Promising Effects of CAR T-Cell Therapy in Refractory Stiff Person Syndrome and a Hopeful Future for All Neuroautoimmunities

Abstract

Chimeric antigen receptor (CAR) T cells are genetically modified T cells expressing CARs, initially developed to recognize tumor antigens and kill cancer cells that evade T-cell recognition. Because of their impressive success in hemato-oncologic malignancies, CAR T cells are being repurposed with redesigned constructs for safety and sustained efficacy to target refractory systemic autoimmune or neurologic diseases. The CD19 CAR T cells—targeting those CD19-positive, antibody-secreting, long-lived plasma cells, and plasmablasts—are now extensively explored in refractory neuroautoimmunities with promising benefits based on case series in patients with myasthenia gravis (MG), stiff person syndrome (SPS), neuromyelitis, myositis, and multiple sclerosis; some patients with MG and SPS with steadily progressive and disabling disease refractory to all available therapies, including rituximab and new biologics, exhibit impressive clinical improvements with long-lasting benefits. The review, triggered by these early results and ongoing trials, addresses what these cells are and why they show effectiveness not only in antibody-mediated B-cell neurologic diseases unresponsive to available anti–B-cell agents but also in patients with nonpathogenic antibodies, implying effects even beyond B cells; points out that CARs are “living cells” penetrating physiologic barriers, such as the blood-brain barrier, expanding within tissues to memory cells ensuring sustained effects; describes the process and challenges of preparing and administering CAR T cells and their safety profile stressing the differences in toxicities when treating autoimmunites vs malignancies; and highlights that CD19 CAR T cells can successfully target even 2 different autoimmune diseases in the same patient, such as SPS and MG, offering promising prospects of changing the therapeutic algorithm in all neuroautoimmunities with potential for achieving even an immune reset shifting immunity to a healthy state.

Introduction

Stiff person syndrome (SPS) is an autoimmune neuronal excitability disorder caused by impaired inhibitory GABAergic neurotransmission associated with high-titer antibodies against glutamic acid decarboxylase, causing muscle stiffness, slow freezing gait, falls, and painful muscle spasms precipitated by external stimuli, collectively leading to disability and need for walking aids.¹ Patients respond to immunotherapy with IVIg based on a controlled trial,² but the benefit declines over time in 30% of the patients³; because only 35% respond to rituximab based on a negative controlled study,⁴ SPS being a progressive disease according to prospective longitudinal assessments,⁵ it can be disabling, necessitating the need for more effective therapies. A recently reported patient with both SPS and AChR-positive myasthenia gravis (MG) had relatively stable disease for 8 years but then progressed being unresponsive to IVIg, plasmapheresis, rituximab, and other immunotherapies⁶; although MG improved, the patient's SPS symptomatology steadily worsened, losing the ability to walk. After 3 months of CD19 chimeric antigen receptor (CAR) T-cell therapy, however, the patient was able to walk independently, exhibiting less stiffness, pains, and spasms with steadily enhancing benefits several months thereafter without the need for any immunotherapy not only for SPS but also for MG; the patient became AChR antibody-negative while their GAD-ab titers were reduced from 320 to 32 IU.⁶

Because this is the second case of refractory SPS responding to the same CD19 CAR T cells,⁷ the credit goes to Prof Ralf Gold's team who pioneered this exciting therapy. My meeting Prof Ralf Gold by chance on October 13, 2023, at the Milan airport lounge after an international Neurology meeting was the impetus that led to a breakthrough for the field; not only did he discuss with me the promising results of their first patient with SPS⁷ but, based on my enthusiastically expressed interest, he also called from the airport the CAR T-cell company in California to connect me with, facilitating since this very exact day the design of the first phase II CAR T-cell study that has been now completed in 26 enrolled patients with SPS.

Such success raises key questions about CAR T cells in all neurologic autoimmunities: What are these cells that reverse the course of an immune disease, such as SPS, from years long of progressive disability to a steadily increasing improvement in just a few months? Why might patients respond to CAR T-targeting B cells but not to other anti-B agents such as rituximab? How did CAR T cells help not 1 but concurrently 2 immunologically different autoimmune diseases, one with pathogenic AChR-ab and another with nonpathogenic GAD-ab? How did just one infusion exert long-term, steadily increasing benefits?

