Background/Purpose:

Reports of a link between bisphosphonate use in patients with osteoporosis and atypical femoral fractures–those occurring in subtrochanteric or diaphyseal locations–have generated significant concern among patients and providers.  Attempts to estimate the degree of increased risk have varied widely with OR’s ranging from 1 to 46.  In addition, clinicians and patients may be overestimating the absolute risk of atypical fracture when considering these highly effective drugs.  To obtain better estimates of these rates and assist clinicians in determining the risk versus benefits, we systematically identified and summarized studies reporting rates of atypical femoral fractures in subjects receiving bisphosphonates for osteoporosis. 

Methods:

With assistance of a reference librarian, we searched MEDLINE (1948-2012), and the Cochrane Library (2012) for studies that reported incidence of atypical femoral fractures among subjects exposed to bisphosphonates for osteoporosis.  We also reviewed reference lists and consulted with experts.  We excluded studies evaluating bisphosphonates in malignancy.  We used no language restrictions.  Two reviewers independently extracted data.  Disagreements were resolved through consensus.  Fracture rates were summarized according to study characteristics.

Results:

Of 257 initial studies identified, 12 met all inclusion criteria and included three randomized controlled trials, six retrospective cohorts, and three population-based case-control studies.  Among a total of 205,466 subjects followed over a range of less than 1 to 10 years of bisphosphonate exposure, there were 1440 subtrochanteric or diaphyseal fractures identified by ICD-9 or 10 coding, of which 160 were also confirmed radiographically.  The incidence of atypical femoral fractures per 1000 patient-years of treatment with a bisphosphonate for osteoporosis ranged from 0.02 to 1 in studies that required radiographic verification of atypical nature, 0.23 to 3.4 in observational studies or randomized controlled trials, and 1.55 to 3.4 in studies of secondary prevention.   

Conclusion:

Currently available evidence suggests the incidence of atypical femoral fractures among patients receiving bisphosphonates for osteoporosis is low.  In addition, even at the highest reported estimate of 3.4 per 1000 patient-years, the rate is considerably lower than estimates for recurrent osteoporotic fractures, which occur at approximately 100 per 1,000 patients per year.  Discussions regarding bisphosphonate therapy should put both benefits and harms in appropriate context so that decisions are driven by evidence, not fear.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/incidence-of-atypical-femur-fractures-associated-with-bisphosphonate-use-for-osteoporosis-a-systematic-review-of-the-literature/