Discovery through preclinical models of patient-derived xenografts in uro-oncology: MURAL PDX collection

The limited availability of clinically-relevant tumor models hinders the translation of new drug targets and promising therapeutics into human clinical trials. Identifying new treatments suitable for heterogeneous tumors that fail current therapies is required to facilitate more accurate patient-selection strategies. We established 58 patient-derived xenograft models (PDXs) from high risk prostate tumors ranging from diagnosis to death, with diverse pathologies and driver mutations. Bulk and single cell RNA sequencing revealed inter- and intra-tumoral heterogeneity among the PDXs, even within adenocarcinoma and neuroendocrine pathologies. PDXs with varying phenotypes could be also propagated and maintained as organoid cultures. From a subset of 18 research-ready PDXs, we performed combination drug testing with the poly ADP ribose polymerase (PARP) inhibitor, talazoparib, using a 1x1x1 experimental design (PDX clinical trial or PCT) to assess the responses to 7 treatments in 8 PDXs. We identified remarkable responders, where PDXs regressed following combination treatment with talazoparib and carboplatin, and confirmed the reproducibility of responses in expansion studies. The Melbourne Urological Research Alliance (MURAL) was formed to govern the distribution of this resource with the research community. We propose that this PDX collection will improve preclinical evaluation of treatment modalities and prioritise effective treatments for clinical trials.