Emergence of a mixed CAF population by FAP-CD3 T cell engager limits therapeutic efficacy

Background: Fibroblast activating protein (FAP) expressing fibroblasts are an attractive target for cancer therapeutic depletion and while preclinical depletion shows success, previous modalities have unsuccessful clinical impact. Here, we wanted to comprehensively understand the tumor microenvironmental changes after FAP+ fibroblast depletion and unravel potential reasons for resistance and vulnerabilities that exist using a FAP targeted T cell engager. To unveil the complex changes that occur in the tumor microenvironment after FAP+ fibroblast depletion, we generated comprehensive single cell RNA-sequencing analysis of FAP+ CAFs depletion within the TME to understand the key population and genetic modulations in all cell subtypes.Methods: A CD3 T cell engager directed against FAP (FAP TcE) was utilized to deplete FAP+ fibroblasts in a preclinical murine model of pancreatic cancer. To understand complex population dynamics upon FAP TcE, we performed single cell RNA-sequencing of treated vs. untreated tumors to unveil population and genetic changes. Results: Administration of FAP TcE resulted in tumor growth control in vivo that was not dependent on T cell priming and egress via draining lymph nodes. After FAP TcE, T cell exhaustion was prevalent with increased T cell exhaustive state (Tex) and the emergence of a T cell progenitor exhausted state (Tpex), yet addition of anti-PD1 failed to enhance tumor efficacy. Within fibroblast populations, FAP TcE depleted FAP+ cancer associated fibroblasts (CAFs); however, depletion is compensated by an emergence of a “mixed CAF” population, potentially limiting the efficacy of FAP TcE. Conclusion: This study highlights the complex and plastic fibroblast changes occurring with stroma targeted therapies that may limit therapeutic efficacy.