Osteoporosis Management

What we know

Osteoporosis is a growing public health concern with significant impact on quality and longevity of life; by age 80, women will have lost ∼30% of their peak bone mass. In 2019, the Endocrine Society had published a guideline on osteoporosis management, but since this publication, romosozumab was approved for use in the United States, Europe, and Canada. Romosozumab has a novel mechanism of action, incorporating both anabolic and antiresorptive properties by blocking the actions of sclerostin, an inhibitor of bone formation. A new guideline update was made to the Endocrine Society recommendations after a systematic review of randomized clinical trials evaluating the efficacy and safety of romosozumab, ^{1 –3} which also includes a more updated discussion on hormonal treatment options.

Guideline recommendations

Postmenopausal women may be stratified by level of fracture risk from low to very high. This fracture risk stratification influences the choice of initial agent and duration for treatment. Very high risk is defined as presence of T-score less than −2.5 with hip or spinal fractures or multiple vertebral fractures. Individuals with very high fracture risk should consider an anabolic agent for initial treatment, including romosozumab for 1 year, or teriparatide/abaloparatide used for 2 years. Use of anabolic therapy should immediately be followed by antiresorptive agents to sustain benefits. The bone mineral density (BMD) gains are more profound if the anabolic drug is administered before antiresorptive agents.

Those who have a high cardiovascular risk (i.e., past history of myocardial infarction or stroke) should avoid romosozumab pending further research. Those stratified as having a high fracture risk can also consider bisphosphonates for 3–5 years or denosumab for 5–10 years, with subsequent reassessment of fracture risk to determine duration of therapy. The recommendations also support the use of hormone therapy (HT) or selective estrogen receptor modulators (SERMs; raloxifene, bazedoxifene) in those who do not tolerate or are not appropriate candidates for other therapies. For those who have low venous thromboembolism risks, HT is a consideration, especially if initiated in individuals with bothersome vasomotor symptoms who are <60 years of age and within 10 years of menopause onset.

In those who are older than age 60 or who do not have bothersome menopausal symptoms, a SERM may be appropriate. Response to treatment may be assessed with repeat BMD measurement at 1–3 years, although bone turnover marker (BTM) monitoring can be considered as an alternative way to identify nonadherence or poor response to therapy (serum C-terminal crosslinking telopeptide for antiresorptive therapy or procollagen type N-terminal propeptide for bone anabolic therapy).

What this changes

The 2020 Endocrine Society guideline provides updated, clinically relevant advice for clinicians on the most appropriate treatment, duration, and monitoring of those at high or very high risk of fracture. The recommendations within this guideline are in line with other subsequently updated 2021 recommendations by the American College of Obstetrics and Gynecology (discussing importance of risk assessment) and by the North American Menopause Society (which includes expanded discussion on use of HT and SERMs for bone health). ^4,5 It is important to note that the routine use of BTMs is not recommended by other guideline recommendations. ⁵ The inclusion of romosozumab, HT, and SERMs to the armamentarium of osteoporosis treatments provides a greater number of options for women.

What we know

After discontinuation of denosumab, bone mass accrued is rapidly lost and BTM can increase to levels above baseline within 12 months, leading to decreased protection from vertebral fracture. Case reports of multiple vertebral fractures after discontinuing denosumab have been published. ⁶ Transitioning to an effective antiresorptive agent to maintain BMD is recommended after denosumab discontinuation. In clinical practice, it is not unusual to see delays when transitioning to a new therapy, especially during the coronavirus disease (COVID-19) pandemic. These studies explore bone loss after denosumab therapy and whether zoledronate use can prevent this loss of BMD.

Study results

Lyu et al. performed a population-based cohort study of a national United Kingdom primary care database, assessing the outcomes related to variation in denosumab injection timing, between 2010 and 2019. The fracture outcomes of three different injection intervals were compared, including on-time (within 4 weeks of recommended dose), short delay (4–16 weeks from suggested date), and long delay (16 weeks–6 months late). Multiple covariates that impact bone health were assessed, and 10-year risks for osteoporosis fracture were calculated using a validated algorithm. Of the final cohort of 2,594 patients, the average age was 76 years, 94% were women, and overall, there was a low rate of comorbid conditions.

However, the patients were at very high risk of fracture, with a mean calculated 10-year osteoporotic fracture risk of 22%. At 6 months, the cumulative risks for vertebral fracture were 2.2, 3.6, and 10.1 in 1,000, respectively, for on-time, short delay, and long delay dosing. The corresponding increases in vertebral fracture risk with short delay was hazard ratio (HR) of 1.48 (confidence interval [CI], 0.58–3.79) versus 3.91 with long delay (CI, 1.62–9.45). There was no difference in fracture risk for hip or nonvertebral fractures.

