Cholangiocarcinoma (CCA) is an aggressive bile duct malignancy with poor five-year survival. Improving preclinical prediction of therapeutic efficacy requires models that retain the native tumour microenvironment. The authors aimed to establish and characterise a patient-derived ex vivo model using precision-cut tissue slices (PCTS) from surgical resections, evaluate its behaviour in culture, and test its utility for pharmacological studies.
PCTS were produced from tumour specimens collected from 25 patients undergoing curative-intent resection for presumed CCA between 2022 and 2024. Slices were cultured ex vivo for up to 15 days. Serial assessments included measures of tissue viability and preservation of histological architecture.
Quantitative proteomics using SWATH-DIA was performed on samples taken at Days 0, 3, 7 and 15 of culture. Proteomic analysis quantified protein abundance changes across the culture period.
A subset of PCTS underwent pharmacological testing. Agents applied included the broad-spectrum apoptosis inducer staurosporine and clinically relevant chemotherapy agents: 5'-deoxy-5-fluorouridine, gemcitabine with and without cisplatin. The source reports dose-dependent and variable responses but does not provide further in-paper numerical details of specific response magnitudes.
PCTS maintained viability through Day 15 of culture, with no significant reduction from baseline viability reported by the authors. Histological evaluation demonstrated retention of tumour architecture over the culture period. Tumour cells remained cytokeratin-19–positive, supporting continued epithelial tumour identity.
Resident immune populations persisted in the slices during culture: CD3+ T cells and CD68+ macrophages were detected across the time course, indicating that key components of the native microenvironment were preserved at least over the 15-day culture interval.
Proteomic profiling quantified 4,578 proteins across sampled time points. The proteomic response to ex vivo culture followed a staged pattern: the number of differentially abundant proteins increased from 8 at Day 3 to 127 at Day 7 and 180 at Day 15, after which changes stabilised according to the authors.
The composition of proteomic changes included an early loss of proteins associated with inflammation and stromal components, a sustained cellular stress response, and evidence of metabolic reprogramming occurring during culture. Despite these culture-induced changes, proteomic signatures that distinguish CCA subtype identity were preserved throughout the 15-day period, indicating maintenance of subtype-specific molecular features in the PCTS model.
The authors tested the pharmacological responsiveness of PCTS to both a broad apoptosis inducer and clinically relevant chemotherapy agents. Staurosporine produced dose-dependent cytotoxicity in the tissue slices, demonstrating that the model can show expected direct cytotoxic effects.
Responses to chemotherapy agents (5'-deoxy-5-fluorouridine, gemcitabine ± cisplatin) were variable across tested slices. The report indicates heterogeneity in ex vivo chemotherapy sensitivity but does not provide detailed per-slice response rates or statistical summaries in the source text provided here.
These findings suggest that PCTS may serve as a platform to assess personalised or tumour-specific drug responses while retaining aspects of the tumour microenvironment that can influence therapeutic effect.
Patient-derived CCA PCTS maintained viability, tumour architecture, cytokeratin-19 expression, and resident immune cells for at least 15 days in culture. Proteomic analysis showed a defined, staged response to ex vivo culture including early loss of inflammatory and stromal proteins, sustained stress responses, and metabolic changes, while CCA subtype proteomic signatures remained intact. The slices also demonstrated pharmacological responsiveness, with clear staurosporine cytotoxicity and variable responses to clinical chemotherapy agents.
The authors declare no competing interests. Funding was provided by the National Centre for the Replacement Refinement and Reduction of Animals in Research and North West Cancer Research. The source article is a preprint on bioRxiv; further methodological or numerical details beyond those reported here were not provided in the source text.