Thousands of patients have benefited from the growing use of cancer immunotherapies. However, the success of these therapies can be highly variable, with some patients experiencing complete tumor remission, while others suffer from progressive disease. Most studies trying to understand this variability have focused on differences in the genetic makeup of the tumors, the status of the patient’s immune system, and/or environmental factors. Here, we explore a more fundamental contributing factor, namely that the heterogeneity of genetically identical cells could account for the stochasticity in the outcome of cancer immunotherapies.
We introduce a simple ex vivo model of tumor/T cell interactions modeling the large Poissonian variations in the immune response. We use a stochasticity-based framework (“cell filleting”) to identify a rare population of naïve T cells (“Spark T cells”) that is necessary and sufficient to spark anti-tumor immune reactions: stochastic fluctuations in the number of spark T cells account for the variability of these reactions. Our framework combines statistical modeling, high-throughput spectral flow cytometry, and machine learning, to further define a gating strategy to identify the Spark T cell progenitor population.
We are currently performing experiments to test the efficacy and the functional significance of the identified immune population in vivo. We are also applying this framework to identify immune populations in human-derived TCR-engineered T cell blasts that have higher efficacy in initiating the response against tumor cells. We envision this framework being applied to identify other relevant immune cell types that act as catalysts for successful cancer immunotherapies.