Introduction:
Atypical femoral fractures are a rare but devastating complication of prolonged anti-resorptive therapy.1, 2 The management is challenging, with some evidence to support the use of teriparatide therapy.3 This case highlights this rare but important risk of bisphosphonate and denosumab use, and reinforces the importance of frequent re-evaluation of the risk versus benefit of prolonged antiresorptive treatment for osteoporosis.
Case summary:
61 year old woman was managed for bilateral atypical femoral fractures in the setting of antiresorptive therapies. She had not had any fractures prior to this. She has a background of osteoporosis diagnosed at the age of 51 based on the results of bone densitometry, after which she was started on risedronate by her GP. After approximately 8 years of treatment with risedronate she was transitioned to denosumab due to concern about the lack of improvement in bone density. After 12 months of therapy with denosumab, she presented to hospital with atraumatic bilateral atypical femoral fractures. For management of her fractures she underwent bilateral femoral intramedullary nail insertion.
The patient’s risk factors for osteoporosis include chronic hepatitis B. She underwent menopause at the age of 47. She had not had any hormone replacement therapy. There was no family history of osteoporosis. She does not smoke or drink alcohol. Her other comorbidities are latent tuberculosis, dyslipidaemia and hypertension. Her medications are atorvastatin 20mg daily and candesartan 8mg daily and colecalciferol 25 micrograms daily.
On further history and examination there were no relevant features to suggest a secondary cause of osteoporosis.
Her previous bone mineral densitometry had been tested 2 years prior to her presentation with atypical femoral fractures. The result had been consistent with osteopaenia, revealing a lumbar spine T score of -2.4, a total hip T score of -0.8, a femoral neck T score of -1.2 and a forearm T score of -0.4.
Relevant investigations at the time of her admission to hospital included a vitamin D level of 93nmol/L, an unremarkable full blood count, normal renal function and electrolytes, mildly elevated liver enzymes (less than 100U/L), normal calcium/magnesium/phosphate. Serum protein electrophoresis and serum light chains were unremarkable. Her parathyroid hormone level was mildly elevated at 10pmol/L and her thyroid function revealed a low TSH of 0.34mU/L with a free T4 level of 12.9pmol/L.
The patient is now recovering from her surgery to address the atypical femoral fractures. She is currently mobilising with crutches. She will be followed up with a view to bridging therapy with oral alendronate therapy for 12-18 months, 9 months after her last dose of denosumab.
Discussion
The decision to initiate antiresorptive therapy in osteoporosis should consider the fracture risk reduction versus the risk of adverse events. Indications for treatment include a history of minimal trauma fracture, a T score of ≤-2.5 or a 10 year risk of hip fracture or major osteoporotic fracture that is above 3% or 10% respectively.4 Bisphosphonates are first line; denosumab can be a feasible alternative therapy in some patients.4, 5
Guidelines recommend that after 3-5 years of bisphosphonate use, the patient’s risk of osteoporotic fractures should be re-evaluated with a view to a ‘drug holiday’ in low risk patients for a period of 2-3 years, given that the risk of adverse effects such as atypical femoral fractures start to rise with long term use.4, 6, 7 A similar principle cannot be applied to denosumab however, given the high incidence of vertebral fractures following discontinuation of this medication. At present we have data on the use of denosumab for up to 10 years. Current guidelines suggest that after 5-10 years of treatment with denosumab, patients could be transitioned to an alternative anti-resorptive agent such as a bisphosphonate, before stopping the denosumab.5, 8
An atypical femoral fracture is defined as a stress fractures of the femoral shaft, with distinct radiological features.3 The true incidence is uncertain as the definition of the fracture has altered in recent years, with one study finding that the incidence was 5.9 per 100,000 patient years as per 2010 criteria, dropping to 2.9 per 100,000 patient years as per 2013 criteria. The incidence rises significantly after 6-10 years of therapy with bisphosphonates.9
Denosumab was approved in Australia by the Therapeutics Goods Administration in 201010 and it is also associated with atypical femoral fractures. Its use in the population is expected to rise given the ease of parenteral administration and the known difficulties with bisphosphonate compliance.11 The FREEDOM extension study revealed 2 cases which occurred after 3 and 7 years of continuous denosumab therapy.12
Management of atypical femoral fractures often involves surgical fixation with an intramedullary nail, particularly if the fracture is complete. The antiresorptive agent must be discontinued and there needs to be optimal calcium and vitamin D intake.3 Teriparatide has also been used to improve bone healing. However, the literature is only for patients who have used bisphosphonate therapy. Furthermore, the data supporting its use is variable, and it largely arises from case reports and small case series. 3, 13
Conclusion:
Whilst rare, atypical femoral fractures are an important adverse event to consider in patients taking bisphosphonates and denosumab, therefore the risk versus benefit of treatment should be periodically re-evaluated. The medical management of atypical femoral fractures is challenging, with there being variable evidence to support the use of teriparatide use. Our case highlights the significant impact of a potentially avoidable complication of bisphosphonate and denosumab use and the challenges of treating atypical femoral fractures.