Bone mineral density assessment by dual-energy X-ray absorptiometry is the most widely available and best validated clinical tool to diagnose osteoporosis, assess fracture risk, and monitor the skeletal effects of medications that reduce fracture risk.1 However, dual-energy X-ray absorptiometry testing remains underused with osteoporosis undertreated. This treatment gap has reached crisis proportions,2 with recent disturbing evidence that hip fracture rates in the United States have stopped their previously observed decades-long decline.3 Treatment with bisphosphonates and patient adherence to medications after fracture remain low and show no signs of improvement.4 It is in this setting that the American College of Physicians (ACP) recently released updated osteoporosis treatment guidelines that included a recommendation “against bone density monitoring during the 5-year pharmacologic treatment period for osteoporosis in women,” although a “weak recommendation; low quality evidence.”5 We fear this recommendation is a disservice to patients and likely will cause more harm than good, with potential to worsen the osteoporosis treatment gap. The ACP Guidelines are in direct conflict with those of the National Osteoporosis Foundation and American Association of Clinical Endocrinologists that advise monitoring with dual-energy X-ray absorptiometry during treatment.6, 7 In addition, the ACP guidelines confuse the literature on monitoring untreated versus treated patients with osteoporosis.
We believe serial monitoring can assist clinicians during pharmacologic treatment of osteoporosis. Although there may be fracture risk reduction in treated women without gains in bone density,8 data also show those who gain bone density while on treatment have greater risk reduction than those who lose bone density.9 In addition, larger increases in bone density are associated with greater reductions in fracture risk.10 The ACP guidelines fail to cite or cite but fail to acknowledge several studies of antiresorptive and anabolic therapy demonstrating these findings.
Miller and colleagues11 examined fracture risk with oral and intravenous ibandronate use. With oral ibandronate, lumbar spine bone density increase at 3 years was associated with fracture rates (relative risk reduction at year 3 for 1% change from baseline: hip, 7.9%; 95% confidence interval [CI], 2.1-13.5 P = .0084]; lumbar spine, 4.7%; 95% CI, −0.1 to 9.3; P = .0565). With intravenous use, lumbar spine bone density increases at years 2 and 3 were significantly associated with vertebral fracture rate (relative risk reduction at year 3 for 1% change from baseline: hip, 11.6%; 95% CI, 7.0-16.0; P < .0001]; lumbar spine, 6.9%; 95% CI, 2.9-10.6; P = .0008). Changes in bone density explained a 23% to 37% of the antifracture effect at 2 and 3 years.11
The ACP guidelines cite an analysis of the Fracture Intervention Trial as evidence to support not monitoring. Although this study found similar fracture risk reductions for women who gained bone density or lost 0% to 4%, this was no longer the case when women lost more than 4%.12 This is an important group to identify because they may require further evaluation or alternate therapy.
Likewise, in a study not cited in the guidelines, Jacques et al13 evaluated the correlation between bone density changes with zoledronic acid fractures fracture rates in 7736 postmenopausal women with osteoporosis in the HORIZON-Pivotal Fracture Trial. They found that the change in total hip bone density over 3 years explained 40% of the vertebral fracture reduction and 61% of the nonvertebral fracture reduction. In the same study, women treated with zoledronic acid who lost bone density over 3 years had a 7.5% incidence of new vertebral fracture, whereas women who gained more bone density than the median (0.032 g/cm2) had an incidence of 2.0%.
Denosumab, a fully humanmonoclonal antibody to RANKL, was Food and Drug Administration approved for the treatment of osteoporosis in 2010. In the FREEDOM trial, changes in total hip bone density induced by denosumab explained 35% and 87% of the treatment effect at 3 years for new or worsening vertebral fractures and nonvertebral fractures, respectively.14 These findings were subsequently confirmed in the extension trial in which fracturereductions observed in year 4 and beyond were correlated with hip bone density attained after 3 years of denosumab.15
In a study of teriparatide, increases in bone density accounted for 30% to 41% of fracture risk reduction.16However, improvements in osteoporosis with anabolic therapy may be related to additional changes in bone strength. Adjunct software is now available with dual-energy X-ray absorptiometry to calculate trabecular bone score, a measure of bone quality, and can be done without exposing the patient to additional radiation.
Last, although the new guidelines only make recommendations for monitoring of women on pharmacologic therapy, the bulk of their justification actually cites studies that involve screening dual-energy X-ray absorptiometry. They cite the Bell study in which bone density at 3 years did not correlate with fracture change.17However, the data in the text appear to be from the study by Gourlay et al,18 which examined the transition time from normal to low bone density for women not receiving pharmacologic therapy. They refer to the study by Berry et al19 in 2013, which also was a serial screening study. Therefore, the bulk of the text supporting recommendations on monitoring women on pharmacologic therapy refers to a different patient population.
Because patients are commonly started on treatment because of low bone mineral density, they are anxious to see follow-up dual-energy X-ray absorptiometry results, just as patients started on antihypertensive medication would like to see what happens to blood pressure with treatment. Who among the ACP Guidelines Committee would want to take a medication for 5 years with no information on treatment effect during that time? Data do show a correlation between the magnitude of bone mineral density increase and reduction in fracture risk, and a significant loss of bone density on treatment warrants further investigation and consideration of alternate therapies. Bone mineral density testing by dual-energy X-ray absorptiometry testing remains inexpensive, noninvasive, and clinically useful to monitor patients treated for osteoporosis.
To read this article in its entirety please visit our website.
-Micol S. Rothman, MD, E. Michael Lewiecki, MD, Paul D. Miller, MD
This article originally appeared in the October 2017 issue of The American Journal of Medicine.