What Are CD19 CAR T Cells

It all started with the need to find ways for the T cells to recognize tumor antigens and kill cancer cells because tumor cells evade T-cell recognition.^8-12 To accomplish this, the patients' T cells were genetically modified to express CARs which contain 4 distinct domains (Figure 1): (1) an antigen-binding domain (A), consisting of the “single-chain variable fragment (scFv)” molecule that binds specifically with very high affinity extracellular cell surface antigens in an major histocompatibiliy complex (MHC)-independent process; for CD19-targeting CAR T cells as used in patients with SPS,^6,7 the scFv is derived from mouse anti-human CD19 monoclonal antibodies ensuring high affinity antigen-specific recognition; (2) a hinge (B) and transmembrane (C) domains that connect the extracellular antigen-binding CD19 to an intracellular signaling domain (D). The hinge domain consists of either CD8, CD28, or IgG1 amino acid sequences that ensure effective access and binding of CAR T cells to targeted antigen; the transmembrane domain, derived from CD28, CD4, or CD8, provides stability in anchoring the CAR into the T-cell receptor (TCR) through “CD3ζ” activation motifs (D). The CD3ζ is the TCR component that mediates T-cell activation and release of cytotoxic molecules, such as perforin or granzyme-B, that kill the targeted cells; and (3) the costimulatory CD28 or 4-1BB (CD137) domains (E), inserted to produce IL-2 leading to proliferation and differentiation of CARs into effector memory T cells after repeated antigen encounter, ensuring their long-term survival and sustained cytotoxicity, being referred to as “serial killers” because they persist for decades.¹³

Figure 1. CD19-Chimeric Antigen Receptor (CAR) T Cells.

Second-generation CAR constructs consist of an antigen-binding single-chain fragment variable (scFv) (A), a hinge region (B), a transmembrane CD8 transverse myelitis domain (C), an intracellular signaling CD3ζ domain (D), and a CD28 costimulatory domain (E). When a CAR binds its CD19 target antigen, the T cell is activated by the CAR and releases cytolytic molecules, such as perforin and granzyme B, which destroy the cells expressing the target antigen. Unlike conventional TCR-mediated T-cell activation, which relies on antigen recognition, activation signaling, and costimulation, the CARs combine all these 3 functions within a single molecule with recognition in an MHC-independent manner (modified version from Motte et al.²⁵). TCR = T-cell receptor.

The process of administering autologous CAR T cells begins with cell apheresis (A) (Figure 2), collecting the patient's lymphocytes, followed by T-cell activation (B) and genetic modification by a DNA lentivirus vector that delivers the genetic sequence necessary for insertion of CAR into the T cells (C). After a 6-week expansion in cultures, when these genetically modified cells have become CAR T cells (D), they are infused to hospitalized patients 2 days after lymphodepletion with fludarabine and cyclophosphamide (E); lymphodepletion is required to reduce the patient's lymphocyte population and limit competition for cytokines, allowing the infused CAR T cells to expand efficiently.^8-12

Figure 2. CAR T-Cell Preparation and Infusion.

The process starts with lymphocytopheresis (A), activation of the removed T cells to make them more responsive to genetic modification (B), introduction of CAR construct into the T cells by viral gene transfer (C), a 6-week CAR T-cell formation and expansion (D), and reinfusion of cells as CARs into the patient (E) (modified version from original provided by and used with permission of Prof Mougiakakos⁸). CAR = chimeric antigen receptor.

Apart from being autologous, CAR T cells can be also allogenic, derived from healthy donors after being genetically modified to express CARs, ready to use, and “off the shelf”, without the need for lymphocytapheresis. Although mechanistically allogenic, these CAR T cells may have the remote potential to trigger graft-versus-host reaction if the host immune cells attack the donor T cells;^8-12 this rare possibility is however further reduced by genome editing, modifying their TCR and MHC molecules ensuring efficacious and readily available therapy with good tolerance, persistence, easy administration, and lower cost compared with the autologous ones.

Rationale for Applying CAR T-Cell Therapy in Neuroautoimmune Diseases

Despite the success of current neuroimmunotherapies not only in SPS but also across the autoimmune neurology spectrum, approximately 20%–30% of patients either do not respond or, after an initial response, the benefit declines because of disease progression.⁸ Furthermore, the efficacy of all immunotherapies has a cycling pattern requiring weekly, biweekly, monthly, or 3–6-month administration cycles, with declining benefit toward the end of each maintenance cycle or with a steadily lesser efficacy which can lead over time to permanent disability.⁸ From the patient's perspective, these cyclic patterns translate in “wearing off” effects in-between doses, some adverse reactions after each injection, and the need for venous access or ports, that collectively prevent their feeling of normalcy in daily life. It is such patients with progressive, therapy-refractory neuroautoimmunities who are suitable candidates for CAR T-cell therapies.

