TerraFlow, A New High Parameter Data Analysis Tool, Reveals Systemic T-cell Exhaustion and Dysfunctional Cytokine Production in Classical Hodgkin Lymphoma

The incredible variety of proteins associated with immune responses presents a major challenge in immune monitoring. When combinations of these proteins are measured, cell types that influence disease can be precisely identified. Here, we introduce TerraFlow, a novel data analysis tool that performs an exhaustive search of disease-associated cell populations from high-parameter flow cytometry experiments. Using a newly generated dataset, from 24-color immune checkpoint-focused and 18-color immune function-focused experiments, we apply TerraFlow to classical Hodgkin lymphoma (cHL), where systemic T-cell immunity has not been investigated in detail. We reveal novel immune perturbations in newly diagnosed cHL, as well as persistent immune perturbations after treatment. Newly diagnosed patients have elevated levels of activated (CD278+), exhausted (e.g., CD366+ and CD152+ phenotypes), and IL17-expressing cells, along with diminished levels of naive and central memory (CD127+) T-cells and fewer IFN{gamma}+ and TNF+ T-cells. Exhaustion signatures are reduced with treatment, but compared to healthy individuals, treated patients still exhibit more activated (CD278+ phenotypes), exhausted (CD366+), and IL17-expressing cells. Notably, TerraFlow identifies more phenotypic differences between patient groups than FlowSOM and CellCNN, often with better predictive power. Finally, we introduce a new non-gating approach for data analysis that obviates the need for time-consuming and subjective setting of fluorescence thresholds. Our results benchmark TerraFlow against common methods, provide mechanistic support for past reports of immune deficiency in cHL, and provide a roadmap for future immunotherapy and biomarker studies.