Validation of a Novel EUS-FNB-Derived Organoid Co-Culture System for Drug Screening in Patients with Pancreatic Cancer

SIMPLE SUMMARY: Pancreatic cancer is a devastating disease with a 5-year survival rate of 12%. Only about 20% of patients are candidates for curative surgery, while the majority receive palliative chemotherapy or best supportive care. Recently, testing chemotherapeutical agents on patient-derived 3-dimensional cultures (organoids) of cancer cells has provided hope that personalized treatment may soon become a reality. However, most of these models only include tumor cells and not cells of the supporting tissue, which have been shown to play a critical role in cancer progression. In this study, we created a co-culture including both tumor and stromal cells from endoscopic ultrasound-guided biopsies and showed that an interaction occurs between the cell types in our model. It may therefore be a step towards better prediction of therapeutic response in the future. We also discuss the limitations of creating these types of models by using endoscopic biopsies from primary tumors. ABSTRACT: Cancer-associated fibroblasts (CAFs) have been shown to impact the chemosensitivity of patient-derived tumor organoids (PDTOs). However, the published literature comparing PDTO response to clinical outcome does not include CAFs in the models. Here, a co-culture model was created using PDTOs and CAFs derived from endoscopic ultrasound-guided fine-needle biopsies (EUS-FNBs) for potential use in drug screening applications. Co-cultures were established, and growth was compared to monocultures using image metrics and a commercially available assay. We were able to establish and expand validated malignant PDTOs from 19.2% of adenocarcinomas from EUS-FNBs. CAFs could be established from 25% of the samples. The viability of PDTOs in the mixed cell co-culture could be isolated using image metrics. The addition of CAFs promoted PDTO growth in half of the established co-cultures. These results show that co-cultures can be established from tiny amounts of tissue provided by EUS-FNB. An increased growth of PDTOs was shown in co-cultures, suggesting that the present setup successfully models CAF–PDTO interaction. Furthermore, we demonstrated that standard validation techniques may be insufficient to detect contamination with normal cells in PDTO cultures established from primary tumor core biopsies.