Questions in osteoporosis management

Case data
You enter the examination room to see Ms Stewart, an 83-year-old long-time patient who is accompanied by her daughter. She has come to follow up with you 1 week after being discharged from hospital, where she was admitted for 9 days following a hip fracture. You review her discharge medications, which now include vitamin D (1000 units daily) but are otherwise unchanged from before her hospital stay. Ms Stewart and her daughter want to know what can be done to prevent another hip fracture. Ms Stewart states that since her hospitalization she has been feeling well and her mobility is improving. She is afraid of having another fall and is using a two-wheeled walker. You review her history and medications.

At age 68, Ms Stewart had a fall while visiting Edmonton. She slipped on ice and sustained a Colles fracture (distal radius) that required 6 weeks in a cast. When she returned to Vancouver she went to see you. Concerned about osteoporosis, you ordered a dual X-ray absorptiometry (DXA) scan to establish bone mineral density (BMD) and opted to repeat the scan at 5-year intervals. Because Ms Stewart’s BMD from this first scan was not in the osteoporotic range you started her on calcium rather than a bisphosphonate. You wonder now if there were other clinical factors you should have considered when assessing Ms Stewart for osteoporosis 15 years ago.

Today Ms Stewart is 5 feet 6 inches tall and weighs 128 pounds. She has a history of hypertension and dyslipidemia as well as mild COPD. She has not had exacerbations requiring prednisone. She is an active smoker. While her first BMD measurement (T score 22.2 at the hip) and most recent measurement (T score 22.4) were not in the osteoporotic range, you refer to the World Health Organization’s Facture Risk Assessment (FRAX) tool and find her risk of a major osteoporotic fracture in 10 years is 21% and her risk of a hip fracture is 8.7%.

After you and Ms Stewart discuss the risks and benefits of treatments to prevent fracture, she opts to start taking oral alendronate (70 mg weekly). You discuss the proper way to take this medication—ingesting it on an empty stomach and remaining upright after—and ask her to come back in 6 weeks to ensure there are no concerns or side effects. Based on her FRAX results, you estimate that this treatment compared with no treatment will reduce Ms Stewart’s risk of a future fragility fracture by approximately 50% over the next 3 to 4 years. You decide to measure her BMD now and again in 3 years to determine if the therapy has improved her BMD.

Discussion
Cases such as this are all too common in our aging population. Osteoporosis occurs in 1 in 4 women and 1 in 8 men over the age of 50 and affects approximately 1.4 million Canadians.[1] Fragility fractures (and specifically hip fractures) in older adults are one of the most significant and feared consequences of osteoporosis. Following a hip fracture, the mortality rate is approximately 20% at 1 year, and an increased mortality rate persists for 5 years.[2,3] For a woman over 50 the lifetime risk of a hip fracture is 12.1%, but most fractures occur later in life at an average age of 83.[4]

Hip fractures are often not the first sign of osteoporosis or of increased risk for fragility fractures, and strategies aimed at identifying patients at risk are key to prevention. Over 80% of all fractures in people 50 years and older are caused by osteoporosis.[5] While BMD is useful for the diagnosis of osteoporosis, a T score of 22.5 is not required to make a diagnosis of osteoporosis. Indeed, 60% of women with fragility fractures have T scores in the nonosteoporotic range (most often in the osteopenic range).6 Clearly there are other factors that need to be considered when assessing patients for their risk of a fracture.

Assessment
The 2010 Osteoporosis Canada guidelines recommend not using BMD alone to calculate a patient’s fracture risk.[7] Instead, the guidelines recommend two well-validated scoring systems, the FRAX tool already mentioned in the case described and the Canadian Association of Radiologists and Osteoporosis Canada (CAROC) tool. Both can be used to evaluate the risk of a fracture and they have excellent concordance.

