THE GREATEST DIMENSION OF PROSTATE CARCINOMA IS A SIMPLE, INEXPENSIVEPREDICTOR OF PROSTATE-SPECIFIC ANTIGEN FAILURE IN RADICAL PROSTATECTOMY SPECIMENS

BACKGROUND. Tumor volume in radical prostatectomies can be determined by several different techniques and appears to predict clinical progre ssion. Greatest tumor dimension and area are easily obtained measures that are both correlated with tumor volume. The authors sought to dete rmine whether greatest tumor dimension and/or area were predictors of prostate specific antigen (PSA) failure in men who underwent radical p rostatectomy for adenocarcinoma of the prostate.METHODS. Fifty-seven m en with prostate carcinoma who underwent surgical resection were follo wed for a median of 27.2 months (range, 1-112 months); 24 (42%) of the se men had PSA failure. Preoperative PSA, Gleason grade, pathologic st age, margin status, and greatest tumor dimension and area were determi ned, and both univariate and multivariate analyses of the outcomes of PSA failure were performed. RESULTS. In the univariate analysis, large r values of greatest tumor dimension and area were strongly associated with increased incidence of PSA failure (P = 0.0001 and 0.0011, respe ctively). The forward stepwise multivariate analysis indicated that gr eatest tumor dimension had marginal statistical significance as a risk factor for PSA failure (P = 0.0577; risk ratio, 1.117). However, in t his series, men with a greatest tumor dimension of less than 1 cm did not experience failure, whereas all patients with a greatest tumor dim ension of more than 2 cm did. CONCLUSIONS. Measuring greatest tumor di mension is a simple, inexpensive predictor of PSA failure in men who h ave undergone radical prostatectomy. Cancer 1998;83:748-52. (C) 1998 A merican Cancer Society.