CAR T centers see preprint linking CD163⁺ monocytes to neurotoxicity risk
New research links CD163⁺ monocytes and CD177⁺ neutrophils to severe ICANS after CAR T therapy, offering a potential path for biomarker-driven care.
Edward Mullen ·

The prevailing wisdom in managing CAR T-cell-associated neurotoxicity (ICANS) centers on broad, symptom-driven interventions. Yet, a recent bioRxiv preprint challenges this empirical model, suggesting that specific cellular biomarkers, rather than diffuse clinical signs, hold the key. This work argues that ICANS is not an undifferentiated inflammatory event, but rather a process traceable to distinct myeloid signatures.
What the preprint actually measured and found
The authors combined mass cytometry, serum proteomics, and single-cell transcriptomic profiling to search for immune correlates of Immune effector cell-associated neurotoxicity syndrome (ICANS) in patients treated with CD19 CAR T cells, and report a distinct myeloid inflammatory signature associated with severe neurotoxicity. Specifically, they observe coordinated expansion of CD163⁺ monocytes together with immature CD177⁺ neutrophils in peripheral blood that coincides with severe ICANS episodes.
The paper frames these populations as potential biomarkers of neurotoxic risk, arguing the signal is robust across the assays used.
Why this reading contradicts routine practice
The prevailing clinical approach to CAR T neurotoxicity management is symptom-based — steroids, supportive care, and broad immune suppression are deployed reactively because ICANS has been heterogeneous and unpredictable. The preprint challenges that by presenting specific, measurable cellular markers rather than an undifferentiated cytokine storm fingerprint.
If those markers consistently precede or discriminate severe ICANS, they could enable earlier, targeted interventions or enrollment criteria that select or de-risk patients before infusion. That prospective diagnostic lens departs from the current empirical orientation in care and R&D.
Limits in the evidence: what the paper does not settle This is a preprint and the claim is unvalidated outside the presented cohort; the authors do not report multicenter replication, nor do they present a prospective validation in an independent clinical series. The mechanistic link between expanded CD163⁺ monocytes or CD177⁺ neutrophils and central nervous system injury remains correlative in the dataset presented.
The assays used (mass cytometry, proteomics, single-cell RNA) are powerful for discovery but can be sensitive to batch effects, sampling timing relative to symptom onset, and patient selection. The paper does not define a diagnostic threshold nor quantify sensitivity or specificity against standard clinical endpoints, so its translational readiness for a clinical decision rule is unproven.
How this could reprice R&D margins in CAR T and supportive care If subsequent work validates these myeloid markers as predictive, pharmaceutical and diagnostic companies could reallocate development spend away from broad, empiric neuroprotectants toward companion diagnostics, targeted anti-myeloid agents, or prophylactic regimens stratified by biomarker status. That would shift margin structures: diagnostics and targeted adjuncts command higher per-patient margins than routine steroids and hospital care, and they reorder R&D portfolios toward earlier-stage biomarker work.
The preprint itself does not address commercialization pathways, reimbursement, or regulatory interactions, but the data pattern it reports is exactly the sort of signal that converts a symptomatic management market into one rentable by diagnostics and targeted biologics.
The skeptic case: plausible counter-reads and outstanding questions A reasonable counter is that myeloid expansions reflect downstream inflammation severity rather than causal, actionable predictors; they might simply be epiphenomena present only after severe ICANS has already begun. Another concern is cohort size and heterogeneity — single-center or convenience-sample effects could produce signatures that fail in broader patient mixes.
Finally, integrating a cell-based biomarker into time-sensitive CAR T workflows faces logistic hurdles: flow or cytometry-based assays have turnaround times and standardization challenges that the paper does not solve. In short, the finding is hypothesis-generating but not yet a clinical pivot.
Near-term implications for hospitals, drug makers, and diagnostics vendors Clinical centers that run CD19 CAR T programs will see immediate pressure from this preprint to prioritize prospective sampling and banking of peripheral myeloid compartments during trials and standard care, because prospective validation is the only path to using the markers clinically. Pharmaceutical sponsors and diagnostics vendors should treat the signal as a catalyst to design prospective substudies: small, adjunctive trials that add high-frequency blood sampling and prespecify CD163/CD177-based analyses.
Regulators and payors will be key gatekeepers; the paper omits discussion of regulatory pathways for a companion diagnostic tied to a cell therapy, and that omission is where commercial translation often stalls.
Observable signals to watch in the next 6 months Watch whether major CAR T sponsors or academic consortia register prospective validation cohorts that include standardized assays for CD163⁺ monocytes and CD177⁺ neutrophils, whether diagnostic vendors file pre-submissions or develop rapid assays tied to those markers, and whether conference programs (including ASH 2025) include multicenter validation data; absence of such activity would weaken the translational case. Also monitor whether hospital cellular-therapy programs change sample-collection protocols or informed-consent language to permit biomarker-driven substudies, because operational uptake is a leading indicator of momentum.
The bioRxiv preprint offers a specific, data-forward counterproposal to symptom-only management: that myeloid cell populations can stratify neurotoxicity risk. It is an unvalidated claim at present, but one that, if replicated, would realign R&D spending and commercial margins toward diagnostic-enabled, targeted interventions rather than broadly applied immunosuppression.
The paper does not, however, address test commercialization, regulatory strategy, or prospective diagnostic performance metrics — those gaps will determine whether this discovery remains an academic signal or becomes a new margin for industry.