Irradiation Induced Neoantigens Sensitize Cancer to Checkpoint Immunotherapy
Danielle M Lussier1, Elise Alspach1, James M White1, Cora D Arthur1, Gavin P Dunn1, Elaine R Mardis3, and Robert D Schreiber1*
1Department of Pathology and Immunology, School of Medicine, Washington University in Saint Louis
2Department of Neurosurgery, School of Medicine, Washington University in Saint Louis
3McDonnell Genome Institute, School of Medicine, Washington University in Saint Louis
Immunotherapies are a promising advance in treating several tumor types, shown to lead to long, durable responses in subsets of patients with tumors, such as melanoma, non-small cell lung cancer, and kidney cancers, that display high mutational landscapes and thus are a rich source of tumor specific neoantigens. However, even with the success of these new therapies, only a subset of cancer patients, especially those with tumors that display high neoantigen loads, benefit from these treatments. Given that irradiation is capable of inducing mutations in DNA, we hypothesized that the use of clinically relevant doses of irradiation may increase the number of mutational neoantigens within poorly immunogenic tumors rendering them more sensitive to immunotherapy via increasing tumor neoantigen load. To test this hypothesis, we subjected a KRAS+/-p53-/- oncogene-driven sarcoma line, we previously characterized as being devoid of neoantigens and nonimmunogenic, to in vitro irradiation and then generated clones from the irradiated tumor line. The individual clones were inoculated into cohorts of syngeneic immunocompetent mice or Rag2-/- immunodeficient mice, and we monitored tumor growth in the presence or absence of checkpoint blocking monoclonal antibodies (a combination of a-CTLA-4 plus a-PD-1). Several KRAS+/-p53-/-irradiated clones showed increased immunogenicity via complete rejection or delayed tumor outgrowth in WT recipients. Using genomic and bioinformatics approaches, we then identified irradiation-induced mutations arising in each clone and predicted which of these were potential tumor-specific neoantigens. As expected, the predicted mutant neoantigens in each clone were unique to that particular clone. The identities of functionally immunogenic neoantigens were validated by staining tumor infiltrating lymphocytes with a family of MHCI-peptide tetramers carrying peptides based on the epitope predictions. Overexpression of validated neoantigens into the nonirradiated nonimmunogenic parental KRAS+/-p53-/- oncogene-driven sarcoma line was sufficient to render it sensitive to checkpoint blockade immunotherapy in vivo. To test whether our approach was clinically relaveant, we irradiated mice bearing parental nonimmunogenic KRAS+/-p53-/- oncogene-driven tumors and found that this treatment resulted in generation of immunogenic tumors that responded to checkpoint blockade immunotherapy. These results demonstrate that irradiation can indeed increase tumor immunogenicity through de novo generation of nascent tumor specific mutant neoantigens in tumors that are typically nonresponsive.