Using PSA to guide timing of androgen deprivation in patients with T0-4 N0-2 M0 prostate cancer not suitable for local curative treatment (EORTC 30891)

EORTC trial 30891 compared immediate versus deferred androgen-deprivation therapy (ADT) in T0-4 N0-2 M0 prostate cancer (PCa). Many patients randomly assigned to deferred ADT did not require ADT because they died before becoming symptomatic. The question arises whether serum prostate-specific antigen (PSA) levels may be used to decide when to initiate ADT in PCa not suitable for local curative treatment. PSA data at baseline, PSA doubling time (PSADT) in patients receiving no ADT, and time to PSA relapse (>2 ng/ml) in patients whose PSA declined to <2 ng/ml within the first year after immediate ADT were analyzed in 939 eligible patients randomly assigned to immediate (n=468) or deferred ADT (n=471). In both arms, patients with a baseline PSA>50 ng/ml were at a>3.5-fold higher risk to die of PCa than patients with a baseline PSA <or=8 ng/ml. If baseline PSA was between 8 and 50 ng/ml, the risk of PCa death was approximately 7.5-fold higher in patients with PSADT <12 mo than in patients with PSADT>12 mo. Time to PSA relapse after response to immediate ADT correlated significantly with baseline PSA, suggesting that baseline PSA may also reflect disease aggressiveness. Patients with a baseline PSA>50 ng/ml and/or a PSADT <12 mo were at increased risk to die from PCa and might have benefited from immediate ADT, whereas patients with a baseline PSA <50 ng/ml and a slow PSADT (>12 mo) were likely to die of causes unrelated to PCa, and thus could be spared the burden of immediate ADT