Prostate biopsy volume indices do not predict for significant Gleason upgrading.

Significant discordance exists between biopsy and matched prostatectomy grades. This study tests the hypothesis that surrogate tumor volume indices available from biopsies could yield an improved prediction of the underlying pathologic Gleason grade. Records of 124 patients who underwent radical prostatectomy were reviewed. Biopsies were characterized by primary and secondary Gleason grade, number of positive cores, and linear tumor length. Surgical specimens were characterized by primary and secondary Gleason grade, organ-confined disease, seminal vesicle invasion, and margins. Biochemical failure (BF) was defined by a postoperative prostate-specific antigen >0.05 ng/mL. There were 28 patients (24%) who experienced biochemical failure. On multivariate analysis, only the pathologic Gleason sum (P = 0.012) and the cumulative tumor length (P = 0.050) were independently associated with BF, and only the cumulative tumor length was associated with nonorgan-confined disease (P = 0.034). For patients with a cumulative tumor length >10 mm, 49% (18 of 37) had nonorgan-confined disease and 37% (13 of 35) had BF compared with 29% (25 of 87) and 19% (15 of 80), respectively, if they had cumulative tumor length < or =10 mm (P <0.034). Overall, an exact match was seen in 39% of biopsy Gleason grades, whereas 21% were downgraded by 1 or more points, and 41% were upgraded by 1 or more points. On univariate or multivariate analysis, none of the biopsy surrogate volume indices examined achieved significance or suggested a trend in predicting for a clinically meaningful grade change. Although indices of tumor volume from prostate needle biopsies independently predict for organ-confined disease and BF after prostatectomy, none predicted for a clinically significant upgrading or downgrading of biopsies. This suggests that the correlation that exists between such volume surrogates and outcomes after surgery reflect tumor volume effects only, independently of any possible association between tumor volume and Gleason grade.