overall association between bisphosphonates and risk of subtrochanteric, femoral shaft, and atypical femur fracture. The effect size was represented by RR with 95% CI. The Cochran's Q statistical test and the I2 test were used to assess heterogeneity in combined studies. Publication bias was checked using the Begg's test and the Egger's test was used for funnel plot asymmetry (Begg and Mazumdar, 1994; Egger et al., 1997). p-Values of overall RR were generated using the Z test. Sensitivity analysis was conducted by removing each study and analyzing the others to ensure no single study was totally responsible for the overall results. The significance level was set at 0.05, and all p-values were two tailed. This study was performed using Comprehensive Meta Analysis software (Version 2.2.046; BIOSTAT, Englewood, NJ).

Results

Finally, 11 references were selected in this study, including 10 cohort studies and 6 case–control studies (Abrahamsen et al., 2009, 2010; Lenart et al., 2009; Black et al., 2010; Hsiao et al., 2011; Kim et al., 2011; Park-Wyllie et al., 2011; Schilcher et al., 2011; Vestergaard et al., 2011; Feldstein et al., 2012; Meier et al., 2012). The total analyses demonstrate that both the fixed-effects and the random-effects models show a significant association between bisphosphonates and risk of subtrochanteric, femoral shaft, and atypical femur fracture (fixed-effects model: adjusted RR of 2.106 [95% CI 1.888–2.349]; random-effects model: adjusted RR of 3.243 (95% CI 2.160–4.870]) (Figs. 1 and 2). As I-squared (I2)=87.535, Q-value=120.334, p-value of heterogeneity<0.001, the random-effects model was used to calculate the pooled estimates of adjusted RR. Under this model, the analysis results show Z-value=5.672, p-value of Test of null (2-Tail)<0.001(Table 1). Sensitivity analysis results demonstrate that when any single study is removed, the remaining results are still significant (Fig. 3). The trim and fill results show that two necessary studies have been missed. After filling these two in the comprehensive analysis, the funnel plot of precision by the log risk ratio show much more symmetry (Fig. 4A, B).

FIG. 1.

FIG. 1.

Open in a new tab

Fixed-model meta-analysis of the relationship between bisphosphonates and risk of subtrochanteric, femoral shaft, and atypical femur fracture. The overall odds ratio (OR) is shown. The OR of each study is marked with a black square. The overall OR is indicated by diamond.

FIG. 2.

FIG. 2.

Open in a new tab

Random-model meta-analysis of the relationship between bisphosphonates and risk of subtrochanteric, femoral shaft, and atypical femur fracture. The overall OR is shown. The OR of each study is marked with a black square. The overall OR is indicated by diamond.

Table 1.

The Statistics of Basic Parameters in the Study

  Effect size and 95% interval Test of null (2-Tail) Heterogeneity Tau squared
Model Number studies Point estimate Lower limit Upper limit Z-value p-Value Q-value df(Q) p-Value I-squared Tau squared Stand error
Fixed 16 2.106 1.888 2.349 13.337 0 120.334 15 0 87.535 0.424 0.308
Random 16 3.243 2.16 4.87 5.672 0 / / /      
Open in a new tab

‘/’ means does not exist.

FIG. 3.

FIG. 3.

Open in a new tab

The sensitivity analysis of the studies for the association between bisphosphonate use and subtrochanteric, femoral shaft, and atypical femur fracture. Point (square) and overall (diamond) estimates are given as risk ratios with 95% confidence interval (CI) (horizontal bar).

FIG. 4.

FIG. 4.

Open in a new tab

(A). Funnel plot with pseudo 95% confidence limits. (B) Funnel plot of all studies with pseudo 95% CI, including the hypothetical studies using the trim and fill method.

Discussion

Bisphosphonates (also called diphosphonates) have been used to treat osteoporosis and similar diseases for several decades. It has been shown that this kind of drug could increase bone mineral density and reduce vertebral and proximal femur fracture risk in patients with osteoporosis. Although bisphosphonates are generally safe and effective, more and more studies indicated that women taking bisphosphonates for osteoporosis have had unusual fractures. Gedmintas et al. summarized these published data in a meta-analysis and suggested that long-term use of bisphosphonates was significantly associated with the risk of osteoporosis-related fractures. However, their study had some limitations and some important statistical analyses were not performed.

In this current study, we conducted a comprehensive analysis for the association between bisphosphonate use and risk of subtrochanteric, femoral shaft, and atypical femur fracture. The funnel plot appears asymmetric (Fig. 4A) suggesting publication bias. Therefore, we performed a trim and fill analysis, which is a simple funnel plot-based method of testing and adjusting for publication bias in meta-analysis. This method uses only an iterative approach and simple symmetry assumptions are easy to implement in practice and to estimate the number of missing studies. Following this method, we found that two missing studies should be filled, theoretically. After the trim and fill statistics were taken, the funnel plot seemed much more symmetrical (Fig. 4A, B). On the other hand, the trim and fill analysis did not change the direction of the results, although the magnitude of the association was somewhat weakened, indicating that the association is not an artifact of unpublished negative studies. Sensitivity analysis (Fig. 3) was conducted by removing each study and analyzing the others to ensure that no single study was totally responsible for the overall results. The results indicate that no heterogeneity existed in the population.