Sølling et al. sought to determine if treatment with zoledronate could blunt BMD losses after denosumab discontinuation, and whether timing of zoledronate therapy had an impact on this outcome. The study was a randomized, open-label interventional study of 61 osteopenic patients who were an average of 67 years of age (85% women). This was the largest randomized controlled trial to assess bone loss after denosumab to date. Patients received a single zoledronate 5 mg infusion 6 or 9 months after the last denosumab injection, followed by monitoring of bone density and BTM. The primary endpoints were changes in lumbar spine BMD at 6 months after infusion of zoledronic acid.

The authors also report the changes in BMD from baseline to 12 months after the last denosumab injection; the study is ongoing. At 6 months, the lumbar spine BMD decreased significantly by 2.1% in the 6-month versus 4.3% in the 9-month infusion group. At 12 months, the decrease in lumbar spine BMD was 4.8% in the 6-month group versus 4.1% for those who had a zoledronate infusion at 9 months after discontinuation of denosumab. Incident vertebral fractures were documented in two women in the 9-month group. Zoledronate infusion did not prevent bone loss completely in all individuals of either group.

What this changes

These studies highlight the importance of adhering to timely administration of denosumab injections, followed by appropriate use of an antiresorptive agent soon after discontinuation. Because of the increased risk of bone loss and fracture with cessation of treatment, delaying the next denosumab dose by more than 4 months was associated with increased risk for vertebral fracture compared with on-time injection. A single dose of zoledronate 5 mg does not appear to be enough to deter bone loss completely in all individuals, whether the infusion occurs at 6 or 9 months after denosumab discontinuation. Because an average 0.25–0.5 standard deviation loss of BMD can be seen, consideration can be given to treat with denosumab to a target treatment BMD above this range before stopping treatment.

What we know

Vitamin K antagonists (warfarin) and direct oral anticoagulants (DOACs), which include a thrombin inhibitor (dabigatran) and factor Xa inhibitors (rivaroxaban, apixaban, and edoxaban), are the first-line choices for long-term anticoagulation treatment in patients with atrial fibrillation (AF). Anticoagulants have been found to increase the risk of osteoporotic fracture, although head-to-head comparisons of risk are unknown.

Study results

This population-based cohort study compared the osteoporotic fracture risk among those with newly diagnosed AF receiving a novel anticoagulant prescription between 2010 and 2017. After excluding patients with transient AF, those younger than 18 years, and those with a history of seizures or bone tumors, previous anticoagulant or HT use, the resultant cohort of patients (n = 23,515, mean age 74.4 years) was stratified by anticoagulant type and evaluated for the composite incidence of nontraumatic hip and vertebral fracture. Analysis through Cox proportional hazard models demonstrated a reduced risk of osteoporotic fracture in DOAC users compared with warfarin users (HRs, 0.62 [CI, 0.41–0.94] for apixaban compared with warfarin, 0.65 [CI, 0.49–0.86] for dabigatran compared with warfarin, and 0.52 [CI, 0.37–0.73] for rivaroxaban compared with warfarin).

Although these results remained consistent within both sexes among the analyses, women were found to have a higher overall incidence of osteoporotic fractures than men, regardless of anticoagulant type. There was no statistically significant difference in fracture risk between the different types of DOACs studied.

What this changes

The results of this study echo similar findings found in several previous studies, indicating that the use of DOACs, regardless of the type, confers a lower risk of osteoporotic fractures than warfarin. ^{7 –9} Although limited in power, this study also suggests that there is no significant difference in fracture risk between apixaban, dabigatran, and rivaroxaban. Clinicians should consider that osteoporotic fracture risk may be reduced by choosing a DOAC instead of warfarin for anticoagulation therapy, especially because patients with AF have a higher incidence of hip fracture than those without. ¹⁰

Tamblyn R, et al. Multinational Investigation of Fracture Risk with Antidepressant Use by Class, Drug, and Indication. J Am Geriatr Soc. 2020;68(7):1494–1503.

What we know

Antidepressants are commonly prescribed in the elderly population and often used for extended periods of time. Although selective serotonin reuptake inhibitors (SSRIs) and serotonin and norepinephrine reuptake inhibitors (SNRIs) are often thought to be safer compared with other antidepressants such as tricyclic antidepressants (TCAs), they have also been associated with increased fall risk (likely due to central nervous system side effects), fracture risk, and bone loss. ^11,12 There has been limited data on head-to-head fracture risk comparisons among different types of antidepressant medications.

Study results

In this multicountry retrospective cohort study conducted between 2009 and 2014, data were extracted from electronic health records of five participating jurisdictions (Quebec, Montreal, Boston, United Kingdom, Taiwan) to identify patients aged 65 years and older treated with antidepressants and evaluate for a primary outcome of fracture incidence. Approximately half of the patients were aged 75 years and older. Multimodal Cox proportional hazard analyses revealed that despite considerable variation among medication choices between jurisdictions, the use of any antidepressant was found to convey a statistically significant increase in fracture risk. In North America, SSRIs were the predominant medications used compared with TCAs in the United Kingdom and Taiwan.

