patient preference. Biopsy reclassification is defined by >2 positive cores, >50% core involvement, or GS >7 (upgrading) at annual sur- veillance biopsy. Cox proportional hazards models of time to biopsy reclassification and upgrading evaluated PSA (PSA ¼2.5 vs. <2.5) and PSAD (PSAD ¼0.09 vs. <0.09), adjusted for age, number of positive cores, maximum core involvement, and year of diagnosis. Sub-ana- lyses evaluated whether the effect of PSAD or PSA varied over time (before 2004 vs. 2004 -2014), or by prostate volume (¼ 45 mm3 vs. <45 mm3), or age at diagnosis (¼65 vs. <65). RESULTS: 1160 patients (median age 66) with at least 1 sur- veillance biopsy entered the AS program from 1995-2014. Median follow-up duration, time to reclassification, and time to upgrading were 4.8, 3.3, and 4.2 years, respectively. On multivariable analysis, risk of biopsy reclassification was associated with both binary PSAD, hazard ratio [HR]¼1.35 (95% CI 1.1-1.6), p¼0.003; and PSA, HR¼1.59 (95% CI 1.17-2.15), p¼0.003, as was risk of upgrading: PSAD, HR¼1.53 (95% CI 1.14-2.06), p¼0.004; and PSA, HR¼1.71, (95% CI 1.08-2.71), p¼0.023. Particularly strong significant associations for both PSAD and PSA were seen in men older than 65 who had prostate volume <45 mm3 (HR¼2.40-3.59). However, adding PSA or PSAD to a base model with age, number of positive cores, and maximum tumor percentage gave only small increases in the c-index, except in men with prostate volume <45 (c-index increased from 0.613 to 0.646 for PSA and to 0.648 for PSAD). CONCLUSIONS: Both PSAD and PSA have statistically sig- nificant predictive value for risks of biopsy reclassification and biopsy upgrading, with neither showing an advantage over the other. However, their utility for risk stratification is particularly important in men older than 65 with prostate volume <45 mm3. Source of Funding: None. MP15-09 BASELINE AND LONGITUDINAL PCA3 PREDICT MORE EXTENSIVE CANCER IN AN ACTIVE SURVEILLANCE POPULATION Jeffrey Tosoian*, Mufaddal Mamawala, Patricia Landis, Sacha Wolf, Jonathan Epstein, Lori Sokoll, Christian Pavlovich, Baltimore, MD INTRODUCTION AND OBJECTIVES: One drawback of active surveillance (AS) is the use of serial prostate biopsy, which is costly and associated with various risks. As such, alternative methods of moni- toring AS patients are needed. Prostate cancer antigen 3 (PCA3) is a urinary biomarker that may improve diagnostic ability in some pop- ulations. Whether PCA3 has a role in AS remains unclear, and its utility as a longitudinal measure has not been evaluated. We sought to evaluate the role of baseline and longitudinal PCA3 values in predicting higher-risk cancer among men in AS. METHODS: A total of 1511 men with favorable-risk (i.e. VLR or LR) cancer have enrolled in the Johns Hopkins AS program. Monitoring includes semiannual PSA and DRE, and annual prostate biopsy for most men; post-DRE urine specimens have been obtained since 2007. To assess the value of repeated PCA3 (Progensa PCA3, Hologic, Inc.) measurements, we identified patients with two urine samples obtained over an interval of three or more years. Initial (iPCA3) and subsequent (sPCA3) PCA3 scores were compared based on grade reclassification (GR, defined as any Gleason score >6) and high-volume (defined as >2 cores or >50% involvement of any core) or bilateral cancer during follow-up. A multivariate model was used to evaluate PCA3 score in conjunction with other pathologic parameters. RESULTS: In total, 294 men met study criteria; median time between PCA3 measures was 5 years (IQR 4-6). Thirty-five (12%) men underwent GR at a median of 7 years (IQR 5-9) after enrollment. Men who underwent GR had a higher median iPCA3 (49.6 vs. 25.5, p¼0.003) and sPCA3 (62.0 vs. 39.0, p¼0.004) than those who did not. While PCA3 score increased over time in the cohort (median iPCA3 vs. sPCA3, 27.1 vs. 41.0, p<0.0001), the rate of increase did not differ by GR status (log-normalized rate 0.07 vs. 0.08, p¼0.88). In a multivariate