The case centers on a patient who, after receiving CAR T cell therapy, developed signs of profound neurological injury. Brain magnetic resonance imaging revealed abnormalities characteristic of immune effector cell-associated neurotoxicity syndrome, universally abbreviated as ICANS: fluid-attenuated inversion recovery sequences showed periventricular white matter hyperintensities involving the atrium and temporal horn of the lateral ventricles, while diffusion-weighted imaging demonstrated areas of cytotoxic edema within the hippocampi on both sides. But what made the case exceptional was the spinal imaging. T2-weighted magnetic resonance images of the spine disclosed a longitudinally extensive spinal cord lesion, with abnormal signal intensity extending over more than three vertebral body segments. Critically, the lesion showed no enhancement after gadolinium administration on post-contrast T1-weighted images.

That combination of findings, a long spinal cord lesion that refuses to light up with contrast, is diagnostically treacherous. In clinical practice, a longitudinally extensive T2 hyperintense lesion in the spinal cord is nonspecific and can be produced by infectious, autoimmune, or vascular processes. Acute autoimmune inflammation of the spinal cord, or ischemia-related breakdown of the blood-spinal cord barrier, typically produces contrast enhancement, because the damaged barrier allows the gadolinium to leak into the cord tissue. A non-enhancing lesion therefore sits uneasily with the classic picture of acute inflammatory myelitis. The Paris team reasoned that, given the concurrent brain lesions suggestive of ICANS and the exclusion of infectious causes, the most likely diagnosis was spinal cord involvement as a component of ICANS itself, a manifestation so rare that only a handful of cases have been described in the literature, and those published reports describe similar non-enhancing imaging features.

The first and most urgent diagnostic step the authors emphasize is the systematic exclusion of infectious myelitis. This is not a bureaucratic precaution. Patients who have undergone CAR T cell therapy are profoundly immunocompromised, both from the lymphodepleting chemotherapy that precedes the infusion and from the therapy’s effects on normal B cells and other immune populations. In such patients, opportunistic infections of the central nervous system are a constant threat. The authors single out viral etiologies that must always be considered: cytomegalovirus, herpes simplex virus, varicella-zoster virus, enteroviruses, and human herpesvirus-6, each of which can produce devastating spinal cord disease in the immunosuppressed host. In their patient, cerebrospinal fluid cultures and multiplex polymerase chain reaction testing were negative, making an infectious etiology unlikely and clearing the way for the diagnosis of ICANS-associated myelopathy.

Understanding why this matters requires appreciating what ICANS is and how it usually behaves. Immune effector cell-associated neurotoxicity syndrome is the most feared non-hematological toxicity of CAR T cell therapy, typically emerging days after infusion as the engineered cells expand and release inflammatory cytokines. It usually announces itself with aphasia, confusion, tremor, and, in severe cases, seizures, cerebral edema, and coma. Aseptic meningitis is a common feature, and brain imaging may show transient abnormalities. The syndrome is graded from I to IV, with grade IV representing life-threatening neurotoxicity. Standard management escalates from corticosteroids to the interleukin-1 receptor antagonist anakinra for severe disease. But the spinal cord has been largely absent from this picture, which is precisely why the Paris case is so valuable to the field.

The timing of spinal cord involvement poses its own diagnostic puzzle, and the authors are candid about this uncertainty. In patients with severe ICANS, concomitant severe encephalopathy frequently precludes a complete neurological examination. A patient who is obtunded or comatose cannot report numbness, weakness, or a sensory level, and the examiner cannot reliably test for the signs that would localize a spinal cord lesion. As a result, it is often impossible to determine when spinal cord symptoms first occur, or whether they emerged before, alongside, or after the encephalopathy. In the reported patient, cerebral and spinal cord involvement probably developed simultaneously, a synchrony that itself supports the idea that both represent facets of a single systemic neurotoxic process rather than two independent complications.

Therapeutically, the patient received the recommended treatment for grade IV ICANS: high-dose corticosteroids combined with anakinra, the interleukin-1 receptor antagonist that has become the mainstay of salvage therapy for steroid-refractory neurotoxicity. Beyond this standard regimen, however, the evidence base evaporates. The authors state plainly that there is no clear evidence supporting a specific therapeutic strategy for ICANS-associated myelopathy. Interventions that neurologists might reach for in other forms of acute myelitis, such as intravenous immunoglobulins or plasmapheresis, have no established role here, because the underlying biology of the spinal cord lesion remains obscure. Larger case series, the authors argue, are needed to define the optimal treatment and the prognosis of this rare manifestation.

