THE 4KSCORE TEST PREDICTS HIGH GRADE …

1 THE 4 KSCORE TEST PREDICTS high GRADE PROSTATE CANCER ON BIOPSY AT PSA LEVELS < 4NG/ML Daniel W. Lin, MD1, Neal D. Shore, MD2, Stephen M. Zappala, MD3, Mohit Mathur, MD PhD*4, Jay R. Newmark, MD4, Grannum R. Sant, MD4 and Mitchell S. Steiner, MD4 1 Seattle, Washington, 2 Myrtle Beach, SC, 3 Andover MA, 4 Miami FL (Presentation to be made by Dr. Lin) Introduction: The clinical value of PSA for prostate cancer early detection has been questioned primarily due to its poor specificity for clinically relevant disease, namely high GRADE prostate cancer. This is especially true at PSA levels lower than 4 ng/mL, generally considered to be within the normal range. However, PSA values between 4 ng/mL have been referred to as the PSA Danger Zone because the risk of prostate cancer in men with PSA within this range is 15-18x higher compared to that in men with PSA< Moreover, a surprising number of otherwise healthy men with PSA less than 4 ng/mL are at risk for prostate cancer metastases and death up to 20 years later.

2 The 4 Kscore Test has been validated to predict an individual man s risk of high GRADE cancer on biopsy, with an area under the curve of This test incorporates: Total PSA (tPSA), Free PSA, Intact PSA, human Kallikrein-2, age, digital rectal examination findings, and previous biopsy information into an algorithm. The objective of this study was to see if the 4 Kscore Test can predict high GRADE (Gleason 7) prostate cancer in men with a tPSA < 4 ng/mL. Methods: In a large prospective study that enrolled 1,012 men who were previously scheduled for prostate biopsy at 26 sites across the United States, 348 men were identified with PSA < 4 ng/mL. The 4 Kscore was calculated, and subsequent biopsy results were retrieved.

3 We evaluated the accuracy of the 4 Kscore as a predictive model for patients with low tPSA by fitting a logistic regression model with high GRADE prostate cancer as the outcome. We examined the mean values of the components of the 4 Kscore in cases with low tPSA and high tPSA. We then examined the ability of the 4 Kscore Test to detect high GRADE cancer at different cutoff values. Results: Of the 348 men with tPSA < 4 ng/mL, 116 ( ) were diagnosed with prostate cancer, and of these 37 ( ) were diagnosed with high GRADE prostate cancer on biopsy. Using a 4 Kscore Test cutoff of , 34 of the 37 ( ) would have been recommended for biopsy; additionally, this cutoff would result in avoidance of 194 ( ) of the 348 biopsies.

4 Two of these 37 cases had a tPSA below 2 ng/mL, both of which had 4 Kscores above the cutoff. Of the 348 cases with low tPSA, only 3 ( ) would have a delayed diagnosis of high GRADE cancer. The 4 Kscore AUC for all cases is The AUC for cases with tPSA < 4 ng/mL is vs for cases with tPSA 4 ng/mL. Conclusions: The 4 Kscore Test (cutoff ) is effective as a predictive instrument in men with tPSA < 4 ng/mL, with a sensitivity of and specificity of The Receiver Operator Curve shows similar AUCs for the 4 Kscore irrespective of whether the tPSA levels are high or low, and the 4 Kscore Test maintains good discrimination for risk of high GRADE prostate cancer in this lower tPSA subpopulation.

5 Source of Funding: OPKO Health, Inc. NOVEL URINE MARKERS FOR DIAGNOSING AND MONITORING NON-INDOLENT PROSTATE CANCER Daniella Bianchi-Frias1, Ilsa M. Coleman1, John S. Banerji3, Mazen Alsinnawi4, Khanh Pham4, Claudio Jeldres4, Roman Gulati3, Jing Xia3, Scott A. Tomlins5,6 , Christopher R. Porter3, and Peter S. Nelson1,2,7 Divisions of Human Biology1, Clinical Research2, and Public Health Sciences3, Seattle, WA. Section of Urology and Renal Transplantation4, Seattle, WA. Departments of Pathology5 and Urology6, Ann Arbor, MI. Department of Medicine7, Seattle, WA. Presentation to be made by Mazen Alsinnawi Introduction: Prostate cancer (PCa) is the most common solid tumor in men. Active surveillance (AS) is an acceptable alternative to immediate treatment for low-risk, Gleason pattern 3 (GP3) PCa.

