Androgen deprivation therapy (ADT) is the Standard systemic therapy for men with advanced prostate cancer. Although initially beneficial, inevitably all patients will progress to the lethal castration-resistant state of the disease. Metastatic castration-resistant prostate cancer (mCRPC) affects approximately 1.500-2.000 men in Belgium every year. Cytotoxic chemotherapy and newer endocrine agents are available, but they are expensive and lack predictive treatment biomarkers. Additionally, there are currently no clear guidelines on treatment choice or treatment sequencing, which leads to a trial-and-error approach in the treatment of men with mCRPC. This will result in an unsustainable situation for prostate cancer care, accompanied with a major cost burden on health care. mCRPC is a heterogeneous disease, where different metastases may have a different molecular build-up, resulting in two types of resistance when treating this disease: intrinsic or acquired. Whereas intrinsic (or primary) resistance is already present before the start of therapy, acquired resistance develops eventually in all patients due to the selection pressure from the administered therapy. Both resistance mechanisms emphasise the need for treatment personalisation, which is the goal of the current project. Our previous Swedish-Belgian collaborative studies, supported by Kom Op Tegen Kanker, have demonstrated how liquid biopsies, such as circulating tumour cells and plasma-derived cell-free DNA (cfDNA) can be an excellent surrogate to interrogate these resistance mechanisms in a non-invasive way. We and others have explored the clinical validity of molecularly-guided therapy selection in mCRPC and have identified biomarker signatures that might help us guide treatment choice. However, the clinical utility of these signatures has not been demonstrated, and requires validation in the context of a randomised controlled trial (RCT). The proposed project application aims towards the initiation of the Prostate Biomarker (ProBio) trial, which is a Swedish- Belgian collaborative effort with several novel approaches. It is adaptive, multi-arm, open-label, multiple assignment, and a biomarker driven phase 3 RCT in patients with mCRPC. Here, we will analyse cfDNA as a real-time liquid biopsy using a prostate-specific biomarker panel, designed and validated to detect 1) mutations in 78 genes, 2) genomic structural rearrangements (GSRs) in 11 prostate cancer-associated genes, 3) genome-wide copy number alterations, 4) 63 microsatellites to infer microsatellite instability (MSI), 4) tumour mutational burden and 5) estimate the circulating tumour DNA (ctDNA) fraction. Using pre-defined biomarker signatures, focused on genomic alterations in AR and TP53, DNArepair deficiency and TMPRSS2-ERG fusions, we will select which new line of systemic therapy will be initiated, and comparing these patients to the standard-of-care. The statistical design of ProBio is novel, since the randomisation probabilities for a given experimental systemic therapy are subjective to change as the trial evolves and learns from prior experience. Secondly, ProBio implements a re-randomization of non-responding patients, providing essential insights in optimal treatment sequencing. The hypothesis is that treatment decisions based on molecular profiling of cfDNA will significantly prolong the progression-free survival (PFS), resulting in increased response rates, which in turn would translate into improved overall survival (OS). Additionally, secondary endpoints encompass Quality of life (QoL), heath economy and adverse event profile comparison between different trials arms. This project could result in the first validation of the clinical utility of the liquid biopsy in the standard-of-care in mCRPC.