Secular Trends in the Pharmacologic Treatment of Osteoporosis and Malignancy-Related Bone Disease from 2009 to 2020

Abstract

Background

New bone-directed therapies, including denosumab, abaloparatide, and romosozumab, emerged during the past decade, and recent trends in use of these therapies are unknown.

Objective

To examine temporal trends in bone-directed therapies.

Design

An open cohort study in a US commercial insurance database, January 2009 to March 2020.

Participants/Interventions

All-users of bone-directed therapies age >50 years, users with osteoporosis, users with malignancies, and patients with recent (within 180 days) fractures at key osteoporotic sites.

Main Measures

The percentage of each cohort with prescription dispensing or medication administration claims for each bone-directed therapy during each quarter of the study period.

Key Results

We analyzed 15.48 million prescription dispensings or medication administration claims from 1.46 million unique individuals (89% women, mean age 69 years). Among all users of bone-directed therapies, alendronate, and zoledronic acid use increased modestly (49 to 63% and 2 to 4%, respectively, during the study period). In contrast, denosumab use increased rapidly after approval in 2010, overtaking use of all other medications except alendronate by 2017 and reaching 16% of users by March 2020. Similar trends were seen in cohorts of osteoporosis, malignancy, and recent fractures. Importantly, use of any bone-directed therapy after fractures was low and declined from 15 to 8%.

Conclusions

Rates of denosumab use outpaced growth of all other bone-directed therapies over the past decade. Treatment rates after osteoporotic fractures were low and declined over time, highlighting major failings in osteoporosis treatment in the US.

INTRODUCTION

Osteoporosis is highly prevalent, affecting approximately 7–12 million adults in the United States (US).^1,2 Osteoporotic fractures are also common, affecting approximately 50% of women and 20% of men over the age of 50 years during their lifetimes.^3,4 Osteoporosis accounts for more than two million fractures annually in the US,¹ and these fractures are in turn associated with substantial healthcare costs, disability, and mortality.^3,5,6,7,8,9 Despite the substantial personal and societal consequences of osteoporosis and the availability of numerous safe and effective medications, osteoporosis remains undertreated, even after major complications such as hip fracture.^10,11,12,13 Further to this, rates of undertreatment have increased in recent years following media coverage of anti-resorptive side effects, such as atypical femur fractures and osteonecrosis of the jaw.^{11,12,13,14,15,16}

Although undertreatment of osteoporosis is well described, little is known about the changes in types of osteoporosis medications used in the US over time. In the past decade, several new medications—denosumab, abaloparatide, and romosozumab—have been approved for osteoporosis treatment (Appendix Tables 1 and 2 in the Supplementary information). Meanwhile, patents for other medications, including alendronate, zoledronic acid, raloxifene, and teriparatide, have expired during this timeframe. Finally, increasing evidence supports the use of anti-resorptive therapy for the prevention of fractures in patients with metastatic bone disease.^17,18,19

Given the evolution of bone-directed therapies for osteoporosis and malignancy over the past decade, we examined secular trends in the use of bone-directed therapies in a commercial insurance database of US adults between 2009 and 2020.

METHODS

Data Source

We used claims data from Optum’s de-identified Clinformatics® Data Mart Database.

The data source contains longitudinal claims from members of commercial and Medicare Advantage health plans and includes outpatient and inpatient claims, diagnosis codes, procedure codes, and pharmacy dispensings. Approximately 77.6 million individuals are included, with representation from across the US. Data are de-identified prior to analysis, so informed consent was not required. The study was approved by the Institutional Review Board of Brigham and Women’s Hospital (2011P002580_186).

