Largest Metabolomic Study of CAR T-Cell Therapy Reveals New Biomarkers for Predicting Severe Neurotoxicity
Key Insights
Researchers from Kite (search), a Gilead company, conducted the most extensive metabolomic study ever in CAR T-cell therapy, analyzing over 3,800 longitudinal samples from patients treated with FDA-approved anti-CD19 (search) therapies.
The study identified clear metabolic signatures that distinguish patients who develop severe neurological events (search), including elevated tryptophan breakdown and increased arginine pathway activity.
Metabolomics-derived pathway scores outperformed traditional inflammatory markers like IL-6 and TNFα (search) in predicting neurotoxicity (search) risk, offering potential for more accurate early warning tools.
Researchers from Kite (search), a Gilead company, have completed the largest metabolomic analysis ever conducted in CAR T-cell therapy, revealing new insights into the biological mechanisms driving severe neurotoxicity (search). The comprehensive study analyzed more than 3,800 longitudinal serum and plasma samples, along with rare cerebrospinal fluid (CSF) samples, from patients treated with FDA-approved anti-CD19 (search) CAR T-cell therapies axicabtagene ciloleucel (axi-cel) and brexucabtagene autoleucel (brexu-cel).
The multi-trial meta-cohort, spanning six clinical studies, enabled investigators to uncover metabolic pathways strongly associated with severe neurological events (search) (NEs), a significant and sometimes life-threatening toxicity of CAR T-cell therapy. Despite the transformative efficacy of CAR T-cell treatments for large B-cell lymphoma (search) (LBCL), mantle cell lymphoma (search), follicular lymphoma (search), and B-cell acute lymphoblastic leukemia (search), the biological drivers of neurotoxicity (search) have remained poorly understood.
Metabolic Signatures Distinguish High-Risk Patients
Using Metabolon (search)'s high-resolution metabolomics platform, researchers identified clear and reproducible metabolic signatures that distinguished patients who developed high-grade NEs (grade ≥ 3). The study leveraged Metabolon's untargeted Global Discovery Panel to provide unprecedented resolution into metabolic dysfunction associated with CAR T-cell therapy.
"This study demonstrates how metabolomics uniquely exposes the biological pathways driving CAR T-cell–associated neurotoxicity (search), insights that are not accessible through proteomics or cytokine profiling alone," said Heino Heyman, Director of Global Field Metabolomics Sciences at Metabolon (search). "By mapping disruptions in tryptophan catabolism, NMDA-linked excitotoxicity, and polyamine metabolism, the analysis not only identified patients at risk for severe neurologic events but also highlighted actionable pathway targets to mitigate toxicity."
Key Metabolic Pathways Linked to Neurotoxicity
The analysis revealed several critical metabolic disruptions associated with severe neurological events (search):
Elevated Tryptophan Breakdown: Patients who developed severe neurological events (search) consistently showed higher breakdown of tryptophan, leading to increased levels of metabolites such as quinolinate. These markers were present both before and after treatment and are strongly associated with heightened neurotoxicity (search) risk.
Increased Arginine Pathway Activity: Severe cases showed a shift in how the body processes arginine, resulting in the production of more urea and acetylated polyamines, such as N1, N12-diacetylspermine. This pattern reflects increased immune system activation and may serve as another indicator of neurotoxicity (search) risk.
Central Nervous System Confirmation: Cerebrospinal fluid samples taken during neurotoxic events showed the same metabolic disruptions observed in blood—higher levels of glutamate and other stress-related metabolites—confirming that these changes directly involve the central nervous system.
Superior Predictive Performance Over Traditional Markers
The study demonstrated that metabolite-based scores built from these biomarkers outperformed standard inflammatory markers such as IL-6 and TNFα (search) in identifying patients at risk for severe neurological events (search). This finding demonstrates the potential for more accurate, metabolomics-driven early warning tools in CAR T-cell therapy monitoring.
Machine-learning models reinforced the importance of the tryptophan-kynurenine pathway for neurotoxicity (search) associated with CAR T-cell therapy. The same metabolites linked to neurotoxicity—such as quinolinate and acetylated polyamines—were also associated with worse disease outcomes, suggesting these biomarkers may have broader prognostic value.
Clinical Implications and Future Applications
"These findings underscore the power of metabolomics to reveal mechanisms that are invisible to genomic, proteomic, and cellular assays alone," said Ro Hastie, CEO of Metabolon (search). "By enabling unprecedented resolution into metabolic dysfunction associated with CAR T-cell therapy, Metabolon's platform has helped identify new biomarkers and potential therapeutic targets to mitigate severe neurotoxicity (search)."
The research provides a functional, pathway-level understanding of safety and patient response, which is crucial for enhancing cell therapy design, monitoring strategies, and clinical outcomes. The identification of actionable pathway targets offers new opportunities for developing interventions to mitigate CAR T-cell therapy-associated neurotoxicity (search) while preserving therapeutic efficacy.