Integrated Clinical-Grade genomic and pathology AI biomarkers for high-risk prostate cancer - Project Summary Each year 375,000 men die of prostate cancer, the majority having received an initial diagnosis of high-risk disease. The combination of radiation therapy (RT) and androgen deprivation therapy (ADT) have been the standard treatment for decades, but recently a phase III trial demonstrated that the addition of newer androgen receptor signaling inhibitors (ARSIs) to the standard therapy significantly improved metastasis-free survival. However, high cost and increased grade 3 toxicities with the addition of ARSIs suggests a need for more careful patient selection to maximize treatment efficacy while minimizing therapy side effects and reducing cost. This team has previously performed key studies validating the Decipher genomic classifier and the Artera artificial intelligence (AI)-based digital pathology tools as risk stratification approaches for patients with localized prostate cancer. This resulted in the designation of Level 1 (the strongest level of evidence) supporting the use of these classifiers in the 2023 NCCN Guidelines for Prostate Cancer Management and the use of these two classifiers independent of each other in the clinic. However, it is still unknown how they compare to each other, particularly in identifying patients who would benefit from the addition of ARSI to RT and standard ADT and whether integrating genomic and pathology AI classifiers can result in improved performance compared to each individual approach. This application will address these unmet needs by: 1) Validating genomic and pathology AI as prognostic biomarkers for high-risk prostate cancer patients; 2) Validating genomic and pathology AI as predictive biomarkers for the addition of ARSIs to RT + ADT; and 3) Developing and validating integrated genomic and pathology AI prognostic biomarkers and predictors of response to the addition of ARSI to RT + ADT. With access to pretreatment tissue samples from four international phase III trials (STAMPEDE, ENZARAD, ATLAS, and PREDICT-RT) and both the Artera AI and Decipher Genomic Classifier tools, this application is uniquely positioned to develop impactful and cost-effective prognostic and predictive clinical-grade biomarkers for management of men with high-risk prostate cancer. The CLIA-compliant platforms enable rapid clinical translation of biomarkers to personalize therapy and thus reduce cost and avoid harmful side effects among patients who might not benefit from addition of ARSI, thus providing a paradigm shift in prostate cancer treatment.
