Background: The 3-arm EMBARK trial found statistically superior and clinically meaningful improvements in metastasis-free survival (MFS) with enzalutamide + leuprolide acetate (enza combo) and enza monotherapy vs placebo + leuprolide (alone) in patients (pts) with high-risk BCR prostate cancer. Tx was suspended at week 37 in 321 (90.9%) pts in the enza combo arm and 240 (67.8%) in the leuprolide-alone arm based on prostate-specific antigen (PSA) response. Outcomes in the enza combo vs leuprolide-alone arms by tx suspension status are reported. Methods: The phase 3 EMBARK study enrolled pts with high-risk BCR, defined as PSA doubling time ≤9 months, and PSA ≥2 ng/mL above nadir post-radiotherapy (RT) or ≥1 ng/mL after radical prostatectomy (RP) ± postoperative RT. Pts were randomized 1:1 to enza 160 mg/day + leuprolide (22.5 mg every 12 weeks) or leuprolide alone. Tx was suspended at week 37 if serum PSA was <0.2 ng/mL at week 36 and restarted when PSA reached ≥2 ng/mL (primary RP) or ≥5 ng/mL (no primary RP). A secondary endpoint was the proportion of pts with undetectable PSA 2 years after suspension. MFS (per blinded independent central review) was analyzed descriptively between tx arms by suspension status. P-values were nominal. Results: The 3-year MFS rate (95% CI) in the suspension group was 94.4% (91.2–96.5%) with enza combo and 90.0% (85.3–93.2%) with leuprolide alone; in the no suspension group, 3-year MFS rates were 76.2% (33.2–93.5%) and 66.9% (55.4–76.1%), respectively. MFS in the suspension group was improved with enza combo vs leuprolide alone (HR 0.470, 95% CI 0.308–0.717; P=0.0003); no difference in MFS was observed in the no suspension group (HR 0.719, 95% CI 0.225–2.295; P=0.5763), although sample size was small for enza combo (n=9). Compared with no suspension, a higher proportion of pts in the suspension group received prior RP and a lower proportion received prior RT (Table). Two years after suspension, 16.8% (95% CI, 12.9–21.4%) of enza combo pts and 9.6% (95% CI, 6.2–14.0%) of leuprolide-alone pts had undetectable PSA (P=0.0089). Conclusions: MFS was improved with enza combo vs leuprolide alone for pts with BCR who suspended tx. No difference in MFS was observed in the no suspension group, but firm conclusions were precluded by the limited number of pts who did not suspend tx with enza combo. Pts who suspended tx had higher rates of prior RP vs those who did not. Pts who had received enza combo prior to tx suspension were more likely to have undetectable PSA 2 years thereafter compared with those who received leuprolide alone. Clinical trial information: NCT02319837.

*Pts who discontinued tx ≤week 37 were balanced across tx arms and excluded from the analysis.