Abstract
This large-scale multicenter study compared the effectiveness and safety of alendronate and raloxifene in over 300 million patients across nine international databases. The primary outcome was the incidence of osteoporotic hip fracture. Results indicated that both treatments provide similar protection against hip fractures (HR 1.03, 95% CI 0.94-1.13). However, alendronate was associated with a significantly higher risk of atypical femoral fractures (HR 1.51) and a slightly higher risk of vertebral fractures (HR 1.07). These findings suggest that raloxifene is a competitive and potentially safer first-line treatment option for osteoporosis.
Introduction
Osteoporosis represents a significant global health challenge, particularly for postmenopausal women. While alendronate, a bisphosphonate, and raloxifene, a selective estrogen receptor modulator (SERM), are both widely prescribed, direct head-to-head comparative evidence has been lacking. Clinical guidelines often favor alendronate based on placebo-controlled trials, yet bisphosphonates are linked to rare but serious adverse events like atypical femoral fractures (AFF). This study sought to provide high-quality real-world evidence to guide clinical decision-making by leveraging a massive federated data network.
Methodology: Large-Scale Observational Research
The researchers utilized a retrospective, new-user cohort design within the Observational Health Data Sciences and Informatics (OHDSI) framework. By employing the OMOP Common Data Model (CDM), the study standardized data across nine diverse healthcare databases, including electronic health records and insurance claims from the US and South Korea, encompassing over 300 million patients.
To minimize bias, the study used:
- Expansive Propensity Score (PS) Modeling: Incorporating all available pre-treatment patient characteristics to account for confounding by indication.
- Negative Control Calibration: Testing the model against 147 unrelated outcomes to adjust for systematic bias and produce robust calibrated p-values.
Key Results: Effectiveness and Safety
The final cohort included 283,586 alendronate patients and 40,463 raloxifene patients.
Primary Outcome: Hip Fracture
The study found no significant difference in the risk of hip fractures between the two cohorts. The summary hazard ratio was 1.03 (95% CI 0.94–1.13), demonstrating parity in primary protection across various database types.
Secondary Outcomes and Safety
- Atypical Femoral Fracture (AFF): Alendronate users faced a 51% higher risk of AFF (HR 1.51, 95% CI 1.23–1.84).
- Vertebral Fracture: Surprisingly, alendronate was associated with a slightly higher risk of vertebral fractures compared to raloxifene (HR 1.07, 95% CI 1.01–1.14).
- Other Safety Concerns: No statistically significant differences were observed for esophageal cancer (HR 0.95) or osteonecrosis of the jaw (ONJ, HR 1.62).
Discussion and Clinical Implications
This study challenges the prevailing preference for alendronate by demonstrating that raloxifene is equally effective at preventing hip fractures while potentially carrying a lower risk of atypical femoral fractures. While limitations exist—such as the lack of Bone Mineral Density (BMD) data and potential channeling bias regarding thromboembolism—the methodological rigor and scale of this research provide strong evidence for raloxifene as a robust alternative in osteoporosis management. Future research should investigate the long-term effects of treatment duration and the impact of "drug holidays" on these comparative outcomes.
Summary updated