The Uniqueness of CD19 CAR T-Cell Therapy

• 1. Profound effects within tissues. In patients currently treated with various anti-B cell monoclonal antibodies, there is a discrepancy between their direct effects on peripheral blood B-cell depletion and lack of tissue penetration.¹⁴ By contrast, CD19 CAR T cells migrate to fundamental disease-specific areas frequently shielded by physiological barriers, such as bone marrow and blood-brain barrier (BBB).^8,10 Lymph node biopsies from patients treated either with CD19 CAR T cells or rituximab show complete B-cell depletion only in the CAR T-cell–treated group, although circulating B cells are eliminated in all patients,¹⁵ supporting their advantage of potentially eliminating even thymic B cells relevant to MG. Their sustained presence within the CNS^10,16 can also mitigate chronic neuroinflammation and neurodegeneration by depleting infiltrating lymphoid follicles, such as the meningeal ectopic ones seen in multiple sclerosis (MS) and the resident B cells driving smoldering compartmentalized CNS inflammation, not currently accessible to approved anti–B-cell antibodies, such as rituximab, ocrelizumab, or ofatumumab.

• 2. Sustained benefits. Therapeutic antibodies have limited half-life; hence, repeated infusions are required. By contrast, CAR T cells are “living cells” that constantly expand, almost without exhaustion, to memory cells that persist for years,^12,13 providing long-term disease control and sustained benefits without the need for retreatment because of declining benefit, raising the question as to whether they can even eradicate autoimmunity.

• 3. Distinct B-cell targeting. Current anti-CD20 monoclonal antibodies, although can target CD20-expressing memory B cells,¹⁷ do not target plasmablasts and long-lived plasma cells, which are responsible for autoantibody production because most of these cells are CD19⁺ but CD20⁻; hence, rituximab has failed in controlled trials not only in SPS but also in AChR-MG. By contrast, CD19 CAR T cells eliminate the broadest range of B cells, including pathogenic plasmablasts and plasma cells making them applicable to targeting all antibodies including the IgG4 subclass. Importantly, the long-lived bone marrow plasma cells and plasmablasts that do not express CD19 (and instead express B cell maturation antigen, CD38, and CD138) are not affected, which explains why CD19-CAR T cells do not affect the reservoir of established humoral immunity protection and preserve the vaccine-induced IgG antibodies, as confirmed in a recently treated patient with SPS⁶ where postvaccinal antibody titers against tetanus, varicella zoster virus, rubella, and mumps remained unchanged.

• 4. Global effects, even beyond B cells. All current biologics and monoclonal antibodies target specific cells or cellular factors, such as T cells, B cells, immunoglobulins, cytokines, or complement, connected to antibody and complement-dependent cytotoxicity or antibody-dependent cellular phagocytosis; by contrast, CAR T cells function more autonomously by exerting global actions with robust therapeutic effects in all key immune factors.⁸ As shown by Zamvil's group,¹⁸ CD19 CAR T cells not only ameliorated experimental allergic encephalomyelitis but also thoroughly depleted B cells in peripheral tissues and CNS, demonstrating that their clinical benefit was independent of antigen-specific B cells and highlighting that CD19 CAR T cells are effective even in the absence of pathogenic antibodies. These data justify the reasoning why CD19-CAR T cells were concurrently effective in patients with GAD-ab–positive SPS and AChR-MG⁶ and have shown effectiveness in chronic inflammatory demyelinating polyneuropathy, myositis, MS, and lupus,¹⁹ where no disease-specific antibodies have been observed, implying that they also exert a role in modifying activated or regulatory T cells.

• 5. Promise for immune reset. The dynamics of CAR T-cell expansion and B-cell ablation lead to a naive non–class-switched B-cell system with the disappearance of circulating plasmablasts and sustained antibody reduction, collectively supporting the view that CD19 CAR T-cell therapy can induce an immune reset.²⁰ Although CAR T cells are directed to both pathogenic and nonpathogenic B cells, the rebooting of the B-cell system by depleting disease-associated B-cell clones that had triggered the original antibody production offer the possibility of even eradicating pathogenic B-cell autoimmunity. Preliminary evidence based on gene and proteomics expression suggests that CD19 CAR T cells have also the potential to modulate cytotoxic CD8+T cells shifting the expression profiles of immune cell subsets toward a healthy state.²¹

