Pancreatic ductal adenocarcinoma (PDAC) remains one of the deadliest cancers, driven by late detection, metastatic potential, and a heterogeneous, highly desmoplastic tumor microenvironment (TME). The PDAC TME contains cancer-associated fibroblasts (CAF), pancreatic stellate cells (PSC), immunosuppressive myeloid populations, and regulatory lymphocytes that collectively promote immune evasion. CAF and PSC produce extracellular matrix and secrete immunomodulatory cytokines and chemokines such as TGF-β and CXCL12, contributing to a dense stroma and spatial exclusion of cytotoxic lymphocytes. Tumor-infiltrating T-cells in PDAC commonly display an exhausted phenotype marked by elevated inhibitory receptors including PD-1, LAG-3, and CTLA-4, with impaired proliferation and effector functions. These features limit responses to immune checkpoint inhibitors (ICI) in most PDAC patients.
The authors aimed to recapitulate key, simultaneous mechanisms of T-cell dysfunction in PDAC within a controlled, scalable, animal-free 3D platform. This approach was designed to model both stromal-mediated spatial exclusion and checkpoint-driven exhaustion to enable mechanistic dissection and functional testing of immunotherapeutic interventions, including patient-specific responses.
Three-dimensional spheroid systems were established to mimic PDAC architecture at increasing complexity. Mono-culture spheroids were generated from the human PANC-1 cell line. Co-culture spheroids combined PANC-1 cells with primary pancreatic stellate cells (PSC) to recreate a fibrotic stromal compartment. Patient-derived spheroids (PDS) were produced from enzymatically dissociated PDAC specimens and contained primary tumor cells together with CAF isolated from resected tumors.
Spheroid formation used ultra-low attachment plates with small amounts of Matrigel to support 3D aggregation. PANC-1 mono-cultures used standardized cell numbers per well; co-culture and PDS methods followed established differential trypsinization and culture conditions to expand tumor and stromal components prior to spheroid assembly.
Peripheral blood mononuclear cells (PBMC) were obtained from anonymized healthy donors and from PDAC patients when autologous assessments were required. T-cell isolation employed Ficoll-Paque separation followed by MACS selection for CD4 and CD8 subsets. Purified T-cells were maintained in RPMI with IL-2 and activated/expanded using CD3/CD28 Dynabeads at a 1:2 bead-to-cell ratio prior to spheroid infiltration.
Infiltration experiments introduced primary human T-cells into preformed spheroids to evaluate entry into tumor cores, checkpoint expression changes, and effector function within the 3D microenvironment.
Resected PDAC tissue and blood were collected with informed consent under institutional ethics approvals. Tumor specimens were minced and enzymatically digested using collagenase IV and DNase, mechanically dissociated, filtered, and cultured to expand tumor cells and CAF. Patient-derived CAF were characterized by αSMA staining; tumor cells were validated by epithelial and lineage markers (e.g., pan-cytokeratin, E-cadherin, EpCAM). Tumor-infiltrating immune cells were isolated from digests via MACS with CD45 selection for downstream flow cytometry and comparison to spheroid-infiltrated T-cells.
The study demonstrates that inclusion of PSC produced a fibrotic barrier around tumor cells that markedly restricted T-cell infiltration into spheroids, modeling the desmoplastic TME of PDAC. Mechanistically, CAF- and PSC-derived CXCL12 signaling through CXCR4 on T-cells promoted chemotactic retention of lymphocytes in stroma-rich regions and contributed to tumor-core exclusion.
Pharmacologic blockade of CXCR4 with AMD3100 significantly enhanced T-cell infiltration into PSC-containing spheroids, supporting the role of the CXCL12–CXCR4 axis in stromal immune exclusion within this 3D platform.
T-cells that infiltrated spheroids acquired a pronounced exhaustion signature. Flow cytometric and phenotypic analyses showed strong upregulation of inhibitory receptors including PD-1, LAG-3, and CTLA-4, closely mirroring checkpoint expression observed on tumor-infiltrating lymphocytes isolated from PDAC patient samples. These changes were associated with reduced effector markers and diminished functional capacity, consistent with the exhausted T-cell state reported in clinical PDAC specimens.
Interventions targeting stroma- and checkpoint-mediated suppression were tested within the spheroid systems. CXCR4 inhibition with AMD3100 increased T-cell access to tumor regions in PSC-containing spheroids, countering stromal exclusion. Separately, treatment with the anti-PD-1 monoclonal antibody pembrolizumab partially restored effector cell functions of infiltrated T-cells within the 3D culture, demonstrating that checkpoint blockade can reverse aspects of functional exhaustion in this platform.
Combining stroma-targeting and checkpoint-blocking strategies within the model recapitulates clinically relevant mechanisms and responses and allows functional assessment of therapeutic combinations in a controlled, patient-relevant context.
The minimalistic 3D spheroid platform captures cytokine- and stroma-driven immunomodulation typically observed in advanced organoid or in vivo systems while remaining scalable and animal-free. Patient-derived spheroids (PDS) reproduced individual patients’ T-cell suppression patterns and therapeutic responsiveness, indicating potential utility for personalized functional testing. By modeling both spatial exclusion and checkpoint-mediated exhaustion concurrently, this system offers a mechanistically faithful tool to dissect TME-driven immune suppression and to accelerate evaluation of combinatorial or patient-tailored immunotherapies for PDAC.
The authors identify stromal CXCL12–CXCR4 signaling and checkpoint upregulation (notably PD-1) as critical determinants of T-cell dysfunction in PDAC. Their 3D spheroid models, including PDS, reproduce key hallmarks of the PDAC immune landscape—T-cell exhaustion, stromal exclusion, and therapeutic responsiveness—and provide a scalable, animal-free platform to investigate mechanisms and test immunotherapeutic strategies. The system supports functional assessment of CXCR4 antagonists like AMD3100 and PD-1 blockade with pembrolizumab, and may inform development and personalization of combination therapies in PDAC.