In addition to a large societal economic burden, osteoporosis-related fractures cause significant morbidity and premature mortality [ 1]. Evidence from multiple large randomized controlled clinical trials show that bisphosphonates significantly reduce the fracture risk in osteoporosis. The adverse effects of bisphosphonates from clinical trials were mostly comparable to those observed with a placebo. However, post-marketing reports indicate potentially serious safety signals that have generated widespread discussion on the safety of these drugs.
One of the most concerning adverse effects associated with bisphosphonate is that of osteonecrosis of the jaw (ONJ). Bisphosphonate-associated ONJ has been reported predominantly in patients with malignancy who received higher and more frequent doses of i.v. bisphosphonate. Fortunately, among patients with osteoporosis receiving bisphosphonate therapy, the incidence of ONJ is estimated to be very low, between 1 in 10 000 and 1 in 100 000 patient-years [ 2]. The major risk factors for developing ONJ are dental disease, poor dental hygiene and invasive dental procedures. Thus it is important to obtain regular oral exams and address major dental issues prior to initiation of bisphosphonate. For a patient taking a bisphosphonate and undergoing an invasive dental procedure, recent guidelines suggest considering stopping treatment temporarily and allowing mucosalization of any oral openings first, although there is a lack of strong evidence to confirm that this strategy will reduce the risk of ONJ occurrence [ 3].
Atypical femur fracture (AFF)—defined as a fracture in the subtrochanteric region of the hip that most commonly occurs without preceding trauma—is another concerning complication of bisphosphonate use that has been widely publicized. However, similar to ONJ, the incidence of AFF is also low. Multiple well-done epidemiologic studies suggest that it is unlikely the incidence of AFF exceeds 0.1%, but the rate appears to increase with longer duration of bisphosphonate treatment [ 4, 5]. The group at highest risk for this outcome is unknown but risk factors for AFF while taking bisphosphonates appear to include Asian ancestry, hypophosphatasia, active RA, glucocorticoid use for >6 months and low 25-hydroxyvitamin D levels [ 6]. Clinicians should vigilantly evaluate any new or persistent pain in the thigh or groin in a patient taking a bisphosphonate and the drug should be discontinued if an atypical fracture or radiographic prodromal stress reaction along the femoral cortex is found [ 4].
Milder adverse events related to bisphosphonate administration are often averted with precautionary measures and tend to be less treatment limiting. To avoid dyspepsia and the less common esophagitis, oral bisphosphonates should be avoided in patients with gastrointestinal disorders that delay oesophageal emptying (e.g. achalasia, stricture, Barrett’s oesophagus). Taking an oral bisphosphonate with a full glass of water and staying upright for 30–60 min will prevent irritation of the oesophageal mucosa. I.v. zoledronic acid provides an alternative to oral bisphosphonates and is an option that promotes greater adherence. However, with i.v. bisphosphonates, infusion-related acute phase reactions with arthralgia and myalgia are seen in up to a third of patients and may last for several days. Less commonly, these infusions can lead to a flu-like illness. This complication is less prevalent in those who have previously taken oral bisphosphonates, or with subsequent i.v. infusions, and can be reduced with the use of prophylactic acetaminophen before the infusion and at regular intervals for several days afterwards. In patients with chronic kidney disease, bisphosphonates, particularly i.v. preparations, should not be prescribed when the glomerular filtration rate is 2 (ergocalciferol) at or around the time of zoledronic acid administration. Clinicians should also encourage their bisphosphonate-using patients to consume appropriate calcium from dietary sources or supplementation.
The concerns over ONJ and AFF in particular have led to questions about the optimal duration of bisphosphonate therapy. Since bisphosphonates reside in the skeleton for a long period, a drug holiday is suggested. The rationale is to avoid unnecessary outcomes associated with long-term use while continuing to obtain benefit from the drug. The Fracture Intervention Trial Long-term Extension study showed that even though BMD declined at hip sites after discontinuation of alendronate following 5 years of treatment, no difference was observed in the rates of non-vertebral fractures in those patients relative to patients who continued treatment for an additional 5 years [ 7]. Recent guidelines recommend reviewing bisphosphonate treatment after 5 years with alendronate, risedronate or ibandronate and after 3 years of treatment with zoledronic acid [ 8]. Drug holidays should be considered in lower-risk patients. Reassessment of fracture risk may be performed with FRAX (https://www.sheffield.ac.uk/FRAX/) and bone density scans after 1.5–3 years. However, patients who remain at high risk of fractures (age >75 years, previous hip or vertebral fracture, occurrence of fracture while on treatment, current glucocorticoid therapy with prednisolone ⩾7.5 mg daily) should continue to receive either continued bisphosphonates or an alternative type of anti-osteoporotic treatment for a longer period of time [ 8].
Even though advances in bone biology have led to the development of newer agents (currently in late-phase development), present treatment choices remain limited and it is likely that bisphosphonates will remain a common first-line option for some time to come. Since serious adverse effects are rare, the benefit:risk ratio for fracture prevention is usually high and clinicians should offer these drugs to all those who may benefit from them and have no clear contraindications. For example, Black and Rosen [ 9] estimated that treating 1000 women for 3 years prevents 100 typical fractures at any site compared with 0.08 atypical fractures. Effective physician–patient communication is imperative. We spend time with patients discussing rare but real risks, reinforce benefits, simplify treatment regimens and tailor therapy to the patient’s preference. The need for continued bisphosphonate treatment demands reassessment by the clinician on a regular basis determined by clinical consideration of a patient’s risk profile assisted by both reassessment of BMD and measurements of bone turnover. There remain many uncertainties about the optimal duration of bisphosphonates, the timing and benefits of drug holidays and when treatment should be reinstated. A lack of data to address these issues necessitates large pragmatic treatment strategy trials to provide an evidence base for decisions around the practice of optimal osteoporosis care. Until then, we rely on well-crafted international guidelines and evidence-based medicine, which also emphasize the importance of patient preference and choice.
We thank Dr. Elizabeth Rahn for her assistance in language editing of the manuscript.
Funding: No specific funding was received from any funding bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article.
Disclosure statement: K.G.S. reports grants and personal fees from Amgen and Merck and personal fees from Lilly and Radius. The other author has declared no conflicts of interest.