Denosumab and bone metastasis-free survival in men with nonmetastatic castration-resistant prostate cancer: exploratory analyses by baseline prostate-specific antigen doubling time

Abstract

Purpose: Denosumab, an anti-RANK ligand monoclonal antibody, significantly increases bone metastasis-free survival (BMFS; hazard ratio [HR], 0.85; P = .028) and delays time to first bone metastasis in men with nonmetastatic castration-resistant prostate cancer (CRPC) and baseline prostate-specific antigen (PSA) ≥ 8.0 ng/mL and/or PSA doubling time (PSADT) ≤ 10.0 months. To identify men at greatest risk for bone metastasis or death, we evaluated relationships between PSA and PSADT with BMFS in the placebo group and the efficacy and safety of denosumab in men with PSADT ≤ 10, ≤ 6, and ≤ 4 months.

Patients and methods: A total of 1,432 men with nonmetastatic CRPC were randomly assigned 1:1 to monthly subcutaneous denosumab 120 mg or placebo. Enrollment began February 2006; primary analysis cutoff was July 2010, when approximately 660 men were anticipated to have developed bone metastases or died.

Results: In the placebo group, shorter BMFS was observed as PSADT decreased below 8 months. In analyses by shorter baseline PSADT, denosumab consistently increased BMFS by a median of 6.0, 7.2, and 7.5 months among men with PSADT ≤ 10 (HR, 0.84; P = .042), ≤ 6 (HR, 0.77; P = .006), and ≤ 4 months (HR, 0.71; P = .004), respectively. Denosumab also consistently increased time to bone metastasis by PSADT subset. No difference in survival was observed between treatment groups for the overall study population or PSADT subsets.

Conclusion: Patients with shorter PSADT are at greater risk for bone metastasis or death. Denosumab consistently improves BMFS in men with shorter PSADT and seems to have the greatest treatment effects in men at high risk for progression.

Fig 1.

CONSORT diagram. PSADT, prostate-specific antigen doubling time. (*) Data published previously. (†) One additional patient assigned to placebo and two patients assigned to denosumab were excluded after random assignment because review activities and oversight of institutional review board were not ensured.

Fig 2.

(A) Relative risk for bone metastasis–free survival (BMFS) over prostate-specific antigen doubling time (PSADT) in placebo group. Cox proportional hazards model was used with natural cubic spline of 6 df for inverse of PSADT. The y-axis represents increasing risk; x-axis represents PSADT at study entry expressed as continuous variable. For practical purposes, x-axis begins at 20 months at origin and goes to doubling time of 2 months. (B) Proportion of patients with BMFS in placebo group among overall population and by PSADT subgroup.

Fig 3.

Bone metastasis–free survival by prostate-specific antigen doubling time (A) ≤ 10, (B) ≤ 6, and (C) ≤ 4 months. HR, hazard ratio.

Fig 4.

Time to first bone metastasis by prostate-specific antigen doubling time (A) ≤ 10, (B) ≤ 6 months, and (C) ≤ 4 months. HR, hazard ratio.

Fig A1.

Results by nonoverlapping prostate-specific antigen (PSA) doubling time subgroups; hazard ratios and 95% CIs from Cox proportional hazards model stratified by random assignment stratification variables.