Progressive Multiple Cranial Neuropathies as a Manifestation of CAR-T Neurotoxicity

Abstract

Objective: To describe a case of delayed onset multiple cranial neuropathies as a manifestation of neurotoxicity after chimeric antigen receptor T-cell (CAR-T) therapy for multiple myeloma. While ICANS following CAR-T is a well-reported complication, it classically presents with encephalopathy, seizures, dysphasia, tremors, headache, and cerebral edema. Isolated unilateral facial neuropathies secondary to CAR-T neurotoxicity have been described, but progressive multiple cranial neuropathies have not. Herein, a 75-year-old male presented with left facial nerve palsy 19 days after initiating CAR-T therapy for multiple myeloma. Contrasted brain MRI showed contralateral right facial nerve enhancement, and his left facial palsy was treated with steroids and valacyclovir for 7 days. The facial palsy persisted and progressed to involve bilateral facial nerves and left cranial nerve VI by 31 days post-CAR-T. Specifically, his exam showed impaired abduction of left eye and nearly absent facial movement. Repeat contrasted MRI brain showed mild enhancement of bilateral facial nerves. Extensive serum and CSF testing was unremarkable. Initial treatment with oral steroids for 7 days was ineffective. Concern regarding the impact of steroids on CAR-T efficacy influenced treatment dose and duration. Anakinra was considered but not given. Subsequent treatment with intravenous high dose steroids, followed by a prolonged prednisone taper, led to resolution of CN VI palsy at 2.5 months from onset (2 weeks after completed therapy), and moderate improvement of bilateral facial palsy 5.5 months from onset (3.5 months after completed therapy). CAR-T neurotoxicity can present with progressive multiple cranial neuropathies. The best treatment of these cases is unknown; however, this patient improved in the context of corticosteroids and facial rehabilitation over a prolonged period of follow-up.

Introduction

Immune effector cell-associated neurotoxicity syndrome (ICANS) after chimeric antigen receptor T-cell (CAR-T) therapy refers to a pathologic process involving the nervous system following immunotherapy. ¹ A wide variety of symptoms have been described in the literature with severity previously outlined by the American Society for Transplantation and Cellular Therapy (ASTCT) scale. ² The domains referenced on this scale pay specific attention to the central nervous system and systemic manifestations, including encephalopathy, seizures, dysphasia (agraphia, early mild expressive aphasia), tremors, headache, and cerebral edema. Often, the development of ICANS is preceded by other immune phenomena, specifically cytokine release syndrome (CRS), representing a supraphysiologic response to immunotherapy that leads to fevers, hypotension, hypoxia, and possibly multi-organ failure.^1,3 Mechanisms have been proposed to explain the development of ICANS, but the principal cause remains unknown. CRS often precedes ICANS, supporting theories that endothelial dysregulation and disruption of the blood-brain barrier allows for T-cell transmigration and high concentration of central nervous system cytokines, leading to CNS inflammation and off-target cellular destruction.^4,5 These theories are further supported by elevated proteins, cytokines, and T cells in CSF analysis of ICANS patients and animal models. ⁵ Alternative theories concentrate on the role of IL-1 in CNS monocyte activation and subsequent IL-6 and other inflammatory biomarker production.^2,5

In addition to the aforementioned symptomatology, others have reported the development of cranial neuropathies as a manifestation of non-ICANS neurotoxicity which may develop from similar inflammatory cascades.^3,6,7 The majority of these involve unilateral or rarely simultaneous bilateral facial nerve palsy at onset without involvement of other cranial nerves, especially after B Cell Maturation Agent (BCMA)-targeted therapy.^3,7 These manifestations are rarely life-threatening but greatly impact quality of life for patients. Herein, we expand the spectrum of CAR-T neurotoxicity manifesting as multiple cranial neuropathies by describing a case of progressive facial diplegia and left abducens palsy following exposure to ciltacabtagene autoleucel in a patient with refractory multiple myeloma.