Study Cohorts

The primary analysis included all individuals over 50 years old who received any bone-directed therapy for any indication between January 1, 2009, and March 31, 2020 (“the primary cohort”). Two subgroup analyses were performed, limiting this cohort to those with osteoporosis based on International Classification of Diseases (ICD)-9 or ICD-10 codes and without malignancy (“the osteoporosis subgroup”) or to those with active malignancy as marked by ICD-9 or ICD-10 codes for malignancies likely to metastasize to bone as placed by oncology providers (“the malignancy subgroup”) within the current or previous quarters relative to the dispensing or administration of bone-directed therapy (Appendix Table 3 in the Supplementary information). These were open cohorts, allowing individuals to enter and leave the cohort quarterly based on changes in prescribing of bone-directed therapy, disenrollment, or death. A pre-planned secondary analysis examined rates of treatment and non-treatment in all women over age 50 years old and men over age 60 years old with ICD-9, ICD-10, or Current Procedural Terminology (CPT) codes for fracture at key osteoporosis sites (vertebrae, hip, wrist, humerus; the “fracture cohort”) in the prior 180 days and who did not have evidence of active malignancy using the criteria above. Sensitivity analyses were performed on the primary cohort to examine bone-directed therapies by age decile (e.g., 50–59, 60–69, 70–79, and >80 years old) and counting treatment effect of denosumab for 2 calendar quarters and zoledronic acid for 4 quarters in individuals with continued enrollment in the database.

Statistical Analyses

We identified all dispensings or administrations of bone-directed therapy (i.e., alendronate, zoledronic acid, other bisphosphonates (ibandronate, pamidronate, risedronate, considered as a single category due to low overall use), raloxifene, denosumab, parathyroid hormone/parathyroid hormone–related protein (PTH/PTHrP) analogs, and romosozumab) between January 1, 2009, and March 31, 2020. For the primary, subgroup, and sensitivity analyses, we calculated the percentage of the study cohort (all of whom used at least one bone-directed medication) with any use of each medication for each calendar quarter of the study period. For the fracture secondary analysis, we calculated the percentage of individuals who were treated with each medication or who were not treated for each calendar quarter of the study period. Analyses were conducted using SAS (Statistical Analysis Software), version 9.4 (SAS Institute Inc., Cary, NC, USA).

Role of the Funding Source

This work was supported by internal funding of the Division of Pharmacoepidemiology and Pharmacoeconomics, Brigham and Women’s Hospital. The authors had complete control over the design, conduct, analysis, and decision to submit the manuscript for publication.

RESULTS

In the primary cohort, a total of 15.48 million unique prescription dispensings or medication administration claims from 1.46 million unique individuals during the study period were analyzed (for reference, there were 4.5 million to 6.7 million individuals older than 50 years enrolled in the database during the study period). Of the 1.46 million individuals receiving bone-directed therapy, 89% were women and 71% were over the age of 65 years old, with a mean age of 69 years. Alendronate was the most common medication used in the primary cohort, with use increasing over time from 49 to 63% by the end of the study period (Fig. 1). Percent of users receiving zoledronic acid increased slightly from 2 to 4%. Use of other bisphosphonates declined steadily from 38 to 10%. By comparison, after its approval in 2010, denosumab use increased rapidly, overtaking use of every other medication except alendronate by early 2017 and reaching 16% of users by the end of the study period. The percentage of individuals treated with raloxifene peaked at 13% in 2012 but declined to 6% by the end of the study period. Use of teriparatide, abaloparatide, and romosozumab remained less than 2% throughout the study period.

Figure 1

Utilization trends of bone-directed treatments in all users in a US commercial insurance database between January 2009 and March 2020. PTH, parathyroid hormone; PTHrP, parathyroid hormone–related protein. Other bisphosphonates include risedronate, ibandronate, and pamidronate.

Two sensitivity analyses were performed in the primary cohort. We first examined use of bone-directed therapies by age decile and found that temporal trends seen in the primary cohort, including the rise in denosumab use, were reproduced in each age decile (Appendix Figure 1 in the Supplementary information). We then counted treatment effect of denosumab for 2 quarters and zoledronic acid for 4 quarters in patients with continued enrollment and found that relative use of zoledronic acid increased slightly (from 2 to 10%) during the study period, but was still outpaced by growth in denosumab from 0 to 25% (Appendix Figure 2 in the Supplementary information).

