Differential Association of Post-Testosterone Recovery PSA Kinetics with Progression and Metastatic Outcomes in Non-metastatic Prostate Cancer.

PURPOSE: Prostate-specific antigen (PSA) kinetics following testosterone recovery remain difficult to interpret in patients treated with definitive radiotherapy (RT) with androgen deprivation therapy (ADT). We evaluated the association between post-testosterone recovery PSA dynamics and progression and metastatic outcomes using Bayesian joint modeling.

METHODS

Individual patient data were pooled from two randomized trials of localized prostate cancer treated with RT and ADT. Serial PSA measurements obtained up to 18 months after testosterone recovery to supracastrate levels (>50 ng/dL) and before progression were analyzed using Bayesian joint models linking longitudinal mixed-effects models of log-transformed PSA with survival submodels for progression-free survival (PFS) and metastasis-free survival (MFS). Associations of PSA slope and cumulative PSA exposure (area under the curve, AUC) with event risk were estimated via shared random effects. Sensitivity analyses used Bayesian cause-specific joint models censoring deaths.

RESULTS

The cohort included 461 patients contributing 1,594 post-recovery PSA measurements with median follow-up of 115 months. PSA slope was re-expressed as hazard ratios (HRs) at selected PSA doubling time (PSADT) values relative to stable PSA. For progression-free survival (PFS), shorter PSADT was associated with directionally higher hazards of progression or death (PSADT 12 months: posterior median HR 1.10; 95% credible interval [CrI], 0.87-1.39; PSADT 6 months: HR 1.20; 95% CrI, 0.75-1.92) with 78% posterior probability of HR>1. For metastasis-free survival (MFS), estimates were directionally opposite (PSADT 12 months: HR 0.80; 95% CrI, 0.61-1.02; PSADT 6 months: HR 0.63; 95% CrI, 0.37-1.04). Cumulative PSA exposure showed no clinically meaningful association with either endpoint. In death-censored Bayesian joint models, the PSADT 12-month HR was 1.27 (95% CrI, 0.98-1.64) for progression and 1.13 (95% CrI, 0.59-2.15) for distant metastasis, with imprecision for distant metastasis reflecting the limited number of metastatic events (n=32).

CONCLUSIONS

PSA slope after testosterone recovery was directionally associated with progression risk. Associations with metastatic outcomes were inconsistent across modeling frameworks and imprecisely estimated, particularly for distant metastasis because of the limited number of events. No definitive inference regarding metastatic risk should be drawn from these analyses.