Case Description

A 75-year-old male with stage III kappa light chain multiple myeloma refractory to daratumumab, pomalidomide, and dexamethasone treatment was lymphodepleted (fludarabine and cyclophosphamide) and 5 days later treated with 140 mL infusion of ciltacabtagene autoleucel (BCMA CAR-T therapy); he subsequently developed multiple cranial neuropathies in a stepwise manner. Before this, he had fever, chills, and rigors 8 days post-CAR-T consistent with mild CRS, which resolved with acetaminophen (tocilizumab not administered). At 19 days post-CAR-T, he developed House-Brackmann grade IV left facial nerve palsy (Figure 1). Review of his lab work showed increased inflammatory markers a few days prior (ferritin 130, CRP 58.6). MRI Brain with and without contrast showed facial nerve enhancement of the contralateral right facial nerve (Figure 2). Weighing the potential detrimental effects of corticosteroids on CAR-T efficacy, he was initially treated with valacyclovir alone, but facial function worsened, requiring the addition of prednisone 60 mg daily for 7 days. Despite this treatment, left facial function worsened after corticosteroids were stopped. He then developed left abducens palsy and right facial nerve palsy on day 31 post-CAR-T. Specifically, his exam showed impaired abduction of left eye and twitches of orbicularis oculi without other facial movement evident (House-Brackmann grade VI bilaterally) (Figure 1). The other pertinent findings of his neurologic exam included left upper extremity ataxia, reduced deep tendon reflexes, and a wide-based gait. A repeat MRI Brain with and without contrast showed mild bilateral facial nerve enhancement without abnormalities evident of the abducens nerve or meninges (Figure 2). Extensive serum and CSF testing was unremarkable. CSF testing included: cell count (WBC 2/uL, RBC 8/uL), protein (45 mg/dL), glucose (79 mg/dL), VZV, HSV, EBV, CMV, meningitis-encephalitis panel, cytology and flow cytometry, autoimmune encephalopathy panel, ACE, OCB. Serum testing included: autoimmune encephalopathy panel, Gq1b, acetylcholine receptor, MuSK antibodies, B1, ANCA, RPR, A1c, thiamine, folate, B12, IL-6, and ferritin. Testing for Lyme was considered but not conducted given the absence of travel and exposure that might confer a higher risk for infection.

Figure 1.

Day 38 post-CAR-T: House Brackmann grade 6 (left), grade 5 (right); left CN VI palsy Day 240 post-CAR-T: grade 1 (left), grade 2 (right); left CN VI resolved

Figure 2.

(A): Post contrast axial T1-weighted MRI of the temporal bones demonstrating symmetric enhancement of For Peer Review bilateral facial nerves within the internal auditory canal fundus (white arrows), labyrinthine segment (yellow arrows), anterior genu and geniculate ganglion (red arrows). (B) Post contrast axial T1-weighted MRI of the temporal bones demonstrating symmetric enhancement of bilateral facial nerves involving the tympanic segment (white arrows)

To balance the need to improve facial function and diplopia with risks associated with undertreated multiple myeloma, inpatient treatment included intravenous methylprednisolone 1000 mg for 3 days, 500 mg for 2 days, and 250 mg for 1 day, with a planned taper of oral prednisone 60 mg for 7 days with reductions of 10 mg each day until discontinuing. Once he reached 10 mg daily, the patient increased prednisone to 20 mg daily because of a headache, which then prompted restarting the prednisone taper previously prescribed (60 mg daily for 7 days, followed by daily reductions of 10 mg until off, for a total of 34 days of steroid exposure). He reported significant impact on his quality of life due to difficulty speaking, managing secretions, disrupting vision, and need to tape his eyelids to see as well as maintain closed position to avoid corneal injury at night. The inability to show facial expressions was also socially limiting and particularly concerning to him. He was referred to otolaryngology and ultimately had neuromuscular facial re-training therapy. He had resolution of abducens palsy at 2.5 months from onset (2 weeks after completed steroids) and significant improvement of bilateral facial palsy 5.5 months from onset (3.5 months after completed steroids) with functional recovery to House-Brackmann II on the right and III on the left. At last follow up 7 months from onset, his facial function was much improved, exhibiting House-Brackmann II bilaterally (Figure-1).

Discussion

Immune effector cell-associated neurotoxicity syndrome (ICANS) and non-ICANS neurotoxicity are complex and incompletely understood complications of CAR-T cell therapy, characterized by a wide spectrum of neurological manifestations. This case highlights a unique presentation with multiple cranial neuropathies after ciltacabtagene autoleucel therapy. Given this is an underreported manifestation, thorough workup to exclude alternative etiologies must be carried out. In combination with their temporal relationship and absence of significant findings on extensive infectious, autoimmune, and oncologic evaluations, including unremarkable CSF analysis and negative serologies, these symptoms were attributable to the CAR-T exposure.

