Bone mineral density testing and the risk of fractures in men initiating androgen-deprivation therapy: Population-based study

238 Background: Androgen deprivation therapy (ADT) is a staple of advanced prostate cancer (PCa) treatment, however several side-effects are associated with its long-term use. Notably, loss of bone mineral density (BMD) is accelerated which increases fracture risk. Although guidelines recommend BMD testing when initiating ADT to properly assess baseline fracture risk, there is limited data to support this recommendation in the PCa patient population. The objective was to examine the association between baseline BMD testing (bBMDT) and the risk of fractures in men initiating ADT for PCa. Methods: A retrospective observational cohort study using data from Quebec public healthcare insurance administrative databases was conducted. The cohort included PCa patients who initiated ADT from 2004-2015 and who received at least one year of ADT treatment. Baseline BMD testing was defined as a BMD test performed from 6 months prior to ADT initiation and up to 3 months after. Patients were categorized as either having received bBMDT or not received bBMDT when initiating ADT. The primary study outcomes were incidence of any fracture and incidence of fractures resulting in hospitalization. Inverse probability of treatment weighting was used to adjust for measured baseline characteristics which included patient demographic variables, comorbidities, risk factors for fractures, and use of other medications affecting bone density. Results: We identified 13,532 patients who initiated ADT, of which 2,070 (15.3%) underwent bBMDT. The unadjusted 5-year incidence of any fracture was 15.1% for patients not receiving bBMDT and 14.0% for patients receiving bBMDT. In adjusted analyses, bBMDT (hazard ratio [HR] 0.92, 95% confidence interval [CI] 0.76-1.12) was not associated with the risk of any fracture. For fractures requiring hospitalization, bBMDT was associated with a lower risk (HR 0.71, 95%CI 0.52-0.98). Furthermore, bBMDT was associated with increased odds of bisphosphonate use within 1 year of ADT initiation among patients who were bisphosphonate-naïve at baseline (odds ratio [OR] 2.03, 95%CI 1.74-2.36). Conclusions: In our study population, bBMDT was associated with a lower risk of fractures resulting in hospitalization. Given the low uptake of bBMDT, additional efforts emphasizing the importance of BMD testing in guidelines may be needed.