Neoepitopes-based vaccines: challenges and perspectives.

First generations of cancer vaccines using shared tumour antigens have been associated with disappointing clinical results. However, the paradigm shift introduced by immune checkpoint inhibitors has led to a renewed interest on anti-tumoural vaccination based on mutation-associated neoantigens. First clinical results are encouraging with some signs of clinical activity associated with induction of a specific immune response. In advanced or metastatic diseases, vaccination may either enhance the response to Programmed cell death 1 (PD-1/-L1) antagonists by increasing the number of effectors within the tumour or induce an anti-tumoural T-cell response in immunologically 'cold' tumours. There is also a strong rationale to use cancer vaccines in an adjuvant setting to induce a long-term control of the residual disease. Prediction of neoepitopes efficiently presented by Human Leukocyte Antigen (HLA) molecules remains a challenge, as well as identification of clonal neoantigens. Some mechanisms of resistance are already identified, such as tumour loss of neoepitopes-presenting HLA class I molecules. In this context, the role of CD4+ T cells induced by different cancer vaccines should be clarified. Finally, although studies have focused on mutated epitopes corresponding to single nucleotide variants, other neoantigens could be of strong interest such as those linked to tumour specific RNA-splicing abnormalities or associated with insertions-deletions.