CAR T Toxicities

Their most common side effect is the cytokine release syndrome (CRS), a systemic inflammatory response that presents with fever, chills, tachycardia, cellular edema, hypotension, and hypoxia associated with elevated inflammatory biomarkers such as C-reactive protein, ferritin, and IL-6, observed within the first week after infusion. CRS occurs because activated CAR-T cells release cytokines, such as IFN-γ, TNF-α, and granulocyte macrophage-colony-stimulating factor which, by stimulating myeloid cells, trigger the release of large amounts of cytokines, such as IL-1β, IL-6, and TNF-α, that cause cytopenia and cardiovascular or respiratory symptoms. CRS is rather common, requiring inpatient monitoring for 1 week; it is, however, more serious in patients with malignancies because of lymphoid cancer environment but less severe in neuroautoimmunities. CRS responds to antipyretics, corticosteroids, and tocilizumab which is the first-line treatment; if CRS is persistent, anakinra—an IL-1-receptor antagonist—is considered. The enlargement of cervical lymph nodes noted in the recently treated patient with SPS⁶—a rare side effect we have also observed—reflects the CAR T-cells–engaging B cells in secondary lymphoid organs before their depletion.^8,19 Immune effector cell-associated neurotoxicity syndrome (ICANS) can occur in approximately 27% of malignancy-treated patients, with approximately 10% of them exhibiting severe neurotoxicity because of increased permeability of the BBB allowing proinflammatory cytokines to enter CNS, triggering cerebral edema requiring steroids and anakinra^8,19,20; however, ICANS has not been reported in the treatment of autoimmune diseases, where the overall toxicity profile of CAR T cells has been quite favorable compared with that of malignancies. Of interest to neuroimmunology, preliminary preclinical data have recently shown that CAR T cells in mice models with systemic cancers may impair cognition by inducing neuroinflammation,²² but whether these findings will be of clinical relevance to human malignancies treated with CAR T is currently unknown.

Prospects of Changing the Therapeutic Algorithm in Neuroautoimmunities

The success of homologous CD19 CAR T-cell therapy, although impressive based on case series, is confounded by accessibility, cost, time-consuming process in transforming the extracted cells to CAR T cells, need for hospitalization and coordinated medical teams with facilities, and expertise in cell therapy and hematology/oncology.⁸ These difficulties are now being steadily overcome with collaborative efforts of industry and academia, highly driven by the unique CAR T-cell effects in reversing refractory autoimmunities with sustained benefits and relative safety, as recently discussed.⁸

Although the experience of CD19 CAR T cells in neurology is limited to only a few case studies in MG, SPS, myositis, MS, neuromyelitis spectrum disorders, and encephalitis,^8,19,23 trials are expanding in large-scale studies generating enthusiasm that CAR T cells will have a major impact in treating neuroautoimmunitites. The 2 SPS cases^6,7 and some promising effects we are witnessing in the ongoing phase II SPS trial not only strengthen this expectation but also expand it to all neuroautoimmunites as a means of reversing progressive disability with inducing long-lasting benefits, offering strong wish and hope that CAR T-cell therapy will change our therapeutic algorithm by shifting the applied therapies in the early phases of disability development and as soon as patients reach refractory status rather than waiting for further disease progression that lowers the chances for substantial improvement; the noted lack of improvement in the stiffness of the left arm immobilized in a patient with SPS for 10 years⁶ is consistent with our view for earlier treatment initiation before the potentially reversible GABAergic dysfunction transitions to irreversible loss of GABAergic neurons and the cumulative deficits become untreatable.

The phenomenon of “1 stone targets 2 birds.” The observations that CD19 CAR T cells can help not 1 but 2 concurrent autoimmunities—each with different mechanism in the same patient, such as SPS and MG,⁶ rheumatoid arthritis and MG,²⁴ or MG and Lambert-Eaton myasthenic syndrome²⁵—offer another exciting dimension because overlapping autoimmunitites occur in 25%–30% of patients with SPS,¹ with similar percentages in other autoimmune neurologic diseases. Such a therapeutic benefit in 2 different and immunologically active diseases, each one contributing to the same patient's overall disability progression, is consistent with the global effect of CD19 CAR T cells even beyond B cells, adding another unique advantage of “1 stone targets 2 birds,” which further strengthens their expanding promising future.

Glossary

BBB

blood-brain barrier

CAR

chimeric antigen receptor

CRS

cytokine release syndrome

ICANS

immune effector cell-associated neurotoxicity syndrome

MG

myasthenia gravis

SPS

stiff-person syndrome

TCR

T-cell receptor

Author Contributions

M.C. Dalakas: drafting/revision of the manuscript for content, including medical writing for content; study concept or design.

Study Funding

The author reports no targeted funding.

Disclosure

The author reports no relevant disclosures. Go to Neurology.org/NN for fulls disclosures.