The effect of Gleason score on the predictive value of prostate-specific antigen doubling time

To evaluate the influence of the pathological Gleason score on the predictive value of the prostate-specific antigen (PSA) doubling time (DT), as this variable predicts a patient's risk of disease progression both before and after definitive therapy for prostate cancer, and there is an inverse correlation between the Gleason score and PSA production.

We evaluated all men treated with radical prostatectomy (RP) between 1990 and 1999 who did not receive neoadjuvant or adjuvant therapy. We identified 2296 patients who had multiple PSA values available before RP, and 1323 who had biochemical recurrence after RP and had at least two PSA values available before starting secondary therapy. Systemic progression and cancer-specific survival (CSS) rates were estimated using the Kaplan-Meier method and Cox proportional hazard regression models.

A PSA DT of <18 vs >18 months predicted a lower 10-year systemic progression-free survival for patients with tumours having a pathological Gleason score of <7 (98% vs 99%, P = 0.005), 7 (82% vs 91%, P = 0.003) and 8-10 (57% vs 73%, P = 0.042). A PSA DT after RP of <12 months was significantly associated with a lower 10-year systemic progression-free survival for patients with tumours having a Gleason score of <7 (77% vs 94%, P < 0.001) and 7 (61% vs 86%, P < 0.001), but not 8-10 (61% vs 75%, P = 0.11). The ability of PSA DT before and after RP to predict systemic progression and CSS decreased with increasing Gleason score.

The PSA DT remains associated with outcome both before and after RP across increasing pathological Gleason scores, although the predictive ability of PSA DT is diminished in Gleason 8-10 cancers.