VAIL, CO USA (UroToday.com) - Introduction and Objective: A number of new diagnostic assays are being developed to improve risk stratification in prostate cancer (PCa), but their performance in “real-life” clinical practice remains to be determined.
The Genomic Prostate Score - GPS (Genomic Health, Inc.) is a biopsy-based 17-gene test that has been clinically validated as a predictor of favorable pathology (Cooperberg et al., AUA 2013), and is intended to help guide use of active surveillance for men with very low, low, and low-intermediate risk PCa. Here we report the early experience with the assay for prostate biopsy specimens tested in the Genomic Health Clinical Laboratory.
Methods: We report on the first 750 patient samples tested since the assay was made available on 5/8/2013 that met clinical and pathology submission criteria. NCCN Risk group classification was provided by submitting physicians. All submitted biopsy samples were centrally reviewed for Gleason score and tumor length at Genomic Health. The 17-gene assay (12 cancer-related genes and 5 reference genes) was performed, and the GPS was calculated.
Results: Of 750 samples that met submission criteria, 99% yielded a GPS result, including 96% of samples at the minimum acceptable tumor length (1 mm). The median patient age was 64 (range 42-89). 25%, 38%, and 30% of patients were classified by the submitting physician as NCCN very low, low, and low-intermediate risk respectively (7% no NCCN designation). A wide range of GPS values were observed within each NCCN risk group: range 3-67 very low, 2-67 for low, and 1-67 for low-intermediate. Similarly, GPS ranged from 2 – 67 for patients with 1 mm samples. In 33% of NCCN low-risk patients, GPS indicated a higher likelihood of favorable pathology than predicted by clinical risk alone, consistent with NCCN very low risk. Conversely, in 10% of NCCN low-risk patients, the assay predicted a lower likelihood of favorable pathology than expected by clinical risk alone, consistent with NCCN intermediate risk.
Conclusion: The specimens tested to date are contemporary very low, low and low-intermediate risk PCa cases with GPS score distribution very similar to that observed in the clinical validation study of the assay. A wide range of GPS values was observed in each NCCN risk group and in tumors with only 1 mm of tumor length, enabling clinically meaningful refinement of risk, which should help physicians and patients make initial treatment decisions more confidently. The high rate of analytical success (96%) with small samples enables broad use of the GPS even with limited available tissue.
Presented by Ketan K. Badani, Bela Denes, Athanasios C Tsiatis, Michael Bonham, H. Jeffrey Lawrence, Tara Maddala, Megan Rothney, Helen Bailey, Joseph M. Anderson, Vivian Tan, F.L. Baehner, Phillip G Febbo, and Gordon Brown at the 24th International Prostate Cancer Update - February 19 - 22, 2014 - Cascade Conference Center - Vail, Colorado USA
Genomic Health, Inc., Redwood City, CA USA
Department of Pathology, University of California, San Francisco, CA USA