Quadriparesis and paraparesis following chimeric antigen receptor T-cell therapy in children and adolescents

Abstract

Immune effector cell-associated neurotoxicity syndrome (ICANS) is a common but potentially severe adverse event associated with chimeric antigen receptor T-cell (CART) therapy, characterized by the development of acute neurologic symptoms following CART infusion. ICANS encompasses a wide clinical spectrum typified by mild to severe encephalopathy, seizures, and/or cerebral edema. As more patients have been treated with CART, new ICANS phenomenology has emerged. We present the clinical course of 5 children who developed acute onset of quadriparesis or paraparesis associated with abnormal brain and/or spine neuroimaging after infusion of CD19- or CD22-directed CART, adverse events not previously reported in children. Orthogonal data from autopsy studies, cerebrospinal fluid (CSF) flow cytometry, and CSF proteomics/cytokine profiling demonstrated chronic white matter destruction, but a notable lack of inflammatory pathologic changes and cell populations. Instead, children with quadriparesis or paraparesis post-CART therapy had lower levels of proinflammatory cytokines, such as interferon gamma, CCL17, CCL23, and CXCL10, than those who did not develop quadriparesis or paraparesis. Taken together, these findings imply a noninflammatory source of this newly described ICANS phenomenon in children. The pathophysiology of some neurologic symptoms following CART may therefore have a more complex etiology than exclusive T-cell activation and excessive cytokine production.

Graphical abstract

Figure 1.

Visual vignettes describing development of paresis in 4 patients with spinal cord abnormalities on neuroimaging. Patient 1: (A) Time line of clinical course. (B) Axial FLAIR images from MRI of the brain showing bilateral, symmetric white matter changes (thick white arrows) with restricted diffusion (diffusion-weighted images not shown). (C) Axial (left) and sagittal (right) T2 images from MRI of the spine showing increased T2 signal in dorsal columns through entire cord (thin red arrows), without enhancement (enhanced images not shown). Patient 2: (D) Time line of clinical course. (E) Axial FLAIR images from MRI of the brain showing changes in the subcortical, deep, and periventricular cerebral white matter (thick white arrows). (F) Axial (left) and sagittal (right) T2 images from MRI of the spine showing increased T2 signal in dorsal columns through entire cord (thin red arrows), without enhancement (enhanced images not shown). Patient 3: (G) Time line of clinical course. (H) Axial FLAIR images from MRI of the brain showing symmetric foci of T2 hyperintensity involving supratentorial white matter (thick white arrow) and brainstem. (I) Axial (left) and sagittal (right) T2 images from MRI of the spine showing diffuse hyperintensity throughout the entire cord, from the brainstem to the conus (thin red arrows), without enhancement (gadolinium-enhanced images not shown). (J) Axial FLAIR images from MRI of the brain showing interval reduction in T2 signal abnormalities from prior study (thin white arrow). (K) Axial (left) and sagittal (right) T2 images from MRI of the spine showing interval progression of mildly expansile T2 hyperintense signal involving most of the spinal cord. Patient 4: (L) Time line of clinical course. (M) Axial FLAIR images from MRI of the brain showing nonspecific nonenhancing white matter T2/FLAIR hyperintensities (thick white arrows). (N) Sagittal T2 image from MRI of the spine showing extensive, patchy, nonenhancing multifocal T2 lesions, some of which were expansile (thin white arrows; contrast images not shown). D/C, discharge; DTR, deep tendon reflex; IVIG, intravenous immunoglobulin; PICU, pediatric intensive care unit; rehab, rehabilitation; WBC, white blood cell.

Figure 1.

