Stereotactic body radiotherapy and the tumor microenvironment in pancreatic ductal adenocarcinoma: from mechanistic remodeling to immunotherapy combinations

Abstract Pancreatic ductal adenocarcinoma (PDAC) remains largely refractory to immunotherapy. Stereotactic body radiotherapy (SBRT) delivers ablative, immunomodulatory doses that may render this immunologically inert tumor susceptible to immune control, although the relevant effects are hard to attribute to radiation alone. Here we synthesize how SBRT reshapes the PDAC microenvironment, applying throughout an evidence hierarchy that separates changes seen after SBRT given without concurrent immunotherapy from those arising in combination therapy. We propose, as a testable hypothesis rather than as an established conclusion, that SBRT produces a dose-tunable balance between activation and suppression rather than either alone. The activating arm is a real but spatially constrained immunostimulatory response, comprising antigen release, dendritic-cell cross-presentation, and cytotoxic T cell recruitment. The suppressive arm is a reversible, radiation-induced program on four coordinated axes: myeloid recruitment, regulatory T cell and transforming growth factor-β activation, desmoplastic stromal exclusion, and adaptive checkpoint induction. Dose and fractionation determine the balance point, but the two arms rest on unequal evidence: activation on resected human specimens, suppression on murine causality outside a single profiling study. The practical implication is that the most promising agent to add to irradiation is the one releasing whichever brake dominates in that tumor. Trial data are consistent with this. SBRT with pembrolizumab and trametinib prolonged survival in locally recurrent resected disease, whereas SBRT with nivolumab, with or without ipilimumab, did not in refractory metastatic disease. We close by outlining a three-layer biomarker strategy for patient selection—spanning spatial tissue architecture, tumor antigenicity, and circulating dynamics, and with the paired-sampling, stratified, and sequence-testing designs needed to adjudicate it.