The CAROC tool has five parameters and stratifies patients into three 10-year fracture risk categories: low, moderate, and high. The tool can be downloaded without charge either as a web-based calculator or mobile app (www.osteoporosis.ca/health-care-professionals/clinical-tools-and-resourc...). According to the CAROC guidelines, all patients who are found to be at high risk (greater than 20% risk of fragility fracture over 10 years) should be offered treatment, while those at low risk (less than 10% risk) can be monitored with close follow-up. Patients with a moderate risk of fracture (10% to 20% risk over 10 years) require further workup and review. The Osteoporosis Canada guidelines recommend that all moderate-risk patients undergo a lateral lumbar spine X-ray to identify any silent or asymptomatic vertebral fractures. Other factors to consider when assessing moderate-risk patients include past use of glucocorticoids, recurrent falls, rapid bone loss, or current treatment with an aromatase inhibitor. The presence of these risk factors suggests the patient may benefit from therapy to prevent fracture, but clinical judgment is key.[7]

The FRAX tool can be used to calculate a 10-year fracture risk score and the free online calculator is simple to use. One advantage of this calculator over the CAROC calculator is that it can be used in multiple countries and does not require a BMD measurement. When BMD is available, both FRAX and CAROC are calculated with femoral neck T scores, even if the spine T-score is lower.

Using either of these risk calculation tools gives the practitioner a better idea of the true fracture risk over the next 10 years, and allows for a more informed discussion regarding treatment options. While useful for guiding decisions regarding therapy, both of these tools have their limitations. Both can be used only in patients older than 40, do not apply to patients already on osteoporosis therapy, and may underestimate risk in young patients. Furthermore, the FRAX requires detailed knowledge of patient risk factors, including glucocorticoid use, prior fracture, parental hip fracture, current smoking, excess alcohol consumption, and rheumatoid arthritis.

Management
The management of osteoporosis includes both pharmacological and nonpharmacological therapies. Exercise, particularly a combination of resistance training and cardiovascular fitness, has been associated with lower rates of hip fracture in older women.[8] There is no evidence that high-intensity exercise is more beneficial than low-intensity exercise, and therefore we recommend that women find an activity that they enjoy doing to ensure long-term compliance. Generally, 30 minutes of exercise 3 days a week is a good starting point for women who are not exercising currently.9 While important for general health and for reducing the risk of fracture, the overall effect of exercise on BMD is low. Further pharmacological treatment is generally recommended.

In the 25 years since effective treatment for osteoporosis has been available, bisphosphonates have been the most widely prescribed, best-studied, and most effective osteoporosis medications. Denosumab was recently added to the list of first-line agents, while selective estrogen receptor modulators (SERMs), calcitonin, and parathyroid hormone (PTH) receptor agonists such as teriparatide remain second-line or niche agents.[7] The most recent Osteoporosis Canada guidelines recommend the following as first-line agents: oral alendronate or risedronate, zoledronic acid by infusion, and denosumab by injection.[7]

Questions about the safety and efficacy of bisphosphonates are common, and newer agents such as denosumab continue to be underutilized because they are unfamiliar to many practitioners. While our intention here is not to provide a comprehensive overview of osteoporosis or its management, and we cannot discuss menopausal hormone therapy, SERMs, or teriparatide, which are not typically first-line therapy and require more in-depth consideration of indications and potential side effects, we can address some common questions regarding the management of osteoporosis.

Should my patient take calcium and vitamin D? The first step toward optimizing bone health is ensuring that a person has the adequate building blocks to maintain healthy bones and prevent bone loss. Calcium and vitamin D are an integral part of bone formation, but unfortunately many patients lack adequate intake of these nutrients. Guidelines differ significantly on the amount of supplemental calcium patients with osteoporosis require, from none up to 1200 mg a day.[10] We recommend calcium from dietary sources—specifically dairy and dairy equivalents—rather than supplementation. However, in patients with little or no dairy intake (less than three servings a day), we recommend supplementation with 500 mg of elemental calcium a day. It is important to recognize that elemental calcium is not the same as total calcium. Different formulations of calcium salts contain different amounts of elemental calcium. For example, calcium carbonate contains 40% elemental calcium, so 1200 mg are needed to provide an intake of 500 mg of elemental calcium. In a recent large Canadian study, the benefits of up to 1000 mg of calcium supplementation (with or without vitamin D) were found to include a modest decrease in fractures as well as a lower mortality rate.[11] The safety of supplemental calcium has been questioned by a few studies that found a small but statistically significant increase in myocardial infarcts in patients using supplemental calcium,[12-14] but the methodology of these studies has been challenged and the results have not been reproduced in other studies.[15,16] 
The rates of vitamin D deficiency in older adults are estimated at between 40% and 100%, with rates being especially high among patients in nursing homes.[17] Osteoporosis Canada recommends that all older patients at risk of fracture take 800 to 2000 international units of vitamin D a day, as this is associated with a more robust response to antiresorptive therapy as well as a lower risk of fracture.[5] This dose is safe, effective at improving bone health, and without risk. Furthermore, this dose will achieve sufficient levels of vitamin D and does not require repeated monitoring of levels unless malabsorption is a concern. It should be noted, however, that supplementation in patients with normal vitamin D levels has not been shown to prevent falls or fractures, as the effect was limited to those with vitamin D deficiency.[17]

