Objectives Adoptive transfer of autologous Tcells specifically targeting neoepitopes derived from tumor mutations has achieved responses in melanoma and other cancers. We sought to identify ovarian cancer neoepitopes.

Methods We collected tumor, blood and ascites specimens from ovarian cancer patients. We combined mass spectrometry analysis of MHC class1 peptidomes with parallel sequencing of whole exome and RNA of cancer cells, to identify neoepitopes. Missense mutations were identified by whole exome sequencing and epitopes confirmed by mass spectrometry from ascites. We selected peptides predicted to have high binding affinity to HLA-A*02:01, for in vitro Tcell priming. We first validated the immunogenicity of these peptides for priming Tcells with synthesized peptides in healthy HLA-A*02:01+ donors.

Results 32 missense mutations were identified in patient’s ascites. 7 epitopes were confirmed by mass spectroscopy; 5 were predicted to have high binding affinity to HLA-A*02:01. 3/5 epitopes induced peptide-specific Tcell responses not cross-reactive with native sequences. However, these 3 peptides failed to induce autologous peptide-specific Tcell response in TILs to patient‘s ascites, nor to autologous tumor cells, suggesting immunosuppression. Interestingly, when autologous tumor cells were pre-treated with IFN-gamma, Tcell responses against tumor cells were observed.

Conclusions We demonstrated effective T cell responses against tumor neoantigens, is not only dependent on stimulation of Tcells, but also on efficient epitope processing and presentation on cell surface. While efforts have been focused on activation of Tcell responses, such as immune checkpoint blockade, our study suggests that strategies to modulate tumor cells to efficiently present neoepitopes should be considered for successful immunotherapy against neoantigens.