In the strictly defined osteoporosis subgroup, a total of 7.93 million unique prescription dispensings or medication administration claims from 948,546 unique individuals during the study period were analyzed. Of these, 92% were women and 76% were over the age of 65 years old, with a mean age of 70 years. Trends in the use of alendronate, zoledronic acid, other bisphosphonates, denosumab, raloxifene, PTH/PTHrP analogs, and romosozumab in the osteoporosis cohort paralleled those in the primary cohort (Fig. 2).

Figure 2

Utilization trends of bone-directed treatments in treated osteoporosis patients in a commercial insurance database between January 2009 and March 2020. PTH, parathyroid hormone; PTHrP, parathyroid hormone–related protein. Other bisphosphonates include risedronate, ibandronate, and pamidronate.

In the strictly defined malignancy subgroup, a total of 945,218 unique prescription dispensings or medication administration claims from 131,557 unique individuals during the study period were analyzed. Of these, 77% were women and 67% were over the age of 65 years old, with a mean age of 68 years. At the beginning of the study period, alendronate, zoledronic acid, and other bisphosphonates were used in 35%, 28%, and 32% of the cohort, respectively (Fig. 3). By the end of the study period, relative use of each of these medications had declined to 24%, 23%, and 5% respectively. In contrast, denosumab use increased rapidly to 18% in January 2012 and 46% by the end of the study period. Use of raloxifene was 7% at the beginning of the study period and 2% by the end of the study period. Use of PTH/PTHrP analogs and romosozumab remained less than 1% throughout the study period.

Figure 3

Utilization trends of bone-directed treatments in treated patients with malignancies likely to metastasize to bone in a commercial insurance database between January 2009 and March 2020. PTH, parathyroid hormone; PTHrP, parathyroid hormone–related protein. Other bisphosphonates include risedronate, ibandronate, and pamidronate.

In the fracture secondary analysis, 1.35 million unique individuals were analyzed, of whom 72% were women and 79% were over the age of 65 years old (mean age 74 years). The majority (>85%) of patients who sustained recent osteoporotic fractures received no bone-directed therapy, and the percentage of patients receiving any treatment in the 90–180 days after fracture decreased from 15% at the beginning of the study period to 8% by the end of the study period (Fig. 4). Absolute use of alendronate, zoledronic acid, other bisphosphonates, raloxifene, and PTH/PTHrP analogs declined, reflecting the overall decline in treatment rates, while absolute use of denosumab increased from 0 to 2% during the study period.

Figure 4

Utilization trends of bone-directed treatments in treated and untreated patients with osteoporotic fractures in a commercial insurance database between January 2009 and March 2020. PTH, parathyroid hormone; PTHrP, parathyroid hormone–related protein. Other bisphosphonates include risedronate, ibandronate, and pamidronate.

Among the patients with a fracture who received treatment in the fracture analysis, relative use of alendronate and zoledronic acid slightly increased over time, from 53% and 3% at the beginning of the study period to 57% and 4% by the end of the study period, respectively (Appendix Figure 3 in the Supplementary information). Use of other bisphosphonates and raloxifene declined over time, from 36% and 7% at the beginning of the study period to 9% and 4% at the end of the study period, respectively. Use of PTH/PTHrP analogs increased slightly from 3 to 6% during the study period, and use of denosumab increased from 0 to 21% during the study period.

DISCUSSION

During the past decade, relative utilization of alendronate and zoledronic acid have slightly increased as measured by percent of all patients treated with bone-directed therapy, while other bisphosphonates and raloxifene have declined. Meanwhile, the relative share of patients receiving denosumab as bone-directed therapy has increased since its introduction in 2010, overtaking use of every other medication except alendronate by early 2017. Use of PTH/PTHrP analogs and romosozumab remains low but increased slightly in the fracture cohort. Importantly, treatment of any kind for those with recent fractures at common osteoporotic sites remains very low (<15%) and has declined over the past decade, indicating a significant gap in the care of patients with osteoporotic fractures.

