CAR T-cell therapy enters progressive multiple sclerosis with an early safety signal
Phase I data presented at MSToronto2026 place CAR T-cell therapy in a new corner of neuroimmunology: progressive multiple sclerosis, a disease in which inflammation and neurodegeneration often advance despite standard treatment. The candidate, azercabtagene zapreleucel (azer-cel), is an “off-the-shelf” allogeneic anti-CD19 product built to deplete peripheral B cells and then allow immune reconstitution. In a disorder where chronically activated B-cell populations can sustain meningeal and systemic immune activity, that mechanism is biologically plausible even in the progressive phase, when disability can deepen long after overt relapses have become less frequent.
The early readout came from TG-Azercel-101, an open-label phase I study involving 18 adults with refractory progressive multiple sclerosis. Each participant underwent lymphodepletion and then received a single infusion of azer-cel, with doses ranging from 30 to 500 × 10^6 CAR T cells. Follow-up was short, with a median duration of 102 days, so the dataset is best viewed as a first clinical look rather than a definitive test of efficacy. Still, the signal was encouraging enough to support the next step in development.
No serious dose-related adverse events emerged.
A short follow-up period did not reveal new toxicity after lymphodepletion and infusion
Safety is the immediate gatekeeper for any cellular therapy in a non-oncologic disease, especially one that requires lymphodepletion before infusion. In this cohort, the absence of serious dose-related adverse events across a dose range that extended to 500 × 10^6 CAR T cells is a meaningful early result. It suggests that the manufacturing approach, conditioning regimen, and cell product can be delivered without an obvious acute toxicity pattern in this small group of patients with refractory disease.
That matters because progressive multiple sclerosis is not a setting where investigators can afford to trade neurologic decline for uncontrolled immune toxicity. CAR T-cell platforms have transformed hematologic oncology, but their movement into autoimmune neurology raises different questions: whether B-cell depletion can be achieved without severe systemic inflammation, and whether immune reset can be obtained without destabilizing a vulnerable neurologic patient population. The first phase I pass through that filter was clean enough to keep the program moving.
The study does not provide the kind of granular adverse-event breakdown that would allow bedside comparisons with other immunotherapies. Even so, “no serious dose-related adverse events” is a narrow but important statement in an early cellular-therapy trial. In a disease defined by accumulated disability, a tolerable first exposure is often the prerequisite for asking the harder question of whether immune reconstitution can alter the course of progression.
Biomarker movement and disability stability hint at biological activity
Beyond safety, investigators reported that most participants remained stable or improved on cerebrospinal fluid biomarker analyses and clinical disability measures. The exact biomarkers are not detailed in the available text, but the direction of change is clinically relevant. In progressive multiple sclerosis, biomarker improvement is attractive because it may reflect a reduction in inflammatory trafficking within the central nervous system, a shift in intrathecal immune activity, or attenuation of downstream tissue injury that feeds disability accumulation.
The CD19 target gives the platform its rationale. By depleting peripheral B cells, azer-cel is designed to interrupt antigen presentation, cytokine signaling, and B-cell–driven immune amplification. In progressive disease, where compartmentalized immune activity may persist even when relapse activity slows, that kind of depletion could blunt the cellular support system that helps sustain chronic inflammation. If immune reconstitution follows without recurrent pathologic activation, the result could be a more durable reset than standard continuous immunosuppression provides.
The disability findings should still be interpreted cautiously. Open-label phase I studies are not built to separate drug effect from expectation, regression to the mean, or the natural variability of clinical exam scoring over short intervals. But the convergence of biomarker stability or improvement with maintained or improved disability measures is more persuasive than either signal alone. In an early study, that kind of concordance is the strongest reason to keep investing in the platform.
Why an allogeneic anti-CD19 platform matters in refractory disease
Azer-cel is not an individualized autologous product. It is described as an allogeneic, off-the-shelf therapy, which carries practical implications for speed, scalability, and access if later-phase data remain favorable. For patients with refractory progressive multiple sclerosis, the appeal is obvious: a treatment that can be delivered without the delay required to manufacture a bespoke product, and one that targets a biologic pathway already implicated in the disease.
The choice of CD19 also matters mechanistically. CD19-directed therapy reaches a broader B-cell compartment than some narrower approaches, potentially affecting cells that participate in antigen presentation and inflammatory persistence. In a condition where progressive decline often reflects more than one immune pathway, broader B-cell depletion may be part of why investigators are exploring cellular therapy at all. The challenge will be to show that the biologic effect is not only measurable in cerebrospinal fluid but also durable enough to change the slope of disability over time.
This is still a first-in-disease experience. The sample was small, the follow-up was limited, and the open-label design leaves plenty of room for uncertainty. None of those limitations diminish the importance of the early safety readout, but they do constrain how far the field can go in interpreting the disability and biomarker changes.
What the phase I result can and cannot claim
TG-Azercel-101 establishes feasibility, not proof of efficacy. It shows that azercabtagene zapreleucel can be delivered to adults with refractory progressive multiple sclerosis after lymphodepletion, with no serious dose-related adverse events seen over a median 102 days of follow-up. It also shows that most participants remained stable or improved on cerebrospinal fluid biomarker analyses and clinical disability measures. Those are early indicators, not final answers.
The next clinical question is straightforward: does the apparent biologic activity persist long enough to matter for a disease measured in months and years rather than days? A larger, controlled study will be needed to determine whether the early biomarker movement tracks with durable slowing of progression, whether the safety profile remains acceptable with longer observation, and whether the treatment effect differs across the dose range used here. Until then, the most defensible reading is narrow but important: a new cellular therapy has cleared an initial safety hurdle in a field that has spent decades searching for a way to reset the immune system without worsening neurologic harm.