Abstract 4034: Immunotherapy generates selective pressure to create an escape tumor with increased susceptibility to treatment with T cell based therapies

Host immune surveillance as a mechanism to promote tumor growth or to suppress tumor development is well studied. Less is known how immuno-selective pressure in the form of immunotherapies can instruct the immune system during this process. We hypothesized that the application of a T cell-dependent selection pressure in the form of a whole tumor vaccine would alter the immunogenic architecture in our transplantable murine melanoma model SB-3123. We generated an escape variant to test this hypothesis. We performed whole-exome and RNA-sequencing of the tumor line SB-3123 and two fresh tumors generated by subcutaneous implantation of SB-3123 to identify mutations and to assess levels of expression. There were 349 mutations shared between the tumor line and fresh tumors and less than 5 mutations were unique to each sample. We then investigated whether tumors that had undergone an additional round of host immunoeditingwould acquire different mutations. To test this, we generated a re-derived tumor cell line from a fresh SB-3123 tumor. Remarkably, the re-derived tissue culture line retained 333 mutations in common with the prior samples and produced only 2 unique mutations. To test the immunogenicity of SB-3123, we vaccinated mice with irradiated SB-3123 and administered a live tumor challenge two weeks after vaccination. This resulted in a profound vaccination response with mice being completely protected from tumor for over 200 days. However, one vaccinated mouse developed a recrudesced tumor at the site of implantation 40 days after challenge and a tumor line was created (SB-3123-esc). We asked whether SB-3123-esc represented an escape variant by immunizing mice with SB-3123 and then challenging with either SB-3123 or SB-3123-esc. Vaccination with SB-3123 did not protect against SB-3123-esc. We also performed sequencing on SB-3123-esc and discovered that 441 nonsynonymous mutations were common to both tumor lines while 80 were unique to SB-3123 and 40 to SB-3123-esc. Furthermore, we were able to determine that SB-3123-esc lost 2 immunogenic neoantigens but had acquired 5. In addition, we identified a murine T cell receptor with specific reactivity for the escape variant but not the parental tumor. Our results demonstrate that the expressed mutations of SB-3123 tumor did not change, regardless of whether we sequenced the tumor line, fresh tumors, or a re-derived tumor cell line. However, the escape tumor had new unique mutations and at least one of these was demonstrably immunogenic; we were able to develop a murine T cell receptor with specific reactivity for the escape variant but not the parental tumor. In conclusion, we posit that these observations offera proofof concept that the process of cancer immune evasion can create new opportunities to treat escape tumors with T cell based therapies. Citation Format: Jenny H. Pan, Suman Vodnala, Robert L. Eil, Zhiya Yu, Jared Gartner, David Clever, RahulRoychoudhuri, Shashank J. Patel, Christopher A. Klebanoff, Madhu Sukumar, Tori Yamamoto, Nicholas P. Restifo. Immunotherapy generates selectivepressure to create an escape tumor with increased susceptibility to treatment with T cell based therapies. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4034.