The main limitation in our study is the varying definition of an atypical femur fracture. In addition, we were limited by the significant heterogeneity of studies, as the total I2=87.535 (Table 1). Many potential factors may cause this heterogeneity, such as patient characteristics, variations in study design and size, type of bisphosphonates and so on. Because of the value of I2, we used a random-effects model to account for both sensitivity analysis and trim and fill analysis. Also, we performed a statistical analysis of the basic parameters of this study (Table 1).

Conclusions

Despite these limitations, our current study, demonstrating two very important analyses—sensitivity analysis and trim and fill analysis—supports the study of Gedmintas et al. (2013), and makes the analysis on the research area of bisphosphonates and risk of subtrochanteric, femoral shaft, and atypical femur fracture complete.

Acknowledgments

This work is supported by grants from the National Natural Science Foundation of China (grant number 31000408) and grants from the Science and Technology Commission of Shanghai municipality (grant number 13ZR1461100).

Author Disclosure Statement

The authors have declared that no competing interests exist.

References

  1. Abrahamsen B, et al. (2009) Subtrochanteric and diaphyseal femur fractures in patients treated with alendronate: a register-based national cohort study. J Bone Miner Res 24:1095–1102 [DOI] [PubMed] [Google Scholar]
  2. Abrahamsen B, et al. (2010) Cumulative alendronate dose and the long-term absolute risk of subtrochanteric and diaphyseal femur fractures: a register-based national cohort analysis. J Clin Endocrinol Metab 95:5258–5265 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Begg CB, Mazumdar M. (1994) Operating characteristics of a rank correlation test for publication bias. Biometrics 50:1088–1101 [PubMed] [Google Scholar]
  4. Black DM, et al. (2006) Effects of continuing or stopping alendronate after 5 years of treatment: the Fracture Intervention Trial Long-term Extension (FLEX): a randomized trial. JAMA 296:2927–2938 [DOI] [PubMed] [Google Scholar]
  5. Black DM, et al. (2010) Bisphosphonates and fractures of the subtrochanteric or diaphyseal femur. N Engl J Med 362:1761–1771 [DOI] [PubMed] [Google Scholar]
  6. Egger M, et al. (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ 315:629–634 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Feldstein A, et al. (2012) Incidence and demography of femur fractures with and without atypical features. J Bone Miner Res 27:977–986 [DOI] [PubMed] [Google Scholar]
  8. Gedmintas L, et al. (2013) Bisphosphonates and risk of subtrochanteric, femoral shaft, and atypical femur fracture: a systematic review and meta-analysis. J Bone Miner Res 28:1729–1737 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Girgis CM, et al. (2010) Atypical femoral fractures and bisphosphonate use. N Engl J Med 362:1848–1849 [DOI] [PubMed] [Google Scholar]
  10. Hsiao FY, et al. (2011) Hip and subtrochanteric or diaphyseal femoral fractures in alendronate users: a 10-year, nationwide retrospective cohort study in Taiwanese women. Clin Ther 33:1659–1667 [DOI] [PubMed] [Google Scholar]
  11. Ioannidis G, et al. (2009) Relation between fractures and mortality: results from the Canadian Multicentre Osteoporosis Study. CMAJ 181:265–271 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Kim SY, et al. (2011) Oral bisphosphonates and risk of subtrochanteric or diaphyseal femur fractures in a population-based cohort. J Bone Miner Res 26:993–1001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Lenart BA, et al. (2009) Association of low-energy femoral fractures with prolonged bisphosphonate use: a case control study. Osteoporos Int 20:1353–1362 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Mashiba T, et al. (2000) Suppressed bone turnover by bisphosphonates increases microdamage accumulation and reduces some biomechanical properties in dog rib. J Bone Miner Res 15:613–620 [DOI] [PubMed] [Google Scholar]
  15. Meier RPH, et al. (2012) Increasing occurrence of atypical femoral fractures associated with bisphosphonate use. Arch Intern Med 172:930–936 [DOI] [PubMed] [Google Scholar]
  16. Odvina CV, et al. (2005) Severely suppressed bone turnover: a potential complication of alendronate therapy. J Clin Endocrinol Metab 90:1294–1301 [DOI] [PubMed] [Google Scholar]
  17. Papaioannou A, et al. (2009) The impact of incident fractures on health-related quality of life: 5 years of data from the Canadian Multicentre Osteoporosis Study. Osteoporos Int 20:703–714 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Park-Wyllie LY, et al. (2011) Bisphosphonate use and the risk of subtrochanteric or femoral shaft fractures in older women. Jama 305:783–789 [DOI] [PubMed] [Google Scholar]
  19. Schilcher J, et al. (2011) Bisphosphonate use and atypical fractures of the femoral shaft. N Engl J Med 364:1728–1737 [DOI] [PubMed] [Google Scholar]
  20. Vestergaard P, et al. (2011) Risk of femoral shaft and subtrochanteric fractures among users of bisphosphonates and raloxifene. Osteoporos Int 22:993–1001 [DOI] [PubMed] [Google Scholar]

Articles from Genetic Testing and Molecular Biomarkers are provided here courtesy of Mary Ann Liebert, Inc.

RESOURCES