A theory of multi-site vaccination for cancer

What distinguishes cancer from other more treatable diseases is perhaps the random, multi-strain nature of the disease. Here we apply tools from statistical mechanics to model cancer vaccine design. The difficulty of controlling cancer by many of the standard therapies has led to substantial interest in control by the immune system. Escape of cancer from the immune system can be viewed as a percolation transition, with the immune system killing of cancer cells the parameter controlling whether the cancers cells proliferate. The model we develop suggests that vaccination with the different strains of cancer in different physical regions leads to an improved immune response against each strain. Our approach captures the recognition characteristics between the T cell receptors and tumor, the primary dynamics due to T cell resource competition, and the secondary dynamics due to competition between escape of tumor cells by epitope mutation and allele loss and elimination of tumor cells by T cells.