Introduction

Prostate cancer (PCa)-associated mortality results from metastasis to bone and resistance to androgen deprivation or cytotoxic therapy. Despite early detection of primary PCa, advanced castration resistant prostate cancer (CRPC) and bone metastases (BM) are detected in 10% of patients already at the time of initial diagnosis. The majority of recurrences might be due to therapy resistant cancer cells with stem cell-like properties (cancer stem cell-like, CSC-like). CSC-like cells are highly tumorigenic and metastatic, however, current treatments target the differentiated tumor bulk cells. Our aim is to determine whether CSC-like cells from metastases tissues are responsive to the standard compounds used for the treatment of the primary tumor type.

Methods

To model metastatic CSC-like cells we have generated organoids (here termed as &[raquo]canceroids&[laquo]) from patient-derived biopsies and established patient-derived xenografts (PDXs). Cytotoxic compounds and androgen inhibitors are being tested on the canceroids from different BM or LN tissues in order to determine whether agents targeting the primary tumor are effective in the metastases. Viability assays and light sheet microscopy imaging are used.

Results

We have successfully derived canceroids from bulk tumor tissues, in particular; primary PCa, bone metastasis from PCa tissue (BM-PCa), established metastatic PDX models LAPC9 and BM18 and lymph node (LN) metastasis from PCa (LAPC4) PDX models. Canceroids maintain key features of the original tumor; the known luminal phenotype of BM18 is evidently maintained in the BM18 canceroids, based on positive cytokeratin (CK)18 and absent CK5 expression. LAPC4 organoids contain CK5 and CK18 cells, in line with the mixed basal and luminal phenotype. BM18 and LAPC9 canceroids are maintained in presence and absence of dihydrotestosterone in order to determine whether they adjust to androgen (in)dependent status of growth. Ongoing studies involve growth response of BM-PCa, BM18, LAPC4 and LAPC9 canceroids to chemotherapeutics (cabacitaxel, docetaxel) and hormone inhibitors (abiraterone, enzalutamide).

Conclusions

Identification of the oncogenic properties of metastatic CSC-like (sub)populations has both prognostic and therapeutic applications. Establishment of CSC-derived organoids from BM tissue is the first step towards routine derivation from metastasis or primary PCa tissues as a potential platform for personalized drug compound evaluation.

Funding

Swiss National Funds