Impact of extrinsic heme on adoptive T cell therapy Free

Chimeric Antigen Receptor (CAR) T cell fitness can be significantly influenced by the metabolic microenvironment. As CAR T cells are often manufactured from autologous patient T cells, circulating metabolites present in the patient's blood may impact the quality of the starting material. One such metabolite is heme (iron-protoporphyrin IX), a vital cofactor involved in numerous cellular processes, including mitochondrial respiration. Heme can be obtained from dietary sources, the extracellular environment, or through endogenous biosynthesis. Microenvironmental heme can also be elevated due to hemolysis. Given its potential abundance in apheresed T cell products, we sought to investigate how exposure to free heme impacts CAR T cell phenotype and function in the context of leukemia. We hypothesized that CAR T cells manufactured in a heme-rich environment may exhibit improved persistence, as heme has been shown to promote oxidative phosphorylation.

We previously found that supplementation of hemin (the oxidized form of heme, hereafter referred to as heme) during healthy donor T cell activation results in a significant, dose-dependent reduction in markers associated with T cell exhaustion, particularly PD-1, LAG3, and TIM-3. Surprisingly, analysis of heme-induced T cell differentiation status revealed a decrease in naïve and central memory subsets, accompanied by an increase in T effector memory cells. These phenotypic changes persisted when T cells were engineered into CAR T cells under heme-rich conditions.

To determine translational relevance, we assayed chronic lymphocytic leukemia (CLL) patient-derived anti-CD19 CAR T cells, as a trademark for this malignancy is widespread T cell dysfunction. CAR T cells manufactured in the presence of heme similarly displayed reduced expression of exhaustion markers and a skewing toward effector memory phenotypes. Given this phenotype, we sought to assess the antitumor activity of heme-manufactured CAR T cells under conditions of continuous antigen stimulation. To evaluate in vivo functionality, we employed a xenograft model in which NOG mice were inoculated with OSU-CLL cells and subsequently treated with either control or hemin-manufactured CAR T cells. Notably, mice receiving control anti-CD19 CAR T cells had significant improvement in survival (P=0.0087), while hemin-manufactured CAR T cells displayed no survival benefit compared to untreated mice (p=0.1320). This finding highlights the critical role of less-differentiated CAR T cell subsets, rather than the reduced expression of exhaustion markers, in driving durable antitumor responses.

Next, in vivo CAR T cell persistence and expansion were analyzed. OSU-CLL-tumor-bearing mice were treated with control or hemin-manufactured CAR T cells, and the CAR T cells were harvested 48 hours later. Flow cytometric analysis revealed a reduction in the absolute number of human CAR T cells in mice treated with heme-supplemented products as compared to control CAR T cells, suggesting impaired in vivo expansion and persistence.

Collectively, these data demonstrate that while the presence of heme during T cell activation and CAR T cell manufacturing reduces the expression of exhaustion markers, it also promotes T cell differentiation and compromises the maintenance of naïve and central memory T cells. This shift in phenotype is associated with diminished antitumor activity and reduced CAR T cell persistence in vivo. These findings have potential clinical implications, particularly for T cell apheresis in patients with elevated systemic heme levels, where the starting T cell fitness is predictive of efficacy. Additionally, these findings also have implications for in vivo CAR T cell manufacturing, where the extracellular metabolic landscape may significantly influence therapeutic efficacy. Moving forward, we aim to identify therapeutic strategies to overcome or inhibit these pathways to enhance CAR T cell function in heme-rich environments.