Purpose: A multivariable analysis to evaluate the potential clinical and pathologic factors that predict for early biochemical failure in patients with pathologically organ-confined and margin-negative disease was performed to define patients who may benefit from adjuvant therapy.
Patients and methods: Three hundred forty-one prostate cancer patients treated with a radical retropubic prostatectomy between January 1989 and June 1995 and found to have pathologically organ-confined and margin-negative disease comprised the study population. A logistic regression multivariable analysis to evaluate the predictive value of the preoperative prostate-specific antigen (PSA) level, pathologic (prostatectomy) Gleason score, and pathologic stage on PSA failure occurring during the first postoperative year was performed.
Results: Predictors of PSA failure during the first postoperative year in patients with pathologically organ-confined disease included pathologic Gleason score > or = 7 (P = .0007) and preoperative PSA level greater than 10 (P < .0001). Corresponding 3-year freedom-from-PSA-failure rates for these pathologic organ-confined patients with both, one, or neither of these factors were 60%, 75% to 84%, and 95%, respectively (P < .0001).
Conclusion: Prostate cancer patients with pathologically organ-confined and margin-negative disease and a preoperative PSA level greater than 10 ng/mL or a pathologic Gleason score > or = 7 have significant decrements in short-term PSA-failure-free survival. Therefore, these patients should be considered for adjuvant therapy in the setting of a phase III clinical trial.