Osteoporosis: To Treat or Not to Treat?

Maria Rita Talla; Stephen J. Gallacher


Age Ageing. 2020;49(3):327-328. 

'To treat or not to treat'—that is the big question for fracture prevention strategies. Over recent decades, evidence has accumulated for the efficacy of drug treatment in patients who are at high risk of fracture which has helped and inform national and international guidelines.[1]

Throughout this time, the definition of osteoporosis has remained constant: bone mineral density (BMD) T-score < –2.5.[2] This cut-off has of course been used to determine the diagnosis. This definition was never meant to be a treatment intervention threshold, however, it was subsequently used in most clinical trials evaluating anti-osteoporosis drug treatments. There are a few exceptions to this. Trials of drug therapies have been carried out in patients with vertebral fractures without 'osteoporotic' BMD levels with oral alendronic acid[3] and risedronate,[4] and in patients' post-hip fracture with intravenous zoledronic acid.[5] In addition, a recent study has demonstrated the anti-fracture efficacy of IV zolendronic acid in preventing fractures in a population of women over the age of 65 with osteopenic BMD.[6]

These various trials have led to confusion around whether or not a given patient has osteoporosis and whether or not treatment should be offered to reduce their fracture risk. To a large extent, the terminology we use is to blame here. It is clear that the World Health Organization (WHO) T-score definition is arbitrary and does not reflect underlying bone pathophysiology. It is also clear that there are people who are at high future fracture risk who may have pathologically abnormal bone but in whom the WHO T-score definition does not apply. What we have here is the term 'osteoporosis' meaning two different things depending on the context. Perhaps in retrospect, it might have been better for the WHO definition to have used different terminology—perhaps even the phrase 'densitometric osteoporosis' might have been better with 'osteoporosis' being used to define people at higher fracture risk (whatever threshold might be set for this)—there of course being significant overlap in these two populations.

The important point is how we make treatment recommendations taking into account these definitions. This is where understanding the treatment evidence base becomes important. A view is often expressed that all patients at high fracture risk should be offered drug therapy often irrespective of their measured BMD. This may sound attractive since we know that the majority of low trauma fractures occur in women with non-osteoporotic, generally osteopenic, BMD. Whilst this may appear appealing, as a concept this fails to recognise the treatment evidence base as it stands.

The risedronate Hip Intervention Program (HIP) trial[4] has tested this idea. In this trial, one group of women who were over age 80 and were defined as being at high fracture risk were treated solely on the basis of their fracture risk (and not BMD—BMD was only measured in a small subset). A second group of younger women (age 70–79) were treated on the basis of having a low measured BMD. In this trial, a significant reduction in hip fracture risk was observed with treatment only in the younger cohort. Patients at high fracture risk treated on the basis of presence of risk factors for hip fracture alone did not see any anti-fracture benefit with drug therapy.

A similar result has been seen in patients treated with alendronic acid. In the Fracture Interventional Trial (FIT),[3] patients were included if BMD T-score was <–2.5. During the trial, densitometric BMD was re-defined meaning that a small subgroup of this trial population was included with 'osteopenic' BMD. A post-hoc analysis of this trial showed that no anti-fracture benefit was observed in women with 'osteopenia'.

The risedronate HIP and FIT trials provide evidence to show that in general anti-osteoporosis drug therapies should be reserved for women with densitometric osteoporosis.

There are exceptions to this statement and this is where some of the confusion around how osteoporosis is defined comes into play. The most important exception relates to patients who have a vertebral fracture(s). In clinical trials,[3,7] patients have been treated with anti-osteoporosis drug therapies without requiring 'osteoporotic' BMD measures. These trials have generally all shown anti-fracture benefit with treatment—hence the patient with vertebral fracture is an exception and should always be considered for anti-osteoporosis treatment.

There may be a similar argument with respect to the 'hip fracture' patient. However, to date, only one trial with IV zoledronic acid has demonstrated anti-fracture benefit in this population.[5]

The recent work of Reid et al.[6] might be thought to challenge much of this viewpoint. This trial studied the use of IV zoledronic acid (given as four infusions over a 6-year period) in a cohort of women whose bone density was generally 'osteopenic'. This population had a calculated average 10-year risk of major fracture using FRAX of 12%. This trial showed significant anti-fracture benefit with this treatment regimen.

The key principle behind all treatment decisions should be an understanding and application of the evidence base. At this point in time starting an anti-osteoporosis therapy should be based upon:

  • Finding people who may be at high risk of fracture. This will be based on the presence of clinical risk factors, which may or may not include previous fracture.[1]

  • Once an individual with fracture risk factor(s) has been identified, then a formal fracture risk calculation should be carried out. Various tools are available to do this such as FRAX or QFracture.

  • Individuals at high fracture risk—a risk threshold should be agreed within any given health economy—should then go forward for measurement of BMD.

  • Individuals then found to have BMD < –2.5 should be considered for anti-osteoporosis therapy.

  • Women over age 65 with osteopenic BMD and a 10-year risk of major fracture of around or in excess of 12% should be considered for a programme of IV zoledronic acid.

This pathway applies the treatment evidence base as it stands and importantly avoids using treatments where the evidence of benefit may be lacking.