AbstractThe successes of autologous CAR T cell therapies in treating hematologic malignancies has spurred interest in applying the same approaches for treating solid tumors. However, challenges associated with antigen heterogeneity, CAR T cell exhaustion, cytokine release syndromes, high production costs, and manufacturing complexities have limited their use in solid tumors to date. To address these issues, we have developed a universal CAR expressing a single ScFv against fluorescein (FL) that together with fluorescein-containing bispecific adapters (tumor tags, TTs) can mediate engagement of the universal CAR T cell with multiple other cell types. The TTs that we currently use include UB-TT170 that targets folate receptors (FR) on cancer cells and tumor infiltrating myeloid cells, UB-TT4xx (prostate-specific membrane antigen [PSMA]), UB-TT2xx (carbonic anhydrase IX [CA IX]), and UB-TT5xx (fibroblast activation protein [FAP] on cancer-associated fibroblasts [CAFs]). In this study, we first test various combinations of TTs in a KB tumor model, where FRα and CA IX are expressed on the cancer cells, and FAP is expressed on CAFs. Results from these studies reveal that concurrent administration of two TTs significantly enhances tumor regression relative to injection of only a single TT. Considered together with previous data showing that optimization of TT dosing frequency and concentration can avoid problems associated with CAR T cell exhaustion and excessive cytokine release, we conclude that universal CAR T cells warrant serious consideration as viable alternatives to fixed specificity CAR T cells for treatment of solid tumors. Another challenge limiting CAR T therapies is their high cost of production. To address this limitation, we have employed a surface-engineered lentiviral particle, VivoVecTM, that is currently undergoing development for production of the CAR T cells in vivo to assess whether they can be exploited to generate human CAR T cells by simple incubation in vitro. For this purpose, we have incubated the viral particles for only 1 hour with unmodified human PBMCs and then immediately injected PBMC suspension into MDA-MB-231 tumor-bearing mice. Importantly, injection of these modified PBMCs into NSG tumor-bearing mice led to eradication of their MDA-MB-231 tumors, but only when the mice were subsequently injected with an appropriate TT. Taken together, the data demonstrate that our universal anti-FL CAR T cell can be produced by brief co-incubation of a lentiviral particle with unfractionated PBMCs and that the resulting unprocessed PBMCs can be immediately exploited for eradication of MDA-MB-231 tumors in NSG mice when combined with the appropriate bispecific adapter (TT). The data further suggest that with additional improvements heterogeneous solid tumors might be treated with in vivo manufactured universal CAR T cells if they are co-administered with appropriately selected tumor tags (TT).Citation Format:Suilan Zheng, Bo Huang, Sudarshan R. Kasireddy, Yashapal Singh, Haiyan Chu, Madduri Srinivasarao, Laurie Beitz, Byoung Ryu, Andrew Scharenberg, Philip Low. In vivo generation of universal CAR T cells for targeting heterologous antigens in solid tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4798.