Percentage of positive biopsy cores as a predictor of clinical outcome in prostate cancer treated with radiotherapy

Purpose: The clinical significance of pretreatment biopsy characteristics is not well understood relative to known prognostic factors. We performed a complete pathology analysis of pretreatment biopsy specimens in an attempt to clarify their impact on clinical outcome following radiotherapy. Materials and Methods: From 1991 to 1999, 160 patients with locally advanced prostate cancer were prospectively treated with external beam radiotherapy in combination with high dose rate brachytherapy at our hospital and had pretreatment biopsy material available for complete pathological review. Patients with pretreatment prostate specific antigen (PSA) 10.0 ng./ml.. or greater, Gleason 7 or greater or clinical stage T2b-T3c NO MO disease were eligible for study entry. Pelvic external beam radiotherapy (46.0 Gy.) was supplemented with 3 (1991 to 1995) or 2 (1995 to 1999) ultrasound guided transperineal interstitial (192)iridium high dose rate implants. The brachytherapy dose was escalated from 5.50 to 10.50 Gy. per implant. All pretreatment biopsy specimen slides from each case were reviewed by a single pathologist (N. S. G.). Median followup was 4.2 years. Biochemical failure was defined as 3 consecutive PSA increases. Results: On univariate analysis pretreatment PSA, Gleason score, clinical T classification, percentage of positive biopsy cores, dose per implant and total radiotherapy dose (equivalent in 2 Gy. fractions) were significantly associated with biochemical failure. Perineural invasion was of borderline significance (p = 0.07). On univariate analysis for clinical failure only Gleason score and percent positive cores were significant. Percent positive cores was associated with biochemical and clinical failure whether analyzed in subgroups or as a continuous variable. Patients with less than 33% positive cores had a 5-year biochemical control rate of 83% and 5-year clinical failure rate of only 7%. Conversely, patients with more than 67% positive cores had a 5-year biochemical control rate of only 57% and 25% had clinical failure at 5 years. Since percent positive cores correlated with biochemical and clinical failure, an analysis was performed to determine which other prognostic factors were associated with percent positive cores. Pretreatment PSA level, Gleason score, clinical T classification and perineural invasion were significantly associated with a higher percent positive cores. Nevertheless, on Cox multiple regression analysis higher percent positive cores, pretreatment PSA and Gleason score remained independently associated with biochemical failure but not T classification. On Cox multiple regression analysis for clinical failure only Gleason score remained independently significant with percent positive cores maintaining borderline significance (p = 0.07). Conclusions: Percent positive pretreatment biopsy cores is a powerful predictor of biochemical and clinical outcome for prostate cancer treated with radiotherapy, independent of other known prognostic factors. Percent positive cores should be seriously considered as a primary factor in risk group stratification for prostate cancer.