Abstract

Cancer immunotherapies such as immune checkpoint inhibitors (CI) are approved for a broad range of solid tumors and selected hematologic malignancies. Clinical response to CIs has been shown to be associated with mutation burden and is predicated on the presence of T cells specific to tumor-associated antigens. Therefore, increasing the frequency of such T cell responses should increase the efficacy of anti-tumor immunity. Individualized neoantigen peptide vaccines are a promising strategy to activate tumor-specific T cell immunity. Next generation sequencing of tumor specimens is now widely available and novel computational algorithms provide the means for the prediction and selection of neoantigens.

We are conducting a preclinical study of a personalized vaccine in an animal model of multiple myeloma (MM). MM is an incurable cancer of plasma cells that resides in the bone marrow (BM) and causes lytic bone disease and other end-organ damage. MM is characterized by an intermediate level of somatic mutation that may yield mutational neoantigens for vaccine design. This investigation pursued two goals: first, to determine if neoantigens are immunogenic, and second, if they induce protective immunity against tumors. To address the immunogenicity of neoantigens, we used the MOPC.315.BM myeloma cell line in the immunocompetent BALB/c mouse background. MOPC.315.BM cells mimic human disease by homing to the BM and causing osteolytic lesions. We performed whole exome sequencing and RNAseq on MOPC.315.BM cells to identify expressed somatic mutations. Neoantigens were then selected based on predicted MHC class I and II binding and expression levels, and corresponding 27-mer peptides were synthesized and tested in vivo for their immunogenicity. Mice were immunized 3 times at day 0, 14 & 21 with pools of peptides in the presence of poly (I:C) adjuvant, and splenocytes were harvested at day 28 to measure peptide-specific T cell activation by IFNg ELISpot. Of the selected peptides, 25 of 66 induce high IFNg T cell responses in the spleen. Interestingly, 88% of these responses were CD4 T cells and 12% were CD8 T cells, consistent with other preclinical and clinical studies where neoantigens predicted to bind MHC-I molecules induced mainly CD4 T cell activation. Humoral immune response specific to the peptides were analyzed by ELISA and showed that 8 of the 25 peptides induce peptide-specific IgG antibodies in the serum.

To determine whether neoantigens generate tumor protective immunity, we challenged vaccinated mice with the MOPC.315.BM cell line. We used a well-characterized antigen lambda 2315 idiotype (Id) produced by MOPC cells and known to activate CD4 T cells. Vaccines using only the Id were found to be weakly immunogenic and failed to elicit protective responses. We developed a new strategy of vaccination by adding a proprietary amino acid sequence comprising a high affinity HSP70 binding motif to the idiotype peptide. Mice vaccinated at days 0, 14 & 21 with the so-modified peptide and poly-(I:C) adjuvant had 4 times higher Id-specific IFNg CD4 T cell response in the spleen compared to mice vaccinated with the unmodified Id-peptide. When challenged with MOPC.315.BM cells after vaccination with the modified Id peptide, 60% of these mice maintained non-detectable Id level in their blood and resisted tumor growth for more than 70 days while unvaccinated mice had very high levels of circulating Id and died around 30 days after tumor injection. Id-mediated cytotoxicity is being investigated in vitro to complement these results. We are conducting tumor protection experiments with the 25 immunogenic peptides identified in the vaccination experiments described in the previous section.

In summary, we showed that neoantigens derived from bioinformatic analysis of tumor genome sequencing are immunogenic and that they mainly activate CD4 T cells in this setting. Additionally, our vaccination strategy using the idiotype peptide showed that CD4 T cell immunity can be protective against tumor growth. We are currently investigating the potential of the neoantigens identified in the immunogenicity study to protect mice against tumor growth and eradicate pre-existing tumors. Our findings indicate that neoantigen vaccination is very promising and that future investigations combining this strategy with CIs and/or immunomodulatory agents may lead to the efficient generation of protective anti-tumor immunity.

Disclosures

Bekri: Agenus Inc: Research Funding. Cho: Ludwig Institute for Cancer Research: Research Funding; Multiple Myeloma Research Foundation: Research Funding; Genentech: Other: advisory board, Research Funding; Bristol Myers-Squibb: Other: advisory board, Research Funding; Agenus, Inc.: Research Funding.

Author notes

*

Asterisk with author names denotes non-ASH members.