model, high-volume (OR¼3.09, p¼0.006) and bilateral (OR¼9.83, p<0.0001) disease significantly predicted GR while PCA3 score was not statistically significant (OR¼1.00, p¼0.13). Nonetheless, on uni- variate analysis PCA3 values predicted high-volume cancer (OR¼1.30, p¼0.006) and the presence of bilateral cancer (OR¼ 1.61, p<0.0001). CONCLUSIONS: In men with favorable-risk cancer on AS, PCA3 scores were associated with higher-risk cancer, while time- related changes in PCA3 score were not. PCA3 score did not inde- pendently predict GR, but did demonstrate a robust association with more extensive disease on biopsy. These findings suggest PCA3 may provide a non-invasive alternative to biopsy for identifying men at higher risk of reclassification during AS. Source of Funding: Reagent support provided by Beckman Coulter. MP15-10 THE PROGNOSTIC SIGNIFICANCE OF A NEGATIVE CONFIRMATORY PROSTATE BIOPSY ON PROGRESSION FOR PATIENTS ON ACTIVE SURVEILLANCE Vishnu Ganesan*, Charles Dai, Yaw Nyame, Daniel Greene, Nima Almassi, Joseph Zabell, Hans Arora, Sam Haywood, Alice Crane, Chad Reichard, Daniel Hettel, Anna Zampini, Ahmed El-Shafei, Robert Stein, Khaled Fareed, Michael Gong, J Stephen Jones, Andrew Stephenson, Eric Klein, Cleveland, OH INTRODUCTION AND OBJECTIVES: Obtaining a repeat prostate biopsy to accurately classify disease after the diagnosis of low risk prostate cancer is a common practice in the contemporary man- agement of men on active surveillance. It is estimated that 20-30% of patients are upgraded on a repeat biopsy. However, a second biopsy may at times yield no detectable cancer. The clinical significance of an entirely negative biopsy in this setting is still unclear. The objective of this study was to see compare outcomes on AS for patients who demonstrate a negative confirmatory biopsy METHODS: We retrospectively reviewed patients who were managed expectantly on active surveillance from 2002 to 2015. We selected patients with an initial diagnosis of Gleason 3+3 disease and a confirmatory biopsy within 1 year of diagnosis. The primary outcome was Type 1 and Type 2 progression, defined as an increase in the Gleason score, and a predominant Gleason pattern 4 or 5 or >50% involvement in cores, respectively. Kaplan-Meier survival curves were used to calculate 1-year and 3-year progression-free survival. On multivariable analysis, the association of negative confirmatory biopsy with progression-free survival was tested with Cox proportional hazards regression modeling. RESULTS: A total of 405 patients were identified who had an initial diagnosis of Gleason 3+3 disease and a confirmatory biopsy within 1 year. 155 (38%) of patients had a confirmatory biopsy with no disease. In the follow-up period, 128 patients (32%) of patients experienced a Type 1 progression and 34 (8%) of patients experi- enced a Type 2 progression. 1-year and 3-year Type 1 progression- free survival for men with and without a negative confirmatory biopsy was 100% vs % 63% and 96% vs 56% (p < 0.0001). For Type 2 progression, the 1-year and 3-year progression free survival for those with and without negative confirmatory biopsy was 100% vs 89% and 100% and 86% (p < 0.0001). On multivariable analysis adjusting for age, initial PSA, initial stage, and initial number of positive cores, the presence of a negative confirmatory biopsy was independently asso- ciated decreased risk of Type 1 progression (HR 0.7, 95% CI 0.6 to 0.9, p ¼0.008) and Type 2 progression (HR 0.12, 95% CI 0.03 to 0.36, p < 0.0001). CONCLUSIONS: More than a third of patients being considered for AS will have a negative confirmatory biopsy. We found that patients with these findings were less likely to progress while on AS. Taking into account initial clinical picture such as age and tumor burden, men who have a negative confirmatory biopsy may benefit from a less rigorous surveillance protocol. Source of Funding: None Vol. 195, No. 4S, Supplement, Friday, May 6, 2016 THE JOURNAL OF UROLOGY â e155