Adoptive T cell therapies (ACTs) are emerging as a promising strategy to treat cancer. Tumor-infiltrating lymphocytes (TILs) are expanded ex vivo, selected for recognition of neoantigens, further expanded and then infused back into patients. This procedure requires extensive culturing and expansion of TILs during which many T cell clonotypes are lost. As tumor-reactive TILs are often exhausted and tend to be overgrown by functional, non-specific T cells in culture, the chance to identify potent tumor-reactive T cells dramatically decreases. Moreover, extensive expansion of T cells diminishes their anti-tumor activity and persistence in the body after adoptive transfers. Thus, improving the fitness of T cells is crucial to increase the success rate of ACTs and make this therapy accessible to a broad spectrum of cancer patients. Our first aim is to increase the fitness of T cells by designing metabolic and pharmacological interventions based on proteomic profiles of TILs from patients with liver cancer. Second, we will use machine-learning algorithms for the extraction of signatures to predict whether TILs grow well in culture, require and respond to metabolic interventions, or cannot be revitalized and do not grow at all. To deal with non-growing T cells, we aim at establishing a microfluidics-based workflow to graft the entire T cell receptor (TCR) repertoire from thousands of non-growing TILs onto fast-growing Jurkat cells. After selecting Jurkat cells that recognize neoantigens, their TCRs will be expressed on naïve T cells obtained from the patient’s blood that are fit and suitable for ACT. This project will contribute to a better understanding of the T cell response to liver cancer and help to increase the success of personalized ACTs for solid tumors.
