AUA 2013 - Session Highlights: Development and validation of the biopsy-based genomic prostate score (GPS) as a predictor of high grade or extracapsular prostate cancer to improve patient selection for active surveillance

SAN DIEGO, CA USA ( - Active surveillance (AS) is gaining momentum in this era of prostate cancer over-diagnosis and overtreatment, and many efforts are on the way to better risk-stratify patients prior to the intervention.

Limitations of the currently practiced prostate-needle biopsy approach include under-sampling, disease multifocalit,y and tumor heterogeneity. Dr. Cooperberg and colleagues from UCSF conducted two large development studies: first to identify genes predictive of clinical recurrence and adverse pathology using the radical prostatectomy specimens, and second to confirm predictive value of these genes in biopsy specimens. They developed a 17-gene genomic prostate score (GPS), which was then validated to predict grade and stage in patients suitable for AS.

auaManually micro-dissected FPE prostate cancer tissue from two spatially distinct tumor specimens was obtained and gene expression quantitated by RT-PCR technique using both radical prostatectomy and prostate biopsy tissue. Findings were validated by assessing GPS on a single needle biopsy for each patient, and association with clinical recurrence and adverse pathology (high grade and/or pT3) was analyzed by Cox proportional hazards model or logistic regression.

The authors analyzed 732 candidate genes from 441 radical prostatectomy specimens and identified 288 genes that were potentially predictive of clinical recurrence independent of the Gleason pattern. Of these genes, 81 were selected for confirmation in the prostate needle biopsy cohort of low / intermediate-risk patients (N=167), and strong association with adverse pathology was demonstrated. Further analyses identified 17 genes encompassing multiple biological pathways (stromal response, cellular organization, androgen, proliferation and reference genes), and a GPS algorithm was developed. Using 395 patients who would be candidates for AS, biopsies with as little as 1 mm of tumor length were then used for validation and demonstrated strong prediction (p < 0.005) of high-grade and/or pT3 disease, after adjusting for CAPRA score or other standard pretreatment factors.

They concluded that the 17-gene GPS, designed to overcome tumor heterogeneity and biopsy under-sampling, has been validated as a biopsy-based predictor of high-grade and/or pT3 disease, thus providing much needed risk-stratification at diagnosis. Although the initial cost of this assessment may be prohibitive for wide-spread use at this time, GPS is potentially a very powerful tool to guide patient care and minimize overtreatment.

Presented by Matthew Cooperberg, Jeffry Simko, Sara Falzarano, Tara Maddala, June Chan, Janet Cowan, Cristina Magi-Galluzzi, Athanasios Tsiatis, Imelda Tenggara-Hunter, Dejan Knezevic, Frederick Baehner, Michael Kattan, Steven Shak; Mark Lee; Eric Klein; Peter Carroll at the American Urological Association (AUA) Annual Meeting - May 4 - 8, 2013 - San Diego Convention Center - San Diego, California USA

Reported for by Serge Ginzburg, MD

View Full AUA 2013 Coverage




email news signup