Visual vignettes describing development of paresis in 4 patients with spinal cord abnormalities on neuroimaging. Patient 1: (A) Time line of clinical course. (B) Axial FLAIR images from MRI of the brain showing bilateral, symmetric white matter changes (thick white arrows) with restricted diffusion (diffusion-weighted images not shown). (C) Axial (left) and sagittal (right) T2 images from MRI of the spine showing increased T2 signal in dorsal columns through entire cord (thin red arrows), without enhancement (enhanced images not shown). Patient 2: (D) Time line of clinical course. (E) Axial FLAIR images from MRI of the brain showing changes in the subcortical, deep, and periventricular cerebral white matter (thick white arrows). (F) Axial (left) and sagittal (right) T2 images from MRI of the spine showing increased T2 signal in dorsal columns through entire cord (thin red arrows), without enhancement (enhanced images not shown). Patient 3: (G) Time line of clinical course. (H) Axial FLAIR images from MRI of the brain showing symmetric foci of T2 hyperintensity involving supratentorial white matter (thick white arrow) and brainstem. (I) Axial (left) and sagittal (right) T2 images from MRI of the spine showing diffuse hyperintensity throughout the entire cord, from the brainstem to the conus (thin red arrows), without enhancement (gadolinium-enhanced images not shown). (J) Axial FLAIR images from MRI of the brain showing interval reduction in T2 signal abnormalities from prior study (thin white arrow). (K) Axial (left) and sagittal (right) T2 images from MRI of the spine showing interval progression of mildly expansile T2 hyperintense signal involving most of the spinal cord. Patient 4: (L) Time line of clinical course. (M) Axial FLAIR images from MRI of the brain showing nonspecific nonenhancing white matter T2/FLAIR hyperintensities (thick white arrows). (N) Sagittal T2 image from MRI of the spine showing extensive, patchy, nonenhancing multifocal T2 lesions, some of which were expansile (thin white arrows; contrast images not shown). D/C, discharge; DTR, deep tendon reflex; IVIG, intravenous immunoglobulin; PICU, pediatric intensive care unit; rehab, rehabilitation; WBC, white blood cell.

Figure 2.

Visual vignette describing development of paresis in 1 patient with brain abnormalities only on neuroimaging. Patient 5: (A) Time line of clinical course. (B) Head computed tomography showing mild lateral ventricle prominence (thin white arrow), but no short-term change. (C) Axial FLAIR images from MRI of the brain showing extensive bilateral areas of abnormal T2 signal intensity involving the white matter of the centrum semiovale, superior corona radiata, and periventricular white matter (thick white arrows), associated with restricted diffusion centrally, but no enhancement (diffusion-weighted and gadolinium-enhanced imaging not shown). (D) Sagittal T2 image from MRI of the spine with normal findings. DTR, deep tendon reflex; PICU, pediatric intensive care unit.

Figure 3.

Autopsy findings from the spinal cord (A-F) of patient 1 treated with CART22 and the brain and spinal cord (G-J) of patient 5 treated with CTL019, with both samples notable for white matter damage and an absence of lymphocytic inflammation. (A) Spinal cord showing pallor throughout all the white matter tracts (hematoxylin and eosin [H&E], 1× objective). (B) White matter of the cord showed extensive vacuolation, whereas the gray matter (bottom right) was relatively well preserved (H&E, 10× objective). (C) Numerous axonal spheroids (white arrows) were seen throughout white matter of the cord, indicating axonal injury (neurofilament immunostain, 20× objective). (D) Few macrophages were present within the white matter (CD68 immunostain, 10× objective). (E) Only extremely rare T lymphocytes (black arrow) were present within cord parenchyma (CD3 immunostain, 10× objective). (F) B lymphocytes were absent within the cord (CD79a immunostain, 10× objective). (G) White matter of the spinal cord was well preserved (H&E, 1× objective). (H) Cerebral cortex showing a region of pallor within the white matter (black arrow) (H&E, 1× objective). (I) Cortical white matter with vacuolation and gliosis (H&E, 20× objective). (J) Axonal swellings were seen within the white matter of the cerebral cortex, indicating axonal injury (β-APP immunostain, 10× objective).

Figure 4.

Correlative studies of cellular populations and proteins from CSF demonstrate a marked absence of inflammatory signature. (A) CSF volume and cell counts for 2 patients with spinal paresis after CART at the time of symptoms and 2 patients who received CART but did not develop symptoms of ICANS. Proportion of cell types determined by flow cytometry of each lineage distribution (B) and T-cell distribution (C) were similar between all 4 patients. Proportion of CD4 (D) and CD8 (E) cells with flow cytometry evidence of activation were similar between all 4 patients. (F) Volcano plot (adjusted P > .05, log fold change >0.5) demonstrating differentially expressed proteins measured from D28 CSF in patients who developed quadriparesis or paraparesis (N = 4) and patients treated with CART19 or CART22 who did not (N = 150). The top 2 differentially expressed proteins were interferon gamma (G) and CCL17 (H), and were lower in patients who developed paresis (N = 4) than in those who developed any other phenomenology of ICANS (N = 78). ID, identifier.