The proposed pathophysiology of ICANS/non-ICANS neurotoxicity to some extent incorporates CRS as a precursor to prime the system.^3-5 Our case is more in accordance with the theory of cytokine dysregulation, involving IL-1 and IL-6 leading to T cell transmigration and monocyte activation as evidenced by CRS preceding neurotoxicity. Though not all neurotoxicity presentations are preceded by CRS, the mechanisms are likely tangential, with inflammatory and immune interactions that go beyond the scope of this report. The median time of onset for CRS was 8 days in CARTITUDE-4 and occurred in 76.1% of those patients. ³ They also showed a median onset of 9.5 days in ICANS with occurrence rate of 20.5%. ³ However, for cranial neuropathies specifically, the literature suggests a median onset of 21 days.^3,6 Our patient’s presentation fits within this timeframe, but the progressive nature and multifocal cranial nerve involvement have not been previously described.

The absence of definitive biomarkers or imaging findings specific to neurotoxic cranial neuropathies poses a diagnostic challenge. We considered a broad differential including neoplastic infiltration, lymphoma, basilar meningitis, Gq1b antibody associated disorders (Miller-Fisher syndrome, rhombencephalitis), infections (syphilis, CMV, HSV, VZV), myasthenia gravis, systemic rheumatologic diseases, and thiamine deficiency. Our patient was not altogether different from the CARTITUDE-4 demographic, except for his age (14 years older than median age in that study); importantly no unique characteristic of this patient was identified as greater risk for developing multiple cranial neuropathies. Future studies should focus on elucidating the precise inflammatory pathways involved and exploring targeted therapeutic options. Additionally, the delayed onset of cranial neuropathies, well beyond the typical timeline for ICANS manifestations, highlights the need for extended vigilance for other neurotoxic sequela in patients undergoing CAR-T therapy. ⁸

A recent review identified several categories of non-ICANS neurotoxicity, highlighting that neurocognitive changes, movement disorders, and cranial neuropathies are more commonly associated with BCMA-targeted CAR-T therapies. ⁸ These complications are not adequately captured by the Immune Effector Cell Encephalopathy (ICE) score or ASTCT grading criteria and have the potential to lead to other forms of disabling neurologic impairments, underscoring the need for additional measures to guide treatment escalation. Careful monitoring and systematic reporting of such events can expand our understanding of non-ICANS manifestations and may help elucidate their underlying mechanisms and potential treatments. ⁸

Treatment options are limited, so steroids remain an initial option in suppressing T-cell activation pathways and promoting T-cell apoptosis. ⁸ Alternatively, the IL-1 inhibitor, anakinra, has also been used in higher grade ICANS cases. Theoretical concern for the impact of high dose steroids on CAR-T efficacy influenced our treatment dose and duration, but studies have shown no significant impact of steroid exposure on CAR-T efficacy or long-term outcomes.^9,10 Anakinra was considered but not given since the patient did not exhibit manifestations of ICANS. ¹¹ Cyclophosphamide and etoposide have been suggested for severe neurotoxicity complications. ⁸ Further studies will be needed to guide appropriate treatment for such cases. Given the prolonged recovery phase, it is unclear if steroids altered the course of recovery. The median recovery for facial palsy is 66 days, with complete recovery in 90% of patients. ⁷ The correlate pathology of idiopathic facial palsies may suggest the possibility of spontaneous recovery, but this cannot be certain, and our case was severely debilitating. Facial neurorehabilitation offers a non-invasive option to aid recovery.

This case expands the recognized spectrum of non-ICANS neurotoxicity and underscores the potential for progressive, multifocal cranial neuropathies in affected patients. While these complications are not typically life-threatening, they can markedly affect quality of life and may require multidisciplinary management. Evaluating for neurocognitive and movement disorders in patients with cranial neuropathy may aid in characterizing the full spectrum of non-ICANS neurotoxicity and refining future diagnostic criteria. ⁸ The best treatment approach remains undefined; in this case, the patient experienced gradual improvement with corticosteroids and facial rehabilitation over extended follow-up.

ORCID iDs

Zachary T. Lazzari https://orcid.org/0000-0001-9401-1656

Spencer K. Hutto https://orcid.org/0000-0002-5644-9638