Longitudinal Imaging and Liquid Biopsy Markers of Prostate Cancer Radiotherapy Outcomes - ABSTRACT For intermediate and high risk prostate cancer patients who are candidates for radiotherapy (RT), the only recognized early clinical response indicator is serum PSA level. There are some risk features used for treatment intensification, but optimized criteria are lacking for recommendations concerning the use and length of androgen deprivation therapy (ADT), and the use of additional systemic agents (e.g., abiraterone, chemotherapy). Approximately 30-40% of patients diagnosed with unfavorable intermediate to high risk nonmetastatic prostate cancer will develop biochemical or clinical progression within 5 years and about half of men with distant metastasis from prostate cancer originated from these groups. There is a critical unmet need to better define risk of progression such that treatment intensity is optimized and progression prevented. We propose to define the role of quantitative multiparametric MRI (mpMRI) features in combination with Circulating Tumor Cells (CTCs) as early markers of patient outcome after RT ± ADT in unique longitudinal clinical studies strategically carried out at the University of Miami and in an external radiotherapy validation cohort from the NCI. Our proposal addresses the problem of prostate tumor heterogeneity by applying an mpMRI 10 point pixel-by-pixel automated habitat risk score (HRS) classification system based on spatially distinct tumor/microenvironment characteristics. The habitat concept was introduced to map and adjust for tumor heterogeneity. The established “absolute” scale of HRS is ideal for monitoring changes longitudinally while reducing MRI vendor effects. In addition, we will apply more detailed quantitative features in models that link the changes in habitat composition and radiomics to liquid biopsy information and clinical outcome measures. The primary early endpoint is freedom from biochemical nadir thresholds at 9 months (FFBN9mo). The secondary longer term endpoint is time to Biochemical and/or Clinical Disease Failure (BCDF). Our group has recognized that liquid biopsy measures have the potential to improve the appraisal of pretreatment risk and longitudinal response to treatment. As such, CTCs, quantified using a microfiltration system, have been measured in our clinical trials at intervals before and after RT±ADT. In this application, we will leverage the unique clinical power of several mature and ongoing prospective RT trials in which mpMRI features (Aim 1), and CTCs (Aim 2) are collected pre-RT and at four time points post-RT. To our knowledge, these trials constitute the largest and most exhaustive resource for investigation of the relationship between quantitative imaging features, pre-Tx tissue genomics, liquid biopsies and RT treatment outcomes. The final goal of the proposal is to develop a multivariable model (Aim 3) that incorporates pretreatment and longitudinal mpMRI features and CTCs to guide physicians’ decisions about intensification and de-intensification of radiation therapy. These models will be validated in a prospective clinical trial (the Miami UAdapt trial).