The Cancer Institute at St. Francis Hospital Oyster bay, NY
E. Obedian1,2, S. K. Cheng3, J. Restrepo1, N. Sheth1, and I. Faiena1; 1James J. Peters Bronx VA Medical Center, Bronx, NY, 2The Cancer Institute at St. Francis Hospital, East Hills, NY, 3Herbert Irving Comprehensive Cancer Center, Columbia University Irving Medical Center, New York, NY
Purpose/Objective(s): Genomic profiling with Decipher or Oncotype (GPS) are two of the most commonly utilized tests to enhance the National Comprehensive Cancer Network (NCCN) prostate cancer risk group and may guide clinicians on active surveillance, surgery, radiotherapy, and/or androgen deprivation therapy recommendations especially in the NCCN intermediate risk group setting. This quality improvement project seeks to assess the hypothesis that there is concordance between Decipher and GPS testing and to determine the primary driver of discordant results. Materials/
Methods: Between 9/2021 and 6/2022, Decipher and GPS scores were requested on the biopsy tissue of 21 prostate cancer patients all from a single hospital. 67% (14 patients) had successful genomic profiling of the same tissue for both Decipher and GPS testing, and all of these were NCCN intermediate risk patients. Patients were subclassified as NCCN favorable intermediate (FI) or unfavorable intermediate (UI) risk based on Clinical T stage, Gleason Score, percent positive cores, and PSA. GPS was stratified by low vs. high risk. Decipher was stratified by low vs. intermediate vs. high risk. The genomic profile scores were compared to each other and to the NCCN risk group. Results: Successful processing of tissue was achieved in 71% (15 patients) of the cohort for Decipher scores and 86% (18 patients) for GPS scores (p=0.14). Of the 14 patients with both GPS and Decipher scores, 5 (36%) were NCCN FI risk and 9 (64%) were NCCN UI risk patients. The overall concordance rate between GPS and Decipher was 64% (60% for FI and 67% for UI). In the 5 patients with NCCN FI, GPS was low for all (100%), while Decipher was low in 3 (60%), intermediate in 1 (20%), and high in 1 (20%). In the 9 patients with NCCN UI, GPS was low in 7 (78%) and high in 2 (22%), while Decipher was low in 4 (44%), intermediate in 2 (22%), and high in 3 (33%). GPS and Decipher scores were both low in 10 (71%) and both intermediate or high in 2 (14%). When GPS was low (n=12), Decipher was low in 7 (58%), intermediate in 3 (25%) and high in 2 (17%). When GPS was high (n=2), Decipher was also high. Conclusion: Both Decipher and Oncotype GPS testing enhanced prostate cancer risk stratification above and beyond the NCCN favorable intermediate or unfavorable intermediate risk group designations. Despite an overall concordance rate of 64%, discordant results were typically related to higher risk scores more commonly identified by Decipher than by Oncotype GPS. Though ongoing NRG randomized trials will analyze the results of intensification or de-intensification of prostate cancer treatment based on genomic profiling with Decipher, further study is necessary in patients with discordant genomic profile scores.
