Dendritic cell (DC)-based vaccines have been largely used in the adjuvant setting for the treatment of cancer, however, despite their proven safety, clinical outcomes still remain modest. In order to improve their efficacy, DC-based vaccines are often combined with one or multiple immunomodulatory agents. However, the selection of the most promising combinations is hampered by the plethora of agents available and the unknown interplay between these different agents. To address this point, we developed a hybrid experimental and computational platform to predict the effects and immunogenicity of dual combinations of stimuli once combined with DC vaccination, based on the experimental data of a variety of assays to monitor different aspects of the immune response after a single stimulus. To assess the stimuli behavior when used as single agents, we first developed an in vitro co-culture system of T cell priming using monocyte-derived DCs loaded with whole tumor lysate to prime autologous peripheral blood mononuclear cells in the presence of the chosen stimuli, as single adjuvants, and characterized the elicited response assessing 18 different phenotypic and functional traits important for an efficient anti-cancer response. We then developed and applied a prediction algorithm, generating a ranking for all possible dual combinations of the different single stimuli considered here. The ranking generated by the prediction tool was then validated with experimental data showing a strong correlation with the predicted scores, confirming that the top ranked conditions globally significantly outperformed the worst conditions. Thus, the method developed here constitutes an innovative tool for the selection of the best immunomodulatory agents to implement in future DC-based vaccines.
Keywords: Algorithm; Cancer; Dendritic cells; Hybrid platform; Immunotherapy; Prediction; Vaccines.