Should my patient take a drug holiday after a certain period of treatment? The question of whether a drug should be discontinued for some time and then restarted has been hotly debated in the literature. First, there is no evidence to support the common belief in a finite timeline for treating osteoporosis. Treatment can continue safely beyond 5 years. Second, it should be noted that a drug holiday applies only to bisphosphonates. This approach does not apply to menopausal hormone replacement therapy, raloxifine, or denosumab. The rationale behind discontinuing bisphosphonates for a drug holiday is that these drugs have such an affinity for binding to bone that a substantial amount of the drug can accumulate over time, leading to a reservoir that continues to be released for months or years after treatment is stopped.[18] When treatment is stopped, the continued presence of bisphosphonate in bone and continued release (and possible reattachment to bone) means there may be some lingering antifracture effect.[19] The optimal time for a drug holiday has not been established. Drug holidays are usually considered after approximately 3 to 5 years of therapy. The Osteoporosis Canada guidelines recommend that patients at high risk of fracture do not take a drug holiday, as the risk of fracture far outweighs the potential side effects. Patients at low risk can be offered a break from therapy, as these patients may not benefit from therapy in the first place. Patients at moderate risk are more challenging when it comes to deciding on a drug holiday, as they are when making the original decision to treat, and may be considered for a drug holiday if they have no previous fragility fracture and a femoral neck BMD greater than 22.5. If a vertebral fracture is found during workup, however, the patient should be considered high risk and therapy should continue.20 The recommended length of a drug holiday is between 1 and 3 years, and the possibility of restarting therapy should be revisited on a yearly basis.20 Studies have shown drug holidays to be technically safe, but an individual patient’s clinical factors must be carefully considered and potential risks must be discussed.[21]

What common side effects of bisphosphonates might my patient experience? The most common complaint from patients on bisphosphonates is dyspepsia. Part of this and other gastrointestinal symptoms may be due to taking the medication improperly (i.e., not remaining upright for 30 minutes after ingestion or failing to take the drug on an empty stomach). While dyspepsia is not a serious side effect, it is one of the most common reasons for non-adherence to a bisphosphonate regimen.[22] Interestingly, there may be an easy solution to this. Alendronate was first approved for use in 2000 and in 2005 a generic version became available. Almost immediately clinicians noticed some patients who had previously tolerated the branded version were experiencing an increase in side effects (usually GI-related) with the new generic tablet.[23] This led to a 20% lower adherence rate after 1 year on generic versus branded bisphosphonates.[24] One possible explanation for the different reactions to the two formulations is that generic bisphosphonate compounds have been found to disintegrate significantly faster in vitro than branded versions, which may increase their contact with the esophageal mucosa and contribute to dyspepsia.[25] Thus, patients who are not tolerating their oral bisphosphonate despite taking it correctly may be considered for a trial of a branded bisphosphonate before abandoning or switching agents. 

What serious side effects of bisphosphonates might my patient experience? While bisphosphonates are highly effective at preventing fragility fractures and tend to be safe and well tolerated, concerns have emerged that long-term use poses an increased risk of atypical fractures and osteonecrosis of the jaw (ONJ). While case studies have raised concerns about these side effects, no evidence of this association is available yet from prospective randomized controlled trials.