During the study period, osteoporosis treatment guidelines recommended the use of bisphosphonates as first-line therapy, with consideration of denosumab as an acceptable “alternative initial treatment.” ^20,21,22 Recently, there has also been increasing consideration of PTH/PTHrP analogs as initial therapy, especially in cases of advanced bone loss or high risk of fracture,^20,23 due to their osteoanabolic characteristics, significant fracture prevention, and concern that their benefit may be blunted after use of anti-resorptive agents.^23,24 In accordance with the guidelines, alendronate and zoledronic acid relative use increased slightly over time, although denosumab use increased to a greater degree. Use of other bisphosphonates declined, possibly related to patent expirations or meta-analyses and observational data suggesting inferior fracture prevention compared to alendronate and zoledronic acid.^25,26 Uptake of PTH analogs and romosozumab remained low, despite randomized controlled trials demonstrating the superiority of teriparatide (a PTH analog)²⁷ and romosozumab²⁸ versus bisphosphonates for reducing vertebral, non-vertebral, and clinical fractures. This lower uptake may be related to higher medication costs (Appendix Table 1 in the Supplementary information), concerns regarding serious side effects (osteosarcoma for teriparatide, cardiovascular events for romosozumab), or debate surrounding these medications’ relative efficacy in prevention of hip fractures. Additionally, the later approval years of abaloparatide and romosozumab (2017 and 2019, respectively) may limit their adoption and guideline recommendations, as well as our ability to detect uptake during the study period.

Regarding the rise in use of denosumab, there were several major studies which may have influenced physician prescribing during this period. Between 2010 and 2012, denosumab was found to be superior to zoledronic acid for prevention of skeletal-related events in patients treated for bone metastases from a variety of solid organ malignancies.^19,29,30,31 For the treatment of postmenopausal osteoporosis, a randomized trial published in 2016 found that denosumab resulted in greater increases in bone mineral density than zoledronic acid.^32,33 No difference in fracture rates was seen in this study, although data suggest that greater improvements in bone density are associated with greater reduction in fracture rates.³⁴ A meta-analysis of 10 randomized controlled trials confirmed that denosumab resulted in greater increases in bone mineral density than bisphosphonates including alendronate (n=7), risedronate (n=1), and zoledronic acid (n=2), although only one trial showed reduction in osteoporotic fractures with denosumab versus alendronate.³⁵ However, in 2017–2018, numerous studies documented an increase in vertebral fractures following abrupt discontinuation of denosumab, ^36,37,38,39 such that recent guideline updates recommend starting an alternate bone-directed therapy, generally a bisphosphonate, upon discontinuation of denosumab.⁴⁰ However, results from early trials to prevent rebound fracture after denosumab discontinuation are conflicting, suggesting that even following denosumab with a bisphosphonate may not completely negate the risk of rebound vertebral fractures,^{41,42,43,44,45} and questions remain regarding the ideal type, timing, and duration of alternate therapy after denosumab discontinuation.

