Background Cancer immunotherapy is a promising new treatment approach that leverages the natural ability of certain immune cells, especially cytotoxic T lymphocytes (CTLs), to identify and eradicate tumour cells. Whilst cancer immunotherapy has proven remarkably successful in haematological cancers, its efficacy against solid tumours, like pancreatic ductal adenocarcinoma (PDAC), an aggressive and deadly type of pancreatic cancer, remains limited. This limitation arises because CTLs must migrate to and infiltrate the tumour mass to mount an effective anti-cancer response. However, PDAC presents an extensive desmoplastic tumour microenvironment (TME), which surrounds and protects the tumour mass from infiltrating CTLs.

Methods We developed a novel 3D in vitro model of PDAC that integrates key cellular components of the TME, including primary cancer-associated fibroblasts (CAFs) and primary CTLs. By embedding CAFs in collagen gel, we replicated some mechanical and biochemical properties observed in vivo, such as collagen remodelling and deposition, increased stiffness, and CAF-collagen crosstalk. Primary CTLs were included at different remodelling stages, allowing them to infiltrate and migrate inside these environments. High-speed confocal microscopy was employed to assess real-time CTL activation, migration patterns, and tumour infiltration efficiency.

Results We evaluated our in vitro model in comparison to in vivo murine tumour tissue. We found increased stiffness within the different remodelled environments, which was similarly observed between normal-to-PDAC in vivo. In poorly remodelled environments, biochemical cues influence overall CTL motility. In highly CAF-remodelled environments, we noticed a switch from biochemical to mechanical cues, with mechanical confinement impacting CTL viability, invasion and overall migration speed. This was further confirmed with Fasudil, a ROCK inhibitor currently in clinical trials, which improved CTL migration and infiltration by perturbating mechanical forces. To further assess the impact of CAF-remodelled environments on CTL functions, we embedded PDAC spheroids in CAF-collagen matrices. We proved that these matrices significantly impair CTL infiltration and eradication of PDAC tumoroids. This suggests that CAFs create an immunosuppressive environment for CTLs, potentially contributing to immune escaping mechanisms in PDAC.

Conclusions Our study provides new insights into the negative effects of the pancreatic tumour microenvironment on CTLs. CAFs create unfavourable environments for CTL viability and motility, shielding the tumour cells from immune detection and eradication. This underscores the need to consider the mutual interaction of CTL with the tumour microenvironment when developing new immunotherapeutic strategies for PDAC. Further, it also presents an inspiring opportunity for future research and the development of innovative solutions to overcome these challenges.