Vascular and Stromal Cells Change Drug Responses in Synovial Sarcoma Organoids
A patient-derived synovial sarcoma collection was extended into a vascularized chip. Paired drug tests showed lower chemotherapy sensitivity in the more complex model and variable responses to adding anlotinib.
Extending a Rare-Tumor Collection
A Fudan University-led collaboration established synovial sarcoma organoids and added stromal and endothelial cells to examine how the tumor environment changes drug response. The study combined a patient-derived model collection, analysis of treated tumor tissue and a vascularized organoid-on-chip system. Avatarget collaborators established the organoids and carried out drug-sensitivity experiments; university investigators conceived, designed and supervised the study and performed statistical analyses.
The team generated 24 organoid lines from 30 patient samples, approximately an 80% establishment rate. Histology, tumor markers and genomic and transcriptomic comparisons supported their relationship to the original tumors. Conventional organoids provided a tumor-cell model, while the chip added compartments that could influence the activity of chemotherapy and antiangiogenic agents.
Adding Fibroblasts and a Vascular Network
Analysis of pretreatment and post-treatment tissue implicated endothelial and stromal compartments in the response to chemotherapy plus anlotinib. The investigators therefore combined patient-derived tumor organoids and cancer-associated fibroblasts with human umbilical vein endothelial cells. The endothelial cells formed interconnected, lumen-containing structures in a perfused chip; dextran distribution supported communication through the vascular-like network.
Anlotinib disrupted that network, reducing vessel area, junction number and total vessel length. These structural measurements gave the assay a vascular endpoint alongside tumor-cell viability. The endothelial cells were a reproducible laboratory source rather than endothelial cells isolated from each patient's tumor.
Comparing the Same Patients Across Model Formats
Matched drug tests showed that the vascularized models were less sensitive to doxorubicin, alone or with ifosfamide, than the corresponding conventional organoids. In three paired models, doxorubicin IC50 values were 0.497, 0.129 and 0.243 μM in conventional organoids, compared with 1.720, 0.840 and 0.517 μM in the vascularized format. The comparison showed that adding the surrounding cell populations changed the apparent pharmacology.
Adding anlotinib increased sensitivity in some vascularized models. For SS23 and SS26, doxorubicin IC50 values fell from 0.840 to 0.173 μM and from 0.517 to 0.183 μM, respectively. Responses remained heterogeneous. In a representative patient who changed from doxorubicin–ifosfamide to doxorubicin–anlotinib, the vascularized assay resembled the clinical pattern of improved tumor shrinkage after the regimen change.
Testing Additional Agents and Transcriptional Effects
Pazopanib and the histone deacetylase inhibitor vorinostat produced different responses in the models. Vorinostat reduced viability in a concentration-dependent manner, while pazopanib sensitivity varied. Expression analysis linked pazopanib to suppression of proliferation-related programs and vorinostat to stress, apoptosis and changes in SS18–SSX-associated transcription. Immune-associated expression signatures were measured in a system that did not contain a functional immune compartment.
The work provides a paired way to study tumor-intrinsic and stromal–vascular contributions to drug activity. The single clinical example was exploratory. The model used umbilical vein endothelial cells, and the cellular proportions were chosen for stable network formation. Altered drug penetration, stromal signaling and matrix remodeling remain possible explanations for its increased resistance; their individual contributions were not resolved by these experiments.