• intermittent androgen deprivation therapy;
  • non-metastatic biochemically relapsed prostate cancer;
  • PSA doubling time;
  • serum testosterone;
  • disease progression



We sought to determine the change of PSA doubling time (PSADT) and its association with disease progression during intermittent androgen deprivation (IAD) therapy for prostate cancer.


Data were retrospectively analyzed in 96 patients with biochemically relapsed prostate cancer (BRPC) treated with IAD since 1995. IAD consisted of LHRH-agonists ± antiandrogen given usually at PSA threshold (ng/ml) of 10–20, for 6–9 months. Cycles were repeated until the development of castration resistance. Mixed effects model was used to study PSADT change over cycles. Multivariate cox regression model was used to identify outcome-associated variables.


Patients received a mean of 2.8 treatment cycles over a mean follow-up time of 71 months. Fifty-seven (59%) remain on treatment and 39 (41%) developed PSA refractoriness (n = 8) or positive scans (n = 31). First off treatment interval PSADT (median 2.3 months) was significantly shorter than the baseline (median 7.34) but remained stable in subsequent cycles. Off treatment interval PSADT adjusted for testosterone recovery (median 3.7) was significantly longer than that based on all PSA determinations (median 2). Factors associated with disease progression were pre-treatment PSADT (≥6 vs. <6), first off treatment interval PSADT (≥3 vs. <3), and PSA nadir during the first treatment interval (<0.1 vs. ≥0.1).


During IAD for BRPC, PSADT becomes shorter, and is associated with testosterone recovery. PSADT before treatment and during the first off treatment interval is associated with disease progression. If prospectively validated these data may guide treatment with IAD and clinical trial design. Prostate 71:1608–1615, 2011. © 2011 Wiley-Liss, Inc.