Nitrogen-containing bisphosphonates (N-BPs; such as alendronate and zoledronate) are commonly used in the treatment of osteoporosis and fracture prevention, in which zoledronate has a proven better efficacy than alendronate. In 2018, our real-world propensity score matched study showed that the use of N-BPs was significantly associated with reduced risk of myocardial infarction and stroke in hip fracture patients. In addition to cardiovascular diseases, both preclinical study and sensitivity analysis also suggest evidence for N-BPs in pneumonia prevention. Moreover, a pragmatic clinical trial is developed to evaluate effect of the tested intervention in real-life routine clinical practice since traditional explanatory radomised controlled trial (RCT) may have poor generalizability due to highly selected patients and controlled environments. This study aims to evaluate if zoledronate reduces risk of pneumonia in hip fracture patients using pragmatic clinical trial approach.

This is an open-label, multi-centre, pragmatic, randomised controlled trial. Patients will be recruited from 4 hospitals, namely Caritas Medical Centre, Prince of Wales Hospital, Queen Mary Hospital, and United Christian Hospital. Age, sex, body mass index, eGFR, history of fracture, chronic respiratory diseases, and other medical history, will be measured and recorded at recruitment.

N-BPs are widely used in the treatment of osteoporosis and fracture prevention. Although alendronate is the first-line antiosteoporosis medication in many countries, it is associated with esophageal and gastrointestinal irritation. In addition, the regimen for alendronate is one tablet per week and careful use of alendronate is required to avoid gastrointestinal irritation (e.g. take the tablet on an empty stomach, stay upright for at least 30 minutes after taking the medication), leading to relative low drug compliance. On the other hand, the regimen of zoledronate is 5mg infusion once a year. Since the route of administration is intravenous infusion, it avoids the gastrointestinal irritation problem. In terms of fracture prevention, zoledronate has a proven better efficacy than alendronate. Most importantly, previous HORIZON recurrent fracture trial showed that zoledronate use reduced risk of mortality in addition to fracture prevention. Thus, zoledronate is considered the most efficacious N-BP in clinical use.

N-BPs could exert extra-skeletal beneficial effects. In 2018, our real-world propensity score matched study (N=34,991) using the Clinical Data Analysis and Reporting System (CDARS), a clinical database managed by the Hong Kong Hospital Authority, showed that use of N-BPs was significantly associated with reduced risk of myocardial infarction and stroke in hip fracture patients. Later in the same year, a randomized controlled trial (RCT) in 2,000 osteopenic women showed that zoledronate may reduce risk of myocardial infarction with an odds ratio of 0.61 (95% CI: 0.36-1.02) after following up for 6 years. They subsequently performed a detailed post-hoc analysis and showed that zoledronate use was significantly associated with reduced risk of myocardial infarction, composite cardiovascular end-point, and fatal stroke. This example demonstrated not only the extra-skeletal effect of N-BPs, but also showed that high-quality real-world data with state-of-the-art statistical method could potentially be useful in causal inference.

In addition to cardiovascular diseases, N-BPs may be useful in preventing pneumonia. Preclinical study showed that N-BPs are anti-inflammatory and may modulate macrophages in response to pneumonia. In addition, pharmacokinetic studies showed that the highest concentration of N-BPs was detected in trachea other than bone after oral ingestion or intravenous infusion of N-BPs. Among various N-BPs, alendronate was detected in the trachea with a concentration of 607 ng/ml 72 hours after oral ingestion, which was almost half of the concentration (1370 ng/ml) detected in vertebrae. Thus, we conducted a real-world population-based propensity score matched study in a cohort of 54,047 hip fracture patients to investigate the association of N-BP use with risk of pneumonia. Among hip fracture patients, N-BP use was significantly associated with 24% risk reduction in pneumonia (hazard ratio [HR]: 0.76; 95% confidence interval [CI]: 0.70-0.83), with an absolute risk difference of 2%. Similar significant association was observed for pneumonia mortality. To further reduce the potential bias of confounding by indication, we performed a sensitivity analysis by including users of other anti-osteoporosis medications in the control group, instead of patients without using any anti-osteoporosis medications. A similar significant association was observed. In agreement with the preclinical studies, we provided evidence for N-BPs in pneumonia prevention. This finding attracted wide media coverage, including Reuters and the New York Times.

Although utilisation of real-world data is an emerging approach in evaluating drug effectiveness, RCT is still considered the gold standard in evaluating drug efficacy. However, traditional explanatory RCT may have poor generalizability due to highly selected patients and controlled environments. Thus, pragmatic clinical trial is developed to evaluate if the tested intervention is effective in real-life routine clinical practice, by reducing bias using randomization and improving generalizability using real-world setting. We aim to evaluate if zoledronate reduces risk of pneumonia in hip fracture patients using pragmatic clinical trial approach.

Inclusion Criteria:

• Male or female ≥ 60 years
• With recent fragility hip fracture at proximal femur
• Have the ability to understand the requirements of the study, provide written informed consent, including consent for the use and discloser of research-related health information, and comply with the study data collection procedures. Provide signed and dated informed consent form

Exclusion Criteria:

• Known to be hypersensitive to any N-BPs
• Estimated glomerular filtration rate (eGFR) < 30 ml per minute per 1.73 m2 of body surface area
• Regular user of anti-osteoporosis medications (including bisphosphonates, denosumab, teriparatides, and raloxifene) or oral or intravenous systemic glucocorticoids in the previous year.
• Subject currently involved in a clinical trial or in an exclusion period following participation in another clinical trial

Individual participant data will be shared upon special request to the principal investigator.