Abstract 1571: Developing antigen-specific Th1 epitope based vaccines by excluding immunosuppressive peptide sequences

Abstract

Abstract We have determined that IGFBP-2 is immunogenic in ovarian cancer. Our aim is to develop a multi-epitope vaccine that will elicit Th1 immunity to IGFBP-2. Antigen specific Th1 can modulate the tumor microenvironment to enhance cross priming, supporting the proliferation of cytotoxic T cells which are capable of eradicating ovarian cancer cells. Although some clinical benefit has been demonstrated with this strategy, recent ex vivo analyses of human PBMC have described the presence of antigen-specific CD4+ T regulatory cell (Tregs) in cancer patients that were not detected in healthy individuals. Thus, we questioned if we could optimize our vaccine to include only Th1-stimulating epitopes. Using a combined scoring system from five algorithms for predicting class II binding to determine Th epitopes, we identified 14 IGFBP-2 peptides. Th1 immunogenicity (IFN-gamma) and potential immunosuppression (IL-10) was evaluated by ELISPOT for 40 different donors. Twenty-two percent of the donors only responded with IL-10 secretion to any peptide, 22% only responded with IFN-gamma and 53% of the patients had a mixed IFN-gamma and IL-10 response. To determine which peptides would induce a predominantly Th1 response in the greatest number of people, we used a ratio of IFN-gamma to IL-10 and analyzed both the magnitude and frequency of ELISPOT responses for each of peptides using the following algorithm: (corrected mean SPW) x (percent of responding donors). The peptides were then ranked from highest IL-10 response to highest IFN-gamma response. Interestingly, 6 of the 14 peptides demonstrated a preference to secrete IFN-gamma over IL-10 and are only located in the N-terminus (amino acids 1-163) of IGFBP-2; the remaining potentially immunosuppressive peptides are located in the C-terminus (amino acids 164-328). Vaccination with p1-163 in MMTV-neu mice demonstrated a robust Th1 response (p=0.03) and concomitant inhibition of tumor growth by 70% compared to adjuvant only control animals, p164-328-vaccinated or full length protein-vaccinated mice (p<0.001 for all). Vaccination with p164-328 or full length protein did not inhibit tumor growth. These data suggest that more effective vaccines can be designed when both Th1 epitopes and immunosuppressive epitopes are screened simultaneously and epitopes that are most likely to induce robust Th1 responses in the majority of individuals can be identified and included as vaccine components. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 1571. doi:1538-7445.AM2012-1571