Abstract
Patients with highly mutated tumors, such as melanoma or smoking-related lung cancer, have higher rates of response to immune checkpoint blockade therapy, perhaps due to increased neoantigen expression. Many chemotherapies are known to be mutagenic, but the impact of standard treatment protocols on mutation burden and neoantigen expression in most human cancers is unknown. We sought to quantify the effect of chemotherapy treatment on computationally predicted neoantigen expression for 28 high grade serous ovarian carcinoma (HGSC) patients with chemotherapy-exposed samples collected in the Australian Ovarian Cancer Study. We find that relapsed HGSC tumors harbor nearly double the predicted expressed neoantigen burden of primary samples, but mutations associated with signatures identified from preclinical models exposed to cisplatin and cyclophosphamide account for only a small part of this increase. In both treated and untreated samples, most predicted neoantigens were associated with COSMIC Mutational Signatures 1, 3, and 8. Our analysis demonstrates a pipeline for extracting mutational signatures from preclinical models and quantifying the impact of specific mutational signatures on neoantigen burden.
