The inclusion and exclusion criteria for the study are shown below.
In this study, low risk (not on active surveillance) or intermediate risk men with prostate cancer were stratified by race and matched by CAPRA-score, and underwent Decipher molecular testing. Patients subsequently underwent standard of care treatment with either radiotherapy or surgery. The endpoint of the study was 2-year PSA failure, and multiple secondary endpoints were incorporated.
The first presented data point was the concordance between self-identified race and genomically inferred genetic ancestry from peripheral blood analysis. There was significant overlap between these two parameters, though 9% of self-identified AAM were classified as having non-AAM ancestry. Additionally, the concordance between biopsy specimen Decipher classification and prostatectomy Decipher classification was just over 80% in both AAM and non-AAM.
Patients in each cohort were well-matched with regards to various clinical parameters such as PSA, biopsy gleason score, NCCN risk category, and CAPRA score. AAM in this study tended to be younger than non-AAM. Patients with unfavorable or high risk by NCCN stratification as well as Decipher classifier tended to be older. Most notably, AAM classified as high genomic risk for metastasis by the Decipher classifier was significantly younger than similarly classified non-AAM.
Treatment type and use of ADT did not differ between the two groups, however, more AAM had pathologic upgrade at the time of prostatectomy relative to their biopsy pathology (25% versus 3.1% in non-AAM, p = 0.01).
When comparing Decipher risk classification and NCCN risk classification, more AAM classified as NCCN low-risk were likely to be classified as Decipher high-risk for metastasis (20% in AAM versus 0% in non-AAM).
To further understand the discordance between clinical risk stratification and genomic risk stratification, the authors calculated risk deviation specifically in the low and favorable-intermediate risk patients, stratified by race. A higher proportion of AAM were re-classified as high genomic risk for distanct metastasis than non-AAM (RR = 3.99, p = 0.02).
Dr. Yamoah concluded his talk by noting the significant age difference between AAM and non-AAM of six years in patients with high genomic risk for distant metastasis. NCCN risk categories, based on clinical parameters, are likely suboptimal in identifying subsets of AAM who are more likely to have clinically aggressive disease. Genomic stratification, whether by Decipher or other means, may help better identify these at-risk patients and partially address disparities in the care of AAM with prostate cancer.
Presented by: Kosj Yamoah, MD, PhD, Section head of Genitourinary radiation oncology and director of radiation oncology cancer health disparities research, H. Lee Moffitt Cancer Center and Research Institute, University of South Florida, Tampa, FL
Written by: Alok Tewari, MD, PhD, Medical Oncologist at the Dana-Farber Cancer Institute, at the 2021 American Society of Clinical Oncology (ASCO) Annual Meeting, Virtual Annual Meeting #ASCO21, June, 4-8, 2021