6 Furthermore, recent reports from Klotz et al in Toronto have demonstrated the outcomes of both intermediate risk, Gleason sum 7, and low risk Gleason 6 PCa in men on AS. Importantly men with intermediate risk fare significantly worse on AS. AS often entails significant risk of under-grading the tumor and repeated biopsies. We hypothesized that transcripts associated with high Gleason GRADE cancers are quantifiable in urine samples and reflect the presence of higher- GRADE non-indolent tumors. Our objective was to develop a qPCR-urine-based assay for the detection of occult high - GRADE cancer GP4 Vs. GP3 disease, in urine samples from men diagnosed with PCa. Methods: We performed a prospective IRB approved evaluation of 53 men undergoing radical prostatectomy (RP) for clinically localized disease.

7 In the operating room prior to RP men underwent a digital rectal exam (DRE) followed by collection of urine by catheterization. Final pathology reviewed by a single GU pathologist was used as the gold standard for determination of final histological GRADE . We had previously identified 4 consistently up-regulated GP4 transcripts (RELN, GRIN3A, RGS5 and LRNN1) Quantitative PCR (qPCR) was performed for the GP4 markers and normalized to KLK3 and RPL13A. By comparing transcriptional profiles of GP3 and GP4 cancers, we identified transcripts differentially expressed between these histologies. To evaluate the ability of the GP4 markers to discriminate between low and high - GRADE tumors, univariate and multivariate logistic regression analyses were performed and the overall significance determined using likelihood ratio tests .

8 Results: Each gene was an independent predictor of high - GRADE (Gleason score 4+3 vs. 3+4) prostate cancer (p< for each gene) with area under receiver operating characteristic curves (AUC) of , , , and When tested as a panel, the 4 genes were significantly associated with high - GRADE cancer (Primary pattern Gleason 4 disease) based on the overall likelihood ratio test (p= ). Furthermore, the 4-gene panel outperformed each gene alone, with an AUC of Conclusions: A novel 4-gene panel assayed by qPCR in men undergoing RP appears to be capable of discriminating clinically relevant high - GRADE (Gleason 4+3) tumors from low- GRADE (Gleason 3+4) tumors using urine sediments.

9 Further studies are needed to confirm these preliminary findings. DoD W81 XWH-13-1-0050 / PC110970 CCSG: 5 P30 CA0154704 RO1: 1 R01 CA181605 COMPARATIVE COST ANALYSIS FOR MANAGEMENT OF LOW RISK PROSTATE CANCER Franklin Gaylis MD, Kevin McGill, Catherine Ball, Hilary Prime, Renee Calabrese LVN, Paul Dato MD, Edward Cohen MD Purpose: Several options are available for the management of low risk Prostate Cancer (PCa). Current guidelines suggest that Active Surveillance (AS) be considered in the management of low risk PCa in addition to surgery and radiation therapy. In this study, we compare the actual costs of managing patients with low risk PCa using AS, Radical Prostatectomy (RP) and Radiation Therapy (RT) as well as patients who started on AS but progressed and received RP or RT.

10 Methods: 144 patients were diagnosed with low risk PCa in 2012 (according to NCCN guidelines) at Genesis Healthcare Partners (GHP). Data pertaining to both risk stratification and treatment were abstracted from both the clinical and practice management domains of the Allscripts Electronic Medical Record. Patients were excluded from analysis if they did not have at least three years of follow-up care or if their treatment was deviant from the three listed above ( receiving a prostatectomy then radiation therapy). The final cohort for analysis consisted of 75 patients. Cost data analysis started at the time of diagnosis and continued up to 3 years following the diagnosis for the period; 2012-2015.