The imaging findings deserve close attention from clinicians who will face similar dilemmas. The pattern reported here, a long T2-hyperintense cord lesion without contrast enhancement, occurring in a patient with hippocampal cytotoxic edema and periventricular white matter changes, effectively constitutes a signature that others can look for. The hippocampal involvement is particularly suggestive, because bilateral cytotoxic edema in the hippocampi has been described in severe CAR T cell neurotoxicity and reflects the vulnerability of this region to the inflammatory and possibly vascular disturbances that ICANS produces. When such brain findings are accompanied by a non-enhancing longitudinal cord lesion in a recently infused patient, ICANS-associated myelopathy moves to the top of the differential diagnosis, provided infection has been rigorously excluded.

The broader significance of the case lies in the growing recognition that CAR T cell neurotoxicity is not confined to the brain. As the therapy expands from refractory lymphomas and multiple myeloma into earlier lines of treatment and into autoimmune diseases, the population of exposed patients is growing rapidly, and rare complications become clinically meaningful at scale. Reports such as this one, published as part of the journal’s imaging series with practical tips for intensivists, serve as sentinels. They alert critical care physicians, neurologists, and hematologists that when a CAR T cell recipient deteriorates neurologically, the evaluation must extend beyond the cranial vault. Spinal imaging, cerebrospinal fluid analysis with multiplex pathogen testing, and careful correlation of brain and cord findings should be part of the diagnostic armamentarium.

For now, the message from the Paris team is one of disciplined caution. Infectious myelitis must be systematically ruled out before attributing a cord lesion to ICANS, because the treatments differ fundamentally and the stakes are enormous. The non-enhancing character of the lesion, while atypical for classic autoimmune myelitis, appears consistent across the few published cases of ICANS-associated myelopathy and may reflect a distinct pathophysiology, perhaps one that does not disrupt the blood-spinal cord barrier in the way conventional inflammation does. Until larger series clarify the mechanism, the optimal treatment, and the long-term prognosis of this unusual complication, the case stands as both a warning and a guide: the engineered immune cells that save lives can, in rare instances, reach beyond the brain, and clinicians must be ready to recognize when they do.

Subject of Research: Acute myelopathy as a rare manifestation of CAR T cell therapy-associated neurotoxicity

Article Title: Acute myelopathy following chimeric antigen receptor (CAR) T cell therapy: an unusual diagnostic challenge

Article References: Bois, A., Provost, C., Decroocq, J., & Benghanem, S. (2026). Acute myelopathy following chimeric antigen receptor (CAR) T cell therapy: an unusual diagnostic challenge. Intensive Care Medicine. https://doi.org/10.1007/s00134-026-08617-w

Image Credits: AI Generated

DOI: 10.1007/s00134-026-08617-w

Keywords: CAR T cell therapy, ICANS, neurotoxicity, myelopathy, spinal cord, immune effector cell, MRI, encephalopathy, anakinra, corticosteroids, immunocompromised, intensive care

X-linked adrenoleukodystrophy is caused by pathogenic variants in the ABCD1 gene, which cripple the cellular machinery that breaks down very long-chain fatty acids. These fatty acids accumulate throughout the body, and in the nervous system they drive progressive destruction of myelin and axons. Roughly 60 percent of male patients develop a leukodystrophy, an inflammatory destruction of brain white matter, while virtually all men and most women eventually develop a slowly progressive disease of the spinal cord that targets the dorsal columns and corticospinal tracts. This spinal cord form, often called myeloneuropathy, erodes the ability to walk over decades, yet its pace varies enormously between individuals and is notoriously difficult to quantify with existing clinical scales.

That variability is precisely the problem the Amsterdam team set out to solve. Traditional outcome measures such as the Expanded Disability Status Scale, or EDSS, and the six-minute walk test change slowly, meaning clinical trials must follow patients for years to detect meaningful differences. Large patient numbers are needed because progression rates differ so widely. A biomarker that could sort patients at baseline into fast and slow progressors would allow smaller, faster, and cheaper trials, and would give clinicians and families a rational basis for decisions about when to intervene.

The study drew on the ongoing Dutch X-ALD cohort, using data collected between 2015 and 2024 at Amsterdam University Medical Centers. The researchers selected 66 adult men with spinal cord disease, excluding those with active leukodystrophy, prior hematopoietic stem cell transplantation, gene therapy, or compassionate use of the drug leriglitazone, because these interventions independently alter disease trajectories. Blood samples were processed within two hours of collection, centrifuged, aliquoted, and frozen at minus 80 degrees Celsius in a biobank. Plasma concentrations of neurofilament light chain and glial fibrillary acidic protein, a marker of astrocyte injury, were measured using single-molecule array technology, an ultra-sensitive immunoassay capable of detecting proteins at femtomolar concentrations. Assays were run by certified technicians blinded to all clinical data, with intra-assay coefficients of variation of just 4.8 percent for NfL and 3.3 percent for GFAP.

