Abstract B33: Neoepitope peptide vaccines and immune checkpoint blockade in a new preclinical ovarian cancer model

Abstract

Abstract Objective: Therapeutic success currently seen with immune checkpoint blockade (ICB) occurs primarily in immunogenic tumors and can be partly explained by the high rates of non-synonymous genetic mutations (single nucleotide variants, SNV) in tumor cells, leading to neoepitope presentations and robust T-cell infiltration. Despite showing mutations in fewer genes (most frequently TP53), high-grade serous ovarian cancer (HGSOC) displays signs of T-cell inflammation and responds, albeit modestly, to ICB. We focus on immune therapy tailored to the genomic characteristics of the tumor, especially to structural aberrations (SAs), like gene fusions. Our hypothesis is that SAs that are expressed as proteins are antigenic and immunogenic, and represent a source of neoantigens for combination immunotherapies that can overcome locoregional immune suppression. DNA structural aberrations can result in intergenic fusions and intragenic rearrangements, such as exon duplications (dup) and deletions (del). Here we focus on identifying gene fusion and exon dup/del-derived neoepitopes based on RNA-seq data for a combination of retrospective and prospective sample collections. Our hypothesis is that these gene rearrangements resulting from SVs may contribute to the immunogenicity and antigenicity of HGSOC. Methods: FusionCatcher was used to evaluate RNA sequencing data for in-frame gene fusions events in banked human HGSOC. HLAminer was used to determine individual haplotypes. Upon detection of fusion events (validated via PCR) we identified the MHC-I and MHC-II restricted peptides with predicted high or intermediate (IC50<500nM) binding affinity. The peptides (with sequences across the fusion point) were synthesized and pulsed onto HLA-matched PBMCs, for in vitro T cells stimulation assay. Following antigen priming and in vitro restimulation, IFNγ production by T cells was measured by ELISA. The efficacy of SA-derived neoepitopes was tested in vivo with two new preclinical mouse models based on MKP-Lung and MKP-Liver cells ovarian cancer cell lines. RNA-seq data of their complete mutanome were used to generate neoepitopes. RNA from normal mouse OSE isolated from healthy mice was used as reference for mutation calling. For vaccine generation, peptides were pulsed onto dendritic cells matured into Th1-inducing phenotype (DC1). Mice were treated with Rat IgG, anti-PDL1, or vaccine + anti PD-L1. Objective response was measured by tumor volume. Results: Our results demonstrate neoepitopes-specific human T-cell responses in vitro. RNA-seq data were available for nine patients; eight MHC-I and five MHC-II neoantigens were synthesized from gene fusions from three different patients represented by six sets of haplotype specific PBMCs. In vitro, neoantigen-treated PBMCs showed increased cell clumping. An objective IFN gamma response on ELISA was seen for six peptides at two different concentrations and five peptides at one concentration. Analysis of RNA-seq of murine tumors shows that the metastatic MKP-Lung cells (which carry molecular signatures associated with recurrent ovarian cancer), have an increased number of mutations (SNVs, indels) including gene fusions, in line with findings from human tumors. Using in silico prediction tools, we identified 16 potential peptide binders (IC50< 500), two with strong and 14 with moderate binding. The transcript variants were validated via PCR. The vaccine comprised the top seven MHC-I restricted peptide candidates with IC50<500Nm and one large (MHC-II) candidate peptide for CD4 T-cell responses. No difference was seen between treatment groups (ANOVA p=0.35). Conclusion: Our results demonstrate that PBMC pulsed with a combination of short and large peptides release increased IFNγ, supporting the proposed vaccine design. Structural variants contribute to tumor immunogenic potential. Immunogenicity data from this project will be further used to refine the target selection for individualized, mutanome-based vaccines. Citation Format: Malcolm S. Ross, Ma Tianzhou, Lixin Zhang, Nolan Priedigkeit, George Tseng, Adrian V. Lee, Robert P. Edwards, Anda M. Vlad. Neoepitope peptide vaccines and immune checkpoint blockade in a new preclinical ovarian cancer model. [abstract]. In: Proceedings of the AACR Conference: Addressing Critical Questions in Ovarian Cancer Research and Treatment; Oct 1-4, 2017; Pittsburgh, PA. Philadelphia (PA): AACR; Clin Cancer Res 2018;24(15_Suppl):Abstract nr B33.