Over the last decade, several new drugs have been approved for metastatic castrate-resistant prostate cancer (mCRPC). Unfortunately, tumours invariably acquire therapy resistance, often within months. There is an immediate need to identify mechanisms of resistance and to develop novel therapies in the setting of mCRPC. The shortage of contemporary preclinical models of mCRPC makes this task challenging. Prostate cancer research is constrained by the limited choice of experimental models, which do not represent the spectrum of tumours in the clinic or current trends in clinical practice.
Using specimens from the CASCADE rapid autopsy program (Alsop et al Nature Biotech, 2016), we established patient derived xenografts (PDX) from men with mCRPC who had failed multiple lines of conventional therapies, including abiraterone acetate, enzalutamide, docetaxel and cabazitaxel. PDXs were generated from different metastases from each patient, providing new tools to study intra-patient tumour heterogeneity. Importantly, data from low coverage whole genome sequencing and RNAseq, showed that the PDXs faithfully recapitulate the genomic and transcriptomic features of the original patient specimens.
We further characterised the androgen receptor axis, which remains the main drug target for treating mCRPC. The PDXs exhibit castration-resistance due to a variety of mechanisms, including AR mutations, expression of constitutively active AR splice variants and loss of AR expression. Accordingly, the PDXs have varying sensitivity to different forms of androgen receptor blockade, such as enzalutamide and galaterone.
These PDXs expand the limited set of experimental models currently available for prostate cancer research. They are a new resource that encompasses tumour heterogeneity for rapidly testing new drugs and studying mechanisms of resistance.