Models, methods and markers: The foundations of personalized oncolytic virotherapy for pancreatic cancer

Pancreatic ductal adenocarcinoma (PDAC) remains one of the deadliest cancers, with limited benefit from current therapies. This thesis investigated oncolytic virotherapies which are capable of selectively lysing tumor cells while activating antitumor immunity. Using various PDAC models, we showed that NDV induced robust infection, replication, and cell death in 2D monolayers and, importantly, in patient-derived organoids (PDOs), with variability that mirrors patient-specific tumor biology. While 3D spheroids highlighted how compact architecture can limit viral penetration, PDOs were more sensitive to NDV infection, replication and induced cell death and preserved the genetic and phenotypic features of the original tumors, making them a powerful model for evaluating NDV efficacy in vitro.
To understand patient-specific differences in OV sensitivity, we assessed four different viruses across ten PDAC organoids and linked their responses to baseline gene-expression profiles. This revealed that developmental and stem-cell–related transcriptional signatures were associated with increased sensitivity to all four viruses. In parallel, we engineered a new NDV variant by introducing two amino acid substitutions near the F protein of NDV that allows production in mammalian cells without the use of eggs. This allows acceleration of GMP-grade production of NDV for future clinical evaluation.
Finally, a literature review of two decades of clinical OV studies showed that treatment success depended on factors such as viral dose, route of delivery, intratumoral replication, and immune activation, but cross-trial result comparison is hindered by inconsistent parameter inclusion across trials. Altogether, this thesis provided experimental, molecular, and translational insights that support a more personalized development path for viro-immunotherapy in pancreatic cancer.