Importance: Increased prostate-specific antigen (PSA) levels after treatment (PSA failure) may have different associations with outcomes for men with locally advanced vs localized prostate cancer.
Objective: To evaluate whether the association between PSA failure and death may be different in locally advanced vs localized prostate cancer.
Design, setting, and participants: This multicenter cohort study included patients from 2 randomized clinical trials. The Dana-Farber Cancer Institute (DFCI) 95-096 trial randomized 206 men with localized prostate cancer from December 1, 1995, to April 15, 2001, whereas the European Organisation for Research and Treatment of Cancer (EORTC) 22961 trial randomized 970 men with locally advanced prostate cancer from October 30, 1997, to May 1, 2002. Data were analyzed from January 1, 2020, to October 31, 2020.
Interventions: The DFCI 95-096 trial randomized men to 0 vs 6 months of androgen deprivation therapy (ADT) with external beam radiotherapy; the EORTC 22961 trial randomized men to 6 vs 36 months of ADT with external beam radiotherapy.
Main outcomes and measures: For each trial, the PSA doubling time (time to doubling of PSA levels) associated with PSA failure was evaluated. The risk of all-cause mortality associated with PSA failure (nadir plus 2 definition) was evaluated after adjustment of baseline covariates and treatment.
Results: This analysis included a total of 1173 men (206 from DFCI 95-096 and 967 with available tumor stage from EORTC 22961; median age, 70.0 [interquartile range (IQR), 65.0-74.0 years). For DFCI 95-096, 161 men died (30 [18.6%] due to prostate cancer) at a median follow-up of 18.2 (IQR, 17.3-18.8) years. Among the 108 men with PSA failure, the median PSA doubling time was 13.0 (IQR, 7.4-31.1) months. For EORTC 22961, 230 men died (75 [32.6%] due to prostate cancer) at a median follow-up of 6.4 (IQR, 6.3-6.6) years. Among 290 men who experienced PSA failure, the median PSA doubling time was 5.0 (IQR, 2.9-8.9) months. Compared with DFCI 95-096, PSA failure was associated with a higher risk of all-cause mortality in EORTC 22961 (adjusted hazard ratios, 3.98 [95% CI, 2.92-5.44]; P < .001 vs 1.51 [95% CI, 1.03-2.23]; P = .04).
Conclusions and relevance: The association of PSA failure with outcomes may differ between locally advanced and localized prostate cancer. This finding supports the study of treatment intensification with the use of novel antiandrogen agents in addition to ADT at the time of PSA failure after treatment for locally advanced disease.
Trial registration: ClinicalTrials.gov Identifiers: NCT00116220 and NCT00003026.