Analysis of clinicopathologic factors predicting outcome after radical prostatectomy

BACKGROUND: A variable biochemical failure rate has been reported for patients undergoing radical prostatectomy. The authors analyzed their 1987-1993 prostatectomy experience retrospectively to stratify the risk of failure in order to appropriately select patients who potentially may benefit from adjuvant therapy. METHODS: A stepwise logistic regression was used to identify variables associated with biochemical failure in 265 patients who underwent radical prostatectomy only. Prostate tumors were examined by one pathologist using 4-mm step sections. Numerous clinicopathologic variables were evaluated, and the neoplasms were subclassified into five pathologic categories based on tumor extent and margin status. Actuarial projections of biochemical failure were created using the Kaplan-Meier method. RESULTS: Pathologically, 56.2% of the tumors were organ-confined with negative margins, 12.8% had a positive surgical margin without evidence of extraprostatic extension (EPE), 24.2% had EPE (17% with negative margins and 7.2% with positive margins), and 6.8% had seminal vesicle involvement. The Gleason score was > or = 7 in 86.4% of the total population. Values for the preoperative prostate specific antigen assay were < or = 4.0 ng/mL in 23.4% of the men and > 10 ng/mL in 27.7%. The overall observed biochemical failure rate in this patient group with a minimum 48 months of follow-up was 15.5%. Overall, stepwise logistic regression analysis revealed that pathologic category was the variable most strongly associated with biochemical failure and that vascular invasion was the only other examined variable associated with failure. CONCLUSIONS: The combination of pathologic category and the prostatectomy Gleason score can stratify a patient's probability of biochemical failure into three distinct groups and can identify the appropriate patients who may benefit from novel adjuvant therapeutic strategies.