Several antidepressants were found to increase fracture risk in some, but not all jurisdictions, including SSRIs (citalopram [HR 1.23 (in Montreal), 1.43 (in Quebec); CI, 1.11–1.36 and 1.11–1.84], sertraline [HR 1.36; CI, 1.10–1.68]) and TCAs (doxepin [HR 1.36; CI, 1.00–1.86], imipramine [HR 1.16; CI, 1.05–1.28]). Although there was no increase in fracture risk found with the SNRI class as a whole, duloxetine was independently associated with a statistically significant increase in fracture risk compared with venlafaxine (HR 1.41; CI 1.06–1.88).

Amitriptyline was either protective or not significantly associated with fractures, except in the United States where higher doses were more commonly prescribed. The SSRIs were associated with a greater fracture risk than TCAs in all jurisdictions, except Taiwan. Although atypicals (i.e., bupropion, mirtazapine, trazodone) as a class were found to increase fracture risk (HR 1.34; CI, 1.14–1.58), results among specific medications varied significantly, making it difficult to draw any conclusions.

What this changes

Antidepressants are known to increase the risk of bone loss, falls, and fractures in the elderly population, who are already at high risk for these potentially debilitating events. The American Geriatric Society (AGS) Beers Criteria, which is utilized to identify medications inappropriately dangerous to geriatric patients, was updated in 2019 to caution use of SSRIs, SNRIs, and TCAs in patients with a history of falls or fractures. ¹² It does not, however, include guidance on which specific antidepressants confer a comparatively higher or lower osteoporotic fracture risk. Within classes, fracture risk for patients may be reduced by selecting an SSRI other than citalopram or sertraline, selecting the SNRI venlafaxine over duloxetine, and using the TCA amitriptyline instead of imipramine or doxepin, when these medications are deemed necessary and appropriate. Using lower dosages of these medications may additionally decrease fracture risk. ¹³

What we know

Bisphosphonates' mechanism of action involves inhibiting osteoclast-mediated bone resorption and thus bone remodeling. Multiple large, randomized clinical trials demonstrated the effectiveness of bisphosphonates in improving BMD, leading to their widespread use. However, reports of a possible correlation between prolonged use of bisphosphonates and low-energy femoral subtrochanteric and shaft fractures (atypical femoral fractures, AFF) have caused concern about their long-term safety and decreased prescribing among clinicians. ^14,15 This study aimed to determine the risk of AFF in correlation with the duration of bisphosphonate use and whether other clinical factors impact this effect.

Study results

This study reviewed the health records of women aged 50 years or older who received bisphosphonates within the Kaiser Permanente Southern California health care system, evaluating the primary outcome of AFF from 2007 to 2017. Potential risk factors explored included age, race/ethnicity, glucocorticoid use, and duration of cumulative bisphosphonate use. Data were obtained from electronic health records with fractures radiographically adjudicated. Multivariable Cox models were used to outline hazard risk ratios for the association between each potential risk factor and atypical fracture.

A total of 196,120 women made up the analytic cohort of those who used bisphosphonates, 59.5% being aged 65 years or older. The cohort was 53.3% White, 24.0% Hispanic, 13.5% Asian, and 5.9% Black. There were 277 AFF (1.74 fractures per 10,000 patient-years) and 9,102 hip fractures (58.9 fractures per 10,000 patient-years) that occurred within this cohort; the highest rates of AFF were in age groups 65–74 years and 75–84 years (2.24 and 2.35 per 10,000 person-years, respectively). Asian women had increased rates of AFF but lower rates for hip fracture compared with other racial/ethnic groups.

Other factors such as short height, heavier weight, and glucocorticoid use of 1 year also increased risk. Prolonged bisphosphonate use of 8 or more years of exposure increased AFF risk (HR 43.51; CI:13.70–138.15). Rates of AFF decreased quickly after discontinuation. However, the overall fracture risks versus benefits of bisphosphonate treatment strongly favored their use based on calculated modeling of benefits. Among Whites, 149 hip fractures and 541 clinical fractures (vertebral and nonvertebral) were prevented after 3 years, although 2 AFFs occurred. Among Asians, 91 hip and 330 other clinical fractures were prevented, whereas 8 AFFs occurred during the same duration of treatment.

What this changes

This study demonstrated a higher risk of AFF with prolonged bisphosphonate usage, especially if used for 5 years or more, but this risk was expeditiously lowered after discontinuation. Asian women had a heightened risk for AFF with bisphosphonate use, although benefits of treatment still outweighed risk for overall fracture reduction. Consistent with guideline recommendations, patients should be offered bisphosphonate treatment for osteoporosis, but clinicians should be aware of risk factors that may heighten AFF risk.