The stratification strategy was deliberately simple. Because NfL levels rise with age even in healthy people, and because no disease-specific reference values exist for adrenoleukodystrophy, the team divided patients using cohort-based thresholds. Men whose baseline plasma NfL was at or above 15.7 picograms per milliliter, the upper quartile, were classified as high; the rest as low. The differences at baseline were striking. The high-NfL group was significantly older, averaging 58.4 years against 36.4 years, and scored far worse on every clinical measure: higher EDSS disability scores, lower scores on the Severity Scoring System for Myelopathy, and dramatically shorter six-minute walk distances, averaging 333 meters compared with 529 meters in the low group.

The longitudinal results were even more consequential. Over up to seven years of follow-up, the high-NfL group showed significant, measurable worsening on the EDSS as early as the first year of follow-up, with an estimated mean change of 0.24 points, and the deterioration continued relentlessly, accumulating to 1.70 points by year seven. The low-NfL group, by contrast, showed minimal and statistically insignificant change over the same period. The difference in progression slopes between the two groups was highly significant, with a beta coefficient of minus 0.27 and a p-value below 0.001. Walking capacity declined significantly only in the high-NfL group, while patient-reported myelopathy scores fell in both groups but did not separate statistically between them.

Because age is such a strong confounder, the researchers performed a sophisticated sensitivity analysis. They log-transformed baseline NfL values, regressed them on age within the cohort, and calculated standardized residuals that capture NfL elevations beyond what would be expected for a given age within the X-ALD population itself. Patients were then grouped into four categories spanning minus one to plus two standard deviations. The gradient was unmistakable: men in the very high residual group showed EDSS worsening of 0.137 points per visit and SSPROM decline of 1.276 points per visit, both highly significant, while the low residual group showed no significant progression on any outcome measure. Pairwise comparisons confirmed that the mean, high, and very high groups differed significantly from one another in their EDSS and SSPROM slopes even after correction for multiple testing. Crucially, the association between higher age-adjusted NfL and faster clinical progression persisted, supporting the interpretation that the protein reflects disease-related axonal injury rather than aging alone.

GFAP told a very different story. Although plasma GFAP levels rose significantly over the study period overall, baseline GFAP stratification showed no significant differences in EDSS, SSPROM, or walking test scores, and no significant differences in progression slopes between high and low groups. The authors conclude that GFAP, which marks astroglial activation rather than axonal loss, has limited prognostic value in this context and may serve better as a complementary marker than a standalone predictor. The contrast is biologically informative: adrenoleukodystrophy-related myelopathy is fundamentally a degeneration of axons and their myelin sheaths, so a cytoskeletal protein shed from damaged axons tracks the disease process more faithfully than a marker of astrocytic response.

The findings dovetail with a growing literature on NfL in other forms of the disease. Previous work by the same group had shown correlations between NfL concentrations and spinal cord disease severity, and pediatric studies have demonstrated that plasma NfL is a sensitive monitor of cerebral lesion development in childhood cerebral adrenoleukodystrophy, with increases of more than 50 percent from baseline serving as a robust indicator of active inflammation. Other researchers have reported NfL elevations in myeloneuropathy patients who later developed the cerebral form, reaching levels comparable to those seen in amyotrophic lateral sclerosis during active disease. The new results extend this framework to the slow, chronic degeneration of the adult spinal cord, showing that the same molecule captures both explosive inflammatory damage and grinding degenerative loss.

The authors acknowledge limitations, including the modest sample size, the use of cohort-derived rather than population-based thresholds, and the insensitivity of clinical scales to early change in a disease where axonal degeneration precedes measurable disability until a critical threshold is crossed. Even so, the effect sizes were substantial and the trajectories separated early. Given that this represents the longest biochemical prospective follow-up in adult men with X-ALD-related myeloneuropathy, the researchers argue that serum NfL should be integrated into clinical trial protocols both as a stratification tool, to balance fast and slow progressors across treatment arms, and as a candidate surrogate outcome measure. For patients living with an unpredictable disease, a number drawn from a routine blood draw may soon transform both research and care.

Subject of Research: Prognostic biomarkers for spinal cord disease progression in X-linked adrenoleukodystrophy

Article Title: Prognostic Value of Neurofilament Light Chain and Glial Fibrillary Acidic Protein in ALD‐Related Myelopathy

Article References: Kabak, E. G., Voermans, M. M. C., Heijst, H., Teunissen, C. E., & Engelen, M. (2026). Prognostic Value of Neurofilament Light Chain and Glial Fibrillary Acidic Protein in ALD‐Related Myelopathy. Annals of Clinical and Translational Neurology, 13(10), 2114-2123. https://doi.org/10.1002/acn3.70386

Image Credits: AI Generated

DOI: 10.1002/acn3.70386

Keywords: X-linked adrenoleukodystrophy, neurofilament light chain, GFAP, biomarkers, myelopathy, spinal cord disease, ABCD1, clinical trials, neurodegeneration, Simoa, prognosis, longitudinal study