Radium-223 with abiraterone: a caution, not a combination

ERA-223 was a randomised, double-blind, placebo-controlled phase 3 trial (Smith et al., sponsored by Bayer) in 806 men with asymptomatic or mildly symptomatic bone-predominant metastatic castration-resistant prostate cancer, comparing radium-223 plus abiraterone acetate/prednisone/prednisolone against placebo plus abiraterone acetate/prednisone/prednisolone. The trial was unblinded prematurely on 17 November 2017, before completion of follow-up, after an unplanned ad hoc analysis showed more fractures and more deaths in the radium-223 arm.

At the primary analysis (data cut-off 15 February 2018), the combination did not improve the primary endpoint of symptomatic skeletal-event-free survival: median 22.3 months with radium-223 versus 26.0 months with placebo (hazard ratio 1.122, 95% CI 0.917-1.374, p=0.26). Overall survival was numerically shorter with radium-223 (30.7 versus 33.3 months; 155 deaths [39%] versus 141 deaths [35%]) though the difference was not statistically significant. The starkest signal was fracture risk: any-grade fractures occurred in 29% of the radium-223 group versus 11% of the placebo group in the extended follow-up analysis (26% versus 10% during the on-treatment period). Independent central review confirmed 76 patients with at least one fracture in the combination arm versus 23 on placebo, with osteoporotic (fragility) fractures nearly tenfold more common (37 versus 4 patients). Fracture risk was substantially reduced, but not eliminated, by concurrent bone-health agents (bisphosphonates or denosumab): 15% versus 7% fracture incidence with and without radium-223 among bone-health-agent users, compared with much higher rates in patients not on a bone-health agent.

On the basis of ERA-223, the investigators explicitly do not recommend radium-223 in combination with abiraterone acetate plus prednisone/prednisolone, and the FDA and EMA revised prescribing information for radium-223 accordingly. This is one of the clearest examples in prostate-cancer radiopharmaceutical practice where two active, individually beneficial agents (radium-223 improved overall survival in ALSYMPCA; abiraterone improves survival in mCRPC) produce net harm when combined, and it is now standard teaching to avoid concurrent radium-223 and abiraterone outside a bone-health-agent-protected, closely monitored context.