Androgen deprivation therapy and radiotherapy with or without cabazitaxel in very-high risk localized prostate cancer: First results of the PEACE-2 randomized phase III trial.

LBA307 Background: Long-term androgen deprivation therapy (ADT) combined with radiotherapy is standard in patients with localized prostate cancer, no detectable metastases, and a high-risk of dissemination. Cabazitaxel improves overall survival in patients with metastatic castration-resistant disease. We hypothesized that earlier use of cabazitaxel may prevent the onset of relapse or death. Methods: PEACE 2 (NCT01952223) is a 2x2 factorial design randomized trial for patients with very high-risk localized prostate cancer (defined by at least 2 criteria among: Gleason 8-10, T3/T4 disease (MRI T stage permitted), and PSA >20 ng/mL) and no detectable metastases on conventional imaging. Eligible patients all received standard of care (SOC), which consisted of prostate only fractionated RT (74-78Gy in 37-39 fractions) and 3 years of ADT. Patients were then randomized 1:1:1:1 to receive SOC only (Arm A), SOC + prophylactic pelvic RT (46-50 Gy in 23-25 fractions, Arm B), SOC + 4 cycles of cabazitaxel (20-25 mg/m2/3 weeks x 4 cycles) prior to RT (Arm C), or SOC + both cabazitaxel and pelvic RT (Arm D). The primary endpoint is clinical progression-free survival (cPFS) with death, metastases and proven local relapses as events. Initially 1048 patients were planned to detect a treatment effect corresponding to a hazard ratio of 0.70 (absolute difference of 7.5% in cPFS at 6 years) for both comparisons. The accrual was closed early but the analysis plan was maintained after the target event number was reached (n=247 planned events), hence maintaining statistical power. Interaction was first tested between cabazitaxel and pelvic RT: if no interaction was detected, hazard ratios (Cox, logrank) for cPFS were reported for ADT-RT vs ADT-RT-cabazitaxel. Database was locked in October 2025. Results: Overall, 761 patients were included from 2013 to 2021 (median age: 67y) from 4 EU countries, with 2 (79%) or 3 (21%) risk factors, and they were randomized to receive cabazitaxel (n=381) or not (n=380). Pet-choline was used as part of baseline imaging in 18%. The median follow-up is about 85 months and no interaction was found between cabazitaxel and pelvic RT. There was no improvement of cPFS with cabazitaxel (HR: 1.11 0.87-1.41; 6-year cPFS rates: 67.2% vs 71.4%). Overall survival rates were greater than 85% in all arms with also no difference. Grade >2 toxicity was reported in 65% and 47% respectively with or without cabazitaxel. Conclusions: Cabazitaxel does not improve cPFS in very-high risk localized prostate cancer. With very few prostate cancer-related deaths observed during the first decade, data from PEACE-2 challenge our current definition of very-high risk localized prostate cancer, especially when defined using modern imaging. Clinical trial information: NCT01952223 .