Back to 2016 Annual Meeting
Association of PSA and Number of Cores Positive to Likelihood Of Adverse Pathology At Radical Prostatectomy Based On A 17 Gene Expression Assay
David Albala, MD1, Michael Bonham, MD, PhD2, Brooke Nightingale, BS2, Anson Tharayanil, BS2, Bela Denes, MD2, Megan Rothney, PhD2, Debbie McCullough, MS2, John Bennett, MPH2, Phillip Febbo, MD2, Athanasios Tsiatis, MD2
1Associated Medical Professionals, Syracuse, NY, 2Genomic Health, Redwood City, CA
BACKGROUND: The overtreatment of prostate cancer and underutilization of active surveillance (AS) in men with Gleason 6 (GS6) cancer on biopsy stems from the uncertainty with current risk instruments. Volume of GS6 tumor and PSA > 10 ng/ml are commonly thought to be correlated with risk and drive treatment decisions. The Genomic Prostate Score (GPS) is a 17-gene biopsy-based RT-PCR assay analytically and clinically validated as an independent predictor of adverse pathology at prostatectomy. We report on the impact of volume of GS 6 disease and PSA at biopsy on the biologic aggressiveness of the disease as measured by GPS.
METHODS: 1,055 pathology reports received at the Genomic Health Inc. clinical lab were retrospectively reviewed to record submitted Gleason score, number of cores positive, and PSA (if available on path report). The GPS was calculated from a single specimen for each case with an associated path report using the validated algorithm of 12 cancer-related and 5 reference genes; median GPS was calculated for each sub-group (≤2 cores positive, > 4 cores positive, PSA< 4 ng/ml, PSA 4-10 ng/ml, PSA>10 ng/ml).
RESULTS: 803 cases (76%) were submitted as GS6. 66% of cases utilized standard 12 core biopsy, 4 % of cases had less than 12 cores and 30 % of cases had more than 12 cores. The median GPS for GS6 cases with ≤2 cores positive was 24 (IQR 16.5 to 30.5) compared to a median GPS of 26(IQR 20 to 34) for 4 or more cores positive. After incorporating the GPS, risk refinement, a change in categorical risk assessment was seen in 24% of cases with ≤2 cores positive and 26% of cases with >4 cores positive. 294 of cases with GS6 also reported a PSA value. 25% of GS6 cases had a PSA <4, 67% were 4-10, and 8% were >10 ng/ml. No significant correlation was seen between PSA and GPS. The median GPS values for cases with a PSA<4, 4-10, and >10 were 25(IQR 16-31), 24(IQR 17-32), and 25(IQR 17-39), respectively. Risk refinement following GPS occurred in 48% of cases with a PSA<10 and 23% with a PSA>10.
CONCLUSIONS: The GPS results highlight the broad spectrum of tumor aggressiveness in a series of 803 biopsies containing GS6 independent of % cores positive and PSA. The degree of risk refinement among all GS6 cases independent of tumor volume or PSA highlight the utility of the GPS to provide predictive information beyond traditional clinical variables used for risk stratification in the management of men with GS 6 disease at the time of diagnosis.
Back to 2016 Annual Meeting