(MSP) assay to validate the results from the European MATLOC study. The cancer prevalence rate of the DOCUMENT cohort was adjusted to 18% to allow direct comparison. The predetermined analytical gene marker cutoffs from MATLOC were used to demonstrate clinical performance. RESULTS: The epigenetic assay resulted in an NPV of 88% (95% confidence interval (CI), 85-91%), a sensitivity of 60% (95% CI, 50-71%) and specificity of 64% (95% CI, 57-70%). Detection of cancer- associated DNA-methylation in the first biopsy was associated with an odds ratio of 2.33 (95% CI, 1.27-4.01; p ¼ 0.004) when corrected for other risk factors such as histopathology (p¼0.040), age (p¼0.046), PSA (p¼0.254) and race (p¼0.758). CONCLUSIONS: The DOCUMENT study validates this epige- netic assay as a significant, independent predictor for the absence of prostate cancer in a repeat biopsy in this US population. These findings are consistent with and confirm the results from the earlier European MATLOC study, demonstrating an NPV of 90% (95% CI, 86-94%). The consistently high NPV of the epigenetic assay is significantly higher than the use of standard histopathology alone. The addition of this test to other commonly known risk factors can be used to reduce unnec- essary repeat prostate biopsies. Source of Funding: MDxHealth MP63-15 THE PROFILE FEASIBILITY STUDY: GENETIC PROSTATE CANCER RISK STRATIFICATION FOR TARGETED SCREENING Rosalind Eeles*, Christos Mikropoulos, Pardeep Kumar, London, United Kingdom; Elena Castro, Madrid, United Kingdom; Elizabeth Bancroft, Natalie Taylor, Tohkir Dadaev, Zsofia Kote-Jarai, London, United Kingdom INTRODUCTION AND OBJECTIVES: Prostate cancer (PC) screening is controversial and a better assessment of individualized PC risk is needed. Several single nucleotide polymorphisms (SNPs) conferring a cumulative risk of PC have been identified. We have explored the potential role of genetic markers for targeted screening in a population with increased risk of PC due to family history (FH). METHODS: PROFILE was developed as a pilot study to determine the feasibility of targeted PC screening using prostatic biopsy (PB) and its association with specific genetic profiles in men with FH. We also evaluated the role of PSA and Diffusion Weighted MRI (DW- MRI) as screening tools. 116 men aged 40-69 with FH of PC were enrolled between October 2009 and December 2012. Cumulative SNP risk scores were calculated by summing 59 risk alleles for each locus using the weighted effect (log-additive model). DW-MRI was performed in 50 patients. Participants were asked to undergo a 10 core PB regardless of baseline PSA RESULTS: Median age 53 yrs (40-69) and median PSA was 1.15. One hundred and two men accepted to undergo a PB as pri- mary PC screening. Twenty-three tumours were found (22.5% of biopsies) as well as seven men diagnosed with atypical small acinar proliferation (ASAP) (6.8%) and 8 men with high-grade prostatic intraepithelial neoplasia (HG-PIN) (7.8%). In total 37.1% received an abnormal result. Of those diagnosed with PC, 41% were intermediate or high risk and required treatment (compared to 24% in general population screening). The predictive performance of DW-MRI, PSA, genetic model and genetic model plus PSA measured by AUC were: 0.85, 0.73, 0.57 and 0.74, respectively. The genetic model performed best in men with a ‘normal’ PSA (<3 ng/ml - AUC 0.63). Analyses of a 78 SNP profile, incorporating the 23 recently identified SNPs is underway. CONCLUSIONS: Our results indicate that PB used for PC screening in men with FH of PC identified a higher proportion of clini- cally significant PCs. The high AUC for DW-MRI would warrant a larger study. The incidence of ASAP is higher in this group than the general population. The SNP risk score was more predictive in men with PSA<3 where PB would not normally be undertaken, therefore an expanded study to investigate the role of genetic profiling in directing PB in PC screening is indicated. Source of Funding: The Ronald and Rita McAulay Foundation, Cancer Research UK, NIHR. MP63-16 IMPORTANCE OF EXTREME APICAL SAMPLING FOR PROSTATE CANCER DIAGNOSIS Ahmed Elshafei*, Ganesh Kartha, Yonghong Li, Cleveland, OH; Ayman S. Moussa, Cairo, Egypt; Tianming Gao, J. Stephen Jones, Cleveland, OH INTRODUCTION AND OBJECTIVES: To assess extreme api- cal sampling on prostate cancer (PCa) detection, plus determination of aggression and disease burden. Outcomes were compared in patients with standard risk vs. predicted high risk of a positive biopsy. METHODS: 3053 patients who underwent initial prostate biopsy from 2000 to 2011 were reviewed. 2521 underwent 12 core while 532 underwent 14 core sampling (2 extra cores from the extreme anterior apex). Patients with additional cores from other areas or additional cores obtained in response to DRE or hypoechoic findings were excluded. Patients were stratified into 2 groups; those with standard suspicion of PCa (elevated prostate specific antigen (PSA)  10ng/ml, normal digital rectal exam (DRE), and no suspicious lesion on trans- rectal ultrasound (TRUS)) and those with higher suspicion PCa (PSA > 10ng/ml and/or abnormal DRE and/or lesion on TRUS). RESULTS: Univariate analysis, as detailed in Table 1, shows the 14 core apical-focus scheme detected more cancer compared to the 12 core scheme in patients with a standard risk of cancer on biopsy (43.09% vs. 36.18%; P¼0.02). On multivariate analysis, PCa detection with 14 core sampling was more likely in all patients (OR 1.339, 95% CI 1.070-1.676) and in men with standard risk (OR 1.334, 95% CI 1.007- 1.769). Table 2. A greater median number of positive cores (3 vs. 2, p¼ 0.04) and a higher maximum cancer % per core (40% vs. 25% p¼0.002) was seen in the 14 core cohort when stratified to standard risk. Gleason7 was more likely to be detected with 14 cores in the standard risk group (55.6% vs. 45.2% p¼0.03). Differences in PCa detection (p¼ 0.53) and Gleason7 (p¼0.10) between 12 and 14 core sampling was not noted in the higher risk group. CONCLUSIONS: Extreme apical sampling increases aggres- sive cancer detection on initial biopsy, especially in patients with standard suspicion of PCa. This area should be intentionally sampled during prostate biopsy, although it may not be as important in those with high suspicion of PCa, where standard 12 core sampling may be sufficient. Table 1: Prostate cancer detection between different schemes stratified by the risk groups Prostate cancer detection 12 core scheme 14 core scheme P- value Overall patients 1055/2521 (41.85%) 246/532 (46.24%) 0.063 Patients with suspicion of having cancer 606/1675 (36.18%) 134/311 (43.09%) 0.02 Patients with higher suspicion of having cancer 449/846 (53.07%) 112/221 (50.68%) 0.53 Table 2: Multivariate analysis of clinico-demographic variables associated with prostate cancer detection on initial biopsy All patients baseline comparison between 12 and 14 core groups (n¼3053) Variables Odds Ratio 95% Confidence Limits P-value Number of Cores 1.339 1.070-1.676 0.0108 Age 1.048 1.036-1.060 <0.0001 e714 THE JOURNAL OF UROLOGY â Vol. 191, No. 4S, Supplement, Monday, May 19, 2014