Bone microarchitecture and strength in men with prostate cancer with and without prevalent vertebral fractures at the initiation of androgen deprivation therapy

Androgen deprivation therapy (ADT) in prostate cancer (PCa) is associated with an increased fracture risk. However, studies on the prevalence of low areal bone mineral density (aBMD) and vertebral fractures (VFs) at the initiation of ADT and data on bone microarchitecture and strength are scarce. The aims of this study were 1) to evaluate bone microarchitecture and strength in men with PCa at the initiation of ADT in a real-world cohort as compared to normative data, and 2) to investigate whether the presence of VF at the initiation of ADT is associated with bone microarchitecture and strength.

The cohort consisted of men with PCa who were referred to our PCa care pathway for fracture prevention at ADT initiation, between January 2019 and October 2023. All men underwent dual-energy X-ray absorptiometry (DXA), X-ray of the spine, and high-resolution peripheral quantitative CT (HR-pQCT) at ADT initiation. VFs were identified morphometrically on X-ray using Genant's VF classification. Men with bone metastases, previous ADT treatment, >1 administration of ADT prior to entering the pathway, or prior or current use of anti-osteoporosis medication were excluded from the study. HR-pQCT parameters were compared with normative data to calculate Z-scores. Linear regression analyses were performed (covariates: age, body mass index, previous fractures, prostate-specific antigen level, time since diagnosis, chronic obstructive pulmonary disease, and rheumatoid arthritis) to study the relationship between VFs and bone microarchitecture.

A total of 256 men (median age: 73.9, IQR: 69.5–78.1 yrs) were evaluated, of whom 10 (3.9 %) had osteoporosis and 85 (33.2 %) at least one prevalent VF. Of the 85 men with ≥1 VF, 46 had 1 VF (54.1 %) and the other 39 (45.9 %) had at least 2 VFs. A total of 473 HR-pQCT scans of sufficient quality were analysed (n = 233 radius, n = 240 tibia). The mean Z-scores were ≥ −1 for all HR-pQCT parameters at both the radius and tibia in the entire cohort as well as when stratified by VF. Linear regression analyses showed that the presence of ≥1VF was associated with a larger trabecular and smaller cortical area, lower total and trabecular BMD, and lower trabecular and cortical thickness of the tibia. There were no significant associations between the HR-pQCT parameters at the distal radius or with femoral neck aBMD.

In conclusion, men with PCa had a bone microarchitecture and strength comparable to normative data at the time of ADT initiation. However, men with at least one prevalent VF had impaired microarchitecture at the tibia compared to those without prevalent VFs.