Percentage of positive biopsy cores to predict distant metastasis of prostate cancer after definitive radiation therapy.

80 Background: To assess the prognostic value of percentage of positive biopsy cores (PPC) and perineural invasion (PNI) in predicting clinical outcome following radiotherapy (RT) for prostate cancer. Methods: One thousand and fifty-six patients with clinical stage T1-T3 N0 M0 prostate cancer, who had ≥ 4 biopsy cores sampled and complete biopsy core data available, were treated with either adaptive image-guided RT (median 75.6 Gy, n=387), low-dose EBRT (median 66.6 Gy, n=393), or EBRT and high-dose rate brachytherapy boost (n=276) at William Beaumont Hospital (1993-2004). Neoadjuvant and/or adjuvant androgen deprivation (AD) were given to 253 patients (24%). Multivariate cox regression analysis included PPC, gleason score, PSA, T stage, PNI, RT dose, androgen deprivation, and age. Biochemical failure (BF) was scored according to the Phoenix definition. Clinical failure (CF) was defined as any locoregional recurrence (LRR) or distant metastasis (DM). Median follow-up was 7.6 years. Results: Median cores sampled was 7, median PPC was 33%, and 18% had PNI. On univariate Cox regression, both PPC and PNI were predicators of biochemical failure and clinical failure (all P<0.05). On multivariate Cox regression, PPC, either as continuous or categorical variable, remained an independent predicator of BF, CF, DM, cause-specific survival, and overall survival (all P<0.05). PPC of >50% was associated with significantly higher DM (HR 4.01, 95% CI 1.86-8.61), and its independent predicative value remained significant whether AD was given or not (all P<0.05). Combining ≤50% vs ≥50% PPC with NCCN risk group stratification demonstrated added prognostic value of DM for intermediate-risk (HR 5.44, 95% CI 1.78-16.6) and high-risk groups (HR 4.39, 95% CI 1.70-2.84), with or without AD (all P<0.05). On multivariate Cox regression, PNI was an independent predicator of LRR only (HR 2.51, 95% CI 1.18-5.33). Conclusions: PPC is an independent and powerful predicator of DM for intermediate- and high-risk prostate cancer, regardless the use of AD. It should be considered for risk stratification and when designing for future trials testing adjuvant treatment after definitive RT for prostate cancer. No significant financial relationships to disclose.