AUTHORS
Mark Raymond Dowling, PhD, Hamish W Scott, MD2, Faiza Zuairia Tabassum3, Tammi Tang4, Irene Lee5, Michael Dickinson, MBBS (Hons), DMedSci FRACP, FRCPA6, Simon Harrison, MBBS, FRCPath, MRCP, PhD, FRACP7, Jane Oliaro8
AFFILIATIONS
1 Peter MacCallum Cancer Centre and Royal Melbourne Hospital, Melbourne, Victoria, Australia, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia, Royal Melbourne Hospital, Melbourne, Victoria, Australia, The University of Melbourne, Melbourne, Victoria, Australia
2 Peter MacCallum Cancer Centre and Royal Melbourne Hospital, Melbourne, Victoria, Australia, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia
3 Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia
4 Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia
5 Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia
6 Peter MacCallum Cancer Centre and Royal Melbourne Hospital, Melbourne, Victoria, Australia, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia, The University of Melbourne, Melbourne, Victoria, Australia
7 Peter MacCallum Cancer Centre and Royal Melbourne Hospital, Melbourne, Victoria, Australia, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia, Royal Melbourne Hospital, Melbourne, Victoria, Australia, The University of Melbourne, Melbourne, Victoria, Australia
8 Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia, The University of Melbourne, Melbourne, Victoria, Australia
ABSTRACT
Introduction: Immune effector cell-associated neurotoxicity syndrome (ICANS) is a potentially devastating complication of chimeric antigen receptor (CAR) T-cell therapy, with underlying mechanisms still poorly understood. Progress has been hindered by the lack of representative preclinical models. Although non-human primates likely most closely reproduce ICANS, their use is limited by cost and scalability, while the translational relevance of existing murine models remains uncertain. To address this knowledge gap and identify novel biological pathways involved in ICANS pathogenesis, we used a high-dimensional slow off-rate modified DNA aptamer (SOMAmer) assay to profile serum from patients receiving axicabtagene ciloleucel (axi-cel). The assay quantifies >11,000 analytes, representing approximately half of the known human proteome with broad coverage of biologic processes. Selected analytes identified in the discovery cohort (DC) were validated using the Meso Scale Discovery (MSD) electrochemiluminescence platform.
Methods: Serum was collected pre-infusion (D0), day +3 (D+3) and day +7 (D+7) after axi-cel. In the DC, samples from 34 patients underwent SomaScan proteomic analysis. Findings were validated using the MSD platform in 69 patients, comprising 33 patients from the DC and an independent validation cohort (VC) of 36 patients. Normalised data were analysed in R using Limma to identify differentially abundant proteins between low-grade and high-grade ICANS at each timepoint.
Results: The DC included 22 patients with high-grade (grade 3–4) and 12 with low-grade (grade 0–1) ICANS. The VC was chosen randomly from available samples across the full axi-cel cohort and included 12 high-grade and 24 low-grade cases. The D+3 sample preceded, or in two cases coincided with, peak ICANS severity (median D+5 in the DC and D+6 in the VC).
No proteins were differentially abundant pre-infusion. In contrast, only a small number of proteins distinguished severe from non-severe ICANS after infusion. At D+3, significant analytes (log2 fold change >1, false discovery rate <0.05) were IL-6, IL-8, monocyte chemoattractant protein-1 (MCP-1), macrophage colony-stimulating factor (M-CSF), CX3CL1/fractalkine, IL-1R4 and NT-proBNP. Interpretation of D+7 findings was confounded by treatment received; of note all patients with high IL-1Rα levels had received anakinra, suggesting the aptamer detected the drug itself.
Given the close relationship between cytokine release syndrome (CRS) and ICANS, we performed a secondary analysis comparing high-grade (grade 2–3) and low-grade (grade 0–1) CRS; grade 2 was chosen as the threshold to more evenly balance the groups given the rarity of grade 3 CRS. MCP-1, IL-6 and IL-8 were also differentially abundant in this analysis, together with apolipoprotein-L2 (Apo-L2), and all D+3 analytes peaked before declining by D+7. In contrast, M-CSF and CX3CL1 were not associated with CRS severity, suggesting these pathways may be integral to ICANS pathophysiology independently of CRS.
For validation, we assessed five core analytes identified in the DC (M-CSF, CX3CL1, IL-6, IL-8 and MCP-1), together with six cytokines implicated in ICANS identified from a literature review (GM-CSF, G-CSF, IL-1β, IL-15, IFNγ and TNFα) and IL-7 as a negative comparator using the MSD platform. As D+3 samples likely reflect early disease biology and are less confounded by treatment than D+7, analyses focused on the D+3 timepoint. SomaScan and MSD measurements were moderately to highly correlated. Hierarchical clustering of absolute D+3 values and D0-to-D+3 fold changes using the core analytes aligned closely with clinical classification of severe CRS and/or ICANS. Each core analyte independently discriminated high-grade from low-grade ICANS. In the VC, area under receiver operator curve (AUROC) values were 0.80 for M-CSF and 0.89 for CX3CL1. Consistent with prior literature, all six additional ICANS-associated cytokines were significant on univariable analysis.
Conclusion: We identified an early serum signature of myeloid and endothelial activation associated with severe ICANS, including M-CSF and CX3CL1/fractalkine, both known mediators of neuroinflammation and microglial activation, and potentially amenable to existing drug inhibition. Importantly, this signature is detectable before the onset of severe ICANS, supporting its potential as an early diagnostic biomarker to identify patients at risk and enable earlier therapeutic intervention.
DISCLOSURES
Hamish W Scott, MD: Kite/Gilead - Honoraria