Fracture risk on abiraterone: the opposite of what the meta-analyses suggested

Design

Secondary analysis of two randomised phase 3 trials with linked administrative data and competing-risk modelling

Population

3,102 of 3,893 patients with high-risk non-metastatic or metastatic prostate cancer in the STAMPEDE platform, randomised 2011-2016

Primary outcome

Five-year cumulative incidence of fracture-related hospitalisation, with death as a competing risk

Effect

M1 disease: 22% with abiraterone versus 30% standard care (SDHR 0.77, 95% CI 0.59-0.99, p=0.04); 28% versus 38% for abiraterone plus enzalutamide (SDHR 0.69, 0.54-0.88, p=0.002). No difference in M0

Men on androgen deprivation therapy fracture more, and previous meta-analyses of randomised trials suggested that adding an androgen receptor pathway inhibitor made that worse. This secondary analysis of two STAMPEDE trials tested that against hospital records rather than trial adverse event reporting.

Between November 2011 and March 2016, 3,893 patients with high-risk non-metastatic or metastatic prostate cancer were randomised to standard of care, standard of care plus abiraterone with prednisolone, or — in a later comparison — that combination plus enzalutamide. Linked Hospital Episode Statistics data were available for 3,102 patients in England, and fracture-related hospitalisations were identified with a prespecified coding framework. Flexible parametric competing-risk models estimated five-year cumulative incidence with death as the competing risk, which matters in a population where many patients die before they can fracture.

In metastatic disease, five-year fracture hospitalisation was lower with abiraterone than standard care alone: 22% against 30% (sub-distribution hazard ratio 0.77, 95% CI 0.59-0.99, p=0.04). With abiraterone plus enzalutamide it was 28% against 38% (SDHR 0.69, 95% CI 0.54-0.88, p=0.002). In non-metastatic disease there was no significant difference.

The authors' explanation is straightforward and plausible: in metastatic disease the fractures are largely pathological, and better disease control means fewer of them. The drug effect on bone density, whatever it is, is outweighed by the effect on the metastases.

• Do not withhold intensification from a metastatic patient on the grounds of fracture risk
• Bone-protective therapy and fracture risk assessment remain indicated — this does not replace them
• The finding applies to metastatic disease; in M0 patients there was no difference either way
• Fracture-related hospitalisation misses fractures managed without admission, which is most of them
• A 22-30% five-year hospitalised fracture rate is high; this is a population that needs bone health addressed

Why it matters

It removes a reason clinicians have used to hold back treatment intensification in metastatic disease.

Don't overread it

This is a secondary analysis using administrative hospitalisation codes, with no baseline bone density and no accounting for bone-protective therapy over time.

The statistics, in plain English

A sub-distribution hazard ratio accounts for the fact that a man who dies cannot later fracture — ignore that and you overestimate the fracture rate in the group that survives longer, which is exactly the trap here. The abiraterone interval (0.59 to 0.99) only just excludes 1.0, so the benefit is real but not large. And the outcome counted only fractures needing hospital admission, so the absolute figures understate the total burden.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.