Our results demonstrate growth in use of denosumab, outpacing other bone-directed therapies, with neither clear jumps in prescribing following positive studies nor clear declines in prescribing following reports of safety concerns (Appendix Table 2 in the Supplementary information). Similarly, data suggesting decreased all-cause mortality in men and women treated with zoledronic acid after hip fracture⁴⁶ and improved disease-free survival with zoledronic acid in select patients with breast cancer¹⁸ did not associate with clear rises in zoledronic acid use in the osteoporosis or malignancy subgroups, respectively. In the malignancy subgroup as in the osteoporosis subgroup, there was a dramatic uptake of denosumab, reaching 18% by 2012 and 46% by 2020, following regulatory approval for this indication in 2011 and the publication of evidence for equivalent or superior prevention of skeletal-related events, relative to bisphosphonates, in breast cancer, prostate cancer, and lung cancer.^17,18,19 Similarly, Gupta et al. observed a rapid rise in denosumab use for multiple myeloma in Medicare claims after the approval of denosumab for prevention of skeletal-related events in multiple myeloma in 2018⁴⁷. The rapid rise in denosumab uptake may be influenced by many factors, such as advertising, physician or pharmacy incentives,^48,49 improved medication adherence, patient or insurance plan preferences, ease of biannual subcutaneous administration in a clinic setting, or better tolerability (i.e., no flu-like symptoms, unlike those observed after zoledronic acid administration). However, denosumab was more expensive than many options available during this period, with estimated annual median standard drug-related costs per patient for bone-directed therapies ranging from $3082 for denosumab to $368 for zoledronic acid to $113 for alendronate in the Optum data set in 2020 (prior to accounting for drug administration costs; Appendix Table 1 in the Supplementary information). Further to this, other parenteral medications, such as zoledronic acid, share many of the benefits of denosumab (e.g., convenience of annual infusions, non-reliance on adherence to oral therapy) and did not see as steep a rise in use during this period.

Given the already very low rates of treatment of osteoporosis, the decline in treatment rates for patients with a recent fracture in this study and others^11,12,13 is especially concerning. An older, widely publicized study showed that after years of rising rates of hip fracture, these rates stabilized and then began to decline starting in approximately 1995,⁵⁰ the year alendronate was approved in the US. However, more recent studies have found either plateauing or increasing fracture rates, especially at the hip, beginning in approximately 2013^51,52,53 and declines in femoral neck bone density in US adults over age 50 beginning in approximately 2007.² It is possible that the decline in treatment for osteoporosis has contributed to rising rates of hip fractures in recent years. Further to this, our study shows that 92% of individuals receiving bone-directed therapy in the primary cohort were women, possibly reflecting that osteoporosis is both underdiagnosed and undertreated in men.^54,55

There are several strengths to this study, including large sample size, availability of pharmacy dispensing and medication administration claims data, and generalizability to commercially insured patients in the US. However, our results must be interpreted in light of the study design. As we examined medication use by calendar quarter, patients receiving medications with administration less frequently than every 3 months (denosumab biannually, zoledronic acid annually) will be relatively undercounted. However, we conducted a sensitivity analysis counting treatment effect of denosumab for 2 quarters and zoledronic acid for 4 quarters and found similar results to our primary analysis. In the fracture cohort, in which individuals receiving treatment greater than 6 months prior to fracture (e.g., with zoledronic acid given annually) but not in the quarters immediately following the fracture will be counted as untreated. Individuals in the fracture cohort may also have a variable look-back period (at least 90 and up to 180 days) depending on the exact timing of their fracture within a calendar quarter. Although this may affect reported relative rates of medication use, it should not significantly affect trends over time. Finally, the method used to identify fractures has not been previously validated, although we systematically collected ICD-9, ICD-10, and procedure codes for fractures at common osteoporotic sites, allowing for comprehensive fracture identification.

Our study demonstrates that alendronate is the most commonly used bone-directed therapy, with low but stable use of zoledronic acid, PTH/PTHrP analogs, and romosozumab and declines in use of other bisphosphonates and raloxifene over the past decade. Conversely, denosumab use has increased in the treatment of both osteoporosis and malignancy despite concerns about cost and rebound vertebral fractures. Finally, overall rates of treatment for secondary fracture prevention are dismally low (<15%) and have declined in the past decade. Ongoing efforts by clinicians, patient advocacy groups, and public health entities should focus on efforts to improve fracture prevention and treatment in at-risk populations, with attention to risk-benefit and cost-effectiveness analyses in the choice of medications used.