While ONJ is a risk for patients on bisphosphonate therapy, it is exceedingly rare. The absolute risk of ONJ when a bisphosphonate is administered for osteoporosis is approximately 1 case per 10 000 person-years.[20] The risk of ONJ is highest in patients using a bisphosphonate in the context of a malignancy, a population outside the scope of this article. Other risk factors for ONJ include obesity, diabetes, oral infection, and dental extractions (although routine dental care does not require modification solely because of bisphosphonate use).[26]

Finally, the concern about atypical femur fracture has made some clinicians hesitate to use bisphosphonates, especially in patients with previous hip fractures. Like ONJ, atypical fractures are exceptionally rare, accounting for less than 1% of all hip and femur fractures.[27] To date, there has been no direct evidence linking the use of bisphosphonates to the occurrence of atypical fractures, although there are a number of case series and cohort analyses demonstrating an association.[28-30] Health Canada’s review of the evidence has shown a slightly increased risk of this type of fracture with bisphosphonate use; however, up to half of atypical fractures occur in people not exposed to bisphosphonates, complicating estimates of bisphosphonate-associated incidence.[27,29] Further complicating our understanding of atypical fractures is that they may be associated more with the presence of osteoporosis than with exposure to bisphosphonates.[31] While an association might exist, the risk must be kept in perspective. The absolute risk of bisphosphonate-associated atypical fracture is between 2 and 78 cases per 100 000 person-years. Thus, for every 100 fractures prevented, one atypical femur fracture may occur.[32]

Should my patient use denosumab instead of a bisphosphonate? Denosumab is a monoclonal antibody that inhibits the action if RANKL, a protein found on osteoclasts involved in cell replication and apoptosis. Denosumab works by preventing osteoclast maturation, which in turn decreases bone resorption, increases bone mineral density, and reduces the risk of fracture.[32,33] There are few studies that directly compare denosumab and oral bisphosphonates. From the available literature, denosumab appears to be superior at improving BMD when compared with bisphosphonates, but at 1 year the rates of fracture remain similar for the two therapies.[34] Denosumab is only available by subcutaneous injection given every 6 months, which may ensure adherence over a prolonged period, assuming the patient is motivated to continue with treatment. Overall, denosumab tends to be well tolerated, with the most frequent side effects being injection site irritation and back and joint discomfort.[35] Dyspepsia is not an issue. Rates of severe adverse reactions such as atypical femur fracture, osteonecrosis of the jaw, and hypocalcemia are low and thought to be similar to the rates for oral bisphosphonates.[36] Denosumab remains under patent at this time so cost is one of the barriers to this medication being used more frequently. If medication cost is not a concern, first-line therapy with denosumab should be considered in older patients who have difficulty with the dosing requirements of oral bisphosphonates or who have markedly impaired renal function.

Summary
Fracture prevention requires the clinician to recognize when a patient is at risk and know that this is not determined by bone mineral density alone. Furthermore, the clinician must recognize that a previous fragility fracture significantly increases the likelihood of a subsequent fracture and should be a trigger to act, ensuring that the first fracture is also the last fracture. By using the CAROC and FRAX tools a clinician can calculate risk accurately and have a meaningful discussion with the patient about the likelihood of a fracture occurring over the next 10 years. Both pharmacological and nonpharmacological treatments can then be considered, as shown in the case of Ms Stewart.

Although the evidence for calcium and vitamin D in preventing fractures is not overwhelming, calcium supplementation in patients who eat a calcium-sparse diet is reasonable. Likewise, vitamin D plays a crucial role in bone formation and remodeling, and supplementation should be considered in all patients at risk of fracture. Finally, bisphosphonates remain the most appropriate first-line therapy to prevent fracture despite questions about their use and concerns about side effects. Overwhelming evidence suggests bisphosphonates are safe, well tolerated, and effective at preventing fragility fractures. Evidence also supports considering denosumab as a first-line therapy in select patients. In all cases, it is imperative that patients at risk of fracture understand they are at risk and know the options available to help prevent fractures.

Competing interests
Authors did not receive any external sources of funding for writing this article. Dr Dian is on the advisory board at Amgen and Merck, and has been a speaker for the board at Amgen, Bayer, Lilly, Merck, and Servier.

This article has been peer reviewed.

Dr Budlovsky is a subspecialty resident in the Division of Geriatric Medicine at the University of British Columbia. Dr Dian is a clinical professor in the Division of Geriatric Medicine at the University of British Columbia and is on active staff at Vancouver General Hospital.