Prostate Cancer (PCa) is the most common malignancy in men but overtreatment is a major public health concern. To address this, Genomic Health Inc. in collaboration with academic and military hospitals validated a 17-gene relative expression assay to predict the probability of adverse pathologic features at the time of radical prostatectomy to improve the safety of active surveillance. Their Oncotype DX (ODX) assay improves the prognostic accuracy beyond standard clinical parameters at the time of biopsy. The ODX GPS is scaled from 0-100 and increases the number of men deemed eligible for active surveillance by decreasing the pre-test probability of adverse pathology and allows patients and providers to confidently elect active surveillance to avoid upfront treatment morbidity. ODX has been validated in largely European American (EA) populations but has yet to be adequately assessed in high-risk African American (AA) men. Our overarching hypothesis is that the prognostic accuracy of Oncotype DX varies by race and by socioeconomic status and genetic ancestry among AAs. This is significant since AAs have the highest risk of PCa and PCa mortality and are more likely to harbor adverse pathology in their prostate despite being eligible for active surveillance by clinical parameters. Risk of aggressive PCa varies across studies in AA men based on socioeconomic factors including insurance, education and income. Moreover, gene expression varies significantly between AA and EA prostate tumors. Given this, ODX can be dangerous for AAs if prediction performance is poorer or differs significantly between AAs and EAs or between subpopulations of AAs with clinically localized prostate cancer (CLPCa). To assess the prediction performance of ODX in AAs with CLPCa, we propose the following Specific Aims: (1) to compare the effectiveness of the ODX GPS as a predictor of adverse pathology in archived specimen from 100 AA and 50 EA men with CLPCa; (2) to compare the prediction performance of the GPS for adverse pathology in archived specimen from 150 AAs with CLPCa from divergent socio-demographic (income, education, insurance status) backgrounds; (3) to assess for variation in GPS prediction performance in 100 previously recruited AAs with CLPCa by degree of genetic European or African ancestry. Should this comparative effectiveness study reveal variations in prediction performance by race, socioeconomic status or genetic ancestry, a calibrated GPS score could be provided. Alternatively, if the ODX assay inadequately measures risk of adverse pathology we prevent its use in men of African ancestry. In either scenario we improve the safety of active surveillance for a high-risk PCa population.
The morbidity associated with the upfront treatment of prostate cancer has led to the promotion of active surveillance to delay or avoid treatment-related complications. However, active surveillance is grossly underutilized due the inability of standard clinical parameters to detect adverse pathologic features in the prostates of eligible men. Our protocol will improve the safety of active surveillance by evaluating the effectiveness of a genomics assay for the prediction of adverse pathology in high-risk African American prostate cancer patients.